diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,70193 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8769177437454527,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.769177437454527e-05,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 3e-06,
+      "loss": 1.1658,
+      "step": 1
+    },
+    {
+      "epoch": 0.00017538354874909053,
+      "grad_norm": 0.04296875,
+      "learning_rate": 6e-06,
+      "loss": 1.1284,
+      "step": 2
+    },
+    {
+      "epoch": 0.0002630753231236358,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 9e-06,
+      "loss": 1.0655,
+      "step": 3
+    },
+    {
+      "epoch": 0.00035076709749818106,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.2e-05,
+      "loss": 1.2313,
+      "step": 4
+    },
+    {
+      "epoch": 0.00043845887187272633,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.5e-05,
+      "loss": 1.1557,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005261506462472716,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.8e-05,
+      "loss": 1.1887,
+      "step": 6
+    },
+    {
+      "epoch": 0.0006138424206218169,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 2.1000000000000002e-05,
+      "loss": 1.1527,
+      "step": 7
+    },
+    {
+      "epoch": 0.0007015341949963621,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.4e-05,
+      "loss": 1.1127,
+      "step": 8
+    },
+    {
+      "epoch": 0.0007892259693709074,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 2.7e-05,
+      "loss": 1.1494,
+      "step": 9
+    },
+    {
+      "epoch": 0.0008769177437454527,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 3e-05,
+      "loss": 1.1626,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009646095181199979,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 1.1429,
+      "step": 11
+    },
+    {
+      "epoch": 0.0010523012924945432,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 3.6e-05,
+      "loss": 1.1566,
+      "step": 12
+    },
+    {
+      "epoch": 0.0011399930668690886,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 3.9e-05,
+      "loss": 1.1611,
+      "step": 13
+    },
+    {
+      "epoch": 0.0012276848412436337,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.2000000000000004e-05,
+      "loss": 1.0719,
+      "step": 14
+    },
+    {
+      "epoch": 0.001315376615618179,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 1.1424,
+      "step": 15
+    },
+    {
+      "epoch": 0.0014030683899927242,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.8e-05,
+      "loss": 1.126,
+      "step": 16
+    },
+    {
+      "epoch": 0.0014907601643672696,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 1.1402,
+      "step": 17
+    },
+    {
+      "epoch": 0.0015784519387418148,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 5.4e-05,
+      "loss": 1.1307,
+      "step": 18
+    },
+    {
+      "epoch": 0.0016661437131163602,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 5.7e-05,
+      "loss": 1.1494,
+      "step": 19
+    },
+    {
+      "epoch": 0.0017538354874909053,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 6e-05,
+      "loss": 1.1678,
+      "step": 20
+    },
+    {
+      "epoch": 0.0018415272618654507,
+      "grad_norm": 0.0546875,
+      "learning_rate": 6.3e-05,
+      "loss": 1.1662,
+      "step": 21
+    },
+    {
+      "epoch": 0.0019292190362399958,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 1.1019,
+      "step": 22
+    },
+    {
+      "epoch": 0.002016910810614541,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 6.9e-05,
+      "loss": 1.1203,
+      "step": 23
+    },
+    {
+      "epoch": 0.0021046025849890864,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 7.2e-05,
+      "loss": 1.1157,
+      "step": 24
+    },
+    {
+      "epoch": 0.0021922943593636317,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.1259,
+      "step": 25
+    },
+    {
+      "epoch": 0.002279986133738177,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.8e-05,
+      "loss": 1.1498,
+      "step": 26
+    },
+    {
+      "epoch": 0.0023676779081127225,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 8.1e-05,
+      "loss": 1.1414,
+      "step": 27
+    },
+    {
+      "epoch": 0.0024553696824872674,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 8.400000000000001e-05,
+      "loss": 1.1709,
+      "step": 28
+    },
+    {
+      "epoch": 0.002543061456861813,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 8.7e-05,
+      "loss": 1.1324,
+      "step": 29
+    },
+    {
+      "epoch": 0.002630753231236358,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 1.1482,
+      "step": 30
+    },
+    {
+      "epoch": 0.0027184450056109036,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 9.3e-05,
+      "loss": 1.1014,
+      "step": 31
+    },
+    {
+      "epoch": 0.0028061367799854485,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 9.6e-05,
+      "loss": 1.0998,
+      "step": 32
+    },
+    {
+      "epoch": 0.002893828554359994,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.1555,
+      "step": 33
+    },
+    {
+      "epoch": 0.0029815203287345392,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 1.1934,
+      "step": 34
+    },
+    {
+      "epoch": 0.0030692121031090846,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00010500000000000002,
+      "loss": 1.1266,
+      "step": 35
+    },
+    {
+      "epoch": 0.0031569038774836296,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.000108,
+      "loss": 1.1031,
+      "step": 36
+    },
+    {
+      "epoch": 0.003244595651858175,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000111,
+      "loss": 1.1517,
+      "step": 37
+    },
+    {
+      "epoch": 0.0033322874262327203,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.000114,
+      "loss": 1.096,
+      "step": 38
+    },
+    {
+      "epoch": 0.0034199792006072657,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000117,
+      "loss": 1.1044,
+      "step": 39
+    },
+    {
+      "epoch": 0.0035076709749818106,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.00012,
+      "loss": 1.0901,
+      "step": 40
+    },
+    {
+      "epoch": 0.003595362749356356,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.000123,
+      "loss": 1.1287,
+      "step": 41
+    },
+    {
+      "epoch": 0.0036830545237309014,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000126,
+      "loss": 1.2277,
+      "step": 42
+    },
+    {
+      "epoch": 0.0037707462981054467,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000129,
+      "loss": 1.168,
+      "step": 43
+    },
+    {
+      "epoch": 0.0038584380724799917,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 1.1054,
+      "step": 44
+    },
+    {
+      "epoch": 0.0039461298468545375,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000135,
+      "loss": 1.1698,
+      "step": 45
+    },
+    {
+      "epoch": 0.004033821621229082,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.000138,
+      "loss": 1.1513,
+      "step": 46
+    },
+    {
+      "epoch": 0.004121513395603627,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000141,
+      "loss": 1.1666,
+      "step": 47
+    },
+    {
+      "epoch": 0.004209205169978173,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.000144,
+      "loss": 1.1025,
+      "step": 48
+    },
+    {
+      "epoch": 0.004296896944352718,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000147,
+      "loss": 1.1536,
+      "step": 49
+    },
+    {
+      "epoch": 0.0043845887187272635,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.1672,
+      "step": 50
+    },
+    {
+      "epoch": 0.004472280493101809,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.000153,
+      "loss": 1.0913,
+      "step": 51
+    },
+    {
+      "epoch": 0.004559972267476354,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.000156,
+      "loss": 1.1205,
+      "step": 52
+    },
+    {
+      "epoch": 0.0046476640418509,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000159,
+      "loss": 1.2006,
+      "step": 53
+    },
+    {
+      "epoch": 0.004735355816225445,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000162,
+      "loss": 1.1862,
+      "step": 54
+    },
+    {
+      "epoch": 0.0048230475905999895,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000165,
+      "loss": 1.2395,
+      "step": 55
+    },
+    {
+      "epoch": 0.004910739364974535,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00016800000000000002,
+      "loss": 1.138,
+      "step": 56
+    },
+    {
+      "epoch": 0.00499843113934908,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.000171,
+      "loss": 1.1346,
+      "step": 57
+    },
+    {
+      "epoch": 0.005086122913723626,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.000174,
+      "loss": 1.0885,
+      "step": 58
+    },
+    {
+      "epoch": 0.005173814688098171,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000177,
+      "loss": 1.1216,
+      "step": 59
+    },
+    {
+      "epoch": 0.005261506462472716,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 1.1571,
+      "step": 60
+    },
+    {
+      "epoch": 0.005349198236847262,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.000183,
+      "loss": 1.1255,
+      "step": 61
+    },
+    {
+      "epoch": 0.005436890011221807,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000186,
+      "loss": 1.1221,
+      "step": 62
+    },
+    {
+      "epoch": 0.005524581785596352,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.000189,
+      "loss": 1.1555,
+      "step": 63
+    },
+    {
+      "epoch": 0.005612273559970897,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000192,
+      "loss": 1.2047,
+      "step": 64
+    },
+    {
+      "epoch": 0.005699965334345442,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00019500000000000002,
+      "loss": 1.1844,
+      "step": 65
+    },
+    {
+      "epoch": 0.005787657108719988,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 1.123,
+      "step": 66
+    },
+    {
+      "epoch": 0.005875348883094533,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.000201,
+      "loss": 1.1913,
+      "step": 67
+    },
+    {
+      "epoch": 0.0059630406574690785,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00020400000000000003,
+      "loss": 1.1745,
+      "step": 68
+    },
+    {
+      "epoch": 0.006050732431843624,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00020700000000000002,
+      "loss": 1.094,
+      "step": 69
+    },
+    {
+      "epoch": 0.006138424206218169,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00021000000000000004,
+      "loss": 1.1598,
+      "step": 70
+    },
+    {
+      "epoch": 0.006226115980592715,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 1.1675,
+      "step": 71
+    },
+    {
+      "epoch": 0.006313807754967259,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000216,
+      "loss": 1.1369,
+      "step": 72
+    },
+    {
+      "epoch": 0.0064014995293418045,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 1.1755,
+      "step": 73
+    },
+    {
+      "epoch": 0.00648919130371635,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000222,
+      "loss": 1.2022,
+      "step": 74
+    },
+    {
+      "epoch": 0.006576883078090895,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.000225,
+      "loss": 1.1722,
+      "step": 75
+    },
+    {
+      "epoch": 0.006664574852465441,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.000228,
+      "loss": 1.0734,
+      "step": 76
+    },
+    {
+      "epoch": 0.006752266626839986,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000231,
+      "loss": 1.1007,
+      "step": 77
+    },
+    {
+      "epoch": 0.006839958401214531,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000234,
+      "loss": 1.0976,
+      "step": 78
+    },
+    {
+      "epoch": 0.006927650175589077,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00023700000000000001,
+      "loss": 1.1234,
+      "step": 79
+    },
+    {
+      "epoch": 0.007015341949963621,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00024,
+      "loss": 1.1001,
+      "step": 80
+    },
+    {
+      "epoch": 0.007103033724338167,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00024300000000000002,
+      "loss": 1.1492,
+      "step": 81
+    },
+    {
+      "epoch": 0.007190725498712712,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000246,
+      "loss": 1.1726,
+      "step": 82
+    },
+    {
+      "epoch": 0.007278417273087257,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00024900000000000004,
+      "loss": 1.1477,
+      "step": 83
+    },
+    {
+      "epoch": 0.007366109047461803,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.000252,
+      "loss": 1.1242,
+      "step": 84
+    },
+    {
+      "epoch": 0.007453800821836348,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000255,
+      "loss": 1.1142,
+      "step": 85
+    },
+    {
+      "epoch": 0.0075414925962108935,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000258,
+      "loss": 1.1349,
+      "step": 86
+    },
+    {
+      "epoch": 0.007629184370585439,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000261,
+      "loss": 1.1336,
+      "step": 87
+    },
+    {
+      "epoch": 0.007716876144959983,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 1.102,
+      "step": 88
+    },
+    {
+      "epoch": 0.007804567919334529,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.000267,
+      "loss": 1.1895,
+      "step": 89
+    },
+    {
+      "epoch": 0.007892259693709075,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00027,
+      "loss": 1.1714,
+      "step": 90
+    },
+    {
+      "epoch": 0.00797995146808362,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000273,
+      "loss": 1.1803,
+      "step": 91
+    },
+    {
+      "epoch": 0.008067643242458164,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000276,
+      "loss": 1.1686,
+      "step": 92
+    },
+    {
+      "epoch": 0.00815533501683271,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000279,
+      "loss": 1.1308,
+      "step": 93
+    },
+    {
+      "epoch": 0.008243026791207255,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000282,
+      "loss": 1.1795,
+      "step": 94
+    },
+    {
+      "epoch": 0.008330718565581801,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000285,
+      "loss": 1.1199,
+      "step": 95
+    },
+    {
+      "epoch": 0.008418410339956345,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000288,
+      "loss": 1.1489,
+      "step": 96
+    },
+    {
+      "epoch": 0.008506102114330892,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000291,
+      "loss": 1.086,
+      "step": 97
+    },
+    {
+      "epoch": 0.008593793888705436,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000294,
+      "loss": 1.1221,
+      "step": 98
+    },
+    {
+      "epoch": 0.008681485663079982,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000297,
+      "loss": 1.128,
+      "step": 99
+    },
+    {
+      "epoch": 0.008769177437454527,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00030000000000000003,
+      "loss": 1.177,
+      "step": 100
+    },
+    {
+      "epoch": 0.008856869211829071,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00030300000000000005,
+      "loss": 1.0933,
+      "step": 101
+    },
+    {
+      "epoch": 0.008944560986203618,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000306,
+      "loss": 1.1342,
+      "step": 102
+    },
+    {
+      "epoch": 0.009032252760578162,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000309,
+      "loss": 1.1317,
+      "step": 103
+    },
+    {
+      "epoch": 0.009119944534952708,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000312,
+      "loss": 1.1696,
+      "step": 104
+    },
+    {
+      "epoch": 0.009207636309327253,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000315,
+      "loss": 1.1273,
+      "step": 105
+    },
+    {
+      "epoch": 0.0092953280837018,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000318,
+      "loss": 1.1595,
+      "step": 106
+    },
+    {
+      "epoch": 0.009383019858076344,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.000321,
+      "loss": 1.1281,
+      "step": 107
+    },
+    {
+      "epoch": 0.00947071163245089,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000324,
+      "loss": 1.1297,
+      "step": 108
+    },
+    {
+      "epoch": 0.009558403406825434,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.000327,
+      "loss": 1.1683,
+      "step": 109
+    },
+    {
+      "epoch": 0.009646095181199979,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00033,
+      "loss": 1.0732,
+      "step": 110
+    },
+    {
+      "epoch": 0.009733786955574525,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000333,
+      "loss": 1.1155,
+      "step": 111
+    },
+    {
+      "epoch": 0.00982147872994907,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00033600000000000004,
+      "loss": 1.1712,
+      "step": 112
+    },
+    {
+      "epoch": 0.009909170504323616,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.000339,
+      "loss": 1.1747,
+      "step": 113
+    },
+    {
+      "epoch": 0.00999686227869816,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000342,
+      "loss": 1.1377,
+      "step": 114
+    },
+    {
+      "epoch": 0.010084554053072707,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00034500000000000004,
+      "loss": 1.1282,
+      "step": 115
+    },
+    {
+      "epoch": 0.010172245827447251,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.000348,
+      "loss": 1.1553,
+      "step": 116
+    },
+    {
+      "epoch": 0.010259937601821796,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000351,
+      "loss": 1.1211,
+      "step": 117
+    },
+    {
+      "epoch": 0.010347629376196342,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.000354,
+      "loss": 1.1568,
+      "step": 118
+    },
+    {
+      "epoch": 0.010435321150570886,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.000357,
+      "loss": 1.1445,
+      "step": 119
+    },
+    {
+      "epoch": 0.010523012924945433,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00035999999999999997,
+      "loss": 1.1153,
+      "step": 120
+    },
+    {
+      "epoch": 0.010610704699319977,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000363,
+      "loss": 1.151,
+      "step": 121
+    },
+    {
+      "epoch": 0.010698396473694523,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.000366,
+      "loss": 1.123,
+      "step": 122
+    },
+    {
+      "epoch": 0.010786088248069068,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000369,
+      "loss": 1.1733,
+      "step": 123
+    },
+    {
+      "epoch": 0.010873780022443614,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.000372,
+      "loss": 1.1694,
+      "step": 124
+    },
+    {
+      "epoch": 0.010961471796818159,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000375,
+      "loss": 1.1659,
+      "step": 125
+    },
+    {
+      "epoch": 0.011049163571192703,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.000378,
+      "loss": 1.1329,
+      "step": 126
+    },
+    {
+      "epoch": 0.01113685534556725,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.000381,
+      "loss": 1.1064,
+      "step": 127
+    },
+    {
+      "epoch": 0.011224547119941794,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000384,
+      "loss": 1.1358,
+      "step": 128
+    },
+    {
+      "epoch": 0.01131223889431634,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00038700000000000003,
+      "loss": 1.1024,
+      "step": 129
+    },
+    {
+      "epoch": 0.011399930668690885,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00039000000000000005,
+      "loss": 1.2044,
+      "step": 130
+    },
+    {
+      "epoch": 0.011487622443065431,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.000393,
+      "loss": 1.1628,
+      "step": 131
+    },
+    {
+      "epoch": 0.011575314217439975,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00039600000000000003,
+      "loss": 1.1833,
+      "step": 132
+    },
+    {
+      "epoch": 0.011663005991814522,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00039900000000000005,
+      "loss": 1.1599,
+      "step": 133
+    },
+    {
+      "epoch": 0.011750697766189066,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000402,
+      "loss": 1.122,
+      "step": 134
+    },
+    {
+      "epoch": 0.01183838954056361,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00040500000000000003,
+      "loss": 1.1688,
+      "step": 135
+    },
+    {
+      "epoch": 0.011926081314938157,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00040800000000000005,
+      "loss": 1.1775,
+      "step": 136
+    },
+    {
+      "epoch": 0.012013773089312701,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000411,
+      "loss": 1.1348,
+      "step": 137
+    },
+    {
+      "epoch": 0.012101464863687248,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00041400000000000003,
+      "loss": 1.2099,
+      "step": 138
+    },
+    {
+      "epoch": 0.012189156638061792,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.00041700000000000005,
+      "loss": 1.1488,
+      "step": 139
+    },
+    {
+      "epoch": 0.012276848412436338,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00042000000000000007,
+      "loss": 1.176,
+      "step": 140
+    },
+    {
+      "epoch": 0.012364540186810883,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000423,
+      "loss": 1.1426,
+      "step": 141
+    },
+    {
+      "epoch": 0.01245223196118543,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00042599999999999995,
+      "loss": 1.1649,
+      "step": 142
+    },
+    {
+      "epoch": 0.012539923735559974,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00042899999999999997,
+      "loss": 1.1576,
+      "step": 143
+    },
+    {
+      "epoch": 0.012627615509934518,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000432,
+      "loss": 1.1381,
+      "step": 144
+    },
+    {
+      "epoch": 0.012715307284309064,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000435,
+      "loss": 1.2093,
+      "step": 145
+    },
+    {
+      "epoch": 0.012802999058683609,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00043799999999999997,
+      "loss": 1.1952,
+      "step": 146
+    },
+    {
+      "epoch": 0.012890690833058155,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000441,
+      "loss": 1.1391,
+      "step": 147
+    },
+    {
+      "epoch": 0.0129783826074327,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000444,
+      "loss": 1.113,
+      "step": 148
+    },
+    {
+      "epoch": 0.013066074381807246,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00044699999999999997,
+      "loss": 1.1297,
+      "step": 149
+    },
+    {
+      "epoch": 0.01315376615618179,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00045,
+      "loss": 1.1373,
+      "step": 150
+    },
+    {
+      "epoch": 0.013241457930556335,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000453,
+      "loss": 1.1788,
+      "step": 151
+    },
+    {
+      "epoch": 0.013329149704930881,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.000456,
+      "loss": 1.1361,
+      "step": 152
+    },
+    {
+      "epoch": 0.013416841479305426,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000459,
+      "loss": 1.0928,
+      "step": 153
+    },
+    {
+      "epoch": 0.013504533253679972,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000462,
+      "loss": 1.1558,
+      "step": 154
+    },
+    {
+      "epoch": 0.013592225028054516,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000465,
+      "loss": 1.1215,
+      "step": 155
+    },
+    {
+      "epoch": 0.013679916802429063,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000468,
+      "loss": 1.1364,
+      "step": 156
+    },
+    {
+      "epoch": 0.013767608576803607,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.000471,
+      "loss": 1.2186,
+      "step": 157
+    },
+    {
+      "epoch": 0.013855300351178153,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00047400000000000003,
+      "loss": 1.132,
+      "step": 158
+    },
+    {
+      "epoch": 0.013942992125552698,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.000477,
+      "loss": 1.1483,
+      "step": 159
+    },
+    {
+      "epoch": 0.014030683899927242,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00048,
+      "loss": 1.1765,
+      "step": 160
+    },
+    {
+      "epoch": 0.014118375674301789,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00048300000000000003,
+      "loss": 1.2102,
+      "step": 161
+    },
+    {
+      "epoch": 0.014206067448676333,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00048600000000000005,
+      "loss": 1.1264,
+      "step": 162
+    },
+    {
+      "epoch": 0.01429375922305088,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004890000000000001,
+      "loss": 1.1673,
+      "step": 163
+    },
+    {
+      "epoch": 0.014381450997425424,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000492,
+      "loss": 1.1293,
+      "step": 164
+    },
+    {
+      "epoch": 0.01446914277179997,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.000495,
+      "loss": 1.214,
+      "step": 165
+    },
+    {
+      "epoch": 0.014556834546174515,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004980000000000001,
+      "loss": 1.1727,
+      "step": 166
+    },
+    {
+      "epoch": 0.014644526320549061,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000501,
+      "loss": 1.1803,
+      "step": 167
+    },
+    {
+      "epoch": 0.014732218094923605,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000504,
+      "loss": 1.1959,
+      "step": 168
+    },
+    {
+      "epoch": 0.01481990986929815,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005070000000000001,
+      "loss": 1.1719,
+      "step": 169
+    },
+    {
+      "epoch": 0.014907601643672696,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00051,
+      "loss": 1.178,
+      "step": 170
+    },
+    {
+      "epoch": 0.01499529341804724,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.000513,
+      "loss": 1.1316,
+      "step": 171
+    },
+    {
+      "epoch": 0.015082985192421787,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000516,
+      "loss": 1.1646,
+      "step": 172
+    },
+    {
+      "epoch": 0.015170676966796331,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005189999999999999,
+      "loss": 1.1202,
+      "step": 173
+    },
+    {
+      "epoch": 0.015258368741170878,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.000522,
+      "loss": 1.1387,
+      "step": 174
+    },
+    {
+      "epoch": 0.015346060515545422,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.000525,
+      "loss": 1.1942,
+      "step": 175
+    },
+    {
+      "epoch": 0.015433752289919967,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005279999999999999,
+      "loss": 1.1539,
+      "step": 176
+    },
+    {
+      "epoch": 0.015521444064294513,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.000531,
+      "loss": 1.1139,
+      "step": 177
+    },
+    {
+      "epoch": 0.015609135838669057,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.000534,
+      "loss": 1.172,
+      "step": 178
+    },
+    {
+      "epoch": 0.015696827613043602,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000537,
+      "loss": 1.1199,
+      "step": 179
+    },
+    {
+      "epoch": 0.01578451938741815,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.00054,
+      "loss": 1.1085,
+      "step": 180
+    },
+    {
+      "epoch": 0.015872211161792694,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.000543,
+      "loss": 1.1067,
+      "step": 181
+    },
+    {
+      "epoch": 0.01595990293616724,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.000546,
+      "loss": 1.2397,
+      "step": 182
+    },
+    {
+      "epoch": 0.016047594710541783,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000549,
+      "loss": 1.1286,
+      "step": 183
+    },
+    {
+      "epoch": 0.016135286484916328,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000552,
+      "loss": 1.1336,
+      "step": 184
+    },
+    {
+      "epoch": 0.016222978259290876,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000555,
+      "loss": 1.1732,
+      "step": 185
+    },
+    {
+      "epoch": 0.01631067003366542,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000558,
+      "loss": 1.1995,
+      "step": 186
+    },
+    {
+      "epoch": 0.016398361808039965,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000561,
+      "loss": 1.1547,
+      "step": 187
+    },
+    {
+      "epoch": 0.01648605358241451,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000564,
+      "loss": 1.1546,
+      "step": 188
+    },
+    {
+      "epoch": 0.016573745356789057,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000567,
+      "loss": 1.134,
+      "step": 189
+    },
+    {
+      "epoch": 0.016661437131163602,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00057,
+      "loss": 1.1755,
+      "step": 190
+    },
+    {
+      "epoch": 0.016749128905538146,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.000573,
+      "loss": 1.2015,
+      "step": 191
+    },
+    {
+      "epoch": 0.01683682067991269,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.000576,
+      "loss": 1.2078,
+      "step": 192
+    },
+    {
+      "epoch": 0.016924512454287235,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000579,
+      "loss": 1.2049,
+      "step": 193
+    },
+    {
+      "epoch": 0.017012204228661783,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.000582,
+      "loss": 1.1235,
+      "step": 194
+    },
+    {
+      "epoch": 0.017099896003036328,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.000585,
+      "loss": 1.193,
+      "step": 195
+    },
+    {
+      "epoch": 0.017187587777410872,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.000588,
+      "loss": 1.1205,
+      "step": 196
+    },
+    {
+      "epoch": 0.017275279551785417,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.000591,
+      "loss": 1.1362,
+      "step": 197
+    },
+    {
+      "epoch": 0.017362971326159965,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000594,
+      "loss": 1.1614,
+      "step": 198
+    },
+    {
+      "epoch": 0.01745066310053451,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005970000000000001,
+      "loss": 1.1948,
+      "step": 199
+    },
+    {
+      "epoch": 0.017538354874909054,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 1.1425,
+      "step": 200
+    },
+    {
+      "epoch": 0.0176260466492836,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.000603,
+      "loss": 1.1436,
+      "step": 201
+    },
+    {
+      "epoch": 0.017713738423658143,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0006060000000000001,
+      "loss": 1.1551,
+      "step": 202
+    },
+    {
+      "epoch": 0.01780143019803269,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0006090000000000001,
+      "loss": 1.1274,
+      "step": 203
+    },
+    {
+      "epoch": 0.017889121972407235,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.000612,
+      "loss": 1.1853,
+      "step": 204
+    },
+    {
+      "epoch": 0.01797681374678178,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000615,
+      "loss": 1.1765,
+      "step": 205
+    },
+    {
+      "epoch": 0.018064505521156324,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.000618,
+      "loss": 1.1268,
+      "step": 206
+    },
+    {
+      "epoch": 0.018152197295530872,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.000621,
+      "loss": 1.1769,
+      "step": 207
+    },
+    {
+      "epoch": 0.018239889069905417,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.000624,
+      "loss": 1.0777,
+      "step": 208
+    },
+    {
+      "epoch": 0.01832758084427996,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.000627,
+      "loss": 1.0759,
+      "step": 209
+    },
+    {
+      "epoch": 0.018415272618654506,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00063,
+      "loss": 1.1398,
+      "step": 210
+    },
+    {
+      "epoch": 0.01850296439302905,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.000633,
+      "loss": 1.1684,
+      "step": 211
+    },
+    {
+      "epoch": 0.0185906561674036,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.000636,
+      "loss": 1.1609,
+      "step": 212
+    },
+    {
+      "epoch": 0.018678347941778143,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000639,
+      "loss": 1.1409,
+      "step": 213
+    },
+    {
+      "epoch": 0.018766039716152687,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.000642,
+      "loss": 1.1721,
+      "step": 214
+    },
+    {
+      "epoch": 0.018853731490527232,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000645,
+      "loss": 1.1046,
+      "step": 215
+    },
+    {
+      "epoch": 0.01894142326490178,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000648,
+      "loss": 1.1549,
+      "step": 216
+    },
+    {
+      "epoch": 0.019029115039276324,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.000651,
+      "loss": 1.1698,
+      "step": 217
+    },
+    {
+      "epoch": 0.01911680681365087,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000654,
+      "loss": 1.1671,
+      "step": 218
+    },
+    {
+      "epoch": 0.019204498588025413,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000657,
+      "loss": 1.1809,
+      "step": 219
+    },
+    {
+      "epoch": 0.019292190362399958,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00066,
+      "loss": 1.1231,
+      "step": 220
+    },
+    {
+      "epoch": 0.019379882136774506,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006630000000000001,
+      "loss": 1.2035,
+      "step": 221
+    },
+    {
+      "epoch": 0.01946757391114905,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000666,
+      "loss": 1.1891,
+      "step": 222
+    },
+    {
+      "epoch": 0.019555265685523595,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000669,
+      "loss": 1.1181,
+      "step": 223
+    },
+    {
+      "epoch": 0.01964295745989814,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006720000000000001,
+      "loss": 1.1088,
+      "step": 224
+    },
+    {
+      "epoch": 0.019730649234272687,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000675,
+      "loss": 1.1524,
+      "step": 225
+    },
+    {
+      "epoch": 0.019818341008647232,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000678,
+      "loss": 1.1337,
+      "step": 226
+    },
+    {
+      "epoch": 0.019906032783021776,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0006810000000000001,
+      "loss": 1.1906,
+      "step": 227
+    },
+    {
+      "epoch": 0.01999372455739632,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000684,
+      "loss": 1.1515,
+      "step": 228
+    },
+    {
+      "epoch": 0.020081416331770865,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000687,
+      "loss": 1.1852,
+      "step": 229
+    },
+    {
+      "epoch": 0.020169108106145413,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0006900000000000001,
+      "loss": 1.161,
+      "step": 230
+    },
+    {
+      "epoch": 0.020256799880519958,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.000693,
+      "loss": 1.1192,
+      "step": 231
+    },
+    {
+      "epoch": 0.020344491654894502,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.000696,
+      "loss": 1.1421,
+      "step": 232
+    },
+    {
+      "epoch": 0.020432183429269047,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0006990000000000001,
+      "loss": 1.1496,
+      "step": 233
+    },
+    {
+      "epoch": 0.02051987520364359,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.000702,
+      "loss": 1.1524,
+      "step": 234
+    },
+    {
+      "epoch": 0.02060756697801814,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.000705,
+      "loss": 1.121,
+      "step": 235
+    },
+    {
+      "epoch": 0.020695258752392684,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000708,
+      "loss": 1.1113,
+      "step": 236
+    },
+    {
+      "epoch": 0.02078295052676723,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0007109999999999999,
+      "loss": 1.1425,
+      "step": 237
+    },
+    {
+      "epoch": 0.020870642301141773,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.000714,
+      "loss": 1.1002,
+      "step": 238
+    },
+    {
+      "epoch": 0.02095833407551632,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.000717,
+      "loss": 1.143,
+      "step": 239
+    },
+    {
+      "epoch": 0.021046025849890865,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0007199999999999999,
+      "loss": 1.1428,
+      "step": 240
+    },
+    {
+      "epoch": 0.02113371762426541,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.000723,
+      "loss": 1.1528,
+      "step": 241
+    },
+    {
+      "epoch": 0.021221409398639954,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000726,
+      "loss": 1.1262,
+      "step": 242
+    },
+    {
+      "epoch": 0.0213091011730145,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000729,
+      "loss": 1.0917,
+      "step": 243
+    },
+    {
+      "epoch": 0.021396792947389047,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000732,
+      "loss": 1.1335,
+      "step": 244
+    },
+    {
+      "epoch": 0.02148448472176359,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.000735,
+      "loss": 1.1724,
+      "step": 245
+    },
+    {
+      "epoch": 0.021572176496138136,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.000738,
+      "loss": 1.1043,
+      "step": 246
+    },
+    {
+      "epoch": 0.02165986827051268,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000741,
+      "loss": 1.1998,
+      "step": 247
+    },
+    {
+      "epoch": 0.02174756004488723,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000744,
+      "loss": 1.1098,
+      "step": 248
+    },
+    {
+      "epoch": 0.021835251819261773,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.000747,
+      "loss": 1.1391,
+      "step": 249
+    },
+    {
+      "epoch": 0.021922943593636317,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00075,
+      "loss": 1.1297,
+      "step": 250
+    },
+    {
+      "epoch": 0.022010635368010862,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000753,
+      "loss": 1.1278,
+      "step": 251
+    },
+    {
+      "epoch": 0.022098327142385406,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000756,
+      "loss": 1.1577,
+      "step": 252
+    },
+    {
+      "epoch": 0.022186018916759954,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.000759,
+      "loss": 1.162,
+      "step": 253
+    },
+    {
+      "epoch": 0.0222737106911345,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.000762,
+      "loss": 1.1489,
+      "step": 254
+    },
+    {
+      "epoch": 0.022361402465509043,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0007650000000000001,
+      "loss": 1.1197,
+      "step": 255
+    },
+    {
+      "epoch": 0.022449094239883588,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000768,
+      "loss": 1.1886,
+      "step": 256
+    },
+    {
+      "epoch": 0.022536786014258136,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000771,
+      "loss": 1.1247,
+      "step": 257
+    },
+    {
+      "epoch": 0.02262447778863268,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0007740000000000001,
+      "loss": 1.0887,
+      "step": 258
+    },
+    {
+      "epoch": 0.022712169563007225,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000777,
+      "loss": 1.1116,
+      "step": 259
+    },
+    {
+      "epoch": 0.02279986133738177,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0007800000000000001,
+      "loss": 1.1466,
+      "step": 260
+    },
+    {
+      "epoch": 0.022887553111756314,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0007830000000000001,
+      "loss": 1.1808,
+      "step": 261
+    },
+    {
+      "epoch": 0.022975244886130862,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000786,
+      "loss": 1.1295,
+      "step": 262
+    },
+    {
+      "epoch": 0.023062936660505406,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0007890000000000001,
+      "loss": 1.1737,
+      "step": 263
+    },
+    {
+      "epoch": 0.02315062843487995,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0007920000000000001,
+      "loss": 1.0943,
+      "step": 264
+    },
+    {
+      "epoch": 0.023238320209254495,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000795,
+      "loss": 1.1239,
+      "step": 265
+    },
+    {
+      "epoch": 0.023326011983629043,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0007980000000000001,
+      "loss": 1.1174,
+      "step": 266
+    },
+    {
+      "epoch": 0.023413703758003588,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008010000000000001,
+      "loss": 1.1444,
+      "step": 267
+    },
+    {
+      "epoch": 0.023501395532378132,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.000804,
+      "loss": 1.1608,
+      "step": 268
+    },
+    {
+      "epoch": 0.023589087306752677,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008070000000000001,
+      "loss": 1.1755,
+      "step": 269
+    },
+    {
+      "epoch": 0.02367677908112722,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008100000000000001,
+      "loss": 1.1465,
+      "step": 270
+    },
+    {
+      "epoch": 0.02376447085550177,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.000813,
+      "loss": 1.1984,
+      "step": 271
+    },
+    {
+      "epoch": 0.023852162629876314,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008160000000000001,
+      "loss": 1.1619,
+      "step": 272
+    },
+    {
+      "epoch": 0.02393985440425086,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008190000000000001,
+      "loss": 1.1327,
+      "step": 273
+    },
+    {
+      "epoch": 0.024027546178625403,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000822,
+      "loss": 1.1771,
+      "step": 274
+    },
+    {
+      "epoch": 0.02411523795299995,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008250000000000001,
+      "loss": 1.1673,
+      "step": 275
+    },
+    {
+      "epoch": 0.024202929727374495,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008280000000000001,
+      "loss": 1.138,
+      "step": 276
+    },
+    {
+      "epoch": 0.02429062150174904,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008310000000000001,
+      "loss": 1.1402,
+      "step": 277
+    },
+    {
+      "epoch": 0.024378313276123584,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008340000000000001,
+      "loss": 1.1657,
+      "step": 278
+    },
+    {
+      "epoch": 0.02446600505049813,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008370000000000001,
+      "loss": 1.1595,
+      "step": 279
+    },
+    {
+      "epoch": 0.024553696824872677,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008400000000000001,
+      "loss": 1.1554,
+      "step": 280
+    },
+    {
+      "epoch": 0.02464138859924722,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008430000000000001,
+      "loss": 1.1981,
+      "step": 281
+    },
+    {
+      "epoch": 0.024729080373621766,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.000846,
+      "loss": 1.1148,
+      "step": 282
+    },
+    {
+      "epoch": 0.02481677214799631,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0008489999999999999,
+      "loss": 1.1277,
+      "step": 283
+    },
+    {
+      "epoch": 0.02490446392237086,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008519999999999999,
+      "loss": 1.1403,
+      "step": 284
+    },
+    {
+      "epoch": 0.024992155696745403,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.000855,
+      "loss": 1.0865,
+      "step": 285
+    },
+    {
+      "epoch": 0.025079847471119947,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008579999999999999,
+      "loss": 1.1986,
+      "step": 286
+    },
+    {
+      "epoch": 0.025167539245494492,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.000861,
+      "loss": 1.1272,
+      "step": 287
+    },
+    {
+      "epoch": 0.025255231019869036,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.000864,
+      "loss": 1.1783,
+      "step": 288
+    },
+    {
+      "epoch": 0.025342922794243584,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008669999999999999,
+      "loss": 1.11,
+      "step": 289
+    },
+    {
+      "epoch": 0.02543061456861813,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00087,
+      "loss": 1.1297,
+      "step": 290
+    },
+    {
+      "epoch": 0.025518306342992673,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000873,
+      "loss": 1.132,
+      "step": 291
+    },
+    {
+      "epoch": 0.025605998117367218,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008759999999999999,
+      "loss": 1.1429,
+      "step": 292
+    },
+    {
+      "epoch": 0.025693689891741762,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.000879,
+      "loss": 1.1246,
+      "step": 293
+    },
+    {
+      "epoch": 0.02578138166611631,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.000882,
+      "loss": 1.1745,
+      "step": 294
+    },
+    {
+      "epoch": 0.025869073440490855,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008849999999999999,
+      "loss": 1.1477,
+      "step": 295
+    },
+    {
+      "epoch": 0.0259567652148654,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000888,
+      "loss": 1.2113,
+      "step": 296
+    },
+    {
+      "epoch": 0.026044456989239944,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000891,
+      "loss": 1.1925,
+      "step": 297
+    },
+    {
+      "epoch": 0.026132148763614492,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008939999999999999,
+      "loss": 1.1713,
+      "step": 298
+    },
+    {
+      "epoch": 0.026219840537989036,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.000897,
+      "loss": 1.1088,
+      "step": 299
+    },
+    {
+      "epoch": 0.02630753231236358,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0009,
+      "loss": 1.1081,
+      "step": 300
+    },
+    {
+      "epoch": 0.026395224086738125,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009029999999999999,
+      "loss": 1.1574,
+      "step": 301
+    },
+    {
+      "epoch": 0.02648291586111267,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000906,
+      "loss": 1.182,
+      "step": 302
+    },
+    {
+      "epoch": 0.026570607635487218,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.000909,
+      "loss": 1.142,
+      "step": 303
+    },
+    {
+      "epoch": 0.026658299409861762,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.000912,
+      "loss": 1.1583,
+      "step": 304
+    },
+    {
+      "epoch": 0.026745991184236307,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000915,
+      "loss": 1.1874,
+      "step": 305
+    },
+    {
+      "epoch": 0.02683368295861085,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.000918,
+      "loss": 1.131,
+      "step": 306
+    },
+    {
+      "epoch": 0.0269213747329854,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.000921,
+      "loss": 1.1403,
+      "step": 307
+    },
+    {
+      "epoch": 0.027009066507359944,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.000924,
+      "loss": 1.1459,
+      "step": 308
+    },
+    {
+      "epoch": 0.02709675828173449,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000927,
+      "loss": 1.12,
+      "step": 309
+    },
+    {
+      "epoch": 0.027184450056109033,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00093,
+      "loss": 1.1903,
+      "step": 310
+    },
+    {
+      "epoch": 0.027272141830483577,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000933,
+      "loss": 1.1237,
+      "step": 311
+    },
+    {
+      "epoch": 0.027359833604858125,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.000936,
+      "loss": 1.1044,
+      "step": 312
+    },
+    {
+      "epoch": 0.02744752537923267,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0009390000000000001,
+      "loss": 1.1263,
+      "step": 313
+    },
+    {
+      "epoch": 0.027535217153607214,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000942,
+      "loss": 1.1294,
+      "step": 314
+    },
+    {
+      "epoch": 0.02762290892798176,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.000945,
+      "loss": 1.1812,
+      "step": 315
+    },
+    {
+      "epoch": 0.027710600702356307,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0009480000000000001,
+      "loss": 1.1673,
+      "step": 316
+    },
+    {
+      "epoch": 0.02779829247673085,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.000951,
+      "loss": 1.1607,
+      "step": 317
+    },
+    {
+      "epoch": 0.027885984251105396,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.000954,
+      "loss": 1.1432,
+      "step": 318
+    },
+    {
+      "epoch": 0.02797367602547994,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0009570000000000001,
+      "loss": 1.15,
+      "step": 319
+    },
+    {
+      "epoch": 0.028061367799854485,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00096,
+      "loss": 1.1851,
+      "step": 320
+    },
+    {
+      "epoch": 0.028149059574229033,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000963,
+      "loss": 1.1291,
+      "step": 321
+    },
+    {
+      "epoch": 0.028236751348603577,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0009660000000000001,
+      "loss": 1.1273,
+      "step": 322
+    },
+    {
+      "epoch": 0.028324443122978122,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.000969,
+      "loss": 1.1779,
+      "step": 323
+    },
+    {
+      "epoch": 0.028412134897352666,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0009720000000000001,
+      "loss": 1.1774,
+      "step": 324
+    },
+    {
+      "epoch": 0.028499826671727214,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0009750000000000001,
+      "loss": 1.1247,
+      "step": 325
+    },
+    {
+      "epoch": 0.02858751844610176,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0009780000000000001,
+      "loss": 1.1964,
+      "step": 326
+    },
+    {
+      "epoch": 0.028675210220476303,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.000981,
+      "loss": 1.0994,
+      "step": 327
+    },
+    {
+      "epoch": 0.028762901994850848,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.000984,
+      "loss": 1.1888,
+      "step": 328
+    },
+    {
+      "epoch": 0.028850593769225392,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.000987,
+      "loss": 1.1212,
+      "step": 329
+    },
+    {
+      "epoch": 0.02893828554359994,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00099,
+      "loss": 1.1633,
+      "step": 330
+    },
+    {
+      "epoch": 0.029025977317974485,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0009930000000000002,
+      "loss": 1.1294,
+      "step": 331
+    },
+    {
+      "epoch": 0.02911366909234903,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0009960000000000001,
+      "loss": 1.191,
+      "step": 332
+    },
+    {
+      "epoch": 0.029201360866723574,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.000999,
+      "loss": 1.1672,
+      "step": 333
+    },
+    {
+      "epoch": 0.029289052641098122,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001002,
+      "loss": 1.1475,
+      "step": 334
+    },
+    {
+      "epoch": 0.029376744415472666,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001005,
+      "loss": 1.1125,
+      "step": 335
+    },
+    {
+      "epoch": 0.02946443618984721,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001008,
+      "loss": 1.1321,
+      "step": 336
+    },
+    {
+      "epoch": 0.029552127964221755,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0010110000000000002,
+      "loss": 1.1477,
+      "step": 337
+    },
+    {
+      "epoch": 0.0296398197385963,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0010140000000000001,
+      "loss": 1.0962,
+      "step": 338
+    },
+    {
+      "epoch": 0.029727511512970848,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0010170000000000001,
+      "loss": 1.2346,
+      "step": 339
+    },
+    {
+      "epoch": 0.029815203287345392,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.00102,
+      "loss": 1.15,
+      "step": 340
+    },
+    {
+      "epoch": 0.029902895061719937,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001023,
+      "loss": 1.1906,
+      "step": 341
+    },
+    {
+      "epoch": 0.02999058683609448,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001026,
+      "loss": 1.1226,
+      "step": 342
+    },
+    {
+      "epoch": 0.030078278610469026,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0010290000000000002,
+      "loss": 1.1437,
+      "step": 343
+    },
+    {
+      "epoch": 0.030165970384843574,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001032,
+      "loss": 1.1805,
+      "step": 344
+    },
+    {
+      "epoch": 0.03025366215921812,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001035,
+      "loss": 1.1995,
+      "step": 345
+    },
+    {
+      "epoch": 0.030341353933592663,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0010379999999999999,
+      "loss": 1.1826,
+      "step": 346
+    },
+    {
+      "epoch": 0.030429045707967207,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001041,
+      "loss": 1.2097,
+      "step": 347
+    },
+    {
+      "epoch": 0.030516737482341755,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.001044,
+      "loss": 1.1366,
+      "step": 348
+    },
+    {
+      "epoch": 0.0306044292567163,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.001047,
+      "loss": 1.0497,
+      "step": 349
+    },
+    {
+      "epoch": 0.030692121031090844,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00105,
+      "loss": 1.1373,
+      "step": 350
+    },
+    {
+      "epoch": 0.03077981280546539,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001053,
+      "loss": 1.154,
+      "step": 351
+    },
+    {
+      "epoch": 0.030867504579839933,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0010559999999999999,
+      "loss": 1.0712,
+      "step": 352
+    },
+    {
+      "epoch": 0.03095519635421448,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001059,
+      "loss": 1.1138,
+      "step": 353
+    },
+    {
+      "epoch": 0.031042888128589026,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001062,
+      "loss": 1.1591,
+      "step": 354
+    },
+    {
+      "epoch": 0.03113057990296357,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001065,
+      "loss": 1.1796,
+      "step": 355
+    },
+    {
+      "epoch": 0.031218271677338115,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001068,
+      "loss": 1.1627,
+      "step": 356
+    },
+    {
+      "epoch": 0.03130596345171266,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001071,
+      "loss": 1.1789,
+      "step": 357
+    },
+    {
+      "epoch": 0.031393655226087204,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.001074,
+      "loss": 1.1437,
+      "step": 358
+    },
+    {
+      "epoch": 0.031481347000461755,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001077,
+      "loss": 1.135,
+      "step": 359
+    },
+    {
+      "epoch": 0.0315690387748363,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00108,
+      "loss": 1.1739,
+      "step": 360
+    },
+    {
+      "epoch": 0.031656730549210844,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001083,
+      "loss": 1.1799,
+      "step": 361
+    },
+    {
+      "epoch": 0.03174442232358539,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001086,
+      "loss": 1.1486,
+      "step": 362
+    },
+    {
+      "epoch": 0.03183211409795993,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.001089,
+      "loss": 1.1124,
+      "step": 363
+    },
+    {
+      "epoch": 0.03191980587233448,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001092,
+      "loss": 1.1203,
+      "step": 364
+    },
+    {
+      "epoch": 0.03200749764670902,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001095,
+      "loss": 1.1634,
+      "step": 365
+    },
+    {
+      "epoch": 0.03209518942108357,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001098,
+      "loss": 1.213,
+      "step": 366
+    },
+    {
+      "epoch": 0.03218288119545811,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001101,
+      "loss": 1.1538,
+      "step": 367
+    },
+    {
+      "epoch": 0.032270572969832656,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001104,
+      "loss": 1.1712,
+      "step": 368
+    },
+    {
+      "epoch": 0.03235826474420721,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001107,
+      "loss": 1.203,
+      "step": 369
+    },
+    {
+      "epoch": 0.03244595651858175,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00111,
+      "loss": 1.1138,
+      "step": 370
+    },
+    {
+      "epoch": 0.032533648292956296,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.001113,
+      "loss": 1.2147,
+      "step": 371
+    },
+    {
+      "epoch": 0.03262134006733084,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001116,
+      "loss": 1.1656,
+      "step": 372
+    },
+    {
+      "epoch": 0.032709031841705385,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.001119,
+      "loss": 1.1468,
+      "step": 373
+    },
+    {
+      "epoch": 0.03279672361607993,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001122,
+      "loss": 1.0946,
+      "step": 374
+    },
+    {
+      "epoch": 0.032884415390454474,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0011250000000000001,
+      "loss": 1.138,
+      "step": 375
+    },
+    {
+      "epoch": 0.03297210716482902,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001128,
+      "loss": 1.0943,
+      "step": 376
+    },
+    {
+      "epoch": 0.033059798939203563,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001131,
+      "loss": 1.175,
+      "step": 377
+    },
+    {
+      "epoch": 0.033147490713578115,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001134,
+      "loss": 1.1694,
+      "step": 378
+    },
+    {
+      "epoch": 0.03323518248795266,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001137,
+      "loss": 1.1661,
+      "step": 379
+    },
+    {
+      "epoch": 0.033322874262327204,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00114,
+      "loss": 1.156,
+      "step": 380
+    },
+    {
+      "epoch": 0.03341056603670175,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0011430000000000001,
+      "loss": 1.1612,
+      "step": 381
+    },
+    {
+      "epoch": 0.03349825781107629,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.001146,
+      "loss": 1.214,
+      "step": 382
+    },
+    {
+      "epoch": 0.03358594958545084,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001149,
+      "loss": 1.1871,
+      "step": 383
+    },
+    {
+      "epoch": 0.03367364135982538,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001152,
+      "loss": 1.1854,
+      "step": 384
+    },
+    {
+      "epoch": 0.033761333134199926,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.001155,
+      "loss": 1.2159,
+      "step": 385
+    },
+    {
+      "epoch": 0.03384902490857447,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001158,
+      "loss": 1.1848,
+      "step": 386
+    },
+    {
+      "epoch": 0.03393671668294902,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0011610000000000001,
+      "loss": 1.1314,
+      "step": 387
+    },
+    {
+      "epoch": 0.03402440845732357,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001164,
+      "loss": 1.0823,
+      "step": 388
+    },
+    {
+      "epoch": 0.03411210023169811,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.001167,
+      "loss": 1.192,
+      "step": 389
+    },
+    {
+      "epoch": 0.034199792006072656,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00117,
+      "loss": 1.1671,
+      "step": 390
+    },
+    {
+      "epoch": 0.0342874837804472,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001173,
+      "loss": 1.199,
+      "step": 391
+    },
+    {
+      "epoch": 0.034375175554821745,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.001176,
+      "loss": 1.1531,
+      "step": 392
+    },
+    {
+      "epoch": 0.03446286732919629,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0011790000000000001,
+      "loss": 1.1779,
+      "step": 393
+    },
+    {
+      "epoch": 0.034550559103570834,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001182,
+      "loss": 1.2149,
+      "step": 394
+    },
+    {
+      "epoch": 0.03463825087794538,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.001185,
+      "loss": 1.1088,
+      "step": 395
+    },
+    {
+      "epoch": 0.03472594265231993,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.001188,
+      "loss": 1.2079,
+      "step": 396
+    },
+    {
+      "epoch": 0.034813634426694474,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001191,
+      "loss": 1.1626,
+      "step": 397
+    },
+    {
+      "epoch": 0.03490132620106902,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0011940000000000002,
+      "loss": 1.1697,
+      "step": 398
+    },
+    {
+      "epoch": 0.03498901797544356,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0011970000000000001,
+      "loss": 1.0707,
+      "step": 399
+    },
+    {
+      "epoch": 0.03507670974981811,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 1.1594,
+      "step": 400
+    },
+    {
+      "epoch": 0.03516440152419265,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.001203,
+      "loss": 1.1704,
+      "step": 401
+    },
+    {
+      "epoch": 0.0352520932985672,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001206,
+      "loss": 1.2377,
+      "step": 402
+    },
+    {
+      "epoch": 0.03533978507294174,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001209,
+      "loss": 1.1689,
+      "step": 403
+    },
+    {
+      "epoch": 0.035427476847316286,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0012120000000000002,
+      "loss": 1.1144,
+      "step": 404
+    },
+    {
+      "epoch": 0.03551516862169084,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0012150000000000002,
+      "loss": 1.1204,
+      "step": 405
+    },
+    {
+      "epoch": 0.03560286039606538,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0012180000000000001,
+      "loss": 1.0814,
+      "step": 406
+    },
+    {
+      "epoch": 0.035690552170439926,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0012209999999999999,
+      "loss": 1.1226,
+      "step": 407
+    },
+    {
+      "epoch": 0.03577824394481447,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001224,
+      "loss": 1.1736,
+      "step": 408
+    },
+    {
+      "epoch": 0.035865935719189015,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001227,
+      "loss": 1.1062,
+      "step": 409
+    },
+    {
+      "epoch": 0.03595362749356356,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00123,
+      "loss": 1.1278,
+      "step": 410
+    },
+    {
+      "epoch": 0.036041319267938104,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001233,
+      "loss": 1.1471,
+      "step": 411
+    },
+    {
+      "epoch": 0.03612901104231265,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001236,
+      "loss": 1.1663,
+      "step": 412
+    },
+    {
+      "epoch": 0.036216702816687193,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012389999999999999,
+      "loss": 1.1849,
+      "step": 413
+    },
+    {
+      "epoch": 0.036304394591061745,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001242,
+      "loss": 1.1569,
+      "step": 414
+    },
+    {
+      "epoch": 0.03639208636543629,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001245,
+      "loss": 1.1768,
+      "step": 415
+    },
+    {
+      "epoch": 0.036479778139810834,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.001248,
+      "loss": 1.1954,
+      "step": 416
+    },
+    {
+      "epoch": 0.03656746991418538,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001251,
+      "loss": 1.1599,
+      "step": 417
+    },
+    {
+      "epoch": 0.03665516168855992,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.001254,
+      "loss": 1.1768,
+      "step": 418
+    },
+    {
+      "epoch": 0.03674285346293447,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0012569999999999999,
+      "loss": 1.1885,
+      "step": 419
+    },
+    {
+      "epoch": 0.03683054523730901,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00126,
+      "loss": 1.1622,
+      "step": 420
+    },
+    {
+      "epoch": 0.036918237011683556,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001263,
+      "loss": 1.1106,
+      "step": 421
+    },
+    {
+      "epoch": 0.0370059287860581,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.001266,
+      "loss": 1.1625,
+      "step": 422
+    },
+    {
+      "epoch": 0.03709362056043265,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001269,
+      "loss": 1.196,
+      "step": 423
+    },
+    {
+      "epoch": 0.0371813123348072,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.001272,
+      "loss": 1.1093,
+      "step": 424
+    },
+    {
+      "epoch": 0.03726900410918174,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.001275,
+      "loss": 1.1551,
+      "step": 425
+    },
+    {
+      "epoch": 0.037356695883556286,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001278,
+      "loss": 1.1507,
+      "step": 426
+    },
+    {
+      "epoch": 0.03744438765793083,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.001281,
+      "loss": 1.1926,
+      "step": 427
+    },
+    {
+      "epoch": 0.037532079432305375,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.001284,
+      "loss": 1.1639,
+      "step": 428
+    },
+    {
+      "epoch": 0.03761977120667992,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001287,
+      "loss": 1.0905,
+      "step": 429
+    },
+    {
+      "epoch": 0.037707462981054464,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00129,
+      "loss": 1.1512,
+      "step": 430
+    },
+    {
+      "epoch": 0.03779515475542901,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001293,
+      "loss": 1.2661,
+      "step": 431
+    },
+    {
+      "epoch": 0.03788284652980356,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001296,
+      "loss": 1.1454,
+      "step": 432
+    },
+    {
+      "epoch": 0.037970538304178104,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001299,
+      "loss": 1.1774,
+      "step": 433
+    },
+    {
+      "epoch": 0.03805823007855265,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.001302,
+      "loss": 1.1833,
+      "step": 434
+    },
+    {
+      "epoch": 0.03814592185292719,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001305,
+      "loss": 1.1726,
+      "step": 435
+    },
+    {
+      "epoch": 0.03823361362730174,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001308,
+      "loss": 1.195,
+      "step": 436
+    },
+    {
+      "epoch": 0.03832130540167628,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001311,
+      "loss": 1.1641,
+      "step": 437
+    },
+    {
+      "epoch": 0.03840899717605083,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001314,
+      "loss": 1.1287,
+      "step": 438
+    },
+    {
+      "epoch": 0.03849668895042537,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001317,
+      "loss": 1.1811,
+      "step": 439
+    },
+    {
+      "epoch": 0.038584380724799916,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00132,
+      "loss": 1.176,
+      "step": 440
+    },
+    {
+      "epoch": 0.03867207249917447,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001323,
+      "loss": 1.1574,
+      "step": 441
+    },
+    {
+      "epoch": 0.03875976427354901,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0013260000000000001,
+      "loss": 1.1575,
+      "step": 442
+    },
+    {
+      "epoch": 0.038847456047923556,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001329,
+      "loss": 1.2033,
+      "step": 443
+    },
+    {
+      "epoch": 0.0389351478222981,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.001332,
+      "loss": 1.0864,
+      "step": 444
+    },
+    {
+      "epoch": 0.039022839596672645,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001335,
+      "loss": 1.1684,
+      "step": 445
+    },
+    {
+      "epoch": 0.03911053137104719,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.001338,
+      "loss": 1.1843,
+      "step": 446
+    },
+    {
+      "epoch": 0.039198223145421734,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.001341,
+      "loss": 1.1668,
+      "step": 447
+    },
+    {
+      "epoch": 0.03928591491979628,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0013440000000000001,
+      "loss": 1.1332,
+      "step": 448
+    },
+    {
+      "epoch": 0.039373606694170823,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.001347,
+      "loss": 1.1679,
+      "step": 449
+    },
+    {
+      "epoch": 0.039461298468545375,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00135,
+      "loss": 1.114,
+      "step": 450
+    },
+    {
+      "epoch": 0.03954899024291992,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001353,
+      "loss": 1.1945,
+      "step": 451
+    },
+    {
+      "epoch": 0.039636682017294464,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001356,
+      "loss": 1.1376,
+      "step": 452
+    },
+    {
+      "epoch": 0.03972437379166901,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001359,
+      "loss": 1.2279,
+      "step": 453
+    },
+    {
+      "epoch": 0.03981206556604355,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0013620000000000001,
+      "loss": 1.1463,
+      "step": 454
+    },
+    {
+      "epoch": 0.0398997573404181,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0013650000000000001,
+      "loss": 1.1059,
+      "step": 455
+    },
+    {
+      "epoch": 0.03998744911479264,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001368,
+      "loss": 1.202,
+      "step": 456
+    },
+    {
+      "epoch": 0.040075140889167186,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001371,
+      "loss": 1.1904,
+      "step": 457
+    },
+    {
+      "epoch": 0.04016283266354173,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.001374,
+      "loss": 1.1002,
+      "step": 458
+    },
+    {
+      "epoch": 0.04025052443791628,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0013770000000000002,
+      "loss": 1.175,
+      "step": 459
+    },
+    {
+      "epoch": 0.04033821621229083,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0013800000000000002,
+      "loss": 1.1055,
+      "step": 460
+    },
+    {
+      "epoch": 0.04042590798666537,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0013830000000000001,
+      "loss": 1.1636,
+      "step": 461
+    },
+    {
+      "epoch": 0.040513599761039916,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001386,
+      "loss": 1.2022,
+      "step": 462
+    },
+    {
+      "epoch": 0.04060129153541446,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.001389,
+      "loss": 1.1555,
+      "step": 463
+    },
+    {
+      "epoch": 0.040688983309789005,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.001392,
+      "loss": 1.127,
+      "step": 464
+    },
+    {
+      "epoch": 0.04077667508416355,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0013950000000000002,
+      "loss": 1.1509,
+      "step": 465
+    },
+    {
+      "epoch": 0.040864366858538094,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0013980000000000002,
+      "loss": 1.1856,
+      "step": 466
+    },
+    {
+      "epoch": 0.04095205863291264,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0014010000000000001,
+      "loss": 1.112,
+      "step": 467
+    },
+    {
+      "epoch": 0.04103975040728718,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.001404,
+      "loss": 1.1798,
+      "step": 468
+    },
+    {
+      "epoch": 0.041127442181661734,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.001407,
+      "loss": 1.157,
+      "step": 469
+    },
+    {
+      "epoch": 0.04121513395603628,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00141,
+      "loss": 1.1253,
+      "step": 470
+    },
+    {
+      "epoch": 0.04130282573041082,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001413,
+      "loss": 1.2278,
+      "step": 471
+    },
+    {
+      "epoch": 0.04139051750478537,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.001416,
+      "loss": 1.1901,
+      "step": 472
+    },
+    {
+      "epoch": 0.04147820927915991,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001419,
+      "loss": 1.1247,
+      "step": 473
+    },
+    {
+      "epoch": 0.04156590105353446,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0014219999999999999,
+      "loss": 1.1976,
+      "step": 474
+    },
+    {
+      "epoch": 0.041653592827909,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001425,
+      "loss": 1.1147,
+      "step": 475
+    },
+    {
+      "epoch": 0.041741284602283546,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001428,
+      "loss": 1.1605,
+      "step": 476
+    },
+    {
+      "epoch": 0.04182897637665809,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001431,
+      "loss": 1.1467,
+      "step": 477
+    },
+    {
+      "epoch": 0.04191666815103264,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001434,
+      "loss": 1.1731,
+      "step": 478
+    },
+    {
+      "epoch": 0.042004359925407186,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001437,
+      "loss": 1.1652,
+      "step": 479
+    },
+    {
+      "epoch": 0.04209205169978173,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0014399999999999999,
+      "loss": 1.126,
+      "step": 480
+    },
+    {
+      "epoch": 0.042179743474156275,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001443,
+      "loss": 1.1246,
+      "step": 481
+    },
+    {
+      "epoch": 0.04226743524853082,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001446,
+      "loss": 1.1049,
+      "step": 482
+    },
+    {
+      "epoch": 0.042355127022905364,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001449,
+      "loss": 1.1786,
+      "step": 483
+    },
+    {
+      "epoch": 0.04244281879727991,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001452,
+      "loss": 1.1095,
+      "step": 484
+    },
+    {
+      "epoch": 0.042530510571654453,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001455,
+      "loss": 1.1608,
+      "step": 485
+    },
+    {
+      "epoch": 0.042618202346029,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001458,
+      "loss": 1.2512,
+      "step": 486
+    },
+    {
+      "epoch": 0.04270589412040355,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.001461,
+      "loss": 1.1778,
+      "step": 487
+    },
+    {
+      "epoch": 0.042793585894778094,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001464,
+      "loss": 1.1793,
+      "step": 488
+    },
+    {
+      "epoch": 0.04288127766915264,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001467,
+      "loss": 1.1419,
+      "step": 489
+    },
+    {
+      "epoch": 0.04296896944352718,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00147,
+      "loss": 1.1623,
+      "step": 490
+    },
+    {
+      "epoch": 0.04305666121790173,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001473,
+      "loss": 1.162,
+      "step": 491
+    },
+    {
+      "epoch": 0.04314435299227627,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.001476,
+      "loss": 1.1999,
+      "step": 492
+    },
+    {
+      "epoch": 0.043232044766650816,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001479,
+      "loss": 1.1679,
+      "step": 493
+    },
+    {
+      "epoch": 0.04331973654102536,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001482,
+      "loss": 1.1739,
+      "step": 494
+    },
+    {
+      "epoch": 0.043407428315399905,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001485,
+      "loss": 1.1403,
+      "step": 495
+    },
+    {
+      "epoch": 0.04349512008977446,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001488,
+      "loss": 1.1612,
+      "step": 496
+    },
+    {
+      "epoch": 0.043582811864149,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001491,
+      "loss": 1.1721,
+      "step": 497
+    },
+    {
+      "epoch": 0.043670503638523546,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001494,
+      "loss": 1.1623,
+      "step": 498
+    },
+    {
+      "epoch": 0.04375819541289809,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001497,
+      "loss": 1.1506,
+      "step": 499
+    },
+    {
+      "epoch": 0.043845887187272635,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0015,
+      "loss": 1.1622,
+      "step": 500
+    },
+    {
+      "epoch": 0.043845887187272635,
+      "eval_loss": 1.1753246784210205,
+      "eval_runtime": 427.0911,
+      "eval_samples_per_second": 33.827,
+      "eval_steps_per_second": 8.457,
+      "step": 500
+    },
+    {
+      "epoch": 0.04393357896164718,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001503,
+      "loss": 1.2038,
+      "step": 501
+    },
+    {
+      "epoch": 0.044021270736021724,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.001506,
+      "loss": 1.1894,
+      "step": 502
+    },
+    {
+      "epoch": 0.04410896251039627,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0015090000000000001,
+      "loss": 1.1718,
+      "step": 503
+    },
+    {
+      "epoch": 0.04419665428477081,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.001512,
+      "loss": 1.1739,
+      "step": 504
+    },
+    {
+      "epoch": 0.044284346059145364,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.001515,
+      "loss": 1.2066,
+      "step": 505
+    },
+    {
+      "epoch": 0.04437203783351991,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.001518,
+      "loss": 1.2004,
+      "step": 506
+    },
+    {
+      "epoch": 0.04445972960789445,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001521,
+      "loss": 1.1137,
+      "step": 507
+    },
+    {
+      "epoch": 0.044547421382269,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001524,
+      "loss": 1.1641,
+      "step": 508
+    },
+    {
+      "epoch": 0.04463511315664354,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0015270000000000001,
+      "loss": 1.1574,
+      "step": 509
+    },
+    {
+      "epoch": 0.04472280493101809,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0015300000000000001,
+      "loss": 1.1937,
+      "step": 510
+    },
+    {
+      "epoch": 0.04481049670539263,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.001533,
+      "loss": 1.1683,
+      "step": 511
+    },
+    {
+      "epoch": 0.044898188479767176,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.001536,
+      "loss": 1.1322,
+      "step": 512
+    },
+    {
+      "epoch": 0.04498588025414172,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001539,
+      "loss": 1.2337,
+      "step": 513
+    },
+    {
+      "epoch": 0.04507357202851627,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001542,
+      "loss": 1.166,
+      "step": 514
+    },
+    {
+      "epoch": 0.045161263802890816,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0015450000000000001,
+      "loss": 1.172,
+      "step": 515
+    },
+    {
+      "epoch": 0.04524895557726536,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0015480000000000001,
+      "loss": 1.1241,
+      "step": 516
+    },
+    {
+      "epoch": 0.045336647351639905,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001551,
+      "loss": 1.1513,
+      "step": 517
+    },
+    {
+      "epoch": 0.04542433912601445,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001554,
+      "loss": 1.1605,
+      "step": 518
+    },
+    {
+      "epoch": 0.045512030900388994,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001557,
+      "loss": 1.1583,
+      "step": 519
+    },
+    {
+      "epoch": 0.04559972267476354,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0015600000000000002,
+      "loss": 1.2352,
+      "step": 520
+    },
+    {
+      "epoch": 0.045687414449138083,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0015630000000000002,
+      "loss": 1.1565,
+      "step": 521
+    },
+    {
+      "epoch": 0.04577510622351263,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0015660000000000001,
+      "loss": 1.1127,
+      "step": 522
+    },
+    {
+      "epoch": 0.04586279799788718,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.001569,
+      "loss": 1.1731,
+      "step": 523
+    },
+    {
+      "epoch": 0.045950489772261724,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.001572,
+      "loss": 1.1597,
+      "step": 524
+    },
+    {
+      "epoch": 0.04603818154663627,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.001575,
+      "loss": 1.1712,
+      "step": 525
+    },
+    {
+      "epoch": 0.04612587332101081,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0015780000000000002,
+      "loss": 1.1518,
+      "step": 526
+    },
+    {
+      "epoch": 0.04621356509538536,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0015810000000000002,
+      "loss": 1.187,
+      "step": 527
+    },
+    {
+      "epoch": 0.0463012568697599,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0015840000000000001,
+      "loss": 1.2328,
+      "step": 528
+    },
+    {
+      "epoch": 0.046388948644134446,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.001587,
+      "loss": 1.1717,
+      "step": 529
+    },
+    {
+      "epoch": 0.04647664041850899,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00159,
+      "loss": 1.195,
+      "step": 530
+    },
+    {
+      "epoch": 0.046564332192883535,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001593,
+      "loss": 1.1765,
+      "step": 531
+    },
+    {
+      "epoch": 0.04665202396725809,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0015960000000000002,
+      "loss": 1.1699,
+      "step": 532
+    },
+    {
+      "epoch": 0.04673971574163263,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0015990000000000002,
+      "loss": 1.2119,
+      "step": 533
+    },
+    {
+      "epoch": 0.046827407516007176,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0016020000000000001,
+      "loss": 1.1685,
+      "step": 534
+    },
+    {
+      "epoch": 0.04691509929038172,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001605,
+      "loss": 1.1741,
+      "step": 535
+    },
+    {
+      "epoch": 0.047002791064756265,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.001608,
+      "loss": 1.1958,
+      "step": 536
+    },
+    {
+      "epoch": 0.04709048283913081,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0016110000000000002,
+      "loss": 1.1116,
+      "step": 537
+    },
+    {
+      "epoch": 0.047178174613505354,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0016140000000000002,
+      "loss": 1.1486,
+      "step": 538
+    },
+    {
+      "epoch": 0.0472658663878799,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0016170000000000002,
+      "loss": 1.2106,
+      "step": 539
+    },
+    {
+      "epoch": 0.04735355816225444,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0016200000000000001,
+      "loss": 1.084,
+      "step": 540
+    },
+    {
+      "epoch": 0.047441249936628994,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.001623,
+      "loss": 1.1961,
+      "step": 541
+    },
+    {
+      "epoch": 0.04752894171100354,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001626,
+      "loss": 1.1434,
+      "step": 542
+    },
+    {
+      "epoch": 0.04761663348537808,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0016290000000000002,
+      "loss": 1.104,
+      "step": 543
+    },
+    {
+      "epoch": 0.04770432525975263,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0016320000000000002,
+      "loss": 1.1166,
+      "step": 544
+    },
+    {
+      "epoch": 0.04779201703412717,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0016350000000000002,
+      "loss": 1.124,
+      "step": 545
+    },
+    {
+      "epoch": 0.04787970880850172,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0016380000000000001,
+      "loss": 1.1609,
+      "step": 546
+    },
+    {
+      "epoch": 0.04796740058287626,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.001641,
+      "loss": 1.2102,
+      "step": 547
+    },
+    {
+      "epoch": 0.048055092357250806,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001644,
+      "loss": 1.1698,
+      "step": 548
+    },
+    {
+      "epoch": 0.04814278413162535,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0016470000000000002,
+      "loss": 1.1735,
+      "step": 549
+    },
+    {
+      "epoch": 0.0482304759059999,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0016500000000000002,
+      "loss": 1.2511,
+      "step": 550
+    },
+    {
+      "epoch": 0.048318167680374446,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0016530000000000002,
+      "loss": 1.2098,
+      "step": 551
+    },
+    {
+      "epoch": 0.04840585945474899,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0016560000000000001,
+      "loss": 1.1615,
+      "step": 552
+    },
+    {
+      "epoch": 0.048493551229123535,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001659,
+      "loss": 1.1997,
+      "step": 553
+    },
+    {
+      "epoch": 0.04858124300349808,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0016620000000000003,
+      "loss": 1.1446,
+      "step": 554
+    },
+    {
+      "epoch": 0.048668934777872624,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0016650000000000002,
+      "loss": 1.2135,
+      "step": 555
+    },
+    {
+      "epoch": 0.04875662655224717,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0016680000000000002,
+      "loss": 1.156,
+      "step": 556
+    },
+    {
+      "epoch": 0.04884431832662171,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0016710000000000002,
+      "loss": 1.208,
+      "step": 557
+    },
+    {
+      "epoch": 0.04893201010099626,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0016740000000000001,
+      "loss": 1.1824,
+      "step": 558
+    },
+    {
+      "epoch": 0.04901970187537081,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001677,
+      "loss": 1.124,
+      "step": 559
+    },
+    {
+      "epoch": 0.049107393649745354,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0016800000000000003,
+      "loss": 1.2047,
+      "step": 560
+    },
+    {
+      "epoch": 0.0491950854241199,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016830000000000003,
+      "loss": 1.174,
+      "step": 561
+    },
+    {
+      "epoch": 0.04928277719849444,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0016860000000000002,
+      "loss": 1.1866,
+      "step": 562
+    },
+    {
+      "epoch": 0.04937046897286899,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.001689,
+      "loss": 1.1495,
+      "step": 563
+    },
+    {
+      "epoch": 0.04945816074724353,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001692,
+      "loss": 1.1851,
+      "step": 564
+    },
+    {
+      "epoch": 0.049545852521618076,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.001695,
+      "loss": 1.1824,
+      "step": 565
+    },
+    {
+      "epoch": 0.04963354429599262,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0016979999999999999,
+      "loss": 1.1339,
+      "step": 566
+    },
+    {
+      "epoch": 0.049721236070367165,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0017009999999999998,
+      "loss": 1.1914,
+      "step": 567
+    },
+    {
+      "epoch": 0.04980892784474172,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017039999999999998,
+      "loss": 1.1541,
+      "step": 568
+    },
+    {
+      "epoch": 0.04989661961911626,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001707,
+      "loss": 1.1928,
+      "step": 569
+    },
+    {
+      "epoch": 0.049984311393490806,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00171,
+      "loss": 1.1804,
+      "step": 570
+    },
+    {
+      "epoch": 0.05007200316786535,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001713,
+      "loss": 1.2177,
+      "step": 571
+    },
+    {
+      "epoch": 0.050159694942239895,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0017159999999999999,
+      "loss": 1.1609,
+      "step": 572
+    },
+    {
+      "epoch": 0.05024738671661444,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0017189999999999998,
+      "loss": 1.1638,
+      "step": 573
+    },
+    {
+      "epoch": 0.050335078490988984,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.001722,
+      "loss": 1.1652,
+      "step": 574
+    },
+    {
+      "epoch": 0.05042277026536353,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001725,
+      "loss": 1.139,
+      "step": 575
+    },
+    {
+      "epoch": 0.05051046203973807,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001728,
+      "loss": 1.1755,
+      "step": 576
+    },
+    {
+      "epoch": 0.05059815381411262,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001731,
+      "loss": 1.1723,
+      "step": 577
+    },
+    {
+      "epoch": 0.05068584558848717,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017339999999999999,
+      "loss": 1.2931,
+      "step": 578
+    },
+    {
+      "epoch": 0.05077353736286171,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0017369999999999998,
+      "loss": 1.2084,
+      "step": 579
+    },
+    {
+      "epoch": 0.05086122913723626,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00174,
+      "loss": 1.1393,
+      "step": 580
+    },
+    {
+      "epoch": 0.0509489209116108,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.001743,
+      "loss": 1.1763,
+      "step": 581
+    },
+    {
+      "epoch": 0.05103661268598535,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001746,
+      "loss": 1.1301,
+      "step": 582
+    },
+    {
+      "epoch": 0.05112430446035989,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.001749,
+      "loss": 1.2042,
+      "step": 583
+    },
+    {
+      "epoch": 0.051211996234734436,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0017519999999999999,
+      "loss": 1.1959,
+      "step": 584
+    },
+    {
+      "epoch": 0.05129968800910898,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0017549999999999998,
+      "loss": 1.1316,
+      "step": 585
+    },
+    {
+      "epoch": 0.051387379783483525,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001758,
+      "loss": 1.2323,
+      "step": 586
+    },
+    {
+      "epoch": 0.051475071557858076,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001761,
+      "loss": 1.1328,
+      "step": 587
+    },
+    {
+      "epoch": 0.05156276333223262,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001764,
+      "loss": 1.1951,
+      "step": 588
+    },
+    {
+      "epoch": 0.051650455106607165,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001767,
+      "loss": 1.1176,
+      "step": 589
+    },
+    {
+      "epoch": 0.05173814688098171,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0017699999999999999,
+      "loss": 1.1789,
+      "step": 590
+    },
+    {
+      "epoch": 0.051825838655356254,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001773,
+      "loss": 1.161,
+      "step": 591
+    },
+    {
+      "epoch": 0.0519135304297308,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001776,
+      "loss": 1.194,
+      "step": 592
+    },
+    {
+      "epoch": 0.05200122220410534,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001779,
+      "loss": 1.2048,
+      "step": 593
+    },
+    {
+      "epoch": 0.05208891397847989,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001782,
+      "loss": 1.2059,
+      "step": 594
+    },
+    {
+      "epoch": 0.05217660575285443,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001785,
+      "loss": 1.1874,
+      "step": 595
+    },
+    {
+      "epoch": 0.052264297527228984,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0017879999999999999,
+      "loss": 1.221,
+      "step": 596
+    },
+    {
+      "epoch": 0.05235198930160353,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001791,
+      "loss": 1.165,
+      "step": 597
+    },
+    {
+      "epoch": 0.05243968107597807,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.001794,
+      "loss": 1.2116,
+      "step": 598
+    },
+    {
+      "epoch": 0.05252737285035262,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001797,
+      "loss": 1.1841,
+      "step": 599
+    },
+    {
+      "epoch": 0.05261506462472716,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0018,
+      "loss": 1.1804,
+      "step": 600
+    },
+    {
+      "epoch": 0.052702756399101706,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001803,
+      "loss": 1.1746,
+      "step": 601
+    },
+    {
+      "epoch": 0.05279044817347625,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0018059999999999999,
+      "loss": 1.2058,
+      "step": 602
+    },
+    {
+      "epoch": 0.052878139947850795,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001809,
+      "loss": 1.1783,
+      "step": 603
+    },
+    {
+      "epoch": 0.05296583172222534,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.001812,
+      "loss": 1.1468,
+      "step": 604
+    },
+    {
+      "epoch": 0.05305352349659989,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.001815,
+      "loss": 1.1409,
+      "step": 605
+    },
+    {
+      "epoch": 0.053141215270974436,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.001818,
+      "loss": 1.1699,
+      "step": 606
+    },
+    {
+      "epoch": 0.05322890704534898,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001821,
+      "loss": 1.1966,
+      "step": 607
+    },
+    {
+      "epoch": 0.053316598819723525,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001824,
+      "loss": 1.1734,
+      "step": 608
+    },
+    {
+      "epoch": 0.05340429059409807,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001827,
+      "loss": 1.183,
+      "step": 609
+    },
+    {
+      "epoch": 0.053491982368472614,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00183,
+      "loss": 1.1807,
+      "step": 610
+    },
+    {
+      "epoch": 0.05357967414284716,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001833,
+      "loss": 1.164,
+      "step": 611
+    },
+    {
+      "epoch": 0.0536673659172217,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001836,
+      "loss": 1.1253,
+      "step": 612
+    },
+    {
+      "epoch": 0.05375505769159625,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.001839,
+      "loss": 1.1593,
+      "step": 613
+    },
+    {
+      "epoch": 0.0538427494659708,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001842,
+      "loss": 1.2045,
+      "step": 614
+    },
+    {
+      "epoch": 0.05393044124034534,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.001845,
+      "loss": 1.1429,
+      "step": 615
+    },
+    {
+      "epoch": 0.05401813301471989,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001848,
+      "loss": 1.1433,
+      "step": 616
+    },
+    {
+      "epoch": 0.05410582478909443,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001851,
+      "loss": 1.1286,
+      "step": 617
+    },
+    {
+      "epoch": 0.05419351656346898,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001854,
+      "loss": 1.1921,
+      "step": 618
+    },
+    {
+      "epoch": 0.05428120833784352,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001857,
+      "loss": 1.1742,
+      "step": 619
+    },
+    {
+      "epoch": 0.054368900112218066,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00186,
+      "loss": 1.2216,
+      "step": 620
+    },
+    {
+      "epoch": 0.05445659188659261,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.001863,
+      "loss": 1.1694,
+      "step": 621
+    },
+    {
+      "epoch": 0.054544283660967155,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001866,
+      "loss": 1.1512,
+      "step": 622
+    },
+    {
+      "epoch": 0.054631975435341706,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001869,
+      "loss": 1.1755,
+      "step": 623
+    },
+    {
+      "epoch": 0.05471966720971625,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001872,
+      "loss": 1.1805,
+      "step": 624
+    },
+    {
+      "epoch": 0.054807358984090795,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001875,
+      "loss": 1.2201,
+      "step": 625
+    },
+    {
+      "epoch": 0.05489505075846534,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0018780000000000001,
+      "loss": 1.255,
+      "step": 626
+    },
+    {
+      "epoch": 0.054982742532839884,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.001881,
+      "loss": 1.1907,
+      "step": 627
+    },
+    {
+      "epoch": 0.05507043430721443,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001884,
+      "loss": 1.1669,
+      "step": 628
+    },
+    {
+      "epoch": 0.05515812608158897,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.001887,
+      "loss": 1.1855,
+      "step": 629
+    },
+    {
+      "epoch": 0.05524581785596352,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00189,
+      "loss": 1.1983,
+      "step": 630
+    },
+    {
+      "epoch": 0.05533350963033806,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0018930000000000002,
+      "loss": 1.2005,
+      "step": 631
+    },
+    {
+      "epoch": 0.055421201404712614,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0018960000000000001,
+      "loss": 1.1762,
+      "step": 632
+    },
+    {
+      "epoch": 0.05550889317908716,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.001899,
+      "loss": 1.2092,
+      "step": 633
+    },
+    {
+      "epoch": 0.0555965849534617,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.001902,
+      "loss": 1.1696,
+      "step": 634
+    },
+    {
+      "epoch": 0.05568427672783625,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.001905,
+      "loss": 1.218,
+      "step": 635
+    },
+    {
+      "epoch": 0.05577196850221079,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.001908,
+      "loss": 1.2202,
+      "step": 636
+    },
+    {
+      "epoch": 0.055859660276585336,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0019110000000000002,
+      "loss": 1.1392,
+      "step": 637
+    },
+    {
+      "epoch": 0.05594735205095988,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0019140000000000001,
+      "loss": 1.1393,
+      "step": 638
+    },
+    {
+      "epoch": 0.056035043825334425,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001917,
+      "loss": 1.1697,
+      "step": 639
+    },
+    {
+      "epoch": 0.05612273559970897,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00192,
+      "loss": 1.1154,
+      "step": 640
+    },
+    {
+      "epoch": 0.05621042737408352,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.001923,
+      "loss": 1.1602,
+      "step": 641
+    },
+    {
+      "epoch": 0.056298119148458066,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001926,
+      "loss": 1.163,
+      "step": 642
+    },
+    {
+      "epoch": 0.05638581092283261,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0019290000000000002,
+      "loss": 1.1692,
+      "step": 643
+    },
+    {
+      "epoch": 0.056473502697207155,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019320000000000001,
+      "loss": 1.1855,
+      "step": 644
+    },
+    {
+      "epoch": 0.0565611944715817,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001935,
+      "loss": 1.1763,
+      "step": 645
+    },
+    {
+      "epoch": 0.056648886245956244,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.001938,
+      "loss": 1.1464,
+      "step": 646
+    },
+    {
+      "epoch": 0.05673657802033079,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001941,
+      "loss": 1.2453,
+      "step": 647
+    },
+    {
+      "epoch": 0.05682426979470533,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0019440000000000002,
+      "loss": 1.1012,
+      "step": 648
+    },
+    {
+      "epoch": 0.05691196156907988,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0019470000000000002,
+      "loss": 1.1887,
+      "step": 649
+    },
+    {
+      "epoch": 0.05699965334345443,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0019500000000000001,
+      "loss": 1.1739,
+      "step": 650
+    },
+    {
+      "epoch": 0.05708734511782897,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.001953,
+      "loss": 1.1621,
+      "step": 651
+    },
+    {
+      "epoch": 0.05717503689220352,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0019560000000000003,
+      "loss": 1.2449,
+      "step": 652
+    },
+    {
+      "epoch": 0.05726272866657806,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0019590000000000002,
+      "loss": 1.2406,
+      "step": 653
+    },
+    {
+      "epoch": 0.05735042044095261,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001962,
+      "loss": 1.2487,
+      "step": 654
+    },
+    {
+      "epoch": 0.05743811221532715,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.001965,
+      "loss": 1.1728,
+      "step": 655
+    },
+    {
+      "epoch": 0.057525803989701696,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.001968,
+      "loss": 1.2062,
+      "step": 656
+    },
+    {
+      "epoch": 0.05761349576407624,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001971,
+      "loss": 1.1946,
+      "step": 657
+    },
+    {
+      "epoch": 0.057701187538450785,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001974,
+      "loss": 1.1645,
+      "step": 658
+    },
+    {
+      "epoch": 0.057788879312825336,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.001977,
+      "loss": 1.1662,
+      "step": 659
+    },
+    {
+      "epoch": 0.05787657108719988,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00198,
+      "loss": 1.1606,
+      "step": 660
+    },
+    {
+      "epoch": 0.057964262861574425,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001983,
+      "loss": 1.1823,
+      "step": 661
+    },
+    {
+      "epoch": 0.05805195463594897,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0019860000000000004,
+      "loss": 1.2349,
+      "step": 662
+    },
+    {
+      "epoch": 0.058139646410323514,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0019890000000000003,
+      "loss": 1.1962,
+      "step": 663
+    },
+    {
+      "epoch": 0.05822733818469806,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019920000000000003,
+      "loss": 1.2109,
+      "step": 664
+    },
+    {
+      "epoch": 0.0583150299590726,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0019950000000000002,
+      "loss": 1.2259,
+      "step": 665
+    },
+    {
+      "epoch": 0.05840272173344715,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.001998,
+      "loss": 1.2124,
+      "step": 666
+    },
+    {
+      "epoch": 0.05849041350782169,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002001,
+      "loss": 1.1655,
+      "step": 667
+    },
+    {
+      "epoch": 0.058578105282196244,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002004,
+      "loss": 1.204,
+      "step": 668
+    },
+    {
+      "epoch": 0.05866579705657079,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002007,
+      "loss": 1.1704,
+      "step": 669
+    },
+    {
+      "epoch": 0.05875348883094533,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00201,
+      "loss": 1.1911,
+      "step": 670
+    },
+    {
+      "epoch": 0.05884118060531988,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002013,
+      "loss": 1.1369,
+      "step": 671
+    },
+    {
+      "epoch": 0.05892887237969442,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002016,
+      "loss": 1.1515,
+      "step": 672
+    },
+    {
+      "epoch": 0.059016564154068966,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002019,
+      "loss": 1.1569,
+      "step": 673
+    },
+    {
+      "epoch": 0.05910425592844351,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0020220000000000004,
+      "loss": 1.178,
+      "step": 674
+    },
+    {
+      "epoch": 0.059191947702818055,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0020250000000000003,
+      "loss": 1.1742,
+      "step": 675
+    },
+    {
+      "epoch": 0.0592796394771926,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0020280000000000003,
+      "loss": 1.2247,
+      "step": 676
+    },
+    {
+      "epoch": 0.059367331251567144,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0020310000000000003,
+      "loss": 1.2672,
+      "step": 677
+    },
+    {
+      "epoch": 0.059455023025941696,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0020340000000000002,
+      "loss": 1.1972,
+      "step": 678
+    },
+    {
+      "epoch": 0.05954271480031624,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002037,
+      "loss": 1.1777,
+      "step": 679
+    },
+    {
+      "epoch": 0.059630406574690785,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00204,
+      "loss": 1.1733,
+      "step": 680
+    },
+    {
+      "epoch": 0.05971809834906533,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002043,
+      "loss": 1.1494,
+      "step": 681
+    },
+    {
+      "epoch": 0.059805790123439874,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002046,
+      "loss": 1.1838,
+      "step": 682
+    },
+    {
+      "epoch": 0.05989348189781442,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002049,
+      "loss": 1.2084,
+      "step": 683
+    },
+    {
+      "epoch": 0.05998117367218896,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002052,
+      "loss": 1.1681,
+      "step": 684
+    },
+    {
+      "epoch": 0.06006886544656351,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0020550000000000004,
+      "loss": 1.2024,
+      "step": 685
+    },
+    {
+      "epoch": 0.06015655722093805,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0020580000000000004,
+      "loss": 1.1547,
+      "step": 686
+    },
+    {
+      "epoch": 0.0602442489953126,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0020610000000000003,
+      "loss": 1.226,
+      "step": 687
+    },
+    {
+      "epoch": 0.06033194076968715,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002064,
+      "loss": 1.1922,
+      "step": 688
+    },
+    {
+      "epoch": 0.06041963254406169,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002067,
+      "loss": 1.1949,
+      "step": 689
+    },
+    {
+      "epoch": 0.06050732431843624,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00207,
+      "loss": 1.2049,
+      "step": 690
+    },
+    {
+      "epoch": 0.06059501609281078,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0020729999999999998,
+      "loss": 1.1883,
+      "step": 691
+    },
+    {
+      "epoch": 0.060682707867185326,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0020759999999999997,
+      "loss": 1.2099,
+      "step": 692
+    },
+    {
+      "epoch": 0.06077039964155987,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0020789999999999997,
+      "loss": 1.1957,
+      "step": 693
+    },
+    {
+      "epoch": 0.060858091415934415,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002082,
+      "loss": 1.2148,
+      "step": 694
+    },
+    {
+      "epoch": 0.06094578319030896,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002085,
+      "loss": 1.1443,
+      "step": 695
+    },
+    {
+      "epoch": 0.06103347496468351,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002088,
+      "loss": 1.1784,
+      "step": 696
+    },
+    {
+      "epoch": 0.061121166739058055,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002091,
+      "loss": 1.1608,
+      "step": 697
+    },
+    {
+      "epoch": 0.0612088585134326,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002094,
+      "loss": 1.1311,
+      "step": 698
+    },
+    {
+      "epoch": 0.061296550287807144,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002097,
+      "loss": 1.2302,
+      "step": 699
+    },
+    {
+      "epoch": 0.06138424206218169,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0021,
+      "loss": 1.2455,
+      "step": 700
+    },
+    {
+      "epoch": 0.06147193383655623,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002103,
+      "loss": 1.2083,
+      "step": 701
+    },
+    {
+      "epoch": 0.06155962561093078,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002106,
+      "loss": 1.1682,
+      "step": 702
+    },
+    {
+      "epoch": 0.06164731738530532,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0021089999999999998,
+      "loss": 1.1818,
+      "step": 703
+    },
+    {
+      "epoch": 0.06173500915967987,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0021119999999999997,
+      "loss": 1.2536,
+      "step": 704
+    },
+    {
+      "epoch": 0.06182270093405442,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002115,
+      "loss": 1.154,
+      "step": 705
+    },
+    {
+      "epoch": 0.06191039270842896,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002118,
+      "loss": 1.1649,
+      "step": 706
+    },
+    {
+      "epoch": 0.06199808448280351,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002121,
+      "loss": 1.2484,
+      "step": 707
+    },
+    {
+      "epoch": 0.06208577625717805,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002124,
+      "loss": 1.2103,
+      "step": 708
+    },
+    {
+      "epoch": 0.062173468031552596,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002127,
+      "loss": 1.1832,
+      "step": 709
+    },
+    {
+      "epoch": 0.06226115980592714,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00213,
+      "loss": 1.1885,
+      "step": 710
+    },
+    {
+      "epoch": 0.062348851580301685,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002133,
+      "loss": 1.2414,
+      "step": 711
+    },
+    {
+      "epoch": 0.06243654335467623,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002136,
+      "loss": 1.2285,
+      "step": 712
+    },
+    {
+      "epoch": 0.06252423512905078,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002139,
+      "loss": 1.2218,
+      "step": 713
+    },
+    {
+      "epoch": 0.06261192690342532,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002142,
+      "loss": 1.201,
+      "step": 714
+    },
+    {
+      "epoch": 0.06269961867779987,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0021449999999999998,
+      "loss": 1.1846,
+      "step": 715
+    },
+    {
+      "epoch": 0.06278731045217441,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002148,
+      "loss": 1.1882,
+      "step": 716
+    },
+    {
+      "epoch": 0.06287500222654896,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002151,
+      "loss": 1.2032,
+      "step": 717
+    },
+    {
+      "epoch": 0.06296269400092351,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002154,
+      "loss": 1.2151,
+      "step": 718
+    },
+    {
+      "epoch": 0.06305038577529805,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002157,
+      "loss": 1.1866,
+      "step": 719
+    },
+    {
+      "epoch": 0.0631380775496726,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00216,
+      "loss": 1.1822,
+      "step": 720
+    },
+    {
+      "epoch": 0.06322576932404714,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002163,
+      "loss": 1.2161,
+      "step": 721
+    },
+    {
+      "epoch": 0.06331346109842169,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002166,
+      "loss": 1.1923,
+      "step": 722
+    },
+    {
+      "epoch": 0.06340115287279623,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002169,
+      "loss": 1.1727,
+      "step": 723
+    },
+    {
+      "epoch": 0.06348884464717078,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002172,
+      "loss": 1.1592,
+      "step": 724
+    },
+    {
+      "epoch": 0.06357653642154532,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002175,
+      "loss": 1.1971,
+      "step": 725
+    },
+    {
+      "epoch": 0.06366422819591987,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002178,
+      "loss": 1.2687,
+      "step": 726
+    },
+    {
+      "epoch": 0.0637519199702944,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0021809999999999998,
+      "loss": 1.2438,
+      "step": 727
+    },
+    {
+      "epoch": 0.06383961174466896,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002184,
+      "loss": 1.1864,
+      "step": 728
+    },
+    {
+      "epoch": 0.06392730351904351,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002187,
+      "loss": 1.1569,
+      "step": 729
+    },
+    {
+      "epoch": 0.06401499529341804,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00219,
+      "loss": 1.203,
+      "step": 730
+    },
+    {
+      "epoch": 0.0641026870677926,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002193,
+      "loss": 1.1815,
+      "step": 731
+    },
+    {
+      "epoch": 0.06419037884216713,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002196,
+      "loss": 1.208,
+      "step": 732
+    },
+    {
+      "epoch": 0.06427807061654169,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002199,
+      "loss": 1.1815,
+      "step": 733
+    },
+    {
+      "epoch": 0.06436576239091622,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002202,
+      "loss": 1.2005,
+      "step": 734
+    },
+    {
+      "epoch": 0.06445345416529077,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002205,
+      "loss": 1.1672,
+      "step": 735
+    },
+    {
+      "epoch": 0.06454114593966531,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002208,
+      "loss": 1.2257,
+      "step": 736
+    },
+    {
+      "epoch": 0.06462883771403986,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002211,
+      "loss": 1.1748,
+      "step": 737
+    },
+    {
+      "epoch": 0.06471652948841441,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002214,
+      "loss": 1.1862,
+      "step": 738
+    },
+    {
+      "epoch": 0.06480422126278895,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0022170000000000002,
+      "loss": 1.1603,
+      "step": 739
+    },
+    {
+      "epoch": 0.0648919130371635,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00222,
+      "loss": 1.2469,
+      "step": 740
+    },
+    {
+      "epoch": 0.06497960481153804,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002223,
+      "loss": 1.2134,
+      "step": 741
+    },
+    {
+      "epoch": 0.06506729658591259,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002226,
+      "loss": 1.1568,
+      "step": 742
+    },
+    {
+      "epoch": 0.06515498836028713,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002229,
+      "loss": 1.1611,
+      "step": 743
+    },
+    {
+      "epoch": 0.06524268013466168,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002232,
+      "loss": 1.2252,
+      "step": 744
+    },
+    {
+      "epoch": 0.06533037190903622,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002235,
+      "loss": 1.2526,
+      "step": 745
+    },
+    {
+      "epoch": 0.06541806368341077,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002238,
+      "loss": 1.2256,
+      "step": 746
+    },
+    {
+      "epoch": 0.06550575545778532,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002241,
+      "loss": 1.1456,
+      "step": 747
+    },
+    {
+      "epoch": 0.06559344723215986,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002244,
+      "loss": 1.1424,
+      "step": 748
+    },
+    {
+      "epoch": 0.06568113900653441,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002247,
+      "loss": 1.1713,
+      "step": 749
+    },
+    {
+      "epoch": 0.06576883078090895,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0022500000000000003,
+      "loss": 1.1616,
+      "step": 750
+    },
+    {
+      "epoch": 0.0658565225552835,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0022530000000000002,
+      "loss": 1.2177,
+      "step": 751
+    },
+    {
+      "epoch": 0.06594421432965804,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002256,
+      "loss": 1.23,
+      "step": 752
+    },
+    {
+      "epoch": 0.06603190610403259,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002259,
+      "loss": 1.1637,
+      "step": 753
+    },
+    {
+      "epoch": 0.06611959787840713,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002262,
+      "loss": 1.1783,
+      "step": 754
+    },
+    {
+      "epoch": 0.06620728965278168,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002265,
+      "loss": 1.158,
+      "step": 755
+    },
+    {
+      "epoch": 0.06629498142715623,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002268,
+      "loss": 1.1594,
+      "step": 756
+    },
+    {
+      "epoch": 0.06638267320153077,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002271,
+      "loss": 1.2228,
+      "step": 757
+    },
+    {
+      "epoch": 0.06647036497590532,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002274,
+      "loss": 1.1687,
+      "step": 758
+    },
+    {
+      "epoch": 0.06655805675027986,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002277,
+      "loss": 1.2647,
+      "step": 759
+    },
+    {
+      "epoch": 0.06664574852465441,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00228,
+      "loss": 1.1637,
+      "step": 760
+    },
+    {
+      "epoch": 0.06673344029902895,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002283,
+      "loss": 1.1811,
+      "step": 761
+    },
+    {
+      "epoch": 0.0668211320734035,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0022860000000000003,
+      "loss": 1.1874,
+      "step": 762
+    },
+    {
+      "epoch": 0.06690882384777803,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0022890000000000002,
+      "loss": 1.1863,
+      "step": 763
+    },
+    {
+      "epoch": 0.06699651562215259,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002292,
+      "loss": 1.195,
+      "step": 764
+    },
+    {
+      "epoch": 0.06708420739652714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002295,
+      "loss": 1.1976,
+      "step": 765
+    },
+    {
+      "epoch": 0.06717189917090167,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002298,
+      "loss": 1.179,
+      "step": 766
+    },
+    {
+      "epoch": 0.06725959094527623,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002301,
+      "loss": 1.2312,
+      "step": 767
+    },
+    {
+      "epoch": 0.06734728271965076,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002304,
+      "loss": 1.2145,
+      "step": 768
+    },
+    {
+      "epoch": 0.06743497449402532,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002307,
+      "loss": 1.1487,
+      "step": 769
+    },
+    {
+      "epoch": 0.06752266626839985,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00231,
+      "loss": 1.1608,
+      "step": 770
+    },
+    {
+      "epoch": 0.0676103580427744,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002313,
+      "loss": 1.1619,
+      "step": 771
+    },
+    {
+      "epoch": 0.06769804981714894,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002316,
+      "loss": 1.1695,
+      "step": 772
+    },
+    {
+      "epoch": 0.0677857415915235,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0023190000000000003,
+      "loss": 1.2122,
+      "step": 773
+    },
+    {
+      "epoch": 0.06787343336589804,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0023220000000000003,
+      "loss": 1.2475,
+      "step": 774
+    },
+    {
+      "epoch": 0.06796112514027258,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0023250000000000002,
+      "loss": 1.1552,
+      "step": 775
+    },
+    {
+      "epoch": 0.06804881691464713,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002328,
+      "loss": 1.1782,
+      "step": 776
+    },
+    {
+      "epoch": 0.06813650868902167,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002331,
+      "loss": 1.2141,
+      "step": 777
+    },
+    {
+      "epoch": 0.06822420046339622,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002334,
+      "loss": 1.225,
+      "step": 778
+    },
+    {
+      "epoch": 0.06831189223777076,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002337,
+      "loss": 1.1926,
+      "step": 779
+    },
+    {
+      "epoch": 0.06839958401214531,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00234,
+      "loss": 1.1816,
+      "step": 780
+    },
+    {
+      "epoch": 0.06848727578651985,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002343,
+      "loss": 1.2141,
+      "step": 781
+    },
+    {
+      "epoch": 0.0685749675608944,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002346,
+      "loss": 1.2328,
+      "step": 782
+    },
+    {
+      "epoch": 0.06866265933526895,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002349,
+      "loss": 1.189,
+      "step": 783
+    },
+    {
+      "epoch": 0.06875035110964349,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002352,
+      "loss": 1.1854,
+      "step": 784
+    },
+    {
+      "epoch": 0.06883804288401804,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0023550000000000003,
+      "loss": 1.219,
+      "step": 785
+    },
+    {
+      "epoch": 0.06892573465839258,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0023580000000000003,
+      "loss": 1.186,
+      "step": 786
+    },
+    {
+      "epoch": 0.06901342643276713,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0023610000000000003,
+      "loss": 1.2529,
+      "step": 787
+    },
+    {
+      "epoch": 0.06910111820714167,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002364,
+      "loss": 1.1786,
+      "step": 788
+    },
+    {
+      "epoch": 0.06918880998151622,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002367,
+      "loss": 1.2026,
+      "step": 789
+    },
+    {
+      "epoch": 0.06927650175589076,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00237,
+      "loss": 1.2605,
+      "step": 790
+    },
+    {
+      "epoch": 0.06936419353026531,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002373,
+      "loss": 1.2177,
+      "step": 791
+    },
+    {
+      "epoch": 0.06945188530463986,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002376,
+      "loss": 1.2597,
+      "step": 792
+    },
+    {
+      "epoch": 0.0695395770790144,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002379,
+      "loss": 1.2239,
+      "step": 793
+    },
+    {
+      "epoch": 0.06962726885338895,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002382,
+      "loss": 1.1098,
+      "step": 794
+    },
+    {
+      "epoch": 0.06971496062776349,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002385,
+      "loss": 1.2398,
+      "step": 795
+    },
+    {
+      "epoch": 0.06980265240213804,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0023880000000000004,
+      "loss": 1.2229,
+      "step": 796
+    },
+    {
+      "epoch": 0.06989034417651258,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0023910000000000003,
+      "loss": 1.1818,
+      "step": 797
+    },
+    {
+      "epoch": 0.06997803595088713,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0023940000000000003,
+      "loss": 1.1801,
+      "step": 798
+    },
+    {
+      "epoch": 0.07006572772526166,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0023970000000000003,
+      "loss": 1.2839,
+      "step": 799
+    },
+    {
+      "epoch": 0.07015341949963622,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 1.17,
+      "step": 800
+    },
+    {
+      "epoch": 0.07024111127401077,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002403,
+      "loss": 1.1887,
+      "step": 801
+    },
+    {
+      "epoch": 0.0703288030483853,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002406,
+      "loss": 1.2965,
+      "step": 802
+    },
+    {
+      "epoch": 0.07041649482275986,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002409,
+      "loss": 1.1965,
+      "step": 803
+    },
+    {
+      "epoch": 0.0705041865971344,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002412,
+      "loss": 1.169,
+      "step": 804
+    },
+    {
+      "epoch": 0.07059187837150895,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002415,
+      "loss": 1.2176,
+      "step": 805
+    },
+    {
+      "epoch": 0.07067957014588348,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002418,
+      "loss": 1.1354,
+      "step": 806
+    },
+    {
+      "epoch": 0.07076726192025803,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0024210000000000004,
+      "loss": 1.1584,
+      "step": 807
+    },
+    {
+      "epoch": 0.07085495369463257,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0024240000000000004,
+      "loss": 1.1809,
+      "step": 808
+    },
+    {
+      "epoch": 0.07094264546900712,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0024270000000000003,
+      "loss": 1.1324,
+      "step": 809
+    },
+    {
+      "epoch": 0.07103033724338167,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0024300000000000003,
+      "loss": 1.2154,
+      "step": 810
+    },
+    {
+      "epoch": 0.07111802901775621,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0024330000000000003,
+      "loss": 1.1983,
+      "step": 811
+    },
+    {
+      "epoch": 0.07120572079213076,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0024360000000000002,
+      "loss": 1.2254,
+      "step": 812
+    },
+    {
+      "epoch": 0.0712934125665053,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0024389999999999998,
+      "loss": 1.165,
+      "step": 813
+    },
+    {
+      "epoch": 0.07138110434087985,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0024419999999999997,
+      "loss": 1.2106,
+      "step": 814
+    },
+    {
+      "epoch": 0.07146879611525439,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0024449999999999997,
+      "loss": 1.1528,
+      "step": 815
+    },
+    {
+      "epoch": 0.07155648788962894,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002448,
+      "loss": 1.1848,
+      "step": 816
+    },
+    {
+      "epoch": 0.07164417966400348,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.002451,
+      "loss": 1.1658,
+      "step": 817
+    },
+    {
+      "epoch": 0.07173187143837803,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002454,
+      "loss": 1.2963,
+      "step": 818
+    },
+    {
+      "epoch": 0.07181956321275258,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002457,
+      "loss": 1.1949,
+      "step": 819
+    },
+    {
+      "epoch": 0.07190725498712712,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00246,
+      "loss": 1.2026,
+      "step": 820
+    },
+    {
+      "epoch": 0.07199494676150167,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002463,
+      "loss": 1.187,
+      "step": 821
+    },
+    {
+      "epoch": 0.07208263853587621,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002466,
+      "loss": 1.2453,
+      "step": 822
+    },
+    {
+      "epoch": 0.07217033031025076,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002469,
+      "loss": 1.2092,
+      "step": 823
+    },
+    {
+      "epoch": 0.0722580220846253,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002472,
+      "loss": 1.2321,
+      "step": 824
+    },
+    {
+      "epoch": 0.07234571385899985,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0024749999999999998,
+      "loss": 1.1999,
+      "step": 825
+    },
+    {
+      "epoch": 0.07243340563337439,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0024779999999999997,
+      "loss": 1.2179,
+      "step": 826
+    },
+    {
+      "epoch": 0.07252109740774894,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002481,
+      "loss": 1.2094,
+      "step": 827
+    },
+    {
+      "epoch": 0.07260878918212349,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002484,
+      "loss": 1.2573,
+      "step": 828
+    },
+    {
+      "epoch": 0.07269648095649803,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002487,
+      "loss": 1.2118,
+      "step": 829
+    },
+    {
+      "epoch": 0.07278417273087258,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00249,
+      "loss": 1.1918,
+      "step": 830
+    },
+    {
+      "epoch": 0.07287186450524712,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002493,
+      "loss": 1.1716,
+      "step": 831
+    },
+    {
+      "epoch": 0.07295955627962167,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002496,
+      "loss": 1.2032,
+      "step": 832
+    },
+    {
+      "epoch": 0.0730472480539962,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002499,
+      "loss": 1.2014,
+      "step": 833
+    },
+    {
+      "epoch": 0.07313493982837076,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002502,
+      "loss": 1.2087,
+      "step": 834
+    },
+    {
+      "epoch": 0.0732226316027453,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002505,
+      "loss": 1.1957,
+      "step": 835
+    },
+    {
+      "epoch": 0.07331032337711985,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002508,
+      "loss": 1.217,
+      "step": 836
+    },
+    {
+      "epoch": 0.0733980151514944,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025109999999999998,
+      "loss": 1.2022,
+      "step": 837
+    },
+    {
+      "epoch": 0.07348570692586893,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0025139999999999997,
+      "loss": 1.2414,
+      "step": 838
+    },
+    {
+      "epoch": 0.07357339870024349,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002517,
+      "loss": 1.2123,
+      "step": 839
+    },
+    {
+      "epoch": 0.07366109047461802,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00252,
+      "loss": 1.1463,
+      "step": 840
+    },
+    {
+      "epoch": 0.07374878224899258,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002523,
+      "loss": 1.24,
+      "step": 841
+    },
+    {
+      "epoch": 0.07383647402336711,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002526,
+      "loss": 1.2643,
+      "step": 842
+    },
+    {
+      "epoch": 0.07392416579774166,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002529,
+      "loss": 1.1958,
+      "step": 843
+    },
+    {
+      "epoch": 0.0740118575721162,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002532,
+      "loss": 1.1996,
+      "step": 844
+    },
+    {
+      "epoch": 0.07409954934649075,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002535,
+      "loss": 1.1694,
+      "step": 845
+    },
+    {
+      "epoch": 0.0741872411208653,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002538,
+      "loss": 1.2552,
+      "step": 846
+    },
+    {
+      "epoch": 0.07427493289523984,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002541,
+      "loss": 1.2317,
+      "step": 847
+    },
+    {
+      "epoch": 0.0743626246696144,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002544,
+      "loss": 1.2529,
+      "step": 848
+    },
+    {
+      "epoch": 0.07445031644398893,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002547,
+      "loss": 1.1682,
+      "step": 849
+    },
+    {
+      "epoch": 0.07453800821836348,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00255,
+      "loss": 1.2095,
+      "step": 850
+    },
+    {
+      "epoch": 0.07462569999273802,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002553,
+      "loss": 1.3158,
+      "step": 851
+    },
+    {
+      "epoch": 0.07471339176711257,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002556,
+      "loss": 1.2173,
+      "step": 852
+    },
+    {
+      "epoch": 0.07480108354148711,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002559,
+      "loss": 1.2856,
+      "step": 853
+    },
+    {
+      "epoch": 0.07488877531586166,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002562,
+      "loss": 1.2333,
+      "step": 854
+    },
+    {
+      "epoch": 0.07497646709023621,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.002565,
+      "loss": 1.2129,
+      "step": 855
+    },
+    {
+      "epoch": 0.07506415886461075,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002568,
+      "loss": 1.2414,
+      "step": 856
+    },
+    {
+      "epoch": 0.0751518506389853,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002571,
+      "loss": 1.2281,
+      "step": 857
+    },
+    {
+      "epoch": 0.07523954241335984,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002574,
+      "loss": 1.2306,
+      "step": 858
+    },
+    {
+      "epoch": 0.07532723418773439,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002577,
+      "loss": 1.2104,
+      "step": 859
+    },
+    {
+      "epoch": 0.07541492596210893,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00258,
+      "loss": 1.2253,
+      "step": 860
+    },
+    {
+      "epoch": 0.07550261773648348,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0025830000000000002,
+      "loss": 1.2411,
+      "step": 861
+    },
+    {
+      "epoch": 0.07559030951085802,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002586,
+      "loss": 1.1738,
+      "step": 862
+    },
+    {
+      "epoch": 0.07567800128523257,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002589,
+      "loss": 1.2206,
+      "step": 863
+    },
+    {
+      "epoch": 0.07576569305960712,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002592,
+      "loss": 1.2041,
+      "step": 864
+    },
+    {
+      "epoch": 0.07585338483398166,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002595,
+      "loss": 1.1726,
+      "step": 865
+    },
+    {
+      "epoch": 0.07594107660835621,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002598,
+      "loss": 1.2984,
+      "step": 866
+    },
+    {
+      "epoch": 0.07602876838273075,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002601,
+      "loss": 1.1888,
+      "step": 867
+    },
+    {
+      "epoch": 0.0761164601571053,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.002604,
+      "loss": 1.1803,
+      "step": 868
+    },
+    {
+      "epoch": 0.07620415193147984,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002607,
+      "loss": 1.1997,
+      "step": 869
+    },
+    {
+      "epoch": 0.07629184370585439,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00261,
+      "loss": 1.2026,
+      "step": 870
+    },
+    {
+      "epoch": 0.07637953548022892,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002613,
+      "loss": 1.2052,
+      "step": 871
+    },
+    {
+      "epoch": 0.07646722725460348,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.002616,
+      "loss": 1.1758,
+      "step": 872
+    },
+    {
+      "epoch": 0.07655491902897803,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0026190000000000002,
+      "loss": 1.2126,
+      "step": 873
+    },
+    {
+      "epoch": 0.07664261080335256,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002622,
+      "loss": 1.1752,
+      "step": 874
+    },
+    {
+      "epoch": 0.07673030257772712,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.002625,
+      "loss": 1.1959,
+      "step": 875
+    },
+    {
+      "epoch": 0.07681799435210165,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002628,
+      "loss": 1.1591,
+      "step": 876
+    },
+    {
+      "epoch": 0.0769056861264762,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002631,
+      "loss": 1.2444,
+      "step": 877
+    },
+    {
+      "epoch": 0.07699337790085074,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002634,
+      "loss": 1.2269,
+      "step": 878
+    },
+    {
+      "epoch": 0.0770810696752253,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002637,
+      "loss": 1.199,
+      "step": 879
+    },
+    {
+      "epoch": 0.07716876144959983,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00264,
+      "loss": 1.2457,
+      "step": 880
+    },
+    {
+      "epoch": 0.07725645322397438,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002643,
+      "loss": 1.1842,
+      "step": 881
+    },
+    {
+      "epoch": 0.07734414499834893,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002646,
+      "loss": 1.1871,
+      "step": 882
+    },
+    {
+      "epoch": 0.07743183677272347,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.002649,
+      "loss": 1.248,
+      "step": 883
+    },
+    {
+      "epoch": 0.07751952854709802,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0026520000000000003,
+      "loss": 1.2379,
+      "step": 884
+    },
+    {
+      "epoch": 0.07760722032147256,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0026550000000000002,
+      "loss": 1.1706,
+      "step": 885
+    },
+    {
+      "epoch": 0.07769491209584711,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002658,
+      "loss": 1.2141,
+      "step": 886
+    },
+    {
+      "epoch": 0.07778260387022165,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002661,
+      "loss": 1.1919,
+      "step": 887
+    },
+    {
+      "epoch": 0.0778702956445962,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002664,
+      "loss": 1.2467,
+      "step": 888
+    },
+    {
+      "epoch": 0.07795798741897074,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002667,
+      "loss": 1.1961,
+      "step": 889
+    },
+    {
+      "epoch": 0.07804567919334529,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00267,
+      "loss": 1.2178,
+      "step": 890
+    },
+    {
+      "epoch": 0.07813337096771984,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002673,
+      "loss": 1.1719,
+      "step": 891
+    },
+    {
+      "epoch": 0.07822106274209438,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002676,
+      "loss": 1.2323,
+      "step": 892
+    },
+    {
+      "epoch": 0.07830875451646893,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002679,
+      "loss": 1.2321,
+      "step": 893
+    },
+    {
+      "epoch": 0.07839644629084347,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002682,
+      "loss": 1.211,
+      "step": 894
+    },
+    {
+      "epoch": 0.07848413806521802,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0026850000000000003,
+      "loss": 1.2103,
+      "step": 895
+    },
+    {
+      "epoch": 0.07857182983959256,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0026880000000000003,
+      "loss": 1.1379,
+      "step": 896
+    },
+    {
+      "epoch": 0.07865952161396711,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0026910000000000002,
+      "loss": 1.2037,
+      "step": 897
+    },
+    {
+      "epoch": 0.07874721338834165,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002694,
+      "loss": 1.2589,
+      "step": 898
+    },
+    {
+      "epoch": 0.0788349051627162,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002697,
+      "loss": 1.1949,
+      "step": 899
+    },
+    {
+      "epoch": 0.07892259693709075,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0027,
+      "loss": 1.182,
+      "step": 900
+    },
+    {
+      "epoch": 0.07901028871146529,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002703,
+      "loss": 1.1464,
+      "step": 901
+    },
+    {
+      "epoch": 0.07909798048583984,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002706,
+      "loss": 1.2589,
+      "step": 902
+    },
+    {
+      "epoch": 0.07918567226021438,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002709,
+      "loss": 1.148,
+      "step": 903
+    },
+    {
+      "epoch": 0.07927336403458893,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002712,
+      "loss": 1.1859,
+      "step": 904
+    },
+    {
+      "epoch": 0.07936105580896347,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002715,
+      "loss": 1.1813,
+      "step": 905
+    },
+    {
+      "epoch": 0.07944874758333802,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002718,
+      "loss": 1.2407,
+      "step": 906
+    },
+    {
+      "epoch": 0.07953643935771255,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0027210000000000003,
+      "loss": 1.2042,
+      "step": 907
+    },
+    {
+      "epoch": 0.0796241311320871,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0027240000000000003,
+      "loss": 1.279,
+      "step": 908
+    },
+    {
+      "epoch": 0.07971182290646166,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0027270000000000003,
+      "loss": 1.2023,
+      "step": 909
+    },
+    {
+      "epoch": 0.0797995146808362,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0027300000000000002,
+      "loss": 1.2134,
+      "step": 910
+    },
+    {
+      "epoch": 0.07988720645521075,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.002733,
+      "loss": 1.2442,
+      "step": 911
+    },
+    {
+      "epoch": 0.07997489822958528,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002736,
+      "loss": 1.2028,
+      "step": 912
+    },
+    {
+      "epoch": 0.08006259000395984,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002739,
+      "loss": 1.2103,
+      "step": 913
+    },
+    {
+      "epoch": 0.08015028177833437,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002742,
+      "loss": 1.1703,
+      "step": 914
+    },
+    {
+      "epoch": 0.08023797355270892,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002745,
+      "loss": 1.2154,
+      "step": 915
+    },
+    {
+      "epoch": 0.08032566532708346,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002748,
+      "loss": 1.2085,
+      "step": 916
+    },
+    {
+      "epoch": 0.08041335710145801,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002751,
+      "loss": 1.2022,
+      "step": 917
+    },
+    {
+      "epoch": 0.08050104887583256,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0027540000000000004,
+      "loss": 1.2548,
+      "step": 918
+    },
+    {
+      "epoch": 0.0805887406502071,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0027570000000000003,
+      "loss": 1.1738,
+      "step": 919
+    },
+    {
+      "epoch": 0.08067643242458165,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0027600000000000003,
+      "loss": 1.1934,
+      "step": 920
+    },
+    {
+      "epoch": 0.08076412419895619,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0027630000000000003,
+      "loss": 1.2174,
+      "step": 921
+    },
+    {
+      "epoch": 0.08085181597333074,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0027660000000000002,
+      "loss": 1.2276,
+      "step": 922
+    },
+    {
+      "epoch": 0.08093950774770528,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002769,
+      "loss": 1.124,
+      "step": 923
+    },
+    {
+      "epoch": 0.08102719952207983,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002772,
+      "loss": 1.2187,
+      "step": 924
+    },
+    {
+      "epoch": 0.08111489129645437,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002775,
+      "loss": 1.2156,
+      "step": 925
+    },
+    {
+      "epoch": 0.08120258307082892,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002778,
+      "loss": 1.2305,
+      "step": 926
+    },
+    {
+      "epoch": 0.08129027484520347,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002781,
+      "loss": 1.236,
+      "step": 927
+    },
+    {
+      "epoch": 0.08137796661957801,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002784,
+      "loss": 1.2164,
+      "step": 928
+    },
+    {
+      "epoch": 0.08146565839395256,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0027870000000000004,
+      "loss": 1.232,
+      "step": 929
+    },
+    {
+      "epoch": 0.0815533501683271,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0027900000000000004,
+      "loss": 1.1919,
+      "step": 930
+    },
+    {
+      "epoch": 0.08164104194270165,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0027930000000000003,
+      "loss": 1.2217,
+      "step": 931
+    },
+    {
+      "epoch": 0.08172873371707619,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0027960000000000003,
+      "loss": 1.24,
+      "step": 932
+    },
+    {
+      "epoch": 0.08181642549145074,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027990000000000003,
+      "loss": 1.1977,
+      "step": 933
+    },
+    {
+      "epoch": 0.08190411726582528,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0028020000000000002,
+      "loss": 1.3008,
+      "step": 934
+    },
+    {
+      "epoch": 0.08199180904019983,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002805,
+      "loss": 1.1938,
+      "step": 935
+    },
+    {
+      "epoch": 0.08207950081457437,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002808,
+      "loss": 1.2427,
+      "step": 936
+    },
+    {
+      "epoch": 0.08216719258894892,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002811,
+      "loss": 1.2226,
+      "step": 937
+    },
+    {
+      "epoch": 0.08225488436332347,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002814,
+      "loss": 1.2063,
+      "step": 938
+    },
+    {
+      "epoch": 0.082342576137698,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002817,
+      "loss": 1.1925,
+      "step": 939
+    },
+    {
+      "epoch": 0.08243026791207256,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00282,
+      "loss": 1.1635,
+      "step": 940
+    },
+    {
+      "epoch": 0.0825179596864471,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002823,
+      "loss": 1.1588,
+      "step": 941
+    },
+    {
+      "epoch": 0.08260565146082165,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002826,
+      "loss": 1.2534,
+      "step": 942
+    },
+    {
+      "epoch": 0.08269334323519618,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002829,
+      "loss": 1.253,
+      "step": 943
+    },
+    {
+      "epoch": 0.08278103500957074,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002832,
+      "loss": 1.1999,
+      "step": 944
+    },
+    {
+      "epoch": 0.08286872678394527,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002835,
+      "loss": 1.2377,
+      "step": 945
+    },
+    {
+      "epoch": 0.08295641855831982,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002838,
+      "loss": 1.1636,
+      "step": 946
+    },
+    {
+      "epoch": 0.08304411033269438,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0028409999999999998,
+      "loss": 1.1766,
+      "step": 947
+    },
+    {
+      "epoch": 0.08313180210706891,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0028439999999999997,
+      "loss": 1.2019,
+      "step": 948
+    },
+    {
+      "epoch": 0.08321949388144347,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002847,
+      "loss": 1.1763,
+      "step": 949
+    },
+    {
+      "epoch": 0.083307185655818,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00285,
+      "loss": 1.2541,
+      "step": 950
+    },
+    {
+      "epoch": 0.08339487743019255,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002853,
+      "loss": 1.2463,
+      "step": 951
+    },
+    {
+      "epoch": 0.08348256920456709,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002856,
+      "loss": 1.2397,
+      "step": 952
+    },
+    {
+      "epoch": 0.08357026097894164,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002859,
+      "loss": 1.2318,
+      "step": 953
+    },
+    {
+      "epoch": 0.08365795275331618,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002862,
+      "loss": 1.2518,
+      "step": 954
+    },
+    {
+      "epoch": 0.08374564452769073,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002865,
+      "loss": 1.2705,
+      "step": 955
+    },
+    {
+      "epoch": 0.08383333630206528,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002868,
+      "loss": 1.2331,
+      "step": 956
+    },
+    {
+      "epoch": 0.08392102807643982,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002871,
+      "loss": 1.2133,
+      "step": 957
+    },
+    {
+      "epoch": 0.08400871985081437,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002874,
+      "loss": 1.1885,
+      "step": 958
+    },
+    {
+      "epoch": 0.08409641162518891,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002877,
+      "loss": 1.1859,
+      "step": 959
+    },
+    {
+      "epoch": 0.08418410339956346,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028799999999999997,
+      "loss": 1.1881,
+      "step": 960
+    },
+    {
+      "epoch": 0.084271795173938,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002883,
+      "loss": 1.2317,
+      "step": 961
+    },
+    {
+      "epoch": 0.08435948694831255,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002886,
+      "loss": 1.212,
+      "step": 962
+    },
+    {
+      "epoch": 0.08444717872268709,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002889,
+      "loss": 1.2385,
+      "step": 963
+    },
+    {
+      "epoch": 0.08453487049706164,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002892,
+      "loss": 1.1706,
+      "step": 964
+    },
+    {
+      "epoch": 0.08462256227143619,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002895,
+      "loss": 1.2452,
+      "step": 965
+    },
+    {
+      "epoch": 0.08471025404581073,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002898,
+      "loss": 1.2055,
+      "step": 966
+    },
+    {
+      "epoch": 0.08479794582018528,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002901,
+      "loss": 1.2252,
+      "step": 967
+    },
+    {
+      "epoch": 0.08488563759455982,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002904,
+      "loss": 1.198,
+      "step": 968
+    },
+    {
+      "epoch": 0.08497332936893437,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002907,
+      "loss": 1.2776,
+      "step": 969
+    },
+    {
+      "epoch": 0.08506102114330891,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00291,
+      "loss": 1.2509,
+      "step": 970
+    },
+    {
+      "epoch": 0.08514871291768346,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002913,
+      "loss": 1.1464,
+      "step": 971
+    },
+    {
+      "epoch": 0.085236404692058,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002916,
+      "loss": 1.2446,
+      "step": 972
+    },
+    {
+      "epoch": 0.08532409646643255,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.002919,
+      "loss": 1.2405,
+      "step": 973
+    },
+    {
+      "epoch": 0.0854117882408071,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002922,
+      "loss": 1.2451,
+      "step": 974
+    },
+    {
+      "epoch": 0.08549948001518164,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.002925,
+      "loss": 1.2476,
+      "step": 975
+    },
+    {
+      "epoch": 0.08558717178955619,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002928,
+      "loss": 1.2005,
+      "step": 976
+    },
+    {
+      "epoch": 0.08567486356393073,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.002931,
+      "loss": 1.2115,
+      "step": 977
+    },
+    {
+      "epoch": 0.08576255533830528,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002934,
+      "loss": 1.1749,
+      "step": 978
+    },
+    {
+      "epoch": 0.08585024711267981,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002937,
+      "loss": 1.2287,
+      "step": 979
+    },
+    {
+      "epoch": 0.08593793888705437,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00294,
+      "loss": 1.1517,
+      "step": 980
+    },
+    {
+      "epoch": 0.0860256306614289,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002943,
+      "loss": 1.244,
+      "step": 981
+    },
+    {
+      "epoch": 0.08611332243580345,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002946,
+      "loss": 1.1973,
+      "step": 982
+    },
+    {
+      "epoch": 0.086201014210178,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0029490000000000002,
+      "loss": 1.2874,
+      "step": 983
+    },
+    {
+      "epoch": 0.08628870598455254,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002952,
+      "loss": 1.2582,
+      "step": 984
+    },
+    {
+      "epoch": 0.0863763977589271,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002955,
+      "loss": 1.2256,
+      "step": 985
+    },
+    {
+      "epoch": 0.08646408953330163,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002958,
+      "loss": 1.219,
+      "step": 986
+    },
+    {
+      "epoch": 0.08655178130767618,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002961,
+      "loss": 1.2202,
+      "step": 987
+    },
+    {
+      "epoch": 0.08663947308205072,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002964,
+      "loss": 1.2885,
+      "step": 988
+    },
+    {
+      "epoch": 0.08672716485642527,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002967,
+      "loss": 1.2178,
+      "step": 989
+    },
+    {
+      "epoch": 0.08681485663079981,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00297,
+      "loss": 1.2682,
+      "step": 990
+    },
+    {
+      "epoch": 0.08690254840517436,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002973,
+      "loss": 1.2171,
+      "step": 991
+    },
+    {
+      "epoch": 0.08699024017954891,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002976,
+      "loss": 1.2959,
+      "step": 992
+    },
+    {
+      "epoch": 0.08707793195392345,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002979,
+      "loss": 1.2449,
+      "step": 993
+    },
+    {
+      "epoch": 0.087165623728298,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002982,
+      "loss": 1.2013,
+      "step": 994
+    },
+    {
+      "epoch": 0.08725331550267254,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0029850000000000002,
+      "loss": 1.1975,
+      "step": 995
+    },
+    {
+      "epoch": 0.08734100727704709,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002988,
+      "loss": 1.2065,
+      "step": 996
+    },
+    {
+      "epoch": 0.08742869905142163,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002991,
+      "loss": 1.2246,
+      "step": 997
+    },
+    {
+      "epoch": 0.08751639082579618,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002994,
+      "loss": 1.3003,
+      "step": 998
+    },
+    {
+      "epoch": 0.08760408260017072,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002997,
+      "loss": 1.1732,
+      "step": 999
+    },
+    {
+      "epoch": 0.08769177437454527,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.003,
+      "loss": 1.2027,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08769177437454527,
+      "eval_loss": 1.2352585792541504,
+      "eval_runtime": 429.1616,
+      "eval_samples_per_second": 33.663,
+      "eval_steps_per_second": 8.416,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08777946614891982,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029999999384417424,
+      "loss": 1.2718,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08786715792329436,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029999997537669756,
+      "loss": 1.2869,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08795484969766891,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002999999445975716,
+      "loss": 1.2037,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08804254147204345,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029999990150679926,
+      "loss": 1.2742,
+      "step": 1004
+    },
+    {
+      "epoch": 0.088130233246418,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.002999998461043843,
+      "loss": 1.2448,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08821792502079254,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029999977839033198,
+      "loss": 1.1709,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08830561679516709,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0029999969836464833,
+      "loss": 1.222,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08839330856954163,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002999996060273407,
+      "loss": 1.2483,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08848100034391618,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029999950137841744,
+      "loss": 1.1605,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08856869211829073,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002999993844178882,
+      "loss": 1.2067,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08865638389266527,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002999992551457636,
+      "loss": 1.2093,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08874407566703982,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002999991135620554,
+      "loss": 1.2135,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08883176744141436,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029999895966677658,
+      "loss": 1.1967,
+      "step": 1013
+    },
+    {
+      "epoch": 0.0889194592157889,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0029999879345994113,
+      "loss": 1.2596,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08900715099016344,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002999986149415642,
+      "loss": 1.205,
+      "step": 1015
+    },
+    {
+      "epoch": 0.089094842764538,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002999984241116621,
+      "loss": 1.2303,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08918253453891253,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029999822097025223,
+      "loss": 1.2616,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08927022631328708,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029999800551735304,
+      "loss": 1.2431,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08935791808766164,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002999977777529843,
+      "loss": 1.199,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08944560986203617,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002999975376771667,
+      "loss": 1.2637,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08953330163641073,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0029999728528992214,
+      "loss": 1.2013,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08962099341078526,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002999970205912737,
+      "loss": 1.2653,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08970868518515981,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002999967435812455,
+      "loss": 1.27,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08979637695953435,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029999645425986265,
+      "loss": 1.2497,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0898840687339089,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029999615262715183,
+      "loss": 1.2822,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08997176050828344,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029999583868314025,
+      "loss": 1.2212,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09005945228265799,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0029999551242785674,
+      "loss": 1.2082,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09014714405703254,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0029999517386133097,
+      "loss": 1.1986,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09023483583140708,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029999482298359386,
+      "loss": 1.2419,
+      "step": 1029
+    },
+    {
+      "epoch": 0.09032252760578163,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002999944597946773,
+      "loss": 1.1901,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09041021938015617,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0029999408429461456,
+      "loss": 1.2331,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09049791115453072,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029999369648343976,
+      "loss": 1.2162,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09058560292890526,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029999329636118837,
+      "loss": 1.1974,
+      "step": 1033
+    },
+    {
+      "epoch": 0.09067329470327981,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029999288392789686,
+      "loss": 1.2497,
+      "step": 1034
+    },
+    {
+      "epoch": 0.09076098647765435,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002999924591836028,
+      "loss": 1.2433,
+      "step": 1035
+    },
+    {
+      "epoch": 0.0908486782520289,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002999920221283449,
+      "loss": 1.2738,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09093637002640345,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002999915727621631,
+      "loss": 1.2344,
+      "step": 1037
+    },
+    {
+      "epoch": 0.09102406180077799,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0029999111108509834,
+      "loss": 1.213,
+      "step": 1038
+    },
+    {
+      "epoch": 0.09111175357515254,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0029999063709719278,
+      "loss": 1.2075,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09119944534952708,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002999901507984895,
+      "loss": 1.2245,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09128713712390163,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029998965218903297,
+      "loss": 1.2,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09137482889827617,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0029998914126886864,
+      "loss": 1.2469,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09146252067265072,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002999886180380431,
+      "loss": 1.2332,
+      "step": 1043
+    },
+    {
+      "epoch": 0.09155021244702526,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029998808249660407,
+      "loss": 1.2035,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09163790422139981,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029998753464460038,
+      "loss": 1.2663,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09172559599577436,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029998697448208205,
+      "loss": 1.2392,
+      "step": 1046
+    },
+    {
+      "epoch": 0.0918132877701489,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0029998640200910006,
+      "loss": 1.2818,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09190097954452345,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002999858172257067,
+      "loss": 1.261,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09198867131889799,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0029998522013195525,
+      "loss": 1.2804,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09207636309327254,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0029998461072790017,
+      "loss": 1.2455,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09216405486764707,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002999839890135971,
+      "loss": 1.22,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09225174664202163,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029998335498910263,
+      "loss": 1.2181,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09233943841639616,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002999827086544747,
+      "loss": 1.2936,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09242713019077071,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0029998205000977213,
+      "loss": 1.215,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09251482196514527,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029998137905505513,
+      "loss": 1.2093,
+      "step": 1055
+    },
+    {
+      "epoch": 0.0926025137395198,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029998069579038476,
+      "loss": 1.2312,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09269020551389436,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029998000021582345,
+      "loss": 1.1841,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09277789728826889,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002999792923314345,
+      "loss": 1.2149,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09286558906264344,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029997857213728257,
+      "loss": 1.2036,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09295328083701798,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029997783963343332,
+      "loss": 1.2765,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09304097261139253,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002999770948199535,
+      "loss": 1.2624,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09312866438576707,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002999763376969111,
+      "loss": 1.2227,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09321635616014162,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002999755682643751,
+      "loss": 1.2304,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09330404793451617,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0029997478652241576,
+      "loss": 1.2698,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09339173970889071,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002999739924711043,
+      "loss": 1.2416,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09347943148326526,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002999731861105132,
+      "loss": 1.2739,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0935671232576398,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.002999723674407159,
+      "loss": 1.241,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09365481503201435,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029997153646178715,
+      "loss": 1.2133,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09374250680638889,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0029997069317380274,
+      "loss": 1.2393,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09383019858076344,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002999698375768395,
+      "loss": 1.2339,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09391789035513798,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002999689696709755,
+      "loss": 1.2435,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09400558212951253,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0029996808945628984,
+      "loss": 1.2748,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09409327390388708,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002999671969328629,
+      "loss": 1.2199,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09418096567826162,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00299966292100776,
+      "loss": 1.2602,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09426865745263617,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0029996537496011166,
+      "loss": 1.2514,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09435634922701071,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0029996444551095352,
+      "loss": 1.2206,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09444404100138526,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002999635037533864,
+      "loss": 1.2731,
+      "step": 1077
+    },
+    {
+      "epoch": 0.0945317327757598,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0029996254968749614,
+      "loss": 1.2356,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09461942455013435,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002999615833133697,
+      "loss": 1.2592,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09470711632450889,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029996060463109535,
+      "loss": 1.1929,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09479480809888344,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002999596136407622,
+      "loss": 1.2117,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09488249987325799,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002999586103424607,
+      "loss": 1.173,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09497019164763253,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0029995759473628227,
+      "loss": 1.2415,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09505788342200708,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0029995656682231964,
+      "loss": 1.233,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09514557519638162,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029995552660066656,
+      "loss": 1.2176,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09523326697075617,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0029995447407141777,
+      "loss": 1.2514,
+      "step": 1086
+    },
+    {
+      "epoch": 0.0953209587451307,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0029995340923466935,
+      "loss": 1.1719,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09540865051950526,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029995233209051835,
+      "loss": 1.2674,
+      "step": 1088
+    },
+    {
+      "epoch": 0.0954963422938798,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002999512426390631,
+      "loss": 1.1974,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09558403406825434,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0029995014088040287,
+      "loss": 1.2171,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0956717258426289,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0029994902681463815,
+      "loss": 1.2789,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09575941761700343,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029994790044187056,
+      "loss": 1.1824,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09584710939137799,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002999467617622028,
+      "loss": 1.1969,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09593480116575252,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002999456107757388,
+      "loss": 1.2936,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09602249294012707,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0029994444748258344,
+      "loss": 1.2661,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09611018471450161,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0029994327188284276,
+      "loss": 1.2637,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09619787648887616,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002999420839766241,
+      "loss": 1.2525,
+      "step": 1097
+    },
+    {
+      "epoch": 0.0962855682632507,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002999408837640357,
+      "loss": 1.2274,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09637326003762525,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029993967124518706,
+      "loss": 1.1619,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0964609518119998,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029993844642018873,
+      "loss": 1.1803,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09654864358637434,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029993720928915245,
+      "loss": 1.2556,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09663633536074889,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029993595985219105,
+      "loss": 1.2952,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09672402713512343,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029993469810941837,
+      "loss": 1.2566,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09681171890949798,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0029993342406094965,
+      "loss": 1.1917,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09689941068387252,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002999321377069009,
+      "loss": 1.2433,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09698710245824707,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029993083904738954,
+      "loss": 1.2338,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09707479423262161,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00299929528082534,
+      "loss": 1.2022,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09716248600699616,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0029992820481245385,
+      "loss": 1.2294,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09725017778137071,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029992686923726963,
+      "loss": 1.2095,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09733786955574525,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029992552135710334,
+      "loss": 1.1755,
+      "step": 1110
+    },
+    {
+      "epoch": 0.0974255613301198,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0029992416117207775,
+      "loss": 1.2073,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09751325310449434,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00299922788682317,
+      "loss": 1.2102,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09760094487886889,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0029992140388794626,
+      "loss": 1.2712,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09768863665324343,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029992000678909173,
+      "loss": 1.2604,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09777632842761798,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002999185973858809,
+      "loss": 1.2931,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09786402020199252,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0029991717567844226,
+      "loss": 1.2079,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09795171197636707,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002999157416669055,
+      "loss": 1.2036,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09803940375074162,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002999142953514014,
+      "loss": 1.2258,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09812709552511616,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002999128367320618,
+      "loss": 1.3019,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09821478729949071,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002999113658090198,
+      "loss": 1.23,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09830247907386525,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002999098825824095,
+      "loss": 1.2882,
+      "step": 1121
+    },
+    {
+      "epoch": 0.0983901708482398,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0029990838705236614,
+      "loss": 1.2154,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09847786262261433,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002999068792190262,
+      "loss": 1.2567,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09856555439698889,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029990535908252713,
+      "loss": 1.2004,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09865324617136342,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0029990382664300754,
+      "loss": 1.1912,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09874093794573797,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0029990228190060722,
+      "loss": 1.2991,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09882862972011253,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0029990072485546705,
+      "loss": 1.2323,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09891632149448706,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00299899155507729,
+      "loss": 1.342,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09900401326886162,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002998975738575362,
+      "loss": 1.2406,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09909170504323615,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002998959799050329,
+      "loss": 1.2276,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0991793968176107,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029989437365036454,
+      "loss": 1.2036,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09926708859198524,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029989275509367747,
+      "loss": 1.2688,
+      "step": 1132
+    },
+    {
+      "epoch": 0.0993547803663598,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0029989112423511933,
+      "loss": 1.2139,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09944247214073433,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0029988948107483896,
+      "loss": 1.2585,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09953016391510888,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0029988782561298608,
+      "loss": 1.2379,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09961785568948343,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002998861578497117,
+      "loss": 1.2192,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09970554746385797,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00299884477785168,
+      "loss": 1.1981,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09979323923823252,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0029988278541950805,
+      "loss": 1.1949,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09988093101260706,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002998810807528863,
+      "loss": 1.2511,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09996862278698161,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029987936378545813,
+      "loss": 1.2232,
+      "step": 1140
+    },
+    {
+      "epoch": 0.10005631456135615,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0029987763451738026,
+      "loss": 1.2326,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1001440063357307,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0029987589294881026,
+      "loss": 1.2861,
+      "step": 1142
+    },
+    {
+      "epoch": 0.10023169811010524,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0029987413907990703,
+      "loss": 1.209,
+      "step": 1143
+    },
+    {
+      "epoch": 0.10031938988447979,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0029987237291083046,
+      "loss": 1.2089,
+      "step": 1144
+    },
+    {
+      "epoch": 0.10040708165885433,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002998705944417417,
+      "loss": 1.241,
+      "step": 1145
+    },
+    {
+      "epoch": 0.10049477343322888,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002998688036728028,
+      "loss": 1.2162,
+      "step": 1146
+    },
+    {
+      "epoch": 0.10058246520760343,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029986700060417723,
+      "loss": 1.1938,
+      "step": 1147
+    },
+    {
+      "epoch": 0.10067015698197797,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029986518523602936,
+      "loss": 1.246,
+      "step": 1148
+    },
+    {
+      "epoch": 0.10075784875635252,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029986335756852474,
+      "loss": 1.231,
+      "step": 1149
+    },
+    {
+      "epoch": 0.10084554053072706,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002998615176018301,
+      "loss": 1.1912,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10093323230510161,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029985966533611313,
+      "loss": 1.2493,
+      "step": 1151
+    },
+    {
+      "epoch": 0.10102092407947615,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029985780077154286,
+      "loss": 1.1881,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1011086158538507,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002998559239082893,
+      "loss": 1.2643,
+      "step": 1153
+    },
+    {
+      "epoch": 0.10119630762822523,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002998540347465236,
+      "loss": 1.2051,
+      "step": 1154
+    },
+    {
+      "epoch": 0.10128399940259979,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029985213328641803,
+      "loss": 1.237,
+      "step": 1155
+    },
+    {
+      "epoch": 0.10137169117697434,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029985021952814604,
+      "loss": 1.1926,
+      "step": 1156
+    },
+    {
+      "epoch": 0.10145938295134888,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0029984829347188212,
+      "loss": 1.2545,
+      "step": 1157
+    },
+    {
+      "epoch": 0.10154707472572343,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00299846355117802,
+      "loss": 1.2598,
+      "step": 1158
+    },
+    {
+      "epoch": 0.10163476650009796,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0029984440446608235,
+      "loss": 1.2418,
+      "step": 1159
+    },
+    {
+      "epoch": 0.10172245827447252,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0029984244151690116,
+      "loss": 1.2688,
+      "step": 1160
+    },
+    {
+      "epoch": 0.10181015004884705,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0029984046627043737,
+      "loss": 1.2354,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1018978418232216,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0029983847872687114,
+      "loss": 1.2647,
+      "step": 1162
+    },
+    {
+      "epoch": 0.10198553359759614,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.002998364788863837,
+      "loss": 1.2424,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1020732253719707,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002998344667491575,
+      "loss": 1.2311,
+      "step": 1164
+    },
+    {
+      "epoch": 0.10216091714634525,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00299832442315376,
+      "loss": 1.2327,
+      "step": 1165
+    },
+    {
+      "epoch": 0.10224860892071978,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029983040558522384,
+      "loss": 1.3207,
+      "step": 1166
+    },
+    {
+      "epoch": 0.10233630069509433,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0029982835655888674,
+      "loss": 1.254,
+      "step": 1167
+    },
+    {
+      "epoch": 0.10242399246946887,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002998262952365516,
+      "loss": 1.1578,
+      "step": 1168
+    },
+    {
+      "epoch": 0.10251168424384342,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029982422161840636,
+      "loss": 1.2216,
+      "step": 1169
+    },
+    {
+      "epoch": 0.10259937601821796,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0029982213570464017,
+      "loss": 1.2352,
+      "step": 1170
+    },
+    {
+      "epoch": 0.10268706779259251,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0029982003749544324,
+      "loss": 1.2474,
+      "step": 1171
+    },
+    {
+      "epoch": 0.10277475956696705,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029981792699100692,
+      "loss": 1.2753,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1028624513413416,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0029981580419152372,
+      "loss": 1.1818,
+      "step": 1173
+    },
+    {
+      "epoch": 0.10295014311571615,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0029981366909718715,
+      "loss": 1.1834,
+      "step": 1174
+    },
+    {
+      "epoch": 0.10303783489009069,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0029981152170819206,
+      "loss": 1.2568,
+      "step": 1175
+    },
+    {
+      "epoch": 0.10312552666446524,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029980936202473416,
+      "loss": 1.2909,
+      "step": 1176
+    },
+    {
+      "epoch": 0.10321321843883978,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002998071900470104,
+      "loss": 1.2636,
+      "step": 1177
+    },
+    {
+      "epoch": 0.10330091021321433,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00299805005775219,
+      "loss": 1.1922,
+      "step": 1178
+    },
+    {
+      "epoch": 0.10338860198758887,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00299802809209559,
+      "loss": 1.207,
+      "step": 1179
+    },
+    {
+      "epoch": 0.10347629376196342,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002998006003502308,
+      "loss": 1.3056,
+      "step": 1180
+    },
+    {
+      "epoch": 0.10356398553633796,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0029979837919743586,
+      "loss": 1.2899,
+      "step": 1181
+    },
+    {
+      "epoch": 0.10365167731071251,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0029979614575137673,
+      "loss": 1.258,
+      "step": 1182
+    },
+    {
+      "epoch": 0.10373936908508706,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00299793900012257,
+      "loss": 1.2957,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1038270608594616,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002997916419802817,
+      "loss": 1.2868,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10391475263383615,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0029978937165565656,
+      "loss": 1.2358,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10400244440821069,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002997870890385886,
+      "loss": 1.2856,
+      "step": 1186
+    },
+    {
+      "epoch": 0.10409013618258524,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002997847941292861,
+      "loss": 1.2968,
+      "step": 1187
+    },
+    {
+      "epoch": 0.10417782795695978,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029978248692795837,
+      "loss": 1.205,
+      "step": 1188
+    },
+    {
+      "epoch": 0.10426551973133433,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029978016743481576,
+      "loss": 1.1977,
+      "step": 1189
+    },
+    {
+      "epoch": 0.10435321150570886,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029977783565006983,
+      "loss": 1.1755,
+      "step": 1190
+    },
+    {
+      "epoch": 0.10444090328008342,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029977549157393316,
+      "loss": 1.1375,
+      "step": 1191
+    },
+    {
+      "epoch": 0.10452859505445797,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029977313520661956,
+      "loss": 1.2817,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1046162868288325,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029977076654834406,
+      "loss": 1.2136,
+      "step": 1193
+    },
+    {
+      "epoch": 0.10470397860320706,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029976838559932248,
+      "loss": 1.2893,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1047916703775816,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0029976599235977206,
+      "loss": 1.2309,
+      "step": 1195
+    },
+    {
+      "epoch": 0.10487936215195615,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029976358682991104,
+      "loss": 1.1881,
+      "step": 1196
+    },
+    {
+      "epoch": 0.10496705392633068,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002997611690099587,
+      "loss": 1.2545,
+      "step": 1197
+    },
+    {
+      "epoch": 0.10505474570070523,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0029975873890013575,
+      "loss": 1.2107,
+      "step": 1198
+    },
+    {
+      "epoch": 0.10514243747507977,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0029975629650066367,
+      "loss": 1.2703,
+      "step": 1199
+    },
+    {
+      "epoch": 0.10523012924945432,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0029975384181176513,
+      "loss": 1.2156,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10531782102382888,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029975137483366416,
+      "loss": 1.2019,
+      "step": 1201
+    },
+    {
+      "epoch": 0.10540551279820341,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0029974889556658568,
+      "loss": 1.1964,
+      "step": 1202
+    },
+    {
+      "epoch": 0.10549320457257796,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0029974640401075575,
+      "loss": 1.2435,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1055808963469525,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002997439001664016,
+      "loss": 1.2428,
+      "step": 1204
+    },
+    {
+      "epoch": 0.10566858812132705,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002997413840337516,
+      "loss": 1.2146,
+      "step": 1205
+    },
+    {
+      "epoch": 0.10575627989570159,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029973885561303524,
+      "loss": 1.1744,
+      "step": 1206
+    },
+    {
+      "epoch": 0.10584397167007614,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0029973631490448306,
+      "loss": 1.2676,
+      "step": 1207
+    },
+    {
+      "epoch": 0.10593166344445068,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029973376190832674,
+      "loss": 1.2675,
+      "step": 1208
+    },
+    {
+      "epoch": 0.10601935521882523,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002997311966247992,
+      "loss": 1.2295,
+      "step": 1209
+    },
+    {
+      "epoch": 0.10610704699319978,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0029972861905413436,
+      "loss": 1.2739,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10619473876757432,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002997260291965672,
+      "loss": 1.1552,
+      "step": 1211
+    },
+    {
+      "epoch": 0.10628243054194887,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029972342705233404,
+      "loss": 1.1411,
+      "step": 1212
+    },
+    {
+      "epoch": 0.10637012231632341,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029972081262167197,
+      "loss": 1.2573,
+      "step": 1213
+    },
+    {
+      "epoch": 0.10645781409069796,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029971818590481974,
+      "loss": 1.2577,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1065455058650725,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029971554690201665,
+      "loss": 1.2148,
+      "step": 1215
+    },
+    {
+      "epoch": 0.10663319763944705,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029971289561350344,
+      "loss": 1.2585,
+      "step": 1216
+    },
+    {
+      "epoch": 0.10672088941382159,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029971023203952192,
+      "loss": 1.2893,
+      "step": 1217
+    },
+    {
+      "epoch": 0.10680858118819614,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00299707556180315,
+      "loss": 1.2542,
+      "step": 1218
+    },
+    {
+      "epoch": 0.10689627296257069,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0029970486803612673,
+      "loss": 1.2508,
+      "step": 1219
+    },
+    {
+      "epoch": 0.10698396473694523,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002997021676072022,
+      "loss": 1.2215,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10707165651131978,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002996994548937877,
+      "loss": 1.2594,
+      "step": 1221
+    },
+    {
+      "epoch": 0.10715934828569432,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002996967298961307,
+      "loss": 1.2044,
+      "step": 1222
+    },
+    {
+      "epoch": 0.10724704006006887,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0029969399261447955,
+      "loss": 1.1786,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1073347318344434,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002996912430490841,
+      "loss": 1.2013,
+      "step": 1224
+    },
+    {
+      "epoch": 0.10742242360881796,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.002996884812001949,
+      "loss": 1.2407,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1075101153831925,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002996857070680639,
+      "loss": 1.2203,
+      "step": 1226
+    },
+    {
+      "epoch": 0.10759780715756705,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002996829206529442,
+      "loss": 1.2675,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1076854989319416,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0029968012195508978,
+      "loss": 1.2556,
+      "step": 1228
+    },
+    {
+      "epoch": 0.10777319070631614,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029967731097475586,
+      "loss": 1.2905,
+      "step": 1229
+    },
+    {
+      "epoch": 0.10786088248069069,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002996744877121989,
+      "loss": 1.2587,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10794857425506522,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029967165216767634,
+      "loss": 1.2594,
+      "step": 1231
+    },
+    {
+      "epoch": 0.10803626602943978,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029966880434144673,
+      "loss": 1.1906,
+      "step": 1232
+    },
+    {
+      "epoch": 0.10812395780381431,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0029966594423376973,
+      "loss": 1.213,
+      "step": 1233
+    },
+    {
+      "epoch": 0.10821164957818886,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002996630718449064,
+      "loss": 1.2473,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1082993413525634,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0029966018717511845,
+      "loss": 1.2035,
+      "step": 1235
+    },
+    {
+      "epoch": 0.10838703312693795,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002996572902246691,
+      "loss": 1.2043,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1084747249013125,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029965438099382245,
+      "loss": 1.2142,
+      "step": 1237
+    },
+    {
+      "epoch": 0.10856241667568704,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029965145948284387,
+      "loss": 1.2318,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1086501084500616,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029964852569199984,
+      "loss": 1.1786,
+      "step": 1239
+    },
+    {
+      "epoch": 0.10873780022443613,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002996455796215578,
+      "loss": 1.2101,
+      "step": 1240
+    },
+    {
+      "epoch": 0.10882549199881068,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0029964262127178654,
+      "loss": 1.2245,
+      "step": 1241
+    },
+    {
+      "epoch": 0.10891318377318522,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0029963965064295573,
+      "loss": 1.1896,
+      "step": 1242
+    },
+    {
+      "epoch": 0.10900087554755977,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002996366677353364,
+      "loss": 1.2298,
+      "step": 1243
+    },
+    {
+      "epoch": 0.10908856732193431,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029963367254920055,
+      "loss": 1.2506,
+      "step": 1244
+    },
+    {
+      "epoch": 0.10917625909630886,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002996306650848213,
+      "loss": 1.2451,
+      "step": 1245
+    },
+    {
+      "epoch": 0.10926395087068341,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0029962764534247287,
+      "loss": 1.2024,
+      "step": 1246
+    },
+    {
+      "epoch": 0.10935164264505795,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029962461332243085,
+      "loss": 1.2491,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1094393344194325,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0029962156902497154,
+      "loss": 1.1692,
+      "step": 1248
+    },
+    {
+      "epoch": 0.10952702619380704,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002996185124503727,
+      "loss": 1.1808,
+      "step": 1249
+    },
+    {
+      "epoch": 0.10961471796818159,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029961544359891302,
+      "loss": 1.2391,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10970240974255613,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0029961236247087234,
+      "loss": 1.2164,
+      "step": 1251
+    },
+    {
+      "epoch": 0.10979010151693068,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0029960926906653185,
+      "loss": 1.2398,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10987779329130522,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029960616338617335,
+      "loss": 1.227,
+      "step": 1253
+    },
+    {
+      "epoch": 0.10996548506567977,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029960304543008034,
+      "loss": 1.1366,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11005317684005432,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00299599915198537,
+      "loss": 1.2744,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11014086861442886,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002995967726918289,
+      "loss": 1.276,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11022856038880341,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002995936179102426,
+      "loss": 1.2089,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11031625216317795,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029959045085406573,
+      "loss": 1.289,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1104039439375525,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002995872715235873,
+      "loss": 1.1645,
+      "step": 1259
+    },
+    {
+      "epoch": 0.11049163571192704,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0029958407991909704,
+      "loss": 1.2306,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11057932748630159,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029958087604088617,
+      "loss": 1.2353,
+      "step": 1261
+    },
+    {
+      "epoch": 0.11066701926067612,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002995776598892468,
+      "loss": 1.2188,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11075471103505068,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0029957443146447224,
+      "loss": 1.2458,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11084240280942523,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029957119076685695,
+      "loss": 1.2121,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11093009458379977,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002995679377966965,
+      "loss": 1.2579,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11101778635817432,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029956467255428743,
+      "loss": 1.2074,
+      "step": 1266
+    },
+    {
+      "epoch": 0.11110547813254885,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029956139503992765,
+      "loss": 1.2233,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1111931699069234,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029955810525391603,
+      "loss": 1.1884,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11128086168129794,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002995548031965526,
+      "loss": 1.2408,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1113685534556725,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0029955148886813836,
+      "loss": 1.2718,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11145624523004703,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029954816226897573,
+      "loss": 1.2472,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11154393700442158,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029954482339936803,
+      "loss": 1.2382,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11163162877879614,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002995414722596198,
+      "loss": 1.2168,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11171932055317067,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0029953810885003655,
+      "loss": 1.2388,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11180701232754522,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029953473317092514,
+      "loss": 1.2328,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11189470410191976,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029953134522259337,
+      "loss": 1.2185,
+      "step": 1276
+    },
+    {
+      "epoch": 0.11198239587629431,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029952794500535014,
+      "loss": 1.2673,
+      "step": 1277
+    },
+    {
+      "epoch": 0.11207008765066885,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029952453251950563,
+      "loss": 1.1878,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1121577794250434,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0029952110776537105,
+      "loss": 1.1887,
+      "step": 1279
+    },
+    {
+      "epoch": 0.11224547119941794,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002995176707432587,
+      "loss": 1.2541,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11233316297379249,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0029951422145348206,
+      "loss": 1.1988,
+      "step": 1281
+    },
+    {
+      "epoch": 0.11242085474816704,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0029951075989635566,
+      "loss": 1.1977,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11250854652254158,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029950728607219513,
+      "loss": 1.2858,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11259623829691613,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0029950379998131744,
+      "loss": 1.2136,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11268393007129067,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0029950030162404035,
+      "loss": 1.1722,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11277162184566522,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029949679100068297,
+      "loss": 1.2412,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11285931362003976,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002994932681115655,
+      "loss": 1.2478,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11294700539441431,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0029948973295700907,
+      "loss": 1.2344,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11303469716878885,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002994861855373362,
+      "loss": 1.3181,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1131223889431634,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002994826258528705,
+      "loss": 1.1635,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11321008071753795,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0029947905390393637,
+      "loss": 1.1966,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11329777249191249,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002994754696908597,
+      "loss": 1.2595,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11338546426628704,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0029947187321396735,
+      "loss": 1.2539,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11347315604066158,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002994682644735873,
+      "loss": 1.2618,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11356084781503613,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029946464347004867,
+      "loss": 1.2937,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11364853958941067,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002994610102036817,
+      "loss": 1.2365,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11373623136378522,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029945736467481767,
+      "loss": 1.2025,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11382392313815975,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002994537068837891,
+      "loss": 1.2138,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1139116149125343,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002994500368309295,
+      "loss": 1.2215,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11399930668690886,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029944635451657365,
+      "loss": 1.2056,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1140869984612834,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002994426599410574,
+      "loss": 1.2978,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11417469023565795,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0029943895310471755,
+      "loss": 1.2251,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11426238201003248,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029943523400789225,
+      "loss": 1.2908,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11435007378440704,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0029943150265092067,
+      "loss": 1.2399,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11443776555878157,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0029942775903414307,
+      "loss": 1.2414,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11452545733315612,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0029942400315790085,
+      "loss": 1.298,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11461314910753066,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0029942023502253657,
+      "loss": 1.2238,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11470084088190521,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029941645462839388,
+      "loss": 1.2058,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11478853265627977,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0029941266197581755,
+      "loss": 1.2297,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1148762244306543,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0029940885706515336,
+      "loss": 1.261,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11496391620502885,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002994050398967484,
+      "loss": 1.2694,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11505160797940339,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002994012104709508,
+      "loss": 1.2114,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11513929975377794,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002993973687881097,
+      "loss": 1.3171,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11522699152815248,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0029939351484857555,
+      "loss": 1.179,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11531468330252703,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0029938964865269977,
+      "loss": 1.1591,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11540237507690157,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029938577020083503,
+      "loss": 1.2616,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11549006685127612,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029938187949333484,
+      "loss": 1.2064,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11557775862565067,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029937797653055423,
+      "loss": 1.2495,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11566545040002521,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00299374061312849,
+      "loss": 1.1988,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11575314217439976,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002993701338405763,
+      "loss": 1.2312,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1158408339487743,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002993661941140943,
+      "loss": 1.2405,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11592852572314885,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002993622421337622,
+      "loss": 1.3058,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11601621749752339,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0029935827789994048,
+      "loss": 1.2511,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11610390927189794,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002993543014129907,
+      "loss": 1.21,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11619160104627248,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0029935031267327543,
+      "loss": 1.275,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11627929282064703,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029934631168115847,
+      "loss": 1.1929,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11636698459502158,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002993422984370047,
+      "loss": 1.2059,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11645467636939612,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002993382729411801,
+      "loss": 1.2447,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11654236814377067,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029933423519405184,
+      "loss": 1.2291,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1166300599181452,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002993301851959881,
+      "loss": 1.1888,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11671775169251976,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002993261229473582,
+      "loss": 1.223,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1168054434668943,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029932204844853273,
+      "loss": 1.2749,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11689313524126885,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029931796169988313,
+      "loss": 1.2093,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11698082701564338,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002993138627017822,
+      "loss": 1.2348,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11706851879001794,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002993097514546037,
+      "loss": 1.2177,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11715621056439249,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029930562795872258,
+      "loss": 1.1629,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11724390233876703,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.002993014922145149,
+      "loss": 1.2504,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11733159411314158,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002992973442223578,
+      "loss": 1.2193,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11741928588751611,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0029929318398262965,
+      "loss": 1.2525,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11750697766189067,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002992890114957098,
+      "loss": 1.2871,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1175946694362652,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029928482676197872,
+      "loss": 1.1741,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11768236121063975,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002992806297818181,
+      "loss": 1.276,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11777005298501429,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002992764205556107,
+      "loss": 1.2048,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11785774475938884,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0029927219908374037,
+      "loss": 1.2242,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1179454365337634,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.002992679653665921,
+      "loss": 1.2332,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11803312830813793,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0029926371940455204,
+      "loss": 1.2667,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11812082008251248,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0029925946119800737,
+      "loss": 1.3023,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11820851185688702,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0029925519074734635,
+      "loss": 1.2397,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11829620363126157,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.002992509080529586,
+      "loss": 1.2721,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11838389540563611,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002992466131152345,
+      "loss": 1.3193,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11847158718001066,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029924230593456596,
+      "loss": 1.1955,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1185592789543852,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002992379865113456,
+      "loss": 1.2117,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11864697072875975,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029923365484596743,
+      "loss": 1.2117,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11873466250313429,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029922931093882645,
+      "loss": 1.2643,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11882235427750884,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002992249547903188,
+      "loss": 1.2068,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11891004605188339,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002992205864008418,
+      "loss": 1.2446,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11899773782625793,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002992162057707938,
+      "loss": 1.2532,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11908542960063248,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0029921181290057425,
+      "loss": 1.2755,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11917312137500702,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0029920740779058393,
+      "loss": 1.2407,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11926081314938157,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029920299044122445,
+      "loss": 1.2351,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11934850492375611,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029919856085289863,
+      "loss": 1.2408,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11943619669813066,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029919411902601055,
+      "loss": 1.2901,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1195238884725052,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002991896649609652,
+      "loss": 1.2612,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11961158024687975,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0029918519865816885,
+      "loss": 1.263,
+      "step": 1364
+    },
+    {
+      "epoch": 0.1196992720212543,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029918072011802872,
+      "loss": 1.1916,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11978696379562884,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002991762293409534,
+      "loss": 1.2163,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11987465557000339,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002991717263273523,
+      "loss": 1.1645,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11996234734437793,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029916721107763606,
+      "loss": 1.1664,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12005003911875248,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0029916268359221655,
+      "loss": 1.295,
+      "step": 1369
+    },
+    {
+      "epoch": 0.12013773089312701,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002991581438715066,
+      "loss": 1.1667,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12022542266750157,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029915359191592036,
+      "loss": 1.1757,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1203131144418761,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002991490277258728,
+      "loss": 1.208,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12040080621625066,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0029914445130178016,
+      "loss": 1.2035,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1204884979906252,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002991398626440599,
+      "loss": 1.225,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12057618976499974,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029913526175313052,
+      "loss": 1.2107,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1206638815393743,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0029913064862941144,
+      "loss": 1.2726,
+      "step": 1376
+    },
+    {
+      "epoch": 0.12075157331374883,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002991260232733235,
+      "loss": 1.214,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12083926508812338,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029912138568528846,
+      "loss": 1.2616,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12092695686249792,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002991167358657293,
+      "loss": 1.2894,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12101464863687247,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0029911207381507,
+      "loss": 1.1904,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12110234041124701,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029910739953373584,
+      "loss": 1.1905,
+      "step": 1381
+    },
+    {
+      "epoch": 0.12119003218562156,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029910271302215304,
+      "loss": 1.2156,
+      "step": 1382
+    },
+    {
+      "epoch": 0.12127772395999611,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029909801428074896,
+      "loss": 1.2372,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12136541573437065,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0029909330330995213,
+      "loss": 1.2494,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1214531075087452,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029908858011019226,
+      "loss": 1.2056,
+      "step": 1385
+    },
+    {
+      "epoch": 0.12154079928311974,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002990838446819,
+      "loss": 1.266,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12162849105749429,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002990790970255072,
+      "loss": 1.2105,
+      "step": 1387
+    },
+    {
+      "epoch": 0.12171618283186883,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0029907433714144696,
+      "loss": 1.1864,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12180387460624338,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0029906956503015325,
+      "loss": 1.2204,
+      "step": 1389
+    },
+    {
+      "epoch": 0.12189156638061792,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002990647806920613,
+      "loss": 1.2312,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12197925815499247,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029905998412760744,
+      "loss": 1.2329,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12206694992936702,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002990551753372291,
+      "loss": 1.1962,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12215464170374156,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0029905035432136484,
+      "loss": 1.1929,
+      "step": 1393
+    },
+    {
+      "epoch": 0.12224233347811611,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029904552108045426,
+      "loss": 1.2269,
+      "step": 1394
+    },
+    {
+      "epoch": 0.12233002525249065,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029904067561493824,
+      "loss": 1.2734,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1224177170268652,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029903581792525866,
+      "loss": 1.1868,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12250540880123974,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029903094801185847,
+      "loss": 1.2306,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12259310057561429,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002990260658751818,
+      "loss": 1.2046,
+      "step": 1398
+    },
+    {
+      "epoch": 0.12268079234998883,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029902117151567394,
+      "loss": 1.2347,
+      "step": 1399
+    },
+    {
+      "epoch": 0.12276848412436338,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002990162649337812,
+      "loss": 1.2616,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12285617589873793,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00299011346129951,
+      "loss": 1.2644,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12294386767311247,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00299006415104632,
+      "loss": 1.2854,
+      "step": 1402
+    },
+    {
+      "epoch": 0.12303155944748702,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00299001471858274,
+      "loss": 1.2626,
+      "step": 1403
+    },
+    {
+      "epoch": 0.12311925122186156,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002989965163913276,
+      "loss": 1.186,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12320694299623611,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002989915487042448,
+      "loss": 1.1977,
+      "step": 1405
+    },
+    {
+      "epoch": 0.12329463477061064,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002989865687974787,
+      "loss": 1.2557,
+      "step": 1406
+    },
+    {
+      "epoch": 0.1233823265449852,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029898157667148334,
+      "loss": 1.2345,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12347001831935973,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0029897657232671408,
+      "loss": 1.2252,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12355771009373429,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002989715557636273,
+      "loss": 1.2605,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12364540186810884,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029896652698268057,
+      "loss": 1.2567,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12373309364248337,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029896148598433227,
+      "loss": 1.2003,
+      "step": 1411
+    },
+    {
+      "epoch": 0.12382078541685793,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029895643276904235,
+      "loss": 1.2087,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12390847719123246,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0029895136733727157,
+      "loss": 1.1963,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12399616896560701,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029894628968948184,
+      "loss": 1.2341,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12408386073998155,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002989411998261363,
+      "loss": 1.2918,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1241715525143561,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0029893609774769908,
+      "loss": 1.2334,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12425924428873064,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029893098345463547,
+      "loss": 1.2939,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12434693606310519,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0029892585694741196,
+      "loss": 1.2374,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12443462783747974,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00298920718226496,
+      "loss": 1.2522,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12452231961185428,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029891556729235626,
+      "loss": 1.2733,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12461001138622883,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002989104041454625,
+      "loss": 1.2333,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12469770316060337,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002989052287862856,
+      "loss": 1.2635,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12478539493497792,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002989000412152975,
+      "loss": 1.2516,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12487308670935246,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029889484143297124,
+      "loss": 1.2372,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12496077848372701,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0029888962943978113,
+      "loss": 1.2064,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12504847025810156,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002988844052362024,
+      "loss": 1.2496,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1251361620324761,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.002988791688227116,
+      "loss": 1.2487,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12522385380685064,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002988739201997862,
+      "loss": 1.2427,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1253115455812252,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.002988686593679048,
+      "loss": 1.152,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12539923735559974,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029886338632754733,
+      "loss": 1.2219,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12548692912997428,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0029885810107919456,
+      "loss": 1.2429,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12557462090434882,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029885280362332853,
+      "loss": 1.2552,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12566231267872338,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0029884749396043237,
+      "loss": 1.2547,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12575000445309792,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0029884217209099023,
+      "loss": 1.1672,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12583769622747246,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029883683801548754,
+      "loss": 1.2381,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12592538800184702,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029883149173441075,
+      "loss": 1.2191,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12601307977622156,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029882613324824737,
+      "loss": 1.2202,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1261007715505961,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002988207625574861,
+      "loss": 1.2144,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12618846332497063,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029881537966261673,
+      "loss": 1.2211,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1262761550993452,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029880998456413023,
+      "loss": 1.2397,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12636384687371974,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002988045772625185,
+      "loss": 1.1851,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12645153864809427,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0029879915775827483,
+      "loss": 1.1996,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1265392304224688,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002987937260518933,
+      "loss": 1.2325,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12662692219684338,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0029878828214386934,
+      "loss": 1.2232,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12671461397121792,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0029878282603469945,
+      "loss": 1.2375,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12680230574559245,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029877735772488117,
+      "loss": 1.2673,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12688999751996702,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029877187721491323,
+      "loss": 1.2063,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12697768929434156,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002987663845052954,
+      "loss": 1.213,
+      "step": 1448
+    },
+    {
+      "epoch": 0.1270653810687161,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002987608795965286,
+      "loss": 1.247,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12715307284309063,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002987553624891149,
+      "loss": 1.2318,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1272407646174652,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029874983318355745,
+      "loss": 1.2509,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12732845639183973,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0029874429168036047,
+      "loss": 1.2573,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12741614816621427,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029873873798002934,
+      "loss": 1.1857,
+      "step": 1453
+    },
+    {
+      "epoch": 0.1275038399405888,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0029873317208307056,
+      "loss": 1.2676,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12759153171496337,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002987275939899917,
+      "loss": 1.2704,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1276792234893379,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002987220037013015,
+      "loss": 1.1953,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12776691526371245,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002987164012175098,
+      "loss": 1.2961,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12785460703808701,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029871078653912744,
+      "loss": 1.2668,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12794229881246155,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0029870515966666654,
+      "loss": 1.2222,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1280299905868361,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002986995206006402,
+      "loss": 1.2353,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12811768236121063,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029869386934156276,
+      "loss": 1.2168,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1282053741355852,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029868820588994957,
+      "loss": 1.1639,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12829306590995973,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029868253024631715,
+      "loss": 1.2447,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12838075768433427,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00298676842411183,
+      "loss": 1.2335,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12846844945870883,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00298671142385066,
+      "loss": 1.222,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12855614123308337,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029866543016848577,
+      "loss": 1.2759,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1286438330074579,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002986597057619634,
+      "loss": 1.1909,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12873152478183245,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0029865396916602094,
+      "loss": 1.2919,
+      "step": 1468
+    },
+    {
+      "epoch": 0.128819216556207,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002986482203811815,
+      "loss": 1.2541,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12890690833058155,
+      "grad_norm": 0.125,
+      "learning_rate": 0.002986424594079694,
+      "loss": 1.2356,
+      "step": 1470
+    },
+    {
+      "epoch": 0.1289946001049561,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0029863668624690995,
+      "loss": 1.2285,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12908229187933062,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029863090089852972,
+      "loss": 1.2105,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1291699836537052,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0029862510336335635,
+      "loss": 1.2278,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12925767542807973,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002986192936419184,
+      "loss": 1.2569,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12934536720245426,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029861347173474584,
+      "loss": 1.2992,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12943305897682883,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0029860763764236963,
+      "loss": 1.2915,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12952075075120337,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029860179136532174,
+      "loss": 1.2971,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1296084425255779,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0029859593290413535,
+      "loss": 1.2245,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12969613429995244,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002985900622593448,
+      "loss": 1.1996,
+      "step": 1479
+    },
+    {
+      "epoch": 0.129783826074327,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029858417943148534,
+      "loss": 1.2459,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12987151784870155,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0029857828442109366,
+      "loss": 1.1602,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12995920962307608,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0029857237722870724,
+      "loss": 1.2804,
+      "step": 1482
+    },
+    {
+      "epoch": 0.13004690139745065,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002985664578548648,
+      "loss": 1.2021,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13013459317182519,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0029856052630010625,
+      "loss": 1.2046,
+      "step": 1484
+    },
+    {
+      "epoch": 0.13022228494619972,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002985545825649725,
+      "loss": 1.2553,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13030997672057426,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029854862665000554,
+      "loss": 1.2347,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13039766849494883,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002985426585557486,
+      "loss": 1.1759,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13048536026932336,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029853667828274593,
+      "loss": 1.2221,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1305730520436979,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0029853068583154292,
+      "loss": 1.2229,
+      "step": 1489
+    },
+    {
+      "epoch": 0.13066074381807244,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029852468120268607,
+      "loss": 1.2474,
+      "step": 1490
+    },
+    {
+      "epoch": 0.130748435592447,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00298518664396723,
+      "loss": 1.276,
+      "step": 1491
+    },
+    {
+      "epoch": 0.13083612736682154,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029851263541420246,
+      "loss": 1.259,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13092381914119608,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0029850659425567413,
+      "loss": 1.2486,
+      "step": 1493
+    },
+    {
+      "epoch": 0.13101151091557064,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029850054092168915,
+      "loss": 1.238,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13109920268994518,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002984944754127994,
+      "loss": 1.233,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13118689446431972,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0029848839772955815,
+      "loss": 1.1471,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13127458623869426,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002984823078725196,
+      "loss": 1.2504,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13136227801306882,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029847620584223917,
+      "loss": 1.1724,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13144996978744336,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0029847009163927333,
+      "loss": 1.2235,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1315376615618179,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029846396526417964,
+      "loss": 1.207,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1315376615618179,
+      "eval_loss": 1.2418609857559204,
+      "eval_runtime": 429.7895,
+      "eval_samples_per_second": 33.614,
+      "eval_steps_per_second": 8.404,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13162535333619246,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002984578267175169,
+      "loss": 1.21,
+      "step": 1501
+    },
+    {
+      "epoch": 0.131713045110567,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0029845167599984487,
+      "loss": 1.2382,
+      "step": 1502
+    },
+    {
+      "epoch": 0.13180073688494154,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002984455131117245,
+      "loss": 1.2137,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13188842865931608,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029843933805371783,
+      "loss": 1.2029,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13197612043369064,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00298433150826388,
+      "loss": 1.2327,
+      "step": 1505
+    },
+    {
+      "epoch": 0.13206381220806518,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029842695143029925,
+      "loss": 1.2136,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13215150398243972,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029842073986601696,
+      "loss": 1.2261,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13223919575681425,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0029841451613410765,
+      "loss": 1.296,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13232688753118882,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029840828023513888,
+      "loss": 1.2243,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13241457930556336,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0029840203216967933,
+      "loss": 1.1826,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1325022710799379,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002983957719382988,
+      "loss": 1.2358,
+      "step": 1511
+    },
+    {
+      "epoch": 0.13258996285431246,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0029838949954156826,
+      "loss": 1.255,
+      "step": 1512
+    },
+    {
+      "epoch": 0.132677654628687,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002983832149800597,
+      "loss": 1.2429,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13276534640306153,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0029837691825434624,
+      "loss": 1.3166,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13285303817743607,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0029837060936500214,
+      "loss": 1.238,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13294072995181064,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.002983642883126028,
+      "loss": 1.246,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13302842172618518,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002983579550977246,
+      "loss": 1.2287,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1331161135005597,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002983516097209452,
+      "loss": 1.2517,
+      "step": 1518
+    },
+    {
+      "epoch": 0.13320380527493428,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002983452521828432,
+      "loss": 1.2551,
+      "step": 1519
+    },
+    {
+      "epoch": 0.13329149704930882,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029833888248399845,
+      "loss": 1.2386,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13337918882368335,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0029833250062499188,
+      "loss": 1.2619,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1334668805980579,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029832610660640536,
+      "loss": 1.3181,
+      "step": 1522
+    },
+    {
+      "epoch": 0.13355457237243246,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029831970042882216,
+      "loss": 1.1977,
+      "step": 1523
+    },
+    {
+      "epoch": 0.133642264146807,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029831328209282645,
+      "loss": 1.2195,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13372995592118153,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029830685159900347,
+      "loss": 1.2581,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13381764769555607,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029830040894793983,
+      "loss": 1.2582,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13390533946993063,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0029829395414022303,
+      "loss": 1.2592,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13399303124430517,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029828748717644167,
+      "loss": 1.2546,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1340807230186797,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029828100805718554,
+      "loss": 1.2101,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13416841479305427,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029827451678304555,
+      "loss": 1.2072,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1342561065674288,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002982680133546137,
+      "loss": 1.27,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13434379834180335,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0029826149777248305,
+      "loss": 1.1899,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1344314901161779,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002982549700372478,
+      "loss": 1.2853,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13451918189055245,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0029824843014950326,
+      "loss": 1.1747,
+      "step": 1534
+    },
+    {
+      "epoch": 0.134606873664927,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002982418781098459,
+      "loss": 1.2496,
+      "step": 1535
+    },
+    {
+      "epoch": 0.13469456543930153,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029823531391887322,
+      "loss": 1.2428,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1347822572136761,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0029822873757718387,
+      "loss": 1.2669,
+      "step": 1537
+    },
+    {
+      "epoch": 0.13486994898805063,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0029822214908537758,
+      "loss": 1.2088,
+      "step": 1538
+    },
+    {
+      "epoch": 0.13495764076242517,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0029821554844405517,
+      "loss": 1.2522,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1350453325367997,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002982089356538187,
+      "loss": 1.2662,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13513302431117427,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.002982023107152711,
+      "loss": 1.2763,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1352207160855488,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029819567362901664,
+      "loss": 1.2297,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13530840785992335,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002981890243956606,
+      "loss": 1.2362,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13539609963429788,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002981823630158094,
+      "loss": 1.244,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13548379140867245,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0029817568949007047,
+      "loss": 1.2439,
+      "step": 1545
+    },
+    {
+      "epoch": 0.135571483183047,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002981690038190524,
+      "loss": 1.229,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13565917495742152,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0029816230600336504,
+      "loss": 1.2143,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1357468667317961,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0029815559604361905,
+      "loss": 1.2115,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13583455850617063,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002981488739404265,
+      "loss": 1.2156,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13592225028054516,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0029814213969440034,
+      "loss": 1.2304,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1360099420549197,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029813539330615477,
+      "loss": 1.264,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13609763382929427,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0029812863477630502,
+      "loss": 1.2226,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1361853256036688,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002981218641054674,
+      "loss": 1.1539,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13627301737804334,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002981150812942595,
+      "loss": 1.1669,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1363607091524179,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0029810828634329973,
+      "loss": 1.2288,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13644840092679245,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002981014792532079,
+      "loss": 1.2686,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13653609270116698,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029809466002460477,
+      "loss": 1.2389,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13662378447554152,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0029808782865811223,
+      "loss": 1.2592,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1367114762499161,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002980809851543533,
+      "loss": 1.3417,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13679916802429062,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0029807412951395203,
+      "loss": 1.2297,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13688685979866516,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0029806726173753372,
+      "loss": 1.2832,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1369745515730397,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0029806038182572463,
+      "loss": 1.2379,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13706224334741426,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0029805348977915216,
+      "loss": 1.258,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1371499351217888,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00298046585598445,
+      "loss": 1.2212,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13723762689616334,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0029803966928423262,
+      "loss": 1.2305,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1373253186705379,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029803274083714582,
+      "loss": 1.2141,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13741301044491244,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029802580025781655,
+      "loss": 1.2364,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13750070221928698,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029801884754687767,
+      "loss": 1.2939,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13758839399366152,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0029801188270496327,
+      "loss": 1.2379,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13767608576803608,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002980049057327086,
+      "loss": 1.1832,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13776377754241062,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002979979166307498,
+      "loss": 1.2445,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13785146931678516,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002979909153997243,
+      "loss": 1.233,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13793916109115972,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002979839020402707,
+      "loss": 1.1687,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13802685286553426,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029797687655302853,
+      "loss": 1.2213,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1381145446399088,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002979698389386385,
+      "loss": 1.3031,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13820223641428334,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002979627891977424,
+      "loss": 1.3023,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1382899281886579,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0029795572733098317,
+      "loss": 1.1969,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13837761996303244,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002979486533390048,
+      "loss": 1.2537,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13846531173740698,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002979415672224525,
+      "loss": 1.2111,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13855300351178151,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029793446898197244,
+      "loss": 1.2083,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13864069528615608,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00297927358618212,
+      "loss": 1.2314,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13872838706053062,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029792023613181957,
+      "loss": 1.1932,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13881607883490515,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029791310152344473,
+      "loss": 1.1545,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13890377060927972,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029790595479373813,
+      "loss": 1.2353,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13899146238365426,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0029789879594335164,
+      "loss": 1.2443,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1390791541580288,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0029789162497293794,
+      "loss": 1.2265,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13916684593240333,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002978844418831511,
+      "loss": 1.2414,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1392545377067779,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029787724667464624,
+      "loss": 1.2488,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13934222948115244,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002978700393480795,
+      "loss": 1.1473,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13942992125552697,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029786281990410823,
+      "loss": 1.3154,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13951761302990154,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029785558834339067,
+      "loss": 1.2534,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13960530480427608,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002978483446665865,
+      "loss": 1.201,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1396929965786506,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029784108887435617,
+      "loss": 1.2566,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13978068835302515,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0029783382096736145,
+      "loss": 1.2412,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13986838012739972,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0029782654094626525,
+      "loss": 1.2333,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13995607190177425,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0029781924881173137,
+      "loss": 1.237,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1400437636761488,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0029781194456442485,
+      "loss": 1.2805,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14013145545052333,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029780462820501188,
+      "loss": 1.2017,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1402191472248979,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029779729973415958,
+      "loss": 1.2132,
+      "step": 1599
+    },
+    {
+      "epoch": 0.14030683899927243,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0029778995915253643,
+      "loss": 1.2458,
+      "step": 1600
+    },
+    {
+      "epoch": 0.14039453077364697,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002977826064608118,
+      "loss": 1.1536,
+      "step": 1601
+    },
+    {
+      "epoch": 0.14048222254802153,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002977752416596562,
+      "loss": 1.1708,
+      "step": 1602
+    },
+    {
+      "epoch": 0.14056991432239607,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002977678647497413,
+      "loss": 1.2385,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1406576060967706,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0029776047573173993,
+      "loss": 1.2509,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14074529787114515,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029775307460632584,
+      "loss": 1.2119,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1408329896455197,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002977456613741741,
+      "loss": 1.2478,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14092068141989425,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002977382360359607,
+      "loss": 1.2545,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1410083731942688,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002977307985923628,
+      "loss": 1.2298,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14109606496864335,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0029772334904405876,
+      "loss": 1.1832,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1411837567430179,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002977158873917279,
+      "loss": 1.2469,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14127144851739243,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029770841363605067,
+      "loss": 1.1923,
+      "step": 1611
+    },
+    {
+      "epoch": 0.14135914029176697,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0029770092777770874,
+      "loss": 1.2275,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14144683206614153,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002976934298173848,
+      "loss": 1.2375,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14153452384051607,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0029768591975576253,
+      "loss": 1.2115,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1416222156148906,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029767839759352694,
+      "loss": 1.2172,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14170990738926514,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029767086333136405,
+      "loss": 1.2543,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1417975991636397,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029766331696996083,
+      "loss": 1.2789,
+      "step": 1617
+    },
+    {
+      "epoch": 0.14188529093801425,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002976557585100056,
+      "loss": 1.277,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14197298271238878,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029764818795218763,
+      "loss": 1.2974,
+      "step": 1619
+    },
+    {
+      "epoch": 0.14206067448676335,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002976406052971974,
+      "loss": 1.263,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1421483662611379,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029763301054572633,
+      "loss": 1.2491,
+      "step": 1621
+    },
+    {
+      "epoch": 0.14223605803551242,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0029762540369846708,
+      "loss": 1.1965,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14232374980988696,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002976177847561134,
+      "loss": 1.1656,
+      "step": 1623
+    },
+    {
+      "epoch": 0.14241144158426153,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029761015371936013,
+      "loss": 1.2432,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14249913335863607,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029760251058890312,
+      "loss": 1.2994,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1425868251330106,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002975948553654395,
+      "loss": 1.3091,
+      "step": 1626
+    },
+    {
+      "epoch": 0.14267451690738517,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002975871880496673,
+      "loss": 1.2049,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1427622086817597,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002975795086422859,
+      "loss": 1.2023,
+      "step": 1628
+    },
+    {
+      "epoch": 0.14284990045613424,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002975718171439955,
+      "loss": 1.2296,
+      "step": 1629
+    },
+    {
+      "epoch": 0.14293759223050878,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002975641135554977,
+      "loss": 1.2299,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14302528400488335,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029755639787749488,
+      "loss": 1.2367,
+      "step": 1631
+    },
+    {
+      "epoch": 0.14311297577925788,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0029754867011069076,
+      "loss": 1.2048,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14320066755363242,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029754093025579015,
+      "loss": 1.2234,
+      "step": 1633
+    },
+    {
+      "epoch": 0.14328835932800696,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002975331783134988,
+      "loss": 1.2113,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14337605110238152,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0029752541428452375,
+      "loss": 1.1891,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14346374287675606,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0029751763816957305,
+      "loss": 1.2037,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1435514346511306,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002975098499693558,
+      "loss": 1.2415,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14363912642550516,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029750204968458233,
+      "loss": 1.2437,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1437268181998797,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0029749423731596394,
+      "loss": 1.1978,
+      "step": 1639
+    },
+    {
+      "epoch": 0.14381450997425424,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.002974864128642132,
+      "loss": 1.2061,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14390220174862878,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002974785763300436,
+      "loss": 1.2559,
+      "step": 1641
+    },
+    {
+      "epoch": 0.14398989352300334,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002974707277141698,
+      "loss": 1.2282,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14407758529737788,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0029746286701730763,
+      "loss": 1.2323,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14416527707175242,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029745499424017395,
+      "loss": 1.2645,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14425296884612698,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002974471093834867,
+      "loss": 1.2442,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14434066062050152,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00297439212447965,
+      "loss": 1.2318,
+      "step": 1646
+    },
+    {
+      "epoch": 0.14442835239487606,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00297431303434329,
+      "loss": 1.2473,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1445160441692506,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002974233823433,
+      "loss": 1.2531,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14460373594362516,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002974154491756004,
+      "loss": 1.2138,
+      "step": 1649
+    },
+    {
+      "epoch": 0.1446914277179997,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002974075039319536,
+      "loss": 1.185,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14477911949237424,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029739954661308437,
+      "loss": 1.1755,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14486681126674877,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002973915772197182,
+      "loss": 1.2024,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14495450304112334,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029738359575258192,
+      "loss": 1.2599,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14504219481549788,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002973756022124035,
+      "loss": 1.2355,
+      "step": 1654
+    },
+    {
+      "epoch": 0.14512988658987241,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0029736759659991186,
+      "loss": 1.2386,
+      "step": 1655
+    },
+    {
+      "epoch": 0.14521757836424698,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0029735957891583713,
+      "loss": 1.1994,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14530527013862152,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029735154916091053,
+      "loss": 1.263,
+      "step": 1657
+    },
+    {
+      "epoch": 0.14539296191299605,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002973435073358643,
+      "loss": 1.2615,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1454806536873706,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029733545344143176,
+      "loss": 1.2311,
+      "step": 1659
+    },
+    {
+      "epoch": 0.14556834546174516,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002973273874783475,
+      "loss": 1.2165,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1456560372361197,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029731930944734723,
+      "loss": 1.2473,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14574372901049423,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029731121934916736,
+      "loss": 1.2503,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14583142078486877,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0029730311718454594,
+      "loss": 1.329,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14591911255924334,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002972950029542218,
+      "loss": 1.2774,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14600680433361787,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002972868766589348,
+      "loss": 1.2674,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1460944961079924,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029727873829942623,
+      "loss": 1.2709,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14618218788236698,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002972705878764382,
+      "loss": 1.2515,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1462698796567415,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029726242539071398,
+      "loss": 1.1644,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14635757143111605,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029725425084299803,
+      "loss": 1.1872,
+      "step": 1669
+    },
+    {
+      "epoch": 0.1464452632054906,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029724606423403577,
+      "loss": 1.2981,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14653295497986515,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029723786556457386,
+      "loss": 1.2188,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1466206467542397,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029722965483535996,
+      "loss": 1.1862,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14670833852861423,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0029722143204714293,
+      "loss": 1.253,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1467960303029888,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0029721319720067262,
+      "loss": 1.2437,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14688372207736333,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002972049502967,
+      "loss": 1.2417,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14697141385173787,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0029719669133597723,
+      "loss": 1.3076,
+      "step": 1676
+    },
+    {
+      "epoch": 0.1470591056261124,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002971884203192575,
+      "loss": 1.255,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14714679740048697,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00297180137247295,
+      "loss": 1.2467,
+      "step": 1678
+    },
+    {
+      "epoch": 0.1472344891748615,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002971718421208453,
+      "loss": 1.2239,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14732218094923605,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002971635349406647,
+      "loss": 1.2074,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14740987272361059,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029715521570751096,
+      "loss": 1.2576,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14749756449798515,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002971468844221427,
+      "loss": 1.2036,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1475852562723597,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029713854108531965,
+      "loss": 1.2238,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14767294804673423,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0029713018569780284,
+      "loss": 1.2164,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1477606398211088,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0029712181826035415,
+      "loss": 1.2165,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14784833159548333,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029711343877373672,
+      "loss": 1.2262,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14793602336985787,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029710504723871474,
+      "loss": 1.201,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1480237151442324,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029709664365605345,
+      "loss": 1.2723,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14811140691860697,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0029708822802651928,
+      "loss": 1.2362,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1481990986929815,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0029707980035087967,
+      "loss": 1.184,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14828679046735604,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002970713606299033,
+      "loss": 1.1951,
+      "step": 1691
+    },
+    {
+      "epoch": 0.1483744822417306,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002970629088643597,
+      "loss": 1.2467,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14846217401610515,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0029705444505501977,
+      "loss": 1.2589,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14854986579047968,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0029704596920265536,
+      "loss": 1.2283,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14863755756485422,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0029703748130803943,
+      "loss": 1.2138,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1487252493392288,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0029702898137194604,
+      "loss": 1.221,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14881294111360333,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0029702046939515036,
+      "loss": 1.235,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14890063288797786,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002970119453784287,
+      "loss": 1.3198,
+      "step": 1698
+    },
+    {
+      "epoch": 0.1489883246623524,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0029700340932255842,
+      "loss": 1.2292,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14907601643672697,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029699486122831795,
+      "loss": 1.2851,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1491637082111015,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002969863010964869,
+      "loss": 1.2093,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14925139998547604,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002969777289278459,
+      "loss": 1.2072,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1493390917598506,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0029696914472317672,
+      "loss": 1.2596,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14942678353422514,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029696054848326226,
+      "loss": 1.2439,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14951447530859968,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002969519402088864,
+      "loss": 1.2596,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14960216708297422,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002969433199008342,
+      "loss": 1.2164,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14968985885734878,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0029693468755989188,
+      "loss": 1.2637,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14977755063172332,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002969260431868466,
+      "loss": 1.1995,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14986524240609786,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002969173867824868,
+      "loss": 1.2292,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14995293418047242,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002969087183476018,
+      "loss": 1.2604,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15004062595484696,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029690003788298224,
+      "loss": 1.2612,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1501283177292215,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002968913453894197,
+      "loss": 1.1911,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15021600950359604,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00296882640867707,
+      "loss": 1.1865,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1503037012779706,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029687392431863788,
+      "loss": 1.2322,
+      "step": 1714
+    },
+    {
+      "epoch": 0.15039139305234514,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029686519574300724,
+      "loss": 1.3256,
+      "step": 1715
+    },
+    {
+      "epoch": 0.15047908482671968,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029685645514161123,
+      "loss": 1.1571,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15056677660109422,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0029684770251524684,
+      "loss": 1.1894,
+      "step": 1717
+    },
+    {
+      "epoch": 0.15065446837546878,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002968389378647124,
+      "loss": 1.2602,
+      "step": 1718
+    },
+    {
+      "epoch": 0.15074216014984332,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029683016119080715,
+      "loss": 1.2948,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15082985192421786,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002968213724943315,
+      "loss": 1.2603,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15091754369859242,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0029681257177608697,
+      "loss": 1.2357,
+      "step": 1721
+    },
+    {
+      "epoch": 0.15100523547296696,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0029680375903687614,
+      "loss": 1.2241,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1510929272473415,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0029679493427750277,
+      "loss": 1.291,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15118061902171603,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002967860974987716,
+      "loss": 1.2624,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1512683107960906,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002967772487014886,
+      "loss": 1.2348,
+      "step": 1725
+    },
+    {
+      "epoch": 0.15135600257046514,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029676838788646066,
+      "loss": 1.2996,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15144369434483967,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0029675951505449593,
+      "loss": 1.2744,
+      "step": 1727
+    },
+    {
+      "epoch": 0.15153138611921424,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002967506302064035,
+      "loss": 1.2476,
+      "step": 1728
+    },
+    {
+      "epoch": 0.15161907789358878,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029674173334299377,
+      "loss": 1.1993,
+      "step": 1729
+    },
+    {
+      "epoch": 0.15170676966796331,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00296732824465078,
+      "loss": 1.3051,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15179446144233785,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029672390357346873,
+      "loss": 1.2253,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15188215321671242,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002967149706689795,
+      "loss": 1.2658,
+      "step": 1732
+    },
+    {
+      "epoch": 0.15196984499108696,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00296706025752425,
+      "loss": 1.184,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1520575367654615,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002966970688246209,
+      "loss": 1.221,
+      "step": 1734
+    },
+    {
+      "epoch": 0.15214522853983603,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029668809988638405,
+      "loss": 1.2489,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1522329203142106,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002966791189385325,
+      "loss": 1.2001,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15232061208858513,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029667012598188526,
+      "loss": 1.2649,
+      "step": 1737
+    },
+    {
+      "epoch": 0.15240830386295967,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0029666112101726237,
+      "loss": 1.1744,
+      "step": 1738
+    },
+    {
+      "epoch": 0.15249599563733424,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002966521040454852,
+      "loss": 1.2393,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15258368741170877,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0029664307506737596,
+      "loss": 1.2344,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1526713791860833,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029663403408375813,
+      "loss": 1.1983,
+      "step": 1741
+    },
+    {
+      "epoch": 0.15275907096045785,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002966249810954561,
+      "loss": 1.2052,
+      "step": 1742
+    },
+    {
+      "epoch": 0.15284676273483241,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002966159161032957,
+      "loss": 1.2302,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15293445450920695,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002966068391081035,
+      "loss": 1.1733,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1530221462835815,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002965977501107073,
+      "loss": 1.1805,
+      "step": 1745
+    },
+    {
+      "epoch": 0.15310983805795605,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029658864911193596,
+      "loss": 1.2527,
+      "step": 1746
+    },
+    {
+      "epoch": 0.1531975298323306,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0029657953611261956,
+      "loss": 1.2283,
+      "step": 1747
+    },
+    {
+      "epoch": 0.15328522160670513,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029657041111358917,
+      "loss": 1.2021,
+      "step": 1748
+    },
+    {
+      "epoch": 0.15337291338107967,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029656127411567686,
+      "loss": 1.2102,
+      "step": 1749
+    },
+    {
+      "epoch": 0.15346060515545423,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0029655212511971603,
+      "loss": 1.2241,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15354829692982877,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0029654296412654096,
+      "loss": 1.2334,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1536359887042033,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029653379113698717,
+      "loss": 1.2512,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15372368047857785,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0029652460615189114,
+      "loss": 1.1526,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1538113722529524,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002965154091720906,
+      "loss": 1.2399,
+      "step": 1754
+    },
+    {
+      "epoch": 0.15389906402732695,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002965062001984242,
+      "loss": 1.2404,
+      "step": 1755
+    },
+    {
+      "epoch": 0.15398675580170149,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029649697923173185,
+      "loss": 1.2426,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15407444757607605,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029648774627285446,
+      "loss": 1.2045,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1541621393504506,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029647850132263403,
+      "loss": 1.3156,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15424983112482513,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002964692443819137,
+      "loss": 1.2339,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15433752289919966,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002964599754515377,
+      "loss": 1.2263,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15442521467357423,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002964506945323512,
+      "loss": 1.2452,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15451290644794877,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002964414016252008,
+      "loss": 1.2067,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1546005982223233,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029643209673093384,
+      "loss": 1.2505,
+      "step": 1763
+    },
+    {
+      "epoch": 0.15468828999669787,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029642277985039893,
+      "loss": 1.3219,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1547759817710724,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029641345098444585,
+      "loss": 1.2168,
+      "step": 1765
+    },
+    {
+      "epoch": 0.15486367354544694,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.002964041101339252,
+      "loss": 1.231,
+      "step": 1766
+    },
+    {
+      "epoch": 0.15495136531982148,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029639475729968893,
+      "loss": 1.19,
+      "step": 1767
+    },
+    {
+      "epoch": 0.15503905709419605,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0029638539248259,
+      "loss": 1.2543,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15512674886857059,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002963760156834825,
+      "loss": 1.2207,
+      "step": 1769
+    },
+    {
+      "epoch": 0.15521444064294512,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0029636662690322143,
+      "loss": 1.2478,
+      "step": 1770
+    },
+    {
+      "epoch": 0.15530213241731966,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029635722614266315,
+      "loss": 1.2565,
+      "step": 1771
+    },
+    {
+      "epoch": 0.15538982419169423,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002963478134026649,
+      "loss": 1.2674,
+      "step": 1772
+    },
+    {
+      "epoch": 0.15547751596606876,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002963383886840852,
+      "loss": 1.2259,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1555652077404433,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002963289519877835,
+      "loss": 1.2103,
+      "step": 1774
+    },
+    {
+      "epoch": 0.15565289951481787,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0029631950331462037,
+      "loss": 1.2993,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1557405912891924,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029631004266545756,
+      "loss": 1.2323,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15582828306356694,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002963005700411578,
+      "loss": 1.2281,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15591597483794148,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00296291085442585,
+      "loss": 1.2804,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15600366661231604,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0029628158887060416,
+      "loss": 1.1703,
+      "step": 1779
+    },
+    {
+      "epoch": 0.15609135838669058,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002962720803260813,
+      "loss": 1.1887,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15617905016106512,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0029626255980988365,
+      "loss": 1.2142,
+      "step": 1781
+    },
+    {
+      "epoch": 0.15626674193543968,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029625302732287934,
+      "loss": 1.1763,
+      "step": 1782
+    },
+    {
+      "epoch": 0.15635443370981422,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029624348286593776,
+      "loss": 1.2937,
+      "step": 1783
+    },
+    {
+      "epoch": 0.15644212548418876,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029623392643992937,
+      "loss": 1.2644,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1565298172585633,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0029622435804572563,
+      "loss": 1.2322,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15661750903293786,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029621477768419927,
+      "loss": 1.2369,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1567052008073124,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002962051853562238,
+      "loss": 1.1507,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15679289258168694,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0029619558106267424,
+      "loss": 1.2517,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15688058435606148,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0029618596480442635,
+      "loss": 1.2031,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15696827613043604,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0029617633658235707,
+      "loss": 1.3067,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15705596790481058,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0029616669639734457,
+      "loss": 1.2568,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15714365967918512,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002961570442502679,
+      "loss": 1.2221,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15723135145355968,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029614738014200745,
+      "loss": 1.3026,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15731904322793422,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029613770407344447,
+      "loss": 1.1919,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15740673500230876,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002961280160454614,
+      "loss": 1.222,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1574944267766833,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029611831605894172,
+      "loss": 1.1845,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15758211855105786,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0029610860411477015,
+      "loss": 1.3043,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1576698103254324,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0029609888021383235,
+      "loss": 1.2524,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15775750209980693,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002960891443570151,
+      "loss": 1.2282,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1578451938741815,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0029607939654520627,
+      "loss": 1.2411,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15793288564855604,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0029606963677929485,
+      "loss": 1.23,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15802057742293057,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0029605986506017093,
+      "loss": 1.2296,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1581082691973051,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029605008138872562,
+      "loss": 1.151,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15819596097167968,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029604028576585124,
+      "loss": 1.1825,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15828365274605422,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00296030478192441,
+      "loss": 1.251,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15837134452042875,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002960206586693895,
+      "loss": 1.1839,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1584590362948033,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002960108271975921,
+      "loss": 1.2336,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15854672806917786,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029600098377794543,
+      "loss": 1.2594,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1586344198435524,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0029599112841134723,
+      "loss": 1.2354,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15872211161792693,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0029598126109869633,
+      "loss": 1.2487,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1588098033923015,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002959713818408925,
+      "loss": 1.2244,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15889749516667603,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0029596149063883677,
+      "loss": 1.1701,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15898518694105057,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0029595158749343114,
+      "loss": 1.2335,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1590728787154251,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029594167240557883,
+      "loss": 1.2393,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15916057048979967,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0029593174537618392,
+      "loss": 1.244,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1592482622641742,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0029592180640615195,
+      "loss": 1.2361,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15933595403854875,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029591185549638914,
+      "loss": 1.2492,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15942364581292331,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002959018926478031,
+      "loss": 1.2262,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15951133758729785,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002958919178613023,
+      "loss": 1.2143,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1595990293616724,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002958819311377965,
+      "loss": 1.2711,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15968672113604693,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002958719324781965,
+      "loss": 1.2125,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1597744129104215,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002958619218834141,
+      "loss": 1.1969,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15986210468479603,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002958518993543622,
+      "loss": 1.2653,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15994979645917057,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002958418648919549,
+      "loss": 1.293,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1600374882335451,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0029583181849710733,
+      "loss": 1.2418,
+      "step": 1825
+    },
+    {
+      "epoch": 0.16012518000791967,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029582176017073558,
+      "loss": 1.2498,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1602128717822942,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00295811689913757,
+      "loss": 1.1711,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16030056355666875,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002958016077270901,
+      "loss": 1.2305,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1603882553310433,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0029579151361165414,
+      "loss": 1.262,
+      "step": 1829
+    },
+    {
+      "epoch": 0.16047594710541785,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029578140756836985,
+      "loss": 1.147,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1605636388797924,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0029577128959815875,
+      "loss": 1.2276,
+      "step": 1831
+    },
+    {
+      "epoch": 0.16065133065416692,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029576115970194362,
+      "loss": 1.2673,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1607390224285415,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029575101788064826,
+      "loss": 1.2303,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16082671420291603,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0029574086413519766,
+      "loss": 1.2291,
+      "step": 1834
+    },
+    {
+      "epoch": 0.16091440597729056,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029573069846651773,
+      "loss": 1.2494,
+      "step": 1835
+    },
+    {
+      "epoch": 0.16100209775166513,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002957205208755356,
+      "loss": 1.1361,
+      "step": 1836
+    },
+    {
+      "epoch": 0.16108978952603967,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0029571033136317937,
+      "loss": 1.2498,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1611774813004142,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002957001299303784,
+      "loss": 1.1754,
+      "step": 1838
+    },
+    {
+      "epoch": 0.16126517307478874,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0029568991657806295,
+      "loss": 1.2336,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1613528648491633,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002956796913071645,
+      "loss": 1.21,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16144055662353785,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002956694541186155,
+      "loss": 1.1894,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16152824839791238,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002956592050133497,
+      "loss": 1.214,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16161594017228692,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0029564894399230165,
+      "loss": 1.2265,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16170363194666149,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029563867105640716,
+      "loss": 1.2602,
+      "step": 1844
+    },
+    {
+      "epoch": 0.16179132372103602,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002956283862066031,
+      "loss": 1.2213,
+      "step": 1845
+    },
+    {
+      "epoch": 0.16187901549541056,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0029561808944382744,
+      "loss": 1.1755,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16196670726978513,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029560778076901926,
+      "loss": 1.2572,
+      "step": 1847
+    },
+    {
+      "epoch": 0.16205439904415966,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0029559746018311857,
+      "loss": 1.2332,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1621420908185342,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002955871276870667,
+      "loss": 1.2821,
+      "step": 1849
+    },
+    {
+      "epoch": 0.16222978259290874,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029557678328180586,
+      "loss": 1.2363,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1623174743672833,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002955664269682795,
+      "loss": 1.2524,
+      "step": 1851
+    },
+    {
+      "epoch": 0.16240516614165784,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029555605874743204,
+      "loss": 1.1926,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16249285791603238,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0029554567862020904,
+      "loss": 1.2426,
+      "step": 1853
+    },
+    {
+      "epoch": 0.16258054969040694,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002955352865875572,
+      "loss": 1.2712,
+      "step": 1854
+    },
+    {
+      "epoch": 0.16266824146478148,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002955248826504241,
+      "loss": 1.3041,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16275593323915602,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002955144668097588,
+      "loss": 1.1817,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16284362501353056,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0029550403906651087,
+      "loss": 1.1873,
+      "step": 1857
+    },
+    {
+      "epoch": 0.16293131678790512,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0029549359942163157,
+      "loss": 1.2236,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16301900856227966,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029548314787607288,
+      "loss": 1.2345,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1631067003366542,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002954726844307879,
+      "loss": 1.2695,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16319439211102874,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029546220908673094,
+      "loss": 1.2535,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1632820838854033,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0029545172184485733,
+      "loss": 1.1989,
+      "step": 1862
+    },
+    {
+      "epoch": 0.16336977565977784,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002954412227061234,
+      "loss": 1.1743,
+      "step": 1863
+    },
+    {
+      "epoch": 0.16345746743415238,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0029543071167148667,
+      "loss": 1.1973,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16354515920852694,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002954201887419058,
+      "loss": 1.1789,
+      "step": 1865
+    },
+    {
+      "epoch": 0.16363285098290148,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002954096539183404,
+      "loss": 1.2212,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16372054275727602,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0029539910720175113,
+      "loss": 1.2066,
+      "step": 1867
+    },
+    {
+      "epoch": 0.16380823453165055,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029538854859309997,
+      "loss": 1.2717,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16389592630602512,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029537797809334977,
+      "loss": 1.1794,
+      "step": 1869
+    },
+    {
+      "epoch": 0.16398361808039966,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002953673957034645,
+      "loss": 1.2161,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1640713098547742,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029535680142440935,
+      "loss": 1.229,
+      "step": 1871
+    },
+    {
+      "epoch": 0.16415900162914873,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029534619525715036,
+      "loss": 1.2013,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1642466934035233,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029533557720265494,
+      "loss": 1.2405,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16433438517789783,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029532494726189126,
+      "loss": 1.2009,
+      "step": 1874
+    },
+    {
+      "epoch": 0.16442207695227237,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002953143054358288,
+      "loss": 1.1971,
+      "step": 1875
+    },
+    {
+      "epoch": 0.16450976872664694,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0029530365172543817,
+      "loss": 1.1584,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16459746050102148,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002952929861316908,
+      "loss": 1.2321,
+      "step": 1877
+    },
+    {
+      "epoch": 0.164685152275396,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002952823086555595,
+      "loss": 1.235,
+      "step": 1878
+    },
+    {
+      "epoch": 0.16477284404977055,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0029527161929801798,
+      "loss": 1.2132,
+      "step": 1879
+    },
+    {
+      "epoch": 0.16486053582414512,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00295260918060041,
+      "loss": 1.2056,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16494822759851965,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029525020494260462,
+      "loss": 1.2095,
+      "step": 1881
+    },
+    {
+      "epoch": 0.1650359193728942,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029523947994668583,
+      "loss": 1.2664,
+      "step": 1882
+    },
+    {
+      "epoch": 0.16512361114726876,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029522874307326263,
+      "loss": 1.2446,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1652113029216433,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.002952179943233142,
+      "loss": 1.2069,
+      "step": 1884
+    },
+    {
+      "epoch": 0.16529899469601783,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002952072336978209,
+      "loss": 1.2048,
+      "step": 1885
+    },
+    {
+      "epoch": 0.16538668647039237,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00295196461197764,
+      "loss": 1.2081,
+      "step": 1886
+    },
+    {
+      "epoch": 0.16547437824476693,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0029518567682412593,
+      "loss": 1.1748,
+      "step": 1887
+    },
+    {
+      "epoch": 0.16556207001914147,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0029517488057789025,
+      "loss": 1.1854,
+      "step": 1888
+    },
+    {
+      "epoch": 0.165649761793516,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002951640724600414,
+      "loss": 1.2574,
+      "step": 1889
+    },
+    {
+      "epoch": 0.16573745356789055,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0029515325247156526,
+      "loss": 1.261,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1658251453422651,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029514242061344847,
+      "loss": 1.245,
+      "step": 1891
+    },
+    {
+      "epoch": 0.16591283711663965,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0029513157688667883,
+      "loss": 1.2528,
+      "step": 1892
+    },
+    {
+      "epoch": 0.1660005288910142,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002951207212922454,
+      "loss": 1.1879,
+      "step": 1893
+    },
+    {
+      "epoch": 0.16608822066538875,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00295109853831138,
+      "loss": 1.2397,
+      "step": 1894
+    },
+    {
+      "epoch": 0.1661759124397633,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029509897450434785,
+      "loss": 1.2009,
+      "step": 1895
+    },
+    {
+      "epoch": 0.16626360421413783,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002950880833128671,
+      "loss": 1.2706,
+      "step": 1896
+    },
+    {
+      "epoch": 0.16635129598851237,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0029507718025768895,
+      "loss": 1.2773,
+      "step": 1897
+    },
+    {
+      "epoch": 0.16643898776288693,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0029506626533980776,
+      "loss": 1.2068,
+      "step": 1898
+    },
+    {
+      "epoch": 0.16652667953726147,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002950553385602189,
+      "loss": 1.2264,
+      "step": 1899
+    },
+    {
+      "epoch": 0.166614371311636,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002950443999199189,
+      "loss": 1.2045,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16670206308601057,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002950334494199054,
+      "loss": 1.2409,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1667897548603851,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029502248706117693,
+      "loss": 1.2565,
+      "step": 1902
+    },
+    {
+      "epoch": 0.16687744663475965,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002950115128447333,
+      "loss": 1.3078,
+      "step": 1903
+    },
+    {
+      "epoch": 0.16696513840913418,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029500052677157535,
+      "loss": 1.1882,
+      "step": 1904
+    },
+    {
+      "epoch": 0.16705283018350875,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0029498952884270493,
+      "loss": 1.2154,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1671405219578833,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00294978519059125,
+      "loss": 1.2375,
+      "step": 1906
+    },
+    {
+      "epoch": 0.16722821373225782,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029496749742183976,
+      "loss": 1.2393,
+      "step": 1907
+    },
+    {
+      "epoch": 0.16731590550663236,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002949564639318542,
+      "loss": 1.2693,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16740359728100693,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029494541859017465,
+      "loss": 1.2,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16749128905538146,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029493436139780838,
+      "loss": 1.2399,
+      "step": 1910
+    },
+    {
+      "epoch": 0.167578980829756,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0029492329235576375,
+      "loss": 1.2153,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16766667260413057,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029491221146505024,
+      "loss": 1.176,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1677543643785051,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0029490111872667846,
+      "loss": 1.1792,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16784205615287964,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029489001414165997,
+      "loss": 1.2286,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16792974792725418,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029487889771100746,
+      "loss": 1.1379,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16801743970162875,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002948677694357348,
+      "loss": 1.1805,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16810513147600328,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002948566293168568,
+      "loss": 1.1697,
+      "step": 1917
+    },
+    {
+      "epoch": 0.16819282325037782,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029484547735538946,
+      "loss": 1.2461,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1682805150247524,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029483431355234973,
+      "loss": 1.229,
+      "step": 1919
+    },
+    {
+      "epoch": 0.16836820679912692,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002948231379087558,
+      "loss": 1.2439,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16845589857350146,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029481195042562686,
+      "loss": 1.2012,
+      "step": 1921
+    },
+    {
+      "epoch": 0.168543590347876,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002948007511039831,
+      "loss": 1.2381,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16863128212225056,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002947895399448459,
+      "loss": 1.2534,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1687189738966251,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0029477831694923774,
+      "loss": 1.2072,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16880666567099964,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029476708211818212,
+      "loss": 1.2774,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16889435744537418,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002947558354527036,
+      "loss": 1.2066,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16898204921974874,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002947445769538278,
+      "loss": 1.2122,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16906974099412328,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0029473330662258155,
+      "loss": 1.2672,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16915743276849782,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002947220244599926,
+      "loss": 1.195,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16924512454287238,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0029471073046709,
+      "loss": 1.1915,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16933281631724692,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0029469942464490353,
+      "loss": 1.2258,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16942050809162146,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002946881069944644,
+      "loss": 1.2318,
+      "step": 1932
+    },
+    {
+      "epoch": 0.169508199865996,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002946767775168047,
+      "loss": 1.2427,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16959589164037056,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002946654362129576,
+      "loss": 1.1934,
+      "step": 1934
+    },
+    {
+      "epoch": 0.1696835834147451,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029465408308395748,
+      "loss": 1.2515,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16977127518911964,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002946427181308397,
+      "loss": 1.2293,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1698589669634942,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029463134135464066,
+      "loss": 1.1644,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16994665873786874,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029461995275639795,
+      "loss": 1.2246,
+      "step": 1938
+    },
+    {
+      "epoch": 0.17003435051224328,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029460855233715013,
+      "loss": 1.2079,
+      "step": 1939
+    },
+    {
+      "epoch": 0.17012204228661781,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029459714009793694,
+      "loss": 1.2012,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17020973406099238,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029458571603979914,
+      "loss": 1.2467,
+      "step": 1941
+    },
+    {
+      "epoch": 0.17029742583536692,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0029457428016377856,
+      "loss": 1.2008,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17038511760974145,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002945628324709181,
+      "loss": 1.2972,
+      "step": 1943
+    },
+    {
+      "epoch": 0.170472809384116,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029455137296226183,
+      "loss": 1.2295,
+      "step": 1944
+    },
+    {
+      "epoch": 0.17056050115849056,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0029453990163885472,
+      "loss": 1.2901,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1706481929328651,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00294528418501743,
+      "loss": 1.2214,
+      "step": 1946
+    },
+    {
+      "epoch": 0.17073588470723963,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0029451692355197396,
+      "loss": 1.2061,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1708235764816142,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0029450541679059573,
+      "loss": 1.2308,
+      "step": 1948
+    },
+    {
+      "epoch": 0.17091126825598874,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029449389821865793,
+      "loss": 1.1401,
+      "step": 1949
+    },
+    {
+      "epoch": 0.17099896003036327,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0029448236783721077,
+      "loss": 1.2164,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1710866518047378,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00294470825647306,
+      "loss": 1.3284,
+      "step": 1951
+    },
+    {
+      "epoch": 0.17117434357911238,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029445927164999616,
+      "loss": 1.2917,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1712620353534869,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0029444770584633488,
+      "loss": 1.2137,
+      "step": 1953
+    },
+    {
+      "epoch": 0.17134972712786145,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029443612823737706,
+      "loss": 1.1714,
+      "step": 1954
+    },
+    {
+      "epoch": 0.17143741890223602,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029442453882417845,
+      "loss": 1.2233,
+      "step": 1955
+    },
+    {
+      "epoch": 0.17152511067661055,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00294412937607796,
+      "loss": 1.2183,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1716128024509851,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002944013245892877,
+      "loss": 1.2503,
+      "step": 1957
+    },
+    {
+      "epoch": 0.17170049422535963,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029438969976971273,
+      "loss": 1.2305,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1717881859997342,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.002943780631501311,
+      "loss": 1.1944,
+      "step": 1959
+    },
+    {
+      "epoch": 0.17187587777410873,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002943664147316041,
+      "loss": 1.247,
+      "step": 1960
+    },
+    {
+      "epoch": 0.17196356954848327,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029435475451519404,
+      "loss": 1.2013,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1720512613228578,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002943430825019643,
+      "loss": 1.3264,
+      "step": 1962
+    },
+    {
+      "epoch": 0.17213895309723237,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002943313986929793,
+      "loss": 1.2173,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1722266448716069,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029431970308930466,
+      "loss": 1.258,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17231433664598145,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0029430799569200687,
+      "loss": 1.2104,
+      "step": 1965
+    },
+    {
+      "epoch": 0.172402028420356,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029429627650215374,
+      "loss": 1.172,
+      "step": 1966
+    },
+    {
+      "epoch": 0.17248972019473055,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0029428454552081393,
+      "loss": 1.2442,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1725774119691051,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002942728027490574,
+      "loss": 1.1975,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17266510374347963,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002942610481879549,
+      "loss": 1.2699,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1727527955178542,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029424928183857846,
+      "loss": 1.2762,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17284048729222873,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002942375037020012,
+      "loss": 1.2617,
+      "step": 1971
+    },
+    {
+      "epoch": 0.17292817906660327,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029422571377929726,
+      "loss": 1.2496,
+      "step": 1972
+    },
+    {
+      "epoch": 0.17301587084097783,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0029421391207154184,
+      "loss": 1.1984,
+      "step": 1973
+    },
+    {
+      "epoch": 0.17310356261535237,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0029420209857981116,
+      "loss": 1.2159,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1731912543897269,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0029419027330518255,
+      "loss": 1.1537,
+      "step": 1975
+    },
+    {
+      "epoch": 0.17327894616410144,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002941784362487346,
+      "loss": 1.2383,
+      "step": 1976
+    },
+    {
+      "epoch": 0.173366637938476,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029416658741154675,
+      "loss": 1.1781,
+      "step": 1977
+    },
+    {
+      "epoch": 0.17345432971285055,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029415472679469954,
+      "loss": 1.2203,
+      "step": 1978
+    },
+    {
+      "epoch": 0.17354202148722508,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0029414285439927465,
+      "loss": 1.2184,
+      "step": 1979
+    },
+    {
+      "epoch": 0.17362971326159962,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0029413097022635486,
+      "loss": 1.2343,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1737174050359742,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002941190742770239,
+      "loss": 1.2458,
+      "step": 1981
+    },
+    {
+      "epoch": 0.17380509681034872,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002941071665523667,
+      "loss": 1.1951,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17389278858472326,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029409524705346917,
+      "loss": 1.2263,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17398048035909783,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002940833157814184,
+      "loss": 1.3174,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17406817213347237,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029407137273730244,
+      "loss": 1.2567,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1741558639078469,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002940594179222105,
+      "loss": 1.3004,
+      "step": 1986
+    },
+    {
+      "epoch": 0.17424355568222144,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002940474513372328,
+      "loss": 1.1722,
+      "step": 1987
+    },
+    {
+      "epoch": 0.174331247456596,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0029403547298346064,
+      "loss": 1.264,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17441893923097054,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0029402348286198653,
+      "loss": 1.204,
+      "step": 1989
+    },
+    {
+      "epoch": 0.17450663100534508,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002940114809739038,
+      "loss": 1.1928,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17459432277971965,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0029399946732030706,
+      "loss": 1.2589,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17468201455409418,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029398744190229187,
+      "loss": 1.1797,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17476970632846872,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00293975404720955,
+      "loss": 1.247,
+      "step": 1993
+    },
+    {
+      "epoch": 0.17485739810284326,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029396335577739413,
+      "loss": 1.1877,
+      "step": 1994
+    },
+    {
+      "epoch": 0.17494508987721782,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029395129507270817,
+      "loss": 1.2879,
+      "step": 1995
+    },
+    {
+      "epoch": 0.17503278165159236,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00293939222607997,
+      "loss": 1.284,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1751204734259669,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002939271383843615,
+      "loss": 1.1872,
+      "step": 1997
+    },
+    {
+      "epoch": 0.17520816520034144,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002939150424029039,
+      "loss": 1.2073,
+      "step": 1998
+    },
+    {
+      "epoch": 0.175295856974716,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002939029346647272,
+      "loss": 1.242,
+      "step": 1999
+    },
+    {
+      "epoch": 0.17538354874909054,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029389081517093556,
+      "loss": 1.2536,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17538354874909054,
+      "eval_loss": 1.2374000549316406,
+      "eval_runtime": 428.5743,
+      "eval_samples_per_second": 33.709,
+      "eval_steps_per_second": 8.428,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17547124052346508,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002938786839226344,
+      "loss": 1.1888,
+      "step": 2001
+    },
+    {
+      "epoch": 0.17555893229783964,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002938665409209299,
+      "loss": 1.2685,
+      "step": 2002
+    },
+    {
+      "epoch": 0.17564662407221418,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029385438616692954,
+      "loss": 1.2854,
+      "step": 2003
+    },
+    {
+      "epoch": 0.17573431584658872,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002938422196617418,
+      "loss": 1.2861,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17582200762096326,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029383004140647625,
+      "loss": 1.2657,
+      "step": 2005
+    },
+    {
+      "epoch": 0.17590969939533782,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002938178514022435,
+      "loss": 1.2393,
+      "step": 2006
+    },
+    {
+      "epoch": 0.17599739116971236,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002938056496501552,
+      "loss": 1.2239,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1760850829440869,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002937934361513242,
+      "loss": 1.2971,
+      "step": 2008
+    },
+    {
+      "epoch": 0.17617277471846146,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002937812109068643,
+      "loss": 1.2387,
+      "step": 2009
+    },
+    {
+      "epoch": 0.176260466492836,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002937689739178904,
+      "loss": 1.2409,
+      "step": 2010
+    },
+    {
+      "epoch": 0.17634815826721054,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002937567251855185,
+      "loss": 1.2703,
+      "step": 2011
+    },
+    {
+      "epoch": 0.17643585004158507,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0029374446471086566,
+      "loss": 1.1958,
+      "step": 2012
+    },
+    {
+      "epoch": 0.17652354181595964,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0029373219249504996,
+      "loss": 1.23,
+      "step": 2013
+    },
+    {
+      "epoch": 0.17661123359033418,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002937199085391907,
+      "loss": 1.2415,
+      "step": 2014
+    },
+    {
+      "epoch": 0.17669892536470871,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00293707612844408,
+      "loss": 1.2129,
+      "step": 2015
+    },
+    {
+      "epoch": 0.17678661713908325,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0029369530541182324,
+      "loss": 1.1714,
+      "step": 2016
+    },
+    {
+      "epoch": 0.17687430891345782,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0029368298624255895,
+      "loss": 1.2369,
+      "step": 2017
+    },
+    {
+      "epoch": 0.17696200068783235,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002936706553377384,
+      "loss": 1.1888,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1770496924622069,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002936583126984863,
+      "loss": 1.2226,
+      "step": 2019
+    },
+    {
+      "epoch": 0.17713738423658146,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002936459583259282,
+      "loss": 1.264,
+      "step": 2020
+    },
+    {
+      "epoch": 0.177225076010956,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002936335922211908,
+      "loss": 1.2089,
+      "step": 2021
+    },
+    {
+      "epoch": 0.17731276778533053,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029362121438540187,
+      "loss": 1.2172,
+      "step": 2022
+    },
+    {
+      "epoch": 0.17740045955970507,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002936088248196902,
+      "loss": 1.2532,
+      "step": 2023
+    },
+    {
+      "epoch": 0.17748815133407964,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002935964235251857,
+      "loss": 1.2628,
+      "step": 2024
+    },
+    {
+      "epoch": 0.17757584310845417,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002935840105030194,
+      "loss": 1.2052,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1776635348828287,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029357158575432324,
+      "loss": 1.26,
+      "step": 2026
+    },
+    {
+      "epoch": 0.17775122665720328,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0029355914928023037,
+      "loss": 1.2588,
+      "step": 2027
+    },
+    {
+      "epoch": 0.1778389184315778,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0029354670108187497,
+      "loss": 1.2044,
+      "step": 2028
+    },
+    {
+      "epoch": 0.17792661020595235,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029353424116039225,
+      "loss": 1.1899,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1780143019803269,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0029352176951691856,
+      "loss": 1.235,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17810199375470145,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029350928615259125,
+      "loss": 1.2379,
+      "step": 2031
+    },
+    {
+      "epoch": 0.178189685529076,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002934967910685488,
+      "loss": 1.3001,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17827737730345053,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002934842842659307,
+      "loss": 1.2333,
+      "step": 2033
+    },
+    {
+      "epoch": 0.17836506907782507,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002934717657458776,
+      "loss": 1.2045,
+      "step": 2034
+    },
+    {
+      "epoch": 0.17845276085219963,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0029345923550953106,
+      "loss": 1.1979,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17854045262657417,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0029344669355803386,
+      "loss": 1.2464,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1786281444009487,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0029343413989252986,
+      "loss": 1.1283,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17871583617532327,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029342157451416376,
+      "loss": 1.3007,
+      "step": 2038
+    },
+    {
+      "epoch": 0.1788035279496978,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002934089974240817,
+      "loss": 1.2194,
+      "step": 2039
+    },
+    {
+      "epoch": 0.17889121972407235,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0029339640862343047,
+      "loss": 1.2209,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17897891149844689,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029338380811335823,
+      "loss": 1.1927,
+      "step": 2041
+    },
+    {
+      "epoch": 0.17906660327282145,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0029337119589501414,
+      "loss": 1.2917,
+      "step": 2042
+    },
+    {
+      "epoch": 0.179154295047196,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0029335857196954836,
+      "loss": 1.3238,
+      "step": 2043
+    },
+    {
+      "epoch": 0.17924198682157053,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002933459363381122,
+      "loss": 1.2641,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1793296785959451,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029333328900185795,
+      "loss": 1.2249,
+      "step": 2045
+    },
+    {
+      "epoch": 0.17941737037031963,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00293320629961939,
+      "loss": 1.2576,
+      "step": 2046
+    },
+    {
+      "epoch": 0.17950506214469417,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002933079592195099,
+      "loss": 1.2239,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1795927539190687,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0029329527677572614,
+      "loss": 1.2142,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17968044569344327,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0029328258263174433,
+      "loss": 1.2847,
+      "step": 2049
+    },
+    {
+      "epoch": 0.1797681374678178,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002932698767887221,
+      "loss": 1.338,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17985582924219234,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0029325715924781825,
+      "loss": 1.2185,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17994352101656688,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002932444300101926,
+      "loss": 1.1705,
+      "step": 2052
+    },
+    {
+      "epoch": 0.18003121279094145,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0029323168907700596,
+      "loss": 1.2895,
+      "step": 2053
+    },
+    {
+      "epoch": 0.18011890456531598,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0029321893644942035,
+      "loss": 1.2526,
+      "step": 2054
+    },
+    {
+      "epoch": 0.18020659633969052,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029320617212859872,
+      "loss": 1.1969,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1802942881140651,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002931933961157051,
+      "loss": 1.1678,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18038197988843963,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002931806084119048,
+      "loss": 1.2011,
+      "step": 2057
+    },
+    {
+      "epoch": 0.18046967166281416,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0029316780901836375,
+      "loss": 1.1828,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1805573634371887,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002931549979362495,
+      "loss": 1.2415,
+      "step": 2059
+    },
+    {
+      "epoch": 0.18064505521156327,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002931421751667302,
+      "loss": 1.2183,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1807327469859378,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002931293407109754,
+      "loss": 1.2753,
+      "step": 2061
+    },
+    {
+      "epoch": 0.18082043876031234,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0029311649457015544,
+      "loss": 1.3004,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1809081305346869,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029310363674544195,
+      "loss": 1.1922,
+      "step": 2063
+    },
+    {
+      "epoch": 0.18099582230906144,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029309076723800743,
+      "loss": 1.2012,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18108351408343598,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0029307788604902565,
+      "loss": 1.2232,
+      "step": 2065
+    },
+    {
+      "epoch": 0.18117120585781052,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002930649931796712,
+      "loss": 1.3227,
+      "step": 2066
+    },
+    {
+      "epoch": 0.18125889763218508,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029305208863112005,
+      "loss": 1.1751,
+      "step": 2067
+    },
+    {
+      "epoch": 0.18134658940655962,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00293039172404549,
+      "loss": 1.1815,
+      "step": 2068
+    },
+    {
+      "epoch": 0.18143428118093416,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029302624450113592,
+      "loss": 1.2264,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1815219729553087,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002930133049220598,
+      "loss": 1.1518,
+      "step": 2070
+    },
+    {
+      "epoch": 0.18160966472968326,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002930003536685008,
+      "loss": 1.217,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1816973565040578,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029298739074164,
+      "loss": 1.1827,
+      "step": 2072
+    },
+    {
+      "epoch": 0.18178504827843234,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002929744161426595,
+      "loss": 1.2134,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1818727400528069,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002929614298727426,
+      "loss": 1.215,
+      "step": 2074
+    },
+    {
+      "epoch": 0.18196043182718144,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0029294843193307368,
+      "loss": 1.2318,
+      "step": 2075
+    },
+    {
+      "epoch": 0.18204812360155598,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029293542232483806,
+      "loss": 1.2634,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18213581537593052,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029292240104922215,
+      "loss": 1.2311,
+      "step": 2077
+    },
+    {
+      "epoch": 0.18222350715030508,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0029290936810741353,
+      "loss": 1.2753,
+      "step": 2078
+    },
+    {
+      "epoch": 0.18231119892467962,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0029289632350060074,
+      "loss": 1.2298,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18239889069905416,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002928832672299734,
+      "loss": 1.2143,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1824865824734287,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002928701992967222,
+      "loss": 1.2469,
+      "step": 2081
+    },
+    {
+      "epoch": 0.18257427424780326,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029285711970203888,
+      "loss": 1.2208,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1826619660221778,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0029284402844711637,
+      "loss": 1.2153,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18274965779655233,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0029283092553314847,
+      "loss": 1.2403,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1828373495709269,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0029281781096133015,
+      "loss": 1.1838,
+      "step": 2085
+    },
+    {
+      "epoch": 0.18292504134530144,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0029280468473285746,
+      "loss": 1.25,
+      "step": 2086
+    },
+    {
+      "epoch": 0.18301273311967597,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002927915468489274,
+      "loss": 1.2227,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1831004248940505,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002927783973107382,
+      "loss": 1.276,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18318811666842508,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029276523611948906,
+      "loss": 1.2114,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18327580844279961,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0029275206327638017,
+      "loss": 1.2702,
+      "step": 2090
+    },
+    {
+      "epoch": 0.18336350021717415,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029273887878261285,
+      "loss": 1.185,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18345119199154872,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002927256826393896,
+      "loss": 1.1702,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18353888376592326,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029271247484791377,
+      "loss": 1.2359,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1836265755402978,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029269925540939,
+      "loss": 1.2313,
+      "step": 2094
+    },
+    {
+      "epoch": 0.18371426731467233,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002926860243250237,
+      "loss": 1.2114,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1838019590890469,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002926727815960216,
+      "loss": 1.2709,
+      "step": 2096
+    },
+    {
+      "epoch": 0.18388965086342143,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0029265952722359154,
+      "loss": 1.2259,
+      "step": 2097
+    },
+    {
+      "epoch": 0.18397734263779597,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029264626120894203,
+      "loss": 1.2063,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1840650344121705,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0029263298355328305,
+      "loss": 1.1828,
+      "step": 2099
+    },
+    {
+      "epoch": 0.18415272618654507,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0029261969425782544,
+      "loss": 1.212,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1842404179609196,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029260639332378117,
+      "loss": 1.1854,
+      "step": 2101
+    },
+    {
+      "epoch": 0.18432810973529415,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002925930807523632,
+      "loss": 1.2632,
+      "step": 2102
+    },
+    {
+      "epoch": 0.18441580150966871,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002925797565447857,
+      "loss": 1.231,
+      "step": 2103
+    },
+    {
+      "epoch": 0.18450349328404325,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029256642070226374,
+      "loss": 1.3015,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1845911850584178,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002925530732260135,
+      "loss": 1.1686,
+      "step": 2105
+    },
+    {
+      "epoch": 0.18467887683279233,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.002925397141172523,
+      "loss": 1.2009,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1847665686071669,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029252634337719836,
+      "loss": 1.2491,
+      "step": 2107
+    },
+    {
+      "epoch": 0.18485426038154143,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0029251296100707117,
+      "loss": 1.252,
+      "step": 2108
+    },
+    {
+      "epoch": 0.18494195215591597,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0029249956700809106,
+      "loss": 1.2084,
+      "step": 2109
+    },
+    {
+      "epoch": 0.18502964393029053,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002924861613814796,
+      "loss": 1.2012,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18511733570466507,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0029247274412845933,
+      "loss": 1.2048,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1852050274790396,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029245931525025384,
+      "loss": 1.3062,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18529271925341415,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002924458747480879,
+      "loss": 1.2684,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1853804110277887,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0029243242262318708,
+      "loss": 1.1959,
+      "step": 2114
+    },
+    {
+      "epoch": 0.18546810280216325,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029241895887677834,
+      "loss": 1.2658,
+      "step": 2115
+    },
+    {
+      "epoch": 0.18555579457653779,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002924054835100895,
+      "loss": 1.241,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18564348635091232,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0029239199652434937,
+      "loss": 1.1998,
+      "step": 2117
+    },
+    {
+      "epoch": 0.1857311781252869,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029237849792078815,
+      "loss": 1.2349,
+      "step": 2118
+    },
+    {
+      "epoch": 0.18581886989966143,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0029236498770063664,
+      "loss": 1.275,
+      "step": 2119
+    },
+    {
+      "epoch": 0.18590656167403596,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002923514658651271,
+      "loss": 1.2772,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18599425344841053,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029233793241549262,
+      "loss": 1.2346,
+      "step": 2121
+    },
+    {
+      "epoch": 0.18608194522278507,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029232438735296742,
+      "loss": 1.2353,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1861696369971596,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002923108306787868,
+      "loss": 1.2354,
+      "step": 2123
+    },
+    {
+      "epoch": 0.18625732877153414,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00292297262394187,
+      "loss": 1.1676,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1863450205459087,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002922836825004056,
+      "loss": 1.1817,
+      "step": 2125
+    },
+    {
+      "epoch": 0.18643271232028324,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002922700909986808,
+      "loss": 1.2602,
+      "step": 2126
+    },
+    {
+      "epoch": 0.18652040409465778,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0029225648789025236,
+      "loss": 1.274,
+      "step": 2127
+    },
+    {
+      "epoch": 0.18660809586903235,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029224287317636067,
+      "loss": 1.2017,
+      "step": 2128
+    },
+    {
+      "epoch": 0.18669578764340689,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.002922292468582475,
+      "loss": 1.2335,
+      "step": 2129
+    },
+    {
+      "epoch": 0.18678347941778142,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029221560893715535,
+      "loss": 1.2363,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18687117119215596,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002922019594143281,
+      "loss": 1.2476,
+      "step": 2131
+    },
+    {
+      "epoch": 0.18695886296653053,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029218829829101055,
+      "loss": 1.2307,
+      "step": 2132
+    },
+    {
+      "epoch": 0.18704655474090506,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002921746255684485,
+      "loss": 1.2079,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1871342465152796,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029216094124788895,
+      "loss": 1.1987,
+      "step": 2134
+    },
+    {
+      "epoch": 0.18722193828965414,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0029214724533057983,
+      "loss": 1.205,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1873096300640287,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002921335378177701,
+      "loss": 1.2035,
+      "step": 2136
+    },
+    {
+      "epoch": 0.18739732183840324,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002921198187107099,
+      "loss": 1.1853,
+      "step": 2137
+    },
+    {
+      "epoch": 0.18748501361277778,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029210608801065047,
+      "loss": 1.2574,
+      "step": 2138
+    },
+    {
+      "epoch": 0.18757270538715234,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002920923457188439,
+      "loss": 1.2154,
+      "step": 2139
+    },
+    {
+      "epoch": 0.18766039716152688,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002920785918365435,
+      "loss": 1.259,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18774808893590142,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002920648263650036,
+      "loss": 1.1876,
+      "step": 2141
+    },
+    {
+      "epoch": 0.18783578071027596,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029205104930547948,
+      "loss": 1.2673,
+      "step": 2142
+    },
+    {
+      "epoch": 0.18792347248465052,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002920372606592277,
+      "loss": 1.246,
+      "step": 2143
+    },
+    {
+      "epoch": 0.18801116425902506,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002920234604275057,
+      "loss": 1.2107,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1880988560333996,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00292009648611572,
+      "loss": 1.255,
+      "step": 2145
+    },
+    {
+      "epoch": 0.18818654780777416,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0029199582521268618,
+      "loss": 1.2306,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1882742395821487,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00291981990232109,
+      "loss": 1.2099,
+      "step": 2147
+    },
+    {
+      "epoch": 0.18836193135652324,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002919681436711021,
+      "loss": 1.2574,
+      "step": 2148
+    },
+    {
+      "epoch": 0.18844962313089778,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029195428553092824,
+      "loss": 1.2209,
+      "step": 2149
+    },
+    {
+      "epoch": 0.18853731490527234,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.002919404158128513,
+      "loss": 1.2061,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18862500667964688,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029192653451813604,
+      "loss": 1.2044,
+      "step": 2151
+    },
+    {
+      "epoch": 0.18871269845402142,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.002919126416480486,
+      "loss": 1.2074,
+      "step": 2152
+    },
+    {
+      "epoch": 0.18880039022839595,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002918987372038558,
+      "loss": 1.2965,
+      "step": 2153
+    },
+    {
+      "epoch": 0.18888808200277052,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002918848211868258,
+      "loss": 1.2164,
+      "step": 2154
+    },
+    {
+      "epoch": 0.18897577377714506,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029187089359822764,
+      "loss": 1.2771,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1890634655515196,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002918569544393315,
+      "loss": 1.2297,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18915115732589416,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002918430037114086,
+      "loss": 1.2593,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1892388491002687,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002918290414157312,
+      "loss": 1.2255,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18932654087464323,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029181506755357263,
+      "loss": 1.2266,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18941423264901777,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0029180108212620726,
+      "loss": 1.2049,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18950192442339234,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0029178708513491053,
+      "loss": 1.2316,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18958961619776687,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0029177307658095897,
+      "loss": 1.2655,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1896773079721414,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0029175905646563007,
+      "loss": 1.2376,
+      "step": 2163
+    },
+    {
+      "epoch": 0.18976499974651598,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0029174502479020247,
+      "loss": 1.2186,
+      "step": 2164
+    },
+    {
+      "epoch": 0.18985269152089052,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002917309815559558,
+      "loss": 1.2951,
+      "step": 2165
+    },
+    {
+      "epoch": 0.18994038329526505,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0029171692676417072,
+      "loss": 1.2102,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1900280750696396,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0029170286041612908,
+      "loss": 1.2002,
+      "step": 2167
+    },
+    {
+      "epoch": 0.19011576684401416,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002916887825131136,
+      "loss": 1.2508,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1902034586183887,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029167469305640824,
+      "loss": 1.2636,
+      "step": 2169
+    },
+    {
+      "epoch": 0.19029115039276323,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0029166059204729787,
+      "loss": 1.2211,
+      "step": 2170
+    },
+    {
+      "epoch": 0.19037884216713777,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002916464794870685,
+      "loss": 1.2183,
+      "step": 2171
+    },
+    {
+      "epoch": 0.19046653394151233,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002916323553770071,
+      "loss": 1.291,
+      "step": 2172
+    },
+    {
+      "epoch": 0.19055422571588687,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0029161821971840185,
+      "loss": 1.2438,
+      "step": 2173
+    },
+    {
+      "epoch": 0.1906419174902614,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0029160407251254176,
+      "loss": 1.218,
+      "step": 2174
+    },
+    {
+      "epoch": 0.19072960926463597,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029158991376071713,
+      "loss": 1.1763,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1908173010390105,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029157574346421918,
+      "loss": 1.2258,
+      "step": 2176
+    },
+    {
+      "epoch": 0.19090499281338505,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0029156156162434012,
+      "loss": 1.2147,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1909926845877596,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0029154736824237334,
+      "loss": 1.2553,
+      "step": 2178
+    },
+    {
+      "epoch": 0.19108037636213415,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002915331633196133,
+      "loss": 1.1916,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1911680681365087,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002915189468573554,
+      "loss": 1.1776,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19125575991088323,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0029150471885689613,
+      "loss": 1.208,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1913434516852578,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002914904793195331,
+      "loss": 1.2131,
+      "step": 2182
+    },
+    {
+      "epoch": 0.19143114345963233,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0029147622824656485,
+      "loss": 1.2452,
+      "step": 2183
+    },
+    {
+      "epoch": 0.19151883523400687,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029146196563929116,
+      "loss": 1.2204,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1916065270083814,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002914476914990126,
+      "loss": 1.2134,
+      "step": 2185
+    },
+    {
+      "epoch": 0.19169421878275597,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0029143340582703098,
+      "loss": 1.2047,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1917819105571305,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029141910862464917,
+      "loss": 1.2427,
+      "step": 2187
+    },
+    {
+      "epoch": 0.19186960233150505,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029140479989317103,
+      "loss": 1.2311,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19195729410587958,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002913904796339014,
+      "loss": 1.2115,
+      "step": 2189
+    },
+    {
+      "epoch": 0.19204498588025415,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002913761478481463,
+      "loss": 1.2186,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1921326776546287,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002913618045372128,
+      "loss": 1.278,
+      "step": 2191
+    },
+    {
+      "epoch": 0.19222036942900322,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029134744970240895,
+      "loss": 1.2644,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1923080612033778,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002913330833450438,
+      "loss": 1.2333,
+      "step": 2193
+    },
+    {
+      "epoch": 0.19239575297775233,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002913187054664276,
+      "loss": 1.2697,
+      "step": 2194
+    },
+    {
+      "epoch": 0.19248344475212686,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0029130431606787153,
+      "loss": 1.2019,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1925711365265014,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029128991515068787,
+      "loss": 1.2012,
+      "step": 2196
+    },
+    {
+      "epoch": 0.19265882830087597,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0029127550271619003,
+      "loss": 1.2528,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1927465200752505,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029126107876569223,
+      "loss": 1.2931,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19283421184962504,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002912466433005101,
+      "loss": 1.2057,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1929219036239996,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029123219632195985,
+      "loss": 1.2672,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19300959539837415,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002912177378313593,
+      "loss": 1.207,
+      "step": 2201
+    },
+    {
+      "epoch": 0.19309728717274868,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002912032678300268,
+      "loss": 1.2218,
+      "step": 2202
+    },
+    {
+      "epoch": 0.19318497894712322,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0029118878631928216,
+      "loss": 1.2283,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19327267072149779,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002911742933004459,
+      "loss": 1.277,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19336036249587232,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0029115978877483984,
+      "loss": 1.18,
+      "step": 2205
+    },
+    {
+      "epoch": 0.19344805427024686,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0029114527274378666,
+      "loss": 1.1876,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1935357460446214,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0029113074520861035,
+      "loss": 1.2212,
+      "step": 2207
+    },
+    {
+      "epoch": 0.19362343781899596,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002911162061706356,
+      "loss": 1.1823,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1937111295933705,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0029110165563118844,
+      "loss": 1.2737,
+      "step": 2209
+    },
+    {
+      "epoch": 0.19379882136774504,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002910870935915958,
+      "loss": 1.2434,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1938865131421196,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029107252005318575,
+      "loss": 1.2678,
+      "step": 2211
+    },
+    {
+      "epoch": 0.19397420491649414,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002910579350172873,
+      "loss": 1.2172,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19406189669086868,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002910433384852306,
+      "loss": 1.2179,
+      "step": 2213
+    },
+    {
+      "epoch": 0.19414958846524322,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029102873045834684,
+      "loss": 1.2059,
+      "step": 2214
+    },
+    {
+      "epoch": 0.19423728023961778,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0029101411093796813,
+      "loss": 1.1747,
+      "step": 2215
+    },
+    {
+      "epoch": 0.19432497201399232,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029099947992542786,
+      "loss": 1.2096,
+      "step": 2216
+    },
+    {
+      "epoch": 0.19441266378836686,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002909848374220603,
+      "loss": 1.1815,
+      "step": 2217
+    },
+    {
+      "epoch": 0.19450035556274142,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0029097018342920077,
+      "loss": 1.2166,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19458804733711596,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0029095551794818568,
+      "loss": 1.2344,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1946757391114905,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002909408409803525,
+      "loss": 1.1877,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19476343088586504,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0029092615252703977,
+      "loss": 1.2411,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1948511226602396,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0029091145258958702,
+      "loss": 1.2854,
+      "step": 2222
+    },
+    {
+      "epoch": 0.19493881443461414,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0029089674116933484,
+      "loss": 1.2076,
+      "step": 2223
+    },
+    {
+      "epoch": 0.19502650620898868,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002908820182676248,
+      "loss": 1.217,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1951141979833632,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002908672838857997,
+      "loss": 1.3098,
+      "step": 2225
+    },
+    {
+      "epoch": 0.19520188975773778,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0029085253802520325,
+      "loss": 1.2397,
+      "step": 2226
+    },
+    {
+      "epoch": 0.19528958153211232,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029083778068718018,
+      "loss": 1.2012,
+      "step": 2227
+    },
+    {
+      "epoch": 0.19537727330648685,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.002908230118730764,
+      "loss": 1.3807,
+      "step": 2228
+    },
+    {
+      "epoch": 0.19546496508086142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002908082315842387,
+      "loss": 1.2038,
+      "step": 2229
+    },
+    {
+      "epoch": 0.19555265685523596,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0029079343982201513,
+      "loss": 1.24,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1956403486296105,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029077863658775447,
+      "loss": 1.2039,
+      "step": 2231
+    },
+    {
+      "epoch": 0.19572804040398503,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0029076382188280695,
+      "loss": 1.1737,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1958157321783596,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002907489957085235,
+      "loss": 1.2468,
+      "step": 2233
+    },
+    {
+      "epoch": 0.19590342395273413,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029073415806625622,
+      "loss": 1.2271,
+      "step": 2234
+    },
+    {
+      "epoch": 0.19599111572710867,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029071930895735835,
+      "loss": 1.2028,
+      "step": 2235
+    },
+    {
+      "epoch": 0.19607880750148324,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0029070444838318406,
+      "loss": 1.2072,
+      "step": 2236
+    },
+    {
+      "epoch": 0.19616649927585778,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0029068957634508855,
+      "loss": 1.1776,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1962541910502323,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002906746928444281,
+      "loss": 1.2094,
+      "step": 2238
+    },
+    {
+      "epoch": 0.19634188282460685,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0029065979788256016,
+      "loss": 1.2517,
+      "step": 2239
+    },
+    {
+      "epoch": 0.19642957459898142,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029064489146084305,
+      "loss": 1.2747,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19651726637335595,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0029062997358063614,
+      "loss": 1.1872,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1966049581477305,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029061504424329997,
+      "loss": 1.2261,
+      "step": 2242
+    },
+    {
+      "epoch": 0.19669264992210503,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0029060010345019604,
+      "loss": 1.2154,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1967803416964796,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0029058515120268693,
+      "loss": 1.1852,
+      "step": 2244
+    },
+    {
+      "epoch": 0.19686803347085413,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0029057018750213623,
+      "loss": 1.2149,
+      "step": 2245
+    },
+    {
+      "epoch": 0.19695572524522867,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0029055521234990856,
+      "loss": 1.2663,
+      "step": 2246
+    },
+    {
+      "epoch": 0.19704341701960323,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029054022574736965,
+      "loss": 1.229,
+      "step": 2247
+    },
+    {
+      "epoch": 0.19713110879397777,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029052522769588618,
+      "loss": 1.2652,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1972188005683523,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002905102181968261,
+      "loss": 1.1608,
+      "step": 2249
+    },
+    {
+      "epoch": 0.19730649234272685,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029049519725155805,
+      "loss": 1.2427,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1973941841171014,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00290480164861452,
+      "loss": 1.2146,
+      "step": 2251
+    },
+    {
+      "epoch": 0.19748187589147595,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029046512102787877,
+      "loss": 1.2384,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1975695676658505,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002904500657522105,
+      "loss": 1.2716,
+      "step": 2253
+    },
+    {
+      "epoch": 0.19765725944022505,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0029043499903582,
+      "loss": 1.2459,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1977449512145996,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002904199208800814,
+      "loss": 1.1976,
+      "step": 2255
+    },
+    {
+      "epoch": 0.19783264298897413,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029040483128636983,
+      "loss": 1.2153,
+      "step": 2256
+    },
+    {
+      "epoch": 0.19792033476334867,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0029038973025606136,
+      "loss": 1.243,
+      "step": 2257
+    },
+    {
+      "epoch": 0.19800802653772323,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0029037461779053315,
+      "loss": 1.1208,
+      "step": 2258
+    },
+    {
+      "epoch": 0.19809571831209777,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002903594938911635,
+      "loss": 1.2481,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1981834100864723,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0029034435855933164,
+      "loss": 1.2173,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19827110186084684,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0029032921179641775,
+      "loss": 1.227,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1983587936352214,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002903140536038034,
+      "loss": 1.2387,
+      "step": 2262
+    },
+    {
+      "epoch": 0.19844648540959595,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002902988839828708,
+      "loss": 1.2604,
+      "step": 2263
+    },
+    {
+      "epoch": 0.19853417718397048,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002902837029350034,
+      "loss": 1.2199,
+      "step": 2264
+    },
+    {
+      "epoch": 0.19862186895834505,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029026851046158575,
+      "loss": 1.2029,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1987095607327196,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029025330656400334,
+      "loss": 1.2838,
+      "step": 2266
+    },
+    {
+      "epoch": 0.19879725250709412,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002902380912436427,
+      "loss": 1.2007,
+      "step": 2267
+    },
+    {
+      "epoch": 0.19888494428146866,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029022286450189143,
+      "loss": 1.2441,
+      "step": 2268
+    },
+    {
+      "epoch": 0.19897263605584323,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002902076263401382,
+      "loss": 1.274,
+      "step": 2269
+    },
+    {
+      "epoch": 0.19906032783021776,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002901923767597726,
+      "loss": 1.2176,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1991480196045923,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0029017711576218553,
+      "loss": 1.2198,
+      "step": 2271
+    },
+    {
+      "epoch": 0.19923571137896687,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029016184334876858,
+      "loss": 1.2703,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1993234031533414,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029014655952091465,
+      "loss": 1.2411,
+      "step": 2273
+    },
+    {
+      "epoch": 0.19941109492771594,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0029013126428001756,
+      "loss": 1.2385,
+      "step": 2274
+    },
+    {
+      "epoch": 0.19949878670209048,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002901159576274721,
+      "loss": 1.2118,
+      "step": 2275
+    },
+    {
+      "epoch": 0.19958647847646505,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002901006395646744,
+      "loss": 1.226,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19967417025083958,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0029008531009302127,
+      "loss": 1.2068,
+      "step": 2277
+    },
+    {
+      "epoch": 0.19976186202521412,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0029006996921391074,
+      "loss": 1.1829,
+      "step": 2278
+    },
+    {
+      "epoch": 0.19984955379958866,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0029005461692874195,
+      "loss": 1.1688,
+      "step": 2279
+    },
+    {
+      "epoch": 0.19993724557396322,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029003925323891487,
+      "loss": 1.249,
+      "step": 2280
+    },
+    {
+      "epoch": 0.20002493734833776,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002900238781458307,
+      "loss": 1.2226,
+      "step": 2281
+    },
+    {
+      "epoch": 0.2001126291227123,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002900084916508916,
+      "loss": 1.2285,
+      "step": 2282
+    },
+    {
+      "epoch": 0.20020032089708686,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002899930937555008,
+      "loss": 1.1471,
+      "step": 2283
+    },
+    {
+      "epoch": 0.2002880126714614,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0028997768446106247,
+      "loss": 1.2034,
+      "step": 2284
+    },
+    {
+      "epoch": 0.20037570444583594,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002899622637689819,
+      "loss": 1.2634,
+      "step": 2285
+    },
+    {
+      "epoch": 0.20046339622021048,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0028994683168066557,
+      "loss": 1.2306,
+      "step": 2286
+    },
+    {
+      "epoch": 0.20055108799458504,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0028993138819752066,
+      "loss": 1.2843,
+      "step": 2287
+    },
+    {
+      "epoch": 0.20063877976895958,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.002899159333209557,
+      "loss": 1.2464,
+      "step": 2288
+    },
+    {
+      "epoch": 0.20072647154333412,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0028990046705238003,
+      "loss": 1.2667,
+      "step": 2289
+    },
+    {
+      "epoch": 0.20081416331770865,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0028988498939320428,
+      "loss": 1.2393,
+      "step": 2290
+    },
+    {
+      "epoch": 0.20090185509208322,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002898695003448398,
+      "loss": 1.2104,
+      "step": 2291
+    },
+    {
+      "epoch": 0.20098954686645776,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0028985399990869927,
+      "loss": 1.2468,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2010772386408323,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002898384880861963,
+      "loss": 1.1991,
+      "step": 2293
+    },
+    {
+      "epoch": 0.20116493041520686,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0028982296487874543,
+      "loss": 1.2474,
+      "step": 2294
+    },
+    {
+      "epoch": 0.2012526221895814,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0028980743028776245,
+      "loss": 1.2428,
+      "step": 2295
+    },
+    {
+      "epoch": 0.20134031396395594,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0028979188431466395,
+      "loss": 1.2533,
+      "step": 2296
+    },
+    {
+      "epoch": 0.20142800573833047,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002897763269608678,
+      "loss": 1.2392,
+      "step": 2297
+    },
+    {
+      "epoch": 0.20151569751270504,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0028976075822779272,
+      "loss": 1.2092,
+      "step": 2298
+    },
+    {
+      "epoch": 0.20160338928707958,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0028974517811685852,
+      "loss": 1.2441,
+      "step": 2299
+    },
+    {
+      "epoch": 0.2016910810614541,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002897295866294862,
+      "loss": 1.2282,
+      "step": 2300
+    },
+    {
+      "epoch": 0.20177877283582868,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0028971398376709746,
+      "loss": 1.2885,
+      "step": 2301
+    },
+    {
+      "epoch": 0.20186646461020322,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002896983695311154,
+      "loss": 1.1914,
+      "step": 2302
+    },
+    {
+      "epoch": 0.20195415638457775,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002896827439229639,
+      "loss": 1.1823,
+      "step": 2303
+    },
+    {
+      "epoch": 0.2020418481589523,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002896671069440681,
+      "loss": 1.1945,
+      "step": 2304
+    },
+    {
+      "epoch": 0.20212953993332686,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0028965145859585393,
+      "loss": 1.1994,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2022172317077014,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0028963579887974853,
+      "loss": 1.2515,
+      "step": 2306
+    },
+    {
+      "epoch": 0.20230492348207593,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0028962012779718002,
+      "loss": 1.1885,
+      "step": 2307
+    },
+    {
+      "epoch": 0.20239261525645047,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028960444534957756,
+      "loss": 1.1365,
+      "step": 2308
+    },
+    {
+      "epoch": 0.20248030703082504,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0028958875153837133,
+      "loss": 1.2862,
+      "step": 2309
+    },
+    {
+      "epoch": 0.20256799880519957,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002895730463649926,
+      "loss": 1.2135,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2026556905795741,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002895573298308736,
+      "loss": 1.1977,
+      "step": 2311
+    },
+    {
+      "epoch": 0.20274338235394868,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0028954160193744773,
+      "loss": 1.2087,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2028310741283232,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002895258626861492,
+      "loss": 1.2667,
+      "step": 2313
+    },
+    {
+      "epoch": 0.20291876590269775,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0028951011207841346,
+      "loss": 1.1602,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2030064576770723,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00289494350115677,
+      "loss": 1.2413,
+      "step": 2315
+    },
+    {
+      "epoch": 0.20309414945144685,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002894785767993771,
+      "loss": 1.2095,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2031818412258214,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0028946279213095233,
+      "loss": 1.2188,
+      "step": 2317
+    },
+    {
+      "epoch": 0.20326953300019593,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0028944699611184225,
+      "loss": 1.2537,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2033572247745705,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002894311887434874,
+      "loss": 1.2358,
+      "step": 2319
+    },
+    {
+      "epoch": 0.20344491654894503,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0028941537002732933,
+      "loss": 1.2175,
+      "step": 2320
+    },
+    {
+      "epoch": 0.20353260832331957,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002893995399648107,
+      "loss": 1.2213,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2036203000976941,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0028938369855737514,
+      "loss": 1.2282,
+      "step": 2322
+    },
+    {
+      "epoch": 0.20370799187206867,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002893678458064674,
+      "loss": 1.2173,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2037956836464432,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002893519817135332,
+      "loss": 1.2252,
+      "step": 2324
+    },
+    {
+      "epoch": 0.20388337542081775,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028933610628001922,
+      "loss": 1.2915,
+      "step": 2325
+    },
+    {
+      "epoch": 0.20397106719519228,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0028932021950737333,
+      "loss": 1.1885,
+      "step": 2326
+    },
+    {
+      "epoch": 0.20405875896956685,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0028930432139704435,
+      "loss": 1.2358,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2041464507439414,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0028928841195048215,
+      "loss": 1.1917,
+      "step": 2328
+    },
+    {
+      "epoch": 0.20423414251831593,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0028927249116913766,
+      "loss": 1.187,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2043218342926905,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0028925655905446277,
+      "loss": 1.1759,
+      "step": 2330
+    },
+    {
+      "epoch": 0.20440952606706503,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002892406156079104,
+      "loss": 1.2177,
+      "step": 2331
+    },
+    {
+      "epoch": 0.20449721784143957,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002892246608309347,
+      "loss": 1.1686,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2045849096158141,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028920869472499056,
+      "loss": 1.2278,
+      "step": 2333
+    },
+    {
+      "epoch": 0.20467260139018867,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0028919271729153416,
+      "loss": 1.2325,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2047602931645632,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002891767285320225,
+      "loss": 1.1896,
+      "step": 2335
+    },
+    {
+      "epoch": 0.20484798493893774,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0028916072844791378,
+      "loss": 1.3283,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2049356767133123,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002891447170406671,
+      "loss": 1.2072,
+      "step": 2337
+    },
+    {
+      "epoch": 0.20502336848768685,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0028912869431174278,
+      "loss": 1.2582,
+      "step": 2338
+    },
+    {
+      "epoch": 0.20511106026206138,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002891126602626019,
+      "loss": 1.2377,
+      "step": 2339
+    },
+    {
+      "epoch": 0.20519875203643592,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0028909661489470682,
+      "loss": 1.2506,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2052864438108105,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002890805582095209,
+      "loss": 1.1812,
+      "step": 2341
+    },
+    {
+      "epoch": 0.20537413558518502,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002890644902085083,
+      "loss": 1.2333,
+      "step": 2342
+    },
+    {
+      "epoch": 0.20546182735955956,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0028904841089313444,
+      "loss": 1.2214,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2055495191339341,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0028903232026486576,
+      "loss": 1.2293,
+      "step": 2344
+    },
+    {
+      "epoch": 0.20563721090830867,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0028901621832516967,
+      "loss": 1.2232,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2057249026826832,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0028900010507551463,
+      "loss": 1.2796,
+      "step": 2346
+    },
+    {
+      "epoch": 0.20581259445705774,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002889839805173701,
+      "loss": 1.2772,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2059002862314323,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002889678446522066,
+      "loss": 1.17,
+      "step": 2348
+    },
+    {
+      "epoch": 0.20598797800580684,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002889516974814957,
+      "loss": 1.1554,
+      "step": 2349
+    },
+    {
+      "epoch": 0.20607566978018138,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0028893553900670994,
+      "loss": 1.2177,
+      "step": 2350
+    },
+    {
+      "epoch": 0.20616336155455592,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00288919369229323,
+      "loss": 1.2588,
+      "step": 2351
+    },
+    {
+      "epoch": 0.20625105332893048,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0028890318815080946,
+      "loss": 1.2717,
+      "step": 2352
+    },
+    {
+      "epoch": 0.20633874510330502,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0028888699577264503,
+      "loss": 1.2739,
+      "step": 2353
+    },
+    {
+      "epoch": 0.20642643687767956,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002888707920963064,
+      "loss": 1.2117,
+      "step": 2354
+    },
+    {
+      "epoch": 0.20651412865205412,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0028885457712327133,
+      "loss": 1.2334,
+      "step": 2355
+    },
+    {
+      "epoch": 0.20660182042642866,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002888383508550185,
+      "loss": 1.2427,
+      "step": 2356
+    },
+    {
+      "epoch": 0.2066895122008032,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0028882211329302773,
+      "loss": 1.2333,
+      "step": 2357
+    },
+    {
+      "epoch": 0.20677720397517774,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0028880586443878,
+      "loss": 1.2249,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2068648957495523,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002887896042937569,
+      "loss": 1.2365,
+      "step": 2359
+    },
+    {
+      "epoch": 0.20695258752392684,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0028877333285944155,
+      "loss": 1.2562,
+      "step": 2360
+    },
+    {
+      "epoch": 0.20704027929830138,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028875705013731764,
+      "loss": 1.1434,
+      "step": 2361
+    },
+    {
+      "epoch": 0.20712797107267591,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0028874075612887037,
+      "loss": 1.2028,
+      "step": 2362
+    },
+    {
+      "epoch": 0.20721566284705048,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002887244508355855,
+      "loss": 1.158,
+      "step": 2363
+    },
+    {
+      "epoch": 0.20730335462142502,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0028870813425895016,
+      "loss": 1.1906,
+      "step": 2364
+    },
+    {
+      "epoch": 0.20739104639579956,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002886918064004523,
+      "loss": 1.2418,
+      "step": 2365
+    },
+    {
+      "epoch": 0.20747873817017412,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0028867546726158095,
+      "loss": 1.2573,
+      "step": 2366
+    },
+    {
+      "epoch": 0.20756642994454866,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028865911684382633,
+      "loss": 1.208,
+      "step": 2367
+    },
+    {
+      "epoch": 0.2076541217189232,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0028864275514867946,
+      "loss": 1.228,
+      "step": 2368
+    },
+    {
+      "epoch": 0.20774181349329773,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0028862638217763245,
+      "loss": 1.2732,
+      "step": 2369
+    },
+    {
+      "epoch": 0.2078295052676723,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0028860999793217856,
+      "loss": 1.1994,
+      "step": 2370
+    },
+    {
+      "epoch": 0.20791719704204684,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00288593602413812,
+      "loss": 1.2048,
+      "step": 2371
+    },
+    {
+      "epoch": 0.20800488881642137,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0028857719562402793,
+      "loss": 1.3161,
+      "step": 2372
+    },
+    {
+      "epoch": 0.20809258059079594,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002885607775643226,
+      "loss": 1.2285,
+      "step": 2373
+    },
+    {
+      "epoch": 0.20818027236517048,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0028854434823619335,
+      "loss": 1.2331,
+      "step": 2374
+    },
+    {
+      "epoch": 0.20826796413954501,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0028852790764113847,
+      "loss": 1.2045,
+      "step": 2375
+    },
+    {
+      "epoch": 0.20835565591391955,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002885114557806573,
+      "loss": 1.1751,
+      "step": 2376
+    },
+    {
+      "epoch": 0.20844334768829412,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002884949926562502,
+      "loss": 1.2258,
+      "step": 2377
+    },
+    {
+      "epoch": 0.20853103946266865,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0028847851826941853,
+      "loss": 1.1964,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2086187312370432,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0028846203262166484,
+      "loss": 1.2436,
+      "step": 2379
+    },
+    {
+      "epoch": 0.20870642301141773,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0028844553571449245,
+      "loss": 1.2333,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2087941147857923,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002884290275494059,
+      "loss": 1.2614,
+      "step": 2381
+    },
+    {
+      "epoch": 0.20888180656016683,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002884125081279106,
+      "loss": 1.1986,
+      "step": 2382
+    },
+    {
+      "epoch": 0.20896949833454137,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002883959774515132,
+      "loss": 1.2024,
+      "step": 2383
+    },
+    {
+      "epoch": 0.20905719010891594,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0028837943552172114,
+      "loss": 1.2517,
+      "step": 2384
+    },
+    {
+      "epoch": 0.20914488188329047,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.002883628823400431,
+      "loss": 1.1813,
+      "step": 2385
+    },
+    {
+      "epoch": 0.209232573657665,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002883463179079887,
+      "loss": 1.2123,
+      "step": 2386
+    },
+    {
+      "epoch": 0.20932026543203955,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0028832974222706845,
+      "loss": 1.1983,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2094079572064141,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002883131552987941,
+      "loss": 1.2064,
+      "step": 2388
+    },
+    {
+      "epoch": 0.20949564898078865,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0028829655712467837,
+      "loss": 1.206,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2095833407551632,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002882799477062349,
+      "loss": 1.208,
+      "step": 2390
+    },
+    {
+      "epoch": 0.20967103252953775,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002882633270449784,
+      "loss": 1.2076,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2097587243039123,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002882466951424247,
+      "loss": 1.2018,
+      "step": 2392
+    },
+    {
+      "epoch": 0.20984641607828683,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0028823005200009056,
+      "loss": 1.1516,
+      "step": 2393
+    },
+    {
+      "epoch": 0.20993410785266137,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002882133976194938,
+      "loss": 1.1705,
+      "step": 2394
+    },
+    {
+      "epoch": 0.21002179962703593,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0028819673200215327,
+      "loss": 1.2489,
+      "step": 2395
+    },
+    {
+      "epoch": 0.21010949140141047,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002881800551495888,
+      "loss": 1.1655,
+      "step": 2396
+    },
+    {
+      "epoch": 0.210197183175785,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002881633670633213,
+      "loss": 1.1966,
+      "step": 2397
+    },
+    {
+      "epoch": 0.21028487495015954,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002881466677448726,
+      "loss": 1.2546,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2103725667245341,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0028812995719576574,
+      "loss": 1.231,
+      "step": 2399
+    },
+    {
+      "epoch": 0.21046025849890865,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0028811323541752466,
+      "loss": 1.3084,
+      "step": 2400
+    },
+    {
+      "epoch": 0.21054795027328319,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0028809650241167434,
+      "loss": 1.2047,
+      "step": 2401
+    },
+    {
+      "epoch": 0.21063564204765775,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0028807975817974076,
+      "loss": 1.2099,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2107233338220323,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002880630027232509,
+      "loss": 1.221,
+      "step": 2403
+    },
+    {
+      "epoch": 0.21081102559640683,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0028804623604373296,
+      "loss": 1.2288,
+      "step": 2404
+    },
+    {
+      "epoch": 0.21089871737078136,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0028802945814271593,
+      "loss": 1.23,
+      "step": 2405
+    },
+    {
+      "epoch": 0.21098640914515593,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.002880126690217299,
+      "loss": 1.2156,
+      "step": 2406
+    },
+    {
+      "epoch": 0.21107410091953047,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00287995868682306,
+      "loss": 1.1559,
+      "step": 2407
+    },
+    {
+      "epoch": 0.211161792693905,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0028797905712597636,
+      "loss": 1.1611,
+      "step": 2408
+    },
+    {
+      "epoch": 0.21124948446827957,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0028796223435427425,
+      "loss": 1.2363,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2113371762426541,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.002879454003687337,
+      "loss": 1.2284,
+      "step": 2410
+    },
+    {
+      "epoch": 0.21142486801702864,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0028792855517089013,
+      "loss": 1.1585,
+      "step": 2411
+    },
+    {
+      "epoch": 0.21151255979140318,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0028791169876227958,
+      "loss": 1.2513,
+      "step": 2412
+    },
+    {
+      "epoch": 0.21160025156577775,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0028789483114443942,
+      "loss": 1.1754,
+      "step": 2413
+    },
+    {
+      "epoch": 0.21168794334015228,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002878779523189079,
+      "loss": 1.2413,
+      "step": 2414
+    },
+    {
+      "epoch": 0.21177563511452682,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0028786106228722436,
+      "loss": 1.2058,
+      "step": 2415
+    },
+    {
+      "epoch": 0.21186332688890136,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002878441610509291,
+      "loss": 1.2768,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21195101866327593,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002878272486115635,
+      "loss": 1.2236,
+      "step": 2417
+    },
+    {
+      "epoch": 0.21203871043765046,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002878103249706699,
+      "loss": 1.1947,
+      "step": 2418
+    },
+    {
+      "epoch": 0.212126402212025,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0028779339012979166,
+      "loss": 1.238,
+      "step": 2419
+    },
+    {
+      "epoch": 0.21221409398639957,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002877764440904732,
+      "loss": 1.1973,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2123017857607741,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0028775948685426,
+      "loss": 1.2345,
+      "step": 2421
+    },
+    {
+      "epoch": 0.21238947753514864,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0028774251842269855,
+      "loss": 1.2017,
+      "step": 2422
+    },
+    {
+      "epoch": 0.21247716930952318,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002877255387973362,
+      "loss": 1.2343,
+      "step": 2423
+    },
+    {
+      "epoch": 0.21256486108389774,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0028770854797972164,
+      "loss": 1.2073,
+      "step": 2424
+    },
+    {
+      "epoch": 0.21265255285827228,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002876915459714042,
+      "loss": 1.1852,
+      "step": 2425
+    },
+    {
+      "epoch": 0.21274024463264682,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0028767453277393448,
+      "loss": 1.2246,
+      "step": 2426
+    },
+    {
+      "epoch": 0.21282793640702138,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0028765750838886414,
+      "loss": 1.1585,
+      "step": 2427
+    },
+    {
+      "epoch": 0.21291562818139592,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0028764047281774558,
+      "loss": 1.1536,
+      "step": 2428
+    },
+    {
+      "epoch": 0.21300331995577046,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0028762342606213254,
+      "loss": 1.2755,
+      "step": 2429
+    },
+    {
+      "epoch": 0.213091011730145,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0028760636812357955,
+      "loss": 1.2515,
+      "step": 2430
+    },
+    {
+      "epoch": 0.21317870350451956,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0028758929900364236,
+      "loss": 1.2335,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2132663952788941,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002875722187038775,
+      "loss": 1.2292,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21335408705326864,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0028755512722584277,
+      "loss": 1.2493,
+      "step": 2433
+    },
+    {
+      "epoch": 0.21344177882764317,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002875380245710968,
+      "loss": 1.2574,
+      "step": 2434
+    },
+    {
+      "epoch": 0.21352947060201774,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002875209107411993,
+      "loss": 1.3211,
+      "step": 2435
+    },
+    {
+      "epoch": 0.21361716237639228,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0028750378573771103,
+      "loss": 1.1894,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21370485415076682,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0028748664956219374,
+      "loss": 1.1815,
+      "step": 2437
+    },
+    {
+      "epoch": 0.21379254592514138,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002874695022162102,
+      "loss": 1.2064,
+      "step": 2438
+    },
+    {
+      "epoch": 0.21388023769951592,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0028745234370132424,
+      "loss": 1.2104,
+      "step": 2439
+    },
+    {
+      "epoch": 0.21396792947389046,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0028743517401910064,
+      "loss": 1.2221,
+      "step": 2440
+    },
+    {
+      "epoch": 0.214055621248265,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0028741799317110518,
+      "loss": 1.2286,
+      "step": 2441
+    },
+    {
+      "epoch": 0.21414331302263956,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002874008011589048,
+      "loss": 1.2943,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2142310047970141,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0028738359798406734,
+      "loss": 1.2198,
+      "step": 2443
+    },
+    {
+      "epoch": 0.21431869657138863,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0028736638364816164,
+      "loss": 1.2397,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2144063883457632,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0028734915815275767,
+      "loss": 1.2015,
+      "step": 2445
+    },
+    {
+      "epoch": 0.21449408012013774,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0028733192149942633,
+      "loss": 1.245,
+      "step": 2446
+    },
+    {
+      "epoch": 0.21458177189451227,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028731467368973956,
+      "loss": 1.2822,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2146694636688868,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0028729741472527026,
+      "loss": 1.243,
+      "step": 2448
+    },
+    {
+      "epoch": 0.21475715544326138,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0028728014460759244,
+      "loss": 1.2323,
+      "step": 2449
+    },
+    {
+      "epoch": 0.21484484721763591,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0028726286333828114,
+      "loss": 1.2198,
+      "step": 2450
+    },
+    {
+      "epoch": 0.21493253899201045,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002872455709189123,
+      "loss": 1.2101,
+      "step": 2451
+    },
+    {
+      "epoch": 0.215020230766385,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00287228267351063,
+      "loss": 1.2838,
+      "step": 2452
+    },
+    {
+      "epoch": 0.21510792254075956,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002872109526363112,
+      "loss": 1.2129,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2151956143151341,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0028719362677623603,
+      "loss": 1.1734,
+      "step": 2454
+    },
+    {
+      "epoch": 0.21528330608950863,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0028717628977241753,
+      "loss": 1.244,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2153709978638832,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0028715894162643675,
+      "loss": 1.2126,
+      "step": 2456
+    },
+    {
+      "epoch": 0.21545868963825773,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0028714158233987593,
+      "loss": 1.2491,
+      "step": 2457
+    },
+    {
+      "epoch": 0.21554638141263227,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002871242119143181,
+      "loss": 1.1801,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2156340731870068,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.002871068303513474,
+      "loss": 1.2637,
+      "step": 2459
+    },
+    {
+      "epoch": 0.21572176496138137,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0028708943765254893,
+      "loss": 1.2972,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2158094567357559,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0028707203381950893,
+      "loss": 1.2826,
+      "step": 2461
+    },
+    {
+      "epoch": 0.21589714851013045,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0028705461885381463,
+      "loss": 1.2108,
+      "step": 2462
+    },
+    {
+      "epoch": 0.21598484028450501,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0028703719275705412,
+      "loss": 1.1843,
+      "step": 2463
+    },
+    {
+      "epoch": 0.21607253205887955,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002870197555308167,
+      "loss": 1.2419,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2161602238332541,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0028700230717669255,
+      "loss": 1.2047,
+      "step": 2465
+    },
+    {
+      "epoch": 0.21624791560762863,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0028698484769627297,
+      "loss": 1.2018,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2163356073820032,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0028696737709115014,
+      "loss": 1.2248,
+      "step": 2467
+    },
+    {
+      "epoch": 0.21642329915637773,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002869498953629174,
+      "loss": 1.2404,
+      "step": 2468
+    },
+    {
+      "epoch": 0.21651099093075227,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00286932402513169,
+      "loss": 1.2839,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2165986827051268,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002869148985435003,
+      "loss": 1.1919,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21668637447950137,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002868973834555075,
+      "loss": 1.2441,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2167740662538759,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0028687985725078806,
+      "loss": 1.2368,
+      "step": 2472
+    },
+    {
+      "epoch": 0.21686175802825045,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0028686231993094026,
+      "loss": 1.1671,
+      "step": 2473
+    },
+    {
+      "epoch": 0.216949449802625,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0028684477149756344,
+      "loss": 1.2183,
+      "step": 2474
+    },
+    {
+      "epoch": 0.21703714157699955,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0028682721195225805,
+      "loss": 1.2049,
+      "step": 2475
+    },
+    {
+      "epoch": 0.21712483335137409,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0028680964129662537,
+      "loss": 1.191,
+      "step": 2476
+    },
+    {
+      "epoch": 0.21721252512574862,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002867920595322679,
+      "loss": 1.1428,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2173002169001232,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00286774466660789,
+      "loss": 1.1859,
+      "step": 2478
+    },
+    {
+      "epoch": 0.21738790867449773,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0028675686268379305,
+      "loss": 1.1801,
+      "step": 2479
+    },
+    {
+      "epoch": 0.21747560044887226,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002867392476028856,
+      "loss": 1.2237,
+      "step": 2480
+    },
+    {
+      "epoch": 0.21756329222324683,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00286721621419673,
+      "loss": 1.1939,
+      "step": 2481
+    },
+    {
+      "epoch": 0.21765098399762137,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0028670398413576277,
+      "loss": 1.2146,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2177386757719959,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0028668633575276332,
+      "loss": 1.2449,
+      "step": 2483
+    },
+    {
+      "epoch": 0.21782636754637044,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002866686762722842,
+      "loss": 1.2002,
+      "step": 2484
+    },
+    {
+      "epoch": 0.217914059320745,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0028665100569593596,
+      "loss": 1.2648,
+      "step": 2485
+    },
+    {
+      "epoch": 0.21800175109511954,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0028663332402533,
+      "loss": 1.2024,
+      "step": 2486
+    },
+    {
+      "epoch": 0.21808944286949408,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0028661563126207885,
+      "loss": 1.2726,
+      "step": 2487
+    },
+    {
+      "epoch": 0.21817713464386862,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0028659792740779607,
+      "loss": 1.2363,
+      "step": 2488
+    },
+    {
+      "epoch": 0.21826482641824319,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0028658021246409627,
+      "loss": 1.2317,
+      "step": 2489
+    },
+    {
+      "epoch": 0.21835251819261772,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002865624864325949,
+      "loss": 1.2804,
+      "step": 2490
+    },
+    {
+      "epoch": 0.21844020996699226,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0028654474931490864,
+      "loss": 1.19,
+      "step": 2491
+    },
+    {
+      "epoch": 0.21852790174136683,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0028652700111265494,
+      "loss": 1.2135,
+      "step": 2492
+    },
+    {
+      "epoch": 0.21861559351574136,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002865092418274525,
+      "loss": 1.2374,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2187032852901159,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0028649147146092087,
+      "loss": 1.2551,
+      "step": 2494
+    },
+    {
+      "epoch": 0.21879097706449044,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0028647369001468066,
+      "loss": 1.2747,
+      "step": 2495
+    },
+    {
+      "epoch": 0.218878668838865,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002864558974903535,
+      "loss": 1.2038,
+      "step": 2496
+    },
+    {
+      "epoch": 0.21896636061323954,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0028643809388956202,
+      "loss": 1.2603,
+      "step": 2497
+    },
+    {
+      "epoch": 0.21905405238761408,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0028642027921392996,
+      "loss": 1.225,
+      "step": 2498
+    },
+    {
+      "epoch": 0.21914174416198862,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0028640245346508177,
+      "loss": 1.2936,
+      "step": 2499
+    },
+    {
+      "epoch": 0.21922943593636318,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028638461664464328,
+      "loss": 1.2442,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21922943593636318,
+      "eval_loss": 1.2314904928207397,
+      "eval_runtime": 428.341,
+      "eval_samples_per_second": 33.728,
+      "eval_steps_per_second": 8.433,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21931712771073772,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002863667687542411,
+      "loss": 1.2263,
+      "step": 2501
+    },
+    {
+      "epoch": 0.21940481948511226,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0028634890979550285,
+      "loss": 1.2238,
+      "step": 2502
+    },
+    {
+      "epoch": 0.21949251125948682,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002863310397700574,
+      "loss": 1.1454,
+      "step": 2503
+    },
+    {
+      "epoch": 0.21958020303386136,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0028631315867953424,
+      "loss": 1.1863,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2196678948082359,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0028629526652556424,
+      "loss": 1.2242,
+      "step": 2505
+    },
+    {
+      "epoch": 0.21975558658261043,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0028627736330977902,
+      "loss": 1.2066,
+      "step": 2506
+    },
+    {
+      "epoch": 0.219843278356985,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0028625944903381134,
+      "loss": 1.1913,
+      "step": 2507
+    },
+    {
+      "epoch": 0.21993097013135954,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0028624152369929492,
+      "loss": 1.2735,
+      "step": 2508
+    },
+    {
+      "epoch": 0.22001866190573408,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0028622358730786457,
+      "loss": 1.1731,
+      "step": 2509
+    },
+    {
+      "epoch": 0.22010635368010864,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00286205639861156,
+      "loss": 1.1977,
+      "step": 2510
+    },
+    {
+      "epoch": 0.22019404545448318,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0028618768136080596,
+      "loss": 1.2318,
+      "step": 2511
+    },
+    {
+      "epoch": 0.22028173722885772,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0028616971180845216,
+      "loss": 1.2453,
+      "step": 2512
+    },
+    {
+      "epoch": 0.22036942900323225,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002861517312057335,
+      "loss": 1.2301,
+      "step": 2513
+    },
+    {
+      "epoch": 0.22045712077760682,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028613373955428964,
+      "loss": 1.238,
+      "step": 2514
+    },
+    {
+      "epoch": 0.22054481255198136,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002861157368557615,
+      "loss": 1.2046,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2206325043263559,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0028609772311179085,
+      "loss": 1.2274,
+      "step": 2516
+    },
+    {
+      "epoch": 0.22072019610073043,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002860796983240204,
+      "loss": 1.1724,
+      "step": 2517
+    },
+    {
+      "epoch": 0.220807887875105,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0028606166249409404,
+      "loss": 1.2101,
+      "step": 2518
+    },
+    {
+      "epoch": 0.22089557964947953,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002860436156236566,
+      "loss": 1.1699,
+      "step": 2519
+    },
+    {
+      "epoch": 0.22098327142385407,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0028602555771435386,
+      "loss": 1.226,
+      "step": 2520
+    },
+    {
+      "epoch": 0.22107096319822864,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002860074887678327,
+      "loss": 1.2821,
+      "step": 2521
+    },
+    {
+      "epoch": 0.22115865497260317,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0028598940878574097,
+      "loss": 1.1798,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2212463467469777,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0028597131776972743,
+      "loss": 1.2217,
+      "step": 2523
+    },
+    {
+      "epoch": 0.22133403852135225,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0028595321572144204,
+      "loss": 1.2743,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22142173029572682,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002859351026425356,
+      "loss": 1.1935,
+      "step": 2525
+    },
+    {
+      "epoch": 0.22150942207010135,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0028591697853466,
+      "loss": 1.2026,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2215971138444759,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002858988433994681,
+      "loss": 1.2264,
+      "step": 2527
+    },
+    {
+      "epoch": 0.22168480561885046,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002858806972386138,
+      "loss": 1.1948,
+      "step": 2528
+    },
+    {
+      "epoch": 0.221772497393225,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0028586254005375194,
+      "loss": 1.2117,
+      "step": 2529
+    },
+    {
+      "epoch": 0.22186018916759953,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0028584437184653844,
+      "loss": 1.247,
+      "step": 2530
+    },
+    {
+      "epoch": 0.22194788094197407,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0028582619261863017,
+      "loss": 1.1657,
+      "step": 2531
+    },
+    {
+      "epoch": 0.22203557271634863,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0028580800237168504,
+      "loss": 1.2557,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22212326449072317,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0028578980110736197,
+      "loss": 1.1681,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2222109562650977,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0028577158882732087,
+      "loss": 1.2274,
+      "step": 2534
+    },
+    {
+      "epoch": 0.22229864803947225,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0028575336553322266,
+      "loss": 1.2279,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2223863398138468,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002857351312267292,
+      "loss": 1.2491,
+      "step": 2536
+    },
+    {
+      "epoch": 0.22247403158822135,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0028571688590950345,
+      "loss": 1.2769,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2225617233625959,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0028569862958320934,
+      "loss": 1.1647,
+      "step": 2538
+    },
+    {
+      "epoch": 0.22264941513697045,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0028568036224951177,
+      "loss": 1.2185,
+      "step": 2539
+    },
+    {
+      "epoch": 0.222737106911345,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0028566208391007674,
+      "loss": 1.2816,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22282479868571953,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002856437945665711,
+      "loss": 1.2133,
+      "step": 2541
+    },
+    {
+      "epoch": 0.22291249046009406,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002856254942206629,
+      "loss": 1.24,
+      "step": 2542
+    },
+    {
+      "epoch": 0.22300018223446863,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0028560718287402097,
+      "loss": 1.1865,
+      "step": 2543
+    },
+    {
+      "epoch": 0.22308787400884317,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002855888605283153,
+      "loss": 1.2732,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2231755657832177,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002855705271852169,
+      "loss": 1.179,
+      "step": 2545
+    },
+    {
+      "epoch": 0.22326325755759227,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002855521828463976,
+      "loss": 1.211,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2233509493319668,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002855338275135305,
+      "loss": 1.1939,
+      "step": 2547
+    },
+    {
+      "epoch": 0.22343864110634135,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002855154611882894,
+      "loss": 1.2526,
+      "step": 2548
+    },
+    {
+      "epoch": 0.22352633288071588,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0028549708387234944,
+      "loss": 1.267,
+      "step": 2549
+    },
+    {
+      "epoch": 0.22361402465509045,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0028547869556738645,
+      "loss": 1.2507,
+      "step": 2550
+    },
+    {
+      "epoch": 0.223701716429465,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0028546029627507744,
+      "loss": 1.256,
+      "step": 2551
+    },
+    {
+      "epoch": 0.22378940820383952,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0028544188599710038,
+      "loss": 1.1852,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22387709997821406,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0028542346473513424,
+      "loss": 1.2165,
+      "step": 2553
+    },
+    {
+      "epoch": 0.22396479175258863,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00285405032490859,
+      "loss": 1.154,
+      "step": 2554
+    },
+    {
+      "epoch": 0.22405248352696316,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0028538658926595558,
+      "loss": 1.1535,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2241401753013377,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0028536813506210602,
+      "loss": 1.2141,
+      "step": 2556
+    },
+    {
+      "epoch": 0.22422786707571227,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028534966988099327,
+      "loss": 1.2095,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2243155588500868,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002853311937243013,
+      "loss": 1.2091,
+      "step": 2558
+    },
+    {
+      "epoch": 0.22440325062446134,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002853127065937151,
+      "loss": 1.2521,
+      "step": 2559
+    },
+    {
+      "epoch": 0.22449094239883588,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0028529420849092066,
+      "loss": 1.2062,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22457863417321045,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0028527569941760493,
+      "loss": 1.17,
+      "step": 2561
+    },
+    {
+      "epoch": 0.22466632594758498,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002852571793754559,
+      "loss": 1.204,
+      "step": 2562
+    },
+    {
+      "epoch": 0.22475401772195952,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0028523864836616257,
+      "loss": 1.207,
+      "step": 2563
+    },
+    {
+      "epoch": 0.22484170949633409,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002852201063914149,
+      "loss": 1.1967,
+      "step": 2564
+    },
+    {
+      "epoch": 0.22492940127070862,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002852015534529039,
+      "loss": 1.1718,
+      "step": 2565
+    },
+    {
+      "epoch": 0.22501709304508316,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0028518298955232148,
+      "loss": 1.2645,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2251047848194577,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0028516441469136067,
+      "loss": 1.242,
+      "step": 2567
+    },
+    {
+      "epoch": 0.22519247659383226,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002851458288717154,
+      "loss": 1.2146,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2252801683682068,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028512723209508078,
+      "loss": 1.2159,
+      "step": 2569
+    },
+    {
+      "epoch": 0.22536786014258134,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002851086243631527,
+      "loss": 1.2566,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22545555191695588,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0028509000567762807,
+      "loss": 1.2297,
+      "step": 2571
+    },
+    {
+      "epoch": 0.22554324369133044,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00285071376040205,
+      "loss": 1.2149,
+      "step": 2572
+    },
+    {
+      "epoch": 0.22563093546570498,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002850527354525823,
+      "loss": 1.1809,
+      "step": 2573
+    },
+    {
+      "epoch": 0.22571862724007952,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002850340839164601,
+      "loss": 1.1465,
+      "step": 2574
+    },
+    {
+      "epoch": 0.22580631901445408,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002850154214335393,
+      "loss": 1.1898,
+      "step": 2575
+    },
+    {
+      "epoch": 0.22589401078882862,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002849967480055219,
+      "loss": 1.1575,
+      "step": 2576
+    },
+    {
+      "epoch": 0.22598170256320316,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0028497806363411084,
+      "loss": 1.2273,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2260693943375777,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002849593683210101,
+      "loss": 1.215,
+      "step": 2578
+    },
+    {
+      "epoch": 0.22615708611195226,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0028494066206792464,
+      "loss": 1.249,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2262447778863268,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0028492194487656046,
+      "loss": 1.2699,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22633246966070134,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0028490321674862445,
+      "loss": 1.2158,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2264201614350759,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0028488447768582463,
+      "loss": 1.2098,
+      "step": 2582
+    },
+    {
+      "epoch": 0.22650785320945044,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0028486572768986985,
+      "loss": 1.2614,
+      "step": 2583
+    },
+    {
+      "epoch": 0.22659554498382498,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0028484696676247022,
+      "loss": 1.2308,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2266832367581995,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0028482819490533657,
+      "loss": 1.2501,
+      "step": 2585
+    },
+    {
+      "epoch": 0.22677092853257408,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0028480941212018084,
+      "loss": 1.2196,
+      "step": 2586
+    },
+    {
+      "epoch": 0.22685862030694862,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0028479061840871608,
+      "loss": 1.2298,
+      "step": 2587
+    },
+    {
+      "epoch": 0.22694631208132315,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0028477181377265614,
+      "loss": 1.2795,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2270340038556977,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00284752998213716,
+      "loss": 1.2464,
+      "step": 2589
+    },
+    {
+      "epoch": 0.22712169563007226,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002847341717336115,
+      "loss": 1.2755,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2272093874044468,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0028471533433405967,
+      "loss": 1.1613,
+      "step": 2591
+    },
+    {
+      "epoch": 0.22729707917882133,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.002846964860167784,
+      "loss": 1.2249,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2273847709531959,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0028467762678348655,
+      "loss": 1.1909,
+      "step": 2593
+    },
+    {
+      "epoch": 0.22747246272757043,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002846587566359041,
+      "loss": 1.1888,
+      "step": 2594
+    },
+    {
+      "epoch": 0.22756015450194497,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0028463987557575195,
+      "loss": 1.2059,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2276478462763195,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0028462098360475204,
+      "loss": 1.2286,
+      "step": 2596
+    },
+    {
+      "epoch": 0.22773553805069408,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0028460208072462715,
+      "loss": 1.2356,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2278232298250686,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002845831669371013,
+      "loss": 1.1843,
+      "step": 2598
+    },
+    {
+      "epoch": 0.22791092159944315,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0028456424224389933,
+      "loss": 1.3018,
+      "step": 2599
+    },
+    {
+      "epoch": 0.22799861337381772,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0028454530664674712,
+      "loss": 1.213,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22808630514819225,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0028452636014737152,
+      "loss": 1.2077,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2281739969225668,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002845074027475005,
+      "loss": 1.1707,
+      "step": 2602
+    },
+    {
+      "epoch": 0.22826168869694133,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0028448843444886275,
+      "loss": 1.2423,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2283493804713159,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002844694552531883,
+      "loss": 1.1616,
+      "step": 2604
+    },
+    {
+      "epoch": 0.22843707224569043,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0028445046516220795,
+      "loss": 1.2025,
+      "step": 2605
+    },
+    {
+      "epoch": 0.22852476402006497,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0028443146417765353,
+      "loss": 1.2347,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2286124557944395,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0028441245230125785,
+      "loss": 1.245,
+      "step": 2607
+    },
+    {
+      "epoch": 0.22870014756881407,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002843934295347548,
+      "loss": 1.2798,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2287878393431886,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002843743958798792,
+      "loss": 1.2723,
+      "step": 2609
+    },
+    {
+      "epoch": 0.22887553111756315,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0028435535133836684,
+      "loss": 1.2685,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2289632228919377,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028433629591195454,
+      "loss": 1.1875,
+      "step": 2611
+    },
+    {
+      "epoch": 0.22905091466631225,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0028431722960238015,
+      "loss": 1.2832,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2291386064406868,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002842981524113825,
+      "loss": 1.1799,
+      "step": 2613
+    },
+    {
+      "epoch": 0.22922629821506132,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002842790643407012,
+      "loss": 1.2775,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2293139899894359,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0028425996539207716,
+      "loss": 1.2276,
+      "step": 2615
+    },
+    {
+      "epoch": 0.22940168176381043,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0028424085556725217,
+      "loss": 1.2903,
+      "step": 2616
+    },
+    {
+      "epoch": 0.22948937353818497,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00284221734867969,
+      "loss": 1.1447,
+      "step": 2617
+    },
+    {
+      "epoch": 0.22957706531255953,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0028420260329597136,
+      "loss": 1.2324,
+      "step": 2618
+    },
+    {
+      "epoch": 0.22966475708693407,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0028418346085300407,
+      "loss": 1.2355,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2297524488613086,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0028416430754081273,
+      "loss": 1.2618,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22984014063568314,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002841451433611442,
+      "loss": 1.2654,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2299278324100577,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0028412596831574616,
+      "loss": 1.2563,
+      "step": 2622
+    },
+    {
+      "epoch": 0.23001552418443225,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0028410678240636737,
+      "loss": 1.2498,
+      "step": 2623
+    },
+    {
+      "epoch": 0.23010321595880678,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0028408758563475745,
+      "loss": 1.2915,
+      "step": 2624
+    },
+    {
+      "epoch": 0.23019090773318132,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002840683780026672,
+      "loss": 1.1496,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2302785995075559,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0028404915951184824,
+      "loss": 1.1934,
+      "step": 2626
+    },
+    {
+      "epoch": 0.23036629128193042,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0028402993016405317,
+      "loss": 1.2236,
+      "step": 2627
+    },
+    {
+      "epoch": 0.23045398305630496,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0028401068996103582,
+      "loss": 1.261,
+      "step": 2628
+    },
+    {
+      "epoch": 0.23054167483067953,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0028399143890455074,
+      "loss": 1.2673,
+      "step": 2629
+    },
+    {
+      "epoch": 0.23062936660505406,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0028397217699635364,
+      "loss": 1.1613,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2307170583794286,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002839529042382011,
+      "loss": 1.2489,
+      "step": 2631
+    },
+    {
+      "epoch": 0.23080475015380314,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002839336206318508,
+      "loss": 1.1916,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2308924419281777,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0028391432617906123,
+      "loss": 1.2008,
+      "step": 2633
+    },
+    {
+      "epoch": 0.23098013370255224,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002838950208815922,
+      "loss": 1.2168,
+      "step": 2634
+    },
+    {
+      "epoch": 0.23106782547692678,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0028387570474120412,
+      "loss": 1.2238,
+      "step": 2635
+    },
+    {
+      "epoch": 0.23115551725130135,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0028385637775965866,
+      "loss": 1.2041,
+      "step": 2636
+    },
+    {
+      "epoch": 0.23124320902567588,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002838370399387184,
+      "loss": 1.2188,
+      "step": 2637
+    },
+    {
+      "epoch": 0.23133090080005042,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.002838176912801468,
+      "loss": 1.1179,
+      "step": 2638
+    },
+    {
+      "epoch": 0.23141859257442496,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002837983317857086,
+      "loss": 1.257,
+      "step": 2639
+    },
+    {
+      "epoch": 0.23150628434879952,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0028377896145716916,
+      "loss": 1.2597,
+      "step": 2640
+    },
+    {
+      "epoch": 0.23159397612317406,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0028375958029629505,
+      "loss": 1.2147,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2316816678975486,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0028374018830485377,
+      "loss": 1.1762,
+      "step": 2642
+    },
+    {
+      "epoch": 0.23176935967192314,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0028372078548461388,
+      "loss": 1.2334,
+      "step": 2643
+    },
+    {
+      "epoch": 0.2318570514462977,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002837013718373448,
+      "loss": 1.1568,
+      "step": 2644
+    },
+    {
+      "epoch": 0.23194474322067224,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0028368194736481706,
+      "loss": 1.2559,
+      "step": 2645
+    },
+    {
+      "epoch": 0.23203243499504678,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002836625120688021,
+      "loss": 1.2163,
+      "step": 2646
+    },
+    {
+      "epoch": 0.23212012676942134,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.002836430659510724,
+      "loss": 1.26,
+      "step": 2647
+    },
+    {
+      "epoch": 0.23220781854379588,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002836236090134013,
+      "loss": 1.1791,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23229551031817042,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002836041412575633,
+      "loss": 1.2183,
+      "step": 2649
+    },
+    {
+      "epoch": 0.23238320209254495,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0028358466268533383,
+      "loss": 1.2223,
+      "step": 2650
+    },
+    {
+      "epoch": 0.23247089386691952,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0028356517329848918,
+      "loss": 1.2644,
+      "step": 2651
+    },
+    {
+      "epoch": 0.23255858564129406,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0028354567309880683,
+      "loss": 1.2642,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2326462774156686,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002835261620880651,
+      "loss": 1.1787,
+      "step": 2653
+    },
+    {
+      "epoch": 0.23273396919004316,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002835066402680434,
+      "loss": 1.254,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2328216609644177,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00283487107640522,
+      "loss": 1.2245,
+      "step": 2655
+    },
+    {
+      "epoch": 0.23290935273879224,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002834675642072823,
+      "loss": 1.1839,
+      "step": 2656
+    },
+    {
+      "epoch": 0.23299704451316677,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0028344800997010654,
+      "loss": 1.2421,
+      "step": 2657
+    },
+    {
+      "epoch": 0.23308473628754134,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00283428444930778,
+      "loss": 1.2176,
+      "step": 2658
+    },
+    {
+      "epoch": 0.23317242806191588,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0028340886909108106,
+      "loss": 1.2124,
+      "step": 2659
+    },
+    {
+      "epoch": 0.2332601198362904,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.002833892824528009,
+      "loss": 1.2466,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23334781161066495,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002833696850177238,
+      "loss": 1.2166,
+      "step": 2661
+    },
+    {
+      "epoch": 0.23343550338503952,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00283350076787637,
+      "loss": 1.203,
+      "step": 2662
+    },
+    {
+      "epoch": 0.23352319515941405,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0028333045776432874,
+      "loss": 1.254,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2336108869337886,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002833108279495882,
+      "loss": 1.2022,
+      "step": 2664
+    },
+    {
+      "epoch": 0.23369857870816316,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0028329118734520555,
+      "loss": 1.2903,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2337862704825377,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028327153595297198,
+      "loss": 1.269,
+      "step": 2666
+    },
+    {
+      "epoch": 0.23387396225691223,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002832518737746797,
+      "loss": 1.129,
+      "step": 2667
+    },
+    {
+      "epoch": 0.23396165403128677,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0028323220081212168,
+      "loss": 1.2234,
+      "step": 2668
+    },
+    {
+      "epoch": 0.23404934580566134,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0028321251706709223,
+      "loss": 1.2084,
+      "step": 2669
+    },
+    {
+      "epoch": 0.23413703758003587,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0028319282254138643,
+      "loss": 1.1985,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2342247293544104,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0028317311723680023,
+      "loss": 1.169,
+      "step": 2671
+    },
+    {
+      "epoch": 0.23431242112878498,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0028315340115513087,
+      "loss": 1.2108,
+      "step": 2672
+    },
+    {
+      "epoch": 0.2344001129031595,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002831336742981763,
+      "loss": 1.2387,
+      "step": 2673
+    },
+    {
+      "epoch": 0.23448780467753405,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0028311393666773563,
+      "loss": 1.1739,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2345754964519086,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0028309418826560884,
+      "loss": 1.238,
+      "step": 2675
+    },
+    {
+      "epoch": 0.23466318822628315,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0028307442909359694,
+      "loss": 1.2088,
+      "step": 2676
+    },
+    {
+      "epoch": 0.2347508800006577,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002830546591535019,
+      "loss": 1.2301,
+      "step": 2677
+    },
+    {
+      "epoch": 0.23483857177503223,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002830348784471267,
+      "loss": 1.2532,
+      "step": 2678
+    },
+    {
+      "epoch": 0.23492626354940677,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0028301508697627527,
+      "loss": 1.2274,
+      "step": 2679
+    },
+    {
+      "epoch": 0.23501395532378133,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0028299528474275263,
+      "loss": 1.2825,
+      "step": 2680
+    },
+    {
+      "epoch": 0.23510164709815587,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0028297547174836457,
+      "loss": 1.1979,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2351893388725304,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028295564799491807,
+      "loss": 1.2234,
+      "step": 2682
+    },
+    {
+      "epoch": 0.23527703064690497,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0028293581348422096,
+      "loss": 1.1451,
+      "step": 2683
+    },
+    {
+      "epoch": 0.2353647224212795,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002829159682180821,
+      "loss": 1.1801,
+      "step": 2684
+    },
+    {
+      "epoch": 0.23545241419565405,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0028289611219831133,
+      "loss": 1.1826,
+      "step": 2685
+    },
+    {
+      "epoch": 0.23554010597002858,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002828762454267195,
+      "loss": 1.2583,
+      "step": 2686
+    },
+    {
+      "epoch": 0.23562779774440315,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002828563679051184,
+      "loss": 1.2453,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2357154895187777,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002828364796353207,
+      "loss": 1.2039,
+      "step": 2688
+    },
+    {
+      "epoch": 0.23580318129315223,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0028281658061914032,
+      "loss": 1.1532,
+      "step": 2689
+    },
+    {
+      "epoch": 0.2358908730675268,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0028279667085839197,
+      "loss": 1.2449,
+      "step": 2690
+    },
+    {
+      "epoch": 0.23597856484190133,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0028277675035489128,
+      "loss": 1.1931,
+      "step": 2691
+    },
+    {
+      "epoch": 0.23606625661627587,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0028275681911045496,
+      "loss": 1.1981,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2361539483906504,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.002827368771269008,
+      "loss": 1.2245,
+      "step": 2693
+    },
+    {
+      "epoch": 0.23624164016502497,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0028271692440604733,
+      "loss": 1.2685,
+      "step": 2694
+    },
+    {
+      "epoch": 0.2363293319393995,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0028269696094971423,
+      "loss": 1.1999,
+      "step": 2695
+    },
+    {
+      "epoch": 0.23641702371377404,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0028267698675972217,
+      "loss": 1.2199,
+      "step": 2696
+    },
+    {
+      "epoch": 0.23650471548814858,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002826570018378927,
+      "loss": 1.1412,
+      "step": 2697
+    },
+    {
+      "epoch": 0.23659240726252315,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0028263700618604832,
+      "loss": 1.2175,
+      "step": 2698
+    },
+    {
+      "epoch": 0.23668009903689768,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0028261699980601275,
+      "loss": 1.2005,
+      "step": 2699
+    },
+    {
+      "epoch": 0.23676779081127222,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0028259698269961033,
+      "loss": 1.1881,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2368554825856468,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0028257695486866674,
+      "loss": 1.2221,
+      "step": 2701
+    },
+    {
+      "epoch": 0.23694317436002132,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0028255691631500837,
+      "loss": 1.229,
+      "step": 2702
+    },
+    {
+      "epoch": 0.23703086613439586,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0028253686704046268,
+      "loss": 1.2126,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2371185579087704,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0028251680704685814,
+      "loss": 1.2506,
+      "step": 2704
+    },
+    {
+      "epoch": 0.23720624968314497,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002824967363360242,
+      "loss": 1.1674,
+      "step": 2705
+    },
+    {
+      "epoch": 0.2372939414575195,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002824766549097912,
+      "loss": 1.2024,
+      "step": 2706
+    },
+    {
+      "epoch": 0.23738163323189404,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0028245656276999053,
+      "loss": 1.1874,
+      "step": 2707
+    },
+    {
+      "epoch": 0.23746932500626858,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028243645991845455,
+      "loss": 1.2237,
+      "step": 2708
+    },
+    {
+      "epoch": 0.23755701678064314,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002824163463570166,
+      "loss": 1.216,
+      "step": 2709
+    },
+    {
+      "epoch": 0.23764470855501768,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0028239622208751096,
+      "loss": 1.2144,
+      "step": 2710
+    },
+    {
+      "epoch": 0.23773240032939222,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002823760871117729,
+      "loss": 1.3082,
+      "step": 2711
+    },
+    {
+      "epoch": 0.23782009210376678,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0028235594143163874,
+      "loss": 1.2421,
+      "step": 2712
+    },
+    {
+      "epoch": 0.23790778387814132,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0028233578504894565,
+      "loss": 1.2216,
+      "step": 2713
+    },
+    {
+      "epoch": 0.23799547565251586,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0028231561796553187,
+      "loss": 1.2552,
+      "step": 2714
+    },
+    {
+      "epoch": 0.2380831674268904,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0028229544018323655,
+      "loss": 1.2286,
+      "step": 2715
+    },
+    {
+      "epoch": 0.23817085920126496,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028227525170389993,
+      "loss": 1.2183,
+      "step": 2716
+    },
+    {
+      "epoch": 0.2382585509756395,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0028225505252936307,
+      "loss": 1.2415,
+      "step": 2717
+    },
+    {
+      "epoch": 0.23834624275001404,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0028223484266146816,
+      "loss": 1.1825,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2384339345243886,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002822146221020582,
+      "loss": 1.2649,
+      "step": 2719
+    },
+    {
+      "epoch": 0.23852162629876314,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0028219439085297726,
+      "loss": 1.2604,
+      "step": 2720
+    },
+    {
+      "epoch": 0.23860931807313768,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0028217414891607046,
+      "loss": 1.2391,
+      "step": 2721
+    },
+    {
+      "epoch": 0.23869700984751221,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0028215389629318377,
+      "loss": 1.2664,
+      "step": 2722
+    },
+    {
+      "epoch": 0.23878470162188678,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0028213363298616413,
+      "loss": 1.2401,
+      "step": 2723
+    },
+    {
+      "epoch": 0.23887239339626132,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0028211335899685957,
+      "loss": 1.1829,
+      "step": 2724
+    },
+    {
+      "epoch": 0.23896008517063586,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00282093074327119,
+      "loss": 1.2116,
+      "step": 2725
+    },
+    {
+      "epoch": 0.2390477769450104,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002820727789787923,
+      "loss": 1.3203,
+      "step": 2726
+    },
+    {
+      "epoch": 0.23913546871938496,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002820524729537304,
+      "loss": 1.2356,
+      "step": 2727
+    },
+    {
+      "epoch": 0.2392231604937595,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0028203215625378507,
+      "loss": 1.2577,
+      "step": 2728
+    },
+    {
+      "epoch": 0.23931085226813403,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0028201182888080933,
+      "loss": 1.1915,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2393985440425086,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0028199149083665675,
+      "loss": 1.1676,
+      "step": 2730
+    },
+    {
+      "epoch": 0.23948623581688314,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0028197114212318223,
+      "loss": 1.2593,
+      "step": 2731
+    },
+    {
+      "epoch": 0.23957392759125767,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002819507827422416,
+      "loss": 1.1877,
+      "step": 2732
+    },
+    {
+      "epoch": 0.2396616193656322,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0028193041269569136,
+      "loss": 1.2384,
+      "step": 2733
+    },
+    {
+      "epoch": 0.23974931114000678,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0028191003198538944,
+      "loss": 1.2381,
+      "step": 2734
+    },
+    {
+      "epoch": 0.23983700291438131,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002818896406131944,
+      "loss": 1.2749,
+      "step": 2735
+    },
+    {
+      "epoch": 0.23992469468875585,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0028186923858096583,
+      "loss": 1.2021,
+      "step": 2736
+    },
+    {
+      "epoch": 0.24001238646313042,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002818488258905644,
+      "loss": 1.2125,
+      "step": 2737
+    },
+    {
+      "epoch": 0.24010007823750495,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0028182840254385175,
+      "loss": 1.2032,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2401877700118795,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028180796854269034,
+      "loss": 1.2294,
+      "step": 2739
+    },
+    {
+      "epoch": 0.24027546178625403,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0028178752388894374,
+      "loss": 1.1691,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2403631535606286,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0028176706858447646,
+      "loss": 1.2234,
+      "step": 2741
+    },
+    {
+      "epoch": 0.24045084533500313,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002817466026311539,
+      "loss": 1.1813,
+      "step": 2742
+    },
+    {
+      "epoch": 0.24053853710937767,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002817261260308426,
+      "loss": 1.1748,
+      "step": 2743
+    },
+    {
+      "epoch": 0.2406262288837522,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0028170563878541,
+      "loss": 1.2314,
+      "step": 2744
+    },
+    {
+      "epoch": 0.24071392065812677,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0028168514089672433,
+      "loss": 1.2772,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2408016124325013,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0028166463236665505,
+      "loss": 1.2319,
+      "step": 2746
+    },
+    {
+      "epoch": 0.24088930420687585,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0028164411319707245,
+      "loss": 1.2018,
+      "step": 2747
+    },
+    {
+      "epoch": 0.2409769959812504,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002816235833898479,
+      "loss": 1.3023,
+      "step": 2748
+    },
+    {
+      "epoch": 0.24106468775562495,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.002816030429468535,
+      "loss": 1.2087,
+      "step": 2749
+    },
+    {
+      "epoch": 0.2411523795299995,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0028158249186996274,
+      "loss": 1.1666,
+      "step": 2750
+    },
+    {
+      "epoch": 0.24124007130437403,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002815619301610496,
+      "loss": 1.2572,
+      "step": 2751
+    },
+    {
+      "epoch": 0.2413277630787486,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002815413578219893,
+      "loss": 1.2339,
+      "step": 2752
+    },
+    {
+      "epoch": 0.24141545485312313,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0028152077485465807,
+      "loss": 1.1894,
+      "step": 2753
+    },
+    {
+      "epoch": 0.24150314662749767,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0028150018126093294,
+      "loss": 1.233,
+      "step": 2754
+    },
+    {
+      "epoch": 0.24159083840187223,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0028147957704269208,
+      "loss": 1.2057,
+      "step": 2755
+    },
+    {
+      "epoch": 0.24167853017624677,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0028145896220181437,
+      "loss": 1.2393,
+      "step": 2756
+    },
+    {
+      "epoch": 0.2417662219506213,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0028143833674018006,
+      "loss": 1.2517,
+      "step": 2757
+    },
+    {
+      "epoch": 0.24185391372499584,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0028141770065966995,
+      "loss": 1.2469,
+      "step": 2758
+    },
+    {
+      "epoch": 0.2419416054993704,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002813970539621661,
+      "loss": 1.2275,
+      "step": 2759
+    },
+    {
+      "epoch": 0.24202929727374495,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002813763966495514,
+      "loss": 1.1829,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24211698904811949,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0028135572872370973,
+      "loss": 1.1754,
+      "step": 2761
+    },
+    {
+      "epoch": 0.24220468082249402,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00281335050186526,
+      "loss": 1.2077,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2422923725968686,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00281314361039886,
+      "loss": 1.2106,
+      "step": 2763
+    },
+    {
+      "epoch": 0.24238006437124313,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0028129366128567656,
+      "loss": 1.2277,
+      "step": 2764
+    },
+    {
+      "epoch": 0.24246775614561766,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002812729509257854,
+      "loss": 1.2201,
+      "step": 2765
+    },
+    {
+      "epoch": 0.24255544791999223,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002812522299621013,
+      "loss": 1.198,
+      "step": 2766
+    },
+    {
+      "epoch": 0.24264313969436677,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0028123149839651388,
+      "loss": 1.2044,
+      "step": 2767
+    },
+    {
+      "epoch": 0.2427308314687413,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002812107562309139,
+      "loss": 1.2222,
+      "step": 2768
+    },
+    {
+      "epoch": 0.24281852324311584,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002811900034671929,
+      "loss": 1.2447,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2429062150174904,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0028116924010724354,
+      "loss": 1.1934,
+      "step": 2770
+    },
+    {
+      "epoch": 0.24299390679186494,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0028114846615295945,
+      "loss": 1.2241,
+      "step": 2771
+    },
+    {
+      "epoch": 0.24308159856623948,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0028112768160623498,
+      "loss": 1.1974,
+      "step": 2772
+    },
+    {
+      "epoch": 0.24316929034061405,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002811068864689658,
+      "loss": 1.2624,
+      "step": 2773
+    },
+    {
+      "epoch": 0.24325698211498858,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0028108608074304824,
+      "loss": 1.2345,
+      "step": 2774
+    },
+    {
+      "epoch": 0.24334467388936312,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0028106526443037985,
+      "loss": 1.2242,
+      "step": 2775
+    },
+    {
+      "epoch": 0.24343236566373766,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.002810444375328589,
+      "loss": 1.204,
+      "step": 2776
+    },
+    {
+      "epoch": 0.24352005743811223,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0028102360005238486,
+      "loss": 1.1605,
+      "step": 2777
+    },
+    {
+      "epoch": 0.24360774921248676,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028100275199085803,
+      "loss": 1.2478,
+      "step": 2778
+    },
+    {
+      "epoch": 0.2436954409868613,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002809818933501796,
+      "loss": 1.1483,
+      "step": 2779
+    },
+    {
+      "epoch": 0.24378313276123584,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00280961024132252,
+      "loss": 1.2006,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2438708245356104,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002809401443389783,
+      "loss": 1.1655,
+      "step": 2781
+    },
+    {
+      "epoch": 0.24395851630998494,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0028091925397226276,
+      "loss": 1.2363,
+      "step": 2782
+    },
+    {
+      "epoch": 0.24404620808435948,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0028089835303401044,
+      "loss": 1.2547,
+      "step": 2783
+    },
+    {
+      "epoch": 0.24413389985873404,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0028087744152612754,
+      "loss": 1.1715,
+      "step": 2784
+    },
+    {
+      "epoch": 0.24422159163310858,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0028085651945052113,
+      "loss": 1.2515,
+      "step": 2785
+    },
+    {
+      "epoch": 0.24430928340748312,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002808355868090992,
+      "loss": 1.2418,
+      "step": 2786
+    },
+    {
+      "epoch": 0.24439697518185766,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002808146436037708,
+      "loss": 1.1804,
+      "step": 2787
+    },
+    {
+      "epoch": 0.24448466695623222,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002807936898364459,
+      "loss": 1.2145,
+      "step": 2788
+    },
+    {
+      "epoch": 0.24457235873060676,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0028077272550903537,
+      "loss": 1.2147,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2446600505049813,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0028075175062345116,
+      "loss": 1.2056,
+      "step": 2790
+    },
+    {
+      "epoch": 0.24474774227935586,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002807307651816061,
+      "loss": 1.1901,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2448354340537304,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00280709769185414,
+      "loss": 1.2624,
+      "step": 2792
+    },
+    {
+      "epoch": 0.24492312582810494,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002806887626367897,
+      "loss": 1.2017,
+      "step": 2793
+    },
+    {
+      "epoch": 0.24501081760247947,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0028066774553764887,
+      "loss": 1.1858,
+      "step": 2794
+    },
+    {
+      "epoch": 0.24509850937685404,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002806467178899082,
+      "loss": 1.1742,
+      "step": 2795
+    },
+    {
+      "epoch": 0.24518620115122858,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0028062567969548546,
+      "loss": 1.2245,
+      "step": 2796
+    },
+    {
+      "epoch": 0.24527389292560312,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002806046309562992,
+      "loss": 1.2189,
+      "step": 2797
+    },
+    {
+      "epoch": 0.24536158469997765,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00280583571674269,
+      "loss": 1.1877,
+      "step": 2798
+    },
+    {
+      "epoch": 0.24544927647435222,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0028056250185131545,
+      "loss": 1.2477,
+      "step": 2799
+    },
+    {
+      "epoch": 0.24553696824872676,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0028054142148936007,
+      "loss": 1.2512,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2456246600231013,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002805203305903253,
+      "loss": 1.1946,
+      "step": 2801
+    },
+    {
+      "epoch": 0.24571235179747586,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0028049922915613463,
+      "loss": 1.1907,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2458000435718504,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0028047811718871236,
+      "loss": 1.2198,
+      "step": 2803
+    },
+    {
+      "epoch": 0.24588773534622493,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0028045699468998396,
+      "loss": 1.2383,
+      "step": 2804
+    },
+    {
+      "epoch": 0.24597542712059947,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0028043586166187568,
+      "loss": 1.2598,
+      "step": 2805
+    },
+    {
+      "epoch": 0.24606311889497404,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002804147181063148,
+      "loss": 1.1921,
+      "step": 2806
+    },
+    {
+      "epoch": 0.24615081066934857,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002803935640252296,
+      "loss": 1.1618,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2462385024437231,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0028037239942054924,
+      "loss": 1.1913,
+      "step": 2808
+    },
+    {
+      "epoch": 0.24632619421809768,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0028035122429420386,
+      "loss": 1.216,
+      "step": 2809
+    },
+    {
+      "epoch": 0.24641388599247221,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002803300386481246,
+      "loss": 1.2054,
+      "step": 2810
+    },
+    {
+      "epoch": 0.24650157776684675,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002803088424842436,
+      "loss": 1.2625,
+      "step": 2811
+    },
+    {
+      "epoch": 0.2465892695412213,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0028028763580449376,
+      "loss": 1.197,
+      "step": 2812
+    },
+    {
+      "epoch": 0.24667696131559586,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002802664186108092,
+      "loss": 1.222,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2467646530899704,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002802451909051248,
+      "loss": 1.2714,
+      "step": 2814
+    },
+    {
+      "epoch": 0.24685234486434493,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002802239526893765,
+      "loss": 1.2253,
+      "step": 2815
+    },
+    {
+      "epoch": 0.24694003663871947,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002802027039655012,
+      "loss": 1.1983,
+      "step": 2816
+    },
+    {
+      "epoch": 0.24702772841309403,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002801814447354366,
+      "loss": 1.1837,
+      "step": 2817
+    },
+    {
+      "epoch": 0.24711542018746857,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0028016017500112162,
+      "loss": 1.2055,
+      "step": 2818
+    },
+    {
+      "epoch": 0.2472031119618431,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0028013889476449596,
+      "loss": 1.2456,
+      "step": 2819
+    },
+    {
+      "epoch": 0.24729080373621767,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0028011760402750037,
+      "loss": 1.2163,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2473784955105922,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028009630279207643,
+      "loss": 1.2467,
+      "step": 2821
+    },
+    {
+      "epoch": 0.24746618728496675,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002800749910601668,
+      "loss": 1.2159,
+      "step": 2822
+    },
+    {
+      "epoch": 0.2475538790593413,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00280053668833715,
+      "loss": 1.2399,
+      "step": 2823
+    },
+    {
+      "epoch": 0.24764157083371585,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0028003233611466572,
+      "loss": 1.2219,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2477292626080904,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0028001099290496426,
+      "loss": 1.2125,
+      "step": 2825
+    },
+    {
+      "epoch": 0.24781695438246493,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027998963920655715,
+      "loss": 1.2351,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2479046461568395,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0027996827502139183,
+      "loss": 1.2392,
+      "step": 2827
+    },
+    {
+      "epoch": 0.24799233793121403,
+      "grad_norm": 0.125,
+      "learning_rate": 0.002799469003514166,
+      "loss": 1.2123,
+      "step": 2828
+    },
+    {
+      "epoch": 0.24808002970558857,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0027992551519858077,
+      "loss": 1.2462,
+      "step": 2829
+    },
+    {
+      "epoch": 0.2481677214799631,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0027990411956483464,
+      "loss": 1.233,
+      "step": 2830
+    },
+    {
+      "epoch": 0.24825541325433767,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0027988271345212945,
+      "loss": 1.2389,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2483431050287122,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027986129686241737,
+      "loss": 1.2273,
+      "step": 2832
+    },
+    {
+      "epoch": 0.24843079680308675,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0027983986979765152,
+      "loss": 1.2232,
+      "step": 2833
+    },
+    {
+      "epoch": 0.24851848857746128,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00279818432259786,
+      "loss": 1.2149,
+      "step": 2834
+    },
+    {
+      "epoch": 0.24860618035183585,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027979698425077584,
+      "loss": 1.2788,
+      "step": 2835
+    },
+    {
+      "epoch": 0.24869387212621039,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0027977552577257707,
+      "loss": 1.2552,
+      "step": 2836
+    },
+    {
+      "epoch": 0.24878156390058492,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0027975405682714666,
+      "loss": 1.2173,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2488692556749595,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0027973257741644247,
+      "loss": 1.1674,
+      "step": 2838
+    },
+    {
+      "epoch": 0.24895694744933403,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002797110875424234,
+      "loss": 1.2148,
+      "step": 2839
+    },
+    {
+      "epoch": 0.24904463922370856,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0027968958720704937,
+      "loss": 1.2162,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2491323309980831,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027966807641228095,
+      "loss": 1.2317,
+      "step": 2841
+    },
+    {
+      "epoch": 0.24922002277245767,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0027964655516008,
+      "loss": 1.2457,
+      "step": 2842
+    },
+    {
+      "epoch": 0.2493077145468322,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0027962502345240917,
+      "loss": 1.1574,
+      "step": 2843
+    },
+    {
+      "epoch": 0.24939540632120674,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002796034812912321,
+      "loss": 1.2483,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2494830980955813,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002795819286785134,
+      "loss": 1.1864,
+      "step": 2845
+    },
+    {
+      "epoch": 0.24957078986995584,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0027956036561621865,
+      "loss": 1.2003,
+      "step": 2846
+    },
+    {
+      "epoch": 0.24965848164433038,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0027953879210631423,
+      "loss": 1.3245,
+      "step": 2847
+    },
+    {
+      "epoch": 0.24974617341870492,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0027951720815076763,
+      "loss": 1.1422,
+      "step": 2848
+    },
+    {
+      "epoch": 0.24983386519307949,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002794956137515473,
+      "loss": 1.2049,
+      "step": 2849
+    },
+    {
+      "epoch": 0.24992155696745402,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002794740089106226,
+      "loss": 1.2318,
+      "step": 2850
+    },
+    {
+      "epoch": 0.25000924874182856,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0027945239362996374,
+      "loss": 1.1833,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2500969405162031,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0027943076791154204,
+      "loss": 1.2134,
+      "step": 2852
+    },
+    {
+      "epoch": 0.25018463229057764,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0027940913175732974,
+      "loss": 1.2269,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2502723240649522,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0027938748516929996,
+      "loss": 1.2345,
+      "step": 2854
+    },
+    {
+      "epoch": 0.25036001583932677,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0027936582814942685,
+      "loss": 1.2196,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2504477076137013,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002793441606996854,
+      "loss": 1.1914,
+      "step": 2856
+    },
+    {
+      "epoch": 0.25053539938807584,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002793224828220517,
+      "loss": 1.1923,
+      "step": 2857
+    },
+    {
+      "epoch": 0.2506230911624504,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0027930079451850267,
+      "loss": 1.2086,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2507107829368249,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0027927909579101625,
+      "loss": 1.2562,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2507984747111995,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002792573866415713,
+      "loss": 1.2113,
+      "step": 2860
+    },
+    {
+      "epoch": 0.250886166485574,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002792356670721477,
+      "loss": 1.1959,
+      "step": 2861
+    },
+    {
+      "epoch": 0.25097385825994856,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002792139370847261,
+      "loss": 1.2234,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2510615500343231,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002791921966812883,
+      "loss": 1.1612,
+      "step": 2863
+    },
+    {
+      "epoch": 0.25114924180869763,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0027917044586381694,
+      "loss": 1.1847,
+      "step": 2864
+    },
+    {
+      "epoch": 0.2512369335830722,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0027914868463429564,
+      "loss": 1.2312,
+      "step": 2865
+    },
+    {
+      "epoch": 0.25132462535744676,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00279126912994709,
+      "loss": 1.1881,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2514123171318213,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0027910513094704247,
+      "loss": 1.2322,
+      "step": 2867
+    },
+    {
+      "epoch": 0.25150000890619584,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002790833384932826,
+      "loss": 1.2592,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2515877006805704,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0027906153563541673,
+      "loss": 1.1932,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2516753924549449,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002790397223754333,
+      "loss": 1.2049,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2517630842293195,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0027901789871532154,
+      "loss": 1.2293,
+      "step": 2871
+    },
+    {
+      "epoch": 0.25185077600369404,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0027899606465707177,
+      "loss": 1.2083,
+      "step": 2872
+    },
+    {
+      "epoch": 0.25193846777806855,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0027897422020267512,
+      "loss": 1.2075,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2520261595524431,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002789523653541239,
+      "loss": 1.2167,
+      "step": 2874
+    },
+    {
+      "epoch": 0.25211385132681763,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002789305001134111,
+      "loss": 1.2596,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2522015431011922,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0027890862448253085,
+      "loss": 1.2052,
+      "step": 2876
+    },
+    {
+      "epoch": 0.25228923487556676,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00278886738463478,
+      "loss": 1.1974,
+      "step": 2877
+    },
+    {
+      "epoch": 0.25237692664994127,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0027886484205824867,
+      "loss": 1.193,
+      "step": 2878
+    },
+    {
+      "epoch": 0.25246461842431583,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002788429352688397,
+      "loss": 1.2423,
+      "step": 2879
+    },
+    {
+      "epoch": 0.2525523101986904,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0027882101809724885,
+      "loss": 1.2184,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2526400019730649,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00278799090545475,
+      "loss": 1.2665,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2527276937474395,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002787771526155179,
+      "loss": 1.1902,
+      "step": 2882
+    },
+    {
+      "epoch": 0.25281538552181404,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0027875520430937816,
+      "loss": 1.1662,
+      "step": 2883
+    },
+    {
+      "epoch": 0.25290307729618855,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0027873324562905743,
+      "loss": 1.1653,
+      "step": 2884
+    },
+    {
+      "epoch": 0.2529907690705631,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0027871127657655833,
+      "loss": 1.1631,
+      "step": 2885
+    },
+    {
+      "epoch": 0.2530784608449376,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0027868929715388433,
+      "loss": 1.2427,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2531661526193122,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0027866730736303994,
+      "loss": 1.2334,
+      "step": 2887
+    },
+    {
+      "epoch": 0.25325384439368676,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0027864530720603056,
+      "loss": 1.2079,
+      "step": 2888
+    },
+    {
+      "epoch": 0.25334153616806127,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0027862329668486246,
+      "loss": 1.2163,
+      "step": 2889
+    },
+    {
+      "epoch": 0.25342922794243583,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0027860127580154313,
+      "loss": 1.2109,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2535169197168104,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0027857924455808062,
+      "loss": 1.1996,
+      "step": 2891
+    },
+    {
+      "epoch": 0.2536046114911849,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0027855720295648424,
+      "loss": 1.1655,
+      "step": 2892
+    },
+    {
+      "epoch": 0.25369230326555947,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002785351509987641,
+      "loss": 1.293,
+      "step": 2893
+    },
+    {
+      "epoch": 0.25377999503993404,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0027851308868693126,
+      "loss": 1.3099,
+      "step": 2894
+    },
+    {
+      "epoch": 0.25386768681430855,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002784910160229978,
+      "loss": 1.1904,
+      "step": 2895
+    },
+    {
+      "epoch": 0.2539553785886831,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002784689330089766,
+      "loss": 1.2164,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2540430703630576,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002784468396468817,
+      "loss": 1.2586,
+      "step": 2897
+    },
+    {
+      "epoch": 0.2541307621374322,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0027842473593872783,
+      "loss": 1.2428,
+      "step": 2898
+    },
+    {
+      "epoch": 0.25421845391180675,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002784026218865309,
+      "loss": 1.1785,
+      "step": 2899
+    },
+    {
+      "epoch": 0.25430614568618126,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0027838049749230754,
+      "loss": 1.2431,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2543938374605558,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0027835836275807552,
+      "loss": 1.1473,
+      "step": 2901
+    },
+    {
+      "epoch": 0.2544815292349304,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0027833621768585345,
+      "loss": 1.201,
+      "step": 2902
+    },
+    {
+      "epoch": 0.2545692210093049,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002783140622776609,
+      "loss": 1.2095,
+      "step": 2903
+    },
+    {
+      "epoch": 0.25465691278367947,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0027829189653551844,
+      "loss": 1.1969,
+      "step": 2904
+    },
+    {
+      "epoch": 0.25474460455805403,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002782697204614474,
+      "loss": 1.2771,
+      "step": 2905
+    },
+    {
+      "epoch": 0.25483229633242854,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0027824753405747025,
+      "loss": 1.2157,
+      "step": 2906
+    },
+    {
+      "epoch": 0.2549199881068031,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002782253373256104,
+      "loss": 1.2889,
+      "step": 2907
+    },
+    {
+      "epoch": 0.2550076798811776,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002782031302678921,
+      "loss": 1.2213,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2550953716555522,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002781809128863405,
+      "loss": 1.198,
+      "step": 2909
+    },
+    {
+      "epoch": 0.25518306342992675,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002781586851829818,
+      "loss": 1.2364,
+      "step": 2910
+    },
+    {
+      "epoch": 0.25527075520430126,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002781364471598432,
+      "loss": 1.2125,
+      "step": 2911
+    },
+    {
+      "epoch": 0.2553584469786758,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0027811419881895263,
+      "loss": 1.2548,
+      "step": 2912
+    },
+    {
+      "epoch": 0.2554461387530504,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002780919401623391,
+      "loss": 1.2392,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2555338305274249,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0027806967119203265,
+      "loss": 1.2197,
+      "step": 2914
+    },
+    {
+      "epoch": 0.25562152230179946,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0027804739191006405,
+      "loss": 1.2204,
+      "step": 2915
+    },
+    {
+      "epoch": 0.25570921407617403,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002780251023184651,
+      "loss": 1.186,
+      "step": 2916
+    },
+    {
+      "epoch": 0.25579690585054854,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0027800280241926864,
+      "loss": 1.1727,
+      "step": 2917
+    },
+    {
+      "epoch": 0.2558845976249231,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002779804922145083,
+      "loss": 1.22,
+      "step": 2918
+    },
+    {
+      "epoch": 0.25597228939929767,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0027795817170621867,
+      "loss": 1.2085,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2560599811736722,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0027793584089643546,
+      "loss": 1.2524,
+      "step": 2920
+    },
+    {
+      "epoch": 0.25614767294804675,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0027791349978719504,
+      "loss": 1.2057,
+      "step": 2921
+    },
+    {
+      "epoch": 0.25623536472242125,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0027789114838053497,
+      "loss": 1.2192,
+      "step": 2922
+    },
+    {
+      "epoch": 0.2563230564967958,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0027786878667849357,
+      "loss": 1.2144,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2564107482711704,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0027784641468311024,
+      "loss": 1.2547,
+      "step": 2924
+    },
+    {
+      "epoch": 0.2564984400455449,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0027782403239642517,
+      "loss": 1.2265,
+      "step": 2925
+    },
+    {
+      "epoch": 0.25658613181991946,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002778016398204796,
+      "loss": 1.2449,
+      "step": 2926
+    },
+    {
+      "epoch": 0.256673823594294,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0027777923695731566,
+      "loss": 1.2019,
+      "step": 2927
+    },
+    {
+      "epoch": 0.25676151536866854,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027775682380897647,
+      "loss": 1.2748,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2568492071430431,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00277734400377506,
+      "loss": 1.1948,
+      "step": 2929
+    },
+    {
+      "epoch": 0.25693689891741767,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0027771196666494928,
+      "loss": 1.2228,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2570245906917922,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0027768952267335214,
+      "loss": 1.1704,
+      "step": 2931
+    },
+    {
+      "epoch": 0.25711228246616674,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002776670684047614,
+      "loss": 1.2227,
+      "step": 2932
+    },
+    {
+      "epoch": 0.25719997424054125,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0027764460386122494,
+      "loss": 1.1527,
+      "step": 2933
+    },
+    {
+      "epoch": 0.2572876660149158,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0027762212904479137,
+      "loss": 1.2376,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2573753577892904,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002775996439575103,
+      "loss": 1.205,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2574630495636649,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0027757714860143245,
+      "loss": 1.2358,
+      "step": 2936
+    },
+    {
+      "epoch": 0.25755074133803946,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002775546429786093,
+      "loss": 1.2129,
+      "step": 2937
+    },
+    {
+      "epoch": 0.257638433112414,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0027753212709109316,
+      "loss": 1.2641,
+      "step": 2938
+    },
+    {
+      "epoch": 0.25772612488678853,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0027750960094093764,
+      "loss": 1.154,
+      "step": 2939
+    },
+    {
+      "epoch": 0.2578138166611631,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002774870645301969,
+      "loss": 1.1743,
+      "step": 2940
+    },
+    {
+      "epoch": 0.25790150843553766,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0027746451786092624,
+      "loss": 1.3064,
+      "step": 2941
+    },
+    {
+      "epoch": 0.2579892002099122,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0027744196093518188,
+      "loss": 1.2278,
+      "step": 2942
+    },
+    {
+      "epoch": 0.25807689198428674,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0027741939375502097,
+      "loss": 1.2385,
+      "step": 2943
+    },
+    {
+      "epoch": 0.25816458375866125,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0027739681632250155,
+      "loss": 1.2469,
+      "step": 2944
+    },
+    {
+      "epoch": 0.2582522755330358,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002773742286396827,
+      "loss": 1.1947,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2583399673074104,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002773516307086242,
+      "loss": 1.2256,
+      "step": 2946
+    },
+    {
+      "epoch": 0.2584276590817849,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027732902253138707,
+      "loss": 1.2535,
+      "step": 2947
+    },
+    {
+      "epoch": 0.25851535085615945,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00277306404110033,
+      "loss": 1.1799,
+      "step": 2948
+    },
+    {
+      "epoch": 0.258603042630534,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0027728377544662484,
+      "loss": 1.2235,
+      "step": 2949
+    },
+    {
+      "epoch": 0.25869073440490853,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0027726113654322616,
+      "loss": 1.1489,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2587784261792831,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002772384874019017,
+      "loss": 1.2553,
+      "step": 2951
+    },
+    {
+      "epoch": 0.25886611795365766,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0027721582802471695,
+      "loss": 1.2271,
+      "step": 2952
+    },
+    {
+      "epoch": 0.25895380972803217,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0027719315841373826,
+      "loss": 1.148,
+      "step": 2953
+    },
+    {
+      "epoch": 0.25904150150240673,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0027717047857103326,
+      "loss": 1.2451,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2591291932767813,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0027714778849867016,
+      "loss": 1.1276,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2592168850511558,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002771250881987182,
+      "loss": 1.2717,
+      "step": 2956
+    },
+    {
+      "epoch": 0.2593045768255304,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002771023776732477,
+      "loss": 1.1861,
+      "step": 2957
+    },
+    {
+      "epoch": 0.2593922685999049,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002770796569243297,
+      "loss": 1.2405,
+      "step": 2958
+    },
+    {
+      "epoch": 0.25947996037427945,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0027705692595403632,
+      "loss": 1.1906,
+      "step": 2959
+    },
+    {
+      "epoch": 0.259567652148654,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002770341847644406,
+      "loss": 1.245,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2596553439230285,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0027701143335761643,
+      "loss": 1.2197,
+      "step": 2961
+    },
+    {
+      "epoch": 0.2597430356974031,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002769886717356387,
+      "loss": 1.1793,
+      "step": 2962
+    },
+    {
+      "epoch": 0.25983072747177766,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002769658999005832,
+      "loss": 1.1586,
+      "step": 2963
+    },
+    {
+      "epoch": 0.25991841924615217,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0027694311785452673,
+      "loss": 1.2168,
+      "step": 2964
+    },
+    {
+      "epoch": 0.26000611102052673,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002769203255995468,
+      "loss": 1.2114,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2600938027949013,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002768975231377221,
+      "loss": 1.319,
+      "step": 2966
+    },
+    {
+      "epoch": 0.2601814945692758,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002768747104711322,
+      "loss": 1.1747,
+      "step": 2967
+    },
+    {
+      "epoch": 0.26026918634365037,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0027685188760185747,
+      "loss": 1.2456,
+      "step": 2968
+    },
+    {
+      "epoch": 0.2603568781180249,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0027682905453197937,
+      "loss": 1.2478,
+      "step": 2969
+    },
+    {
+      "epoch": 0.26044456989239945,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0027680621126358017,
+      "loss": 1.2655,
+      "step": 2970
+    },
+    {
+      "epoch": 0.260532261666774,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002767833577987431,
+      "loss": 1.2369,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2606199534411485,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0027676049413955244,
+      "loss": 1.2384,
+      "step": 2972
+    },
+    {
+      "epoch": 0.2607076452155231,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002767376202880931,
+      "loss": 1.184,
+      "step": 2973
+    },
+    {
+      "epoch": 0.26079533698989765,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002767147362464514,
+      "loss": 1.2164,
+      "step": 2974
+    },
+    {
+      "epoch": 0.26088302876427216,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0027669184201671395,
+      "loss": 1.2302,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2609707205386467,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0027666893760096897,
+      "loss": 1.2051,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2610584123130213,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0027664602300130517,
+      "loss": 1.2319,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2611461040873958,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0027662309821981225,
+      "loss": 1.2215,
+      "step": 2978
+    },
+    {
+      "epoch": 0.26123379586177037,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002766001632585809,
+      "loss": 1.2031,
+      "step": 2979
+    },
+    {
+      "epoch": 0.2613214876361449,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0027657721811970275,
+      "loss": 1.1895,
+      "step": 2980
+    },
+    {
+      "epoch": 0.26140917941051944,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0027655426280527038,
+      "loss": 1.1953,
+      "step": 2981
+    },
+    {
+      "epoch": 0.261496871184894,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0027653129731737723,
+      "loss": 1.2263,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2615845629592685,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0027650832165811764,
+      "loss": 1.211,
+      "step": 2983
+    },
+    {
+      "epoch": 0.2616722547336431,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0027648533582958702,
+      "loss": 1.2051,
+      "step": 2984
+    },
+    {
+      "epoch": 0.26175994650801765,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002764623398338815,
+      "loss": 1.2248,
+      "step": 2985
+    },
+    {
+      "epoch": 0.26184763828239216,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0027643933367309838,
+      "loss": 1.159,
+      "step": 2986
+    },
+    {
+      "epoch": 0.2619353300567667,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027641631734933564,
+      "loss": 1.292,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2620230218311413,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0027639329086469244,
+      "loss": 1.2101,
+      "step": 2988
+    },
+    {
+      "epoch": 0.2621107136055158,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002763702542212687,
+      "loss": 1.2107,
+      "step": 2989
+    },
+    {
+      "epoch": 0.26219840537989036,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002763472074211652,
+      "loss": 1.2438,
+      "step": 2990
+    },
+    {
+      "epoch": 0.26228609715426493,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0027632415046648386,
+      "loss": 1.2295,
+      "step": 2991
+    },
+    {
+      "epoch": 0.26237378892863944,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0027630108335932745,
+      "loss": 1.1904,
+      "step": 2992
+    },
+    {
+      "epoch": 0.262461480703014,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0027627800610179943,
+      "loss": 1.2132,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2625491724773885,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002762549186960046,
+      "loss": 1.2255,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2626368642517631,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0027623182114404838,
+      "loss": 1.1881,
+      "step": 2995
+    },
+    {
+      "epoch": 0.26272455602613765,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027620871344803716,
+      "loss": 1.211,
+      "step": 2996
+    },
+    {
+      "epoch": 0.26281224780051216,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0027618559561007837,
+      "loss": 1.2015,
+      "step": 2997
+    },
+    {
+      "epoch": 0.2628999395748867,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0027616246763228036,
+      "loss": 1.2013,
+      "step": 2998
+    },
+    {
+      "epoch": 0.2629876313492613,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0027613932951675224,
+      "loss": 1.2283,
+      "step": 2999
+    },
+    {
+      "epoch": 0.2630753231236358,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0027611618126560417,
+      "loss": 1.2163,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2630753231236358,
+      "eval_loss": 1.2262382507324219,
+      "eval_runtime": 428.7851,
+      "eval_samples_per_second": 33.693,
+      "eval_steps_per_second": 8.424,
+      "step": 3000
+    },
+    {
+      "epoch": 0.26316301489801036,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002760930228809472,
+      "loss": 1.2914,
+      "step": 3001
+    },
+    {
+      "epoch": 0.2632507066723849,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002760698543648933,
+      "loss": 1.199,
+      "step": 3002
+    },
+    {
+      "epoch": 0.26333839844675944,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002760466757195554,
+      "loss": 1.2503,
+      "step": 3003
+    },
+    {
+      "epoch": 0.263426090221134,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0027602348694704738,
+      "loss": 1.1814,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2635137819955085,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0027600028804948395,
+      "loss": 1.2841,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2636014737698831,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0027597707902898084,
+      "loss": 1.2287,
+      "step": 3006
+    },
+    {
+      "epoch": 0.26368916554425764,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002759538598876545,
+      "loss": 1.197,
+      "step": 3007
+    },
+    {
+      "epoch": 0.26377685731863215,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0027593063062762266,
+      "loss": 1.23,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2638645490930067,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002759073912510036,
+      "loss": 1.278,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2639522408673813,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002758841417599168,
+      "loss": 1.2244,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2640399326417558,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0027586088215648254,
+      "loss": 1.1954,
+      "step": 3011
+    },
+    {
+      "epoch": 0.26412762441613036,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00275837612442822,
+      "loss": 1.2664,
+      "step": 3012
+    },
+    {
+      "epoch": 0.2642153161905049,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002758143326210573,
+      "loss": 1.1921,
+      "step": 3013
+    },
+    {
+      "epoch": 0.26430300796487943,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0027579104269331157,
+      "loss": 1.231,
+      "step": 3014
+    },
+    {
+      "epoch": 0.264390699739254,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0027576774266170876,
+      "loss": 1.2155,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2644783915136285,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0027574443252837374,
+      "loss": 1.1872,
+      "step": 3016
+    },
+    {
+      "epoch": 0.2645660832880031,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0027572111229543232,
+      "loss": 1.1784,
+      "step": 3017
+    },
+    {
+      "epoch": 0.26465377506237764,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0027569778196501137,
+      "loss": 1.1992,
+      "step": 3018
+    },
+    {
+      "epoch": 0.26474146683675215,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0027567444153923843,
+      "loss": 1.2336,
+      "step": 3019
+    },
+    {
+      "epoch": 0.2648291586111267,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0027565109102024206,
+      "loss": 1.2378,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2649168503855013,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027562773041015193,
+      "loss": 1.2282,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2650045421598758,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002756043597110983,
+      "loss": 1.2049,
+      "step": 3022
+    },
+    {
+      "epoch": 0.26509223393425035,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0027558097892521265,
+      "loss": 1.2567,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2651799257086249,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002755575880546272,
+      "loss": 1.2378,
+      "step": 3024
+    },
+    {
+      "epoch": 0.26526761748299943,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0027553418710147505,
+      "loss": 1.2389,
+      "step": 3025
+    },
+    {
+      "epoch": 0.265355309257374,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002755107760678904,
+      "loss": 1.2203,
+      "step": 3026
+    },
+    {
+      "epoch": 0.26544300103174856,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002754873549560083,
+      "loss": 1.1923,
+      "step": 3027
+    },
+    {
+      "epoch": 0.26553069280612307,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002754639237679646,
+      "loss": 1.2175,
+      "step": 3028
+    },
+    {
+      "epoch": 0.26561838458049764,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027544048250589623,
+      "loss": 1.2011,
+      "step": 3029
+    },
+    {
+      "epoch": 0.26570607635487214,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002754170311719409,
+      "loss": 1.2183,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2657937681292467,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0027539356976823744,
+      "loss": 1.1718,
+      "step": 3031
+    },
+    {
+      "epoch": 0.2658814599036213,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0027537009829692537,
+      "loss": 1.1575,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2659691516779958,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0027534661676014527,
+      "loss": 1.2039,
+      "step": 3033
+    },
+    {
+      "epoch": 0.26605684345237035,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002753231251600386,
+      "loss": 1.2451,
+      "step": 3034
+    },
+    {
+      "epoch": 0.2661445352267449,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0027529962349874767,
+      "loss": 1.2158,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2662322270011194,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002752761117784158,
+      "loss": 1.2443,
+      "step": 3036
+    },
+    {
+      "epoch": 0.266319918775494,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0027525259000118723,
+      "loss": 1.2247,
+      "step": 3037
+    },
+    {
+      "epoch": 0.26640761054986856,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002752290581692071,
+      "loss": 1.1676,
+      "step": 3038
+    },
+    {
+      "epoch": 0.26649530232424307,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002752055162846214,
+      "loss": 1.1866,
+      "step": 3039
+    },
+    {
+      "epoch": 0.26658299409861763,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002751819643495771,
+      "loss": 1.2442,
+      "step": 3040
+    },
+    {
+      "epoch": 0.26667068587299214,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002751584023662221,
+      "loss": 1.1324,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2667583776473667,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0027513483033670513,
+      "loss": 1.2184,
+      "step": 3042
+    },
+    {
+      "epoch": 0.26684606942174127,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00275111248263176,
+      "loss": 1.2118,
+      "step": 3043
+    },
+    {
+      "epoch": 0.2669337611961158,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0027508765614778527,
+      "loss": 1.1679,
+      "step": 3044
+    },
+    {
+      "epoch": 0.26702145297049035,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0027506405399268445,
+      "loss": 1.1669,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2671091447448649,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002750404418000261,
+      "loss": 1.2071,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2671968365192394,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002750168195719635,
+      "loss": 1.3092,
+      "step": 3047
+    },
+    {
+      "epoch": 0.267284528293614,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0027499318731065096,
+      "loss": 1.195,
+      "step": 3048
+    },
+    {
+      "epoch": 0.26737222006798855,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002749695450182437,
+      "loss": 1.2021,
+      "step": 3049
+    },
+    {
+      "epoch": 0.26745991184236306,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0027494589269689775,
+      "loss": 1.1933,
+      "step": 3050
+    },
+    {
+      "epoch": 0.26754760361673763,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0027492223034877024,
+      "loss": 1.2187,
+      "step": 3051
+    },
+    {
+      "epoch": 0.26763529539111214,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0027489855797601915,
+      "loss": 1.1839,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2677229871654867,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002748748755808032,
+      "loss": 1.1965,
+      "step": 3053
+    },
+    {
+      "epoch": 0.26781067893986127,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0027485118316528225,
+      "loss": 1.2002,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2678983707142358,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0027482748073161703,
+      "loss": 1.2611,
+      "step": 3055
+    },
+    {
+      "epoch": 0.26798606248861034,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00274803768281969,
+      "loss": 1.246,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2680737542629849,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.002747800458185008,
+      "loss": 1.2235,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2681614460373594,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0027475631334337586,
+      "loss": 1.1927,
+      "step": 3058
+    },
+    {
+      "epoch": 0.268249137811734,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0027473257085875837,
+      "loss": 1.2451,
+      "step": 3059
+    },
+    {
+      "epoch": 0.26833682958610855,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0027470881836681375,
+      "loss": 1.195,
+      "step": 3060
+    },
+    {
+      "epoch": 0.26842452136048306,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0027468505586970816,
+      "loss": 1.2327,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2685122131348576,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002746612833696086,
+      "loss": 1.2268,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2685999049092322,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00274637500868683,
+      "loss": 1.1877,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2686875966836067,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0027461370836910047,
+      "loss": 1.2453,
+      "step": 3064
+    },
+    {
+      "epoch": 0.26877528845798127,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002745899058730306,
+      "loss": 1.1945,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2688629802323558,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002745660933826443,
+      "loss": 1.2393,
+      "step": 3066
+    },
+    {
+      "epoch": 0.26895067200673034,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0027454227090011308,
+      "loss": 1.2457,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2690383637811049,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0027451843842760957,
+      "loss": 1.1826,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2691260555554794,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0027449459596730714,
+      "loss": 1.2722,
+      "step": 3069
+    },
+    {
+      "epoch": 0.269213747329854,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002744707435213803,
+      "loss": 1.2205,
+      "step": 3070
+    },
+    {
+      "epoch": 0.26930143910422855,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002744468810920042,
+      "loss": 1.2162,
+      "step": 3071
+    },
+    {
+      "epoch": 0.26938913087860306,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0027442300868135506,
+      "loss": 1.2269,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2694768226529776,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0027439912629161,
+      "loss": 1.2111,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2695645144273522,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002743752339249471,
+      "loss": 1.2563,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2696522062017267,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0027435133158354515,
+      "loss": 1.218,
+      "step": 3075
+    },
+    {
+      "epoch": 0.26973989797610126,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0027432741926958406,
+      "loss": 1.2171,
+      "step": 3076
+    },
+    {
+      "epoch": 0.26982758975047577,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0027430349698524463,
+      "loss": 1.1564,
+      "step": 3077
+    },
+    {
+      "epoch": 0.26991528152485034,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002742795647327083,
+      "loss": 1.277,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2700029732992249,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0027425562251415787,
+      "loss": 1.2364,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2700906650735994,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002742316703317767,
+      "loss": 1.2028,
+      "step": 3080
+    },
+    {
+      "epoch": 0.270178356847974,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027420770818774916,
+      "loss": 1.2521,
+      "step": 3081
+    },
+    {
+      "epoch": 0.27026604862234854,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002741837360842606,
+      "loss": 1.269,
+      "step": 3082
+    },
+    {
+      "epoch": 0.27035374039672305,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002741597540234971,
+      "loss": 1.2486,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2704414321710976,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0027413576200764583,
+      "loss": 1.1683,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2705291239454722,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0027411176003889486,
+      "loss": 1.2407,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2706168157198467,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0027408774811943297,
+      "loss": 1.1782,
+      "step": 3086
+    },
+    {
+      "epoch": 0.27070450749422126,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0027406372625145012,
+      "loss": 1.2137,
+      "step": 3087
+    },
+    {
+      "epoch": 0.27079219926859577,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0027403969443713693,
+      "loss": 1.1782,
+      "step": 3088
+    },
+    {
+      "epoch": 0.27087989104297033,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0027401565267868512,
+      "loss": 1.2019,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2709675828173449,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002739916009782872,
+      "loss": 1.2204,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2710552745917194,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0027396753933813666,
+      "loss": 1.194,
+      "step": 3091
+    },
+    {
+      "epoch": 0.271142966366094,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.002739434677604278,
+      "loss": 1.259,
+      "step": 3092
+    },
+    {
+      "epoch": 0.27123065814046854,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002739193862473559,
+      "loss": 1.1818,
+      "step": 3093
+    },
+    {
+      "epoch": 0.27131834991484305,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002738952948011172,
+      "loss": 1.2055,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2714060416892176,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0027387119342390867,
+      "loss": 1.1827,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2714937334635922,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002738470821179284,
+      "loss": 1.1592,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2715814252379667,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002738229608853752,
+      "loss": 1.2316,
+      "step": 3097
+    },
+    {
+      "epoch": 0.27166911701234125,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0027379882972844894,
+      "loss": 1.2042,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2717568087867158,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002737746886493502,
+      "loss": 1.168,
+      "step": 3099
+    },
+    {
+      "epoch": 0.27184450056109033,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0027375053765028067,
+      "loss": 1.2092,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2719321923354649,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0027372637673344284,
+      "loss": 1.2274,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2720198841098394,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0027370220590104016,
+      "loss": 1.2464,
+      "step": 3102
+    },
+    {
+      "epoch": 0.27210757588421397,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0027367802515527694,
+      "loss": 1.1675,
+      "step": 3103
+    },
+    {
+      "epoch": 0.27219526765858854,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002736538344983583,
+      "loss": 1.2383,
+      "step": 3104
+    },
+    {
+      "epoch": 0.27228295943296305,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0027362963393249046,
+      "loss": 1.2322,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2723706512073376,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002736054234598805,
+      "loss": 1.1892,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2724583429817122,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0027358120308273626,
+      "loss": 1.2003,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2725460347560867,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002735569728032665,
+      "loss": 1.2341,
+      "step": 3108
+    },
+    {
+      "epoch": 0.27263372653046125,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0027353273262368115,
+      "loss": 1.2852,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2727214183048358,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0027350848254619076,
+      "loss": 1.1885,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2728091100792103,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0027348422257300687,
+      "loss": 1.1697,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2728968018535849,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027345995270634193,
+      "loss": 1.1881,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2729844936279594,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0027343567294840928,
+      "loss": 1.2431,
+      "step": 3113
+    },
+    {
+      "epoch": 0.27307218540233397,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0027341138330142316,
+      "loss": 1.2331,
+      "step": 3114
+    },
+    {
+      "epoch": 0.27315987717670853,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0027338708376759883,
+      "loss": 1.1984,
+      "step": 3115
+    },
+    {
+      "epoch": 0.27324756895108304,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002733627743491522,
+      "loss": 1.1997,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2733352607254576,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0027333845504830027,
+      "loss": 1.1682,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2734229524998322,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002733141258672609,
+      "loss": 1.2168,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2735106442742067,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0027328978680825292,
+      "loss": 1.1736,
+      "step": 3119
+    },
+    {
+      "epoch": 0.27359833604858125,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027326543787349594,
+      "loss": 1.2487,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2736860278229558,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0027324107906521043,
+      "loss": 1.1617,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2737737195973303,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00273216710385618,
+      "loss": 1.2255,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2738614113717049,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002731923318369409,
+      "loss": 1.2132,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2739491031460794,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0027316794342140244,
+      "loss": 1.2828,
+      "step": 3124
+    },
+    {
+      "epoch": 0.27403679492045396,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0027314354514122686,
+      "loss": 1.2747,
+      "step": 3125
+    },
+    {
+      "epoch": 0.27412448669482853,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0027311913699863905,
+      "loss": 1.1524,
+      "step": 3126
+    },
+    {
+      "epoch": 0.27421217846920304,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002730947189958651,
+      "loss": 1.2235,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2742998702435776,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002730702911351317,
+      "loss": 1.2542,
+      "step": 3128
+    },
+    {
+      "epoch": 0.27438756201795217,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027304585341866687,
+      "loss": 1.212,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2744752537923267,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0027302140584869903,
+      "loss": 1.1597,
+      "step": 3130
+    },
+    {
+      "epoch": 0.27456294556670124,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0027299694842745793,
+      "loss": 1.2361,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2746506373410758,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002729724811571739,
+      "loss": 1.1911,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2747383291154503,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002729480040400783,
+      "loss": 1.2191,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2748260208898249,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002729235170784033,
+      "loss": 1.1737,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2749137126641994,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0027289902027438227,
+      "loss": 1.1687,
+      "step": 3135
+    },
+    {
+      "epoch": 0.27500140443857396,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0027287451363024917,
+      "loss": 1.1117,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2750890962129485,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002728499971482388,
+      "loss": 1.2436,
+      "step": 3137
+    },
+    {
+      "epoch": 0.27517678798732303,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0027282547083058713,
+      "loss": 1.2373,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2752644797616976,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0027280093467953086,
+      "loss": 1.2195,
+      "step": 3139
+    },
+    {
+      "epoch": 0.27535217153607217,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027277638869730764,
+      "loss": 1.1732,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2754398633104467,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0027275183288615603,
+      "loss": 1.203,
+      "step": 3141
+    },
+    {
+      "epoch": 0.27552755508482124,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002727272672483154,
+      "loss": 1.1519,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2756152468591958,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0027270269178602605,
+      "loss": 1.157,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2757029386335703,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002726781065015293,
+      "loss": 1.1774,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2757906304079449,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002726535113970672,
+      "loss": 1.2341,
+      "step": 3145
+    },
+    {
+      "epoch": 0.27587832218231945,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0027262890647488276,
+      "loss": 1.2086,
+      "step": 3146
+    },
+    {
+      "epoch": 0.27596601395669396,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.002726042917372199,
+      "loss": 1.1979,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2760537057310685,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0027257966718632343,
+      "loss": 1.1944,
+      "step": 3148
+    },
+    {
+      "epoch": 0.27614139750544303,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0027255503282443898,
+      "loss": 1.2485,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2762290892798176,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0027253038865381316,
+      "loss": 1.1847,
+      "step": 3150
+    },
+    {
+      "epoch": 0.27631678105419216,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0027250573467669354,
+      "loss": 1.2059,
+      "step": 3151
+    },
+    {
+      "epoch": 0.27640447282856667,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0027248107089532844,
+      "loss": 1.2377,
+      "step": 3152
+    },
+    {
+      "epoch": 0.27649216460294124,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002724563973119671,
+      "loss": 1.2669,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2765798563773158,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0027243171392885976,
+      "loss": 1.2524,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2766675481516903,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002724070207482575,
+      "loss": 1.1837,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2767552399260649,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002723823177724121,
+      "loss": 1.1579,
+      "step": 3156
+    },
+    {
+      "epoch": 0.27684293170043944,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0027235760500357656,
+      "loss": 1.2003,
+      "step": 3157
+    },
+    {
+      "epoch": 0.27693062347481395,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002723328824440046,
+      "loss": 1.1946,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2770183152491885,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0027230815009595083,
+      "loss": 1.1877,
+      "step": 3159
+    },
+    {
+      "epoch": 0.27710600702356303,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002722834079616708,
+      "loss": 1.1973,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2771936987979376,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0027225865604342094,
+      "loss": 1.162,
+      "step": 3161
+    },
+    {
+      "epoch": 0.27728139057231216,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0027223389434345847,
+      "loss": 1.2216,
+      "step": 3162
+    },
+    {
+      "epoch": 0.27736908234668667,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002722091228640417,
+      "loss": 1.206,
+      "step": 3163
+    },
+    {
+      "epoch": 0.27745677412106123,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0027218434160742972,
+      "loss": 1.2357,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2775444658954358,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0027215955057588244,
+      "loss": 1.1718,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2776321576698103,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0027213474977166078,
+      "loss": 1.2912,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2777198494441849,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0027210993919702647,
+      "loss": 1.2118,
+      "step": 3167
+    },
+    {
+      "epoch": 0.27780754121855944,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002720851188542423,
+      "loss": 1.2573,
+      "step": 3168
+    },
+    {
+      "epoch": 0.27789523299293395,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0027206028874557165,
+      "loss": 1.1541,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2779829247673085,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0027203544887327907,
+      "loss": 1.294,
+      "step": 3170
+    },
+    {
+      "epoch": 0.278070616541683,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0027201059923962984,
+      "loss": 1.2322,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2781583083160576,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.002719857398468902,
+      "loss": 1.2167,
+      "step": 3172
+    },
+    {
+      "epoch": 0.27824600009043216,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0027196087069732734,
+      "loss": 1.2141,
+      "step": 3173
+    },
+    {
+      "epoch": 0.27833369186480666,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002719359917932091,
+      "loss": 1.2145,
+      "step": 3174
+    },
+    {
+      "epoch": 0.27842138363918123,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002719111031368045,
+      "loss": 1.2392,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2785090754135558,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002718862047303833,
+      "loss": 1.2095,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2785967671879303,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0027186129657621617,
+      "loss": 1.1472,
+      "step": 3177
+    },
+    {
+      "epoch": 0.27868445896230487,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002718363786765746,
+      "loss": 1.1702,
+      "step": 3178
+    },
+    {
+      "epoch": 0.27877215073667944,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0027181145103373112,
+      "loss": 1.2109,
+      "step": 3179
+    },
+    {
+      "epoch": 0.27885984251105395,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0027178651364995906,
+      "loss": 1.2122,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2789475342854285,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0027176156652753258,
+      "loss": 1.1962,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2790352260598031,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0027173660966872686,
+      "loss": 1.1764,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2791229178341776,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002717116430758178,
+      "loss": 1.2399,
+      "step": 3183
+    },
+    {
+      "epoch": 0.27921060960855215,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0027168666675108253,
+      "loss": 1.2251,
+      "step": 3184
+    },
+    {
+      "epoch": 0.27929830138292666,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027166168069679857,
+      "loss": 1.1583,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2793859931573012,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0027163668491524463,
+      "loss": 1.2055,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2794736849316758,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002716116794087004,
+      "loss": 1.2419,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2795613767060503,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0027158666417944617,
+      "loss": 1.2447,
+      "step": 3188
+    },
+    {
+      "epoch": 0.27964906848042487,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002715616392297633,
+      "loss": 1.2102,
+      "step": 3189
+    },
+    {
+      "epoch": 0.27973676025479943,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0027153660456193406,
+      "loss": 1.2258,
+      "step": 3190
+    },
+    {
+      "epoch": 0.27982445202917394,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002715115601782415,
+      "loss": 1.1866,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2799121438035485,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027148650608096957,
+      "loss": 1.2743,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2799998355779231,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002714614422724032,
+      "loss": 1.2109,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2800875273522976,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0027143636875482818,
+      "loss": 1.1708,
+      "step": 3194
+    },
+    {
+      "epoch": 0.28017521912667215,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00271411285530531,
+      "loss": 1.1406,
+      "step": 3195
+    },
+    {
+      "epoch": 0.28026291090104666,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0027138619260179933,
+      "loss": 1.2615,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2803506026754212,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002713610899709216,
+      "loss": 1.1542,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2804382944497958,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0027133597764018693,
+      "loss": 1.2493,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2805259862241703,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002713108556118856,
+      "loss": 1.1763,
+      "step": 3199
+    },
+    {
+      "epoch": 0.28061367799854486,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002712857238883087,
+      "loss": 1.1743,
+      "step": 3200
+    },
+    {
+      "epoch": 0.28070136977291943,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0027126058247174826,
+      "loss": 1.1941,
+      "step": 3201
+    },
+    {
+      "epoch": 0.28078906154729394,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0027123543136449687,
+      "loss": 1.2341,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2808767533216685,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002712102705688484,
+      "loss": 1.2625,
+      "step": 3203
+    },
+    {
+      "epoch": 0.28096444509604307,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002711851000870975,
+      "loss": 1.1388,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2810521368704176,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0027115991992153953,
+      "loss": 1.1956,
+      "step": 3205
+    },
+    {
+      "epoch": 0.28113982864479214,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027113473007447098,
+      "loss": 1.187,
+      "step": 3206
+    },
+    {
+      "epoch": 0.28122752041916665,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00271109530548189,
+      "loss": 1.243,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2813152121935412,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002710843213449917,
+      "loss": 1.1944,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2814029039679158,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0027105910246717825,
+      "loss": 1.1856,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2814905957422903,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002710338739170484,
+      "loss": 1.1999,
+      "step": 3210
+    },
+    {
+      "epoch": 0.28157828751666486,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002710086356969029,
+      "loss": 1.2298,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2816659792910394,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002709833878090436,
+      "loss": 1.166,
+      "step": 3212
+    },
+    {
+      "epoch": 0.28175367106541394,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0027095813025577284,
+      "loss": 1.2138,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2818413628397885,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002709328630393942,
+      "loss": 1.2286,
+      "step": 3214
+    },
+    {
+      "epoch": 0.28192905461416307,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0027090758616221183,
+      "loss": 1.1712,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2820167463885376,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0027088229962653108,
+      "loss": 1.2474,
+      "step": 3216
+    },
+    {
+      "epoch": 0.28210443816291214,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002708570034346579,
+      "loss": 1.1507,
+      "step": 3217
+    },
+    {
+      "epoch": 0.2821921299372867,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002708316975888993,
+      "loss": 1.167,
+      "step": 3218
+    },
+    {
+      "epoch": 0.2822798217116612,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0027080638209156303,
+      "loss": 1.1839,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2823675134860358,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0027078105694495792,
+      "loss": 1.1495,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2824552052604103,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002707557221513935,
+      "loss": 1.219,
+      "step": 3221
+    },
+    {
+      "epoch": 0.28254289703478486,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0027073037771318015,
+      "loss": 1.1879,
+      "step": 3222
+    },
+    {
+      "epoch": 0.2826305888091594,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002707050236326293,
+      "loss": 1.1974,
+      "step": 3223
+    },
+    {
+      "epoch": 0.28271828058353393,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0027067965991205323,
+      "loss": 1.2573,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2828059723579085,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002706542865537649,
+      "loss": 1.2002,
+      "step": 3225
+    },
+    {
+      "epoch": 0.28289366413228306,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002706289035600785,
+      "loss": 1.2129,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2829813559066576,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002706035109333087,
+      "loss": 1.2783,
+      "step": 3227
+    },
+    {
+      "epoch": 0.28306904768103214,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002705781086757713,
+      "loss": 1.3446,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2831567394554067,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00270552696789783,
+      "loss": 1.1802,
+      "step": 3229
+    },
+    {
+      "epoch": 0.2832444312297812,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027052727527766118,
+      "loss": 1.2642,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2833321230041558,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002705018441417243,
+      "loss": 1.2062,
+      "step": 3231
+    },
+    {
+      "epoch": 0.2834198147785303,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0027047640338429156,
+      "loss": 1.1977,
+      "step": 3232
+    },
+    {
+      "epoch": 0.28350750655290485,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0027045095300768317,
+      "loss": 1.2683,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2835951983272794,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0027042549301422004,
+      "loss": 1.2624,
+      "step": 3234
+    },
+    {
+      "epoch": 0.28368289010165393,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0027040002340622407,
+      "loss": 1.1793,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2837705818760285,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0027037454418601813,
+      "loss": 1.2047,
+      "step": 3236
+    },
+    {
+      "epoch": 0.28385827365040306,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0027034905535592573,
+      "loss": 1.2205,
+      "step": 3237
+    },
+    {
+      "epoch": 0.28394596542477757,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002703235569182715,
+      "loss": 1.2565,
+      "step": 3238
+    },
+    {
+      "epoch": 0.28403365719915213,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0027029804887538076,
+      "loss": 1.1787,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2841213489735267,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0027027253122957973,
+      "loss": 1.1655,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2842090407479012,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0027024700398319565,
+      "loss": 1.2108,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2842967325222758,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002702214671385565,
+      "loss": 1.2202,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2843844242966503,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002701959206979912,
+      "loss": 1.1777,
+      "step": 3243
+    },
+    {
+      "epoch": 0.28447211607102485,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027017036466382943,
+      "loss": 1.2016,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2845598078453994,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002701447990384019,
+      "loss": 1.2301,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2846474996197739,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0027011922382404016,
+      "loss": 1.2284,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2847351913941485,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0027009363902307654,
+      "loss": 1.2319,
+      "step": 3247
+    },
+    {
+      "epoch": 0.28482288316852306,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002700680446378444,
+      "loss": 1.2084,
+      "step": 3248
+    },
+    {
+      "epoch": 0.28491057494289757,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002700424406706778,
+      "loss": 1.2149,
+      "step": 3249
+    },
+    {
+      "epoch": 0.28499826671727213,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0027001682712391173,
+      "loss": 1.1756,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2850859584916467,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0026999120399988216,
+      "loss": 1.1716,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2851736502660212,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0026996557130092584,
+      "loss": 1.2102,
+      "step": 3252
+    },
+    {
+      "epoch": 0.28526134204039577,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0026993992902938036,
+      "loss": 1.2236,
+      "step": 3253
+    },
+    {
+      "epoch": 0.28534903381477034,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0026991427718758426,
+      "loss": 1.2205,
+      "step": 3254
+    },
+    {
+      "epoch": 0.28543672558914485,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002698886157778769,
+      "loss": 1.182,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2855244173635194,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002698629448025986,
+      "loss": 1.1327,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2856121091378939,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0026983726426409037,
+      "loss": 1.1479,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2856998009122685,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002698115741646943,
+      "loss": 1.1775,
+      "step": 3258
+    },
+    {
+      "epoch": 0.28578749268664305,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026978587450675324,
+      "loss": 1.1728,
+      "step": 3259
+    },
+    {
+      "epoch": 0.28587518446101756,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00269760165292611,
+      "loss": 1.2283,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2859628762353921,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00269734446524612,
+      "loss": 1.1522,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2860505680097667,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0026970871820510187,
+      "loss": 1.2,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2861382597841412,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00269682980336427,
+      "loss": 1.2017,
+      "step": 3263
+    },
+    {
+      "epoch": 0.28622595155851577,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0026965723292093455,
+      "loss": 1.2271,
+      "step": 3264
+    },
+    {
+      "epoch": 0.28631364333289033,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002696314759609726,
+      "loss": 1.1372,
+      "step": 3265
+    },
+    {
+      "epoch": 0.28640133510726484,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0026960570945889013,
+      "loss": 1.1639,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2864890268816394,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0026957993341703702,
+      "loss": 1.1657,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2865767186560139,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026955414783776397,
+      "loss": 1.2948,
+      "step": 3268
+    },
+    {
+      "epoch": 0.2866644104303885,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002695283527234225,
+      "loss": 1.2274,
+      "step": 3269
+    },
+    {
+      "epoch": 0.28675210220476305,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0026950254807636517,
+      "loss": 1.1702,
+      "step": 3270
+    },
+    {
+      "epoch": 0.28683979397913756,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0026947673389894518,
+      "loss": 1.2168,
+      "step": 3271
+    },
+    {
+      "epoch": 0.2869274857535121,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002694509101935168,
+      "loss": 1.1172,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2870151775278867,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0026942507696243495,
+      "loss": 1.2251,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2871028693022612,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002693992342080558,
+      "loss": 1.2433,
+      "step": 3274
+    },
+    {
+      "epoch": 0.28719056107663576,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002693733819327359,
+      "loss": 1.2236,
+      "step": 3275
+    },
+    {
+      "epoch": 0.28727825285101033,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00269347520138833,
+      "loss": 1.1721,
+      "step": 3276
+    },
+    {
+      "epoch": 0.28736594462538484,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0026932164882870566,
+      "loss": 1.2073,
+      "step": 3277
+    },
+    {
+      "epoch": 0.2874536363997594,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0026929576800471325,
+      "loss": 1.2224,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2875413281741339,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00269269877669216,
+      "loss": 1.2111,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2876290199485085,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0026924397782457517,
+      "loss": 1.1894,
+      "step": 3280
+    },
+    {
+      "epoch": 0.28771671172288305,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002692180684731526,
+      "loss": 1.2371,
+      "step": 3281
+    },
+    {
+      "epoch": 0.28780440349725755,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0026919214961731125,
+      "loss": 1.1913,
+      "step": 3282
+    },
+    {
+      "epoch": 0.2878920952716321,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002691662212594148,
+      "loss": 1.2234,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2879797870460067,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0026914028340182788,
+      "loss": 1.1804,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2880674788203812,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0026911433604691593,
+      "loss": 1.1984,
+      "step": 3285
+    },
+    {
+      "epoch": 0.28815517059475576,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0026908837919704534,
+      "loss": 1.2014,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2882428623691303,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002690624128545833,
+      "loss": 1.214,
+      "step": 3287
+    },
+    {
+      "epoch": 0.28833055414350484,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0026903643702189776,
+      "loss": 1.2022,
+      "step": 3288
+    },
+    {
+      "epoch": 0.2884182459178794,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0026901045170135782,
+      "loss": 1.1854,
+      "step": 3289
+    },
+    {
+      "epoch": 0.28850593769225397,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0026898445689533313,
+      "loss": 1.2531,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2885936294666285,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002689584526061944,
+      "loss": 1.1876,
+      "step": 3291
+    },
+    {
+      "epoch": 0.28868132124100304,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0026893243883631315,
+      "loss": 1.2351,
+      "step": 3292
+    },
+    {
+      "epoch": 0.28876901301537755,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0026890641558806184,
+      "loss": 1.2028,
+      "step": 3293
+    },
+    {
+      "epoch": 0.2888567047897521,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0026888038286381364,
+      "loss": 1.2048,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2889443965641267,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002688543406659427,
+      "loss": 1.1785,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2890320883385012,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026882828899682395,
+      "loss": 1.1823,
+      "step": 3296
+    },
+    {
+      "epoch": 0.28911978011287576,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002688022278588332,
+      "loss": 1.2352,
+      "step": 3297
+    },
+    {
+      "epoch": 0.2892074718872503,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0026877615725434735,
+      "loss": 1.2245,
+      "step": 3298
+    },
+    {
+      "epoch": 0.28929516366162483,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002687500771857439,
+      "loss": 1.2448,
+      "step": 3299
+    },
+    {
+      "epoch": 0.2893828554359994,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002687239876554011,
+      "loss": 1.2425,
+      "step": 3300
+    },
+    {
+      "epoch": 0.28947054721037396,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002686978886656985,
+      "loss": 1.2207,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2895582389847485,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002686717802190161,
+      "loss": 1.1927,
+      "step": 3302
+    },
+    {
+      "epoch": 0.28964593075912304,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00268645662317735,
+      "loss": 1.2497,
+      "step": 3303
+    },
+    {
+      "epoch": 0.28973362253349755,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00268619534964237,
+      "loss": 1.185,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2898213143078721,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00268593398160905,
+      "loss": 1.1985,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2899090060822467,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0026856725191012234,
+      "loss": 1.176,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2899966978566212,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0026854109621427384,
+      "loss": 1.2014,
+      "step": 3307
+    },
+    {
+      "epoch": 0.29008438963099575,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0026851493107574458,
+      "loss": 1.2225,
+      "step": 3308
+    },
+    {
+      "epoch": 0.2901720814053703,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002684887564969208,
+      "loss": 1.2287,
+      "step": 3309
+    },
+    {
+      "epoch": 0.29025977317974483,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002684625724801896,
+      "loss": 1.1743,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2903474649541194,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002684363790279389,
+      "loss": 1.2122,
+      "step": 3311
+    },
+    {
+      "epoch": 0.29043515672849396,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0026841017614255744,
+      "loss": 1.2261,
+      "step": 3312
+    },
+    {
+      "epoch": 0.29052284850286847,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002683839638264349,
+      "loss": 1.2134,
+      "step": 3313
+    },
+    {
+      "epoch": 0.29061054027724303,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0026835774208196174,
+      "loss": 1.2287,
+      "step": 3314
+    },
+    {
+      "epoch": 0.29069823205161754,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0026833151091152927,
+      "loss": 1.1888,
+      "step": 3315
+    },
+    {
+      "epoch": 0.2907859238259921,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002683052703175298,
+      "loss": 1.2475,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2908736156003667,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0026827902030235637,
+      "loss": 1.2276,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2909613073747412,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002682527608684028,
+      "loss": 1.1959,
+      "step": 3318
+    },
+    {
+      "epoch": 0.29104899914911575,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002682264920180641,
+      "loss": 1.2085,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2911366909234903,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0026820021375373575,
+      "loss": 1.2389,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2912243826978648,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0026817392607781433,
+      "loss": 1.2551,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2913120744722394,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002681476289926972,
+      "loss": 1.2067,
+      "step": 3322
+    },
+    {
+      "epoch": 0.29139976624661396,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0026812132250078257,
+      "loss": 1.2492,
+      "step": 3323
+    },
+    {
+      "epoch": 0.29148745802098847,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0026809500660446954,
+      "loss": 1.1585,
+      "step": 3324
+    },
+    {
+      "epoch": 0.29157514979536303,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0026806868130615806,
+      "loss": 1.2225,
+      "step": 3325
+    },
+    {
+      "epoch": 0.29166284156973754,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002680423466082489,
+      "loss": 1.2486,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2917505333441121,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0026801600251314364,
+      "loss": 1.1767,
+      "step": 3327
+    },
+    {
+      "epoch": 0.29183822511848667,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00267989649023245,
+      "loss": 1.2791,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2919259168928612,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026796328614095625,
+      "loss": 1.19,
+      "step": 3329
+    },
+    {
+      "epoch": 0.29201360866723575,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002679369138686816,
+      "loss": 1.1732,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2921013004416103,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0026791053220882603,
+      "loss": 1.1711,
+      "step": 3331
+    },
+    {
+      "epoch": 0.2921889922159848,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0026788414116379567,
+      "loss": 1.2378,
+      "step": 3332
+    },
+    {
+      "epoch": 0.2922766839903594,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002678577407359972,
+      "loss": 1.2414,
+      "step": 3333
+    },
+    {
+      "epoch": 0.29236437576473395,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0026783133092783836,
+      "loss": 1.1671,
+      "step": 3334
+    },
+    {
+      "epoch": 0.29245206753910846,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002678049117417275,
+      "loss": 1.2867,
+      "step": 3335
+    },
+    {
+      "epoch": 0.292539759313483,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0026777848318007415,
+      "loss": 1.1654,
+      "step": 3336
+    },
+    {
+      "epoch": 0.2926274510878576,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0026775204524528843,
+      "loss": 1.1988,
+      "step": 3337
+    },
+    {
+      "epoch": 0.2927151428622321,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0026772559793978145,
+      "loss": 1.188,
+      "step": 3338
+    },
+    {
+      "epoch": 0.29280283463660667,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002676991412659651,
+      "loss": 1.1591,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2928905264109812,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0026767267522625217,
+      "loss": 1.1497,
+      "step": 3340
+    },
+    {
+      "epoch": 0.29297821818535574,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0026764619982305635,
+      "loss": 1.2675,
+      "step": 3341
+    },
+    {
+      "epoch": 0.2930659099597303,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0026761971505879206,
+      "loss": 1.1982,
+      "step": 3342
+    },
+    {
+      "epoch": 0.2931536017341048,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002675932209358747,
+      "loss": 1.2707,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2932412935084794,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.002675667174567204,
+      "loss": 1.1987,
+      "step": 3344
+    },
+    {
+      "epoch": 0.29332898528285395,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026754020462374625,
+      "loss": 1.2035,
+      "step": 3345
+    },
+    {
+      "epoch": 0.29341667705722846,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0026751368243937016,
+      "loss": 1.1739,
+      "step": 3346
+    },
+    {
+      "epoch": 0.293504368831603,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0026748715090601084,
+      "loss": 1.2262,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2935920606059776,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00267460610026088,
+      "loss": 1.194,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2936797523803521,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0026743405980202196,
+      "loss": 1.2192,
+      "step": 3349
+    },
+    {
+      "epoch": 0.29376744415472666,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002674075002362341,
+      "loss": 1.2768,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2938551359291012,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002673809313311466,
+      "loss": 1.202,
+      "step": 3351
+    },
+    {
+      "epoch": 0.29394282770347574,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002673543530891824,
+      "loss": 1.1624,
+      "step": 3352
+    },
+    {
+      "epoch": 0.2940305194778503,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0026732776551276548,
+      "loss": 1.1743,
+      "step": 3353
+    },
+    {
+      "epoch": 0.2941182112522248,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0026730116860432048,
+      "loss": 1.2838,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2942059030265994,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00267274562366273,
+      "loss": 1.1809,
+      "step": 3355
+    },
+    {
+      "epoch": 0.29429359480097395,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002672479468010495,
+      "loss": 1.1581,
+      "step": 3356
+    },
+    {
+      "epoch": 0.29438128657534846,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0026722132191107717,
+      "loss": 1.2221,
+      "step": 3357
+    },
+    {
+      "epoch": 0.294468978349723,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0026719468769878414,
+      "loss": 1.1325,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2945566701240976,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002671680441665994,
+      "loss": 1.2125,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2946443618984721,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002671413913169528,
+      "loss": 1.1941,
+      "step": 3360
+    },
+    {
+      "epoch": 0.29473205367284666,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00267114729152275,
+      "loss": 1.2597,
+      "step": 3361
+    },
+    {
+      "epoch": 0.29481974544722117,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0026708805767499755,
+      "loss": 1.2055,
+      "step": 3362
+    },
+    {
+      "epoch": 0.29490743722159574,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0026706137688755267,
+      "loss": 1.2154,
+      "step": 3363
+    },
+    {
+      "epoch": 0.2949951289959703,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0026703468679237377,
+      "loss": 1.2426,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2950828207703448,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002670079873918948,
+      "loss": 1.2014,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2951705125447194,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0026698127868855074,
+      "loss": 1.127,
+      "step": 3366
+    },
+    {
+      "epoch": 0.29525820431909394,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002669545606847773,
+      "loss": 1.1882,
+      "step": 3367
+    },
+    {
+      "epoch": 0.29534589609346845,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0026692783338301117,
+      "loss": 1.1638,
+      "step": 3368
+    },
+    {
+      "epoch": 0.295433587867843,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002669010967856897,
+      "loss": 1.1676,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2955212796422176,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002668743508952513,
+      "loss": 1.2229,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2956089714165921,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00266847595714135,
+      "loss": 1.2415,
+      "step": 3371
+    },
+    {
+      "epoch": 0.29569666319096666,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00266820831244781,
+      "loss": 1.2129,
+      "step": 3372
+    },
+    {
+      "epoch": 0.2957843549653412,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0026679405748963,
+      "loss": 1.1786,
+      "step": 3373
+    },
+    {
+      "epoch": 0.29587204673971573,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0026676727445112372,
+      "loss": 1.2688,
+      "step": 3374
+    },
+    {
+      "epoch": 0.2959597385140903,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0026674048213170473,
+      "loss": 1.1816,
+      "step": 3375
+    },
+    {
+      "epoch": 0.2960474302884648,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0026671368053381643,
+      "loss": 1.2277,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2961351220628394,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0026668686965990304,
+      "loss": 1.2234,
+      "step": 3377
+    },
+    {
+      "epoch": 0.29622281383721394,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002666600495124096,
+      "loss": 1.2073,
+      "step": 3378
+    },
+    {
+      "epoch": 0.29631050561158845,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002666332200937821,
+      "loss": 1.1695,
+      "step": 3379
+    },
+    {
+      "epoch": 0.296398197385963,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.002666063814064673,
+      "loss": 1.1747,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2964858891603376,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0026657953345291282,
+      "loss": 1.2501,
+      "step": 3381
+    },
+    {
+      "epoch": 0.2965735809347121,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0026655267623556713,
+      "loss": 1.1619,
+      "step": 3382
+    },
+    {
+      "epoch": 0.29666127270908665,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002665258097568795,
+      "loss": 1.226,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2967489644834612,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002664989340193001,
+      "loss": 1.2084,
+      "step": 3384
+    },
+    {
+      "epoch": 0.29683665625783573,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002664720490252799,
+      "loss": 1.2736,
+      "step": 3385
+    },
+    {
+      "epoch": 0.2969243480322103,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026644515477727082,
+      "loss": 1.1786,
+      "step": 3386
+    },
+    {
+      "epoch": 0.2970120398065848,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002664182512777255,
+      "loss": 1.1718,
+      "step": 3387
+    },
+    {
+      "epoch": 0.29709973158095937,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002663913385290975,
+      "loss": 1.194,
+      "step": 3388
+    },
+    {
+      "epoch": 0.29718742335533394,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002663644165338411,
+      "loss": 1.1978,
+      "step": 3389
+    },
+    {
+      "epoch": 0.29727511512970844,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0026633748529441164,
+      "loss": 1.2503,
+      "step": 3390
+    },
+    {
+      "epoch": 0.297362806904083,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002663105448132651,
+      "loss": 1.2191,
+      "step": 3391
+    },
+    {
+      "epoch": 0.2974504986784576,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0026628359509285843,
+      "loss": 1.1757,
+      "step": 3392
+    },
+    {
+      "epoch": 0.2975381904528321,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0026625663613564932,
+      "loss": 1.1931,
+      "step": 3393
+    },
+    {
+      "epoch": 0.29762588222720665,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0026622966794409638,
+      "loss": 1.2169,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2977135740015812,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0026620269052065907,
+      "loss": 1.1785,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2978012657759557,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0026617570386779765,
+      "loss": 1.1919,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2978889575503303,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002661487079879732,
+      "loss": 1.2444,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2979766493247048,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002661217028836477,
+      "loss": 1.1825,
+      "step": 3398
+    },
+    {
+      "epoch": 0.29806434109907937,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0026609468855728394,
+      "loss": 1.2251,
+      "step": 3399
+    },
+    {
+      "epoch": 0.29815203287345393,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0026606766501134555,
+      "loss": 1.1071,
+      "step": 3400
+    },
+    {
+      "epoch": 0.29823972464782844,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00266040632248297,
+      "loss": 1.1983,
+      "step": 3401
+    },
+    {
+      "epoch": 0.298327416422203,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0026601359027060363,
+      "loss": 1.2267,
+      "step": 3402
+    },
+    {
+      "epoch": 0.29841510819657757,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0026598653908073166,
+      "loss": 1.2106,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2985027999709521,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0026595947868114796,
+      "loss": 1.1717,
+      "step": 3404
+    },
+    {
+      "epoch": 0.29859049174532665,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0026593240907432044,
+      "loss": 1.2218,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2986781835197012,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002659053302627178,
+      "loss": 1.1307,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2987658752940757,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0026587824224880946,
+      "loss": 1.2031,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2988535670684503,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002658511450350659,
+      "loss": 1.2752,
+      "step": 3408
+    },
+    {
+      "epoch": 0.29894125884282485,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0026582403862395825,
+      "loss": 1.2815,
+      "step": 3409
+    },
+    {
+      "epoch": 0.29902895061719936,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0026579692301795855,
+      "loss": 1.2104,
+      "step": 3410
+    },
+    {
+      "epoch": 0.29911664239157393,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0026576979821953963,
+      "loss": 1.2245,
+      "step": 3411
+    },
+    {
+      "epoch": 0.29920433416594844,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0026574266423117525,
+      "loss": 1.1756,
+      "step": 3412
+    },
+    {
+      "epoch": 0.299292025940323,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0026571552105534003,
+      "loss": 1.2237,
+      "step": 3413
+    },
+    {
+      "epoch": 0.29937971771469757,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026568836869450925,
+      "loss": 1.2208,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2994674094890721,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0026566120715115914,
+      "loss": 1.2336,
+      "step": 3415
+    },
+    {
+      "epoch": 0.29955510126344664,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0026563403642776684,
+      "loss": 1.1912,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2996427930378212,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002656068565268102,
+      "loss": 1.1874,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2997304848121957,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002655796674507679,
+      "loss": 1.252,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2998181765865703,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0026555246920211962,
+      "loss": 1.2643,
+      "step": 3419
+    },
+    {
+      "epoch": 0.29990586836094485,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0026552526178334574,
+      "loss": 1.2865,
+      "step": 3420
+    },
+    {
+      "epoch": 0.29999356013531936,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0026549804519692744,
+      "loss": 1.2064,
+      "step": 3421
+    },
+    {
+      "epoch": 0.3000812519096939,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0026547081944534694,
+      "loss": 1.2559,
+      "step": 3422
+    },
+    {
+      "epoch": 0.30016894368406843,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0026544358453108705,
+      "loss": 1.1246,
+      "step": 3423
+    },
+    {
+      "epoch": 0.300256635458443,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0026541634045663147,
+      "loss": 1.1764,
+      "step": 3424
+    },
+    {
+      "epoch": 0.30034432723281757,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0026538908722446492,
+      "loss": 1.2398,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3004320190071921,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0026536182483707276,
+      "loss": 1.2528,
+      "step": 3426
+    },
+    {
+      "epoch": 0.30051971078156664,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002653345532969412,
+      "loss": 1.1639,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3006074025559412,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0026530727260655747,
+      "loss": 1.2688,
+      "step": 3428
+    },
+    {
+      "epoch": 0.3006950943303157,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002652799827684094,
+      "loss": 1.1951,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3007827861046903,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0026525268378498577,
+      "loss": 1.1979,
+      "step": 3430
+    },
+    {
+      "epoch": 0.30087047787906485,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0026522537565877614,
+      "loss": 1.1978,
+      "step": 3431
+    },
+    {
+      "epoch": 0.30095816965343936,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00265198058392271,
+      "loss": 1.2794,
+      "step": 3432
+    },
+    {
+      "epoch": 0.3010458614278139,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0026517073198796164,
+      "loss": 1.1975,
+      "step": 3433
+    },
+    {
+      "epoch": 0.30113355320218843,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002651433964483401,
+      "loss": 1.2226,
+      "step": 3434
+    },
+    {
+      "epoch": 0.301221244976563,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0026511605177589933,
+      "loss": 1.1545,
+      "step": 3435
+    },
+    {
+      "epoch": 0.30130893675093756,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0026508869797313302,
+      "loss": 1.2286,
+      "step": 3436
+    },
+    {
+      "epoch": 0.30139662852531207,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002650613350425359,
+      "loss": 1.2159,
+      "step": 3437
+    },
+    {
+      "epoch": 0.30148432029968664,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002650339629866033,
+      "loss": 1.2146,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3015720120740612,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0026500658180783152,
+      "loss": 1.2633,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3016597038484357,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002649791915087176,
+      "loss": 1.2285,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3017473956228103,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0026495179209175958,
+      "loss": 1.1456,
+      "step": 3441
+    },
+    {
+      "epoch": 0.30183508739718484,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0026492438355945613,
+      "loss": 1.2146,
+      "step": 3442
+    },
+    {
+      "epoch": 0.30192277917155935,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0026489696591430684,
+      "loss": 1.1624,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3020104709459339,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0026486953915881213,
+      "loss": 1.2009,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3020981627203085,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0026484210329547325,
+      "loss": 1.1907,
+      "step": 3445
+    },
+    {
+      "epoch": 0.302185854494683,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002648146583267923,
+      "loss": 1.1996,
+      "step": 3446
+    },
+    {
+      "epoch": 0.30227354626905756,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002647872042552722,
+      "loss": 1.1837,
+      "step": 3447
+    },
+    {
+      "epoch": 0.30236123804343207,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0026475974108341664,
+      "loss": 1.1951,
+      "step": 3448
+    },
+    {
+      "epoch": 0.30244892981780663,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002647322688137303,
+      "loss": 1.2407,
+      "step": 3449
+    },
+    {
+      "epoch": 0.3025366215921812,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002647047874487184,
+      "loss": 1.1998,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3026243133665557,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0026467729699088736,
+      "loss": 1.2456,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3027120051409303,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002646497974427441,
+      "loss": 1.1674,
+      "step": 3452
+    },
+    {
+      "epoch": 0.30279969691530484,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0026462228880679657,
+      "loss": 1.221,
+      "step": 3453
+    },
+    {
+      "epoch": 0.30288738868967935,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0026459477108555346,
+      "loss": 1.2694,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3029750804640539,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0026456724428152436,
+      "loss": 1.1538,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3030627722384285,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002645397083972196,
+      "loss": 1.2451,
+      "step": 3456
+    },
+    {
+      "epoch": 0.303150464012803,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002645121634351504,
+      "loss": 1.1821,
+      "step": 3457
+    },
+    {
+      "epoch": 0.30323815578717755,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002644846093978287,
+      "loss": 1.2361,
+      "step": 3458
+    },
+    {
+      "epoch": 0.30332584756155206,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0026445704628776756,
+      "loss": 1.2057,
+      "step": 3459
+    },
+    {
+      "epoch": 0.30341353933592663,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002644294741074805,
+      "loss": 1.2027,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3035012311103012,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002644018928594821,
+      "loss": 1.1607,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3035889228846757,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002643743025462876,
+      "loss": 1.1377,
+      "step": 3462
+    },
+    {
+      "epoch": 0.30367661465905027,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002643467031704133,
+      "loss": 1.1791,
+      "step": 3463
+    },
+    {
+      "epoch": 0.30376430643342484,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0026431909473437613,
+      "loss": 1.211,
+      "step": 3464
+    },
+    {
+      "epoch": 0.30385199820779935,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0026429147724069393,
+      "loss": 1.2137,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3039396899821739,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0026426385069188525,
+      "loss": 1.2094,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3040273817565485,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0026423621509046973,
+      "loss": 1.2065,
+      "step": 3467
+    },
+    {
+      "epoch": 0.304115073530923,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026420857043896752,
+      "loss": 1.149,
+      "step": 3468
+    },
+    {
+      "epoch": 0.30420276530529755,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002641809167398998,
+      "loss": 1.2014,
+      "step": 3469
+    },
+    {
+      "epoch": 0.30429045707967206,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002641532539957885,
+      "loss": 1.1623,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3043781488540466,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0026412558220915644,
+      "loss": 1.186,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3044658406284212,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0026409790138252716,
+      "loss": 1.2664,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3045535324027957,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0026407021151842506,
+      "loss": 1.2906,
+      "step": 3473
+    },
+    {
+      "epoch": 0.30464122417717027,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002640425126193754,
+      "loss": 1.2422,
+      "step": 3474
+    },
+    {
+      "epoch": 0.30472891595154483,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002640148046879043,
+      "loss": 1.2103,
+      "step": 3475
+    },
+    {
+      "epoch": 0.30481660772591934,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.002639870877265386,
+      "loss": 1.2288,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3049042995002939,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0026395936173780605,
+      "loss": 1.1978,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3049919912746685,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0026393162672423516,
+      "loss": 1.2104,
+      "step": 3478
+    },
+    {
+      "epoch": 0.305079683049043,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002639038826883553,
+      "loss": 1.1958,
+      "step": 3479
+    },
+    {
+      "epoch": 0.30516737482341755,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002638761296326967,
+      "loss": 1.2448,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3052550665977921,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0026384836755979036,
+      "loss": 1.226,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3053427583721666,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00263820596472168,
+      "loss": 1.2219,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3054304501465412,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002637928163723623,
+      "loss": 1.1962,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3055181419209157,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0026376502726290683,
+      "loss": 1.2776,
+      "step": 3484
+    },
+    {
+      "epoch": 0.30560583369529026,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0026373722914633585,
+      "loss": 1.2405,
+      "step": 3485
+    },
+    {
+      "epoch": 0.30569352546966483,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026370942202518445,
+      "loss": 1.1839,
+      "step": 3486
+    },
+    {
+      "epoch": 0.30578121724403934,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002636816059019886,
+      "loss": 1.2358,
+      "step": 3487
+    },
+    {
+      "epoch": 0.3058689090184139,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00263653780779285,
+      "loss": 1.213,
+      "step": 3488
+    },
+    {
+      "epoch": 0.30595660079278847,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002636259466596113,
+      "loss": 1.2303,
+      "step": 3489
+    },
+    {
+      "epoch": 0.306044292567163,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0026359810354550585,
+      "loss": 1.1842,
+      "step": 3490
+    },
+    {
+      "epoch": 0.30613198434153754,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002635702514395079,
+      "loss": 1.1999,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3062196761159121,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002635423903441574,
+      "loss": 1.2215,
+      "step": 3492
+    },
+    {
+      "epoch": 0.3063073678902866,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002635145202619954,
+      "loss": 1.1474,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3063950596646612,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0026348664119556337,
+      "loss": 1.1982,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3064827514390357,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00263458753147404,
+      "loss": 1.2316,
+      "step": 3495
+    },
+    {
+      "epoch": 0.30657044321341026,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0026343085612006044,
+      "loss": 1.2483,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3066581349877848,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0026340295011607695,
+      "loss": 1.2385,
+      "step": 3497
+    },
+    {
+      "epoch": 0.30674582676215933,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0026337503513799844,
+      "loss": 1.2339,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3068335185365339,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0026334711118837055,
+      "loss": 1.2372,
+      "step": 3499
+    },
+    {
+      "epoch": 0.30692121031090847,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0026331917826974013,
+      "loss": 1.1969,
+      "step": 3500
+    },
+    {
+      "epoch": 0.30692121031090847,
+      "eval_loss": 1.2185108661651611,
+      "eval_runtime": 429.5269,
+      "eval_samples_per_second": 33.635,
+      "eval_steps_per_second": 8.409,
+      "step": 3500
+    },
+    {
+      "epoch": 0.307008902085283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0026329123638465443,
+      "loss": 1.2152,
+      "step": 3501
+    },
+    {
+      "epoch": 0.30709659385965754,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0026326328553566173,
+      "loss": 1.2259,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3071842856340321,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0026323532572531103,
+      "loss": 1.2248,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3072719774084066,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026320735695615225,
+      "loss": 1.2517,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3073596691827812,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00263179379230736,
+      "loss": 1.1755,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3074473609571557,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002631513925516138,
+      "loss": 1.1728,
+      "step": 3506
+    },
+    {
+      "epoch": 0.30753505273153026,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00263123396921338,
+      "loss": 1.1534,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3076227445059048,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002630953923424617,
+      "loss": 1.2124,
+      "step": 3508
+    },
+    {
+      "epoch": 0.30771043628027933,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0026306737881753887,
+      "loss": 1.2559,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3077981280546539,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0026303935634912423,
+      "loss": 1.1662,
+      "step": 3510
+    },
+    {
+      "epoch": 0.30788581982902846,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0026301132493977333,
+      "loss": 1.2256,
+      "step": 3511
+    },
+    {
+      "epoch": 0.30797351160340297,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0026298328459204265,
+      "loss": 1.1922,
+      "step": 3512
+    },
+    {
+      "epoch": 0.30806120337777754,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002629552353084894,
+      "loss": 1.2255,
+      "step": 3513
+    },
+    {
+      "epoch": 0.3081488951521521,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0026292717709167153,
+      "loss": 1.2495,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3082365869265266,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002628991099441479,
+      "loss": 1.2112,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3083242787009012,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002628710338684782,
+      "loss": 1.1647,
+      "step": 3516
+    },
+    {
+      "epoch": 0.30841197047527574,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002628429488672228,
+      "loss": 1.2251,
+      "step": 3517
+    },
+    {
+      "epoch": 0.30849966224965025,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0026281485494294306,
+      "loss": 1.1533,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3085873540240248,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002627867520982011,
+      "loss": 1.206,
+      "step": 3519
+    },
+    {
+      "epoch": 0.30867504579839933,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0026275864033555975,
+      "loss": 1.1906,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3087627375727739,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026273051965758284,
+      "loss": 1.1816,
+      "step": 3521
+    },
+    {
+      "epoch": 0.30885042934714846,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0026270239006683477,
+      "loss": 1.1512,
+      "step": 3522
+    },
+    {
+      "epoch": 0.30893812112152297,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002626742515658809,
+      "loss": 1.1593,
+      "step": 3523
+    },
+    {
+      "epoch": 0.30902581289589753,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0026264610415728746,
+      "loss": 1.2025,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3091135046702721,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0026261794784362144,
+      "loss": 1.2547,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3092011964446466,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002625897826274505,
+      "loss": 1.1875,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3092888882190212,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0026256160851134335,
+      "loss": 1.2426,
+      "step": 3527
+    },
+    {
+      "epoch": 0.30937657999339574,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0026253342549786935,
+      "loss": 1.2302,
+      "step": 3528
+    },
+    {
+      "epoch": 0.30946427176777025,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002625052335895988,
+      "loss": 1.1756,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3095519635421448,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0026247703278910253,
+      "loss": 1.256,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3096396553165193,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002624488230989526,
+      "loss": 1.2101,
+      "step": 3531
+    },
+    {
+      "epoch": 0.3097273470908939,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026242060452172154,
+      "loss": 1.1964,
+      "step": 3532
+    },
+    {
+      "epoch": 0.30981503886526846,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.002623923770599828,
+      "loss": 1.1782,
+      "step": 3533
+    },
+    {
+      "epoch": 0.30990273063964296,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0026236414071631074,
+      "loss": 1.264,
+      "step": 3534
+    },
+    {
+      "epoch": 0.30999042241401753,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0026233589549328035,
+      "loss": 1.1518,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3100781141883921,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0026230764139346754,
+      "loss": 1.1993,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3101658059627666,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002622793784194491,
+      "loss": 1.2011,
+      "step": 3537
+    },
+    {
+      "epoch": 0.31025349773714117,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002622511065738024,
+      "loss": 1.2317,
+      "step": 3538
+    },
+    {
+      "epoch": 0.31034118951151574,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002622228258591059,
+      "loss": 1.2117,
+      "step": 3539
+    },
+    {
+      "epoch": 0.31042888128589025,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026219453627793865,
+      "loss": 1.2107,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3105165730602648,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002621662378328806,
+      "loss": 1.2337,
+      "step": 3541
+    },
+    {
+      "epoch": 0.3106042648346393,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002621379305265125,
+      "loss": 1.2475,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3106919566090139,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0026210961436141586,
+      "loss": 1.2156,
+      "step": 3543
+    },
+    {
+      "epoch": 0.31077964838338845,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002620812893401731,
+      "loss": 1.1861,
+      "step": 3544
+    },
+    {
+      "epoch": 0.31086734015776296,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002620529554653674,
+      "loss": 1.1988,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3109550319321375,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002620246127395826,
+      "loss": 1.2996,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3110427237065121,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002619962611654037,
+      "loss": 1.1323,
+      "step": 3547
+    },
+    {
+      "epoch": 0.3111304154808866,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002619679007454161,
+      "loss": 1.1999,
+      "step": 3548
+    },
+    {
+      "epoch": 0.31121810725526117,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002619395314822063,
+      "loss": 1.1919,
+      "step": 3549
+    },
+    {
+      "epoch": 0.31130579902963573,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002619111533783615,
+      "loss": 1.165,
+      "step": 3550
+    },
+    {
+      "epoch": 0.31139349080401024,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0026188276643646963,
+      "loss": 1.1591,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3114811825783848,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0026185437065911962,
+      "loss": 1.2056,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3115688743527593,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0026182596604890102,
+      "loss": 1.1502,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3116565661271339,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0026179755260840425,
+      "loss": 1.2779,
+      "step": 3554
+    },
+    {
+      "epoch": 0.31174425790150845,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002617691303402206,
+      "loss": 1.226,
+      "step": 3555
+    },
+    {
+      "epoch": 0.31183194967588296,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00261740699246942,
+      "loss": 1.2453,
+      "step": 3556
+    },
+    {
+      "epoch": 0.3119196414502575,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0026171225933116138,
+      "loss": 1.2301,
+      "step": 3557
+    },
+    {
+      "epoch": 0.3120073332246321,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0026168381059547244,
+      "loss": 1.2133,
+      "step": 3558
+    },
+    {
+      "epoch": 0.3120950249990066,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002616553530424695,
+      "loss": 1.1595,
+      "step": 3559
+    },
+    {
+      "epoch": 0.31218271677338116,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0026162688667474795,
+      "loss": 1.2748,
+      "step": 3560
+    },
+    {
+      "epoch": 0.31227040854775573,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0026159841149490373,
+      "loss": 1.1683,
+      "step": 3561
+    },
+    {
+      "epoch": 0.31235810032213024,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0026156992750553374,
+      "loss": 1.1743,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3124457920965048,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002615414347092356,
+      "loss": 1.2337,
+      "step": 3563
+    },
+    {
+      "epoch": 0.31253348387087937,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002615129331086079,
+      "loss": 1.2287,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3126211756452539,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0026148442270624984,
+      "loss": 1.2678,
+      "step": 3565
+    },
+    {
+      "epoch": 0.31270886741962844,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0026145590350476152,
+      "loss": 1.1903,
+      "step": 3566
+    },
+    {
+      "epoch": 0.31279655919400295,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0026142737550674377,
+      "loss": 1.1978,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3128842509683775,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0026139883871479836,
+      "loss": 1.2058,
+      "step": 3568
+    },
+    {
+      "epoch": 0.3129719427427521,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0026137029313152765,
+      "loss": 1.2322,
+      "step": 3569
+    },
+    {
+      "epoch": 0.3130596345171266,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0026134173875953497,
+      "loss": 1.1886,
+      "step": 3570
+    },
+    {
+      "epoch": 0.31314732629150116,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0026131317560142446,
+      "loss": 1.1288,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3132350180658757,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0026128460365980096,
+      "loss": 1.2598,
+      "step": 3572
+    },
+    {
+      "epoch": 0.31332270984025024,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026125602293727017,
+      "loss": 1.1925,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3134104016146248,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0026122743343643855,
+      "loss": 1.1877,
+      "step": 3574
+    },
+    {
+      "epoch": 0.31349809338899937,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0026119883515991348,
+      "loss": 1.1627,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3135857851633739,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002611702281103029,
+      "loss": 1.1539,
+      "step": 3576
+    },
+    {
+      "epoch": 0.31367347693774844,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026114161229021577,
+      "loss": 1.1983,
+      "step": 3577
+    },
+    {
+      "epoch": 0.31376116871212295,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0026111298770226184,
+      "loss": 1.2329,
+      "step": 3578
+    },
+    {
+      "epoch": 0.3138488604864975,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002610843543490515,
+      "loss": 1.1944,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3139365522608721,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0026105571223319613,
+      "loss": 1.1938,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3140242440352466,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0026102706135730774,
+      "loss": 1.2268,
+      "step": 3581
+    },
+    {
+      "epoch": 0.31411193580962116,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0026099840172399925,
+      "loss": 1.1975,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3141996275839957,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0026096973333588434,
+      "loss": 1.2414,
+      "step": 3583
+    },
+    {
+      "epoch": 0.31428731935837023,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0026094105619557753,
+      "loss": 1.128,
+      "step": 3584
+    },
+    {
+      "epoch": 0.3143750111327448,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0026091237030569404,
+      "loss": 1.1955,
+      "step": 3585
+    },
+    {
+      "epoch": 0.31446270290711936,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0026088367566885,
+      "loss": 1.2278,
+      "step": 3586
+    },
+    {
+      "epoch": 0.3145503946814939,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0026085497228766218,
+      "loss": 1.2631,
+      "step": 3587
+    },
+    {
+      "epoch": 0.31463808645586844,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002608262601647484,
+      "loss": 1.1882,
+      "step": 3588
+    },
+    {
+      "epoch": 0.31472577823024295,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0026079753930272707,
+      "loss": 1.2073,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3148134700046175,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002607688097042174,
+      "loss": 1.1955,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3149011617789921,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.002607400713718396,
+      "loss": 1.2355,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3149888535533666,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002607113243082144,
+      "loss": 1.25,
+      "step": 3592
+    },
+    {
+      "epoch": 0.31507654532774115,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0026068256851596344,
+      "loss": 1.1725,
+      "step": 3593
+    },
+    {
+      "epoch": 0.3151642371021157,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0026065380399770935,
+      "loss": 1.2078,
+      "step": 3594
+    },
+    {
+      "epoch": 0.31525192887649023,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0026062503075607517,
+      "loss": 1.2682,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3153396206508648,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002605962487936851,
+      "loss": 1.1699,
+      "step": 3596
+    },
+    {
+      "epoch": 0.31542731242523936,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0026056745811316386,
+      "loss": 1.1881,
+      "step": 3597
+    },
+    {
+      "epoch": 0.31551500419961387,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002605386587171372,
+      "loss": 1.19,
+      "step": 3598
+    },
+    {
+      "epoch": 0.31560269597398843,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0026050985060823146,
+      "loss": 1.2089,
+      "step": 3599
+    },
+    {
+      "epoch": 0.315690387748363,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002604810337890739,
+      "loss": 1.1829,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3157780795227375,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002604522082622926,
+      "loss": 1.2315,
+      "step": 3601
+    },
+    {
+      "epoch": 0.3158657712971121,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0026042337403051627,
+      "loss": 1.2491,
+      "step": 3602
+    },
+    {
+      "epoch": 0.3159534630714866,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002603945310963746,
+      "loss": 1.2127,
+      "step": 3603
+    },
+    {
+      "epoch": 0.31604115484586115,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002603656794624979,
+      "loss": 1.1826,
+      "step": 3604
+    },
+    {
+      "epoch": 0.3161288466202357,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0026033681913151746,
+      "loss": 1.2227,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3162165383946102,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002603079501060652,
+      "loss": 1.2382,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3163042301689848,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026027907238877395,
+      "loss": 1.1835,
+      "step": 3607
+    },
+    {
+      "epoch": 0.31639192194335936,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002602501859822772,
+      "loss": 1.2064,
+      "step": 3608
+    },
+    {
+      "epoch": 0.31647961371773387,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0026022129088920944,
+      "loss": 1.1548,
+      "step": 3609
+    },
+    {
+      "epoch": 0.31656730549210843,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002601923871122057,
+      "loss": 1.1669,
+      "step": 3610
+    },
+    {
+      "epoch": 0.316654997266483,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0026016347465390205,
+      "loss": 1.2062,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3167426890408575,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0026013455351693517,
+      "loss": 1.2366,
+      "step": 3612
+    },
+    {
+      "epoch": 0.31683038081523207,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002601056237039426,
+      "loss": 1.1988,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3169180725896066,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0026007668521756265,
+      "loss": 1.2078,
+      "step": 3614
+    },
+    {
+      "epoch": 0.31700576436398115,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0026004773806043444,
+      "loss": 1.265,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3170934561383557,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0026001878223519794,
+      "loss": 1.2037,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3171811479127302,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0025998981774449374,
+      "loss": 1.2352,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3172688396871048,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0025996084459096336,
+      "loss": 1.1897,
+      "step": 3618
+    },
+    {
+      "epoch": 0.31735653146147935,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002599318627772491,
+      "loss": 1.1959,
+      "step": 3619
+    },
+    {
+      "epoch": 0.31744422323585386,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0025990287230599402,
+      "loss": 1.2534,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3175319150102284,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0025987387317984198,
+      "loss": 1.1965,
+      "step": 3621
+    },
+    {
+      "epoch": 0.317619606784603,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0025984486540143767,
+      "loss": 1.1945,
+      "step": 3622
+    },
+    {
+      "epoch": 0.3177072985589775,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002598158489734264,
+      "loss": 1.2277,
+      "step": 3623
+    },
+    {
+      "epoch": 0.31779499033335207,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0025978682389845454,
+      "loss": 1.2335,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3178826821077266,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0025975779017916897,
+      "loss": 1.188,
+      "step": 3625
+    },
+    {
+      "epoch": 0.31797037388210114,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002597287478182176,
+      "loss": 1.1692,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3180580656564757,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002596996968182489,
+      "loss": 1.2101,
+      "step": 3627
+    },
+    {
+      "epoch": 0.3181457574308502,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0025967063718191244,
+      "loss": 1.1681,
+      "step": 3628
+    },
+    {
+      "epoch": 0.3182334492052248,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025964156891185816,
+      "loss": 1.1548,
+      "step": 3629
+    },
+    {
+      "epoch": 0.31832114097959935,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002596124920107371,
+      "loss": 1.2024,
+      "step": 3630
+    },
+    {
+      "epoch": 0.31840883275397386,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0025958340648120103,
+      "loss": 1.2057,
+      "step": 3631
+    },
+    {
+      "epoch": 0.3184965245283484,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002595543123259025,
+      "loss": 1.2492,
+      "step": 3632
+    },
+    {
+      "epoch": 0.318584216302723,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0025952520954749475,
+      "loss": 1.1861,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3186719080770975,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002594960981486319,
+      "loss": 1.2378,
+      "step": 3634
+    },
+    {
+      "epoch": 0.31875959985147206,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.002594669781319688,
+      "loss": 1.2508,
+      "step": 3635
+    },
+    {
+      "epoch": 0.31884729162584663,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0025943784950016122,
+      "loss": 1.2439,
+      "step": 3636
+    },
+    {
+      "epoch": 0.31893498340022114,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025940871225586556,
+      "loss": 1.1863,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3190226751745957,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00259379566401739,
+      "loss": 1.2902,
+      "step": 3638
+    },
+    {
+      "epoch": 0.3191103669489702,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002593504119404397,
+      "loss": 1.1508,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3191980587233448,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002593212488746263,
+      "loss": 1.2137,
+      "step": 3640
+    },
+    {
+      "epoch": 0.31928575049771935,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002592920772069586,
+      "loss": 1.2045,
+      "step": 3641
+    },
+    {
+      "epoch": 0.31937344227209385,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0025926289694009675,
+      "loss": 1.1515,
+      "step": 3642
+    },
+    {
+      "epoch": 0.3194611340464684,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0025923370807670217,
+      "loss": 1.2219,
+      "step": 3643
+    },
+    {
+      "epoch": 0.319548825820843,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002592045106194366,
+      "loss": 1.1632,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3196365175952175,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0025917530457096284,
+      "loss": 1.224,
+      "step": 3645
+    },
+    {
+      "epoch": 0.31972420936959206,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0025914608993394444,
+      "loss": 1.1729,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3198119011439666,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0025911686671104563,
+      "loss": 1.1572,
+      "step": 3647
+    },
+    {
+      "epoch": 0.31989959291834114,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002590876349049316,
+      "loss": 1.2318,
+      "step": 3648
+    },
+    {
+      "epoch": 0.3199872846927157,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0025905839451826817,
+      "loss": 1.2542,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3200749764670902,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0025902914555372193,
+      "loss": 1.2318,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3201626682414648,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002589998880139603,
+      "loss": 1.201,
+      "step": 3651
+    },
+    {
+      "epoch": 0.32025036001583934,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002589706219016516,
+      "loss": 1.322,
+      "step": 3652
+    },
+    {
+      "epoch": 0.32033805179021385,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002589413472194648,
+      "loss": 1.1562,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3204257435645884,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002589120639700696,
+      "loss": 1.1722,
+      "step": 3654
+    },
+    {
+      "epoch": 0.320513435338963,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002588827721561366,
+      "loss": 1.1912,
+      "step": 3655
+    },
+    {
+      "epoch": 0.3206011271133375,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0025885347178033714,
+      "loss": 1.2354,
+      "step": 3656
+    },
+    {
+      "epoch": 0.32068881888771206,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002588241628453433,
+      "loss": 1.2044,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3207765106620866,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0025879484535382804,
+      "loss": 1.2094,
+      "step": 3658
+    },
+    {
+      "epoch": 0.32086420243646113,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0025876551930846506,
+      "loss": 1.3157,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3209518942108357,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0025873618471192873,
+      "loss": 1.2161,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3210395859852102,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002587068415668943,
+      "loss": 1.226,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3211272777595848,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0025867748987603786,
+      "loss": 1.1852,
+      "step": 3662
+    },
+    {
+      "epoch": 0.32121496953395934,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0025864812964203616,
+      "loss": 1.171,
+      "step": 3663
+    },
+    {
+      "epoch": 0.32130266130833385,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0025861876086756682,
+      "loss": 1.2078,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3213903530827084,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0025858938355530817,
+      "loss": 1.2128,
+      "step": 3665
+    },
+    {
+      "epoch": 0.321478044857083,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002585599977079393,
+      "loss": 1.182,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3215657366314575,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002585306033281402,
+      "loss": 1.1406,
+      "step": 3667
+    },
+    {
+      "epoch": 0.32165342840583205,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002585012004185915,
+      "loss": 1.1854,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3217411201802066,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002584717889819747,
+      "loss": 1.2627,
+      "step": 3669
+    },
+    {
+      "epoch": 0.32182881195458113,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0025844236902097203,
+      "loss": 1.3122,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3219165037289557,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002584129405382665,
+      "loss": 1.1991,
+      "step": 3671
+    },
+    {
+      "epoch": 0.32200419550333026,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0025838350353654194,
+      "loss": 1.2346,
+      "step": 3672
+    },
+    {
+      "epoch": 0.32209188727770477,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00258354058018483,
+      "loss": 1.1088,
+      "step": 3673
+    },
+    {
+      "epoch": 0.32217957905207933,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0025832460398677493,
+      "loss": 1.1771,
+      "step": 3674
+    },
+    {
+      "epoch": 0.32226727082645384,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0025829514144410383,
+      "loss": 1.1773,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3223549626008284,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002582656703931567,
+      "loss": 1.181,
+      "step": 3676
+    },
+    {
+      "epoch": 0.322442654375203,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002582361908366212,
+      "loss": 1.2021,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3225303461495775,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002582067027771858,
+      "loss": 1.2342,
+      "step": 3678
+    },
+    {
+      "epoch": 0.32261803792395205,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002581772062175397,
+      "loss": 1.176,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3227057296983266,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00258147701160373,
+      "loss": 1.2086,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3227934214727011,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0025811818760837634,
+      "loss": 1.2242,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3228811132470757,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0025808866556424137,
+      "loss": 1.2311,
+      "step": 3682
+    },
+    {
+      "epoch": 0.32296880502145026,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002580591350306604,
+      "loss": 1.2864,
+      "step": 3683
+    },
+    {
+      "epoch": 0.32305649679582477,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0025802959601032654,
+      "loss": 1.2212,
+      "step": 3684
+    },
+    {
+      "epoch": 0.32314418857019933,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002580000485059337,
+      "loss": 1.1889,
+      "step": 3685
+    },
+    {
+      "epoch": 0.32323188034457384,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002579704925201765,
+      "loss": 1.2047,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3233195721189484,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0025794092805575043,
+      "loss": 1.2856,
+      "step": 3687
+    },
+    {
+      "epoch": 0.32340726389332297,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0025791135511535153,
+      "loss": 1.1939,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3234949556676975,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0025788177370167703,
+      "loss": 1.1674,
+      "step": 3689
+    },
+    {
+      "epoch": 0.32358264744207205,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025785218381742447,
+      "loss": 1.1952,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3236703392164466,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0025782258546529243,
+      "loss": 1.1901,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3237580309908211,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002577929786479803,
+      "loss": 1.205,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3238457227651957,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0025776336336818797,
+      "loss": 1.2244,
+      "step": 3693
+    },
+    {
+      "epoch": 0.32393341453957025,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002577337396286164,
+      "loss": 1.2313,
+      "step": 3694
+    },
+    {
+      "epoch": 0.32402110631394476,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0025770410743196715,
+      "loss": 1.1898,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3241087980883193,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0025767446678094264,
+      "loss": 1.196,
+      "step": 3696
+    },
+    {
+      "epoch": 0.32419648986269384,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0025764481767824596,
+      "loss": 1.186,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3242841816370684,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002576151601265811,
+      "loss": 1.2333,
+      "step": 3698
+    },
+    {
+      "epoch": 0.32437187341144297,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0025758549412865277,
+      "loss": 1.2599,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3244595651858175,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0025755581968716634,
+      "loss": 1.2609,
+      "step": 3700
+    },
+    {
+      "epoch": 0.32454725696019204,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002575261368048281,
+      "loss": 1.2195,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3246349487345666,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0025749644548434497,
+      "loss": 1.1784,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3247226405089411,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0025746674572842483,
+      "loss": 1.1971,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3248103322833157,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002574370375397762,
+      "loss": 1.1759,
+      "step": 3704
+    },
+    {
+      "epoch": 0.32489802405769025,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0025740732092110834,
+      "loss": 1.1879,
+      "step": 3705
+    },
+    {
+      "epoch": 0.32498571583206476,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0025737759587513144,
+      "loss": 1.2027,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3250734076064393,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002573478624045562,
+      "loss": 1.2157,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3251610993808139,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002573181205120943,
+      "loss": 1.1949,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3252487911551884,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0025728837020045814,
+      "loss": 1.2089,
+      "step": 3709
+    },
+    {
+      "epoch": 0.32533648292956296,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0025725861147236086,
+      "loss": 1.2611,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3254241747039375,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0025722884433051637,
+      "loss": 1.1872,
+      "step": 3711
+    },
+    {
+      "epoch": 0.32551186647831204,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002571990687776394,
+      "loss": 1.2293,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3255995582526866,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002571692848164453,
+      "loss": 1.1905,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3256872500270611,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002571394924496504,
+      "loss": 1.2039,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3257749418014357,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002571096916799716,
+      "loss": 1.2498,
+      "step": 3715
+    },
+    {
+      "epoch": 0.32586263357581025,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002570798825101268,
+      "loss": 1.2058,
+      "step": 3716
+    },
+    {
+      "epoch": 0.32595032535018476,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.002570500649428344,
+      "loss": 1.1403,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3260380171245593,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002570202389808136,
+      "loss": 1.2172,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3261257088989339,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0025699040462678464,
+      "loss": 1.2432,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3262134006733084,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002569605618834682,
+      "loss": 1.1899,
+      "step": 3720
+    },
+    {
+      "epoch": 0.32630109244768296,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0025693071075358594,
+      "loss": 1.2001,
+      "step": 3721
+    },
+    {
+      "epoch": 0.32638878422205747,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002569008512398602,
+      "loss": 1.1853,
+      "step": 3722
+    },
+    {
+      "epoch": 0.32647647599643204,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0025687098334501402,
+      "loss": 1.2448,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3265641677708066,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002568411070717714,
+      "loss": 1.1032,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3266518595451811,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002568112224228568,
+      "loss": 1.1911,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3267395513195557,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002567813294009958,
+      "loss": 1.2242,
+      "step": 3726
+    },
+    {
+      "epoch": 0.32682724309393024,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002567514280089144,
+      "loss": 1.246,
+      "step": 3727
+    },
+    {
+      "epoch": 0.32691493486830475,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0025672151824933974,
+      "loss": 1.1884,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3270026266426793,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002566916001249993,
+      "loss": 1.189,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3270903184170539,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002566616736386217,
+      "loss": 1.18,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3271780101914284,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0025663173879293603,
+      "loss": 1.2033,
+      "step": 3731
+    },
+    {
+      "epoch": 0.32726570196580296,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.002566017955906724,
+      "loss": 1.1358,
+      "step": 3732
+    },
+    {
+      "epoch": 0.32735339374017747,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002565718440345614,
+      "loss": 1.3151,
+      "step": 3733
+    },
+    {
+      "epoch": 0.32744108551455203,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002565418841273347,
+      "loss": 1.1943,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3275287772889266,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002565119158717244,
+      "loss": 1.2604,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3276164690633011,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002564819392704636,
+      "loss": 1.1909,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3277041608376757,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0025645195432628617,
+      "loss": 1.2115,
+      "step": 3737
+    },
+    {
+      "epoch": 0.32779185261205024,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0025642196104192654,
+      "loss": 1.2065,
+      "step": 3738
+    },
+    {
+      "epoch": 0.32787954438642475,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002563919594201201,
+      "loss": 1.151,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3279672361607993,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002563619494636029,
+      "loss": 1.2012,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3280549279351739,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002563319311751117,
+      "loss": 1.1898,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3281426197095484,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0025630190455738417,
+      "loss": 1.1545,
+      "step": 3742
+    },
+    {
+      "epoch": 0.32823031148392295,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0025627186961315867,
+      "loss": 1.2197,
+      "step": 3743
+    },
+    {
+      "epoch": 0.32831800325829746,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0025624182634517425,
+      "loss": 1.1481,
+      "step": 3744
+    },
+    {
+      "epoch": 0.32840569503267203,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025621177475617083,
+      "loss": 1.2178,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3284933868070466,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00256181714848889,
+      "loss": 1.2462,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3285810785814211,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0025615164662607013,
+      "loss": 1.1869,
+      "step": 3747
+    },
+    {
+      "epoch": 0.32866877035579567,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0025612157009045644,
+      "loss": 1.1707,
+      "step": 3748
+    },
+    {
+      "epoch": 0.32875646213017024,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0025609148524479075,
+      "loss": 1.1469,
+      "step": 3749
+    },
+    {
+      "epoch": 0.32884415390454474,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002560613920918168,
+      "loss": 1.2231,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3289318456789193,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0025603129063427894,
+      "loss": 1.1746,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3290195374532939,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0025600118087492235,
+      "loss": 1.1842,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3291072292276684,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00255971062816493,
+      "loss": 1.196,
+      "step": 3753
+    },
+    {
+      "epoch": 0.32919492100204295,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025594093646173757,
+      "loss": 1.197,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3292826127764175,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0025591080181340355,
+      "loss": 1.1952,
+      "step": 3755
+    },
+    {
+      "epoch": 0.329370304550792,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0025588065887423898,
+      "loss": 1.2179,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3294579963251666,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00255850507646993,
+      "loss": 1.1883,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3295456880995411,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0025582034813441522,
+      "loss": 1.1891,
+      "step": 3758
+    },
+    {
+      "epoch": 0.32963337987391567,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0025579018033925617,
+      "loss": 1.1942,
+      "step": 3759
+    },
+    {
+      "epoch": 0.32972107164829023,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0025576000426426704,
+      "loss": 1.1921,
+      "step": 3760
+    },
+    {
+      "epoch": 0.32980876342266474,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002557298199121998,
+      "loss": 1.2723,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3298964551970393,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002556996272858072,
+      "loss": 1.2081,
+      "step": 3762
+    },
+    {
+      "epoch": 0.32998414697141387,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0025566942638784273,
+      "loss": 1.2951,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3300718387457884,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0025563921722106067,
+      "loss": 1.1731,
+      "step": 3764
+    },
+    {
+      "epoch": 0.33015953052016295,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0025560899978821594,
+      "loss": 1.2414,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3302472222945375,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0025557877409206438,
+      "loss": 1.2604,
+      "step": 3766
+    },
+    {
+      "epoch": 0.330334914068912,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002555485401353624,
+      "loss": 1.1924,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3304226058432866,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002555182979208674,
+      "loss": 1.2083,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3305102976176611,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002554880474513372,
+      "loss": 1.2136,
+      "step": 3769
+    },
+    {
+      "epoch": 0.33059798939203566,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0025545778872953073,
+      "loss": 1.2399,
+      "step": 3770
+    },
+    {
+      "epoch": 0.33068568116641023,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002554275217582074,
+      "loss": 1.2094,
+      "step": 3771
+    },
+    {
+      "epoch": 0.33077337294078474,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002553972465401275,
+      "loss": 1.2202,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3308610647151593,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002553669630780521,
+      "loss": 1.2021,
+      "step": 3773
+    },
+    {
+      "epoch": 0.33094875648953387,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00255336671374743,
+      "loss": 1.1743,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3310364482639084,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002553063714329626,
+      "loss": 1.2048,
+      "step": 3775
+    },
+    {
+      "epoch": 0.33112414003828294,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0025527606325547432,
+      "loss": 1.2572,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3312118318126575,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002552457468450421,
+      "loss": 1.1929,
+      "step": 3777
+    },
+    {
+      "epoch": 0.331299523587032,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002552154222044307,
+      "loss": 1.259,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3313872153614066,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0025518508933640572,
+      "loss": 1.2576,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3314749071357811,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002551547482437334,
+      "loss": 1.205,
+      "step": 3780
+    },
+    {
+      "epoch": 0.33156259891015566,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0025512439892918072,
+      "loss": 1.1563,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3316502906845302,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0025509404139551555,
+      "loss": 1.2718,
+      "step": 3782
+    },
+    {
+      "epoch": 0.33173798245890473,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002550636756455064,
+      "loss": 1.1583,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3318256742332793,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0025503330168192254,
+      "loss": 1.2247,
+      "step": 3784
+    },
+    {
+      "epoch": 0.33191336600765386,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0025500291950753392,
+      "loss": 1.2187,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3320010577820284,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0025497252912511145,
+      "loss": 1.1745,
+      "step": 3786
+    },
+    {
+      "epoch": 0.33208874955640294,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0025494213053742654,
+      "loss": 1.2522,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3321764413307775,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0025491172374725157,
+      "loss": 1.2106,
+      "step": 3788
+    },
+    {
+      "epoch": 0.332264133105152,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0025488130875735947,
+      "loss": 1.1873,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3323518248795266,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0025485088557052403,
+      "loss": 1.189,
+      "step": 3790
+    },
+    {
+      "epoch": 0.33243951665390115,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0025482045418951976,
+      "loss": 1.2034,
+      "step": 3791
+    },
+    {
+      "epoch": 0.33252720842827566,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0025479001461712193,
+      "loss": 1.2346,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3326149002026502,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002547595668561066,
+      "loss": 1.1971,
+      "step": 3793
+    },
+    {
+      "epoch": 0.33270259197702473,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002547291109092505,
+      "loss": 1.17,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3327902837513993,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002546986467793311,
+      "loss": 1.1872,
+      "step": 3795
+    },
+    {
+      "epoch": 0.33287797552577386,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0025466817446912664,
+      "loss": 1.1714,
+      "step": 3796
+    },
+    {
+      "epoch": 0.33296566730014837,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0025463769398141613,
+      "loss": 1.2313,
+      "step": 3797
+    },
+    {
+      "epoch": 0.33305335907452294,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002546072053189793,
+      "loss": 1.2128,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3331410508488975,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002545767084845968,
+      "loss": 1.2158,
+      "step": 3799
+    },
+    {
+      "epoch": 0.333228742623272,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002545462034810496,
+      "loss": 1.2332,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3333164343976466,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002545156903111199,
+      "loss": 1.226,
+      "step": 3801
+    },
+    {
+      "epoch": 0.33340412617202114,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002544851689775903,
+      "loss": 1.1971,
+      "step": 3802
+    },
+    {
+      "epoch": 0.33349181794639565,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025445463948324423,
+      "loss": 1.1959,
+      "step": 3803
+    },
+    {
+      "epoch": 0.3335795097207702,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00254424101830866,
+      "loss": 1.216,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3336672014951447,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002543935560232406,
+      "loss": 1.2151,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3337548932695193,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002543630020631536,
+      "loss": 1.2374,
+      "step": 3806
+    },
+    {
+      "epoch": 0.33384258504389386,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0025433243995339147,
+      "loss": 1.1865,
+      "step": 3807
+    },
+    {
+      "epoch": 0.33393027681826837,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025430186969674145,
+      "loss": 1.2509,
+      "step": 3808
+    },
+    {
+      "epoch": 0.33401796859264293,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025427129129599146,
+      "loss": 1.2154,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3341056603670175,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0025424070475393015,
+      "loss": 1.2328,
+      "step": 3810
+    },
+    {
+      "epoch": 0.334193352141392,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0025421011007334695,
+      "loss": 1.2164,
+      "step": 3811
+    },
+    {
+      "epoch": 0.3342810439157666,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0025417950725703203,
+      "loss": 1.2128,
+      "step": 3812
+    },
+    {
+      "epoch": 0.33436873569014114,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0025414889630777622,
+      "loss": 1.1729,
+      "step": 3813
+    },
+    {
+      "epoch": 0.33445642746451565,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002541182772283712,
+      "loss": 1.1531,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3345441192388902,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0025408765002160934,
+      "loss": 1.155,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3346318110132647,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0025405701469028382,
+      "loss": 1.2115,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3347195027876393,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0025402637123718846,
+      "loss": 1.2175,
+      "step": 3817
+    },
+    {
+      "epoch": 0.33480719456201385,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0025399571966511787,
+      "loss": 1.2269,
+      "step": 3818
+    },
+    {
+      "epoch": 0.33489488633638836,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025396505997686736,
+      "loss": 1.1774,
+      "step": 3819
+    },
+    {
+      "epoch": 0.33498257811076293,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00253934392175233,
+      "loss": 1.1798,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3350702698851375,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0025390371626301173,
+      "loss": 1.201,
+      "step": 3821
+    },
+    {
+      "epoch": 0.335157961659512,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00253873032243001,
+      "loss": 1.2311,
+      "step": 3822
+    },
+    {
+      "epoch": 0.33524565343388657,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002538423401179992,
+      "loss": 1.2466,
+      "step": 3823
+    },
+    {
+      "epoch": 0.33533334520826114,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0025381163989080533,
+      "loss": 1.2409,
+      "step": 3824
+    },
+    {
+      "epoch": 0.33542103698263565,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002537809315642191,
+      "loss": 1.2727,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3355087287570102,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002537502151410411,
+      "loss": 1.1955,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3355964205313848,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0025371949062407266,
+      "loss": 1.2691,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3356841123057593,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002536887580161157,
+      "loss": 1.1967,
+      "step": 3828
+    },
+    {
+      "epoch": 0.33577180408013385,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002536580173199729,
+      "loss": 1.1999,
+      "step": 3829
+    },
+    {
+      "epoch": 0.33585949585450836,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002536272685384478,
+      "loss": 1.2114,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3359471876288829,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002535965116743446,
+      "loss": 1.2365,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3360348794032575,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0025356574673046826,
+      "loss": 1.1783,
+      "step": 3832
+    },
+    {
+      "epoch": 0.336122571177632,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002535349737096244,
+      "loss": 1.1973,
+      "step": 3833
+    },
+    {
+      "epoch": 0.33621026295200657,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025350419261461962,
+      "loss": 1.1656,
+      "step": 3834
+    },
+    {
+      "epoch": 0.33629795472638113,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002534734034482608,
+      "loss": 1.2391,
+      "step": 3835
+    },
+    {
+      "epoch": 0.33638564650075564,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0025344260621335607,
+      "loss": 1.1438,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3364733382751302,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0025341180091271393,
+      "loss": 1.2051,
+      "step": 3837
+    },
+    {
+      "epoch": 0.3365610300495048,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0025338098754914377,
+      "loss": 1.1704,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3366487218238793,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0025335016612545573,
+      "loss": 1.1827,
+      "step": 3839
+    },
+    {
+      "epoch": 0.33673641359825385,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002533193366444606,
+      "loss": 1.2276,
+      "step": 3840
+    },
+    {
+      "epoch": 0.33682410537262836,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025328849910896995,
+      "loss": 1.2361,
+      "step": 3841
+    },
+    {
+      "epoch": 0.3369117971470029,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.002532576535217961,
+      "loss": 1.1686,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3369994889213775,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0025322679988575214,
+      "loss": 1.2593,
+      "step": 3843
+    },
+    {
+      "epoch": 0.337087180695752,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0025319593820365177,
+      "loss": 1.1464,
+      "step": 3844
+    },
+    {
+      "epoch": 0.33717487247012656,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0025316506847830947,
+      "loss": 1.2232,
+      "step": 3845
+    },
+    {
+      "epoch": 0.33726256424450113,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.002531341907125405,
+      "loss": 1.1705,
+      "step": 3846
+    },
+    {
+      "epoch": 0.33735025601887564,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002531033049091609,
+      "loss": 1.1531,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3374379477932502,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002530724110709873,
+      "loss": 1.1871,
+      "step": 3848
+    },
+    {
+      "epoch": 0.33752563956762477,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0025304150920083717,
+      "loss": 1.228,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3376133313419993,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0025301059930152876,
+      "loss": 1.2601,
+      "step": 3850
+    },
+    {
+      "epoch": 0.33770102311637384,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002529796813758808,
+      "loss": 1.2028,
+      "step": 3851
+    },
+    {
+      "epoch": 0.33778871489074835,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0025294875542671304,
+      "loss": 1.1734,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3378764066651229,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0025291782145684585,
+      "loss": 1.1952,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3379640984394975,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0025288687946910028,
+      "loss": 1.2404,
+      "step": 3854
+    },
+    {
+      "epoch": 0.338051790213872,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0025285592946629816,
+      "loss": 1.2223,
+      "step": 3855
+    },
+    {
+      "epoch": 0.33813948198824656,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002528249714512621,
+      "loss": 1.2059,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3382271737626211,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0025279400542681536,
+      "loss": 1.1894,
+      "step": 3857
+    },
+    {
+      "epoch": 0.33831486553699563,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00252763031395782,
+      "loss": 1.2305,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3384025573113702,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002527320493609867,
+      "loss": 1.2219,
+      "step": 3859
+    },
+    {
+      "epoch": 0.33849024908574477,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002527010593252549,
+      "loss": 1.1756,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3385779408601193,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025267006129141303,
+      "loss": 1.1833,
+      "step": 3861
+    },
+    {
+      "epoch": 0.33866563263449384,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0025263905526228783,
+      "loss": 1.169,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3387533244088684,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00252608041240707,
+      "loss": 1.2373,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3388410161832429,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00252577019229499,
+      "loss": 1.1814,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3389287079576175,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025254598923149293,
+      "loss": 1.2448,
+      "step": 3865
+    },
+    {
+      "epoch": 0.339016399731992,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002525149512495186,
+      "loss": 1.211,
+      "step": 3866
+    },
+    {
+      "epoch": 0.33910409150636656,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0025248390528640665,
+      "loss": 1.1562,
+      "step": 3867
+    },
+    {
+      "epoch": 0.3391917832807411,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002524528513449884,
+      "loss": 1.2184,
+      "step": 3868
+    },
+    {
+      "epoch": 0.33927947505511563,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0025242178942809584,
+      "loss": 1.169,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3393671668294902,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002523907195385618,
+      "loss": 1.1899,
+      "step": 3870
+    },
+    {
+      "epoch": 0.33945485860386476,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002523596416792197,
+      "loss": 1.2032,
+      "step": 3871
+    },
+    {
+      "epoch": 0.33954255037823927,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002523285558529038,
+      "loss": 1.2399,
+      "step": 3872
+    },
+    {
+      "epoch": 0.33963024215261384,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0025229746206244905,
+      "loss": 1.2305,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3397179339269884,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0025226636031069117,
+      "loss": 1.2,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3398056257013629,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002522352506004664,
+      "loss": 1.2619,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3398933174757375,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0025220413293461203,
+      "loss": 1.184,
+      "step": 3876
+    },
+    {
+      "epoch": 0.339981009250112,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.002521730073159659,
+      "loss": 1.1855,
+      "step": 3877
+    },
+    {
+      "epoch": 0.34006870102448655,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0025214187374736647,
+      "loss": 1.1762,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3401563927988611,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002521107322316531,
+      "loss": 1.225,
+      "step": 3879
+    },
+    {
+      "epoch": 0.34024408457323563,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002520795827716659,
+      "loss": 1.2175,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3403317763476102,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002520484253702455,
+      "loss": 1.2177,
+      "step": 3881
+    },
+    {
+      "epoch": 0.34041946812198476,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0025201726003023337,
+      "loss": 1.1474,
+      "step": 3882
+    },
+    {
+      "epoch": 0.34050715989635927,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0025198608675447184,
+      "loss": 1.2183,
+      "step": 3883
+    },
+    {
+      "epoch": 0.34059485167073383,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002519549055458037,
+      "loss": 1.258,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3406825434451084,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0025192371640707267,
+      "loss": 1.2241,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3407702352194829,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002518925193411231,
+      "loss": 1.2153,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3408579269938575,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002518613143508001,
+      "loss": 1.1849,
+      "step": 3887
+    },
+    {
+      "epoch": 0.340945618768232,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002518301014389494,
+      "loss": 1.1763,
+      "step": 3888
+    },
+    {
+      "epoch": 0.34103331054260655,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002517988806084176,
+      "loss": 1.2465,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3411210023169811,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00251767651862052,
+      "loss": 1.1731,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3412086940913556,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0025173641520270054,
+      "loss": 1.1931,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3412963858657302,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0025170517063321197,
+      "loss": 1.1758,
+      "step": 3892
+    },
+    {
+      "epoch": 0.34138407764010475,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0025167391815643556,
+      "loss": 1.165,
+      "step": 3893
+    },
+    {
+      "epoch": 0.34147176941447926,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0025164265777522163,
+      "loss": 1.1666,
+      "step": 3894
+    },
+    {
+      "epoch": 0.34155946118885383,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0025161138949242098,
+      "loss": 1.1887,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3416471529632284,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0025158011331088514,
+      "loss": 1.2408,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3417348447376029,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0025154882923346658,
+      "loss": 1.2252,
+      "step": 3897
+    },
+    {
+      "epoch": 0.34182253651197747,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0025151753726301808,
+      "loss": 1.2157,
+      "step": 3898
+    },
+    {
+      "epoch": 0.34191022828635204,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002514862374023936,
+      "loss": 1.2426,
+      "step": 3899
+    },
+    {
+      "epoch": 0.34199792006072655,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0025145492965444756,
+      "loss": 1.2843,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3420856118351011,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0025142361402203505,
+      "loss": 1.1935,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3421733036094756,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0025139229050801207,
+      "loss": 1.1665,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3422609953838502,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0025136095911523523,
+      "loss": 1.1793,
+      "step": 3903
+    },
+    {
+      "epoch": 0.34234868715822475,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002513296198465619,
+      "loss": 1.1819,
+      "step": 3904
+    },
+    {
+      "epoch": 0.34243637893259926,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0025129827270485,
+      "loss": 1.2102,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3425240707069738,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0025126691769295846,
+      "loss": 1.2154,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3426117624813484,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002512355548137467,
+      "loss": 1.2066,
+      "step": 3907
+    },
+    {
+      "epoch": 0.3426994542557229,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00251204184070075,
+      "loss": 1.1542,
+      "step": 3908
+    },
+    {
+      "epoch": 0.34278714603009747,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0025117280546480417,
+      "loss": 1.1944,
+      "step": 3909
+    },
+    {
+      "epoch": 0.34287483780447203,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0025114141900079603,
+      "loss": 1.1956,
+      "step": 3910
+    },
+    {
+      "epoch": 0.34296252957884654,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002511100246809128,
+      "loss": 1.2062,
+      "step": 3911
+    },
+    {
+      "epoch": 0.3430502213532211,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0025107862250801764,
+      "loss": 1.2028,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3431379131275956,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002510472124849743,
+      "loss": 1.2182,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3432256049019702,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002510157946146473,
+      "loss": 1.1644,
+      "step": 3914
+    },
+    {
+      "epoch": 0.34331329667634475,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0025098436889990187,
+      "loss": 1.2194,
+      "step": 3915
+    },
+    {
+      "epoch": 0.34340098845071926,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0025095293534360395,
+      "loss": 1.2093,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3434886802250938,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025092149394862024,
+      "loss": 1.2212,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3435763719994684,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0025089004471781813,
+      "loss": 1.1971,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3436640637738429,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0025085858765406567,
+      "loss": 1.1094,
+      "step": 3919
+    },
+    {
+      "epoch": 0.34375175554821746,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002508271227602316,
+      "loss": 1.2273,
+      "step": 3920
+    },
+    {
+      "epoch": 0.34383944732259203,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002507956500391855,
+      "loss": 1.217,
+      "step": 3921
+    },
+    {
+      "epoch": 0.34392713909696654,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0025076416949379758,
+      "loss": 1.2243,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3440148308713411,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0025073268112693887,
+      "loss": 1.1952,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3441025226457156,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002507011849414809,
+      "loss": 1.1919,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3441902144200902,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002506696809402961,
+      "loss": 1.2253,
+      "step": 3925
+    },
+    {
+      "epoch": 0.34427790619446474,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002506381691262576,
+      "loss": 1.2012,
+      "step": 3926
+    },
+    {
+      "epoch": 0.34436559796883925,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0025060664950223915,
+      "loss": 1.1712,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3444532897432138,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002505751220711152,
+      "loss": 1.1488,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3445409815175884,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0025054358683576105,
+      "loss": 1.1479,
+      "step": 3929
+    },
+    {
+      "epoch": 0.3446286732919629,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0025051204379905262,
+      "loss": 1.2139,
+      "step": 3930
+    },
+    {
+      "epoch": 0.34471636506633746,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0025048049296386658,
+      "loss": 1.2481,
+      "step": 3931
+    },
+    {
+      "epoch": 0.344804056840712,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0025044893433308017,
+      "loss": 1.1538,
+      "step": 3932
+    },
+    {
+      "epoch": 0.34489174861508654,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025041736790957156,
+      "loss": 1.194,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3449794403894611,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0025038579369621953,
+      "loss": 1.1806,
+      "step": 3934
+    },
+    {
+      "epoch": 0.34506713216383567,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002503542116959035,
+      "loss": 1.245,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3451548239382102,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0025032262191150376,
+      "loss": 1.1789,
+      "step": 3936
+    },
+    {
+      "epoch": 0.34524251571258474,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0025029102434590113,
+      "loss": 1.2087,
+      "step": 3937
+    },
+    {
+      "epoch": 0.34533020748695925,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002502594190019773,
+      "loss": 1.19,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3454178992613338,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0025022780588261446,
+      "loss": 1.1945,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3455055910357084,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025019618499069578,
+      "loss": 1.1989,
+      "step": 3940
+    },
+    {
+      "epoch": 0.3455932828100829,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00250164556329105,
+      "loss": 1.1456,
+      "step": 3941
+    },
+    {
+      "epoch": 0.34568097458445746,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0025013291990072647,
+      "loss": 1.1782,
+      "step": 3942
+    },
+    {
+      "epoch": 0.345768666358832,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025010127570844545,
+      "loss": 1.2424,
+      "step": 3943
+    },
+    {
+      "epoch": 0.34585635813320653,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002500696237551478,
+      "loss": 1.1906,
+      "step": 3944
+    },
+    {
+      "epoch": 0.3459440499075811,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0025003796404372003,
+      "loss": 1.1946,
+      "step": 3945
+    },
+    {
+      "epoch": 0.34603174168195566,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0025000629657704943,
+      "loss": 1.1226,
+      "step": 3946
+    },
+    {
+      "epoch": 0.34611943345633017,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0024997462135802413,
+      "loss": 1.233,
+      "step": 3947
+    },
+    {
+      "epoch": 0.34620712523070474,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002499429383895326,
+      "loss": 1.2009,
+      "step": 3948
+    },
+    {
+      "epoch": 0.34629481700507925,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0024991124767446446,
+      "loss": 1.2342,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3463825087794538,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024987954921570966,
+      "loss": 1.1713,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3464702005538284,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0024984784301615914,
+      "loss": 1.1788,
+      "step": 3951
+    },
+    {
+      "epoch": 0.3465578923282029,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0024981612907870434,
+      "loss": 1.1809,
+      "step": 3952
+    },
+    {
+      "epoch": 0.34664558410257745,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0024978440740623755,
+      "loss": 1.2026,
+      "step": 3953
+    },
+    {
+      "epoch": 0.346733275876952,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002497526780016517,
+      "loss": 1.1263,
+      "step": 3954
+    },
+    {
+      "epoch": 0.34682096765132653,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002497209408678403,
+      "loss": 1.1873,
+      "step": 3955
+    },
+    {
+      "epoch": 0.3469086594257011,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024968919600769786,
+      "loss": 1.2177,
+      "step": 3956
+    },
+    {
+      "epoch": 0.34699635120007566,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0024965744342411943,
+      "loss": 1.2806,
+      "step": 3957
+    },
+    {
+      "epoch": 0.34708404297445017,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002496256831200006,
+      "loss": 1.2113,
+      "step": 3958
+    },
+    {
+      "epoch": 0.34717173474882473,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024959391509823798,
+      "loss": 1.2125,
+      "step": 3959
+    },
+    {
+      "epoch": 0.34725942652319924,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0024956213936172863,
+      "loss": 1.2437,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3473471182975738,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002495303559133706,
+      "loss": 1.2195,
+      "step": 3961
+    },
+    {
+      "epoch": 0.3474348100719484,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002494985647560622,
+      "loss": 1.2298,
+      "step": 3962
+    },
+    {
+      "epoch": 0.3475225018463229,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0024946676589270283,
+      "loss": 1.2144,
+      "step": 3963
+    },
+    {
+      "epoch": 0.34761019362069745,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0024943495932619247,
+      "loss": 1.2405,
+      "step": 3964
+    },
+    {
+      "epoch": 0.347697885395072,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0024940314505943187,
+      "loss": 1.1543,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3477855771694465,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002493713230953222,
+      "loss": 1.2461,
+      "step": 3966
+    },
+    {
+      "epoch": 0.3478732689438211,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002493394934367658,
+      "loss": 1.1611,
+      "step": 3967
+    },
+    {
+      "epoch": 0.34796096071819566,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002493076560866652,
+      "loss": 1.2033,
+      "step": 3968
+    },
+    {
+      "epoch": 0.34804865249257017,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0024927581104792406,
+      "loss": 1.1924,
+      "step": 3969
+    },
+    {
+      "epoch": 0.34813634426694473,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0024924395832344654,
+      "loss": 1.2847,
+      "step": 3970
+    },
+    {
+      "epoch": 0.34822403604131924,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0024921209791613744,
+      "loss": 1.1938,
+      "step": 3971
+    },
+    {
+      "epoch": 0.3483117278156938,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002491802298289024,
+      "loss": 1.2244,
+      "step": 3972
+    },
+    {
+      "epoch": 0.34839941959006837,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0024914835406464773,
+      "loss": 1.2591,
+      "step": 3973
+    },
+    {
+      "epoch": 0.3484871113644429,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002491164706262804,
+      "loss": 1.1762,
+      "step": 3974
+    },
+    {
+      "epoch": 0.34857480313881745,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002490845795167081,
+      "loss": 1.2163,
+      "step": 3975
+    },
+    {
+      "epoch": 0.348662494913192,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0024905268073883925,
+      "loss": 1.2662,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3487501866875665,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002490207742955828,
+      "loss": 1.208,
+      "step": 3977
+    },
+    {
+      "epoch": 0.3488378784619411,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0024898886018984868,
+      "loss": 1.2244,
+      "step": 3978
+    },
+    {
+      "epoch": 0.34892557023631565,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0024895693842454736,
+      "loss": 1.1687,
+      "step": 3979
+    },
+    {
+      "epoch": 0.34901326201069016,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002489250090025899,
+      "loss": 1.169,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3491009537850647,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002488930719268884,
+      "loss": 1.1745,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3491886455594393,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024886112720035524,
+      "loss": 1.1344,
+      "step": 3982
+    },
+    {
+      "epoch": 0.3492763373338138,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0024882917482590377,
+      "loss": 1.1921,
+      "step": 3983
+    },
+    {
+      "epoch": 0.34936402910818837,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0024879721480644793,
+      "loss": 1.1843,
+      "step": 3984
+    },
+    {
+      "epoch": 0.3494517208825629,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0024876524714490247,
+      "loss": 1.2066,
+      "step": 3985
+    },
+    {
+      "epoch": 0.34953941265693744,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002487332718441827,
+      "loss": 1.1675,
+      "step": 3986
+    },
+    {
+      "epoch": 0.349627104431312,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0024870128890720467,
+      "loss": 1.1768,
+      "step": 3987
+    },
+    {
+      "epoch": 0.3497147962056865,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002486692983368852,
+      "loss": 1.2232,
+      "step": 3988
+    },
+    {
+      "epoch": 0.3498024879800611,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0024863730013614167,
+      "loss": 1.1969,
+      "step": 3989
+    },
+    {
+      "epoch": 0.34989017975443565,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0024860529430789226,
+      "loss": 1.1865,
+      "step": 3990
+    },
+    {
+      "epoch": 0.34997787152881016,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0024857328085505586,
+      "loss": 1.1836,
+      "step": 3991
+    },
+    {
+      "epoch": 0.3500655633031847,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00248541259780552,
+      "loss": 1.1888,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3501532550775593,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0024850923108730096,
+      "loss": 1.1849,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3502409468519338,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0024847719477822353,
+      "loss": 1.2671,
+      "step": 3994
+    },
+    {
+      "epoch": 0.35032863862630836,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0024844515085624143,
+      "loss": 1.2104,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3504163304006829,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002484130993242771,
+      "loss": 1.1946,
+      "step": 3996
+    },
+    {
+      "epoch": 0.35050402217505744,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002483810401852534,
+      "loss": 1.1576,
+      "step": 3997
+    },
+    {
+      "epoch": 0.350591713949432,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00248348973442094,
+      "loss": 1.2008,
+      "step": 3998
+    },
+    {
+      "epoch": 0.3506794057238065,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0024831689909772343,
+      "loss": 1.2239,
+      "step": 3999
+    },
+    {
+      "epoch": 0.3507670974981811,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002482848171550668,
+      "loss": 1.2102,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3507670974981811,
+      "eval_loss": 1.2098665237426758,
+      "eval_runtime": 429.0914,
+      "eval_samples_per_second": 33.669,
+      "eval_steps_per_second": 8.418,
+      "step": 4000
+    },
+    {
+      "epoch": 0.35085478927255564,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0024825272761704975,
+      "loss": 1.1917,
+      "step": 4001
+    },
+    {
+      "epoch": 0.35094248104693015,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0024822063048659886,
+      "loss": 1.2528,
+      "step": 4002
+    },
+    {
+      "epoch": 0.3510301728213047,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0024818852576664135,
+      "loss": 1.1819,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3511178645956793,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0024815641346010507,
+      "loss": 1.1739,
+      "step": 4004
+    },
+    {
+      "epoch": 0.3512055563700538,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002481242935699185,
+      "loss": 1.211,
+      "step": 4005
+    },
+    {
+      "epoch": 0.35129324814442836,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00248092166099011,
+      "loss": 1.1861,
+      "step": 4006
+    },
+    {
+      "epoch": 0.35138093991880287,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0024806003105031236,
+      "loss": 1.2647,
+      "step": 4007
+    },
+    {
+      "epoch": 0.35146863169317744,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0024802788842675336,
+      "loss": 1.1157,
+      "step": 4008
+    },
+    {
+      "epoch": 0.351556323467552,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0024799573823126525,
+      "loss": 1.236,
+      "step": 4009
+    },
+    {
+      "epoch": 0.3516440152419265,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002479635804667801,
+      "loss": 1.1695,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3517317070163011,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0024793141513623053,
+      "loss": 1.1794,
+      "step": 4011
+    },
+    {
+      "epoch": 0.35181939879067564,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0024789924224255006,
+      "loss": 1.255,
+      "step": 4012
+    },
+    {
+      "epoch": 0.35190709056505015,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0024786706178867267,
+      "loss": 1.1786,
+      "step": 4013
+    },
+    {
+      "epoch": 0.3519947823394247,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0024783487377753313,
+      "loss": 1.2247,
+      "step": 4014
+    },
+    {
+      "epoch": 0.3520824741137993,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00247802678212067,
+      "loss": 1.2098,
+      "step": 4015
+    },
+    {
+      "epoch": 0.3521701658881738,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002477704750952103,
+      "loss": 1.2144,
+      "step": 4016
+    },
+    {
+      "epoch": 0.35225785766254836,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0024773826442989998,
+      "loss": 1.2155,
+      "step": 4017
+    },
+    {
+      "epoch": 0.3523455494369229,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0024770604621907354,
+      "loss": 1.2848,
+      "step": 4018
+    },
+    {
+      "epoch": 0.35243324121129743,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002476738204656692,
+      "loss": 1.2201,
+      "step": 4019
+    },
+    {
+      "epoch": 0.352520932985672,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0024764158717262576,
+      "loss": 1.2494,
+      "step": 4020
+    },
+    {
+      "epoch": 0.3526086247600465,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0024760934634288304,
+      "loss": 1.2228,
+      "step": 4021
+    },
+    {
+      "epoch": 0.3526963165344211,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002475770979793811,
+      "loss": 1.1534,
+      "step": 4022
+    },
+    {
+      "epoch": 0.35278400830879564,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0024754484208506103,
+      "loss": 1.1702,
+      "step": 4023
+    },
+    {
+      "epoch": 0.35287170008317015,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0024751257866286445,
+      "loss": 1.231,
+      "step": 4024
+    },
+    {
+      "epoch": 0.3529593918575447,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0024748030771573368,
+      "loss": 1.183,
+      "step": 4025
+    },
+    {
+      "epoch": 0.3530470836319193,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002474480292466118,
+      "loss": 1.1784,
+      "step": 4026
+    },
+    {
+      "epoch": 0.3531347754062938,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0024741574325844247,
+      "loss": 1.1829,
+      "step": 4027
+    },
+    {
+      "epoch": 0.35322246718066835,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002473834497541701,
+      "loss": 1.1981,
+      "step": 4028
+    },
+    {
+      "epoch": 0.3533101589550429,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002473511487367398,
+      "loss": 1.2156,
+      "step": 4029
+    },
+    {
+      "epoch": 0.35339785072941743,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0024731884020909737,
+      "loss": 1.1608,
+      "step": 4030
+    },
+    {
+      "epoch": 0.353485542503792,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0024728652417418916,
+      "loss": 1.2088,
+      "step": 4031
+    },
+    {
+      "epoch": 0.3535732342781665,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0024725420063496234,
+      "loss": 1.1753,
+      "step": 4032
+    },
+    {
+      "epoch": 0.35366092605254107,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002472218695943648,
+      "loss": 1.1922,
+      "step": 4033
+    },
+    {
+      "epoch": 0.35374861782691563,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00247189531055345,
+      "loss": 1.1724,
+      "step": 4034
+    },
+    {
+      "epoch": 0.35383630960129014,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0024715718502085213,
+      "loss": 1.1824,
+      "step": 4035
+    },
+    {
+      "epoch": 0.3539240013756647,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0024712483149383615,
+      "loss": 1.176,
+      "step": 4036
+    },
+    {
+      "epoch": 0.3540116931500393,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.002470924704772474,
+      "loss": 1.2307,
+      "step": 4037
+    },
+    {
+      "epoch": 0.3540993849244138,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002470601019740374,
+      "loss": 1.1297,
+      "step": 4038
+    },
+    {
+      "epoch": 0.35418707669878835,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002470277259871579,
+      "loss": 1.208,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3542747684731629,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002469953425195615,
+      "loss": 1.2233,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3543624602475374,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002469629515742015,
+      "loss": 1.1881,
+      "step": 4041
+    },
+    {
+      "epoch": 0.354450152021912,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00246930553154032,
+      "loss": 1.2324,
+      "step": 4042
+    },
+    {
+      "epoch": 0.3545378437962865,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0024689814726200753,
+      "loss": 1.1999,
+      "step": 4043
+    },
+    {
+      "epoch": 0.35462553557066107,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002468657339010834,
+      "loss": 1.1888,
+      "step": 4044
+    },
+    {
+      "epoch": 0.35471322734503563,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002468333130742157,
+      "loss": 1.1923,
+      "step": 4045
+    },
+    {
+      "epoch": 0.35480091911941014,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0024680088478436116,
+      "loss": 1.2299,
+      "step": 4046
+    },
+    {
+      "epoch": 0.3548886108937847,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00246768449034477,
+      "loss": 1.1515,
+      "step": 4047
+    },
+    {
+      "epoch": 0.35497630266815927,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002467360058275214,
+      "loss": 1.2143,
+      "step": 4048
+    },
+    {
+      "epoch": 0.3550639944425338,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002467035551664531,
+      "loss": 1.2015,
+      "step": 4049
+    },
+    {
+      "epoch": 0.35515168621690835,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0024667109705423145,
+      "loss": 1.2023,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3552393779912829,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002466386314938166,
+      "loss": 1.1811,
+      "step": 4051
+    },
+    {
+      "epoch": 0.3553270697656574,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0024660615848816933,
+      "loss": 1.1924,
+      "step": 4052
+    },
+    {
+      "epoch": 0.355414761540032,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0024657367804025105,
+      "loss": 1.2035,
+      "step": 4053
+    },
+    {
+      "epoch": 0.35550245331440655,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002465411901530239,
+      "loss": 1.1671,
+      "step": 4054
+    },
+    {
+      "epoch": 0.35559014508878106,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002465086948294507,
+      "loss": 1.2076,
+      "step": 4055
+    },
+    {
+      "epoch": 0.3556778368631556,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00246476192072495,
+      "loss": 1.175,
+      "step": 4056
+    },
+    {
+      "epoch": 0.35576552863753014,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0024644368188512086,
+      "loss": 1.1606,
+      "step": 4057
+    },
+    {
+      "epoch": 0.3558532204119047,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0024641116427029314,
+      "loss": 1.1418,
+      "step": 4058
+    },
+    {
+      "epoch": 0.35594091218627927,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0024637863923097753,
+      "loss": 1.1985,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3560286039606538,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0024634610677013997,
+      "loss": 1.1962,
+      "step": 4060
+    },
+    {
+      "epoch": 0.35611629573502834,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0024631356689074754,
+      "loss": 1.1436,
+      "step": 4061
+    },
+    {
+      "epoch": 0.3562039875094029,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.002462810195957677,
+      "loss": 1.1962,
+      "step": 4062
+    },
+    {
+      "epoch": 0.3562916792837774,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002462484648881687,
+      "loss": 1.1953,
+      "step": 4063
+    },
+    {
+      "epoch": 0.356379371058152,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002462159027709194,
+      "loss": 1.2035,
+      "step": 4064
+    },
+    {
+      "epoch": 0.35646706283252655,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0024618333324698954,
+      "loss": 1.2382,
+      "step": 4065
+    },
+    {
+      "epoch": 0.35655475460690106,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0024615075631934913,
+      "loss": 1.2347,
+      "step": 4066
+    },
+    {
+      "epoch": 0.3566424463812756,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002461181719909693,
+      "loss": 1.2166,
+      "step": 4067
+    },
+    {
+      "epoch": 0.35673013815565013,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0024608558026482158,
+      "loss": 1.217,
+      "step": 4068
+    },
+    {
+      "epoch": 0.3568178299300247,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0024605298114387824,
+      "loss": 1.1648,
+      "step": 4069
+    },
+    {
+      "epoch": 0.35690552170439926,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002460203746311123,
+      "loss": 1.1743,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3569932134787738,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0024598776072949735,
+      "loss": 1.2306,
+      "step": 4071
+    },
+    {
+      "epoch": 0.35708090525314834,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002459551394420077,
+      "loss": 1.1495,
+      "step": 4072
+    },
+    {
+      "epoch": 0.3571685970275229,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0024592251077161825,
+      "loss": 1.1746,
+      "step": 4073
+    },
+    {
+      "epoch": 0.3572562888018974,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0024588987472130472,
+      "loss": 1.2092,
+      "step": 4074
+    },
+    {
+      "epoch": 0.357343980576272,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002458572312940435,
+      "loss": 1.2551,
+      "step": 4075
+    },
+    {
+      "epoch": 0.35743167235064655,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0024582458049281146,
+      "loss": 1.1381,
+      "step": 4076
+    },
+    {
+      "epoch": 0.35751936412502106,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0024579192232058638,
+      "loss": 1.2417,
+      "step": 4077
+    },
+    {
+      "epoch": 0.3576070558993956,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002457592567803465,
+      "loss": 1.1758,
+      "step": 4078
+    },
+    {
+      "epoch": 0.35769474767377013,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002457265838750709,
+      "loss": 1.2283,
+      "step": 4079
+    },
+    {
+      "epoch": 0.3577824394481447,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002456939036077392,
+      "loss": 1.2084,
+      "step": 4080
+    },
+    {
+      "epoch": 0.35787013122251926,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0024566121598133183,
+      "loss": 1.1438,
+      "step": 4081
+    },
+    {
+      "epoch": 0.35795782299689377,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0024562852099882983,
+      "loss": 1.2152,
+      "step": 4082
+    },
+    {
+      "epoch": 0.35804551477126834,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002455958186632148,
+      "loss": 1.2063,
+      "step": 4083
+    },
+    {
+      "epoch": 0.3581332065456429,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0024556310897746922,
+      "loss": 1.2277,
+      "step": 4084
+    },
+    {
+      "epoch": 0.3582208983200174,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0024553039194457602,
+      "loss": 1.1478,
+      "step": 4085
+    },
+    {
+      "epoch": 0.358308590094392,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00245497667567519,
+      "loss": 1.2032,
+      "step": 4086
+    },
+    {
+      "epoch": 0.35839628186876654,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002454649358492825,
+      "loss": 1.1712,
+      "step": 4087
+    },
+    {
+      "epoch": 0.35848397364314105,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0024543219679285153,
+      "loss": 1.1696,
+      "step": 4088
+    },
+    {
+      "epoch": 0.3585716654175156,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002453994504012119,
+      "loss": 1.2239,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3586593571918902,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002453666966773499,
+      "loss": 1.1189,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3587470489662647,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002453339356242527,
+      "loss": 1.2107,
+      "step": 4091
+    },
+    {
+      "epoch": 0.35883474074063926,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.002453011672449079,
+      "loss": 1.1908,
+      "step": 4092
+    },
+    {
+      "epoch": 0.35892243251501377,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024526839154230392,
+      "loss": 1.2217,
+      "step": 4093
+    },
+    {
+      "epoch": 0.35901012428938833,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0024523560851942984,
+      "loss": 1.2338,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3590978160637629,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0024520281817927542,
+      "loss": 1.2004,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3591855078381374,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0024517002052483103,
+      "loss": 1.1644,
+      "step": 4096
+    },
+    {
+      "epoch": 0.359273199612512,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0024513721555908773,
+      "loss": 1.1475,
+      "step": 4097
+    },
+    {
+      "epoch": 0.35936089138688654,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002451044032850372,
+      "loss": 1.2503,
+      "step": 4098
+    },
+    {
+      "epoch": 0.35944858316126105,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0024507158370567195,
+      "loss": 1.1619,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3595362749356356,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0024503875682398493,
+      "loss": 1.1601,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3596239667100102,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002450059226429699,
+      "loss": 1.159,
+      "step": 4101
+    },
+    {
+      "epoch": 0.3597116584843847,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.002449730811656213,
+      "loss": 1.2004,
+      "step": 4102
+    },
+    {
+      "epoch": 0.35979935025875925,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0024494023239493407,
+      "loss": 1.153,
+      "step": 4103
+    },
+    {
+      "epoch": 0.35988704203313376,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002449073763339041,
+      "loss": 1.1981,
+      "step": 4104
+    },
+    {
+      "epoch": 0.35997473380750833,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0024487451298552763,
+      "loss": 1.2117,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3600624255818829,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002448416423528018,
+      "loss": 1.1941,
+      "step": 4106
+    },
+    {
+      "epoch": 0.3601501173562574,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0024480876443872426,
+      "loss": 1.148,
+      "step": 4107
+    },
+    {
+      "epoch": 0.36023780913063197,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002447758792462934,
+      "loss": 1.2041,
+      "step": 4108
+    },
+    {
+      "epoch": 0.36032550090500653,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002447429867785084,
+      "loss": 1.2234,
+      "step": 4109
+    },
+    {
+      "epoch": 0.36041319267938104,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0024471008703836877,
+      "loss": 1.1527,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3605008844537556,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00244677180028875,
+      "loss": 1.2224,
+      "step": 4111
+    },
+    {
+      "epoch": 0.3605885762281302,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002446442657530281,
+      "loss": 1.1912,
+      "step": 4112
+    },
+    {
+      "epoch": 0.3606762680025047,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024461134421382975,
+      "loss": 1.1434,
+      "step": 4113
+    },
+    {
+      "epoch": 0.36076395977687925,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002445784154142824,
+      "loss": 1.1893,
+      "step": 4114
+    },
+    {
+      "epoch": 0.36085165155125376,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0024454547935738894,
+      "loss": 1.1631,
+      "step": 4115
+    },
+    {
+      "epoch": 0.3609393433256283,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002445125360461531,
+      "loss": 1.191,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3610270351000029,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002444795854835792,
+      "loss": 1.1859,
+      "step": 4117
+    },
+    {
+      "epoch": 0.3611147268743774,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0024444662767267233,
+      "loss": 1.1426,
+      "step": 4118
+    },
+    {
+      "epoch": 0.36120241864875197,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002444136626164381,
+      "loss": 1.1814,
+      "step": 4119
+    },
+    {
+      "epoch": 0.36129011042312653,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002443806903178829,
+      "loss": 1.2003,
+      "step": 4120
+    },
+    {
+      "epoch": 0.36137780219750104,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0024434771078001355,
+      "loss": 1.1729,
+      "step": 4121
+    },
+    {
+      "epoch": 0.3614654939718756,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0024431472400583787,
+      "loss": 1.1712,
+      "step": 4122
+    },
+    {
+      "epoch": 0.36155318574625017,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0024428172999836407,
+      "loss": 1.1327,
+      "step": 4123
+    },
+    {
+      "epoch": 0.3616408775206247,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002442487287606012,
+      "loss": 1.185,
+      "step": 4124
+    },
+    {
+      "epoch": 0.36172856929499925,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002442157202955588,
+      "loss": 1.2227,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3618162610693738,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002441827046062473,
+      "loss": 1.222,
+      "step": 4126
+    },
+    {
+      "epoch": 0.3619039528437483,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002441496816956775,
+      "loss": 1.2575,
+      "step": 4127
+    },
+    {
+      "epoch": 0.3619916446181229,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00244116651566861,
+      "loss": 1.1627,
+      "step": 4128
+    },
+    {
+      "epoch": 0.3620793363924974,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0024408361422281017,
+      "loss": 1.2641,
+      "step": 4129
+    },
+    {
+      "epoch": 0.36216702816687196,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024405056966653785,
+      "loss": 1.1924,
+      "step": 4130
+    },
+    {
+      "epoch": 0.36225471994124653,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002440175179010577,
+      "loss": 1.229,
+      "step": 4131
+    },
+    {
+      "epoch": 0.36234241171562104,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002439844589293838,
+      "loss": 1.1924,
+      "step": 4132
+    },
+    {
+      "epoch": 0.3624301034899956,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0024395139275453126,
+      "loss": 1.1221,
+      "step": 4133
+    },
+    {
+      "epoch": 0.36251779526437017,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0024391831937951546,
+      "loss": 1.2069,
+      "step": 4134
+    },
+    {
+      "epoch": 0.3626054870387447,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002438852388073526,
+      "loss": 1.1549,
+      "step": 4135
+    },
+    {
+      "epoch": 0.36269317881311924,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002438521510410597,
+      "loss": 1.2077,
+      "step": 4136
+    },
+    {
+      "epoch": 0.3627808705874938,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002438190560836541,
+      "loss": 1.2363,
+      "step": 4137
+    },
+    {
+      "epoch": 0.3628685623618683,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0024378595393815413,
+      "loss": 1.1664,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3629562541362429,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002437528446075785,
+      "loss": 1.2101,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3630439459106174,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0024371972809494676,
+      "loss": 1.1892,
+      "step": 4140
+    },
+    {
+      "epoch": 0.36313163768499196,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00243686604403279,
+      "loss": 1.2095,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3632193294593665,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002436534735355961,
+      "loss": 1.2164,
+      "step": 4142
+    },
+    {
+      "epoch": 0.36330702123374103,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0024362033549491942,
+      "loss": 1.2092,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3633947130081156,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0024358719028427112,
+      "loss": 1.1836,
+      "step": 4144
+    },
+    {
+      "epoch": 0.36348240478249016,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0024355403790667393,
+      "loss": 1.2167,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3635700965568647,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0024352087836515125,
+      "loss": 1.1796,
+      "step": 4146
+    },
+    {
+      "epoch": 0.36365778833123924,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0024348771166272714,
+      "loss": 1.1778,
+      "step": 4147
+    },
+    {
+      "epoch": 0.3637454801056138,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0024345453780242644,
+      "loss": 1.2545,
+      "step": 4148
+    },
+    {
+      "epoch": 0.3638331718799883,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0024342135678727434,
+      "loss": 1.1539,
+      "step": 4149
+    },
+    {
+      "epoch": 0.3639208636543629,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0024338816862029696,
+      "loss": 1.2414,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3640085554287374,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0024335497330452097,
+      "loss": 1.2082,
+      "step": 4151
+    },
+    {
+      "epoch": 0.36409624720311196,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0024332177084297364,
+      "loss": 1.1856,
+      "step": 4152
+    },
+    {
+      "epoch": 0.3641839389774865,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0024328856123868306,
+      "loss": 1.1912,
+      "step": 4153
+    },
+    {
+      "epoch": 0.36427163075186103,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002432553444946778,
+      "loss": 1.2292,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3643593225262356,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002432221206139871,
+      "loss": 1.225,
+      "step": 4155
+    },
+    {
+      "epoch": 0.36444701430061016,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0024318888959964096,
+      "loss": 1.1632,
+      "step": 4156
+    },
+    {
+      "epoch": 0.36453470607498467,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0024315565145466987,
+      "loss": 1.2808,
+      "step": 4157
+    },
+    {
+      "epoch": 0.36462239784935924,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002431224061821052,
+      "loss": 1.2205,
+      "step": 4158
+    },
+    {
+      "epoch": 0.3647100896237338,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0024308915378497874,
+      "loss": 1.2312,
+      "step": 4159
+    },
+    {
+      "epoch": 0.3647977813981083,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0024305589426632302,
+      "loss": 1.1715,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3648854731724829,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0024302262762917123,
+      "loss": 1.2137,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3649731649468574,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002429893538765572,
+      "loss": 1.2057,
+      "step": 4162
+    },
+    {
+      "epoch": 0.36506085672123195,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002429560730115155,
+      "loss": 1.2417,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3651485484956065,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002429227850370811,
+      "loss": 1.1862,
+      "step": 4164
+    },
+    {
+      "epoch": 0.365236240269981,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0024288948995628994,
+      "loss": 1.2219,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3653239320443556,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0024285618777217834,
+      "loss": 1.2605,
+      "step": 4166
+    },
+    {
+      "epoch": 0.36541162381873016,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002428228784877834,
+      "loss": 1.1386,
+      "step": 4167
+    },
+    {
+      "epoch": 0.36549931559310467,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002427895621061428,
+      "loss": 1.1934,
+      "step": 4168
+    },
+    {
+      "epoch": 0.36558700736747923,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00242756238630295,
+      "loss": 1.1323,
+      "step": 4169
+    },
+    {
+      "epoch": 0.3656746991418538,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002427229080632789,
+      "loss": 1.2227,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3657623909162283,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002426895704081343,
+      "loss": 1.1657,
+      "step": 4171
+    },
+    {
+      "epoch": 0.3658500826906029,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002426562256679014,
+      "loss": 1.2945,
+      "step": 4172
+    },
+    {
+      "epoch": 0.36593777446497744,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002426228738456212,
+      "loss": 1.189,
+      "step": 4173
+    },
+    {
+      "epoch": 0.36602546623935195,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0024258951494433526,
+      "loss": 1.1634,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3661131580137265,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0024255614896708587,
+      "loss": 1.2358,
+      "step": 4175
+    },
+    {
+      "epoch": 0.366200849788101,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0024252277591691592,
+      "loss": 1.2129,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3662885415624756,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002424893957968689,
+      "loss": 1.1566,
+      "step": 4177
+    },
+    {
+      "epoch": 0.36637623333685015,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00242456008609989,
+      "loss": 1.2257,
+      "step": 4178
+    },
+    {
+      "epoch": 0.36646392511122466,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0024242261435932114,
+      "loss": 1.1792,
+      "step": 4179
+    },
+    {
+      "epoch": 0.36655161688559923,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0024238921304791067,
+      "loss": 1.1862,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3666393086599738,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0024235580467880376,
+      "loss": 1.1976,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3667270004343483,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002423223892550472,
+      "loss": 1.1824,
+      "step": 4182
+    },
+    {
+      "epoch": 0.36681469220872287,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0024228896677968826,
+      "loss": 1.2103,
+      "step": 4183
+    },
+    {
+      "epoch": 0.36690238398309744,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002422555372557751,
+      "loss": 1.2096,
+      "step": 4184
+    },
+    {
+      "epoch": 0.36699007575747195,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0024222210068635635,
+      "loss": 1.1659,
+      "step": 4185
+    },
+    {
+      "epoch": 0.3670777675318465,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0024218865707448145,
+      "loss": 1.2103,
+      "step": 4186
+    },
+    {
+      "epoch": 0.367165459306221,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0024215520642320025,
+      "loss": 1.1318,
+      "step": 4187
+    },
+    {
+      "epoch": 0.3672531510805956,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0024212174873556335,
+      "loss": 1.1934,
+      "step": 4188
+    },
+    {
+      "epoch": 0.36734084285497015,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002420882840146221,
+      "loss": 1.2553,
+      "step": 4189
+    },
+    {
+      "epoch": 0.36742853462934466,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0024205481226342836,
+      "loss": 1.1886,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3675162264037192,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0024202133348503463,
+      "loss": 1.1369,
+      "step": 4191
+    },
+    {
+      "epoch": 0.3676039181780938,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002419878476824942,
+      "loss": 1.1955,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3676916099524683,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0024195435485886065,
+      "loss": 1.2252,
+      "step": 4193
+    },
+    {
+      "epoch": 0.36777930172684287,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0024192085501718867,
+      "loss": 1.2567,
+      "step": 4194
+    },
+    {
+      "epoch": 0.36786699350121743,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002418873481605333,
+      "loss": 1.2069,
+      "step": 4195
+    },
+    {
+      "epoch": 0.36795468527559194,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0024185383429195027,
+      "loss": 1.1973,
+      "step": 4196
+    },
+    {
+      "epoch": 0.3680423770499665,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0024182031341449593,
+      "loss": 1.2204,
+      "step": 4197
+    },
+    {
+      "epoch": 0.368130068824341,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0024178678553122735,
+      "loss": 1.1874,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3682177605987156,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002417532506452021,
+      "loss": 1.2007,
+      "step": 4199
+    },
+    {
+      "epoch": 0.36830545237309015,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0024171970875947854,
+      "loss": 1.2274,
+      "step": 4200
+    },
+    {
+      "epoch": 0.36839314414746466,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0024168615987711557,
+      "loss": 1.186,
+      "step": 4201
+    },
+    {
+      "epoch": 0.3684808359218392,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0024165260400117283,
+      "loss": 1.2233,
+      "step": 4202
+    },
+    {
+      "epoch": 0.3685685276962138,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0024161904113471043,
+      "loss": 1.2034,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3686562194705883,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0024158547128078933,
+      "loss": 1.1557,
+      "step": 4204
+    },
+    {
+      "epoch": 0.36874391124496286,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002415518944424709,
+      "loss": 1.2064,
+      "step": 4205
+    },
+    {
+      "epoch": 0.36883160301933743,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002415183106228174,
+      "loss": 1.1588,
+      "step": 4206
+    },
+    {
+      "epoch": 0.36891929479371194,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0024148471982489146,
+      "loss": 1.2246,
+      "step": 4207
+    },
+    {
+      "epoch": 0.3690069865680865,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0024145112205175645,
+      "loss": 1.2152,
+      "step": 4208
+    },
+    {
+      "epoch": 0.36909467834246107,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0024141751730647654,
+      "loss": 1.187,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3691823701168356,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0024138390559211628,
+      "loss": 1.2557,
+      "step": 4210
+    },
+    {
+      "epoch": 0.36927006189121014,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0024135028691174105,
+      "loss": 1.2463,
+      "step": 4211
+    },
+    {
+      "epoch": 0.36935775366558465,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0024131666126841673,
+      "loss": 1.1331,
+      "step": 4212
+    },
+    {
+      "epoch": 0.3694454454399592,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002412830286652099,
+      "loss": 1.1649,
+      "step": 4213
+    },
+    {
+      "epoch": 0.3695331372143338,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0024124938910518785,
+      "loss": 1.1731,
+      "step": 4214
+    },
+    {
+      "epoch": 0.3696208289887083,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002412157425914183,
+      "loss": 1.2278,
+      "step": 4215
+    },
+    {
+      "epoch": 0.36970852076308286,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0024118208912696977,
+      "loss": 1.1799,
+      "step": 4216
+    },
+    {
+      "epoch": 0.3697962125374574,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0024114842871491137,
+      "loss": 1.19,
+      "step": 4217
+    },
+    {
+      "epoch": 0.36988390431183193,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0024111476135831285,
+      "loss": 1.1761,
+      "step": 4218
+    },
+    {
+      "epoch": 0.3699715960862065,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002410810870602446,
+      "loss": 1.2772,
+      "step": 4219
+    },
+    {
+      "epoch": 0.37005928786058107,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.002410474058237776,
+      "loss": 1.1806,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3701469796349556,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002410137176519835,
+      "loss": 1.1487,
+      "step": 4221
+    },
+    {
+      "epoch": 0.37023467140933014,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002409800225479346,
+      "loss": 1.242,
+      "step": 4222
+    },
+    {
+      "epoch": 0.37032236318370465,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0024094632051470375,
+      "loss": 1.2088,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3704100549580792,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002409126115553645,
+      "loss": 1.1997,
+      "step": 4224
+    },
+    {
+      "epoch": 0.3704977467324538,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0024087889567299113,
+      "loss": 1.1768,
+      "step": 4225
+    },
+    {
+      "epoch": 0.3705854385068283,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.002408451728706583,
+      "loss": 1.2187,
+      "step": 4226
+    },
+    {
+      "epoch": 0.37067313028120286,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002408114431514415,
+      "loss": 1.1996,
+      "step": 4227
+    },
+    {
+      "epoch": 0.3707608220555774,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0024077770651841682,
+      "loss": 1.2108,
+      "step": 4228
+    },
+    {
+      "epoch": 0.37084851382995193,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002407439629746609,
+      "loss": 1.1962,
+      "step": 4229
+    },
+    {
+      "epoch": 0.3709362056043265,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0024071021252325105,
+      "loss": 1.2024,
+      "step": 4230
+    },
+    {
+      "epoch": 0.37102389737870106,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024067645516726533,
+      "loss": 1.1889,
+      "step": 4231
+    },
+    {
+      "epoch": 0.37111158915307557,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0024064269090978228,
+      "loss": 1.2876,
+      "step": 4232
+    },
+    {
+      "epoch": 0.37119928092745014,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00240608919753881,
+      "loss": 1.2961,
+      "step": 4233
+    },
+    {
+      "epoch": 0.37128697270182465,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002405751417026415,
+      "loss": 1.2306,
+      "step": 4234
+    },
+    {
+      "epoch": 0.3713746644761992,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002405413567591441,
+      "loss": 1.2144,
+      "step": 4235
+    },
+    {
+      "epoch": 0.3714623562505738,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0024050756492647,
+      "loss": 1.2143,
+      "step": 4236
+    },
+    {
+      "epoch": 0.3715500480249483,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002404737662077009,
+      "loss": 1.2175,
+      "step": 4237
+    },
+    {
+      "epoch": 0.37163773979932285,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002404399606059192,
+      "loss": 1.1913,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3717254315736974,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002404061481242078,
+      "loss": 1.2011,
+      "step": 4239
+    },
+    {
+      "epoch": 0.37181312334807193,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0024037232876565036,
+      "loss": 1.1957,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3719008151224465,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002403385025333311,
+      "loss": 1.1492,
+      "step": 4241
+    },
+    {
+      "epoch": 0.37198850689682106,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0024030466943033495,
+      "loss": 1.2376,
+      "step": 4242
+    },
+    {
+      "epoch": 0.37207619867119557,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002402708294597472,
+      "loss": 1.1493,
+      "step": 4243
+    },
+    {
+      "epoch": 0.37216389044557013,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0024023698262465422,
+      "loss": 1.16,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3722515822199447,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0024020312892814267,
+      "loss": 1.1355,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3723392739943192,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0024016926837329986,
+      "loss": 1.1972,
+      "step": 4246
+    },
+    {
+      "epoch": 0.3724269657686938,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0024013540096321385,
+      "loss": 1.1791,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3725146575430683,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0024010152670097323,
+      "loss": 1.2102,
+      "step": 4248
+    },
+    {
+      "epoch": 0.37260234931744285,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002400676455896672,
+      "loss": 1.2079,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3726900410918174,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0024003375763238573,
+      "loss": 1.2348,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3727777328661919,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0023999986283221924,
+      "loss": 1.164,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3728654246405665,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002399659611922589,
+      "loss": 1.2384,
+      "step": 4252
+    },
+    {
+      "epoch": 0.37295311641494105,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0023993205271559638,
+      "loss": 1.1913,
+      "step": 4253
+    },
+    {
+      "epoch": 0.37304080818931556,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002398981374053241,
+      "loss": 1.1624,
+      "step": 4254
+    },
+    {
+      "epoch": 0.37312849996369013,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0023986421526453502,
+      "loss": 1.2163,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3732161917380647,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002398302862963228,
+      "loss": 1.1505,
+      "step": 4256
+    },
+    {
+      "epoch": 0.3733038835124392,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002397963505037816,
+      "loss": 1.174,
+      "step": 4257
+    },
+    {
+      "epoch": 0.37339157528681377,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002397624078900064,
+      "loss": 1.2025,
+      "step": 4258
+    },
+    {
+      "epoch": 0.3734792670611883,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0023972845845809257,
+      "loss": 1.2265,
+      "step": 4259
+    },
+    {
+      "epoch": 0.37356695883556285,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002396945022111362,
+      "loss": 1.2228,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3736546506099374,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0023966053915223406,
+      "loss": 1.1801,
+      "step": 4261
+    },
+    {
+      "epoch": 0.3737423423843119,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002396265692844835,
+      "loss": 1.135,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3738300341586865,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002395925926109825,
+      "loss": 1.277,
+      "step": 4263
+    },
+    {
+      "epoch": 0.37391772593306105,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0023955860913482956,
+      "loss": 1.2154,
+      "step": 4264
+    },
+    {
+      "epoch": 0.37400541770743556,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0023952461885912396,
+      "loss": 1.2157,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3740931094818101,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002394906217869655,
+      "loss": 1.2412,
+      "step": 4266
+    },
+    {
+      "epoch": 0.3741808012561847,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0023945661792145463,
+      "loss": 1.1501,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3742684930305592,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0023942260726569245,
+      "loss": 1.1912,
+      "step": 4268
+    },
+    {
+      "epoch": 0.37435618480493377,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002393885898227805,
+      "loss": 1.1913,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3744438765793083,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0023935456559582135,
+      "loss": 1.171,
+      "step": 4270
+    },
+    {
+      "epoch": 0.37453156835368284,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0023932053458791762,
+      "loss": 1.1646,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3746192601280574,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0023928649680217302,
+      "loss": 1.2058,
+      "step": 4272
+    },
+    {
+      "epoch": 0.3747069519024319,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0023925245224169175,
+      "loss": 1.1787,
+      "step": 4273
+    },
+    {
+      "epoch": 0.3747946436768065,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023921840090957848,
+      "loss": 1.2181,
+      "step": 4274
+    },
+    {
+      "epoch": 0.37488233545118105,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0023918434280893863,
+      "loss": 1.2035,
+      "step": 4275
+    },
+    {
+      "epoch": 0.37497002722555556,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0023915027794287826,
+      "loss": 1.1696,
+      "step": 4276
+    },
+    {
+      "epoch": 0.3750577189999301,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002391162063145039,
+      "loss": 1.1982,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3751454107743047,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002390821279269229,
+      "loss": 1.2026,
+      "step": 4278
+    },
+    {
+      "epoch": 0.3752331025486792,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002390480427832431,
+      "loss": 1.1698,
+      "step": 4279
+    },
+    {
+      "epoch": 0.37532079432305376,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0023901395088657285,
+      "loss": 1.1955,
+      "step": 4280
+    },
+    {
+      "epoch": 0.37540848609742833,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002389798522400214,
+      "loss": 1.1708,
+      "step": 4281
+    },
+    {
+      "epoch": 0.37549617787180284,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0023894574684669846,
+      "loss": 1.1912,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3755838696461774,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0023891163470971424,
+      "loss": 1.2321,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3756715614205519,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0023887751583217976,
+      "loss": 1.1552,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3757592531949265,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002388433902172065,
+      "loss": 1.1564,
+      "step": 4285
+    },
+    {
+      "epoch": 0.37584694496930104,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0023880925786790674,
+      "loss": 1.1676,
+      "step": 4286
+    },
+    {
+      "epoch": 0.37593463674367555,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0023877511878739317,
+      "loss": 1.2311,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3760223285180501,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0023874097297877923,
+      "loss": 1.2259,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3761100202924247,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002387068204451789,
+      "loss": 1.1995,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3761977120667992,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002386726611897068,
+      "loss": 1.1985,
+      "step": 4290
+    },
+    {
+      "epoch": 0.37628540384117376,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0023863849521547826,
+      "loss": 1.1792,
+      "step": 4291
+    },
+    {
+      "epoch": 0.3763730956155483,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0023860432252560903,
+      "loss": 1.1654,
+      "step": 4292
+    },
+    {
+      "epoch": 0.37646078738992284,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002385701431232155,
+      "loss": 1.2859,
+      "step": 4293
+    },
+    {
+      "epoch": 0.3765484791642974,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0023853595701141496,
+      "loss": 1.1748,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3766361709386719,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0023850176419332493,
+      "loss": 1.1836,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3767238627130465,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023846756467206373,
+      "loss": 1.2103,
+      "step": 4296
+    },
+    {
+      "epoch": 0.37681155448742104,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0023843335845075025,
+      "loss": 1.2232,
+      "step": 4297
+    },
+    {
+      "epoch": 0.37689924626179555,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002383991455325041,
+      "loss": 1.1305,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3769869380361701,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0023836492592044533,
+      "loss": 1.2417,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3770746298105447,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0023833069961769478,
+      "loss": 1.2355,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3771623215849192,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0023829646662737367,
+      "loss": 1.1322,
+      "step": 4301
+    },
+    {
+      "epoch": 0.37725001335929376,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0023826222695260395,
+      "loss": 1.2267,
+      "step": 4302
+    },
+    {
+      "epoch": 0.3773377051336683,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0023822798059650835,
+      "loss": 1.2061,
+      "step": 4303
+    },
+    {
+      "epoch": 0.37742539690804283,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0023819372756220996,
+      "loss": 1.1751,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3775130886824174,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002381594678528325,
+      "loss": 1.1727,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3776007804567919,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0023812520147150048,
+      "loss": 1.1614,
+      "step": 4306
+    },
+    {
+      "epoch": 0.37768847223116647,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0023809092842133886,
+      "loss": 1.1658,
+      "step": 4307
+    },
+    {
+      "epoch": 0.37777616400554104,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0023805664870547322,
+      "loss": 1.208,
+      "step": 4308
+    },
+    {
+      "epoch": 0.37786385577991555,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0023802236232702985,
+      "loss": 1.2419,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3779515475542901,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0023798806928913554,
+      "loss": 1.1895,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3780392393286647,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002379537695949177,
+      "loss": 1.2066,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3781269311030392,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002379194632475044,
+      "loss": 1.2361,
+      "step": 4312
+    },
+    {
+      "epoch": 0.37821462287741375,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002378851502500243,
+      "loss": 1.2504,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3783023146517883,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0023785083060560664,
+      "loss": 1.1892,
+      "step": 4314
+    },
+    {
+      "epoch": 0.37839000642616283,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002378165043173813,
+      "loss": 1.175,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3784776982005374,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0023778217138847877,
+      "loss": 1.1791,
+      "step": 4316
+    },
+    {
+      "epoch": 0.37856538997491196,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002377478318220301,
+      "loss": 1.1963,
+      "step": 4317
+    },
+    {
+      "epoch": 0.37865308174928647,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0023771348562116696,
+      "loss": 1.2317,
+      "step": 4318
+    },
+    {
+      "epoch": 0.37874077352366103,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0023767913278902163,
+      "loss": 1.1678,
+      "step": 4319
+    },
+    {
+      "epoch": 0.37882846529803554,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0023764477332872697,
+      "loss": 1.1759,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3789161570724101,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002376104072434165,
+      "loss": 1.188,
+      "step": 4321
+    },
+    {
+      "epoch": 0.3790038488467847,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002375760345362244,
+      "loss": 1.1986,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3790915406211592,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0023754165521028526,
+      "loss": 1.2153,
+      "step": 4323
+    },
+    {
+      "epoch": 0.37917923239553375,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0023750726926873443,
+      "loss": 1.1526,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3792669241699083,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002374728767147079,
+      "loss": 1.1888,
+      "step": 4325
+    },
+    {
+      "epoch": 0.3793546159442828,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00237438477551342,
+      "loss": 1.235,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3794423077186574,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0023740407178177406,
+      "loss": 1.167,
+      "step": 4327
+    },
+    {
+      "epoch": 0.37952999949303196,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0023736965940914156,
+      "loss": 1.222,
+      "step": 4328
+    },
+    {
+      "epoch": 0.37961769126740647,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00237335240436583,
+      "loss": 1.2199,
+      "step": 4329
+    },
+    {
+      "epoch": 0.37970538304178103,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0023730081486723724,
+      "loss": 1.2215,
+      "step": 4330
+    },
+    {
+      "epoch": 0.37979307481615554,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002372663827042438,
+      "loss": 1.2528,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3798807665905301,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002372319439507428,
+      "loss": 1.2067,
+      "step": 4332
+    },
+    {
+      "epoch": 0.37996845836490467,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00237197498609875,
+      "loss": 1.1944,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3800561501392792,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0023716304668478174,
+      "loss": 1.1633,
+      "step": 4334
+    },
+    {
+      "epoch": 0.38014384191365375,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0023712858817860486,
+      "loss": 1.2245,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3802315336880283,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0023709412309448692,
+      "loss": 1.2107,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3803192254624028,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0023705965143557112,
+      "loss": 1.1796,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3804069172367774,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002370251732050011,
+      "loss": 1.2441,
+      "step": 4338
+    },
+    {
+      "epoch": 0.38049460901115195,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0023699068840592123,
+      "loss": 1.1879,
+      "step": 4339
+    },
+    {
+      "epoch": 0.38058230078552646,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002369561970414764,
+      "loss": 1.1904,
+      "step": 4340
+    },
+    {
+      "epoch": 0.380669992559901,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0023692169911481214,
+      "loss": 1.2185,
+      "step": 4341
+    },
+    {
+      "epoch": 0.38075768433427554,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002368871946290746,
+      "loss": 1.1401,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3808453761086501,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002368526835874105,
+      "loss": 1.2125,
+      "step": 4343
+    },
+    {
+      "epoch": 0.38093306788302467,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0023681816599296717,
+      "loss": 1.2209,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3810207596573992,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0023678364184889244,
+      "loss": 1.2007,
+      "step": 4345
+    },
+    {
+      "epoch": 0.38110845143177374,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0023674911115833487,
+      "loss": 1.1789,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3811961432061483,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0023671457392444363,
+      "loss": 1.163,
+      "step": 4347
+    },
+    {
+      "epoch": 0.3812838349805228,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002366800301503684,
+      "loss": 1.1625,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3813715267548974,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0023664547983925943,
+      "loss": 1.1849,
+      "step": 4349
+    },
+    {
+      "epoch": 0.38145921852927195,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0023661092299426764,
+      "loss": 1.2015,
+      "step": 4350
+    },
+    {
+      "epoch": 0.38154691030364646,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0023657635961854455,
+      "loss": 1.1743,
+      "step": 4351
+    },
+    {
+      "epoch": 0.381634602078021,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0023654178971524225,
+      "loss": 1.1563,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3817222938523956,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002365072132875134,
+      "loss": 1.1945,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3818099856267701,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0023647263033851128,
+      "loss": 1.1769,
+      "step": 4354
+    },
+    {
+      "epoch": 0.38189767740114466,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023643804087138975,
+      "loss": 1.2177,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3819853691755192,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0023640344488930335,
+      "loss": 1.1471,
+      "step": 4356
+    },
+    {
+      "epoch": 0.38207306094989374,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002363688423954071,
+      "loss": 1.1954,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3821607527242683,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0023633423339285667,
+      "loss": 1.2234,
+      "step": 4358
+    },
+    {
+      "epoch": 0.3822484444986428,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0023629961788480826,
+      "loss": 1.2009,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3823361362730174,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0023626499587441883,
+      "loss": 1.1995,
+      "step": 4360
+    },
+    {
+      "epoch": 0.38242382804739194,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002362303673648457,
+      "loss": 1.1698,
+      "step": 4361
+    },
+    {
+      "epoch": 0.38251151982176645,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0023619573235924697,
+      "loss": 1.2508,
+      "step": 4362
+    },
+    {
+      "epoch": 0.382599211596141,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0023616109086078124,
+      "loss": 1.142,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3826869033705156,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0023612644287260764,
+      "loss": 1.1909,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3827745951448901,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002360917883978861,
+      "loss": 1.217,
+      "step": 4365
+    },
+    {
+      "epoch": 0.38286228691926466,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0023605712743977702,
+      "loss": 1.1299,
+      "step": 4366
+    },
+    {
+      "epoch": 0.38294997869363917,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0023602246000144133,
+      "loss": 1.2982,
+      "step": 4367
+    },
+    {
+      "epoch": 0.38303767046801374,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0023598778608604066,
+      "loss": 1.2395,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3831253622423883,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002359531056967372,
+      "loss": 1.1757,
+      "step": 4369
+    },
+    {
+      "epoch": 0.3832130540167628,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0023591841883669356,
+      "loss": 1.1603,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3833007457911374,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0023588372550907323,
+      "loss": 1.2002,
+      "step": 4371
+    },
+    {
+      "epoch": 0.38338843756551194,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0023584902571704013,
+      "loss": 1.1883,
+      "step": 4372
+    },
+    {
+      "epoch": 0.38347612933988645,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0023581431946375873,
+      "loss": 1.1866,
+      "step": 4373
+    },
+    {
+      "epoch": 0.383563821114261,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002357796067523943,
+      "loss": 1.1918,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3836515128886356,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002357448875861124,
+      "loss": 1.1796,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3837392046630101,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0023571016196807936,
+      "loss": 1.1758,
+      "step": 4376
+    },
+    {
+      "epoch": 0.38382689643738466,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0023567542990146208,
+      "loss": 1.1749,
+      "step": 4377
+    },
+    {
+      "epoch": 0.38391458821175917,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0023564069138942807,
+      "loss": 1.1758,
+      "step": 4378
+    },
+    {
+      "epoch": 0.38400227998613373,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0023560594643514537,
+      "loss": 1.2357,
+      "step": 4379
+    },
+    {
+      "epoch": 0.3840899717605083,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002355711950417826,
+      "loss": 1.1883,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3841776635348828,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0023553643721250903,
+      "loss": 1.1925,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3842653553092574,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002355016729504945,
+      "loss": 1.1407,
+      "step": 4382
+    },
+    {
+      "epoch": 0.38435304708363194,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0023546690225890936,
+      "loss": 1.1061,
+      "step": 4383
+    },
+    {
+      "epoch": 0.38444073885800645,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0023543212514092467,
+      "loss": 1.1525,
+      "step": 4384
+    },
+    {
+      "epoch": 0.384528430632381,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0023539734159971197,
+      "loss": 1.1491,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3846161224067556,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002353625516384434,
+      "loss": 1.2202,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3847038141811301,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0023532775526029183,
+      "loss": 1.2202,
+      "step": 4387
+    },
+    {
+      "epoch": 0.38479150595550465,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0023529295246843052,
+      "loss": 1.2204,
+      "step": 4388
+    },
+    {
+      "epoch": 0.38487919772987916,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002352581432660334,
+      "loss": 1.2238,
+      "step": 4389
+    },
+    {
+      "epoch": 0.38496688950425373,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0023522332765627493,
+      "loss": 1.1857,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3850545812786283,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0023518850564233033,
+      "loss": 1.1549,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3851422730530028,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0023515367722737514,
+      "loss": 1.1915,
+      "step": 4392
+    },
+    {
+      "epoch": 0.38522996482737737,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0023511884241458575,
+      "loss": 1.2433,
+      "step": 4393
+    },
+    {
+      "epoch": 0.38531765660175193,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0023508400120713898,
+      "loss": 1.2096,
+      "step": 4394
+    },
+    {
+      "epoch": 0.38540534837612644,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0023504915360821215,
+      "loss": 1.2293,
+      "step": 4395
+    },
+    {
+      "epoch": 0.385493040150501,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0023501429962098333,
+      "loss": 1.2381,
+      "step": 4396
+    },
+    {
+      "epoch": 0.3855807319248756,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0023497943924863118,
+      "loss": 1.1618,
+      "step": 4397
+    },
+    {
+      "epoch": 0.3856684236992501,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002349445724943348,
+      "loss": 1.1659,
+      "step": 4398
+    },
+    {
+      "epoch": 0.38575611547362465,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002349096993612739,
+      "loss": 1.245,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3858438072479992,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00234874819852629,
+      "loss": 1.2502,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3859314990223737,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0023483993397158086,
+      "loss": 1.159,
+      "step": 4401
+    },
+    {
+      "epoch": 0.3860191907967483,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002348050417213111,
+      "loss": 1.2358,
+      "step": 4402
+    },
+    {
+      "epoch": 0.3861068825711228,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002347701431050016,
+      "loss": 1.1797,
+      "step": 4403
+    },
+    {
+      "epoch": 0.38619457434549737,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0023473523812583526,
+      "loss": 1.1931,
+      "step": 4404
+    },
+    {
+      "epoch": 0.38628226611987193,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0023470032678699518,
+      "loss": 1.1632,
+      "step": 4405
+    },
+    {
+      "epoch": 0.38636995789424644,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0023466540909166523,
+      "loss": 1.1668,
+      "step": 4406
+    },
+    {
+      "epoch": 0.386457649668621,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0023463048504302982,
+      "loss": 1.2111,
+      "step": 4407
+    },
+    {
+      "epoch": 0.38654534144299557,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002345955546442739,
+      "loss": 1.2508,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3866330332173701,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002345606178985831,
+      "loss": 1.2213,
+      "step": 4409
+    },
+    {
+      "epoch": 0.38672072499174465,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002345256748091435,
+      "loss": 1.1249,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3868084167661192,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0023449072537914182,
+      "loss": 1.2265,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3868961085404937,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002344557696117654,
+      "loss": 1.165,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3869838003148683,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.002344208075102021,
+      "loss": 1.2191,
+      "step": 4413
+    },
+    {
+      "epoch": 0.3870714920892428,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0023438583907764033,
+      "loss": 1.2516,
+      "step": 4414
+    },
+    {
+      "epoch": 0.38715918386361736,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002343508643172692,
+      "loss": 1.1963,
+      "step": 4415
+    },
+    {
+      "epoch": 0.3872468756379919,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0023431588323227824,
+      "loss": 1.2059,
+      "step": 4416
+    },
+    {
+      "epoch": 0.38733456741236644,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0023428089582585767,
+      "loss": 1.2504,
+      "step": 4417
+    },
+    {
+      "epoch": 0.387422259186741,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002342459021011983,
+      "loss": 1.1835,
+      "step": 4418
+    },
+    {
+      "epoch": 0.38750995096111557,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0023421090206149128,
+      "loss": 1.1912,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3875976427354901,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002341758957099288,
+      "loss": 1.173,
+      "step": 4420
+    },
+    {
+      "epoch": 0.38768533450986464,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.002341408830497031,
+      "loss": 1.2235,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3877730262842392,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0023410586408400746,
+      "loss": 1.2032,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3878607180586137,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002340708388160353,
+      "loss": 1.2516,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3879484098329883,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0023403580724898104,
+      "loss": 1.1956,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3880361016073628,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002340007693860393,
+      "loss": 1.151,
+      "step": 4425
+    },
+    {
+      "epoch": 0.38812379338173736,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002339657252304056,
+      "loss": 1.2015,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3882114851561119,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002339306747852757,
+      "loss": 1.1914,
+      "step": 4427
+    },
+    {
+      "epoch": 0.38829917693048643,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0023389561805384628,
+      "loss": 1.2163,
+      "step": 4428
+    },
+    {
+      "epoch": 0.388386868704861,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002338605550393143,
+      "loss": 1.1853,
+      "step": 4429
+    },
+    {
+      "epoch": 0.38847456047923556,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002338254857448775,
+      "loss": 1.2342,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3885622522536101,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00233790410173734,
+      "loss": 1.2106,
+      "step": 4431
+    },
+    {
+      "epoch": 0.38864994402798464,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0023375532832908275,
+      "loss": 1.2122,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3887376358023592,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.00233720240214123,
+      "loss": 1.144,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3888253275767337,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0023368514583205478,
+      "loss": 1.2472,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3889130193511083,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0023365004518607854,
+      "loss": 1.1903,
+      "step": 4435
+    },
+    {
+      "epoch": 0.38900071112548285,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002336149382793954,
+      "loss": 1.1802,
+      "step": 4436
+    },
+    {
+      "epoch": 0.38908840289985736,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0023357982511520705,
+      "loss": 1.2091,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3891760946742319,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0023354470569671567,
+      "loss": 1.2581,
+      "step": 4438
+    },
+    {
+      "epoch": 0.38926378644860643,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0023350958002712405,
+      "loss": 1.2063,
+      "step": 4439
+    },
+    {
+      "epoch": 0.389351478222981,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0023347444810963565,
+      "loss": 1.1633,
+      "step": 4440
+    },
+    {
+      "epoch": 0.38943916999735556,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.002334393099474543,
+      "loss": 1.1685,
+      "step": 4441
+    },
+    {
+      "epoch": 0.38952686177173007,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0023340416554378457,
+      "loss": 1.1964,
+      "step": 4442
+    },
+    {
+      "epoch": 0.38961455354610464,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0023336901490183153,
+      "loss": 1.2225,
+      "step": 4443
+    },
+    {
+      "epoch": 0.3897022453204792,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0023333385802480084,
+      "loss": 1.2217,
+      "step": 4444
+    },
+    {
+      "epoch": 0.3897899370948537,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0023329869491589862,
+      "loss": 1.1519,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3898776288692283,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0023326352557833186,
+      "loss": 1.1687,
+      "step": 4446
+    },
+    {
+      "epoch": 0.38996532064360284,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.002332283500153077,
+      "loss": 1.1582,
+      "step": 4447
+    },
+    {
+      "epoch": 0.39005301241797735,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0023319316823003414,
+      "loss": 1.1993,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3901407041923519,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0023315798022571976,
+      "loss": 1.2458,
+      "step": 4449
+    },
+    {
+      "epoch": 0.3902283959667264,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0023312278600557345,
+      "loss": 1.1381,
+      "step": 4450
+    },
+    {
+      "epoch": 0.390316087741101,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0023308758557280497,
+      "loss": 1.1344,
+      "step": 4451
+    },
+    {
+      "epoch": 0.39040377951547556,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002330523789306244,
+      "loss": 1.1845,
+      "step": 4452
+    },
+    {
+      "epoch": 0.39049147128985007,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002330171660822426,
+      "loss": 1.1683,
+      "step": 4453
+    },
+    {
+      "epoch": 0.39057916306422463,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0023298194703087085,
+      "loss": 1.1937,
+      "step": 4454
+    },
+    {
+      "epoch": 0.3906668548385992,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00232946721779721,
+      "loss": 1.206,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3907545466129737,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0023291149033200552,
+      "loss": 1.1844,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3908422383873483,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002328762526909374,
+      "loss": 1.188,
+      "step": 4457
+    },
+    {
+      "epoch": 0.39092993016172284,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002328410088597303,
+      "loss": 1.2321,
+      "step": 4458
+    },
+    {
+      "epoch": 0.39101762193609735,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.002328057588415983,
+      "loss": 1.1785,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3911053137104719,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0023277050263975614,
+      "loss": 1.2156,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3911930054848464,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002327352402574191,
+      "loss": 1.1433,
+      "step": 4461
+    },
+    {
+      "epoch": 0.391280697259221,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0023269997169780305,
+      "loss": 1.2131,
+      "step": 4462
+    },
+    {
+      "epoch": 0.39136838903359555,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0023266469696412427,
+      "loss": 1.205,
+      "step": 4463
+    },
+    {
+      "epoch": 0.39145608080797006,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0023262941605959983,
+      "loss": 1.1669,
+      "step": 4464
+    },
+    {
+      "epoch": 0.39154377258234463,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0023259412898744723,
+      "loss": 1.1667,
+      "step": 4465
+    },
+    {
+      "epoch": 0.3916314643567192,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002325588357508846,
+      "loss": 1.2205,
+      "step": 4466
+    },
+    {
+      "epoch": 0.3917191561310937,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0023252353635313046,
+      "loss": 1.3022,
+      "step": 4467
+    },
+    {
+      "epoch": 0.39180684790546827,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002324882307974042,
+      "loss": 1.1683,
+      "step": 4468
+    },
+    {
+      "epoch": 0.39189453967984283,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002324529190869254,
+      "loss": 1.1594,
+      "step": 4469
+    },
+    {
+      "epoch": 0.39198223145421734,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0023241760122491464,
+      "loss": 1.196,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3920699232285919,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0023238227721459268,
+      "loss": 1.18,
+      "step": 4471
+    },
+    {
+      "epoch": 0.3921576150029665,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002323469470591809,
+      "loss": 1.1662,
+      "step": 4472
+    },
+    {
+      "epoch": 0.392245306777341,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0023231161076190145,
+      "loss": 1.214,
+      "step": 4473
+    },
+    {
+      "epoch": 0.39233299855171555,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0023227626832597686,
+      "loss": 1.2058,
+      "step": 4474
+    },
+    {
+      "epoch": 0.39242069032609006,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0023224091975463028,
+      "loss": 1.2157,
+      "step": 4475
+    },
+    {
+      "epoch": 0.3925083821004646,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0023220556505108535,
+      "loss": 1.2262,
+      "step": 4476
+    },
+    {
+      "epoch": 0.3925960738748392,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002321702042185664,
+      "loss": 1.1816,
+      "step": 4477
+    },
+    {
+      "epoch": 0.3926837656492137,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0023213483726029827,
+      "loss": 1.1763,
+      "step": 4478
+    },
+    {
+      "epoch": 0.39277145742358827,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0023209946417950627,
+      "loss": 1.1918,
+      "step": 4479
+    },
+    {
+      "epoch": 0.39285914919796283,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002320640849794163,
+      "loss": 1.124,
+      "step": 4480
+    },
+    {
+      "epoch": 0.39294684097233734,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0023202869966325495,
+      "loss": 1.2206,
+      "step": 4481
+    },
+    {
+      "epoch": 0.3930345327467119,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002319933082342492,
+      "loss": 1.2051,
+      "step": 4482
+    },
+    {
+      "epoch": 0.39312222452108647,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002319579106956267,
+      "loss": 1.2819,
+      "step": 4483
+    },
+    {
+      "epoch": 0.393209916295461,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0023192250705061563,
+      "loss": 1.2025,
+      "step": 4484
+    },
+    {
+      "epoch": 0.39329760806983555,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0023188709730244464,
+      "loss": 1.1656,
+      "step": 4485
+    },
+    {
+      "epoch": 0.39338529984421006,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0023185168145434306,
+      "loss": 1.259,
+      "step": 4486
+    },
+    {
+      "epoch": 0.3934729916185846,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0023181625950954073,
+      "loss": 1.2147,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3935606833929592,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0023178083147126793,
+      "loss": 1.2315,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3936483751673337,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0023174539734275574,
+      "loss": 1.1948,
+      "step": 4489
+    },
+    {
+      "epoch": 0.39373606694170826,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0023170995712723567,
+      "loss": 1.1979,
+      "step": 4490
+    },
+    {
+      "epoch": 0.39382375871608283,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0023167451082793966,
+      "loss": 1.2401,
+      "step": 4491
+    },
+    {
+      "epoch": 0.39391145049045734,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0023163905844810037,
+      "loss": 1.2192,
+      "step": 4492
+    },
+    {
+      "epoch": 0.3939991422648319,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0023160359999095104,
+      "loss": 1.183,
+      "step": 4493
+    },
+    {
+      "epoch": 0.39408683403920647,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0023156813545972526,
+      "loss": 1.154,
+      "step": 4494
+    },
+    {
+      "epoch": 0.394174525813581,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0023153266485765743,
+      "loss": 1.2804,
+      "step": 4495
+    },
+    {
+      "epoch": 0.39426221758795554,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0023149718818798225,
+      "loss": 1.1826,
+      "step": 4496
+    },
+    {
+      "epoch": 0.39434990936233005,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0023146170545393524,
+      "loss": 1.1874,
+      "step": 4497
+    },
+    {
+      "epoch": 0.3944376011367046,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002314262166587522,
+      "loss": 1.1727,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3945252929110792,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0023139072180566973,
+      "loss": 1.2275,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3946129846854537,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002313552208979248,
+      "loss": 1.1908,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3946129846854537,
+      "eval_loss": 1.2015705108642578,
+      "eval_runtime": 452.923,
+      "eval_samples_per_second": 31.897,
+      "eval_steps_per_second": 7.975,
+      "step": 4500
+    },
+    {
+      "epoch": 0.39470067645982826,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00231319713938755,
+      "loss": 1.1734,
+      "step": 4501
+    },
+    {
+      "epoch": 0.3947883682342028,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0023128420093139848,
+      "loss": 1.1641,
+      "step": 4502
+    },
+    {
+      "epoch": 0.39487606000857733,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0023124868187909403,
+      "loss": 1.1557,
+      "step": 4503
+    },
+    {
+      "epoch": 0.3949637517829519,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0023121315678508074,
+      "loss": 1.1579,
+      "step": 4504
+    },
+    {
+      "epoch": 0.39505144355732646,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002311776256525985,
+      "loss": 1.1644,
+      "step": 4505
+    },
+    {
+      "epoch": 0.395139135331701,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0023114208848488763,
+      "loss": 1.1869,
+      "step": 4506
+    },
+    {
+      "epoch": 0.39522682710607554,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0023110654528518903,
+      "loss": 1.2109,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3953145188804501,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0023107099605674418,
+      "loss": 1.2391,
+      "step": 4508
+    },
+    {
+      "epoch": 0.3954022106548246,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00231035440802795,
+      "loss": 1.2041,
+      "step": 4509
+    },
+    {
+      "epoch": 0.3954899024291992,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002309998795265841,
+      "loss": 1.2319,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3955775942035737,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002309643122313546,
+      "loss": 1.2718,
+      "step": 4511
+    },
+    {
+      "epoch": 0.39566528597794826,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002309287389203501,
+      "loss": 1.2,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3957529777523228,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0023089315959681483,
+      "loss": 1.2292,
+      "step": 4513
+    },
+    {
+      "epoch": 0.39584066952669733,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0023085757426399346,
+      "loss": 1.2115,
+      "step": 4514
+    },
+    {
+      "epoch": 0.3959283613010719,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0023082198292513137,
+      "loss": 1.1656,
+      "step": 4515
+    },
+    {
+      "epoch": 0.39601605307544646,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002307863855834743,
+      "loss": 1.2009,
+      "step": 4516
+    },
+    {
+      "epoch": 0.39610374484982097,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0023075078224226876,
+      "loss": 1.1837,
+      "step": 4517
+    },
+    {
+      "epoch": 0.39619143662419554,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0023071517290476156,
+      "loss": 1.174,
+      "step": 4518
+    },
+    {
+      "epoch": 0.3962791283985701,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0023067955757420026,
+      "loss": 1.1315,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3963668201729446,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0023064393625383288,
+      "loss": 1.1989,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3964545119473192,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0023060830894690793,
+      "loss": 1.178,
+      "step": 4521
+    },
+    {
+      "epoch": 0.3965422037216937,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002305726756566746,
+      "loss": 1.1502,
+      "step": 4522
+    },
+    {
+      "epoch": 0.39662989549606825,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0023053703638638246,
+      "loss": 1.2172,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3967175872704428,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0023050139113928187,
+      "loss": 1.1811,
+      "step": 4524
+    },
+    {
+      "epoch": 0.3968052790448173,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0023046573991862347,
+      "loss": 1.1775,
+      "step": 4525
+    },
+    {
+      "epoch": 0.3968929708191919,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002304300827276586,
+      "loss": 1.1931,
+      "step": 4526
+    },
+    {
+      "epoch": 0.39698066259356646,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0023039441956963906,
+      "loss": 1.2344,
+      "step": 4527
+    },
+    {
+      "epoch": 0.39706835436794097,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002303587504478173,
+      "loss": 1.2004,
+      "step": 4528
+    },
+    {
+      "epoch": 0.39715604614231553,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002303230753654462,
+      "loss": 1.2093,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3972437379166901,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0023028739432577926,
+      "loss": 1.1811,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3973314296910646,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002302517073320705,
+      "loss": 1.0967,
+      "step": 4531
+    },
+    {
+      "epoch": 0.3974191214654392,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0023021601438757444,
+      "loss": 1.2165,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3975068132398137,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002301803154955463,
+      "loss": 1.1675,
+      "step": 4533
+    },
+    {
+      "epoch": 0.39759450501418825,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002301446106592416,
+      "loss": 1.1931,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3976821967885628,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0023010889988191656,
+      "loss": 1.2032,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3977698885629373,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0023007318316682797,
+      "loss": 1.1837,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3978575803373119,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0023003746051723306,
+      "loss": 1.1994,
+      "step": 4537
+    },
+    {
+      "epoch": 0.39794527211168645,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002300017319363896,
+      "loss": 1.1971,
+      "step": 4538
+    },
+    {
+      "epoch": 0.39803296388606096,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0022996599742755607,
+      "loss": 1.1948,
+      "step": 4539
+    },
+    {
+      "epoch": 0.39812065566043553,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002299302569939912,
+      "loss": 1.1891,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3982083474348101,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0022989451063895453,
+      "loss": 1.188,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3982960392091846,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0022985875836570606,
+      "loss": 1.1454,
+      "step": 4542
+    },
+    {
+      "epoch": 0.39838373098355917,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002298230001775062,
+      "loss": 1.136,
+      "step": 4543
+    },
+    {
+      "epoch": 0.39847142275793374,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002297872360776161,
+      "loss": 1.2469,
+      "step": 4544
+    },
+    {
+      "epoch": 0.39855911453230825,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002297514660692973,
+      "loss": 1.2082,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3986468063066828,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0022971569015581197,
+      "loss": 1.1765,
+      "step": 4546
+    },
+    {
+      "epoch": 0.3987344980810573,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0022967990834042275,
+      "loss": 1.1949,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3988221898554319,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0022964412062639284,
+      "loss": 1.1734,
+      "step": 4548
+    },
+    {
+      "epoch": 0.39890988162980645,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00229608327016986,
+      "loss": 1.2095,
+      "step": 4549
+    },
+    {
+      "epoch": 0.39899757340418096,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0022957252751546657,
+      "loss": 1.1749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3990852651785555,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0022953672212509928,
+      "loss": 1.2266,
+      "step": 4551
+    },
+    {
+      "epoch": 0.3991729569529301,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002295009108491496,
+      "loss": 1.1793,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3992606487273046,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0022946509369088327,
+      "loss": 1.1168,
+      "step": 4553
+    },
+    {
+      "epoch": 0.39934834050167917,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0022942927065356684,
+      "loss": 1.2145,
+      "step": 4554
+    },
+    {
+      "epoch": 0.39943603227605373,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.002293934417404673,
+      "loss": 1.1621,
+      "step": 4555
+    },
+    {
+      "epoch": 0.39952372405042824,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0022935760695485204,
+      "loss": 1.1743,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3996114158248028,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002293217662999892,
+      "loss": 1.2235,
+      "step": 4557
+    },
+    {
+      "epoch": 0.3996991075991773,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0022928591977914733,
+      "loss": 1.2135,
+      "step": 4558
+    },
+    {
+      "epoch": 0.3997867993735519,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.002292500673955955,
+      "loss": 1.1473,
+      "step": 4559
+    },
+    {
+      "epoch": 0.39987449114792645,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0022921420915260338,
+      "loss": 1.186,
+      "step": 4560
+    },
+    {
+      "epoch": 0.39996218292230096,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0022917834505344114,
+      "loss": 1.1226,
+      "step": 4561
+    },
+    {
+      "epoch": 0.4000498746966755,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0022914247510137952,
+      "loss": 1.1331,
+      "step": 4562
+    },
+    {
+      "epoch": 0.4001375664710501,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002291065992996898,
+      "loss": 1.1763,
+      "step": 4563
+    },
+    {
+      "epoch": 0.4002252582454246,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022907071765164372,
+      "loss": 1.1668,
+      "step": 4564
+    },
+    {
+      "epoch": 0.40031295001979916,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0022903483016051347,
+      "loss": 1.2092,
+      "step": 4565
+    },
+    {
+      "epoch": 0.40040064179417373,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002289989368295721,
+      "loss": 1.1662,
+      "step": 4566
+    },
+    {
+      "epoch": 0.40048833356854824,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002289630376620929,
+      "loss": 1.2176,
+      "step": 4567
+    },
+    {
+      "epoch": 0.4005760253429228,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002289271326613498,
+      "loss": 1.183,
+      "step": 4568
+    },
+    {
+      "epoch": 0.4006637171172973,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002288912218306172,
+      "loss": 1.1982,
+      "step": 4569
+    },
+    {
+      "epoch": 0.4007514088916719,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002288553051731701,
+      "loss": 1.2129,
+      "step": 4570
+    },
+    {
+      "epoch": 0.40083910066604644,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0022881938269228405,
+      "loss": 1.2557,
+      "step": 4571
+    },
+    {
+      "epoch": 0.40092679244042095,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00228783454391235,
+      "loss": 1.1269,
+      "step": 4572
+    },
+    {
+      "epoch": 0.4010144842147955,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002287475202732996,
+      "loss": 1.1852,
+      "step": 4573
+    },
+    {
+      "epoch": 0.4011021759891701,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0022871158034175484,
+      "loss": 1.2035,
+      "step": 4574
+    },
+    {
+      "epoch": 0.4011898677635446,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0022867563459987856,
+      "loss": 1.2152,
+      "step": 4575
+    },
+    {
+      "epoch": 0.40127755953791916,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022863968305094867,
+      "loss": 1.1928,
+      "step": 4576
+    },
+    {
+      "epoch": 0.4013652513122937,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00228603725698244,
+      "loss": 1.1563,
+      "step": 4577
+    },
+    {
+      "epoch": 0.40145294308666823,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0022856776254504374,
+      "loss": 1.1701,
+      "step": 4578
+    },
+    {
+      "epoch": 0.4015406348610428,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0022853179359462765,
+      "loss": 1.2062,
+      "step": 4579
+    },
+    {
+      "epoch": 0.4016283266354173,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00228495818850276,
+      "loss": 1.19,
+      "step": 4580
+    },
+    {
+      "epoch": 0.4017160184097919,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0022845983831526954,
+      "loss": 1.2226,
+      "step": 4581
+    },
+    {
+      "epoch": 0.40180371018416644,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002284238519928897,
+      "loss": 1.1557,
+      "step": 4582
+    },
+    {
+      "epoch": 0.40189140195854095,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0022838785988641822,
+      "loss": 1.1986,
+      "step": 4583
+    },
+    {
+      "epoch": 0.4019790937329155,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.002283518619991376,
+      "loss": 1.189,
+      "step": 4584
+    },
+    {
+      "epoch": 0.4020667855072901,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002283158583343307,
+      "loss": 1.2521,
+      "step": 4585
+    },
+    {
+      "epoch": 0.4021544772816646,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0022827984889528096,
+      "loss": 1.2153,
+      "step": 4586
+    },
+    {
+      "epoch": 0.40224216905603916,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002282438336852723,
+      "loss": 1.1644,
+      "step": 4587
+    },
+    {
+      "epoch": 0.4023298608304137,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0022820781270758944,
+      "loss": 1.2138,
+      "step": 4588
+    },
+    {
+      "epoch": 0.40241755260478823,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0022817178596551708,
+      "loss": 1.2278,
+      "step": 4589
+    },
+    {
+      "epoch": 0.4025052443791628,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0022813575346234097,
+      "loss": 1.1527,
+      "step": 4590
+    },
+    {
+      "epoch": 0.40259293615353736,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002280997152013471,
+      "loss": 1.2198,
+      "step": 4591
+    },
+    {
+      "epoch": 0.40268062792791187,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0022806367118582216,
+      "loss": 1.1774,
+      "step": 4592
+    },
+    {
+      "epoch": 0.40276831970228644,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002280276214190531,
+      "loss": 1.201,
+      "step": 4593
+    },
+    {
+      "epoch": 0.40285601147666095,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0022799156590432774,
+      "loss": 1.149,
+      "step": 4594
+    },
+    {
+      "epoch": 0.4029437032510355,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0022795550464493417,
+      "loss": 1.1615,
+      "step": 4595
+    },
+    {
+      "epoch": 0.4030313950254101,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0022791943764416114,
+      "loss": 1.2375,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4031190867997846,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0022788336490529777,
+      "loss": 1.1875,
+      "step": 4597
+    },
+    {
+      "epoch": 0.40320677857415915,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0022784728643163387,
+      "loss": 1.2051,
+      "step": 4598
+    },
+    {
+      "epoch": 0.4032944703485337,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002278112022264597,
+      "loss": 1.2171,
+      "step": 4599
+    },
+    {
+      "epoch": 0.4033821621229082,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0022777511229306597,
+      "loss": 1.1783,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4034698538972828,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0022773901663474413,
+      "loss": 1.139,
+      "step": 4601
+    },
+    {
+      "epoch": 0.40355754567165736,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002277029152547859,
+      "loss": 1.1809,
+      "step": 4602
+    },
+    {
+      "epoch": 0.40364523744603187,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002276668081564837,
+      "loss": 1.1842,
+      "step": 4603
+    },
+    {
+      "epoch": 0.40373292922040643,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0022763069534313027,
+      "loss": 1.229,
+      "step": 4604
+    },
+    {
+      "epoch": 0.40382062099478094,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.002275945768180192,
+      "loss": 1.2391,
+      "step": 4605
+    },
+    {
+      "epoch": 0.4039083127691555,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002275584525844443,
+      "loss": 1.2351,
+      "step": 4606
+    },
+    {
+      "epoch": 0.4039960045435301,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0022752232264569997,
+      "loss": 1.2537,
+      "step": 4607
+    },
+    {
+      "epoch": 0.4040836963179046,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0022748618700508126,
+      "loss": 1.1731,
+      "step": 4608
+    },
+    {
+      "epoch": 0.40417138809227915,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002274500456658836,
+      "loss": 1.2042,
+      "step": 4609
+    },
+    {
+      "epoch": 0.4042590798666537,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0022741389863140293,
+      "loss": 1.1686,
+      "step": 4610
+    },
+    {
+      "epoch": 0.4043467716410282,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.002273777459049359,
+      "loss": 1.139,
+      "step": 4611
+    },
+    {
+      "epoch": 0.4044344634154028,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0022734158748977944,
+      "loss": 1.176,
+      "step": 4612
+    },
+    {
+      "epoch": 0.40452215518977735,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0022730542338923114,
+      "loss": 1.2132,
+      "step": 4613
+    },
+    {
+      "epoch": 0.40460984696415186,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0022726925360658902,
+      "loss": 1.1708,
+      "step": 4614
+    },
+    {
+      "epoch": 0.40469753873852643,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0022723307814515184,
+      "loss": 1.1551,
+      "step": 4615
+    },
+    {
+      "epoch": 0.40478523051290094,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0022719689700821852,
+      "loss": 1.1605,
+      "step": 4616
+    },
+    {
+      "epoch": 0.4048729222872755,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022716071019908872,
+      "loss": 1.2081,
+      "step": 4617
+    },
+    {
+      "epoch": 0.40496061406165007,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0022712451772106262,
+      "loss": 1.2364,
+      "step": 4618
+    },
+    {
+      "epoch": 0.4050483058360246,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002270883195774409,
+      "loss": 1.2434,
+      "step": 4619
+    },
+    {
+      "epoch": 0.40513599761039915,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002270521157715247,
+      "loss": 1.189,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4052236893847737,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0022701590630661577,
+      "loss": 1.1543,
+      "step": 4621
+    },
+    {
+      "epoch": 0.4053113811591482,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0022697969118601626,
+      "loss": 1.1583,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4053990729335228,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0022694347041302887,
+      "loss": 1.2092,
+      "step": 4623
+    },
+    {
+      "epoch": 0.40548676470789735,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0022690724399095693,
+      "loss": 1.2342,
+      "step": 4624
+    },
+    {
+      "epoch": 0.40557445648227186,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022687101192310414,
+      "loss": 1.1907,
+      "step": 4625
+    },
+    {
+      "epoch": 0.4056621482566464,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0022683477421277477,
+      "loss": 1.2634,
+      "step": 4626
+    },
+    {
+      "epoch": 0.405749840031021,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0022679853086327363,
+      "loss": 1.2333,
+      "step": 4627
+    },
+    {
+      "epoch": 0.4058375318053955,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00226762281877906,
+      "loss": 1.1411,
+      "step": 4628
+    },
+    {
+      "epoch": 0.40592522357977007,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002267260272599777,
+      "loss": 1.2024,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4060129153541446,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002266897670127951,
+      "loss": 1.1921,
+      "step": 4630
+    },
+    {
+      "epoch": 0.40610060712851914,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0022665350113966498,
+      "loss": 1.2146,
+      "step": 4631
+    },
+    {
+      "epoch": 0.4061882989028937,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0022661722964389466,
+      "loss": 1.1346,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4062759906772682,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0022658095252879214,
+      "loss": 1.1703,
+      "step": 4633
+    },
+    {
+      "epoch": 0.4063636824516428,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0022654466979766565,
+      "loss": 1.1633,
+      "step": 4634
+    },
+    {
+      "epoch": 0.40645137422601735,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0022650838145382417,
+      "loss": 1.1517,
+      "step": 4635
+    },
+    {
+      "epoch": 0.40653906600039186,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002264720875005771,
+      "loss": 1.2363,
+      "step": 4636
+    },
+    {
+      "epoch": 0.4066267577747664,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0022643578794123436,
+      "loss": 1.2092,
+      "step": 4637
+    },
+    {
+      "epoch": 0.406714449549141,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0022639948277910628,
+      "loss": 1.2184,
+      "step": 4638
+    },
+    {
+      "epoch": 0.4068021413235155,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0022636317201750397,
+      "loss": 1.1903,
+      "step": 4639
+    },
+    {
+      "epoch": 0.40688983309789006,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0022632685565973872,
+      "loss": 1.1191,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4069775248722646,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0022629053370912255,
+      "loss": 1.1288,
+      "step": 4641
+    },
+    {
+      "epoch": 0.40706521664663914,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002262542061689679,
+      "loss": 1.2015,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4071529084210137,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0022621787304258783,
+      "loss": 1.2369,
+      "step": 4643
+    },
+    {
+      "epoch": 0.4072406001953882,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0022618153433329575,
+      "loss": 1.2255,
+      "step": 4644
+    },
+    {
+      "epoch": 0.4073282919697628,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0022614519004440573,
+      "loss": 1.201,
+      "step": 4645
+    },
+    {
+      "epoch": 0.40741598374413734,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0022610884017923217,
+      "loss": 1.1814,
+      "step": 4646
+    },
+    {
+      "epoch": 0.40750367551851185,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002260724847410901,
+      "loss": 1.212,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4075913672928864,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022603612373329517,
+      "loss": 1.2029,
+      "step": 4648
+    },
+    {
+      "epoch": 0.407679059067261,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0022599975715916327,
+      "loss": 1.1847,
+      "step": 4649
+    },
+    {
+      "epoch": 0.4077667508416355,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00225963385022011,
+      "loss": 1.2206,
+      "step": 4650
+    },
+    {
+      "epoch": 0.40785444261601006,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002259270073251553,
+      "loss": 1.1588,
+      "step": 4651
+    },
+    {
+      "epoch": 0.40794213439038457,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0022589062407191397,
+      "loss": 1.2069,
+      "step": 4652
+    },
+    {
+      "epoch": 0.40802982616475914,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002258542352656048,
+      "loss": 1.1887,
+      "step": 4653
+    },
+    {
+      "epoch": 0.4081175179391337,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0022581784090954654,
+      "loss": 1.1534,
+      "step": 4654
+    },
+    {
+      "epoch": 0.4082052097135082,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.002257814410070581,
+      "loss": 1.2116,
+      "step": 4655
+    },
+    {
+      "epoch": 0.4082929014878828,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0022574503556145923,
+      "loss": 1.1905,
+      "step": 4656
+    },
+    {
+      "epoch": 0.40838059326225734,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0022570862457606993,
+      "loss": 1.2149,
+      "step": 4657
+    },
+    {
+      "epoch": 0.40846828503663185,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0022567220805421073,
+      "loss": 1.1825,
+      "step": 4658
+    },
+    {
+      "epoch": 0.4085559768110064,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0022563578599920277,
+      "loss": 1.1967,
+      "step": 4659
+    },
+    {
+      "epoch": 0.408643668585381,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002255993584143677,
+      "loss": 1.1686,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4087313603597555,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002255629253030276,
+      "loss": 1.1378,
+      "step": 4661
+    },
+    {
+      "epoch": 0.40881905213413006,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00225526486668505,
+      "loss": 1.1769,
+      "step": 4662
+    },
+    {
+      "epoch": 0.4089067439085046,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0022549004251412315,
+      "loss": 1.2198,
+      "step": 4663
+    },
+    {
+      "epoch": 0.40899443568287913,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0022545359284320554,
+      "loss": 1.2296,
+      "step": 4664
+    },
+    {
+      "epoch": 0.4090821274572537,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0022541713765907626,
+      "loss": 1.2574,
+      "step": 4665
+    },
+    {
+      "epoch": 0.4091698192316282,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0022538067696506003,
+      "loss": 1.1995,
+      "step": 4666
+    },
+    {
+      "epoch": 0.40925751100600277,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0022534421076448197,
+      "loss": 1.1303,
+      "step": 4667
+    },
+    {
+      "epoch": 0.40934520278037734,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022530773906066764,
+      "loss": 1.2421,
+      "step": 4668
+    },
+    {
+      "epoch": 0.40943289455475185,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0022527126185694314,
+      "loss": 1.2229,
+      "step": 4669
+    },
+    {
+      "epoch": 0.4095205863291264,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002252347791566352,
+      "loss": 1.1714,
+      "step": 4670
+    },
+    {
+      "epoch": 0.409608278103501,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002251982909630709,
+      "loss": 1.1859,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4096959698778755,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0022516179727957784,
+      "loss": 1.163,
+      "step": 4672
+    },
+    {
+      "epoch": 0.40978366165225005,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002251252981094842,
+      "loss": 1.211,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4098713534266246,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002250887934561186,
+      "loss": 1.2209,
+      "step": 4674
+    },
+    {
+      "epoch": 0.40995904520099913,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002250522833228101,
+      "loss": 1.1775,
+      "step": 4675
+    },
+    {
+      "epoch": 0.4100467369753737,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002250157677128884,
+      "loss": 1.1566,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4101344287497482,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0022497924662968362,
+      "loss": 1.1601,
+      "step": 4677
+    },
+    {
+      "epoch": 0.41022212052412277,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002249427200765264,
+      "loss": 1.1969,
+      "step": 4678
+    },
+    {
+      "epoch": 0.41030981229849733,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0022490618805674784,
+      "loss": 1.1831,
+      "step": 4679
+    },
+    {
+      "epoch": 0.41039750407287184,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002248696505736796,
+      "loss": 1.2024,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4104851958472464,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0022483310763065376,
+      "loss": 1.1946,
+      "step": 4681
+    },
+    {
+      "epoch": 0.410572887621621,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00224796559231003,
+      "loss": 1.1869,
+      "step": 4682
+    },
+    {
+      "epoch": 0.4106605793959955,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0022476000537806034,
+      "loss": 1.2125,
+      "step": 4683
+    },
+    {
+      "epoch": 0.41074827117037005,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0022472344607515947,
+      "loss": 1.1468,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4108359629447446,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002246868813256345,
+      "loss": 1.1662,
+      "step": 4685
+    },
+    {
+      "epoch": 0.4109236547191191,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0022465031113282005,
+      "loss": 1.2063,
+      "step": 4686
+    },
+    {
+      "epoch": 0.4110113464934937,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002246137355000512,
+      "loss": 1.1611,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4110990382678682,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002245771544306636,
+      "loss": 1.1623,
+      "step": 4688
+    },
+    {
+      "epoch": 0.41118673004224277,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002245405679279932,
+      "loss": 1.1834,
+      "step": 4689
+    },
+    {
+      "epoch": 0.41127442181661733,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002245039759953768,
+      "loss": 1.2068,
+      "step": 4690
+    },
+    {
+      "epoch": 0.41136211359099184,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0022446737863615145,
+      "loss": 1.1509,
+      "step": 4691
+    },
+    {
+      "epoch": 0.4114498053653664,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002244307758536546,
+      "loss": 1.1721,
+      "step": 4692
+    },
+    {
+      "epoch": 0.41153749713974097,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002243941676512244,
+      "loss": 1.1564,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4116251889141155,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.002243575540321995,
+      "loss": 1.1817,
+      "step": 4694
+    },
+    {
+      "epoch": 0.41171288068849005,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002243209349999189,
+      "loss": 1.1845,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4118005724628646,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002242843105577221,
+      "loss": 1.1246,
+      "step": 4696
+    },
+    {
+      "epoch": 0.4118882642372391,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0022424768070894922,
+      "loss": 1.1809,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4119759560116137,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0022421104545694084,
+      "loss": 1.1924,
+      "step": 4698
+    },
+    {
+      "epoch": 0.41206364778598825,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00224174404805038,
+      "loss": 1.1817,
+      "step": 4699
+    },
+    {
+      "epoch": 0.41215133956036276,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0022413775875658216,
+      "loss": 1.178,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4122390313347373,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0022410110731491536,
+      "loss": 1.1739,
+      "step": 4701
+    },
+    {
+      "epoch": 0.41232672310911184,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0022406445048338015,
+      "loss": 1.1165,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4124144148834864,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002240277882653196,
+      "loss": 1.1801,
+      "step": 4703
+    },
+    {
+      "epoch": 0.41250210665786097,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0022399112066407707,
+      "loss": 1.1847,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4125897984322355,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0022395444768299666,
+      "loss": 1.1788,
+      "step": 4705
+    },
+    {
+      "epoch": 0.41267749020661004,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0022391776932542276,
+      "loss": 1.2305,
+      "step": 4706
+    },
+    {
+      "epoch": 0.4127651819809846,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002238810855947004,
+      "loss": 1.2245,
+      "step": 4707
+    },
+    {
+      "epoch": 0.4128528737553591,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002238443964941751,
+      "loss": 1.2323,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4129405655297337,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002238077020271927,
+      "loss": 1.1676,
+      "step": 4709
+    },
+    {
+      "epoch": 0.41302825730410825,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002237710021970997,
+      "loss": 1.1967,
+      "step": 4710
+    },
+    {
+      "epoch": 0.41311594907848276,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00223734297007243,
+      "loss": 1.1317,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4132036408528573,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0022369758646097007,
+      "loss": 1.1089,
+      "step": 4712
+    },
+    {
+      "epoch": 0.41329133262723183,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002236608705616287,
+      "loss": 1.233,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4133790244016064,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0022362414931256748,
+      "loss": 1.1774,
+      "step": 4714
+    },
+    {
+      "epoch": 0.41346671617598096,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002235874227171351,
+      "loss": 1.154,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4135544079503555,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0022355069077868105,
+      "loss": 1.2476,
+      "step": 4716
+    },
+    {
+      "epoch": 0.41364209972473004,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0022351395350055513,
+      "loss": 1.1992,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4137297914991046,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002234772108861077,
+      "loss": 1.1768,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4138174832734791,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.002234404629386896,
+      "loss": 1.1977,
+      "step": 4719
+    },
+    {
+      "epoch": 0.4139051750478537,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002234037096616521,
+      "loss": 1.1995,
+      "step": 4720
+    },
+    {
+      "epoch": 0.41399286682222824,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0022336695105834715,
+      "loss": 1.1652,
+      "step": 4721
+    },
+    {
+      "epoch": 0.41408055859660275,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0022333018713212686,
+      "loss": 1.1718,
+      "step": 4722
+    },
+    {
+      "epoch": 0.4141682503709773,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002232934178863441,
+      "loss": 1.1668,
+      "step": 4723
+    },
+    {
+      "epoch": 0.41425594214535183,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0022325664332435214,
+      "loss": 1.2276,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4143436339197264,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0022321986344950467,
+      "loss": 1.1783,
+      "step": 4725
+    },
+    {
+      "epoch": 0.41443132569410096,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022318307826515596,
+      "loss": 1.211,
+      "step": 4726
+    },
+    {
+      "epoch": 0.41451901746847547,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0022314628777466068,
+      "loss": 1.2016,
+      "step": 4727
+    },
+    {
+      "epoch": 0.41460670924285004,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0022310949198137407,
+      "loss": 1.1836,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4146944010172246,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002230726908886518,
+      "loss": 1.177,
+      "step": 4729
+    },
+    {
+      "epoch": 0.4147820927915991,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0022303588449985,
+      "loss": 1.1801,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4148697845659737,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0022299907281832544,
+      "loss": 1.1692,
+      "step": 4731
+    },
+    {
+      "epoch": 0.41495747634034824,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002229622558474351,
+      "loss": 1.1842,
+      "step": 4732
+    },
+    {
+      "epoch": 0.41504516811472275,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022292543359053673,
+      "loss": 1.192,
+      "step": 4733
+    },
+    {
+      "epoch": 0.4151328598890973,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0022288860605098825,
+      "loss": 1.1334,
+      "step": 4734
+    },
+    {
+      "epoch": 0.4152205516634719,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0022285177323214836,
+      "loss": 1.1655,
+      "step": 4735
+    },
+    {
+      "epoch": 0.4153082434378464,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0022281493513737613,
+      "loss": 1.2125,
+      "step": 4736
+    },
+    {
+      "epoch": 0.41539593521222096,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0022277809177003102,
+      "loss": 1.2054,
+      "step": 4737
+    },
+    {
+      "epoch": 0.41548362698659547,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0022274124313347316,
+      "loss": 1.1648,
+      "step": 4738
+    },
+    {
+      "epoch": 0.41557131876097003,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002227043892310629,
+      "loss": 1.1985,
+      "step": 4739
+    },
+    {
+      "epoch": 0.4156590105353446,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002226675300661614,
+      "loss": 1.1521,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4157467023097191,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0022263066564212996,
+      "loss": 1.137,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4158343940840937,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0022259379596233065,
+      "loss": 1.2149,
+      "step": 4742
+    },
+    {
+      "epoch": 0.41592208585846824,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0022255692103012575,
+      "loss": 1.2328,
+      "step": 4743
+    },
+    {
+      "epoch": 0.41600977763284275,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0022252004084887824,
+      "loss": 1.2007,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4160974694072173,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0022248315542195154,
+      "loss": 1.2149,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4161851611815919,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002224462647527094,
+      "loss": 1.1812,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4162728529559664,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0022240936884451628,
+      "loss": 1.2071,
+      "step": 4747
+    },
+    {
+      "epoch": 0.41636054473034095,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002223724677007369,
+      "loss": 1.2017,
+      "step": 4748
+    },
+    {
+      "epoch": 0.41644823650471546,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0022233556132473654,
+      "loss": 1.1635,
+      "step": 4749
+    },
+    {
+      "epoch": 0.41653592827909003,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0022229864971988105,
+      "loss": 1.1337,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4166236200534646,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002222617328895366,
+      "loss": 1.1905,
+      "step": 4751
+    },
+    {
+      "epoch": 0.4167113118278391,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002222248108370699,
+      "loss": 1.1773,
+      "step": 4752
+    },
+    {
+      "epoch": 0.41679900360221367,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0022218788356584817,
+      "loss": 1.229,
+      "step": 4753
+    },
+    {
+      "epoch": 0.41688669537658823,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002221509510792391,
+      "loss": 1.1346,
+      "step": 4754
+    },
+    {
+      "epoch": 0.41697438715096274,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002221140133806109,
+      "loss": 1.1473,
+      "step": 4755
+    },
+    {
+      "epoch": 0.4170620789253373,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0022207707047333203,
+      "loss": 1.1951,
+      "step": 4756
+    },
+    {
+      "epoch": 0.4171497706997119,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022204012236077173,
+      "loss": 1.2017,
+      "step": 4757
+    },
+    {
+      "epoch": 0.4172374624740864,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002220031690462995,
+      "loss": 1.1507,
+      "step": 4758
+    },
+    {
+      "epoch": 0.41732515424846095,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0022196621053328543,
+      "loss": 1.1191,
+      "step": 4759
+    },
+    {
+      "epoch": 0.41741284602283546,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022192924682509995,
+      "loss": 1.1995,
+      "step": 4760
+    },
+    {
+      "epoch": 0.41750053779721,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002218922779251142,
+      "loss": 1.1791,
+      "step": 4761
+    },
+    {
+      "epoch": 0.4175882295715846,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0022185530383669948,
+      "loss": 1.1984,
+      "step": 4762
+    },
+    {
+      "epoch": 0.4176759213459591,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002218183245632279,
+      "loss": 1.1598,
+      "step": 4763
+    },
+    {
+      "epoch": 0.41776361312033367,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0022178134010807176,
+      "loss": 1.1963,
+      "step": 4764
+    },
+    {
+      "epoch": 0.41785130489470823,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00221744350474604,
+      "loss": 1.1663,
+      "step": 4765
+    },
+    {
+      "epoch": 0.41793899666908274,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0022170735566619795,
+      "loss": 1.1885,
+      "step": 4766
+    },
+    {
+      "epoch": 0.4180266884434573,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0022167035568622746,
+      "loss": 1.2168,
+      "step": 4767
+    },
+    {
+      "epoch": 0.41811438021783187,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002216333505380668,
+      "loss": 1.1708,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4182020719922064,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002215963402250908,
+      "loss": 1.1844,
+      "step": 4769
+    },
+    {
+      "epoch": 0.41828976376658095,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0022155932475067465,
+      "loss": 1.1693,
+      "step": 4770
+    },
+    {
+      "epoch": 0.4183774555409555,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0022152230411819408,
+      "loss": 1.2251,
+      "step": 4771
+    },
+    {
+      "epoch": 0.41846514731533,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0022148527833102533,
+      "loss": 1.1673,
+      "step": 4772
+    },
+    {
+      "epoch": 0.4185528390897046,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0022144824739254495,
+      "loss": 1.264,
+      "step": 4773
+    },
+    {
+      "epoch": 0.4186405308640791,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.002214112113061302,
+      "loss": 1.2484,
+      "step": 4774
+    },
+    {
+      "epoch": 0.41872822263845366,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002213741700751586,
+      "loss": 1.224,
+      "step": 4775
+    },
+    {
+      "epoch": 0.4188159144128282,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.002213371237030082,
+      "loss": 1.2451,
+      "step": 4776
+    },
+    {
+      "epoch": 0.41890360618720274,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0022130007219305753,
+      "loss": 1.2618,
+      "step": 4777
+    },
+    {
+      "epoch": 0.4189912979615773,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0022126301554868566,
+      "loss": 1.2119,
+      "step": 4778
+    },
+    {
+      "epoch": 0.41907898973595187,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0022122595377327193,
+      "loss": 1.156,
+      "step": 4779
+    },
+    {
+      "epoch": 0.4191666815103264,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0022118888687019645,
+      "loss": 1.1695,
+      "step": 4780
+    },
+    {
+      "epoch": 0.41925437328470094,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0022115181484283946,
+      "loss": 1.1473,
+      "step": 4781
+    },
+    {
+      "epoch": 0.4193420650590755,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0022111473769458197,
+      "loss": 1.2108,
+      "step": 4782
+    },
+    {
+      "epoch": 0.41942975683345,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0022107765542880528,
+      "loss": 1.1664,
+      "step": 4783
+    },
+    {
+      "epoch": 0.4195174486078246,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002210405680488911,
+      "loss": 1.1956,
+      "step": 4784
+    },
+    {
+      "epoch": 0.4196051403821991,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002210034755582219,
+      "loss": 1.2688,
+      "step": 4785
+    },
+    {
+      "epoch": 0.41969283215657366,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0022096637796018017,
+      "loss": 1.1942,
+      "step": 4786
+    },
+    {
+      "epoch": 0.4197805239309482,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002209292752581493,
+      "loss": 1.1719,
+      "step": 4787
+    },
+    {
+      "epoch": 0.41986821570532273,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0022089216745551287,
+      "loss": 1.1974,
+      "step": 4788
+    },
+    {
+      "epoch": 0.4199559074796973,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0022085505455565514,
+      "loss": 1.1684,
+      "step": 4789
+    },
+    {
+      "epoch": 0.42004359925407186,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0022081793656196056,
+      "loss": 1.2018,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4201312910284464,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002207808134778143,
+      "loss": 1.1625,
+      "step": 4791
+    },
+    {
+      "epoch": 0.42021898280282094,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002207436853066018,
+      "loss": 1.1541,
+      "step": 4792
+    },
+    {
+      "epoch": 0.4203066745771955,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0022070655205170915,
+      "loss": 1.1677,
+      "step": 4793
+    },
+    {
+      "epoch": 0.42039436635157,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0022066941371652275,
+      "loss": 1.1971,
+      "step": 4794
+    },
+    {
+      "epoch": 0.4204820581259446,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.002206322703044295,
+      "loss": 1.1807,
+      "step": 4795
+    },
+    {
+      "epoch": 0.4205697499003191,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0022059512181881683,
+      "loss": 1.2052,
+      "step": 4796
+    },
+    {
+      "epoch": 0.42065744167469366,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0022055796826307254,
+      "loss": 1.1778,
+      "step": 4797
+    },
+    {
+      "epoch": 0.4207451334490682,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00220520809640585,
+      "loss": 1.1647,
+      "step": 4798
+    },
+    {
+      "epoch": 0.42083282522344273,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002204836459547429,
+      "loss": 1.2059,
+      "step": 4799
+    },
+    {
+      "epoch": 0.4209205169978173,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0022044647720893557,
+      "loss": 1.1849,
+      "step": 4800
+    },
+    {
+      "epoch": 0.42100820877219186,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0022040930340655267,
+      "loss": 1.2287,
+      "step": 4801
+    },
+    {
+      "epoch": 0.42109590054656637,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002203721245509843,
+      "loss": 1.1924,
+      "step": 4802
+    },
+    {
+      "epoch": 0.42118359232094094,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002203349406456211,
+      "loss": 1.1663,
+      "step": 4803
+    },
+    {
+      "epoch": 0.4212712840953155,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002202977516938542,
+      "loss": 1.1895,
+      "step": 4804
+    },
+    {
+      "epoch": 0.42135897586969,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002202605576990751,
+      "loss": 1.1886,
+      "step": 4805
+    },
+    {
+      "epoch": 0.4214466676440646,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0022022335866467573,
+      "loss": 1.1478,
+      "step": 4806
+    },
+    {
+      "epoch": 0.4215343594184391,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0022018615459404867,
+      "loss": 1.1595,
+      "step": 4807
+    },
+    {
+      "epoch": 0.42162205119281365,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0022014894549058674,
+      "loss": 1.1607,
+      "step": 4808
+    },
+    {
+      "epoch": 0.4217097429671882,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0022011173135768336,
+      "loss": 1.2429,
+      "step": 4809
+    },
+    {
+      "epoch": 0.4217974347415627,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0022007451219873235,
+      "loss": 1.1888,
+      "step": 4810
+    },
+    {
+      "epoch": 0.4218851265159373,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00220037288017128,
+      "loss": 1.1911,
+      "step": 4811
+    },
+    {
+      "epoch": 0.42197281829031186,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0022000005881626507,
+      "loss": 1.2029,
+      "step": 4812
+    },
+    {
+      "epoch": 0.42206051006468637,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002199628245995387,
+      "loss": 1.2099,
+      "step": 4813
+    },
+    {
+      "epoch": 0.42214820183906093,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002199255853703447,
+      "loss": 1.1832,
+      "step": 4814
+    },
+    {
+      "epoch": 0.4222358936134355,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002198883411320791,
+      "loss": 1.2283,
+      "step": 4815
+    },
+    {
+      "epoch": 0.42232358538781,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.002198510918881384,
+      "loss": 1.208,
+      "step": 4816
+    },
+    {
+      "epoch": 0.4224112771621846,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0021981383764191975,
+      "loss": 1.146,
+      "step": 4817
+    },
+    {
+      "epoch": 0.42249896893655914,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0021977657839682067,
+      "loss": 1.2165,
+      "step": 4818
+    },
+    {
+      "epoch": 0.42258666071093365,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0021973931415623896,
+      "loss": 1.2154,
+      "step": 4819
+    },
+    {
+      "epoch": 0.4226743524853082,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0021970204492357316,
+      "loss": 1.1796,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4227620442596827,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0021966477070222206,
+      "loss": 1.1872,
+      "step": 4821
+    },
+    {
+      "epoch": 0.4228497360340573,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00219627491495585,
+      "loss": 1.1753,
+      "step": 4822
+    },
+    {
+      "epoch": 0.42293742780843185,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0021959020730706173,
+      "loss": 1.1361,
+      "step": 4823
+    },
+    {
+      "epoch": 0.42302511958280636,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0021955291814005243,
+      "loss": 1.2429,
+      "step": 4824
+    },
+    {
+      "epoch": 0.42311281135718093,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.002195156239979579,
+      "loss": 1.1787,
+      "step": 4825
+    },
+    {
+      "epoch": 0.4232005031315555,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0021947832488417914,
+      "loss": 1.1422,
+      "step": 4826
+    },
+    {
+      "epoch": 0.42328819490593,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0021944102080211783,
+      "loss": 1.1502,
+      "step": 4827
+    },
+    {
+      "epoch": 0.42337588668030457,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002194037117551759,
+      "loss": 1.1687,
+      "step": 4828
+    },
+    {
+      "epoch": 0.42346357845467913,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00219366397746756,
+      "loss": 1.1991,
+      "step": 4829
+    },
+    {
+      "epoch": 0.42355127022905364,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0021932907878026084,
+      "loss": 1.2254,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4236389620034282,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0021929175485909404,
+      "loss": 1.2396,
+      "step": 4831
+    },
+    {
+      "epoch": 0.4237266537778027,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002192544259866593,
+      "loss": 1.1812,
+      "step": 4832
+    },
+    {
+      "epoch": 0.4238143455521773,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0021921709216636102,
+      "loss": 1.2107,
+      "step": 4833
+    },
+    {
+      "epoch": 0.42390203732655185,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002191797534016039,
+      "loss": 1.2215,
+      "step": 4834
+    },
+    {
+      "epoch": 0.42398972910092636,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0021914240969579304,
+      "loss": 1.2262,
+      "step": 4835
+    },
+    {
+      "epoch": 0.4240774208753009,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0021910506105233428,
+      "loss": 1.2075,
+      "step": 4836
+    },
+    {
+      "epoch": 0.4241651126496755,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0021906770747463357,
+      "loss": 1.1603,
+      "step": 4837
+    },
+    {
+      "epoch": 0.42425280442405,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002190303489660975,
+      "loss": 1.1846,
+      "step": 4838
+    },
+    {
+      "epoch": 0.42434049619842457,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0021899298553013312,
+      "loss": 1.2529,
+      "step": 4839
+    },
+    {
+      "epoch": 0.42442818797279913,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0021895561717014786,
+      "loss": 1.1983,
+      "step": 4840
+    },
+    {
+      "epoch": 0.42451587974717364,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0021891824388954955,
+      "loss": 1.2227,
+      "step": 4841
+    },
+    {
+      "epoch": 0.4246035715215482,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.002188808656917466,
+      "loss": 1.1087,
+      "step": 4842
+    },
+    {
+      "epoch": 0.4246912632959227,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002188434825801478,
+      "loss": 1.2238,
+      "step": 4843
+    },
+    {
+      "epoch": 0.4247789550702973,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002188060945581624,
+      "loss": 1.1898,
+      "step": 4844
+    },
+    {
+      "epoch": 0.42486664684467185,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002187687016292001,
+      "loss": 1.1934,
+      "step": 4845
+    },
+    {
+      "epoch": 0.42495433861904636,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0021873130379667094,
+      "loss": 1.1658,
+      "step": 4846
+    },
+    {
+      "epoch": 0.4250420303934209,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0021869390106398563,
+      "loss": 1.1675,
+      "step": 4847
+    },
+    {
+      "epoch": 0.4251297221677955,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0021865649343455512,
+      "loss": 1.2309,
+      "step": 4848
+    },
+    {
+      "epoch": 0.42521741394217,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0021861908091179096,
+      "loss": 1.2049,
+      "step": 4849
+    },
+    {
+      "epoch": 0.42530510571654456,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00218581663499105,
+      "loss": 1.1983,
+      "step": 4850
+    },
+    {
+      "epoch": 0.42539279749091913,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0021854424119990973,
+      "loss": 1.2293,
+      "step": 4851
+    },
+    {
+      "epoch": 0.42548048926529364,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0021850681401761785,
+      "loss": 1.1584,
+      "step": 4852
+    },
+    {
+      "epoch": 0.4255681810396682,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0021846938195564255,
+      "loss": 1.1314,
+      "step": 4853
+    },
+    {
+      "epoch": 0.42565587281404277,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0021843194501739776,
+      "loss": 1.1285,
+      "step": 4854
+    },
+    {
+      "epoch": 0.4257435645884173,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0021839450320629747,
+      "loss": 1.1013,
+      "step": 4855
+    },
+    {
+      "epoch": 0.42583125636279184,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002183570565257564,
+      "loss": 1.1649,
+      "step": 4856
+    },
+    {
+      "epoch": 0.42591894813716635,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0021831960497918942,
+      "loss": 1.231,
+      "step": 4857
+    },
+    {
+      "epoch": 0.4260066399115409,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0021828214857001213,
+      "loss": 1.1807,
+      "step": 4858
+    },
+    {
+      "epoch": 0.4260943316859155,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002182446873016405,
+      "loss": 1.1603,
+      "step": 4859
+    },
+    {
+      "epoch": 0.42618202346029,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0021820722117749077,
+      "loss": 1.1681,
+      "step": 4860
+    },
+    {
+      "epoch": 0.42626971523466456,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002181697502009799,
+      "loss": 1.1983,
+      "step": 4861
+    },
+    {
+      "epoch": 0.4263574070090391,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00218132274375525,
+      "loss": 1.1868,
+      "step": 4862
+    },
+    {
+      "epoch": 0.42644509878341363,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0021809479370454386,
+      "loss": 1.1626,
+      "step": 4863
+    },
+    {
+      "epoch": 0.4265327905577882,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0021805730819145458,
+      "loss": 1.1947,
+      "step": 4864
+    },
+    {
+      "epoch": 0.42662048233216276,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002180198178396758,
+      "loss": 1.2031,
+      "step": 4865
+    },
+    {
+      "epoch": 0.4267081741065373,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0021798232265262643,
+      "loss": 1.1203,
+      "step": 4866
+    },
+    {
+      "epoch": 0.42679586588091184,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0021794482263372606,
+      "loss": 1.1481,
+      "step": 4867
+    },
+    {
+      "epoch": 0.42688355765528635,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0021790731778639455,
+      "loss": 1.199,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4269712494296609,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0021786980811405224,
+      "loss": 1.1754,
+      "step": 4869
+    },
+    {
+      "epoch": 0.4270589412040355,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0021783229362011985,
+      "loss": 1.1882,
+      "step": 4870
+    },
+    {
+      "epoch": 0.42714663297841,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0021779477430801876,
+      "loss": 1.1599,
+      "step": 4871
+    },
+    {
+      "epoch": 0.42723432475278456,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0021775725018117046,
+      "loss": 1.1987,
+      "step": 4872
+    },
+    {
+      "epoch": 0.4273220165271591,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002177197212429972,
+      "loss": 1.1567,
+      "step": 4873
+    },
+    {
+      "epoch": 0.42740970830153363,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0021768218749692136,
+      "loss": 1.1874,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4274974000759082,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0021764464894636617,
+      "loss": 1.241,
+      "step": 4875
+    },
+    {
+      "epoch": 0.42758509185028276,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0021760710559475477,
+      "loss": 1.1836,
+      "step": 4876
+    },
+    {
+      "epoch": 0.42767278362465727,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002175695574455112,
+      "loss": 1.1566,
+      "step": 4877
+    },
+    {
+      "epoch": 0.42776047539903184,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002175320045020597,
+      "loss": 1.1462,
+      "step": 4878
+    },
+    {
+      "epoch": 0.42784816717340635,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0021749444676782497,
+      "loss": 1.2024,
+      "step": 4879
+    },
+    {
+      "epoch": 0.4279358589477809,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002174568842462322,
+      "loss": 1.2287,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4280235507221555,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0021741931694070704,
+      "loss": 1.2622,
+      "step": 4881
+    },
+    {
+      "epoch": 0.42811124249653,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002173817448546755,
+      "loss": 1.1871,
+      "step": 4882
+    },
+    {
+      "epoch": 0.42819893427090455,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00217344167991564,
+      "loss": 1.1956,
+      "step": 4883
+    },
+    {
+      "epoch": 0.4282866260452791,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0021730658635479958,
+      "loss": 1.1308,
+      "step": 4884
+    },
+    {
+      "epoch": 0.4283743178196536,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0021726899994780947,
+      "loss": 1.2655,
+      "step": 4885
+    },
+    {
+      "epoch": 0.4284620095940282,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002172314087740215,
+      "loss": 1.1991,
+      "step": 4886
+    },
+    {
+      "epoch": 0.42854970136840276,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002171938128368639,
+      "loss": 1.1886,
+      "step": 4887
+    },
+    {
+      "epoch": 0.42863739314277727,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0021715621213976525,
+      "loss": 1.1792,
+      "step": 4888
+    },
+    {
+      "epoch": 0.42872508491715183,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002171186066861548,
+      "loss": 1.1964,
+      "step": 4889
+    },
+    {
+      "epoch": 0.4288127766915264,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0021708099647946184,
+      "loss": 1.1481,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4289004684659009,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0021704338152311654,
+      "loss": 1.192,
+      "step": 4891
+    },
+    {
+      "epoch": 0.4289881602402755,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002170057618205491,
+      "loss": 1.2446,
+      "step": 4892
+    },
+    {
+      "epoch": 0.42907585201465,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002169681373751906,
+      "loss": 1.1884,
+      "step": 4893
+    },
+    {
+      "epoch": 0.42916354378902455,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00216930508190472,
+      "loss": 1.2559,
+      "step": 4894
+    },
+    {
+      "epoch": 0.4292512355633991,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0021689287426982506,
+      "loss": 1.1841,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4293389273377736,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0021685523561668207,
+      "loss": 1.2165,
+      "step": 4896
+    },
+    {
+      "epoch": 0.4294266191121482,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002168175922344754,
+      "loss": 1.2199,
+      "step": 4897
+    },
+    {
+      "epoch": 0.42951431088652275,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002167799441266381,
+      "loss": 1.1957,
+      "step": 4898
+    },
+    {
+      "epoch": 0.42960200266089726,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002167422912966035,
+      "loss": 1.2114,
+      "step": 4899
+    },
+    {
+      "epoch": 0.42968969443527183,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0021670463374780556,
+      "loss": 1.1836,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4297773862096464,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002166669714836785,
+      "loss": 1.1967,
+      "step": 4901
+    },
+    {
+      "epoch": 0.4298650779840209,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0021662930450765706,
+      "loss": 1.1702,
+      "step": 4902
+    },
+    {
+      "epoch": 0.42995276975839547,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002165916328231763,
+      "loss": 1.1802,
+      "step": 4903
+    },
+    {
+      "epoch": 0.43004046153277,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0021655395643367177,
+      "loss": 1.1757,
+      "step": 4904
+    },
+    {
+      "epoch": 0.43012815330714455,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0021651627534257953,
+      "loss": 1.1943,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4302158450815191,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0021647858955333595,
+      "loss": 1.1604,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4303035368558936,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.002164408990693779,
+      "loss": 1.1118,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4303912286302682,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0021640320389414262,
+      "loss": 1.2022,
+      "step": 4908
+    },
+    {
+      "epoch": 0.43047892040464275,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002163655040310679,
+      "loss": 1.1361,
+      "step": 4909
+    },
+    {
+      "epoch": 0.43056661217901726,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002163277994835918,
+      "loss": 1.2088,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4306543039533918,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002162900902551529,
+      "loss": 1.1912,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4307419957277664,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002162523763491901,
+      "loss": 1.1541,
+      "step": 4912
+    },
+    {
+      "epoch": 0.4308296875021409,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0021621465776914296,
+      "loss": 1.2025,
+      "step": 4913
+    },
+    {
+      "epoch": 0.43091737927651547,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0021617693451845125,
+      "loss": 1.1347,
+      "step": 4914
+    },
+    {
+      "epoch": 0.43100507105089,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0021613920660055515,
+      "loss": 1.1467,
+      "step": 4915
+    },
+    {
+      "epoch": 0.43109276282526454,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0021610147401889546,
+      "loss": 1.1837,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4311804545996391,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.002160637367769133,
+      "loss": 1.1613,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4312681463740136,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002160259948780501,
+      "loss": 1.1833,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4313558381483882,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0021598824832574793,
+      "loss": 1.2067,
+      "step": 4919
+    },
+    {
+      "epoch": 0.43144352992276275,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0021595049712344913,
+      "loss": 1.2314,
+      "step": 4920
+    },
+    {
+      "epoch": 0.43153122169713726,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.002159127412745965,
+      "loss": 1.1826,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4316189134715118,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002158749807826333,
+      "loss": 1.2132,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4317066052458864,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002158372156510032,
+      "loss": 1.1945,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4317942970202609,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0021579944588315026,
+      "loss": 1.2673,
+      "step": 4924
+    },
+    {
+      "epoch": 0.43188198879463546,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00215761671482519,
+      "loss": 1.163,
+      "step": 4925
+    },
+    {
+      "epoch": 0.43196968056901003,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0021572389245255435,
+      "loss": 1.2013,
+      "step": 4926
+    },
+    {
+      "epoch": 0.43205737234338454,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002156861087967017,
+      "loss": 1.1685,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4321450641177591,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002156483205184066,
+      "loss": 1.152,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4322327558921336,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0021561052762111566,
+      "loss": 1.181,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4323204476665082,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002155727301082751,
+      "loss": 1.2436,
+      "step": 4930
+    },
+    {
+      "epoch": 0.43240813944088274,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002155349279833321,
+      "loss": 1.1037,
+      "step": 4931
+    },
+    {
+      "epoch": 0.43249583121525725,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0021549712124973417,
+      "loss": 1.1749,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4325835229896318,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002154593099109291,
+      "loss": 1.1924,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4326712147640064,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002154214939703652,
+      "loss": 1.1579,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4327589065383809,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0021538367343149132,
+      "loss": 1.1479,
+      "step": 4935
+    },
+    {
+      "epoch": 0.43284659831275546,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002153458482977564,
+      "loss": 1.1738,
+      "step": 4936
+    },
+    {
+      "epoch": 0.43293429008713,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0021530801857261015,
+      "loss": 1.2221,
+      "step": 4937
+    },
+    {
+      "epoch": 0.43302198186150453,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0021527018425950245,
+      "loss": 1.1382,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4331096736358791,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0021523234536188368,
+      "loss": 1.198,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4331973654102536,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002151945018832047,
+      "loss": 1.1888,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4332850571846282,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0021515665382691672,
+      "loss": 1.2007,
+      "step": 4941
+    },
+    {
+      "epoch": 0.43337274895900274,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002151188011964714,
+      "loss": 1.1962,
+      "step": 4942
+    },
+    {
+      "epoch": 0.43346044073337725,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002150809439953208,
+      "loss": 1.2089,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4335481325077518,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0021504308222691742,
+      "loss": 1.1976,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4336358242821264,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002150052158947141,
+      "loss": 1.1382,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4337235160565009,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0021496734500216414,
+      "loss": 1.1823,
+      "step": 4946
+    },
+    {
+      "epoch": 0.43381120783087546,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.002149294695527214,
+      "loss": 1.1644,
+      "step": 4947
+    },
+    {
+      "epoch": 0.43389889960525,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0021489158954983985,
+      "loss": 1.2014,
+      "step": 4948
+    },
+    {
+      "epoch": 0.43398659137962453,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0021485370499697413,
+      "loss": 1.1426,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4340742831539991,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002148158158975792,
+      "loss": 1.1685,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4341619749283736,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002147779222551105,
+      "loss": 1.1762,
+      "step": 4951
+    },
+    {
+      "epoch": 0.43424966670274817,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.002147400240730238,
+      "loss": 1.1752,
+      "step": 4952
+    },
+    {
+      "epoch": 0.43433735847712274,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0021470212135477527,
+      "loss": 1.1845,
+      "step": 4953
+    },
+    {
+      "epoch": 0.43442505025149725,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002146642141038216,
+      "loss": 1.1951,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4345127420258718,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0021462630232361984,
+      "loss": 1.1493,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4346004338002464,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0021458838601762734,
+      "loss": 1.2026,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4346881255746209,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0021455046518930212,
+      "loss": 1.1481,
+      "step": 4957
+    },
+    {
+      "epoch": 0.43477581734899545,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002145125398421023,
+      "loss": 1.198,
+      "step": 4958
+    },
+    {
+      "epoch": 0.43486350912337,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002144746099794867,
+      "loss": 1.1794,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4349512008977445,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0021443667560491447,
+      "loss": 1.1483,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4350388926721191,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00214398736721845,
+      "loss": 1.1619,
+      "step": 4961
+    },
+    {
+      "epoch": 0.43512658444649366,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0021436079333373823,
+      "loss": 1.1995,
+      "step": 4962
+    },
+    {
+      "epoch": 0.43521427622086817,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002143228454440545,
+      "loss": 1.1979,
+      "step": 4963
+    },
+    {
+      "epoch": 0.43530196799524273,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002142848930562547,
+      "loss": 1.1015,
+      "step": 4964
+    },
+    {
+      "epoch": 0.43538965976961724,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0021424693617379978,
+      "loss": 1.1995,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4354773515439918,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0021420897480015147,
+      "loss": 1.2003,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4355650433183664,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0021417100893877166,
+      "loss": 1.2204,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4356527350927409,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002141330385931228,
+      "loss": 1.1681,
+      "step": 4968
+    },
+    {
+      "epoch": 0.43574042686711545,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0021409506376666764,
+      "loss": 1.1656,
+      "step": 4969
+    },
+    {
+      "epoch": 0.43582811864149,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0021405708446286943,
+      "loss": 1.1585,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4359158104158645,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002140191006851917,
+      "loss": 1.155,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4360035021902391,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0021398111243709854,
+      "loss": 1.1486,
+      "step": 4972
+    },
+    {
+      "epoch": 0.43609119396461365,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002139431197220544,
+      "loss": 1.2036,
+      "step": 4973
+    },
+    {
+      "epoch": 0.43617888573898816,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0021390512254352407,
+      "loss": 1.1558,
+      "step": 4974
+    },
+    {
+      "epoch": 0.43626657751336273,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0021386712090497285,
+      "loss": 1.1237,
+      "step": 4975
+    },
+    {
+      "epoch": 0.43635426928773724,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0021382911480986627,
+      "loss": 1.2211,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4364419610621118,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002137911042616706,
+      "loss": 1.18,
+      "step": 4977
+    },
+    {
+      "epoch": 0.43652965283648637,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0021375308926385207,
+      "loss": 1.2392,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4366173446108609,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002137150698198777,
+      "loss": 1.1148,
+      "step": 4979
+    },
+    {
+      "epoch": 0.43670503638523545,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0021367704593321466,
+      "loss": 1.1992,
+      "step": 4980
+    },
+    {
+      "epoch": 0.43679272815961,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.002136390176073307,
+      "loss": 1.1714,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4368804199339845,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0021360098484569396,
+      "loss": 1.1588,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4369681117083591,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002135629476517728,
+      "loss": 1.265,
+      "step": 4983
+    },
+    {
+      "epoch": 0.43705580348273365,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0021352490602903626,
+      "loss": 1.1021,
+      "step": 4984
+    },
+    {
+      "epoch": 0.43714349525710816,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0021348685998095348,
+      "loss": 1.1914,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4372311870314827,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002134488095109943,
+      "loss": 1.1854,
+      "step": 4986
+    },
+    {
+      "epoch": 0.43731887880585724,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.002134107546226287,
+      "loss": 1.1916,
+      "step": 4987
+    },
+    {
+      "epoch": 0.4374065705802318,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002133726953193273,
+      "loss": 1.1949,
+      "step": 4988
+    },
+    {
+      "epoch": 0.43749426235460637,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0021333463160456093,
+      "loss": 1.1549,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4375819541289809,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002132965634818009,
+      "loss": 1.2031,
+      "step": 4990
+    },
+    {
+      "epoch": 0.43766964590335544,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0021325849095451903,
+      "loss": 1.2499,
+      "step": 4991
+    },
+    {
+      "epoch": 0.43775733767773,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0021322041402618733,
+      "loss": 1.2724,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4378450294521045,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002131823327002784,
+      "loss": 1.237,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4379327212264791,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0021314424698026502,
+      "loss": 1.1802,
+      "step": 4994
+    },
+    {
+      "epoch": 0.43802041300085365,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002131061568696207,
+      "loss": 1.2068,
+      "step": 4995
+    },
+    {
+      "epoch": 0.43810810477522816,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0021306806237181893,
+      "loss": 1.1827,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4381957965496027,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.002130299634903341,
+      "loss": 1.1625,
+      "step": 4997
+    },
+    {
+      "epoch": 0.43828348832397723,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0021299186022864045,
+      "loss": 1.1956,
+      "step": 4998
+    },
+    {
+      "epoch": 0.4383711800983518,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0021295375259021312,
+      "loss": 1.1608,
+      "step": 4999
+    },
+    {
+      "epoch": 0.43845887187272636,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002129156405785273,
+      "loss": 1.1611,
+      "step": 5000
+    },
+    {
+      "epoch": 0.43845887187272636,
+      "eval_loss": 1.1928588151931763,
+      "eval_runtime": 428.8018,
+      "eval_samples_per_second": 33.692,
+      "eval_steps_per_second": 8.423,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4385465636471009,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002128775241970588,
+      "loss": 1.2164,
+      "step": 5001
+    },
+    {
+      "epoch": 0.43863425542147544,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002128394034492837,
+      "loss": 1.1381,
+      "step": 5002
+    },
+    {
+      "epoch": 0.43872194719585,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0021280127833867846,
+      "loss": 1.1917,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4388096389702245,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0021276314886872006,
+      "loss": 1.1573,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4388973307445991,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0021272501504288575,
+      "loss": 1.1918,
+      "step": 5005
+    },
+    {
+      "epoch": 0.43898502251897364,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0021268687686465328,
+      "loss": 1.1492,
+      "step": 5006
+    },
+    {
+      "epoch": 0.43907271429334815,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002126487343375008,
+      "loss": 1.1615,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4391604060677227,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002126105874649067,
+      "loss": 1.1276,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4392480978420973,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002125724362503499,
+      "loss": 1.1589,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4393357896164718,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002125342806973098,
+      "loss": 1.1759,
+      "step": 5010
+    },
+    {
+      "epoch": 0.43942348139084636,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0021249612080926593,
+      "loss": 1.2428,
+      "step": 5011
+    },
+    {
+      "epoch": 0.43951117316522087,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0021245795658969852,
+      "loss": 1.1711,
+      "step": 5012
+    },
+    {
+      "epoch": 0.43959886493959544,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00212419788042088,
+      "loss": 1.1816,
+      "step": 5013
+    },
+    {
+      "epoch": 0.43968655671397,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0021238161516991512,
+      "loss": 1.1812,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4397742484883445,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002123434379766613,
+      "loss": 1.2394,
+      "step": 5015
+    },
+    {
+      "epoch": 0.4398619402627191,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0021230525646580814,
+      "loss": 1.1626,
+      "step": 5016
+    },
+    {
+      "epoch": 0.43994963203709364,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0021226707064083772,
+      "loss": 1.1584,
+      "step": 5017
+    },
+    {
+      "epoch": 0.44003732381146815,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002122288805052325,
+      "loss": 1.2637,
+      "step": 5018
+    },
+    {
+      "epoch": 0.4401250155858427,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002121906860624753,
+      "loss": 1.1642,
+      "step": 5019
+    },
+    {
+      "epoch": 0.4402127073602173,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0021215248731604936,
+      "loss": 1.167,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4403003991345918,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.002121142842694383,
+      "loss": 1.1973,
+      "step": 5021
+    },
+    {
+      "epoch": 0.44038809090896636,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002120760769261261,
+      "loss": 1.1185,
+      "step": 5022
+    },
+    {
+      "epoch": 0.44047578268334087,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0021203786528959717,
+      "loss": 1.2457,
+      "step": 5023
+    },
+    {
+      "epoch": 0.44056347445771543,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0021199964936333645,
+      "loss": 1.1167,
+      "step": 5024
+    },
+    {
+      "epoch": 0.44065116623209,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00211961429150829,
+      "loss": 1.1496,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4407388580064645,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0021192320465556044,
+      "loss": 1.1974,
+      "step": 5026
+    },
+    {
+      "epoch": 0.44082654978083907,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0021188497588101674,
+      "loss": 1.204,
+      "step": 5027
+    },
+    {
+      "epoch": 0.44091424155521364,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002118467428306843,
+      "loss": 1.1446,
+      "step": 5028
+    },
+    {
+      "epoch": 0.44100193332958815,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0021180850550804983,
+      "loss": 1.2387,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4410896251039627,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0021177026391660044,
+      "loss": 1.1794,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4411773168783373,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.002117320180598238,
+      "loss": 1.2515,
+      "step": 5031
+    },
+    {
+      "epoch": 0.4412650086527118,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0021169376794120767,
+      "loss": 1.2179,
+      "step": 5032
+    },
+    {
+      "epoch": 0.44135270042708635,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0021165551356424053,
+      "loss": 1.1867,
+      "step": 5033
+    },
+    {
+      "epoch": 0.44144039220146086,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0021161725493241096,
+      "loss": 1.2549,
+      "step": 5034
+    },
+    {
+      "epoch": 0.44152808397583543,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002115789920492081,
+      "loss": 1.15,
+      "step": 5035
+    },
+    {
+      "epoch": 0.44161577575021,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0021154072491812136,
+      "loss": 1.216,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4417034675245845,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0021150245354264073,
+      "loss": 1.1308,
+      "step": 5037
+    },
+    {
+      "epoch": 0.44179115929895907,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002114641779262563,
+      "loss": 1.2273,
+      "step": 5038
+    },
+    {
+      "epoch": 0.44187885107333363,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002114258980724588,
+      "loss": 1.23,
+      "step": 5039
+    },
+    {
+      "epoch": 0.44196654284770814,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002113876139847393,
+      "loss": 1.19,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4420542346220827,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0021134932566658905,
+      "loss": 1.1712,
+      "step": 5041
+    },
+    {
+      "epoch": 0.4421419263964573,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002113110331215001,
+      "loss": 1.177,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4422296181708318,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0021127273635296435,
+      "loss": 1.2157,
+      "step": 5043
+    },
+    {
+      "epoch": 0.44231730994520635,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0021123443536447454,
+      "loss": 1.1866,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4424050017195809,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0021119613015952356,
+      "loss": 1.1193,
+      "step": 5045
+    },
+    {
+      "epoch": 0.4424926934939554,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002111578207416048,
+      "loss": 1.1052,
+      "step": 5046
+    },
+    {
+      "epoch": 0.44258038526833,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0021111950711421197,
+      "loss": 1.2488,
+      "step": 5047
+    },
+    {
+      "epoch": 0.4426680770427045,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0021108118928083912,
+      "loss": 1.2453,
+      "step": 5048
+    },
+    {
+      "epoch": 0.44275576881707907,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0021104286724498073,
+      "loss": 1.2521,
+      "step": 5049
+    },
+    {
+      "epoch": 0.44284346059145363,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0021100454101013176,
+      "loss": 1.1598,
+      "step": 5050
+    },
+    {
+      "epoch": 0.44293115236582814,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0021096621057978735,
+      "loss": 1.1938,
+      "step": 5051
+    },
+    {
+      "epoch": 0.4430188441402027,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002109278759574432,
+      "loss": 1.1463,
+      "step": 5052
+    },
+    {
+      "epoch": 0.44310653591457727,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002108895371465954,
+      "loss": 1.1279,
+      "step": 5053
+    },
+    {
+      "epoch": 0.4431942276889518,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002108511941507402,
+      "loss": 1.1682,
+      "step": 5054
+    },
+    {
+      "epoch": 0.44328191946332635,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.002108128469733745,
+      "loss": 1.1934,
+      "step": 5055
+    },
+    {
+      "epoch": 0.4433696112377009,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0021077449561799537,
+      "loss": 1.1987,
+      "step": 5056
+    },
+    {
+      "epoch": 0.4434573030120754,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002107361400881005,
+      "loss": 1.1953,
+      "step": 5057
+    },
+    {
+      "epoch": 0.44354499478645,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0021069778038718765,
+      "loss": 1.2273,
+      "step": 5058
+    },
+    {
+      "epoch": 0.4436326865608245,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002106594165187552,
+      "loss": 1.1607,
+      "step": 5059
+    },
+    {
+      "epoch": 0.44372037833519906,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002106210484863018,
+      "loss": 1.1598,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4438080701095736,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0021058267629332664,
+      "loss": 1.2293,
+      "step": 5061
+    },
+    {
+      "epoch": 0.44389576188394814,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00210544299943329,
+      "loss": 1.2128,
+      "step": 5062
+    },
+    {
+      "epoch": 0.4439834536583227,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.002105059194398088,
+      "loss": 1.1795,
+      "step": 5063
+    },
+    {
+      "epoch": 0.44407114543269727,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0021046753478626626,
+      "loss": 1.217,
+      "step": 5064
+    },
+    {
+      "epoch": 0.4441588372070718,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002104291459862018,
+      "loss": 1.1526,
+      "step": 5065
+    },
+    {
+      "epoch": 0.44424652898144634,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0021039075304311664,
+      "loss": 1.2137,
+      "step": 5066
+    },
+    {
+      "epoch": 0.4443342207558209,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002103523559605119,
+      "loss": 1.2328,
+      "step": 5067
+    },
+    {
+      "epoch": 0.4444219125301954,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002103139547418894,
+      "loss": 1.1448,
+      "step": 5068
+    },
+    {
+      "epoch": 0.44450960430457,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002102755493907512,
+      "loss": 1.1743,
+      "step": 5069
+    },
+    {
+      "epoch": 0.4445972960789445,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002102371399105998,
+      "loss": 1.2346,
+      "step": 5070
+    },
+    {
+      "epoch": 0.44468498785331906,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00210198726304938,
+      "loss": 1.1625,
+      "step": 5071
+    },
+    {
+      "epoch": 0.4447726796276936,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0021016030857726902,
+      "loss": 1.1455,
+      "step": 5072
+    },
+    {
+      "epoch": 0.44486037140206813,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002101218867310965,
+      "loss": 1.1301,
+      "step": 5073
+    },
+    {
+      "epoch": 0.4449480631764427,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0021008346076992436,
+      "loss": 1.1499,
+      "step": 5074
+    },
+    {
+      "epoch": 0.44503575495081726,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0021004503069725704,
+      "loss": 1.1597,
+      "step": 5075
+    },
+    {
+      "epoch": 0.4451234467251918,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0021000659651659913,
+      "loss": 1.1519,
+      "step": 5076
+    },
+    {
+      "epoch": 0.44521113849956634,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0020996815823145585,
+      "loss": 1.1608,
+      "step": 5077
+    },
+    {
+      "epoch": 0.4452988302739409,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.002099297158453326,
+      "loss": 1.2376,
+      "step": 5078
+    },
+    {
+      "epoch": 0.4453865220483154,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0020989126936173524,
+      "loss": 1.1929,
+      "step": 5079
+    },
+    {
+      "epoch": 0.44547421382269,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0020985281878417006,
+      "loss": 1.2082,
+      "step": 5080
+    },
+    {
+      "epoch": 0.44556190559706454,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002098143641161436,
+      "loss": 1.2023,
+      "step": 5081
+    },
+    {
+      "epoch": 0.44564959737143905,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002097759053611627,
+      "loss": 1.216,
+      "step": 5082
+    },
+    {
+      "epoch": 0.4457372891458136,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0020973744252273494,
+      "loss": 1.1852,
+      "step": 5083
+    },
+    {
+      "epoch": 0.44582498092018813,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0020969897560436787,
+      "loss": 1.2432,
+      "step": 5084
+    },
+    {
+      "epoch": 0.4459126726945627,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002096605046095696,
+      "loss": 1.2106,
+      "step": 5085
+    },
+    {
+      "epoch": 0.44600036446893726,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0020962202954184863,
+      "loss": 1.1587,
+      "step": 5086
+    },
+    {
+      "epoch": 0.44608805624331177,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.002095835504047137,
+      "loss": 1.1774,
+      "step": 5087
+    },
+    {
+      "epoch": 0.44617574801768634,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0020954506720167417,
+      "loss": 1.1778,
+      "step": 5088
+    },
+    {
+      "epoch": 0.4462634397920609,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.002095065799362394,
+      "loss": 1.1658,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4463511315664354,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0020946808861191948,
+      "loss": 1.1408,
+      "step": 5090
+    },
+    {
+      "epoch": 0.44643882334081,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002094295932322247,
+      "loss": 1.1782,
+      "step": 5091
+    },
+    {
+      "epoch": 0.44652651511518454,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0020939109380066568,
+      "loss": 1.1967,
+      "step": 5092
+    },
+    {
+      "epoch": 0.44661420688955905,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0020935259032075345,
+      "loss": 1.1205,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4467018986639336,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.002093140827959995,
+      "loss": 1.1427,
+      "step": 5094
+    },
+    {
+      "epoch": 0.4467895904383081,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0020927557122991563,
+      "loss": 1.1853,
+      "step": 5095
+    },
+    {
+      "epoch": 0.4468772822126827,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00209237055626014,
+      "loss": 1.126,
+      "step": 5096
+    },
+    {
+      "epoch": 0.44696497398705726,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0020919853598780703,
+      "loss": 1.1208,
+      "step": 5097
+    },
+    {
+      "epoch": 0.44705266576143177,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002091600123188077,
+      "loss": 1.2101,
+      "step": 5098
+    },
+    {
+      "epoch": 0.44714035753580633,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002091214846225292,
+      "loss": 1.1911,
+      "step": 5099
+    },
+    {
+      "epoch": 0.4472280493101809,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0020908295290248528,
+      "loss": 1.1845,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4473157410845554,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002090444171621898,
+      "loss": 1.2307,
+      "step": 5101
+    },
+    {
+      "epoch": 0.44740343285893,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020900587740515716,
+      "loss": 1.1727,
+      "step": 5102
+    },
+    {
+      "epoch": 0.44749112463330454,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0020896733363490218,
+      "loss": 1.1761,
+      "step": 5103
+    },
+    {
+      "epoch": 0.44757881640767905,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0020892878585493974,
+      "loss": 1.2429,
+      "step": 5104
+    },
+    {
+      "epoch": 0.4476665081820536,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0020889023406878557,
+      "loss": 1.2114,
+      "step": 5105
+    },
+    {
+      "epoch": 0.4477541999564281,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002088516782799553,
+      "loss": 1.2543,
+      "step": 5106
+    },
+    {
+      "epoch": 0.4478418917308027,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0020881311849196517,
+      "loss": 1.2172,
+      "step": 5107
+    },
+    {
+      "epoch": 0.44792958350517725,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0020877455470833177,
+      "loss": 1.1936,
+      "step": 5108
+    },
+    {
+      "epoch": 0.44801727527955176,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002087359869325719,
+      "loss": 1.2439,
+      "step": 5109
+    },
+    {
+      "epoch": 0.44810496705392633,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00208697415168203,
+      "loss": 1.2146,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4481926588283009,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0020865883941874257,
+      "loss": 1.1201,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4482803506026754,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.002086202596877088,
+      "loss": 1.1521,
+      "step": 5112
+    },
+    {
+      "epoch": 0.44836804237704997,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0020858167597861987,
+      "loss": 1.1847,
+      "step": 5113
+    },
+    {
+      "epoch": 0.44845573415142453,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0020854308829499463,
+      "loss": 1.2262,
+      "step": 5114
+    },
+    {
+      "epoch": 0.44854342592579904,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020850449664035213,
+      "loss": 1.1292,
+      "step": 5115
+    },
+    {
+      "epoch": 0.4486311177001736,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002084659010182118,
+      "loss": 1.2237,
+      "step": 5116
+    },
+    {
+      "epoch": 0.4487188094745482,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0020842730143209357,
+      "loss": 1.2155,
+      "step": 5117
+    },
+    {
+      "epoch": 0.4488065012489227,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0020838869788551754,
+      "loss": 1.1337,
+      "step": 5118
+    },
+    {
+      "epoch": 0.44889419302329725,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0020835009038200425,
+      "loss": 1.173,
+      "step": 5119
+    },
+    {
+      "epoch": 0.44898188479767176,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.002083114789250746,
+      "loss": 1.1426,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4490695765720463,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0020827286351824997,
+      "loss": 1.115,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4491572683464209,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0020823424416505187,
+      "loss": 1.1145,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4492449601207954,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002081956208690022,
+      "loss": 1.1264,
+      "step": 5123
+    },
+    {
+      "epoch": 0.44933265189516997,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0020815699363362357,
+      "loss": 1.1835,
+      "step": 5124
+    },
+    {
+      "epoch": 0.44942034366954453,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0020811836246243842,
+      "loss": 1.1939,
+      "step": 5125
+    },
+    {
+      "epoch": 0.44950803544391904,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0020807972735896996,
+      "loss": 1.2149,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4495957272182936,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0020804108832674162,
+      "loss": 1.2162,
+      "step": 5127
+    },
+    {
+      "epoch": 0.44968341899266817,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0020800244536927706,
+      "loss": 1.1516,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4497711107670427,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0020796379849010054,
+      "loss": 1.1648,
+      "step": 5129
+    },
+    {
+      "epoch": 0.44985880254141725,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002079251476927365,
+      "loss": 1.146,
+      "step": 5130
+    },
+    {
+      "epoch": 0.44994649431579176,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0020788649298070978,
+      "loss": 1.1734,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4500341860901663,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002078478343575456,
+      "loss": 1.1656,
+      "step": 5132
+    },
+    {
+      "epoch": 0.4501218778645409,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0020780917182676955,
+      "loss": 1.2233,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4502095696389154,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002077705053919076,
+      "loss": 1.1914,
+      "step": 5134
+    },
+    {
+      "epoch": 0.45029726141328996,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002077318350564859,
+      "loss": 1.1821,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4503849531876645,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002076931608240312,
+      "loss": 1.2019,
+      "step": 5136
+    },
+    {
+      "epoch": 0.45047264496203904,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002076544826980704,
+      "loss": 1.1701,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4505603367364136,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002076158006821309,
+      "loss": 1.1287,
+      "step": 5138
+    },
+    {
+      "epoch": 0.45064802851078817,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0020757711477974046,
+      "loss": 1.2193,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4507357202851627,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0020753842499442697,
+      "loss": 1.2386,
+      "step": 5140
+    },
+    {
+      "epoch": 0.45082341205953724,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0020749973132971897,
+      "loss": 1.1771,
+      "step": 5141
+    },
+    {
+      "epoch": 0.45091110383391175,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002074610337891452,
+      "loss": 1.1241,
+      "step": 5142
+    },
+    {
+      "epoch": 0.4509987956082863,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0020742233237623473,
+      "loss": 1.1377,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4510864873826609,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0020738362709451707,
+      "loss": 1.1751,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4511741791570354,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002073449179475221,
+      "loss": 1.1675,
+      "step": 5145
+    },
+    {
+      "epoch": 0.45126187093140996,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0020730620493877978,
+      "loss": 1.1716,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4513495627057845,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002072674880718209,
+      "loss": 1.1977,
+      "step": 5147
+    },
+    {
+      "epoch": 0.45143725448015903,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0020722876735017617,
+      "loss": 1.1476,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4515249462545336,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002071900427773769,
+      "loss": 1.2445,
+      "step": 5149
+    },
+    {
+      "epoch": 0.45161263802890816,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020715131435695465,
+      "loss": 1.1702,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4517003298032827,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0020711258209244127,
+      "loss": 1.138,
+      "step": 5151
+    },
+    {
+      "epoch": 0.45178802157765724,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002070738459873692,
+      "loss": 1.1649,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4518757133520318,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0020703510604527095,
+      "loss": 1.1415,
+      "step": 5153
+    },
+    {
+      "epoch": 0.4519634051264063,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0020699636226967954,
+      "loss": 1.1556,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4520510969007809,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002069576146641283,
+      "loss": 1.1622,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4521387886751554,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020691886323215094,
+      "loss": 1.1854,
+      "step": 5156
+    },
+    {
+      "epoch": 0.45222648044952996,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002068801079772815,
+      "loss": 1.1665,
+      "step": 5157
+    },
+    {
+      "epoch": 0.4523141722239045,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0020684134890305425,
+      "loss": 1.2226,
+      "step": 5158
+    },
+    {
+      "epoch": 0.45240186399827903,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.00206802586013004,
+      "loss": 1.206,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4524895557726536,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0020676381931066584,
+      "loss": 1.2103,
+      "step": 5160
+    },
+    {
+      "epoch": 0.45257724754702816,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002067250487995752,
+      "loss": 1.1733,
+      "step": 5161
+    },
+    {
+      "epoch": 0.45266493932140267,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0020668627448326778,
+      "loss": 1.1935,
+      "step": 5162
+    },
+    {
+      "epoch": 0.45275263109577724,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0020664749636527973,
+      "loss": 1.1838,
+      "step": 5163
+    },
+    {
+      "epoch": 0.4528403228701518,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0020660871444914764,
+      "loss": 1.1831,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4529280146445263,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0020656992873840803,
+      "loss": 1.2171,
+      "step": 5165
+    },
+    {
+      "epoch": 0.4530157064189009,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0020653113923659837,
+      "loss": 1.2114,
+      "step": 5166
+    },
+    {
+      "epoch": 0.4531033981932754,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0020649234594725596,
+      "loss": 1.172,
+      "step": 5167
+    },
+    {
+      "epoch": 0.45319108996764995,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020645354887391873,
+      "loss": 1.1647,
+      "step": 5168
+    },
+    {
+      "epoch": 0.4532787817420245,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0020641474802012482,
+      "loss": 1.2189,
+      "step": 5169
+    },
+    {
+      "epoch": 0.453366473516399,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0020637594338941286,
+      "loss": 1.167,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4534541652907736,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002063371349853216,
+      "loss": 1.1042,
+      "step": 5171
+    },
+    {
+      "epoch": 0.45354185706514816,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002062983228113904,
+      "loss": 1.1901,
+      "step": 5172
+    },
+    {
+      "epoch": 0.45362954883952267,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0020625950687115876,
+      "loss": 1.1705,
+      "step": 5173
+    },
+    {
+      "epoch": 0.45371724061389723,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.002062206871681666,
+      "loss": 1.1499,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4538049323882718,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002061818637059542,
+      "loss": 1.1395,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4538926241626463,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002061430364880621,
+      "loss": 1.1806,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4539803159370209,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002061042055180313,
+      "loss": 1.2428,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4540680077113954,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0020606537079940306,
+      "loss": 1.177,
+      "step": 5178
+    },
+    {
+      "epoch": 0.45415569948576995,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00206026532335719,
+      "loss": 1.1747,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4542433912601445,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002059876901305211,
+      "loss": 1.1627,
+      "step": 5180
+    },
+    {
+      "epoch": 0.454331083034519,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0020594884418735173,
+      "loss": 1.1975,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4544187748088936,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0020590999450975345,
+      "loss": 1.1292,
+      "step": 5182
+    },
+    {
+      "epoch": 0.45450646658326815,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0020587114110126926,
+      "loss": 1.2401,
+      "step": 5183
+    },
+    {
+      "epoch": 0.45459415835764266,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0020583228396544254,
+      "loss": 1.1774,
+      "step": 5184
+    },
+    {
+      "epoch": 0.45468185013201723,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.002057934231058169,
+      "loss": 1.2155,
+      "step": 5185
+    },
+    {
+      "epoch": 0.4547695419063918,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002057545585259364,
+      "loss": 1.1958,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4548572336807663,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0020571569022934536,
+      "loss": 1.1075,
+      "step": 5187
+    },
+    {
+      "epoch": 0.45494492545514087,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0020567681821958843,
+      "loss": 1.1641,
+      "step": 5188
+    },
+    {
+      "epoch": 0.45503261722951543,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0020563794250021074,
+      "loss": 1.1856,
+      "step": 5189
+    },
+    {
+      "epoch": 0.45512030900388994,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0020559906307475767,
+      "loss": 1.1647,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4552080007782645,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0020556017994677474,
+      "loss": 1.186,
+      "step": 5191
+    },
+    {
+      "epoch": 0.455295692552639,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002055212931198081,
+      "loss": 1.1897,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4553833843270136,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0020548240259740417,
+      "loss": 1.1956,
+      "step": 5193
+    },
+    {
+      "epoch": 0.45547107610138815,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0020544350838310964,
+      "loss": 1.2482,
+      "step": 5194
+    },
+    {
+      "epoch": 0.45555876787576266,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002054046104804715,
+      "loss": 1.2236,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4556464596501372,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0020536570889303726,
+      "loss": 1.2121,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4557341514245118,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0020532680362435455,
+      "loss": 1.2086,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4558218431988863,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0020528789467797147,
+      "loss": 1.1654,
+      "step": 5198
+    },
+    {
+      "epoch": 0.45590953497326087,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002052489820574364,
+      "loss": 1.1572,
+      "step": 5199
+    },
+    {
+      "epoch": 0.45599722674763543,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00205210065766298,
+      "loss": 1.12,
+      "step": 5200
+    },
+    {
+      "epoch": 0.45608491852200994,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0020517114580810543,
+      "loss": 1.1034,
+      "step": 5201
+    },
+    {
+      "epoch": 0.4561726102963845,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0020513222218640813,
+      "loss": 1.203,
+      "step": 5202
+    },
+    {
+      "epoch": 0.456260302070759,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002050932949047557,
+      "loss": 1.2056,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4563479938451336,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0020505436396669827,
+      "loss": 1.2128,
+      "step": 5204
+    },
+    {
+      "epoch": 0.45643568561950815,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.002050154293757863,
+      "loss": 1.1468,
+      "step": 5205
+    },
+    {
+      "epoch": 0.45652337739388266,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002049764911355704,
+      "loss": 1.1581,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4566110691682572,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0020493754924960175,
+      "loss": 1.222,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4566987609426318,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002048986037214317,
+      "loss": 1.1168,
+      "step": 5208
+    },
+    {
+      "epoch": 0.4567864527170063,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0020485965455461197,
+      "loss": 1.1635,
+      "step": 5209
+    },
+    {
+      "epoch": 0.45687414449138086,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0020482070175269467,
+      "loss": 1.1941,
+      "step": 5210
+    },
+    {
+      "epoch": 0.45696183626575543,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0020478174531923213,
+      "loss": 1.1895,
+      "step": 5211
+    },
+    {
+      "epoch": 0.45704952804012994,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002047427852577772,
+      "loss": 1.1612,
+      "step": 5212
+    },
+    {
+      "epoch": 0.4571372198145045,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0020470382157188275,
+      "loss": 1.1698,
+      "step": 5213
+    },
+    {
+      "epoch": 0.457224911588879,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0020466485426510234,
+      "loss": 1.1466,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4573126033632536,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002046258833409896,
+      "loss": 1.1319,
+      "step": 5215
+    },
+    {
+      "epoch": 0.45740029513762814,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0020458690880309854,
+      "loss": 1.2753,
+      "step": 5216
+    },
+    {
+      "epoch": 0.45748798691200265,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002045479306549836,
+      "loss": 1.2515,
+      "step": 5217
+    },
+    {
+      "epoch": 0.4575756786863772,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002045089489001995,
+      "loss": 1.242,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4576633704607518,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002044699635423013,
+      "loss": 1.1769,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4577510622351263,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0020443097458484427,
+      "loss": 1.2065,
+      "step": 5220
+    },
+    {
+      "epoch": 0.45783875400950086,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0020439198203138415,
+      "loss": 1.1994,
+      "step": 5221
+    },
+    {
+      "epoch": 0.4579264457838754,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0020435298588547695,
+      "loss": 1.1597,
+      "step": 5222
+    },
+    {
+      "epoch": 0.45801413755824993,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0020431398615067906,
+      "loss": 1.1836,
+      "step": 5223
+    },
+    {
+      "epoch": 0.4581018293326245,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0020427498283054706,
+      "loss": 1.1805,
+      "step": 5224
+    },
+    {
+      "epoch": 0.458189521106999,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002042359759286381,
+      "loss": 1.1553,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4582772128813736,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0020419696544850935,
+      "loss": 1.2174,
+      "step": 5226
+    },
+    {
+      "epoch": 0.45836490465574814,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0020415795139371857,
+      "loss": 1.1505,
+      "step": 5227
+    },
+    {
+      "epoch": 0.45845259643012265,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0020411893376782366,
+      "loss": 1.149,
+      "step": 5228
+    },
+    {
+      "epoch": 0.4585402882044972,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00204079912574383,
+      "loss": 1.1881,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4586279799788718,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0020404088781695523,
+      "loss": 1.2628,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4587156717532463,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0020400185949909925,
+      "loss": 1.2306,
+      "step": 5231
+    },
+    {
+      "epoch": 0.45880336352762086,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002039628276243744,
+      "loss": 1.1691,
+      "step": 5232
+    },
+    {
+      "epoch": 0.4588910553019954,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002039237921963402,
+      "loss": 1.1744,
+      "step": 5233
+    },
+    {
+      "epoch": 0.45897874707636993,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0020388475321855668,
+      "loss": 1.2127,
+      "step": 5234
+    },
+    {
+      "epoch": 0.4590664388507445,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0020384571069458402,
+      "loss": 1.1825,
+      "step": 5235
+    },
+    {
+      "epoch": 0.45915413062511906,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002038066646279828,
+      "loss": 1.185,
+      "step": 5236
+    },
+    {
+      "epoch": 0.45924182239949357,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00203767615022314,
+      "loss": 1.1696,
+      "step": 5237
+    },
+    {
+      "epoch": 0.45932951417386814,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0020372856188113878,
+      "loss": 1.1827,
+      "step": 5238
+    },
+    {
+      "epoch": 0.45941720594824265,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0020368950520801865,
+      "loss": 1.1566,
+      "step": 5239
+    },
+    {
+      "epoch": 0.4595048977226172,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0020365044500651555,
+      "loss": 1.1677,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4595925894969918,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0020361138128019165,
+      "loss": 1.1707,
+      "step": 5241
+    },
+    {
+      "epoch": 0.4596802812713663,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002035723140326094,
+      "loss": 1.1899,
+      "step": 5242
+    },
+    {
+      "epoch": 0.45976797304574085,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002035332432673318,
+      "loss": 1.1317,
+      "step": 5243
+    },
+    {
+      "epoch": 0.4598556648201154,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0020349416898792177,
+      "loss": 1.257,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4599433565944899,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0020345509119794295,
+      "loss": 1.1831,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4600310483688645,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0020341600990095906,
+      "loss": 1.2278,
+      "step": 5246
+    },
+    {
+      "epoch": 0.46011874014323906,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0020337692510053428,
+      "loss": 1.1497,
+      "step": 5247
+    },
+    {
+      "epoch": 0.46020643191761357,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0020333783680023293,
+      "loss": 1.1507,
+      "step": 5248
+    },
+    {
+      "epoch": 0.46029412369198813,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0020329874500361993,
+      "loss": 1.2228,
+      "step": 5249
+    },
+    {
+      "epoch": 0.46038181546636264,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002032596497142602,
+      "loss": 1.1897,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4604695072407372,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002032205509357192,
+      "loss": 1.1816,
+      "step": 5251
+    },
+    {
+      "epoch": 0.4605571990151118,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002031814486715626,
+      "loss": 1.171,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4606448907894863,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002031423429253564,
+      "loss": 1.2107,
+      "step": 5253
+    },
+    {
+      "epoch": 0.46073258256386085,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0020310323370066707,
+      "loss": 1.1908,
+      "step": 5254
+    },
+    {
+      "epoch": 0.4608202743382354,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020306412100106115,
+      "loss": 1.1246,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4609079661126099,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0020302500483010567,
+      "loss": 1.1751,
+      "step": 5256
+    },
+    {
+      "epoch": 0.4609956578869845,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.002029858851913679,
+      "loss": 1.1497,
+      "step": 5257
+    },
+    {
+      "epoch": 0.46108334966135905,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002029467620884155,
+      "loss": 1.1777,
+      "step": 5258
+    },
+    {
+      "epoch": 0.46117104143573356,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002029076355248163,
+      "loss": 1.2054,
+      "step": 5259
+    },
+    {
+      "epoch": 0.46125873321010813,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002028685055041386,
+      "loss": 1.1278,
+      "step": 5260
+    },
+    {
+      "epoch": 0.46134642498448264,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0020282937202995097,
+      "loss": 1.1275,
+      "step": 5261
+    },
+    {
+      "epoch": 0.4614341167588572,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.002027902351058223,
+      "loss": 1.2013,
+      "step": 5262
+    },
+    {
+      "epoch": 0.46152180853323177,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0020275109473532173,
+      "loss": 1.1834,
+      "step": 5263
+    },
+    {
+      "epoch": 0.4616095003076063,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0020271195092201873,
+      "loss": 1.1776,
+      "step": 5264
+    },
+    {
+      "epoch": 0.46169719208198085,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0020267280366948323,
+      "loss": 1.2324,
+      "step": 5265
+    },
+    {
+      "epoch": 0.4617848838563554,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002026336529812852,
+      "loss": 1.1748,
+      "step": 5266
+    },
+    {
+      "epoch": 0.4618725756307299,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0020259449886099526,
+      "loss": 1.1385,
+      "step": 5267
+    },
+    {
+      "epoch": 0.4619602674051045,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0020255534131218404,
+      "loss": 1.1508,
+      "step": 5268
+    },
+    {
+      "epoch": 0.46204795917947905,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002025161803384226,
+      "loss": 1.159,
+      "step": 5269
+    },
+    {
+      "epoch": 0.46213565095385356,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002024770159432824,
+      "loss": 1.1769,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4622233427282281,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020243784813033503,
+      "loss": 1.2052,
+      "step": 5271
+    },
+    {
+      "epoch": 0.4623110345026027,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0020239867690315266,
+      "loss": 1.111,
+      "step": 5272
+    },
+    {
+      "epoch": 0.4623987262769772,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.002023595022653074,
+      "loss": 1.1928,
+      "step": 5273
+    },
+    {
+      "epoch": 0.46248641805135177,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002023203242203721,
+      "loss": 1.1503,
+      "step": 5274
+    },
+    {
+      "epoch": 0.4625741098257263,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0020228114277191945,
+      "loss": 1.2033,
+      "step": 5275
+    },
+    {
+      "epoch": 0.46266180160010084,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002022419579235229,
+      "loss": 1.2272,
+      "step": 5276
+    },
+    {
+      "epoch": 0.4627494933744754,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0020220276967875587,
+      "loss": 1.1007,
+      "step": 5277
+    },
+    {
+      "epoch": 0.4628371851488499,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0020216357804119234,
+      "loss": 1.1784,
+      "step": 5278
+    },
+    {
+      "epoch": 0.4629248769232245,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0020212438301440636,
+      "loss": 1.2131,
+      "step": 5279
+    },
+    {
+      "epoch": 0.46301256869759905,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.002020851846019726,
+      "loss": 1.1642,
+      "step": 5280
+    },
+    {
+      "epoch": 0.46310026047197356,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002020459828074656,
+      "loss": 1.1888,
+      "step": 5281
+    },
+    {
+      "epoch": 0.4631879522463481,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.002020067776344606,
+      "loss": 1.0788,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4632756440207227,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0020196756908653314,
+      "loss": 1.2003,
+      "step": 5283
+    },
+    {
+      "epoch": 0.4633633357950972,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0020192835716725875,
+      "loss": 1.2512,
+      "step": 5284
+    },
+    {
+      "epoch": 0.46345102756947176,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0020188914188021354,
+      "loss": 1.1943,
+      "step": 5285
+    },
+    {
+      "epoch": 0.4635387193438463,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0020184992322897378,
+      "loss": 1.1704,
+      "step": 5286
+    },
+    {
+      "epoch": 0.46362641111822084,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0020181070121711615,
+      "loss": 1.1282,
+      "step": 5287
+    },
+    {
+      "epoch": 0.4637141028925954,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0020177147584821763,
+      "loss": 1.1656,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4638017946669699,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.002017322471258554,
+      "loss": 1.1281,
+      "step": 5289
+    },
+    {
+      "epoch": 0.4638894864413445,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0020169301505360707,
+      "loss": 1.1506,
+      "step": 5290
+    },
+    {
+      "epoch": 0.46397717821571904,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0020165377963505047,
+      "loss": 1.2141,
+      "step": 5291
+    },
+    {
+      "epoch": 0.46406486999009355,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002016145408737639,
+      "loss": 1.1921,
+      "step": 5292
+    },
+    {
+      "epoch": 0.4641525617644681,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002015752987733256,
+      "loss": 1.2243,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4642402535388427,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002015360533373145,
+      "loss": 1.2605,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4643279453132172,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0020149680456930962,
+      "loss": 1.197,
+      "step": 5295
+    },
+    {
+      "epoch": 0.46441563708759176,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002014575524728904,
+      "loss": 1.1938,
+      "step": 5296
+    },
+    {
+      "epoch": 0.46450332886196627,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0020141829705163654,
+      "loss": 1.1694,
+      "step": 5297
+    },
+    {
+      "epoch": 0.46459102063634083,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00201379038309128,
+      "loss": 1.1704,
+      "step": 5298
+    },
+    {
+      "epoch": 0.4646787124107154,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00201339776248945,
+      "loss": 1.1751,
+      "step": 5299
+    },
+    {
+      "epoch": 0.4647664041850899,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0020130051087466835,
+      "loss": 1.1607,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4648540959594645,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0020126124218987876,
+      "loss": 1.1614,
+      "step": 5301
+    },
+    {
+      "epoch": 0.46494178773383904,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002012219701981575,
+      "loss": 1.1535,
+      "step": 5302
+    },
+    {
+      "epoch": 0.46502947950821355,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00201182694903086,
+      "loss": 1.1149,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4651171712825881,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0020114341630824615,
+      "loss": 1.1813,
+      "step": 5304
+    },
+    {
+      "epoch": 0.4652048630569627,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002011041344172201,
+      "loss": 1.1763,
+      "step": 5305
+    },
+    {
+      "epoch": 0.4652925548313372,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002010648492335901,
+      "loss": 1.1094,
+      "step": 5306
+    },
+    {
+      "epoch": 0.46538024660571176,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0020102556076093906,
+      "loss": 1.1906,
+      "step": 5307
+    },
+    {
+      "epoch": 0.4654679383800863,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.002009862690028498,
+      "loss": 1.218,
+      "step": 5308
+    },
+    {
+      "epoch": 0.46555563015446083,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0020094697396290575,
+      "loss": 1.2312,
+      "step": 5309
+    },
+    {
+      "epoch": 0.4656433219288354,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0020090767564469045,
+      "loss": 1.1815,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4657310137032099,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.002008683740517878,
+      "loss": 1.2485,
+      "step": 5311
+    },
+    {
+      "epoch": 0.46581870547758447,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.002008290691877821,
+      "loss": 1.2737,
+      "step": 5312
+    },
+    {
+      "epoch": 0.46590639725195904,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0020078976105625773,
+      "loss": 1.1637,
+      "step": 5313
+    },
+    {
+      "epoch": 0.46599408902633355,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0020075044966079954,
+      "loss": 1.1778,
+      "step": 5314
+    },
+    {
+      "epoch": 0.4660817808007081,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0020071113500499263,
+      "loss": 1.1683,
+      "step": 5315
+    },
+    {
+      "epoch": 0.4661694725750827,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.002006718170924224,
+      "loss": 1.1503,
+      "step": 5316
+    },
+    {
+      "epoch": 0.4662571643494572,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0020063249592667458,
+      "loss": 1.2301,
+      "step": 5317
+    },
+    {
+      "epoch": 0.46634485612383175,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0020059317151133508,
+      "loss": 1.1232,
+      "step": 5318
+    },
+    {
+      "epoch": 0.4664325478982063,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002005538438499902,
+      "loss": 1.2855,
+      "step": 5319
+    },
+    {
+      "epoch": 0.4665202396725808,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002005145129462266,
+      "loss": 1.1561,
+      "step": 5320
+    },
+    {
+      "epoch": 0.4666079314469554,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0020047517880363106,
+      "loss": 1.1757,
+      "step": 5321
+    },
+    {
+      "epoch": 0.4666956232213299,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002004358414257908,
+      "loss": 1.1372,
+      "step": 5322
+    },
+    {
+      "epoch": 0.46678331499570447,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002003965008162932,
+      "loss": 1.2237,
+      "step": 5323
+    },
+    {
+      "epoch": 0.46687100677007903,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002003571569787262,
+      "loss": 1.1755,
+      "step": 5324
+    },
+    {
+      "epoch": 0.46695869854445354,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0020031780991667776,
+      "loss": 1.1883,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4670463903188281,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002002784596337362,
+      "loss": 1.2047,
+      "step": 5326
+    },
+    {
+      "epoch": 0.4671340820932027,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002002391061334902,
+      "loss": 1.2028,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4672217738675772,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002001997494195287,
+      "loss": 1.1238,
+      "step": 5328
+    },
+    {
+      "epoch": 0.46730946564195175,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002001603894954409,
+      "loss": 1.1169,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4673971574163263,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.002001210263648163,
+      "loss": 1.1403,
+      "step": 5330
+    },
+    {
+      "epoch": 0.4674848491907008,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0020008166003124485,
+      "loss": 1.1873,
+      "step": 5331
+    },
+    {
+      "epoch": 0.4675725409650754,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002000422904983165,
+      "loss": 1.1732,
+      "step": 5332
+    },
+    {
+      "epoch": 0.4676602327394499,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.002000029177696218,
+      "loss": 1.1879,
+      "step": 5333
+    },
+    {
+      "epoch": 0.46774792451382446,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001999635418487513,
+      "loss": 1.2192,
+      "step": 5334
+    },
+    {
+      "epoch": 0.46783561628819903,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.00199924162739296,
+      "loss": 1.1529,
+      "step": 5335
+    },
+    {
+      "epoch": 0.46792330806257354,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001998847804448473,
+      "loss": 1.1675,
+      "step": 5336
+    },
+    {
+      "epoch": 0.4680109998369481,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001998453949689966,
+      "loss": 1.251,
+      "step": 5337
+    },
+    {
+      "epoch": 0.46809869161132267,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0019980600631533587,
+      "loss": 1.1895,
+      "step": 5338
+    },
+    {
+      "epoch": 0.4681863833856972,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.001997666144874572,
+      "loss": 1.1433,
+      "step": 5339
+    },
+    {
+      "epoch": 0.46827407516007175,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0019972721948895308,
+      "loss": 1.1403,
+      "step": 5340
+    },
+    {
+      "epoch": 0.4683617669344463,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0019968782132341613,
+      "loss": 1.2059,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4684494587088208,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001996484199944395,
+      "loss": 1.1731,
+      "step": 5342
+    },
+    {
+      "epoch": 0.4685371504831954,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0019960901550561635,
+      "loss": 1.1422,
+      "step": 5343
+    },
+    {
+      "epoch": 0.46862484225756995,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.001995696078605403,
+      "loss": 1.185,
+      "step": 5344
+    },
+    {
+      "epoch": 0.46871253403194446,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0019953019706280537,
+      "loss": 1.1869,
+      "step": 5345
+    },
+    {
+      "epoch": 0.468800225806319,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.001994907831160055,
+      "loss": 1.1721,
+      "step": 5346
+    },
+    {
+      "epoch": 0.46888791758069354,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0019945136602373533,
+      "loss": 1.2218,
+      "step": 5347
+    },
+    {
+      "epoch": 0.4689756093550681,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.001994119457895894,
+      "loss": 1.1583,
+      "step": 5348
+    },
+    {
+      "epoch": 0.46906330112944267,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001993725224171629,
+      "loss": 1.1574,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4691509929038172,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001993330959100511,
+      "loss": 1.1463,
+      "step": 5350
+    },
+    {
+      "epoch": 0.46923868467819174,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0019929366627184967,
+      "loss": 1.2273,
+      "step": 5351
+    },
+    {
+      "epoch": 0.4693263764525663,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0019925423350615427,
+      "loss": 1.2788,
+      "step": 5352
+    },
+    {
+      "epoch": 0.4694140682269408,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0019921479761656124,
+      "loss": 1.1574,
+      "step": 5353
+    },
+    {
+      "epoch": 0.4695017600013154,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00199175358606667,
+      "loss": 1.1618,
+      "step": 5354
+    },
+    {
+      "epoch": 0.46958945177568995,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.001991359164800683,
+      "loss": 1.1096,
+      "step": 5355
+    },
+    {
+      "epoch": 0.46967714355006446,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0019909647124036214,
+      "loss": 1.2371,
+      "step": 5356
+    },
+    {
+      "epoch": 0.469764835324439,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0019905702289114576,
+      "loss": 1.1693,
+      "step": 5357
+    },
+    {
+      "epoch": 0.46985252709881353,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0019901757143601685,
+      "loss": 1.2045,
+      "step": 5358
+    },
+    {
+      "epoch": 0.4699402188731881,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0019897811687857327,
+      "loss": 1.2171,
+      "step": 5359
+    },
+    {
+      "epoch": 0.47002791064756266,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0019893865922241314,
+      "loss": 1.2155,
+      "step": 5360
+    },
+    {
+      "epoch": 0.4701156024219372,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0019889919847113483,
+      "loss": 1.1602,
+      "step": 5361
+    },
+    {
+      "epoch": 0.47020329419631174,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001988597346283372,
+      "loss": 1.1818,
+      "step": 5362
+    },
+    {
+      "epoch": 0.4702909859706863,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001988202676976192,
+      "loss": 1.1876,
+      "step": 5363
+    },
+    {
+      "epoch": 0.4703786777450608,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0019878079768258,
+      "loss": 1.2834,
+      "step": 5364
+    },
+    {
+      "epoch": 0.4704663695194354,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0019874132458681934,
+      "loss": 1.2145,
+      "step": 5365
+    },
+    {
+      "epoch": 0.47055406129380994,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0019870184841393693,
+      "loss": 1.1583,
+      "step": 5366
+    },
+    {
+      "epoch": 0.47064175306818445,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0019866236916753297,
+      "loss": 1.1915,
+      "step": 5367
+    },
+    {
+      "epoch": 0.470729444842559,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0019862288685120787,
+      "loss": 1.1231,
+      "step": 5368
+    },
+    {
+      "epoch": 0.47081713661693353,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0019858340146856226,
+      "loss": 1.1834,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4709048283913081,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001985439130231971,
+      "loss": 1.1888,
+      "step": 5370
+    },
+    {
+      "epoch": 0.47099252016568266,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001985044215187137,
+      "loss": 1.1398,
+      "step": 5371
+    },
+    {
+      "epoch": 0.47108021194005717,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001984649269587135,
+      "loss": 1.1932,
+      "step": 5372
+    },
+    {
+      "epoch": 0.47116790371443174,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0019842542934679843,
+      "loss": 1.2102,
+      "step": 5373
+    },
+    {
+      "epoch": 0.4712555954888063,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001983859286865704,
+      "loss": 1.1804,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4713432872631808,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0019834642498163194,
+      "loss": 1.1624,
+      "step": 5375
+    },
+    {
+      "epoch": 0.4714309790375554,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0019830691823558557,
+      "loss": 1.1903,
+      "step": 5376
+    },
+    {
+      "epoch": 0.47151867081192994,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0019826740845203424,
+      "loss": 1.171,
+      "step": 5377
+    },
+    {
+      "epoch": 0.47160636258630445,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.001982278956345811,
+      "loss": 1.1894,
+      "step": 5378
+    },
+    {
+      "epoch": 0.471694054360679,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0019818837978682973,
+      "loss": 1.169,
+      "step": 5379
+    },
+    {
+      "epoch": 0.4717817461350536,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0019814886091238377,
+      "loss": 1.2408,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4718694379094281,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0019810933901484723,
+      "loss": 1.1323,
+      "step": 5381
+    },
+    {
+      "epoch": 0.47195712968380266,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001980698140978245,
+      "loss": 1.1996,
+      "step": 5382
+    },
+    {
+      "epoch": 0.47204482145817717,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0019803028616492005,
+      "loss": 1.1505,
+      "step": 5383
+    },
+    {
+      "epoch": 0.47213251323255173,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001979907552197388,
+      "loss": 1.16,
+      "step": 5384
+    },
+    {
+      "epoch": 0.4722202050069263,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0019795122126588585,
+      "loss": 1.1524,
+      "step": 5385
+    },
+    {
+      "epoch": 0.4723078967813008,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0019791168430696652,
+      "loss": 1.2051,
+      "step": 5386
+    },
+    {
+      "epoch": 0.47239558855567537,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.001978721443465866,
+      "loss": 1.1658,
+      "step": 5387
+    },
+    {
+      "epoch": 0.47248328033004994,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0019783260138835196,
+      "loss": 1.2829,
+      "step": 5388
+    },
+    {
+      "epoch": 0.47257097210442445,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001977930554358688,
+      "loss": 1.1749,
+      "step": 5389
+    },
+    {
+      "epoch": 0.472658663878799,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0019775350649274366,
+      "loss": 1.1354,
+      "step": 5390
+    },
+    {
+      "epoch": 0.4727463556531736,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0019771395456258333,
+      "loss": 1.1723,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4728340474275481,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0019767439964899474,
+      "loss": 1.1718,
+      "step": 5392
+    },
+    {
+      "epoch": 0.47292173920192265,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001976348417555853,
+      "loss": 1.194,
+      "step": 5393
+    },
+    {
+      "epoch": 0.47300943097629716,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.001975952808859625,
+      "loss": 1.1772,
+      "step": 5394
+    },
+    {
+      "epoch": 0.47309712275067173,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0019755571704373424,
+      "loss": 1.1829,
+      "step": 5395
+    },
+    {
+      "epoch": 0.4731848145250463,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0019751615023250856,
+      "loss": 1.202,
+      "step": 5396
+    },
+    {
+      "epoch": 0.4732725062994208,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0019747658045589397,
+      "loss": 1.1656,
+      "step": 5397
+    },
+    {
+      "epoch": 0.47336019807379537,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0019743700771749913,
+      "loss": 1.1332,
+      "step": 5398
+    },
+    {
+      "epoch": 0.47344788984816993,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0019739743202093285,
+      "loss": 1.1497,
+      "step": 5399
+    },
+    {
+      "epoch": 0.47353558162254444,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0019735785336980447,
+      "loss": 1.1316,
+      "step": 5400
+    },
+    {
+      "epoch": 0.473623273396919,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001973182717677233,
+      "loss": 1.1576,
+      "step": 5401
+    },
+    {
+      "epoch": 0.4737109651712936,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0019727868721829923,
+      "loss": 1.1987,
+      "step": 5402
+    },
+    {
+      "epoch": 0.4737986569456681,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001972390997251422,
+      "loss": 1.2304,
+      "step": 5403
+    },
+    {
+      "epoch": 0.47388634872004265,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0019719950929186253,
+      "loss": 1.2244,
+      "step": 5404
+    },
+    {
+      "epoch": 0.47397404049441716,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001971599159220707,
+      "loss": 1.1542,
+      "step": 5405
+    },
+    {
+      "epoch": 0.4740617322687917,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0019712031961937756,
+      "loss": 1.1553,
+      "step": 5406
+    },
+    {
+      "epoch": 0.4741494240431663,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0019708072038739422,
+      "loss": 1.1995,
+      "step": 5407
+    },
+    {
+      "epoch": 0.4742371158175408,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00197041118229732,
+      "loss": 1.2107,
+      "step": 5408
+    },
+    {
+      "epoch": 0.47432480759191537,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0019700151315000247,
+      "loss": 1.1998,
+      "step": 5409
+    },
+    {
+      "epoch": 0.47441249936628993,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001969619051518176,
+      "loss": 1.1977,
+      "step": 5410
+    },
+    {
+      "epoch": 0.47450019114066444,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001969222942387895,
+      "loss": 1.1719,
+      "step": 5411
+    },
+    {
+      "epoch": 0.474587882915039,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001968826804145305,
+      "loss": 1.1768,
+      "step": 5412
+    },
+    {
+      "epoch": 0.47467557468941357,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001968430636826534,
+      "loss": 1.1132,
+      "step": 5413
+    },
+    {
+      "epoch": 0.4747632664637881,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0019680344404677105,
+      "loss": 1.1454,
+      "step": 5414
+    },
+    {
+      "epoch": 0.47485095823816265,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0019676382151049675,
+      "loss": 1.2117,
+      "step": 5415
+    },
+    {
+      "epoch": 0.47493865001253716,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001967241960774439,
+      "loss": 1.247,
+      "step": 5416
+    },
+    {
+      "epoch": 0.4750263417869117,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.001966845677512263,
+      "loss": 1.1018,
+      "step": 5417
+    },
+    {
+      "epoch": 0.4751140335612863,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001966449365354579,
+      "loss": 1.1767,
+      "step": 5418
+    },
+    {
+      "epoch": 0.4752017253356608,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0019660530243375295,
+      "loss": 1.165,
+      "step": 5419
+    },
+    {
+      "epoch": 0.47528941711003536,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0019656566544972603,
+      "loss": 1.1854,
+      "step": 5420
+    },
+    {
+      "epoch": 0.4753771088844099,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001965260255869919,
+      "loss": 1.1484,
+      "step": 5421
+    },
+    {
+      "epoch": 0.47546480065878444,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001964863828491656,
+      "loss": 1.1347,
+      "step": 5422
+    },
+    {
+      "epoch": 0.475552492433159,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0019644673723986246,
+      "loss": 1.1231,
+      "step": 5423
+    },
+    {
+      "epoch": 0.47564018420753357,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001964070887626981,
+      "loss": 1.1108,
+      "step": 5424
+    },
+    {
+      "epoch": 0.4757278759819081,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0019636743742128832,
+      "loss": 1.1696,
+      "step": 5425
+    },
+    {
+      "epoch": 0.47581556775628264,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0019632778321924923,
+      "loss": 1.1713,
+      "step": 5426
+    },
+    {
+      "epoch": 0.4759032595306572,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001962881261601972,
+      "loss": 1.1142,
+      "step": 5427
+    },
+    {
+      "epoch": 0.4759909513050317,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0019624846624774874,
+      "loss": 1.1895,
+      "step": 5428
+    },
+    {
+      "epoch": 0.4760786430794063,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0019620880348552093,
+      "loss": 1.1786,
+      "step": 5429
+    },
+    {
+      "epoch": 0.4761663348537808,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0019616913787713075,
+      "loss": 1.1989,
+      "step": 5430
+    },
+    {
+      "epoch": 0.47625402662815536,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001961294694261957,
+      "loss": 1.2001,
+      "step": 5431
+    },
+    {
+      "epoch": 0.4763417184025299,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.001960897981363333,
+      "loss": 1.1706,
+      "step": 5432
+    },
+    {
+      "epoch": 0.47642941017690443,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0019605012401116167,
+      "loss": 1.2116,
+      "step": 5433
+    },
+    {
+      "epoch": 0.476517101951279,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019601044705429885,
+      "loss": 1.253,
+      "step": 5434
+    },
+    {
+      "epoch": 0.47660479372565356,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001959707672693633,
+      "loss": 1.2066,
+      "step": 5435
+    },
+    {
+      "epoch": 0.4766924855000281,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001959310846599738,
+      "loss": 1.2256,
+      "step": 5436
+    },
+    {
+      "epoch": 0.47678017727440264,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0019589139922974916,
+      "loss": 1.2722,
+      "step": 5437
+    },
+    {
+      "epoch": 0.4768678690487772,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0019585171098230867,
+      "loss": 1.1843,
+      "step": 5438
+    },
+    {
+      "epoch": 0.4769555608231517,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0019581201992127173,
+      "loss": 1.1852,
+      "step": 5439
+    },
+    {
+      "epoch": 0.4770432525975263,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001957723260502582,
+      "loss": 1.1629,
+      "step": 5440
+    },
+    {
+      "epoch": 0.4771309443719008,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001957326293728879,
+      "loss": 1.1815,
+      "step": 5441
+    },
+    {
+      "epoch": 0.47721863614627535,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001956929298927812,
+      "loss": 1.2097,
+      "step": 5442
+    },
+    {
+      "epoch": 0.4773063279206499,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0019565322761355844,
+      "loss": 1.1467,
+      "step": 5443
+    },
+    {
+      "epoch": 0.47739401969502443,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0019561352253884058,
+      "loss": 1.1584,
+      "step": 5444
+    },
+    {
+      "epoch": 0.477481711469399,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001955738146722484,
+      "loss": 1.1315,
+      "step": 5445
+    },
+    {
+      "epoch": 0.47756940324377356,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0019553410401740323,
+      "loss": 1.1802,
+      "step": 5446
+    },
+    {
+      "epoch": 0.47765709501814807,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0019549439057792665,
+      "loss": 1.1644,
+      "step": 5447
+    },
+    {
+      "epoch": 0.47774478679252264,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0019545467435744036,
+      "loss": 1.138,
+      "step": 5448
+    },
+    {
+      "epoch": 0.4778324785668972,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0019541495535956637,
+      "loss": 1.1394,
+      "step": 5449
+    },
+    {
+      "epoch": 0.4779201703412717,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0019537523358792697,
+      "loss": 1.1708,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4780078621156463,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0019533550904614473,
+      "loss": 1.1702,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4780955538900208,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0019529578173784234,
+      "loss": 1.1644,
+      "step": 5452
+    },
+    {
+      "epoch": 0.47818324566439535,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0019525605166664285,
+      "loss": 1.1508,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4782709374387699,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0019521631883616956,
+      "loss": 1.1322,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4783586292131444,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0019517658325004603,
+      "loss": 1.1928,
+      "step": 5455
+    },
+    {
+      "epoch": 0.478446320987519,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0019513684491189595,
+      "loss": 1.1934,
+      "step": 5456
+    },
+    {
+      "epoch": 0.47853401276189356,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0019509710382534346,
+      "loss": 1.1654,
+      "step": 5457
+    },
+    {
+      "epoch": 0.47862170453626807,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0019505735999401275,
+      "loss": 1.1728,
+      "step": 5458
+    },
+    {
+      "epoch": 0.47870939631064263,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0019501761342152844,
+      "loss": 1.1689,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4787970880850172,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0019497786411151524,
+      "loss": 1.2006,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4788847798593917,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0019493811206759827,
+      "loss": 1.1617,
+      "step": 5461
+    },
+    {
+      "epoch": 0.4789724716337663,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0019489835729340273,
+      "loss": 1.2235,
+      "step": 5462
+    },
+    {
+      "epoch": 0.47906016340814084,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.001948585997925542,
+      "loss": 1.1347,
+      "step": 5463
+    },
+    {
+      "epoch": 0.47914785518251535,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0019481883956867846,
+      "loss": 1.1704,
+      "step": 5464
+    },
+    {
+      "epoch": 0.4792355469568899,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001947790766254015,
+      "loss": 1.1714,
+      "step": 5465
+    },
+    {
+      "epoch": 0.4793232387312644,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0019473931096634963,
+      "loss": 1.2116,
+      "step": 5466
+    },
+    {
+      "epoch": 0.479410930505639,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0019469954259514939,
+      "loss": 1.2081,
+      "step": 5467
+    },
+    {
+      "epoch": 0.47949862228001355,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0019465977151542752,
+      "loss": 1.1785,
+      "step": 5468
+    },
+    {
+      "epoch": 0.47958631405438806,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0019461999773081108,
+      "loss": 1.1121,
+      "step": 5469
+    },
+    {
+      "epoch": 0.47967400582876263,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0019458022124492727,
+      "loss": 1.1052,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4797616976031372,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0019454044206140368,
+      "loss": 1.1249,
+      "step": 5471
+    },
+    {
+      "epoch": 0.4798493893775117,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0019450066018386798,
+      "loss": 1.2033,
+      "step": 5472
+    },
+    {
+      "epoch": 0.47993708115188627,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001944608756159483,
+      "loss": 1.118,
+      "step": 5473
+    },
+    {
+      "epoch": 0.48002477292626083,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0019442108836127275,
+      "loss": 1.189,
+      "step": 5474
+    },
+    {
+      "epoch": 0.48011246470063534,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0019438129842346992,
+      "loss": 1.1394,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4802001564750099,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001943415058061685,
+      "loss": 1.1106,
+      "step": 5476
+    },
+    {
+      "epoch": 0.4802878482493844,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0019430171051299753,
+      "loss": 1.1651,
+      "step": 5477
+    },
+    {
+      "epoch": 0.480375540023759,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.001942619125475862,
+      "loss": 1.1356,
+      "step": 5478
+    },
+    {
+      "epoch": 0.48046323179813355,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0019422211191356396,
+      "loss": 1.1435,
+      "step": 5479
+    },
+    {
+      "epoch": 0.48055092357250806,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0019418230861456055,
+      "loss": 1.1986,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4806386153468826,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0019414250265420594,
+      "loss": 1.1725,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4807263071212572,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0019410269403613031,
+      "loss": 1.1358,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4808139988956317,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019406288276396415,
+      "loss": 1.1441,
+      "step": 5483
+    },
+    {
+      "epoch": 0.48090169067000627,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0019402306884133806,
+      "loss": 1.1403,
+      "step": 5484
+    },
+    {
+      "epoch": 0.48098938244438083,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0019398325227188305,
+      "loss": 1.1858,
+      "step": 5485
+    },
+    {
+      "epoch": 0.48107707421875534,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0019394343305923025,
+      "loss": 1.2389,
+      "step": 5486
+    },
+    {
+      "epoch": 0.4811647659931299,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0019390361120701105,
+      "loss": 1.212,
+      "step": 5487
+    },
+    {
+      "epoch": 0.4812524577675044,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0019386378671885712,
+      "loss": 1.1715,
+      "step": 5488
+    },
+    {
+      "epoch": 0.481340149541879,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001938239595984004,
+      "loss": 1.1256,
+      "step": 5489
+    },
+    {
+      "epoch": 0.48142784131625355,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001937841298492729,
+      "loss": 1.1571,
+      "step": 5490
+    },
+    {
+      "epoch": 0.48151553309062806,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0019374429747510712,
+      "loss": 1.1195,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4816032248650026,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0019370446247953558,
+      "loss": 1.132,
+      "step": 5492
+    },
+    {
+      "epoch": 0.4816909166393772,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0019366462486619118,
+      "loss": 1.1541,
+      "step": 5493
+    },
+    {
+      "epoch": 0.4817786084137517,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0019362478463870697,
+      "loss": 1.2695,
+      "step": 5494
+    },
+    {
+      "epoch": 0.48186630018812626,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0019358494180071635,
+      "loss": 1.1426,
+      "step": 5495
+    },
+    {
+      "epoch": 0.4819539919625008,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0019354509635585277,
+      "loss": 1.2493,
+      "step": 5496
+    },
+    {
+      "epoch": 0.48204168373687534,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0019350524830775014,
+      "loss": 1.2335,
+      "step": 5497
+    },
+    {
+      "epoch": 0.4821293755112499,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0019346539766004242,
+      "loss": 1.1303,
+      "step": 5498
+    },
+    {
+      "epoch": 0.48221706728562447,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0019342554441636388,
+      "loss": 1.1783,
+      "step": 5499
+    },
+    {
+      "epoch": 0.482304759059999,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.001933856885803491,
+      "loss": 1.2055,
+      "step": 5500
+    },
+    {
+      "epoch": 0.482304759059999,
+      "eval_loss": 1.1834715604782104,
+      "eval_runtime": 428.3748,
+      "eval_samples_per_second": 33.725,
+      "eval_steps_per_second": 8.432,
+      "step": 5500
+    },
+    {
+      "epoch": 0.48239245083437354,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001933458301556328,
+      "loss": 1.1604,
+      "step": 5501
+    },
+    {
+      "epoch": 0.48248014260874805,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0019330596914585,
+      "loss": 1.1278,
+      "step": 5502
+    },
+    {
+      "epoch": 0.4825678343831226,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0019326610555463588,
+      "loss": 1.133,
+      "step": 5503
+    },
+    {
+      "epoch": 0.4826555261574972,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001932262393856259,
+      "loss": 1.1856,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4827432179318717,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0019318637064245571,
+      "loss": 1.1695,
+      "step": 5505
+    },
+    {
+      "epoch": 0.48283090970624626,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.001931464993287613,
+      "loss": 1.1824,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4829186014806208,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001931066254481788,
+      "loss": 1.1753,
+      "step": 5507
+    },
+    {
+      "epoch": 0.48300629325499533,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0019306674900434464,
+      "loss": 1.1074,
+      "step": 5508
+    },
+    {
+      "epoch": 0.4830939850293699,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.001930268700008954,
+      "loss": 1.1815,
+      "step": 5509
+    },
+    {
+      "epoch": 0.48318167680374446,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0019298698844146798,
+      "loss": 1.1893,
+      "step": 5510
+    },
+    {
+      "epoch": 0.483269368578119,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019294710432969946,
+      "loss": 1.1488,
+      "step": 5511
+    },
+    {
+      "epoch": 0.48335706035249354,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001929072176692272,
+      "loss": 1.1032,
+      "step": 5512
+    },
+    {
+      "epoch": 0.48344475212686805,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001928673284636887,
+      "loss": 1.2194,
+      "step": 5513
+    },
+    {
+      "epoch": 0.4835324439012426,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0019282743671672174,
+      "loss": 1.1615,
+      "step": 5514
+    },
+    {
+      "epoch": 0.4836201356756172,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0019278754243196449,
+      "loss": 1.1459,
+      "step": 5515
+    },
+    {
+      "epoch": 0.4837078274499917,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0019274764561305502,
+      "loss": 1.1523,
+      "step": 5516
+    },
+    {
+      "epoch": 0.48379551922436625,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0019270774626363194,
+      "loss": 1.2307,
+      "step": 5517
+    },
+    {
+      "epoch": 0.4838832109987408,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0019266784438733387,
+      "loss": 1.1652,
+      "step": 5518
+    },
+    {
+      "epoch": 0.48397090277311533,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001926279399877999,
+      "loss": 1.1742,
+      "step": 5519
+    },
+    {
+      "epoch": 0.4840585945474899,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.001925880330686691,
+      "loss": 1.2336,
+      "step": 5520
+    },
+    {
+      "epoch": 0.48414628632186446,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0019254812363358087,
+      "loss": 1.1945,
+      "step": 5521
+    },
+    {
+      "epoch": 0.48423397809623897,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0019250821168617493,
+      "loss": 1.1766,
+      "step": 5522
+    },
+    {
+      "epoch": 0.48432166987061354,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0019246829723009102,
+      "loss": 1.1674,
+      "step": 5523
+    },
+    {
+      "epoch": 0.48440936164498805,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0019242838026896935,
+      "loss": 1.1515,
+      "step": 5524
+    },
+    {
+      "epoch": 0.4844970534193626,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0019238846080645022,
+      "loss": 1.1357,
+      "step": 5525
+    },
+    {
+      "epoch": 0.4845847451937372,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0019234853884617417,
+      "loss": 1.1147,
+      "step": 5526
+    },
+    {
+      "epoch": 0.4846724369681117,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0019230861439178194,
+      "loss": 1.1837,
+      "step": 5527
+    },
+    {
+      "epoch": 0.48476012874248625,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0019226868744691462,
+      "loss": 1.228,
+      "step": 5528
+    },
+    {
+      "epoch": 0.4848478205168608,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001922287580152134,
+      "loss": 1.1785,
+      "step": 5529
+    },
+    {
+      "epoch": 0.4849355122912353,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001921888261003197,
+      "loss": 1.1373,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4850232040656099,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0019214889170587528,
+      "loss": 1.1214,
+      "step": 5531
+    },
+    {
+      "epoch": 0.48511089583998446,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00192108954835522,
+      "loss": 1.2121,
+      "step": 5532
+    },
+    {
+      "epoch": 0.48519858761435897,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0019206901549290202,
+      "loss": 1.1636,
+      "step": 5533
+    },
+    {
+      "epoch": 0.48528627938873353,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.001920290736816577,
+      "loss": 1.1863,
+      "step": 5534
+    },
+    {
+      "epoch": 0.4853739711631081,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0019198912940543167,
+      "loss": 1.1505,
+      "step": 5535
+    },
+    {
+      "epoch": 0.4854616629374826,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0019194918266786667,
+      "loss": 1.2243,
+      "step": 5536
+    },
+    {
+      "epoch": 0.4855493547118572,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0019190923347260586,
+      "loss": 1.1104,
+      "step": 5537
+    },
+    {
+      "epoch": 0.4856370464862317,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001918692818232923,
+      "loss": 1.2346,
+      "step": 5538
+    },
+    {
+      "epoch": 0.48572473826060625,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.001918293277235697,
+      "loss": 1.1643,
+      "step": 5539
+    },
+    {
+      "epoch": 0.4858124300349808,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0019178937117708163,
+      "loss": 1.1439,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4859001218093553,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0019174941218747204,
+      "loss": 1.1538,
+      "step": 5541
+    },
+    {
+      "epoch": 0.4859878135837299,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0019170945075838515,
+      "loss": 1.1435,
+      "step": 5542
+    },
+    {
+      "epoch": 0.48607550535810445,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0019166948689346528,
+      "loss": 1.1268,
+      "step": 5543
+    },
+    {
+      "epoch": 0.48616319713247896,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0019162952059635706,
+      "loss": 1.204,
+      "step": 5544
+    },
+    {
+      "epoch": 0.48625088890685353,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0019158955187070536,
+      "loss": 1.1307,
+      "step": 5545
+    },
+    {
+      "epoch": 0.4863385806812281,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0019154958072015512,
+      "loss": 1.2306,
+      "step": 5546
+    },
+    {
+      "epoch": 0.4864262724556026,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0019150960714835162,
+      "loss": 1.2036,
+      "step": 5547
+    },
+    {
+      "epoch": 0.48651396422997717,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0019146963115894045,
+      "loss": 1.1675,
+      "step": 5548
+    },
+    {
+      "epoch": 0.4866016560043517,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0019142965275556717,
+      "loss": 1.1721,
+      "step": 5549
+    },
+    {
+      "epoch": 0.48668934777872624,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0019138967194187785,
+      "loss": 1.1753,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4867770395531008,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0019134968872151856,
+      "loss": 1.1615,
+      "step": 5551
+    },
+    {
+      "epoch": 0.4868647313274753,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001913097030981357,
+      "loss": 1.159,
+      "step": 5552
+    },
+    {
+      "epoch": 0.4869524231018499,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0019126971507537577,
+      "loss": 1.1597,
+      "step": 5553
+    },
+    {
+      "epoch": 0.48704011487622445,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0019122972465688568,
+      "loss": 1.1994,
+      "step": 5554
+    },
+    {
+      "epoch": 0.48712780665059896,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0019118973184631244,
+      "loss": 1.1902,
+      "step": 5555
+    },
+    {
+      "epoch": 0.4872154984249735,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001911497366473032,
+      "loss": 1.1935,
+      "step": 5556
+    },
+    {
+      "epoch": 0.4873031901993481,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0019110973906350556,
+      "loss": 1.169,
+      "step": 5557
+    },
+    {
+      "epoch": 0.4873908819737226,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0019106973909856708,
+      "loss": 1.1402,
+      "step": 5558
+    },
+    {
+      "epoch": 0.48747857374809717,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0019102973675613568,
+      "loss": 1.1357,
+      "step": 5559
+    },
+    {
+      "epoch": 0.4875662655224717,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0019098973203985948,
+      "loss": 1.1902,
+      "step": 5560
+    },
+    {
+      "epoch": 0.48765395729684624,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0019094972495338684,
+      "loss": 1.1765,
+      "step": 5561
+    },
+    {
+      "epoch": 0.4877416490712208,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0019090971550036627,
+      "loss": 1.1093,
+      "step": 5562
+    },
+    {
+      "epoch": 0.4878293408455953,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0019086970368444654,
+      "loss": 1.1568,
+      "step": 5563
+    },
+    {
+      "epoch": 0.4879170326199699,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001908296895092766,
+      "loss": 1.1526,
+      "step": 5564
+    },
+    {
+      "epoch": 0.48800472439434445,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0019078967297850562,
+      "loss": 1.2091,
+      "step": 5565
+    },
+    {
+      "epoch": 0.48809241616871896,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0019074965409578307,
+      "loss": 1.226,
+      "step": 5566
+    },
+    {
+      "epoch": 0.4881801079430935,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0019070963286475858,
+      "loss": 1.1625,
+      "step": 5567
+    },
+    {
+      "epoch": 0.4882677997174681,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.001906696092890819,
+      "loss": 1.1822,
+      "step": 5568
+    },
+    {
+      "epoch": 0.4883554914918426,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0019062958337240311,
+      "loss": 1.1377,
+      "step": 5569
+    },
+    {
+      "epoch": 0.48844318326621716,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0019058955511837246,
+      "loss": 1.1434,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4885308750405917,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.001905495245306405,
+      "loss": 1.1226,
+      "step": 5571
+    },
+    {
+      "epoch": 0.48861856681496624,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0019050949161285781,
+      "loss": 1.1819,
+      "step": 5572
+    },
+    {
+      "epoch": 0.4887062585893408,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.001904694563686754,
+      "loss": 1.152,
+      "step": 5573
+    },
+    {
+      "epoch": 0.4887939503637153,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0019042941880174425,
+      "loss": 1.2202,
+      "step": 5574
+    },
+    {
+      "epoch": 0.4888816421380899,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0019038937891571578,
+      "loss": 1.2113,
+      "step": 5575
+    },
+    {
+      "epoch": 0.48896933391246444,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.001903493367142415,
+      "loss": 1.1988,
+      "step": 5576
+    },
+    {
+      "epoch": 0.48905702568683895,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0019030929220097317,
+      "loss": 1.1405,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4891447174612135,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0019026924537956266,
+      "loss": 1.1408,
+      "step": 5578
+    },
+    {
+      "epoch": 0.4892324092355881,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001902291962536623,
+      "loss": 1.1574,
+      "step": 5579
+    },
+    {
+      "epoch": 0.4893201010099626,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0019018914482692433,
+      "loss": 1.1742,
+      "step": 5580
+    },
+    {
+      "epoch": 0.48940779278433716,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0019014909110300132,
+      "loss": 1.2344,
+      "step": 5581
+    },
+    {
+      "epoch": 0.4894954845587117,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0019010903508554617,
+      "loss": 1.1444,
+      "step": 5582
+    },
+    {
+      "epoch": 0.48958317633308623,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0019006897677821182,
+      "loss": 1.209,
+      "step": 5583
+    },
+    {
+      "epoch": 0.4896708681074608,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0019002891618465152,
+      "loss": 1.1928,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4897585598818353,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0018998885330851866,
+      "loss": 1.0958,
+      "step": 5585
+    },
+    {
+      "epoch": 0.4898462516562099,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0018994878815346695,
+      "loss": 1.1861,
+      "step": 5586
+    },
+    {
+      "epoch": 0.48993394343058444,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0018990872072315009,
+      "loss": 1.1376,
+      "step": 5587
+    },
+    {
+      "epoch": 0.49002163520495895,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0018986865102122226,
+      "loss": 1.1151,
+      "step": 5588
+    },
+    {
+      "epoch": 0.4901093269793335,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001898285790513376,
+      "loss": 1.1595,
+      "step": 5589
+    },
+    {
+      "epoch": 0.4901970187537081,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001897885048171507,
+      "loss": 1.1631,
+      "step": 5590
+    },
+    {
+      "epoch": 0.4902847105280826,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0018974842832231603,
+      "loss": 1.1789,
+      "step": 5591
+    },
+    {
+      "epoch": 0.49037240230245716,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001897083495704887,
+      "loss": 1.1362,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4904600940768317,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001896682685653236,
+      "loss": 1.1265,
+      "step": 5593
+    },
+    {
+      "epoch": 0.49054778585120623,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0018962818531047612,
+      "loss": 1.1857,
+      "step": 5594
+    },
+    {
+      "epoch": 0.4906354776255808,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001895880998096017,
+      "loss": 1.0938,
+      "step": 5595
+    },
+    {
+      "epoch": 0.4907231693999553,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0018954801206635607,
+      "loss": 1.1917,
+      "step": 5596
+    },
+    {
+      "epoch": 0.49081086117432987,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0018950792208439515,
+      "loss": 1.1448,
+      "step": 5597
+    },
+    {
+      "epoch": 0.49089855294870444,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0018946782986737495,
+      "loss": 1.1591,
+      "step": 5598
+    },
+    {
+      "epoch": 0.49098624472307895,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001894277354189518,
+      "loss": 1.1763,
+      "step": 5599
+    },
+    {
+      "epoch": 0.4910739364974535,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0018938763874278227,
+      "loss": 1.159,
+      "step": 5600
+    },
+    {
+      "epoch": 0.4911616282718281,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0018934753984252309,
+      "loss": 1.1632,
+      "step": 5601
+    },
+    {
+      "epoch": 0.4912493200462026,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0018930743872183104,
+      "loss": 1.1699,
+      "step": 5602
+    },
+    {
+      "epoch": 0.49133701182057715,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.001892673353843634,
+      "loss": 1.1563,
+      "step": 5603
+    },
+    {
+      "epoch": 0.4914247035949517,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0018922722983377737,
+      "loss": 1.1437,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4915123953693262,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0018918712207373059,
+      "loss": 1.1474,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4916000871437008,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0018914701210788066,
+      "loss": 1.2111,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4916877789180753,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0018910689993988559,
+      "loss": 1.1501,
+      "step": 5607
+    },
+    {
+      "epoch": 0.49177547069244987,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0018906678557340344,
+      "loss": 1.1374,
+      "step": 5608
+    },
+    {
+      "epoch": 0.49186316246682443,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0018902666901209257,
+      "loss": 1.1501,
+      "step": 5609
+    },
+    {
+      "epoch": 0.49195085424119894,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0018898655025961155,
+      "loss": 1.1997,
+      "step": 5610
+    },
+    {
+      "epoch": 0.4920385460155735,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0018894642931961904,
+      "loss": 1.1713,
+      "step": 5611
+    },
+    {
+      "epoch": 0.4921262377899481,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0018890630619577402,
+      "loss": 1.2223,
+      "step": 5612
+    },
+    {
+      "epoch": 0.4922139295643226,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0018886618089173559,
+      "loss": 1.1973,
+      "step": 5613
+    },
+    {
+      "epoch": 0.49230162133869715,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0018882605341116305,
+      "loss": 1.1854,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4923893131130717,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0018878592375771595,
+      "loss": 1.1021,
+      "step": 5615
+    },
+    {
+      "epoch": 0.4924770048874462,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00188745791935054,
+      "loss": 1.2012,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4925646966618208,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0018870565794683715,
+      "loss": 1.0881,
+      "step": 5617
+    },
+    {
+      "epoch": 0.49265238843619535,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0018866552179672548,
+      "loss": 1.1713,
+      "step": 5618
+    },
+    {
+      "epoch": 0.49274008021056986,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001886253834883793,
+      "loss": 1.133,
+      "step": 5619
+    },
+    {
+      "epoch": 0.49282777198494443,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001885852430254591,
+      "loss": 1.1271,
+      "step": 5620
+    },
+    {
+      "epoch": 0.49291546375931894,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.001885451004116257,
+      "loss": 1.1898,
+      "step": 5621
+    },
+    {
+      "epoch": 0.4930031555336935,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0018850495565053992,
+      "loss": 1.1786,
+      "step": 5622
+    },
+    {
+      "epoch": 0.49309084730806807,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001884648087458628,
+      "loss": 1.1575,
+      "step": 5623
+    },
+    {
+      "epoch": 0.4931785390824426,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0018842465970125572,
+      "loss": 1.2206,
+      "step": 5624
+    },
+    {
+      "epoch": 0.49326623085681714,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0018838450852038018,
+      "loss": 1.1465,
+      "step": 5625
+    },
+    {
+      "epoch": 0.4933539226311917,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001883443552068978,
+      "loss": 1.1343,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4934416144055662,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0018830419976447047,
+      "loss": 1.1781,
+      "step": 5627
+    },
+    {
+      "epoch": 0.4935293061799408,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.001882640421967603,
+      "loss": 1.157,
+      "step": 5628
+    },
+    {
+      "epoch": 0.49361699795431535,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0018822388250742954,
+      "loss": 1.1914,
+      "step": 5629
+    },
+    {
+      "epoch": 0.49370468972868986,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0018818372070014073,
+      "loss": 1.1222,
+      "step": 5630
+    },
+    {
+      "epoch": 0.4937923815030644,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0018814355677855632,
+      "loss": 1.1642,
+      "step": 5631
+    },
+    {
+      "epoch": 0.49388007327743894,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0018810339074633931,
+      "loss": 1.217,
+      "step": 5632
+    },
+    {
+      "epoch": 0.4939677650518135,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001880632226071527,
+      "loss": 1.173,
+      "step": 5633
+    },
+    {
+      "epoch": 0.49405545682618807,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0018802305236465973,
+      "loss": 1.1862,
+      "step": 5634
+    },
+    {
+      "epoch": 0.4941431486005626,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0018798288002252381,
+      "loss": 1.1268,
+      "step": 5635
+    },
+    {
+      "epoch": 0.49423084037493714,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0018794270558440858,
+      "loss": 1.1543,
+      "step": 5636
+    },
+    {
+      "epoch": 0.4943185321493117,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001879025290539778,
+      "loss": 1.1281,
+      "step": 5637
+    },
+    {
+      "epoch": 0.4944062239236862,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0018786235043489552,
+      "loss": 1.1674,
+      "step": 5638
+    },
+    {
+      "epoch": 0.4944939156980608,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0018782216973082593,
+      "loss": 1.2175,
+      "step": 5639
+    },
+    {
+      "epoch": 0.49458160747243535,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0018778198694543332,
+      "loss": 1.2016,
+      "step": 5640
+    },
+    {
+      "epoch": 0.49466929924680986,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0018774180208238232,
+      "loss": 1.1328,
+      "step": 5641
+    },
+    {
+      "epoch": 0.4947569910211844,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0018770161514533772,
+      "loss": 1.2101,
+      "step": 5642
+    },
+    {
+      "epoch": 0.49484468279555893,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0018766142613796438,
+      "loss": 1.1545,
+      "step": 5643
+    },
+    {
+      "epoch": 0.4949323745699335,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001876212350639275,
+      "loss": 1.1378,
+      "step": 5644
+    },
+    {
+      "epoch": 0.49502006634430806,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0018758104192689235,
+      "loss": 1.1898,
+      "step": 5645
+    },
+    {
+      "epoch": 0.4951077581186826,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.001875408467305245,
+      "loss": 1.2606,
+      "step": 5646
+    },
+    {
+      "epoch": 0.49519544989305714,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001875006494784896,
+      "loss": 1.1996,
+      "step": 5647
+    },
+    {
+      "epoch": 0.4952831416674317,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0018746045017445359,
+      "loss": 1.1701,
+      "step": 5648
+    },
+    {
+      "epoch": 0.4953708334418062,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0018742024882208242,
+      "loss": 1.1601,
+      "step": 5649
+    },
+    {
+      "epoch": 0.4954585252161808,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0018738004542504252,
+      "loss": 1.166,
+      "step": 5650
+    },
+    {
+      "epoch": 0.49554621699055534,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001873398399870002,
+      "loss": 1.1934,
+      "step": 5651
+    },
+    {
+      "epoch": 0.49563390876492985,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0018729963251162217,
+      "loss": 1.1283,
+      "step": 5652
+    },
+    {
+      "epoch": 0.4957216005393044,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001872594230025752,
+      "loss": 1.1343,
+      "step": 5653
+    },
+    {
+      "epoch": 0.495809292313679,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0018721921146352635,
+      "loss": 1.1333,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4958969840880535,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0018717899789814274,
+      "loss": 1.1926,
+      "step": 5655
+    },
+    {
+      "epoch": 0.49598467586242806,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0018713878231009177,
+      "loss": 1.1943,
+      "step": 5656
+    },
+    {
+      "epoch": 0.49607236763680257,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0018709856470304103,
+      "loss": 1.171,
+      "step": 5657
+    },
+    {
+      "epoch": 0.49616005941117713,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0018705834508065823,
+      "loss": 1.1957,
+      "step": 5658
+    },
+    {
+      "epoch": 0.4962477511855517,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.001870181234466113,
+      "loss": 1.1959,
+      "step": 5659
+    },
+    {
+      "epoch": 0.4963354429599262,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0018697789980456833,
+      "loss": 1.1978,
+      "step": 5660
+    },
+    {
+      "epoch": 0.4964231347343008,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0018693767415819769,
+      "loss": 1.2062,
+      "step": 5661
+    },
+    {
+      "epoch": 0.49651082650867534,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0018689744651116773,
+      "loss": 1.1486,
+      "step": 5662
+    },
+    {
+      "epoch": 0.49659851828304985,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0018685721686714724,
+      "loss": 1.1489,
+      "step": 5663
+    },
+    {
+      "epoch": 0.4966862100574244,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0018681698522980497,
+      "loss": 1.1639,
+      "step": 5664
+    },
+    {
+      "epoch": 0.496773901831799,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0018677675160280992,
+      "loss": 1.165,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4968615936061735,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001867365159898314,
+      "loss": 1.1317,
+      "step": 5666
+    },
+    {
+      "epoch": 0.49694928538054806,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.001866962783945387,
+      "loss": 1.1819,
+      "step": 5667
+    },
+    {
+      "epoch": 0.49703697715492257,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0018665603882060144,
+      "loss": 1.1626,
+      "step": 5668
+    },
+    {
+      "epoch": 0.49712466892929713,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001866157972716893,
+      "loss": 1.1676,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4972123607036717,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0018657555375147232,
+      "loss": 1.1656,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4973000524780462,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018653530826362047,
+      "loss": 1.1896,
+      "step": 5671
+    },
+    {
+      "epoch": 0.49738774425242077,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0018649506081180418,
+      "loss": 1.1408,
+      "step": 5672
+    },
+    {
+      "epoch": 0.49747543602679534,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0018645481139969377,
+      "loss": 1.1446,
+      "step": 5673
+    },
+    {
+      "epoch": 0.49756312780116985,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0018641456003096002,
+      "loss": 1.1638,
+      "step": 5674
+    },
+    {
+      "epoch": 0.4976508195755444,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001863743067092736,
+      "loss": 1.1686,
+      "step": 5675
+    },
+    {
+      "epoch": 0.497738511349919,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0018633405143830557,
+      "loss": 1.1248,
+      "step": 5676
+    },
+    {
+      "epoch": 0.4978262031242935,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0018629379422172715,
+      "loss": 1.1806,
+      "step": 5677
+    },
+    {
+      "epoch": 0.49791389489866805,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0018625353506320967,
+      "loss": 1.1103,
+      "step": 5678
+    },
+    {
+      "epoch": 0.49800158667304256,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0018621327396642468,
+      "loss": 1.1912,
+      "step": 5679
+    },
+    {
+      "epoch": 0.4980892784474171,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001861730109350438,
+      "loss": 1.2087,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4981769702217917,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0018613274597273912,
+      "loss": 1.2358,
+      "step": 5681
+    },
+    {
+      "epoch": 0.4982646619961662,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0018609247908318244,
+      "loss": 1.1343,
+      "step": 5682
+    },
+    {
+      "epoch": 0.49835235377054077,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0018605221027004615,
+      "loss": 1.1756,
+      "step": 5683
+    },
+    {
+      "epoch": 0.49844004554491533,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0018601193953700261,
+      "loss": 1.1562,
+      "step": 5684
+    },
+    {
+      "epoch": 0.49852773731928984,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0018597166688772446,
+      "loss": 1.1202,
+      "step": 5685
+    },
+    {
+      "epoch": 0.4986154290936644,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.001859313923258844,
+      "loss": 1.1693,
+      "step": 5686
+    },
+    {
+      "epoch": 0.498703120868039,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018589111585515538,
+      "loss": 1.179,
+      "step": 5687
+    },
+    {
+      "epoch": 0.4987908126424135,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0018585083747921056,
+      "loss": 1.144,
+      "step": 5688
+    },
+    {
+      "epoch": 0.49887850441678805,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018581055720172318,
+      "loss": 1.1149,
+      "step": 5689
+    },
+    {
+      "epoch": 0.4989661961911626,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001857702750263667,
+      "loss": 1.1048,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4990538879655371,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0018572999095681476,
+      "loss": 1.2094,
+      "step": 5691
+    },
+    {
+      "epoch": 0.4991415797399117,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0018568970499674119,
+      "loss": 1.2009,
+      "step": 5692
+    },
+    {
+      "epoch": 0.4992292715142862,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0018564941714981987,
+      "loss": 1.1597,
+      "step": 5693
+    },
+    {
+      "epoch": 0.49931696328866076,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0018560912741972504,
+      "loss": 1.2405,
+      "step": 5694
+    },
+    {
+      "epoch": 0.49940465506303533,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0018556883581013098,
+      "loss": 1.2059,
+      "step": 5695
+    },
+    {
+      "epoch": 0.49949234683740984,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0018552854232471222,
+      "loss": 1.1652,
+      "step": 5696
+    },
+    {
+      "epoch": 0.4995800386117844,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0018548824696714334,
+      "loss": 1.1485,
+      "step": 5697
+    },
+    {
+      "epoch": 0.49966773038615897,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001854479497410993,
+      "loss": 1.2003,
+      "step": 5698
+    },
+    {
+      "epoch": 0.4997554221605335,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0018540765065025498,
+      "loss": 1.1798,
+      "step": 5699
+    },
+    {
+      "epoch": 0.49984311393490805,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.001853673496982856,
+      "loss": 1.1801,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4999308057092826,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0018532704688886657,
+      "loss": 1.1632,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5000184974836571,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0018528674222567328,
+      "loss": 1.2307,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5001061892580316,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.001852464357123815,
+      "loss": 1.2115,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5001938810324063,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0018520612735266702,
+      "loss": 1.1467,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5002815728067808,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001851658171502059,
+      "loss": 1.1028,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5003692645811553,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0018512550510867431,
+      "loss": 1.1291,
+      "step": 5706
+    },
+    {
+      "epoch": 0.5004569563555299,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.001850851912317486,
+      "loss": 1.1355,
+      "step": 5707
+    },
+    {
+      "epoch": 0.5005446481299044,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0018504487552310535,
+      "loss": 1.119,
+      "step": 5708
+    },
+    {
+      "epoch": 0.5006323399042789,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0018500455798642114,
+      "loss": 1.1954,
+      "step": 5709
+    },
+    {
+      "epoch": 0.5007200316786535,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001849642386253729,
+      "loss": 1.2129,
+      "step": 5710
+    },
+    {
+      "epoch": 0.500807723453028,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001849239174436376,
+      "loss": 1.1345,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5008954152274026,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0018488359444489252,
+      "loss": 1.182,
+      "step": 5712
+    },
+    {
+      "epoch": 0.5009831070017772,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0018484326963281493,
+      "loss": 1.192,
+      "step": 5713
+    },
+    {
+      "epoch": 0.5010707987761517,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001848029430110824,
+      "loss": 1.1425,
+      "step": 5714
+    },
+    {
+      "epoch": 0.5011584905505262,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0018476261458337255,
+      "loss": 1.1525,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5012461823249008,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0018472228435336328,
+      "loss": 1.1223,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5013338740992753,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0018468195232473263,
+      "loss": 1.1235,
+      "step": 5717
+    },
+    {
+      "epoch": 0.5014215658736498,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0018464161850115864,
+      "loss": 1.1074,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5015092576480245,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0018460128288631983,
+      "loss": 1.199,
+      "step": 5719
+    },
+    {
+      "epoch": 0.501596949422399,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001845609454838946,
+      "loss": 1.1578,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5016846411967735,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0018452060629756165,
+      "loss": 1.1661,
+      "step": 5721
+    },
+    {
+      "epoch": 0.501772332971148,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0018448026533099975,
+      "loss": 1.2222,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5018600247455226,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0018443992258788805,
+      "loss": 1.1709,
+      "step": 5723
+    },
+    {
+      "epoch": 0.5019477165198971,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.001843995780719055,
+      "loss": 1.1649,
+      "step": 5724
+    },
+    {
+      "epoch": 0.5020354082942716,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0018435923178673152,
+      "loss": 1.173,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5021231000686462,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0018431888373604559,
+      "loss": 1.2286,
+      "step": 5726
+    },
+    {
+      "epoch": 0.5022107918430208,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0018427853392352731,
+      "loss": 1.1544,
+      "step": 5727
+    },
+    {
+      "epoch": 0.5022984836173953,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0018423818235285654,
+      "loss": 1.1648,
+      "step": 5728
+    },
+    {
+      "epoch": 0.5023861753917699,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0018419782902771315,
+      "loss": 1.1256,
+      "step": 5729
+    },
+    {
+      "epoch": 0.5024738671661444,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0018415747395177737,
+      "loss": 1.1708,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5025615589405189,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0018411711712872938,
+      "loss": 1.1864,
+      "step": 5731
+    },
+    {
+      "epoch": 0.5026492507148935,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001840767585622497,
+      "loss": 1.1364,
+      "step": 5732
+    },
+    {
+      "epoch": 0.502736942489268,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0018403639825601886,
+      "loss": 1.1742,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5028246342636425,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0018399603621371764,
+      "loss": 1.1361,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5029123260380172,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0018395567243902696,
+      "loss": 1.1774,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5030000178123917,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0018391530693562788,
+      "loss": 1.0974,
+      "step": 5736
+    },
+    {
+      "epoch": 0.5030877095867662,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0018387493970720165,
+      "loss": 1.2111,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5031754013611408,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0018383457075742962,
+      "loss": 1.1339,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5032630931355153,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0018379420008999337,
+      "loss": 1.1121,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5033507849098898,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0018375382770857462,
+      "loss": 1.2145,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5034384766842644,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0018371345361685514,
+      "loss": 1.1664,
+      "step": 5741
+    },
+    {
+      "epoch": 0.503526168458639,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0018367307781851701,
+      "loss": 1.1216,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5036138602330135,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0018363270031724244,
+      "loss": 1.1755,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5037015520073881,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0018359232111671367,
+      "loss": 1.1663,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5037892437817626,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0018355194022061322,
+      "loss": 1.154,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5038769355561371,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0018351155763262365,
+      "loss": 1.1686,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5039646273305116,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001834711733564279,
+      "loss": 1.155,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5040523191048862,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0018343078739570876,
+      "loss": 1.166,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5041400108792607,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.001833903997541494,
+      "loss": 1.2249,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5042277026536353,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0018335001043543312,
+      "loss": 1.1903,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5043153944280099,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0018330961944324315,
+      "loss": 1.1876,
+      "step": 5751
+    },
+    {
+      "epoch": 0.5044030862023844,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0018326922678126322,
+      "loss": 1.1586,
+      "step": 5752
+    },
+    {
+      "epoch": 0.5044907779767589,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0018322883245317696,
+      "loss": 1.2384,
+      "step": 5753
+    },
+    {
+      "epoch": 0.5045784697511335,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0018318843646266823,
+      "loss": 1.2219,
+      "step": 5754
+    },
+    {
+      "epoch": 0.504666161525508,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0018314803881342104,
+      "loss": 1.1529,
+      "step": 5755
+    },
+    {
+      "epoch": 0.5047538532998825,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0018310763950911964,
+      "loss": 1.1267,
+      "step": 5756
+    },
+    {
+      "epoch": 0.5048415450742572,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0018306723855344817,
+      "loss": 1.2313,
+      "step": 5757
+    },
+    {
+      "epoch": 0.5049292368486317,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0018302683595009122,
+      "loss": 1.161,
+      "step": 5758
+    },
+    {
+      "epoch": 0.5050169286230062,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0018298643170273333,
+      "loss": 1.1566,
+      "step": 5759
+    },
+    {
+      "epoch": 0.5051046203973808,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0018294602581505933,
+      "loss": 1.2054,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5051923121717553,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0018290561829075415,
+      "loss": 1.2319,
+      "step": 5761
+    },
+    {
+      "epoch": 0.5052800039461298,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0018286520913350273,
+      "loss": 1.1668,
+      "step": 5762
+    },
+    {
+      "epoch": 0.5053676957205044,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001828247983469904,
+      "loss": 1.1772,
+      "step": 5763
+    },
+    {
+      "epoch": 0.505455387494879,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0018278438593490238,
+      "loss": 1.2506,
+      "step": 5764
+    },
+    {
+      "epoch": 0.5055430792692535,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0018274397190092438,
+      "loss": 1.1798,
+      "step": 5765
+    },
+    {
+      "epoch": 0.5056307710436281,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0018270355624874196,
+      "loss": 1.2016,
+      "step": 5766
+    },
+    {
+      "epoch": 0.5057184628180026,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.001826631389820409,
+      "loss": 1.1744,
+      "step": 5767
+    },
+    {
+      "epoch": 0.5058061545923771,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0018262272010450707,
+      "loss": 1.1798,
+      "step": 5768
+    },
+    {
+      "epoch": 0.5058938463667516,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0018258229961982675,
+      "loss": 1.1568,
+      "step": 5769
+    },
+    {
+      "epoch": 0.5059815381411262,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0018254187753168605,
+      "loss": 1.225,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5060692299155007,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.001825014538437714,
+      "loss": 1.1546,
+      "step": 5771
+    },
+    {
+      "epoch": 0.5061569216898752,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0018246102855976934,
+      "loss": 1.1973,
+      "step": 5772
+    },
+    {
+      "epoch": 0.5062446134642499,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0018242060168336656,
+      "loss": 1.135,
+      "step": 5773
+    },
+    {
+      "epoch": 0.5063323052386244,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0018238017321824994,
+      "loss": 1.1889,
+      "step": 5774
+    },
+    {
+      "epoch": 0.5064199970129989,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0018233974316810631,
+      "loss": 1.1415,
+      "step": 5775
+    },
+    {
+      "epoch": 0.5065076887873735,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0018229931153662293,
+      "loss": 1.1746,
+      "step": 5776
+    },
+    {
+      "epoch": 0.506595380561748,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0018225887832748694,
+      "loss": 1.1884,
+      "step": 5777
+    },
+    {
+      "epoch": 0.5066830723361225,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0018221844354438582,
+      "loss": 1.2115,
+      "step": 5778
+    },
+    {
+      "epoch": 0.5067707641104972,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0018217800719100706,
+      "loss": 1.1784,
+      "step": 5779
+    },
+    {
+      "epoch": 0.5068584558848717,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0018213756927103846,
+      "loss": 1.1439,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5069461476592462,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0018209712978816772,
+      "loss": 1.1463,
+      "step": 5781
+    },
+    {
+      "epoch": 0.5070338394336208,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0018205668874608288,
+      "loss": 1.1772,
+      "step": 5782
+    },
+    {
+      "epoch": 0.5071215312079953,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0018201624614847213,
+      "loss": 1.1902,
+      "step": 5783
+    },
+    {
+      "epoch": 0.5072092229823698,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0018197580199902361,
+      "loss": 1.1643,
+      "step": 5784
+    },
+    {
+      "epoch": 0.5072969147567444,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0018193535630142577,
+      "loss": 1.1249,
+      "step": 5785
+    },
+    {
+      "epoch": 0.5073846065311189,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0018189490905936713,
+      "loss": 1.1695,
+      "step": 5786
+    },
+    {
+      "epoch": 0.5074722983054935,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0018185446027653644,
+      "loss": 1.1812,
+      "step": 5787
+    },
+    {
+      "epoch": 0.5075599900798681,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0018181400995662244,
+      "loss": 1.1458,
+      "step": 5788
+    },
+    {
+      "epoch": 0.5076476818542426,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018177355810331417,
+      "loss": 1.0803,
+      "step": 5789
+    },
+    {
+      "epoch": 0.5077353736286171,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0018173310472030068,
+      "loss": 1.1709,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5078230654029917,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.001816926498112713,
+      "loss": 1.1859,
+      "step": 5791
+    },
+    {
+      "epoch": 0.5079107571773662,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0018165219337991528,
+      "loss": 1.1926,
+      "step": 5792
+    },
+    {
+      "epoch": 0.5079984489517407,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0018161173542992222,
+      "loss": 1.2352,
+      "step": 5793
+    },
+    {
+      "epoch": 0.5080861407261152,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0018157127596498182,
+      "loss": 1.1169,
+      "step": 5794
+    },
+    {
+      "epoch": 0.5081738325004899,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0018153081498878375,
+      "loss": 1.152,
+      "step": 5795
+    },
+    {
+      "epoch": 0.5082615242748644,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0018149035250501807,
+      "loss": 1.1563,
+      "step": 5796
+    },
+    {
+      "epoch": 0.5083492160492389,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.001814498885173748,
+      "loss": 1.168,
+      "step": 5797
+    },
+    {
+      "epoch": 0.5084369078236135,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001814094230295442,
+      "loss": 1.1188,
+      "step": 5798
+    },
+    {
+      "epoch": 0.508524599597988,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0018136895604521656,
+      "loss": 1.1922,
+      "step": 5799
+    },
+    {
+      "epoch": 0.5086122913723625,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0018132848756808237,
+      "loss": 1.1117,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5086999831467371,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001812880176018323,
+      "loss": 1.2078,
+      "step": 5801
+    },
+    {
+      "epoch": 0.5087876749211117,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0018124754615015705,
+      "loss": 1.1425,
+      "step": 5802
+    },
+    {
+      "epoch": 0.5088753666954862,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018120707321674751,
+      "loss": 1.1504,
+      "step": 5803
+    },
+    {
+      "epoch": 0.5089630584698608,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0018116659880529472,
+      "loss": 1.1957,
+      "step": 5804
+    },
+    {
+      "epoch": 0.5090507502442353,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0018112612291948994,
+      "loss": 1.1273,
+      "step": 5805
+    },
+    {
+      "epoch": 0.5091384420186098,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0018108564556302425,
+      "loss": 1.155,
+      "step": 5806
+    },
+    {
+      "epoch": 0.5092261337929844,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0018104516673958931,
+      "loss": 1.1446,
+      "step": 5807
+    },
+    {
+      "epoch": 0.5093138255673589,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001810046864528766,
+      "loss": 1.1482,
+      "step": 5808
+    },
+    {
+      "epoch": 0.5094015173417334,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0018096420470657777,
+      "loss": 1.1814,
+      "step": 5809
+    },
+    {
+      "epoch": 0.5094892091161081,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0018092372150438464,
+      "loss": 1.2135,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5095769008904826,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001808832368499892,
+      "loss": 1.1625,
+      "step": 5811
+    },
+    {
+      "epoch": 0.5096645926648571,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0018084275074708364,
+      "loss": 1.186,
+      "step": 5812
+    },
+    {
+      "epoch": 0.5097522844392317,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0018080226319936005,
+      "loss": 1.1213,
+      "step": 5813
+    },
+    {
+      "epoch": 0.5098399762136062,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0018076177421051088,
+      "loss": 1.139,
+      "step": 5814
+    },
+    {
+      "epoch": 0.5099276679879807,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001807212837842286,
+      "loss": 1.2451,
+      "step": 5815
+    },
+    {
+      "epoch": 0.5100153597623552,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001806807919242058,
+      "loss": 1.1949,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5101030515367299,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0018064029863413527,
+      "loss": 1.1656,
+      "step": 5817
+    },
+    {
+      "epoch": 0.5101907433111044,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0018059980391770993,
+      "loss": 1.1062,
+      "step": 5818
+    },
+    {
+      "epoch": 0.5102784350854789,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0018055930777862265,
+      "loss": 1.1592,
+      "step": 5819
+    },
+    {
+      "epoch": 0.5103661268598535,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0018051881022056672,
+      "loss": 1.1576,
+      "step": 5820
+    },
+    {
+      "epoch": 0.510453818634228,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0018047831124723534,
+      "loss": 1.1511,
+      "step": 5821
+    },
+    {
+      "epoch": 0.5105415104086025,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0018043781086232195,
+      "loss": 1.1957,
+      "step": 5822
+    },
+    {
+      "epoch": 0.5106292021829771,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0018039730906952004,
+      "loss": 1.1676,
+      "step": 5823
+    },
+    {
+      "epoch": 0.5107168939573516,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0018035680587252328,
+      "loss": 1.2499,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5108045857317262,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018031630127502553,
+      "loss": 1.1384,
+      "step": 5825
+    },
+    {
+      "epoch": 0.5108922775061008,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0018027579528072058,
+      "loss": 1.1521,
+      "step": 5826
+    },
+    {
+      "epoch": 0.5109799692804753,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0018023528789330257,
+      "loss": 1.208,
+      "step": 5827
+    },
+    {
+      "epoch": 0.5110676610548498,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0018019477911646558,
+      "loss": 1.1748,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5111553528292244,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0018015426895390402,
+      "loss": 1.1991,
+      "step": 5829
+    },
+    {
+      "epoch": 0.5112430446035989,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001801137574093122,
+      "loss": 1.1822,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5113307363779734,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0018007324448638482,
+      "loss": 1.1399,
+      "step": 5831
+    },
+    {
+      "epoch": 0.5114184281523481,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0018003273018881631,
+      "loss": 1.2482,
+      "step": 5832
+    },
+    {
+      "epoch": 0.5115061199267226,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0017999221452030174,
+      "loss": 1.2276,
+      "step": 5833
+    },
+    {
+      "epoch": 0.5115938117010971,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0017995169748453584,
+      "loss": 1.1534,
+      "step": 5834
+    },
+    {
+      "epoch": 0.5116815034754717,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0017991117908521374,
+      "loss": 1.1785,
+      "step": 5835
+    },
+    {
+      "epoch": 0.5117691952498462,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0017987065932603065,
+      "loss": 1.1173,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5118568870242207,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0017983013821068175,
+      "loss": 1.1945,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5119445787985953,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0017978961574286255,
+      "loss": 1.2266,
+      "step": 5838
+    },
+    {
+      "epoch": 0.5120322705729698,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0017974909192626858,
+      "loss": 1.2387,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5121199623473444,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0017970856676459547,
+      "loss": 1.1835,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5122076541217189,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0017966804026153906,
+      "loss": 1.162,
+      "step": 5841
+    },
+    {
+      "epoch": 0.5122953458960935,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001796275124207953,
+      "loss": 1.2007,
+      "step": 5842
+    },
+    {
+      "epoch": 0.512383037670468,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001795869832460601,
+      "loss": 1.2048,
+      "step": 5843
+    },
+    {
+      "epoch": 0.5124707294448425,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001795464527410297,
+      "loss": 1.1028,
+      "step": 5844
+    },
+    {
+      "epoch": 0.5125584212192171,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0017950592090940035,
+      "loss": 1.1461,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5126461129935916,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0017946538775486848,
+      "loss": 1.1955,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5127338047679662,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0017942485328113057,
+      "loss": 1.2078,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5128214965423408,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0017938431749188327,
+      "loss": 1.134,
+      "step": 5848
+    },
+    {
+      "epoch": 0.5129091883167153,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0017934378039082342,
+      "loss": 1.1885,
+      "step": 5849
+    },
+    {
+      "epoch": 0.5129968800910898,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0017930324198164775,
+      "loss": 1.1321,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5130845718654644,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0017926270226805334,
+      "loss": 1.125,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5131722636398389,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0017922216125373733,
+      "loss": 1.1672,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5132599554142134,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0017918161894239692,
+      "loss": 1.0977,
+      "step": 5853
+    },
+    {
+      "epoch": 0.513347647188588,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0017914107533772948,
+      "loss": 1.2088,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5134353389629626,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0017910053044343248,
+      "loss": 1.1303,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5135230307373371,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001790599842632035,
+      "loss": 1.1778,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5136107225117117,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001790194368007402,
+      "loss": 1.2318,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5136984142860862,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0017897888805974052,
+      "loss": 1.2004,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5137861060604607,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0017893833804390235,
+      "loss": 1.1498,
+      "step": 5859
+    },
+    {
+      "epoch": 0.5138737978348353,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017889778675692367,
+      "loss": 1.1901,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5139614896092098,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0017885723420250272,
+      "loss": 1.1503,
+      "step": 5861
+    },
+    {
+      "epoch": 0.5140491813835844,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0017881668038433785,
+      "loss": 1.1612,
+      "step": 5862
+    },
+    {
+      "epoch": 0.514136873157959,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0017877612530612732,
+      "loss": 1.2007,
+      "step": 5863
+    },
+    {
+      "epoch": 0.5142245649323335,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001787355689715697,
+      "loss": 1.2136,
+      "step": 5864
+    },
+    {
+      "epoch": 0.514312256706708,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0017869501138436372,
+      "loss": 1.1456,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5143999484810825,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00178654452548208,
+      "loss": 1.1664,
+      "step": 5866
+    },
+    {
+      "epoch": 0.5144876402554571,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0017861389246680156,
+      "loss": 1.0804,
+      "step": 5867
+    },
+    {
+      "epoch": 0.5145753320298316,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.001785733311438432,
+      "loss": 1.198,
+      "step": 5868
+    },
+    {
+      "epoch": 0.5146630238042061,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0017853276858303209,
+      "loss": 1.1833,
+      "step": 5869
+    },
+    {
+      "epoch": 0.5147507155785808,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0017849220478806738,
+      "loss": 1.1416,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5148384073529553,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0017845163976264846,
+      "loss": 1.17,
+      "step": 5871
+    },
+    {
+      "epoch": 0.5149260991273298,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0017841107351047472,
+      "loss": 1.1496,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5150137909017044,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017837050603524568,
+      "loss": 1.1755,
+      "step": 5873
+    },
+    {
+      "epoch": 0.5151014826760789,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0017832993734066102,
+      "loss": 1.2067,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5151891744504534,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.001782893674304205,
+      "loss": 1.1894,
+      "step": 5875
+    },
+    {
+      "epoch": 0.515276866224828,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0017824879630822397,
+      "loss": 1.1842,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5153645579992026,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0017820822397777137,
+      "loss": 1.1571,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5154522497735771,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001781676504427629,
+      "loss": 1.146,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5155399415479517,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0017812707570689867,
+      "loss": 1.1602,
+      "step": 5879
+    },
+    {
+      "epoch": 0.5156276333223262,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0017808649977387903,
+      "loss": 1.205,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5157153250967007,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0017804592264740438,
+      "loss": 1.2132,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5158030168710753,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001780053443311753,
+      "loss": 1.1783,
+      "step": 5882
+    },
+    {
+      "epoch": 0.5158907086454498,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0017796476482889233,
+      "loss": 1.2132,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5159784004198243,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0017792418414425634,
+      "loss": 1.118,
+      "step": 5884
+    },
+    {
+      "epoch": 0.516066092194199,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0017788360228096812,
+      "loss": 1.125,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5161537839685735,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0017784301924272856,
+      "loss": 1.1587,
+      "step": 5886
+    },
+    {
+      "epoch": 0.516241475742948,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0017780243503323888,
+      "loss": 1.2328,
+      "step": 5887
+    },
+    {
+      "epoch": 0.5163291675173225,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001777618496562001,
+      "loss": 1.1395,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5164168592916971,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0017772126311531363,
+      "loss": 1.174,
+      "step": 5889
+    },
+    {
+      "epoch": 0.5165045510660716,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0017768067541428078,
+      "loss": 1.2274,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5165922428404461,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001776400865568031,
+      "loss": 1.1515,
+      "step": 5891
+    },
+    {
+      "epoch": 0.5166799346148208,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0017759949654658207,
+      "loss": 1.1719,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5167676263891953,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0017755890538731958,
+      "loss": 1.1532,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5168553181635698,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017751831308271732,
+      "loss": 1.2069,
+      "step": 5894
+    },
+    {
+      "epoch": 0.5169430099379444,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017747771963647716,
+      "loss": 1.1432,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5170307017123189,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0017743712505230123,
+      "loss": 1.2202,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5171183934866934,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0017739652933389158,
+      "loss": 1.1342,
+      "step": 5897
+    },
+    {
+      "epoch": 0.517206085261068,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0017735593248495048,
+      "loss": 1.2046,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5172937770354425,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.001773153345091802,
+      "loss": 1.1277,
+      "step": 5899
+    },
+    {
+      "epoch": 0.5173814688098171,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0017727473541028326,
+      "loss": 1.1266,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5174691605841917,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0017723413519196212,
+      "loss": 1.2666,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5175568523585662,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0017719353385791943,
+      "loss": 1.1376,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5176445441329407,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.001771529314118579,
+      "loss": 1.1263,
+      "step": 5903
+    },
+    {
+      "epoch": 0.5177322359073153,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0017711232785748042,
+      "loss": 1.1738,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5178199276816898,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0017707172319848992,
+      "loss": 1.1593,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5179076194560643,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0017703111743858939,
+      "loss": 1.1658,
+      "step": 5906
+    },
+    {
+      "epoch": 0.517995311230439,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0017699051058148207,
+      "loss": 1.1897,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5180830030048135,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0017694990263087103,
+      "loss": 1.1685,
+      "step": 5908
+    },
+    {
+      "epoch": 0.518170694779188,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001769092935904598,
+      "loss": 1.2157,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5182583865535626,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0017686868346395177,
+      "loss": 1.1714,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5183460783279371,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0017682807225505037,
+      "loss": 1.2533,
+      "step": 5911
+    },
+    {
+      "epoch": 0.5184337701023116,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0017678745996745936,
+      "loss": 1.1112,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5185214618766861,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001767468466048824,
+      "loss": 1.0989,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5186091536510608,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0017670623217102338,
+      "loss": 1.1558,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5186968454254353,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0017666561666958619,
+      "loss": 1.1009,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5187845371998098,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0017662500010427488,
+      "loss": 1.1835,
+      "step": 5916
+    },
+    {
+      "epoch": 0.5188722289741844,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0017658438247879354,
+      "loss": 1.1869,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5189599207485589,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001765437637968465,
+      "loss": 1.1594,
+      "step": 5918
+    },
+    {
+      "epoch": 0.5190476125229334,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0017650314406213795,
+      "loss": 1.1837,
+      "step": 5919
+    },
+    {
+      "epoch": 0.519135304297308,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0017646252327837236,
+      "loss": 1.1915,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5192229960716825,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0017642190144925422,
+      "loss": 1.2011,
+      "step": 5921
+    },
+    {
+      "epoch": 0.519310687846057,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001763812785784882,
+      "loss": 1.1376,
+      "step": 5922
+    },
+    {
+      "epoch": 0.5193983796204317,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.001763406546697789,
+      "loss": 1.168,
+      "step": 5923
+    },
+    {
+      "epoch": 0.5194860713948062,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.001763000297268312,
+      "loss": 1.1994,
+      "step": 5924
+    },
+    {
+      "epoch": 0.5195737631691807,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0017625940375334997,
+      "loss": 1.1415,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5196614549435553,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0017621877675304016,
+      "loss": 1.1392,
+      "step": 5926
+    },
+    {
+      "epoch": 0.5197491467179298,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001761781487296069,
+      "loss": 1.1923,
+      "step": 5927
+    },
+    {
+      "epoch": 0.5198368384923043,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.001761375196867553,
+      "loss": 1.211,
+      "step": 5928
+    },
+    {
+      "epoch": 0.519924530266679,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0017609688962819069,
+      "loss": 1.161,
+      "step": 5929
+    },
+    {
+      "epoch": 0.5200122220410535,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0017605625855761833,
+      "loss": 1.1306,
+      "step": 5930
+    },
+    {
+      "epoch": 0.520099913815428,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0017601562647874378,
+      "loss": 1.134,
+      "step": 5931
+    },
+    {
+      "epoch": 0.5201876055898026,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0017597499339527252,
+      "loss": 1.163,
+      "step": 5932
+    },
+    {
+      "epoch": 0.5202752973641771,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001759343593109102,
+      "loss": 1.1614,
+      "step": 5933
+    },
+    {
+      "epoch": 0.5203629891385516,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0017589372422936252,
+      "loss": 1.1535,
+      "step": 5934
+    },
+    {
+      "epoch": 0.5204506809129262,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0017585308815433536,
+      "loss": 1.1652,
+      "step": 5935
+    },
+    {
+      "epoch": 0.5205383726873007,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0017581245108953454,
+      "loss": 1.1668,
+      "step": 5936
+    },
+    {
+      "epoch": 0.5206260644616753,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001757718130386661,
+      "loss": 1.1596,
+      "step": 5937
+    },
+    {
+      "epoch": 0.5207137562360498,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.001757311740054361,
+      "loss": 1.1838,
+      "step": 5938
+    },
+    {
+      "epoch": 0.5208014480104244,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0017569053399355075,
+      "loss": 1.25,
+      "step": 5939
+    },
+    {
+      "epoch": 0.5208891397847989,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001756498930067163,
+      "loss": 1.1483,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5209768315591734,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001756092510486391,
+      "loss": 1.1425,
+      "step": 5941
+    },
+    {
+      "epoch": 0.521064523333548,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0017556860812302558,
+      "loss": 1.1259,
+      "step": 5942
+    },
+    {
+      "epoch": 0.5211522151079225,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0017552796423358228,
+      "loss": 1.2174,
+      "step": 5943
+    },
+    {
+      "epoch": 0.521239906882297,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0017548731938401587,
+      "loss": 1.1255,
+      "step": 5944
+    },
+    {
+      "epoch": 0.5213275986566717,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0017544667357803296,
+      "loss": 1.2208,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5214152904310462,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0017540602681934037,
+      "loss": 1.1505,
+      "step": 5946
+    },
+    {
+      "epoch": 0.5215029822054207,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0017536537911164505,
+      "loss": 1.2064,
+      "step": 5947
+    },
+    {
+      "epoch": 0.5215906739797953,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0017532473045865386,
+      "loss": 1.1832,
+      "step": 5948
+    },
+    {
+      "epoch": 0.5216783657541698,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0017528408086407395,
+      "loss": 1.1826,
+      "step": 5949
+    },
+    {
+      "epoch": 0.5217660575285443,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001752434303316124,
+      "loss": 1.1855,
+      "step": 5950
+    },
+    {
+      "epoch": 0.521853749302919,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0017520277886497642,
+      "loss": 1.1448,
+      "step": 5951
+    },
+    {
+      "epoch": 0.5219414410772935,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017516212646787339,
+      "loss": 1.1677,
+      "step": 5952
+    },
+    {
+      "epoch": 0.522029132851668,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001751214731440106,
+      "loss": 1.1793,
+      "step": 5953
+    },
+    {
+      "epoch": 0.5221168246260426,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0017508081889709565,
+      "loss": 1.1917,
+      "step": 5954
+    },
+    {
+      "epoch": 0.5222045164004171,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00175040163730836,
+      "loss": 1.2077,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5222922081747916,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0017499950764893937,
+      "loss": 1.1763,
+      "step": 5956
+    },
+    {
+      "epoch": 0.5223798999491662,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0017495885065511342,
+      "loss": 1.1958,
+      "step": 5957
+    },
+    {
+      "epoch": 0.5224675917235407,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0017491819275306602,
+      "loss": 1.1691,
+      "step": 5958
+    },
+    {
+      "epoch": 0.5225552834979152,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0017487753394650498,
+      "loss": 1.1708,
+      "step": 5959
+    },
+    {
+      "epoch": 0.5226429752722898,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0017483687423913842,
+      "loss": 1.1886,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5227306670466644,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0017479621363467432,
+      "loss": 1.1713,
+      "step": 5961
+    },
+    {
+      "epoch": 0.5228183588210389,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0017475555213682084,
+      "loss": 1.1956,
+      "step": 5962
+    },
+    {
+      "epoch": 0.5229060505954134,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0017471488974928613,
+      "loss": 1.1315,
+      "step": 5963
+    },
+    {
+      "epoch": 0.522993742369788,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0017467422647577861,
+      "loss": 1.1548,
+      "step": 5964
+    },
+    {
+      "epoch": 0.5230814341441625,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0017463356232000657,
+      "loss": 1.1998,
+      "step": 5965
+    },
+    {
+      "epoch": 0.523169125918537,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0017459289728567853,
+      "loss": 1.1694,
+      "step": 5966
+    },
+    {
+      "epoch": 0.5232568176929117,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00174552231376503,
+      "loss": 1.1099,
+      "step": 5967
+    },
+    {
+      "epoch": 0.5233445094672862,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0017451156459618864,
+      "loss": 1.1725,
+      "step": 5968
+    },
+    {
+      "epoch": 0.5234322012416607,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001744708969484442,
+      "loss": 1.163,
+      "step": 5969
+    },
+    {
+      "epoch": 0.5235198930160353,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.001744302284369783,
+      "loss": 1.1783,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5236075847904098,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001743895590655,
+      "loss": 1.1826,
+      "step": 5971
+    },
+    {
+      "epoch": 0.5236952765647843,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.001743488888377181,
+      "loss": 1.2116,
+      "step": 5972
+    },
+    {
+      "epoch": 0.5237829683391589,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001743082177573417,
+      "loss": 1.1832,
+      "step": 5973
+    },
+    {
+      "epoch": 0.5238706601135334,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0017426754582807987,
+      "loss": 1.1156,
+      "step": 5974
+    },
+    {
+      "epoch": 0.523958351887908,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0017422687305364175,
+      "loss": 1.1881,
+      "step": 5975
+    },
+    {
+      "epoch": 0.5240460436622826,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0017418619943773664,
+      "loss": 1.1664,
+      "step": 5976
+    },
+    {
+      "epoch": 0.5241337354366571,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0017414552498407389,
+      "loss": 1.2041,
+      "step": 5977
+    },
+    {
+      "epoch": 0.5242214272110316,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0017410484969636287,
+      "loss": 1.1877,
+      "step": 5978
+    },
+    {
+      "epoch": 0.5243091189854062,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017406417357831304,
+      "loss": 1.087,
+      "step": 5979
+    },
+    {
+      "epoch": 0.5243968107597807,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0017402349663363395,
+      "loss": 1.1674,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5244845025341552,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0017398281886603529,
+      "loss": 1.1599,
+      "step": 5981
+    },
+    {
+      "epoch": 0.5245721943085299,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0017394214027922675,
+      "loss": 1.2198,
+      "step": 5982
+    },
+    {
+      "epoch": 0.5246598860829044,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0017390146087691808,
+      "loss": 1.1682,
+      "step": 5983
+    },
+    {
+      "epoch": 0.5247475778572789,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0017386078066281916,
+      "loss": 1.2219,
+      "step": 5984
+    },
+    {
+      "epoch": 0.5248352696316534,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001738200996406399,
+      "loss": 1.1836,
+      "step": 5985
+    },
+    {
+      "epoch": 0.524922961406028,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.001737794178140904,
+      "loss": 1.1525,
+      "step": 5986
+    },
+    {
+      "epoch": 0.5250106531804025,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0017373873518688058,
+      "loss": 1.189,
+      "step": 5987
+    },
+    {
+      "epoch": 0.525098344954777,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0017369805176272069,
+      "loss": 1.2234,
+      "step": 5988
+    },
+    {
+      "epoch": 0.5251860367291517,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0017365736754532096,
+      "loss": 1.1839,
+      "step": 5989
+    },
+    {
+      "epoch": 0.5252737285035262,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001736166825383916,
+      "loss": 1.1544,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5253614202779007,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0017357599674564309,
+      "loss": 1.1166,
+      "step": 5991
+    },
+    {
+      "epoch": 0.5254491120522753,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.001735353101707858,
+      "loss": 1.1344,
+      "step": 5992
+    },
+    {
+      "epoch": 0.5255368038266498,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0017349462281753022,
+      "loss": 1.1934,
+      "step": 5993
+    },
+    {
+      "epoch": 0.5256244956010243,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0017345393468958703,
+      "loss": 1.1764,
+      "step": 5994
+    },
+    {
+      "epoch": 0.5257121873753989,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0017341324579066683,
+      "loss": 1.1373,
+      "step": 5995
+    },
+    {
+      "epoch": 0.5257998791497734,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0017337255612448023,
+      "loss": 1.156,
+      "step": 5996
+    },
+    {
+      "epoch": 0.525887570924148,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0017333186569473823,
+      "loss": 1.1707,
+      "step": 5997
+    },
+    {
+      "epoch": 0.5259752626985226,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0017329117450515154,
+      "loss": 1.1315,
+      "step": 5998
+    },
+    {
+      "epoch": 0.5260629544728971,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0017325048255943112,
+      "loss": 1.1613,
+      "step": 5999
+    },
+    {
+      "epoch": 0.5261506462472716,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0017320978986128802,
+      "loss": 1.1743,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5261506462472716,
+      "eval_loss": 1.1750010251998901,
+      "eval_runtime": 428.5284,
+      "eval_samples_per_second": 33.713,
+      "eval_steps_per_second": 8.429,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5262383380216462,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0017316909641443329,
+      "loss": 1.1835,
+      "step": 6001
+    },
+    {
+      "epoch": 0.5263260297960207,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0017312840222257803,
+      "loss": 1.1314,
+      "step": 6002
+    },
+    {
+      "epoch": 0.5264137215703952,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001730877072894335,
+      "loss": 1.1287,
+      "step": 6003
+    },
+    {
+      "epoch": 0.5265014133447699,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001730470116187109,
+      "loss": 1.1468,
+      "step": 6004
+    },
+    {
+      "epoch": 0.5265891051191444,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0017300631521412165,
+      "loss": 1.1755,
+      "step": 6005
+    },
+    {
+      "epoch": 0.5266767968935189,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0017296561807937708,
+      "loss": 1.1622,
+      "step": 6006
+    },
+    {
+      "epoch": 0.5267644886678935,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001729249202181887,
+      "loss": 1.2125,
+      "step": 6007
+    },
+    {
+      "epoch": 0.526852180442268,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0017288422163426814,
+      "loss": 1.1493,
+      "step": 6008
+    },
+    {
+      "epoch": 0.5269398722166425,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0017284352233132682,
+      "loss": 1.2026,
+      "step": 6009
+    },
+    {
+      "epoch": 0.527027563991017,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001728028223130766,
+      "loss": 1.1812,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5271152557653916,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001727621215832291,
+      "loss": 1.1587,
+      "step": 6011
+    },
+    {
+      "epoch": 0.5272029475397662,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001727214201454961,
+      "loss": 1.1317,
+      "step": 6012
+    },
+    {
+      "epoch": 0.5272906393141407,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0017268071800358956,
+      "loss": 1.1448,
+      "step": 6013
+    },
+    {
+      "epoch": 0.5273783310885153,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0017264001516122136,
+      "loss": 1.2169,
+      "step": 6014
+    },
+    {
+      "epoch": 0.5274660228628898,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0017259931162210349,
+      "loss": 1.1486,
+      "step": 6015
+    },
+    {
+      "epoch": 0.5275537146372643,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0017255860738994802,
+      "loss": 1.1254,
+      "step": 6016
+    },
+    {
+      "epoch": 0.5276414064116389,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0017251790246846706,
+      "loss": 1.1978,
+      "step": 6017
+    },
+    {
+      "epoch": 0.5277290981860134,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.001724771968613728,
+      "loss": 1.1444,
+      "step": 6018
+    },
+    {
+      "epoch": 0.5278167899603879,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0017243649057237752,
+      "loss": 1.2183,
+      "step": 6019
+    },
+    {
+      "epoch": 0.5279044817347626,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001723957836051935,
+      "loss": 1.1968,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5279921735091371,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0017235507596353307,
+      "loss": 1.1243,
+      "step": 6021
+    },
+    {
+      "epoch": 0.5280798652835116,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0017231436765110868,
+      "loss": 1.1691,
+      "step": 6022
+    },
+    {
+      "epoch": 0.5281675570578862,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.001722736586716329,
+      "loss": 1.1984,
+      "step": 6023
+    },
+    {
+      "epoch": 0.5282552488322607,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0017223294902881819,
+      "loss": 1.11,
+      "step": 6024
+    },
+    {
+      "epoch": 0.5283429406066352,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0017219223872637721,
+      "loss": 1.1576,
+      "step": 6025
+    },
+    {
+      "epoch": 0.5284306323810098,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0017215152776802265,
+      "loss": 1.1859,
+      "step": 6026
+    },
+    {
+      "epoch": 0.5285183241553844,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0017211081615746717,
+      "loss": 1.1828,
+      "step": 6027
+    },
+    {
+      "epoch": 0.5286060159297589,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0017207010389842367,
+      "loss": 1.1305,
+      "step": 6028
+    },
+    {
+      "epoch": 0.5286937077041335,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0017202939099460492,
+      "loss": 1.1895,
+      "step": 6029
+    },
+    {
+      "epoch": 0.528781399478508,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0017198867744972386,
+      "loss": 1.1717,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5288690912528825,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0017194796326749342,
+      "loss": 1.2066,
+      "step": 6031
+    },
+    {
+      "epoch": 0.528956783027257,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001719072484516267,
+      "loss": 1.1947,
+      "step": 6032
+    },
+    {
+      "epoch": 0.5290444748016316,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0017186653300583677,
+      "loss": 1.1564,
+      "step": 6033
+    },
+    {
+      "epoch": 0.5291321665760061,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0017182581693383672,
+      "loss": 1.1513,
+      "step": 6034
+    },
+    {
+      "epoch": 0.5292198583503807,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001717851002393398,
+      "loss": 1.1876,
+      "step": 6035
+    },
+    {
+      "epoch": 0.5293075501247553,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0017174438292605922,
+      "loss": 1.1827,
+      "step": 6036
+    },
+    {
+      "epoch": 0.5293952418991298,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0017170366499770842,
+      "loss": 1.2665,
+      "step": 6037
+    },
+    {
+      "epoch": 0.5294829336735043,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0017166294645800063,
+      "loss": 1.1667,
+      "step": 6038
+    },
+    {
+      "epoch": 0.5295706254478789,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001716222273106493,
+      "loss": 1.1468,
+      "step": 6039
+    },
+    {
+      "epoch": 0.5296583172222534,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0017158150755936794,
+      "loss": 1.1207,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5297460089966279,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017154078720787004,
+      "loss": 1.1247,
+      "step": 6041
+    },
+    {
+      "epoch": 0.5298337007710026,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0017150006625986933,
+      "loss": 1.113,
+      "step": 6042
+    },
+    {
+      "epoch": 0.5299213925453771,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0017145934471907932,
+      "loss": 1.1456,
+      "step": 6043
+    },
+    {
+      "epoch": 0.5300090843197516,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0017141862258921372,
+      "loss": 1.1194,
+      "step": 6044
+    },
+    {
+      "epoch": 0.5300967760941262,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0017137789987398632,
+      "loss": 1.1798,
+      "step": 6045
+    },
+    {
+      "epoch": 0.5301844678685007,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0017133717657711097,
+      "loss": 1.1831,
+      "step": 6046
+    },
+    {
+      "epoch": 0.5302721596428752,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0017129645270230138,
+      "loss": 1.1408,
+      "step": 6047
+    },
+    {
+      "epoch": 0.5303598514172498,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0017125572825327162,
+      "loss": 1.1667,
+      "step": 6048
+    },
+    {
+      "epoch": 0.5304475431916243,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0017121500323373558,
+      "loss": 1.1412,
+      "step": 6049
+    },
+    {
+      "epoch": 0.5305352349659989,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001711742776474073,
+      "loss": 1.0961,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5306229267403735,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0017113355149800077,
+      "loss": 1.1647,
+      "step": 6051
+    },
+    {
+      "epoch": 0.530710618514748,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0017109282478923024,
+      "loss": 1.1693,
+      "step": 6052
+    },
+    {
+      "epoch": 0.5307983102891225,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0017105209752480979,
+      "loss": 1.1893,
+      "step": 6053
+    },
+    {
+      "epoch": 0.5308860020634971,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001710113697084537,
+      "loss": 1.1365,
+      "step": 6054
+    },
+    {
+      "epoch": 0.5309736938378716,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.001709706413438762,
+      "loss": 1.1853,
+      "step": 6055
+    },
+    {
+      "epoch": 0.5310613856122461,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0017092991243479161,
+      "loss": 1.1712,
+      "step": 6056
+    },
+    {
+      "epoch": 0.5311490773866206,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0017088918298491435,
+      "loss": 1.1231,
+      "step": 6057
+    },
+    {
+      "epoch": 0.5312367691609953,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017084845299795875,
+      "loss": 1.2111,
+      "step": 6058
+    },
+    {
+      "epoch": 0.5313244609353698,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0017080772247763938,
+      "loss": 1.205,
+      "step": 6059
+    },
+    {
+      "epoch": 0.5314121527097443,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0017076699142767068,
+      "loss": 1.1468,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5314998444841189,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0017072625985176727,
+      "loss": 1.0453,
+      "step": 6061
+    },
+    {
+      "epoch": 0.5315875362584934,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0017068552775364374,
+      "loss": 1.223,
+      "step": 6062
+    },
+    {
+      "epoch": 0.5316752280328679,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0017064479513701474,
+      "loss": 1.1231,
+      "step": 6063
+    },
+    {
+      "epoch": 0.5317629198072426,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0017060406200559502,
+      "loss": 1.1232,
+      "step": 6064
+    },
+    {
+      "epoch": 0.5318506115816171,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.001705633283630993,
+      "loss": 1.1985,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5319383033559916,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001705225942132424,
+      "loss": 1.1715,
+      "step": 6066
+    },
+    {
+      "epoch": 0.5320259951303662,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0017048185955973914,
+      "loss": 1.236,
+      "step": 6067
+    },
+    {
+      "epoch": 0.5321136869047407,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0017044112440630448,
+      "loss": 1.0972,
+      "step": 6068
+    },
+    {
+      "epoch": 0.5322013786791152,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0017040038875665328,
+      "loss": 1.1446,
+      "step": 6069
+    },
+    {
+      "epoch": 0.5322890704534898,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001703596526145006,
+      "loss": 1.1626,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5323767622278643,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0017031891598356145,
+      "loss": 1.1777,
+      "step": 6071
+    },
+    {
+      "epoch": 0.5324644540022389,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0017027817886755086,
+      "loss": 1.1236,
+      "step": 6072
+    },
+    {
+      "epoch": 0.5325521457766135,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0017023744127018397,
+      "loss": 1.1375,
+      "step": 6073
+    },
+    {
+      "epoch": 0.532639837550988,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0017019670319517598,
+      "loss": 1.1734,
+      "step": 6074
+    },
+    {
+      "epoch": 0.5327275293253625,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0017015596464624208,
+      "loss": 1.2084,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5328152210997371,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001701152256270975,
+      "loss": 1.1874,
+      "step": 6076
+    },
+    {
+      "epoch": 0.5329029128741116,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0017007448614145754,
+      "loss": 1.1296,
+      "step": 6077
+    },
+    {
+      "epoch": 0.5329906046484861,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0017003374619303754,
+      "loss": 1.1186,
+      "step": 6078
+    },
+    {
+      "epoch": 0.5330782964228606,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0016999300578555295,
+      "loss": 1.1818,
+      "step": 6079
+    },
+    {
+      "epoch": 0.5331659881972353,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0016995226492271904,
+      "loss": 1.179,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5332536799716098,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0016991152360825137,
+      "loss": 1.2066,
+      "step": 6081
+    },
+    {
+      "epoch": 0.5333413717459843,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0016987078184586546,
+      "loss": 1.2245,
+      "step": 6082
+    },
+    {
+      "epoch": 0.5334290635203589,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0016983003963927678,
+      "loss": 1.1402,
+      "step": 6083
+    },
+    {
+      "epoch": 0.5335167552947334,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0016978929699220098,
+      "loss": 1.1463,
+      "step": 6084
+    },
+    {
+      "epoch": 0.5336044470691079,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0016974855390835364,
+      "loss": 1.1086,
+      "step": 6085
+    },
+    {
+      "epoch": 0.5336921388434825,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0016970781039145045,
+      "loss": 1.2318,
+      "step": 6086
+    },
+    {
+      "epoch": 0.533779830617857,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016966706644520708,
+      "loss": 1.1516,
+      "step": 6087
+    },
+    {
+      "epoch": 0.5338675223922316,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0016962632207333937,
+      "loss": 1.1711,
+      "step": 6088
+    },
+    {
+      "epoch": 0.5339552141666062,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0016958557727956296,
+      "loss": 1.1232,
+      "step": 6089
+    },
+    {
+      "epoch": 0.5340429059409807,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001695448320675938,
+      "loss": 1.2053,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5341305977153552,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0016950408644114765,
+      "loss": 1.2034,
+      "step": 6091
+    },
+    {
+      "epoch": 0.5342182894897298,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016946334040394044,
+      "loss": 1.1347,
+      "step": 6092
+    },
+    {
+      "epoch": 0.5343059812641043,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0016942259395968811,
+      "loss": 1.1686,
+      "step": 6093
+    },
+    {
+      "epoch": 0.5343936730384788,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001693818471121066,
+      "loss": 1.1984,
+      "step": 6094
+    },
+    {
+      "epoch": 0.5344813648128535,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0016934109986491198,
+      "loss": 1.1266,
+      "step": 6095
+    },
+    {
+      "epoch": 0.534569056587228,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001693003522218203,
+      "loss": 1.116,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5346567483616025,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016925960418654753,
+      "loss": 1.168,
+      "step": 6097
+    },
+    {
+      "epoch": 0.5347444401359771,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0016921885576280985,
+      "loss": 1.1856,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5348321319103516,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0016917810695432347,
+      "loss": 1.1623,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5349198236847261,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001691373577648045,
+      "loss": 1.1376,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5350075154591007,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0016909660819796916,
+      "loss": 1.1783,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5350952072334753,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016905585825753374,
+      "loss": 1.1521,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5351828990078498,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0016901510794721454,
+      "loss": 1.2314,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5352705907822243,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0016897435727072781,
+      "loss": 1.1634,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5353582825565989,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0016893360623179005,
+      "loss": 1.1481,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5354459743309734,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0016889285483411747,
+      "loss": 1.1689,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5355336661053479,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0016885210308142658,
+      "loss": 1.1625,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5356213578797225,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0016881135097743388,
+      "loss": 1.1612,
+      "step": 6108
+    },
+    {
+      "epoch": 0.535709049654097,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0016877059852585577,
+      "loss": 1.158,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5357967414284716,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0016872984573040883,
+      "loss": 1.1712,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5358844332028462,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0016868909259480955,
+      "loss": 1.1936,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5359721249772207,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001686483391227746,
+      "loss": 1.2165,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5360598167515952,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0016860758531802055,
+      "loss": 1.1493,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5361475085259698,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0016856683118426398,
+      "loss": 1.1423,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5362352003003443,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0016852607672522165,
+      "loss": 1.1809,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5363228920747188,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0016848532194461017,
+      "loss": 1.1543,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5364105838490935,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0016844456684614638,
+      "loss": 1.1536,
+      "step": 6117
+    },
+    {
+      "epoch": 0.536498275623468,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0016840381143354696,
+      "loss": 1.175,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5365859673978425,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0016836305571052879,
+      "loss": 1.1169,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5366736591722171,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0016832229968080854,
+      "loss": 1.1344,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5367613509465916,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016828154334810317,
+      "loss": 1.1725,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5368490427209661,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0016824078671612954,
+      "loss": 1.1193,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5369367344953407,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0016820002978860445,
+      "loss": 1.1892,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5370244262697152,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0016815927256924501,
+      "loss": 1.1075,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5371121180440898,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016811851506176795,
+      "loss": 1.1074,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5371998098184644,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0016807775726989046,
+      "loss": 1.1119,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5372875015928389,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001680369991973294,
+      "loss": 1.1076,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5373751933672134,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001679962408478019,
+      "loss": 1.208,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5374628851415879,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001679554822250249,
+      "loss": 1.1426,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5375505769159625,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0016791472333271564,
+      "loss": 1.1756,
+      "step": 6130
+    },
+    {
+      "epoch": 0.537638268690337,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0016787396417459108,
+      "loss": 1.1522,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5377259604647115,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.001678332047543684,
+      "loss": 1.125,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5378136522390862,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0016779244507576477,
+      "loss": 1.1209,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5379013440134607,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0016775168514249735,
+      "loss": 1.1574,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5379890357878352,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0016771092495828338,
+      "loss": 1.1463,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5380767275622098,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0016767016452684,
+      "loss": 1.1986,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5381644193365843,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001676294038518846,
+      "loss": 1.2035,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5382521111109588,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0016758864293713429,
+      "loss": 1.2038,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5383398028853335,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0016754788178630646,
+      "loss": 1.1651,
+      "step": 6139
+    },
+    {
+      "epoch": 0.538427494659708,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0016750712040311832,
+      "loss": 1.1258,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5385151864340825,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0016746635879128734,
+      "loss": 1.1763,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5386028782084571,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0016742559695453075,
+      "loss": 1.1739,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5386905699828316,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0016738483489656603,
+      "loss": 1.1347,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5387782617572061,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016734407262111047,
+      "loss": 1.13,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5388659535315807,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0016730331013188162,
+      "loss": 1.1444,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5389536453059552,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0016726254743259677,
+      "loss": 1.1331,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5390413370803298,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0016722178452697343,
+      "loss": 1.1253,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5391290288547044,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0016718102141872913,
+      "loss": 1.1512,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5392167206290789,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0016714025811158126,
+      "loss": 1.087,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5393044124034534,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.001670994946092474,
+      "loss": 1.1883,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5393921041778279,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0016705873091544502,
+      "loss": 1.0774,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5394797959522025,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0016701796703389173,
+      "loss": 1.1376,
+      "step": 6152
+    },
+    {
+      "epoch": 0.539567487726577,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00166977202968305,
+      "loss": 1.1503,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5396551795009515,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0016693643872240252,
+      "loss": 1.1777,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5397428712753262,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0016689567429990185,
+      "loss": 1.2055,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5398305630497007,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0016685490970452053,
+      "loss": 1.171,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5399182548240752,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0016681414493997624,
+      "loss": 1.1589,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5400059465984498,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.001667733800099866,
+      "loss": 1.1231,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5400936383728243,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0016673261491826931,
+      "loss": 1.1593,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5401813301471988,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016669184966854197,
+      "loss": 1.1881,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5402690219215734,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016665108426452237,
+      "loss": 1.1642,
+      "step": 6161
+    },
+    {
+      "epoch": 0.540356713695948,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0016661031870992809,
+      "loss": 1.1686,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5404444054703225,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016656955300847695,
+      "loss": 1.1913,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5405320972446971,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001665287871638866,
+      "loss": 1.1588,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5406197890190716,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.001664880211798748,
+      "loss": 1.1768,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5407074807934461,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001664472550601593,
+      "loss": 1.2207,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5407951725678207,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0016640648880845792,
+      "loss": 1.1983,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5408828643421952,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0016636572242848837,
+      "loss": 1.1084,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5409705561165697,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0016632495592396842,
+      "loss": 1.1205,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5410582478909444,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0016628418929861596,
+      "loss": 1.2338,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5411459396653189,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.001662434225561487,
+      "loss": 1.1588,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5412336314396934,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0016620265570028462,
+      "loss": 1.182,
+      "step": 6172
+    },
+    {
+      "epoch": 0.541321323214068,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016616188873474133,
+      "loss": 1.1188,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5414090149884425,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0016612112166323688,
+      "loss": 1.1278,
+      "step": 6174
+    },
+    {
+      "epoch": 0.541496706762817,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0016608035448948896,
+      "loss": 1.188,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5415843985371915,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001660395872172155,
+      "loss": 1.1424,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5416720903115662,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.001659988198501344,
+      "loss": 1.1199,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5417597820859407,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0016595805239196346,
+      "loss": 1.167,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5418474738603152,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0016591728484642069,
+      "loss": 1.2168,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5419351656346898,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0016587651721722385,
+      "loss": 1.1293,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5420228574090643,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0016583574950809095,
+      "loss": 1.1482,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5421105491834388,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0016579498172273983,
+      "loss": 1.1544,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5421982409578134,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0016575421386488842,
+      "loss": 1.1129,
+      "step": 6183
+    },
+    {
+      "epoch": 0.542285932732188,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0016571344593825465,
+      "loss": 1.1664,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5423736245065625,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0016567267794655649,
+      "loss": 1.1701,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5424613162809371,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0016563190989351175,
+      "loss": 1.1784,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5425490080553116,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0016559114178283853,
+      "loss": 1.185,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5426366998296861,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001655503736182547,
+      "loss": 1.1976,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5427243916040607,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0016550960540347823,
+      "loss": 1.1419,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5428120833784352,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0016546883714222706,
+      "loss": 1.2035,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5428997751528097,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0016542806883821913,
+      "loss": 1.1298,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5429874669271844,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0016538730049517243,
+      "loss": 1.159,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5430751587015589,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001653465321168049,
+      "loss": 1.1406,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5431628504759334,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0016530576370683459,
+      "loss": 1.2321,
+      "step": 6194
+    },
+    {
+      "epoch": 0.543250542250308,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0016526499526897935,
+      "loss": 1.1539,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5433382340246825,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001652242268069573,
+      "loss": 1.1746,
+      "step": 6196
+    },
+    {
+      "epoch": 0.543425925799057,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0016518345832448628,
+      "loss": 1.1594,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5435136175734316,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0016514268982528442,
+      "loss": 1.1595,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5436013093478061,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.001651019213130695,
+      "loss": 1.1401,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5436890011221807,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0016506115279155968,
+      "loss": 1.1848,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5437766928965552,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0016502038426447292,
+      "loss": 1.1586,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5438643846709298,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0016497961573552712,
+      "loss": 1.1362,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5439520764453043,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0016493884720844034,
+      "loss": 1.1869,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5440397682196788,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.001648980786869305,
+      "loss": 1.1318,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5441274599940534,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0016485731017471567,
+      "loss": 1.2033,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5442151517684279,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016481654167551372,
+      "loss": 1.1822,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5443028435428025,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0016477577319304274,
+      "loss": 1.1775,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5443905353171771,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0016473500473102062,
+      "loss": 1.1539,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5444782270915516,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0016469423629316545,
+      "loss": 1.1698,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5445659188659261,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0016465346788319516,
+      "loss": 1.1285,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5446536106403007,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0016461269950482763,
+      "loss": 1.1193,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5447413024146752,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016457193116178096,
+      "loss": 1.1623,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5448289941890497,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00164531162857773,
+      "loss": 1.1908,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5449166859634244,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0016449039459652183,
+      "loss": 1.1165,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5450043777377989,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0016444962638174534,
+      "loss": 1.1602,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5450920695121734,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0016440885821716147,
+      "loss": 1.1961,
+      "step": 6216
+    },
+    {
+      "epoch": 0.545179761286548,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0016436809010648824,
+      "loss": 1.1823,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5452674530609225,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0016432732205344355,
+      "loss": 1.1681,
+      "step": 6218
+    },
+    {
+      "epoch": 0.545355144835297,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0016428655406174539,
+      "loss": 1.1674,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5454428366096716,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.001642457861351116,
+      "loss": 1.1003,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5455305283840461,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001642050182772602,
+      "loss": 1.1554,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5456182201584207,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0016416425049190905,
+      "loss": 1.1374,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5457059119327952,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0016412348278277615,
+      "loss": 1.1634,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5457936037071698,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0016408271515357931,
+      "loss": 1.1173,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5458812954815443,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0016404194760803654,
+      "loss": 1.1473,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5459689872559188,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016400118014986559,
+      "loss": 1.0985,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5460566790302934,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0016396041278278455,
+      "loss": 1.1385,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5461443708046679,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001639196455105111,
+      "loss": 1.1643,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5462320625790424,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0016387887833676318,
+      "loss": 1.1503,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5463197543534171,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.001638381112652587,
+      "loss": 1.1416,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5464074461277916,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0016379734429971544,
+      "loss": 1.1623,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5464951379021661,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001637565774438513,
+      "loss": 1.1652,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5465828296765407,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0016371581070138408,
+      "loss": 1.1798,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5466705214509152,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0016367504407603162,
+      "loss": 1.1315,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5467582132252897,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0016363427757151168,
+      "loss": 1.21,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5468459049996643,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0016359351119154216,
+      "loss": 1.1493,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5469335967740389,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.001635527449398407,
+      "loss": 1.1037,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5470212885484134,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0016351197882012523,
+      "loss": 1.1149,
+      "step": 6238
+    },
+    {
+      "epoch": 0.547108980322788,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0016347121283611342,
+      "loss": 1.2202,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5471966720971625,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0016343044699152311,
+      "loss": 1.1511,
+      "step": 6240
+    },
+    {
+      "epoch": 0.547284363871537,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001633896812900719,
+      "loss": 1.12,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5473720556459116,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0016334891573547762,
+      "loss": 1.145,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5474597474202861,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0016330815033145803,
+      "loss": 1.2593,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5475474391946606,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0016326738508173075,
+      "loss": 1.2497,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5476351309690353,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0016322661999001345,
+      "loss": 1.2126,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5477228227434098,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0016318585506002383,
+      "loss": 1.1608,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5478105145177843,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0016314509029547951,
+      "loss": 1.1759,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5478982062921588,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0016310432570009823,
+      "loss": 1.1361,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5479858980665334,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0016306356127759752,
+      "loss": 1.1059,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5480735898409079,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0016302279703169502,
+      "loss": 1.1806,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5481612816152824,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016298203296610829,
+      "loss": 1.1273,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5482489733896571,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0016294126908455503,
+      "loss": 1.1676,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5483366651640316,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0016290050539075262,
+      "loss": 1.1975,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5484243569384061,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016285974188841878,
+      "loss": 1.1697,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5485120487127807,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.001628189785812709,
+      "loss": 1.163,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5485997404871552,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0016277821547302656,
+      "loss": 1.1997,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5486874322615297,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0016273745256740323,
+      "loss": 1.1365,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5487751240359043,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001626966898681184,
+      "loss": 1.1422,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5488628158102788,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0016265592737888953,
+      "loss": 1.1426,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5489505075846534,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0016261516510343397,
+      "loss": 1.1918,
+      "step": 6260
+    },
+    {
+      "epoch": 0.549038199359028,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0016257440304546927,
+      "loss": 1.1411,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5491258911334025,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0016253364120871275,
+      "loss": 1.1349,
+      "step": 6262
+    },
+    {
+      "epoch": 0.549213582907777,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0016249287959688174,
+      "loss": 1.1885,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5493012746821516,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.001624521182136936,
+      "loss": 1.1467,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5493889664565261,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0016241135706286578,
+      "loss": 1.1815,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5494766582309006,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0016237059614811545,
+      "loss": 1.139,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5495643500052753,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0016232983547316002,
+      "loss": 1.1717,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5496520417796498,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0016228907504171664,
+      "loss": 1.1009,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5497397335540243,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0016224831485750267,
+      "loss": 1.1357,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5498274253283988,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0016220755492423527,
+      "loss": 1.1442,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5499151171027734,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001621667952456316,
+      "loss": 1.1538,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5500028088771479,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0016212603582540894,
+      "loss": 1.1771,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5500905006515224,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0016208527666728438,
+      "loss": 1.1882,
+      "step": 6273
+    },
+    {
+      "epoch": 0.550178192425897,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001620445177749751,
+      "loss": 1.135,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5502658842002716,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.001620037591521981,
+      "loss": 1.1446,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5503535759746461,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001619630008026706,
+      "loss": 1.214,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5504412677490207,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0016192224273010956,
+      "loss": 1.1768,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5505289595233952,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0016188148493823207,
+      "loss": 1.2054,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5506166512977697,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.001618407274307551,
+      "loss": 1.1467,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5507043430721443,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016179997021139555,
+      "loss": 1.1381,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5507920348465188,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0016175921328387053,
+      "loss": 1.1963,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5508797266208934,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0016171845665189687,
+      "loss": 1.1804,
+      "step": 6282
+    },
+    {
+      "epoch": 0.550967418395268,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001616777003191915,
+      "loss": 1.1913,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5510551101696425,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0016163694428947128,
+      "loss": 1.1054,
+      "step": 6284
+    },
+    {
+      "epoch": 0.551142801944017,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0016159618856645306,
+      "loss": 1.1939,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5512304937183916,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0016155543315385362,
+      "loss": 1.1101,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5513181854927661,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016151467805538982,
+      "loss": 1.1556,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5514058772671406,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001614739232747784,
+      "loss": 1.1463,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5514935690415153,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0016143316881573606,
+      "loss": 1.1454,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5515812608158898,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001613924146819795,
+      "loss": 1.1797,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5516689525902643,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0016135166087722541,
+      "loss": 1.1276,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5517566443646389,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0016131090740519045,
+      "loss": 1.1592,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5518443361390134,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0016127015426959119,
+      "loss": 1.1215,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5519320279133879,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0016122940147414422,
+      "loss": 1.1942,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5520197196877624,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016118864902256618,
+      "loss": 1.255,
+      "step": 6295
+    },
+    {
+      "epoch": 0.552107411462137,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0016114789691857347,
+      "loss": 1.2133,
+      "step": 6296
+    },
+    {
+      "epoch": 0.5521951032365116,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0016110714516588257,
+      "loss": 1.1311,
+      "step": 6297
+    },
+    {
+      "epoch": 0.5522827950108861,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0016106639376821004,
+      "loss": 1.173,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5523704867852607,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0016102564272927223,
+      "loss": 1.1729,
+      "step": 6299
+    },
+    {
+      "epoch": 0.5524581785596352,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001609848920527855,
+      "loss": 1.1597,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5525458703340097,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016094414174246628,
+      "loss": 1.2032,
+      "step": 6301
+    },
+    {
+      "epoch": 0.5526335621083843,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0016090339180203084,
+      "loss": 1.2268,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5527212538827588,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0016086264223519554,
+      "loss": 1.1406,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5528089456571333,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0016082189304567657,
+      "loss": 1.1744,
+      "step": 6304
+    },
+    {
+      "epoch": 0.552896637431508,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0016078114423719015,
+      "loss": 1.1374,
+      "step": 6305
+    },
+    {
+      "epoch": 0.5529843292058825,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001607403958134525,
+      "loss": 1.1721,
+      "step": 6306
+    },
+    {
+      "epoch": 0.553072020980257,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0016069964777817975,
+      "loss": 1.1685,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5531597127546316,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0016065890013508802,
+      "loss": 1.1869,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5532474045290061,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001606181528878934,
+      "loss": 1.1795,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5533350963033806,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0016057740604031193,
+      "loss": 1.2014,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5534227880777552,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0016053665959605958,
+      "loss": 1.1581,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5535104798521298,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0016049591355885246,
+      "loss": 1.1228,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5535981716265043,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0016045516793240626,
+      "loss": 1.1614,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5536858634008789,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001604144227204371,
+      "loss": 1.1347,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5537735551752534,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001603736779266607,
+      "loss": 1.1603,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5538612469496279,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0016033293355479294,
+      "loss": 1.1814,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5539489387240025,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0016029218960854957,
+      "loss": 1.1171,
+      "step": 6317
+    },
+    {
+      "epoch": 0.554036630498377,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.001602514460916464,
+      "loss": 1.116,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5541243222727515,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0016021070300779906,
+      "loss": 1.1405,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5542120140471261,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0016016996036072326,
+      "loss": 1.1298,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5542997058215007,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0016012921815413458,
+      "loss": 1.1875,
+      "step": 6321
+    },
+    {
+      "epoch": 0.5543873975958752,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0016008847639174867,
+      "loss": 1.1621,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5544750893702497,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0016004773507728098,
+      "loss": 1.1173,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5545627811446243,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0016000699421444711,
+      "loss": 1.1682,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5546504729189988,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0015996625380696246,
+      "loss": 1.1739,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5547381646933733,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0015992551385854246,
+      "loss": 1.1476,
+      "step": 6326
+    },
+    {
+      "epoch": 0.554825856467748,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.001598847743729025,
+      "loss": 1.1328,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5549135482421225,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0015984403535375792,
+      "loss": 1.2484,
+      "step": 6328
+    },
+    {
+      "epoch": 0.555001240016497,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0015980329680482406,
+      "loss": 1.1521,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5550889317908716,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.001597625587298161,
+      "loss": 1.1866,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5551766235652461,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.001597218211324492,
+      "loss": 1.1049,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5552643153396206,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0015968108401643861,
+      "loss": 1.1274,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5553520071139952,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001596403473854994,
+      "loss": 1.1283,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5554396988883697,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0015959961124334674,
+      "loss": 1.1537,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5555273906627443,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0015955887559369556,
+      "loss": 1.0928,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5556150824371189,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0015951814044026086,
+      "loss": 1.1594,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5557027742114934,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0015947740578675764,
+      "loss": 1.1512,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5557904659858679,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0015943667163690074,
+      "loss": 1.1504,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5558781577602425,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0015939593799440502,
+      "loss": 1.1153,
+      "step": 6339
+    },
+    {
+      "epoch": 0.555965849534617,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0015935520486298528,
+      "loss": 1.1693,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5560535413089915,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0015931447224635628,
+      "loss": 1.2001,
+      "step": 6341
+    },
+    {
+      "epoch": 0.556141233083366,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0015927374014823273,
+      "loss": 1.1423,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5562289248577407,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0015923300857232931,
+      "loss": 1.1328,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5563166166321152,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0015919227752236066,
+      "loss": 1.1577,
+      "step": 6344
+    },
+    {
+      "epoch": 0.5564043084064897,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0015915154700204123,
+      "loss": 1.1518,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5564920001808643,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001591108170150857,
+      "loss": 1.1548,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5565796919552388,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0015907008756520843,
+      "loss": 1.2151,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5566673837296133,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0015902935865612384,
+      "loss": 1.1432,
+      "step": 6348
+    },
+    {
+      "epoch": 0.556755075503988,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0015898863029154635,
+      "loss": 1.081,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5568427672783625,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0015894790247519021,
+      "loss": 1.1976,
+      "step": 6350
+    },
+    {
+      "epoch": 0.556930459052737,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0015890717521076976,
+      "loss": 1.2402,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5570181508271116,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0015886644850199923,
+      "loss": 1.1517,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5571058426014861,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0015882572235259276,
+      "loss": 1.1267,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5571935343758606,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0015878499676626448,
+      "loss": 1.249,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5572812261502352,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0015874427174672843,
+      "loss": 1.1747,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5573689179246097,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0015870354729769867,
+      "loss": 1.1809,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5574566096989843,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001586628234228891,
+      "loss": 1.1567,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5575443014733589,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0015862210012601368,
+      "loss": 1.1509,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5576319932477334,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0015858137741078628,
+      "loss": 1.1619,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5577196850221079,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0015854065528092072,
+      "loss": 1.1199,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5578073767964825,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0015849993374013067,
+      "loss": 1.1476,
+      "step": 6361
+    },
+    {
+      "epoch": 0.557895068570857,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001584592127921299,
+      "loss": 1.181,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5579827603452315,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0015841849244063214,
+      "loss": 1.1602,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5580704521196062,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0015837777268935076,
+      "loss": 1.1877,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5581581438939807,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0015833705354199943,
+      "loss": 1.1736,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5582458356683552,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0015829633500229164,
+      "loss": 1.1355,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5583335274427297,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0015825561707394078,
+      "loss": 1.1769,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5584212192171043,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0015821489976066024,
+      "loss": 1.1726,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5585089109914788,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0015817418306616332,
+      "loss": 1.157,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5585966027658533,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0015813346699416328,
+      "loss": 1.1732,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5586842945402279,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0015809275154837332,
+      "loss": 1.1495,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5587719863146025,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001580520367325066,
+      "loss": 1.2056,
+      "step": 6372
+    },
+    {
+      "epoch": 0.558859678088977,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0015801132255027618,
+      "loss": 1.1562,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5589473698633516,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001579706090053951,
+      "loss": 1.1355,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5590350616377261,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0015792989610157637,
+      "loss": 1.1778,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5591227534121006,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0015788918384253283,
+      "loss": 1.1674,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5592104451864752,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0015784847223197734,
+      "loss": 1.216,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5592981369608497,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001578077612736228,
+      "loss": 1.1611,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5593858287352242,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001577670509711818,
+      "loss": 1.1882,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5594735205095989,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0015772634132836714,
+      "loss": 1.1964,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5595612122839734,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0015768563234889134,
+      "loss": 1.1675,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5596489040583479,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0015764492403646697,
+      "loss": 1.184,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5597365958327225,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0015760421639480656,
+      "loss": 1.1824,
+      "step": 6383
+    },
+    {
+      "epoch": 0.559824287607097,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001575635094276225,
+      "loss": 1.1667,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5599119793814715,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0015752280313862722,
+      "loss": 1.1928,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5599996711558461,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0015748209753153294,
+      "loss": 1.1528,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5600873629302207,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0015744139261005203,
+      "loss": 1.1643,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5601750547045952,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0015740068837789653,
+      "loss": 1.2557,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5602627464789698,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0015735998483877869,
+      "loss": 1.1257,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5603504382533443,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0015731928199641044,
+      "loss": 1.115,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5604381300277188,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0015727857985450393,
+      "loss": 1.1729,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5605258218020933,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0015723787841677094,
+      "loss": 1.1319,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5606135135764679,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0015719717768692342,
+      "loss": 1.1822,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5607012053508424,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0015715647766867318,
+      "loss": 1.1574,
+      "step": 6394
+    },
+    {
+      "epoch": 0.560788897125217,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0015711577836573188,
+      "loss": 1.1939,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5608765888995916,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0015707507978181129,
+      "loss": 1.1728,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5609642806739661,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0015703438192062294,
+      "loss": 1.1533,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5610519724483406,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0015699368478587841,
+      "loss": 1.1258,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5611396642227152,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0015695298838128916,
+      "loss": 1.1756,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5612273559970897,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0015691229271056657,
+      "loss": 1.131,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5613150477714642,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0015687159777742201,
+      "loss": 1.1651,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5614027395458389,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0015683090358556675,
+      "loss": 1.1808,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5614904313202134,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0015679021013871202,
+      "loss": 1.2043,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5615781230945879,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001567495174405689,
+      "loss": 1.1812,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5616658148689625,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001567088254948485,
+      "loss": 1.2238,
+      "step": 6405
+    },
+    {
+      "epoch": 0.561753506643337,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.001566681343052618,
+      "loss": 1.1851,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5618411984177115,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0015662744387551976,
+      "loss": 1.1656,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5619288901920861,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015658675420933321,
+      "loss": 1.1281,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5620165819664606,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00156546065310413,
+      "loss": 1.1559,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5621042737408352,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0015650537718246977,
+      "loss": 1.1692,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5621919655152098,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0015646468982921423,
+      "loss": 1.1127,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5622796572895843,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0015642400325435693,
+      "loss": 1.1707,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5623673490639588,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.001563833174616084,
+      "loss": 1.1612,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5624550408383333,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.001563426324546791,
+      "loss": 1.2445,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5625427326127079,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0015630194823727935,
+      "loss": 1.2119,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5626304243870824,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0015626126481311949,
+      "loss": 1.1181,
+      "step": 6416
+    },
+    {
+      "epoch": 0.562718116161457,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0015622058218590967,
+      "loss": 1.1875,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5628058079358316,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0015617990035936011,
+      "loss": 1.1615,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5628934997102061,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0015613921933718088,
+      "loss": 1.2038,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5629811914845806,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0015609853912308196,
+      "loss": 1.1852,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5630688832589552,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0015605785972077329,
+      "loss": 1.1208,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5631565750333297,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0015601718113396475,
+      "loss": 1.184,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5632442668077042,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0015597650336636607,
+      "loss": 1.1531,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5633319585820789,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00155935826421687,
+      "loss": 1.116,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5634196503564534,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0015589515030363715,
+      "loss": 1.2072,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5635073421308279,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0015585447501592613,
+      "loss": 1.0931,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5635950339052025,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0015581380056226334,
+      "loss": 1.1671,
+      "step": 6427
+    },
+    {
+      "epoch": 0.563682725679577,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0015577312694635824,
+      "loss": 1.0912,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5637704174539515,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0015573245417192015,
+      "loss": 1.1193,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5638581092283261,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001556917822426583,
+      "loss": 1.1215,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5639458010027006,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0015565111116228197,
+      "loss": 1.1901,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5640334927770752,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0015561044093450006,
+      "loss": 1.1357,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5641211845514498,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0015556977156302172,
+      "loss": 1.2096,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5642088763258243,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0015552910305155587,
+      "loss": 1.1243,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5642965681001988,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0015548843540381136,
+      "loss": 1.1539,
+      "step": 6435
+    },
+    {
+      "epoch": 0.5643842598745734,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0015544776862349704,
+      "loss": 1.1485,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5644719516489479,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.001554071027143215,
+      "loss": 1.1624,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5645596434233224,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0015536643767999347,
+      "loss": 1.1725,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5646473351976969,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0015532577352422143,
+      "loss": 1.1978,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5647350269720716,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0015528511025071387,
+      "loss": 1.1565,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5648227187464461,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001552444478631792,
+      "loss": 1.1567,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5649104105208206,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001552037863653257,
+      "loss": 1.1374,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5649981022951952,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0015516312576086158,
+      "loss": 1.1368,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5650857940695697,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0015512246605349502,
+      "loss": 1.1687,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5651734858439442,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0015508180724693398,
+      "loss": 1.1501,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5652611776183188,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0015504114934488662,
+      "loss": 1.1564,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5653488693926934,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0015500049235106068,
+      "loss": 1.1258,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5654365611670679,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0015495983626916403,
+      "loss": 1.1579,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5655242529414425,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0015491918110290444,
+      "loss": 1.1412,
+      "step": 6449
+    },
+    {
+      "epoch": 0.565611944715817,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0015487852685598942,
+      "loss": 1.1863,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5656996364901915,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0015483787353212668,
+      "loss": 1.1376,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5657873282645661,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.001547972211350236,
+      "loss": 1.1793,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5658750200389406,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001547565696683877,
+      "loss": 1.1658,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5659627118133151,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.001547159191359261,
+      "loss": 1.1813,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5660504035876898,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0015467526954134616,
+      "loss": 1.1809,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5661380953620643,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0015463462088835499,
+      "loss": 1.1622,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5662257871364388,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0015459397318065965,
+      "loss": 1.1251,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5663134789108134,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0015455332642196709,
+      "loss": 1.14,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5664011706851879,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0015451268061598415,
+      "loss": 1.1875,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5664888624595624,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0015447203576641772,
+      "loss": 1.1589,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5665765542339369,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001544313918769744,
+      "loss": 1.1756,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5666642460083116,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.001543907489513609,
+      "loss": 1.1708,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5667519377826861,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001543501069932837,
+      "loss": 1.1668,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5668396295570606,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0015430946600644925,
+      "loss": 1.1647,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5669273213314352,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0015426882599456394,
+      "loss": 1.239,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5670150131058097,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0015422818696133394,
+      "loss": 1.1028,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5671027048801842,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001541875489104655,
+      "loss": 1.1704,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5671903966545588,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.001541469118456647,
+      "loss": 1.2021,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5672780884289333,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015410627577063752,
+      "loss": 1.1609,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5673657802033079,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.001540656406890898,
+      "loss": 1.1175,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5674534719776825,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0015402500660472752,
+      "loss": 1.1487,
+      "step": 6471
+    },
+    {
+      "epoch": 0.567541163752057,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0015398437352125624,
+      "loss": 1.2183,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5676288555264315,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0015394374144238167,
+      "loss": 1.0994,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5677165473008061,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0015390311037180935,
+      "loss": 1.1235,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5678042390751806,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0015386248031324474,
+      "loss": 1.1824,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5678919308495551,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0015382185127039312,
+      "loss": 1.1131,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5679796226239298,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0015378122324695989,
+      "loss": 1.1558,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5680673143983043,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0015374059624665005,
+      "loss": 1.1311,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5681550061726788,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0015369997027316884,
+      "loss": 1.1621,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5682426979470534,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001536593453302211,
+      "loss": 1.1048,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5683303897214279,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0015361872142151186,
+      "loss": 1.1657,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5684180814958024,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0015357809855074582,
+      "loss": 1.1411,
+      "step": 6482
+    },
+    {
+      "epoch": 0.568505773270177,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001535374767216277,
+      "loss": 1.1566,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5685934650445515,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0015349685593786214,
+      "loss": 1.0925,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5686811568189261,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0015345623620315356,
+      "loss": 1.1369,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5687688485933006,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001534156175212065,
+      "loss": 1.2089,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5688565403676752,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015337499989572518,
+      "loss": 1.1559,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5689442321420497,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0015333438333041383,
+      "loss": 1.1969,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5690319239164242,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0015329376782897666,
+      "loss": 1.1695,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5691196156907988,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001532531533951176,
+      "loss": 1.1342,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5692073074651733,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0015321254003254068,
+      "loss": 1.1141,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5692949992395478,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015317192774494963,
+      "loss": 1.1399,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5693826910139225,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0015313131653604828,
+      "loss": 1.1925,
+      "step": 6493
+    },
+    {
+      "epoch": 0.569470382788297,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001530907064095402,
+      "loss": 1.1537,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5695580745626715,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0015305009736912897,
+      "loss": 1.1201,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5696457663370461,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0015300948941851797,
+      "loss": 1.1923,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5697334581114206,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0015296888256141061,
+      "loss": 1.1451,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5698211498857951,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001529282768015101,
+      "loss": 1.1178,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5699088416601698,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0015288767214251965,
+      "loss": 1.0961,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5699965334345443,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0015284706858814214,
+      "loss": 1.1519,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5699965334345443,
+      "eval_loss": 1.1673738956451416,
+      "eval_runtime": 428.6209,
+      "eval_samples_per_second": 33.706,
+      "eval_steps_per_second": 8.427,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5700842252089188,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0015280646614208061,
+      "loss": 1.1473,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5701719169832934,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0015276586480803794,
+      "loss": 1.1324,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5702596087576679,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015272526458971678,
+      "loss": 1.1759,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5703473005320424,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001526846654908198,
+      "loss": 1.1068,
+      "step": 6504
+    },
+    {
+      "epoch": 0.570434992306417,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0015264406751504952,
+      "loss": 1.1239,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5705226840807915,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0015260347066610841,
+      "loss": 1.1505,
+      "step": 6506
+    },
+    {
+      "epoch": 0.570610375855166,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0015256287494769879,
+      "loss": 1.1984,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5706980676295407,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0015252228036352288,
+      "loss": 1.19,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5707857594039152,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0015248168691728273,
+      "loss": 1.0757,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5708734511782897,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0015244109461268046,
+      "loss": 1.1678,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5709611429526642,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0015240050345341793,
+      "loss": 1.2016,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5710488347270388,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.001523599134431969,
+      "loss": 1.1747,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5711365265014133,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0015231932458571924,
+      "loss": 1.12,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5712242182757878,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0015227873688468637,
+      "loss": 1.0671,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5713119100501625,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0015223815034379991,
+      "loss": 1.1723,
+      "step": 6515
+    },
+    {
+      "epoch": 0.571399601824537,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.001521975649667612,
+      "loss": 1.1501,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5714872935989115,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0015215698075727146,
+      "loss": 1.1721,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5715749853732861,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0015211639771903196,
+      "loss": 1.2306,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5716626771476606,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001520758158557437,
+      "loss": 1.1278,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5717503689220351,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001520352351711077,
+      "loss": 1.1728,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5718380606964097,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0015199465566882475,
+      "loss": 1.1506,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5719257524707843,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0015195407735259564,
+      "loss": 1.1374,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5720134442451588,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0015191350022612101,
+      "loss": 1.1839,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5721011360195334,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0015187292429310138,
+      "loss": 1.1396,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5721888277939079,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0015183234955723714,
+      "loss": 1.1299,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5722765195682824,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0015179177602222865,
+      "loss": 1.1961,
+      "step": 6526
+    },
+    {
+      "epoch": 0.572364211342657,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0015175120369177607,
+      "loss": 1.1657,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5724519031170315,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0015171063256957957,
+      "loss": 1.1423,
+      "step": 6528
+    },
+    {
+      "epoch": 0.572539594891406,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0015167006265933898,
+      "loss": 1.2199,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5726272866657807,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0015162949396475432,
+      "loss": 1.1024,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5727149784401552,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0015158892648952527,
+      "loss": 1.0885,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5728026702145297,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0015154836023735156,
+      "loss": 1.2168,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5728903619889042,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0015150779521193266,
+      "loss": 1.1093,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5729780537632788,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0015146723141696797,
+      "loss": 1.2044,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5730657455376533,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0015142666885615689,
+      "loss": 1.211,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5731534373120278,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001513861075331985,
+      "loss": 1.1308,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5732411290864025,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0015134554745179199,
+      "loss": 1.1379,
+      "step": 6537
+    },
+    {
+      "epoch": 0.573328820860777,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0015130498861563632,
+      "loss": 1.1147,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5734165126351515,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001512644310284303,
+      "loss": 1.2587,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5735042044095261,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0015122387469387272,
+      "loss": 1.1573,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5735918961839006,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001511833196156622,
+      "loss": 1.1539,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5736795879582751,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001511427657974973,
+      "loss": 1.1387,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5737672797326497,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0015110221324307635,
+      "loss": 1.1737,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5738549715070242,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001510616619560977,
+      "loss": 1.1281,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5739426632813988,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0015102111194025948,
+      "loss": 1.153,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5740303550557734,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001509805631992598,
+      "loss": 1.1813,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5741180468301479,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0015094001573679654,
+      "loss": 1.1682,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5742057386045224,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0015089946955656752,
+      "loss": 1.1304,
+      "step": 6548
+    },
+    {
+      "epoch": 0.574293430378897,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0015085892466227058,
+      "loss": 1.1013,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5743811221532715,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001508183810576031,
+      "loss": 1.1896,
+      "step": 6550
+    },
+    {
+      "epoch": 0.574468813927646,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001507778387462627,
+      "loss": 1.1672,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5745565057020207,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0015073729773194666,
+      "loss": 1.1243,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5746441974763952,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0015069675801835229,
+      "loss": 1.1437,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5747318892507697,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0015065621960917664,
+      "loss": 1.1048,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5748195810251443,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0015061568250811675,
+      "loss": 1.1311,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5749072727995188,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0015057514671886947,
+      "loss": 1.1287,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5749949645738933,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0015053461224513157,
+      "loss": 1.2245,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5750826563482678,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0015049407909059967,
+      "loss": 1.1767,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5751703481226424,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0015045354725897035,
+      "loss": 1.1353,
+      "step": 6559
+    },
+    {
+      "epoch": 0.575258039897017,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.001504130167539399,
+      "loss": 1.1364,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5753457316713915,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001503724875792048,
+      "loss": 1.0904,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5754334234457661,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0015033195973846094,
+      "loss": 1.113,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5755211152201406,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0015029143323540453,
+      "loss": 1.151,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5756088069945151,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0015025090807373146,
+      "loss": 1.1726,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5756964987688897,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0015021038425713745,
+      "loss": 1.1951,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5757841905432642,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0015016986178931835,
+      "loss": 1.1474,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5758718823176388,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0015012934067396946,
+      "loss": 1.1454,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5759595740920134,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0015008882091478626,
+      "loss": 1.1748,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5760472658663879,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.001500483025154642,
+      "loss": 1.1602,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5761349576407624,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001500077854796983,
+      "loss": 1.18,
+      "step": 6570
+    },
+    {
+      "epoch": 0.576222649415137,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0014996726981118369,
+      "loss": 1.1792,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5763103411895115,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0014992675551361522,
+      "loss": 1.1792,
+      "step": 6572
+    },
+    {
+      "epoch": 0.576398032963886,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.001498862425906878,
+      "loss": 1.1694,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5764857247382607,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00149845731046096,
+      "loss": 1.1887,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5765734165126352,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0014980522088353441,
+      "loss": 1.1877,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5766611082870097,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0014976471210669745,
+      "loss": 1.2154,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5767488000613843,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014972420471927946,
+      "loss": 1.1618,
+      "step": 6577
+    },
+    {
+      "epoch": 0.5768364918357588,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0014968369872497451,
+      "loss": 1.122,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5769241836101333,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0014964319412747672,
+      "loss": 1.0998,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5770118753845079,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0014960269093047996,
+      "loss": 1.1228,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5770995671588824,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001495621891376781,
+      "loss": 1.1265,
+      "step": 6581
+    },
+    {
+      "epoch": 0.577187258933257,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0014952168875276466,
+      "loss": 1.1559,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5772749507076315,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0014948118977943334,
+      "loss": 1.1404,
+      "step": 6583
+    },
+    {
+      "epoch": 0.5773626424820061,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0014944069222137739,
+      "loss": 1.1058,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5774503342563806,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0014940019608229015,
+      "loss": 1.097,
+      "step": 6585
+    },
+    {
+      "epoch": 0.5775380260307551,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0014935970136586475,
+      "loss": 1.1689,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5776257178051297,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001493192080757942,
+      "loss": 1.1258,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5777134095795042,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0014927871621577143,
+      "loss": 1.1758,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5778011013538787,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0014923822578948912,
+      "loss": 1.1779,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5778887931282534,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0014919773680063997,
+      "loss": 1.1491,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5779764849026279,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0014915724925291638,
+      "loss": 1.1562,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5780641766770024,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001491167631500108,
+      "loss": 1.2082,
+      "step": 6592
+    },
+    {
+      "epoch": 0.578151868451377,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0014907627849561538,
+      "loss": 1.159,
+      "step": 6593
+    },
+    {
+      "epoch": 0.5782395602257515,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001490357952934223,
+      "loss": 1.1082,
+      "step": 6594
+    },
+    {
+      "epoch": 0.578327252000126,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0014899531354712345,
+      "loss": 1.0696,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5784149437745006,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0014895483326041068,
+      "loss": 1.14,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5785026355488752,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0014891435443697572,
+      "loss": 1.1619,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5785903273232497,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001488738770805101,
+      "loss": 1.217,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5786780190976243,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0014883340119470525,
+      "loss": 1.1591,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5787657108719988,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0014879292678325255,
+      "loss": 1.1941,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5788534026463733,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0014875245384984304,
+      "loss": 1.1776,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5789410944207479,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0014871198239816777,
+      "loss": 1.1477,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5790287861951224,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0014867151243191765,
+      "loss": 1.1426,
+      "step": 6603
+    },
+    {
+      "epoch": 0.579116477969497,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0014863104395478348,
+      "loss": 1.1707,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5792041697438715,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0014859057697045584,
+      "loss": 1.1674,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5792918615182461,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001485501114826252,
+      "loss": 1.1671,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5793795532926206,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0014850964749498193,
+      "loss": 1.1706,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5794672450669951,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0014846918501121627,
+      "loss": 1.1243,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5795549368413697,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0014842872403501824,
+      "loss": 1.1424,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5796426286157442,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0014838826457007782,
+      "loss": 1.1872,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5797303203901187,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0014834780662008472,
+      "loss": 1.1556,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5798180121644934,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0014830735018872875,
+      "loss": 1.1546,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5799057039388679,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0014826689527969931,
+      "loss": 1.1947,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5799933957132424,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0014822644189668584,
+      "loss": 1.1886,
+      "step": 6614
+    },
+    {
+      "epoch": 0.580081087487617,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0014818599004337756,
+      "loss": 1.1989,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5801687792619915,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.001481455397234636,
+      "loss": 1.1278,
+      "step": 6616
+    },
+    {
+      "epoch": 0.580256471036366,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0014810509094063291,
+      "loss": 1.1574,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5803441628107406,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.001480646436985743,
+      "loss": 1.184,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5804318545851151,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0014802419800097647,
+      "loss": 1.1273,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5805195463594897,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0014798375385152792,
+      "loss": 1.1614,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5806072381338643,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0014794331125391712,
+      "loss": 1.1572,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5806949299082388,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001479028702118323,
+      "loss": 1.114,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5807826216826133,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0014786243072896158,
+      "loss": 1.1975,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5808703134569879,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0014782199280899296,
+      "loss": 1.1623,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5809580052313624,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001477815564556142,
+      "loss": 1.1613,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5810456970057369,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001477411216725131,
+      "loss": 1.2044,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5811333887801116,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.001477006884633771,
+      "loss": 1.1286,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5812210805544861,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.001476602568318937,
+      "loss": 1.1279,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5813087723288606,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0014761982678175008,
+      "loss": 1.1675,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5813964641032351,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0014757939831663341,
+      "loss": 1.1099,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5814841558776097,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001475389714402306,
+      "loss": 1.1385,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5815718476519842,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0014749854615622858,
+      "loss": 1.1145,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5816595394263587,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0014745812246831397,
+      "loss": 1.171,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5817472312007334,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0014741770038017334,
+      "loss": 1.1571,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5818349229751079,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0014737727989549295,
+      "loss": 1.07,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5819226147494824,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001473368610179592,
+      "loss": 1.1658,
+      "step": 6636
+    },
+    {
+      "epoch": 0.582010306523857,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001472964437512581,
+      "loss": 1.1448,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5820979982982315,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0014725602809907562,
+      "loss": 1.1345,
+      "step": 6638
+    },
+    {
+      "epoch": 0.582185690072606,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0014721561406509762,
+      "loss": 1.1409,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5822733818469806,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0014717520165300967,
+      "loss": 1.162,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5823610736213551,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0014713479086649734,
+      "loss": 1.1569,
+      "step": 6641
+    },
+    {
+      "epoch": 0.5824487653957297,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001470943817092459,
+      "loss": 1.1567,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5825364571701043,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001470539741849407,
+      "loss": 1.1385,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5826241489444788,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0014701356829726667,
+      "loss": 1.1444,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5827118407188533,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0014697316404990882,
+      "loss": 1.1737,
+      "step": 6645
+    },
+    {
+      "epoch": 0.5827995324932279,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0014693276144655187,
+      "loss": 1.1545,
+      "step": 6646
+    },
+    {
+      "epoch": 0.5828872242676024,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001468923604908804,
+      "loss": 1.1808,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5829749160419769,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.001468519611865789,
+      "loss": 1.175,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5830626078163516,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0014681156353733176,
+      "loss": 1.1567,
+      "step": 6649
+    },
+    {
+      "epoch": 0.5831502995907261,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0014677116754682302,
+      "loss": 1.1398,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5832379913651006,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001467307732187368,
+      "loss": 1.1527,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5833256831394751,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0014669038055675687,
+      "loss": 1.1722,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5834133749138497,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0014664998956456696,
+      "loss": 1.1584,
+      "step": 6653
+    },
+    {
+      "epoch": 0.5835010666882242,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001466096002458506,
+      "loss": 1.1718,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5835887584625987,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0014656921260429129,
+      "loss": 1.1168,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5836764502369733,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0014652882664357217,
+      "loss": 1.1906,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5837641420113479,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0014648844236737637,
+      "loss": 1.1417,
+      "step": 6657
+    },
+    {
+      "epoch": 0.5838518337857224,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0014644805977938682,
+      "loss": 1.1412,
+      "step": 6658
+    },
+    {
+      "epoch": 0.583939525560097,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0014640767888328637,
+      "loss": 1.173,
+      "step": 6659
+    },
+    {
+      "epoch": 0.5840272173344715,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0014636729968275758,
+      "loss": 1.1438,
+      "step": 6660
+    },
+    {
+      "epoch": 0.584114909108846,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0014632692218148299,
+      "loss": 1.1821,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5842026008832206,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0014628654638314486,
+      "loss": 1.1465,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5842902926575951,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0014624617229142543,
+      "loss": 1.1731,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5843779844319696,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0014620579991000665,
+      "loss": 1.1443,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5844656762063443,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001461654292425704,
+      "loss": 1.1478,
+      "step": 6665
+    },
+    {
+      "epoch": 0.5845533679807188,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0014612506029279839,
+      "loss": 1.1897,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5846410597550933,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014608469306437216,
+      "loss": 1.2188,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5847287515294679,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.001460443275609731,
+      "loss": 1.0837,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5848164433038424,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014600396378628242,
+      "loss": 1.1272,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5849041350782169,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001459636017439812,
+      "loss": 1.1098,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5849918268525915,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0014592324143775036,
+      "loss": 1.1893,
+      "step": 6671
+    },
+    {
+      "epoch": 0.585079518626966,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0014588288287127067,
+      "loss": 1.1451,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5851672104013406,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0014584252604822267,
+      "loss": 1.1472,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5852549021757152,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0014580217097228687,
+      "loss": 1.1446,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5853425939500897,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001457618176471435,
+      "loss": 1.1293,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5854302857244642,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0014572146607647269,
+      "loss": 1.1971,
+      "step": 6676
+    },
+    {
+      "epoch": 0.5855179774988387,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0014568111626395448,
+      "loss": 1.1173,
+      "step": 6677
+    },
+    {
+      "epoch": 0.5856056692732133,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0014564076821326852,
+      "loss": 1.154,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5856933610475878,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0014560042192809455,
+      "loss": 1.0997,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5857810528219624,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00145560077412112,
+      "loss": 1.1704,
+      "step": 6680
+    },
+    {
+      "epoch": 0.585868744596337,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0014551973466900023,
+      "loss": 1.1583,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5859564363707115,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0014547939370243834,
+      "loss": 1.152,
+      "step": 6682
+    },
+    {
+      "epoch": 0.586044128145086,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001454390545161054,
+      "loss": 1.1871,
+      "step": 6683
+    },
+    {
+      "epoch": 0.5861318199194606,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0014539871711368017,
+      "loss": 1.158,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5862195116938351,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0014535838149884136,
+      "loss": 1.2047,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5863072034682096,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0014531804767526748,
+      "loss": 1.1308,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5863948952425843,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0014527771564663676,
+      "loss": 1.2001,
+      "step": 6687
+    },
+    {
+      "epoch": 0.5864825870169588,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.001452373854166275,
+      "loss": 1.1281,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5865702787913333,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0014519705698891763,
+      "loss": 1.1435,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5866579705657079,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001451567303671851,
+      "loss": 1.1643,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5867456623400824,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001451164055551075,
+      "loss": 1.1658,
+      "step": 6691
+    },
+    {
+      "epoch": 0.5868333541144569,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001450760825563624,
+      "loss": 1.1337,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5869210458888315,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0014503576137462712,
+      "loss": 1.1698,
+      "step": 6693
+    },
+    {
+      "epoch": 0.587008737663206,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0014499544201357888,
+      "loss": 1.1794,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5870964294375806,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014495512447689467,
+      "loss": 1.1848,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5871841212119552,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001449148087682514,
+      "loss": 1.1707,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5872718129863297,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0014487449489132569,
+      "loss": 1.1317,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5873595047607042,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001448341828497941,
+      "loss": 1.1461,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5874471965350788,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00144793872647333,
+      "loss": 1.178,
+      "step": 6699
+    },
+    {
+      "epoch": 0.5875348883094533,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001447535642876185,
+      "loss": 1.1699,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5876225800838278,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0014471325777432671,
+      "loss": 1.1414,
+      "step": 6701
+    },
+    {
+      "epoch": 0.5877102718582023,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001446729531111335,
+      "loss": 1.1146,
+      "step": 6702
+    },
+    {
+      "epoch": 0.587797963632577,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001446326503017144,
+      "loss": 1.1659,
+      "step": 6703
+    },
+    {
+      "epoch": 0.5878856554069515,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0014459234934974504,
+      "loss": 1.1631,
+      "step": 6704
+    },
+    {
+      "epoch": 0.587973347181326,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0014455205025890071,
+      "loss": 1.1495,
+      "step": 6705
+    },
+    {
+      "epoch": 0.5880610389557006,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0014451175303285665,
+      "loss": 1.1394,
+      "step": 6706
+    },
+    {
+      "epoch": 0.5881487307300751,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0014447145767528778,
+      "loss": 1.1444,
+      "step": 6707
+    },
+    {
+      "epoch": 0.5882364225044496,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0014443116418986902,
+      "loss": 1.1674,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5883241142788243,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00144390872580275,
+      "loss": 1.2399,
+      "step": 6709
+    },
+    {
+      "epoch": 0.5884118060531988,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0014435058285018018,
+      "loss": 1.1569,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5884994978275733,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0014431029500325885,
+      "loss": 1.13,
+      "step": 6711
+    },
+    {
+      "epoch": 0.5885871896019479,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0014427000904318526,
+      "loss": 1.1484,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5886748813763224,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0014422972497363333,
+      "loss": 1.12,
+      "step": 6713
+    },
+    {
+      "epoch": 0.5887625731506969,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0014418944279827686,
+      "loss": 1.1149,
+      "step": 6714
+    },
+    {
+      "epoch": 0.5888502649250715,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0014414916252078944,
+      "loss": 1.0776,
+      "step": 6715
+    },
+    {
+      "epoch": 0.588937956699446,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0014410888414484464,
+      "loss": 1.1272,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5890256484738206,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001440686076741156,
+      "loss": 1.2084,
+      "step": 6717
+    },
+    {
+      "epoch": 0.5891133402481952,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0014402833311227558,
+      "loss": 1.1663,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5892010320225697,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0014398806046299747,
+      "loss": 1.1657,
+      "step": 6719
+    },
+    {
+      "epoch": 0.5892887237969442,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0014394778972995391,
+      "loss": 1.1435,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5893764155713188,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0014390752091681764,
+      "loss": 1.1213,
+      "step": 6721
+    },
+    {
+      "epoch": 0.5894641073456933,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0014386725402726094,
+      "loss": 1.0895,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5895517991200678,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.001438269890649562,
+      "loss": 1.1917,
+      "step": 6723
+    },
+    {
+      "epoch": 0.5896394908944423,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0014378672603357536,
+      "loss": 1.1347,
+      "step": 6724
+    },
+    {
+      "epoch": 0.589727182668817,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0014374646493679037,
+      "loss": 1.1518,
+      "step": 6725
+    },
+    {
+      "epoch": 0.5898148744431915,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0014370620577827285,
+      "loss": 1.1036,
+      "step": 6726
+    },
+    {
+      "epoch": 0.589902566217566,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0014366594856169445,
+      "loss": 1.1732,
+      "step": 6727
+    },
+    {
+      "epoch": 0.5899902579919406,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0014362569329072646,
+      "loss": 1.1447,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5900779497663151,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0014358543996904004,
+      "loss": 1.136,
+      "step": 6729
+    },
+    {
+      "epoch": 0.5901656415406896,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0014354518860030623,
+      "loss": 1.1331,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5902533333150642,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0014350493918819582,
+      "loss": 1.139,
+      "step": 6731
+    },
+    {
+      "epoch": 0.5903410250894388,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0014346469173637949,
+      "loss": 1.1862,
+      "step": 6732
+    },
+    {
+      "epoch": 0.5904287168638133,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0014342444624852767,
+      "loss": 1.1603,
+      "step": 6733
+    },
+    {
+      "epoch": 0.5905164086381879,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0014338420272831066,
+      "loss": 1.1724,
+      "step": 6734
+    },
+    {
+      "epoch": 0.5906041004125624,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0014334396117939856,
+      "loss": 1.1519,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5906917921869369,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0014330372160546135,
+      "loss": 1.1185,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5907794839613115,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0014326348401016867,
+      "loss": 1.1543,
+      "step": 6737
+    },
+    {
+      "epoch": 0.590867175735686,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0014322324839719008,
+      "loss": 1.1257,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5909548675100605,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0014318301477019512,
+      "loss": 1.1664,
+      "step": 6739
+    },
+    {
+      "epoch": 0.5910425592844352,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001431427831328528,
+      "loss": 1.1064,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5911302510588097,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0014310255348883231,
+      "loss": 1.1521,
+      "step": 6741
+    },
+    {
+      "epoch": 0.5912179428331842,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0014306232584180235,
+      "loss": 1.1836,
+      "step": 6742
+    },
+    {
+      "epoch": 0.5913056346075588,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001430221001954317,
+      "loss": 1.161,
+      "step": 6743
+    },
+    {
+      "epoch": 0.5913933263819333,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0014298187655338872,
+      "loss": 1.1726,
+      "step": 6744
+    },
+    {
+      "epoch": 0.5914810181563078,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0014294165491934181,
+      "loss": 1.1718,
+      "step": 6745
+    },
+    {
+      "epoch": 0.5915687099306824,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0014290143529695901,
+      "loss": 1.1438,
+      "step": 6746
+    },
+    {
+      "epoch": 0.591656401705057,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0014286121768990827,
+      "loss": 1.1601,
+      "step": 6747
+    },
+    {
+      "epoch": 0.5917440934794315,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0014282100210185728,
+      "loss": 1.1424,
+      "step": 6748
+    },
+    {
+      "epoch": 0.591831785253806,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0014278078853647372,
+      "loss": 1.2146,
+      "step": 6749
+    },
+    {
+      "epoch": 0.5919194770281806,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.001427405769974248,
+      "loss": 1.1657,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5920071688025551,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0014270036748837787,
+      "loss": 1.1562,
+      "step": 6751
+    },
+    {
+      "epoch": 0.5920948605769296,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001426601600129998,
+      "loss": 1.1418,
+      "step": 6752
+    },
+    {
+      "epoch": 0.5921825523513042,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0014261995457495755,
+      "loss": 1.1452,
+      "step": 6753
+    },
+    {
+      "epoch": 0.5922702441256787,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0014257975117791762,
+      "loss": 1.1553,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5923579359000533,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0014253954982554652,
+      "loss": 1.2102,
+      "step": 6755
+    },
+    {
+      "epoch": 0.5924456276744279,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0014249935052151044,
+      "loss": 1.1401,
+      "step": 6756
+    },
+    {
+      "epoch": 0.5925333194488024,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0014245915326947553,
+      "loss": 1.1454,
+      "step": 6757
+    },
+    {
+      "epoch": 0.5926210112231769,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0014241895807310767,
+      "loss": 1.1654,
+      "step": 6758
+    },
+    {
+      "epoch": 0.5927087029975515,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0014237876493607255,
+      "loss": 1.149,
+      "step": 6759
+    },
+    {
+      "epoch": 0.592796394771926,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0014233857386203562,
+      "loss": 1.1226,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5928840865463005,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001422983848546623,
+      "loss": 1.1733,
+      "step": 6761
+    },
+    {
+      "epoch": 0.5929717783206752,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0014225819791761768,
+      "loss": 1.2006,
+      "step": 6762
+    },
+    {
+      "epoch": 0.5930594700950497,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001422180130545667,
+      "loss": 1.1561,
+      "step": 6763
+    },
+    {
+      "epoch": 0.5931471618694242,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0014217783026917409,
+      "loss": 1.1179,
+      "step": 6764
+    },
+    {
+      "epoch": 0.5932348536437988,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001421376495651045,
+      "loss": 1.1723,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5933225454181733,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0014209747094602217,
+      "loss": 1.1122,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5934102371925478,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0014205729441559144,
+      "loss": 1.1787,
+      "step": 6767
+    },
+    {
+      "epoch": 0.5934979289669224,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0014201711997747619,
+      "loss": 1.1629,
+      "step": 6768
+    },
+    {
+      "epoch": 0.593585620741297,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0014197694763534029,
+      "loss": 1.1946,
+      "step": 6769
+    },
+    {
+      "epoch": 0.5936733125156715,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0014193677739284735,
+      "loss": 1.1731,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5937610042900461,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0014189660925366075,
+      "loss": 1.1258,
+      "step": 6771
+    },
+    {
+      "epoch": 0.5938486960644206,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0014185644322144376,
+      "loss": 1.1935,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5939363878387951,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014181627929985936,
+      "loss": 1.1101,
+      "step": 6773
+    },
+    {
+      "epoch": 0.5940240796131696,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014177611749257048,
+      "loss": 1.1173,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5941117713875442,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001417359578032397,
+      "loss": 1.172,
+      "step": 6775
+    },
+    {
+      "epoch": 0.5941994631619187,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0014169580023552955,
+      "loss": 1.1546,
+      "step": 6776
+    },
+    {
+      "epoch": 0.5942871549362932,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0014165564479310224,
+      "loss": 1.1183,
+      "step": 6777
+    },
+    {
+      "epoch": 0.5943748467106679,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0014161549147961986,
+      "loss": 1.1704,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5944625384850424,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0014157534029874427,
+      "loss": 1.1868,
+      "step": 6779
+    },
+    {
+      "epoch": 0.5945502302594169,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0014153519125413723,
+      "loss": 1.2295,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5946379220337915,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0014149504434946012,
+      "loss": 1.1211,
+      "step": 6781
+    },
+    {
+      "epoch": 0.594725613808166,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.001414548995883743,
+      "loss": 1.1301,
+      "step": 6782
+    },
+    {
+      "epoch": 0.5948133055825405,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0014141475697454086,
+      "loss": 1.1345,
+      "step": 6783
+    },
+    {
+      "epoch": 0.5949009973569152,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001413746165116207,
+      "loss": 1.1623,
+      "step": 6784
+    },
+    {
+      "epoch": 0.5949886891312897,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0014133447820327454,
+      "loss": 1.2222,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5950763809056642,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0014129434205316285,
+      "loss": 1.1341,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5951640726800388,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0014125420806494603,
+      "loss": 1.1263,
+      "step": 6787
+    },
+    {
+      "epoch": 0.5952517644544133,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0014121407624228409,
+      "loss": 1.0746,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5953394562287878,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0014117394658883697,
+      "loss": 1.1617,
+      "step": 6789
+    },
+    {
+      "epoch": 0.5954271480031624,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0014113381910826445,
+      "loss": 1.194,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5955148397775369,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0014109369380422602,
+      "loss": 1.1704,
+      "step": 6791
+    },
+    {
+      "epoch": 0.5956025315519115,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0014105357068038096,
+      "loss": 1.1137,
+      "step": 6792
+    },
+    {
+      "epoch": 0.5956902233262861,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0014101344974038847,
+      "loss": 1.1629,
+      "step": 6793
+    },
+    {
+      "epoch": 0.5957779151006606,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0014097333098790743,
+      "loss": 1.1639,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5958656068750351,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0014093321442659656,
+      "loss": 1.1031,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5959532986494096,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0014089310006011446,
+      "loss": 1.1601,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5960409904237842,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0014085298789211936,
+      "loss": 1.1727,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5961286821981587,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0014081287792626945,
+      "loss": 1.1469,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5962163739725332,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.001407727701662226,
+      "loss": 1.214,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5963040657469079,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0014073266461563662,
+      "loss": 1.1608,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5963917575212824,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0014069256127816896,
+      "loss": 1.152,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5964794492956569,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0014065246015747697,
+      "loss": 1.1424,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5965671410700315,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0014061236125721777,
+      "loss": 1.1651,
+      "step": 6803
+    },
+    {
+      "epoch": 0.596654832844406,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0014057226458104822,
+      "loss": 1.0881,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5967425246187805,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0014053217013262514,
+      "loss": 1.1691,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5968302163931551,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0014049207791560494,
+      "loss": 1.1236,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5969179081675297,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00140451987933644,
+      "loss": 1.1734,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5970055999419042,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0014041190019039832,
+      "loss": 1.1557,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5970932917162788,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0014037181468952394,
+      "loss": 1.1358,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5971809834906533,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0014033173143467644,
+      "loss": 1.1523,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5972686752650278,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0014029165042951134,
+      "loss": 1.1464,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5973563670394024,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0014025157167768397,
+      "loss": 1.0916,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5974440588137769,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0014021149518284935,
+      "loss": 1.1473,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5975317505881514,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001401714209486624,
+      "loss": 1.1554,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5976194423625261,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0014013134897877778,
+      "loss": 1.1871,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5977071341369006,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.001400912792768499,
+      "loss": 1.1417,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5977948259112751,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0014005121184653309,
+      "loss": 1.1749,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5978825176856497,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0014001114669148131,
+      "loss": 1.2356,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5979702094600242,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0013997108381534845,
+      "loss": 1.18,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5980579012343987,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001399310232217882,
+      "loss": 1.18,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5981455930087732,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001398909649144539,
+      "loss": 1.1364,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5982332847831479,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0013985090889699872,
+      "loss": 1.165,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5983209765575224,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0013981085517307578,
+      "loss": 1.1581,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5984086683318969,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0013977080374633776,
+      "loss": 1.1309,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5984963601062715,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0013973075462043738,
+      "loss": 1.2084,
+      "step": 6825
+    },
+    {
+      "epoch": 0.598584051880646,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0013969070779902687,
+      "loss": 1.1102,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5986717436550205,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0013965066328575852,
+      "loss": 1.1294,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5987594354293951,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0013961062108428421,
+      "loss": 1.1641,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5988471272037696,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013957058119825577,
+      "loss": 1.0509,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5989348189781442,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0013953054363132465,
+      "loss": 1.087,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5990225107525188,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001394905083871422,
+      "loss": 1.1567,
+      "step": 6831
+    },
+    {
+      "epoch": 0.5991102025268933,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0013945047546935951,
+      "loss": 1.1862,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5991978943012678,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0013941044488162753,
+      "loss": 1.1225,
+      "step": 6833
+    },
+    {
+      "epoch": 0.5992855860756424,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.001393704166275969,
+      "loss": 1.1721,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5993732778500169,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.001393303907109181,
+      "loss": 1.1816,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5994609696243914,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0013929036713524147,
+      "loss": 1.1805,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5995486613987661,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0013925034590421695,
+      "loss": 1.1387,
+      "step": 6837
+    },
+    {
+      "epoch": 0.5996363531731406,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0013921032702149444,
+      "loss": 1.2183,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5997240449475151,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0013917031049072346,
+      "loss": 1.1666,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5998117367218897,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0013913029631555352,
+      "loss": 1.2364,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5998994284962642,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.001390902844996338,
+      "loss": 1.1031,
+      "step": 6841
+    },
+    {
+      "epoch": 0.5999871202706387,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0013905027504661316,
+      "loss": 1.1253,
+      "step": 6842
+    },
+    {
+      "epoch": 0.6000748120450133,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0013901026796014051,
+      "loss": 1.1478,
+      "step": 6843
+    },
+    {
+      "epoch": 0.6001625038193878,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0013897026324386436,
+      "loss": 1.1008,
+      "step": 6844
+    },
+    {
+      "epoch": 0.6002501955937624,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0013893026090143299,
+      "loss": 1.2723,
+      "step": 6845
+    },
+    {
+      "epoch": 0.6003378873681369,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0013889026093649448,
+      "loss": 1.1811,
+      "step": 6846
+    },
+    {
+      "epoch": 0.6004255791425115,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0013885026335269679,
+      "loss": 1.1386,
+      "step": 6847
+    },
+    {
+      "epoch": 0.600513270916886,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001388102681536876,
+      "loss": 1.1425,
+      "step": 6848
+    },
+    {
+      "epoch": 0.6006009626912605,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0013877027534311434,
+      "loss": 1.1828,
+      "step": 6849
+    },
+    {
+      "epoch": 0.6006886544656351,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0013873028492462423,
+      "loss": 1.1733,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6007763462400096,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0013869029690186435,
+      "loss": 1.168,
+      "step": 6851
+    },
+    {
+      "epoch": 0.6008640380143841,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013865031127848143,
+      "loss": 1.0719,
+      "step": 6852
+    },
+    {
+      "epoch": 0.6009517297887588,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0013861032805812215,
+      "loss": 1.1597,
+      "step": 6853
+    },
+    {
+      "epoch": 0.6010394215631333,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0013857034724443287,
+      "loss": 1.1289,
+      "step": 6854
+    },
+    {
+      "epoch": 0.6011271133375078,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0013853036884105964,
+      "loss": 1.1274,
+      "step": 6855
+    },
+    {
+      "epoch": 0.6012148051118824,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0013849039285164844,
+      "loss": 1.1377,
+      "step": 6856
+    },
+    {
+      "epoch": 0.6013024968862569,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013845041927984495,
+      "loss": 1.2187,
+      "step": 6857
+    },
+    {
+      "epoch": 0.6013901886606314,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0013841044812929473,
+      "loss": 1.1356,
+      "step": 6858
+    },
+    {
+      "epoch": 0.601477880435006,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0013837047940364296,
+      "loss": 1.148,
+      "step": 6859
+    },
+    {
+      "epoch": 0.6015655722093806,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0013833051310653474,
+      "loss": 1.1463,
+      "step": 6860
+    },
+    {
+      "epoch": 0.6016532639837551,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0013829054924161485,
+      "loss": 1.1551,
+      "step": 6861
+    },
+    {
+      "epoch": 0.6017409557581297,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013825058781252798,
+      "loss": 1.1378,
+      "step": 6862
+    },
+    {
+      "epoch": 0.6018286475325042,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0013821062882291841,
+      "loss": 1.1071,
+      "step": 6863
+    },
+    {
+      "epoch": 0.6019163393068787,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0013817067227643034,
+      "loss": 1.1459,
+      "step": 6864
+    },
+    {
+      "epoch": 0.6020040310812533,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0013813071817670769,
+      "loss": 1.1573,
+      "step": 6865
+    },
+    {
+      "epoch": 0.6020917228556278,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0013809076652739416,
+      "loss": 1.156,
+      "step": 6866
+    },
+    {
+      "epoch": 0.6021794146300024,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0013805081733213333,
+      "loss": 1.1583,
+      "step": 6867
+    },
+    {
+      "epoch": 0.6022671064043769,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013801087059456833,
+      "loss": 1.1862,
+      "step": 6868
+    },
+    {
+      "epoch": 0.6023547981787515,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001379709263183423,
+      "loss": 1.1801,
+      "step": 6869
+    },
+    {
+      "epoch": 0.602442489953126,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0013793098450709798,
+      "loss": 1.1178,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6025301817275005,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0013789104516447803,
+      "loss": 1.1991,
+      "step": 6871
+    },
+    {
+      "epoch": 0.6026178735018751,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0013785110829412479,
+      "loss": 1.0913,
+      "step": 6872
+    },
+    {
+      "epoch": 0.6027055652762496,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.001378111738996803,
+      "loss": 1.1135,
+      "step": 6873
+    },
+    {
+      "epoch": 0.6027932570506241,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0013777124198478667,
+      "loss": 1.129,
+      "step": 6874
+    },
+    {
+      "epoch": 0.6028809488249988,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0013773131255308542,
+      "loss": 1.1324,
+      "step": 6875
+    },
+    {
+      "epoch": 0.6029686405993733,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0013769138560821806,
+      "loss": 1.1065,
+      "step": 6876
+    },
+    {
+      "epoch": 0.6030563323737478,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0013765146115382587,
+      "loss": 1.1309,
+      "step": 6877
+    },
+    {
+      "epoch": 0.6031440241481224,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013761153919354982,
+      "loss": 1.1465,
+      "step": 6878
+    },
+    {
+      "epoch": 0.6032317159224969,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0013757161973103065,
+      "loss": 1.2025,
+      "step": 6879
+    },
+    {
+      "epoch": 0.6033194076968714,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0013753170276990898,
+      "loss": 1.1748,
+      "step": 6880
+    },
+    {
+      "epoch": 0.603407099471246,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0013749178831382514,
+      "loss": 1.136,
+      "step": 6881
+    },
+    {
+      "epoch": 0.6034947912456206,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0013745187636641917,
+      "loss": 1.1192,
+      "step": 6882
+    },
+    {
+      "epoch": 0.6035824830199951,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0013741196693133092,
+      "loss": 1.1606,
+      "step": 6883
+    },
+    {
+      "epoch": 0.6036701747943697,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001373720600122001,
+      "loss": 1.1478,
+      "step": 6884
+    },
+    {
+      "epoch": 0.6037578665687442,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001373321556126661,
+      "loss": 1.0844,
+      "step": 6885
+    },
+    {
+      "epoch": 0.6038455583431187,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001372922537363681,
+      "loss": 1.1169,
+      "step": 6886
+    },
+    {
+      "epoch": 0.6039332501174933,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0013725235438694498,
+      "loss": 1.2101,
+      "step": 6887
+    },
+    {
+      "epoch": 0.6040209418918678,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0013721245756803558,
+      "loss": 1.1506,
+      "step": 6888
+    },
+    {
+      "epoch": 0.6041086336662423,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0013717256328327828,
+      "loss": 1.2201,
+      "step": 6889
+    },
+    {
+      "epoch": 0.604196325440617,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0013713267153631135,
+      "loss": 1.1558,
+      "step": 6890
+    },
+    {
+      "epoch": 0.6042840172149915,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0013709278233077287,
+      "loss": 1.0958,
+      "step": 6891
+    },
+    {
+      "epoch": 0.604371708989366,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0013705289567030056,
+      "loss": 1.1832,
+      "step": 6892
+    },
+    {
+      "epoch": 0.6044594007637405,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0013701301155853202,
+      "loss": 1.1421,
+      "step": 6893
+    },
+    {
+      "epoch": 0.6045470925381151,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0013697312999910464,
+      "loss": 1.1767,
+      "step": 6894
+    },
+    {
+      "epoch": 0.6046347843124896,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001369332509956554,
+      "loss": 1.281,
+      "step": 6895
+    },
+    {
+      "epoch": 0.6047224760868641,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0013689337455182122,
+      "loss": 1.1507,
+      "step": 6896
+    },
+    {
+      "epoch": 0.6048101678612388,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0013685350067123873,
+      "loss": 1.1057,
+      "step": 6897
+    },
+    {
+      "epoch": 0.6048978596356133,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0013681362935754435,
+      "loss": 1.126,
+      "step": 6898
+    },
+    {
+      "epoch": 0.6049855514099878,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0013677376061437417,
+      "loss": 1.1488,
+      "step": 6899
+    },
+    {
+      "epoch": 0.6050732431843624,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0013673389444536417,
+      "loss": 1.1829,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6051609349587369,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0013669403085415,
+      "loss": 1.2205,
+      "step": 6901
+    },
+    {
+      "epoch": 0.6052486267331114,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0013665416984436718,
+      "loss": 1.1251,
+      "step": 6902
+    },
+    {
+      "epoch": 0.605336318507486,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0013661431141965088,
+      "loss": 1.1633,
+      "step": 6903
+    },
+    {
+      "epoch": 0.6054240102818605,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0013657445558363612,
+      "loss": 1.153,
+      "step": 6904
+    },
+    {
+      "epoch": 0.6055117020562351,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0013653460233995767,
+      "loss": 1.1302,
+      "step": 6905
+    },
+    {
+      "epoch": 0.6055993938306097,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0013649475169224992,
+      "loss": 1.1041,
+      "step": 6906
+    },
+    {
+      "epoch": 0.6056870856049842,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0013645490364414727,
+      "loss": 1.1579,
+      "step": 6907
+    },
+    {
+      "epoch": 0.6057747773793587,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001364150581992837,
+      "loss": 1.1275,
+      "step": 6908
+    },
+    {
+      "epoch": 0.6058624691537333,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0013637521536129303,
+      "loss": 1.1988,
+      "step": 6909
+    },
+    {
+      "epoch": 0.6059501609281078,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0013633537513380886,
+      "loss": 1.1497,
+      "step": 6910
+    },
+    {
+      "epoch": 0.6060378527024823,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0013629553752046446,
+      "loss": 1.2842,
+      "step": 6911
+    },
+    {
+      "epoch": 0.606125544476857,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0013625570252489292,
+      "loss": 1.1625,
+      "step": 6912
+    },
+    {
+      "epoch": 0.6062132362512315,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0013621587015072714,
+      "loss": 1.1526,
+      "step": 6913
+    },
+    {
+      "epoch": 0.606300928025606,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0013617604040159967,
+      "loss": 1.1649,
+      "step": 6914
+    },
+    {
+      "epoch": 0.6063886197999805,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0013613621328114292,
+      "loss": 1.1654,
+      "step": 6915
+    },
+    {
+      "epoch": 0.6064763115743551,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0013609638879298897,
+      "loss": 1.1149,
+      "step": 6916
+    },
+    {
+      "epoch": 0.6065640033487296,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0013605656694076975,
+      "loss": 1.1657,
+      "step": 6917
+    },
+    {
+      "epoch": 0.6066516951231041,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0013601674772811697,
+      "loss": 1.1605,
+      "step": 6918
+    },
+    {
+      "epoch": 0.6067393868974787,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001359769311586619,
+      "loss": 1.154,
+      "step": 6919
+    },
+    {
+      "epoch": 0.6068270786718533,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0013593711723603585,
+      "loss": 1.1168,
+      "step": 6920
+    },
+    {
+      "epoch": 0.6069147704462278,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0013589730596386964,
+      "loss": 1.169,
+      "step": 6921
+    },
+    {
+      "epoch": 0.6070024622206024,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001358574973457941,
+      "loss": 1.1701,
+      "step": 6922
+    },
+    {
+      "epoch": 0.6070901539949769,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0013581769138543947,
+      "loss": 1.1789,
+      "step": 6923
+    },
+    {
+      "epoch": 0.6071778457693514,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0013577788808643606,
+      "loss": 1.1562,
+      "step": 6924
+    },
+    {
+      "epoch": 0.607265537543726,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0013573808745241385,
+      "loss": 1.1661,
+      "step": 6925
+    },
+    {
+      "epoch": 0.6073532293181005,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001356982894870025,
+      "loss": 1.1582,
+      "step": 6926
+    },
+    {
+      "epoch": 0.607440921092475,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0013565849419383152,
+      "loss": 1.139,
+      "step": 6927
+    },
+    {
+      "epoch": 0.6075286128668497,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0013561870157653012,
+      "loss": 1.1224,
+      "step": 6928
+    },
+    {
+      "epoch": 0.6076163046412242,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001355789116387273,
+      "loss": 1.0936,
+      "step": 6929
+    },
+    {
+      "epoch": 0.6077039964155987,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0013553912438405175,
+      "loss": 1.1331,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6077916881899733,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0013549933981613202,
+      "loss": 1.1388,
+      "step": 6931
+    },
+    {
+      "epoch": 0.6078793799643478,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0013545955793859636,
+      "loss": 1.2244,
+      "step": 6932
+    },
+    {
+      "epoch": 0.6079670717387223,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0013541977875507277,
+      "loss": 1.1788,
+      "step": 6933
+    },
+    {
+      "epoch": 0.608054763513097,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0013538000226918894,
+      "loss": 1.1546,
+      "step": 6934
+    },
+    {
+      "epoch": 0.6081424552874715,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013534022848457252,
+      "loss": 1.1026,
+      "step": 6935
+    },
+    {
+      "epoch": 0.608230147061846,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001353004574048506,
+      "loss": 1.2076,
+      "step": 6936
+    },
+    {
+      "epoch": 0.6083178388362206,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0013526068903365039,
+      "loss": 1.1365,
+      "step": 6937
+    },
+    {
+      "epoch": 0.6084055306105951,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.001352209233745985,
+      "loss": 1.1471,
+      "step": 6938
+    },
+    {
+      "epoch": 0.6084932223849696,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001351811604313216,
+      "loss": 1.1913,
+      "step": 6939
+    },
+    {
+      "epoch": 0.6085809141593441,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0013514140020744587,
+      "loss": 1.1136,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6086686059337187,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0013510164270659729,
+      "loss": 1.1642,
+      "step": 6941
+    },
+    {
+      "epoch": 0.6087562977080933,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.001350618879324018,
+      "loss": 1.0913,
+      "step": 6942
+    },
+    {
+      "epoch": 0.6088439894824678,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0013502213588848476,
+      "loss": 1.1066,
+      "step": 6943
+    },
+    {
+      "epoch": 0.6089316812568424,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001349823865784716,
+      "loss": 1.2153,
+      "step": 6944
+    },
+    {
+      "epoch": 0.6090193730312169,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0013494264000598725,
+      "loss": 1.1746,
+      "step": 6945
+    },
+    {
+      "epoch": 0.6091070648055914,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0013490289617465656,
+      "loss": 1.1373,
+      "step": 6946
+    },
+    {
+      "epoch": 0.609194756579966,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0013486315508810407,
+      "loss": 1.1602,
+      "step": 6947
+    },
+    {
+      "epoch": 0.6092824483543405,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00134823416749954,
+      "loss": 1.1731,
+      "step": 6948
+    },
+    {
+      "epoch": 0.609370140128715,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0013478368116383044,
+      "loss": 1.138,
+      "step": 6949
+    },
+    {
+      "epoch": 0.6094578319030897,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0013474394833335715,
+      "loss": 1.071,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6095455236774642,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.001347042182621577,
+      "loss": 1.1846,
+      "step": 6951
+    },
+    {
+      "epoch": 0.6096332154518387,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001346644909538553,
+      "loss": 1.1784,
+      "step": 6952
+    },
+    {
+      "epoch": 0.6097209072262133,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00134624766412073,
+      "loss": 1.1712,
+      "step": 6953
+    },
+    {
+      "epoch": 0.6098085990005878,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0013458504464043358,
+      "loss": 1.125,
+      "step": 6954
+    },
+    {
+      "epoch": 0.6098962907749623,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0013454532564255966,
+      "loss": 1.1824,
+      "step": 6955
+    },
+    {
+      "epoch": 0.609983982549337,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.001345056094220734,
+      "loss": 1.1512,
+      "step": 6956
+    },
+    {
+      "epoch": 0.6100716743237115,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0013446589598259677,
+      "loss": 1.1292,
+      "step": 6957
+    },
+    {
+      "epoch": 0.610159366098086,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013442618532775168,
+      "loss": 1.1319,
+      "step": 6958
+    },
+    {
+      "epoch": 0.6102470578724606,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0013438647746115947,
+      "loss": 1.1561,
+      "step": 6959
+    },
+    {
+      "epoch": 0.6103347496468351,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013434677238644158,
+      "loss": 1.1709,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6104224414212096,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0013430707010721884,
+      "loss": 1.1731,
+      "step": 6961
+    },
+    {
+      "epoch": 0.6105101331955842,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0013426737062711212,
+      "loss": 1.1645,
+      "step": 6962
+    },
+    {
+      "epoch": 0.6105978249699587,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0013422767394974183,
+      "loss": 1.1936,
+      "step": 6963
+    },
+    {
+      "epoch": 0.6106855167443332,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.001341879800787283,
+      "loss": 1.1678,
+      "step": 6964
+    },
+    {
+      "epoch": 0.6107732085187078,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001341482890176914,
+      "loss": 1.1192,
+      "step": 6965
+    },
+    {
+      "epoch": 0.6108609002930824,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0013410860077025088,
+      "loss": 1.1294,
+      "step": 6966
+    },
+    {
+      "epoch": 0.6109485920674569,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0013406891534002625,
+      "loss": 1.1399,
+      "step": 6967
+    },
+    {
+      "epoch": 0.6110362838418314,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013402923273063672,
+      "loss": 1.1161,
+      "step": 6968
+    },
+    {
+      "epoch": 0.611123975616206,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013398955294570115,
+      "loss": 1.2074,
+      "step": 6969
+    },
+    {
+      "epoch": 0.6112116673905805,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0013394987598883833,
+      "loss": 1.132,
+      "step": 6970
+    },
+    {
+      "epoch": 0.611299359164955,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0013391020186366666,
+      "loss": 1.1569,
+      "step": 6971
+    },
+    {
+      "epoch": 0.6113870509393297,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0013387053057380433,
+      "loss": 1.2112,
+      "step": 6972
+    },
+    {
+      "epoch": 0.6114747427137042,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013383086212286931,
+      "loss": 1.1633,
+      "step": 6973
+    },
+    {
+      "epoch": 0.6115624344880787,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0013379119651447916,
+      "loss": 1.1634,
+      "step": 6974
+    },
+    {
+      "epoch": 0.6116501262624533,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0013375153375225126,
+      "loss": 1.1658,
+      "step": 6975
+    },
+    {
+      "epoch": 0.6117378180368278,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001337118738398029,
+      "loss": 1.1337,
+      "step": 6976
+    },
+    {
+      "epoch": 0.6118255098112023,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.001336722167807508,
+      "loss": 1.2006,
+      "step": 6977
+    },
+    {
+      "epoch": 0.6119132015855769,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.001336325625787117,
+      "loss": 1.149,
+      "step": 6978
+    },
+    {
+      "epoch": 0.6120008933599514,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0013359291123730191,
+      "loss": 1.1229,
+      "step": 6979
+    },
+    {
+      "epoch": 0.612088585134326,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0013355326276013756,
+      "loss": 1.1943,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6121762769087006,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0013351361715083442,
+      "loss": 1.1539,
+      "step": 6981
+    },
+    {
+      "epoch": 0.6122639686830751,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0013347397441300815,
+      "loss": 1.1314,
+      "step": 6982
+    },
+    {
+      "epoch": 0.6123516604574496,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0013343433455027401,
+      "loss": 1.1423,
+      "step": 6983
+    },
+    {
+      "epoch": 0.6124393522318242,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0013339469756624709,
+      "loss": 1.1411,
+      "step": 6984
+    },
+    {
+      "epoch": 0.6125270440061987,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0013335506346454213,
+      "loss": 1.2055,
+      "step": 6985
+    },
+    {
+      "epoch": 0.6126147357805732,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.001333154322487737,
+      "loss": 1.1789,
+      "step": 6986
+    },
+    {
+      "epoch": 0.6127024275549477,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.001332758039225561,
+      "loss": 1.089,
+      "step": 6987
+    },
+    {
+      "epoch": 0.6127901193293224,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0013323617848950327,
+      "loss": 1.1798,
+      "step": 6988
+    },
+    {
+      "epoch": 0.6128778111036969,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001331965559532289,
+      "loss": 1.1761,
+      "step": 6989
+    },
+    {
+      "epoch": 0.6129655028780714,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0013315693631734665,
+      "loss": 1.2367,
+      "step": 6990
+    },
+    {
+      "epoch": 0.613053194652446,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013311731958546957,
+      "loss": 1.15,
+      "step": 6991
+    },
+    {
+      "epoch": 0.6131408864268205,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0013307770576121056,
+      "loss": 1.1308,
+      "step": 6992
+    },
+    {
+      "epoch": 0.613228578201195,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0013303809484818243,
+      "loss": 1.1075,
+      "step": 6993
+    },
+    {
+      "epoch": 0.6133162699755697,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001329984868499975,
+      "loss": 1.1455,
+      "step": 6994
+    },
+    {
+      "epoch": 0.6134039617499442,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0013295888177026803,
+      "loss": 1.126,
+      "step": 6995
+    },
+    {
+      "epoch": 0.6134916535243187,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0013291927961260577,
+      "loss": 1.1293,
+      "step": 6996
+    },
+    {
+      "epoch": 0.6135793452986933,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0013287968038062242,
+      "loss": 1.1212,
+      "step": 6997
+    },
+    {
+      "epoch": 0.6136670370730678,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0013284008407792927,
+      "loss": 1.1112,
+      "step": 6998
+    },
+    {
+      "epoch": 0.6137547288474423,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0013280049070813745,
+      "loss": 1.1704,
+      "step": 6999
+    },
+    {
+      "epoch": 0.6138424206218169,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.001327609002748578,
+      "loss": 1.1397,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6138424206218169,
+      "eval_loss": 1.160142183303833,
+      "eval_runtime": 429.0326,
+      "eval_samples_per_second": 33.673,
+      "eval_steps_per_second": 8.419,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6139301123961914,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0013272131278170075,
+      "loss": 1.1611,
+      "step": 7001
+    },
+    {
+      "epoch": 0.614017804170566,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.001326817282322767,
+      "loss": 1.1525,
+      "step": 7002
+    },
+    {
+      "epoch": 0.6141054959449406,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013264214663019558,
+      "loss": 1.1286,
+      "step": 7003
+    },
+    {
+      "epoch": 0.6141931877193151,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0013260256797906714,
+      "loss": 1.1521,
+      "step": 7004
+    },
+    {
+      "epoch": 0.6142808794936896,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001325629922825009,
+      "loss": 1.1163,
+      "step": 7005
+    },
+    {
+      "epoch": 0.6143685712680642,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00132523419544106,
+      "loss": 1.1486,
+      "step": 7006
+    },
+    {
+      "epoch": 0.6144562630424387,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0013248384976749146,
+      "loss": 1.1185,
+      "step": 7007
+    },
+    {
+      "epoch": 0.6145439548168132,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013244428295626585,
+      "loss": 1.1294,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6146316465911879,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001324047191140376,
+      "loss": 1.1348,
+      "step": 7009
+    },
+    {
+      "epoch": 0.6147193383655624,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0013236515824441475,
+      "loss": 1.1134,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6148070301399369,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001323256003510053,
+      "loss": 1.1252,
+      "step": 7011
+    },
+    {
+      "epoch": 0.6148947219143114,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.001322860454374167,
+      "loss": 1.22,
+      "step": 7012
+    },
+    {
+      "epoch": 0.614982413688686,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0013224649350725632,
+      "loss": 1.1248,
+      "step": 7013
+    },
+    {
+      "epoch": 0.6150701054630605,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013220694456413118,
+      "loss": 1.161,
+      "step": 7014
+    },
+    {
+      "epoch": 0.615157797237435,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0013216739861164808,
+      "loss": 1.165,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6152454890118096,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.001321278556534134,
+      "loss": 1.1638,
+      "step": 7016
+    },
+    {
+      "epoch": 0.6153331807861842,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013208831569303347,
+      "loss": 1.175,
+      "step": 7017
+    },
+    {
+      "epoch": 0.6154208725605587,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.001320487787341142,
+      "loss": 1.2447,
+      "step": 7018
+    },
+    {
+      "epoch": 0.6155085643349333,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001320092447802612,
+      "loss": 1.1454,
+      "step": 7019
+    },
+    {
+      "epoch": 0.6155962561093078,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013196971383507992,
+      "loss": 1.1323,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6156839478836823,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0013193018590217551,
+      "loss": 1.1776,
+      "step": 7021
+    },
+    {
+      "epoch": 0.6157716396580569,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0013189066098515275,
+      "loss": 1.2058,
+      "step": 7022
+    },
+    {
+      "epoch": 0.6158593314324314,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013185113908761627,
+      "loss": 1.1369,
+      "step": 7023
+    },
+    {
+      "epoch": 0.6159470232068059,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0013181162021317033,
+      "loss": 1.1576,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6160347149811806,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001317721043654189,
+      "loss": 1.1829,
+      "step": 7025
+    },
+    {
+      "epoch": 0.6161224067555551,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001317325915479658,
+      "loss": 1.1451,
+      "step": 7026
+    },
+    {
+      "epoch": 0.6162100985299296,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001316930817644145,
+      "loss": 1.1137,
+      "step": 7027
+    },
+    {
+      "epoch": 0.6162977903043042,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001316535750183681,
+      "loss": 1.1314,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6163854820786787,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0013161407131342961,
+      "loss": 1.0983,
+      "step": 7029
+    },
+    {
+      "epoch": 0.6164731738530532,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0013157457065320162,
+      "loss": 1.1452,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6165608656274278,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0013153507304128652,
+      "loss": 1.0899,
+      "step": 7031
+    },
+    {
+      "epoch": 0.6166485574018024,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0013149557848128634,
+      "loss": 1.2623,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6167362491761769,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0013145608697680295,
+      "loss": 1.1314,
+      "step": 7033
+    },
+    {
+      "epoch": 0.6168239409505515,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.001314165985314378,
+      "loss": 1.1606,
+      "step": 7034
+    },
+    {
+      "epoch": 0.616911632724926,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0013137711314879217,
+      "loss": 1.12,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6169993244993005,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0013133763083246703,
+      "loss": 1.2337,
+      "step": 7036
+    },
+    {
+      "epoch": 0.617087016273675,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0013129815158606309,
+      "loss": 1.2394,
+      "step": 7037
+    },
+    {
+      "epoch": 0.6171747080480496,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0013125867541318068,
+      "loss": 1.1556,
+      "step": 7038
+    },
+    {
+      "epoch": 0.6172623998224241,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0013121920231741998,
+      "loss": 1.1269,
+      "step": 7039
+    },
+    {
+      "epoch": 0.6173500915967987,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0013117973230238084,
+      "loss": 1.2225,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6174377833711733,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0013114026537166283,
+      "loss": 1.2224,
+      "step": 7041
+    },
+    {
+      "epoch": 0.6175254751455478,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.001311008015288652,
+      "loss": 1.1639,
+      "step": 7042
+    },
+    {
+      "epoch": 0.6176131669199223,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.001310613407775869,
+      "loss": 1.1485,
+      "step": 7043
+    },
+    {
+      "epoch": 0.6177008586942969,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001310218831214268,
+      "loss": 1.1462,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6177885504686714,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0013098242856398315,
+      "loss": 1.1115,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6178762422430459,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0013094297710885424,
+      "loss": 1.1209,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6179639340174206,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001309035287596379,
+      "loss": 1.218,
+      "step": 7047
+    },
+    {
+      "epoch": 0.6180516257917951,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013086408351993175,
+      "loss": 1.1006,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6181393175661696,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0013082464139333298,
+      "loss": 1.1715,
+      "step": 7049
+    },
+    {
+      "epoch": 0.6182270093405442,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001307852023834388,
+      "loss": 1.146,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6183147011149187,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013074576649384575,
+      "loss": 1.1791,
+      "step": 7051
+    },
+    {
+      "epoch": 0.6184023928892932,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0013070633372815037,
+      "loss": 1.1356,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6184900846636678,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0013066690408994888,
+      "loss": 1.1568,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6185777764380423,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0013062747758283706,
+      "loss": 1.1428,
+      "step": 7054
+    },
+    {
+      "epoch": 0.6186654682124169,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001305880542104106,
+      "loss": 1.1859,
+      "step": 7055
+    },
+    {
+      "epoch": 0.6187531599867915,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0013054863397626473,
+      "loss": 1.225,
+      "step": 7056
+    },
+    {
+      "epoch": 0.618840851761166,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0013050921688399452,
+      "loss": 1.1398,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6189285435355405,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0013046980293719471,
+      "loss": 1.1559,
+      "step": 7058
+    },
+    {
+      "epoch": 0.619016235309915,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001304303921394597,
+      "loss": 1.1,
+      "step": 7059
+    },
+    {
+      "epoch": 0.6191039270842896,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0013039098449438371,
+      "loss": 1.1826,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6191916188586641,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0013035158000556056,
+      "loss": 1.1743,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6192793106330386,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0013031217867658389,
+      "loss": 1.1808,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6193670024074133,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0013027278051104694,
+      "loss": 1.1274,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6194546941817878,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0013023338551254283,
+      "loss": 1.1416,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6195423859561623,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0013019399368466415,
+      "loss": 1.1597,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6196300777305369,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0013015460503100346,
+      "loss": 1.1999,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6197177695049114,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0013011521955515275,
+      "loss": 1.1818,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6198054612792859,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0013007583726070402,
+      "loss": 1.2153,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6198931530536606,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0013003645815124875,
+      "loss": 1.2075,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6199808448280351,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0012999708223037827,
+      "loss": 1.1377,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6200685366024096,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.001299577095016835,
+      "loss": 1.1461,
+      "step": 7071
+    },
+    {
+      "epoch": 0.6201562283767842,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.001299183399687552,
+      "loss": 1.2033,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6202439201511587,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0012987897363518368,
+      "loss": 1.0833,
+      "step": 7073
+    },
+    {
+      "epoch": 0.6203316119255332,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0012983961050455916,
+      "loss": 1.1631,
+      "step": 7074
+    },
+    {
+      "epoch": 0.6204193036999078,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0012980025058047138,
+      "loss": 1.1793,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6205069954742823,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0012976089386650983,
+      "loss": 1.1512,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6205946872486569,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0012972154036626385,
+      "loss": 1.1151,
+      "step": 7077
+    },
+    {
+      "epoch": 0.6206823790230315,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0012968219008332228,
+      "loss": 1.2146,
+      "step": 7078
+    },
+    {
+      "epoch": 0.620770070797406,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0012964284302127381,
+      "loss": 1.1997,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6208577625717805,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0012960349918370677,
+      "loss": 1.1165,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6209454543461551,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0012956415857420924,
+      "loss": 1.1553,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6210331461205296,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00129524821196369,
+      "loss": 1.1352,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6211208378949041,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0012948548705377344,
+      "loss": 1.0914,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6212085296692786,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0012944615615000984,
+      "loss": 1.1849,
+      "step": 7084
+    },
+    {
+      "epoch": 0.6212962214436533,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0012940682848866494,
+      "loss": 1.1317,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6213839132180278,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0012936750407332544,
+      "loss": 1.163,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6214716049924023,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0012932818290757758,
+      "loss": 1.1418,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6215592967667769,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0012928886499500736,
+      "loss": 1.1355,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6216469885411514,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012924955033920046,
+      "loss": 1.1441,
+      "step": 7089
+    },
+    {
+      "epoch": 0.6217346803155259,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001292102389437423,
+      "loss": 1.1688,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6218223720899005,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001291709308122179,
+      "loss": 1.1275,
+      "step": 7091
+    },
+    {
+      "epoch": 0.621910063864275,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001291316259482122,
+      "loss": 1.1624,
+      "step": 7092
+    },
+    {
+      "epoch": 0.6219977556386496,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.001290923243553096,
+      "loss": 1.1486,
+      "step": 7093
+    },
+    {
+      "epoch": 0.6220854474130242,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0012905302603709427,
+      "loss": 1.1929,
+      "step": 7094
+    },
+    {
+      "epoch": 0.6221731391873987,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001290137309971502,
+      "loss": 1.2062,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6222608309617732,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0012897443923906096,
+      "loss": 1.1259,
+      "step": 7096
+    },
+    {
+      "epoch": 0.6223485227361478,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001289351507664099,
+      "loss": 1.1629,
+      "step": 7097
+    },
+    {
+      "epoch": 0.6224362145105223,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0012889586558277992,
+      "loss": 1.1184,
+      "step": 7098
+    },
+    {
+      "epoch": 0.6225239062848968,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0012885658369175383,
+      "loss": 1.1373,
+      "step": 7099
+    },
+    {
+      "epoch": 0.6226115980592715,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0012881730509691402,
+      "loss": 1.2196,
+      "step": 7100
+    },
+    {
+      "epoch": 0.622699289833646,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0012877802980184256,
+      "loss": 1.1609,
+      "step": 7101
+    },
+    {
+      "epoch": 0.6227869816080205,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0012873875781012126,
+      "loss": 1.1309,
+      "step": 7102
+    },
+    {
+      "epoch": 0.6228746733823951,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0012869948912533169,
+      "loss": 1.2125,
+      "step": 7103
+    },
+    {
+      "epoch": 0.6229623651567696,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0012866022375105494,
+      "loss": 1.1824,
+      "step": 7104
+    },
+    {
+      "epoch": 0.6230500569311441,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0012862096169087203,
+      "loss": 1.1788,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6231377487055186,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0012858170294836346,
+      "loss": 1.1006,
+      "step": 7106
+    },
+    {
+      "epoch": 0.6232254404798933,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0012854244752710956,
+      "loss": 1.1616,
+      "step": 7107
+    },
+    {
+      "epoch": 0.6233131322542678,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0012850319543069042,
+      "loss": 1.1207,
+      "step": 7108
+    },
+    {
+      "epoch": 0.6234008240286423,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0012846394666268556,
+      "loss": 1.17,
+      "step": 7109
+    },
+    {
+      "epoch": 0.6234885158030169,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0012842470122667443,
+      "loss": 1.1747,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6235762075773914,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0012838545912623618,
+      "loss": 1.1349,
+      "step": 7111
+    },
+    {
+      "epoch": 0.6236638993517659,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0012834622036494952,
+      "loss": 1.195,
+      "step": 7112
+    },
+    {
+      "epoch": 0.6237515911261405,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012830698494639295,
+      "loss": 1.0944,
+      "step": 7113
+    },
+    {
+      "epoch": 0.623839282900515,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001282677528741446,
+      "loss": 1.125,
+      "step": 7114
+    },
+    {
+      "epoch": 0.6239269746748896,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0012822852415178244,
+      "loss": 1.1442,
+      "step": 7115
+    },
+    {
+      "epoch": 0.6240146664492642,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0012818929878288384,
+      "loss": 1.1898,
+      "step": 7116
+    },
+    {
+      "epoch": 0.6241023582236387,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0012815007677102624,
+      "loss": 1.1723,
+      "step": 7117
+    },
+    {
+      "epoch": 0.6241900499980132,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0012811085811978648,
+      "loss": 1.2259,
+      "step": 7118
+    },
+    {
+      "epoch": 0.6242777417723878,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0012807164283274127,
+      "loss": 1.108,
+      "step": 7119
+    },
+    {
+      "epoch": 0.6243654335467623,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0012803243091346684,
+      "loss": 1.1421,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6244531253211368,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0012799322236553936,
+      "loss": 1.153,
+      "step": 7121
+    },
+    {
+      "epoch": 0.6245408170955115,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001279540171925344,
+      "loss": 1.1382,
+      "step": 7122
+    },
+    {
+      "epoch": 0.624628508869886,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0012791481539802748,
+      "loss": 1.1608,
+      "step": 7123
+    },
+    {
+      "epoch": 0.6247162006442605,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0012787561698559362,
+      "loss": 1.1136,
+      "step": 7124
+    },
+    {
+      "epoch": 0.6248038924186351,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.001278364219588077,
+      "loss": 1.2437,
+      "step": 7125
+    },
+    {
+      "epoch": 0.6248915841930096,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0012779723032124417,
+      "loss": 1.1347,
+      "step": 7126
+    },
+    {
+      "epoch": 0.6249792759673841,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0012775804207647717,
+      "loss": 1.1173,
+      "step": 7127
+    },
+    {
+      "epoch": 0.6250669677417587,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001277188572280806,
+      "loss": 1.1661,
+      "step": 7128
+    },
+    {
+      "epoch": 0.6251546595161332,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00127679675779628,
+      "loss": 1.1156,
+      "step": 7129
+    },
+    {
+      "epoch": 0.6252423512905078,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.001276404977346926,
+      "loss": 1.087,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6253300430648823,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012760132309684738,
+      "loss": 1.1409,
+      "step": 7131
+    },
+    {
+      "epoch": 0.6254177348392569,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0012756215186966497,
+      "loss": 1.1705,
+      "step": 7132
+    },
+    {
+      "epoch": 0.6255054266136314,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012752298405671763,
+      "loss": 1.1707,
+      "step": 7133
+    },
+    {
+      "epoch": 0.6255931183880059,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0012748381966157742,
+      "loss": 1.1785,
+      "step": 7134
+    },
+    {
+      "epoch": 0.6256808101623805,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0012744465868781605,
+      "loss": 1.1098,
+      "step": 7135
+    },
+    {
+      "epoch": 0.625768501936755,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0012740550113900478,
+      "loss": 1.0921,
+      "step": 7136
+    },
+    {
+      "epoch": 0.6258561937111295,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001273663470187148,
+      "loss": 1.145,
+      "step": 7137
+    },
+    {
+      "epoch": 0.6259438854855042,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001273271963305168,
+      "loss": 1.1525,
+      "step": 7138
+    },
+    {
+      "epoch": 0.6260315772598787,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0012728804907798127,
+      "loss": 1.1481,
+      "step": 7139
+    },
+    {
+      "epoch": 0.6261192690342532,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001272489052646783,
+      "loss": 1.1546,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6262069608086278,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.001272097648941777,
+      "loss": 1.1193,
+      "step": 7141
+    },
+    {
+      "epoch": 0.6262946525830023,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0012717062797004905,
+      "loss": 1.254,
+      "step": 7142
+    },
+    {
+      "epoch": 0.6263823443573768,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0012713149449586141,
+      "loss": 1.1184,
+      "step": 7143
+    },
+    {
+      "epoch": 0.6264700361317515,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0012709236447518374,
+      "loss": 1.1392,
+      "step": 7144
+    },
+    {
+      "epoch": 0.626557727906126,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0012705323791158455,
+      "loss": 1.1812,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6266454196805005,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0012701411480863212,
+      "loss": 1.1901,
+      "step": 7146
+    },
+    {
+      "epoch": 0.6267331114548751,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0012697499516989435,
+      "loss": 1.1907,
+      "step": 7147
+    },
+    {
+      "epoch": 0.6268208032292496,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0012693587899893887,
+      "loss": 1.1438,
+      "step": 7148
+    },
+    {
+      "epoch": 0.6269084950036241,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0012689676629933295,
+      "loss": 1.1743,
+      "step": 7149
+    },
+    {
+      "epoch": 0.6269961867779987,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001268576570746436,
+      "loss": 1.1376,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6270838785523732,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0012681855132843744,
+      "loss": 1.1764,
+      "step": 7151
+    },
+    {
+      "epoch": 0.6271715703267478,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0012677944906428085,
+      "loss": 1.1481,
+      "step": 7152
+    },
+    {
+      "epoch": 0.6272592621011224,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0012674035028573981,
+      "loss": 1.1515,
+      "step": 7153
+    },
+    {
+      "epoch": 0.6273469538754969,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0012670125499638011,
+      "loss": 1.138,
+      "step": 7154
+    },
+    {
+      "epoch": 0.6274346456498714,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0012666216319976702,
+      "loss": 1.1233,
+      "step": 7155
+    },
+    {
+      "epoch": 0.6275223374242459,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0012662307489946572,
+      "loss": 1.1358,
+      "step": 7156
+    },
+    {
+      "epoch": 0.6276100291986205,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0012658399009904092,
+      "loss": 1.1568,
+      "step": 7157
+    },
+    {
+      "epoch": 0.627697720972995,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0012654490880205705,
+      "loss": 1.1594,
+      "step": 7158
+    },
+    {
+      "epoch": 0.6277854127473695,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0012650583101207827,
+      "loss": 1.1967,
+      "step": 7159
+    },
+    {
+      "epoch": 0.6278731045217442,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0012646675673266827,
+      "loss": 1.1724,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6279607962961187,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001264276859673906,
+      "loss": 1.0737,
+      "step": 7161
+    },
+    {
+      "epoch": 0.6280484880704932,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0012638861871980837,
+      "loss": 1.189,
+      "step": 7162
+    },
+    {
+      "epoch": 0.6281361798448678,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0012634955499348447,
+      "loss": 1.1351,
+      "step": 7163
+    },
+    {
+      "epoch": 0.6282238716192423,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012631049479198137,
+      "loss": 1.1214,
+      "step": 7164
+    },
+    {
+      "epoch": 0.6283115633936168,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0012627143811886124,
+      "loss": 1.1078,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6283992551679914,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0012623238497768603,
+      "loss": 1.1501,
+      "step": 7166
+    },
+    {
+      "epoch": 0.628486946942366,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0012619333537201717,
+      "loss": 1.1204,
+      "step": 7167
+    },
+    {
+      "epoch": 0.6285746387167405,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0012615428930541602,
+      "loss": 1.1181,
+      "step": 7168
+    },
+    {
+      "epoch": 0.6286623304911151,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0012611524678144332,
+      "loss": 1.1198,
+      "step": 7169
+    },
+    {
+      "epoch": 0.6287500222654896,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001260762078036598,
+      "loss": 1.1633,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6288377140398641,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0012603717237562562,
+      "loss": 1.1155,
+      "step": 7171
+    },
+    {
+      "epoch": 0.6289254058142387,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0012599814050090075,
+      "loss": 1.1685,
+      "step": 7172
+    },
+    {
+      "epoch": 0.6290130975886132,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0012595911218304476,
+      "loss": 1.1616,
+      "step": 7173
+    },
+    {
+      "epoch": 0.6291007893629877,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00125920087425617,
+      "loss": 1.1229,
+      "step": 7174
+    },
+    {
+      "epoch": 0.6291884811373624,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0012588106623217632,
+      "loss": 1.1745,
+      "step": 7175
+    },
+    {
+      "epoch": 0.6292761729117369,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001258420486062815,
+      "loss": 1.1212,
+      "step": 7176
+    },
+    {
+      "epoch": 0.6293638646861114,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0012580303455149073,
+      "loss": 1.1196,
+      "step": 7177
+    },
+    {
+      "epoch": 0.6294515564604859,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0012576402407136198,
+      "loss": 1.0892,
+      "step": 7178
+    },
+    {
+      "epoch": 0.6295392482348605,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0012572501716945298,
+      "loss": 1.1233,
+      "step": 7179
+    },
+    {
+      "epoch": 0.629626940009235,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012568601384932102,
+      "loss": 1.1242,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6297146317836095,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0012564701411452311,
+      "loss": 1.2031,
+      "step": 7181
+    },
+    {
+      "epoch": 0.6298023235579842,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0012560801796861587,
+      "loss": 1.1606,
+      "step": 7182
+    },
+    {
+      "epoch": 0.6298900153323587,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0012556902541515577,
+      "loss": 1.1174,
+      "step": 7183
+    },
+    {
+      "epoch": 0.6299777071067332,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0012553003645769873,
+      "loss": 1.1626,
+      "step": 7184
+    },
+    {
+      "epoch": 0.6300653988811078,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0012549105109980051,
+      "loss": 1.0986,
+      "step": 7185
+    },
+    {
+      "epoch": 0.6301530906554823,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001254520693450164,
+      "loss": 1.1397,
+      "step": 7186
+    },
+    {
+      "epoch": 0.6302407824298568,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0012541309119690146,
+      "loss": 1.0868,
+      "step": 7187
+    },
+    {
+      "epoch": 0.6303284742042314,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0012537411665901046,
+      "loss": 1.1596,
+      "step": 7188
+    },
+    {
+      "epoch": 0.630416165978606,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012533514573489766,
+      "loss": 1.0838,
+      "step": 7189
+    },
+    {
+      "epoch": 0.6305038577529805,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0012529617842811725,
+      "loss": 1.115,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6305915495273551,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001252572147422228,
+      "loss": 1.1466,
+      "step": 7191
+    },
+    {
+      "epoch": 0.6306792413017296,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012521825468076783,
+      "loss": 1.1557,
+      "step": 7192
+    },
+    {
+      "epoch": 0.6307669330761041,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.001251792982473054,
+      "loss": 1.114,
+      "step": 7193
+    },
+    {
+      "epoch": 0.6308546248504787,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0012514034544538805,
+      "loss": 1.0619,
+      "step": 7194
+    },
+    {
+      "epoch": 0.6309423166248532,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0012510139627856835,
+      "loss": 1.1903,
+      "step": 7195
+    },
+    {
+      "epoch": 0.6310300083992277,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0012506245075039827,
+      "loss": 1.1645,
+      "step": 7196
+    },
+    {
+      "epoch": 0.6311177001736024,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0012502350886442963,
+      "loss": 1.1891,
+      "step": 7197
+    },
+    {
+      "epoch": 0.6312053919479769,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0012498457062421378,
+      "loss": 1.1088,
+      "step": 7198
+    },
+    {
+      "epoch": 0.6312930837223514,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0012494563603330175,
+      "loss": 1.2219,
+      "step": 7199
+    },
+    {
+      "epoch": 0.631380775496726,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0012490670509524434,
+      "loss": 1.1092,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6314684672711005,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0012486777781359195,
+      "loss": 1.2014,
+      "step": 7201
+    },
+    {
+      "epoch": 0.631556159045475,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0012482885419189459,
+      "loss": 1.1901,
+      "step": 7202
+    },
+    {
+      "epoch": 0.6316438508198495,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012478993423370205,
+      "loss": 1.1738,
+      "step": 7203
+    },
+    {
+      "epoch": 0.6317315425942241,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0012475101794256365,
+      "loss": 1.1432,
+      "step": 7204
+    },
+    {
+      "epoch": 0.6318192343685987,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0012471210532202857,
+      "loss": 1.1686,
+      "step": 7205
+    },
+    {
+      "epoch": 0.6319069261429732,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0012467319637564543,
+      "loss": 1.1431,
+      "step": 7206
+    },
+    {
+      "epoch": 0.6319946179173478,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0012463429110696272,
+      "loss": 1.1414,
+      "step": 7207
+    },
+    {
+      "epoch": 0.6320823096917223,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0012459538951952844,
+      "loss": 1.1469,
+      "step": 7208
+    },
+    {
+      "epoch": 0.6321700014660968,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0012455649161689033,
+      "loss": 1.1107,
+      "step": 7209
+    },
+    {
+      "epoch": 0.6322576932404714,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001245175974025959,
+      "loss": 1.1063,
+      "step": 7210
+    },
+    {
+      "epoch": 0.6323453850148459,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0012447870688019193,
+      "loss": 1.1451,
+      "step": 7211
+    },
+    {
+      "epoch": 0.6324330767892204,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0012443982005322534,
+      "loss": 1.1692,
+      "step": 7212
+    },
+    {
+      "epoch": 0.6325207685635951,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012440093692524241,
+      "loss": 1.1065,
+      "step": 7213
+    },
+    {
+      "epoch": 0.6326084603379696,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0012436205749978928,
+      "loss": 1.0812,
+      "step": 7214
+    },
+    {
+      "epoch": 0.6326961521123441,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0012432318178041157,
+      "loss": 1.1128,
+      "step": 7215
+    },
+    {
+      "epoch": 0.6327838438867187,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0012428430977065466,
+      "loss": 1.1139,
+      "step": 7216
+    },
+    {
+      "epoch": 0.6328715356610932,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012424544147406364,
+      "loss": 1.0918,
+      "step": 7217
+    },
+    {
+      "epoch": 0.6329592274354677,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0012420657689418315,
+      "loss": 1.142,
+      "step": 7218
+    },
+    {
+      "epoch": 0.6330469192098424,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0012416771603455752,
+      "loss": 1.1275,
+      "step": 7219
+    },
+    {
+      "epoch": 0.6331346109842169,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0012412885889873074,
+      "loss": 1.1352,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6332223027585914,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001240900054902466,
+      "loss": 1.1584,
+      "step": 7221
+    },
+    {
+      "epoch": 0.633309994532966,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0012405115581264829,
+      "loss": 1.1348,
+      "step": 7222
+    },
+    {
+      "epoch": 0.6333976863073405,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0012401230986947885,
+      "loss": 1.1354,
+      "step": 7223
+    },
+    {
+      "epoch": 0.633485378081715,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0012397346766428099,
+      "loss": 1.1097,
+      "step": 7224
+    },
+    {
+      "epoch": 0.6335730698560896,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0012393462920059696,
+      "loss": 1.168,
+      "step": 7225
+    },
+    {
+      "epoch": 0.6336607616304641,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0012389579448196868,
+      "loss": 1.1622,
+      "step": 7226
+    },
+    {
+      "epoch": 0.6337484534048387,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0012385696351193795,
+      "loss": 1.1357,
+      "step": 7227
+    },
+    {
+      "epoch": 0.6338361451792132,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012381813629404586,
+      "loss": 1.0921,
+      "step": 7228
+    },
+    {
+      "epoch": 0.6339238369535878,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0012377931283183345,
+      "loss": 1.194,
+      "step": 7229
+    },
+    {
+      "epoch": 0.6340115287279623,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0012374049312884128,
+      "loss": 1.1458,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6340992205023368,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0012370167718860963,
+      "loss": 1.1036,
+      "step": 7231
+    },
+    {
+      "epoch": 0.6341869122767114,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0012366286501467843,
+      "loss": 1.0911,
+      "step": 7232
+    },
+    {
+      "epoch": 0.6342746040510859,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012362405661058718,
+      "loss": 1.1664,
+      "step": 7233
+    },
+    {
+      "epoch": 0.6343622958254604,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0012358525197987522,
+      "loss": 1.1619,
+      "step": 7234
+    },
+    {
+      "epoch": 0.6344499875998351,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0012354645112608131,
+      "loss": 1.0955,
+      "step": 7235
+    },
+    {
+      "epoch": 0.6345376793742096,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0012350765405274406,
+      "loss": 1.1524,
+      "step": 7236
+    },
+    {
+      "epoch": 0.6346253711485841,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0012346886076340167,
+      "loss": 1.2004,
+      "step": 7237
+    },
+    {
+      "epoch": 0.6347130629229587,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0012343007126159196,
+      "loss": 1.1758,
+      "step": 7238
+    },
+    {
+      "epoch": 0.6348007546973332,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0012339128555085245,
+      "loss": 1.122,
+      "step": 7239
+    },
+    {
+      "epoch": 0.6348884464717077,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0012335250363472024,
+      "loss": 1.1258,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6349761382460823,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001233137255167322,
+      "loss": 1.1516,
+      "step": 7241
+    },
+    {
+      "epoch": 0.6350638300204569,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001232749512004248,
+      "loss": 1.1319,
+      "step": 7242
+    },
+    {
+      "epoch": 0.6351515217948314,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0012323618068933412,
+      "loss": 1.1533,
+      "step": 7243
+    },
+    {
+      "epoch": 0.635239213569206,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0012319741398699603,
+      "loss": 1.1871,
+      "step": 7244
+    },
+    {
+      "epoch": 0.6353269053435805,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0012315865109694577,
+      "loss": 1.1659,
+      "step": 7245
+    },
+    {
+      "epoch": 0.635414597117955,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0012311989202271859,
+      "loss": 1.1952,
+      "step": 7246
+    },
+    {
+      "epoch": 0.6355022888923296,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0012308113676784906,
+      "loss": 1.1106,
+      "step": 7247
+    },
+    {
+      "epoch": 0.6355899806667041,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001230423853358717,
+      "loss": 1.1539,
+      "step": 7248
+    },
+    {
+      "epoch": 0.6356776724410786,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0012300363773032048,
+      "loss": 1.1665,
+      "step": 7249
+    },
+    {
+      "epoch": 0.6357653642154532,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001229648939547291,
+      "loss": 1.2366,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6358530559898278,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001229261540126308,
+      "loss": 1.1753,
+      "step": 7251
+    },
+    {
+      "epoch": 0.6359407477642023,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0012288741790755873,
+      "loss": 1.1539,
+      "step": 7252
+    },
+    {
+      "epoch": 0.6360284395385768,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0012284868564304539,
+      "loss": 1.1168,
+      "step": 7253
+    },
+    {
+      "epoch": 0.6361161313129514,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0012280995722262313,
+      "loss": 1.1243,
+      "step": 7254
+    },
+    {
+      "epoch": 0.6362038230873259,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0012277123264982385,
+      "loss": 1.18,
+      "step": 7255
+    },
+    {
+      "epoch": 0.6362915148617004,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0012273251192817913,
+      "loss": 1.179,
+      "step": 7256
+    },
+    {
+      "epoch": 0.6363792066360751,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.001226937950612202,
+      "loss": 1.0844,
+      "step": 7257
+    },
+    {
+      "epoch": 0.6364668984104496,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0012265508205247797,
+      "loss": 1.2039,
+      "step": 7258
+    },
+    {
+      "epoch": 0.6365545901848241,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0012261637290548293,
+      "loss": 1.1832,
+      "step": 7259
+    },
+    {
+      "epoch": 0.6366422819591987,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0012257766762376527,
+      "loss": 1.1512,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6367299737335732,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012253896621085487,
+      "loss": 1.1915,
+      "step": 7261
+    },
+    {
+      "epoch": 0.6368176655079477,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0012250026867028105,
+      "loss": 1.1925,
+      "step": 7262
+    },
+    {
+      "epoch": 0.6369053572823223,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0012246157500557307,
+      "loss": 1.1498,
+      "step": 7263
+    },
+    {
+      "epoch": 0.6369930490566968,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0012242288522025962,
+      "loss": 1.1403,
+      "step": 7264
+    },
+    {
+      "epoch": 0.6370807408310714,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012238419931786912,
+      "loss": 1.1598,
+      "step": 7265
+    },
+    {
+      "epoch": 0.637168432605446,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0012234551730192962,
+      "loss": 1.1859,
+      "step": 7266
+    },
+    {
+      "epoch": 0.6372561243798205,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0012230683917596884,
+      "loss": 1.1658,
+      "step": 7267
+    },
+    {
+      "epoch": 0.637343816154195,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0012226816494351413,
+      "loss": 1.164,
+      "step": 7268
+    },
+    {
+      "epoch": 0.6374315079285696,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001222294946080924,
+      "loss": 1.1382,
+      "step": 7269
+    },
+    {
+      "epoch": 0.6375191997029441,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0012219082817323042,
+      "loss": 1.1505,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6376068914773186,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0012215216564245438,
+      "loss": 1.1261,
+      "step": 7271
+    },
+    {
+      "epoch": 0.6376945832516933,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0012211350701929024,
+      "loss": 1.2135,
+      "step": 7272
+    },
+    {
+      "epoch": 0.6377822750260678,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0012207485230726351,
+      "loss": 1.1571,
+      "step": 7273
+    },
+    {
+      "epoch": 0.6378699668004423,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001220362015098995,
+      "loss": 1.11,
+      "step": 7274
+    },
+    {
+      "epoch": 0.6379576585748168,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0012199755463072292,
+      "loss": 1.1104,
+      "step": 7275
+    },
+    {
+      "epoch": 0.6380453503491914,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0012195891167325844,
+      "loss": 1.1428,
+      "step": 7276
+    },
+    {
+      "epoch": 0.6381330421235659,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0012192027264103001,
+      "loss": 1.1711,
+      "step": 7277
+    },
+    {
+      "epoch": 0.6382207338979404,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0012188163753756157,
+      "loss": 1.145,
+      "step": 7278
+    },
+    {
+      "epoch": 0.638308425672315,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001218430063663765,
+      "loss": 1.18,
+      "step": 7279
+    },
+    {
+      "epoch": 0.6383961174466896,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.001218043791309978,
+      "loss": 1.1248,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6384838092210641,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0012176575583494821,
+      "loss": 1.1487,
+      "step": 7281
+    },
+    {
+      "epoch": 0.6385715009954387,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0012172713648175007,
+      "loss": 1.1715,
+      "step": 7282
+    },
+    {
+      "epoch": 0.6386591927698132,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0012168852107492538,
+      "loss": 1.1345,
+      "step": 7283
+    },
+    {
+      "epoch": 0.6387468845441877,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0012164990961799579,
+      "loss": 1.1468,
+      "step": 7284
+    },
+    {
+      "epoch": 0.6388345763185623,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001216113021144825,
+      "loss": 1.2466,
+      "step": 7285
+    },
+    {
+      "epoch": 0.6389222680929368,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0012157269856790647,
+      "loss": 1.1641,
+      "step": 7286
+    },
+    {
+      "epoch": 0.6390099598673114,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0012153409898178823,
+      "loss": 1.1354,
+      "step": 7287
+    },
+    {
+      "epoch": 0.639097651641686,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001214955033596479,
+      "loss": 1.1205,
+      "step": 7288
+    },
+    {
+      "epoch": 0.6391853434160605,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001214569117050054,
+      "loss": 1.1036,
+      "step": 7289
+    },
+    {
+      "epoch": 0.639273035190435,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0012141832402138017,
+      "loss": 1.133,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6393607269648096,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0012137974031229126,
+      "loss": 1.1065,
+      "step": 7291
+    },
+    {
+      "epoch": 0.6394484187391841,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001213411605812574,
+      "loss": 1.119,
+      "step": 7292
+    },
+    {
+      "epoch": 0.6395361105135586,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0012130258483179703,
+      "loss": 1.1227,
+      "step": 7293
+    },
+    {
+      "epoch": 0.6396238022879333,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.001212640130674281,
+      "loss": 1.1007,
+      "step": 7294
+    },
+    {
+      "epoch": 0.6397114940623078,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0012122544529166832,
+      "loss": 1.159,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6397991858366823,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0012118688150803492,
+      "loss": 1.1491,
+      "step": 7296
+    },
+    {
+      "epoch": 0.6398868776110568,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0012114832172004477,
+      "loss": 1.1251,
+      "step": 7297
+    },
+    {
+      "epoch": 0.6399745693854314,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0012110976593121448,
+      "loss": 1.1067,
+      "step": 7298
+    },
+    {
+      "epoch": 0.6400622611598059,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0012107121414506024,
+      "loss": 1.1571,
+      "step": 7299
+    },
+    {
+      "epoch": 0.6401499529341804,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0012103266636509789,
+      "loss": 1.0881,
+      "step": 7300
+    },
+    {
+      "epoch": 0.640237644708555,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0012099412259484288,
+      "loss": 1.0955,
+      "step": 7301
+    },
+    {
+      "epoch": 0.6403253364829296,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0012095558283781022,
+      "loss": 1.1745,
+      "step": 7302
+    },
+    {
+      "epoch": 0.6404130282573041,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0012091704709751479,
+      "loss": 1.1186,
+      "step": 7303
+    },
+    {
+      "epoch": 0.6405007200316787,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001208785153774708,
+      "loss": 1.0849,
+      "step": 7304
+    },
+    {
+      "epoch": 0.6405884118060532,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0012083998768119234,
+      "loss": 1.1483,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6406761035804277,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00120801464012193,
+      "loss": 1.1379,
+      "step": 7306
+    },
+    {
+      "epoch": 0.6407637953548023,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0012076294437398606,
+      "loss": 1.1658,
+      "step": 7307
+    },
+    {
+      "epoch": 0.6408514871291768,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0012072442877008435,
+      "loss": 1.1676,
+      "step": 7308
+    },
+    {
+      "epoch": 0.6409391789035513,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0012068591720400049,
+      "loss": 1.173,
+      "step": 7309
+    },
+    {
+      "epoch": 0.641026870677926,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0012064740967924655,
+      "loss": 1.1585,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6411145624523005,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0012060890619933437,
+      "loss": 1.1952,
+      "step": 7311
+    },
+    {
+      "epoch": 0.641202254226675,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0012057040676777537,
+      "loss": 1.1012,
+      "step": 7312
+    },
+    {
+      "epoch": 0.6412899460010496,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0012053191138808057,
+      "loss": 1.1457,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6413776377754241,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0012049342006376065,
+      "loss": 1.1382,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6414653295497986,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001204549327983259,
+      "loss": 1.1487,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6415530213241732,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001204164495952863,
+      "loss": 1.1955,
+      "step": 7316
+    },
+    {
+      "epoch": 0.6416407130985478,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0012037797045815139,
+      "loss": 1.1475,
+      "step": 7317
+    },
+    {
+      "epoch": 0.6417284048729223,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0012033949539043045,
+      "loss": 1.1787,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6418160966472969,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0012030102439563217,
+      "loss": 1.1177,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6419037884216714,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.001202625574772651,
+      "loss": 1.1268,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6419914801960459,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0012022409463883729,
+      "loss": 1.1627,
+      "step": 7321
+    },
+    {
+      "epoch": 0.6420791719704204,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0012018563588385648,
+      "loss": 1.136,
+      "step": 7322
+    },
+    {
+      "epoch": 0.642166863744795,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0012014718121582996,
+      "loss": 1.1379,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6422545555191695,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.001201087306382647,
+      "loss": 1.1718,
+      "step": 7324
+    },
+    {
+      "epoch": 0.642342247293544,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001200702841546674,
+      "loss": 1.0806,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6424299390679187,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012003184176854415,
+      "loss": 1.1306,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6425176308422932,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0011999340348340084,
+      "loss": 1.1772,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6426053226166677,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0011995496930274298,
+      "loss": 1.0925,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6426930143910423,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011991653923007564,
+      "loss": 1.1413,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6427807061654168,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0011987811326890356,
+      "loss": 1.1961,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6428683979397913,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0011983969142273102,
+      "loss": 1.1505,
+      "step": 7331
+    },
+    {
+      "epoch": 0.642956089714166,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0011980127369506206,
+      "loss": 1.1326,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6430437814885405,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0011976286008940022,
+      "loss": 1.129,
+      "step": 7333
+    },
+    {
+      "epoch": 0.643131473262915,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0011972445060924885,
+      "loss": 1.147,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6432191650372896,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0011968604525811062,
+      "loss": 1.1447,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6433068568116641,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001196476440394881,
+      "loss": 1.1126,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6433945485860386,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0011960924695688338,
+      "loss": 1.1944,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6434822403604132,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0011957085401379816,
+      "loss": 1.1263,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6435699321347877,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.001195324652137338,
+      "loss": 1.1204,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6436576239091623,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0011949408056019125,
+      "loss": 1.0916,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6437453156835369,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00119455700056671,
+      "loss": 1.0883,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6438330074579114,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0011941732370667338,
+      "loss": 1.1117,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6439206992322859,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0011937895151369817,
+      "loss": 1.197,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6440083910066605,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0011934058348124484,
+      "loss": 1.1885,
+      "step": 7344
+    },
+    {
+      "epoch": 0.644096082781035,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0011930221961281241,
+      "loss": 1.1365,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6441837745554095,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.001192638599118996,
+      "loss": 1.1672,
+      "step": 7346
+    },
+    {
+      "epoch": 0.644271466329784,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0011922550438200465,
+      "loss": 1.1877,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6443591581041587,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0011918715302662555,
+      "loss": 1.179,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6444468498785332,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0011914880584925984,
+      "loss": 1.1783,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6445345416529077,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0011911046285340468,
+      "loss": 1.1803,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6446222334272823,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0011907212404255682,
+      "loss": 1.1334,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6447099252016568,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001190337894202127,
+      "loss": 1.0958,
+      "step": 7352
+    },
+    {
+      "epoch": 0.6447976169760313,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001189954589898683,
+      "loss": 1.1119,
+      "step": 7353
+    },
+    {
+      "epoch": 0.644885308750406,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001189571327550193,
+      "loss": 1.177,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6449730005247805,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0011891881071916092,
+      "loss": 1.1989,
+      "step": 7355
+    },
+    {
+      "epoch": 0.645060692299155,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001188804928857881,
+      "loss": 1.1495,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6451483840735296,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0011884217925839519,
+      "loss": 1.1861,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6452360758479041,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0011880386984047643,
+      "loss": 1.1457,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6453237676222786,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0011876556463552544,
+      "loss": 1.1559,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6454114593966532,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0011872726364703567,
+      "loss": 1.1569,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6454991511710277,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0011868896687849994,
+      "loss": 1.1296,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6455868429454023,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0011865067433341093,
+      "loss": 1.1266,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6456745347197769,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001186123860152608,
+      "loss": 1.176,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6457622264941514,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0011857410192754124,
+      "loss": 1.1469,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6458499182685259,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0011853582207374375,
+      "loss": 1.1658,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6459376100429005,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0011849754645735936,
+      "loss": 1.1004,
+      "step": 7366
+    },
+    {
+      "epoch": 0.646025301817275,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0011845927508187869,
+      "loss": 1.1537,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6461129935916495,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0011842100795079196,
+      "loss": 1.1099,
+      "step": 7368
+    },
+    {
+      "epoch": 0.646200685366024,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0011838274506758906,
+      "loss": 1.1873,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6462883771403987,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001183444864357595,
+      "loss": 1.1034,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6463760689147732,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.001183062320587923,
+      "loss": 1.0943,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6464637606891477,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0011826798194017622,
+      "loss": 1.1303,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6465514524635223,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0011822973608339956,
+      "loss": 1.1401,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6466391442378968,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001181914944919502,
+      "loss": 1.1241,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6467268360122713,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0011815325716931573,
+      "loss": 1.0903,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6468145277866459,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001181150241189833,
+      "loss": 1.1764,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6469022195610205,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0011807679534443956,
+      "loss": 1.1862,
+      "step": 7377
+    },
+    {
+      "epoch": 0.646989911335395,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0011803857084917101,
+      "loss": 1.1054,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6470776031097696,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0011800035063666353,
+      "loss": 1.1412,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6471652948841441,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.001179621347104028,
+      "loss": 1.1472,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6472529866585186,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0011792392307387396,
+      "loss": 1.1687,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6473406784328932,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001178857157305618,
+      "loss": 1.1452,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6474283702072677,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0011784751268395068,
+      "loss": 1.1099,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6475160619816422,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0011780931393752473,
+      "loss": 1.1205,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6476037537560169,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.001177711194947675,
+      "loss": 1.0971,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6476914455303914,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0011773292935916228,
+      "loss": 1.1128,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6477791373047659,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0011769474353419188,
+      "loss": 1.1382,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6478668290791405,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.001176565620233387,
+      "loss": 1.1079,
+      "step": 7388
+    },
+    {
+      "epoch": 0.647954520853515,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.001176183848300849,
+      "loss": 1.1177,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6480422126278895,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0011758021195791206,
+      "loss": 1.1391,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6481299044022641,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0011754204341030152,
+      "loss": 1.1471,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6482175961766387,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0011750387919073405,
+      "loss": 1.1572,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6483052879510132,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0011746571930269024,
+      "loss": 1.1009,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6483929797253877,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0011742756374965007,
+      "loss": 1.1104,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6484806714997623,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0011738941253509334,
+      "loss": 1.1098,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6485683632741368,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0011735126566249927,
+      "loss": 1.1253,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6486560550485113,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0011731312313534674,
+      "loss": 1.1893,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6487437468228859,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0011727498495711427,
+      "loss": 1.1991,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6488314385972604,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0011723685113127998,
+      "loss": 1.1673,
+      "step": 7399
+    },
+    {
+      "epoch": 0.648919130371635,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0011719872166132156,
+      "loss": 1.1671,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6490068221460096,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0011716059655071634,
+      "loss": 1.1135,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6490945139203841,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0011712247580294122,
+      "loss": 1.1561,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6491822056947586,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001170843594214727,
+      "loss": 1.1035,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6492698974691332,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0011704624740978694,
+      "loss": 1.1649,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6493575892435077,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0011700813977135955,
+      "loss": 1.1874,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6494452810178822,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0011697003650966596,
+      "loss": 1.1339,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6495329727922569,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0011693193762818109,
+      "loss": 1.0972,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6496206645666314,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001168938431303794,
+      "loss": 1.1548,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6497083563410059,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00116855753019735,
+      "loss": 1.1354,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6497960481153805,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0011681766729972165,
+      "loss": 1.1527,
+      "step": 7410
+    },
+    {
+      "epoch": 0.649883739889755,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001167795859738127,
+      "loss": 1.1623,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6499714316641295,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0011674150904548099,
+      "loss": 1.1526,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6500591234385041,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0011670343651819912,
+      "loss": 1.1589,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6501468152128786,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0011666536839543913,
+      "loss": 1.1982,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6502345069872532,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0011662730468067271,
+      "loss": 1.1712,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6503221987616278,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0011658924537737132,
+      "loss": 1.1435,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6504098905360023,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0011655119048900574,
+      "loss": 1.1254,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6504975823103768,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0011651314001904654,
+      "loss": 1.1376,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6505852740847513,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0011647509397096378,
+      "loss": 1.1826,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6506729658591259,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001164370523482272,
+      "loss": 1.1737,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6507606576335004,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0011639901515430604,
+      "loss": 1.1658,
+      "step": 7421
+    },
+    {
+      "epoch": 0.650848349407875,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0011636098239266928,
+      "loss": 1.1536,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6509360411822496,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0011632295406678534,
+      "loss": 1.11,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6510237329566241,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0011628493018012236,
+      "loss": 1.1376,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6511114247309986,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0011624691073614797,
+      "loss": 1.1156,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6511991165053732,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001162088957383295,
+      "loss": 1.168,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6512868082797477,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0011617088519013368,
+      "loss": 1.1781,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6513745000541222,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.001161328790950272,
+      "loss": 1.117,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6514621918284969,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.001160948774564759,
+      "loss": 1.1634,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6515498836028714,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0011605688027794563,
+      "loss": 1.1364,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6516375753772459,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0011601888756290152,
+      "loss": 1.1498,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6517252671516205,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0011598089931480833,
+      "loss": 1.2154,
+      "step": 7432
+    },
+    {
+      "epoch": 0.651812958925995,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0011594291553713065,
+      "loss": 1.1365,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6519006507003695,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0011590493623333238,
+      "loss": 1.086,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6519883424747441,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011586696140687722,
+      "loss": 1.1691,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6520760342491186,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0011582899106122833,
+      "loss": 1.1705,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6521637260234932,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011579102519984855,
+      "loss": 1.1443,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6522514177978678,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0011575306382620022,
+      "loss": 1.1842,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6523391095722423,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0011571510694374533,
+      "loss": 1.1573,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6524268013466168,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0011567715455594548,
+      "loss": 1.1478,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6525144931209913,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001156392066662618,
+      "loss": 1.1261,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6526021848953659,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0011560126327815503,
+      "loss": 1.1636,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6526898766697404,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0011556332439508555,
+      "loss": 1.1806,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6527775684441149,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0011552539002051326,
+      "loss": 1.0957,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6528652602184896,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0011548746015789767,
+      "loss": 1.1387,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6529529519928641,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0011544953481069794,
+      "loss": 1.1024,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6530406437672386,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001154116139823727,
+      "loss": 1.1576,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6531283355416132,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0011537369767638025,
+      "loss": 1.1527,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6532160273159877,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011533578589617845,
+      "loss": 1.1353,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6533037190903622,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0011529787864522475,
+      "loss": 1.1612,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6533914108647368,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0011525997592697625,
+      "loss": 1.1254,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6534791026391114,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.001152220777448895,
+      "loss": 1.1283,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6535667944134859,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.001151841841024208,
+      "loss": 1.1263,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6536544861878605,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001151462950030259,
+      "loss": 1.1161,
+      "step": 7454
+    },
+    {
+      "epoch": 0.653742177962235,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001151084104501602,
+      "loss": 1.1439,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6538298697366095,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0011507053044727865,
+      "loss": 1.1595,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6539175615109841,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0011503265499783588,
+      "loss": 1.2589,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6540052532853586,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0011499478410528592,
+      "loss": 1.1445,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6540929450597331,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.001149569177730826,
+      "loss": 1.1383,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6541806368341078,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.001149190560046792,
+      "loss": 1.1329,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6542683286084823,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0011488119880352858,
+      "loss": 1.1326,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6543560203828568,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0011484334617308325,
+      "loss": 1.0822,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6544437121572314,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001148054981167953,
+      "loss": 1.1518,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6545314039316059,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0011476765463811634,
+      "loss": 1.0807,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6546190957059804,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0011472981574049764,
+      "loss": 1.2032,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6547067874803549,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001146919814273899,
+      "loss": 1.1183,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6547944792547296,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0011465415170224362,
+      "loss": 1.1071,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6548821710291041,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0011461632656850872,
+      "loss": 1.1514,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6549698628034786,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0011457850602963479,
+      "loss": 1.1467,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6550575545778532,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0011454069008907093,
+      "loss": 1.0881,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6551452463522277,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0011450287875026592,
+      "loss": 1.1201,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6552329381266022,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0011446507201666792,
+      "loss": 1.1515,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6553206299009768,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0011442726989172498,
+      "loss": 1.1147,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6554083216753513,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0011438947237888443,
+      "loss": 1.0955,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6554960134497259,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0011435167948159336,
+      "loss": 1.1173,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6555837052241005,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0011431389120329838,
+      "loss": 1.1149,
+      "step": 7476
+    },
+    {
+      "epoch": 0.655671396998475,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0011427610754744567,
+      "loss": 1.1912,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6557590887728495,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0011423832851748098,
+      "loss": 1.0817,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6558467805472241,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0011420055411684976,
+      "loss": 1.1004,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6559344723215986,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0011416278434899679,
+      "loss": 1.0793,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6560221640959731,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011412501921736673,
+      "loss": 1.1459,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6561098558703478,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0011408725872540354,
+      "loss": 1.1544,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6561975476447223,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0011404950287655093,
+      "loss": 1.1389,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6562852394190968,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0011401175167425216,
+      "loss": 1.1658,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6563729311934714,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0011397400512194993,
+      "loss": 1.093,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6564606229678459,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0011393626322308674,
+      "loss": 1.1317,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6565483147422204,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0011389852598110454,
+      "loss": 1.0885,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6566360065165949,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0011386079339944487,
+      "loss": 1.1275,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6567236982909695,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0011382306548154883,
+      "loss": 1.053,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6568113900653441,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0011378534223085706,
+      "loss": 1.1561,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6568990818397186,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001137476236508099,
+      "loss": 1.147,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6569867736140932,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001137099097448471,
+      "loss": 1.1732,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6570744653884677,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0011367220051640822,
+      "loss": 1.1763,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6571621571628422,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001136344959689321,
+      "loss": 1.1622,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6572498489372168,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0011359679610585735,
+      "loss": 1.1408,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6573375407115913,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001135591009306221,
+      "loss": 1.1771,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6574252324859658,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0011352141044666405,
+      "loss": 1.0935,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6575129242603405,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0011348372465742051,
+      "loss": 1.1131,
+      "step": 7498
+    },
+    {
+      "epoch": 0.657600616034715,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001134460435663283,
+      "loss": 1.1851,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6576883078090895,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0011340836717682377,
+      "loss": 1.1702,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6576883078090895,
+      "eval_loss": 1.1537786722183228,
+      "eval_runtime": 428.7397,
+      "eval_samples_per_second": 33.696,
+      "eval_steps_per_second": 8.425,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6577759995834641,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00113370695492343,
+      "loss": 1.1107,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6578636913578386,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.001133330285163215,
+      "loss": 1.1474,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6579513831322131,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0011329536625219441,
+      "loss": 1.1292,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6580390749065878,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001132577087033965,
+      "loss": 1.1498,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6581267666809623,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0011322005587336196,
+      "loss": 1.1462,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6582144584553368,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0011318240776552466,
+      "loss": 1.1104,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6583021502297114,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0011314476438331797,
+      "loss": 1.1768,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6583898420040859,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0011310712573017496,
+      "loss": 1.1191,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6584775337784604,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0011306949180952806,
+      "loss": 1.1077,
+      "step": 7509
+    },
+    {
+      "epoch": 0.658565225552835,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.001130318626248095,
+      "loss": 1.1859,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6586529173272095,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0011299423817945087,
+      "loss": 1.1059,
+      "step": 7511
+    },
+    {
+      "epoch": 0.658740609101584,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0011295661847688346,
+      "loss": 1.1432,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6588283008759586,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0011291900352053815,
+      "loss": 1.1918,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6589159926503332,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0011288139331384526,
+      "loss": 1.1305,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6590036844247077,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0011284378786023477,
+      "loss": 1.1144,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6590913761990822,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0011280618716313616,
+      "loss": 1.1225,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6591790679734568,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0011276859122597853,
+      "loss": 1.1641,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6592667597478313,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0011273100005219057,
+      "loss": 1.1089,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6593544515222058,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0011269341364520047,
+      "loss": 1.1493,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6594421432965805,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00112655832008436,
+      "loss": 1.1403,
+      "step": 7520
+    },
+    {
+      "epoch": 0.659529835070955,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0011261825514532454,
+      "loss": 1.1257,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6596175268453295,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00112580683059293,
+      "loss": 1.1204,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6597052186197041,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0011254311575376779,
+      "loss": 1.1292,
+      "step": 7523
+    },
+    {
+      "epoch": 0.6597929103940786,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0011250555323217507,
+      "loss": 1.1125,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6598806021684531,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0011246799549794032,
+      "loss": 1.156,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6599682939428277,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0011243044255448884,
+      "loss": 1.1471,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6600559857172023,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0011239289440524523,
+      "loss": 1.1821,
+      "step": 7527
+    },
+    {
+      "epoch": 0.6601436774915768,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001123553510536339,
+      "loss": 1.1382,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6602313692659514,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001123178125030786,
+      "loss": 1.1097,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6603190610403259,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0011228027875700285,
+      "loss": 1.1387,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6604067528147004,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0011224274981882954,
+      "loss": 1.1625,
+      "step": 7531
+    },
+    {
+      "epoch": 0.660494444589075,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0011220522569198135,
+      "loss": 1.1831,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6605821363634495,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0011216770637988017,
+      "loss": 1.1244,
+      "step": 7533
+    },
+    {
+      "epoch": 0.660669828137824,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001121301918859478,
+      "loss": 1.1305,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6607575199121987,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.001120926822136055,
+      "loss": 1.1568,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6608452116865732,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0011205517736627398,
+      "loss": 1.1298,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6609329034609477,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0011201767734737359,
+      "loss": 1.1092,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6610205952353222,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0011198018216032428,
+      "loss": 1.118,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6611082870096968,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0011194269180854546,
+      "loss": 1.0887,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6611959787840713,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0011190520629545619,
+      "loss": 1.1448,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6612836705584458,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0011186772562447502,
+      "loss": 1.1428,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6613713623328205,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0011183024979902014,
+      "loss": 1.1733,
+      "step": 7542
+    },
+    {
+      "epoch": 0.661459054107195,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.001117927788225092,
+      "loss": 1.1467,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6615467458815695,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0011175531269835952,
+      "loss": 1.2025,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6616344376559441,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0011171785142998784,
+      "loss": 1.1294,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6617221294303186,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0011168039502081055,
+      "loss": 1.1198,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6618098212046931,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.001116429434742436,
+      "loss": 1.1813,
+      "step": 7547
+    },
+    {
+      "epoch": 0.6618975129790677,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.001116054967937025,
+      "loss": 1.1119,
+      "step": 7548
+    },
+    {
+      "epoch": 0.6619852047534422,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0011156805498260226,
+      "loss": 1.1233,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6620728965278168,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0011153061804435743,
+      "loss": 1.1417,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6621605883021914,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0011149318598238226,
+      "loss": 1.1121,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6622482800765659,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0011145575880009034,
+      "loss": 1.2055,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6623359718509404,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0011141833650089499,
+      "loss": 1.1156,
+      "step": 7553
+    },
+    {
+      "epoch": 0.662423663625315,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011138091908820906,
+      "loss": 1.1669,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6625113553996895,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001113435065654449,
+      "loss": 1.1295,
+      "step": 7555
+    },
+    {
+      "epoch": 0.662599047174064,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0011130609893601439,
+      "loss": 1.1014,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6626867389484387,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0011126869620332908,
+      "loss": 1.1427,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6627744307228132,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0011123129837079992,
+      "loss": 1.1361,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6628621224971877,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001111939054418376,
+      "loss": 1.1818,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6629498142715622,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0011115651741985218,
+      "loss": 1.1468,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6630375060459368,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001111191343082534,
+      "loss": 1.1611,
+      "step": 7561
+    },
+    {
+      "epoch": 0.6631251978203113,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011108175611045045,
+      "loss": 1.1583,
+      "step": 7562
+    },
+    {
+      "epoch": 0.6632128895946858,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0011104438282985218,
+      "loss": 1.1376,
+      "step": 7563
+    },
+    {
+      "epoch": 0.6633005813690604,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.001110070144698669,
+      "loss": 1.1165,
+      "step": 7564
+    },
+    {
+      "epoch": 0.663388273143435,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0011096965103390246,
+      "loss": 1.1535,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6634759649178095,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0011093229252536647,
+      "loss": 1.1352,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6635636566921841,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0011089493894766579,
+      "loss": 1.1082,
+      "step": 7567
+    },
+    {
+      "epoch": 0.6636513484665586,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0011085759030420698,
+      "loss": 1.1556,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6637390402409331,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001108202465983962,
+      "loss": 1.1511,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6638267320153077,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0011078290783363904,
+      "loss": 1.1481,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6639144237896822,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0011074557401334073,
+      "loss": 1.152,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6640021155640567,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00110708245140906,
+      "loss": 1.1514,
+      "step": 7572
+    },
+    {
+      "epoch": 0.6640898073384314,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0011067092121973916,
+      "loss": 1.1632,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6641774991128059,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0011063360225324406,
+      "loss": 1.0995,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6642651908871804,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.001105962882448241,
+      "loss": 1.1576,
+      "step": 7575
+    },
+    {
+      "epoch": 0.664352882661555,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0011055897919788219,
+      "loss": 1.1013,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6644405744359295,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0011052167511582088,
+      "loss": 1.2086,
+      "step": 7577
+    },
+    {
+      "epoch": 0.664528266210304,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0011048437600204213,
+      "loss": 1.0804,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6646159579846787,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0011044708185994757,
+      "loss": 1.157,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6647036497590532,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001104097926929383,
+      "loss": 1.1547,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6647913415334277,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0011037250850441502,
+      "loss": 1.1102,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6648790333078023,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0011033522929777793,
+      "loss": 1.1337,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6649667250821768,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0011029795507642686,
+      "loss": 1.1331,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6650544168565513,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0011026068584376106,
+      "loss": 1.0685,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6651421086309258,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0011022342160317941,
+      "loss": 1.1808,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6652298004053004,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0011018616235808029,
+      "loss": 1.147,
+      "step": 7586
+    },
+    {
+      "epoch": 0.665317492179675,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0011014890811186161,
+      "loss": 1.1395,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6654051839540495,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011011165886792098,
+      "loss": 1.1371,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6654928757284241,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0011007441462965535,
+      "loss": 1.1849,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6655805675027986,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001100371754004613,
+      "loss": 1.0766,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6656682592771731,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0010999994118373495,
+      "loss": 1.1364,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6657559510515477,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0010996271198287202,
+      "loss": 1.1228,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6658436428259222,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0010992548780126767,
+      "loss": 1.0905,
+      "step": 7593
+    },
+    {
+      "epoch": 0.6659313346002967,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010988826864231664,
+      "loss": 1.1159,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6660190263746714,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0010985105450941328,
+      "loss": 1.1433,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6661067181490459,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0010981384540595133,
+      "loss": 1.1362,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6661944099234204,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0010977664133532425,
+      "loss": 1.1315,
+      "step": 7597
+    },
+    {
+      "epoch": 0.666282101697795,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0010973944230092492,
+      "loss": 1.1408,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6663697934721695,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001097022483061458,
+      "loss": 1.1242,
+      "step": 7599
+    },
+    {
+      "epoch": 0.666457485246544,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010966505935437892,
+      "loss": 1.1186,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6665451770209186,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0010962787544901574,
+      "loss": 1.141,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6666328687952932,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.001095906965934474,
+      "loss": 1.1332,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6667205605696677,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0010955352279106443,
+      "loss": 1.147,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6668082523440423,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0010951635404525708,
+      "loss": 1.1097,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6668959441184168,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0010947919035941504,
+      "loss": 1.1853,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6669836358927913,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0010944203173692748,
+      "loss": 1.1445,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6670713276671659,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001094048781811832,
+      "loss": 1.1052,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6671590194415404,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0010936772969557052,
+      "loss": 1.1009,
+      "step": 7608
+    },
+    {
+      "epoch": 0.667246711215915,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010933058628347727,
+      "loss": 1.1267,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6673344029902895,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.001092934479482909,
+      "loss": 1.1422,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6674220947646641,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.001092563146933982,
+      "loss": 1.1047,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6675097865390386,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0010921918652218574,
+      "loss": 1.1516,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6675974783134131,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010918206343803944,
+      "loss": 1.1182,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6676851700877877,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0010914494544434488,
+      "loss": 1.1549,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6677728618621622,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.001091078325444871,
+      "loss": 1.1092,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6678605536365367,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0010907072474185071,
+      "loss": 1.0999,
+      "step": 7616
+    },
+    {
+      "epoch": 0.6679482454109114,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001090336220398199,
+      "loss": 1.1294,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6680359371852859,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.001089965244417782,
+      "loss": 1.1129,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6681236289596604,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0010895943195110894,
+      "loss": 1.1492,
+      "step": 7619
+    },
+    {
+      "epoch": 0.668211320734035,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010892234457119479,
+      "loss": 1.1355,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6682990125084095,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0010888526230541805,
+      "loss": 1.096,
+      "step": 7621
+    },
+    {
+      "epoch": 0.668386704282784,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0010884818515716056,
+      "loss": 1.1133,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6684743960571586,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010881111312980361,
+      "loss": 1.1504,
+      "step": 7623
+    },
+    {
+      "epoch": 0.6685620878315331,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.001087740462267281,
+      "loss": 1.1186,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6686497796059077,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001087369844513144,
+      "loss": 1.2002,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6687374713802823,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0010869992780694246,
+      "loss": 1.1043,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6688251631546568,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0010866287629699187,
+      "loss": 1.1646,
+      "step": 7627
+    },
+    {
+      "epoch": 0.6689128549290313,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0010862582992484146,
+      "loss": 1.1227,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6690005467034059,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0010858878869386976,
+      "loss": 1.0949,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6690882384777804,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00108551752607455,
+      "loss": 1.1251,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6691759302521549,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0010851472166897467,
+      "loss": 1.1154,
+      "step": 7631
+    },
+    {
+      "epoch": 0.6692636220265294,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001084776958818059,
+      "loss": 1.1771,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6693513138009041,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0010844067524932535,
+      "loss": 1.1586,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6694390055752786,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0010840365977490924,
+      "loss": 1.1525,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6695266973496531,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0010836664946193323,
+      "loss": 1.1427,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6696143891240277,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0010832964431377263,
+      "loss": 1.1019,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6697020808984022,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001082926443338021,
+      "loss": 1.0758,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6697897726727767,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0010825564952539604,
+      "loss": 1.1533,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6698774644471513,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0010821865989192826,
+      "loss": 1.1247,
+      "step": 7639
+    },
+    {
+      "epoch": 0.6699651562215259,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0010818167543677217,
+      "loss": 1.1344,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6700528479959004,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0010814469616330054,
+      "loss": 1.1003,
+      "step": 7641
+    },
+    {
+      "epoch": 0.670140539770275,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0010810772207488586,
+      "loss": 1.1638,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6702282315446495,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010807075317490003,
+      "loss": 1.1342,
+      "step": 7643
+    },
+    {
+      "epoch": 0.670315923319024,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0010803378946671466,
+      "loss": 1.1401,
+      "step": 7644
+    },
+    {
+      "epoch": 0.6704036150933986,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0010799683095370051,
+      "loss": 1.1781,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6704913068677731,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.001079598776392283,
+      "loss": 1.1657,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6705789986421477,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0010792292952666795,
+      "loss": 1.1179,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6706666904165223,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001078859866193891,
+      "loss": 1.1252,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6707543821908968,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0010784904892076088,
+      "loss": 1.1337,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6708420739652713,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0010781211643415183,
+      "loss": 1.1129,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6709297657396459,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0010777518916293015,
+      "loss": 1.1351,
+      "step": 7651
+    },
+    {
+      "epoch": 0.6710174575140204,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0010773826711046345,
+      "loss": 1.1485,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6711051492883949,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0010770135028011897,
+      "loss": 1.1596,
+      "step": 7653
+    },
+    {
+      "epoch": 0.6711928410627696,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.001076644386752635,
+      "loss": 1.114,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6712805328371441,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0010762753229926316,
+      "loss": 1.0893,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6713682246115186,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0010759063115548374,
+      "loss": 1.1654,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6714559163858931,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0010755373524729056,
+      "loss": 1.1385,
+      "step": 7657
+    },
+    {
+      "epoch": 0.6715436081602677,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0010751684457804848,
+      "loss": 1.1198,
+      "step": 7658
+    },
+    {
+      "epoch": 0.6716312999346422,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0010747995915112174,
+      "loss": 1.1009,
+      "step": 7659
+    },
+    {
+      "epoch": 0.6717189917090167,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010744307896987425,
+      "loss": 1.1458,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6718066834833913,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0010740620403766937,
+      "loss": 1.1854,
+      "step": 7661
+    },
+    {
+      "epoch": 0.6718943752577659,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0010736933435787006,
+      "loss": 1.1357,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6719820670321404,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010733246993383863,
+      "loss": 1.1288,
+      "step": 7663
+    },
+    {
+      "epoch": 0.672069758806515,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0010729561076893708,
+      "loss": 1.1204,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6721574505808895,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0010725875686652681,
+      "loss": 1.1841,
+      "step": 7665
+    },
+    {
+      "epoch": 0.672245142355264,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00107221908229969,
+      "loss": 1.1166,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6723328341296386,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001071850648626239,
+      "loss": 1.1447,
+      "step": 7667
+    },
+    {
+      "epoch": 0.6724205259040131,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0010714822676785168,
+      "loss": 1.131,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6725082176783876,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001071113939490118,
+      "loss": 1.0941,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6725959094527623,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0010707456640946334,
+      "loss": 1.1172,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6726836012271368,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0010703774415256496,
+      "loss": 1.2103,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6727712930015113,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0010700092718167462,
+      "loss": 1.1058,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6728589847758859,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0010696411550014998,
+      "loss": 1.0723,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6729466765502604,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010692730911134821,
+      "loss": 1.1582,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6730343683246349,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0010689050801862595,
+      "loss": 1.0875,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6731220600990095,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0010685371222533939,
+      "loss": 1.1061,
+      "step": 7676
+    },
+    {
+      "epoch": 0.673209751873384,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0010681692173484412,
+      "loss": 1.1086,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6732974436477586,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0010678013655049537,
+      "loss": 1.1059,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6733851354221332,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0010674335667564788,
+      "loss": 1.1736,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6734728271965077,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0010670658211365592,
+      "loss": 1.1791,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6735605189708822,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0010666981286787316,
+      "loss": 1.1354,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6736482107452567,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0010663304894165287,
+      "loss": 1.1206,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6737359025196313,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0010659629033834786,
+      "loss": 1.1854,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6738235942940058,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0010655953706131044,
+      "loss": 1.0993,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6739112860683804,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010652278911389237,
+      "loss": 1.1392,
+      "step": 7685
+    },
+    {
+      "epoch": 0.673998977842755,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0010648604649944494,
+      "loss": 1.1346,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6740866696171295,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00106449309221319,
+      "loss": 1.0955,
+      "step": 7687
+    },
+    {
+      "epoch": 0.674174361391504,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001064125772828649,
+      "loss": 1.1305,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6742620531658786,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0010637585068743256,
+      "loss": 1.0967,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6743497449402531,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001063391294383713,
+      "loss": 1.1476,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6744374367146276,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0010630241353902995,
+      "loss": 1.1854,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6745251284890023,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00106265702992757,
+      "loss": 1.1513,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6746128202633768,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0010622899780290036,
+      "loss": 1.1292,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6747005120377513,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010619229797280735,
+      "loss": 1.1361,
+      "step": 7694
+    },
+    {
+      "epoch": 0.6747882038121259,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010615560350582492,
+      "loss": 1.1391,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6748758955865004,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0010611891440529954,
+      "loss": 1.1188,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6749635873608749,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0010608223067457726,
+      "loss": 1.1187,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6750512791352495,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0010604555231700338,
+      "loss": 1.1863,
+      "step": 7698
+    },
+    {
+      "epoch": 0.675138970909624,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.001060088793359229,
+      "loss": 1.1254,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6752266626839986,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0010597221173468039,
+      "loss": 1.1119,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6753143544583732,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010593554951661983,
+      "loss": 1.1324,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6754020462327477,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.001058988926850847,
+      "loss": 1.1005,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6754897380071222,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0010586224124341792,
+      "loss": 1.1295,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6755774297814967,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010582559519496206,
+      "loss": 1.1078,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6756651215558713,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0010578895454305915,
+      "loss": 1.1968,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6757528133302458,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001057523192910508,
+      "loss": 1.1119,
+      "step": 7706
+    },
+    {
+      "epoch": 0.6758405051046203,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0010571568944227796,
+      "loss": 1.0514,
+      "step": 7707
+    },
+    {
+      "epoch": 0.675928196878995,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0010567906500008116,
+      "loss": 1.1403,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6760158886533695,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0010564244596780053,
+      "loss": 1.1622,
+      "step": 7709
+    },
+    {
+      "epoch": 0.676103580427744,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001056058323487756,
+      "loss": 1.1068,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6761912722021186,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0010556922414634545,
+      "loss": 1.1646,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6762789639764931,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0010553262136384863,
+      "loss": 1.1194,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6763666557508676,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010549602400462318,
+      "loss": 1.1536,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6764543475252422,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010545943207200675,
+      "loss": 1.1052,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6765420392996168,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0010542284556933647,
+      "loss": 1.1866,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6766297310739913,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0010538626449994883,
+      "loss": 1.1243,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6767174228483659,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010534968886717995,
+      "loss": 1.1034,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6768051146227404,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010531311867436556,
+      "loss": 1.1111,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6768928063971149,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010527655392484053,
+      "loss": 1.1518,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6769804981714895,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0010523999462193974,
+      "loss": 1.1437,
+      "step": 7720
+    },
+    {
+      "epoch": 0.677068189945864,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010520344076899708,
+      "loss": 1.1442,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6771558817202386,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0010516689236934626,
+      "loss": 1.0603,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6772435734946132,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0010513034942632043,
+      "loss": 1.1319,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6773312652689877,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0010509381194325218,
+      "loss": 1.0905,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6774189570433622,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001050572799234736,
+      "loss": 1.1628,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6775066488177368,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0010502075337031635,
+      "loss": 1.1714,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6775943405921113,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.001049842322871116,
+      "loss": 1.1309,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6776820323664858,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010494771667718989,
+      "loss": 1.223,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6777697241408603,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0010491120654388146,
+      "loss": 1.1736,
+      "step": 7729
+    },
+    {
+      "epoch": 0.677857415915235,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001048747018905158,
+      "loss": 1.1612,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6779451076896095,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0010483820272042216,
+      "loss": 1.1541,
+      "step": 7731
+    },
+    {
+      "epoch": 0.678032799463984,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010480170903692909,
+      "loss": 1.1749,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6781204912383586,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0010476522084336482,
+      "loss": 1.1399,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6782081830127331,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0010472873814305684,
+      "loss": 1.1426,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6782958747871076,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0010469226093933245,
+      "loss": 1.1757,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6783835665614822,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001046557892355181,
+      "loss": 1.0712,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6784712583358568,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0010461932303493999,
+      "loss": 1.1702,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6785589501102313,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010458286234092378,
+      "loss": 1.1565,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6786466418846059,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0010454640715679452,
+      "loss": 1.0842,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6787343336589804,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001045099574858769,
+      "loss": 1.1267,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6788220254333549,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0010447351333149498,
+      "loss": 1.1461,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6789097172077295,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0010443707469697245,
+      "loss": 1.1502,
+      "step": 7742
+    },
+    {
+      "epoch": 0.678997408982104,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001044006415856323,
+      "loss": 1.1999,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6790851007564785,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.001043642140007972,
+      "loss": 1.1392,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6791727925308532,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0010432779194578926,
+      "loss": 1.1665,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6792604843052277,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0010429137542393014,
+      "loss": 1.1236,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6793481760796022,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.001042549644385408,
+      "loss": 1.1717,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6794358678539768,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010421855899294188,
+      "loss": 1.1073,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6795235596283513,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0010418215909045348,
+      "loss": 1.1137,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6796112514027258,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010414576473439523,
+      "loss": 1.1321,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6796989431771003,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0010410937592808607,
+      "loss": 1.0864,
+      "step": 7751
+    },
+    {
+      "epoch": 0.679786634951475,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001040729926748447,
+      "loss": 1.1572,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6798743267258495,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010403661497798908,
+      "loss": 1.1088,
+      "step": 7753
+    },
+    {
+      "epoch": 0.679962018500224,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0010400024284083677,
+      "loss": 1.1618,
+      "step": 7754
+    },
+    {
+      "epoch": 0.6800497102745986,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0010396387626670494,
+      "loss": 1.1616,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6801374020489731,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0010392751525890992,
+      "loss": 1.1111,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6802250938233476,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0010389115982076787,
+      "loss": 1.1503,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6803127855977222,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0010385480995559432,
+      "loss": 1.1506,
+      "step": 7758
+    },
+    {
+      "epoch": 0.6804004773720967,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0010381846566670425,
+      "loss": 1.0544,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6804881691464713,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.001037821269574122,
+      "loss": 1.1589,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6805758609208459,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0010374579383103208,
+      "loss": 1.1284,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6806635526952204,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010370946629087747,
+      "loss": 1.0949,
+      "step": 7762
+    },
+    {
+      "epoch": 0.6807512444695949,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010367314434026127,
+      "loss": 1.0876,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6808389362439695,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001036368279824961,
+      "loss": 1.1807,
+      "step": 7764
+    },
+    {
+      "epoch": 0.680926628018344,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001036005172208937,
+      "loss": 1.1351,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6810143197927185,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0010356421205876566,
+      "loss": 1.182,
+      "step": 7766
+    },
+    {
+      "epoch": 0.6811020115670932,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0010352791249942288,
+      "loss": 1.1423,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6811897033414677,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010349161854617582,
+      "loss": 1.2074,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6812773951158422,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0010345533020233442,
+      "loss": 1.1345,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6813650868902168,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0010341904747120793,
+      "loss": 1.1493,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6814527786645913,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0010338277035610536,
+      "loss": 1.1132,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6815404704389658,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0010334649886033507,
+      "loss": 1.0994,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6816281622133404,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0010331023298720497,
+      "loss": 1.2056,
+      "step": 7773
+    },
+    {
+      "epoch": 0.681715853987715,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0010327397274002234,
+      "loss": 1.1879,
+      "step": 7774
+    },
+    {
+      "epoch": 0.6818035457620895,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00103237718122094,
+      "loss": 1.1233,
+      "step": 7775
+    },
+    {
+      "epoch": 0.681891237536464,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001032014691367264,
+      "loss": 1.1102,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6819789293108386,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0010316522578722527,
+      "loss": 1.1536,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6820666210852131,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.001031289880768959,
+      "loss": 1.1407,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6821543128595876,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001030927560090431,
+      "loss": 1.1882,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6822420046339622,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.001030565295869711,
+      "loss": 1.1005,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6823296964083367,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001030203088139838,
+      "loss": 1.1542,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6824173881827112,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0010298409369338425,
+      "loss": 1.2198,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6825050799570859,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.001029478842284753,
+      "loss": 1.1803,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6825927717314604,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0010291168042255907,
+      "loss": 1.1074,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6826804635058349,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0010287548227893735,
+      "loss": 1.1227,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6827681552802095,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0010283928980091132,
+      "loss": 1.1154,
+      "step": 7786
+    },
+    {
+      "epoch": 0.682855847054584,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0010280310299178158,
+      "loss": 1.0879,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6829435388289585,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0010276692185484822,
+      "loss": 1.1283,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6830312306033332,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0010273074639341095,
+      "loss": 1.1631,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6831189223777077,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0010269457661076888,
+      "loss": 1.1593,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6832066141520822,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0010265841251022058,
+      "loss": 1.1204,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6832943059264568,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0010262225409506413,
+      "loss": 1.1523,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6833819977008313,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0010258610136859704,
+      "loss": 1.1178,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6834696894752058,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0010254995433411642,
+      "loss": 1.0623,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6835573812495804,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001025138129949188,
+      "loss": 1.1637,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6836450730239549,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0010247767735430003,
+      "loss": 1.1403,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6837327647983295,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0010244154741555574,
+      "loss": 1.1479,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6838204565727041,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0010240542318198079,
+      "loss": 1.1136,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6839081483470786,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0010236930465686972,
+      "loss": 1.1899,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6839958401214531,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0010233319184351636,
+      "loss": 1.1164,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6840835318958276,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.001022970847452141,
+      "loss": 1.1056,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6841712236702022,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0010226098336525587,
+      "loss": 1.1916,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6842589154445767,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0010222488770693402,
+      "loss": 1.1306,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6843466072189512,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.001021887977735404,
+      "loss": 1.0922,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6844342989933259,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0010215271356836617,
+      "loss": 1.1178,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6845219907677004,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0010211663509470227,
+      "loss": 1.1672,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6846096825420749,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0010208056235583888,
+      "loss": 1.193,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6846973743164495,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0010204449535506585,
+      "loss": 1.1681,
+      "step": 7808
+    },
+    {
+      "epoch": 0.684785066090824,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010200843409567228,
+      "loss": 1.1995,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6848727578651985,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.001019723785809469,
+      "loss": 1.142,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6849604496395731,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.001019363288141779,
+      "loss": 1.1184,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6850481414139477,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010190028479865286,
+      "loss": 1.1465,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6851358331883222,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0010186424653765905,
+      "loss": 1.139,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6852235249626968,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0010182821403448294,
+      "loss": 1.1413,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6853112167370713,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001017921872924106,
+      "loss": 1.129,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6853989085114458,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0010175616631472762,
+      "loss": 1.1217,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6854866002858204,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0010172015110471906,
+      "loss": 1.1116,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6855742920601949,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0010168414166566932,
+      "loss": 1.0954,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6856619838345694,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0010164813800086238,
+      "loss": 1.1899,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6857496756089441,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010161214011358182,
+      "loss": 1.1614,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6858373673833186,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0010157614800711035,
+      "loss": 1.1321,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6859250591576931,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0010154016168473054,
+      "loss": 1.1145,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6860127509320676,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0010150418114972409,
+      "loss": 1.1202,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6861004427064422,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0010146820640537237,
+      "loss": 1.1988,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6861881344808167,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010143223745495626,
+      "loss": 1.1131,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6862758262551912,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0010139627430175604,
+      "loss": 1.1964,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6863635180295659,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010136031694905137,
+      "loss": 1.1726,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6864512098039404,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.001013243654001215,
+      "loss": 1.1455,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6865389015783149,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010128841965824512,
+      "loss": 1.1605,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6866265933526895,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0010125247972670045,
+      "loss": 1.1724,
+      "step": 7830
+    },
+    {
+      "epoch": 0.686714285127064,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010121654560876503,
+      "loss": 1.1608,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6868019769014385,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0010118061730771599,
+      "loss": 1.1072,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6868896686758131,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.001011446948268299,
+      "loss": 1.1588,
+      "step": 7833
+    },
+    {
+      "epoch": 0.6869773604501876,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0010110877816938285,
+      "loss": 1.1428,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6870650522245622,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0010107286733865024,
+      "loss": 1.1657,
+      "step": 7835
+    },
+    {
+      "epoch": 0.6871527439989368,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.001010369623379071,
+      "loss": 1.1331,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6872404357733113,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0010100106317042794,
+      "loss": 1.1471,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6873281275476858,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0010096516983948652,
+      "loss": 1.1486,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6874158193220604,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0010092928234835634,
+      "loss": 1.1833,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6875035110964349,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0010089340070031027,
+      "loss": 1.1089,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6875912028708094,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001008575248986205,
+      "loss": 1.1568,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6876788946451841,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0010082165494655886,
+      "loss": 1.0997,
+      "step": 7842
+    },
+    {
+      "epoch": 0.6877665864195586,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0010078579084739664,
+      "loss": 1.1167,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6878542781939331,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0010074993260440455,
+      "loss": 1.1459,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6879419699683077,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0010071408022085272,
+      "loss": 1.2438,
+      "step": 7845
+    },
+    {
+      "epoch": 0.6880296617426822,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0010067823370001083,
+      "loss": 1.1847,
+      "step": 7846
+    },
+    {
+      "epoch": 0.6881173535170567,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0010064239304514795,
+      "loss": 1.0766,
+      "step": 7847
+    },
+    {
+      "epoch": 0.6882050452914312,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0010060655825953274,
+      "loss": 1.1505,
+      "step": 7848
+    },
+    {
+      "epoch": 0.6882927370658058,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0010057072934643316,
+      "loss": 1.1462,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6883804288401804,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0010053490630911673,
+      "loss": 1.183,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6884681206145549,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0010049908915085043,
+      "loss": 1.0956,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6885558123889295,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0010046327787490074,
+      "loss": 1.1504,
+      "step": 7852
+    },
+    {
+      "epoch": 0.688643504163304,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0010042747248453345,
+      "loss": 1.1546,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6887311959376785,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0010039167298301402,
+      "loss": 1.0873,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6888188877120531,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001003558793736072,
+      "loss": 1.1588,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6889065794864276,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0010032009165957727,
+      "loss": 1.1666,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6889942712608021,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001002843098441881,
+      "loss": 1.153,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6890819630351768,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0010024853393070274,
+      "loss": 1.1272,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6891696548095513,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010021276392238392,
+      "loss": 1.1569,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6892573465839258,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.001001769998224938,
+      "loss": 1.149,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6893450383583004,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00100141241634294,
+      "loss": 1.1102,
+      "step": 7861
+    },
+    {
+      "epoch": 0.6894327301326749,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0010010548936104549,
+      "loss": 1.1673,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6895204219070494,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.001000697430060088,
+      "loss": 1.113,
+      "step": 7863
+    },
+    {
+      "epoch": 0.689608113681424,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0010003400257244395,
+      "loss": 1.1013,
+      "step": 7864
+    },
+    {
+      "epoch": 0.6896958054557986,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0009999826806361036,
+      "loss": 1.1037,
+      "step": 7865
+    },
+    {
+      "epoch": 0.6897834972301731,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0009996253948276698,
+      "loss": 1.1803,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6898711890045477,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009992681683317203,
+      "loss": 1.1183,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6899588807789222,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0009989110011808342,
+      "loss": 1.1339,
+      "step": 7868
+    },
+    {
+      "epoch": 0.6900465725532967,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0009985538934075837,
+      "loss": 1.0644,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6901342643276713,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0009981968450445372,
+      "loss": 1.1196,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6902219561020458,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000997839856124256,
+      "loss": 1.1922,
+      "step": 7871
+    },
+    {
+      "epoch": 0.6903096478764204,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009974829266792954,
+      "loss": 1.1288,
+      "step": 7872
+    },
+    {
+      "epoch": 0.6903973396507949,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0009971260567422078,
+      "loss": 1.1156,
+      "step": 7873
+    },
+    {
+      "epoch": 0.6904850314251695,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0009967692463455384,
+      "loss": 1.1175,
+      "step": 7874
+    },
+    {
+      "epoch": 0.690572723199544,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0009964124955218277,
+      "loss": 1.1035,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6906604149739185,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009960558043036099,
+      "loss": 1.0896,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6907481067482931,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0009956991727234144,
+      "loss": 1.0858,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6908357985226676,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0009953426008137655,
+      "loss": 1.1113,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6909234902970421,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009949860886071817,
+      "loss": 1.1667,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6910111820714168,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0009946296361361751,
+      "loss": 1.1316,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6910988738457913,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009942732434332544,
+      "loss": 1.1224,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6911865656201658,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.000993916910530921,
+      "loss": 1.16,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6912742573945404,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009935606374616717,
+      "loss": 1.1298,
+      "step": 7883
+    },
+    {
+      "epoch": 0.6913619491689149,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009932044242579976,
+      "loss": 1.2187,
+      "step": 7884
+    },
+    {
+      "epoch": 0.6914496409432894,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009928482709523844,
+      "loss": 1.1766,
+      "step": 7885
+    },
+    {
+      "epoch": 0.691537332717664,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0009924921775773124,
+      "loss": 1.1219,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6916250244920386,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.000992136144165257,
+      "loss": 1.1552,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6917127162664131,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0009917801707486872,
+      "loss": 1.1048,
+      "step": 7888
+    },
+    {
+      "epoch": 0.6918004080407877,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0009914242573600656,
+      "loss": 1.1212,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6918880998151622,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0009910684040318524,
+      "loss": 1.1564,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6919757915895367,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009907126107964988,
+      "loss": 1.1832,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6920634833639113,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0009903568776864539,
+      "loss": 1.2012,
+      "step": 7892
+    },
+    {
+      "epoch": 0.6921511751382858,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.000990001204734159,
+      "loss": 1.1097,
+      "step": 7893
+    },
+    {
+      "epoch": 0.6922388669126603,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00098964559197205,
+      "loss": 1.1303,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6923265586870349,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0009892900394325584,
+      "loss": 1.1133,
+      "step": 7895
+    },
+    {
+      "epoch": 0.6924142504614095,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0009889345471481095,
+      "loss": 1.1557,
+      "step": 7896
+    },
+    {
+      "epoch": 0.692501942235784,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0009885791151511242,
+      "loss": 1.1435,
+      "step": 7897
+    },
+    {
+      "epoch": 0.6925896340101585,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009882237434740153,
+      "loss": 1.1548,
+      "step": 7898
+    },
+    {
+      "epoch": 0.6926773257845331,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0009878684321491928,
+      "loss": 1.2064,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6927650175589076,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0009875131812090597,
+      "loss": 1.0812,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6928527093332821,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0009871579906860152,
+      "loss": 1.1228,
+      "step": 7901
+    },
+    {
+      "epoch": 0.6929404011076568,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009868028606124502,
+      "loss": 1.1644,
+      "step": 7902
+    },
+    {
+      "epoch": 0.6930280928820313,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000986447791020752,
+      "loss": 1.1241,
+      "step": 7903
+    },
+    {
+      "epoch": 0.6931157846564058,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0009860927819433027,
+      "loss": 1.1308,
+      "step": 7904
+    },
+    {
+      "epoch": 0.6932034764307804,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0009857378334124781,
+      "loss": 1.1352,
+      "step": 7905
+    },
+    {
+      "epoch": 0.6932911682051549,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009853829454606483,
+      "loss": 1.1337,
+      "step": 7906
+    },
+    {
+      "epoch": 0.6933788599795294,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009850281181201777,
+      "loss": 1.1306,
+      "step": 7907
+    },
+    {
+      "epoch": 0.693466551753904,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0009846733514234263,
+      "loss": 1.1354,
+      "step": 7908
+    },
+    {
+      "epoch": 0.6935542435282785,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0009843186454027476,
+      "loss": 1.1131,
+      "step": 7909
+    },
+    {
+      "epoch": 0.6936419353026531,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009839640000904905,
+      "loss": 1.1682,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6937296270770277,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009836094155189965,
+      "loss": 1.1239,
+      "step": 7911
+    },
+    {
+      "epoch": 0.6938173188514022,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0009832548917206039,
+      "loss": 1.1814,
+      "step": 7912
+    },
+    {
+      "epoch": 0.6939050106257767,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0009829004287276438,
+      "loss": 1.0868,
+      "step": 7913
+    },
+    {
+      "epoch": 0.6939927024001513,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000982546026572443,
+      "loss": 1.1181,
+      "step": 7914
+    },
+    {
+      "epoch": 0.6940803941745258,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000982191685287321,
+      "loss": 1.1627,
+      "step": 7915
+    },
+    {
+      "epoch": 0.6941680859489003,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0009818374049045933,
+      "loss": 1.0883,
+      "step": 7916
+    },
+    {
+      "epoch": 0.694255777723275,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009814831854565696,
+      "loss": 1.2146,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6943434694976495,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0009811290269755534,
+      "loss": 1.0834,
+      "step": 7918
+    },
+    {
+      "epoch": 0.694431161272024,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0009807749294938439,
+      "loss": 1.0952,
+      "step": 7919
+    },
+    {
+      "epoch": 0.6945188530463985,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0009804208930437328,
+      "loss": 1.1411,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6946065448207731,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009800669176575076,
+      "loss": 1.1176,
+      "step": 7921
+    },
+    {
+      "epoch": 0.6946942365951476,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0009797130033674505,
+      "loss": 1.1095,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6947819283695221,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009793591502058369,
+      "loss": 1.173,
+      "step": 7923
+    },
+    {
+      "epoch": 0.6948696201438967,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0009790053582049382,
+      "loss": 1.1187,
+      "step": 7924
+    },
+    {
+      "epoch": 0.6949573119182713,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0009786516273970177,
+      "loss": 1.1486,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6950450036926458,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000978297957814336,
+      "loss": 1.1093,
+      "step": 7926
+    },
+    {
+      "epoch": 0.6951326954670204,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0009779443494891463,
+      "loss": 1.191,
+      "step": 7927
+    },
+    {
+      "epoch": 0.6952203872413949,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000977590802453698,
+      "loss": 1.152,
+      "step": 7928
+    },
+    {
+      "epoch": 0.6953080790157694,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0009772373167402319,
+      "loss": 1.1854,
+      "step": 7929
+    },
+    {
+      "epoch": 0.695395770790144,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0009768838923809857,
+      "loss": 1.152,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6954834625645185,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0009765305294081909,
+      "loss": 1.101,
+      "step": 7931
+    },
+    {
+      "epoch": 0.695571154338893,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.000976177227854074,
+      "loss": 1.1091,
+      "step": 7932
+    },
+    {
+      "epoch": 0.6956588461132677,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0009758239877508537,
+      "loss": 1.0926,
+      "step": 7933
+    },
+    {
+      "epoch": 0.6957465378876422,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009754708091307454,
+      "loss": 1.1195,
+      "step": 7934
+    },
+    {
+      "epoch": 0.6958342296620167,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009751176920259581,
+      "loss": 1.1517,
+      "step": 7935
+    },
+    {
+      "epoch": 0.6959219214363913,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0009747646364686956,
+      "loss": 1.0778,
+      "step": 7936
+    },
+    {
+      "epoch": 0.6960096132107658,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0009744116424911546,
+      "loss": 1.1607,
+      "step": 7937
+    },
+    {
+      "epoch": 0.6960973049851403,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0009740587101255278,
+      "loss": 1.1132,
+      "step": 7938
+    },
+    {
+      "epoch": 0.696184996759515,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.000973705839404002,
+      "loss": 1.1237,
+      "step": 7939
+    },
+    {
+      "epoch": 0.6962726885338895,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0009733530303587575,
+      "loss": 1.1683,
+      "step": 7940
+    },
+    {
+      "epoch": 0.696360380308264,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0009730002830219703,
+      "loss": 1.1732,
+      "step": 7941
+    },
+    {
+      "epoch": 0.6964480720826385,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0009726475974258092,
+      "loss": 1.1425,
+      "step": 7942
+    },
+    {
+      "epoch": 0.6965357638570131,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0009722949736024386,
+      "loss": 1.1575,
+      "step": 7943
+    },
+    {
+      "epoch": 0.6966234556313876,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0009719424115840169,
+      "loss": 1.1174,
+      "step": 7944
+    },
+    {
+      "epoch": 0.6967111474057621,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000971589911402697,
+      "loss": 1.0691,
+      "step": 7945
+    },
+    {
+      "epoch": 0.6967988391801367,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0009712374730906264,
+      "loss": 1.1336,
+      "step": 7946
+    },
+    {
+      "epoch": 0.6968865309545113,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0009708850966799453,
+      "loss": 1.1688,
+      "step": 7947
+    },
+    {
+      "epoch": 0.6969742227288858,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009705327822027904,
+      "loss": 1.0877,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6970619145032604,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0009701805296912918,
+      "loss": 1.1268,
+      "step": 7949
+    },
+    {
+      "epoch": 0.6971496062776349,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0009698283391775742,
+      "loss": 1.1476,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6972372980520094,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009694762106937559,
+      "loss": 1.099,
+      "step": 7951
+    },
+    {
+      "epoch": 0.697324989826384,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009691241442719503,
+      "loss": 1.1352,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6974126816007585,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0009687721399442652,
+      "loss": 1.1027,
+      "step": 7953
+    },
+    {
+      "epoch": 0.697500373375133,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009684201977428027,
+      "loss": 1.136,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6975880651495077,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0009680683176996585,
+      "loss": 1.1475,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6976757569238822,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009677164998469235,
+      "loss": 1.1675,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6977634486982567,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000967364744216682,
+      "loss": 1.1738,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6978511404726313,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0009670130508410137,
+      "loss": 1.134,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6979388322470058,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0009666614197519926,
+      "loss": 1.0882,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6980265240213803,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0009663098509816852,
+      "loss": 1.1686,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6981142157957549,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0009659583445621545,
+      "loss": 1.1708,
+      "step": 7961
+    },
+    {
+      "epoch": 0.6982019075701295,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0009656069005254571,
+      "loss": 1.1807,
+      "step": 7962
+    },
+    {
+      "epoch": 0.698289599344504,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.000965255518903644,
+      "loss": 1.0855,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6983772911188786,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0009649041997287594,
+      "loss": 1.1701,
+      "step": 7964
+    },
+    {
+      "epoch": 0.6984649828932531,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0009645529430328435,
+      "loss": 1.11,
+      "step": 7965
+    },
+    {
+      "epoch": 0.6985526746676276,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0009642017488479296,
+      "loss": 1.1784,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6986403664420021,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0009638506172060461,
+      "loss": 1.1526,
+      "step": 7967
+    },
+    {
+      "epoch": 0.6987280582163767,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0009634995481392147,
+      "loss": 1.138,
+      "step": 7968
+    },
+    {
+      "epoch": 0.6988157499907512,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009631485416794522,
+      "loss": 1.1574,
+      "step": 7969
+    },
+    {
+      "epoch": 0.6989034417651258,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009627975978587698,
+      "loss": 1.1471,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6989911335395004,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0009624467167091724,
+      "loss": 1.1316,
+      "step": 7971
+    },
+    {
+      "epoch": 0.6990788253138749,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0009620958982626598,
+      "loss": 1.1147,
+      "step": 7972
+    },
+    {
+      "epoch": 0.6991665170882494,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0009617451425512258,
+      "loss": 1.1545,
+      "step": 7973
+    },
+    {
+      "epoch": 0.699254208862624,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009613944496068575,
+      "loss": 1.1029,
+      "step": 7974
+    },
+    {
+      "epoch": 0.6993419006369985,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0009610438194615375,
+      "loss": 1.1226,
+      "step": 7975
+    },
+    {
+      "epoch": 0.699429592411373,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009606932521472428,
+      "loss": 1.1437,
+      "step": 7976
+    },
+    {
+      "epoch": 0.6995172841857477,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009603427476959449,
+      "loss": 1.0961,
+      "step": 7977
+    },
+    {
+      "epoch": 0.6996049759601222,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0009599923061396073,
+      "loss": 1.1154,
+      "step": 7978
+    },
+    {
+      "epoch": 0.6996926677344967,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00095964192751019,
+      "loss": 1.1959,
+      "step": 7979
+    },
+    {
+      "epoch": 0.6997803595088713,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009592916118396468,
+      "loss": 1.1459,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6998680512832458,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0009589413591599262,
+      "loss": 1.1298,
+      "step": 7981
+    },
+    {
+      "epoch": 0.6999557430576203,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0009585911695029691,
+      "loss": 1.1605,
+      "step": 7982
+    },
+    {
+      "epoch": 0.7000434348319949,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0009582410429007125,
+      "loss": 1.0921,
+      "step": 7983
+    },
+    {
+      "epoch": 0.7001311266063694,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0009578909793850871,
+      "loss": 1.1072,
+      "step": 7984
+    },
+    {
+      "epoch": 0.700218818380744,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009575409789880179,
+      "loss": 1.1983,
+      "step": 7985
+    },
+    {
+      "epoch": 0.7003065101551186,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0009571910417414236,
+      "loss": 1.1749,
+      "step": 7986
+    },
+    {
+      "epoch": 0.7003942019294931,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0009568411676772178,
+      "loss": 1.1587,
+      "step": 7987
+    },
+    {
+      "epoch": 0.7004818937038676,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000956491356827308,
+      "loss": 1.1172,
+      "step": 7988
+    },
+    {
+      "epoch": 0.7005695854782422,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009561416092235969,
+      "loss": 1.1097,
+      "step": 7989
+    },
+    {
+      "epoch": 0.7006572772526167,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009557919248979796,
+      "loss": 1.0974,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7007449690269912,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0009554423038823464,
+      "loss": 1.1613,
+      "step": 7991
+    },
+    {
+      "epoch": 0.7008326608013657,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009550927462085818,
+      "loss": 1.1511,
+      "step": 7992
+    },
+    {
+      "epoch": 0.7009203525757404,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0009547432519085652,
+      "loss": 1.1124,
+      "step": 7993
+    },
+    {
+      "epoch": 0.7010080443501149,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009543938210141695,
+      "loss": 1.0719,
+      "step": 7994
+    },
+    {
+      "epoch": 0.7010957361244894,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009540444535572611,
+      "loss": 1.1304,
+      "step": 7995
+    },
+    {
+      "epoch": 0.701183427898864,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0009536951495697022,
+      "loss": 1.2113,
+      "step": 7996
+    },
+    {
+      "epoch": 0.7012711196732385,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009533459090833479,
+      "loss": 1.1697,
+      "step": 7997
+    },
+    {
+      "epoch": 0.701358811447613,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0009529967321300489,
+      "loss": 1.0975,
+      "step": 7998
+    },
+    {
+      "epoch": 0.7014465032219876,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0009526476187416481,
+      "loss": 1.1787,
+      "step": 7999
+    },
+    {
+      "epoch": 0.7015341949963622,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0009522985689499842,
+      "loss": 1.1633,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7015341949963622,
+      "eval_loss": 1.148275375366211,
+      "eval_runtime": 429.1936,
+      "eval_samples_per_second": 33.661,
+      "eval_steps_per_second": 8.416,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7016218867707367,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0009519495827868897,
+      "loss": 1.113,
+      "step": 8001
+    },
+    {
+      "epoch": 0.7017095785451113,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0009516006602841915,
+      "loss": 1.0907,
+      "step": 8002
+    },
+    {
+      "epoch": 0.7017972703194858,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0009512518014737104,
+      "loss": 1.1518,
+      "step": 8003
+    },
+    {
+      "epoch": 0.7018849620938603,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0009509030063872606,
+      "loss": 1.1587,
+      "step": 8004
+    },
+    {
+      "epoch": 0.7019726538682349,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0009505542750566523,
+      "loss": 1.1565,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7020603456426094,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0009502056075136886,
+      "loss": 1.1193,
+      "step": 8006
+    },
+    {
+      "epoch": 0.702148037416984,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0009498570037901668,
+      "loss": 1.1086,
+      "step": 8007
+    },
+    {
+      "epoch": 0.7022357291913586,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009495084639178792,
+      "loss": 1.1417,
+      "step": 8008
+    },
+    {
+      "epoch": 0.7023234209657331,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0009491599879286109,
+      "loss": 1.1422,
+      "step": 8009
+    },
+    {
+      "epoch": 0.7024111127401076,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009488115758541424,
+      "loss": 1.1405,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7024988045144822,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0009484632277262482,
+      "loss": 1.1527,
+      "step": 8011
+    },
+    {
+      "epoch": 0.7025864962888567,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.000948114943576697,
+      "loss": 1.1399,
+      "step": 8012
+    },
+    {
+      "epoch": 0.7026741880632312,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0009477667234372506,
+      "loss": 1.1465,
+      "step": 8013
+    },
+    {
+      "epoch": 0.7027618798376057,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0009474185673396664,
+      "loss": 1.1091,
+      "step": 8014
+    },
+    {
+      "epoch": 0.7028495716119804,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000947070475315695,
+      "loss": 1.0887,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7029372633863549,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0009467224473970822,
+      "loss": 1.1258,
+      "step": 8016
+    },
+    {
+      "epoch": 0.7030249551607294,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000946374483615566,
+      "loss": 1.1599,
+      "step": 8017
+    },
+    {
+      "epoch": 0.703112646935104,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0009460265840028807,
+      "loss": 1.1551,
+      "step": 8018
+    },
+    {
+      "epoch": 0.7032003387094785,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0009456787485907534,
+      "loss": 1.1649,
+      "step": 8019
+    },
+    {
+      "epoch": 0.703288030483853,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0009453309774109069,
+      "loss": 1.1577,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7033757222582276,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009449832704950555,
+      "loss": 1.2459,
+      "step": 8021
+    },
+    {
+      "epoch": 0.7034634140326022,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009446356278749097,
+      "loss": 1.1152,
+      "step": 8022
+    },
+    {
+      "epoch": 0.7035511058069767,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009442880495821742,
+      "loss": 1.146,
+      "step": 8023
+    },
+    {
+      "epoch": 0.7036387975813513,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0009439405356485468,
+      "loss": 1.1322,
+      "step": 8024
+    },
+    {
+      "epoch": 0.7037264893557258,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0009435930861057199,
+      "loss": 1.1139,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7038141811301003,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009432457009853794,
+      "loss": 1.2094,
+      "step": 8026
+    },
+    {
+      "epoch": 0.7039018729044749,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000942898380319207,
+      "loss": 1.1713,
+      "step": 8027
+    },
+    {
+      "epoch": 0.7039895646788494,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0009425511241388765,
+      "loss": 1.1217,
+      "step": 8028
+    },
+    {
+      "epoch": 0.7040772564532239,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0009422039324760573,
+      "loss": 1.1332,
+      "step": 8029
+    },
+    {
+      "epoch": 0.7041649482275986,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009418568053624129,
+      "loss": 1.1754,
+      "step": 8030
+    },
+    {
+      "epoch": 0.7042526400019731,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000941509742829599,
+      "loss": 1.1638,
+      "step": 8031
+    },
+    {
+      "epoch": 0.7043403317763476,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009411627449092679,
+      "loss": 1.2097,
+      "step": 8032
+    },
+    {
+      "epoch": 0.7044280235507222,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0009408158116330645,
+      "loss": 1.2393,
+      "step": 8033
+    },
+    {
+      "epoch": 0.7045157153250967,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0009404689430326288,
+      "loss": 1.1448,
+      "step": 8034
+    },
+    {
+      "epoch": 0.7046034070994712,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0009401221391395935,
+      "loss": 1.1948,
+      "step": 8035
+    },
+    {
+      "epoch": 0.7046910988738458,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0009397753999855864,
+      "loss": 1.1226,
+      "step": 8036
+    },
+    {
+      "epoch": 0.7047787906482204,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0009394287256022294,
+      "loss": 1.1006,
+      "step": 8037
+    },
+    {
+      "epoch": 0.7048664824225949,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0009390821160211388,
+      "loss": 1.176,
+      "step": 8038
+    },
+    {
+      "epoch": 0.7049541741969694,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009387355712739236,
+      "loss": 1.0653,
+      "step": 8039
+    },
+    {
+      "epoch": 0.705041865971344,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0009383890913921885,
+      "loss": 1.1019,
+      "step": 8040
+    },
+    {
+      "epoch": 0.7051295577457185,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0009380426764075312,
+      "loss": 1.1791,
+      "step": 8041
+    },
+    {
+      "epoch": 0.705217249520093,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009376963263515432,
+      "loss": 1.1534,
+      "step": 8042
+    },
+    {
+      "epoch": 0.7053049412944676,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0009373500412558124,
+      "loss": 1.0992,
+      "step": 8043
+    },
+    {
+      "epoch": 0.7053926330688421,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009370038211519177,
+      "loss": 1.1575,
+      "step": 8044
+    },
+    {
+      "epoch": 0.7054803248432167,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0009366576660714337,
+      "loss": 1.1553,
+      "step": 8045
+    },
+    {
+      "epoch": 0.7055680166175913,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0009363115760459292,
+      "loss": 1.2077,
+      "step": 8046
+    },
+    {
+      "epoch": 0.7056557083919658,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.000935965551106967,
+      "loss": 1.1095,
+      "step": 8047
+    },
+    {
+      "epoch": 0.7057434001663403,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0009356195912861026,
+      "loss": 1.161,
+      "step": 8048
+    },
+    {
+      "epoch": 0.7058310919407149,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0009352736966148875,
+      "loss": 1.1469,
+      "step": 8049
+    },
+    {
+      "epoch": 0.7059187837150894,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009349278671248664,
+      "loss": 1.1284,
+      "step": 8050
+    },
+    {
+      "epoch": 0.7060064754894639,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009345821028475783,
+      "loss": 1.1485,
+      "step": 8051
+    },
+    {
+      "epoch": 0.7060941672638386,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0009342364038145549,
+      "loss": 1.111,
+      "step": 8052
+    },
+    {
+      "epoch": 0.7061818590382131,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0009338907700573236,
+      "loss": 1.1661,
+      "step": 8053
+    },
+    {
+      "epoch": 0.7062695508125876,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0009335452016074058,
+      "loss": 1.1296,
+      "step": 8054
+    },
+    {
+      "epoch": 0.7063572425869622,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0009331996984963158,
+      "loss": 1.1191,
+      "step": 8055
+    },
+    {
+      "epoch": 0.7064449343613367,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0009328542607555637,
+      "loss": 1.169,
+      "step": 8056
+    },
+    {
+      "epoch": 0.7065326261357112,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009325088884166514,
+      "loss": 1.1024,
+      "step": 8057
+    },
+    {
+      "epoch": 0.7066203179100858,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.000932163581511076,
+      "loss": 1.1052,
+      "step": 8058
+    },
+    {
+      "epoch": 0.7067080096844603,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009318183400703287,
+      "loss": 1.1811,
+      "step": 8059
+    },
+    {
+      "epoch": 0.7067957014588349,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0009314731641258951,
+      "loss": 1.1666,
+      "step": 8060
+    },
+    {
+      "epoch": 0.7068833932332095,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009311280537092543,
+      "loss": 1.1674,
+      "step": 8061
+    },
+    {
+      "epoch": 0.706971085007584,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000930783008851879,
+      "loss": 1.1915,
+      "step": 8062
+    },
+    {
+      "epoch": 0.7070587767819585,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0009304380295852364,
+      "loss": 1.1089,
+      "step": 8063
+    },
+    {
+      "epoch": 0.707146468556333,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.000930093115940788,
+      "loss": 1.1882,
+      "step": 8064
+    },
+    {
+      "epoch": 0.7072341603307076,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0009297482679499894,
+      "loss": 1.157,
+      "step": 8065
+    },
+    {
+      "epoch": 0.7073218521050821,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009294034856442891,
+      "loss": 1.0451,
+      "step": 8066
+    },
+    {
+      "epoch": 0.7074095438794566,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009290587690551309,
+      "loss": 1.2029,
+      "step": 8067
+    },
+    {
+      "epoch": 0.7074972356538313,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0009287141182139516,
+      "loss": 1.122,
+      "step": 8068
+    },
+    {
+      "epoch": 0.7075849274282058,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009283695331521831,
+      "loss": 1.1352,
+      "step": 8069
+    },
+    {
+      "epoch": 0.7076726192025803,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00092802501390125,
+      "loss": 1.1998,
+      "step": 8070
+    },
+    {
+      "epoch": 0.7077603109769549,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009276805604925718,
+      "loss": 1.0744,
+      "step": 8071
+    },
+    {
+      "epoch": 0.7078480027513294,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0009273361729575618,
+      "loss": 1.1178,
+      "step": 8072
+    },
+    {
+      "epoch": 0.7079356945257039,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0009269918513276278,
+      "loss": 1.1662,
+      "step": 8073
+    },
+    {
+      "epoch": 0.7080233863000785,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0009266475956341707,
+      "loss": 1.132,
+      "step": 8074
+    },
+    {
+      "epoch": 0.7081110780744531,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0009263034059085848,
+      "loss": 1.144,
+      "step": 8075
+    },
+    {
+      "epoch": 0.7081987698488276,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009259592821822603,
+      "loss": 1.1731,
+      "step": 8076
+    },
+    {
+      "epoch": 0.7082864616232022,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0009256152244865801,
+      "loss": 1.1816,
+      "step": 8077
+    },
+    {
+      "epoch": 0.7083741533975767,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009252712328529216,
+      "loss": 1.1394,
+      "step": 8078
+    },
+    {
+      "epoch": 0.7084618451719512,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009249273073126556,
+      "loss": 1.1457,
+      "step": 8079
+    },
+    {
+      "epoch": 0.7085495369463258,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0009245834478971474,
+      "loss": 1.1595,
+      "step": 8080
+    },
+    {
+      "epoch": 0.7086372287207003,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0009242396546377561,
+      "loss": 1.1558,
+      "step": 8081
+    },
+    {
+      "epoch": 0.7087249204950749,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0009238959275658348,
+      "loss": 1.1925,
+      "step": 8082
+    },
+    {
+      "epoch": 0.7088126122694495,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0009235522667127307,
+      "loss": 1.16,
+      "step": 8083
+    },
+    {
+      "epoch": 0.708900304043824,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0009232086721097844,
+      "loss": 1.1077,
+      "step": 8084
+    },
+    {
+      "epoch": 0.7089879958181985,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0009228651437883308,
+      "loss": 1.1019,
+      "step": 8085
+    },
+    {
+      "epoch": 0.709075687592573,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009225216817796989,
+      "loss": 1.1505,
+      "step": 8086
+    },
+    {
+      "epoch": 0.7091633793669476,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0009221782861152125,
+      "loss": 1.0998,
+      "step": 8087
+    },
+    {
+      "epoch": 0.7092510711413221,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0009218349568261868,
+      "loss": 1.1861,
+      "step": 8088
+    },
+    {
+      "epoch": 0.7093387629156966,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0009214916939439335,
+      "loss": 1.1069,
+      "step": 8089
+    },
+    {
+      "epoch": 0.7094264546900713,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009211484974997568,
+      "loss": 1.1707,
+      "step": 8090
+    },
+    {
+      "epoch": 0.7095141464644458,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000920805367524956,
+      "loss": 1.0944,
+      "step": 8091
+    },
+    {
+      "epoch": 0.7096018382388203,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009204623040508237,
+      "loss": 1.1431,
+      "step": 8092
+    },
+    {
+      "epoch": 0.7096895300131949,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009201193071086453,
+      "loss": 1.1128,
+      "step": 8093
+    },
+    {
+      "epoch": 0.7097772217875694,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0009197763767297018,
+      "loss": 1.1169,
+      "step": 8094
+    },
+    {
+      "epoch": 0.7098649135619439,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0009194335129452678,
+      "loss": 1.1213,
+      "step": 8095
+    },
+    {
+      "epoch": 0.7099526053363185,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000919090715786612,
+      "loss": 1.1245,
+      "step": 8096
+    },
+    {
+      "epoch": 0.710040297110693,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0009187479852849954,
+      "loss": 1.1705,
+      "step": 8097
+    },
+    {
+      "epoch": 0.7101279888850676,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009184053214716751,
+      "loss": 1.1951,
+      "step": 8098
+    },
+    {
+      "epoch": 0.7102156806594422,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0009180627243779006,
+      "loss": 1.1532,
+      "step": 8099
+    },
+    {
+      "epoch": 0.7103033724338167,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0009177201940349168,
+      "loss": 1.1359,
+      "step": 8100
+    },
+    {
+      "epoch": 0.7103910642081912,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0009173777304739603,
+      "loss": 1.1392,
+      "step": 8101
+    },
+    {
+      "epoch": 0.7104787559825658,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0009170353337262638,
+      "loss": 1.1034,
+      "step": 8102
+    },
+    {
+      "epoch": 0.7105664477569403,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0009166930038230526,
+      "loss": 1.1511,
+      "step": 8103
+    },
+    {
+      "epoch": 0.7106541395313148,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0009163507407955469,
+      "loss": 1.1505,
+      "step": 8104
+    },
+    {
+      "epoch": 0.7107418313056895,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009160085446749592,
+      "loss": 1.2235,
+      "step": 8105
+    },
+    {
+      "epoch": 0.710829523080064,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009156664154924974,
+      "loss": 1.1778,
+      "step": 8106
+    },
+    {
+      "epoch": 0.7109172148544385,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0009153243532793629,
+      "loss": 1.1308,
+      "step": 8107
+    },
+    {
+      "epoch": 0.7110049066288131,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0009149823580667515,
+      "loss": 1.1633,
+      "step": 8108
+    },
+    {
+      "epoch": 0.7110925984031876,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009146404298858508,
+      "loss": 1.1336,
+      "step": 8109
+    },
+    {
+      "epoch": 0.7111802901775621,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0009142985687678451,
+      "loss": 1.0819,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7112679819519366,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0009139567747439105,
+      "loss": 1.101,
+      "step": 8111
+    },
+    {
+      "epoch": 0.7113556737263113,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009136150478452177,
+      "loss": 1.1218,
+      "step": 8112
+    },
+    {
+      "epoch": 0.7114433655006858,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0009132733881029317,
+      "loss": 1.1529,
+      "step": 8113
+    },
+    {
+      "epoch": 0.7115310572750603,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0009129317955482114,
+      "loss": 1.1269,
+      "step": 8114
+    },
+    {
+      "epoch": 0.7116187490494349,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000912590270212208,
+      "loss": 1.1572,
+      "step": 8115
+    },
+    {
+      "epoch": 0.7117064408238094,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0009122488121260683,
+      "loss": 1.1743,
+      "step": 8116
+    },
+    {
+      "epoch": 0.7117941325981839,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0009119074213209327,
+      "loss": 1.0585,
+      "step": 8117
+    },
+    {
+      "epoch": 0.7118818243725585,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.000911566097827935,
+      "loss": 1.1368,
+      "step": 8118
+    },
+    {
+      "epoch": 0.711969516146933,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0009112248416782027,
+      "loss": 1.1222,
+      "step": 8119
+    },
+    {
+      "epoch": 0.7120572079213076,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0009108836529028577,
+      "loss": 1.1348,
+      "step": 8120
+    },
+    {
+      "epoch": 0.7121448996956822,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0009105425315330153,
+      "loss": 1.1329,
+      "step": 8121
+    },
+    {
+      "epoch": 0.7122325914700567,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0009102014775997859,
+      "loss": 1.1363,
+      "step": 8122
+    },
+    {
+      "epoch": 0.7123202832444312,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0009098604911342713,
+      "loss": 1.174,
+      "step": 8123
+    },
+    {
+      "epoch": 0.7124079750188058,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0009095195721675692,
+      "loss": 1.1272,
+      "step": 8124
+    },
+    {
+      "epoch": 0.7124956667931803,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0009091787207307712,
+      "loss": 1.1392,
+      "step": 8125
+    },
+    {
+      "epoch": 0.7125833585675548,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0009088379368549609,
+      "loss": 1.1088,
+      "step": 8126
+    },
+    {
+      "epoch": 0.7126710503419295,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0009084972205712181,
+      "loss": 1.1153,
+      "step": 8127
+    },
+    {
+      "epoch": 0.712758742116304,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0009081565719106139,
+      "loss": 1.1241,
+      "step": 8128
+    },
+    {
+      "epoch": 0.7128464338906785,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0009078159909042154,
+      "loss": 1.1815,
+      "step": 8129
+    },
+    {
+      "epoch": 0.7129341256650531,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009074754775830826,
+      "loss": 1.1789,
+      "step": 8130
+    },
+    {
+      "epoch": 0.7130218174394276,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00090713503197827,
+      "loss": 1.1215,
+      "step": 8131
+    },
+    {
+      "epoch": 0.7131095092138021,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009067946541208241,
+      "loss": 1.1278,
+      "step": 8132
+    },
+    {
+      "epoch": 0.7131972009881766,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0009064543440417873,
+      "loss": 1.1864,
+      "step": 8133
+    },
+    {
+      "epoch": 0.7132848927625512,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0009061141017721948,
+      "loss": 1.1689,
+      "step": 8134
+    },
+    {
+      "epoch": 0.7133725845369258,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009057739273430759,
+      "loss": 1.1247,
+      "step": 8135
+    },
+    {
+      "epoch": 0.7134602763113003,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0009054338207854541,
+      "loss": 1.0898,
+      "step": 8136
+    },
+    {
+      "epoch": 0.7135479680856749,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009050937821303452,
+      "loss": 1.0819,
+      "step": 8137
+    },
+    {
+      "epoch": 0.7136356598600494,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0009047538114087606,
+      "loss": 1.1521,
+      "step": 8138
+    },
+    {
+      "epoch": 0.7137233516344239,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009044139086517044,
+      "loss": 1.1245,
+      "step": 8139
+    },
+    {
+      "epoch": 0.7138110434087985,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009040740738901755,
+      "loss": 1.1293,
+      "step": 8140
+    },
+    {
+      "epoch": 0.713898735183173,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000903734307155165,
+      "loss": 1.0882,
+      "step": 8141
+    },
+    {
+      "epoch": 0.7139864269575475,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0009033946084776598,
+      "loss": 1.059,
+      "step": 8142
+    },
+    {
+      "epoch": 0.7140741187319222,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0009030549778886385,
+      "loss": 1.1671,
+      "step": 8143
+    },
+    {
+      "epoch": 0.7141618105062967,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009027154154190748,
+      "loss": 1.0945,
+      "step": 8144
+    },
+    {
+      "epoch": 0.7142495022806712,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0009023759210999367,
+      "loss": 1.1182,
+      "step": 8145
+    },
+    {
+      "epoch": 0.7143371940550458,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009020364949621839,
+      "loss": 1.1655,
+      "step": 8146
+    },
+    {
+      "epoch": 0.7144248858294203,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009016971370367721,
+      "loss": 1.1338,
+      "step": 8147
+    },
+    {
+      "epoch": 0.7145125776037948,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009013578473546498,
+      "loss": 1.1451,
+      "step": 8148
+    },
+    {
+      "epoch": 0.7146002693781695,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0009010186259467594,
+      "loss": 1.1313,
+      "step": 8149
+    },
+    {
+      "epoch": 0.714687961152544,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0009006794728440365,
+      "loss": 1.0902,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7147756529269185,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0009003403880774114,
+      "loss": 1.1349,
+      "step": 8151
+    },
+    {
+      "epoch": 0.7148633447012931,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0009000013716778074,
+      "loss": 1.061,
+      "step": 8152
+    },
+    {
+      "epoch": 0.7149510364756676,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008996624236761431,
+      "loss": 1.1497,
+      "step": 8153
+    },
+    {
+      "epoch": 0.7150387282500421,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008993235441033282,
+      "loss": 1.1541,
+      "step": 8154
+    },
+    {
+      "epoch": 0.7151264200244167,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008989847329902678,
+      "loss": 1.1061,
+      "step": 8155
+    },
+    {
+      "epoch": 0.7152141117987912,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008986459903678615,
+      "loss": 1.1606,
+      "step": 8156
+    },
+    {
+      "epoch": 0.7153018035731658,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008983073162670016,
+      "loss": 1.1395,
+      "step": 8157
+    },
+    {
+      "epoch": 0.7153894953475403,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008979687107185733,
+      "loss": 1.1943,
+      "step": 8158
+    },
+    {
+      "epoch": 0.7154771871219149,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008976301737534579,
+      "loss": 1.1356,
+      "step": 8159
+    },
+    {
+      "epoch": 0.7155648788962894,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008972917054025278,
+      "loss": 1.1445,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7156525706706639,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0008969533056966512,
+      "loss": 1.1333,
+      "step": 8161
+    },
+    {
+      "epoch": 0.7157402624450385,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.000896614974666689,
+      "loss": 1.0955,
+      "step": 8162
+    },
+    {
+      "epoch": 0.715827954219413,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008962767123434967,
+      "loss": 1.1486,
+      "step": 8163
+    },
+    {
+      "epoch": 0.7159156459937875,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0008959385187579222,
+      "loss": 1.0904,
+      "step": 8164
+    },
+    {
+      "epoch": 0.7160033377681622,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008956003939408082,
+      "loss": 1.1531,
+      "step": 8165
+    },
+    {
+      "epoch": 0.7160910295425367,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008952623379229907,
+      "loss": 1.1658,
+      "step": 8166
+    },
+    {
+      "epoch": 0.7161787213169112,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0008949243507353002,
+      "loss": 1.1237,
+      "step": 8167
+    },
+    {
+      "epoch": 0.7162664130912858,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008945864324085591,
+      "loss": 1.1002,
+      "step": 8168
+    },
+    {
+      "epoch": 0.7163541048656603,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008942485829735853,
+      "loss": 1.1475,
+      "step": 8169
+    },
+    {
+      "epoch": 0.7164417966400348,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008939108024611899,
+      "loss": 1.1785,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7165294884144094,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008935730909021779,
+      "loss": 1.0951,
+      "step": 8171
+    },
+    {
+      "epoch": 0.716617180188784,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008932354483273467,
+      "loss": 1.1179,
+      "step": 8172
+    },
+    {
+      "epoch": 0.7167048719631585,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008928978747674892,
+      "loss": 1.1436,
+      "step": 8173
+    },
+    {
+      "epoch": 0.7167925637375331,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000892560370253391,
+      "loss": 1.1567,
+      "step": 8174
+    },
+    {
+      "epoch": 0.7168802555119076,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008922229348158322,
+      "loss": 1.124,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7169679472862821,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0008918855684855856,
+      "loss": 1.111,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7170556390606567,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008915482712934176,
+      "loss": 1.1377,
+      "step": 8177
+    },
+    {
+      "epoch": 0.7171433308350312,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008912110432700892,
+      "loss": 1.1149,
+      "step": 8178
+    },
+    {
+      "epoch": 0.7172310226094057,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008908738844463548,
+      "loss": 1.1796,
+      "step": 8179
+    },
+    {
+      "epoch": 0.7173187143837804,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008905367948529632,
+      "loss": 1.1376,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7174064061581549,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008901997745206547,
+      "loss": 1.1642,
+      "step": 8181
+    },
+    {
+      "epoch": 0.7174940979325294,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008898628234801653,
+      "loss": 1.1191,
+      "step": 8182
+    },
+    {
+      "epoch": 0.7175817897069039,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008895259417622244,
+      "loss": 1.1662,
+      "step": 8183
+    },
+    {
+      "epoch": 0.7176694814812785,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008891891293975548,
+      "loss": 1.1617,
+      "step": 8184
+    },
+    {
+      "epoch": 0.717757173255653,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008888523864168718,
+      "loss": 1.1331,
+      "step": 8185
+    },
+    {
+      "epoch": 0.7178448650300275,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008885157128508867,
+      "loss": 1.2114,
+      "step": 8186
+    },
+    {
+      "epoch": 0.7179325568044022,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008881791087303027,
+      "loss": 1.1043,
+      "step": 8187
+    },
+    {
+      "epoch": 0.7180202485787767,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008878425740858172,
+      "loss": 1.1973,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7181079403531512,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.000887506108948122,
+      "loss": 1.1264,
+      "step": 8189
+    },
+    {
+      "epoch": 0.7181956321275258,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008871697133479008,
+      "loss": 1.1412,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7182833239019003,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008868333873158326,
+      "loss": 1.1072,
+      "step": 8191
+    },
+    {
+      "epoch": 0.7183710156762748,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008864971308825894,
+      "loss": 1.1245,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7184587074506494,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008861609440788372,
+      "loss": 1.1238,
+      "step": 8193
+    },
+    {
+      "epoch": 0.718546399225024,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0008858248269352352,
+      "loss": 1.1278,
+      "step": 8194
+    },
+    {
+      "epoch": 0.7186340909993985,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008854887794824357,
+      "loss": 1.1412,
+      "step": 8195
+    },
+    {
+      "epoch": 0.7187217827737731,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008851528017510859,
+      "loss": 1.137,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7188094745481476,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008848168937718263,
+      "loss": 1.1153,
+      "step": 8197
+    },
+    {
+      "epoch": 0.7188971663225221,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0008844810555752912,
+      "loss": 1.1245,
+      "step": 8198
+    },
+    {
+      "epoch": 0.7189848580968967,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008841452871921069,
+      "loss": 1.1581,
+      "step": 8199
+    },
+    {
+      "epoch": 0.7190725498712712,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008838095886528956,
+      "loss": 1.1094,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7191602416456457,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008834739599882718,
+      "loss": 1.1283,
+      "step": 8201
+    },
+    {
+      "epoch": 0.7192479334200204,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008831384012288445,
+      "loss": 1.1577,
+      "step": 8202
+    },
+    {
+      "epoch": 0.7193356251943949,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.000882802912405215,
+      "loss": 1.1222,
+      "step": 8203
+    },
+    {
+      "epoch": 0.7194233169687694,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008824674935479792,
+      "loss": 1.1365,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7195110087431439,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0008821321446877269,
+      "loss": 1.119,
+      "step": 8205
+    },
+    {
+      "epoch": 0.7195987005175185,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008817968658550411,
+      "loss": 1.1557,
+      "step": 8206
+    },
+    {
+      "epoch": 0.719686392291893,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008814616570804974,
+      "loss": 1.0611,
+      "step": 8207
+    },
+    {
+      "epoch": 0.7197740840662675,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0008811265183946669,
+      "loss": 1.1068,
+      "step": 8208
+    },
+    {
+      "epoch": 0.7198617758406421,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008807914498281131,
+      "loss": 1.1046,
+      "step": 8209
+    },
+    {
+      "epoch": 0.7199494676150167,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008804564514113935,
+      "loss": 1.1496,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7200371593893912,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008801215231750592,
+      "loss": 1.0697,
+      "step": 8211
+    },
+    {
+      "epoch": 0.7201248511637658,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0008797866651496541,
+      "loss": 1.1424,
+      "step": 8212
+    },
+    {
+      "epoch": 0.7202125429381403,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0008794518773657166,
+      "loss": 1.1677,
+      "step": 8213
+    },
+    {
+      "epoch": 0.7203002347125148,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008791171598537791,
+      "loss": 1.1895,
+      "step": 8214
+    },
+    {
+      "epoch": 0.7203879264868894,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0008787825126443666,
+      "loss": 1.1127,
+      "step": 8215
+    },
+    {
+      "epoch": 0.7204756182612639,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.000878447935767998,
+      "loss": 1.1392,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7205633100356384,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0008781134292551858,
+      "loss": 1.0924,
+      "step": 8217
+    },
+    {
+      "epoch": 0.7206510018100131,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008777789931364363,
+      "loss": 1.1449,
+      "step": 8218
+    },
+    {
+      "epoch": 0.7207386935843876,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.000877444627442249,
+      "loss": 1.1213,
+      "step": 8219
+    },
+    {
+      "epoch": 0.7208263853587621,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008771103322031177,
+      "loss": 1.1579,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7209140771331367,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0008767761074495287,
+      "loss": 1.1059,
+      "step": 8221
+    },
+    {
+      "epoch": 0.7210017689075112,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0008764419532119625,
+      "loss": 1.1148,
+      "step": 8222
+    },
+    {
+      "epoch": 0.7210894606818857,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008761078695208931,
+      "loss": 1.0967,
+      "step": 8223
+    },
+    {
+      "epoch": 0.7211771524562604,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008757738564067888,
+      "loss": 1.1421,
+      "step": 8224
+    },
+    {
+      "epoch": 0.7212648442306349,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008754399139001098,
+      "loss": 1.1592,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7213525360050094,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000875106042031311,
+      "loss": 1.1389,
+      "step": 8226
+    },
+    {
+      "epoch": 0.721440227779384,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0008747722408308413,
+      "loss": 1.1119,
+      "step": 8227
+    },
+    {
+      "epoch": 0.7215279195537585,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0008744385103291415,
+      "loss": 1.0484,
+      "step": 8228
+    },
+    {
+      "epoch": 0.721615611328133,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000874104850556648,
+      "loss": 1.1597,
+      "step": 8229
+    },
+    {
+      "epoch": 0.7217033031025075,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008737712615437887,
+      "loss": 1.2046,
+      "step": 8230
+    },
+    {
+      "epoch": 0.7217909948768821,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0008734377433209864,
+      "loss": 1.109,
+      "step": 8231
+    },
+    {
+      "epoch": 0.7218786866512567,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008731042959186573,
+      "loss": 1.1816,
+      "step": 8232
+    },
+    {
+      "epoch": 0.7219663784256312,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008727709193672113,
+      "loss": 1.1684,
+      "step": 8233
+    },
+    {
+      "epoch": 0.7220540702000058,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0008724376136970505,
+      "loss": 1.1199,
+      "step": 8234
+    },
+    {
+      "epoch": 0.7221417619743803,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008721043789385723,
+      "loss": 1.1543,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7222294537487548,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0008717712151221663,
+      "loss": 1.1432,
+      "step": 8236
+    },
+    {
+      "epoch": 0.7223171455231294,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0008714381222782172,
+      "loss": 1.1117,
+      "step": 8237
+    },
+    {
+      "epoch": 0.7224048372975039,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.000871105100437101,
+      "loss": 1.1142,
+      "step": 8238
+    },
+    {
+      "epoch": 0.7224925290718784,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008707721496291888,
+      "loss": 1.1357,
+      "step": 8239
+    },
+    {
+      "epoch": 0.7225802208462531,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008704392698848451,
+      "loss": 1.1223,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7226679126206276,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008701064612344277,
+      "loss": 1.107,
+      "step": 8241
+    },
+    {
+      "epoch": 0.7227556043950021,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000869773723708288,
+      "loss": 1.157,
+      "step": 8242
+    },
+    {
+      "epoch": 0.7228432961693767,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008694410573367703,
+      "loss": 1.1855,
+      "step": 8243
+    },
+    {
+      "epoch": 0.7229309879437512,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008691084621502134,
+      "loss": 1.141,
+      "step": 8244
+    },
+    {
+      "epoch": 0.7230186797181257,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0008687759381789485,
+      "loss": 1.105,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7231063714925003,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008684434854533014,
+      "loss": 1.1051,
+      "step": 8246
+    },
+    {
+      "epoch": 0.7231940632668749,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0008681111040035913,
+      "loss": 1.114,
+      "step": 8247
+    },
+    {
+      "epoch": 0.7232817550412494,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008677787938601296,
+      "loss": 1.1324,
+      "step": 8248
+    },
+    {
+      "epoch": 0.723369446815624,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0008674465550532224,
+      "loss": 1.1451,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7234571385899985,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008671143876131694,
+      "loss": 1.089,
+      "step": 8250
+    },
+    {
+      "epoch": 0.723544830364373,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0008667822915702639,
+      "loss": 1.132,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7236325221387476,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0008664502669547908,
+      "loss": 1.1319,
+      "step": 8252
+    },
+    {
+      "epoch": 0.7237202139131221,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008661183137970306,
+      "loss": 1.0913,
+      "step": 8253
+    },
+    {
+      "epoch": 0.7238079056874966,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008657864321272567,
+      "loss": 1.1528,
+      "step": 8254
+    },
+    {
+      "epoch": 0.7238955974618712,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0008654546219757363,
+      "loss": 1.1565,
+      "step": 8255
+    },
+    {
+      "epoch": 0.7239832892362458,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008651228833727283,
+      "loss": 1.1178,
+      "step": 8256
+    },
+    {
+      "epoch": 0.7240709810106203,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0008647912163484876,
+      "loss": 1.068,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7241586727849948,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008644596209332608,
+      "loss": 1.2285,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7242463645593694,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008641280971572892,
+      "loss": 1.137,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7243340563337439,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008637966450508061,
+      "loss": 1.1765,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7244217481081184,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008634652646440396,
+      "loss": 1.1462,
+      "step": 8261
+    },
+    {
+      "epoch": 0.7245094398824931,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008631339559672101,
+      "loss": 1.1458,
+      "step": 8262
+    },
+    {
+      "epoch": 0.7245971316568676,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008628027190505326,
+      "loss": 1.128,
+      "step": 8263
+    },
+    {
+      "epoch": 0.7246848234312421,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008624715539242154,
+      "loss": 1.1438,
+      "step": 8264
+    },
+    {
+      "epoch": 0.7247725152056167,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008621404606184591,
+      "loss": 1.1505,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7248602069799912,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008618094391634589,
+      "loss": 1.1551,
+      "step": 8266
+    },
+    {
+      "epoch": 0.7249478987543657,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000861478489589403,
+      "loss": 1.1217,
+      "step": 8267
+    },
+    {
+      "epoch": 0.7250355905287403,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.000861147611926474,
+      "loss": 1.1507,
+      "step": 8268
+    },
+    {
+      "epoch": 0.7251232823031148,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000860816806204846,
+      "loss": 1.1517,
+      "step": 8269
+    },
+    {
+      "epoch": 0.7252109740774894,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008604860724546878,
+      "loss": 1.2101,
+      "step": 8270
+    },
+    {
+      "epoch": 0.725298665851864,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008601554107061616,
+      "loss": 1.0934,
+      "step": 8271
+    },
+    {
+      "epoch": 0.7253863576262385,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008598248209894231,
+      "loss": 1.1272,
+      "step": 8272
+    },
+    {
+      "epoch": 0.725474049400613,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008594943033346216,
+      "loss": 1.133,
+      "step": 8273
+    },
+    {
+      "epoch": 0.7255617411749876,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008591638577718984,
+      "loss": 1.1075,
+      "step": 8274
+    },
+    {
+      "epoch": 0.7256494329493621,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008588334843313899,
+      "loss": 1.1525,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7257371247237366,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.000858503183043225,
+      "loss": 1.1431,
+      "step": 8276
+    },
+    {
+      "epoch": 0.7258248164981111,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0008581729539375273,
+      "loss": 1.1179,
+      "step": 8277
+    },
+    {
+      "epoch": 0.7259125082724858,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0008578427970444122,
+      "loss": 1.0911,
+      "step": 8278
+    },
+    {
+      "epoch": 0.7260002000468603,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008575127123939884,
+      "loss": 1.12,
+      "step": 8279
+    },
+    {
+      "epoch": 0.7260878918212348,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0008571827000163596,
+      "loss": 1.0798,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7261755835956094,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0008568527599416216,
+      "loss": 1.1519,
+      "step": 8281
+    },
+    {
+      "epoch": 0.7262632753699839,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008565228921998652,
+      "loss": 1.1446,
+      "step": 8282
+    },
+    {
+      "epoch": 0.7263509671443584,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0008561930968211721,
+      "loss": 1.0922,
+      "step": 8283
+    },
+    {
+      "epoch": 0.726438658918733,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008558633738356193,
+      "loss": 1.0825,
+      "step": 8284
+    },
+    {
+      "epoch": 0.7265263506931076,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008555337232732768,
+      "loss": 1.15,
+      "step": 8285
+    },
+    {
+      "epoch": 0.7266140424674821,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0008552041451642083,
+      "loss": 1.2071,
+      "step": 8286
+    },
+    {
+      "epoch": 0.7267017342418567,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008548746395384695,
+      "loss": 1.1457,
+      "step": 8287
+    },
+    {
+      "epoch": 0.7267894260162312,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.000854545206426111,
+      "loss": 1.1105,
+      "step": 8288
+    },
+    {
+      "epoch": 0.7268771177906057,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008542158458571763,
+      "loss": 1.1139,
+      "step": 8289
+    },
+    {
+      "epoch": 0.7269648095649803,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0008538865578617025,
+      "loss": 1.1922,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7270525013393548,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.000853557342469719,
+      "loss": 1.1537,
+      "step": 8291
+    },
+    {
+      "epoch": 0.7271401931137293,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0008532281997112497,
+      "loss": 1.1147,
+      "step": 8292
+    },
+    {
+      "epoch": 0.727227884888104,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008528991296163122,
+      "loss": 1.164,
+      "step": 8293
+    },
+    {
+      "epoch": 0.7273155766624785,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0008525701322149165,
+      "loss": 1.1176,
+      "step": 8294
+    },
+    {
+      "epoch": 0.727403268436853,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0008522412075370663,
+      "loss": 1.1592,
+      "step": 8295
+    },
+    {
+      "epoch": 0.7274909602112276,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.000851912355612758,
+      "loss": 1.1747,
+      "step": 8296
+    },
+    {
+      "epoch": 0.7275786519856021,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0008515835764719826,
+      "loss": 1.129,
+      "step": 8297
+    },
+    {
+      "epoch": 0.7276663437599766,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.000851254870144724,
+      "loss": 1.0943,
+      "step": 8298
+    },
+    {
+      "epoch": 0.7277540355343513,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0008509262366609594,
+      "loss": 1.1115,
+      "step": 8299
+    },
+    {
+      "epoch": 0.7278417273087258,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008505976760506596,
+      "loss": 1.1402,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7279294190831003,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008502691883437876,
+      "loss": 1.1216,
+      "step": 8301
+    },
+    {
+      "epoch": 0.7280171108574748,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008499407735703011,
+      "loss": 1.1382,
+      "step": 8302
+    },
+    {
+      "epoch": 0.7281048026318494,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0008496124317601507,
+      "loss": 1.1686,
+      "step": 8303
+    },
+    {
+      "epoch": 0.7281924944062239,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000849284162943281,
+      "loss": 1.1039,
+      "step": 8304
+    },
+    {
+      "epoch": 0.7282801861805984,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008489559671496278,
+      "loss": 1.1901,
+      "step": 8305
+    },
+    {
+      "epoch": 0.728367877954973,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0008486278444091227,
+      "loss": 1.1143,
+      "step": 8306
+    },
+    {
+      "epoch": 0.7284555697293476,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008482997947516895,
+      "loss": 1.1289,
+      "step": 8307
+    },
+    {
+      "epoch": 0.7285432615037221,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0008479718182072458,
+      "loss": 1.168,
+      "step": 8308
+    },
+    {
+      "epoch": 0.7286309532780967,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008476439148057015,
+      "loss": 1.098,
+      "step": 8309
+    },
+    {
+      "epoch": 0.7287186450524712,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000847316084576961,
+      "loss": 1.1238,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7288063368268457,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0008469883275509217,
+      "loss": 1.1249,
+      "step": 8311
+    },
+    {
+      "epoch": 0.7288940286012203,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008466606437574734,
+      "loss": 1.1589,
+      "step": 8312
+    },
+    {
+      "epoch": 0.7289817203755948,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008463330332265012,
+      "loss": 1.1232,
+      "step": 8313
+    },
+    {
+      "epoch": 0.7290694121499693,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008460054959878815,
+      "loss": 1.0965,
+      "step": 8314
+    },
+    {
+      "epoch": 0.729157103924344,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008456780320714849,
+      "loss": 1.0877,
+      "step": 8315
+    },
+    {
+      "epoch": 0.7292447956987185,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008453506415071752,
+      "loss": 1.1332,
+      "step": 8316
+    },
+    {
+      "epoch": 0.729332487473093,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008450233243248105,
+      "loss": 1.1446,
+      "step": 8317
+    },
+    {
+      "epoch": 0.7294201792474676,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00084469608055424,
+      "loss": 1.1769,
+      "step": 8318
+    },
+    {
+      "epoch": 0.7295078710218421,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008443689102253083,
+      "loss": 1.0746,
+      "step": 8319
+    },
+    {
+      "epoch": 0.7295955627962166,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0008440418133678519,
+      "loss": 1.129,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7296832545705912,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0008437147900117022,
+      "loss": 1.1121,
+      "step": 8321
+    },
+    {
+      "epoch": 0.7297709463449658,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008433878401866818,
+      "loss": 1.1072,
+      "step": 8322
+    },
+    {
+      "epoch": 0.7298586381193403,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008430609639226079,
+      "loss": 1.1073,
+      "step": 8323
+    },
+    {
+      "epoch": 0.7299463298937148,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008427341612492915,
+      "loss": 1.1259,
+      "step": 8324
+    },
+    {
+      "epoch": 0.7300340216680894,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008424074321965351,
+      "loss": 1.0869,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7301217134424639,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008420807767941369,
+      "loss": 1.14,
+      "step": 8326
+    },
+    {
+      "epoch": 0.7302094052168384,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0008417541950718856,
+      "loss": 1.1342,
+      "step": 8327
+    },
+    {
+      "epoch": 0.730297096991213,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008414276870595657,
+      "loss": 1.1242,
+      "step": 8328
+    },
+    {
+      "epoch": 0.7303847887655875,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 0.000841101252786953,
+      "loss": 1.08,
+      "step": 8329
+    },
+    {
+      "epoch": 0.730472480539962,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008407748922838179,
+      "loss": 1.0879,
+      "step": 8330
+    },
+    {
+      "epoch": 0.7305601723143367,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0008404486055799239,
+      "loss": 1.1088,
+      "step": 8331
+    },
+    {
+      "epoch": 0.7306478640887112,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000840122392705027,
+      "loss": 1.1781,
+      "step": 8332
+    },
+    {
+      "epoch": 0.7307355558630857,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0008397962536888772,
+      "loss": 1.0915,
+      "step": 8333
+    },
+    {
+      "epoch": 0.7308232476374603,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008394701885612175,
+      "loss": 1.1008,
+      "step": 8334
+    },
+    {
+      "epoch": 0.7309109394118348,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008391441973517845,
+      "loss": 1.1234,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7309986311862093,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0008388182800903074,
+      "loss": 1.1562,
+      "step": 8336
+    },
+    {
+      "epoch": 0.731086322960584,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008384924368065088,
+      "loss": 1.2029,
+      "step": 8337
+    },
+    {
+      "epoch": 0.7311740147349585,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008381666675301051,
+      "loss": 1.1526,
+      "step": 8338
+    },
+    {
+      "epoch": 0.731261706509333,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008378409722908061,
+      "loss": 1.1425,
+      "step": 8339
+    },
+    {
+      "epoch": 0.7313493982837076,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008375153511183134,
+      "loss": 1.1728,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7314370900580821,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008371898040423233,
+      "loss": 1.1045,
+      "step": 8341
+    },
+    {
+      "epoch": 0.7315247818324566,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008368643310925246,
+      "loss": 1.2176,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7316124736068312,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008365389322986005,
+      "loss": 1.0988,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7317001653812057,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008362136076902251,
+      "loss": 1.14,
+      "step": 8344
+    },
+    {
+      "epoch": 0.7317878571555803,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0008358883572970687,
+      "loss": 1.1626,
+      "step": 8345
+    },
+    {
+      "epoch": 0.7318755489299549,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008355631811487919,
+      "loss": 1.1319,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7319632407043294,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008352380792750505,
+      "loss": 1.1266,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7320509324787039,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008349130517054934,
+      "loss": 1.1362,
+      "step": 8348
+    },
+    {
+      "epoch": 0.7321386242530784,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0008345880984697614,
+      "loss": 1.1265,
+      "step": 8349
+    },
+    {
+      "epoch": 0.732226316027453,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008342632195974901,
+      "loss": 1.1361,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7323140078018275,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008339384151183068,
+      "loss": 1.1975,
+      "step": 8351
+    },
+    {
+      "epoch": 0.732401699576202,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008336136850618338,
+      "loss": 1.1185,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7324893913505767,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008332890294576857,
+      "loss": 1.1775,
+      "step": 8353
+    },
+    {
+      "epoch": 0.7325770831249512,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0008329644483354693,
+      "loss": 1.1219,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7326647748993257,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008326399417247859,
+      "loss": 1.0847,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7327524666737003,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008323155096552299,
+      "loss": 1.1591,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7328401584480748,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008319911521563891,
+      "loss": 1.1414,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7329278502224493,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.000831666869257843,
+      "loss": 1.1184,
+      "step": 8358
+    },
+    {
+      "epoch": 0.733015541996824,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000831342660989166,
+      "loss": 1.137,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7331032337711985,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008310185273799247,
+      "loss": 1.1031,
+      "step": 8360
+    },
+    {
+      "epoch": 0.733190925545573,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008306944684596803,
+      "loss": 1.1634,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7332786173199476,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008303704842579851,
+      "loss": 1.176,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7333663090943221,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008300465748043857,
+      "loss": 1.1661,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7334540008686966,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0008297227401284215,
+      "loss": 1.1129,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7335416926430712,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008293989802596265,
+      "loss": 1.1117,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7336293844174457,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0008290752952275261,
+      "loss": 1.1694,
+      "step": 8366
+    },
+    {
+      "epoch": 0.7337170761918202,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0008287516850616392,
+      "loss": 1.128,
+      "step": 8367
+    },
+    {
+      "epoch": 0.7338047679661949,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008284281497914789,
+      "loss": 1.1343,
+      "step": 8368
+    },
+    {
+      "epoch": 0.7338924597405694,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008281046894465502,
+      "loss": 1.1606,
+      "step": 8369
+    },
+    {
+      "epoch": 0.7339801515149439,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008277813040563523,
+      "loss": 1.1728,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7340678432893185,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008274579936503768,
+      "loss": 1.0915,
+      "step": 8371
+    },
+    {
+      "epoch": 0.734155535063693,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000827134758258109,
+      "loss": 1.222,
+      "step": 8372
+    },
+    {
+      "epoch": 0.7342432268380675,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008268115979090268,
+      "loss": 1.1566,
+      "step": 8373
+    },
+    {
+      "epoch": 0.734330918612442,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0008264885126326023,
+      "loss": 1.0803,
+      "step": 8374
+    },
+    {
+      "epoch": 0.7344186103868167,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008261655024582992,
+      "loss": 1.1495,
+      "step": 8375
+    },
+    {
+      "epoch": 0.7345063021611912,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008258425674155756,
+      "loss": 1.1461,
+      "step": 8376
+    },
+    {
+      "epoch": 0.7345939939355657,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0008255197075338822,
+      "loss": 1.0912,
+      "step": 8377
+    },
+    {
+      "epoch": 0.7346816857099403,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008251969228426631,
+      "loss": 1.1583,
+      "step": 8378
+    },
+    {
+      "epoch": 0.7347693774843148,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008248742133713559,
+      "loss": 1.1411,
+      "step": 8379
+    },
+    {
+      "epoch": 0.7348570692586893,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008245515791493904,
+      "loss": 1.1343,
+      "step": 8380
+    },
+    {
+      "epoch": 0.7349447610330639,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008242290202061893,
+      "loss": 1.0828,
+      "step": 8381
+    },
+    {
+      "epoch": 0.7350324528074385,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008239065365711702,
+      "loss": 1.1374,
+      "step": 8382
+    },
+    {
+      "epoch": 0.735120144581813,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0008235841282737422,
+      "loss": 1.1263,
+      "step": 8383
+    },
+    {
+      "epoch": 0.7352078363561876,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0008232617953433088,
+      "loss": 1.0992,
+      "step": 8384
+    },
+    {
+      "epoch": 0.7352955281305621,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0008229395378092651,
+      "loss": 1.1467,
+      "step": 8385
+    },
+    {
+      "epoch": 0.7353832199049366,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0008226173557010004,
+      "loss": 1.1255,
+      "step": 8386
+    },
+    {
+      "epoch": 0.7354709116793112,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000822295249047897,
+      "loss": 1.1335,
+      "step": 8387
+    },
+    {
+      "epoch": 0.7355586034536857,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0008219732178793305,
+      "loss": 1.128,
+      "step": 8388
+    },
+    {
+      "epoch": 0.7356462952280602,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008216512622246689,
+      "loss": 1.1116,
+      "step": 8389
+    },
+    {
+      "epoch": 0.7357339870024349,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008213293821132736,
+      "loss": 1.1942,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7358216787768094,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008210075775744994,
+      "loss": 1.1393,
+      "step": 8391
+    },
+    {
+      "epoch": 0.7359093705511839,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0008206858486376945,
+      "loss": 1.1297,
+      "step": 8392
+    },
+    {
+      "epoch": 0.7359970623255585,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.000820364195332199,
+      "loss": 1.1,
+      "step": 8393
+    },
+    {
+      "epoch": 0.736084754099933,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008200426176873472,
+      "loss": 1.1379,
+      "step": 8394
+    },
+    {
+      "epoch": 0.7361724458743075,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008197211157324662,
+      "loss": 1.1564,
+      "step": 8395
+    },
+    {
+      "epoch": 0.736260137648682,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008193996894968765,
+      "loss": 1.1245,
+      "step": 8396
+    },
+    {
+      "epoch": 0.7363478294230567,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008190783390098908,
+      "loss": 1.1679,
+      "step": 8397
+    },
+    {
+      "epoch": 0.7364355211974312,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008187570643008152,
+      "loss": 1.1779,
+      "step": 8398
+    },
+    {
+      "epoch": 0.7365232129718057,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008184358653989495,
+      "loss": 1.1787,
+      "step": 8399
+    },
+    {
+      "epoch": 0.7366109047461803,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008181147423335864,
+      "loss": 1.1855,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7366985965205548,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0008177936951340116,
+      "loss": 1.1093,
+      "step": 8401
+    },
+    {
+      "epoch": 0.7367862882949293,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.000817472723829503,
+      "loss": 1.0963,
+      "step": 8402
+    },
+    {
+      "epoch": 0.7368739800693039,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008171518284493327,
+      "loss": 1.1415,
+      "step": 8403
+    },
+    {
+      "epoch": 0.7369616718436784,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008168310090227657,
+      "loss": 1.1224,
+      "step": 8404
+    },
+    {
+      "epoch": 0.737049363618053,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00081651026557906,
+      "loss": 1.2081,
+      "step": 8405
+    },
+    {
+      "epoch": 0.7371370553924276,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008161895981474669,
+      "loss": 1.0961,
+      "step": 8406
+    },
+    {
+      "epoch": 0.7372247471668021,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008158690067572293,
+      "loss": 1.1095,
+      "step": 8407
+    },
+    {
+      "epoch": 0.7373124389411766,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008155484914375853,
+      "loss": 1.139,
+      "step": 8408
+    },
+    {
+      "epoch": 0.7374001307155512,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0008152280522177645,
+      "loss": 1.1328,
+      "step": 8409
+    },
+    {
+      "epoch": 0.7374878224899257,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.000814907689126991,
+      "loss": 1.1528,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7375755142643002,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008145874021944798,
+      "loss": 1.1254,
+      "step": 8411
+    },
+    {
+      "epoch": 0.7376632060386749,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008142671914494413,
+      "loss": 1.1362,
+      "step": 8412
+    },
+    {
+      "epoch": 0.7377508978130494,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008139470569210777,
+      "loss": 1.1209,
+      "step": 8413
+    },
+    {
+      "epoch": 0.7378385895874239,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008136269986385838,
+      "loss": 1.161,
+      "step": 8414
+    },
+    {
+      "epoch": 0.7379262813617985,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000813307016631149,
+      "loss": 1.1072,
+      "step": 8415
+    },
+    {
+      "epoch": 0.738013973136173,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008129871109279537,
+      "loss": 1.1165,
+      "step": 8416
+    },
+    {
+      "epoch": 0.7381016649105475,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0008126672815581736,
+      "loss": 1.1038,
+      "step": 8417
+    },
+    {
+      "epoch": 0.7381893566849221,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0008123475285509755,
+      "loss": 1.1735,
+      "step": 8418
+    },
+    {
+      "epoch": 0.7382770484592966,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000812027851935521,
+      "loss": 1.0937,
+      "step": 8419
+    },
+    {
+      "epoch": 0.7383647402336712,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0008117082517409628,
+      "loss": 1.149,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7384524320080457,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008113887279964481,
+      "loss": 1.1326,
+      "step": 8421
+    },
+    {
+      "epoch": 0.7385401237824203,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008110692807311162,
+      "loss": 1.121,
+      "step": 8422
+    },
+    {
+      "epoch": 0.7386278155567948,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0008107499099741008,
+      "loss": 1.1381,
+      "step": 8423
+    },
+    {
+      "epoch": 0.7387155073311693,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0008104306157545266,
+      "loss": 1.1566,
+      "step": 8424
+    },
+    {
+      "epoch": 0.7388031991055439,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008101113981015132,
+      "loss": 1.1154,
+      "step": 8425
+    },
+    {
+      "epoch": 0.7388908908799184,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0008097922570441718,
+      "loss": 1.1304,
+      "step": 8426
+    },
+    {
+      "epoch": 0.738978582654293,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0008094731926116082,
+      "loss": 1.2247,
+      "step": 8427
+    },
+    {
+      "epoch": 0.7390662744286676,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000809154204832919,
+      "loss": 1.1973,
+      "step": 8428
+    },
+    {
+      "epoch": 0.7391539662030421,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008088352937371959,
+      "loss": 1.1957,
+      "step": 8429
+    },
+    {
+      "epoch": 0.7392416579774166,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.000808516459353523,
+      "loss": 1.1166,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7393293497517912,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0008081977017109762,
+      "loss": 1.1123,
+      "step": 8431
+    },
+    {
+      "epoch": 0.7394170415261657,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0008078790208386261,
+      "loss": 1.1656,
+      "step": 8432
+    },
+    {
+      "epoch": 0.7395047333005402,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008075604167655353,
+      "loss": 1.1343,
+      "step": 8433
+    },
+    {
+      "epoch": 0.7395924250749148,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0008072418895207595,
+      "loss": 1.1157,
+      "step": 8434
+    },
+    {
+      "epoch": 0.7396801168492894,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008069234391333479,
+      "loss": 1.1455,
+      "step": 8435
+    },
+    {
+      "epoch": 0.7397678086236639,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0008066050656323424,
+      "loss": 1.1309,
+      "step": 8436
+    },
+    {
+      "epoch": 0.7398555003980385,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0008062867690467781,
+      "loss": 1.131,
+      "step": 8437
+    },
+    {
+      "epoch": 0.739943192172413,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008059685494056819,
+      "loss": 1.0857,
+      "step": 8438
+    },
+    {
+      "epoch": 0.7400308839467875,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008056504067380751,
+      "loss": 1.1586,
+      "step": 8439
+    },
+    {
+      "epoch": 0.7401185757211621,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0008053323410729716,
+      "loss": 1.1498,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7402062674955366,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008050143524393784,
+      "loss": 1.0999,
+      "step": 8441
+    },
+    {
+      "epoch": 0.7402939592699112,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008046964408662946,
+      "loss": 1.1921,
+      "step": 8442
+    },
+    {
+      "epoch": 0.7403816510442858,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008043786063827134,
+      "loss": 1.1246,
+      "step": 8443
+    },
+    {
+      "epoch": 0.7404693428186603,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00080406084901762,
+      "loss": 1.115,
+      "step": 8444
+    },
+    {
+      "epoch": 0.7405570345930348,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.000803743168799994,
+      "loss": 1.1294,
+      "step": 8445
+    },
+    {
+      "epoch": 0.7406447263674093,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008034255657588061,
+      "loss": 1.1278,
+      "step": 8446
+    },
+    {
+      "epoch": 0.7407324181417839,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008031080399230215,
+      "loss": 1.1057,
+      "step": 8447
+    },
+    {
+      "epoch": 0.7408201099161584,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0008027905913215971,
+      "loss": 1.0668,
+      "step": 8448
+    },
+    {
+      "epoch": 0.7409078016905329,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008024732199834835,
+      "loss": 1.124,
+      "step": 8449
+    },
+    {
+      "epoch": 0.7409954934649076,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008021559259376249,
+      "loss": 1.1174,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7410831852392821,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008018387092129568,
+      "loss": 1.1505,
+      "step": 8451
+    },
+    {
+      "epoch": 0.7411708770136566,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008015215698384087,
+      "loss": 1.1067,
+      "step": 8452
+    },
+    {
+      "epoch": 0.7412585687880312,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008012045078429033,
+      "loss": 1.0593,
+      "step": 8453
+    },
+    {
+      "epoch": 0.7413462605624057,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000800887523255356,
+      "loss": 1.1849,
+      "step": 8454
+    },
+    {
+      "epoch": 0.7414339523367802,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008005706161046742,
+      "loss": 1.1211,
+      "step": 8455
+    },
+    {
+      "epoch": 0.7415216441111548,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008002537864197594,
+      "loss": 1.1241,
+      "step": 8456
+    },
+    {
+      "epoch": 0.7416093358855294,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0007999370342295055,
+      "loss": 1.1306,
+      "step": 8457
+    },
+    {
+      "epoch": 0.7416970276599039,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007996203595627999,
+      "loss": 1.1567,
+      "step": 8458
+    },
+    {
+      "epoch": 0.7417847194342785,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0007993037624485226,
+      "loss": 1.1228,
+      "step": 8459
+    },
+    {
+      "epoch": 0.741872411208653,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0007989872429155458,
+      "loss": 1.1281,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7419601029830275,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0007986708009927353,
+      "loss": 1.138,
+      "step": 8461
+    },
+    {
+      "epoch": 0.7420477947574021,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0007983544367089503,
+      "loss": 1.134,
+      "step": 8462
+    },
+    {
+      "epoch": 0.7421354865317766,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007980381500930423,
+      "loss": 1.117,
+      "step": 8463
+    },
+    {
+      "epoch": 0.7422231783061511,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007977219411738559,
+      "loss": 1.1518,
+      "step": 8464
+    },
+    {
+      "epoch": 0.7423108700805258,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000797405809980228,
+      "loss": 1.0851,
+      "step": 8465
+    },
+    {
+      "epoch": 0.7423985618549003,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000797089756540989,
+      "loss": 1.1438,
+      "step": 8466
+    },
+    {
+      "epoch": 0.7424862536292748,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0007967737808849625,
+      "loss": 1.1092,
+      "step": 8467
+    },
+    {
+      "epoch": 0.7425739454036493,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0007964578830409652,
+      "loss": 1.1362,
+      "step": 8468
+    },
+    {
+      "epoch": 0.7426616371780239,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0007961420630378048,
+      "loss": 1.1725,
+      "step": 8469
+    },
+    {
+      "epoch": 0.7427493289523984,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007958263209042844,
+      "loss": 1.0755,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7428370207267729,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007955106566691984,
+      "loss": 1.2014,
+      "step": 8471
+    },
+    {
+      "epoch": 0.7429247125011476,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0007951950703613351,
+      "loss": 1.1219,
+      "step": 8472
+    },
+    {
+      "epoch": 0.7430124042755221,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0007948795620094741,
+      "loss": 1.1353,
+      "step": 8473
+    },
+    {
+      "epoch": 0.7431000960498966,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0007945641316423896,
+      "loss": 1.1229,
+      "step": 8474
+    },
+    {
+      "epoch": 0.7431877878242712,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000794248779288848,
+      "loss": 1.2321,
+      "step": 8475
+    },
+    {
+      "epoch": 0.7432754795986457,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007939335049776092,
+      "loss": 1.1559,
+      "step": 8476
+    },
+    {
+      "epoch": 0.7433631713730202,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0007936183087374244,
+      "loss": 1.087,
+      "step": 8477
+    },
+    {
+      "epoch": 0.7434508631473948,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007933031905970388,
+      "loss": 1.1271,
+      "step": 8478
+    },
+    {
+      "epoch": 0.7435385549217693,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0007929881505851909,
+      "loss": 1.1422,
+      "step": 8479
+    },
+    {
+      "epoch": 0.7436262466961439,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0007926731887306117,
+      "loss": 1.1394,
+      "step": 8480
+    },
+    {
+      "epoch": 0.7437139384705185,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0007923583050620244,
+      "loss": 1.1279,
+      "step": 8481
+    },
+    {
+      "epoch": 0.743801630244893,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0007920434996081456,
+      "loss": 1.1016,
+      "step": 8482
+    },
+    {
+      "epoch": 0.7438893220192675,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007917287723976844,
+      "loss": 1.1227,
+      "step": 8483
+    },
+    {
+      "epoch": 0.7439770137936421,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0007914141234593438,
+      "loss": 1.1154,
+      "step": 8484
+    },
+    {
+      "epoch": 0.7440647055680166,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0007910995528218192,
+      "loss": 1.1578,
+      "step": 8485
+    },
+    {
+      "epoch": 0.7441523973423911,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007907850605137974,
+      "loss": 1.1411,
+      "step": 8486
+    },
+    {
+      "epoch": 0.7442400891167658,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007904706465639604,
+      "loss": 1.1278,
+      "step": 8487
+    },
+    {
+      "epoch": 0.7443277808911403,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0007901563110009813,
+      "loss": 1.1096,
+      "step": 8488
+    },
+    {
+      "epoch": 0.7444154726655148,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0007898420538535271,
+      "loss": 1.0725,
+      "step": 8489
+    },
+    {
+      "epoch": 0.7445031644398894,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007895278751502575,
+      "loss": 1.1455,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7445908562142639,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000789213774919824,
+      "loss": 1.1188,
+      "step": 8491
+    },
+    {
+      "epoch": 0.7446785479886384,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007888997531908719,
+      "loss": 1.1308,
+      "step": 8492
+    },
+    {
+      "epoch": 0.7447662397630129,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0007885858099920397,
+      "loss": 1.1698,
+      "step": 8493
+    },
+    {
+      "epoch": 0.7448539315373875,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007882719453519581,
+      "loss": 1.1693,
+      "step": 8494
+    },
+    {
+      "epoch": 0.7449416233117621,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0007879581592992503,
+      "loss": 1.1556,
+      "step": 8495
+    },
+    {
+      "epoch": 0.7450293150861366,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.000787644451862533,
+      "loss": 1.103,
+      "step": 8496
+    },
+    {
+      "epoch": 0.7451170068605112,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007873308230704152,
+      "loss": 1.0997,
+      "step": 8497
+    },
+    {
+      "epoch": 0.7452046986348857,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0007870172729515003,
+      "loss": 1.1354,
+      "step": 8498
+    },
+    {
+      "epoch": 0.7452923904092602,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.000786703801534382,
+      "loss": 1.2128,
+      "step": 8499
+    },
+    {
+      "epoch": 0.7453800821836348,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000786390408847648,
+      "loss": 1.1248,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7453800821836348,
+      "eval_loss": 1.1438930034637451,
+      "eval_runtime": 429.1939,
+      "eval_samples_per_second": 33.661,
+      "eval_steps_per_second": 8.416,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7454677739580093,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007860770949198794,
+      "loss": 1.0943,
+      "step": 8501
+    },
+    {
+      "epoch": 0.7455554657323838,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0007857638597796496,
+      "loss": 1.2627,
+      "step": 8502
+    },
+    {
+      "epoch": 0.7456431575067585,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0007854507034555251,
+      "loss": 1.1442,
+      "step": 8503
+    },
+    {
+      "epoch": 0.745730849281133,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007851376259760642,
+      "loss": 1.1325,
+      "step": 8504
+    },
+    {
+      "epoch": 0.7458185410555075,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007848246273698191,
+      "loss": 1.1798,
+      "step": 8505
+    },
+    {
+      "epoch": 0.7459062328298821,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0007845117076653347,
+      "loss": 1.1366,
+      "step": 8506
+    },
+    {
+      "epoch": 0.7459939246042566,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0007841988668911487,
+      "loss": 1.0675,
+      "step": 8507
+    },
+    {
+      "epoch": 0.7460816163786311,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0007838861050757905,
+      "loss": 1.0947,
+      "step": 8508
+    },
+    {
+      "epoch": 0.7461693081530058,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0007835734222477838,
+      "loss": 1.1269,
+      "step": 8509
+    },
+    {
+      "epoch": 0.7462569999273803,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.000783260818435644,
+      "loss": 1.1438,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7463446917017548,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007829482936678805,
+      "loss": 1.1764,
+      "step": 8511
+    },
+    {
+      "epoch": 0.7464323834761294,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007826358479729945,
+      "loss": 1.0977,
+      "step": 8512
+    },
+    {
+      "epoch": 0.7465200752505039,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007823234813794796,
+      "loss": 1.12,
+      "step": 8513
+    },
+    {
+      "epoch": 0.7466077670248784,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007820111939158236,
+      "loss": 1.1032,
+      "step": 8514
+    },
+    {
+      "epoch": 0.7466954587992529,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0007816989856105061,
+      "loss": 1.1272,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7467831505736275,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0007813868564919994,
+      "loss": 1.1701,
+      "step": 8516
+    },
+    {
+      "epoch": 0.746870842348002,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0007810748065887694,
+      "loss": 1.0965,
+      "step": 8517
+    },
+    {
+      "epoch": 0.7469585341223766,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0007807628359292736,
+      "loss": 1.1156,
+      "step": 8518
+    },
+    {
+      "epoch": 0.7470462258967512,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0007804509445419631,
+      "loss": 1.0932,
+      "step": 8519
+    },
+    {
+      "epoch": 0.7471339176711257,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0007801391324552819,
+      "loss": 1.1866,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7472216094455002,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007798273996976664,
+      "loss": 1.0996,
+      "step": 8521
+    },
+    {
+      "epoch": 0.7473093012198748,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0007795157462975456,
+      "loss": 1.1634,
+      "step": 8522
+    },
+    {
+      "epoch": 0.7473969929942493,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0007792041722833415,
+      "loss": 1.0959,
+      "step": 8523
+    },
+    {
+      "epoch": 0.7474846847686238,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0007788926776834689,
+      "loss": 1.0954,
+      "step": 8524
+    },
+    {
+      "epoch": 0.7475723765429985,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0007785812625263358,
+      "loss": 1.0653,
+      "step": 8525
+    },
+    {
+      "epoch": 0.747660068317373,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007782699268403415,
+      "loss": 1.1369,
+      "step": 8526
+    },
+    {
+      "epoch": 0.7477477600917475,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0007779586706538797,
+      "loss": 1.1252,
+      "step": 8527
+    },
+    {
+      "epoch": 0.7478354518661221,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007776474939953359,
+      "loss": 1.0726,
+      "step": 8528
+    },
+    {
+      "epoch": 0.7479231436404966,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0007773363968930889,
+      "loss": 1.1477,
+      "step": 8529
+    },
+    {
+      "epoch": 0.7480108354148711,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007770253793755095,
+      "loss": 1.1386,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7480985271892457,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007767144414709619,
+      "loss": 1.1553,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7481862189636203,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007764035832078033,
+      "loss": 1.1119,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7482739107379948,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0007760928046143824,
+      "loss": 1.1262,
+      "step": 8533
+    },
+    {
+      "epoch": 0.7483616025123694,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0007757821057190418,
+      "loss": 1.2086,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7484492942867439,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0007754714865501163,
+      "loss": 1.2241,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7485369860611184,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0007751609471359337,
+      "loss": 1.0654,
+      "step": 8536
+    },
+    {
+      "epoch": 0.748624677835493,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0007748504875048141,
+      "loss": 1.1249,
+      "step": 8537
+    },
+    {
+      "epoch": 0.7487123696098675,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007745401076850715,
+      "loss": 1.1814,
+      "step": 8538
+    },
+    {
+      "epoch": 0.748800061384242,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0007742298077050105,
+      "loss": 1.0929,
+      "step": 8539
+    },
+    {
+      "epoch": 0.7488877531586166,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007739195875929303,
+      "loss": 1.1113,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7489754449329912,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000773609447377122,
+      "loss": 1.0623,
+      "step": 8541
+    },
+    {
+      "epoch": 0.7490631367073657,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007732993870858698,
+      "loss": 1.146,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7491508284817402,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007729894067474508,
+      "loss": 1.1399,
+      "step": 8543
+    },
+    {
+      "epoch": 0.7492385202561148,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0007726795063901334,
+      "loss": 1.1926,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7493262120304893,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007723696860421803,
+      "loss": 1.1333,
+      "step": 8545
+    },
+    {
+      "epoch": 0.7494139038048638,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000772059945731846,
+      "loss": 1.0736,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7495015955792385,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.000771750285487379,
+      "loss": 1.1819,
+      "step": 8547
+    },
+    {
+      "epoch": 0.749589287353613,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007714407053370184,
+      "loss": 1.1212,
+      "step": 8548
+    },
+    {
+      "epoch": 0.7496769791279875,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0007711312053089975,
+      "loss": 1.1662,
+      "step": 8549
+    },
+    {
+      "epoch": 0.7497646709023621,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0007708217854315418,
+      "loss": 1.0975,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7498523626767366,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0007705124457328697,
+      "loss": 1.1772,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7499400544511111,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007702031862411923,
+      "loss": 1.119,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7500277462254857,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000769894006984713,
+      "loss": 1.126,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7501154379998602,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007695849079916282,
+      "loss": 1.146,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7502031297742348,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0007692758892901269,
+      "loss": 1.1341,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7502908215486094,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0007689669509083916,
+      "loss": 1.1656,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7503785133229839,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007686580928745952,
+      "loss": 1.1114,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7504662050973584,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0007683493152169057,
+      "loss": 1.1849,
+      "step": 8558
+    },
+    {
+      "epoch": 0.750553896871733,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007680406179634828,
+      "loss": 1.1453,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7506415886461075,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007677320011424789,
+      "loss": 1.1095,
+      "step": 8560
+    },
+    {
+      "epoch": 0.750729280420482,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007674234647820389,
+      "loss": 1.1356,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7508169721948567,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007671150089103004,
+      "loss": 1.0642,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7509046639692312,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007668066335553938,
+      "loss": 1.1231,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7509923557436057,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007664983387454429,
+      "loss": 1.1691,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7510800475179802,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0007661901245085622,
+      "loss": 1.1368,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7511677392923548,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 0.0007658819908728612,
+      "loss": 1.1001,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7512554310667293,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0007655739378664396,
+      "loss": 1.1745,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7513431228411038,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0007652659655173922,
+      "loss": 1.147,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7514308146154784,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007649580738538042,
+      "loss": 1.1246,
+      "step": 8569
+    },
+    {
+      "epoch": 0.751518506389853,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000764650262903756,
+      "loss": 1.1035,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7516061981642275,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0007643425326953176,
+      "loss": 1.0869,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7516938899386021,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0007640348832565541,
+      "loss": 1.1414,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7517815817129766,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007637273146155221,
+      "loss": 1.1834,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7518692734873511,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0007634198268002715,
+      "loss": 1.223,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7519569652617257,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0007631124198388436,
+      "loss": 1.1055,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7520446570361002,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0007628050937592736,
+      "loss": 1.1015,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7521323488104747,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007624978485895887,
+      "loss": 1.0945,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7522200405848494,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007621906843578094,
+      "loss": 1.1515,
+      "step": 8578
+    },
+    {
+      "epoch": 0.7523077323592239,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0007618836010919472,
+      "loss": 1.0851,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7523954241335984,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0007615765988200079,
+      "loss": 1.0855,
+      "step": 8580
+    },
+    {
+      "epoch": 0.752483115907973,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0007612696775699896,
+      "loss": 1.1743,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7525708076823475,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.000760962837369883,
+      "loss": 1.1705,
+      "step": 8582
+    },
+    {
+      "epoch": 0.752658499456722,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0007606560782476702,
+      "loss": 1.1679,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7527461912310967,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007603494002313268,
+      "loss": 1.1942,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7528338830054712,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007600428033488219,
+      "loss": 1.1375,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7529215747798457,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0007597362876281157,
+      "loss": 1.0941,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7530092665542202,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007594298530971621,
+      "loss": 1.1157,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7530969583285948,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0007591234997839066,
+      "loss": 1.0968,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7531846501029693,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0007588172277162883,
+      "loss": 1.0576,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7532723418773438,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0007585110369222381,
+      "loss": 1.1316,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7533600336517184,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0007582049274296804,
+      "loss": 1.1364,
+      "step": 8591
+    },
+    {
+      "epoch": 0.753447725426093,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0007578988992665307,
+      "loss": 1.1809,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7535354172004675,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0007575929524606984,
+      "loss": 1.0927,
+      "step": 8593
+    },
+    {
+      "epoch": 0.7536231089748421,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007572870870400853,
+      "loss": 1.0951,
+      "step": 8594
+    },
+    {
+      "epoch": 0.7537108007492166,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0007569813030325851,
+      "loss": 1.1027,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7537984925235911,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007566756004660853,
+      "loss": 1.1168,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7538861842979657,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0007563699793684643,
+      "loss": 1.085,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7539738760723402,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007560644397675943,
+      "loss": 1.1056,
+      "step": 8598
+    },
+    {
+      "epoch": 0.7540615678467147,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00075575898169134,
+      "loss": 1.1167,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7541492596210894,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0007554536051675578,
+      "loss": 1.1951,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7542369513954639,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0007551483102240978,
+      "loss": 1.136,
+      "step": 8601
+    },
+    {
+      "epoch": 0.7543246431698384,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0007548430968888015,
+      "loss": 1.0868,
+      "step": 8602
+    },
+    {
+      "epoch": 0.754412334944213,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007545379651895038,
+      "loss": 1.1547,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7545000267185875,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0007542329151540323,
+      "loss": 1.1059,
+      "step": 8604
+    },
+    {
+      "epoch": 0.754587718492962,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007539279468102069,
+      "loss": 1.1105,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7546754102673366,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007536230601858392,
+      "loss": 1.2443,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7547631020417112,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000753318255308734,
+      "loss": 1.1553,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7548507938160857,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0007530135322066896,
+      "loss": 1.1524,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7549384855904603,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007527088909074958,
+      "loss": 1.1766,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7550261773648348,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007524043314389343,
+      "loss": 1.1565,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7551138691392093,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0007520998538287807,
+      "loss": 1.1142,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7552015609135838,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0007517954581048026,
+      "loss": 1.0841,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7552892526879584,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0007514911442947602,
+      "loss": 1.1583,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7553769444623329,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0007511869124264057,
+      "loss": 1.0915,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7554646362367075,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0007508827625274845,
+      "loss": 1.1135,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7555523280110821,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007505786946257348,
+      "loss": 1.1438,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7556400197854566,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007502747087488856,
+      "loss": 1.1148,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7557277115598311,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0007499708049246611,
+      "loss": 1.1461,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7558154033342057,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007496669831807751,
+      "loss": 1.1236,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7559030951085802,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0007493632435449362,
+      "loss": 1.1408,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7559907868829547,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007490595860448444,
+      "loss": 1.1919,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7560784786573294,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007487560107081926,
+      "loss": 1.1325,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7561661704317039,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007484525175626666,
+      "loss": 1.1023,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7562538622060784,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007481491066359432,
+      "loss": 1.1197,
+      "step": 8624
+    },
+    {
+      "epoch": 0.756341553980453,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.000747845777955693,
+      "loss": 1.1252,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7564292457548275,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007475425315495791,
+      "loss": 1.057,
+      "step": 8626
+    },
+    {
+      "epoch": 0.756516937529202,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0007472393674452572,
+      "loss": 1.1083,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7566046293035766,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007469362856703739,
+      "loss": 1.158,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7566923210779511,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0007466332862525701,
+      "loss": 1.1221,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7567800128523257,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0007463303692194786,
+      "loss": 1.1119,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7568677046267003,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007460275345987251,
+      "loss": 1.166,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7569553964010748,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007457247824179264,
+      "loss": 1.1657,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7570430881754493,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0007454221127046934,
+      "loss": 1.0879,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7571307799498239,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007451195254866285,
+      "loss": 1.1409,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7572184717241984,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007448170207913267,
+      "loss": 1.0777,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7573061634985729,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007445145986463763,
+      "loss": 1.1177,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7573938552729474,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0007442122590793567,
+      "loss": 1.1204,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7574815470473221,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0007439100021178407,
+      "loss": 1.1544,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7575692388216966,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007436078277893935,
+      "loss": 1.1013,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7576569305960711,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0007433057361215731,
+      "loss": 1.0742,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7577446223704457,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0007430037271419282,
+      "loss": 1.1808,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7578323141448202,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0007427018008780022,
+      "loss": 1.1043,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7579200059191947,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0007423999573573298,
+      "loss": 1.1825,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7580076976935693,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007420981966074388,
+      "loss": 1.1525,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7580953894679439,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.000741796518655848,
+      "loss": 1.0803,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7581830812423184,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007414949235300701,
+      "loss": 1.0897,
+      "step": 8646
+    },
+    {
+      "epoch": 0.758270773016693,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0007411934112576101,
+      "loss": 1.1099,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7583584647910675,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0007408919818659648,
+      "loss": 1.1116,
+      "step": 8648
+    },
+    {
+      "epoch": 0.758446156565442,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007405906353826243,
+      "loss": 1.1428,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7585338483398166,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0007402893718350703,
+      "loss": 1.1617,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7586215401141911,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0007399881912507767,
+      "loss": 1.0983,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7587092318885656,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007396870936572109,
+      "loss": 1.1031,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7587969236629403,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0007393860790818322,
+      "loss": 1.1795,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7588846154373148,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0007390851475520927,
+      "loss": 1.1228,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7589723072116893,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000738784299095436,
+      "loss": 1.1171,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7590599989860639,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007384835337392987,
+      "loss": 1.1796,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7591476907604384,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0007381828515111102,
+      "loss": 1.1423,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7592353825348129,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007378822524382922,
+      "loss": 1.1962,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7593230743091874,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007375817365482577,
+      "loss": 1.1874,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7594107660835621,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007372813038684134,
+      "loss": 1.185,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7594984578579366,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.000736980954426158,
+      "loss": 1.1391,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7595861496323111,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0007366806882488832,
+      "loss": 1.0968,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7596738414066857,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0007363805053639714,
+      "loss": 1.0923,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7597615331810602,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.000736080405798799,
+      "loss": 1.165,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7598492249554347,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0007357803895807343,
+      "loss": 1.1829,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7599369167298093,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0007354804567371385,
+      "loss": 1.1658,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7600246085041839,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0007351806072953643,
+      "loss": 1.1244,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7601123002785584,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0007348808412827565,
+      "loss": 1.0674,
+      "step": 8668
+    },
+    {
+      "epoch": 0.760199992052933,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007345811587266537,
+      "loss": 1.1089,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7602876838273075,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000734281559654386,
+      "loss": 1.1791,
+      "step": 8670
+    },
+    {
+      "epoch": 0.760375375601682,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007339820440932767,
+      "loss": 1.1221,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7604630673760566,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0007336826120706398,
+      "loss": 1.1652,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7605507591504311,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007333832636137832,
+      "loss": 1.1084,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7606384509248056,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007330839987500069,
+      "loss": 1.1681,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7607261426991803,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0007327848175066026,
+      "loss": 1.1883,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7608138344735548,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0007324857199108558,
+      "loss": 1.1365,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7609015262479293,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0007321867059900424,
+      "loss": 1.175,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7609892180223039,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.000731887775771432,
+      "loss": 1.1831,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7610769097966784,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0007315889292822862,
+      "loss": 1.0899,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7611646015710529,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0007312901665498598,
+      "loss": 1.1483,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7612522933454275,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0007309914876013982,
+      "loss": 1.1419,
+      "step": 8681
+    },
+    {
+      "epoch": 0.761339985119802,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0007306928924641404,
+      "loss": 1.0812,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7614276768941766,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007303943811653176,
+      "loss": 1.1408,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7615153686685511,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000730095953732154,
+      "loss": 1.1484,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7616030604429257,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007297976101918644,
+      "loss": 1.103,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7616907522173002,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007294993505716569,
+      "loss": 1.1436,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7617784439916747,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007292011748987324,
+      "loss": 1.12,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7618661357660493,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0007289030832002837,
+      "loss": 1.2152,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7619538275404238,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0007286050755034965,
+      "loss": 1.1363,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7620415193147984,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007283071518355473,
+      "loss": 1.0707,
+      "step": 8690
+    },
+    {
+      "epoch": 0.762129211089173,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007280093122236065,
+      "loss": 1.147,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7622169028635475,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0007277115566948364,
+      "loss": 1.165,
+      "step": 8692
+    },
+    {
+      "epoch": 0.762304594637922,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0007274138852763919,
+      "loss": 1.0908,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7623922864122966,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000727116297995419,
+      "loss": 1.1214,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7624799781866711,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0007268187948790573,
+      "loss": 1.1461,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7625676699610456,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0007265213759544382,
+      "loss": 1.1707,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7626553617354203,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0007262240412486862,
+      "loss": 1.1722,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7627430535097948,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007259267907889167,
+      "loss": 1.1185,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7628307452841693,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000725629624602238,
+      "loss": 1.1577,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7629184370585439,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0007253325427157514,
+      "loss": 1.1558,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7630061288329184,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0007250355451565506,
+      "loss": 1.1316,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7630938206072929,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007247386319517201,
+      "loss": 1.0985,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7631815123816675,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0007244418031283373,
+      "loss": 1.1352,
+      "step": 8703
+    },
+    {
+      "epoch": 0.763269204156042,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0007241450587134728,
+      "loss": 1.1275,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7633568959304166,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0007238483987341889,
+      "loss": 1.0956,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7634445877047912,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007235518232175402,
+      "loss": 1.1325,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7635322794791657,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000723255332190574,
+      "loss": 1.1158,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7636199712535402,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0007229589256803287,
+      "loss": 1.1521,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7637076630279147,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0007226626037138363,
+      "loss": 1.1167,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7637953548022893,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0007223663663181205,
+      "loss": 1.1655,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7638830465766638,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007220702135201978,
+      "loss": 1.1254,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7639707383510383,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007217741453470757,
+      "loss": 1.121,
+      "step": 8712
+    },
+    {
+      "epoch": 0.764058430125413,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0007214781618257555,
+      "loss": 1.158,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7641461218997875,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007211822629832298,
+      "loss": 1.1561,
+      "step": 8714
+    },
+    {
+      "epoch": 0.764233813674162,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0007208864488464846,
+      "loss": 1.1309,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7643215054485366,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0007205907194424961,
+      "loss": 1.1532,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7644091972229111,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007202950747982349,
+      "loss": 1.1659,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7644968889972856,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007199995149406633,
+      "loss": 1.2157,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7645845807716602,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007197040398967347,
+      "loss": 1.1732,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7646722725460348,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0007194086496933966,
+      "loss": 1.0747,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7647599643204093,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0007191133443575868,
+      "loss": 1.1001,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7648476560947839,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000718818123916237,
+      "loss": 1.1789,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7649353478691584,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007185229883962705,
+      "loss": 1.1394,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7650230396435329,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000718227937824603,
+      "loss": 1.1502,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7651107314179075,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000717932972228142,
+      "loss": 1.0985,
+      "step": 8725
+    },
+    {
+      "epoch": 0.765198423192282,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007176380916337878,
+      "loss": 1.1711,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7652861149666565,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.000717343296068433,
+      "loss": 1.1504,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7653738067410312,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007170485855589616,
+      "loss": 1.093,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7654614985154057,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007167539601322514,
+      "loss": 1.0963,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7655491902897802,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0007164594198151704,
+      "loss": 1.194,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7656368820641547,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0007161649646345804,
+      "loss": 1.0734,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7657245738385293,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.000715870594617335,
+      "loss": 1.0952,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7658122656129038,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007155763097902803,
+      "loss": 1.1707,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7658999573872783,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007152821101802534,
+      "loss": 1.1272,
+      "step": 8734
+    },
+    {
+      "epoch": 0.765987649161653,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007149879958140857,
+      "loss": 1.1936,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7660753409360275,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007146939667185986,
+      "loss": 1.1045,
+      "step": 8736
+    },
+    {
+      "epoch": 0.766163032710402,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007144000229206072,
+      "loss": 1.1193,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7662507244847766,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.000714106164446919,
+      "loss": 1.1672,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7663384162591511,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000713812391324332,
+      "loss": 1.1292,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7664261080335256,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0007135187035796382,
+      "loss": 1.0682,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7665137998079002,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007132251012396212,
+      "loss": 1.1377,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7666014915822748,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.000712931584331057,
+      "loss": 1.1428,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7666891833566493,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0007126381528807131,
+      "loss": 1.0837,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7667768751310239,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007123448069153497,
+      "loss": 1.1044,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7668645669053984,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0007120515464617194,
+      "loss": 1.1089,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7669522586797729,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007117583715465672,
+      "loss": 1.1258,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7670399504541475,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0007114652821966289,
+      "loss": 1.1643,
+      "step": 8747
+    },
+    {
+      "epoch": 0.767127642228522,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0007111722784386342,
+      "loss": 1.1111,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7672153340028965,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.000710879360299304,
+      "loss": 1.1626,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7673030257772712,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0007105865278053524,
+      "loss": 1.1738,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7673907175516457,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0007102937809834839,
+      "loss": 1.0734,
+      "step": 8751
+    },
+    {
+      "epoch": 0.7674784093260202,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0007100011198603971,
+      "loss": 1.1216,
+      "step": 8752
+    },
+    {
+      "epoch": 0.7675661011003948,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007097085444627812,
+      "loss": 1.1679,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7676537928747693,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0007094160548173187,
+      "loss": 1.1109,
+      "step": 8754
+    },
+    {
+      "epoch": 0.7677414846491438,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0007091236509506843,
+      "loss": 1.1382,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7678291764235183,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0007088313328895436,
+      "loss": 1.1719,
+      "step": 8756
+    },
+    {
+      "epoch": 0.767916868197893,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0007085391006605557,
+      "loss": 1.1093,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7680045599722675,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0007082469542903716,
+      "loss": 1.1367,
+      "step": 8758
+    },
+    {
+      "epoch": 0.768092251746642,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000707954893805634,
+      "loss": 1.095,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7681799435210166,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0007076629192329788,
+      "loss": 1.1202,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7682676352953911,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0007073710305990322,
+      "loss": 1.1109,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7683553270697656,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0007070792279304142,
+      "loss": 1.1824,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7684430188441402,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0007067875112537366,
+      "loss": 1.1529,
+      "step": 8763
+    },
+    {
+      "epoch": 0.7685307106185147,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007064958805956035,
+      "loss": 1.0925,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7686184023928893,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007062043359826098,
+      "loss": 1.1174,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7687060941672639,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007059128774413444,
+      "loss": 1.1572,
+      "step": 8766
+    },
+    {
+      "epoch": 0.7687937859416384,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0007056215049983875,
+      "loss": 1.1326,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7688814777160129,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0007053302186803117,
+      "loss": 1.1491,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7689691694903875,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007050390185136814,
+      "loss": 1.1296,
+      "step": 8769
+    },
+    {
+      "epoch": 0.769056861264762,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000704747904525053,
+      "loss": 1.1264,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7691445530391365,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007044568767409753,
+      "loss": 1.0936,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7692322448135112,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0007041659351879894,
+      "loss": 1.1458,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7693199365878857,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0007038750798926293,
+      "loss": 1.1524,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7694076283622602,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007035843108814189,
+      "loss": 1.0978,
+      "step": 8774
+    },
+    {
+      "epoch": 0.7694953201366348,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007032936281808761,
+      "loss": 1.1188,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7695830119110093,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0007030030318175109,
+      "loss": 1.0903,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7696707036853838,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0007027125218178244,
+      "loss": 1.2298,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7697583954597583,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007024220982083104,
+      "loss": 1.1756,
+      "step": 8778
+    },
+    {
+      "epoch": 0.769846087234133,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0007021317610154549,
+      "loss": 1.082,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7699337790085075,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007018415102657358,
+      "loss": 1.1471,
+      "step": 8780
+    },
+    {
+      "epoch": 0.770021470782882,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0007015513459856239,
+      "loss": 1.1169,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7701091625572566,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0007012612682015802,
+      "loss": 1.0852,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7701968543316311,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0007009712769400597,
+      "loss": 1.105,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7702845461060056,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007006813722275089,
+      "loss": 1.1304,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7703722378803802,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0007003915540903663,
+      "loss": 1.1833,
+      "step": 8785
+    },
+    {
+      "epoch": 0.7704599296547547,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0007001018225550631,
+      "loss": 1.1155,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7705476214291292,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0006998121776480213,
+      "loss": 1.1064,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7706353132035039,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006995226193956558,
+      "loss": 1.1066,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7707230049778784,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006992331478243736,
+      "loss": 1.0869,
+      "step": 8789
+    },
+    {
+      "epoch": 0.7708106967522529,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0006989437629605739,
+      "loss": 1.1609,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7708983885266275,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0006986544648306485,
+      "loss": 1.1664,
+      "step": 8791
+    },
+    {
+      "epoch": 0.770986080301002,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0006983652534609795,
+      "loss": 1.1653,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7710737720753765,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006980761288779427,
+      "loss": 1.0771,
+      "step": 8793
+    },
+    {
+      "epoch": 0.7711614638497511,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0006977870911079055,
+      "loss": 1.1566,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7712491556241257,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006974981401772282,
+      "loss": 1.1304,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7713368473985002,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006972092761122609,
+      "loss": 1.1605,
+      "step": 8796
+    },
+    {
+      "epoch": 0.7714245391728748,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0006969204989393481,
+      "loss": 1.1734,
+      "step": 8797
+    },
+    {
+      "epoch": 0.7715122309472493,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0006966318086848254,
+      "loss": 1.1063,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7715999227216238,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006963432053750212,
+      "loss": 1.1034,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7716876144959984,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006960546890362544,
+      "loss": 1.1835,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7717753062703729,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0006957662596948371,
+      "loss": 1.1541,
+      "step": 8801
+    },
+    {
+      "epoch": 0.7718629980447475,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006954779173770738,
+      "loss": 1.1353,
+      "step": 8802
+    },
+    {
+      "epoch": 0.771950689819122,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006951896621092607,
+      "loss": 1.0771,
+      "step": 8803
+    },
+    {
+      "epoch": 0.7720383815934966,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006949014939176858,
+      "loss": 1.1346,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7721260733678711,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0006946134128286283,
+      "loss": 1.0905,
+      "step": 8805
+    },
+    {
+      "epoch": 0.7722137651422456,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0006943254188683614,
+      "loss": 1.2018,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7723014569166202,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0006940375120631494,
+      "loss": 1.1055,
+      "step": 8807
+    },
+    {
+      "epoch": 0.7723891486909947,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0006937496924392486,
+      "loss": 1.1228,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7724768404653692,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006934619600229071,
+      "loss": 1.1385,
+      "step": 8809
+    },
+    {
+      "epoch": 0.7725645322397439,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0006931743148403655,
+      "loss": 1.2018,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7726522240141184,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006928867569178563,
+      "loss": 1.1125,
+      "step": 8811
+    },
+    {
+      "epoch": 0.7727399157884929,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0006925992862816041,
+      "loss": 1.0908,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7728276075628675,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000692311902957826,
+      "loss": 1.1161,
+      "step": 8813
+    },
+    {
+      "epoch": 0.772915299337242,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0006920246069727296,
+      "loss": 1.1903,
+      "step": 8814
+    },
+    {
+      "epoch": 0.7730029911116165,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.000691737398352516,
+      "loss": 1.1246,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7730906828859911,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0006914502771233779,
+      "loss": 1.1663,
+      "step": 8816
+    },
+    {
+      "epoch": 0.7731783746603657,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0006911632433115006,
+      "loss": 1.1501,
+      "step": 8817
+    },
+    {
+      "epoch": 0.7732660664347402,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0006908762969430599,
+      "loss": 1.1549,
+      "step": 8818
+    },
+    {
+      "epoch": 0.7733537582091148,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006905894380442247,
+      "loss": 1.1227,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7734414499834893,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006903026666411566,
+      "loss": 1.1738,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7735291417578638,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0006900159827600075,
+      "loss": 1.154,
+      "step": 8821
+    },
+    {
+      "epoch": 0.7736168335322384,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000689729386426923,
+      "loss": 1.0917,
+      "step": 8822
+    },
+    {
+      "epoch": 0.7737045253066129,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000689442877668039,
+      "loss": 1.115,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7737922170809874,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006891564565094851,
+      "loss": 1.1118,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7738799088553621,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006888701229773818,
+      "loss": 1.1249,
+      "step": 8825
+    },
+    {
+      "epoch": 0.7739676006297366,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0006885838770978425,
+      "loss": 1.0973,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7740552924041111,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006882977188969714,
+      "loss": 1.1404,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7741429841784856,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006880116484008658,
+      "loss": 1.1479,
+      "step": 8828
+    },
+    {
+      "epoch": 0.7742306759528602,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0006877256656356144,
+      "loss": 1.1088,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7743183677272347,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006874397706272987,
+      "loss": 1.0518,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7744060595016092,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0006871539634019905,
+      "loss": 1.1634,
+      "step": 8831
+    },
+    {
+      "epoch": 0.7744937512759839,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006868682439857553,
+      "loss": 1.1367,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7745814430503584,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00068658261240465,
+      "loss": 1.0738,
+      "step": 8833
+    },
+    {
+      "epoch": 0.7746691348247329,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006862970686847238,
+      "loss": 1.1291,
+      "step": 8834
+    },
+    {
+      "epoch": 0.7747568265991075,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006860116128520168,
+      "loss": 1.1778,
+      "step": 8835
+    },
+    {
+      "epoch": 0.774844518373482,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006857262449325622,
+      "loss": 1.1855,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7749322101478565,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006854409649523852,
+      "loss": 1.1499,
+      "step": 8837
+    },
+    {
+      "epoch": 0.7750199019222311,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0006851557729375016,
+      "loss": 1.1886,
+      "step": 8838
+    },
+    {
+      "epoch": 0.7751075936966056,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.000684870668913921,
+      "loss": 1.0963,
+      "step": 8839
+    },
+    {
+      "epoch": 0.7751952854709802,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006845856529076441,
+      "loss": 1.0804,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7752829772453548,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006843007249446632,
+      "loss": 1.1898,
+      "step": 8841
+    },
+    {
+      "epoch": 0.7753706690197293,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0006840158850509633,
+      "loss": 1.1674,
+      "step": 8842
+    },
+    {
+      "epoch": 0.7754583607941038,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006837311332525209,
+      "loss": 1.1466,
+      "step": 8843
+    },
+    {
+      "epoch": 0.7755460525684784,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0006834464695753051,
+      "loss": 1.0985,
+      "step": 8844
+    },
+    {
+      "epoch": 0.7756337443428529,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006831618940452759,
+      "loss": 1.1517,
+      "step": 8845
+    },
+    {
+      "epoch": 0.7757214361172274,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000682877406688386,
+      "loss": 1.1631,
+      "step": 8846
+    },
+    {
+      "epoch": 0.7758091278916021,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00068259300753058,
+      "loss": 1.1483,
+      "step": 8847
+    },
+    {
+      "epoch": 0.7758968196659766,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006823086965977947,
+      "loss": 1.126,
+      "step": 8848
+    },
+    {
+      "epoch": 0.7759845114403511,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006820244739159577,
+      "loss": 1.1596,
+      "step": 8849
+    },
+    {
+      "epoch": 0.7760722032147256,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006817403395109901,
+      "loss": 1.1319,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7761598949891002,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0006814562934088039,
+      "loss": 1.1175,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7762475867634747,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0006811723356353039,
+      "loss": 1.1591,
+      "step": 8852
+    },
+    {
+      "epoch": 0.7763352785378492,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006808884662163853,
+      "loss": 1.1414,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7764229703122238,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006806046851779374,
+      "loss": 1.1277,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7765106620865984,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006803209925458393,
+      "loss": 1.119,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7765983538609729,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0006800373883459634,
+      "loss": 1.1687,
+      "step": 8856
+    },
+    {
+      "epoch": 0.7766860456353475,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006797538726041743,
+      "loss": 1.1359,
+      "step": 8857
+    },
+    {
+      "epoch": 0.776773737409722,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006794704453463266,
+      "loss": 1.0774,
+      "step": 8858
+    },
+    {
+      "epoch": 0.7768614291840965,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0006791871065982692,
+      "loss": 1.1939,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7769491209584711,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006789038563858415,
+      "loss": 1.1459,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7770368127328456,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0006786206947348756,
+      "loss": 1.099,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7771245045072201,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0006783376216711943,
+      "loss": 1.1047,
+      "step": 8862
+    },
+    {
+      "epoch": 0.7772121962815948,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0006780546372206135,
+      "loss": 1.1139,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7772998880559693,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0006777717414089407,
+      "loss": 1.0861,
+      "step": 8864
+    },
+    {
+      "epoch": 0.7773875798303438,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006774889342619756,
+      "loss": 1.1896,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7774752716047184,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006772062158055093,
+      "loss": 1.1577,
+      "step": 8866
+    },
+    {
+      "epoch": 0.7775629633790929,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006769235860653245,
+      "loss": 1.1313,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7776506551534674,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006766410450671967,
+      "loss": 1.1603,
+      "step": 8868
+    },
+    {
+      "epoch": 0.777738346927842,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006763585928368928,
+      "loss": 1.1262,
+      "step": 8869
+    },
+    {
+      "epoch": 0.7778260387022166,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006760762294001722,
+      "loss": 1.1301,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7779137304765911,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0006757939547827853,
+      "loss": 1.0957,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7780014222509657,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0006755117690104746,
+      "loss": 1.0357,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7780891140253402,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0006752296721089749,
+      "loss": 1.1092,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7781768057997147,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006749476641040125,
+      "loss": 1.1159,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7782644975740892,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006746657450213067,
+      "loss": 1.1468,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7783521893484638,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0006743839148865666,
+      "loss": 1.1189,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7784398811228384,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.000674102173725495,
+      "loss": 1.1277,
+      "step": 8877
+    },
+    {
+      "epoch": 0.7785275728972129,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006738205215637859,
+      "loss": 1.0777,
+      "step": 8878
+    },
+    {
+      "epoch": 0.7786152646715875,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006735389584271256,
+      "loss": 1.16,
+      "step": 8879
+    },
+    {
+      "epoch": 0.778702956445962,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0006732574843411912,
+      "loss": 1.1116,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7787906482203365,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006729760993316529,
+      "loss": 1.1335,
+      "step": 8881
+    },
+    {
+      "epoch": 0.7788783399947111,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006726948034241719,
+      "loss": 1.1627,
+      "step": 8882
+    },
+    {
+      "epoch": 0.7789660317690856,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0006724135966444025,
+      "loss": 1.1483,
+      "step": 8883
+    },
+    {
+      "epoch": 0.7790537235434601,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000672132479017989,
+      "loss": 1.1891,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7791414153178348,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006718514505705689,
+      "loss": 1.1026,
+      "step": 8885
+    },
+    {
+      "epoch": 0.7792291070922093,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006715705113277717,
+      "loss": 1.0936,
+      "step": 8886
+    },
+    {
+      "epoch": 0.7793167988665838,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006712896613152185,
+      "loss": 1.1819,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7794044906409584,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0006710089005585215,
+      "loss": 1.1403,
+      "step": 8888
+    },
+    {
+      "epoch": 0.7794921824153329,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0006707282290832851,
+      "loss": 1.1289,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7795798741897074,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006704476469151061,
+      "loss": 1.1788,
+      "step": 8890
+    },
+    {
+      "epoch": 0.779667565964082,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006701671540795733,
+      "loss": 1.1299,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7797552577384566,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006698867506022666,
+      "loss": 1.1352,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7798429495128311,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006696064365087584,
+      "loss": 1.1155,
+      "step": 8893
+    },
+    {
+      "epoch": 0.7799306412872057,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006693262118246119,
+      "loss": 1.1214,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7800183330615802,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0006690460765753833,
+      "loss": 1.1451,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7801060248359547,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006687660307866201,
+      "loss": 1.1226,
+      "step": 8896
+    },
+    {
+      "epoch": 0.7801937166103293,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006684860744838624,
+      "loss": 1.1296,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7802814083847038,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0006682062076926405,
+      "loss": 1.1234,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7803691001590783,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006679264304384779,
+      "loss": 1.1472,
+      "step": 8899
+    },
+    {
+      "epoch": 0.7804567919334529,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0006676467427468896,
+      "loss": 1.018,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7805444837078275,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000667367144643383,
+      "loss": 1.1223,
+      "step": 8901
+    },
+    {
+      "epoch": 0.780632175482202,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0006670876361534558,
+      "loss": 1.1032,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7807198672565765,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0006668082173025985,
+      "loss": 1.1337,
+      "step": 8903
+    },
+    {
+      "epoch": 0.7808075590309511,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006665288881162945,
+      "loss": 1.1529,
+      "step": 8904
+    },
+    {
+      "epoch": 0.7808952508053256,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0006662496486200163,
+      "loss": 1.1562,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7809829425797001,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006659704988392312,
+      "loss": 1.0984,
+      "step": 8906
+    },
+    {
+      "epoch": 0.7810706343540748,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.000665691438799396,
+      "loss": 1.0548,
+      "step": 8907
+    },
+    {
+      "epoch": 0.7811583261284493,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006654124685259604,
+      "loss": 1.1327,
+      "step": 8908
+    },
+    {
+      "epoch": 0.7812460179028238,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0006651335880443661,
+      "loss": 1.1161,
+      "step": 8909
+    },
+    {
+      "epoch": 0.7813337096771984,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006648547973800464,
+      "loss": 1.1093,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7814214014515729,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006645760965584258,
+      "loss": 1.0665,
+      "step": 8911
+    },
+    {
+      "epoch": 0.7815090932259474,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0006642974856049214,
+      "loss": 1.1322,
+      "step": 8912
+    },
+    {
+      "epoch": 0.781596785000322,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006640189645449417,
+      "loss": 1.166,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7816844767746965,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0006637405334038874,
+      "loss": 1.0959,
+      "step": 8914
+    },
+    {
+      "epoch": 0.7817721685490711,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0006634621922071501,
+      "loss": 1.1888,
+      "step": 8915
+    },
+    {
+      "epoch": 0.7818598603234457,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0006631839409801142,
+      "loss": 1.209,
+      "step": 8916
+    },
+    {
+      "epoch": 0.7819475520978202,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0006629057797481553,
+      "loss": 1.1704,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7820352438721947,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0006626277085366413,
+      "loss": 1.0844,
+      "step": 8918
+    },
+    {
+      "epoch": 0.7821229356465693,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006623497273709317,
+      "loss": 1.163,
+      "step": 8919
+    },
+    {
+      "epoch": 0.7822106274209438,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.000662071836276377,
+      "loss": 1.2092,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7822983191953183,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0006617940352783207,
+      "loss": 1.2079,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7823860109696928,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0006615163244020972,
+      "loss": 1.1588,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7824737027440675,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000661238703673033,
+      "loss": 1.0975,
+      "step": 8923
+    },
+    {
+      "epoch": 0.782561394518442,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000660961173116447,
+      "loss": 1.1209,
+      "step": 8924
+    },
+    {
+      "epoch": 0.7826490862928165,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006606837327576484,
+      "loss": 1.0952,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7827367780671911,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006604063826219395,
+      "loss": 1.1019,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7828244698415656,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000660129122734614,
+      "loss": 1.0934,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7829121616159401,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006598519531209573,
+      "loss": 1.1145,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7829998533903147,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006595748738062462,
+      "loss": 1.1256,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7830875451646893,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006592978848157497,
+      "loss": 1.1401,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7831752369390638,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006590209861747288,
+      "loss": 1.1675,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7832629287134384,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006587441779084362,
+      "loss": 1.1613,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7833506204878129,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.000658467460042115,
+      "loss": 1.1007,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7834383122621874,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.000658190832601002,
+      "loss": 1.129,
+      "step": 8934
+    },
+    {
+      "epoch": 0.783526004036562,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006579142956103247,
+      "loss": 1.1367,
+      "step": 8935
+    },
+    {
+      "epoch": 0.7836136958109365,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006576378490953029,
+      "loss": 1.1572,
+      "step": 8936
+    },
+    {
+      "epoch": 0.783701387585311,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006573614930811473,
+      "loss": 1.12,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7837890793596857,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006570852275930613,
+      "loss": 1.1414,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7838767711340602,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0006568090526562389,
+      "loss": 1.1604,
+      "step": 8939
+    },
+    {
+      "epoch": 0.7839644629084347,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006565329682958669,
+      "loss": 1.2276,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7840521546828093,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006562569745371243,
+      "loss": 1.1197,
+      "step": 8941
+    },
+    {
+      "epoch": 0.7841398464571838,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006559810714051797,
+      "loss": 1.1756,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7842275382315583,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0006557052589251952,
+      "loss": 1.1851,
+      "step": 8943
+    },
+    {
+      "epoch": 0.784315230005933,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006554295371223246,
+      "loss": 1.1202,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7844029217803075,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006551539060217125,
+      "loss": 1.1427,
+      "step": 8945
+    },
+    {
+      "epoch": 0.784490613554682,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006548783656484966,
+      "loss": 1.116,
+      "step": 8946
+    },
+    {
+      "epoch": 0.7845783053290565,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006546029160278043,
+      "loss": 1.1071,
+      "step": 8947
+    },
+    {
+      "epoch": 0.7846659971034311,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0006543275571847567,
+      "loss": 1.1113,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7847536888778056,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006540522891444652,
+      "loss": 1.0965,
+      "step": 8949
+    },
+    {
+      "epoch": 0.7848413806521801,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0006537771119320345,
+      "loss": 1.1677,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7849290724265547,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006535020255725592,
+      "loss": 1.1401,
+      "step": 8951
+    },
+    {
+      "epoch": 0.7850167642009293,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0006532270300911266,
+      "loss": 1.1321,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7851044559753038,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006529521255128155,
+      "loss": 1.1679,
+      "step": 8953
+    },
+    {
+      "epoch": 0.7851921477496784,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006526773118626974,
+      "loss": 1.07,
+      "step": 8954
+    },
+    {
+      "epoch": 0.7852798395240529,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006524025891658337,
+      "loss": 1.1298,
+      "step": 8955
+    },
+    {
+      "epoch": 0.7853675312984274,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006521279574472782,
+      "loss": 1.1264,
+      "step": 8956
+    },
+    {
+      "epoch": 0.785455223072802,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0006518534167320769,
+      "loss": 1.1025,
+      "step": 8957
+    },
+    {
+      "epoch": 0.7855429148471765,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0006515789670452675,
+      "loss": 1.1231,
+      "step": 8958
+    },
+    {
+      "epoch": 0.785630606621551,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006513046084118792,
+      "loss": 1.1028,
+      "step": 8959
+    },
+    {
+      "epoch": 0.7857182983959257,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006510303408569321,
+      "loss": 1.1045,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7858059901703002,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0006507561644054391,
+      "loss": 1.1486,
+      "step": 8961
+    },
+    {
+      "epoch": 0.7858936819446747,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0006504820790824042,
+      "loss": 1.1269,
+      "step": 8962
+    },
+    {
+      "epoch": 0.7859813737190493,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.000650208084912824,
+      "loss": 1.1354,
+      "step": 8963
+    },
+    {
+      "epoch": 0.7860690654934238,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006499341819216853,
+      "loss": 1.1562,
+      "step": 8964
+    },
+    {
+      "epoch": 0.7861567572677983,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006496603701339672,
+      "loss": 1.1993,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7862444490421729,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006493866495746412,
+      "loss": 1.0762,
+      "step": 8966
+    },
+    {
+      "epoch": 0.7863321408165475,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00064911302026867,
+      "loss": 1.1173,
+      "step": 8967
+    },
+    {
+      "epoch": 0.786419832590922,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006488394822410072,
+      "loss": 1.1532,
+      "step": 8968
+    },
+    {
+      "epoch": 0.7865075243652965,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.000648566035516599,
+      "loss": 1.0801,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7865952161396711,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006482926801203835,
+      "loss": 1.0817,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7866829079140456,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006480194160772894,
+      "loss": 1.1365,
+      "step": 8971
+    },
+    {
+      "epoch": 0.7867705996884201,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0006477462434122389,
+      "loss": 1.1035,
+      "step": 8972
+    },
+    {
+      "epoch": 0.7868582914627947,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0006474731621501429,
+      "loss": 1.0941,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7869459832371692,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0006472001723159063,
+      "loss": 1.0893,
+      "step": 8974
+    },
+    {
+      "epoch": 0.7870336750115438,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006469272739344254,
+      "loss": 1.1202,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7871213667859184,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006466544670305877,
+      "loss": 1.1172,
+      "step": 8976
+    },
+    {
+      "epoch": 0.7872090585602929,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006463817516292729,
+      "loss": 1.1236,
+      "step": 8977
+    },
+    {
+      "epoch": 0.7872967503346674,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0006461091277553512,
+      "loss": 1.1162,
+      "step": 8978
+    },
+    {
+      "epoch": 0.787384442109042,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0006458365954336854,
+      "loss": 1.1215,
+      "step": 8979
+    },
+    {
+      "epoch": 0.7874721338834165,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0006455641546891298,
+      "loss": 1.1223,
+      "step": 8980
+    },
+    {
+      "epoch": 0.787559825657791,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0006452918055465309,
+      "loss": 1.1076,
+      "step": 8981
+    },
+    {
+      "epoch": 0.7876475174321657,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0006450195480307253,
+      "loss": 1.1247,
+      "step": 8982
+    },
+    {
+      "epoch": 0.7877352092065402,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0006447473821665426,
+      "loss": 1.1358,
+      "step": 8983
+    },
+    {
+      "epoch": 0.7878229009809147,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006444753079788034,
+      "loss": 1.1315,
+      "step": 8984
+    },
+    {
+      "epoch": 0.7879105927552893,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0006442033254923209,
+      "loss": 1.1581,
+      "step": 8985
+    },
+    {
+      "epoch": 0.7879982845296638,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0006439314347318983,
+      "loss": 1.1199,
+      "step": 8986
+    },
+    {
+      "epoch": 0.7880859763040383,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006436596357223316,
+      "loss": 1.1397,
+      "step": 8987
+    },
+    {
+      "epoch": 0.7881736680784129,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0006433879284884083,
+      "loss": 1.1308,
+      "step": 8988
+    },
+    {
+      "epoch": 0.7882613598527874,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0006431163130549077,
+      "loss": 1.1227,
+      "step": 8989
+    },
+    {
+      "epoch": 0.788349051627162,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0006428447894466003,
+      "loss": 1.1527,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7884367434015366,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006425733576882475,
+      "loss": 1.1154,
+      "step": 8991
+    },
+    {
+      "epoch": 0.7885244351759111,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006423020178046039,
+      "loss": 1.102,
+      "step": 8992
+    },
+    {
+      "epoch": 0.7886121269502856,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000642030769820415,
+      "loss": 1.1666,
+      "step": 8993
+    },
+    {
+      "epoch": 0.7886998187246601,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0006417596137604182,
+      "loss": 1.2063,
+      "step": 8994
+    },
+    {
+      "epoch": 0.7887875104990347,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0006414885496493414,
+      "loss": 1.1439,
+      "step": 8995
+    },
+    {
+      "epoch": 0.7888752022734092,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006412175775119055,
+      "loss": 1.1385,
+      "step": 8996
+    },
+    {
+      "epoch": 0.7889628940477837,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006409466973728223,
+      "loss": 1.0893,
+      "step": 8997
+    },
+    {
+      "epoch": 0.7890505858221584,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006406759092567956,
+      "loss": 1.1224,
+      "step": 8998
+    },
+    {
+      "epoch": 0.7891382775965329,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006404052131885206,
+      "loss": 1.164,
+      "step": 8999
+    },
+    {
+      "epoch": 0.7892259693709074,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006401346091926837,
+      "loss": 1.1306,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7892259693709074,
+      "eval_loss": 1.1407471895217896,
+      "eval_runtime": 428.9889,
+      "eval_samples_per_second": 33.677,
+      "eval_steps_per_second": 8.42,
+      "step": 9000
+    },
+    {
+      "epoch": 0.789313661145282,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006398640972939633,
+      "loss": 1.1118,
+      "step": 9001
+    },
+    {
+      "epoch": 0.7894013529196565,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006395936775170298,
+      "loss": 1.117,
+      "step": 9002
+    },
+    {
+      "epoch": 0.789489044694031,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006393233498865447,
+      "loss": 1.1224,
+      "step": 9003
+    },
+    {
+      "epoch": 0.7895767364684056,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0006390531144271609,
+      "loss": 1.1093,
+      "step": 9004
+    },
+    {
+      "epoch": 0.7896644282427802,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0006387829711635231,
+      "loss": 1.187,
+      "step": 9005
+    },
+    {
+      "epoch": 0.7897521200171547,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0006385129201202683,
+      "loss": 1.1548,
+      "step": 9006
+    },
+    {
+      "epoch": 0.7898398117915293,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006382429613220237,
+      "loss": 1.095,
+      "step": 9007
+    },
+    {
+      "epoch": 0.7899275035659038,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006379730947934097,
+      "loss": 1.1489,
+      "step": 9008
+    },
+    {
+      "epoch": 0.7900151953402783,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006377033205590364,
+      "loss": 1.1344,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7901028871146529,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006374336386435071,
+      "loss": 1.1541,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7901905788890274,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006371640490714158,
+      "loss": 1.148,
+      "step": 9011
+    },
+    {
+      "epoch": 0.790278270663402,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0006368945518673493,
+      "loss": 1.1606,
+      "step": 9012
+    },
+    {
+      "epoch": 0.7903659624377766,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0006366251470558838,
+      "loss": 1.1188,
+      "step": 9013
+    },
+    {
+      "epoch": 0.7904536542121511,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006363558346615891,
+      "loss": 1.1534,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7905413459865256,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006360866147090253,
+      "loss": 1.1252,
+      "step": 9015
+    },
+    {
+      "epoch": 0.7906290377609002,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006358174872227453,
+      "loss": 1.1249,
+      "step": 9016
+    },
+    {
+      "epoch": 0.7907167295352747,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000635548452227292,
+      "loss": 1.1153,
+      "step": 9017
+    },
+    {
+      "epoch": 0.7908044213096492,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.000635279509747201,
+      "loss": 1.0609,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7908921130840237,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006350106598069993,
+      "loss": 1.1268,
+      "step": 9019
+    },
+    {
+      "epoch": 0.7909798048583984,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006347419024312056,
+      "loss": 1.1,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7910674966327729,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006344732376443292,
+      "loss": 1.2022,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7911551884071474,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0006342046654708717,
+      "loss": 1.0776,
+      "step": 9022
+    },
+    {
+      "epoch": 0.791242880181522,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006339361859353271,
+      "loss": 1.1075,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7913305719558965,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006336677990621791,
+      "loss": 1.0695,
+      "step": 9024
+    },
+    {
+      "epoch": 0.791418263730271,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006333995048759043,
+      "loss": 1.1209,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7915059555046456,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00063313130340097,
+      "loss": 1.1308,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7915936472790202,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0006328631946618359,
+      "loss": 1.166,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7916813390533947,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006325951786829526,
+      "loss": 1.1079,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7917690308277693,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006323272554887625,
+      "loss": 1.1442,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7918567226021438,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006320594251037003,
+      "loss": 1.1177,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7919444143765183,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0006317916875521901,
+      "loss": 1.1549,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7920321061508929,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0006315240428586495,
+      "loss": 1.1457,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7921197979252674,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006312564910474871,
+      "loss": 1.1217,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7922074896996419,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006309890321431033,
+      "loss": 1.143,
+      "step": 9034
+    },
+    {
+      "epoch": 0.7922951814740166,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006307216661698887,
+      "loss": 1.1722,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7923828732483911,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0006304543931522268,
+      "loss": 1.1168,
+      "step": 9036
+    },
+    {
+      "epoch": 0.7924705650227656,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006301872131144926,
+      "loss": 1.1261,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7925582567971402,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006299201260810521,
+      "loss": 1.0978,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7926459485715147,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0006296531320762623,
+      "loss": 1.1366,
+      "step": 9039
+    },
+    {
+      "epoch": 0.7927336403458892,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0006293862311244734,
+      "loss": 1.1998,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7928213321202637,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006291194232500251,
+      "loss": 1.1143,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7929090238946384,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006288527084772498,
+      "loss": 1.1374,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7929967156690129,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006285860868304723,
+      "loss": 1.1488,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7930844074433874,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.000628319558334006,
+      "loss": 1.09,
+      "step": 9044
+    },
+    {
+      "epoch": 0.793172099217762,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0006280531230121588,
+      "loss": 1.1228,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7932597909921365,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0006277867808892285,
+      "loss": 1.1171,
+      "step": 9046
+    },
+    {
+      "epoch": 0.793347482766511,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0006275205319895055,
+      "loss": 1.0605,
+      "step": 9047
+    },
+    {
+      "epoch": 0.7934351745408856,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00062725437633727,
+      "loss": 1.1171,
+      "step": 9048
+    },
+    {
+      "epoch": 0.7935228663152601,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006269883139567951,
+      "loss": 1.1789,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7936105580896347,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0006267223448723452,
+      "loss": 1.1001,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7936982498640093,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006264564691081762,
+      "loss": 1.1068,
+      "step": 9051
+    },
+    {
+      "epoch": 0.7937859416383838,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006261906866885346,
+      "loss": 1.0892,
+      "step": 9052
+    },
+    {
+      "epoch": 0.7938736334127583,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006259249976376594,
+      "loss": 1.0549,
+      "step": 9053
+    },
+    {
+      "epoch": 0.7939613251871329,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006256594019797808,
+      "loss": 1.1802,
+      "step": 9054
+    },
+    {
+      "epoch": 0.7940490169615074,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006253938997391204,
+      "loss": 1.1563,
+      "step": 9055
+    },
+    {
+      "epoch": 0.7941367087358819,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006251284909398918,
+      "loss": 1.1438,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7942244005102566,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.000624863175606299,
+      "loss": 1.0949,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7943120922846311,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0006245979537625377,
+      "loss": 1.1091,
+      "step": 9058
+    },
+    {
+      "epoch": 0.7943997840590056,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0006243328254327962,
+      "loss": 1.1325,
+      "step": 9059
+    },
+    {
+      "epoch": 0.7944874758333802,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0006240677906412533,
+      "loss": 1.1239,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7945751676077547,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006238028494120798,
+      "loss": 1.1056,
+      "step": 9061
+    },
+    {
+      "epoch": 0.7946628593821292,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0006235380017694367,
+      "loss": 1.0702,
+      "step": 9062
+    },
+    {
+      "epoch": 0.7947505511565038,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006232732477374782,
+      "loss": 1.1675,
+      "step": 9063
+    },
+    {
+      "epoch": 0.7948382429308783,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006230085873403491,
+      "loss": 1.1573,
+      "step": 9064
+    },
+    {
+      "epoch": 0.7949259347052529,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.000622744020602186,
+      "loss": 1.0541,
+      "step": 9065
+    },
+    {
+      "epoch": 0.7950136264796274,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0006224795475471159,
+      "loss": 1.1356,
+      "step": 9066
+    },
+    {
+      "epoch": 0.795101318254002,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006222151681992587,
+      "loss": 1.0819,
+      "step": 9067
+    },
+    {
+      "epoch": 0.7951890100283765,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0006219508825827249,
+      "loss": 1.116,
+      "step": 9068
+    },
+    {
+      "epoch": 0.795276701802751,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0006216866907216171,
+      "loss": 1.1154,
+      "step": 9069
+    },
+    {
+      "epoch": 0.7953643935771256,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0006214225926400281,
+      "loss": 1.1331,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7954520853515001,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0006211585883620435,
+      "loss": 1.1331,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7955397771258746,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006208946779117395,
+      "loss": 1.1622,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7956274689002493,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0006206308613131848,
+      "loss": 1.2091,
+      "step": 9073
+    },
+    {
+      "epoch": 0.7957151606746238,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0006203671385904381,
+      "loss": 1.1037,
+      "step": 9074
+    },
+    {
+      "epoch": 0.7958028524489983,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0006201035097675501,
+      "loss": 1.1571,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7958905442233729,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0006198399748685633,
+      "loss": 1.1815,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7959782359977474,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006195765339175115,
+      "loss": 1.1223,
+      "step": 9077
+    },
+    {
+      "epoch": 0.7960659277721219,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006193131869384202,
+      "loss": 1.2083,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7961536195464965,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0006190499339553051,
+      "loss": 1.1192,
+      "step": 9079
+    },
+    {
+      "epoch": 0.7962413113208711,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0006187867749921746,
+      "loss": 1.1091,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7963290030952456,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006185237100730281,
+      "loss": 1.1815,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7964166948696202,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0006182607392218566,
+      "loss": 1.1268,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7965043866439947,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006179978624626427,
+      "loss": 1.0693,
+      "step": 9083
+    },
+    {
+      "epoch": 0.7965920784183692,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006177350798193591,
+      "loss": 1.1332,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7966797701927438,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006174723913159717,
+      "loss": 1.1883,
+      "step": 9085
+    },
+    {
+      "epoch": 0.7967674619671183,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006172097969764366,
+      "loss": 1.1484,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7968551537414928,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0006169472968247023,
+      "loss": 1.1208,
+      "step": 9087
+    },
+    {
+      "epoch": 0.7969428455158675,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0006166848908847073,
+      "loss": 1.1419,
+      "step": 9088
+    },
+    {
+      "epoch": 0.797030537290242,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006164225791803828,
+      "loss": 1.1159,
+      "step": 9089
+    },
+    {
+      "epoch": 0.7971182290646165,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006161603617356508,
+      "loss": 1.0835,
+      "step": 9090
+    },
+    {
+      "epoch": 0.797205920838991,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006158982385744254,
+      "loss": 1.0899,
+      "step": 9091
+    },
+    {
+      "epoch": 0.7972936126133656,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000615636209720611,
+      "loss": 1.2,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7973813043877401,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.000615374275198104,
+      "loss": 1.1592,
+      "step": 9093
+    },
+    {
+      "epoch": 0.7974689961621146,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0006151124350307921,
+      "loss": 1.1291,
+      "step": 9094
+    },
+    {
+      "epoch": 0.7975566879364893,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006148506892425545,
+      "loss": 1.0863,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7976443797108638,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0006145890378572623,
+      "loss": 1.0934,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7977320714852383,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006143274808987765,
+      "loss": 1.1264,
+      "step": 9097
+    },
+    {
+      "epoch": 0.7978197632596129,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006140660183909508,
+      "loss": 1.1944,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7979074550339874,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0006138046503576302,
+      "loss": 1.0747,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7979951468083619,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006135433768226507,
+      "loss": 1.098,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7980828385827365,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0006132821978098393,
+      "loss": 1.1361,
+      "step": 9101
+    },
+    {
+      "epoch": 0.798170530357111,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006130211133430153,
+      "loss": 1.1153,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7982582221314856,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006127601234459888,
+      "loss": 1.1259,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7983459139058602,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006124992281425617,
+      "loss": 1.162,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7984336056802347,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0006122384274565263,
+      "loss": 1.1153,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7985212974546092,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0006119777214116676,
+      "loss": 1.0784,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7986089892289838,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0006117171100317608,
+      "loss": 1.1493,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7986966810033583,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006114565933405736,
+      "loss": 1.1809,
+      "step": 9108
+    },
+    {
+      "epoch": 0.7987843727777328,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000611196171361864,
+      "loss": 1.1241,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7988720645521075,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006109358441193821,
+      "loss": 1.0945,
+      "step": 9110
+    },
+    {
+      "epoch": 0.798959756326482,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0006106756116368685,
+      "loss": 1.1021,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7990474481008565,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0006104154739380562,
+      "loss": 1.12,
+      "step": 9112
+    },
+    {
+      "epoch": 0.799135139875231,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006101554310466689,
+      "loss": 1.1035,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7992228316496056,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0006098954829864224,
+      "loss": 1.1626,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7993105234239801,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006096356297810224,
+      "loss": 1.0878,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7993982151983546,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006093758714541674,
+      "loss": 1.149,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7994859069727293,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006091162080295463,
+      "loss": 1.2016,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7995735987471038,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006088566395308407,
+      "loss": 1.1447,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7996612905214783,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0006085971659817213,
+      "loss": 1.1134,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7997489822958529,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006083377874058521,
+      "loss": 1.1025,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7998366740702274,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006080785038268876,
+      "loss": 1.1129,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7999243658446019,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006078193152684742,
+      "loss": 1.1528,
+      "step": 9122
+    },
+    {
+      "epoch": 0.8000120576189765,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006075602217542485,
+      "loss": 1.1656,
+      "step": 9123
+    },
+    {
+      "epoch": 0.800099749393351,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006073012233078396,
+      "loss": 1.1705,
+      "step": 9124
+    },
+    {
+      "epoch": 0.8001874411677256,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006070423199528677,
+      "loss": 1.0995,
+      "step": 9125
+    },
+    {
+      "epoch": 0.8002751329421002,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0006067835117129436,
+      "loss": 1.1841,
+      "step": 9126
+    },
+    {
+      "epoch": 0.8003628247164747,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0006065247986116705,
+      "loss": 1.1183,
+      "step": 9127
+    },
+    {
+      "epoch": 0.8004505164908492,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006062661806726416,
+      "loss": 1.1949,
+      "step": 9128
+    },
+    {
+      "epoch": 0.8005382082652238,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0006060076579194428,
+      "loss": 1.1408,
+      "step": 9129
+    },
+    {
+      "epoch": 0.8006259000395983,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0006057492303756503,
+      "loss": 1.1182,
+      "step": 9130
+    },
+    {
+      "epoch": 0.8007135918139728,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006054908980648327,
+      "loss": 1.1501,
+      "step": 9131
+    },
+    {
+      "epoch": 0.8008012835883475,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006052326610105486,
+      "loss": 1.1055,
+      "step": 9132
+    },
+    {
+      "epoch": 0.800888975362722,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0006049745192363486,
+      "loss": 1.1328,
+      "step": 9133
+    },
+    {
+      "epoch": 0.8009766671370965,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0006047164727657749,
+      "loss": 1.0939,
+      "step": 9134
+    },
+    {
+      "epoch": 0.8010643589114711,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0006044585216223603,
+      "loss": 1.1942,
+      "step": 9135
+    },
+    {
+      "epoch": 0.8011520506858456,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00060420066582963,
+      "loss": 1.1072,
+      "step": 9136
+    },
+    {
+      "epoch": 0.8012397424602201,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0006039429054110987,
+      "loss": 1.1332,
+      "step": 9137
+    },
+    {
+      "epoch": 0.8013274342345946,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0006036852403902743,
+      "loss": 1.0691,
+      "step": 9138
+    },
+    {
+      "epoch": 0.8014151260089692,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006034276707906548,
+      "loss": 1.1077,
+      "step": 9139
+    },
+    {
+      "epoch": 0.8015028177833438,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0006031701966357304,
+      "loss": 1.1056,
+      "step": 9140
+    },
+    {
+      "epoch": 0.8015905095577183,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006029128179489813,
+      "loss": 1.1231,
+      "step": 9141
+    },
+    {
+      "epoch": 0.8016782013320929,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006026555347538806,
+      "loss": 1.1495,
+      "step": 9142
+    },
+    {
+      "epoch": 0.8017658931064674,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000602398347073891,
+      "loss": 1.1651,
+      "step": 9143
+    },
+    {
+      "epoch": 0.8018535848808419,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006021412549324677,
+      "loss": 1.0998,
+      "step": 9144
+    },
+    {
+      "epoch": 0.8019412766552165,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006018842583530574,
+      "loss": 1.1035,
+      "step": 9145
+    },
+    {
+      "epoch": 0.802028968429591,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0006016273573590966,
+      "loss": 1.1187,
+      "step": 9146
+    },
+    {
+      "epoch": 0.8021166602039655,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006013705519740145,
+      "loss": 1.1263,
+      "step": 9147
+    },
+    {
+      "epoch": 0.8022043519783402,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006011138422212309,
+      "loss": 1.1447,
+      "step": 9148
+    },
+    {
+      "epoch": 0.8022920437527147,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0006008572281241578,
+      "loss": 1.1089,
+      "step": 9149
+    },
+    {
+      "epoch": 0.8023797355270892,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0006006007097061967,
+      "loss": 1.1459,
+      "step": 9150
+    },
+    {
+      "epoch": 0.8024674273014638,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0006003442869907417,
+      "loss": 1.1257,
+      "step": 9151
+    },
+    {
+      "epoch": 0.8025551190758383,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0006000879600011782,
+      "loss": 1.1322,
+      "step": 9152
+    },
+    {
+      "epoch": 0.8026428108502128,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005998317287608828,
+      "loss": 1.1317,
+      "step": 9153
+    },
+    {
+      "epoch": 0.8027305026245874,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005995755932932224,
+      "loss": 1.0847,
+      "step": 9154
+    },
+    {
+      "epoch": 0.802818194398962,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000599319553621556,
+      "loss": 1.1207,
+      "step": 9155
+    },
+    {
+      "epoch": 0.8029058861733365,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005990636097692343,
+      "loss": 1.154,
+      "step": 9156
+    },
+    {
+      "epoch": 0.8029935779477111,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005988077617595986,
+      "loss": 1.1267,
+      "step": 9157
+    },
+    {
+      "epoch": 0.8030812697220856,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005985520096159812,
+      "loss": 1.1921,
+      "step": 9158
+    },
+    {
+      "epoch": 0.8031689614964601,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005982963533617064,
+      "loss": 1.1132,
+      "step": 9159
+    },
+    {
+      "epoch": 0.8032566532708346,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005980407930200888,
+      "loss": 1.0891,
+      "step": 9160
+    },
+    {
+      "epoch": 0.8033443450452092,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005977853286144353,
+      "loss": 1.0859,
+      "step": 9161
+    },
+    {
+      "epoch": 0.8034320368195838,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005975299601680437,
+      "loss": 1.1116,
+      "step": 9162
+    },
+    {
+      "epoch": 0.8035197285939583,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000597274687704203,
+      "loss": 1.1149,
+      "step": 9163
+    },
+    {
+      "epoch": 0.8036074203683329,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005970195112461929,
+      "loss": 1.0943,
+      "step": 9164
+    },
+    {
+      "epoch": 0.8036951121427074,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005967644308172852,
+      "loss": 1.1193,
+      "step": 9165
+    },
+    {
+      "epoch": 0.8037828039170819,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005965094464407424,
+      "loss": 1.1208,
+      "step": 9166
+    },
+    {
+      "epoch": 0.8038704956914565,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0005962545581398189,
+      "loss": 1.0602,
+      "step": 9167
+    },
+    {
+      "epoch": 0.803958187465831,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.000595999765937759,
+      "loss": 1.1482,
+      "step": 9168
+    },
+    {
+      "epoch": 0.8040458792402055,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005957450698577997,
+      "loss": 1.1856,
+      "step": 9169
+    },
+    {
+      "epoch": 0.8041335710145802,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0005954904699231684,
+      "loss": 1.0747,
+      "step": 9170
+    },
+    {
+      "epoch": 0.8042212627889547,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005952359661570845,
+      "loss": 1.1007,
+      "step": 9171
+    },
+    {
+      "epoch": 0.8043089545633292,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005949815585827572,
+      "loss": 1.1138,
+      "step": 9172
+    },
+    {
+      "epoch": 0.8043966463377038,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005947272472233882,
+      "loss": 1.1271,
+      "step": 9173
+    },
+    {
+      "epoch": 0.8044843381120783,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005944730321021701,
+      "loss": 1.1044,
+      "step": 9174
+    },
+    {
+      "epoch": 0.8045720298864528,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005942189132422871,
+      "loss": 1.1568,
+      "step": 9175
+    },
+    {
+      "epoch": 0.8046597216608274,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005939648906669137,
+      "loss": 1.0934,
+      "step": 9176
+    },
+    {
+      "epoch": 0.804747413435202,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005937109643992157,
+      "loss": 1.1795,
+      "step": 9177
+    },
+    {
+      "epoch": 0.8048351052095765,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005934571344623509,
+      "loss": 1.0788,
+      "step": 9178
+    },
+    {
+      "epoch": 0.8049227969839511,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005932034008794681,
+      "loss": 1.109,
+      "step": 9179
+    },
+    {
+      "epoch": 0.8050104887583256,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005929497636737073,
+      "loss": 1.1257,
+      "step": 9180
+    },
+    {
+      "epoch": 0.8050981805327001,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000592696222868199,
+      "loss": 1.1558,
+      "step": 9181
+    },
+    {
+      "epoch": 0.8051858723070747,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005924427784860657,
+      "loss": 1.1571,
+      "step": 9182
+    },
+    {
+      "epoch": 0.8052735640814492,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.000592189430550421,
+      "loss": 1.1009,
+      "step": 9183
+    },
+    {
+      "epoch": 0.8053612558558237,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005919361790843698,
+      "loss": 1.1374,
+      "step": 9184
+    },
+    {
+      "epoch": 0.8054489476301983,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005916830241110074,
+      "loss": 1.1207,
+      "step": 9185
+    },
+    {
+      "epoch": 0.8055366394045729,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005914299656534211,
+      "loss": 1.1193,
+      "step": 9186
+    },
+    {
+      "epoch": 0.8056243311789474,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0005911770037346893,
+      "loss": 1.1663,
+      "step": 9187
+    },
+    {
+      "epoch": 0.8057120229533219,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005909241383778814,
+      "loss": 1.1316,
+      "step": 9188
+    },
+    {
+      "epoch": 0.8057997147276965,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005906713696060585,
+      "loss": 1.1531,
+      "step": 9189
+    },
+    {
+      "epoch": 0.805887406502071,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005904186974422717,
+      "loss": 1.0923,
+      "step": 9190
+    },
+    {
+      "epoch": 0.8059750982764455,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005901661219095644,
+      "loss": 1.1188,
+      "step": 9191
+    },
+    {
+      "epoch": 0.8060627900508202,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005899136430309711,
+      "loss": 1.124,
+      "step": 9192
+    },
+    {
+      "epoch": 0.8061504818251947,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005896612608295167,
+      "loss": 1.1534,
+      "step": 9193
+    },
+    {
+      "epoch": 0.8062381735995692,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005894089753282184,
+      "loss": 1.0778,
+      "step": 9194
+    },
+    {
+      "epoch": 0.8063258653739438,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005891567865500832,
+      "loss": 1.0945,
+      "step": 9195
+    },
+    {
+      "epoch": 0.8064135571483183,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005889046945181106,
+      "loss": 1.1654,
+      "step": 9196
+    },
+    {
+      "epoch": 0.8065012489226928,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005886526992552904,
+      "loss": 1.1498,
+      "step": 9197
+    },
+    {
+      "epoch": 0.8065889406970674,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.000588400800784605,
+      "loss": 1.1059,
+      "step": 9198
+    },
+    {
+      "epoch": 0.806676632471442,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005881489991290252,
+      "loss": 1.1426,
+      "step": 9199
+    },
+    {
+      "epoch": 0.8067643242458165,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0005878972943115159,
+      "loss": 1.1368,
+      "step": 9200
+    },
+    {
+      "epoch": 0.8068520160201911,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005876456863550315,
+      "loss": 1.1687,
+      "step": 9201
+    },
+    {
+      "epoch": 0.8069397077945656,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005873941752825183,
+      "loss": 1.1419,
+      "step": 9202
+    },
+    {
+      "epoch": 0.8070273995689401,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.000587142761116913,
+      "loss": 1.0956,
+      "step": 9203
+    },
+    {
+      "epoch": 0.8071150913433147,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000586891443881144,
+      "loss": 1.0782,
+      "step": 9204
+    },
+    {
+      "epoch": 0.8072027831176892,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005866402235981309,
+      "loss": 1.1099,
+      "step": 9205
+    },
+    {
+      "epoch": 0.8072904748920637,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005863891002907849,
+      "loss": 1.1006,
+      "step": 9206
+    },
+    {
+      "epoch": 0.8073781666664384,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005861380739820065,
+      "loss": 1.1861,
+      "step": 9207
+    },
+    {
+      "epoch": 0.8074658584408129,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005858871446946897,
+      "loss": 1.0721,
+      "step": 9208
+    },
+    {
+      "epoch": 0.8075535502151874,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005856363124517189,
+      "loss": 1.1253,
+      "step": 9209
+    },
+    {
+      "epoch": 0.8076412419895619,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005853855772759681,
+      "loss": 1.124,
+      "step": 9210
+    },
+    {
+      "epoch": 0.8077289337639365,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005851349391903046,
+      "loss": 1.0994,
+      "step": 9211
+    },
+    {
+      "epoch": 0.807816625538311,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005848843982175857,
+      "loss": 1.1379,
+      "step": 9212
+    },
+    {
+      "epoch": 0.8079043173126855,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 0.0005846339543806599,
+      "loss": 1.1196,
+      "step": 9213
+    },
+    {
+      "epoch": 0.8079920090870601,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0005843836077023671,
+      "loss": 1.0609,
+      "step": 9214
+    },
+    {
+      "epoch": 0.8080797008614347,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005841333582055387,
+      "loss": 1.0578,
+      "step": 9215
+    },
+    {
+      "epoch": 0.8081673926358092,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005838832059129967,
+      "loss": 1.0787,
+      "step": 9216
+    },
+    {
+      "epoch": 0.8082550844101838,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005836331508475537,
+      "loss": 1.0827,
+      "step": 9217
+    },
+    {
+      "epoch": 0.8083427761845583,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005833831930320147,
+      "loss": 1.0956,
+      "step": 9218
+    },
+    {
+      "epoch": 0.8084304679589328,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005831333324891751,
+      "loss": 1.1561,
+      "step": 9219
+    },
+    {
+      "epoch": 0.8085181597333074,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005828835692418217,
+      "loss": 1.1686,
+      "step": 9220
+    },
+    {
+      "epoch": 0.8086058515076819,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005826339033127317,
+      "loss": 1.0974,
+      "step": 9221
+    },
+    {
+      "epoch": 0.8086935432820564,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005823843347246743,
+      "loss": 1.159,
+      "step": 9222
+    },
+    {
+      "epoch": 0.8087812350564311,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005821348635004096,
+      "loss": 1.1683,
+      "step": 9223
+    },
+    {
+      "epoch": 0.8088689268308056,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0005818854896626891,
+      "loss": 1.1019,
+      "step": 9224
+    },
+    {
+      "epoch": 0.8089566186051801,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.000581636213234254,
+      "loss": 1.1119,
+      "step": 9225
+    },
+    {
+      "epoch": 0.8090443103795547,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.000581387034237839,
+      "loss": 1.1439,
+      "step": 9226
+    },
+    {
+      "epoch": 0.8091320021539292,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0005811379526961673,
+      "loss": 1.1232,
+      "step": 9227
+    },
+    {
+      "epoch": 0.8092196939283037,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005808889686319549,
+      "loss": 1.1207,
+      "step": 9228
+    },
+    {
+      "epoch": 0.8093073857026784,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005806400820679094,
+      "loss": 1.1605,
+      "step": 9229
+    },
+    {
+      "epoch": 0.8093950774770529,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005803912930267272,
+      "loss": 1.1491,
+      "step": 9230
+    },
+    {
+      "epoch": 0.8094827692514274,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.000580142601531098,
+      "loss": 1.0774,
+      "step": 9231
+    },
+    {
+      "epoch": 0.8095704610258019,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005798940076037018,
+      "loss": 1.1035,
+      "step": 9232
+    },
+    {
+      "epoch": 0.8096581528001765,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00057964551126721,
+      "loss": 1.0883,
+      "step": 9233
+    },
+    {
+      "epoch": 0.809745844574551,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000579397112544284,
+      "loss": 1.1322,
+      "step": 9234
+    },
+    {
+      "epoch": 0.8098335363489255,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005791488114575775,
+      "loss": 1.1261,
+      "step": 9235
+    },
+    {
+      "epoch": 0.8099212281233001,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005789006080297353,
+      "loss": 1.1373,
+      "step": 9236
+    },
+    {
+      "epoch": 0.8100089198976747,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005786525022833928,
+      "loss": 1.1358,
+      "step": 9237
+    },
+    {
+      "epoch": 0.8100966116720492,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000578404494241176,
+      "loss": 1.1212,
+      "step": 9238
+    },
+    {
+      "epoch": 0.8101843034464238,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005781565839257031,
+      "loss": 1.1261,
+      "step": 9239
+    },
+    {
+      "epoch": 0.8102719952207983,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0005779087713595828,
+      "loss": 1.1094,
+      "step": 9240
+    },
+    {
+      "epoch": 0.8103596869951728,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005776610565654149,
+      "loss": 1.1065,
+      "step": 9241
+    },
+    {
+      "epoch": 0.8104473787695474,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005774134395657908,
+      "loss": 1.127,
+      "step": 9242
+    },
+    {
+      "epoch": 0.8105350705439219,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005771659203832921,
+      "loss": 1.1426,
+      "step": 9243
+    },
+    {
+      "epoch": 0.8106227623182964,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005769184990404917,
+      "loss": 1.141,
+      "step": 9244
+    },
+    {
+      "epoch": 0.8107104540926711,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000576671175559954,
+      "loss": 1.1165,
+      "step": 9245
+    },
+    {
+      "epoch": 0.8107981458670456,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005764239499642345,
+      "loss": 1.1643,
+      "step": 9246
+    },
+    {
+      "epoch": 0.8108858376414201,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005761768222758794,
+      "loss": 1.1296,
+      "step": 9247
+    },
+    {
+      "epoch": 0.8109735294157947,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005759297925174258,
+      "loss": 1.1311,
+      "step": 9248
+    },
+    {
+      "epoch": 0.8110612211901692,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005756828607114023,
+      "loss": 1.0759,
+      "step": 9249
+    },
+    {
+      "epoch": 0.8111489129645437,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005754360268803287,
+      "loss": 1.2053,
+      "step": 9250
+    },
+    {
+      "epoch": 0.8112366047389183,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005751892910467158,
+      "loss": 1.1632,
+      "step": 9251
+    },
+    {
+      "epoch": 0.8113242965132929,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005749426532330647,
+      "loss": 1.1217,
+      "step": 9252
+    },
+    {
+      "epoch": 0.8114119882876674,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005746961134618681,
+      "loss": 1.1207,
+      "step": 9253
+    },
+    {
+      "epoch": 0.811499680062042,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005744496717556105,
+      "loss": 1.1801,
+      "step": 9254
+    },
+    {
+      "epoch": 0.8115873718364165,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005742033281367664,
+      "loss": 1.0634,
+      "step": 9255
+    },
+    {
+      "epoch": 0.811675063610791,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005739570826278014,
+      "loss": 1.103,
+      "step": 9256
+    },
+    {
+      "epoch": 0.8117627553851655,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005737109352511724,
+      "loss": 1.1117,
+      "step": 9257
+    },
+    {
+      "epoch": 0.8118504471595401,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005734648860293278,
+      "loss": 1.1345,
+      "step": 9258
+    },
+    {
+      "epoch": 0.8119381389339146,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005732189349847071,
+      "loss": 1.0801,
+      "step": 9259
+    },
+    {
+      "epoch": 0.8120258307082892,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005729730821397396,
+      "loss": 1.1172,
+      "step": 9260
+    },
+    {
+      "epoch": 0.8121135224826638,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005727273275168463,
+      "loss": 1.0871,
+      "step": 9261
+    },
+    {
+      "epoch": 0.8122012142570383,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005724816711384401,
+      "loss": 1.1154,
+      "step": 9262
+    },
+    {
+      "epoch": 0.8122889060314128,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005722361130269236,
+      "loss": 1.1348,
+      "step": 9263
+    },
+    {
+      "epoch": 0.8123765978057874,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005719906532046916,
+      "loss": 1.1395,
+      "step": 9264
+    },
+    {
+      "epoch": 0.8124642895801619,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000571745291694129,
+      "loss": 1.1329,
+      "step": 9265
+    },
+    {
+      "epoch": 0.8125519813545364,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005715000285176123,
+      "loss": 1.1948,
+      "step": 9266
+    },
+    {
+      "epoch": 0.812639673128911,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005712548636975088,
+      "loss": 1.1354,
+      "step": 9267
+    },
+    {
+      "epoch": 0.8127273649032856,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005710097972561773,
+      "loss": 1.1652,
+      "step": 9268
+    },
+    {
+      "epoch": 0.8128150566776601,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005707648292159668,
+      "loss": 1.0781,
+      "step": 9269
+    },
+    {
+      "epoch": 0.8129027484520347,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005705199595992176,
+      "loss": 1.1954,
+      "step": 9270
+    },
+    {
+      "epoch": 0.8129904402264092,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005702751884282614,
+      "loss": 1.0935,
+      "step": 9271
+    },
+    {
+      "epoch": 0.8130781320007837,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005700305157254207,
+      "loss": 1.1218,
+      "step": 9272
+    },
+    {
+      "epoch": 0.8131658237751583,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005697859415130095,
+      "loss": 1.1236,
+      "step": 9273
+    },
+    {
+      "epoch": 0.8132535155495328,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005695414658133315,
+      "loss": 1.1597,
+      "step": 9274
+    },
+    {
+      "epoch": 0.8133412073239074,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005692970886486825,
+      "loss": 1.1672,
+      "step": 9275
+    },
+    {
+      "epoch": 0.813428899098282,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005690528100413494,
+      "loss": 1.0898,
+      "step": 9276
+    },
+    {
+      "epoch": 0.8135165908726565,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005688086300136099,
+      "loss": 1.0864,
+      "step": 9277
+    },
+    {
+      "epoch": 0.813604282647031,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005685645485877322,
+      "loss": 1.083,
+      "step": 9278
+    },
+    {
+      "epoch": 0.8136919744214056,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005683205657859756,
+      "loss": 1.1697,
+      "step": 9279
+    },
+    {
+      "epoch": 0.8137796661957801,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.000568076681630591,
+      "loss": 1.1222,
+      "step": 9280
+    },
+    {
+      "epoch": 0.8138673579701546,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005678328961438201,
+      "loss": 1.1486,
+      "step": 9281
+    },
+    {
+      "epoch": 0.8139550497445291,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.000567589209347896,
+      "loss": 1.169,
+      "step": 9282
+    },
+    {
+      "epoch": 0.8140427415189038,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005673456212650412,
+      "loss": 1.1248,
+      "step": 9283
+    },
+    {
+      "epoch": 0.8141304332932783,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005671021319174709,
+      "loss": 1.1091,
+      "step": 9284
+    },
+    {
+      "epoch": 0.8142181250676528,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005668587413273908,
+      "loss": 1.1363,
+      "step": 9285
+    },
+    {
+      "epoch": 0.8143058168420274,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005666154495169978,
+      "loss": 1.1442,
+      "step": 9286
+    },
+    {
+      "epoch": 0.8143935086164019,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005663722565084784,
+      "loss": 1.1048,
+      "step": 9287
+    },
+    {
+      "epoch": 0.8144812003907764,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.000566129162324012,
+      "loss": 1.101,
+      "step": 9288
+    },
+    {
+      "epoch": 0.814568892165151,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0005658861669857682,
+      "loss": 1.157,
+      "step": 9289
+    },
+    {
+      "epoch": 0.8146565839395256,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005656432705159075,
+      "loss": 1.1482,
+      "step": 9290
+    },
+    {
+      "epoch": 0.8147442757139001,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005654004729365809,
+      "loss": 1.1618,
+      "step": 9291
+    },
+    {
+      "epoch": 0.8148319674882747,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005651577742699313,
+      "loss": 1.1374,
+      "step": 9292
+    },
+    {
+      "epoch": 0.8149196592626492,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005649151745380921,
+      "loss": 1.085,
+      "step": 9293
+    },
+    {
+      "epoch": 0.8150073510370237,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005646726737631885,
+      "loss": 1.1414,
+      "step": 9294
+    },
+    {
+      "epoch": 0.8150950428113983,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000564430271967335,
+      "loss": 1.1432,
+      "step": 9295
+    },
+    {
+      "epoch": 0.8151827345857728,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005641879691726379,
+      "loss": 1.1774,
+      "step": 9296
+    },
+    {
+      "epoch": 0.8152704263601473,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005639457654011954,
+      "loss": 1.1015,
+      "step": 9297
+    },
+    {
+      "epoch": 0.815358118134522,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005637036606750953,
+      "loss": 1.1158,
+      "step": 9298
+    },
+    {
+      "epoch": 0.8154458099088965,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005634616550164169,
+      "loss": 1.1498,
+      "step": 9299
+    },
+    {
+      "epoch": 0.815533501683271,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005632197484472312,
+      "loss": 1.1196,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8156211934576456,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005629779409895983,
+      "loss": 1.1293,
+      "step": 9301
+    },
+    {
+      "epoch": 0.8157088852320201,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005627362326655715,
+      "loss": 1.1439,
+      "step": 9302
+    },
+    {
+      "epoch": 0.8157965770063946,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005624946234971932,
+      "loss": 1.1481,
+      "step": 9303
+    },
+    {
+      "epoch": 0.8158842687807691,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005622531135064983,
+      "loss": 1.1299,
+      "step": 9304
+    },
+    {
+      "epoch": 0.8159719605551438,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005620117027155112,
+      "loss": 1.0746,
+      "step": 9305
+    },
+    {
+      "epoch": 0.8160596523295183,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005617703911462482,
+      "loss": 1.1264,
+      "step": 9306
+    },
+    {
+      "epoch": 0.8161473441038928,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005615291788207161,
+      "loss": 1.1398,
+      "step": 9307
+    },
+    {
+      "epoch": 0.8162350358782674,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005612880657609133,
+      "loss": 1.1568,
+      "step": 9308
+    },
+    {
+      "epoch": 0.8163227276526419,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005610470519888281,
+      "loss": 1.1489,
+      "step": 9309
+    },
+    {
+      "epoch": 0.8164104194270164,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005608061375264408,
+      "loss": 1.123,
+      "step": 9310
+    },
+    {
+      "epoch": 0.816498111201391,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005605653223957223,
+      "loss": 1.203,
+      "step": 9311
+    },
+    {
+      "epoch": 0.8165858029757656,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005603246066186336,
+      "loss": 1.2096,
+      "step": 9312
+    },
+    {
+      "epoch": 0.8166734947501401,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005600839902171284,
+      "loss": 1.1389,
+      "step": 9313
+    },
+    {
+      "epoch": 0.8167611865245147,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000559843473213149,
+      "loss": 1.1223,
+      "step": 9314
+    },
+    {
+      "epoch": 0.8168488782988892,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005596030556286309,
+      "loss": 1.1524,
+      "step": 9315
+    },
+    {
+      "epoch": 0.8169365700732637,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005593627374854991,
+      "loss": 1.1613,
+      "step": 9316
+    },
+    {
+      "epoch": 0.8170242618476383,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005591225188056705,
+      "loss": 1.1139,
+      "step": 9317
+    },
+    {
+      "epoch": 0.8171119536220128,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000558882399611052,
+      "loss": 1.1525,
+      "step": 9318
+    },
+    {
+      "epoch": 0.8171996453963873,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005586423799235417,
+      "loss": 1.2005,
+      "step": 9319
+    },
+    {
+      "epoch": 0.817287337170762,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005584024597650292,
+      "loss": 1.1648,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8173750289451365,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005581626391573947,
+      "loss": 1.1723,
+      "step": 9321
+    },
+    {
+      "epoch": 0.817462720719511,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005579229181225084,
+      "loss": 1.1581,
+      "step": 9322
+    },
+    {
+      "epoch": 0.8175504124938856,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005576832966822331,
+      "loss": 1.0623,
+      "step": 9323
+    },
+    {
+      "epoch": 0.8176381042682601,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005574437748584212,
+      "loss": 1.1253,
+      "step": 9324
+    },
+    {
+      "epoch": 0.8177257960426346,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005572043526729165,
+      "loss": 1.1632,
+      "step": 9325
+    },
+    {
+      "epoch": 0.8178134878170092,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005569650301475544,
+      "loss": 1.108,
+      "step": 9326
+    },
+    {
+      "epoch": 0.8179011795913838,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005567258073041594,
+      "loss": 1.1252,
+      "step": 9327
+    },
+    {
+      "epoch": 0.8179888713657583,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005564866841645489,
+      "loss": 1.1345,
+      "step": 9328
+    },
+    {
+      "epoch": 0.8180765631401328,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005562476607505294,
+      "loss": 1.1327,
+      "step": 9329
+    },
+    {
+      "epoch": 0.8181642549145074,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005560087370839,
+      "loss": 1.1051,
+      "step": 9330
+    },
+    {
+      "epoch": 0.8182519466888819,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005557699131864498,
+      "loss": 1.0976,
+      "step": 9331
+    },
+    {
+      "epoch": 0.8183396384632564,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005555311890799587,
+      "loss": 1.1253,
+      "step": 9332
+    },
+    {
+      "epoch": 0.818427330237631,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005552925647861973,
+      "loss": 1.1574,
+      "step": 9333
+    },
+    {
+      "epoch": 0.8185150220120055,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0005550540403269286,
+      "loss": 1.0956,
+      "step": 9334
+    },
+    {
+      "epoch": 0.81860271378638,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005548156157239046,
+      "loss": 1.1553,
+      "step": 9335
+    },
+    {
+      "epoch": 0.8186904055607547,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005545772909988693,
+      "loss": 1.1972,
+      "step": 9336
+    },
+    {
+      "epoch": 0.8187780973351292,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005543390661735571,
+      "loss": 1.1036,
+      "step": 9337
+    },
+    {
+      "epoch": 0.8188657891095037,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005541009412696937,
+      "loss": 1.1856,
+      "step": 9338
+    },
+    {
+      "epoch": 0.8189534808838783,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005538629163089956,
+      "loss": 1.1341,
+      "step": 9339
+    },
+    {
+      "epoch": 0.8190411726582528,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005536249913131697,
+      "loss": 1.1511,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8191288644326273,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005533871663039143,
+      "loss": 1.1279,
+      "step": 9341
+    },
+    {
+      "epoch": 0.819216556207002,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005531494413029184,
+      "loss": 1.1228,
+      "step": 9342
+    },
+    {
+      "epoch": 0.8193042479813765,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005529118163318624,
+      "loss": 1.1713,
+      "step": 9343
+    },
+    {
+      "epoch": 0.819391939755751,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005526742914124162,
+      "loss": 1.1058,
+      "step": 9344
+    },
+    {
+      "epoch": 0.8194796315301256,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005524368665662421,
+      "loss": 1.1127,
+      "step": 9345
+    },
+    {
+      "epoch": 0.8195673233045001,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005521995418149922,
+      "loss": 1.1265,
+      "step": 9346
+    },
+    {
+      "epoch": 0.8196550150788746,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005519623171803103,
+      "loss": 1.0708,
+      "step": 9347
+    },
+    {
+      "epoch": 0.8197427068532492,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005517251926838306,
+      "loss": 1.1559,
+      "step": 9348
+    },
+    {
+      "epoch": 0.8198303986276237,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005514881683471777,
+      "loss": 1.1404,
+      "step": 9349
+    },
+    {
+      "epoch": 0.8199180904019983,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0005512512441919682,
+      "loss": 1.0509,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8200057821763728,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0005510144202398088,
+      "loss": 1.0908,
+      "step": 9351
+    },
+    {
+      "epoch": 0.8200934739507474,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005507776965122974,
+      "loss": 1.1447,
+      "step": 9352
+    },
+    {
+      "epoch": 0.8201811657251219,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005505410730310227,
+      "loss": 1.1395,
+      "step": 9353
+    },
+    {
+      "epoch": 0.8202688574994964,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005503045498175636,
+      "loss": 1.0863,
+      "step": 9354
+    },
+    {
+      "epoch": 0.820356549273871,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005500681268934906,
+      "loss": 1.1353,
+      "step": 9355
+    },
+    {
+      "epoch": 0.8204442410482455,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000549831804280365,
+      "loss": 1.1011,
+      "step": 9356
+    },
+    {
+      "epoch": 0.82053193282262,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005495955819997393,
+      "loss": 1.0913,
+      "step": 9357
+    },
+    {
+      "epoch": 0.8206196245969947,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005493594600731552,
+      "loss": 1.1645,
+      "step": 9358
+    },
+    {
+      "epoch": 0.8207073163713692,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005491234385221473,
+      "loss": 1.1237,
+      "step": 9359
+    },
+    {
+      "epoch": 0.8207950081457437,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005488875173682398,
+      "loss": 1.1263,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8208826999201183,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005486516966329485,
+      "loss": 1.1365,
+      "step": 9361
+    },
+    {
+      "epoch": 0.8209703916944928,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0005484159763377794,
+      "loss": 1.0689,
+      "step": 9362
+    },
+    {
+      "epoch": 0.8210580834688673,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005481803565042292,
+      "loss": 1.1867,
+      "step": 9363
+    },
+    {
+      "epoch": 0.821145775243242,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005479448371537863,
+      "loss": 1.1531,
+      "step": 9364
+    },
+    {
+      "epoch": 0.8212334670176165,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005477094183079292,
+      "loss": 1.0918,
+      "step": 9365
+    },
+    {
+      "epoch": 0.821321158791991,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005474740999881279,
+      "loss": 1.1724,
+      "step": 9366
+    },
+    {
+      "epoch": 0.8214088505663656,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005472388822158422,
+      "loss": 1.1108,
+      "step": 9367
+    },
+    {
+      "epoch": 0.8214965423407401,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005470037650125235,
+      "loss": 1.1216,
+      "step": 9368
+    },
+    {
+      "epoch": 0.8215842341151146,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005467687483996143,
+      "loss": 1.1639,
+      "step": 9369
+    },
+    {
+      "epoch": 0.8216719258894892,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005465338323985476,
+      "loss": 1.1134,
+      "step": 9370
+    },
+    {
+      "epoch": 0.8217596176638637,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005462990170307465,
+      "loss": 1.0993,
+      "step": 9371
+    },
+    {
+      "epoch": 0.8218473094382382,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005460643023176258,
+      "loss": 1.1033,
+      "step": 9372
+    },
+    {
+      "epoch": 0.8219350012126129,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005458296882805908,
+      "loss": 1.1338,
+      "step": 9373
+    },
+    {
+      "epoch": 0.8220226929869874,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005455951749410382,
+      "loss": 1.0877,
+      "step": 9374
+    },
+    {
+      "epoch": 0.8221103847613619,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005453607623203545,
+      "loss": 1.1239,
+      "step": 9375
+    },
+    {
+      "epoch": 0.8221980765357364,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005451264504399174,
+      "loss": 1.1959,
+      "step": 9376
+    },
+    {
+      "epoch": 0.822285768310111,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.000544892239321096,
+      "loss": 1.1579,
+      "step": 9377
+    },
+    {
+      "epoch": 0.8223734600844855,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005446581289852494,
+      "loss": 1.1079,
+      "step": 9378
+    },
+    {
+      "epoch": 0.82246115185886,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005444241194537287,
+      "loss": 1.2099,
+      "step": 9379
+    },
+    {
+      "epoch": 0.8225488436332347,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0005441902107478737,
+      "loss": 1.0921,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8226365354076092,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005439564028890169,
+      "loss": 1.0934,
+      "step": 9381
+    },
+    {
+      "epoch": 0.8227242271819837,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005437226958984808,
+      "loss": 1.1486,
+      "step": 9382
+    },
+    {
+      "epoch": 0.8228119189563583,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005434890897975791,
+      "loss": 1.1206,
+      "step": 9383
+    },
+    {
+      "epoch": 0.8228996107307328,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005432555846076162,
+      "loss": 1.1473,
+      "step": 9384
+    },
+    {
+      "epoch": 0.8229873025051073,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005430221803498867,
+      "loss": 1.0677,
+      "step": 9385
+    },
+    {
+      "epoch": 0.8230749942794819,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005427888770456766,
+      "loss": 1.1731,
+      "step": 9386
+    },
+    {
+      "epoch": 0.8231626860538565,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005425556747162627,
+      "loss": 1.1009,
+      "step": 9387
+    },
+    {
+      "epoch": 0.823250377828231,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005423225733829128,
+      "loss": 1.1717,
+      "step": 9388
+    },
+    {
+      "epoch": 0.8233380696026056,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005420895730668844,
+      "loss": 1.1596,
+      "step": 9389
+    },
+    {
+      "epoch": 0.8234257613769801,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005418566737894268,
+      "loss": 1.1548,
+      "step": 9390
+    },
+    {
+      "epoch": 0.8235134531513546,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00054162387557178,
+      "loss": 1.1612,
+      "step": 9391
+    },
+    {
+      "epoch": 0.8236011449257292,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005413911784351747,
+      "loss": 1.1693,
+      "step": 9392
+    },
+    {
+      "epoch": 0.8236888367001037,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005411585824008318,
+      "loss": 1.1328,
+      "step": 9393
+    },
+    {
+      "epoch": 0.8237765284744782,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005409260874899638,
+      "loss": 1.1164,
+      "step": 9394
+    },
+    {
+      "epoch": 0.8238642202488529,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005406936937237735,
+      "loss": 1.1757,
+      "step": 9395
+    },
+    {
+      "epoch": 0.8239519120232274,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000540461401123455,
+      "loss": 1.1175,
+      "step": 9396
+    },
+    {
+      "epoch": 0.8240396037976019,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005402292097101924,
+      "loss": 1.1401,
+      "step": 9397
+    },
+    {
+      "epoch": 0.8241272955719765,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005399971195051607,
+      "loss": 1.1603,
+      "step": 9398
+    },
+    {
+      "epoch": 0.824214987346351,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005397651305295263,
+      "loss": 1.1254,
+      "step": 9399
+    },
+    {
+      "epoch": 0.8243026791207255,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005395332428044458,
+      "loss": 1.1166,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8243903708951,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005393014563510675,
+      "loss": 1.0996,
+      "step": 9401
+    },
+    {
+      "epoch": 0.8244780626694747,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005390697711905287,
+      "loss": 1.1909,
+      "step": 9402
+    },
+    {
+      "epoch": 0.8245657544438492,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005388381873439587,
+      "loss": 1.0715,
+      "step": 9403
+    },
+    {
+      "epoch": 0.8246534462182237,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0005386067048324778,
+      "loss": 1.0782,
+      "step": 9404
+    },
+    {
+      "epoch": 0.8247411379925983,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005383753236771964,
+      "loss": 1.1912,
+      "step": 9405
+    },
+    {
+      "epoch": 0.8248288297669728,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005381440438992161,
+      "loss": 1.0922,
+      "step": 9406
+    },
+    {
+      "epoch": 0.8249165215413473,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005379128655196284,
+      "loss": 1.1881,
+      "step": 9407
+    },
+    {
+      "epoch": 0.8250042133157219,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005376817885595164,
+      "loss": 1.1163,
+      "step": 9408
+    },
+    {
+      "epoch": 0.8250919050900964,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.000537450813039954,
+      "loss": 1.164,
+      "step": 9409
+    },
+    {
+      "epoch": 0.825179596864471,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005372199389820058,
+      "loss": 1.1099,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8252672886388456,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000536989166406726,
+      "loss": 1.0769,
+      "step": 9411
+    },
+    {
+      "epoch": 0.8253549804132201,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005367584953351612,
+      "loss": 1.2007,
+      "step": 9412
+    },
+    {
+      "epoch": 0.8254426721875946,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000536527925788348,
+      "loss": 1.1401,
+      "step": 9413
+    },
+    {
+      "epoch": 0.8255303639619692,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005362974577873134,
+      "loss": 1.181,
+      "step": 9414
+    },
+    {
+      "epoch": 0.8256180557363437,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005360670913530759,
+      "loss": 1.1372,
+      "step": 9415
+    },
+    {
+      "epoch": 0.8257057475107182,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005358368265066437,
+      "loss": 1.1081,
+      "step": 9416
+    },
+    {
+      "epoch": 0.8257934392850929,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005356066632690165,
+      "loss": 1.1504,
+      "step": 9417
+    },
+    {
+      "epoch": 0.8258811310594674,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005353766016611852,
+      "loss": 1.162,
+      "step": 9418
+    },
+    {
+      "epoch": 0.8259688228338419,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005351466417041305,
+      "loss": 1.1629,
+      "step": 9419
+    },
+    {
+      "epoch": 0.8260565146082165,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005349167834188238,
+      "loss": 1.1,
+      "step": 9420
+    },
+    {
+      "epoch": 0.826144206382591,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005346870268262281,
+      "loss": 1.1229,
+      "step": 9421
+    },
+    {
+      "epoch": 0.8262318981569655,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005344573719472962,
+      "loss": 1.1627,
+      "step": 9422
+    },
+    {
+      "epoch": 0.82631958993134,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005342278188029727,
+      "loss": 1.1363,
+      "step": 9423
+    },
+    {
+      "epoch": 0.8264072817057146,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005339983674141912,
+      "loss": 1.1401,
+      "step": 9424
+    },
+    {
+      "epoch": 0.8264949734800892,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005337690178018778,
+      "loss": 1.1354,
+      "step": 9425
+    },
+    {
+      "epoch": 0.8265826652544637,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005335397699869485,
+      "loss": 1.1857,
+      "step": 9426
+    },
+    {
+      "epoch": 0.8266703570288383,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005333106239903104,
+      "loss": 1.1318,
+      "step": 9427
+    },
+    {
+      "epoch": 0.8267580488032128,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005330815798328601,
+      "loss": 1.1494,
+      "step": 9428
+    },
+    {
+      "epoch": 0.8268457405775873,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005328526375354866,
+      "loss": 1.1129,
+      "step": 9429
+    },
+    {
+      "epoch": 0.8269334323519619,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.000532623797119069,
+      "loss": 1.1157,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8270211241263364,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0005323950586044763,
+      "loss": 1.1061,
+      "step": 9431
+    },
+    {
+      "epoch": 0.827108815900711,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000532166422012569,
+      "loss": 1.111,
+      "step": 9432
+    },
+    {
+      "epoch": 0.8271965076750856,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0005319378873641986,
+      "loss": 1.0943,
+      "step": 9433
+    },
+    {
+      "epoch": 0.8272841994494601,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005317094546802066,
+      "loss": 1.0969,
+      "step": 9434
+    },
+    {
+      "epoch": 0.8273718912238346,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005314811239814253,
+      "loss": 1.1013,
+      "step": 9435
+    },
+    {
+      "epoch": 0.8274595829982092,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005312528952886781,
+      "loss": 1.1361,
+      "step": 9436
+    },
+    {
+      "epoch": 0.8275472747725837,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005310247686227791,
+      "loss": 1.1385,
+      "step": 9437
+    },
+    {
+      "epoch": 0.8276349665469582,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005307967440045324,
+      "loss": 1.1161,
+      "step": 9438
+    },
+    {
+      "epoch": 0.8277226583213328,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0005305688214547333,
+      "loss": 1.1054,
+      "step": 9439
+    },
+    {
+      "epoch": 0.8278103500957074,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0005303410009941679,
+      "loss": 1.1476,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8278980418700819,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0005301132826436131,
+      "loss": 1.096,
+      "step": 9441
+    },
+    {
+      "epoch": 0.8279857336444565,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005298856664238358,
+      "loss": 1.1903,
+      "step": 9442
+    },
+    {
+      "epoch": 0.828073425418831,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005296581523555939,
+      "loss": 1.1288,
+      "step": 9443
+    },
+    {
+      "epoch": 0.8281611171932055,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005294307404596366,
+      "loss": 1.1086,
+      "step": 9444
+    },
+    {
+      "epoch": 0.8282488089675801,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005292034307567032,
+      "loss": 1.091,
+      "step": 9445
+    },
+    {
+      "epoch": 0.8283365007419546,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005289762232675233,
+      "loss": 1.1515,
+      "step": 9446
+    },
+    {
+      "epoch": 0.8284241925163291,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005287491180128183,
+      "loss": 1.2079,
+      "step": 9447
+    },
+    {
+      "epoch": 0.8285118842907037,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005285221150132989,
+      "loss": 1.1706,
+      "step": 9448
+    },
+    {
+      "epoch": 0.8285995760650783,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005282952142896676,
+      "loss": 1.1141,
+      "step": 9449
+    },
+    {
+      "epoch": 0.8286872678394528,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005280684158626174,
+      "loss": 1.0987,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8287749596138273,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000527841719752831,
+      "loss": 1.1808,
+      "step": 9451
+    },
+    {
+      "epoch": 0.8288626513882019,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000527615125980983,
+      "loss": 1.1396,
+      "step": 9452
+    },
+    {
+      "epoch": 0.8289503431625764,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.000527388634567738,
+      "loss": 1.1436,
+      "step": 9453
+    },
+    {
+      "epoch": 0.8290380349369509,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005271622455337519,
+      "loss": 1.1633,
+      "step": 9454
+    },
+    {
+      "epoch": 0.8291257267113256,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00052693595889967,
+      "loss": 1.1088,
+      "step": 9455
+    },
+    {
+      "epoch": 0.8292134184857001,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005267097746861296,
+      "loss": 1.1379,
+      "step": 9456
+    },
+    {
+      "epoch": 0.8293011102600746,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005264836929137582,
+      "loss": 1.1337,
+      "step": 9457
+    },
+    {
+      "epoch": 0.8293888020344492,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005262577136031732,
+      "loss": 1.146,
+      "step": 9458
+    },
+    {
+      "epoch": 0.8294764938088237,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005260318367749843,
+      "loss": 1.1384,
+      "step": 9459
+    },
+    {
+      "epoch": 0.8295641855831982,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005258060624497902,
+      "loss": 1.1017,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8296518773575728,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005255803906481811,
+      "loss": 1.1231,
+      "step": 9461
+    },
+    {
+      "epoch": 0.8297395691319474,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005253548213907376,
+      "loss": 1.1503,
+      "step": 9462
+    },
+    {
+      "epoch": 0.8298272609063219,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005251293546980314,
+      "loss": 1.1401,
+      "step": 9463
+    },
+    {
+      "epoch": 0.8299149526806965,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005249039905906242,
+      "loss": 1.1423,
+      "step": 9464
+    },
+    {
+      "epoch": 0.830002644455071,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0005246787290890685,
+      "loss": 1.1809,
+      "step": 9465
+    },
+    {
+      "epoch": 0.8300903362294455,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005244535702139076,
+      "loss": 1.147,
+      "step": 9466
+    },
+    {
+      "epoch": 0.8301780280038201,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005242285139856755,
+      "loss": 1.1633,
+      "step": 9467
+    },
+    {
+      "epoch": 0.8302657197781946,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005240035604248971,
+      "loss": 1.0903,
+      "step": 9468
+    },
+    {
+      "epoch": 0.8303534115525691,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005237787095520868,
+      "loss": 1.1272,
+      "step": 9469
+    },
+    {
+      "epoch": 0.8304411033269438,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005235539613877512,
+      "loss": 1.0585,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8305287951013183,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000523329315952386,
+      "loss": 1.0976,
+      "step": 9471
+    },
+    {
+      "epoch": 0.8306164868756928,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0005231047732664793,
+      "loss": 1.1317,
+      "step": 9472
+    },
+    {
+      "epoch": 0.8307041786500673,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005228803333505078,
+      "loss": 1.1194,
+      "step": 9473
+    },
+    {
+      "epoch": 0.8307918704244419,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005226559962249403,
+      "loss": 1.2008,
+      "step": 9474
+    },
+    {
+      "epoch": 0.8308795621988164,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005224317619102357,
+      "loss": 1.1224,
+      "step": 9475
+    },
+    {
+      "epoch": 0.8309672539731909,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005222076304268439,
+      "loss": 1.1111,
+      "step": 9476
+    },
+    {
+      "epoch": 0.8310549457475656,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005219836017952044,
+      "loss": 1.1378,
+      "step": 9477
+    },
+    {
+      "epoch": 0.8311426375219401,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005217596760357485,
+      "loss": 1.1888,
+      "step": 9478
+    },
+    {
+      "epoch": 0.8312303292963146,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005215358531688978,
+      "loss": 1.1202,
+      "step": 9479
+    },
+    {
+      "epoch": 0.8313180210706892,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005213121332150644,
+      "loss": 1.1081,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8314057128450637,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005210885161946508,
+      "loss": 1.106,
+      "step": 9481
+    },
+    {
+      "epoch": 0.8314934046194382,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0005208650021280498,
+      "loss": 1.0938,
+      "step": 9482
+    },
+    {
+      "epoch": 0.8315810963938128,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005206415910356457,
+      "loss": 1.122,
+      "step": 9483
+    },
+    {
+      "epoch": 0.8316687881681873,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005204182829378132,
+      "loss": 1.1575,
+      "step": 9484
+    },
+    {
+      "epoch": 0.8317564799425619,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005201950778549177,
+      "loss": 1.0674,
+      "step": 9485
+    },
+    {
+      "epoch": 0.8318441717169365,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005199719758073141,
+      "loss": 1.1363,
+      "step": 9486
+    },
+    {
+      "epoch": 0.831931863491311,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005197489768153492,
+      "loss": 1.112,
+      "step": 9487
+    },
+    {
+      "epoch": 0.8320195552656855,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005195260808993599,
+      "loss": 1.1528,
+      "step": 9488
+    },
+    {
+      "epoch": 0.8321072470400601,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005193032880796737,
+      "loss": 1.177,
+      "step": 9489
+    },
+    {
+      "epoch": 0.8321949388144346,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005190805983766091,
+      "loss": 1.1562,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8322826305888091,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0005188580118104739,
+      "loss": 1.1115,
+      "step": 9491
+    },
+    {
+      "epoch": 0.8323703223631838,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005186355284015682,
+      "loss": 1.1875,
+      "step": 9492
+    },
+    {
+      "epoch": 0.8324580141375583,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005184131481701819,
+      "loss": 1.0686,
+      "step": 9493
+    },
+    {
+      "epoch": 0.8325457059119328,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0005181908711365953,
+      "loss": 1.1052,
+      "step": 9494
+    },
+    {
+      "epoch": 0.8326333976863073,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0005179686973210793,
+      "loss": 1.1258,
+      "step": 9495
+    },
+    {
+      "epoch": 0.8327210894606819,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005177466267438958,
+      "loss": 1.1104,
+      "step": 9496
+    },
+    {
+      "epoch": 0.8328087812350564,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005175246594252974,
+      "loss": 1.1036,
+      "step": 9497
+    },
+    {
+      "epoch": 0.8328964730094309,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005173027953855262,
+      "loss": 1.1065,
+      "step": 9498
+    },
+    {
+      "epoch": 0.8329841647838055,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005170810346448164,
+      "loss": 1.1399,
+      "step": 9499
+    },
+    {
+      "epoch": 0.8330718565581801,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0005168593772233912,
+      "loss": 1.0726,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8330718565581801,
+      "eval_loss": 1.138482689857483,
+      "eval_runtime": 428.9247,
+      "eval_samples_per_second": 33.682,
+      "eval_steps_per_second": 8.421,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8331595483325546,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0005166378231414657,
+      "loss": 1.1238,
+      "step": 9501
+    },
+    {
+      "epoch": 0.8332472401069292,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.000516416372419245,
+      "loss": 1.237,
+      "step": 9502
+    },
+    {
+      "epoch": 0.8333349318813037,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.000516195025076925,
+      "loss": 1.1987,
+      "step": 9503
+    },
+    {
+      "epoch": 0.8334226236556782,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005159737811346916,
+      "loss": 1.1143,
+      "step": 9504
+    },
+    {
+      "epoch": 0.8335103154300528,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005157526406127218,
+      "loss": 1.1351,
+      "step": 9505
+    },
+    {
+      "epoch": 0.8335980072044273,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005155316035311834,
+      "loss": 1.075,
+      "step": 9506
+    },
+    {
+      "epoch": 0.8336856989788018,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005153106699102341,
+      "loss": 1.1192,
+      "step": 9507
+    },
+    {
+      "epoch": 0.8337733907531765,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005150898397700223,
+      "loss": 1.0909,
+      "step": 9508
+    },
+    {
+      "epoch": 0.833861082527551,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005148691131306874,
+      "loss": 1.137,
+      "step": 9509
+    },
+    {
+      "epoch": 0.8339487743019255,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0005146484900123591,
+      "loss": 1.1156,
+      "step": 9510
+    },
+    {
+      "epoch": 0.8340364660763001,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005144279704351576,
+      "loss": 1.169,
+      "step": 9511
+    },
+    {
+      "epoch": 0.8341241578506746,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005142075544191941,
+      "loss": 1.1143,
+      "step": 9512
+    },
+    {
+      "epoch": 0.8342118496250491,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005139872419845691,
+      "loss": 1.1044,
+      "step": 9513
+    },
+    {
+      "epoch": 0.8342995413994237,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005137670331513755,
+      "loss": 1.1542,
+      "step": 9514
+    },
+    {
+      "epoch": 0.8343872331737983,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000513546927939695,
+      "loss": 1.1188,
+      "step": 9515
+    },
+    {
+      "epoch": 0.8344749249481728,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000513326926369601,
+      "loss": 1.1767,
+      "step": 9516
+    },
+    {
+      "epoch": 0.8345626167225474,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.000513107028461157,
+      "loss": 1.1572,
+      "step": 9517
+    },
+    {
+      "epoch": 0.8346503084969219,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000512887234234417,
+      "loss": 1.1116,
+      "step": 9518
+    },
+    {
+      "epoch": 0.8347380002712964,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000512667543709426,
+      "loss": 1.146,
+      "step": 9519
+    },
+    {
+      "epoch": 0.8348256920456709,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005124479569062187,
+      "loss": 1.1614,
+      "step": 9520
+    },
+    {
+      "epoch": 0.8349133838200455,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005122284738448217,
+      "loss": 1.1036,
+      "step": 9521
+    },
+    {
+      "epoch": 0.83500107559442,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005120090945452501,
+      "loss": 1.1164,
+      "step": 9522
+    },
+    {
+      "epoch": 0.8350887673687946,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005117898190275117,
+      "loss": 1.162,
+      "step": 9523
+    },
+    {
+      "epoch": 0.8351764591431692,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005115706473116034,
+      "loss": 1.1027,
+      "step": 9524
+    },
+    {
+      "epoch": 0.8352641509175437,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005113515794175137,
+      "loss": 1.0984,
+      "step": 9525
+    },
+    {
+      "epoch": 0.8353518426919182,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005111326153652199,
+      "loss": 1.1134,
+      "step": 9526
+    },
+    {
+      "epoch": 0.8354395344662928,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005109137551746918,
+      "loss": 1.101,
+      "step": 9527
+    },
+    {
+      "epoch": 0.8355272262406673,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0005106949988658888,
+      "loss": 1.1233,
+      "step": 9528
+    },
+    {
+      "epoch": 0.8356149180150418,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005104763464587609,
+      "loss": 1.1347,
+      "step": 9529
+    },
+    {
+      "epoch": 0.8357026097894165,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005102577979732485,
+      "loss": 1.0752,
+      "step": 9530
+    },
+    {
+      "epoch": 0.835790301563791,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005100393534292828,
+      "loss": 1.1701,
+      "step": 9531
+    },
+    {
+      "epoch": 0.8358779933381655,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005098210128467851,
+      "loss": 1.156,
+      "step": 9532
+    },
+    {
+      "epoch": 0.8359656851125401,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005096027762456675,
+      "loss": 1.209,
+      "step": 9533
+    },
+    {
+      "epoch": 0.8360533768869146,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005093846436458331,
+      "loss": 1.1422,
+      "step": 9534
+    },
+    {
+      "epoch": 0.8361410686612891,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005091666150671745,
+      "loss": 1.1057,
+      "step": 9535
+    },
+    {
+      "epoch": 0.8362287604356637,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005089486905295756,
+      "loss": 1.1676,
+      "step": 9536
+    },
+    {
+      "epoch": 0.8363164522100383,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005087308700529103,
+      "loss": 1.0809,
+      "step": 9537
+    },
+    {
+      "epoch": 0.8364041439844128,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000508513153657044,
+      "loss": 1.0877,
+      "step": 9538
+    },
+    {
+      "epoch": 0.8364918357587874,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.000508295541361831,
+      "loss": 1.1148,
+      "step": 9539
+    },
+    {
+      "epoch": 0.8365795275331619,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0005080780331871175,
+      "loss": 1.1498,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8366672193075364,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005078606291527392,
+      "loss": 1.1405,
+      "step": 9541
+    },
+    {
+      "epoch": 0.836754911081911,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005076433292785232,
+      "loss": 1.1299,
+      "step": 9542
+    },
+    {
+      "epoch": 0.8368426028562855,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005074261335842869,
+      "loss": 1.1186,
+      "step": 9543
+    },
+    {
+      "epoch": 0.83693029463066,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005072090420898375,
+      "loss": 1.1578,
+      "step": 9544
+    },
+    {
+      "epoch": 0.8370179864050346,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005069920548149734,
+      "loss": 1.1367,
+      "step": 9545
+    },
+    {
+      "epoch": 0.8371056781794092,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000506775171779483,
+      "loss": 1.1157,
+      "step": 9546
+    },
+    {
+      "epoch": 0.8371933699537837,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005065583930031461,
+      "loss": 1.0922,
+      "step": 9547
+    },
+    {
+      "epoch": 0.8372810617281582,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.000506341718505732,
+      "loss": 1.0878,
+      "step": 9548
+    },
+    {
+      "epoch": 0.8373687535025328,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005061251483070005,
+      "loss": 1.1328,
+      "step": 9549
+    },
+    {
+      "epoch": 0.8374564452769073,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005059086824267027,
+      "loss": 1.0989,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8375441370512818,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005056923208845795,
+      "loss": 1.0961,
+      "step": 9551
+    },
+    {
+      "epoch": 0.8376318288256565,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000505476063700363,
+      "loss": 1.1205,
+      "step": 9552
+    },
+    {
+      "epoch": 0.837719520600031,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0005052599108937746,
+      "loss": 1.1181,
+      "step": 9553
+    },
+    {
+      "epoch": 0.8378072123744055,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005050438624845271,
+      "loss": 1.1301,
+      "step": 9554
+    },
+    {
+      "epoch": 0.8378949041487801,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005048279184923238,
+      "loss": 1.0896,
+      "step": 9555
+    },
+    {
+      "epoch": 0.8379825959231546,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005046120789368583,
+      "loss": 1.1376,
+      "step": 9556
+    },
+    {
+      "epoch": 0.8380702876975291,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005043963438378141,
+      "loss": 1.1014,
+      "step": 9557
+    },
+    {
+      "epoch": 0.8381579794719037,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.000504180713214866,
+      "loss": 1.1675,
+      "step": 9558
+    },
+    {
+      "epoch": 0.8382456712462782,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005039651870876788,
+      "loss": 1.1099,
+      "step": 9559
+    },
+    {
+      "epoch": 0.8383333630206528,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005037497654759085,
+      "loss": 1.1203,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8384210547950274,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005035344483992002,
+      "loss": 1.1286,
+      "step": 9561
+    },
+    {
+      "epoch": 0.8385087465694019,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0005033192358771907,
+      "loss": 1.1347,
+      "step": 9562
+    },
+    {
+      "epoch": 0.8385964383437764,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.0005031041279295066,
+      "loss": 1.0812,
+      "step": 9563
+    },
+    {
+      "epoch": 0.838684130118151,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005028891245757655,
+      "loss": 1.1524,
+      "step": 9564
+    },
+    {
+      "epoch": 0.8387718218925255,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005026742258355755,
+      "loss": 1.1591,
+      "step": 9565
+    },
+    {
+      "epoch": 0.8388595136669,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005024594317285337,
+      "loss": 1.1186,
+      "step": 9566
+    },
+    {
+      "epoch": 0.8389472054412745,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0005022447422742295,
+      "loss": 1.0775,
+      "step": 9567
+    },
+    {
+      "epoch": 0.8390348972156492,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005020301574922418,
+      "loss": 1.1206,
+      "step": 9568
+    },
+    {
+      "epoch": 0.8391225889900237,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005018156774021402,
+      "loss": 1.1334,
+      "step": 9569
+    },
+    {
+      "epoch": 0.8392102807643982,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005016013020234851,
+      "loss": 1.0894,
+      "step": 9570
+    },
+    {
+      "epoch": 0.8392979725387728,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005013870313758265,
+      "loss": 1.12,
+      "step": 9571
+    },
+    {
+      "epoch": 0.8393856643131473,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005011728654787053,
+      "loss": 1.1436,
+      "step": 9572
+    },
+    {
+      "epoch": 0.8394733560875218,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005009588043516533,
+      "loss": 1.1481,
+      "step": 9573
+    },
+    {
+      "epoch": 0.8395610478618964,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005007448480141923,
+      "loss": 1.0871,
+      "step": 9574
+    },
+    {
+      "epoch": 0.839648739636271,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005005309964858342,
+      "loss": 1.095,
+      "step": 9575
+    },
+    {
+      "epoch": 0.8397364314106455,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005003172497860817,
+      "loss": 1.1169,
+      "step": 9576
+    },
+    {
+      "epoch": 0.8398241231850201,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005001036079344283,
+      "loss": 1.1642,
+      "step": 9577
+    },
+    {
+      "epoch": 0.8399118149593946,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004998900709503576,
+      "loss": 1.0762,
+      "step": 9578
+    },
+    {
+      "epoch": 0.8399995067337691,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0004996766388533432,
+      "loss": 1.1409,
+      "step": 9579
+    },
+    {
+      "epoch": 0.8400871985081437,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0004994633116628496,
+      "loss": 1.1026,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8401748902825182,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0004992500893983321,
+      "loss": 1.0856,
+      "step": 9581
+    },
+    {
+      "epoch": 0.8402625820568927,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000499036972079236,
+      "loss": 1.0976,
+      "step": 9582
+    },
+    {
+      "epoch": 0.8403502738312674,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004988239597249969,
+      "loss": 1.1184,
+      "step": 9583
+    },
+    {
+      "epoch": 0.8404379656056419,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004986110523550407,
+      "loss": 1.12,
+      "step": 9584
+    },
+    {
+      "epoch": 0.8405256573800164,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004983982499887841,
+      "loss": 1.15,
+      "step": 9585
+    },
+    {
+      "epoch": 0.840613349154391,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004981855526456343,
+      "loss": 1.1946,
+      "step": 9586
+    },
+    {
+      "epoch": 0.8407010409287655,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000497972960344989,
+      "loss": 1.1289,
+      "step": 9587
+    },
+    {
+      "epoch": 0.84078873270314,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0004977604731062356,
+      "loss": 1.1008,
+      "step": 9588
+    },
+    {
+      "epoch": 0.8408764244775147,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0004975480909487522,
+      "loss": 1.1392,
+      "step": 9589
+    },
+    {
+      "epoch": 0.8409641162518892,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0004973358138919083,
+      "loss": 1.086,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8410518080262637,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004971236419550626,
+      "loss": 1.1561,
+      "step": 9591
+    },
+    {
+      "epoch": 0.8411394998006382,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0004969115751575643,
+      "loss": 1.176,
+      "step": 9592
+    },
+    {
+      "epoch": 0.8412271915750128,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004966996135187539,
+      "loss": 1.1336,
+      "step": 9593
+    },
+    {
+      "epoch": 0.8413148833493873,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0004964877570579615,
+      "loss": 1.1391,
+      "step": 9594
+    },
+    {
+      "epoch": 0.8414025751237618,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004962760057945076,
+      "loss": 1.0273,
+      "step": 9595
+    },
+    {
+      "epoch": 0.8414902668981364,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004960643597477042,
+      "loss": 1.1679,
+      "step": 9596
+    },
+    {
+      "epoch": 0.841577958672511,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.000495852818936852,
+      "loss": 1.1532,
+      "step": 9597
+    },
+    {
+      "epoch": 0.8416656504468855,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0004956413833812432,
+      "loss": 1.1456,
+      "step": 9598
+    },
+    {
+      "epoch": 0.8417533422212601,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0004954300531001606,
+      "loss": 1.0768,
+      "step": 9599
+    },
+    {
+      "epoch": 0.8418410339956346,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0004952188281128764,
+      "loss": 1.1052,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8419287257700091,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0004950077084386542,
+      "loss": 1.0954,
+      "step": 9601
+    },
+    {
+      "epoch": 0.8420164175443837,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0004947966940967471,
+      "loss": 1.0906,
+      "step": 9602
+    },
+    {
+      "epoch": 0.8421041093187582,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004945857851063997,
+      "loss": 1.0763,
+      "step": 9603
+    },
+    {
+      "epoch": 0.8421918010931327,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004943749814868457,
+      "loss": 1.1233,
+      "step": 9604
+    },
+    {
+      "epoch": 0.8422794928675074,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004941642832573105,
+      "loss": 1.1657,
+      "step": 9605
+    },
+    {
+      "epoch": 0.8423671846418819,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004939536904370086,
+      "loss": 1.1243,
+      "step": 9606
+    },
+    {
+      "epoch": 0.8424548764162564,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0004937432030451456,
+      "loss": 1.0712,
+      "step": 9607
+    },
+    {
+      "epoch": 0.842542568190631,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.000493532821100918,
+      "loss": 1.1154,
+      "step": 9608
+    },
+    {
+      "epoch": 0.8426302599650055,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004933225446235118,
+      "loss": 1.1331,
+      "step": 9609
+    },
+    {
+      "epoch": 0.84271795173938,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004931123736321033,
+      "loss": 1.0916,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8428056435137546,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0004929023081458599,
+      "loss": 1.1041,
+      "step": 9611
+    },
+    {
+      "epoch": 0.8428933352881292,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004926923481839388,
+      "loss": 1.1601,
+      "step": 9612
+    },
+    {
+      "epoch": 0.8429810270625037,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0004924824937654885,
+      "loss": 1.106,
+      "step": 9613
+    },
+    {
+      "epoch": 0.8430687188368782,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004922727449096462,
+      "loss": 1.1532,
+      "step": 9614
+    },
+    {
+      "epoch": 0.8431564106112528,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000492063101635541,
+      "loss": 1.1599,
+      "step": 9615
+    },
+    {
+      "epoch": 0.8432441023856273,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0004918535639622921,
+      "loss": 1.043,
+      "step": 9616
+    },
+    {
+      "epoch": 0.8433317941600018,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0004916441319090079,
+      "loss": 1.1345,
+      "step": 9617
+    },
+    {
+      "epoch": 0.8434194859343764,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0004914348054947889,
+      "loss": 1.149,
+      "step": 9618
+    },
+    {
+      "epoch": 0.8435071777087509,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004912255847387248,
+      "loss": 1.1142,
+      "step": 9619
+    },
+    {
+      "epoch": 0.8435948694831255,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004910164696598958,
+      "loss": 1.1558,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8436825612575001,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004908074602773728,
+      "loss": 1.0948,
+      "step": 9621
+    },
+    {
+      "epoch": 0.8437702530318746,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0004905985566102171,
+      "loss": 1.1124,
+      "step": 9622
+    },
+    {
+      "epoch": 0.8438579448062491,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0004903897586774804,
+      "loss": 1.1841,
+      "step": 9623
+    },
+    {
+      "epoch": 0.8439456365806237,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0004901810664982038,
+      "loss": 1.1179,
+      "step": 9624
+    },
+    {
+      "epoch": 0.8440333283549982,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0004899724800914199,
+      "loss": 1.0956,
+      "step": 9625
+    },
+    {
+      "epoch": 0.8441210201293727,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0004897639994761513,
+      "loss": 1.1108,
+      "step": 9626
+    },
+    {
+      "epoch": 0.8442087119037474,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 0.000489555624671411,
+      "loss": 1.1081,
+      "step": 9627
+    },
+    {
+      "epoch": 0.8442964036781219,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004893473556962018,
+      "loss": 1.118,
+      "step": 9628
+    },
+    {
+      "epoch": 0.8443840954524964,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0004891391925695176,
+      "loss": 1.1352,
+      "step": 9629
+    },
+    {
+      "epoch": 0.844471787226871,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004889311353103422,
+      "loss": 1.1591,
+      "step": 9630
+    },
+    {
+      "epoch": 0.8445594790012455,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004887231839376503,
+      "loss": 1.1755,
+      "step": 9631
+    },
+    {
+      "epoch": 0.84464717077562,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0004885153384704059,
+      "loss": 1.1208,
+      "step": 9632
+    },
+    {
+      "epoch": 0.8447348625499946,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0004883075989275646,
+      "loss": 1.0899,
+      "step": 9633
+    },
+    {
+      "epoch": 0.8448225543243691,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004880999653280712,
+      "loss": 1.1314,
+      "step": 9634
+    },
+    {
+      "epoch": 0.8449102460987437,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00048789243769086133,
+      "loss": 1.1664,
+      "step": 9635
+    },
+    {
+      "epoch": 0.8449979378731183,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0004876850160348616,
+      "loss": 1.1229,
+      "step": 9636
+    },
+    {
+      "epoch": 0.8450856296474928,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00048747770037898744,
+      "loss": 1.1354,
+      "step": 9637
+    },
+    {
+      "epoch": 0.8451733214218673,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0004872704907421462,
+      "loss": 1.1488,
+      "step": 9638
+    },
+    {
+      "epoch": 0.8452610131962418,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00048706338714323444,
+      "loss": 1.0833,
+      "step": 9639
+    },
+    {
+      "epoch": 0.8453487049706164,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004868563896011401,
+      "loss": 1.1524,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8454363967449909,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00048664949813474,
+      "loss": 1.0742,
+      "step": 9641
+    },
+    {
+      "epoch": 0.8455240885193654,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00048644271276290253,
+      "loss": 1.1666,
+      "step": 9642
+    },
+    {
+      "epoch": 0.8456117802937401,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00048623603350448605,
+      "loss": 1.1208,
+      "step": 9643
+    },
+    {
+      "epoch": 0.8456994720681146,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00048602946037833926,
+      "loss": 1.1022,
+      "step": 9644
+    },
+    {
+      "epoch": 0.8457871638424891,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00048582299340330066,
+      "loss": 1.1217,
+      "step": 9645
+    },
+    {
+      "epoch": 0.8458748556168637,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004856166325981997,
+      "loss": 1.1677,
+      "step": 9646
+    },
+    {
+      "epoch": 0.8459625473912382,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00048541037798185597,
+      "loss": 1.1098,
+      "step": 9647
+    },
+    {
+      "epoch": 0.8460502391656127,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.00048520422957307944,
+      "loss": 1.0784,
+      "step": 9648
+    },
+    {
+      "epoch": 0.8461379309399873,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0004849981873906707,
+      "loss": 1.1096,
+      "step": 9649
+    },
+    {
+      "epoch": 0.8462256227143619,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0004847922514534197,
+      "loss": 1.1904,
+      "step": 9650
+    },
+    {
+      "epoch": 0.8463133144887364,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00048458642178010705,
+      "loss": 1.102,
+      "step": 9651
+    },
+    {
+      "epoch": 0.846401006263111,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004843806983895044,
+      "loss": 1.1791,
+      "step": 9652
+    },
+    {
+      "epoch": 0.8464886980374855,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004841750813003729,
+      "loss": 1.1488,
+      "step": 9653
+    },
+    {
+      "epoch": 0.84657638981186,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00048396957053146474,
+      "loss": 1.1663,
+      "step": 9654
+    },
+    {
+      "epoch": 0.8466640815862346,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00048376416610152134,
+      "loss": 1.0483,
+      "step": 9655
+    },
+    {
+      "epoch": 0.8467517733606091,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00048355886802927545,
+      "loss": 1.1396,
+      "step": 9656
+    },
+    {
+      "epoch": 0.8468394651349836,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00048335367633344965,
+      "loss": 1.0971,
+      "step": 9657
+    },
+    {
+      "epoch": 0.8469271569093583,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00048314859103275705,
+      "loss": 1.1176,
+      "step": 9658
+    },
+    {
+      "epoch": 0.8470148486837328,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004829436121459006,
+      "loss": 1.1512,
+      "step": 9659
+    },
+    {
+      "epoch": 0.8471025404581073,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004827387396915738,
+      "loss": 1.1656,
+      "step": 9660
+    },
+    {
+      "epoch": 0.8471902322324819,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004825339736884608,
+      "loss": 1.1271,
+      "step": 9661
+    },
+    {
+      "epoch": 0.8472779240068564,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0004823293141552358,
+      "loss": 1.0985,
+      "step": 9662
+    },
+    {
+      "epoch": 0.8473656157812309,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0004821247611105628,
+      "loss": 1.1543,
+      "step": 9663
+    },
+    {
+      "epoch": 0.8474533075556054,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004819203145730968,
+      "loss": 1.0924,
+      "step": 9664
+    },
+    {
+      "epoch": 0.8475409993299801,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004817159745614826,
+      "loss": 1.1021,
+      "step": 9665
+    },
+    {
+      "epoch": 0.8476286911043546,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.000481511741094356,
+      "loss": 1.0923,
+      "step": 9666
+    },
+    {
+      "epoch": 0.8477163828787291,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000481307614190342,
+      "loss": 1.1214,
+      "step": 9667
+    },
+    {
+      "epoch": 0.8478040746531037,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0004811035938680565,
+      "loss": 1.1401,
+      "step": 9668
+    },
+    {
+      "epoch": 0.8478917664274782,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004808996801461058,
+      "loss": 1.111,
+      "step": 9669
+    },
+    {
+      "epoch": 0.8479794582018527,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.00048069587304308607,
+      "loss": 1.1538,
+      "step": 9670
+    },
+    {
+      "epoch": 0.8480671499762273,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00048049217257758466,
+      "loss": 1.1297,
+      "step": 9671
+    },
+    {
+      "epoch": 0.8481548417506019,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0004802885787681776,
+      "loss": 1.1236,
+      "step": 9672
+    },
+    {
+      "epoch": 0.8482425335249764,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004800850916334326,
+      "loss": 1.1562,
+      "step": 9673
+    },
+    {
+      "epoch": 0.848330225299351,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00047988171119190716,
+      "loss": 1.146,
+      "step": 9674
+    },
+    {
+      "epoch": 0.8484179170737255,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00047967843746214906,
+      "loss": 1.1751,
+      "step": 9675
+    },
+    {
+      "epoch": 0.8485056088481,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004794752704626964,
+      "loss": 1.1753,
+      "step": 9676
+    },
+    {
+      "epoch": 0.8485933006224746,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.00047927221021207717,
+      "loss": 1.0936,
+      "step": 9677
+    },
+    {
+      "epoch": 0.8486809923968491,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00047906925672881024,
+      "loss": 1.2118,
+      "step": 9678
+    },
+    {
+      "epoch": 0.8487686841712236,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004788664100314043,
+      "loss": 1.1683,
+      "step": 9679
+    },
+    {
+      "epoch": 0.8488563759455983,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0004786636701383587,
+      "loss": 1.1196,
+      "step": 9680
+    },
+    {
+      "epoch": 0.8489440677199728,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004784610370681625,
+      "loss": 1.1122,
+      "step": 9681
+    },
+    {
+      "epoch": 0.8490317594943473,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004782585108392953,
+      "loss": 1.1635,
+      "step": 9682
+    },
+    {
+      "epoch": 0.8491194512687219,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000478056091470227,
+      "loss": 1.0639,
+      "step": 9683
+    },
+    {
+      "epoch": 0.8492071430430964,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004778537789794182,
+      "loss": 1.0846,
+      "step": 9684
+    },
+    {
+      "epoch": 0.8492948348174709,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004776515733853189,
+      "loss": 1.1454,
+      "step": 9685
+    },
+    {
+      "epoch": 0.8493825265918454,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0004774494747063694,
+      "loss": 1.1479,
+      "step": 9686
+    },
+    {
+      "epoch": 0.84947021836622,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00047724748296100076,
+      "loss": 1.1241,
+      "step": 9687
+    },
+    {
+      "epoch": 0.8495579101405946,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 0.00047704559816763435,
+      "loss": 1.0953,
+      "step": 9688
+    },
+    {
+      "epoch": 0.8496456019149691,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.00047684382034468163,
+      "loss": 1.0851,
+      "step": 9689
+    },
+    {
+      "epoch": 0.8497332936893437,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004766421495105437,
+      "loss": 1.1509,
+      "step": 9690
+    },
+    {
+      "epoch": 0.8498209854637182,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004764405856836128,
+      "loss": 1.1505,
+      "step": 9691
+    },
+    {
+      "epoch": 0.8499086772380927,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004762391288822709,
+      "loss": 1.1223,
+      "step": 9692
+    },
+    {
+      "epoch": 0.8499963690124673,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004760377791248908,
+      "loss": 1.1802,
+      "step": 9693
+    },
+    {
+      "epoch": 0.8500840607868418,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004758365364298344,
+      "loss": 1.1412,
+      "step": 9694
+    },
+    {
+      "epoch": 0.8501717525612164,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00047563540081545463,
+      "loss": 1.1682,
+      "step": 9695
+    },
+    {
+      "epoch": 0.850259444335591,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0004754343723000947,
+      "loss": 1.1476,
+      "step": 9696
+    },
+    {
+      "epoch": 0.8503471361099655,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0004752334509020883,
+      "loss": 1.0911,
+      "step": 9697
+    },
+    {
+      "epoch": 0.85043482788434,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004750326366397583,
+      "loss": 1.1304,
+      "step": 9698
+    },
+    {
+      "epoch": 0.8505225196587146,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004748319295314185,
+      "loss": 1.0771,
+      "step": 9699
+    },
+    {
+      "epoch": 0.8506102114330891,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004746313295953731,
+      "loss": 1.1032,
+      "step": 9700
+    },
+    {
+      "epoch": 0.8506979032074636,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004744308368499166,
+      "loss": 1.085,
+      "step": 9701
+    },
+    {
+      "epoch": 0.8507855949818383,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00047423045131333297,
+      "loss": 1.1626,
+      "step": 9702
+    },
+    {
+      "epoch": 0.8508732867562128,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004740301730038968,
+      "loss": 1.1094,
+      "step": 9703
+    },
+    {
+      "epoch": 0.8509609785305873,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004738300019398729,
+      "loss": 1.2059,
+      "step": 9704
+    },
+    {
+      "epoch": 0.8510486703049619,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004736299381395167,
+      "loss": 1.1351,
+      "step": 9705
+    },
+    {
+      "epoch": 0.8511363620793364,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004734299816210734,
+      "loss": 1.0825,
+      "step": 9706
+    },
+    {
+      "epoch": 0.8512240538537109,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004732301324027786,
+      "loss": 1.1388,
+      "step": 9707
+    },
+    {
+      "epoch": 0.8513117456280855,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004730303905028578,
+      "loss": 1.1918,
+      "step": 9708
+    },
+    {
+      "epoch": 0.85139943740246,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00047283075593952683,
+      "loss": 1.1522,
+      "step": 9709
+    },
+    {
+      "epoch": 0.8514871291768346,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00047263122873099226,
+      "loss": 1.1861,
+      "step": 9710
+    },
+    {
+      "epoch": 0.8515748209512091,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004724318088954505,
+      "loss": 1.0792,
+      "step": 9711
+    },
+    {
+      "epoch": 0.8516625127255837,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004722324964510876,
+      "loss": 1.0793,
+      "step": 9712
+    },
+    {
+      "epoch": 0.8517502044999582,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004720332914160805,
+      "loss": 1.1432,
+      "step": 9713
+    },
+    {
+      "epoch": 0.8518378962743327,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00047183419380859653,
+      "loss": 1.103,
+      "step": 9714
+    },
+    {
+      "epoch": 0.8519255880487073,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004716352036467929,
+      "loss": 1.2073,
+      "step": 9715
+    },
+    {
+      "epoch": 0.8520132798230818,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 0.00047143632094881634,
+      "loss": 1.1112,
+      "step": 9716
+    },
+    {
+      "epoch": 0.8521009715974563,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00047123754573280503,
+      "loss": 1.1106,
+      "step": 9717
+    },
+    {
+      "epoch": 0.852188663371831,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.00047103887801688686,
+      "loss": 1.156,
+      "step": 9718
+    },
+    {
+      "epoch": 0.8522763551462055,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00047084031781917927,
+      "loss": 1.099,
+      "step": 9719
+    },
+    {
+      "epoch": 0.85236404692058,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004706418651577909,
+      "loss": 1.0966,
+      "step": 9720
+    },
+    {
+      "epoch": 0.8524517386949546,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0004704435200508196,
+      "loss": 1.1417,
+      "step": 9721
+    },
+    {
+      "epoch": 0.8525394304693291,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0004702452825163544,
+      "loss": 1.0499,
+      "step": 9722
+    },
+    {
+      "epoch": 0.8526271222437036,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004700471525724738,
+      "loss": 1.1095,
+      "step": 9723
+    },
+    {
+      "epoch": 0.8527148140180782,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004698491302372473,
+      "loss": 1.1664,
+      "step": 9724
+    },
+    {
+      "epoch": 0.8528025057924528,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004696512155287332,
+      "loss": 1.1363,
+      "step": 9725
+    },
+    {
+      "epoch": 0.8528901975668273,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004694534084649812,
+      "loss": 1.153,
+      "step": 9726
+    },
+    {
+      "epoch": 0.8529778893412019,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00046925570906403065,
+      "loss": 1.1675,
+      "step": 9727
+    },
+    {
+      "epoch": 0.8530655811155764,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004690581173439115,
+      "loss": 1.1347,
+      "step": 9728
+    },
+    {
+      "epoch": 0.8531532728899509,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0004688606333226437,
+      "loss": 1.1007,
+      "step": 9729
+    },
+    {
+      "epoch": 0.8532409646643255,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00046866325701823685,
+      "loss": 1.1525,
+      "step": 9730
+    },
+    {
+      "epoch": 0.8533286564387,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004684659884486912,
+      "loss": 1.1732,
+      "step": 9731
+    },
+    {
+      "epoch": 0.8534163482130745,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004682688276319974,
+      "loss": 1.1699,
+      "step": 9732
+    },
+    {
+      "epoch": 0.8535040399874492,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0004680717745861361,
+      "loss": 1.1483,
+      "step": 9733
+    },
+    {
+      "epoch": 0.8535917317618237,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00046787482932907765,
+      "loss": 1.0927,
+      "step": 9734
+    },
+    {
+      "epoch": 0.8536794235361982,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.00046767799187878343,
+      "loss": 1.1192,
+      "step": 9735
+    },
+    {
+      "epoch": 0.8537671153105727,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004674812622532038,
+      "loss": 1.1248,
+      "step": 9736
+    },
+    {
+      "epoch": 0.8538548070849473,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.00046728464047028043,
+      "loss": 1.1261,
+      "step": 9737
+    },
+    {
+      "epoch": 0.8539424988593218,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00046708812654794486,
+      "loss": 1.1831,
+      "step": 9738
+    },
+    {
+      "epoch": 0.8540301906336963,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0004668917205041183,
+      "loss": 1.082,
+      "step": 9739
+    },
+    {
+      "epoch": 0.854117882408071,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004666954223567128,
+      "loss": 1.0797,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8542055741824455,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00046649923212362994,
+      "loss": 1.1097,
+      "step": 9741
+    },
+    {
+      "epoch": 0.85429326595682,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00046630314982276216,
+      "loss": 1.0978,
+      "step": 9742
+    },
+    {
+      "epoch": 0.8543809577311946,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0004661071754719912,
+      "loss": 1.1392,
+      "step": 9743
+    },
+    {
+      "epoch": 0.8544686495055691,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004659113090891897,
+      "loss": 1.0998,
+      "step": 9744
+    },
+    {
+      "epoch": 0.8545563412799436,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004657155506922199,
+      "loss": 1.1715,
+      "step": 9745
+    },
+    {
+      "epoch": 0.8546440330543182,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004655199002989351,
+      "loss": 1.1198,
+      "step": 9746
+    },
+    {
+      "epoch": 0.8547317248286928,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00046532435792717746,
+      "loss": 1.1117,
+      "step": 9747
+    },
+    {
+      "epoch": 0.8548194166030673,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00046512892359478,
+      "loss": 1.1627,
+      "step": 9748
+    },
+    {
+      "epoch": 0.8549071083774419,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004649335973195661,
+      "loss": 1.0906,
+      "step": 9749
+    },
+    {
+      "epoch": 0.8549948001518164,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0004647383791193491,
+      "loss": 1.1372,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8550824919261909,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004645432690119319,
+      "loss": 1.1458,
+      "step": 9751
+    },
+    {
+      "epoch": 0.8551701837005655,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004643482670151086,
+      "loss": 1.1657,
+      "step": 9752
+    },
+    {
+      "epoch": 0.85525787547494,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00046415337314666236,
+      "loss": 1.095,
+      "step": 9753
+    },
+    {
+      "epoch": 0.8553455672493145,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004639585874243672,
+      "loss": 1.0849,
+      "step": 9754
+    },
+    {
+      "epoch": 0.8554332590236892,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004637639098659875,
+      "loss": 1.1733,
+      "step": 9755
+    },
+    {
+      "epoch": 0.8555209507980637,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004635693404892765,
+      "loss": 1.0902,
+      "step": 9756
+    },
+    {
+      "epoch": 0.8556086425724382,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004633748793119791,
+      "loss": 1.1505,
+      "step": 9757
+    },
+    {
+      "epoch": 0.8556963343468127,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00046318052635182936,
+      "loss": 1.1375,
+      "step": 9758
+    },
+    {
+      "epoch": 0.8557840261211873,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00046298628162655186,
+      "loss": 1.151,
+      "step": 9759
+    },
+    {
+      "epoch": 0.8558717178955618,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0004627921451538613,
+      "loss": 1.1007,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8559594096699363,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.00046259811695146234,
+      "loss": 1.1091,
+      "step": 9761
+    },
+    {
+      "epoch": 0.856047101444311,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0004624041970370497,
+      "loss": 1.1137,
+      "step": 9762
+    },
+    {
+      "epoch": 0.8561347932186855,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00046221038542830863,
+      "loss": 1.1804,
+      "step": 9763
+    },
+    {
+      "epoch": 0.85622248499306,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004620166821429143,
+      "loss": 1.0772,
+      "step": 9764
+    },
+    {
+      "epoch": 0.8563101767674346,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00046182308719853175,
+      "loss": 1.1436,
+      "step": 9765
+    },
+    {
+      "epoch": 0.8563978685418091,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004616296006128162,
+      "loss": 1.1759,
+      "step": 9766
+    },
+    {
+      "epoch": 0.8564855603161836,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004614362224034133,
+      "loss": 1.1921,
+      "step": 9767
+    },
+    {
+      "epoch": 0.8565732520905582,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0004612429525879591,
+      "loss": 1.1232,
+      "step": 9768
+    },
+    {
+      "epoch": 0.8566609438649327,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00046104979118407864,
+      "loss": 1.1529,
+      "step": 9769
+    },
+    {
+      "epoch": 0.8567486356393073,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004608567382093878,
+      "loss": 1.1025,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8568363274136819,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00046066379368149264,
+      "loss": 1.1908,
+      "step": 9771
+    },
+    {
+      "epoch": 0.8569240191880564,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0004604709576179894,
+      "loss": 1.1124,
+      "step": 9772
+    },
+    {
+      "epoch": 0.8570117109624309,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00046027823003646416,
+      "loss": 1.1596,
+      "step": 9773
+    },
+    {
+      "epoch": 0.8570994027368055,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004600856109544928,
+      "loss": 1.0672,
+      "step": 9774
+    },
+    {
+      "epoch": 0.85718709451118,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.00045989310038964203,
+      "loss": 1.1121,
+      "step": 9775
+    },
+    {
+      "epoch": 0.8572747862855545,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00045970069835946824,
+      "loss": 1.1125,
+      "step": 9776
+    },
+    {
+      "epoch": 0.8573624780599292,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00045950840488151836,
+      "loss": 1.1799,
+      "step": 9777
+    },
+    {
+      "epoch": 0.8574501698343037,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00045931621997332843,
+      "loss": 1.0867,
+      "step": 9778
+    },
+    {
+      "epoch": 0.8575378616086782,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00045912414365242545,
+      "loss": 1.1374,
+      "step": 9779
+    },
+    {
+      "epoch": 0.8576255533830528,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00045893217593632633,
+      "loss": 1.0963,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8577132451574273,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00045874031684253817,
+      "loss": 1.0821,
+      "step": 9781
+    },
+    {
+      "epoch": 0.8578009369318018,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.000458548566388558,
+      "loss": 1.0814,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8578886287061763,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00045835692459187267,
+      "loss": 1.1456,
+      "step": 9783
+    },
+    {
+      "epoch": 0.857976320480551,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00045816539146995965,
+      "loss": 1.1459,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8580640122549255,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004579739670402864,
+      "loss": 1.1463,
+      "step": 9785
+    },
+    {
+      "epoch": 0.8581517040293,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.00045778265132030986,
+      "loss": 1.1535,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8582393958036746,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0004575914443274782,
+      "loss": 1.084,
+      "step": 9787
+    },
+    {
+      "epoch": 0.8583270875780491,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.00045740034607922827,
+      "loss": 1.1102,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8584147793524236,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00045720935659298824,
+      "loss": 1.102,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8585024711267982,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0004570184758861755,
+      "loss": 1.1132,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8585901629011727,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004568277039761985,
+      "loss": 1.1168,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8586778546755472,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0004566370408804545,
+      "loss": 1.0892,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8587655464499219,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004564464866163317,
+      "loss": 1.093,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8588532382242964,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004562560412012082,
+      "loss": 1.1019,
+      "step": 9794
+    },
+    {
+      "epoch": 0.8589409299986709,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0004560657046524523,
+      "loss": 1.1642,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8590286217730455,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.00045587547698742174,
+      "loss": 1.1228,
+      "step": 9796
+    },
+    {
+      "epoch": 0.85911631354742,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00045568535822346505,
+      "loss": 1.148,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8592040053217945,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 0.0004554953483779208,
+      "loss": 1.1241,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8592916970961691,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00045530544746811717,
+      "loss": 1.1176,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8593793888705437,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004551156555113725,
+      "loss": 1.1434,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8594670806449182,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004549259725249955,
+      "loss": 1.101,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8595547724192928,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004547363985262849,
+      "loss": 1.1217,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8596424641936673,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00045454693353252907,
+      "loss": 1.1048,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8597301559680418,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.00045435757756100696,
+      "loss": 1.1397,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8598178477424163,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000454168330628987,
+      "loss": 1.112,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8599055395167909,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00045397919275372826,
+      "loss": 1.1584,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8599932312911654,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00045379016395247977,
+      "loss": 1.1453,
+      "step": 9807
+    },
+    {
+      "epoch": 0.86008092306554,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0004536012442424805,
+      "loss": 1.1543,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8601686148399146,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.00045341243364095896,
+      "loss": 1.1633,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8602563066142891,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00045322373216513456,
+      "loss": 1.1127,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8603439983886636,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 0.00045303513983221635,
+      "loss": 1.0812,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8604316901630382,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004528466566594034,
+      "loss": 1.1023,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8605193819374127,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004526582826638852,
+      "loss": 1.1241,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8606070737117872,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0004524700178628405,
+      "loss": 1.1551,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8606947654861619,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00045228186227343865,
+      "loss": 1.132,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8607824572605364,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004520938159128392,
+      "loss": 1.137,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8608701490349109,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00045190587879819145,
+      "loss": 1.1322,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8609578408092855,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004517180509466344,
+      "loss": 1.112,
+      "step": 9818
+    },
+    {
+      "epoch": 0.86104553258366,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00045153033237529814,
+      "loss": 1.1188,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8611332243580345,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00045134272310130147,
+      "loss": 1.137,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8612209161324091,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00045115522314175406,
+      "loss": 1.1229,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8613086079067837,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00045096783251375594,
+      "loss": 1.1675,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8613962996811582,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004507805512343958,
+      "loss": 1.096,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8614839914555328,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004505933793207536,
+      "loss": 1.1094,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8615716832299073,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00045040631678989904,
+      "loss": 1.0766,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8616593750042818,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004502193636588918,
+      "loss": 1.1525,
+      "step": 9826
+    },
+    {
+      "epoch": 0.8617470667786564,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0004500325199447812,
+      "loss": 1.1101,
+      "step": 9827
+    },
+    {
+      "epoch": 0.8618347585530309,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 0.0004498457856646071,
+      "loss": 1.0606,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8619224503274054,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.000449659160835399,
+      "loss": 1.1601,
+      "step": 9829
+    },
+    {
+      "epoch": 0.86201014210178,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00044947264547417733,
+      "loss": 1.1422,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8620978338761546,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004492862395979508,
+      "loss": 1.0959,
+      "step": 9831
+    },
+    {
+      "epoch": 0.8621855256505291,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00044909994322371957,
+      "loss": 1.1758,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8622732174249036,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00044891375636847344,
+      "loss": 1.142,
+      "step": 9833
+    },
+    {
+      "epoch": 0.8623609091992782,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004487276790491925,
+      "loss": 1.1204,
+      "step": 9834
+    },
+    {
+      "epoch": 0.8624486009736527,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00044854171128284567,
+      "loss": 1.1317,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8625362927480272,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004483558530863936,
+      "loss": 1.1388,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8626239845224019,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0004481701044767855,
+      "loss": 1.0911,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8627116762967764,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00044798446547096143,
+      "loss": 1.1727,
+      "step": 9838
+    },
+    {
+      "epoch": 0.8627993680711509,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00044779893608585104,
+      "loss": 1.0929,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8628870598455255,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0004476135163383745,
+      "loss": 1.1326,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8629747516199,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000447428206245441,
+      "loss": 1.1008,
+      "step": 9841
+    },
+    {
+      "epoch": 0.8630624433942745,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00044724300582395053,
+      "loss": 1.1546,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8631501351686491,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00044705791509079334,
+      "loss": 1.15,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8632378269430236,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.000446872934062849,
+      "loss": 1.1296,
+      "step": 9844
+    },
+    {
+      "epoch": 0.8633255187173982,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0004466880627569869,
+      "loss": 1.1438,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8634132104917728,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00044650330119006725,
+      "loss": 1.1049,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8635009022661473,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004463186493789398,
+      "loss": 1.1744,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8635885940405218,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00044613410734044436,
+      "loss": 1.105,
+      "step": 9848
+    },
+    {
+      "epoch": 0.8636762858148964,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00044594967509141025,
+      "loss": 1.118,
+      "step": 9849
+    },
+    {
+      "epoch": 0.8637639775892709,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00044576535264865776,
+      "loss": 1.1352,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8638516693636454,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00044558114002899606,
+      "loss": 1.0946,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8639393611380201,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004453970372492257,
+      "loss": 1.1595,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8640270529123946,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00044521304432613594,
+      "loss": 1.0893,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8641147446867691,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00044502916127650593,
+      "loss": 1.1185,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8642024364611436,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0004448453881171059,
+      "loss": 1.124,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8642901282355182,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004446617248646954,
+      "loss": 1.0964,
+      "step": 9856
+    },
+    {
+      "epoch": 0.8643778200098927,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0004444781715360243,
+      "loss": 1.1098,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8644655117842672,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00044429472814783155,
+      "loss": 1.1513,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8645532035586418,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004441113947168471,
+      "loss": 1.0964,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8646408953330164,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004439281712597906,
+      "loss": 1.1486,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8647285871073909,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00044374505779337154,
+      "loss": 1.1763,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8648162788817655,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000443562054334289,
+      "loss": 1.1796,
+      "step": 9862
+    },
+    {
+      "epoch": 0.86490397065614,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004433791608992328,
+      "loss": 1.1487,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8649916624305145,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004431963775048823,
+      "loss": 1.0923,
+      "step": 9864
+    },
+    {
+      "epoch": 0.8650793542048891,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00044301370416790684,
+      "loss": 1.1753,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8651670459792636,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00044283114090496584,
+      "loss": 1.1221,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8652547377536381,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004426486877327082,
+      "loss": 1.0957,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8653424295280128,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004424663446677736,
+      "loss": 1.1225,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8654301213023873,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0004422841117267911,
+      "loss": 1.0965,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8655178130767618,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00044210198892638027,
+      "loss": 1.171,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8656055048511364,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0004419199762831498,
+      "loss": 1.1254,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8656931966255109,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0004417380738136985,
+      "loss": 1.1511,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8657808883998854,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004415562815346159,
+      "loss": 1.1471,
+      "step": 9873
+    },
+    {
+      "epoch": 0.86586858017426,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004413745994624807,
+      "loss": 1.1009,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8659562719486346,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004411930276138624,
+      "loss": 1.1512,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8660439637230091,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.00044101156600531916,
+      "loss": 1.0685,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8661316554973836,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004408302146534001,
+      "loss": 1.1239,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8662193472717582,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004406489735746439,
+      "loss": 1.1116,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8663070390461327,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00044046784278557976,
+      "loss": 1.1166,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8663947308205072,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00044028682230272577,
+      "loss": 1.1129,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8664824225948818,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004401059121425905,
+      "loss": 1.101,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8665701143692564,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0004399251123216728,
+      "loss": 1.1271,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8666578061436309,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004397444228564615,
+      "loss": 1.1343,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8667454979180055,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004395638437634341,
+      "loss": 1.0843,
+      "step": 9884
+    },
+    {
+      "epoch": 0.86683318969238,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004393833750590596,
+      "loss": 1.113,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8669208814667545,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00043920301675979596,
+      "loss": 1.1092,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8670085732411291,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00043902276888209196,
+      "loss": 1.127,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8670962650155036,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00043884263144238514,
+      "loss": 1.1439,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8671839567898781,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00043866260445710355,
+      "loss": 1.1047,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8672716485642528,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004384826879426653,
+      "loss": 1.1111,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8673593403386273,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00043830288191547864,
+      "loss": 1.0754,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8674470321130018,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00043812318639194084,
+      "loss": 1.1161,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8675347238873764,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00043794360138844045,
+      "loss": 1.1981,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8676224156617509,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004377641269213545,
+      "loss": 1.1418,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8677101074361254,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00043758476300705064,
+      "loss": 1.1135,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8677977992105,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004374055096618868,
+      "loss": 1.0697,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8678854909848746,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00043722636690221025,
+      "loss": 1.1285,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8679731827592491,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00043704733474435797,
+      "loss": 1.1349,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8680608745336237,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00043686841320465764,
+      "loss": 1.1651,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8681485663079982,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00043668960229942635,
+      "loss": 1.1073,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8682362580823727,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004365109020449715,
+      "loss": 1.1426,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8683239498567472,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004363323124575895,
+      "loss": 1.1544,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8684116416311218,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00043615383355356737,
+      "loss": 1.0973,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8684993334054963,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0004359754653491825,
+      "loss": 1.1119,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8685870251798709,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.00043579720786070096,
+      "loss": 1.1324,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8686747169542455,
+      "grad_norm": 0.041015625,
+      "learning_rate": 0.0004356190611043798,
+      "loss": 1.1055,
+      "step": 9906
+    },
+    {
+      "epoch": 0.86876240872862,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00043544102509646506,
+      "loss": 1.1625,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8688501005029945,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00043526309985319355,
+      "loss": 1.1062,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8689377922773691,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00043508528539079144,
+      "loss": 1.1309,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8690254840517436,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00043490758172547525,
+      "loss": 1.1133,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8691131758261181,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004347299888734507,
+      "loss": 1.0884,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8692008676004928,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.000434552506850914,
+      "loss": 1.118,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8692885593748673,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00043437513567405085,
+      "loss": 1.1172,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8693762511492418,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00043419787535903763,
+      "loss": 1.0849,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8694639429236164,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004340207259220393,
+      "loss": 1.1097,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8695516346979909,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004338436873792116,
+      "loss": 1.1156,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8696393264723654,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00043366675974670026,
+      "loss": 1.1471,
+      "step": 9917
+    },
+    {
+      "epoch": 0.86972701824674,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.00043348994304064054,
+      "loss": 1.1024,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8698147100211145,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004333132372771577,
+      "loss": 1.1382,
+      "step": 9919
+    },
+    {
+      "epoch": 0.869902401795489,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.00043313664247236663,
+      "loss": 1.1343,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8699900935698637,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00043296015864237265,
+      "loss": 1.1069,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8700777853442382,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00043278378580327006,
+      "loss": 1.1137,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8701654771186127,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0004326075239711441,
+      "loss": 1.1276,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8702531688929873,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0004324313731620695,
+      "loss": 1.162,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8703408606673618,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004322553333921103,
+      "loss": 1.1152,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8704285524417363,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004320794046773212,
+      "loss": 1.1247,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8705162442161108,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0004319035870337462,
+      "loss": 1.1184,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8706039359904855,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00043172788047741984,
+      "loss": 1.1429,
+      "step": 9928
+    },
+    {
+      "epoch": 0.87069162776486,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004315522850243657,
+      "loss": 1.1238,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8707793195392345,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004313768006905975,
+      "loss": 1.1444,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8708670113136091,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0004312014274921195,
+      "loss": 1.1181,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8709547030879836,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000431026165444925,
+      "loss": 1.1409,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8710423948623581,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004308510145649974,
+      "loss": 1.1649,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8711300866367327,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00043067597486830993,
+      "loss": 1.1273,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8712177784111073,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000430501046370826,
+      "loss": 1.1104,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8713054701854818,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00043032622908849874,
+      "loss": 1.1337,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8713931619598564,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004301515230372704,
+      "loss": 1.0986,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8714808537342309,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004299769282330745,
+      "loss": 1.152,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8715685455086054,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004298024446918331,
+      "loss": 1.154,
+      "step": 9939
+    },
+    {
+      "epoch": 0.87165623728298,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004296280724294587,
+      "loss": 1.1612,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8717439290573545,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 0.00042945381146185414,
+      "loss": 1.1055,
+      "step": 9941
+    },
+    {
+      "epoch": 0.871831620831729,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004292796618049107,
+      "loss": 1.1523,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8719193126061037,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00042910562347451083,
+      "loss": 1.1656,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8720070043804782,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004289316964865265,
+      "loss": 1.1781,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8720946961548527,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004287578808568191,
+      "loss": 1.2174,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8721823879292273,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0004285841766012408,
+      "loss": 1.1036,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8722700797036018,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.0004284105837356322,
+      "loss": 1.1524,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8723577714779763,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00042823710227582495,
+      "loss": 1.0991,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8724454632523508,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00042806373223763984,
+      "loss": 1.1533,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8725331550267255,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00042789047363688833,
+      "loss": 1.1264,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8726208468011,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00042771732648937035,
+      "loss": 1.0901,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8727085385754745,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004275442908108771,
+      "loss": 1.2054,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8727962303498491,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00042737136661718855,
+      "loss": 1.09,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8728839221242236,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0004271985539240755,
+      "loss": 1.1582,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8729716138985981,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0004270258527472979,
+      "loss": 1.0873,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8730593056729727,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0004268532631026049,
+      "loss": 1.12,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8731469974473473,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00042668078500573677,
+      "loss": 1.1726,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8732346892217218,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00042650841847242325,
+      "loss": 1.1801,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8733223809960964,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00042633616351838365,
+      "loss": 1.1355,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8734100727704709,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004261640201593269,
+      "loss": 1.1412,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8734977645448454,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00042599198841095223,
+      "loss": 1.1516,
+      "step": 9961
+    },
+    {
+      "epoch": 0.87358545631922,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0004258200682889482,
+      "loss": 1.1193,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8736731480935945,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0004256482598089943,
+      "loss": 1.1439,
+      "step": 9963
+    },
+    {
+      "epoch": 0.873760839867969,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00042547656298675806,
+      "loss": 1.1135,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8738485316423437,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00042530497783789826,
+      "loss": 1.0825,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8739362234167182,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0004251335043780628,
+      "loss": 1.0954,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8740239151910927,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004249621426228902,
+      "loss": 1.0629,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8741116069654673,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0004247908925880073,
+      "loss": 1.1624,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8741992987398418,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004246197542890323,
+      "loss": 1.1488,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8742869905142163,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00042444872774157226,
+      "loss": 1.1421,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8743746822885909,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004242778129612247,
+      "loss": 1.1597,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8744623740629655,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004241070099635768,
+      "loss": 1.1248,
+      "step": 9972
+    },
+    {
+      "epoch": 0.87455006583734,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004239363187642045,
+      "loss": 1.0638,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8746377576117145,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004237657393786749,
+      "loss": 1.135,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8747254493860891,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00042359527182254426,
+      "loss": 1.1566,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8748131411604636,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.00042342491611135896,
+      "loss": 1.1713,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8749008329348381,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00042325467226065505,
+      "loss": 1.1344,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8749885247092127,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0004230845402859581,
+      "loss": 1.1154,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8750762164835872,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0004229145202027839,
+      "loss": 1.1036,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8751639082579618,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.00042274461202663745,
+      "loss": 1.1386,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8752516000323364,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0004225748157730148,
+      "loss": 1.1012,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8753392918067109,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.00042240513145739986,
+      "loss": 1.1533,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8754269835810854,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0004222355590952681,
+      "loss": 1.1253,
+      "step": 9983
+    },
+    {
+      "epoch": 0.87551467535546,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0004220660987020838,
+      "loss": 1.0936,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8756023671298345,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0004218967502933016,
+      "loss": 1.0833,
+      "step": 9985
+    },
+    {
+      "epoch": 0.875690058904209,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00042172751388436534,
+      "loss": 1.1584,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8757777506785837,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000421558389490709,
+      "loss": 1.1314,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8758654424529582,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0004213893771277562,
+      "loss": 1.1364,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8759531342273327,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00042122047681092097,
+      "loss": 1.1391,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8760408260017073,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00042105168855560617,
+      "loss": 1.1301,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8761285177760818,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00042088301237720456,
+      "loss": 1.1312,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8762162095504563,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00042071444829109927,
+      "loss": 1.1921,
+      "step": 9992
+    },
+    {
+      "epoch": 0.8763039013248309,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00042054599631266283,
+      "loss": 1.0855,
+      "step": 9993
+    },
+    {
+      "epoch": 0.8763915930992054,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00042037765645725817,
+      "loss": 1.1294,
+      "step": 9994
+    },
+    {
+      "epoch": 0.87647928487358,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 0.0004202094287402365,
+      "loss": 1.0818,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8765669766479545,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00042004131317694034,
+      "loss": 1.1676,
+      "step": 9996
+    },
+    {
+      "epoch": 0.8766546684223291,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00041987330978270125,
+      "loss": 1.1169,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8767423601967036,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0004197054185728409,
+      "loss": 1.09,
+      "step": 9998
+    },
+    {
+      "epoch": 0.8768300519710781,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00041953763956267045,
+      "loss": 1.0852,
+      "step": 9999
+    },
+    {
+      "epoch": 0.8769177437454527,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004193699727674907,
+      "loss": 1.1104,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8769177437454527,
+      "eval_loss": 1.1368882656097412,
+      "eval_runtime": 439.2361,
+      "eval_samples_per_second": 32.891,
+      "eval_steps_per_second": 8.223,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11403,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.948783792128e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}