diff --git "a/full/checkpoint-13464/trainer_state.json" "b/full/checkpoint-13464/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/full/checkpoint-13464/trainer_state.json"
@@ -0,0 +1,18877 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999628653124883,
+  "eval_steps": 500,
+  "global_step": 13464,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003713468751160459,
+      "grad_norm": 24.08954620361328,
+      "learning_rate": 3.094059405940594e-07,
+      "loss": 1.1527,
+      "step": 5
+    },
+    {
+      "epoch": 0.0007426937502320918,
+      "grad_norm": 18.06231689453125,
+      "learning_rate": 6.188118811881188e-07,
+      "loss": 1.1166,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011140406253481376,
+      "grad_norm": 20.671451568603516,
+      "learning_rate": 9.282178217821782e-07,
+      "loss": 1.0603,
+      "step": 15
+    },
+    {
+      "epoch": 0.0014853875004641837,
+      "grad_norm": 6.015936374664307,
+      "learning_rate": 1.2376237623762377e-06,
+      "loss": 0.9175,
+      "step": 20
+    },
+    {
+      "epoch": 0.0018567343755802295,
+      "grad_norm": 3.1561789512634277,
+      "learning_rate": 1.5470297029702971e-06,
+      "loss": 0.8301,
+      "step": 25
+    },
+    {
+      "epoch": 0.002228081250696275,
+      "grad_norm": 2.571364641189575,
+      "learning_rate": 1.8564356435643564e-06,
+      "loss": 0.747,
+      "step": 30
+    },
+    {
+      "epoch": 0.0025994281258123215,
+      "grad_norm": 2.3243155479431152,
+      "learning_rate": 2.1658415841584156e-06,
+      "loss": 0.6899,
+      "step": 35
+    },
+    {
+      "epoch": 0.0029707750009283673,
+      "grad_norm": 2.2084057331085205,
+      "learning_rate": 2.4752475247524753e-06,
+      "loss": 0.692,
+      "step": 40
+    },
+    {
+      "epoch": 0.003342121876044413,
+      "grad_norm": 2.460026979446411,
+      "learning_rate": 2.784653465346535e-06,
+      "loss": 0.6368,
+      "step": 45
+    },
+    {
+      "epoch": 0.003713468751160459,
+      "grad_norm": 2.309441566467285,
+      "learning_rate": 3.0940594059405943e-06,
+      "loss": 0.5975,
+      "step": 50
+    },
+    {
+      "epoch": 0.004084815626276505,
+      "grad_norm": 2.027996301651001,
+      "learning_rate": 3.403465346534654e-06,
+      "loss": 0.6295,
+      "step": 55
+    },
+    {
+      "epoch": 0.00445616250139255,
+      "grad_norm": 1.7956230640411377,
+      "learning_rate": 3.7128712871287128e-06,
+      "loss": 0.5895,
+      "step": 60
+    },
+    {
+      "epoch": 0.004827509376508597,
+      "grad_norm": 1.8671722412109375,
+      "learning_rate": 4.0222772277227725e-06,
+      "loss": 0.5851,
+      "step": 65
+    },
+    {
+      "epoch": 0.005198856251624643,
+      "grad_norm": 1.6895778179168701,
+      "learning_rate": 4.331683168316831e-06,
+      "loss": 0.5727,
+      "step": 70
+    },
+    {
+      "epoch": 0.005570203126740688,
+      "grad_norm": 1.6208025217056274,
+      "learning_rate": 4.641089108910891e-06,
+      "loss": 0.562,
+      "step": 75
+    },
+    {
+      "epoch": 0.005941550001856735,
+      "grad_norm": 1.7942975759506226,
+      "learning_rate": 4.950495049504951e-06,
+      "loss": 0.579,
+      "step": 80
+    },
+    {
+      "epoch": 0.00631289687697278,
+      "grad_norm": 1.7987550497055054,
+      "learning_rate": 5.25990099009901e-06,
+      "loss": 0.5846,
+      "step": 85
+    },
+    {
+      "epoch": 0.006684243752088826,
+      "grad_norm": 1.887704849243164,
+      "learning_rate": 5.56930693069307e-06,
+      "loss": 0.5673,
+      "step": 90
+    },
+    {
+      "epoch": 0.007055590627204872,
+      "grad_norm": 2.0477488040924072,
+      "learning_rate": 5.878712871287129e-06,
+      "loss": 0.5562,
+      "step": 95
+    },
+    {
+      "epoch": 0.007426937502320918,
+      "grad_norm": 2.3760573863983154,
+      "learning_rate": 6.1881188118811885e-06,
+      "loss": 0.5614,
+      "step": 100
+    },
+    {
+      "epoch": 0.007798284377436964,
+      "grad_norm": 2.2995216846466064,
+      "learning_rate": 6.497524752475248e-06,
+      "loss": 0.5513,
+      "step": 105
+    },
+    {
+      "epoch": 0.00816963125255301,
+      "grad_norm": 2.010840654373169,
+      "learning_rate": 6.806930693069308e-06,
+      "loss": 0.5673,
+      "step": 110
+    },
+    {
+      "epoch": 0.008540978127669055,
+      "grad_norm": 1.9840739965438843,
+      "learning_rate": 7.116336633663366e-06,
+      "loss": 0.5649,
+      "step": 115
+    },
+    {
+      "epoch": 0.0089123250027851,
+      "grad_norm": 2.1075551509857178,
+      "learning_rate": 7.4257425742574256e-06,
+      "loss": 0.545,
+      "step": 120
+    },
+    {
+      "epoch": 0.009283671877901148,
+      "grad_norm": 1.8998537063598633,
+      "learning_rate": 7.735148514851485e-06,
+      "loss": 0.5478,
+      "step": 125
+    },
+    {
+      "epoch": 0.009655018753017193,
+      "grad_norm": 1.6922427415847778,
+      "learning_rate": 8.044554455445545e-06,
+      "loss": 0.5526,
+      "step": 130
+    },
+    {
+      "epoch": 0.010026365628133239,
+      "grad_norm": 1.8979038000106812,
+      "learning_rate": 8.353960396039605e-06,
+      "loss": 0.5557,
+      "step": 135
+    },
+    {
+      "epoch": 0.010397712503249286,
+      "grad_norm": 1.866027593612671,
+      "learning_rate": 8.663366336633663e-06,
+      "loss": 0.535,
+      "step": 140
+    },
+    {
+      "epoch": 0.010769059378365331,
+      "grad_norm": 2.00050950050354,
+      "learning_rate": 8.972772277227722e-06,
+      "loss": 0.5548,
+      "step": 145
+    },
+    {
+      "epoch": 0.011140406253481377,
+      "grad_norm": 1.6254732608795166,
+      "learning_rate": 9.282178217821782e-06,
+      "loss": 0.5264,
+      "step": 150
+    },
+    {
+      "epoch": 0.011511753128597422,
+      "grad_norm": 1.7823671102523804,
+      "learning_rate": 9.591584158415842e-06,
+      "loss": 0.554,
+      "step": 155
+    },
+    {
+      "epoch": 0.01188310000371347,
+      "grad_norm": 1.4978529214859009,
+      "learning_rate": 9.900990099009901e-06,
+      "loss": 0.5328,
+      "step": 160
+    },
+    {
+      "epoch": 0.012254446878829515,
+      "grad_norm": 1.9153281450271606,
+      "learning_rate": 1.0210396039603961e-05,
+      "loss": 0.5349,
+      "step": 165
+    },
+    {
+      "epoch": 0.01262579375394556,
+      "grad_norm": 1.415780782699585,
+      "learning_rate": 1.051980198019802e-05,
+      "loss": 0.5395,
+      "step": 170
+    },
+    {
+      "epoch": 0.012997140629061606,
+      "grad_norm": 1.6551861763000488,
+      "learning_rate": 1.082920792079208e-05,
+      "loss": 0.5429,
+      "step": 175
+    },
+    {
+      "epoch": 0.013368487504177653,
+      "grad_norm": 1.8297767639160156,
+      "learning_rate": 1.113861386138614e-05,
+      "loss": 0.536,
+      "step": 180
+    },
+    {
+      "epoch": 0.013739834379293698,
+      "grad_norm": 1.7368695735931396,
+      "learning_rate": 1.14480198019802e-05,
+      "loss": 0.5343,
+      "step": 185
+    },
+    {
+      "epoch": 0.014111181254409744,
+      "grad_norm": 1.7614681720733643,
+      "learning_rate": 1.1757425742574258e-05,
+      "loss": 0.532,
+      "step": 190
+    },
+    {
+      "epoch": 0.01448252812952579,
+      "grad_norm": 1.63772714138031,
+      "learning_rate": 1.2066831683168317e-05,
+      "loss": 0.5602,
+      "step": 195
+    },
+    {
+      "epoch": 0.014853875004641836,
+      "grad_norm": 1.5544772148132324,
+      "learning_rate": 1.2376237623762377e-05,
+      "loss": 0.5387,
+      "step": 200
+    },
+    {
+      "epoch": 0.015225221879757882,
+      "grad_norm": 1.7135066986083984,
+      "learning_rate": 1.2685643564356437e-05,
+      "loss": 0.532,
+      "step": 205
+    },
+    {
+      "epoch": 0.015596568754873927,
+      "grad_norm": 1.9367787837982178,
+      "learning_rate": 1.2995049504950496e-05,
+      "loss": 0.5353,
+      "step": 210
+    },
+    {
+      "epoch": 0.015967915629989973,
+      "grad_norm": 1.8191226720809937,
+      "learning_rate": 1.3304455445544556e-05,
+      "loss": 0.53,
+      "step": 215
+    },
+    {
+      "epoch": 0.01633926250510602,
+      "grad_norm": 1.752224087715149,
+      "learning_rate": 1.3613861386138616e-05,
+      "loss": 0.5535,
+      "step": 220
+    },
+    {
+      "epoch": 0.016710609380222067,
+      "grad_norm": 1.6079806089401245,
+      "learning_rate": 1.3923267326732675e-05,
+      "loss": 0.5679,
+      "step": 225
+    },
+    {
+      "epoch": 0.01708195625533811,
+      "grad_norm": 1.429037094116211,
+      "learning_rate": 1.4232673267326732e-05,
+      "loss": 0.5179,
+      "step": 230
+    },
+    {
+      "epoch": 0.017453303130454158,
+      "grad_norm": 1.5963634252548218,
+      "learning_rate": 1.4542079207920791e-05,
+      "loss": 0.5224,
+      "step": 235
+    },
+    {
+      "epoch": 0.0178246500055702,
+      "grad_norm": 2.0120062828063965,
+      "learning_rate": 1.4851485148514851e-05,
+      "loss": 0.5274,
+      "step": 240
+    },
+    {
+      "epoch": 0.01819599688068625,
+      "grad_norm": 1.9131743907928467,
+      "learning_rate": 1.516089108910891e-05,
+      "loss": 0.5365,
+      "step": 245
+    },
+    {
+      "epoch": 0.018567343755802296,
+      "grad_norm": 1.6677278280258179,
+      "learning_rate": 1.547029702970297e-05,
+      "loss": 0.5378,
+      "step": 250
+    },
+    {
+      "epoch": 0.01893869063091834,
+      "grad_norm": 1.643923282623291,
+      "learning_rate": 1.577970297029703e-05,
+      "loss": 0.5427,
+      "step": 255
+    },
+    {
+      "epoch": 0.019310037506034387,
+      "grad_norm": 1.5207113027572632,
+      "learning_rate": 1.608910891089109e-05,
+      "loss": 0.543,
+      "step": 260
+    },
+    {
+      "epoch": 0.019681384381150434,
+      "grad_norm": 1.5055848360061646,
+      "learning_rate": 1.639851485148515e-05,
+      "loss": 0.5276,
+      "step": 265
+    },
+    {
+      "epoch": 0.020052731256266478,
+      "grad_norm": 1.609848141670227,
+      "learning_rate": 1.670792079207921e-05,
+      "loss": 0.5295,
+      "step": 270
+    },
+    {
+      "epoch": 0.020424078131382525,
+      "grad_norm": 1.8507963418960571,
+      "learning_rate": 1.701732673267327e-05,
+      "loss": 0.5399,
+      "step": 275
+    },
+    {
+      "epoch": 0.020795425006498572,
+      "grad_norm": 1.620498538017273,
+      "learning_rate": 1.7326732673267325e-05,
+      "loss": 0.5126,
+      "step": 280
+    },
+    {
+      "epoch": 0.021166771881614616,
+      "grad_norm": 1.9648469686508179,
+      "learning_rate": 1.7636138613861385e-05,
+      "loss": 0.5409,
+      "step": 285
+    },
+    {
+      "epoch": 0.021538118756730663,
+      "grad_norm": 1.8357899188995361,
+      "learning_rate": 1.7945544554455445e-05,
+      "loss": 0.5291,
+      "step": 290
+    },
+    {
+      "epoch": 0.021909465631846706,
+      "grad_norm": 1.7807576656341553,
+      "learning_rate": 1.8254950495049504e-05,
+      "loss": 0.5415,
+      "step": 295
+    },
+    {
+      "epoch": 0.022280812506962754,
+      "grad_norm": 1.6266732215881348,
+      "learning_rate": 1.8564356435643564e-05,
+      "loss": 0.5351,
+      "step": 300
+    },
+    {
+      "epoch": 0.0226521593820788,
+      "grad_norm": 1.704314947128296,
+      "learning_rate": 1.8873762376237624e-05,
+      "loss": 0.5197,
+      "step": 305
+    },
+    {
+      "epoch": 0.023023506257194844,
+      "grad_norm": 1.6191048622131348,
+      "learning_rate": 1.9183168316831683e-05,
+      "loss": 0.5324,
+      "step": 310
+    },
+    {
+      "epoch": 0.02339485313231089,
+      "grad_norm": 1.5764744281768799,
+      "learning_rate": 1.9492574257425743e-05,
+      "loss": 0.5155,
+      "step": 315
+    },
+    {
+      "epoch": 0.02376620000742694,
+      "grad_norm": 1.7684473991394043,
+      "learning_rate": 1.9801980198019803e-05,
+      "loss": 0.5385,
+      "step": 320
+    },
+    {
+      "epoch": 0.024137546882542982,
+      "grad_norm": 1.5683865547180176,
+      "learning_rate": 2.0111386138613862e-05,
+      "loss": 0.5161,
+      "step": 325
+    },
+    {
+      "epoch": 0.02450889375765903,
+      "grad_norm": 1.4367133378982544,
+      "learning_rate": 2.0420792079207922e-05,
+      "loss": 0.5179,
+      "step": 330
+    },
+    {
+      "epoch": 0.024880240632775077,
+      "grad_norm": 1.755611538887024,
+      "learning_rate": 2.073019801980198e-05,
+      "loss": 0.5299,
+      "step": 335
+    },
+    {
+      "epoch": 0.02525158750789112,
+      "grad_norm": 1.6052712202072144,
+      "learning_rate": 2.103960396039604e-05,
+      "loss": 0.5318,
+      "step": 340
+    },
+    {
+      "epoch": 0.025622934383007168,
+      "grad_norm": 1.6675814390182495,
+      "learning_rate": 2.13490099009901e-05,
+      "loss": 0.5473,
+      "step": 345
+    },
+    {
+      "epoch": 0.02599428125812321,
+      "grad_norm": 1.628082036972046,
+      "learning_rate": 2.165841584158416e-05,
+      "loss": 0.5245,
+      "step": 350
+    },
+    {
+      "epoch": 0.02636562813323926,
+      "grad_norm": 1.5258337259292603,
+      "learning_rate": 2.196782178217822e-05,
+      "loss": 0.5231,
+      "step": 355
+    },
+    {
+      "epoch": 0.026736975008355306,
+      "grad_norm": 1.8839592933654785,
+      "learning_rate": 2.227722772277228e-05,
+      "loss": 0.5127,
+      "step": 360
+    },
+    {
+      "epoch": 0.02710832188347135,
+      "grad_norm": 1.5799691677093506,
+      "learning_rate": 2.258663366336634e-05,
+      "loss": 0.5396,
+      "step": 365
+    },
+    {
+      "epoch": 0.027479668758587397,
+      "grad_norm": 1.3579293489456177,
+      "learning_rate": 2.28960396039604e-05,
+      "loss": 0.5211,
+      "step": 370
+    },
+    {
+      "epoch": 0.027851015633703444,
+      "grad_norm": 1.8711540699005127,
+      "learning_rate": 2.320544554455446e-05,
+      "loss": 0.5274,
+      "step": 375
+    },
+    {
+      "epoch": 0.028222362508819487,
+      "grad_norm": 1.3717608451843262,
+      "learning_rate": 2.3514851485148515e-05,
+      "loss": 0.5333,
+      "step": 380
+    },
+    {
+      "epoch": 0.028593709383935535,
+      "grad_norm": 1.8865386247634888,
+      "learning_rate": 2.3824257425742575e-05,
+      "loss": 0.5255,
+      "step": 385
+    },
+    {
+      "epoch": 0.02896505625905158,
+      "grad_norm": 1.7165244817733765,
+      "learning_rate": 2.4133663366336635e-05,
+      "loss": 0.5313,
+      "step": 390
+    },
+    {
+      "epoch": 0.029336403134167625,
+      "grad_norm": 1.7152857780456543,
+      "learning_rate": 2.4443069306930694e-05,
+      "loss": 0.5368,
+      "step": 395
+    },
+    {
+      "epoch": 0.029707750009283673,
+      "grad_norm": 1.3566906452178955,
+      "learning_rate": 2.4752475247524754e-05,
+      "loss": 0.5333,
+      "step": 400
+    },
+    {
+      "epoch": 0.030079096884399716,
+      "grad_norm": 1.4016120433807373,
+      "learning_rate": 2.5061881188118814e-05,
+      "loss": 0.4903,
+      "step": 405
+    },
+    {
+      "epoch": 0.030450443759515763,
+      "grad_norm": 1.3507463932037354,
+      "learning_rate": 2.5371287128712873e-05,
+      "loss": 0.5223,
+      "step": 410
+    },
+    {
+      "epoch": 0.03082179063463181,
+      "grad_norm": 1.451769232749939,
+      "learning_rate": 2.5680693069306933e-05,
+      "loss": 0.5069,
+      "step": 415
+    },
+    {
+      "epoch": 0.031193137509747854,
+      "grad_norm": 1.7525213956832886,
+      "learning_rate": 2.5990099009900993e-05,
+      "loss": 0.5088,
+      "step": 420
+    },
+    {
+      "epoch": 0.0315644843848639,
+      "grad_norm": 1.8567416667938232,
+      "learning_rate": 2.6299504950495053e-05,
+      "loss": 0.5191,
+      "step": 425
+    },
+    {
+      "epoch": 0.031935831259979945,
+      "grad_norm": 1.7284778356552124,
+      "learning_rate": 2.6608910891089112e-05,
+      "loss": 0.5325,
+      "step": 430
+    },
+    {
+      "epoch": 0.032307178135095996,
+      "grad_norm": 1.5363997220993042,
+      "learning_rate": 2.6918316831683172e-05,
+      "loss": 0.5364,
+      "step": 435
+    },
+    {
+      "epoch": 0.03267852501021204,
+      "grad_norm": 1.3674659729003906,
+      "learning_rate": 2.722772277227723e-05,
+      "loss": 0.519,
+      "step": 440
+    },
+    {
+      "epoch": 0.03304987188532808,
+      "grad_norm": 1.5923279523849487,
+      "learning_rate": 2.753712871287129e-05,
+      "loss": 0.5007,
+      "step": 445
+    },
+    {
+      "epoch": 0.033421218760444134,
+      "grad_norm": 1.5102838277816772,
+      "learning_rate": 2.784653465346535e-05,
+      "loss": 0.5195,
+      "step": 450
+    },
+    {
+      "epoch": 0.03379256563556018,
+      "grad_norm": 1.2837610244750977,
+      "learning_rate": 2.8155940594059404e-05,
+      "loss": 0.5125,
+      "step": 455
+    },
+    {
+      "epoch": 0.03416391251067622,
+      "grad_norm": 1.592606782913208,
+      "learning_rate": 2.8465346534653464e-05,
+      "loss": 0.516,
+      "step": 460
+    },
+    {
+      "epoch": 0.03453525938579227,
+      "grad_norm": 1.7333269119262695,
+      "learning_rate": 2.8774752475247523e-05,
+      "loss": 0.5314,
+      "step": 465
+    },
+    {
+      "epoch": 0.034906606260908316,
+      "grad_norm": 1.5925847291946411,
+      "learning_rate": 2.9084158415841583e-05,
+      "loss": 0.5196,
+      "step": 470
+    },
+    {
+      "epoch": 0.03527795313602436,
+      "grad_norm": 1.5458823442459106,
+      "learning_rate": 2.9393564356435643e-05,
+      "loss": 0.5271,
+      "step": 475
+    },
+    {
+      "epoch": 0.0356493000111404,
+      "grad_norm": 1.2263989448547363,
+      "learning_rate": 2.9702970297029702e-05,
+      "loss": 0.5217,
+      "step": 480
+    },
+    {
+      "epoch": 0.036020646886256454,
+      "grad_norm": 1.3630470037460327,
+      "learning_rate": 3.0012376237623762e-05,
+      "loss": 0.5026,
+      "step": 485
+    },
+    {
+      "epoch": 0.0363919937613725,
+      "grad_norm": 1.3696426153182983,
+      "learning_rate": 3.032178217821782e-05,
+      "loss": 0.5317,
+      "step": 490
+    },
+    {
+      "epoch": 0.03676334063648854,
+      "grad_norm": 1.7865773439407349,
+      "learning_rate": 3.063118811881188e-05,
+      "loss": 0.5163,
+      "step": 495
+    },
+    {
+      "epoch": 0.03713468751160459,
+      "grad_norm": 1.4488468170166016,
+      "learning_rate": 3.094059405940594e-05,
+      "loss": 0.4999,
+      "step": 500
+    },
+    {
+      "epoch": 0.037506034386720635,
+      "grad_norm": 1.3591450452804565,
+      "learning_rate": 3.125e-05,
+      "loss": 0.5287,
+      "step": 505
+    },
+    {
+      "epoch": 0.03787738126183668,
+      "grad_norm": 1.3478190898895264,
+      "learning_rate": 3.155940594059406e-05,
+      "loss": 0.5325,
+      "step": 510
+    },
+    {
+      "epoch": 0.03824872813695273,
+      "grad_norm": 1.257960319519043,
+      "learning_rate": 3.186881188118812e-05,
+      "loss": 0.5381,
+      "step": 515
+    },
+    {
+      "epoch": 0.03862007501206877,
+      "grad_norm": 1.758743166923523,
+      "learning_rate": 3.217821782178218e-05,
+      "loss": 0.5584,
+      "step": 520
+    },
+    {
+      "epoch": 0.03899142188718482,
+      "grad_norm": 1.5418094396591187,
+      "learning_rate": 3.248762376237624e-05,
+      "loss": 0.5378,
+      "step": 525
+    },
+    {
+      "epoch": 0.03936276876230087,
+      "grad_norm": 3.621204376220703,
+      "learning_rate": 3.27970297029703e-05,
+      "loss": 0.5402,
+      "step": 530
+    },
+    {
+      "epoch": 0.03973411563741691,
+      "grad_norm": 1.2378004789352417,
+      "learning_rate": 3.310643564356436e-05,
+      "loss": 0.5234,
+      "step": 535
+    },
+    {
+      "epoch": 0.040105462512532955,
+      "grad_norm": 1.4689031839370728,
+      "learning_rate": 3.341584158415842e-05,
+      "loss": 0.5305,
+      "step": 540
+    },
+    {
+      "epoch": 0.040476809387649006,
+      "grad_norm": 1.4394994974136353,
+      "learning_rate": 3.372524752475248e-05,
+      "loss": 0.5413,
+      "step": 545
+    },
+    {
+      "epoch": 0.04084815626276505,
+      "grad_norm": 1.2873084545135498,
+      "learning_rate": 3.403465346534654e-05,
+      "loss": 0.563,
+      "step": 550
+    },
+    {
+      "epoch": 0.04121950313788109,
+      "grad_norm": 1.5888473987579346,
+      "learning_rate": 3.43440594059406e-05,
+      "loss": 0.5219,
+      "step": 555
+    },
+    {
+      "epoch": 0.041590850012997144,
+      "grad_norm": 1.2506392002105713,
+      "learning_rate": 3.465346534653465e-05,
+      "loss": 0.5213,
+      "step": 560
+    },
+    {
+      "epoch": 0.04196219688811319,
+      "grad_norm": 1.3455543518066406,
+      "learning_rate": 3.496287128712871e-05,
+      "loss": 0.5112,
+      "step": 565
+    },
+    {
+      "epoch": 0.04233354376322923,
+      "grad_norm": 1.3935933113098145,
+      "learning_rate": 3.527227722772277e-05,
+      "loss": 0.5205,
+      "step": 570
+    },
+    {
+      "epoch": 0.04270489063834528,
+      "grad_norm": 1.4280612468719482,
+      "learning_rate": 3.558168316831683e-05,
+      "loss": 0.5188,
+      "step": 575
+    },
+    {
+      "epoch": 0.043076237513461325,
+      "grad_norm": 1.0775330066680908,
+      "learning_rate": 3.589108910891089e-05,
+      "loss": 0.5062,
+      "step": 580
+    },
+    {
+      "epoch": 0.04344758438857737,
+      "grad_norm": 1.1990082263946533,
+      "learning_rate": 3.620049504950495e-05,
+      "loss": 0.5318,
+      "step": 585
+    },
+    {
+      "epoch": 0.04381893126369341,
+      "grad_norm": 1.9189587831497192,
+      "learning_rate": 3.650990099009901e-05,
+      "loss": 0.5046,
+      "step": 590
+    },
+    {
+      "epoch": 0.04419027813880946,
+      "grad_norm": 1.3169859647750854,
+      "learning_rate": 3.681930693069307e-05,
+      "loss": 0.5214,
+      "step": 595
+    },
+    {
+      "epoch": 0.04456162501392551,
+      "grad_norm": 1.3218770027160645,
+      "learning_rate": 3.712871287128713e-05,
+      "loss": 0.5586,
+      "step": 600
+    },
+    {
+      "epoch": 0.04493297188904155,
+      "grad_norm": 1.255031704902649,
+      "learning_rate": 3.743811881188119e-05,
+      "loss": 0.5327,
+      "step": 605
+    },
+    {
+      "epoch": 0.0453043187641576,
+      "grad_norm": 1.5894232988357544,
+      "learning_rate": 3.774752475247525e-05,
+      "loss": 0.5205,
+      "step": 610
+    },
+    {
+      "epoch": 0.045675665639273645,
+      "grad_norm": 1.4050710201263428,
+      "learning_rate": 3.805693069306931e-05,
+      "loss": 0.5393,
+      "step": 615
+    },
+    {
+      "epoch": 0.04604701251438969,
+      "grad_norm": 1.4929163455963135,
+      "learning_rate": 3.8366336633663367e-05,
+      "loss": 0.5024,
+      "step": 620
+    },
+    {
+      "epoch": 0.04641835938950574,
+      "grad_norm": 1.336452841758728,
+      "learning_rate": 3.8675742574257426e-05,
+      "loss": 0.5212,
+      "step": 625
+    },
+    {
+      "epoch": 0.04678970626462178,
+      "grad_norm": 1.2623363733291626,
+      "learning_rate": 3.8985148514851486e-05,
+      "loss": 0.5135,
+      "step": 630
+    },
+    {
+      "epoch": 0.04716105313973783,
+      "grad_norm": 1.336139440536499,
+      "learning_rate": 3.9294554455445546e-05,
+      "loss": 0.5288,
+      "step": 635
+    },
+    {
+      "epoch": 0.04753240001485388,
+      "grad_norm": 1.4025272130966187,
+      "learning_rate": 3.9603960396039605e-05,
+      "loss": 0.5192,
+      "step": 640
+    },
+    {
+      "epoch": 0.04790374688996992,
+      "grad_norm": 1.3830080032348633,
+      "learning_rate": 3.9913366336633665e-05,
+      "loss": 0.5229,
+      "step": 645
+    },
+    {
+      "epoch": 0.048275093765085965,
+      "grad_norm": 1.624871850013733,
+      "learning_rate": 4.0222772277227725e-05,
+      "loss": 0.5078,
+      "step": 650
+    },
+    {
+      "epoch": 0.048646440640202016,
+      "grad_norm": 1.2770441770553589,
+      "learning_rate": 4.0532178217821784e-05,
+      "loss": 0.5115,
+      "step": 655
+    },
+    {
+      "epoch": 0.04901778751531806,
+      "grad_norm": 1.5466053485870361,
+      "learning_rate": 4.0841584158415844e-05,
+      "loss": 0.5047,
+      "step": 660
+    },
+    {
+      "epoch": 0.0493891343904341,
+      "grad_norm": 1.1815299987792969,
+      "learning_rate": 4.1150990099009904e-05,
+      "loss": 0.5196,
+      "step": 665
+    },
+    {
+      "epoch": 0.049760481265550154,
+      "grad_norm": 1.5421866178512573,
+      "learning_rate": 4.146039603960396e-05,
+      "loss": 0.4923,
+      "step": 670
+    },
+    {
+      "epoch": 0.0501318281406662,
+      "grad_norm": 1.328102707862854,
+      "learning_rate": 4.176980198019802e-05,
+      "loss": 0.5092,
+      "step": 675
+    },
+    {
+      "epoch": 0.05050317501578224,
+      "grad_norm": 1.354792594909668,
+      "learning_rate": 4.207920792079208e-05,
+      "loss": 0.5084,
+      "step": 680
+    },
+    {
+      "epoch": 0.050874521890898285,
+      "grad_norm": 1.2935703992843628,
+      "learning_rate": 4.238861386138614e-05,
+      "loss": 0.5262,
+      "step": 685
+    },
+    {
+      "epoch": 0.051245868766014335,
+      "grad_norm": 1.5065386295318604,
+      "learning_rate": 4.26980198019802e-05,
+      "loss": 0.5239,
+      "step": 690
+    },
+    {
+      "epoch": 0.05161721564113038,
+      "grad_norm": 1.410735845565796,
+      "learning_rate": 4.300742574257426e-05,
+      "loss": 0.5291,
+      "step": 695
+    },
+    {
+      "epoch": 0.05198856251624642,
+      "grad_norm": 1.1437729597091675,
+      "learning_rate": 4.331683168316832e-05,
+      "loss": 0.5144,
+      "step": 700
+    },
+    {
+      "epoch": 0.05235990939136247,
+      "grad_norm": 1.7676151990890503,
+      "learning_rate": 4.362623762376238e-05,
+      "loss": 0.5213,
+      "step": 705
+    },
+    {
+      "epoch": 0.05273125626647852,
+      "grad_norm": 1.1369547843933105,
+      "learning_rate": 4.393564356435644e-05,
+      "loss": 0.5246,
+      "step": 710
+    },
+    {
+      "epoch": 0.05310260314159456,
+      "grad_norm": 1.37111496925354,
+      "learning_rate": 4.42450495049505e-05,
+      "loss": 0.5051,
+      "step": 715
+    },
+    {
+      "epoch": 0.05347395001671061,
+      "grad_norm": 1.3732452392578125,
+      "learning_rate": 4.455445544554456e-05,
+      "loss": 0.5288,
+      "step": 720
+    },
+    {
+      "epoch": 0.053845296891826655,
+      "grad_norm": 1.937808632850647,
+      "learning_rate": 4.486386138613862e-05,
+      "loss": 0.5245,
+      "step": 725
+    },
+    {
+      "epoch": 0.0542166437669427,
+      "grad_norm": 1.255505084991455,
+      "learning_rate": 4.517326732673268e-05,
+      "loss": 0.5184,
+      "step": 730
+    },
+    {
+      "epoch": 0.05458799064205875,
+      "grad_norm": 1.2280532121658325,
+      "learning_rate": 4.548267326732674e-05,
+      "loss": 0.5227,
+      "step": 735
+    },
+    {
+      "epoch": 0.05495933751717479,
+      "grad_norm": 1.4528734683990479,
+      "learning_rate": 4.57920792079208e-05,
+      "loss": 0.5056,
+      "step": 740
+    },
+    {
+      "epoch": 0.05533068439229084,
+      "grad_norm": 1.3214168548583984,
+      "learning_rate": 4.610148514851486e-05,
+      "loss": 0.5231,
+      "step": 745
+    },
+    {
+      "epoch": 0.05570203126740689,
+      "grad_norm": 1.2361901998519897,
+      "learning_rate": 4.641089108910892e-05,
+      "loss": 0.5375,
+      "step": 750
+    },
+    {
+      "epoch": 0.05607337814252293,
+      "grad_norm": 1.4549564123153687,
+      "learning_rate": 4.672029702970298e-05,
+      "loss": 0.5249,
+      "step": 755
+    },
+    {
+      "epoch": 0.056444725017638975,
+      "grad_norm": 1.2583074569702148,
+      "learning_rate": 4.702970297029703e-05,
+      "loss": 0.522,
+      "step": 760
+    },
+    {
+      "epoch": 0.056816071892755025,
+      "grad_norm": 1.1710424423217773,
+      "learning_rate": 4.733910891089109e-05,
+      "loss": 0.4936,
+      "step": 765
+    },
+    {
+      "epoch": 0.05718741876787107,
+      "grad_norm": 1.3571052551269531,
+      "learning_rate": 4.764851485148515e-05,
+      "loss": 0.5418,
+      "step": 770
+    },
+    {
+      "epoch": 0.05755876564298711,
+      "grad_norm": 1.2731226682662964,
+      "learning_rate": 4.795792079207921e-05,
+      "loss": 0.5191,
+      "step": 775
+    },
+    {
+      "epoch": 0.05793011251810316,
+      "grad_norm": 1.5696614980697632,
+      "learning_rate": 4.826732673267327e-05,
+      "loss": 0.5168,
+      "step": 780
+    },
+    {
+      "epoch": 0.05830145939321921,
+      "grad_norm": 1.2827826738357544,
+      "learning_rate": 4.857673267326733e-05,
+      "loss": 0.512,
+      "step": 785
+    },
+    {
+      "epoch": 0.05867280626833525,
+      "grad_norm": 2.312535524368286,
+      "learning_rate": 4.888613861386139e-05,
+      "loss": 0.5248,
+      "step": 790
+    },
+    {
+      "epoch": 0.059044153143451295,
+      "grad_norm": 1.1396629810333252,
+      "learning_rate": 4.919554455445545e-05,
+      "loss": 0.5347,
+      "step": 795
+    },
+    {
+      "epoch": 0.059415500018567345,
+      "grad_norm": 1.181564211845398,
+      "learning_rate": 4.950495049504951e-05,
+      "loss": 0.5062,
+      "step": 800
+    },
+    {
+      "epoch": 0.05978684689368339,
+      "grad_norm": 1.0706778764724731,
+      "learning_rate": 4.981435643564357e-05,
+      "loss": 0.5154,
+      "step": 805
+    },
+    {
+      "epoch": 0.06015819376879943,
+      "grad_norm": 1.4956251382827759,
+      "learning_rate": 4.9999999276691775e-05,
+      "loss": 0.5154,
+      "step": 810
+    },
+    {
+      "epoch": 0.06052954064391548,
+      "grad_norm": 1.0542585849761963,
+      "learning_rate": 4.9999991139474754e-05,
+      "loss": 0.5341,
+      "step": 815
+    },
+    {
+      "epoch": 0.06090088751903153,
+      "grad_norm": 1.3105566501617432,
+      "learning_rate": 4.999997396090837e-05,
+      "loss": 0.5366,
+      "step": 820
+    },
+    {
+      "epoch": 0.06127223439414757,
+      "grad_norm": 1.2557010650634766,
+      "learning_rate": 4.9999947740998846e-05,
+      "loss": 0.5292,
+      "step": 825
+    },
+    {
+      "epoch": 0.06164358126926362,
+      "grad_norm": 1.1770594120025635,
+      "learning_rate": 4.999991247975567e-05,
+      "loss": 0.5266,
+      "step": 830
+    },
+    {
+      "epoch": 0.062014928144379665,
+      "grad_norm": 1.2295761108398438,
+      "learning_rate": 4.9999868177191576e-05,
+      "loss": 0.5223,
+      "step": 835
+    },
+    {
+      "epoch": 0.06238627501949571,
+      "grad_norm": 1.2842494249343872,
+      "learning_rate": 4.999981483332261e-05,
+      "loss": 0.5138,
+      "step": 840
+    },
+    {
+      "epoch": 0.06275762189461176,
+      "grad_norm": 1.5931482315063477,
+      "learning_rate": 4.999975244816805e-05,
+      "loss": 0.5125,
+      "step": 845
+    },
+    {
+      "epoch": 0.0631289687697278,
+      "grad_norm": 1.422558069229126,
+      "learning_rate": 4.9999681021750454e-05,
+      "loss": 0.5139,
+      "step": 850
+    },
+    {
+      "epoch": 0.06350031564484385,
+      "grad_norm": 1.165947675704956,
+      "learning_rate": 4.999960055409566e-05,
+      "loss": 0.5069,
+      "step": 855
+    },
+    {
+      "epoch": 0.06387166251995989,
+      "grad_norm": 1.1719167232513428,
+      "learning_rate": 4.999951104523277e-05,
+      "loss": 0.5198,
+      "step": 860
+    },
+    {
+      "epoch": 0.06424300939507593,
+      "grad_norm": 1.5366706848144531,
+      "learning_rate": 4.999941249519416e-05,
+      "loss": 0.5169,
+      "step": 865
+    },
+    {
+      "epoch": 0.06461435627019199,
+      "grad_norm": 1.3433315753936768,
+      "learning_rate": 4.999930490401546e-05,
+      "loss": 0.5239,
+      "step": 870
+    },
+    {
+      "epoch": 0.06498570314530804,
+      "grad_norm": 1.2835320234298706,
+      "learning_rate": 4.9999188271735584e-05,
+      "loss": 0.5341,
+      "step": 875
+    },
+    {
+      "epoch": 0.06535705002042408,
+      "grad_norm": 1.7535203695297241,
+      "learning_rate": 4.9999062598396715e-05,
+      "loss": 0.4885,
+      "step": 880
+    },
+    {
+      "epoch": 0.06572839689554012,
+      "grad_norm": 1.2752184867858887,
+      "learning_rate": 4.999892788404431e-05,
+      "loss": 0.5412,
+      "step": 885
+    },
+    {
+      "epoch": 0.06609974377065617,
+      "grad_norm": 1.509055733680725,
+      "learning_rate": 4.999878412872708e-05,
+      "loss": 0.526,
+      "step": 890
+    },
+    {
+      "epoch": 0.06647109064577221,
+      "grad_norm": 1.215212106704712,
+      "learning_rate": 4.9998631332497014e-05,
+      "loss": 0.5281,
+      "step": 895
+    },
+    {
+      "epoch": 0.06684243752088827,
+      "grad_norm": 1.4890586137771606,
+      "learning_rate": 4.999846949540937e-05,
+      "loss": 0.5112,
+      "step": 900
+    },
+    {
+      "epoch": 0.06721378439600431,
+      "grad_norm": 1.1772491931915283,
+      "learning_rate": 4.999829861752269e-05,
+      "loss": 0.5101,
+      "step": 905
+    },
+    {
+      "epoch": 0.06758513127112036,
+      "grad_norm": 1.0531741380691528,
+      "learning_rate": 4.999811869889876e-05,
+      "loss": 0.4913,
+      "step": 910
+    },
+    {
+      "epoch": 0.0679564781462364,
+      "grad_norm": 1.1728847026824951,
+      "learning_rate": 4.999792973960266e-05,
+      "loss": 0.5091,
+      "step": 915
+    },
+    {
+      "epoch": 0.06832782502135244,
+      "grad_norm": 1.3163584470748901,
+      "learning_rate": 4.99977317397027e-05,
+      "loss": 0.534,
+      "step": 920
+    },
+    {
+      "epoch": 0.06869917189646849,
+      "grad_norm": 1.0669679641723633,
+      "learning_rate": 4.9997524699270526e-05,
+      "loss": 0.5161,
+      "step": 925
+    },
+    {
+      "epoch": 0.06907051877158454,
+      "grad_norm": 1.5822666883468628,
+      "learning_rate": 4.999730861838099e-05,
+      "loss": 0.5384,
+      "step": 930
+    },
+    {
+      "epoch": 0.06944186564670059,
+      "grad_norm": 1.1044566631317139,
+      "learning_rate": 4.999708349711224e-05,
+      "loss": 0.5078,
+      "step": 935
+    },
+    {
+      "epoch": 0.06981321252181663,
+      "grad_norm": 1.54874849319458,
+      "learning_rate": 4.99968493355457e-05,
+      "loss": 0.5163,
+      "step": 940
+    },
+    {
+      "epoch": 0.07018455939693267,
+      "grad_norm": 1.185843586921692,
+      "learning_rate": 4.999660613376606e-05,
+      "loss": 0.5176,
+      "step": 945
+    },
+    {
+      "epoch": 0.07055590627204872,
+      "grad_norm": 1.1341667175292969,
+      "learning_rate": 4.999635389186126e-05,
+      "loss": 0.5192,
+      "step": 950
+    },
+    {
+      "epoch": 0.07092725314716476,
+      "grad_norm": 1.001858115196228,
+      "learning_rate": 4.999609260992253e-05,
+      "loss": 0.5224,
+      "step": 955
+    },
+    {
+      "epoch": 0.0712986000222808,
+      "grad_norm": 1.3810193538665771,
+      "learning_rate": 4.999582228804437e-05,
+      "loss": 0.4965,
+      "step": 960
+    },
+    {
+      "epoch": 0.07166994689739686,
+      "grad_norm": 1.026790738105774,
+      "learning_rate": 4.9995542926324536e-05,
+      "loss": 0.5086,
+      "step": 965
+    },
+    {
+      "epoch": 0.07204129377251291,
+      "grad_norm": 1.3299849033355713,
+      "learning_rate": 4.999525452486407e-05,
+      "loss": 0.5177,
+      "step": 970
+    },
+    {
+      "epoch": 0.07241264064762895,
+      "grad_norm": 1.235654354095459,
+      "learning_rate": 4.9994957083767256e-05,
+      "loss": 0.5109,
+      "step": 975
+    },
+    {
+      "epoch": 0.072783987522745,
+      "grad_norm": 1.2094359397888184,
+      "learning_rate": 4.9994650603141676e-05,
+      "loss": 0.5019,
+      "step": 980
+    },
+    {
+      "epoch": 0.07315533439786104,
+      "grad_norm": 1.2392818927764893,
+      "learning_rate": 4.999433508309817e-05,
+      "loss": 0.52,
+      "step": 985
+    },
+    {
+      "epoch": 0.07352668127297708,
+      "grad_norm": 1.0825366973876953,
+      "learning_rate": 4.999401052375085e-05,
+      "loss": 0.491,
+      "step": 990
+    },
+    {
+      "epoch": 0.07389802814809314,
+      "grad_norm": 1.3078283071517944,
+      "learning_rate": 4.999367692521709e-05,
+      "loss": 0.4992,
+      "step": 995
+    },
+    {
+      "epoch": 0.07426937502320918,
+      "grad_norm": 1.2533962726593018,
+      "learning_rate": 4.9993334287617546e-05,
+      "loss": 0.5161,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07464072189832523,
+      "grad_norm": 1.3753759860992432,
+      "learning_rate": 4.999298261107611e-05,
+      "loss": 0.526,
+      "step": 1005
+    },
+    {
+      "epoch": 0.07501206877344127,
+      "grad_norm": 1.099392056465149,
+      "learning_rate": 4.999262189571999e-05,
+      "loss": 0.5163,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07538341564855731,
+      "grad_norm": 1.0913268327713013,
+      "learning_rate": 4.999225214167964e-05,
+      "loss": 0.5052,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07575476252367336,
+      "grad_norm": 1.2516167163848877,
+      "learning_rate": 4.999187334908877e-05,
+      "loss": 0.5093,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07612610939878942,
+      "grad_norm": 1.0062530040740967,
+      "learning_rate": 4.999148551808438e-05,
+      "loss": 0.4984,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07649745627390546,
+      "grad_norm": 0.9695545434951782,
+      "learning_rate": 4.999108864880673e-05,
+      "loss": 0.5041,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0768688031490215,
+      "grad_norm": 1.2652006149291992,
+      "learning_rate": 4.999068274139934e-05,
+      "loss": 0.525,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07724015002413755,
+      "grad_norm": 1.2803031206130981,
+      "learning_rate": 4.999026779600903e-05,
+      "loss": 0.5169,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07761149689925359,
+      "grad_norm": 1.0694555044174194,
+      "learning_rate": 4.9989843812785845e-05,
+      "loss": 0.532,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07798284377436963,
+      "grad_norm": 1.0950549840927124,
+      "learning_rate": 4.998941079188313e-05,
+      "loss": 0.5061,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07835419064948568,
+      "grad_norm": 1.2538092136383057,
+      "learning_rate": 4.998896873345749e-05,
+      "loss": 0.521,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07872553752460174,
+      "grad_norm": 1.1405869722366333,
+      "learning_rate": 4.998851763766879e-05,
+      "loss": 0.5175,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07909688439971778,
+      "grad_norm": 1.5427117347717285,
+      "learning_rate": 4.998805750468017e-05,
+      "loss": 0.5416,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07946823127483382,
+      "grad_norm": 1.0638082027435303,
+      "learning_rate": 4.998758833465805e-05,
+      "loss": 0.4665,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07983957814994987,
+      "grad_norm": 1.1549831628799438,
+      "learning_rate": 4.998711012777211e-05,
+      "loss": 0.5117,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08021092502506591,
+      "grad_norm": 1.106373906135559,
+      "learning_rate": 4.998662288419528e-05,
+      "loss": 0.5101,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08058227190018195,
+      "grad_norm": 1.6457034349441528,
+      "learning_rate": 4.998612660410378e-05,
+      "loss": 0.4911,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08095361877529801,
+      "grad_norm": 0.9553409814834595,
+      "learning_rate": 4.998562128767709e-05,
+      "loss": 0.5055,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08132496565041406,
+      "grad_norm": 0.9238666892051697,
+      "learning_rate": 4.998510693509797e-05,
+      "loss": 0.5128,
+      "step": 1095
+    },
+    {
+      "epoch": 0.0816963125255301,
+      "grad_norm": 0.9115836024284363,
+      "learning_rate": 4.998458354655242e-05,
+      "loss": 0.5021,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08206765940064614,
+      "grad_norm": 0.9943495392799377,
+      "learning_rate": 4.998405112222974e-05,
+      "loss": 0.4871,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08243900627576219,
+      "grad_norm": 1.342914342880249,
+      "learning_rate": 4.998350966232248e-05,
+      "loss": 0.5015,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08281035315087823,
+      "grad_norm": 1.1267045736312866,
+      "learning_rate": 4.998295916702646e-05,
+      "loss": 0.4942,
+      "step": 1115
+    },
+    {
+      "epoch": 0.08318170002599429,
+      "grad_norm": 1.411318302154541,
+      "learning_rate": 4.998239963654077e-05,
+      "loss": 0.5256,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08355304690111033,
+      "grad_norm": 1.043161153793335,
+      "learning_rate": 4.9981831071067766e-05,
+      "loss": 0.4977,
+      "step": 1125
+    },
+    {
+      "epoch": 0.08392439377622637,
+      "grad_norm": 1.0562570095062256,
+      "learning_rate": 4.9981253470813066e-05,
+      "loss": 0.4864,
+      "step": 1130
+    },
+    {
+      "epoch": 0.08429574065134242,
+      "grad_norm": 1.0544936656951904,
+      "learning_rate": 4.9980666835985565e-05,
+      "loss": 0.4805,
+      "step": 1135
+    },
+    {
+      "epoch": 0.08466708752645846,
+      "grad_norm": 1.0702053308486938,
+      "learning_rate": 4.998007116679744e-05,
+      "loss": 0.4813,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0850384344015745,
+      "grad_norm": 1.0946789979934692,
+      "learning_rate": 4.9979466463464084e-05,
+      "loss": 0.5028,
+      "step": 1145
+    },
+    {
+      "epoch": 0.08540978127669056,
+      "grad_norm": 1.2464505434036255,
+      "learning_rate": 4.997885272620422e-05,
+      "loss": 0.5108,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08578112815180661,
+      "grad_norm": 1.092294692993164,
+      "learning_rate": 4.997822995523979e-05,
+      "loss": 0.4844,
+      "step": 1155
+    },
+    {
+      "epoch": 0.08615247502692265,
+      "grad_norm": 1.2743663787841797,
+      "learning_rate": 4.9977598150796024e-05,
+      "loss": 0.4984,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0865238219020387,
+      "grad_norm": 1.2585391998291016,
+      "learning_rate": 4.997695731310143e-05,
+      "loss": 0.4927,
+      "step": 1165
+    },
+    {
+      "epoch": 0.08689516877715474,
+      "grad_norm": 0.9504044055938721,
+      "learning_rate": 4.997630744238775e-05,
+      "loss": 0.5196,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08726651565227078,
+      "grad_norm": 1.1162606477737427,
+      "learning_rate": 4.9975648538890026e-05,
+      "loss": 0.4981,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08763786252738683,
+      "grad_norm": 0.9510879516601562,
+      "learning_rate": 4.997498060284655e-05,
+      "loss": 0.5058,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08800920940250288,
+      "grad_norm": 1.524391770362854,
+      "learning_rate": 4.9974303634498884e-05,
+      "loss": 0.4991,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08838055627761893,
+      "grad_norm": 0.9613252878189087,
+      "learning_rate": 4.997361763409185e-05,
+      "loss": 0.4902,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08875190315273497,
+      "grad_norm": 0.8819934725761414,
+      "learning_rate": 4.997292260187356e-05,
+      "loss": 0.4917,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08912325002785101,
+      "grad_norm": 1.2517043352127075,
+      "learning_rate": 4.997221853809536e-05,
+      "loss": 0.4978,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08949459690296706,
+      "grad_norm": 1.4532949924468994,
+      "learning_rate": 4.997150544301188e-05,
+      "loss": 0.5176,
+      "step": 1205
+    },
+    {
+      "epoch": 0.0898659437780831,
+      "grad_norm": 0.9775285720825195,
+      "learning_rate": 4.997078331688101e-05,
+      "loss": 0.4947,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09023729065319916,
+      "grad_norm": 1.055336833000183,
+      "learning_rate": 4.997005215996392e-05,
+      "loss": 0.4804,
+      "step": 1215
+    },
+    {
+      "epoch": 0.0906086375283152,
+      "grad_norm": 1.0310511589050293,
+      "learning_rate": 4.996931197252503e-05,
+      "loss": 0.4905,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09097998440343125,
+      "grad_norm": 1.1748933792114258,
+      "learning_rate": 4.996856275483204e-05,
+      "loss": 0.5215,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09135133127854729,
+      "grad_norm": 1.008360505104065,
+      "learning_rate": 4.996780450715589e-05,
+      "loss": 0.5094,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09172267815366333,
+      "grad_norm": 1.1358569860458374,
+      "learning_rate": 4.996703722977082e-05,
+      "loss": 0.4978,
+      "step": 1235
+    },
+    {
+      "epoch": 0.09209402502877938,
+      "grad_norm": 1.0410033464431763,
+      "learning_rate": 4.996626092295431e-05,
+      "loss": 0.4902,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09246537190389544,
+      "grad_norm": 1.212699294090271,
+      "learning_rate": 4.9965475586987123e-05,
+      "loss": 0.5013,
+      "step": 1245
+    },
+    {
+      "epoch": 0.09283671877901148,
+      "grad_norm": 1.0074434280395508,
+      "learning_rate": 4.996468122215326e-05,
+      "loss": 0.4745,
+      "step": 1250
+    },
+    {
+      "epoch": 0.09320806565412752,
+      "grad_norm": 1.273102045059204,
+      "learning_rate": 4.996387782874003e-05,
+      "loss": 0.5078,
+      "step": 1255
+    },
+    {
+      "epoch": 0.09357941252924357,
+      "grad_norm": 1.1507134437561035,
+      "learning_rate": 4.9963065407037977e-05,
+      "loss": 0.5149,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09395075940435961,
+      "grad_norm": 1.203317642211914,
+      "learning_rate": 4.99622439573409e-05,
+      "loss": 0.4843,
+      "step": 1265
+    },
+    {
+      "epoch": 0.09432210627947565,
+      "grad_norm": 1.1259766817092896,
+      "learning_rate": 4.9961413479945905e-05,
+      "loss": 0.5034,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0946934531545917,
+      "grad_norm": 1.205767273902893,
+      "learning_rate": 4.9960573975153325e-05,
+      "loss": 0.5017,
+      "step": 1275
+    },
+    {
+      "epoch": 0.09506480002970776,
+      "grad_norm": 1.0495589971542358,
+      "learning_rate": 4.9959725443266765e-05,
+      "loss": 0.5004,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0954361469048238,
+      "grad_norm": 1.2640390396118164,
+      "learning_rate": 4.995886788459311e-05,
+      "loss": 0.4858,
+      "step": 1285
+    },
+    {
+      "epoch": 0.09580749377993984,
+      "grad_norm": 1.2448623180389404,
+      "learning_rate": 4.9958001299442486e-05,
+      "loss": 0.5001,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09617884065505589,
+      "grad_norm": 0.8900221586227417,
+      "learning_rate": 4.995712568812832e-05,
+      "loss": 0.472,
+      "step": 1295
+    },
+    {
+      "epoch": 0.09655018753017193,
+      "grad_norm": 0.8876100182533264,
+      "learning_rate": 4.995624105096725e-05,
+      "loss": 0.498,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09692153440528797,
+      "grad_norm": 0.8986858129501343,
+      "learning_rate": 4.995534738827923e-05,
+      "loss": 0.4935,
+      "step": 1305
+    },
+    {
+      "epoch": 0.09729288128040403,
+      "grad_norm": 1.0120513439178467,
+      "learning_rate": 4.995444470038746e-05,
+      "loss": 0.491,
+      "step": 1310
+    },
+    {
+      "epoch": 0.09766422815552007,
+      "grad_norm": 1.1924999952316284,
+      "learning_rate": 4.995353298761839e-05,
+      "loss": 0.4768,
+      "step": 1315
+    },
+    {
+      "epoch": 0.09803557503063612,
+      "grad_norm": 1.1135913133621216,
+      "learning_rate": 4.9952612250301746e-05,
+      "loss": 0.4949,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09840692190575216,
+      "grad_norm": 1.3137718439102173,
+      "learning_rate": 4.995168248877051e-05,
+      "loss": 0.4907,
+      "step": 1325
+    },
+    {
+      "epoch": 0.0987782687808682,
+      "grad_norm": 1.22038733959198,
+      "learning_rate": 4.9950743703360954e-05,
+      "loss": 0.4961,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09914961565598425,
+      "grad_norm": 1.185892939567566,
+      "learning_rate": 4.9949795894412576e-05,
+      "loss": 0.4842,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09952096253110031,
+      "grad_norm": 0.8760488629341125,
+      "learning_rate": 4.994883906226816e-05,
+      "loss": 0.4797,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09989230940621635,
+      "grad_norm": 1.1870434284210205,
+      "learning_rate": 4.9947873207273756e-05,
+      "loss": 0.4775,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1002636562813324,
+      "grad_norm": 1.0038819313049316,
+      "learning_rate": 4.994689832977866e-05,
+      "loss": 0.4744,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10063500315644844,
+      "grad_norm": 1.1821181774139404,
+      "learning_rate": 4.994591443013543e-05,
+      "loss": 0.4957,
+      "step": 1355
+    },
+    {
+      "epoch": 0.10100635003156448,
+      "grad_norm": 0.9416809678077698,
+      "learning_rate": 4.994492150869992e-05,
+      "loss": 0.4998,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10137769690668053,
+      "grad_norm": 0.8790668845176697,
+      "learning_rate": 4.9943919565831216e-05,
+      "loss": 0.5038,
+      "step": 1365
+    },
+    {
+      "epoch": 0.10174904378179657,
+      "grad_norm": 0.9455973505973816,
+      "learning_rate": 4.9942908601891666e-05,
+      "loss": 0.4791,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10212039065691263,
+      "grad_norm": 1.012926697731018,
+      "learning_rate": 4.99418886172469e-05,
+      "loss": 0.4505,
+      "step": 1375
+    },
+    {
+      "epoch": 0.10249173753202867,
+      "grad_norm": 1.208351492881775,
+      "learning_rate": 4.99408596122658e-05,
+      "loss": 0.4861,
+      "step": 1380
+    },
+    {
+      "epoch": 0.10286308440714471,
+      "grad_norm": 1.1314637660980225,
+      "learning_rate": 4.99398215873205e-05,
+      "loss": 0.4873,
+      "step": 1385
+    },
+    {
+      "epoch": 0.10323443128226076,
+      "grad_norm": 1.5732420682907104,
+      "learning_rate": 4.9938774542786416e-05,
+      "loss": 0.4915,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1036057781573768,
+      "grad_norm": 1.0597292184829712,
+      "learning_rate": 4.993771847904221e-05,
+      "loss": 0.4978,
+      "step": 1395
+    },
+    {
+      "epoch": 0.10397712503249285,
+      "grad_norm": 1.0806019306182861,
+      "learning_rate": 4.9936653396469814e-05,
+      "loss": 0.4827,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1043484719076089,
+      "grad_norm": 0.9781217575073242,
+      "learning_rate": 4.9935579295454415e-05,
+      "loss": 0.4542,
+      "step": 1405
+    },
+    {
+      "epoch": 0.10471981878272495,
+      "grad_norm": 1.0065959692001343,
+      "learning_rate": 4.993449617638447e-05,
+      "loss": 0.5061,
+      "step": 1410
+    },
+    {
+      "epoch": 0.10509116565784099,
+      "grad_norm": 0.883406400680542,
+      "learning_rate": 4.9933404039651696e-05,
+      "loss": 0.4943,
+      "step": 1415
+    },
+    {
+      "epoch": 0.10546251253295703,
+      "grad_norm": 1.0783164501190186,
+      "learning_rate": 4.993230288565107e-05,
+      "loss": 0.4971,
+      "step": 1420
+    },
+    {
+      "epoch": 0.10583385940807308,
+      "grad_norm": 1.1411856412887573,
+      "learning_rate": 4.993119271478082e-05,
+      "loss": 0.4713,
+      "step": 1425
+    },
+    {
+      "epoch": 0.10620520628318912,
+      "grad_norm": 1.077597737312317,
+      "learning_rate": 4.993007352744245e-05,
+      "loss": 0.4882,
+      "step": 1430
+    },
+    {
+      "epoch": 0.10657655315830518,
+      "grad_norm": 1.2164273262023926,
+      "learning_rate": 4.992894532404071e-05,
+      "loss": 0.478,
+      "step": 1435
+    },
+    {
+      "epoch": 0.10694790003342122,
+      "grad_norm": 1.0068302154541016,
+      "learning_rate": 4.9927808104983644e-05,
+      "loss": 0.4889,
+      "step": 1440
+    },
+    {
+      "epoch": 0.10731924690853727,
+      "grad_norm": 1.2516838312149048,
+      "learning_rate": 4.99266618706825e-05,
+      "loss": 0.4866,
+      "step": 1445
+    },
+    {
+      "epoch": 0.10769059378365331,
+      "grad_norm": 0.9896201491355896,
+      "learning_rate": 4.992550662155184e-05,
+      "loss": 0.4707,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10806194065876935,
+      "grad_norm": 0.8557565808296204,
+      "learning_rate": 4.992434235800946e-05,
+      "loss": 0.4912,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1084332875338854,
+      "grad_norm": 0.9181938171386719,
+      "learning_rate": 4.992316908047641e-05,
+      "loss": 0.4949,
+      "step": 1460
+    },
+    {
+      "epoch": 0.10880463440900145,
+      "grad_norm": 0.9550036787986755,
+      "learning_rate": 4.9921986789377016e-05,
+      "loss": 0.4766,
+      "step": 1465
+    },
+    {
+      "epoch": 0.1091759812841175,
+      "grad_norm": 1.0532774925231934,
+      "learning_rate": 4.992079548513887e-05,
+      "loss": 0.4732,
+      "step": 1470
+    },
+    {
+      "epoch": 0.10954732815923354,
+      "grad_norm": 1.107776403427124,
+      "learning_rate": 4.99195951681928e-05,
+      "loss": 0.4779,
+      "step": 1475
+    },
+    {
+      "epoch": 0.10991867503434959,
+      "grad_norm": 1.1875584125518799,
+      "learning_rate": 4.991838583897291e-05,
+      "loss": 0.5138,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11029002190946563,
+      "grad_norm": 0.963405966758728,
+      "learning_rate": 4.991716749791656e-05,
+      "loss": 0.4846,
+      "step": 1485
+    },
+    {
+      "epoch": 0.11066136878458167,
+      "grad_norm": 1.337599277496338,
+      "learning_rate": 4.991594014546437e-05,
+      "loss": 0.482,
+      "step": 1490
+    },
+    {
+      "epoch": 0.11103271565969772,
+      "grad_norm": 1.094760775566101,
+      "learning_rate": 4.991470378206021e-05,
+      "loss": 0.4878,
+      "step": 1495
+    },
+    {
+      "epoch": 0.11140406253481377,
+      "grad_norm": 1.0118831396102905,
+      "learning_rate": 4.9913458408151216e-05,
+      "loss": 0.4795,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11177540940992982,
+      "grad_norm": 1.1335307359695435,
+      "learning_rate": 4.991220402418778e-05,
+      "loss": 0.4705,
+      "step": 1505
+    },
+    {
+      "epoch": 0.11214675628504586,
+      "grad_norm": 0.9686943292617798,
+      "learning_rate": 4.991094063062357e-05,
+      "loss": 0.4788,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1125181031601619,
+      "grad_norm": 1.356569766998291,
+      "learning_rate": 4.9909668227915485e-05,
+      "loss": 0.4903,
+      "step": 1515
+    },
+    {
+      "epoch": 0.11288945003527795,
+      "grad_norm": 1.1215263605117798,
+      "learning_rate": 4.99083868165237e-05,
+      "loss": 0.4612,
+      "step": 1520
+    },
+    {
+      "epoch": 0.113260796910394,
+      "grad_norm": 0.9811176061630249,
+      "learning_rate": 4.9907096396911634e-05,
+      "loss": 0.4978,
+      "step": 1525
+    },
+    {
+      "epoch": 0.11363214378551005,
+      "grad_norm": 1.0463788509368896,
+      "learning_rate": 4.9905796969545985e-05,
+      "loss": 0.4743,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1140034906606261,
+      "grad_norm": 0.8583905696868896,
+      "learning_rate": 4.9904488534896695e-05,
+      "loss": 0.5038,
+      "step": 1535
+    },
+    {
+      "epoch": 0.11437483753574214,
+      "grad_norm": 1.0710617303848267,
+      "learning_rate": 4.990317109343695e-05,
+      "loss": 0.4723,
+      "step": 1540
+    },
+    {
+      "epoch": 0.11474618441085818,
+      "grad_norm": 1.47568678855896,
+      "learning_rate": 4.9901844645643225e-05,
+      "loss": 0.4758,
+      "step": 1545
+    },
+    {
+      "epoch": 0.11511753128597423,
+      "grad_norm": 1.0400784015655518,
+      "learning_rate": 4.990050919199522e-05,
+      "loss": 0.4836,
+      "step": 1550
+    },
+    {
+      "epoch": 0.11548887816109027,
+      "grad_norm": 0.974274754524231,
+      "learning_rate": 4.989916473297592e-05,
+      "loss": 0.4874,
+      "step": 1555
+    },
+    {
+      "epoch": 0.11586022503620633,
+      "grad_norm": 2.123159646987915,
+      "learning_rate": 4.9897811269071554e-05,
+      "loss": 0.471,
+      "step": 1560
+    },
+    {
+      "epoch": 0.11623157191132237,
+      "grad_norm": 0.9718955159187317,
+      "learning_rate": 4.98964488007716e-05,
+      "loss": 0.4726,
+      "step": 1565
+    },
+    {
+      "epoch": 0.11660291878643841,
+      "grad_norm": 1.31617271900177,
+      "learning_rate": 4.989507732856881e-05,
+      "loss": 0.4713,
+      "step": 1570
+    },
+    {
+      "epoch": 0.11697426566155446,
+      "grad_norm": 1.1256753206253052,
+      "learning_rate": 4.989369685295916e-05,
+      "loss": 0.4697,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1173456125366705,
+      "grad_norm": 1.2796781063079834,
+      "learning_rate": 4.9892307374441935e-05,
+      "loss": 0.5017,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11771695941178655,
+      "grad_norm": 1.045142650604248,
+      "learning_rate": 4.989090889351963e-05,
+      "loss": 0.4939,
+      "step": 1585
+    },
+    {
+      "epoch": 0.11808830628690259,
+      "grad_norm": 1.194403886795044,
+      "learning_rate": 4.9889501410698006e-05,
+      "loss": 0.483,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11845965316201865,
+      "grad_norm": 1.3747239112854004,
+      "learning_rate": 4.98880849264861e-05,
+      "loss": 0.5081,
+      "step": 1595
+    },
+    {
+      "epoch": 0.11883100003713469,
+      "grad_norm": 1.1211419105529785,
+      "learning_rate": 4.988665944139618e-05,
+      "loss": 0.48,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11920234691225073,
+      "grad_norm": 1.2679734230041504,
+      "learning_rate": 4.988522495594378e-05,
+      "loss": 0.4886,
+      "step": 1605
+    },
+    {
+      "epoch": 0.11957369378736678,
+      "grad_norm": 0.9416651725769043,
+      "learning_rate": 4.9883781470647684e-05,
+      "loss": 0.4769,
+      "step": 1610
+    },
+    {
+      "epoch": 0.11994504066248282,
+      "grad_norm": 1.0857641696929932,
+      "learning_rate": 4.9882328986029934e-05,
+      "loss": 0.4767,
+      "step": 1615
+    },
+    {
+      "epoch": 0.12031638753759887,
+      "grad_norm": 1.0913755893707275,
+      "learning_rate": 4.9880867502615834e-05,
+      "loss": 0.4845,
+      "step": 1620
+    },
+    {
+      "epoch": 0.12068773441271492,
+      "grad_norm": 1.094753623008728,
+      "learning_rate": 4.987939702093394e-05,
+      "loss": 0.4514,
+      "step": 1625
+    },
+    {
+      "epoch": 0.12105908128783097,
+      "grad_norm": 1.1173486709594727,
+      "learning_rate": 4.9877917541516036e-05,
+      "loss": 0.4695,
+      "step": 1630
+    },
+    {
+      "epoch": 0.12143042816294701,
+      "grad_norm": 1.1658018827438354,
+      "learning_rate": 4.987642906489721e-05,
+      "loss": 0.5047,
+      "step": 1635
+    },
+    {
+      "epoch": 0.12180177503806305,
+      "grad_norm": 1.165580153465271,
+      "learning_rate": 4.9874931591615756e-05,
+      "loss": 0.493,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1221731219131791,
+      "grad_norm": 1.222952127456665,
+      "learning_rate": 4.9873425122213245e-05,
+      "loss": 0.5104,
+      "step": 1645
+    },
+    {
+      "epoch": 0.12254446878829514,
+      "grad_norm": 0.9523534178733826,
+      "learning_rate": 4.98719096572345e-05,
+      "loss": 0.4683,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1229158156634112,
+      "grad_norm": 0.9762274622917175,
+      "learning_rate": 4.987038519722759e-05,
+      "loss": 0.4804,
+      "step": 1655
+    },
+    {
+      "epoch": 0.12328716253852724,
+      "grad_norm": 1.119864821434021,
+      "learning_rate": 4.986885174274386e-05,
+      "loss": 0.4881,
+      "step": 1660
+    },
+    {
+      "epoch": 0.12365850941364329,
+      "grad_norm": 0.6928736567497253,
+      "learning_rate": 4.9867309294337865e-05,
+      "loss": 0.4629,
+      "step": 1665
+    },
+    {
+      "epoch": 0.12402985628875933,
+      "grad_norm": 0.9195057153701782,
+      "learning_rate": 4.986575785256745e-05,
+      "loss": 0.4444,
+      "step": 1670
+    },
+    {
+      "epoch": 0.12440120316387537,
+      "grad_norm": 0.9007054567337036,
+      "learning_rate": 4.9864197417993705e-05,
+      "loss": 0.4845,
+      "step": 1675
+    },
+    {
+      "epoch": 0.12477255003899142,
+      "grad_norm": 1.1238789558410645,
+      "learning_rate": 4.9862627991180967e-05,
+      "loss": 0.4726,
+      "step": 1680
+    },
+    {
+      "epoch": 0.12514389691410746,
+      "grad_norm": 0.964441180229187,
+      "learning_rate": 4.9861049572696814e-05,
+      "loss": 0.4785,
+      "step": 1685
+    },
+    {
+      "epoch": 0.12551524378922352,
+      "grad_norm": 0.9361527562141418,
+      "learning_rate": 4.98594621631121e-05,
+      "loss": 0.4457,
+      "step": 1690
+    },
+    {
+      "epoch": 0.12588659066433955,
+      "grad_norm": 0.9168547987937927,
+      "learning_rate": 4.985786576300091e-05,
+      "loss": 0.4703,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1262579375394556,
+      "grad_norm": 1.0740379095077515,
+      "learning_rate": 4.9856260372940587e-05,
+      "loss": 0.4661,
+      "step": 1700
+    },
+    {
+      "epoch": 0.12662928441457166,
+      "grad_norm": 0.9852397441864014,
+      "learning_rate": 4.985464599351174e-05,
+      "loss": 0.458,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1270006312896877,
+      "grad_norm": 0.8920067548751831,
+      "learning_rate": 4.98530226252982e-05,
+      "loss": 0.4639,
+      "step": 1710
+    },
+    {
+      "epoch": 0.12737197816480375,
+      "grad_norm": 0.983961820602417,
+      "learning_rate": 4.985139026888708e-05,
+      "loss": 0.4877,
+      "step": 1715
+    },
+    {
+      "epoch": 0.12774332503991978,
+      "grad_norm": 1.1460301876068115,
+      "learning_rate": 4.984974892486872e-05,
+      "loss": 0.4696,
+      "step": 1720
+    },
+    {
+      "epoch": 0.12811467191503584,
+      "grad_norm": 1.1181328296661377,
+      "learning_rate": 4.984809859383671e-05,
+      "loss": 0.4756,
+      "step": 1725
+    },
+    {
+      "epoch": 0.12848601879015187,
+      "grad_norm": 0.9542262554168701,
+      "learning_rate": 4.984643927638792e-05,
+      "loss": 0.4715,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12885736566526793,
+      "grad_norm": 1.1598364114761353,
+      "learning_rate": 4.9844770973122426e-05,
+      "loss": 0.4744,
+      "step": 1735
+    },
+    {
+      "epoch": 0.12922871254038398,
+      "grad_norm": 1.019036889076233,
+      "learning_rate": 4.98430936846436e-05,
+      "loss": 0.4886,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1296000594155,
+      "grad_norm": 1.1101309061050415,
+      "learning_rate": 4.984140741155802e-05,
+      "loss": 0.4813,
+      "step": 1745
+    },
+    {
+      "epoch": 0.12997140629061607,
+      "grad_norm": 1.0255146026611328,
+      "learning_rate": 4.983971215447555e-05,
+      "loss": 0.4749,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1303427531657321,
+      "grad_norm": 0.8294594287872314,
+      "learning_rate": 4.983800791400927e-05,
+      "loss": 0.4626,
+      "step": 1755
+    },
+    {
+      "epoch": 0.13071410004084816,
+      "grad_norm": 1.257464051246643,
+      "learning_rate": 4.983629469077554e-05,
+      "loss": 0.4851,
+      "step": 1760
+    },
+    {
+      "epoch": 0.13108544691596422,
+      "grad_norm": 0.9600897431373596,
+      "learning_rate": 4.983457248539395e-05,
+      "loss": 0.474,
+      "step": 1765
+    },
+    {
+      "epoch": 0.13145679379108025,
+      "grad_norm": 1.1232261657714844,
+      "learning_rate": 4.983284129848733e-05,
+      "loss": 0.4704,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1318281406661963,
+      "grad_norm": 0.901701807975769,
+      "learning_rate": 4.983110113068179e-05,
+      "loss": 0.4659,
+      "step": 1775
+    },
+    {
+      "epoch": 0.13219948754131233,
+      "grad_norm": 0.829624354839325,
+      "learning_rate": 4.982935198260666e-05,
+      "loss": 0.463,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1325708344164284,
+      "grad_norm": 1.1589818000793457,
+      "learning_rate": 4.982759385489453e-05,
+      "loss": 0.4614,
+      "step": 1785
+    },
+    {
+      "epoch": 0.13294218129154442,
+      "grad_norm": 0.9396231770515442,
+      "learning_rate": 4.9825826748181225e-05,
+      "loss": 0.4658,
+      "step": 1790
+    },
+    {
+      "epoch": 0.13331352816666048,
+      "grad_norm": 1.2056121826171875,
+      "learning_rate": 4.982405066310584e-05,
+      "loss": 0.4618,
+      "step": 1795
+    },
+    {
+      "epoch": 0.13368487504177654,
+      "grad_norm": 1.2107425928115845,
+      "learning_rate": 4.982226560031069e-05,
+      "loss": 0.4813,
+      "step": 1800
+    },
+    {
+      "epoch": 0.13405622191689257,
+      "grad_norm": 1.4222759008407593,
+      "learning_rate": 4.9820471560441365e-05,
+      "loss": 0.4693,
+      "step": 1805
+    },
+    {
+      "epoch": 0.13442756879200862,
+      "grad_norm": 1.177914023399353,
+      "learning_rate": 4.9818668544146675e-05,
+      "loss": 0.4696,
+      "step": 1810
+    },
+    {
+      "epoch": 0.13479891566712465,
+      "grad_norm": 1.003184199333191,
+      "learning_rate": 4.981685655207869e-05,
+      "loss": 0.4646,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1351702625422407,
+      "grad_norm": 1.2326256036758423,
+      "learning_rate": 4.981503558489272e-05,
+      "loss": 0.4656,
+      "step": 1820
+    },
+    {
+      "epoch": 0.13554160941735674,
+      "grad_norm": 0.9597446322441101,
+      "learning_rate": 4.9813205643247343e-05,
+      "loss": 0.4609,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1359129562924728,
+      "grad_norm": 1.1886649131774902,
+      "learning_rate": 4.981136672780434e-05,
+      "loss": 0.472,
+      "step": 1830
+    },
+    {
+      "epoch": 0.13628430316758885,
+      "grad_norm": 0.9577616453170776,
+      "learning_rate": 4.9809518839228786e-05,
+      "loss": 0.4688,
+      "step": 1835
+    },
+    {
+      "epoch": 0.13665565004270488,
+      "grad_norm": 1.0367263555526733,
+      "learning_rate": 4.9807661978188966e-05,
+      "loss": 0.4562,
+      "step": 1840
+    },
+    {
+      "epoch": 0.13702699691782094,
+      "grad_norm": 0.9544695019721985,
+      "learning_rate": 4.980579614535642e-05,
+      "loss": 0.4729,
+      "step": 1845
+    },
+    {
+      "epoch": 0.13739834379293697,
+      "grad_norm": 1.0527535676956177,
+      "learning_rate": 4.9803921341405943e-05,
+      "loss": 0.4644,
+      "step": 1850
+    },
+    {
+      "epoch": 0.13776969066805303,
+      "grad_norm": 1.109802007675171,
+      "learning_rate": 4.980203756701555e-05,
+      "loss": 0.4876,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1381410375431691,
+      "grad_norm": 1.127073049545288,
+      "learning_rate": 4.9800144822866526e-05,
+      "loss": 0.4829,
+      "step": 1860
+    },
+    {
+      "epoch": 0.13851238441828512,
+      "grad_norm": 1.0131291151046753,
+      "learning_rate": 4.979824310964339e-05,
+      "loss": 0.4933,
+      "step": 1865
+    },
+    {
+      "epoch": 0.13888373129340117,
+      "grad_norm": 0.9607428908348083,
+      "learning_rate": 4.979633242803391e-05,
+      "loss": 0.4698,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1392550781685172,
+      "grad_norm": 0.9015031456947327,
+      "learning_rate": 4.979441277872907e-05,
+      "loss": 0.4914,
+      "step": 1875
+    },
+    {
+      "epoch": 0.13962642504363326,
+      "grad_norm": 1.4103434085845947,
+      "learning_rate": 4.9792484162423146e-05,
+      "loss": 0.4834,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1399977719187493,
+      "grad_norm": 0.8119305372238159,
+      "learning_rate": 4.979054657981361e-05,
+      "loss": 0.4593,
+      "step": 1885
+    },
+    {
+      "epoch": 0.14036911879386535,
+      "grad_norm": 0.9538850784301758,
+      "learning_rate": 4.978860003160121e-05,
+      "loss": 0.4475,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1407404656689814,
+      "grad_norm": 0.964397668838501,
+      "learning_rate": 4.978664451848992e-05,
+      "loss": 0.4568,
+      "step": 1895
+    },
+    {
+      "epoch": 0.14111181254409744,
+      "grad_norm": 0.9824573397636414,
+      "learning_rate": 4.9784680041186946e-05,
+      "loss": 0.4499,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1414831594192135,
+      "grad_norm": 2.0519871711730957,
+      "learning_rate": 4.9782706600402774e-05,
+      "loss": 0.4711,
+      "step": 1905
+    },
+    {
+      "epoch": 0.14185450629432952,
+      "grad_norm": 1.1351428031921387,
+      "learning_rate": 4.978072419685108e-05,
+      "loss": 0.4851,
+      "step": 1910
+    },
+    {
+      "epoch": 0.14222585316944558,
+      "grad_norm": 1.1103254556655884,
+      "learning_rate": 4.977873283124883e-05,
+      "loss": 0.4622,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1425972000445616,
+      "grad_norm": 1.0085701942443848,
+      "learning_rate": 4.97767325043162e-05,
+      "loss": 0.4777,
+      "step": 1920
+    },
+    {
+      "epoch": 0.14296854691967767,
+      "grad_norm": 0.9454324245452881,
+      "learning_rate": 4.977472321677662e-05,
+      "loss": 0.4542,
+      "step": 1925
+    },
+    {
+      "epoch": 0.14333989379479373,
+      "grad_norm": 1.1343188285827637,
+      "learning_rate": 4.9772704969356746e-05,
+      "loss": 0.4716,
+      "step": 1930
+    },
+    {
+      "epoch": 0.14371124066990976,
+      "grad_norm": 1.3752365112304688,
+      "learning_rate": 4.97706777627865e-05,
+      "loss": 0.4778,
+      "step": 1935
+    },
+    {
+      "epoch": 0.14408258754502581,
+      "grad_norm": 1.1367682218551636,
+      "learning_rate": 4.976864159779903e-05,
+      "loss": 0.4674,
+      "step": 1940
+    },
+    {
+      "epoch": 0.14445393442014184,
+      "grad_norm": 1.1333668231964111,
+      "learning_rate": 4.97665964751307e-05,
+      "loss": 0.4848,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1448252812952579,
+      "grad_norm": 0.9810808300971985,
+      "learning_rate": 4.976454239552117e-05,
+      "loss": 0.4754,
+      "step": 1950
+    },
+    {
+      "epoch": 0.14519662817037396,
+      "grad_norm": 1.1838676929473877,
+      "learning_rate": 4.976247935971328e-05,
+      "loss": 0.4632,
+      "step": 1955
+    },
+    {
+      "epoch": 0.14556797504549,
+      "grad_norm": 0.9803593754768372,
+      "learning_rate": 4.976040736845316e-05,
+      "loss": 0.4817,
+      "step": 1960
+    },
+    {
+      "epoch": 0.14593932192060605,
+      "grad_norm": 1.4662213325500488,
+      "learning_rate": 4.975832642249012e-05,
+      "loss": 0.4718,
+      "step": 1965
+    },
+    {
+      "epoch": 0.14631066879572208,
+      "grad_norm": 1.4118518829345703,
+      "learning_rate": 4.975623652257677e-05,
+      "loss": 0.4812,
+      "step": 1970
+    },
+    {
+      "epoch": 0.14668201567083813,
+      "grad_norm": 0.9826330542564392,
+      "learning_rate": 4.975413766946893e-05,
+      "loss": 0.4664,
+      "step": 1975
+    },
+    {
+      "epoch": 0.14705336254595416,
+      "grad_norm": 1.0626474618911743,
+      "learning_rate": 4.975202986392565e-05,
+      "loss": 0.4695,
+      "step": 1980
+    },
+    {
+      "epoch": 0.14742470942107022,
+      "grad_norm": 0.8978970646858215,
+      "learning_rate": 4.974991310670922e-05,
+      "loss": 0.4836,
+      "step": 1985
+    },
+    {
+      "epoch": 0.14779605629618628,
+      "grad_norm": 0.993985116481781,
+      "learning_rate": 4.97477873985852e-05,
+      "loss": 0.4761,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1481674031713023,
+      "grad_norm": 0.9552356004714966,
+      "learning_rate": 4.974565274032233e-05,
+      "loss": 0.4765,
+      "step": 1995
+    },
+    {
+      "epoch": 0.14853875004641837,
+      "grad_norm": 0.9018028974533081,
+      "learning_rate": 4.974350913269263e-05,
+      "loss": 0.4664,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1489100969215344,
+      "grad_norm": 0.978186845779419,
+      "learning_rate": 4.9741356576471356e-05,
+      "loss": 0.4541,
+      "step": 2005
+    },
+    {
+      "epoch": 0.14928144379665045,
+      "grad_norm": 1.0590962171554565,
+      "learning_rate": 4.973919507243697e-05,
+      "loss": 0.4698,
+      "step": 2010
+    },
+    {
+      "epoch": 0.14965279067176648,
+      "grad_norm": 0.9982547163963318,
+      "learning_rate": 4.973702462137121e-05,
+      "loss": 0.4669,
+      "step": 2015
+    },
+    {
+      "epoch": 0.15002413754688254,
+      "grad_norm": 1.0058673620224,
+      "learning_rate": 4.9734845224059e-05,
+      "loss": 0.4572,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1503954844219986,
+      "grad_norm": 0.9145234823226929,
+      "learning_rate": 4.973265688128855e-05,
+      "loss": 0.4683,
+      "step": 2025
+    },
+    {
+      "epoch": 0.15076683129711463,
+      "grad_norm": 0.8498237133026123,
+      "learning_rate": 4.9730459593851285e-05,
+      "loss": 0.4686,
+      "step": 2030
+    },
+    {
+      "epoch": 0.15113817817223069,
+      "grad_norm": 1.5811172723770142,
+      "learning_rate": 4.972825336254185e-05,
+      "loss": 0.4608,
+      "step": 2035
+    },
+    {
+      "epoch": 0.15150952504734672,
+      "grad_norm": 1.1382615566253662,
+      "learning_rate": 4.972603818815814e-05,
+      "loss": 0.4698,
+      "step": 2040
+    },
+    {
+      "epoch": 0.15188087192246277,
+      "grad_norm": 1.0171360969543457,
+      "learning_rate": 4.9723814071501295e-05,
+      "loss": 0.4559,
+      "step": 2045
+    },
+    {
+      "epoch": 0.15225221879757883,
+      "grad_norm": 1.2327966690063477,
+      "learning_rate": 4.972158101337566e-05,
+      "loss": 0.4462,
+      "step": 2050
+    },
+    {
+      "epoch": 0.15262356567269486,
+      "grad_norm": 1.284108281135559,
+      "learning_rate": 4.971933901458883e-05,
+      "loss": 0.4561,
+      "step": 2055
+    },
+    {
+      "epoch": 0.15299491254781092,
+      "grad_norm": 1.175051212310791,
+      "learning_rate": 4.971708807595165e-05,
+      "loss": 0.4632,
+      "step": 2060
+    },
+    {
+      "epoch": 0.15336625942292695,
+      "grad_norm": 1.0316301584243774,
+      "learning_rate": 4.9714828198278165e-05,
+      "loss": 0.4601,
+      "step": 2065
+    },
+    {
+      "epoch": 0.153737606298043,
+      "grad_norm": 0.9318085312843323,
+      "learning_rate": 4.971255938238567e-05,
+      "loss": 0.4735,
+      "step": 2070
+    },
+    {
+      "epoch": 0.15410895317315904,
+      "grad_norm": 0.8723039031028748,
+      "learning_rate": 4.971028162909469e-05,
+      "loss": 0.4442,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1544803000482751,
+      "grad_norm": 1.4766545295715332,
+      "learning_rate": 4.9707994939228997e-05,
+      "loss": 0.4601,
+      "step": 2080
+    },
+    {
+      "epoch": 0.15485164692339115,
+      "grad_norm": 1.0068985223770142,
+      "learning_rate": 4.970569931361557e-05,
+      "loss": 0.4633,
+      "step": 2085
+    },
+    {
+      "epoch": 0.15522299379850718,
+      "grad_norm": 0.9444260001182556,
+      "learning_rate": 4.970339475308464e-05,
+      "loss": 0.458,
+      "step": 2090
+    },
+    {
+      "epoch": 0.15559434067362324,
+      "grad_norm": 0.7326513528823853,
+      "learning_rate": 4.970108125846966e-05,
+      "loss": 0.4568,
+      "step": 2095
+    },
+    {
+      "epoch": 0.15596568754873927,
+      "grad_norm": 1.3498059511184692,
+      "learning_rate": 4.969875883060731e-05,
+      "loss": 0.4804,
+      "step": 2100
+    },
+    {
+      "epoch": 0.15633703442385533,
+      "grad_norm": 1.2519272565841675,
+      "learning_rate": 4.9696427470337495e-05,
+      "loss": 0.454,
+      "step": 2105
+    },
+    {
+      "epoch": 0.15670838129897136,
+      "grad_norm": 0.9399036169052124,
+      "learning_rate": 4.969408717850338e-05,
+      "loss": 0.4411,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1570797281740874,
+      "grad_norm": 1.2885676622390747,
+      "learning_rate": 4.9691737955951334e-05,
+      "loss": 0.4688,
+      "step": 2115
+    },
+    {
+      "epoch": 0.15745107504920347,
+      "grad_norm": 0.9798744320869446,
+      "learning_rate": 4.968937980353097e-05,
+      "loss": 0.469,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1578224219243195,
+      "grad_norm": 0.9649677872657776,
+      "learning_rate": 4.9687012722095106e-05,
+      "loss": 0.468,
+      "step": 2125
+    },
+    {
+      "epoch": 0.15819376879943556,
+      "grad_norm": 1.1075035333633423,
+      "learning_rate": 4.9684636712499816e-05,
+      "loss": 0.437,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1585651156745516,
+      "grad_norm": 1.0285245180130005,
+      "learning_rate": 4.9682251775604397e-05,
+      "loss": 0.4469,
+      "step": 2135
+    },
+    {
+      "epoch": 0.15893646254966765,
+      "grad_norm": 1.5093187093734741,
+      "learning_rate": 4.9679857912271365e-05,
+      "loss": 0.4684,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1593078094247837,
+      "grad_norm": 1.210630178451538,
+      "learning_rate": 4.967745512336648e-05,
+      "loss": 0.466,
+      "step": 2145
+    },
+    {
+      "epoch": 0.15967915629989973,
+      "grad_norm": 1.0899977684020996,
+      "learning_rate": 4.967504340975871e-05,
+      "loss": 0.4438,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1600505031750158,
+      "grad_norm": 0.985546350479126,
+      "learning_rate": 4.9672622772320274e-05,
+      "loss": 0.463,
+      "step": 2155
+    },
+    {
+      "epoch": 0.16042185005013182,
+      "grad_norm": 1.167546033859253,
+      "learning_rate": 4.967019321192659e-05,
+      "loss": 0.4543,
+      "step": 2160
+    },
+    {
+      "epoch": 0.16079319692524788,
+      "grad_norm": 1.2297065258026123,
+      "learning_rate": 4.966775472945633e-05,
+      "loss": 0.4369,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1611645438003639,
+      "grad_norm": 1.4039182662963867,
+      "learning_rate": 4.9665307325791375e-05,
+      "loss": 0.4446,
+      "step": 2170
+    },
+    {
+      "epoch": 0.16153589067547997,
+      "grad_norm": 0.9450597167015076,
+      "learning_rate": 4.966285100181684e-05,
+      "loss": 0.4493,
+      "step": 2175
+    },
+    {
+      "epoch": 0.16190723755059602,
+      "grad_norm": 0.9438601136207581,
+      "learning_rate": 4.966038575842107e-05,
+      "loss": 0.4618,
+      "step": 2180
+    },
+    {
+      "epoch": 0.16227858442571205,
+      "grad_norm": 0.9899676442146301,
+      "learning_rate": 4.965791159649562e-05,
+      "loss": 0.4851,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1626499313008281,
+      "grad_norm": 1.3799790143966675,
+      "learning_rate": 4.965542851693528e-05,
+      "loss": 0.4661,
+      "step": 2190
+    },
+    {
+      "epoch": 0.16302127817594414,
+      "grad_norm": 0.9580912590026855,
+      "learning_rate": 4.965293652063808e-05,
+      "loss": 0.4495,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1633926250510602,
+      "grad_norm": 1.1697057485580444,
+      "learning_rate": 4.9650435608505255e-05,
+      "loss": 0.4805,
+      "step": 2200
+    },
+    {
+      "epoch": 0.16376397192617625,
+      "grad_norm": 0.8966139554977417,
+      "learning_rate": 4.964792578144126e-05,
+      "loss": 0.4365,
+      "step": 2205
+    },
+    {
+      "epoch": 0.16413531880129228,
+      "grad_norm": 0.918121874332428,
+      "learning_rate": 4.96454070403538e-05,
+      "loss": 0.4669,
+      "step": 2210
+    },
+    {
+      "epoch": 0.16450666567640834,
+      "grad_norm": 0.9863184690475464,
+      "learning_rate": 4.964287938615378e-05,
+      "loss": 0.4437,
+      "step": 2215
+    },
+    {
+      "epoch": 0.16487801255152437,
+      "grad_norm": 1.2543003559112549,
+      "learning_rate": 4.964034281975534e-05,
+      "loss": 0.471,
+      "step": 2220
+    },
+    {
+      "epoch": 0.16524935942664043,
+      "grad_norm": 0.7972759008407593,
+      "learning_rate": 4.963779734207582e-05,
+      "loss": 0.448,
+      "step": 2225
+    },
+    {
+      "epoch": 0.16562070630175646,
+      "grad_norm": 1.0038282871246338,
+      "learning_rate": 4.963524295403583e-05,
+      "loss": 0.4569,
+      "step": 2230
+    },
+    {
+      "epoch": 0.16599205317687252,
+      "grad_norm": 1.3752259016036987,
+      "learning_rate": 4.9632679656559164e-05,
+      "loss": 0.4662,
+      "step": 2235
+    },
+    {
+      "epoch": 0.16636340005198857,
+      "grad_norm": 1.423388123512268,
+      "learning_rate": 4.963010745057285e-05,
+      "loss": 0.4842,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1667347469271046,
+      "grad_norm": 1.0444741249084473,
+      "learning_rate": 4.962752633700714e-05,
+      "loss": 0.4554,
+      "step": 2245
+    },
+    {
+      "epoch": 0.16710609380222066,
+      "grad_norm": 0.9717627167701721,
+      "learning_rate": 4.962493631679549e-05,
+      "loss": 0.4636,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1674774406773367,
+      "grad_norm": 1.248321771621704,
+      "learning_rate": 4.9622337390874604e-05,
+      "loss": 0.4451,
+      "step": 2255
+    },
+    {
+      "epoch": 0.16784878755245275,
+      "grad_norm": 0.9144636988639832,
+      "learning_rate": 4.9619729560184394e-05,
+      "loss": 0.4491,
+      "step": 2260
+    },
+    {
+      "epoch": 0.16822013442756878,
+      "grad_norm": 0.9530754685401917,
+      "learning_rate": 4.961711282566799e-05,
+      "loss": 0.4665,
+      "step": 2265
+    },
+    {
+      "epoch": 0.16859148130268484,
+      "grad_norm": 1.0826373100280762,
+      "learning_rate": 4.961448718827175e-05,
+      "loss": 0.4386,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1689628281778009,
+      "grad_norm": 0.8597514033317566,
+      "learning_rate": 4.961185264894523e-05,
+      "loss": 0.4703,
+      "step": 2275
+    },
+    {
+      "epoch": 0.16933417505291692,
+      "grad_norm": 0.9985958933830261,
+      "learning_rate": 4.960920920864124e-05,
+      "loss": 0.4422,
+      "step": 2280
+    },
+    {
+      "epoch": 0.16970552192803298,
+      "grad_norm": 1.0870766639709473,
+      "learning_rate": 4.960655686831579e-05,
+      "loss": 0.4793,
+      "step": 2285
+    },
+    {
+      "epoch": 0.170076868803149,
+      "grad_norm": 1.083158016204834,
+      "learning_rate": 4.96038956289281e-05,
+      "loss": 0.4528,
+      "step": 2290
+    },
+    {
+      "epoch": 0.17044821567826507,
+      "grad_norm": 1.491381049156189,
+      "learning_rate": 4.960122549144062e-05,
+      "loss": 0.4299,
+      "step": 2295
+    },
+    {
+      "epoch": 0.17081956255338113,
+      "grad_norm": 0.8504036068916321,
+      "learning_rate": 4.959854645681902e-05,
+      "loss": 0.4569,
+      "step": 2300
+    },
+    {
+      "epoch": 0.17119090942849716,
+      "grad_norm": 1.2399927377700806,
+      "learning_rate": 4.959585852603218e-05,
+      "loss": 0.4547,
+      "step": 2305
+    },
+    {
+      "epoch": 0.17156225630361321,
+      "grad_norm": 0.8506380915641785,
+      "learning_rate": 4.959316170005221e-05,
+      "loss": 0.4448,
+      "step": 2310
+    },
+    {
+      "epoch": 0.17193360317872924,
+      "grad_norm": 1.13820219039917,
+      "learning_rate": 4.9590455979854424e-05,
+      "loss": 0.462,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1723049500538453,
+      "grad_norm": 1.0776944160461426,
+      "learning_rate": 4.9587741366417344e-05,
+      "loss": 0.4556,
+      "step": 2320
+    },
+    {
+      "epoch": 0.17267629692896133,
+      "grad_norm": 1.0773634910583496,
+      "learning_rate": 4.9585017860722735e-05,
+      "loss": 0.4379,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1730476438040774,
+      "grad_norm": 1.044816255569458,
+      "learning_rate": 4.958228546375557e-05,
+      "loss": 0.4758,
+      "step": 2330
+    },
+    {
+      "epoch": 0.17341899067919345,
+      "grad_norm": 0.9068227410316467,
+      "learning_rate": 4.957954417650401e-05,
+      "loss": 0.4528,
+      "step": 2335
+    },
+    {
+      "epoch": 0.17379033755430948,
+      "grad_norm": 0.9294278621673584,
+      "learning_rate": 4.957679399995948e-05,
+      "loss": 0.4642,
+      "step": 2340
+    },
+    {
+      "epoch": 0.17416168442942553,
+      "grad_norm": 0.9256554245948792,
+      "learning_rate": 4.9574034935116564e-05,
+      "loss": 0.4621,
+      "step": 2345
+    },
+    {
+      "epoch": 0.17453303130454156,
+      "grad_norm": 1.2020440101623535,
+      "learning_rate": 4.957126698297311e-05,
+      "loss": 0.4491,
+      "step": 2350
+    },
+    {
+      "epoch": 0.17490437817965762,
+      "grad_norm": 0.8281359672546387,
+      "learning_rate": 4.9568490144530144e-05,
+      "loss": 0.4485,
+      "step": 2355
+    },
+    {
+      "epoch": 0.17527572505477365,
+      "grad_norm": 0.8755785822868347,
+      "learning_rate": 4.956570442079194e-05,
+      "loss": 0.457,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1756470719298897,
+      "grad_norm": 1.081264853477478,
+      "learning_rate": 4.956290981276595e-05,
+      "loss": 0.4571,
+      "step": 2365
+    },
+    {
+      "epoch": 0.17601841880500577,
+      "grad_norm": 0.8732466697692871,
+      "learning_rate": 4.9560106321462864e-05,
+      "loss": 0.4475,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1763897656801218,
+      "grad_norm": 0.7618255019187927,
+      "learning_rate": 4.955729394789657e-05,
+      "loss": 0.4696,
+      "step": 2375
+    },
+    {
+      "epoch": 0.17676111255523785,
+      "grad_norm": 0.9234897494316101,
+      "learning_rate": 4.955447269308418e-05,
+      "loss": 0.4112,
+      "step": 2380
+    },
+    {
+      "epoch": 0.17713245943035388,
+      "grad_norm": 1.0670307874679565,
+      "learning_rate": 4.955164255804601e-05,
+      "loss": 0.4223,
+      "step": 2385
+    },
+    {
+      "epoch": 0.17750380630546994,
+      "grad_norm": 1.4194025993347168,
+      "learning_rate": 4.954880354380559e-05,
+      "loss": 0.4294,
+      "step": 2390
+    },
+    {
+      "epoch": 0.177875153180586,
+      "grad_norm": 0.8653207421302795,
+      "learning_rate": 4.954595565138966e-05,
+      "loss": 0.4607,
+      "step": 2395
+    },
+    {
+      "epoch": 0.17824650005570203,
+      "grad_norm": 1.3651710748672485,
+      "learning_rate": 4.954309888182818e-05,
+      "loss": 0.4508,
+      "step": 2400
+    },
+    {
+      "epoch": 0.17861784693081809,
+      "grad_norm": 1.020920991897583,
+      "learning_rate": 4.954023323615429e-05,
+      "loss": 0.4659,
+      "step": 2405
+    },
+    {
+      "epoch": 0.17898919380593412,
+      "grad_norm": 1.022684097290039,
+      "learning_rate": 4.953735871540439e-05,
+      "loss": 0.4464,
+      "step": 2410
+    },
+    {
+      "epoch": 0.17936054068105017,
+      "grad_norm": 1.0919467210769653,
+      "learning_rate": 4.953447532061805e-05,
+      "loss": 0.4756,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1797318875561662,
+      "grad_norm": 0.9148445129394531,
+      "learning_rate": 4.9531583052838056e-05,
+      "loss": 0.4427,
+      "step": 2420
+    },
+    {
+      "epoch": 0.18010323443128226,
+      "grad_norm": 0.9530015587806702,
+      "learning_rate": 4.952868191311042e-05,
+      "loss": 0.445,
+      "step": 2425
+    },
+    {
+      "epoch": 0.18047458130639832,
+      "grad_norm": 1.0264039039611816,
+      "learning_rate": 4.952577190248434e-05,
+      "loss": 0.4512,
+      "step": 2430
+    },
+    {
+      "epoch": 0.18084592818151435,
+      "grad_norm": 1.0021376609802246,
+      "learning_rate": 4.9522853022012236e-05,
+      "loss": 0.4604,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1812172750566304,
+      "grad_norm": 1.5464372634887695,
+      "learning_rate": 4.951992527274975e-05,
+      "loss": 0.4836,
+      "step": 2440
+    },
+    {
+      "epoch": 0.18158862193174644,
+      "grad_norm": 1.0679705142974854,
+      "learning_rate": 4.951698865575569e-05,
+      "loss": 0.4829,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1819599688068625,
+      "grad_norm": 1.0598608255386353,
+      "learning_rate": 4.951404317209211e-05,
+      "loss": 0.4587,
+      "step": 2450
+    },
+    {
+      "epoch": 0.18233131568197852,
+      "grad_norm": 1.095999002456665,
+      "learning_rate": 4.9511088822824246e-05,
+      "loss": 0.4487,
+      "step": 2455
+    },
+    {
+      "epoch": 0.18270266255709458,
+      "grad_norm": 0.9862808585166931,
+      "learning_rate": 4.950812560902056e-05,
+      "loss": 0.4516,
+      "step": 2460
+    },
+    {
+      "epoch": 0.18307400943221064,
+      "grad_norm": 1.0196229219436646,
+      "learning_rate": 4.9505153531752715e-05,
+      "loss": 0.4521,
+      "step": 2465
+    },
+    {
+      "epoch": 0.18344535630732667,
+      "grad_norm": 1.2567567825317383,
+      "learning_rate": 4.950217259209555e-05,
+      "loss": 0.4607,
+      "step": 2470
+    },
+    {
+      "epoch": 0.18381670318244273,
+      "grad_norm": 1.211969256401062,
+      "learning_rate": 4.9499182791127166e-05,
+      "loss": 0.4295,
+      "step": 2475
+    },
+    {
+      "epoch": 0.18418805005755876,
+      "grad_norm": 0.8847858905792236,
+      "learning_rate": 4.9496184129928815e-05,
+      "loss": 0.4231,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1845593969326748,
+      "grad_norm": 1.00296151638031,
+      "learning_rate": 4.949317660958499e-05,
+      "loss": 0.4302,
+      "step": 2485
+    },
+    {
+      "epoch": 0.18493074380779087,
+      "grad_norm": 1.3998608589172363,
+      "learning_rate": 4.949016023118336e-05,
+      "loss": 0.4551,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1853020906829069,
+      "grad_norm": 1.3031240701675415,
+      "learning_rate": 4.948713499581481e-05,
+      "loss": 0.475,
+      "step": 2495
+    },
+    {
+      "epoch": 0.18567343755802296,
+      "grad_norm": 1.0290464162826538,
+      "learning_rate": 4.9484100904573436e-05,
+      "loss": 0.4522,
+      "step": 2500
+    },
+    {
+      "epoch": 0.186044784433139,
+      "grad_norm": 1.1162320375442505,
+      "learning_rate": 4.948105795855653e-05,
+      "loss": 0.4417,
+      "step": 2505
+    },
+    {
+      "epoch": 0.18641613130825505,
+      "grad_norm": 1.0466334819793701,
+      "learning_rate": 4.9478006158864586e-05,
+      "loss": 0.4514,
+      "step": 2510
+    },
+    {
+      "epoch": 0.18678747818337108,
+      "grad_norm": 0.9525560140609741,
+      "learning_rate": 4.94749455066013e-05,
+      "loss": 0.4335,
+      "step": 2515
+    },
+    {
+      "epoch": 0.18715882505848713,
+      "grad_norm": 0.918469250202179,
+      "learning_rate": 4.947187600287356e-05,
+      "loss": 0.4488,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1875301719336032,
+      "grad_norm": 0.9201804995536804,
+      "learning_rate": 4.946879764879149e-05,
+      "loss": 0.4598,
+      "step": 2525
+    },
+    {
+      "epoch": 0.18790151880871922,
+      "grad_norm": 0.810072660446167,
+      "learning_rate": 4.946571044546835e-05,
+      "loss": 0.4531,
+      "step": 2530
+    },
+    {
+      "epoch": 0.18827286568383528,
+      "grad_norm": 1.2535103559494019,
+      "learning_rate": 4.946261439402067e-05,
+      "loss": 0.4832,
+      "step": 2535
+    },
+    {
+      "epoch": 0.1886442125589513,
+      "grad_norm": 0.7918863892555237,
+      "learning_rate": 4.945950949556814e-05,
+      "loss": 0.4433,
+      "step": 2540
+    },
+    {
+      "epoch": 0.18901555943406737,
+      "grad_norm": 0.9122186899185181,
+      "learning_rate": 4.945639575123366e-05,
+      "loss": 0.451,
+      "step": 2545
+    },
+    {
+      "epoch": 0.1893869063091834,
+      "grad_norm": 0.9057549834251404,
+      "learning_rate": 4.9453273162143325e-05,
+      "loss": 0.4345,
+      "step": 2550
+    },
+    {
+      "epoch": 0.18975825318429945,
+      "grad_norm": 0.932822585105896,
+      "learning_rate": 4.945014172942644e-05,
+      "loss": 0.4801,
+      "step": 2555
+    },
+    {
+      "epoch": 0.1901296000594155,
+      "grad_norm": 1.1034599542617798,
+      "learning_rate": 4.9447001454215494e-05,
+      "loss": 0.4444,
+      "step": 2560
+    },
+    {
+      "epoch": 0.19050094693453154,
+      "grad_norm": 1.0966523885726929,
+      "learning_rate": 4.944385233764618e-05,
+      "loss": 0.4456,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1908722938096476,
+      "grad_norm": 1.0594526529312134,
+      "learning_rate": 4.94406943808574e-05,
+      "loss": 0.4359,
+      "step": 2570
+    },
+    {
+      "epoch": 0.19124364068476363,
+      "grad_norm": 0.8708404898643494,
+      "learning_rate": 4.9437527584991226e-05,
+      "loss": 0.4377,
+      "step": 2575
+    },
+    {
+      "epoch": 0.19161498755987968,
+      "grad_norm": 0.972554087638855,
+      "learning_rate": 4.9434351951192956e-05,
+      "loss": 0.444,
+      "step": 2580
+    },
+    {
+      "epoch": 0.19198633443499574,
+      "grad_norm": 1.0811034440994263,
+      "learning_rate": 4.943116748061106e-05,
+      "loss": 0.4336,
+      "step": 2585
+    },
+    {
+      "epoch": 0.19235768131011177,
+      "grad_norm": 0.8871290683746338,
+      "learning_rate": 4.942797417439722e-05,
+      "loss": 0.4537,
+      "step": 2590
+    },
+    {
+      "epoch": 0.19272902818522783,
+      "grad_norm": 1.0862644910812378,
+      "learning_rate": 4.942477203370631e-05,
+      "loss": 0.4406,
+      "step": 2595
+    },
+    {
+      "epoch": 0.19310037506034386,
+      "grad_norm": 0.9374284744262695,
+      "learning_rate": 4.9421561059696397e-05,
+      "loss": 0.4473,
+      "step": 2600
+    },
+    {
+      "epoch": 0.19347172193545992,
+      "grad_norm": 1.007503867149353,
+      "learning_rate": 4.941834125352874e-05,
+      "loss": 0.447,
+      "step": 2605
+    },
+    {
+      "epoch": 0.19384306881057595,
+      "grad_norm": 1.0377479791641235,
+      "learning_rate": 4.941511261636779e-05,
+      "loss": 0.4608,
+      "step": 2610
+    },
+    {
+      "epoch": 0.194214415685692,
+      "grad_norm": 1.0334117412567139,
+      "learning_rate": 4.941187514938122e-05,
+      "loss": 0.4527,
+      "step": 2615
+    },
+    {
+      "epoch": 0.19458576256080806,
+      "grad_norm": 1.2403839826583862,
+      "learning_rate": 4.940862885373984e-05,
+      "loss": 0.4355,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1949571094359241,
+      "grad_norm": 0.9511928558349609,
+      "learning_rate": 4.940537373061771e-05,
+      "loss": 0.4372,
+      "step": 2625
+    },
+    {
+      "epoch": 0.19532845631104015,
+      "grad_norm": 0.9966602325439453,
+      "learning_rate": 4.940210978119205e-05,
+      "loss": 0.4451,
+      "step": 2630
+    },
+    {
+      "epoch": 0.19569980318615618,
+      "grad_norm": 0.8612943291664124,
+      "learning_rate": 4.9398837006643286e-05,
+      "loss": 0.4587,
+      "step": 2635
+    },
+    {
+      "epoch": 0.19607115006127224,
+      "grad_norm": 1.1878662109375,
+      "learning_rate": 4.939555540815502e-05,
+      "loss": 0.468,
+      "step": 2640
+    },
+    {
+      "epoch": 0.19644249693638827,
+      "grad_norm": 0.9826651215553284,
+      "learning_rate": 4.939226498691407e-05,
+      "loss": 0.4819,
+      "step": 2645
+    },
+    {
+      "epoch": 0.19681384381150432,
+      "grad_norm": 1.0273140668869019,
+      "learning_rate": 4.9388965744110416e-05,
+      "loss": 0.4426,
+      "step": 2650
+    },
+    {
+      "epoch": 0.19718519068662038,
+      "grad_norm": 1.0519393682479858,
+      "learning_rate": 4.9385657680937246e-05,
+      "loss": 0.456,
+      "step": 2655
+    },
+    {
+      "epoch": 0.1975565375617364,
+      "grad_norm": 0.8215546607971191,
+      "learning_rate": 4.938234079859094e-05,
+      "loss": 0.4444,
+      "step": 2660
+    },
+    {
+      "epoch": 0.19792788443685247,
+      "grad_norm": 0.9723137021064758,
+      "learning_rate": 4.937901509827107e-05,
+      "loss": 0.4561,
+      "step": 2665
+    },
+    {
+      "epoch": 0.1982992313119685,
+      "grad_norm": 1.0705112218856812,
+      "learning_rate": 4.9375680581180375e-05,
+      "loss": 0.4512,
+      "step": 2670
+    },
+    {
+      "epoch": 0.19867057818708456,
+      "grad_norm": 0.9131028652191162,
+      "learning_rate": 4.937233724852479e-05,
+      "loss": 0.4496,
+      "step": 2675
+    },
+    {
+      "epoch": 0.19904192506220061,
+      "grad_norm": 0.9114557504653931,
+      "learning_rate": 4.936898510151346e-05,
+      "loss": 0.4638,
+      "step": 2680
+    },
+    {
+      "epoch": 0.19941327193731664,
+      "grad_norm": 0.7485722899436951,
+      "learning_rate": 4.9365624141358706e-05,
+      "loss": 0.4399,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1997846188124327,
+      "grad_norm": 1.0070992708206177,
+      "learning_rate": 4.936225436927602e-05,
+      "loss": 0.4228,
+      "step": 2690
+    },
+    {
+      "epoch": 0.20015596568754873,
+      "grad_norm": 1.330584168434143,
+      "learning_rate": 4.93588757864841e-05,
+      "loss": 0.438,
+      "step": 2695
+    },
+    {
+      "epoch": 0.2005273125626648,
+      "grad_norm": 1.0673860311508179,
+      "learning_rate": 4.935548839420483e-05,
+      "loss": 0.4643,
+      "step": 2700
+    },
+    {
+      "epoch": 0.20089865943778082,
+      "grad_norm": 1.4085111618041992,
+      "learning_rate": 4.935209219366326e-05,
+      "loss": 0.4395,
+      "step": 2705
+    },
+    {
+      "epoch": 0.20127000631289688,
+      "grad_norm": 0.8586167693138123,
+      "learning_rate": 4.934868718608765e-05,
+      "loss": 0.4578,
+      "step": 2710
+    },
+    {
+      "epoch": 0.20164135318801293,
+      "grad_norm": 1.0641870498657227,
+      "learning_rate": 4.9345273372709435e-05,
+      "loss": 0.446,
+      "step": 2715
+    },
+    {
+      "epoch": 0.20201270006312896,
+      "grad_norm": 0.8288577198982239,
+      "learning_rate": 4.934185075476323e-05,
+      "loss": 0.4267,
+      "step": 2720
+    },
+    {
+      "epoch": 0.20238404693824502,
+      "grad_norm": 0.9544481039047241,
+      "learning_rate": 4.933841933348684e-05,
+      "loss": 0.4445,
+      "step": 2725
+    },
+    {
+      "epoch": 0.20275539381336105,
+      "grad_norm": 0.9621331095695496,
+      "learning_rate": 4.933497911012126e-05,
+      "loss": 0.4276,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2031267406884771,
+      "grad_norm": 0.9601415395736694,
+      "learning_rate": 4.933153008591065e-05,
+      "loss": 0.4706,
+      "step": 2735
+    },
+    {
+      "epoch": 0.20349808756359314,
+      "grad_norm": 1.2368193864822388,
+      "learning_rate": 4.932807226210237e-05,
+      "loss": 0.4305,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2038694344387092,
+      "grad_norm": 1.1494923830032349,
+      "learning_rate": 4.932460563994695e-05,
+      "loss": 0.4529,
+      "step": 2745
+    },
+    {
+      "epoch": 0.20424078131382525,
+      "grad_norm": 0.9575107097625732,
+      "learning_rate": 4.932113022069812e-05,
+      "loss": 0.4683,
+      "step": 2750
+    },
+    {
+      "epoch": 0.20461212818894128,
+      "grad_norm": 1.0877399444580078,
+      "learning_rate": 4.931764600561277e-05,
+      "loss": 0.4704,
+      "step": 2755
+    },
+    {
+      "epoch": 0.20498347506405734,
+      "grad_norm": 0.8758722543716431,
+      "learning_rate": 4.931415299595098e-05,
+      "loss": 0.4592,
+      "step": 2760
+    },
+    {
+      "epoch": 0.20535482193917337,
+      "grad_norm": 1.0770483016967773,
+      "learning_rate": 4.9310651192976014e-05,
+      "loss": 0.4485,
+      "step": 2765
+    },
+    {
+      "epoch": 0.20572616881428943,
+      "grad_norm": 0.9201017022132874,
+      "learning_rate": 4.9307140597954314e-05,
+      "loss": 0.4404,
+      "step": 2770
+    },
+    {
+      "epoch": 0.20609751568940549,
+      "grad_norm": 0.8507365584373474,
+      "learning_rate": 4.9303621212155506e-05,
+      "loss": 0.4565,
+      "step": 2775
+    },
+    {
+      "epoch": 0.20646886256452152,
+      "grad_norm": 0.9411526322364807,
+      "learning_rate": 4.9300093036852384e-05,
+      "loss": 0.4393,
+      "step": 2780
+    },
+    {
+      "epoch": 0.20684020943963757,
+      "grad_norm": 1.0703768730163574,
+      "learning_rate": 4.9296556073320925e-05,
+      "loss": 0.4297,
+      "step": 2785
+    },
+    {
+      "epoch": 0.2072115563147536,
+      "grad_norm": 1.1285152435302734,
+      "learning_rate": 4.92930103228403e-05,
+      "loss": 0.4269,
+      "step": 2790
+    },
+    {
+      "epoch": 0.20758290318986966,
+      "grad_norm": 1.2548874616622925,
+      "learning_rate": 4.928945578669282e-05,
+      "loss": 0.4486,
+      "step": 2795
+    },
+    {
+      "epoch": 0.2079542500649857,
+      "grad_norm": 0.824675440788269,
+      "learning_rate": 4.928589246616402e-05,
+      "loss": 0.4543,
+      "step": 2800
+    },
+    {
+      "epoch": 0.20832559694010175,
+      "grad_norm": 1.1596276760101318,
+      "learning_rate": 4.928232036254257e-05,
+      "loss": 0.4291,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2086969438152178,
+      "grad_norm": 0.7505876421928406,
+      "learning_rate": 4.9278739477120364e-05,
+      "loss": 0.4275,
+      "step": 2810
+    },
+    {
+      "epoch": 0.20906829069033384,
+      "grad_norm": 1.3590447902679443,
+      "learning_rate": 4.9275149811192414e-05,
+      "loss": 0.4727,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2094396375654499,
+      "grad_norm": 0.9690061807632446,
+      "learning_rate": 4.9271551366056954e-05,
+      "loss": 0.4522,
+      "step": 2820
+    },
+    {
+      "epoch": 0.20981098444056592,
+      "grad_norm": 1.082969307899475,
+      "learning_rate": 4.9267944143015364e-05,
+      "loss": 0.4563,
+      "step": 2825
+    },
+    {
+      "epoch": 0.21018233131568198,
+      "grad_norm": 0.8886339664459229,
+      "learning_rate": 4.926432814337223e-05,
+      "loss": 0.4441,
+      "step": 2830
+    },
+    {
+      "epoch": 0.21055367819079804,
+      "grad_norm": 0.8670248985290527,
+      "learning_rate": 4.926070336843528e-05,
+      "loss": 0.4481,
+      "step": 2835
+    },
+    {
+      "epoch": 0.21092502506591407,
+      "grad_norm": 0.8443616032600403,
+      "learning_rate": 4.925706981951543e-05,
+      "loss": 0.4475,
+      "step": 2840
+    },
+    {
+      "epoch": 0.21129637194103013,
+      "grad_norm": 0.9135586023330688,
+      "learning_rate": 4.925342749792677e-05,
+      "loss": 0.4618,
+      "step": 2845
+    },
+    {
+      "epoch": 0.21166771881614616,
+      "grad_norm": 1.1337090730667114,
+      "learning_rate": 4.924977640498657e-05,
+      "loss": 0.4454,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2120390656912622,
+      "grad_norm": 0.8965758085250854,
+      "learning_rate": 4.9246116542015244e-05,
+      "loss": 0.4554,
+      "step": 2855
+    },
+    {
+      "epoch": 0.21241041256637824,
+      "grad_norm": 0.9783238768577576,
+      "learning_rate": 4.92424479103364e-05,
+      "loss": 0.441,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2127817594414943,
+      "grad_norm": 0.8907028436660767,
+      "learning_rate": 4.9238770511276824e-05,
+      "loss": 0.436,
+      "step": 2865
+    },
+    {
+      "epoch": 0.21315310631661036,
+      "grad_norm": 1.0488508939743042,
+      "learning_rate": 4.9235084346166464e-05,
+      "loss": 0.4247,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2135244531917264,
+      "grad_norm": 0.9586307406425476,
+      "learning_rate": 4.923138941633841e-05,
+      "loss": 0.4512,
+      "step": 2875
+    },
+    {
+      "epoch": 0.21389580006684245,
+      "grad_norm": 1.152602195739746,
+      "learning_rate": 4.922768572312899e-05,
+      "loss": 0.4569,
+      "step": 2880
+    },
+    {
+      "epoch": 0.21426714694195848,
+      "grad_norm": 1.1480371952056885,
+      "learning_rate": 4.922397326787763e-05,
+      "loss": 0.4287,
+      "step": 2885
+    },
+    {
+      "epoch": 0.21463849381707453,
+      "grad_norm": 1.0389635562896729,
+      "learning_rate": 4.9220252051926964e-05,
+      "loss": 0.4553,
+      "step": 2890
+    },
+    {
+      "epoch": 0.21500984069219056,
+      "grad_norm": 0.8871331214904785,
+      "learning_rate": 4.9216522076622776e-05,
+      "loss": 0.4579,
+      "step": 2895
+    },
+    {
+      "epoch": 0.21538118756730662,
+      "grad_norm": 0.7509046196937561,
+      "learning_rate": 4.921278334331404e-05,
+      "loss": 0.4398,
+      "step": 2900
+    },
+    {
+      "epoch": 0.21575253444242268,
+      "grad_norm": 1.3356949090957642,
+      "learning_rate": 4.920903585335288e-05,
+      "loss": 0.4378,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2161238813175387,
+      "grad_norm": 0.9830528497695923,
+      "learning_rate": 4.920527960809459e-05,
+      "loss": 0.4623,
+      "step": 2910
+    },
+    {
+      "epoch": 0.21649522819265477,
+      "grad_norm": 1.6082857847213745,
+      "learning_rate": 4.9201514608897635e-05,
+      "loss": 0.4316,
+      "step": 2915
+    },
+    {
+      "epoch": 0.2168665750677708,
+      "grad_norm": 0.8056818842887878,
+      "learning_rate": 4.9197740857123645e-05,
+      "loss": 0.4257,
+      "step": 2920
+    },
+    {
+      "epoch": 0.21723792194288685,
+      "grad_norm": 0.8548802733421326,
+      "learning_rate": 4.919395835413739e-05,
+      "loss": 0.4382,
+      "step": 2925
+    },
+    {
+      "epoch": 0.2176092688180029,
+      "grad_norm": 1.2386173009872437,
+      "learning_rate": 4.919016710130685e-05,
+      "loss": 0.4366,
+      "step": 2930
+    },
+    {
+      "epoch": 0.21798061569311894,
+      "grad_norm": 1.0536203384399414,
+      "learning_rate": 4.918636710000315e-05,
+      "loss": 0.4362,
+      "step": 2935
+    },
+    {
+      "epoch": 0.218351962568235,
+      "grad_norm": 1.0775947570800781,
+      "learning_rate": 4.918255835160056e-05,
+      "loss": 0.4301,
+      "step": 2940
+    },
+    {
+      "epoch": 0.21872330944335103,
+      "grad_norm": 1.2517681121826172,
+      "learning_rate": 4.9178740857476545e-05,
+      "loss": 0.4007,
+      "step": 2945
+    },
+    {
+      "epoch": 0.21909465631846708,
+      "grad_norm": 0.9359631538391113,
+      "learning_rate": 4.9174914619011706e-05,
+      "loss": 0.4347,
+      "step": 2950
+    },
+    {
+      "epoch": 0.21946600319358311,
+      "grad_norm": 1.2085964679718018,
+      "learning_rate": 4.9171079637589824e-05,
+      "loss": 0.447,
+      "step": 2955
+    },
+    {
+      "epoch": 0.21983735006869917,
+      "grad_norm": 1.3199816942214966,
+      "learning_rate": 4.9167235914597826e-05,
+      "loss": 0.4585,
+      "step": 2960
+    },
+    {
+      "epoch": 0.22020869694381523,
+      "grad_norm": 1.0172200202941895,
+      "learning_rate": 4.916338345142582e-05,
+      "loss": 0.427,
+      "step": 2965
+    },
+    {
+      "epoch": 0.22058004381893126,
+      "grad_norm": 0.9850211143493652,
+      "learning_rate": 4.915952224946707e-05,
+      "loss": 0.4373,
+      "step": 2970
+    },
+    {
+      "epoch": 0.22095139069404732,
+      "grad_norm": 0.7755311727523804,
+      "learning_rate": 4.9155652310117976e-05,
+      "loss": 0.4038,
+      "step": 2975
+    },
+    {
+      "epoch": 0.22132273756916335,
+      "grad_norm": 1.2828562259674072,
+      "learning_rate": 4.9151773634778145e-05,
+      "loss": 0.4351,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2216940844442794,
+      "grad_norm": 0.7583413124084473,
+      "learning_rate": 4.914788622485028e-05,
+      "loss": 0.425,
+      "step": 2985
+    },
+    {
+      "epoch": 0.22206543131939543,
+      "grad_norm": 0.9321625828742981,
+      "learning_rate": 4.9143990081740314e-05,
+      "loss": 0.4344,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2224367781945115,
+      "grad_norm": 1.1756302118301392,
+      "learning_rate": 4.914008520685729e-05,
+      "loss": 0.4594,
+      "step": 2995
+    },
+    {
+      "epoch": 0.22280812506962755,
+      "grad_norm": 1.289210557937622,
+      "learning_rate": 4.913617160161341e-05,
+      "loss": 0.4596,
+      "step": 3000
+    },
+    {
+      "epoch": 0.22317947194474358,
+      "grad_norm": 0.8851030468940735,
+      "learning_rate": 4.913224926742407e-05,
+      "loss": 0.4429,
+      "step": 3005
+    },
+    {
+      "epoch": 0.22355081881985964,
+      "grad_norm": 0.9165136218070984,
+      "learning_rate": 4.912831820570777e-05,
+      "loss": 0.444,
+      "step": 3010
+    },
+    {
+      "epoch": 0.22392216569497567,
+      "grad_norm": 1.077898621559143,
+      "learning_rate": 4.912437841788622e-05,
+      "loss": 0.4279,
+      "step": 3015
+    },
+    {
+      "epoch": 0.22429351257009172,
+      "grad_norm": 0.9308784008026123,
+      "learning_rate": 4.912042990538424e-05,
+      "loss": 0.4445,
+      "step": 3020
+    },
+    {
+      "epoch": 0.22466485944520778,
+      "grad_norm": 0.9209590554237366,
+      "learning_rate": 4.911647266962984e-05,
+      "loss": 0.447,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2250362063203238,
+      "grad_norm": 0.7767472267150879,
+      "learning_rate": 4.9112506712054165e-05,
+      "loss": 0.4474,
+      "step": 3030
+    },
+    {
+      "epoch": 0.22540755319543987,
+      "grad_norm": 1.1049778461456299,
+      "learning_rate": 4.910853203409152e-05,
+      "loss": 0.4665,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2257789000705559,
+      "grad_norm": 0.803619921207428,
+      "learning_rate": 4.910454863717936e-05,
+      "loss": 0.4522,
+      "step": 3040
+    },
+    {
+      "epoch": 0.22615024694567196,
+      "grad_norm": 1.1416994333267212,
+      "learning_rate": 4.91005565227583e-05,
+      "loss": 0.4509,
+      "step": 3045
+    },
+    {
+      "epoch": 0.226521593820788,
+      "grad_norm": 0.9570928812026978,
+      "learning_rate": 4.90965556922721e-05,
+      "loss": 0.4321,
+      "step": 3050
+    },
+    {
+      "epoch": 0.22689294069590404,
+      "grad_norm": 0.979243814945221,
+      "learning_rate": 4.9092546147167696e-05,
+      "loss": 0.4453,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2272642875710201,
+      "grad_norm": 1.013358473777771,
+      "learning_rate": 4.908852788889514e-05,
+      "loss": 0.45,
+      "step": 3060
+    },
+    {
+      "epoch": 0.22763563444613613,
+      "grad_norm": 1.1085175275802612,
+      "learning_rate": 4.908450091890765e-05,
+      "loss": 0.442,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2280069813212522,
+      "grad_norm": 0.9845399260520935,
+      "learning_rate": 4.90804652386616e-05,
+      "loss": 0.4279,
+      "step": 3070
+    },
+    {
+      "epoch": 0.22837832819636822,
+      "grad_norm": 0.9485452771186829,
+      "learning_rate": 4.907642084961651e-05,
+      "loss": 0.4463,
+      "step": 3075
+    },
+    {
+      "epoch": 0.22874967507148428,
+      "grad_norm": 1.0089235305786133,
+      "learning_rate": 4.907236775323505e-05,
+      "loss": 0.4399,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2291210219466003,
+      "grad_norm": 1.4188969135284424,
+      "learning_rate": 4.906830595098304e-05,
+      "loss": 0.4251,
+      "step": 3085
+    },
+    {
+      "epoch": 0.22949236882171636,
+      "grad_norm": 1.038788080215454,
+      "learning_rate": 4.9064235444329455e-05,
+      "loss": 0.4493,
+      "step": 3090
+    },
+    {
+      "epoch": 0.22986371569683242,
+      "grad_norm": 1.14411461353302,
+      "learning_rate": 4.9060156234746394e-05,
+      "loss": 0.4607,
+      "step": 3095
+    },
+    {
+      "epoch": 0.23023506257194845,
+      "grad_norm": 1.0400830507278442,
+      "learning_rate": 4.905606832370912e-05,
+      "loss": 0.427,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2306064094470645,
+      "grad_norm": 0.7747494578361511,
+      "learning_rate": 4.905197171269607e-05,
+      "loss": 0.4404,
+      "step": 3105
+    },
+    {
+      "epoch": 0.23097775632218054,
+      "grad_norm": 0.9267304539680481,
+      "learning_rate": 4.904786640318876e-05,
+      "loss": 0.4644,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2313491031972966,
+      "grad_norm": 1.1972533464431763,
+      "learning_rate": 4.904375239667193e-05,
+      "loss": 0.4521,
+      "step": 3115
+    },
+    {
+      "epoch": 0.23172045007241265,
+      "grad_norm": 0.8945311307907104,
+      "learning_rate": 4.903962969463341e-05,
+      "loss": 0.4202,
+      "step": 3120
+    },
+    {
+      "epoch": 0.23209179694752868,
+      "grad_norm": 0.8536193370819092,
+      "learning_rate": 4.9035498298564175e-05,
+      "loss": 0.4504,
+      "step": 3125
+    },
+    {
+      "epoch": 0.23246314382264474,
+      "grad_norm": 0.886146605014801,
+      "learning_rate": 4.903135820995839e-05,
+      "loss": 0.4499,
+      "step": 3130
+    },
+    {
+      "epoch": 0.23283449069776077,
+      "grad_norm": 0.8584529161453247,
+      "learning_rate": 4.9027209430313325e-05,
+      "loss": 0.4066,
+      "step": 3135
+    },
+    {
+      "epoch": 0.23320583757287683,
+      "grad_norm": 2.4391167163848877,
+      "learning_rate": 4.9023051961129404e-05,
+      "loss": 0.4334,
+      "step": 3140
+    },
+    {
+      "epoch": 0.23357718444799286,
+      "grad_norm": 1.1037994623184204,
+      "learning_rate": 4.9018885803910186e-05,
+      "loss": 0.4382,
+      "step": 3145
+    },
+    {
+      "epoch": 0.23394853132310892,
+      "grad_norm": 0.8490164279937744,
+      "learning_rate": 4.9014710960162394e-05,
+      "loss": 0.4394,
+      "step": 3150
+    },
+    {
+      "epoch": 0.23431987819822497,
+      "grad_norm": 1.031388282775879,
+      "learning_rate": 4.901052743139586e-05,
+      "loss": 0.4654,
+      "step": 3155
+    },
+    {
+      "epoch": 0.234691225073341,
+      "grad_norm": 0.8660055994987488,
+      "learning_rate": 4.9006335219123576e-05,
+      "loss": 0.4347,
+      "step": 3160
+    },
+    {
+      "epoch": 0.23506257194845706,
+      "grad_norm": 0.9747069478034973,
+      "learning_rate": 4.900213432486168e-05,
+      "loss": 0.4359,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2354339188235731,
+      "grad_norm": 0.9307319521903992,
+      "learning_rate": 4.8997924750129444e-05,
+      "loss": 0.4363,
+      "step": 3170
+    },
+    {
+      "epoch": 0.23580526569868915,
+      "grad_norm": 1.0029345750808716,
+      "learning_rate": 4.899370649644928e-05,
+      "loss": 0.4549,
+      "step": 3175
+    },
+    {
+      "epoch": 0.23617661257380518,
+      "grad_norm": 1.0878911018371582,
+      "learning_rate": 4.898947956534672e-05,
+      "loss": 0.4557,
+      "step": 3180
+    },
+    {
+      "epoch": 0.23654795944892124,
+      "grad_norm": 1.0974375009536743,
+      "learning_rate": 4.898524395835047e-05,
+      "loss": 0.4555,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2369193063240373,
+      "grad_norm": 1.0777721405029297,
+      "learning_rate": 4.8980999676992345e-05,
+      "loss": 0.4376,
+      "step": 3190
+    },
+    {
+      "epoch": 0.23729065319915332,
+      "grad_norm": 0.9894423484802246,
+      "learning_rate": 4.897674672280731e-05,
+      "loss": 0.4296,
+      "step": 3195
+    },
+    {
+      "epoch": 0.23766200007426938,
+      "grad_norm": 0.9134079813957214,
+      "learning_rate": 4.897248509733346e-05,
+      "loss": 0.4359,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2380333469493854,
+      "grad_norm": 1.1131818294525146,
+      "learning_rate": 4.8968214802112034e-05,
+      "loss": 0.4543,
+      "step": 3205
+    },
+    {
+      "epoch": 0.23840469382450147,
+      "grad_norm": 1.0237282514572144,
+      "learning_rate": 4.89639358386874e-05,
+      "loss": 0.43,
+      "step": 3210
+    },
+    {
+      "epoch": 0.23877604069961753,
+      "grad_norm": 0.9851337671279907,
+      "learning_rate": 4.895964820860705e-05,
+      "loss": 0.4319,
+      "step": 3215
+    },
+    {
+      "epoch": 0.23914738757473356,
+      "grad_norm": 0.8571179509162903,
+      "learning_rate": 4.8955351913421656e-05,
+      "loss": 0.4303,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2395187344498496,
+      "grad_norm": 0.9178614616394043,
+      "learning_rate": 4.895104695468495e-05,
+      "loss": 0.4199,
+      "step": 3225
+    },
+    {
+      "epoch": 0.23989008132496564,
+      "grad_norm": 0.7542501091957092,
+      "learning_rate": 4.894673333395387e-05,
+      "loss": 0.4335,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2402614282000817,
+      "grad_norm": 0.9010379910469055,
+      "learning_rate": 4.894241105278844e-05,
+      "loss": 0.4334,
+      "step": 3235
+    },
+    {
+      "epoch": 0.24063277507519773,
+      "grad_norm": 1.1573511362075806,
+      "learning_rate": 4.893808011275183e-05,
+      "loss": 0.4202,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2410041219503138,
+      "grad_norm": 0.9232560992240906,
+      "learning_rate": 4.893374051541035e-05,
+      "loss": 0.4202,
+      "step": 3245
+    },
+    {
+      "epoch": 0.24137546882542985,
+      "grad_norm": 0.9102317094802856,
+      "learning_rate": 4.8929392262333426e-05,
+      "loss": 0.4414,
+      "step": 3250
+    },
+    {
+      "epoch": 0.24174681570054588,
+      "grad_norm": 0.8203946948051453,
+      "learning_rate": 4.892503535509363e-05,
+      "loss": 0.4186,
+      "step": 3255
+    },
+    {
+      "epoch": 0.24211816257566193,
+      "grad_norm": 0.802352786064148,
+      "learning_rate": 4.892066979526664e-05,
+      "loss": 0.4426,
+      "step": 3260
+    },
+    {
+      "epoch": 0.24248950945077796,
+      "grad_norm": 0.8278765678405762,
+      "learning_rate": 4.891629558443129e-05,
+      "loss": 0.4114,
+      "step": 3265
+    },
+    {
+      "epoch": 0.24286085632589402,
+      "grad_norm": 1.7212625741958618,
+      "learning_rate": 4.891191272416953e-05,
+      "loss": 0.438,
+      "step": 3270
+    },
+    {
+      "epoch": 0.24323220320101005,
+      "grad_norm": 5.098775386810303,
+      "learning_rate": 4.8907521216066446e-05,
+      "loss": 0.458,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2436035500761261,
+      "grad_norm": 0.9874866604804993,
+      "learning_rate": 4.890312106171023e-05,
+      "loss": 0.4227,
+      "step": 3280
+    },
+    {
+      "epoch": 0.24397489695124217,
+      "grad_norm": 1.151456594467163,
+      "learning_rate": 4.8898712262692225e-05,
+      "loss": 0.4593,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2443462438263582,
+      "grad_norm": 0.8587061762809753,
+      "learning_rate": 4.889429482060689e-05,
+      "loss": 0.4161,
+      "step": 3290
+    },
+    {
+      "epoch": 0.24471759070147425,
+      "grad_norm": 1.2171800136566162,
+      "learning_rate": 4.888986873705181e-05,
+      "loss": 0.4358,
+      "step": 3295
+    },
+    {
+      "epoch": 0.24508893757659028,
+      "grad_norm": 0.8855735063552856,
+      "learning_rate": 4.888543401362771e-05,
+      "loss": 0.4328,
+      "step": 3300
+    },
+    {
+      "epoch": 0.24546028445170634,
+      "grad_norm": 1.1113088130950928,
+      "learning_rate": 4.88809906519384e-05,
+      "loss": 0.4259,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2458316313268224,
+      "grad_norm": 6.572192668914795,
+      "learning_rate": 4.887653865359085e-05,
+      "loss": 0.4463,
+      "step": 3310
+    },
+    {
+      "epoch": 0.24620297820193843,
+      "grad_norm": 1.3304423093795776,
+      "learning_rate": 4.887207802019515e-05,
+      "loss": 0.4284,
+      "step": 3315
+    },
+    {
+      "epoch": 0.24657432507705448,
+      "grad_norm": 0.8548965454101562,
+      "learning_rate": 4.88676087533645e-05,
+      "loss": 0.4335,
+      "step": 3320
+    },
+    {
+      "epoch": 0.24694567195217051,
+      "grad_norm": 1.166911244392395,
+      "learning_rate": 4.886313085471524e-05,
+      "loss": 0.4364,
+      "step": 3325
+    },
+    {
+      "epoch": 0.24731701882728657,
+      "grad_norm": 1.1613044738769531,
+      "learning_rate": 4.885864432586681e-05,
+      "loss": 0.4326,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2476883657024026,
+      "grad_norm": 0.9552707076072693,
+      "learning_rate": 4.8854149168441776e-05,
+      "loss": 0.4353,
+      "step": 3335
+    },
+    {
+      "epoch": 0.24805971257751866,
+      "grad_norm": 0.9253180027008057,
+      "learning_rate": 4.8849645384065843e-05,
+      "loss": 0.4177,
+      "step": 3340
+    },
+    {
+      "epoch": 0.24843105945263472,
+      "grad_norm": 0.9358575344085693,
+      "learning_rate": 4.884513297436782e-05,
+      "loss": 0.4393,
+      "step": 3345
+    },
+    {
+      "epoch": 0.24880240632775075,
+      "grad_norm": 0.9707365036010742,
+      "learning_rate": 4.884061194097963e-05,
+      "loss": 0.4308,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2491737532028668,
+      "grad_norm": 0.8588376641273499,
+      "learning_rate": 4.883608228553635e-05,
+      "loss": 0.4028,
+      "step": 3355
+    },
+    {
+      "epoch": 0.24954510007798283,
+      "grad_norm": 0.9038694500923157,
+      "learning_rate": 4.88315440096761e-05,
+      "loss": 0.4524,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2499164469530989,
+      "grad_norm": 1.0825928449630737,
+      "learning_rate": 4.882699711504021e-05,
+      "loss": 0.4372,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2502877938282149,
+      "grad_norm": 1.158320665359497,
+      "learning_rate": 4.882244160327306e-05,
+      "loss": 0.418,
+      "step": 3370
+    },
+    {
+      "epoch": 0.250659140703331,
+      "grad_norm": 1.276110291481018,
+      "learning_rate": 4.881787747602218e-05,
+      "loss": 0.4545,
+      "step": 3375
+    },
+    {
+      "epoch": 0.25103048757844704,
+      "grad_norm": 1.0024749040603638,
+      "learning_rate": 4.8813304734938206e-05,
+      "loss": 0.4674,
+      "step": 3380
+    },
+    {
+      "epoch": 0.25140183445356307,
+      "grad_norm": 1.1805554628372192,
+      "learning_rate": 4.8808723381674884e-05,
+      "loss": 0.4638,
+      "step": 3385
+    },
+    {
+      "epoch": 0.2517731813286791,
+      "grad_norm": 1.3252657651901245,
+      "learning_rate": 4.880413341788907e-05,
+      "loss": 0.4241,
+      "step": 3390
+    },
+    {
+      "epoch": 0.2521445282037952,
+      "grad_norm": 1.0752530097961426,
+      "learning_rate": 4.879953484524076e-05,
+      "loss": 0.4416,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2525158750789112,
+      "grad_norm": 1.014775276184082,
+      "learning_rate": 4.8794927665393044e-05,
+      "loss": 0.4295,
+      "step": 3400
+    },
+    {
+      "epoch": 0.25288722195402724,
+      "grad_norm": 0.9509404301643372,
+      "learning_rate": 4.879031188001211e-05,
+      "loss": 0.427,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2532585688291433,
+      "grad_norm": 1.2720471620559692,
+      "learning_rate": 4.87856874907673e-05,
+      "loss": 0.4238,
+      "step": 3410
+    },
+    {
+      "epoch": 0.25362991570425936,
+      "grad_norm": 1.0840013027191162,
+      "learning_rate": 4.878105449933103e-05,
+      "loss": 0.4166,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2540012625793754,
+      "grad_norm": 0.956425666809082,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.4203,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2543726094544914,
+      "grad_norm": 1.113987684249878,
+      "learning_rate": 4.877176271658939e-05,
+      "loss": 0.4318,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2547439563296075,
+      "grad_norm": 1.0058467388153076,
+      "learning_rate": 4.8767103928644434e-05,
+      "loss": 0.4475,
+      "step": 3430
+    },
+    {
+      "epoch": 0.25511530320472353,
+      "grad_norm": 1.0395872592926025,
+      "learning_rate": 4.876243654522883e-05,
+      "loss": 0.4322,
+      "step": 3435
+    },
+    {
+      "epoch": 0.25548665007983956,
+      "grad_norm": 0.9146986603736877,
+      "learning_rate": 4.8757760568030574e-05,
+      "loss": 0.4309,
+      "step": 3440
+    },
+    {
+      "epoch": 0.25585799695495565,
+      "grad_norm": 0.9177193641662598,
+      "learning_rate": 4.875307599874075e-05,
+      "loss": 0.4369,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2562293438300717,
+      "grad_norm": 0.9752681255340576,
+      "learning_rate": 4.8748382839053544e-05,
+      "loss": 0.4264,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2566006907051877,
+      "grad_norm": 0.9808909296989441,
+      "learning_rate": 4.8743681090666265e-05,
+      "loss": 0.4313,
+      "step": 3455
+    },
+    {
+      "epoch": 0.25697203758030374,
+      "grad_norm": 0.8328979015350342,
+      "learning_rate": 4.873897075527931e-05,
+      "loss": 0.4325,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2573433844554198,
+      "grad_norm": 0.925728976726532,
+      "learning_rate": 4.87342518345962e-05,
+      "loss": 0.4305,
+      "step": 3465
+    },
+    {
+      "epoch": 0.25771473133053585,
+      "grad_norm": 0.9330899119377136,
+      "learning_rate": 4.872952433032355e-05,
+      "loss": 0.4218,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2580860782056519,
+      "grad_norm": 0.9173476696014404,
+      "learning_rate": 4.8724788244171076e-05,
+      "loss": 0.4265,
+      "step": 3475
+    },
+    {
+      "epoch": 0.25845742508076797,
+      "grad_norm": 1.3077492713928223,
+      "learning_rate": 4.872004357785161e-05,
+      "loss": 0.4421,
+      "step": 3480
+    },
+    {
+      "epoch": 0.258828771955884,
+      "grad_norm": 1.0578168630599976,
+      "learning_rate": 4.8715290333081073e-05,
+      "loss": 0.4426,
+      "step": 3485
+    },
+    {
+      "epoch": 0.259200118831,
+      "grad_norm": 0.7287719249725342,
+      "learning_rate": 4.87105285115785e-05,
+      "loss": 0.4407,
+      "step": 3490
+    },
+    {
+      "epoch": 0.25957146570611606,
+      "grad_norm": 0.7943789958953857,
+      "learning_rate": 4.8705758115066026e-05,
+      "loss": 0.4297,
+      "step": 3495
+    },
+    {
+      "epoch": 0.25994281258123214,
+      "grad_norm": 0.9521101713180542,
+      "learning_rate": 4.870097914526888e-05,
+      "loss": 0.4287,
+      "step": 3500
+    },
+    {
+      "epoch": 0.26031415945634817,
+      "grad_norm": 1.0764399766921997,
+      "learning_rate": 4.8696191603915394e-05,
+      "loss": 0.4339,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2606855063314642,
+      "grad_norm": 1.1681718826293945,
+      "learning_rate": 4.8691395492737e-05,
+      "loss": 0.4451,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2610568532065803,
+      "grad_norm": 0.9852285981178284,
+      "learning_rate": 4.8686590813468245e-05,
+      "loss": 0.4569,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2614282000816963,
+      "grad_norm": 0.9874268770217896,
+      "learning_rate": 4.868177756784676e-05,
+      "loss": 0.4192,
+      "step": 3520
+    },
+    {
+      "epoch": 0.26179954695681235,
+      "grad_norm": 1.0080324411392212,
+      "learning_rate": 4.867695575761324e-05,
+      "loss": 0.4347,
+      "step": 3525
+    },
+    {
+      "epoch": 0.26217089383192843,
+      "grad_norm": 0.9174315929412842,
+      "learning_rate": 4.867212538451156e-05,
+      "loss": 0.4194,
+      "step": 3530
+    },
+    {
+      "epoch": 0.26254224070704446,
+      "grad_norm": 1.1453148126602173,
+      "learning_rate": 4.866728645028862e-05,
+      "loss": 0.4419,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2629135875821605,
+      "grad_norm": 0.9571831226348877,
+      "learning_rate": 4.866243895669445e-05,
+      "loss": 0.4443,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2632849344572765,
+      "grad_norm": 0.7948081493377686,
+      "learning_rate": 4.8657582905482155e-05,
+      "loss": 0.4199,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2636562813323926,
+      "grad_norm": 0.9427078366279602,
+      "learning_rate": 4.8652718298407944e-05,
+      "loss": 0.4481,
+      "step": 3550
+    },
+    {
+      "epoch": 0.26402762820750864,
+      "grad_norm": 0.9686035513877869,
+      "learning_rate": 4.8647845137231143e-05,
+      "loss": 0.4431,
+      "step": 3555
+    },
+    {
+      "epoch": 0.26439897508262467,
+      "grad_norm": 1.433406949043274,
+      "learning_rate": 4.8642963423714124e-05,
+      "loss": 0.4271,
+      "step": 3560
+    },
+    {
+      "epoch": 0.26477032195774075,
+      "grad_norm": 1.2971220016479492,
+      "learning_rate": 4.86380731596224e-05,
+      "loss": 0.4361,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2651416688328568,
+      "grad_norm": 1.1293293237686157,
+      "learning_rate": 4.8633174346724544e-05,
+      "loss": 0.4512,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2655130157079728,
+      "grad_norm": 0.8800034523010254,
+      "learning_rate": 4.8628266986792235e-05,
+      "loss": 0.4351,
+      "step": 3575
+    },
+    {
+      "epoch": 0.26588436258308884,
+      "grad_norm": 0.7938306331634521,
+      "learning_rate": 4.862335108160024e-05,
+      "loss": 0.4311,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2662557094582049,
+      "grad_norm": 0.768816351890564,
+      "learning_rate": 4.861842663292641e-05,
+      "loss": 0.4229,
+      "step": 3585
+    },
+    {
+      "epoch": 0.26662705633332096,
+      "grad_norm": 0.7999665141105652,
+      "learning_rate": 4.86134936425517e-05,
+      "loss": 0.4431,
+      "step": 3590
+    },
+    {
+      "epoch": 0.266998403208437,
+      "grad_norm": 0.9142205119132996,
+      "learning_rate": 4.860855211226015e-05,
+      "loss": 0.4048,
+      "step": 3595
+    },
+    {
+      "epoch": 0.26736975008355307,
+      "grad_norm": 0.9010916352272034,
+      "learning_rate": 4.860360204383888e-05,
+      "loss": 0.4178,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2677410969586691,
+      "grad_norm": 0.7956218719482422,
+      "learning_rate": 4.859864343907809e-05,
+      "loss": 0.4451,
+      "step": 3605
+    },
+    {
+      "epoch": 0.26811244383378513,
+      "grad_norm": 0.9147886633872986,
+      "learning_rate": 4.8593676299771096e-05,
+      "loss": 0.4474,
+      "step": 3610
+    },
+    {
+      "epoch": 0.26848379070890116,
+      "grad_norm": 1.0486589670181274,
+      "learning_rate": 4.858870062771428e-05,
+      "loss": 0.4374,
+      "step": 3615
+    },
+    {
+      "epoch": 0.26885513758401725,
+      "grad_norm": 0.9408405423164368,
+      "learning_rate": 4.8583716424707115e-05,
+      "loss": 0.4216,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2692264844591333,
+      "grad_norm": 1.010460615158081,
+      "learning_rate": 4.8578723692552155e-05,
+      "loss": 0.4178,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2695978313342493,
+      "grad_norm": 0.8639962673187256,
+      "learning_rate": 4.857372243305504e-05,
+      "loss": 0.4383,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2699691782093654,
+      "grad_norm": 0.9397608041763306,
+      "learning_rate": 4.856871264802451e-05,
+      "loss": 0.4627,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2703405250844814,
+      "grad_norm": 0.8951156735420227,
+      "learning_rate": 4.856369433927235e-05,
+      "loss": 0.4427,
+      "step": 3640
+    },
+    {
+      "epoch": 0.27071187195959745,
+      "grad_norm": 2.018953561782837,
+      "learning_rate": 4.855866750861348e-05,
+      "loss": 0.4281,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2710832188347135,
+      "grad_norm": 1.3016127347946167,
+      "learning_rate": 4.855363215786585e-05,
+      "loss": 0.4172,
+      "step": 3650
+    },
+    {
+      "epoch": 0.27145456570982957,
+      "grad_norm": 0.979545533657074,
+      "learning_rate": 4.854858828885053e-05,
+      "loss": 0.4509,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2718259125849456,
+      "grad_norm": 0.9157105088233948,
+      "learning_rate": 4.8543535903391656e-05,
+      "loss": 0.4373,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2721972594600616,
+      "grad_norm": 0.8544624447822571,
+      "learning_rate": 4.853847500331643e-05,
+      "loss": 0.4296,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2725686063351777,
+      "grad_norm": 1.124678134918213,
+      "learning_rate": 4.853340559045516e-05,
+      "loss": 0.4154,
+      "step": 3670
+    },
+    {
+      "epoch": 0.27293995321029374,
+      "grad_norm": 0.9993969202041626,
+      "learning_rate": 4.852832766664122e-05,
+      "loss": 0.4447,
+      "step": 3675
+    },
+    {
+      "epoch": 0.27331130008540977,
+      "grad_norm": 0.8751777410507202,
+      "learning_rate": 4.8523241233711056e-05,
+      "loss": 0.4452,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2736826469605258,
+      "grad_norm": 1.4159234762191772,
+      "learning_rate": 4.85181462935042e-05,
+      "loss": 0.4125,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2740539938356419,
+      "grad_norm": 0.9526045918464661,
+      "learning_rate": 4.851304284786326e-05,
+      "loss": 0.4295,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2744253407107579,
+      "grad_norm": 0.8319068551063538,
+      "learning_rate": 4.8507930898633915e-05,
+      "loss": 0.4304,
+      "step": 3695
+    },
+    {
+      "epoch": 0.27479668758587394,
+      "grad_norm": 0.9949585795402527,
+      "learning_rate": 4.850281044766492e-05,
+      "loss": 0.4138,
+      "step": 3700
+    },
+    {
+      "epoch": 0.27516803446099003,
+      "grad_norm": 0.8212401866912842,
+      "learning_rate": 4.849768149680811e-05,
+      "loss": 0.4204,
+      "step": 3705
+    },
+    {
+      "epoch": 0.27553938133610606,
+      "grad_norm": 1.044382095336914,
+      "learning_rate": 4.84925440479184e-05,
+      "loss": 0.4224,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2759107282112221,
+      "grad_norm": 0.9368321299552917,
+      "learning_rate": 4.8487398102853754e-05,
+      "loss": 0.4434,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2762820750863382,
+      "grad_norm": 1.0776258707046509,
+      "learning_rate": 4.848224366347522e-05,
+      "loss": 0.4255,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2766534219614542,
+      "grad_norm": 0.8507828712463379,
+      "learning_rate": 4.847708073164695e-05,
+      "loss": 0.446,
+      "step": 3725
+    },
+    {
+      "epoch": 0.27702476883657023,
+      "grad_norm": 0.9348880648612976,
+      "learning_rate": 4.847190930923612e-05,
+      "loss": 0.3984,
+      "step": 3730
+    },
+    {
+      "epoch": 0.27739611571168626,
+      "grad_norm": 2.778374195098877,
+      "learning_rate": 4.8466729398113e-05,
+      "loss": 0.4018,
+      "step": 3735
+    },
+    {
+      "epoch": 0.27776746258680235,
+      "grad_norm": 0.9289271235466003,
+      "learning_rate": 4.8461541000150916e-05,
+      "loss": 0.4328,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2781388094619184,
+      "grad_norm": 0.7580967545509338,
+      "learning_rate": 4.8456344117226296e-05,
+      "loss": 0.404,
+      "step": 3745
+    },
+    {
+      "epoch": 0.2785101563370344,
+      "grad_norm": 0.7272197008132935,
+      "learning_rate": 4.8451138751218584e-05,
+      "loss": 0.442,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2788815032121505,
+      "grad_norm": 0.8026915192604065,
+      "learning_rate": 4.844592490401035e-05,
+      "loss": 0.4272,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2792528500872665,
+      "grad_norm": 0.8243690133094788,
+      "learning_rate": 4.8440702577487186e-05,
+      "loss": 0.4389,
+      "step": 3760
+    },
+    {
+      "epoch": 0.27962419696238255,
+      "grad_norm": 0.8651421070098877,
+      "learning_rate": 4.843547177353778e-05,
+      "loss": 0.4177,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2799955438374986,
+      "grad_norm": 2.0537188053131104,
+      "learning_rate": 4.8430232494053864e-05,
+      "loss": 0.4236,
+      "step": 3770
+    },
+    {
+      "epoch": 0.28036689071261467,
+      "grad_norm": 1.3934059143066406,
+      "learning_rate": 4.842498474093025e-05,
+      "loss": 0.4236,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2807382375877307,
+      "grad_norm": 1.1123522520065308,
+      "learning_rate": 4.84197285160648e-05,
+      "loss": 0.4236,
+      "step": 3780
+    },
+    {
+      "epoch": 0.28110958446284673,
+      "grad_norm": 1.1888607740402222,
+      "learning_rate": 4.841446382135846e-05,
+      "loss": 0.4224,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2814809313379628,
+      "grad_norm": 0.9268679618835449,
+      "learning_rate": 4.840919065871522e-05,
+      "loss": 0.4528,
+      "step": 3790
+    },
+    {
+      "epoch": 0.28185227821307884,
+      "grad_norm": 1.1020976305007935,
+      "learning_rate": 4.840390903004216e-05,
+      "loss": 0.427,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2822236250881949,
+      "grad_norm": 1.011352300643921,
+      "learning_rate": 4.839861893724937e-05,
+      "loss": 0.4511,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2825949719633109,
+      "grad_norm": 0.9751265645027161,
+      "learning_rate": 4.8393320382250076e-05,
+      "loss": 0.4277,
+      "step": 3805
+    },
+    {
+      "epoch": 0.282966318838427,
+      "grad_norm": 1.1336296796798706,
+      "learning_rate": 4.8388013366960484e-05,
+      "loss": 0.4254,
+      "step": 3810
+    },
+    {
+      "epoch": 0.283337665713543,
+      "grad_norm": 1.0664730072021484,
+      "learning_rate": 4.838269789329991e-05,
+      "loss": 0.3992,
+      "step": 3815
+    },
+    {
+      "epoch": 0.28370901258865905,
+      "grad_norm": 0.9408633708953857,
+      "learning_rate": 4.837737396319073e-05,
+      "loss": 0.4228,
+      "step": 3820
+    },
+    {
+      "epoch": 0.28408035946377513,
+      "grad_norm": 0.9652064442634583,
+      "learning_rate": 4.837204157855835e-05,
+      "loss": 0.4141,
+      "step": 3825
+    },
+    {
+      "epoch": 0.28445170633889116,
+      "grad_norm": 1.0405441522598267,
+      "learning_rate": 4.836670074133125e-05,
+      "loss": 0.4437,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2848230532140072,
+      "grad_norm": 1.0543575286865234,
+      "learning_rate": 4.8361351453440975e-05,
+      "loss": 0.4349,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2851944000891232,
+      "grad_norm": 1.3610877990722656,
+      "learning_rate": 4.83559937168221e-05,
+      "loss": 0.4469,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2855657469642393,
+      "grad_norm": 1.1551353931427002,
+      "learning_rate": 4.8350627533412295e-05,
+      "loss": 0.4183,
+      "step": 3845
+    },
+    {
+      "epoch": 0.28593709383935534,
+      "grad_norm": 0.8394594788551331,
+      "learning_rate": 4.834525290515224e-05,
+      "loss": 0.4219,
+      "step": 3850
+    },
+    {
+      "epoch": 0.28630844071447137,
+      "grad_norm": 0.9839069843292236,
+      "learning_rate": 4.833986983398571e-05,
+      "loss": 0.4195,
+      "step": 3855
+    },
+    {
+      "epoch": 0.28667978758958745,
+      "grad_norm": 0.9199720621109009,
+      "learning_rate": 4.8334478321859505e-05,
+      "loss": 0.4211,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2870511344647035,
+      "grad_norm": 0.8119776844978333,
+      "learning_rate": 4.832907837072349e-05,
+      "loss": 0.4187,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2874224813398195,
+      "grad_norm": 1.8330432176589966,
+      "learning_rate": 4.832366998253058e-05,
+      "loss": 0.4458,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2877938282149356,
+      "grad_norm": 1.2934834957122803,
+      "learning_rate": 4.831825315923674e-05,
+      "loss": 0.4275,
+      "step": 3875
+    },
+    {
+      "epoch": 0.28816517509005163,
+      "grad_norm": 0.8350692987442017,
+      "learning_rate": 4.8312827902800976e-05,
+      "loss": 0.4392,
+      "step": 3880
+    },
+    {
+      "epoch": 0.28853652196516766,
+      "grad_norm": 0.9420866966247559,
+      "learning_rate": 4.8307394215185374e-05,
+      "loss": 0.4119,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2889078688402837,
+      "grad_norm": 1.0160822868347168,
+      "learning_rate": 4.830195209835504e-05,
+      "loss": 0.4138,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2892792157153998,
+      "grad_norm": 1.0691471099853516,
+      "learning_rate": 4.829650155427813e-05,
+      "loss": 0.4323,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2896505625905158,
+      "grad_norm": 1.0200157165527344,
+      "learning_rate": 4.829104258492587e-05,
+      "loss": 0.4123,
+      "step": 3900
+    },
+    {
+      "epoch": 0.29002190946563183,
+      "grad_norm": 0.7049094438552856,
+      "learning_rate": 4.828557519227251e-05,
+      "loss": 0.42,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2903932563407479,
+      "grad_norm": 1.1828627586364746,
+      "learning_rate": 4.828009937829535e-05,
+      "loss": 0.4265,
+      "step": 3910
+    },
+    {
+      "epoch": 0.29076460321586395,
+      "grad_norm": 1.1924457550048828,
+      "learning_rate": 4.827461514497475e-05,
+      "loss": 0.434,
+      "step": 3915
+    },
+    {
+      "epoch": 0.29113595009098,
+      "grad_norm": 1.2100766897201538,
+      "learning_rate": 4.826912249429411e-05,
+      "loss": 0.4307,
+      "step": 3920
+    },
+    {
+      "epoch": 0.291507296966096,
+      "grad_norm": 0.9846333861351013,
+      "learning_rate": 4.8263621428239855e-05,
+      "loss": 0.4366,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2918786438412121,
+      "grad_norm": 1.0882083177566528,
+      "learning_rate": 4.8258111948801484e-05,
+      "loss": 0.4128,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2922499907163281,
+      "grad_norm": 0.8791163563728333,
+      "learning_rate": 4.825259405797151e-05,
+      "loss": 0.4417,
+      "step": 3935
+    },
+    {
+      "epoch": 0.29262133759144415,
+      "grad_norm": 1.0271704196929932,
+      "learning_rate": 4.82470677577455e-05,
+      "loss": 0.4214,
+      "step": 3940
+    },
+    {
+      "epoch": 0.29299268446656024,
+      "grad_norm": 1.3489196300506592,
+      "learning_rate": 4.8241533050122076e-05,
+      "loss": 0.4208,
+      "step": 3945
+    },
+    {
+      "epoch": 0.29336403134167627,
+      "grad_norm": 0.7574217915534973,
+      "learning_rate": 4.8235989937102874e-05,
+      "loss": 0.4211,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2937353782167923,
+      "grad_norm": 0.928737461566925,
+      "learning_rate": 4.823043842069259e-05,
+      "loss": 0.4397,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2941067250919083,
+      "grad_norm": 0.9212475419044495,
+      "learning_rate": 4.822487850289896e-05,
+      "loss": 0.4156,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2944780719670244,
+      "grad_norm": 0.77180016040802,
+      "learning_rate": 4.821931018573274e-05,
+      "loss": 0.4262,
+      "step": 3965
+    },
+    {
+      "epoch": 0.29484941884214044,
+      "grad_norm": 1.0247732400894165,
+      "learning_rate": 4.821373347120773e-05,
+      "loss": 0.4208,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2952207657172565,
+      "grad_norm": 0.9615873694419861,
+      "learning_rate": 4.820814836134079e-05,
+      "loss": 0.4056,
+      "step": 3975
+    },
+    {
+      "epoch": 0.29559211259237256,
+      "grad_norm": 0.9379513263702393,
+      "learning_rate": 4.820255485815178e-05,
+      "loss": 0.4463,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2959634594674886,
+      "grad_norm": 1.0274722576141357,
+      "learning_rate": 4.819695296366362e-05,
+      "loss": 0.4299,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2963348063426046,
+      "grad_norm": 1.0676618814468384,
+      "learning_rate": 4.819134267990226e-05,
+      "loss": 0.4361,
+      "step": 3990
+    },
+    {
+      "epoch": 0.29670615321772065,
+      "grad_norm": 0.9245099425315857,
+      "learning_rate": 4.818572400889669e-05,
+      "loss": 0.4248,
+      "step": 3995
+    },
+    {
+      "epoch": 0.29707750009283673,
+      "grad_norm": 1.0777188539505005,
+      "learning_rate": 4.8180096952678914e-05,
+      "loss": 0.4382,
+      "step": 4000
+    },
+    {
+      "epoch": 0.29744884696795276,
+      "grad_norm": 0.8690295815467834,
+      "learning_rate": 4.817446151328397e-05,
+      "loss": 0.4226,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2978201938430688,
+      "grad_norm": 0.9178885221481323,
+      "learning_rate": 4.816881769274997e-05,
+      "loss": 0.4394,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2981915407181849,
+      "grad_norm": 0.9835901856422424,
+      "learning_rate": 4.816316549311799e-05,
+      "loss": 0.4227,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2985628875933009,
+      "grad_norm": 0.893937349319458,
+      "learning_rate": 4.815750491643219e-05,
+      "loss": 0.4159,
+      "step": 4020
+    },
+    {
+      "epoch": 0.29893423446841694,
+      "grad_norm": 1.1184569597244263,
+      "learning_rate": 4.815183596473975e-05,
+      "loss": 0.4322,
+      "step": 4025
+    },
+    {
+      "epoch": 0.29930558134353297,
+      "grad_norm": 0.9903736710548401,
+      "learning_rate": 4.814615864009085e-05,
+      "loss": 0.4254,
+      "step": 4030
+    },
+    {
+      "epoch": 0.29967692821864905,
+      "grad_norm": 0.8701817393302917,
+      "learning_rate": 4.814047294453872e-05,
+      "loss": 0.4144,
+      "step": 4035
+    },
+    {
+      "epoch": 0.3000482750937651,
+      "grad_norm": 0.9406540393829346,
+      "learning_rate": 4.813477888013962e-05,
+      "loss": 0.4346,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3004196219688811,
+      "grad_norm": 1.1688896417617798,
+      "learning_rate": 4.8129076448952826e-05,
+      "loss": 0.4492,
+      "step": 4045
+    },
+    {
+      "epoch": 0.3007909688439972,
+      "grad_norm": 1.02422034740448,
+      "learning_rate": 4.812336565304066e-05,
+      "loss": 0.418,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3011623157191132,
+      "grad_norm": 0.9559808969497681,
+      "learning_rate": 4.811764649446845e-05,
+      "loss": 0.4208,
+      "step": 4055
+    },
+    {
+      "epoch": 0.30153366259422926,
+      "grad_norm": 0.6793580651283264,
+      "learning_rate": 4.811191897530454e-05,
+      "loss": 0.4155,
+      "step": 4060
+    },
+    {
+      "epoch": 0.30190500946934534,
+      "grad_norm": 1.0068315267562866,
+      "learning_rate": 4.8106183097620314e-05,
+      "loss": 0.4447,
+      "step": 4065
+    },
+    {
+      "epoch": 0.30227635634446137,
+      "grad_norm": 0.9851054549217224,
+      "learning_rate": 4.8100438863490196e-05,
+      "loss": 0.4442,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3026477032195774,
+      "grad_norm": 1.5529272556304932,
+      "learning_rate": 4.80946862749916e-05,
+      "loss": 0.4025,
+      "step": 4075
+    },
+    {
+      "epoch": 0.30301905009469343,
+      "grad_norm": 0.726374626159668,
+      "learning_rate": 4.808892533420496e-05,
+      "loss": 0.4047,
+      "step": 4080
+    },
+    {
+      "epoch": 0.3033903969698095,
+      "grad_norm": 0.9681001901626587,
+      "learning_rate": 4.808315604321375e-05,
+      "loss": 0.4129,
+      "step": 4085
+    },
+    {
+      "epoch": 0.30376174384492555,
+      "grad_norm": 0.8776106238365173,
+      "learning_rate": 4.807737840410447e-05,
+      "loss": 0.4335,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3041330907200416,
+      "grad_norm": 1.3474504947662354,
+      "learning_rate": 4.807159241896662e-05,
+      "loss": 0.432,
+      "step": 4095
+    },
+    {
+      "epoch": 0.30450443759515766,
+      "grad_norm": 0.9012182950973511,
+      "learning_rate": 4.8065798089892724e-05,
+      "loss": 0.4367,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3048757844702737,
+      "grad_norm": 1.0132664442062378,
+      "learning_rate": 4.8059995418978325e-05,
+      "loss": 0.4111,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3052471313453897,
+      "grad_norm": 0.7376078963279724,
+      "learning_rate": 4.8054184408321984e-05,
+      "loss": 0.4415,
+      "step": 4110
+    },
+    {
+      "epoch": 0.30561847822050575,
+      "grad_norm": 0.8478325009346008,
+      "learning_rate": 4.804836506002527e-05,
+      "loss": 0.4358,
+      "step": 4115
+    },
+    {
+      "epoch": 0.30598982509562184,
+      "grad_norm": 0.7889040112495422,
+      "learning_rate": 4.8042537376192784e-05,
+      "loss": 0.4049,
+      "step": 4120
+    },
+    {
+      "epoch": 0.30636117197073787,
+      "grad_norm": 2.196556806564331,
+      "learning_rate": 4.803670135893213e-05,
+      "loss": 0.3927,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3067325188458539,
+      "grad_norm": 0.6706188321113586,
+      "learning_rate": 4.8030857010353924e-05,
+      "loss": 0.4148,
+      "step": 4130
+    },
+    {
+      "epoch": 0.30710386572097,
+      "grad_norm": 0.9582722187042236,
+      "learning_rate": 4.80250043325718e-05,
+      "loss": 0.4262,
+      "step": 4135
+    },
+    {
+      "epoch": 0.307475212596086,
+      "grad_norm": 0.8332164287567139,
+      "learning_rate": 4.801914332770241e-05,
+      "loss": 0.4455,
+      "step": 4140
+    },
+    {
+      "epoch": 0.30784655947120204,
+      "grad_norm": 1.1163699626922607,
+      "learning_rate": 4.8013273997865394e-05,
+      "loss": 0.4344,
+      "step": 4145
+    },
+    {
+      "epoch": 0.30821790634631807,
+      "grad_norm": 0.8440271019935608,
+      "learning_rate": 4.800739634518343e-05,
+      "loss": 0.3971,
+      "step": 4150
+    },
+    {
+      "epoch": 0.30858925322143416,
+      "grad_norm": 0.8289729356765747,
+      "learning_rate": 4.800151037178219e-05,
+      "loss": 0.4024,
+      "step": 4155
+    },
+    {
+      "epoch": 0.3089606000965502,
+      "grad_norm": 0.844292163848877,
+      "learning_rate": 4.799561607979037e-05,
+      "loss": 0.4333,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3093319469716662,
+      "grad_norm": 0.743328332901001,
+      "learning_rate": 4.798971347133966e-05,
+      "loss": 0.4371,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3097032938467823,
+      "grad_norm": 1.2272253036499023,
+      "learning_rate": 4.798380254856476e-05,
+      "loss": 0.4197,
+      "step": 4170
+    },
+    {
+      "epoch": 0.31007464072189833,
+      "grad_norm": 0.8877312541007996,
+      "learning_rate": 4.797788331360339e-05,
+      "loss": 0.4121,
+      "step": 4175
+    },
+    {
+      "epoch": 0.31044598759701436,
+      "grad_norm": 1.0048223733901978,
+      "learning_rate": 4.7971955768596244e-05,
+      "loss": 0.4314,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3108173344721304,
+      "grad_norm": 1.184412956237793,
+      "learning_rate": 4.796601991568706e-05,
+      "loss": 0.4307,
+      "step": 4185
+    },
+    {
+      "epoch": 0.3111886813472465,
+      "grad_norm": 0.7950173020362854,
+      "learning_rate": 4.796007575702256e-05,
+      "loss": 0.4165,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3115600282223625,
+      "grad_norm": 0.9427579641342163,
+      "learning_rate": 4.7954123294752475e-05,
+      "loss": 0.4202,
+      "step": 4195
+    },
+    {
+      "epoch": 0.31193137509747854,
+      "grad_norm": 0.8798120021820068,
+      "learning_rate": 4.794816253102953e-05,
+      "loss": 0.4276,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3123027219725946,
+      "grad_norm": 0.9066138863563538,
+      "learning_rate": 4.7942193468009465e-05,
+      "loss": 0.4223,
+      "step": 4205
+    },
+    {
+      "epoch": 0.31267406884771065,
+      "grad_norm": 1.109160304069519,
+      "learning_rate": 4.793621610785102e-05,
+      "loss": 0.3984,
+      "step": 4210
+    },
+    {
+      "epoch": 0.3130454157228267,
+      "grad_norm": 0.9013233184814453,
+      "learning_rate": 4.793023045271592e-05,
+      "loss": 0.4128,
+      "step": 4215
+    },
+    {
+      "epoch": 0.3134167625979427,
+      "grad_norm": 0.8852935433387756,
+      "learning_rate": 4.792423650476892e-05,
+      "loss": 0.4275,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3137881094730588,
+      "grad_norm": 1.0032018423080444,
+      "learning_rate": 4.791823426617774e-05,
+      "loss": 0.4299,
+      "step": 4225
+    },
+    {
+      "epoch": 0.3141594563481748,
+      "grad_norm": 1.0188161134719849,
+      "learning_rate": 4.7912223739113125e-05,
+      "loss": 0.4151,
+      "step": 4230
+    },
+    {
+      "epoch": 0.31453080322329086,
+      "grad_norm": 0.8745348453521729,
+      "learning_rate": 4.7906204925748786e-05,
+      "loss": 0.425,
+      "step": 4235
+    },
+    {
+      "epoch": 0.31490215009840694,
+      "grad_norm": 0.8320831060409546,
+      "learning_rate": 4.790017782826148e-05,
+      "loss": 0.4199,
+      "step": 4240
+    },
+    {
+      "epoch": 0.31527349697352297,
+      "grad_norm": 0.8643554449081421,
+      "learning_rate": 4.789414244883092e-05,
+      "loss": 0.4582,
+      "step": 4245
+    },
+    {
+      "epoch": 0.315644843848639,
+      "grad_norm": 0.9545555114746094,
+      "learning_rate": 4.788809878963982e-05,
+      "loss": 0.4064,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3160161907237551,
+      "grad_norm": 1.0125293731689453,
+      "learning_rate": 4.7882046852873896e-05,
+      "loss": 0.4186,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3163875375988711,
+      "grad_norm": 1.3028231859207153,
+      "learning_rate": 4.787598664072186e-05,
+      "loss": 0.4336,
+      "step": 4260
+    },
+    {
+      "epoch": 0.31675888447398715,
+      "grad_norm": 0.8226160407066345,
+      "learning_rate": 4.786991815537542e-05,
+      "loss": 0.4105,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3171302313491032,
+      "grad_norm": 1.0225698947906494,
+      "learning_rate": 4.786384139902924e-05,
+      "loss": 0.4341,
+      "step": 4270
+    },
+    {
+      "epoch": 0.31750157822421926,
+      "grad_norm": 0.7511369585990906,
+      "learning_rate": 4.7857756373881035e-05,
+      "loss": 0.429,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3178729250993353,
+      "grad_norm": 0.9852497577667236,
+      "learning_rate": 4.7851663082131466e-05,
+      "loss": 0.4497,
+      "step": 4280
+    },
+    {
+      "epoch": 0.3182442719744513,
+      "grad_norm": 0.9103526473045349,
+      "learning_rate": 4.784556152598419e-05,
+      "loss": 0.4116,
+      "step": 4285
+    },
+    {
+      "epoch": 0.3186156188495674,
+      "grad_norm": 0.908010721206665,
+      "learning_rate": 4.7839451707645884e-05,
+      "loss": 0.4548,
+      "step": 4290
+    },
+    {
+      "epoch": 0.31898696572468344,
+      "grad_norm": 0.8548882603645325,
+      "learning_rate": 4.783333362932616e-05,
+      "loss": 0.4009,
+      "step": 4295
+    },
+    {
+      "epoch": 0.31935831259979947,
+      "grad_norm": 1.011780858039856,
+      "learning_rate": 4.7827207293237664e-05,
+      "loss": 0.4284,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3197296594749155,
+      "grad_norm": 0.7785174250602722,
+      "learning_rate": 4.7821072701596007e-05,
+      "loss": 0.4125,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3201010063500316,
+      "grad_norm": 0.8328904509544373,
+      "learning_rate": 4.781492985661979e-05,
+      "loss": 0.4225,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3204723532251476,
+      "grad_norm": 1.5389479398727417,
+      "learning_rate": 4.780877876053059e-05,
+      "loss": 0.4086,
+      "step": 4315
+    },
+    {
+      "epoch": 0.32084370010026364,
+      "grad_norm": 1.0274771451950073,
+      "learning_rate": 4.780261941555298e-05,
+      "loss": 0.4263,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3212150469753797,
+      "grad_norm": 0.9804433584213257,
+      "learning_rate": 4.7796451823914526e-05,
+      "loss": 0.4599,
+      "step": 4325
+    },
+    {
+      "epoch": 0.32158639385049576,
+      "grad_norm": 0.793917179107666,
+      "learning_rate": 4.7790275987845745e-05,
+      "loss": 0.4156,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3219577407256118,
+      "grad_norm": 0.8779480457305908,
+      "learning_rate": 4.778409190958016e-05,
+      "loss": 0.4314,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3223290876007278,
+      "grad_norm": 0.8802932500839233,
+      "learning_rate": 4.777789959135427e-05,
+      "loss": 0.4211,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3227004344758439,
+      "grad_norm": 0.8226904273033142,
+      "learning_rate": 4.777169903540755e-05,
+      "loss": 0.4152,
+      "step": 4345
+    },
+    {
+      "epoch": 0.32307178135095993,
+      "grad_norm": 0.8297529816627502,
+      "learning_rate": 4.776549024398245e-05,
+      "loss": 0.4169,
+      "step": 4350
+    },
+    {
+      "epoch": 0.32344312822607596,
+      "grad_norm": 1.039888858795166,
+      "learning_rate": 4.775927321932442e-05,
+      "loss": 0.4451,
+      "step": 4355
+    },
+    {
+      "epoch": 0.32381447510119205,
+      "grad_norm": 1.020605206489563,
+      "learning_rate": 4.7753047963681856e-05,
+      "loss": 0.4209,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3241858219763081,
+      "grad_norm": 1.1356257200241089,
+      "learning_rate": 4.774681447930616e-05,
+      "loss": 0.4108,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3245571688514241,
+      "grad_norm": 1.8272466659545898,
+      "learning_rate": 4.77405727684517e-05,
+      "loss": 0.4172,
+      "step": 4370
+    },
+    {
+      "epoch": 0.32492851572654013,
+      "grad_norm": 0.9962214231491089,
+      "learning_rate": 4.77343228333758e-05,
+      "loss": 0.4147,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3252998626016562,
+      "grad_norm": 0.873292863368988,
+      "learning_rate": 4.772806467633878e-05,
+      "loss": 0.4377,
+      "step": 4380
+    },
+    {
+      "epoch": 0.32567120947677225,
+      "grad_norm": 0.8075531721115112,
+      "learning_rate": 4.772179829960394e-05,
+      "loss": 0.4211,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3260425563518883,
+      "grad_norm": 1.0832219123840332,
+      "learning_rate": 4.771552370543753e-05,
+      "loss": 0.4312,
+      "step": 4390
+    },
+    {
+      "epoch": 0.32641390322700436,
+      "grad_norm": 0.9082911610603333,
+      "learning_rate": 4.7709240896108786e-05,
+      "loss": 0.438,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3267852501021204,
+      "grad_norm": 0.7049622535705566,
+      "learning_rate": 4.770294987388991e-05,
+      "loss": 0.4371,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3271565969772364,
+      "grad_norm": 1.050377607345581,
+      "learning_rate": 4.769665064105608e-05,
+      "loss": 0.4126,
+      "step": 4405
+    },
+    {
+      "epoch": 0.3275279438523525,
+      "grad_norm": 0.9951369762420654,
+      "learning_rate": 4.7690343199885434e-05,
+      "loss": 0.431,
+      "step": 4410
+    },
+    {
+      "epoch": 0.32789929072746854,
+      "grad_norm": 0.8130521774291992,
+      "learning_rate": 4.768402755265908e-05,
+      "loss": 0.4099,
+      "step": 4415
+    },
+    {
+      "epoch": 0.32827063760258457,
+      "grad_norm": 0.8695419430732727,
+      "learning_rate": 4.7677703701661115e-05,
+      "loss": 0.4242,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3286419844777006,
+      "grad_norm": 0.8330603241920471,
+      "learning_rate": 4.767137164917857e-05,
+      "loss": 0.4007,
+      "step": 4425
+    },
+    {
+      "epoch": 0.3290133313528167,
+      "grad_norm": 0.8372355699539185,
+      "learning_rate": 4.766503139750147e-05,
+      "loss": 0.4218,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3293846782279327,
+      "grad_norm": 0.835241436958313,
+      "learning_rate": 4.765868294892278e-05,
+      "loss": 0.4127,
+      "step": 4435
+    },
+    {
+      "epoch": 0.32975602510304874,
+      "grad_norm": 1.2005548477172852,
+      "learning_rate": 4.765232630573845e-05,
+      "loss": 0.4205,
+      "step": 4440
+    },
+    {
+      "epoch": 0.33012737197816483,
+      "grad_norm": 0.9409222602844238,
+      "learning_rate": 4.764596147024738e-05,
+      "loss": 0.4084,
+      "step": 4445
+    },
+    {
+      "epoch": 0.33049871885328086,
+      "grad_norm": 0.9494869709014893,
+      "learning_rate": 4.7639588444751445e-05,
+      "loss": 0.4241,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3308700657283969,
+      "grad_norm": 0.992162823677063,
+      "learning_rate": 4.763320723155548e-05,
+      "loss": 0.4082,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3312414126035129,
+      "grad_norm": 1.012285590171814,
+      "learning_rate": 4.7626817832967256e-05,
+      "loss": 0.4194,
+      "step": 4460
+    },
+    {
+      "epoch": 0.331612759478629,
+      "grad_norm": 0.860953688621521,
+      "learning_rate": 4.762042025129755e-05,
+      "loss": 0.3837,
+      "step": 4465
+    },
+    {
+      "epoch": 0.33198410635374503,
+      "grad_norm": 0.8259625434875488,
+      "learning_rate": 4.7614014488860065e-05,
+      "loss": 0.4119,
+      "step": 4470
+    },
+    {
+      "epoch": 0.33235545322886106,
+      "grad_norm": 0.7920925617218018,
+      "learning_rate": 4.7607600547971466e-05,
+      "loss": 0.4372,
+      "step": 4475
+    },
+    {
+      "epoch": 0.33272680010397715,
+      "grad_norm": 0.9090449213981628,
+      "learning_rate": 4.760117843095139e-05,
+      "loss": 0.4074,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3330981469790932,
+      "grad_norm": 0.8305264711380005,
+      "learning_rate": 4.7594748140122416e-05,
+      "loss": 0.406,
+      "step": 4485
+    },
+    {
+      "epoch": 0.3334694938542092,
+      "grad_norm": 0.8922953009605408,
+      "learning_rate": 4.758830967781008e-05,
+      "loss": 0.424,
+      "step": 4490
+    },
+    {
+      "epoch": 0.33384084072932524,
+      "grad_norm": 0.7428418397903442,
+      "learning_rate": 4.758186304634289e-05,
+      "loss": 0.4096,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3342121876044413,
+      "grad_norm": 0.6855220794677734,
+      "learning_rate": 4.757540824805229e-05,
+      "loss": 0.4262,
+      "step": 4500
+    },
+    {
+      "epoch": 0.33458353447955735,
+      "grad_norm": 0.8831934332847595,
+      "learning_rate": 4.756894528527268e-05,
+      "loss": 0.429,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3349548813546734,
+      "grad_norm": 0.7570902705192566,
+      "learning_rate": 4.756247416034143e-05,
+      "loss": 0.4068,
+      "step": 4510
+    },
+    {
+      "epoch": 0.33532622822978947,
+      "grad_norm": 0.8471874594688416,
+      "learning_rate": 4.755599487559884e-05,
+      "loss": 0.4319,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3356975751049055,
+      "grad_norm": 1.5127962827682495,
+      "learning_rate": 4.754950743338817e-05,
+      "loss": 0.3788,
+      "step": 4520
+    },
+    {
+      "epoch": 0.33606892198002153,
+      "grad_norm": 0.8634641170501709,
+      "learning_rate": 4.754301183605563e-05,
+      "loss": 0.4188,
+      "step": 4525
+    },
+    {
+      "epoch": 0.33644026885513756,
+      "grad_norm": 0.9390023350715637,
+      "learning_rate": 4.753650808595038e-05,
+      "loss": 0.3925,
+      "step": 4530
+    },
+    {
+      "epoch": 0.33681161573025364,
+      "grad_norm": 1.002371072769165,
+      "learning_rate": 4.752999618542453e-05,
+      "loss": 0.4376,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3371829626053697,
+      "grad_norm": 1.0945231914520264,
+      "learning_rate": 4.752347613683313e-05,
+      "loss": 0.4435,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3375543094804857,
+      "grad_norm": 1.3020802736282349,
+      "learning_rate": 4.75169479425342e-05,
+      "loss": 0.4297,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3379256563556018,
+      "grad_norm": 0.9557328820228577,
+      "learning_rate": 4.751041160488866e-05,
+      "loss": 0.4348,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3382970032307178,
+      "grad_norm": 0.7302908301353455,
+      "learning_rate": 4.7503867126260426e-05,
+      "loss": 0.4073,
+      "step": 4555
+    },
+    {
+      "epoch": 0.33866835010583385,
+      "grad_norm": 0.9016184210777283,
+      "learning_rate": 4.749731450901632e-05,
+      "loss": 0.4192,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3390396969809499,
+      "grad_norm": 1.1325968503952026,
+      "learning_rate": 4.749075375552614e-05,
+      "loss": 0.4138,
+      "step": 4565
+    },
+    {
+      "epoch": 0.33941104385606596,
+      "grad_norm": 0.7631894946098328,
+      "learning_rate": 4.7484184868162595e-05,
+      "loss": 0.3958,
+      "step": 4570
+    },
+    {
+      "epoch": 0.339782390731182,
+      "grad_norm": 0.9254847764968872,
+      "learning_rate": 4.7477607849301356e-05,
+      "loss": 0.4121,
+      "step": 4575
+    },
+    {
+      "epoch": 0.340153737606298,
+      "grad_norm": 0.8892576694488525,
+      "learning_rate": 4.747102270132103e-05,
+      "loss": 0.4099,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3405250844814141,
+      "grad_norm": 1.2878713607788086,
+      "learning_rate": 4.7464429426603155e-05,
+      "loss": 0.4052,
+      "step": 4585
+    },
+    {
+      "epoch": 0.34089643135653014,
+      "grad_norm": 1.045745611190796,
+      "learning_rate": 4.7457828027532225e-05,
+      "loss": 0.4015,
+      "step": 4590
+    },
+    {
+      "epoch": 0.34126777823164617,
+      "grad_norm": 0.7936729192733765,
+      "learning_rate": 4.7451218506495656e-05,
+      "loss": 0.4264,
+      "step": 4595
+    },
+    {
+      "epoch": 0.34163912510676225,
+      "grad_norm": 1.1288927793502808,
+      "learning_rate": 4.744460086588382e-05,
+      "loss": 0.457,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3420104719818783,
+      "grad_norm": 0.7576532363891602,
+      "learning_rate": 4.743797510809e-05,
+      "loss": 0.415,
+      "step": 4605
+    },
+    {
+      "epoch": 0.3423818188569943,
+      "grad_norm": 0.8422873616218567,
+      "learning_rate": 4.743134123551043e-05,
+      "loss": 0.4191,
+      "step": 4610
+    },
+    {
+      "epoch": 0.34275316573211034,
+      "grad_norm": 1.4032751321792603,
+      "learning_rate": 4.742469925054429e-05,
+      "loss": 0.4125,
+      "step": 4615
+    },
+    {
+      "epoch": 0.34312451260722643,
+      "grad_norm": 0.7778531312942505,
+      "learning_rate": 4.741804915559367e-05,
+      "loss": 0.4079,
+      "step": 4620
+    },
+    {
+      "epoch": 0.34349585948234246,
+      "grad_norm": 1.4162076711654663,
+      "learning_rate": 4.74113909530636e-05,
+      "loss": 0.4409,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3438672063574585,
+      "grad_norm": 1.6434004306793213,
+      "learning_rate": 4.7404724645362056e-05,
+      "loss": 0.4113,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3442385532325746,
+      "grad_norm": 0.9791908264160156,
+      "learning_rate": 4.7398050234899935e-05,
+      "loss": 0.4103,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3446099001076906,
+      "grad_norm": 0.8746637105941772,
+      "learning_rate": 4.739136772409106e-05,
+      "loss": 0.4281,
+      "step": 4640
+    },
+    {
+      "epoch": 0.34498124698280663,
+      "grad_norm": 1.060185194015503,
+      "learning_rate": 4.73846771153522e-05,
+      "loss": 0.4098,
+      "step": 4645
+    },
+    {
+      "epoch": 0.34535259385792266,
+      "grad_norm": 0.9599495530128479,
+      "learning_rate": 4.737797841110302e-05,
+      "loss": 0.4181,
+      "step": 4650
+    },
+    {
+      "epoch": 0.34572394073303875,
+      "grad_norm": 1.1338365077972412,
+      "learning_rate": 4.737127161376616e-05,
+      "loss": 0.4168,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3460952876081548,
+      "grad_norm": 1.0875575542449951,
+      "learning_rate": 4.7364556725767127e-05,
+      "loss": 0.4383,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3464666344832708,
+      "grad_norm": 0.8746302127838135,
+      "learning_rate": 4.735783374953442e-05,
+      "loss": 0.4245,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3468379813583869,
+      "grad_norm": 1.142159104347229,
+      "learning_rate": 4.7351102687499415e-05,
+      "loss": 0.4159,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3472093282335029,
+      "grad_norm": 1.0561567544937134,
+      "learning_rate": 4.734436354209644e-05,
+      "loss": 0.4076,
+      "step": 4675
+    },
+    {
+      "epoch": 0.34758067510861895,
+      "grad_norm": 0.7917932868003845,
+      "learning_rate": 4.733761631576271e-05,
+      "loss": 0.4199,
+      "step": 4680
+    },
+    {
+      "epoch": 0.347952021983735,
+      "grad_norm": 0.8924756646156311,
+      "learning_rate": 4.733086101093842e-05,
+      "loss": 0.3952,
+      "step": 4685
+    },
+    {
+      "epoch": 0.34832336885885107,
+      "grad_norm": 0.9237497448921204,
+      "learning_rate": 4.7324097630066625e-05,
+      "loss": 0.4241,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3486947157339671,
+      "grad_norm": 0.8856948614120483,
+      "learning_rate": 4.7317326175593345e-05,
+      "loss": 0.4269,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3490660626090831,
+      "grad_norm": 0.9089455604553223,
+      "learning_rate": 4.73105466499675e-05,
+      "loss": 0.3952,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3494374094841992,
+      "grad_norm": 0.948407769203186,
+      "learning_rate": 4.730375905564093e-05,
+      "loss": 0.421,
+      "step": 4705
+    },
+    {
+      "epoch": 0.34980875635931524,
+      "grad_norm": 0.658133864402771,
+      "learning_rate": 4.729696339506841e-05,
+      "loss": 0.4207,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3501801032344313,
+      "grad_norm": 1.0339378118515015,
+      "learning_rate": 4.72901596707076e-05,
+      "loss": 0.4139,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3505514501095473,
+      "grad_norm": 0.9219155311584473,
+      "learning_rate": 4.7283347885019106e-05,
+      "loss": 0.4305,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3509227969846634,
+      "grad_norm": 1.0298707485198975,
+      "learning_rate": 4.727652804046644e-05,
+      "loss": 0.4122,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3512941438597794,
+      "grad_norm": 0.7566863894462585,
+      "learning_rate": 4.7269700139516016e-05,
+      "loss": 0.433,
+      "step": 4730
+    },
+    {
+      "epoch": 0.35166549073489545,
+      "grad_norm": 0.851755678653717,
+      "learning_rate": 4.726286418463718e-05,
+      "loss": 0.4055,
+      "step": 4735
+    },
+    {
+      "epoch": 0.35203683761001153,
+      "grad_norm": 1.0216217041015625,
+      "learning_rate": 4.725602017830219e-05,
+      "loss": 0.431,
+      "step": 4740
+    },
+    {
+      "epoch": 0.35240818448512756,
+      "grad_norm": 0.917681097984314,
+      "learning_rate": 4.7249168122986196e-05,
+      "loss": 0.421,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3527795313602436,
+      "grad_norm": 1.4520493745803833,
+      "learning_rate": 4.7242308021167274e-05,
+      "loss": 0.4081,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3531508782353596,
+      "grad_norm": 0.8711174130439758,
+      "learning_rate": 4.7235439875326414e-05,
+      "loss": 0.4266,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3535222251104757,
+      "grad_norm": 0.8265815377235413,
+      "learning_rate": 4.7228563687947504e-05,
+      "loss": 0.4126,
+      "step": 4760
+    },
+    {
+      "epoch": 0.35389357198559174,
+      "grad_norm": 0.9316338896751404,
+      "learning_rate": 4.722167946151735e-05,
+      "loss": 0.406,
+      "step": 4765
+    },
+    {
+      "epoch": 0.35426491886070777,
+      "grad_norm": 0.9665477871894836,
+      "learning_rate": 4.7214787198525656e-05,
+      "loss": 0.4188,
+      "step": 4770
+    },
+    {
+      "epoch": 0.35463626573582385,
+      "grad_norm": 0.9457529187202454,
+      "learning_rate": 4.720788690146504e-05,
+      "loss": 0.4481,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3550076126109399,
+      "grad_norm": 0.9023556113243103,
+      "learning_rate": 4.7200978572831026e-05,
+      "loss": 0.4059,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3553789594860559,
+      "grad_norm": 0.9159427881240845,
+      "learning_rate": 4.719406221512203e-05,
+      "loss": 0.4029,
+      "step": 4785
+    },
+    {
+      "epoch": 0.355750306361172,
+      "grad_norm": 0.7920024394989014,
+      "learning_rate": 4.718713783083939e-05,
+      "loss": 0.4157,
+      "step": 4790
+    },
+    {
+      "epoch": 0.356121653236288,
+      "grad_norm": 0.8898520469665527,
+      "learning_rate": 4.7180205422487335e-05,
+      "loss": 0.4111,
+      "step": 4795
+    },
+    {
+      "epoch": 0.35649300011140406,
+      "grad_norm": 1.610779881477356,
+      "learning_rate": 4.7173264992572994e-05,
+      "loss": 0.4243,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3568643469865201,
+      "grad_norm": 1.1176823377609253,
+      "learning_rate": 4.7166316543606406e-05,
+      "loss": 0.4032,
+      "step": 4805
+    },
+    {
+      "epoch": 0.35723569386163617,
+      "grad_norm": 0.8275334239006042,
+      "learning_rate": 4.715936007810051e-05,
+      "loss": 0.3972,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3576070407367522,
+      "grad_norm": 0.8289036750793457,
+      "learning_rate": 4.715239559857114e-05,
+      "loss": 0.4108,
+      "step": 4815
+    },
+    {
+      "epoch": 0.35797838761186823,
+      "grad_norm": 0.825310230255127,
+      "learning_rate": 4.714542310753702e-05,
+      "loss": 0.4153,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3583497344869843,
+      "grad_norm": 0.7950535416603088,
+      "learning_rate": 4.713844260751978e-05,
+      "loss": 0.4071,
+      "step": 4825
+    },
+    {
+      "epoch": 0.35872108136210035,
+      "grad_norm": 1.4191958904266357,
+      "learning_rate": 4.713145410104396e-05,
+      "loss": 0.4023,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3590924282372164,
+      "grad_norm": 0.9486410021781921,
+      "learning_rate": 4.712445759063697e-05,
+      "loss": 0.4136,
+      "step": 4835
+    },
+    {
+      "epoch": 0.3594637751123324,
+      "grad_norm": 1.385170817375183,
+      "learning_rate": 4.711745307882914e-05,
+      "loss": 0.4231,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3598351219874485,
+      "grad_norm": 0.7949105501174927,
+      "learning_rate": 4.711044056815366e-05,
+      "loss": 0.4139,
+      "step": 4845
+    },
+    {
+      "epoch": 0.3602064688625645,
+      "grad_norm": 0.8897971510887146,
+      "learning_rate": 4.710342006114665e-05,
+      "loss": 0.4398,
+      "step": 4850
+    },
+    {
+      "epoch": 0.36057781573768055,
+      "grad_norm": 1.0138311386108398,
+      "learning_rate": 4.70963915603471e-05,
+      "loss": 0.413,
+      "step": 4855
+    },
+    {
+      "epoch": 0.36094916261279664,
+      "grad_norm": 1.127977728843689,
+      "learning_rate": 4.708935506829689e-05,
+      "loss": 0.42,
+      "step": 4860
+    },
+    {
+      "epoch": 0.36132050948791267,
+      "grad_norm": 0.8652618527412415,
+      "learning_rate": 4.708231058754081e-05,
+      "loss": 0.4238,
+      "step": 4865
+    },
+    {
+      "epoch": 0.3616918563630287,
+      "grad_norm": 0.8667855858802795,
+      "learning_rate": 4.707525812062651e-05,
+      "loss": 0.3933,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3620632032381447,
+      "grad_norm": 0.8137786388397217,
+      "learning_rate": 4.706819767010455e-05,
+      "loss": 0.4253,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3624345501132608,
+      "grad_norm": 0.837216317653656,
+      "learning_rate": 4.7061129238528376e-05,
+      "loss": 0.4101,
+      "step": 4880
+    },
+    {
+      "epoch": 0.36280589698837684,
+      "grad_norm": 0.8621715903282166,
+      "learning_rate": 4.705405282845431e-05,
+      "loss": 0.3916,
+      "step": 4885
+    },
+    {
+      "epoch": 0.36317724386349287,
+      "grad_norm": 0.9200319647789001,
+      "learning_rate": 4.7046968442441566e-05,
+      "loss": 0.4334,
+      "step": 4890
+    },
+    {
+      "epoch": 0.36354859073860896,
+      "grad_norm": 0.9625046849250793,
+      "learning_rate": 4.7039876083052235e-05,
+      "loss": 0.4364,
+      "step": 4895
+    },
+    {
+      "epoch": 0.363919937613725,
+      "grad_norm": 0.9593774676322937,
+      "learning_rate": 4.703277575285131e-05,
+      "loss": 0.405,
+      "step": 4900
+    },
+    {
+      "epoch": 0.364291284488841,
+      "grad_norm": 0.9800950884819031,
+      "learning_rate": 4.702566745440665e-05,
+      "loss": 0.4131,
+      "step": 4905
+    },
+    {
+      "epoch": 0.36466263136395705,
+      "grad_norm": 0.808862566947937,
+      "learning_rate": 4.7018551190289e-05,
+      "loss": 0.402,
+      "step": 4910
+    },
+    {
+      "epoch": 0.36503397823907313,
+      "grad_norm": 0.8395322561264038,
+      "learning_rate": 4.701142696307198e-05,
+      "loss": 0.3979,
+      "step": 4915
+    },
+    {
+      "epoch": 0.36540532511418916,
+      "grad_norm": 0.9485448598861694,
+      "learning_rate": 4.70042947753321e-05,
+      "loss": 0.3904,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3657766719893052,
+      "grad_norm": 1.0159146785736084,
+      "learning_rate": 4.6997154629648743e-05,
+      "loss": 0.4085,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3661480188644213,
+      "grad_norm": 1.6333436965942383,
+      "learning_rate": 4.699000652860418e-05,
+      "loss": 0.404,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3665193657395373,
+      "grad_norm": 0.9616270065307617,
+      "learning_rate": 4.6982850474783546e-05,
+      "loss": 0.4199,
+      "step": 4935
+    },
+    {
+      "epoch": 0.36689071261465334,
+      "grad_norm": 0.8749721646308899,
+      "learning_rate": 4.6975686470774854e-05,
+      "loss": 0.4358,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3672620594897694,
+      "grad_norm": 0.7861535549163818,
+      "learning_rate": 4.6968514519169e-05,
+      "loss": 0.3966,
+      "step": 4945
+    },
+    {
+      "epoch": 0.36763340636488545,
+      "grad_norm": 0.9194499254226685,
+      "learning_rate": 4.696133462255975e-05,
+      "loss": 0.4247,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3680047532400015,
+      "grad_norm": 0.8600870370864868,
+      "learning_rate": 4.695414678354374e-05,
+      "loss": 0.3944,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3683761001151175,
+      "grad_norm": 0.7348113059997559,
+      "learning_rate": 4.6946951004720476e-05,
+      "loss": 0.416,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3687474469902336,
+      "grad_norm": 0.9184551239013672,
+      "learning_rate": 4.693974728869235e-05,
+      "loss": 0.3881,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3691187938653496,
+      "grad_norm": 0.8617504239082336,
+      "learning_rate": 4.693253563806462e-05,
+      "loss": 0.4277,
+      "step": 4970
+    },
+    {
+      "epoch": 0.36949014074046566,
+      "grad_norm": 0.8256643414497375,
+      "learning_rate": 4.69253160554454e-05,
+      "loss": 0.4238,
+      "step": 4975
+    },
+    {
+      "epoch": 0.36986148761558174,
+      "grad_norm": 0.9334145188331604,
+      "learning_rate": 4.691808854344568e-05,
+      "loss": 0.4065,
+      "step": 4980
+    },
+    {
+      "epoch": 0.37023283449069777,
+      "grad_norm": 1.2154117822647095,
+      "learning_rate": 4.691085310467932e-05,
+      "loss": 0.4309,
+      "step": 4985
+    },
+    {
+      "epoch": 0.3706041813658138,
+      "grad_norm": 1.2641429901123047,
+      "learning_rate": 4.690360974176306e-05,
+      "loss": 0.4277,
+      "step": 4990
+    },
+    {
+      "epoch": 0.37097552824092983,
+      "grad_norm": 1.158447265625,
+      "learning_rate": 4.6896358457316475e-05,
+      "loss": 0.4309,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3713468751160459,
+      "grad_norm": 0.7376827001571655,
+      "learning_rate": 4.6889099253962035e-05,
+      "loss": 0.4129,
+      "step": 5000
+    },
+    {
+      "epoch": 0.37171822199116195,
+      "grad_norm": 0.9000616669654846,
+      "learning_rate": 4.688183213432504e-05,
+      "loss": 0.4151,
+      "step": 5005
+    },
+    {
+      "epoch": 0.372089568866278,
+      "grad_norm": 0.8609789609909058,
+      "learning_rate": 4.687455710103369e-05,
+      "loss": 0.4264,
+      "step": 5010
+    },
+    {
+      "epoch": 0.37246091574139406,
+      "grad_norm": 0.9813238978385925,
+      "learning_rate": 4.686727415671903e-05,
+      "loss": 0.4125,
+      "step": 5015
+    },
+    {
+      "epoch": 0.3728322626165101,
+      "grad_norm": 0.7004589438438416,
+      "learning_rate": 4.6859983304014974e-05,
+      "loss": 0.4049,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3732036094916261,
+      "grad_norm": 0.7324821949005127,
+      "learning_rate": 4.685268454555827e-05,
+      "loss": 0.4229,
+      "step": 5025
+    },
+    {
+      "epoch": 0.37357495636674215,
+      "grad_norm": 0.8856626152992249,
+      "learning_rate": 4.684537788398855e-05,
+      "loss": 0.4101,
+      "step": 5030
+    },
+    {
+      "epoch": 0.37394630324185824,
+      "grad_norm": 1.6421722173690796,
+      "learning_rate": 4.683806332194831e-05,
+      "loss": 0.3939,
+      "step": 5035
+    },
+    {
+      "epoch": 0.37431765011697427,
+      "grad_norm": 0.740895688533783,
+      "learning_rate": 4.683074086208288e-05,
+      "loss": 0.4111,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3746889969920903,
+      "grad_norm": 1.21027410030365,
+      "learning_rate": 4.682341050704046e-05,
+      "loss": 0.4195,
+      "step": 5045
+    },
+    {
+      "epoch": 0.3750603438672064,
+      "grad_norm": 0.8727134466171265,
+      "learning_rate": 4.68160722594721e-05,
+      "loss": 0.4091,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3754316907423224,
+      "grad_norm": 0.8802042603492737,
+      "learning_rate": 4.680872612203171e-05,
+      "loss": 0.3945,
+      "step": 5055
+    },
+    {
+      "epoch": 0.37580303761743844,
+      "grad_norm": 0.8701076507568359,
+      "learning_rate": 4.680137209737606e-05,
+      "loss": 0.3914,
+      "step": 5060
+    },
+    {
+      "epoch": 0.37617438449255447,
+      "grad_norm": 0.9685407280921936,
+      "learning_rate": 4.679401018816475e-05,
+      "loss": 0.4192,
+      "step": 5065
+    },
+    {
+      "epoch": 0.37654573136767056,
+      "grad_norm": 0.839486300945282,
+      "learning_rate": 4.6786640397060246e-05,
+      "loss": 0.4029,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3769170782427866,
+      "grad_norm": 0.8863376975059509,
+      "learning_rate": 4.6779262726727866e-05,
+      "loss": 0.3989,
+      "step": 5075
+    },
+    {
+      "epoch": 0.3772884251179026,
+      "grad_norm": 0.9725920557975769,
+      "learning_rate": 4.677187717983578e-05,
+      "loss": 0.4042,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3776597719930187,
+      "grad_norm": 0.9129449129104614,
+      "learning_rate": 4.6764483759055e-05,
+      "loss": 0.4086,
+      "step": 5085
+    },
+    {
+      "epoch": 0.37803111886813473,
+      "grad_norm": 1.042917251586914,
+      "learning_rate": 4.675708246705938e-05,
+      "loss": 0.3807,
+      "step": 5090
+    },
+    {
+      "epoch": 0.37840246574325076,
+      "grad_norm": 0.8052343130111694,
+      "learning_rate": 4.674967330652562e-05,
+      "loss": 0.4202,
+      "step": 5095
+    },
+    {
+      "epoch": 0.3787738126183668,
+      "grad_norm": 0.8487399220466614,
+      "learning_rate": 4.67422562801333e-05,
+      "loss": 0.4092,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3791451594934829,
+      "grad_norm": 1.2970117330551147,
+      "learning_rate": 4.6734831390564795e-05,
+      "loss": 0.4255,
+      "step": 5105
+    },
+    {
+      "epoch": 0.3795165063685989,
+      "grad_norm": 0.9775636196136475,
+      "learning_rate": 4.672739864050535e-05,
+      "loss": 0.4067,
+      "step": 5110
+    },
+    {
+      "epoch": 0.37988785324371493,
+      "grad_norm": 1.285401463508606,
+      "learning_rate": 4.671995803264306e-05,
+      "loss": 0.4423,
+      "step": 5115
+    },
+    {
+      "epoch": 0.380259200118831,
+      "grad_norm": 0.7425358891487122,
+      "learning_rate": 4.671250956966885e-05,
+      "loss": 0.4119,
+      "step": 5120
+    },
+    {
+      "epoch": 0.38063054699394705,
+      "grad_norm": 0.8824949860572815,
+      "learning_rate": 4.670505325427648e-05,
+      "loss": 0.4094,
+      "step": 5125
+    },
+    {
+      "epoch": 0.3810018938690631,
+      "grad_norm": 0.8295382261276245,
+      "learning_rate": 4.669758908916255e-05,
+      "loss": 0.3983,
+      "step": 5130
+    },
+    {
+      "epoch": 0.38137324074417916,
+      "grad_norm": 0.7965675592422485,
+      "learning_rate": 4.6690117077026515e-05,
+      "loss": 0.4033,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3817445876192952,
+      "grad_norm": 0.7937360405921936,
+      "learning_rate": 4.6682637220570666e-05,
+      "loss": 0.3942,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3821159344944112,
+      "grad_norm": 0.892683744430542,
+      "learning_rate": 4.6675149522500124e-05,
+      "loss": 0.4026,
+      "step": 5145
+    },
+    {
+      "epoch": 0.38248728136952725,
+      "grad_norm": 0.8494230508804321,
+      "learning_rate": 4.666765398552283e-05,
+      "loss": 0.4035,
+      "step": 5150
+    },
+    {
+      "epoch": 0.38285862824464334,
+      "grad_norm": 1.0645548105239868,
+      "learning_rate": 4.666015061234958e-05,
+      "loss": 0.4324,
+      "step": 5155
+    },
+    {
+      "epoch": 0.38322997511975937,
+      "grad_norm": 1.1099061965942383,
+      "learning_rate": 4.665263940569401e-05,
+      "loss": 0.4147,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3836013219948754,
+      "grad_norm": 0.864147961139679,
+      "learning_rate": 4.664512036827258e-05,
+      "loss": 0.4311,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3839726688699915,
+      "grad_norm": 0.6989151835441589,
+      "learning_rate": 4.663759350280456e-05,
+      "loss": 0.4137,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3843440157451075,
+      "grad_norm": 0.997937798500061,
+      "learning_rate": 4.663005881201209e-05,
+      "loss": 0.3688,
+      "step": 5175
+    },
+    {
+      "epoch": 0.38471536262022354,
+      "grad_norm": 0.9270450472831726,
+      "learning_rate": 4.6622516298620124e-05,
+      "loss": 0.4192,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3850867094953396,
+      "grad_norm": 0.8299030661582947,
+      "learning_rate": 4.661496596535643e-05,
+      "loss": 0.4072,
+      "step": 5185
+    },
+    {
+      "epoch": 0.38545805637045566,
+      "grad_norm": 1.146391749382019,
+      "learning_rate": 4.660740781495162e-05,
+      "loss": 0.3983,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3858294032455717,
+      "grad_norm": 0.8312609791755676,
+      "learning_rate": 4.659984185013913e-05,
+      "loss": 0.4204,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3862007501206877,
+      "grad_norm": 0.8791617751121521,
+      "learning_rate": 4.659226807365523e-05,
+      "loss": 0.4005,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3865720969958038,
+      "grad_norm": 0.8427607417106628,
+      "learning_rate": 4.6584686488239e-05,
+      "loss": 0.3986,
+      "step": 5205
+    },
+    {
+      "epoch": 0.38694344387091983,
+      "grad_norm": 0.7755858302116394,
+      "learning_rate": 4.657709709663237e-05,
+      "loss": 0.41,
+      "step": 5210
+    },
+    {
+      "epoch": 0.38731479074603586,
+      "grad_norm": 0.9663941264152527,
+      "learning_rate": 4.656949990158004e-05,
+      "loss": 0.3973,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3876861376211519,
+      "grad_norm": 0.8808470964431763,
+      "learning_rate": 4.656189490582959e-05,
+      "loss": 0.4142,
+      "step": 5220
+    },
+    {
+      "epoch": 0.388057484496268,
+      "grad_norm": 0.835097074508667,
+      "learning_rate": 4.65542821121314e-05,
+      "loss": 0.3907,
+      "step": 5225
+    },
+    {
+      "epoch": 0.388428831371384,
+      "grad_norm": 1.165091872215271,
+      "learning_rate": 4.654666152323866e-05,
+      "loss": 0.4102,
+      "step": 5230
+    },
+    {
+      "epoch": 0.38880017824650004,
+      "grad_norm": 0.9790161848068237,
+      "learning_rate": 4.653903314190738e-05,
+      "loss": 0.4125,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3891715251216161,
+      "grad_norm": 0.8133743405342102,
+      "learning_rate": 4.653139697089641e-05,
+      "loss": 0.417,
+      "step": 5240
+    },
+    {
+      "epoch": 0.38954287199673215,
+      "grad_norm": 0.9274176955223083,
+      "learning_rate": 4.652375301296739e-05,
+      "loss": 0.4218,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3899142188718482,
+      "grad_norm": 0.8492559194564819,
+      "learning_rate": 4.6516101270884805e-05,
+      "loss": 0.3966,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3902855657469642,
+      "grad_norm": 0.9852241277694702,
+      "learning_rate": 4.650844174741592e-05,
+      "loss": 0.4035,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3906569126220803,
+      "grad_norm": 0.9053683876991272,
+      "learning_rate": 4.6500774445330854e-05,
+      "loss": 0.3954,
+      "step": 5260
+    },
+    {
+      "epoch": 0.39102825949719633,
+      "grad_norm": 1.2327396869659424,
+      "learning_rate": 4.64930993674025e-05,
+      "loss": 0.3793,
+      "step": 5265
+    },
+    {
+      "epoch": 0.39139960637231236,
+      "grad_norm": 0.9179367423057556,
+      "learning_rate": 4.648541651640659e-05,
+      "loss": 0.4146,
+      "step": 5270
+    },
+    {
+      "epoch": 0.39177095324742844,
+      "grad_norm": 0.8724064826965332,
+      "learning_rate": 4.647772589512165e-05,
+      "loss": 0.4214,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3921423001225445,
+      "grad_norm": 0.8555846810340881,
+      "learning_rate": 4.6470027506329036e-05,
+      "loss": 0.4041,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3925136469976605,
+      "grad_norm": 0.804630696773529,
+      "learning_rate": 4.64623213528129e-05,
+      "loss": 0.4092,
+      "step": 5285
+    },
+    {
+      "epoch": 0.39288499387277653,
+      "grad_norm": 1.16206693649292,
+      "learning_rate": 4.64546074373602e-05,
+      "loss": 0.4095,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3932563407478926,
+      "grad_norm": 1.0643155574798584,
+      "learning_rate": 4.644688576276071e-05,
+      "loss": 0.4197,
+      "step": 5295
+    },
+    {
+      "epoch": 0.39362768762300865,
+      "grad_norm": 0.9959713220596313,
+      "learning_rate": 4.643915633180699e-05,
+      "loss": 0.4022,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3939990344981247,
+      "grad_norm": 0.8475741744041443,
+      "learning_rate": 4.6431419147294445e-05,
+      "loss": 0.4099,
+      "step": 5305
+    },
+    {
+      "epoch": 0.39437038137324076,
+      "grad_norm": 0.7796865701675415,
+      "learning_rate": 4.642367421202124e-05,
+      "loss": 0.4044,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3947417282483568,
+      "grad_norm": 0.9301022291183472,
+      "learning_rate": 4.641592152878837e-05,
+      "loss": 0.4104,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3951130751234728,
+      "grad_norm": 1.0044193267822266,
+      "learning_rate": 4.640816110039962e-05,
+      "loss": 0.4028,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3954844219985889,
+      "grad_norm": 1.0422751903533936,
+      "learning_rate": 4.6400392929661596e-05,
+      "loss": 0.3784,
+      "step": 5325
+    },
+    {
+      "epoch": 0.39585576887370494,
+      "grad_norm": 0.7252130508422852,
+      "learning_rate": 4.6392617019383665e-05,
+      "loss": 0.4118,
+      "step": 5330
+    },
+    {
+      "epoch": 0.39622711574882097,
+      "grad_norm": 1.1372671127319336,
+      "learning_rate": 4.638483337237803e-05,
+      "loss": 0.4297,
+      "step": 5335
+    },
+    {
+      "epoch": 0.396598462623937,
+      "grad_norm": 0.917350709438324,
+      "learning_rate": 4.6377041991459694e-05,
+      "loss": 0.4082,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3969698094990531,
+      "grad_norm": 0.8461329340934753,
+      "learning_rate": 4.636924287944641e-05,
+      "loss": 0.4239,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3973411563741691,
+      "grad_norm": 0.823368489742279,
+      "learning_rate": 4.636143603915878e-05,
+      "loss": 0.4517,
+      "step": 5350
+    },
+    {
+      "epoch": 0.39771250324928514,
+      "grad_norm": 0.99635910987854,
+      "learning_rate": 4.635362147342018e-05,
+      "loss": 0.4001,
+      "step": 5355
+    },
+    {
+      "epoch": 0.39808385012440123,
+      "grad_norm": 0.8141667246818542,
+      "learning_rate": 4.634579918505677e-05,
+      "loss": 0.4269,
+      "step": 5360
+    },
+    {
+      "epoch": 0.39845519699951726,
+      "grad_norm": 0.9809877276420593,
+      "learning_rate": 4.633796917689751e-05,
+      "loss": 0.4372,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3988265438746333,
+      "grad_norm": 0.8991804122924805,
+      "learning_rate": 4.6330131451774174e-05,
+      "loss": 0.4092,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3991978907497493,
+      "grad_norm": 0.8061063885688782,
+      "learning_rate": 4.6322286012521285e-05,
+      "loss": 0.4262,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3995692376248654,
+      "grad_norm": 0.8821228742599487,
+      "learning_rate": 4.63144328619762e-05,
+      "loss": 0.4151,
+      "step": 5380
+    },
+    {
+      "epoch": 0.39994058449998143,
+      "grad_norm": 1.0889192819595337,
+      "learning_rate": 4.630657200297902e-05,
+      "loss": 0.4131,
+      "step": 5385
+    },
+    {
+      "epoch": 0.40031193137509746,
+      "grad_norm": 1.7080798149108887,
+      "learning_rate": 4.629870343837268e-05,
+      "loss": 0.409,
+      "step": 5390
+    },
+    {
+      "epoch": 0.40068327825021355,
+      "grad_norm": 1.1223076581954956,
+      "learning_rate": 4.629082717100286e-05,
+      "loss": 0.4065,
+      "step": 5395
+    },
+    {
+      "epoch": 0.4010546251253296,
+      "grad_norm": 0.7243461608886719,
+      "learning_rate": 4.628294320371805e-05,
+      "loss": 0.3867,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4014259720004456,
+      "grad_norm": 0.7852326035499573,
+      "learning_rate": 4.6275051539369526e-05,
+      "loss": 0.3952,
+      "step": 5405
+    },
+    {
+      "epoch": 0.40179731887556164,
+      "grad_norm": 0.8796769976615906,
+      "learning_rate": 4.626715218081134e-05,
+      "loss": 0.4228,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4021686657506777,
+      "grad_norm": 0.9907186031341553,
+      "learning_rate": 4.625924513090031e-05,
+      "loss": 0.4161,
+      "step": 5415
+    },
+    {
+      "epoch": 0.40254001262579375,
+      "grad_norm": 1.3190866708755493,
+      "learning_rate": 4.625133039249607e-05,
+      "loss": 0.4085,
+      "step": 5420
+    },
+    {
+      "epoch": 0.4029113595009098,
+      "grad_norm": 1.0041900873184204,
+      "learning_rate": 4.624340796846102e-05,
+      "loss": 0.4146,
+      "step": 5425
+    },
+    {
+      "epoch": 0.40328270637602587,
+      "grad_norm": 0.9058469533920288,
+      "learning_rate": 4.6235477861660324e-05,
+      "loss": 0.4104,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4036540532511419,
+      "grad_norm": 0.8792639970779419,
+      "learning_rate": 4.622754007496195e-05,
+      "loss": 0.4051,
+      "step": 5435
+    },
+    {
+      "epoch": 0.4040254001262579,
+      "grad_norm": 1.6717143058776855,
+      "learning_rate": 4.6219594611236624e-05,
+      "loss": 0.4286,
+      "step": 5440
+    },
+    {
+      "epoch": 0.40439674700137396,
+      "grad_norm": 3.575216293334961,
+      "learning_rate": 4.621164147335786e-05,
+      "loss": 0.4006,
+      "step": 5445
+    },
+    {
+      "epoch": 0.40476809387649004,
+      "grad_norm": 1.1020781993865967,
+      "learning_rate": 4.6203680664201934e-05,
+      "loss": 0.4101,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4051394407516061,
+      "grad_norm": 0.995681643486023,
+      "learning_rate": 4.619571218664791e-05,
+      "loss": 0.4096,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4055107876267221,
+      "grad_norm": 1.0068079233169556,
+      "learning_rate": 4.6187736043577626e-05,
+      "loss": 0.4048,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4058821345018382,
+      "grad_norm": 0.9563468098640442,
+      "learning_rate": 4.617975223787568e-05,
+      "loss": 0.4009,
+      "step": 5465
+    },
+    {
+      "epoch": 0.4062534813769542,
+      "grad_norm": 1.1841391324996948,
+      "learning_rate": 4.617176077242945e-05,
+      "loss": 0.4124,
+      "step": 5470
+    },
+    {
+      "epoch": 0.40662482825207025,
+      "grad_norm": 0.7610587477684021,
+      "learning_rate": 4.616376165012909e-05,
+      "loss": 0.3878,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4069961751271863,
+      "grad_norm": 0.8032383918762207,
+      "learning_rate": 4.615575487386751e-05,
+      "loss": 0.393,
+      "step": 5480
+    },
+    {
+      "epoch": 0.40736752200230236,
+      "grad_norm": 1.00172758102417,
+      "learning_rate": 4.614774044654038e-05,
+      "loss": 0.4099,
+      "step": 5485
+    },
+    {
+      "epoch": 0.4077388688774184,
+      "grad_norm": 0.9631612300872803,
+      "learning_rate": 4.613971837104617e-05,
+      "loss": 0.4009,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4081102157525344,
+      "grad_norm": 0.8837024569511414,
+      "learning_rate": 4.6131688650286084e-05,
+      "loss": 0.4009,
+      "step": 5495
+    },
+    {
+      "epoch": 0.4084815626276505,
+      "grad_norm": 1.4335302114486694,
+      "learning_rate": 4.612365128716412e-05,
+      "loss": 0.4,
+      "step": 5500
+    },
+    {
+      "epoch": 0.40885290950276654,
+      "grad_norm": 0.8482183814048767,
+      "learning_rate": 4.6115606284587e-05,
+      "loss": 0.4027,
+      "step": 5505
+    },
+    {
+      "epoch": 0.40922425637788257,
+      "grad_norm": 0.9031336903572083,
+      "learning_rate": 4.610755364546425e-05,
+      "loss": 0.4131,
+      "step": 5510
+    },
+    {
+      "epoch": 0.40959560325299865,
+      "grad_norm": 0.7115298509597778,
+      "learning_rate": 4.609949337270813e-05,
+      "loss": 0.408,
+      "step": 5515
+    },
+    {
+      "epoch": 0.4099669501281147,
+      "grad_norm": 0.9542083144187927,
+      "learning_rate": 4.609142546923369e-05,
+      "loss": 0.3955,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4103382970032307,
+      "grad_norm": 0.8037524223327637,
+      "learning_rate": 4.608334993795869e-05,
+      "loss": 0.4065,
+      "step": 5525
+    },
+    {
+      "epoch": 0.41070964387834674,
+      "grad_norm": 0.8711167573928833,
+      "learning_rate": 4.6075266781803695e-05,
+      "loss": 0.4007,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4110809907534628,
+      "grad_norm": 1.1807262897491455,
+      "learning_rate": 4.6067176003692006e-05,
+      "loss": 0.3878,
+      "step": 5535
+    },
+    {
+      "epoch": 0.41145233762857886,
+      "grad_norm": 1.0701298713684082,
+      "learning_rate": 4.60590776065497e-05,
+      "loss": 0.3864,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4118236845036949,
+      "grad_norm": 1.0017831325531006,
+      "learning_rate": 4.605097159330558e-05,
+      "loss": 0.4227,
+      "step": 5545
+    },
+    {
+      "epoch": 0.41219503137881097,
+      "grad_norm": 0.8837273716926575,
+      "learning_rate": 4.6042857966891235e-05,
+      "loss": 0.4079,
+      "step": 5550
+    },
+    {
+      "epoch": 0.412566378253927,
+      "grad_norm": 0.8195706605911255,
+      "learning_rate": 4.6034736730240975e-05,
+      "loss": 0.4209,
+      "step": 5555
+    },
+    {
+      "epoch": 0.41293772512904303,
+      "grad_norm": 0.9010388255119324,
+      "learning_rate": 4.6026607886291884e-05,
+      "loss": 0.3947,
+      "step": 5560
+    },
+    {
+      "epoch": 0.41330907200415906,
+      "grad_norm": 0.9195128679275513,
+      "learning_rate": 4.60184714379838e-05,
+      "loss": 0.4229,
+      "step": 5565
+    },
+    {
+      "epoch": 0.41368041887927515,
+      "grad_norm": 1.1202746629714966,
+      "learning_rate": 4.6010327388259286e-05,
+      "loss": 0.4052,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4140517657543912,
+      "grad_norm": 0.7508527636528015,
+      "learning_rate": 4.600217574006369e-05,
+      "loss": 0.394,
+      "step": 5575
+    },
+    {
+      "epoch": 0.4144231126295072,
+      "grad_norm": 1.0335317850112915,
+      "learning_rate": 4.599401649634507e-05,
+      "loss": 0.4025,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4147944595046233,
+      "grad_norm": 0.7544407844543457,
+      "learning_rate": 4.598584966005426e-05,
+      "loss": 0.4082,
+      "step": 5585
+    },
+    {
+      "epoch": 0.4151658063797393,
+      "grad_norm": 1.0011963844299316,
+      "learning_rate": 4.597767523414483e-05,
+      "loss": 0.4141,
+      "step": 5590
+    },
+    {
+      "epoch": 0.41553715325485535,
+      "grad_norm": 0.7946156859397888,
+      "learning_rate": 4.5969493221573114e-05,
+      "loss": 0.4134,
+      "step": 5595
+    },
+    {
+      "epoch": 0.4159085001299714,
+      "grad_norm": 0.9330390691757202,
+      "learning_rate": 4.596130362529813e-05,
+      "loss": 0.4074,
+      "step": 5600
+    },
+    {
+      "epoch": 0.41627984700508747,
+      "grad_norm": 0.9563977718353271,
+      "learning_rate": 4.59531064482817e-05,
+      "loss": 0.4138,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4166511938802035,
+      "grad_norm": 0.8964327573776245,
+      "learning_rate": 4.5944901693488375e-05,
+      "loss": 0.3875,
+      "step": 5610
+    },
+    {
+      "epoch": 0.4170225407553195,
+      "grad_norm": 0.9961819052696228,
+      "learning_rate": 4.593668936388543e-05,
+      "loss": 0.4063,
+      "step": 5615
+    },
+    {
+      "epoch": 0.4173938876304356,
+      "grad_norm": 1.026484489440918,
+      "learning_rate": 4.5928469462442883e-05,
+      "loss": 0.4261,
+      "step": 5620
+    },
+    {
+      "epoch": 0.41776523450555164,
+      "grad_norm": 0.7321872711181641,
+      "learning_rate": 4.592024199213349e-05,
+      "loss": 0.4019,
+      "step": 5625
+    },
+    {
+      "epoch": 0.41813658138066767,
+      "grad_norm": 1.0808043479919434,
+      "learning_rate": 4.5912006955932775e-05,
+      "loss": 0.4064,
+      "step": 5630
+    },
+    {
+      "epoch": 0.4185079282557837,
+      "grad_norm": 0.987091600894928,
+      "learning_rate": 4.590376435681895e-05,
+      "loss": 0.4116,
+      "step": 5635
+    },
+    {
+      "epoch": 0.4188792751308998,
+      "grad_norm": 0.9258447289466858,
+      "learning_rate": 4.589551419777299e-05,
+      "loss": 0.4051,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4192506220060158,
+      "grad_norm": 0.9010351896286011,
+      "learning_rate": 4.588725648177861e-05,
+      "loss": 0.4052,
+      "step": 5645
+    },
+    {
+      "epoch": 0.41962196888113185,
+      "grad_norm": 1.3273558616638184,
+      "learning_rate": 4.5878991211822226e-05,
+      "loss": 0.4098,
+      "step": 5650
+    },
+    {
+      "epoch": 0.41999331575624793,
+      "grad_norm": 1.0378541946411133,
+      "learning_rate": 4.587071839089302e-05,
+      "loss": 0.4009,
+      "step": 5655
+    },
+    {
+      "epoch": 0.42036466263136396,
+      "grad_norm": 1.0806586742401123,
+      "learning_rate": 4.586243802198289e-05,
+      "loss": 0.3911,
+      "step": 5660
+    },
+    {
+      "epoch": 0.42073600950648,
+      "grad_norm": 1.4784213304519653,
+      "learning_rate": 4.585415010808646e-05,
+      "loss": 0.3955,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4211073563815961,
+      "grad_norm": 1.1250252723693848,
+      "learning_rate": 4.58458546522011e-05,
+      "loss": 0.4094,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4214787032567121,
+      "grad_norm": 1.0189043283462524,
+      "learning_rate": 4.583755165732689e-05,
+      "loss": 0.4241,
+      "step": 5675
+    },
+    {
+      "epoch": 0.42185005013182814,
+      "grad_norm": 1.0218403339385986,
+      "learning_rate": 4.582924112646664e-05,
+      "loss": 0.3957,
+      "step": 5680
+    },
+    {
+      "epoch": 0.42222139700694417,
+      "grad_norm": 1.075886607170105,
+      "learning_rate": 4.582092306262588e-05,
+      "loss": 0.3923,
+      "step": 5685
+    },
+    {
+      "epoch": 0.42259274388206025,
+      "grad_norm": 1.131255030632019,
+      "learning_rate": 4.581259746881288e-05,
+      "loss": 0.4157,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4229640907571763,
+      "grad_norm": 0.8696762323379517,
+      "learning_rate": 4.580426434803864e-05,
+      "loss": 0.3969,
+      "step": 5695
+    },
+    {
+      "epoch": 0.4233354376322923,
+      "grad_norm": 0.8163397312164307,
+      "learning_rate": 4.579592370331684e-05,
+      "loss": 0.3903,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4237067845074084,
+      "grad_norm": 0.7770934700965881,
+      "learning_rate": 4.578757553766392e-05,
+      "loss": 0.3879,
+      "step": 5705
+    },
+    {
+      "epoch": 0.4240781313825244,
+      "grad_norm": 1.1027836799621582,
+      "learning_rate": 4.5779219854099034e-05,
+      "loss": 0.3994,
+      "step": 5710
+    },
+    {
+      "epoch": 0.42444947825764046,
+      "grad_norm": 1.2394556999206543,
+      "learning_rate": 4.577085665564404e-05,
+      "loss": 0.4253,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4248208251327565,
+      "grad_norm": 0.9389461874961853,
+      "learning_rate": 4.576248594532354e-05,
+      "loss": 0.4327,
+      "step": 5720
+    },
+    {
+      "epoch": 0.42519217200787257,
+      "grad_norm": 1.0077425241470337,
+      "learning_rate": 4.575410772616481e-05,
+      "loss": 0.4028,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4255635188829886,
+      "grad_norm": 1.0668855905532837,
+      "learning_rate": 4.5745722001197886e-05,
+      "loss": 0.3972,
+      "step": 5730
+    },
+    {
+      "epoch": 0.42593486575810463,
+      "grad_norm": 0.8116600513458252,
+      "learning_rate": 4.57373287734555e-05,
+      "loss": 0.4024,
+      "step": 5735
+    },
+    {
+      "epoch": 0.4263062126332207,
+      "grad_norm": 0.8789268136024475,
+      "learning_rate": 4.5728928045973084e-05,
+      "loss": 0.424,
+      "step": 5740
+    },
+    {
+      "epoch": 0.42667755950833675,
+      "grad_norm": 1.0181138515472412,
+      "learning_rate": 4.5720519821788806e-05,
+      "loss": 0.4101,
+      "step": 5745
+    },
+    {
+      "epoch": 0.4270489063834528,
+      "grad_norm": 0.6793472170829773,
+      "learning_rate": 4.571210410394353e-05,
+      "loss": 0.3875,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4274202532585688,
+      "grad_norm": 0.7746472954750061,
+      "learning_rate": 4.570368089548084e-05,
+      "loss": 0.4065,
+      "step": 5755
+    },
+    {
+      "epoch": 0.4277916001336849,
+      "grad_norm": 0.8889456391334534,
+      "learning_rate": 4.569525019944701e-05,
+      "loss": 0.4063,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4281629470088009,
+      "grad_norm": 0.8817050457000732,
+      "learning_rate": 4.568681201889107e-05,
+      "loss": 0.4093,
+      "step": 5765
+    },
+    {
+      "epoch": 0.42853429388391695,
+      "grad_norm": 0.9229766726493835,
+      "learning_rate": 4.567836635686468e-05,
+      "loss": 0.3724,
+      "step": 5770
+    },
+    {
+      "epoch": 0.42890564075903304,
+      "grad_norm": 1.0542956590652466,
+      "learning_rate": 4.5669913216422275e-05,
+      "loss": 0.408,
+      "step": 5775
+    },
+    {
+      "epoch": 0.42927698763414907,
+      "grad_norm": 0.7305135130882263,
+      "learning_rate": 4.566145260062096e-05,
+      "loss": 0.3989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4296483345092651,
+      "grad_norm": 1.04397451877594,
+      "learning_rate": 4.565298451252055e-05,
+      "loss": 0.4075,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4300196813843811,
+      "grad_norm": 1.1163504123687744,
+      "learning_rate": 4.5644508955183566e-05,
+      "loss": 0.3953,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4303910282594972,
+      "grad_norm": 0.823851466178894,
+      "learning_rate": 4.5636025931675225e-05,
+      "loss": 0.4168,
+      "step": 5795
+    },
+    {
+      "epoch": 0.43076237513461324,
+      "grad_norm": 0.857819676399231,
+      "learning_rate": 4.562753544506345e-05,
+      "loss": 0.4047,
+      "step": 5800
+    },
+    {
+      "epoch": 0.43113372200972927,
+      "grad_norm": 0.8693041205406189,
+      "learning_rate": 4.5619037498418865e-05,
+      "loss": 0.4103,
+      "step": 5805
+    },
+    {
+      "epoch": 0.43150506888484536,
+      "grad_norm": 1.0778625011444092,
+      "learning_rate": 4.561053209481478e-05,
+      "loss": 0.4136,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4318764157599614,
+      "grad_norm": 0.9446026086807251,
+      "learning_rate": 4.5602019237327214e-05,
+      "loss": 0.4157,
+      "step": 5815
+    },
+    {
+      "epoch": 0.4322477626350774,
+      "grad_norm": 1.0272636413574219,
+      "learning_rate": 4.5593498929034875e-05,
+      "loss": 0.3854,
+      "step": 5820
+    },
+    {
+      "epoch": 0.43261910951019344,
+      "grad_norm": 1.2057569026947021,
+      "learning_rate": 4.5584971173019163e-05,
+      "loss": 0.3711,
+      "step": 5825
+    },
+    {
+      "epoch": 0.43299045638530953,
+      "grad_norm": 0.9038252234458923,
+      "learning_rate": 4.5576435972364186e-05,
+      "loss": 0.4098,
+      "step": 5830
+    },
+    {
+      "epoch": 0.43336180326042556,
+      "grad_norm": 0.7703349590301514,
+      "learning_rate": 4.556789333015673e-05,
+      "loss": 0.4069,
+      "step": 5835
+    },
+    {
+      "epoch": 0.4337331501355416,
+      "grad_norm": 0.9097495079040527,
+      "learning_rate": 4.5559343249486264e-05,
+      "loss": 0.4208,
+      "step": 5840
+    },
+    {
+      "epoch": 0.4341044970106577,
+      "grad_norm": 1.0321422815322876,
+      "learning_rate": 4.555078573344498e-05,
+      "loss": 0.3767,
+      "step": 5845
+    },
+    {
+      "epoch": 0.4344758438857737,
+      "grad_norm": 0.9304498434066772,
+      "learning_rate": 4.554222078512772e-05,
+      "loss": 0.403,
+      "step": 5850
+    },
+    {
+      "epoch": 0.43484719076088973,
+      "grad_norm": 0.9338169693946838,
+      "learning_rate": 4.553364840763204e-05,
+      "loss": 0.4111,
+      "step": 5855
+    },
+    {
+      "epoch": 0.4352185376360058,
+      "grad_norm": 0.8401198387145996,
+      "learning_rate": 4.5525068604058193e-05,
+      "loss": 0.3897,
+      "step": 5860
+    },
+    {
+      "epoch": 0.43558988451112185,
+      "grad_norm": 1.1248438358306885,
+      "learning_rate": 4.551648137750908e-05,
+      "loss": 0.3891,
+      "step": 5865
+    },
+    {
+      "epoch": 0.4359612313862379,
+      "grad_norm": 0.7891606092453003,
+      "learning_rate": 4.550788673109031e-05,
+      "loss": 0.399,
+      "step": 5870
+    },
+    {
+      "epoch": 0.4363325782613539,
+      "grad_norm": 0.7357394695281982,
+      "learning_rate": 4.549928466791017e-05,
+      "loss": 0.4071,
+      "step": 5875
+    },
+    {
+      "epoch": 0.43670392513647,
+      "grad_norm": 1.5309947729110718,
+      "learning_rate": 4.5490675191079635e-05,
+      "loss": 0.4035,
+      "step": 5880
+    },
+    {
+      "epoch": 0.437075272011586,
+      "grad_norm": 0.8977563977241516,
+      "learning_rate": 4.548205830371236e-05,
+      "loss": 0.4206,
+      "step": 5885
+    },
+    {
+      "epoch": 0.43744661888670205,
+      "grad_norm": 0.9142330288887024,
+      "learning_rate": 4.5473434008924675e-05,
+      "loss": 0.4124,
+      "step": 5890
+    },
+    {
+      "epoch": 0.43781796576181814,
+      "grad_norm": 0.9033256769180298,
+      "learning_rate": 4.5464802309835597e-05,
+      "loss": 0.393,
+      "step": 5895
+    },
+    {
+      "epoch": 0.43818931263693417,
+      "grad_norm": 1.1107105016708374,
+      "learning_rate": 4.5456163209566806e-05,
+      "loss": 0.4233,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4385606595120502,
+      "grad_norm": 1.1025406122207642,
+      "learning_rate": 4.544751671124267e-05,
+      "loss": 0.4062,
+      "step": 5905
+    },
+    {
+      "epoch": 0.43893200638716623,
+      "grad_norm": 0.7960245013237,
+      "learning_rate": 4.543886281799023e-05,
+      "loss": 0.4196,
+      "step": 5910
+    },
+    {
+      "epoch": 0.4393033532622823,
+      "grad_norm": 0.8743290901184082,
+      "learning_rate": 4.543020153293922e-05,
+      "loss": 0.3964,
+      "step": 5915
+    },
+    {
+      "epoch": 0.43967470013739834,
+      "grad_norm": 0.8431464433670044,
+      "learning_rate": 4.5421532859222e-05,
+      "loss": 0.4154,
+      "step": 5920
+    },
+    {
+      "epoch": 0.4400460470125144,
+      "grad_norm": 1.0902396440505981,
+      "learning_rate": 4.541285679997365e-05,
+      "loss": 0.4144,
+      "step": 5925
+    },
+    {
+      "epoch": 0.44041739388763046,
+      "grad_norm": 0.6638891100883484,
+      "learning_rate": 4.54041733583319e-05,
+      "loss": 0.3975,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4407887407627465,
+      "grad_norm": 0.7656953930854797,
+      "learning_rate": 4.5395482537437154e-05,
+      "loss": 0.3902,
+      "step": 5935
+    },
+    {
+      "epoch": 0.4411600876378625,
+      "grad_norm": 0.8354926109313965,
+      "learning_rate": 4.538678434043246e-05,
+      "loss": 0.3898,
+      "step": 5940
+    },
+    {
+      "epoch": 0.44153143451297855,
+      "grad_norm": 1.043817400932312,
+      "learning_rate": 4.537807877046359e-05,
+      "loss": 0.399,
+      "step": 5945
+    },
+    {
+      "epoch": 0.44190278138809463,
+      "grad_norm": 1.3780771493911743,
+      "learning_rate": 4.536936583067895e-05,
+      "loss": 0.412,
+      "step": 5950
+    },
+    {
+      "epoch": 0.44227412826321066,
+      "grad_norm": 0.8868488669395447,
+      "learning_rate": 4.5360645524229576e-05,
+      "loss": 0.4185,
+      "step": 5955
+    },
+    {
+      "epoch": 0.4426454751383267,
+      "grad_norm": 0.7088744044303894,
+      "learning_rate": 4.535191785426922e-05,
+      "loss": 0.3828,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4430168220134428,
+      "grad_norm": 0.7177873849868774,
+      "learning_rate": 4.534318282395429e-05,
+      "loss": 0.3717,
+      "step": 5965
+    },
+    {
+      "epoch": 0.4433881688885588,
+      "grad_norm": 0.8855662941932678,
+      "learning_rate": 4.5334440436443837e-05,
+      "loss": 0.3969,
+      "step": 5970
+    },
+    {
+      "epoch": 0.44375951576367484,
+      "grad_norm": 0.7701443433761597,
+      "learning_rate": 4.532569069489958e-05,
+      "loss": 0.4083,
+      "step": 5975
+    },
+    {
+      "epoch": 0.44413086263879087,
+      "grad_norm": 1.0092111825942993,
+      "learning_rate": 4.53169336024859e-05,
+      "loss": 0.3955,
+      "step": 5980
+    },
+    {
+      "epoch": 0.44450220951390695,
+      "grad_norm": 1.2627179622650146,
+      "learning_rate": 4.530816916236984e-05,
+      "loss": 0.4155,
+      "step": 5985
+    },
+    {
+      "epoch": 0.444873556389023,
+      "grad_norm": 1.103607177734375,
+      "learning_rate": 4.529939737772109e-05,
+      "loss": 0.3959,
+      "step": 5990
+    },
+    {
+      "epoch": 0.445244903264139,
+      "grad_norm": 0.8310053944587708,
+      "learning_rate": 4.5290618251711994e-05,
+      "loss": 0.4192,
+      "step": 5995
+    },
+    {
+      "epoch": 0.4456162501392551,
+      "grad_norm": 0.9397602677345276,
+      "learning_rate": 4.528183178751758e-05,
+      "loss": 0.3736,
+      "step": 6000
+    },
+    {
+      "epoch": 0.44598759701437113,
+      "grad_norm": 0.8605632185935974,
+      "learning_rate": 4.5273037988315495e-05,
+      "loss": 0.3898,
+      "step": 6005
+    },
+    {
+      "epoch": 0.44635894388948716,
+      "grad_norm": 0.933535635471344,
+      "learning_rate": 4.5264236857286055e-05,
+      "loss": 0.4008,
+      "step": 6010
+    },
+    {
+      "epoch": 0.4467302907646032,
+      "grad_norm": 0.8004394769668579,
+      "learning_rate": 4.525542839761222e-05,
+      "loss": 0.4177,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4471016376397193,
+      "grad_norm": 1.1060221195220947,
+      "learning_rate": 4.5246612612479625e-05,
+      "loss": 0.4068,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4474729845148353,
+      "grad_norm": 0.959685742855072,
+      "learning_rate": 4.52377895050765e-05,
+      "loss": 0.4194,
+      "step": 6025
+    },
+    {
+      "epoch": 0.44784433138995133,
+      "grad_norm": 0.8318073749542236,
+      "learning_rate": 4.522895907859379e-05,
+      "loss": 0.3829,
+      "step": 6030
+    },
+    {
+      "epoch": 0.4482156782650674,
+      "grad_norm": 0.7173587679862976,
+      "learning_rate": 4.522012133622504e-05,
+      "loss": 0.3829,
+      "step": 6035
+    },
+    {
+      "epoch": 0.44858702514018345,
+      "grad_norm": 0.8138423562049866,
+      "learning_rate": 4.521127628116646e-05,
+      "loss": 0.3977,
+      "step": 6040
+    },
+    {
+      "epoch": 0.4489583720152995,
+      "grad_norm": 0.7958429455757141,
+      "learning_rate": 4.52024239166169e-05,
+      "loss": 0.3835,
+      "step": 6045
+    },
+    {
+      "epoch": 0.44932971889041556,
+      "grad_norm": 0.779323399066925,
+      "learning_rate": 4.519356424577786e-05,
+      "loss": 0.3854,
+      "step": 6050
+    },
+    {
+      "epoch": 0.4497010657655316,
+      "grad_norm": 0.73027503490448,
+      "learning_rate": 4.518469727185346e-05,
+      "loss": 0.3968,
+      "step": 6055
+    },
+    {
+      "epoch": 0.4500724126406476,
+      "grad_norm": 0.8514652252197266,
+      "learning_rate": 4.5175822998050494e-05,
+      "loss": 0.3736,
+      "step": 6060
+    },
+    {
+      "epoch": 0.45044375951576365,
+      "grad_norm": 0.7881259322166443,
+      "learning_rate": 4.516694142757837e-05,
+      "loss": 0.4076,
+      "step": 6065
+    },
+    {
+      "epoch": 0.45081510639087974,
+      "grad_norm": 0.7900510430335999,
+      "learning_rate": 4.515805256364914e-05,
+      "loss": 0.3974,
+      "step": 6070
+    },
+    {
+      "epoch": 0.45118645326599577,
+      "grad_norm": 0.983288586139679,
+      "learning_rate": 4.5149156409477524e-05,
+      "loss": 0.3985,
+      "step": 6075
+    },
+    {
+      "epoch": 0.4515578001411118,
+      "grad_norm": 1.7363369464874268,
+      "learning_rate": 4.514025296828082e-05,
+      "loss": 0.3949,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4519291470162279,
+      "grad_norm": 0.9478625655174255,
+      "learning_rate": 4.513134224327901e-05,
+      "loss": 0.4051,
+      "step": 6085
+    },
+    {
+      "epoch": 0.4523004938913439,
+      "grad_norm": 0.7871227264404297,
+      "learning_rate": 4.5122424237694697e-05,
+      "loss": 0.4028,
+      "step": 6090
+    },
+    {
+      "epoch": 0.45267184076645994,
+      "grad_norm": 1.424820899963379,
+      "learning_rate": 4.5113498954753104e-05,
+      "loss": 0.4083,
+      "step": 6095
+    },
+    {
+      "epoch": 0.453043187641576,
+      "grad_norm": 1.3722267150878906,
+      "learning_rate": 4.51045663976821e-05,
+      "loss": 0.4084,
+      "step": 6100
+    },
+    {
+      "epoch": 0.45341453451669206,
+      "grad_norm": 0.8124660849571228,
+      "learning_rate": 4.509562656971219e-05,
+      "loss": 0.3827,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4537858813918081,
+      "grad_norm": 0.9037453532218933,
+      "learning_rate": 4.5086679474076486e-05,
+      "loss": 0.3787,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4541572282669241,
+      "grad_norm": 0.808563768863678,
+      "learning_rate": 4.5077725114010755e-05,
+      "loss": 0.3986,
+      "step": 6115
+    },
+    {
+      "epoch": 0.4545285751420402,
+      "grad_norm": 0.8395090103149414,
+      "learning_rate": 4.506876349275337e-05,
+      "loss": 0.3942,
+      "step": 6120
+    },
+    {
+      "epoch": 0.45489992201715623,
+      "grad_norm": 0.9864828586578369,
+      "learning_rate": 4.5059794613545336e-05,
+      "loss": 0.3934,
+      "step": 6125
+    },
+    {
+      "epoch": 0.45527126889227226,
+      "grad_norm": 0.7316352128982544,
+      "learning_rate": 4.505081847963029e-05,
+      "loss": 0.3583,
+      "step": 6130
+    },
+    {
+      "epoch": 0.4556426157673883,
+      "grad_norm": 0.8625379800796509,
+      "learning_rate": 4.504183509425448e-05,
+      "loss": 0.3955,
+      "step": 6135
+    },
+    {
+      "epoch": 0.4560139626425044,
+      "grad_norm": 1.1295115947723389,
+      "learning_rate": 4.50328444606668e-05,
+      "loss": 0.3918,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4563853095176204,
+      "grad_norm": 0.8367231488227844,
+      "learning_rate": 4.502384658211873e-05,
+      "loss": 0.4002,
+      "step": 6145
+    },
+    {
+      "epoch": 0.45675665639273644,
+      "grad_norm": 2.5678558349609375,
+      "learning_rate": 4.5014841461864407e-05,
+      "loss": 0.4113,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4571280032678525,
+      "grad_norm": 1.087945580482483,
+      "learning_rate": 4.500582910316056e-05,
+      "loss": 0.4016,
+      "step": 6155
+    },
+    {
+      "epoch": 0.45749935014296855,
+      "grad_norm": 0.8971551060676575,
+      "learning_rate": 4.499680950926654e-05,
+      "loss": 0.4118,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4578706970180846,
+      "grad_norm": 0.935817539691925,
+      "learning_rate": 4.4987782683444336e-05,
+      "loss": 0.3853,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4582420438932006,
+      "grad_norm": 2.3204643726348877,
+      "learning_rate": 4.497874862895852e-05,
+      "loss": 0.3958,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4586133907683167,
+      "grad_norm": 0.8300501108169556,
+      "learning_rate": 4.4969707349076314e-05,
+      "loss": 0.4185,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4589847376434327,
+      "grad_norm": 0.8631926774978638,
+      "learning_rate": 4.4960658847067516e-05,
+      "loss": 0.4038,
+      "step": 6180
+    },
+    {
+      "epoch": 0.45935608451854876,
+      "grad_norm": 0.9046585559844971,
+      "learning_rate": 4.4951603126204555e-05,
+      "loss": 0.3812,
+      "step": 6185
+    },
+    {
+      "epoch": 0.45972743139366484,
+      "grad_norm": 1.0047301054000854,
+      "learning_rate": 4.494254018976248e-05,
+      "loss": 0.4147,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4600987782687809,
+      "grad_norm": 0.6368455290794373,
+      "learning_rate": 4.493347004101894e-05,
+      "loss": 0.398,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4604701251438969,
+      "grad_norm": 0.8941465616226196,
+      "learning_rate": 4.4924392683254185e-05,
+      "loss": 0.4039,
+      "step": 6200
+    },
+    {
+      "epoch": 0.460841472019013,
+      "grad_norm": 0.8189155459403992,
+      "learning_rate": 4.4915308119751074e-05,
+      "loss": 0.4083,
+      "step": 6205
+    },
+    {
+      "epoch": 0.461212818894129,
+      "grad_norm": 1.35800302028656,
+      "learning_rate": 4.490621635379509e-05,
+      "loss": 0.3816,
+      "step": 6210
+    },
+    {
+      "epoch": 0.46158416576924505,
+      "grad_norm": 1.2649494409561157,
+      "learning_rate": 4.4897117388674296e-05,
+      "loss": 0.3959,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4619555126443611,
+      "grad_norm": 0.909601628780365,
+      "learning_rate": 4.488801122767937e-05,
+      "loss": 0.3561,
+      "step": 6220
+    },
+    {
+      "epoch": 0.46232685951947716,
+      "grad_norm": 0.8168251514434814,
+      "learning_rate": 4.4878897874103605e-05,
+      "loss": 0.3912,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4626982063945932,
+      "grad_norm": 0.7770623564720154,
+      "learning_rate": 4.4869777331242865e-05,
+      "loss": 0.4195,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4630695532697092,
+      "grad_norm": 0.8942532539367676,
+      "learning_rate": 4.4860649602395645e-05,
+      "loss": 0.4062,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4634409001448253,
+      "grad_norm": 0.7446513175964355,
+      "learning_rate": 4.485151469086303e-05,
+      "loss": 0.3851,
+      "step": 6240
+    },
+    {
+      "epoch": 0.46381224701994134,
+      "grad_norm": 1.5065016746520996,
+      "learning_rate": 4.484237259994867e-05,
+      "loss": 0.3893,
+      "step": 6245
+    },
+    {
+      "epoch": 0.46418359389505737,
+      "grad_norm": 0.8029677271842957,
+      "learning_rate": 4.483322333295887e-05,
+      "loss": 0.4138,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4645549407701734,
+      "grad_norm": 1.67062246799469,
+      "learning_rate": 4.4824066893202494e-05,
+      "loss": 0.3852,
+      "step": 6255
+    },
+    {
+      "epoch": 0.4649262876452895,
+      "grad_norm": 0.7975144386291504,
+      "learning_rate": 4.4814903283990996e-05,
+      "loss": 0.374,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4652976345204055,
+      "grad_norm": 1.0474958419799805,
+      "learning_rate": 4.4805732508638444e-05,
+      "loss": 0.3746,
+      "step": 6265
+    },
+    {
+      "epoch": 0.46566898139552154,
+      "grad_norm": 1.08444082736969,
+      "learning_rate": 4.479655457046147e-05,
+      "loss": 0.3937,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4660403282706376,
+      "grad_norm": 0.8867533802986145,
+      "learning_rate": 4.4787369472779326e-05,
+      "loss": 0.4087,
+      "step": 6275
+    },
+    {
+      "epoch": 0.46641167514575366,
+      "grad_norm": 0.954857349395752,
+      "learning_rate": 4.4778177218913844e-05,
+      "loss": 0.4139,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4667830220208697,
+      "grad_norm": 1.2051773071289062,
+      "learning_rate": 4.476897781218943e-05,
+      "loss": 0.3953,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4671543688959857,
+      "grad_norm": 0.8374889492988586,
+      "learning_rate": 4.475977125593309e-05,
+      "loss": 0.3984,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4675257157711018,
+      "grad_norm": 0.756777286529541,
+      "learning_rate": 4.475055755347443e-05,
+      "loss": 0.3882,
+      "step": 6295
+    },
+    {
+      "epoch": 0.46789706264621783,
+      "grad_norm": 0.6392567753791809,
+      "learning_rate": 4.474133670814559e-05,
+      "loss": 0.4093,
+      "step": 6300
+    },
+    {
+      "epoch": 0.46826840952133386,
+      "grad_norm": 2.0606844425201416,
+      "learning_rate": 4.4732108723281346e-05,
+      "loss": 0.3742,
+      "step": 6305
+    },
+    {
+      "epoch": 0.46863975639644995,
+      "grad_norm": 0.7938241958618164,
+      "learning_rate": 4.472287360221904e-05,
+      "loss": 0.3869,
+      "step": 6310
+    },
+    {
+      "epoch": 0.469011103271566,
+      "grad_norm": 1.1123112440109253,
+      "learning_rate": 4.471363134829858e-05,
+      "loss": 0.4054,
+      "step": 6315
+    },
+    {
+      "epoch": 0.469382450146682,
+      "grad_norm": 0.8488892316818237,
+      "learning_rate": 4.4704381964862474e-05,
+      "loss": 0.4049,
+      "step": 6320
+    },
+    {
+      "epoch": 0.46975379702179804,
+      "grad_norm": 0.8527318239212036,
+      "learning_rate": 4.4695125455255794e-05,
+      "loss": 0.4112,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4701251438969141,
+      "grad_norm": 0.9174249172210693,
+      "learning_rate": 4.46858618228262e-05,
+      "loss": 0.3904,
+      "step": 6330
+    },
+    {
+      "epoch": 0.47049649077203015,
+      "grad_norm": 1.0107132196426392,
+      "learning_rate": 4.4676591070923926e-05,
+      "loss": 0.389,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4708678376471462,
+      "grad_norm": 1.0217775106430054,
+      "learning_rate": 4.466731320290176e-05,
+      "loss": 0.405,
+      "step": 6340
+    },
+    {
+      "epoch": 0.47123918452226227,
+      "grad_norm": 1.0906291007995605,
+      "learning_rate": 4.46580282221151e-05,
+      "loss": 0.4112,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4716105313973783,
+      "grad_norm": 0.6523454785346985,
+      "learning_rate": 4.464873613192189e-05,
+      "loss": 0.3853,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4719818782724943,
+      "grad_norm": 1.7757562398910522,
+      "learning_rate": 4.4639436935682646e-05,
+      "loss": 0.3861,
+      "step": 6355
+    },
+    {
+      "epoch": 0.47235322514761036,
+      "grad_norm": 0.9793681502342224,
+      "learning_rate": 4.463013063676047e-05,
+      "loss": 0.4017,
+      "step": 6360
+    },
+    {
+      "epoch": 0.47272457202272644,
+      "grad_norm": 0.6784321069717407,
+      "learning_rate": 4.462081723852102e-05,
+      "loss": 0.3991,
+      "step": 6365
+    },
+    {
+      "epoch": 0.47309591889784247,
+      "grad_norm": 0.7017379999160767,
+      "learning_rate": 4.461149674433253e-05,
+      "loss": 0.4008,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4734672657729585,
+      "grad_norm": 0.8638973832130432,
+      "learning_rate": 4.460216915756578e-05,
+      "loss": 0.4009,
+      "step": 6375
+    },
+    {
+      "epoch": 0.4738386126480746,
+      "grad_norm": 0.7157192826271057,
+      "learning_rate": 4.4592834481594146e-05,
+      "loss": 0.3994,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4742099595231906,
+      "grad_norm": 0.721689760684967,
+      "learning_rate": 4.4583492719793544e-05,
+      "loss": 0.4016,
+      "step": 6385
+    },
+    {
+      "epoch": 0.47458130639830665,
+      "grad_norm": 0.8904692530632019,
+      "learning_rate": 4.4574143875542465e-05,
+      "loss": 0.4127,
+      "step": 6390
+    },
+    {
+      "epoch": 0.47495265327342273,
+      "grad_norm": 1.060440182685852,
+      "learning_rate": 4.456478795222195e-05,
+      "loss": 0.4021,
+      "step": 6395
+    },
+    {
+      "epoch": 0.47532400014853876,
+      "grad_norm": 0.7294143438339233,
+      "learning_rate": 4.455542495321561e-05,
+      "loss": 0.3877,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4756953470236548,
+      "grad_norm": 0.9109171628952026,
+      "learning_rate": 4.4546054881909615e-05,
+      "loss": 0.4181,
+      "step": 6405
+    },
+    {
+      "epoch": 0.4760666938987708,
+      "grad_norm": 0.8929049372673035,
+      "learning_rate": 4.453667774169269e-05,
+      "loss": 0.3796,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4764380407738869,
+      "grad_norm": 0.9146767258644104,
+      "learning_rate": 4.452729353595611e-05,
+      "loss": 0.3882,
+      "step": 6415
+    },
+    {
+      "epoch": 0.47680938764900294,
+      "grad_norm": 0.8753120303153992,
+      "learning_rate": 4.4517902268093715e-05,
+      "loss": 0.3875,
+      "step": 6420
+    },
+    {
+      "epoch": 0.47718073452411897,
+      "grad_norm": 0.7614051103591919,
+      "learning_rate": 4.45085039415019e-05,
+      "loss": 0.4054,
+      "step": 6425
+    },
+    {
+      "epoch": 0.47755208139923505,
+      "grad_norm": 1.0136183500289917,
+      "learning_rate": 4.44990985595796e-05,
+      "loss": 0.3927,
+      "step": 6430
+    },
+    {
+      "epoch": 0.4779234282743511,
+      "grad_norm": 0.9799590706825256,
+      "learning_rate": 4.4489686125728316e-05,
+      "loss": 0.3908,
+      "step": 6435
+    },
+    {
+      "epoch": 0.4782947751494671,
+      "grad_norm": 0.7973908185958862,
+      "learning_rate": 4.44802666433521e-05,
+      "loss": 0.3715,
+      "step": 6440
+    },
+    {
+      "epoch": 0.47866612202458314,
+      "grad_norm": 0.8001505732536316,
+      "learning_rate": 4.447084011585752e-05,
+      "loss": 0.4111,
+      "step": 6445
+    },
+    {
+      "epoch": 0.4790374688996992,
+      "grad_norm": 0.6913841366767883,
+      "learning_rate": 4.446140654665375e-05,
+      "loss": 0.3877,
+      "step": 6450
+    },
+    {
+      "epoch": 0.47940881577481526,
+      "grad_norm": 1.518284797668457,
+      "learning_rate": 4.445196593915246e-05,
+      "loss": 0.4072,
+      "step": 6455
+    },
+    {
+      "epoch": 0.4797801626499313,
+      "grad_norm": 0.6852445006370544,
+      "learning_rate": 4.4442518296767896e-05,
+      "loss": 0.4089,
+      "step": 6460
+    },
+    {
+      "epoch": 0.48015150952504737,
+      "grad_norm": 0.8990404605865479,
+      "learning_rate": 4.4433063622916824e-05,
+      "loss": 0.3974,
+      "step": 6465
+    },
+    {
+      "epoch": 0.4805228564001634,
+      "grad_norm": 0.9696192741394043,
+      "learning_rate": 4.4423601921018574e-05,
+      "loss": 0.398,
+      "step": 6470
+    },
+    {
+      "epoch": 0.48089420327527943,
+      "grad_norm": 0.8919362425804138,
+      "learning_rate": 4.4414133194495e-05,
+      "loss": 0.3827,
+      "step": 6475
+    },
+    {
+      "epoch": 0.48126555015039546,
+      "grad_norm": 0.8759976625442505,
+      "learning_rate": 4.4404657446770524e-05,
+      "loss": 0.3855,
+      "step": 6480
+    },
+    {
+      "epoch": 0.48163689702551155,
+      "grad_norm": 0.8538511395454407,
+      "learning_rate": 4.4395174681272076e-05,
+      "loss": 0.3788,
+      "step": 6485
+    },
+    {
+      "epoch": 0.4820082439006276,
+      "grad_norm": 0.8503746390342712,
+      "learning_rate": 4.438568490142914e-05,
+      "loss": 0.3966,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4823795907757436,
+      "grad_norm": 1.078658103942871,
+      "learning_rate": 4.437618811067373e-05,
+      "loss": 0.4003,
+      "step": 6495
+    },
+    {
+      "epoch": 0.4827509376508597,
+      "grad_norm": 0.9684752225875854,
+      "learning_rate": 4.43666843124404e-05,
+      "loss": 0.3998,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4831222845259757,
+      "grad_norm": 3.1359939575195312,
+      "learning_rate": 4.435717351016624e-05,
+      "loss": 0.4117,
+      "step": 6505
+    },
+    {
+      "epoch": 0.48349363140109175,
+      "grad_norm": 0.8860180974006653,
+      "learning_rate": 4.4347655707290874e-05,
+      "loss": 0.3987,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4838649782762078,
+      "grad_norm": 0.9404197931289673,
+      "learning_rate": 4.4338130907256445e-05,
+      "loss": 0.3865,
+      "step": 6515
+    },
+    {
+      "epoch": 0.48423632515132387,
+      "grad_norm": 1.2225645780563354,
+      "learning_rate": 4.432859911350765e-05,
+      "loss": 0.3799,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4846076720264399,
+      "grad_norm": 1.038920521736145,
+      "learning_rate": 4.431906032949169e-05,
+      "loss": 0.3978,
+      "step": 6525
+    },
+    {
+      "epoch": 0.4849790189015559,
+      "grad_norm": 1.3525471687316895,
+      "learning_rate": 4.4309514558658315e-05,
+      "loss": 0.4046,
+      "step": 6530
+    },
+    {
+      "epoch": 0.485350365776672,
+      "grad_norm": 0.8096221685409546,
+      "learning_rate": 4.4299961804459776e-05,
+      "loss": 0.4029,
+      "step": 6535
+    },
+    {
+      "epoch": 0.48572171265178804,
+      "grad_norm": 0.9344236850738525,
+      "learning_rate": 4.429040207035088e-05,
+      "loss": 0.41,
+      "step": 6540
+    },
+    {
+      "epoch": 0.48609305952690407,
+      "grad_norm": 0.9042350053787231,
+      "learning_rate": 4.428083535978894e-05,
+      "loss": 0.4128,
+      "step": 6545
+    },
+    {
+      "epoch": 0.4864644064020201,
+      "grad_norm": 1.1597720384597778,
+      "learning_rate": 4.42712616762338e-05,
+      "loss": 0.3787,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4868357532771362,
+      "grad_norm": 1.200708270072937,
+      "learning_rate": 4.426168102314781e-05,
+      "loss": 0.3972,
+      "step": 6555
+    },
+    {
+      "epoch": 0.4872071001522522,
+      "grad_norm": 1.0330638885498047,
+      "learning_rate": 4.4252093403995864e-05,
+      "loss": 0.3953,
+      "step": 6560
+    },
+    {
+      "epoch": 0.48757844702736824,
+      "grad_norm": 1.0347399711608887,
+      "learning_rate": 4.424249882224536e-05,
+      "loss": 0.4076,
+      "step": 6565
+    },
+    {
+      "epoch": 0.48794979390248433,
+      "grad_norm": 0.6727543473243713,
+      "learning_rate": 4.423289728136621e-05,
+      "loss": 0.4036,
+      "step": 6570
+    },
+    {
+      "epoch": 0.48832114077760036,
+      "grad_norm": 1.5079855918884277,
+      "learning_rate": 4.422328878483087e-05,
+      "loss": 0.3964,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4886924876527164,
+      "grad_norm": 0.678868293762207,
+      "learning_rate": 4.421367333611428e-05,
+      "loss": 0.3871,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4890638345278325,
+      "grad_norm": 1.0084245204925537,
+      "learning_rate": 4.42040509386939e-05,
+      "loss": 0.3853,
+      "step": 6585
+    },
+    {
+      "epoch": 0.4894351814029485,
+      "grad_norm": 0.8365580439567566,
+      "learning_rate": 4.4194421596049715e-05,
+      "loss": 0.4266,
+      "step": 6590
+    },
+    {
+      "epoch": 0.48980652827806453,
+      "grad_norm": 1.2573304176330566,
+      "learning_rate": 4.418478531166423e-05,
+      "loss": 0.3802,
+      "step": 6595
+    },
+    {
+      "epoch": 0.49017787515318056,
+      "grad_norm": 0.9323030114173889,
+      "learning_rate": 4.417514208902242e-05,
+      "loss": 0.4013,
+      "step": 6600
+    },
+    {
+      "epoch": 0.49054922202829665,
+      "grad_norm": 0.7535553574562073,
+      "learning_rate": 4.416549193161182e-05,
+      "loss": 0.3918,
+      "step": 6605
+    },
+    {
+      "epoch": 0.4909205689034127,
+      "grad_norm": 0.871611475944519,
+      "learning_rate": 4.415583484292245e-05,
+      "loss": 0.3729,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4912919157785287,
+      "grad_norm": 0.9813038110733032,
+      "learning_rate": 4.4146170826446806e-05,
+      "loss": 0.3681,
+      "step": 6615
+    },
+    {
+      "epoch": 0.4916632626536448,
+      "grad_norm": 0.9408519268035889,
+      "learning_rate": 4.413649988567995e-05,
+      "loss": 0.4136,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4920346095287608,
+      "grad_norm": 0.8955950140953064,
+      "learning_rate": 4.4126822024119405e-05,
+      "loss": 0.3629,
+      "step": 6625
+    },
+    {
+      "epoch": 0.49240595640387685,
+      "grad_norm": 1.0907254219055176,
+      "learning_rate": 4.4117137245265206e-05,
+      "loss": 0.401,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4927773032789929,
+      "grad_norm": 0.7147353291511536,
+      "learning_rate": 4.41074455526199e-05,
+      "loss": 0.3912,
+      "step": 6635
+    },
+    {
+      "epoch": 0.49314865015410897,
+      "grad_norm": 1.297784447669983,
+      "learning_rate": 4.409774694968853e-05,
+      "loss": 0.4182,
+      "step": 6640
+    },
+    {
+      "epoch": 0.493519997029225,
+      "grad_norm": 0.7881303429603577,
+      "learning_rate": 4.408804143997862e-05,
+      "loss": 0.3971,
+      "step": 6645
+    },
+    {
+      "epoch": 0.49389134390434103,
+      "grad_norm": 1.1137521266937256,
+      "learning_rate": 4.4078329027000226e-05,
+      "loss": 0.3955,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4942626907794571,
+      "grad_norm": 0.8524259328842163,
+      "learning_rate": 4.406860971426587e-05,
+      "loss": 0.3625,
+      "step": 6655
+    },
+    {
+      "epoch": 0.49463403765457314,
+      "grad_norm": 0.8187363743782043,
+      "learning_rate": 4.4058883505290586e-05,
+      "loss": 0.4043,
+      "step": 6660
+    },
+    {
+      "epoch": 0.4950053845296892,
+      "grad_norm": 1.0168330669403076,
+      "learning_rate": 4.40491504035919e-05,
+      "loss": 0.3851,
+      "step": 6665
+    },
+    {
+      "epoch": 0.4953767314048052,
+      "grad_norm": 0.7828294038772583,
+      "learning_rate": 4.403941041268982e-05,
+      "loss": 0.4131,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4957480782799213,
+      "grad_norm": 0.966033399105072,
+      "learning_rate": 4.402966353610686e-05,
+      "loss": 0.3716,
+      "step": 6675
+    },
+    {
+      "epoch": 0.4961194251550373,
+      "grad_norm": 2.529012680053711,
+      "learning_rate": 4.401990977736802e-05,
+      "loss": 0.3974,
+      "step": 6680
+    },
+    {
+      "epoch": 0.49649077203015335,
+      "grad_norm": 0.6573621034622192,
+      "learning_rate": 4.401014914000078e-05,
+      "loss": 0.3743,
+      "step": 6685
+    },
+    {
+      "epoch": 0.49686211890526943,
+      "grad_norm": 0.9817414879798889,
+      "learning_rate": 4.4000381627535114e-05,
+      "loss": 0.3809,
+      "step": 6690
+    },
+    {
+      "epoch": 0.49723346578038546,
+      "grad_norm": 0.7202504873275757,
+      "learning_rate": 4.399060724350349e-05,
+      "loss": 0.4057,
+      "step": 6695
+    },
+    {
+      "epoch": 0.4976048126555015,
+      "grad_norm": 0.9702993035316467,
+      "learning_rate": 4.398082599144085e-05,
+      "loss": 0.3853,
+      "step": 6700
+    },
+    {
+      "epoch": 0.4979761595306175,
+      "grad_norm": 0.7524059414863586,
+      "learning_rate": 4.397103787488462e-05,
+      "loss": 0.3919,
+      "step": 6705
+    },
+    {
+      "epoch": 0.4983475064057336,
+      "grad_norm": 0.7320048213005066,
+      "learning_rate": 4.396124289737472e-05,
+      "loss": 0.3848,
+      "step": 6710
+    },
+    {
+      "epoch": 0.49871885328084964,
+      "grad_norm": 2.150871753692627,
+      "learning_rate": 4.395144106245355e-05,
+      "loss": 0.3875,
+      "step": 6715
+    },
+    {
+      "epoch": 0.49909020015596567,
+      "grad_norm": 0.8532472252845764,
+      "learning_rate": 4.394163237366596e-05,
+      "loss": 0.4054,
+      "step": 6720
+    },
+    {
+      "epoch": 0.49946154703108175,
+      "grad_norm": 0.9899107813835144,
+      "learning_rate": 4.393181683455933e-05,
+      "loss": 0.3891,
+      "step": 6725
+    },
+    {
+      "epoch": 0.4998328939061978,
+      "grad_norm": 1.153525471687317,
+      "learning_rate": 4.392199444868347e-05,
+      "loss": 0.4071,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5002042407813139,
+      "grad_norm": 0.8968061208724976,
+      "learning_rate": 4.3912165219590703e-05,
+      "loss": 0.3756,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5005755876564298,
+      "grad_norm": 0.8012394905090332,
+      "learning_rate": 4.390232915083579e-05,
+      "loss": 0.4128,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5009469345315459,
+      "grad_norm": 1.1018304824829102,
+      "learning_rate": 4.389248624597599e-05,
+      "loss": 0.3938,
+      "step": 6745
+    },
+    {
+      "epoch": 0.501318281406662,
+      "grad_norm": 0.9241147637367249,
+      "learning_rate": 4.3882636508571054e-05,
+      "loss": 0.3906,
+      "step": 6750
+    },
+    {
+      "epoch": 0.501689628281778,
+      "grad_norm": 0.9045403599739075,
+      "learning_rate": 4.3872779942183154e-05,
+      "loss": 0.3939,
+      "step": 6755
+    },
+    {
+      "epoch": 0.5020609751568941,
+      "grad_norm": 0.6724326014518738,
+      "learning_rate": 4.3862916550376964e-05,
+      "loss": 0.4002,
+      "step": 6760
+    },
+    {
+      "epoch": 0.50243232203201,
+      "grad_norm": 0.8588703870773315,
+      "learning_rate": 4.385304633671962e-05,
+      "loss": 0.3885,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5028036689071261,
+      "grad_norm": 0.9330534934997559,
+      "learning_rate": 4.3843169304780726e-05,
+      "loss": 0.3891,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5031750157822422,
+      "grad_norm": 1.3435461521148682,
+      "learning_rate": 4.3833285458132364e-05,
+      "loss": 0.3868,
+      "step": 6775
+    },
+    {
+      "epoch": 0.5035463626573582,
+      "grad_norm": 0.771759033203125,
+      "learning_rate": 4.3823394800349046e-05,
+      "loss": 0.3999,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5039177095324743,
+      "grad_norm": 0.7695342302322388,
+      "learning_rate": 4.381349733500778e-05,
+      "loss": 0.3824,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5042890564075904,
+      "grad_norm": 0.8571486473083496,
+      "learning_rate": 4.380359306568802e-05,
+      "loss": 0.3814,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5046604032827063,
+      "grad_norm": 0.8740094304084778,
+      "learning_rate": 4.379368199597169e-05,
+      "loss": 0.3879,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5050317501578224,
+      "grad_norm": 0.8141562342643738,
+      "learning_rate": 4.378376412944317e-05,
+      "loss": 0.378,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5054030970329385,
+      "grad_norm": 1.1338756084442139,
+      "learning_rate": 4.37738394696893e-05,
+      "loss": 0.4075,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5057744439080545,
+      "grad_norm": 1.0660085678100586,
+      "learning_rate": 4.376390802029937e-05,
+      "loss": 0.3672,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5061457907831706,
+      "grad_norm": 1.1899162530899048,
+      "learning_rate": 4.3753969784865123e-05,
+      "loss": 0.3949,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5065171376582867,
+      "grad_norm": 0.7957924008369446,
+      "learning_rate": 4.374402476698077e-05,
+      "loss": 0.3909,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5068884845334026,
+      "grad_norm": 0.8344728350639343,
+      "learning_rate": 4.3734072970242965e-05,
+      "loss": 0.3837,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5072598314085187,
+      "grad_norm": 0.9248842000961304,
+      "learning_rate": 4.372411439825082e-05,
+      "loss": 0.3793,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5076311782836348,
+      "grad_norm": 0.811426043510437,
+      "learning_rate": 4.371414905460589e-05,
+      "loss": 0.4047,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5080025251587508,
+      "grad_norm": 2.3013105392456055,
+      "learning_rate": 4.370417694291218e-05,
+      "loss": 0.38,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5083738720338669,
+      "grad_norm": 0.911200225353241,
+      "learning_rate": 4.369419806677615e-05,
+      "loss": 0.4152,
+      "step": 6845
+    },
+    {
+      "epoch": 0.5087452189089828,
+      "grad_norm": 1.1220375299453735,
+      "learning_rate": 4.368421242980669e-05,
+      "loss": 0.3996,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5091165657840989,
+      "grad_norm": 0.8643706440925598,
+      "learning_rate": 4.3674220035615165e-05,
+      "loss": 0.3748,
+      "step": 6855
+    },
+    {
+      "epoch": 0.509487912659215,
+      "grad_norm": 0.8339560627937317,
+      "learning_rate": 4.3664220887815364e-05,
+      "loss": 0.3778,
+      "step": 6860
+    },
+    {
+      "epoch": 0.509859259534331,
+      "grad_norm": 0.9615778923034668,
+      "learning_rate": 4.3654214990023504e-05,
+      "loss": 0.3862,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5102306064094471,
+      "grad_norm": 1.6474332809448242,
+      "learning_rate": 4.364420234585827e-05,
+      "loss": 0.4069,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5106019532845631,
+      "grad_norm": 0.6933843493461609,
+      "learning_rate": 4.3634182958940766e-05,
+      "loss": 0.3902,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5109733001596791,
+      "grad_norm": 0.7939101457595825,
+      "learning_rate": 4.362415683289456e-05,
+      "loss": 0.3915,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5113446470347952,
+      "grad_norm": 0.9869307279586792,
+      "learning_rate": 4.361412397134562e-05,
+      "loss": 0.3807,
+      "step": 6885
+    },
+    {
+      "epoch": 0.5117159939099113,
+      "grad_norm": 0.9772399663925171,
+      "learning_rate": 4.360408437792239e-05,
+      "loss": 0.3938,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5120873407850273,
+      "grad_norm": 0.8522541522979736,
+      "learning_rate": 4.3594038056255734e-05,
+      "loss": 0.3959,
+      "step": 6895
+    },
+    {
+      "epoch": 0.5124586876601434,
+      "grad_norm": 0.6819554567337036,
+      "learning_rate": 4.358398500997893e-05,
+      "loss": 0.3797,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5128300345352594,
+      "grad_norm": 0.656989574432373,
+      "learning_rate": 4.357392524272771e-05,
+      "loss": 0.3808,
+      "step": 6905
+    },
+    {
+      "epoch": 0.5132013814103754,
+      "grad_norm": 1.0240429639816284,
+      "learning_rate": 4.356385875814023e-05,
+      "loss": 0.391,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5135727282854915,
+      "grad_norm": 0.9838010668754578,
+      "learning_rate": 4.3553785559857074e-05,
+      "loss": 0.3858,
+      "step": 6915
+    },
+    {
+      "epoch": 0.5139440751606075,
+      "grad_norm": 0.9715674519538879,
+      "learning_rate": 4.354370565152126e-05,
+      "loss": 0.3989,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5143154220357236,
+      "grad_norm": 0.8658883571624756,
+      "learning_rate": 4.3533619036778216e-05,
+      "loss": 0.3905,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5146867689108396,
+      "grad_norm": 1.0177209377288818,
+      "learning_rate": 4.352352571927583e-05,
+      "loss": 0.3768,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5150581157859556,
+      "grad_norm": 0.9606013298034668,
+      "learning_rate": 4.3513425702664365e-05,
+      "loss": 0.3922,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5154294626610717,
+      "grad_norm": 1.0270155668258667,
+      "learning_rate": 4.350331899059655e-05,
+      "loss": 0.3806,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5158008095361878,
+      "grad_norm": 0.8382008671760559,
+      "learning_rate": 4.349320558672752e-05,
+      "loss": 0.3759,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5161721564113038,
+      "grad_norm": 0.8334744572639465,
+      "learning_rate": 4.348308549471481e-05,
+      "loss": 0.3932,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5165435032864198,
+      "grad_norm": 0.9420961737632751,
+      "learning_rate": 4.347295871821843e-05,
+      "loss": 0.3849,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5169148501615359,
+      "grad_norm": 1.008453130722046,
+      "learning_rate": 4.346282526090073e-05,
+      "loss": 0.3829,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5172861970366519,
+      "grad_norm": 0.9556795358657837,
+      "learning_rate": 4.345268512642654e-05,
+      "loss": 0.4101,
+      "step": 6965
+    },
+    {
+      "epoch": 0.517657543911768,
+      "grad_norm": 0.8188214302062988,
+      "learning_rate": 4.344253831846308e-05,
+      "loss": 0.3921,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5180288907868841,
+      "grad_norm": 0.8770802617073059,
+      "learning_rate": 4.3432384840679974e-05,
+      "loss": 0.3884,
+      "step": 6975
+    },
+    {
+      "epoch": 0.518400237662,
+      "grad_norm": 0.8020485639572144,
+      "learning_rate": 4.342222469674927e-05,
+      "loss": 0.3975,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5187715845371161,
+      "grad_norm": 0.8058055639266968,
+      "learning_rate": 4.3412057890345435e-05,
+      "loss": 0.3952,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5191429314122321,
+      "grad_norm": 1.0167996883392334,
+      "learning_rate": 4.3401884425145336e-05,
+      "loss": 0.3838,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5195142782873482,
+      "grad_norm": 0.8972017168998718,
+      "learning_rate": 4.339170430482824e-05,
+      "loss": 0.3993,
+      "step": 6995
+    },
+    {
+      "epoch": 0.5198856251624643,
+      "grad_norm": 0.9788426756858826,
+      "learning_rate": 4.338151753307583e-05,
+      "loss": 0.3954,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5202569720375803,
+      "grad_norm": 0.8452064990997314,
+      "learning_rate": 4.3371324113572205e-05,
+      "loss": 0.3961,
+      "step": 7005
+    },
+    {
+      "epoch": 0.5206283189126963,
+      "grad_norm": 0.7179628014564514,
+      "learning_rate": 4.3361124050003844e-05,
+      "loss": 0.394,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5209996657878124,
+      "grad_norm": 0.9873393177986145,
+      "learning_rate": 4.335091734605965e-05,
+      "loss": 0.4077,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5213710126629284,
+      "grad_norm": 0.9680303931236267,
+      "learning_rate": 4.334070400543091e-05,
+      "loss": 0.4056,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5217423595380445,
+      "grad_norm": 1.9106343984603882,
+      "learning_rate": 4.3330484031811336e-05,
+      "loss": 0.3935,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5221137064131606,
+      "grad_norm": 0.9799425601959229,
+      "learning_rate": 4.332025742889701e-05,
+      "loss": 0.3763,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5224850532882765,
+      "grad_norm": 0.9366163611412048,
+      "learning_rate": 4.331002420038642e-05,
+      "loss": 0.3861,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5228564001633926,
+      "grad_norm": 0.8720141053199768,
+      "learning_rate": 4.329978434998048e-05,
+      "loss": 0.3852,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5232277470385087,
+      "grad_norm": 0.8284469842910767,
+      "learning_rate": 4.3289537881382445e-05,
+      "loss": 0.3941,
+      "step": 7045
+    },
+    {
+      "epoch": 0.5235990939136247,
+      "grad_norm": 0.9448261857032776,
+      "learning_rate": 4.327928479829801e-05,
+      "loss": 0.3827,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5239704407887408,
+      "grad_norm": 0.9553066492080688,
+      "learning_rate": 4.326902510443524e-05,
+      "loss": 0.3927,
+      "step": 7055
+    },
+    {
+      "epoch": 0.5243417876638569,
+      "grad_norm": 0.9436995983123779,
+      "learning_rate": 4.325875880350461e-05,
+      "loss": 0.4134,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5247131345389728,
+      "grad_norm": 2.108335494995117,
+      "learning_rate": 4.3248485899218935e-05,
+      "loss": 0.3881,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5250844814140889,
+      "grad_norm": 1.2432762384414673,
+      "learning_rate": 4.323820639529348e-05,
+      "loss": 0.3905,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5254558282892049,
+      "grad_norm": 0.7592414021492004,
+      "learning_rate": 4.3227920295445866e-05,
+      "loss": 0.3824,
+      "step": 7075
+    },
+    {
+      "epoch": 0.525827175164321,
+      "grad_norm": 1.110966682434082,
+      "learning_rate": 4.32176276033961e-05,
+      "loss": 0.3664,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5261985220394371,
+      "grad_norm": 0.959549069404602,
+      "learning_rate": 4.320732832286657e-05,
+      "loss": 0.3825,
+      "step": 7085
+    },
+    {
+      "epoch": 0.526569868914553,
+      "grad_norm": 0.8384568691253662,
+      "learning_rate": 4.3197022457582065e-05,
+      "loss": 0.4071,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5269412157896691,
+      "grad_norm": 0.8591688275337219,
+      "learning_rate": 4.318671001126974e-05,
+      "loss": 0.3991,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5273125626647852,
+      "grad_norm": 0.9322898983955383,
+      "learning_rate": 4.3176390987659134e-05,
+      "loss": 0.3788,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5276839095399012,
+      "grad_norm": 0.8797940611839294,
+      "learning_rate": 4.3166065390482146e-05,
+      "loss": 0.4064,
+      "step": 7105
+    },
+    {
+      "epoch": 0.5280552564150173,
+      "grad_norm": 0.886881411075592,
+      "learning_rate": 4.3155733223473103e-05,
+      "loss": 0.3957,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5284266032901334,
+      "grad_norm": 0.8280029296875,
+      "learning_rate": 4.314539449036865e-05,
+      "loss": 0.4041,
+      "step": 7115
+    },
+    {
+      "epoch": 0.5287979501652493,
+      "grad_norm": 0.8642339706420898,
+      "learning_rate": 4.313504919490785e-05,
+      "loss": 0.3824,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5291692970403654,
+      "grad_norm": 0.8907843232154846,
+      "learning_rate": 4.31246973408321e-05,
+      "loss": 0.3912,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5295406439154815,
+      "grad_norm": 0.8841026425361633,
+      "learning_rate": 4.311433893188521e-05,
+      "loss": 0.3776,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5299119907905975,
+      "grad_norm": 0.8547298312187195,
+      "learning_rate": 4.310397397181334e-05,
+      "loss": 0.3878,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5302833376657136,
+      "grad_norm": 1.2953871488571167,
+      "learning_rate": 4.3093602464365e-05,
+      "loss": 0.3881,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5306546845408295,
+      "grad_norm": 0.9269512295722961,
+      "learning_rate": 4.308322441329112e-05,
+      "loss": 0.3754,
+      "step": 7145
+    },
+    {
+      "epoch": 0.5310260314159456,
+      "grad_norm": 1.1527509689331055,
+      "learning_rate": 4.307283982234494e-05,
+      "loss": 0.3692,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5313973782910617,
+      "grad_norm": 0.9051923751831055,
+      "learning_rate": 4.306244869528209e-05,
+      "loss": 0.4169,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5317687251661777,
+      "grad_norm": 0.8554239869117737,
+      "learning_rate": 4.305205103586058e-05,
+      "loss": 0.3821,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5321400720412938,
+      "grad_norm": 1.063449501991272,
+      "learning_rate": 4.304164684784076e-05,
+      "loss": 0.3977,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5325114189164099,
+      "grad_norm": 0.9604543447494507,
+      "learning_rate": 4.303123613498534e-05,
+      "loss": 0.3864,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5328827657915258,
+      "grad_norm": 1.005906581878662,
+      "learning_rate": 4.30208189010594e-05,
+      "loss": 0.3756,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5332541126666419,
+      "grad_norm": 0.8495503067970276,
+      "learning_rate": 4.301039514983038e-05,
+      "loss": 0.3763,
+      "step": 7180
+    },
+    {
+      "epoch": 0.533625459541758,
+      "grad_norm": 0.8146620988845825,
+      "learning_rate": 4.2999964885068064e-05,
+      "loss": 0.3729,
+      "step": 7185
+    },
+    {
+      "epoch": 0.533996806416874,
+      "grad_norm": 0.6939537525177002,
+      "learning_rate": 4.298952811054462e-05,
+      "loss": 0.38,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5343681532919901,
+      "grad_norm": 0.955191433429718,
+      "learning_rate": 4.297908483003452e-05,
+      "loss": 0.3648,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5347395001671061,
+      "grad_norm": 0.9314118027687073,
+      "learning_rate": 4.296863504731464e-05,
+      "loss": 0.3953,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5351108470422221,
+      "grad_norm": 1.5681297779083252,
+      "learning_rate": 4.295817876616419e-05,
+      "loss": 0.4045,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5354821939173382,
+      "grad_norm": 0.8491237163543701,
+      "learning_rate": 4.2947715990364715e-05,
+      "loss": 0.4016,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5358535407924543,
+      "grad_norm": 0.940566897392273,
+      "learning_rate": 4.293724672370012e-05,
+      "loss": 0.3824,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5362248876675703,
+      "grad_norm": 1.5019309520721436,
+      "learning_rate": 4.292677096995667e-05,
+      "loss": 0.3823,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5365962345426863,
+      "grad_norm": 0.7174615263938904,
+      "learning_rate": 4.291628873292294e-05,
+      "loss": 0.3854,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5369675814178023,
+      "grad_norm": 0.901176929473877,
+      "learning_rate": 4.290580001638991e-05,
+      "loss": 0.3872,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5373389282929184,
+      "grad_norm": 0.8742217421531677,
+      "learning_rate": 4.289530482415084e-05,
+      "loss": 0.3791,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5377102751680345,
+      "grad_norm": 0.939732551574707,
+      "learning_rate": 4.288480316000137e-05,
+      "loss": 0.3676,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5380816220431505,
+      "grad_norm": 0.8793368339538574,
+      "learning_rate": 4.2874295027739465e-05,
+      "loss": 0.3813,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5384529689182666,
+      "grad_norm": 1.093706488609314,
+      "learning_rate": 4.2863780431165445e-05,
+      "loss": 0.3734,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5388243157933826,
+      "grad_norm": 0.855695903301239,
+      "learning_rate": 4.285325937408194e-05,
+      "loss": 0.3642,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5391956626684986,
+      "grad_norm": 0.9116891026496887,
+      "learning_rate": 4.2842731860293946e-05,
+      "loss": 0.3662,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5395670095436147,
+      "grad_norm": 0.9691513776779175,
+      "learning_rate": 4.283219789360878e-05,
+      "loss": 0.3793,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5399383564187308,
+      "grad_norm": 0.7580130696296692,
+      "learning_rate": 4.2821657477836086e-05,
+      "loss": 0.3971,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5403097032938468,
+      "grad_norm": 0.8524672389030457,
+      "learning_rate": 4.2811110616787866e-05,
+      "loss": 0.3985,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5406810501689628,
+      "grad_norm": 0.7902607917785645,
+      "learning_rate": 4.280055731427841e-05,
+      "loss": 0.3844,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5410523970440789,
+      "grad_norm": 0.7604623436927795,
+      "learning_rate": 4.278999757412439e-05,
+      "loss": 0.3699,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5414237439191949,
+      "grad_norm": 0.951298713684082,
+      "learning_rate": 4.2779431400144765e-05,
+      "loss": 0.4153,
+      "step": 7290
+    },
+    {
+      "epoch": 0.541795090794311,
+      "grad_norm": 1.0854779481887817,
+      "learning_rate": 4.2768858796160834e-05,
+      "loss": 0.4008,
+      "step": 7295
+    },
+    {
+      "epoch": 0.542166437669427,
+      "grad_norm": 0.7753530740737915,
+      "learning_rate": 4.2758279765996224e-05,
+      "loss": 0.3824,
+      "step": 7300
+    },
+    {
+      "epoch": 0.542537784544543,
+      "grad_norm": 0.8362932801246643,
+      "learning_rate": 4.2747694313476886e-05,
+      "loss": 0.3856,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5429091314196591,
+      "grad_norm": 1.1193138360977173,
+      "learning_rate": 4.27371024424311e-05,
+      "loss": 0.3749,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5432804782947751,
+      "grad_norm": 0.9703423380851746,
+      "learning_rate": 4.272650415668944e-05,
+      "loss": 0.3666,
+      "step": 7315
+    },
+    {
+      "epoch": 0.5436518251698912,
+      "grad_norm": 0.6996259093284607,
+      "learning_rate": 4.2715899460084844e-05,
+      "loss": 0.3828,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5440231720450073,
+      "grad_norm": 1.0184882879257202,
+      "learning_rate": 4.2705288356452525e-05,
+      "loss": 0.376,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5443945189201232,
+      "grad_norm": 1.1657757759094238,
+      "learning_rate": 4.269467084963004e-05,
+      "loss": 0.3993,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5447658657952393,
+      "grad_norm": 0.9135143160820007,
+      "learning_rate": 4.268404694345726e-05,
+      "loss": 0.3784,
+      "step": 7335
+    },
+    {
+      "epoch": 0.5451372126703554,
+      "grad_norm": 0.8046631217002869,
+      "learning_rate": 4.2673416641776346e-05,
+      "loss": 0.3873,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5455085595454714,
+      "grad_norm": 1.1384739875793457,
+      "learning_rate": 4.266277994843181e-05,
+      "loss": 0.3937,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5458799064205875,
+      "grad_norm": 0.7930871248245239,
+      "learning_rate": 4.2652136867270444e-05,
+      "loss": 0.3867,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5462512532957036,
+      "grad_norm": 0.9256702661514282,
+      "learning_rate": 4.264148740214136e-05,
+      "loss": 0.3996,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5466226001708195,
+      "grad_norm": 1.0180842876434326,
+      "learning_rate": 4.2630831556896e-05,
+      "loss": 0.3955,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5469939470459356,
+      "grad_norm": 1.0560667514801025,
+      "learning_rate": 4.262016933538806e-05,
+      "loss": 0.3968,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5473652939210516,
+      "grad_norm": 0.8952111005783081,
+      "learning_rate": 4.260950074147361e-05,
+      "loss": 0.3786,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5477366407961677,
+      "grad_norm": 0.8378480672836304,
+      "learning_rate": 4.2598825779010975e-05,
+      "loss": 0.3595,
+      "step": 7375
+    },
+    {
+      "epoch": 0.5481079876712838,
+      "grad_norm": 0.9558596014976501,
+      "learning_rate": 4.25881444518608e-05,
+      "loss": 0.3788,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5484793345463997,
+      "grad_norm": 1.0218459367752075,
+      "learning_rate": 4.257745676388603e-05,
+      "loss": 0.3883,
+      "step": 7385
+    },
+    {
+      "epoch": 0.5488506814215158,
+      "grad_norm": 0.7036837339401245,
+      "learning_rate": 4.256676271895191e-05,
+      "loss": 0.378,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5492220282966319,
+      "grad_norm": 0.8147867321968079,
+      "learning_rate": 4.2556062320926e-05,
+      "loss": 0.3931,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5495933751717479,
+      "grad_norm": 0.9215070009231567,
+      "learning_rate": 4.254535557367811e-05,
+      "loss": 0.3993,
+      "step": 7400
+    },
+    {
+      "epoch": 0.549964722046864,
+      "grad_norm": 0.9754782319068909,
+      "learning_rate": 4.2534642481080425e-05,
+      "loss": 0.3986,
+      "step": 7405
+    },
+    {
+      "epoch": 0.5503360689219801,
+      "grad_norm": 0.9843300580978394,
+      "learning_rate": 4.2523923047007344e-05,
+      "loss": 0.3794,
+      "step": 7410
+    },
+    {
+      "epoch": 0.550707415797096,
+      "grad_norm": 0.6864088773727417,
+      "learning_rate": 4.25131972753356e-05,
+      "loss": 0.3876,
+      "step": 7415
+    },
+    {
+      "epoch": 0.5510787626722121,
+      "grad_norm": 0.8335707187652588,
+      "learning_rate": 4.250246516994422e-05,
+      "loss": 0.3679,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5514501095473282,
+      "grad_norm": 0.8459278345108032,
+      "learning_rate": 4.2491726734714504e-05,
+      "loss": 0.3871,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5518214564224442,
+      "grad_norm": 0.9613873362541199,
+      "learning_rate": 4.2480981973530064e-05,
+      "loss": 0.3964,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5521928032975603,
+      "grad_norm": 0.766555666923523,
+      "learning_rate": 4.247023089027678e-05,
+      "loss": 0.3756,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5525641501726763,
+      "grad_norm": 0.7106664180755615,
+      "learning_rate": 4.245947348884282e-05,
+      "loss": 0.3887,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5529354970477923,
+      "grad_norm": 0.6389920711517334,
+      "learning_rate": 4.2448709773118654e-05,
+      "loss": 0.3658,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5533068439229084,
+      "grad_norm": 1.4300941228866577,
+      "learning_rate": 4.2437939746997024e-05,
+      "loss": 0.3866,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5536781907980244,
+      "grad_norm": 0.7473590970039368,
+      "learning_rate": 4.242716341437294e-05,
+      "loss": 0.3692,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5540495376731405,
+      "grad_norm": 0.8290305137634277,
+      "learning_rate": 4.241638077914372e-05,
+      "loss": 0.3846,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5544208845482566,
+      "grad_norm": 1.0392992496490479,
+      "learning_rate": 4.2405591845208934e-05,
+      "loss": 0.3814,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5547922314233725,
+      "grad_norm": 0.7894637584686279,
+      "learning_rate": 4.239479661647046e-05,
+      "loss": 0.3628,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5551635782984886,
+      "grad_norm": 0.9801110029220581,
+      "learning_rate": 4.238399509683243e-05,
+      "loss": 0.3801,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5555349251736047,
+      "grad_norm": 0.751207172870636,
+      "learning_rate": 4.237318729020127e-05,
+      "loss": 0.3844,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5559062720487207,
+      "grad_norm": 1.1230332851409912,
+      "learning_rate": 4.236237320048564e-05,
+      "loss": 0.4014,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5562776189238368,
+      "grad_norm": 1.1641552448272705,
+      "learning_rate": 4.235155283159652e-05,
+      "loss": 0.3794,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5566489657989528,
+      "grad_norm": 1.1932958364486694,
+      "learning_rate": 4.234072618744714e-05,
+      "loss": 0.3856,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5570203126740688,
+      "grad_norm": 1.112725019454956,
+      "learning_rate": 4.232989327195299e-05,
+      "loss": 0.3993,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5573916595491849,
+      "grad_norm": 1.0324194431304932,
+      "learning_rate": 4.231905408903185e-05,
+      "loss": 0.3835,
+      "step": 7505
+    },
+    {
+      "epoch": 0.557763006424301,
+      "grad_norm": 0.749826192855835,
+      "learning_rate": 4.230820864260375e-05,
+      "loss": 0.3364,
+      "step": 7510
+    },
+    {
+      "epoch": 0.558134353299417,
+      "grad_norm": 1.167183756828308,
+      "learning_rate": 4.229735693659099e-05,
+      "loss": 0.3906,
+      "step": 7515
+    },
+    {
+      "epoch": 0.558505700174533,
+      "grad_norm": 0.7182947993278503,
+      "learning_rate": 4.228649897491814e-05,
+      "loss": 0.3719,
+      "step": 7520
+    },
+    {
+      "epoch": 0.558877047049649,
+      "grad_norm": 1.1357501745224,
+      "learning_rate": 4.2275634761512014e-05,
+      "loss": 0.3986,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5592483939247651,
+      "grad_norm": 0.790942370891571,
+      "learning_rate": 4.2264764300301705e-05,
+      "loss": 0.3771,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5596197407998812,
+      "grad_norm": 0.7060715556144714,
+      "learning_rate": 4.2253887595218566e-05,
+      "loss": 0.3708,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5599910876749972,
+      "grad_norm": 0.9133544564247131,
+      "learning_rate": 4.22430046501962e-05,
+      "loss": 0.3865,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5603624345501133,
+      "grad_norm": 0.8340880870819092,
+      "learning_rate": 4.2232115469170465e-05,
+      "loss": 0.3893,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5607337814252293,
+      "grad_norm": 0.8766905069351196,
+      "learning_rate": 4.2221220056079466e-05,
+      "loss": 0.3837,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5611051283003453,
+      "grad_norm": 0.7138217091560364,
+      "learning_rate": 4.22103184148636e-05,
+      "loss": 0.3927,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5614764751754614,
+      "grad_norm": 1.0313303470611572,
+      "learning_rate": 4.219941054946547e-05,
+      "loss": 0.3939,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5618478220505775,
+      "grad_norm": 0.6473879218101501,
+      "learning_rate": 4.218849646382996e-05,
+      "loss": 0.3788,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5622191689256935,
+      "grad_norm": 0.748626172542572,
+      "learning_rate": 4.217757616190419e-05,
+      "loss": 0.3893,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5625905158008095,
+      "grad_norm": 0.7278555035591125,
+      "learning_rate": 4.2166649647637534e-05,
+      "loss": 0.3706,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5629618626759256,
+      "grad_norm": 0.8890153169631958,
+      "learning_rate": 4.215571692498161e-05,
+      "loss": 0.3737,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5633332095510416,
+      "grad_norm": 0.92683345079422,
+      "learning_rate": 4.214477799789028e-05,
+      "loss": 0.3809,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5637045564261577,
+      "grad_norm": 0.8576033711433411,
+      "learning_rate": 4.213383287031966e-05,
+      "loss": 0.3891,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5640759033012738,
+      "grad_norm": 0.9028112888336182,
+      "learning_rate": 4.2122881546228086e-05,
+      "loss": 0.3653,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5644472501763897,
+      "grad_norm": 0.6952714920043945,
+      "learning_rate": 4.211192402957616e-05,
+      "loss": 0.3993,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5648185970515058,
+      "grad_norm": 0.8100974559783936,
+      "learning_rate": 4.210096032432671e-05,
+      "loss": 0.3873,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5651899439266218,
+      "grad_norm": 1.0961740016937256,
+      "learning_rate": 4.20899904344448e-05,
+      "loss": 0.3911,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5655612908017379,
+      "grad_norm": 0.9107250571250916,
+      "learning_rate": 4.207901436389775e-05,
+      "loss": 0.3912,
+      "step": 7615
+    },
+    {
+      "epoch": 0.565932637676854,
+      "grad_norm": 0.7377946376800537,
+      "learning_rate": 4.206803211665509e-05,
+      "loss": 0.3896,
+      "step": 7620
+    },
+    {
+      "epoch": 0.56630398455197,
+      "grad_norm": 0.9254267811775208,
+      "learning_rate": 4.205704369668859e-05,
+      "loss": 0.3793,
+      "step": 7625
+    },
+    {
+      "epoch": 0.566675331427086,
+      "grad_norm": 0.8499208092689514,
+      "learning_rate": 4.204604910797228e-05,
+      "loss": 0.3603,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5670466783022021,
+      "grad_norm": 0.9021883606910706,
+      "learning_rate": 4.203504835448235e-05,
+      "loss": 0.3699,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5674180251773181,
+      "grad_norm": 1.3291375637054443,
+      "learning_rate": 4.2024041440197326e-05,
+      "loss": 0.4044,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5677893720524342,
+      "grad_norm": 0.8598208427429199,
+      "learning_rate": 4.201302836909787e-05,
+      "loss": 0.3752,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5681607189275503,
+      "grad_norm": 1.3929486274719238,
+      "learning_rate": 4.200200914516691e-05,
+      "loss": 0.4129,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5685320658026662,
+      "grad_norm": 0.9411462545394897,
+      "learning_rate": 4.199098377238959e-05,
+      "loss": 0.3784,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5689034126777823,
+      "grad_norm": 0.9103113412857056,
+      "learning_rate": 4.1979952254753294e-05,
+      "loss": 0.3861,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5692747595528984,
+      "grad_norm": 0.8931922316551208,
+      "learning_rate": 4.19689145962476e-05,
+      "loss": 0.3887,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5696461064280144,
+      "grad_norm": 0.7689836621284485,
+      "learning_rate": 4.1957870800864344e-05,
+      "loss": 0.3652,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5700174533031305,
+      "grad_norm": 0.8733877539634705,
+      "learning_rate": 4.194682087259754e-05,
+      "loss": 0.3929,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5703888001782464,
+      "grad_norm": 1.7723708152770996,
+      "learning_rate": 4.193576481544344e-05,
+      "loss": 0.3832,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5707601470533625,
+      "grad_norm": 1.078263759613037,
+      "learning_rate": 4.192470263340051e-05,
+      "loss": 0.4004,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5711314939284786,
+      "grad_norm": 1.061034083366394,
+      "learning_rate": 4.191363433046946e-05,
+      "loss": 0.385,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5715028408035946,
+      "grad_norm": 0.9127119183540344,
+      "learning_rate": 4.190255991065316e-05,
+      "loss": 0.369,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5718741876787107,
+      "grad_norm": 1.017187237739563,
+      "learning_rate": 4.189147937795673e-05,
+      "loss": 0.3792,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5722455345538268,
+      "grad_norm": 0.9224918484687805,
+      "learning_rate": 4.188039273638749e-05,
+      "loss": 0.3978,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5726168814289427,
+      "grad_norm": 1.0088601112365723,
+      "learning_rate": 4.186929998995497e-05,
+      "loss": 0.3739,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5729882283040588,
+      "grad_norm": 1.0696107149124146,
+      "learning_rate": 4.1858201142670904e-05,
+      "loss": 0.3929,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5733595751791749,
+      "grad_norm": 0.8687728047370911,
+      "learning_rate": 4.1847096198549236e-05,
+      "loss": 0.3683,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5737309220542909,
+      "grad_norm": 0.7542261481285095,
+      "learning_rate": 4.183598516160612e-05,
+      "loss": 0.3854,
+      "step": 7725
+    },
+    {
+      "epoch": 0.574102268929407,
+      "grad_norm": 0.8737427592277527,
+      "learning_rate": 4.18248680358599e-05,
+      "loss": 0.3918,
+      "step": 7730
+    },
+    {
+      "epoch": 0.574473615804523,
+      "grad_norm": 0.7520382404327393,
+      "learning_rate": 4.1813744825331126e-05,
+      "loss": 0.3836,
+      "step": 7735
+    },
+    {
+      "epoch": 0.574844962679639,
+      "grad_norm": 1.4893221855163574,
+      "learning_rate": 4.1802615534042563e-05,
+      "loss": 0.3679,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5752163095547551,
+      "grad_norm": 1.0394209623336792,
+      "learning_rate": 4.179148016601917e-05,
+      "loss": 0.3835,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5755876564298712,
+      "grad_norm": 0.9864911437034607,
+      "learning_rate": 4.178033872528809e-05,
+      "loss": 0.3737,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5759590033049872,
+      "grad_norm": 0.7848846316337585,
+      "learning_rate": 4.176919121587867e-05,
+      "loss": 0.3839,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5763303501801033,
+      "grad_norm": 0.6822853088378906,
+      "learning_rate": 4.175803764182245e-05,
+      "loss": 0.3695,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5767016970552192,
+      "grad_norm": 1.83903169631958,
+      "learning_rate": 4.1746878007153164e-05,
+      "loss": 0.3493,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5770730439303353,
+      "grad_norm": 0.9079355001449585,
+      "learning_rate": 4.173571231590676e-05,
+      "loss": 0.3889,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5774443908054514,
+      "grad_norm": 0.8775737285614014,
+      "learning_rate": 4.172454057212134e-05,
+      "loss": 0.3736,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5778157376805674,
+      "grad_norm": 0.8449175953865051,
+      "learning_rate": 4.171336277983721e-05,
+      "loss": 0.3912,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5781870845556835,
+      "grad_norm": 0.9498149752616882,
+      "learning_rate": 4.170217894309686e-05,
+      "loss": 0.3749,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5785584314307995,
+      "grad_norm": 0.8782013058662415,
+      "learning_rate": 4.1690989065944985e-05,
+      "loss": 0.3888,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5789297783059155,
+      "grad_norm": 0.9666025042533875,
+      "learning_rate": 4.167979315242844e-05,
+      "loss": 0.3881,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5793011251810316,
+      "grad_norm": 1.0839999914169312,
+      "learning_rate": 4.1668591206596274e-05,
+      "loss": 0.3891,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5796724720561477,
+      "grad_norm": 0.8281628489494324,
+      "learning_rate": 4.165738323249972e-05,
+      "loss": 0.3787,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5800438189312637,
+      "grad_norm": 0.7159334421157837,
+      "learning_rate": 4.1646169234192185e-05,
+      "loss": 0.4055,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5804151658063798,
+      "grad_norm": 0.9558557868003845,
+      "learning_rate": 4.163494921572926e-05,
+      "loss": 0.398,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5807865126814958,
+      "grad_norm": 0.7024179100990295,
+      "learning_rate": 4.162372318116871e-05,
+      "loss": 0.3903,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5811578595566118,
+      "grad_norm": 0.9543454647064209,
+      "learning_rate": 4.161249113457047e-05,
+      "loss": 0.3796,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5815292064317279,
+      "grad_norm": 0.8931434750556946,
+      "learning_rate": 4.160125307999667e-05,
+      "loss": 0.3798,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5819005533068439,
+      "grad_norm": 0.9350513219833374,
+      "learning_rate": 4.1590009021511585e-05,
+      "loss": 0.3922,
+      "step": 7835
+    },
+    {
+      "epoch": 0.58227190018196,
+      "grad_norm": 1.3297828435897827,
+      "learning_rate": 4.157875896318168e-05,
+      "loss": 0.384,
+      "step": 7840
+    },
+    {
+      "epoch": 0.582643247057076,
+      "grad_norm": 0.8016966581344604,
+      "learning_rate": 4.156750290907559e-05,
+      "loss": 0.3897,
+      "step": 7845
+    },
+    {
+      "epoch": 0.583014593932192,
+      "grad_norm": 0.7308679223060608,
+      "learning_rate": 4.155624086326409e-05,
+      "loss": 0.384,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5833859408073081,
+      "grad_norm": 0.9664947390556335,
+      "learning_rate": 4.1544972829820173e-05,
+      "loss": 0.3586,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5837572876824242,
+      "grad_norm": 0.7831148505210876,
+      "learning_rate": 4.153369881281897e-05,
+      "loss": 0.3751,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5841286345575402,
+      "grad_norm": 0.8598526120185852,
+      "learning_rate": 4.152241881633775e-05,
+      "loss": 0.3967,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5844999814326562,
+      "grad_norm": 0.9544880986213684,
+      "learning_rate": 4.151113284445599e-05,
+      "loss": 0.3772,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5848713283077723,
+      "grad_norm": 0.9176838397979736,
+      "learning_rate": 4.14998409012553e-05,
+      "loss": 0.3836,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5852426751828883,
+      "grad_norm": 0.7144293785095215,
+      "learning_rate": 4.148854299081947e-05,
+      "loss": 0.3762,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5856140220580044,
+      "grad_norm": 0.9049531817436218,
+      "learning_rate": 4.147723911723441e-05,
+      "loss": 0.3935,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5859853689331205,
+      "grad_norm": 1.0131149291992188,
+      "learning_rate": 4.146592928458824e-05,
+      "loss": 0.3899,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5863567158082365,
+      "grad_norm": 0.9420373439788818,
+      "learning_rate": 4.1454613496971195e-05,
+      "loss": 0.3769,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5867280626833525,
+      "grad_norm": 1.0125765800476074,
+      "learning_rate": 4.144329175847567e-05,
+      "loss": 0.3655,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5870994095584685,
+      "grad_norm": 0.810465395450592,
+      "learning_rate": 4.143196407319623e-05,
+      "loss": 0.3802,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5874707564335846,
+      "grad_norm": 0.9369770884513855,
+      "learning_rate": 4.142063044522958e-05,
+      "loss": 0.3688,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5878421033087007,
+      "grad_norm": 0.9551800489425659,
+      "learning_rate": 4.140929087867456e-05,
+      "loss": 0.3758,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5882134501838167,
+      "grad_norm": 1.0710747241973877,
+      "learning_rate": 4.139794537763217e-05,
+      "loss": 0.3665,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5885847970589327,
+      "grad_norm": 1.1802254915237427,
+      "learning_rate": 4.138659394620558e-05,
+      "loss": 0.3828,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5889561439340488,
+      "grad_norm": 0.9443197846412659,
+      "learning_rate": 4.137523658850005e-05,
+      "loss": 0.3853,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5893274908091648,
+      "grad_norm": 0.7773070931434631,
+      "learning_rate": 4.136387330862303e-05,
+      "loss": 0.3952,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5896988376842809,
+      "grad_norm": 0.8189634084701538,
+      "learning_rate": 4.135250411068411e-05,
+      "loss": 0.3939,
+      "step": 7940
+    },
+    {
+      "epoch": 0.590070184559397,
+      "grad_norm": 0.9138241410255432,
+      "learning_rate": 4.1341128998795e-05,
+      "loss": 0.4005,
+      "step": 7945
+    },
+    {
+      "epoch": 0.590441531434513,
+      "grad_norm": 0.9136779308319092,
+      "learning_rate": 4.132974797706954e-05,
+      "loss": 0.3808,
+      "step": 7950
+    },
+    {
+      "epoch": 0.590812878309629,
+      "grad_norm": 1.363004446029663,
+      "learning_rate": 4.131836104962373e-05,
+      "loss": 0.3793,
+      "step": 7955
+    },
+    {
+      "epoch": 0.5911842251847451,
+      "grad_norm": 0.699315071105957,
+      "learning_rate": 4.1306968220575716e-05,
+      "loss": 0.3825,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5915555720598611,
+      "grad_norm": 0.9782199859619141,
+      "learning_rate": 4.1295569494045736e-05,
+      "loss": 0.3861,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5919269189349772,
+      "grad_norm": 0.991134524345398,
+      "learning_rate": 4.128416487415621e-05,
+      "loss": 0.374,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5922982658100933,
+      "grad_norm": 0.8225498199462891,
+      "learning_rate": 4.127275436503165e-05,
+      "loss": 0.3808,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5926696126852092,
+      "grad_norm": 1.1791242361068726,
+      "learning_rate": 4.126133797079872e-05,
+      "loss": 0.3901,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5930409595603253,
+      "grad_norm": 0.8043322563171387,
+      "learning_rate": 4.1249915695586195e-05,
+      "loss": 0.3776,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5934123064354413,
+      "grad_norm": 0.9959288835525513,
+      "learning_rate": 4.123848754352501e-05,
+      "loss": 0.3846,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5937836533105574,
+      "grad_norm": 1.1933143138885498,
+      "learning_rate": 4.122705351874818e-05,
+      "loss": 0.3712,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5941550001856735,
+      "grad_norm": 1.0833144187927246,
+      "learning_rate": 4.121561362539087e-05,
+      "loss": 0.397,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5945263470607894,
+      "grad_norm": 0.8315593600273132,
+      "learning_rate": 4.120416786759038e-05,
+      "loss": 0.3924,
+      "step": 8005
+    },
+    {
+      "epoch": 0.5948976939359055,
+      "grad_norm": 0.6777327656745911,
+      "learning_rate": 4.119271624948611e-05,
+      "loss": 0.3754,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5952690408110216,
+      "grad_norm": 0.9331532120704651,
+      "learning_rate": 4.1181258775219566e-05,
+      "loss": 0.3777,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5956403876861376,
+      "grad_norm": 0.8980804085731506,
+      "learning_rate": 4.1169795448934414e-05,
+      "loss": 0.4035,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5960117345612537,
+      "grad_norm": 0.9153735041618347,
+      "learning_rate": 4.1158326274776404e-05,
+      "loss": 0.3697,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5963830814363698,
+      "grad_norm": 0.8011136651039124,
+      "learning_rate": 4.114685125689341e-05,
+      "loss": 0.3739,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5967544283114857,
+      "grad_norm": 0.7510383725166321,
+      "learning_rate": 4.113537039943542e-05,
+      "loss": 0.3825,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5971257751866018,
+      "grad_norm": 0.9823917746543884,
+      "learning_rate": 4.112388370655453e-05,
+      "loss": 0.3692,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5974971220617179,
+      "grad_norm": 0.7274696826934814,
+      "learning_rate": 4.111239118240495e-05,
+      "loss": 0.3599,
+      "step": 8045
+    },
+    {
+      "epoch": 0.5978684689368339,
+      "grad_norm": 0.8090460300445557,
+      "learning_rate": 4.110089283114301e-05,
+      "loss": 0.3787,
+      "step": 8050
+    },
+    {
+      "epoch": 0.59823981581195,
+      "grad_norm": 0.7262414693832397,
+      "learning_rate": 4.1089388656927116e-05,
+      "loss": 0.3765,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5986111626870659,
+      "grad_norm": 0.8816661834716797,
+      "learning_rate": 4.107787866391782e-05,
+      "loss": 0.3741,
+      "step": 8060
+    },
+    {
+      "epoch": 0.598982509562182,
+      "grad_norm": 0.8424727916717529,
+      "learning_rate": 4.106636285627774e-05,
+      "loss": 0.3892,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5993538564372981,
+      "grad_norm": 0.7614096999168396,
+      "learning_rate": 4.105484123817163e-05,
+      "loss": 0.4107,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5997252033124141,
+      "grad_norm": 0.758870005607605,
+      "learning_rate": 4.1043313813766314e-05,
+      "loss": 0.3829,
+      "step": 8075
+    },
+    {
+      "epoch": 0.6000965501875302,
+      "grad_norm": 0.8257849812507629,
+      "learning_rate": 4.103178058723075e-05,
+      "loss": 0.3938,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6004678970626462,
+      "grad_norm": 2.6828434467315674,
+      "learning_rate": 4.102024156273597e-05,
+      "loss": 0.4119,
+      "step": 8085
+    },
+    {
+      "epoch": 0.6008392439377622,
+      "grad_norm": 0.8033463954925537,
+      "learning_rate": 4.100869674445511e-05,
+      "loss": 0.3991,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6012105908128783,
+      "grad_norm": 0.9146053791046143,
+      "learning_rate": 4.099714613656339e-05,
+      "loss": 0.3687,
+      "step": 8095
+    },
+    {
+      "epoch": 0.6015819376879944,
+      "grad_norm": 0.7598369121551514,
+      "learning_rate": 4.098558974323816e-05,
+      "loss": 0.3657,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6019532845631104,
+      "grad_norm": 1.4062983989715576,
+      "learning_rate": 4.09740275686588e-05,
+      "loss": 0.4008,
+      "step": 8105
+    },
+    {
+      "epoch": 0.6023246314382265,
+      "grad_norm": 0.8199751973152161,
+      "learning_rate": 4.096245961700685e-05,
+      "loss": 0.3914,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6026959783133425,
+      "grad_norm": 0.7408627271652222,
+      "learning_rate": 4.0950885892465904e-05,
+      "loss": 0.3501,
+      "step": 8115
+    },
+    {
+      "epoch": 0.6030673251884585,
+      "grad_norm": 0.9060487151145935,
+      "learning_rate": 4.093930639922163e-05,
+      "loss": 0.3818,
+      "step": 8120
+    },
+    {
+      "epoch": 0.6034386720635746,
+      "grad_norm": 0.8455791473388672,
+      "learning_rate": 4.0927721141461816e-05,
+      "loss": 0.3913,
+      "step": 8125
+    },
+    {
+      "epoch": 0.6038100189386907,
+      "grad_norm": 0.8929671049118042,
+      "learning_rate": 4.091613012337629e-05,
+      "loss": 0.367,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6041813658138067,
+      "grad_norm": 0.8208998441696167,
+      "learning_rate": 4.0904533349157035e-05,
+      "loss": 0.3663,
+      "step": 8135
+    },
+    {
+      "epoch": 0.6045527126889227,
+      "grad_norm": 0.7038164138793945,
+      "learning_rate": 4.089293082299805e-05,
+      "loss": 0.3583,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6049240595640387,
+      "grad_norm": 0.8460705876350403,
+      "learning_rate": 4.088132254909542e-05,
+      "loss": 0.3791,
+      "step": 8145
+    },
+    {
+      "epoch": 0.6052954064391548,
+      "grad_norm": 0.8331727981567383,
+      "learning_rate": 4.0869708531647346e-05,
+      "loss": 0.3842,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6056667533142709,
+      "grad_norm": 0.7929322123527527,
+      "learning_rate": 4.085808877485407e-05,
+      "loss": 0.3791,
+      "step": 8155
+    },
+    {
+      "epoch": 0.6060381001893869,
+      "grad_norm": 0.850964367389679,
+      "learning_rate": 4.0846463282917944e-05,
+      "loss": 0.3855,
+      "step": 8160
+    },
+    {
+      "epoch": 0.606409447064503,
+      "grad_norm": 0.7293786406517029,
+      "learning_rate": 4.083483206004336e-05,
+      "loss": 0.3748,
+      "step": 8165
+    },
+    {
+      "epoch": 0.606780793939619,
+      "grad_norm": 0.9626237750053406,
+      "learning_rate": 4.0823195110436805e-05,
+      "loss": 0.367,
+      "step": 8170
+    },
+    {
+      "epoch": 0.607152140814735,
+      "grad_norm": 0.9357216954231262,
+      "learning_rate": 4.0811552438306824e-05,
+      "loss": 0.387,
+      "step": 8175
+    },
+    {
+      "epoch": 0.6075234876898511,
+      "grad_norm": 0.8385303616523743,
+      "learning_rate": 4.079990404786404e-05,
+      "loss": 0.3803,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6078948345649672,
+      "grad_norm": 0.8900553584098816,
+      "learning_rate": 4.078824994332114e-05,
+      "loss": 0.4066,
+      "step": 8185
+    },
+    {
+      "epoch": 0.6082661814400832,
+      "grad_norm": 0.906819760799408,
+      "learning_rate": 4.0776590128892886e-05,
+      "loss": 0.3848,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6086375283151992,
+      "grad_norm": 0.7580803632736206,
+      "learning_rate": 4.076492460879609e-05,
+      "loss": 0.3857,
+      "step": 8195
+    },
+    {
+      "epoch": 0.6090088751903153,
+      "grad_norm": 0.9620898365974426,
+      "learning_rate": 4.0753253387249626e-05,
+      "loss": 0.3959,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6093802220654313,
+      "grad_norm": 0.7553649544715881,
+      "learning_rate": 4.074157646847446e-05,
+      "loss": 0.3752,
+      "step": 8205
+    },
+    {
+      "epoch": 0.6097515689405474,
+      "grad_norm": 1.101928472518921,
+      "learning_rate": 4.072989385669358e-05,
+      "loss": 0.3945,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6101229158156634,
+      "grad_norm": 0.6902221441268921,
+      "learning_rate": 4.071820555613206e-05,
+      "loss": 0.3756,
+      "step": 8215
+    },
+    {
+      "epoch": 0.6104942626907794,
+      "grad_norm": 0.7055659294128418,
+      "learning_rate": 4.070651157101701e-05,
+      "loss": 0.3882,
+      "step": 8220
+    },
+    {
+      "epoch": 0.6108656095658955,
+      "grad_norm": 0.929861843585968,
+      "learning_rate": 4.0694811905577634e-05,
+      "loss": 0.4055,
+      "step": 8225
+    },
+    {
+      "epoch": 0.6112369564410115,
+      "grad_norm": 1.9215669631958008,
+      "learning_rate": 4.0683106564045135e-05,
+      "loss": 0.3598,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6116083033161276,
+      "grad_norm": 0.7936436533927917,
+      "learning_rate": 4.067139555065281e-05,
+      "loss": 0.3895,
+      "step": 8235
+    },
+    {
+      "epoch": 0.6119796501912437,
+      "grad_norm": 0.8062562346458435,
+      "learning_rate": 4.065967886963601e-05,
+      "loss": 0.3681,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6123509970663596,
+      "grad_norm": 0.983604371547699,
+      "learning_rate": 4.0647956525232094e-05,
+      "loss": 0.3759,
+      "step": 8245
+    },
+    {
+      "epoch": 0.6127223439414757,
+      "grad_norm": 1.1448338031768799,
+      "learning_rate": 4.063622852168051e-05,
+      "loss": 0.3793,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6130936908165918,
+      "grad_norm": 0.9411412477493286,
+      "learning_rate": 4.0624494863222736e-05,
+      "loss": 0.3812,
+      "step": 8255
+    },
+    {
+      "epoch": 0.6134650376917078,
+      "grad_norm": 0.7395769953727722,
+      "learning_rate": 4.061275555410229e-05,
+      "loss": 0.3808,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6138363845668239,
+      "grad_norm": 0.8813555240631104,
+      "learning_rate": 4.060101059856476e-05,
+      "loss": 0.3736,
+      "step": 8265
+    },
+    {
+      "epoch": 0.61420773144194,
+      "grad_norm": 0.7604705095291138,
+      "learning_rate": 4.0589260000857745e-05,
+      "loss": 0.3914,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6145790783170559,
+      "grad_norm": 1.0517133474349976,
+      "learning_rate": 4.0577503765230893e-05,
+      "loss": 0.3586,
+      "step": 8275
+    },
+    {
+      "epoch": 0.614950425192172,
+      "grad_norm": 0.8700730204582214,
+      "learning_rate": 4.05657418959359e-05,
+      "loss": 0.383,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6153217720672881,
+      "grad_norm": 0.8465512990951538,
+      "learning_rate": 4.05539743972265e-05,
+      "loss": 0.3552,
+      "step": 8285
+    },
+    {
+      "epoch": 0.6156931189424041,
+      "grad_norm": 1.0167089700698853,
+      "learning_rate": 4.0542201273358446e-05,
+      "loss": 0.4038,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6160644658175202,
+      "grad_norm": 0.9293932914733887,
+      "learning_rate": 4.053042252858955e-05,
+      "loss": 0.3929,
+      "step": 8295
+    },
+    {
+      "epoch": 0.6164358126926361,
+      "grad_norm": 0.8165609240531921,
+      "learning_rate": 4.051863816717964e-05,
+      "loss": 0.3744,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6168071595677522,
+      "grad_norm": 0.885839581489563,
+      "learning_rate": 4.050684819339056e-05,
+      "loss": 0.3824,
+      "step": 8305
+    },
+    {
+      "epoch": 0.6171785064428683,
+      "grad_norm": 1.0093748569488525,
+      "learning_rate": 4.049505261148622e-05,
+      "loss": 0.3811,
+      "step": 8310
+    },
+    {
+      "epoch": 0.6175498533179843,
+      "grad_norm": 0.9099218845367432,
+      "learning_rate": 4.048325142573254e-05,
+      "loss": 0.3781,
+      "step": 8315
+    },
+    {
+      "epoch": 0.6179212001931004,
+      "grad_norm": 0.7306044101715088,
+      "learning_rate": 4.047144464039746e-05,
+      "loss": 0.3505,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6182925470682165,
+      "grad_norm": 0.8570384383201599,
+      "learning_rate": 4.045963225975097e-05,
+      "loss": 0.3927,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6186638939433324,
+      "grad_norm": 0.8275445699691772,
+      "learning_rate": 4.0447814288065044e-05,
+      "loss": 0.3814,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6190352408184485,
+      "grad_norm": 0.906373918056488,
+      "learning_rate": 4.0435990729613706e-05,
+      "loss": 0.365,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6194065876935646,
+      "grad_norm": 0.8798776865005493,
+      "learning_rate": 4.0424161588673e-05,
+      "loss": 0.3808,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6197779345686806,
+      "grad_norm": 1.0505601167678833,
+      "learning_rate": 4.041232686952098e-05,
+      "loss": 0.4154,
+      "step": 8345
+    },
+    {
+      "epoch": 0.6201492814437967,
+      "grad_norm": 0.8700944185256958,
+      "learning_rate": 4.0400486576437724e-05,
+      "loss": 0.402,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6205206283189127,
+      "grad_norm": 0.9342040419578552,
+      "learning_rate": 4.038864071370532e-05,
+      "loss": 0.3918,
+      "step": 8355
+    },
+    {
+      "epoch": 0.6208919751940287,
+      "grad_norm": 0.7223671674728394,
+      "learning_rate": 4.037678928560786e-05,
+      "loss": 0.3933,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6212633220691448,
+      "grad_norm": 1.1166205406188965,
+      "learning_rate": 4.036493229643149e-05,
+      "loss": 0.408,
+      "step": 8365
+    },
+    {
+      "epoch": 0.6216346689442608,
+      "grad_norm": 1.0232312679290771,
+      "learning_rate": 4.0353069750464314e-05,
+      "loss": 0.3843,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6220060158193769,
+      "grad_norm": 0.9179908037185669,
+      "learning_rate": 4.0341201651996485e-05,
+      "loss": 0.4008,
+      "step": 8375
+    },
+    {
+      "epoch": 0.622377362694493,
+      "grad_norm": 0.725074291229248,
+      "learning_rate": 4.032932800532014e-05,
+      "loss": 0.3791,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6227487095696089,
+      "grad_norm": 0.7987901568412781,
+      "learning_rate": 4.031744881472944e-05,
+      "loss": 0.3634,
+      "step": 8385
+    },
+    {
+      "epoch": 0.623120056444725,
+      "grad_norm": 0.7891753315925598,
+      "learning_rate": 4.0305564084520536e-05,
+      "loss": 0.3629,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6234914033198411,
+      "grad_norm": 0.9482506513595581,
+      "learning_rate": 4.029367381899159e-05,
+      "loss": 0.3832,
+      "step": 8395
+    },
+    {
+      "epoch": 0.6238627501949571,
+      "grad_norm": 0.7741977572441101,
+      "learning_rate": 4.028177802244277e-05,
+      "loss": 0.3547,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6242340970700732,
+      "grad_norm": 0.6711797714233398,
+      "learning_rate": 4.0269876699176245e-05,
+      "loss": 0.394,
+      "step": 8405
+    },
+    {
+      "epoch": 0.6246054439451892,
+      "grad_norm": 1.1592880487442017,
+      "learning_rate": 4.0257969853496156e-05,
+      "loss": 0.407,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6249767908203052,
+      "grad_norm": 0.9707858562469482,
+      "learning_rate": 4.0246057489708686e-05,
+      "loss": 0.3682,
+      "step": 8415
+    },
+    {
+      "epoch": 0.6253481376954213,
+      "grad_norm": 0.9577823281288147,
+      "learning_rate": 4.023413961212198e-05,
+      "loss": 0.3788,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6257194845705374,
+      "grad_norm": 0.9387092590332031,
+      "learning_rate": 4.0222216225046184e-05,
+      "loss": 0.3768,
+      "step": 8425
+    },
+    {
+      "epoch": 0.6260908314456534,
+      "grad_norm": 1.0328980684280396,
+      "learning_rate": 4.021028733279344e-05,
+      "loss": 0.3651,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6264621783207694,
+      "grad_norm": 1.5044817924499512,
+      "learning_rate": 4.0198352939677894e-05,
+      "loss": 0.3812,
+      "step": 8435
+    },
+    {
+      "epoch": 0.6268335251958854,
+      "grad_norm": 0.967638373374939,
+      "learning_rate": 4.0186413050015657e-05,
+      "loss": 0.3899,
+      "step": 8440
+    },
+    {
+      "epoch": 0.6272048720710015,
+      "grad_norm": 0.8514472246170044,
+      "learning_rate": 4.017446766812484e-05,
+      "loss": 0.3891,
+      "step": 8445
+    },
+    {
+      "epoch": 0.6275762189461176,
+      "grad_norm": 0.7428901791572571,
+      "learning_rate": 4.0162516798325535e-05,
+      "loss": 0.3713,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6279475658212336,
+      "grad_norm": 0.8475235104560852,
+      "learning_rate": 4.0150560444939826e-05,
+      "loss": 0.3736,
+      "step": 8455
+    },
+    {
+      "epoch": 0.6283189126963497,
+      "grad_norm": 0.8051154017448425,
+      "learning_rate": 4.013859861229179e-05,
+      "loss": 0.3801,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6286902595714657,
+      "grad_norm": 0.7596623301506042,
+      "learning_rate": 4.012663130470746e-05,
+      "loss": 0.382,
+      "step": 8465
+    },
+    {
+      "epoch": 0.6290616064465817,
+      "grad_norm": 0.9572256803512573,
+      "learning_rate": 4.0114658526514856e-05,
+      "loss": 0.3771,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6294329533216978,
+      "grad_norm": 0.9228476285934448,
+      "learning_rate": 4.0102680282044e-05,
+      "loss": 0.3848,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6298043001968139,
+      "grad_norm": 0.8364796042442322,
+      "learning_rate": 4.009069657562686e-05,
+      "loss": 0.3729,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6301756470719299,
+      "grad_norm": 1.0168267488479614,
+      "learning_rate": 4.007870741159739e-05,
+      "loss": 0.3632,
+      "step": 8485
+    },
+    {
+      "epoch": 0.6305469939470459,
+      "grad_norm": 0.8414918780326843,
+      "learning_rate": 4.006671279429154e-05,
+      "loss": 0.3705,
+      "step": 8490
+    },
+    {
+      "epoch": 0.630918340822162,
+      "grad_norm": 0.9206081032752991,
+      "learning_rate": 4.0054712728047194e-05,
+      "loss": 0.3769,
+      "step": 8495
+    },
+    {
+      "epoch": 0.631289687697278,
+      "grad_norm": 0.93706214427948,
+      "learning_rate": 4.0042707217204226e-05,
+      "loss": 0.3867,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6316610345723941,
+      "grad_norm": 1.0076346397399902,
+      "learning_rate": 4.0030696266104486e-05,
+      "loss": 0.391,
+      "step": 8505
+    },
+    {
+      "epoch": 0.6320323814475102,
+      "grad_norm": 0.9548998475074768,
+      "learning_rate": 4.001867987909178e-05,
+      "loss": 0.3764,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6324037283226261,
+      "grad_norm": 1.0261924266815186,
+      "learning_rate": 4.000665806051188e-05,
+      "loss": 0.3743,
+      "step": 8515
+    },
+    {
+      "epoch": 0.6327750751977422,
+      "grad_norm": 1.048486351966858,
+      "learning_rate": 3.999463081471254e-05,
+      "loss": 0.3913,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6331464220728582,
+      "grad_norm": 1.449408769607544,
+      "learning_rate": 3.998259814604345e-05,
+      "loss": 0.3814,
+      "step": 8525
+    },
+    {
+      "epoch": 0.6335177689479743,
+      "grad_norm": 0.6910377144813538,
+      "learning_rate": 3.997056005885628e-05,
+      "loss": 0.3701,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6338891158230904,
+      "grad_norm": 2.5074048042297363,
+      "learning_rate": 3.9958516557504635e-05,
+      "loss": 0.4016,
+      "step": 8535
+    },
+    {
+      "epoch": 0.6342604626982064,
+      "grad_norm": 0.9100797772407532,
+      "learning_rate": 3.994646764634413e-05,
+      "loss": 0.368,
+      "step": 8540
+    },
+    {
+      "epoch": 0.6346318095733224,
+      "grad_norm": 0.9319835901260376,
+      "learning_rate": 3.993441332973228e-05,
+      "loss": 0.3704,
+      "step": 8545
+    },
+    {
+      "epoch": 0.6350031564484385,
+      "grad_norm": 1.1303112506866455,
+      "learning_rate": 3.9922353612028576e-05,
+      "loss": 0.3715,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6353745033235545,
+      "grad_norm": 0.8505479693412781,
+      "learning_rate": 3.991028849759448e-05,
+      "loss": 0.3563,
+      "step": 8555
+    },
+    {
+      "epoch": 0.6357458501986706,
+      "grad_norm": 0.7193156480789185,
+      "learning_rate": 3.9898217990793375e-05,
+      "loss": 0.3668,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6361171970737867,
+      "grad_norm": 0.8877584934234619,
+      "learning_rate": 3.9886142095990624e-05,
+      "loss": 0.3501,
+      "step": 8565
+    },
+    {
+      "epoch": 0.6364885439489026,
+      "grad_norm": 0.8598123788833618,
+      "learning_rate": 3.987406081755351e-05,
+      "loss": 0.3898,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6368598908240187,
+      "grad_norm": 0.7827747464179993,
+      "learning_rate": 3.9861974159851286e-05,
+      "loss": 0.3747,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6372312376991348,
+      "grad_norm": 1.0073925256729126,
+      "learning_rate": 3.984988212725514e-05,
+      "loss": 0.3731,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6376025845742508,
+      "grad_norm": 0.9102515578269958,
+      "learning_rate": 3.983778472413821e-05,
+      "loss": 0.3815,
+      "step": 8585
+    },
+    {
+      "epoch": 0.6379739314493669,
+      "grad_norm": 0.9365094900131226,
+      "learning_rate": 3.982568195487555e-05,
+      "loss": 0.3836,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6383452783244828,
+      "grad_norm": 0.9244111776351929,
+      "learning_rate": 3.9813573823844194e-05,
+      "loss": 0.3957,
+      "step": 8595
+    },
+    {
+      "epoch": 0.6387166251995989,
+      "grad_norm": 1.1187372207641602,
+      "learning_rate": 3.98014603354231e-05,
+      "loss": 0.4007,
+      "step": 8600
+    },
+    {
+      "epoch": 0.639087972074715,
+      "grad_norm": 0.9645322561264038,
+      "learning_rate": 3.978934149399315e-05,
+      "loss": 0.3768,
+      "step": 8605
+    },
+    {
+      "epoch": 0.639459318949831,
+      "grad_norm": 1.0242455005645752,
+      "learning_rate": 3.977721730393719e-05,
+      "loss": 0.3664,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6398306658249471,
+      "grad_norm": 0.9759249687194824,
+      "learning_rate": 3.976508776963996e-05,
+      "loss": 0.3875,
+      "step": 8615
+    },
+    {
+      "epoch": 0.6402020127000632,
+      "grad_norm": 0.7980156540870667,
+      "learning_rate": 3.975295289548817e-05,
+      "loss": 0.3865,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6405733595751791,
+      "grad_norm": 0.6709896326065063,
+      "learning_rate": 3.9740812685870445e-05,
+      "loss": 0.3523,
+      "step": 8625
+    },
+    {
+      "epoch": 0.6409447064502952,
+      "grad_norm": 0.8203238248825073,
+      "learning_rate": 3.972866714517734e-05,
+      "loss": 0.3508,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6413160533254113,
+      "grad_norm": 1.0209792852401733,
+      "learning_rate": 3.9716516277801344e-05,
+      "loss": 0.358,
+      "step": 8635
+    },
+    {
+      "epoch": 0.6416874002005273,
+      "grad_norm": 1.0689033269882202,
+      "learning_rate": 3.970436008813686e-05,
+      "loss": 0.3918,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6420587470756434,
+      "grad_norm": 1.2434110641479492,
+      "learning_rate": 3.969219858058022e-05,
+      "loss": 0.382,
+      "step": 8645
+    },
+    {
+      "epoch": 0.6424300939507595,
+      "grad_norm": 0.8145067691802979,
+      "learning_rate": 3.96800317595297e-05,
+      "loss": 0.3613,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6428014408258754,
+      "grad_norm": 1.0757936239242554,
+      "learning_rate": 3.9667859629385475e-05,
+      "loss": 0.3759,
+      "step": 8655
+    },
+    {
+      "epoch": 0.6431727877009915,
+      "grad_norm": 1.010427713394165,
+      "learning_rate": 3.9655682194549636e-05,
+      "loss": 0.3739,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6435441345761076,
+      "grad_norm": 0.7364680767059326,
+      "learning_rate": 3.96434994594262e-05,
+      "loss": 0.348,
+      "step": 8665
+    },
+    {
+      "epoch": 0.6439154814512236,
+      "grad_norm": 0.7643406391143799,
+      "learning_rate": 3.963131142842113e-05,
+      "loss": 0.3514,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6442868283263397,
+      "grad_norm": 0.9164127707481384,
+      "learning_rate": 3.961911810594226e-05,
+      "loss": 0.364,
+      "step": 8675
+    },
+    {
+      "epoch": 0.6446581752014556,
+      "grad_norm": 0.7336118221282959,
+      "learning_rate": 3.960691949639934e-05,
+      "loss": 0.3876,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6450295220765717,
+      "grad_norm": 0.728316605091095,
+      "learning_rate": 3.959471560420407e-05,
+      "loss": 0.3902,
+      "step": 8685
+    },
+    {
+      "epoch": 0.6454008689516878,
+      "grad_norm": 0.73708575963974,
+      "learning_rate": 3.958250643377003e-05,
+      "loss": 0.3751,
+      "step": 8690
+    },
+    {
+      "epoch": 0.6457722158268038,
+      "grad_norm": 0.708919107913971,
+      "learning_rate": 3.957029198951271e-05,
+      "loss": 0.3802,
+      "step": 8695
+    },
+    {
+      "epoch": 0.6461435627019199,
+      "grad_norm": 0.9214386343955994,
+      "learning_rate": 3.955807227584953e-05,
+      "loss": 0.3688,
+      "step": 8700
+    },
+    {
+      "epoch": 0.646514909577036,
+      "grad_norm": 0.8379608392715454,
+      "learning_rate": 3.9545847297199784e-05,
+      "loss": 0.3778,
+      "step": 8705
+    },
+    {
+      "epoch": 0.6468862564521519,
+      "grad_norm": 0.7550622820854187,
+      "learning_rate": 3.953361705798469e-05,
+      "loss": 0.3723,
+      "step": 8710
+    },
+    {
+      "epoch": 0.647257603327268,
+      "grad_norm": 0.9238272309303284,
+      "learning_rate": 3.952138156262736e-05,
+      "loss": 0.3716,
+      "step": 8715
+    },
+    {
+      "epoch": 0.6476289502023841,
+      "grad_norm": 0.844140887260437,
+      "learning_rate": 3.9509140815552826e-05,
+      "loss": 0.3875,
+      "step": 8720
+    },
+    {
+      "epoch": 0.6480002970775001,
+      "grad_norm": 0.643217146396637,
+      "learning_rate": 3.9496894821187995e-05,
+      "loss": 0.3714,
+      "step": 8725
+    },
+    {
+      "epoch": 0.6483716439526162,
+      "grad_norm": 0.7092486619949341,
+      "learning_rate": 3.948464358396168e-05,
+      "loss": 0.3826,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6487429908277322,
+      "grad_norm": 0.761620044708252,
+      "learning_rate": 3.947238710830459e-05,
+      "loss": 0.3705,
+      "step": 8735
+    },
+    {
+      "epoch": 0.6491143377028482,
+      "grad_norm": 0.7885283827781677,
+      "learning_rate": 3.946012539864933e-05,
+      "loss": 0.3824,
+      "step": 8740
+    },
+    {
+      "epoch": 0.6494856845779643,
+      "grad_norm": 0.7201789021492004,
+      "learning_rate": 3.94478584594304e-05,
+      "loss": 0.3892,
+      "step": 8745
+    },
+    {
+      "epoch": 0.6498570314530803,
+      "grad_norm": 0.8360629081726074,
+      "learning_rate": 3.943558629508418e-05,
+      "loss": 0.3712,
+      "step": 8750
+    },
+    {
+      "epoch": 0.6502283783281964,
+      "grad_norm": 0.5975351333618164,
+      "learning_rate": 3.9423308910048974e-05,
+      "loss": 0.373,
+      "step": 8755
+    },
+    {
+      "epoch": 0.6505997252033124,
+      "grad_norm": 0.7709315419197083,
+      "learning_rate": 3.941102630876492e-05,
+      "loss": 0.3562,
+      "step": 8760
+    },
+    {
+      "epoch": 0.6509710720784284,
+      "grad_norm": 0.7547560930252075,
+      "learning_rate": 3.9398738495674096e-05,
+      "loss": 0.3672,
+      "step": 8765
+    },
+    {
+      "epoch": 0.6513424189535445,
+      "grad_norm": 0.7901609539985657,
+      "learning_rate": 3.938644547522041e-05,
+      "loss": 0.3691,
+      "step": 8770
+    },
+    {
+      "epoch": 0.6517137658286606,
+      "grad_norm": 0.9239474534988403,
+      "learning_rate": 3.9374147251849704e-05,
+      "loss": 0.3717,
+      "step": 8775
+    },
+    {
+      "epoch": 0.6520851127037766,
+      "grad_norm": 0.9300574064254761,
+      "learning_rate": 3.936184383000968e-05,
+      "loss": 0.3667,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6524564595788926,
+      "grad_norm": 0.8764330148696899,
+      "learning_rate": 3.934953521414992e-05,
+      "loss": 0.3529,
+      "step": 8785
+    },
+    {
+      "epoch": 0.6528278064540087,
+      "grad_norm": 0.8157888650894165,
+      "learning_rate": 3.933722140872188e-05,
+      "loss": 0.3712,
+      "step": 8790
+    },
+    {
+      "epoch": 0.6531991533291247,
+      "grad_norm": 0.8766961097717285,
+      "learning_rate": 3.93249024181789e-05,
+      "loss": 0.3557,
+      "step": 8795
+    },
+    {
+      "epoch": 0.6535705002042408,
+      "grad_norm": 0.8514896631240845,
+      "learning_rate": 3.931257824697619e-05,
+      "loss": 0.3978,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6539418470793569,
+      "grad_norm": 1.4988986253738403,
+      "learning_rate": 3.9300248899570855e-05,
+      "loss": 0.3756,
+      "step": 8805
+    },
+    {
+      "epoch": 0.6543131939544728,
+      "grad_norm": 1.069677710533142,
+      "learning_rate": 3.928791438042184e-05,
+      "loss": 0.3773,
+      "step": 8810
+    },
+    {
+      "epoch": 0.6546845408295889,
+      "grad_norm": 0.7891872525215149,
+      "learning_rate": 3.9275574693989966e-05,
+      "loss": 0.3627,
+      "step": 8815
+    },
+    {
+      "epoch": 0.655055887704705,
+      "grad_norm": 1.4193977117538452,
+      "learning_rate": 3.926322984473795e-05,
+      "loss": 0.3792,
+      "step": 8820
+    },
+    {
+      "epoch": 0.655427234579821,
+      "grad_norm": 0.7989646792411804,
+      "learning_rate": 3.925087983713034e-05,
+      "loss": 0.3705,
+      "step": 8825
+    },
+    {
+      "epoch": 0.6557985814549371,
+      "grad_norm": 0.8563501238822937,
+      "learning_rate": 3.923852467563358e-05,
+      "loss": 0.3847,
+      "step": 8830
+    },
+    {
+      "epoch": 0.656169928330053,
+      "grad_norm": 0.8602546453475952,
+      "learning_rate": 3.922616436471597e-05,
+      "loss": 0.3738,
+      "step": 8835
+    },
+    {
+      "epoch": 0.6565412752051691,
+      "grad_norm": 0.9216306209564209,
+      "learning_rate": 3.921379890884764e-05,
+      "loss": 0.3612,
+      "step": 8840
+    },
+    {
+      "epoch": 0.6569126220802852,
+      "grad_norm": 0.8325752019882202,
+      "learning_rate": 3.920142831250063e-05,
+      "loss": 0.3593,
+      "step": 8845
+    },
+    {
+      "epoch": 0.6572839689554012,
+      "grad_norm": 0.9701927304267883,
+      "learning_rate": 3.918905258014882e-05,
+      "loss": 0.3981,
+      "step": 8850
+    },
+    {
+      "epoch": 0.6576553158305173,
+      "grad_norm": 0.781143069267273,
+      "learning_rate": 3.917667171626793e-05,
+      "loss": 0.3792,
+      "step": 8855
+    },
+    {
+      "epoch": 0.6580266627056334,
+      "grad_norm": 0.7671363353729248,
+      "learning_rate": 3.916428572533555e-05,
+      "loss": 0.3833,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6583980095807493,
+      "grad_norm": 0.7129806876182556,
+      "learning_rate": 3.915189461183114e-05,
+      "loss": 0.3648,
+      "step": 8865
+    },
+    {
+      "epoch": 0.6587693564558654,
+      "grad_norm": 0.9245615005493164,
+      "learning_rate": 3.9139498380235985e-05,
+      "loss": 0.3976,
+      "step": 8870
+    },
+    {
+      "epoch": 0.6591407033309815,
+      "grad_norm": 0.8696265816688538,
+      "learning_rate": 3.9127097035033234e-05,
+      "loss": 0.3717,
+      "step": 8875
+    },
+    {
+      "epoch": 0.6595120502060975,
+      "grad_norm": 0.9935540556907654,
+      "learning_rate": 3.911469058070789e-05,
+      "loss": 0.3638,
+      "step": 8880
+    },
+    {
+      "epoch": 0.6598833970812136,
+      "grad_norm": 0.8621585369110107,
+      "learning_rate": 3.910227902174679e-05,
+      "loss": 0.3765,
+      "step": 8885
+    },
+    {
+      "epoch": 0.6602547439563297,
+      "grad_norm": 0.7071739435195923,
+      "learning_rate": 3.908986236263863e-05,
+      "loss": 0.3521,
+      "step": 8890
+    },
+    {
+      "epoch": 0.6606260908314456,
+      "grad_norm": 1.5234918594360352,
+      "learning_rate": 3.907744060787395e-05,
+      "loss": 0.3993,
+      "step": 8895
+    },
+    {
+      "epoch": 0.6609974377065617,
+      "grad_norm": 1.445349097251892,
+      "learning_rate": 3.906501376194512e-05,
+      "loss": 0.371,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6613687845816777,
+      "grad_norm": 0.8245334625244141,
+      "learning_rate": 3.905258182934637e-05,
+      "loss": 0.4072,
+      "step": 8905
+    },
+    {
+      "epoch": 0.6617401314567938,
+      "grad_norm": 0.80816251039505,
+      "learning_rate": 3.904014481457374e-05,
+      "loss": 0.3406,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6621114783319099,
+      "grad_norm": 0.8921861052513123,
+      "learning_rate": 3.902770272212516e-05,
+      "loss": 0.3795,
+      "step": 8915
+    },
+    {
+      "epoch": 0.6624828252070258,
+      "grad_norm": 1.2531591653823853,
+      "learning_rate": 3.901525555650033e-05,
+      "loss": 0.3854,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6628541720821419,
+      "grad_norm": 0.9966285824775696,
+      "learning_rate": 3.900280332220083e-05,
+      "loss": 0.3734,
+      "step": 8925
+    },
+    {
+      "epoch": 0.663225518957258,
+      "grad_norm": 0.8290938138961792,
+      "learning_rate": 3.899034602373007e-05,
+      "loss": 0.3667,
+      "step": 8930
+    },
+    {
+      "epoch": 0.663596865832374,
+      "grad_norm": 0.7914450764656067,
+      "learning_rate": 3.8977883665593284e-05,
+      "loss": 0.387,
+      "step": 8935
+    },
+    {
+      "epoch": 0.6639682127074901,
+      "grad_norm": 0.899835467338562,
+      "learning_rate": 3.896541625229752e-05,
+      "loss": 0.364,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6643395595826062,
+      "grad_norm": 0.8077391982078552,
+      "learning_rate": 3.895294378835168e-05,
+      "loss": 0.3782,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6647109064577221,
+      "grad_norm": 0.8813906908035278,
+      "learning_rate": 3.8940466278266476e-05,
+      "loss": 0.3549,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6650822533328382,
+      "grad_norm": 0.7896718978881836,
+      "learning_rate": 3.892798372655446e-05,
+      "loss": 0.3783,
+      "step": 8955
+    },
+    {
+      "epoch": 0.6654536002079543,
+      "grad_norm": 1.244232416152954,
+      "learning_rate": 3.891549613772999e-05,
+      "loss": 0.372,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6658249470830703,
+      "grad_norm": 0.776960015296936,
+      "learning_rate": 3.8903003516309256e-05,
+      "loss": 0.3989,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6661962939581864,
+      "grad_norm": 0.8111973404884338,
+      "learning_rate": 3.889050586681027e-05,
+      "loss": 0.385,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6665676408333023,
+      "grad_norm": 0.7727485299110413,
+      "learning_rate": 3.887800319375284e-05,
+      "loss": 0.358,
+      "step": 8975
+    },
+    {
+      "epoch": 0.6669389877084184,
+      "grad_norm": 0.8484100103378296,
+      "learning_rate": 3.886549550165863e-05,
+      "loss": 0.3947,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6673103345835345,
+      "grad_norm": 0.9307198524475098,
+      "learning_rate": 3.88529827950511e-05,
+      "loss": 0.3645,
+      "step": 8985
+    },
+    {
+      "epoch": 0.6676816814586505,
+      "grad_norm": 0.9293243288993835,
+      "learning_rate": 3.8840465078455505e-05,
+      "loss": 0.3702,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6680530283337666,
+      "grad_norm": 0.8592212796211243,
+      "learning_rate": 3.882794235639894e-05,
+      "loss": 0.3785,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6684243752088826,
+      "grad_norm": 0.7336313128471375,
+      "learning_rate": 3.881541463341029e-05,
+      "loss": 0.3673,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6687957220839986,
+      "grad_norm": 0.9456973671913147,
+      "learning_rate": 3.880288191402026e-05,
+      "loss": 0.3889,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6691670689591147,
+      "grad_norm": 0.9240425825119019,
+      "learning_rate": 3.8790344202761365e-05,
+      "loss": 0.3637,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6695384158342308,
+      "grad_norm": 1.105124831199646,
+      "learning_rate": 3.877780150416792e-05,
+      "loss": 0.3795,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6699097627093468,
+      "grad_norm": 0.8785472512245178,
+      "learning_rate": 3.8765253822776036e-05,
+      "loss": 0.3639,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6702811095844629,
+      "grad_norm": 1.0078473091125488,
+      "learning_rate": 3.875270116312364e-05,
+      "loss": 0.3979,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6706524564595789,
+      "grad_norm": 0.7979200482368469,
+      "learning_rate": 3.874014352975045e-05,
+      "loss": 0.3602,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6710238033346949,
+      "grad_norm": 0.7981281876564026,
+      "learning_rate": 3.872758092719798e-05,
+      "loss": 0.36,
+      "step": 9035
+    },
+    {
+      "epoch": 0.671395150209811,
+      "grad_norm": 2.0684547424316406,
+      "learning_rate": 3.871501336000957e-05,
+      "loss": 0.3873,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6717664970849271,
+      "grad_norm": 0.6863536238670349,
+      "learning_rate": 3.87024408327303e-05,
+      "loss": 0.3738,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6721378439600431,
+      "grad_norm": 0.7692870497703552,
+      "learning_rate": 3.8689863349907084e-05,
+      "loss": 0.3639,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6725091908351591,
+      "grad_norm": 0.8045924305915833,
+      "learning_rate": 3.8677280916088645e-05,
+      "loss": 0.3557,
+      "step": 9055
+    },
+    {
+      "epoch": 0.6728805377102751,
+      "grad_norm": 1.2326745986938477,
+      "learning_rate": 3.866469353582545e-05,
+      "loss": 0.3831,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6732518845853912,
+      "grad_norm": 0.9144732356071472,
+      "learning_rate": 3.865210121366977e-05,
+      "loss": 0.3797,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6736232314605073,
+      "grad_norm": 1.0086545944213867,
+      "learning_rate": 3.863950395417569e-05,
+      "loss": 0.3989,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6739945783356233,
+      "grad_norm": 1.0512826442718506,
+      "learning_rate": 3.8626901761899045e-05,
+      "loss": 0.3837,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6743659252107393,
+      "grad_norm": 0.9185574054718018,
+      "learning_rate": 3.861429464139748e-05,
+      "loss": 0.369,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6747372720858554,
+      "grad_norm": 0.9505536556243896,
+      "learning_rate": 3.860168259723041e-05,
+      "loss": 0.365,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6751086189609714,
+      "grad_norm": 0.8058850169181824,
+      "learning_rate": 3.8589065633959035e-05,
+      "loss": 0.361,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6754799658360875,
+      "grad_norm": 0.8129885792732239,
+      "learning_rate": 3.857644375614631e-05,
+      "loss": 0.3766,
+      "step": 9095
+    },
+    {
+      "epoch": 0.6758513127112036,
+      "grad_norm": 0.641776978969574,
+      "learning_rate": 3.8563816968357016e-05,
+      "loss": 0.3641,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6762226595863196,
+      "grad_norm": 0.8404680490493774,
+      "learning_rate": 3.855118527515767e-05,
+      "loss": 0.361,
+      "step": 9105
+    },
+    {
+      "epoch": 0.6765940064614356,
+      "grad_norm": 0.8596828579902649,
+      "learning_rate": 3.853854868111658e-05,
+      "loss": 0.3523,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6769653533365517,
+      "grad_norm": 0.7938897013664246,
+      "learning_rate": 3.8525907190803816e-05,
+      "loss": 0.3705,
+      "step": 9115
+    },
+    {
+      "epoch": 0.6773367002116677,
+      "grad_norm": 0.9502366781234741,
+      "learning_rate": 3.851326080879123e-05,
+      "loss": 0.3702,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6777080470867838,
+      "grad_norm": 0.7413920760154724,
+      "learning_rate": 3.850060953965244e-05,
+      "loss": 0.369,
+      "step": 9125
+    },
+    {
+      "epoch": 0.6780793939618998,
+      "grad_norm": 0.9958828687667847,
+      "learning_rate": 3.848795338796283e-05,
+      "loss": 0.3825,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6784507408370158,
+      "grad_norm": 0.7100253105163574,
+      "learning_rate": 3.847529235829953e-05,
+      "loss": 0.352,
+      "step": 9135
+    },
+    {
+      "epoch": 0.6788220877121319,
+      "grad_norm": 1.3226805925369263,
+      "learning_rate": 3.846262645524148e-05,
+      "loss": 0.3683,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6791934345872479,
+      "grad_norm": 0.9204307198524475,
+      "learning_rate": 3.8449955683369346e-05,
+      "loss": 0.3678,
+      "step": 9145
+    },
+    {
+      "epoch": 0.679564781462364,
+      "grad_norm": 1.0092697143554688,
+      "learning_rate": 3.8437280047265554e-05,
+      "loss": 0.3743,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6799361283374801,
+      "grad_norm": 1.3019070625305176,
+      "learning_rate": 3.8424599551514314e-05,
+      "loss": 0.3702,
+      "step": 9155
+    },
+    {
+      "epoch": 0.680307475212596,
+      "grad_norm": 0.7268680334091187,
+      "learning_rate": 3.841191420070158e-05,
+      "loss": 0.3635,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6806788220877121,
+      "grad_norm": 0.938025176525116,
+      "learning_rate": 3.839922399941505e-05,
+      "loss": 0.3854,
+      "step": 9165
+    },
+    {
+      "epoch": 0.6810501689628282,
+      "grad_norm": 0.8636729717254639,
+      "learning_rate": 3.838652895224419e-05,
+      "loss": 0.3751,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6814215158379442,
+      "grad_norm": 0.7120767831802368,
+      "learning_rate": 3.837382906378022e-05,
+      "loss": 0.3751,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6817928627130603,
+      "grad_norm": 0.9305611848831177,
+      "learning_rate": 3.8361124338616105e-05,
+      "loss": 0.3635,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6821642095881764,
+      "grad_norm": 0.6582894325256348,
+      "learning_rate": 3.834841478134656e-05,
+      "loss": 0.3625,
+      "step": 9185
+    },
+    {
+      "epoch": 0.6825355564632923,
+      "grad_norm": 0.856611967086792,
+      "learning_rate": 3.833570039656805e-05,
+      "loss": 0.3726,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6829069033384084,
+      "grad_norm": 0.9939899444580078,
+      "learning_rate": 3.8322981188878786e-05,
+      "loss": 0.3685,
+      "step": 9195
+    },
+    {
+      "epoch": 0.6832782502135245,
+      "grad_norm": 0.9134172797203064,
+      "learning_rate": 3.8310257162878724e-05,
+      "loss": 0.3914,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6836495970886405,
+      "grad_norm": 0.9030920267105103,
+      "learning_rate": 3.829752832316955e-05,
+      "loss": 0.3782,
+      "step": 9205
+    },
+    {
+      "epoch": 0.6840209439637566,
+      "grad_norm": 0.9032113552093506,
+      "learning_rate": 3.828479467435471e-05,
+      "loss": 0.3646,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6843922908388725,
+      "grad_norm": 1.0852000713348389,
+      "learning_rate": 3.8272056221039375e-05,
+      "loss": 0.3481,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6847636377139886,
+      "grad_norm": 0.8781152367591858,
+      "learning_rate": 3.8259312967830465e-05,
+      "loss": 0.3631,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6851349845891047,
+      "grad_norm": 0.7677597999572754,
+      "learning_rate": 3.824656491933662e-05,
+      "loss": 0.3524,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6855063314642207,
+      "grad_norm": 0.7801468968391418,
+      "learning_rate": 3.823381208016823e-05,
+      "loss": 0.3917,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6858776783393368,
+      "grad_norm": 0.7930712103843689,
+      "learning_rate": 3.822105445493742e-05,
+      "loss": 0.391,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6862490252144529,
+      "grad_norm": 0.8240649104118347,
+      "learning_rate": 3.820829204825802e-05,
+      "loss": 0.3703,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6866203720895688,
+      "grad_norm": 0.7643125057220459,
+      "learning_rate": 3.819552486474562e-05,
+      "loss": 0.3791,
+      "step": 9245
+    },
+    {
+      "epoch": 0.6869917189646849,
+      "grad_norm": 0.9027881622314453,
+      "learning_rate": 3.818275290901753e-05,
+      "loss": 0.3388,
+      "step": 9250
+    },
+    {
+      "epoch": 0.687363065839801,
+      "grad_norm": 0.9473441243171692,
+      "learning_rate": 3.816997618569276e-05,
+      "loss": 0.3743,
+      "step": 9255
+    },
+    {
+      "epoch": 0.687734412714917,
+      "grad_norm": 0.8707664012908936,
+      "learning_rate": 3.8157194699392074e-05,
+      "loss": 0.3624,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6881057595900331,
+      "grad_norm": 0.9271723628044128,
+      "learning_rate": 3.814440845473795e-05,
+      "loss": 0.3659,
+      "step": 9265
+    },
+    {
+      "epoch": 0.6884771064651491,
+      "grad_norm": 0.7871653437614441,
+      "learning_rate": 3.813161745635459e-05,
+      "loss": 0.3624,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6888484533402651,
+      "grad_norm": 0.8163115978240967,
+      "learning_rate": 3.8118821708867903e-05,
+      "loss": 0.3694,
+      "step": 9275
+    },
+    {
+      "epoch": 0.6892198002153812,
+      "grad_norm": 1.2728971242904663,
+      "learning_rate": 3.810602121690553e-05,
+      "loss": 0.3728,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6895911470904972,
+      "grad_norm": 0.8345783352851868,
+      "learning_rate": 3.809321598509682e-05,
+      "loss": 0.3616,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6899624939656133,
+      "grad_norm": 0.7567407488822937,
+      "learning_rate": 3.808040601807283e-05,
+      "loss": 0.3727,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6903338408407294,
+      "grad_norm": 0.8576236963272095,
+      "learning_rate": 3.806759132046635e-05,
+      "loss": 0.3884,
+      "step": 9295
+    },
+    {
+      "epoch": 0.6907051877158453,
+      "grad_norm": 0.7449585199356079,
+      "learning_rate": 3.805477189691186e-05,
+      "loss": 0.3729,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6910765345909614,
+      "grad_norm": 0.9378268122673035,
+      "learning_rate": 3.804194775204556e-05,
+      "loss": 0.36,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6914478814660775,
+      "grad_norm": 1.0872933864593506,
+      "learning_rate": 3.802911889050534e-05,
+      "loss": 0.3761,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6918192283411935,
+      "grad_norm": 0.8560049533843994,
+      "learning_rate": 3.8016285316930845e-05,
+      "loss": 0.3716,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6921905752163096,
+      "grad_norm": 1.188044786453247,
+      "learning_rate": 3.800344703596335e-05,
+      "loss": 0.3558,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6925619220914256,
+      "grad_norm": 0.8799236416816711,
+      "learning_rate": 3.79906040522459e-05,
+      "loss": 0.3653,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6929332689665416,
+      "grad_norm": 1.0614551305770874,
+      "learning_rate": 3.79777563704232e-05,
+      "loss": 0.3722,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6933046158416577,
+      "grad_norm": 0.8129174113273621,
+      "learning_rate": 3.796490399514167e-05,
+      "loss": 0.3597,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6936759627167738,
+      "grad_norm": 0.9150453209877014,
+      "learning_rate": 3.7952046931049425e-05,
+      "loss": 0.375,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6940473095918898,
+      "grad_norm": 0.8379879593849182,
+      "learning_rate": 3.7939185182796266e-05,
+      "loss": 0.3811,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6944186564670058,
+      "grad_norm": 1.0485833883285522,
+      "learning_rate": 3.7926318755033716e-05,
+      "loss": 0.3758,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6947900033421219,
+      "grad_norm": 0.7375094890594482,
+      "learning_rate": 3.7913447652414955e-05,
+      "loss": 0.4004,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6951613502172379,
+      "grad_norm": 0.8998247385025024,
+      "learning_rate": 3.7900571879594874e-05,
+      "loss": 0.3741,
+      "step": 9360
+    },
+    {
+      "epoch": 0.695532697092354,
+      "grad_norm": 0.7974969148635864,
+      "learning_rate": 3.7887691441230055e-05,
+      "loss": 0.3839,
+      "step": 9365
+    },
+    {
+      "epoch": 0.69590404396747,
+      "grad_norm": 0.7805307507514954,
+      "learning_rate": 3.787480634197875e-05,
+      "loss": 0.3626,
+      "step": 9370
+    },
+    {
+      "epoch": 0.696275390842586,
+      "grad_norm": 0.8351677656173706,
+      "learning_rate": 3.786191658650092e-05,
+      "loss": 0.368,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6966467377177021,
+      "grad_norm": 0.9680357575416565,
+      "learning_rate": 3.784902217945819e-05,
+      "loss": 0.4092,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6970180845928181,
+      "grad_norm": 0.7669398784637451,
+      "learning_rate": 3.7836123125513875e-05,
+      "loss": 0.3497,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6973894314679342,
+      "grad_norm": 0.7332344055175781,
+      "learning_rate": 3.782321942933299e-05,
+      "loss": 0.3643,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6977607783430503,
+      "grad_norm": 0.7292071580886841,
+      "learning_rate": 3.7810311095582176e-05,
+      "loss": 0.3805,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6981321252181663,
+      "grad_norm": 0.8352912664413452,
+      "learning_rate": 3.7797398128929805e-05,
+      "loss": 0.3792,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6985034720932823,
+      "grad_norm": 0.7182777523994446,
+      "learning_rate": 3.778448053404591e-05,
+      "loss": 0.3643,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6988748189683984,
+      "grad_norm": 1.0204569101333618,
+      "learning_rate": 3.777155831560217e-05,
+      "loss": 0.3696,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6992461658435144,
+      "grad_norm": 1.1198655366897583,
+      "learning_rate": 3.775863147827198e-05,
+      "loss": 0.3755,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6996175127186305,
+      "grad_norm": 0.8152322173118591,
+      "learning_rate": 3.774570002673038e-05,
+      "loss": 0.3673,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6999888595937466,
+      "grad_norm": 0.6712133288383484,
+      "learning_rate": 3.7732763965654075e-05,
+      "loss": 0.3435,
+      "step": 9425
+    },
+    {
+      "epoch": 0.7003602064688625,
+      "grad_norm": 1.092829704284668,
+      "learning_rate": 3.7719823299721443e-05,
+      "loss": 0.3627,
+      "step": 9430
+    },
+    {
+      "epoch": 0.7007315533439786,
+      "grad_norm": 1.0296307802200317,
+      "learning_rate": 3.7706878033612534e-05,
+      "loss": 0.3787,
+      "step": 9435
+    },
+    {
+      "epoch": 0.7011029002190946,
+      "grad_norm": 0.8601410388946533,
+      "learning_rate": 3.7693928172009066e-05,
+      "loss": 0.3877,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7014742470942107,
+      "grad_norm": 1.8063080310821533,
+      "learning_rate": 3.7680973719594394e-05,
+      "loss": 0.3596,
+      "step": 9445
+    },
+    {
+      "epoch": 0.7018455939693268,
+      "grad_norm": 1.0898480415344238,
+      "learning_rate": 3.766801468105355e-05,
+      "loss": 0.3585,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7022169408444427,
+      "grad_norm": 0.840470552444458,
+      "learning_rate": 3.7655051061073234e-05,
+      "loss": 0.3789,
+      "step": 9455
+    },
+    {
+      "epoch": 0.7025882877195588,
+      "grad_norm": 0.8825913667678833,
+      "learning_rate": 3.764208286434179e-05,
+      "loss": 0.375,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7029596345946749,
+      "grad_norm": 0.7693654894828796,
+      "learning_rate": 3.762911009554921e-05,
+      "loss": 0.3469,
+      "step": 9465
+    },
+    {
+      "epoch": 0.7033309814697909,
+      "grad_norm": 0.7990744113922119,
+      "learning_rate": 3.7616132759387156e-05,
+      "loss": 0.3498,
+      "step": 9470
+    },
+    {
+      "epoch": 0.703702328344907,
+      "grad_norm": 0.8645015954971313,
+      "learning_rate": 3.760315086054893e-05,
+      "loss": 0.3721,
+      "step": 9475
+    },
+    {
+      "epoch": 0.7040736752200231,
+      "grad_norm": 0.8150882720947266,
+      "learning_rate": 3.75901644037295e-05,
+      "loss": 0.369,
+      "step": 9480
+    },
+    {
+      "epoch": 0.704445022095139,
+      "grad_norm": 0.7890775203704834,
+      "learning_rate": 3.757717339362546e-05,
+      "loss": 0.3891,
+      "step": 9485
+    },
+    {
+      "epoch": 0.7048163689702551,
+      "grad_norm": 0.7995883226394653,
+      "learning_rate": 3.756417783493507e-05,
+      "loss": 0.3625,
+      "step": 9490
+    },
+    {
+      "epoch": 0.7051877158453712,
+      "grad_norm": 1.114989161491394,
+      "learning_rate": 3.755117773235823e-05,
+      "loss": 0.3743,
+      "step": 9495
+    },
+    {
+      "epoch": 0.7055590627204872,
+      "grad_norm": 1.1512959003448486,
+      "learning_rate": 3.753817309059646e-05,
+      "loss": 0.3785,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7059304095956033,
+      "grad_norm": 0.7948722839355469,
+      "learning_rate": 3.752516391435297e-05,
+      "loss": 0.3863,
+      "step": 9505
+    },
+    {
+      "epoch": 0.7063017564707192,
+      "grad_norm": 0.9194521307945251,
+      "learning_rate": 3.751215020833255e-05,
+      "loss": 0.3619,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7066731033458353,
+      "grad_norm": 0.8141660690307617,
+      "learning_rate": 3.749913197724169e-05,
+      "loss": 0.3721,
+      "step": 9515
+    },
+    {
+      "epoch": 0.7070444502209514,
+      "grad_norm": 0.7911691665649414,
+      "learning_rate": 3.748610922578848e-05,
+      "loss": 0.3782,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7074157970960674,
+      "grad_norm": 0.9009805917739868,
+      "learning_rate": 3.747308195868263e-05,
+      "loss": 0.3323,
+      "step": 9525
+    },
+    {
+      "epoch": 0.7077871439711835,
+      "grad_norm": 1.0365558862686157,
+      "learning_rate": 3.746005018063554e-05,
+      "loss": 0.368,
+      "step": 9530
+    },
+    {
+      "epoch": 0.7081584908462996,
+      "grad_norm": 0.8006228804588318,
+      "learning_rate": 3.744701389636018e-05,
+      "loss": 0.3686,
+      "step": 9535
+    },
+    {
+      "epoch": 0.7085298377214155,
+      "grad_norm": 0.87027907371521,
+      "learning_rate": 3.743397311057118e-05,
+      "loss": 0.349,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7089011845965316,
+      "grad_norm": 0.8917665481567383,
+      "learning_rate": 3.742092782798479e-05,
+      "loss": 0.3655,
+      "step": 9545
+    },
+    {
+      "epoch": 0.7092725314716477,
+      "grad_norm": 1.0651155710220337,
+      "learning_rate": 3.740787805331891e-05,
+      "loss": 0.3842,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7096438783467637,
+      "grad_norm": 1.054017424583435,
+      "learning_rate": 3.739482379129302e-05,
+      "loss": 0.3916,
+      "step": 9555
+    },
+    {
+      "epoch": 0.7100152252218798,
+      "grad_norm": 0.6938828229904175,
+      "learning_rate": 3.738176504662826e-05,
+      "loss": 0.3645,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7103865720969958,
+      "grad_norm": 0.7110164165496826,
+      "learning_rate": 3.736870182404738e-05,
+      "loss": 0.356,
+      "step": 9565
+    },
+    {
+      "epoch": 0.7107579189721118,
+      "grad_norm": 1.1128145456314087,
+      "learning_rate": 3.735563412827475e-05,
+      "loss": 0.3733,
+      "step": 9570
+    },
+    {
+      "epoch": 0.7111292658472279,
+      "grad_norm": 0.9033927917480469,
+      "learning_rate": 3.734256196403634e-05,
+      "loss": 0.3504,
+      "step": 9575
+    },
+    {
+      "epoch": 0.711500612722344,
+      "grad_norm": 0.88112473487854,
+      "learning_rate": 3.7329485336059774e-05,
+      "loss": 0.3631,
+      "step": 9580
+    },
+    {
+      "epoch": 0.71187195959746,
+      "grad_norm": 0.7752759456634521,
+      "learning_rate": 3.731640424907425e-05,
+      "loss": 0.3491,
+      "step": 9585
+    },
+    {
+      "epoch": 0.712243306472576,
+      "grad_norm": 1.040805697441101,
+      "learning_rate": 3.7303318707810595e-05,
+      "loss": 0.3647,
+      "step": 9590
+    },
+    {
+      "epoch": 0.712614653347692,
+      "grad_norm": 0.7947229146957397,
+      "learning_rate": 3.7290228717001275e-05,
+      "loss": 0.3535,
+      "step": 9595
+    },
+    {
+      "epoch": 0.7129860002228081,
+      "grad_norm": 0.8026042580604553,
+      "learning_rate": 3.72771342813803e-05,
+      "loss": 0.3818,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7133573470979242,
+      "grad_norm": 0.9147016406059265,
+      "learning_rate": 3.7264035405683364e-05,
+      "loss": 0.3651,
+      "step": 9605
+    },
+    {
+      "epoch": 0.7137286939730402,
+      "grad_norm": 1.2373228073120117,
+      "learning_rate": 3.7250932094647706e-05,
+      "loss": 0.3719,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7141000408481563,
+      "grad_norm": 0.8862619996070862,
+      "learning_rate": 3.72378243530122e-05,
+      "loss": 0.364,
+      "step": 9615
+    },
+    {
+      "epoch": 0.7144713877232723,
+      "grad_norm": 0.8791931867599487,
+      "learning_rate": 3.722471218551732e-05,
+      "loss": 0.3795,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7148427345983883,
+      "grad_norm": 0.9781063199043274,
+      "learning_rate": 3.7211595596905115e-05,
+      "loss": 0.3658,
+      "step": 9625
+    },
+    {
+      "epoch": 0.7152140814735044,
+      "grad_norm": 0.7280610799789429,
+      "learning_rate": 3.719847459191928e-05,
+      "loss": 0.3742,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7155854283486205,
+      "grad_norm": 0.8904832601547241,
+      "learning_rate": 3.718534917530505e-05,
+      "loss": 0.3596,
+      "step": 9635
+    },
+    {
+      "epoch": 0.7159567752237365,
+      "grad_norm": 1.0527021884918213,
+      "learning_rate": 3.7172219351809315e-05,
+      "loss": 0.3633,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7163281220988525,
+      "grad_norm": 1.0968849658966064,
+      "learning_rate": 3.715908512618051e-05,
+      "loss": 0.3743,
+      "step": 9645
+    },
+    {
+      "epoch": 0.7166994689739686,
+      "grad_norm": 0.9742524027824402,
+      "learning_rate": 3.714594650316869e-05,
+      "loss": 0.3759,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7170708158490846,
+      "grad_norm": 0.8875055909156799,
+      "learning_rate": 3.71328034875255e-05,
+      "loss": 0.358,
+      "step": 9655
+    },
+    {
+      "epoch": 0.7174421627242007,
+      "grad_norm": 0.8786531686782837,
+      "learning_rate": 3.711965608400414e-05,
+      "loss": 0.3813,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7178135095993167,
+      "grad_norm": 0.9650137424468994,
+      "learning_rate": 3.710650429735946e-05,
+      "loss": 0.3534,
+      "step": 9665
+    },
+    {
+      "epoch": 0.7181848564744328,
+      "grad_norm": 0.7877172231674194,
+      "learning_rate": 3.709334813234782e-05,
+      "loss": 0.3575,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7185562033495488,
+      "grad_norm": 1.950656533241272,
+      "learning_rate": 3.708018759372722e-05,
+      "loss": 0.3697,
+      "step": 9675
+    },
+    {
+      "epoch": 0.7189275502246648,
+      "grad_norm": 0.8109143376350403,
+      "learning_rate": 3.706702268625721e-05,
+      "loss": 0.3487,
+      "step": 9680
+    },
+    {
+      "epoch": 0.7192988970997809,
+      "grad_norm": 0.8184154033660889,
+      "learning_rate": 3.705385341469896e-05,
+      "loss": 0.3634,
+      "step": 9685
+    },
+    {
+      "epoch": 0.719670243974897,
+      "grad_norm": 0.960198700428009,
+      "learning_rate": 3.704067978381517e-05,
+      "loss": 0.3768,
+      "step": 9690
+    },
+    {
+      "epoch": 0.720041590850013,
+      "grad_norm": 0.9191488027572632,
+      "learning_rate": 3.702750179837013e-05,
+      "loss": 0.3566,
+      "step": 9695
+    },
+    {
+      "epoch": 0.720412937725129,
+      "grad_norm": 1.1562496423721313,
+      "learning_rate": 3.701431946312974e-05,
+      "loss": 0.3661,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7207842846002451,
+      "grad_norm": 0.9678395390510559,
+      "learning_rate": 3.7001132782861414e-05,
+      "loss": 0.3659,
+      "step": 9705
+    },
+    {
+      "epoch": 0.7211556314753611,
+      "grad_norm": 0.8770809769630432,
+      "learning_rate": 3.69879417623342e-05,
+      "loss": 0.3867,
+      "step": 9710
+    },
+    {
+      "epoch": 0.7215269783504772,
+      "grad_norm": 0.9320460557937622,
+      "learning_rate": 3.697474640631866e-05,
+      "loss": 0.358,
+      "step": 9715
+    },
+    {
+      "epoch": 0.7218983252255933,
+      "grad_norm": 0.8323056101799011,
+      "learning_rate": 3.6961546719586955e-05,
+      "loss": 0.3941,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7222696721007092,
+      "grad_norm": 0.7591274976730347,
+      "learning_rate": 3.6948342706912815e-05,
+      "loss": 0.3754,
+      "step": 9725
+    },
+    {
+      "epoch": 0.7226410189758253,
+      "grad_norm": 0.9951517581939697,
+      "learning_rate": 3.693513437307152e-05,
+      "loss": 0.3804,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7230123658509414,
+      "grad_norm": 1.128585696220398,
+      "learning_rate": 3.6921921722839906e-05,
+      "loss": 0.3622,
+      "step": 9735
+    },
+    {
+      "epoch": 0.7233837127260574,
+      "grad_norm": 1.0597444772720337,
+      "learning_rate": 3.69087047609964e-05,
+      "loss": 0.3773,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7237550596011735,
+      "grad_norm": 2.1655824184417725,
+      "learning_rate": 3.689548349232097e-05,
+      "loss": 0.3649,
+      "step": 9745
+    },
+    {
+      "epoch": 0.7241264064762895,
+      "grad_norm": 0.9280144572257996,
+      "learning_rate": 3.688225792159512e-05,
+      "loss": 0.3645,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7244977533514055,
+      "grad_norm": 0.9158397912979126,
+      "learning_rate": 3.6869028053601965e-05,
+      "loss": 0.3593,
+      "step": 9755
+    },
+    {
+      "epoch": 0.7248691002265216,
+      "grad_norm": 0.9126660823822021,
+      "learning_rate": 3.685579389312612e-05,
+      "loss": 0.3907,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7252404471016376,
+      "grad_norm": 0.7070202827453613,
+      "learning_rate": 3.6842555444953775e-05,
+      "loss": 0.3507,
+      "step": 9765
+    },
+    {
+      "epoch": 0.7256117939767537,
+      "grad_norm": 0.9516451954841614,
+      "learning_rate": 3.682931271387267e-05,
+      "loss": 0.3622,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7259831408518698,
+      "grad_norm": 0.8496583700180054,
+      "learning_rate": 3.68160657046721e-05,
+      "loss": 0.3653,
+      "step": 9775
+    },
+    {
+      "epoch": 0.7263544877269857,
+      "grad_norm": 0.7504469156265259,
+      "learning_rate": 3.680281442214288e-05,
+      "loss": 0.3347,
+      "step": 9780
+    },
+    {
+      "epoch": 0.7267258346021018,
+      "grad_norm": 0.9300464391708374,
+      "learning_rate": 3.678955887107742e-05,
+      "loss": 0.3443,
+      "step": 9785
+    },
+    {
+      "epoch": 0.7270971814772179,
+      "grad_norm": 0.9748815298080444,
+      "learning_rate": 3.677629905626962e-05,
+      "loss": 0.3651,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7274685283523339,
+      "grad_norm": 1.010619878768921,
+      "learning_rate": 3.676303498251496e-05,
+      "loss": 0.3774,
+      "step": 9795
+    },
+    {
+      "epoch": 0.72783987522745,
+      "grad_norm": 0.8692498803138733,
+      "learning_rate": 3.6749766654610445e-05,
+      "loss": 0.3602,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7282112221025661,
+      "grad_norm": 1.0647649765014648,
+      "learning_rate": 3.673649407735461e-05,
+      "loss": 0.3669,
+      "step": 9805
+    },
+    {
+      "epoch": 0.728582568977682,
+      "grad_norm": 0.9087816476821899,
+      "learning_rate": 3.672321725554756e-05,
+      "loss": 0.3725,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7289539158527981,
+      "grad_norm": 0.9249609708786011,
+      "learning_rate": 3.6709936193990885e-05,
+      "loss": 0.3845,
+      "step": 9815
+    },
+    {
+      "epoch": 0.7293252627279141,
+      "grad_norm": 0.8808562159538269,
+      "learning_rate": 3.669665089748775e-05,
+      "loss": 0.3579,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7296966096030302,
+      "grad_norm": 0.9480738639831543,
+      "learning_rate": 3.668336137084283e-05,
+      "loss": 0.3585,
+      "step": 9825
+    },
+    {
+      "epoch": 0.7300679564781463,
+      "grad_norm": 1.0477834939956665,
+      "learning_rate": 3.667006761886235e-05,
+      "loss": 0.3685,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7304393033532622,
+      "grad_norm": 0.8944914937019348,
+      "learning_rate": 3.6656769646354037e-05,
+      "loss": 0.3764,
+      "step": 9835
+    },
+    {
+      "epoch": 0.7308106502283783,
+      "grad_norm": 0.7739643454551697,
+      "learning_rate": 3.664346745812717e-05,
+      "loss": 0.3564,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7311819971034944,
+      "grad_norm": 0.7765239477157593,
+      "learning_rate": 3.663016105899252e-05,
+      "loss": 0.3697,
+      "step": 9845
+    },
+    {
+      "epoch": 0.7315533439786104,
+      "grad_norm": 0.8198454976081848,
+      "learning_rate": 3.6616850453762416e-05,
+      "loss": 0.3508,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7319246908537265,
+      "grad_norm": 0.7845231294631958,
+      "learning_rate": 3.6603535647250684e-05,
+      "loss": 0.3629,
+      "step": 9855
+    },
+    {
+      "epoch": 0.7322960377288426,
+      "grad_norm": 0.9282823204994202,
+      "learning_rate": 3.659021664427269e-05,
+      "loss": 0.3622,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7326673846039585,
+      "grad_norm": 0.9496755599975586,
+      "learning_rate": 3.65768934496453e-05,
+      "loss": 0.3445,
+      "step": 9865
+    },
+    {
+      "epoch": 0.7330387314790746,
+      "grad_norm": 0.7232083082199097,
+      "learning_rate": 3.6563566068186895e-05,
+      "loss": 0.3561,
+      "step": 9870
+    },
+    {
+      "epoch": 0.7334100783541907,
+      "grad_norm": 1.0331436395645142,
+      "learning_rate": 3.6550234504717396e-05,
+      "loss": 0.3706,
+      "step": 9875
+    },
+    {
+      "epoch": 0.7337814252293067,
+      "grad_norm": 0.9413464069366455,
+      "learning_rate": 3.6536898764058196e-05,
+      "loss": 0.3433,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7341527721044228,
+      "grad_norm": 0.9857883453369141,
+      "learning_rate": 3.652355885103223e-05,
+      "loss": 0.3581,
+      "step": 9885
+    },
+    {
+      "epoch": 0.7345241189795388,
+      "grad_norm": 0.8081353306770325,
+      "learning_rate": 3.651021477046393e-05,
+      "loss": 0.3914,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7348954658546548,
+      "grad_norm": 1.0811405181884766,
+      "learning_rate": 3.649686652717925e-05,
+      "loss": 0.3712,
+      "step": 9895
+    },
+    {
+      "epoch": 0.7352668127297709,
+      "grad_norm": 0.816540002822876,
+      "learning_rate": 3.6483514126005626e-05,
+      "loss": 0.3834,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7356381596048869,
+      "grad_norm": 0.7943872213363647,
+      "learning_rate": 3.6470157571772004e-05,
+      "loss": 0.3805,
+      "step": 9905
+    },
+    {
+      "epoch": 0.736009506480003,
+      "grad_norm": 0.9035302400588989,
+      "learning_rate": 3.645679686930883e-05,
+      "loss": 0.3471,
+      "step": 9910
+    },
+    {
+      "epoch": 0.736380853355119,
+      "grad_norm": 0.8763010501861572,
+      "learning_rate": 3.644343202344808e-05,
+      "loss": 0.3741,
+      "step": 9915
+    },
+    {
+      "epoch": 0.736752200230235,
+      "grad_norm": 0.7111252546310425,
+      "learning_rate": 3.64300630390232e-05,
+      "loss": 0.3737,
+      "step": 9920
+    },
+    {
+      "epoch": 0.7371235471053511,
+      "grad_norm": 1.0100241899490356,
+      "learning_rate": 3.641668992086913e-05,
+      "loss": 0.3761,
+      "step": 9925
+    },
+    {
+      "epoch": 0.7374948939804672,
+      "grad_norm": 1.498679757118225,
+      "learning_rate": 3.640331267382231e-05,
+      "loss": 0.3887,
+      "step": 9930
+    },
+    {
+      "epoch": 0.7378662408555832,
+      "grad_norm": 1.0976613759994507,
+      "learning_rate": 3.638993130272068e-05,
+      "loss": 0.357,
+      "step": 9935
+    },
+    {
+      "epoch": 0.7382375877306993,
+      "grad_norm": 0.8685204982757568,
+      "learning_rate": 3.6376545812403675e-05,
+      "loss": 0.3666,
+      "step": 9940
+    },
+    {
+      "epoch": 0.7386089346058153,
+      "grad_norm": 0.7913703322410583,
+      "learning_rate": 3.63631562077122e-05,
+      "loss": 0.3803,
+      "step": 9945
+    },
+    {
+      "epoch": 0.7389802814809313,
+      "grad_norm": 0.8833486437797546,
+      "learning_rate": 3.634976249348867e-05,
+      "loss": 0.3465,
+      "step": 9950
+    },
+    {
+      "epoch": 0.7393516283560474,
+      "grad_norm": 1.0097178220748901,
+      "learning_rate": 3.633636467457697e-05,
+      "loss": 0.3666,
+      "step": 9955
+    },
+    {
+      "epoch": 0.7397229752311635,
+      "grad_norm": 0.7995867133140564,
+      "learning_rate": 3.632296275582249e-05,
+      "loss": 0.3568,
+      "step": 9960
+    },
+    {
+      "epoch": 0.7400943221062795,
+      "grad_norm": 0.8641396760940552,
+      "learning_rate": 3.6309556742072075e-05,
+      "loss": 0.3428,
+      "step": 9965
+    },
+    {
+      "epoch": 0.7404656689813955,
+      "grad_norm": 0.9114529490470886,
+      "learning_rate": 3.629614663817407e-05,
+      "loss": 0.3701,
+      "step": 9970
+    },
+    {
+      "epoch": 0.7408370158565115,
+      "grad_norm": 0.8092938661575317,
+      "learning_rate": 3.628273244897828e-05,
+      "loss": 0.3676,
+      "step": 9975
+    },
+    {
+      "epoch": 0.7412083627316276,
+      "grad_norm": 0.8112850785255432,
+      "learning_rate": 3.6269314179336025e-05,
+      "loss": 0.3839,
+      "step": 9980
+    },
+    {
+      "epoch": 0.7415797096067437,
+      "grad_norm": 0.7114982604980469,
+      "learning_rate": 3.625589183410008e-05,
+      "loss": 0.3701,
+      "step": 9985
+    },
+    {
+      "epoch": 0.7419510564818597,
+      "grad_norm": 1.0398838520050049,
+      "learning_rate": 3.6242465418124657e-05,
+      "loss": 0.3586,
+      "step": 9990
+    },
+    {
+      "epoch": 0.7423224033569757,
+      "grad_norm": 0.7869742512702942,
+      "learning_rate": 3.62290349362655e-05,
+      "loss": 0.3749,
+      "step": 9995
+    },
+    {
+      "epoch": 0.7426937502320918,
+      "grad_norm": 1.0306918621063232,
+      "learning_rate": 3.621560039337979e-05,
+      "loss": 0.3514,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7430650971072078,
+      "grad_norm": 1.0042853355407715,
+      "learning_rate": 3.62021617943262e-05,
+      "loss": 0.3929,
+      "step": 10005
+    },
+    {
+      "epoch": 0.7434364439823239,
+      "grad_norm": 0.8658770322799683,
+      "learning_rate": 3.618871914396483e-05,
+      "loss": 0.3775,
+      "step": 10010
+    },
+    {
+      "epoch": 0.74380779085744,
+      "grad_norm": 0.8654706478118896,
+      "learning_rate": 3.6175272447157294e-05,
+      "loss": 0.3726,
+      "step": 10015
+    },
+    {
+      "epoch": 0.744179137732556,
+      "grad_norm": 1.4344186782836914,
+      "learning_rate": 3.616182170876662e-05,
+      "loss": 0.3462,
+      "step": 10020
+    },
+    {
+      "epoch": 0.744550484607672,
+      "grad_norm": 0.9275874495506287,
+      "learning_rate": 3.614836693365734e-05,
+      "loss": 0.3717,
+      "step": 10025
+    },
+    {
+      "epoch": 0.7449218314827881,
+      "grad_norm": 0.8803759813308716,
+      "learning_rate": 3.6134908126695435e-05,
+      "loss": 0.3425,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7452931783579041,
+      "grad_norm": 0.8547605872154236,
+      "learning_rate": 3.6121445292748314e-05,
+      "loss": 0.3505,
+      "step": 10035
+    },
+    {
+      "epoch": 0.7456645252330202,
+      "grad_norm": 0.7750731110572815,
+      "learning_rate": 3.6107978436684875e-05,
+      "loss": 0.3654,
+      "step": 10040
+    },
+    {
+      "epoch": 0.7460358721081362,
+      "grad_norm": 0.8395034670829773,
+      "learning_rate": 3.6094507563375475e-05,
+      "loss": 0.3687,
+      "step": 10045
+    },
+    {
+      "epoch": 0.7464072189832522,
+      "grad_norm": 0.6695213913917542,
+      "learning_rate": 3.60810326776919e-05,
+      "loss": 0.3758,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7467785658583683,
+      "grad_norm": 0.8508323431015015,
+      "learning_rate": 3.60675537845074e-05,
+      "loss": 0.3948,
+      "step": 10055
+    },
+    {
+      "epoch": 0.7471499127334843,
+      "grad_norm": 1.0589313507080078,
+      "learning_rate": 3.6054070888696664e-05,
+      "loss": 0.3818,
+      "step": 10060
+    },
+    {
+      "epoch": 0.7475212596086004,
+      "grad_norm": 0.6998316049575806,
+      "learning_rate": 3.6040583995135854e-05,
+      "loss": 0.3619,
+      "step": 10065
+    },
+    {
+      "epoch": 0.7478926064837165,
+      "grad_norm": 0.9173612594604492,
+      "learning_rate": 3.602709310870255e-05,
+      "loss": 0.3511,
+      "step": 10070
+    },
+    {
+      "epoch": 0.7482639533588324,
+      "grad_norm": 0.8886153697967529,
+      "learning_rate": 3.601359823427578e-05,
+      "loss": 0.3543,
+      "step": 10075
+    },
+    {
+      "epoch": 0.7486353002339485,
+      "grad_norm": 0.9810105562210083,
+      "learning_rate": 3.600009937673603e-05,
+      "loss": 0.3851,
+      "step": 10080
+    },
+    {
+      "epoch": 0.7490066471090646,
+      "grad_norm": 0.9224115014076233,
+      "learning_rate": 3.598659654096521e-05,
+      "loss": 0.3566,
+      "step": 10085
+    },
+    {
+      "epoch": 0.7493779939841806,
+      "grad_norm": 0.849329948425293,
+      "learning_rate": 3.5973089731846684e-05,
+      "loss": 0.3489,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7497493408592967,
+      "grad_norm": 0.952434241771698,
+      "learning_rate": 3.595957895426524e-05,
+      "loss": 0.3836,
+      "step": 10095
+    },
+    {
+      "epoch": 0.7501206877344128,
+      "grad_norm": 1.2084769010543823,
+      "learning_rate": 3.594606421310711e-05,
+      "loss": 0.3452,
+      "step": 10100
+    },
+    {
+      "epoch": 0.7504920346095287,
+      "grad_norm": 0.6959002017974854,
+      "learning_rate": 3.5932545513259945e-05,
+      "loss": 0.3544,
+      "step": 10105
+    },
+    {
+      "epoch": 0.7508633814846448,
+      "grad_norm": 1.344814658164978,
+      "learning_rate": 3.5919022859612863e-05,
+      "loss": 0.3726,
+      "step": 10110
+    },
+    {
+      "epoch": 0.7512347283597609,
+      "grad_norm": 0.7053912281990051,
+      "learning_rate": 3.590549625705635e-05,
+      "loss": 0.3885,
+      "step": 10115
+    },
+    {
+      "epoch": 0.7516060752348769,
+      "grad_norm": 0.876451849937439,
+      "learning_rate": 3.58919657104824e-05,
+      "loss": 0.3642,
+      "step": 10120
+    },
+    {
+      "epoch": 0.751977422109993,
+      "grad_norm": 0.8557884097099304,
+      "learning_rate": 3.5878431224784347e-05,
+      "loss": 0.3821,
+      "step": 10125
+    },
+    {
+      "epoch": 0.7523487689851089,
+      "grad_norm": 0.8507999181747437,
+      "learning_rate": 3.586489280485703e-05,
+      "loss": 0.3623,
+      "step": 10130
+    },
+    {
+      "epoch": 0.752720115860225,
+      "grad_norm": 0.8015534281730652,
+      "learning_rate": 3.585135045559665e-05,
+      "loss": 0.3374,
+      "step": 10135
+    },
+    {
+      "epoch": 0.7530914627353411,
+      "grad_norm": 2.207430124282837,
+      "learning_rate": 3.583780418190086e-05,
+      "loss": 0.3622,
+      "step": 10140
+    },
+    {
+      "epoch": 0.7534628096104571,
+      "grad_norm": 0.9500503540039062,
+      "learning_rate": 3.5824253988668735e-05,
+      "loss": 0.3705,
+      "step": 10145
+    },
+    {
+      "epoch": 0.7538341564855732,
+      "grad_norm": 0.865135908126831,
+      "learning_rate": 3.5810699880800746e-05,
+      "loss": 0.3576,
+      "step": 10150
+    },
+    {
+      "epoch": 0.7542055033606893,
+      "grad_norm": 0.8576253652572632,
+      "learning_rate": 3.5797141863198803e-05,
+      "loss": 0.3718,
+      "step": 10155
+    },
+    {
+      "epoch": 0.7545768502358052,
+      "grad_norm": 0.9014632105827332,
+      "learning_rate": 3.5783579940766206e-05,
+      "loss": 0.3571,
+      "step": 10160
+    },
+    {
+      "epoch": 0.7549481971109213,
+      "grad_norm": 0.9770298004150391,
+      "learning_rate": 3.577001411840769e-05,
+      "loss": 0.3772,
+      "step": 10165
+    },
+    {
+      "epoch": 0.7553195439860374,
+      "grad_norm": 0.7506251931190491,
+      "learning_rate": 3.575644440102937e-05,
+      "loss": 0.3662,
+      "step": 10170
+    },
+    {
+      "epoch": 0.7556908908611534,
+      "grad_norm": 0.8840534090995789,
+      "learning_rate": 3.574287079353882e-05,
+      "loss": 0.3676,
+      "step": 10175
+    },
+    {
+      "epoch": 0.7560622377362695,
+      "grad_norm": 0.7646118998527527,
+      "learning_rate": 3.572929330084498e-05,
+      "loss": 0.3312,
+      "step": 10180
+    },
+    {
+      "epoch": 0.7564335846113855,
+      "grad_norm": 0.7132297158241272,
+      "learning_rate": 3.57157119278582e-05,
+      "loss": 0.3462,
+      "step": 10185
+    },
+    {
+      "epoch": 0.7568049314865015,
+      "grad_norm": 1.2096678018569946,
+      "learning_rate": 3.570212667949023e-05,
+      "loss": 0.3675,
+      "step": 10190
+    },
+    {
+      "epoch": 0.7571762783616176,
+      "grad_norm": 0.8162215948104858,
+      "learning_rate": 3.5688537560654255e-05,
+      "loss": 0.353,
+      "step": 10195
+    },
+    {
+      "epoch": 0.7575476252367336,
+      "grad_norm": 1.0601283311843872,
+      "learning_rate": 3.5674944576264825e-05,
+      "loss": 0.3683,
+      "step": 10200
+    },
+    {
+      "epoch": 0.7579189721118497,
+      "grad_norm": 1.0272165536880493,
+      "learning_rate": 3.566134773123789e-05,
+      "loss": 0.364,
+      "step": 10205
+    },
+    {
+      "epoch": 0.7582903189869658,
+      "grad_norm": 0.7758386731147766,
+      "learning_rate": 3.564774703049081e-05,
+      "loss": 0.3654,
+      "step": 10210
+    },
+    {
+      "epoch": 0.7586616658620817,
+      "grad_norm": 0.952804684638977,
+      "learning_rate": 3.563414247894234e-05,
+      "loss": 0.3627,
+      "step": 10215
+    },
+    {
+      "epoch": 0.7590330127371978,
+      "grad_norm": 0.8867418169975281,
+      "learning_rate": 3.562053408151262e-05,
+      "loss": 0.3566,
+      "step": 10220
+    },
+    {
+      "epoch": 0.7594043596123139,
+      "grad_norm": 1.061828851699829,
+      "learning_rate": 3.5606921843123176e-05,
+      "loss": 0.3703,
+      "step": 10225
+    },
+    {
+      "epoch": 0.7597757064874299,
+      "grad_norm": 0.7243791818618774,
+      "learning_rate": 3.559330576869694e-05,
+      "loss": 0.366,
+      "step": 10230
+    },
+    {
+      "epoch": 0.760147053362546,
+      "grad_norm": 0.8533843159675598,
+      "learning_rate": 3.5579685863158205e-05,
+      "loss": 0.3852,
+      "step": 10235
+    },
+    {
+      "epoch": 0.760518400237662,
+      "grad_norm": 0.939856767654419,
+      "learning_rate": 3.556606213143269e-05,
+      "loss": 0.376,
+      "step": 10240
+    },
+    {
+      "epoch": 0.760889747112778,
+      "grad_norm": 0.6458791494369507,
+      "learning_rate": 3.5552434578447455e-05,
+      "loss": 0.3485,
+      "step": 10245
+    },
+    {
+      "epoch": 0.7612610939878941,
+      "grad_norm": 0.7483659386634827,
+      "learning_rate": 3.553880320913096e-05,
+      "loss": 0.3484,
+      "step": 10250
+    },
+    {
+      "epoch": 0.7616324408630102,
+      "grad_norm": 0.778672993183136,
+      "learning_rate": 3.5525168028413045e-05,
+      "loss": 0.36,
+      "step": 10255
+    },
+    {
+      "epoch": 0.7620037877381262,
+      "grad_norm": 0.8349382877349854,
+      "learning_rate": 3.5511529041224944e-05,
+      "loss": 0.3571,
+      "step": 10260
+    },
+    {
+      "epoch": 0.7623751346132422,
+      "grad_norm": 3.6593878269195557,
+      "learning_rate": 3.5497886252499254e-05,
+      "loss": 0.3588,
+      "step": 10265
+    },
+    {
+      "epoch": 0.7627464814883583,
+      "grad_norm": 0.8269204497337341,
+      "learning_rate": 3.5484239667169914e-05,
+      "loss": 0.3642,
+      "step": 10270
+    },
+    {
+      "epoch": 0.7631178283634743,
+      "grad_norm": 0.8218369483947754,
+      "learning_rate": 3.54705892901723e-05,
+      "loss": 0.3452,
+      "step": 10275
+    },
+    {
+      "epoch": 0.7634891752385904,
+      "grad_norm": 0.901314377784729,
+      "learning_rate": 3.545693512644311e-05,
+      "loss": 0.3567,
+      "step": 10280
+    },
+    {
+      "epoch": 0.7638605221137064,
+      "grad_norm": 0.8213258981704712,
+      "learning_rate": 3.544327718092044e-05,
+      "loss": 0.3669,
+      "step": 10285
+    },
+    {
+      "epoch": 0.7642318689888224,
+      "grad_norm": 0.8209224343299866,
+      "learning_rate": 3.5429615458543744e-05,
+      "loss": 0.3553,
+      "step": 10290
+    },
+    {
+      "epoch": 0.7646032158639385,
+      "grad_norm": 0.7939985990524292,
+      "learning_rate": 3.541594996425381e-05,
+      "loss": 0.3555,
+      "step": 10295
+    },
+    {
+      "epoch": 0.7649745627390545,
+      "grad_norm": 1.008037805557251,
+      "learning_rate": 3.540228070299286e-05,
+      "loss": 0.38,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7653459096141706,
+      "grad_norm": 0.847432553768158,
+      "learning_rate": 3.538860767970441e-05,
+      "loss": 0.3463,
+      "step": 10305
+    },
+    {
+      "epoch": 0.7657172564892867,
+      "grad_norm": 0.7961945533752441,
+      "learning_rate": 3.537493089933338e-05,
+      "loss": 0.3802,
+      "step": 10310
+    },
+    {
+      "epoch": 0.7660886033644027,
+      "grad_norm": 0.9736582636833191,
+      "learning_rate": 3.536125036682603e-05,
+      "loss": 0.3424,
+      "step": 10315
+    },
+    {
+      "epoch": 0.7664599502395187,
+      "grad_norm": 0.9529137015342712,
+      "learning_rate": 3.534756608712998e-05,
+      "loss": 0.3605,
+      "step": 10320
+    },
+    {
+      "epoch": 0.7668312971146348,
+      "grad_norm": 0.9066136479377747,
+      "learning_rate": 3.533387806519419e-05,
+      "loss": 0.3548,
+      "step": 10325
+    },
+    {
+      "epoch": 0.7672026439897508,
+      "grad_norm": 1.0122712850570679,
+      "learning_rate": 3.5320186305969025e-05,
+      "loss": 0.3543,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7675739908648669,
+      "grad_norm": 0.9614221453666687,
+      "learning_rate": 3.530649081440614e-05,
+      "loss": 0.37,
+      "step": 10335
+    },
+    {
+      "epoch": 0.767945337739983,
+      "grad_norm": 0.8859171271324158,
+      "learning_rate": 3.529279159545856e-05,
+      "loss": 0.3566,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7683166846150989,
+      "grad_norm": 0.7675598859786987,
+      "learning_rate": 3.527908865408069e-05,
+      "loss": 0.3432,
+      "step": 10345
+    },
+    {
+      "epoch": 0.768688031490215,
+      "grad_norm": 0.7545421719551086,
+      "learning_rate": 3.526538199522823e-05,
+      "loss": 0.3661,
+      "step": 10350
+    },
+    {
+      "epoch": 0.769059378365331,
+      "grad_norm": 0.7817773818969727,
+      "learning_rate": 3.525167162385827e-05,
+      "loss": 0.3485,
+      "step": 10355
+    },
+    {
+      "epoch": 0.7694307252404471,
+      "grad_norm": 0.9702377915382385,
+      "learning_rate": 3.523795754492921e-05,
+      "loss": 0.3737,
+      "step": 10360
+    },
+    {
+      "epoch": 0.7698020721155632,
+      "grad_norm": 0.8555597066879272,
+      "learning_rate": 3.522423976340079e-05,
+      "loss": 0.3569,
+      "step": 10365
+    },
+    {
+      "epoch": 0.7701734189906791,
+      "grad_norm": 0.8786563277244568,
+      "learning_rate": 3.521051828423413e-05,
+      "loss": 0.3417,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7705447658657952,
+      "grad_norm": 0.9140200018882751,
+      "learning_rate": 3.519679311239164e-05,
+      "loss": 0.352,
+      "step": 10375
+    },
+    {
+      "epoch": 0.7709161127409113,
+      "grad_norm": 0.8808820843696594,
+      "learning_rate": 3.518306425283709e-05,
+      "loss": 0.3401,
+      "step": 10380
+    },
+    {
+      "epoch": 0.7712874596160273,
+      "grad_norm": 0.671355664730072,
+      "learning_rate": 3.516933171053558e-05,
+      "loss": 0.352,
+      "step": 10385
+    },
+    {
+      "epoch": 0.7716588064911434,
+      "grad_norm": 0.9198750853538513,
+      "learning_rate": 3.515559549045354e-05,
+      "loss": 0.3553,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7720301533662595,
+      "grad_norm": 1.168434500694275,
+      "learning_rate": 3.514185559755873e-05,
+      "loss": 0.3712,
+      "step": 10395
+    },
+    {
+      "epoch": 0.7724015002413754,
+      "grad_norm": 0.8529664278030396,
+      "learning_rate": 3.5128112036820244e-05,
+      "loss": 0.3673,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7727728471164915,
+      "grad_norm": 0.7546283006668091,
+      "learning_rate": 3.511436481320848e-05,
+      "loss": 0.3659,
+      "step": 10405
+    },
+    {
+      "epoch": 0.7731441939916076,
+      "grad_norm": 1.0193859338760376,
+      "learning_rate": 3.5100613931695194e-05,
+      "loss": 0.3635,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7735155408667236,
+      "grad_norm": 0.854232132434845,
+      "learning_rate": 3.5086859397253444e-05,
+      "loss": 0.3754,
+      "step": 10415
+    },
+    {
+      "epoch": 0.7738868877418397,
+      "grad_norm": 1.3594534397125244,
+      "learning_rate": 3.507310121485761e-05,
+      "loss": 0.3633,
+      "step": 10420
+    },
+    {
+      "epoch": 0.7742582346169556,
+      "grad_norm": 0.8981972336769104,
+      "learning_rate": 3.50593393894834e-05,
+      "loss": 0.3307,
+      "step": 10425
+    },
+    {
+      "epoch": 0.7746295814920717,
+      "grad_norm": 0.9904119372367859,
+      "learning_rate": 3.5045573926107825e-05,
+      "loss": 0.3626,
+      "step": 10430
+    },
+    {
+      "epoch": 0.7750009283671878,
+      "grad_norm": 0.9676375985145569,
+      "learning_rate": 3.503180482970925e-05,
+      "loss": 0.3656,
+      "step": 10435
+    },
+    {
+      "epoch": 0.7753722752423038,
+      "grad_norm": 0.6548202037811279,
+      "learning_rate": 3.5018032105267286e-05,
+      "loss": 0.3438,
+      "step": 10440
+    },
+    {
+      "epoch": 0.7757436221174199,
+      "grad_norm": 0.964228093624115,
+      "learning_rate": 3.500425575776293e-05,
+      "loss": 0.3788,
+      "step": 10445
+    },
+    {
+      "epoch": 0.776114968992536,
+      "grad_norm": 0.8802219033241272,
+      "learning_rate": 3.499047579217842e-05,
+      "loss": 0.3342,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7764863158676519,
+      "grad_norm": 0.9338513612747192,
+      "learning_rate": 3.497669221349736e-05,
+      "loss": 0.3521,
+      "step": 10455
+    },
+    {
+      "epoch": 0.776857662742768,
+      "grad_norm": 0.8177879452705383,
+      "learning_rate": 3.496290502670464e-05,
+      "loss": 0.3687,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7772290096178841,
+      "grad_norm": 1.0469391345977783,
+      "learning_rate": 3.4949114236786446e-05,
+      "loss": 0.3767,
+      "step": 10465
+    },
+    {
+      "epoch": 0.7776003564930001,
+      "grad_norm": 0.7534080147743225,
+      "learning_rate": 3.493531984873027e-05,
+      "loss": 0.3473,
+      "step": 10470
+    },
+    {
+      "epoch": 0.7779717033681162,
+      "grad_norm": 0.7704299092292786,
+      "learning_rate": 3.4921521867524916e-05,
+      "loss": 0.3758,
+      "step": 10475
+    },
+    {
+      "epoch": 0.7783430502432322,
+      "grad_norm": 0.7549129128456116,
+      "learning_rate": 3.4907720298160476e-05,
+      "loss": 0.3571,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7787143971183482,
+      "grad_norm": 0.9326322674751282,
+      "learning_rate": 3.4893915145628344e-05,
+      "loss": 0.387,
+      "step": 10485
+    },
+    {
+      "epoch": 0.7790857439934643,
+      "grad_norm": 0.7050992250442505,
+      "learning_rate": 3.488010641492121e-05,
+      "loss": 0.3553,
+      "step": 10490
+    },
+    {
+      "epoch": 0.7794570908685804,
+      "grad_norm": 0.6863691806793213,
+      "learning_rate": 3.486629411103306e-05,
+      "loss": 0.3583,
+      "step": 10495
+    },
+    {
+      "epoch": 0.7798284377436964,
+      "grad_norm": 1.0876379013061523,
+      "learning_rate": 3.485247823895918e-05,
+      "loss": 0.3854,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7801997846188125,
+      "grad_norm": 0.8426483273506165,
+      "learning_rate": 3.483865880369612e-05,
+      "loss": 0.3433,
+      "step": 10505
+    },
+    {
+      "epoch": 0.7805711314939284,
+      "grad_norm": 0.9481852650642395,
+      "learning_rate": 3.4824835810241734e-05,
+      "loss": 0.3431,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7809424783690445,
+      "grad_norm": 0.8300859332084656,
+      "learning_rate": 3.4811009263595177e-05,
+      "loss": 0.3626,
+      "step": 10515
+    },
+    {
+      "epoch": 0.7813138252441606,
+      "grad_norm": 0.8380314707756042,
+      "learning_rate": 3.4797179168756877e-05,
+      "loss": 0.3528,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7816851721192766,
+      "grad_norm": 1.0434871912002563,
+      "learning_rate": 3.4783345530728537e-05,
+      "loss": 0.3725,
+      "step": 10525
+    },
+    {
+      "epoch": 0.7820565189943927,
+      "grad_norm": 0.7483533024787903,
+      "learning_rate": 3.476950835451315e-05,
+      "loss": 0.3664,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7824278658695087,
+      "grad_norm": 0.7801994681358337,
+      "learning_rate": 3.475566764511499e-05,
+      "loss": 0.3779,
+      "step": 10535
+    },
+    {
+      "epoch": 0.7827992127446247,
+      "grad_norm": 0.6180443167686462,
+      "learning_rate": 3.4741823407539606e-05,
+      "loss": 0.3626,
+      "step": 10540
+    },
+    {
+      "epoch": 0.7831705596197408,
+      "grad_norm": 0.8744031190872192,
+      "learning_rate": 3.4727975646793825e-05,
+      "loss": 0.3513,
+      "step": 10545
+    },
+    {
+      "epoch": 0.7835419064948569,
+      "grad_norm": 0.8767526745796204,
+      "learning_rate": 3.4714124367885734e-05,
+      "loss": 0.3712,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7839132533699729,
+      "grad_norm": 0.7508691549301147,
+      "learning_rate": 3.4700269575824726e-05,
+      "loss": 0.349,
+      "step": 10555
+    },
+    {
+      "epoch": 0.784284600245089,
+      "grad_norm": 0.9903153777122498,
+      "learning_rate": 3.468641127562143e-05,
+      "loss": 0.37,
+      "step": 10560
+    },
+    {
+      "epoch": 0.784655947120205,
+      "grad_norm": 0.978623628616333,
+      "learning_rate": 3.4672549472287754e-05,
+      "loss": 0.3566,
+      "step": 10565
+    },
+    {
+      "epoch": 0.785027293995321,
+      "grad_norm": 1.4273784160614014,
+      "learning_rate": 3.4658684170836886e-05,
+      "loss": 0.3743,
+      "step": 10570
+    },
+    {
+      "epoch": 0.7853986408704371,
+      "grad_norm": 1.0664880275726318,
+      "learning_rate": 3.4644815376283265e-05,
+      "loss": 0.3281,
+      "step": 10575
+    },
+    {
+      "epoch": 0.7857699877455531,
+      "grad_norm": 0.9014327526092529,
+      "learning_rate": 3.4630943093642594e-05,
+      "loss": 0.3581,
+      "step": 10580
+    },
+    {
+      "epoch": 0.7861413346206692,
+      "grad_norm": 0.755337119102478,
+      "learning_rate": 3.461706732793184e-05,
+      "loss": 0.3556,
+      "step": 10585
+    },
+    {
+      "epoch": 0.7865126814957852,
+      "grad_norm": 0.9834188222885132,
+      "learning_rate": 3.4603188084169245e-05,
+      "loss": 0.3587,
+      "step": 10590
+    },
+    {
+      "epoch": 0.7868840283709012,
+      "grad_norm": 1.2598179578781128,
+      "learning_rate": 3.4589305367374274e-05,
+      "loss": 0.3572,
+      "step": 10595
+    },
+    {
+      "epoch": 0.7872553752460173,
+      "grad_norm": 1.7740238904953003,
+      "learning_rate": 3.457541918256768e-05,
+      "loss": 0.3806,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7876267221211334,
+      "grad_norm": 0.8469778895378113,
+      "learning_rate": 3.456152953477146e-05,
+      "loss": 0.365,
+      "step": 10605
+    },
+    {
+      "epoch": 0.7879980689962494,
+      "grad_norm": 0.807431697845459,
+      "learning_rate": 3.454763642900886e-05,
+      "loss": 0.3728,
+      "step": 10610
+    },
+    {
+      "epoch": 0.7883694158713654,
+      "grad_norm": 0.8008100986480713,
+      "learning_rate": 3.453373987030438e-05,
+      "loss": 0.3608,
+      "step": 10615
+    },
+    {
+      "epoch": 0.7887407627464815,
+      "grad_norm": 0.8345623016357422,
+      "learning_rate": 3.451983986368376e-05,
+      "loss": 0.3362,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7891121096215975,
+      "grad_norm": 1.0271682739257812,
+      "learning_rate": 3.4505936414174e-05,
+      "loss": 0.3582,
+      "step": 10625
+    },
+    {
+      "epoch": 0.7894834564967136,
+      "grad_norm": 1.199553370475769,
+      "learning_rate": 3.4492029526803346e-05,
+      "loss": 0.3765,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7898548033718297,
+      "grad_norm": 0.9158863425254822,
+      "learning_rate": 3.447811920660127e-05,
+      "loss": 0.3664,
+      "step": 10635
+    },
+    {
+      "epoch": 0.7902261502469456,
+      "grad_norm": 0.8040124773979187,
+      "learning_rate": 3.44642054585985e-05,
+      "loss": 0.3441,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7905974971220617,
+      "grad_norm": 0.8067423105239868,
+      "learning_rate": 3.4450288287827005e-05,
+      "loss": 0.3788,
+      "step": 10645
+    },
+    {
+      "epoch": 0.7909688439971778,
+      "grad_norm": 2.4240920543670654,
+      "learning_rate": 3.4436367699319974e-05,
+      "loss": 0.3677,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7913401908722938,
+      "grad_norm": 1.0094491243362427,
+      "learning_rate": 3.442244369811186e-05,
+      "loss": 0.3594,
+      "step": 10655
+    },
+    {
+      "epoch": 0.7917115377474099,
+      "grad_norm": 0.9552943706512451,
+      "learning_rate": 3.4408516289238324e-05,
+      "loss": 0.36,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7920828846225259,
+      "grad_norm": 1.0734007358551025,
+      "learning_rate": 3.4394585477736276e-05,
+      "loss": 0.3355,
+      "step": 10665
+    },
+    {
+      "epoch": 0.7924542314976419,
+      "grad_norm": 0.8186981678009033,
+      "learning_rate": 3.438065126864385e-05,
+      "loss": 0.3445,
+      "step": 10670
+    },
+    {
+      "epoch": 0.792825578372758,
+      "grad_norm": 0.7257325053215027,
+      "learning_rate": 3.436671366700042e-05,
+      "loss": 0.3586,
+      "step": 10675
+    },
+    {
+      "epoch": 0.793196925247874,
+      "grad_norm": 0.977857768535614,
+      "learning_rate": 3.435277267784656e-05,
+      "loss": 0.3387,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7935682721229901,
+      "grad_norm": 0.7728844285011292,
+      "learning_rate": 3.433882830622409e-05,
+      "loss": 0.3505,
+      "step": 10685
+    },
+    {
+      "epoch": 0.7939396189981062,
+      "grad_norm": 0.8265921473503113,
+      "learning_rate": 3.4324880557176056e-05,
+      "loss": 0.364,
+      "step": 10690
+    },
+    {
+      "epoch": 0.7943109658732221,
+      "grad_norm": 0.9378109574317932,
+      "learning_rate": 3.431092943574671e-05,
+      "loss": 0.3594,
+      "step": 10695
+    },
+    {
+      "epoch": 0.7946823127483382,
+      "grad_norm": 0.8733883500099182,
+      "learning_rate": 3.429697494698154e-05,
+      "loss": 0.3641,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7950536596234543,
+      "grad_norm": 1.0986816883087158,
+      "learning_rate": 3.428301709592724e-05,
+      "loss": 0.3487,
+      "step": 10705
+    },
+    {
+      "epoch": 0.7954250064985703,
+      "grad_norm": 0.8749418258666992,
+      "learning_rate": 3.426905588763172e-05,
+      "loss": 0.3728,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7957963533736864,
+      "grad_norm": 1.04216730594635,
+      "learning_rate": 3.4255091327144127e-05,
+      "loss": 0.3761,
+      "step": 10715
+    },
+    {
+      "epoch": 0.7961677002488025,
+      "grad_norm": 0.9614870548248291,
+      "learning_rate": 3.424112341951478e-05,
+      "loss": 0.3657,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7965390471239184,
+      "grad_norm": 0.7303864359855652,
+      "learning_rate": 3.422715216979524e-05,
+      "loss": 0.3684,
+      "step": 10725
+    },
+    {
+      "epoch": 0.7969103939990345,
+      "grad_norm": 0.6995201706886292,
+      "learning_rate": 3.421317758303826e-05,
+      "loss": 0.3465,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7972817408741505,
+      "grad_norm": 0.9378324747085571,
+      "learning_rate": 3.419919966429782e-05,
+      "loss": 0.3479,
+      "step": 10735
+    },
+    {
+      "epoch": 0.7976530877492666,
+      "grad_norm": 0.9413684010505676,
+      "learning_rate": 3.4185218418629075e-05,
+      "loss": 0.3701,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7980244346243827,
+      "grad_norm": 0.8533821105957031,
+      "learning_rate": 3.417123385108841e-05,
+      "loss": 0.3616,
+      "step": 10745
+    },
+    {
+      "epoch": 0.7983957814994986,
+      "grad_norm": 0.814765453338623,
+      "learning_rate": 3.415724596673341e-05,
+      "loss": 0.3559,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7987671283746147,
+      "grad_norm": 0.8759481310844421,
+      "learning_rate": 3.4143254770622826e-05,
+      "loss": 0.3601,
+      "step": 10755
+    },
+    {
+      "epoch": 0.7991384752497308,
+      "grad_norm": 1.0086429119110107,
+      "learning_rate": 3.4129260267816655e-05,
+      "loss": 0.3771,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7995098221248468,
+      "grad_norm": 1.0653283596038818,
+      "learning_rate": 3.411526246337605e-05,
+      "loss": 0.3615,
+      "step": 10765
+    },
+    {
+      "epoch": 0.7998811689999629,
+      "grad_norm": 0.8258479237556458,
+      "learning_rate": 3.410126136236339e-05,
+      "loss": 0.3426,
+      "step": 10770
+    },
+    {
+      "epoch": 0.800252515875079,
+      "grad_norm": 0.8926478624343872,
+      "learning_rate": 3.4087256969842216e-05,
+      "loss": 0.344,
+      "step": 10775
+    },
+    {
+      "epoch": 0.8006238627501949,
+      "grad_norm": 0.9824770092964172,
+      "learning_rate": 3.4073249290877286e-05,
+      "loss": 0.364,
+      "step": 10780
+    },
+    {
+      "epoch": 0.800995209625311,
+      "grad_norm": 1.6393191814422607,
+      "learning_rate": 3.4059238330534514e-05,
+      "loss": 0.3641,
+      "step": 10785
+    },
+    {
+      "epoch": 0.8013665565004271,
+      "grad_norm": 0.7166811227798462,
+      "learning_rate": 3.404522409388104e-05,
+      "loss": 0.3355,
+      "step": 10790
+    },
+    {
+      "epoch": 0.8017379033755431,
+      "grad_norm": 0.9480621218681335,
+      "learning_rate": 3.4031206585985176e-05,
+      "loss": 0.3473,
+      "step": 10795
+    },
+    {
+      "epoch": 0.8021092502506592,
+      "grad_norm": 1.0220279693603516,
+      "learning_rate": 3.4017185811916394e-05,
+      "loss": 0.3558,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8024805971257752,
+      "grad_norm": 0.7704471945762634,
+      "learning_rate": 3.400316177674538e-05,
+      "loss": 0.3577,
+      "step": 10805
+    },
+    {
+      "epoch": 0.8028519440008912,
+      "grad_norm": 0.6855241656303406,
+      "learning_rate": 3.3989134485543974e-05,
+      "loss": 0.3412,
+      "step": 10810
+    },
+    {
+      "epoch": 0.8032232908760073,
+      "grad_norm": 0.973976194858551,
+      "learning_rate": 3.3975103943385196e-05,
+      "loss": 0.3509,
+      "step": 10815
+    },
+    {
+      "epoch": 0.8035946377511233,
+      "grad_norm": 0.7488546371459961,
+      "learning_rate": 3.396107015534326e-05,
+      "loss": 0.3363,
+      "step": 10820
+    },
+    {
+      "epoch": 0.8039659846262394,
+      "grad_norm": 0.8770419359207153,
+      "learning_rate": 3.3947033126493545e-05,
+      "loss": 0.3624,
+      "step": 10825
+    },
+    {
+      "epoch": 0.8043373315013554,
+      "grad_norm": 0.8518621921539307,
+      "learning_rate": 3.393299286191259e-05,
+      "loss": 0.3495,
+      "step": 10830
+    },
+    {
+      "epoch": 0.8047086783764714,
+      "grad_norm": 1.432323932647705,
+      "learning_rate": 3.391894936667813e-05,
+      "loss": 0.3705,
+      "step": 10835
+    },
+    {
+      "epoch": 0.8050800252515875,
+      "grad_norm": 1.0123671293258667,
+      "learning_rate": 3.390490264586903e-05,
+      "loss": 0.3776,
+      "step": 10840
+    },
+    {
+      "epoch": 0.8054513721267036,
+      "grad_norm": 1.0251203775405884,
+      "learning_rate": 3.3890852704565366e-05,
+      "loss": 0.3833,
+      "step": 10845
+    },
+    {
+      "epoch": 0.8058227190018196,
+      "grad_norm": 0.7330724000930786,
+      "learning_rate": 3.387679954784836e-05,
+      "loss": 0.3492,
+      "step": 10850
+    },
+    {
+      "epoch": 0.8061940658769357,
+      "grad_norm": 1.0848844051361084,
+      "learning_rate": 3.386274318080036e-05,
+      "loss": 0.368,
+      "step": 10855
+    },
+    {
+      "epoch": 0.8065654127520517,
+      "grad_norm": 1.0718929767608643,
+      "learning_rate": 3.3848683608504946e-05,
+      "loss": 0.3608,
+      "step": 10860
+    },
+    {
+      "epoch": 0.8069367596271677,
+      "grad_norm": 1.0632661581039429,
+      "learning_rate": 3.383462083604679e-05,
+      "loss": 0.336,
+      "step": 10865
+    },
+    {
+      "epoch": 0.8073081065022838,
+      "grad_norm": 0.7139385342597961,
+      "learning_rate": 3.3820554868511785e-05,
+      "loss": 0.3594,
+      "step": 10870
+    },
+    {
+      "epoch": 0.8076794533773999,
+      "grad_norm": 0.798427402973175,
+      "learning_rate": 3.380648571098692e-05,
+      "loss": 0.3541,
+      "step": 10875
+    },
+    {
+      "epoch": 0.8080508002525159,
+      "grad_norm": 0.9843793511390686,
+      "learning_rate": 3.379241336856036e-05,
+      "loss": 0.3692,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8084221471276319,
+      "grad_norm": 0.8368058800697327,
+      "learning_rate": 3.377833784632144e-05,
+      "loss": 0.351,
+      "step": 10885
+    },
+    {
+      "epoch": 0.8087934940027479,
+      "grad_norm": 0.5817992687225342,
+      "learning_rate": 3.376425914936063e-05,
+      "loss": 0.3451,
+      "step": 10890
+    },
+    {
+      "epoch": 0.809164840877864,
+      "grad_norm": 0.9955894351005554,
+      "learning_rate": 3.3750177282769554e-05,
+      "loss": 0.3556,
+      "step": 10895
+    },
+    {
+      "epoch": 0.8095361877529801,
+      "grad_norm": 0.924485445022583,
+      "learning_rate": 3.373609225164095e-05,
+      "loss": 0.356,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8099075346280961,
+      "grad_norm": 0.7800727486610413,
+      "learning_rate": 3.3722004061068756e-05,
+      "loss": 0.3646,
+      "step": 10905
+    },
+    {
+      "epoch": 0.8102788815032121,
+      "grad_norm": 0.8341777920722961,
+      "learning_rate": 3.370791271614801e-05,
+      "loss": 0.3562,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8106502283783282,
+      "grad_norm": 0.8623719215393066,
+      "learning_rate": 3.369381822197491e-05,
+      "loss": 0.3579,
+      "step": 10915
+    },
+    {
+      "epoch": 0.8110215752534442,
+      "grad_norm": 0.9438233375549316,
+      "learning_rate": 3.367972058364678e-05,
+      "loss": 0.3598,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8113929221285603,
+      "grad_norm": 0.8673635721206665,
+      "learning_rate": 3.36656198062621e-05,
+      "loss": 0.376,
+      "step": 10925
+    },
+    {
+      "epoch": 0.8117642690036764,
+      "grad_norm": 0.9977436661720276,
+      "learning_rate": 3.365151589492046e-05,
+      "loss": 0.36,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8121356158787923,
+      "grad_norm": 2.1114859580993652,
+      "learning_rate": 3.363740885472261e-05,
+      "loss": 0.3643,
+      "step": 10935
+    },
+    {
+      "epoch": 0.8125069627539084,
+      "grad_norm": 0.8585381507873535,
+      "learning_rate": 3.362329869077042e-05,
+      "loss": 0.3441,
+      "step": 10940
+    },
+    {
+      "epoch": 0.8128783096290245,
+      "grad_norm": 1.0983209609985352,
+      "learning_rate": 3.360918540816687e-05,
+      "loss": 0.3689,
+      "step": 10945
+    },
+    {
+      "epoch": 0.8132496565041405,
+      "grad_norm": 0.9772186875343323,
+      "learning_rate": 3.359506901201611e-05,
+      "loss": 0.3568,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8136210033792566,
+      "grad_norm": 0.8578617572784424,
+      "learning_rate": 3.358094950742338e-05,
+      "loss": 0.3572,
+      "step": 10955
+    },
+    {
+      "epoch": 0.8139923502543726,
+      "grad_norm": 0.8567529916763306,
+      "learning_rate": 3.356682689949506e-05,
+      "loss": 0.3464,
+      "step": 10960
+    },
+    {
+      "epoch": 0.8143636971294886,
+      "grad_norm": 1.8338557481765747,
+      "learning_rate": 3.3552701193338636e-05,
+      "loss": 0.3542,
+      "step": 10965
+    },
+    {
+      "epoch": 0.8147350440046047,
+      "grad_norm": 1.243157148361206,
+      "learning_rate": 3.353857239406275e-05,
+      "loss": 0.3616,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8151063908797207,
+      "grad_norm": 1.1987615823745728,
+      "learning_rate": 3.3524440506777125e-05,
+      "loss": 0.3598,
+      "step": 10975
+    },
+    {
+      "epoch": 0.8154777377548368,
+      "grad_norm": 0.8914827704429626,
+      "learning_rate": 3.3510305536592625e-05,
+      "loss": 0.3743,
+      "step": 10980
+    },
+    {
+      "epoch": 0.8158490846299529,
+      "grad_norm": 0.7617464661598206,
+      "learning_rate": 3.349616748862122e-05,
+      "loss": 0.3407,
+      "step": 10985
+    },
+    {
+      "epoch": 0.8162204315050688,
+      "grad_norm": 0.8160666227340698,
+      "learning_rate": 3.3482026367975974e-05,
+      "loss": 0.35,
+      "step": 10990
+    },
+    {
+      "epoch": 0.8165917783801849,
+      "grad_norm": 0.81715327501297,
+      "learning_rate": 3.346788217977111e-05,
+      "loss": 0.3788,
+      "step": 10995
+    },
+    {
+      "epoch": 0.816963125255301,
+      "grad_norm": 0.7353161573410034,
+      "learning_rate": 3.345373492912191e-05,
+      "loss": 0.3429,
+      "step": 11000
+    },
+    {
+      "epoch": 0.817334472130417,
+      "grad_norm": 0.8322049975395203,
+      "learning_rate": 3.34395846211448e-05,
+      "loss": 0.3609,
+      "step": 11005
+    },
+    {
+      "epoch": 0.8177058190055331,
+      "grad_norm": 0.7142672538757324,
+      "learning_rate": 3.3425431260957276e-05,
+      "loss": 0.3516,
+      "step": 11010
+    },
+    {
+      "epoch": 0.8180771658806492,
+      "grad_norm": 0.8758529424667358,
+      "learning_rate": 3.3411274853677994e-05,
+      "loss": 0.3598,
+      "step": 11015
+    },
+    {
+      "epoch": 0.8184485127557651,
+      "grad_norm": 1.3628263473510742,
+      "learning_rate": 3.3397115404426635e-05,
+      "loss": 0.3707,
+      "step": 11020
+    },
+    {
+      "epoch": 0.8188198596308812,
+      "grad_norm": 1.0877119302749634,
+      "learning_rate": 3.338295291832405e-05,
+      "loss": 0.3616,
+      "step": 11025
+    },
+    {
+      "epoch": 0.8191912065059973,
+      "grad_norm": 0.7904345393180847,
+      "learning_rate": 3.336878740049215e-05,
+      "loss": 0.3661,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8195625533811133,
+      "grad_norm": 0.85951167345047,
+      "learning_rate": 3.335461885605397e-05,
+      "loss": 0.3643,
+      "step": 11035
+    },
+    {
+      "epoch": 0.8199339002562294,
+      "grad_norm": 1.0981451272964478,
+      "learning_rate": 3.334044729013359e-05,
+      "loss": 0.3426,
+      "step": 11040
+    },
+    {
+      "epoch": 0.8203052471313453,
+      "grad_norm": 0.7961001396179199,
+      "learning_rate": 3.3326272707856235e-05,
+      "loss": 0.3458,
+      "step": 11045
+    },
+    {
+      "epoch": 0.8206765940064614,
+      "grad_norm": 0.8995984196662903,
+      "learning_rate": 3.33120951143482e-05,
+      "loss": 0.3545,
+      "step": 11050
+    },
+    {
+      "epoch": 0.8210479408815775,
+      "grad_norm": 0.6880924701690674,
+      "learning_rate": 3.329791451473687e-05,
+      "loss": 0.3446,
+      "step": 11055
+    },
+    {
+      "epoch": 0.8214192877566935,
+      "grad_norm": 0.7782796621322632,
+      "learning_rate": 3.3283730914150716e-05,
+      "loss": 0.3611,
+      "step": 11060
+    },
+    {
+      "epoch": 0.8217906346318096,
+      "grad_norm": 1.008087158203125,
+      "learning_rate": 3.326954431771929e-05,
+      "loss": 0.3652,
+      "step": 11065
+    },
+    {
+      "epoch": 0.8221619815069257,
+      "grad_norm": 0.6520247459411621,
+      "learning_rate": 3.325535473057324e-05,
+      "loss": 0.3516,
+      "step": 11070
+    },
+    {
+      "epoch": 0.8225333283820416,
+      "grad_norm": 0.7982839345932007,
+      "learning_rate": 3.3241162157844284e-05,
+      "loss": 0.3556,
+      "step": 11075
+    },
+    {
+      "epoch": 0.8229046752571577,
+      "grad_norm": 3.1263599395751953,
+      "learning_rate": 3.322696660466523e-05,
+      "loss": 0.3483,
+      "step": 11080
+    },
+    {
+      "epoch": 0.8232760221322738,
+      "grad_norm": 0.9678716659545898,
+      "learning_rate": 3.321276807616995e-05,
+      "loss": 0.3566,
+      "step": 11085
+    },
+    {
+      "epoch": 0.8236473690073898,
+      "grad_norm": 0.8444531559944153,
+      "learning_rate": 3.31985665774934e-05,
+      "loss": 0.3521,
+      "step": 11090
+    },
+    {
+      "epoch": 0.8240187158825059,
+      "grad_norm": 1.0266430377960205,
+      "learning_rate": 3.318436211377162e-05,
+      "loss": 0.3575,
+      "step": 11095
+    },
+    {
+      "epoch": 0.8243900627576219,
+      "grad_norm": 0.8156036734580994,
+      "learning_rate": 3.31701546901417e-05,
+      "loss": 0.3489,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8247614096327379,
+      "grad_norm": 0.7811394333839417,
+      "learning_rate": 3.315594431174183e-05,
+      "loss": 0.3729,
+      "step": 11105
+    },
+    {
+      "epoch": 0.825132756507854,
+      "grad_norm": 0.8737289905548096,
+      "learning_rate": 3.314173098371123e-05,
+      "loss": 0.3448,
+      "step": 11110
+    },
+    {
+      "epoch": 0.82550410338297,
+      "grad_norm": 0.9332969188690186,
+      "learning_rate": 3.312751471119023e-05,
+      "loss": 0.3762,
+      "step": 11115
+    },
+    {
+      "epoch": 0.8258754502580861,
+      "grad_norm": 1.149572730064392,
+      "learning_rate": 3.311329549932018e-05,
+      "loss": 0.3539,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8262467971332021,
+      "grad_norm": 0.8576069474220276,
+      "learning_rate": 3.309907335324354e-05,
+      "loss": 0.3521,
+      "step": 11125
+    },
+    {
+      "epoch": 0.8266181440083181,
+      "grad_norm": 1.0628324747085571,
+      "learning_rate": 3.30848482781038e-05,
+      "loss": 0.3798,
+      "step": 11130
+    },
+    {
+      "epoch": 0.8269894908834342,
+      "grad_norm": 0.8770873546600342,
+      "learning_rate": 3.30706202790455e-05,
+      "loss": 0.3445,
+      "step": 11135
+    },
+    {
+      "epoch": 0.8273608377585503,
+      "grad_norm": 0.8677864074707031,
+      "learning_rate": 3.305638936121428e-05,
+      "loss": 0.3895,
+      "step": 11140
+    },
+    {
+      "epoch": 0.8277321846336663,
+      "grad_norm": 0.8235073089599609,
+      "learning_rate": 3.304215552975679e-05,
+      "loss": 0.3337,
+      "step": 11145
+    },
+    {
+      "epoch": 0.8281035315087824,
+      "grad_norm": 0.9545353055000305,
+      "learning_rate": 3.302791878982077e-05,
+      "loss": 0.3528,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8284748783838984,
+      "grad_norm": 0.8804544806480408,
+      "learning_rate": 3.3013679146554984e-05,
+      "loss": 0.3609,
+      "step": 11155
+    },
+    {
+      "epoch": 0.8288462252590144,
+      "grad_norm": 0.8000383973121643,
+      "learning_rate": 3.299943660510925e-05,
+      "loss": 0.3584,
+      "step": 11160
+    },
+    {
+      "epoch": 0.8292175721341305,
+      "grad_norm": 1.1502083539962769,
+      "learning_rate": 3.2985191170634454e-05,
+      "loss": 0.3894,
+      "step": 11165
+    },
+    {
+      "epoch": 0.8295889190092466,
+      "grad_norm": 0.7234513759613037,
+      "learning_rate": 3.297094284828252e-05,
+      "loss": 0.3795,
+      "step": 11170
+    },
+    {
+      "epoch": 0.8299602658843626,
+      "grad_norm": 0.9550766348838806,
+      "learning_rate": 3.2956691643206406e-05,
+      "loss": 0.3402,
+      "step": 11175
+    },
+    {
+      "epoch": 0.8303316127594786,
+      "grad_norm": 0.9498189687728882,
+      "learning_rate": 3.2942437560560116e-05,
+      "loss": 0.3531,
+      "step": 11180
+    },
+    {
+      "epoch": 0.8307029596345947,
+      "grad_norm": 0.9834123253822327,
+      "learning_rate": 3.2928180605498697e-05,
+      "loss": 0.3597,
+      "step": 11185
+    },
+    {
+      "epoch": 0.8310743065097107,
+      "grad_norm": 0.7517110109329224,
+      "learning_rate": 3.2913920783178235e-05,
+      "loss": 0.35,
+      "step": 11190
+    },
+    {
+      "epoch": 0.8314456533848268,
+      "grad_norm": 0.841060221195221,
+      "learning_rate": 3.289965809875587e-05,
+      "loss": 0.3565,
+      "step": 11195
+    },
+    {
+      "epoch": 0.8318170002599428,
+      "grad_norm": 0.7667614817619324,
+      "learning_rate": 3.2885392557389726e-05,
+      "loss": 0.346,
+      "step": 11200
+    },
+    {
+      "epoch": 0.8321883471350588,
+      "grad_norm": 3.239955186843872,
+      "learning_rate": 3.2871124164239035e-05,
+      "loss": 0.359,
+      "step": 11205
+    },
+    {
+      "epoch": 0.8325596940101749,
+      "grad_norm": 1.0403164625167847,
+      "learning_rate": 3.2856852924463986e-05,
+      "loss": 0.3428,
+      "step": 11210
+    },
+    {
+      "epoch": 0.8329310408852909,
+      "grad_norm": 0.7579136490821838,
+      "learning_rate": 3.2842578843225855e-05,
+      "loss": 0.3331,
+      "step": 11215
+    },
+    {
+      "epoch": 0.833302387760407,
+      "grad_norm": 0.9784735441207886,
+      "learning_rate": 3.282830192568691e-05,
+      "loss": 0.3483,
+      "step": 11220
+    },
+    {
+      "epoch": 0.8336737346355231,
+      "grad_norm": 1.0152697563171387,
+      "learning_rate": 3.281402217701045e-05,
+      "loss": 0.3629,
+      "step": 11225
+    },
+    {
+      "epoch": 0.834045081510639,
+      "grad_norm": 0.6538757085800171,
+      "learning_rate": 3.279973960236083e-05,
+      "loss": 0.3553,
+      "step": 11230
+    },
+    {
+      "epoch": 0.8344164283857551,
+      "grad_norm": 1.0720804929733276,
+      "learning_rate": 3.2785454206903377e-05,
+      "loss": 0.348,
+      "step": 11235
+    },
+    {
+      "epoch": 0.8347877752608712,
+      "grad_norm": 0.8411654829978943,
+      "learning_rate": 3.277116599580448e-05,
+      "loss": 0.3688,
+      "step": 11240
+    },
+    {
+      "epoch": 0.8351591221359872,
+      "grad_norm": 0.9355876445770264,
+      "learning_rate": 3.2756874974231506e-05,
+      "loss": 0.3592,
+      "step": 11245
+    },
+    {
+      "epoch": 0.8355304690111033,
+      "grad_norm": 0.9845572710037231,
+      "learning_rate": 3.274258114735288e-05,
+      "loss": 0.3513,
+      "step": 11250
+    },
+    {
+      "epoch": 0.8359018158862194,
+      "grad_norm": 0.9073336720466614,
+      "learning_rate": 3.2728284520338024e-05,
+      "loss": 0.3639,
+      "step": 11255
+    },
+    {
+      "epoch": 0.8362731627613353,
+      "grad_norm": 0.9223979115486145,
+      "learning_rate": 3.271398509835737e-05,
+      "loss": 0.3617,
+      "step": 11260
+    },
+    {
+      "epoch": 0.8366445096364514,
+      "grad_norm": 0.7728599309921265,
+      "learning_rate": 3.269968288658236e-05,
+      "loss": 0.3576,
+      "step": 11265
+    },
+    {
+      "epoch": 0.8370158565115674,
+      "grad_norm": 0.7334360480308533,
+      "learning_rate": 3.2685377890185435e-05,
+      "loss": 0.3591,
+      "step": 11270
+    },
+    {
+      "epoch": 0.8373872033866835,
+      "grad_norm": 0.9434806108474731,
+      "learning_rate": 3.267107011434009e-05,
+      "loss": 0.3894,
+      "step": 11275
+    },
+    {
+      "epoch": 0.8377585502617996,
+      "grad_norm": 0.7240009307861328,
+      "learning_rate": 3.2656759564220745e-05,
+      "loss": 0.3637,
+      "step": 11280
+    },
+    {
+      "epoch": 0.8381298971369155,
+      "grad_norm": 1.0840269327163696,
+      "learning_rate": 3.264244624500291e-05,
+      "loss": 0.3681,
+      "step": 11285
+    },
+    {
+      "epoch": 0.8385012440120316,
+      "grad_norm": 0.882512629032135,
+      "learning_rate": 3.2628130161863025e-05,
+      "loss": 0.3489,
+      "step": 11290
+    },
+    {
+      "epoch": 0.8388725908871477,
+      "grad_norm": 0.8570889234542847,
+      "learning_rate": 3.261381131997859e-05,
+      "loss": 0.3558,
+      "step": 11295
+    },
+    {
+      "epoch": 0.8392439377622637,
+      "grad_norm": 0.7707122564315796,
+      "learning_rate": 3.2599489724528044e-05,
+      "loss": 0.3436,
+      "step": 11300
+    },
+    {
+      "epoch": 0.8396152846373798,
+      "grad_norm": 0.9872932434082031,
+      "learning_rate": 3.2585165380690866e-05,
+      "loss": 0.3636,
+      "step": 11305
+    },
+    {
+      "epoch": 0.8399866315124959,
+      "grad_norm": 0.8050171136856079,
+      "learning_rate": 3.257083829364751e-05,
+      "loss": 0.3381,
+      "step": 11310
+    },
+    {
+      "epoch": 0.8403579783876118,
+      "grad_norm": 0.7836553454399109,
+      "learning_rate": 3.255650846857943e-05,
+      "loss": 0.3601,
+      "step": 11315
+    },
+    {
+      "epoch": 0.8407293252627279,
+      "grad_norm": 1.6859378814697266,
+      "learning_rate": 3.254217591066906e-05,
+      "loss": 0.3414,
+      "step": 11320
+    },
+    {
+      "epoch": 0.841100672137844,
+      "grad_norm": 0.9329738616943359,
+      "learning_rate": 3.252784062509984e-05,
+      "loss": 0.3492,
+      "step": 11325
+    },
+    {
+      "epoch": 0.84147201901296,
+      "grad_norm": 1.1672630310058594,
+      "learning_rate": 3.251350261705617e-05,
+      "loss": 0.3535,
+      "step": 11330
+    },
+    {
+      "epoch": 0.8418433658880761,
+      "grad_norm": 0.9215666055679321,
+      "learning_rate": 3.249916189172345e-05,
+      "loss": 0.3356,
+      "step": 11335
+    },
+    {
+      "epoch": 0.8422147127631922,
+      "grad_norm": 0.7470521330833435,
+      "learning_rate": 3.2484818454288076e-05,
+      "loss": 0.3672,
+      "step": 11340
+    },
+    {
+      "epoch": 0.8425860596383081,
+      "grad_norm": 1.0789822340011597,
+      "learning_rate": 3.24704723099374e-05,
+      "loss": 0.39,
+      "step": 11345
+    },
+    {
+      "epoch": 0.8429574065134242,
+      "grad_norm": 0.7468339204788208,
+      "learning_rate": 3.245612346385977e-05,
+      "loss": 0.3556,
+      "step": 11350
+    },
+    {
+      "epoch": 0.8433287533885402,
+      "grad_norm": 0.8107490539550781,
+      "learning_rate": 3.2441771921244505e-05,
+      "loss": 0.3363,
+      "step": 11355
+    },
+    {
+      "epoch": 0.8437001002636563,
+      "grad_norm": 0.977986752986908,
+      "learning_rate": 3.242741768728188e-05,
+      "loss": 0.3455,
+      "step": 11360
+    },
+    {
+      "epoch": 0.8440714471387724,
+      "grad_norm": 0.829646110534668,
+      "learning_rate": 3.24130607671632e-05,
+      "loss": 0.3369,
+      "step": 11365
+    },
+    {
+      "epoch": 0.8444427940138883,
+      "grad_norm": 0.96271812915802,
+      "learning_rate": 3.239870116608067e-05,
+      "loss": 0.3568,
+      "step": 11370
+    },
+    {
+      "epoch": 0.8448141408890044,
+      "grad_norm": 0.8084895014762878,
+      "learning_rate": 3.238433888922751e-05,
+      "loss": 0.3448,
+      "step": 11375
+    },
+    {
+      "epoch": 0.8451854877641205,
+      "grad_norm": 0.7734755873680115,
+      "learning_rate": 3.2369973941797905e-05,
+      "loss": 0.3631,
+      "step": 11380
+    },
+    {
+      "epoch": 0.8455568346392365,
+      "grad_norm": 0.798600971698761,
+      "learning_rate": 3.235560632898699e-05,
+      "loss": 0.3513,
+      "step": 11385
+    },
+    {
+      "epoch": 0.8459281815143526,
+      "grad_norm": 0.9249498248100281,
+      "learning_rate": 3.234123605599086e-05,
+      "loss": 0.3686,
+      "step": 11390
+    },
+    {
+      "epoch": 0.8462995283894686,
+      "grad_norm": 0.7547146081924438,
+      "learning_rate": 3.23268631280066e-05,
+      "loss": 0.346,
+      "step": 11395
+    },
+    {
+      "epoch": 0.8466708752645846,
+      "grad_norm": 0.8783119916915894,
+      "learning_rate": 3.231248755023224e-05,
+      "loss": 0.3561,
+      "step": 11400
+    },
+    {
+      "epoch": 0.8470422221397007,
+      "grad_norm": 0.72531658411026,
+      "learning_rate": 3.229810932786675e-05,
+      "loss": 0.3535,
+      "step": 11405
+    },
+    {
+      "epoch": 0.8474135690148168,
+      "grad_norm": 0.8411027193069458,
+      "learning_rate": 3.228372846611009e-05,
+      "loss": 0.3365,
+      "step": 11410
+    },
+    {
+      "epoch": 0.8477849158899328,
+      "grad_norm": 1.3037042617797852,
+      "learning_rate": 3.226934497016314e-05,
+      "loss": 0.3557,
+      "step": 11415
+    },
+    {
+      "epoch": 0.8481562627650489,
+      "grad_norm": 0.9049678444862366,
+      "learning_rate": 3.2254958845227765e-05,
+      "loss": 0.3461,
+      "step": 11420
+    },
+    {
+      "epoch": 0.8485276096401648,
+      "grad_norm": 0.9194682836532593,
+      "learning_rate": 3.224057009650676e-05,
+      "loss": 0.3731,
+      "step": 11425
+    },
+    {
+      "epoch": 0.8488989565152809,
+      "grad_norm": 0.8797911405563354,
+      "learning_rate": 3.222617872920389e-05,
+      "loss": 0.3695,
+      "step": 11430
+    },
+    {
+      "epoch": 0.849270303390397,
+      "grad_norm": 0.9936988353729248,
+      "learning_rate": 3.2211784748523813e-05,
+      "loss": 0.354,
+      "step": 11435
+    },
+    {
+      "epoch": 0.849641650265513,
+      "grad_norm": 0.8593586683273315,
+      "learning_rate": 3.219738815967222e-05,
+      "loss": 0.3863,
+      "step": 11440
+    },
+    {
+      "epoch": 0.8500129971406291,
+      "grad_norm": 1.028411865234375,
+      "learning_rate": 3.218298896785566e-05,
+      "loss": 0.3556,
+      "step": 11445
+    },
+    {
+      "epoch": 0.8503843440157451,
+      "grad_norm": 1.2228703498840332,
+      "learning_rate": 3.2168587178281686e-05,
+      "loss": 0.344,
+      "step": 11450
+    },
+    {
+      "epoch": 0.8507556908908611,
+      "grad_norm": 0.8291895985603333,
+      "learning_rate": 3.215418279615874e-05,
+      "loss": 0.3277,
+      "step": 11455
+    },
+    {
+      "epoch": 0.8511270377659772,
+      "grad_norm": 0.8701698780059814,
+      "learning_rate": 3.2139775826696233e-05,
+      "loss": 0.3926,
+      "step": 11460
+    },
+    {
+      "epoch": 0.8514983846410933,
+      "grad_norm": 0.8717911243438721,
+      "learning_rate": 3.212536627510451e-05,
+      "loss": 0.3421,
+      "step": 11465
+    },
+    {
+      "epoch": 0.8518697315162093,
+      "grad_norm": 1.0229847431182861,
+      "learning_rate": 3.211095414659485e-05,
+      "loss": 0.3591,
+      "step": 11470
+    },
+    {
+      "epoch": 0.8522410783913253,
+      "grad_norm": 0.666601300239563,
+      "learning_rate": 3.209653944637945e-05,
+      "loss": 0.3442,
+      "step": 11475
+    },
+    {
+      "epoch": 0.8526124252664414,
+      "grad_norm": 0.7465684413909912,
+      "learning_rate": 3.2082122179671445e-05,
+      "loss": 0.3585,
+      "step": 11480
+    },
+    {
+      "epoch": 0.8529837721415574,
+      "grad_norm": 0.9072179794311523,
+      "learning_rate": 3.2067702351684905e-05,
+      "loss": 0.3356,
+      "step": 11485
+    },
+    {
+      "epoch": 0.8533551190166735,
+      "grad_norm": 0.8667060136795044,
+      "learning_rate": 3.205327996763481e-05,
+      "loss": 0.355,
+      "step": 11490
+    },
+    {
+      "epoch": 0.8537264658917895,
+      "grad_norm": 0.9172279238700867,
+      "learning_rate": 3.203885503273708e-05,
+      "loss": 0.3482,
+      "step": 11495
+    },
+    {
+      "epoch": 0.8540978127669056,
+      "grad_norm": 0.8449558615684509,
+      "learning_rate": 3.2024427552208546e-05,
+      "loss": 0.362,
+      "step": 11500
+    },
+    {
+      "epoch": 0.8544691596420216,
+      "grad_norm": 0.8456041216850281,
+      "learning_rate": 3.2009997531266985e-05,
+      "loss": 0.3489,
+      "step": 11505
+    },
+    {
+      "epoch": 0.8548405065171376,
+      "grad_norm": 0.8721354007720947,
+      "learning_rate": 3.199556497513105e-05,
+      "loss": 0.348,
+      "step": 11510
+    },
+    {
+      "epoch": 0.8552118533922537,
+      "grad_norm": 0.7469504475593567,
+      "learning_rate": 3.198112988902035e-05,
+      "loss": 0.3365,
+      "step": 11515
+    },
+    {
+      "epoch": 0.8555832002673698,
+      "grad_norm": 0.8014267086982727,
+      "learning_rate": 3.1966692278155386e-05,
+      "loss": 0.3464,
+      "step": 11520
+    },
+    {
+      "epoch": 0.8559545471424858,
+      "grad_norm": 0.6369099617004395,
+      "learning_rate": 3.1952252147757563e-05,
+      "loss": 0.3464,
+      "step": 11525
+    },
+    {
+      "epoch": 0.8563258940176018,
+      "grad_norm": 0.9981203079223633,
+      "learning_rate": 3.193780950304925e-05,
+      "loss": 0.3571,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8566972408927179,
+      "grad_norm": 0.6954818964004517,
+      "learning_rate": 3.192336434925367e-05,
+      "loss": 0.3616,
+      "step": 11535
+    },
+    {
+      "epoch": 0.8570685877678339,
+      "grad_norm": 1.202691674232483,
+      "learning_rate": 3.190891669159497e-05,
+      "loss": 0.3253,
+      "step": 11540
+    },
+    {
+      "epoch": 0.85743993464295,
+      "grad_norm": 0.6270232200622559,
+      "learning_rate": 3.189446653529819e-05,
+      "loss": 0.3458,
+      "step": 11545
+    },
+    {
+      "epoch": 0.8578112815180661,
+      "grad_norm": 1.0812908411026,
+      "learning_rate": 3.188001388558932e-05,
+      "loss": 0.3713,
+      "step": 11550
+    },
+    {
+      "epoch": 0.858182628393182,
+      "grad_norm": 1.0152004957199097,
+      "learning_rate": 3.1865558747695194e-05,
+      "loss": 0.3417,
+      "step": 11555
+    },
+    {
+      "epoch": 0.8585539752682981,
+      "grad_norm": 0.9745574593544006,
+      "learning_rate": 3.1851101126843584e-05,
+      "loss": 0.3675,
+      "step": 11560
+    },
+    {
+      "epoch": 0.8589253221434142,
+      "grad_norm": 0.8742450475692749,
+      "learning_rate": 3.183664102826316e-05,
+      "loss": 0.3524,
+      "step": 11565
+    },
+    {
+      "epoch": 0.8592966690185302,
+      "grad_norm": 0.8508179783821106,
+      "learning_rate": 3.182217845718344e-05,
+      "loss": 0.3647,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8596680158936463,
+      "grad_norm": 0.8023908734321594,
+      "learning_rate": 3.180771341883491e-05,
+      "loss": 0.3531,
+      "step": 11575
+    },
+    {
+      "epoch": 0.8600393627687622,
+      "grad_norm": 1.2750799655914307,
+      "learning_rate": 3.1793245918448886e-05,
+      "loss": 0.3619,
+      "step": 11580
+    },
+    {
+      "epoch": 0.8604107096438783,
+      "grad_norm": 0.8956974744796753,
+      "learning_rate": 3.177877596125761e-05,
+      "loss": 0.3629,
+      "step": 11585
+    },
+    {
+      "epoch": 0.8607820565189944,
+      "grad_norm": 0.8610901236534119,
+      "learning_rate": 3.17643035524942e-05,
+      "loss": 0.348,
+      "step": 11590
+    },
+    {
+      "epoch": 0.8611534033941104,
+      "grad_norm": 0.8622403144836426,
+      "learning_rate": 3.1749828697392667e-05,
+      "loss": 0.3444,
+      "step": 11595
+    },
+    {
+      "epoch": 0.8615247502692265,
+      "grad_norm": 0.9162015318870544,
+      "learning_rate": 3.173535140118789e-05,
+      "loss": 0.362,
+      "step": 11600
+    },
+    {
+      "epoch": 0.8618960971443426,
+      "grad_norm": 0.7178031802177429,
+      "learning_rate": 3.1720871669115646e-05,
+      "loss": 0.3558,
+      "step": 11605
+    },
+    {
+      "epoch": 0.8622674440194585,
+      "grad_norm": 0.9224133491516113,
+      "learning_rate": 3.1706389506412595e-05,
+      "loss": 0.3828,
+      "step": 11610
+    },
+    {
+      "epoch": 0.8626387908945746,
+      "grad_norm": 0.7860289812088013,
+      "learning_rate": 3.1691904918316263e-05,
+      "loss": 0.3344,
+      "step": 11615
+    },
+    {
+      "epoch": 0.8630101377696907,
+      "grad_norm": 0.9802098274230957,
+      "learning_rate": 3.167741791006507e-05,
+      "loss": 0.3834,
+      "step": 11620
+    },
+    {
+      "epoch": 0.8633814846448067,
+      "grad_norm": 0.9482443332672119,
+      "learning_rate": 3.166292848689829e-05,
+      "loss": 0.3722,
+      "step": 11625
+    },
+    {
+      "epoch": 0.8637528315199228,
+      "grad_norm": 1.1732617616653442,
+      "learning_rate": 3.16484366540561e-05,
+      "loss": 0.3598,
+      "step": 11630
+    },
+    {
+      "epoch": 0.8641241783950389,
+      "grad_norm": 1.0246295928955078,
+      "learning_rate": 3.16339424167795e-05,
+      "loss": 0.3534,
+      "step": 11635
+    },
+    {
+      "epoch": 0.8644955252701548,
+      "grad_norm": 0.8778313994407654,
+      "learning_rate": 3.1619445780310435e-05,
+      "loss": 0.3376,
+      "step": 11640
+    },
+    {
+      "epoch": 0.8648668721452709,
+      "grad_norm": 1.0736422538757324,
+      "learning_rate": 3.1604946749891636e-05,
+      "loss": 0.3479,
+      "step": 11645
+    },
+    {
+      "epoch": 0.8652382190203869,
+      "grad_norm": 0.9888156056404114,
+      "learning_rate": 3.159044533076675e-05,
+      "loss": 0.3558,
+      "step": 11650
+    },
+    {
+      "epoch": 0.865609565895503,
+      "grad_norm": 0.7955016493797302,
+      "learning_rate": 3.157594152818027e-05,
+      "loss": 0.3526,
+      "step": 11655
+    },
+    {
+      "epoch": 0.8659809127706191,
+      "grad_norm": 0.8220962882041931,
+      "learning_rate": 3.1561435347377564e-05,
+      "loss": 0.3507,
+      "step": 11660
+    },
+    {
+      "epoch": 0.866352259645735,
+      "grad_norm": 0.8022737503051758,
+      "learning_rate": 3.154692679360485e-05,
+      "loss": 0.3493,
+      "step": 11665
+    },
+    {
+      "epoch": 0.8667236065208511,
+      "grad_norm": 1.158897042274475,
+      "learning_rate": 3.1532415872109206e-05,
+      "loss": 0.3636,
+      "step": 11670
+    },
+    {
+      "epoch": 0.8670949533959672,
+      "grad_norm": 0.8414539098739624,
+      "learning_rate": 3.151790258813855e-05,
+      "loss": 0.3466,
+      "step": 11675
+    },
+    {
+      "epoch": 0.8674663002710832,
+      "grad_norm": 0.7647371888160706,
+      "learning_rate": 3.15033869469417e-05,
+      "loss": 0.3444,
+      "step": 11680
+    },
+    {
+      "epoch": 0.8678376471461993,
+      "grad_norm": 0.8496689796447754,
+      "learning_rate": 3.1488868953768276e-05,
+      "loss": 0.3827,
+      "step": 11685
+    },
+    {
+      "epoch": 0.8682089940213154,
+      "grad_norm": 0.9408618211746216,
+      "learning_rate": 3.1474348613868774e-05,
+      "loss": 0.3402,
+      "step": 11690
+    },
+    {
+      "epoch": 0.8685803408964313,
+      "grad_norm": 0.9130736589431763,
+      "learning_rate": 3.1459825932494535e-05,
+      "loss": 0.3402,
+      "step": 11695
+    },
+    {
+      "epoch": 0.8689516877715474,
+      "grad_norm": 0.8409910798072815,
+      "learning_rate": 3.1445300914897744e-05,
+      "loss": 0.3453,
+      "step": 11700
+    },
+    {
+      "epoch": 0.8693230346466635,
+      "grad_norm": 0.8987837433815002,
+      "learning_rate": 3.1430773566331436e-05,
+      "loss": 0.3364,
+      "step": 11705
+    },
+    {
+      "epoch": 0.8696943815217795,
+      "grad_norm": 2.895118474960327,
+      "learning_rate": 3.14162438920495e-05,
+      "loss": 0.3431,
+      "step": 11710
+    },
+    {
+      "epoch": 0.8700657283968956,
+      "grad_norm": 0.8010272979736328,
+      "learning_rate": 3.140171189730662e-05,
+      "loss": 0.3428,
+      "step": 11715
+    },
+    {
+      "epoch": 0.8704370752720116,
+      "grad_norm": 1.2577528953552246,
+      "learning_rate": 3.138717758735839e-05,
+      "loss": 0.3701,
+      "step": 11720
+    },
+    {
+      "epoch": 0.8708084221471276,
+      "grad_norm": 0.8167540431022644,
+      "learning_rate": 3.137264096746117e-05,
+      "loss": 0.36,
+      "step": 11725
+    },
+    {
+      "epoch": 0.8711797690222437,
+      "grad_norm": 0.7406752705574036,
+      "learning_rate": 3.135810204287221e-05,
+      "loss": 0.3473,
+      "step": 11730
+    },
+    {
+      "epoch": 0.8715511158973597,
+      "grad_norm": 0.8588741421699524,
+      "learning_rate": 3.134356081884956e-05,
+      "loss": 0.3361,
+      "step": 11735
+    },
+    {
+      "epoch": 0.8719224627724758,
+      "grad_norm": 0.8938350677490234,
+      "learning_rate": 3.132901730065211e-05,
+      "loss": 0.3431,
+      "step": 11740
+    },
+    {
+      "epoch": 0.8722938096475918,
+      "grad_norm": 0.9436230063438416,
+      "learning_rate": 3.13144714935396e-05,
+      "loss": 0.357,
+      "step": 11745
+    },
+    {
+      "epoch": 0.8726651565227078,
+      "grad_norm": 1.0733808279037476,
+      "learning_rate": 3.129992340277256e-05,
+      "loss": 0.3459,
+      "step": 11750
+    },
+    {
+      "epoch": 0.8730365033978239,
+      "grad_norm": 0.7322360873222351,
+      "learning_rate": 3.1285373033612385e-05,
+      "loss": 0.3645,
+      "step": 11755
+    },
+    {
+      "epoch": 0.87340785027294,
+      "grad_norm": 1.1299097537994385,
+      "learning_rate": 3.127082039132126e-05,
+      "loss": 0.3319,
+      "step": 11760
+    },
+    {
+      "epoch": 0.873779197148056,
+      "grad_norm": 0.7460024952888489,
+      "learning_rate": 3.125626548116222e-05,
+      "loss": 0.3266,
+      "step": 11765
+    },
+    {
+      "epoch": 0.874150544023172,
+      "grad_norm": 0.807644248008728,
+      "learning_rate": 3.124170830839911e-05,
+      "loss": 0.3627,
+      "step": 11770
+    },
+    {
+      "epoch": 0.8745218908982881,
+      "grad_norm": 0.7130464315414429,
+      "learning_rate": 3.1227148878296584e-05,
+      "loss": 0.3616,
+      "step": 11775
+    },
+    {
+      "epoch": 0.8748932377734041,
+      "grad_norm": 0.840950608253479,
+      "learning_rate": 3.12125871961201e-05,
+      "loss": 0.3755,
+      "step": 11780
+    },
+    {
+      "epoch": 0.8752645846485202,
+      "grad_norm": 0.8666706681251526,
+      "learning_rate": 3.1198023267135995e-05,
+      "loss": 0.3671,
+      "step": 11785
+    },
+    {
+      "epoch": 0.8756359315236363,
+      "grad_norm": 0.848078727722168,
+      "learning_rate": 3.118345709661134e-05,
+      "loss": 0.3648,
+      "step": 11790
+    },
+    {
+      "epoch": 0.8760072783987523,
+      "grad_norm": 0.7801305651664734,
+      "learning_rate": 3.116888868981405e-05,
+      "loss": 0.3276,
+      "step": 11795
+    },
+    {
+      "epoch": 0.8763786252738683,
+      "grad_norm": 0.855327844619751,
+      "learning_rate": 3.115431805201287e-05,
+      "loss": 0.361,
+      "step": 11800
+    },
+    {
+      "epoch": 0.8767499721489843,
+      "grad_norm": 0.7050216794013977,
+      "learning_rate": 3.113974518847731e-05,
+      "loss": 0.3358,
+      "step": 11805
+    },
+    {
+      "epoch": 0.8771213190241004,
+      "grad_norm": 0.7866954207420349,
+      "learning_rate": 3.112517010447773e-05,
+      "loss": 0.3489,
+      "step": 11810
+    },
+    {
+      "epoch": 0.8774926658992165,
+      "grad_norm": 1.1569349765777588,
+      "learning_rate": 3.1110592805285246e-05,
+      "loss": 0.3434,
+      "step": 11815
+    },
+    {
+      "epoch": 0.8778640127743325,
+      "grad_norm": 1.1610534191131592,
+      "learning_rate": 3.109601329617181e-05,
+      "loss": 0.3598,
+      "step": 11820
+    },
+    {
+      "epoch": 0.8782353596494485,
+      "grad_norm": 1.4828381538391113,
+      "learning_rate": 3.1081431582410146e-05,
+      "loss": 0.3625,
+      "step": 11825
+    },
+    {
+      "epoch": 0.8786067065245646,
+      "grad_norm": 0.9669677019119263,
+      "learning_rate": 3.106684766927382e-05,
+      "loss": 0.386,
+      "step": 11830
+    },
+    {
+      "epoch": 0.8789780533996806,
+      "grad_norm": 0.7989631295204163,
+      "learning_rate": 3.105226156203713e-05,
+      "loss": 0.3315,
+      "step": 11835
+    },
+    {
+      "epoch": 0.8793494002747967,
+      "grad_norm": 0.7316701412200928,
+      "learning_rate": 3.1037673265975226e-05,
+      "loss": 0.3785,
+      "step": 11840
+    },
+    {
+      "epoch": 0.8797207471499128,
+      "grad_norm": 2.4976911544799805,
+      "learning_rate": 3.102308278636402e-05,
+      "loss": 0.3629,
+      "step": 11845
+    },
+    {
+      "epoch": 0.8800920940250287,
+      "grad_norm": 0.879723846912384,
+      "learning_rate": 3.1008490128480206e-05,
+      "loss": 0.3411,
+      "step": 11850
+    },
+    {
+      "epoch": 0.8804634409001448,
+      "grad_norm": 0.8403365612030029,
+      "learning_rate": 3.0993895297601293e-05,
+      "loss": 0.3689,
+      "step": 11855
+    },
+    {
+      "epoch": 0.8808347877752609,
+      "grad_norm": 0.8075715899467468,
+      "learning_rate": 3.097929829900556e-05,
+      "loss": 0.3557,
+      "step": 11860
+    },
+    {
+      "epoch": 0.8812061346503769,
+      "grad_norm": 1.033172845840454,
+      "learning_rate": 3.096469913797206e-05,
+      "loss": 0.3461,
+      "step": 11865
+    },
+    {
+      "epoch": 0.881577481525493,
+      "grad_norm": 0.8531133532524109,
+      "learning_rate": 3.095009781978066e-05,
+      "loss": 0.3493,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8819488284006091,
+      "grad_norm": 0.8683094382286072,
+      "learning_rate": 3.0935494349711964e-05,
+      "loss": 0.3624,
+      "step": 11875
+    },
+    {
+      "epoch": 0.882320175275725,
+      "grad_norm": 1.5701484680175781,
+      "learning_rate": 3.092088873304739e-05,
+      "loss": 0.3572,
+      "step": 11880
+    },
+    {
+      "epoch": 0.8826915221508411,
+      "grad_norm": 0.9219105243682861,
+      "learning_rate": 3.090628097506911e-05,
+      "loss": 0.3525,
+      "step": 11885
+    },
+    {
+      "epoch": 0.8830628690259571,
+      "grad_norm": 0.8252973556518555,
+      "learning_rate": 3.08916710810601e-05,
+      "loss": 0.3462,
+      "step": 11890
+    },
+    {
+      "epoch": 0.8834342159010732,
+      "grad_norm": 0.8419556617736816,
+      "learning_rate": 3.087705905630406e-05,
+      "loss": 0.3503,
+      "step": 11895
+    },
+    {
+      "epoch": 0.8838055627761893,
+      "grad_norm": 0.8596616387367249,
+      "learning_rate": 3.0862444906085514e-05,
+      "loss": 0.3658,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8841769096513052,
+      "grad_norm": 1.017411231994629,
+      "learning_rate": 3.084782863568971e-05,
+      "loss": 0.3496,
+      "step": 11905
+    },
+    {
+      "epoch": 0.8845482565264213,
+      "grad_norm": 0.9587264060974121,
+      "learning_rate": 3.0833210250402694e-05,
+      "loss": 0.3452,
+      "step": 11910
+    },
+    {
+      "epoch": 0.8849196034015374,
+      "grad_norm": 1.0870927572250366,
+      "learning_rate": 3.0818589755511254e-05,
+      "loss": 0.3624,
+      "step": 11915
+    },
+    {
+      "epoch": 0.8852909502766534,
+      "grad_norm": 0.9153867959976196,
+      "learning_rate": 3.080396715630297e-05,
+      "loss": 0.3623,
+      "step": 11920
+    },
+    {
+      "epoch": 0.8856622971517695,
+      "grad_norm": 0.8542145490646362,
+      "learning_rate": 3.078934245806615e-05,
+      "loss": 0.378,
+      "step": 11925
+    },
+    {
+      "epoch": 0.8860336440268856,
+      "grad_norm": 0.6727744340896606,
+      "learning_rate": 3.077471566608988e-05,
+      "loss": 0.3439,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8864049909020015,
+      "grad_norm": 0.9169203042984009,
+      "learning_rate": 3.0760086785664e-05,
+      "loss": 0.3537,
+      "step": 11935
+    },
+    {
+      "epoch": 0.8867763377771176,
+      "grad_norm": 0.7953917980194092,
+      "learning_rate": 3.07454558220791e-05,
+      "loss": 0.3408,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8871476846522337,
+      "grad_norm": 0.7179327607154846,
+      "learning_rate": 3.073082278062654e-05,
+      "loss": 0.34,
+      "step": 11945
+    },
+    {
+      "epoch": 0.8875190315273497,
+      "grad_norm": 0.9073119163513184,
+      "learning_rate": 3.07161876665984e-05,
+      "loss": 0.357,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8878903784024658,
+      "grad_norm": 0.7440044283866882,
+      "learning_rate": 3.070155048528756e-05,
+      "loss": 0.3719,
+      "step": 11955
+    },
+    {
+      "epoch": 0.8882617252775817,
+      "grad_norm": 0.7663770914077759,
+      "learning_rate": 3.068691124198758e-05,
+      "loss": 0.3634,
+      "step": 11960
+    },
+    {
+      "epoch": 0.8886330721526978,
+      "grad_norm": 0.795628011226654,
+      "learning_rate": 3.0672269941992825e-05,
+      "loss": 0.3358,
+      "step": 11965
+    },
+    {
+      "epoch": 0.8890044190278139,
+      "grad_norm": 0.9965798854827881,
+      "learning_rate": 3.065762659059838e-05,
+      "loss": 0.3331,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8893757659029299,
+      "grad_norm": 0.8984972238540649,
+      "learning_rate": 3.0642981193100065e-05,
+      "loss": 0.3521,
+      "step": 11975
+    },
+    {
+      "epoch": 0.889747112778046,
+      "grad_norm": 0.859411358833313,
+      "learning_rate": 3.0628333754794444e-05,
+      "loss": 0.3375,
+      "step": 11980
+    },
+    {
+      "epoch": 0.890118459653162,
+      "grad_norm": 1.0164400339126587,
+      "learning_rate": 3.061368428097884e-05,
+      "loss": 0.3496,
+      "step": 11985
+    },
+    {
+      "epoch": 0.890489806528278,
+      "grad_norm": 1.0231250524520874,
+      "learning_rate": 3.0599032776951295e-05,
+      "loss": 0.3414,
+      "step": 11990
+    },
+    {
+      "epoch": 0.8908611534033941,
+      "grad_norm": 0.7204227447509766,
+      "learning_rate": 3.0584379248010556e-05,
+      "loss": 0.3552,
+      "step": 11995
+    },
+    {
+      "epoch": 0.8912325002785102,
+      "grad_norm": 0.8453019261360168,
+      "learning_rate": 3.056972369945617e-05,
+      "loss": 0.3494,
+      "step": 12000
+    },
+    {
+      "epoch": 0.8916038471536262,
+      "grad_norm": 0.8141784071922302,
+      "learning_rate": 3.0555066136588354e-05,
+      "loss": 0.3554,
+      "step": 12005
+    },
+    {
+      "epoch": 0.8919751940287423,
+      "grad_norm": 1.678991436958313,
+      "learning_rate": 3.054040656470809e-05,
+      "loss": 0.3733,
+      "step": 12010
+    },
+    {
+      "epoch": 0.8923465409038583,
+      "grad_norm": 0.8809152841567993,
+      "learning_rate": 3.0525744989117044e-05,
+      "loss": 0.3491,
+      "step": 12015
+    },
+    {
+      "epoch": 0.8927178877789743,
+      "grad_norm": 0.784815788269043,
+      "learning_rate": 3.051108141511767e-05,
+      "loss": 0.3444,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8930892346540904,
+      "grad_norm": 0.919638991355896,
+      "learning_rate": 3.049641584801309e-05,
+      "loss": 0.3502,
+      "step": 12025
+    },
+    {
+      "epoch": 0.8934605815292064,
+      "grad_norm": 0.757986307144165,
+      "learning_rate": 3.048174829310717e-05,
+      "loss": 0.3472,
+      "step": 12030
+    },
+    {
+      "epoch": 0.8938319284043225,
+      "grad_norm": 0.7388538122177124,
+      "learning_rate": 3.046707875570449e-05,
+      "loss": 0.3509,
+      "step": 12035
+    },
+    {
+      "epoch": 0.8942032752794385,
+      "grad_norm": 2.4140920639038086,
+      "learning_rate": 3.0452407241110347e-05,
+      "loss": 0.3668,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8945746221545545,
+      "grad_norm": 0.9365836381912231,
+      "learning_rate": 3.0437733754630772e-05,
+      "loss": 0.3037,
+      "step": 12045
+    },
+    {
+      "epoch": 0.8949459690296706,
+      "grad_norm": 0.6372706294059753,
+      "learning_rate": 3.0423058301572466e-05,
+      "loss": 0.3356,
+      "step": 12050
+    },
+    {
+      "epoch": 0.8953173159047867,
+      "grad_norm": 0.8278931975364685,
+      "learning_rate": 3.0408380887242882e-05,
+      "loss": 0.3504,
+      "step": 12055
+    },
+    {
+      "epoch": 0.8956886627799027,
+      "grad_norm": 1.124632716178894,
+      "learning_rate": 3.0393701516950157e-05,
+      "loss": 0.373,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8960600096550188,
+      "grad_norm": 0.8332802653312683,
+      "learning_rate": 3.0379020196003156e-05,
+      "loss": 0.3443,
+      "step": 12065
+    },
+    {
+      "epoch": 0.8964313565301348,
+      "grad_norm": 0.7034924030303955,
+      "learning_rate": 3.0364336929711435e-05,
+      "loss": 0.3516,
+      "step": 12070
+    },
+    {
+      "epoch": 0.8968027034052508,
+      "grad_norm": 0.9156379699707031,
+      "learning_rate": 3.0349651723385254e-05,
+      "loss": 0.3379,
+      "step": 12075
+    },
+    {
+      "epoch": 0.8971740502803669,
+      "grad_norm": 0.7082470655441284,
+      "learning_rate": 3.0334964582335583e-05,
+      "loss": 0.3254,
+      "step": 12080
+    },
+    {
+      "epoch": 0.897545397155483,
+      "grad_norm": 0.9504019618034363,
+      "learning_rate": 3.0320275511874085e-05,
+      "loss": 0.3355,
+      "step": 12085
+    },
+    {
+      "epoch": 0.897916744030599,
+      "grad_norm": 1.6386775970458984,
+      "learning_rate": 3.030558451731313e-05,
+      "loss": 0.3538,
+      "step": 12090
+    },
+    {
+      "epoch": 0.898288090905715,
+      "grad_norm": 0.7984548807144165,
+      "learning_rate": 3.029089160396576e-05,
+      "loss": 0.344,
+      "step": 12095
+    },
+    {
+      "epoch": 0.8986594377808311,
+      "grad_norm": 0.9346609115600586,
+      "learning_rate": 3.0276196777145742e-05,
+      "loss": 0.3587,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8990307846559471,
+      "grad_norm": 0.9744558334350586,
+      "learning_rate": 3.0261500042167507e-05,
+      "loss": 0.3305,
+      "step": 12105
+    },
+    {
+      "epoch": 0.8994021315310632,
+      "grad_norm": 1.0569168329238892,
+      "learning_rate": 3.0246801404346203e-05,
+      "loss": 0.3552,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8997734784061792,
+      "grad_norm": 1.1052289009094238,
+      "learning_rate": 3.0232100868997648e-05,
+      "loss": 0.3634,
+      "step": 12115
+    },
+    {
+      "epoch": 0.9001448252812952,
+      "grad_norm": 0.998342752456665,
+      "learning_rate": 3.0217398441438343e-05,
+      "loss": 0.358,
+      "step": 12120
+    },
+    {
+      "epoch": 0.9005161721564113,
+      "grad_norm": 0.9586290717124939,
+      "learning_rate": 3.0202694126985494e-05,
+      "loss": 0.3349,
+      "step": 12125
+    },
+    {
+      "epoch": 0.9008875190315273,
+      "grad_norm": 0.7554903030395508,
+      "learning_rate": 3.018798793095697e-05,
+      "loss": 0.3456,
+      "step": 12130
+    },
+    {
+      "epoch": 0.9012588659066434,
+      "grad_norm": 0.7605744004249573,
+      "learning_rate": 3.0173279858671333e-05,
+      "loss": 0.3497,
+      "step": 12135
+    },
+    {
+      "epoch": 0.9016302127817595,
+      "grad_norm": 0.8068292737007141,
+      "learning_rate": 3.0158569915447803e-05,
+      "loss": 0.3484,
+      "step": 12140
+    },
+    {
+      "epoch": 0.9020015596568755,
+      "grad_norm": 0.8266493082046509,
+      "learning_rate": 3.0143858106606303e-05,
+      "loss": 0.3199,
+      "step": 12145
+    },
+    {
+      "epoch": 0.9023729065319915,
+      "grad_norm": 0.8617991209030151,
+      "learning_rate": 3.012914443746742e-05,
+      "loss": 0.3528,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9027442534071076,
+      "grad_norm": 0.800369918346405,
+      "learning_rate": 3.011442891335241e-05,
+      "loss": 0.3501,
+      "step": 12155
+    },
+    {
+      "epoch": 0.9031156002822236,
+      "grad_norm": 0.5995765328407288,
+      "learning_rate": 3.0099711539583202e-05,
+      "loss": 0.3415,
+      "step": 12160
+    },
+    {
+      "epoch": 0.9034869471573397,
+      "grad_norm": 0.892348051071167,
+      "learning_rate": 3.0084992321482396e-05,
+      "loss": 0.3692,
+      "step": 12165
+    },
+    {
+      "epoch": 0.9038582940324558,
+      "grad_norm": 0.904771089553833,
+      "learning_rate": 3.007027126437325e-05,
+      "loss": 0.3472,
+      "step": 12170
+    },
+    {
+      "epoch": 0.9042296409075717,
+      "grad_norm": 0.8912681341171265,
+      "learning_rate": 3.0055548373579712e-05,
+      "loss": 0.3379,
+      "step": 12175
+    },
+    {
+      "epoch": 0.9046009877826878,
+      "grad_norm": 0.6829692125320435,
+      "learning_rate": 3.0040823654426357e-05,
+      "loss": 0.3496,
+      "step": 12180
+    },
+    {
+      "epoch": 0.9049723346578038,
+      "grad_norm": 0.8899698853492737,
+      "learning_rate": 3.0026097112238454e-05,
+      "loss": 0.3404,
+      "step": 12185
+    },
+    {
+      "epoch": 0.9053436815329199,
+      "grad_norm": 0.7954959273338318,
+      "learning_rate": 3.0011368752341905e-05,
+      "loss": 0.3657,
+      "step": 12190
+    },
+    {
+      "epoch": 0.905715028408036,
+      "grad_norm": 0.8103697896003723,
+      "learning_rate": 2.999663858006329e-05,
+      "loss": 0.325,
+      "step": 12195
+    },
+    {
+      "epoch": 0.906086375283152,
+      "grad_norm": 0.7728240489959717,
+      "learning_rate": 2.9981906600729842e-05,
+      "loss": 0.3404,
+      "step": 12200
+    },
+    {
+      "epoch": 0.906457722158268,
+      "grad_norm": 0.873102605342865,
+      "learning_rate": 2.9967172819669433e-05,
+      "loss": 0.3357,
+      "step": 12205
+    },
+    {
+      "epoch": 0.9068290690333841,
+      "grad_norm": 1.4603787660598755,
+      "learning_rate": 2.9952437242210592e-05,
+      "loss": 0.349,
+      "step": 12210
+    },
+    {
+      "epoch": 0.9072004159085001,
+      "grad_norm": 0.7289944887161255,
+      "learning_rate": 2.99376998736825e-05,
+      "loss": 0.3354,
+      "step": 12215
+    },
+    {
+      "epoch": 0.9075717627836162,
+      "grad_norm": 0.7822965383529663,
+      "learning_rate": 2.9922960719415e-05,
+      "loss": 0.3467,
+      "step": 12220
+    },
+    {
+      "epoch": 0.9079431096587323,
+      "grad_norm": 0.859673023223877,
+      "learning_rate": 2.9908219784738568e-05,
+      "loss": 0.3443,
+      "step": 12225
+    },
+    {
+      "epoch": 0.9083144565338482,
+      "grad_norm": 0.7056594491004944,
+      "learning_rate": 2.98934770749843e-05,
+      "loss": 0.3503,
+      "step": 12230
+    },
+    {
+      "epoch": 0.9086858034089643,
+      "grad_norm": 0.8911407589912415,
+      "learning_rate": 2.9878732595483982e-05,
+      "loss": 0.3383,
+      "step": 12235
+    },
+    {
+      "epoch": 0.9090571502840804,
+      "grad_norm": 0.878011167049408,
+      "learning_rate": 2.9863986351570006e-05,
+      "loss": 0.3527,
+      "step": 12240
+    },
+    {
+      "epoch": 0.9094284971591964,
+      "grad_norm": 0.8897885680198669,
+      "learning_rate": 2.9849238348575416e-05,
+      "loss": 0.3577,
+      "step": 12245
+    },
+    {
+      "epoch": 0.9097998440343125,
+      "grad_norm": 1.0172691345214844,
+      "learning_rate": 2.9834488591833874e-05,
+      "loss": 0.3301,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9101711909094286,
+      "grad_norm": 0.8597069978713989,
+      "learning_rate": 2.981973708667971e-05,
+      "loss": 0.3324,
+      "step": 12255
+    },
+    {
+      "epoch": 0.9105425377845445,
+      "grad_norm": 0.8964053392410278,
+      "learning_rate": 2.980498383844785e-05,
+      "loss": 0.3381,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9109138846596606,
+      "grad_norm": 0.81219881772995,
+      "learning_rate": 2.979022885247387e-05,
+      "loss": 0.3557,
+      "step": 12265
+    },
+    {
+      "epoch": 0.9112852315347766,
+      "grad_norm": 0.7895968556404114,
+      "learning_rate": 2.9775472134093986e-05,
+      "loss": 0.34,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9116565784098927,
+      "grad_norm": 1.065226674079895,
+      "learning_rate": 2.9760713688645003e-05,
+      "loss": 0.3473,
+      "step": 12275
+    },
+    {
+      "epoch": 0.9120279252850088,
+      "grad_norm": 1.1710779666900635,
+      "learning_rate": 2.9745953521464393e-05,
+      "loss": 0.3425,
+      "step": 12280
+    },
+    {
+      "epoch": 0.9123992721601247,
+      "grad_norm": 0.6871535778045654,
+      "learning_rate": 2.9731191637890222e-05,
+      "loss": 0.352,
+      "step": 12285
+    },
+    {
+      "epoch": 0.9127706190352408,
+      "grad_norm": 0.8434427380561829,
+      "learning_rate": 2.9716428043261187e-05,
+      "loss": 0.3521,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9131419659103569,
+      "grad_norm": 0.8964160680770874,
+      "learning_rate": 2.9701662742916586e-05,
+      "loss": 0.3217,
+      "step": 12295
+    },
+    {
+      "epoch": 0.9135133127854729,
+      "grad_norm": 1.7727447748184204,
+      "learning_rate": 2.9686895742196386e-05,
+      "loss": 0.3556,
+      "step": 12300
+    },
+    {
+      "epoch": 0.913884659660589,
+      "grad_norm": 0.7401552796363831,
+      "learning_rate": 2.9672127046441106e-05,
+      "loss": 0.3542,
+      "step": 12305
+    },
+    {
+      "epoch": 0.914256006535705,
+      "grad_norm": 0.8029780983924866,
+      "learning_rate": 2.965735666099191e-05,
+      "loss": 0.335,
+      "step": 12310
+    },
+    {
+      "epoch": 0.914627353410821,
+      "grad_norm": 1.061109185218811,
+      "learning_rate": 2.9642584591190575e-05,
+      "loss": 0.3502,
+      "step": 12315
+    },
+    {
+      "epoch": 0.9149987002859371,
+      "grad_norm": 0.9626049399375916,
+      "learning_rate": 2.9627810842379472e-05,
+      "loss": 0.3564,
+      "step": 12320
+    },
+    {
+      "epoch": 0.9153700471610532,
+      "grad_norm": 0.9086316227912903,
+      "learning_rate": 2.9613035419901603e-05,
+      "loss": 0.3414,
+      "step": 12325
+    },
+    {
+      "epoch": 0.9157413940361692,
+      "grad_norm": 1.798221230506897,
+      "learning_rate": 2.9598258329100546e-05,
+      "loss": 0.362,
+      "step": 12330
+    },
+    {
+      "epoch": 0.9161127409112853,
+      "grad_norm": 0.7969675064086914,
+      "learning_rate": 2.9583479575320504e-05,
+      "loss": 0.3382,
+      "step": 12335
+    },
+    {
+      "epoch": 0.9164840877864012,
+      "grad_norm": 0.6091347336769104,
+      "learning_rate": 2.956869916390626e-05,
+      "loss": 0.3676,
+      "step": 12340
+    },
+    {
+      "epoch": 0.9168554346615173,
+      "grad_norm": 0.639950156211853,
+      "learning_rate": 2.9553917100203234e-05,
+      "loss": 0.3228,
+      "step": 12345
+    },
+    {
+      "epoch": 0.9172267815366334,
+      "grad_norm": 0.7365955114364624,
+      "learning_rate": 2.9539133389557395e-05,
+      "loss": 0.3592,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9175981284117494,
+      "grad_norm": 0.8247587084770203,
+      "learning_rate": 2.952434803731535e-05,
+      "loss": 0.3462,
+      "step": 12355
+    },
+    {
+      "epoch": 0.9179694752868655,
+      "grad_norm": 1.6064798831939697,
+      "learning_rate": 2.9509561048824276e-05,
+      "loss": 0.3507,
+      "step": 12360
+    },
+    {
+      "epoch": 0.9183408221619815,
+      "grad_norm": 0.7756306529045105,
+      "learning_rate": 2.949477242943195e-05,
+      "loss": 0.3549,
+      "step": 12365
+    },
+    {
+      "epoch": 0.9187121690370975,
+      "grad_norm": 1.0088223218917847,
+      "learning_rate": 2.9479982184486737e-05,
+      "loss": 0.3677,
+      "step": 12370
+    },
+    {
+      "epoch": 0.9190835159122136,
+      "grad_norm": 0.9376280307769775,
+      "learning_rate": 2.9465190319337587e-05,
+      "loss": 0.3292,
+      "step": 12375
+    },
+    {
+      "epoch": 0.9194548627873297,
+      "grad_norm": 0.7487239837646484,
+      "learning_rate": 2.9450396839334042e-05,
+      "loss": 0.3398,
+      "step": 12380
+    },
+    {
+      "epoch": 0.9198262096624457,
+      "grad_norm": 0.8737534284591675,
+      "learning_rate": 2.943560174982622e-05,
+      "loss": 0.3441,
+      "step": 12385
+    },
+    {
+      "epoch": 0.9201975565375617,
+      "grad_norm": 0.8645554780960083,
+      "learning_rate": 2.9420805056164835e-05,
+      "loss": 0.3467,
+      "step": 12390
+    },
+    {
+      "epoch": 0.9205689034126778,
+      "grad_norm": 0.8328379392623901,
+      "learning_rate": 2.940600676370116e-05,
+      "loss": 0.3392,
+      "step": 12395
+    },
+    {
+      "epoch": 0.9209402502877938,
+      "grad_norm": 0.6697943210601807,
+      "learning_rate": 2.9391206877787064e-05,
+      "loss": 0.3419,
+      "step": 12400
+    },
+    {
+      "epoch": 0.9213115971629099,
+      "grad_norm": 0.9482589364051819,
+      "learning_rate": 2.9376405403774988e-05,
+      "loss": 0.3467,
+      "step": 12405
+    },
+    {
+      "epoch": 0.921682944038026,
+      "grad_norm": 0.9211759567260742,
+      "learning_rate": 2.9361602347017946e-05,
+      "loss": 0.3485,
+      "step": 12410
+    },
+    {
+      "epoch": 0.922054290913142,
+      "grad_norm": 1.0176403522491455,
+      "learning_rate": 2.9346797712869522e-05,
+      "loss": 0.36,
+      "step": 12415
+    },
+    {
+      "epoch": 0.922425637788258,
+      "grad_norm": 0.8227538466453552,
+      "learning_rate": 2.9331991506683876e-05,
+      "loss": 0.3462,
+      "step": 12420
+    },
+    {
+      "epoch": 0.922796984663374,
+      "grad_norm": 0.8057060241699219,
+      "learning_rate": 2.9317183733815723e-05,
+      "loss": 0.3587,
+      "step": 12425
+    },
+    {
+      "epoch": 0.9231683315384901,
+      "grad_norm": 0.8205479383468628,
+      "learning_rate": 2.9302374399620364e-05,
+      "loss": 0.3842,
+      "step": 12430
+    },
+    {
+      "epoch": 0.9235396784136062,
+      "grad_norm": 0.9048784375190735,
+      "learning_rate": 2.9287563509453658e-05,
+      "loss": 0.3531,
+      "step": 12435
+    },
+    {
+      "epoch": 0.9239110252887222,
+      "grad_norm": 0.8701086044311523,
+      "learning_rate": 2.9272751068672015e-05,
+      "loss": 0.3555,
+      "step": 12440
+    },
+    {
+      "epoch": 0.9242823721638382,
+      "grad_norm": 1.1633310317993164,
+      "learning_rate": 2.9257937082632426e-05,
+      "loss": 0.349,
+      "step": 12445
+    },
+    {
+      "epoch": 0.9246537190389543,
+      "grad_norm": 0.9627925157546997,
+      "learning_rate": 2.9243121556692417e-05,
+      "loss": 0.3371,
+      "step": 12450
+    },
+    {
+      "epoch": 0.9250250659140703,
+      "grad_norm": 0.828937828540802,
+      "learning_rate": 2.922830449621009e-05,
+      "loss": 0.3361,
+      "step": 12455
+    },
+    {
+      "epoch": 0.9253964127891864,
+      "grad_norm": 1.0228971242904663,
+      "learning_rate": 2.9213485906544103e-05,
+      "loss": 0.3445,
+      "step": 12460
+    },
+    {
+      "epoch": 0.9257677596643025,
+      "grad_norm": 1.1212095022201538,
+      "learning_rate": 2.9198665793053648e-05,
+      "loss": 0.3366,
+      "step": 12465
+    },
+    {
+      "epoch": 0.9261391065394184,
+      "grad_norm": 1.1539406776428223,
+      "learning_rate": 2.9183844161098484e-05,
+      "loss": 0.3714,
+      "step": 12470
+    },
+    {
+      "epoch": 0.9265104534145345,
+      "grad_norm": 0.8389959335327148,
+      "learning_rate": 2.9169021016038917e-05,
+      "loss": 0.3501,
+      "step": 12475
+    },
+    {
+      "epoch": 0.9268818002896506,
+      "grad_norm": 0.9025770425796509,
+      "learning_rate": 2.9154196363235804e-05,
+      "loss": 0.3364,
+      "step": 12480
+    },
+    {
+      "epoch": 0.9272531471647666,
+      "grad_norm": 0.9710972309112549,
+      "learning_rate": 2.913937020805052e-05,
+      "loss": 0.3577,
+      "step": 12485
+    },
+    {
+      "epoch": 0.9276244940398827,
+      "grad_norm": 1.1810095310211182,
+      "learning_rate": 2.912454255584503e-05,
+      "loss": 0.3565,
+      "step": 12490
+    },
+    {
+      "epoch": 0.9279958409149986,
+      "grad_norm": 1.0007275342941284,
+      "learning_rate": 2.9109713411981798e-05,
+      "loss": 0.3399,
+      "step": 12495
+    },
+    {
+      "epoch": 0.9283671877901147,
+      "grad_norm": 1.5194200277328491,
+      "learning_rate": 2.909488278182385e-05,
+      "loss": 0.3367,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9287385346652308,
+      "grad_norm": 0.5930137038230896,
+      "learning_rate": 2.908005067073476e-05,
+      "loss": 0.3312,
+      "step": 12505
+    },
+    {
+      "epoch": 0.9291098815403468,
+      "grad_norm": 0.9861831665039062,
+      "learning_rate": 2.9065217084078588e-05,
+      "loss": 0.3504,
+      "step": 12510
+    },
+    {
+      "epoch": 0.9294812284154629,
+      "grad_norm": 3.8240902423858643,
+      "learning_rate": 2.905038202721999e-05,
+      "loss": 0.3466,
+      "step": 12515
+    },
+    {
+      "epoch": 0.929852575290579,
+      "grad_norm": 1.0071384906768799,
+      "learning_rate": 2.903554550552412e-05,
+      "loss": 0.333,
+      "step": 12520
+    },
+    {
+      "epoch": 0.9302239221656949,
+      "grad_norm": 0.8163401484489441,
+      "learning_rate": 2.902070752435666e-05,
+      "loss": 0.3312,
+      "step": 12525
+    },
+    {
+      "epoch": 0.930595269040811,
+      "grad_norm": 0.9900504350662231,
+      "learning_rate": 2.900586808908382e-05,
+      "loss": 0.3588,
+      "step": 12530
+    },
+    {
+      "epoch": 0.9309666159159271,
+      "grad_norm": 0.9772970676422119,
+      "learning_rate": 2.8991027205072363e-05,
+      "loss": 0.3598,
+      "step": 12535
+    },
+    {
+      "epoch": 0.9313379627910431,
+      "grad_norm": 1.0091187953948975,
+      "learning_rate": 2.8976184877689545e-05,
+      "loss": 0.3455,
+      "step": 12540
+    },
+    {
+      "epoch": 0.9317093096661592,
+      "grad_norm": 3.244717836380005,
+      "learning_rate": 2.8961341112303147e-05,
+      "loss": 0.3511,
+      "step": 12545
+    },
+    {
+      "epoch": 0.9320806565412753,
+      "grad_norm": 0.766475260257721,
+      "learning_rate": 2.894649591428149e-05,
+      "loss": 0.336,
+      "step": 12550
+    },
+    {
+      "epoch": 0.9324520034163912,
+      "grad_norm": 0.9685466885566711,
+      "learning_rate": 2.893164928899339e-05,
+      "loss": 0.341,
+      "step": 12555
+    },
+    {
+      "epoch": 0.9328233502915073,
+      "grad_norm": 0.8565616607666016,
+      "learning_rate": 2.8916801241808205e-05,
+      "loss": 0.3481,
+      "step": 12560
+    },
+    {
+      "epoch": 0.9331946971666233,
+      "grad_norm": 0.896621823310852,
+      "learning_rate": 2.8901951778095783e-05,
+      "loss": 0.3516,
+      "step": 12565
+    },
+    {
+      "epoch": 0.9335660440417394,
+      "grad_norm": 0.9232330322265625,
+      "learning_rate": 2.888710090322649e-05,
+      "loss": 0.3601,
+      "step": 12570
+    },
+    {
+      "epoch": 0.9339373909168555,
+      "grad_norm": 0.8799304962158203,
+      "learning_rate": 2.88722486225712e-05,
+      "loss": 0.337,
+      "step": 12575
+    },
+    {
+      "epoch": 0.9343087377919714,
+      "grad_norm": 1.2841633558273315,
+      "learning_rate": 2.8857394941501325e-05,
+      "loss": 0.3623,
+      "step": 12580
+    },
+    {
+      "epoch": 0.9346800846670875,
+      "grad_norm": 0.7310310006141663,
+      "learning_rate": 2.8842539865388734e-05,
+      "loss": 0.3434,
+      "step": 12585
+    },
+    {
+      "epoch": 0.9350514315422036,
+      "grad_norm": 0.8738952875137329,
+      "learning_rate": 2.8827683399605843e-05,
+      "loss": 0.3628,
+      "step": 12590
+    },
+    {
+      "epoch": 0.9354227784173196,
+      "grad_norm": 0.8512542247772217,
+      "learning_rate": 2.8812825549525545e-05,
+      "loss": 0.3302,
+      "step": 12595
+    },
+    {
+      "epoch": 0.9357941252924357,
+      "grad_norm": 0.8368507027626038,
+      "learning_rate": 2.8797966320521246e-05,
+      "loss": 0.3378,
+      "step": 12600
+    },
+    {
+      "epoch": 0.9361654721675517,
+      "grad_norm": 0.7401739954948425,
+      "learning_rate": 2.878310571796685e-05,
+      "loss": 0.3359,
+      "step": 12605
+    },
+    {
+      "epoch": 0.9365368190426677,
+      "grad_norm": 0.7290681004524231,
+      "learning_rate": 2.876824374723675e-05,
+      "loss": 0.3379,
+      "step": 12610
+    },
+    {
+      "epoch": 0.9369081659177838,
+      "grad_norm": 0.8211921453475952,
+      "learning_rate": 2.8753380413705843e-05,
+      "loss": 0.3525,
+      "step": 12615
+    },
+    {
+      "epoch": 0.9372795127928999,
+      "grad_norm": 0.9918500781059265,
+      "learning_rate": 2.8738515722749504e-05,
+      "loss": 0.3552,
+      "step": 12620
+    },
+    {
+      "epoch": 0.9376508596680159,
+      "grad_norm": 0.908306360244751,
+      "learning_rate": 2.8723649679743626e-05,
+      "loss": 0.3535,
+      "step": 12625
+    },
+    {
+      "epoch": 0.938022206543132,
+      "grad_norm": 0.9392675161361694,
+      "learning_rate": 2.8708782290064562e-05,
+      "loss": 0.3341,
+      "step": 12630
+    },
+    {
+      "epoch": 0.938393553418248,
+      "grad_norm": 0.6768072247505188,
+      "learning_rate": 2.8693913559089164e-05,
+      "loss": 0.3272,
+      "step": 12635
+    },
+    {
+      "epoch": 0.938764900293364,
+      "grad_norm": 0.8711260557174683,
+      "learning_rate": 2.8679043492194778e-05,
+      "loss": 0.3469,
+      "step": 12640
+    },
+    {
+      "epoch": 0.9391362471684801,
+      "grad_norm": 0.9255607724189758,
+      "learning_rate": 2.8664172094759218e-05,
+      "loss": 0.3684,
+      "step": 12645
+    },
+    {
+      "epoch": 0.9395075940435961,
+      "grad_norm": 1.1237608194351196,
+      "learning_rate": 2.86492993721608e-05,
+      "loss": 0.3568,
+      "step": 12650
+    },
+    {
+      "epoch": 0.9398789409187122,
+      "grad_norm": 0.95046466588974,
+      "learning_rate": 2.863442532977828e-05,
+      "loss": 0.3538,
+      "step": 12655
+    },
+    {
+      "epoch": 0.9402502877938282,
+      "grad_norm": 0.8607051968574524,
+      "learning_rate": 2.8619549972990934e-05,
+      "loss": 0.3388,
+      "step": 12660
+    },
+    {
+      "epoch": 0.9406216346689442,
+      "grad_norm": 0.989203929901123,
+      "learning_rate": 2.8604673307178487e-05,
+      "loss": 0.3608,
+      "step": 12665
+    },
+    {
+      "epoch": 0.9409929815440603,
+      "grad_norm": 0.9122014045715332,
+      "learning_rate": 2.858979533772116e-05,
+      "loss": 0.3398,
+      "step": 12670
+    },
+    {
+      "epoch": 0.9413643284191764,
+      "grad_norm": 1.0959396362304688,
+      "learning_rate": 2.8574916069999613e-05,
+      "loss": 0.3634,
+      "step": 12675
+    },
+    {
+      "epoch": 0.9417356752942924,
+      "grad_norm": 1.000762939453125,
+      "learning_rate": 2.856003550939501e-05,
+      "loss": 0.3404,
+      "step": 12680
+    },
+    {
+      "epoch": 0.9421070221694084,
+      "grad_norm": 0.8134162425994873,
+      "learning_rate": 2.854515366128896e-05,
+      "loss": 0.3614,
+      "step": 12685
+    },
+    {
+      "epoch": 0.9424783690445245,
+      "grad_norm": 0.659251868724823,
+      "learning_rate": 2.853027053106354e-05,
+      "loss": 0.3293,
+      "step": 12690
+    },
+    {
+      "epoch": 0.9428497159196405,
+      "grad_norm": 1.0323694944381714,
+      "learning_rate": 2.851538612410131e-05,
+      "loss": 0.3517,
+      "step": 12695
+    },
+    {
+      "epoch": 0.9432210627947566,
+      "grad_norm": 1.0935428142547607,
+      "learning_rate": 2.850050044578526e-05,
+      "loss": 0.3391,
+      "step": 12700
+    },
+    {
+      "epoch": 0.9435924096698727,
+      "grad_norm": 1.014157772064209,
+      "learning_rate": 2.8485613501498866e-05,
+      "loss": 0.3281,
+      "step": 12705
+    },
+    {
+      "epoch": 0.9439637565449887,
+      "grad_norm": 0.9155251383781433,
+      "learning_rate": 2.8470725296626046e-05,
+      "loss": 0.3377,
+      "step": 12710
+    },
+    {
+      "epoch": 0.9443351034201047,
+      "grad_norm": 0.9557784795761108,
+      "learning_rate": 2.845583583655119e-05,
+      "loss": 0.3406,
+      "step": 12715
+    },
+    {
+      "epoch": 0.9447064502952207,
+      "grad_norm": 0.9360071420669556,
+      "learning_rate": 2.8440945126659124e-05,
+      "loss": 0.3551,
+      "step": 12720
+    },
+    {
+      "epoch": 0.9450777971703368,
+      "grad_norm": 1.0895956754684448,
+      "learning_rate": 2.842605317233514e-05,
+      "loss": 0.3259,
+      "step": 12725
+    },
+    {
+      "epoch": 0.9454491440454529,
+      "grad_norm": 0.8507657051086426,
+      "learning_rate": 2.8411159978964964e-05,
+      "loss": 0.3472,
+      "step": 12730
+    },
+    {
+      "epoch": 0.9458204909205689,
+      "grad_norm": 0.7691348791122437,
+      "learning_rate": 2.8396265551934792e-05,
+      "loss": 0.3593,
+      "step": 12735
+    },
+    {
+      "epoch": 0.9461918377956849,
+      "grad_norm": 0.6096264123916626,
+      "learning_rate": 2.8381369896631254e-05,
+      "loss": 0.3183,
+      "step": 12740
+    },
+    {
+      "epoch": 0.946563184670801,
+      "grad_norm": 0.7602872848510742,
+      "learning_rate": 2.8366473018441415e-05,
+      "loss": 0.3286,
+      "step": 12745
+    },
+    {
+      "epoch": 0.946934531545917,
+      "grad_norm": 0.9890530705451965,
+      "learning_rate": 2.8351574922752804e-05,
+      "loss": 0.324,
+      "step": 12750
+    },
+    {
+      "epoch": 0.9473058784210331,
+      "grad_norm": 0.8396008014678955,
+      "learning_rate": 2.8336675614953368e-05,
+      "loss": 0.3404,
+      "step": 12755
+    },
+    {
+      "epoch": 0.9476772252961492,
+      "grad_norm": 1.0501019954681396,
+      "learning_rate": 2.8321775100431513e-05,
+      "loss": 0.3322,
+      "step": 12760
+    },
+    {
+      "epoch": 0.9480485721712651,
+      "grad_norm": 1.1446385383605957,
+      "learning_rate": 2.8306873384576066e-05,
+      "loss": 0.3323,
+      "step": 12765
+    },
+    {
+      "epoch": 0.9484199190463812,
+      "grad_norm": 0.8426380753517151,
+      "learning_rate": 2.8291970472776292e-05,
+      "loss": 0.3485,
+      "step": 12770
+    },
+    {
+      "epoch": 0.9487912659214973,
+      "grad_norm": 0.9239525198936462,
+      "learning_rate": 2.8277066370421888e-05,
+      "loss": 0.3329,
+      "step": 12775
+    },
+    {
+      "epoch": 0.9491626127966133,
+      "grad_norm": 1.0131057500839233,
+      "learning_rate": 2.826216108290299e-05,
+      "loss": 0.3416,
+      "step": 12780
+    },
+    {
+      "epoch": 0.9495339596717294,
+      "grad_norm": 1.0407359600067139,
+      "learning_rate": 2.8247254615610157e-05,
+      "loss": 0.3636,
+      "step": 12785
+    },
+    {
+      "epoch": 0.9499053065468455,
+      "grad_norm": 1.0886719226837158,
+      "learning_rate": 2.823234697393436e-05,
+      "loss": 0.3371,
+      "step": 12790
+    },
+    {
+      "epoch": 0.9502766534219614,
+      "grad_norm": 0.6153767704963684,
+      "learning_rate": 2.8217438163267036e-05,
+      "loss": 0.3319,
+      "step": 12795
+    },
+    {
+      "epoch": 0.9506480002970775,
+      "grad_norm": 0.9892891049385071,
+      "learning_rate": 2.820252818899999e-05,
+      "loss": 0.353,
+      "step": 12800
+    },
+    {
+      "epoch": 0.9510193471721935,
+      "grad_norm": 0.9184731841087341,
+      "learning_rate": 2.8187617056525494e-05,
+      "loss": 0.3163,
+      "step": 12805
+    },
+    {
+      "epoch": 0.9513906940473096,
+      "grad_norm": 1.0722240209579468,
+      "learning_rate": 2.8172704771236204e-05,
+      "loss": 0.3558,
+      "step": 12810
+    },
+    {
+      "epoch": 0.9517620409224257,
+      "grad_norm": 1.0215551853179932,
+      "learning_rate": 2.8157791338525224e-05,
+      "loss": 0.3558,
+      "step": 12815
+    },
+    {
+      "epoch": 0.9521333877975416,
+      "grad_norm": 1.0462008714675903,
+      "learning_rate": 2.8142876763786053e-05,
+      "loss": 0.3504,
+      "step": 12820
+    },
+    {
+      "epoch": 0.9525047346726577,
+      "grad_norm": 0.8154160380363464,
+      "learning_rate": 2.81279610524126e-05,
+      "loss": 0.3436,
+      "step": 12825
+    },
+    {
+      "epoch": 0.9528760815477738,
+      "grad_norm": 0.8203123807907104,
+      "learning_rate": 2.811304420979921e-05,
+      "loss": 0.3453,
+      "step": 12830
+    },
+    {
+      "epoch": 0.9532474284228898,
+      "grad_norm": 1.075543999671936,
+      "learning_rate": 2.80981262413406e-05,
+      "loss": 0.3419,
+      "step": 12835
+    },
+    {
+      "epoch": 0.9536187752980059,
+      "grad_norm": 0.8190999627113342,
+      "learning_rate": 2.808320715243194e-05,
+      "loss": 0.3439,
+      "step": 12840
+    },
+    {
+      "epoch": 0.953990122173122,
+      "grad_norm": 0.8523265719413757,
+      "learning_rate": 2.8068286948468753e-05,
+      "loss": 0.343,
+      "step": 12845
+    },
+    {
+      "epoch": 0.9543614690482379,
+      "grad_norm": 0.9448350071907043,
+      "learning_rate": 2.8053365634847012e-05,
+      "loss": 0.3251,
+      "step": 12850
+    },
+    {
+      "epoch": 0.954732815923354,
+      "grad_norm": 0.7018768191337585,
+      "learning_rate": 2.803844321696305e-05,
+      "loss": 0.3526,
+      "step": 12855
+    },
+    {
+      "epoch": 0.9551041627984701,
+      "grad_norm": 0.7716934680938721,
+      "learning_rate": 2.8023519700213652e-05,
+      "loss": 0.3447,
+      "step": 12860
+    },
+    {
+      "epoch": 0.9554755096735861,
+      "grad_norm": 0.7590059638023376,
+      "learning_rate": 2.8008595089995937e-05,
+      "loss": 0.3506,
+      "step": 12865
+    },
+    {
+      "epoch": 0.9558468565487022,
+      "grad_norm": 1.1159205436706543,
+      "learning_rate": 2.799366939170747e-05,
+      "loss": 0.3472,
+      "step": 12870
+    },
+    {
+      "epoch": 0.9562182034238181,
+      "grad_norm": 1.0658485889434814,
+      "learning_rate": 2.797874261074619e-05,
+      "loss": 0.3587,
+      "step": 12875
+    },
+    {
+      "epoch": 0.9565895502989342,
+      "grad_norm": 0.8533363342285156,
+      "learning_rate": 2.7963814752510413e-05,
+      "loss": 0.3248,
+      "step": 12880
+    },
+    {
+      "epoch": 0.9569608971740503,
+      "grad_norm": 0.8562402129173279,
+      "learning_rate": 2.794888582239889e-05,
+      "loss": 0.3433,
+      "step": 12885
+    },
+    {
+      "epoch": 0.9573322440491663,
+      "grad_norm": 0.9736987948417664,
+      "learning_rate": 2.7933955825810704e-05,
+      "loss": 0.3518,
+      "step": 12890
+    },
+    {
+      "epoch": 0.9577035909242824,
+      "grad_norm": 0.722210705280304,
+      "learning_rate": 2.791902476814535e-05,
+      "loss": 0.3424,
+      "step": 12895
+    },
+    {
+      "epoch": 0.9580749377993985,
+      "grad_norm": 3.026717185974121,
+      "learning_rate": 2.790409265480272e-05,
+      "loss": 0.3496,
+      "step": 12900
+    },
+    {
+      "epoch": 0.9584462846745144,
+      "grad_norm": 0.774347722530365,
+      "learning_rate": 2.7889159491183075e-05,
+      "loss": 0.3407,
+      "step": 12905
+    },
+    {
+      "epoch": 0.9588176315496305,
+      "grad_norm": 0.7414963245391846,
+      "learning_rate": 2.7874225282687038e-05,
+      "loss": 0.3331,
+      "step": 12910
+    },
+    {
+      "epoch": 0.9591889784247466,
+      "grad_norm": 0.8461934924125671,
+      "learning_rate": 2.785929003471564e-05,
+      "loss": 0.3518,
+      "step": 12915
+    },
+    {
+      "epoch": 0.9595603252998626,
+      "grad_norm": 0.9420278668403625,
+      "learning_rate": 2.784435375267027e-05,
+      "loss": 0.3518,
+      "step": 12920
+    },
+    {
+      "epoch": 0.9599316721749787,
+      "grad_norm": 0.6645112037658691,
+      "learning_rate": 2.7829416441952693e-05,
+      "loss": 0.3604,
+      "step": 12925
+    },
+    {
+      "epoch": 0.9603030190500947,
+      "grad_norm": 0.8086858987808228,
+      "learning_rate": 2.7814478107965065e-05,
+      "loss": 0.3518,
+      "step": 12930
+    },
+    {
+      "epoch": 0.9606743659252107,
+      "grad_norm": 0.7968643307685852,
+      "learning_rate": 2.779953875610987e-05,
+      "loss": 0.344,
+      "step": 12935
+    },
+    {
+      "epoch": 0.9610457128003268,
+      "grad_norm": 1.1970291137695312,
+      "learning_rate": 2.7784598391790007e-05,
+      "loss": 0.3269,
+      "step": 12940
+    },
+    {
+      "epoch": 0.9614170596754429,
+      "grad_norm": 0.9078313708305359,
+      "learning_rate": 2.776965702040871e-05,
+      "loss": 0.324,
+      "step": 12945
+    },
+    {
+      "epoch": 0.9617884065505589,
+      "grad_norm": 0.7471829056739807,
+      "learning_rate": 2.7754714647369594e-05,
+      "loss": 0.3151,
+      "step": 12950
+    },
+    {
+      "epoch": 0.962159753425675,
+      "grad_norm": 1.185411810874939,
+      "learning_rate": 2.7739771278076616e-05,
+      "loss": 0.3452,
+      "step": 12955
+    },
+    {
+      "epoch": 0.9625311003007909,
+      "grad_norm": 0.7399412393569946,
+      "learning_rate": 2.772482691793412e-05,
+      "loss": 0.3411,
+      "step": 12960
+    },
+    {
+      "epoch": 0.962902447175907,
+      "grad_norm": 0.9101039171218872,
+      "learning_rate": 2.770988157234679e-05,
+      "loss": 0.3404,
+      "step": 12965
+    },
+    {
+      "epoch": 0.9632737940510231,
+      "grad_norm": 0.9467521905899048,
+      "learning_rate": 2.769493524671967e-05,
+      "loss": 0.3403,
+      "step": 12970
+    },
+    {
+      "epoch": 0.9636451409261391,
+      "grad_norm": 2.10199236869812,
+      "learning_rate": 2.7679987946458173e-05,
+      "loss": 0.3473,
+      "step": 12975
+    },
+    {
+      "epoch": 0.9640164878012552,
+      "grad_norm": 0.9057947397232056,
+      "learning_rate": 2.7665039676968034e-05,
+      "loss": 0.3513,
+      "step": 12980
+    },
+    {
+      "epoch": 0.9643878346763712,
+      "grad_norm": 1.0547771453857422,
+      "learning_rate": 2.7650090443655364e-05,
+      "loss": 0.35,
+      "step": 12985
+    },
+    {
+      "epoch": 0.9647591815514872,
+      "grad_norm": 0.9510254263877869,
+      "learning_rate": 2.763514025192661e-05,
+      "loss": 0.3523,
+      "step": 12990
+    },
+    {
+      "epoch": 0.9651305284266033,
+      "grad_norm": 0.816395103931427,
+      "learning_rate": 2.762018910718858e-05,
+      "loss": 0.3532,
+      "step": 12995
+    },
+    {
+      "epoch": 0.9655018753017194,
+      "grad_norm": 1.1673964262008667,
+      "learning_rate": 2.7605237014848405e-05,
+      "loss": 0.3483,
+      "step": 13000
+    },
+    {
+      "epoch": 0.9658732221768354,
+      "grad_norm": 0.9841897487640381,
+      "learning_rate": 2.759028398031358e-05,
+      "loss": 0.3267,
+      "step": 13005
+    },
+    {
+      "epoch": 0.9662445690519514,
+      "grad_norm": 0.8482001423835754,
+      "learning_rate": 2.7575330008991924e-05,
+      "loss": 0.3519,
+      "step": 13010
+    },
+    {
+      "epoch": 0.9666159159270675,
+      "grad_norm": 0.7443394660949707,
+      "learning_rate": 2.7560375106291603e-05,
+      "loss": 0.3301,
+      "step": 13015
+    },
+    {
+      "epoch": 0.9669872628021835,
+      "grad_norm": 0.9038524627685547,
+      "learning_rate": 2.754541927762113e-05,
+      "loss": 0.3627,
+      "step": 13020
+    },
+    {
+      "epoch": 0.9673586096772996,
+      "grad_norm": 0.995887279510498,
+      "learning_rate": 2.7530462528389327e-05,
+      "loss": 0.3343,
+      "step": 13025
+    },
+    {
+      "epoch": 0.9677299565524156,
+      "grad_norm": 0.9123496413230896,
+      "learning_rate": 2.7515504864005372e-05,
+      "loss": 0.2947,
+      "step": 13030
+    },
+    {
+      "epoch": 0.9681013034275316,
+      "grad_norm": 0.6203708052635193,
+      "learning_rate": 2.750054628987876e-05,
+      "loss": 0.3362,
+      "step": 13035
+    },
+    {
+      "epoch": 0.9684726503026477,
+      "grad_norm": 0.9952647686004639,
+      "learning_rate": 2.7485586811419333e-05,
+      "loss": 0.3443,
+      "step": 13040
+    },
+    {
+      "epoch": 0.9688439971777637,
+      "grad_norm": 1.0715820789337158,
+      "learning_rate": 2.7470626434037235e-05,
+      "loss": 0.3621,
+      "step": 13045
+    },
+    {
+      "epoch": 0.9692153440528798,
+      "grad_norm": 0.9002212285995483,
+      "learning_rate": 2.7455665163142946e-05,
+      "loss": 0.3603,
+      "step": 13050
+    },
+    {
+      "epoch": 0.9695866909279959,
+      "grad_norm": 0.8712630271911621,
+      "learning_rate": 2.7440703004147278e-05,
+      "loss": 0.3416,
+      "step": 13055
+    },
+    {
+      "epoch": 0.9699580378031118,
+      "grad_norm": 1.128656029701233,
+      "learning_rate": 2.7425739962461357e-05,
+      "loss": 0.367,
+      "step": 13060
+    },
+    {
+      "epoch": 0.9703293846782279,
+      "grad_norm": 0.9522350430488586,
+      "learning_rate": 2.7410776043496632e-05,
+      "loss": 0.3442,
+      "step": 13065
+    },
+    {
+      "epoch": 0.970700731553344,
+      "grad_norm": 0.844645619392395,
+      "learning_rate": 2.739581125266486e-05,
+      "loss": 0.3489,
+      "step": 13070
+    },
+    {
+      "epoch": 0.97107207842846,
+      "grad_norm": 0.7419381141662598,
+      "learning_rate": 2.7380845595378114e-05,
+      "loss": 0.3405,
+      "step": 13075
+    },
+    {
+      "epoch": 0.9714434253035761,
+      "grad_norm": 0.6905413269996643,
+      "learning_rate": 2.7365879077048794e-05,
+      "loss": 0.3489,
+      "step": 13080
+    },
+    {
+      "epoch": 0.9718147721786922,
+      "grad_norm": 0.9001625776290894,
+      "learning_rate": 2.7350911703089604e-05,
+      "loss": 0.3444,
+      "step": 13085
+    },
+    {
+      "epoch": 0.9721861190538081,
+      "grad_norm": 0.9108909368515015,
+      "learning_rate": 2.7335943478913544e-05,
+      "loss": 0.3484,
+      "step": 13090
+    },
+    {
+      "epoch": 0.9725574659289242,
+      "grad_norm": 1.0819025039672852,
+      "learning_rate": 2.7320974409933952e-05,
+      "loss": 0.3543,
+      "step": 13095
+    },
+    {
+      "epoch": 0.9729288128040402,
+      "grad_norm": 0.7084153890609741,
+      "learning_rate": 2.730600450156444e-05,
+      "loss": 0.3328,
+      "step": 13100
+    },
+    {
+      "epoch": 0.9733001596791563,
+      "grad_norm": 0.8965533971786499,
+      "learning_rate": 2.729103375921894e-05,
+      "loss": 0.3462,
+      "step": 13105
+    },
+    {
+      "epoch": 0.9736715065542724,
+      "grad_norm": 0.8042412996292114,
+      "learning_rate": 2.727606218831168e-05,
+      "loss": 0.3398,
+      "step": 13110
+    },
+    {
+      "epoch": 0.9740428534293883,
+      "grad_norm": 0.8511387705802917,
+      "learning_rate": 2.7261089794257193e-05,
+      "loss": 0.3338,
+      "step": 13115
+    },
+    {
+      "epoch": 0.9744142003045044,
+      "grad_norm": 0.7709136605262756,
+      "learning_rate": 2.724611658247031e-05,
+      "loss": 0.3528,
+      "step": 13120
+    },
+    {
+      "epoch": 0.9747855471796205,
+      "grad_norm": 0.8876646161079407,
+      "learning_rate": 2.7231142558366147e-05,
+      "loss": 0.3516,
+      "step": 13125
+    },
+    {
+      "epoch": 0.9751568940547365,
+      "grad_norm": 0.7100489735603333,
+      "learning_rate": 2.721616772736013e-05,
+      "loss": 0.3352,
+      "step": 13130
+    },
+    {
+      "epoch": 0.9755282409298526,
+      "grad_norm": 1.010686993598938,
+      "learning_rate": 2.720119209486795e-05,
+      "loss": 0.3296,
+      "step": 13135
+    },
+    {
+      "epoch": 0.9758995878049687,
+      "grad_norm": 0.8019929528236389,
+      "learning_rate": 2.7186215666305632e-05,
+      "loss": 0.3338,
+      "step": 13140
+    },
+    {
+      "epoch": 0.9762709346800846,
+      "grad_norm": 0.9189656376838684,
+      "learning_rate": 2.7171238447089447e-05,
+      "loss": 0.3603,
+      "step": 13145
+    },
+    {
+      "epoch": 0.9766422815552007,
+      "grad_norm": 0.7699278593063354,
+      "learning_rate": 2.7156260442635972e-05,
+      "loss": 0.3331,
+      "step": 13150
+    },
+    {
+      "epoch": 0.9770136284303168,
+      "grad_norm": 1.2535781860351562,
+      "learning_rate": 2.714128165836205e-05,
+      "loss": 0.3477,
+      "step": 13155
+    },
+    {
+      "epoch": 0.9773849753054328,
+      "grad_norm": 0.8146220445632935,
+      "learning_rate": 2.7126302099684848e-05,
+      "loss": 0.3317,
+      "step": 13160
+    },
+    {
+      "epoch": 0.9777563221805489,
+      "grad_norm": 0.9014816880226135,
+      "learning_rate": 2.711132177202176e-05,
+      "loss": 0.3539,
+      "step": 13165
+    },
+    {
+      "epoch": 0.978127669055665,
+      "grad_norm": 0.7516365647315979,
+      "learning_rate": 2.7096340680790493e-05,
+      "loss": 0.3425,
+      "step": 13170
+    },
+    {
+      "epoch": 0.9784990159307809,
+      "grad_norm": 0.7220563292503357,
+      "learning_rate": 2.708135883140902e-05,
+      "loss": 0.3408,
+      "step": 13175
+    },
+    {
+      "epoch": 0.978870362805897,
+      "grad_norm": 0.8894364237785339,
+      "learning_rate": 2.7066376229295583e-05,
+      "loss": 0.341,
+      "step": 13180
+    },
+    {
+      "epoch": 0.979241709681013,
+      "grad_norm": 0.8864320516586304,
+      "learning_rate": 2.7051392879868713e-05,
+      "loss": 0.3284,
+      "step": 13185
+    },
+    {
+      "epoch": 0.9796130565561291,
+      "grad_norm": 0.9717690944671631,
+      "learning_rate": 2.7036408788547192e-05,
+      "loss": 0.3369,
+      "step": 13190
+    },
+    {
+      "epoch": 0.9799844034312452,
+      "grad_norm": 0.8697006106376648,
+      "learning_rate": 2.7021423960750076e-05,
+      "loss": 0.3451,
+      "step": 13195
+    },
+    {
+      "epoch": 0.9803557503063611,
+      "grad_norm": 0.8000594973564148,
+      "learning_rate": 2.700643840189669e-05,
+      "loss": 0.3215,
+      "step": 13200
+    },
+    {
+      "epoch": 0.9807270971814772,
+      "grad_norm": 1.589208960533142,
+      "learning_rate": 2.699145211740663e-05,
+      "loss": 0.3277,
+      "step": 13205
+    },
+    {
+      "epoch": 0.9810984440565933,
+      "grad_norm": 1.129319429397583,
+      "learning_rate": 2.6976465112699735e-05,
+      "loss": 0.3415,
+      "step": 13210
+    },
+    {
+      "epoch": 0.9814697909317093,
+      "grad_norm": 1.2226186990737915,
+      "learning_rate": 2.6961477393196126e-05,
+      "loss": 0.3737,
+      "step": 13215
+    },
+    {
+      "epoch": 0.9818411378068254,
+      "grad_norm": 0.8157563209533691,
+      "learning_rate": 2.6946488964316176e-05,
+      "loss": 0.3411,
+      "step": 13220
+    },
+    {
+      "epoch": 0.9822124846819414,
+      "grad_norm": 0.9670863747596741,
+      "learning_rate": 2.69314998314805e-05,
+      "loss": 0.3426,
+      "step": 13225
+    },
+    {
+      "epoch": 0.9825838315570574,
+      "grad_norm": 1.0690017938613892,
+      "learning_rate": 2.6916510000109995e-05,
+      "loss": 0.3566,
+      "step": 13230
+    },
+    {
+      "epoch": 0.9829551784321735,
+      "grad_norm": 0.968449056148529,
+      "learning_rate": 2.690151947562578e-05,
+      "loss": 0.3323,
+      "step": 13235
+    },
+    {
+      "epoch": 0.9833265253072896,
+      "grad_norm": 0.9085156321525574,
+      "learning_rate": 2.6886528263449247e-05,
+      "loss": 0.3504,
+      "step": 13240
+    },
+    {
+      "epoch": 0.9836978721824056,
+      "grad_norm": 0.8193844556808472,
+      "learning_rate": 2.6871536369002026e-05,
+      "loss": 0.3132,
+      "step": 13245
+    },
+    {
+      "epoch": 0.9840692190575216,
+      "grad_norm": 0.6981024146080017,
+      "learning_rate": 2.6856543797706008e-05,
+      "loss": 0.3433,
+      "step": 13250
+    },
+    {
+      "epoch": 0.9844405659326376,
+      "grad_norm": 1.0682088136672974,
+      "learning_rate": 2.6841550554983296e-05,
+      "loss": 0.3288,
+      "step": 13255
+    },
+    {
+      "epoch": 0.9848119128077537,
+      "grad_norm": 1.944576382637024,
+      "learning_rate": 2.682655664625628e-05,
+      "loss": 0.3428,
+      "step": 13260
+    },
+    {
+      "epoch": 0.9851832596828698,
+      "grad_norm": 0.9447241425514221,
+      "learning_rate": 2.681156207694756e-05,
+      "loss": 0.3309,
+      "step": 13265
+    },
+    {
+      "epoch": 0.9855546065579858,
+      "grad_norm": 0.6979337334632874,
+      "learning_rate": 2.6796566852479976e-05,
+      "loss": 0.3218,
+      "step": 13270
+    },
+    {
+      "epoch": 0.9859259534331019,
+      "grad_norm": 0.914198637008667,
+      "learning_rate": 2.6781570978276632e-05,
+      "loss": 0.3447,
+      "step": 13275
+    },
+    {
+      "epoch": 0.9862973003082179,
+      "grad_norm": 0.8899309039115906,
+      "learning_rate": 2.6766574459760828e-05,
+      "loss": 0.3661,
+      "step": 13280
+    },
+    {
+      "epoch": 0.9866686471833339,
+      "grad_norm": 1.063792109489441,
+      "learning_rate": 2.6751577302356122e-05,
+      "loss": 0.3394,
+      "step": 13285
+    },
+    {
+      "epoch": 0.98703999405845,
+      "grad_norm": 0.8097315430641174,
+      "learning_rate": 2.6736579511486297e-05,
+      "loss": 0.3542,
+      "step": 13290
+    },
+    {
+      "epoch": 0.9874113409335661,
+      "grad_norm": 0.6995810270309448,
+      "learning_rate": 2.672158109257538e-05,
+      "loss": 0.3578,
+      "step": 13295
+    },
+    {
+      "epoch": 0.9877826878086821,
+      "grad_norm": 0.7139712572097778,
+      "learning_rate": 2.670658205104759e-05,
+      "loss": 0.3285,
+      "step": 13300
+    },
+    {
+      "epoch": 0.9881540346837981,
+      "grad_norm": 0.9237804412841797,
+      "learning_rate": 2.6691582392327397e-05,
+      "loss": 0.3341,
+      "step": 13305
+    },
+    {
+      "epoch": 0.9885253815589142,
+      "grad_norm": 0.980902910232544,
+      "learning_rate": 2.6676582121839495e-05,
+      "loss": 0.333,
+      "step": 13310
+    },
+    {
+      "epoch": 0.9888967284340302,
+      "grad_norm": 0.8618834018707275,
+      "learning_rate": 2.6661581245008793e-05,
+      "loss": 0.3346,
+      "step": 13315
+    },
+    {
+      "epoch": 0.9892680753091463,
+      "grad_norm": 0.8352036476135254,
+      "learning_rate": 2.6646579767260417e-05,
+      "loss": 0.3466,
+      "step": 13320
+    },
+    {
+      "epoch": 0.9896394221842624,
+      "grad_norm": 1.0906356573104858,
+      "learning_rate": 2.6631577694019706e-05,
+      "loss": 0.3279,
+      "step": 13325
+    },
+    {
+      "epoch": 0.9900107690593783,
+      "grad_norm": 1.1875203847885132,
+      "learning_rate": 2.6616575030712243e-05,
+      "loss": 0.3378,
+      "step": 13330
+    },
+    {
+      "epoch": 0.9903821159344944,
+      "grad_norm": 0.7007048726081848,
+      "learning_rate": 2.660157178276378e-05,
+      "loss": 0.3276,
+      "step": 13335
+    },
+    {
+      "epoch": 0.9907534628096104,
+      "grad_norm": 1.1638414859771729,
+      "learning_rate": 2.658656795560031e-05,
+      "loss": 0.313,
+      "step": 13340
+    },
+    {
+      "epoch": 0.9911248096847265,
+      "grad_norm": 0.9487748742103577,
+      "learning_rate": 2.6571563554648033e-05,
+      "loss": 0.3097,
+      "step": 13345
+    },
+    {
+      "epoch": 0.9914961565598426,
+      "grad_norm": 0.878073513507843,
+      "learning_rate": 2.6556558585333347e-05,
+      "loss": 0.3296,
+      "step": 13350
+    },
+    {
+      "epoch": 0.9918675034349586,
+      "grad_norm": 0.6810947060585022,
+      "learning_rate": 2.6541553053082862e-05,
+      "loss": 0.3447,
+      "step": 13355
+    },
+    {
+      "epoch": 0.9922388503100746,
+      "grad_norm": 0.7682046890258789,
+      "learning_rate": 2.6526546963323395e-05,
+      "loss": 0.33,
+      "step": 13360
+    },
+    {
+      "epoch": 0.9926101971851907,
+      "grad_norm": 0.7372832298278809,
+      "learning_rate": 2.6511540321481958e-05,
+      "loss": 0.3364,
+      "step": 13365
+    },
+    {
+      "epoch": 0.9929815440603067,
+      "grad_norm": 1.080094575881958,
+      "learning_rate": 2.6496533132985758e-05,
+      "loss": 0.3333,
+      "step": 13370
+    },
+    {
+      "epoch": 0.9933528909354228,
+      "grad_norm": 0.8818208575248718,
+      "learning_rate": 2.6481525403262214e-05,
+      "loss": 0.3462,
+      "step": 13375
+    },
+    {
+      "epoch": 0.9937242378105389,
+      "grad_norm": 0.8809915781021118,
+      "learning_rate": 2.6466517137738932e-05,
+      "loss": 0.3592,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9940955846856548,
+      "grad_norm": 1.348819613456726,
+      "learning_rate": 2.645150834184371e-05,
+      "loss": 0.33,
+      "step": 13385
+    },
+    {
+      "epoch": 0.9944669315607709,
+      "grad_norm": 0.714433491230011,
+      "learning_rate": 2.6436499021004534e-05,
+      "loss": 0.3374,
+      "step": 13390
+    },
+    {
+      "epoch": 0.994838278435887,
+      "grad_norm": 0.7457917928695679,
+      "learning_rate": 2.6421489180649604e-05,
+      "loss": 0.3482,
+      "step": 13395
+    },
+    {
+      "epoch": 0.995209625311003,
+      "grad_norm": 0.7141998410224915,
+      "learning_rate": 2.6406478826207275e-05,
+      "loss": 0.3359,
+      "step": 13400
+    },
+    {
+      "epoch": 0.9955809721861191,
+      "grad_norm": 0.9509761333465576,
+      "learning_rate": 2.6391467963106113e-05,
+      "loss": 0.3523,
+      "step": 13405
+    },
+    {
+      "epoch": 0.995952319061235,
+      "grad_norm": 0.9249238967895508,
+      "learning_rate": 2.6376456596774858e-05,
+      "loss": 0.3352,
+      "step": 13410
+    },
+    {
+      "epoch": 0.9963236659363511,
+      "grad_norm": 0.913701057434082,
+      "learning_rate": 2.636144473264241e-05,
+      "loss": 0.3387,
+      "step": 13415
+    },
+    {
+      "epoch": 0.9966950128114672,
+      "grad_norm": 0.7925592660903931,
+      "learning_rate": 2.634643237613791e-05,
+      "loss": 0.3378,
+      "step": 13420
+    },
+    {
+      "epoch": 0.9970663596865832,
+      "grad_norm": 1.0819846391677856,
+      "learning_rate": 2.6331419532690603e-05,
+      "loss": 0.3384,
+      "step": 13425
+    },
+    {
+      "epoch": 0.9974377065616993,
+      "grad_norm": 0.7865422964096069,
+      "learning_rate": 2.6316406207729972e-05,
+      "loss": 0.347,
+      "step": 13430
+    },
+    {
+      "epoch": 0.9978090534368154,
+      "grad_norm": 0.7669934034347534,
+      "learning_rate": 2.630139240668562e-05,
+      "loss": 0.3611,
+      "step": 13435
+    },
+    {
+      "epoch": 0.9981804003119313,
+      "grad_norm": 0.7105921506881714,
+      "learning_rate": 2.628637813498738e-05,
+      "loss": 0.3548,
+      "step": 13440
+    },
+    {
+      "epoch": 0.9985517471870474,
+      "grad_norm": 0.9623938798904419,
+      "learning_rate": 2.6271363398065206e-05,
+      "loss": 0.3565,
+      "step": 13445
+    },
+    {
+      "epoch": 0.9989230940621635,
+      "grad_norm": 0.8119087219238281,
+      "learning_rate": 2.625634820134924e-05,
+      "loss": 0.3341,
+      "step": 13450
+    },
+    {
+      "epoch": 0.9992944409372795,
+      "grad_norm": 0.8674593567848206,
+      "learning_rate": 2.6241332550269794e-05,
+      "loss": 0.3478,
+      "step": 13455
+    },
+    {
+      "epoch": 0.9996657878123956,
+      "grad_norm": 1.0126354694366455,
+      "learning_rate": 2.6226316450257338e-05,
+      "loss": 0.3165,
+      "step": 13460
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 26928,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.530315460862699e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}