InternVL-reasoning-2b / trainer_state.json
dyang39's picture
Upload trainer_state.json with huggingface_hub
e16b720 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9756244357508277,
"eval_steps": 200,
"global_step": 309,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.000000000000001e-06,
"loss": 3.5752,
"step": 1
},
{
"epoch": 0.02,
"learning_rate": 8.000000000000001e-06,
"loss": 3.4389,
"step": 2
},
{
"epoch": 0.03,
"learning_rate": 1.2e-05,
"loss": 2.2415,
"step": 3
},
{
"epoch": 0.04,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.6566,
"step": 4
},
{
"epoch": 0.05,
"learning_rate": 2e-05,
"loss": 1.5019,
"step": 5
},
{
"epoch": 0.06,
"learning_rate": 2.4e-05,
"loss": 1.409,
"step": 6
},
{
"epoch": 0.07,
"learning_rate": 2.8e-05,
"loss": 1.4669,
"step": 7
},
{
"epoch": 0.08,
"learning_rate": 3.2000000000000005e-05,
"loss": 1.9948,
"step": 8
},
{
"epoch": 0.09,
"learning_rate": 3.6e-05,
"loss": 1.3069,
"step": 9
},
{
"epoch": 0.1,
"learning_rate": 4e-05,
"loss": 1.276,
"step": 10
},
{
"epoch": 0.11,
"learning_rate": 3.9998896039909675e-05,
"loss": 1.2514,
"step": 11
},
{
"epoch": 0.12,
"learning_rate": 3.999558428151149e-05,
"loss": 1.4888,
"step": 12
},
{
"epoch": 0.13,
"learning_rate": 3.999006509041036e-05,
"loss": 1.4359,
"step": 13
},
{
"epoch": 0.13,
"learning_rate": 3.998233907590294e-05,
"loss": 1.225,
"step": 14
},
{
"epoch": 0.14,
"learning_rate": 3.997240709091041e-05,
"loss": 1.2955,
"step": 15
},
{
"epoch": 0.15,
"learning_rate": 3.996027023188427e-05,
"loss": 1.2106,
"step": 16
},
{
"epoch": 0.16,
"learning_rate": 3.994592983868533e-05,
"loss": 1.1727,
"step": 17
},
{
"epoch": 0.17,
"learning_rate": 3.9929387494435746e-05,
"loss": 1.21,
"step": 18
},
{
"epoch": 0.18,
"learning_rate": 3.9910645025344324e-05,
"loss": 1.1913,
"step": 19
},
{
"epoch": 0.19,
"learning_rate": 3.988970450050483e-05,
"loss": 1.1689,
"step": 20
},
{
"epoch": 0.2,
"learning_rate": 3.986656823166766e-05,
"loss": 1.1868,
"step": 21
},
{
"epoch": 0.21,
"learning_rate": 3.984123877298454e-05,
"loss": 1.139,
"step": 22
},
{
"epoch": 0.22,
"learning_rate": 3.981371892072661e-05,
"loss": 1.1681,
"step": 23
},
{
"epoch": 0.23,
"learning_rate": 3.978401171297576e-05,
"loss": 1.1059,
"step": 24
},
{
"epoch": 0.24,
"learning_rate": 3.9752120429289135e-05,
"loss": 1.1138,
"step": 25
},
{
"epoch": 0.25,
"learning_rate": 3.9718048590337186e-05,
"loss": 1.0705,
"step": 26
},
{
"epoch": 0.26,
"learning_rate": 3.9681799957514965e-05,
"loss": 1.1247,
"step": 27
},
{
"epoch": 0.27,
"learning_rate": 3.964337853252685e-05,
"loss": 1.1524,
"step": 28
},
{
"epoch": 0.28,
"learning_rate": 3.960278855694484e-05,
"loss": 1.0491,
"step": 29
},
{
"epoch": 0.29,
"learning_rate": 3.956003451174023e-05,
"loss": 1.1492,
"step": 30
},
{
"epoch": 0.3,
"learning_rate": 3.9515121116788985e-05,
"loss": 1.1623,
"step": 31
},
{
"epoch": 0.31,
"learning_rate": 3.946805333035067e-05,
"loss": 1.1024,
"step": 32
},
{
"epoch": 0.32,
"learning_rate": 3.9418836348521045e-05,
"loss": 1.0908,
"step": 33
},
{
"epoch": 0.33,
"learning_rate": 3.9367475604658485e-05,
"loss": 1.1505,
"step": 34
},
{
"epoch": 0.34,
"learning_rate": 3.931397676878415e-05,
"loss": 1.0795,
"step": 35
},
{
"epoch": 0.35,
"learning_rate": 3.925834574695599e-05,
"loss": 1.0738,
"step": 36
},
{
"epoch": 0.36,
"learning_rate": 3.92005886806168e-05,
"loss": 1.0591,
"step": 37
},
{
"epoch": 0.37,
"learning_rate": 3.9140711945916197e-05,
"loss": 1.0694,
"step": 38
},
{
"epoch": 0.38,
"learning_rate": 3.907872215300672e-05,
"loss": 1.0729,
"step": 39
},
{
"epoch": 0.39,
"learning_rate": 3.9014626145314115e-05,
"loss": 1.1091,
"step": 40
},
{
"epoch": 0.39,
"learning_rate": 3.8948430998781824e-05,
"loss": 1.0049,
"step": 41
},
{
"epoch": 0.4,
"learning_rate": 3.888014402108984e-05,
"loss": 1.0333,
"step": 42
},
{
"epoch": 0.41,
"learning_rate": 3.8809772750847964e-05,
"loss": 1.0432,
"step": 43
},
{
"epoch": 0.42,
"learning_rate": 3.87373249567636e-05,
"loss": 0.9931,
"step": 44
},
{
"epoch": 0.43,
"learning_rate": 3.8662808636784056e-05,
"loss": 1.0425,
"step": 45
},
{
"epoch": 0.44,
"learning_rate": 3.8586232017213675e-05,
"loss": 1.0287,
"step": 46
},
{
"epoch": 0.45,
"learning_rate": 3.850760355180564e-05,
"loss": 0.9849,
"step": 47
},
{
"epoch": 0.46,
"learning_rate": 3.842693192082874e-05,
"loss": 1.0096,
"step": 48
},
{
"epoch": 0.47,
"learning_rate": 3.834422603010906e-05,
"loss": 0.9763,
"step": 49
},
{
"epoch": 0.48,
"learning_rate": 3.8259495010046876e-05,
"loss": 0.9852,
"step": 50
},
{
"epoch": 0.49,
"learning_rate": 3.8172748214608624e-05,
"loss": 1.0338,
"step": 51
},
{
"epoch": 0.5,
"learning_rate": 3.808399522029433e-05,
"loss": 0.9748,
"step": 52
},
{
"epoch": 0.51,
"learning_rate": 3.7993245825080364e-05,
"loss": 0.9784,
"step": 53
},
{
"epoch": 0.52,
"learning_rate": 3.790051004733775e-05,
"loss": 0.9736,
"step": 54
},
{
"epoch": 0.53,
"learning_rate": 3.7805798124726283e-05,
"loss": 1.0221,
"step": 55
},
{
"epoch": 0.54,
"learning_rate": 3.7709120513064196e-05,
"loss": 0.9352,
"step": 56
},
{
"epoch": 0.55,
"learning_rate": 3.7610487885174e-05,
"loss": 0.9668,
"step": 57
},
{
"epoch": 0.56,
"learning_rate": 3.7509911129704165e-05,
"loss": 0.9478,
"step": 58
},
{
"epoch": 0.57,
"learning_rate": 3.74074013499271e-05,
"loss": 0.941,
"step": 59
},
{
"epoch": 0.58,
"learning_rate": 3.730296986251337e-05,
"loss": 0.9212,
"step": 60
},
{
"epoch": 0.59,
"learning_rate": 3.7196628196282415e-05,
"loss": 0.8714,
"step": 61
},
{
"epoch": 0.6,
"learning_rate": 3.7088388090929776e-05,
"loss": 0.948,
"step": 62
},
{
"epoch": 0.61,
"learning_rate": 3.69782614957311e-05,
"loss": 0.9212,
"step": 63
},
{
"epoch": 0.62,
"learning_rate": 3.686626056822298e-05,
"loss": 0.8565,
"step": 64
},
{
"epoch": 0.63,
"learning_rate": 3.675239767286083e-05,
"loss": 0.9639,
"step": 65
},
{
"epoch": 0.64,
"learning_rate": 3.6636685379653875e-05,
"loss": 0.877,
"step": 66
},
{
"epoch": 0.65,
"learning_rate": 3.651913646277747e-05,
"loss": 0.8497,
"step": 67
},
{
"epoch": 0.65,
"learning_rate": 3.639976389916292e-05,
"loss": 0.8182,
"step": 68
},
{
"epoch": 0.66,
"learning_rate": 3.627858086706483e-05,
"loss": 0.853,
"step": 69
},
{
"epoch": 0.67,
"learning_rate": 3.61556007446063e-05,
"loss": 0.7592,
"step": 70
},
{
"epoch": 0.68,
"learning_rate": 3.603083710830205e-05,
"loss": 0.9022,
"step": 71
},
{
"epoch": 0.69,
"learning_rate": 3.5904303731559584e-05,
"loss": 0.8494,
"step": 72
},
{
"epoch": 0.7,
"learning_rate": 3.5776014583158714e-05,
"loss": 0.7637,
"step": 73
},
{
"epoch": 0.71,
"learning_rate": 3.564598382570943e-05,
"loss": 0.8176,
"step": 74
},
{
"epoch": 0.72,
"learning_rate": 3.55142258140884e-05,
"loss": 0.8738,
"step": 75
},
{
"epoch": 0.73,
"learning_rate": 3.538075509385427e-05,
"loss": 0.7815,
"step": 76
},
{
"epoch": 0.74,
"learning_rate": 3.524558639964187e-05,
"loss": 0.8885,
"step": 77
},
{
"epoch": 0.75,
"learning_rate": 3.5108734653535587e-05,
"loss": 0.8237,
"step": 78
},
{
"epoch": 0.76,
"learning_rate": 3.497021496342203e-05,
"loss": 0.7413,
"step": 79
},
{
"epoch": 0.77,
"learning_rate": 3.4830042621322145e-05,
"loss": 0.7746,
"step": 80
},
{
"epoch": 0.78,
"learning_rate": 3.468823310170309e-05,
"loss": 0.8366,
"step": 81
},
{
"epoch": 0.79,
"learning_rate": 3.4544802059769864e-05,
"loss": 0.76,
"step": 82
},
{
"epoch": 0.8,
"learning_rate": 3.439976532973708e-05,
"loss": 0.8352,
"step": 83
},
{
"epoch": 0.81,
"learning_rate": 3.425313892308089e-05,
"loss": 0.7205,
"step": 84
},
{
"epoch": 0.82,
"learning_rate": 3.41049390267714e-05,
"loss": 0.7648,
"step": 85
},
{
"epoch": 0.83,
"learning_rate": 3.395518200148571e-05,
"loss": 0.8244,
"step": 86
},
{
"epoch": 0.84,
"learning_rate": 3.380388437980173e-05,
"loss": 0.7936,
"step": 87
},
{
"epoch": 0.85,
"learning_rate": 3.365106286437309e-05,
"loss": 0.7051,
"step": 88
},
{
"epoch": 0.86,
"learning_rate": 3.3496734326085154e-05,
"loss": 0.7281,
"step": 89
},
{
"epoch": 0.87,
"learning_rate": 3.3340915802192655e-05,
"loss": 0.8165,
"step": 90
},
{
"epoch": 0.88,
"learning_rate": 3.318362449443876e-05,
"loss": 0.7686,
"step": 91
},
{
"epoch": 0.89,
"learning_rate": 3.3024877767156094e-05,
"loss": 0.7015,
"step": 92
},
{
"epoch": 0.9,
"learning_rate": 3.2864693145349795e-05,
"loss": 0.707,
"step": 93
},
{
"epoch": 0.91,
"learning_rate": 3.2703088312762825e-05,
"loss": 0.6862,
"step": 94
},
{
"epoch": 0.91,
"learning_rate": 3.254008110992373e-05,
"loss": 0.646,
"step": 95
},
{
"epoch": 0.92,
"learning_rate": 3.237568953217717e-05,
"loss": 0.5646,
"step": 96
},
{
"epoch": 0.93,
"learning_rate": 3.220993172769723e-05,
"loss": 0.6573,
"step": 97
},
{
"epoch": 0.94,
"learning_rate": 3.204282599548399e-05,
"loss": 0.631,
"step": 98
},
{
"epoch": 0.95,
"learning_rate": 3.187439078334338e-05,
"loss": 0.5593,
"step": 99
},
{
"epoch": 0.96,
"learning_rate": 3.170464468585059e-05,
"loss": 0.6312,
"step": 100
},
{
"epoch": 0.97,
"learning_rate": 3.153360644229735e-05,
"loss": 0.6508,
"step": 101
},
{
"epoch": 0.98,
"learning_rate": 3.136129493462312e-05,
"loss": 0.6596,
"step": 102
},
{
"epoch": 0.99,
"learning_rate": 3.1187729185330665e-05,
"loss": 0.6439,
"step": 103
},
{
"epoch": 1.0,
"learning_rate": 3.101292835538602e-05,
"loss": 0.6206,
"step": 104
},
{
"epoch": 1.01,
"learning_rate": 3.083691174210318e-05,
"loss": 0.486,
"step": 105
},
{
"epoch": 1.02,
"learning_rate": 3.065969877701378e-05,
"loss": 0.5089,
"step": 106
},
{
"epoch": 1.03,
"learning_rate": 3.04813090237219e-05,
"loss": 0.4347,
"step": 107
},
{
"epoch": 1.04,
"learning_rate": 3.030176217574437e-05,
"loss": 0.4901,
"step": 108
},
{
"epoch": 1.05,
"learning_rate": 3.0121078054336633e-05,
"loss": 0.4452,
"step": 109
},
{
"epoch": 1.06,
"learning_rate": 2.9939276606304592e-05,
"loss": 0.4698,
"step": 110
},
{
"epoch": 1.07,
"learning_rate": 2.975637790180255e-05,
"loss": 0.4264,
"step": 111
},
{
"epoch": 1.08,
"learning_rate": 2.9572402132117533e-05,
"loss": 0.4965,
"step": 112
},
{
"epoch": 1.09,
"learning_rate": 2.9387369607440277e-05,
"loss": 0.4725,
"step": 113
},
{
"epoch": 1.1,
"learning_rate": 2.9201300754623046e-05,
"loss": 0.4696,
"step": 114
},
{
"epoch": 1.11,
"learning_rate": 2.9014216114924595e-05,
"loss": 0.4647,
"step": 115
},
{
"epoch": 1.12,
"learning_rate": 2.8826136341742504e-05,
"loss": 0.4576,
"step": 116
},
{
"epoch": 1.13,
"learning_rate": 2.863708219833311e-05,
"loss": 0.4064,
"step": 117
},
{
"epoch": 1.14,
"learning_rate": 2.8447074555519335e-05,
"loss": 0.5083,
"step": 118
},
{
"epoch": 1.15,
"learning_rate": 2.825613438938663e-05,
"loss": 0.4381,
"step": 119
},
{
"epoch": 1.16,
"learning_rate": 2.8064282778967296e-05,
"loss": 0.4829,
"step": 120
},
{
"epoch": 1.17,
"learning_rate": 2.7871540903913465e-05,
"loss": 0.4532,
"step": 121
},
{
"epoch": 1.17,
"learning_rate": 2.7677930042158893e-05,
"loss": 0.4142,
"step": 122
},
{
"epoch": 1.18,
"learning_rate": 2.7483471567570045e-05,
"loss": 0.4921,
"step": 123
},
{
"epoch": 1.19,
"learning_rate": 2.7288186947586426e-05,
"loss": 0.4045,
"step": 124
},
{
"epoch": 1.2,
"learning_rate": 2.7092097740850712e-05,
"loss": 0.4199,
"step": 125
},
{
"epoch": 1.21,
"learning_rate": 2.6895225594828743e-05,
"loss": 0.4419,
"step": 126
},
{
"epoch": 1.22,
"learning_rate": 2.6697592243419723e-05,
"loss": 0.4105,
"step": 127
},
{
"epoch": 1.23,
"learning_rate": 2.6499219504556907e-05,
"loss": 0.3799,
"step": 128
},
{
"epoch": 1.24,
"learning_rate": 2.630012927779896e-05,
"loss": 0.4399,
"step": 129
},
{
"epoch": 1.25,
"learning_rate": 2.6100343541912353e-05,
"loss": 0.3672,
"step": 130
},
{
"epoch": 1.26,
"learning_rate": 2.5899884352444994e-05,
"loss": 0.4002,
"step": 131
},
{
"epoch": 1.27,
"learning_rate": 2.5698773839291373e-05,
"loss": 0.4816,
"step": 132
},
{
"epoch": 1.28,
"learning_rate": 2.549703420424951e-05,
"loss": 0.4849,
"step": 133
},
{
"epoch": 1.29,
"learning_rate": 2.5294687718569994e-05,
"loss": 0.3788,
"step": 134
},
{
"epoch": 1.3,
"learning_rate": 2.5091756720497266e-05,
"loss": 0.3748,
"step": 135
},
{
"epoch": 1.31,
"learning_rate": 2.4888263612803637e-05,
"loss": 0.3789,
"step": 136
},
{
"epoch": 1.32,
"learning_rate": 2.468423086031605e-05,
"loss": 0.4272,
"step": 137
},
{
"epoch": 1.33,
"learning_rate": 2.44796809874361e-05,
"loss": 0.4927,
"step": 138
},
{
"epoch": 1.34,
"learning_rate": 2.4274636575653398e-05,
"loss": 0.4966,
"step": 139
},
{
"epoch": 1.35,
"learning_rate": 2.4069120261052682e-05,
"loss": 0.2791,
"step": 140
},
{
"epoch": 1.36,
"learning_rate": 2.3863154731814867e-05,
"loss": 0.4203,
"step": 141
},
{
"epoch": 1.37,
"learning_rate": 2.365676272571239e-05,
"loss": 0.389,
"step": 142
},
{
"epoch": 1.38,
"learning_rate": 2.344996702759901e-05,
"loss": 0.4018,
"step": 143
},
{
"epoch": 1.39,
"learning_rate": 2.3242790466894494e-05,
"loss": 0.4103,
"step": 144
},
{
"epoch": 1.4,
"learning_rate": 2.303525591506431e-05,
"loss": 0.3833,
"step": 145
},
{
"epoch": 1.41,
"learning_rate": 2.2827386283094707e-05,
"loss": 0.3272,
"step": 146
},
{
"epoch": 1.42,
"learning_rate": 2.2619204518963452e-05,
"loss": 0.3897,
"step": 147
},
{
"epoch": 1.43,
"learning_rate": 2.2410733605106462e-05,
"loss": 0.3893,
"step": 148
},
{
"epoch": 1.43,
"learning_rate": 2.2201996555880633e-05,
"loss": 0.3954,
"step": 149
},
{
"epoch": 1.44,
"learning_rate": 2.199301641502313e-05,
"loss": 0.4332,
"step": 150
},
{
"epoch": 1.45,
"learning_rate": 2.178381625310748e-05,
"loss": 0.4815,
"step": 151
},
{
"epoch": 1.46,
"learning_rate": 2.157441916499663e-05,
"loss": 0.4959,
"step": 152
},
{
"epoch": 1.47,
"learning_rate": 2.1364848267293424e-05,
"loss": 0.2849,
"step": 153
},
{
"epoch": 1.48,
"learning_rate": 2.115512669578857e-05,
"loss": 0.4259,
"step": 154
},
{
"epoch": 1.49,
"learning_rate": 2.0945277602906574e-05,
"loss": 0.4646,
"step": 155
},
{
"epoch": 1.5,
"learning_rate": 2.0735324155149795e-05,
"loss": 0.3296,
"step": 156
},
{
"epoch": 1.51,
"learning_rate": 2.0525289530540945e-05,
"loss": 0.3316,
"step": 157
},
{
"epoch": 1.52,
"learning_rate": 2.031519691606433e-05,
"loss": 0.2673,
"step": 158
},
{
"epoch": 1.53,
"learning_rate": 2.0105069505106126e-05,
"loss": 0.3335,
"step": 159
},
{
"epoch": 1.54,
"learning_rate": 1.9894930494893884e-05,
"loss": 0.4136,
"step": 160
},
{
"epoch": 1.55,
"learning_rate": 1.9684803083935676e-05,
"loss": 0.3702,
"step": 161
},
{
"epoch": 1.56,
"learning_rate": 1.9474710469459062e-05,
"loss": 0.322,
"step": 162
},
{
"epoch": 1.57,
"learning_rate": 1.926467584485021e-05,
"loss": 0.2832,
"step": 163
},
{
"epoch": 1.58,
"learning_rate": 1.905472239709343e-05,
"loss": 0.3917,
"step": 164
},
{
"epoch": 1.59,
"learning_rate": 1.8844873304211435e-05,
"loss": 0.3641,
"step": 165
},
{
"epoch": 1.6,
"learning_rate": 1.8635151732706586e-05,
"loss": 0.3516,
"step": 166
},
{
"epoch": 1.61,
"learning_rate": 1.8425580835003376e-05,
"loss": 0.2568,
"step": 167
},
{
"epoch": 1.62,
"learning_rate": 1.8216183746892528e-05,
"loss": 0.4459,
"step": 168
},
{
"epoch": 1.63,
"learning_rate": 1.8006983584976877e-05,
"loss": 0.4058,
"step": 169
},
{
"epoch": 1.64,
"learning_rate": 1.7798003444119374e-05,
"loss": 0.4351,
"step": 170
},
{
"epoch": 1.65,
"learning_rate": 1.758926639489354e-05,
"loss": 0.4268,
"step": 171
},
{
"epoch": 1.66,
"learning_rate": 1.7380795481036558e-05,
"loss": 0.2704,
"step": 172
},
{
"epoch": 1.67,
"learning_rate": 1.7172613716905306e-05,
"loss": 0.3372,
"step": 173
},
{
"epoch": 1.68,
"learning_rate": 1.69647440849357e-05,
"loss": 0.3816,
"step": 174
},
{
"epoch": 1.69,
"learning_rate": 1.6757209533105512e-05,
"loss": 0.3816,
"step": 175
},
{
"epoch": 1.69,
"learning_rate": 1.6550032972400996e-05,
"loss": 0.3937,
"step": 176
},
{
"epoch": 1.7,
"learning_rate": 1.634323727428762e-05,
"loss": 0.3147,
"step": 177
},
{
"epoch": 1.71,
"learning_rate": 1.613684526818514e-05,
"loss": 0.3302,
"step": 178
},
{
"epoch": 1.72,
"learning_rate": 1.5930879738947328e-05,
"loss": 0.4236,
"step": 179
},
{
"epoch": 1.73,
"learning_rate": 1.572536342434661e-05,
"loss": 0.2248,
"step": 180
},
{
"epoch": 1.74,
"learning_rate": 1.552031901256391e-05,
"loss": 0.3379,
"step": 181
},
{
"epoch": 1.75,
"learning_rate": 1.531576913968396e-05,
"loss": 0.3467,
"step": 182
},
{
"epoch": 1.76,
"learning_rate": 1.5111736387196377e-05,
"loss": 0.3314,
"step": 183
},
{
"epoch": 1.77,
"learning_rate": 1.4908243279502741e-05,
"loss": 0.3386,
"step": 184
},
{
"epoch": 1.78,
"learning_rate": 1.4705312281430013e-05,
"loss": 0.3905,
"step": 185
},
{
"epoch": 1.79,
"learning_rate": 1.4502965795750487e-05,
"loss": 0.4172,
"step": 186
},
{
"epoch": 1.8,
"learning_rate": 1.430122616070863e-05,
"loss": 0.3216,
"step": 187
},
{
"epoch": 1.81,
"learning_rate": 1.4100115647555006e-05,
"loss": 0.2909,
"step": 188
},
{
"epoch": 1.82,
"learning_rate": 1.3899656458087647e-05,
"loss": 0.272,
"step": 189
},
{
"epoch": 1.83,
"learning_rate": 1.3699870722201043e-05,
"loss": 0.3334,
"step": 190
},
{
"epoch": 1.84,
"learning_rate": 1.3500780495443098e-05,
"loss": 0.3541,
"step": 191
},
{
"epoch": 1.85,
"learning_rate": 1.3302407756580278e-05,
"loss": 0.3599,
"step": 192
},
{
"epoch": 1.86,
"learning_rate": 1.3104774405171259e-05,
"loss": 0.2508,
"step": 193
},
{
"epoch": 1.87,
"learning_rate": 1.2907902259149287e-05,
"loss": 0.3862,
"step": 194
},
{
"epoch": 1.88,
"learning_rate": 1.2711813052413575e-05,
"loss": 0.3328,
"step": 195
},
{
"epoch": 1.89,
"learning_rate": 1.2516528432429955e-05,
"loss": 0.2619,
"step": 196
},
{
"epoch": 1.9,
"learning_rate": 1.232206995784111e-05,
"loss": 0.3766,
"step": 197
},
{
"epoch": 1.91,
"learning_rate": 1.2128459096086544e-05,
"loss": 0.3733,
"step": 198
},
{
"epoch": 1.92,
"learning_rate": 1.1935717221032707e-05,
"loss": 0.2327,
"step": 199
},
{
"epoch": 1.93,
"learning_rate": 1.1743865610613377e-05,
"loss": 0.3572,
"step": 200
},
{
"epoch": 1.94,
"learning_rate": 1.1552925444480674e-05,
"loss": 0.1715,
"step": 201
},
{
"epoch": 1.95,
"learning_rate": 1.1362917801666895e-05,
"loss": 0.2646,
"step": 202
},
{
"epoch": 1.95,
"learning_rate": 1.11738636582575e-05,
"loss": 0.2298,
"step": 203
},
{
"epoch": 1.96,
"learning_rate": 1.0985783885075407e-05,
"loss": 0.282,
"step": 204
},
{
"epoch": 1.97,
"learning_rate": 1.0798699245376959e-05,
"loss": 0.3746,
"step": 205
},
{
"epoch": 1.98,
"learning_rate": 1.0612630392559728e-05,
"loss": 0.318,
"step": 206
},
{
"epoch": 1.99,
"learning_rate": 1.0427597867882474e-05,
"loss": 0.2953,
"step": 207
},
{
"epoch": 2.0,
"learning_rate": 1.0243622098197456e-05,
"loss": 0.2542,
"step": 208
},
{
"epoch": 2.01,
"learning_rate": 1.0060723393695411e-05,
"loss": 0.2674,
"step": 209
},
{
"epoch": 2.02,
"learning_rate": 9.878921945663368e-06,
"loss": 0.2679,
"step": 210
},
{
"epoch": 2.03,
"learning_rate": 9.698237824255634e-06,
"loss": 0.2404,
"step": 211
},
{
"epoch": 2.04,
"learning_rate": 9.518690976278108e-06,
"loss": 0.227,
"step": 212
},
{
"epoch": 2.05,
"learning_rate": 9.340301222986232e-06,
"loss": 0.1782,
"step": 213
},
{
"epoch": 2.06,
"learning_rate": 9.163088257896825e-06,
"loss": 0.2436,
"step": 214
},
{
"epoch": 2.07,
"learning_rate": 8.987071644613985e-06,
"loss": 0.253,
"step": 215
},
{
"epoch": 2.08,
"learning_rate": 8.812270814669338e-06,
"loss": 0.2591,
"step": 216
},
{
"epoch": 2.09,
"learning_rate": 8.638705065376887e-06,
"loss": 0.2712,
"step": 217
},
{
"epoch": 2.1,
"learning_rate": 8.466393557702659e-06,
"loss": 0.2411,
"step": 218
},
{
"epoch": 2.11,
"learning_rate": 8.295355314149413e-06,
"loss": 0.2139,
"step": 219
},
{
"epoch": 2.12,
"learning_rate": 8.125609216656627e-06,
"loss": 0.2237,
"step": 220
},
{
"epoch": 2.13,
"learning_rate": 7.957174004516015e-06,
"loss": 0.1817,
"step": 221
},
{
"epoch": 2.14,
"learning_rate": 7.790068272302776e-06,
"loss": 0.1816,
"step": 222
},
{
"epoch": 2.15,
"learning_rate": 7.624310467822833e-06,
"loss": 0.2413,
"step": 223
},
{
"epoch": 2.16,
"learning_rate": 7.459918890076272e-06,
"loss": 0.2405,
"step": 224
},
{
"epoch": 2.17,
"learning_rate": 7.296911687237187e-06,
"loss": 0.2459,
"step": 225
},
{
"epoch": 2.18,
"learning_rate": 7.1353068546502144e-06,
"loss": 0.2203,
"step": 226
},
{
"epoch": 2.19,
"learning_rate": 6.975122232843916e-06,
"loss": 0.1892,
"step": 227
},
{
"epoch": 2.2,
"learning_rate": 6.816375505561248e-06,
"loss": 0.191,
"step": 228
},
{
"epoch": 2.21,
"learning_rate": 6.659084197807348e-06,
"loss": 0.2545,
"step": 229
},
{
"epoch": 2.21,
"learning_rate": 6.503265673914849e-06,
"loss": 0.2312,
"step": 230
},
{
"epoch": 2.22,
"learning_rate": 6.348937135626922e-06,
"loss": 0.1733,
"step": 231
},
{
"epoch": 2.23,
"learning_rate": 6.196115620198271e-06,
"loss": 0.176,
"step": 232
},
{
"epoch": 2.24,
"learning_rate": 6.044817998514296e-06,
"loss": 0.2399,
"step": 233
},
{
"epoch": 2.25,
"learning_rate": 5.895060973228606e-06,
"loss": 0.2473,
"step": 234
},
{
"epoch": 2.26,
"learning_rate": 5.7468610769191195e-06,
"loss": 0.248,
"step": 235
},
{
"epoch": 2.27,
"learning_rate": 5.600234670262925e-06,
"loss": 0.2179,
"step": 236
},
{
"epoch": 2.28,
"learning_rate": 5.455197940230137e-06,
"loss": 0.277,
"step": 237
},
{
"epoch": 2.29,
"learning_rate": 5.311766898296915e-06,
"loss": 0.221,
"step": 238
},
{
"epoch": 2.3,
"learning_rate": 5.169957378677859e-06,
"loss": 0.2477,
"step": 239
},
{
"epoch": 2.31,
"learning_rate": 5.029785036577976e-06,
"loss": 0.2348,
"step": 240
},
{
"epoch": 2.32,
"learning_rate": 4.891265346464416e-06,
"loss": 0.1883,
"step": 241
},
{
"epoch": 2.33,
"learning_rate": 4.7544136003581365e-06,
"loss": 0.2551,
"step": 242
},
{
"epoch": 2.34,
"learning_rate": 4.619244906145734e-06,
"loss": 0.2097,
"step": 243
},
{
"epoch": 2.35,
"learning_rate": 4.4857741859116024e-06,
"loss": 0.2101,
"step": 244
},
{
"epoch": 2.36,
"learning_rate": 4.354016174290572e-06,
"loss": 0.2218,
"step": 245
},
{
"epoch": 2.37,
"learning_rate": 4.223985416841292e-06,
"loss": 0.3078,
"step": 246
},
{
"epoch": 2.38,
"learning_rate": 4.095696268440426e-06,
"loss": 0.2115,
"step": 247
},
{
"epoch": 2.39,
"learning_rate": 3.969162891697962e-06,
"loss": 0.1939,
"step": 248
},
{
"epoch": 2.4,
"learning_rate": 3.844399255393705e-06,
"loss": 0.1181,
"step": 249
},
{
"epoch": 2.41,
"learning_rate": 3.7214191329351735e-06,
"loss": 0.2248,
"step": 250
},
{
"epoch": 2.42,
"learning_rate": 3.6002361008370802e-06,
"loss": 0.1661,
"step": 251
},
{
"epoch": 2.43,
"learning_rate": 3.4808635372225276e-06,
"loss": 0.1613,
"step": 252
},
{
"epoch": 2.44,
"learning_rate": 3.3633146203461275e-06,
"loss": 0.226,
"step": 253
},
{
"epoch": 2.45,
"learning_rate": 3.2476023271391698e-06,
"loss": 0.1599,
"step": 254
},
{
"epoch": 2.46,
"learning_rate": 3.1337394317770208e-06,
"loss": 0.1809,
"step": 255
},
{
"epoch": 2.47,
"learning_rate": 3.021738504268905e-06,
"loss": 0.1874,
"step": 256
},
{
"epoch": 2.47,
"learning_rate": 2.911611909070229e-06,
"loss": 0.1993,
"step": 257
},
{
"epoch": 2.48,
"learning_rate": 2.8033718037175915e-06,
"loss": 0.2099,
"step": 258
},
{
"epoch": 2.49,
"learning_rate": 2.6970301374866337e-06,
"loss": 0.1388,
"step": 259
},
{
"epoch": 2.5,
"learning_rate": 2.5925986500729083e-06,
"loss": 0.2015,
"step": 260
},
{
"epoch": 2.51,
"learning_rate": 2.490088870295839e-06,
"loss": 0.233,
"step": 261
},
{
"epoch": 2.52,
"learning_rate": 2.3895121148260027e-06,
"loss": 0.2557,
"step": 262
},
{
"epoch": 2.53,
"learning_rate": 2.2908794869358044e-06,
"loss": 0.1154,
"step": 263
},
{
"epoch": 2.54,
"learning_rate": 2.1942018752737227e-06,
"loss": 0.1874,
"step": 264
},
{
"epoch": 2.55,
"learning_rate": 2.099489952662248e-06,
"loss": 0.2139,
"step": 265
},
{
"epoch": 2.56,
"learning_rate": 2.0067541749196453e-06,
"loss": 0.2447,
"step": 266
},
{
"epoch": 2.57,
"learning_rate": 1.916004779705669e-06,
"loss": 0.2365,
"step": 267
},
{
"epoch": 2.58,
"learning_rate": 1.8272517853913775e-06,
"loss": 0.2033,
"step": 268
},
{
"epoch": 2.59,
"learning_rate": 1.740504989953129e-06,
"loss": 0.2648,
"step": 269
},
{
"epoch": 2.6,
"learning_rate": 1.6557739698909436e-06,
"loss": 0.1375,
"step": 270
},
{
"epoch": 2.61,
"learning_rate": 1.573068079171265e-06,
"loss": 0.1661,
"step": 271
},
{
"epoch": 2.62,
"learning_rate": 1.4923964481943599e-06,
"loss": 0.2003,
"step": 272
},
{
"epoch": 2.63,
"learning_rate": 1.4137679827863293e-06,
"loss": 0.2477,
"step": 273
},
{
"epoch": 2.64,
"learning_rate": 1.3371913632159506e-06,
"loss": 0.2486,
"step": 274
},
{
"epoch": 2.65,
"learning_rate": 1.2626750432364077e-06,
"loss": 0.1322,
"step": 275
},
{
"epoch": 2.66,
"learning_rate": 1.1902272491520362e-06,
"loss": 0.2583,
"step": 276
},
{
"epoch": 2.67,
"learning_rate": 1.119855978910165e-06,
"loss": 0.2542,
"step": 277
},
{
"epoch": 2.68,
"learning_rate": 1.0515690012181823e-06,
"loss": 0.2442,
"step": 278
},
{
"epoch": 2.69,
"learning_rate": 9.853738546858893e-07,
"loss": 0.2323,
"step": 279
},
{
"epoch": 2.7,
"learning_rate": 9.212778469932848e-07,
"loss": 0.2138,
"step": 280
},
{
"epoch": 2.71,
"learning_rate": 8.592880540838111e-07,
"loss": 0.183,
"step": 281
},
{
"epoch": 2.72,
"learning_rate": 7.994113193832076e-07,
"loss": 0.2025,
"step": 282
},
{
"epoch": 2.73,
"learning_rate": 7.416542530440174e-07,
"loss": 0.215,
"step": 283
},
{
"epoch": 2.73,
"learning_rate": 6.860232312158554e-07,
"loss": 0.2263,
"step": 284
},
{
"epoch": 2.74,
"learning_rate": 6.325243953415117e-07,
"loss": 0.2348,
"step": 285
},
{
"epoch": 2.75,
"learning_rate": 5.811636514789598e-07,
"loss": 0.1654,
"step": 286
},
{
"epoch": 2.76,
"learning_rate": 5.31946669649337e-07,
"loss": 0.1634,
"step": 287
},
{
"epoch": 2.77,
"learning_rate": 4.848788832110151e-07,
"loss": 0.2984,
"step": 288
},
{
"epoch": 2.78,
"learning_rate": 4.399654882597726e-07,
"loss": 0.216,
"step": 289
},
{
"epoch": 2.79,
"learning_rate": 3.972114430551632e-07,
"loss": 0.1855,
"step": 290
},
{
"epoch": 2.8,
"learning_rate": 3.5662146747315054e-07,
"loss": 0.138,
"step": 291
},
{
"epoch": 2.81,
"learning_rate": 3.1820004248503957e-07,
"loss": 0.2327,
"step": 292
},
{
"epoch": 2.82,
"learning_rate": 2.8195140966281285e-07,
"loss": 0.19,
"step": 293
},
{
"epoch": 2.83,
"learning_rate": 2.478795707108672e-07,
"loss": 0.2633,
"step": 294
},
{
"epoch": 2.84,
"learning_rate": 2.1598828702424467e-07,
"loss": 0.252,
"step": 295
},
{
"epoch": 2.85,
"learning_rate": 1.862810792733849e-07,
"loss": 0.1745,
"step": 296
},
{
"epoch": 2.86,
"learning_rate": 1.5876122701546481e-07,
"loss": 0.2287,
"step": 297
},
{
"epoch": 2.87,
"learning_rate": 1.3343176833234161e-07,
"loss": 0.2116,
"step": 298
},
{
"epoch": 2.88,
"learning_rate": 1.1029549949516549e-07,
"loss": 0.227,
"step": 299
},
{
"epoch": 2.89,
"learning_rate": 8.935497465567989e-08,
"loss": 0.2794,
"step": 300
},
{
"epoch": 2.9,
"learning_rate": 7.06125055642537e-08,
"loss": 0.2103,
"step": 301
},
{
"epoch": 2.91,
"learning_rate": 5.4070161314676574e-08,
"loss": 0.2188,
"step": 302
},
{
"epoch": 2.92,
"learning_rate": 3.972976811573048e-08,
"loss": 0.2068,
"step": 303
},
{
"epoch": 2.93,
"learning_rate": 2.7592909089593224e-08,
"loss": 0.2937,
"step": 304
},
{
"epoch": 2.94,
"learning_rate": 1.766092409706266e-08,
"loss": 0.3126,
"step": 305
},
{
"epoch": 2.95,
"learning_rate": 9.934909589646157e-09,
"loss": 0.1625,
"step": 306
},
{
"epoch": 2.96,
"learning_rate": 4.415718488510745e-09,
"loss": 0.2465,
"step": 307
},
{
"epoch": 2.97,
"learning_rate": 1.1039600903250958e-09,
"loss": 0.2177,
"step": 308
},
{
"epoch": 2.98,
"learning_rate": 0.0,
"loss": 0.2132,
"step": 309
},
{
"epoch": 2.98,
"step": 309,
"total_flos": 1.3696627367060921e+21,
"train_loss": 0.5500956740242378,
"train_runtime": 5438.4326,
"train_samples_per_second": 5.498,
"train_steps_per_second": 0.057
}
],
"logging_steps": 1.0,
"max_steps": 309,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 200,
"total_flos": 1.3696627367060921e+21,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}