whisper-medium-ml-rmft / trainer_state.json
Kush0610's picture
Upload whisper-medium-reverse-ml-mft-1-1-1 model
a793bac verified
{
"best_global_step": 2448,
"best_metric": 0.0535544753074646,
"best_model_checkpoint": "./models/whisper-medium-reverse-ml-mft-1-1-1/checkpoint-2448",
"epoch": 4.0,
"eval_steps": 816,
"global_step": 3264,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0,
"eval_loss": 0.15549185872077942,
"eval_runtime": 37.6186,
"eval_samples_per_second": 133.338,
"eval_steps_per_second": 1.063,
"step": 0
},
{
"epoch": 0.0024509803921568627,
"grad_norm": 1.5625,
"learning_rate": 3.067484662576688e-08,
"loss": 0.124,
"step": 2
},
{
"epoch": 0.004901960784313725,
"grad_norm": 2.015625,
"learning_rate": 9.202453987730062e-08,
"loss": 0.1223,
"step": 4
},
{
"epoch": 0.007352941176470588,
"grad_norm": 1.359375,
"learning_rate": 1.5337423312883438e-07,
"loss": 0.1117,
"step": 6
},
{
"epoch": 0.00980392156862745,
"grad_norm": 1.5234375,
"learning_rate": 2.1472392638036812e-07,
"loss": 0.1288,
"step": 8
},
{
"epoch": 0.012254901960784314,
"grad_norm": 1.5234375,
"learning_rate": 2.760736196319019e-07,
"loss": 0.1241,
"step": 10
},
{
"epoch": 0.014705882352941176,
"grad_norm": 2.078125,
"learning_rate": 3.3742331288343563e-07,
"loss": 0.1258,
"step": 12
},
{
"epoch": 0.01715686274509804,
"grad_norm": 1.578125,
"learning_rate": 3.9877300613496937e-07,
"loss": 0.1465,
"step": 14
},
{
"epoch": 0.0196078431372549,
"grad_norm": 1.4921875,
"learning_rate": 4.601226993865031e-07,
"loss": 0.1402,
"step": 16
},
{
"epoch": 0.022058823529411766,
"grad_norm": 1.640625,
"learning_rate": 5.214723926380368e-07,
"loss": 0.1325,
"step": 18
},
{
"epoch": 0.024509803921568627,
"grad_norm": 1.6328125,
"learning_rate": 5.828220858895705e-07,
"loss": 0.1395,
"step": 20
},
{
"epoch": 0.02696078431372549,
"grad_norm": 1.671875,
"learning_rate": 6.441717791411044e-07,
"loss": 0.1311,
"step": 22
},
{
"epoch": 0.029411764705882353,
"grad_norm": 1.546875,
"learning_rate": 7.05521472392638e-07,
"loss": 0.1303,
"step": 24
},
{
"epoch": 0.031862745098039214,
"grad_norm": 1.703125,
"learning_rate": 7.668711656441719e-07,
"loss": 0.1327,
"step": 26
},
{
"epoch": 0.03431372549019608,
"grad_norm": 1.3984375,
"learning_rate": 8.282208588957055e-07,
"loss": 0.1223,
"step": 28
},
{
"epoch": 0.03676470588235294,
"grad_norm": 1.6171875,
"learning_rate": 8.895705521472393e-07,
"loss": 0.1348,
"step": 30
},
{
"epoch": 0.0392156862745098,
"grad_norm": 1.515625,
"learning_rate": 9.509202453987732e-07,
"loss": 0.1251,
"step": 32
},
{
"epoch": 0.041666666666666664,
"grad_norm": 1.5703125,
"learning_rate": 1.0122699386503068e-06,
"loss": 0.1224,
"step": 34
},
{
"epoch": 0.04411764705882353,
"grad_norm": 1.6484375,
"learning_rate": 1.0736196319018406e-06,
"loss": 0.1309,
"step": 36
},
{
"epoch": 0.04656862745098039,
"grad_norm": 1.5859375,
"learning_rate": 1.1349693251533743e-06,
"loss": 0.1334,
"step": 38
},
{
"epoch": 0.049019607843137254,
"grad_norm": 1.875,
"learning_rate": 1.196319018404908e-06,
"loss": 0.1313,
"step": 40
},
{
"epoch": 0.051470588235294115,
"grad_norm": 1.78125,
"learning_rate": 1.2576687116564418e-06,
"loss": 0.1291,
"step": 42
},
{
"epoch": 0.05392156862745098,
"grad_norm": 1.59375,
"learning_rate": 1.3190184049079755e-06,
"loss": 0.1183,
"step": 44
},
{
"epoch": 0.056372549019607844,
"grad_norm": 1.9765625,
"learning_rate": 1.3803680981595095e-06,
"loss": 0.1433,
"step": 46
},
{
"epoch": 0.058823529411764705,
"grad_norm": 1.6484375,
"learning_rate": 1.441717791411043e-06,
"loss": 0.1311,
"step": 48
},
{
"epoch": 0.061274509803921566,
"grad_norm": 1.5390625,
"learning_rate": 1.5030674846625768e-06,
"loss": 0.1187,
"step": 50
},
{
"epoch": 0.06372549019607843,
"grad_norm": 1.4609375,
"learning_rate": 1.5644171779141107e-06,
"loss": 0.1273,
"step": 52
},
{
"epoch": 0.0661764705882353,
"grad_norm": 1.546875,
"learning_rate": 1.6257668711656445e-06,
"loss": 0.1245,
"step": 54
},
{
"epoch": 0.06862745098039216,
"grad_norm": 1.3828125,
"learning_rate": 1.687116564417178e-06,
"loss": 0.1278,
"step": 56
},
{
"epoch": 0.07107843137254902,
"grad_norm": 1.484375,
"learning_rate": 1.7484662576687117e-06,
"loss": 0.1252,
"step": 58
},
{
"epoch": 0.07352941176470588,
"grad_norm": 1.4609375,
"learning_rate": 1.8098159509202457e-06,
"loss": 0.1266,
"step": 60
},
{
"epoch": 0.07598039215686274,
"grad_norm": 1.3984375,
"learning_rate": 1.8711656441717794e-06,
"loss": 0.1221,
"step": 62
},
{
"epoch": 0.0784313725490196,
"grad_norm": 1.4296875,
"learning_rate": 1.932515337423313e-06,
"loss": 0.1215,
"step": 64
},
{
"epoch": 0.08088235294117647,
"grad_norm": 1.609375,
"learning_rate": 1.9938650306748465e-06,
"loss": 0.1428,
"step": 66
},
{
"epoch": 0.08333333333333333,
"grad_norm": 1.515625,
"learning_rate": 2.0552147239263804e-06,
"loss": 0.1162,
"step": 68
},
{
"epoch": 0.0857843137254902,
"grad_norm": 1.6171875,
"learning_rate": 2.1165644171779144e-06,
"loss": 0.1228,
"step": 70
},
{
"epoch": 0.08823529411764706,
"grad_norm": 1.546875,
"learning_rate": 2.177914110429448e-06,
"loss": 0.1333,
"step": 72
},
{
"epoch": 0.09068627450980392,
"grad_norm": 1.8203125,
"learning_rate": 2.239263803680982e-06,
"loss": 0.1173,
"step": 74
},
{
"epoch": 0.09313725490196079,
"grad_norm": 1.59375,
"learning_rate": 2.3006134969325154e-06,
"loss": 0.1169,
"step": 76
},
{
"epoch": 0.09558823529411764,
"grad_norm": 1.3359375,
"learning_rate": 2.3619631901840493e-06,
"loss": 0.1206,
"step": 78
},
{
"epoch": 0.09803921568627451,
"grad_norm": 1.515625,
"learning_rate": 2.423312883435583e-06,
"loss": 0.127,
"step": 80
},
{
"epoch": 0.10049019607843138,
"grad_norm": 1.4375,
"learning_rate": 2.484662576687117e-06,
"loss": 0.1158,
"step": 82
},
{
"epoch": 0.10294117647058823,
"grad_norm": 1.4375,
"learning_rate": 2.5460122699386504e-06,
"loss": 0.1368,
"step": 84
},
{
"epoch": 0.1053921568627451,
"grad_norm": 1.21875,
"learning_rate": 2.6073619631901843e-06,
"loss": 0.1087,
"step": 86
},
{
"epoch": 0.10784313725490197,
"grad_norm": 1.5546875,
"learning_rate": 2.6687116564417183e-06,
"loss": 0.1097,
"step": 88
},
{
"epoch": 0.11029411764705882,
"grad_norm": 1.4609375,
"learning_rate": 2.7300613496932514e-06,
"loss": 0.1185,
"step": 90
},
{
"epoch": 0.11274509803921569,
"grad_norm": 1.328125,
"learning_rate": 2.7914110429447853e-06,
"loss": 0.1084,
"step": 92
},
{
"epoch": 0.11519607843137254,
"grad_norm": 1.046875,
"learning_rate": 2.8527607361963193e-06,
"loss": 0.1217,
"step": 94
},
{
"epoch": 0.11764705882352941,
"grad_norm": 1.1015625,
"learning_rate": 2.914110429447853e-06,
"loss": 0.1096,
"step": 96
},
{
"epoch": 0.12009803921568628,
"grad_norm": 1.5546875,
"learning_rate": 2.9754601226993867e-06,
"loss": 0.1076,
"step": 98
},
{
"epoch": 0.12254901960784313,
"grad_norm": 0.921875,
"learning_rate": 3.0368098159509207e-06,
"loss": 0.0991,
"step": 100
},
{
"epoch": 0.125,
"grad_norm": 1.1171875,
"learning_rate": 3.0981595092024542e-06,
"loss": 0.1098,
"step": 102
},
{
"epoch": 0.12745098039215685,
"grad_norm": 1.25,
"learning_rate": 3.159509202453988e-06,
"loss": 0.1084,
"step": 104
},
{
"epoch": 0.12990196078431374,
"grad_norm": 1.265625,
"learning_rate": 3.2208588957055213e-06,
"loss": 0.1047,
"step": 106
},
{
"epoch": 0.1323529411764706,
"grad_norm": 1.296875,
"learning_rate": 3.2822085889570552e-06,
"loss": 0.1058,
"step": 108
},
{
"epoch": 0.13480392156862744,
"grad_norm": 1.1640625,
"learning_rate": 3.343558282208589e-06,
"loss": 0.0949,
"step": 110
},
{
"epoch": 0.13725490196078433,
"grad_norm": 1.4140625,
"learning_rate": 3.4049079754601227e-06,
"loss": 0.1214,
"step": 112
},
{
"epoch": 0.13970588235294118,
"grad_norm": 1.3046875,
"learning_rate": 3.4662576687116567e-06,
"loss": 0.1041,
"step": 114
},
{
"epoch": 0.14215686274509803,
"grad_norm": 1.3203125,
"learning_rate": 3.5276073619631906e-06,
"loss": 0.1127,
"step": 116
},
{
"epoch": 0.14460784313725492,
"grad_norm": 0.91796875,
"learning_rate": 3.588957055214724e-06,
"loss": 0.0946,
"step": 118
},
{
"epoch": 0.14705882352941177,
"grad_norm": 1.0703125,
"learning_rate": 3.650306748466258e-06,
"loss": 0.0931,
"step": 120
},
{
"epoch": 0.14950980392156862,
"grad_norm": 1.296875,
"learning_rate": 3.711656441717792e-06,
"loss": 0.0935,
"step": 122
},
{
"epoch": 0.15196078431372548,
"grad_norm": 1.109375,
"learning_rate": 3.773006134969325e-06,
"loss": 0.1132,
"step": 124
},
{
"epoch": 0.15441176470588236,
"grad_norm": 0.94140625,
"learning_rate": 3.834355828220859e-06,
"loss": 0.107,
"step": 126
},
{
"epoch": 0.1568627450980392,
"grad_norm": 1.171875,
"learning_rate": 3.8957055214723935e-06,
"loss": 0.0869,
"step": 128
},
{
"epoch": 0.15931372549019607,
"grad_norm": 0.9609375,
"learning_rate": 3.957055214723927e-06,
"loss": 0.0927,
"step": 130
},
{
"epoch": 0.16176470588235295,
"grad_norm": 0.9375,
"learning_rate": 4.0184049079754606e-06,
"loss": 0.1006,
"step": 132
},
{
"epoch": 0.1642156862745098,
"grad_norm": 0.8515625,
"learning_rate": 4.079754601226994e-06,
"loss": 0.09,
"step": 134
},
{
"epoch": 0.16666666666666666,
"grad_norm": 0.72265625,
"learning_rate": 4.141104294478528e-06,
"loss": 0.0873,
"step": 136
},
{
"epoch": 0.16911764705882354,
"grad_norm": 0.96484375,
"learning_rate": 4.202453987730062e-06,
"loss": 0.0933,
"step": 138
},
{
"epoch": 0.1715686274509804,
"grad_norm": 0.88671875,
"learning_rate": 4.2638036809815955e-06,
"loss": 0.0946,
"step": 140
},
{
"epoch": 0.17401960784313725,
"grad_norm": 0.671875,
"learning_rate": 4.325153374233129e-06,
"loss": 0.09,
"step": 142
},
{
"epoch": 0.17647058823529413,
"grad_norm": 0.65625,
"learning_rate": 4.3865030674846634e-06,
"loss": 0.0832,
"step": 144
},
{
"epoch": 0.17892156862745098,
"grad_norm": 0.66796875,
"learning_rate": 4.447852760736197e-06,
"loss": 0.093,
"step": 146
},
{
"epoch": 0.18137254901960784,
"grad_norm": 0.74609375,
"learning_rate": 4.5092024539877305e-06,
"loss": 0.0851,
"step": 148
},
{
"epoch": 0.18382352941176472,
"grad_norm": 1.0078125,
"learning_rate": 4.570552147239265e-06,
"loss": 0.0938,
"step": 150
},
{
"epoch": 0.18627450980392157,
"grad_norm": 0.828125,
"learning_rate": 4.6319018404907975e-06,
"loss": 0.081,
"step": 152
},
{
"epoch": 0.18872549019607843,
"grad_norm": 0.6875,
"learning_rate": 4.693251533742332e-06,
"loss": 0.0719,
"step": 154
},
{
"epoch": 0.19117647058823528,
"grad_norm": 0.58203125,
"learning_rate": 4.7546012269938654e-06,
"loss": 0.0779,
"step": 156
},
{
"epoch": 0.19362745098039216,
"grad_norm": 0.72265625,
"learning_rate": 4.815950920245399e-06,
"loss": 0.0804,
"step": 158
},
{
"epoch": 0.19607843137254902,
"grad_norm": 0.671875,
"learning_rate": 4.877300613496933e-06,
"loss": 0.0862,
"step": 160
},
{
"epoch": 0.19852941176470587,
"grad_norm": 0.63671875,
"learning_rate": 4.938650306748467e-06,
"loss": 0.0898,
"step": 162
},
{
"epoch": 0.20098039215686275,
"grad_norm": 0.62890625,
"learning_rate": 5e-06,
"loss": 0.0732,
"step": 164
},
{
"epoch": 0.2034313725490196,
"grad_norm": 0.58984375,
"learning_rate": 5.061349693251534e-06,
"loss": 0.076,
"step": 166
},
{
"epoch": 0.20588235294117646,
"grad_norm": 0.62109375,
"learning_rate": 5.122699386503068e-06,
"loss": 0.0797,
"step": 168
},
{
"epoch": 0.20833333333333334,
"grad_norm": 0.5234375,
"learning_rate": 5.184049079754602e-06,
"loss": 0.078,
"step": 170
},
{
"epoch": 0.2107843137254902,
"grad_norm": 0.515625,
"learning_rate": 5.245398773006135e-06,
"loss": 0.0729,
"step": 172
},
{
"epoch": 0.21323529411764705,
"grad_norm": 0.57421875,
"learning_rate": 5.30674846625767e-06,
"loss": 0.0786,
"step": 174
},
{
"epoch": 0.21568627450980393,
"grad_norm": 0.640625,
"learning_rate": 5.368098159509203e-06,
"loss": 0.0819,
"step": 176
},
{
"epoch": 0.2181372549019608,
"grad_norm": 0.51171875,
"learning_rate": 5.429447852760736e-06,
"loss": 0.0771,
"step": 178
},
{
"epoch": 0.22058823529411764,
"grad_norm": 0.482421875,
"learning_rate": 5.490797546012271e-06,
"loss": 0.073,
"step": 180
},
{
"epoch": 0.22303921568627452,
"grad_norm": 0.43359375,
"learning_rate": 5.552147239263804e-06,
"loss": 0.088,
"step": 182
},
{
"epoch": 0.22549019607843138,
"grad_norm": 0.53515625,
"learning_rate": 5.613496932515337e-06,
"loss": 0.0837,
"step": 184
},
{
"epoch": 0.22794117647058823,
"grad_norm": 0.443359375,
"learning_rate": 5.674846625766872e-06,
"loss": 0.0732,
"step": 186
},
{
"epoch": 0.23039215686274508,
"grad_norm": 0.44921875,
"learning_rate": 5.736196319018405e-06,
"loss": 0.0789,
"step": 188
},
{
"epoch": 0.23284313725490197,
"grad_norm": 0.455078125,
"learning_rate": 5.797546012269939e-06,
"loss": 0.0711,
"step": 190
},
{
"epoch": 0.23529411764705882,
"grad_norm": 0.57421875,
"learning_rate": 5.858895705521472e-06,
"loss": 0.0696,
"step": 192
},
{
"epoch": 0.23774509803921567,
"grad_norm": 0.5703125,
"learning_rate": 5.920245398773007e-06,
"loss": 0.0713,
"step": 194
},
{
"epoch": 0.24019607843137256,
"grad_norm": 0.52734375,
"learning_rate": 5.98159509202454e-06,
"loss": 0.0816,
"step": 196
},
{
"epoch": 0.2426470588235294,
"grad_norm": 0.515625,
"learning_rate": 6.042944785276074e-06,
"loss": 0.0706,
"step": 198
},
{
"epoch": 0.24509803921568626,
"grad_norm": 0.455078125,
"learning_rate": 6.104294478527608e-06,
"loss": 0.0752,
"step": 200
},
{
"epoch": 0.24754901960784315,
"grad_norm": 0.453125,
"learning_rate": 6.165644171779142e-06,
"loss": 0.0723,
"step": 202
},
{
"epoch": 0.25,
"grad_norm": 0.470703125,
"learning_rate": 6.226993865030675e-06,
"loss": 0.0759,
"step": 204
},
{
"epoch": 0.25245098039215685,
"grad_norm": 0.3984375,
"learning_rate": 6.28834355828221e-06,
"loss": 0.0746,
"step": 206
},
{
"epoch": 0.2549019607843137,
"grad_norm": 0.4375,
"learning_rate": 6.349693251533743e-06,
"loss": 0.0705,
"step": 208
},
{
"epoch": 0.25735294117647056,
"grad_norm": 0.50390625,
"learning_rate": 6.411042944785276e-06,
"loss": 0.0756,
"step": 210
},
{
"epoch": 0.25980392156862747,
"grad_norm": 0.435546875,
"learning_rate": 6.472392638036811e-06,
"loss": 0.0709,
"step": 212
},
{
"epoch": 0.2622549019607843,
"grad_norm": 0.5078125,
"learning_rate": 6.533742331288344e-06,
"loss": 0.0698,
"step": 214
},
{
"epoch": 0.2647058823529412,
"grad_norm": 0.5234375,
"learning_rate": 6.595092024539877e-06,
"loss": 0.0693,
"step": 216
},
{
"epoch": 0.26715686274509803,
"grad_norm": 0.419921875,
"learning_rate": 6.656441717791412e-06,
"loss": 0.0643,
"step": 218
},
{
"epoch": 0.2696078431372549,
"grad_norm": 0.46484375,
"learning_rate": 6.717791411042945e-06,
"loss": 0.0713,
"step": 220
},
{
"epoch": 0.27205882352941174,
"grad_norm": 0.4140625,
"learning_rate": 6.779141104294479e-06,
"loss": 0.0777,
"step": 222
},
{
"epoch": 0.27450980392156865,
"grad_norm": 0.4453125,
"learning_rate": 6.840490797546013e-06,
"loss": 0.0596,
"step": 224
},
{
"epoch": 0.2769607843137255,
"grad_norm": 0.373046875,
"learning_rate": 6.901840490797547e-06,
"loss": 0.0658,
"step": 226
},
{
"epoch": 0.27941176470588236,
"grad_norm": 0.384765625,
"learning_rate": 6.96319018404908e-06,
"loss": 0.0708,
"step": 228
},
{
"epoch": 0.2818627450980392,
"grad_norm": 0.3828125,
"learning_rate": 7.0245398773006145e-06,
"loss": 0.0759,
"step": 230
},
{
"epoch": 0.28431372549019607,
"grad_norm": 0.498046875,
"learning_rate": 7.085889570552148e-06,
"loss": 0.0671,
"step": 232
},
{
"epoch": 0.2867647058823529,
"grad_norm": 0.3359375,
"learning_rate": 7.1472392638036816e-06,
"loss": 0.0716,
"step": 234
},
{
"epoch": 0.28921568627450983,
"grad_norm": 0.37890625,
"learning_rate": 7.208588957055215e-06,
"loss": 0.0642,
"step": 236
},
{
"epoch": 0.2916666666666667,
"grad_norm": 0.361328125,
"learning_rate": 7.2699386503067495e-06,
"loss": 0.0593,
"step": 238
},
{
"epoch": 0.29411764705882354,
"grad_norm": 0.3125,
"learning_rate": 7.331288343558283e-06,
"loss": 0.0593,
"step": 240
},
{
"epoch": 0.2965686274509804,
"grad_norm": 0.318359375,
"learning_rate": 7.392638036809816e-06,
"loss": 0.0623,
"step": 242
},
{
"epoch": 0.29901960784313725,
"grad_norm": 0.46484375,
"learning_rate": 7.453987730061351e-06,
"loss": 0.0677,
"step": 244
},
{
"epoch": 0.3014705882352941,
"grad_norm": 0.419921875,
"learning_rate": 7.5153374233128836e-06,
"loss": 0.0648,
"step": 246
},
{
"epoch": 0.30392156862745096,
"grad_norm": 0.44921875,
"learning_rate": 7.576687116564417e-06,
"loss": 0.0739,
"step": 248
},
{
"epoch": 0.30637254901960786,
"grad_norm": 0.345703125,
"learning_rate": 7.638036809815951e-06,
"loss": 0.0659,
"step": 250
},
{
"epoch": 0.3088235294117647,
"grad_norm": 0.416015625,
"learning_rate": 7.699386503067485e-06,
"loss": 0.0663,
"step": 252
},
{
"epoch": 0.3112745098039216,
"grad_norm": 0.388671875,
"learning_rate": 7.760736196319019e-06,
"loss": 0.064,
"step": 254
},
{
"epoch": 0.3137254901960784,
"grad_norm": 0.318359375,
"learning_rate": 7.822085889570554e-06,
"loss": 0.0592,
"step": 256
},
{
"epoch": 0.3161764705882353,
"grad_norm": 0.40234375,
"learning_rate": 7.883435582822087e-06,
"loss": 0.0644,
"step": 258
},
{
"epoch": 0.31862745098039214,
"grad_norm": 0.337890625,
"learning_rate": 7.944785276073619e-06,
"loss": 0.0621,
"step": 260
},
{
"epoch": 0.32107843137254904,
"grad_norm": 0.3828125,
"learning_rate": 8.006134969325154e-06,
"loss": 0.0574,
"step": 262
},
{
"epoch": 0.3235294117647059,
"grad_norm": 0.341796875,
"learning_rate": 8.067484662576688e-06,
"loss": 0.065,
"step": 264
},
{
"epoch": 0.32598039215686275,
"grad_norm": 0.33984375,
"learning_rate": 8.128834355828221e-06,
"loss": 0.0619,
"step": 266
},
{
"epoch": 0.3284313725490196,
"grad_norm": 0.306640625,
"learning_rate": 8.190184049079755e-06,
"loss": 0.057,
"step": 268
},
{
"epoch": 0.33088235294117646,
"grad_norm": 0.361328125,
"learning_rate": 8.251533742331288e-06,
"loss": 0.0614,
"step": 270
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.33984375,
"learning_rate": 8.312883435582822e-06,
"loss": 0.0685,
"step": 272
},
{
"epoch": 0.33578431372549017,
"grad_norm": 0.357421875,
"learning_rate": 8.374233128834357e-06,
"loss": 0.0658,
"step": 274
},
{
"epoch": 0.3382352941176471,
"grad_norm": 0.373046875,
"learning_rate": 8.43558282208589e-06,
"loss": 0.0642,
"step": 276
},
{
"epoch": 0.34068627450980393,
"grad_norm": 0.328125,
"learning_rate": 8.496932515337424e-06,
"loss": 0.0739,
"step": 278
},
{
"epoch": 0.3431372549019608,
"grad_norm": 0.287109375,
"learning_rate": 8.558282208588958e-06,
"loss": 0.0586,
"step": 280
},
{
"epoch": 0.34558823529411764,
"grad_norm": 0.32421875,
"learning_rate": 8.619631901840491e-06,
"loss": 0.0611,
"step": 282
},
{
"epoch": 0.3480392156862745,
"grad_norm": 0.310546875,
"learning_rate": 8.680981595092025e-06,
"loss": 0.0572,
"step": 284
},
{
"epoch": 0.35049019607843135,
"grad_norm": 0.416015625,
"learning_rate": 8.742331288343558e-06,
"loss": 0.0679,
"step": 286
},
{
"epoch": 0.35294117647058826,
"grad_norm": 0.369140625,
"learning_rate": 8.803680981595094e-06,
"loss": 0.0594,
"step": 288
},
{
"epoch": 0.3553921568627451,
"grad_norm": 0.4296875,
"learning_rate": 8.865030674846627e-06,
"loss": 0.0672,
"step": 290
},
{
"epoch": 0.35784313725490197,
"grad_norm": 0.3125,
"learning_rate": 8.926380368098159e-06,
"loss": 0.0566,
"step": 292
},
{
"epoch": 0.3602941176470588,
"grad_norm": 0.3515625,
"learning_rate": 8.987730061349694e-06,
"loss": 0.0636,
"step": 294
},
{
"epoch": 0.3627450980392157,
"grad_norm": 0.373046875,
"learning_rate": 9.049079754601228e-06,
"loss": 0.0688,
"step": 296
},
{
"epoch": 0.36519607843137253,
"grad_norm": 0.271484375,
"learning_rate": 9.110429447852761e-06,
"loss": 0.0617,
"step": 298
},
{
"epoch": 0.36764705882352944,
"grad_norm": 0.310546875,
"learning_rate": 9.171779141104295e-06,
"loss": 0.0597,
"step": 300
},
{
"epoch": 0.3700980392156863,
"grad_norm": 0.408203125,
"learning_rate": 9.233128834355828e-06,
"loss": 0.07,
"step": 302
},
{
"epoch": 0.37254901960784315,
"grad_norm": 0.34765625,
"learning_rate": 9.294478527607362e-06,
"loss": 0.0594,
"step": 304
},
{
"epoch": 0.375,
"grad_norm": 0.43359375,
"learning_rate": 9.355828220858897e-06,
"loss": 0.0539,
"step": 306
},
{
"epoch": 0.37745098039215685,
"grad_norm": 0.3359375,
"learning_rate": 9.41717791411043e-06,
"loss": 0.0658,
"step": 308
},
{
"epoch": 0.3799019607843137,
"grad_norm": 0.310546875,
"learning_rate": 9.478527607361964e-06,
"loss": 0.0618,
"step": 310
},
{
"epoch": 0.38235294117647056,
"grad_norm": 0.32421875,
"learning_rate": 9.539877300613498e-06,
"loss": 0.055,
"step": 312
},
{
"epoch": 0.38480392156862747,
"grad_norm": 0.2890625,
"learning_rate": 9.601226993865031e-06,
"loss": 0.0583,
"step": 314
},
{
"epoch": 0.3872549019607843,
"grad_norm": 0.326171875,
"learning_rate": 9.662576687116565e-06,
"loss": 0.0599,
"step": 316
},
{
"epoch": 0.3897058823529412,
"grad_norm": 0.306640625,
"learning_rate": 9.7239263803681e-06,
"loss": 0.0604,
"step": 318
},
{
"epoch": 0.39215686274509803,
"grad_norm": 0.298828125,
"learning_rate": 9.785276073619633e-06,
"loss": 0.0556,
"step": 320
},
{
"epoch": 0.3946078431372549,
"grad_norm": 0.380859375,
"learning_rate": 9.846625766871167e-06,
"loss": 0.0643,
"step": 322
},
{
"epoch": 0.39705882352941174,
"grad_norm": 0.302734375,
"learning_rate": 9.9079754601227e-06,
"loss": 0.0572,
"step": 324
},
{
"epoch": 0.39950980392156865,
"grad_norm": 0.400390625,
"learning_rate": 9.969325153374234e-06,
"loss": 0.0597,
"step": 326
},
{
"epoch": 0.4019607843137255,
"grad_norm": 0.267578125,
"learning_rate": 9.999997141513604e-06,
"loss": 0.0509,
"step": 328
},
{
"epoch": 0.40441176470588236,
"grad_norm": 0.396484375,
"learning_rate": 9.999974273642042e-06,
"loss": 0.0603,
"step": 330
},
{
"epoch": 0.4068627450980392,
"grad_norm": 0.365234375,
"learning_rate": 9.999928538003506e-06,
"loss": 0.0644,
"step": 332
},
{
"epoch": 0.40931372549019607,
"grad_norm": 0.326171875,
"learning_rate": 9.999859934807169e-06,
"loss": 0.0587,
"step": 334
},
{
"epoch": 0.4117647058823529,
"grad_norm": 0.271484375,
"learning_rate": 9.999768464366794e-06,
"loss": 0.0519,
"step": 336
},
{
"epoch": 0.41421568627450983,
"grad_norm": 0.267578125,
"learning_rate": 9.99965412710073e-06,
"loss": 0.0508,
"step": 338
},
{
"epoch": 0.4166666666666667,
"grad_norm": 0.349609375,
"learning_rate": 9.999516923531906e-06,
"loss": 0.0645,
"step": 340
},
{
"epoch": 0.41911764705882354,
"grad_norm": 0.30078125,
"learning_rate": 9.999356854287834e-06,
"loss": 0.0569,
"step": 342
},
{
"epoch": 0.4215686274509804,
"grad_norm": 0.29296875,
"learning_rate": 9.999173920100601e-06,
"loss": 0.0594,
"step": 344
},
{
"epoch": 0.42401960784313725,
"grad_norm": 0.349609375,
"learning_rate": 9.998968121806873e-06,
"loss": 0.0564,
"step": 346
},
{
"epoch": 0.4264705882352941,
"grad_norm": 0.251953125,
"learning_rate": 9.998739460347882e-06,
"loss": 0.0573,
"step": 348
},
{
"epoch": 0.42892156862745096,
"grad_norm": 0.2734375,
"learning_rate": 9.99848793676943e-06,
"loss": 0.0607,
"step": 350
},
{
"epoch": 0.43137254901960786,
"grad_norm": 0.390625,
"learning_rate": 9.99821355222188e-06,
"loss": 0.0602,
"step": 352
},
{
"epoch": 0.4338235294117647,
"grad_norm": 0.30859375,
"learning_rate": 9.99791630796015e-06,
"loss": 0.0617,
"step": 354
},
{
"epoch": 0.4362745098039216,
"grad_norm": 0.328125,
"learning_rate": 9.997596205343709e-06,
"loss": 0.0565,
"step": 356
},
{
"epoch": 0.4387254901960784,
"grad_norm": 0.333984375,
"learning_rate": 9.997253245836573e-06,
"loss": 0.0524,
"step": 358
},
{
"epoch": 0.4411764705882353,
"grad_norm": 0.388671875,
"learning_rate": 9.996887431007291e-06,
"loss": 0.0569,
"step": 360
},
{
"epoch": 0.44362745098039214,
"grad_norm": 0.298828125,
"learning_rate": 9.996498762528947e-06,
"loss": 0.0564,
"step": 362
},
{
"epoch": 0.44607843137254904,
"grad_norm": 0.341796875,
"learning_rate": 9.996087242179147e-06,
"loss": 0.0526,
"step": 364
},
{
"epoch": 0.4485294117647059,
"grad_norm": 0.330078125,
"learning_rate": 9.995652871840006e-06,
"loss": 0.0547,
"step": 366
},
{
"epoch": 0.45098039215686275,
"grad_norm": 0.255859375,
"learning_rate": 9.995195653498157e-06,
"loss": 0.0581,
"step": 368
},
{
"epoch": 0.4534313725490196,
"grad_norm": 0.43359375,
"learning_rate": 9.99471558924472e-06,
"loss": 0.0563,
"step": 370
},
{
"epoch": 0.45588235294117646,
"grad_norm": 0.306640625,
"learning_rate": 9.994212681275306e-06,
"loss": 0.0611,
"step": 372
},
{
"epoch": 0.4583333333333333,
"grad_norm": 0.365234375,
"learning_rate": 9.993686931890003e-06,
"loss": 0.0602,
"step": 374
},
{
"epoch": 0.46078431372549017,
"grad_norm": 0.32421875,
"learning_rate": 9.993138343493364e-06,
"loss": 0.057,
"step": 376
},
{
"epoch": 0.4632352941176471,
"grad_norm": 0.36328125,
"learning_rate": 9.992566918594406e-06,
"loss": 0.0579,
"step": 378
},
{
"epoch": 0.46568627450980393,
"grad_norm": 0.2470703125,
"learning_rate": 9.99197265980658e-06,
"loss": 0.0554,
"step": 380
},
{
"epoch": 0.4681372549019608,
"grad_norm": 0.314453125,
"learning_rate": 9.991355569847776e-06,
"loss": 0.0613,
"step": 382
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.408203125,
"learning_rate": 9.990715651540303e-06,
"loss": 0.0643,
"step": 384
},
{
"epoch": 0.4730392156862745,
"grad_norm": 0.275390625,
"learning_rate": 9.990052907810875e-06,
"loss": 0.0585,
"step": 386
},
{
"epoch": 0.47549019607843135,
"grad_norm": 0.30859375,
"learning_rate": 9.989367341690602e-06,
"loss": 0.0549,
"step": 388
},
{
"epoch": 0.47794117647058826,
"grad_norm": 0.3125,
"learning_rate": 9.988658956314973e-06,
"loss": 0.0599,
"step": 390
},
{
"epoch": 0.4803921568627451,
"grad_norm": 0.314453125,
"learning_rate": 9.987927754923844e-06,
"loss": 0.0556,
"step": 392
},
{
"epoch": 0.48284313725490197,
"grad_norm": 0.30078125,
"learning_rate": 9.98717374086142e-06,
"loss": 0.0487,
"step": 394
},
{
"epoch": 0.4852941176470588,
"grad_norm": 0.271484375,
"learning_rate": 9.986396917576244e-06,
"loss": 0.0517,
"step": 396
},
{
"epoch": 0.4877450980392157,
"grad_norm": 0.298828125,
"learning_rate": 9.985597288621174e-06,
"loss": 0.0453,
"step": 398
},
{
"epoch": 0.49019607843137253,
"grad_norm": 0.35546875,
"learning_rate": 9.984774857653377e-06,
"loss": 0.053,
"step": 400
},
{
"epoch": 0.49264705882352944,
"grad_norm": 0.2490234375,
"learning_rate": 9.983929628434305e-06,
"loss": 0.0526,
"step": 402
},
{
"epoch": 0.4950980392156863,
"grad_norm": 0.302734375,
"learning_rate": 9.983061604829676e-06,
"loss": 0.0523,
"step": 404
},
{
"epoch": 0.49754901960784315,
"grad_norm": 0.326171875,
"learning_rate": 9.982170790809464e-06,
"loss": 0.0576,
"step": 406
},
{
"epoch": 0.5,
"grad_norm": 0.35546875,
"learning_rate": 9.981257190447875e-06,
"loss": 0.0548,
"step": 408
},
{
"epoch": 0.5024509803921569,
"grad_norm": 0.333984375,
"learning_rate": 9.980320807923333e-06,
"loss": 0.0572,
"step": 410
},
{
"epoch": 0.5049019607843137,
"grad_norm": 0.3203125,
"learning_rate": 9.979361647518453e-06,
"loss": 0.0579,
"step": 412
},
{
"epoch": 0.5073529411764706,
"grad_norm": 0.2578125,
"learning_rate": 9.978379713620029e-06,
"loss": 0.057,
"step": 414
},
{
"epoch": 0.5098039215686274,
"grad_norm": 0.33203125,
"learning_rate": 9.977375010719013e-06,
"loss": 0.0476,
"step": 416
},
{
"epoch": 0.5122549019607843,
"grad_norm": 0.275390625,
"learning_rate": 9.976347543410487e-06,
"loss": 0.0529,
"step": 418
},
{
"epoch": 0.5147058823529411,
"grad_norm": 0.275390625,
"learning_rate": 9.975297316393658e-06,
"loss": 0.0525,
"step": 420
},
{
"epoch": 0.5171568627450981,
"grad_norm": 0.2734375,
"learning_rate": 9.974224334471814e-06,
"loss": 0.0576,
"step": 422
},
{
"epoch": 0.5196078431372549,
"grad_norm": 0.287109375,
"learning_rate": 9.973128602552324e-06,
"loss": 0.055,
"step": 424
},
{
"epoch": 0.5220588235294118,
"grad_norm": 0.33203125,
"learning_rate": 9.9720101256466e-06,
"loss": 0.0532,
"step": 426
},
{
"epoch": 0.5245098039215687,
"grad_norm": 0.310546875,
"learning_rate": 9.970868908870087e-06,
"loss": 0.0497,
"step": 428
},
{
"epoch": 0.5269607843137255,
"grad_norm": 0.23828125,
"learning_rate": 9.969704957442221e-06,
"loss": 0.0518,
"step": 430
},
{
"epoch": 0.5294117647058824,
"grad_norm": 0.345703125,
"learning_rate": 9.968518276686428e-06,
"loss": 0.0649,
"step": 432
},
{
"epoch": 0.5318627450980392,
"grad_norm": 0.2421875,
"learning_rate": 9.96730887203008e-06,
"loss": 0.052,
"step": 434
},
{
"epoch": 0.5343137254901961,
"grad_norm": 0.291015625,
"learning_rate": 9.966076749004485e-06,
"loss": 0.0547,
"step": 436
},
{
"epoch": 0.5367647058823529,
"grad_norm": 0.30078125,
"learning_rate": 9.964821913244853e-06,
"loss": 0.0528,
"step": 438
},
{
"epoch": 0.5392156862745098,
"grad_norm": 0.30078125,
"learning_rate": 9.96354437049027e-06,
"loss": 0.0589,
"step": 440
},
{
"epoch": 0.5416666666666666,
"grad_norm": 0.337890625,
"learning_rate": 9.962244126583678e-06,
"loss": 0.0528,
"step": 442
},
{
"epoch": 0.5441176470588235,
"grad_norm": 0.26953125,
"learning_rate": 9.960921187471841e-06,
"loss": 0.0583,
"step": 444
},
{
"epoch": 0.5465686274509803,
"grad_norm": 0.302734375,
"learning_rate": 9.959575559205323e-06,
"loss": 0.0564,
"step": 446
},
{
"epoch": 0.5490196078431373,
"grad_norm": 0.2412109375,
"learning_rate": 9.958207247938458e-06,
"loss": 0.0524,
"step": 448
},
{
"epoch": 0.5514705882352942,
"grad_norm": 0.3203125,
"learning_rate": 9.956816259929322e-06,
"loss": 0.0497,
"step": 450
},
{
"epoch": 0.553921568627451,
"grad_norm": 0.31640625,
"learning_rate": 9.955402601539709e-06,
"loss": 0.0584,
"step": 452
},
{
"epoch": 0.5563725490196079,
"grad_norm": 0.318359375,
"learning_rate": 9.953966279235092e-06,
"loss": 0.0583,
"step": 454
},
{
"epoch": 0.5588235294117647,
"grad_norm": 0.392578125,
"learning_rate": 9.952507299584602e-06,
"loss": 0.0558,
"step": 456
},
{
"epoch": 0.5612745098039216,
"grad_norm": 0.36328125,
"learning_rate": 9.951025669260996e-06,
"loss": 0.0661,
"step": 458
},
{
"epoch": 0.5637254901960784,
"grad_norm": 0.3046875,
"learning_rate": 9.949521395040623e-06,
"loss": 0.0573,
"step": 460
},
{
"epoch": 0.5661764705882353,
"grad_norm": 0.33203125,
"learning_rate": 9.947994483803394e-06,
"loss": 0.0563,
"step": 462
},
{
"epoch": 0.5686274509803921,
"grad_norm": 0.2470703125,
"learning_rate": 9.94644494253276e-06,
"loss": 0.0546,
"step": 464
},
{
"epoch": 0.571078431372549,
"grad_norm": 0.29296875,
"learning_rate": 9.944872778315663e-06,
"loss": 0.0572,
"step": 466
},
{
"epoch": 0.5735294117647058,
"grad_norm": 0.271484375,
"learning_rate": 9.94327799834252e-06,
"loss": 0.0544,
"step": 468
},
{
"epoch": 0.5759803921568627,
"grad_norm": 0.3203125,
"learning_rate": 9.94166060990718e-06,
"loss": 0.0691,
"step": 470
},
{
"epoch": 0.5784313725490197,
"grad_norm": 0.28515625,
"learning_rate": 9.940020620406891e-06,
"loss": 0.0527,
"step": 472
},
{
"epoch": 0.5808823529411765,
"grad_norm": 0.2431640625,
"learning_rate": 9.938358037342272e-06,
"loss": 0.0504,
"step": 474
},
{
"epoch": 0.5833333333333334,
"grad_norm": 0.318359375,
"learning_rate": 9.936672868317277e-06,
"loss": 0.0522,
"step": 476
},
{
"epoch": 0.5857843137254902,
"grad_norm": 0.3046875,
"learning_rate": 9.934965121039153e-06,
"loss": 0.0522,
"step": 478
},
{
"epoch": 0.5882352941176471,
"grad_norm": 0.376953125,
"learning_rate": 9.933234803318415e-06,
"loss": 0.0601,
"step": 480
},
{
"epoch": 0.5906862745098039,
"grad_norm": 0.263671875,
"learning_rate": 9.931481923068804e-06,
"loss": 0.0559,
"step": 482
},
{
"epoch": 0.5931372549019608,
"grad_norm": 0.298828125,
"learning_rate": 9.92970648830725e-06,
"loss": 0.0563,
"step": 484
},
{
"epoch": 0.5955882352941176,
"grad_norm": 0.30859375,
"learning_rate": 9.927908507153847e-06,
"loss": 0.0461,
"step": 486
},
{
"epoch": 0.5980392156862745,
"grad_norm": 0.244140625,
"learning_rate": 9.926087987831794e-06,
"loss": 0.0592,
"step": 488
},
{
"epoch": 0.6004901960784313,
"grad_norm": 0.259765625,
"learning_rate": 9.92424493866738e-06,
"loss": 0.0543,
"step": 490
},
{
"epoch": 0.6029411764705882,
"grad_norm": 0.263671875,
"learning_rate": 9.922379368089928e-06,
"loss": 0.0559,
"step": 492
},
{
"epoch": 0.6053921568627451,
"grad_norm": 0.251953125,
"learning_rate": 9.920491284631773e-06,
"loss": 0.0522,
"step": 494
},
{
"epoch": 0.6078431372549019,
"grad_norm": 0.353515625,
"learning_rate": 9.918580696928206e-06,
"loss": 0.0608,
"step": 496
},
{
"epoch": 0.6102941176470589,
"grad_norm": 0.2099609375,
"learning_rate": 9.91664761371745e-06,
"loss": 0.0463,
"step": 498
},
{
"epoch": 0.6127450980392157,
"grad_norm": 0.275390625,
"learning_rate": 9.914692043840608e-06,
"loss": 0.0498,
"step": 500
},
{
"epoch": 0.6151960784313726,
"grad_norm": 0.330078125,
"learning_rate": 9.912713996241633e-06,
"loss": 0.0551,
"step": 502
},
{
"epoch": 0.6176470588235294,
"grad_norm": 0.26953125,
"learning_rate": 9.91071347996727e-06,
"loss": 0.051,
"step": 504
},
{
"epoch": 0.6200980392156863,
"grad_norm": 0.271484375,
"learning_rate": 9.90869050416704e-06,
"loss": 0.0488,
"step": 506
},
{
"epoch": 0.6225490196078431,
"grad_norm": 0.34375,
"learning_rate": 9.906645078093179e-06,
"loss": 0.0553,
"step": 508
},
{
"epoch": 0.625,
"grad_norm": 0.2578125,
"learning_rate": 9.904577211100597e-06,
"loss": 0.0564,
"step": 510
},
{
"epoch": 0.6274509803921569,
"grad_norm": 0.275390625,
"learning_rate": 9.902486912646846e-06,
"loss": 0.0543,
"step": 512
},
{
"epoch": 0.6299019607843137,
"grad_norm": 0.263671875,
"learning_rate": 9.900374192292064e-06,
"loss": 0.0581,
"step": 514
},
{
"epoch": 0.6323529411764706,
"grad_norm": 0.28515625,
"learning_rate": 9.898239059698943e-06,
"loss": 0.0488,
"step": 516
},
{
"epoch": 0.6348039215686274,
"grad_norm": 0.2734375,
"learning_rate": 9.896081524632673e-06,
"loss": 0.0515,
"step": 518
},
{
"epoch": 0.6372549019607843,
"grad_norm": 0.369140625,
"learning_rate": 9.893901596960906e-06,
"loss": 0.0544,
"step": 520
},
{
"epoch": 0.6397058823529411,
"grad_norm": 0.287109375,
"learning_rate": 9.891699286653714e-06,
"loss": 0.0545,
"step": 522
},
{
"epoch": 0.6421568627450981,
"grad_norm": 0.2431640625,
"learning_rate": 9.88947460378353e-06,
"loss": 0.0492,
"step": 524
},
{
"epoch": 0.6446078431372549,
"grad_norm": 0.267578125,
"learning_rate": 9.88722755852511e-06,
"loss": 0.0574,
"step": 526
},
{
"epoch": 0.6470588235294118,
"grad_norm": 0.28515625,
"learning_rate": 9.884958161155492e-06,
"loss": 0.0476,
"step": 528
},
{
"epoch": 0.6495098039215687,
"grad_norm": 0.28515625,
"learning_rate": 9.882666422053935e-06,
"loss": 0.0595,
"step": 530
},
{
"epoch": 0.6519607843137255,
"grad_norm": 0.2734375,
"learning_rate": 9.880352351701888e-06,
"loss": 0.0532,
"step": 532
},
{
"epoch": 0.6544117647058824,
"grad_norm": 0.279296875,
"learning_rate": 9.87801596068293e-06,
"loss": 0.0572,
"step": 534
},
{
"epoch": 0.6568627450980392,
"grad_norm": 0.259765625,
"learning_rate": 9.875657259682722e-06,
"loss": 0.0546,
"step": 536
},
{
"epoch": 0.6593137254901961,
"grad_norm": 0.259765625,
"learning_rate": 9.873276259488967e-06,
"loss": 0.0476,
"step": 538
},
{
"epoch": 0.6617647058823529,
"grad_norm": 0.259765625,
"learning_rate": 9.87087297099135e-06,
"loss": 0.0527,
"step": 540
},
{
"epoch": 0.6642156862745098,
"grad_norm": 0.3203125,
"learning_rate": 9.868447405181499e-06,
"loss": 0.0609,
"step": 542
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.3203125,
"learning_rate": 9.865999573152922e-06,
"loss": 0.0515,
"step": 544
},
{
"epoch": 0.6691176470588235,
"grad_norm": 0.3203125,
"learning_rate": 9.86352948610097e-06,
"loss": 0.0524,
"step": 546
},
{
"epoch": 0.6715686274509803,
"grad_norm": 0.349609375,
"learning_rate": 9.861037155322777e-06,
"loss": 0.0505,
"step": 548
},
{
"epoch": 0.6740196078431373,
"grad_norm": 0.265625,
"learning_rate": 9.858522592217208e-06,
"loss": 0.0483,
"step": 550
},
{
"epoch": 0.6764705882352942,
"grad_norm": 0.29296875,
"learning_rate": 9.85598580828481e-06,
"loss": 0.0517,
"step": 552
},
{
"epoch": 0.678921568627451,
"grad_norm": 0.25390625,
"learning_rate": 9.85342681512776e-06,
"loss": 0.0534,
"step": 554
},
{
"epoch": 0.6813725490196079,
"grad_norm": 0.287109375,
"learning_rate": 9.85084562444981e-06,
"loss": 0.06,
"step": 556
},
{
"epoch": 0.6838235294117647,
"grad_norm": 0.255859375,
"learning_rate": 9.848242248056234e-06,
"loss": 0.0473,
"step": 558
},
{
"epoch": 0.6862745098039216,
"grad_norm": 0.328125,
"learning_rate": 9.845616697853775e-06,
"loss": 0.0521,
"step": 560
},
{
"epoch": 0.6887254901960784,
"grad_norm": 0.2255859375,
"learning_rate": 9.84296898585059e-06,
"loss": 0.048,
"step": 562
},
{
"epoch": 0.6911764705882353,
"grad_norm": 0.330078125,
"learning_rate": 9.84029912415619e-06,
"loss": 0.0514,
"step": 564
},
{
"epoch": 0.6936274509803921,
"grad_norm": 0.353515625,
"learning_rate": 9.837607124981397e-06,
"loss": 0.0521,
"step": 566
},
{
"epoch": 0.696078431372549,
"grad_norm": 0.296875,
"learning_rate": 9.834893000638272e-06,
"loss": 0.0531,
"step": 568
},
{
"epoch": 0.6985294117647058,
"grad_norm": 0.2734375,
"learning_rate": 9.832156763540075e-06,
"loss": 0.0529,
"step": 570
},
{
"epoch": 0.7009803921568627,
"grad_norm": 0.296875,
"learning_rate": 9.829398426201196e-06,
"loss": 0.0605,
"step": 572
},
{
"epoch": 0.7034313725490197,
"grad_norm": 0.28125,
"learning_rate": 9.826618001237101e-06,
"loss": 0.0535,
"step": 574
},
{
"epoch": 0.7058823529411765,
"grad_norm": 0.28125,
"learning_rate": 9.82381550136428e-06,
"loss": 0.055,
"step": 576
},
{
"epoch": 0.7083333333333334,
"grad_norm": 0.35546875,
"learning_rate": 9.82099093940018e-06,
"loss": 0.0475,
"step": 578
},
{
"epoch": 0.7107843137254902,
"grad_norm": 0.361328125,
"learning_rate": 9.818144328263154e-06,
"loss": 0.0548,
"step": 580
},
{
"epoch": 0.7132352941176471,
"grad_norm": 0.328125,
"learning_rate": 9.815275680972396e-06,
"loss": 0.0499,
"step": 582
},
{
"epoch": 0.7156862745098039,
"grad_norm": 0.3125,
"learning_rate": 9.812385010647885e-06,
"loss": 0.0556,
"step": 584
},
{
"epoch": 0.7181372549019608,
"grad_norm": 0.275390625,
"learning_rate": 9.809472330510323e-06,
"loss": 0.0541,
"step": 586
},
{
"epoch": 0.7205882352941176,
"grad_norm": 0.234375,
"learning_rate": 9.80653765388108e-06,
"loss": 0.0499,
"step": 588
},
{
"epoch": 0.7230392156862745,
"grad_norm": 0.298828125,
"learning_rate": 9.803580994182122e-06,
"loss": 0.0482,
"step": 590
},
{
"epoch": 0.7254901960784313,
"grad_norm": 0.30859375,
"learning_rate": 9.800602364935962e-06,
"loss": 0.0518,
"step": 592
},
{
"epoch": 0.7279411764705882,
"grad_norm": 0.26953125,
"learning_rate": 9.797601779765589e-06,
"loss": 0.0477,
"step": 594
},
{
"epoch": 0.7303921568627451,
"grad_norm": 0.29296875,
"learning_rate": 9.794579252394409e-06,
"loss": 0.0519,
"step": 596
},
{
"epoch": 0.7328431372549019,
"grad_norm": 0.2490234375,
"learning_rate": 9.791534796646185e-06,
"loss": 0.0512,
"step": 598
},
{
"epoch": 0.7352941176470589,
"grad_norm": 0.3125,
"learning_rate": 9.788468426444968e-06,
"loss": 0.0554,
"step": 600
},
{
"epoch": 0.7377450980392157,
"grad_norm": 0.357421875,
"learning_rate": 9.78538015581504e-06,
"loss": 0.0569,
"step": 602
},
{
"epoch": 0.7401960784313726,
"grad_norm": 0.322265625,
"learning_rate": 9.782269998880842e-06,
"loss": 0.0492,
"step": 604
},
{
"epoch": 0.7426470588235294,
"grad_norm": 0.265625,
"learning_rate": 9.779137969866918e-06,
"loss": 0.0534,
"step": 606
},
{
"epoch": 0.7450980392156863,
"grad_norm": 0.251953125,
"learning_rate": 9.775984083097844e-06,
"loss": 0.0565,
"step": 608
},
{
"epoch": 0.7475490196078431,
"grad_norm": 0.26171875,
"learning_rate": 9.772808352998163e-06,
"loss": 0.0548,
"step": 610
},
{
"epoch": 0.75,
"grad_norm": 0.2734375,
"learning_rate": 9.769610794092319e-06,
"loss": 0.0485,
"step": 612
},
{
"epoch": 0.7524509803921569,
"grad_norm": 0.27734375,
"learning_rate": 9.766391421004598e-06,
"loss": 0.0514,
"step": 614
},
{
"epoch": 0.7549019607843137,
"grad_norm": 0.275390625,
"learning_rate": 9.763150248459044e-06,
"loss": 0.0534,
"step": 616
},
{
"epoch": 0.7573529411764706,
"grad_norm": 0.373046875,
"learning_rate": 9.759887291279415e-06,
"loss": 0.0626,
"step": 618
},
{
"epoch": 0.7598039215686274,
"grad_norm": 0.234375,
"learning_rate": 9.756602564389094e-06,
"loss": 0.0476,
"step": 620
},
{
"epoch": 0.7622549019607843,
"grad_norm": 0.373046875,
"learning_rate": 9.753296082811033e-06,
"loss": 0.0474,
"step": 622
},
{
"epoch": 0.7647058823529411,
"grad_norm": 0.267578125,
"learning_rate": 9.749967861667677e-06,
"loss": 0.0523,
"step": 624
},
{
"epoch": 0.7671568627450981,
"grad_norm": 0.2353515625,
"learning_rate": 9.746617916180906e-06,
"loss": 0.0527,
"step": 626
},
{
"epoch": 0.7696078431372549,
"grad_norm": 0.2734375,
"learning_rate": 9.74324626167195e-06,
"loss": 0.0479,
"step": 628
},
{
"epoch": 0.7720588235294118,
"grad_norm": 0.271484375,
"learning_rate": 9.739852913561332e-06,
"loss": 0.0435,
"step": 630
},
{
"epoch": 0.7745098039215687,
"grad_norm": 0.248046875,
"learning_rate": 9.73643788736879e-06,
"loss": 0.0523,
"step": 632
},
{
"epoch": 0.7769607843137255,
"grad_norm": 0.263671875,
"learning_rate": 9.73300119871321e-06,
"loss": 0.0582,
"step": 634
},
{
"epoch": 0.7794117647058824,
"grad_norm": 0.34765625,
"learning_rate": 9.729542863312548e-06,
"loss": 0.0585,
"step": 636
},
{
"epoch": 0.7818627450980392,
"grad_norm": 0.30859375,
"learning_rate": 9.726062896983774e-06,
"loss": 0.047,
"step": 638
},
{
"epoch": 0.7843137254901961,
"grad_norm": 0.35546875,
"learning_rate": 9.722561315642776e-06,
"loss": 0.0505,
"step": 640
},
{
"epoch": 0.7867647058823529,
"grad_norm": 0.298828125,
"learning_rate": 9.719038135304306e-06,
"loss": 0.0518,
"step": 642
},
{
"epoch": 0.7892156862745098,
"grad_norm": 0.232421875,
"learning_rate": 9.715493372081904e-06,
"loss": 0.0509,
"step": 644
},
{
"epoch": 0.7916666666666666,
"grad_norm": 0.2333984375,
"learning_rate": 9.711927042187812e-06,
"loss": 0.0477,
"step": 646
},
{
"epoch": 0.7941176470588235,
"grad_norm": 0.287109375,
"learning_rate": 9.70833916193292e-06,
"loss": 0.0509,
"step": 648
},
{
"epoch": 0.7965686274509803,
"grad_norm": 0.251953125,
"learning_rate": 9.704729747726669e-06,
"loss": 0.044,
"step": 650
},
{
"epoch": 0.7990196078431373,
"grad_norm": 0.26953125,
"learning_rate": 9.701098816076995e-06,
"loss": 0.045,
"step": 652
},
{
"epoch": 0.8014705882352942,
"grad_norm": 0.2392578125,
"learning_rate": 9.697446383590245e-06,
"loss": 0.053,
"step": 654
},
{
"epoch": 0.803921568627451,
"grad_norm": 0.244140625,
"learning_rate": 9.693772466971097e-06,
"loss": 0.0545,
"step": 656
},
{
"epoch": 0.8063725490196079,
"grad_norm": 0.3125,
"learning_rate": 9.690077083022493e-06,
"loss": 0.056,
"step": 658
},
{
"epoch": 0.8088235294117647,
"grad_norm": 0.318359375,
"learning_rate": 9.686360248645554e-06,
"loss": 0.0502,
"step": 660
},
{
"epoch": 0.8112745098039216,
"grad_norm": 0.263671875,
"learning_rate": 9.682621980839511e-06,
"loss": 0.0468,
"step": 662
},
{
"epoch": 0.8137254901960784,
"grad_norm": 0.26171875,
"learning_rate": 9.678862296701615e-06,
"loss": 0.0548,
"step": 664
},
{
"epoch": 0.8161764705882353,
"grad_norm": 0.427734375,
"learning_rate": 9.675081213427076e-06,
"loss": 0.0557,
"step": 666
},
{
"epoch": 0.8186274509803921,
"grad_norm": 0.25390625,
"learning_rate": 9.671278748308963e-06,
"loss": 0.0479,
"step": 668
},
{
"epoch": 0.821078431372549,
"grad_norm": 0.2421875,
"learning_rate": 9.667454918738148e-06,
"loss": 0.0502,
"step": 670
},
{
"epoch": 0.8235294117647058,
"grad_norm": 0.259765625,
"learning_rate": 9.663609742203208e-06,
"loss": 0.0558,
"step": 672
},
{
"epoch": 0.8259803921568627,
"grad_norm": 0.2392578125,
"learning_rate": 9.659743236290353e-06,
"loss": 0.0487,
"step": 674
},
{
"epoch": 0.8284313725490197,
"grad_norm": 0.294921875,
"learning_rate": 9.655855418683345e-06,
"loss": 0.0469,
"step": 676
},
{
"epoch": 0.8308823529411765,
"grad_norm": 0.3046875,
"learning_rate": 9.651946307163417e-06,
"loss": 0.0528,
"step": 678
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.267578125,
"learning_rate": 9.648015919609194e-06,
"loss": 0.0471,
"step": 680
},
{
"epoch": 0.8357843137254902,
"grad_norm": 0.3359375,
"learning_rate": 9.644064273996603e-06,
"loss": 0.0477,
"step": 682
},
{
"epoch": 0.8382352941176471,
"grad_norm": 0.30078125,
"learning_rate": 9.6400913883988e-06,
"loss": 0.0486,
"step": 684
},
{
"epoch": 0.8406862745098039,
"grad_norm": 0.271484375,
"learning_rate": 9.636097280986081e-06,
"loss": 0.0511,
"step": 686
},
{
"epoch": 0.8431372549019608,
"grad_norm": 0.2451171875,
"learning_rate": 9.632081970025808e-06,
"loss": 0.0479,
"step": 688
},
{
"epoch": 0.8455882352941176,
"grad_norm": 0.2099609375,
"learning_rate": 9.62804547388231e-06,
"loss": 0.0471,
"step": 690
},
{
"epoch": 0.8480392156862745,
"grad_norm": 0.318359375,
"learning_rate": 9.623987811016814e-06,
"loss": 0.0528,
"step": 692
},
{
"epoch": 0.8504901960784313,
"grad_norm": 0.333984375,
"learning_rate": 9.619908999987356e-06,
"loss": 0.0469,
"step": 694
},
{
"epoch": 0.8529411764705882,
"grad_norm": 0.2265625,
"learning_rate": 9.615809059448687e-06,
"loss": 0.0438,
"step": 696
},
{
"epoch": 0.8553921568627451,
"grad_norm": 0.2890625,
"learning_rate": 9.611688008152205e-06,
"loss": 0.0521,
"step": 698
},
{
"epoch": 0.8578431372549019,
"grad_norm": 0.302734375,
"learning_rate": 9.607545864945851e-06,
"loss": 0.056,
"step": 700
},
{
"epoch": 0.8602941176470589,
"grad_norm": 0.25390625,
"learning_rate": 9.60338264877404e-06,
"loss": 0.0495,
"step": 702
},
{
"epoch": 0.8627450980392157,
"grad_norm": 0.21875,
"learning_rate": 9.599198378677559e-06,
"loss": 0.0487,
"step": 704
},
{
"epoch": 0.8651960784313726,
"grad_norm": 0.255859375,
"learning_rate": 9.59499307379349e-06,
"loss": 0.0465,
"step": 706
},
{
"epoch": 0.8676470588235294,
"grad_norm": 0.2431640625,
"learning_rate": 9.590766753355117e-06,
"loss": 0.0472,
"step": 708
},
{
"epoch": 0.8700980392156863,
"grad_norm": 0.31640625,
"learning_rate": 9.586519436691843e-06,
"loss": 0.049,
"step": 710
},
{
"epoch": 0.8725490196078431,
"grad_norm": 0.279296875,
"learning_rate": 9.582251143229098e-06,
"loss": 0.049,
"step": 712
},
{
"epoch": 0.875,
"grad_norm": 0.306640625,
"learning_rate": 9.577961892488247e-06,
"loss": 0.0497,
"step": 714
},
{
"epoch": 0.8774509803921569,
"grad_norm": 0.365234375,
"learning_rate": 9.573651704086512e-06,
"loss": 0.0498,
"step": 716
},
{
"epoch": 0.8799019607843137,
"grad_norm": 0.251953125,
"learning_rate": 9.569320597736871e-06,
"loss": 0.0567,
"step": 718
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.2451171875,
"learning_rate": 9.564968593247971e-06,
"loss": 0.0489,
"step": 720
},
{
"epoch": 0.8848039215686274,
"grad_norm": 0.267578125,
"learning_rate": 9.56059571052404e-06,
"loss": 0.047,
"step": 722
},
{
"epoch": 0.8872549019607843,
"grad_norm": 0.27734375,
"learning_rate": 9.556201969564789e-06,
"loss": 0.0522,
"step": 724
},
{
"epoch": 0.8897058823529411,
"grad_norm": 0.26171875,
"learning_rate": 9.551787390465336e-06,
"loss": 0.0497,
"step": 726
},
{
"epoch": 0.8921568627450981,
"grad_norm": 0.259765625,
"learning_rate": 9.547351993416099e-06,
"loss": 0.0476,
"step": 728
},
{
"epoch": 0.8946078431372549,
"grad_norm": 0.28515625,
"learning_rate": 9.542895798702702e-06,
"loss": 0.0538,
"step": 730
},
{
"epoch": 0.8970588235294118,
"grad_norm": 0.373046875,
"learning_rate": 9.538418826705899e-06,
"loss": 0.0461,
"step": 732
},
{
"epoch": 0.8995098039215687,
"grad_norm": 0.30078125,
"learning_rate": 9.533921097901462e-06,
"loss": 0.0531,
"step": 734
},
{
"epoch": 0.9019607843137255,
"grad_norm": 0.283203125,
"learning_rate": 9.529402632860104e-06,
"loss": 0.0551,
"step": 736
},
{
"epoch": 0.9044117647058824,
"grad_norm": 0.283203125,
"learning_rate": 9.524863452247368e-06,
"loss": 0.0506,
"step": 738
},
{
"epoch": 0.9068627450980392,
"grad_norm": 0.294921875,
"learning_rate": 9.52030357682355e-06,
"loss": 0.0508,
"step": 740
},
{
"epoch": 0.9093137254901961,
"grad_norm": 0.25390625,
"learning_rate": 9.515723027443587e-06,
"loss": 0.0518,
"step": 742
},
{
"epoch": 0.9117647058823529,
"grad_norm": 0.30078125,
"learning_rate": 9.511121825056978e-06,
"loss": 0.0495,
"step": 744
},
{
"epoch": 0.9142156862745098,
"grad_norm": 0.275390625,
"learning_rate": 9.506499990707672e-06,
"loss": 0.0547,
"step": 746
},
{
"epoch": 0.9166666666666666,
"grad_norm": 0.326171875,
"learning_rate": 9.501857545533987e-06,
"loss": 0.0501,
"step": 748
},
{
"epoch": 0.9191176470588235,
"grad_norm": 0.30078125,
"learning_rate": 9.497194510768502e-06,
"loss": 0.0517,
"step": 750
},
{
"epoch": 0.9215686274509803,
"grad_norm": 0.26953125,
"learning_rate": 9.492510907737966e-06,
"loss": 0.0451,
"step": 752
},
{
"epoch": 0.9240196078431373,
"grad_norm": 0.283203125,
"learning_rate": 9.487806757863194e-06,
"loss": 0.0568,
"step": 754
},
{
"epoch": 0.9264705882352942,
"grad_norm": 0.302734375,
"learning_rate": 9.483082082658984e-06,
"loss": 0.0525,
"step": 756
},
{
"epoch": 0.928921568627451,
"grad_norm": 0.3203125,
"learning_rate": 9.478336903733996e-06,
"loss": 0.0542,
"step": 758
},
{
"epoch": 0.9313725490196079,
"grad_norm": 0.3359375,
"learning_rate": 9.473571242790672e-06,
"loss": 0.0537,
"step": 760
},
{
"epoch": 0.9338235294117647,
"grad_norm": 0.25,
"learning_rate": 9.468785121625131e-06,
"loss": 0.0552,
"step": 762
},
{
"epoch": 0.9362745098039216,
"grad_norm": 0.265625,
"learning_rate": 9.463978562127065e-06,
"loss": 0.0458,
"step": 764
},
{
"epoch": 0.9387254901960784,
"grad_norm": 0.322265625,
"learning_rate": 9.459151586279643e-06,
"loss": 0.0495,
"step": 766
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.220703125,
"learning_rate": 9.454304216159411e-06,
"loss": 0.0512,
"step": 768
},
{
"epoch": 0.9436274509803921,
"grad_norm": 0.265625,
"learning_rate": 9.44943647393619e-06,
"loss": 0.052,
"step": 770
},
{
"epoch": 0.946078431372549,
"grad_norm": 0.259765625,
"learning_rate": 9.444548381872972e-06,
"loss": 0.0532,
"step": 772
},
{
"epoch": 0.9485294117647058,
"grad_norm": 0.30859375,
"learning_rate": 9.439639962325822e-06,
"loss": 0.049,
"step": 774
},
{
"epoch": 0.9509803921568627,
"grad_norm": 0.32421875,
"learning_rate": 9.434711237743776e-06,
"loss": 0.0489,
"step": 776
},
{
"epoch": 0.9534313725490197,
"grad_norm": 0.283203125,
"learning_rate": 9.429762230668734e-06,
"loss": 0.0497,
"step": 778
},
{
"epoch": 0.9558823529411765,
"grad_norm": 0.333984375,
"learning_rate": 9.42479296373536e-06,
"loss": 0.059,
"step": 780
},
{
"epoch": 0.9583333333333334,
"grad_norm": 0.265625,
"learning_rate": 9.41980345967098e-06,
"loss": 0.0496,
"step": 782
},
{
"epoch": 0.9607843137254902,
"grad_norm": 0.265625,
"learning_rate": 9.414793741295472e-06,
"loss": 0.0512,
"step": 784
},
{
"epoch": 0.9632352941176471,
"grad_norm": 0.25390625,
"learning_rate": 9.409763831521173e-06,
"loss": 0.0541,
"step": 786
},
{
"epoch": 0.9656862745098039,
"grad_norm": 0.26953125,
"learning_rate": 9.404713753352757e-06,
"loss": 0.0495,
"step": 788
},
{
"epoch": 0.9681372549019608,
"grad_norm": 0.34375,
"learning_rate": 9.39964352988715e-06,
"loss": 0.0551,
"step": 790
},
{
"epoch": 0.9705882352941176,
"grad_norm": 0.240234375,
"learning_rate": 9.394553184313407e-06,
"loss": 0.0507,
"step": 792
},
{
"epoch": 0.9730392156862745,
"grad_norm": 0.25,
"learning_rate": 9.389442739912612e-06,
"loss": 0.0433,
"step": 794
},
{
"epoch": 0.9754901960784313,
"grad_norm": 0.298828125,
"learning_rate": 9.38431222005778e-06,
"loss": 0.0549,
"step": 796
},
{
"epoch": 0.9779411764705882,
"grad_norm": 0.287109375,
"learning_rate": 9.379161648213737e-06,
"loss": 0.0508,
"step": 798
},
{
"epoch": 0.9803921568627451,
"grad_norm": 0.22265625,
"learning_rate": 9.373991047937017e-06,
"loss": 0.0481,
"step": 800
},
{
"epoch": 0.9828431372549019,
"grad_norm": 0.3125,
"learning_rate": 9.368800442875763e-06,
"loss": 0.0484,
"step": 802
},
{
"epoch": 0.9852941176470589,
"grad_norm": 0.263671875,
"learning_rate": 9.363589856769602e-06,
"loss": 0.0429,
"step": 804
},
{
"epoch": 0.9877450980392157,
"grad_norm": 0.228515625,
"learning_rate": 9.358359313449553e-06,
"loss": 0.0476,
"step": 806
},
{
"epoch": 0.9901960784313726,
"grad_norm": 0.310546875,
"learning_rate": 9.353108836837907e-06,
"loss": 0.0594,
"step": 808
},
{
"epoch": 0.9926470588235294,
"grad_norm": 0.25390625,
"learning_rate": 9.347838450948123e-06,
"loss": 0.0462,
"step": 810
},
{
"epoch": 0.9950980392156863,
"grad_norm": 0.314453125,
"learning_rate": 9.342548179884716e-06,
"loss": 0.057,
"step": 812
},
{
"epoch": 0.9975490196078431,
"grad_norm": 0.263671875,
"learning_rate": 9.337238047843148e-06,
"loss": 0.0555,
"step": 814
},
{
"epoch": 1.0,
"grad_norm": 0.431640625,
"learning_rate": 9.331908079109718e-06,
"loss": 0.0484,
"step": 816
},
{
"epoch": 1.0,
"eval_loss": 0.05779758095741272,
"eval_runtime": 37.3876,
"eval_samples_per_second": 134.162,
"eval_steps_per_second": 1.07,
"step": 816
},
{
"epoch": 1.0024509803921569,
"grad_norm": 0.2578125,
"learning_rate": 9.326558298061446e-06,
"loss": 0.0465,
"step": 818
},
{
"epoch": 1.0049019607843137,
"grad_norm": 0.296875,
"learning_rate": 9.321188729165967e-06,
"loss": 0.0449,
"step": 820
},
{
"epoch": 1.0073529411764706,
"grad_norm": 0.294921875,
"learning_rate": 9.315799396981417e-06,
"loss": 0.0447,
"step": 822
},
{
"epoch": 1.0098039215686274,
"grad_norm": 0.275390625,
"learning_rate": 9.310390326156324e-06,
"loss": 0.0533,
"step": 824
},
{
"epoch": 1.0122549019607843,
"grad_norm": 0.322265625,
"learning_rate": 9.304961541429487e-06,
"loss": 0.0496,
"step": 826
},
{
"epoch": 1.0147058823529411,
"grad_norm": 0.2060546875,
"learning_rate": 9.299513067629872e-06,
"loss": 0.0439,
"step": 828
},
{
"epoch": 1.017156862745098,
"grad_norm": 0.26171875,
"learning_rate": 9.294044929676493e-06,
"loss": 0.0585,
"step": 830
},
{
"epoch": 1.0196078431372548,
"grad_norm": 0.302734375,
"learning_rate": 9.2885571525783e-06,
"loss": 0.0562,
"step": 832
},
{
"epoch": 1.0220588235294117,
"grad_norm": 0.26171875,
"learning_rate": 9.283049761434059e-06,
"loss": 0.0531,
"step": 834
},
{
"epoch": 1.0245098039215685,
"grad_norm": 0.310546875,
"learning_rate": 9.277522781432254e-06,
"loss": 0.0523,
"step": 836
},
{
"epoch": 1.0269607843137254,
"grad_norm": 0.32421875,
"learning_rate": 9.271976237850951e-06,
"loss": 0.0488,
"step": 838
},
{
"epoch": 1.0294117647058822,
"grad_norm": 0.22265625,
"learning_rate": 9.26641015605769e-06,
"loss": 0.0551,
"step": 840
},
{
"epoch": 1.031862745098039,
"grad_norm": 0.2392578125,
"learning_rate": 9.260824561509378e-06,
"loss": 0.0507,
"step": 842
},
{
"epoch": 1.0343137254901962,
"grad_norm": 0.25,
"learning_rate": 9.255219479752163e-06,
"loss": 0.0472,
"step": 844
},
{
"epoch": 1.036764705882353,
"grad_norm": 0.27734375,
"learning_rate": 9.249594936421312e-06,
"loss": 0.0521,
"step": 846
},
{
"epoch": 1.0392156862745099,
"grad_norm": 0.26171875,
"learning_rate": 9.24395095724111e-06,
"loss": 0.0497,
"step": 848
},
{
"epoch": 1.0416666666666667,
"grad_norm": 0.25390625,
"learning_rate": 9.238287568024732e-06,
"loss": 0.0485,
"step": 850
},
{
"epoch": 1.0441176470588236,
"grad_norm": 0.220703125,
"learning_rate": 9.23260479467412e-06,
"loss": 0.0501,
"step": 852
},
{
"epoch": 1.0465686274509804,
"grad_norm": 0.2451171875,
"learning_rate": 9.226902663179877e-06,
"loss": 0.0515,
"step": 854
},
{
"epoch": 1.0490196078431373,
"grad_norm": 0.30859375,
"learning_rate": 9.221181199621141e-06,
"loss": 0.0521,
"step": 856
},
{
"epoch": 1.0514705882352942,
"grad_norm": 0.353515625,
"learning_rate": 9.215440430165463e-06,
"loss": 0.0476,
"step": 858
},
{
"epoch": 1.053921568627451,
"grad_norm": 0.255859375,
"learning_rate": 9.209680381068698e-06,
"loss": 0.044,
"step": 860
},
{
"epoch": 1.0563725490196079,
"grad_norm": 0.36328125,
"learning_rate": 9.203901078674868e-06,
"loss": 0.0551,
"step": 862
},
{
"epoch": 1.0588235294117647,
"grad_norm": 0.279296875,
"learning_rate": 9.198102549416062e-06,
"loss": 0.0482,
"step": 864
},
{
"epoch": 1.0612745098039216,
"grad_norm": 0.2314453125,
"learning_rate": 9.192284819812298e-06,
"loss": 0.0454,
"step": 866
},
{
"epoch": 1.0637254901960784,
"grad_norm": 0.294921875,
"learning_rate": 9.18644791647141e-06,
"loss": 0.0524,
"step": 868
},
{
"epoch": 1.0661764705882353,
"grad_norm": 0.248046875,
"learning_rate": 9.180591866088924e-06,
"loss": 0.0502,
"step": 870
},
{
"epoch": 1.0686274509803921,
"grad_norm": 0.279296875,
"learning_rate": 9.174716695447935e-06,
"loss": 0.0509,
"step": 872
},
{
"epoch": 1.071078431372549,
"grad_norm": 0.2470703125,
"learning_rate": 9.168822431418992e-06,
"loss": 0.0496,
"step": 874
},
{
"epoch": 1.0735294117647058,
"grad_norm": 0.2177734375,
"learning_rate": 9.162909100959962e-06,
"loss": 0.0521,
"step": 876
},
{
"epoch": 1.0759803921568627,
"grad_norm": 0.25,
"learning_rate": 9.156976731115919e-06,
"loss": 0.0529,
"step": 878
},
{
"epoch": 1.0784313725490196,
"grad_norm": 0.25390625,
"learning_rate": 9.15102534901901e-06,
"loss": 0.0494,
"step": 880
},
{
"epoch": 1.0808823529411764,
"grad_norm": 0.279296875,
"learning_rate": 9.145054981888342e-06,
"loss": 0.0589,
"step": 882
},
{
"epoch": 1.0833333333333333,
"grad_norm": 0.2490234375,
"learning_rate": 9.139065657029845e-06,
"loss": 0.0437,
"step": 884
},
{
"epoch": 1.0857843137254901,
"grad_norm": 0.25390625,
"learning_rate": 9.13305740183616e-06,
"loss": 0.0472,
"step": 886
},
{
"epoch": 1.088235294117647,
"grad_norm": 0.27734375,
"learning_rate": 9.127030243786502e-06,
"loss": 0.0539,
"step": 888
},
{
"epoch": 1.0906862745098038,
"grad_norm": 0.23828125,
"learning_rate": 9.120984210446544e-06,
"loss": 0.0469,
"step": 890
},
{
"epoch": 1.093137254901961,
"grad_norm": 0.267578125,
"learning_rate": 9.114919329468283e-06,
"loss": 0.0478,
"step": 892
},
{
"epoch": 1.0955882352941178,
"grad_norm": 0.24609375,
"learning_rate": 9.108835628589919e-06,
"loss": 0.0506,
"step": 894
},
{
"epoch": 1.0980392156862746,
"grad_norm": 0.236328125,
"learning_rate": 9.102733135635727e-06,
"loss": 0.0557,
"step": 896
},
{
"epoch": 1.1004901960784315,
"grad_norm": 0.28515625,
"learning_rate": 9.096611878515926e-06,
"loss": 0.0482,
"step": 898
},
{
"epoch": 1.1029411764705883,
"grad_norm": 0.2734375,
"learning_rate": 9.09047188522656e-06,
"loss": 0.0565,
"step": 900
},
{
"epoch": 1.1053921568627452,
"grad_norm": 0.29296875,
"learning_rate": 9.084313183849356e-06,
"loss": 0.0445,
"step": 902
},
{
"epoch": 1.107843137254902,
"grad_norm": 0.29296875,
"learning_rate": 9.078135802551611e-06,
"loss": 0.0456,
"step": 904
},
{
"epoch": 1.1102941176470589,
"grad_norm": 0.21875,
"learning_rate": 9.071939769586054e-06,
"loss": 0.0526,
"step": 906
},
{
"epoch": 1.1127450980392157,
"grad_norm": 0.2197265625,
"learning_rate": 9.065725113290718e-06,
"loss": 0.0469,
"step": 908
},
{
"epoch": 1.1151960784313726,
"grad_norm": 0.318359375,
"learning_rate": 9.059491862088811e-06,
"loss": 0.0567,
"step": 910
},
{
"epoch": 1.1176470588235294,
"grad_norm": 0.2197265625,
"learning_rate": 9.053240044488587e-06,
"loss": 0.0513,
"step": 912
},
{
"epoch": 1.1200980392156863,
"grad_norm": 0.32421875,
"learning_rate": 9.046969689083215e-06,
"loss": 0.048,
"step": 914
},
{
"epoch": 1.1225490196078431,
"grad_norm": 0.265625,
"learning_rate": 9.040680824550648e-06,
"loss": 0.0469,
"step": 916
},
{
"epoch": 1.125,
"grad_norm": 0.251953125,
"learning_rate": 9.034373479653491e-06,
"loss": 0.0491,
"step": 918
},
{
"epoch": 1.1274509803921569,
"grad_norm": 0.2470703125,
"learning_rate": 9.028047683238872e-06,
"loss": 0.0483,
"step": 920
},
{
"epoch": 1.1299019607843137,
"grad_norm": 0.2578125,
"learning_rate": 9.021703464238308e-06,
"loss": 0.0507,
"step": 922
},
{
"epoch": 1.1323529411764706,
"grad_norm": 0.2451171875,
"learning_rate": 9.01534085166757e-06,
"loss": 0.051,
"step": 924
},
{
"epoch": 1.1348039215686274,
"grad_norm": 0.2373046875,
"learning_rate": 9.008959874626559e-06,
"loss": 0.0472,
"step": 926
},
{
"epoch": 1.1372549019607843,
"grad_norm": 0.30859375,
"learning_rate": 9.002560562299164e-06,
"loss": 0.0558,
"step": 928
},
{
"epoch": 1.1397058823529411,
"grad_norm": 0.2431640625,
"learning_rate": 8.99614294395313e-06,
"loss": 0.0471,
"step": 930
},
{
"epoch": 1.142156862745098,
"grad_norm": 0.27734375,
"learning_rate": 8.989707048939931e-06,
"loss": 0.059,
"step": 932
},
{
"epoch": 1.1446078431372548,
"grad_norm": 0.2255859375,
"learning_rate": 8.983252906694626e-06,
"loss": 0.0441,
"step": 934
},
{
"epoch": 1.1470588235294117,
"grad_norm": 0.220703125,
"learning_rate": 8.97678054673573e-06,
"loss": 0.0446,
"step": 936
},
{
"epoch": 1.1495098039215685,
"grad_norm": 0.294921875,
"learning_rate": 8.970289998665083e-06,
"loss": 0.0477,
"step": 938
},
{
"epoch": 1.1519607843137254,
"grad_norm": 0.28125,
"learning_rate": 8.963781292167701e-06,
"loss": 0.0545,
"step": 940
},
{
"epoch": 1.1544117647058822,
"grad_norm": 0.2451171875,
"learning_rate": 8.957254457011659e-06,
"loss": 0.0551,
"step": 942
},
{
"epoch": 1.156862745098039,
"grad_norm": 0.26171875,
"learning_rate": 8.950709523047936e-06,
"loss": 0.0434,
"step": 944
},
{
"epoch": 1.159313725490196,
"grad_norm": 0.216796875,
"learning_rate": 8.944146520210293e-06,
"loss": 0.0482,
"step": 946
},
{
"epoch": 1.161764705882353,
"grad_norm": 0.33203125,
"learning_rate": 8.937565478515127e-06,
"loss": 0.0514,
"step": 948
},
{
"epoch": 1.1642156862745099,
"grad_norm": 0.251953125,
"learning_rate": 8.930966428061339e-06,
"loss": 0.0465,
"step": 950
},
{
"epoch": 1.1666666666666667,
"grad_norm": 0.25390625,
"learning_rate": 8.924349399030194e-06,
"loss": 0.0482,
"step": 952
},
{
"epoch": 1.1691176470588236,
"grad_norm": 0.2421875,
"learning_rate": 8.917714421685183e-06,
"loss": 0.048,
"step": 954
},
{
"epoch": 1.1715686274509804,
"grad_norm": 0.26171875,
"learning_rate": 8.911061526371886e-06,
"loss": 0.0531,
"step": 956
},
{
"epoch": 1.1740196078431373,
"grad_norm": 0.240234375,
"learning_rate": 8.90439074351783e-06,
"loss": 0.0488,
"step": 958
},
{
"epoch": 1.1764705882352942,
"grad_norm": 0.2578125,
"learning_rate": 8.897702103632355e-06,
"loss": 0.0469,
"step": 960
},
{
"epoch": 1.178921568627451,
"grad_norm": 0.2578125,
"learning_rate": 8.89099563730647e-06,
"loss": 0.0516,
"step": 962
},
{
"epoch": 1.1813725490196079,
"grad_norm": 0.255859375,
"learning_rate": 8.884271375212714e-06,
"loss": 0.0484,
"step": 964
},
{
"epoch": 1.1838235294117647,
"grad_norm": 0.26953125,
"learning_rate": 8.877529348105019e-06,
"loss": 0.0523,
"step": 966
},
{
"epoch": 1.1862745098039216,
"grad_norm": 0.20703125,
"learning_rate": 8.87076958681856e-06,
"loss": 0.045,
"step": 968
},
{
"epoch": 1.1887254901960784,
"grad_norm": 0.2470703125,
"learning_rate": 8.86399212226963e-06,
"loss": 0.042,
"step": 970
},
{
"epoch": 1.1911764705882353,
"grad_norm": 0.2734375,
"learning_rate": 8.857196985455481e-06,
"loss": 0.047,
"step": 972
},
{
"epoch": 1.1936274509803921,
"grad_norm": 0.240234375,
"learning_rate": 8.850384207454195e-06,
"loss": 0.0444,
"step": 974
},
{
"epoch": 1.196078431372549,
"grad_norm": 0.26171875,
"learning_rate": 8.84355381942454e-06,
"loss": 0.0487,
"step": 976
},
{
"epoch": 1.1985294117647058,
"grad_norm": 0.228515625,
"learning_rate": 8.836705852605814e-06,
"loss": 0.0532,
"step": 978
},
{
"epoch": 1.2009803921568627,
"grad_norm": 0.2431640625,
"learning_rate": 8.829840338317726e-06,
"loss": 0.0459,
"step": 980
},
{
"epoch": 1.2034313725490196,
"grad_norm": 0.212890625,
"learning_rate": 8.822957307960232e-06,
"loss": 0.0443,
"step": 982
},
{
"epoch": 1.2058823529411764,
"grad_norm": 0.2109375,
"learning_rate": 8.8160567930134e-06,
"loss": 0.0516,
"step": 984
},
{
"epoch": 1.2083333333333333,
"grad_norm": 0.2099609375,
"learning_rate": 8.809138825037268e-06,
"loss": 0.0471,
"step": 986
},
{
"epoch": 1.2107843137254901,
"grad_norm": 0.2060546875,
"learning_rate": 8.80220343567169e-06,
"loss": 0.0471,
"step": 988
},
{
"epoch": 1.213235294117647,
"grad_norm": 0.24609375,
"learning_rate": 8.795250656636207e-06,
"loss": 0.0466,
"step": 990
},
{
"epoch": 1.215686274509804,
"grad_norm": 0.28125,
"learning_rate": 8.788280519729891e-06,
"loss": 0.0518,
"step": 992
},
{
"epoch": 1.218137254901961,
"grad_norm": 0.28515625,
"learning_rate": 8.781293056831192e-06,
"loss": 0.0503,
"step": 994
},
{
"epoch": 1.2205882352941178,
"grad_norm": 0.26953125,
"learning_rate": 8.774288299897816e-06,
"loss": 0.0483,
"step": 996
},
{
"epoch": 1.2230392156862746,
"grad_norm": 0.23828125,
"learning_rate": 8.767266280966551e-06,
"loss": 0.0529,
"step": 998
},
{
"epoch": 1.2254901960784315,
"grad_norm": 0.251953125,
"learning_rate": 8.760227032153147e-06,
"loss": 0.0516,
"step": 1000
},
{
"epoch": 1.2279411764705883,
"grad_norm": 0.2119140625,
"learning_rate": 8.753170585652148e-06,
"loss": 0.0481,
"step": 1002
},
{
"epoch": 1.2303921568627452,
"grad_norm": 0.279296875,
"learning_rate": 8.746096973736755e-06,
"loss": 0.0508,
"step": 1004
},
{
"epoch": 1.232843137254902,
"grad_norm": 0.25390625,
"learning_rate": 8.739006228758675e-06,
"loss": 0.0481,
"step": 1006
},
{
"epoch": 1.2352941176470589,
"grad_norm": 0.28125,
"learning_rate": 8.731898383147976e-06,
"loss": 0.0459,
"step": 1008
},
{
"epoch": 1.2377450980392157,
"grad_norm": 0.2431640625,
"learning_rate": 8.724773469412936e-06,
"loss": 0.0461,
"step": 1010
},
{
"epoch": 1.2401960784313726,
"grad_norm": 0.2734375,
"learning_rate": 8.7176315201399e-06,
"loss": 0.0507,
"step": 1012
},
{
"epoch": 1.2426470588235294,
"grad_norm": 0.2578125,
"learning_rate": 8.710472567993117e-06,
"loss": 0.0436,
"step": 1014
},
{
"epoch": 1.2450980392156863,
"grad_norm": 0.2451171875,
"learning_rate": 8.70329664571461e-06,
"loss": 0.05,
"step": 1016
},
{
"epoch": 1.2475490196078431,
"grad_norm": 0.310546875,
"learning_rate": 8.696103786124009e-06,
"loss": 0.0487,
"step": 1018
},
{
"epoch": 1.25,
"grad_norm": 0.2451171875,
"learning_rate": 8.68889402211841e-06,
"loss": 0.0485,
"step": 1020
},
{
"epoch": 1.2524509803921569,
"grad_norm": 0.263671875,
"learning_rate": 8.681667386672223e-06,
"loss": 0.0495,
"step": 1022
},
{
"epoch": 1.2549019607843137,
"grad_norm": 0.267578125,
"learning_rate": 8.674423912837025e-06,
"loss": 0.0495,
"step": 1024
},
{
"epoch": 1.2573529411764706,
"grad_norm": 0.263671875,
"learning_rate": 8.667163633741397e-06,
"loss": 0.0503,
"step": 1026
},
{
"epoch": 1.2598039215686274,
"grad_norm": 0.27734375,
"learning_rate": 8.659886582590787e-06,
"loss": 0.0484,
"step": 1028
},
{
"epoch": 1.2622549019607843,
"grad_norm": 0.263671875,
"learning_rate": 8.652592792667346e-06,
"loss": 0.049,
"step": 1030
},
{
"epoch": 1.2647058823529411,
"grad_norm": 0.291015625,
"learning_rate": 8.645282297329785e-06,
"loss": 0.0481,
"step": 1032
},
{
"epoch": 1.267156862745098,
"grad_norm": 0.259765625,
"learning_rate": 8.637955130013213e-06,
"loss": 0.0451,
"step": 1034
},
{
"epoch": 1.2696078431372548,
"grad_norm": 0.2373046875,
"learning_rate": 8.630611324228996e-06,
"loss": 0.0491,
"step": 1036
},
{
"epoch": 1.2720588235294117,
"grad_norm": 0.26953125,
"learning_rate": 8.623250913564598e-06,
"loss": 0.0513,
"step": 1038
},
{
"epoch": 1.2745098039215685,
"grad_norm": 0.265625,
"learning_rate": 8.615873931683417e-06,
"loss": 0.0434,
"step": 1040
},
{
"epoch": 1.2769607843137254,
"grad_norm": 0.22265625,
"learning_rate": 8.608480412324652e-06,
"loss": 0.0473,
"step": 1042
},
{
"epoch": 1.2794117647058822,
"grad_norm": 0.251953125,
"learning_rate": 8.60107038930313e-06,
"loss": 0.05,
"step": 1044
},
{
"epoch": 1.281862745098039,
"grad_norm": 0.23828125,
"learning_rate": 8.593643896509163e-06,
"loss": 0.0536,
"step": 1046
},
{
"epoch": 1.284313725490196,
"grad_norm": 0.25,
"learning_rate": 8.586200967908385e-06,
"loss": 0.0482,
"step": 1048
},
{
"epoch": 1.2867647058823528,
"grad_norm": 0.25,
"learning_rate": 8.578741637541605e-06,
"loss": 0.0498,
"step": 1050
},
{
"epoch": 1.2892156862745099,
"grad_norm": 0.208984375,
"learning_rate": 8.571265939524642e-06,
"loss": 0.0461,
"step": 1052
},
{
"epoch": 1.2916666666666667,
"grad_norm": 0.23046875,
"learning_rate": 8.563773908048177e-06,
"loss": 0.0439,
"step": 1054
},
{
"epoch": 1.2941176470588236,
"grad_norm": 0.23046875,
"learning_rate": 8.556265577377593e-06,
"loss": 0.0433,
"step": 1056
},
{
"epoch": 1.2965686274509804,
"grad_norm": 0.212890625,
"learning_rate": 8.548740981852816e-06,
"loss": 0.0462,
"step": 1058
},
{
"epoch": 1.2990196078431373,
"grad_norm": 0.296875,
"learning_rate": 8.541200155888166e-06,
"loss": 0.051,
"step": 1060
},
{
"epoch": 1.3014705882352942,
"grad_norm": 0.2451171875,
"learning_rate": 8.533643133972186e-06,
"loss": 0.0463,
"step": 1062
},
{
"epoch": 1.303921568627451,
"grad_norm": 0.27734375,
"learning_rate": 8.5260699506675e-06,
"loss": 0.0513,
"step": 1064
},
{
"epoch": 1.3063725490196079,
"grad_norm": 0.2412109375,
"learning_rate": 8.518480640610644e-06,
"loss": 0.0478,
"step": 1066
},
{
"epoch": 1.3088235294117647,
"grad_norm": 0.279296875,
"learning_rate": 8.510875238511911e-06,
"loss": 0.0498,
"step": 1068
},
{
"epoch": 1.3112745098039216,
"grad_norm": 0.267578125,
"learning_rate": 8.503253779155193e-06,
"loss": 0.047,
"step": 1070
},
{
"epoch": 1.3137254901960784,
"grad_norm": 0.205078125,
"learning_rate": 8.495616297397819e-06,
"loss": 0.0447,
"step": 1072
},
{
"epoch": 1.3161764705882353,
"grad_norm": 0.263671875,
"learning_rate": 8.487962828170403e-06,
"loss": 0.048,
"step": 1074
},
{
"epoch": 1.3186274509803921,
"grad_norm": 0.232421875,
"learning_rate": 8.480293406476671e-06,
"loss": 0.0481,
"step": 1076
},
{
"epoch": 1.321078431372549,
"grad_norm": 0.26171875,
"learning_rate": 8.472608067393317e-06,
"loss": 0.0429,
"step": 1078
},
{
"epoch": 1.3235294117647058,
"grad_norm": 0.2412109375,
"learning_rate": 8.464906846069827e-06,
"loss": 0.0497,
"step": 1080
},
{
"epoch": 1.3259803921568627,
"grad_norm": 0.2275390625,
"learning_rate": 8.457189777728333e-06,
"loss": 0.046,
"step": 1082
},
{
"epoch": 1.3284313725490196,
"grad_norm": 0.21484375,
"learning_rate": 8.449456897663438e-06,
"loss": 0.0416,
"step": 1084
},
{
"epoch": 1.3308823529411764,
"grad_norm": 0.259765625,
"learning_rate": 8.441708241242062e-06,
"loss": 0.0463,
"step": 1086
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.263671875,
"learning_rate": 8.433943843903286e-06,
"loss": 0.0512,
"step": 1088
},
{
"epoch": 1.3357843137254901,
"grad_norm": 0.21484375,
"learning_rate": 8.426163741158175e-06,
"loss": 0.0481,
"step": 1090
},
{
"epoch": 1.3382352941176472,
"grad_norm": 0.2734375,
"learning_rate": 8.41836796858963e-06,
"loss": 0.0483,
"step": 1092
},
{
"epoch": 1.340686274509804,
"grad_norm": 0.2216796875,
"learning_rate": 8.410556561852212e-06,
"loss": 0.0557,
"step": 1094
},
{
"epoch": 1.343137254901961,
"grad_norm": 0.2177734375,
"learning_rate": 8.402729556671995e-06,
"loss": 0.045,
"step": 1096
},
{
"epoch": 1.3455882352941178,
"grad_norm": 0.25390625,
"learning_rate": 8.39488698884639e-06,
"loss": 0.0488,
"step": 1098
},
{
"epoch": 1.3480392156862746,
"grad_norm": 0.2353515625,
"learning_rate": 8.38702889424398e-06,
"loss": 0.0443,
"step": 1100
},
{
"epoch": 1.3504901960784315,
"grad_norm": 0.27734375,
"learning_rate": 8.379155308804368e-06,
"loss": 0.0494,
"step": 1102
},
{
"epoch": 1.3529411764705883,
"grad_norm": 0.232421875,
"learning_rate": 8.371266268538002e-06,
"loss": 0.0453,
"step": 1104
},
{
"epoch": 1.3553921568627452,
"grad_norm": 0.310546875,
"learning_rate": 8.363361809526013e-06,
"loss": 0.053,
"step": 1106
},
{
"epoch": 1.357843137254902,
"grad_norm": 0.2333984375,
"learning_rate": 8.355441967920056e-06,
"loss": 0.045,
"step": 1108
},
{
"epoch": 1.3602941176470589,
"grad_norm": 0.2314453125,
"learning_rate": 8.347506779942136e-06,
"loss": 0.0507,
"step": 1110
},
{
"epoch": 1.3627450980392157,
"grad_norm": 0.28125,
"learning_rate": 8.33955628188444e-06,
"loss": 0.0524,
"step": 1112
},
{
"epoch": 1.3651960784313726,
"grad_norm": 0.22265625,
"learning_rate": 8.331590510109188e-06,
"loss": 0.0456,
"step": 1114
},
{
"epoch": 1.3676470588235294,
"grad_norm": 0.220703125,
"learning_rate": 8.323609501048446e-06,
"loss": 0.0459,
"step": 1116
},
{
"epoch": 1.3700980392156863,
"grad_norm": 0.259765625,
"learning_rate": 8.315613291203977e-06,
"loss": 0.0533,
"step": 1118
},
{
"epoch": 1.3725490196078431,
"grad_norm": 0.251953125,
"learning_rate": 8.30760191714706e-06,
"loss": 0.0475,
"step": 1120
},
{
"epoch": 1.375,
"grad_norm": 0.3203125,
"learning_rate": 8.299575415518327e-06,
"loss": 0.0435,
"step": 1122
},
{
"epoch": 1.3774509803921569,
"grad_norm": 0.27734375,
"learning_rate": 8.291533823027607e-06,
"loss": 0.0504,
"step": 1124
},
{
"epoch": 1.3799019607843137,
"grad_norm": 0.2294921875,
"learning_rate": 8.283477176453738e-06,
"loss": 0.0497,
"step": 1126
},
{
"epoch": 1.3823529411764706,
"grad_norm": 0.25390625,
"learning_rate": 8.275405512644412e-06,
"loss": 0.0457,
"step": 1128
},
{
"epoch": 1.3848039215686274,
"grad_norm": 0.220703125,
"learning_rate": 8.267318868516007e-06,
"loss": 0.0482,
"step": 1130
},
{
"epoch": 1.3872549019607843,
"grad_norm": 0.25,
"learning_rate": 8.259217281053411e-06,
"loss": 0.0475,
"step": 1132
},
{
"epoch": 1.3897058823529411,
"grad_norm": 0.2333984375,
"learning_rate": 8.25110078730986e-06,
"loss": 0.0485,
"step": 1134
},
{
"epoch": 1.392156862745098,
"grad_norm": 0.21875,
"learning_rate": 8.242969424406757e-06,
"loss": 0.0452,
"step": 1136
},
{
"epoch": 1.3946078431372548,
"grad_norm": 0.31640625,
"learning_rate": 8.234823229533519e-06,
"loss": 0.0516,
"step": 1138
},
{
"epoch": 1.3970588235294117,
"grad_norm": 0.25390625,
"learning_rate": 8.226662239947396e-06,
"loss": 0.0467,
"step": 1140
},
{
"epoch": 1.3995098039215685,
"grad_norm": 0.25390625,
"learning_rate": 8.2184864929733e-06,
"loss": 0.0482,
"step": 1142
},
{
"epoch": 1.4019607843137254,
"grad_norm": 0.203125,
"learning_rate": 8.21029602600364e-06,
"loss": 0.0426,
"step": 1144
},
{
"epoch": 1.4044117647058822,
"grad_norm": 0.251953125,
"learning_rate": 8.202090876498144e-06,
"loss": 0.0472,
"step": 1146
},
{
"epoch": 1.406862745098039,
"grad_norm": 0.26953125,
"learning_rate": 8.193871081983698e-06,
"loss": 0.051,
"step": 1148
},
{
"epoch": 1.409313725490196,
"grad_norm": 0.2578125,
"learning_rate": 8.185636680054161e-06,
"loss": 0.0483,
"step": 1150
},
{
"epoch": 1.4117647058823528,
"grad_norm": 0.2353515625,
"learning_rate": 8.177387708370208e-06,
"loss": 0.0428,
"step": 1152
},
{
"epoch": 1.4142156862745099,
"grad_norm": 0.208984375,
"learning_rate": 8.169124204659142e-06,
"loss": 0.0431,
"step": 1154
},
{
"epoch": 1.4166666666666667,
"grad_norm": 0.25390625,
"learning_rate": 8.160846206714734e-06,
"loss": 0.0534,
"step": 1156
},
{
"epoch": 1.4191176470588236,
"grad_norm": 0.2421875,
"learning_rate": 8.152553752397043e-06,
"loss": 0.0481,
"step": 1158
},
{
"epoch": 1.4215686274509804,
"grad_norm": 0.263671875,
"learning_rate": 8.14424687963225e-06,
"loss": 0.049,
"step": 1160
},
{
"epoch": 1.4240196078431373,
"grad_norm": 0.291015625,
"learning_rate": 8.135925626412474e-06,
"loss": 0.0469,
"step": 1162
},
{
"epoch": 1.4264705882352942,
"grad_norm": 0.197265625,
"learning_rate": 8.127590030795608e-06,
"loss": 0.0484,
"step": 1164
},
{
"epoch": 1.428921568627451,
"grad_norm": 0.255859375,
"learning_rate": 8.11924013090514e-06,
"loss": 0.0497,
"step": 1166
},
{
"epoch": 1.4313725490196079,
"grad_norm": 0.306640625,
"learning_rate": 8.110875964929976e-06,
"loss": 0.0496,
"step": 1168
},
{
"epoch": 1.4338235294117647,
"grad_norm": 0.25,
"learning_rate": 8.102497571124276e-06,
"loss": 0.0516,
"step": 1170
},
{
"epoch": 1.4362745098039216,
"grad_norm": 0.302734375,
"learning_rate": 8.09410498780727e-06,
"loss": 0.0475,
"step": 1172
},
{
"epoch": 1.4387254901960784,
"grad_norm": 0.29296875,
"learning_rate": 8.085698253363081e-06,
"loss": 0.0439,
"step": 1174
},
{
"epoch": 1.4411764705882353,
"grad_norm": 0.36328125,
"learning_rate": 8.077277406240556e-06,
"loss": 0.0481,
"step": 1176
},
{
"epoch": 1.4436274509803921,
"grad_norm": 0.2353515625,
"learning_rate": 8.068842484953086e-06,
"loss": 0.0456,
"step": 1178
},
{
"epoch": 1.446078431372549,
"grad_norm": 0.279296875,
"learning_rate": 8.060393528078435e-06,
"loss": 0.0421,
"step": 1180
},
{
"epoch": 1.4485294117647058,
"grad_norm": 0.30078125,
"learning_rate": 8.051930574258554e-06,
"loss": 0.0456,
"step": 1182
},
{
"epoch": 1.4509803921568627,
"grad_norm": 0.2138671875,
"learning_rate": 8.043453662199416e-06,
"loss": 0.0496,
"step": 1184
},
{
"epoch": 1.4534313725490196,
"grad_norm": 0.390625,
"learning_rate": 8.03496283067083e-06,
"loss": 0.0451,
"step": 1186
},
{
"epoch": 1.4558823529411764,
"grad_norm": 0.265625,
"learning_rate": 8.026458118506264e-06,
"loss": 0.0507,
"step": 1188
},
{
"epoch": 1.4583333333333333,
"grad_norm": 0.265625,
"learning_rate": 8.017939564602679e-06,
"loss": 0.0482,
"step": 1190
},
{
"epoch": 1.4607843137254901,
"grad_norm": 0.2490234375,
"learning_rate": 8.009407207920333e-06,
"loss": 0.0466,
"step": 1192
},
{
"epoch": 1.4632352941176472,
"grad_norm": 0.30859375,
"learning_rate": 8.000861087482615e-06,
"loss": 0.0468,
"step": 1194
},
{
"epoch": 1.465686274509804,
"grad_norm": 0.197265625,
"learning_rate": 7.992301242375866e-06,
"loss": 0.0461,
"step": 1196
},
{
"epoch": 1.468137254901961,
"grad_norm": 0.2890625,
"learning_rate": 7.983727711749194e-06,
"loss": 0.0508,
"step": 1198
},
{
"epoch": 1.4705882352941178,
"grad_norm": 0.33984375,
"learning_rate": 7.975140534814304e-06,
"loss": 0.0545,
"step": 1200
},
{
"epoch": 1.4730392156862746,
"grad_norm": 0.2333984375,
"learning_rate": 7.96653975084531e-06,
"loss": 0.0489,
"step": 1202
},
{
"epoch": 1.4754901960784315,
"grad_norm": 0.26953125,
"learning_rate": 7.957925399178556e-06,
"loss": 0.0472,
"step": 1204
},
{
"epoch": 1.4779411764705883,
"grad_norm": 0.2431640625,
"learning_rate": 7.949297519212444e-06,
"loss": 0.0494,
"step": 1206
},
{
"epoch": 1.4803921568627452,
"grad_norm": 0.294921875,
"learning_rate": 7.940656150407247e-06,
"loss": 0.048,
"step": 1208
},
{
"epoch": 1.482843137254902,
"grad_norm": 0.310546875,
"learning_rate": 7.932001332284929e-06,
"loss": 0.042,
"step": 1210
},
{
"epoch": 1.4852941176470589,
"grad_norm": 0.2255859375,
"learning_rate": 7.923333104428967e-06,
"loss": 0.0434,
"step": 1212
},
{
"epoch": 1.4877450980392157,
"grad_norm": 0.2216796875,
"learning_rate": 7.91465150648417e-06,
"loss": 0.0374,
"step": 1214
},
{
"epoch": 1.4901960784313726,
"grad_norm": 0.310546875,
"learning_rate": 7.90595657815649e-06,
"loss": 0.0457,
"step": 1216
},
{
"epoch": 1.4926470588235294,
"grad_norm": 0.212890625,
"learning_rate": 7.897248359212852e-06,
"loss": 0.0445,
"step": 1218
},
{
"epoch": 1.4950980392156863,
"grad_norm": 0.234375,
"learning_rate": 7.888526889480967e-06,
"loss": 0.045,
"step": 1220
},
{
"epoch": 1.4975490196078431,
"grad_norm": 0.296875,
"learning_rate": 7.879792208849146e-06,
"loss": 0.0496,
"step": 1222
},
{
"epoch": 1.5,
"grad_norm": 0.287109375,
"learning_rate": 7.871044357266124e-06,
"loss": 0.0465,
"step": 1224
},
{
"epoch": 1.5024509803921569,
"grad_norm": 0.291015625,
"learning_rate": 7.862283374740874e-06,
"loss": 0.0486,
"step": 1226
},
{
"epoch": 1.5049019607843137,
"grad_norm": 0.265625,
"learning_rate": 7.85350930134242e-06,
"loss": 0.0485,
"step": 1228
},
{
"epoch": 1.5073529411764706,
"grad_norm": 0.2314453125,
"learning_rate": 7.844722177199663e-06,
"loss": 0.0483,
"step": 1230
},
{
"epoch": 1.5098039215686274,
"grad_norm": 0.283203125,
"learning_rate": 7.835922042501194e-06,
"loss": 0.0404,
"step": 1232
},
{
"epoch": 1.5122549019607843,
"grad_norm": 0.234375,
"learning_rate": 7.827108937495102e-06,
"loss": 0.0465,
"step": 1234
},
{
"epoch": 1.5147058823529411,
"grad_norm": 0.21484375,
"learning_rate": 7.818282902488806e-06,
"loss": 0.0448,
"step": 1236
},
{
"epoch": 1.517156862745098,
"grad_norm": 0.2255859375,
"learning_rate": 7.809443977848849e-06,
"loss": 0.0504,
"step": 1238
},
{
"epoch": 1.5196078431372548,
"grad_norm": 0.240234375,
"learning_rate": 7.800592204000739e-06,
"loss": 0.0478,
"step": 1240
},
{
"epoch": 1.5220588235294117,
"grad_norm": 0.294921875,
"learning_rate": 7.791727621428741e-06,
"loss": 0.0464,
"step": 1242
},
{
"epoch": 1.5245098039215685,
"grad_norm": 0.26171875,
"learning_rate": 7.782850270675705e-06,
"loss": 0.0435,
"step": 1244
},
{
"epoch": 1.5269607843137254,
"grad_norm": 0.205078125,
"learning_rate": 7.773960192342882e-06,
"loss": 0.0448,
"step": 1246
},
{
"epoch": 1.5294117647058822,
"grad_norm": 0.30078125,
"learning_rate": 7.765057427089726e-06,
"loss": 0.0533,
"step": 1248
},
{
"epoch": 1.531862745098039,
"grad_norm": 0.2080078125,
"learning_rate": 7.75614201563372e-06,
"loss": 0.0448,
"step": 1250
},
{
"epoch": 1.534313725490196,
"grad_norm": 0.234375,
"learning_rate": 7.74721399875018e-06,
"loss": 0.0453,
"step": 1252
},
{
"epoch": 1.5367647058823528,
"grad_norm": 0.2353515625,
"learning_rate": 7.73827341727208e-06,
"loss": 0.0458,
"step": 1254
},
{
"epoch": 1.5392156862745097,
"grad_norm": 0.25390625,
"learning_rate": 7.729320312089862e-06,
"loss": 0.0506,
"step": 1256
},
{
"epoch": 1.5416666666666665,
"grad_norm": 0.3359375,
"learning_rate": 7.720354724151237e-06,
"loss": 0.0467,
"step": 1258
},
{
"epoch": 1.5441176470588234,
"grad_norm": 0.224609375,
"learning_rate": 7.71137669446101e-06,
"loss": 0.0498,
"step": 1260
},
{
"epoch": 1.5465686274509802,
"grad_norm": 0.2734375,
"learning_rate": 7.702386264080894e-06,
"loss": 0.0486,
"step": 1262
},
{
"epoch": 1.5490196078431373,
"grad_norm": 0.2265625,
"learning_rate": 7.69338347412931e-06,
"loss": 0.0438,
"step": 1264
},
{
"epoch": 1.5514705882352942,
"grad_norm": 0.279296875,
"learning_rate": 7.684368365781213e-06,
"loss": 0.0425,
"step": 1266
},
{
"epoch": 1.553921568627451,
"grad_norm": 0.294921875,
"learning_rate": 7.675340980267896e-06,
"loss": 0.0506,
"step": 1268
},
{
"epoch": 1.5563725490196079,
"grad_norm": 0.275390625,
"learning_rate": 7.666301358876798e-06,
"loss": 0.0495,
"step": 1270
},
{
"epoch": 1.5588235294117647,
"grad_norm": 0.3515625,
"learning_rate": 7.657249542951324e-06,
"loss": 0.048,
"step": 1272
},
{
"epoch": 1.5612745098039216,
"grad_norm": 0.318359375,
"learning_rate": 7.648185573890651e-06,
"loss": 0.0555,
"step": 1274
},
{
"epoch": 1.5637254901960784,
"grad_norm": 0.255859375,
"learning_rate": 7.639109493149537e-06,
"loss": 0.0495,
"step": 1276
},
{
"epoch": 1.5661764705882353,
"grad_norm": 0.265625,
"learning_rate": 7.630021342238139e-06,
"loss": 0.0475,
"step": 1278
},
{
"epoch": 1.5686274509803921,
"grad_norm": 0.201171875,
"learning_rate": 7.620921162721811e-06,
"loss": 0.0466,
"step": 1280
},
{
"epoch": 1.571078431372549,
"grad_norm": 0.2470703125,
"learning_rate": 7.611808996220923e-06,
"loss": 0.0502,
"step": 1282
},
{
"epoch": 1.5735294117647058,
"grad_norm": 0.228515625,
"learning_rate": 7.602684884410674e-06,
"loss": 0.0471,
"step": 1284
},
{
"epoch": 1.5759803921568627,
"grad_norm": 0.279296875,
"learning_rate": 7.593548869020887e-06,
"loss": 0.0589,
"step": 1286
},
{
"epoch": 1.5784313725490198,
"grad_norm": 0.23046875,
"learning_rate": 7.584400991835834e-06,
"loss": 0.0461,
"step": 1288
},
{
"epoch": 1.5808823529411766,
"grad_norm": 0.2138671875,
"learning_rate": 7.575241294694033e-06,
"loss": 0.0449,
"step": 1290
},
{
"epoch": 1.5833333333333335,
"grad_norm": 0.291015625,
"learning_rate": 7.566069819488066e-06,
"loss": 0.0461,
"step": 1292
},
{
"epoch": 1.5857843137254903,
"grad_norm": 0.287109375,
"learning_rate": 7.556886608164378e-06,
"loss": 0.0442,
"step": 1294
},
{
"epoch": 1.5882352941176472,
"grad_norm": 0.291015625,
"learning_rate": 7.547691702723095e-06,
"loss": 0.0499,
"step": 1296
},
{
"epoch": 1.590686274509804,
"grad_norm": 0.2353515625,
"learning_rate": 7.53848514521782e-06,
"loss": 0.0484,
"step": 1298
},
{
"epoch": 1.593137254901961,
"grad_norm": 0.2734375,
"learning_rate": 7.529266977755456e-06,
"loss": 0.0489,
"step": 1300
},
{
"epoch": 1.5955882352941178,
"grad_norm": 0.25390625,
"learning_rate": 7.520037242496e-06,
"loss": 0.0407,
"step": 1302
},
{
"epoch": 1.5980392156862746,
"grad_norm": 0.23046875,
"learning_rate": 7.510795981652356e-06,
"loss": 0.0511,
"step": 1304
},
{
"epoch": 1.6004901960784315,
"grad_norm": 0.236328125,
"learning_rate": 7.501543237490141e-06,
"loss": 0.0471,
"step": 1306
},
{
"epoch": 1.6029411764705883,
"grad_norm": 0.232421875,
"learning_rate": 7.4922790523274935e-06,
"loss": 0.0495,
"step": 1308
},
{
"epoch": 1.6053921568627452,
"grad_norm": 0.2421875,
"learning_rate": 7.483003468534874e-06,
"loss": 0.0457,
"step": 1310
},
{
"epoch": 1.607843137254902,
"grad_norm": 0.29296875,
"learning_rate": 7.473716528534883e-06,
"loss": 0.0532,
"step": 1312
},
{
"epoch": 1.6102941176470589,
"grad_norm": 0.185546875,
"learning_rate": 7.46441827480205e-06,
"loss": 0.0408,
"step": 1314
},
{
"epoch": 1.6127450980392157,
"grad_norm": 0.26171875,
"learning_rate": 7.455108749862658e-06,
"loss": 0.0447,
"step": 1316
},
{
"epoch": 1.6151960784313726,
"grad_norm": 0.31640625,
"learning_rate": 7.445787996294532e-06,
"loss": 0.0484,
"step": 1318
},
{
"epoch": 1.6176470588235294,
"grad_norm": 0.251953125,
"learning_rate": 7.43645605672686e-06,
"loss": 0.0446,
"step": 1320
},
{
"epoch": 1.6200980392156863,
"grad_norm": 0.22265625,
"learning_rate": 7.4271129738399805e-06,
"loss": 0.0424,
"step": 1322
},
{
"epoch": 1.6225490196078431,
"grad_norm": 0.291015625,
"learning_rate": 7.4177587903652035e-06,
"loss": 0.0482,
"step": 1324
},
{
"epoch": 1.625,
"grad_norm": 0.23046875,
"learning_rate": 7.4083935490846096e-06,
"loss": 0.0497,
"step": 1326
},
{
"epoch": 1.6274509803921569,
"grad_norm": 0.251953125,
"learning_rate": 7.399017292830848e-06,
"loss": 0.048,
"step": 1328
},
{
"epoch": 1.6299019607843137,
"grad_norm": 0.2265625,
"learning_rate": 7.389630064486946e-06,
"loss": 0.0504,
"step": 1330
},
{
"epoch": 1.6323529411764706,
"grad_norm": 0.267578125,
"learning_rate": 7.380231906986117e-06,
"loss": 0.0433,
"step": 1332
},
{
"epoch": 1.6348039215686274,
"grad_norm": 0.251953125,
"learning_rate": 7.370822863311558e-06,
"loss": 0.0457,
"step": 1334
},
{
"epoch": 1.6372549019607843,
"grad_norm": 0.328125,
"learning_rate": 7.361402976496251e-06,
"loss": 0.0469,
"step": 1336
},
{
"epoch": 1.6397058823529411,
"grad_norm": 0.271484375,
"learning_rate": 7.351972289622775e-06,
"loss": 0.0489,
"step": 1338
},
{
"epoch": 1.642156862745098,
"grad_norm": 0.2177734375,
"learning_rate": 7.342530845823103e-06,
"loss": 0.0441,
"step": 1340
},
{
"epoch": 1.6446078431372548,
"grad_norm": 0.2578125,
"learning_rate": 7.333078688278402e-06,
"loss": 0.052,
"step": 1342
},
{
"epoch": 1.6470588235294117,
"grad_norm": 0.255859375,
"learning_rate": 7.323615860218844e-06,
"loss": 0.0411,
"step": 1344
},
{
"epoch": 1.6495098039215685,
"grad_norm": 0.263671875,
"learning_rate": 7.314142404923399e-06,
"loss": 0.0522,
"step": 1346
},
{
"epoch": 1.6519607843137254,
"grad_norm": 0.25,
"learning_rate": 7.3046583657196445e-06,
"loss": 0.0477,
"step": 1348
},
{
"epoch": 1.6544117647058822,
"grad_norm": 0.25,
"learning_rate": 7.295163785983563e-06,
"loss": 0.0489,
"step": 1350
},
{
"epoch": 1.656862745098039,
"grad_norm": 0.2431640625,
"learning_rate": 7.2856587091393475e-06,
"loss": 0.0485,
"step": 1352
},
{
"epoch": 1.659313725490196,
"grad_norm": 0.2490234375,
"learning_rate": 7.276143178659195e-06,
"loss": 0.0426,
"step": 1354
},
{
"epoch": 1.6617647058823528,
"grad_norm": 0.224609375,
"learning_rate": 7.2666172380631184e-06,
"loss": 0.0461,
"step": 1356
},
{
"epoch": 1.6642156862745097,
"grad_norm": 0.294921875,
"learning_rate": 7.257080930918737e-06,
"loss": 0.0538,
"step": 1358
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.28515625,
"learning_rate": 7.2475343008410884e-06,
"loss": 0.0455,
"step": 1360
},
{
"epoch": 1.6691176470588234,
"grad_norm": 0.29296875,
"learning_rate": 7.237977391492419e-06,
"loss": 0.047,
"step": 1362
},
{
"epoch": 1.6715686274509802,
"grad_norm": 0.302734375,
"learning_rate": 7.228410246581987e-06,
"loss": 0.0446,
"step": 1364
},
{
"epoch": 1.6740196078431373,
"grad_norm": 0.248046875,
"learning_rate": 7.2188329098658685e-06,
"loss": 0.0431,
"step": 1366
},
{
"epoch": 1.6764705882352942,
"grad_norm": 0.2451171875,
"learning_rate": 7.2092454251467445e-06,
"loss": 0.0461,
"step": 1368
},
{
"epoch": 1.678921568627451,
"grad_norm": 0.2353515625,
"learning_rate": 7.199647836273721e-06,
"loss": 0.048,
"step": 1370
},
{
"epoch": 1.6813725490196079,
"grad_norm": 0.251953125,
"learning_rate": 7.190040187142103e-06,
"loss": 0.0535,
"step": 1372
},
{
"epoch": 1.6838235294117647,
"grad_norm": 0.2158203125,
"learning_rate": 7.180422521693217e-06,
"loss": 0.0424,
"step": 1374
},
{
"epoch": 1.6862745098039216,
"grad_norm": 0.263671875,
"learning_rate": 7.1707948839141935e-06,
"loss": 0.0458,
"step": 1376
},
{
"epoch": 1.6887254901960784,
"grad_norm": 0.2021484375,
"learning_rate": 7.161157317837775e-06,
"loss": 0.0425,
"step": 1378
},
{
"epoch": 1.6911764705882353,
"grad_norm": 0.287109375,
"learning_rate": 7.1515098675421125e-06,
"loss": 0.0454,
"step": 1380
},
{
"epoch": 1.6936274509803921,
"grad_norm": 0.2578125,
"learning_rate": 7.1418525771505605e-06,
"loss": 0.0455,
"step": 1382
},
{
"epoch": 1.696078431372549,
"grad_norm": 0.275390625,
"learning_rate": 7.13218549083148e-06,
"loss": 0.0472,
"step": 1384
},
{
"epoch": 1.6985294117647058,
"grad_norm": 0.255859375,
"learning_rate": 7.122508652798033e-06,
"loss": 0.0471,
"step": 1386
},
{
"epoch": 1.7009803921568627,
"grad_norm": 0.26171875,
"learning_rate": 7.112822107307986e-06,
"loss": 0.053,
"step": 1388
},
{
"epoch": 1.7034313725490198,
"grad_norm": 0.244140625,
"learning_rate": 7.1031258986634965e-06,
"loss": 0.0471,
"step": 1390
},
{
"epoch": 1.7058823529411766,
"grad_norm": 0.26171875,
"learning_rate": 7.093420071210922e-06,
"loss": 0.0484,
"step": 1392
},
{
"epoch": 1.7083333333333335,
"grad_norm": 0.345703125,
"learning_rate": 7.083704669340611e-06,
"loss": 0.0427,
"step": 1394
},
{
"epoch": 1.7107843137254903,
"grad_norm": 0.330078125,
"learning_rate": 7.0739797374867e-06,
"loss": 0.0481,
"step": 1396
},
{
"epoch": 1.7132352941176472,
"grad_norm": 0.30859375,
"learning_rate": 7.0642453201269145e-06,
"loss": 0.0447,
"step": 1398
},
{
"epoch": 1.715686274509804,
"grad_norm": 0.30078125,
"learning_rate": 7.054501461782361e-06,
"loss": 0.0496,
"step": 1400
},
{
"epoch": 1.718137254901961,
"grad_norm": 0.25,
"learning_rate": 7.044748207017327e-06,
"loss": 0.0495,
"step": 1402
},
{
"epoch": 1.7205882352941178,
"grad_norm": 0.2265625,
"learning_rate": 7.034985600439068e-06,
"loss": 0.0454,
"step": 1404
},
{
"epoch": 1.7230392156862746,
"grad_norm": 0.2421875,
"learning_rate": 7.0252136866976205e-06,
"loss": 0.0425,
"step": 1406
},
{
"epoch": 1.7254901960784315,
"grad_norm": 0.279296875,
"learning_rate": 7.015432510485584e-06,
"loss": 0.0465,
"step": 1408
},
{
"epoch": 1.7279411764705883,
"grad_norm": 0.2421875,
"learning_rate": 7.005642116537919e-06,
"loss": 0.0435,
"step": 1410
},
{
"epoch": 1.7303921568627452,
"grad_norm": 0.263671875,
"learning_rate": 6.9958425496317435e-06,
"loss": 0.047,
"step": 1412
},
{
"epoch": 1.732843137254902,
"grad_norm": 0.23046875,
"learning_rate": 6.986033854586134e-06,
"loss": 0.0468,
"step": 1414
},
{
"epoch": 1.7352941176470589,
"grad_norm": 0.279296875,
"learning_rate": 6.9762160762619116e-06,
"loss": 0.0502,
"step": 1416
},
{
"epoch": 1.7377450980392157,
"grad_norm": 0.326171875,
"learning_rate": 6.966389259561438e-06,
"loss": 0.0503,
"step": 1418
},
{
"epoch": 1.7401960784313726,
"grad_norm": 0.267578125,
"learning_rate": 6.956553449428418e-06,
"loss": 0.0445,
"step": 1420
},
{
"epoch": 1.7426470588235294,
"grad_norm": 0.2333984375,
"learning_rate": 6.946708690847684e-06,
"loss": 0.0468,
"step": 1422
},
{
"epoch": 1.7450980392156863,
"grad_norm": 0.232421875,
"learning_rate": 6.936855028844996e-06,
"loss": 0.0507,
"step": 1424
},
{
"epoch": 1.7475490196078431,
"grad_norm": 0.2421875,
"learning_rate": 6.926992508486839e-06,
"loss": 0.0485,
"step": 1426
},
{
"epoch": 1.75,
"grad_norm": 0.2353515625,
"learning_rate": 6.9171211748802055e-06,
"loss": 0.0434,
"step": 1428
},
{
"epoch": 1.7524509803921569,
"grad_norm": 0.2578125,
"learning_rate": 6.907241073172399e-06,
"loss": 0.0466,
"step": 1430
},
{
"epoch": 1.7549019607843137,
"grad_norm": 0.244140625,
"learning_rate": 6.897352248550828e-06,
"loss": 0.0483,
"step": 1432
},
{
"epoch": 1.7573529411764706,
"grad_norm": 0.34765625,
"learning_rate": 6.887454746242787e-06,
"loss": 0.0554,
"step": 1434
},
{
"epoch": 1.7598039215686274,
"grad_norm": 0.2119140625,
"learning_rate": 6.87754861151527e-06,
"loss": 0.0427,
"step": 1436
},
{
"epoch": 1.7622549019607843,
"grad_norm": 0.373046875,
"learning_rate": 6.867633889674743e-06,
"loss": 0.0433,
"step": 1438
},
{
"epoch": 1.7647058823529411,
"grad_norm": 0.23046875,
"learning_rate": 6.857710626066949e-06,
"loss": 0.0472,
"step": 1440
},
{
"epoch": 1.767156862745098,
"grad_norm": 0.21484375,
"learning_rate": 6.847778866076697e-06,
"loss": 0.0476,
"step": 1442
},
{
"epoch": 1.7696078431372548,
"grad_norm": 0.240234375,
"learning_rate": 6.837838655127659e-06,
"loss": 0.0434,
"step": 1444
},
{
"epoch": 1.7720588235294117,
"grad_norm": 0.25390625,
"learning_rate": 6.827890038682151e-06,
"loss": 0.0394,
"step": 1446
},
{
"epoch": 1.7745098039215685,
"grad_norm": 0.2353515625,
"learning_rate": 6.817933062240937e-06,
"loss": 0.0475,
"step": 1448
},
{
"epoch": 1.7769607843137254,
"grad_norm": 0.25,
"learning_rate": 6.8079677713430136e-06,
"loss": 0.052,
"step": 1450
},
{
"epoch": 1.7794117647058822,
"grad_norm": 0.318359375,
"learning_rate": 6.797994211565405e-06,
"loss": 0.053,
"step": 1452
},
{
"epoch": 1.781862745098039,
"grad_norm": 0.283203125,
"learning_rate": 6.788012428522956e-06,
"loss": 0.0425,
"step": 1454
},
{
"epoch": 1.784313725490196,
"grad_norm": 0.322265625,
"learning_rate": 6.7780224678681194e-06,
"loss": 0.0459,
"step": 1456
},
{
"epoch": 1.7867647058823528,
"grad_norm": 0.287109375,
"learning_rate": 6.768024375290747e-06,
"loss": 0.0468,
"step": 1458
},
{
"epoch": 1.7892156862745097,
"grad_norm": 0.2216796875,
"learning_rate": 6.758018196517885e-06,
"loss": 0.0462,
"step": 1460
},
{
"epoch": 1.7916666666666665,
"grad_norm": 0.23828125,
"learning_rate": 6.7480039773135605e-06,
"loss": 0.043,
"step": 1462
},
{
"epoch": 1.7941176470588234,
"grad_norm": 0.2490234375,
"learning_rate": 6.737981763478578e-06,
"loss": 0.0459,
"step": 1464
},
{
"epoch": 1.7965686274509802,
"grad_norm": 0.22265625,
"learning_rate": 6.727951600850302e-06,
"loss": 0.0403,
"step": 1466
},
{
"epoch": 1.7990196078431373,
"grad_norm": 0.263671875,
"learning_rate": 6.717913535302453e-06,
"loss": 0.0412,
"step": 1468
},
{
"epoch": 1.8014705882352942,
"grad_norm": 0.2431640625,
"learning_rate": 6.7078676127448965e-06,
"loss": 0.0482,
"step": 1470
},
{
"epoch": 1.803921568627451,
"grad_norm": 0.228515625,
"learning_rate": 6.697813879123429e-06,
"loss": 0.0502,
"step": 1472
},
{
"epoch": 1.8063725490196079,
"grad_norm": 0.287109375,
"learning_rate": 6.687752380419579e-06,
"loss": 0.0505,
"step": 1474
},
{
"epoch": 1.8088235294117647,
"grad_norm": 0.275390625,
"learning_rate": 6.677683162650383e-06,
"loss": 0.0459,
"step": 1476
},
{
"epoch": 1.8112745098039216,
"grad_norm": 0.23828125,
"learning_rate": 6.667606271868183e-06,
"loss": 0.0423,
"step": 1478
},
{
"epoch": 1.8137254901960784,
"grad_norm": 0.2421875,
"learning_rate": 6.657521754160415e-06,
"loss": 0.0494,
"step": 1480
},
{
"epoch": 1.8161764705882353,
"grad_norm": 0.40625,
"learning_rate": 6.6474296556493945e-06,
"loss": 0.0509,
"step": 1482
},
{
"epoch": 1.8186274509803921,
"grad_norm": 0.236328125,
"learning_rate": 6.637330022492112e-06,
"loss": 0.0441,
"step": 1484
},
{
"epoch": 1.821078431372549,
"grad_norm": 0.2197265625,
"learning_rate": 6.627222900880017e-06,
"loss": 0.0464,
"step": 1486
},
{
"epoch": 1.8235294117647058,
"grad_norm": 0.234375,
"learning_rate": 6.617108337038807e-06,
"loss": 0.0501,
"step": 1488
},
{
"epoch": 1.8259803921568627,
"grad_norm": 0.2255859375,
"learning_rate": 6.606986377228216e-06,
"loss": 0.0446,
"step": 1490
},
{
"epoch": 1.8284313725490198,
"grad_norm": 0.25390625,
"learning_rate": 6.59685706774181e-06,
"loss": 0.0437,
"step": 1492
},
{
"epoch": 1.8308823529411766,
"grad_norm": 0.302734375,
"learning_rate": 6.586720454906762e-06,
"loss": 0.0482,
"step": 1494
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.2373046875,
"learning_rate": 6.576576585083651e-06,
"loss": 0.0433,
"step": 1496
},
{
"epoch": 1.8357843137254903,
"grad_norm": 0.318359375,
"learning_rate": 6.566425504666249e-06,
"loss": 0.0439,
"step": 1498
},
{
"epoch": 1.8382352941176472,
"grad_norm": 0.267578125,
"learning_rate": 6.556267260081297e-06,
"loss": 0.0443,
"step": 1500
},
{
"epoch": 1.840686274509804,
"grad_norm": 0.25,
"learning_rate": 6.5461018977883135e-06,
"loss": 0.0464,
"step": 1502
},
{
"epoch": 1.843137254901961,
"grad_norm": 0.240234375,
"learning_rate": 6.535929464279364e-06,
"loss": 0.0442,
"step": 1504
},
{
"epoch": 1.8455882352941178,
"grad_norm": 0.1865234375,
"learning_rate": 6.525750006078854e-06,
"loss": 0.0436,
"step": 1506
},
{
"epoch": 1.8480392156862746,
"grad_norm": 0.322265625,
"learning_rate": 6.515563569743318e-06,
"loss": 0.0483,
"step": 1508
},
{
"epoch": 1.8504901960784315,
"grad_norm": 0.314453125,
"learning_rate": 6.50537020186121e-06,
"loss": 0.0428,
"step": 1510
},
{
"epoch": 1.8529411764705883,
"grad_norm": 0.2099609375,
"learning_rate": 6.495169949052679e-06,
"loss": 0.0403,
"step": 1512
},
{
"epoch": 1.8553921568627452,
"grad_norm": 0.271484375,
"learning_rate": 6.484962857969366e-06,
"loss": 0.0483,
"step": 1514
},
{
"epoch": 1.857843137254902,
"grad_norm": 0.271484375,
"learning_rate": 6.474748975294189e-06,
"loss": 0.0505,
"step": 1516
},
{
"epoch": 1.8602941176470589,
"grad_norm": 0.2412109375,
"learning_rate": 6.4645283477411235e-06,
"loss": 0.0454,
"step": 1518
},
{
"epoch": 1.8627450980392157,
"grad_norm": 0.212890625,
"learning_rate": 6.454301022054999e-06,
"loss": 0.0447,
"step": 1520
},
{
"epoch": 1.8651960784313726,
"grad_norm": 0.2373046875,
"learning_rate": 6.444067045011275e-06,
"loss": 0.0431,
"step": 1522
},
{
"epoch": 1.8676470588235294,
"grad_norm": 0.2255859375,
"learning_rate": 6.433826463415832e-06,
"loss": 0.0441,
"step": 1524
},
{
"epoch": 1.8700980392156863,
"grad_norm": 0.283203125,
"learning_rate": 6.423579324104758e-06,
"loss": 0.0449,
"step": 1526
},
{
"epoch": 1.8725490196078431,
"grad_norm": 0.265625,
"learning_rate": 6.413325673944133e-06,
"loss": 0.0447,
"step": 1528
},
{
"epoch": 1.875,
"grad_norm": 0.2490234375,
"learning_rate": 6.403065559829816e-06,
"loss": 0.0451,
"step": 1530
},
{
"epoch": 1.8774509803921569,
"grad_norm": 0.341796875,
"learning_rate": 6.392799028687227e-06,
"loss": 0.0458,
"step": 1532
},
{
"epoch": 1.8799019607843137,
"grad_norm": 0.2412109375,
"learning_rate": 6.382526127471136e-06,
"loss": 0.0521,
"step": 1534
},
{
"epoch": 1.8823529411764706,
"grad_norm": 0.23828125,
"learning_rate": 6.372246903165445e-06,
"loss": 0.0452,
"step": 1536
},
{
"epoch": 1.8848039215686274,
"grad_norm": 0.2578125,
"learning_rate": 6.361961402782981e-06,
"loss": 0.044,
"step": 1538
},
{
"epoch": 1.8872549019607843,
"grad_norm": 0.251953125,
"learning_rate": 6.35166967336527e-06,
"loss": 0.048,
"step": 1540
},
{
"epoch": 1.8897058823529411,
"grad_norm": 0.25390625,
"learning_rate": 6.3413717619823265e-06,
"loss": 0.0458,
"step": 1542
},
{
"epoch": 1.892156862745098,
"grad_norm": 0.2412109375,
"learning_rate": 6.331067715732443e-06,
"loss": 0.0446,
"step": 1544
},
{
"epoch": 1.8946078431372548,
"grad_norm": 0.2734375,
"learning_rate": 6.320757581741965e-06,
"loss": 0.0498,
"step": 1546
},
{
"epoch": 1.8970588235294117,
"grad_norm": 0.345703125,
"learning_rate": 6.310441407165085e-06,
"loss": 0.0429,
"step": 1548
},
{
"epoch": 1.8995098039215685,
"grad_norm": 0.27734375,
"learning_rate": 6.300119239183621e-06,
"loss": 0.0485,
"step": 1550
},
{
"epoch": 1.9019607843137254,
"grad_norm": 0.2333984375,
"learning_rate": 6.289791125006801e-06,
"loss": 0.0508,
"step": 1552
},
{
"epoch": 1.9044117647058822,
"grad_norm": 0.28125,
"learning_rate": 6.279457111871052e-06,
"loss": 0.0462,
"step": 1554
},
{
"epoch": 1.906862745098039,
"grad_norm": 0.2734375,
"learning_rate": 6.269117247039775e-06,
"loss": 0.0474,
"step": 1556
},
{
"epoch": 1.909313725490196,
"grad_norm": 0.2353515625,
"learning_rate": 6.258771577803141e-06,
"loss": 0.0479,
"step": 1558
},
{
"epoch": 1.9117647058823528,
"grad_norm": 0.29296875,
"learning_rate": 6.248420151477862e-06,
"loss": 0.0465,
"step": 1560
},
{
"epoch": 1.9142156862745097,
"grad_norm": 0.25390625,
"learning_rate": 6.238063015406982e-06,
"loss": 0.0503,
"step": 1562
},
{
"epoch": 1.9166666666666665,
"grad_norm": 0.30859375,
"learning_rate": 6.227700216959661e-06,
"loss": 0.0465,
"step": 1564
},
{
"epoch": 1.9191176470588234,
"grad_norm": 0.27734375,
"learning_rate": 6.2173318035309525e-06,
"loss": 0.0469,
"step": 1566
},
{
"epoch": 1.9215686274509802,
"grad_norm": 0.263671875,
"learning_rate": 6.206957822541596e-06,
"loss": 0.0418,
"step": 1568
},
{
"epoch": 1.9240196078431373,
"grad_norm": 0.2734375,
"learning_rate": 6.1965783214377895e-06,
"loss": 0.0524,
"step": 1570
},
{
"epoch": 1.9264705882352942,
"grad_norm": 0.30859375,
"learning_rate": 6.1861933476909805e-06,
"loss": 0.048,
"step": 1572
},
{
"epoch": 1.928921568627451,
"grad_norm": 0.30859375,
"learning_rate": 6.175802948797645e-06,
"loss": 0.0505,
"step": 1574
},
{
"epoch": 1.9313725490196079,
"grad_norm": 0.310546875,
"learning_rate": 6.165407172279073e-06,
"loss": 0.0501,
"step": 1576
},
{
"epoch": 1.9338235294117647,
"grad_norm": 0.25390625,
"learning_rate": 6.1550060656811464e-06,
"loss": 0.0513,
"step": 1578
},
{
"epoch": 1.9362745098039216,
"grad_norm": 0.2490234375,
"learning_rate": 6.144599676574128e-06,
"loss": 0.0424,
"step": 1580
},
{
"epoch": 1.9387254901960784,
"grad_norm": 0.294921875,
"learning_rate": 6.1341880525524365e-06,
"loss": 0.0455,
"step": 1582
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.21484375,
"learning_rate": 6.123771241234437e-06,
"loss": 0.0462,
"step": 1584
},
{
"epoch": 1.9436274509803921,
"grad_norm": 0.255859375,
"learning_rate": 6.113349290262216e-06,
"loss": 0.0476,
"step": 1586
},
{
"epoch": 1.946078431372549,
"grad_norm": 0.2578125,
"learning_rate": 6.1029222473013705e-06,
"loss": 0.0487,
"step": 1588
},
{
"epoch": 1.9485294117647058,
"grad_norm": 0.287109375,
"learning_rate": 6.0924901600407825e-06,
"loss": 0.0456,
"step": 1590
},
{
"epoch": 1.9509803921568627,
"grad_norm": 0.30078125,
"learning_rate": 6.082053076192405e-06,
"loss": 0.0454,
"step": 1592
},
{
"epoch": 1.9534313725490198,
"grad_norm": 0.263671875,
"learning_rate": 6.071611043491044e-06,
"loss": 0.0459,
"step": 1594
},
{
"epoch": 1.9558823529411766,
"grad_norm": 0.306640625,
"learning_rate": 6.061164109694139e-06,
"loss": 0.0554,
"step": 1596
},
{
"epoch": 1.9583333333333335,
"grad_norm": 0.2578125,
"learning_rate": 6.050712322581546e-06,
"loss": 0.0463,
"step": 1598
},
{
"epoch": 1.9607843137254903,
"grad_norm": 0.25390625,
"learning_rate": 6.040255729955318e-06,
"loss": 0.0478,
"step": 1600
},
{
"epoch": 1.9632352941176472,
"grad_norm": 0.244140625,
"learning_rate": 6.029794379639484e-06,
"loss": 0.0493,
"step": 1602
},
{
"epoch": 1.965686274509804,
"grad_norm": 0.259765625,
"learning_rate": 6.019328319479836e-06,
"loss": 0.0458,
"step": 1604
},
{
"epoch": 1.968137254901961,
"grad_norm": 0.32421875,
"learning_rate": 6.008857597343704e-06,
"loss": 0.0509,
"step": 1606
},
{
"epoch": 1.9705882352941178,
"grad_norm": 0.234375,
"learning_rate": 5.998382261119744e-06,
"loss": 0.0473,
"step": 1608
},
{
"epoch": 1.9730392156862746,
"grad_norm": 0.2412109375,
"learning_rate": 5.987902358717709e-06,
"loss": 0.0406,
"step": 1610
},
{
"epoch": 1.9754901960784315,
"grad_norm": 0.275390625,
"learning_rate": 5.97741793806824e-06,
"loss": 0.0508,
"step": 1612
},
{
"epoch": 1.9779411764705883,
"grad_norm": 0.2890625,
"learning_rate": 5.966929047122641e-06,
"loss": 0.0469,
"step": 1614
},
{
"epoch": 1.9803921568627452,
"grad_norm": 0.21875,
"learning_rate": 5.956435733852662e-06,
"loss": 0.0452,
"step": 1616
},
{
"epoch": 1.982843137254902,
"grad_norm": 0.3046875,
"learning_rate": 5.945938046250277e-06,
"loss": 0.0452,
"step": 1618
},
{
"epoch": 1.9852941176470589,
"grad_norm": 0.2421875,
"learning_rate": 5.935436032327471e-06,
"loss": 0.0404,
"step": 1620
},
{
"epoch": 1.9877450980392157,
"grad_norm": 0.220703125,
"learning_rate": 5.924929740116008e-06,
"loss": 0.045,
"step": 1622
},
{
"epoch": 1.9901960784313726,
"grad_norm": 0.3203125,
"learning_rate": 5.914419217667228e-06,
"loss": 0.0556,
"step": 1624
},
{
"epoch": 1.9926470588235294,
"grad_norm": 0.25390625,
"learning_rate": 5.903904513051812e-06,
"loss": 0.0429,
"step": 1626
},
{
"epoch": 1.9950980392156863,
"grad_norm": 0.294921875,
"learning_rate": 5.893385674359572e-06,
"loss": 0.0531,
"step": 1628
},
{
"epoch": 1.9975490196078431,
"grad_norm": 0.248046875,
"learning_rate": 5.882862749699223e-06,
"loss": 0.051,
"step": 1630
},
{
"epoch": 2.0,
"grad_norm": 0.408203125,
"learning_rate": 5.872335787198173e-06,
"loss": 0.045,
"step": 1632
},
{
"epoch": 2.0,
"eval_loss": 0.05398834869265556,
"eval_runtime": 37.4119,
"eval_samples_per_second": 134.075,
"eval_steps_per_second": 1.069,
"step": 1632
},
{
"epoch": 2.002450980392157,
"grad_norm": 0.2578125,
"learning_rate": 5.861804835002292e-06,
"loss": 0.0441,
"step": 1634
},
{
"epoch": 2.0049019607843137,
"grad_norm": 0.28125,
"learning_rate": 5.8512699412757045e-06,
"loss": 0.0421,
"step": 1636
},
{
"epoch": 2.0073529411764706,
"grad_norm": 0.27734375,
"learning_rate": 5.840731154200554e-06,
"loss": 0.0417,
"step": 1638
},
{
"epoch": 2.0098039215686274,
"grad_norm": 0.265625,
"learning_rate": 5.830188521976794e-06,
"loss": 0.0498,
"step": 1640
},
{
"epoch": 2.0122549019607843,
"grad_norm": 0.30859375,
"learning_rate": 5.819642092821964e-06,
"loss": 0.0463,
"step": 1642
},
{
"epoch": 2.014705882352941,
"grad_norm": 0.1953125,
"learning_rate": 5.809091914970971e-06,
"loss": 0.0412,
"step": 1644
},
{
"epoch": 2.017156862745098,
"grad_norm": 0.263671875,
"learning_rate": 5.798538036675863e-06,
"loss": 0.054,
"step": 1646
},
{
"epoch": 2.019607843137255,
"grad_norm": 0.2890625,
"learning_rate": 5.787980506205615e-06,
"loss": 0.0529,
"step": 1648
},
{
"epoch": 2.0220588235294117,
"grad_norm": 0.25,
"learning_rate": 5.777419371845905e-06,
"loss": 0.0492,
"step": 1650
},
{
"epoch": 2.0245098039215685,
"grad_norm": 0.2734375,
"learning_rate": 5.766854681898893e-06,
"loss": 0.049,
"step": 1652
},
{
"epoch": 2.0269607843137254,
"grad_norm": 0.302734375,
"learning_rate": 5.7562864846830016e-06,
"loss": 0.0455,
"step": 1654
},
{
"epoch": 2.0294117647058822,
"grad_norm": 0.2158203125,
"learning_rate": 5.745714828532691e-06,
"loss": 0.0518,
"step": 1656
},
{
"epoch": 2.031862745098039,
"grad_norm": 0.232421875,
"learning_rate": 5.73513976179825e-06,
"loss": 0.0477,
"step": 1658
},
{
"epoch": 2.034313725490196,
"grad_norm": 0.24609375,
"learning_rate": 5.724561332845553e-06,
"loss": 0.0445,
"step": 1660
},
{
"epoch": 2.036764705882353,
"grad_norm": 0.283203125,
"learning_rate": 5.713979590055864e-06,
"loss": 0.0483,
"step": 1662
},
{
"epoch": 2.0392156862745097,
"grad_norm": 0.2490234375,
"learning_rate": 5.703394581825595e-06,
"loss": 0.0468,
"step": 1664
},
{
"epoch": 2.0416666666666665,
"grad_norm": 0.25390625,
"learning_rate": 5.6928063565660955e-06,
"loss": 0.046,
"step": 1666
},
{
"epoch": 2.0441176470588234,
"grad_norm": 0.2099609375,
"learning_rate": 5.68221496270343e-06,
"loss": 0.0468,
"step": 1668
},
{
"epoch": 2.0465686274509802,
"grad_norm": 0.2353515625,
"learning_rate": 5.671620448678148e-06,
"loss": 0.0487,
"step": 1670
},
{
"epoch": 2.049019607843137,
"grad_norm": 0.294921875,
"learning_rate": 5.661022862945081e-06,
"loss": 0.0492,
"step": 1672
},
{
"epoch": 2.051470588235294,
"grad_norm": 0.328125,
"learning_rate": 5.650422253973099e-06,
"loss": 0.0446,
"step": 1674
},
{
"epoch": 2.053921568627451,
"grad_norm": 0.251953125,
"learning_rate": 5.639818670244904e-06,
"loss": 0.0416,
"step": 1676
},
{
"epoch": 2.0563725490196076,
"grad_norm": 0.35546875,
"learning_rate": 5.6292121602567995e-06,
"loss": 0.0519,
"step": 1678
},
{
"epoch": 2.0588235294117645,
"grad_norm": 0.271484375,
"learning_rate": 5.6186027725184775e-06,
"loss": 0.0453,
"step": 1680
},
{
"epoch": 2.0612745098039214,
"grad_norm": 0.2216796875,
"learning_rate": 5.607990555552788e-06,
"loss": 0.043,
"step": 1682
},
{
"epoch": 2.063725490196078,
"grad_norm": 0.28125,
"learning_rate": 5.597375557895522e-06,
"loss": 0.0492,
"step": 1684
},
{
"epoch": 2.0661764705882355,
"grad_norm": 0.236328125,
"learning_rate": 5.586757828095187e-06,
"loss": 0.0473,
"step": 1686
},
{
"epoch": 2.0686274509803924,
"grad_norm": 0.255859375,
"learning_rate": 5.576137414712788e-06,
"loss": 0.0479,
"step": 1688
},
{
"epoch": 2.071078431372549,
"grad_norm": 0.2451171875,
"learning_rate": 5.565514366321602e-06,
"loss": 0.0467,
"step": 1690
},
{
"epoch": 2.073529411764706,
"grad_norm": 0.220703125,
"learning_rate": 5.5548887315069575e-06,
"loss": 0.0491,
"step": 1692
},
{
"epoch": 2.075980392156863,
"grad_norm": 0.24609375,
"learning_rate": 5.544260558866015e-06,
"loss": 0.0503,
"step": 1694
},
{
"epoch": 2.0784313725490198,
"grad_norm": 0.25390625,
"learning_rate": 5.533629897007536e-06,
"loss": 0.0465,
"step": 1696
},
{
"epoch": 2.0808823529411766,
"grad_norm": 0.267578125,
"learning_rate": 5.522996794551673e-06,
"loss": 0.0551,
"step": 1698
},
{
"epoch": 2.0833333333333335,
"grad_norm": 0.2373046875,
"learning_rate": 5.512361300129738e-06,
"loss": 0.0409,
"step": 1700
},
{
"epoch": 2.0857843137254903,
"grad_norm": 0.228515625,
"learning_rate": 5.501723462383982e-06,
"loss": 0.0442,
"step": 1702
},
{
"epoch": 2.088235294117647,
"grad_norm": 0.279296875,
"learning_rate": 5.491083329967375e-06,
"loss": 0.0508,
"step": 1704
},
{
"epoch": 2.090686274509804,
"grad_norm": 0.228515625,
"learning_rate": 5.48044095154338e-06,
"loss": 0.0447,
"step": 1706
},
{
"epoch": 2.093137254901961,
"grad_norm": 0.2578125,
"learning_rate": 5.469796375785734e-06,
"loss": 0.0452,
"step": 1708
},
{
"epoch": 2.0955882352941178,
"grad_norm": 0.2431640625,
"learning_rate": 5.459149651378224e-06,
"loss": 0.0475,
"step": 1710
},
{
"epoch": 2.0980392156862746,
"grad_norm": 0.232421875,
"learning_rate": 5.4485008270144615e-06,
"loss": 0.0526,
"step": 1712
},
{
"epoch": 2.1004901960784315,
"grad_norm": 0.2890625,
"learning_rate": 5.437849951397665e-06,
"loss": 0.0454,
"step": 1714
},
{
"epoch": 2.1029411764705883,
"grad_norm": 0.259765625,
"learning_rate": 5.427197073240433e-06,
"loss": 0.0528,
"step": 1716
},
{
"epoch": 2.105392156862745,
"grad_norm": 0.279296875,
"learning_rate": 5.416542241264524e-06,
"loss": 0.0421,
"step": 1718
},
{
"epoch": 2.107843137254902,
"grad_norm": 0.30078125,
"learning_rate": 5.405885504200631e-06,
"loss": 0.0435,
"step": 1720
},
{
"epoch": 2.110294117647059,
"grad_norm": 0.2138671875,
"learning_rate": 5.395226910788162e-06,
"loss": 0.0495,
"step": 1722
},
{
"epoch": 2.1127450980392157,
"grad_norm": 0.21484375,
"learning_rate": 5.384566509775011e-06,
"loss": 0.0441,
"step": 1724
},
{
"epoch": 2.1151960784313726,
"grad_norm": 0.30859375,
"learning_rate": 5.373904349917343e-06,
"loss": 0.054,
"step": 1726
},
{
"epoch": 2.1176470588235294,
"grad_norm": 0.2109375,
"learning_rate": 5.3632404799793706e-06,
"loss": 0.049,
"step": 1728
},
{
"epoch": 2.1200980392156863,
"grad_norm": 0.318359375,
"learning_rate": 5.3525749487331184e-06,
"loss": 0.0453,
"step": 1730
},
{
"epoch": 2.122549019607843,
"grad_norm": 0.251953125,
"learning_rate": 5.341907804958215e-06,
"loss": 0.0444,
"step": 1732
},
{
"epoch": 2.125,
"grad_norm": 0.2431640625,
"learning_rate": 5.331239097441664e-06,
"loss": 0.0462,
"step": 1734
},
{
"epoch": 2.127450980392157,
"grad_norm": 0.23046875,
"learning_rate": 5.320568874977618e-06,
"loss": 0.0457,
"step": 1736
},
{
"epoch": 2.1299019607843137,
"grad_norm": 0.2490234375,
"learning_rate": 5.3098971863671646e-06,
"loss": 0.0482,
"step": 1738
},
{
"epoch": 2.1323529411764706,
"grad_norm": 0.2421875,
"learning_rate": 5.299224080418088e-06,
"loss": 0.0485,
"step": 1740
},
{
"epoch": 2.1348039215686274,
"grad_norm": 0.2333984375,
"learning_rate": 5.288549605944662e-06,
"loss": 0.0448,
"step": 1742
},
{
"epoch": 2.1372549019607843,
"grad_norm": 0.310546875,
"learning_rate": 5.277873811767415e-06,
"loss": 0.0524,
"step": 1744
},
{
"epoch": 2.139705882352941,
"grad_norm": 0.2353515625,
"learning_rate": 5.267196746712914e-06,
"loss": 0.0443,
"step": 1746
},
{
"epoch": 2.142156862745098,
"grad_norm": 0.265625,
"learning_rate": 5.256518459613538e-06,
"loss": 0.056,
"step": 1748
},
{
"epoch": 2.144607843137255,
"grad_norm": 0.212890625,
"learning_rate": 5.245838999307253e-06,
"loss": 0.0417,
"step": 1750
},
{
"epoch": 2.1470588235294117,
"grad_norm": 0.216796875,
"learning_rate": 5.2351584146373936e-06,
"loss": 0.0425,
"step": 1752
},
{
"epoch": 2.1495098039215685,
"grad_norm": 0.2890625,
"learning_rate": 5.224476754452433e-06,
"loss": 0.0456,
"step": 1754
},
{
"epoch": 2.1519607843137254,
"grad_norm": 0.259765625,
"learning_rate": 5.21379406760577e-06,
"loss": 0.0509,
"step": 1756
},
{
"epoch": 2.1544117647058822,
"grad_norm": 0.2373046875,
"learning_rate": 5.203110402955491e-06,
"loss": 0.0522,
"step": 1758
},
{
"epoch": 2.156862745098039,
"grad_norm": 0.259765625,
"learning_rate": 5.19242580936416e-06,
"loss": 0.0412,
"step": 1760
},
{
"epoch": 2.159313725490196,
"grad_norm": 0.208984375,
"learning_rate": 5.181740335698586e-06,
"loss": 0.046,
"step": 1762
},
{
"epoch": 2.161764705882353,
"grad_norm": 0.310546875,
"learning_rate": 5.171054030829605e-06,
"loss": 0.0488,
"step": 1764
},
{
"epoch": 2.1642156862745097,
"grad_norm": 0.2412109375,
"learning_rate": 5.1603669436318584e-06,
"loss": 0.0445,
"step": 1766
},
{
"epoch": 2.1666666666666665,
"grad_norm": 0.25,
"learning_rate": 5.1496791229835555e-06,
"loss": 0.0461,
"step": 1768
},
{
"epoch": 2.1691176470588234,
"grad_norm": 0.244140625,
"learning_rate": 5.1389906177662705e-06,
"loss": 0.0457,
"step": 1770
},
{
"epoch": 2.1715686274509802,
"grad_norm": 0.24609375,
"learning_rate": 5.128301476864703e-06,
"loss": 0.0505,
"step": 1772
},
{
"epoch": 2.174019607843137,
"grad_norm": 0.236328125,
"learning_rate": 5.1176117491664615e-06,
"loss": 0.0464,
"step": 1774
},
{
"epoch": 2.176470588235294,
"grad_norm": 0.248046875,
"learning_rate": 5.106921483561837e-06,
"loss": 0.0447,
"step": 1776
},
{
"epoch": 2.178921568627451,
"grad_norm": 0.2578125,
"learning_rate": 5.096230728943583e-06,
"loss": 0.0488,
"step": 1778
},
{
"epoch": 2.1813725490196076,
"grad_norm": 0.25390625,
"learning_rate": 5.085539534206687e-06,
"loss": 0.0458,
"step": 1780
},
{
"epoch": 2.1838235294117645,
"grad_norm": 0.251953125,
"learning_rate": 5.07484794824815e-06,
"loss": 0.0502,
"step": 1782
},
{
"epoch": 2.186274509803922,
"grad_norm": 0.208984375,
"learning_rate": 5.064156019966765e-06,
"loss": 0.043,
"step": 1784
},
{
"epoch": 2.188725490196078,
"grad_norm": 0.232421875,
"learning_rate": 5.053463798262888e-06,
"loss": 0.0401,
"step": 1786
},
{
"epoch": 2.1911764705882355,
"grad_norm": 0.27734375,
"learning_rate": 5.0427713320382164e-06,
"loss": 0.0451,
"step": 1788
},
{
"epoch": 2.1936274509803924,
"grad_norm": 0.2314453125,
"learning_rate": 5.032078670195567e-06,
"loss": 0.0423,
"step": 1790
},
{
"epoch": 2.196078431372549,
"grad_norm": 0.255859375,
"learning_rate": 5.021385861638651e-06,
"loss": 0.0457,
"step": 1792
},
{
"epoch": 2.198529411764706,
"grad_norm": 0.2216796875,
"learning_rate": 5.010692955271853e-06,
"loss": 0.0504,
"step": 1794
},
{
"epoch": 2.200980392156863,
"grad_norm": 0.236328125,
"learning_rate": 5e-06,
"loss": 0.0438,
"step": 1796
},
{
"epoch": 2.2034313725490198,
"grad_norm": 0.2060546875,
"learning_rate": 4.989307044728148e-06,
"loss": 0.0419,
"step": 1798
},
{
"epoch": 2.2058823529411766,
"grad_norm": 0.20703125,
"learning_rate": 4.9786141383613505e-06,
"loss": 0.0493,
"step": 1800
},
{
"epoch": 2.2083333333333335,
"grad_norm": 0.2109375,
"learning_rate": 4.967921329804434e-06,
"loss": 0.0446,
"step": 1802
},
{
"epoch": 2.2107843137254903,
"grad_norm": 0.203125,
"learning_rate": 4.957228667961786e-06,
"loss": 0.0452,
"step": 1804
},
{
"epoch": 2.213235294117647,
"grad_norm": 0.2412109375,
"learning_rate": 4.946536201737114e-06,
"loss": 0.0446,
"step": 1806
},
{
"epoch": 2.215686274509804,
"grad_norm": 0.283203125,
"learning_rate": 4.935843980033235e-06,
"loss": 0.0496,
"step": 1808
},
{
"epoch": 2.218137254901961,
"grad_norm": 0.291015625,
"learning_rate": 4.9251520517518515e-06,
"loss": 0.0483,
"step": 1810
},
{
"epoch": 2.2205882352941178,
"grad_norm": 0.271484375,
"learning_rate": 4.914460465793314e-06,
"loss": 0.0463,
"step": 1812
},
{
"epoch": 2.2230392156862746,
"grad_norm": 0.236328125,
"learning_rate": 4.903769271056419e-06,
"loss": 0.0501,
"step": 1814
},
{
"epoch": 2.2254901960784315,
"grad_norm": 0.25,
"learning_rate": 4.893078516438164e-06,
"loss": 0.0488,
"step": 1816
},
{
"epoch": 2.2279411764705883,
"grad_norm": 0.2119140625,
"learning_rate": 4.882388250833539e-06,
"loss": 0.046,
"step": 1818
},
{
"epoch": 2.230392156862745,
"grad_norm": 0.271484375,
"learning_rate": 4.871698523135298e-06,
"loss": 0.0485,
"step": 1820
},
{
"epoch": 2.232843137254902,
"grad_norm": 0.25390625,
"learning_rate": 4.86100938223373e-06,
"loss": 0.046,
"step": 1822
},
{
"epoch": 2.235294117647059,
"grad_norm": 0.2734375,
"learning_rate": 4.850320877016446e-06,
"loss": 0.044,
"step": 1824
},
{
"epoch": 2.2377450980392157,
"grad_norm": 0.2294921875,
"learning_rate": 4.839633056368144e-06,
"loss": 0.0439,
"step": 1826
},
{
"epoch": 2.2401960784313726,
"grad_norm": 0.26953125,
"learning_rate": 4.828945969170396e-06,
"loss": 0.0481,
"step": 1828
},
{
"epoch": 2.2426470588235294,
"grad_norm": 0.2451171875,
"learning_rate": 4.818259664301416e-06,
"loss": 0.0412,
"step": 1830
},
{
"epoch": 2.2450980392156863,
"grad_norm": 0.2373046875,
"learning_rate": 4.807574190635842e-06,
"loss": 0.0474,
"step": 1832
},
{
"epoch": 2.247549019607843,
"grad_norm": 0.296875,
"learning_rate": 4.796889597044511e-06,
"loss": 0.0465,
"step": 1834
},
{
"epoch": 2.25,
"grad_norm": 0.2431640625,
"learning_rate": 4.786205932394232e-06,
"loss": 0.0461,
"step": 1836
},
{
"epoch": 2.252450980392157,
"grad_norm": 0.263671875,
"learning_rate": 4.775523245547567e-06,
"loss": 0.0475,
"step": 1838
},
{
"epoch": 2.2549019607843137,
"grad_norm": 0.255859375,
"learning_rate": 4.764841585362609e-06,
"loss": 0.0475,
"step": 1840
},
{
"epoch": 2.2573529411764706,
"grad_norm": 0.259765625,
"learning_rate": 4.754161000692748e-06,
"loss": 0.0479,
"step": 1842
},
{
"epoch": 2.2598039215686274,
"grad_norm": 0.265625,
"learning_rate": 4.743481540386463e-06,
"loss": 0.0462,
"step": 1844
},
{
"epoch": 2.2622549019607843,
"grad_norm": 0.25390625,
"learning_rate": 4.732803253287087e-06,
"loss": 0.0473,
"step": 1846
},
{
"epoch": 2.264705882352941,
"grad_norm": 0.279296875,
"learning_rate": 4.722126188232586e-06,
"loss": 0.0462,
"step": 1848
},
{
"epoch": 2.267156862745098,
"grad_norm": 0.2578125,
"learning_rate": 4.711450394055341e-06,
"loss": 0.0434,
"step": 1850
},
{
"epoch": 2.269607843137255,
"grad_norm": 0.2333984375,
"learning_rate": 4.700775919581914e-06,
"loss": 0.0471,
"step": 1852
},
{
"epoch": 2.2720588235294117,
"grad_norm": 0.263671875,
"learning_rate": 4.690102813632836e-06,
"loss": 0.0486,
"step": 1854
},
{
"epoch": 2.2745098039215685,
"grad_norm": 0.26171875,
"learning_rate": 4.679431125022383e-06,
"loss": 0.0419,
"step": 1856
},
{
"epoch": 2.2769607843137254,
"grad_norm": 0.2109375,
"learning_rate": 4.668760902558338e-06,
"loss": 0.0455,
"step": 1858
},
{
"epoch": 2.2794117647058822,
"grad_norm": 0.24609375,
"learning_rate": 4.658092195041788e-06,
"loss": 0.0479,
"step": 1860
},
{
"epoch": 2.281862745098039,
"grad_norm": 0.2197265625,
"learning_rate": 4.647425051266884e-06,
"loss": 0.0514,
"step": 1862
},
{
"epoch": 2.284313725490196,
"grad_norm": 0.240234375,
"learning_rate": 4.63675952002063e-06,
"loss": 0.0463,
"step": 1864
},
{
"epoch": 2.286764705882353,
"grad_norm": 0.2431640625,
"learning_rate": 4.6260956500826574e-06,
"loss": 0.0475,
"step": 1866
},
{
"epoch": 2.2892156862745097,
"grad_norm": 0.2060546875,
"learning_rate": 4.615433490224991e-06,
"loss": 0.0444,
"step": 1868
},
{
"epoch": 2.2916666666666665,
"grad_norm": 0.2333984375,
"learning_rate": 4.604773089211842e-06,
"loss": 0.0423,
"step": 1870
},
{
"epoch": 2.2941176470588234,
"grad_norm": 0.2333984375,
"learning_rate": 4.594114495799371e-06,
"loss": 0.0417,
"step": 1872
},
{
"epoch": 2.2965686274509802,
"grad_norm": 0.208984375,
"learning_rate": 4.583457758735477e-06,
"loss": 0.0444,
"step": 1874
},
{
"epoch": 2.299019607843137,
"grad_norm": 0.28515625,
"learning_rate": 4.572802926759568e-06,
"loss": 0.0492,
"step": 1876
},
{
"epoch": 2.301470588235294,
"grad_norm": 0.2392578125,
"learning_rate": 4.562150048602336e-06,
"loss": 0.0442,
"step": 1878
},
{
"epoch": 2.303921568627451,
"grad_norm": 0.26171875,
"learning_rate": 4.55149917298554e-06,
"loss": 0.049,
"step": 1880
},
{
"epoch": 2.306372549019608,
"grad_norm": 0.2314453125,
"learning_rate": 4.540850348621777e-06,
"loss": 0.0458,
"step": 1882
},
{
"epoch": 2.3088235294117645,
"grad_norm": 0.267578125,
"learning_rate": 4.530203624214267e-06,
"loss": 0.0481,
"step": 1884
},
{
"epoch": 2.311274509803922,
"grad_norm": 0.259765625,
"learning_rate": 4.519559048456622e-06,
"loss": 0.0453,
"step": 1886
},
{
"epoch": 2.313725490196078,
"grad_norm": 0.1982421875,
"learning_rate": 4.508916670032627e-06,
"loss": 0.0429,
"step": 1888
},
{
"epoch": 2.3161764705882355,
"grad_norm": 0.2421875,
"learning_rate": 4.4982765376160205e-06,
"loss": 0.0461,
"step": 1890
},
{
"epoch": 2.318627450980392,
"grad_norm": 0.234375,
"learning_rate": 4.487638699870264e-06,
"loss": 0.0467,
"step": 1892
},
{
"epoch": 2.321078431372549,
"grad_norm": 0.26171875,
"learning_rate": 4.477003205448328e-06,
"loss": 0.0412,
"step": 1894
},
{
"epoch": 2.323529411764706,
"grad_norm": 0.251953125,
"learning_rate": 4.466370102992466e-06,
"loss": 0.0479,
"step": 1896
},
{
"epoch": 2.325980392156863,
"grad_norm": 0.2216796875,
"learning_rate": 4.455739441133987e-06,
"loss": 0.0444,
"step": 1898
},
{
"epoch": 2.3284313725490198,
"grad_norm": 0.2060546875,
"learning_rate": 4.4451112684930424e-06,
"loss": 0.0399,
"step": 1900
},
{
"epoch": 2.3308823529411766,
"grad_norm": 0.255859375,
"learning_rate": 4.4344856336784e-06,
"loss": 0.0445,
"step": 1902
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.25390625,
"learning_rate": 4.423862585287213e-06,
"loss": 0.049,
"step": 1904
},
{
"epoch": 2.3357843137254903,
"grad_norm": 0.2119140625,
"learning_rate": 4.413242171904815e-06,
"loss": 0.0461,
"step": 1906
},
{
"epoch": 2.338235294117647,
"grad_norm": 0.2734375,
"learning_rate": 4.402624442104479e-06,
"loss": 0.0466,
"step": 1908
},
{
"epoch": 2.340686274509804,
"grad_norm": 0.216796875,
"learning_rate": 4.392009444447213e-06,
"loss": 0.0535,
"step": 1910
},
{
"epoch": 2.343137254901961,
"grad_norm": 0.2119140625,
"learning_rate": 4.381397227481523e-06,
"loss": 0.0433,
"step": 1912
},
{
"epoch": 2.3455882352941178,
"grad_norm": 0.25390625,
"learning_rate": 4.3707878397432004e-06,
"loss": 0.0469,
"step": 1914
},
{
"epoch": 2.3480392156862746,
"grad_norm": 0.232421875,
"learning_rate": 4.360181329755099e-06,
"loss": 0.0427,
"step": 1916
},
{
"epoch": 2.3504901960784315,
"grad_norm": 0.271484375,
"learning_rate": 4.349577746026902e-06,
"loss": 0.047,
"step": 1918
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.224609375,
"learning_rate": 4.33897713705492e-06,
"loss": 0.0433,
"step": 1920
},
{
"epoch": 2.355392156862745,
"grad_norm": 0.306640625,
"learning_rate": 4.328379551321853e-06,
"loss": 0.051,
"step": 1922
},
{
"epoch": 2.357843137254902,
"grad_norm": 0.236328125,
"learning_rate": 4.317785037296573e-06,
"loss": 0.0435,
"step": 1924
},
{
"epoch": 2.360294117647059,
"grad_norm": 0.228515625,
"learning_rate": 4.307193643433907e-06,
"loss": 0.0487,
"step": 1926
},
{
"epoch": 2.3627450980392157,
"grad_norm": 0.271484375,
"learning_rate": 4.296605418174406e-06,
"loss": 0.0503,
"step": 1928
},
{
"epoch": 2.3651960784313726,
"grad_norm": 0.22265625,
"learning_rate": 4.286020409944137e-06,
"loss": 0.0436,
"step": 1930
},
{
"epoch": 2.3676470588235294,
"grad_norm": 0.212890625,
"learning_rate": 4.2754386671544475e-06,
"loss": 0.0443,
"step": 1932
},
{
"epoch": 2.3700980392156863,
"grad_norm": 0.255859375,
"learning_rate": 4.264860238201752e-06,
"loss": 0.0509,
"step": 1934
},
{
"epoch": 2.372549019607843,
"grad_norm": 0.2470703125,
"learning_rate": 4.2542851714673095e-06,
"loss": 0.0458,
"step": 1936
},
{
"epoch": 2.375,
"grad_norm": 0.30078125,
"learning_rate": 4.243713515317001e-06,
"loss": 0.0419,
"step": 1938
},
{
"epoch": 2.377450980392157,
"grad_norm": 0.271484375,
"learning_rate": 4.233145318101109e-06,
"loss": 0.0483,
"step": 1940
},
{
"epoch": 2.3799019607843137,
"grad_norm": 0.2275390625,
"learning_rate": 4.222580628154097e-06,
"loss": 0.0479,
"step": 1942
},
{
"epoch": 2.3823529411764706,
"grad_norm": 0.25390625,
"learning_rate": 4.2120194937943856e-06,
"loss": 0.0443,
"step": 1944
},
{
"epoch": 2.3848039215686274,
"grad_norm": 0.2216796875,
"learning_rate": 4.20146196332414e-06,
"loss": 0.0468,
"step": 1946
},
{
"epoch": 2.3872549019607843,
"grad_norm": 0.25390625,
"learning_rate": 4.190908085029031e-06,
"loss": 0.0457,
"step": 1948
},
{
"epoch": 2.389705882352941,
"grad_norm": 0.232421875,
"learning_rate": 4.180357907178037e-06,
"loss": 0.0468,
"step": 1950
},
{
"epoch": 2.392156862745098,
"grad_norm": 0.220703125,
"learning_rate": 4.1698114780232085e-06,
"loss": 0.0438,
"step": 1952
},
{
"epoch": 2.394607843137255,
"grad_norm": 0.3125,
"learning_rate": 4.159268845799448e-06,
"loss": 0.0496,
"step": 1954
},
{
"epoch": 2.3970588235294117,
"grad_norm": 0.259765625,
"learning_rate": 4.148730058724298e-06,
"loss": 0.0451,
"step": 1956
},
{
"epoch": 2.3995098039215685,
"grad_norm": 0.2392578125,
"learning_rate": 4.1381951649977085e-06,
"loss": 0.0465,
"step": 1958
},
{
"epoch": 2.4019607843137254,
"grad_norm": 0.2021484375,
"learning_rate": 4.127664212801828e-06,
"loss": 0.0416,
"step": 1960
},
{
"epoch": 2.4044117647058822,
"grad_norm": 0.23828125,
"learning_rate": 4.117137250300779e-06,
"loss": 0.0453,
"step": 1962
},
{
"epoch": 2.406862745098039,
"grad_norm": 0.251953125,
"learning_rate": 4.10661432564043e-06,
"loss": 0.0491,
"step": 1964
},
{
"epoch": 2.409313725490196,
"grad_norm": 0.263671875,
"learning_rate": 4.096095486948189e-06,
"loss": 0.0466,
"step": 1966
},
{
"epoch": 2.411764705882353,
"grad_norm": 0.23046875,
"learning_rate": 4.085580782332773e-06,
"loss": 0.0414,
"step": 1968
},
{
"epoch": 2.4142156862745097,
"grad_norm": 0.2060546875,
"learning_rate": 4.075070259883992e-06,
"loss": 0.042,
"step": 1970
},
{
"epoch": 2.4166666666666665,
"grad_norm": 0.240234375,
"learning_rate": 4.064563967672533e-06,
"loss": 0.0517,
"step": 1972
},
{
"epoch": 2.4191176470588234,
"grad_norm": 0.2392578125,
"learning_rate": 4.054061953749724e-06,
"loss": 0.0467,
"step": 1974
},
{
"epoch": 2.4215686274509802,
"grad_norm": 0.26953125,
"learning_rate": 4.043564266147339e-06,
"loss": 0.0477,
"step": 1976
},
{
"epoch": 2.424019607843137,
"grad_norm": 0.279296875,
"learning_rate": 4.033070952877362e-06,
"loss": 0.0455,
"step": 1978
},
{
"epoch": 2.426470588235294,
"grad_norm": 0.1982421875,
"learning_rate": 4.022582061931761e-06,
"loss": 0.0471,
"step": 1980
},
{
"epoch": 2.428921568627451,
"grad_norm": 0.2578125,
"learning_rate": 4.0120976412822935e-06,
"loss": 0.0481,
"step": 1982
},
{
"epoch": 2.431372549019608,
"grad_norm": 0.298828125,
"learning_rate": 4.0016177388802575e-06,
"loss": 0.0481,
"step": 1984
},
{
"epoch": 2.4338235294117645,
"grad_norm": 0.2392578125,
"learning_rate": 3.991142402656296e-06,
"loss": 0.0499,
"step": 1986
},
{
"epoch": 2.436274509803922,
"grad_norm": 0.296875,
"learning_rate": 3.980671680520166e-06,
"loss": 0.0461,
"step": 1988
},
{
"epoch": 2.438725490196078,
"grad_norm": 0.287109375,
"learning_rate": 3.970205620360517e-06,
"loss": 0.0425,
"step": 1990
},
{
"epoch": 2.4411764705882355,
"grad_norm": 0.349609375,
"learning_rate": 3.959744270044684e-06,
"loss": 0.0468,
"step": 1992
},
{
"epoch": 2.443627450980392,
"grad_norm": 0.2255859375,
"learning_rate": 3.949287677418456e-06,
"loss": 0.0442,
"step": 1994
},
{
"epoch": 2.446078431372549,
"grad_norm": 0.29296875,
"learning_rate": 3.938835890305862e-06,
"loss": 0.0406,
"step": 1996
},
{
"epoch": 2.448529411764706,
"grad_norm": 0.291015625,
"learning_rate": 3.928388956508958e-06,
"loss": 0.0442,
"step": 1998
},
{
"epoch": 2.450980392156863,
"grad_norm": 0.2119140625,
"learning_rate": 3.917946923807596e-06,
"loss": 0.0482,
"step": 2000
},
{
"epoch": 2.4534313725490198,
"grad_norm": 0.380859375,
"learning_rate": 3.90750983995922e-06,
"loss": 0.0432,
"step": 2002
},
{
"epoch": 2.4558823529411766,
"grad_norm": 0.26171875,
"learning_rate": 3.89707775269863e-06,
"loss": 0.0492,
"step": 2004
},
{
"epoch": 2.4583333333333335,
"grad_norm": 0.2578125,
"learning_rate": 3.886650709737784e-06,
"loss": 0.0465,
"step": 2006
},
{
"epoch": 2.4607843137254903,
"grad_norm": 0.2392578125,
"learning_rate": 3.876228758765566e-06,
"loss": 0.0451,
"step": 2008
},
{
"epoch": 2.463235294117647,
"grad_norm": 0.287109375,
"learning_rate": 3.865811947447565e-06,
"loss": 0.0451,
"step": 2010
},
{
"epoch": 2.465686274509804,
"grad_norm": 0.19140625,
"learning_rate": 3.8554003234258754e-06,
"loss": 0.0447,
"step": 2012
},
{
"epoch": 2.468137254901961,
"grad_norm": 0.29296875,
"learning_rate": 3.844993934318854e-06,
"loss": 0.0492,
"step": 2014
},
{
"epoch": 2.4705882352941178,
"grad_norm": 0.3203125,
"learning_rate": 3.8345928277209275e-06,
"loss": 0.0531,
"step": 2016
},
{
"epoch": 2.4730392156862746,
"grad_norm": 0.2314453125,
"learning_rate": 3.824197051202356e-06,
"loss": 0.0472,
"step": 2018
},
{
"epoch": 2.4754901960784315,
"grad_norm": 0.263671875,
"learning_rate": 3.81380665230902e-06,
"loss": 0.0461,
"step": 2020
},
{
"epoch": 2.4779411764705883,
"grad_norm": 0.2353515625,
"learning_rate": 3.803421678562213e-06,
"loss": 0.0479,
"step": 2022
},
{
"epoch": 2.480392156862745,
"grad_norm": 0.294921875,
"learning_rate": 3.7930421774584057e-06,
"loss": 0.0467,
"step": 2024
},
{
"epoch": 2.482843137254902,
"grad_norm": 0.306640625,
"learning_rate": 3.7826681964690484e-06,
"loss": 0.0409,
"step": 2026
},
{
"epoch": 2.485294117647059,
"grad_norm": 0.228515625,
"learning_rate": 3.7722997830403417e-06,
"loss": 0.0421,
"step": 2028
},
{
"epoch": 2.4877450980392157,
"grad_norm": 0.21875,
"learning_rate": 3.7619369845930195e-06,
"loss": 0.0364,
"step": 2030
},
{
"epoch": 2.4901960784313726,
"grad_norm": 0.294921875,
"learning_rate": 3.7515798485221407e-06,
"loss": 0.0444,
"step": 2032
},
{
"epoch": 2.4926470588235294,
"grad_norm": 0.216796875,
"learning_rate": 3.7412284221968596e-06,
"loss": 0.0432,
"step": 2034
},
{
"epoch": 2.4950980392156863,
"grad_norm": 0.2314453125,
"learning_rate": 3.7308827529602243e-06,
"loss": 0.0438,
"step": 2036
},
{
"epoch": 2.497549019607843,
"grad_norm": 0.30078125,
"learning_rate": 3.72054288812895e-06,
"loss": 0.0482,
"step": 2038
},
{
"epoch": 2.5,
"grad_norm": 0.26953125,
"learning_rate": 3.7102088749931998e-06,
"loss": 0.0454,
"step": 2040
},
{
"epoch": 2.502450980392157,
"grad_norm": 0.2890625,
"learning_rate": 3.69988076081638e-06,
"loss": 0.0473,
"step": 2042
},
{
"epoch": 2.5049019607843137,
"grad_norm": 0.244140625,
"learning_rate": 3.6895585928349165e-06,
"loss": 0.0471,
"step": 2044
},
{
"epoch": 2.5073529411764706,
"grad_norm": 0.236328125,
"learning_rate": 3.679242418258036e-06,
"loss": 0.047,
"step": 2046
},
{
"epoch": 2.5098039215686274,
"grad_norm": 0.263671875,
"learning_rate": 3.668932284267559e-06,
"loss": 0.0392,
"step": 2048
},
{
"epoch": 2.5122549019607843,
"grad_norm": 0.2314453125,
"learning_rate": 3.6586282380176748e-06,
"loss": 0.0453,
"step": 2050
},
{
"epoch": 2.514705882352941,
"grad_norm": 0.2197265625,
"learning_rate": 3.648330326634731e-06,
"loss": 0.0435,
"step": 2052
},
{
"epoch": 2.517156862745098,
"grad_norm": 0.2265625,
"learning_rate": 3.6380385972170195e-06,
"loss": 0.0492,
"step": 2054
},
{
"epoch": 2.519607843137255,
"grad_norm": 0.244140625,
"learning_rate": 3.6277530968345552e-06,
"loss": 0.0465,
"step": 2056
},
{
"epoch": 2.5220588235294117,
"grad_norm": 0.2890625,
"learning_rate": 3.617473872528867e-06,
"loss": 0.0452,
"step": 2058
},
{
"epoch": 2.5245098039215685,
"grad_norm": 0.244140625,
"learning_rate": 3.6072009713127743e-06,
"loss": 0.0424,
"step": 2060
},
{
"epoch": 2.5269607843137254,
"grad_norm": 0.220703125,
"learning_rate": 3.5969344401701844e-06,
"loss": 0.0437,
"step": 2062
},
{
"epoch": 2.5294117647058822,
"grad_norm": 0.2734375,
"learning_rate": 3.586674326055868e-06,
"loss": 0.0511,
"step": 2064
},
{
"epoch": 2.531862745098039,
"grad_norm": 0.208984375,
"learning_rate": 3.5764206758952426e-06,
"loss": 0.0437,
"step": 2066
},
{
"epoch": 2.534313725490196,
"grad_norm": 0.2177734375,
"learning_rate": 3.56617353658417e-06,
"loss": 0.044,
"step": 2068
},
{
"epoch": 2.536764705882353,
"grad_norm": 0.2158203125,
"learning_rate": 3.5559329549887268e-06,
"loss": 0.0447,
"step": 2070
},
{
"epoch": 2.5392156862745097,
"grad_norm": 0.251953125,
"learning_rate": 3.5456989779450014e-06,
"loss": 0.0493,
"step": 2072
},
{
"epoch": 2.5416666666666665,
"grad_norm": 0.322265625,
"learning_rate": 3.5354716522588774e-06,
"loss": 0.0456,
"step": 2074
},
{
"epoch": 2.5441176470588234,
"grad_norm": 0.2158203125,
"learning_rate": 3.5252510247058124e-06,
"loss": 0.0484,
"step": 2076
},
{
"epoch": 2.5465686274509802,
"grad_norm": 0.279296875,
"learning_rate": 3.515037142030636e-06,
"loss": 0.0474,
"step": 2078
},
{
"epoch": 2.549019607843137,
"grad_norm": 0.23046875,
"learning_rate": 3.5048300509473228e-06,
"loss": 0.0424,
"step": 2080
},
{
"epoch": 2.5514705882352944,
"grad_norm": 0.267578125,
"learning_rate": 3.4946297981387913e-06,
"loss": 0.0412,
"step": 2082
},
{
"epoch": 2.553921568627451,
"grad_norm": 0.283203125,
"learning_rate": 3.484436430256683e-06,
"loss": 0.0492,
"step": 2084
},
{
"epoch": 2.556372549019608,
"grad_norm": 0.275390625,
"learning_rate": 3.474249993921147e-06,
"loss": 0.048,
"step": 2086
},
{
"epoch": 2.5588235294117645,
"grad_norm": 0.349609375,
"learning_rate": 3.4640705357206382e-06,
"loss": 0.0468,
"step": 2088
},
{
"epoch": 2.561274509803922,
"grad_norm": 0.32421875,
"learning_rate": 3.4538981022116873e-06,
"loss": 0.0536,
"step": 2090
},
{
"epoch": 2.563725490196078,
"grad_norm": 0.259765625,
"learning_rate": 3.4437327399187025e-06,
"loss": 0.0481,
"step": 2092
},
{
"epoch": 2.5661764705882355,
"grad_norm": 0.265625,
"learning_rate": 3.4335744953337542e-06,
"loss": 0.0461,
"step": 2094
},
{
"epoch": 2.568627450980392,
"grad_norm": 0.2021484375,
"learning_rate": 3.423423414916349e-06,
"loss": 0.0454,
"step": 2096
},
{
"epoch": 2.571078431372549,
"grad_norm": 0.25390625,
"learning_rate": 3.4132795450932395e-06,
"loss": 0.0491,
"step": 2098
},
{
"epoch": 2.5735294117647056,
"grad_norm": 0.2216796875,
"learning_rate": 3.4031429322581917e-06,
"loss": 0.0458,
"step": 2100
},
{
"epoch": 2.575980392156863,
"grad_norm": 0.291015625,
"learning_rate": 3.3930136227717843e-06,
"loss": 0.057,
"step": 2102
},
{
"epoch": 2.5784313725490198,
"grad_norm": 0.234375,
"learning_rate": 3.382891662961195e-06,
"loss": 0.0449,
"step": 2104
},
{
"epoch": 2.5808823529411766,
"grad_norm": 0.2138671875,
"learning_rate": 3.372777099119985e-06,
"loss": 0.0439,
"step": 2106
},
{
"epoch": 2.5833333333333335,
"grad_norm": 0.2734375,
"learning_rate": 3.3626699775078884e-06,
"loss": 0.045,
"step": 2108
},
{
"epoch": 2.5857843137254903,
"grad_norm": 0.2734375,
"learning_rate": 3.3525703443506063e-06,
"loss": 0.0429,
"step": 2110
},
{
"epoch": 2.588235294117647,
"grad_norm": 0.2890625,
"learning_rate": 3.342478245839587e-06,
"loss": 0.0482,
"step": 2112
},
{
"epoch": 2.590686274509804,
"grad_norm": 0.244140625,
"learning_rate": 3.3323937281318184e-06,
"loss": 0.0474,
"step": 2114
},
{
"epoch": 2.593137254901961,
"grad_norm": 0.259765625,
"learning_rate": 3.322316837349618e-06,
"loss": 0.0478,
"step": 2116
},
{
"epoch": 2.5955882352941178,
"grad_norm": 0.2490234375,
"learning_rate": 3.312247619580421e-06,
"loss": 0.0397,
"step": 2118
},
{
"epoch": 2.5980392156862746,
"grad_norm": 0.2265625,
"learning_rate": 3.302186120876572e-06,
"loss": 0.0496,
"step": 2120
},
{
"epoch": 2.6004901960784315,
"grad_norm": 0.236328125,
"learning_rate": 3.292132387255106e-06,
"loss": 0.0459,
"step": 2122
},
{
"epoch": 2.6029411764705883,
"grad_norm": 0.234375,
"learning_rate": 3.2820864646975493e-06,
"loss": 0.0485,
"step": 2124
},
{
"epoch": 2.605392156862745,
"grad_norm": 0.2470703125,
"learning_rate": 3.2720483991497e-06,
"loss": 0.0446,
"step": 2126
},
{
"epoch": 2.607843137254902,
"grad_norm": 0.2734375,
"learning_rate": 3.2620182365214224e-06,
"loss": 0.052,
"step": 2128
},
{
"epoch": 2.610294117647059,
"grad_norm": 0.1884765625,
"learning_rate": 3.251996022686441e-06,
"loss": 0.0399,
"step": 2130
},
{
"epoch": 2.6127450980392157,
"grad_norm": 0.263671875,
"learning_rate": 3.241981803482116e-06,
"loss": 0.0438,
"step": 2132
},
{
"epoch": 2.6151960784313726,
"grad_norm": 0.30859375,
"learning_rate": 3.2319756247092552e-06,
"loss": 0.0472,
"step": 2134
},
{
"epoch": 2.6176470588235294,
"grad_norm": 0.259765625,
"learning_rate": 3.221977532131882e-06,
"loss": 0.0434,
"step": 2136
},
{
"epoch": 2.6200980392156863,
"grad_norm": 0.2294921875,
"learning_rate": 3.2119875714770433e-06,
"loss": 0.0415,
"step": 2138
},
{
"epoch": 2.622549019607843,
"grad_norm": 0.271484375,
"learning_rate": 3.2020057884345963e-06,
"loss": 0.0468,
"step": 2140
},
{
"epoch": 2.625,
"grad_norm": 0.2265625,
"learning_rate": 3.1920322286569877e-06,
"loss": 0.0485,
"step": 2142
},
{
"epoch": 2.627450980392157,
"grad_norm": 0.244140625,
"learning_rate": 3.182066937759066e-06,
"loss": 0.0469,
"step": 2144
},
{
"epoch": 2.6299019607843137,
"grad_norm": 0.216796875,
"learning_rate": 3.172109961317851e-06,
"loss": 0.0489,
"step": 2146
},
{
"epoch": 2.6323529411764706,
"grad_norm": 0.275390625,
"learning_rate": 3.1621613448723414e-06,
"loss": 0.0424,
"step": 2148
},
{
"epoch": 2.6348039215686274,
"grad_norm": 0.2490234375,
"learning_rate": 3.152221133923303e-06,
"loss": 0.0446,
"step": 2150
},
{
"epoch": 2.6372549019607843,
"grad_norm": 0.337890625,
"learning_rate": 3.1422893739330524e-06,
"loss": 0.0455,
"step": 2152
},
{
"epoch": 2.639705882352941,
"grad_norm": 0.263671875,
"learning_rate": 3.132366110325259e-06,
"loss": 0.048,
"step": 2154
},
{
"epoch": 2.642156862745098,
"grad_norm": 0.21484375,
"learning_rate": 3.122451388484732e-06,
"loss": 0.0433,
"step": 2156
},
{
"epoch": 2.644607843137255,
"grad_norm": 0.25390625,
"learning_rate": 3.112545253757213e-06,
"loss": 0.051,
"step": 2158
},
{
"epoch": 2.6470588235294117,
"grad_norm": 0.2470703125,
"learning_rate": 3.102647751449174e-06,
"loss": 0.04,
"step": 2160
},
{
"epoch": 2.6495098039215685,
"grad_norm": 0.25390625,
"learning_rate": 3.0927589268276013e-06,
"loss": 0.0509,
"step": 2162
},
{
"epoch": 2.6519607843137254,
"grad_norm": 0.2431640625,
"learning_rate": 3.082878825119796e-06,
"loss": 0.0466,
"step": 2164
},
{
"epoch": 2.6544117647058822,
"grad_norm": 0.244140625,
"learning_rate": 3.073007491513162e-06,
"loss": 0.0473,
"step": 2166
},
{
"epoch": 2.656862745098039,
"grad_norm": 0.244140625,
"learning_rate": 3.063144971155004e-06,
"loss": 0.0474,
"step": 2168
},
{
"epoch": 2.659313725490196,
"grad_norm": 0.251953125,
"learning_rate": 3.0532913091523187e-06,
"loss": 0.0417,
"step": 2170
},
{
"epoch": 2.661764705882353,
"grad_norm": 0.2177734375,
"learning_rate": 3.043446550571584e-06,
"loss": 0.045,
"step": 2172
},
{
"epoch": 2.6642156862745097,
"grad_norm": 0.296875,
"learning_rate": 3.0336107404385632e-06,
"loss": 0.0525,
"step": 2174
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.283203125,
"learning_rate": 3.0237839237380905e-06,
"loss": 0.0445,
"step": 2176
},
{
"epoch": 2.6691176470588234,
"grad_norm": 0.279296875,
"learning_rate": 3.013966145413866e-06,
"loss": 0.0461,
"step": 2178
},
{
"epoch": 2.6715686274509802,
"grad_norm": 0.28515625,
"learning_rate": 3.004157450368258e-06,
"loss": 0.0436,
"step": 2180
},
{
"epoch": 2.674019607843137,
"grad_norm": 0.2451171875,
"learning_rate": 2.994357883462083e-06,
"loss": 0.0423,
"step": 2182
},
{
"epoch": 2.6764705882352944,
"grad_norm": 0.248046875,
"learning_rate": 2.9845674895144165e-06,
"loss": 0.0453,
"step": 2184
},
{
"epoch": 2.678921568627451,
"grad_norm": 0.2392578125,
"learning_rate": 2.9747863133023803e-06,
"loss": 0.047,
"step": 2186
},
{
"epoch": 2.681372549019608,
"grad_norm": 0.2470703125,
"learning_rate": 2.9650143995609326e-06,
"loss": 0.0524,
"step": 2188
},
{
"epoch": 2.6838235294117645,
"grad_norm": 0.201171875,
"learning_rate": 2.9552517929826763e-06,
"loss": 0.0415,
"step": 2190
},
{
"epoch": 2.686274509803922,
"grad_norm": 0.2392578125,
"learning_rate": 2.945498538217639e-06,
"loss": 0.0448,
"step": 2192
},
{
"epoch": 2.688725490196078,
"grad_norm": 0.1953125,
"learning_rate": 2.935754679873085e-06,
"loss": 0.0415,
"step": 2194
},
{
"epoch": 2.6911764705882355,
"grad_norm": 0.2734375,
"learning_rate": 2.9260202625133016e-06,
"loss": 0.0444,
"step": 2196
},
{
"epoch": 2.693627450980392,
"grad_norm": 0.2451171875,
"learning_rate": 2.9162953306593895e-06,
"loss": 0.0444,
"step": 2198
},
{
"epoch": 2.696078431372549,
"grad_norm": 0.291015625,
"learning_rate": 2.9065799287890797e-06,
"loss": 0.0462,
"step": 2200
},
{
"epoch": 2.6985294117647056,
"grad_norm": 0.240234375,
"learning_rate": 2.896874101336504e-06,
"loss": 0.0462,
"step": 2202
},
{
"epoch": 2.700980392156863,
"grad_norm": 0.2578125,
"learning_rate": 2.8871778926920136e-06,
"loss": 0.0518,
"step": 2204
},
{
"epoch": 2.7034313725490198,
"grad_norm": 0.2333984375,
"learning_rate": 2.8774913472019666e-06,
"loss": 0.0461,
"step": 2206
},
{
"epoch": 2.7058823529411766,
"grad_norm": 0.26171875,
"learning_rate": 2.86781450916852e-06,
"loss": 0.0475,
"step": 2208
},
{
"epoch": 2.7083333333333335,
"grad_norm": 0.337890625,
"learning_rate": 2.858147422849441e-06,
"loss": 0.0418,
"step": 2210
},
{
"epoch": 2.7107843137254903,
"grad_norm": 0.328125,
"learning_rate": 2.8484901324578883e-06,
"loss": 0.047,
"step": 2212
},
{
"epoch": 2.713235294117647,
"grad_norm": 0.302734375,
"learning_rate": 2.838842682162224e-06,
"loss": 0.0437,
"step": 2214
},
{
"epoch": 2.715686274509804,
"grad_norm": 0.3046875,
"learning_rate": 2.8292051160858077e-06,
"loss": 0.0485,
"step": 2216
},
{
"epoch": 2.718137254901961,
"grad_norm": 0.248046875,
"learning_rate": 2.819577478306783e-06,
"loss": 0.0486,
"step": 2218
},
{
"epoch": 2.7205882352941178,
"grad_norm": 0.2265625,
"learning_rate": 2.809959812857898e-06,
"loss": 0.0446,
"step": 2220
},
{
"epoch": 2.7230392156862746,
"grad_norm": 0.236328125,
"learning_rate": 2.800352163726281e-06,
"loss": 0.0414,
"step": 2222
},
{
"epoch": 2.7254901960784315,
"grad_norm": 0.275390625,
"learning_rate": 2.790754574853255e-06,
"loss": 0.0457,
"step": 2224
},
{
"epoch": 2.7279411764705883,
"grad_norm": 0.2373046875,
"learning_rate": 2.781167090134135e-06,
"loss": 0.0427,
"step": 2226
},
{
"epoch": 2.730392156862745,
"grad_norm": 0.259765625,
"learning_rate": 2.7715897534180136e-06,
"loss": 0.046,
"step": 2228
},
{
"epoch": 2.732843137254902,
"grad_norm": 0.228515625,
"learning_rate": 2.7620226085075835e-06,
"loss": 0.046,
"step": 2230
},
{
"epoch": 2.735294117647059,
"grad_norm": 0.279296875,
"learning_rate": 2.752465699158913e-06,
"loss": 0.0494,
"step": 2232
},
{
"epoch": 2.7377450980392157,
"grad_norm": 0.32421875,
"learning_rate": 2.7429190690812636e-06,
"loss": 0.0493,
"step": 2234
},
{
"epoch": 2.7401960784313726,
"grad_norm": 0.263671875,
"learning_rate": 2.733382761936884e-06,
"loss": 0.0437,
"step": 2236
},
{
"epoch": 2.7426470588235294,
"grad_norm": 0.228515625,
"learning_rate": 2.723856821340806e-06,
"loss": 0.0458,
"step": 2238
},
{
"epoch": 2.7450980392156863,
"grad_norm": 0.232421875,
"learning_rate": 2.714341290860655e-06,
"loss": 0.0496,
"step": 2240
},
{
"epoch": 2.747549019607843,
"grad_norm": 0.2412109375,
"learning_rate": 2.7048362140164373e-06,
"loss": 0.0475,
"step": 2242
},
{
"epoch": 2.75,
"grad_norm": 0.2294921875,
"learning_rate": 2.6953416342803564e-06,
"loss": 0.0425,
"step": 2244
},
{
"epoch": 2.752450980392157,
"grad_norm": 0.259765625,
"learning_rate": 2.685857595076603e-06,
"loss": 0.0458,
"step": 2246
},
{
"epoch": 2.7549019607843137,
"grad_norm": 0.2373046875,
"learning_rate": 2.6763841397811576e-06,
"loss": 0.0474,
"step": 2248
},
{
"epoch": 2.7573529411764706,
"grad_norm": 0.3515625,
"learning_rate": 2.6669213117215985e-06,
"loss": 0.054,
"step": 2250
},
{
"epoch": 2.7598039215686274,
"grad_norm": 0.2119140625,
"learning_rate": 2.657469154176899e-06,
"loss": 0.0418,
"step": 2252
},
{
"epoch": 2.7622549019607843,
"grad_norm": 0.380859375,
"learning_rate": 2.6480277103772255e-06,
"loss": 0.0426,
"step": 2254
},
{
"epoch": 2.764705882352941,
"grad_norm": 0.224609375,
"learning_rate": 2.638597023503751e-06,
"loss": 0.0463,
"step": 2256
},
{
"epoch": 2.767156862745098,
"grad_norm": 0.2119140625,
"learning_rate": 2.6291771366884445e-06,
"loss": 0.0469,
"step": 2258
},
{
"epoch": 2.769607843137255,
"grad_norm": 0.2294921875,
"learning_rate": 2.6197680930138834e-06,
"loss": 0.0426,
"step": 2260
},
{
"epoch": 2.7720588235294117,
"grad_norm": 0.255859375,
"learning_rate": 2.6103699355130562e-06,
"loss": 0.0387,
"step": 2262
},
{
"epoch": 2.7745098039215685,
"grad_norm": 0.244140625,
"learning_rate": 2.600982707169154e-06,
"loss": 0.0466,
"step": 2264
},
{
"epoch": 2.7769607843137254,
"grad_norm": 0.25390625,
"learning_rate": 2.5916064509153925e-06,
"loss": 0.0509,
"step": 2266
},
{
"epoch": 2.7794117647058822,
"grad_norm": 0.30078125,
"learning_rate": 2.582241209634797e-06,
"loss": 0.052,
"step": 2268
},
{
"epoch": 2.781862745098039,
"grad_norm": 0.26953125,
"learning_rate": 2.5728870261600203e-06,
"loss": 0.0418,
"step": 2270
},
{
"epoch": 2.784313725490196,
"grad_norm": 0.30859375,
"learning_rate": 2.563543943273143e-06,
"loss": 0.0451,
"step": 2272
},
{
"epoch": 2.786764705882353,
"grad_norm": 0.275390625,
"learning_rate": 2.5542120037054684e-06,
"loss": 0.046,
"step": 2274
},
{
"epoch": 2.7892156862745097,
"grad_norm": 0.2236328125,
"learning_rate": 2.5448912501373458e-06,
"loss": 0.0454,
"step": 2276
},
{
"epoch": 2.7916666666666665,
"grad_norm": 0.2314453125,
"learning_rate": 2.535581725197952e-06,
"loss": 0.0423,
"step": 2278
},
{
"epoch": 2.7941176470588234,
"grad_norm": 0.2412109375,
"learning_rate": 2.5262834714651195e-06,
"loss": 0.045,
"step": 2280
},
{
"epoch": 2.7965686274509802,
"grad_norm": 0.21875,
"learning_rate": 2.5169965314651283e-06,
"loss": 0.0396,
"step": 2282
},
{
"epoch": 2.799019607843137,
"grad_norm": 0.26953125,
"learning_rate": 2.5077209476725094e-06,
"loss": 0.0405,
"step": 2284
},
{
"epoch": 2.8014705882352944,
"grad_norm": 0.240234375,
"learning_rate": 2.498456762509862e-06,
"loss": 0.0474,
"step": 2286
},
{
"epoch": 2.803921568627451,
"grad_norm": 0.224609375,
"learning_rate": 2.4892040183476462e-06,
"loss": 0.0494,
"step": 2288
},
{
"epoch": 2.806372549019608,
"grad_norm": 0.2734375,
"learning_rate": 2.4799627575040014e-06,
"loss": 0.0496,
"step": 2290
},
{
"epoch": 2.8088235294117645,
"grad_norm": 0.267578125,
"learning_rate": 2.4707330222445465e-06,
"loss": 0.0451,
"step": 2292
},
{
"epoch": 2.811274509803922,
"grad_norm": 0.228515625,
"learning_rate": 2.4615148547821815e-06,
"loss": 0.0415,
"step": 2294
},
{
"epoch": 2.813725490196078,
"grad_norm": 0.2421875,
"learning_rate": 2.4523082972769075e-06,
"loss": 0.0486,
"step": 2296
},
{
"epoch": 2.8161764705882355,
"grad_norm": 0.388671875,
"learning_rate": 2.443113391835624e-06,
"loss": 0.0501,
"step": 2298
},
{
"epoch": 2.818627450980392,
"grad_norm": 0.23046875,
"learning_rate": 2.433930180511936e-06,
"loss": 0.0436,
"step": 2300
},
{
"epoch": 2.821078431372549,
"grad_norm": 0.2158203125,
"learning_rate": 2.4247587053059674e-06,
"loss": 0.0457,
"step": 2302
},
{
"epoch": 2.8235294117647056,
"grad_norm": 0.236328125,
"learning_rate": 2.4155990081641685e-06,
"loss": 0.0493,
"step": 2304
},
{
"epoch": 2.825980392156863,
"grad_norm": 0.2216796875,
"learning_rate": 2.4064511309791143e-06,
"loss": 0.0439,
"step": 2306
},
{
"epoch": 2.8284313725490198,
"grad_norm": 0.25390625,
"learning_rate": 2.3973151155893276e-06,
"loss": 0.0431,
"step": 2308
},
{
"epoch": 2.8308823529411766,
"grad_norm": 0.298828125,
"learning_rate": 2.3881910037790784e-06,
"loss": 0.0474,
"step": 2310
},
{
"epoch": 2.8333333333333335,
"grad_norm": 0.2314453125,
"learning_rate": 2.3790788372781915e-06,
"loss": 0.0427,
"step": 2312
},
{
"epoch": 2.8357843137254903,
"grad_norm": 0.30078125,
"learning_rate": 2.3699786577618622e-06,
"loss": 0.0432,
"step": 2314
},
{
"epoch": 2.838235294117647,
"grad_norm": 0.26171875,
"learning_rate": 2.360890506850464e-06,
"loss": 0.0435,
"step": 2316
},
{
"epoch": 2.840686274509804,
"grad_norm": 0.2470703125,
"learning_rate": 2.3518144261093495e-06,
"loss": 0.0456,
"step": 2318
},
{
"epoch": 2.843137254901961,
"grad_norm": 0.2314453125,
"learning_rate": 2.342750457048675e-06,
"loss": 0.0435,
"step": 2320
},
{
"epoch": 2.8455882352941178,
"grad_norm": 0.185546875,
"learning_rate": 2.3336986411232025e-06,
"loss": 0.043,
"step": 2322
},
{
"epoch": 2.8480392156862746,
"grad_norm": 0.318359375,
"learning_rate": 2.3246590197321038e-06,
"loss": 0.0476,
"step": 2324
},
{
"epoch": 2.8504901960784315,
"grad_norm": 0.314453125,
"learning_rate": 2.3156316342187853e-06,
"loss": 0.0421,
"step": 2326
},
{
"epoch": 2.8529411764705883,
"grad_norm": 0.2080078125,
"learning_rate": 2.3066165258706904e-06,
"loss": 0.0397,
"step": 2328
},
{
"epoch": 2.855392156862745,
"grad_norm": 0.27734375,
"learning_rate": 2.297613735919107e-06,
"loss": 0.0476,
"step": 2330
},
{
"epoch": 2.857843137254902,
"grad_norm": 0.2734375,
"learning_rate": 2.2886233055389914e-06,
"loss": 0.0498,
"step": 2332
},
{
"epoch": 2.860294117647059,
"grad_norm": 0.248046875,
"learning_rate": 2.279645275848764e-06,
"loss": 0.0447,
"step": 2334
},
{
"epoch": 2.8627450980392157,
"grad_norm": 0.21875,
"learning_rate": 2.270679687910138e-06,
"loss": 0.044,
"step": 2336
},
{
"epoch": 2.8651960784313726,
"grad_norm": 0.2412109375,
"learning_rate": 2.26172658272792e-06,
"loss": 0.0425,
"step": 2338
},
{
"epoch": 2.8676470588235294,
"grad_norm": 0.2158203125,
"learning_rate": 2.2527860012498216e-06,
"loss": 0.0436,
"step": 2340
},
{
"epoch": 2.8700980392156863,
"grad_norm": 0.26953125,
"learning_rate": 2.243857984366284e-06,
"loss": 0.0443,
"step": 2342
},
{
"epoch": 2.872549019607843,
"grad_norm": 0.25390625,
"learning_rate": 2.2349425729102752e-06,
"loss": 0.044,
"step": 2344
},
{
"epoch": 2.875,
"grad_norm": 0.251953125,
"learning_rate": 2.2260398076571176e-06,
"loss": 0.0444,
"step": 2346
},
{
"epoch": 2.877450980392157,
"grad_norm": 0.328125,
"learning_rate": 2.217149729324295e-06,
"loss": 0.0452,
"step": 2348
},
{
"epoch": 2.8799019607843137,
"grad_norm": 0.240234375,
"learning_rate": 2.2082723785712603e-06,
"loss": 0.0513,
"step": 2350
},
{
"epoch": 2.8823529411764706,
"grad_norm": 0.2314453125,
"learning_rate": 2.199407795999263e-06,
"loss": 0.0445,
"step": 2352
},
{
"epoch": 2.8848039215686274,
"grad_norm": 0.255859375,
"learning_rate": 2.190556022151152e-06,
"loss": 0.0435,
"step": 2354
},
{
"epoch": 2.8872549019607843,
"grad_norm": 0.2470703125,
"learning_rate": 2.1817170975111957e-06,
"loss": 0.0473,
"step": 2356
},
{
"epoch": 2.889705882352941,
"grad_norm": 0.2578125,
"learning_rate": 2.172891062504899e-06,
"loss": 0.0452,
"step": 2358
},
{
"epoch": 2.892156862745098,
"grad_norm": 0.2392578125,
"learning_rate": 2.164077957498807e-06,
"loss": 0.0441,
"step": 2360
},
{
"epoch": 2.894607843137255,
"grad_norm": 0.275390625,
"learning_rate": 2.1552778228003383e-06,
"loss": 0.0492,
"step": 2362
},
{
"epoch": 2.8970588235294117,
"grad_norm": 0.34765625,
"learning_rate": 2.146490698657582e-06,
"loss": 0.0424,
"step": 2364
},
{
"epoch": 2.8995098039215685,
"grad_norm": 0.26953125,
"learning_rate": 2.1377166252591276e-06,
"loss": 0.0477,
"step": 2366
},
{
"epoch": 2.9019607843137254,
"grad_norm": 0.2265625,
"learning_rate": 2.128955642733877e-06,
"loss": 0.0501,
"step": 2368
},
{
"epoch": 2.9044117647058822,
"grad_norm": 0.27734375,
"learning_rate": 2.120207791150854e-06,
"loss": 0.0454,
"step": 2370
},
{
"epoch": 2.906862745098039,
"grad_norm": 0.271484375,
"learning_rate": 2.1114731105190352e-06,
"loss": 0.0468,
"step": 2372
},
{
"epoch": 2.909313725490196,
"grad_norm": 0.232421875,
"learning_rate": 2.102751640787149e-06,
"loss": 0.0472,
"step": 2374
},
{
"epoch": 2.911764705882353,
"grad_norm": 0.2890625,
"learning_rate": 2.0940434218435115e-06,
"loss": 0.046,
"step": 2376
},
{
"epoch": 2.9142156862745097,
"grad_norm": 0.251953125,
"learning_rate": 2.085348493515833e-06,
"loss": 0.0496,
"step": 2378
},
{
"epoch": 2.9166666666666665,
"grad_norm": 0.314453125,
"learning_rate": 2.0766668955710334e-06,
"loss": 0.0459,
"step": 2380
},
{
"epoch": 2.9191176470588234,
"grad_norm": 0.275390625,
"learning_rate": 2.0679986677150714e-06,
"loss": 0.0461,
"step": 2382
},
{
"epoch": 2.9215686274509802,
"grad_norm": 0.267578125,
"learning_rate": 2.059343849592755e-06,
"loss": 0.0413,
"step": 2384
},
{
"epoch": 2.924019607843137,
"grad_norm": 0.275390625,
"learning_rate": 2.050702480787557e-06,
"loss": 0.0517,
"step": 2386
},
{
"epoch": 2.9264705882352944,
"grad_norm": 0.3203125,
"learning_rate": 2.0420746008214463e-06,
"loss": 0.0474,
"step": 2388
},
{
"epoch": 2.928921568627451,
"grad_norm": 0.310546875,
"learning_rate": 2.033460249154692e-06,
"loss": 0.0499,
"step": 2390
},
{
"epoch": 2.931372549019608,
"grad_norm": 0.2890625,
"learning_rate": 2.0248594651856964e-06,
"loss": 0.0495,
"step": 2392
},
{
"epoch": 2.9338235294117645,
"grad_norm": 0.263671875,
"learning_rate": 2.0162722882508072e-06,
"loss": 0.0507,
"step": 2394
},
{
"epoch": 2.936274509803922,
"grad_norm": 0.2373046875,
"learning_rate": 2.0076987576241365e-06,
"loss": 0.0418,
"step": 2396
},
{
"epoch": 2.938725490196078,
"grad_norm": 0.287109375,
"learning_rate": 1.9991389125173882e-06,
"loss": 0.0449,
"step": 2398
},
{
"epoch": 2.9411764705882355,
"grad_norm": 0.212890625,
"learning_rate": 1.9905927920796704e-06,
"loss": 0.0453,
"step": 2400
},
{
"epoch": 2.943627450980392,
"grad_norm": 0.244140625,
"learning_rate": 1.9820604353973226e-06,
"loss": 0.047,
"step": 2402
},
{
"epoch": 2.946078431372549,
"grad_norm": 0.2490234375,
"learning_rate": 1.973541881493738e-06,
"loss": 0.048,
"step": 2404
},
{
"epoch": 2.9485294117647056,
"grad_norm": 0.26953125,
"learning_rate": 1.965037169329173e-06,
"loss": 0.045,
"step": 2406
},
{
"epoch": 2.950980392156863,
"grad_norm": 0.287109375,
"learning_rate": 1.9565463378005873e-06,
"loss": 0.0448,
"step": 2408
},
{
"epoch": 2.9534313725490198,
"grad_norm": 0.25390625,
"learning_rate": 1.9480694257414486e-06,
"loss": 0.0453,
"step": 2410
},
{
"epoch": 2.9558823529411766,
"grad_norm": 0.318359375,
"learning_rate": 1.9396064719215675e-06,
"loss": 0.0548,
"step": 2412
},
{
"epoch": 2.9583333333333335,
"grad_norm": 0.2578125,
"learning_rate": 1.9311575150469146e-06,
"loss": 0.0458,
"step": 2414
},
{
"epoch": 2.9607843137254903,
"grad_norm": 0.251953125,
"learning_rate": 1.9227225937594463e-06,
"loss": 0.0473,
"step": 2416
},
{
"epoch": 2.963235294117647,
"grad_norm": 0.234375,
"learning_rate": 1.9143017466369203e-06,
"loss": 0.0485,
"step": 2418
},
{
"epoch": 2.965686274509804,
"grad_norm": 0.271484375,
"learning_rate": 1.90589501219273e-06,
"loss": 0.0453,
"step": 2420
},
{
"epoch": 2.968137254901961,
"grad_norm": 0.33984375,
"learning_rate": 1.8975024288757238e-06,
"loss": 0.0502,
"step": 2422
},
{
"epoch": 2.9705882352941178,
"grad_norm": 0.2236328125,
"learning_rate": 1.889124035070024e-06,
"loss": 0.0468,
"step": 2424
},
{
"epoch": 2.9730392156862746,
"grad_norm": 0.23046875,
"learning_rate": 1.8807598690948614e-06,
"loss": 0.0402,
"step": 2426
},
{
"epoch": 2.9754901960784315,
"grad_norm": 0.271484375,
"learning_rate": 1.8724099692043934e-06,
"loss": 0.0502,
"step": 2428
},
{
"epoch": 2.9779411764705883,
"grad_norm": 0.275390625,
"learning_rate": 1.8640743735875266e-06,
"loss": 0.0463,
"step": 2430
},
{
"epoch": 2.980392156862745,
"grad_norm": 0.220703125,
"learning_rate": 1.8557531203677504e-06,
"loss": 0.0447,
"step": 2432
},
{
"epoch": 2.982843137254902,
"grad_norm": 0.30078125,
"learning_rate": 1.8474462476029575e-06,
"loss": 0.0447,
"step": 2434
},
{
"epoch": 2.985294117647059,
"grad_norm": 0.2470703125,
"learning_rate": 1.8391537932852677e-06,
"loss": 0.0399,
"step": 2436
},
{
"epoch": 2.9877450980392157,
"grad_norm": 0.212890625,
"learning_rate": 1.8308757953408601e-06,
"loss": 0.0445,
"step": 2438
},
{
"epoch": 2.9901960784313726,
"grad_norm": 0.310546875,
"learning_rate": 1.8226122916297933e-06,
"loss": 0.055,
"step": 2440
},
{
"epoch": 2.9926470588235294,
"grad_norm": 0.255859375,
"learning_rate": 1.8143633199458383e-06,
"loss": 0.0424,
"step": 2442
},
{
"epoch": 2.9950980392156863,
"grad_norm": 0.30078125,
"learning_rate": 1.806128918016304e-06,
"loss": 0.0524,
"step": 2444
},
{
"epoch": 2.997549019607843,
"grad_norm": 0.251953125,
"learning_rate": 1.7979091235018564e-06,
"loss": 0.0503,
"step": 2446
},
{
"epoch": 3.0,
"grad_norm": 0.40234375,
"learning_rate": 1.7897039739963607e-06,
"loss": 0.0443,
"step": 2448
},
{
"epoch": 3.0,
"eval_loss": 0.0535544753074646,
"eval_runtime": 37.6568,
"eval_samples_per_second": 133.203,
"eval_steps_per_second": 1.062,
"step": 2448
},
{
"epoch": 3.002450980392157,
"grad_norm": 0.255859375,
"learning_rate": 1.7815135070267008e-06,
"loss": 0.0437,
"step": 2450
},
{
"epoch": 3.0049019607843137,
"grad_norm": 0.283203125,
"learning_rate": 1.7733377600526041e-06,
"loss": 0.0417,
"step": 2452
},
{
"epoch": 3.0073529411764706,
"grad_norm": 0.27734375,
"learning_rate": 1.7651767704664819e-06,
"loss": 0.0412,
"step": 2454
},
{
"epoch": 3.0098039215686274,
"grad_norm": 0.263671875,
"learning_rate": 1.7570305755932438e-06,
"loss": 0.0494,
"step": 2456
},
{
"epoch": 3.0122549019607843,
"grad_norm": 0.306640625,
"learning_rate": 1.748899212690141e-06,
"loss": 0.0458,
"step": 2458
},
{
"epoch": 3.014705882352941,
"grad_norm": 0.1943359375,
"learning_rate": 1.7407827189465893e-06,
"loss": 0.0407,
"step": 2460
},
{
"epoch": 3.017156862745098,
"grad_norm": 0.2490234375,
"learning_rate": 1.7326811314839931e-06,
"loss": 0.0533,
"step": 2462
},
{
"epoch": 3.019607843137255,
"grad_norm": 0.2890625,
"learning_rate": 1.7245944873555893e-06,
"loss": 0.0524,
"step": 2464
},
{
"epoch": 3.0220588235294117,
"grad_norm": 0.2490234375,
"learning_rate": 1.716522823546264e-06,
"loss": 0.0486,
"step": 2466
},
{
"epoch": 3.0245098039215685,
"grad_norm": 0.26953125,
"learning_rate": 1.7084661769723942e-06,
"loss": 0.0484,
"step": 2468
},
{
"epoch": 3.0269607843137254,
"grad_norm": 0.298828125,
"learning_rate": 1.700424584481674e-06,
"loss": 0.045,
"step": 2470
},
{
"epoch": 3.0294117647058822,
"grad_norm": 0.21484375,
"learning_rate": 1.6923980828529424e-06,
"loss": 0.0513,
"step": 2472
},
{
"epoch": 3.031862745098039,
"grad_norm": 0.2314453125,
"learning_rate": 1.6843867087960252e-06,
"loss": 0.0471,
"step": 2474
},
{
"epoch": 3.034313725490196,
"grad_norm": 0.244140625,
"learning_rate": 1.6763904989515545e-06,
"loss": 0.0441,
"step": 2476
},
{
"epoch": 3.036764705882353,
"grad_norm": 0.283203125,
"learning_rate": 1.6684094898908131e-06,
"loss": 0.0477,
"step": 2478
},
{
"epoch": 3.0392156862745097,
"grad_norm": 0.2470703125,
"learning_rate": 1.6604437181155613e-06,
"loss": 0.0463,
"step": 2480
},
{
"epoch": 3.0416666666666665,
"grad_norm": 0.255859375,
"learning_rate": 1.652493220057866e-06,
"loss": 0.0456,
"step": 2482
},
{
"epoch": 3.0441176470588234,
"grad_norm": 0.208984375,
"learning_rate": 1.6445580320799441e-06,
"loss": 0.0463,
"step": 2484
},
{
"epoch": 3.0465686274509802,
"grad_norm": 0.234375,
"learning_rate": 1.636638190473987e-06,
"loss": 0.0483,
"step": 2486
},
{
"epoch": 3.049019607843137,
"grad_norm": 0.291015625,
"learning_rate": 1.628733731461999e-06,
"loss": 0.0488,
"step": 2488
},
{
"epoch": 3.051470588235294,
"grad_norm": 0.328125,
"learning_rate": 1.6208446911956344e-06,
"loss": 0.0441,
"step": 2490
},
{
"epoch": 3.053921568627451,
"grad_norm": 0.251953125,
"learning_rate": 1.6129711057560215e-06,
"loss": 0.0411,
"step": 2492
},
{
"epoch": 3.0563725490196076,
"grad_norm": 0.3515625,
"learning_rate": 1.6051130111536134e-06,
"loss": 0.0514,
"step": 2494
},
{
"epoch": 3.0588235294117645,
"grad_norm": 0.2734375,
"learning_rate": 1.5972704433280063e-06,
"loss": 0.0449,
"step": 2496
},
{
"epoch": 3.0612745098039214,
"grad_norm": 0.2197265625,
"learning_rate": 1.589443438147789e-06,
"loss": 0.0425,
"step": 2498
},
{
"epoch": 3.063725490196078,
"grad_norm": 0.279296875,
"learning_rate": 1.581632031410374e-06,
"loss": 0.0487,
"step": 2500
},
{
"epoch": 3.0661764705882355,
"grad_norm": 0.2353515625,
"learning_rate": 1.5738362588418265e-06,
"loss": 0.0469,
"step": 2502
},
{
"epoch": 3.0686274509803924,
"grad_norm": 0.255859375,
"learning_rate": 1.5660561560967169e-06,
"loss": 0.0475,
"step": 2504
},
{
"epoch": 3.071078431372549,
"grad_norm": 0.24609375,
"learning_rate": 1.5582917587579394e-06,
"loss": 0.0463,
"step": 2506
},
{
"epoch": 3.073529411764706,
"grad_norm": 0.224609375,
"learning_rate": 1.5505431023365642e-06,
"loss": 0.0488,
"step": 2508
},
{
"epoch": 3.075980392156863,
"grad_norm": 0.248046875,
"learning_rate": 1.5428102222716696e-06,
"loss": 0.0499,
"step": 2510
},
{
"epoch": 3.0784313725490198,
"grad_norm": 0.255859375,
"learning_rate": 1.5350931539301734e-06,
"loss": 0.0461,
"step": 2512
},
{
"epoch": 3.0808823529411766,
"grad_norm": 0.263671875,
"learning_rate": 1.527391932606684e-06,
"loss": 0.0545,
"step": 2514
},
{
"epoch": 3.0833333333333335,
"grad_norm": 0.234375,
"learning_rate": 1.5197065935233301e-06,
"loss": 0.0405,
"step": 2516
},
{
"epoch": 3.0857843137254903,
"grad_norm": 0.2197265625,
"learning_rate": 1.5120371718295984e-06,
"loss": 0.0438,
"step": 2518
},
{
"epoch": 3.088235294117647,
"grad_norm": 0.28515625,
"learning_rate": 1.5043837026021823e-06,
"loss": 0.0503,
"step": 2520
},
{
"epoch": 3.090686274509804,
"grad_norm": 0.224609375,
"learning_rate": 1.4967462208448092e-06,
"loss": 0.0443,
"step": 2522
},
{
"epoch": 3.093137254901961,
"grad_norm": 0.25390625,
"learning_rate": 1.4891247614880904e-06,
"loss": 0.0448,
"step": 2524
},
{
"epoch": 3.0955882352941178,
"grad_norm": 0.244140625,
"learning_rate": 1.4815193593893583e-06,
"loss": 0.047,
"step": 2526
},
{
"epoch": 3.0980392156862746,
"grad_norm": 0.2314453125,
"learning_rate": 1.473930049332502e-06,
"loss": 0.0521,
"step": 2528
},
{
"epoch": 3.1004901960784315,
"grad_norm": 0.28515625,
"learning_rate": 1.4663568660278154e-06,
"loss": 0.0451,
"step": 2530
},
{
"epoch": 3.1029411764705883,
"grad_norm": 0.259765625,
"learning_rate": 1.4587998441118356e-06,
"loss": 0.0523,
"step": 2532
},
{
"epoch": 3.105392156862745,
"grad_norm": 0.28125,
"learning_rate": 1.4512590181471853e-06,
"loss": 0.0418,
"step": 2534
},
{
"epoch": 3.107843137254902,
"grad_norm": 0.298828125,
"learning_rate": 1.4437344226224083e-06,
"loss": 0.0432,
"step": 2536
},
{
"epoch": 3.110294117647059,
"grad_norm": 0.2138671875,
"learning_rate": 1.4362260919518234e-06,
"loss": 0.0491,
"step": 2538
},
{
"epoch": 3.1127450980392157,
"grad_norm": 0.220703125,
"learning_rate": 1.4287340604753596e-06,
"loss": 0.0437,
"step": 2540
},
{
"epoch": 3.1151960784313726,
"grad_norm": 0.298828125,
"learning_rate": 1.4212583624583964e-06,
"loss": 0.0536,
"step": 2542
},
{
"epoch": 3.1176470588235294,
"grad_norm": 0.2109375,
"learning_rate": 1.413799032091615e-06,
"loss": 0.0486,
"step": 2544
},
{
"epoch": 3.1200980392156863,
"grad_norm": 0.31640625,
"learning_rate": 1.4063561034908385e-06,
"loss": 0.045,
"step": 2546
},
{
"epoch": 3.122549019607843,
"grad_norm": 0.2578125,
"learning_rate": 1.3989296106968702e-06,
"loss": 0.0441,
"step": 2548
},
{
"epoch": 3.125,
"grad_norm": 0.244140625,
"learning_rate": 1.3915195876753495e-06,
"loss": 0.0458,
"step": 2550
},
{
"epoch": 3.127450980392157,
"grad_norm": 0.23046875,
"learning_rate": 1.3841260683165836e-06,
"loss": 0.0453,
"step": 2552
},
{
"epoch": 3.1299019607843137,
"grad_norm": 0.25,
"learning_rate": 1.3767490864354034e-06,
"loss": 0.0479,
"step": 2554
},
{
"epoch": 3.1323529411764706,
"grad_norm": 0.244140625,
"learning_rate": 1.369388675771004e-06,
"loss": 0.0481,
"step": 2556
},
{
"epoch": 3.1348039215686274,
"grad_norm": 0.23828125,
"learning_rate": 1.3620448699867877e-06,
"loss": 0.0444,
"step": 2558
},
{
"epoch": 3.1372549019607843,
"grad_norm": 0.31640625,
"learning_rate": 1.354717702670218e-06,
"loss": 0.0519,
"step": 2560
},
{
"epoch": 3.139705882352941,
"grad_norm": 0.2333984375,
"learning_rate": 1.347407207332655e-06,
"loss": 0.0439,
"step": 2562
},
{
"epoch": 3.142156862745098,
"grad_norm": 0.263671875,
"learning_rate": 1.3401134174092128e-06,
"loss": 0.0555,
"step": 2564
},
{
"epoch": 3.144607843137255,
"grad_norm": 0.2109375,
"learning_rate": 1.332836366258603e-06,
"loss": 0.0413,
"step": 2566
},
{
"epoch": 3.1470588235294117,
"grad_norm": 0.21875,
"learning_rate": 1.3255760871629753e-06,
"loss": 0.0422,
"step": 2568
},
{
"epoch": 3.1495098039215685,
"grad_norm": 0.30078125,
"learning_rate": 1.3183326133277774e-06,
"loss": 0.0453,
"step": 2570
},
{
"epoch": 3.1519607843137254,
"grad_norm": 0.25390625,
"learning_rate": 1.3111059778815915e-06,
"loss": 0.0503,
"step": 2572
},
{
"epoch": 3.1544117647058822,
"grad_norm": 0.2412109375,
"learning_rate": 1.3038962138759924e-06,
"loss": 0.0518,
"step": 2574
},
{
"epoch": 3.156862745098039,
"grad_norm": 0.25,
"learning_rate": 1.2967033542853918e-06,
"loss": 0.0409,
"step": 2576
},
{
"epoch": 3.159313725490196,
"grad_norm": 0.2080078125,
"learning_rate": 1.2895274320068829e-06,
"loss": 0.0457,
"step": 2578
},
{
"epoch": 3.161764705882353,
"grad_norm": 0.30078125,
"learning_rate": 1.2823684798601021e-06,
"loss": 0.0484,
"step": 2580
},
{
"epoch": 3.1642156862745097,
"grad_norm": 0.232421875,
"learning_rate": 1.2752265305870637e-06,
"loss": 0.0442,
"step": 2582
},
{
"epoch": 3.1666666666666665,
"grad_norm": 0.2373046875,
"learning_rate": 1.2681016168520244e-06,
"loss": 0.0458,
"step": 2584
},
{
"epoch": 3.1691176470588234,
"grad_norm": 0.251953125,
"learning_rate": 1.2609937712413267e-06,
"loss": 0.0453,
"step": 2586
},
{
"epoch": 3.1715686274509802,
"grad_norm": 0.251953125,
"learning_rate": 1.2539030262632461e-06,
"loss": 0.0502,
"step": 2588
},
{
"epoch": 3.174019607843137,
"grad_norm": 0.23046875,
"learning_rate": 1.2468294143478515e-06,
"loss": 0.046,
"step": 2590
},
{
"epoch": 3.176470588235294,
"grad_norm": 0.2392578125,
"learning_rate": 1.2397729678468534e-06,
"loss": 0.0443,
"step": 2592
},
{
"epoch": 3.178921568627451,
"grad_norm": 0.25390625,
"learning_rate": 1.2327337190334488e-06,
"loss": 0.0484,
"step": 2594
},
{
"epoch": 3.1813725490196076,
"grad_norm": 0.2451171875,
"learning_rate": 1.2257117001021874e-06,
"loss": 0.0455,
"step": 2596
},
{
"epoch": 3.1838235294117645,
"grad_norm": 0.259765625,
"learning_rate": 1.2187069431688092e-06,
"loss": 0.0499,
"step": 2598
},
{
"epoch": 3.186274509803922,
"grad_norm": 0.2099609375,
"learning_rate": 1.211719480270111e-06,
"loss": 0.0427,
"step": 2600
},
{
"epoch": 3.188725490196078,
"grad_norm": 0.2373046875,
"learning_rate": 1.2047493433637935e-06,
"loss": 0.0399,
"step": 2602
},
{
"epoch": 3.1911764705882355,
"grad_norm": 0.275390625,
"learning_rate": 1.1977965643283106e-06,
"loss": 0.0448,
"step": 2604
},
{
"epoch": 3.1936274509803924,
"grad_norm": 0.2294921875,
"learning_rate": 1.1908611749627352e-06,
"loss": 0.042,
"step": 2606
},
{
"epoch": 3.196078431372549,
"grad_norm": 0.25390625,
"learning_rate": 1.1839432069866014e-06,
"loss": 0.0453,
"step": 2608
},
{
"epoch": 3.198529411764706,
"grad_norm": 0.22265625,
"learning_rate": 1.1770426920397688e-06,
"loss": 0.0501,
"step": 2610
},
{
"epoch": 3.200980392156863,
"grad_norm": 0.236328125,
"learning_rate": 1.1701596616822752e-06,
"loss": 0.0435,
"step": 2612
},
{
"epoch": 3.2034313725490198,
"grad_norm": 0.205078125,
"learning_rate": 1.1632941473941867e-06,
"loss": 0.0416,
"step": 2614
},
{
"epoch": 3.2058823529411766,
"grad_norm": 0.2060546875,
"learning_rate": 1.1564461805754635e-06,
"loss": 0.049,
"step": 2616
},
{
"epoch": 3.2083333333333335,
"grad_norm": 0.2060546875,
"learning_rate": 1.1496157925458057e-06,
"loss": 0.0442,
"step": 2618
},
{
"epoch": 3.2107843137254903,
"grad_norm": 0.203125,
"learning_rate": 1.1428030145445207e-06,
"loss": 0.0449,
"step": 2620
},
{
"epoch": 3.213235294117647,
"grad_norm": 0.23828125,
"learning_rate": 1.1360078777303729e-06,
"loss": 0.0443,
"step": 2622
},
{
"epoch": 3.215686274509804,
"grad_norm": 0.28125,
"learning_rate": 1.1292304131814419e-06,
"loss": 0.0493,
"step": 2624
},
{
"epoch": 3.218137254901961,
"grad_norm": 0.283203125,
"learning_rate": 1.1224706518949845e-06,
"loss": 0.048,
"step": 2626
},
{
"epoch": 3.2205882352941178,
"grad_norm": 0.26953125,
"learning_rate": 1.1157286247872873e-06,
"loss": 0.046,
"step": 2628
},
{
"epoch": 3.2230392156862746,
"grad_norm": 0.236328125,
"learning_rate": 1.109004362693531e-06,
"loss": 0.0498,
"step": 2630
},
{
"epoch": 3.2254901960784315,
"grad_norm": 0.2490234375,
"learning_rate": 1.1022978963676462e-06,
"loss": 0.0484,
"step": 2632
},
{
"epoch": 3.2279411764705883,
"grad_norm": 0.21484375,
"learning_rate": 1.0956092564821712e-06,
"loss": 0.0458,
"step": 2634
},
{
"epoch": 3.230392156862745,
"grad_norm": 0.2734375,
"learning_rate": 1.088938473628115e-06,
"loss": 0.0483,
"step": 2636
},
{
"epoch": 3.232843137254902,
"grad_norm": 0.25390625,
"learning_rate": 1.0822855783148184e-06,
"loss": 0.0458,
"step": 2638
},
{
"epoch": 3.235294117647059,
"grad_norm": 0.26953125,
"learning_rate": 1.0756506009698075e-06,
"loss": 0.0437,
"step": 2640
},
{
"epoch": 3.2377450980392157,
"grad_norm": 0.228515625,
"learning_rate": 1.0690335719386618e-06,
"loss": 0.0436,
"step": 2642
},
{
"epoch": 3.2401960784313726,
"grad_norm": 0.267578125,
"learning_rate": 1.0624345214848746e-06,
"loss": 0.0478,
"step": 2644
},
{
"epoch": 3.2426470588235294,
"grad_norm": 0.2421875,
"learning_rate": 1.0558534797897086e-06,
"loss": 0.0409,
"step": 2646
},
{
"epoch": 3.2450980392156863,
"grad_norm": 0.236328125,
"learning_rate": 1.0492904769520646e-06,
"loss": 0.0471,
"step": 2648
},
{
"epoch": 3.247549019607843,
"grad_norm": 0.298828125,
"learning_rate": 1.0427455429883426e-06,
"loss": 0.0462,
"step": 2650
},
{
"epoch": 3.25,
"grad_norm": 0.2451171875,
"learning_rate": 1.0362187078322994e-06,
"loss": 0.0457,
"step": 2652
},
{
"epoch": 3.252450980392157,
"grad_norm": 0.267578125,
"learning_rate": 1.0297100013349181e-06,
"loss": 0.0473,
"step": 2654
},
{
"epoch": 3.2549019607843137,
"grad_norm": 0.259765625,
"learning_rate": 1.0232194532642698e-06,
"loss": 0.0472,
"step": 2656
},
{
"epoch": 3.2573529411764706,
"grad_norm": 0.263671875,
"learning_rate": 1.016747093305376e-06,
"loss": 0.0477,
"step": 2658
},
{
"epoch": 3.2598039215686274,
"grad_norm": 0.26171875,
"learning_rate": 1.0102929510600701e-06,
"loss": 0.0459,
"step": 2660
},
{
"epoch": 3.2622549019607843,
"grad_norm": 0.251953125,
"learning_rate": 1.0038570560468707e-06,
"loss": 0.0471,
"step": 2662
},
{
"epoch": 3.264705882352941,
"grad_norm": 0.271484375,
"learning_rate": 9.974394377008367e-07,
"loss": 0.0459,
"step": 2664
},
{
"epoch": 3.267156862745098,
"grad_norm": 0.259765625,
"learning_rate": 9.910401253734402e-07,
"loss": 0.0431,
"step": 2666
},
{
"epoch": 3.269607843137255,
"grad_norm": 0.234375,
"learning_rate": 9.846591483324296e-07,
"loss": 0.0467,
"step": 2668
},
{
"epoch": 3.2720588235294117,
"grad_norm": 0.265625,
"learning_rate": 9.782965357616925e-07,
"loss": 0.0484,
"step": 2670
},
{
"epoch": 3.2745098039215685,
"grad_norm": 0.26171875,
"learning_rate": 9.719523167611283e-07,
"loss": 0.0417,
"step": 2672
},
{
"epoch": 3.2769607843137254,
"grad_norm": 0.2099609375,
"learning_rate": 9.656265203465093e-07,
"loss": 0.0453,
"step": 2674
},
{
"epoch": 3.2794117647058822,
"grad_norm": 0.2451171875,
"learning_rate": 9.593191754493519e-07,
"loss": 0.0476,
"step": 2676
},
{
"epoch": 3.281862745098039,
"grad_norm": 0.2177734375,
"learning_rate": 9.53030310916786e-07,
"loss": 0.0511,
"step": 2678
},
{
"epoch": 3.284313725490196,
"grad_norm": 0.234375,
"learning_rate": 9.467599555114137e-07,
"loss": 0.0461,
"step": 2680
},
{
"epoch": 3.286764705882353,
"grad_norm": 0.2421875,
"learning_rate": 9.405081379111907e-07,
"loss": 0.0472,
"step": 2682
},
{
"epoch": 3.2892156862745097,
"grad_norm": 0.2060546875,
"learning_rate": 9.342748867092832e-07,
"loss": 0.0442,
"step": 2684
},
{
"epoch": 3.2916666666666665,
"grad_norm": 0.234375,
"learning_rate": 9.280602304139457e-07,
"loss": 0.0422,
"step": 2686
},
{
"epoch": 3.2941176470588234,
"grad_norm": 0.232421875,
"learning_rate": 9.218641974483894e-07,
"loss": 0.0415,
"step": 2688
},
{
"epoch": 3.2965686274509802,
"grad_norm": 0.2099609375,
"learning_rate": 9.156868161506449e-07,
"loss": 0.0442,
"step": 2690
},
{
"epoch": 3.299019607843137,
"grad_norm": 0.287109375,
"learning_rate": 9.095281147734425e-07,
"loss": 0.049,
"step": 2692
},
{
"epoch": 3.301470588235294,
"grad_norm": 0.2392578125,
"learning_rate": 9.033881214840745e-07,
"loss": 0.044,
"step": 2694
},
{
"epoch": 3.303921568627451,
"grad_norm": 0.2578125,
"learning_rate": 8.97266864364274e-07,
"loss": 0.0487,
"step": 2696
},
{
"epoch": 3.306372549019608,
"grad_norm": 0.228515625,
"learning_rate": 8.911643714100821e-07,
"loss": 0.0456,
"step": 2698
},
{
"epoch": 3.3088235294117645,
"grad_norm": 0.265625,
"learning_rate": 8.850806705317183e-07,
"loss": 0.048,
"step": 2700
},
{
"epoch": 3.311274509803922,
"grad_norm": 0.259765625,
"learning_rate": 8.790157895534579e-07,
"loss": 0.0451,
"step": 2702
},
{
"epoch": 3.313725490196078,
"grad_norm": 0.2001953125,
"learning_rate": 8.729697562134992e-07,
"loss": 0.0428,
"step": 2704
},
{
"epoch": 3.3161764705882355,
"grad_norm": 0.2353515625,
"learning_rate": 8.669425981638413e-07,
"loss": 0.0459,
"step": 2706
},
{
"epoch": 3.318627450980392,
"grad_norm": 0.2392578125,
"learning_rate": 8.609343429701561e-07,
"loss": 0.0466,
"step": 2708
},
{
"epoch": 3.321078431372549,
"grad_norm": 0.265625,
"learning_rate": 8.549450181116598e-07,
"loss": 0.0411,
"step": 2710
},
{
"epoch": 3.323529411764706,
"grad_norm": 0.255859375,
"learning_rate": 8.489746509809915e-07,
"loss": 0.0477,
"step": 2712
},
{
"epoch": 3.325980392156863,
"grad_norm": 0.220703125,
"learning_rate": 8.430232688840828e-07,
"loss": 0.0442,
"step": 2714
},
{
"epoch": 3.3284313725490198,
"grad_norm": 0.203125,
"learning_rate": 8.370908990400384e-07,
"loss": 0.0398,
"step": 2716
},
{
"epoch": 3.3308823529411766,
"grad_norm": 0.26171875,
"learning_rate": 8.311775685810097e-07,
"loss": 0.0442,
"step": 2718
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.25390625,
"learning_rate": 8.252833045520659e-07,
"loss": 0.0487,
"step": 2720
},
{
"epoch": 3.3357843137254903,
"grad_norm": 0.21484375,
"learning_rate": 8.194081339110776e-07,
"loss": 0.0459,
"step": 2722
},
{
"epoch": 3.338235294117647,
"grad_norm": 0.2734375,
"learning_rate": 8.13552083528592e-07,
"loss": 0.0464,
"step": 2724
},
{
"epoch": 3.340686274509804,
"grad_norm": 0.2158203125,
"learning_rate": 8.077151801877032e-07,
"loss": 0.0533,
"step": 2726
},
{
"epoch": 3.343137254901961,
"grad_norm": 0.2109375,
"learning_rate": 8.018974505839394e-07,
"loss": 0.0431,
"step": 2728
},
{
"epoch": 3.3455882352941178,
"grad_norm": 0.25390625,
"learning_rate": 7.96098921325133e-07,
"loss": 0.0467,
"step": 2730
},
{
"epoch": 3.3480392156862746,
"grad_norm": 0.234375,
"learning_rate": 7.903196189313039e-07,
"loss": 0.0426,
"step": 2732
},
{
"epoch": 3.3504901960784315,
"grad_norm": 0.265625,
"learning_rate": 7.845595698345382e-07,
"loss": 0.0467,
"step": 2734
},
{
"epoch": 3.3529411764705883,
"grad_norm": 0.22265625,
"learning_rate": 7.788188003788611e-07,
"loss": 0.0432,
"step": 2736
},
{
"epoch": 3.355392156862745,
"grad_norm": 0.314453125,
"learning_rate": 7.73097336820125e-07,
"loss": 0.0508,
"step": 2738
},
{
"epoch": 3.357843137254902,
"grad_norm": 0.2412109375,
"learning_rate": 7.673952053258821e-07,
"loss": 0.0433,
"step": 2740
},
{
"epoch": 3.360294117647059,
"grad_norm": 0.232421875,
"learning_rate": 7.6171243197527e-07,
"loss": 0.0484,
"step": 2742
},
{
"epoch": 3.3627450980392157,
"grad_norm": 0.267578125,
"learning_rate": 7.560490427588907e-07,
"loss": 0.05,
"step": 2744
},
{
"epoch": 3.3651960784313726,
"grad_norm": 0.2177734375,
"learning_rate": 7.504050635786891e-07,
"loss": 0.0433,
"step": 2746
},
{
"epoch": 3.3676470588235294,
"grad_norm": 0.21484375,
"learning_rate": 7.447805202478392e-07,
"loss": 0.044,
"step": 2748
},
{
"epoch": 3.3700980392156863,
"grad_norm": 0.25390625,
"learning_rate": 7.391754384906224e-07,
"loss": 0.0505,
"step": 2750
},
{
"epoch": 3.372549019607843,
"grad_norm": 0.2470703125,
"learning_rate": 7.335898439423106e-07,
"loss": 0.0456,
"step": 2752
},
{
"epoch": 3.375,
"grad_norm": 0.298828125,
"learning_rate": 7.28023762149051e-07,
"loss": 0.0418,
"step": 2754
},
{
"epoch": 3.377450980392157,
"grad_norm": 0.271484375,
"learning_rate": 7.22477218567747e-07,
"loss": 0.0481,
"step": 2756
},
{
"epoch": 3.3799019607843137,
"grad_norm": 0.2255859375,
"learning_rate": 7.16950238565941e-07,
"loss": 0.0477,
"step": 2758
},
{
"epoch": 3.3823529411764706,
"grad_norm": 0.255859375,
"learning_rate": 7.114428474217022e-07,
"loss": 0.0441,
"step": 2760
},
{
"epoch": 3.3848039215686274,
"grad_norm": 0.2236328125,
"learning_rate": 7.059550703235085e-07,
"loss": 0.0466,
"step": 2762
},
{
"epoch": 3.3872549019607843,
"grad_norm": 0.251953125,
"learning_rate": 7.004869323701286e-07,
"loss": 0.0455,
"step": 2764
},
{
"epoch": 3.389705882352941,
"grad_norm": 0.2314453125,
"learning_rate": 6.950384585705133e-07,
"loss": 0.0466,
"step": 2766
},
{
"epoch": 3.392156862745098,
"grad_norm": 0.220703125,
"learning_rate": 6.896096738436775e-07,
"loss": 0.0436,
"step": 2768
},
{
"epoch": 3.394607843137255,
"grad_norm": 0.306640625,
"learning_rate": 6.842006030185833e-07,
"loss": 0.0494,
"step": 2770
},
{
"epoch": 3.3970588235294117,
"grad_norm": 0.26171875,
"learning_rate": 6.788112708340345e-07,
"loss": 0.0449,
"step": 2772
},
{
"epoch": 3.3995098039215685,
"grad_norm": 0.2373046875,
"learning_rate": 6.734417019385564e-07,
"loss": 0.0463,
"step": 2774
},
{
"epoch": 3.4019607843137254,
"grad_norm": 0.2021484375,
"learning_rate": 6.680919208902831e-07,
"loss": 0.0414,
"step": 2776
},
{
"epoch": 3.4044117647058822,
"grad_norm": 0.23828125,
"learning_rate": 6.627619521568529e-07,
"loss": 0.0451,
"step": 2778
},
{
"epoch": 3.406862745098039,
"grad_norm": 0.25,
"learning_rate": 6.574518201152852e-07,
"loss": 0.0489,
"step": 2780
},
{
"epoch": 3.409313725490196,
"grad_norm": 0.263671875,
"learning_rate": 6.521615490518778e-07,
"loss": 0.0464,
"step": 2782
},
{
"epoch": 3.411764705882353,
"grad_norm": 0.228515625,
"learning_rate": 6.46891163162095e-07,
"loss": 0.0413,
"step": 2784
},
{
"epoch": 3.4142156862745097,
"grad_norm": 0.2060546875,
"learning_rate": 6.416406865504482e-07,
"loss": 0.0418,
"step": 2786
},
{
"epoch": 3.4166666666666665,
"grad_norm": 0.23828125,
"learning_rate": 6.364101432303987e-07,
"loss": 0.0514,
"step": 2788
},
{
"epoch": 3.4191176470588234,
"grad_norm": 0.2412109375,
"learning_rate": 6.311995571242385e-07,
"loss": 0.0465,
"step": 2790
},
{
"epoch": 3.4215686274509802,
"grad_norm": 0.267578125,
"learning_rate": 6.26008952062983e-07,
"loss": 0.0475,
"step": 2792
},
{
"epoch": 3.424019607843137,
"grad_norm": 0.275390625,
"learning_rate": 6.208383517862648e-07,
"loss": 0.0454,
"step": 2794
},
{
"epoch": 3.426470588235294,
"grad_norm": 0.1982421875,
"learning_rate": 6.156877799422212e-07,
"loss": 0.0469,
"step": 2796
},
{
"epoch": 3.428921568627451,
"grad_norm": 0.25390625,
"learning_rate": 6.105572600873883e-07,
"loss": 0.048,
"step": 2798
},
{
"epoch": 3.431372549019608,
"grad_norm": 0.296875,
"learning_rate": 6.054468156865955e-07,
"loss": 0.0479,
"step": 2800
},
{
"epoch": 3.4338235294117645,
"grad_norm": 0.240234375,
"learning_rate": 6.003564701128506e-07,
"loss": 0.0497,
"step": 2802
},
{
"epoch": 3.436274509803922,
"grad_norm": 0.294921875,
"learning_rate": 5.952862466472436e-07,
"loss": 0.0459,
"step": 2804
},
{
"epoch": 3.438725490196078,
"grad_norm": 0.2890625,
"learning_rate": 5.902361684788288e-07,
"loss": 0.0424,
"step": 2806
},
{
"epoch": 3.4411764705882355,
"grad_norm": 0.3515625,
"learning_rate": 5.852062587045282e-07,
"loss": 0.0467,
"step": 2808
},
{
"epoch": 3.443627450980392,
"grad_norm": 0.2255859375,
"learning_rate": 5.801965403290221e-07,
"loss": 0.044,
"step": 2810
},
{
"epoch": 3.446078431372549,
"grad_norm": 0.30859375,
"learning_rate": 5.752070362646417e-07,
"loss": 0.0405,
"step": 2812
},
{
"epoch": 3.448529411764706,
"grad_norm": 0.291015625,
"learning_rate": 5.702377693312683e-07,
"loss": 0.0441,
"step": 2814
},
{
"epoch": 3.450980392156863,
"grad_norm": 0.212890625,
"learning_rate": 5.652887622562259e-07,
"loss": 0.0481,
"step": 2816
},
{
"epoch": 3.4534313725490198,
"grad_norm": 0.3828125,
"learning_rate": 5.603600376741791e-07,
"loss": 0.043,
"step": 2818
},
{
"epoch": 3.4558823529411766,
"grad_norm": 0.26171875,
"learning_rate": 5.554516181270303e-07,
"loss": 0.049,
"step": 2820
},
{
"epoch": 3.4583333333333335,
"grad_norm": 0.26171875,
"learning_rate": 5.505635260638114e-07,
"loss": 0.0463,
"step": 2822
},
{
"epoch": 3.4607843137254903,
"grad_norm": 0.236328125,
"learning_rate": 5.456957838405902e-07,
"loss": 0.0449,
"step": 2824
},
{
"epoch": 3.463235294117647,
"grad_norm": 0.28125,
"learning_rate": 5.408484137203579e-07,
"loss": 0.0448,
"step": 2826
},
{
"epoch": 3.465686274509804,
"grad_norm": 0.193359375,
"learning_rate": 5.360214378729361e-07,
"loss": 0.0445,
"step": 2828
},
{
"epoch": 3.468137254901961,
"grad_norm": 0.296875,
"learning_rate": 5.3121487837487e-07,
"loss": 0.049,
"step": 2830
},
{
"epoch": 3.4705882352941178,
"grad_norm": 0.31640625,
"learning_rate": 5.264287572093285e-07,
"loss": 0.0529,
"step": 2832
},
{
"epoch": 3.4730392156862746,
"grad_norm": 0.2265625,
"learning_rate": 5.216630962660057e-07,
"loss": 0.0471,
"step": 2834
},
{
"epoch": 3.4754901960784315,
"grad_norm": 0.263671875,
"learning_rate": 5.169179173410178e-07,
"loss": 0.046,
"step": 2836
},
{
"epoch": 3.4779411764705883,
"grad_norm": 0.232421875,
"learning_rate": 5.121932421368059e-07,
"loss": 0.0478,
"step": 2838
},
{
"epoch": 3.480392156862745,
"grad_norm": 0.30078125,
"learning_rate": 5.074890922620368e-07,
"loss": 0.0465,
"step": 2840
},
{
"epoch": 3.482843137254902,
"grad_norm": 0.296875,
"learning_rate": 5.028054892314993e-07,
"loss": 0.0407,
"step": 2842
},
{
"epoch": 3.485294117647059,
"grad_norm": 0.2275390625,
"learning_rate": 4.981424544660147e-07,
"loss": 0.042,
"step": 2844
},
{
"epoch": 3.4877450980392157,
"grad_norm": 0.22265625,
"learning_rate": 4.935000092923292e-07,
"loss": 0.0363,
"step": 2846
},
{
"epoch": 3.4901960784313726,
"grad_norm": 0.2890625,
"learning_rate": 4.888781749430238e-07,
"loss": 0.0441,
"step": 2848
},
{
"epoch": 3.4926470588235294,
"grad_norm": 0.2158203125,
"learning_rate": 4.842769725564139e-07,
"loss": 0.0431,
"step": 2850
},
{
"epoch": 3.4950980392156863,
"grad_norm": 0.232421875,
"learning_rate": 4.796964231764517e-07,
"loss": 0.0437,
"step": 2852
},
{
"epoch": 3.497549019607843,
"grad_norm": 0.30078125,
"learning_rate": 4.7513654775263316e-07,
"loss": 0.048,
"step": 2854
},
{
"epoch": 3.5,
"grad_norm": 0.26953125,
"learning_rate": 4.705973671398978e-07,
"loss": 0.0453,
"step": 2856
},
{
"epoch": 3.502450980392157,
"grad_norm": 0.2890625,
"learning_rate": 4.6607890209853846e-07,
"loss": 0.0471,
"step": 2858
},
{
"epoch": 3.5049019607843137,
"grad_norm": 0.240234375,
"learning_rate": 4.6158117329410334e-07,
"loss": 0.0469,
"step": 2860
},
{
"epoch": 3.5073529411764706,
"grad_norm": 0.234375,
"learning_rate": 4.571042012972993e-07,
"loss": 0.0469,
"step": 2862
},
{
"epoch": 3.5098039215686274,
"grad_norm": 0.259765625,
"learning_rate": 4.5264800658390284e-07,
"loss": 0.0391,
"step": 2864
},
{
"epoch": 3.5122549019607843,
"grad_norm": 0.23046875,
"learning_rate": 4.482126095346645e-07,
"loss": 0.0452,
"step": 2866
},
{
"epoch": 3.514705882352941,
"grad_norm": 0.2197265625,
"learning_rate": 4.4379803043521187e-07,
"loss": 0.0434,
"step": 2868
},
{
"epoch": 3.517156862745098,
"grad_norm": 0.2236328125,
"learning_rate": 4.394042894759626e-07,
"loss": 0.049,
"step": 2870
},
{
"epoch": 3.519607843137255,
"grad_norm": 0.244140625,
"learning_rate": 4.3503140675203e-07,
"loss": 0.0464,
"step": 2872
},
{
"epoch": 3.5220588235294117,
"grad_norm": 0.287109375,
"learning_rate": 4.3067940226313e-07,
"loss": 0.0451,
"step": 2874
},
{
"epoch": 3.5245098039215685,
"grad_norm": 0.240234375,
"learning_rate": 4.2634829591348793e-07,
"loss": 0.0423,
"step": 2876
},
{
"epoch": 3.5269607843137254,
"grad_norm": 0.22265625,
"learning_rate": 4.220381075117524e-07,
"loss": 0.0436,
"step": 2878
},
{
"epoch": 3.5294117647058822,
"grad_norm": 0.271484375,
"learning_rate": 4.177488567709037e-07,
"loss": 0.0509,
"step": 2880
},
{
"epoch": 3.531862745098039,
"grad_norm": 0.2080078125,
"learning_rate": 4.1348056330815736e-07,
"loss": 0.0435,
"step": 2882
},
{
"epoch": 3.534313725490196,
"grad_norm": 0.216796875,
"learning_rate": 4.0923324664488286e-07,
"loss": 0.0438,
"step": 2884
},
{
"epoch": 3.536764705882353,
"grad_norm": 0.216796875,
"learning_rate": 4.0500692620651094e-07,
"loss": 0.0446,
"step": 2886
},
{
"epoch": 3.5392156862745097,
"grad_norm": 0.25390625,
"learning_rate": 4.008016213224408e-07,
"loss": 0.0492,
"step": 2888
},
{
"epoch": 3.5416666666666665,
"grad_norm": 0.32421875,
"learning_rate": 3.966173512259602e-07,
"loss": 0.0456,
"step": 2890
},
{
"epoch": 3.5441176470588234,
"grad_norm": 0.2138671875,
"learning_rate": 3.9245413505414887e-07,
"loss": 0.0482,
"step": 2892
},
{
"epoch": 3.5465686274509802,
"grad_norm": 0.275390625,
"learning_rate": 3.883119918477962e-07,
"loss": 0.0473,
"step": 2894
},
{
"epoch": 3.549019607843137,
"grad_norm": 0.2275390625,
"learning_rate": 3.841909405513139e-07,
"loss": 0.0423,
"step": 2896
},
{
"epoch": 3.5514705882352944,
"grad_norm": 0.26953125,
"learning_rate": 3.800910000126462e-07,
"loss": 0.0411,
"step": 2898
},
{
"epoch": 3.553921568627451,
"grad_norm": 0.28515625,
"learning_rate": 3.760121889831869e-07,
"loss": 0.049,
"step": 2900
},
{
"epoch": 3.556372549019608,
"grad_norm": 0.2734375,
"learning_rate": 3.719545261176916e-07,
"loss": 0.0478,
"step": 2902
},
{
"epoch": 3.5588235294117645,
"grad_norm": 0.345703125,
"learning_rate": 3.679180299741936e-07,
"loss": 0.0467,
"step": 2904
},
{
"epoch": 3.561274509803922,
"grad_norm": 0.322265625,
"learning_rate": 3.639027190139194e-07,
"loss": 0.0534,
"step": 2906
},
{
"epoch": 3.563725490196078,
"grad_norm": 0.259765625,
"learning_rate": 3.5990861160120126e-07,
"loss": 0.048,
"step": 2908
},
{
"epoch": 3.5661764705882355,
"grad_norm": 0.265625,
"learning_rate": 3.5593572600339856e-07,
"loss": 0.046,
"step": 2910
},
{
"epoch": 3.568627450980392,
"grad_norm": 0.201171875,
"learning_rate": 3.519840803908064e-07,
"loss": 0.0453,
"step": 2912
},
{
"epoch": 3.571078431372549,
"grad_norm": 0.255859375,
"learning_rate": 3.480536928365824e-07,
"loss": 0.049,
"step": 2914
},
{
"epoch": 3.5735294117647056,
"grad_norm": 0.220703125,
"learning_rate": 3.4414458131665607e-07,
"loss": 0.0457,
"step": 2916
},
{
"epoch": 3.575980392156863,
"grad_norm": 0.291015625,
"learning_rate": 3.4025676370964824e-07,
"loss": 0.0569,
"step": 2918
},
{
"epoch": 3.5784313725490198,
"grad_norm": 0.236328125,
"learning_rate": 3.3639025779679367e-07,
"loss": 0.0448,
"step": 2920
},
{
"epoch": 3.5808823529411766,
"grad_norm": 0.2158203125,
"learning_rate": 3.3254508126185303e-07,
"loss": 0.0439,
"step": 2922
},
{
"epoch": 3.5833333333333335,
"grad_norm": 0.271484375,
"learning_rate": 3.287212516910371e-07,
"loss": 0.0449,
"step": 2924
},
{
"epoch": 3.5857843137254903,
"grad_norm": 0.271484375,
"learning_rate": 3.2491878657292643e-07,
"loss": 0.0428,
"step": 2926
},
{
"epoch": 3.588235294117647,
"grad_norm": 0.2890625,
"learning_rate": 3.2113770329838536e-07,
"loss": 0.048,
"step": 2928
},
{
"epoch": 3.590686274509804,
"grad_norm": 0.244140625,
"learning_rate": 3.1737801916049026e-07,
"loss": 0.0472,
"step": 2930
},
{
"epoch": 3.593137254901961,
"grad_norm": 0.255859375,
"learning_rate": 3.136397513544465e-07,
"loss": 0.0476,
"step": 2932
},
{
"epoch": 3.5955882352941178,
"grad_norm": 0.25390625,
"learning_rate": 3.099229169775081e-07,
"loss": 0.0396,
"step": 2934
},
{
"epoch": 3.5980392156862746,
"grad_norm": 0.2216796875,
"learning_rate": 3.0622753302890384e-07,
"loss": 0.0495,
"step": 2936
},
{
"epoch": 3.6004901960784315,
"grad_norm": 0.2333984375,
"learning_rate": 3.025536164097559e-07,
"loss": 0.0459,
"step": 2938
},
{
"epoch": 3.6029411764705883,
"grad_norm": 0.234375,
"learning_rate": 2.9890118392300493e-07,
"loss": 0.0484,
"step": 2940
},
{
"epoch": 3.605392156862745,
"grad_norm": 0.2490234375,
"learning_rate": 2.9527025227333196e-07,
"loss": 0.0445,
"step": 2942
},
{
"epoch": 3.607843137254902,
"grad_norm": 0.2734375,
"learning_rate": 2.9166083806708154e-07,
"loss": 0.052,
"step": 2944
},
{
"epoch": 3.610294117647059,
"grad_norm": 0.1875,
"learning_rate": 2.880729578121888e-07,
"loss": 0.0398,
"step": 2946
},
{
"epoch": 3.6127450980392157,
"grad_norm": 0.263671875,
"learning_rate": 2.84506627918098e-07,
"loss": 0.0438,
"step": 2948
},
{
"epoch": 3.6151960784313726,
"grad_norm": 0.306640625,
"learning_rate": 2.8096186469569466e-07,
"loss": 0.0472,
"step": 2950
},
{
"epoch": 3.6176470588235294,
"grad_norm": 0.259765625,
"learning_rate": 2.774386843572263e-07,
"loss": 0.0434,
"step": 2952
},
{
"epoch": 3.6200980392156863,
"grad_norm": 0.23046875,
"learning_rate": 2.7393710301622746e-07,
"loss": 0.0414,
"step": 2954
},
{
"epoch": 3.622549019607843,
"grad_norm": 0.271484375,
"learning_rate": 2.70457136687452e-07,
"loss": 0.0467,
"step": 2956
},
{
"epoch": 3.625,
"grad_norm": 0.2255859375,
"learning_rate": 2.669988012867919e-07,
"loss": 0.0484,
"step": 2958
},
{
"epoch": 3.627450980392157,
"grad_norm": 0.244140625,
"learning_rate": 2.6356211263121043e-07,
"loss": 0.0468,
"step": 2960
},
{
"epoch": 3.6299019607843137,
"grad_norm": 0.21875,
"learning_rate": 2.60147086438669e-07,
"loss": 0.0488,
"step": 2962
},
{
"epoch": 3.6323529411764706,
"grad_norm": 0.2734375,
"learning_rate": 2.567537383280505e-07,
"loss": 0.0423,
"step": 2964
},
{
"epoch": 3.6348039215686274,
"grad_norm": 0.248046875,
"learning_rate": 2.533820838190959e-07,
"loss": 0.0446,
"step": 2966
},
{
"epoch": 3.6372549019607843,
"grad_norm": 0.33984375,
"learning_rate": 2.50032138332324e-07,
"loss": 0.0455,
"step": 2968
},
{
"epoch": 3.639705882352941,
"grad_norm": 0.26171875,
"learning_rate": 2.46703917188969e-07,
"loss": 0.0479,
"step": 2970
},
{
"epoch": 3.642156862745098,
"grad_norm": 0.21484375,
"learning_rate": 2.433974356109081e-07,
"loss": 0.0433,
"step": 2972
},
{
"epoch": 3.644607843137255,
"grad_norm": 0.255859375,
"learning_rate": 2.4011270872058625e-07,
"loss": 0.0509,
"step": 2974
},
{
"epoch": 3.6470588235294117,
"grad_norm": 0.244140625,
"learning_rate": 2.3684975154095625e-07,
"loss": 0.0399,
"step": 2976
},
{
"epoch": 3.6495098039215685,
"grad_norm": 0.251953125,
"learning_rate": 2.3360857899540457e-07,
"loss": 0.0507,
"step": 2978
},
{
"epoch": 3.6519607843137254,
"grad_norm": 0.244140625,
"learning_rate": 2.303892059076812e-07,
"loss": 0.0466,
"step": 2980
},
{
"epoch": 3.6544117647058822,
"grad_norm": 0.2412109375,
"learning_rate": 2.271916470018376e-07,
"loss": 0.0471,
"step": 2982
},
{
"epoch": 3.656862745098039,
"grad_norm": 0.2451171875,
"learning_rate": 2.2401591690215663e-07,
"loss": 0.0473,
"step": 2984
},
{
"epoch": 3.659313725490196,
"grad_norm": 0.251953125,
"learning_rate": 2.208620301330816e-07,
"loss": 0.0417,
"step": 2986
},
{
"epoch": 3.661764705882353,
"grad_norm": 0.216796875,
"learning_rate": 2.1773000111915798e-07,
"loss": 0.045,
"step": 2988
},
{
"epoch": 3.6642156862745097,
"grad_norm": 0.296875,
"learning_rate": 2.1461984418496062e-07,
"loss": 0.0524,
"step": 2990
},
{
"epoch": 3.6666666666666665,
"grad_norm": 0.283203125,
"learning_rate": 2.1153157355503274e-07,
"loss": 0.0443,
"step": 2992
},
{
"epoch": 3.6691176470588234,
"grad_norm": 0.279296875,
"learning_rate": 2.084652033538165e-07,
"loss": 0.046,
"step": 2994
},
{
"epoch": 3.6715686274509802,
"grad_norm": 0.28125,
"learning_rate": 2.0542074760559193e-07,
"loss": 0.0435,
"step": 2996
},
{
"epoch": 3.674019607843137,
"grad_norm": 0.24609375,
"learning_rate": 2.023982202344127e-07,
"loss": 0.0422,
"step": 2998
},
{
"epoch": 3.6764705882352944,
"grad_norm": 0.251953125,
"learning_rate": 1.993976350640392e-07,
"loss": 0.0452,
"step": 3000
},
{
"epoch": 3.678921568627451,
"grad_norm": 0.2412109375,
"learning_rate": 1.9641900581787887e-07,
"loss": 0.047,
"step": 3002
},
{
"epoch": 3.681372549019608,
"grad_norm": 0.24609375,
"learning_rate": 1.9346234611892112e-07,
"loss": 0.0523,
"step": 3004
},
{
"epoch": 3.6838235294117645,
"grad_norm": 0.1982421875,
"learning_rate": 1.9052766948967738e-07,
"loss": 0.0414,
"step": 3006
},
{
"epoch": 3.686274509803922,
"grad_norm": 0.2333984375,
"learning_rate": 1.8761498935211675e-07,
"loss": 0.0447,
"step": 3008
},
{
"epoch": 3.688725490196078,
"grad_norm": 0.1943359375,
"learning_rate": 1.8472431902760546e-07,
"loss": 0.0414,
"step": 3010
},
{
"epoch": 3.6911764705882355,
"grad_norm": 0.26953125,
"learning_rate": 1.818556717368475e-07,
"loss": 0.0443,
"step": 3012
},
{
"epoch": 3.693627450980392,
"grad_norm": 0.2431640625,
"learning_rate": 1.790090605998207e-07,
"loss": 0.0443,
"step": 3014
},
{
"epoch": 3.696078431372549,
"grad_norm": 0.294921875,
"learning_rate": 1.7618449863572085e-07,
"loss": 0.0461,
"step": 3016
},
{
"epoch": 3.6985294117647056,
"grad_norm": 0.23828125,
"learning_rate": 1.7338199876289984e-07,
"loss": 0.0461,
"step": 3018
},
{
"epoch": 3.700980392156863,
"grad_norm": 0.26171875,
"learning_rate": 1.7060157379880537e-07,
"loss": 0.0517,
"step": 3020
},
{
"epoch": 3.7034313725490198,
"grad_norm": 0.23046875,
"learning_rate": 1.6784323645992585e-07,
"loss": 0.046,
"step": 3022
},
{
"epoch": 3.7058823529411766,
"grad_norm": 0.26171875,
"learning_rate": 1.6510699936172825e-07,
"loss": 0.0474,
"step": 3024
},
{
"epoch": 3.7083333333333335,
"grad_norm": 0.333984375,
"learning_rate": 1.623928750186038e-07,
"loss": 0.0418,
"step": 3026
},
{
"epoch": 3.7107843137254903,
"grad_norm": 0.33203125,
"learning_rate": 1.597008758438101e-07,
"loss": 0.0469,
"step": 3028
},
{
"epoch": 3.713235294117647,
"grad_norm": 0.30078125,
"learning_rate": 1.570310141494108e-07,
"loss": 0.0436,
"step": 3030
},
{
"epoch": 3.715686274509804,
"grad_norm": 0.30078125,
"learning_rate": 1.5438330214622544e-07,
"loss": 0.0484,
"step": 3032
},
{
"epoch": 3.718137254901961,
"grad_norm": 0.2470703125,
"learning_rate": 1.5175775194376686e-07,
"loss": 0.0485,
"step": 3034
},
{
"epoch": 3.7205882352941178,
"grad_norm": 0.224609375,
"learning_rate": 1.491543755501912e-07,
"loss": 0.0445,
"step": 3036
},
{
"epoch": 3.7230392156862746,
"grad_norm": 0.236328125,
"learning_rate": 1.4657318487224182e-07,
"loss": 0.0413,
"step": 3038
},
{
"epoch": 3.7254901960784315,
"grad_norm": 0.27734375,
"learning_rate": 1.4401419171519215e-07,
"loss": 0.0457,
"step": 3040
},
{
"epoch": 3.7279411764705883,
"grad_norm": 0.2421875,
"learning_rate": 1.4147740778279406e-07,
"loss": 0.0426,
"step": 3042
},
{
"epoch": 3.730392156862745,
"grad_norm": 0.26171875,
"learning_rate": 1.3896284467722398e-07,
"loss": 0.046,
"step": 3044
},
{
"epoch": 3.732843137254902,
"grad_norm": 0.2294921875,
"learning_rate": 1.3647051389903022e-07,
"loss": 0.0459,
"step": 3046
},
{
"epoch": 3.735294117647059,
"grad_norm": 0.275390625,
"learning_rate": 1.3400042684707902e-07,
"loss": 0.0494,
"step": 3048
},
{
"epoch": 3.7377450980392157,
"grad_norm": 0.326171875,
"learning_rate": 1.3155259481850313e-07,
"loss": 0.0492,
"step": 3050
},
{
"epoch": 3.7401960784313726,
"grad_norm": 0.265625,
"learning_rate": 1.2912702900865158e-07,
"loss": 0.0437,
"step": 3052
},
{
"epoch": 3.7426470588235294,
"grad_norm": 0.228515625,
"learning_rate": 1.2672374051103554e-07,
"loss": 0.0456,
"step": 3054
},
{
"epoch": 3.7450980392156863,
"grad_norm": 0.2314453125,
"learning_rate": 1.2434274031727922e-07,
"loss": 0.0495,
"step": 3056
},
{
"epoch": 3.747549019607843,
"grad_norm": 0.24609375,
"learning_rate": 1.219840393170718e-07,
"loss": 0.0474,
"step": 3058
},
{
"epoch": 3.75,
"grad_norm": 0.2255859375,
"learning_rate": 1.1964764829811238e-07,
"loss": 0.0424,
"step": 3060
},
{
"epoch": 3.752450980392157,
"grad_norm": 0.2578125,
"learning_rate": 1.1733357794606548e-07,
"loss": 0.0458,
"step": 3062
},
{
"epoch": 3.7549019607843137,
"grad_norm": 0.2333984375,
"learning_rate": 1.1504183884451014e-07,
"loss": 0.0474,
"step": 3064
},
{
"epoch": 3.7573529411764706,
"grad_norm": 0.35546875,
"learning_rate": 1.1277244147489096e-07,
"loss": 0.054,
"step": 3066
},
{
"epoch": 3.7598039215686274,
"grad_norm": 0.2119140625,
"learning_rate": 1.1052539621647207e-07,
"loss": 0.0417,
"step": 3068
},
{
"epoch": 3.7622549019607843,
"grad_norm": 0.388671875,
"learning_rate": 1.0830071334628655e-07,
"loss": 0.0426,
"step": 3070
},
{
"epoch": 3.764705882352941,
"grad_norm": 0.22265625,
"learning_rate": 1.060984030390938e-07,
"loss": 0.0463,
"step": 3072
},
{
"epoch": 3.767156862745098,
"grad_norm": 0.21484375,
"learning_rate": 1.0391847536732946e-07,
"loss": 0.0468,
"step": 3074
},
{
"epoch": 3.769607843137255,
"grad_norm": 0.228515625,
"learning_rate": 1.0176094030105943e-07,
"loss": 0.0425,
"step": 3076
},
{
"epoch": 3.7720588235294117,
"grad_norm": 0.251953125,
"learning_rate": 9.962580770793706e-08,
"loss": 0.0387,
"step": 3078
},
{
"epoch": 3.7745098039215685,
"grad_norm": 0.25,
"learning_rate": 9.751308735315545e-08,
"loss": 0.0466,
"step": 3080
},
{
"epoch": 3.7769607843137254,
"grad_norm": 0.26171875,
"learning_rate": 9.54227888994036e-08,
"loss": 0.0508,
"step": 3082
},
{
"epoch": 3.7794117647058822,
"grad_norm": 0.291015625,
"learning_rate": 9.33549219068225e-08,
"loss": 0.0519,
"step": 3084
},
{
"epoch": 3.781862745098039,
"grad_norm": 0.26171875,
"learning_rate": 9.130949583296023e-08,
"loss": 0.0417,
"step": 3086
},
{
"epoch": 3.784313725490196,
"grad_norm": 0.306640625,
"learning_rate": 8.928652003273087e-08,
"loss": 0.0451,
"step": 3088
},
{
"epoch": 3.786764705882353,
"grad_norm": 0.26953125,
"learning_rate": 8.728600375837004e-08,
"loss": 0.046,
"step": 3090
},
{
"epoch": 3.7892156862745097,
"grad_norm": 0.2216796875,
"learning_rate": 8.530795615939225e-08,
"loss": 0.0453,
"step": 3092
},
{
"epoch": 3.7916666666666665,
"grad_norm": 0.2333984375,
"learning_rate": 8.335238628255027e-08,
"loss": 0.0423,
"step": 3094
},
{
"epoch": 3.7941176470588234,
"grad_norm": 0.2421875,
"learning_rate": 8.141930307179468e-08,
"loss": 0.045,
"step": 3096
},
{
"epoch": 3.7965686274509802,
"grad_norm": 0.220703125,
"learning_rate": 7.950871536822891e-08,
"loss": 0.0395,
"step": 3098
},
{
"epoch": 3.799019607843137,
"grad_norm": 0.26953125,
"learning_rate": 7.762063191007252e-08,
"loss": 0.0406,
"step": 3100
},
{
"epoch": 3.8014705882352944,
"grad_norm": 0.2373046875,
"learning_rate": 7.57550613326219e-08,
"loss": 0.0473,
"step": 3102
},
{
"epoch": 3.803921568627451,
"grad_norm": 0.2255859375,
"learning_rate": 7.391201216820687e-08,
"loss": 0.0494,
"step": 3104
},
{
"epoch": 3.806372549019608,
"grad_norm": 0.2734375,
"learning_rate": 7.209149284615413e-08,
"loss": 0.0496,
"step": 3106
},
{
"epoch": 3.8088235294117645,
"grad_norm": 0.26953125,
"learning_rate": 7.029351169274945e-08,
"loss": 0.0451,
"step": 3108
},
{
"epoch": 3.811274509803922,
"grad_norm": 0.232421875,
"learning_rate": 6.851807693119772e-08,
"loss": 0.0415,
"step": 3110
},
{
"epoch": 3.813725490196078,
"grad_norm": 0.2392578125,
"learning_rate": 6.676519668158576e-08,
"loss": 0.0485,
"step": 3112
},
{
"epoch": 3.8161764705882355,
"grad_norm": 0.396484375,
"learning_rate": 6.503487896084792e-08,
"loss": 0.05,
"step": 3114
},
{
"epoch": 3.818627450980392,
"grad_norm": 0.232421875,
"learning_rate": 6.332713168272386e-08,
"loss": 0.0435,
"step": 3116
},
{
"epoch": 3.821078431372549,
"grad_norm": 0.216796875,
"learning_rate": 6.164196265772804e-08,
"loss": 0.0457,
"step": 3118
},
{
"epoch": 3.8235294117647056,
"grad_norm": 0.2353515625,
"learning_rate": 5.997937959310974e-08,
"loss": 0.0492,
"step": 3120
},
{
"epoch": 3.825980392156863,
"grad_norm": 0.22265625,
"learning_rate": 5.833939009282086e-08,
"loss": 0.0438,
"step": 3122
},
{
"epoch": 3.8284313725490198,
"grad_norm": 0.2470703125,
"learning_rate": 5.672200165748043e-08,
"loss": 0.0431,
"step": 3124
},
{
"epoch": 3.8308823529411766,
"grad_norm": 0.3046875,
"learning_rate": 5.5127221684337376e-08,
"loss": 0.0474,
"step": 3126
},
{
"epoch": 3.8333333333333335,
"grad_norm": 0.2265625,
"learning_rate": 5.355505746724166e-08,
"loss": 0.0427,
"step": 3128
},
{
"epoch": 3.8357843137254903,
"grad_norm": 0.298828125,
"learning_rate": 5.2005516196607094e-08,
"loss": 0.0432,
"step": 3130
},
{
"epoch": 3.838235294117647,
"grad_norm": 0.255859375,
"learning_rate": 5.047860495937917e-08,
"loss": 0.0435,
"step": 3132
},
{
"epoch": 3.840686274509804,
"grad_norm": 0.2490234375,
"learning_rate": 4.897433073900504e-08,
"loss": 0.0456,
"step": 3134
},
{
"epoch": 3.843137254901961,
"grad_norm": 0.2314453125,
"learning_rate": 4.7492700415398e-08,
"loss": 0.0435,
"step": 3136
},
{
"epoch": 3.8455882352941178,
"grad_norm": 0.1865234375,
"learning_rate": 4.6033720764908084e-08,
"loss": 0.0429,
"step": 3138
},
{
"epoch": 3.8480392156862746,
"grad_norm": 0.314453125,
"learning_rate": 4.4597398460290985e-08,
"loss": 0.0475,
"step": 3140
},
{
"epoch": 3.8504901960784315,
"grad_norm": 0.306640625,
"learning_rate": 4.3183740070677496e-08,
"loss": 0.0421,
"step": 3142
},
{
"epoch": 3.8529411764705883,
"grad_norm": 0.2060546875,
"learning_rate": 4.179275206154354e-08,
"loss": 0.0397,
"step": 3144
},
{
"epoch": 3.855392156862745,
"grad_norm": 0.26953125,
"learning_rate": 4.0424440794678555e-08,
"loss": 0.0476,
"step": 3146
},
{
"epoch": 3.857843137254902,
"grad_norm": 0.271484375,
"learning_rate": 3.907881252816048e-08,
"loss": 0.0498,
"step": 3148
},
{
"epoch": 3.860294117647059,
"grad_norm": 0.244140625,
"learning_rate": 3.775587341632303e-08,
"loss": 0.0447,
"step": 3150
},
{
"epoch": 3.8627450980392157,
"grad_norm": 0.21484375,
"learning_rate": 3.645562950973014e-08,
"loss": 0.0439,
"step": 3152
},
{
"epoch": 3.8651960784313726,
"grad_norm": 0.240234375,
"learning_rate": 3.517808675514767e-08,
"loss": 0.0425,
"step": 3154
},
{
"epoch": 3.8676470588235294,
"grad_norm": 0.2197265625,
"learning_rate": 3.3923250995515076e-08,
"loss": 0.0436,
"step": 3156
},
{
"epoch": 3.8700980392156863,
"grad_norm": 0.2734375,
"learning_rate": 3.269112796992102e-08,
"loss": 0.0442,
"step": 3158
},
{
"epoch": 3.872549019607843,
"grad_norm": 0.2578125,
"learning_rate": 3.148172331357447e-08,
"loss": 0.044,
"step": 3160
},
{
"epoch": 3.875,
"grad_norm": 0.251953125,
"learning_rate": 3.029504255778026e-08,
"loss": 0.0444,
"step": 3162
},
{
"epoch": 3.877450980392157,
"grad_norm": 0.328125,
"learning_rate": 2.91310911299153e-08,
"loss": 0.0452,
"step": 3164
},
{
"epoch": 3.8799019607843137,
"grad_norm": 0.240234375,
"learning_rate": 2.7989874353399616e-08,
"loss": 0.0513,
"step": 3166
},
{
"epoch": 3.8823529411764706,
"grad_norm": 0.234375,
"learning_rate": 2.6871397447676418e-08,
"loss": 0.0445,
"step": 3168
},
{
"epoch": 3.8848039215686274,
"grad_norm": 0.255859375,
"learning_rate": 2.5775665528186556e-08,
"loss": 0.0434,
"step": 3170
},
{
"epoch": 3.8872549019607843,
"grad_norm": 0.2470703125,
"learning_rate": 2.4702683606343537e-08,
"loss": 0.0472,
"step": 3172
},
{
"epoch": 3.889705882352941,
"grad_norm": 0.2578125,
"learning_rate": 2.3652456589512983e-08,
"loss": 0.0452,
"step": 3174
},
{
"epoch": 3.892156862745098,
"grad_norm": 0.2373046875,
"learning_rate": 2.2624989280988773e-08,
"loss": 0.044,
"step": 3176
},
{
"epoch": 3.894607843137255,
"grad_norm": 0.275390625,
"learning_rate": 2.1620286379971932e-08,
"loss": 0.0491,
"step": 3178
},
{
"epoch": 3.8970588235294117,
"grad_norm": 0.33984375,
"learning_rate": 2.063835248154844e-08,
"loss": 0.0424,
"step": 3180
},
{
"epoch": 3.8995098039215685,
"grad_norm": 0.265625,
"learning_rate": 1.967919207666813e-08,
"loss": 0.0477,
"step": 3182
},
{
"epoch": 3.9019607843137254,
"grad_norm": 0.228515625,
"learning_rate": 1.8742809552125264e-08,
"loss": 0.0501,
"step": 3184
},
{
"epoch": 3.9044117647058822,
"grad_norm": 0.275390625,
"learning_rate": 1.782920919053688e-08,
"loss": 0.0454,
"step": 3186
},
{
"epoch": 3.906862745098039,
"grad_norm": 0.2734375,
"learning_rate": 1.6938395170325583e-08,
"loss": 0.0468,
"step": 3188
},
{
"epoch": 3.909313725490196,
"grad_norm": 0.23046875,
"learning_rate": 1.6070371565696797e-08,
"loss": 0.0472,
"step": 3190
},
{
"epoch": 3.911764705882353,
"grad_norm": 0.287109375,
"learning_rate": 1.52251423466232e-08,
"loss": 0.0459,
"step": 3192
},
{
"epoch": 3.9142156862745097,
"grad_norm": 0.251953125,
"learning_rate": 1.4402711378826429e-08,
"loss": 0.0496,
"step": 3194
},
{
"epoch": 3.9166666666666665,
"grad_norm": 0.310546875,
"learning_rate": 1.3603082423757629e-08,
"loss": 0.0459,
"step": 3196
},
{
"epoch": 3.9191176470588234,
"grad_norm": 0.275390625,
"learning_rate": 1.2826259138580266e-08,
"loss": 0.0461,
"step": 3198
},
{
"epoch": 3.9215686274509802,
"grad_norm": 0.267578125,
"learning_rate": 1.2072245076156786e-08,
"loss": 0.0413,
"step": 3200
},
{
"epoch": 3.924019607843137,
"grad_norm": 0.2734375,
"learning_rate": 1.1341043685028087e-08,
"loss": 0.0516,
"step": 3202
},
{
"epoch": 3.9264705882352944,
"grad_norm": 0.318359375,
"learning_rate": 1.0632658309399636e-08,
"loss": 0.0474,
"step": 3204
},
{
"epoch": 3.928921568627451,
"grad_norm": 0.310546875,
"learning_rate": 9.947092189126483e-09,
"loss": 0.0498,
"step": 3206
},
{
"epoch": 3.931372549019608,
"grad_norm": 0.29296875,
"learning_rate": 9.284348459698277e-09,
"loss": 0.0495,
"step": 3208
},
{
"epoch": 3.9338235294117645,
"grad_norm": 0.26171875,
"learning_rate": 8.64443015222427e-09,
"loss": 0.0506,
"step": 3210
},
{
"epoch": 3.936274509803922,
"grad_norm": 0.2392578125,
"learning_rate": 8.027340193420551e-09,
"loss": 0.0418,
"step": 3212
},
{
"epoch": 3.938725490196078,
"grad_norm": 0.2890625,
"learning_rate": 7.433081405594511e-09,
"loss": 0.0448,
"step": 3214
},
{
"epoch": 3.9411764705882355,
"grad_norm": 0.2138671875,
"learning_rate": 6.861656506635395e-09,
"loss": 0.0452,
"step": 3216
},
{
"epoch": 3.943627450980392,
"grad_norm": 0.244140625,
"learning_rate": 6.313068109998766e-09,
"loss": 0.0469,
"step": 3218
},
{
"epoch": 3.946078431372549,
"grad_norm": 0.2490234375,
"learning_rate": 5.787318724695401e-09,
"loss": 0.048,
"step": 3220
},
{
"epoch": 3.9485294117647056,
"grad_norm": 0.26953125,
"learning_rate": 5.284410755280745e-09,
"loss": 0.0449,
"step": 3222
},
{
"epoch": 3.950980392156863,
"grad_norm": 0.2890625,
"learning_rate": 4.804346501843249e-09,
"loss": 0.0448,
"step": 3224
},
{
"epoch": 3.9534313725490198,
"grad_norm": 0.25,
"learning_rate": 4.347128159993829e-09,
"loss": 0.0453,
"step": 3226
},
{
"epoch": 3.9558823529411766,
"grad_norm": 0.3125,
"learning_rate": 3.912757820854762e-09,
"loss": 0.0547,
"step": 3228
},
{
"epoch": 3.9583333333333335,
"grad_norm": 0.2578125,
"learning_rate": 3.5012374710535757e-09,
"loss": 0.0457,
"step": 3230
},
{
"epoch": 3.9607843137254903,
"grad_norm": 0.25,
"learning_rate": 3.1125689927097303e-09,
"loss": 0.0473,
"step": 3232
},
{
"epoch": 3.963235294117647,
"grad_norm": 0.2333984375,
"learning_rate": 2.746754163428511e-09,
"loss": 0.0485,
"step": 3234
},
{
"epoch": 3.965686274509804,
"grad_norm": 0.271484375,
"learning_rate": 2.4037946562915914e-09,
"loss": 0.0453,
"step": 3236
},
{
"epoch": 3.968137254901961,
"grad_norm": 0.33203125,
"learning_rate": 2.083692039850926e-09,
"loss": 0.0501,
"step": 3238
},
{
"epoch": 3.9705882352941178,
"grad_norm": 0.2236328125,
"learning_rate": 1.786447778120981e-09,
"loss": 0.0467,
"step": 3240
},
{
"epoch": 3.9730392156862746,
"grad_norm": 0.232421875,
"learning_rate": 1.512063230570404e-09,
"loss": 0.0401,
"step": 3242
},
{
"epoch": 3.9754901960784315,
"grad_norm": 0.2734375,
"learning_rate": 1.2605396521192526e-09,
"loss": 0.0501,
"step": 3244
},
{
"epoch": 3.9779411764705883,
"grad_norm": 0.279296875,
"learning_rate": 1.0318781931284438e-09,
"loss": 0.0463,
"step": 3246
},
{
"epoch": 3.980392156862745,
"grad_norm": 0.220703125,
"learning_rate": 8.260798993997566e-10,
"loss": 0.0447,
"step": 3248
},
{
"epoch": 3.982843137254902,
"grad_norm": 0.302734375,
"learning_rate": 6.431457121675034e-10,
"loss": 0.0447,
"step": 3250
},
{
"epoch": 3.985294117647059,
"grad_norm": 0.2470703125,
"learning_rate": 4.830764680946453e-10,
"loss": 0.0399,
"step": 3252
},
{
"epoch": 3.9877450980392157,
"grad_norm": 0.2109375,
"learning_rate": 3.4587289927112687e-10,
"loss": 0.0445,
"step": 3254
},
{
"epoch": 3.9901960784313726,
"grad_norm": 0.310546875,
"learning_rate": 2.3153563320721385e-10,
"loss": 0.055,
"step": 3256
},
{
"epoch": 3.9926470588235294,
"grad_norm": 0.25390625,
"learning_rate": 1.4006519283349397e-10,
"loss": 0.0424,
"step": 3258
},
{
"epoch": 3.9950980392156863,
"grad_norm": 0.30078125,
"learning_rate": 7.146199649643581e-11,
"loss": 0.0524,
"step": 3260
},
{
"epoch": 3.997549019607843,
"grad_norm": 0.251953125,
"learning_rate": 2.572635795949907e-11,
"loss": 0.0502,
"step": 3262
},
{
"epoch": 4.0,
"grad_norm": 0.40234375,
"learning_rate": 2.858486396473126e-12,
"loss": 0.0443,
"step": 3264
},
{
"epoch": 4.0,
"eval_loss": 0.05356348678469658,
"eval_runtime": 37.3527,
"eval_samples_per_second": 134.288,
"eval_steps_per_second": 1.071,
"step": 3264
},
{
"epoch": 4.0,
"step": 3264,
"total_flos": 4.2537587819544576e+20,
"train_loss": 0.0519512448569431,
"train_runtime": 5570.9779,
"train_samples_per_second": 74.994,
"train_steps_per_second": 0.586
}
],
"logging_steps": 2,
"max_steps": 3264,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 816,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.2537587819544576e+20,
"train_batch_size": 128,
"trial_name": null,
"trial_params": null
}