{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 10.0,
"eval_steps": 500,
"global_step": 520,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.019230769230769232,
"grad_norm": 2.859375,
"learning_rate": 3.846153846153847e-06,
"loss": 3.0191,
"step": 1
},
{
"epoch": 0.09615384615384616,
"grad_norm": 2.25,
"learning_rate": 1.923076923076923e-05,
"loss": 3.0172,
"step": 5
},
{
"epoch": 0.19230769230769232,
"grad_norm": 1.9453125,
"learning_rate": 3.846153846153846e-05,
"loss": 2.9954,
"step": 10
},
{
"epoch": 0.28846153846153844,
"grad_norm": 3.28125,
"learning_rate": 5.769230769230769e-05,
"loss": 2.8517,
"step": 15
},
{
"epoch": 0.38461538461538464,
"grad_norm": 2.328125,
"learning_rate": 7.692307692307693e-05,
"loss": 2.6083,
"step": 20
},
{
"epoch": 0.4807692307692308,
"grad_norm": 1.296875,
"learning_rate": 9.615384615384617e-05,
"loss": 2.3838,
"step": 25
},
{
"epoch": 0.5769230769230769,
"grad_norm": 8.6875,
"learning_rate": 0.00011538461538461538,
"loss": 2.1596,
"step": 30
},
{
"epoch": 0.6730769230769231,
"grad_norm": 0.90234375,
"learning_rate": 0.00013461538461538464,
"loss": 1.9352,
"step": 35
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.9375,
"learning_rate": 0.00015384615384615385,
"loss": 1.7534,
"step": 40
},
{
"epoch": 0.8653846153846154,
"grad_norm": 0.65625,
"learning_rate": 0.0001730769230769231,
"loss": 1.6127,
"step": 45
},
{
"epoch": 0.9615384615384616,
"grad_norm": 1.0,
"learning_rate": 0.00019230769230769233,
"loss": 1.4893,
"step": 50
},
{
"epoch": 1.0,
"eval_loss": 2.595663547515869,
"eval_runtime": 0.5487,
"eval_samples_per_second": 18.225,
"eval_steps_per_second": 1.823,
"step": 52
},
{
"epoch": 1.0576923076923077,
"grad_norm": 0.37109375,
"learning_rate": 0.00019997972289848503,
"loss": 1.4008,
"step": 55
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.73828125,
"learning_rate": 0.00019985583705641418,
"loss": 1.3355,
"step": 60
},
{
"epoch": 1.25,
"grad_norm": 0.232421875,
"learning_rate": 0.00019961946980917456,
"loss": 1.2736,
"step": 65
},
{
"epoch": 1.3461538461538463,
"grad_norm": 0.443359375,
"learning_rate": 0.0001992708874098054,
"loss": 1.2402,
"step": 70
},
{
"epoch": 1.4423076923076923,
"grad_norm": 0.5,
"learning_rate": 0.0001988104825147528,
"loss": 1.2168,
"step": 75
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.58984375,
"learning_rate": 0.00019823877374156647,
"loss": 1.2052,
"step": 80
},
{
"epoch": 1.6346153846153846,
"grad_norm": 0.921875,
"learning_rate": 0.00019755640508470942,
"loss": 1.1812,
"step": 85
},
{
"epoch": 1.7307692307692308,
"grad_norm": 0.2578125,
"learning_rate": 0.00019676414519013781,
"loss": 1.1707,
"step": 90
},
{
"epoch": 1.8269230769230769,
"grad_norm": 0.82421875,
"learning_rate": 0.00019586288648946947,
"loss": 1.1671,
"step": 95
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.271484375,
"learning_rate": 0.00019485364419471454,
"loss": 1.143,
"step": 100
},
{
"epoch": 2.0,
"eval_loss": 2.5093514919281006,
"eval_runtime": 0.551,
"eval_samples_per_second": 18.149,
"eval_steps_per_second": 1.815,
"step": 104
},
{
"epoch": 2.019230769230769,
"grad_norm": 0.30078125,
"learning_rate": 0.00019373755515470254,
"loss": 1.1342,
"step": 105
},
{
"epoch": 2.1153846153846154,
"grad_norm": 0.416015625,
"learning_rate": 0.00019251587657449236,
"loss": 1.111,
"step": 110
},
{
"epoch": 2.2115384615384617,
"grad_norm": 0.3515625,
"learning_rate": 0.00019118998459920902,
"loss": 1.104,
"step": 115
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.36328125,
"learning_rate": 0.0001897613727639014,
"loss": 1.0995,
"step": 120
},
{
"epoch": 2.4038461538461537,
"grad_norm": 0.5625,
"learning_rate": 0.0001882316503111678,
"loss": 1.0938,
"step": 125
},
{
"epoch": 2.5,
"grad_norm": 0.86328125,
"learning_rate": 0.00018660254037844388,
"loss": 1.098,
"step": 130
},
{
"epoch": 2.5961538461538463,
"grad_norm": 0.54296875,
"learning_rate": 0.00018487587805699526,
"loss": 1.0819,
"step": 135
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.419921875,
"learning_rate": 0.00018305360832480117,
"loss": 1.0878,
"step": 140
},
{
"epoch": 2.7884615384615383,
"grad_norm": 0.416015625,
"learning_rate": 0.00018113778385565733,
"loss": 1.0805,
"step": 145
},
{
"epoch": 2.8846153846153846,
"grad_norm": 0.306640625,
"learning_rate": 0.0001791305627069662,
"loss": 1.0708,
"step": 150
},
{
"epoch": 2.980769230769231,
"grad_norm": 0.267578125,
"learning_rate": 0.00017703420588881946,
"loss": 1.0743,
"step": 155
},
{
"epoch": 3.0,
"eval_loss": 2.4954726696014404,
"eval_runtime": 0.693,
"eval_samples_per_second": 14.43,
"eval_steps_per_second": 1.443,
"step": 156
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.3203125,
"learning_rate": 0.00017485107481711012,
"loss": 1.0613,
"step": 160
},
{
"epoch": 3.173076923076923,
"grad_norm": 0.267578125,
"learning_rate": 0.00017258362865354426,
"loss": 1.0387,
"step": 165
},
{
"epoch": 3.269230769230769,
"grad_norm": 0.287109375,
"learning_rate": 0.00017023442153554777,
"loss": 1.0335,
"step": 170
},
{
"epoch": 3.3653846153846154,
"grad_norm": 0.361328125,
"learning_rate": 0.0001678060996991891,
"loss": 1.0387,
"step": 175
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.318359375,
"learning_rate": 0.0001653013984983585,
"loss": 1.0532,
"step": 180
},
{
"epoch": 3.5576923076923075,
"grad_norm": 0.5,
"learning_rate": 0.00016272313932356162,
"loss": 1.0507,
"step": 185
},
{
"epoch": 3.6538461538461537,
"grad_norm": 0.376953125,
"learning_rate": 0.0001600742264237979,
"loss": 1.0327,
"step": 190
},
{
"epoch": 3.75,
"grad_norm": 0.30859375,
"learning_rate": 0.0001573576436351046,
"loss": 1.0336,
"step": 195
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.63671875,
"learning_rate": 0.00015457645101945046,
"loss": 1.0366,
"step": 200
},
{
"epoch": 3.9423076923076925,
"grad_norm": 0.40625,
"learning_rate": 0.00015173378141776568,
"loss": 1.0412,
"step": 205
},
{
"epoch": 4.0,
"eval_loss": 2.4909958839416504,
"eval_runtime": 0.5333,
"eval_samples_per_second": 18.75,
"eval_steps_per_second": 1.875,
"step": 208
},
{
"epoch": 4.038461538461538,
"grad_norm": 0.322265625,
"learning_rate": 0.00014883283692099112,
"loss": 1.0274,
"step": 210
},
{
"epoch": 4.134615384615385,
"grad_norm": 0.490234375,
"learning_rate": 0.00014587688526312143,
"loss": 1.0145,
"step": 215
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.345703125,
"learning_rate": 0.00014286925614030542,
"loss": 1.0097,
"step": 220
},
{
"epoch": 4.326923076923077,
"grad_norm": 0.306640625,
"learning_rate": 0.0001398133374601501,
"loss": 1.0137,
"step": 225
},
{
"epoch": 4.423076923076923,
"grad_norm": 0.427734375,
"learning_rate": 0.00013671257152545277,
"loss": 1.0119,
"step": 230
},
{
"epoch": 4.519230769230769,
"grad_norm": 1.0625,
"learning_rate": 0.0001335704511566605,
"loss": 1.0112,
"step": 235
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.53125,
"learning_rate": 0.0001303905157574247,
"loss": 1.0102,
"step": 240
},
{
"epoch": 4.711538461538462,
"grad_norm": 0.373046875,
"learning_rate": 0.00012717634732768243,
"loss": 1.0055,
"step": 245
},
{
"epoch": 4.8076923076923075,
"grad_norm": 0.36328125,
"learning_rate": 0.0001239315664287558,
"loss": 1.0024,
"step": 250
},
{
"epoch": 4.903846153846154,
"grad_norm": 0.326171875,
"learning_rate": 0.00012065982810501404,
"loss": 1.0077,
"step": 255
},
{
"epoch": 5.0,
"grad_norm": 0.328125,
"learning_rate": 0.00011736481776669306,
"loss": 1.0117,
"step": 260
},
{
"epoch": 5.0,
"eval_loss": 2.5050010681152344,
"eval_runtime": 0.5358,
"eval_samples_per_second": 18.663,
"eval_steps_per_second": 1.866,
"step": 260
},
{
"epoch": 5.096153846153846,
"grad_norm": 0.44921875,
"learning_rate": 0.00011405024703850929,
"loss": 0.9883,
"step": 265
},
{
"epoch": 5.1923076923076925,
"grad_norm": 0.478515625,
"learning_rate": 0.00011071984957874479,
"loss": 0.9915,
"step": 270
},
{
"epoch": 5.288461538461538,
"grad_norm": 0.45703125,
"learning_rate": 0.00010737737687351284,
"loss": 0.9871,
"step": 275
},
{
"epoch": 5.384615384615385,
"grad_norm": 0.3828125,
"learning_rate": 0.00010402659401094152,
"loss": 0.9854,
"step": 280
},
{
"epoch": 5.480769230769231,
"grad_norm": 0.36328125,
"learning_rate": 0.00010067127544003563,
"loss": 0.9884,
"step": 285
},
{
"epoch": 5.576923076923077,
"grad_norm": 0.380859375,
"learning_rate": 9.73152007189939e-05,
"loss": 0.9908,
"step": 290
},
{
"epoch": 5.673076923076923,
"grad_norm": 0.33203125,
"learning_rate": 9.396215025777139e-05,
"loss": 0.9891,
"step": 295
},
{
"epoch": 5.769230769230769,
"grad_norm": 0.30078125,
"learning_rate": 9.061590105968208e-05,
"loss": 0.982,
"step": 300
},
{
"epoch": 5.865384615384615,
"grad_norm": 0.404296875,
"learning_rate": 8.728022246683894e-05,
"loss": 0.9928,
"step": 305
},
{
"epoch": 5.961538461538462,
"grad_norm": 0.369140625,
"learning_rate": 8.395887191422397e-05,
"loss": 0.9886,
"step": 310
},
{
"epoch": 6.0,
"eval_loss": 2.5159404277801514,
"eval_runtime": 0.5352,
"eval_samples_per_second": 18.685,
"eval_steps_per_second": 1.869,
"step": 312
},
{
"epoch": 6.0576923076923075,
"grad_norm": 0.3046875,
"learning_rate": 8.065559069717088e-05,
"loss": 0.9754,
"step": 315
},
{
"epoch": 6.153846153846154,
"grad_norm": 0.345703125,
"learning_rate": 7.73740997570278e-05,
"loss": 0.9761,
"step": 320
},
{
"epoch": 6.25,
"grad_norm": 0.41796875,
"learning_rate": 7.411809548974792e-05,
"loss": 0.9699,
"step": 325
},
{
"epoch": 6.346153846153846,
"grad_norm": 0.421875,
"learning_rate": 7.089124558212871e-05,
"loss": 0.9879,
"step": 330
},
{
"epoch": 6.4423076923076925,
"grad_norm": 0.48046875,
"learning_rate": 6.769718488039023e-05,
"loss": 0.9686,
"step": 335
},
{
"epoch": 6.538461538461538,
"grad_norm": 0.4296875,
"learning_rate": 6.453951129574644e-05,
"loss": 0.9671,
"step": 340
},
{
"epoch": 6.634615384615385,
"grad_norm": 0.3359375,
"learning_rate": 6.142178175158149e-05,
"loss": 0.9746,
"step": 345
},
{
"epoch": 6.730769230769231,
"grad_norm": 0.322265625,
"learning_rate": 5.834750817679606e-05,
"loss": 0.9724,
"step": 350
},
{
"epoch": 6.826923076923077,
"grad_norm": 0.31640625,
"learning_rate": 5.5320153549837415e-05,
"loss": 0.9788,
"step": 355
},
{
"epoch": 6.923076923076923,
"grad_norm": 0.330078125,
"learning_rate": 5.234312799786921e-05,
"loss": 0.9711,
"step": 360
},
{
"epoch": 7.0,
"eval_loss": 2.5233545303344727,
"eval_runtime": 0.5403,
"eval_samples_per_second": 18.508,
"eval_steps_per_second": 1.851,
"step": 364
},
{
"epoch": 7.019230769230769,
"grad_norm": 0.353515625,
"learning_rate": 4.9419784955474524e-05,
"loss": 0.9709,
"step": 365
},
{
"epoch": 7.115384615384615,
"grad_norm": 0.3203125,
"learning_rate": 4.6553417387219886e-05,
"loss": 0.9553,
"step": 370
},
{
"epoch": 7.211538461538462,
"grad_norm": 0.310546875,
"learning_rate": 4.374725407833532e-05,
"loss": 0.9601,
"step": 375
},
{
"epoch": 7.3076923076923075,
"grad_norm": 0.30078125,
"learning_rate": 4.100445599768774e-05,
"loss": 0.9674,
"step": 380
},
{
"epoch": 7.403846153846154,
"grad_norm": 0.4140625,
"learning_rate": 3.832811273714569e-05,
"loss": 0.9669,
"step": 385
},
{
"epoch": 7.5,
"grad_norm": 0.37109375,
"learning_rate": 3.5721239031346066e-05,
"loss": 0.9633,
"step": 390
},
{
"epoch": 7.596153846153846,
"grad_norm": 0.328125,
"learning_rate": 3.318677136178228e-05,
"loss": 0.9557,
"step": 395
},
{
"epoch": 7.6923076923076925,
"grad_norm": 0.37109375,
"learning_rate": 3.072756464904006e-05,
"loss": 0.966,
"step": 400
},
{
"epoch": 7.788461538461538,
"grad_norm": 0.341796875,
"learning_rate": 2.8346389036906828e-05,
"loss": 0.9638,
"step": 405
},
{
"epoch": 7.884615384615385,
"grad_norm": 0.33984375,
"learning_rate": 2.6045926771976303e-05,
"loss": 0.9654,
"step": 410
},
{
"epoch": 7.980769230769231,
"grad_norm": 0.310546875,
"learning_rate": 2.382876918226409e-05,
"loss": 0.9637,
"step": 415
},
{
"epoch": 8.0,
"eval_loss": 2.5294220447540283,
"eval_runtime": 0.536,
"eval_samples_per_second": 18.656,
"eval_steps_per_second": 1.866,
"step": 416
},
{
"epoch": 8.076923076923077,
"grad_norm": 0.310546875,
"learning_rate": 2.1697413758237784e-05,
"loss": 0.9582,
"step": 420
},
{
"epoch": 8.173076923076923,
"grad_norm": 0.30859375,
"learning_rate": 1.965426133954854e-05,
"loss": 0.9599,
"step": 425
},
{
"epoch": 8.26923076923077,
"grad_norm": 0.30078125,
"learning_rate": 1.7701613410634365e-05,
"loss": 0.959,
"step": 430
},
{
"epoch": 8.365384615384615,
"grad_norm": 0.314453125,
"learning_rate": 1.584166950824061e-05,
"loss": 0.9586,
"step": 435
},
{
"epoch": 8.461538461538462,
"grad_norm": 0.30859375,
"learning_rate": 1.4076524743778319e-05,
"loss": 0.9512,
"step": 440
},
{
"epoch": 8.557692307692308,
"grad_norm": 0.29296875,
"learning_rate": 1.2408167443311214e-05,
"loss": 0.9549,
"step": 445
},
{
"epoch": 8.653846153846153,
"grad_norm": 0.302734375,
"learning_rate": 1.083847690782972e-05,
"loss": 0.955,
"step": 450
},
{
"epoch": 8.75,
"grad_norm": 0.326171875,
"learning_rate": 9.369221296335006e-06,
"loss": 0.9558,
"step": 455
},
{
"epoch": 8.846153846153847,
"grad_norm": 0.310546875,
"learning_rate": 8.002055634117578e-06,
"loss": 0.9628,
"step": 460
},
{
"epoch": 8.942307692307692,
"grad_norm": 0.298828125,
"learning_rate": 6.738519948473976e-06,
"loss": 0.9604,
"step": 465
},
{
"epoch": 9.0,
"eval_loss": 2.5298619270324707,
"eval_runtime": 0.5355,
"eval_samples_per_second": 18.673,
"eval_steps_per_second": 1.867,
"step": 468
},
{
"epoch": 9.038461538461538,
"grad_norm": 0.2890625,
"learning_rate": 5.580037533961546e-06,
"loss": 0.9639,
"step": 470
},
{
"epoch": 9.134615384615385,
"grad_norm": 0.294921875,
"learning_rate": 4.527913349145441e-06,
"loss": 0.9503,
"step": 475
},
{
"epoch": 9.23076923076923,
"grad_norm": 0.296875,
"learning_rate": 3.5833325466437694e-06,
"loss": 0.9569,
"step": 480
},
{
"epoch": 9.326923076923077,
"grad_norm": 0.3046875,
"learning_rate": 2.7473591381266708e-06,
"loss": 0.9571,
"step": 485
},
{
"epoch": 9.423076923076923,
"grad_norm": 0.302734375,
"learning_rate": 2.0209347957732328e-06,
"loss": 0.9528,
"step": 490
},
{
"epoch": 9.51923076923077,
"grad_norm": 0.30859375,
"learning_rate": 1.404877791536141e-06,
"loss": 0.9599,
"step": 495
},
{
"epoch": 9.615384615384615,
"grad_norm": 0.3125,
"learning_rate": 8.998820754091531e-07,
"loss": 0.956,
"step": 500
},
{
"epoch": 9.711538461538462,
"grad_norm": 0.30078125,
"learning_rate": 5.065164937354428e-07,
"loss": 0.9578,
"step": 505
},
{
"epoch": 9.807692307692308,
"grad_norm": 0.314453125,
"learning_rate": 2.2522414843748618e-07,
"loss": 0.9612,
"step": 510
},
{
"epoch": 9.903846153846153,
"grad_norm": 0.29296875,
"learning_rate": 5.632189789027687e-08,
"loss": 0.9593,
"step": 515
},
{
"epoch": 10.0,
"grad_norm": 0.298828125,
"learning_rate": 0.0,
"loss": 0.9523,
"step": 520
},
{
"epoch": 10.0,
"eval_loss": 2.5308985710144043,
"eval_runtime": 0.5343,
"eval_samples_per_second": 18.716,
"eval_steps_per_second": 1.872,
"step": 520
},
{
"epoch": 10.0,
"step": 520,
"total_flos": 4.072455240941568e+17,
"train_loss": 1.1447187717144305,
"train_runtime": 1794.4362,
"train_samples_per_second": 18.541,
"train_steps_per_second": 0.29
}
],
"logging_steps": 5,
"max_steps": 520,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 100,
"total_flos": 4.072455240941568e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}