737 lines
15 KiB
JSON
737 lines
15 KiB
JSON
{
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.7012007389162562,
|
|
"eval_steps": 500,
|
|
"global_step": 570,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.0,
|
|
"learning_rate": 1.999992515597253e-05,
|
|
"loss": 1.5474,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"learning_rate": 1.9998128955328933e-05,
|
|
"loss": 1.182,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 1.999251652147735e-05,
|
|
"loss": 1.0465,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"learning_rate": 1.9983164798668152e-05,
|
|
"loss": 1.0136,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"learning_rate": 1.997007728639956e-05,
|
|
"loss": 1.0103,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"learning_rate": 1.9953258882135594e-05,
|
|
"loss": 0.9887,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"learning_rate": 1.9932715879473385e-05,
|
|
"loss": 0.9841,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"learning_rate": 1.990845596578807e-05,
|
|
"loss": 0.9791,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.9880488219356086e-05,
|
|
"loss": 0.9779,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"learning_rate": 1.9848823105958023e-05,
|
|
"loss": 0.9757,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"learning_rate": 1.981347247496222e-05,
|
|
"loss": 0.9683,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"learning_rate": 1.9774449554890615e-05,
|
|
"loss": 0.967,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"learning_rate": 1.973176894846855e-05,
|
|
"loss": 0.9689,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.968544662716026e-05,
|
|
"loss": 0.9559,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"learning_rate": 1.963549992519223e-05,
|
|
"loss": 0.9524,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"learning_rate": 1.9581947533066577e-05,
|
|
"loss": 0.9517,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"learning_rate": 1.9524809490566878e-05,
|
|
"loss": 0.9569,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"learning_rate": 1.9464107179259126e-05,
|
|
"loss": 0.9547,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"learning_rate": 1.939986331449053e-05,
|
|
"loss": 0.9612,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"learning_rate": 1.9332101936889262e-05,
|
|
"loss": 0.9497,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"learning_rate": 1.926084840336821e-05,
|
|
"loss": 0.9406,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"learning_rate": 1.918612937763622e-05,
|
|
"loss": 0.9517,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"learning_rate": 1.910797282022027e-05,
|
|
"loss": 0.9667,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"learning_rate": 1.9026407978002415e-05,
|
|
"loss": 0.9505,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"learning_rate": 1.894146537327533e-05,
|
|
"loss": 0.9409,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"learning_rate": 1.8853176792320598e-05,
|
|
"loss": 0.946,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"learning_rate": 1.8761575273514005e-05,
|
|
"loss": 0.9455,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"learning_rate": 1.866669509496227e-05,
|
|
"loss": 0.949,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"learning_rate": 1.8568571761675893e-05,
|
|
"loss": 0.9583,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"learning_rate": 1.8467241992282842e-05,
|
|
"loss": 0.9387,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"learning_rate": 1.8362743705288127e-05,
|
|
"loss": 0.9492,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"learning_rate": 1.825511600488435e-05,
|
|
"loss": 0.9404,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"learning_rate": 1.814439916631857e-05,
|
|
"loss": 0.9419,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"learning_rate": 1.803063462082095e-05,
|
|
"loss": 0.9337,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.791386494010081e-05,
|
|
"loss": 0.936,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"learning_rate": 1.7794133820415916e-05,
|
|
"loss": 0.942,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"learning_rate": 1.7671486066220965e-05,
|
|
"loss": 0.9367,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"learning_rate": 1.7545967573401336e-05,
|
|
"loss": 0.9333,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"learning_rate": 1.7417625312098453e-05,
|
|
"loss": 0.9438,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.728650730913312e-05,
|
|
"loss": 0.9453,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"learning_rate": 1.7152662630033506e-05,
|
|
"loss": 0.941,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"learning_rate": 1.7016141360674307e-05,
|
|
"loss": 0.9361,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"learning_rate": 1.6876994588534234e-05,
|
|
"loss": 0.9346,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"learning_rate": 1.6735274383578593e-05,
|
|
"loss": 0.9343,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"learning_rate": 1.659103377877423e-05,
|
|
"loss": 0.9369,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"learning_rate": 1.6444326750244145e-05,
|
|
"loss": 0.9483,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"learning_rate": 1.629520819706912e-05,
|
|
"loss": 0.935,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"learning_rate": 1.614373392074402e-05,
|
|
"loss": 0.9382,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"learning_rate": 1.598996060429634e-05,
|
|
"loss": 0.9345,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.5833945791074943e-05,
|
|
"loss": 0.9369,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"learning_rate": 1.56757478632168e-05,
|
|
"loss": 0.9393,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"learning_rate": 1.5515426019799898e-05,
|
|
"loss": 0.9406,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.5353040254690396e-05,
|
|
"loss": 0.9368,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"learning_rate": 1.5188651334092373e-05,
|
|
"loss": 0.9229,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.5022320773808612e-05,
|
|
"loss": 0.9341,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"learning_rate": 1.4854110816220805e-05,
|
|
"loss": 0.9323,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"learning_rate": 1.4684084406997903e-05,
|
|
"loss": 0.9345,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"learning_rate": 1.4512305171541287e-05,
|
|
"loss": 0.9312,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"eval_loss": 0.9350774884223938,
|
|
"eval_runtime": 652.1015,
|
|
"eval_samples_per_second": 35.439,
|
|
"eval_steps_per_second": 0.555,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"learning_rate": 1.4338837391175582e-05,
|
|
"loss": 0.9066,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"learning_rate": 1.4163745979093997e-05,
|
|
"loss": 0.9055,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"learning_rate": 1.3987096456067236e-05,
|
|
"loss": 0.8936,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"learning_rate": 1.3808954925925042e-05,
|
|
"loss": 0.9135,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"learning_rate": 1.3629388050819547e-05,
|
|
"loss": 0.9013,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"learning_rate": 1.3448463026279706e-05,
|
|
"loss": 0.8976,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"learning_rate": 1.3266247556066122e-05,
|
|
"loss": 0.8975,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"learning_rate": 1.3082809826835707e-05,
|
|
"loss": 0.9009,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"learning_rate": 1.2898218482625606e-05,
|
|
"loss": 0.8994,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"learning_rate": 1.2712542599165998e-05,
|
|
"loss": 0.8943,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"learning_rate": 1.252585165803135e-05,
|
|
"loss": 0.8928,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"learning_rate": 1.233821552063977e-05,
|
|
"loss": 0.8967,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"learning_rate": 1.2149704402110243e-05,
|
|
"loss": 0.8833,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"learning_rate": 1.1960388844987529e-05,
|
|
"loss": 0.8832,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"learning_rate": 1.1770339692844484e-05,
|
|
"loss": 0.8823,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"learning_rate": 1.157962806377179e-05,
|
|
"loss": 0.8875,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"learning_rate": 1.1388325323764889e-05,
|
|
"loss": 0.8855,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"learning_rate": 1.1196503060018243e-05,
|
|
"loss": 0.8897,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"learning_rate": 1.1004233054136726e-05,
|
|
"loss": 0.8806,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"learning_rate": 1.0811587255274313e-05,
|
|
"loss": 0.8747,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"learning_rate": 1.0618637753210086e-05,
|
|
"loss": 0.8863,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"learning_rate": 1.0425456751371564e-05,
|
|
"loss": 0.8936,
|
|
"step": 395
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"learning_rate": 1.0232116539815558e-05,
|
|
"loss": 0.8833,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"learning_rate": 1.0038689468176575e-05,
|
|
"loss": 0.8783,
|
|
"step": 405
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"learning_rate": 9.845247918592937e-06,
|
|
"loss": 0.8799,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"learning_rate": 9.651864278620752e-06,
|
|
"loss": 0.8795,
|
|
"step": 415
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"learning_rate": 9.458610914145826e-06,
|
|
"loss": 0.882,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"learning_rate": 9.26556014230371e-06,
|
|
"loss": 0.8911,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"learning_rate": 9.072784204417995e-06,
|
|
"loss": 0.8765,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"learning_rate": 8.880355238966923e-06,
|
|
"loss": 0.8852,
|
|
"step": 435
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"learning_rate": 8.688345254588579e-06,
|
|
"loss": 0.874,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"learning_rate": 8.496826103134572e-06,
|
|
"loss": 0.8766,
|
|
"step": 445
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"learning_rate": 8.305869452782446e-06,
|
|
"loss": 0.8686,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"learning_rate": 8.115546761216822e-06,
|
|
"loss": 0.8749,
|
|
"step": 455
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"learning_rate": 7.92592924888925e-06,
|
|
"loss": 0.8771,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"learning_rate": 7.73708787236694e-06,
|
|
"loss": 0.876,
|
|
"step": 465
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"learning_rate": 7.549093297780133e-06,
|
|
"loss": 0.8727,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"learning_rate": 7.362015874378219e-06,
|
|
"loss": 0.8772,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"learning_rate": 7.175925608204428e-06,
|
|
"loss": 0.8825,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"learning_rate": 6.990892135898937e-06,
|
|
"loss": 0.8822,
|
|
"step": 485
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"learning_rate": 6.806984698640202e-06,
|
|
"loss": 0.8743,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"learning_rate": 6.624272116234319e-06,
|
|
"loss": 0.8738,
|
|
"step": 495
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"learning_rate": 6.442822761362015e-06,
|
|
"loss": 0.8782,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"learning_rate": 6.262704533992995e-06,
|
|
"loss": 0.8742,
|
|
"step": 505
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"learning_rate": 6.083984835977154e-06,
|
|
"loss": 0.8874,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"learning_rate": 5.9067305458222054e-06,
|
|
"loss": 0.8786,
|
|
"step": 515
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"learning_rate": 5.731007993667155e-06,
|
|
"loss": 0.8789,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"learning_rate": 5.556882936460966e-06,
|
|
"loss": 0.8759,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"learning_rate": 5.38442053335571e-06,
|
|
"loss": 0.8791,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"learning_rate": 5.213685321323451e-06,
|
|
"loss": 0.8808,
|
|
"step": 535
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"learning_rate": 5.044741191005908e-06,
|
|
"loss": 0.8846,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"learning_rate": 4.877651362806038e-06,
|
|
"loss": 0.8833,
|
|
"step": 545
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"learning_rate": 4.712478363230361e-06,
|
|
"loss": 0.8654,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"learning_rate": 4.5492840014910145e-06,
|
|
"loss": 0.88,
|
|
"step": 555
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"learning_rate": 4.388129346376177e-06,
|
|
"loss": 0.8759,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"learning_rate": 4.22907470339758e-06,
|
|
"loss": 0.8815,
|
|
"step": 565
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"learning_rate": 4.07217959222365e-06,
|
|
"loss": 0.8747,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"eval_loss": 0.9257117509841919,
|
|
"eval_runtime": 652.8923,
|
|
"eval_samples_per_second": 35.396,
|
|
"eval_steps_per_second": 0.554,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"step": 570,
|
|
"total_flos": 955556403609600.0,
|
|
"train_loss": 0.9215259830156962,
|
|
"train_runtime": 46918.2161,
|
|
"train_samples_per_second": 8.861,
|
|
"train_steps_per_second": 0.017
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 812,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 2,
|
|
"save_steps": 500,
|
|
"total_flos": 955556403609600.0,
|
|
"train_batch_size": 8,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|