695 lines
16 KiB
JSON
695 lines
16 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 3.0,
|
|
"eval_steps": 500,
|
|
"global_step": 938,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.032,
|
|
"grad_norm": 5.314220428466797,
|
|
"learning_rate": 9.574468085106384e-07,
|
|
"loss": 0.8663,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.064,
|
|
"grad_norm": 2.9840216636657715,
|
|
"learning_rate": 2.021276595744681e-06,
|
|
"loss": 0.7951,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.096,
|
|
"grad_norm": 1.4162436723709106,
|
|
"learning_rate": 3.0851063829787237e-06,
|
|
"loss": 0.7093,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.128,
|
|
"grad_norm": 1.0216368436813354,
|
|
"learning_rate": 4.148936170212766e-06,
|
|
"loss": 0.6386,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"grad_norm": 0.7919172644615173,
|
|
"learning_rate": 5.212765957446809e-06,
|
|
"loss": 0.6059,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.192,
|
|
"grad_norm": 0.7189238667488098,
|
|
"learning_rate": 6.276595744680851e-06,
|
|
"loss": 0.5966,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.224,
|
|
"grad_norm": 0.8121111392974854,
|
|
"learning_rate": 7.340425531914894e-06,
|
|
"loss": 0.572,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.256,
|
|
"grad_norm": 0.7291969656944275,
|
|
"learning_rate": 8.404255319148937e-06,
|
|
"loss": 0.5621,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.288,
|
|
"grad_norm": 0.8778895139694214,
|
|
"learning_rate": 9.46808510638298e-06,
|
|
"loss": 0.5576,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"grad_norm": 0.725179135799408,
|
|
"learning_rate": 9.999136119166803e-06,
|
|
"loss": 0.5505,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.352,
|
|
"grad_norm": 0.7167001962661743,
|
|
"learning_rate": 9.9922268634943e-06,
|
|
"loss": 0.5527,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.384,
|
|
"grad_norm": 0.6954172849655151,
|
|
"learning_rate": 9.978417901361958e-06,
|
|
"loss": 0.5533,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.416,
|
|
"grad_norm": 0.6993622183799744,
|
|
"learning_rate": 9.95772831799724e-06,
|
|
"loss": 0.5537,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.448,
|
|
"grad_norm": 0.739738404750824,
|
|
"learning_rate": 9.930186708264902e-06,
|
|
"loss": 0.5432,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"grad_norm": 0.7163270115852356,
|
|
"learning_rate": 9.895831137146319e-06,
|
|
"loss": 0.5293,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.512,
|
|
"grad_norm": 0.7983338236808777,
|
|
"learning_rate": 9.854709087130261e-06,
|
|
"loss": 0.5444,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.544,
|
|
"grad_norm": 0.8286071419715881,
|
|
"learning_rate": 9.80687739258782e-06,
|
|
"loss": 0.5227,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.576,
|
|
"grad_norm": 0.6619768142700195,
|
|
"learning_rate": 9.7524021612222e-06,
|
|
"loss": 0.5313,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.608,
|
|
"grad_norm": 0.6172661185264587,
|
|
"learning_rate": 9.691358682701927e-06,
|
|
"loss": 0.5313,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"grad_norm": 0.6922529339790344,
|
|
"learning_rate": 9.623831324603755e-06,
|
|
"loss": 0.5263,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.672,
|
|
"grad_norm": 0.7383220195770264,
|
|
"learning_rate": 9.549913415809084e-06,
|
|
"loss": 0.536,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.704,
|
|
"grad_norm": 0.9165273308753967,
|
|
"learning_rate": 9.469707117515068e-06,
|
|
"loss": 0.532,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.736,
|
|
"grad_norm": 0.637290894985199,
|
|
"learning_rate": 9.383323282038632e-06,
|
|
"loss": 0.5247,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.768,
|
|
"grad_norm": 0.7203924655914307,
|
|
"learning_rate": 9.29088129960862e-06,
|
|
"loss": 0.5367,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"grad_norm": 0.7599261403083801,
|
|
"learning_rate": 9.192508933357753e-06,
|
|
"loss": 0.535,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.832,
|
|
"grad_norm": 0.7834025025367737,
|
|
"learning_rate": 9.088342142742493e-06,
|
|
"loss": 0.528,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.864,
|
|
"grad_norm": 0.6317589282989502,
|
|
"learning_rate": 8.978524895634842e-06,
|
|
"loss": 0.5359,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.896,
|
|
"grad_norm": 0.6225187182426453,
|
|
"learning_rate": 8.86320896934581e-06,
|
|
"loss": 0.5039,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.928,
|
|
"grad_norm": 0.6623832583427429,
|
|
"learning_rate": 8.742553740855507e-06,
|
|
"loss": 0.5277,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"grad_norm": 0.7472139596939087,
|
|
"learning_rate": 8.616725966539831e-06,
|
|
"loss": 0.5213,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.992,
|
|
"grad_norm": 0.6038188934326172,
|
|
"learning_rate": 8.485899551698166e-06,
|
|
"loss": 0.5248,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 1.0224,
|
|
"grad_norm": 0.6474160552024841,
|
|
"learning_rate": 8.350255310200611e-06,
|
|
"loss": 0.4818,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 1.0544,
|
|
"grad_norm": 0.6729207634925842,
|
|
"learning_rate": 8.209980714586955e-06,
|
|
"loss": 0.4563,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 1.0864,
|
|
"grad_norm": 0.6256211400032043,
|
|
"learning_rate": 8.065269636962765e-06,
|
|
"loss": 0.4494,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 1.1184,
|
|
"grad_norm": 0.6500005722045898,
|
|
"learning_rate": 7.916322081050708e-06,
|
|
"loss": 0.4492,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 1.1504,
|
|
"grad_norm": 0.5361557006835938,
|
|
"learning_rate": 7.76334390576742e-06,
|
|
"loss": 0.4539,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 1.1824,
|
|
"grad_norm": 0.6814967393875122,
|
|
"learning_rate": 7.60654654070796e-06,
|
|
"loss": 0.4433,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 1.2144,
|
|
"grad_norm": 0.6009787917137146,
|
|
"learning_rate": 7.446146693931111e-06,
|
|
"loss": 0.4536,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 1.2464,
|
|
"grad_norm": 0.6081900000572205,
|
|
"learning_rate": 7.282366052449351e-06,
|
|
"loss": 0.4623,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 1.2784,
|
|
"grad_norm": 0.6118571758270264,
|
|
"learning_rate": 7.115430975837457e-06,
|
|
"loss": 0.4437,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 1.3104,
|
|
"grad_norm": 0.5691370368003845,
|
|
"learning_rate": 6.945572183383229e-06,
|
|
"loss": 0.4448,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 1.3424,
|
|
"grad_norm": 0.5975236296653748,
|
|
"learning_rate": 6.773024435212678e-06,
|
|
"loss": 0.4491,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 1.3744,
|
|
"grad_norm": 0.5852976441383362,
|
|
"learning_rate": 6.598026207830428e-06,
|
|
"loss": 0.4482,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 1.4064,
|
|
"grad_norm": 0.4990760087966919,
|
|
"learning_rate": 6.4208193645237314e-06,
|
|
"loss": 0.4564,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 1.4384000000000001,
|
|
"grad_norm": 0.6382614374160767,
|
|
"learning_rate": 6.241648821085666e-06,
|
|
"loss": 0.454,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 1.4704,
|
|
"grad_norm": 0.5935061573982239,
|
|
"learning_rate": 6.060762207319479e-06,
|
|
"loss": 0.4504,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 1.5024,
|
|
"grad_norm": 0.48114678263664246,
|
|
"learning_rate": 5.878409524791931e-06,
|
|
"loss": 0.4444,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 1.5344,
|
|
"grad_norm": 0.5724011063575745,
|
|
"learning_rate": 5.694842801308651e-06,
|
|
"loss": 0.444,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 1.5664,
|
|
"grad_norm": 0.5499016642570496,
|
|
"learning_rate": 5.510315742589042e-06,
|
|
"loss": 0.4311,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 1.5984,
|
|
"grad_norm": 0.7091147303581238,
|
|
"learning_rate": 5.325083381622165e-06,
|
|
"loss": 0.4596,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 1.6303999999999998,
|
|
"grad_norm": 0.5529297590255737,
|
|
"learning_rate": 5.139401726188208e-06,
|
|
"loss": 0.4459,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 1.6623999999999999,
|
|
"grad_norm": 0.6218659281730652,
|
|
"learning_rate": 4.953527405032723e-06,
|
|
"loss": 0.4525,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 1.6944,
|
|
"grad_norm": 0.599751353263855,
|
|
"learning_rate": 4.767717313182611e-06,
|
|
"loss": 0.4604,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 1.7264,
|
|
"grad_norm": 0.5663022994995117,
|
|
"learning_rate": 4.582228256894093e-06,
|
|
"loss": 0.4418,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 1.7584,
|
|
"grad_norm": 0.48218056559562683,
|
|
"learning_rate": 4.397316598723385e-06,
|
|
"loss": 0.4548,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 1.7904,
|
|
"grad_norm": 0.6282760500907898,
|
|
"learning_rate": 4.2132379032105695e-06,
|
|
"loss": 0.4385,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 1.8224,
|
|
"grad_norm": 0.5830026268959045,
|
|
"learning_rate": 4.030246583666437e-06,
|
|
"loss": 0.4507,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 1.8544,
|
|
"grad_norm": 0.5871529579162598,
|
|
"learning_rate": 3.848595550550401e-06,
|
|
"loss": 0.4547,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 1.8864,
|
|
"grad_norm": 0.483813613653183,
|
|
"learning_rate": 3.668535861925509e-06,
|
|
"loss": 0.4402,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 1.9184,
|
|
"grad_norm": 0.5708866715431213,
|
|
"learning_rate": 3.4903163764736104e-06,
|
|
"loss": 0.4446,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 1.9504000000000001,
|
|
"grad_norm": 0.6678810119628906,
|
|
"learning_rate": 3.314183409550293e-06,
|
|
"loss": 0.3824,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 1.9824000000000002,
|
|
"grad_norm": 0.5966857075691223,
|
|
"learning_rate": 3.140380392754901e-06,
|
|
"loss": 0.3858,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 2.016,
|
|
"grad_norm": 0.6602267622947693,
|
|
"learning_rate": 2.969147537486175e-06,
|
|
"loss": 0.4165,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 2.048,
|
|
"grad_norm": 0.5915480256080627,
|
|
"learning_rate": 2.800721502948506e-06,
|
|
"loss": 0.3785,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"grad_norm": 0.5668565034866333,
|
|
"learning_rate": 2.635335069067617e-06,
|
|
"loss": 0.3911,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 2.112,
|
|
"grad_norm": 0.5256936550140381,
|
|
"learning_rate": 2.4732168147677927e-06,
|
|
"loss": 0.3997,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 2.144,
|
|
"grad_norm": 0.5043026804924011,
|
|
"learning_rate": 2.314590802055232e-06,
|
|
"loss": 0.3896,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 2.176,
|
|
"grad_norm": 0.5051006078720093,
|
|
"learning_rate": 2.159676266344222e-06,
|
|
"loss": 0.3966,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 2.208,
|
|
"grad_norm": 0.5608514547348022,
|
|
"learning_rate": 2.0086873134540626e-06,
|
|
"loss": 0.3791,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"grad_norm": 0.620739758014679,
|
|
"learning_rate": 1.8618326236955908e-06,
|
|
"loss": 0.3924,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 2.2720000000000002,
|
|
"grad_norm": 0.5500799417495728,
|
|
"learning_rate": 1.7193151634562071e-06,
|
|
"loss": 0.3884,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 2.304,
|
|
"grad_norm": 0.5139204859733582,
|
|
"learning_rate": 1.581331904682089e-06,
|
|
"loss": 0.3825,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 2.336,
|
|
"grad_norm": 0.5042279958724976,
|
|
"learning_rate": 1.4480735526452427e-06,
|
|
"loss": 0.391,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 2.368,
|
|
"grad_norm": 0.5282170176506042,
|
|
"learning_rate": 1.319724282371664e-06,
|
|
"loss": 0.3858,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"grad_norm": 0.5384218692779541,
|
|
"learning_rate": 1.1964614840949002e-06,
|
|
"loss": 0.3903,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 2.432,
|
|
"grad_norm": 0.4508267045021057,
|
|
"learning_rate": 1.078455518086784e-06,
|
|
"loss": 0.3849,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 2.464,
|
|
"grad_norm": 0.49801719188690186,
|
|
"learning_rate": 9.658694792042284e-07,
|
|
"loss": 0.3902,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 2.496,
|
|
"grad_norm": 0.47622281312942505,
|
|
"learning_rate": 8.58858971477457e-07,
|
|
"loss": 0.3774,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 2.528,
|
|
"grad_norm": 0.4954831302165985,
|
|
"learning_rate": 7.575718930512516e-07,
|
|
"loss": 0.3899,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"grad_norm": 0.5186880230903625,
|
|
"learning_rate": 6.621482317764105e-07,
|
|
"loss": 0.3954,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 2.592,
|
|
"grad_norm": 0.4855698049068451,
|
|
"learning_rate": 5.727198717339511e-07,
|
|
"loss": 0.3809,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 2.624,
|
|
"grad_norm": 0.5997398495674133,
|
|
"learning_rate": 4.894104109594466e-07,
|
|
"loss": 0.3881,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 2.656,
|
|
"grad_norm": 0.5895832180976868,
|
|
"learning_rate": 4.123349906194357e-07,
|
|
"loss": 0.3869,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 2.6879999999999997,
|
|
"grad_norm": 0.5051674842834473,
|
|
"learning_rate": 3.416001358759635e-07,
|
|
"loss": 0.3859,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 2.7199999999999998,
|
|
"grad_norm": 0.5103119611740112,
|
|
"learning_rate": 2.7730360865923954e-07,
|
|
"loss": 0.3855,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 2.752,
|
|
"grad_norm": 0.5123039484024048,
|
|
"learning_rate": 2.1953427255185122e-07,
|
|
"loss": 0.384,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 2.784,
|
|
"grad_norm": 0.5374960899353027,
|
|
"learning_rate": 1.6837196997130434e-07,
|
|
"loss": 0.3866,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 2.816,
|
|
"grad_norm": 0.5188184976577759,
|
|
"learning_rate": 1.2388741182062348e-07,
|
|
"loss": 0.3924,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 2.848,
|
|
"grad_norm": 0.5518755316734314,
|
|
"learning_rate": 8.614207975952083e-08,
|
|
"loss": 0.391,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"grad_norm": 0.5774142742156982,
|
|
"learning_rate": 5.518814123121885e-08,
|
|
"loss": 0.3651,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 2.912,
|
|
"grad_norm": 0.4625570476055145,
|
|
"learning_rate": 3.10683773623488e-08,
|
|
"loss": 0.3891,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 2.944,
|
|
"grad_norm": 0.5453880429267883,
|
|
"learning_rate": 1.3816123835588835e-08,
|
|
"loss": 0.3982,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 2.976,
|
|
"grad_norm": 0.48320916295051575,
|
|
"learning_rate": 3.4552248167507576e-09,
|
|
"loss": 0.3864,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"step": 938,
|
|
"total_flos": 1.2370288879507341e+19,
|
|
"train_loss": 0.13962188102543227,
|
|
"train_runtime": 7247.2658,
|
|
"train_samples_per_second": 4.139,
|
|
"train_steps_per_second": 0.13
|
|
}
|
|
],
|
|
"logging_steps": 10,
|
|
"max_steps": 939,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 3,
|
|
"save_steps": 200,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.2370288879507341e+19,
|
|
"train_batch_size": 2,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|