1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.033488496701383076,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.74211368560791,
|
|
"epoch": 0.00033488496701383075,
|
|
"grad_norm": 4.0,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7607,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 10113.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.742099857330322,
|
|
"epoch": 0.0006697699340276615,
|
|
"grad_norm": 4.40625,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7537,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 19598.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742137718200684,
|
|
"epoch": 0.0010046549010414923,
|
|
"grad_norm": 4.1875,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.7462,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 29705.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742128372192383,
|
|
"epoch": 0.001339539868055323,
|
|
"grad_norm": 5.09375,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.7119,
|
|
"mean_token_accuracy": 0.00036584193585440514,
|
|
"num_tokens": 39871.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742036437988281,
|
|
"epoch": 0.0016744248350691537,
|
|
"grad_norm": 4.09375,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.6391,
|
|
"mean_token_accuracy": 0.0023467960068956017,
|
|
"num_tokens": 49805.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.741703224182128,
|
|
"epoch": 0.0020093098020829846,
|
|
"grad_norm": 3.734375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.5416,
|
|
"mean_token_accuracy": 0.027319939527660608,
|
|
"num_tokens": 59845.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.740657615661622,
|
|
"epoch": 0.0023441947690968153,
|
|
"grad_norm": 3.28125,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.4647,
|
|
"mean_token_accuracy": 0.03426684513688087,
|
|
"num_tokens": 69836.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.737139415740966,
|
|
"epoch": 0.002679079736110646,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.3382,
|
|
"mean_token_accuracy": 0.040677194483578205,
|
|
"num_tokens": 80088.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.727961826324464,
|
|
"epoch": 0.0030139647031244767,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.2362,
|
|
"mean_token_accuracy": 0.04172689206898213,
|
|
"num_tokens": 89458.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.714122295379639,
|
|
"epoch": 0.0033488496701383074,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.1927,
|
|
"mean_token_accuracy": 0.04110763110220432,
|
|
"num_tokens": 99523.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.702536869049073,
|
|
"epoch": 0.003683734637152138,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.1602,
|
|
"mean_token_accuracy": 0.04160156771540642,
|
|
"num_tokens": 110080.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.694610786437988,
|
|
"epoch": 0.004018619604165969,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 10.0683,
|
|
"mean_token_accuracy": 0.043702947348356246,
|
|
"num_tokens": 119412.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.687895107269288,
|
|
"epoch": 0.0043535045711798,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 10.0255,
|
|
"mean_token_accuracy": 0.04217404723167419,
|
|
"num_tokens": 129432.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.679110527038574,
|
|
"epoch": 0.0046883895381936306,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.9755,
|
|
"mean_token_accuracy": 0.048846100270748136,
|
|
"num_tokens": 139756.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.670030307769775,
|
|
"epoch": 0.005023274505207461,
|
|
"grad_norm": 2.140625,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.96,
|
|
"mean_token_accuracy": 0.045024541579186915,
|
|
"num_tokens": 148846.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.662967872619628,
|
|
"epoch": 0.005358159472221292,
|
|
"grad_norm": 1.515625,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.8931,
|
|
"mean_token_accuracy": 0.04923531375825405,
|
|
"num_tokens": 158647.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.654012298583984,
|
|
"epoch": 0.005693044439235123,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.8077,
|
|
"mean_token_accuracy": 0.04514540694653988,
|
|
"num_tokens": 167701.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.641143226623536,
|
|
"epoch": 0.006027929406248953,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.7848,
|
|
"mean_token_accuracy": 0.04806430526077747,
|
|
"num_tokens": 178428.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.615053653717041,
|
|
"epoch": 0.006362814373262784,
|
|
"grad_norm": 1.515625,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.7243,
|
|
"mean_token_accuracy": 0.04973807036876678,
|
|
"num_tokens": 187737.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.592188262939453,
|
|
"epoch": 0.006697699340276615,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.7193,
|
|
"mean_token_accuracy": 0.05120275542140007,
|
|
"num_tokens": 198374.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.56552038192749,
|
|
"epoch": 0.007032584307290445,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.6436,
|
|
"mean_token_accuracy": 0.05372070819139481,
|
|
"num_tokens": 208483.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.523834037780762,
|
|
"epoch": 0.007367469274304276,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.5595,
|
|
"mean_token_accuracy": 0.050585757568478586,
|
|
"num_tokens": 218715.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.464954853057861,
|
|
"epoch": 0.007702354241318108,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.3861,
|
|
"mean_token_accuracy": 0.0644374880939722,
|
|
"num_tokens": 227777.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.395084571838378,
|
|
"epoch": 0.008037239208331938,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 9.3999,
|
|
"mean_token_accuracy": 0.056889945082366465,
|
|
"num_tokens": 238192.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.354606533050537,
|
|
"epoch": 0.008372124175345769,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 9.2388,
|
|
"mean_token_accuracy": 0.06168701723217964,
|
|
"num_tokens": 247259.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.306151485443115,
|
|
"epoch": 0.0087070091423596,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 9.2634,
|
|
"mean_token_accuracy": 0.05518925692886114,
|
|
"num_tokens": 258100.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.27144947052002,
|
|
"epoch": 0.00904189410937343,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 9.1445,
|
|
"mean_token_accuracy": 0.061012673750519754,
|
|
"num_tokens": 267620.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.183198928833008,
|
|
"epoch": 0.009376779076387261,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 9.0077,
|
|
"mean_token_accuracy": 0.06654397621750832,
|
|
"num_tokens": 277654.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.065981578826904,
|
|
"epoch": 0.009711664043401092,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 9.0075,
|
|
"mean_token_accuracy": 0.05672478042542935,
|
|
"num_tokens": 288195.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.022527408599853,
|
|
"epoch": 0.010046549010414923,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.9382,
|
|
"mean_token_accuracy": 0.05744555629789829,
|
|
"num_tokens": 298860.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.896670055389404,
|
|
"epoch": 0.010381433977428753,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.8435,
|
|
"mean_token_accuracy": 0.060239487513899806,
|
|
"num_tokens": 307799.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.799404430389405,
|
|
"epoch": 0.010716318944442584,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.7928,
|
|
"mean_token_accuracy": 0.05781174004077912,
|
|
"num_tokens": 318257.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.696899318695069,
|
|
"epoch": 0.011051203911456415,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.7536,
|
|
"mean_token_accuracy": 0.05891339629888535,
|
|
"num_tokens": 329065.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.617450332641601,
|
|
"epoch": 0.011386088878470245,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.6708,
|
|
"mean_token_accuracy": 0.05866752900183201,
|
|
"num_tokens": 338809.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.496949386596679,
|
|
"epoch": 0.011720973845484076,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.6339,
|
|
"mean_token_accuracy": 0.05763169266283512,
|
|
"num_tokens": 349318.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 9.336700916290283,
|
|
"epoch": 0.012055858812497907,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.5925,
|
|
"mean_token_accuracy": 0.060683618485927585,
|
|
"num_tokens": 358760.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 9.284276008605957,
|
|
"epoch": 0.012390743779511737,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.4743,
|
|
"mean_token_accuracy": 0.060685327649116515,
|
|
"num_tokens": 367589.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 9.175062084197998,
|
|
"epoch": 0.012725628746525568,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.4897,
|
|
"mean_token_accuracy": 0.06429480686783791,
|
|
"num_tokens": 377012.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 9.043092918395995,
|
|
"epoch": 0.013060513713539399,
|
|
"grad_norm": 0.71484375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.4493,
|
|
"mean_token_accuracy": 0.06497252248227596,
|
|
"num_tokens": 386791.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 9.034938144683839,
|
|
"epoch": 0.01339539868055323,
|
|
"grad_norm": 0.7734375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.5439,
|
|
"mean_token_accuracy": 0.057699060812592504,
|
|
"num_tokens": 396620.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 9.003680610656739,
|
|
"epoch": 0.01373028364756706,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.508,
|
|
"mean_token_accuracy": 0.05839668586850166,
|
|
"num_tokens": 406384.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.98305311203003,
|
|
"epoch": 0.01406516861458089,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.4619,
|
|
"mean_token_accuracy": 0.06313387341797352,
|
|
"num_tokens": 416481.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.884008121490478,
|
|
"epoch": 0.014400053581594721,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.4978,
|
|
"mean_token_accuracy": 0.06396310143172741,
|
|
"num_tokens": 425697.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.91770076751709,
|
|
"epoch": 0.014734938548608552,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.4105,
|
|
"mean_token_accuracy": 0.06354090645909309,
|
|
"num_tokens": 435836.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.9373610496521,
|
|
"epoch": 0.015069823515622383,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.4483,
|
|
"mean_token_accuracy": 0.06257182881236076,
|
|
"num_tokens": 445926.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.854573345184326,
|
|
"epoch": 0.015404708482636215,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.4182,
|
|
"mean_token_accuracy": 0.06113107353448868,
|
|
"num_tokens": 455594.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.783684635162354,
|
|
"epoch": 0.015739593449650046,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.4544,
|
|
"mean_token_accuracy": 0.0631777100265026,
|
|
"num_tokens": 465928.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.864254093170166,
|
|
"epoch": 0.016074478416663877,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.3855,
|
|
"mean_token_accuracy": 0.0648330245167017,
|
|
"num_tokens": 475946.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.782320022583008,
|
|
"epoch": 0.016409363383677707,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.4001,
|
|
"mean_token_accuracy": 0.06609157137572766,
|
|
"num_tokens": 485285.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.789734363555908,
|
|
"epoch": 0.016744248350691538,
|
|
"grad_norm": 0.7734375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.3957,
|
|
"mean_token_accuracy": 0.06265997141599655,
|
|
"num_tokens": 496516.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.808907222747802,
|
|
"epoch": 0.01707913331770537,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.3818,
|
|
"mean_token_accuracy": 0.06361959911882878,
|
|
"num_tokens": 506330.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.74974603652954,
|
|
"epoch": 0.0174140182847192,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.4313,
|
|
"mean_token_accuracy": 0.0675040852278471,
|
|
"num_tokens": 515825.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.786038494110107,
|
|
"epoch": 0.01774890325173303,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.3628,
|
|
"mean_token_accuracy": 0.06703696362674236,
|
|
"num_tokens": 526303.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.757538318634033,
|
|
"epoch": 0.01808378821874686,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.4598,
|
|
"mean_token_accuracy": 0.06852960139513016,
|
|
"num_tokens": 536805.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.789075469970703,
|
|
"epoch": 0.01841867318576069,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.3273,
|
|
"mean_token_accuracy": 0.07035987079143524,
|
|
"num_tokens": 546381.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.71272668838501,
|
|
"epoch": 0.018753558152774522,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.3438,
|
|
"mean_token_accuracy": 0.06675126180052757,
|
|
"num_tokens": 556278.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.705748081207275,
|
|
"epoch": 0.019088443119788353,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.3522,
|
|
"mean_token_accuracy": 0.07065789699554444,
|
|
"num_tokens": 565312.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.733772277832031,
|
|
"epoch": 0.019423328086802184,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.319,
|
|
"mean_token_accuracy": 0.07612328678369522,
|
|
"num_tokens": 575304.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.657926940917969,
|
|
"epoch": 0.019758213053816014,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.2978,
|
|
"mean_token_accuracy": 0.06872594207525254,
|
|
"num_tokens": 585262.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.737933158874512,
|
|
"epoch": 0.020093098020829845,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.3275,
|
|
"mean_token_accuracy": 0.0715562004595995,
|
|
"num_tokens": 595297.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.71183385848999,
|
|
"epoch": 0.020427982987843676,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.3075,
|
|
"mean_token_accuracy": 0.07006572633981704,
|
|
"num_tokens": 604354.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.618387031555176,
|
|
"epoch": 0.020762867954857506,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.4043,
|
|
"mean_token_accuracy": 0.0666431374847889,
|
|
"num_tokens": 616283.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.783961200714112,
|
|
"epoch": 0.021097752921871337,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.322,
|
|
"mean_token_accuracy": 0.06938845105469227,
|
|
"num_tokens": 626059.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.627076625823975,
|
|
"epoch": 0.021432637888885168,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.2453,
|
|
"mean_token_accuracy": 0.07488406710326671,
|
|
"num_tokens": 634483.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.68146858215332,
|
|
"epoch": 0.021767522855899,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.2387,
|
|
"mean_token_accuracy": 0.07477111965417862,
|
|
"num_tokens": 643111.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.574335098266602,
|
|
"epoch": 0.02210240782291283,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.2554,
|
|
"mean_token_accuracy": 0.07630948387086392,
|
|
"num_tokens": 653650.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.697111225128173,
|
|
"epoch": 0.02243729278992666,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 8.252,
|
|
"mean_token_accuracy": 0.08444391936063766,
|
|
"num_tokens": 663143.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.565497016906738,
|
|
"epoch": 0.02277217775694049,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.2062,
|
|
"mean_token_accuracy": 0.07809413745999336,
|
|
"num_tokens": 672434.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.655457496643066,
|
|
"epoch": 0.02310706272395432,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 8.2254,
|
|
"mean_token_accuracy": 0.0773940771818161,
|
|
"num_tokens": 682973.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.558607482910157,
|
|
"epoch": 0.023441947690968152,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 8.2771,
|
|
"mean_token_accuracy": 0.07662329860031605,
|
|
"num_tokens": 692136.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.668216800689697,
|
|
"epoch": 0.023776832657981983,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 8.1824,
|
|
"mean_token_accuracy": 0.08042559921741485,
|
|
"num_tokens": 702257.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.555757141113281,
|
|
"epoch": 0.024111717624995813,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 8.1992,
|
|
"mean_token_accuracy": 0.07901003137230873,
|
|
"num_tokens": 712444.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.65919303894043,
|
|
"epoch": 0.024446602592009644,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.3348,
|
|
"mean_token_accuracy": 0.0759161863476038,
|
|
"num_tokens": 721614.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.671118354797363,
|
|
"epoch": 0.024781487559023475,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 8.3212,
|
|
"mean_token_accuracy": 0.07223852351307869,
|
|
"num_tokens": 731189.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.591705894470214,
|
|
"epoch": 0.025116372526037305,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 8.2213,
|
|
"mean_token_accuracy": 0.08042538389563561,
|
|
"num_tokens": 741555.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.528362846374511,
|
|
"epoch": 0.025451257493051136,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 8.1569,
|
|
"mean_token_accuracy": 0.08212503418326378,
|
|
"num_tokens": 751560.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.550797271728516,
|
|
"epoch": 0.025786142460064967,
|
|
"grad_norm": 0.80078125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 8.2041,
|
|
"mean_token_accuracy": 0.07558429762721061,
|
|
"num_tokens": 761471.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.614824962615966,
|
|
"epoch": 0.026121027427078797,
|
|
"grad_norm": 0.83984375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 8.2015,
|
|
"mean_token_accuracy": 0.07853580340743065,
|
|
"num_tokens": 771565.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.519598484039307,
|
|
"epoch": 0.026455912394092628,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 8.1537,
|
|
"mean_token_accuracy": 0.07887223958969117,
|
|
"num_tokens": 781248.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.60389404296875,
|
|
"epoch": 0.02679079736110646,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 8.2014,
|
|
"mean_token_accuracy": 0.075946843624115,
|
|
"num_tokens": 790907.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.511116981506348,
|
|
"epoch": 0.02712568232812029,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000202,
|
|
"loss": 8.2054,
|
|
"mean_token_accuracy": 0.07778018116950988,
|
|
"num_tokens": 800073.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.635944366455078,
|
|
"epoch": 0.02746056729513412,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 8.1885,
|
|
"mean_token_accuracy": 0.07892166152596473,
|
|
"num_tokens": 809782.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.39308671951294,
|
|
"epoch": 0.02779545226214795,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 8.1461,
|
|
"mean_token_accuracy": 0.07830736264586449,
|
|
"num_tokens": 820329.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.558973217010498,
|
|
"epoch": 0.02813033722916178,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 8.0641,
|
|
"mean_token_accuracy": 0.08698545768857002,
|
|
"num_tokens": 829405.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.460458278656006,
|
|
"epoch": 0.028465222196175612,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 8.12,
|
|
"mean_token_accuracy": 0.077915009111166,
|
|
"num_tokens": 838700.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.557555007934571,
|
|
"epoch": 0.028800107163189443,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 8.1831,
|
|
"mean_token_accuracy": 0.08165658637881279,
|
|
"num_tokens": 848892.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.484703731536865,
|
|
"epoch": 0.029134992130203274,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 8.1107,
|
|
"mean_token_accuracy": 0.08097934126853942,
|
|
"num_tokens": 857737.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.450378036499023,
|
|
"epoch": 0.029469877097217104,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 8.043,
|
|
"mean_token_accuracy": 0.0830142006278038,
|
|
"num_tokens": 867758.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.483449363708496,
|
|
"epoch": 0.029804762064230935,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 8.0694,
|
|
"mean_token_accuracy": 0.0823376640677452,
|
|
"num_tokens": 877273.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.511912822723389,
|
|
"epoch": 0.030139647031244766,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 8.1597,
|
|
"mean_token_accuracy": 0.0831824667751789,
|
|
"num_tokens": 887109.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.448967266082764,
|
|
"epoch": 0.030474531998258596,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 8.0778,
|
|
"mean_token_accuracy": 0.08146599680185318,
|
|
"num_tokens": 896157.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.426976585388184,
|
|
"epoch": 0.03080941696527243,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 8.0909,
|
|
"mean_token_accuracy": 0.07872407361865044,
|
|
"num_tokens": 907231.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.366917896270753,
|
|
"epoch": 0.03114430193228626,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.997,
|
|
"mean_token_accuracy": 0.08565031662583351,
|
|
"num_tokens": 916756.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 8.496291255950927,
|
|
"epoch": 0.03147918689930009,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 8.17,
|
|
"mean_token_accuracy": 0.0742259133607149,
|
|
"num_tokens": 927714.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.47745065689087,
|
|
"epoch": 0.03181407186631392,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 8.005,
|
|
"mean_token_accuracy": 0.0867729477584362,
|
|
"num_tokens": 936620.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.397373008728028,
|
|
"epoch": 0.03214895683332775,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 8.0433,
|
|
"mean_token_accuracy": 0.08084618747234344,
|
|
"num_tokens": 945942.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.336866855621338,
|
|
"epoch": 0.032483841800341584,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000242,
|
|
"loss": 8.0883,
|
|
"mean_token_accuracy": 0.08643392398953438,
|
|
"num_tokens": 956685.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.401565170288086,
|
|
"epoch": 0.032818726767355415,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 8.0316,
|
|
"mean_token_accuracy": 0.08411954194307328,
|
|
"num_tokens": 967453.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.446605682373047,
|
|
"epoch": 0.033153611734369245,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 8.0546,
|
|
"mean_token_accuracy": 0.083030104637146,
|
|
"num_tokens": 978220.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.329526329040528,
|
|
"epoch": 0.033488496701383076,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 8.0053,
|
|
"mean_token_accuracy": 0.08742882534861565,
|
|
"num_tokens": 988120.0,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 4000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 233515288166400.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|