00e53abac2
Model: fpadovani/nld-latn-100mb-after-ppt-shuff-dyck-10mb-ckpt500_seed3407 Source: Original Platform
1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.45433893684688775,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.7617779731750485,
|
|
"epoch": 0.004543389368468878,
|
|
"grad_norm": 12.625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.4213,
|
|
"mean_token_accuracy": 0.00039867511950433256,
|
|
"num_tokens": 10156.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.762024164199829,
|
|
"epoch": 0.009086778736937756,
|
|
"grad_norm": 14.375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.3898,
|
|
"mean_token_accuracy": 8.278145687654615e-05,
|
|
"num_tokens": 20669.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.801386165618896,
|
|
"epoch": 0.013630168105406633,
|
|
"grad_norm": 16.875,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.1381,
|
|
"mean_token_accuracy": 0.00035612244391813874,
|
|
"num_tokens": 30952.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 5.016745567321777,
|
|
"epoch": 0.01817355747387551,
|
|
"grad_norm": 28.125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.5526,
|
|
"mean_token_accuracy": 0.0001651436381507665,
|
|
"num_tokens": 41429.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 7.075106763839722,
|
|
"epoch": 0.02271694684234439,
|
|
"grad_norm": 19.75,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 11.9297,
|
|
"mean_token_accuracy": 0.00024400593247264624,
|
|
"num_tokens": 52166.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.367570781707764,
|
|
"epoch": 0.027260336210813266,
|
|
"grad_norm": 3.546875,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.747,
|
|
"mean_token_accuracy": 0.004502167762257159,
|
|
"num_tokens": 61670.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.692459964752198,
|
|
"epoch": 0.03180372557928214,
|
|
"grad_norm": 3.0625,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.4764,
|
|
"mean_token_accuracy": 0.01658898275345564,
|
|
"num_tokens": 71702.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.684140300750732,
|
|
"epoch": 0.03634711494775102,
|
|
"grad_norm": 2.8125,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.1391,
|
|
"mean_token_accuracy": 0.01516802804544568,
|
|
"num_tokens": 82840.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.62061595916748,
|
|
"epoch": 0.0408905043162199,
|
|
"grad_norm": 2.578125,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.8367,
|
|
"mean_token_accuracy": 0.03998391982167959,
|
|
"num_tokens": 92290.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.570560836791993,
|
|
"epoch": 0.04543389368468878,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.6874,
|
|
"mean_token_accuracy": 0.04625674188137054,
|
|
"num_tokens": 101842.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.557878017425537,
|
|
"epoch": 0.049977283053157656,
|
|
"grad_norm": 1.921875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.6545,
|
|
"mean_token_accuracy": 0.043805907294154164,
|
|
"num_tokens": 114030.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.589654445648193,
|
|
"epoch": 0.05452067242162653,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.5177,
|
|
"mean_token_accuracy": 0.04366073869168759,
|
|
"num_tokens": 124349.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.59289197921753,
|
|
"epoch": 0.059064061790095414,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.3936,
|
|
"mean_token_accuracy": 0.044536291435360906,
|
|
"num_tokens": 133534.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.57084617614746,
|
|
"epoch": 0.06360745115856428,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.3062,
|
|
"mean_token_accuracy": 0.041312586329877374,
|
|
"num_tokens": 144383.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.55383882522583,
|
|
"epoch": 0.06815084052703317,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.2925,
|
|
"mean_token_accuracy": 0.04508417844772339,
|
|
"num_tokens": 155467.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.524444103240967,
|
|
"epoch": 0.07269422989550205,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.1809,
|
|
"mean_token_accuracy": 0.045108040794730186,
|
|
"num_tokens": 165903.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.497136688232422,
|
|
"epoch": 0.07723761926397092,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.0687,
|
|
"mean_token_accuracy": 0.04266926664859057,
|
|
"num_tokens": 175812.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.480645847320556,
|
|
"epoch": 0.0817810086324398,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.9456,
|
|
"mean_token_accuracy": 0.0453936118632555,
|
|
"num_tokens": 186357.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.449639892578125,
|
|
"epoch": 0.08632439800090867,
|
|
"grad_norm": 1.9765625,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.8607,
|
|
"mean_token_accuracy": 0.05368485823273659,
|
|
"num_tokens": 196124.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.369837188720703,
|
|
"epoch": 0.09086778736937756,
|
|
"grad_norm": 1.953125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.7331,
|
|
"mean_token_accuracy": 0.056761689856648444,
|
|
"num_tokens": 205655.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.268289566040039,
|
|
"epoch": 0.09541117673784644,
|
|
"grad_norm": 1.8203125,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.6528,
|
|
"mean_token_accuracy": 0.05442145057022572,
|
|
"num_tokens": 215870.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.158163261413574,
|
|
"epoch": 0.09995456610631531,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.522,
|
|
"mean_token_accuracy": 0.05480169989168644,
|
|
"num_tokens": 225027.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.10291576385498,
|
|
"epoch": 0.10449795547478419,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.411,
|
|
"mean_token_accuracy": 0.05560419633984566,
|
|
"num_tokens": 236352.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.994441699981689,
|
|
"epoch": 0.10904134484325306,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.2637,
|
|
"mean_token_accuracy": 0.05579149127006531,
|
|
"num_tokens": 246713.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.80837516784668,
|
|
"epoch": 0.11358473421172194,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.0811,
|
|
"mean_token_accuracy": 0.057105731964111325,
|
|
"num_tokens": 256950.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.595962619781494,
|
|
"epoch": 0.11812812358019083,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.0346,
|
|
"mean_token_accuracy": 0.05215160399675369,
|
|
"num_tokens": 267345.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.378811168670655,
|
|
"epoch": 0.1226715129486597,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.8305,
|
|
"mean_token_accuracy": 0.059673656895756724,
|
|
"num_tokens": 276782.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.09063777923584,
|
|
"epoch": 0.12721490231712856,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.7044,
|
|
"mean_token_accuracy": 0.06308954879641533,
|
|
"num_tokens": 287024.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.829156970977783,
|
|
"epoch": 0.13175829168559747,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.7479,
|
|
"mean_token_accuracy": 0.06074633449316025,
|
|
"num_tokens": 296597.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.660122108459472,
|
|
"epoch": 0.13630168105406634,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.5571,
|
|
"mean_token_accuracy": 0.06737130098044872,
|
|
"num_tokens": 307709.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.354702949523926,
|
|
"epoch": 0.14084507042253522,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.5917,
|
|
"mean_token_accuracy": 0.07087288200855255,
|
|
"num_tokens": 317976.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.30783452987671,
|
|
"epoch": 0.1453884597910041,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.462,
|
|
"mean_token_accuracy": 0.0719795323908329,
|
|
"num_tokens": 327319.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.110205173492432,
|
|
"epoch": 0.14993184915947297,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.4455,
|
|
"mean_token_accuracy": 0.07500142268836499,
|
|
"num_tokens": 337337.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 7.971628332138062,
|
|
"epoch": 0.15447523852794184,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.4838,
|
|
"mean_token_accuracy": 0.06365859620273114,
|
|
"num_tokens": 347779.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 7.962876701354981,
|
|
"epoch": 0.15901862789641072,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.5019,
|
|
"mean_token_accuracy": 0.0738484513014555,
|
|
"num_tokens": 358088.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 7.972894763946533,
|
|
"epoch": 0.1635620172648796,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.4157,
|
|
"mean_token_accuracy": 0.08021429628133774,
|
|
"num_tokens": 368314.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 7.827826261520386,
|
|
"epoch": 0.16810540663334847,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.3112,
|
|
"mean_token_accuracy": 0.08391624316573143,
|
|
"num_tokens": 378406.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 7.8609589576721195,
|
|
"epoch": 0.17264879600181735,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.3269,
|
|
"mean_token_accuracy": 0.08003454320132733,
|
|
"num_tokens": 388912.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 7.787754869461059,
|
|
"epoch": 0.17719218537028622,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.3277,
|
|
"mean_token_accuracy": 0.08325541540980338,
|
|
"num_tokens": 399290.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.774484252929687,
|
|
"epoch": 0.18173557473875512,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.242,
|
|
"mean_token_accuracy": 0.08362203016877175,
|
|
"num_tokens": 409757.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.7403558731079105,
|
|
"epoch": 0.186278964107224,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.4135,
|
|
"mean_token_accuracy": 0.08218215703964234,
|
|
"num_tokens": 420182.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.780586194992066,
|
|
"epoch": 0.19082235347569287,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.2991,
|
|
"mean_token_accuracy": 0.07898701690137386,
|
|
"num_tokens": 429897.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.728657341003418,
|
|
"epoch": 0.19536574284416175,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.1992,
|
|
"mean_token_accuracy": 0.08197759091854095,
|
|
"num_tokens": 441131.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.609148836135864,
|
|
"epoch": 0.19990913221263062,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.1791,
|
|
"mean_token_accuracy": 0.08322597444057464,
|
|
"num_tokens": 451894.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.640337753295898,
|
|
"epoch": 0.2044525215810995,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.1004,
|
|
"mean_token_accuracy": 0.08846881911158562,
|
|
"num_tokens": 462613.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.625602769851684,
|
|
"epoch": 0.20899591094956838,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.2065,
|
|
"mean_token_accuracy": 0.08791121542453766,
|
|
"num_tokens": 471606.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.539081668853759,
|
|
"epoch": 0.21353930031803725,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.1044,
|
|
"mean_token_accuracy": 0.09239566922187806,
|
|
"num_tokens": 481559.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.6146509647369385,
|
|
"epoch": 0.21808268968650613,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.123,
|
|
"mean_token_accuracy": 0.08494666814804078,
|
|
"num_tokens": 492766.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.549022054672241,
|
|
"epoch": 0.222626079054975,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.1902,
|
|
"mean_token_accuracy": 0.08542409986257553,
|
|
"num_tokens": 503357.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.545317697525024,
|
|
"epoch": 0.22716946842344388,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.0221,
|
|
"mean_token_accuracy": 0.0891069382429123,
|
|
"num_tokens": 513098.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.549406671524048,
|
|
"epoch": 0.23171285779191278,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.1016,
|
|
"mean_token_accuracy": 0.08548958823084832,
|
|
"num_tokens": 522807.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.560564756393433,
|
|
"epoch": 0.23625624716038165,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.1193,
|
|
"mean_token_accuracy": 0.08555800542235374,
|
|
"num_tokens": 532670.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.523266553878784,
|
|
"epoch": 0.24079963652885053,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.0547,
|
|
"mean_token_accuracy": 0.08636701181530952,
|
|
"num_tokens": 543647.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.513870096206665,
|
|
"epoch": 0.2453430258973194,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.095,
|
|
"mean_token_accuracy": 0.09233981817960739,
|
|
"num_tokens": 553047.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.446543216705322,
|
|
"epoch": 0.24988641526578828,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.1552,
|
|
"mean_token_accuracy": 0.09209886938333511,
|
|
"num_tokens": 563413.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.474071645736695,
|
|
"epoch": 0.25442980463425713,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.1034,
|
|
"mean_token_accuracy": 0.08880794793367386,
|
|
"num_tokens": 573790.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.525774335861206,
|
|
"epoch": 0.25897319400272606,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.0109,
|
|
"mean_token_accuracy": 0.08925304189324379,
|
|
"num_tokens": 582958.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.388642263412476,
|
|
"epoch": 0.26351658337119493,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 6.9668,
|
|
"mean_token_accuracy": 0.0980121873319149,
|
|
"num_tokens": 592886.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.4521583080291744,
|
|
"epoch": 0.2680599727396638,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.013,
|
|
"mean_token_accuracy": 0.09019459635019303,
|
|
"num_tokens": 603692.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.435794734954834,
|
|
"epoch": 0.2726033621081327,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.0484,
|
|
"mean_token_accuracy": 0.09026054739952087,
|
|
"num_tokens": 613534.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.424719905853271,
|
|
"epoch": 0.27714675147660156,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.9995,
|
|
"mean_token_accuracy": 0.09151231348514557,
|
|
"num_tokens": 623862.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.514718866348266,
|
|
"epoch": 0.28169014084507044,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.0863,
|
|
"mean_token_accuracy": 0.08810076713562012,
|
|
"num_tokens": 633759.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.361165714263916,
|
|
"epoch": 0.2862335302135393,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 6.9699,
|
|
"mean_token_accuracy": 0.09024499729275703,
|
|
"num_tokens": 643553.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.373637962341308,
|
|
"epoch": 0.2907769195820082,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.8834,
|
|
"mean_token_accuracy": 0.09637065082788468,
|
|
"num_tokens": 653714.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.3554847717285154,
|
|
"epoch": 0.29532030895047706,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.9076,
|
|
"mean_token_accuracy": 0.09380478709936142,
|
|
"num_tokens": 662817.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.294995498657227,
|
|
"epoch": 0.29986369831894594,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.9342,
|
|
"mean_token_accuracy": 0.0921212948858738,
|
|
"num_tokens": 672381.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.2491357803344725,
|
|
"epoch": 0.3044070876874148,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.8159,
|
|
"mean_token_accuracy": 0.09338896572589875,
|
|
"num_tokens": 682525.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.309194135665893,
|
|
"epoch": 0.3089504770558837,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.91,
|
|
"mean_token_accuracy": 0.09586632996797562,
|
|
"num_tokens": 691826.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.192427301406861,
|
|
"epoch": 0.31349386642435256,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.8849,
|
|
"mean_token_accuracy": 0.10212687999010087,
|
|
"num_tokens": 701319.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.242300987243652,
|
|
"epoch": 0.31803725579282144,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.9811,
|
|
"mean_token_accuracy": 0.0917223945260048,
|
|
"num_tokens": 710888.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.351683664321899,
|
|
"epoch": 0.3225806451612903,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.9156,
|
|
"mean_token_accuracy": 0.09456708803772926,
|
|
"num_tokens": 722516.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.288315916061402,
|
|
"epoch": 0.3271240345297592,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.8905,
|
|
"mean_token_accuracy": 0.09565110653638839,
|
|
"num_tokens": 733058.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.249158096313477,
|
|
"epoch": 0.33166742389822806,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.959,
|
|
"mean_token_accuracy": 0.09808666259050369,
|
|
"num_tokens": 744832.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.257345390319824,
|
|
"epoch": 0.33621081326669694,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.8852,
|
|
"mean_token_accuracy": 0.09970412328839302,
|
|
"num_tokens": 755387.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.181437587738037,
|
|
"epoch": 0.3407542026351658,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.8328,
|
|
"mean_token_accuracy": 0.0964855819940567,
|
|
"num_tokens": 764455.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.1537618160247805,
|
|
"epoch": 0.3452975920036347,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.8678,
|
|
"mean_token_accuracy": 0.09778558984398841,
|
|
"num_tokens": 774513.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.281053781509399,
|
|
"epoch": 0.34984098137210357,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.9626,
|
|
"mean_token_accuracy": 0.09827063530683518,
|
|
"num_tokens": 785540.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.215037727355957,
|
|
"epoch": 0.35438437074057244,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.8204,
|
|
"mean_token_accuracy": 0.09672441035509109,
|
|
"num_tokens": 795233.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.275153923034668,
|
|
"epoch": 0.35892776010904137,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.8705,
|
|
"mean_token_accuracy": 0.09623145386576652,
|
|
"num_tokens": 804855.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.20860447883606,
|
|
"epoch": 0.36347114947751025,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.8266,
|
|
"mean_token_accuracy": 0.10111764296889306,
|
|
"num_tokens": 814818.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.1098264217376705,
|
|
"epoch": 0.3680145388459791,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.8627,
|
|
"mean_token_accuracy": 0.10245592221617698,
|
|
"num_tokens": 824622.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.1824994564056395,
|
|
"epoch": 0.372557928214448,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.7427,
|
|
"mean_token_accuracy": 0.10741053223609924,
|
|
"num_tokens": 834813.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.061726331710815,
|
|
"epoch": 0.3771013175829169,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.8567,
|
|
"mean_token_accuracy": 0.10019884258508682,
|
|
"num_tokens": 844887.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.173247957229615,
|
|
"epoch": 0.38164470695138575,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.8447,
|
|
"mean_token_accuracy": 0.10034631565213203,
|
|
"num_tokens": 855430.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.076598262786865,
|
|
"epoch": 0.3861880963198546,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.6801,
|
|
"mean_token_accuracy": 0.10056375861167907,
|
|
"num_tokens": 866052.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.042910146713257,
|
|
"epoch": 0.3907314856883235,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.8356,
|
|
"mean_token_accuracy": 0.09405270889401436,
|
|
"num_tokens": 877961.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.111201524734497,
|
|
"epoch": 0.3952748750567924,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.8019,
|
|
"mean_token_accuracy": 0.10706538185477257,
|
|
"num_tokens": 888386.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.103235197067261,
|
|
"epoch": 0.39981826442526125,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.8377,
|
|
"mean_token_accuracy": 0.10170447900891304,
|
|
"num_tokens": 898272.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.130263900756836,
|
|
"epoch": 0.4043616537937301,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.7814,
|
|
"mean_token_accuracy": 0.1046774573624134,
|
|
"num_tokens": 908317.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.089518404006958,
|
|
"epoch": 0.408905043162199,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.7756,
|
|
"mean_token_accuracy": 0.09967579692602158,
|
|
"num_tokens": 917970.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.043964099884033,
|
|
"epoch": 0.4134484325306679,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.8119,
|
|
"mean_token_accuracy": 0.0967039294540882,
|
|
"num_tokens": 928132.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.043747615814209,
|
|
"epoch": 0.41799182189913675,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.7437,
|
|
"mean_token_accuracy": 0.10632947832345963,
|
|
"num_tokens": 938092.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.126165199279785,
|
|
"epoch": 0.4225352112676056,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.8483,
|
|
"mean_token_accuracy": 0.09798347502946854,
|
|
"num_tokens": 948650.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 6.976731872558593,
|
|
"epoch": 0.4270786006360745,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.77,
|
|
"mean_token_accuracy": 0.10464980453252792,
|
|
"num_tokens": 960195.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.047196292877198,
|
|
"epoch": 0.4316219900045434,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.7843,
|
|
"mean_token_accuracy": 0.10333740636706353,
|
|
"num_tokens": 970608.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.10983419418335,
|
|
"epoch": 0.43616537937301225,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.7567,
|
|
"mean_token_accuracy": 0.10173246711492538,
|
|
"num_tokens": 981053.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 6.920445919036865,
|
|
"epoch": 0.4407087687414811,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.6903,
|
|
"mean_token_accuracy": 0.10761705115437507,
|
|
"num_tokens": 991560.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 6.985012483596802,
|
|
"epoch": 0.44525215810995,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.7318,
|
|
"mean_token_accuracy": 0.09935669675469398,
|
|
"num_tokens": 1002008.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.016558361053467,
|
|
"epoch": 0.4497955474784189,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.7297,
|
|
"mean_token_accuracy": 0.11214710623025895,
|
|
"num_tokens": 1012349.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 6.973590946197509,
|
|
"epoch": 0.45433893684688775,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.7202,
|
|
"mean_token_accuracy": 0.10605021119117737,
|
|
"num_tokens": 1022886.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.45433893684688775,
|
|
"eval_entropy": 6.977851612623348,
|
|
"eval_loss": 6.848958492279053,
|
|
"eval_mean_token_accuracy": 0.10486862519539372,
|
|
"eval_num_tokens": 1022886.0,
|
|
"eval_runtime": 2.6621,
|
|
"eval_samples_per_second": 1290.327,
|
|
"eval_steps_per_second": 161.526,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 11000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 10,
|
|
"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": 1493455435776000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|