1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.5931198102016607,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.6509840965271,
|
|
"epoch": 0.005931198102016607,
|
|
"grad_norm": 11.4375,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7955,
|
|
"mean_token_accuracy": 0.00017467249417677522,
|
|
"num_tokens": 12647.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.650976181030273,
|
|
"epoch": 0.011862396204033215,
|
|
"grad_norm": 10.875,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7449,
|
|
"mean_token_accuracy": 0.0003256943717133254,
|
|
"num_tokens": 26668.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.65047550201416,
|
|
"epoch": 0.017793594306049824,
|
|
"grad_norm": 8.5,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.5392,
|
|
"mean_token_accuracy": 0.011320227157557382,
|
|
"num_tokens": 41021.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.645245170593261,
|
|
"epoch": 0.02372479240806643,
|
|
"grad_norm": 5.6875,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.2512,
|
|
"mean_token_accuracy": 0.03331172131001949,
|
|
"num_tokens": 56626.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.620162200927734,
|
|
"epoch": 0.02965599051008304,
|
|
"grad_norm": 3.46875,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 9.9999,
|
|
"mean_token_accuracy": 0.03278177715837956,
|
|
"num_tokens": 72914.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.59833517074585,
|
|
"epoch": 0.03558718861209965,
|
|
"grad_norm": 2.75,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.8507,
|
|
"mean_token_accuracy": 0.03321463130414486,
|
|
"num_tokens": 88514.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.59533166885376,
|
|
"epoch": 0.04151838671411625,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.7557,
|
|
"mean_token_accuracy": 0.03128496501594782,
|
|
"num_tokens": 101256.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.598009204864502,
|
|
"epoch": 0.04744958481613286,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.7229,
|
|
"mean_token_accuracy": 0.03391152955591679,
|
|
"num_tokens": 114459.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.583034133911132,
|
|
"epoch": 0.05338078291814947,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.7096,
|
|
"mean_token_accuracy": 0.031900213845074174,
|
|
"num_tokens": 129858.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.583144092559815,
|
|
"epoch": 0.05931198102016608,
|
|
"grad_norm": 2.421875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.6353,
|
|
"mean_token_accuracy": 0.03371269796043634,
|
|
"num_tokens": 143486.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.582204055786132,
|
|
"epoch": 0.06524317912218268,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.5795,
|
|
"mean_token_accuracy": 0.035220497474074366,
|
|
"num_tokens": 156003.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.569349002838134,
|
|
"epoch": 0.0711743772241993,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.5437,
|
|
"mean_token_accuracy": 0.034727090038359165,
|
|
"num_tokens": 170988.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.552780246734619,
|
|
"epoch": 0.0771055753262159,
|
|
"grad_norm": 2.15625,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.4193,
|
|
"mean_token_accuracy": 0.03448135014623403,
|
|
"num_tokens": 187165.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.542837047576905,
|
|
"epoch": 0.0830367734282325,
|
|
"grad_norm": 2.125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.368,
|
|
"mean_token_accuracy": 0.03228826243430376,
|
|
"num_tokens": 202072.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.534712505340575,
|
|
"epoch": 0.08896797153024912,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.2131,
|
|
"mean_token_accuracy": 0.034324941039085385,
|
|
"num_tokens": 215375.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.50705509185791,
|
|
"epoch": 0.09489916963226572,
|
|
"grad_norm": 1.921875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.1257,
|
|
"mean_token_accuracy": 0.03353528436273336,
|
|
"num_tokens": 228749.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.478670978546143,
|
|
"epoch": 0.10083036773428232,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.0183,
|
|
"mean_token_accuracy": 0.03314766474068165,
|
|
"num_tokens": 242387.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.458618545532227,
|
|
"epoch": 0.10676156583629894,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.9252,
|
|
"mean_token_accuracy": 0.03416665960103273,
|
|
"num_tokens": 257793.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.419714641571044,
|
|
"epoch": 0.11269276393831554,
|
|
"grad_norm": 2.125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.8134,
|
|
"mean_token_accuracy": 0.036843133717775346,
|
|
"num_tokens": 271187.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.339355945587158,
|
|
"epoch": 0.11862396204033215,
|
|
"grad_norm": 1.921875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.6836,
|
|
"mean_token_accuracy": 0.04052316285669803,
|
|
"num_tokens": 284937.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.289305877685546,
|
|
"epoch": 0.12455516014234876,
|
|
"grad_norm": 2.140625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.523,
|
|
"mean_token_accuracy": 0.042971277609467506,
|
|
"num_tokens": 300108.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.171844291687012,
|
|
"epoch": 0.13048635824436536,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.4071,
|
|
"mean_token_accuracy": 0.0435908630490303,
|
|
"num_tokens": 316760.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.090029907226562,
|
|
"epoch": 0.13641755634638197,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.2269,
|
|
"mean_token_accuracy": 0.04757884480059147,
|
|
"num_tokens": 329705.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.968478775024414,
|
|
"epoch": 0.1423487544483986,
|
|
"grad_norm": 1.8046875,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.1048,
|
|
"mean_token_accuracy": 0.05177154988050461,
|
|
"num_tokens": 342947.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.810734272003174,
|
|
"epoch": 0.14827995255041518,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 7.957,
|
|
"mean_token_accuracy": 0.05589810460805893,
|
|
"num_tokens": 358397.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.629272270202637,
|
|
"epoch": 0.1542111506524318,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.8543,
|
|
"mean_token_accuracy": 0.058608030527830125,
|
|
"num_tokens": 370369.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.431285285949707,
|
|
"epoch": 0.1601423487544484,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.7282,
|
|
"mean_token_accuracy": 0.06019565127789974,
|
|
"num_tokens": 386896.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.166491031646729,
|
|
"epoch": 0.166073546856465,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.571,
|
|
"mean_token_accuracy": 0.06539936549961567,
|
|
"num_tokens": 400433.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.998315811157227,
|
|
"epoch": 0.17200474495848161,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.5193,
|
|
"mean_token_accuracy": 0.06127564944326878,
|
|
"num_tokens": 415092.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.760649585723877,
|
|
"epoch": 0.17793594306049823,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.3847,
|
|
"mean_token_accuracy": 0.06551449336111545,
|
|
"num_tokens": 429793.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.53032569885254,
|
|
"epoch": 0.18386714116251482,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.3618,
|
|
"mean_token_accuracy": 0.07152369543910027,
|
|
"num_tokens": 442618.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.432501602172852,
|
|
"epoch": 0.18979833926453143,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.2263,
|
|
"mean_token_accuracy": 0.0709031268954277,
|
|
"num_tokens": 454976.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.288888359069825,
|
|
"epoch": 0.19572953736654805,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.2693,
|
|
"mean_token_accuracy": 0.06482412219047547,
|
|
"num_tokens": 467523.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.135635375976562,
|
|
"epoch": 0.20166073546856464,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.1729,
|
|
"mean_token_accuracy": 0.07200234681367874,
|
|
"num_tokens": 481391.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.090270137786865,
|
|
"epoch": 0.20759193357058126,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.147,
|
|
"mean_token_accuracy": 0.06691372841596603,
|
|
"num_tokens": 496697.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.008182764053345,
|
|
"epoch": 0.21352313167259787,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.1704,
|
|
"mean_token_accuracy": 0.07276545092463493,
|
|
"num_tokens": 510091.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 7.948391151428223,
|
|
"epoch": 0.21945432977461446,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.1071,
|
|
"mean_token_accuracy": 0.0736262746155262,
|
|
"num_tokens": 522168.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 7.860538482666016,
|
|
"epoch": 0.22538552787663108,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.0333,
|
|
"mean_token_accuracy": 0.0750058189034462,
|
|
"num_tokens": 535853.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 7.789449167251587,
|
|
"epoch": 0.2313167259786477,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.0532,
|
|
"mean_token_accuracy": 0.07566255033016205,
|
|
"num_tokens": 551082.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.820679521560669,
|
|
"epoch": 0.2372479240806643,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.0735,
|
|
"mean_token_accuracy": 0.0763503935188055,
|
|
"num_tokens": 567181.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.715967988967895,
|
|
"epoch": 0.2431791221826809,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 6.9882,
|
|
"mean_token_accuracy": 0.0811008334159851,
|
|
"num_tokens": 580689.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.671104192733765,
|
|
"epoch": 0.2491103202846975,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 6.9075,
|
|
"mean_token_accuracy": 0.08624595254659653,
|
|
"num_tokens": 594590.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.669063520431519,
|
|
"epoch": 0.25504151838671413,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 6.9659,
|
|
"mean_token_accuracy": 0.07704100087285042,
|
|
"num_tokens": 610577.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.621254777908325,
|
|
"epoch": 0.2609727164887307,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 6.9894,
|
|
"mean_token_accuracy": 0.0770975686609745,
|
|
"num_tokens": 627066.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.582563304901123,
|
|
"epoch": 0.2669039145907473,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 6.9863,
|
|
"mean_token_accuracy": 0.08030139058828353,
|
|
"num_tokens": 642601.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.5748522758483885,
|
|
"epoch": 0.27283511269276395,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 6.8436,
|
|
"mean_token_accuracy": 0.08795318827033043,
|
|
"num_tokens": 655615.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.51125283241272,
|
|
"epoch": 0.27876631079478054,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 6.8805,
|
|
"mean_token_accuracy": 0.08227636963129044,
|
|
"num_tokens": 670218.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.581772565841675,
|
|
"epoch": 0.2846975088967972,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 6.8693,
|
|
"mean_token_accuracy": 0.08509634733200074,
|
|
"num_tokens": 683661.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.414237833023071,
|
|
"epoch": 0.29062870699881377,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000122,
|
|
"loss": 6.8471,
|
|
"mean_token_accuracy": 0.08358772993087768,
|
|
"num_tokens": 699767.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.437202262878418,
|
|
"epoch": 0.29655990510083036,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 6.774,
|
|
"mean_token_accuracy": 0.09326352775096894,
|
|
"num_tokens": 713579.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.383161449432373,
|
|
"epoch": 0.302491103202847,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 6.8019,
|
|
"mean_token_accuracy": 0.08484338372945785,
|
|
"num_tokens": 727490.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.414105653762817,
|
|
"epoch": 0.3084223013048636,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 6.8822,
|
|
"mean_token_accuracy": 0.08486307561397552,
|
|
"num_tokens": 742477.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.386582565307617,
|
|
"epoch": 0.3143534994068802,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 6.7436,
|
|
"mean_token_accuracy": 0.08875370100140571,
|
|
"num_tokens": 757772.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.385796451568604,
|
|
"epoch": 0.3202846975088968,
|
|
"grad_norm": 1.515625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 6.795,
|
|
"mean_token_accuracy": 0.08812053054571152,
|
|
"num_tokens": 772898.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.329386568069458,
|
|
"epoch": 0.3262158956109134,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 6.7697,
|
|
"mean_token_accuracy": 0.08715710267424584,
|
|
"num_tokens": 788395.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.318710279464722,
|
|
"epoch": 0.33214709371293,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 6.83,
|
|
"mean_token_accuracy": 0.08070869520306587,
|
|
"num_tokens": 803413.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.188418292999268,
|
|
"epoch": 0.33807829181494664,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 6.6976,
|
|
"mean_token_accuracy": 0.09588895291090012,
|
|
"num_tokens": 817775.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.389742612838745,
|
|
"epoch": 0.34400948991696323,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 6.8092,
|
|
"mean_token_accuracy": 0.08903076201677322,
|
|
"num_tokens": 831379.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.300806474685669,
|
|
"epoch": 0.3499406880189798,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 6.7387,
|
|
"mean_token_accuracy": 0.08791814595460892,
|
|
"num_tokens": 846596.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.254555749893188,
|
|
"epoch": 0.35587188612099646,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 6.7438,
|
|
"mean_token_accuracy": 0.08821789696812629,
|
|
"num_tokens": 862063.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.275745964050293,
|
|
"epoch": 0.36180308422301305,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.6954,
|
|
"mean_token_accuracy": 0.08966124951839446,
|
|
"num_tokens": 876841.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.221111488342285,
|
|
"epoch": 0.36773428232502964,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 6.7751,
|
|
"mean_token_accuracy": 0.09116524010896683,
|
|
"num_tokens": 890728.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.206338787078858,
|
|
"epoch": 0.3736654804270463,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 6.6956,
|
|
"mean_token_accuracy": 0.09070762842893601,
|
|
"num_tokens": 905599.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.1817542552948,
|
|
"epoch": 0.37959667852906287,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.6483,
|
|
"mean_token_accuracy": 0.09768397733569145,
|
|
"num_tokens": 918139.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.167359304428101,
|
|
"epoch": 0.38552787663107946,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.7114,
|
|
"mean_token_accuracy": 0.09328882694244385,
|
|
"num_tokens": 931666.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.196693468093872,
|
|
"epoch": 0.3914590747330961,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.6605,
|
|
"mean_token_accuracy": 0.09237670227885246,
|
|
"num_tokens": 946286.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.148892021179199,
|
|
"epoch": 0.3973902728351127,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.6029,
|
|
"mean_token_accuracy": 0.09802859723567962,
|
|
"num_tokens": 959689.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.091142225265503,
|
|
"epoch": 0.4033214709371293,
|
|
"grad_norm": 1.53125,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.6481,
|
|
"mean_token_accuracy": 0.09793345853686333,
|
|
"num_tokens": 972707.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.103200006484985,
|
|
"epoch": 0.4092526690391459,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.6694,
|
|
"mean_token_accuracy": 0.09691287726163864,
|
|
"num_tokens": 988010.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.109372472763061,
|
|
"epoch": 0.4151838671411625,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.7001,
|
|
"mean_token_accuracy": 0.0866947390139103,
|
|
"num_tokens": 1003823.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.018935489654541,
|
|
"epoch": 0.4211150652431791,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.4583,
|
|
"mean_token_accuracy": 0.10792901739478111,
|
|
"num_tokens": 1018276.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.151296997070313,
|
|
"epoch": 0.42704626334519574,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.6782,
|
|
"mean_token_accuracy": 0.09361589401960373,
|
|
"num_tokens": 1031242.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.0902214527130125,
|
|
"epoch": 0.43297746144721233,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.5786,
|
|
"mean_token_accuracy": 0.09998262599110604,
|
|
"num_tokens": 1043940.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 6.944676065444947,
|
|
"epoch": 0.4389086595492289,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.508,
|
|
"mean_token_accuracy": 0.09759898036718369,
|
|
"num_tokens": 1058656.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.117189264297485,
|
|
"epoch": 0.44483985765124556,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.6092,
|
|
"mean_token_accuracy": 0.09189791083335877,
|
|
"num_tokens": 1073128.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 6.948076343536377,
|
|
"epoch": 0.45077105575326215,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.5416,
|
|
"mean_token_accuracy": 0.09809970185160637,
|
|
"num_tokens": 1086467.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.076186227798462,
|
|
"epoch": 0.45670225385527874,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.6616,
|
|
"mean_token_accuracy": 0.10060962289571762,
|
|
"num_tokens": 1100356.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.028731060028076,
|
|
"epoch": 0.4626334519572954,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.5814,
|
|
"mean_token_accuracy": 0.10232764557003975,
|
|
"num_tokens": 1115877.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.04663872718811,
|
|
"epoch": 0.46856465005931197,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.5466,
|
|
"mean_token_accuracy": 0.10219399407505989,
|
|
"num_tokens": 1131415.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 6.8535503387451175,
|
|
"epoch": 0.4744958481613286,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.5137,
|
|
"mean_token_accuracy": 0.09734094068408013,
|
|
"num_tokens": 1145903.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 6.879596471786499,
|
|
"epoch": 0.4804270462633452,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.4397,
|
|
"mean_token_accuracy": 0.0997365266084671,
|
|
"num_tokens": 1161168.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 6.921687078475952,
|
|
"epoch": 0.4863582443653618,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.5196,
|
|
"mean_token_accuracy": 0.09855128526687622,
|
|
"num_tokens": 1178155.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 6.924295806884766,
|
|
"epoch": 0.49228944246737844,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.3697,
|
|
"mean_token_accuracy": 0.1099881686270237,
|
|
"num_tokens": 1190830.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 6.827713060379028,
|
|
"epoch": 0.498220640569395,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.3842,
|
|
"mean_token_accuracy": 0.10342236086726189,
|
|
"num_tokens": 1205260.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 6.883217096328735,
|
|
"epoch": 0.5041518386714117,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.431,
|
|
"mean_token_accuracy": 0.10836012288928032,
|
|
"num_tokens": 1218856.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 6.832102727890015,
|
|
"epoch": 0.5100830367734283,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.5014,
|
|
"mean_token_accuracy": 0.10428635627031327,
|
|
"num_tokens": 1234669.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 6.830685758590699,
|
|
"epoch": 0.5160142348754448,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.4289,
|
|
"mean_token_accuracy": 0.1057804524898529,
|
|
"num_tokens": 1249514.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 6.837182474136353,
|
|
"epoch": 0.5219454329774614,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.3638,
|
|
"mean_token_accuracy": 0.1109524667263031,
|
|
"num_tokens": 1262311.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 6.767211008071899,
|
|
"epoch": 0.527876631079478,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.3519,
|
|
"mean_token_accuracy": 0.10807539224624634,
|
|
"num_tokens": 1277681.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 6.781484460830688,
|
|
"epoch": 0.5338078291814946,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.3968,
|
|
"mean_token_accuracy": 0.10917404815554618,
|
|
"num_tokens": 1292300.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 6.775958061218262,
|
|
"epoch": 0.5397390272835113,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.4142,
|
|
"mean_token_accuracy": 0.10393262058496475,
|
|
"num_tokens": 1307426.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 6.787585258483887,
|
|
"epoch": 0.5456702253855279,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.4414,
|
|
"mean_token_accuracy": 0.09955934658646584,
|
|
"num_tokens": 1322167.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 6.75556001663208,
|
|
"epoch": 0.5516014234875445,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.3126,
|
|
"mean_token_accuracy": 0.11029193922877312,
|
|
"num_tokens": 1335343.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 6.719364500045776,
|
|
"epoch": 0.5575326215895611,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.3129,
|
|
"mean_token_accuracy": 0.11020243093371392,
|
|
"num_tokens": 1349078.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 6.670861291885376,
|
|
"epoch": 0.5634638196915777,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.2388,
|
|
"mean_token_accuracy": 0.11813571006059646,
|
|
"num_tokens": 1362030.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 6.790364742279053,
|
|
"epoch": 0.5693950177935944,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.3965,
|
|
"mean_token_accuracy": 0.10456062778830529,
|
|
"num_tokens": 1378303.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 6.773904657363891,
|
|
"epoch": 0.575326215895611,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.3906,
|
|
"mean_token_accuracy": 0.10592009201645851,
|
|
"num_tokens": 1392585.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 6.805166101455688,
|
|
"epoch": 0.5812574139976275,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.3973,
|
|
"mean_token_accuracy": 0.11140659600496292,
|
|
"num_tokens": 1406962.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 6.573492240905762,
|
|
"epoch": 0.5871886120996441,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.2853,
|
|
"mean_token_accuracy": 0.11129999980330467,
|
|
"num_tokens": 1419493.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 6.735859537124634,
|
|
"epoch": 0.5931198102016607,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.3007,
|
|
"mean_token_accuracy": 0.10790175497531891,
|
|
"num_tokens": 1435373.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.5931198102016607,
|
|
"eval_entropy": 6.67236985896994,
|
|
"eval_loss": 6.441715717315674,
|
|
"eval_mean_token_accuracy": 0.10966419813685988,
|
|
"eval_num_tokens": 1435373.0,
|
|
"eval_runtime": 2.4776,
|
|
"eval_samples_per_second": 1052.652,
|
|
"eval_steps_per_second": 131.581,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 8420,
|
|
"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": 2556432248832000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|