2119ed983f
Model: fpadovani/urd-arab-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": 0.5931198102016607,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.805317306518555,
|
|
"epoch": 0.005931198102016607,
|
|
"grad_norm": 16.25,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.4312,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 12647.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.825564956665039,
|
|
"epoch": 0.011862396204033215,
|
|
"grad_norm": 16.875,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.4169,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 26668.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.845508098602295,
|
|
"epoch": 0.017793594306049824,
|
|
"grad_norm": 24.75,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.1564,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 41021.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 5.1169312477111815,
|
|
"epoch": 0.02372479240806643,
|
|
"grad_norm": 33.75,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.5366,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 56626.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 7.628379774093628,
|
|
"epoch": 0.02965599051008304,
|
|
"grad_norm": 15.25,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 11.8077,
|
|
"mean_token_accuracy": 0.00010419512109365315,
|
|
"num_tokens": 72914.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.455063152313233,
|
|
"epoch": 0.03558718861209965,
|
|
"grad_norm": 3.21875,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.841,
|
|
"mean_token_accuracy": 0.0054948762990534306,
|
|
"num_tokens": 88514.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.646884059906006,
|
|
"epoch": 0.04151838671411625,
|
|
"grad_norm": 3.03125,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.5696,
|
|
"mean_token_accuracy": 0.01355562973767519,
|
|
"num_tokens": 101256.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.619774627685548,
|
|
"epoch": 0.04744958481613286,
|
|
"grad_norm": 2.4375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.282,
|
|
"mean_token_accuracy": 0.012707393150776625,
|
|
"num_tokens": 114459.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.61210651397705,
|
|
"epoch": 0.05338078291814947,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.0287,
|
|
"mean_token_accuracy": 0.01491355374455452,
|
|
"num_tokens": 129858.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.597074794769288,
|
|
"epoch": 0.05931198102016608,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.8185,
|
|
"mean_token_accuracy": 0.0331686332821846,
|
|
"num_tokens": 143486.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.579236507415771,
|
|
"epoch": 0.06524317912218268,
|
|
"grad_norm": 1.9609375,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.6912,
|
|
"mean_token_accuracy": 0.035220497474074366,
|
|
"num_tokens": 156003.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.577684020996093,
|
|
"epoch": 0.0711743772241993,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.6506,
|
|
"mean_token_accuracy": 0.034727090038359165,
|
|
"num_tokens": 170988.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.5711012840271,
|
|
"epoch": 0.0771055753262159,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.5328,
|
|
"mean_token_accuracy": 0.03448135014623403,
|
|
"num_tokens": 187165.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.558809852600097,
|
|
"epoch": 0.0830367734282325,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.4798,
|
|
"mean_token_accuracy": 0.03228826243430376,
|
|
"num_tokens": 202072.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.549299335479736,
|
|
"epoch": 0.08896797153024912,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.3281,
|
|
"mean_token_accuracy": 0.034324941039085385,
|
|
"num_tokens": 215375.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.532067012786865,
|
|
"epoch": 0.09489916963226572,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.2492,
|
|
"mean_token_accuracy": 0.03353528436273336,
|
|
"num_tokens": 228749.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.505972194671632,
|
|
"epoch": 0.10083036773428232,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.1494,
|
|
"mean_token_accuracy": 0.03314766474068165,
|
|
"num_tokens": 242387.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.478415584564209,
|
|
"epoch": 0.10676156583629894,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.0705,
|
|
"mean_token_accuracy": 0.032026905380189416,
|
|
"num_tokens": 257793.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.454620742797852,
|
|
"epoch": 0.11269276393831554,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.9712,
|
|
"mean_token_accuracy": 0.0325484286993742,
|
|
"num_tokens": 271187.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.40716381072998,
|
|
"epoch": 0.11862396204033215,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.8474,
|
|
"mean_token_accuracy": 0.03217122312635183,
|
|
"num_tokens": 284937.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.335246372222901,
|
|
"epoch": 0.12455516014234876,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.6893,
|
|
"mean_token_accuracy": 0.03255892116576433,
|
|
"num_tokens": 300108.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.246189594268799,
|
|
"epoch": 0.13048635824436536,
|
|
"grad_norm": 1.53125,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.5681,
|
|
"mean_token_accuracy": 0.031065495498478413,
|
|
"num_tokens": 316760.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.13212080001831,
|
|
"epoch": 0.13641755634638197,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.3819,
|
|
"mean_token_accuracy": 0.03049417305737734,
|
|
"num_tokens": 329705.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.97801332473755,
|
|
"epoch": 0.1423487544483986,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.244,
|
|
"mean_token_accuracy": 0.03179512321949005,
|
|
"num_tokens": 342947.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.771716022491455,
|
|
"epoch": 0.14827995255041518,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.0865,
|
|
"mean_token_accuracy": 0.03442887142300606,
|
|
"num_tokens": 358397.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.533180236816406,
|
|
"epoch": 0.1542111506524318,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.9643,
|
|
"mean_token_accuracy": 0.03413730468600988,
|
|
"num_tokens": 370369.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.256175804138184,
|
|
"epoch": 0.1601423487544484,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.8421,
|
|
"mean_token_accuracy": 0.03181219045072794,
|
|
"num_tokens": 386896.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 8.96011724472046,
|
|
"epoch": 0.166073546856465,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.6981,
|
|
"mean_token_accuracy": 0.03533113058656454,
|
|
"num_tokens": 400433.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.675125217437744,
|
|
"epoch": 0.17200474495848161,
|
|
"grad_norm": 0.7890625,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.6507,
|
|
"mean_token_accuracy": 0.030981262400746345,
|
|
"num_tokens": 415092.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.432178497314453,
|
|
"epoch": 0.17793594306049823,
|
|
"grad_norm": 0.7109375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.5489,
|
|
"mean_token_accuracy": 0.033769184909760955,
|
|
"num_tokens": 429793.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.235133266448974,
|
|
"epoch": 0.18386714116251482,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.5575,
|
|
"mean_token_accuracy": 0.03205921929329634,
|
|
"num_tokens": 442618.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.142054271697997,
|
|
"epoch": 0.18979833926453143,
|
|
"grad_norm": 0.75,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.4449,
|
|
"mean_token_accuracy": 0.03399460650980472,
|
|
"num_tokens": 454976.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.025918197631835,
|
|
"epoch": 0.19572953736654805,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.5037,
|
|
"mean_token_accuracy": 0.03331909943372011,
|
|
"num_tokens": 467523.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 7.9379983901977536,
|
|
"epoch": 0.20166073546856464,
|
|
"grad_norm": 0.6796875,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.4373,
|
|
"mean_token_accuracy": 0.03239843733608723,
|
|
"num_tokens": 481391.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 7.869046401977539,
|
|
"epoch": 0.20759193357058126,
|
|
"grad_norm": 0.67578125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.4399,
|
|
"mean_token_accuracy": 0.0311466746032238,
|
|
"num_tokens": 496697.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 7.866212034225464,
|
|
"epoch": 0.21352313167259787,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.4743,
|
|
"mean_token_accuracy": 0.027226756513118743,
|
|
"num_tokens": 510091.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 7.788731241226197,
|
|
"epoch": 0.21945432977461446,
|
|
"grad_norm": 0.703125,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.426,
|
|
"mean_token_accuracy": 0.03152833003550768,
|
|
"num_tokens": 522168.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 7.754296016693115,
|
|
"epoch": 0.22538552787663108,
|
|
"grad_norm": 0.71875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.3943,
|
|
"mean_token_accuracy": 0.03427231907844543,
|
|
"num_tokens": 535853.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 7.75240159034729,
|
|
"epoch": 0.2313167259786477,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.4255,
|
|
"mean_token_accuracy": 0.03329353462904692,
|
|
"num_tokens": 551082.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.700812101364136,
|
|
"epoch": 0.2372479240806643,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.4629,
|
|
"mean_token_accuracy": 0.035965221747756,
|
|
"num_tokens": 567181.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.696683025360107,
|
|
"epoch": 0.2431791221826809,
|
|
"grad_norm": 0.71484375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.4112,
|
|
"mean_token_accuracy": 0.035206197388470174,
|
|
"num_tokens": 580689.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.712547016143799,
|
|
"epoch": 0.2491103202846975,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.3345,
|
|
"mean_token_accuracy": 0.03781173955649138,
|
|
"num_tokens": 594590.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.769494247436524,
|
|
"epoch": 0.25504151838671413,
|
|
"grad_norm": 10.5,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.3905,
|
|
"mean_token_accuracy": 0.03799592889845371,
|
|
"num_tokens": 610577.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.6959569454193115,
|
|
"epoch": 0.2609727164887307,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.384,
|
|
"mean_token_accuracy": 0.037612661719322205,
|
|
"num_tokens": 627066.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.781623411178589,
|
|
"epoch": 0.2669039145907473,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.3672,
|
|
"mean_token_accuracy": 0.04785094037652016,
|
|
"num_tokens": 642601.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.77043514251709,
|
|
"epoch": 0.27283511269276395,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.2531,
|
|
"mean_token_accuracy": 0.05052996054291725,
|
|
"num_tokens": 655615.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.747775220870972,
|
|
"epoch": 0.27876631079478054,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.2647,
|
|
"mean_token_accuracy": 0.05221507623791695,
|
|
"num_tokens": 670218.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.743488121032715,
|
|
"epoch": 0.2846975088967972,
|
|
"grad_norm": 0.75390625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.2593,
|
|
"mean_token_accuracy": 0.05101203434169292,
|
|
"num_tokens": 683661.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.713605117797852,
|
|
"epoch": 0.29062870699881377,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.2135,
|
|
"mean_token_accuracy": 0.05287150628864765,
|
|
"num_tokens": 699767.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.695834493637085,
|
|
"epoch": 0.29655990510083036,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.1762,
|
|
"mean_token_accuracy": 0.05634650327265263,
|
|
"num_tokens": 713579.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.629615783691406,
|
|
"epoch": 0.302491103202847,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.1678,
|
|
"mean_token_accuracy": 0.05482946485280991,
|
|
"num_tokens": 727490.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.607522439956665,
|
|
"epoch": 0.3084223013048636,
|
|
"grad_norm": 0.80078125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.2348,
|
|
"mean_token_accuracy": 0.05509466864168644,
|
|
"num_tokens": 742477.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.624451637268066,
|
|
"epoch": 0.3143534994068802,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.113,
|
|
"mean_token_accuracy": 0.05751051269471645,
|
|
"num_tokens": 757772.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.647220134735107,
|
|
"epoch": 0.3202846975088968,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.1596,
|
|
"mean_token_accuracy": 0.05049732141196728,
|
|
"num_tokens": 772898.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.600799989700318,
|
|
"epoch": 0.3262158956109134,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.1154,
|
|
"mean_token_accuracy": 0.05433249212801457,
|
|
"num_tokens": 788395.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.563078927993774,
|
|
"epoch": 0.33214709371293,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.1648,
|
|
"mean_token_accuracy": 0.06050356179475784,
|
|
"num_tokens": 803413.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.517913770675659,
|
|
"epoch": 0.33807829181494664,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.0248,
|
|
"mean_token_accuracy": 0.06933637075126171,
|
|
"num_tokens": 817775.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.566510915756226,
|
|
"epoch": 0.34400948991696323,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.1421,
|
|
"mean_token_accuracy": 0.06264986135065556,
|
|
"num_tokens": 831379.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.551775217056274,
|
|
"epoch": 0.3499406880189798,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.0759,
|
|
"mean_token_accuracy": 0.05862759165465832,
|
|
"num_tokens": 846596.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.553944540023804,
|
|
"epoch": 0.35587188612099646,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.0666,
|
|
"mean_token_accuracy": 0.05912046544253826,
|
|
"num_tokens": 862063.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.582914543151856,
|
|
"epoch": 0.36180308422301305,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.9961,
|
|
"mean_token_accuracy": 0.06986662931740284,
|
|
"num_tokens": 876841.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.488122177124024,
|
|
"epoch": 0.36773428232502964,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.0603,
|
|
"mean_token_accuracy": 0.06794393062591553,
|
|
"num_tokens": 890728.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.442899990081787,
|
|
"epoch": 0.3736654804270463,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 6.983,
|
|
"mean_token_accuracy": 0.0702366728335619,
|
|
"num_tokens": 905599.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.451259899139404,
|
|
"epoch": 0.37959667852906287,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.9367,
|
|
"mean_token_accuracy": 0.07826149612665176,
|
|
"num_tokens": 918139.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.405993700027466,
|
|
"epoch": 0.38552787663107946,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.9943,
|
|
"mean_token_accuracy": 0.0733895305544138,
|
|
"num_tokens": 931666.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.446602392196655,
|
|
"epoch": 0.3914590747330961,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.9418,
|
|
"mean_token_accuracy": 0.0699479129165411,
|
|
"num_tokens": 946286.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.44321928024292,
|
|
"epoch": 0.3973902728351127,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.9048,
|
|
"mean_token_accuracy": 0.0711387813091278,
|
|
"num_tokens": 959689.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.383141279220581,
|
|
"epoch": 0.4033214709371293,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.9099,
|
|
"mean_token_accuracy": 0.07436129115521908,
|
|
"num_tokens": 972707.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.326433992385864,
|
|
"epoch": 0.4092526690391459,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.9298,
|
|
"mean_token_accuracy": 0.07282942235469818,
|
|
"num_tokens": 988010.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.3913586139678955,
|
|
"epoch": 0.4151838671411625,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.9494,
|
|
"mean_token_accuracy": 0.0702156439423561,
|
|
"num_tokens": 1003823.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.271407747268677,
|
|
"epoch": 0.4211150652431791,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.7257,
|
|
"mean_token_accuracy": 0.08785916194319725,
|
|
"num_tokens": 1018276.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.315984010696411,
|
|
"epoch": 0.42704626334519574,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.9177,
|
|
"mean_token_accuracy": 0.07590937204658985,
|
|
"num_tokens": 1031242.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.36967716217041,
|
|
"epoch": 0.43297746144721233,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.826,
|
|
"mean_token_accuracy": 0.08421060293912888,
|
|
"num_tokens": 1043940.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.209809446334839,
|
|
"epoch": 0.4389086595492289,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.7337,
|
|
"mean_token_accuracy": 0.08058795779943466,
|
|
"num_tokens": 1058656.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.326260042190552,
|
|
"epoch": 0.44483985765124556,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.8599,
|
|
"mean_token_accuracy": 0.07443175241351127,
|
|
"num_tokens": 1073128.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.216764450073242,
|
|
"epoch": 0.45077105575326215,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.7831,
|
|
"mean_token_accuracy": 0.08439998105168342,
|
|
"num_tokens": 1086467.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.274557256698609,
|
|
"epoch": 0.45670225385527874,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.9052,
|
|
"mean_token_accuracy": 0.08290659114718438,
|
|
"num_tokens": 1100356.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.288721132278442,
|
|
"epoch": 0.4626334519572954,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.8286,
|
|
"mean_token_accuracy": 0.08163000270724297,
|
|
"num_tokens": 1115877.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.271324348449707,
|
|
"epoch": 0.46856465005931197,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.7874,
|
|
"mean_token_accuracy": 0.08328207209706306,
|
|
"num_tokens": 1131415.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.159351873397827,
|
|
"epoch": 0.4744958481613286,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.7132,
|
|
"mean_token_accuracy": 0.08293095007538795,
|
|
"num_tokens": 1145903.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.103127574920654,
|
|
"epoch": 0.4804270462633452,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.67,
|
|
"mean_token_accuracy": 0.08018745929002762,
|
|
"num_tokens": 1161168.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.14245753288269,
|
|
"epoch": 0.4863582443653618,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.7324,
|
|
"mean_token_accuracy": 0.08255215212702752,
|
|
"num_tokens": 1178155.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.131292104721069,
|
|
"epoch": 0.49228944246737844,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.5728,
|
|
"mean_token_accuracy": 0.09345637336373329,
|
|
"num_tokens": 1190830.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.046381425857544,
|
|
"epoch": 0.498220640569395,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.5814,
|
|
"mean_token_accuracy": 0.08820187523961068,
|
|
"num_tokens": 1205260.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.083016061782837,
|
|
"epoch": 0.5041518386714117,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.6316,
|
|
"mean_token_accuracy": 0.09104111194610595,
|
|
"num_tokens": 1218856.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.034869623184204,
|
|
"epoch": 0.5100830367734283,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.6934,
|
|
"mean_token_accuracy": 0.09010339081287384,
|
|
"num_tokens": 1234669.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.032301235198974,
|
|
"epoch": 0.5160142348754448,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.6293,
|
|
"mean_token_accuracy": 0.09128119722008705,
|
|
"num_tokens": 1249514.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.032708358764649,
|
|
"epoch": 0.5219454329774614,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.5626,
|
|
"mean_token_accuracy": 0.0961005687713623,
|
|
"num_tokens": 1262311.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 6.9413275718688965,
|
|
"epoch": 0.527876631079478,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.5488,
|
|
"mean_token_accuracy": 0.09156981036067009,
|
|
"num_tokens": 1277681.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.003400182723999,
|
|
"epoch": 0.5338078291814946,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.5879,
|
|
"mean_token_accuracy": 0.09069180116057396,
|
|
"num_tokens": 1292300.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.015186309814453,
|
|
"epoch": 0.5397390272835113,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.595,
|
|
"mean_token_accuracy": 0.08999119028449058,
|
|
"num_tokens": 1307426.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 6.900823402404785,
|
|
"epoch": 0.5456702253855279,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.6162,
|
|
"mean_token_accuracy": 0.09043457433581352,
|
|
"num_tokens": 1322167.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 6.999922370910644,
|
|
"epoch": 0.5516014234875445,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.4976,
|
|
"mean_token_accuracy": 0.09610431045293807,
|
|
"num_tokens": 1335343.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 6.857682752609253,
|
|
"epoch": 0.5575326215895611,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.4976,
|
|
"mean_token_accuracy": 0.09826786443591118,
|
|
"num_tokens": 1349078.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 6.899160289764405,
|
|
"epoch": 0.5634638196915777,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.4224,
|
|
"mean_token_accuracy": 0.1049638457596302,
|
|
"num_tokens": 1362030.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 6.993446207046508,
|
|
"epoch": 0.5693950177935944,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.5763,
|
|
"mean_token_accuracy": 0.08985362276434898,
|
|
"num_tokens": 1378303.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 6.935125350952148,
|
|
"epoch": 0.575326215895611,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.5767,
|
|
"mean_token_accuracy": 0.09316942617297172,
|
|
"num_tokens": 1392585.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 6.9703733921051025,
|
|
"epoch": 0.5812574139976275,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.5814,
|
|
"mean_token_accuracy": 0.09292712658643723,
|
|
"num_tokens": 1406962.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 6.836954164505005,
|
|
"epoch": 0.5871886120996441,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.4605,
|
|
"mean_token_accuracy": 0.09848191663622856,
|
|
"num_tokens": 1419493.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 6.868243980407715,
|
|
"epoch": 0.5931198102016607,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.4801,
|
|
"mean_token_accuracy": 0.09710933044552802,
|
|
"num_tokens": 1435373.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.5931198102016607,
|
|
"eval_entropy": 6.868733419231111,
|
|
"eval_loss": 6.604183673858643,
|
|
"eval_mean_token_accuracy": 0.09870429635276458,
|
|
"eval_num_tokens": 1435373.0,
|
|
"eval_runtime": 2.5304,
|
|
"eval_samples_per_second": 1030.674,
|
|
"eval_steps_per_second": 128.834,
|
|
"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
|
|
}
|