1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.03269469692015955,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.742586898803712,
|
|
"epoch": 0.0003269469692015955,
|
|
"grad_norm": 4.25,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7625,
|
|
"mean_token_accuracy": 8.849557489156722e-05,
|
|
"num_tokens": 9940.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.74257230758667,
|
|
"epoch": 0.000653893938403191,
|
|
"grad_norm": 4.1875,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.75,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 19783.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742555141448975,
|
|
"epoch": 0.0009808409076047864,
|
|
"grad_norm": 4.625,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.7384,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 30189.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742575740814209,
|
|
"epoch": 0.001307787876806382,
|
|
"grad_norm": 4.1875,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.7067,
|
|
"mean_token_accuracy": 0.0007516760495491325,
|
|
"num_tokens": 40463.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742465591430664,
|
|
"epoch": 0.0016347348460079775,
|
|
"grad_norm": 4.03125,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.6408,
|
|
"mean_token_accuracy": 0.010044214146910235,
|
|
"num_tokens": 50919.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.742095851898194,
|
|
"epoch": 0.001961681815209573,
|
|
"grad_norm": 3.578125,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.5408,
|
|
"mean_token_accuracy": 0.027823284920305013,
|
|
"num_tokens": 61272.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.740856456756593,
|
|
"epoch": 0.0022886287844111684,
|
|
"grad_norm": 3.078125,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.4323,
|
|
"mean_token_accuracy": 0.03470553252846002,
|
|
"num_tokens": 70877.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.73651065826416,
|
|
"epoch": 0.002615575753612764,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.3205,
|
|
"mean_token_accuracy": 0.040680671110749245,
|
|
"num_tokens": 80600.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.727530670166015,
|
|
"epoch": 0.0029425227228143595,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.2476,
|
|
"mean_token_accuracy": 0.03953301161527634,
|
|
"num_tokens": 91002.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.717786026000976,
|
|
"epoch": 0.003269469692015955,
|
|
"grad_norm": 1.921875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.1594,
|
|
"mean_token_accuracy": 0.04255215115845203,
|
|
"num_tokens": 101156.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.710540103912354,
|
|
"epoch": 0.0035964166612175506,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.1182,
|
|
"mean_token_accuracy": 0.04170279260724783,
|
|
"num_tokens": 111896.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.705566215515137,
|
|
"epoch": 0.003923363630419146,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.9881,
|
|
"mean_token_accuracy": 0.04833919741213322,
|
|
"num_tokens": 120899.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.698173141479492,
|
|
"epoch": 0.004250310599620741,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 10.0328,
|
|
"mean_token_accuracy": 0.04022225625813007,
|
|
"num_tokens": 131504.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.691369152069091,
|
|
"epoch": 0.004577257568822337,
|
|
"grad_norm": 1.7890625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.9577,
|
|
"mean_token_accuracy": 0.04434355236589908,
|
|
"num_tokens": 142365.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.686608409881591,
|
|
"epoch": 0.004904204538023932,
|
|
"grad_norm": 1.5625,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.9256,
|
|
"mean_token_accuracy": 0.041293222457170486,
|
|
"num_tokens": 153481.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.679593467712403,
|
|
"epoch": 0.005231151507225528,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.8544,
|
|
"mean_token_accuracy": 0.041375156864523886,
|
|
"num_tokens": 163941.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.669306468963622,
|
|
"epoch": 0.0055580984764271235,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.7936,
|
|
"mean_token_accuracy": 0.04275329895317555,
|
|
"num_tokens": 173992.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.655621910095215,
|
|
"epoch": 0.005885045445628719,
|
|
"grad_norm": 1.515625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.7624,
|
|
"mean_token_accuracy": 0.0399011442437768,
|
|
"num_tokens": 184474.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.642310905456544,
|
|
"epoch": 0.006211992414830315,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.7029,
|
|
"mean_token_accuracy": 0.0412175253033638,
|
|
"num_tokens": 194998.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.629393005371094,
|
|
"epoch": 0.00653893938403191,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.5401,
|
|
"mean_token_accuracy": 0.05521246679127216,
|
|
"num_tokens": 203327.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.600548839569091,
|
|
"epoch": 0.006865886353233506,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.5267,
|
|
"mean_token_accuracy": 0.05421698726713657,
|
|
"num_tokens": 213921.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.55351276397705,
|
|
"epoch": 0.007192833322435101,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.4417,
|
|
"mean_token_accuracy": 0.05835340581834316,
|
|
"num_tokens": 225790.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.493978023529053,
|
|
"epoch": 0.007519780291636697,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.3301,
|
|
"mean_token_accuracy": 0.056574657931923865,
|
|
"num_tokens": 235689.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.437990379333495,
|
|
"epoch": 0.007846727260838291,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 9.2562,
|
|
"mean_token_accuracy": 0.054686100035905835,
|
|
"num_tokens": 246525.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.380927467346192,
|
|
"epoch": 0.008173674230039887,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 9.1809,
|
|
"mean_token_accuracy": 0.056523428112268445,
|
|
"num_tokens": 256642.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.325669860839843,
|
|
"epoch": 0.008500621199241483,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 9.1399,
|
|
"mean_token_accuracy": 0.052699505165219304,
|
|
"num_tokens": 268145.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.2711612701416,
|
|
"epoch": 0.008827568168443078,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 9.02,
|
|
"mean_token_accuracy": 0.050835156068205836,
|
|
"num_tokens": 279783.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.188286304473877,
|
|
"epoch": 0.009154515137644674,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.9063,
|
|
"mean_token_accuracy": 0.05600942634046078,
|
|
"num_tokens": 289723.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.073064708709717,
|
|
"epoch": 0.00948146210684627,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.8015,
|
|
"mean_token_accuracy": 0.05663799084722996,
|
|
"num_tokens": 300002.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.98648281097412,
|
|
"epoch": 0.009808409076047865,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.777,
|
|
"mean_token_accuracy": 0.05806488022208214,
|
|
"num_tokens": 310041.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.891202545166015,
|
|
"epoch": 0.01013535604524946,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.6997,
|
|
"mean_token_accuracy": 0.05627730451524258,
|
|
"num_tokens": 321469.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.748745250701905,
|
|
"epoch": 0.010462303014451056,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.5702,
|
|
"mean_token_accuracy": 0.05877697579562664,
|
|
"num_tokens": 330796.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.617728519439698,
|
|
"epoch": 0.010789249983652651,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.5724,
|
|
"mean_token_accuracy": 0.05753467902541161,
|
|
"num_tokens": 341402.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.483910274505615,
|
|
"epoch": 0.011116196952854247,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.5709,
|
|
"mean_token_accuracy": 0.058061816915869716,
|
|
"num_tokens": 351717.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.382616806030274,
|
|
"epoch": 0.011443143922055843,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.4663,
|
|
"mean_token_accuracy": 0.057271933555603026,
|
|
"num_tokens": 361787.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 9.258338737487794,
|
|
"epoch": 0.011770090891257438,
|
|
"grad_norm": 0.82421875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.4114,
|
|
"mean_token_accuracy": 0.05387798622250557,
|
|
"num_tokens": 371827.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 9.150267505645752,
|
|
"epoch": 0.012097037860459034,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.4122,
|
|
"mean_token_accuracy": 0.06042315401136875,
|
|
"num_tokens": 382150.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 9.016548824310302,
|
|
"epoch": 0.01242398482966063,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.3316,
|
|
"mean_token_accuracy": 0.06260293386876584,
|
|
"num_tokens": 392151.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.979074478149414,
|
|
"epoch": 0.012750931798862225,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.3421,
|
|
"mean_token_accuracy": 0.057670443505048755,
|
|
"num_tokens": 402779.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.890014362335204,
|
|
"epoch": 0.01307787876806382,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.3466,
|
|
"mean_token_accuracy": 0.06248428300023079,
|
|
"num_tokens": 413500.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.904778671264648,
|
|
"epoch": 0.013404825737265416,
|
|
"grad_norm": 0.8984375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.2835,
|
|
"mean_token_accuracy": 0.06201527528464794,
|
|
"num_tokens": 423054.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.856226634979247,
|
|
"epoch": 0.013731772706467011,
|
|
"grad_norm": 0.8203125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.3133,
|
|
"mean_token_accuracy": 0.05689386166632175,
|
|
"num_tokens": 433002.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.779247093200684,
|
|
"epoch": 0.014058719675668607,
|
|
"grad_norm": 0.72265625,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.295,
|
|
"mean_token_accuracy": 0.05818233527243137,
|
|
"num_tokens": 443492.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.777221393585204,
|
|
"epoch": 0.014385666644870202,
|
|
"grad_norm": 0.796875,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.3091,
|
|
"mean_token_accuracy": 0.062193362042307854,
|
|
"num_tokens": 453001.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.772535514831542,
|
|
"epoch": 0.014712613614071798,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.2306,
|
|
"mean_token_accuracy": 0.06282332614064216,
|
|
"num_tokens": 462916.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.702757263183594,
|
|
"epoch": 0.015039560583273394,
|
|
"grad_norm": 0.75390625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.2694,
|
|
"mean_token_accuracy": 0.060815002769231796,
|
|
"num_tokens": 473549.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.745165348052979,
|
|
"epoch": 0.015366507552474989,
|
|
"grad_norm": 0.8046875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.2862,
|
|
"mean_token_accuracy": 0.05802113525569439,
|
|
"num_tokens": 484472.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.76063575744629,
|
|
"epoch": 0.015693454521676583,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.2251,
|
|
"mean_token_accuracy": 0.06674037426710129,
|
|
"num_tokens": 494202.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.625239944458007,
|
|
"epoch": 0.01602040149087818,
|
|
"grad_norm": 0.78515625,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.1602,
|
|
"mean_token_accuracy": 0.0650412030518055,
|
|
"num_tokens": 504635.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.66549301147461,
|
|
"epoch": 0.016347348460079774,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.1881,
|
|
"mean_token_accuracy": 0.0613247200846672,
|
|
"num_tokens": 514423.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.592597770690919,
|
|
"epoch": 0.01667429542928137,
|
|
"grad_norm": 0.82421875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.2256,
|
|
"mean_token_accuracy": 0.06521738171577454,
|
|
"num_tokens": 525820.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.587825107574464,
|
|
"epoch": 0.017001242398482965,
|
|
"grad_norm": 0.7734375,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.1931,
|
|
"mean_token_accuracy": 0.06806896105408669,
|
|
"num_tokens": 536732.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.599881744384765,
|
|
"epoch": 0.01732818936768456,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.1843,
|
|
"mean_token_accuracy": 0.06989322565495967,
|
|
"num_tokens": 546230.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.689265823364257,
|
|
"epoch": 0.017655136336886156,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.1967,
|
|
"mean_token_accuracy": 0.0643424428999424,
|
|
"num_tokens": 556078.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.557295227050782,
|
|
"epoch": 0.017982083306087752,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.178,
|
|
"mean_token_accuracy": 0.06810705624520778,
|
|
"num_tokens": 565385.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.595024585723877,
|
|
"epoch": 0.018309030275289347,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.237,
|
|
"mean_token_accuracy": 0.0639534905552864,
|
|
"num_tokens": 576066.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.57234001159668,
|
|
"epoch": 0.018635977244490943,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.1585,
|
|
"mean_token_accuracy": 0.0638256598263979,
|
|
"num_tokens": 587472.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.60384578704834,
|
|
"epoch": 0.01896292421369254,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.1544,
|
|
"mean_token_accuracy": 0.06674780026078224,
|
|
"num_tokens": 597400.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.587763404846191,
|
|
"epoch": 0.019289871182894134,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.2128,
|
|
"mean_token_accuracy": 0.06767871156334877,
|
|
"num_tokens": 606856.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.573587322235108,
|
|
"epoch": 0.01961681815209573,
|
|
"grad_norm": 0.83984375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.1472,
|
|
"mean_token_accuracy": 0.07218570187687874,
|
|
"num_tokens": 617209.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.5530442237854,
|
|
"epoch": 0.019943765121297325,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.153,
|
|
"mean_token_accuracy": 0.07159969210624695,
|
|
"num_tokens": 626826.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.517343521118164,
|
|
"epoch": 0.02027071209049892,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.1166,
|
|
"mean_token_accuracy": 0.07488984875380993,
|
|
"num_tokens": 636836.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.550188350677491,
|
|
"epoch": 0.020597659059700516,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.2061,
|
|
"mean_token_accuracy": 0.06659409068524838,
|
|
"num_tokens": 648177.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.58387565612793,
|
|
"epoch": 0.02092460602890211,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.1801,
|
|
"mean_token_accuracy": 0.06891246885061264,
|
|
"num_tokens": 658365.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.48437042236328,
|
|
"epoch": 0.021251552998103707,
|
|
"grad_norm": 1.7890625,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.0879,
|
|
"mean_token_accuracy": 0.06931497268378735,
|
|
"num_tokens": 668980.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.547167491912841,
|
|
"epoch": 0.021578499967305303,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.1331,
|
|
"mean_token_accuracy": 0.07338778600096703,
|
|
"num_tokens": 679406.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.523040771484375,
|
|
"epoch": 0.0219054469365069,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 8.0915,
|
|
"mean_token_accuracy": 0.07461029291152954,
|
|
"num_tokens": 689871.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.489326858520508,
|
|
"epoch": 0.022232393905708494,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.0457,
|
|
"mean_token_accuracy": 0.07580887749791146,
|
|
"num_tokens": 700969.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.469206142425538,
|
|
"epoch": 0.02255934087491009,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 8.0714,
|
|
"mean_token_accuracy": 0.07064862996339798,
|
|
"num_tokens": 711153.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.425196266174316,
|
|
"epoch": 0.022886287844111685,
|
|
"grad_norm": 0.83984375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 8.0605,
|
|
"mean_token_accuracy": 0.07229152098298072,
|
|
"num_tokens": 721475.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.458955097198487,
|
|
"epoch": 0.02321323481331328,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.000177,
|
|
"loss": 8.1029,
|
|
"mean_token_accuracy": 0.07589908726513386,
|
|
"num_tokens": 732823.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.45833330154419,
|
|
"epoch": 0.023540181782514876,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 8.0594,
|
|
"mean_token_accuracy": 0.07507128641009331,
|
|
"num_tokens": 743853.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.486702823638916,
|
|
"epoch": 0.02386712875171647,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.1181,
|
|
"mean_token_accuracy": 0.07701903395354748,
|
|
"num_tokens": 754579.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.394766998291015,
|
|
"epoch": 0.024194075720918067,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 8.0894,
|
|
"mean_token_accuracy": 0.07040258683264256,
|
|
"num_tokens": 764836.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.46137580871582,
|
|
"epoch": 0.024521022690119663,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 8.0219,
|
|
"mean_token_accuracy": 0.07322987243533134,
|
|
"num_tokens": 773961.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.43176326751709,
|
|
"epoch": 0.02484796965932126,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.9916,
|
|
"mean_token_accuracy": 0.0786767140030861,
|
|
"num_tokens": 784331.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.325694942474366,
|
|
"epoch": 0.025174916628522854,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.9694,
|
|
"mean_token_accuracy": 0.07703853137791157,
|
|
"num_tokens": 794724.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.46515941619873,
|
|
"epoch": 0.02550186359772445,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.984,
|
|
"mean_token_accuracy": 0.0744760014116764,
|
|
"num_tokens": 804680.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.296853351593018,
|
|
"epoch": 0.025828810566926045,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.9805,
|
|
"mean_token_accuracy": 0.07433952577412128,
|
|
"num_tokens": 815066.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.427121257781982,
|
|
"epoch": 0.02615575753612764,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.9862,
|
|
"mean_token_accuracy": 0.07774959728121758,
|
|
"num_tokens": 824958.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.403383159637452,
|
|
"epoch": 0.026482704505329236,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.000202,
|
|
"loss": 8.0499,
|
|
"mean_token_accuracy": 0.07819209098815919,
|
|
"num_tokens": 835128.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.332460689544678,
|
|
"epoch": 0.02680965147453083,
|
|
"grad_norm": 0.78515625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.9763,
|
|
"mean_token_accuracy": 0.07198684960603714,
|
|
"num_tokens": 844590.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.386160278320313,
|
|
"epoch": 0.027136598443732427,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.9597,
|
|
"mean_token_accuracy": 0.07920237854123116,
|
|
"num_tokens": 854914.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.396887683868409,
|
|
"epoch": 0.027463545412934023,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.9929,
|
|
"mean_token_accuracy": 0.0751864355057478,
|
|
"num_tokens": 866171.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.301899719238282,
|
|
"epoch": 0.027790492382135618,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.9806,
|
|
"mean_token_accuracy": 0.07633630074560642,
|
|
"num_tokens": 876603.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.322303676605225,
|
|
"epoch": 0.028117439351337214,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.913,
|
|
"mean_token_accuracy": 0.07756362929940223,
|
|
"num_tokens": 887124.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.30418529510498,
|
|
"epoch": 0.02844438632053881,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.9104,
|
|
"mean_token_accuracy": 0.08119071945548058,
|
|
"num_tokens": 897275.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.311804008483886,
|
|
"epoch": 0.028771333289740405,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.8925,
|
|
"mean_token_accuracy": 0.08083942905068398,
|
|
"num_tokens": 907174.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.347180366516113,
|
|
"epoch": 0.029098280258942,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.8892,
|
|
"mean_token_accuracy": 0.08076051622629166,
|
|
"num_tokens": 917126.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.24964075088501,
|
|
"epoch": 0.029425227228143596,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.8899,
|
|
"mean_token_accuracy": 0.08073717057704925,
|
|
"num_tokens": 926758.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.3399188041687,
|
|
"epoch": 0.02975217419734519,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.8954,
|
|
"mean_token_accuracy": 0.08044968619942665,
|
|
"num_tokens": 936973.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.194979095458985,
|
|
"epoch": 0.030079121166546787,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.8588,
|
|
"mean_token_accuracy": 0.07642363831400871,
|
|
"num_tokens": 946845.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.322082328796387,
|
|
"epoch": 0.030406068135748383,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.8846,
|
|
"mean_token_accuracy": 0.08190244510769844,
|
|
"num_tokens": 956904.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 8.182524299621582,
|
|
"epoch": 0.030733015104949978,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.8873,
|
|
"mean_token_accuracy": 0.07898930385708809,
|
|
"num_tokens": 966133.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.294102954864503,
|
|
"epoch": 0.031059962074151574,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.926,
|
|
"mean_token_accuracy": 0.07419689372181892,
|
|
"num_tokens": 976889.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.3031400680542,
|
|
"epoch": 0.031386909043353166,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.9145,
|
|
"mean_token_accuracy": 0.081867391243577,
|
|
"num_tokens": 987385.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.210045433044433,
|
|
"epoch": 0.03171385601255476,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.885,
|
|
"mean_token_accuracy": 0.07997245490550994,
|
|
"num_tokens": 998116.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.263574028015137,
|
|
"epoch": 0.03204080298175636,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.8633,
|
|
"mean_token_accuracy": 0.0772053673863411,
|
|
"num_tokens": 1007412.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.196787071228027,
|
|
"epoch": 0.03236774995095795,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.8959,
|
|
"mean_token_accuracy": 0.07670100852847099,
|
|
"num_tokens": 1016843.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.263703727722168,
|
|
"epoch": 0.03269469692015955,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.8932,
|
|
"mean_token_accuracy": 0.0816141352057457,
|
|
"num_tokens": 1027285.0,
|
|
"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": 239505405050880.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|