ec2fe396ab
Model: fpadovani/tur-latn-10mb-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": 0.4972650422675286,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.810139894485474,
|
|
"epoch": 0.004972650422675286,
|
|
"grad_norm": 21.625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.2169,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 9089.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.796057510375976,
|
|
"epoch": 0.009945300845350571,
|
|
"grad_norm": 22.0,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.3796,
|
|
"mean_token_accuracy": 0.0001728883828036487,
|
|
"num_tokens": 21650.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.851445960998535,
|
|
"epoch": 0.014917951268025857,
|
|
"grad_norm": 24.75,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.047,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 31067.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 4.946325588226318,
|
|
"epoch": 0.019890601690701143,
|
|
"grad_norm": 32.0,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.8176,
|
|
"mean_token_accuracy": 0.0005868025938980282,
|
|
"num_tokens": 40816.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 5.410843563079834,
|
|
"epoch": 0.02486325211337643,
|
|
"grad_norm": 40.25,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 13.0547,
|
|
"mean_token_accuracy": 0.0003202892781700939,
|
|
"num_tokens": 51065.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 7.1249548435211185,
|
|
"epoch": 0.029835902536051714,
|
|
"grad_norm": 30.125,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 11.8315,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 60539.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.19781551361084,
|
|
"epoch": 0.034808552958727,
|
|
"grad_norm": 3.3125,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.8532,
|
|
"mean_token_accuracy": 0.0021006516821216793,
|
|
"num_tokens": 71010.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.736362266540528,
|
|
"epoch": 0.039781203381402286,
|
|
"grad_norm": 3.125,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.6601,
|
|
"mean_token_accuracy": 0.012861686293035746,
|
|
"num_tokens": 81394.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.72396240234375,
|
|
"epoch": 0.04475385380407757,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.5728,
|
|
"mean_token_accuracy": 0.015258754882961512,
|
|
"num_tokens": 92491.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.611907863616944,
|
|
"epoch": 0.04972650422675286,
|
|
"grad_norm": 2.546875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.3994,
|
|
"mean_token_accuracy": 0.01761980978772044,
|
|
"num_tokens": 102113.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.650904369354247,
|
|
"epoch": 0.05469915464942814,
|
|
"grad_norm": 2.578125,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.3041,
|
|
"mean_token_accuracy": 0.01719568707048893,
|
|
"num_tokens": 112211.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.714541339874268,
|
|
"epoch": 0.05967180507210343,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 10.1449,
|
|
"mean_token_accuracy": 0.02930256463587284,
|
|
"num_tokens": 122166.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.710609722137452,
|
|
"epoch": 0.06464445549477872,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 10.0419,
|
|
"mean_token_accuracy": 0.04249542821198702,
|
|
"num_tokens": 131016.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.672384357452392,
|
|
"epoch": 0.069617105917454,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.9842,
|
|
"mean_token_accuracy": 0.03610398881137371,
|
|
"num_tokens": 141369.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.683287620544434,
|
|
"epoch": 0.07458975634012929,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.9433,
|
|
"mean_token_accuracy": 0.051322261616587636,
|
|
"num_tokens": 150939.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.683112144470215,
|
|
"epoch": 0.07956240676280457,
|
|
"grad_norm": 1.5546875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.9042,
|
|
"mean_token_accuracy": 0.050704098865389825,
|
|
"num_tokens": 159981.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.666728496551514,
|
|
"epoch": 0.08453505718547986,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.8301,
|
|
"mean_token_accuracy": 0.0480174720287323,
|
|
"num_tokens": 170139.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.6644793510437,
|
|
"epoch": 0.08950770760815514,
|
|
"grad_norm": 1.7265625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.767,
|
|
"mean_token_accuracy": 0.05141008589416742,
|
|
"num_tokens": 180746.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.65702543258667,
|
|
"epoch": 0.09448035803083044,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.6469,
|
|
"mean_token_accuracy": 0.057331408560276034,
|
|
"num_tokens": 190250.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.622721958160401,
|
|
"epoch": 0.09945300845350571,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.6532,
|
|
"mean_token_accuracy": 0.04415315557271242,
|
|
"num_tokens": 201185.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.592253017425538,
|
|
"epoch": 0.10442565887618101,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.5616,
|
|
"mean_token_accuracy": 0.05255695134401321,
|
|
"num_tokens": 211476.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.56272497177124,
|
|
"epoch": 0.10939830929885629,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.4667,
|
|
"mean_token_accuracy": 0.05462103560566902,
|
|
"num_tokens": 220965.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.503788185119628,
|
|
"epoch": 0.11437095972153158,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.4022,
|
|
"mean_token_accuracy": 0.04950843863189221,
|
|
"num_tokens": 231539.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.478209018707275,
|
|
"epoch": 0.11934361014420686,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 9.3051,
|
|
"mean_token_accuracy": 0.05306029766798019,
|
|
"num_tokens": 241206.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.38660650253296,
|
|
"epoch": 0.12431626056688215,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 9.2202,
|
|
"mean_token_accuracy": 0.05398030839860439,
|
|
"num_tokens": 250629.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.396561527252198,
|
|
"epoch": 0.12928891098955744,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 9.1838,
|
|
"mean_token_accuracy": 0.0500319853425026,
|
|
"num_tokens": 260705.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.283761501312256,
|
|
"epoch": 0.13426156141223272,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 9.1281,
|
|
"mean_token_accuracy": 0.0536313846707344,
|
|
"num_tokens": 270499.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.26081371307373,
|
|
"epoch": 0.139234211834908,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 9.0683,
|
|
"mean_token_accuracy": 0.05152654591947794,
|
|
"num_tokens": 279998.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.157192802429199,
|
|
"epoch": 0.14420686225758328,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.9399,
|
|
"mean_token_accuracy": 0.05353470556437969,
|
|
"num_tokens": 290322.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.07140588760376,
|
|
"epoch": 0.14917951268025859,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.858,
|
|
"mean_token_accuracy": 0.059231129661202434,
|
|
"num_tokens": 299893.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.94995050430298,
|
|
"epoch": 0.15415216310293386,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.7989,
|
|
"mean_token_accuracy": 0.05670776516199112,
|
|
"num_tokens": 309191.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.815523719787597,
|
|
"epoch": 0.15912481352560914,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.7036,
|
|
"mean_token_accuracy": 0.057601967453956605,
|
|
"num_tokens": 318494.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.753715515136719,
|
|
"epoch": 0.16409746394828442,
|
|
"grad_norm": 0.83984375,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.6798,
|
|
"mean_token_accuracy": 0.05443971864879131,
|
|
"num_tokens": 328299.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.610096073150634,
|
|
"epoch": 0.16907011437095973,
|
|
"grad_norm": 0.7578125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.6986,
|
|
"mean_token_accuracy": 0.05396176092326641,
|
|
"num_tokens": 338878.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.495657157897949,
|
|
"epoch": 0.174042764793635,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.5977,
|
|
"mean_token_accuracy": 0.05288307964801788,
|
|
"num_tokens": 348508.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 9.364479351043702,
|
|
"epoch": 0.1790154152163103,
|
|
"grad_norm": 0.75,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.5384,
|
|
"mean_token_accuracy": 0.060325421020388605,
|
|
"num_tokens": 358233.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 9.208246898651122,
|
|
"epoch": 0.1839880656389856,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.4774,
|
|
"mean_token_accuracy": 0.05886298380792141,
|
|
"num_tokens": 367297.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 9.103068447113037,
|
|
"epoch": 0.18896071606166087,
|
|
"grad_norm": 0.671875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.471,
|
|
"mean_token_accuracy": 0.052080392837524414,
|
|
"num_tokens": 376880.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 9.05744981765747,
|
|
"epoch": 0.19393336648433615,
|
|
"grad_norm": 0.8359375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.4617,
|
|
"mean_token_accuracy": 0.05685535296797752,
|
|
"num_tokens": 386806.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.958063793182372,
|
|
"epoch": 0.19890601690701143,
|
|
"grad_norm": 0.8984375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.4048,
|
|
"mean_token_accuracy": 0.059261714294552806,
|
|
"num_tokens": 396161.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.898060321807861,
|
|
"epoch": 0.20387866732968674,
|
|
"grad_norm": 0.8046875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.4628,
|
|
"mean_token_accuracy": 0.057381686940789225,
|
|
"num_tokens": 405643.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.85878095626831,
|
|
"epoch": 0.20885131775236201,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.4071,
|
|
"mean_token_accuracy": 0.0663915179669857,
|
|
"num_tokens": 414367.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.785406398773194,
|
|
"epoch": 0.2138239681750373,
|
|
"grad_norm": 0.734375,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.4544,
|
|
"mean_token_accuracy": 0.05882962830364704,
|
|
"num_tokens": 424430.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.815994358062744,
|
|
"epoch": 0.21879661859771257,
|
|
"grad_norm": 0.83203125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.4285,
|
|
"mean_token_accuracy": 0.06256655529141426,
|
|
"num_tokens": 435176.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.771424484252929,
|
|
"epoch": 0.22376926902038788,
|
|
"grad_norm": 0.734375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.4408,
|
|
"mean_token_accuracy": 0.060541344434022905,
|
|
"num_tokens": 445611.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.745358943939209,
|
|
"epoch": 0.22874191944306316,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.3751,
|
|
"mean_token_accuracy": 0.0634678304195404,
|
|
"num_tokens": 455143.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.766716575622558,
|
|
"epoch": 0.23371456986573844,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.4738,
|
|
"mean_token_accuracy": 0.06072634160518646,
|
|
"num_tokens": 465507.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.770550727844238,
|
|
"epoch": 0.23868722028841372,
|
|
"grad_norm": 0.796875,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.4583,
|
|
"mean_token_accuracy": 0.06104334816336632,
|
|
"num_tokens": 476152.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.77504940032959,
|
|
"epoch": 0.24365987071108902,
|
|
"grad_norm": 0.7890625,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.3132,
|
|
"mean_token_accuracy": 0.06499341167509556,
|
|
"num_tokens": 486341.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.655905055999757,
|
|
"epoch": 0.2486325211337643,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.2634,
|
|
"mean_token_accuracy": 0.06762947700917721,
|
|
"num_tokens": 495915.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.659723567962647,
|
|
"epoch": 0.2536051715564396,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.3514,
|
|
"mean_token_accuracy": 0.06988421343266964,
|
|
"num_tokens": 505666.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.692667961120605,
|
|
"epoch": 0.2585778219791149,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.3023,
|
|
"mean_token_accuracy": 0.0659954596310854,
|
|
"num_tokens": 515411.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.63499994277954,
|
|
"epoch": 0.26355047240179014,
|
|
"grad_norm": 0.78125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.3494,
|
|
"mean_token_accuracy": 0.06087326742708683,
|
|
"num_tokens": 525018.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.6294114112854,
|
|
"epoch": 0.26852312282446544,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.3103,
|
|
"mean_token_accuracy": 0.06277070567011833,
|
|
"num_tokens": 535056.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.69818925857544,
|
|
"epoch": 0.27349577324714075,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.309,
|
|
"mean_token_accuracy": 0.06765819452703,
|
|
"num_tokens": 545115.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.642760372161865,
|
|
"epoch": 0.278468423669816,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.3213,
|
|
"mean_token_accuracy": 0.06374539136886596,
|
|
"num_tokens": 554429.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.534232425689698,
|
|
"epoch": 0.2834410740924913,
|
|
"grad_norm": 0.6640625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.2617,
|
|
"mean_token_accuracy": 0.0756473757326603,
|
|
"num_tokens": 563357.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.696459770202637,
|
|
"epoch": 0.28841372451516656,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.3499,
|
|
"mean_token_accuracy": 0.0666810642927885,
|
|
"num_tokens": 574480.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.571897983551025,
|
|
"epoch": 0.29338637493784187,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.2731,
|
|
"mean_token_accuracy": 0.06942101046442986,
|
|
"num_tokens": 583722.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.63041067123413,
|
|
"epoch": 0.29835902536051717,
|
|
"grad_norm": 0.66015625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.2406,
|
|
"mean_token_accuracy": 0.06790350154042243,
|
|
"num_tokens": 593644.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.592833423614502,
|
|
"epoch": 0.3033316757831924,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.2753,
|
|
"mean_token_accuracy": 0.06820234544575214,
|
|
"num_tokens": 603547.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.531950759887696,
|
|
"epoch": 0.30830432620586773,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.3059,
|
|
"mean_token_accuracy": 0.06576361991465092,
|
|
"num_tokens": 613880.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.58628578186035,
|
|
"epoch": 0.31327697662854304,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.288,
|
|
"mean_token_accuracy": 0.06338743269443511,
|
|
"num_tokens": 624213.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.631004810333252,
|
|
"epoch": 0.3182496270512183,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.3069,
|
|
"mean_token_accuracy": 0.06901536360383034,
|
|
"num_tokens": 634350.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.613157272338867,
|
|
"epoch": 0.3232222774738936,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.3072,
|
|
"mean_token_accuracy": 0.06461950577795506,
|
|
"num_tokens": 644040.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.611722564697265,
|
|
"epoch": 0.32819492789656884,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.2689,
|
|
"mean_token_accuracy": 0.070689857006073,
|
|
"num_tokens": 652915.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.554406261444091,
|
|
"epoch": 0.33316757831924415,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 8.2508,
|
|
"mean_token_accuracy": 0.06666112951934337,
|
|
"num_tokens": 662665.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.609964942932129,
|
|
"epoch": 0.33814022874191946,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.2467,
|
|
"mean_token_accuracy": 0.07032966800034046,
|
|
"num_tokens": 671667.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.529174995422363,
|
|
"epoch": 0.3431128791645947,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 8.2799,
|
|
"mean_token_accuracy": 0.06516398154199124,
|
|
"num_tokens": 680432.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.62266035079956,
|
|
"epoch": 0.34808552958727,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 8.2702,
|
|
"mean_token_accuracy": 0.06741996183991432,
|
|
"num_tokens": 690184.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.603531074523925,
|
|
"epoch": 0.3530581800099453,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 8.159,
|
|
"mean_token_accuracy": 0.07316870428621769,
|
|
"num_tokens": 699147.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.518417072296142,
|
|
"epoch": 0.3580308304326206,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 8.1926,
|
|
"mean_token_accuracy": 0.0727127481251955,
|
|
"num_tokens": 708696.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.589232635498046,
|
|
"epoch": 0.3630034808552959,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.1562,
|
|
"mean_token_accuracy": 0.06915070340037346,
|
|
"num_tokens": 718438.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.530152130126954,
|
|
"epoch": 0.3679761312779712,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 8.2959,
|
|
"mean_token_accuracy": 0.06483390927314758,
|
|
"num_tokens": 728157.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.543238544464112,
|
|
"epoch": 0.37294878170064644,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 8.1901,
|
|
"mean_token_accuracy": 0.07307061739265919,
|
|
"num_tokens": 737007.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.549563598632812,
|
|
"epoch": 0.37792143212332174,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 8.1671,
|
|
"mean_token_accuracy": 0.07229796946048736,
|
|
"num_tokens": 746873.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.522933387756348,
|
|
"epoch": 0.382894082545997,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 8.0514,
|
|
"mean_token_accuracy": 0.08149674609303474,
|
|
"num_tokens": 755486.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.475645160675048,
|
|
"epoch": 0.3878667329686723,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 8.1482,
|
|
"mean_token_accuracy": 0.07295819409191609,
|
|
"num_tokens": 765481.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.463741111755372,
|
|
"epoch": 0.3928393833913476,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 8.1135,
|
|
"mean_token_accuracy": 0.07425905987620354,
|
|
"num_tokens": 774335.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.596048164367676,
|
|
"epoch": 0.39781203381402286,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 8.2392,
|
|
"mean_token_accuracy": 0.07124232538044453,
|
|
"num_tokens": 784307.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.549865531921387,
|
|
"epoch": 0.40278468423669817,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 8.2106,
|
|
"mean_token_accuracy": 0.07055418565869331,
|
|
"num_tokens": 795058.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.556127262115478,
|
|
"epoch": 0.40775733465937347,
|
|
"grad_norm": 2.890625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 8.1755,
|
|
"mean_token_accuracy": 0.07405509985983372,
|
|
"num_tokens": 805687.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.495815467834472,
|
|
"epoch": 0.4127299850820487,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 8.1541,
|
|
"mean_token_accuracy": 0.07714710012078285,
|
|
"num_tokens": 815138.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.47137746810913,
|
|
"epoch": 0.41770263550472403,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 8.1172,
|
|
"mean_token_accuracy": 0.07492221929132939,
|
|
"num_tokens": 825009.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.46223554611206,
|
|
"epoch": 0.4226752859273993,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 8.1004,
|
|
"mean_token_accuracy": 0.07772790230810642,
|
|
"num_tokens": 834708.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.475219249725342,
|
|
"epoch": 0.4276479363500746,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 8.1753,
|
|
"mean_token_accuracy": 0.07542566247284413,
|
|
"num_tokens": 843888.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.477521610260009,
|
|
"epoch": 0.4326205867727499,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 8.1222,
|
|
"mean_token_accuracy": 0.07436901330947876,
|
|
"num_tokens": 853057.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.48756446838379,
|
|
"epoch": 0.43759323719542514,
|
|
"grad_norm": 1.53125,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 8.1761,
|
|
"mean_token_accuracy": 0.07114730812609196,
|
|
"num_tokens": 862701.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.38951072692871,
|
|
"epoch": 0.44256588761810045,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 8.0672,
|
|
"mean_token_accuracy": 0.07624398544430733,
|
|
"num_tokens": 872129.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.414244651794434,
|
|
"epoch": 0.44753853804077576,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 8.0759,
|
|
"mean_token_accuracy": 0.0760790079832077,
|
|
"num_tokens": 883166.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.476319980621337,
|
|
"epoch": 0.452511188463451,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 8.0367,
|
|
"mean_token_accuracy": 0.08530011028051376,
|
|
"num_tokens": 894159.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.479471683502197,
|
|
"epoch": 0.4574838388861263,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 8.0962,
|
|
"mean_token_accuracy": 0.07649504765868187,
|
|
"num_tokens": 904588.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.296580982208251,
|
|
"epoch": 0.46245648930880157,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 8.0798,
|
|
"mean_token_accuracy": 0.0781316339969635,
|
|
"num_tokens": 915554.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 8.486116504669189,
|
|
"epoch": 0.4674291397314769,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 8.0891,
|
|
"mean_token_accuracy": 0.07881890460848809,
|
|
"num_tokens": 925546.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.337780475616455,
|
|
"epoch": 0.4724017901541522,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 8.0598,
|
|
"mean_token_accuracy": 0.07934539914131164,
|
|
"num_tokens": 936615.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.46543378829956,
|
|
"epoch": 0.47737444057682743,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 8.0858,
|
|
"mean_token_accuracy": 0.07611360922455787,
|
|
"num_tokens": 947374.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.334126091003418,
|
|
"epoch": 0.48234709099950274,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 8.0012,
|
|
"mean_token_accuracy": 0.08066745214164257,
|
|
"num_tokens": 957772.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.434980773925782,
|
|
"epoch": 0.48731974142217804,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 8.0842,
|
|
"mean_token_accuracy": 0.07925943285226822,
|
|
"num_tokens": 967724.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.30348539352417,
|
|
"epoch": 0.4922923918448533,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.9835,
|
|
"mean_token_accuracy": 0.07751845642924308,
|
|
"num_tokens": 976497.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.317158412933349,
|
|
"epoch": 0.4972650422675286,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.9707,
|
|
"mean_token_accuracy": 0.08364193737506867,
|
|
"num_tokens": 986207.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.4972650422675286,
|
|
"eval_entropy": 8.216110460219845,
|
|
"eval_loss": 8.090629577636719,
|
|
"eval_mean_token_accuracy": 0.07677195617748846,
|
|
"eval_num_tokens": 986207.0,
|
|
"eval_runtime": 1.9669,
|
|
"eval_samples_per_second": 1761.677,
|
|
"eval_steps_per_second": 220.654,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 10050,
|
|
"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": 238121673523200.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|