b5f745adf6
Model: fpadovani/dan-latn-100mb-after-ppt-shuff-dyck-10mb-ckpt500_seed3407 Source: Original Platform
1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.41841004184100417,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.768266344070435,
|
|
"epoch": 0.0041841004184100415,
|
|
"grad_norm": 12.875,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.3834,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 9848.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.7789147853851315,
|
|
"epoch": 0.008368200836820083,
|
|
"grad_norm": 14.3125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.2092,
|
|
"mean_token_accuracy": 0.00013605442363768817,
|
|
"num_tokens": 18196.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.821842193603516,
|
|
"epoch": 0.012552301255230125,
|
|
"grad_norm": 17.25,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.0541,
|
|
"mean_token_accuracy": 0.0002630572533234954,
|
|
"num_tokens": 26927.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 5.096258974075317,
|
|
"epoch": 0.016736401673640166,
|
|
"grad_norm": 27.125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.3612,
|
|
"mean_token_accuracy": 0.00012658227933570743,
|
|
"num_tokens": 34987.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 7.207125568389893,
|
|
"epoch": 0.02092050209205021,
|
|
"grad_norm": 19.5,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 11.9388,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 44936.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.419103336334228,
|
|
"epoch": 0.02510460251046025,
|
|
"grad_norm": 4.25,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.8146,
|
|
"mean_token_accuracy": 0.0020074134168680756,
|
|
"num_tokens": 54562.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.69917049407959,
|
|
"epoch": 0.029288702928870293,
|
|
"grad_norm": 3.296875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.5537,
|
|
"mean_token_accuracy": 0.0153768889605999,
|
|
"num_tokens": 63844.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.692293643951416,
|
|
"epoch": 0.03347280334728033,
|
|
"grad_norm": 2.71875,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.2761,
|
|
"mean_token_accuracy": 0.017915738467127084,
|
|
"num_tokens": 74331.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.646587371826172,
|
|
"epoch": 0.03765690376569038,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.904,
|
|
"mean_token_accuracy": 0.028478659689426422,
|
|
"num_tokens": 83028.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.585913181304932,
|
|
"epoch": 0.04184100418410042,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.6977,
|
|
"mean_token_accuracy": 0.04541164524853229,
|
|
"num_tokens": 91395.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.563683032989502,
|
|
"epoch": 0.04602510460251046,
|
|
"grad_norm": 2.421875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.5786,
|
|
"mean_token_accuracy": 0.048107235319912435,
|
|
"num_tokens": 101868.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.548775959014893,
|
|
"epoch": 0.0502092050209205,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.4624,
|
|
"mean_token_accuracy": 0.04703106284141541,
|
|
"num_tokens": 110403.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.547953414916993,
|
|
"epoch": 0.05439330543933055,
|
|
"grad_norm": 1.9609375,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.5071,
|
|
"mean_token_accuracy": 0.041691455990076065,
|
|
"num_tokens": 119557.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.555264663696288,
|
|
"epoch": 0.058577405857740586,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.3084,
|
|
"mean_token_accuracy": 0.04668764472007751,
|
|
"num_tokens": 128903.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.545580863952637,
|
|
"epoch": 0.06276150627615062,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.3022,
|
|
"mean_token_accuracy": 0.04386243894696236,
|
|
"num_tokens": 137630.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.519606494903565,
|
|
"epoch": 0.06694560669456066,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.12,
|
|
"mean_token_accuracy": 0.05209527462720871,
|
|
"num_tokens": 146335.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.482771396636963,
|
|
"epoch": 0.07112970711297072,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.0283,
|
|
"mean_token_accuracy": 0.05407022833824158,
|
|
"num_tokens": 155766.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.450575256347657,
|
|
"epoch": 0.07531380753138076,
|
|
"grad_norm": 1.890625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.9477,
|
|
"mean_token_accuracy": 0.05165062099695206,
|
|
"num_tokens": 165844.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.425101280212402,
|
|
"epoch": 0.0794979079497908,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.8842,
|
|
"mean_token_accuracy": 0.05370258502662182,
|
|
"num_tokens": 174858.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.32920789718628,
|
|
"epoch": 0.08368200836820083,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.7167,
|
|
"mean_token_accuracy": 0.05981660895049572,
|
|
"num_tokens": 184020.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.181023502349854,
|
|
"epoch": 0.08786610878661087,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.5856,
|
|
"mean_token_accuracy": 0.05290363691747189,
|
|
"num_tokens": 193714.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.198339366912842,
|
|
"epoch": 0.09205020920502092,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.5126,
|
|
"mean_token_accuracy": 0.0547657772898674,
|
|
"num_tokens": 203139.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.105471420288087,
|
|
"epoch": 0.09623430962343096,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.3484,
|
|
"mean_token_accuracy": 0.05634176060557365,
|
|
"num_tokens": 212446.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.886044597625732,
|
|
"epoch": 0.100418410041841,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.1725,
|
|
"mean_token_accuracy": 0.06103258766233921,
|
|
"num_tokens": 220910.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.764387893676759,
|
|
"epoch": 0.10460251046025104,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.096,
|
|
"mean_token_accuracy": 0.053500480577349664,
|
|
"num_tokens": 230522.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.598654460906982,
|
|
"epoch": 0.1087866108786611,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.918,
|
|
"mean_token_accuracy": 0.05621385425329208,
|
|
"num_tokens": 240092.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.302083587646484,
|
|
"epoch": 0.11297071129707113,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.8502,
|
|
"mean_token_accuracy": 0.05275089591741562,
|
|
"num_tokens": 249478.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.09952507019043,
|
|
"epoch": 0.11715481171548117,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.6323,
|
|
"mean_token_accuracy": 0.06236257366836071,
|
|
"num_tokens": 257714.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.845372676849365,
|
|
"epoch": 0.12133891213389121,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.672,
|
|
"mean_token_accuracy": 0.05868846960365772,
|
|
"num_tokens": 266569.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.660739707946778,
|
|
"epoch": 0.12552301255230125,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.5683,
|
|
"mean_token_accuracy": 0.056071584671735765,
|
|
"num_tokens": 276688.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.484380149841309,
|
|
"epoch": 0.1297071129707113,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.4186,
|
|
"mean_token_accuracy": 0.06363016776740552,
|
|
"num_tokens": 285430.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.278593254089355,
|
|
"epoch": 0.13389121338912133,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.4137,
|
|
"mean_token_accuracy": 0.06225575767457485,
|
|
"num_tokens": 295727.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.146096420288085,
|
|
"epoch": 0.13807531380753138,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.3823,
|
|
"mean_token_accuracy": 0.0689712069928646,
|
|
"num_tokens": 304854.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 7.988676738739014,
|
|
"epoch": 0.14225941422594143,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.3536,
|
|
"mean_token_accuracy": 0.07053346484899521,
|
|
"num_tokens": 313579.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 7.9785551071167,
|
|
"epoch": 0.14644351464435146,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.4132,
|
|
"mean_token_accuracy": 0.07007367871701717,
|
|
"num_tokens": 323310.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 7.922500085830689,
|
|
"epoch": 0.1506276150627615,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.2859,
|
|
"mean_token_accuracy": 0.0698877926915884,
|
|
"num_tokens": 331720.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 7.924635601043701,
|
|
"epoch": 0.15481171548117154,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.24,
|
|
"mean_token_accuracy": 0.07415937893092632,
|
|
"num_tokens": 340331.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 7.836748027801514,
|
|
"epoch": 0.1589958158995816,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.2493,
|
|
"mean_token_accuracy": 0.07680297121405602,
|
|
"num_tokens": 349677.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 7.846647310256958,
|
|
"epoch": 0.16317991631799164,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.2344,
|
|
"mean_token_accuracy": 0.06987173184752464,
|
|
"num_tokens": 358674.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.753779888153076,
|
|
"epoch": 0.16736401673640167,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.2296,
|
|
"mean_token_accuracy": 0.072166907787323,
|
|
"num_tokens": 368258.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.816751956939697,
|
|
"epoch": 0.17154811715481172,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.1436,
|
|
"mean_token_accuracy": 0.07943947166204453,
|
|
"num_tokens": 377083.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.698982191085816,
|
|
"epoch": 0.17573221757322174,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.1951,
|
|
"mean_token_accuracy": 0.07452086620032787,
|
|
"num_tokens": 385456.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.709808778762818,
|
|
"epoch": 0.1799163179916318,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.0971,
|
|
"mean_token_accuracy": 0.0767411794513464,
|
|
"num_tokens": 395295.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.692369031906128,
|
|
"epoch": 0.18410041841004185,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.1406,
|
|
"mean_token_accuracy": 0.07074209861457348,
|
|
"num_tokens": 405023.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.62271146774292,
|
|
"epoch": 0.18828451882845187,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.1623,
|
|
"mean_token_accuracy": 0.07522214204072952,
|
|
"num_tokens": 414224.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.621326494216919,
|
|
"epoch": 0.19246861924686193,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.0582,
|
|
"mean_token_accuracy": 0.0759123720228672,
|
|
"num_tokens": 423951.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.696398591995239,
|
|
"epoch": 0.19665271966527198,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.0805,
|
|
"mean_token_accuracy": 0.07548626773059368,
|
|
"num_tokens": 432427.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.593403148651123,
|
|
"epoch": 0.200836820083682,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.0032,
|
|
"mean_token_accuracy": 0.0808411881327629,
|
|
"num_tokens": 441560.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.589205026626587,
|
|
"epoch": 0.20502092050209206,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.092,
|
|
"mean_token_accuracy": 0.0739415742456913,
|
|
"num_tokens": 450364.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.541353750228882,
|
|
"epoch": 0.20920502092050208,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.0399,
|
|
"mean_token_accuracy": 0.0743038758635521,
|
|
"num_tokens": 460076.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.461675310134888,
|
|
"epoch": 0.21338912133891214,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000127,
|
|
"loss": 6.997,
|
|
"mean_token_accuracy": 0.07904429398477078,
|
|
"num_tokens": 468718.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.617227411270141,
|
|
"epoch": 0.2175732217573222,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.0585,
|
|
"mean_token_accuracy": 0.07864544317126274,
|
|
"num_tokens": 478436.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.475518226623535,
|
|
"epoch": 0.2217573221757322,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.0628,
|
|
"mean_token_accuracy": 0.07705017924308777,
|
|
"num_tokens": 486833.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.514133024215698,
|
|
"epoch": 0.22594142259414227,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 6.9359,
|
|
"mean_token_accuracy": 0.08766618892550468,
|
|
"num_tokens": 495119.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.481584882736206,
|
|
"epoch": 0.2301255230125523,
|
|
"grad_norm": 1.921875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.1039,
|
|
"mean_token_accuracy": 0.08408359885215759,
|
|
"num_tokens": 505396.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.400933456420899,
|
|
"epoch": 0.23430962343096234,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 6.9713,
|
|
"mean_token_accuracy": 0.08110813125967979,
|
|
"num_tokens": 515259.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.482255935668945,
|
|
"epoch": 0.2384937238493724,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 6.9695,
|
|
"mean_token_accuracy": 0.08075925037264824,
|
|
"num_tokens": 523608.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.380628252029419,
|
|
"epoch": 0.24267782426778242,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 6.9614,
|
|
"mean_token_accuracy": 0.08079174906015396,
|
|
"num_tokens": 534613.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.431663274765015,
|
|
"epoch": 0.24686192468619247,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 6.9779,
|
|
"mean_token_accuracy": 0.08681405410170555,
|
|
"num_tokens": 543347.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.367568683624268,
|
|
"epoch": 0.2510460251046025,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 6.9373,
|
|
"mean_token_accuracy": 0.08151624575257302,
|
|
"num_tokens": 552371.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.431672620773315,
|
|
"epoch": 0.25523012552301255,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.8709,
|
|
"mean_token_accuracy": 0.08783738985657692,
|
|
"num_tokens": 560742.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.372213315963745,
|
|
"epoch": 0.2594142259414226,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 6.9672,
|
|
"mean_token_accuracy": 0.08577736541628837,
|
|
"num_tokens": 569960.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.366195917129517,
|
|
"epoch": 0.26359832635983266,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 6.935,
|
|
"mean_token_accuracy": 0.08148402348160744,
|
|
"num_tokens": 578859.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.361504554748535,
|
|
"epoch": 0.26778242677824265,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.9323,
|
|
"mean_token_accuracy": 0.08335322514176369,
|
|
"num_tokens": 588146.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.262340354919433,
|
|
"epoch": 0.2719665271966527,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.8837,
|
|
"mean_token_accuracy": 0.08676013201475144,
|
|
"num_tokens": 597268.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.286304664611817,
|
|
"epoch": 0.27615062761506276,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.8065,
|
|
"mean_token_accuracy": 0.09255077987909317,
|
|
"num_tokens": 606199.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.354407405853271,
|
|
"epoch": 0.2803347280334728,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.9221,
|
|
"mean_token_accuracy": 0.07822546139359474,
|
|
"num_tokens": 615966.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.328773784637451,
|
|
"epoch": 0.28451882845188287,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.8936,
|
|
"mean_token_accuracy": 0.08315553963184356,
|
|
"num_tokens": 625689.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.292076635360718,
|
|
"epoch": 0.28870292887029286,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.9258,
|
|
"mean_token_accuracy": 0.08275936022400857,
|
|
"num_tokens": 634877.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.311211967468262,
|
|
"epoch": 0.2928870292887029,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.8606,
|
|
"mean_token_accuracy": 0.08924969360232353,
|
|
"num_tokens": 644042.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.213941431045532,
|
|
"epoch": 0.29707112970711297,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.8394,
|
|
"mean_token_accuracy": 0.08540538996458054,
|
|
"num_tokens": 653072.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.225866985321045,
|
|
"epoch": 0.301255230125523,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.8411,
|
|
"mean_token_accuracy": 0.08791597038507462,
|
|
"num_tokens": 663686.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.237736654281616,
|
|
"epoch": 0.3054393305439331,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.774,
|
|
"mean_token_accuracy": 0.0887495793402195,
|
|
"num_tokens": 672293.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.061402034759522,
|
|
"epoch": 0.30962343096234307,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.7483,
|
|
"mean_token_accuracy": 0.09100355952978134,
|
|
"num_tokens": 681758.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.178809404373169,
|
|
"epoch": 0.3138075313807531,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.7736,
|
|
"mean_token_accuracy": 0.09110940396785736,
|
|
"num_tokens": 690727.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.200173044204712,
|
|
"epoch": 0.3179916317991632,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.7292,
|
|
"mean_token_accuracy": 0.08735076114535331,
|
|
"num_tokens": 701210.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.065630149841309,
|
|
"epoch": 0.32217573221757323,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.7168,
|
|
"mean_token_accuracy": 0.09401161968708038,
|
|
"num_tokens": 710869.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.200049877166748,
|
|
"epoch": 0.3263598326359833,
|
|
"grad_norm": 1.71875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.8296,
|
|
"mean_token_accuracy": 0.08618490770459175,
|
|
"num_tokens": 721455.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.104738330841064,
|
|
"epoch": 0.3305439330543933,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.7259,
|
|
"mean_token_accuracy": 0.09676406383514405,
|
|
"num_tokens": 730917.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.051578664779663,
|
|
"epoch": 0.33472803347280333,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.7446,
|
|
"mean_token_accuracy": 0.09235388711094857,
|
|
"num_tokens": 739302.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.140300941467285,
|
|
"epoch": 0.3389121338912134,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.7141,
|
|
"mean_token_accuracy": 0.08636429682374,
|
|
"num_tokens": 748785.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.019905233383179,
|
|
"epoch": 0.34309623430962344,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.6677,
|
|
"mean_token_accuracy": 0.09201755002140999,
|
|
"num_tokens": 757215.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.04399356842041,
|
|
"epoch": 0.3472803347280335,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.5945,
|
|
"mean_token_accuracy": 0.09794526174664497,
|
|
"num_tokens": 765682.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.094897365570068,
|
|
"epoch": 0.3514644351464435,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.7956,
|
|
"mean_token_accuracy": 0.08788713663816453,
|
|
"num_tokens": 774814.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.0791693210601805,
|
|
"epoch": 0.35564853556485354,
|
|
"grad_norm": 1.5625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.7096,
|
|
"mean_token_accuracy": 0.09370236843824387,
|
|
"num_tokens": 784019.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 6.943140363693237,
|
|
"epoch": 0.3598326359832636,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.5796,
|
|
"mean_token_accuracy": 0.09640125781297684,
|
|
"num_tokens": 793237.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.025385618209839,
|
|
"epoch": 0.36401673640167365,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.6309,
|
|
"mean_token_accuracy": 0.09518588036298752,
|
|
"num_tokens": 801548.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.028780126571656,
|
|
"epoch": 0.3682008368200837,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.6643,
|
|
"mean_token_accuracy": 0.09595847800374031,
|
|
"num_tokens": 810974.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 6.958902835845947,
|
|
"epoch": 0.3723849372384937,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.5681,
|
|
"mean_token_accuracy": 0.09765126407146454,
|
|
"num_tokens": 820023.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 6.949744319915771,
|
|
"epoch": 0.37656903765690375,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.6524,
|
|
"mean_token_accuracy": 0.09723440259695053,
|
|
"num_tokens": 828297.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.0269598960876465,
|
|
"epoch": 0.3807531380753138,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.7028,
|
|
"mean_token_accuracy": 0.08575191274285317,
|
|
"num_tokens": 838268.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.016048622131348,
|
|
"epoch": 0.38493723849372385,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.623,
|
|
"mean_token_accuracy": 0.09345634952187538,
|
|
"num_tokens": 846491.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 6.9051624774932865,
|
|
"epoch": 0.3891213389121339,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.5585,
|
|
"mean_token_accuracy": 0.10123546048998833,
|
|
"num_tokens": 856501.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 6.930023145675659,
|
|
"epoch": 0.39330543933054396,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.6253,
|
|
"mean_token_accuracy": 0.09574304819107056,
|
|
"num_tokens": 867054.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 6.9083311557769775,
|
|
"epoch": 0.39748953974895396,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.6316,
|
|
"mean_token_accuracy": 0.09689460024237632,
|
|
"num_tokens": 875956.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 6.80507345199585,
|
|
"epoch": 0.401673640167364,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.5549,
|
|
"mean_token_accuracy": 0.09844572842121124,
|
|
"num_tokens": 884427.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 6.988656139373779,
|
|
"epoch": 0.40585774058577406,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.6402,
|
|
"mean_token_accuracy": 0.10031970292329788,
|
|
"num_tokens": 894181.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 6.854498291015625,
|
|
"epoch": 0.4100418410041841,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.4914,
|
|
"mean_token_accuracy": 0.10208117067813874,
|
|
"num_tokens": 904076.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 6.82157187461853,
|
|
"epoch": 0.41422594142259417,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.6406,
|
|
"mean_token_accuracy": 0.10029161274433136,
|
|
"num_tokens": 913915.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 6.97306637763977,
|
|
"epoch": 0.41841004184100417,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.6935,
|
|
"mean_token_accuracy": 0.09564930200576782,
|
|
"num_tokens": 922876.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.41841004184100417,
|
|
"eval_entropy": 6.967948366527908,
|
|
"eval_loss": 6.639475345611572,
|
|
"eval_mean_token_accuracy": 0.09896371554743293,
|
|
"eval_num_tokens": 922876.0,
|
|
"eval_runtime": 2.8273,
|
|
"eval_samples_per_second": 1381.864,
|
|
"eval_steps_per_second": 172.954,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 11950,
|
|
"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": 1351337066496000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|