fasthypernet-6 / checkpoint-1300 /trainer_state.json
aixk's picture
Upload folder using huggingface_hub
53cf616 verified
Raw
History Blame Contribute Delete
22.7 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.1999923079881543,
"eval_steps": 500,
"global_step": 1300,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0015384023691396484,
"grad_norm": 11396.142578125,
"learning_rate": 3e-07,
"loss": 8.8759,
"step": 10
},
{
"epoch": 0.003076804738279297,
"grad_norm": 689.1935424804688,
"learning_rate": 1.2999999999999998e-06,
"loss": 8.2644,
"step": 20
},
{
"epoch": 0.004615207107418946,
"grad_norm": 454.8230285644531,
"learning_rate": 2.2999999999999996e-06,
"loss": 7.2145,
"step": 30
},
{
"epoch": 0.006153609476558594,
"grad_norm": 169.3773193359375,
"learning_rate": 3.2999999999999993e-06,
"loss": 6.899,
"step": 40
},
{
"epoch": 0.007692011845698243,
"grad_norm": 121.77935028076172,
"learning_rate": 4.2999999999999995e-06,
"loss": 6.7803,
"step": 50
},
{
"epoch": 0.009230414214837892,
"grad_norm": 1289.1287841796875,
"learning_rate": 5.3e-06,
"loss": 6.5956,
"step": 60
},
{
"epoch": 0.010768816583977539,
"grad_norm": 157.25286865234375,
"learning_rate": 6.3e-06,
"loss": 6.6402,
"step": 70
},
{
"epoch": 0.012307218953117188,
"grad_norm": 384.85919189453125,
"learning_rate": 7.299999999999999e-06,
"loss": 6.4978,
"step": 80
},
{
"epoch": 0.013845621322256836,
"grad_norm": 301.0735778808594,
"learning_rate": 8.299999999999998e-06,
"loss": 6.3964,
"step": 90
},
{
"epoch": 0.015384023691396485,
"grad_norm": 226.9570770263672,
"learning_rate": 9.299999999999999e-06,
"loss": 6.3804,
"step": 100
},
{
"epoch": 0.016922426060536132,
"grad_norm": 101.4676513671875,
"learning_rate": 1.03e-05,
"loss": 6.3698,
"step": 110
},
{
"epoch": 0.018460828429675783,
"grad_norm": 230.73281860351562,
"learning_rate": 1.1299999999999999e-05,
"loss": 6.4247,
"step": 120
},
{
"epoch": 0.01999923079881543,
"grad_norm": 124.10091400146484,
"learning_rate": 1.2299999999999999e-05,
"loss": 6.3109,
"step": 130
},
{
"epoch": 0.021537633167955077,
"grad_norm": 97.28553009033203,
"learning_rate": 1.33e-05,
"loss": 6.3895,
"step": 140
},
{
"epoch": 0.023076035537094728,
"grad_norm": 182.08811950683594,
"learning_rate": 1.43e-05,
"loss": 6.4089,
"step": 150
},
{
"epoch": 0.024614437906234375,
"grad_norm": 433.91973876953125,
"learning_rate": 1.53e-05,
"loss": 6.3014,
"step": 160
},
{
"epoch": 0.026152840275374026,
"grad_norm": 352.3630065917969,
"learning_rate": 1.6299999999999996e-05,
"loss": 6.3332,
"step": 170
},
{
"epoch": 0.027691242644513673,
"grad_norm": 74.03328704833984,
"learning_rate": 1.7299999999999997e-05,
"loss": 6.315,
"step": 180
},
{
"epoch": 0.02922964501365332,
"grad_norm": 264.071044921875,
"learning_rate": 1.8299999999999998e-05,
"loss": 6.3367,
"step": 190
},
{
"epoch": 0.03076804738279297,
"grad_norm": 92.3805160522461,
"learning_rate": 1.93e-05,
"loss": 6.306,
"step": 200
},
{
"epoch": 0.03230644975193262,
"grad_norm": 105.78662872314453,
"learning_rate": 2.03e-05,
"loss": 6.272,
"step": 210
},
{
"epoch": 0.033844852121072265,
"grad_norm": 34.507057189941406,
"learning_rate": 2.1299999999999996e-05,
"loss": 6.2943,
"step": 220
},
{
"epoch": 0.035383254490211916,
"grad_norm": 29.95108985900879,
"learning_rate": 2.23e-05,
"loss": 6.2231,
"step": 230
},
{
"epoch": 0.036921656859351566,
"grad_norm": 105.78109741210938,
"learning_rate": 2.3299999999999997e-05,
"loss": 6.2223,
"step": 240
},
{
"epoch": 0.03846005922849121,
"grad_norm": 115.77278900146484,
"learning_rate": 2.4299999999999998e-05,
"loss": 6.1873,
"step": 250
},
{
"epoch": 0.03999846159763086,
"grad_norm": 117.57398223876953,
"learning_rate": 2.53e-05,
"loss": 6.1889,
"step": 260
},
{
"epoch": 0.04153686396677051,
"grad_norm": 27.769681930541992,
"learning_rate": 2.63e-05,
"loss": 6.1833,
"step": 270
},
{
"epoch": 0.043075266335910155,
"grad_norm": 50.23863220214844,
"learning_rate": 2.7299999999999996e-05,
"loss": 6.1718,
"step": 280
},
{
"epoch": 0.044613668705049805,
"grad_norm": 16.450408935546875,
"learning_rate": 2.83e-05,
"loss": 6.0966,
"step": 290
},
{
"epoch": 0.046152071074189456,
"grad_norm": 28.899629592895508,
"learning_rate": 2.9299999999999997e-05,
"loss": 6.146,
"step": 300
},
{
"epoch": 0.0476904734433291,
"grad_norm": 47.915279388427734,
"learning_rate": 3.0299999999999998e-05,
"loss": 6.1493,
"step": 310
},
{
"epoch": 0.04922887581246875,
"grad_norm": 27.1001033782959,
"learning_rate": 3.1299999999999995e-05,
"loss": 6.1186,
"step": 320
},
{
"epoch": 0.0507672781816084,
"grad_norm": 22.059106826782227,
"learning_rate": 3.229999999999999e-05,
"loss": 6.0607,
"step": 330
},
{
"epoch": 0.05230568055074805,
"grad_norm": 19.7325496673584,
"learning_rate": 3.3299999999999996e-05,
"loss": 6.1196,
"step": 340
},
{
"epoch": 0.053844082919887695,
"grad_norm": 100.60061645507812,
"learning_rate": 3.4299999999999993e-05,
"loss": 6.1176,
"step": 350
},
{
"epoch": 0.055382485289027346,
"grad_norm": 18.382129669189453,
"learning_rate": 3.53e-05,
"loss": 6.1501,
"step": 360
},
{
"epoch": 0.056920887658166996,
"grad_norm": 31.890426635742188,
"learning_rate": 3.6299999999999995e-05,
"loss": 6.0863,
"step": 370
},
{
"epoch": 0.05845929002730664,
"grad_norm": 22.61912727355957,
"learning_rate": 3.73e-05,
"loss": 6.0543,
"step": 380
},
{
"epoch": 0.05999769239644629,
"grad_norm": 30.638811111450195,
"learning_rate": 3.83e-05,
"loss": 5.989,
"step": 390
},
{
"epoch": 0.06153609476558594,
"grad_norm": 45.842227935791016,
"learning_rate": 3.93e-05,
"loss": 6.0216,
"step": 400
},
{
"epoch": 0.06307449713472559,
"grad_norm": 61.62559127807617,
"learning_rate": 4.03e-05,
"loss": 6.0232,
"step": 410
},
{
"epoch": 0.06461289950386524,
"grad_norm": 24.91366958618164,
"learning_rate": 4.1299999999999994e-05,
"loss": 6.0153,
"step": 420
},
{
"epoch": 0.06615130187300489,
"grad_norm": 22.554046630859375,
"learning_rate": 4.229999999999999e-05,
"loss": 5.9533,
"step": 430
},
{
"epoch": 0.06768970424214453,
"grad_norm": 22.699480056762695,
"learning_rate": 4.3299999999999995e-05,
"loss": 5.9179,
"step": 440
},
{
"epoch": 0.06922810661128419,
"grad_norm": 17.466867446899414,
"learning_rate": 4.43e-05,
"loss": 5.9714,
"step": 450
},
{
"epoch": 0.07076650898042383,
"grad_norm": 20.46624183654785,
"learning_rate": 4.5299999999999997e-05,
"loss": 5.9621,
"step": 460
},
{
"epoch": 0.07230491134956347,
"grad_norm": 15.140242576599121,
"learning_rate": 4.6299999999999994e-05,
"loss": 5.94,
"step": 470
},
{
"epoch": 0.07384331371870313,
"grad_norm": 23.270614624023438,
"learning_rate": 4.73e-05,
"loss": 5.9387,
"step": 480
},
{
"epoch": 0.07538171608784278,
"grad_norm": 23.621036529541016,
"learning_rate": 4.8299999999999995e-05,
"loss": 5.9683,
"step": 490
},
{
"epoch": 0.07692011845698242,
"grad_norm": 16.960304260253906,
"learning_rate": 4.929999999999999e-05,
"loss": 5.9136,
"step": 500
},
{
"epoch": 0.07845852082612208,
"grad_norm": 17.167306900024414,
"learning_rate": 5.0299999999999996e-05,
"loss": 5.9011,
"step": 510
},
{
"epoch": 0.07999692319526172,
"grad_norm": 28.44086456298828,
"learning_rate": 5.13e-05,
"loss": 5.88,
"step": 520
},
{
"epoch": 0.08153532556440136,
"grad_norm": 19.138957977294922,
"learning_rate": 5.23e-05,
"loss": 5.8525,
"step": 530
},
{
"epoch": 0.08307372793354102,
"grad_norm": 23.37527847290039,
"learning_rate": 5.3299999999999995e-05,
"loss": 5.9051,
"step": 540
},
{
"epoch": 0.08461213030268067,
"grad_norm": 14.438946723937988,
"learning_rate": 5.429999999999999e-05,
"loss": 5.8024,
"step": 550
},
{
"epoch": 0.08615053267182031,
"grad_norm": 17.277740478515625,
"learning_rate": 5.529999999999999e-05,
"loss": 5.764,
"step": 560
},
{
"epoch": 0.08768893504095997,
"grad_norm": 18.269975662231445,
"learning_rate": 5.63e-05,
"loss": 5.8667,
"step": 570
},
{
"epoch": 0.08922733741009961,
"grad_norm": 17.265398025512695,
"learning_rate": 5.73e-05,
"loss": 5.7963,
"step": 580
},
{
"epoch": 0.09076573977923925,
"grad_norm": 20.076812744140625,
"learning_rate": 5.8299999999999994e-05,
"loss": 5.7426,
"step": 590
},
{
"epoch": 0.09230414214837891,
"grad_norm": 25.95227813720703,
"learning_rate": 5.929999999999999e-05,
"loss": 5.7171,
"step": 600
},
{
"epoch": 0.09384254451751856,
"grad_norm": 21.3321590423584,
"learning_rate": 6.0299999999999995e-05,
"loss": 5.7266,
"step": 610
},
{
"epoch": 0.0953809468866582,
"grad_norm": 33.217811584472656,
"learning_rate": 6.13e-05,
"loss": 5.6971,
"step": 620
},
{
"epoch": 0.09691934925579786,
"grad_norm": 16.669315338134766,
"learning_rate": 6.23e-05,
"loss": 5.7298,
"step": 630
},
{
"epoch": 0.0984577516249375,
"grad_norm": 14.09073543548584,
"learning_rate": 6.33e-05,
"loss": 5.7124,
"step": 640
},
{
"epoch": 0.09999615399407714,
"grad_norm": 19.583980560302734,
"learning_rate": 6.429999999999999e-05,
"loss": 5.7241,
"step": 650
},
{
"epoch": 0.1015345563632168,
"grad_norm": 24.256364822387695,
"learning_rate": 6.53e-05,
"loss": 5.7128,
"step": 660
},
{
"epoch": 0.10307295873235645,
"grad_norm": 23.34236717224121,
"learning_rate": 6.63e-05,
"loss": 5.5961,
"step": 670
},
{
"epoch": 0.1046113611014961,
"grad_norm": 20.654870986938477,
"learning_rate": 6.73e-05,
"loss": 5.5613,
"step": 680
},
{
"epoch": 0.10614976347063575,
"grad_norm": 30.639785766601562,
"learning_rate": 6.829999999999999e-05,
"loss": 5.6384,
"step": 690
},
{
"epoch": 0.10768816583977539,
"grad_norm": 15.175138473510742,
"learning_rate": 6.93e-05,
"loss": 5.5526,
"step": 700
},
{
"epoch": 0.10922656820891505,
"grad_norm": 20.746421813964844,
"learning_rate": 7.03e-05,
"loss": 5.6075,
"step": 710
},
{
"epoch": 0.11076497057805469,
"grad_norm": 14.452651023864746,
"learning_rate": 7.13e-05,
"loss": 5.6052,
"step": 720
},
{
"epoch": 0.11230337294719434,
"grad_norm": 15.90872859954834,
"learning_rate": 7.23e-05,
"loss": 5.5681,
"step": 730
},
{
"epoch": 0.11384177531633399,
"grad_norm": 19.600723266601562,
"learning_rate": 7.329999999999999e-05,
"loss": 5.5155,
"step": 740
},
{
"epoch": 0.11538017768547364,
"grad_norm": 15.309789657592773,
"learning_rate": 7.429999999999999e-05,
"loss": 5.5764,
"step": 750
},
{
"epoch": 0.11691858005461328,
"grad_norm": 19.05169677734375,
"learning_rate": 7.529999999999999e-05,
"loss": 5.4731,
"step": 760
},
{
"epoch": 0.11845698242375294,
"grad_norm": 19.819692611694336,
"learning_rate": 7.63e-05,
"loss": 5.3959,
"step": 770
},
{
"epoch": 0.11999538479289258,
"grad_norm": 19.908710479736328,
"learning_rate": 7.73e-05,
"loss": 5.43,
"step": 780
},
{
"epoch": 0.12153378716203223,
"grad_norm": 22.767356872558594,
"learning_rate": 7.829999999999999e-05,
"loss": 5.432,
"step": 790
},
{
"epoch": 0.12307218953117188,
"grad_norm": 14.213640213012695,
"learning_rate": 7.929999999999999e-05,
"loss": 5.4079,
"step": 800
},
{
"epoch": 0.12461059190031153,
"grad_norm": 16.46720314025879,
"learning_rate": 8.029999999999999e-05,
"loss": 5.4254,
"step": 810
},
{
"epoch": 0.12614899426945117,
"grad_norm": 19.044301986694336,
"learning_rate": 8.13e-05,
"loss": 5.3561,
"step": 820
},
{
"epoch": 0.12768739663859083,
"grad_norm": 23.400588989257812,
"learning_rate": 8.23e-05,
"loss": 5.4002,
"step": 830
},
{
"epoch": 0.12922579900773049,
"grad_norm": 17.108936309814453,
"learning_rate": 8.329999999999999e-05,
"loss": 5.3148,
"step": 840
},
{
"epoch": 0.13076420137687012,
"grad_norm": 13.769838333129883,
"learning_rate": 8.43e-05,
"loss": 5.2701,
"step": 850
},
{
"epoch": 0.13230260374600977,
"grad_norm": 17.94414520263672,
"learning_rate": 8.529999999999999e-05,
"loss": 5.3379,
"step": 860
},
{
"epoch": 0.13384100611514943,
"grad_norm": 15.634041786193848,
"learning_rate": 8.63e-05,
"loss": 5.2263,
"step": 870
},
{
"epoch": 0.13537940848428906,
"grad_norm": 16.440696716308594,
"learning_rate": 8.729999999999998e-05,
"loss": 5.2153,
"step": 880
},
{
"epoch": 0.13691781085342872,
"grad_norm": 14.667311668395996,
"learning_rate": 8.829999999999999e-05,
"loss": 5.2754,
"step": 890
},
{
"epoch": 0.13845621322256838,
"grad_norm": 12.900784492492676,
"learning_rate": 8.93e-05,
"loss": 5.2061,
"step": 900
},
{
"epoch": 0.139994615591708,
"grad_norm": 15.827488899230957,
"learning_rate": 9.029999999999999e-05,
"loss": 5.2572,
"step": 910
},
{
"epoch": 0.14153301796084766,
"grad_norm": 15.8163480758667,
"learning_rate": 9.13e-05,
"loss": 5.1689,
"step": 920
},
{
"epoch": 0.14307142032998732,
"grad_norm": 13.711785316467285,
"learning_rate": 9.229999999999998e-05,
"loss": 5.1061,
"step": 930
},
{
"epoch": 0.14460982269912695,
"grad_norm": 20.83718490600586,
"learning_rate": 9.329999999999999e-05,
"loss": 5.2162,
"step": 940
},
{
"epoch": 0.1461482250682666,
"grad_norm": 17.33615493774414,
"learning_rate": 9.43e-05,
"loss": 5.0812,
"step": 950
},
{
"epoch": 0.14768662743740626,
"grad_norm": 16.08768081665039,
"learning_rate": 9.529999999999999e-05,
"loss": 5.0975,
"step": 960
},
{
"epoch": 0.1492250298065459,
"grad_norm": 16.137598037719727,
"learning_rate": 9.63e-05,
"loss": 5.0941,
"step": 970
},
{
"epoch": 0.15076343217568555,
"grad_norm": 14.446563720703125,
"learning_rate": 9.729999999999998e-05,
"loss": 5.1043,
"step": 980
},
{
"epoch": 0.1523018345448252,
"grad_norm": 14.645479202270508,
"learning_rate": 9.829999999999999e-05,
"loss": 5.056,
"step": 990
},
{
"epoch": 0.15384023691396484,
"grad_norm": 12.247152328491211,
"learning_rate": 9.93e-05,
"loss": 5.0355,
"step": 1000
},
{
"epoch": 0.1553786392831045,
"grad_norm": 17.88384437561035,
"learning_rate": 0.00010029999999999998,
"loss": 4.97,
"step": 1010
},
{
"epoch": 0.15691704165224415,
"grad_norm": 17.317882537841797,
"learning_rate": 0.0001013,
"loss": 4.9802,
"step": 1020
},
{
"epoch": 0.15845544402138378,
"grad_norm": 12.667706489562988,
"learning_rate": 0.00010229999999999999,
"loss": 5.0057,
"step": 1030
},
{
"epoch": 0.15999384639052344,
"grad_norm": 15.451706886291504,
"learning_rate": 0.00010329999999999999,
"loss": 4.9228,
"step": 1040
},
{
"epoch": 0.1615322487596631,
"grad_norm": 14.22465705871582,
"learning_rate": 0.0001043,
"loss": 5.0062,
"step": 1050
},
{
"epoch": 0.16307065112880273,
"grad_norm": 15.487550735473633,
"learning_rate": 0.00010529999999999998,
"loss": 4.9031,
"step": 1060
},
{
"epoch": 0.1646090534979424,
"grad_norm": 13.079480171203613,
"learning_rate": 0.0001063,
"loss": 4.8262,
"step": 1070
},
{
"epoch": 0.16614745586708204,
"grad_norm": 14.858409881591797,
"learning_rate": 0.00010729999999999999,
"loss": 4.8865,
"step": 1080
},
{
"epoch": 0.16768585823622167,
"grad_norm": 15.18587589263916,
"learning_rate": 0.00010829999999999999,
"loss": 4.8683,
"step": 1090
},
{
"epoch": 0.16922426060536133,
"grad_norm": 13.830065727233887,
"learning_rate": 0.00010929999999999999,
"loss": 4.7915,
"step": 1100
},
{
"epoch": 0.170762662974501,
"grad_norm": 13.956897735595703,
"learning_rate": 0.00011029999999999998,
"loss": 4.8107,
"step": 1110
},
{
"epoch": 0.17230106534364062,
"grad_norm": 14.276887893676758,
"learning_rate": 0.0001113,
"loss": 4.7619,
"step": 1120
},
{
"epoch": 0.17383946771278028,
"grad_norm": 16.42158317565918,
"learning_rate": 0.00011229999999999999,
"loss": 4.7659,
"step": 1130
},
{
"epoch": 0.17537787008191993,
"grad_norm": 15.550186157226562,
"learning_rate": 0.00011329999999999999,
"loss": 4.706,
"step": 1140
},
{
"epoch": 0.17691627245105956,
"grad_norm": 13.871357917785645,
"learning_rate": 0.00011429999999999999,
"loss": 4.7335,
"step": 1150
},
{
"epoch": 0.17845467482019922,
"grad_norm": 13.904778480529785,
"learning_rate": 0.0001153,
"loss": 4.6234,
"step": 1160
},
{
"epoch": 0.17999307718933888,
"grad_norm": 12.691954612731934,
"learning_rate": 0.00011629999999999998,
"loss": 4.664,
"step": 1170
},
{
"epoch": 0.1815314795584785,
"grad_norm": 13.613678932189941,
"learning_rate": 0.00011729999999999999,
"loss": 4.6153,
"step": 1180
},
{
"epoch": 0.18306988192761817,
"grad_norm": 19.315826416015625,
"learning_rate": 0.00011829999999999999,
"loss": 4.6475,
"step": 1190
},
{
"epoch": 0.18460828429675782,
"grad_norm": 12.154134750366211,
"learning_rate": 0.00011929999999999999,
"loss": 4.6046,
"step": 1200
},
{
"epoch": 0.18614668666589745,
"grad_norm": 12.934707641601562,
"learning_rate": 0.0001203,
"loss": 4.6061,
"step": 1210
},
{
"epoch": 0.1876850890350371,
"grad_norm": 14.66080379486084,
"learning_rate": 0.00012129999999999998,
"loss": 4.5456,
"step": 1220
},
{
"epoch": 0.18922349140417677,
"grad_norm": 14.203516006469727,
"learning_rate": 0.0001223,
"loss": 4.5529,
"step": 1230
},
{
"epoch": 0.1907618937733164,
"grad_norm": 16.82846450805664,
"learning_rate": 0.0001233,
"loss": 4.5334,
"step": 1240
},
{
"epoch": 0.19230029614245606,
"grad_norm": 16.63568687438965,
"learning_rate": 0.00012429999999999999,
"loss": 4.4209,
"step": 1250
},
{
"epoch": 0.19383869851159571,
"grad_norm": 13.418827056884766,
"learning_rate": 0.00012529999999999998,
"loss": 4.4474,
"step": 1260
},
{
"epoch": 0.19537710088073534,
"grad_norm": 14.576326370239258,
"learning_rate": 0.00012629999999999998,
"loss": 4.4721,
"step": 1270
},
{
"epoch": 0.196915503249875,
"grad_norm": 14.08271312713623,
"learning_rate": 0.0001273,
"loss": 4.4659,
"step": 1280
},
{
"epoch": 0.19845390561901466,
"grad_norm": 15.0554780960083,
"learning_rate": 0.00012829999999999997,
"loss": 4.4189,
"step": 1290
},
{
"epoch": 0.1999923079881543,
"grad_norm": 13.966469764709473,
"learning_rate": 0.0001293,
"loss": 4.3879,
"step": 1300
}
],
"logging_steps": 10,
"max_steps": 100000,
"num_input_tokens_seen": 0,
"num_train_epochs": 16,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.504746637180928e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}