static_patch8_qwen / checkpoint-5000 /trainer_state.json
Chaeseung's picture
Upload folder using huggingface_hub
b83538c verified
Raw
History Blame Contribute Delete
41.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.8202524504618856,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01128270220717862,
"grad_norm": 872.0,
"learning_rate": 5.9375e-07,
"loss": 43.164,
"step": 20
},
{
"epoch": 0.02256540441435724,
"grad_norm": 462.0,
"learning_rate": 1.21875e-06,
"loss": 42.8795,
"step": 40
},
{
"epoch": 0.03384810662153586,
"grad_norm": 496.0,
"learning_rate": 1.8437500000000003e-06,
"loss": 41.8627,
"step": 60
},
{
"epoch": 0.04513080882871448,
"grad_norm": 272.0,
"learning_rate": 2.4687500000000003e-06,
"loss": 39.8411,
"step": 80
},
{
"epoch": 0.056413511035893094,
"grad_norm": 400.0,
"learning_rate": 3.0937500000000002e-06,
"loss": 37.796,
"step": 100
},
{
"epoch": 0.06769621324307172,
"grad_norm": 280.0,
"learning_rate": 3.7187500000000006e-06,
"loss": 36.7293,
"step": 120
},
{
"epoch": 0.07897891545025033,
"grad_norm": 332.0,
"learning_rate": 4.3437500000000006e-06,
"loss": 36.0082,
"step": 140
},
{
"epoch": 0.09026161765742896,
"grad_norm": 282.0,
"learning_rate": 4.9687500000000005e-06,
"loss": 35.8703,
"step": 160
},
{
"epoch": 0.10154431986460757,
"grad_norm": 240.0,
"learning_rate": 4.981585578600505e-06,
"loss": 34.2551,
"step": 180
},
{
"epoch": 0.11282702207178619,
"grad_norm": 183.0,
"learning_rate": 4.96220197712735e-06,
"loss": 33.9093,
"step": 200
},
{
"epoch": 0.12410972427896481,
"grad_norm": 2024.0,
"learning_rate": 4.942818375654197e-06,
"loss": 34.4173,
"step": 220
},
{
"epoch": 0.13539242648614344,
"grad_norm": 182.0,
"learning_rate": 4.9234347741810434e-06,
"loss": 34.3021,
"step": 240
},
{
"epoch": 0.14667512869332205,
"grad_norm": 214.0,
"learning_rate": 4.90405117270789e-06,
"loss": 34.0691,
"step": 260
},
{
"epoch": 0.15795783090050067,
"grad_norm": 172.0,
"learning_rate": 4.884667571234736e-06,
"loss": 33.5895,
"step": 280
},
{
"epoch": 0.16924053310767928,
"grad_norm": 204.0,
"learning_rate": 4.865283969761582e-06,
"loss": 33.6394,
"step": 300
},
{
"epoch": 0.18052323531485792,
"grad_norm": 192.0,
"learning_rate": 4.845900368288428e-06,
"loss": 33.614,
"step": 320
},
{
"epoch": 0.19180593752203653,
"grad_norm": 209.0,
"learning_rate": 4.826516766815275e-06,
"loss": 34.4056,
"step": 340
},
{
"epoch": 0.20308863972921515,
"grad_norm": 201.0,
"learning_rate": 4.807133165342121e-06,
"loss": 33.5044,
"step": 360
},
{
"epoch": 0.21437134193639376,
"grad_norm": 175.0,
"learning_rate": 4.787749563868967e-06,
"loss": 32.7744,
"step": 380
},
{
"epoch": 0.22565404414357237,
"grad_norm": 171.0,
"learning_rate": 4.768365962395813e-06,
"loss": 33.6279,
"step": 400
},
{
"epoch": 0.23693674635075102,
"grad_norm": 171.0,
"learning_rate": 4.74898236092266e-06,
"loss": 33.6311,
"step": 420
},
{
"epoch": 0.24821944855792963,
"grad_norm": 217.0,
"learning_rate": 4.729598759449507e-06,
"loss": 33.1385,
"step": 440
},
{
"epoch": 0.25950215076510824,
"grad_norm": 198.0,
"learning_rate": 4.710215157976352e-06,
"loss": 33.3726,
"step": 460
},
{
"epoch": 0.2707848529722869,
"grad_norm": 188.0,
"learning_rate": 4.690831556503198e-06,
"loss": 33.4124,
"step": 480
},
{
"epoch": 0.28206755517946547,
"grad_norm": 209.0,
"learning_rate": 4.671447955030045e-06,
"loss": 33.477,
"step": 500
},
{
"epoch": 0.2933502573866441,
"grad_norm": 184.0,
"learning_rate": 4.652064353556891e-06,
"loss": 32.9445,
"step": 520
},
{
"epoch": 0.3046329595938227,
"grad_norm": 183.0,
"learning_rate": 4.632680752083738e-06,
"loss": 34.1471,
"step": 540
},
{
"epoch": 0.31591566180100134,
"grad_norm": 190.0,
"learning_rate": 4.613297150610584e-06,
"loss": 34.0529,
"step": 560
},
{
"epoch": 0.32719836400818,
"grad_norm": 177.0,
"learning_rate": 4.59391354913743e-06,
"loss": 33.6535,
"step": 580
},
{
"epoch": 0.33848106621535856,
"grad_norm": 194.0,
"learning_rate": 4.574529947664277e-06,
"loss": 33.4617,
"step": 600
},
{
"epoch": 0.3497637684225372,
"grad_norm": 240.0,
"learning_rate": 4.555146346191123e-06,
"loss": 33.9342,
"step": 620
},
{
"epoch": 0.36104647062971584,
"grad_norm": 182.0,
"learning_rate": 4.535762744717969e-06,
"loss": 34.2629,
"step": 640
},
{
"epoch": 0.37232917283689443,
"grad_norm": 198.0,
"learning_rate": 4.516379143244815e-06,
"loss": 33.0204,
"step": 660
},
{
"epoch": 0.38361187504407307,
"grad_norm": 178.0,
"learning_rate": 4.4969955417716615e-06,
"loss": 33.6778,
"step": 680
},
{
"epoch": 0.39489457725125165,
"grad_norm": 201.0,
"learning_rate": 4.477611940298508e-06,
"loss": 34.1614,
"step": 700
},
{
"epoch": 0.4061772794584303,
"grad_norm": 191.0,
"learning_rate": 4.458228338825354e-06,
"loss": 34.1801,
"step": 720
},
{
"epoch": 0.41745998166560894,
"grad_norm": 191.0,
"learning_rate": 4.4388447373522e-06,
"loss": 34.2427,
"step": 740
},
{
"epoch": 0.4287426838727875,
"grad_norm": 218.0,
"learning_rate": 4.4194611358790465e-06,
"loss": 34.6679,
"step": 760
},
{
"epoch": 0.44002538607996616,
"grad_norm": 187.0,
"learning_rate": 4.400077534405894e-06,
"loss": 33.7898,
"step": 780
},
{
"epoch": 0.45130808828714475,
"grad_norm": 221.0,
"learning_rate": 4.380693932932739e-06,
"loss": 33.6092,
"step": 800
},
{
"epoch": 0.4625907904943234,
"grad_norm": 210.0,
"learning_rate": 4.361310331459585e-06,
"loss": 34.7121,
"step": 820
},
{
"epoch": 0.47387349270150203,
"grad_norm": 304.0,
"learning_rate": 4.341926729986432e-06,
"loss": 34.2487,
"step": 840
},
{
"epoch": 0.4851561949086806,
"grad_norm": 235.0,
"learning_rate": 4.322543128513278e-06,
"loss": 33.6595,
"step": 860
},
{
"epoch": 0.49643889711585926,
"grad_norm": 184.0,
"learning_rate": 4.303159527040125e-06,
"loss": 34.8067,
"step": 880
},
{
"epoch": 0.5077215993230378,
"grad_norm": 189.0,
"learning_rate": 4.283775925566971e-06,
"loss": 33.984,
"step": 900
},
{
"epoch": 0.5190043015302165,
"grad_norm": 186.0,
"learning_rate": 4.264392324093816e-06,
"loss": 33.8903,
"step": 920
},
{
"epoch": 0.5302870037373951,
"grad_norm": 205.0,
"learning_rate": 4.2450087226206635e-06,
"loss": 34.0346,
"step": 940
},
{
"epoch": 0.5415697059445738,
"grad_norm": 184.0,
"learning_rate": 4.22562512114751e-06,
"loss": 34.2572,
"step": 960
},
{
"epoch": 0.5528524081517523,
"grad_norm": 221.0,
"learning_rate": 4.206241519674356e-06,
"loss": 34.4286,
"step": 980
},
{
"epoch": 0.5641351103589309,
"grad_norm": 172.0,
"learning_rate": 4.186857918201202e-06,
"loss": 34.3633,
"step": 1000
},
{
"epoch": 0.5754178125661096,
"grad_norm": 202.0,
"learning_rate": 4.1674743167280484e-06,
"loss": 34.7848,
"step": 1020
},
{
"epoch": 0.5867005147732882,
"grad_norm": 224.0,
"learning_rate": 4.148090715254895e-06,
"loss": 34.2222,
"step": 1040
},
{
"epoch": 0.5979832169804669,
"grad_norm": 197.0,
"learning_rate": 4.128707113781741e-06,
"loss": 34.3363,
"step": 1060
},
{
"epoch": 0.6092659191876454,
"grad_norm": 213.0,
"learning_rate": 4.109323512308587e-06,
"loss": 33.8178,
"step": 1080
},
{
"epoch": 0.620548621394824,
"grad_norm": 206.0,
"learning_rate": 4.089939910835433e-06,
"loss": 35.2568,
"step": 1100
},
{
"epoch": 0.6318313236020027,
"grad_norm": 218.0,
"learning_rate": 4.07055630936228e-06,
"loss": 33.7571,
"step": 1120
},
{
"epoch": 0.6431140258091813,
"grad_norm": 207.0,
"learning_rate": 4.051172707889126e-06,
"loss": 34.126,
"step": 1140
},
{
"epoch": 0.65439672801636,
"grad_norm": 162.0,
"learning_rate": 4.031789106415972e-06,
"loss": 34.2224,
"step": 1160
},
{
"epoch": 0.6656794302235386,
"grad_norm": 175.0,
"learning_rate": 4.012405504942818e-06,
"loss": 34.8685,
"step": 1180
},
{
"epoch": 0.6769621324307171,
"grad_norm": 198.0,
"learning_rate": 3.993021903469665e-06,
"loss": 34.5534,
"step": 1200
},
{
"epoch": 0.6882448346378958,
"grad_norm": 200.0,
"learning_rate": 3.973638301996512e-06,
"loss": 35.2566,
"step": 1220
},
{
"epoch": 0.6995275368450744,
"grad_norm": 205.0,
"learning_rate": 3.954254700523358e-06,
"loss": 35.0842,
"step": 1240
},
{
"epoch": 0.710810239052253,
"grad_norm": 181.0,
"learning_rate": 3.934871099050203e-06,
"loss": 34.4665,
"step": 1260
},
{
"epoch": 0.7220929412594317,
"grad_norm": 176.0,
"learning_rate": 3.91548749757705e-06,
"loss": 34.4321,
"step": 1280
},
{
"epoch": 0.7333756434666102,
"grad_norm": 239.0,
"learning_rate": 3.896103896103897e-06,
"loss": 34.4197,
"step": 1300
},
{
"epoch": 0.7446583456737889,
"grad_norm": 215.0,
"learning_rate": 3.876720294630743e-06,
"loss": 34.4867,
"step": 1320
},
{
"epoch": 0.7559410478809675,
"grad_norm": 186.0,
"learning_rate": 3.857336693157589e-06,
"loss": 34.6796,
"step": 1340
},
{
"epoch": 0.7672237500881461,
"grad_norm": 183.0,
"learning_rate": 3.837953091684435e-06,
"loss": 34.2202,
"step": 1360
},
{
"epoch": 0.7785064522953248,
"grad_norm": 192.0,
"learning_rate": 3.818569490211282e-06,
"loss": 34.5157,
"step": 1380
},
{
"epoch": 0.7897891545025033,
"grad_norm": 173.0,
"learning_rate": 3.799185888738128e-06,
"loss": 34.5374,
"step": 1400
},
{
"epoch": 0.801071856709682,
"grad_norm": 183.0,
"learning_rate": 3.779802287264974e-06,
"loss": 34.1974,
"step": 1420
},
{
"epoch": 0.8123545589168606,
"grad_norm": 171.0,
"learning_rate": 3.7604186857918207e-06,
"loss": 34.8075,
"step": 1440
},
{
"epoch": 0.8236372611240392,
"grad_norm": 179.0,
"learning_rate": 3.7410350843186666e-06,
"loss": 34.806,
"step": 1460
},
{
"epoch": 0.8349199633312179,
"grad_norm": 170.0,
"learning_rate": 3.7216514828455132e-06,
"loss": 34.8223,
"step": 1480
},
{
"epoch": 0.8462026655383964,
"grad_norm": 214.0,
"learning_rate": 3.7022678813723595e-06,
"loss": 34.7757,
"step": 1500
},
{
"epoch": 0.857485367745575,
"grad_norm": 182.0,
"learning_rate": 3.6828842798992053e-06,
"loss": 34.1471,
"step": 1520
},
{
"epoch": 0.8687680699527537,
"grad_norm": 183.0,
"learning_rate": 3.663500678426052e-06,
"loss": 34.6106,
"step": 1540
},
{
"epoch": 0.8800507721599323,
"grad_norm": 204.0,
"learning_rate": 3.644117076952898e-06,
"loss": 35.2777,
"step": 1560
},
{
"epoch": 0.891333474367111,
"grad_norm": 200.0,
"learning_rate": 3.624733475479744e-06,
"loss": 35.1206,
"step": 1580
},
{
"epoch": 0.9026161765742895,
"grad_norm": 201.0,
"learning_rate": 3.6053498740065907e-06,
"loss": 34.9908,
"step": 1600
},
{
"epoch": 0.9138988787814681,
"grad_norm": 214.0,
"learning_rate": 3.585966272533437e-06,
"loss": 34.5719,
"step": 1620
},
{
"epoch": 0.9251815809886468,
"grad_norm": 188.0,
"learning_rate": 3.5665826710602836e-06,
"loss": 34.7131,
"step": 1640
},
{
"epoch": 0.9364642831958254,
"grad_norm": 177.0,
"learning_rate": 3.5471990695871294e-06,
"loss": 34.7563,
"step": 1660
},
{
"epoch": 0.9477469854030041,
"grad_norm": 207.0,
"learning_rate": 3.527815468113976e-06,
"loss": 35.2203,
"step": 1680
},
{
"epoch": 0.9590296876101826,
"grad_norm": 186.0,
"learning_rate": 3.5084318666408223e-06,
"loss": 34.9196,
"step": 1700
},
{
"epoch": 0.9703123898173612,
"grad_norm": 214.0,
"learning_rate": 3.489048265167668e-06,
"loss": 34.4439,
"step": 1720
},
{
"epoch": 0.9815950920245399,
"grad_norm": 144.0,
"learning_rate": 3.4696646636945147e-06,
"loss": 34.8649,
"step": 1740
},
{
"epoch": 0.9928777942317185,
"grad_norm": 194.0,
"learning_rate": 3.450281062221361e-06,
"loss": 34.3781,
"step": 1760
},
{
"epoch": 1.0039489457725126,
"grad_norm": 168.0,
"learning_rate": 3.4308974607482072e-06,
"loss": 33.7822,
"step": 1780
},
{
"epoch": 1.0152316479796912,
"grad_norm": 194.0,
"learning_rate": 3.4115138592750535e-06,
"loss": 33.925,
"step": 1800
},
{
"epoch": 1.0265143501868697,
"grad_norm": 188.0,
"learning_rate": 3.3921302578019e-06,
"loss": 33.321,
"step": 1820
},
{
"epoch": 1.0377970523940483,
"grad_norm": 154.0,
"learning_rate": 3.3727466563287464e-06,
"loss": 33.7997,
"step": 1840
},
{
"epoch": 1.049079754601227,
"grad_norm": 153.0,
"learning_rate": 3.353363054855592e-06,
"loss": 33.3133,
"step": 1860
},
{
"epoch": 1.0603624568084056,
"grad_norm": 219.0,
"learning_rate": 3.333979453382439e-06,
"loss": 34.4807,
"step": 1880
},
{
"epoch": 1.0716451590155842,
"grad_norm": 167.0,
"learning_rate": 3.314595851909285e-06,
"loss": 34.1661,
"step": 1900
},
{
"epoch": 1.0829278612227629,
"grad_norm": 212.0,
"learning_rate": 3.2952122504361313e-06,
"loss": 34.0704,
"step": 1920
},
{
"epoch": 1.0942105634299415,
"grad_norm": 164.0,
"learning_rate": 3.2758286489629776e-06,
"loss": 33.3962,
"step": 1940
},
{
"epoch": 1.1054932656371201,
"grad_norm": 180.0,
"learning_rate": 3.2564450474898242e-06,
"loss": 33.2866,
"step": 1960
},
{
"epoch": 1.1167759678442988,
"grad_norm": 151.0,
"learning_rate": 3.23706144601667e-06,
"loss": 33.3367,
"step": 1980
},
{
"epoch": 1.1280586700514774,
"grad_norm": 222.0,
"learning_rate": 3.2176778445435163e-06,
"loss": 33.9858,
"step": 2000
},
{
"epoch": 1.1393413722586558,
"grad_norm": 166.0,
"learning_rate": 3.198294243070363e-06,
"loss": 33.587,
"step": 2020
},
{
"epoch": 1.1506240744658345,
"grad_norm": 176.0,
"learning_rate": 3.1789106415972088e-06,
"loss": 33.2809,
"step": 2040
},
{
"epoch": 1.1619067766730131,
"grad_norm": 205.0,
"learning_rate": 3.159527040124055e-06,
"loss": 33.6528,
"step": 2060
},
{
"epoch": 1.1731894788801918,
"grad_norm": 190.0,
"learning_rate": 3.1401434386509017e-06,
"loss": 33.4909,
"step": 2080
},
{
"epoch": 1.1844721810873704,
"grad_norm": 175.0,
"learning_rate": 3.120759837177748e-06,
"loss": 33.5192,
"step": 2100
},
{
"epoch": 1.195754883294549,
"grad_norm": 187.0,
"learning_rate": 3.101376235704594e-06,
"loss": 33.9459,
"step": 2120
},
{
"epoch": 1.2070375855017277,
"grad_norm": 159.0,
"learning_rate": 3.0819926342314404e-06,
"loss": 34.1256,
"step": 2140
},
{
"epoch": 1.2183202877089063,
"grad_norm": 174.0,
"learning_rate": 3.062609032758287e-06,
"loss": 33.5218,
"step": 2160
},
{
"epoch": 1.229602989916085,
"grad_norm": 189.0,
"learning_rate": 3.043225431285133e-06,
"loss": 34.4926,
"step": 2180
},
{
"epoch": 1.2408856921232636,
"grad_norm": 169.0,
"learning_rate": 3.023841829811979e-06,
"loss": 34.1501,
"step": 2200
},
{
"epoch": 1.2521683943304422,
"grad_norm": 157.0,
"learning_rate": 3.0044582283388258e-06,
"loss": 32.9258,
"step": 2220
},
{
"epoch": 1.2634510965376209,
"grad_norm": 162.0,
"learning_rate": 2.9850746268656716e-06,
"loss": 33.7786,
"step": 2240
},
{
"epoch": 1.2747337987447993,
"grad_norm": 189.0,
"learning_rate": 2.9656910253925182e-06,
"loss": 33.3599,
"step": 2260
},
{
"epoch": 1.286016500951978,
"grad_norm": 168.0,
"learning_rate": 2.9463074239193645e-06,
"loss": 34.213,
"step": 2280
},
{
"epoch": 1.2972992031591566,
"grad_norm": 167.0,
"learning_rate": 2.926923822446211e-06,
"loss": 33.6998,
"step": 2300
},
{
"epoch": 1.3085819053663352,
"grad_norm": 166.0,
"learning_rate": 2.907540220973057e-06,
"loss": 33.907,
"step": 2320
},
{
"epoch": 1.3198646075735139,
"grad_norm": 226.0,
"learning_rate": 2.888156619499903e-06,
"loss": 34.1545,
"step": 2340
},
{
"epoch": 1.3311473097806925,
"grad_norm": 157.0,
"learning_rate": 2.86877301802675e-06,
"loss": 33.681,
"step": 2360
},
{
"epoch": 1.3424300119878712,
"grad_norm": 222.0,
"learning_rate": 2.8493894165535957e-06,
"loss": 33.7286,
"step": 2380
},
{
"epoch": 1.3537127141950498,
"grad_norm": 199.0,
"learning_rate": 2.8300058150804423e-06,
"loss": 33.9236,
"step": 2400
},
{
"epoch": 1.3649954164022282,
"grad_norm": 171.0,
"learning_rate": 2.8106222136072886e-06,
"loss": 33.7998,
"step": 2420
},
{
"epoch": 1.3762781186094069,
"grad_norm": 182.0,
"learning_rate": 2.7912386121341344e-06,
"loss": 33.6602,
"step": 2440
},
{
"epoch": 1.3875608208165855,
"grad_norm": 231.0,
"learning_rate": 2.771855010660981e-06,
"loss": 34.5877,
"step": 2460
},
{
"epoch": 1.3988435230237641,
"grad_norm": 174.0,
"learning_rate": 2.7524714091878273e-06,
"loss": 33.8736,
"step": 2480
},
{
"epoch": 1.4101262252309428,
"grad_norm": 189.0,
"learning_rate": 2.733087807714674e-06,
"loss": 33.3423,
"step": 2500
},
{
"epoch": 1.4214089274381214,
"grad_norm": 168.0,
"learning_rate": 2.7137042062415198e-06,
"loss": 34.1842,
"step": 2520
},
{
"epoch": 1.4326916296453,
"grad_norm": 204.0,
"learning_rate": 2.694320604768366e-06,
"loss": 33.48,
"step": 2540
},
{
"epoch": 1.4439743318524787,
"grad_norm": 190.0,
"learning_rate": 2.6749370032952127e-06,
"loss": 34.2732,
"step": 2560
},
{
"epoch": 1.4552570340596573,
"grad_norm": 166.0,
"learning_rate": 2.6555534018220585e-06,
"loss": 33.3832,
"step": 2580
},
{
"epoch": 1.466539736266836,
"grad_norm": 181.0,
"learning_rate": 2.636169800348905e-06,
"loss": 34.4397,
"step": 2600
},
{
"epoch": 1.4778224384740146,
"grad_norm": 172.0,
"learning_rate": 2.6167861988757514e-06,
"loss": 34.0473,
"step": 2620
},
{
"epoch": 1.4891051406811933,
"grad_norm": 182.0,
"learning_rate": 2.597402597402597e-06,
"loss": 34.2556,
"step": 2640
},
{
"epoch": 1.500387842888372,
"grad_norm": 201.0,
"learning_rate": 2.578018995929444e-06,
"loss": 34.0902,
"step": 2660
},
{
"epoch": 1.5116705450955505,
"grad_norm": 300.0,
"learning_rate": 2.55863539445629e-06,
"loss": 33.6671,
"step": 2680
},
{
"epoch": 1.522953247302729,
"grad_norm": 146.0,
"learning_rate": 2.5392517929831368e-06,
"loss": 34.1371,
"step": 2700
},
{
"epoch": 1.5342359495099076,
"grad_norm": 183.0,
"learning_rate": 2.5198681915099826e-06,
"loss": 34.0674,
"step": 2720
},
{
"epoch": 1.5455186517170862,
"grad_norm": 181.0,
"learning_rate": 2.5004845900368292e-06,
"loss": 34.5481,
"step": 2740
},
{
"epoch": 1.5568013539242649,
"grad_norm": 181.0,
"learning_rate": 2.481100988563675e-06,
"loss": 33.6689,
"step": 2760
},
{
"epoch": 1.5680840561314435,
"grad_norm": 172.0,
"learning_rate": 2.4617173870905217e-06,
"loss": 34.5104,
"step": 2780
},
{
"epoch": 1.579366758338622,
"grad_norm": 164.0,
"learning_rate": 2.442333785617368e-06,
"loss": 34.001,
"step": 2800
},
{
"epoch": 1.5906494605458006,
"grad_norm": 195.0,
"learning_rate": 2.422950184144214e-06,
"loss": 34.2375,
"step": 2820
},
{
"epoch": 1.6019321627529792,
"grad_norm": 162.0,
"learning_rate": 2.4035665826710604e-06,
"loss": 33.939,
"step": 2840
},
{
"epoch": 1.6132148649601579,
"grad_norm": 190.0,
"learning_rate": 2.3841829811979067e-06,
"loss": 33.6411,
"step": 2860
},
{
"epoch": 1.6244975671673365,
"grad_norm": 212.0,
"learning_rate": 2.3647993797247533e-06,
"loss": 33.7762,
"step": 2880
},
{
"epoch": 1.6357802693745152,
"grad_norm": 167.0,
"learning_rate": 2.345415778251599e-06,
"loss": 33.9127,
"step": 2900
},
{
"epoch": 1.6470629715816938,
"grad_norm": 248.0,
"learning_rate": 2.3260321767784454e-06,
"loss": 34.6926,
"step": 2920
},
{
"epoch": 1.6583456737888724,
"grad_norm": 155.0,
"learning_rate": 2.306648575305292e-06,
"loss": 34.5797,
"step": 2940
},
{
"epoch": 1.669628375996051,
"grad_norm": 189.0,
"learning_rate": 2.2872649738321383e-06,
"loss": 34.1331,
"step": 2960
},
{
"epoch": 1.6809110782032297,
"grad_norm": 183.0,
"learning_rate": 2.2678813723589845e-06,
"loss": 34.5708,
"step": 2980
},
{
"epoch": 1.6921937804104084,
"grad_norm": 209.0,
"learning_rate": 2.2484977708858308e-06,
"loss": 34.1141,
"step": 3000
},
{
"epoch": 1.703476482617587,
"grad_norm": 215.0,
"learning_rate": 2.229114169412677e-06,
"loss": 34.3492,
"step": 3020
},
{
"epoch": 1.7147591848247656,
"grad_norm": 185.0,
"learning_rate": 2.2097305679395232e-06,
"loss": 34.3873,
"step": 3040
},
{
"epoch": 1.7260418870319443,
"grad_norm": 195.0,
"learning_rate": 2.1903469664663695e-06,
"loss": 33.9293,
"step": 3060
},
{
"epoch": 1.737324589239123,
"grad_norm": 173.0,
"learning_rate": 2.170963364993216e-06,
"loss": 34.1302,
"step": 3080
},
{
"epoch": 1.7486072914463013,
"grad_norm": 162.0,
"learning_rate": 2.1515797635200624e-06,
"loss": 33.6191,
"step": 3100
},
{
"epoch": 1.75988999365348,
"grad_norm": 155.0,
"learning_rate": 2.132196162046908e-06,
"loss": 34.4905,
"step": 3120
},
{
"epoch": 1.7711726958606586,
"grad_norm": 184.0,
"learning_rate": 2.112812560573755e-06,
"loss": 34.0467,
"step": 3140
},
{
"epoch": 1.7824553980678373,
"grad_norm": 194.0,
"learning_rate": 2.093428959100601e-06,
"loss": 34.0904,
"step": 3160
},
{
"epoch": 1.793738100275016,
"grad_norm": 153.0,
"learning_rate": 2.0740453576274473e-06,
"loss": 34.1613,
"step": 3180
},
{
"epoch": 1.8050208024821943,
"grad_norm": 177.0,
"learning_rate": 2.0546617561542936e-06,
"loss": 34.3794,
"step": 3200
},
{
"epoch": 1.816303504689373,
"grad_norm": 164.0,
"learning_rate": 2.03527815468114e-06,
"loss": 34.107,
"step": 3220
},
{
"epoch": 1.8275862068965516,
"grad_norm": 181.0,
"learning_rate": 2.015894553207986e-06,
"loss": 33.9846,
"step": 3240
},
{
"epoch": 1.8388689091037302,
"grad_norm": 205.0,
"learning_rate": 1.9965109517348323e-06,
"loss": 34.2268,
"step": 3260
},
{
"epoch": 1.8501516113109089,
"grad_norm": 221.0,
"learning_rate": 1.977127350261679e-06,
"loss": 34.8003,
"step": 3280
},
{
"epoch": 1.8614343135180875,
"grad_norm": 185.0,
"learning_rate": 1.957743748788525e-06,
"loss": 34.1929,
"step": 3300
},
{
"epoch": 1.8727170157252662,
"grad_norm": 178.0,
"learning_rate": 1.9383601473153714e-06,
"loss": 34.8628,
"step": 3320
},
{
"epoch": 1.8839997179324448,
"grad_norm": 171.0,
"learning_rate": 1.9189765458422177e-06,
"loss": 33.7048,
"step": 3340
},
{
"epoch": 1.8952824201396234,
"grad_norm": 274.0,
"learning_rate": 1.899592944369064e-06,
"loss": 33.5702,
"step": 3360
},
{
"epoch": 1.906565122346802,
"grad_norm": 173.0,
"learning_rate": 1.8802093428959104e-06,
"loss": 34.5673,
"step": 3380
},
{
"epoch": 1.9178478245539807,
"grad_norm": 191.0,
"learning_rate": 1.8608257414227566e-06,
"loss": 34.9713,
"step": 3400
},
{
"epoch": 1.9291305267611594,
"grad_norm": 189.0,
"learning_rate": 1.8414421399496026e-06,
"loss": 33.9675,
"step": 3420
},
{
"epoch": 1.940413228968338,
"grad_norm": 158.0,
"learning_rate": 1.822058538476449e-06,
"loss": 34.4765,
"step": 3440
},
{
"epoch": 1.9516959311755167,
"grad_norm": 196.0,
"learning_rate": 1.8026749370032953e-06,
"loss": 34.3523,
"step": 3460
},
{
"epoch": 1.9629786333826953,
"grad_norm": 188.0,
"learning_rate": 1.7832913355301418e-06,
"loss": 34.8046,
"step": 3480
},
{
"epoch": 1.974261335589874,
"grad_norm": 200.0,
"learning_rate": 1.763907734056988e-06,
"loss": 35.0217,
"step": 3500
},
{
"epoch": 1.9855440377970524,
"grad_norm": 215.0,
"learning_rate": 1.744524132583834e-06,
"loss": 34.9497,
"step": 3520
},
{
"epoch": 1.996826740004231,
"grad_norm": 177.0,
"learning_rate": 1.7251405311106805e-06,
"loss": 34.5596,
"step": 3540
},
{
"epoch": 2.007897891545025,
"grad_norm": 163.0,
"learning_rate": 1.7057569296375267e-06,
"loss": 33.9491,
"step": 3560
},
{
"epoch": 2.019180593752204,
"grad_norm": 166.0,
"learning_rate": 1.6863733281643732e-06,
"loss": 33.7916,
"step": 3580
},
{
"epoch": 2.0304632959593825,
"grad_norm": 232.0,
"learning_rate": 1.6669897266912194e-06,
"loss": 33.9634,
"step": 3600
},
{
"epoch": 2.041745998166561,
"grad_norm": 211.0,
"learning_rate": 1.6476061252180657e-06,
"loss": 34.2085,
"step": 3620
},
{
"epoch": 2.0530287003737393,
"grad_norm": 237.0,
"learning_rate": 1.6282225237449121e-06,
"loss": 34.0196,
"step": 3640
},
{
"epoch": 2.064311402580918,
"grad_norm": 192.0,
"learning_rate": 1.6088389222717581e-06,
"loss": 34.5528,
"step": 3660
},
{
"epoch": 2.0755941047880966,
"grad_norm": 173.0,
"learning_rate": 1.5894553207986044e-06,
"loss": 33.1571,
"step": 3680
},
{
"epoch": 2.0868768069952752,
"grad_norm": 196.0,
"learning_rate": 1.5700717193254508e-06,
"loss": 33.8429,
"step": 3700
},
{
"epoch": 2.098159509202454,
"grad_norm": 203.0,
"learning_rate": 1.550688117852297e-06,
"loss": 33.1751,
"step": 3720
},
{
"epoch": 2.1094422114096325,
"grad_norm": 170.0,
"learning_rate": 1.5313045163791435e-06,
"loss": 34.0564,
"step": 3740
},
{
"epoch": 2.120724913616811,
"grad_norm": 194.0,
"learning_rate": 1.5119209149059895e-06,
"loss": 33.8536,
"step": 3760
},
{
"epoch": 2.13200761582399,
"grad_norm": 179.0,
"learning_rate": 1.4925373134328358e-06,
"loss": 33.399,
"step": 3780
},
{
"epoch": 2.1432903180311684,
"grad_norm": 206.0,
"learning_rate": 1.4731537119596822e-06,
"loss": 34.201,
"step": 3800
},
{
"epoch": 2.154573020238347,
"grad_norm": 171.0,
"learning_rate": 1.4537701104865285e-06,
"loss": 33.347,
"step": 3820
},
{
"epoch": 2.1658557224455257,
"grad_norm": 195.0,
"learning_rate": 1.434386509013375e-06,
"loss": 34.2486,
"step": 3840
},
{
"epoch": 2.1771384246527044,
"grad_norm": 187.0,
"learning_rate": 1.4150029075402212e-06,
"loss": 33.5421,
"step": 3860
},
{
"epoch": 2.188421126859883,
"grad_norm": 184.0,
"learning_rate": 1.3956193060670672e-06,
"loss": 33.8024,
"step": 3880
},
{
"epoch": 2.1997038290670616,
"grad_norm": 188.0,
"learning_rate": 1.3762357045939136e-06,
"loss": 33.4759,
"step": 3900
},
{
"epoch": 2.2109865312742403,
"grad_norm": 185.0,
"learning_rate": 1.3568521031207599e-06,
"loss": 33.8726,
"step": 3920
},
{
"epoch": 2.222269233481419,
"grad_norm": 225.0,
"learning_rate": 1.3374685016476063e-06,
"loss": 34.1284,
"step": 3940
},
{
"epoch": 2.2335519356885976,
"grad_norm": 156.0,
"learning_rate": 1.3180849001744526e-06,
"loss": 33.9871,
"step": 3960
},
{
"epoch": 2.244834637895776,
"grad_norm": 154.0,
"learning_rate": 1.2987012987012986e-06,
"loss": 33.6454,
"step": 3980
},
{
"epoch": 2.256117340102955,
"grad_norm": 173.0,
"learning_rate": 1.279317697228145e-06,
"loss": 33.5879,
"step": 4000
},
{
"epoch": 2.2674000423101335,
"grad_norm": 217.0,
"learning_rate": 1.2599340957549913e-06,
"loss": 33.7071,
"step": 4020
},
{
"epoch": 2.2786827445173117,
"grad_norm": 170.0,
"learning_rate": 1.2405504942818375e-06,
"loss": 34.0824,
"step": 4040
},
{
"epoch": 2.2899654467244908,
"grad_norm": 174.0,
"learning_rate": 1.221166892808684e-06,
"loss": 33.9394,
"step": 4060
},
{
"epoch": 2.301248148931669,
"grad_norm": 189.0,
"learning_rate": 1.2017832913355302e-06,
"loss": 33.8776,
"step": 4080
},
{
"epoch": 2.3125308511388476,
"grad_norm": 203.0,
"learning_rate": 1.1823996898623767e-06,
"loss": 34.6775,
"step": 4100
},
{
"epoch": 2.3238135533460262,
"grad_norm": 215.0,
"learning_rate": 1.1630160883892227e-06,
"loss": 34.0411,
"step": 4120
},
{
"epoch": 2.335096255553205,
"grad_norm": 224.0,
"learning_rate": 1.1436324869160691e-06,
"loss": 33.1014,
"step": 4140
},
{
"epoch": 2.3463789577603835,
"grad_norm": 194.0,
"learning_rate": 1.1242488854429154e-06,
"loss": 33.892,
"step": 4160
},
{
"epoch": 2.357661659967562,
"grad_norm": 168.0,
"learning_rate": 1.1048652839697616e-06,
"loss": 34.1314,
"step": 4180
},
{
"epoch": 2.368944362174741,
"grad_norm": 206.0,
"learning_rate": 1.085481682496608e-06,
"loss": 34.1024,
"step": 4200
},
{
"epoch": 2.3802270643819194,
"grad_norm": 160.0,
"learning_rate": 1.066098081023454e-06,
"loss": 34.1692,
"step": 4220
},
{
"epoch": 2.391509766589098,
"grad_norm": 185.0,
"learning_rate": 1.0467144795503006e-06,
"loss": 33.2612,
"step": 4240
},
{
"epoch": 2.4027924687962767,
"grad_norm": 206.0,
"learning_rate": 1.0273308780771468e-06,
"loss": 33.7068,
"step": 4260
},
{
"epoch": 2.4140751710034554,
"grad_norm": 223.0,
"learning_rate": 1.007947276603993e-06,
"loss": 33.9393,
"step": 4280
},
{
"epoch": 2.425357873210634,
"grad_norm": 171.0,
"learning_rate": 9.885636751308395e-07,
"loss": 33.9153,
"step": 4300
},
{
"epoch": 2.4366405754178126,
"grad_norm": 179.0,
"learning_rate": 9.691800736576857e-07,
"loss": 34.3688,
"step": 4320
},
{
"epoch": 2.4479232776249913,
"grad_norm": 185.0,
"learning_rate": 9.49796472184532e-07,
"loss": 34.2571,
"step": 4340
},
{
"epoch": 2.45920597983217,
"grad_norm": 196.0,
"learning_rate": 9.304128707113783e-07,
"loss": 34.3851,
"step": 4360
},
{
"epoch": 2.4704886820393486,
"grad_norm": 178.0,
"learning_rate": 9.110292692382245e-07,
"loss": 33.9845,
"step": 4380
},
{
"epoch": 2.481771384246527,
"grad_norm": 174.0,
"learning_rate": 8.916456677650709e-07,
"loss": 33.7759,
"step": 4400
},
{
"epoch": 2.493054086453706,
"grad_norm": 196.0,
"learning_rate": 8.72262066291917e-07,
"loss": 34.3738,
"step": 4420
},
{
"epoch": 2.5043367886608845,
"grad_norm": 195.0,
"learning_rate": 8.528784648187634e-07,
"loss": 33.7357,
"step": 4440
},
{
"epoch": 2.5156194908680627,
"grad_norm": 191.0,
"learning_rate": 8.334948633456097e-07,
"loss": 33.7378,
"step": 4460
},
{
"epoch": 2.5269021930752418,
"grad_norm": 183.0,
"learning_rate": 8.141112618724561e-07,
"loss": 33.7745,
"step": 4480
},
{
"epoch": 2.53818489528242,
"grad_norm": 232.0,
"learning_rate": 7.947276603993022e-07,
"loss": 34.2662,
"step": 4500
},
{
"epoch": 2.5494675974895986,
"grad_norm": 221.0,
"learning_rate": 7.753440589261485e-07,
"loss": 34.5754,
"step": 4520
},
{
"epoch": 2.5607502996967773,
"grad_norm": 192.0,
"learning_rate": 7.559604574529948e-07,
"loss": 33.759,
"step": 4540
},
{
"epoch": 2.572033001903956,
"grad_norm": 212.0,
"learning_rate": 7.365768559798411e-07,
"loss": 34.1167,
"step": 4560
},
{
"epoch": 2.5833157041111345,
"grad_norm": 161.0,
"learning_rate": 7.171932545066875e-07,
"loss": 34.7971,
"step": 4580
},
{
"epoch": 2.594598406318313,
"grad_norm": 160.0,
"learning_rate": 6.978096530335336e-07,
"loss": 34.9587,
"step": 4600
},
{
"epoch": 2.605881108525492,
"grad_norm": 236.0,
"learning_rate": 6.784260515603799e-07,
"loss": 34.3313,
"step": 4620
},
{
"epoch": 2.6171638107326705,
"grad_norm": 182.0,
"learning_rate": 6.590424500872263e-07,
"loss": 33.5634,
"step": 4640
},
{
"epoch": 2.628446512939849,
"grad_norm": 168.0,
"learning_rate": 6.396588486140725e-07,
"loss": 34.4282,
"step": 4660
},
{
"epoch": 2.6397292151470277,
"grad_norm": 184.0,
"learning_rate": 6.202752471409188e-07,
"loss": 34.1346,
"step": 4680
},
{
"epoch": 2.6510119173542064,
"grad_norm": 186.0,
"learning_rate": 6.008916456677651e-07,
"loss": 34.5073,
"step": 4700
},
{
"epoch": 2.662294619561385,
"grad_norm": 178.0,
"learning_rate": 5.815080441946113e-07,
"loss": 33.6183,
"step": 4720
},
{
"epoch": 2.6735773217685637,
"grad_norm": 167.0,
"learning_rate": 5.621244427214577e-07,
"loss": 33.6038,
"step": 4740
},
{
"epoch": 2.6848600239757423,
"grad_norm": 179.0,
"learning_rate": 5.42740841248304e-07,
"loss": 34.0521,
"step": 4760
},
{
"epoch": 2.696142726182921,
"grad_norm": 174.0,
"learning_rate": 5.233572397751503e-07,
"loss": 33.989,
"step": 4780
},
{
"epoch": 2.7074254283900996,
"grad_norm": 213.0,
"learning_rate": 5.039736383019965e-07,
"loss": 34.2236,
"step": 4800
},
{
"epoch": 2.7187081305972782,
"grad_norm": 164.0,
"learning_rate": 4.845900368288429e-07,
"loss": 33.7335,
"step": 4820
},
{
"epoch": 2.7299908328044564,
"grad_norm": 189.0,
"learning_rate": 4.6520643535568915e-07,
"loss": 34.3171,
"step": 4840
},
{
"epoch": 2.7412735350116355,
"grad_norm": 164.0,
"learning_rate": 4.4582283388253544e-07,
"loss": 34.4135,
"step": 4860
},
{
"epoch": 2.7525562372188137,
"grad_norm": 157.0,
"learning_rate": 4.264392324093817e-07,
"loss": 33.7103,
"step": 4880
},
{
"epoch": 2.763838939425993,
"grad_norm": 173.0,
"learning_rate": 4.0705563093622803e-07,
"loss": 34.1047,
"step": 4900
},
{
"epoch": 2.775121641633171,
"grad_norm": 189.0,
"learning_rate": 3.8767202946307427e-07,
"loss": 34.0697,
"step": 4920
},
{
"epoch": 2.7864043438403496,
"grad_norm": 168.0,
"learning_rate": 3.6828842798992056e-07,
"loss": 33.9304,
"step": 4940
},
{
"epoch": 2.7976870460475283,
"grad_norm": 206.0,
"learning_rate": 3.489048265167668e-07,
"loss": 34.381,
"step": 4960
},
{
"epoch": 2.808969748254707,
"grad_norm": 189.0,
"learning_rate": 3.2952122504361314e-07,
"loss": 33.7121,
"step": 4980
},
{
"epoch": 2.8202524504618856,
"grad_norm": 202.0,
"learning_rate": 3.101376235704594e-07,
"loss": 33.5941,
"step": 5000
}
],
"logging_steps": 20,
"max_steps": 5319,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 0.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}