| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.993127147766323, | |
| "eval_steps": 500, | |
| "global_step": 290, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006872852233676976, | |
| "grad_norm": 11.911341495631659, | |
| "learning_rate": 3.4482758620689656e-07, | |
| "loss": 0.3151, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013745704467353952, | |
| "grad_norm": 14.023610009584813, | |
| "learning_rate": 6.896551724137931e-07, | |
| "loss": 0.3217, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020618556701030927, | |
| "grad_norm": 13.706714014844485, | |
| "learning_rate": 1.0344827586206898e-06, | |
| "loss": 0.3966, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.027491408934707903, | |
| "grad_norm": 10.766913532858457, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 0.3373, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03436426116838488, | |
| "grad_norm": 9.494862403997999, | |
| "learning_rate": 1.724137931034483e-06, | |
| "loss": 0.2601, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.041237113402061855, | |
| "grad_norm": 7.9796005781017465, | |
| "learning_rate": 2.0689655172413796e-06, | |
| "loss": 0.2296, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.048109965635738834, | |
| "grad_norm": 5.031277691829482, | |
| "learning_rate": 2.4137931034482762e-06, | |
| "loss": 0.1925, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.054982817869415807, | |
| "grad_norm": 5.034929869179998, | |
| "learning_rate": 2.7586206896551725e-06, | |
| "loss": 0.2165, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.061855670103092786, | |
| "grad_norm": 3.8000906869499067, | |
| "learning_rate": 3.103448275862069e-06, | |
| "loss": 0.1462, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06872852233676977, | |
| "grad_norm": 2.9124924064644375, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.1562, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07560137457044673, | |
| "grad_norm": 2.3691642241007065, | |
| "learning_rate": 3.793103448275862e-06, | |
| "loss": 0.1605, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08247422680412371, | |
| "grad_norm": 4.101340880026747, | |
| "learning_rate": 4.137931034482759e-06, | |
| "loss": 0.151, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08934707903780069, | |
| "grad_norm": 3.483298221413338, | |
| "learning_rate": 4.482758620689656e-06, | |
| "loss": 0.1504, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09621993127147767, | |
| "grad_norm": 2.1924981464095405, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 0.1205, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10309278350515463, | |
| "grad_norm": 2.99252439296641, | |
| "learning_rate": 5.172413793103449e-06, | |
| "loss": 0.1451, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10996563573883161, | |
| "grad_norm": 3.0085251257849532, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 0.1647, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11683848797250859, | |
| "grad_norm": 2.488559923295055, | |
| "learning_rate": 5.862068965517242e-06, | |
| "loss": 0.1244, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12371134020618557, | |
| "grad_norm": 2.4443090104864575, | |
| "learning_rate": 6.206896551724138e-06, | |
| "loss": 0.1172, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13058419243986255, | |
| "grad_norm": 1.731480845454143, | |
| "learning_rate": 6.551724137931035e-06, | |
| "loss": 0.0954, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13745704467353953, | |
| "grad_norm": 3.2942531263560393, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 0.1398, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14432989690721648, | |
| "grad_norm": 3.7928242389522655, | |
| "learning_rate": 7.241379310344828e-06, | |
| "loss": 0.1051, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15120274914089346, | |
| "grad_norm": 1.68475639236392, | |
| "learning_rate": 7.586206896551724e-06, | |
| "loss": 0.0874, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15807560137457044, | |
| "grad_norm": 2.040905258991334, | |
| "learning_rate": 7.93103448275862e-06, | |
| "loss": 0.1161, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16494845360824742, | |
| "grad_norm": 2.7528599761784514, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 0.1291, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1718213058419244, | |
| "grad_norm": 2.35204956485846, | |
| "learning_rate": 8.620689655172414e-06, | |
| "loss": 0.1328, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.17869415807560138, | |
| "grad_norm": 2.468292928791487, | |
| "learning_rate": 8.965517241379312e-06, | |
| "loss": 0.1274, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18556701030927836, | |
| "grad_norm": 1.6463074957872081, | |
| "learning_rate": 9.310344827586207e-06, | |
| "loss": 0.1199, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19243986254295534, | |
| "grad_norm": 2.5185479302329457, | |
| "learning_rate": 9.655172413793105e-06, | |
| "loss": 0.1051, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19931271477663232, | |
| "grad_norm": 2.390346345099224, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1351, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20618556701030927, | |
| "grad_norm": 3.1372283321356296, | |
| "learning_rate": 9.999637795788383e-06, | |
| "loss": 0.1459, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21305841924398625, | |
| "grad_norm": 2.894796251328901, | |
| "learning_rate": 9.99855123563029e-06, | |
| "loss": 0.1374, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21993127147766323, | |
| "grad_norm": 1.8614532628537759, | |
| "learning_rate": 9.996740476948386e-06, | |
| "loss": 0.141, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2268041237113402, | |
| "grad_norm": 2.771156267025066, | |
| "learning_rate": 9.994205782088438e-06, | |
| "loss": 0.1122, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.23367697594501718, | |
| "grad_norm": 2.08411373687462, | |
| "learning_rate": 9.990947518281312e-06, | |
| "loss": 0.1243, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24054982817869416, | |
| "grad_norm": 1.549196081138641, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 0.0745, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24742268041237114, | |
| "grad_norm": 1.4403136866707111, | |
| "learning_rate": 9.982262276840002e-06, | |
| "loss": 0.0964, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2542955326460481, | |
| "grad_norm": 4.392782871770487, | |
| "learning_rate": 9.976836557538234e-06, | |
| "loss": 0.1273, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2611683848797251, | |
| "grad_norm": 3.3046759882659584, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.113, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26804123711340205, | |
| "grad_norm": 1.5315885741338178, | |
| "learning_rate": 9.963822852095344e-06, | |
| "loss": 0.0814, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.27491408934707906, | |
| "grad_norm": 1.7435499109150194, | |
| "learning_rate": 9.95623675140179e-06, | |
| "loss": 0.0907, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.281786941580756, | |
| "grad_norm": 1.3136132377430776, | |
| "learning_rate": 9.947932582778188e-06, | |
| "loss": 0.1187, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.28865979381443296, | |
| "grad_norm": 1.9704073469885035, | |
| "learning_rate": 9.938911549346473e-06, | |
| "loss": 0.0939, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.29553264604810997, | |
| "grad_norm": 1.4142468112118425, | |
| "learning_rate": 9.929174958089167e-06, | |
| "loss": 0.093, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3024054982817869, | |
| "grad_norm": 1.1929321559845192, | |
| "learning_rate": 9.918724219660013e-06, | |
| "loss": 0.09, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30927835051546393, | |
| "grad_norm": 2.638764001995451, | |
| "learning_rate": 9.907560848179607e-06, | |
| "loss": 0.0963, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3161512027491409, | |
| "grad_norm": 1.3856129718845853, | |
| "learning_rate": 9.895686461016007e-06, | |
| "loss": 0.1222, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3230240549828179, | |
| "grad_norm": 1.4829448988797076, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.0971, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32989690721649484, | |
| "grad_norm": 2.0191304714032134, | |
| "learning_rate": 9.869811623928001e-06, | |
| "loss": 0.1205, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.33676975945017185, | |
| "grad_norm": 2.0075605455877406, | |
| "learning_rate": 9.855814922793583e-06, | |
| "loss": 0.0961, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3436426116838488, | |
| "grad_norm": 2.04263005705328, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 0.1181, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35051546391752575, | |
| "grad_norm": 1.2609929928110724, | |
| "learning_rate": 9.82571309437831e-06, | |
| "loss": 0.0954, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.35738831615120276, | |
| "grad_norm": 1.2891980244125067, | |
| "learning_rate": 9.809612328301071e-06, | |
| "loss": 0.0837, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3642611683848797, | |
| "grad_norm": 0.9021044305254864, | |
| "learning_rate": 9.792814737487207e-06, | |
| "loss": 0.0924, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3711340206185567, | |
| "grad_norm": 1.008447342585738, | |
| "learning_rate": 9.775322755599979e-06, | |
| "loss": 0.0875, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37800687285223367, | |
| "grad_norm": 0.9868645149108727, | |
| "learning_rate": 9.757138916907184e-06, | |
| "loss": 0.0977, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3848797250859107, | |
| "grad_norm": 1.1701737786184419, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.1439, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3917525773195876, | |
| "grad_norm": 1.1471882173922878, | |
| "learning_rate": 9.718706306981332e-06, | |
| "loss": 0.1411, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39862542955326463, | |
| "grad_norm": 1.6608908423233688, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.1016, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4054982817869416, | |
| "grad_norm": 1.0001617077968141, | |
| "learning_rate": 9.677539179628005e-06, | |
| "loss": 0.1022, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.41237113402061853, | |
| "grad_norm": 0.83900735235759, | |
| "learning_rate": 9.655937565570124e-06, | |
| "loss": 0.0698, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.41924398625429554, | |
| "grad_norm": 1.23497707759338, | |
| "learning_rate": 9.63366139143413e-06, | |
| "loss": 0.1148, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4261168384879725, | |
| "grad_norm": 1.4153421835575986, | |
| "learning_rate": 9.610713884629667e-06, | |
| "loss": 0.1184, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4329896907216495, | |
| "grad_norm": 1.3370125599465301, | |
| "learning_rate": 9.587098369830171e-06, | |
| "loss": 0.0875, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43986254295532645, | |
| "grad_norm": 1.2886674998782797, | |
| "learning_rate": 9.562818268491216e-06, | |
| "loss": 0.1026, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.44673539518900346, | |
| "grad_norm": 1.6270581846121255, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.124, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4536082474226804, | |
| "grad_norm": 0.9005472951916546, | |
| "learning_rate": 9.512278472939627e-06, | |
| "loss": 0.0871, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.46048109965635736, | |
| "grad_norm": 2.0638277439439574, | |
| "learning_rate": 9.486026101017711e-06, | |
| "loss": 0.0869, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.46735395189003437, | |
| "grad_norm": 2.0865052014487757, | |
| "learning_rate": 9.459123786076911e-06, | |
| "loss": 0.1168, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4742268041237113, | |
| "grad_norm": 1.2663145946537608, | |
| "learning_rate": 9.431575425769938e-06, | |
| "loss": 0.0875, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.48109965635738833, | |
| "grad_norm": 1.130315902352012, | |
| "learning_rate": 9.40338501134964e-06, | |
| "loss": 0.0713, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4879725085910653, | |
| "grad_norm": 0.8888648768029372, | |
| "learning_rate": 9.374556627090749e-06, | |
| "loss": 0.0715, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4948453608247423, | |
| "grad_norm": 1.3969842397558478, | |
| "learning_rate": 9.345094449698143e-06, | |
| "loss": 0.1086, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5017182130584192, | |
| "grad_norm": 0.7491368899044253, | |
| "learning_rate": 9.315002747701716e-06, | |
| "loss": 0.0714, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5085910652920962, | |
| "grad_norm": 1.2150355310916483, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.0824, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5154639175257731, | |
| "grad_norm": 0.8007945996706918, | |
| "learning_rate": 9.252948299418255e-06, | |
| "loss": 0.0972, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5223367697594502, | |
| "grad_norm": 0.8194050750954162, | |
| "learning_rate": 9.220994543684225e-06, | |
| "loss": 0.0748, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5292096219931272, | |
| "grad_norm": 0.7739683444236579, | |
| "learning_rate": 9.188429243149824e-06, | |
| "loss": 0.0672, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5360824742268041, | |
| "grad_norm": 0.9754020046294775, | |
| "learning_rate": 9.155257115930651e-06, | |
| "loss": 0.08, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5429553264604811, | |
| "grad_norm": 0.9561399477952047, | |
| "learning_rate": 9.121482968060384e-06, | |
| "loss": 0.0923, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5498281786941581, | |
| "grad_norm": 1.5389553814843786, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 0.1099, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5567010309278351, | |
| "grad_norm": 1.0032908939840397, | |
| "learning_rate": 9.052148269901145e-06, | |
| "loss": 0.0832, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.563573883161512, | |
| "grad_norm": 1.8074586339758751, | |
| "learning_rate": 9.01659776494005e-06, | |
| "loss": 0.081, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.570446735395189, | |
| "grad_norm": 0.6964132515418937, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.0775, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5773195876288659, | |
| "grad_norm": 1.2800895685410214, | |
| "learning_rate": 8.943756195593916e-06, | |
| "loss": 0.0892, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.584192439862543, | |
| "grad_norm": 1.5580218470383653, | |
| "learning_rate": 8.90647568461816e-06, | |
| "loss": 0.084, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5910652920962199, | |
| "grad_norm": 1.1178731871347416, | |
| "learning_rate": 8.868629196864182e-06, | |
| "loss": 0.0986, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5979381443298969, | |
| "grad_norm": 1.1299486832378287, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.0869, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6048109965635738, | |
| "grad_norm": 0.9911446233291148, | |
| "learning_rate": 8.791260305278434e-06, | |
| "loss": 0.0973, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6116838487972509, | |
| "grad_norm": 0.8185510402193774, | |
| "learning_rate": 8.751749110782013e-06, | |
| "loss": 0.0822, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6185567010309279, | |
| "grad_norm": 0.6475370169126121, | |
| "learning_rate": 8.71169435655405e-06, | |
| "loss": 0.0613, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6254295532646048, | |
| "grad_norm": 1.1525837420673424, | |
| "learning_rate": 8.671101845794816e-06, | |
| "loss": 0.0786, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6323024054982818, | |
| "grad_norm": 0.8803865912051703, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.0932, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6391752577319587, | |
| "grad_norm": 1.123639509835482, | |
| "learning_rate": 8.588327156186915e-06, | |
| "loss": 0.0886, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6460481099656358, | |
| "grad_norm": 0.9511797057910161, | |
| "learning_rate": 8.546156969874723e-06, | |
| "loss": 0.0908, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6529209621993127, | |
| "grad_norm": 1.1268166346373218, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 0.0707, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6597938144329897, | |
| "grad_norm": 0.9970183599383561, | |
| "learning_rate": 8.460281461786848e-06, | |
| "loss": 0.0938, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.2609006641675766, | |
| "learning_rate": 8.416588581799447e-06, | |
| "loss": 0.0829, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6735395189003437, | |
| "grad_norm": 1.1294421065157962, | |
| "learning_rate": 8.372400700702569e-06, | |
| "loss": 0.0563, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6804123711340206, | |
| "grad_norm": 0.7824698770489632, | |
| "learning_rate": 8.327724220510873e-06, | |
| "loss": 0.0803, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6872852233676976, | |
| "grad_norm": 1.257452489653872, | |
| "learning_rate": 8.282565614028068e-06, | |
| "loss": 0.0954, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6941580756013745, | |
| "grad_norm": 1.1689346986973006, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.0966, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7010309278350515, | |
| "grad_norm": 0.9083340874471744, | |
| "learning_rate": 8.19082826171243e-06, | |
| "loss": 0.0797, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7079037800687286, | |
| "grad_norm": 1.6886520820272226, | |
| "learning_rate": 8.144262806941743e-06, | |
| "loss": 0.0904, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7147766323024055, | |
| "grad_norm": 1.0800227849123984, | |
| "learning_rate": 8.097241806078616e-06, | |
| "loss": 0.0999, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7216494845360825, | |
| "grad_norm": 0.9433214940053437, | |
| "learning_rate": 8.049772071604864e-06, | |
| "loss": 0.0739, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7285223367697594, | |
| "grad_norm": 0.940105894110171, | |
| "learning_rate": 8.001860481015594e-06, | |
| "loss": 0.0951, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7353951890034365, | |
| "grad_norm": 1.0204068764913392, | |
| "learning_rate": 7.953513975822755e-06, | |
| "loss": 0.0812, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7422680412371134, | |
| "grad_norm": 0.8633860480108828, | |
| "learning_rate": 7.904739560549475e-06, | |
| "loss": 0.0966, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7491408934707904, | |
| "grad_norm": 1.158569659689947, | |
| "learning_rate": 7.855544301715203e-06, | |
| "loss": 0.0723, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7560137457044673, | |
| "grad_norm": 1.3512019802948045, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.0939, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7628865979381443, | |
| "grad_norm": 0.8242430726415078, | |
| "learning_rate": 7.755919823271466e-06, | |
| "loss": 0.108, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7697594501718213, | |
| "grad_norm": 1.3248702517651234, | |
| "learning_rate": 7.70550503742427e-06, | |
| "loss": 0.0776, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7766323024054983, | |
| "grad_norm": 0.6656921455785727, | |
| "learning_rate": 7.654698273449435e-06, | |
| "loss": 0.0655, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7835051546391752, | |
| "grad_norm": 1.135719591174255, | |
| "learning_rate": 7.603506892316513e-06, | |
| "loss": 0.0671, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7903780068728522, | |
| "grad_norm": 1.0958384674082662, | |
| "learning_rate": 7.551938310719043e-06, | |
| "loss": 0.1053, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7972508591065293, | |
| "grad_norm": 0.6447102969546324, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.097, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8041237113402062, | |
| "grad_norm": 0.6494313472798022, | |
| "learning_rate": 7.447699485069342e-06, | |
| "loss": 0.0587, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8109965635738832, | |
| "grad_norm": 0.8718931936540331, | |
| "learning_rate": 7.395044343313777e-06, | |
| "loss": 0.0875, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8178694158075601, | |
| "grad_norm": 0.965244519578607, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.0808, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8247422680412371, | |
| "grad_norm": 0.6991417405514189, | |
| "learning_rate": 7.288700744664167e-06, | |
| "loss": 0.0821, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8316151202749141, | |
| "grad_norm": 0.8139440971561749, | |
| "learning_rate": 7.235027695009846e-06, | |
| "loss": 0.0585, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8384879725085911, | |
| "grad_norm": 1.2366511654313435, | |
| "learning_rate": 7.181030830777838e-06, | |
| "loss": 0.0768, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.845360824742268, | |
| "grad_norm": 0.6043375233447036, | |
| "learning_rate": 7.1267179751248005e-06, | |
| "loss": 0.0622, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.852233676975945, | |
| "grad_norm": 1.2645127477841958, | |
| "learning_rate": 7.0720969969887595e-06, | |
| "loss": 0.066, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8591065292096219, | |
| "grad_norm": 0.8441099029992272, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.0723, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.865979381443299, | |
| "grad_norm": 0.9246533446003287, | |
| "learning_rate": 6.961962371079752e-06, | |
| "loss": 0.109, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.872852233676976, | |
| "grad_norm": 1.1435308451255666, | |
| "learning_rate": 6.906464679796927e-06, | |
| "loss": 0.1011, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8797250859106529, | |
| "grad_norm": 0.9701142660236807, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.0842, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8865979381443299, | |
| "grad_norm": 0.8261093803081156, | |
| "learning_rate": 6.79464874240473e-06, | |
| "loss": 0.0699, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8934707903780069, | |
| "grad_norm": 0.8516435899198158, | |
| "learning_rate": 6.7383466963767386e-06, | |
| "loss": 0.1014, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9003436426116839, | |
| "grad_norm": 0.6584764692755108, | |
| "learning_rate": 6.681792795750876e-06, | |
| "loss": 0.0638, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9072164948453608, | |
| "grad_norm": 0.9506573053336792, | |
| "learning_rate": 6.624995234151539e-06, | |
| "loss": 0.0675, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9140893470790378, | |
| "grad_norm": 0.7359533575978741, | |
| "learning_rate": 6.567962240505136e-06, | |
| "loss": 0.0646, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9209621993127147, | |
| "grad_norm": 0.6599722796648425, | |
| "learning_rate": 6.510702077847864e-06, | |
| "loss": 0.0765, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9278350515463918, | |
| "grad_norm": 0.7545308484516775, | |
| "learning_rate": 6.453223042128556e-06, | |
| "loss": 0.0946, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9347079037800687, | |
| "grad_norm": 0.6760323398144172, | |
| "learning_rate": 6.395533461006736e-06, | |
| "loss": 0.0689, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9415807560137457, | |
| "grad_norm": 0.8630370048812374, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.0921, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9484536082474226, | |
| "grad_norm": 0.574545712833708, | |
| "learning_rate": 6.2795561245035895e-06, | |
| "loss": 0.0733, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9553264604810997, | |
| "grad_norm": 0.7507187184479845, | |
| "learning_rate": 6.221285172114156e-06, | |
| "loss": 0.0521, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9621993127147767, | |
| "grad_norm": 0.6872375434803506, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.0676, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9690721649484536, | |
| "grad_norm": 0.6543996218326336, | |
| "learning_rate": 6.104220909805162e-06, | |
| "loss": 0.0594, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9759450171821306, | |
| "grad_norm": 1.0319319500429935, | |
| "learning_rate": 6.045444560353136e-06, | |
| "loss": 0.0839, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9828178694158075, | |
| "grad_norm": 1.049582113193371, | |
| "learning_rate": 5.986516745132e-06, | |
| "loss": 0.0993, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9896907216494846, | |
| "grad_norm": 0.6715813418477135, | |
| "learning_rate": 5.927446001702899e-06, | |
| "loss": 0.0657, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "grad_norm": 1.0096017833690385, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.0863, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "eval_loss": 0.0730990469455719, | |
| "eval_runtime": 28.706, | |
| "eval_samples_per_second": 34.104, | |
| "eval_steps_per_second": 1.08, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0034364261168385, | |
| "grad_norm": 0.3792818768528503, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.0435, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0103092783505154, | |
| "grad_norm": 0.48207234427337675, | |
| "learning_rate": 5.749461880951966e-06, | |
| "loss": 0.0384, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0171821305841924, | |
| "grad_norm": 1.323805674605071, | |
| "learning_rate": 5.689905195840216e-06, | |
| "loss": 0.0693, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0240549828178693, | |
| "grad_norm": 0.7541402069157691, | |
| "learning_rate": 5.630248556101448e-06, | |
| "loss": 0.0584, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0309278350515463, | |
| "grad_norm": 0.510249035189346, | |
| "learning_rate": 5.570500604890124e-06, | |
| "loss": 0.069, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0378006872852235, | |
| "grad_norm": 0.7143541498524518, | |
| "learning_rate": 5.510669998590074e-06, | |
| "loss": 0.0609, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0446735395189004, | |
| "grad_norm": 0.6871778078635953, | |
| "learning_rate": 5.450765405560328e-06, | |
| "loss": 0.0724, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0515463917525774, | |
| "grad_norm": 1.5482443279859988, | |
| "learning_rate": 5.390795504879243e-06, | |
| "loss": 0.0635, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0584192439862543, | |
| "grad_norm": 0.9571497177743679, | |
| "learning_rate": 5.330768985087059e-06, | |
| "loss": 0.0648, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0652920962199313, | |
| "grad_norm": 0.6848343739049801, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.0595, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0721649484536082, | |
| "grad_norm": 0.6350357028922724, | |
| "learning_rate": 5.2105808820857126e-06, | |
| "loss": 0.0635, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0790378006872852, | |
| "grad_norm": 0.8773191653236156, | |
| "learning_rate": 5.150436711931387e-06, | |
| "loss": 0.0656, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0859106529209621, | |
| "grad_norm": 0.8015401149294459, | |
| "learning_rate": 5.090270746252803e-06, | |
| "loss": 0.0757, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0927835051546393, | |
| "grad_norm": 0.6864605970041614, | |
| "learning_rate": 5.030091701996428e-06, | |
| "loss": 0.0742, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0996563573883162, | |
| "grad_norm": 0.5945561169981439, | |
| "learning_rate": 4.9699082980035735e-06, | |
| "loss": 0.0535, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1065292096219932, | |
| "grad_norm": 0.6630258988992135, | |
| "learning_rate": 4.909729253747197e-06, | |
| "loss": 0.0527, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.1134020618556701, | |
| "grad_norm": 0.8987883737056725, | |
| "learning_rate": 4.8495632880686155e-06, | |
| "loss": 0.0798, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.120274914089347, | |
| "grad_norm": 0.8778336956344429, | |
| "learning_rate": 4.789419117914288e-06, | |
| "loss": 0.0578, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.127147766323024, | |
| "grad_norm": 0.6601894461094396, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.0487, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.134020618556701, | |
| "grad_norm": 1.015258423406782, | |
| "learning_rate": 4.6692310149129425e-06, | |
| "loss": 0.0671, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.140893470790378, | |
| "grad_norm": 0.9060845362684526, | |
| "learning_rate": 4.609204495120759e-06, | |
| "loss": 0.0614, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.147766323024055, | |
| "grad_norm": 0.7459044717511902, | |
| "learning_rate": 4.549234594439674e-06, | |
| "loss": 0.0576, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1546391752577319, | |
| "grad_norm": 1.1213777381576666, | |
| "learning_rate": 4.489330001409929e-06, | |
| "loss": 0.0554, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.161512027491409, | |
| "grad_norm": 0.8405329315763191, | |
| "learning_rate": 4.429499395109877e-06, | |
| "loss": 0.06, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.168384879725086, | |
| "grad_norm": 0.6504677223215201, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.0481, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.175257731958763, | |
| "grad_norm": 0.8290928052957108, | |
| "learning_rate": 4.310094804159784e-06, | |
| "loss": 0.0635, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1821305841924399, | |
| "grad_norm": 0.6509843342740524, | |
| "learning_rate": 4.250538119048036e-06, | |
| "loss": 0.0608, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1890034364261168, | |
| "grad_norm": 0.5977601870239029, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.0495, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1958762886597938, | |
| "grad_norm": 0.5966070728973987, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.056, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.2027491408934707, | |
| "grad_norm": 0.6532057059944705, | |
| "learning_rate": 4.072553998297103e-06, | |
| "loss": 0.0557, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2096219931271477, | |
| "grad_norm": 0.6784281742226108, | |
| "learning_rate": 4.013483254868001e-06, | |
| "loss": 0.0629, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.2164948453608249, | |
| "grad_norm": 0.7491167085718105, | |
| "learning_rate": 3.9545554396468655e-06, | |
| "loss": 0.0641, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.2233676975945018, | |
| "grad_norm": 0.6883309064296732, | |
| "learning_rate": 3.89577909019484e-06, | |
| "loss": 0.052, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2302405498281788, | |
| "grad_norm": 0.751751517059043, | |
| "learning_rate": 3.8371627221284495e-06, | |
| "loss": 0.0736, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2371134020618557, | |
| "grad_norm": 1.0052652291730961, | |
| "learning_rate": 3.7787148278858453e-06, | |
| "loss": 0.0665, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2439862542955327, | |
| "grad_norm": 0.9796435936948683, | |
| "learning_rate": 3.7204438754964113e-06, | |
| "loss": 0.0751, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2508591065292096, | |
| "grad_norm": 0.6176321701014562, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.0639, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2577319587628866, | |
| "grad_norm": 0.8252507997013273, | |
| "learning_rate": 3.6044665389932663e-06, | |
| "loss": 0.0728, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.2646048109965635, | |
| "grad_norm": 0.9080726988314038, | |
| "learning_rate": 3.5467769578714455e-06, | |
| "loss": 0.0868, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2714776632302405, | |
| "grad_norm": 0.7803960364764928, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.0761, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2783505154639174, | |
| "grad_norm": 0.4124973291118846, | |
| "learning_rate": 3.432037759494867e-06, | |
| "loss": 0.038, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2852233676975944, | |
| "grad_norm": 0.8885782616442511, | |
| "learning_rate": 3.375004765848463e-06, | |
| "loss": 0.0724, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2920962199312716, | |
| "grad_norm": 0.7560927449933819, | |
| "learning_rate": 3.3182072042491244e-06, | |
| "loss": 0.0705, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2989690721649485, | |
| "grad_norm": 0.5714322429037654, | |
| "learning_rate": 3.2616533036232635e-06, | |
| "loss": 0.0583, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3058419243986255, | |
| "grad_norm": 0.5451943008756435, | |
| "learning_rate": 3.205351257595272e-06, | |
| "loss": 0.0538, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3127147766323024, | |
| "grad_norm": 0.7019156613780937, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.0567, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.3195876288659794, | |
| "grad_norm": 0.7616019502902455, | |
| "learning_rate": 3.093535320203074e-06, | |
| "loss": 0.0762, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.3264604810996563, | |
| "grad_norm": 0.5845383056327452, | |
| "learning_rate": 3.0380376289202497e-06, | |
| "loss": 0.0702, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.6383335634442417, | |
| "learning_rate": 2.982824190050958e-06, | |
| "loss": 0.0669, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3402061855670104, | |
| "grad_norm": 0.5171824169611532, | |
| "learning_rate": 2.927903003011241e-06, | |
| "loss": 0.0508, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3470790378006874, | |
| "grad_norm": 0.61118699277787, | |
| "learning_rate": 2.8732820248752016e-06, | |
| "loss": 0.0702, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3539518900343643, | |
| "grad_norm": 0.5965506449757368, | |
| "learning_rate": 2.8189691692221627e-06, | |
| "loss": 0.052, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3608247422680413, | |
| "grad_norm": 0.5304098262176566, | |
| "learning_rate": 2.7649723049901554e-06, | |
| "loss": 0.0465, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.3676975945017182, | |
| "grad_norm": 0.572758842385705, | |
| "learning_rate": 2.711299255335833e-06, | |
| "loss": 0.0785, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3745704467353952, | |
| "grad_norm": 0.5656522872427878, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.059, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3814432989690721, | |
| "grad_norm": 0.7280624825333891, | |
| "learning_rate": 2.6049556566862234e-06, | |
| "loss": 0.079, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.388316151202749, | |
| "grad_norm": 0.4865849420944666, | |
| "learning_rate": 2.552300514930657e-06, | |
| "loss": 0.0591, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.395189003436426, | |
| "grad_norm": 0.7395203471226048, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0557, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.402061855670103, | |
| "grad_norm": 0.5521743916491753, | |
| "learning_rate": 2.4480616892809593e-06, | |
| "loss": 0.0518, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.40893470790378, | |
| "grad_norm": 0.5546190128532582, | |
| "learning_rate": 2.396493107683488e-06, | |
| "loss": 0.0582, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4158075601374571, | |
| "grad_norm": 0.5865755538294753, | |
| "learning_rate": 2.345301726550567e-06, | |
| "loss": 0.0656, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.422680412371134, | |
| "grad_norm": 0.8010225976277919, | |
| "learning_rate": 2.2944949625757295e-06, | |
| "loss": 0.0843, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.429553264604811, | |
| "grad_norm": 0.45726468260382025, | |
| "learning_rate": 2.244080176728536e-06, | |
| "loss": 0.0495, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.436426116838488, | |
| "grad_norm": 0.7162602003130447, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.0619, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.443298969072165, | |
| "grad_norm": 0.5672015315741569, | |
| "learning_rate": 2.1444556982847996e-06, | |
| "loss": 0.0708, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4501718213058419, | |
| "grad_norm": 0.4922591905848309, | |
| "learning_rate": 2.095260439450526e-06, | |
| "loss": 0.0556, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.4570446735395188, | |
| "grad_norm": 0.5105592569177255, | |
| "learning_rate": 2.0464860241772454e-06, | |
| "loss": 0.0496, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.463917525773196, | |
| "grad_norm": 0.42801907733420835, | |
| "learning_rate": 1.998139518984409e-06, | |
| "loss": 0.0418, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.470790378006873, | |
| "grad_norm": 0.5154589855739227, | |
| "learning_rate": 1.9502279283951363e-06, | |
| "loss": 0.0421, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.47766323024055, | |
| "grad_norm": 0.49641223738548934, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.0418, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.4845360824742269, | |
| "grad_norm": 0.8411343137726398, | |
| "learning_rate": 1.8557371930582579e-06, | |
| "loss": 0.0868, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4914089347079038, | |
| "grad_norm": 0.6940116973003474, | |
| "learning_rate": 1.8091717382875723e-06, | |
| "loss": 0.0775, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.4982817869415808, | |
| "grad_norm": 0.5148356129419294, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.0506, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.5051546391752577, | |
| "grad_norm": 0.5528030783793042, | |
| "learning_rate": 1.7174343859719334e-06, | |
| "loss": 0.0633, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.5120274914089347, | |
| "grad_norm": 0.4861660085487089, | |
| "learning_rate": 1.6722757794891287e-06, | |
| "loss": 0.0495, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5189003436426116, | |
| "grad_norm": 0.6041081070589223, | |
| "learning_rate": 1.627599299297431e-06, | |
| "loss": 0.0581, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.5257731958762886, | |
| "grad_norm": 0.552939461710149, | |
| "learning_rate": 1.5834114182005544e-06, | |
| "loss": 0.0565, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.5326460481099655, | |
| "grad_norm": 0.615554875648061, | |
| "learning_rate": 1.5397185382131524e-06, | |
| "loss": 0.0566, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.5395189003436425, | |
| "grad_norm": 0.5260312169871831, | |
| "learning_rate": 1.4965269896332884e-06, | |
| "loss": 0.0504, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.5463917525773194, | |
| "grad_norm": 0.5830772839525655, | |
| "learning_rate": 1.4538430301252783e-06, | |
| "loss": 0.0485, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5532646048109966, | |
| "grad_norm": 0.7036659618557127, | |
| "learning_rate": 1.411672843813086e-06, | |
| "loss": 0.0476, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.5601374570446735, | |
| "grad_norm": 0.7115804971686124, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.0723, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.5670103092783505, | |
| "grad_norm": 0.8723638397953115, | |
| "learning_rate": 1.3288981542051844e-06, | |
| "loss": 0.0546, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.5738831615120275, | |
| "grad_norm": 0.5529337228641197, | |
| "learning_rate": 1.2883056434459506e-06, | |
| "loss": 0.0503, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.5807560137457046, | |
| "grad_norm": 0.5117329489961148, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.0502, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5876288659793816, | |
| "grad_norm": 0.654668537650446, | |
| "learning_rate": 1.2087396947215678e-06, | |
| "loss": 0.0662, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.5945017182130585, | |
| "grad_norm": 0.5778351597185641, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0525, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.6013745704467355, | |
| "grad_norm": 0.5783805074353839, | |
| "learning_rate": 1.1313708031358183e-06, | |
| "loss": 0.0516, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.6082474226804124, | |
| "grad_norm": 0.5109833692562927, | |
| "learning_rate": 1.0935243153818437e-06, | |
| "loss": 0.0435, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.6151202749140894, | |
| "grad_norm": 0.6411790362572815, | |
| "learning_rate": 1.0562438044060846e-06, | |
| "loss": 0.0621, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6219931271477663, | |
| "grad_norm": 0.6674969284239922, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.0464, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.6288659793814433, | |
| "grad_norm": 0.5052937779674858, | |
| "learning_rate": 9.834022350599538e-07, | |
| "loss": 0.0456, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.6357388316151202, | |
| "grad_norm": 0.4500722667678127, | |
| "learning_rate": 9.47851730098856e-07, | |
| "loss": 0.0475, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.6426116838487972, | |
| "grad_norm": 0.7332831818934408, | |
| "learning_rate": 9.128883072055411e-07, | |
| "loss": 0.0714, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.6494845360824741, | |
| "grad_norm": 0.5671617185996944, | |
| "learning_rate": 8.785170319396174e-07, | |
| "loss": 0.0554, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.656357388316151, | |
| "grad_norm": 0.4765818004960149, | |
| "learning_rate": 8.447428840693489e-07, | |
| "loss": 0.0451, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.663230240549828, | |
| "grad_norm": 0.514720276970688, | |
| "learning_rate": 8.115707568501768e-07, | |
| "loss": 0.0409, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.670103092783505, | |
| "grad_norm": 0.6181901684653612, | |
| "learning_rate": 7.790054563157745e-07, | |
| "loss": 0.0667, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.6769759450171822, | |
| "grad_norm": 0.5221247139473564, | |
| "learning_rate": 7.470517005817473e-07, | |
| "loss": 0.056, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.6838487972508591, | |
| "grad_norm": 0.7521230565222528, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.0573, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.690721649484536, | |
| "grad_norm": 0.8647542041509214, | |
| "learning_rate": 6.849972522982845e-07, | |
| "loss": 0.0791, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.697594501718213, | |
| "grad_norm": 0.553439556229882, | |
| "learning_rate": 6.549055503018575e-07, | |
| "loss": 0.0641, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.7044673539518902, | |
| "grad_norm": 0.6120805979997885, | |
| "learning_rate": 6.254433729092518e-07, | |
| "loss": 0.0598, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.7113402061855671, | |
| "grad_norm": 0.6564283489570599, | |
| "learning_rate": 5.966149886503614e-07, | |
| "loss": 0.0625, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.718213058419244, | |
| "grad_norm": 0.8308307470821662, | |
| "learning_rate": 5.684245742300625e-07, | |
| "loss": 0.0805, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.725085910652921, | |
| "grad_norm": 0.5533842358159983, | |
| "learning_rate": 5.408762139230889e-07, | |
| "loss": 0.0515, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.731958762886598, | |
| "grad_norm": 0.6925181988395948, | |
| "learning_rate": 5.139738989822901e-07, | |
| "loss": 0.0792, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.738831615120275, | |
| "grad_norm": 0.7007456928839768, | |
| "learning_rate": 4.877215270603752e-07, | |
| "loss": 0.0921, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.745704467353952, | |
| "grad_norm": 0.6030922743018291, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.0601, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.7525773195876289, | |
| "grad_norm": 0.542791083389827, | |
| "learning_rate": 4.371817315087845e-07, | |
| "loss": 0.0483, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.7594501718213058, | |
| "grad_norm": 0.5743106264920087, | |
| "learning_rate": 4.1290163016982855e-07, | |
| "loss": 0.062, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.7663230240549828, | |
| "grad_norm": 0.5197888858726925, | |
| "learning_rate": 3.8928611537033424e-07, | |
| "loss": 0.0465, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.7731958762886597, | |
| "grad_norm": 0.5272091850947849, | |
| "learning_rate": 3.663386085658693e-07, | |
| "loss": 0.0541, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.7800687285223367, | |
| "grad_norm": 0.5803359558179969, | |
| "learning_rate": 3.4406243442987765e-07, | |
| "loss": 0.0521, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.7869415807560136, | |
| "grad_norm": 0.6367742593932908, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.0659, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7938144329896906, | |
| "grad_norm": 0.5390042094629793, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.0514, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.8006872852233677, | |
| "grad_norm": 0.6332565296264808, | |
| "learning_rate": 2.812936930186688e-07, | |
| "loss": 0.0659, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.8075601374570447, | |
| "grad_norm": 0.5201101110771652, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.054, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.8144329896907216, | |
| "grad_norm": 0.5891581874064752, | |
| "learning_rate": 2.428610830928152e-07, | |
| "loss": 0.0446, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.8213058419243986, | |
| "grad_norm": 0.46281419422288056, | |
| "learning_rate": 2.2467724440002336e-07, | |
| "loss": 0.0493, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8281786941580758, | |
| "grad_norm": 0.456056578492629, | |
| "learning_rate": 2.0718526251279346e-07, | |
| "loss": 0.0508, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.8350515463917527, | |
| "grad_norm": 0.6488965295362575, | |
| "learning_rate": 1.9038767169893058e-07, | |
| "loss": 0.0713, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.8419243986254297, | |
| "grad_norm": 0.43964927758350936, | |
| "learning_rate": 1.7428690562169003e-07, | |
| "loss": 0.0496, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.8487972508591066, | |
| "grad_norm": 0.5835089963380599, | |
| "learning_rate": 1.5888529698718347e-07, | |
| "loss": 0.0618, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.8556701030927836, | |
| "grad_norm": 0.5603306086318071, | |
| "learning_rate": 1.4418507720641794e-07, | |
| "loss": 0.0718, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8625429553264605, | |
| "grad_norm": 0.5671360254010153, | |
| "learning_rate": 1.3018837607199909e-07, | |
| "loss": 0.0452, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.8694158075601375, | |
| "grad_norm": 0.5512752012857427, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.0504, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.8762886597938144, | |
| "grad_norm": 0.6216583128853239, | |
| "learning_rate": 1.0431353898399388e-07, | |
| "loss": 0.0608, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.8831615120274914, | |
| "grad_norm": 0.4748958652242861, | |
| "learning_rate": 9.243915182039431e-08, | |
| "loss": 0.053, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.8900343642611683, | |
| "grad_norm": 0.7614440067081399, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.0782, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8969072164948453, | |
| "grad_norm": 0.5447897718360889, | |
| "learning_rate": 7.082504191083417e-08, | |
| "loss": 0.041, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.9037800687285222, | |
| "grad_norm": 0.5104719326544138, | |
| "learning_rate": 6.108845065352864e-08, | |
| "loss": 0.0664, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.9106529209621992, | |
| "grad_norm": 0.4783963620489169, | |
| "learning_rate": 5.206741722181385e-08, | |
| "loss": 0.0486, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.9175257731958761, | |
| "grad_norm": 0.5408213302348288, | |
| "learning_rate": 4.376324859820924e-08, | |
| "loss": 0.0671, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.9243986254295533, | |
| "grad_norm": 0.5976950656989035, | |
| "learning_rate": 3.617714790465576e-08, | |
| "loss": 0.0621, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9312714776632303, | |
| "grad_norm": 0.48880888039435383, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.0612, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.9381443298969072, | |
| "grad_norm": 0.5542235264656635, | |
| "learning_rate": 2.3163442461766604e-08, | |
| "loss": 0.077, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.9450171821305842, | |
| "grad_norm": 0.5301882493731837, | |
| "learning_rate": 1.7737723159999e-08, | |
| "loss": 0.0471, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.9518900343642611, | |
| "grad_norm": 0.7105746618136043, | |
| "learning_rate": 1.3033842410251074e-08, | |
| "loss": 0.0596, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.9587628865979383, | |
| "grad_norm": 0.6600370287170537, | |
| "learning_rate": 9.052481718690998e-09, | |
| "loss": 0.0463, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9656357388316152, | |
| "grad_norm": 0.6888269038670338, | |
| "learning_rate": 5.794217911562205e-09, | |
| "loss": 0.0875, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.9725085910652922, | |
| "grad_norm": 0.6350304012767739, | |
| "learning_rate": 3.2595230516152543e-09, | |
| "loss": 0.0663, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.9793814432989691, | |
| "grad_norm": 0.35895733633770877, | |
| "learning_rate": 1.4487643697103092e-09, | |
| "loss": 0.0324, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.986254295532646, | |
| "grad_norm": 0.45234394039475084, | |
| "learning_rate": 3.6220421161692333e-10, | |
| "loss": 0.05, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "grad_norm": 0.5551103292054969, | |
| "learning_rate": 0.0, | |
| "loss": 0.0667, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "eval_loss": 0.061805013567209244, | |
| "eval_runtime": 28.4562, | |
| "eval_samples_per_second": 34.404, | |
| "eval_steps_per_second": 1.089, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "step": 290, | |
| "total_flos": 8.743958405568922e+16, | |
| "train_loss": 0.08316308311347304, | |
| "train_runtime": 3283.3111, | |
| "train_samples_per_second": 11.329, | |
| "train_steps_per_second": 0.088 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 290, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.743958405568922e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |