| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.4985835694050991, | |
| "eval_steps": 500, | |
| "global_step": 88, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0056657223796034, | |
| "grad_norm": 24.973131796915897, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.8537, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0113314447592068, | |
| "grad_norm": 32.79573813738381, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.0212, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0169971671388102, | |
| "grad_norm": 23.800880905805656, | |
| "learning_rate": 3e-06, | |
| "loss": 2.1456, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0226628895184136, | |
| "grad_norm": 19.091198715081358, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.9808, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.028328611898016998, | |
| "grad_norm": 14.124470348172405, | |
| "learning_rate": 5e-06, | |
| "loss": 2.1825, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0339943342776204, | |
| "grad_norm": 11.461608032959802, | |
| "learning_rate": 6e-06, | |
| "loss": 1.6353, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.039660056657223795, | |
| "grad_norm": 10.354681496346823, | |
| "learning_rate": 7e-06, | |
| "loss": 1.9076, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0453257790368272, | |
| "grad_norm": 10.167669680172194, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4754, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05099150141643059, | |
| "grad_norm": 7.5541696713086255, | |
| "learning_rate": 9e-06, | |
| "loss": 1.6213, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.056657223796033995, | |
| "grad_norm": 4.087852973173369, | |
| "learning_rate": 1e-05, | |
| "loss": 1.5217, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06232294617563739, | |
| "grad_norm": 4.071392878063137, | |
| "learning_rate": 9.999948174819623e-06, | |
| "loss": 1.6551, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0679886685552408, | |
| "grad_norm": 5.2075402015034, | |
| "learning_rate": 9.999792700352826e-06, | |
| "loss": 1.4474, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07365439093484419, | |
| "grad_norm": 3.6492933345906637, | |
| "learning_rate": 9.999533579822611e-06, | |
| "loss": 1.5585, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07932011331444759, | |
| "grad_norm": 6.482920810973195, | |
| "learning_rate": 9.999170818600562e-06, | |
| "loss": 1.3317, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08498583569405099, | |
| "grad_norm": 4.137365745831386, | |
| "learning_rate": 9.998704424206747e-06, | |
| "loss": 1.4029, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0906515580736544, | |
| "grad_norm": 4.745717244720069, | |
| "learning_rate": 9.998134406309555e-06, | |
| "loss": 1.6586, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09631728045325778, | |
| "grad_norm": 5.4377770096801346, | |
| "learning_rate": 9.997460776725497e-06, | |
| "loss": 1.365, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.10198300283286119, | |
| "grad_norm": 3.317130182493388, | |
| "learning_rate": 9.996683549418964e-06, | |
| "loss": 1.4956, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10764872521246459, | |
| "grad_norm": 1.7845609616841893, | |
| "learning_rate": 9.995802740501933e-06, | |
| "loss": 1.3472, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.11331444759206799, | |
| "grad_norm": 14.387033772755194, | |
| "learning_rate": 9.994818368233639e-06, | |
| "loss": 1.4116, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11898016997167139, | |
| "grad_norm": 6.920700020611593, | |
| "learning_rate": 9.993730453020187e-06, | |
| "loss": 1.2776, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12464589235127478, | |
| "grad_norm": 6.05951274644599, | |
| "learning_rate": 9.99253901741414e-06, | |
| "loss": 1.4433, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.13031161473087818, | |
| "grad_norm": 3.0541449788715935, | |
| "learning_rate": 9.991244086114046e-06, | |
| "loss": 1.3396, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1359773371104816, | |
| "grad_norm": 1.8438099140328046, | |
| "learning_rate": 9.989845685963917e-06, | |
| "loss": 1.3061, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.141643059490085, | |
| "grad_norm": 4.048301070320613, | |
| "learning_rate": 9.988343845952697e-06, | |
| "loss": 1.2283, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14730878186968838, | |
| "grad_norm": 3.5627296346591457, | |
| "learning_rate": 9.986738597213633e-06, | |
| "loss": 1.2865, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1529745042492918, | |
| "grad_norm": 2.237494567304501, | |
| "learning_rate": 9.98502997302365e-06, | |
| "loss": 1.3233, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15864022662889518, | |
| "grad_norm": 3.479719952104877, | |
| "learning_rate": 9.983218008802648e-06, | |
| "loss": 1.3033, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1643059490084986, | |
| "grad_norm": 2.066121083229141, | |
| "learning_rate": 9.98130274211278e-06, | |
| "loss": 1.3326, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.16997167138810199, | |
| "grad_norm": 4.090684571263736, | |
| "learning_rate": 9.979284212657658e-06, | |
| "loss": 1.3102, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17563739376770537, | |
| "grad_norm": 2.369637256277251, | |
| "learning_rate": 9.977162462281544e-06, | |
| "loss": 1.4067, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1813031161473088, | |
| "grad_norm": 1.4378564529803546, | |
| "learning_rate": 9.97493753496848e-06, | |
| "loss": 1.2409, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.18696883852691218, | |
| "grad_norm": 1.810353068849482, | |
| "learning_rate": 9.972609476841368e-06, | |
| "loss": 1.2659, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.19263456090651557, | |
| "grad_norm": 2.954930884156565, | |
| "learning_rate": 9.970178336161018e-06, | |
| "loss": 1.3727, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.19830028328611898, | |
| "grad_norm": 2.053307140265503, | |
| "learning_rate": 9.967644163325157e-06, | |
| "loss": 1.3463, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.20396600566572237, | |
| "grad_norm": 1.8032124432327943, | |
| "learning_rate": 9.965007010867366e-06, | |
| "loss": 1.1998, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2096317280453258, | |
| "grad_norm": 1.4952983263862012, | |
| "learning_rate": 9.962266933456008e-06, | |
| "loss": 1.2829, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.21529745042492918, | |
| "grad_norm": 1.3649794008291625, | |
| "learning_rate": 9.959423987893086e-06, | |
| "loss": 1.2056, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.22096317280453256, | |
| "grad_norm": 1.4380773398306634, | |
| "learning_rate": 9.956478233113066e-06, | |
| "loss": 1.29, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.22662889518413598, | |
| "grad_norm": 1.6072540934424309, | |
| "learning_rate": 9.953429730181653e-06, | |
| "loss": 1.2593, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.23229461756373937, | |
| "grad_norm": 1.6010739399694889, | |
| "learning_rate": 9.95027854229454e-06, | |
| "loss": 1.2117, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.23796033994334279, | |
| "grad_norm": 1.2474393925785745, | |
| "learning_rate": 9.947024734776076e-06, | |
| "loss": 1.2022, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.24362606232294617, | |
| "grad_norm": 1.4019264249340568, | |
| "learning_rate": 9.943668375077926e-06, | |
| "loss": 1.2365, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.24929178470254956, | |
| "grad_norm": 1.5087040675714003, | |
| "learning_rate": 9.940209532777666e-06, | |
| "loss": 1.274, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.254957507082153, | |
| "grad_norm": 1.1953570915609946, | |
| "learning_rate": 9.93664827957735e-06, | |
| "loss": 1.2526, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.26062322946175637, | |
| "grad_norm": 1.4826450819224886, | |
| "learning_rate": 9.932984689302012e-06, | |
| "loss": 1.1978, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.26628895184135976, | |
| "grad_norm": 1.1937833972167977, | |
| "learning_rate": 9.929218837898143e-06, | |
| "loss": 1.1816, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2719546742209632, | |
| "grad_norm": 1.1238100782353855, | |
| "learning_rate": 9.925350803432112e-06, | |
| "loss": 1.1931, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2776203966005666, | |
| "grad_norm": 1.3338900623153498, | |
| "learning_rate": 9.921380666088558e-06, | |
| "loss": 1.1978, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.28328611898017, | |
| "grad_norm": 1.3236848667289738, | |
| "learning_rate": 9.917308508168712e-06, | |
| "loss": 1.2551, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.28895184135977336, | |
| "grad_norm": 1.425578635546673, | |
| "learning_rate": 9.913134414088698e-06, | |
| "loss": 1.2441, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.29461756373937675, | |
| "grad_norm": 1.171581674684746, | |
| "learning_rate": 9.908858470377793e-06, | |
| "loss": 1.2369, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3002832861189802, | |
| "grad_norm": 1.1564744150302062, | |
| "learning_rate": 9.904480765676617e-06, | |
| "loss": 1.209, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3059490084985836, | |
| "grad_norm": 1.1357504524893798, | |
| "learning_rate": 9.9000013907353e-06, | |
| "loss": 1.2152, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.311614730878187, | |
| "grad_norm": 1.0498825437855333, | |
| "learning_rate": 9.895420438411616e-06, | |
| "loss": 1.2043, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.31728045325779036, | |
| "grad_norm": 1.6465219316145685, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 1.2289, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.32294617563739375, | |
| "grad_norm": 1.711551232749367, | |
| "learning_rate": 9.885954183574753e-06, | |
| "loss": 1.1831, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3286118980169972, | |
| "grad_norm": 1.2636664413259953, | |
| "learning_rate": 9.881069077297724e-06, | |
| "loss": 1.2061, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3342776203966006, | |
| "grad_norm": 1.4260407982081962, | |
| "learning_rate": 9.876082786106546e-06, | |
| "loss": 1.1998, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.33994334277620397, | |
| "grad_norm": 1.95604739866899, | |
| "learning_rate": 9.870995413367397e-06, | |
| "loss": 1.2215, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.34560906515580736, | |
| "grad_norm": 1.2316545141521473, | |
| "learning_rate": 9.865807064541878e-06, | |
| "loss": 1.1599, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.35127478753541075, | |
| "grad_norm": 1.1178440688886253, | |
| "learning_rate": 9.860517847184837e-06, | |
| "loss": 1.1907, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.35694050991501414, | |
| "grad_norm": 1.305376049095191, | |
| "learning_rate": 9.855127870942131e-06, | |
| "loss": 1.1474, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3626062322946176, | |
| "grad_norm": 1.0495122657744762, | |
| "learning_rate": 9.849637247548356e-06, | |
| "loss": 1.2424, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.36827195467422097, | |
| "grad_norm": 1.141538926125254, | |
| "learning_rate": 9.844046090824533e-06, | |
| "loss": 1.1689, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.37393767705382436, | |
| "grad_norm": 1.26961257521241, | |
| "learning_rate": 9.83835451667574e-06, | |
| "loss": 1.2106, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.37960339943342775, | |
| "grad_norm": 1.081533609255719, | |
| "learning_rate": 9.832562643088724e-06, | |
| "loss": 1.1834, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.38526912181303113, | |
| "grad_norm": 1.443083776392187, | |
| "learning_rate": 9.826670590129442e-06, | |
| "loss": 1.1505, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3909348441926346, | |
| "grad_norm": 1.135777382976375, | |
| "learning_rate": 9.820678479940573e-06, | |
| "loss": 1.1489, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.39660056657223797, | |
| "grad_norm": 1.8779005247112062, | |
| "learning_rate": 9.814586436738998e-06, | |
| "loss": 1.1643, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.40226628895184136, | |
| "grad_norm": 1.7980060811236744, | |
| "learning_rate": 9.808394586813209e-06, | |
| "loss": 1.1594, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.40793201133144474, | |
| "grad_norm": 2.572405910372765, | |
| "learning_rate": 9.802103058520704e-06, | |
| "loss": 1.1854, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.41359773371104813, | |
| "grad_norm": 2.0253448122778606, | |
| "learning_rate": 9.795711982285317e-06, | |
| "loss": 1.1826, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4192634560906516, | |
| "grad_norm": 6.483254642683073, | |
| "learning_rate": 9.78922149059452e-06, | |
| "loss": 1.1646, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.42492917847025496, | |
| "grad_norm": 1.2964281102887218, | |
| "learning_rate": 9.782631717996675e-06, | |
| "loss": 1.2379, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.43059490084985835, | |
| "grad_norm": 1.9517402996335103, | |
| "learning_rate": 9.775942801098241e-06, | |
| "loss": 1.164, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.43626062322946174, | |
| "grad_norm": 3.064531007561859, | |
| "learning_rate": 9.76915487856095e-06, | |
| "loss": 1.1418, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.44192634560906513, | |
| "grad_norm": 1.5009905490397355, | |
| "learning_rate": 9.762268091098926e-06, | |
| "loss": 1.1653, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4475920679886686, | |
| "grad_norm": 1.104518219439204, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 1.2025, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.45325779036827196, | |
| "grad_norm": 7.807500502849419, | |
| "learning_rate": 9.748198494501598e-06, | |
| "loss": 1.148, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.45892351274787535, | |
| "grad_norm": 6.196503908242147, | |
| "learning_rate": 9.741015977030046e-06, | |
| "loss": 1.1819, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.46458923512747874, | |
| "grad_norm": 2.2714978855142736, | |
| "learning_rate": 9.733735177955219e-06, | |
| "loss": 1.1907, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4702549575070821, | |
| "grad_norm": 1.834743890260826, | |
| "learning_rate": 9.72635624820861e-06, | |
| "loss": 1.1381, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.47592067988668557, | |
| "grad_norm": 1.28470626171519, | |
| "learning_rate": 9.71887934075596e-06, | |
| "loss": 1.2079, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.48158640226628896, | |
| "grad_norm": 6.197048819949928, | |
| "learning_rate": 9.711304610594104e-06, | |
| "loss": 1.1272, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.48725212464589235, | |
| "grad_norm": 3.412508821399008, | |
| "learning_rate": 9.703632214747742e-06, | |
| "loss": 1.2382, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.49291784702549574, | |
| "grad_norm": 1.57336480270559, | |
| "learning_rate": 9.695862312266195e-06, | |
| "loss": 1.157, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4985835694050991, | |
| "grad_norm": 7.383065472181884, | |
| "learning_rate": 9.687995064220102e-06, | |
| "loss": 1.1684, | |
| "step": 88 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 700, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 88, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.300089449218048e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |