| { | |
| "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": 13.345214197245895, | |
| "learning_rate": 3.4482758620689656e-07, | |
| "loss": 0.3093, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013745704467353952, | |
| "grad_norm": 12.249021263990924, | |
| "learning_rate": 6.896551724137931e-07, | |
| "loss": 0.3119, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020618556701030927, | |
| "grad_norm": 11.49649478797453, | |
| "learning_rate": 1.0344827586206898e-06, | |
| "loss": 0.2713, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.027491408934707903, | |
| "grad_norm": 10.99705450605852, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 0.3341, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03436426116838488, | |
| "grad_norm": 10.272718388143904, | |
| "learning_rate": 1.724137931034483e-06, | |
| "loss": 0.2736, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.041237113402061855, | |
| "grad_norm": 8.248676176882142, | |
| "learning_rate": 2.0689655172413796e-06, | |
| "loss": 0.2205, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.048109965635738834, | |
| "grad_norm": 4.8852411256237325, | |
| "learning_rate": 2.4137931034482762e-06, | |
| "loss": 0.1904, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.054982817869415807, | |
| "grad_norm": 5.615836145355622, | |
| "learning_rate": 2.7586206896551725e-06, | |
| "loss": 0.2093, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.061855670103092786, | |
| "grad_norm": 4.295001880258592, | |
| "learning_rate": 3.103448275862069e-06, | |
| "loss": 0.1953, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06872852233676977, | |
| "grad_norm": 3.2825548271549745, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.146, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07560137457044673, | |
| "grad_norm": 2.5508942900147598, | |
| "learning_rate": 3.793103448275862e-06, | |
| "loss": 0.1738, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08247422680412371, | |
| "grad_norm": 7.427359395591835, | |
| "learning_rate": 4.137931034482759e-06, | |
| "loss": 0.1945, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08934707903780069, | |
| "grad_norm": 3.074668604987824, | |
| "learning_rate": 4.482758620689656e-06, | |
| "loss": 0.1455, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09621993127147767, | |
| "grad_norm": 3.9676981540389873, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 0.1831, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10309278350515463, | |
| "grad_norm": 2.586033608008397, | |
| "learning_rate": 5.172413793103449e-06, | |
| "loss": 0.1434, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10996563573883161, | |
| "grad_norm": 3.8365084868937003, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 0.1304, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11683848797250859, | |
| "grad_norm": 2.4186281982775277, | |
| "learning_rate": 5.862068965517242e-06, | |
| "loss": 0.1469, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12371134020618557, | |
| "grad_norm": 2.4172297480005116, | |
| "learning_rate": 6.206896551724138e-06, | |
| "loss": 0.1281, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13058419243986255, | |
| "grad_norm": 2.0122824189206723, | |
| "learning_rate": 6.551724137931035e-06, | |
| "loss": 0.0946, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13745704467353953, | |
| "grad_norm": 3.128658170670005, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 0.1158, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14432989690721648, | |
| "grad_norm": 1.789362827428083, | |
| "learning_rate": 7.241379310344828e-06, | |
| "loss": 0.1408, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15120274914089346, | |
| "grad_norm": 1.741952362218679, | |
| "learning_rate": 7.586206896551724e-06, | |
| "loss": 0.1175, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15807560137457044, | |
| "grad_norm": 1.8867611900792864, | |
| "learning_rate": 7.93103448275862e-06, | |
| "loss": 0.1266, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16494845360824742, | |
| "grad_norm": 2.2879379723149076, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 0.1039, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1718213058419244, | |
| "grad_norm": 3.315181989144902, | |
| "learning_rate": 8.620689655172414e-06, | |
| "loss": 0.1105, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.17869415807560138, | |
| "grad_norm": 1.5834323731334103, | |
| "learning_rate": 8.965517241379312e-06, | |
| "loss": 0.0949, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18556701030927836, | |
| "grad_norm": 2.077663073352962, | |
| "learning_rate": 9.310344827586207e-06, | |
| "loss": 0.1047, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19243986254295534, | |
| "grad_norm": 1.6882617257115937, | |
| "learning_rate": 9.655172413793105e-06, | |
| "loss": 0.118, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19931271477663232, | |
| "grad_norm": 1.5682872681531057, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1059, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20618556701030927, | |
| "grad_norm": 4.414553432016193, | |
| "learning_rate": 9.999637795788383e-06, | |
| "loss": 0.1481, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21305841924398625, | |
| "grad_norm": 4.238937260231911, | |
| "learning_rate": 9.99855123563029e-06, | |
| "loss": 0.1075, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21993127147766323, | |
| "grad_norm": 3.0441967557817673, | |
| "learning_rate": 9.996740476948386e-06, | |
| "loss": 0.1266, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2268041237113402, | |
| "grad_norm": 1.565915970733602, | |
| "learning_rate": 9.994205782088438e-06, | |
| "loss": 0.1127, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.23367697594501718, | |
| "grad_norm": 2.3181822016629474, | |
| "learning_rate": 9.990947518281312e-06, | |
| "loss": 0.12, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24054982817869416, | |
| "grad_norm": 2.052251118378025, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 0.1002, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24742268041237114, | |
| "grad_norm": 1.4537686585245448, | |
| "learning_rate": 9.982262276840002e-06, | |
| "loss": 0.1092, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2542955326460481, | |
| "grad_norm": 1.7680937137380806, | |
| "learning_rate": 9.976836557538234e-06, | |
| "loss": 0.1395, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2611683848797251, | |
| "grad_norm": 2.0593524697227332, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.1169, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26804123711340205, | |
| "grad_norm": 1.9733940657942892, | |
| "learning_rate": 9.963822852095344e-06, | |
| "loss": 0.0954, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.27491408934707906, | |
| "grad_norm": 1.2904048681755698, | |
| "learning_rate": 9.95623675140179e-06, | |
| "loss": 0.1116, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.281786941580756, | |
| "grad_norm": 1.904875763159912, | |
| "learning_rate": 9.947932582778188e-06, | |
| "loss": 0.1275, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.28865979381443296, | |
| "grad_norm": 1.7128689866675502, | |
| "learning_rate": 9.938911549346473e-06, | |
| "loss": 0.1054, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.29553264604810997, | |
| "grad_norm": 1.8719076873766436, | |
| "learning_rate": 9.929174958089167e-06, | |
| "loss": 0.0753, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3024054982817869, | |
| "grad_norm": 1.295490979599785, | |
| "learning_rate": 9.918724219660013e-06, | |
| "loss": 0.1056, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30927835051546393, | |
| "grad_norm": 1.995863155464365, | |
| "learning_rate": 9.907560848179607e-06, | |
| "loss": 0.1172, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3161512027491409, | |
| "grad_norm": 1.5818625687456236, | |
| "learning_rate": 9.895686461016007e-06, | |
| "loss": 0.0924, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3230240549828179, | |
| "grad_norm": 2.381163513512712, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.1004, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32989690721649484, | |
| "grad_norm": 1.5493255499019554, | |
| "learning_rate": 9.869811623928001e-06, | |
| "loss": 0.0838, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.33676975945017185, | |
| "grad_norm": 1.7042145931444017, | |
| "learning_rate": 9.855814922793583e-06, | |
| "loss": 0.0925, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3436426116838488, | |
| "grad_norm": 1.7404286715367254, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 0.08, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35051546391752575, | |
| "grad_norm": 1.4892054433508841, | |
| "learning_rate": 9.82571309437831e-06, | |
| "loss": 0.1119, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.35738831615120276, | |
| "grad_norm": 2.3932462419312186, | |
| "learning_rate": 9.809612328301071e-06, | |
| "loss": 0.1309, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3642611683848797, | |
| "grad_norm": 1.2909299759823298, | |
| "learning_rate": 9.792814737487207e-06, | |
| "loss": 0.0826, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3711340206185567, | |
| "grad_norm": 2.212046037479329, | |
| "learning_rate": 9.775322755599979e-06, | |
| "loss": 0.1331, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37800687285223367, | |
| "grad_norm": 1.6196725387822164, | |
| "learning_rate": 9.757138916907184e-06, | |
| "loss": 0.0804, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3848797250859107, | |
| "grad_norm": 1.1384734406191421, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.0846, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3917525773195876, | |
| "grad_norm": 1.7599683872266434, | |
| "learning_rate": 9.718706306981332e-06, | |
| "loss": 0.091, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39862542955326463, | |
| "grad_norm": 1.328417830470734, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.0864, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4054982817869416, | |
| "grad_norm": 1.2938474097335424, | |
| "learning_rate": 9.677539179628005e-06, | |
| "loss": 0.1164, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.41237113402061853, | |
| "grad_norm": 1.18180879989636, | |
| "learning_rate": 9.655937565570124e-06, | |
| "loss": 0.0942, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.41924398625429554, | |
| "grad_norm": 1.036947655872308, | |
| "learning_rate": 9.63366139143413e-06, | |
| "loss": 0.0955, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4261168384879725, | |
| "grad_norm": 1.3791489638686576, | |
| "learning_rate": 9.610713884629667e-06, | |
| "loss": 0.1378, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4329896907216495, | |
| "grad_norm": 0.8455943762549145, | |
| "learning_rate": 9.587098369830171e-06, | |
| "loss": 0.0902, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43986254295532645, | |
| "grad_norm": 1.0364817345124417, | |
| "learning_rate": 9.562818268491216e-06, | |
| "loss": 0.0891, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.44673539518900346, | |
| "grad_norm": 0.9145706414839548, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.0993, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4536082474226804, | |
| "grad_norm": 1.2307002590229683, | |
| "learning_rate": 9.512278472939627e-06, | |
| "loss": 0.0684, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.46048109965635736, | |
| "grad_norm": 3.248646955467502, | |
| "learning_rate": 9.486026101017711e-06, | |
| "loss": 0.0762, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.46735395189003437, | |
| "grad_norm": 1.063052020842025, | |
| "learning_rate": 9.459123786076911e-06, | |
| "loss": 0.0981, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4742268041237113, | |
| "grad_norm": 0.8697160868002, | |
| "learning_rate": 9.431575425769938e-06, | |
| "loss": 0.0742, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.48109965635738833, | |
| "grad_norm": 1.1592746453294174, | |
| "learning_rate": 9.40338501134964e-06, | |
| "loss": 0.128, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4879725085910653, | |
| "grad_norm": 1.1094656802415568, | |
| "learning_rate": 9.374556627090749e-06, | |
| "loss": 0.0848, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4948453608247423, | |
| "grad_norm": 1.0684985482189204, | |
| "learning_rate": 9.345094449698143e-06, | |
| "loss": 0.0946, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5017182130584192, | |
| "grad_norm": 1.3314170233186653, | |
| "learning_rate": 9.315002747701716e-06, | |
| "loss": 0.0802, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5085910652920962, | |
| "grad_norm": 1.4306876516200886, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.0752, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5154639175257731, | |
| "grad_norm": 1.2048018409662151, | |
| "learning_rate": 9.252948299418255e-06, | |
| "loss": 0.0829, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5223367697594502, | |
| "grad_norm": 1.3293078344890725, | |
| "learning_rate": 9.220994543684225e-06, | |
| "loss": 0.0931, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5292096219931272, | |
| "grad_norm": 1.0614233232986774, | |
| "learning_rate": 9.188429243149824e-06, | |
| "loss": 0.1064, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5360824742268041, | |
| "grad_norm": 1.0841808051085717, | |
| "learning_rate": 9.155257115930651e-06, | |
| "loss": 0.0743, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5429553264604811, | |
| "grad_norm": 1.1325476576242235, | |
| "learning_rate": 9.121482968060384e-06, | |
| "loss": 0.0813, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5498281786941581, | |
| "grad_norm": 1.2569421767754902, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 0.1244, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5567010309278351, | |
| "grad_norm": 0.8239743727485099, | |
| "learning_rate": 9.052148269901145e-06, | |
| "loss": 0.0745, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.563573883161512, | |
| "grad_norm": 0.901988148458852, | |
| "learning_rate": 9.01659776494005e-06, | |
| "loss": 0.084, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.570446735395189, | |
| "grad_norm": 1.475769093950794, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.1194, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5773195876288659, | |
| "grad_norm": 0.6883362380086342, | |
| "learning_rate": 8.943756195593916e-06, | |
| "loss": 0.0676, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.584192439862543, | |
| "grad_norm": 0.916699408949792, | |
| "learning_rate": 8.90647568461816e-06, | |
| "loss": 0.0724, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5910652920962199, | |
| "grad_norm": 0.843282856725552, | |
| "learning_rate": 8.868629196864182e-06, | |
| "loss": 0.0953, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5979381443298969, | |
| "grad_norm": 0.9800172659562766, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.1002, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6048109965635738, | |
| "grad_norm": 0.9473164596227117, | |
| "learning_rate": 8.791260305278434e-06, | |
| "loss": 0.0784, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6116838487972509, | |
| "grad_norm": 1.113941306186895, | |
| "learning_rate": 8.751749110782013e-06, | |
| "loss": 0.1011, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6185567010309279, | |
| "grad_norm": 0.8133757605577656, | |
| "learning_rate": 8.71169435655405e-06, | |
| "loss": 0.0758, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6254295532646048, | |
| "grad_norm": 0.9965704208663702, | |
| "learning_rate": 8.671101845794816e-06, | |
| "loss": 0.0964, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6323024054982818, | |
| "grad_norm": 0.9563113095353781, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.1058, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6391752577319587, | |
| "grad_norm": 0.578119288013714, | |
| "learning_rate": 8.588327156186915e-06, | |
| "loss": 0.0647, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6460481099656358, | |
| "grad_norm": 0.9762870948117027, | |
| "learning_rate": 8.546156969874723e-06, | |
| "loss": 0.0803, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6529209621993127, | |
| "grad_norm": 0.824669613807026, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 0.095, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6597938144329897, | |
| "grad_norm": 1.3980066883113789, | |
| "learning_rate": 8.460281461786848e-06, | |
| "loss": 0.113, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.1861256711176218, | |
| "learning_rate": 8.416588581799447e-06, | |
| "loss": 0.1037, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6735395189003437, | |
| "grad_norm": 1.121535316883938, | |
| "learning_rate": 8.372400700702569e-06, | |
| "loss": 0.0992, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6804123711340206, | |
| "grad_norm": 0.6355127288784651, | |
| "learning_rate": 8.327724220510873e-06, | |
| "loss": 0.0735, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6872852233676976, | |
| "grad_norm": 0.9147552926213036, | |
| "learning_rate": 8.282565614028068e-06, | |
| "loss": 0.0843, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6941580756013745, | |
| "grad_norm": 0.9346419658075052, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.0855, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7010309278350515, | |
| "grad_norm": 1.1799679874603957, | |
| "learning_rate": 8.19082826171243e-06, | |
| "loss": 0.0915, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7079037800687286, | |
| "grad_norm": 1.0639293612662772, | |
| "learning_rate": 8.144262806941743e-06, | |
| "loss": 0.0661, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7147766323024055, | |
| "grad_norm": 1.0020414931224566, | |
| "learning_rate": 8.097241806078616e-06, | |
| "loss": 0.0758, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7216494845360825, | |
| "grad_norm": 1.21947350536526, | |
| "learning_rate": 8.049772071604864e-06, | |
| "loss": 0.0884, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7285223367697594, | |
| "grad_norm": 1.1574684204166596, | |
| "learning_rate": 8.001860481015594e-06, | |
| "loss": 0.0603, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7353951890034365, | |
| "grad_norm": 0.7452557060229312, | |
| "learning_rate": 7.953513975822755e-06, | |
| "loss": 0.0705, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7422680412371134, | |
| "grad_norm": 1.2415899111794266, | |
| "learning_rate": 7.904739560549475e-06, | |
| "loss": 0.105, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7491408934707904, | |
| "grad_norm": 0.9155195031641292, | |
| "learning_rate": 7.855544301715203e-06, | |
| "loss": 0.0842, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7560137457044673, | |
| "grad_norm": 1.4258092337993582, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.0999, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7628865979381443, | |
| "grad_norm": 0.7310126712756093, | |
| "learning_rate": 7.755919823271466e-06, | |
| "loss": 0.0754, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7697594501718213, | |
| "grad_norm": 1.1946989492511406, | |
| "learning_rate": 7.70550503742427e-06, | |
| "loss": 0.1138, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7766323024054983, | |
| "grad_norm": 0.7959308512646723, | |
| "learning_rate": 7.654698273449435e-06, | |
| "loss": 0.0903, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7835051546391752, | |
| "grad_norm": 1.1237178964816936, | |
| "learning_rate": 7.603506892316513e-06, | |
| "loss": 0.1068, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7903780068728522, | |
| "grad_norm": 0.9861051959516751, | |
| "learning_rate": 7.551938310719043e-06, | |
| "loss": 0.0878, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7972508591065293, | |
| "grad_norm": 1.1534002165137225, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0743, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8041237113402062, | |
| "grad_norm": 1.6747193293802427, | |
| "learning_rate": 7.447699485069342e-06, | |
| "loss": 0.0927, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8109965635738832, | |
| "grad_norm": 0.9828220699895688, | |
| "learning_rate": 7.395044343313777e-06, | |
| "loss": 0.0925, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8178694158075601, | |
| "grad_norm": 1.2699548213673144, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.0656, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8247422680412371, | |
| "grad_norm": 1.482374374759964, | |
| "learning_rate": 7.288700744664167e-06, | |
| "loss": 0.0788, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8316151202749141, | |
| "grad_norm": 0.8145500193220969, | |
| "learning_rate": 7.235027695009846e-06, | |
| "loss": 0.0693, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8384879725085911, | |
| "grad_norm": 0.8322499618921504, | |
| "learning_rate": 7.181030830777838e-06, | |
| "loss": 0.0789, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.845360824742268, | |
| "grad_norm": 0.5923486542549129, | |
| "learning_rate": 7.1267179751248005e-06, | |
| "loss": 0.0491, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.852233676975945, | |
| "grad_norm": 1.114479812776146, | |
| "learning_rate": 7.0720969969887595e-06, | |
| "loss": 0.0783, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8591065292096219, | |
| "grad_norm": 0.8415271939464942, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.1023, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.865979381443299, | |
| "grad_norm": 0.9670960727161972, | |
| "learning_rate": 6.961962371079752e-06, | |
| "loss": 0.0802, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.872852233676976, | |
| "grad_norm": 0.5868172521645618, | |
| "learning_rate": 6.906464679796927e-06, | |
| "loss": 0.0654, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8797250859106529, | |
| "grad_norm": 1.1189751307152005, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.0824, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8865979381443299, | |
| "grad_norm": 0.8315376522697373, | |
| "learning_rate": 6.79464874240473e-06, | |
| "loss": 0.0757, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8934707903780069, | |
| "grad_norm": 0.7605448062221472, | |
| "learning_rate": 6.7383466963767386e-06, | |
| "loss": 0.0704, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9003436426116839, | |
| "grad_norm": 1.0350123055773908, | |
| "learning_rate": 6.681792795750876e-06, | |
| "loss": 0.0865, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9072164948453608, | |
| "grad_norm": 0.5877475602787419, | |
| "learning_rate": 6.624995234151539e-06, | |
| "loss": 0.0603, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9140893470790378, | |
| "grad_norm": 0.9077747418504163, | |
| "learning_rate": 6.567962240505136e-06, | |
| "loss": 0.0879, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9209621993127147, | |
| "grad_norm": 1.1399760162988553, | |
| "learning_rate": 6.510702077847864e-06, | |
| "loss": 0.0855, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9278350515463918, | |
| "grad_norm": 0.8710491208292915, | |
| "learning_rate": 6.453223042128556e-06, | |
| "loss": 0.0738, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9347079037800687, | |
| "grad_norm": 1.5315294019402452, | |
| "learning_rate": 6.395533461006736e-06, | |
| "loss": 0.0846, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9415807560137457, | |
| "grad_norm": 0.6926898610907182, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.0641, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9484536082474226, | |
| "grad_norm": 1.2509579295881588, | |
| "learning_rate": 6.2795561245035895e-06, | |
| "loss": 0.0917, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9553264604810997, | |
| "grad_norm": 0.7303867094596694, | |
| "learning_rate": 6.221285172114156e-06, | |
| "loss": 0.0525, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9621993127147767, | |
| "grad_norm": 0.7095650683135653, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.0679, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9690721649484536, | |
| "grad_norm": 0.681190163884521, | |
| "learning_rate": 6.104220909805162e-06, | |
| "loss": 0.0769, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9759450171821306, | |
| "grad_norm": 0.67794563705019, | |
| "learning_rate": 6.045444560353136e-06, | |
| "loss": 0.0818, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9828178694158075, | |
| "grad_norm": 1.0430032378123106, | |
| "learning_rate": 5.986516745132e-06, | |
| "loss": 0.0862, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9896907216494846, | |
| "grad_norm": 0.6693707295010333, | |
| "learning_rate": 5.927446001702899e-06, | |
| "loss": 0.0629, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "grad_norm": 0.9354206683394047, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.0949, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "eval_loss": 0.07529699802398682, | |
| "eval_runtime": 28.9095, | |
| "eval_samples_per_second": 33.864, | |
| "eval_steps_per_second": 1.072, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0034364261168385, | |
| "grad_norm": 0.5703912526978481, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.0588, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0103092783505154, | |
| "grad_norm": 0.8697655476763948, | |
| "learning_rate": 5.749461880951966e-06, | |
| "loss": 0.0739, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0171821305841924, | |
| "grad_norm": 0.7388127068166545, | |
| "learning_rate": 5.689905195840216e-06, | |
| "loss": 0.0449, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0240549828178693, | |
| "grad_norm": 0.7429816879374691, | |
| "learning_rate": 5.630248556101448e-06, | |
| "loss": 0.0629, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0309278350515463, | |
| "grad_norm": 0.81173215687706, | |
| "learning_rate": 5.570500604890124e-06, | |
| "loss": 0.0802, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0378006872852235, | |
| "grad_norm": 0.5016553256715115, | |
| "learning_rate": 5.510669998590074e-06, | |
| "loss": 0.0388, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0446735395189004, | |
| "grad_norm": 0.7104815594961017, | |
| "learning_rate": 5.450765405560328e-06, | |
| "loss": 0.0607, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0515463917525774, | |
| "grad_norm": 0.7352046418628161, | |
| "learning_rate": 5.390795504879243e-06, | |
| "loss": 0.0606, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0584192439862543, | |
| "grad_norm": 0.9552894460690632, | |
| "learning_rate": 5.330768985087059e-06, | |
| "loss": 0.0738, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0652920962199313, | |
| "grad_norm": 0.7534982301250649, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.0993, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0721649484536082, | |
| "grad_norm": 0.9277931636306036, | |
| "learning_rate": 5.2105808820857126e-06, | |
| "loss": 0.0751, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0790378006872852, | |
| "grad_norm": 0.8158254871396599, | |
| "learning_rate": 5.150436711931387e-06, | |
| "loss": 0.0645, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0859106529209621, | |
| "grad_norm": 0.7077585135379243, | |
| "learning_rate": 5.090270746252803e-06, | |
| "loss": 0.062, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0927835051546393, | |
| "grad_norm": 0.5885155173281448, | |
| "learning_rate": 5.030091701996428e-06, | |
| "loss": 0.0493, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0996563573883162, | |
| "grad_norm": 0.9034728237089428, | |
| "learning_rate": 4.9699082980035735e-06, | |
| "loss": 0.0645, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1065292096219932, | |
| "grad_norm": 0.8352242596831572, | |
| "learning_rate": 4.909729253747197e-06, | |
| "loss": 0.0696, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.1134020618556701, | |
| "grad_norm": 0.6838209698400846, | |
| "learning_rate": 4.8495632880686155e-06, | |
| "loss": 0.0543, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.120274914089347, | |
| "grad_norm": 0.7074113487575592, | |
| "learning_rate": 4.789419117914288e-06, | |
| "loss": 0.0537, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.127147766323024, | |
| "grad_norm": 0.6561094016081127, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.0501, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.134020618556701, | |
| "grad_norm": 0.6187571128750288, | |
| "learning_rate": 4.6692310149129425e-06, | |
| "loss": 0.0389, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.140893470790378, | |
| "grad_norm": 2.173840283190834, | |
| "learning_rate": 4.609204495120759e-06, | |
| "loss": 0.0774, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.147766323024055, | |
| "grad_norm": 0.615595936936975, | |
| "learning_rate": 4.549234594439674e-06, | |
| "loss": 0.0488, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1546391752577319, | |
| "grad_norm": 0.787444775466672, | |
| "learning_rate": 4.489330001409929e-06, | |
| "loss": 0.0607, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.161512027491409, | |
| "grad_norm": 0.5267430413256998, | |
| "learning_rate": 4.429499395109877e-06, | |
| "loss": 0.0631, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.168384879725086, | |
| "grad_norm": 0.5987274004812723, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.0714, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.175257731958763, | |
| "grad_norm": 0.6840745613955029, | |
| "learning_rate": 4.310094804159784e-06, | |
| "loss": 0.0717, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1821305841924399, | |
| "grad_norm": 0.693501686709217, | |
| "learning_rate": 4.250538119048036e-06, | |
| "loss": 0.0595, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1890034364261168, | |
| "grad_norm": 0.5469655780619754, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.0527, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1958762886597938, | |
| "grad_norm": 0.5553781103640738, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.0589, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.2027491408934707, | |
| "grad_norm": 0.5038987456529297, | |
| "learning_rate": 4.072553998297103e-06, | |
| "loss": 0.0588, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2096219931271477, | |
| "grad_norm": 0.8019892222377671, | |
| "learning_rate": 4.013483254868001e-06, | |
| "loss": 0.0763, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.2164948453608249, | |
| "grad_norm": 0.9382214504383896, | |
| "learning_rate": 3.9545554396468655e-06, | |
| "loss": 0.1074, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.2233676975945018, | |
| "grad_norm": 0.540226527722293, | |
| "learning_rate": 3.89577909019484e-06, | |
| "loss": 0.0592, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2302405498281788, | |
| "grad_norm": 0.8678372259590639, | |
| "learning_rate": 3.8371627221284495e-06, | |
| "loss": 0.0525, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2371134020618557, | |
| "grad_norm": 1.0696138624517977, | |
| "learning_rate": 3.7787148278858453e-06, | |
| "loss": 0.0794, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2439862542955327, | |
| "grad_norm": 0.7129407435300503, | |
| "learning_rate": 3.7204438754964113e-06, | |
| "loss": 0.0685, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2508591065292096, | |
| "grad_norm": 0.8138308603975348, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.0681, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2577319587628866, | |
| "grad_norm": 0.8711788985387782, | |
| "learning_rate": 3.6044665389932663e-06, | |
| "loss": 0.0794, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.2646048109965635, | |
| "grad_norm": 0.663379516796088, | |
| "learning_rate": 3.5467769578714455e-06, | |
| "loss": 0.0592, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2714776632302405, | |
| "grad_norm": 0.6406176289851182, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.0556, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2783505154639174, | |
| "grad_norm": 0.8490733271265373, | |
| "learning_rate": 3.432037759494867e-06, | |
| "loss": 0.0556, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2852233676975944, | |
| "grad_norm": 0.808921808504203, | |
| "learning_rate": 3.375004765848463e-06, | |
| "loss": 0.0516, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2920962199312716, | |
| "grad_norm": 1.0674945201805925, | |
| "learning_rate": 3.3182072042491244e-06, | |
| "loss": 0.0752, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2989690721649485, | |
| "grad_norm": 0.6357296022355372, | |
| "learning_rate": 3.2616533036232635e-06, | |
| "loss": 0.0561, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3058419243986255, | |
| "grad_norm": 1.0447237988044837, | |
| "learning_rate": 3.205351257595272e-06, | |
| "loss": 0.0706, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3127147766323024, | |
| "grad_norm": 0.6169076770106159, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.0509, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.3195876288659794, | |
| "grad_norm": 0.9040797513348718, | |
| "learning_rate": 3.093535320203074e-06, | |
| "loss": 0.065, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.3264604810996563, | |
| "grad_norm": 0.5153103372327345, | |
| "learning_rate": 3.0380376289202497e-06, | |
| "loss": 0.0501, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.7461208037303872, | |
| "learning_rate": 2.982824190050958e-06, | |
| "loss": 0.0615, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3402061855670104, | |
| "grad_norm": 0.563922704827601, | |
| "learning_rate": 2.927903003011241e-06, | |
| "loss": 0.0602, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3470790378006874, | |
| "grad_norm": 0.9583411037622643, | |
| "learning_rate": 2.8732820248752016e-06, | |
| "loss": 0.0476, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3539518900343643, | |
| "grad_norm": 0.8626488725569136, | |
| "learning_rate": 2.8189691692221627e-06, | |
| "loss": 0.0847, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3608247422680413, | |
| "grad_norm": 0.5497606016548258, | |
| "learning_rate": 2.7649723049901554e-06, | |
| "loss": 0.0603, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.3676975945017182, | |
| "grad_norm": 0.6713151538956121, | |
| "learning_rate": 2.711299255335833e-06, | |
| "loss": 0.065, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3745704467353952, | |
| "grad_norm": 0.8443130736334474, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.071, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3814432989690721, | |
| "grad_norm": 0.7540201110775319, | |
| "learning_rate": 2.6049556566862234e-06, | |
| "loss": 0.0632, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.388316151202749, | |
| "grad_norm": 0.6041359379694171, | |
| "learning_rate": 2.552300514930657e-06, | |
| "loss": 0.0593, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.395189003436426, | |
| "grad_norm": 0.71567491054432, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0827, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.402061855670103, | |
| "grad_norm": 0.5113561981438247, | |
| "learning_rate": 2.4480616892809593e-06, | |
| "loss": 0.0462, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.40893470790378, | |
| "grad_norm": 0.6046002004133768, | |
| "learning_rate": 2.396493107683488e-06, | |
| "loss": 0.0543, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4158075601374571, | |
| "grad_norm": 0.5596260557749365, | |
| "learning_rate": 2.345301726550567e-06, | |
| "loss": 0.0629, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.422680412371134, | |
| "grad_norm": 0.7291100061575513, | |
| "learning_rate": 2.2944949625757295e-06, | |
| "loss": 0.066, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.429553264604811, | |
| "grad_norm": 0.6130368802054614, | |
| "learning_rate": 2.244080176728536e-06, | |
| "loss": 0.0802, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.436426116838488, | |
| "grad_norm": 0.6563802875972382, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.0637, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.443298969072165, | |
| "grad_norm": 0.611838739795478, | |
| "learning_rate": 2.1444556982847996e-06, | |
| "loss": 0.0726, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4501718213058419, | |
| "grad_norm": 0.4722693738337201, | |
| "learning_rate": 2.095260439450526e-06, | |
| "loss": 0.051, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.4570446735395188, | |
| "grad_norm": 0.5716420160225314, | |
| "learning_rate": 2.0464860241772454e-06, | |
| "loss": 0.072, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.463917525773196, | |
| "grad_norm": 0.5553361499805398, | |
| "learning_rate": 1.998139518984409e-06, | |
| "loss": 0.0453, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.470790378006873, | |
| "grad_norm": 0.7318302865904475, | |
| "learning_rate": 1.9502279283951363e-06, | |
| "loss": 0.0702, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.47766323024055, | |
| "grad_norm": 0.5265066548304935, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.0382, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.4845360824742269, | |
| "grad_norm": 0.5480173486009389, | |
| "learning_rate": 1.8557371930582579e-06, | |
| "loss": 0.0587, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4914089347079038, | |
| "grad_norm": 0.6040597613949248, | |
| "learning_rate": 1.8091717382875723e-06, | |
| "loss": 0.0541, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.4982817869415808, | |
| "grad_norm": 0.5359810480342913, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.06, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.5051546391752577, | |
| "grad_norm": 0.531397870606965, | |
| "learning_rate": 1.7174343859719334e-06, | |
| "loss": 0.0355, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.5120274914089347, | |
| "grad_norm": 0.8031875399961053, | |
| "learning_rate": 1.6722757794891287e-06, | |
| "loss": 0.0789, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5189003436426116, | |
| "grad_norm": 0.5555062490091698, | |
| "learning_rate": 1.627599299297431e-06, | |
| "loss": 0.0462, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.5257731958762886, | |
| "grad_norm": 0.7113640676067442, | |
| "learning_rate": 1.5834114182005544e-06, | |
| "loss": 0.0781, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.5326460481099655, | |
| "grad_norm": 0.48719927766961096, | |
| "learning_rate": 1.5397185382131524e-06, | |
| "loss": 0.0666, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.5395189003436425, | |
| "grad_norm": 0.47530306151804913, | |
| "learning_rate": 1.4965269896332884e-06, | |
| "loss": 0.0306, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.5463917525773194, | |
| "grad_norm": 0.8074808440044254, | |
| "learning_rate": 1.4538430301252783e-06, | |
| "loss": 0.0814, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5532646048109966, | |
| "grad_norm": 0.6736256014926072, | |
| "learning_rate": 1.411672843813086e-06, | |
| "loss": 0.0503, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.5601374570446735, | |
| "grad_norm": 0.6747346086567719, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.0691, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.5670103092783505, | |
| "grad_norm": 0.6157431911602551, | |
| "learning_rate": 1.3288981542051844e-06, | |
| "loss": 0.0819, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.5738831615120275, | |
| "grad_norm": 0.6480535254610911, | |
| "learning_rate": 1.2883056434459506e-06, | |
| "loss": 0.0565, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.5807560137457046, | |
| "grad_norm": 0.5571895683809527, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.0468, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5876288659793816, | |
| "grad_norm": 0.6535006058396844, | |
| "learning_rate": 1.2087396947215678e-06, | |
| "loss": 0.0568, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.5945017182130585, | |
| "grad_norm": 0.5095126802390112, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0502, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.6013745704467355, | |
| "grad_norm": 0.7704991096131333, | |
| "learning_rate": 1.1313708031358183e-06, | |
| "loss": 0.0572, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.6082474226804124, | |
| "grad_norm": 0.5564450651032732, | |
| "learning_rate": 1.0935243153818437e-06, | |
| "loss": 0.0449, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.6151202749140894, | |
| "grad_norm": 0.6232916139585358, | |
| "learning_rate": 1.0562438044060846e-06, | |
| "loss": 0.0667, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6219931271477663, | |
| "grad_norm": 0.6366113284841454, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.0716, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.6288659793814433, | |
| "grad_norm": 0.45702956352644725, | |
| "learning_rate": 9.834022350599538e-07, | |
| "loss": 0.0424, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.6357388316151202, | |
| "grad_norm": 0.7193105814417473, | |
| "learning_rate": 9.47851730098856e-07, | |
| "loss": 0.0889, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.6426116838487972, | |
| "grad_norm": 0.5404686535107154, | |
| "learning_rate": 9.128883072055411e-07, | |
| "loss": 0.0557, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.6494845360824741, | |
| "grad_norm": 0.537412699687681, | |
| "learning_rate": 8.785170319396174e-07, | |
| "loss": 0.0594, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.656357388316151, | |
| "grad_norm": 0.7034720515361111, | |
| "learning_rate": 8.447428840693489e-07, | |
| "loss": 0.0686, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.663230240549828, | |
| "grad_norm": 0.5232115086991386, | |
| "learning_rate": 8.115707568501768e-07, | |
| "loss": 0.0627, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.670103092783505, | |
| "grad_norm": 0.5572695329245831, | |
| "learning_rate": 7.790054563157745e-07, | |
| "loss": 0.066, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.6769759450171822, | |
| "grad_norm": 0.6363117187522811, | |
| "learning_rate": 7.470517005817473e-07, | |
| "loss": 0.0657, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.6838487972508591, | |
| "grad_norm": 0.566310221392119, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.0574, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.690721649484536, | |
| "grad_norm": 1.0356568477840196, | |
| "learning_rate": 6.849972522982845e-07, | |
| "loss": 0.0894, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.697594501718213, | |
| "grad_norm": 0.47909121121794535, | |
| "learning_rate": 6.549055503018575e-07, | |
| "loss": 0.0537, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.7044673539518902, | |
| "grad_norm": 0.6436104576934767, | |
| "learning_rate": 6.254433729092518e-07, | |
| "loss": 0.0488, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.7113402061855671, | |
| "grad_norm": 0.6672674342926674, | |
| "learning_rate": 5.966149886503614e-07, | |
| "loss": 0.0536, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.718213058419244, | |
| "grad_norm": 0.5941998095293833, | |
| "learning_rate": 5.684245742300625e-07, | |
| "loss": 0.0487, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.725085910652921, | |
| "grad_norm": 0.7399836666359046, | |
| "learning_rate": 5.408762139230889e-07, | |
| "loss": 0.0705, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.731958762886598, | |
| "grad_norm": 0.7206171046865145, | |
| "learning_rate": 5.139738989822901e-07, | |
| "loss": 0.0666, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.738831615120275, | |
| "grad_norm": 0.611284469491028, | |
| "learning_rate": 4.877215270603752e-07, | |
| "loss": 0.049, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.745704467353952, | |
| "grad_norm": 0.4490376679924448, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.0391, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.7525773195876289, | |
| "grad_norm": 0.4883351408158052, | |
| "learning_rate": 4.371817315087845e-07, | |
| "loss": 0.0474, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.7594501718213058, | |
| "grad_norm": 0.6275538945812621, | |
| "learning_rate": 4.1290163016982855e-07, | |
| "loss": 0.0428, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.7663230240549828, | |
| "grad_norm": 0.6428361451692163, | |
| "learning_rate": 3.8928611537033424e-07, | |
| "loss": 0.0611, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.7731958762886597, | |
| "grad_norm": 0.4728018984468074, | |
| "learning_rate": 3.663386085658693e-07, | |
| "loss": 0.0417, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.7800687285223367, | |
| "grad_norm": 0.6118883122173401, | |
| "learning_rate": 3.4406243442987765e-07, | |
| "loss": 0.0706, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.7869415807560136, | |
| "grad_norm": 0.6971392659991725, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.0657, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7938144329896906, | |
| "grad_norm": 0.8003943172849397, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.0645, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.8006872852233677, | |
| "grad_norm": 0.5053347292286738, | |
| "learning_rate": 2.812936930186688e-07, | |
| "loss": 0.0469, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.8075601374570447, | |
| "grad_norm": 0.5277830702042456, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.0479, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.8144329896907216, | |
| "grad_norm": 0.5645294143913049, | |
| "learning_rate": 2.428610830928152e-07, | |
| "loss": 0.054, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.8213058419243986, | |
| "grad_norm": 0.5641631161326522, | |
| "learning_rate": 2.2467724440002336e-07, | |
| "loss": 0.057, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8281786941580758, | |
| "grad_norm": 0.5181045285085284, | |
| "learning_rate": 2.0718526251279346e-07, | |
| "loss": 0.0509, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.8350515463917527, | |
| "grad_norm": 0.5129540682526712, | |
| "learning_rate": 1.9038767169893058e-07, | |
| "loss": 0.0484, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.8419243986254297, | |
| "grad_norm": 0.5996895483521816, | |
| "learning_rate": 1.7428690562169003e-07, | |
| "loss": 0.0629, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.8487972508591066, | |
| "grad_norm": 0.5958675862118761, | |
| "learning_rate": 1.5888529698718347e-07, | |
| "loss": 0.0538, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.8556701030927836, | |
| "grad_norm": 0.5229872637425641, | |
| "learning_rate": 1.4418507720641794e-07, | |
| "loss": 0.0533, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8625429553264605, | |
| "grad_norm": 0.5553379457486358, | |
| "learning_rate": 1.3018837607199909e-07, | |
| "loss": 0.0516, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.8694158075601375, | |
| "grad_norm": 0.53365466375145, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.0498, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.8762886597938144, | |
| "grad_norm": 0.5226361091242422, | |
| "learning_rate": 1.0431353898399388e-07, | |
| "loss": 0.0542, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.8831615120274914, | |
| "grad_norm": 0.5163621708965945, | |
| "learning_rate": 9.243915182039431e-08, | |
| "loss": 0.0453, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.8900343642611683, | |
| "grad_norm": 0.4739921193694733, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.0436, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8969072164948453, | |
| "grad_norm": 0.4407754245174464, | |
| "learning_rate": 7.082504191083417e-08, | |
| "loss": 0.0534, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.9037800687285222, | |
| "grad_norm": 0.6535923265463632, | |
| "learning_rate": 6.108845065352864e-08, | |
| "loss": 0.0574, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.9106529209621992, | |
| "grad_norm": 0.6460446765064554, | |
| "learning_rate": 5.206741722181385e-08, | |
| "loss": 0.0569, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.9175257731958761, | |
| "grad_norm": 0.5963347832438871, | |
| "learning_rate": 4.376324859820924e-08, | |
| "loss": 0.0582, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.9243986254295533, | |
| "grad_norm": 0.5194981933954234, | |
| "learning_rate": 3.617714790465576e-08, | |
| "loss": 0.0512, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9312714776632303, | |
| "grad_norm": 0.4902924513256396, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.0471, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.9381443298969072, | |
| "grad_norm": 0.4965516792645092, | |
| "learning_rate": 2.3163442461766604e-08, | |
| "loss": 0.0469, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.9450171821305842, | |
| "grad_norm": 0.6432301842500561, | |
| "learning_rate": 1.7737723159999e-08, | |
| "loss": 0.0624, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.9518900343642611, | |
| "grad_norm": 0.4826349649450335, | |
| "learning_rate": 1.3033842410251074e-08, | |
| "loss": 0.0458, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.9587628865979383, | |
| "grad_norm": 0.6661201710073558, | |
| "learning_rate": 9.052481718690998e-09, | |
| "loss": 0.0539, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9656357388316152, | |
| "grad_norm": 0.5054996531431956, | |
| "learning_rate": 5.794217911562205e-09, | |
| "loss": 0.0532, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.9725085910652922, | |
| "grad_norm": 0.44671732869672487, | |
| "learning_rate": 3.2595230516152543e-09, | |
| "loss": 0.045, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.9793814432989691, | |
| "grad_norm": 0.6406418425928955, | |
| "learning_rate": 1.4487643697103092e-09, | |
| "loss": 0.0633, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.986254295532646, | |
| "grad_norm": 0.40692202273679495, | |
| "learning_rate": 3.6220421161692333e-10, | |
| "loss": 0.0363, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "grad_norm": 0.5992265549711294, | |
| "learning_rate": 0.0, | |
| "loss": 0.065, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "eval_loss": 0.06222495436668396, | |
| "eval_runtime": 28.8534, | |
| "eval_samples_per_second": 33.93, | |
| "eval_steps_per_second": 1.074, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "step": 290, | |
| "total_flos": 8.781509506919629e+16, | |
| "train_loss": 0.08320249695459317, | |
| "train_runtime": 3374.6623, | |
| "train_samples_per_second": 11.022, | |
| "train_steps_per_second": 0.086 | |
| } | |
| ], | |
| "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": 8.781509506919629e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |