| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 320, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00625, | |
| "grad_norm": 0.40599126920339407, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.508, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0125, | |
| "grad_norm": 0.16174971893093387, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.3989, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01875, | |
| "grad_norm": 0.13980982904548378, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.3526, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 0.27727799449785184, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.3927, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03125, | |
| "grad_norm": 0.1157104063128156, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 0.3232, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0375, | |
| "grad_norm": 0.14945724236967864, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 0.3486, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04375, | |
| "grad_norm": 0.1806281329991288, | |
| "learning_rate": 2.1875e-05, | |
| "loss": 0.3894, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.16431928934147372, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3606, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05625, | |
| "grad_norm": 0.15979437230246274, | |
| "learning_rate": 2.8125000000000003e-05, | |
| "loss": 0.3323, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 0.2004445206673931, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.3689, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06875, | |
| "grad_norm": 0.20835456635890684, | |
| "learning_rate": 3.4375e-05, | |
| "loss": 0.3825, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 0.18507055166360947, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.3368, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08125, | |
| "grad_norm": 0.21673005959559813, | |
| "learning_rate": 4.0625000000000005e-05, | |
| "loss": 0.3345, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0875, | |
| "grad_norm": 0.17167000281963693, | |
| "learning_rate": 4.375e-05, | |
| "loss": 0.3176, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09375, | |
| "grad_norm": 0.19654075830086185, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 0.3136, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.16056560874254397, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2986, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.10625, | |
| "grad_norm": 0.14196579272732793, | |
| "learning_rate": 5.3125000000000004e-05, | |
| "loss": 0.2691, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1125, | |
| "grad_norm": 0.15400147007847648, | |
| "learning_rate": 5.6250000000000005e-05, | |
| "loss": 0.3044, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.11875, | |
| "grad_norm": 0.10944808517977976, | |
| "learning_rate": 5.9375e-05, | |
| "loss": 0.232, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.11702109032620692, | |
| "learning_rate": 6.25e-05, | |
| "loss": 0.237, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13125, | |
| "grad_norm": 0.13599177017428848, | |
| "learning_rate": 6.562500000000001e-05, | |
| "loss": 0.2708, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1375, | |
| "grad_norm": 0.10113783542128998, | |
| "learning_rate": 6.875e-05, | |
| "loss": 0.2334, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.14375, | |
| "grad_norm": 0.07723777905374925, | |
| "learning_rate": 7.1875e-05, | |
| "loss": 0.1994, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.08703618197718174, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.2253, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.15625, | |
| "grad_norm": 0.11206045033884303, | |
| "learning_rate": 7.8125e-05, | |
| "loss": 0.2677, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1625, | |
| "grad_norm": 0.08484274071341459, | |
| "learning_rate": 8.125000000000001e-05, | |
| "loss": 0.1909, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.16875, | |
| "grad_norm": 0.13523622669439594, | |
| "learning_rate": 8.4375e-05, | |
| "loss": 0.2352, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 0.0979564733902834, | |
| "learning_rate": 8.75e-05, | |
| "loss": 0.2312, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.18125, | |
| "grad_norm": 0.10257710940833924, | |
| "learning_rate": 9.062500000000001e-05, | |
| "loss": 0.2571, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 0.09612386608372053, | |
| "learning_rate": 9.375e-05, | |
| "loss": 0.2181, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19375, | |
| "grad_norm": 0.08223867478377872, | |
| "learning_rate": 9.687500000000001e-05, | |
| "loss": 0.1768, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.08986243609911292, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1995, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.20625, | |
| "grad_norm": 0.22662504165100472, | |
| "learning_rate": 9.999702525000749e-05, | |
| "loss": 0.2344, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2125, | |
| "grad_norm": 0.11694951154367994, | |
| "learning_rate": 9.998810135399546e-05, | |
| "loss": 0.1581, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.21875, | |
| "grad_norm": 0.10469730042736397, | |
| "learning_rate": 9.997322937381829e-05, | |
| "loss": 0.2129, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 0.09098907134196822, | |
| "learning_rate": 9.99524110790929e-05, | |
| "loss": 0.1837, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.23125, | |
| "grad_norm": 0.10818262125011802, | |
| "learning_rate": 9.992564894698816e-05, | |
| "loss": 0.1881, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2375, | |
| "grad_norm": 0.09073616923032347, | |
| "learning_rate": 9.989294616193017e-05, | |
| "loss": 0.1837, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.24375, | |
| "grad_norm": 0.08876555004490387, | |
| "learning_rate": 9.985430661522333e-05, | |
| "loss": 0.1918, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.1046229015794259, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.1823, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.25625, | |
| "grad_norm": 0.10749279775724821, | |
| "learning_rate": 9.975923633360985e-05, | |
| "loss": 0.2191, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2625, | |
| "grad_norm": 0.07655181308042079, | |
| "learning_rate": 9.970281691111598e-05, | |
| "loss": 0.1671, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.26875, | |
| "grad_norm": 0.06987142581471417, | |
| "learning_rate": 9.964048335045275e-05, | |
| "loss": 0.1445, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 0.08578476236740849, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 0.1883, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.28125, | |
| "grad_norm": 0.08735769756820697, | |
| "learning_rate": 9.949810418574039e-05, | |
| "loss": 0.1912, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2875, | |
| "grad_norm": 0.08454191122193365, | |
| "learning_rate": 9.941807552338804e-05, | |
| "loss": 0.1781, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.29375, | |
| "grad_norm": 0.07561417915639863, | |
| "learning_rate": 9.933216660424395e-05, | |
| "loss": 0.1741, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.0908845677740239, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.1862, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.30625, | |
| "grad_norm": 0.08957984885995234, | |
| "learning_rate": 9.914274958326505e-05, | |
| "loss": 0.1948, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 0.06721787175218842, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.1387, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.31875, | |
| "grad_norm": 0.07776287708078782, | |
| "learning_rate": 9.892994327504693e-05, | |
| "loss": 0.1897, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 0.0792915804134436, | |
| "learning_rate": 9.881480035599667e-05, | |
| "loss": 0.1878, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.33125, | |
| "grad_norm": 0.11195588138897727, | |
| "learning_rate": 9.869384896386668e-05, | |
| "loss": 0.194, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3375, | |
| "grad_norm": 0.10744243347143223, | |
| "learning_rate": 9.856710349066307e-05, | |
| "loss": 0.1881, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.34375, | |
| "grad_norm": 0.1162711120696835, | |
| "learning_rate": 9.843457901782967e-05, | |
| "loss": 0.2129, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.08320344631618402, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.187, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.35625, | |
| "grad_norm": 0.08853925634789721, | |
| "learning_rate": 9.815225683538814e-05, | |
| "loss": 0.1952, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3625, | |
| "grad_norm": 0.08064278407146641, | |
| "learning_rate": 9.800249271929645e-05, | |
| "loss": 0.1722, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.36875, | |
| "grad_norm": 0.08627992209736293, | |
| "learning_rate": 9.784701678661045e-05, | |
| "loss": 0.1852, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.08472663945575681, | |
| "learning_rate": 9.768584753741134e-05, | |
| "loss": 0.1861, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.38125, | |
| "grad_norm": 0.0706516311545354, | |
| "learning_rate": 9.751900414922805e-05, | |
| "loss": 0.1681, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3875, | |
| "grad_norm": 0.07830536169386694, | |
| "learning_rate": 9.73465064747553e-05, | |
| "loss": 0.1768, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.39375, | |
| "grad_norm": 0.07870447814947025, | |
| "learning_rate": 9.716837503949127e-05, | |
| "loss": 0.1592, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.11475798771048969, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.2007, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.40625, | |
| "grad_norm": 0.09295937705194224, | |
| "learning_rate": 9.67952963378663e-05, | |
| "loss": 0.1499, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4125, | |
| "grad_norm": 0.09326386653332064, | |
| "learning_rate": 9.660039346413994e-05, | |
| "loss": 0.1702, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.41875, | |
| "grad_norm": 0.08131260921217849, | |
| "learning_rate": 9.639994560960923e-05, | |
| "loss": 0.172, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 0.07365872216537292, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.1363, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.43125, | |
| "grad_norm": 0.0834833164127378, | |
| "learning_rate": 9.598251102025461e-05, | |
| "loss": 0.1664, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 0.07655054737496132, | |
| "learning_rate": 9.576557395597236e-05, | |
| "loss": 0.1439, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.44375, | |
| "grad_norm": 0.09641298517325338, | |
| "learning_rate": 9.554319124605879e-05, | |
| "loss": 0.1526, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.07985464053948435, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.1514, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.45625, | |
| "grad_norm": 0.08227789931433875, | |
| "learning_rate": 9.508219537944081e-05, | |
| "loss": 0.1481, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4625, | |
| "grad_norm": 0.07992014265398506, | |
| "learning_rate": 9.484363707663442e-05, | |
| "loss": 0.1552, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.46875, | |
| "grad_norm": 0.08428886519848619, | |
| "learning_rate": 9.459974282946571e-05, | |
| "loss": 0.1533, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 0.07999237487858113, | |
| "learning_rate": 9.435054165891109e-05, | |
| "loss": 0.15, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.48125, | |
| "grad_norm": 0.08663635431835072, | |
| "learning_rate": 9.409606321741775e-05, | |
| "loss": 0.1633, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4875, | |
| "grad_norm": 0.10421286796237919, | |
| "learning_rate": 9.38363377853754e-05, | |
| "loss": 0.1675, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.49375, | |
| "grad_norm": 0.08371956698091174, | |
| "learning_rate": 9.357139626751308e-05, | |
| "loss": 0.1649, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.11576971658648634, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.1818, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.50625, | |
| "grad_norm": 0.09180205457837495, | |
| "learning_rate": 9.302599169280395e-05, | |
| "loss": 0.1535, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5125, | |
| "grad_norm": 0.11143842983351335, | |
| "learning_rate": 9.274559353364734e-05, | |
| "loss": 0.1573, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.51875, | |
| "grad_norm": 0.08382222694709042, | |
| "learning_rate": 9.246010907632895e-05, | |
| "loss": 0.1552, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.08169820823619627, | |
| "learning_rate": 9.21695722906443e-05, | |
| "loss": 0.1348, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.53125, | |
| "grad_norm": 0.09264402945807464, | |
| "learning_rate": 9.18740177475654e-05, | |
| "loss": 0.1541, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5375, | |
| "grad_norm": 0.12035173607425367, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.1584, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.54375, | |
| "grad_norm": 0.0933034733005768, | |
| "learning_rate": 9.126799665424319e-05, | |
| "loss": 0.1558, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.09983789678085002, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.178, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.55625, | |
| "grad_norm": 0.08824386686645358, | |
| "learning_rate": 9.064233422958077e-05, | |
| "loss": 0.1382, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 0.12496408760397422, | |
| "learning_rate": 9.032223021337414e-05, | |
| "loss": 0.2028, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.56875, | |
| "grad_norm": 0.08896745045700023, | |
| "learning_rate": 8.999732825500648e-05, | |
| "loss": 0.1564, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.1051267783567327, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 0.1748, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.58125, | |
| "grad_norm": 0.1068446272490391, | |
| "learning_rate": 8.933328571843084e-05, | |
| "loss": 0.166, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5875, | |
| "grad_norm": 0.09693703827165154, | |
| "learning_rate": 8.899422415464409e-05, | |
| "loss": 0.1371, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.59375, | |
| "grad_norm": 0.11397584234437574, | |
| "learning_rate": 8.865052266813685e-05, | |
| "loss": 0.1695, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.08309347621522221, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.1397, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.60625, | |
| "grad_norm": 0.09441073972154676, | |
| "learning_rate": 8.79493640623581e-05, | |
| "loss": 0.1475, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6125, | |
| "grad_norm": 0.10017885648608761, | |
| "learning_rate": 8.759199037394887e-05, | |
| "loss": 0.1455, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.61875, | |
| "grad_norm": 0.09389095525738553, | |
| "learning_rate": 8.723014361461632e-05, | |
| "loss": 0.155, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.08794088850259763, | |
| "learning_rate": 8.68638668405062e-05, | |
| "loss": 0.14, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.63125, | |
| "grad_norm": 0.08768231254617773, | |
| "learning_rate": 8.649320363489179e-05, | |
| "loss": 0.1268, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6375, | |
| "grad_norm": 0.10989824440042859, | |
| "learning_rate": 8.611819810298778e-05, | |
| "loss": 0.174, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.64375, | |
| "grad_norm": 0.09207703015396494, | |
| "learning_rate": 8.573889486670233e-05, | |
| "loss": 0.1314, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.10433836170510401, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.1563, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.65625, | |
| "grad_norm": 0.1394424066715557, | |
| "learning_rate": 8.496757632016836e-05, | |
| "loss": 0.2076, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6625, | |
| "grad_norm": 0.09605746900221901, | |
| "learning_rate": 8.457565278911348e-05, | |
| "loss": 0.1516, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.66875, | |
| "grad_norm": 0.09934008188904551, | |
| "learning_rate": 8.417961510114356e-05, | |
| "loss": 0.163, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.096551403883122, | |
| "learning_rate": 8.377951038078302e-05, | |
| "loss": 0.1351, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.68125, | |
| "grad_norm": 0.09261714349099234, | |
| "learning_rate": 8.337538623649237e-05, | |
| "loss": 0.1181, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6875, | |
| "grad_norm": 0.09869545197682032, | |
| "learning_rate": 8.296729075500344e-05, | |
| "loss": 0.1404, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.69375, | |
| "grad_norm": 0.17566063934542003, | |
| "learning_rate": 8.255527249559746e-05, | |
| "loss": 0.1746, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.11448964041640074, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.1724, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.70625, | |
| "grad_norm": 0.09940589194936707, | |
| "learning_rate": 8.171966420818228e-05, | |
| "loss": 0.1469, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7125, | |
| "grad_norm": 0.09560062944611337, | |
| "learning_rate": 8.129617360920296e-05, | |
| "loss": 0.1692, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.71875, | |
| "grad_norm": 0.11078503759259131, | |
| "learning_rate": 8.086895907853526e-05, | |
| "loss": 0.145, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.10258735833050488, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 0.1401, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.73125, | |
| "grad_norm": 0.09522206054276056, | |
| "learning_rate": 8.000356199622405e-05, | |
| "loss": 0.1289, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7375, | |
| "grad_norm": 0.13908178309980257, | |
| "learning_rate": 7.956548241817912e-05, | |
| "loss": 0.1191, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.74375, | |
| "grad_norm": 0.10074203716975204, | |
| "learning_rate": 7.912388484339012e-05, | |
| "loss": 0.1679, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.09900042130140317, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 0.1455, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.75625, | |
| "grad_norm": 0.09784130256143893, | |
| "learning_rate": 7.823034629871503e-05, | |
| "loss": 0.1387, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7625, | |
| "grad_norm": 0.10874950463645312, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1595, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.76875, | |
| "grad_norm": 0.108024736321919, | |
| "learning_rate": 7.732337163815217e-05, | |
| "loss": 0.1591, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.10774049198035004, | |
| "learning_rate": 7.68649804173412e-05, | |
| "loss": 0.1623, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.78125, | |
| "grad_norm": 0.1071299904224286, | |
| "learning_rate": 7.64033925325184e-05, | |
| "loss": 0.155, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7875, | |
| "grad_norm": 0.1000917142455837, | |
| "learning_rate": 7.593866290802608e-05, | |
| "loss": 0.1524, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.79375, | |
| "grad_norm": 0.10821955303820206, | |
| "learning_rate": 7.54708468420421e-05, | |
| "loss": 0.1599, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.09585433780364112, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.1554, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.80625, | |
| "grad_norm": 0.12861896583829888, | |
| "learning_rate": 7.45261784079654e-05, | |
| "loss": 0.1538, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "grad_norm": 0.09906883135040825, | |
| "learning_rate": 7.404943844596939e-05, | |
| "loss": 0.1525, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.81875, | |
| "grad_norm": 0.09731521778705561, | |
| "learning_rate": 7.35698368412999e-05, | |
| "loss": 0.1478, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.09827888495925427, | |
| "learning_rate": 7.308743066175172e-05, | |
| "loss": 0.1484, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.83125, | |
| "grad_norm": 0.09864917828015088, | |
| "learning_rate": 7.2602277308836e-05, | |
| "loss": 0.124, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8375, | |
| "grad_norm": 0.1002828586258035, | |
| "learning_rate": 7.211443451095007e-05, | |
| "loss": 0.1422, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.84375, | |
| "grad_norm": 0.10620289407411102, | |
| "learning_rate": 7.162396031650831e-05, | |
| "loss": 0.1488, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.09917275385305373, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 0.1227, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.85625, | |
| "grad_norm": 0.09676283023246729, | |
| "learning_rate": 7.063535149021973e-05, | |
| "loss": 0.1397, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8625, | |
| "grad_norm": 0.10405765108275118, | |
| "learning_rate": 7.013733449293687e-05, | |
| "loss": 0.1489, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.86875, | |
| "grad_norm": 0.11729246349725461, | |
| "learning_rate": 6.96369213542287e-05, | |
| "loss": 0.1625, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.11001117463075423, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1575, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.88125, | |
| "grad_norm": 0.09836686384061544, | |
| "learning_rate": 6.862914510720515e-05, | |
| "loss": 0.1367, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8875, | |
| "grad_norm": 0.09762698241197115, | |
| "learning_rate": 6.812190191418508e-05, | |
| "loss": 0.1269, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.89375, | |
| "grad_norm": 0.11873455243886127, | |
| "learning_rate": 6.761250239606169e-05, | |
| "loss": 0.1653, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.11143312549675602, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 0.163, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.90625, | |
| "grad_norm": 0.10224042656105548, | |
| "learning_rate": 6.658747708766762e-05, | |
| "loss": 0.1372, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9125, | |
| "grad_norm": 0.10537786627093436, | |
| "learning_rate": 6.607197326515808e-05, | |
| "loss": 0.1666, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.91875, | |
| "grad_norm": 0.11598573107229065, | |
| "learning_rate": 6.555455703855454e-05, | |
| "loss": 0.1608, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.09995067555095581, | |
| "learning_rate": 6.503528997521366e-05, | |
| "loss": 0.1358, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.93125, | |
| "grad_norm": 0.1013311198834551, | |
| "learning_rate": 6.451423386272312e-05, | |
| "loss": 0.1467, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 0.12955305425483832, | |
| "learning_rate": 6.399145070154961e-05, | |
| "loss": 0.1602, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.94375, | |
| "grad_norm": 0.11967014014153346, | |
| "learning_rate": 6.346700269766132e-05, | |
| "loss": 0.1491, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.09715431421285034, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 0.1503, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.95625, | |
| "grad_norm": 0.0993416996456672, | |
| "learning_rate": 6.241336196868582e-05, | |
| "loss": 0.1394, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9625, | |
| "grad_norm": 0.11282985988320719, | |
| "learning_rate": 6.188429461630866e-05, | |
| "loss": 0.1471, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.96875, | |
| "grad_norm": 0.0953828839581251, | |
| "learning_rate": 6.135381315171867e-05, | |
| "loss": 0.1163, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.10585616381949887, | |
| "learning_rate": 6.0821980696905146e-05, | |
| "loss": 0.1414, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.98125, | |
| "grad_norm": 0.10303380839923655, | |
| "learning_rate": 6.0288860534611745e-05, | |
| "loss": 0.1371, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.9875, | |
| "grad_norm": 0.10581039986719955, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1591, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.99375, | |
| "grad_norm": 0.08949422220549504, | |
| "learning_rate": 5.9219010977133173e-05, | |
| "loss": 0.1136, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.2566631074675719, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.1355, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.00625, | |
| "grad_norm": 0.10977904119859523, | |
| "learning_rate": 5.814477366972945e-05, | |
| "loss": 0.1389, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0125, | |
| "grad_norm": 0.09358588046249755, | |
| "learning_rate": 5.7606169309495836e-05, | |
| "loss": 0.1278, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.01875, | |
| "grad_norm": 0.1016639716973101, | |
| "learning_rate": 5.706665989117839e-05, | |
| "loss": 0.1266, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.025, | |
| "grad_norm": 0.10447824270529592, | |
| "learning_rate": 5.6526309611002594e-05, | |
| "loss": 0.1344, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.03125, | |
| "grad_norm": 0.09539752784620499, | |
| "learning_rate": 5.5985182765248126e-05, | |
| "loss": 0.11, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.0375, | |
| "grad_norm": 0.09870524777017345, | |
| "learning_rate": 5.544334374259823e-05, | |
| "loss": 0.1177, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.04375, | |
| "grad_norm": 0.1319083652806534, | |
| "learning_rate": 5.490085701647805e-05, | |
| "loss": 0.1519, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.11490589915219332, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 0.1388, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.05625, | |
| "grad_norm": 0.10960359392101734, | |
| "learning_rate": 5.381419872519763e-05, | |
| "loss": 0.1315, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0625, | |
| "grad_norm": 0.11911874080836313, | |
| "learning_rate": 5.327015646150716e-05, | |
| "loss": 0.1513, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.06875, | |
| "grad_norm": 0.11667323294115564, | |
| "learning_rate": 5.2725725081900325e-05, | |
| "loss": 0.1265, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.075, | |
| "grad_norm": 0.12009873927124597, | |
| "learning_rate": 5.218096936826681e-05, | |
| "loss": 0.1346, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.08125, | |
| "grad_norm": 0.1096220905597413, | |
| "learning_rate": 5.1635954141088813e-05, | |
| "loss": 0.1129, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.0875, | |
| "grad_norm": 0.10881175717833912, | |
| "learning_rate": 5.1090744251728064e-05, | |
| "loss": 0.1031, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.09375, | |
| "grad_norm": 0.13152370066828684, | |
| "learning_rate": 5.054540457470912e-05, | |
| "loss": 0.1093, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.1363760209998771, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1283, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.10625, | |
| "grad_norm": 0.11792037030369434, | |
| "learning_rate": 4.945459542529089e-05, | |
| "loss": 0.1163, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1125, | |
| "grad_norm": 0.1291497029388642, | |
| "learning_rate": 4.890925574827195e-05, | |
| "loss": 0.1188, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.11875, | |
| "grad_norm": 0.15349633438432128, | |
| "learning_rate": 4.83640458589112e-05, | |
| "loss": 0.1742, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 0.13584411062076804, | |
| "learning_rate": 4.781903063173321e-05, | |
| "loss": 0.1217, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.13125, | |
| "grad_norm": 0.11830432386637907, | |
| "learning_rate": 4.727427491809968e-05, | |
| "loss": 0.1047, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.1375, | |
| "grad_norm": 0.13406559358836506, | |
| "learning_rate": 4.6729843538492847e-05, | |
| "loss": 0.1288, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.14375, | |
| "grad_norm": 0.12419804398498213, | |
| "learning_rate": 4.618580127480238e-05, | |
| "loss": 0.1123, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.11882741893557454, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 0.1096, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.15625, | |
| "grad_norm": 0.13277371284528894, | |
| "learning_rate": 4.509914298352197e-05, | |
| "loss": 0.1237, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1625, | |
| "grad_norm": 0.13106816929984538, | |
| "learning_rate": 4.4556656257401786e-05, | |
| "loss": 0.1256, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.16875, | |
| "grad_norm": 0.10042628903496502, | |
| "learning_rate": 4.4014817234751885e-05, | |
| "loss": 0.087, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.175, | |
| "grad_norm": 0.12858065744554853, | |
| "learning_rate": 4.347369038899744e-05, | |
| "loss": 0.132, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.18125, | |
| "grad_norm": 0.13353933963282785, | |
| "learning_rate": 4.2933340108821644e-05, | |
| "loss": 0.1342, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.1875, | |
| "grad_norm": 0.12535560223690698, | |
| "learning_rate": 4.239383069050417e-05, | |
| "loss": 0.1144, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.19375, | |
| "grad_norm": 0.11919252614077346, | |
| "learning_rate": 4.185522633027057e-05, | |
| "loss": 0.11, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.12595711384481822, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.1256, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.20625, | |
| "grad_norm": 0.140189285907938, | |
| "learning_rate": 4.078098902286683e-05, | |
| "loss": 0.1387, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2125, | |
| "grad_norm": 0.188759066767972, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1154, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.21875, | |
| "grad_norm": 0.12510322910576527, | |
| "learning_rate": 3.971113946538826e-05, | |
| "loss": 0.1277, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.225, | |
| "grad_norm": 0.13577535319914885, | |
| "learning_rate": 3.917801930309486e-05, | |
| "loss": 0.1245, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.23125, | |
| "grad_norm": 0.11279041105321824, | |
| "learning_rate": 3.864618684828134e-05, | |
| "loss": 0.1149, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2375, | |
| "grad_norm": 0.11597618526997733, | |
| "learning_rate": 3.8115705383691355e-05, | |
| "loss": 0.1137, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.24375, | |
| "grad_norm": 0.21159784242538673, | |
| "learning_rate": 3.758663803131418e-05, | |
| "loss": 0.1411, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.12706926447242534, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 0.117, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 0.13555637001991272, | |
| "eval_runtime": 15.1829, | |
| "eval_samples_per_second": 0.461, | |
| "eval_steps_per_second": 0.132, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.25625, | |
| "grad_norm": 0.140712680168253, | |
| "learning_rate": 3.65329973023387e-05, | |
| "loss": 0.1317, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.2625, | |
| "grad_norm": 0.14063508556547588, | |
| "learning_rate": 3.60085492984504e-05, | |
| "loss": 0.1485, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.26875, | |
| "grad_norm": 0.12420246892648465, | |
| "learning_rate": 3.5485766137276894e-05, | |
| "loss": 0.127, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.275, | |
| "grad_norm": 0.1375293323917167, | |
| "learning_rate": 3.4964710024786354e-05, | |
| "loss": 0.1298, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.28125, | |
| "grad_norm": 0.12567976945004108, | |
| "learning_rate": 3.4445442961445464e-05, | |
| "loss": 0.1218, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.2875, | |
| "grad_norm": 0.10902064060724474, | |
| "learning_rate": 3.392802673484193e-05, | |
| "loss": 0.0911, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.29375, | |
| "grad_norm": 0.1299840428381004, | |
| "learning_rate": 3.341252291233241e-05, | |
| "loss": 0.1105, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.14686027926986012, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 0.157, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.30625, | |
| "grad_norm": 0.13310525982308075, | |
| "learning_rate": 3.2387497603938326e-05, | |
| "loss": 0.1167, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3125, | |
| "grad_norm": 0.23056279523041534, | |
| "learning_rate": 3.1878098085814924e-05, | |
| "loss": 0.1267, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.31875, | |
| "grad_norm": 0.13379086959343525, | |
| "learning_rate": 3.137085489279485e-05, | |
| "loss": 0.1405, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.325, | |
| "grad_norm": 0.16429012844656798, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.1454, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.33125, | |
| "grad_norm": 0.1472540477988046, | |
| "learning_rate": 3.0363078645771303e-05, | |
| "loss": 0.1228, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3375, | |
| "grad_norm": 0.12810467911662524, | |
| "learning_rate": 2.9862665507063147e-05, | |
| "loss": 0.1155, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.34375, | |
| "grad_norm": 0.1349885611023629, | |
| "learning_rate": 2.936464850978027e-05, | |
| "loss": 0.114, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.2226264712891022, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 0.131, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.35625, | |
| "grad_norm": 0.14284289832580171, | |
| "learning_rate": 2.8376039683491686e-05, | |
| "loss": 0.118, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.3625, | |
| "grad_norm": 0.14670574239643006, | |
| "learning_rate": 2.7885565489049946e-05, | |
| "loss": 0.1206, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.36875, | |
| "grad_norm": 0.13807081126340223, | |
| "learning_rate": 2.7397722691164018e-05, | |
| "loss": 0.1196, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.1414426520091727, | |
| "learning_rate": 2.6912569338248315e-05, | |
| "loss": 0.1257, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.38125, | |
| "grad_norm": 0.12974232596021115, | |
| "learning_rate": 2.6430163158700115e-05, | |
| "loss": 0.1078, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.3875, | |
| "grad_norm": 0.12697025279359767, | |
| "learning_rate": 2.595056155403063e-05, | |
| "loss": 0.1039, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.39375, | |
| "grad_norm": 0.1291839881974509, | |
| "learning_rate": 2.54738215920346e-05, | |
| "loss": 0.1044, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.13740214702279938, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.1152, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.40625, | |
| "grad_norm": 0.1384802768122724, | |
| "learning_rate": 2.4529153157957913e-05, | |
| "loss": 0.1234, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4125, | |
| "grad_norm": 0.13354781137733693, | |
| "learning_rate": 2.4061337091973918e-05, | |
| "loss": 0.1193, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.41875, | |
| "grad_norm": 0.1771648365806598, | |
| "learning_rate": 2.3596607467481603e-05, | |
| "loss": 0.1452, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.425, | |
| "grad_norm": 0.11638362220378876, | |
| "learning_rate": 2.3135019582658802e-05, | |
| "loss": 0.0936, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.43125, | |
| "grad_norm": 0.13555782216280093, | |
| "learning_rate": 2.2676628361847836e-05, | |
| "loss": 0.1194, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.4375, | |
| "grad_norm": 0.13796797972293184, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.1328, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.44375, | |
| "grad_norm": 0.12977028041039126, | |
| "learning_rate": 2.176965370128498e-05, | |
| "loss": 0.1079, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.13740927778792686, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 0.1225, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.45625, | |
| "grad_norm": 0.15309969236799242, | |
| "learning_rate": 2.08761151566099e-05, | |
| "loss": 0.1521, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.4625, | |
| "grad_norm": 0.16080840678597888, | |
| "learning_rate": 2.0434517581820896e-05, | |
| "loss": 0.1214, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.46875, | |
| "grad_norm": 0.1462626524834281, | |
| "learning_rate": 1.999643800377596e-05, | |
| "loss": 0.1305, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.475, | |
| "grad_norm": 0.13708479721295597, | |
| "learning_rate": 1.9561928549563968e-05, | |
| "loss": 0.1259, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.48125, | |
| "grad_norm": 0.14485439862293895, | |
| "learning_rate": 1.913104092146476e-05, | |
| "loss": 0.1313, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.4875, | |
| "grad_norm": 0.1305624984599595, | |
| "learning_rate": 1.8703826390797048e-05, | |
| "loss": 0.1082, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.49375, | |
| "grad_norm": 0.12983841773778187, | |
| "learning_rate": 1.8280335791817733e-05, | |
| "loss": 0.1254, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.1305817772014699, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 0.0976, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.50625, | |
| "grad_norm": 0.13523687056968922, | |
| "learning_rate": 1.7444727504402553e-05, | |
| "loss": 0.118, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.5125, | |
| "grad_norm": 0.13276802064818327, | |
| "learning_rate": 1.703270924499656e-05, | |
| "loss": 0.1221, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.51875, | |
| "grad_norm": 0.14813379345753966, | |
| "learning_rate": 1.662461376350764e-05, | |
| "loss": 0.1415, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.525, | |
| "grad_norm": 0.1373736965280858, | |
| "learning_rate": 1.622048961921699e-05, | |
| "loss": 0.1241, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.53125, | |
| "grad_norm": 0.12987502342407337, | |
| "learning_rate": 1.5820384898856434e-05, | |
| "loss": 0.1109, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5375, | |
| "grad_norm": 0.14097871387133398, | |
| "learning_rate": 1.5424347210886538e-05, | |
| "loss": 0.1171, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.54375, | |
| "grad_norm": 0.12911360333125718, | |
| "learning_rate": 1.5032423679831642e-05, | |
| "loss": 0.1087, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.13452146268445594, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.1228, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.55625, | |
| "grad_norm": 0.16983759281577876, | |
| "learning_rate": 1.4261105133297692e-05, | |
| "loss": 0.1217, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 0.13444059283520182, | |
| "learning_rate": 1.3881801897012225e-05, | |
| "loss": 0.1114, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.56875, | |
| "grad_norm": 0.12954454455265535, | |
| "learning_rate": 1.3506796365108232e-05, | |
| "loss": 0.1022, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.575, | |
| "grad_norm": 0.17591202784634585, | |
| "learning_rate": 1.3136133159493802e-05, | |
| "loss": 0.1114, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.58125, | |
| "grad_norm": 0.1321105098614603, | |
| "learning_rate": 1.2769856385383688e-05, | |
| "loss": 0.1183, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.5875, | |
| "grad_norm": 0.1432184606258264, | |
| "learning_rate": 1.2408009626051137e-05, | |
| "loss": 0.1152, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.59375, | |
| "grad_norm": 0.1652313311913505, | |
| "learning_rate": 1.2050635937641908e-05, | |
| "loss": 0.0925, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.1363797254860259, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 0.1283, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.60625, | |
| "grad_norm": 0.14243039726380746, | |
| "learning_rate": 1.134947733186315e-05, | |
| "loss": 0.13, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.6125, | |
| "grad_norm": 0.1407260083641188, | |
| "learning_rate": 1.100577584535592e-05, | |
| "loss": 0.1107, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.61875, | |
| "grad_norm": 0.1359874658608621, | |
| "learning_rate": 1.0666714281569151e-05, | |
| "loss": 0.1061, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 0.1619598596563152, | |
| "learning_rate": 1.0332332985438248e-05, | |
| "loss": 0.1151, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.63125, | |
| "grad_norm": 0.14589312780062294, | |
| "learning_rate": 1.000267174499352e-05, | |
| "loss": 0.1288, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6375, | |
| "grad_norm": 0.1448713969935771, | |
| "learning_rate": 9.677769786625867e-06, | |
| "loss": 0.1096, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.64375, | |
| "grad_norm": 0.15204585922412353, | |
| "learning_rate": 9.357665770419244e-06, | |
| "loss": 0.0998, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.132551181762264, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 0.1083, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.65625, | |
| "grad_norm": 0.17245754726915724, | |
| "learning_rate": 8.732003345756811e-06, | |
| "loss": 0.118, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.6625, | |
| "grad_norm": 0.13788212679850506, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.1088, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.66875, | |
| "grad_norm": 0.16823545202946685, | |
| "learning_rate": 8.125982252434611e-06, | |
| "loss": 0.1239, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.675, | |
| "grad_norm": 0.13554434267304058, | |
| "learning_rate": 7.830427709355725e-06, | |
| "loss": 0.1164, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.68125, | |
| "grad_norm": 0.12270084667138997, | |
| "learning_rate": 7.539890923671062e-06, | |
| "loss": 0.0855, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.6875, | |
| "grad_norm": 0.1610350461050514, | |
| "learning_rate": 7.2544064663526815e-06, | |
| "loss": 0.1395, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.69375, | |
| "grad_norm": 0.13371983154995107, | |
| "learning_rate": 6.974008307196056e-06, | |
| "loss": 0.1081, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.15579745002988243, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.1358, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.70625, | |
| "grad_norm": 0.13901889620059626, | |
| "learning_rate": 6.428603732486937e-06, | |
| "loss": 0.1208, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.7125, | |
| "grad_norm": 0.15932824101956905, | |
| "learning_rate": 6.163662214624616e-06, | |
| "loss": 0.1208, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.71875, | |
| "grad_norm": 0.14809539420100104, | |
| "learning_rate": 5.903936782582253e-06, | |
| "loss": 0.0961, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.725, | |
| "grad_norm": 0.14762764130305206, | |
| "learning_rate": 5.649458341088915e-06, | |
| "loss": 0.1328, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.73125, | |
| "grad_norm": 0.14698627279133403, | |
| "learning_rate": 5.400257170534295e-06, | |
| "loss": 0.1137, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.7375, | |
| "grad_norm": 0.12766172584371513, | |
| "learning_rate": 5.156362923365588e-06, | |
| "loss": 0.0963, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.74375, | |
| "grad_norm": 0.15550557920695354, | |
| "learning_rate": 4.917804620559202e-06, | |
| "loss": 0.1299, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.13371200089329044, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 0.1003, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.75625, | |
| "grad_norm": 0.19718769962541566, | |
| "learning_rate": 4.456808753941205e-06, | |
| "loss": 0.1005, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.7625, | |
| "grad_norm": 0.14609277866526107, | |
| "learning_rate": 4.234426044027645e-06, | |
| "loss": 0.1184, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.76875, | |
| "grad_norm": 0.13846373587108526, | |
| "learning_rate": 4.017488979745387e-06, | |
| "loss": 0.1083, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.775, | |
| "grad_norm": 0.16152586345422645, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.1349, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.78125, | |
| "grad_norm": 0.15220108167136218, | |
| "learning_rate": 3.600054390390778e-06, | |
| "loss": 0.1189, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.7875, | |
| "grad_norm": 0.14825944302091063, | |
| "learning_rate": 3.3996065358600782e-06, | |
| "loss": 0.1124, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.79375, | |
| "grad_norm": 0.17101334397597814, | |
| "learning_rate": 3.2047036621337236e-06, | |
| "loss": 0.1242, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.15468868193258223, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.1269, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.80625, | |
| "grad_norm": 0.13742574401888047, | |
| "learning_rate": 2.8316249605087386e-06, | |
| "loss": 0.108, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.8125, | |
| "grad_norm": 0.12723180285999527, | |
| "learning_rate": 2.653493525244721e-06, | |
| "loss": 0.104, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.81875, | |
| "grad_norm": 0.15299833624657141, | |
| "learning_rate": 2.4809958507719444e-06, | |
| "loss": 0.1227, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.825, | |
| "grad_norm": 0.153187221540368, | |
| "learning_rate": 2.314152462588659e-06, | |
| "loss": 0.1122, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.83125, | |
| "grad_norm": 0.1566217766687287, | |
| "learning_rate": 2.152983213389559e-06, | |
| "loss": 0.1206, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.8375, | |
| "grad_norm": 0.1619426370013769, | |
| "learning_rate": 1.99750728070357e-06, | |
| "loss": 0.0849, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.84375, | |
| "grad_norm": 0.1440771616743465, | |
| "learning_rate": 1.8477431646118648e-06, | |
| "loss": 0.1168, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.1266403405941791, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 0.1002, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.85625, | |
| "grad_norm": 0.14490235679395352, | |
| "learning_rate": 1.565420982170346e-06, | |
| "loss": 0.1204, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.8625, | |
| "grad_norm": 0.14214631954361082, | |
| "learning_rate": 1.4328965093369283e-06, | |
| "loss": 0.1052, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.86875, | |
| "grad_norm": 0.15125342239263603, | |
| "learning_rate": 1.3061510361333185e-06, | |
| "loss": 0.1185, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 0.14873736163107223, | |
| "learning_rate": 1.1851996440033319e-06, | |
| "loss": 0.1372, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.88125, | |
| "grad_norm": 0.15823839872382495, | |
| "learning_rate": 1.0700567249530834e-06, | |
| "loss": 0.1374, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.8875, | |
| "grad_norm": 0.1477147886939706, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.1265, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.89375, | |
| "grad_norm": 0.149870799299156, | |
| "learning_rate": 8.572504167349449e-07, | |
| "loss": 0.1156, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.13907104929191907, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 0.1131, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.90625, | |
| "grad_norm": 0.1438881017950769, | |
| "learning_rate": 6.678333957560512e-07, | |
| "loss": 0.1212, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.9125, | |
| "grad_norm": 0.14174173341102103, | |
| "learning_rate": 5.81924476611967e-07, | |
| "loss": 0.1249, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.91875, | |
| "grad_norm": 0.1409441874264998, | |
| "learning_rate": 5.018958142596065e-07, | |
| "loss": 0.1123, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.925, | |
| "grad_norm": 0.16658637594296666, | |
| "learning_rate": 4.277569313094809e-07, | |
| "loss": 0.1017, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.93125, | |
| "grad_norm": 0.1767438653947937, | |
| "learning_rate": 3.59516649547248e-07, | |
| "loss": 0.1265, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.9375, | |
| "grad_norm": 0.15957908409005475, | |
| "learning_rate": 2.971830888840177e-07, | |
| "loss": 0.139, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.94375, | |
| "grad_norm": 0.13358713671440797, | |
| "learning_rate": 2.407636663901591e-07, | |
| "loss": 0.1015, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.14596552013446734, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 0.1294, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.95625, | |
| "grad_norm": 0.15582989274250142, | |
| "learning_rate": 1.4569338477666838e-07, | |
| "loss": 0.1282, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.9625, | |
| "grad_norm": 0.14404601592111216, | |
| "learning_rate": 1.0705383806982606e-07, | |
| "loss": 0.1103, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.96875, | |
| "grad_norm": 0.1591129940081992, | |
| "learning_rate": 7.43510530118452e-08, | |
| "loss": 0.1114, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.975, | |
| "grad_norm": 0.14377087428625243, | |
| "learning_rate": 4.7588920907110094e-08, | |
| "loss": 0.1202, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.98125, | |
| "grad_norm": 0.14890980505746196, | |
| "learning_rate": 2.6770626181715773e-08, | |
| "loss": 0.1172, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.9875, | |
| "grad_norm": 0.1536343383847314, | |
| "learning_rate": 1.189864600454338e-08, | |
| "loss": 0.1239, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.99375, | |
| "grad_norm": 0.15806075985905502, | |
| "learning_rate": 2.974749992512571e-09, | |
| "loss": 0.1275, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.14605676768659354, | |
| "learning_rate": 0.0, | |
| "loss": 0.1243, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 320, | |
| "total_flos": 1567061767618560.0, | |
| "train_loss": 0.15366401586215944, | |
| "train_runtime": 5710.924, | |
| "train_samples_per_second": 0.224, | |
| "train_steps_per_second": 0.056 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 320, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1567061767618560.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |