| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.821256038647343, | |
| "eval_steps": 500, | |
| "global_step": 170, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004830917874396135, | |
| "grad_norm": 0.7645118236541748, | |
| "learning_rate": 5e-06, | |
| "loss": 1.259, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00966183574879227, | |
| "grad_norm": 0.9345910549163818, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4787, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.014492753623188406, | |
| "grad_norm": 0.9917318224906921, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.5453, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01932367149758454, | |
| "grad_norm": 1.0239824056625366, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5964, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.024154589371980676, | |
| "grad_norm": 0.9726951718330383, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.5687, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.028985507246376812, | |
| "grad_norm": 0.7599917650222778, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5249, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.033816425120772944, | |
| "grad_norm": 0.5268093347549438, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.4637, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.03864734299516908, | |
| "grad_norm": 0.5739946365356445, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4514, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.043478260869565216, | |
| "grad_norm": 0.6630675792694092, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.442, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04830917874396135, | |
| "grad_norm": 0.5699703097343445, | |
| "learning_rate": 5e-05, | |
| "loss": 1.4091, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05314009661835749, | |
| "grad_norm": 0.4952673017978668, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.3877, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.057971014492753624, | |
| "grad_norm": 0.5180989503860474, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3707, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06280193236714976, | |
| "grad_norm": 0.41192442178726196, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.3599, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.06763285024154589, | |
| "grad_norm": 0.28801196813583374, | |
| "learning_rate": 7e-05, | |
| "loss": 1.3484, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.07246376811594203, | |
| "grad_norm": 0.2618640959262848, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.3382, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.07729468599033816, | |
| "grad_norm": 0.2657703161239624, | |
| "learning_rate": 8e-05, | |
| "loss": 1.335, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0821256038647343, | |
| "grad_norm": 0.2432931512594223, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.323, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 0.24172987043857574, | |
| "learning_rate": 9e-05, | |
| "loss": 1.3317, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.09178743961352658, | |
| "grad_norm": 0.26086804270744324, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.3163, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 0.2007642686367035, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2877, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.10144927536231885, | |
| "grad_norm": 0.2327784299850464, | |
| "learning_rate": 9.946524064171123e-05, | |
| "loss": 1.2899, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.10628019323671498, | |
| "grad_norm": 0.20648740231990814, | |
| "learning_rate": 9.893048128342246e-05, | |
| "loss": 1.287, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.22094646096229553, | |
| "learning_rate": 9.83957219251337e-05, | |
| "loss": 1.2873, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.11594202898550725, | |
| "grad_norm": 0.18131175637245178, | |
| "learning_rate": 9.786096256684493e-05, | |
| "loss": 1.2594, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.12077294685990338, | |
| "grad_norm": 0.16657911241054535, | |
| "learning_rate": 9.732620320855615e-05, | |
| "loss": 1.2639, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.12560386473429952, | |
| "grad_norm": 0.1740303933620453, | |
| "learning_rate": 9.679144385026739e-05, | |
| "loss": 1.2603, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.13043478260869565, | |
| "grad_norm": 0.15640808641910553, | |
| "learning_rate": 9.625668449197861e-05, | |
| "loss": 1.274, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.13526570048309178, | |
| "grad_norm": 0.16680403053760529, | |
| "learning_rate": 9.572192513368984e-05, | |
| "loss": 1.2724, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.14009661835748793, | |
| "grad_norm": 0.15997755527496338, | |
| "learning_rate": 9.518716577540108e-05, | |
| "loss": 1.2672, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.14492753623188406, | |
| "grad_norm": 0.15305301547050476, | |
| "learning_rate": 9.46524064171123e-05, | |
| "loss": 1.2718, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1497584541062802, | |
| "grad_norm": 0.14839769899845123, | |
| "learning_rate": 9.411764705882353e-05, | |
| "loss": 1.2707, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.15458937198067632, | |
| "grad_norm": 0.14878958463668823, | |
| "learning_rate": 9.358288770053476e-05, | |
| "loss": 1.2648, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.15942028985507245, | |
| "grad_norm": 0.17154482007026672, | |
| "learning_rate": 9.3048128342246e-05, | |
| "loss": 1.2641, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1642512077294686, | |
| "grad_norm": 0.1447138488292694, | |
| "learning_rate": 9.251336898395723e-05, | |
| "loss": 1.27, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.16908212560386474, | |
| "grad_norm": 0.1631896197795868, | |
| "learning_rate": 9.197860962566846e-05, | |
| "loss": 1.276, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.14892889559268951, | |
| "learning_rate": 9.144385026737968e-05, | |
| "loss": 1.2747, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.178743961352657, | |
| "grad_norm": 0.1588708907365799, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 1.276, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.18357487922705315, | |
| "grad_norm": 0.151743546128273, | |
| "learning_rate": 9.037433155080214e-05, | |
| "loss": 1.2788, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.18840579710144928, | |
| "grad_norm": 0.15703994035720825, | |
| "learning_rate": 8.983957219251337e-05, | |
| "loss": 1.2936, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 0.1660437434911728, | |
| "learning_rate": 8.930481283422461e-05, | |
| "loss": 1.2824, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.19806763285024154, | |
| "grad_norm": 0.15268553793430328, | |
| "learning_rate": 8.877005347593583e-05, | |
| "loss": 1.3056, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2028985507246377, | |
| "grad_norm": 0.1577601134777069, | |
| "learning_rate": 8.823529411764706e-05, | |
| "loss": 1.3142, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.20772946859903382, | |
| "grad_norm": 0.16757714748382568, | |
| "learning_rate": 8.770053475935829e-05, | |
| "loss": 1.3389, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.21256038647342995, | |
| "grad_norm": 0.1712018847465515, | |
| "learning_rate": 8.716577540106952e-05, | |
| "loss": 1.3437, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 0.1829441487789154, | |
| "learning_rate": 8.663101604278076e-05, | |
| "loss": 1.358, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.20615732669830322, | |
| "learning_rate": 8.609625668449198e-05, | |
| "loss": 1.3878, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.22705314009661837, | |
| "grad_norm": 0.23940807580947876, | |
| "learning_rate": 8.556149732620321e-05, | |
| "loss": 1.4699, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2318840579710145, | |
| "grad_norm": 0.41468575596809387, | |
| "learning_rate": 8.502673796791443e-05, | |
| "loss": 1.4893, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.23671497584541062, | |
| "grad_norm": 0.5656126737594604, | |
| "learning_rate": 8.449197860962568e-05, | |
| "loss": 1.4967, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.24154589371980675, | |
| "grad_norm": 4.2125325202941895, | |
| "learning_rate": 8.39572192513369e-05, | |
| "loss": 1.6295, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2463768115942029, | |
| "grad_norm": 1.100631833076477, | |
| "learning_rate": 8.342245989304814e-05, | |
| "loss": 1.0523, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.25120772946859904, | |
| "grad_norm": 0.44898825883865356, | |
| "learning_rate": 8.288770053475936e-05, | |
| "loss": 1.106, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2560386473429952, | |
| "grad_norm": 0.2860921025276184, | |
| "learning_rate": 8.23529411764706e-05, | |
| "loss": 1.1398, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 0.28824105858802795, | |
| "learning_rate": 8.181818181818183e-05, | |
| "loss": 1.1547, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.26570048309178745, | |
| "grad_norm": 0.32123416662216187, | |
| "learning_rate": 8.128342245989305e-05, | |
| "loss": 1.1646, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.27053140096618356, | |
| "grad_norm": 0.2752850353717804, | |
| "learning_rate": 8.074866310160429e-05, | |
| "loss": 1.1752, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2753623188405797, | |
| "grad_norm": 0.22371803224086761, | |
| "learning_rate": 8.021390374331551e-05, | |
| "loss": 1.1934, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.28019323671497587, | |
| "grad_norm": 0.23126192390918732, | |
| "learning_rate": 7.967914438502674e-05, | |
| "loss": 1.2057, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.28502415458937197, | |
| "grad_norm": 0.24694480001926422, | |
| "learning_rate": 7.914438502673798e-05, | |
| "loss": 1.1858, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 0.20105589926242828, | |
| "learning_rate": 7.86096256684492e-05, | |
| "loss": 1.202, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2946859903381642, | |
| "grad_norm": 0.15975232422351837, | |
| "learning_rate": 7.807486631016043e-05, | |
| "loss": 1.2014, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2995169082125604, | |
| "grad_norm": 0.17269295454025269, | |
| "learning_rate": 7.754010695187165e-05, | |
| "loss": 1.1878, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.30434782608695654, | |
| "grad_norm": 0.1990584284067154, | |
| "learning_rate": 7.700534759358289e-05, | |
| "loss": 1.1951, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.30917874396135264, | |
| "grad_norm": 0.18062998354434967, | |
| "learning_rate": 7.647058823529411e-05, | |
| "loss": 1.1978, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.3140096618357488, | |
| "grad_norm": 0.15606802701950073, | |
| "learning_rate": 7.593582887700536e-05, | |
| "loss": 1.2153, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.3188405797101449, | |
| "grad_norm": 0.1434660404920578, | |
| "learning_rate": 7.540106951871658e-05, | |
| "loss": 1.2074, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.32367149758454106, | |
| "grad_norm": 0.1473468840122223, | |
| "learning_rate": 7.486631016042782e-05, | |
| "loss": 1.1962, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.3285024154589372, | |
| "grad_norm": 0.1452961415052414, | |
| "learning_rate": 7.433155080213904e-05, | |
| "loss": 1.209, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.14672888815402985, | |
| "learning_rate": 7.379679144385027e-05, | |
| "loss": 1.2126, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.33816425120772947, | |
| "grad_norm": 0.14124266803264618, | |
| "learning_rate": 7.326203208556151e-05, | |
| "loss": 1.2111, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.34299516908212563, | |
| "grad_norm": 0.13139352202415466, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 1.216, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.1385214626789093, | |
| "learning_rate": 7.219251336898396e-05, | |
| "loss": 1.2199, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3526570048309179, | |
| "grad_norm": 0.1288975179195404, | |
| "learning_rate": 7.165775401069518e-05, | |
| "loss": 1.201, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.357487922705314, | |
| "grad_norm": 0.13003361225128174, | |
| "learning_rate": 7.112299465240642e-05, | |
| "loss": 1.2186, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.36231884057971014, | |
| "grad_norm": 0.13762855529785156, | |
| "learning_rate": 7.058823529411765e-05, | |
| "loss": 1.2209, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3671497584541063, | |
| "grad_norm": 0.13935087621212006, | |
| "learning_rate": 7.005347593582889e-05, | |
| "loss": 1.2219, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3719806763285024, | |
| "grad_norm": 0.13384683430194855, | |
| "learning_rate": 6.951871657754011e-05, | |
| "loss": 1.2419, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.37681159420289856, | |
| "grad_norm": 0.12453139573335648, | |
| "learning_rate": 6.898395721925133e-05, | |
| "loss": 1.2154, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.38164251207729466, | |
| "grad_norm": 0.13903535902500153, | |
| "learning_rate": 6.844919786096257e-05, | |
| "loss": 1.2378, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 0.13833968341350555, | |
| "learning_rate": 6.79144385026738e-05, | |
| "loss": 1.2244, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.391304347826087, | |
| "grad_norm": 0.13052114844322205, | |
| "learning_rate": 6.737967914438504e-05, | |
| "loss": 1.226, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3961352657004831, | |
| "grad_norm": 0.13437196612358093, | |
| "learning_rate": 6.684491978609626e-05, | |
| "loss": 1.2457, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.40096618357487923, | |
| "grad_norm": 0.13693881034851074, | |
| "learning_rate": 6.631016042780749e-05, | |
| "loss": 1.2384, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4057971014492754, | |
| "grad_norm": 0.13426737487316132, | |
| "learning_rate": 6.577540106951871e-05, | |
| "loss": 1.2427, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4106280193236715, | |
| "grad_norm": 0.13844414055347443, | |
| "learning_rate": 6.524064171122995e-05, | |
| "loss": 1.2661, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.41545893719806765, | |
| "grad_norm": 0.13957887887954712, | |
| "learning_rate": 6.470588235294118e-05, | |
| "loss": 1.2696, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.42028985507246375, | |
| "grad_norm": 0.13465899229049683, | |
| "learning_rate": 6.41711229946524e-05, | |
| "loss": 1.2591, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4251207729468599, | |
| "grad_norm": 0.1441555917263031, | |
| "learning_rate": 6.363636363636364e-05, | |
| "loss": 1.2766, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.42995169082125606, | |
| "grad_norm": 0.1500505656003952, | |
| "learning_rate": 6.310160427807486e-05, | |
| "loss": 1.2887, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 0.15137352049350739, | |
| "learning_rate": 6.25668449197861e-05, | |
| "loss": 1.2792, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4396135265700483, | |
| "grad_norm": 0.15071454644203186, | |
| "learning_rate": 6.203208556149733e-05, | |
| "loss": 1.2876, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.1570783108472824, | |
| "learning_rate": 6.149732620320857e-05, | |
| "loss": 1.3223, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.4492753623188406, | |
| "grad_norm": 0.16483648121356964, | |
| "learning_rate": 6.096256684491979e-05, | |
| "loss": 1.3285, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.45410628019323673, | |
| "grad_norm": 0.1700102537870407, | |
| "learning_rate": 6.0427807486631016e-05, | |
| "loss": 1.3349, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.45893719806763283, | |
| "grad_norm": 0.1778935194015503, | |
| "learning_rate": 5.9893048128342244e-05, | |
| "loss": 1.3644, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.463768115942029, | |
| "grad_norm": 0.19471201300621033, | |
| "learning_rate": 5.9358288770053486e-05, | |
| "loss": 1.4028, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.46859903381642515, | |
| "grad_norm": 0.25646305084228516, | |
| "learning_rate": 5.882352941176471e-05, | |
| "loss": 1.4694, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.47342995169082125, | |
| "grad_norm": 0.277474045753479, | |
| "learning_rate": 5.8288770053475936e-05, | |
| "loss": 1.5052, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4782608695652174, | |
| "grad_norm": 0.33792170882225037, | |
| "learning_rate": 5.7754010695187164e-05, | |
| "loss": 1.5414, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 0.6432341933250427, | |
| "learning_rate": 5.721925133689839e-05, | |
| "loss": 1.6046, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.48792270531400966, | |
| "grad_norm": 0.5846565365791321, | |
| "learning_rate": 5.6684491978609634e-05, | |
| "loss": 0.9549, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.4927536231884058, | |
| "grad_norm": 0.49914559721946716, | |
| "learning_rate": 5.614973262032086e-05, | |
| "loss": 1.0978, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.4975845410628019, | |
| "grad_norm": 0.33365777134895325, | |
| "learning_rate": 5.561497326203209e-05, | |
| "loss": 1.1242, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5024154589371981, | |
| "grad_norm": 0.7763075828552246, | |
| "learning_rate": 5.508021390374332e-05, | |
| "loss": 1.1525, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5072463768115942, | |
| "grad_norm": 0.2733679413795471, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 1.1607, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5120772946859904, | |
| "grad_norm": 0.26070019602775574, | |
| "learning_rate": 5.401069518716578e-05, | |
| "loss": 1.1609, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5169082125603864, | |
| "grad_norm": 0.2708110809326172, | |
| "learning_rate": 5.347593582887701e-05, | |
| "loss": 1.1815, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.26362475752830505, | |
| "learning_rate": 5.294117647058824e-05, | |
| "loss": 1.1943, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.5265700483091788, | |
| "grad_norm": 0.21209686994552612, | |
| "learning_rate": 5.2406417112299466e-05, | |
| "loss": 1.1944, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.5314009661835749, | |
| "grad_norm": 0.21895644068717957, | |
| "learning_rate": 5.1871657754010694e-05, | |
| "loss": 1.192, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5362318840579711, | |
| "grad_norm": 0.20762521028518677, | |
| "learning_rate": 5.1336898395721935e-05, | |
| "loss": 1.1987, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.5410628019323671, | |
| "grad_norm": 0.18395845592021942, | |
| "learning_rate": 5.0802139037433164e-05, | |
| "loss": 1.1925, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.5458937198067633, | |
| "grad_norm": 0.1653558313846588, | |
| "learning_rate": 5.026737967914439e-05, | |
| "loss": 1.1961, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.5507246376811594, | |
| "grad_norm": 0.1628160923719406, | |
| "learning_rate": 4.973262032085561e-05, | |
| "loss": 1.1993, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.17022235691547394, | |
| "learning_rate": 4.919786096256685e-05, | |
| "loss": 1.2101, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5603864734299517, | |
| "grad_norm": 0.1710771918296814, | |
| "learning_rate": 4.8663101604278076e-05, | |
| "loss": 1.2059, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.5652173913043478, | |
| "grad_norm": 0.17160499095916748, | |
| "learning_rate": 4.8128342245989304e-05, | |
| "loss": 1.206, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.5700483091787439, | |
| "grad_norm": 0.15871083736419678, | |
| "learning_rate": 4.759358288770054e-05, | |
| "loss": 1.2067, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.5748792270531401, | |
| "grad_norm": 0.14880803227424622, | |
| "learning_rate": 4.705882352941177e-05, | |
| "loss": 1.218, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 0.14413942396640778, | |
| "learning_rate": 4.6524064171123e-05, | |
| "loss": 1.1958, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5845410628019324, | |
| "grad_norm": 0.14965958893299103, | |
| "learning_rate": 4.598930481283423e-05, | |
| "loss": 1.2206, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.5893719806763285, | |
| "grad_norm": 0.14546802639961243, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 1.2107, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5942028985507246, | |
| "grad_norm": 0.14043672382831573, | |
| "learning_rate": 4.491978609625669e-05, | |
| "loss": 1.2059, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5990338164251208, | |
| "grad_norm": 0.1403893083333969, | |
| "learning_rate": 4.4385026737967915e-05, | |
| "loss": 1.2327, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6038647342995169, | |
| "grad_norm": 0.13266952335834503, | |
| "learning_rate": 4.385026737967914e-05, | |
| "loss": 1.194, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 0.1347023993730545, | |
| "learning_rate": 4.331550802139038e-05, | |
| "loss": 1.1951, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.6135265700483091, | |
| "grad_norm": 0.13116984069347382, | |
| "learning_rate": 4.2780748663101606e-05, | |
| "loss": 1.2206, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.6183574879227053, | |
| "grad_norm": 0.14027127623558044, | |
| "learning_rate": 4.224598930481284e-05, | |
| "loss": 1.2304, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.6231884057971014, | |
| "grad_norm": 0.13990454375743866, | |
| "learning_rate": 4.171122994652407e-05, | |
| "loss": 1.2229, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.6280193236714976, | |
| "grad_norm": 0.13515028357505798, | |
| "learning_rate": 4.11764705882353e-05, | |
| "loss": 1.2191, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6328502415458938, | |
| "grad_norm": 0.1329352706670761, | |
| "learning_rate": 4.0641711229946525e-05, | |
| "loss": 1.2308, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.6376811594202898, | |
| "grad_norm": 0.13061358034610748, | |
| "learning_rate": 4.0106951871657754e-05, | |
| "loss": 1.2446, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.642512077294686, | |
| "grad_norm": 0.13514551520347595, | |
| "learning_rate": 3.957219251336899e-05, | |
| "loss": 1.2425, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.6473429951690821, | |
| "grad_norm": 0.13962285220623016, | |
| "learning_rate": 3.903743315508022e-05, | |
| "loss": 1.2574, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 0.14081381261348724, | |
| "learning_rate": 3.8502673796791445e-05, | |
| "loss": 1.247, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.6570048309178744, | |
| "grad_norm": 0.1396479606628418, | |
| "learning_rate": 3.796791443850268e-05, | |
| "loss": 1.2379, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.6618357487922706, | |
| "grad_norm": 0.13990682363510132, | |
| "learning_rate": 3.743315508021391e-05, | |
| "loss": 1.2637, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.1405869573354721, | |
| "learning_rate": 3.6898395721925136e-05, | |
| "loss": 1.2458, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.6714975845410628, | |
| "grad_norm": 0.14569957554340363, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.2755, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 0.14868015050888062, | |
| "learning_rate": 3.582887700534759e-05, | |
| "loss": 1.2836, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6811594202898551, | |
| "grad_norm": 0.1531868726015091, | |
| "learning_rate": 3.529411764705883e-05, | |
| "loss": 1.295, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.6859903381642513, | |
| "grad_norm": 0.16108393669128418, | |
| "learning_rate": 3.4759358288770055e-05, | |
| "loss": 1.3239, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.6908212560386473, | |
| "grad_norm": 0.1609143316745758, | |
| "learning_rate": 3.4224598930481284e-05, | |
| "loss": 1.3301, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.16705213487148285, | |
| "learning_rate": 3.368983957219252e-05, | |
| "loss": 1.3484, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.7004830917874396, | |
| "grad_norm": 0.18058659136295319, | |
| "learning_rate": 3.3155080213903747e-05, | |
| "loss": 1.3794, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7053140096618358, | |
| "grad_norm": 0.20221418142318726, | |
| "learning_rate": 3.2620320855614975e-05, | |
| "loss": 1.4019, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.7101449275362319, | |
| "grad_norm": 0.24968379735946655, | |
| "learning_rate": 3.20855614973262e-05, | |
| "loss": 1.4674, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.714975845410628, | |
| "grad_norm": 0.3043461740016937, | |
| "learning_rate": 3.155080213903743e-05, | |
| "loss": 1.4972, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.7198067632850241, | |
| "grad_norm": 0.33808770775794983, | |
| "learning_rate": 3.1016042780748666e-05, | |
| "loss": 1.5237, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.7246376811594203, | |
| "grad_norm": 0.5125290155410767, | |
| "learning_rate": 3.0481283422459894e-05, | |
| "loss": 1.5533, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7294685990338164, | |
| "grad_norm": 0.25133025646209717, | |
| "learning_rate": 2.9946524064171122e-05, | |
| "loss": 0.8978, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.7342995169082126, | |
| "grad_norm": 0.30135515332221985, | |
| "learning_rate": 2.9411764705882354e-05, | |
| "loss": 1.0278, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.7391304347826086, | |
| "grad_norm": 0.2961145043373108, | |
| "learning_rate": 2.8877005347593582e-05, | |
| "loss": 1.1024, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.7439613526570048, | |
| "grad_norm": 0.273294597864151, | |
| "learning_rate": 2.8342245989304817e-05, | |
| "loss": 1.1373, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.748792270531401, | |
| "grad_norm": 0.23799936473369598, | |
| "learning_rate": 2.7807486631016045e-05, | |
| "loss": 1.1419, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.7536231884057971, | |
| "grad_norm": 0.21061824262142181, | |
| "learning_rate": 2.7272727272727273e-05, | |
| "loss": 1.1512, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.7584541062801933, | |
| "grad_norm": 0.21470795571804047, | |
| "learning_rate": 2.6737967914438505e-05, | |
| "loss": 1.1616, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.7632850241545893, | |
| "grad_norm": 0.21231332421302795, | |
| "learning_rate": 2.6203208556149733e-05, | |
| "loss": 1.1662, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.7681159420289855, | |
| "grad_norm": 0.20699279010295868, | |
| "learning_rate": 2.5668449197860968e-05, | |
| "loss": 1.1716, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 0.20150579512119293, | |
| "learning_rate": 2.5133689839572196e-05, | |
| "loss": 1.1836, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.1880892962217331, | |
| "learning_rate": 2.4598930481283424e-05, | |
| "loss": 1.1794, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 0.17840184271335602, | |
| "learning_rate": 2.4064171122994652e-05, | |
| "loss": 1.1757, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.7874396135265701, | |
| "grad_norm": 0.18409621715545654, | |
| "learning_rate": 2.3529411764705884e-05, | |
| "loss": 1.1935, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.7922705314009661, | |
| "grad_norm": 0.18802137672901154, | |
| "learning_rate": 2.2994652406417115e-05, | |
| "loss": 1.1795, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.7971014492753623, | |
| "grad_norm": 0.1941538006067276, | |
| "learning_rate": 2.2459893048128343e-05, | |
| "loss": 1.2069, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8019323671497585, | |
| "grad_norm": 0.18578243255615234, | |
| "learning_rate": 2.192513368983957e-05, | |
| "loss": 1.1965, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.8067632850241546, | |
| "grad_norm": 0.17622320353984833, | |
| "learning_rate": 2.1390374331550803e-05, | |
| "loss": 1.1839, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.8115942028985508, | |
| "grad_norm": 0.16080059111118317, | |
| "learning_rate": 2.0855614973262035e-05, | |
| "loss": 1.1926, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.8164251207729468, | |
| "grad_norm": 0.14835765957832336, | |
| "learning_rate": 2.0320855614973263e-05, | |
| "loss": 1.1958, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.821256038647343, | |
| "grad_norm": 0.14560697972774506, | |
| "learning_rate": 1.9786096256684494e-05, | |
| "loss": 1.2047, | |
| "step": 170 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 207, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.493930894749139e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |