diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,11234 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.654315221339976, + "eval_steps": 500, + "global_step": 1600, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0016606985313197364, + "grad_norm": 13.191475868225098, + "learning_rate": 0.0, + "loss": 0.5561, + "step": 1 + }, + { + "epoch": 0.003321397062639473, + "grad_norm": 24.887184143066406, + "learning_rate": 1.818181818181818e-06, + "loss": 0.5758, + "step": 2 + }, + { + "epoch": 0.004982095593959209, + "grad_norm": 14.736196517944336, + "learning_rate": 3.636363636363636e-06, + "loss": 0.5803, + "step": 3 + }, + { + "epoch": 0.006642794125278946, + "grad_norm": 21.936140060424805, + "learning_rate": 5.4545454545454545e-06, + "loss": 0.6029, + "step": 4 + }, + { + "epoch": 0.008303492656598683, + "grad_norm": 28.519376754760742, + "learning_rate": 7.272727272727272e-06, + "loss": 0.5618, + "step": 5 + }, + { + "epoch": 0.009964191187918418, + "grad_norm": 40.16019821166992, + "learning_rate": 9.090909090909091e-06, + "loss": 0.579, + "step": 6 + }, + { + "epoch": 0.011624889719238155, + "grad_norm": 35.93324279785156, + "learning_rate": 1.0909090909090909e-05, + "loss": 0.5292, + "step": 7 + }, + { + "epoch": 0.013285588250557891, + "grad_norm": 6.860420227050781, + "learning_rate": 1.2727272727272727e-05, + "loss": 0.5657, + "step": 8 + }, + { + "epoch": 0.014946286781877626, + "grad_norm": 39.96791076660156, + "learning_rate": 1.4545454545454545e-05, + "loss": 0.5369, + "step": 9 + }, + { + "epoch": 0.016606985313197365, + "grad_norm": 21.117414474487305, + "learning_rate": 1.6363636363636366e-05, + "loss": 0.5191, + "step": 10 + }, + { + "epoch": 0.0182676838445171, + "grad_norm": 14.313277244567871, + "learning_rate": 1.8181818181818182e-05, + "loss": 0.5064, + "step": 11 + }, + { + "epoch": 0.019928382375836835, + "grad_norm": 19.6678524017334, + "learning_rate": 2e-05, + "loss": 0.4869, + "step": 12 + }, + { + "epoch": 0.021589080907156572, + "grad_norm": 14.446533203125, + "learning_rate": 2.1818181818181818e-05, + "loss": 0.4978, + "step": 13 + }, + { + "epoch": 0.02324977943847631, + "grad_norm": 8.615384101867676, + "learning_rate": 2.3636363636363637e-05, + "loss": 0.4631, + "step": 14 + }, + { + "epoch": 0.024910477969796046, + "grad_norm": 10.880105018615723, + "learning_rate": 2.5454545454545454e-05, + "loss": 0.4866, + "step": 15 + }, + { + "epoch": 0.026571176501115783, + "grad_norm": 78.6583251953125, + "learning_rate": 2.7272727272727273e-05, + "loss": 0.4749, + "step": 16 + }, + { + "epoch": 0.02823187503243552, + "grad_norm": 5.913409233093262, + "learning_rate": 2.909090909090909e-05, + "loss": 0.4226, + "step": 17 + }, + { + "epoch": 0.029892573563755253, + "grad_norm": 70.15821838378906, + "learning_rate": 3.090909090909091e-05, + "loss": 0.4456, + "step": 18 + }, + { + "epoch": 0.03155327209507499, + "grad_norm": 15.988476753234863, + "learning_rate": 3.272727272727273e-05, + "loss": 0.446, + "step": 19 + }, + { + "epoch": 0.03321397062639473, + "grad_norm": 9.39255142211914, + "learning_rate": 3.454545454545455e-05, + "loss": 0.4187, + "step": 20 + }, + { + "epoch": 0.034874669157714464, + "grad_norm": 7.129188537597656, + "learning_rate": 3.6363636363636364e-05, + "loss": 0.4394, + "step": 21 + }, + { + "epoch": 0.0365353676890342, + "grad_norm": 12.609400749206543, + "learning_rate": 3.818181818181819e-05, + "loss": 0.4331, + "step": 22 + }, + { + "epoch": 0.03819606622035394, + "grad_norm": 13.4182710647583, + "learning_rate": 4e-05, + "loss": 0.4485, + "step": 23 + }, + { + "epoch": 0.03985676475167367, + "grad_norm": 40.47924041748047, + "learning_rate": 4.181818181818182e-05, + "loss": 0.4465, + "step": 24 + }, + { + "epoch": 0.04151746328299341, + "grad_norm": 25.087459564208984, + "learning_rate": 4.3636363636363636e-05, + "loss": 0.4271, + "step": 25 + }, + { + "epoch": 0.043178161814313144, + "grad_norm": 14.203847885131836, + "learning_rate": 4.545454545454546e-05, + "loss": 0.4207, + "step": 26 + }, + { + "epoch": 0.044838860345632885, + "grad_norm": 52.52724838256836, + "learning_rate": 4.7272727272727275e-05, + "loss": 0.4096, + "step": 27 + }, + { + "epoch": 0.04649955887695262, + "grad_norm": 55.11048889160156, + "learning_rate": 4.909090909090909e-05, + "loss": 0.4185, + "step": 28 + }, + { + "epoch": 0.04816025740827235, + "grad_norm": 17.722021102905273, + "learning_rate": 5.090909090909091e-05, + "loss": 0.3929, + "step": 29 + }, + { + "epoch": 0.04982095593959209, + "grad_norm": 5.971854209899902, + "learning_rate": 5.272727272727272e-05, + "loss": 0.4051, + "step": 30 + }, + { + "epoch": 0.051481654470911825, + "grad_norm": 213.95367431640625, + "learning_rate": 5.4545454545454546e-05, + "loss": 0.4352, + "step": 31 + }, + { + "epoch": 0.053142353002231565, + "grad_norm": 21.540088653564453, + "learning_rate": 5.636363636363636e-05, + "loss": 0.4039, + "step": 32 + }, + { + "epoch": 0.0548030515335513, + "grad_norm": 38.425453186035156, + "learning_rate": 5.818181818181818e-05, + "loss": 0.3927, + "step": 33 + }, + { + "epoch": 0.05646375006487104, + "grad_norm": 13.704586029052734, + "learning_rate": 6e-05, + "loss": 0.3909, + "step": 34 + }, + { + "epoch": 0.05812444859619077, + "grad_norm": 35.628753662109375, + "learning_rate": 6.181818181818182e-05, + "loss": 0.3844, + "step": 35 + }, + { + "epoch": 0.059785147127510506, + "grad_norm": 19.170928955078125, + "learning_rate": 6.363636363636364e-05, + "loss": 0.3973, + "step": 36 + }, + { + "epoch": 0.061445845658830246, + "grad_norm": 30.053909301757812, + "learning_rate": 6.545454545454546e-05, + "loss": 0.3816, + "step": 37 + }, + { + "epoch": 0.06310654419014998, + "grad_norm": 14.09982967376709, + "learning_rate": 6.727272727272727e-05, + "loss": 0.3674, + "step": 38 + }, + { + "epoch": 0.06476724272146972, + "grad_norm": 23.357311248779297, + "learning_rate": 6.90909090909091e-05, + "loss": 0.3658, + "step": 39 + }, + { + "epoch": 0.06642794125278946, + "grad_norm": 38.456756591796875, + "learning_rate": 7.090909090909092e-05, + "loss": 0.3885, + "step": 40 + }, + { + "epoch": 0.06808863978410919, + "grad_norm": 26.597867965698242, + "learning_rate": 7.272727272727273e-05, + "loss": 0.4027, + "step": 41 + }, + { + "epoch": 0.06974933831542893, + "grad_norm": 78.06511688232422, + "learning_rate": 7.454545454545455e-05, + "loss": 0.4209, + "step": 42 + }, + { + "epoch": 0.07141003684674867, + "grad_norm": 14.040410041809082, + "learning_rate": 7.636363636363637e-05, + "loss": 0.3833, + "step": 43 + }, + { + "epoch": 0.0730707353780684, + "grad_norm": 34.664710998535156, + "learning_rate": 7.818181818181818e-05, + "loss": 0.3862, + "step": 44 + }, + { + "epoch": 0.07473143390938813, + "grad_norm": 23.512224197387695, + "learning_rate": 8e-05, + "loss": 0.3821, + "step": 45 + }, + { + "epoch": 0.07639213244070787, + "grad_norm": 21.267244338989258, + "learning_rate": 8.181818181818183e-05, + "loss": 0.3886, + "step": 46 + }, + { + "epoch": 0.07805283097202761, + "grad_norm": 13.587517738342285, + "learning_rate": 8.363636363636364e-05, + "loss": 0.3997, + "step": 47 + }, + { + "epoch": 0.07971352950334734, + "grad_norm": 32.17759323120117, + "learning_rate": 8.545454545454545e-05, + "loss": 0.3757, + "step": 48 + }, + { + "epoch": 0.08137422803466708, + "grad_norm": 97.13513946533203, + "learning_rate": 8.727272727272727e-05, + "loss": 0.3978, + "step": 49 + }, + { + "epoch": 0.08303492656598682, + "grad_norm": 22.572778701782227, + "learning_rate": 8.90909090909091e-05, + "loss": 0.3705, + "step": 50 + }, + { + "epoch": 0.08469562509730655, + "grad_norm": 19.494312286376953, + "learning_rate": 9.090909090909092e-05, + "loss": 0.3753, + "step": 51 + }, + { + "epoch": 0.08635632362862629, + "grad_norm": 13.419837951660156, + "learning_rate": 9.272727272727273e-05, + "loss": 0.3754, + "step": 52 + }, + { + "epoch": 0.08801702215994603, + "grad_norm": 33.55635070800781, + "learning_rate": 9.454545454545455e-05, + "loss": 0.3708, + "step": 53 + }, + { + "epoch": 0.08967772069126577, + "grad_norm": 47.15851974487305, + "learning_rate": 9.636363636363637e-05, + "loss": 0.3799, + "step": 54 + }, + { + "epoch": 0.0913384192225855, + "grad_norm": 49.44485855102539, + "learning_rate": 9.818181818181818e-05, + "loss": 0.3894, + "step": 55 + }, + { + "epoch": 0.09299911775390524, + "grad_norm": 10.229188919067383, + "learning_rate": 0.0001, + "loss": 0.3341, + "step": 56 + }, + { + "epoch": 0.09465981628522498, + "grad_norm": 68.67826080322266, + "learning_rate": 9.999991979887508e-05, + "loss": 0.3623, + "step": 57 + }, + { + "epoch": 0.0963205148165447, + "grad_norm": 46.58010482788086, + "learning_rate": 9.999967919575757e-05, + "loss": 0.3513, + "step": 58 + }, + { + "epoch": 0.09798121334786444, + "grad_norm": 19.71407699584961, + "learning_rate": 9.999927819141937e-05, + "loss": 0.3756, + "step": 59 + }, + { + "epoch": 0.09964191187918418, + "grad_norm": 114.99296569824219, + "learning_rate": 9.999871678714689e-05, + "loss": 0.3814, + "step": 60 + }, + { + "epoch": 0.10130261041050392, + "grad_norm": 235.51441955566406, + "learning_rate": 9.999799498474118e-05, + "loss": 0.3709, + "step": 61 + }, + { + "epoch": 0.10296330894182365, + "grad_norm": 58.31473159790039, + "learning_rate": 9.999711278651777e-05, + "loss": 0.3335, + "step": 62 + }, + { + "epoch": 0.10462400747314339, + "grad_norm": 38.7056884765625, + "learning_rate": 9.999607019530681e-05, + "loss": 0.3833, + "step": 63 + }, + { + "epoch": 0.10628470600446313, + "grad_norm": 29.422666549682617, + "learning_rate": 9.999486721445299e-05, + "loss": 0.3883, + "step": 64 + }, + { + "epoch": 0.10794540453578286, + "grad_norm": 55.9561653137207, + "learning_rate": 9.999350384781552e-05, + "loss": 0.3657, + "step": 65 + }, + { + "epoch": 0.1096061030671026, + "grad_norm": 19.884254455566406, + "learning_rate": 9.999198009976812e-05, + "loss": 0.3287, + "step": 66 + }, + { + "epoch": 0.11126680159842234, + "grad_norm": 26.66572380065918, + "learning_rate": 9.999029597519907e-05, + "loss": 0.3396, + "step": 67 + }, + { + "epoch": 0.11292750012974208, + "grad_norm": 38.8479118347168, + "learning_rate": 9.998845147951113e-05, + "loss": 0.3448, + "step": 68 + }, + { + "epoch": 0.1145881986610618, + "grad_norm": 15.002911567687988, + "learning_rate": 9.998644661862148e-05, + "loss": 0.3527, + "step": 69 + }, + { + "epoch": 0.11624889719238155, + "grad_norm": 23.062976837158203, + "learning_rate": 9.998428139896183e-05, + "loss": 0.3383, + "step": 70 + }, + { + "epoch": 0.11790959572370129, + "grad_norm": 12.6146821975708, + "learning_rate": 9.99819558274783e-05, + "loss": 0.3332, + "step": 71 + }, + { + "epoch": 0.11957029425502101, + "grad_norm": 25.191553115844727, + "learning_rate": 9.997946991163143e-05, + "loss": 0.3659, + "step": 72 + }, + { + "epoch": 0.12123099278634075, + "grad_norm": 29.208486557006836, + "learning_rate": 9.997682365939616e-05, + "loss": 0.3645, + "step": 73 + }, + { + "epoch": 0.12289169131766049, + "grad_norm": 92.01864624023438, + "learning_rate": 9.997401707926175e-05, + "loss": 0.361, + "step": 74 + }, + { + "epoch": 0.12455238984898023, + "grad_norm": 19.42112159729004, + "learning_rate": 9.997105018023187e-05, + "loss": 0.3521, + "step": 75 + }, + { + "epoch": 0.12621308838029996, + "grad_norm": 17.35445785522461, + "learning_rate": 9.996792297182447e-05, + "loss": 0.3652, + "step": 76 + }, + { + "epoch": 0.1278737869116197, + "grad_norm": 18.6886043548584, + "learning_rate": 9.996463546407174e-05, + "loss": 0.3373, + "step": 77 + }, + { + "epoch": 0.12953448544293944, + "grad_norm": 21.951919555664062, + "learning_rate": 9.996118766752018e-05, + "loss": 0.3341, + "step": 78 + }, + { + "epoch": 0.13119518397425917, + "grad_norm": 10.152750968933105, + "learning_rate": 9.995757959323047e-05, + "loss": 0.359, + "step": 79 + }, + { + "epoch": 0.13285588250557892, + "grad_norm": 12.917724609375, + "learning_rate": 9.995381125277748e-05, + "loss": 0.3547, + "step": 80 + }, + { + "epoch": 0.13451658103689865, + "grad_norm": 27.07183074951172, + "learning_rate": 9.99498826582502e-05, + "loss": 0.3279, + "step": 81 + }, + { + "epoch": 0.13617727956821837, + "grad_norm": 50.98102951049805, + "learning_rate": 9.994579382225175e-05, + "loss": 0.336, + "step": 82 + }, + { + "epoch": 0.13783797809953813, + "grad_norm": 30.40250015258789, + "learning_rate": 9.994154475789931e-05, + "loss": 0.3508, + "step": 83 + }, + { + "epoch": 0.13949867663085785, + "grad_norm": 63.462032318115234, + "learning_rate": 9.993713547882404e-05, + "loss": 0.3427, + "step": 84 + }, + { + "epoch": 0.14115937516217758, + "grad_norm": 15.641836166381836, + "learning_rate": 9.993256599917113e-05, + "loss": 0.3639, + "step": 85 + }, + { + "epoch": 0.14282007369349733, + "grad_norm": 15.584700584411621, + "learning_rate": 9.992783633359967e-05, + "loss": 0.3628, + "step": 86 + }, + { + "epoch": 0.14448077222481706, + "grad_norm": 18.697765350341797, + "learning_rate": 9.992294649728265e-05, + "loss": 0.3484, + "step": 87 + }, + { + "epoch": 0.1461414707561368, + "grad_norm": 32.01298522949219, + "learning_rate": 9.991789650590686e-05, + "loss": 0.3386, + "step": 88 + }, + { + "epoch": 0.14780216928745654, + "grad_norm": 21.03476905822754, + "learning_rate": 9.991268637567292e-05, + "loss": 0.3503, + "step": 89 + }, + { + "epoch": 0.14946286781877627, + "grad_norm": 34.96243667602539, + "learning_rate": 9.990731612329516e-05, + "loss": 0.3343, + "step": 90 + }, + { + "epoch": 0.15112356635009602, + "grad_norm": 230.6318817138672, + "learning_rate": 9.990178576600157e-05, + "loss": 0.3478, + "step": 91 + }, + { + "epoch": 0.15278426488141575, + "grad_norm": 76.9327163696289, + "learning_rate": 9.989609532153381e-05, + "loss": 0.3131, + "step": 92 + }, + { + "epoch": 0.15444496341273548, + "grad_norm": 23.221542358398438, + "learning_rate": 9.989024480814708e-05, + "loss": 0.3657, + "step": 93 + }, + { + "epoch": 0.15610566194405523, + "grad_norm": 14.934358596801758, + "learning_rate": 9.988423424461009e-05, + "loss": 0.3445, + "step": 94 + }, + { + "epoch": 0.15776636047537496, + "grad_norm": 18.620206832885742, + "learning_rate": 9.987806365020497e-05, + "loss": 0.3562, + "step": 95 + }, + { + "epoch": 0.15942705900669468, + "grad_norm": 45.51302719116211, + "learning_rate": 9.987173304472733e-05, + "loss": 0.3321, + "step": 96 + }, + { + "epoch": 0.16108775753801444, + "grad_norm": 30.344011306762695, + "learning_rate": 9.986524244848596e-05, + "loss": 0.3505, + "step": 97 + }, + { + "epoch": 0.16274845606933416, + "grad_norm": 92.91533660888672, + "learning_rate": 9.985859188230302e-05, + "loss": 0.354, + "step": 98 + }, + { + "epoch": 0.1644091546006539, + "grad_norm": 23.510000228881836, + "learning_rate": 9.985178136751383e-05, + "loss": 0.3461, + "step": 99 + }, + { + "epoch": 0.16606985313197364, + "grad_norm": 20.6863956451416, + "learning_rate": 9.984481092596683e-05, + "loss": 0.3425, + "step": 100 + }, + { + "epoch": 0.16773055166329337, + "grad_norm": 15.736349105834961, + "learning_rate": 9.983768058002349e-05, + "loss": 0.313, + "step": 101 + }, + { + "epoch": 0.1693912501946131, + "grad_norm": 26.916696548461914, + "learning_rate": 9.983039035255831e-05, + "loss": 0.3547, + "step": 102 + }, + { + "epoch": 0.17105194872593285, + "grad_norm": 14.109468460083008, + "learning_rate": 9.982294026695865e-05, + "loss": 0.3637, + "step": 103 + }, + { + "epoch": 0.17271264725725258, + "grad_norm": 14.722790718078613, + "learning_rate": 9.981533034712472e-05, + "loss": 0.3471, + "step": 104 + }, + { + "epoch": 0.17437334578857233, + "grad_norm": 44.82086181640625, + "learning_rate": 9.980756061746948e-05, + "loss": 0.3681, + "step": 105 + }, + { + "epoch": 0.17603404431989206, + "grad_norm": 9.19070053100586, + "learning_rate": 9.979963110291857e-05, + "loss": 0.3518, + "step": 106 + }, + { + "epoch": 0.17769474285121178, + "grad_norm": 27.845748901367188, + "learning_rate": 9.979154182891024e-05, + "loss": 0.3431, + "step": 107 + }, + { + "epoch": 0.17935544138253154, + "grad_norm": 11.760729789733887, + "learning_rate": 9.978329282139525e-05, + "loss": 0.3362, + "step": 108 + }, + { + "epoch": 0.18101613991385127, + "grad_norm": 22.305797576904297, + "learning_rate": 9.97748841068368e-05, + "loss": 0.3462, + "step": 109 + }, + { + "epoch": 0.182676838445171, + "grad_norm": 42.350955963134766, + "learning_rate": 9.976631571221038e-05, + "loss": 0.3271, + "step": 110 + }, + { + "epoch": 0.18433753697649075, + "grad_norm": 25.78905487060547, + "learning_rate": 9.975758766500382e-05, + "loss": 0.3481, + "step": 111 + }, + { + "epoch": 0.18599823550781047, + "grad_norm": 24.677139282226562, + "learning_rate": 9.974869999321707e-05, + "loss": 0.3439, + "step": 112 + }, + { + "epoch": 0.1876589340391302, + "grad_norm": 159.541748046875, + "learning_rate": 9.973965272536219e-05, + "loss": 0.3388, + "step": 113 + }, + { + "epoch": 0.18931963257044995, + "grad_norm": 28.44233512878418, + "learning_rate": 9.973044589046325e-05, + "loss": 0.3296, + "step": 114 + }, + { + "epoch": 0.19098033110176968, + "grad_norm": 95.02351379394531, + "learning_rate": 9.972107951805614e-05, + "loss": 0.3551, + "step": 115 + }, + { + "epoch": 0.1926410296330894, + "grad_norm": 11.964502334594727, + "learning_rate": 9.971155363818862e-05, + "loss": 0.3465, + "step": 116 + }, + { + "epoch": 0.19430172816440916, + "grad_norm": 60.92937469482422, + "learning_rate": 9.970186828142017e-05, + "loss": 0.3482, + "step": 117 + }, + { + "epoch": 0.1959624266957289, + "grad_norm": 37.06871795654297, + "learning_rate": 9.96920234788218e-05, + "loss": 0.3612, + "step": 118 + }, + { + "epoch": 0.19762312522704864, + "grad_norm": 9.348763465881348, + "learning_rate": 9.968201926197613e-05, + "loss": 0.3228, + "step": 119 + }, + { + "epoch": 0.19928382375836837, + "grad_norm": 55.865013122558594, + "learning_rate": 9.967185566297712e-05, + "loss": 0.3433, + "step": 120 + }, + { + "epoch": 0.2009445222896881, + "grad_norm": 35.589839935302734, + "learning_rate": 9.966153271443004e-05, + "loss": 0.3444, + "step": 121 + }, + { + "epoch": 0.20260522082100785, + "grad_norm": 29.946760177612305, + "learning_rate": 9.965105044945138e-05, + "loss": 0.3361, + "step": 122 + }, + { + "epoch": 0.20426591935232757, + "grad_norm": 16.42388343811035, + "learning_rate": 9.964040890166872e-05, + "loss": 0.3079, + "step": 123 + }, + { + "epoch": 0.2059266178836473, + "grad_norm": 26.32118034362793, + "learning_rate": 9.962960810522064e-05, + "loss": 0.3153, + "step": 124 + }, + { + "epoch": 0.20758731641496705, + "grad_norm": 53.86672592163086, + "learning_rate": 9.961864809475654e-05, + "loss": 0.3246, + "step": 125 + }, + { + "epoch": 0.20924801494628678, + "grad_norm": 9.0363130569458, + "learning_rate": 9.960752890543667e-05, + "loss": 0.3243, + "step": 126 + }, + { + "epoch": 0.2109087134776065, + "grad_norm": 33.26886749267578, + "learning_rate": 9.959625057293187e-05, + "loss": 0.342, + "step": 127 + }, + { + "epoch": 0.21256941200892626, + "grad_norm": 23.017858505249023, + "learning_rate": 9.958481313342354e-05, + "loss": 0.3396, + "step": 128 + }, + { + "epoch": 0.214230110540246, + "grad_norm": 281.400634765625, + "learning_rate": 9.95732166236035e-05, + "loss": 0.3465, + "step": 129 + }, + { + "epoch": 0.21589080907156571, + "grad_norm": 19.316598892211914, + "learning_rate": 9.956146108067388e-05, + "loss": 0.3438, + "step": 130 + }, + { + "epoch": 0.21755150760288547, + "grad_norm": 15.318970680236816, + "learning_rate": 9.954954654234699e-05, + "loss": 0.3333, + "step": 131 + }, + { + "epoch": 0.2192122061342052, + "grad_norm": 36.29383850097656, + "learning_rate": 9.953747304684518e-05, + "loss": 0.3578, + "step": 132 + }, + { + "epoch": 0.22087290466552495, + "grad_norm": 24.523378372192383, + "learning_rate": 9.95252406329008e-05, + "loss": 0.3366, + "step": 133 + }, + { + "epoch": 0.22253360319684468, + "grad_norm": 78.71446990966797, + "learning_rate": 9.951284933975598e-05, + "loss": 0.3507, + "step": 134 + }, + { + "epoch": 0.2241943017281644, + "grad_norm": 14.822957992553711, + "learning_rate": 9.950029920716252e-05, + "loss": 0.3467, + "step": 135 + }, + { + "epoch": 0.22585500025948416, + "grad_norm": 31.55790901184082, + "learning_rate": 9.948759027538185e-05, + "loss": 0.3351, + "step": 136 + }, + { + "epoch": 0.22751569879080388, + "grad_norm": 76.93319702148438, + "learning_rate": 9.947472258518478e-05, + "loss": 0.3406, + "step": 137 + }, + { + "epoch": 0.2291763973221236, + "grad_norm": 45.591285705566406, + "learning_rate": 9.946169617785141e-05, + "loss": 0.3369, + "step": 138 + }, + { + "epoch": 0.23083709585344336, + "grad_norm": 18.619455337524414, + "learning_rate": 9.944851109517107e-05, + "loss": 0.3382, + "step": 139 + }, + { + "epoch": 0.2324977943847631, + "grad_norm": 49.328182220458984, + "learning_rate": 9.943516737944209e-05, + "loss": 0.3161, + "step": 140 + }, + { + "epoch": 0.23415849291608282, + "grad_norm": 41.13008117675781, + "learning_rate": 9.942166507347173e-05, + "loss": 0.3446, + "step": 141 + }, + { + "epoch": 0.23581919144740257, + "grad_norm": 28.75891876220703, + "learning_rate": 9.940800422057596e-05, + "loss": 0.3452, + "step": 142 + }, + { + "epoch": 0.2374798899787223, + "grad_norm": 91.85528564453125, + "learning_rate": 9.939418486457945e-05, + "loss": 0.3418, + "step": 143 + }, + { + "epoch": 0.23914058851004202, + "grad_norm": 24.160757064819336, + "learning_rate": 9.938020704981528e-05, + "loss": 0.3168, + "step": 144 + }, + { + "epoch": 0.24080128704136178, + "grad_norm": 50.253143310546875, + "learning_rate": 9.936607082112492e-05, + "loss": 0.334, + "step": 145 + }, + { + "epoch": 0.2424619855726815, + "grad_norm": 14.691905975341797, + "learning_rate": 9.935177622385804e-05, + "loss": 0.3411, + "step": 146 + }, + { + "epoch": 0.24412268410400126, + "grad_norm": 20.2746639251709, + "learning_rate": 9.933732330387234e-05, + "loss": 0.3375, + "step": 147 + }, + { + "epoch": 0.24578338263532098, + "grad_norm": 15.6539306640625, + "learning_rate": 9.932271210753343e-05, + "loss": 0.3489, + "step": 148 + }, + { + "epoch": 0.2474440811666407, + "grad_norm": 24.095462799072266, + "learning_rate": 9.93079426817147e-05, + "loss": 0.3379, + "step": 149 + }, + { + "epoch": 0.24910477969796047, + "grad_norm": 14.961967468261719, + "learning_rate": 9.929301507379715e-05, + "loss": 0.3085, + "step": 150 + }, + { + "epoch": 0.2507654782292802, + "grad_norm": 152.29129028320312, + "learning_rate": 9.927792933166917e-05, + "loss": 0.3287, + "step": 151 + }, + { + "epoch": 0.2524261767605999, + "grad_norm": 29.937938690185547, + "learning_rate": 9.926268550372654e-05, + "loss": 0.3253, + "step": 152 + }, + { + "epoch": 0.25408687529191964, + "grad_norm": 13.421550750732422, + "learning_rate": 9.924728363887212e-05, + "loss": 0.3278, + "step": 153 + }, + { + "epoch": 0.2557475738232394, + "grad_norm": 24.944440841674805, + "learning_rate": 9.923172378651581e-05, + "loss": 0.3316, + "step": 154 + }, + { + "epoch": 0.25740827235455915, + "grad_norm": 118.63322448730469, + "learning_rate": 9.92160059965743e-05, + "loss": 0.3531, + "step": 155 + }, + { + "epoch": 0.2590689708858789, + "grad_norm": 34.51362228393555, + "learning_rate": 9.920013031947097e-05, + "loss": 0.3319, + "step": 156 + }, + { + "epoch": 0.2607296694171986, + "grad_norm": 32.96346664428711, + "learning_rate": 9.918409680613569e-05, + "loss": 0.3607, + "step": 157 + }, + { + "epoch": 0.26239036794851833, + "grad_norm": 21.493207931518555, + "learning_rate": 9.916790550800474e-05, + "loss": 0.3389, + "step": 158 + }, + { + "epoch": 0.26405106647983806, + "grad_norm": 72.81291198730469, + "learning_rate": 9.915155647702048e-05, + "loss": 0.339, + "step": 159 + }, + { + "epoch": 0.26571176501115784, + "grad_norm": 19.447999954223633, + "learning_rate": 9.913504976563137e-05, + "loss": 0.3245, + "step": 160 + }, + { + "epoch": 0.26737246354247757, + "grad_norm": 16.12717628479004, + "learning_rate": 9.911838542679167e-05, + "loss": 0.293, + "step": 161 + }, + { + "epoch": 0.2690331620737973, + "grad_norm": 35.158451080322266, + "learning_rate": 9.910156351396134e-05, + "loss": 0.349, + "step": 162 + }, + { + "epoch": 0.270693860605117, + "grad_norm": 30.566896438598633, + "learning_rate": 9.908458408110582e-05, + "loss": 0.3272, + "step": 163 + }, + { + "epoch": 0.27235455913643675, + "grad_norm": 22.65450096130371, + "learning_rate": 9.90674471826959e-05, + "loss": 0.3335, + "step": 164 + }, + { + "epoch": 0.2740152576677565, + "grad_norm": 13.14643669128418, + "learning_rate": 9.90501528737075e-05, + "loss": 0.3223, + "step": 165 + }, + { + "epoch": 0.27567595619907626, + "grad_norm": 56.709999084472656, + "learning_rate": 9.903270120962159e-05, + "loss": 0.3487, + "step": 166 + }, + { + "epoch": 0.277336654730396, + "grad_norm": 21.17612075805664, + "learning_rate": 9.901509224642385e-05, + "loss": 0.34, + "step": 167 + }, + { + "epoch": 0.2789973532617157, + "grad_norm": 297.5917663574219, + "learning_rate": 9.899732604060464e-05, + "loss": 0.33, + "step": 168 + }, + { + "epoch": 0.28065805179303543, + "grad_norm": 96.20665740966797, + "learning_rate": 9.897940264915875e-05, + "loss": 0.3154, + "step": 169 + }, + { + "epoch": 0.28231875032435516, + "grad_norm": 30.58903694152832, + "learning_rate": 9.896132212958524e-05, + "loss": 0.3409, + "step": 170 + }, + { + "epoch": 0.28397944885567494, + "grad_norm": 36.77935791015625, + "learning_rate": 9.894308453988721e-05, + "loss": 0.321, + "step": 171 + }, + { + "epoch": 0.28564014738699467, + "grad_norm": 14.269662857055664, + "learning_rate": 9.892468993857166e-05, + "loss": 0.3185, + "step": 172 + }, + { + "epoch": 0.2873008459183144, + "grad_norm": 20.358572006225586, + "learning_rate": 9.890613838464934e-05, + "loss": 0.3105, + "step": 173 + }, + { + "epoch": 0.2889615444496341, + "grad_norm": 17.754365921020508, + "learning_rate": 9.888742993763442e-05, + "loss": 0.3124, + "step": 174 + }, + { + "epoch": 0.29062224298095385, + "grad_norm": 55.279319763183594, + "learning_rate": 9.886856465754448e-05, + "loss": 0.3135, + "step": 175 + }, + { + "epoch": 0.2922829415122736, + "grad_norm": 34.28364944458008, + "learning_rate": 9.884954260490017e-05, + "loss": 0.3065, + "step": 176 + }, + { + "epoch": 0.29394364004359336, + "grad_norm": 30.18939971923828, + "learning_rate": 9.883036384072509e-05, + "loss": 0.3159, + "step": 177 + }, + { + "epoch": 0.2956043385749131, + "grad_norm": 70.1207504272461, + "learning_rate": 9.881102842654557e-05, + "loss": 0.3432, + "step": 178 + }, + { + "epoch": 0.2972650371062328, + "grad_norm": 12.422219276428223, + "learning_rate": 9.87915364243905e-05, + "loss": 0.3487, + "step": 179 + }, + { + "epoch": 0.29892573563755254, + "grad_norm": 20.446584701538086, + "learning_rate": 9.87718878967911e-05, + "loss": 0.3346, + "step": 180 + }, + { + "epoch": 0.30058643416887226, + "grad_norm": 64.20825958251953, + "learning_rate": 9.875208290678073e-05, + "loss": 0.3325, + "step": 181 + }, + { + "epoch": 0.30224713270019204, + "grad_norm": 14.690919876098633, + "learning_rate": 9.873212151789468e-05, + "loss": 0.3269, + "step": 182 + }, + { + "epoch": 0.30390783123151177, + "grad_norm": 30.525850296020508, + "learning_rate": 9.871200379417e-05, + "loss": 0.3107, + "step": 183 + }, + { + "epoch": 0.3055685297628315, + "grad_norm": 29.644283294677734, + "learning_rate": 9.869172980014525e-05, + "loss": 0.3477, + "step": 184 + }, + { + "epoch": 0.3072292282941512, + "grad_norm": 404.232666015625, + "learning_rate": 9.86712996008603e-05, + "loss": 0.3285, + "step": 185 + }, + { + "epoch": 0.30888992682547095, + "grad_norm": 20.1916561126709, + "learning_rate": 9.865071326185618e-05, + "loss": 0.325, + "step": 186 + }, + { + "epoch": 0.3105506253567907, + "grad_norm": 15.538655281066895, + "learning_rate": 9.862997084917476e-05, + "loss": 0.3392, + "step": 187 + }, + { + "epoch": 0.31221132388811046, + "grad_norm": 19.837736129760742, + "learning_rate": 9.860907242935863e-05, + "loss": 0.3096, + "step": 188 + }, + { + "epoch": 0.3138720224194302, + "grad_norm": 27.80318260192871, + "learning_rate": 9.858801806945088e-05, + "loss": 0.3257, + "step": 189 + }, + { + "epoch": 0.3155327209507499, + "grad_norm": 35.97450256347656, + "learning_rate": 9.856680783699484e-05, + "loss": 0.3243, + "step": 190 + }, + { + "epoch": 0.31719341948206964, + "grad_norm": 11.300904273986816, + "learning_rate": 9.85454418000339e-05, + "loss": 0.3308, + "step": 191 + }, + { + "epoch": 0.31885411801338936, + "grad_norm": 36.19790267944336, + "learning_rate": 9.852392002711125e-05, + "loss": 0.3253, + "step": 192 + }, + { + "epoch": 0.3205148165447091, + "grad_norm": 18.56528091430664, + "learning_rate": 9.850224258726972e-05, + "loss": 0.3028, + "step": 193 + }, + { + "epoch": 0.3221755150760289, + "grad_norm": 404.9875793457031, + "learning_rate": 9.848040955005148e-05, + "loss": 0.3069, + "step": 194 + }, + { + "epoch": 0.3238362136073486, + "grad_norm": 28.360759735107422, + "learning_rate": 9.845842098549795e-05, + "loss": 0.3167, + "step": 195 + }, + { + "epoch": 0.3254969121386683, + "grad_norm": 79.17635345458984, + "learning_rate": 9.843627696414939e-05, + "loss": 0.3435, + "step": 196 + }, + { + "epoch": 0.32715761066998805, + "grad_norm": 66.2103271484375, + "learning_rate": 9.841397755704483e-05, + "loss": 0.324, + "step": 197 + }, + { + "epoch": 0.3288183092013078, + "grad_norm": 15.748398780822754, + "learning_rate": 9.839152283572178e-05, + "loss": 0.3144, + "step": 198 + }, + { + "epoch": 0.33047900773262756, + "grad_norm": 93.86544799804688, + "learning_rate": 9.836891287221599e-05, + "loss": 0.3031, + "step": 199 + }, + { + "epoch": 0.3321397062639473, + "grad_norm": 323.493896484375, + "learning_rate": 9.834614773906124e-05, + "loss": 0.3517, + "step": 200 + }, + { + "epoch": 0.333800404795267, + "grad_norm": 41.129146575927734, + "learning_rate": 9.83232275092891e-05, + "loss": 0.333, + "step": 201 + }, + { + "epoch": 0.33546110332658674, + "grad_norm": 13.682321548461914, + "learning_rate": 9.830015225642873e-05, + "loss": 0.3081, + "step": 202 + }, + { + "epoch": 0.33712180185790647, + "grad_norm": 90.22856903076172, + "learning_rate": 9.827692205450651e-05, + "loss": 0.3474, + "step": 203 + }, + { + "epoch": 0.3387825003892262, + "grad_norm": 32.902252197265625, + "learning_rate": 9.825353697804602e-05, + "loss": 0.3325, + "step": 204 + }, + { + "epoch": 0.340443198920546, + "grad_norm": 72.50242614746094, + "learning_rate": 9.822999710206764e-05, + "loss": 0.3403, + "step": 205 + }, + { + "epoch": 0.3421038974518657, + "grad_norm": 42.025611877441406, + "learning_rate": 9.820630250208835e-05, + "loss": 0.3283, + "step": 206 + }, + { + "epoch": 0.34376459598318543, + "grad_norm": 63.42696762084961, + "learning_rate": 9.818245325412148e-05, + "loss": 0.3074, + "step": 207 + }, + { + "epoch": 0.34542529451450515, + "grad_norm": 21.447444915771484, + "learning_rate": 9.815844943467649e-05, + "loss": 0.2909, + "step": 208 + }, + { + "epoch": 0.3470859930458249, + "grad_norm": 38.42819595336914, + "learning_rate": 9.813429112075872e-05, + "loss": 0.3148, + "step": 209 + }, + { + "epoch": 0.34874669157714466, + "grad_norm": 143.58738708496094, + "learning_rate": 9.810997838986913e-05, + "loss": 0.3317, + "step": 210 + }, + { + "epoch": 0.3504073901084644, + "grad_norm": 14.622925758361816, + "learning_rate": 9.808551132000404e-05, + "loss": 0.3334, + "step": 211 + }, + { + "epoch": 0.3520680886397841, + "grad_norm": 41.87562942504883, + "learning_rate": 9.806088998965494e-05, + "loss": 0.3438, + "step": 212 + }, + { + "epoch": 0.35372878717110384, + "grad_norm": 14.850021362304688, + "learning_rate": 9.803611447780814e-05, + "loss": 0.3147, + "step": 213 + }, + { + "epoch": 0.35538948570242357, + "grad_norm": 19.952550888061523, + "learning_rate": 9.801118486394461e-05, + "loss": 0.3348, + "step": 214 + }, + { + "epoch": 0.3570501842337433, + "grad_norm": 14.695087432861328, + "learning_rate": 9.798610122803966e-05, + "loss": 0.3301, + "step": 215 + }, + { + "epoch": 0.3587108827650631, + "grad_norm": 119.33077239990234, + "learning_rate": 9.796086365056272e-05, + "loss": 0.3353, + "step": 216 + }, + { + "epoch": 0.3603715812963828, + "grad_norm": 24.40523910522461, + "learning_rate": 9.793547221247709e-05, + "loss": 0.3222, + "step": 217 + }, + { + "epoch": 0.36203227982770253, + "grad_norm": 26.061222076416016, + "learning_rate": 9.790992699523967e-05, + "loss": 0.3299, + "step": 218 + }, + { + "epoch": 0.36369297835902226, + "grad_norm": 14.160664558410645, + "learning_rate": 9.78842280808006e-05, + "loss": 0.3554, + "step": 219 + }, + { + "epoch": 0.365353676890342, + "grad_norm": 25.48128318786621, + "learning_rate": 9.785837555160321e-05, + "loss": 0.3054, + "step": 220 + }, + { + "epoch": 0.3670143754216617, + "grad_norm": 21.467037200927734, + "learning_rate": 9.783236949058356e-05, + "loss": 0.3152, + "step": 221 + }, + { + "epoch": 0.3686750739529815, + "grad_norm": 41.671390533447266, + "learning_rate": 9.780620998117025e-05, + "loss": 0.3095, + "step": 222 + }, + { + "epoch": 0.3703357724843012, + "grad_norm": 10.983573913574219, + "learning_rate": 9.777989710728418e-05, + "loss": 0.3558, + "step": 223 + }, + { + "epoch": 0.37199647101562094, + "grad_norm": 15.790307998657227, + "learning_rate": 9.775343095333824e-05, + "loss": 0.3062, + "step": 224 + }, + { + "epoch": 0.37365716954694067, + "grad_norm": 12.44168472290039, + "learning_rate": 9.772681160423703e-05, + "loss": 0.3364, + "step": 225 + }, + { + "epoch": 0.3753178680782604, + "grad_norm": 25.46112060546875, + "learning_rate": 9.77000391453766e-05, + "loss": 0.3203, + "step": 226 + }, + { + "epoch": 0.3769785666095802, + "grad_norm": 24.347057342529297, + "learning_rate": 9.767311366264425e-05, + "loss": 0.3104, + "step": 227 + }, + { + "epoch": 0.3786392651408999, + "grad_norm": 27.665260314941406, + "learning_rate": 9.76460352424181e-05, + "loss": 0.3309, + "step": 228 + }, + { + "epoch": 0.38029996367221963, + "grad_norm": 10.520977973937988, + "learning_rate": 9.761880397156695e-05, + "loss": 0.3194, + "step": 229 + }, + { + "epoch": 0.38196066220353936, + "grad_norm": 18.125877380371094, + "learning_rate": 9.759141993744995e-05, + "loss": 0.317, + "step": 230 + }, + { + "epoch": 0.3836213607348591, + "grad_norm": 18.901464462280273, + "learning_rate": 9.756388322791631e-05, + "loss": 0.3188, + "step": 231 + }, + { + "epoch": 0.3852820592661788, + "grad_norm": 36.56083297729492, + "learning_rate": 9.753619393130504e-05, + "loss": 0.3139, + "step": 232 + }, + { + "epoch": 0.3869427577974986, + "grad_norm": 165.3616943359375, + "learning_rate": 9.750835213644464e-05, + "loss": 0.3323, + "step": 233 + }, + { + "epoch": 0.3886034563288183, + "grad_norm": 61.09331130981445, + "learning_rate": 9.748035793265285e-05, + "loss": 0.3238, + "step": 234 + }, + { + "epoch": 0.39026415486013805, + "grad_norm": 14.617741584777832, + "learning_rate": 9.745221140973632e-05, + "loss": 0.3467, + "step": 235 + }, + { + "epoch": 0.3919248533914578, + "grad_norm": 25.66883087158203, + "learning_rate": 9.742391265799037e-05, + "loss": 0.3154, + "step": 236 + }, + { + "epoch": 0.3935855519227775, + "grad_norm": 20.4726505279541, + "learning_rate": 9.739546176819867e-05, + "loss": 0.3225, + "step": 237 + }, + { + "epoch": 0.3952462504540973, + "grad_norm": 40.93400192260742, + "learning_rate": 9.736685883163297e-05, + "loss": 0.3281, + "step": 238 + }, + { + "epoch": 0.396906948985417, + "grad_norm": 311.4083251953125, + "learning_rate": 9.733810394005274e-05, + "loss": 0.3362, + "step": 239 + }, + { + "epoch": 0.39856764751673673, + "grad_norm": 22.18343734741211, + "learning_rate": 9.730919718570502e-05, + "loss": 0.323, + "step": 240 + }, + { + "epoch": 0.40022834604805646, + "grad_norm": 31.630779266357422, + "learning_rate": 9.728013866132392e-05, + "loss": 0.3277, + "step": 241 + }, + { + "epoch": 0.4018890445793762, + "grad_norm": 12.531517028808594, + "learning_rate": 9.725092846013053e-05, + "loss": 0.3182, + "step": 242 + }, + { + "epoch": 0.4035497431106959, + "grad_norm": 31.334840774536133, + "learning_rate": 9.722156667583248e-05, + "loss": 0.3463, + "step": 243 + }, + { + "epoch": 0.4052104416420157, + "grad_norm": 48.84646987915039, + "learning_rate": 9.71920534026237e-05, + "loss": 0.3239, + "step": 244 + }, + { + "epoch": 0.4068711401733354, + "grad_norm": 16.593637466430664, + "learning_rate": 9.716238873518409e-05, + "loss": 0.315, + "step": 245 + }, + { + "epoch": 0.40853183870465515, + "grad_norm": 19.017454147338867, + "learning_rate": 9.713257276867924e-05, + "loss": 0.3129, + "step": 246 + }, + { + "epoch": 0.4101925372359749, + "grad_norm": 25.900310516357422, + "learning_rate": 9.710260559876011e-05, + "loss": 0.3123, + "step": 247 + }, + { + "epoch": 0.4118532357672946, + "grad_norm": 31.709957122802734, + "learning_rate": 9.707248732156275e-05, + "loss": 0.3044, + "step": 248 + }, + { + "epoch": 0.4135139342986143, + "grad_norm": 28.169206619262695, + "learning_rate": 9.704221803370792e-05, + "loss": 0.325, + "step": 249 + }, + { + "epoch": 0.4151746328299341, + "grad_norm": 24.24585723876953, + "learning_rate": 9.701179783230087e-05, + "loss": 0.3054, + "step": 250 + }, + { + "epoch": 0.41683533136125384, + "grad_norm": 12.39324951171875, + "learning_rate": 9.698122681493098e-05, + "loss": 0.296, + "step": 251 + }, + { + "epoch": 0.41849602989257356, + "grad_norm": 22.066993713378906, + "learning_rate": 9.695050507967144e-05, + "loss": 0.3165, + "step": 252 + }, + { + "epoch": 0.4201567284238933, + "grad_norm": 80.2625732421875, + "learning_rate": 9.691963272507898e-05, + "loss": 0.3263, + "step": 253 + }, + { + "epoch": 0.421817426955213, + "grad_norm": 28.812179565429688, + "learning_rate": 9.688860985019348e-05, + "loss": 0.3329, + "step": 254 + }, + { + "epoch": 0.4234781254865328, + "grad_norm": 80.37259674072266, + "learning_rate": 9.685743655453772e-05, + "loss": 0.3182, + "step": 255 + }, + { + "epoch": 0.4251388240178525, + "grad_norm": 45.12711715698242, + "learning_rate": 9.682611293811703e-05, + "loss": 0.3121, + "step": 256 + }, + { + "epoch": 0.42679952254917225, + "grad_norm": 11.43038272857666, + "learning_rate": 9.6794639101419e-05, + "loss": 0.2963, + "step": 257 + }, + { + "epoch": 0.428460221080492, + "grad_norm": 125.04705047607422, + "learning_rate": 9.67630151454131e-05, + "loss": 0.3113, + "step": 258 + }, + { + "epoch": 0.4301209196118117, + "grad_norm": 27.13665771484375, + "learning_rate": 9.673124117155043e-05, + "loss": 0.3089, + "step": 259 + }, + { + "epoch": 0.43178161814313143, + "grad_norm": 16.769529342651367, + "learning_rate": 9.669931728176328e-05, + "loss": 0.3138, + "step": 260 + }, + { + "epoch": 0.4334423166744512, + "grad_norm": 10.994537353515625, + "learning_rate": 9.666724357846498e-05, + "loss": 0.3123, + "step": 261 + }, + { + "epoch": 0.43510301520577094, + "grad_norm": 41.37866973876953, + "learning_rate": 9.663502016454936e-05, + "loss": 0.3097, + "step": 262 + }, + { + "epoch": 0.43676371373709066, + "grad_norm": 82.15670776367188, + "learning_rate": 9.660264714339063e-05, + "loss": 0.3329, + "step": 263 + }, + { + "epoch": 0.4384244122684104, + "grad_norm": 12.925142288208008, + "learning_rate": 9.657012461884285e-05, + "loss": 0.2965, + "step": 264 + }, + { + "epoch": 0.4400851107997301, + "grad_norm": 13.891080856323242, + "learning_rate": 9.653745269523977e-05, + "loss": 0.3504, + "step": 265 + }, + { + "epoch": 0.4417458093310499, + "grad_norm": 18.489437103271484, + "learning_rate": 9.650463147739442e-05, + "loss": 0.3248, + "step": 266 + }, + { + "epoch": 0.4434065078623696, + "grad_norm": 11.54362678527832, + "learning_rate": 9.647166107059868e-05, + "loss": 0.3163, + "step": 267 + }, + { + "epoch": 0.44506720639368935, + "grad_norm": 19.45928382873535, + "learning_rate": 9.643854158062315e-05, + "loss": 0.3329, + "step": 268 + }, + { + "epoch": 0.4467279049250091, + "grad_norm": 20.200828552246094, + "learning_rate": 9.640527311371662e-05, + "loss": 0.3373, + "step": 269 + }, + { + "epoch": 0.4483886034563288, + "grad_norm": 12.717829704284668, + "learning_rate": 9.637185577660582e-05, + "loss": 0.3246, + "step": 270 + }, + { + "epoch": 0.45004930198764853, + "grad_norm": 162.41033935546875, + "learning_rate": 9.63382896764951e-05, + "loss": 0.3495, + "step": 271 + }, + { + "epoch": 0.4517100005189683, + "grad_norm": 10.170032501220703, + "learning_rate": 9.630457492106601e-05, + "loss": 0.3059, + "step": 272 + }, + { + "epoch": 0.45337069905028804, + "grad_norm": 48.172279357910156, + "learning_rate": 9.627071161847698e-05, + "loss": 0.3275, + "step": 273 + }, + { + "epoch": 0.45503139758160777, + "grad_norm": 46.062015533447266, + "learning_rate": 9.623669987736305e-05, + "loss": 0.3063, + "step": 274 + }, + { + "epoch": 0.4566920961129275, + "grad_norm": 39.71500015258789, + "learning_rate": 9.620253980683538e-05, + "loss": 0.3077, + "step": 275 + }, + { + "epoch": 0.4583527946442472, + "grad_norm": 83.11470794677734, + "learning_rate": 9.616823151648103e-05, + "loss": 0.3212, + "step": 276 + }, + { + "epoch": 0.46001349317556695, + "grad_norm": 328.2074890136719, + "learning_rate": 9.613377511636253e-05, + "loss": 0.3303, + "step": 277 + }, + { + "epoch": 0.4616741917068867, + "grad_norm": 84.802734375, + "learning_rate": 9.609917071701755e-05, + "loss": 0.3139, + "step": 278 + }, + { + "epoch": 0.46333489023820645, + "grad_norm": 13.519368171691895, + "learning_rate": 9.60644184294586e-05, + "loss": 0.3133, + "step": 279 + }, + { + "epoch": 0.4649955887695262, + "grad_norm": 33.07325744628906, + "learning_rate": 9.602951836517255e-05, + "loss": 0.3203, + "step": 280 + }, + { + "epoch": 0.4666562873008459, + "grad_norm": 29.523157119750977, + "learning_rate": 9.599447063612037e-05, + "loss": 0.3154, + "step": 281 + }, + { + "epoch": 0.46831698583216563, + "grad_norm": 28.713653564453125, + "learning_rate": 9.59592753547368e-05, + "loss": 0.308, + "step": 282 + }, + { + "epoch": 0.4699776843634854, + "grad_norm": 11.922811508178711, + "learning_rate": 9.592393263392982e-05, + "loss": 0.2987, + "step": 283 + }, + { + "epoch": 0.47163838289480514, + "grad_norm": 10.632648468017578, + "learning_rate": 9.588844258708052e-05, + "loss": 0.3011, + "step": 284 + }, + { + "epoch": 0.47329908142612487, + "grad_norm": 17.077857971191406, + "learning_rate": 9.585280532804255e-05, + "loss": 0.322, + "step": 285 + }, + { + "epoch": 0.4749597799574446, + "grad_norm": 20.53763771057129, + "learning_rate": 9.581702097114185e-05, + "loss": 0.3215, + "step": 286 + }, + { + "epoch": 0.4766204784887643, + "grad_norm": 16.649274826049805, + "learning_rate": 9.578108963117623e-05, + "loss": 0.3085, + "step": 287 + }, + { + "epoch": 0.47828117702008405, + "grad_norm": 10.379050254821777, + "learning_rate": 9.574501142341504e-05, + "loss": 0.3358, + "step": 288 + }, + { + "epoch": 0.47994187555140383, + "grad_norm": 38.27133560180664, + "learning_rate": 9.570878646359882e-05, + "loss": 0.3157, + "step": 289 + }, + { + "epoch": 0.48160257408272356, + "grad_norm": 43.45355987548828, + "learning_rate": 9.567241486793886e-05, + "loss": 0.3394, + "step": 290 + }, + { + "epoch": 0.4832632726140433, + "grad_norm": 32.14323425292969, + "learning_rate": 9.563589675311686e-05, + "loss": 0.3084, + "step": 291 + }, + { + "epoch": 0.484923971145363, + "grad_norm": 15.220690727233887, + "learning_rate": 9.55992322362846e-05, + "loss": 0.3187, + "step": 292 + }, + { + "epoch": 0.48658466967668274, + "grad_norm": 16.599641799926758, + "learning_rate": 9.556242143506347e-05, + "loss": 0.2902, + "step": 293 + }, + { + "epoch": 0.4882453682080025, + "grad_norm": 8.873578071594238, + "learning_rate": 9.552546446754422e-05, + "loss": 0.2988, + "step": 294 + }, + { + "epoch": 0.48990606673932224, + "grad_norm": 41.431846618652344, + "learning_rate": 9.548836145228644e-05, + "loss": 0.3352, + "step": 295 + }, + { + "epoch": 0.49156676527064197, + "grad_norm": 15.18100357055664, + "learning_rate": 9.545111250831825e-05, + "loss": 0.3119, + "step": 296 + }, + { + "epoch": 0.4932274638019617, + "grad_norm": 19.954790115356445, + "learning_rate": 9.541371775513597e-05, + "loss": 0.3342, + "step": 297 + }, + { + "epoch": 0.4948881623332814, + "grad_norm": 22.85648536682129, + "learning_rate": 9.537617731270365e-05, + "loss": 0.3189, + "step": 298 + }, + { + "epoch": 0.49654886086460115, + "grad_norm": 82.43091583251953, + "learning_rate": 9.53384913014527e-05, + "loss": 0.3155, + "step": 299 + }, + { + "epoch": 0.49820955939592093, + "grad_norm": 72.58283996582031, + "learning_rate": 9.530065984228156e-05, + "loss": 0.333, + "step": 300 + }, + { + "epoch": 0.49987025792724066, + "grad_norm": 56.16787338256836, + "learning_rate": 9.526268305655523e-05, + "loss": 0.3073, + "step": 301 + }, + { + "epoch": 0.5015309564585604, + "grad_norm": 16.05253028869629, + "learning_rate": 9.522456106610498e-05, + "loss": 0.3323, + "step": 302 + }, + { + "epoch": 0.5031916549898802, + "grad_norm": 13.669990539550781, + "learning_rate": 9.518629399322784e-05, + "loss": 0.321, + "step": 303 + }, + { + "epoch": 0.5048523535211998, + "grad_norm": 143.93814086914062, + "learning_rate": 9.514788196068632e-05, + "loss": 0.2954, + "step": 304 + }, + { + "epoch": 0.5065130520525196, + "grad_norm": 8.028514862060547, + "learning_rate": 9.510932509170793e-05, + "loss": 0.3144, + "step": 305 + }, + { + "epoch": 0.5081737505838393, + "grad_norm": 26.369863510131836, + "learning_rate": 9.507062350998486e-05, + "loss": 0.3242, + "step": 306 + }, + { + "epoch": 0.5098344491151591, + "grad_norm": 29.293994903564453, + "learning_rate": 9.503177733967352e-05, + "loss": 0.2978, + "step": 307 + }, + { + "epoch": 0.5114951476464789, + "grad_norm": 41.09022521972656, + "learning_rate": 9.499278670539418e-05, + "loss": 0.2918, + "step": 308 + }, + { + "epoch": 0.5131558461777985, + "grad_norm": 21.166954040527344, + "learning_rate": 9.495365173223052e-05, + "loss": 0.3123, + "step": 309 + }, + { + "epoch": 0.5148165447091183, + "grad_norm": 16.879657745361328, + "learning_rate": 9.491437254572933e-05, + "loss": 0.3024, + "step": 310 + }, + { + "epoch": 0.516477243240438, + "grad_norm": 9.364297866821289, + "learning_rate": 9.487494927189998e-05, + "loss": 0.3113, + "step": 311 + }, + { + "epoch": 0.5181379417717578, + "grad_norm": 13.823391914367676, + "learning_rate": 9.483538203721414e-05, + "loss": 0.3281, + "step": 312 + }, + { + "epoch": 0.5197986403030774, + "grad_norm": 10.906787872314453, + "learning_rate": 9.479567096860524e-05, + "loss": 0.3097, + "step": 313 + }, + { + "epoch": 0.5214593388343972, + "grad_norm": 16.87537384033203, + "learning_rate": 9.47558161934682e-05, + "loss": 0.3303, + "step": 314 + }, + { + "epoch": 0.523120037365717, + "grad_norm": 15.128436088562012, + "learning_rate": 9.471581783965891e-05, + "loss": 0.3099, + "step": 315 + }, + { + "epoch": 0.5247807358970367, + "grad_norm": 15.296138763427734, + "learning_rate": 9.467567603549393e-05, + "loss": 0.2922, + "step": 316 + }, + { + "epoch": 0.5264414344283564, + "grad_norm": 26.614158630371094, + "learning_rate": 9.463539090974993e-05, + "loss": 0.3316, + "step": 317 + }, + { + "epoch": 0.5281021329596761, + "grad_norm": 41.7516975402832, + "learning_rate": 9.459496259166342e-05, + "loss": 0.3016, + "step": 318 + }, + { + "epoch": 0.5297628314909959, + "grad_norm": 32.64419937133789, + "learning_rate": 9.455439121093029e-05, + "loss": 0.3112, + "step": 319 + }, + { + "epoch": 0.5314235300223157, + "grad_norm": 66.82496643066406, + "learning_rate": 9.451367689770532e-05, + "loss": 0.3278, + "step": 320 + }, + { + "epoch": 0.5330842285536354, + "grad_norm": 20.874038696289062, + "learning_rate": 9.447281978260187e-05, + "loss": 0.32, + "step": 321 + }, + { + "epoch": 0.5347449270849551, + "grad_norm": 70.21574401855469, + "learning_rate": 9.443181999669142e-05, + "loss": 0.3101, + "step": 322 + }, + { + "epoch": 0.5364056256162748, + "grad_norm": 15.442045211791992, + "learning_rate": 9.43906776715031e-05, + "loss": 0.3257, + "step": 323 + }, + { + "epoch": 0.5380663241475946, + "grad_norm": 77.55043029785156, + "learning_rate": 9.434939293902335e-05, + "loss": 0.3168, + "step": 324 + }, + { + "epoch": 0.5397270226789144, + "grad_norm": 12.460114479064941, + "learning_rate": 9.430796593169546e-05, + "loss": 0.3322, + "step": 325 + }, + { + "epoch": 0.541387721210234, + "grad_norm": 11.248091697692871, + "learning_rate": 9.426639678241913e-05, + "loss": 0.3208, + "step": 326 + }, + { + "epoch": 0.5430484197415538, + "grad_norm": 30.707332611083984, + "learning_rate": 9.422468562455005e-05, + "loss": 0.3264, + "step": 327 + }, + { + "epoch": 0.5447091182728735, + "grad_norm": 16.947376251220703, + "learning_rate": 9.41828325918995e-05, + "loss": 0.3025, + "step": 328 + }, + { + "epoch": 0.5463698168041933, + "grad_norm": 21.29446029663086, + "learning_rate": 9.414083781873391e-05, + "loss": 0.325, + "step": 329 + }, + { + "epoch": 0.548030515335513, + "grad_norm": 12.905153274536133, + "learning_rate": 9.409870143977437e-05, + "loss": 0.2891, + "step": 330 + }, + { + "epoch": 0.5496912138668327, + "grad_norm": 9.222719192504883, + "learning_rate": 9.405642359019628e-05, + "loss": 0.3124, + "step": 331 + }, + { + "epoch": 0.5513519123981525, + "grad_norm": 84.41971588134766, + "learning_rate": 9.401400440562892e-05, + "loss": 0.2968, + "step": 332 + }, + { + "epoch": 0.5530126109294722, + "grad_norm": 8.102546691894531, + "learning_rate": 9.397144402215493e-05, + "loss": 0.3009, + "step": 333 + }, + { + "epoch": 0.554673309460792, + "grad_norm": 16.299074172973633, + "learning_rate": 9.392874257630988e-05, + "loss": 0.3156, + "step": 334 + }, + { + "epoch": 0.5563340079921116, + "grad_norm": 10.908889770507812, + "learning_rate": 9.388590020508201e-05, + "loss": 0.2771, + "step": 335 + }, + { + "epoch": 0.5579947065234314, + "grad_norm": 37.578548431396484, + "learning_rate": 9.384291704591153e-05, + "loss": 0.324, + "step": 336 + }, + { + "epoch": 0.5596554050547512, + "grad_norm": 98.8309097290039, + "learning_rate": 9.379979323669036e-05, + "loss": 0.3144, + "step": 337 + }, + { + "epoch": 0.5613161035860709, + "grad_norm": 123.9063491821289, + "learning_rate": 9.375652891576163e-05, + "loss": 0.2922, + "step": 338 + }, + { + "epoch": 0.5629768021173907, + "grad_norm": 40.04903793334961, + "learning_rate": 9.37131242219192e-05, + "loss": 0.2919, + "step": 339 + }, + { + "epoch": 0.5646375006487103, + "grad_norm": 30.752716064453125, + "learning_rate": 9.366957929440731e-05, + "loss": 0.3177, + "step": 340 + }, + { + "epoch": 0.5662981991800301, + "grad_norm": 14.724492073059082, + "learning_rate": 9.362589427292004e-05, + "loss": 0.3196, + "step": 341 + }, + { + "epoch": 0.5679588977113499, + "grad_norm": 38.862152099609375, + "learning_rate": 9.358206929760087e-05, + "loss": 0.313, + "step": 342 + }, + { + "epoch": 0.5696195962426696, + "grad_norm": 83.72445678710938, + "learning_rate": 9.353810450904234e-05, + "loss": 0.3187, + "step": 343 + }, + { + "epoch": 0.5712802947739893, + "grad_norm": 41.9790153503418, + "learning_rate": 9.349400004828547e-05, + "loss": 0.322, + "step": 344 + }, + { + "epoch": 0.572940993305309, + "grad_norm": 15.165878295898438, + "learning_rate": 9.344975605681932e-05, + "loss": 0.3211, + "step": 345 + }, + { + "epoch": 0.5746016918366288, + "grad_norm": 20.742877960205078, + "learning_rate": 9.340537267658063e-05, + "loss": 0.3108, + "step": 346 + }, + { + "epoch": 0.5762623903679486, + "grad_norm": 17.051950454711914, + "learning_rate": 9.336085004995327e-05, + "loss": 0.3165, + "step": 347 + }, + { + "epoch": 0.5779230888992682, + "grad_norm": 10.417563438415527, + "learning_rate": 9.331618831976783e-05, + "loss": 0.3139, + "step": 348 + }, + { + "epoch": 0.579583787430588, + "grad_norm": 18.770387649536133, + "learning_rate": 9.327138762930114e-05, + "loss": 0.2999, + "step": 349 + }, + { + "epoch": 0.5812444859619077, + "grad_norm": 10.266261100769043, + "learning_rate": 9.322644812227588e-05, + "loss": 0.314, + "step": 350 + }, + { + "epoch": 0.5829051844932275, + "grad_norm": 20.65653419494629, + "learning_rate": 9.318136994285996e-05, + "loss": 0.2831, + "step": 351 + }, + { + "epoch": 0.5845658830245472, + "grad_norm": 47.8186149597168, + "learning_rate": 9.31361532356662e-05, + "loss": 0.2987, + "step": 352 + }, + { + "epoch": 0.5862265815558669, + "grad_norm": 31.965633392333984, + "learning_rate": 9.309079814575187e-05, + "loss": 0.3051, + "step": 353 + }, + { + "epoch": 0.5878872800871867, + "grad_norm": 11.494510650634766, + "learning_rate": 9.304530481861813e-05, + "loss": 0.3483, + "step": 354 + }, + { + "epoch": 0.5895479786185064, + "grad_norm": 20.134716033935547, + "learning_rate": 9.299967340020961e-05, + "loss": 0.3001, + "step": 355 + }, + { + "epoch": 0.5912086771498262, + "grad_norm": 19.050811767578125, + "learning_rate": 9.295390403691395e-05, + "loss": 0.3057, + "step": 356 + }, + { + "epoch": 0.5928693756811458, + "grad_norm": 20.096332550048828, + "learning_rate": 9.290799687556136e-05, + "loss": 0.2807, + "step": 357 + }, + { + "epoch": 0.5945300742124656, + "grad_norm": 31.533571243286133, + "learning_rate": 9.286195206342403e-05, + "loss": 0.3222, + "step": 358 + }, + { + "epoch": 0.5961907727437854, + "grad_norm": 39.516456604003906, + "learning_rate": 9.281576974821582e-05, + "loss": 0.3148, + "step": 359 + }, + { + "epoch": 0.5978514712751051, + "grad_norm": 63.63636016845703, + "learning_rate": 9.276945007809166e-05, + "loss": 0.3071, + "step": 360 + }, + { + "epoch": 0.5995121698064249, + "grad_norm": 25.25543785095215, + "learning_rate": 9.272299320164716e-05, + "loss": 0.3111, + "step": 361 + }, + { + "epoch": 0.6011728683377445, + "grad_norm": 21.786413192749023, + "learning_rate": 9.267639926791805e-05, + "loss": 0.3175, + "step": 362 + }, + { + "epoch": 0.6028335668690643, + "grad_norm": 37.87742233276367, + "learning_rate": 9.262966842637977e-05, + "loss": 0.3068, + "step": 363 + }, + { + "epoch": 0.6044942654003841, + "grad_norm": 57.43134307861328, + "learning_rate": 9.258280082694694e-05, + "loss": 0.3342, + "step": 364 + }, + { + "epoch": 0.6061549639317038, + "grad_norm": 17.7891845703125, + "learning_rate": 9.253579661997298e-05, + "loss": 0.2971, + "step": 365 + }, + { + "epoch": 0.6078156624630235, + "grad_norm": 51.276214599609375, + "learning_rate": 9.248865595624947e-05, + "loss": 0.315, + "step": 366 + }, + { + "epoch": 0.6094763609943432, + "grad_norm": 37.59144973754883, + "learning_rate": 9.244137898700578e-05, + "loss": 0.3222, + "step": 367 + }, + { + "epoch": 0.611137059525663, + "grad_norm": 9.40713119506836, + "learning_rate": 9.239396586390854e-05, + "loss": 0.3071, + "step": 368 + }, + { + "epoch": 0.6127977580569827, + "grad_norm": 12.5270414352417, + "learning_rate": 9.234641673906121e-05, + "loss": 0.3009, + "step": 369 + }, + { + "epoch": 0.6144584565883024, + "grad_norm": 29.016000747680664, + "learning_rate": 9.229873176500351e-05, + "loss": 0.3078, + "step": 370 + }, + { + "epoch": 0.6161191551196222, + "grad_norm": 19.265480041503906, + "learning_rate": 9.2250911094711e-05, + "loss": 0.2885, + "step": 371 + }, + { + "epoch": 0.6177798536509419, + "grad_norm": 252.14859008789062, + "learning_rate": 9.22029548815945e-05, + "loss": 0.2998, + "step": 372 + }, + { + "epoch": 0.6194405521822617, + "grad_norm": 10.757695198059082, + "learning_rate": 9.215486327949974e-05, + "loss": 0.3173, + "step": 373 + }, + { + "epoch": 0.6211012507135814, + "grad_norm": 10.930907249450684, + "learning_rate": 9.210663644270671e-05, + "loss": 0.3065, + "step": 374 + }, + { + "epoch": 0.6227619492449011, + "grad_norm": 64.07231903076172, + "learning_rate": 9.205827452592932e-05, + "loss": 0.2784, + "step": 375 + }, + { + "epoch": 0.6244226477762209, + "grad_norm": 24.454105377197266, + "learning_rate": 9.200977768431474e-05, + "loss": 0.2876, + "step": 376 + }, + { + "epoch": 0.6260833463075406, + "grad_norm": 32.62709045410156, + "learning_rate": 9.196114607344301e-05, + "loss": 0.3075, + "step": 377 + }, + { + "epoch": 0.6277440448388604, + "grad_norm": 51.31664276123047, + "learning_rate": 9.191237984932656e-05, + "loss": 0.3016, + "step": 378 + }, + { + "epoch": 0.62940474337018, + "grad_norm": 8.667896270751953, + "learning_rate": 9.18634791684096e-05, + "loss": 0.3026, + "step": 379 + }, + { + "epoch": 0.6310654419014998, + "grad_norm": 21.055294036865234, + "learning_rate": 9.181444418756773e-05, + "loss": 0.3158, + "step": 380 + }, + { + "epoch": 0.6327261404328196, + "grad_norm": 19.304393768310547, + "learning_rate": 9.17652750641074e-05, + "loss": 0.3075, + "step": 381 + }, + { + "epoch": 0.6343868389641393, + "grad_norm": 78.29962921142578, + "learning_rate": 9.171597195576531e-05, + "loss": 0.3152, + "step": 382 + }, + { + "epoch": 0.6360475374954591, + "grad_norm": 30.911787033081055, + "learning_rate": 9.16665350207081e-05, + "loss": 0.3139, + "step": 383 + }, + { + "epoch": 0.6377082360267787, + "grad_norm": 59.8438720703125, + "learning_rate": 9.161696441753167e-05, + "loss": 0.311, + "step": 384 + }, + { + "epoch": 0.6393689345580985, + "grad_norm": 15.886602401733398, + "learning_rate": 9.156726030526073e-05, + "loss": 0.316, + "step": 385 + }, + { + "epoch": 0.6410296330894182, + "grad_norm": 19.520782470703125, + "learning_rate": 9.151742284334833e-05, + "loss": 0.3018, + "step": 386 + }, + { + "epoch": 0.642690331620738, + "grad_norm": 17.10700798034668, + "learning_rate": 9.146745219167527e-05, + "loss": 0.3165, + "step": 387 + }, + { + "epoch": 0.6443510301520577, + "grad_norm": 8.914456367492676, + "learning_rate": 9.141734851054968e-05, + "loss": 0.3059, + "step": 388 + }, + { + "epoch": 0.6460117286833774, + "grad_norm": 6.80755090713501, + "learning_rate": 9.136711196070639e-05, + "loss": 0.3257, + "step": 389 + }, + { + "epoch": 0.6476724272146972, + "grad_norm": 28.446495056152344, + "learning_rate": 9.131674270330653e-05, + "loss": 0.3065, + "step": 390 + }, + { + "epoch": 0.6493331257460169, + "grad_norm": 25.92674446105957, + "learning_rate": 9.126624089993694e-05, + "loss": 0.2975, + "step": 391 + }, + { + "epoch": 0.6509938242773367, + "grad_norm": 15.610758781433105, + "learning_rate": 9.121560671260968e-05, + "loss": 0.3179, + "step": 392 + }, + { + "epoch": 0.6526545228086564, + "grad_norm": 12.867768287658691, + "learning_rate": 9.116484030376149e-05, + "loss": 0.3234, + "step": 393 + }, + { + "epoch": 0.6543152213399761, + "grad_norm": 23.178924560546875, + "learning_rate": 9.111394183625332e-05, + "loss": 0.2968, + "step": 394 + }, + { + "epoch": 0.6559759198712959, + "grad_norm": 10.973976135253906, + "learning_rate": 9.106291147336971e-05, + "loss": 0.3129, + "step": 395 + }, + { + "epoch": 0.6576366184026156, + "grad_norm": 41.91653060913086, + "learning_rate": 9.10117493788184e-05, + "loss": 0.2994, + "step": 396 + }, + { + "epoch": 0.6592973169339353, + "grad_norm": 37.21331024169922, + "learning_rate": 9.096045571672965e-05, + "loss": 0.3072, + "step": 397 + }, + { + "epoch": 0.6609580154652551, + "grad_norm": 30.14063835144043, + "learning_rate": 9.090903065165586e-05, + "loss": 0.2966, + "step": 398 + }, + { + "epoch": 0.6626187139965748, + "grad_norm": 67.18895721435547, + "learning_rate": 9.085747434857095e-05, + "loss": 0.3145, + "step": 399 + }, + { + "epoch": 0.6642794125278946, + "grad_norm": 46.18681716918945, + "learning_rate": 9.080578697286985e-05, + "loss": 0.3023, + "step": 400 + }, + { + "epoch": 0.6659401110592142, + "grad_norm": 10.217764854431152, + "learning_rate": 9.075396869036801e-05, + "loss": 0.3192, + "step": 401 + }, + { + "epoch": 0.667600809590534, + "grad_norm": 417.52044677734375, + "learning_rate": 9.070201966730079e-05, + "loss": 0.2859, + "step": 402 + }, + { + "epoch": 0.6692615081218538, + "grad_norm": 26.046894073486328, + "learning_rate": 9.0649940070323e-05, + "loss": 0.3043, + "step": 403 + }, + { + "epoch": 0.6709222066531735, + "grad_norm": 34.24388122558594, + "learning_rate": 9.059773006650832e-05, + "loss": 0.2983, + "step": 404 + }, + { + "epoch": 0.6725829051844933, + "grad_norm": 13.467427253723145, + "learning_rate": 9.054538982334882e-05, + "loss": 0.2854, + "step": 405 + }, + { + "epoch": 0.6742436037158129, + "grad_norm": 28.355485916137695, + "learning_rate": 9.049291950875434e-05, + "loss": 0.3306, + "step": 406 + }, + { + "epoch": 0.6759043022471327, + "grad_norm": 9.620882034301758, + "learning_rate": 9.0440319291052e-05, + "loss": 0.3153, + "step": 407 + }, + { + "epoch": 0.6775650007784524, + "grad_norm": 11.979683876037598, + "learning_rate": 9.038758933898565e-05, + "loss": 0.3032, + "step": 408 + }, + { + "epoch": 0.6792256993097722, + "grad_norm": 185.97085571289062, + "learning_rate": 9.03347298217154e-05, + "loss": 0.327, + "step": 409 + }, + { + "epoch": 0.680886397841092, + "grad_norm": 76.22681427001953, + "learning_rate": 9.028174090881694e-05, + "loss": 0.2875, + "step": 410 + }, + { + "epoch": 0.6825470963724116, + "grad_norm": 18.044015884399414, + "learning_rate": 9.022862277028107e-05, + "loss": 0.317, + "step": 411 + }, + { + "epoch": 0.6842077949037314, + "grad_norm": 68.02251434326172, + "learning_rate": 9.017537557651316e-05, + "loss": 0.3086, + "step": 412 + }, + { + "epoch": 0.6858684934350511, + "grad_norm": 17.809236526489258, + "learning_rate": 9.012199949833264e-05, + "loss": 0.2943, + "step": 413 + }, + { + "epoch": 0.6875291919663709, + "grad_norm": 18.80552101135254, + "learning_rate": 9.006849470697236e-05, + "loss": 0.3134, + "step": 414 + }, + { + "epoch": 0.6891898904976906, + "grad_norm": 9.522897720336914, + "learning_rate": 9.001486137407809e-05, + "loss": 0.3197, + "step": 415 + }, + { + "epoch": 0.6908505890290103, + "grad_norm": 44.19784927368164, + "learning_rate": 8.996109967170795e-05, + "loss": 0.3134, + "step": 416 + }, + { + "epoch": 0.6925112875603301, + "grad_norm": 37.70685958862305, + "learning_rate": 8.990720977233194e-05, + "loss": 0.3135, + "step": 417 + }, + { + "epoch": 0.6941719860916498, + "grad_norm": 22.376129150390625, + "learning_rate": 8.985319184883128e-05, + "loss": 0.3198, + "step": 418 + }, + { + "epoch": 0.6958326846229695, + "grad_norm": 38.35197448730469, + "learning_rate": 8.979904607449787e-05, + "loss": 0.3093, + "step": 419 + }, + { + "epoch": 0.6974933831542893, + "grad_norm": 38.13330841064453, + "learning_rate": 8.974477262303382e-05, + "loss": 0.3017, + "step": 420 + }, + { + "epoch": 0.699154081685609, + "grad_norm": 8.704827308654785, + "learning_rate": 8.969037166855078e-05, + "loss": 0.3089, + "step": 421 + }, + { + "epoch": 0.7008147802169288, + "grad_norm": 77.3583984375, + "learning_rate": 8.963584338556948e-05, + "loss": 0.3043, + "step": 422 + }, + { + "epoch": 0.7024754787482484, + "grad_norm": 59.96723175048828, + "learning_rate": 8.95811879490191e-05, + "loss": 0.2985, + "step": 423 + }, + { + "epoch": 0.7041361772795682, + "grad_norm": 56.64527893066406, + "learning_rate": 8.952640553423672e-05, + "loss": 0.3369, + "step": 424 + }, + { + "epoch": 0.7057968758108879, + "grad_norm": 37.24342727661133, + "learning_rate": 8.947149631696682e-05, + "loss": 0.3117, + "step": 425 + }, + { + "epoch": 0.7074575743422077, + "grad_norm": 137.05282592773438, + "learning_rate": 8.941646047336064e-05, + "loss": 0.2958, + "step": 426 + }, + { + "epoch": 0.7091182728735275, + "grad_norm": 34.54379653930664, + "learning_rate": 8.936129817997562e-05, + "loss": 0.2983, + "step": 427 + }, + { + "epoch": 0.7107789714048471, + "grad_norm": 17.301864624023438, + "learning_rate": 8.930600961377489e-05, + "loss": 0.336, + "step": 428 + }, + { + "epoch": 0.7124396699361669, + "grad_norm": 83.31056213378906, + "learning_rate": 8.925059495212667e-05, + "loss": 0.3098, + "step": 429 + }, + { + "epoch": 0.7141003684674866, + "grad_norm": 13.80082893371582, + "learning_rate": 8.919505437280367e-05, + "loss": 0.2986, + "step": 430 + }, + { + "epoch": 0.7157610669988064, + "grad_norm": 9.891674995422363, + "learning_rate": 8.913938805398257e-05, + "loss": 0.3033, + "step": 431 + }, + { + "epoch": 0.7174217655301262, + "grad_norm": 20.786617279052734, + "learning_rate": 8.908359617424345e-05, + "loss": 0.2966, + "step": 432 + }, + { + "epoch": 0.7190824640614458, + "grad_norm": 11.70949935913086, + "learning_rate": 8.902767891256912e-05, + "loss": 0.3106, + "step": 433 + }, + { + "epoch": 0.7207431625927656, + "grad_norm": 20.220985412597656, + "learning_rate": 8.897163644834472e-05, + "loss": 0.2826, + "step": 434 + }, + { + "epoch": 0.7224038611240853, + "grad_norm": 94.52996063232422, + "learning_rate": 8.891546896135698e-05, + "loss": 0.3326, + "step": 435 + }, + { + "epoch": 0.7240645596554051, + "grad_norm": 14.206509590148926, + "learning_rate": 8.885917663179372e-05, + "loss": 0.3113, + "step": 436 + }, + { + "epoch": 0.7257252581867248, + "grad_norm": 9.238081932067871, + "learning_rate": 8.880275964024327e-05, + "loss": 0.3085, + "step": 437 + }, + { + "epoch": 0.7273859567180445, + "grad_norm": 16.078716278076172, + "learning_rate": 8.874621816769388e-05, + "loss": 0.3045, + "step": 438 + }, + { + "epoch": 0.7290466552493643, + "grad_norm": 226.89561462402344, + "learning_rate": 8.868955239553316e-05, + "loss": 0.3137, + "step": 439 + }, + { + "epoch": 0.730707353780684, + "grad_norm": 19.5429630279541, + "learning_rate": 8.863276250554739e-05, + "loss": 0.2968, + "step": 440 + }, + { + "epoch": 0.7323680523120037, + "grad_norm": 10.357669830322266, + "learning_rate": 8.857584867992116e-05, + "loss": 0.3218, + "step": 441 + }, + { + "epoch": 0.7340287508433234, + "grad_norm": 52.78630447387695, + "learning_rate": 8.851881110123657e-05, + "loss": 0.3225, + "step": 442 + }, + { + "epoch": 0.7356894493746432, + "grad_norm": 21.655752182006836, + "learning_rate": 8.846164995247272e-05, + "loss": 0.3077, + "step": 443 + }, + { + "epoch": 0.737350147905963, + "grad_norm": 31.048274993896484, + "learning_rate": 8.840436541700516e-05, + "loss": 0.313, + "step": 444 + }, + { + "epoch": 0.7390108464372827, + "grad_norm": 19.989078521728516, + "learning_rate": 8.834695767860525e-05, + "loss": 0.3181, + "step": 445 + }, + { + "epoch": 0.7406715449686024, + "grad_norm": 8.132134437561035, + "learning_rate": 8.82894269214396e-05, + "loss": 0.3246, + "step": 446 + }, + { + "epoch": 0.7423322434999221, + "grad_norm": 20.8494930267334, + "learning_rate": 8.823177333006947e-05, + "loss": 0.3094, + "step": 447 + }, + { + "epoch": 0.7439929420312419, + "grad_norm": 191.60983276367188, + "learning_rate": 8.817399708945017e-05, + "loss": 0.3023, + "step": 448 + }, + { + "epoch": 0.7456536405625617, + "grad_norm": 43.99884033203125, + "learning_rate": 8.811609838493049e-05, + "loss": 0.3052, + "step": 449 + }, + { + "epoch": 0.7473143390938813, + "grad_norm": 19.382110595703125, + "learning_rate": 8.805807740225206e-05, + "loss": 0.3363, + "step": 450 + }, + { + "epoch": 0.7489750376252011, + "grad_norm": 27.73593521118164, + "learning_rate": 8.799993432754884e-05, + "loss": 0.3034, + "step": 451 + }, + { + "epoch": 0.7506357361565208, + "grad_norm": 55.53531265258789, + "learning_rate": 8.79416693473464e-05, + "loss": 0.3143, + "step": 452 + }, + { + "epoch": 0.7522964346878406, + "grad_norm": 36.05241394042969, + "learning_rate": 8.788328264856143e-05, + "loss": 0.2937, + "step": 453 + }, + { + "epoch": 0.7539571332191604, + "grad_norm": 24.02420425415039, + "learning_rate": 8.782477441850106e-05, + "loss": 0.2891, + "step": 454 + }, + { + "epoch": 0.75561783175048, + "grad_norm": 23.53626823425293, + "learning_rate": 8.776614484486237e-05, + "loss": 0.3085, + "step": 455 + }, + { + "epoch": 0.7572785302817998, + "grad_norm": 21.195833206176758, + "learning_rate": 8.770739411573163e-05, + "loss": 0.2826, + "step": 456 + }, + { + "epoch": 0.7589392288131195, + "grad_norm": 49.207733154296875, + "learning_rate": 8.764852241958386e-05, + "loss": 0.276, + "step": 457 + }, + { + "epoch": 0.7605999273444393, + "grad_norm": 10.246028900146484, + "learning_rate": 8.758952994528208e-05, + "loss": 0.3145, + "step": 458 + }, + { + "epoch": 0.762260625875759, + "grad_norm": 90.51976776123047, + "learning_rate": 8.75304168820768e-05, + "loss": 0.3293, + "step": 459 + }, + { + "epoch": 0.7639213244070787, + "grad_norm": 50.458980560302734, + "learning_rate": 8.747118341960542e-05, + "loss": 0.3052, + "step": 460 + }, + { + "epoch": 0.7655820229383985, + "grad_norm": 45.22133255004883, + "learning_rate": 8.741182974789151e-05, + "loss": 0.3233, + "step": 461 + }, + { + "epoch": 0.7672427214697182, + "grad_norm": 15.685876846313477, + "learning_rate": 8.735235605734435e-05, + "loss": 0.3001, + "step": 462 + }, + { + "epoch": 0.768903420001038, + "grad_norm": 70.90806579589844, + "learning_rate": 8.729276253875823e-05, + "loss": 0.2961, + "step": 463 + }, + { + "epoch": 0.7705641185323576, + "grad_norm": 197.4364013671875, + "learning_rate": 8.723304938331178e-05, + "loss": 0.2954, + "step": 464 + }, + { + "epoch": 0.7722248170636774, + "grad_norm": 29.27678108215332, + "learning_rate": 8.717321678256755e-05, + "loss": 0.294, + "step": 465 + }, + { + "epoch": 0.7738855155949972, + "grad_norm": 19.00115966796875, + "learning_rate": 8.711326492847119e-05, + "loss": 0.2973, + "step": 466 + }, + { + "epoch": 0.7755462141263169, + "grad_norm": 16.248558044433594, + "learning_rate": 8.705319401335093e-05, + "loss": 0.3069, + "step": 467 + }, + { + "epoch": 0.7772069126576366, + "grad_norm": 15.289970397949219, + "learning_rate": 8.6993004229917e-05, + "loss": 0.3069, + "step": 468 + }, + { + "epoch": 0.7788676111889563, + "grad_norm": 15.599447250366211, + "learning_rate": 8.69326957712609e-05, + "loss": 0.3039, + "step": 469 + }, + { + "epoch": 0.7805283097202761, + "grad_norm": 28.66286849975586, + "learning_rate": 8.687226883085493e-05, + "loss": 0.2926, + "step": 470 + }, + { + "epoch": 0.7821890082515959, + "grad_norm": 11.616494178771973, + "learning_rate": 8.681172360255139e-05, + "loss": 0.2965, + "step": 471 + }, + { + "epoch": 0.7838497067829155, + "grad_norm": 8.636465072631836, + "learning_rate": 8.675106028058209e-05, + "loss": 0.3105, + "step": 472 + }, + { + "epoch": 0.7855104053142353, + "grad_norm": 22.446075439453125, + "learning_rate": 8.669027905955773e-05, + "loss": 0.322, + "step": 473 + }, + { + "epoch": 0.787171103845555, + "grad_norm": 26.431161880493164, + "learning_rate": 8.66293801344672e-05, + "loss": 0.2909, + "step": 474 + }, + { + "epoch": 0.7888318023768748, + "grad_norm": 9.34886360168457, + "learning_rate": 8.656836370067696e-05, + "loss": 0.3115, + "step": 475 + }, + { + "epoch": 0.7904925009081946, + "grad_norm": 10.392329216003418, + "learning_rate": 8.650722995393051e-05, + "loss": 0.3032, + "step": 476 + }, + { + "epoch": 0.7921531994395142, + "grad_norm": 16.351160049438477, + "learning_rate": 8.644597909034762e-05, + "loss": 0.2805, + "step": 477 + }, + { + "epoch": 0.793813897970834, + "grad_norm": 11.851846694946289, + "learning_rate": 8.638461130642386e-05, + "loss": 0.302, + "step": 478 + }, + { + "epoch": 0.7954745965021537, + "grad_norm": 13.054351806640625, + "learning_rate": 8.632312679902983e-05, + "loss": 0.3173, + "step": 479 + }, + { + "epoch": 0.7971352950334735, + "grad_norm": 13.70720386505127, + "learning_rate": 8.626152576541057e-05, + "loss": 0.3124, + "step": 480 + }, + { + "epoch": 0.7987959935647931, + "grad_norm": 15.25360107421875, + "learning_rate": 8.6199808403185e-05, + "loss": 0.3206, + "step": 481 + }, + { + "epoch": 0.8004566920961129, + "grad_norm": 8.500411987304688, + "learning_rate": 8.61379749103452e-05, + "loss": 0.2843, + "step": 482 + }, + { + "epoch": 0.8021173906274327, + "grad_norm": 93.16901397705078, + "learning_rate": 8.607602548525575e-05, + "loss": 0.323, + "step": 483 + }, + { + "epoch": 0.8037780891587524, + "grad_norm": 51.738250732421875, + "learning_rate": 8.601396032665323e-05, + "loss": 0.2966, + "step": 484 + }, + { + "epoch": 0.8054387876900722, + "grad_norm": 18.019519805908203, + "learning_rate": 8.595177963364547e-05, + "loss": 0.3195, + "step": 485 + }, + { + "epoch": 0.8070994862213918, + "grad_norm": 9.113760948181152, + "learning_rate": 8.58894836057109e-05, + "loss": 0.2879, + "step": 486 + }, + { + "epoch": 0.8087601847527116, + "grad_norm": 31.719749450683594, + "learning_rate": 8.5827072442698e-05, + "loss": 0.2877, + "step": 487 + }, + { + "epoch": 0.8104208832840314, + "grad_norm": 48.240909576416016, + "learning_rate": 8.576454634482458e-05, + "loss": 0.3141, + "step": 488 + }, + { + "epoch": 0.8120815818153511, + "grad_norm": 23.986005783081055, + "learning_rate": 8.570190551267721e-05, + "loss": 0.2991, + "step": 489 + }, + { + "epoch": 0.8137422803466708, + "grad_norm": 14.616287231445312, + "learning_rate": 8.563915014721044e-05, + "loss": 0.2967, + "step": 490 + }, + { + "epoch": 0.8154029788779905, + "grad_norm": 32.80522155761719, + "learning_rate": 8.557628044974635e-05, + "loss": 0.3066, + "step": 491 + }, + { + "epoch": 0.8170636774093103, + "grad_norm": 12.438755989074707, + "learning_rate": 8.551329662197373e-05, + "loss": 0.298, + "step": 492 + }, + { + "epoch": 0.8187243759406301, + "grad_norm": 18.6926212310791, + "learning_rate": 8.545019886594757e-05, + "loss": 0.3053, + "step": 493 + }, + { + "epoch": 0.8203850744719497, + "grad_norm": 90.16177368164062, + "learning_rate": 8.538698738408828e-05, + "loss": 0.3226, + "step": 494 + }, + { + "epoch": 0.8220457730032695, + "grad_norm": 28.3646240234375, + "learning_rate": 8.532366237918115e-05, + "loss": 0.3031, + "step": 495 + }, + { + "epoch": 0.8237064715345892, + "grad_norm": 12.563741683959961, + "learning_rate": 8.526022405437563e-05, + "loss": 0.2943, + "step": 496 + }, + { + "epoch": 0.825367170065909, + "grad_norm": 92.90933227539062, + "learning_rate": 8.519667261318474e-05, + "loss": 0.2923, + "step": 497 + }, + { + "epoch": 0.8270278685972287, + "grad_norm": 39.55075454711914, + "learning_rate": 8.513300825948436e-05, + "loss": 0.3128, + "step": 498 + }, + { + "epoch": 0.8286885671285484, + "grad_norm": 25.202983856201172, + "learning_rate": 8.50692311975126e-05, + "loss": 0.3119, + "step": 499 + }, + { + "epoch": 0.8303492656598682, + "grad_norm": 14.300535202026367, + "learning_rate": 8.500534163186913e-05, + "loss": 0.3338, + "step": 500 + }, + { + "epoch": 0.8320099641911879, + "grad_norm": 19.413766860961914, + "learning_rate": 8.494133976751456e-05, + "loss": 0.3261, + "step": 501 + }, + { + "epoch": 0.8336706627225077, + "grad_norm": 38.7458381652832, + "learning_rate": 8.487722580976976e-05, + "loss": 0.307, + "step": 502 + }, + { + "epoch": 0.8353313612538273, + "grad_norm": 15.395964622497559, + "learning_rate": 8.48129999643152e-05, + "loss": 0.3306, + "step": 503 + }, + { + "epoch": 0.8369920597851471, + "grad_norm": 10.697406768798828, + "learning_rate": 8.474866243719023e-05, + "loss": 0.3242, + "step": 504 + }, + { + "epoch": 0.8386527583164669, + "grad_norm": 15.346046447753906, + "learning_rate": 8.46842134347926e-05, + "loss": 0.3014, + "step": 505 + }, + { + "epoch": 0.8403134568477866, + "grad_norm": 14.454690933227539, + "learning_rate": 8.461965316387758e-05, + "loss": 0.304, + "step": 506 + }, + { + "epoch": 0.8419741553791064, + "grad_norm": 13.33731746673584, + "learning_rate": 8.455498183155741e-05, + "loss": 0.3166, + "step": 507 + }, + { + "epoch": 0.843634853910426, + "grad_norm": 18.96268081665039, + "learning_rate": 8.449019964530064e-05, + "loss": 0.3152, + "step": 508 + }, + { + "epoch": 0.8452955524417458, + "grad_norm": 32.26592254638672, + "learning_rate": 8.442530681293147e-05, + "loss": 0.3136, + "step": 509 + }, + { + "epoch": 0.8469562509730656, + "grad_norm": 66.2313232421875, + "learning_rate": 8.436030354262899e-05, + "loss": 0.2956, + "step": 510 + }, + { + "epoch": 0.8486169495043853, + "grad_norm": 18.61186981201172, + "learning_rate": 8.429519004292663e-05, + "loss": 0.3015, + "step": 511 + }, + { + "epoch": 0.850277648035705, + "grad_norm": 40.44683837890625, + "learning_rate": 8.422996652271142e-05, + "loss": 0.3052, + "step": 512 + }, + { + "epoch": 0.8519383465670247, + "grad_norm": 101.628662109375, + "learning_rate": 8.416463319122335e-05, + "loss": 0.2621, + "step": 513 + }, + { + "epoch": 0.8535990450983445, + "grad_norm": 216.36434936523438, + "learning_rate": 8.409919025805471e-05, + "loss": 0.2885, + "step": 514 + }, + { + "epoch": 0.8552597436296643, + "grad_norm": 23.783138275146484, + "learning_rate": 8.403363793314934e-05, + "loss": 0.3114, + "step": 515 + }, + { + "epoch": 0.856920442160984, + "grad_norm": 72.62994384765625, + "learning_rate": 8.396797642680205e-05, + "loss": 0.3092, + "step": 516 + }, + { + "epoch": 0.8585811406923037, + "grad_norm": 38.82360076904297, + "learning_rate": 8.390220594965795e-05, + "loss": 0.2772, + "step": 517 + }, + { + "epoch": 0.8602418392236234, + "grad_norm": 21.858617782592773, + "learning_rate": 8.383632671271163e-05, + "loss": 0.3067, + "step": 518 + }, + { + "epoch": 0.8619025377549432, + "grad_norm": 12.71886920928955, + "learning_rate": 8.377033892730668e-05, + "loss": 0.3048, + "step": 519 + }, + { + "epoch": 0.8635632362862629, + "grad_norm": 25.45595932006836, + "learning_rate": 8.37042428051349e-05, + "loss": 0.2966, + "step": 520 + }, + { + "epoch": 0.8652239348175826, + "grad_norm": 82.44416046142578, + "learning_rate": 8.363803855823557e-05, + "loss": 0.2903, + "step": 521 + }, + { + "epoch": 0.8668846333489024, + "grad_norm": 62.75419235229492, + "learning_rate": 8.357172639899497e-05, + "loss": 0.2738, + "step": 522 + }, + { + "epoch": 0.8685453318802221, + "grad_norm": 45.436100006103516, + "learning_rate": 8.350530654014541e-05, + "loss": 0.3127, + "step": 523 + }, + { + "epoch": 0.8702060304115419, + "grad_norm": 26.62226104736328, + "learning_rate": 8.343877919476486e-05, + "loss": 0.3179, + "step": 524 + }, + { + "epoch": 0.8718667289428615, + "grad_norm": 15.89708423614502, + "learning_rate": 8.337214457627597e-05, + "loss": 0.3103, + "step": 525 + }, + { + "epoch": 0.8735274274741813, + "grad_norm": 233.15623474121094, + "learning_rate": 8.330540289844566e-05, + "loss": 0.296, + "step": 526 + }, + { + "epoch": 0.8751881260055011, + "grad_norm": 29.75141716003418, + "learning_rate": 8.323855437538419e-05, + "loss": 0.3111, + "step": 527 + }, + { + "epoch": 0.8768488245368208, + "grad_norm": 17.699806213378906, + "learning_rate": 8.317159922154462e-05, + "loss": 0.3115, + "step": 528 + }, + { + "epoch": 0.8785095230681406, + "grad_norm": 10.695178031921387, + "learning_rate": 8.310453765172214e-05, + "loss": 0.3192, + "step": 529 + }, + { + "epoch": 0.8801702215994602, + "grad_norm": 35.274574279785156, + "learning_rate": 8.303736988105326e-05, + "loss": 0.2942, + "step": 530 + }, + { + "epoch": 0.88183092013078, + "grad_norm": 30.4547119140625, + "learning_rate": 8.29700961250152e-05, + "loss": 0.3074, + "step": 531 + }, + { + "epoch": 0.8834916186620998, + "grad_norm": 16.834020614624023, + "learning_rate": 8.290271659942524e-05, + "loss": 0.2884, + "step": 532 + }, + { + "epoch": 0.8851523171934195, + "grad_norm": 51.659488677978516, + "learning_rate": 8.283523152043987e-05, + "loss": 0.2899, + "step": 533 + }, + { + "epoch": 0.8868130157247393, + "grad_norm": 24.517641067504883, + "learning_rate": 8.276764110455431e-05, + "loss": 0.28, + "step": 534 + }, + { + "epoch": 0.8884737142560589, + "grad_norm": 13.82650089263916, + "learning_rate": 8.269994556860161e-05, + "loss": 0.2855, + "step": 535 + }, + { + "epoch": 0.8901344127873787, + "grad_norm": 42.67622375488281, + "learning_rate": 8.263214512975212e-05, + "loss": 0.3272, + "step": 536 + }, + { + "epoch": 0.8917951113186984, + "grad_norm": 17.009784698486328, + "learning_rate": 8.256424000551273e-05, + "loss": 0.312, + "step": 537 + }, + { + "epoch": 0.8934558098500182, + "grad_norm": 23.562740325927734, + "learning_rate": 8.249623041372607e-05, + "loss": 0.3135, + "step": 538 + }, + { + "epoch": 0.8951165083813379, + "grad_norm": 17.051776885986328, + "learning_rate": 8.242811657257003e-05, + "loss": 0.2932, + "step": 539 + }, + { + "epoch": 0.8967772069126576, + "grad_norm": 17.782793045043945, + "learning_rate": 8.235989870055683e-05, + "loss": 0.2845, + "step": 540 + }, + { + "epoch": 0.8984379054439774, + "grad_norm": 9.99154281616211, + "learning_rate": 8.229157701653251e-05, + "loss": 0.2965, + "step": 541 + }, + { + "epoch": 0.9000986039752971, + "grad_norm": 16.476123809814453, + "learning_rate": 8.222315173967608e-05, + "loss": 0.2997, + "step": 542 + }, + { + "epoch": 0.9017593025066168, + "grad_norm": 56.15431213378906, + "learning_rate": 8.215462308949893e-05, + "loss": 0.3019, + "step": 543 + }, + { + "epoch": 0.9034200010379366, + "grad_norm": 20.351224899291992, + "learning_rate": 8.208599128584402e-05, + "loss": 0.299, + "step": 544 + }, + { + "epoch": 0.9050806995692563, + "grad_norm": 20.28563690185547, + "learning_rate": 8.201725654888531e-05, + "loss": 0.2779, + "step": 545 + }, + { + "epoch": 0.9067413981005761, + "grad_norm": 31.785614013671875, + "learning_rate": 8.19484190991269e-05, + "loss": 0.3013, + "step": 546 + }, + { + "epoch": 0.9084020966318958, + "grad_norm": 163.7185821533203, + "learning_rate": 8.187947915740242e-05, + "loss": 0.3223, + "step": 547 + }, + { + "epoch": 0.9100627951632155, + "grad_norm": 34.03445053100586, + "learning_rate": 8.181043694487432e-05, + "loss": 0.2851, + "step": 548 + }, + { + "epoch": 0.9117234936945353, + "grad_norm": 22.548782348632812, + "learning_rate": 8.174129268303313e-05, + "loss": 0.3, + "step": 549 + }, + { + "epoch": 0.913384192225855, + "grad_norm": 25.545127868652344, + "learning_rate": 8.167204659369672e-05, + "loss": 0.2983, + "step": 550 + }, + { + "epoch": 0.9150448907571748, + "grad_norm": 40.361663818359375, + "learning_rate": 8.16026988990097e-05, + "loss": 0.2985, + "step": 551 + }, + { + "epoch": 0.9167055892884944, + "grad_norm": 12.812162399291992, + "learning_rate": 8.153324982144258e-05, + "loss": 0.3072, + "step": 552 + }, + { + "epoch": 0.9183662878198142, + "grad_norm": 11.464404106140137, + "learning_rate": 8.146369958379111e-05, + "loss": 0.3063, + "step": 553 + }, + { + "epoch": 0.9200269863511339, + "grad_norm": 26.899913787841797, + "learning_rate": 8.139404840917562e-05, + "loss": 0.3165, + "step": 554 + }, + { + "epoch": 0.9216876848824537, + "grad_norm": 24.372446060180664, + "learning_rate": 8.132429652104018e-05, + "loss": 0.3202, + "step": 555 + }, + { + "epoch": 0.9233483834137735, + "grad_norm": 26.897872924804688, + "learning_rate": 8.1254444143152e-05, + "loss": 0.3059, + "step": 556 + }, + { + "epoch": 0.9250090819450931, + "grad_norm": 21.11618423461914, + "learning_rate": 8.118449149960065e-05, + "loss": 0.2844, + "step": 557 + }, + { + "epoch": 0.9266697804764129, + "grad_norm": 18.354814529418945, + "learning_rate": 8.111443881479733e-05, + "loss": 0.2838, + "step": 558 + }, + { + "epoch": 0.9283304790077326, + "grad_norm": 73.88275146484375, + "learning_rate": 8.104428631347425e-05, + "loss": 0.3014, + "step": 559 + }, + { + "epoch": 0.9299911775390524, + "grad_norm": 29.243141174316406, + "learning_rate": 8.097403422068377e-05, + "loss": 0.3009, + "step": 560 + }, + { + "epoch": 0.9316518760703721, + "grad_norm": 32.62166976928711, + "learning_rate": 8.090368276179776e-05, + "loss": 0.3046, + "step": 561 + }, + { + "epoch": 0.9333125746016918, + "grad_norm": 21.50398063659668, + "learning_rate": 8.083323216250687e-05, + "loss": 0.2741, + "step": 562 + }, + { + "epoch": 0.9349732731330116, + "grad_norm": 31.77344512939453, + "learning_rate": 8.07626826488198e-05, + "loss": 0.2847, + "step": 563 + }, + { + "epoch": 0.9366339716643313, + "grad_norm": 16.91570472717285, + "learning_rate": 8.069203444706256e-05, + "loss": 0.3398, + "step": 564 + }, + { + "epoch": 0.938294670195651, + "grad_norm": 8.741859436035156, + "learning_rate": 8.062128778387776e-05, + "loss": 0.2788, + "step": 565 + }, + { + "epoch": 0.9399553687269708, + "grad_norm": 54.93818283081055, + "learning_rate": 8.055044288622386e-05, + "loss": 0.3252, + "step": 566 + }, + { + "epoch": 0.9416160672582905, + "grad_norm": 10.991079330444336, + "learning_rate": 8.047949998137451e-05, + "loss": 0.2925, + "step": 567 + }, + { + "epoch": 0.9432767657896103, + "grad_norm": 23.592836380004883, + "learning_rate": 8.040845929691773e-05, + "loss": 0.2731, + "step": 568 + }, + { + "epoch": 0.94493746432093, + "grad_norm": 50.829315185546875, + "learning_rate": 8.033732106075523e-05, + "loss": 0.3142, + "step": 569 + }, + { + "epoch": 0.9465981628522497, + "grad_norm": 13.88403034210205, + "learning_rate": 8.026608550110166e-05, + "loss": 0.2893, + "step": 570 + }, + { + "epoch": 0.9482588613835695, + "grad_norm": 38.30795669555664, + "learning_rate": 8.019475284648392e-05, + "loss": 0.2737, + "step": 571 + }, + { + "epoch": 0.9499195599148892, + "grad_norm": 62.13677978515625, + "learning_rate": 8.012332332574039e-05, + "loss": 0.2894, + "step": 572 + }, + { + "epoch": 0.951580258446209, + "grad_norm": 9.613969802856445, + "learning_rate": 8.005179716802014e-05, + "loss": 0.2955, + "step": 573 + }, + { + "epoch": 0.9532409569775286, + "grad_norm": 16.622901916503906, + "learning_rate": 7.998017460278234e-05, + "loss": 0.2902, + "step": 574 + }, + { + "epoch": 0.9549016555088484, + "grad_norm": 7.570230484008789, + "learning_rate": 7.990845585979538e-05, + "loss": 0.2993, + "step": 575 + }, + { + "epoch": 0.9565623540401681, + "grad_norm": 17.529743194580078, + "learning_rate": 7.983664116913624e-05, + "loss": 0.2847, + "step": 576 + }, + { + "epoch": 0.9582230525714879, + "grad_norm": 15.870035171508789, + "learning_rate": 7.976473076118967e-05, + "loss": 0.3092, + "step": 577 + }, + { + "epoch": 0.9598837511028077, + "grad_norm": 23.450817108154297, + "learning_rate": 7.969272486664745e-05, + "loss": 0.3062, + "step": 578 + }, + { + "epoch": 0.9615444496341273, + "grad_norm": 20.107789993286133, + "learning_rate": 7.96206237165078e-05, + "loss": 0.2968, + "step": 579 + }, + { + "epoch": 0.9632051481654471, + "grad_norm": 19.919832229614258, + "learning_rate": 7.954842754207442e-05, + "loss": 0.279, + "step": 580 + }, + { + "epoch": 0.9648658466967668, + "grad_norm": 19.68936538696289, + "learning_rate": 7.947613657495589e-05, + "loss": 0.2871, + "step": 581 + }, + { + "epoch": 0.9665265452280866, + "grad_norm": 29.387771606445312, + "learning_rate": 7.940375104706488e-05, + "loss": 0.3021, + "step": 582 + }, + { + "epoch": 0.9681872437594063, + "grad_norm": 33.55486297607422, + "learning_rate": 7.933127119061742e-05, + "loss": 0.2861, + "step": 583 + }, + { + "epoch": 0.969847942290726, + "grad_norm": 19.263111114501953, + "learning_rate": 7.925869723813217e-05, + "loss": 0.3111, + "step": 584 + }, + { + "epoch": 0.9715086408220458, + "grad_norm": 15.75847339630127, + "learning_rate": 7.91860294224296e-05, + "loss": 0.2787, + "step": 585 + }, + { + "epoch": 0.9731693393533655, + "grad_norm": 9.852529525756836, + "learning_rate": 7.911326797663138e-05, + "loss": 0.2864, + "step": 586 + }, + { + "epoch": 0.9748300378846853, + "grad_norm": 7.465989112854004, + "learning_rate": 7.904041313415945e-05, + "loss": 0.284, + "step": 587 + }, + { + "epoch": 0.976490736416005, + "grad_norm": 12.43015193939209, + "learning_rate": 7.896746512873546e-05, + "loss": 0.2614, + "step": 588 + }, + { + "epoch": 0.9781514349473247, + "grad_norm": 11.797941207885742, + "learning_rate": 7.889442419437989e-05, + "loss": 0.3279, + "step": 589 + }, + { + "epoch": 0.9798121334786445, + "grad_norm": 16.83205795288086, + "learning_rate": 7.882129056541133e-05, + "loss": 0.2699, + "step": 590 + }, + { + "epoch": 0.9814728320099642, + "grad_norm": 30.030778884887695, + "learning_rate": 7.874806447644576e-05, + "loss": 0.3064, + "step": 591 + }, + { + "epoch": 0.9831335305412839, + "grad_norm": 51.789127349853516, + "learning_rate": 7.867474616239576e-05, + "loss": 0.2998, + "step": 592 + }, + { + "epoch": 0.9847942290726036, + "grad_norm": 43.69134521484375, + "learning_rate": 7.860133585846978e-05, + "loss": 0.2917, + "step": 593 + }, + { + "epoch": 0.9864549276039234, + "grad_norm": 28.270532608032227, + "learning_rate": 7.852783380017142e-05, + "loss": 0.3149, + "step": 594 + }, + { + "epoch": 0.9881156261352432, + "grad_norm": 44.79228591918945, + "learning_rate": 7.845424022329852e-05, + "loss": 0.2881, + "step": 595 + }, + { + "epoch": 0.9897763246665628, + "grad_norm": 22.45023536682129, + "learning_rate": 7.838055536394264e-05, + "loss": 0.2847, + "step": 596 + }, + { + "epoch": 0.9914370231978826, + "grad_norm": 12.348166465759277, + "learning_rate": 7.83067794584881e-05, + "loss": 0.3042, + "step": 597 + }, + { + "epoch": 0.9930977217292023, + "grad_norm": 20.62948226928711, + "learning_rate": 7.823291274361136e-05, + "loss": 0.3029, + "step": 598 + }, + { + "epoch": 0.9947584202605221, + "grad_norm": 16.2228946685791, + "learning_rate": 7.815895545628011e-05, + "loss": 0.3245, + "step": 599 + }, + { + "epoch": 0.9964191187918419, + "grad_norm": 152.15866088867188, + "learning_rate": 7.808490783375269e-05, + "loss": 0.3033, + "step": 600 + }, + { + "epoch": 0.9980798173231615, + "grad_norm": 10.669971466064453, + "learning_rate": 7.801077011357718e-05, + "loss": 0.2714, + "step": 601 + }, + { + "epoch": 0.9997405158544813, + "grad_norm": 33.99671173095703, + "learning_rate": 7.793654253359075e-05, + "loss": 0.2868, + "step": 602 + }, + { + "epoch": 1.0, + "grad_norm": 19.02952003479004, + "learning_rate": 7.786222533191882e-05, + "loss": 0.2835, + "step": 603 + }, + { + "epoch": 1.0016606985313197, + "grad_norm": 67.08675384521484, + "learning_rate": 7.778781874697429e-05, + "loss": 0.2806, + "step": 604 + }, + { + "epoch": 1.0033213970626396, + "grad_norm": 66.4101333618164, + "learning_rate": 7.771332301745685e-05, + "loss": 0.3183, + "step": 605 + }, + { + "epoch": 1.0049820955939592, + "grad_norm": 10.34506893157959, + "learning_rate": 7.763873838235215e-05, + "loss": 0.3084, + "step": 606 + }, + { + "epoch": 1.006642794125279, + "grad_norm": 17.519620895385742, + "learning_rate": 7.756406508093105e-05, + "loss": 0.2746, + "step": 607 + }, + { + "epoch": 1.0083034926565986, + "grad_norm": 16.470495223999023, + "learning_rate": 7.748930335274886e-05, + "loss": 0.2969, + "step": 608 + }, + { + "epoch": 1.0099641911879185, + "grad_norm": 52.280517578125, + "learning_rate": 7.74144534376446e-05, + "loss": 0.3093, + "step": 609 + }, + { + "epoch": 1.0116248897192381, + "grad_norm": 35.919151306152344, + "learning_rate": 7.733951557574011e-05, + "loss": 0.2942, + "step": 610 + }, + { + "epoch": 1.0132855882505578, + "grad_norm": 20.74170684814453, + "learning_rate": 7.726449000743945e-05, + "loss": 0.2714, + "step": 611 + }, + { + "epoch": 1.0149462867818777, + "grad_norm": 16.128490447998047, + "learning_rate": 7.718937697342803e-05, + "loss": 0.2793, + "step": 612 + }, + { + "epoch": 1.0166069853131974, + "grad_norm": 116.78807830810547, + "learning_rate": 7.711417671467183e-05, + "loss": 0.2893, + "step": 613 + }, + { + "epoch": 1.018267683844517, + "grad_norm": 223.0644073486328, + "learning_rate": 7.703888947241668e-05, + "loss": 0.2964, + "step": 614 + }, + { + "epoch": 1.019928382375837, + "grad_norm": 14.410882949829102, + "learning_rate": 7.696351548818742e-05, + "loss": 0.3051, + "step": 615 + }, + { + "epoch": 1.0215890809071566, + "grad_norm": 10.32640266418457, + "learning_rate": 7.68880550037872e-05, + "loss": 0.283, + "step": 616 + }, + { + "epoch": 1.0232497794384763, + "grad_norm": 16.771568298339844, + "learning_rate": 7.681250826129663e-05, + "loss": 0.2927, + "step": 617 + }, + { + "epoch": 1.024910477969796, + "grad_norm": 17.918943405151367, + "learning_rate": 7.673687550307307e-05, + "loss": 0.3049, + "step": 618 + }, + { + "epoch": 1.0265711765011158, + "grad_norm": 18.209108352661133, + "learning_rate": 7.66611569717498e-05, + "loss": 0.2884, + "step": 619 + }, + { + "epoch": 1.0282318750324355, + "grad_norm": 63.16032409667969, + "learning_rate": 7.658535291023532e-05, + "loss": 0.2819, + "step": 620 + }, + { + "epoch": 1.0298925735637552, + "grad_norm": 14.381697654724121, + "learning_rate": 7.650946356171241e-05, + "loss": 0.2859, + "step": 621 + }, + { + "epoch": 1.031553272095075, + "grad_norm": 27.722930908203125, + "learning_rate": 7.643348916963755e-05, + "loss": 0.3175, + "step": 622 + }, + { + "epoch": 1.0332139706263948, + "grad_norm": 208.41160583496094, + "learning_rate": 7.635742997774e-05, + "loss": 0.291, + "step": 623 + }, + { + "epoch": 1.0348746691577144, + "grad_norm": 18.575511932373047, + "learning_rate": 7.628128623002109e-05, + "loss": 0.2741, + "step": 624 + }, + { + "epoch": 1.036535367689034, + "grad_norm": 24.20387077331543, + "learning_rate": 7.620505817075334e-05, + "loss": 0.3056, + "step": 625 + }, + { + "epoch": 1.038196066220354, + "grad_norm": 7.227424621582031, + "learning_rate": 7.612874604447984e-05, + "loss": 0.2905, + "step": 626 + }, + { + "epoch": 1.0398567647516737, + "grad_norm": 17.7248592376709, + "learning_rate": 7.60523500960133e-05, + "loss": 0.2899, + "step": 627 + }, + { + "epoch": 1.0415174632829933, + "grad_norm": 16.93624496459961, + "learning_rate": 7.597587057043539e-05, + "loss": 0.2944, + "step": 628 + }, + { + "epoch": 1.0431781618143132, + "grad_norm": 12.364616394042969, + "learning_rate": 7.589930771309583e-05, + "loss": 0.3004, + "step": 629 + }, + { + "epoch": 1.044838860345633, + "grad_norm": 100.5436019897461, + "learning_rate": 7.582266176961175e-05, + "loss": 0.3008, + "step": 630 + }, + { + "epoch": 1.0464995588769526, + "grad_norm": 22.728607177734375, + "learning_rate": 7.574593298586675e-05, + "loss": 0.2691, + "step": 631 + }, + { + "epoch": 1.0481602574082725, + "grad_norm": 55.61294174194336, + "learning_rate": 7.566912160801023e-05, + "loss": 0.2877, + "step": 632 + }, + { + "epoch": 1.0498209559395921, + "grad_norm": 11.758903503417969, + "learning_rate": 7.559222788245656e-05, + "loss": 0.2953, + "step": 633 + }, + { + "epoch": 1.0514816544709118, + "grad_norm": 26.880233764648438, + "learning_rate": 7.551525205588427e-05, + "loss": 0.3095, + "step": 634 + }, + { + "epoch": 1.0531423530022315, + "grad_norm": 31.562231063842773, + "learning_rate": 7.543819437523527e-05, + "loss": 0.2822, + "step": 635 + }, + { + "epoch": 1.0548030515335514, + "grad_norm": 16.721778869628906, + "learning_rate": 7.536105508771407e-05, + "loss": 0.3034, + "step": 636 + }, + { + "epoch": 1.056463750064871, + "grad_norm": 9.964337348937988, + "learning_rate": 7.528383444078695e-05, + "loss": 0.2729, + "step": 637 + }, + { + "epoch": 1.0581244485961907, + "grad_norm": 373.1859436035156, + "learning_rate": 7.520653268218128e-05, + "loss": 0.2807, + "step": 638 + }, + { + "epoch": 1.0597851471275106, + "grad_norm": 36.72458267211914, + "learning_rate": 7.51291500598845e-05, + "loss": 0.3009, + "step": 639 + }, + { + "epoch": 1.0614458456588303, + "grad_norm": 10.133912086486816, + "learning_rate": 7.50516868221436e-05, + "loss": 0.2937, + "step": 640 + }, + { + "epoch": 1.06310654419015, + "grad_norm": 30.149831771850586, + "learning_rate": 7.497414321746414e-05, + "loss": 0.2994, + "step": 641 + }, + { + "epoch": 1.0647672427214698, + "grad_norm": 10.886618614196777, + "learning_rate": 7.489651949460945e-05, + "loss": 0.2843, + "step": 642 + }, + { + "epoch": 1.0664279412527895, + "grad_norm": 166.30923461914062, + "learning_rate": 7.481881590259997e-05, + "loss": 0.2861, + "step": 643 + }, + { + "epoch": 1.0680886397841092, + "grad_norm": 8.131608963012695, + "learning_rate": 7.474103269071229e-05, + "loss": 0.3018, + "step": 644 + }, + { + "epoch": 1.0697493383154288, + "grad_norm": 23.057449340820312, + "learning_rate": 7.466317010847844e-05, + "loss": 0.3117, + "step": 645 + }, + { + "epoch": 1.0714100368467487, + "grad_norm": 33.74464416503906, + "learning_rate": 7.458522840568513e-05, + "loss": 0.3023, + "step": 646 + }, + { + "epoch": 1.0730707353780684, + "grad_norm": 10.909245491027832, + "learning_rate": 7.450720783237282e-05, + "loss": 0.3109, + "step": 647 + }, + { + "epoch": 1.074731433909388, + "grad_norm": 20.060121536254883, + "learning_rate": 7.442910863883503e-05, + "loss": 0.315, + "step": 648 + }, + { + "epoch": 1.076392132440708, + "grad_norm": 28.669387817382812, + "learning_rate": 7.435093107561748e-05, + "loss": 0.3135, + "step": 649 + }, + { + "epoch": 1.0780528309720276, + "grad_norm": 124.63036346435547, + "learning_rate": 7.427267539351733e-05, + "loss": 0.2938, + "step": 650 + }, + { + "epoch": 1.0797135295033473, + "grad_norm": 14.605134010314941, + "learning_rate": 7.41943418435823e-05, + "loss": 0.2814, + "step": 651 + }, + { + "epoch": 1.081374228034667, + "grad_norm": 12.903773307800293, + "learning_rate": 7.411593067710997e-05, + "loss": 0.3049, + "step": 652 + }, + { + "epoch": 1.0830349265659869, + "grad_norm": 62.30295944213867, + "learning_rate": 7.403744214564687e-05, + "loss": 0.2754, + "step": 653 + }, + { + "epoch": 1.0846956250973065, + "grad_norm": 79.19766998291016, + "learning_rate": 7.395887650098776e-05, + "loss": 0.2796, + "step": 654 + }, + { + "epoch": 1.0863563236286262, + "grad_norm": 21.715784072875977, + "learning_rate": 7.388023399517475e-05, + "loss": 0.321, + "step": 655 + }, + { + "epoch": 1.0880170221599461, + "grad_norm": 15.694341659545898, + "learning_rate": 7.380151488049652e-05, + "loss": 0.2916, + "step": 656 + }, + { + "epoch": 1.0896777206912658, + "grad_norm": 24.067087173461914, + "learning_rate": 7.372271940948759e-05, + "loss": 0.3033, + "step": 657 + }, + { + "epoch": 1.0913384192225855, + "grad_norm": 43.47443389892578, + "learning_rate": 7.36438478349273e-05, + "loss": 0.2964, + "step": 658 + }, + { + "epoch": 1.0929991177539051, + "grad_norm": 16.10695457458496, + "learning_rate": 7.356490040983927e-05, + "loss": 0.3108, + "step": 659 + }, + { + "epoch": 1.094659816285225, + "grad_norm": 329.49432373046875, + "learning_rate": 7.348587738749037e-05, + "loss": 0.2907, + "step": 660 + }, + { + "epoch": 1.0963205148165447, + "grad_norm": 47.651634216308594, + "learning_rate": 7.340677902139001e-05, + "loss": 0.2919, + "step": 661 + }, + { + "epoch": 1.0979812133478644, + "grad_norm": 11.220695495605469, + "learning_rate": 7.33276055652893e-05, + "loss": 0.3099, + "step": 662 + }, + { + "epoch": 1.0996419118791843, + "grad_norm": 32.71885681152344, + "learning_rate": 7.324835727318027e-05, + "loss": 0.2922, + "step": 663 + }, + { + "epoch": 1.101302610410504, + "grad_norm": 168.6045379638672, + "learning_rate": 7.3169034399295e-05, + "loss": 0.2957, + "step": 664 + }, + { + "epoch": 1.1029633089418236, + "grad_norm": 15.729631423950195, + "learning_rate": 7.308963719810483e-05, + "loss": 0.2903, + "step": 665 + }, + { + "epoch": 1.1046240074731435, + "grad_norm": 45.95412063598633, + "learning_rate": 7.301016592431957e-05, + "loss": 0.2783, + "step": 666 + }, + { + "epoch": 1.1062847060044632, + "grad_norm": 14.088940620422363, + "learning_rate": 7.293062083288663e-05, + "loss": 0.2821, + "step": 667 + }, + { + "epoch": 1.1079454045357828, + "grad_norm": 15.857392311096191, + "learning_rate": 7.285100217899022e-05, + "loss": 0.3021, + "step": 668 + }, + { + "epoch": 1.1096061030671025, + "grad_norm": 15.9371919631958, + "learning_rate": 7.277131021805061e-05, + "loss": 0.294, + "step": 669 + }, + { + "epoch": 1.1112668015984224, + "grad_norm": 8.502978324890137, + "learning_rate": 7.269154520572318e-05, + "loss": 0.2815, + "step": 670 + }, + { + "epoch": 1.112927500129742, + "grad_norm": 128.68911743164062, + "learning_rate": 7.261170739789766e-05, + "loss": 0.302, + "step": 671 + }, + { + "epoch": 1.1145881986610617, + "grad_norm": 13.39405632019043, + "learning_rate": 7.253179705069735e-05, + "loss": 0.2978, + "step": 672 + }, + { + "epoch": 1.1162488971923816, + "grad_norm": 17.800065994262695, + "learning_rate": 7.245181442047824e-05, + "loss": 0.296, + "step": 673 + }, + { + "epoch": 1.1179095957237013, + "grad_norm": 19.4841365814209, + "learning_rate": 7.23717597638282e-05, + "loss": 0.287, + "step": 674 + }, + { + "epoch": 1.119570294255021, + "grad_norm": 16.200912475585938, + "learning_rate": 7.229163333756615e-05, + "loss": 0.2808, + "step": 675 + }, + { + "epoch": 1.1212309927863409, + "grad_norm": 36.189964294433594, + "learning_rate": 7.221143539874132e-05, + "loss": 0.3103, + "step": 676 + }, + { + "epoch": 1.1228916913176605, + "grad_norm": 15.907554626464844, + "learning_rate": 7.213116620463226e-05, + "loss": 0.2728, + "step": 677 + }, + { + "epoch": 1.1245523898489802, + "grad_norm": 30.8767147064209, + "learning_rate": 7.20508260127462e-05, + "loss": 0.2986, + "step": 678 + }, + { + "epoch": 1.1262130883802999, + "grad_norm": 48.497589111328125, + "learning_rate": 7.197041508081803e-05, + "loss": 0.3131, + "step": 679 + }, + { + "epoch": 1.1278737869116198, + "grad_norm": 14.5043363571167, + "learning_rate": 7.18899336668097e-05, + "loss": 0.2993, + "step": 680 + }, + { + "epoch": 1.1295344854429394, + "grad_norm": 23.169178009033203, + "learning_rate": 7.180938202890916e-05, + "loss": 0.2926, + "step": 681 + }, + { + "epoch": 1.131195183974259, + "grad_norm": 229.18008422851562, + "learning_rate": 7.172876042552974e-05, + "loss": 0.3, + "step": 682 + }, + { + "epoch": 1.132855882505579, + "grad_norm": 72.39683532714844, + "learning_rate": 7.164806911530911e-05, + "loss": 0.3083, + "step": 683 + }, + { + "epoch": 1.1345165810368987, + "grad_norm": 67.89056396484375, + "learning_rate": 7.156730835710867e-05, + "loss": 0.2874, + "step": 684 + }, + { + "epoch": 1.1361772795682183, + "grad_norm": 13.902090072631836, + "learning_rate": 7.148647841001257e-05, + "loss": 0.2872, + "step": 685 + }, + { + "epoch": 1.137837978099538, + "grad_norm": 9.960161209106445, + "learning_rate": 7.140557953332687e-05, + "loss": 0.2907, + "step": 686 + }, + { + "epoch": 1.139498676630858, + "grad_norm": 25.784360885620117, + "learning_rate": 7.132461198657886e-05, + "loss": 0.2896, + "step": 687 + }, + { + "epoch": 1.1411593751621776, + "grad_norm": 71.54449462890625, + "learning_rate": 7.124357602951602e-05, + "loss": 0.2924, + "step": 688 + }, + { + "epoch": 1.1428200736934973, + "grad_norm": 11.410158157348633, + "learning_rate": 7.11624719221054e-05, + "loss": 0.259, + "step": 689 + }, + { + "epoch": 1.1444807722248171, + "grad_norm": 16.570106506347656, + "learning_rate": 7.10812999245326e-05, + "loss": 0.2776, + "step": 690 + }, + { + "epoch": 1.1461414707561368, + "grad_norm": 21.61848258972168, + "learning_rate": 7.100006029720103e-05, + "loss": 0.3109, + "step": 691 + }, + { + "epoch": 1.1478021692874565, + "grad_norm": 20.25295066833496, + "learning_rate": 7.09187533007311e-05, + "loss": 0.2814, + "step": 692 + }, + { + "epoch": 1.1494628678187762, + "grad_norm": 56.11478805541992, + "learning_rate": 7.083737919595929e-05, + "loss": 0.2974, + "step": 693 + }, + { + "epoch": 1.151123566350096, + "grad_norm": 11.665583610534668, + "learning_rate": 7.07559382439374e-05, + "loss": 0.279, + "step": 694 + }, + { + "epoch": 1.1527842648814157, + "grad_norm": 21.889516830444336, + "learning_rate": 7.067443070593165e-05, + "loss": 0.2931, + "step": 695 + }, + { + "epoch": 1.1544449634127354, + "grad_norm": 25.720151901245117, + "learning_rate": 7.059285684342192e-05, + "loss": 0.2728, + "step": 696 + }, + { + "epoch": 1.1561056619440553, + "grad_norm": 49.9734001159668, + "learning_rate": 7.051121691810081e-05, + "loss": 0.2833, + "step": 697 + }, + { + "epoch": 1.157766360475375, + "grad_norm": 13.756512641906738, + "learning_rate": 7.042951119187288e-05, + "loss": 0.2959, + "step": 698 + }, + { + "epoch": 1.1594270590066946, + "grad_norm": 30.945308685302734, + "learning_rate": 7.034773992685377e-05, + "loss": 0.2861, + "step": 699 + }, + { + "epoch": 1.1610877575380145, + "grad_norm": 34.149967193603516, + "learning_rate": 7.026590338536938e-05, + "loss": 0.2913, + "step": 700 + }, + { + "epoch": 1.1627484560693342, + "grad_norm": 42.5093879699707, + "learning_rate": 7.018400182995504e-05, + "loss": 0.3094, + "step": 701 + }, + { + "epoch": 1.1644091546006539, + "grad_norm": 33.95546340942383, + "learning_rate": 7.010203552335458e-05, + "loss": 0.2937, + "step": 702 + }, + { + "epoch": 1.1660698531319738, + "grad_norm": 46.43781661987305, + "learning_rate": 7.002000472851964e-05, + "loss": 0.2863, + "step": 703 + }, + { + "epoch": 1.1677305516632934, + "grad_norm": 26.104015350341797, + "learning_rate": 6.99379097086087e-05, + "loss": 0.3001, + "step": 704 + }, + { + "epoch": 1.169391250194613, + "grad_norm": 27.750532150268555, + "learning_rate": 6.985575072698625e-05, + "loss": 0.2771, + "step": 705 + }, + { + "epoch": 1.1710519487259328, + "grad_norm": 64.4836654663086, + "learning_rate": 6.977352804722202e-05, + "loss": 0.3063, + "step": 706 + }, + { + "epoch": 1.1727126472572527, + "grad_norm": 11.415141105651855, + "learning_rate": 6.969124193309005e-05, + "loss": 0.2935, + "step": 707 + }, + { + "epoch": 1.1743733457885723, + "grad_norm": 53.069496154785156, + "learning_rate": 6.960889264856791e-05, + "loss": 0.2822, + "step": 708 + }, + { + "epoch": 1.176034044319892, + "grad_norm": 51.70155715942383, + "learning_rate": 6.952648045783583e-05, + "loss": 0.2997, + "step": 709 + }, + { + "epoch": 1.177694742851212, + "grad_norm": 25.220409393310547, + "learning_rate": 6.944400562527578e-05, + "loss": 0.2731, + "step": 710 + }, + { + "epoch": 1.1793554413825316, + "grad_norm": 35.78030014038086, + "learning_rate": 6.936146841547076e-05, + "loss": 0.2822, + "step": 711 + }, + { + "epoch": 1.1810161399138512, + "grad_norm": 46.63139724731445, + "learning_rate": 6.927886909320387e-05, + "loss": 0.2894, + "step": 712 + }, + { + "epoch": 1.182676838445171, + "grad_norm": 17.879621505737305, + "learning_rate": 6.919620792345743e-05, + "loss": 0.2862, + "step": 713 + }, + { + "epoch": 1.1843375369764908, + "grad_norm": 12.535906791687012, + "learning_rate": 6.911348517141217e-05, + "loss": 0.2714, + "step": 714 + }, + { + "epoch": 1.1859982355078105, + "grad_norm": 33.826412200927734, + "learning_rate": 6.903070110244645e-05, + "loss": 0.2923, + "step": 715 + }, + { + "epoch": 1.1876589340391301, + "grad_norm": 18.073890686035156, + "learning_rate": 6.894785598213527e-05, + "loss": 0.2749, + "step": 716 + }, + { + "epoch": 1.18931963257045, + "grad_norm": 28.41618537902832, + "learning_rate": 6.88649500762495e-05, + "loss": 0.2576, + "step": 717 + }, + { + "epoch": 1.1909803311017697, + "grad_norm": 70.42158508300781, + "learning_rate": 6.878198365075498e-05, + "loss": 0.288, + "step": 718 + }, + { + "epoch": 1.1926410296330894, + "grad_norm": 128.601318359375, + "learning_rate": 6.869895697181181e-05, + "loss": 0.2819, + "step": 719 + }, + { + "epoch": 1.194301728164409, + "grad_norm": 46.425025939941406, + "learning_rate": 6.861587030577326e-05, + "loss": 0.3028, + "step": 720 + }, + { + "epoch": 1.195962426695729, + "grad_norm": 17.589296340942383, + "learning_rate": 6.853272391918511e-05, + "loss": 0.2982, + "step": 721 + }, + { + "epoch": 1.1976231252270486, + "grad_norm": 15.980798721313477, + "learning_rate": 6.84495180787847e-05, + "loss": 0.2861, + "step": 722 + }, + { + "epoch": 1.1992838237583683, + "grad_norm": 32.1770133972168, + "learning_rate": 6.836625305150013e-05, + "loss": 0.2928, + "step": 723 + }, + { + "epoch": 1.2009445222896882, + "grad_norm": 237.07186889648438, + "learning_rate": 6.828292910444934e-05, + "loss": 0.3006, + "step": 724 + }, + { + "epoch": 1.2026052208210078, + "grad_norm": 17.907649993896484, + "learning_rate": 6.819954650493929e-05, + "loss": 0.2982, + "step": 725 + }, + { + "epoch": 1.2042659193523275, + "grad_norm": 31.907760620117188, + "learning_rate": 6.811610552046512e-05, + "loss": 0.2794, + "step": 726 + }, + { + "epoch": 1.2059266178836472, + "grad_norm": 20.636262893676758, + "learning_rate": 6.803260641870929e-05, + "loss": 0.2871, + "step": 727 + }, + { + "epoch": 1.207587316414967, + "grad_norm": 31.23724937438965, + "learning_rate": 6.794904946754064e-05, + "loss": 0.286, + "step": 728 + }, + { + "epoch": 1.2092480149462868, + "grad_norm": 24.60127830505371, + "learning_rate": 6.786543493501364e-05, + "loss": 0.3027, + "step": 729 + }, + { + "epoch": 1.2109087134776064, + "grad_norm": 13.422248840332031, + "learning_rate": 6.778176308936746e-05, + "loss": 0.305, + "step": 730 + }, + { + "epoch": 1.2125694120089263, + "grad_norm": 19.39386749267578, + "learning_rate": 6.769803419902518e-05, + "loss": 0.3089, + "step": 731 + }, + { + "epoch": 1.214230110540246, + "grad_norm": 9.790982246398926, + "learning_rate": 6.761424853259283e-05, + "loss": 0.3013, + "step": 732 + }, + { + "epoch": 1.2158908090715657, + "grad_norm": 9.754476547241211, + "learning_rate": 6.753040635885859e-05, + "loss": 0.256, + "step": 733 + }, + { + "epoch": 1.2175515076028856, + "grad_norm": 25.753896713256836, + "learning_rate": 6.744650794679193e-05, + "loss": 0.2883, + "step": 734 + }, + { + "epoch": 1.2192122061342052, + "grad_norm": 81.93810272216797, + "learning_rate": 6.736255356554271e-05, + "loss": 0.3122, + "step": 735 + }, + { + "epoch": 1.220872904665525, + "grad_norm": 51.9580078125, + "learning_rate": 6.727854348444042e-05, + "loss": 0.2652, + "step": 736 + }, + { + "epoch": 1.2225336031968448, + "grad_norm": 25.365999221801758, + "learning_rate": 6.719447797299313e-05, + "loss": 0.2897, + "step": 737 + }, + { + "epoch": 1.2241943017281645, + "grad_norm": 10.5918550491333, + "learning_rate": 6.711035730088681e-05, + "loss": 0.2911, + "step": 738 + }, + { + "epoch": 1.2258550002594841, + "grad_norm": 10.157291412353516, + "learning_rate": 6.702618173798434e-05, + "loss": 0.2916, + "step": 739 + }, + { + "epoch": 1.2275156987908038, + "grad_norm": 16.369253158569336, + "learning_rate": 6.694195155432472e-05, + "loss": 0.2949, + "step": 740 + }, + { + "epoch": 1.2291763973221237, + "grad_norm": 14.09495735168457, + "learning_rate": 6.685766702012219e-05, + "loss": 0.2883, + "step": 741 + }, + { + "epoch": 1.2308370958534434, + "grad_norm": 8.436043739318848, + "learning_rate": 6.67733284057653e-05, + "loss": 0.2804, + "step": 742 + }, + { + "epoch": 1.232497794384763, + "grad_norm": 35.84096145629883, + "learning_rate": 6.668893598181614e-05, + "loss": 0.2487, + "step": 743 + }, + { + "epoch": 1.234158492916083, + "grad_norm": 14.134810447692871, + "learning_rate": 6.660449001900939e-05, + "loss": 0.2659, + "step": 744 + }, + { + "epoch": 1.2358191914474026, + "grad_norm": 11.42489242553711, + "learning_rate": 6.651999078825151e-05, + "loss": 0.2672, + "step": 745 + }, + { + "epoch": 1.2374798899787223, + "grad_norm": 24.68474578857422, + "learning_rate": 6.643543856061981e-05, + "loss": 0.2848, + "step": 746 + }, + { + "epoch": 1.239140588510042, + "grad_norm": 34.61080551147461, + "learning_rate": 6.635083360736167e-05, + "loss": 0.2755, + "step": 747 + }, + { + "epoch": 1.2408012870413618, + "grad_norm": 104.77276611328125, + "learning_rate": 6.626617619989358e-05, + "loss": 0.2874, + "step": 748 + }, + { + "epoch": 1.2424619855726815, + "grad_norm": 17.170560836791992, + "learning_rate": 6.61814666098003e-05, + "loss": 0.2891, + "step": 749 + }, + { + "epoch": 1.2441226841040012, + "grad_norm": 22.896869659423828, + "learning_rate": 6.609670510883402e-05, + "loss": 0.2901, + "step": 750 + }, + { + "epoch": 1.245783382635321, + "grad_norm": 94.8833999633789, + "learning_rate": 6.601189196891346e-05, + "loss": 0.2928, + "step": 751 + }, + { + "epoch": 1.2474440811666407, + "grad_norm": 36.6009407043457, + "learning_rate": 6.592702746212294e-05, + "loss": 0.2849, + "step": 752 + }, + { + "epoch": 1.2491047796979604, + "grad_norm": 26.896333694458008, + "learning_rate": 6.584211186071168e-05, + "loss": 0.2775, + "step": 753 + }, + { + "epoch": 1.25076547822928, + "grad_norm": 26.26393699645996, + "learning_rate": 6.575714543709271e-05, + "loss": 0.2928, + "step": 754 + }, + { + "epoch": 1.2524261767606, + "grad_norm": 19.48758316040039, + "learning_rate": 6.567212846384215e-05, + "loss": 0.2759, + "step": 755 + }, + { + "epoch": 1.2540868752919196, + "grad_norm": 106.55645751953125, + "learning_rate": 6.558706121369828e-05, + "loss": 0.3019, + "step": 756 + }, + { + "epoch": 1.2557475738232395, + "grad_norm": 35.15691375732422, + "learning_rate": 6.550194395956065e-05, + "loss": 0.3039, + "step": 757 + }, + { + "epoch": 1.2574082723545592, + "grad_norm": 19.721324920654297, + "learning_rate": 6.541677697448926e-05, + "loss": 0.2852, + "step": 758 + }, + { + "epoch": 1.2590689708858789, + "grad_norm": 31.74245834350586, + "learning_rate": 6.533156053170363e-05, + "loss": 0.2818, + "step": 759 + }, + { + "epoch": 1.2607296694171986, + "grad_norm": 91.82173919677734, + "learning_rate": 6.524629490458193e-05, + "loss": 0.2845, + "step": 760 + }, + { + "epoch": 1.2623903679485182, + "grad_norm": 33.08683395385742, + "learning_rate": 6.516098036666013e-05, + "loss": 0.2898, + "step": 761 + }, + { + "epoch": 1.2640510664798381, + "grad_norm": 11.8512601852417, + "learning_rate": 6.50756171916311e-05, + "loss": 0.3024, + "step": 762 + }, + { + "epoch": 1.2657117650111578, + "grad_norm": 15.28557300567627, + "learning_rate": 6.49902056533438e-05, + "loss": 0.2973, + "step": 763 + }, + { + "epoch": 1.2673724635424777, + "grad_norm": 12.084609031677246, + "learning_rate": 6.490474602580224e-05, + "loss": 0.2677, + "step": 764 + }, + { + "epoch": 1.2690331620737973, + "grad_norm": 1329.02880859375, + "learning_rate": 6.481923858316472e-05, + "loss": 0.2898, + "step": 765 + }, + { + "epoch": 1.270693860605117, + "grad_norm": 441.1099853515625, + "learning_rate": 6.473368359974303e-05, + "loss": 0.2753, + "step": 766 + }, + { + "epoch": 1.2723545591364367, + "grad_norm": 392.98577880859375, + "learning_rate": 6.464808135000138e-05, + "loss": 0.3083, + "step": 767 + }, + { + "epoch": 1.2740152576677564, + "grad_norm": 25.587831497192383, + "learning_rate": 6.456243210855564e-05, + "loss": 0.2857, + "step": 768 + }, + { + "epoch": 1.2756759561990763, + "grad_norm": 43.282230377197266, + "learning_rate": 6.44767361501724e-05, + "loss": 0.3014, + "step": 769 + }, + { + "epoch": 1.277336654730396, + "grad_norm": 44.518028259277344, + "learning_rate": 6.43909937497682e-05, + "loss": 0.306, + "step": 770 + }, + { + "epoch": 1.2789973532617158, + "grad_norm": 16.484920501708984, + "learning_rate": 6.43052051824085e-05, + "loss": 0.2881, + "step": 771 + }, + { + "epoch": 1.2806580517930355, + "grad_norm": 23.33445930480957, + "learning_rate": 6.421937072330686e-05, + "loss": 0.2936, + "step": 772 + }, + { + "epoch": 1.2823187503243552, + "grad_norm": 46.49196243286133, + "learning_rate": 6.413349064782412e-05, + "loss": 0.2866, + "step": 773 + }, + { + "epoch": 1.2839794488556748, + "grad_norm": 16.946788787841797, + "learning_rate": 6.404756523146742e-05, + "loss": 0.29, + "step": 774 + }, + { + "epoch": 1.2856401473869947, + "grad_norm": 14.870948791503906, + "learning_rate": 6.396159474988935e-05, + "loss": 0.2958, + "step": 775 + }, + { + "epoch": 1.2873008459183144, + "grad_norm": 161.6600341796875, + "learning_rate": 6.387557947888708e-05, + "loss": 0.2649, + "step": 776 + }, + { + "epoch": 1.288961544449634, + "grad_norm": 9.838664054870605, + "learning_rate": 6.37895196944015e-05, + "loss": 0.2962, + "step": 777 + }, + { + "epoch": 1.290622242980954, + "grad_norm": 157.66258239746094, + "learning_rate": 6.370341567251625e-05, + "loss": 0.2994, + "step": 778 + }, + { + "epoch": 1.2922829415122736, + "grad_norm": 11.306894302368164, + "learning_rate": 6.361726768945689e-05, + "loss": 0.2611, + "step": 779 + }, + { + "epoch": 1.2939436400435933, + "grad_norm": 23.204843521118164, + "learning_rate": 6.353107602159004e-05, + "loss": 0.2743, + "step": 780 + }, + { + "epoch": 1.295604338574913, + "grad_norm": 46.965538024902344, + "learning_rate": 6.344484094542247e-05, + "loss": 0.2936, + "step": 781 + }, + { + "epoch": 1.2972650371062329, + "grad_norm": 27.95403480529785, + "learning_rate": 6.335856273760016e-05, + "loss": 0.2769, + "step": 782 + }, + { + "epoch": 1.2989257356375525, + "grad_norm": 25.462093353271484, + "learning_rate": 6.327224167490749e-05, + "loss": 0.2952, + "step": 783 + }, + { + "epoch": 1.3005864341688722, + "grad_norm": 21.219467163085938, + "learning_rate": 6.318587803426632e-05, + "loss": 0.2674, + "step": 784 + }, + { + "epoch": 1.302247132700192, + "grad_norm": 11.112910270690918, + "learning_rate": 6.309947209273508e-05, + "loss": 0.2619, + "step": 785 + }, + { + "epoch": 1.3039078312315118, + "grad_norm": 17.56656837463379, + "learning_rate": 6.301302412750794e-05, + "loss": 0.2917, + "step": 786 + }, + { + "epoch": 1.3055685297628314, + "grad_norm": 33.908077239990234, + "learning_rate": 6.292653441591383e-05, + "loss": 0.3162, + "step": 787 + }, + { + "epoch": 1.3072292282941511, + "grad_norm": 12.80300235748291, + "learning_rate": 6.284000323541565e-05, + "loss": 0.2746, + "step": 788 + }, + { + "epoch": 1.308889926825471, + "grad_norm": 14.64559268951416, + "learning_rate": 6.275343086360934e-05, + "loss": 0.2943, + "step": 789 + }, + { + "epoch": 1.3105506253567907, + "grad_norm": 37.15105438232422, + "learning_rate": 6.266681757822296e-05, + "loss": 0.2908, + "step": 790 + }, + { + "epoch": 1.3122113238881106, + "grad_norm": 18.574054718017578, + "learning_rate": 6.258016365711583e-05, + "loss": 0.2861, + "step": 791 + }, + { + "epoch": 1.3138720224194302, + "grad_norm": 30.741130828857422, + "learning_rate": 6.249346937827758e-05, + "loss": 0.2794, + "step": 792 + }, + { + "epoch": 1.31553272095075, + "grad_norm": 51.434967041015625, + "learning_rate": 6.240673501982742e-05, + "loss": 0.2972, + "step": 793 + }, + { + "epoch": 1.3171934194820696, + "grad_norm": 28.587820053100586, + "learning_rate": 6.231996086001303e-05, + "loss": 0.2828, + "step": 794 + }, + { + "epoch": 1.3188541180133893, + "grad_norm": 9.587411880493164, + "learning_rate": 6.223314717720984e-05, + "loss": 0.2822, + "step": 795 + }, + { + "epoch": 1.3205148165447091, + "grad_norm": 17.34294319152832, + "learning_rate": 6.214629424992007e-05, + "loss": 0.2904, + "step": 796 + }, + { + "epoch": 1.3221755150760288, + "grad_norm": 16.9996337890625, + "learning_rate": 6.205940235677177e-05, + "loss": 0.3177, + "step": 797 + }, + { + "epoch": 1.3238362136073487, + "grad_norm": 13.364457130432129, + "learning_rate": 6.197247177651808e-05, + "loss": 0.2835, + "step": 798 + }, + { + "epoch": 1.3254969121386684, + "grad_norm": 9.773347854614258, + "learning_rate": 6.188550278803622e-05, + "loss": 0.2908, + "step": 799 + }, + { + "epoch": 1.327157610669988, + "grad_norm": 20.065311431884766, + "learning_rate": 6.179849567032657e-05, + "loss": 0.3076, + "step": 800 + }, + { + "epoch": 1.3288183092013077, + "grad_norm": 58.05887222290039, + "learning_rate": 6.171145070251194e-05, + "loss": 0.2922, + "step": 801 + }, + { + "epoch": 1.3304790077326276, + "grad_norm": 37.12381362915039, + "learning_rate": 6.162436816383646e-05, + "loss": 0.2631, + "step": 802 + }, + { + "epoch": 1.3321397062639473, + "grad_norm": 17.7030086517334, + "learning_rate": 6.153724833366483e-05, + "loss": 0.2832, + "step": 803 + }, + { + "epoch": 1.333800404795267, + "grad_norm": 13.268787384033203, + "learning_rate": 6.145009149148143e-05, + "loss": 0.2855, + "step": 804 + }, + { + "epoch": 1.3354611033265869, + "grad_norm": 25.01466178894043, + "learning_rate": 6.136289791688929e-05, + "loss": 0.2934, + "step": 805 + }, + { + "epoch": 1.3371218018579065, + "grad_norm": 25.221342086791992, + "learning_rate": 6.127566788960934e-05, + "loss": 0.2907, + "step": 806 + }, + { + "epoch": 1.3387825003892262, + "grad_norm": 25.98119354248047, + "learning_rate": 6.118840168947942e-05, + "loss": 0.3115, + "step": 807 + }, + { + "epoch": 1.3404431989205459, + "grad_norm": 31.095993041992188, + "learning_rate": 6.110109959645344e-05, + "loss": 0.2756, + "step": 808 + }, + { + "epoch": 1.3421038974518658, + "grad_norm": 15.839167594909668, + "learning_rate": 6.101376189060043e-05, + "loss": 0.2797, + "step": 809 + }, + { + "epoch": 1.3437645959831854, + "grad_norm": 11.755404472351074, + "learning_rate": 6.092638885210369e-05, + "loss": 0.2699, + "step": 810 + }, + { + "epoch": 1.345425294514505, + "grad_norm": 23.13016128540039, + "learning_rate": 6.083898076125985e-05, + "loss": 0.2959, + "step": 811 + }, + { + "epoch": 1.347085993045825, + "grad_norm": 44.43939971923828, + "learning_rate": 6.0751537898478005e-05, + "loss": 0.2962, + "step": 812 + }, + { + "epoch": 1.3487466915771447, + "grad_norm": 46.97671127319336, + "learning_rate": 6.066406054427881e-05, + "loss": 0.2802, + "step": 813 + }, + { + "epoch": 1.3504073901084643, + "grad_norm": 17.91573143005371, + "learning_rate": 6.057654897929353e-05, + "loss": 0.2978, + "step": 814 + }, + { + "epoch": 1.352068088639784, + "grad_norm": 20.49462127685547, + "learning_rate": 6.048900348426319e-05, + "loss": 0.2977, + "step": 815 + }, + { + "epoch": 1.353728787171104, + "grad_norm": 8.59965705871582, + "learning_rate": 6.0401424340037704e-05, + "loss": 0.2681, + "step": 816 + }, + { + "epoch": 1.3553894857024236, + "grad_norm": 35.38010025024414, + "learning_rate": 6.0313811827574915e-05, + "loss": 0.2839, + "step": 817 + }, + { + "epoch": 1.3570501842337432, + "grad_norm": 20.218486785888672, + "learning_rate": 6.0226166227939686e-05, + "loss": 0.3071, + "step": 818 + }, + { + "epoch": 1.3587108827650631, + "grad_norm": 13.453107833862305, + "learning_rate": 6.013848782230303e-05, + "loss": 0.2987, + "step": 819 + }, + { + "epoch": 1.3603715812963828, + "grad_norm": 16.644697189331055, + "learning_rate": 6.0050776891941264e-05, + "loss": 0.2879, + "step": 820 + }, + { + "epoch": 1.3620322798277025, + "grad_norm": 14.217223167419434, + "learning_rate": 5.996303371823495e-05, + "loss": 0.3009, + "step": 821 + }, + { + "epoch": 1.3636929783590221, + "grad_norm": 17.42101287841797, + "learning_rate": 5.987525858266818e-05, + "loss": 0.2774, + "step": 822 + }, + { + "epoch": 1.365353676890342, + "grad_norm": 26.677465438842773, + "learning_rate": 5.9787451766827495e-05, + "loss": 0.2972, + "step": 823 + }, + { + "epoch": 1.3670143754216617, + "grad_norm": 46.07740783691406, + "learning_rate": 5.9699613552401145e-05, + "loss": 0.2897, + "step": 824 + }, + { + "epoch": 1.3686750739529816, + "grad_norm": 9.926246643066406, + "learning_rate": 5.961174422117808e-05, + "loss": 0.2988, + "step": 825 + }, + { + "epoch": 1.3703357724843013, + "grad_norm": 24.13726806640625, + "learning_rate": 5.9523844055047026e-05, + "loss": 0.3004, + "step": 826 + }, + { + "epoch": 1.371996471015621, + "grad_norm": 25.25496482849121, + "learning_rate": 5.943591333599571e-05, + "loss": 0.2945, + "step": 827 + }, + { + "epoch": 1.3736571695469406, + "grad_norm": 21.481048583984375, + "learning_rate": 5.934795234610982e-05, + "loss": 0.2883, + "step": 828 + }, + { + "epoch": 1.3753178680782603, + "grad_norm": 9.13387680053711, + "learning_rate": 5.925996136757217e-05, + "loss": 0.2932, + "step": 829 + }, + { + "epoch": 1.3769785666095802, + "grad_norm": 24.65291976928711, + "learning_rate": 5.9171940682661784e-05, + "loss": 0.2808, + "step": 830 + }, + { + "epoch": 1.3786392651408998, + "grad_norm": 19.52792739868164, + "learning_rate": 5.9083890573752964e-05, + "loss": 0.2621, + "step": 831 + }, + { + "epoch": 1.3802999636722197, + "grad_norm": 111.61624145507812, + "learning_rate": 5.899581132331445e-05, + "loss": 0.2773, + "step": 832 + }, + { + "epoch": 1.3819606622035394, + "grad_norm": 33.516902923583984, + "learning_rate": 5.8907703213908427e-05, + "loss": 0.269, + "step": 833 + }, + { + "epoch": 1.383621360734859, + "grad_norm": 51.55702590942383, + "learning_rate": 5.881956652818966e-05, + "loss": 0.2822, + "step": 834 + }, + { + "epoch": 1.3852820592661788, + "grad_norm": 19.665721893310547, + "learning_rate": 5.873140154890462e-05, + "loss": 0.3109, + "step": 835 + }, + { + "epoch": 1.3869427577974986, + "grad_norm": 26.804651260375977, + "learning_rate": 5.8643208558890516e-05, + "loss": 0.268, + "step": 836 + }, + { + "epoch": 1.3886034563288183, + "grad_norm": 90.68706512451172, + "learning_rate": 5.855498784107445e-05, + "loss": 0.2764, + "step": 837 + }, + { + "epoch": 1.390264154860138, + "grad_norm": 14.730112075805664, + "learning_rate": 5.846673967847243e-05, + "loss": 0.2677, + "step": 838 + }, + { + "epoch": 1.3919248533914579, + "grad_norm": 18.75749969482422, + "learning_rate": 5.837846435418853e-05, + "loss": 0.2836, + "step": 839 + }, + { + "epoch": 1.3935855519227776, + "grad_norm": 72.0309066772461, + "learning_rate": 5.8290162151414e-05, + "loss": 0.2967, + "step": 840 + }, + { + "epoch": 1.3952462504540972, + "grad_norm": 12.728450775146484, + "learning_rate": 5.820183335342624e-05, + "loss": 0.2742, + "step": 841 + }, + { + "epoch": 1.396906948985417, + "grad_norm": 20.818998336791992, + "learning_rate": 5.811347824358801e-05, + "loss": 0.2848, + "step": 842 + }, + { + "epoch": 1.3985676475167368, + "grad_norm": 17.97675132751465, + "learning_rate": 5.802509710534652e-05, + "loss": 0.2637, + "step": 843 + }, + { + "epoch": 1.4002283460480565, + "grad_norm": 19.154029846191406, + "learning_rate": 5.793669022223238e-05, + "loss": 0.286, + "step": 844 + }, + { + "epoch": 1.4018890445793761, + "grad_norm": 38.61789321899414, + "learning_rate": 5.7848257877858904e-05, + "loss": 0.2981, + "step": 845 + }, + { + "epoch": 1.403549743110696, + "grad_norm": 17.977771759033203, + "learning_rate": 5.7759800355920977e-05, + "loss": 0.2885, + "step": 846 + }, + { + "epoch": 1.4052104416420157, + "grad_norm": 43.73091125488281, + "learning_rate": 5.767131794019437e-05, + "loss": 0.3001, + "step": 847 + }, + { + "epoch": 1.4068711401733354, + "grad_norm": 36.181053161621094, + "learning_rate": 5.758281091453461e-05, + "loss": 0.289, + "step": 848 + }, + { + "epoch": 1.408531838704655, + "grad_norm": 115.1836166381836, + "learning_rate": 5.749427956287623e-05, + "loss": 0.2903, + "step": 849 + }, + { + "epoch": 1.410192537235975, + "grad_norm": 18.840404510498047, + "learning_rate": 5.7405724169231786e-05, + "loss": 0.2872, + "step": 850 + }, + { + "epoch": 1.4118532357672946, + "grad_norm": 23.613710403442383, + "learning_rate": 5.731714501769098e-05, + "loss": 0.2894, + "step": 851 + }, + { + "epoch": 1.4135139342986143, + "grad_norm": 10.212331771850586, + "learning_rate": 5.72285423924197e-05, + "loss": 0.2778, + "step": 852 + }, + { + "epoch": 1.4151746328299342, + "grad_norm": 39.11152648925781, + "learning_rate": 5.713991657765916e-05, + "loss": 0.2596, + "step": 853 + }, + { + "epoch": 1.4168353313612538, + "grad_norm": 30.35797882080078, + "learning_rate": 5.7051267857724966e-05, + "loss": 0.2973, + "step": 854 + }, + { + "epoch": 1.4184960298925735, + "grad_norm": 22.555871963500977, + "learning_rate": 5.6962596517006195e-05, + "loss": 0.2775, + "step": 855 + }, + { + "epoch": 1.4201567284238932, + "grad_norm": 19.49358367919922, + "learning_rate": 5.68739028399645e-05, + "loss": 0.2904, + "step": 856 + }, + { + "epoch": 1.421817426955213, + "grad_norm": 14.129653930664062, + "learning_rate": 5.67851871111332e-05, + "loss": 0.2616, + "step": 857 + }, + { + "epoch": 1.4234781254865327, + "grad_norm": 11.473206520080566, + "learning_rate": 5.669644961511631e-05, + "loss": 0.2806, + "step": 858 + }, + { + "epoch": 1.4251388240178526, + "grad_norm": 10.121034622192383, + "learning_rate": 5.660769063658775e-05, + "loss": 0.2793, + "step": 859 + }, + { + "epoch": 1.4267995225491723, + "grad_norm": 15.14218521118164, + "learning_rate": 5.651891046029031e-05, + "loss": 0.3146, + "step": 860 + }, + { + "epoch": 1.428460221080492, + "grad_norm": 105.05244445800781, + "learning_rate": 5.643010937103476e-05, + "loss": 0.2841, + "step": 861 + }, + { + "epoch": 1.4301209196118116, + "grad_norm": 23.236011505126953, + "learning_rate": 5.634128765369902e-05, + "loss": 0.2944, + "step": 862 + }, + { + "epoch": 1.4317816181431313, + "grad_norm": 50.88520050048828, + "learning_rate": 5.6252445593227134e-05, + "loss": 0.2859, + "step": 863 + }, + { + "epoch": 1.4334423166744512, + "grad_norm": 16.57847785949707, + "learning_rate": 5.616358347462847e-05, + "loss": 0.2764, + "step": 864 + }, + { + "epoch": 1.4351030152057709, + "grad_norm": 117.3586654663086, + "learning_rate": 5.607470158297665e-05, + "loss": 0.2769, + "step": 865 + }, + { + "epoch": 1.4367637137370908, + "grad_norm": 16.760282516479492, + "learning_rate": 5.598580020340881e-05, + "loss": 0.2894, + "step": 866 + }, + { + "epoch": 1.4384244122684104, + "grad_norm": 18.6272029876709, + "learning_rate": 5.5896879621124584e-05, + "loss": 0.3072, + "step": 867 + }, + { + "epoch": 1.4400851107997301, + "grad_norm": 20.149240493774414, + "learning_rate": 5.580794012138518e-05, + "loss": 0.299, + "step": 868 + }, + { + "epoch": 1.4417458093310498, + "grad_norm": 26.97914695739746, + "learning_rate": 5.571898198951253e-05, + "loss": 0.2925, + "step": 869 + }, + { + "epoch": 1.4434065078623697, + "grad_norm": 23.113183975219727, + "learning_rate": 5.563000551088833e-05, + "loss": 0.2942, + "step": 870 + }, + { + "epoch": 1.4450672063936894, + "grad_norm": 16.684524536132812, + "learning_rate": 5.55410109709531e-05, + "loss": 0.2873, + "step": 871 + }, + { + "epoch": 1.446727904925009, + "grad_norm": 13.51911735534668, + "learning_rate": 5.545199865520536e-05, + "loss": 0.2768, + "step": 872 + }, + { + "epoch": 1.448388603456329, + "grad_norm": 26.504819869995117, + "learning_rate": 5.5362968849200604e-05, + "loss": 0.2762, + "step": 873 + }, + { + "epoch": 1.4500493019876486, + "grad_norm": 26.66693115234375, + "learning_rate": 5.527392183855046e-05, + "loss": 0.2691, + "step": 874 + }, + { + "epoch": 1.4517100005189683, + "grad_norm": 238.01905822753906, + "learning_rate": 5.518485790892175e-05, + "loss": 0.2755, + "step": 875 + }, + { + "epoch": 1.453370699050288, + "grad_norm": 35.2214241027832, + "learning_rate": 5.509577734603555e-05, + "loss": 0.2787, + "step": 876 + }, + { + "epoch": 1.4550313975816078, + "grad_norm": 77.63333892822266, + "learning_rate": 5.5006680435666336e-05, + "loss": 0.3035, + "step": 877 + }, + { + "epoch": 1.4566920961129275, + "grad_norm": 45.94542694091797, + "learning_rate": 5.4917567463640994e-05, + "loss": 0.3067, + "step": 878 + }, + { + "epoch": 1.4583527946442472, + "grad_norm": 57.478336334228516, + "learning_rate": 5.482843871583797e-05, + "loss": 0.2688, + "step": 879 + }, + { + "epoch": 1.460013493175567, + "grad_norm": 15.525650978088379, + "learning_rate": 5.4739294478186274e-05, + "loss": 0.2594, + "step": 880 + }, + { + "epoch": 1.4616741917068867, + "grad_norm": 106.072998046875, + "learning_rate": 5.465013503666462e-05, + "loss": 0.2987, + "step": 881 + }, + { + "epoch": 1.4633348902382064, + "grad_norm": 152.6387481689453, + "learning_rate": 5.456096067730054e-05, + "loss": 0.2604, + "step": 882 + }, + { + "epoch": 1.464995588769526, + "grad_norm": 21.95989227294922, + "learning_rate": 5.447177168616937e-05, + "loss": 0.265, + "step": 883 + }, + { + "epoch": 1.466656287300846, + "grad_norm": 17.603464126586914, + "learning_rate": 5.4382568349393416e-05, + "loss": 0.2775, + "step": 884 + }, + { + "epoch": 1.4683169858321656, + "grad_norm": 25.807437896728516, + "learning_rate": 5.429335095314099e-05, + "loss": 0.2959, + "step": 885 + }, + { + "epoch": 1.4699776843634855, + "grad_norm": 13.686755180358887, + "learning_rate": 5.4204119783625515e-05, + "loss": 0.294, + "step": 886 + }, + { + "epoch": 1.4716383828948052, + "grad_norm": 22.57857322692871, + "learning_rate": 5.411487512710462e-05, + "loss": 0.2882, + "step": 887 + }, + { + "epoch": 1.4732990814261249, + "grad_norm": 27.24214744567871, + "learning_rate": 5.402561726987915e-05, + "loss": 0.276, + "step": 888 + }, + { + "epoch": 1.4749597799574445, + "grad_norm": 48.67518997192383, + "learning_rate": 5.3936346498292325e-05, + "loss": 0.3048, + "step": 889 + }, + { + "epoch": 1.4766204784887642, + "grad_norm": 29.256378173828125, + "learning_rate": 5.384706309872881e-05, + "loss": 0.2893, + "step": 890 + }, + { + "epoch": 1.478281177020084, + "grad_norm": 12.49537181854248, + "learning_rate": 5.3757767357613785e-05, + "loss": 0.2722, + "step": 891 + }, + { + "epoch": 1.4799418755514038, + "grad_norm": 110.85940551757812, + "learning_rate": 5.366845956141198e-05, + "loss": 0.2682, + "step": 892 + }, + { + "epoch": 1.4816025740827237, + "grad_norm": 22.701759338378906, + "learning_rate": 5.3579139996626816e-05, + "loss": 0.2936, + "step": 893 + }, + { + "epoch": 1.4832632726140433, + "grad_norm": 10.952596664428711, + "learning_rate": 5.348980894979952e-05, + "loss": 0.2907, + "step": 894 + }, + { + "epoch": 1.484923971145363, + "grad_norm": 58.649009704589844, + "learning_rate": 5.340046670750806e-05, + "loss": 0.2897, + "step": 895 + }, + { + "epoch": 1.4865846696766827, + "grad_norm": 11.12337589263916, + "learning_rate": 5.3311113556366396e-05, + "loss": 0.2896, + "step": 896 + }, + { + "epoch": 1.4882453682080026, + "grad_norm": 19.447628021240234, + "learning_rate": 5.322174978302345e-05, + "loss": 0.2907, + "step": 897 + }, + { + "epoch": 1.4899060667393222, + "grad_norm": 9.741415977478027, + "learning_rate": 5.3132375674162235e-05, + "loss": 0.2706, + "step": 898 + }, + { + "epoch": 1.491566765270642, + "grad_norm": 145.7992401123047, + "learning_rate": 5.304299151649892e-05, + "loss": 0.2901, + "step": 899 + }, + { + "epoch": 1.4932274638019618, + "grad_norm": 30.227359771728516, + "learning_rate": 5.2953597596781866e-05, + "loss": 0.2875, + "step": 900 + }, + { + "epoch": 1.4948881623332815, + "grad_norm": 23.82413673400879, + "learning_rate": 5.286419420179084e-05, + "loss": 0.2831, + "step": 901 + }, + { + "epoch": 1.4965488608646011, + "grad_norm": 166.69720458984375, + "learning_rate": 5.277478161833593e-05, + "loss": 0.3017, + "step": 902 + }, + { + "epoch": 1.4982095593959208, + "grad_norm": 7.857309818267822, + "learning_rate": 5.268536013325672e-05, + "loss": 0.2563, + "step": 903 + }, + { + "epoch": 1.4998702579272407, + "grad_norm": 38.910926818847656, + "learning_rate": 5.2595930033421383e-05, + "loss": 0.2964, + "step": 904 + }, + { + "epoch": 1.5015309564585604, + "grad_norm": 15.284276008605957, + "learning_rate": 5.250649160572567e-05, + "loss": 0.3087, + "step": 905 + }, + { + "epoch": 1.5031916549898803, + "grad_norm": 10.590121269226074, + "learning_rate": 5.2417045137092126e-05, + "loss": 0.2737, + "step": 906 + }, + { + "epoch": 1.5048523535212, + "grad_norm": 14.894525527954102, + "learning_rate": 5.232759091446902e-05, + "loss": 0.2829, + "step": 907 + }, + { + "epoch": 1.5065130520525196, + "grad_norm": 12.605585098266602, + "learning_rate": 5.22381292248295e-05, + "loss": 0.2906, + "step": 908 + }, + { + "epoch": 1.5081737505838393, + "grad_norm": 27.536279678344727, + "learning_rate": 5.214866035517074e-05, + "loss": 0.2945, + "step": 909 + }, + { + "epoch": 1.509834449115159, + "grad_norm": 18.08845329284668, + "learning_rate": 5.205918459251287e-05, + "loss": 0.296, + "step": 910 + }, + { + "epoch": 1.5114951476464789, + "grad_norm": 183.22830200195312, + "learning_rate": 5.1969702223898174e-05, + "loss": 0.2845, + "step": 911 + }, + { + "epoch": 1.5131558461777985, + "grad_norm": 14.84533405303955, + "learning_rate": 5.18802135363901e-05, + "loss": 0.2961, + "step": 912 + }, + { + "epoch": 1.5148165447091184, + "grad_norm": 136.57386779785156, + "learning_rate": 5.1790718817072404e-05, + "loss": 0.2949, + "step": 913 + }, + { + "epoch": 1.516477243240438, + "grad_norm": 47.80552673339844, + "learning_rate": 5.170121835304817e-05, + "loss": 0.2773, + "step": 914 + }, + { + "epoch": 1.5181379417717578, + "grad_norm": 29.221660614013672, + "learning_rate": 5.161171243143891e-05, + "loss": 0.2775, + "step": 915 + }, + { + "epoch": 1.5197986403030774, + "grad_norm": 14.316165924072266, + "learning_rate": 5.1522201339383634e-05, + "loss": 0.2898, + "step": 916 + }, + { + "epoch": 1.521459338834397, + "grad_norm": 21.883058547973633, + "learning_rate": 5.143268536403798e-05, + "loss": 0.2833, + "step": 917 + }, + { + "epoch": 1.523120037365717, + "grad_norm": 31.403982162475586, + "learning_rate": 5.134316479257321e-05, + "loss": 0.2895, + "step": 918 + }, + { + "epoch": 1.5247807358970367, + "grad_norm": 38.35506057739258, + "learning_rate": 5.125363991217536e-05, + "loss": 0.3044, + "step": 919 + }, + { + "epoch": 1.5264414344283566, + "grad_norm": 19.42740249633789, + "learning_rate": 5.116411101004422e-05, + "loss": 0.266, + "step": 920 + }, + { + "epoch": 1.5281021329596762, + "grad_norm": 14.785085678100586, + "learning_rate": 5.1074578373392615e-05, + "loss": 0.298, + "step": 921 + }, + { + "epoch": 1.529762831490996, + "grad_norm": 11.301557540893555, + "learning_rate": 5.0985042289445214e-05, + "loss": 0.2865, + "step": 922 + }, + { + "epoch": 1.5314235300223156, + "grad_norm": 23.628995895385742, + "learning_rate": 5.0895503045437845e-05, + "loss": 0.2639, + "step": 923 + }, + { + "epoch": 1.5330842285536352, + "grad_norm": 14.561969757080078, + "learning_rate": 5.08059609286164e-05, + "loss": 0.2686, + "step": 924 + }, + { + "epoch": 1.5347449270849551, + "grad_norm": 19.491962432861328, + "learning_rate": 5.071641622623602e-05, + "loss": 0.2711, + "step": 925 + }, + { + "epoch": 1.5364056256162748, + "grad_norm": 12.267081260681152, + "learning_rate": 5.062686922556017e-05, + "loss": 0.3125, + "step": 926 + }, + { + "epoch": 1.5380663241475947, + "grad_norm": 14.002256393432617, + "learning_rate": 5.053732021385963e-05, + "loss": 0.2772, + "step": 927 + }, + { + "epoch": 1.5397270226789144, + "grad_norm": 82.95569610595703, + "learning_rate": 5.044776947841165e-05, + "loss": 0.2553, + "step": 928 + }, + { + "epoch": 1.541387721210234, + "grad_norm": 56.53269577026367, + "learning_rate": 5.0358217306499057e-05, + "loss": 0.2951, + "step": 929 + }, + { + "epoch": 1.5430484197415537, + "grad_norm": 102.66098022460938, + "learning_rate": 5.026866398540922e-05, + "loss": 0.3053, + "step": 930 + }, + { + "epoch": 1.5447091182728734, + "grad_norm": 9.849115371704102, + "learning_rate": 5.0179109802433213e-05, + "loss": 0.2887, + "step": 931 + }, + { + "epoch": 1.5463698168041933, + "grad_norm": 40.51551818847656, + "learning_rate": 5.008955504486491e-05, + "loss": 0.2952, + "step": 932 + }, + { + "epoch": 1.548030515335513, + "grad_norm": 16.44049835205078, + "learning_rate": 5e-05, + "loss": 0.2942, + "step": 933 + }, + { + "epoch": 1.5496912138668328, + "grad_norm": 58.39529800415039, + "learning_rate": 4.9910444955135096e-05, + "loss": 0.3037, + "step": 934 + }, + { + "epoch": 1.5513519123981525, + "grad_norm": 14.203904151916504, + "learning_rate": 4.982089019756679e-05, + "loss": 0.284, + "step": 935 + }, + { + "epoch": 1.5530126109294722, + "grad_norm": 51.40688705444336, + "learning_rate": 4.973133601459079e-05, + "loss": 0.2792, + "step": 936 + }, + { + "epoch": 1.5546733094607919, + "grad_norm": 40.534080505371094, + "learning_rate": 4.9641782693500955e-05, + "loss": 0.2706, + "step": 937 + }, + { + "epoch": 1.5563340079921115, + "grad_norm": 103.09912872314453, + "learning_rate": 4.955223052158835e-05, + "loss": 0.2764, + "step": 938 + }, + { + "epoch": 1.5579947065234314, + "grad_norm": 25.620494842529297, + "learning_rate": 4.9462679786140385e-05, + "loss": 0.2747, + "step": 939 + }, + { + "epoch": 1.5596554050547513, + "grad_norm": 16.26183319091797, + "learning_rate": 4.937313077443985e-05, + "loss": 0.2776, + "step": 940 + }, + { + "epoch": 1.561316103586071, + "grad_norm": 21.900814056396484, + "learning_rate": 4.928358377376398e-05, + "loss": 0.3061, + "step": 941 + }, + { + "epoch": 1.5629768021173907, + "grad_norm": 15.629485130310059, + "learning_rate": 4.919403907138361e-05, + "loss": 0.3067, + "step": 942 + }, + { + "epoch": 1.5646375006487103, + "grad_norm": 34.60820388793945, + "learning_rate": 4.910449695456216e-05, + "loss": 0.3004, + "step": 943 + }, + { + "epoch": 1.56629819918003, + "grad_norm": 31.900197982788086, + "learning_rate": 4.901495771055479e-05, + "loss": 0.2865, + "step": 944 + }, + { + "epoch": 1.5679588977113499, + "grad_norm": 25.575546264648438, + "learning_rate": 4.892542162660741e-05, + "loss": 0.2839, + "step": 945 + }, + { + "epoch": 1.5696195962426696, + "grad_norm": 53.05941390991211, + "learning_rate": 4.8835888989955776e-05, + "loss": 0.2714, + "step": 946 + }, + { + "epoch": 1.5712802947739894, + "grad_norm": 29.1506290435791, + "learning_rate": 4.874636008782466e-05, + "loss": 0.2828, + "step": 947 + }, + { + "epoch": 1.5729409933053091, + "grad_norm": 29.428146362304688, + "learning_rate": 4.8656835207426795e-05, + "loss": 0.2962, + "step": 948 + }, + { + "epoch": 1.5746016918366288, + "grad_norm": 643.2196655273438, + "learning_rate": 4.856731463596204e-05, + "loss": 0.2685, + "step": 949 + }, + { + "epoch": 1.5762623903679485, + "grad_norm": 51.18458557128906, + "learning_rate": 4.847779866061638e-05, + "loss": 0.2816, + "step": 950 + }, + { + "epoch": 1.5779230888992681, + "grad_norm": 19.216337203979492, + "learning_rate": 4.8388287568561105e-05, + "loss": 0.2869, + "step": 951 + }, + { + "epoch": 1.579583787430588, + "grad_norm": 9.958168983459473, + "learning_rate": 4.829878164695183e-05, + "loss": 0.2868, + "step": 952 + }, + { + "epoch": 1.5812444859619077, + "grad_norm": 15.129812240600586, + "learning_rate": 4.820928118292761e-05, + "loss": 0.2946, + "step": 953 + }, + { + "epoch": 1.5829051844932276, + "grad_norm": 13.884086608886719, + "learning_rate": 4.8119786463609905e-05, + "loss": 0.2758, + "step": 954 + }, + { + "epoch": 1.5845658830245473, + "grad_norm": 15.602642059326172, + "learning_rate": 4.8030297776101844e-05, + "loss": 0.2817, + "step": 955 + }, + { + "epoch": 1.586226581555867, + "grad_norm": 59.39334487915039, + "learning_rate": 4.7940815407487135e-05, + "loss": 0.2971, + "step": 956 + }, + { + "epoch": 1.5878872800871866, + "grad_norm": 75.20787048339844, + "learning_rate": 4.785133964482927e-05, + "loss": 0.2871, + "step": 957 + }, + { + "epoch": 1.5895479786185063, + "grad_norm": 17.12590217590332, + "learning_rate": 4.776187077517051e-05, + "loss": 0.2637, + "step": 958 + }, + { + "epoch": 1.5912086771498262, + "grad_norm": 14.124485969543457, + "learning_rate": 4.7672409085531e-05, + "loss": 0.2921, + "step": 959 + }, + { + "epoch": 1.5928693756811458, + "grad_norm": 25.335908889770508, + "learning_rate": 4.758295486290789e-05, + "loss": 0.2742, + "step": 960 + }, + { + "epoch": 1.5945300742124657, + "grad_norm": 17.80637550354004, + "learning_rate": 4.7493508394274325e-05, + "loss": 0.2959, + "step": 961 + }, + { + "epoch": 1.5961907727437854, + "grad_norm": 61.27547073364258, + "learning_rate": 4.7404069966578635e-05, + "loss": 0.2803, + "step": 962 + }, + { + "epoch": 1.597851471275105, + "grad_norm": 10.976715087890625, + "learning_rate": 4.7314639866743285e-05, + "loss": 0.2949, + "step": 963 + }, + { + "epoch": 1.5995121698064247, + "grad_norm": 11.593228340148926, + "learning_rate": 4.7225218381664094e-05, + "loss": 0.2861, + "step": 964 + }, + { + "epoch": 1.6011728683377444, + "grad_norm": 11.976430892944336, + "learning_rate": 4.7135805798209186e-05, + "loss": 0.2739, + "step": 965 + }, + { + "epoch": 1.6028335668690643, + "grad_norm": 14.272520065307617, + "learning_rate": 4.704640240321814e-05, + "loss": 0.2571, + "step": 966 + }, + { + "epoch": 1.6044942654003842, + "grad_norm": 11.826343536376953, + "learning_rate": 4.69570084835011e-05, + "loss": 0.2832, + "step": 967 + }, + { + "epoch": 1.6061549639317039, + "grad_norm": 78.57592010498047, + "learning_rate": 4.6867624325837777e-05, + "loss": 0.2978, + "step": 968 + }, + { + "epoch": 1.6078156624630235, + "grad_norm": 29.39203453063965, + "learning_rate": 4.6778250216976565e-05, + "loss": 0.2915, + "step": 969 + }, + { + "epoch": 1.6094763609943432, + "grad_norm": 22.652767181396484, + "learning_rate": 4.668888644363361e-05, + "loss": 0.2755, + "step": 970 + }, + { + "epoch": 1.6111370595256629, + "grad_norm": 15.469453811645508, + "learning_rate": 4.6599533292491946e-05, + "loss": 0.3017, + "step": 971 + }, + { + "epoch": 1.6127977580569826, + "grad_norm": 19.697040557861328, + "learning_rate": 4.6510191050200495e-05, + "loss": 0.2823, + "step": 972 + }, + { + "epoch": 1.6144584565883024, + "grad_norm": 14.159893035888672, + "learning_rate": 4.642086000337319e-05, + "loss": 0.2567, + "step": 973 + }, + { + "epoch": 1.6161191551196223, + "grad_norm": 6.8621907234191895, + "learning_rate": 4.633154043858804e-05, + "loss": 0.3165, + "step": 974 + }, + { + "epoch": 1.617779853650942, + "grad_norm": 48.186275482177734, + "learning_rate": 4.624223264238622e-05, + "loss": 0.281, + "step": 975 + }, + { + "epoch": 1.6194405521822617, + "grad_norm": 8.274702072143555, + "learning_rate": 4.6152936901271186e-05, + "loss": 0.2941, + "step": 976 + }, + { + "epoch": 1.6211012507135814, + "grad_norm": 7.512510299682617, + "learning_rate": 4.6063653501707686e-05, + "loss": 0.2849, + "step": 977 + }, + { + "epoch": 1.622761949244901, + "grad_norm": 38.67232894897461, + "learning_rate": 4.597438273012087e-05, + "loss": 0.2843, + "step": 978 + }, + { + "epoch": 1.624422647776221, + "grad_norm": 12.054966926574707, + "learning_rate": 4.58851248728954e-05, + "loss": 0.2862, + "step": 979 + }, + { + "epoch": 1.6260833463075406, + "grad_norm": 21.105966567993164, + "learning_rate": 4.579588021637448e-05, + "loss": 0.2787, + "step": 980 + }, + { + "epoch": 1.6277440448388605, + "grad_norm": 12.212367057800293, + "learning_rate": 4.5706649046859016e-05, + "loss": 0.2787, + "step": 981 + }, + { + "epoch": 1.6294047433701802, + "grad_norm": 32.4214973449707, + "learning_rate": 4.561743165060659e-05, + "loss": 0.303, + "step": 982 + }, + { + "epoch": 1.6310654419014998, + "grad_norm": 25.607418060302734, + "learning_rate": 4.5528228313830646e-05, + "loss": 0.2736, + "step": 983 + }, + { + "epoch": 1.6327261404328195, + "grad_norm": 19.603490829467773, + "learning_rate": 4.5439039322699485e-05, + "loss": 0.2938, + "step": 984 + }, + { + "epoch": 1.6343868389641392, + "grad_norm": 19.48715591430664, + "learning_rate": 4.534986496333539e-05, + "loss": 0.2537, + "step": 985 + }, + { + "epoch": 1.636047537495459, + "grad_norm": 126.12147521972656, + "learning_rate": 4.526070552181374e-05, + "loss": 0.2757, + "step": 986 + }, + { + "epoch": 1.6377082360267787, + "grad_norm": 7.359575271606445, + "learning_rate": 4.5171561284162044e-05, + "loss": 0.2766, + "step": 987 + }, + { + "epoch": 1.6393689345580986, + "grad_norm": 48.8001708984375, + "learning_rate": 4.508243253635902e-05, + "loss": 0.2811, + "step": 988 + }, + { + "epoch": 1.6410296330894183, + "grad_norm": 18.930316925048828, + "learning_rate": 4.4993319564333675e-05, + "loss": 0.2917, + "step": 989 + }, + { + "epoch": 1.642690331620738, + "grad_norm": 15.069755554199219, + "learning_rate": 4.490422265396445e-05, + "loss": 0.2813, + "step": 990 + }, + { + "epoch": 1.6443510301520576, + "grad_norm": 47.264564514160156, + "learning_rate": 4.481514209107827e-05, + "loss": 0.2856, + "step": 991 + }, + { + "epoch": 1.6460117286833773, + "grad_norm": 16.947811126708984, + "learning_rate": 4.472607816144956e-05, + "loss": 0.2865, + "step": 992 + }, + { + "epoch": 1.6476724272146972, + "grad_norm": 80.86692810058594, + "learning_rate": 4.463703115079941e-05, + "loss": 0.2783, + "step": 993 + }, + { + "epoch": 1.6493331257460169, + "grad_norm": 35.8098258972168, + "learning_rate": 4.454800134479466e-05, + "loss": 0.2531, + "step": 994 + }, + { + "epoch": 1.6509938242773368, + "grad_norm": 48.696353912353516, + "learning_rate": 4.445898902904689e-05, + "loss": 0.2937, + "step": 995 + }, + { + "epoch": 1.6526545228086564, + "grad_norm": 25.220211029052734, + "learning_rate": 4.436999448911168e-05, + "loss": 0.2969, + "step": 996 + }, + { + "epoch": 1.654315221339976, + "grad_norm": 17.5574893951416, + "learning_rate": 4.428101801048748e-05, + "loss": 0.2892, + "step": 997 + }, + { + "epoch": 1.6559759198712958, + "grad_norm": 57.936744689941406, + "learning_rate": 4.419205987861483e-05, + "loss": 0.2985, + "step": 998 + }, + { + "epoch": 1.6576366184026154, + "grad_norm": 37.452117919921875, + "learning_rate": 4.4103120378875434e-05, + "loss": 0.2907, + "step": 999 + }, + { + "epoch": 1.6592973169339353, + "grad_norm": 10.426920890808105, + "learning_rate": 4.4014199796591193e-05, + "loss": 0.276, + "step": 1000 + }, + { + "epoch": 1.6609580154652552, + "grad_norm": 13.090839385986328, + "learning_rate": 4.392529841702336e-05, + "loss": 0.2706, + "step": 1001 + }, + { + "epoch": 1.662618713996575, + "grad_norm": 10.039213180541992, + "learning_rate": 4.383641652537155e-05, + "loss": 0.2779, + "step": 1002 + }, + { + "epoch": 1.6642794125278946, + "grad_norm": 44.60422134399414, + "learning_rate": 4.374755440677287e-05, + "loss": 0.28, + "step": 1003 + }, + { + "epoch": 1.6659401110592142, + "grad_norm": 16.090087890625, + "learning_rate": 4.3658712346301006e-05, + "loss": 0.3198, + "step": 1004 + }, + { + "epoch": 1.667600809590534, + "grad_norm": 10.215002059936523, + "learning_rate": 4.356989062896525e-05, + "loss": 0.263, + "step": 1005 + }, + { + "epoch": 1.6692615081218538, + "grad_norm": 19.978612899780273, + "learning_rate": 4.3481089539709705e-05, + "loss": 0.2816, + "step": 1006 + }, + { + "epoch": 1.6709222066531735, + "grad_norm": 11.551654815673828, + "learning_rate": 4.339230936341226e-05, + "loss": 0.2845, + "step": 1007 + }, + { + "epoch": 1.6725829051844934, + "grad_norm": 62.54079818725586, + "learning_rate": 4.33035503848837e-05, + "loss": 0.2798, + "step": 1008 + }, + { + "epoch": 1.674243603715813, + "grad_norm": 11.756832122802734, + "learning_rate": 4.321481288886682e-05, + "loss": 0.2729, + "step": 1009 + }, + { + "epoch": 1.6759043022471327, + "grad_norm": 90.3555679321289, + "learning_rate": 4.31260971600355e-05, + "loss": 0.2921, + "step": 1010 + }, + { + "epoch": 1.6775650007784524, + "grad_norm": 38.02375030517578, + "learning_rate": 4.303740348299381e-05, + "loss": 0.2605, + "step": 1011 + }, + { + "epoch": 1.679225699309772, + "grad_norm": 15.210528373718262, + "learning_rate": 4.294873214227505e-05, + "loss": 0.2571, + "step": 1012 + }, + { + "epoch": 1.680886397841092, + "grad_norm": 53.868194580078125, + "learning_rate": 4.2860083422340846e-05, + "loss": 0.2959, + "step": 1013 + }, + { + "epoch": 1.6825470963724116, + "grad_norm": 27.089468002319336, + "learning_rate": 4.2771457607580315e-05, + "loss": 0.2863, + "step": 1014 + }, + { + "epoch": 1.6842077949037315, + "grad_norm": 16.614988327026367, + "learning_rate": 4.268285498230903e-05, + "loss": 0.2635, + "step": 1015 + }, + { + "epoch": 1.6858684934350512, + "grad_norm": 21.056549072265625, + "learning_rate": 4.259427583076822e-05, + "loss": 0.2999, + "step": 1016 + }, + { + "epoch": 1.6875291919663709, + "grad_norm": 681.7119750976562, + "learning_rate": 4.2505720437123783e-05, + "loss": 0.2891, + "step": 1017 + }, + { + "epoch": 1.6891898904976905, + "grad_norm": 41.42094802856445, + "learning_rate": 4.241718908546541e-05, + "loss": 0.272, + "step": 1018 + }, + { + "epoch": 1.6908505890290102, + "grad_norm": 23.537466049194336, + "learning_rate": 4.232868205980566e-05, + "loss": 0.2763, + "step": 1019 + }, + { + "epoch": 1.69251128756033, + "grad_norm": 9.240571022033691, + "learning_rate": 4.224019964407902e-05, + "loss": 0.2801, + "step": 1020 + }, + { + "epoch": 1.6941719860916498, + "grad_norm": 16.439271926879883, + "learning_rate": 4.21517421221411e-05, + "loss": 0.2595, + "step": 1021 + }, + { + "epoch": 1.6958326846229697, + "grad_norm": 20.681480407714844, + "learning_rate": 4.2063309777767626e-05, + "loss": 0.2673, + "step": 1022 + }, + { + "epoch": 1.6974933831542893, + "grad_norm": 8.995615005493164, + "learning_rate": 4.197490289465351e-05, + "loss": 0.28, + "step": 1023 + }, + { + "epoch": 1.699154081685609, + "grad_norm": 8.972556114196777, + "learning_rate": 4.188652175641199e-05, + "loss": 0.286, + "step": 1024 + }, + { + "epoch": 1.7008147802169287, + "grad_norm": 35.38237380981445, + "learning_rate": 4.179816664657377e-05, + "loss": 0.275, + "step": 1025 + }, + { + "epoch": 1.7024754787482483, + "grad_norm": 19.524112701416016, + "learning_rate": 4.170983784858601e-05, + "loss": 0.2913, + "step": 1026 + }, + { + "epoch": 1.7041361772795682, + "grad_norm": 45.598541259765625, + "learning_rate": 4.162153564581148e-05, + "loss": 0.2669, + "step": 1027 + }, + { + "epoch": 1.705796875810888, + "grad_norm": 31.71568489074707, + "learning_rate": 4.153326032152758e-05, + "loss": 0.3152, + "step": 1028 + }, + { + "epoch": 1.7074575743422078, + "grad_norm": 10.269993782043457, + "learning_rate": 4.144501215892557e-05, + "loss": 0.2741, + "step": 1029 + }, + { + "epoch": 1.7091182728735275, + "grad_norm": 13.3892183303833, + "learning_rate": 4.135679144110948e-05, + "loss": 0.2779, + "step": 1030 + }, + { + "epoch": 1.7107789714048471, + "grad_norm": 19.40374755859375, + "learning_rate": 4.126859845109539e-05, + "loss": 0.2744, + "step": 1031 + }, + { + "epoch": 1.7124396699361668, + "grad_norm": 33.533958435058594, + "learning_rate": 4.118043347181035e-05, + "loss": 0.2714, + "step": 1032 + }, + { + "epoch": 1.7141003684674865, + "grad_norm": 14.68758773803711, + "learning_rate": 4.109229678609159e-05, + "loss": 0.2915, + "step": 1033 + }, + { + "epoch": 1.7157610669988064, + "grad_norm": 45.297080993652344, + "learning_rate": 4.1004188676685565e-05, + "loss": 0.2834, + "step": 1034 + }, + { + "epoch": 1.7174217655301263, + "grad_norm": 13.071781158447266, + "learning_rate": 4.0916109426247034e-05, + "loss": 0.3161, + "step": 1035 + }, + { + "epoch": 1.719082464061446, + "grad_norm": 8.669297218322754, + "learning_rate": 4.082805931733822e-05, + "loss": 0.2842, + "step": 1036 + }, + { + "epoch": 1.7207431625927656, + "grad_norm": 8.116327285766602, + "learning_rate": 4.074003863242784e-05, + "loss": 0.2664, + "step": 1037 + }, + { + "epoch": 1.7224038611240853, + "grad_norm": 18.077898025512695, + "learning_rate": 4.06520476538902e-05, + "loss": 0.2928, + "step": 1038 + }, + { + "epoch": 1.724064559655405, + "grad_norm": 22.521236419677734, + "learning_rate": 4.0564086664004294e-05, + "loss": 0.3005, + "step": 1039 + }, + { + "epoch": 1.7257252581867248, + "grad_norm": 11.232478141784668, + "learning_rate": 4.0476155944952986e-05, + "loss": 0.2731, + "step": 1040 + }, + { + "epoch": 1.7273859567180445, + "grad_norm": 23.123260498046875, + "learning_rate": 4.038825577882194e-05, + "loss": 0.2905, + "step": 1041 + }, + { + "epoch": 1.7290466552493644, + "grad_norm": 7.232083320617676, + "learning_rate": 4.030038644759887e-05, + "loss": 0.2517, + "step": 1042 + }, + { + "epoch": 1.730707353780684, + "grad_norm": 39.02933120727539, + "learning_rate": 4.021254823317252e-05, + "loss": 0.2685, + "step": 1043 + }, + { + "epoch": 1.7323680523120037, + "grad_norm": 72.39202880859375, + "learning_rate": 4.0124741417331837e-05, + "loss": 0.2909, + "step": 1044 + }, + { + "epoch": 1.7340287508433234, + "grad_norm": 162.18104553222656, + "learning_rate": 4.0036966281765046e-05, + "loss": 0.2817, + "step": 1045 + }, + { + "epoch": 1.735689449374643, + "grad_norm": 7.645057678222656, + "learning_rate": 3.994922310805875e-05, + "loss": 0.2791, + "step": 1046 + }, + { + "epoch": 1.737350147905963, + "grad_norm": 67.43783569335938, + "learning_rate": 3.986151217769698e-05, + "loss": 0.3013, + "step": 1047 + }, + { + "epoch": 1.7390108464372827, + "grad_norm": 24.866619110107422, + "learning_rate": 3.977383377206033e-05, + "loss": 0.2759, + "step": 1048 + }, + { + "epoch": 1.7406715449686025, + "grad_norm": 16.723388671875, + "learning_rate": 3.968618817242509e-05, + "loss": 0.2699, + "step": 1049 + }, + { + "epoch": 1.7423322434999222, + "grad_norm": 10.21795654296875, + "learning_rate": 3.9598575659962294e-05, + "loss": 0.2747, + "step": 1050 + }, + { + "epoch": 1.7439929420312419, + "grad_norm": 253.9533233642578, + "learning_rate": 3.951099651573682e-05, + "loss": 0.2969, + "step": 1051 + }, + { + "epoch": 1.7456536405625616, + "grad_norm": 149.43836975097656, + "learning_rate": 3.942345102070649e-05, + "loss": 0.2717, + "step": 1052 + }, + { + "epoch": 1.7473143390938812, + "grad_norm": 10.979608535766602, + "learning_rate": 3.933593945572121e-05, + "loss": 0.2866, + "step": 1053 + }, + { + "epoch": 1.7489750376252011, + "grad_norm": 11.278124809265137, + "learning_rate": 3.9248462101521986e-05, + "loss": 0.2948, + "step": 1054 + }, + { + "epoch": 1.7506357361565208, + "grad_norm": 29.37539291381836, + "learning_rate": 3.916101923874015e-05, + "loss": 0.2946, + "step": 1055 + }, + { + "epoch": 1.7522964346878407, + "grad_norm": 9.42124080657959, + "learning_rate": 3.907361114789632e-05, + "loss": 0.3163, + "step": 1056 + }, + { + "epoch": 1.7539571332191604, + "grad_norm": 14.308135032653809, + "learning_rate": 3.8986238109399584e-05, + "loss": 0.2605, + "step": 1057 + }, + { + "epoch": 1.75561783175048, + "grad_norm": 15.659673690795898, + "learning_rate": 3.889890040354658e-05, + "loss": 0.2615, + "step": 1058 + }, + { + "epoch": 1.7572785302817997, + "grad_norm": 15.974221229553223, + "learning_rate": 3.8811598310520585e-05, + "loss": 0.2848, + "step": 1059 + }, + { + "epoch": 1.7589392288131194, + "grad_norm": 10.16062068939209, + "learning_rate": 3.8724332110390665e-05, + "loss": 0.3013, + "step": 1060 + }, + { + "epoch": 1.7605999273444393, + "grad_norm": 10.001726150512695, + "learning_rate": 3.863710208311072e-05, + "loss": 0.278, + "step": 1061 + }, + { + "epoch": 1.7622606258757592, + "grad_norm": 18.319469451904297, + "learning_rate": 3.854990850851859e-05, + "loss": 0.2814, + "step": 1062 + }, + { + "epoch": 1.7639213244070788, + "grad_norm": 65.80924987792969, + "learning_rate": 3.846275166633517e-05, + "loss": 0.2782, + "step": 1063 + }, + { + "epoch": 1.7655820229383985, + "grad_norm": 13.594179153442383, + "learning_rate": 3.837563183616355e-05, + "loss": 0.2663, + "step": 1064 + }, + { + "epoch": 1.7672427214697182, + "grad_norm": 80.1303939819336, + "learning_rate": 3.828854929748807e-05, + "loss": 0.2744, + "step": 1065 + }, + { + "epoch": 1.7689034200010378, + "grad_norm": 19.906957626342773, + "learning_rate": 3.820150432967344e-05, + "loss": 0.2721, + "step": 1066 + }, + { + "epoch": 1.7705641185323575, + "grad_norm": 29.24519157409668, + "learning_rate": 3.81144972119638e-05, + "loss": 0.2596, + "step": 1067 + }, + { + "epoch": 1.7722248170636774, + "grad_norm": 20.713769912719727, + "learning_rate": 3.8027528223481936e-05, + "loss": 0.2548, + "step": 1068 + }, + { + "epoch": 1.7738855155949973, + "grad_norm": 15.758745193481445, + "learning_rate": 3.7940597643228234e-05, + "loss": 0.2839, + "step": 1069 + }, + { + "epoch": 1.775546214126317, + "grad_norm": 18.031856536865234, + "learning_rate": 3.785370575007995e-05, + "loss": 0.2809, + "step": 1070 + }, + { + "epoch": 1.7772069126576366, + "grad_norm": 44.73027420043945, + "learning_rate": 3.776685282279016e-05, + "loss": 0.2745, + "step": 1071 + }, + { + "epoch": 1.7788676111889563, + "grad_norm": 436.51556396484375, + "learning_rate": 3.7680039139986986e-05, + "loss": 0.2727, + "step": 1072 + }, + { + "epoch": 1.780528309720276, + "grad_norm": 14.402546882629395, + "learning_rate": 3.759326498017261e-05, + "loss": 0.2882, + "step": 1073 + }, + { + "epoch": 1.7821890082515959, + "grad_norm": 134.99525451660156, + "learning_rate": 3.750653062172243e-05, + "loss": 0.2689, + "step": 1074 + }, + { + "epoch": 1.7838497067829155, + "grad_norm": 11.556097030639648, + "learning_rate": 3.7419836342884184e-05, + "loss": 0.2804, + "step": 1075 + }, + { + "epoch": 1.7855104053142354, + "grad_norm": 44.521366119384766, + "learning_rate": 3.733318242177705e-05, + "loss": 0.2849, + "step": 1076 + }, + { + "epoch": 1.787171103845555, + "grad_norm": 23.42568016052246, + "learning_rate": 3.724656913639066e-05, + "loss": 0.2792, + "step": 1077 + }, + { + "epoch": 1.7888318023768748, + "grad_norm": 29.743019104003906, + "learning_rate": 3.715999676458435e-05, + "loss": 0.2749, + "step": 1078 + }, + { + "epoch": 1.7904925009081945, + "grad_norm": 82.91143035888672, + "learning_rate": 3.7073465584086186e-05, + "loss": 0.3039, + "step": 1079 + }, + { + "epoch": 1.7921531994395141, + "grad_norm": 76.2025375366211, + "learning_rate": 3.698697587249208e-05, + "loss": 0.2982, + "step": 1080 + }, + { + "epoch": 1.793813897970834, + "grad_norm": 10.867552757263184, + "learning_rate": 3.6900527907264936e-05, + "loss": 0.2759, + "step": 1081 + }, + { + "epoch": 1.7954745965021537, + "grad_norm": 22.257034301757812, + "learning_rate": 3.681412196573369e-05, + "loss": 0.2957, + "step": 1082 + }, + { + "epoch": 1.7971352950334736, + "grad_norm": 9.685708999633789, + "learning_rate": 3.672775832509253e-05, + "loss": 0.2831, + "step": 1083 + }, + { + "epoch": 1.7987959935647932, + "grad_norm": 13.656418800354004, + "learning_rate": 3.664143726239984e-05, + "loss": 0.2978, + "step": 1084 + }, + { + "epoch": 1.800456692096113, + "grad_norm": 650.8150024414062, + "learning_rate": 3.655515905457754e-05, + "loss": 0.2923, + "step": 1085 + }, + { + "epoch": 1.8021173906274326, + "grad_norm": 16.22422218322754, + "learning_rate": 3.646892397840997e-05, + "loss": 0.2897, + "step": 1086 + }, + { + "epoch": 1.8037780891587523, + "grad_norm": 7.556303024291992, + "learning_rate": 3.638273231054313e-05, + "loss": 0.2841, + "step": 1087 + }, + { + "epoch": 1.8054387876900722, + "grad_norm": 31.054821014404297, + "learning_rate": 3.6296584327483785e-05, + "loss": 0.2795, + "step": 1088 + }, + { + "epoch": 1.8070994862213918, + "grad_norm": 20.195293426513672, + "learning_rate": 3.621048030559851e-05, + "loss": 0.2612, + "step": 1089 + }, + { + "epoch": 1.8087601847527117, + "grad_norm": 18.975658416748047, + "learning_rate": 3.612442052111292e-05, + "loss": 0.2642, + "step": 1090 + }, + { + "epoch": 1.8104208832840314, + "grad_norm": 28.037137985229492, + "learning_rate": 3.603840525011066e-05, + "loss": 0.2758, + "step": 1091 + }, + { + "epoch": 1.812081581815351, + "grad_norm": 10.855775833129883, + "learning_rate": 3.59524347685326e-05, + "loss": 0.2637, + "step": 1092 + }, + { + "epoch": 1.8137422803466707, + "grad_norm": 6.196507930755615, + "learning_rate": 3.58665093521759e-05, + "loss": 0.2668, + "step": 1093 + }, + { + "epoch": 1.8154029788779904, + "grad_norm": 9.843585968017578, + "learning_rate": 3.5780629276693145e-05, + "loss": 0.276, + "step": 1094 + }, + { + "epoch": 1.8170636774093103, + "grad_norm": 17.883283615112305, + "learning_rate": 3.569479481759152e-05, + "loss": 0.2542, + "step": 1095 + }, + { + "epoch": 1.8187243759406302, + "grad_norm": 15.545995712280273, + "learning_rate": 3.5609006250231815e-05, + "loss": 0.2707, + "step": 1096 + }, + { + "epoch": 1.8203850744719499, + "grad_norm": 7.960692882537842, + "learning_rate": 3.552326384982761e-05, + "loss": 0.2663, + "step": 1097 + }, + { + "epoch": 1.8220457730032695, + "grad_norm": 28.72408103942871, + "learning_rate": 3.543756789144439e-05, + "loss": 0.2805, + "step": 1098 + }, + { + "epoch": 1.8237064715345892, + "grad_norm": 177.16275024414062, + "learning_rate": 3.535191864999862e-05, + "loss": 0.2801, + "step": 1099 + }, + { + "epoch": 1.8253671700659089, + "grad_norm": 32.54202651977539, + "learning_rate": 3.526631640025697e-05, + "loss": 0.2756, + "step": 1100 + }, + { + "epoch": 1.8270278685972285, + "grad_norm": 21.565338134765625, + "learning_rate": 3.518076141683528e-05, + "loss": 0.2421, + "step": 1101 + }, + { + "epoch": 1.8286885671285484, + "grad_norm": 19.470008850097656, + "learning_rate": 3.509525397419778e-05, + "loss": 0.2919, + "step": 1102 + }, + { + "epoch": 1.8303492656598683, + "grad_norm": 106.8519287109375, + "learning_rate": 3.50097943466562e-05, + "loss": 0.2638, + "step": 1103 + }, + { + "epoch": 1.832009964191188, + "grad_norm": 16.89085578918457, + "learning_rate": 3.492438280836888e-05, + "loss": 0.2447, + "step": 1104 + }, + { + "epoch": 1.8336706627225077, + "grad_norm": 33.42323303222656, + "learning_rate": 3.483901963333988e-05, + "loss": 0.2902, + "step": 1105 + }, + { + "epoch": 1.8353313612538273, + "grad_norm": 12.57965087890625, + "learning_rate": 3.4753705095418085e-05, + "loss": 0.2807, + "step": 1106 + }, + { + "epoch": 1.836992059785147, + "grad_norm": 19.403825759887695, + "learning_rate": 3.46684394682964e-05, + "loss": 0.291, + "step": 1107 + }, + { + "epoch": 1.838652758316467, + "grad_norm": 129.98965454101562, + "learning_rate": 3.458322302551075e-05, + "loss": 0.2853, + "step": 1108 + }, + { + "epoch": 1.8403134568477866, + "grad_norm": 10.069053649902344, + "learning_rate": 3.4498056040439366e-05, + "loss": 0.2727, + "step": 1109 + }, + { + "epoch": 1.8419741553791065, + "grad_norm": 7.090850830078125, + "learning_rate": 3.441293878630174e-05, + "loss": 0.2967, + "step": 1110 + }, + { + "epoch": 1.8436348539104261, + "grad_norm": 18.833589553833008, + "learning_rate": 3.432787153615788e-05, + "loss": 0.2734, + "step": 1111 + }, + { + "epoch": 1.8452955524417458, + "grad_norm": 11.524720191955566, + "learning_rate": 3.424285456290732e-05, + "loss": 0.3023, + "step": 1112 + }, + { + "epoch": 1.8469562509730655, + "grad_norm": 54.83722686767578, + "learning_rate": 3.415788813928833e-05, + "loss": 0.276, + "step": 1113 + }, + { + "epoch": 1.8486169495043852, + "grad_norm": 22.467540740966797, + "learning_rate": 3.4072972537877056e-05, + "loss": 0.2775, + "step": 1114 + }, + { + "epoch": 1.850277648035705, + "grad_norm": 13.082950592041016, + "learning_rate": 3.3988108031086566e-05, + "loss": 0.2971, + "step": 1115 + }, + { + "epoch": 1.8519383465670247, + "grad_norm": 16.470565795898438, + "learning_rate": 3.3903294891165994e-05, + "loss": 0.2861, + "step": 1116 + }, + { + "epoch": 1.8535990450983446, + "grad_norm": 26.090900421142578, + "learning_rate": 3.3818533390199705e-05, + "loss": 0.2499, + "step": 1117 + }, + { + "epoch": 1.8552597436296643, + "grad_norm": 17.931398391723633, + "learning_rate": 3.3733823800106436e-05, + "loss": 0.3034, + "step": 1118 + }, + { + "epoch": 1.856920442160984, + "grad_norm": 23.2215633392334, + "learning_rate": 3.364916639263833e-05, + "loss": 0.2773, + "step": 1119 + }, + { + "epoch": 1.8585811406923036, + "grad_norm": 12.329863548278809, + "learning_rate": 3.35645614393802e-05, + "loss": 0.2539, + "step": 1120 + }, + { + "epoch": 1.8602418392236233, + "grad_norm": 15.17630672454834, + "learning_rate": 3.348000921174851e-05, + "loss": 0.2669, + "step": 1121 + }, + { + "epoch": 1.8619025377549432, + "grad_norm": 42.21493911743164, + "learning_rate": 3.339550998099063e-05, + "loss": 0.2943, + "step": 1122 + }, + { + "epoch": 1.8635632362862629, + "grad_norm": 42.26921844482422, + "learning_rate": 3.331106401818387e-05, + "loss": 0.2955, + "step": 1123 + }, + { + "epoch": 1.8652239348175828, + "grad_norm": 55.217376708984375, + "learning_rate": 3.32266715942347e-05, + "loss": 0.276, + "step": 1124 + }, + { + "epoch": 1.8668846333489024, + "grad_norm": 25.220335006713867, + "learning_rate": 3.3142332979877824e-05, + "loss": 0.2662, + "step": 1125 + }, + { + "epoch": 1.868545331880222, + "grad_norm": 24.88803482055664, + "learning_rate": 3.305804844567528e-05, + "loss": 0.2736, + "step": 1126 + }, + { + "epoch": 1.8702060304115418, + "grad_norm": 13.316408157348633, + "learning_rate": 3.297381826201568e-05, + "loss": 0.2652, + "step": 1127 + }, + { + "epoch": 1.8718667289428614, + "grad_norm": 311.9528503417969, + "learning_rate": 3.28896426991132e-05, + "loss": 0.2906, + "step": 1128 + }, + { + "epoch": 1.8735274274741813, + "grad_norm": 50.26872253417969, + "learning_rate": 3.2805522027006876e-05, + "loss": 0.2808, + "step": 1129 + }, + { + "epoch": 1.8751881260055012, + "grad_norm": 11.797974586486816, + "learning_rate": 3.272145651555959e-05, + "loss": 0.2573, + "step": 1130 + }, + { + "epoch": 1.876848824536821, + "grad_norm": 15.524652481079102, + "learning_rate": 3.263744643445729e-05, + "loss": 0.2772, + "step": 1131 + }, + { + "epoch": 1.8785095230681406, + "grad_norm": 44.341796875, + "learning_rate": 3.25534920532081e-05, + "loss": 0.2793, + "step": 1132 + }, + { + "epoch": 1.8801702215994602, + "grad_norm": 19.346965789794922, + "learning_rate": 3.246959364114143e-05, + "loss": 0.2872, + "step": 1133 + }, + { + "epoch": 1.88183092013078, + "grad_norm": 8.47995662689209, + "learning_rate": 3.2385751467407177e-05, + "loss": 0.2926, + "step": 1134 + }, + { + "epoch": 1.8834916186620998, + "grad_norm": 28.525705337524414, + "learning_rate": 3.2301965800974834e-05, + "loss": 0.2958, + "step": 1135 + }, + { + "epoch": 1.8851523171934195, + "grad_norm": 15.42277717590332, + "learning_rate": 3.2218236910632546e-05, + "loss": 0.2818, + "step": 1136 + }, + { + "epoch": 1.8868130157247394, + "grad_norm": 9.878376007080078, + "learning_rate": 3.213456506498638e-05, + "loss": 0.2911, + "step": 1137 + }, + { + "epoch": 1.888473714256059, + "grad_norm": 26.60796356201172, + "learning_rate": 3.205095053245937e-05, + "loss": 0.2962, + "step": 1138 + }, + { + "epoch": 1.8901344127873787, + "grad_norm": 16.922300338745117, + "learning_rate": 3.1967393581290715e-05, + "loss": 0.2773, + "step": 1139 + }, + { + "epoch": 1.8917951113186984, + "grad_norm": 33.345733642578125, + "learning_rate": 3.188389447953488e-05, + "loss": 0.2679, + "step": 1140 + }, + { + "epoch": 1.893455809850018, + "grad_norm": 11.762744903564453, + "learning_rate": 3.1800453495060725e-05, + "loss": 0.2659, + "step": 1141 + }, + { + "epoch": 1.895116508381338, + "grad_norm": 24.39533042907715, + "learning_rate": 3.171707089555068e-05, + "loss": 0.2699, + "step": 1142 + }, + { + "epoch": 1.8967772069126576, + "grad_norm": 18.04019546508789, + "learning_rate": 3.163374694849988e-05, + "loss": 0.2637, + "step": 1143 + }, + { + "epoch": 1.8984379054439775, + "grad_norm": 26.92572784423828, + "learning_rate": 3.1550481921215304e-05, + "loss": 0.2814, + "step": 1144 + }, + { + "epoch": 1.9000986039752972, + "grad_norm": 17.512197494506836, + "learning_rate": 3.1467276080814896e-05, + "loss": 0.2959, + "step": 1145 + }, + { + "epoch": 1.9017593025066168, + "grad_norm": 18.834636688232422, + "learning_rate": 3.138412969422675e-05, + "loss": 0.274, + "step": 1146 + }, + { + "epoch": 1.9034200010379365, + "grad_norm": 61.84907531738281, + "learning_rate": 3.130104302818821e-05, + "loss": 0.3004, + "step": 1147 + }, + { + "epoch": 1.9050806995692562, + "grad_norm": 14.55224609375, + "learning_rate": 3.1218016349245015e-05, + "loss": 0.2812, + "step": 1148 + }, + { + "epoch": 1.906741398100576, + "grad_norm": 19.832944869995117, + "learning_rate": 3.113504992375052e-05, + "loss": 0.2762, + "step": 1149 + }, + { + "epoch": 1.9084020966318958, + "grad_norm": 23.48835563659668, + "learning_rate": 3.1052144017864746e-05, + "loss": 0.2985, + "step": 1150 + }, + { + "epoch": 1.9100627951632156, + "grad_norm": 47.911258697509766, + "learning_rate": 3.096929889755356e-05, + "loss": 0.2736, + "step": 1151 + }, + { + "epoch": 1.9117234936945353, + "grad_norm": 19.487102508544922, + "learning_rate": 3.088651482858783e-05, + "loss": 0.2765, + "step": 1152 + }, + { + "epoch": 1.913384192225855, + "grad_norm": 27.229175567626953, + "learning_rate": 3.080379207654258e-05, + "loss": 0.2743, + "step": 1153 + }, + { + "epoch": 1.9150448907571747, + "grad_norm": 20.660064697265625, + "learning_rate": 3.072113090679614e-05, + "loss": 0.2727, + "step": 1154 + }, + { + "epoch": 1.9167055892884943, + "grad_norm": 46.127830505371094, + "learning_rate": 3.0638531584529244e-05, + "loss": 0.2731, + "step": 1155 + }, + { + "epoch": 1.9183662878198142, + "grad_norm": 17.219501495361328, + "learning_rate": 3.0555994374724225e-05, + "loss": 0.2772, + "step": 1156 + }, + { + "epoch": 1.920026986351134, + "grad_norm": 66.9420394897461, + "learning_rate": 3.0473519542164198e-05, + "loss": 0.294, + "step": 1157 + }, + { + "epoch": 1.9216876848824538, + "grad_norm": 11.458773612976074, + "learning_rate": 3.0391107351432086e-05, + "loss": 0.2624, + "step": 1158 + }, + { + "epoch": 1.9233483834137735, + "grad_norm": 7.661323070526123, + "learning_rate": 3.030875806690996e-05, + "loss": 0.2524, + "step": 1159 + }, + { + "epoch": 1.9250090819450931, + "grad_norm": 12.583983421325684, + "learning_rate": 3.022647195277799e-05, + "loss": 0.2728, + "step": 1160 + }, + { + "epoch": 1.9266697804764128, + "grad_norm": 12.300309181213379, + "learning_rate": 3.0144249273013768e-05, + "loss": 0.3115, + "step": 1161 + }, + { + "epoch": 1.9283304790077325, + "grad_norm": 32.204551696777344, + "learning_rate": 3.0062090291391325e-05, + "loss": 0.2927, + "step": 1162 + }, + { + "epoch": 1.9299911775390524, + "grad_norm": 12.028011322021484, + "learning_rate": 2.997999527148036e-05, + "loss": 0.2775, + "step": 1163 + }, + { + "epoch": 1.9316518760703723, + "grad_norm": 11.328278541564941, + "learning_rate": 2.9897964476645425e-05, + "loss": 0.2556, + "step": 1164 + }, + { + "epoch": 1.933312574601692, + "grad_norm": 15.071605682373047, + "learning_rate": 2.981599817004498e-05, + "loss": 0.273, + "step": 1165 + }, + { + "epoch": 1.9349732731330116, + "grad_norm": 9.165623664855957, + "learning_rate": 2.9734096614630637e-05, + "loss": 0.2808, + "step": 1166 + }, + { + "epoch": 1.9366339716643313, + "grad_norm": 44.23814010620117, + "learning_rate": 2.9652260073146243e-05, + "loss": 0.2815, + "step": 1167 + }, + { + "epoch": 1.938294670195651, + "grad_norm": 10.883672714233398, + "learning_rate": 2.957048880812714e-05, + "loss": 0.2559, + "step": 1168 + }, + { + "epoch": 1.9399553687269708, + "grad_norm": 33.23820495605469, + "learning_rate": 2.94887830818992e-05, + "loss": 0.2642, + "step": 1169 + }, + { + "epoch": 1.9416160672582905, + "grad_norm": 16.232196807861328, + "learning_rate": 2.94071431565781e-05, + "loss": 0.2905, + "step": 1170 + }, + { + "epoch": 1.9432767657896104, + "grad_norm": 9.370057106018066, + "learning_rate": 2.9325569294068366e-05, + "loss": 0.2539, + "step": 1171 + }, + { + "epoch": 1.94493746432093, + "grad_norm": 26.445266723632812, + "learning_rate": 2.924406175606261e-05, + "loss": 0.2745, + "step": 1172 + }, + { + "epoch": 1.9465981628522497, + "grad_norm": 32.306819915771484, + "learning_rate": 2.9162620804040715e-05, + "loss": 0.2928, + "step": 1173 + }, + { + "epoch": 1.9482588613835694, + "grad_norm": 20.964317321777344, + "learning_rate": 2.9081246699268905e-05, + "loss": 0.2558, + "step": 1174 + }, + { + "epoch": 1.949919559914889, + "grad_norm": 15.256046295166016, + "learning_rate": 2.8999939702798974e-05, + "loss": 0.2459, + "step": 1175 + }, + { + "epoch": 1.951580258446209, + "grad_norm": 24.95514488220215, + "learning_rate": 2.891870007546742e-05, + "loss": 0.2895, + "step": 1176 + }, + { + "epoch": 1.9532409569775286, + "grad_norm": 21.605392456054688, + "learning_rate": 2.8837528077894603e-05, + "loss": 0.2769, + "step": 1177 + }, + { + "epoch": 1.9549016555088485, + "grad_norm": 22.20134735107422, + "learning_rate": 2.8756423970483965e-05, + "loss": 0.2553, + "step": 1178 + }, + { + "epoch": 1.9565623540401682, + "grad_norm": 70.49386596679688, + "learning_rate": 2.867538801342115e-05, + "loss": 0.2755, + "step": 1179 + }, + { + "epoch": 1.9582230525714879, + "grad_norm": 11.37830638885498, + "learning_rate": 2.8594420466673134e-05, + "loss": 0.2873, + "step": 1180 + }, + { + "epoch": 1.9598837511028075, + "grad_norm": 14.251688957214355, + "learning_rate": 2.851352158998746e-05, + "loss": 0.2459, + "step": 1181 + }, + { + "epoch": 1.9615444496341272, + "grad_norm": 40.83670425415039, + "learning_rate": 2.843269164289133e-05, + "loss": 0.281, + "step": 1182 + }, + { + "epoch": 1.9632051481654471, + "grad_norm": 16.484617233276367, + "learning_rate": 2.835193088469089e-05, + "loss": 0.2656, + "step": 1183 + }, + { + "epoch": 1.9648658466967668, + "grad_norm": 16.4002628326416, + "learning_rate": 2.827123957447029e-05, + "loss": 0.2699, + "step": 1184 + }, + { + "epoch": 1.9665265452280867, + "grad_norm": 22.61533546447754, + "learning_rate": 2.8190617971090837e-05, + "loss": 0.2922, + "step": 1185 + }, + { + "epoch": 1.9681872437594063, + "grad_norm": 265.6081237792969, + "learning_rate": 2.8110066333190317e-05, + "loss": 0.2641, + "step": 1186 + }, + { + "epoch": 1.969847942290726, + "grad_norm": 23.90536117553711, + "learning_rate": 2.8029584919181963e-05, + "loss": 0.2835, + "step": 1187 + }, + { + "epoch": 1.9715086408220457, + "grad_norm": 18.159608840942383, + "learning_rate": 2.794917398725382e-05, + "loss": 0.275, + "step": 1188 + }, + { + "epoch": 1.9731693393533654, + "grad_norm": 32.7334098815918, + "learning_rate": 2.7868833795367743e-05, + "loss": 0.2778, + "step": 1189 + }, + { + "epoch": 1.9748300378846853, + "grad_norm": 43.110103607177734, + "learning_rate": 2.7788564601258703e-05, + "loss": 0.3004, + "step": 1190 + }, + { + "epoch": 1.9764907364160051, + "grad_norm": 24.01694107055664, + "learning_rate": 2.7708366662433864e-05, + "loss": 0.274, + "step": 1191 + }, + { + "epoch": 1.9781514349473248, + "grad_norm": 124.2526626586914, + "learning_rate": 2.762824023617182e-05, + "loss": 0.2676, + "step": 1192 + }, + { + "epoch": 1.9798121334786445, + "grad_norm": 34.311153411865234, + "learning_rate": 2.754818557952178e-05, + "loss": 0.2754, + "step": 1193 + }, + { + "epoch": 1.9814728320099642, + "grad_norm": 6.2869954109191895, + "learning_rate": 2.7468202949302653e-05, + "loss": 0.2674, + "step": 1194 + }, + { + "epoch": 1.9831335305412838, + "grad_norm": 86.01805877685547, + "learning_rate": 2.738829260210235e-05, + "loss": 0.2757, + "step": 1195 + }, + { + "epoch": 1.9847942290726035, + "grad_norm": 20.675840377807617, + "learning_rate": 2.7308454794276837e-05, + "loss": 0.2678, + "step": 1196 + }, + { + "epoch": 1.9864549276039234, + "grad_norm": 427.17095947265625, + "learning_rate": 2.722868978194939e-05, + "loss": 0.2867, + "step": 1197 + }, + { + "epoch": 1.9881156261352433, + "grad_norm": 52.369598388671875, + "learning_rate": 2.714899782100978e-05, + "loss": 0.2867, + "step": 1198 + }, + { + "epoch": 1.989776324666563, + "grad_norm": 118.34003448486328, + "learning_rate": 2.7069379167113395e-05, + "loss": 0.2715, + "step": 1199 + }, + { + "epoch": 1.9914370231978826, + "grad_norm": 96.16859436035156, + "learning_rate": 2.6989834075680452e-05, + "loss": 0.2805, + "step": 1200 + }, + { + "epoch": 1.9930977217292023, + "grad_norm": 16.232057571411133, + "learning_rate": 2.6910362801895173e-05, + "loss": 0.2881, + "step": 1201 + }, + { + "epoch": 1.994758420260522, + "grad_norm": 14.908853530883789, + "learning_rate": 2.6830965600704994e-05, + "loss": 0.2909, + "step": 1202 + }, + { + "epoch": 1.9964191187918419, + "grad_norm": 20.73726463317871, + "learning_rate": 2.6751642726819725e-05, + "loss": 0.2582, + "step": 1203 + }, + { + "epoch": 1.9980798173231615, + "grad_norm": 12.7951078414917, + "learning_rate": 2.66723944347107e-05, + "loss": 0.2711, + "step": 1204 + }, + { + "epoch": 1.9997405158544814, + "grad_norm": 29.738046646118164, + "learning_rate": 2.659322097861001e-05, + "loss": 0.288, + "step": 1205 + }, + { + "epoch": 2.0, + "grad_norm": 61.80087661743164, + "learning_rate": 2.651412261250965e-05, + "loss": 0.3155, + "step": 1206 + }, + { + "epoch": 2.0016606985313197, + "grad_norm": 17.413053512573242, + "learning_rate": 2.643509959016074e-05, + "loss": 0.2682, + "step": 1207 + }, + { + "epoch": 2.0033213970626393, + "grad_norm": 39.304439544677734, + "learning_rate": 2.635615216507271e-05, + "loss": 0.2556, + "step": 1208 + }, + { + "epoch": 2.004982095593959, + "grad_norm": 13.959939956665039, + "learning_rate": 2.627728059051242e-05, + "loss": 0.2843, + "step": 1209 + }, + { + "epoch": 2.006642794125279, + "grad_norm": 23.269983291625977, + "learning_rate": 2.619848511950348e-05, + "loss": 0.2821, + "step": 1210 + }, + { + "epoch": 2.008303492656599, + "grad_norm": 23.32681655883789, + "learning_rate": 2.6119766004825276e-05, + "loss": 0.2797, + "step": 1211 + }, + { + "epoch": 2.0099641911879185, + "grad_norm": 25.73476219177246, + "learning_rate": 2.6041123499012242e-05, + "loss": 0.2667, + "step": 1212 + }, + { + "epoch": 2.011624889719238, + "grad_norm": 9.890053749084473, + "learning_rate": 2.596255785435313e-05, + "loss": 0.2546, + "step": 1213 + }, + { + "epoch": 2.013285588250558, + "grad_norm": 44.863746643066406, + "learning_rate": 2.5884069322890047e-05, + "loss": 0.2807, + "step": 1214 + }, + { + "epoch": 2.0149462867818775, + "grad_norm": 31.812408447265625, + "learning_rate": 2.5805658156417716e-05, + "loss": 0.2619, + "step": 1215 + }, + { + "epoch": 2.016606985313197, + "grad_norm": 13.861759185791016, + "learning_rate": 2.5727324606482684e-05, + "loss": 0.2551, + "step": 1216 + }, + { + "epoch": 2.0182676838445173, + "grad_norm": 292.68963623046875, + "learning_rate": 2.5649068924382513e-05, + "loss": 0.2492, + "step": 1217 + }, + { + "epoch": 2.019928382375837, + "grad_norm": 15.022615432739258, + "learning_rate": 2.5570891361164974e-05, + "loss": 0.2935, + "step": 1218 + }, + { + "epoch": 2.0215890809071566, + "grad_norm": 33.31991958618164, + "learning_rate": 2.549279216762719e-05, + "loss": 0.2292, + "step": 1219 + }, + { + "epoch": 2.0232497794384763, + "grad_norm": 7.302118301391602, + "learning_rate": 2.541477159431489e-05, + "loss": 0.2773, + "step": 1220 + }, + { + "epoch": 2.024910477969796, + "grad_norm": 10.284438133239746, + "learning_rate": 2.5336829891521574e-05, + "loss": 0.2467, + "step": 1221 + }, + { + "epoch": 2.0265711765011156, + "grad_norm": 18.24944496154785, + "learning_rate": 2.5258967309287728e-05, + "loss": 0.2563, + "step": 1222 + }, + { + "epoch": 2.0282318750324357, + "grad_norm": 230.6899871826172, + "learning_rate": 2.5181184097400046e-05, + "loss": 0.2711, + "step": 1223 + }, + { + "epoch": 2.0298925735637554, + "grad_norm": 17.670936584472656, + "learning_rate": 2.5103480505390543e-05, + "loss": 0.2607, + "step": 1224 + }, + { + "epoch": 2.031553272095075, + "grad_norm": 10.486580848693848, + "learning_rate": 2.502585678253587e-05, + "loss": 0.2923, + "step": 1225 + }, + { + "epoch": 2.0332139706263948, + "grad_norm": 15.968626022338867, + "learning_rate": 2.4948313177856402e-05, + "loss": 0.2629, + "step": 1226 + }, + { + "epoch": 2.0348746691577144, + "grad_norm": 11.306639671325684, + "learning_rate": 2.4870849940115498e-05, + "loss": 0.2703, + "step": 1227 + }, + { + "epoch": 2.036535367689034, + "grad_norm": 197.81805419921875, + "learning_rate": 2.479346731781874e-05, + "loss": 0.3032, + "step": 1228 + }, + { + "epoch": 2.0381960662203538, + "grad_norm": 17.972484588623047, + "learning_rate": 2.4716165559213056e-05, + "loss": 0.275, + "step": 1229 + }, + { + "epoch": 2.039856764751674, + "grad_norm": 18.936050415039062, + "learning_rate": 2.463894491228596e-05, + "loss": 0.2802, + "step": 1230 + }, + { + "epoch": 2.0415174632829935, + "grad_norm": 16.643709182739258, + "learning_rate": 2.456180562476474e-05, + "loss": 0.2715, + "step": 1231 + }, + { + "epoch": 2.043178161814313, + "grad_norm": 57.196083068847656, + "learning_rate": 2.4484747944115742e-05, + "loss": 0.2365, + "step": 1232 + }, + { + "epoch": 2.044838860345633, + "grad_norm": 16.08137321472168, + "learning_rate": 2.4407772117543437e-05, + "loss": 0.265, + "step": 1233 + }, + { + "epoch": 2.0464995588769526, + "grad_norm": 26.36579132080078, + "learning_rate": 2.4330878391989776e-05, + "loss": 0.2638, + "step": 1234 + }, + { + "epoch": 2.0481602574082722, + "grad_norm": 18.793617248535156, + "learning_rate": 2.425406701413327e-05, + "loss": 0.2389, + "step": 1235 + }, + { + "epoch": 2.049820955939592, + "grad_norm": 20.406641006469727, + "learning_rate": 2.417733823038828e-05, + "loss": 0.2856, + "step": 1236 + }, + { + "epoch": 2.051481654470912, + "grad_norm": 95.74354553222656, + "learning_rate": 2.4100692286904173e-05, + "loss": 0.285, + "step": 1237 + }, + { + "epoch": 2.0531423530022317, + "grad_norm": 98.55819702148438, + "learning_rate": 2.4024129429564623e-05, + "loss": 0.25, + "step": 1238 + }, + { + "epoch": 2.0548030515335514, + "grad_norm": 21.065919876098633, + "learning_rate": 2.394764990398671e-05, + "loss": 0.2762, + "step": 1239 + }, + { + "epoch": 2.056463750064871, + "grad_norm": 15.908268928527832, + "learning_rate": 2.3871253955520162e-05, + "loss": 0.2653, + "step": 1240 + }, + { + "epoch": 2.0581244485961907, + "grad_norm": 15.789944648742676, + "learning_rate": 2.3794941829246648e-05, + "loss": 0.2653, + "step": 1241 + }, + { + "epoch": 2.0597851471275104, + "grad_norm": 10.54181957244873, + "learning_rate": 2.3718713769978922e-05, + "loss": 0.2351, + "step": 1242 + }, + { + "epoch": 2.06144584565883, + "grad_norm": 12.108061790466309, + "learning_rate": 2.3642570022260003e-05, + "loss": 0.267, + "step": 1243 + }, + { + "epoch": 2.06310654419015, + "grad_norm": 23.205421447753906, + "learning_rate": 2.356651083036246e-05, + "loss": 0.2506, + "step": 1244 + }, + { + "epoch": 2.06476724272147, + "grad_norm": 26.426939010620117, + "learning_rate": 2.3490536438287607e-05, + "loss": 0.2778, + "step": 1245 + }, + { + "epoch": 2.0664279412527895, + "grad_norm": 51.35211181640625, + "learning_rate": 2.3414647089764696e-05, + "loss": 0.2646, + "step": 1246 + }, + { + "epoch": 2.068088639784109, + "grad_norm": 15.948187828063965, + "learning_rate": 2.33388430282502e-05, + "loss": 0.2632, + "step": 1247 + }, + { + "epoch": 2.069749338315429, + "grad_norm": 75.65266418457031, + "learning_rate": 2.3263124496926935e-05, + "loss": 0.2864, + "step": 1248 + }, + { + "epoch": 2.0714100368467485, + "grad_norm": 68.45854187011719, + "learning_rate": 2.3187491738703383e-05, + "loss": 0.2787, + "step": 1249 + }, + { + "epoch": 2.073070735378068, + "grad_norm": 12.738755226135254, + "learning_rate": 2.311194499621282e-05, + "loss": 0.2579, + "step": 1250 + }, + { + "epoch": 2.0747314339093883, + "grad_norm": 16.639389038085938, + "learning_rate": 2.3036484511812582e-05, + "loss": 0.249, + "step": 1251 + }, + { + "epoch": 2.076392132440708, + "grad_norm": 11.296469688415527, + "learning_rate": 2.2961110527583328e-05, + "loss": 0.2693, + "step": 1252 + }, + { + "epoch": 2.0780528309720276, + "grad_norm": 17.459983825683594, + "learning_rate": 2.2885823285328174e-05, + "loss": 0.2676, + "step": 1253 + }, + { + "epoch": 2.0797135295033473, + "grad_norm": 15.04466724395752, + "learning_rate": 2.2810623026571987e-05, + "loss": 0.2564, + "step": 1254 + }, + { + "epoch": 2.081374228034667, + "grad_norm": 1093.754638671875, + "learning_rate": 2.2735509992560554e-05, + "loss": 0.2612, + "step": 1255 + }, + { + "epoch": 2.0830349265659867, + "grad_norm": 11.760372161865234, + "learning_rate": 2.266048442425989e-05, + "loss": 0.2707, + "step": 1256 + }, + { + "epoch": 2.0846956250973063, + "grad_norm": 10.40364933013916, + "learning_rate": 2.2585546562355414e-05, + "loss": 0.2789, + "step": 1257 + }, + { + "epoch": 2.0863563236286264, + "grad_norm": 61.76397705078125, + "learning_rate": 2.251069664725114e-05, + "loss": 0.288, + "step": 1258 + }, + { + "epoch": 2.088017022159946, + "grad_norm": 19.892759323120117, + "learning_rate": 2.2435934919068962e-05, + "loss": 0.2832, + "step": 1259 + }, + { + "epoch": 2.089677720691266, + "grad_norm": 15.674310684204102, + "learning_rate": 2.2361261617647872e-05, + "loss": 0.276, + "step": 1260 + }, + { + "epoch": 2.0913384192225855, + "grad_norm": 37.02360534667969, + "learning_rate": 2.2286676982543154e-05, + "loss": 0.2461, + "step": 1261 + }, + { + "epoch": 2.092999117753905, + "grad_norm": 11.663684844970703, + "learning_rate": 2.221218125302572e-05, + "loss": 0.2613, + "step": 1262 + }, + { + "epoch": 2.094659816285225, + "grad_norm": 11.333456039428711, + "learning_rate": 2.213777466808118e-05, + "loss": 0.2771, + "step": 1263 + }, + { + "epoch": 2.096320514816545, + "grad_norm": 16.587526321411133, + "learning_rate": 2.206345746640925e-05, + "loss": 0.2816, + "step": 1264 + }, + { + "epoch": 2.0979812133478646, + "grad_norm": 78.1726303100586, + "learning_rate": 2.198922988642283e-05, + "loss": 0.2641, + "step": 1265 + }, + { + "epoch": 2.0996419118791843, + "grad_norm": 29.875125885009766, + "learning_rate": 2.191509216624732e-05, + "loss": 0.2893, + "step": 1266 + }, + { + "epoch": 2.101302610410504, + "grad_norm": 29.735599517822266, + "learning_rate": 2.1841044543719902e-05, + "loss": 0.2781, + "step": 1267 + }, + { + "epoch": 2.1029633089418236, + "grad_norm": 20.6688175201416, + "learning_rate": 2.1767087256388662e-05, + "loss": 0.2827, + "step": 1268 + }, + { + "epoch": 2.1046240074731433, + "grad_norm": 14.732158660888672, + "learning_rate": 2.1693220541511906e-05, + "loss": 0.2738, + "step": 1269 + }, + { + "epoch": 2.106284706004463, + "grad_norm": 28.433513641357422, + "learning_rate": 2.161944463605736e-05, + "loss": 0.2783, + "step": 1270 + }, + { + "epoch": 2.107945404535783, + "grad_norm": 35.21087646484375, + "learning_rate": 2.1545759776701485e-05, + "loss": 0.2814, + "step": 1271 + }, + { + "epoch": 2.1096061030671027, + "grad_norm": 26.468786239624023, + "learning_rate": 2.1472166199828592e-05, + "loss": 0.2842, + "step": 1272 + }, + { + "epoch": 2.1112668015984224, + "grad_norm": 30.4160213470459, + "learning_rate": 2.1398664141530217e-05, + "loss": 0.2754, + "step": 1273 + }, + { + "epoch": 2.112927500129742, + "grad_norm": 13.977611541748047, + "learning_rate": 2.1325253837604253e-05, + "loss": 0.2686, + "step": 1274 + }, + { + "epoch": 2.1145881986610617, + "grad_norm": 90.86012268066406, + "learning_rate": 2.1251935523554268e-05, + "loss": 0.257, + "step": 1275 + }, + { + "epoch": 2.1162488971923814, + "grad_norm": 13.951212882995605, + "learning_rate": 2.1178709434588677e-05, + "loss": 0.257, + "step": 1276 + }, + { + "epoch": 2.117909595723701, + "grad_norm": 8.633283615112305, + "learning_rate": 2.1105575805620126e-05, + "loss": 0.242, + "step": 1277 + }, + { + "epoch": 2.119570294255021, + "grad_norm": 13.20469856262207, + "learning_rate": 2.1032534871264553e-05, + "loss": 0.263, + "step": 1278 + }, + { + "epoch": 2.121230992786341, + "grad_norm": 24.980571746826172, + "learning_rate": 2.095958686584055e-05, + "loss": 0.2591, + "step": 1279 + }, + { + "epoch": 2.1228916913176605, + "grad_norm": 10.948518753051758, + "learning_rate": 2.0886732023368638e-05, + "loss": 0.2608, + "step": 1280 + }, + { + "epoch": 2.12455238984898, + "grad_norm": 15.760869026184082, + "learning_rate": 2.081397057757039e-05, + "loss": 0.2825, + "step": 1281 + }, + { + "epoch": 2.1262130883803, + "grad_norm": 96.2586669921875, + "learning_rate": 2.0741302761867836e-05, + "loss": 0.262, + "step": 1282 + }, + { + "epoch": 2.1278737869116195, + "grad_norm": 28.27165985107422, + "learning_rate": 2.0668728809382588e-05, + "loss": 0.2769, + "step": 1283 + }, + { + "epoch": 2.1295344854429397, + "grad_norm": 6.3706183433532715, + "learning_rate": 2.0596248952935138e-05, + "loss": 0.247, + "step": 1284 + }, + { + "epoch": 2.1311951839742593, + "grad_norm": 9.746777534484863, + "learning_rate": 2.0523863425044132e-05, + "loss": 0.2597, + "step": 1285 + }, + { + "epoch": 2.132855882505579, + "grad_norm": 24.117761611938477, + "learning_rate": 2.0451572457925587e-05, + "loss": 0.2852, + "step": 1286 + }, + { + "epoch": 2.1345165810368987, + "grad_norm": 8.490272521972656, + "learning_rate": 2.0379376283492198e-05, + "loss": 0.2576, + "step": 1287 + }, + { + "epoch": 2.1361772795682183, + "grad_norm": 16.885372161865234, + "learning_rate": 2.0307275133352543e-05, + "loss": 0.2844, + "step": 1288 + }, + { + "epoch": 2.137837978099538, + "grad_norm": 90.52965545654297, + "learning_rate": 2.0235269238810356e-05, + "loss": 0.2636, + "step": 1289 + }, + { + "epoch": 2.1394986766308577, + "grad_norm": 25.42213249206543, + "learning_rate": 2.0163358830863782e-05, + "loss": 0.2693, + "step": 1290 + }, + { + "epoch": 2.141159375162178, + "grad_norm": 16.58472442626953, + "learning_rate": 2.009154414020462e-05, + "loss": 0.2566, + "step": 1291 + }, + { + "epoch": 2.1428200736934975, + "grad_norm": 14.825553894042969, + "learning_rate": 2.0019825397217672e-05, + "loss": 0.2668, + "step": 1292 + }, + { + "epoch": 2.144480772224817, + "grad_norm": 14.335012435913086, + "learning_rate": 1.994820283197988e-05, + "loss": 0.2972, + "step": 1293 + }, + { + "epoch": 2.146141470756137, + "grad_norm": 37.96998596191406, + "learning_rate": 1.9876676674259625e-05, + "loss": 0.2668, + "step": 1294 + }, + { + "epoch": 2.1478021692874565, + "grad_norm": 47.23372268676758, + "learning_rate": 1.9805247153516082e-05, + "loss": 0.2696, + "step": 1295 + }, + { + "epoch": 2.149462867818776, + "grad_norm": 26.076147079467773, + "learning_rate": 1.9733914498898336e-05, + "loss": 0.2547, + "step": 1296 + }, + { + "epoch": 2.151123566350096, + "grad_norm": 18.419475555419922, + "learning_rate": 1.9662678939244776e-05, + "loss": 0.2846, + "step": 1297 + }, + { + "epoch": 2.152784264881416, + "grad_norm": 63.934608459472656, + "learning_rate": 1.9591540703082283e-05, + "loss": 0.257, + "step": 1298 + }, + { + "epoch": 2.1544449634127356, + "grad_norm": 13.389415740966797, + "learning_rate": 1.9520500018625503e-05, + "loss": 0.2712, + "step": 1299 + }, + { + "epoch": 2.1561056619440553, + "grad_norm": 19.486148834228516, + "learning_rate": 1.9449557113776156e-05, + "loss": 0.2719, + "step": 1300 + }, + { + "epoch": 2.157766360475375, + "grad_norm": 32.610694885253906, + "learning_rate": 1.9378712216122256e-05, + "loss": 0.2764, + "step": 1301 + }, + { + "epoch": 2.1594270590066946, + "grad_norm": 24.50152015686035, + "learning_rate": 1.930796555293744e-05, + "loss": 0.259, + "step": 1302 + }, + { + "epoch": 2.1610877575380143, + "grad_norm": 10.53490161895752, + "learning_rate": 1.92373173511802e-05, + "loss": 0.2987, + "step": 1303 + }, + { + "epoch": 2.162748456069334, + "grad_norm": 126.9980239868164, + "learning_rate": 1.9166767837493138e-05, + "loss": 0.2643, + "step": 1304 + }, + { + "epoch": 2.164409154600654, + "grad_norm": 10.409600257873535, + "learning_rate": 1.9096317238202238e-05, + "loss": 0.2643, + "step": 1305 + }, + { + "epoch": 2.1660698531319738, + "grad_norm": 21.069969177246094, + "learning_rate": 1.9025965779316236e-05, + "loss": 0.2721, + "step": 1306 + }, + { + "epoch": 2.1677305516632934, + "grad_norm": 21.844518661499023, + "learning_rate": 1.895571368652576e-05, + "loss": 0.2761, + "step": 1307 + }, + { + "epoch": 2.169391250194613, + "grad_norm": 8.465849876403809, + "learning_rate": 1.8885561185202684e-05, + "loss": 0.2793, + "step": 1308 + }, + { + "epoch": 2.1710519487259328, + "grad_norm": 33.49427795410156, + "learning_rate": 1.8815508500399365e-05, + "loss": 0.2808, + "step": 1309 + }, + { + "epoch": 2.1727126472572524, + "grad_norm": 101.14215087890625, + "learning_rate": 1.8745555856848012e-05, + "loss": 0.2564, + "step": 1310 + }, + { + "epoch": 2.174373345788572, + "grad_norm": 105.88152313232422, + "learning_rate": 1.8675703478959818e-05, + "loss": 0.2803, + "step": 1311 + }, + { + "epoch": 2.1760340443198922, + "grad_norm": 13.287734031677246, + "learning_rate": 1.8605951590824383e-05, + "loss": 0.2848, + "step": 1312 + }, + { + "epoch": 2.177694742851212, + "grad_norm": 137.82272338867188, + "learning_rate": 1.853630041620889e-05, + "loss": 0.2922, + "step": 1313 + }, + { + "epoch": 2.1793554413825316, + "grad_norm": 14.095038414001465, + "learning_rate": 1.846675017855744e-05, + "loss": 0.2746, + "step": 1314 + }, + { + "epoch": 2.1810161399138512, + "grad_norm": 55.31532287597656, + "learning_rate": 1.8397301100990305e-05, + "loss": 0.2812, + "step": 1315 + }, + { + "epoch": 2.182676838445171, + "grad_norm": 19.303455352783203, + "learning_rate": 1.832795340630329e-05, + "loss": 0.2916, + "step": 1316 + }, + { + "epoch": 2.1843375369764906, + "grad_norm": 44.89860153198242, + "learning_rate": 1.82587073169669e-05, + "loss": 0.2691, + "step": 1317 + }, + { + "epoch": 2.1859982355078103, + "grad_norm": 54.613319396972656, + "learning_rate": 1.8189563055125692e-05, + "loss": 0.2754, + "step": 1318 + }, + { + "epoch": 2.1876589340391304, + "grad_norm": 15.545280456542969, + "learning_rate": 1.8120520842597593e-05, + "loss": 0.2658, + "step": 1319 + }, + { + "epoch": 2.18931963257045, + "grad_norm": 15.21256160736084, + "learning_rate": 1.805158090087311e-05, + "loss": 0.2648, + "step": 1320 + }, + { + "epoch": 2.1909803311017697, + "grad_norm": 25.193960189819336, + "learning_rate": 1.79827434511147e-05, + "loss": 0.2706, + "step": 1321 + }, + { + "epoch": 2.1926410296330894, + "grad_norm": 64.63079833984375, + "learning_rate": 1.7914008714155982e-05, + "loss": 0.2763, + "step": 1322 + }, + { + "epoch": 2.194301728164409, + "grad_norm": 10.006000518798828, + "learning_rate": 1.7845376910501087e-05, + "loss": 0.3045, + "step": 1323 + }, + { + "epoch": 2.1959624266957287, + "grad_norm": 18.25226402282715, + "learning_rate": 1.7776848260323935e-05, + "loss": 0.2674, + "step": 1324 + }, + { + "epoch": 2.197623125227049, + "grad_norm": 26.94184684753418, + "learning_rate": 1.77084229834675e-05, + "loss": 0.2675, + "step": 1325 + }, + { + "epoch": 2.1992838237583685, + "grad_norm": 9.828961372375488, + "learning_rate": 1.7640101299443162e-05, + "loss": 0.284, + "step": 1326 + }, + { + "epoch": 2.200944522289688, + "grad_norm": 12.98893928527832, + "learning_rate": 1.757188342742998e-05, + "loss": 0.2755, + "step": 1327 + }, + { + "epoch": 2.202605220821008, + "grad_norm": 21.647016525268555, + "learning_rate": 1.7503769586273934e-05, + "loss": 0.2891, + "step": 1328 + }, + { + "epoch": 2.2042659193523275, + "grad_norm": 14.959050178527832, + "learning_rate": 1.7435759994487293e-05, + "loss": 0.266, + "step": 1329 + }, + { + "epoch": 2.205926617883647, + "grad_norm": 9.576624870300293, + "learning_rate": 1.7367854870247875e-05, + "loss": 0.2411, + "step": 1330 + }, + { + "epoch": 2.207587316414967, + "grad_norm": 9.505250930786133, + "learning_rate": 1.7300054431398398e-05, + "loss": 0.2773, + "step": 1331 + }, + { + "epoch": 2.209248014946287, + "grad_norm": 29.796367645263672, + "learning_rate": 1.723235889544572e-05, + "loss": 0.2563, + "step": 1332 + }, + { + "epoch": 2.2109087134776066, + "grad_norm": 46.41476058959961, + "learning_rate": 1.716476847956014e-05, + "loss": 0.2746, + "step": 1333 + }, + { + "epoch": 2.2125694120089263, + "grad_norm": 27.295066833496094, + "learning_rate": 1.7097283400574783e-05, + "loss": 0.2789, + "step": 1334 + }, + { + "epoch": 2.214230110540246, + "grad_norm": 84.90970611572266, + "learning_rate": 1.7029903874984786e-05, + "loss": 0.2652, + "step": 1335 + }, + { + "epoch": 2.2158908090715657, + "grad_norm": 28.05855369567871, + "learning_rate": 1.6962630118946744e-05, + "loss": 0.2697, + "step": 1336 + }, + { + "epoch": 2.2175515076028853, + "grad_norm": 40.68085861206055, + "learning_rate": 1.6895462348277868e-05, + "loss": 0.2717, + "step": 1337 + }, + { + "epoch": 2.219212206134205, + "grad_norm": 10.003382682800293, + "learning_rate": 1.6828400778455388e-05, + "loss": 0.2604, + "step": 1338 + }, + { + "epoch": 2.220872904665525, + "grad_norm": 8.966263771057129, + "learning_rate": 1.676144562461584e-05, + "loss": 0.2632, + "step": 1339 + }, + { + "epoch": 2.222533603196845, + "grad_norm": 18.03782844543457, + "learning_rate": 1.6694597101554354e-05, + "loss": 0.2801, + "step": 1340 + }, + { + "epoch": 2.2241943017281645, + "grad_norm": 230.7095489501953, + "learning_rate": 1.6627855423724014e-05, + "loss": 0.2524, + "step": 1341 + }, + { + "epoch": 2.225855000259484, + "grad_norm": 12.939144134521484, + "learning_rate": 1.6561220805235146e-05, + "loss": 0.2531, + "step": 1342 + }, + { + "epoch": 2.227515698790804, + "grad_norm": 25.774147033691406, + "learning_rate": 1.6494693459854588e-05, + "loss": 0.2394, + "step": 1343 + }, + { + "epoch": 2.2291763973221235, + "grad_norm": 16.560794830322266, + "learning_rate": 1.642827360100505e-05, + "loss": 0.273, + "step": 1344 + }, + { + "epoch": 2.2308370958534436, + "grad_norm": 12.314764976501465, + "learning_rate": 1.6361961441764423e-05, + "loss": 0.2734, + "step": 1345 + }, + { + "epoch": 2.2324977943847633, + "grad_norm": 25.500205993652344, + "learning_rate": 1.629575719486512e-05, + "loss": 0.2782, + "step": 1346 + }, + { + "epoch": 2.234158492916083, + "grad_norm": 11.771465301513672, + "learning_rate": 1.622966107269333e-05, + "loss": 0.2738, + "step": 1347 + }, + { + "epoch": 2.2358191914474026, + "grad_norm": 7.9409356117248535, + "learning_rate": 1.6163673287288377e-05, + "loss": 0.2705, + "step": 1348 + }, + { + "epoch": 2.2374798899787223, + "grad_norm": 9.508504867553711, + "learning_rate": 1.6097794050342076e-05, + "loss": 0.2783, + "step": 1349 + }, + { + "epoch": 2.239140588510042, + "grad_norm": 13.581822395324707, + "learning_rate": 1.6032023573197947e-05, + "loss": 0.273, + "step": 1350 + }, + { + "epoch": 2.2408012870413616, + "grad_norm": 9.67588996887207, + "learning_rate": 1.5966362066850672e-05, + "loss": 0.28, + "step": 1351 + }, + { + "epoch": 2.2424619855726817, + "grad_norm": 23.675989151000977, + "learning_rate": 1.590080974194531e-05, + "loss": 0.2776, + "step": 1352 + }, + { + "epoch": 2.2441226841040014, + "grad_norm": 10.242881774902344, + "learning_rate": 1.5835366808776658e-05, + "loss": 0.2831, + "step": 1353 + }, + { + "epoch": 2.245783382635321, + "grad_norm": 11.923916816711426, + "learning_rate": 1.57700334772886e-05, + "loss": 0.2593, + "step": 1354 + }, + { + "epoch": 2.2474440811666407, + "grad_norm": 39.941890716552734, + "learning_rate": 1.570480995707338e-05, + "loss": 0.2551, + "step": 1355 + }, + { + "epoch": 2.2491047796979604, + "grad_norm": 11.768372535705566, + "learning_rate": 1.5639696457371023e-05, + "loss": 0.2768, + "step": 1356 + }, + { + "epoch": 2.25076547822928, + "grad_norm": 12.35759449005127, + "learning_rate": 1.5574693187068534e-05, + "loss": 0.292, + "step": 1357 + }, + { + "epoch": 2.2524261767605998, + "grad_norm": 16.811208724975586, + "learning_rate": 1.5509800354699355e-05, + "loss": 0.2617, + "step": 1358 + }, + { + "epoch": 2.25408687529192, + "grad_norm": 34.322750091552734, + "learning_rate": 1.5445018168442603e-05, + "loss": 0.2736, + "step": 1359 + }, + { + "epoch": 2.2557475738232395, + "grad_norm": 10.676634788513184, + "learning_rate": 1.538034683612243e-05, + "loss": 0.2603, + "step": 1360 + }, + { + "epoch": 2.257408272354559, + "grad_norm": 10.121386528015137, + "learning_rate": 1.53157865652074e-05, + "loss": 0.2646, + "step": 1361 + }, + { + "epoch": 2.259068970885879, + "grad_norm": 18.305124282836914, + "learning_rate": 1.525133756280977e-05, + "loss": 0.2458, + "step": 1362 + }, + { + "epoch": 2.2607296694171986, + "grad_norm": 104.64045715332031, + "learning_rate": 1.518700003568483e-05, + "loss": 0.275, + "step": 1363 + }, + { + "epoch": 2.262390367948518, + "grad_norm": 22.398454666137695, + "learning_rate": 1.5122774190230243e-05, + "loss": 0.2627, + "step": 1364 + }, + { + "epoch": 2.264051066479838, + "grad_norm": 25.227514266967773, + "learning_rate": 1.5058660232485434e-05, + "loss": 0.2646, + "step": 1365 + }, + { + "epoch": 2.265711765011158, + "grad_norm": 13.016712188720703, + "learning_rate": 1.4994658368130875e-05, + "loss": 0.2659, + "step": 1366 + }, + { + "epoch": 2.2673724635424777, + "grad_norm": 71.8415756225586, + "learning_rate": 1.4930768802487415e-05, + "loss": 0.2656, + "step": 1367 + }, + { + "epoch": 2.2690331620737973, + "grad_norm": 57.14645004272461, + "learning_rate": 1.4866991740515651e-05, + "loss": 0.2724, + "step": 1368 + }, + { + "epoch": 2.270693860605117, + "grad_norm": 53.880001068115234, + "learning_rate": 1.4803327386815275e-05, + "loss": 0.2361, + "step": 1369 + }, + { + "epoch": 2.2723545591364367, + "grad_norm": 16.154911041259766, + "learning_rate": 1.4739775945624373e-05, + "loss": 0.2594, + "step": 1370 + }, + { + "epoch": 2.2740152576677564, + "grad_norm": 21.753944396972656, + "learning_rate": 1.467633762081887e-05, + "loss": 0.2581, + "step": 1371 + }, + { + "epoch": 2.275675956199076, + "grad_norm": 17.82083511352539, + "learning_rate": 1.461301261591172e-05, + "loss": 0.2777, + "step": 1372 + }, + { + "epoch": 2.277336654730396, + "grad_norm": 17.128704071044922, + "learning_rate": 1.4549801134052437e-05, + "loss": 0.2637, + "step": 1373 + }, + { + "epoch": 2.278997353261716, + "grad_norm": 20.422931671142578, + "learning_rate": 1.4486703378026255e-05, + "loss": 0.2626, + "step": 1374 + }, + { + "epoch": 2.2806580517930355, + "grad_norm": 279.8073425292969, + "learning_rate": 1.442371955025365e-05, + "loss": 0.2827, + "step": 1375 + }, + { + "epoch": 2.282318750324355, + "grad_norm": 10.705789566040039, + "learning_rate": 1.4360849852789565e-05, + "loss": 0.2789, + "step": 1376 + }, + { + "epoch": 2.283979448855675, + "grad_norm": 20.468900680541992, + "learning_rate": 1.4298094487322811e-05, + "loss": 0.2835, + "step": 1377 + }, + { + "epoch": 2.2856401473869945, + "grad_norm": 25.806373596191406, + "learning_rate": 1.4235453655175429e-05, + "loss": 0.2593, + "step": 1378 + }, + { + "epoch": 2.287300845918314, + "grad_norm": 50.75285720825195, + "learning_rate": 1.4172927557302007e-05, + "loss": 0.2734, + "step": 1379 + }, + { + "epoch": 2.2889615444496343, + "grad_norm": 189.87539672851562, + "learning_rate": 1.4110516394289102e-05, + "loss": 0.2616, + "step": 1380 + }, + { + "epoch": 2.290622242980954, + "grad_norm": 110.41610717773438, + "learning_rate": 1.4048220366354542e-05, + "loss": 0.267, + "step": 1381 + }, + { + "epoch": 2.2922829415122736, + "grad_norm": 22.71442413330078, + "learning_rate": 1.398603967334678e-05, + "loss": 0.2772, + "step": 1382 + }, + { + "epoch": 2.2939436400435933, + "grad_norm": 104.33234405517578, + "learning_rate": 1.392397451474427e-05, + "loss": 0.2926, + "step": 1383 + }, + { + "epoch": 2.295604338574913, + "grad_norm": 12.76983642578125, + "learning_rate": 1.3862025089654823e-05, + "loss": 0.2796, + "step": 1384 + }, + { + "epoch": 2.2972650371062326, + "grad_norm": 12.67192268371582, + "learning_rate": 1.380019159681501e-05, + "loss": 0.2674, + "step": 1385 + }, + { + "epoch": 2.2989257356375523, + "grad_norm": 13.048188209533691, + "learning_rate": 1.3738474234589444e-05, + "loss": 0.2567, + "step": 1386 + }, + { + "epoch": 2.3005864341688724, + "grad_norm": 36.5905876159668, + "learning_rate": 1.367687320097018e-05, + "loss": 0.282, + "step": 1387 + }, + { + "epoch": 2.302247132700192, + "grad_norm": 13.732542991638184, + "learning_rate": 1.3615388693576148e-05, + "loss": 0.2507, + "step": 1388 + }, + { + "epoch": 2.3039078312315118, + "grad_norm": 13.254141807556152, + "learning_rate": 1.3554020909652376e-05, + "loss": 0.2547, + "step": 1389 + }, + { + "epoch": 2.3055685297628314, + "grad_norm": 26.69158172607422, + "learning_rate": 1.3492770046069509e-05, + "loss": 0.286, + "step": 1390 + }, + { + "epoch": 2.307229228294151, + "grad_norm": 16.04448127746582, + "learning_rate": 1.3431636299323053e-05, + "loss": 0.2573, + "step": 1391 + }, + { + "epoch": 2.308889926825471, + "grad_norm": 17.15887451171875, + "learning_rate": 1.3370619865532823e-05, + "loss": 0.2562, + "step": 1392 + }, + { + "epoch": 2.3105506253567905, + "grad_norm": 13.622615814208984, + "learning_rate": 1.3309720940442282e-05, + "loss": 0.2824, + "step": 1393 + }, + { + "epoch": 2.3122113238881106, + "grad_norm": 17.415517807006836, + "learning_rate": 1.3248939719417913e-05, + "loss": 0.2612, + "step": 1394 + }, + { + "epoch": 2.3138720224194302, + "grad_norm": 41.98634719848633, + "learning_rate": 1.3188276397448635e-05, + "loss": 0.2681, + "step": 1395 + }, + { + "epoch": 2.31553272095075, + "grad_norm": 31.89238929748535, + "learning_rate": 1.312773116914508e-05, + "loss": 0.2969, + "step": 1396 + }, + { + "epoch": 2.3171934194820696, + "grad_norm": 66.38880157470703, + "learning_rate": 1.3067304228739096e-05, + "loss": 0.2746, + "step": 1397 + }, + { + "epoch": 2.3188541180133893, + "grad_norm": 27.34821891784668, + "learning_rate": 1.3006995770083013e-05, + "loss": 0.2735, + "step": 1398 + }, + { + "epoch": 2.320514816544709, + "grad_norm": 23.84678077697754, + "learning_rate": 1.2946805986649074e-05, + "loss": 0.2692, + "step": 1399 + }, + { + "epoch": 2.322175515076029, + "grad_norm": 9.86176872253418, + "learning_rate": 1.2886735071528828e-05, + "loss": 0.2747, + "step": 1400 + }, + { + "epoch": 2.3238362136073487, + "grad_norm": 24.083120346069336, + "learning_rate": 1.2826783217432464e-05, + "loss": 0.2701, + "step": 1401 + }, + { + "epoch": 2.3254969121386684, + "grad_norm": 22.610305786132812, + "learning_rate": 5.812931398191593e-05, + "loss": 0.2862, + "step": 1402 + }, + { + "epoch": 2.327157610669988, + "grad_norm": 18.712230682373047, + "learning_rate": 5.807630328666265e-05, + "loss": 0.2699, + "step": 1403 + }, + { + "epoch": 2.3288183092013077, + "grad_norm": 7.991389751434326, + "learning_rate": 5.8023283268368536e-05, + "loss": 0.2492, + "step": 1404 + }, + { + "epoch": 2.3304790077326274, + "grad_norm": 18.018320083618164, + "learning_rate": 5.797025398823829e-05, + "loss": 0.2578, + "step": 1405 + }, + { + "epoch": 2.3321397062639475, + "grad_norm": 63.561912536621094, + "learning_rate": 5.791721550748732e-05, + "loss": 0.2978, + "step": 1406 + }, + { + "epoch": 2.333800404795267, + "grad_norm": 13.887791633605957, + "learning_rate": 5.786416788734165e-05, + "loss": 0.2776, + "step": 1407 + }, + { + "epoch": 2.335461103326587, + "grad_norm": 39.19808578491211, + "learning_rate": 5.781111118903785e-05, + "loss": 0.2493, + "step": 1408 + }, + { + "epoch": 2.3371218018579065, + "grad_norm": 152.4955291748047, + "learning_rate": 5.7758045473822954e-05, + "loss": 0.2803, + "step": 1409 + }, + { + "epoch": 2.338782500389226, + "grad_norm": 28.978900909423828, + "learning_rate": 5.770497080295445e-05, + "loss": 0.2802, + "step": 1410 + }, + { + "epoch": 2.340443198920546, + "grad_norm": 455.5882568359375, + "learning_rate": 5.765188723770013e-05, + "loss": 0.3016, + "step": 1411 + }, + { + "epoch": 2.3421038974518655, + "grad_norm": 15.944208145141602, + "learning_rate": 5.759879483933803e-05, + "loss": 0.274, + "step": 1412 + }, + { + "epoch": 2.3437645959831857, + "grad_norm": 12.824470520019531, + "learning_rate": 5.754569366915648e-05, + "loss": 0.2592, + "step": 1413 + }, + { + "epoch": 2.3454252945145053, + "grad_norm": 12.322883605957031, + "learning_rate": 5.749258378845379e-05, + "loss": 0.2482, + "step": 1414 + }, + { + "epoch": 2.347085993045825, + "grad_norm": 23.785053253173828, + "learning_rate": 5.743946525853845e-05, + "loss": 0.2613, + "step": 1415 + }, + { + "epoch": 2.3487466915771447, + "grad_norm": 19.648841857910156, + "learning_rate": 5.738633814072888e-05, + "loss": 0.283, + "step": 1416 + }, + { + "epoch": 2.3504073901084643, + "grad_norm": 27.808544158935547, + "learning_rate": 5.7333202496353413e-05, + "loss": 0.2783, + "step": 1417 + }, + { + "epoch": 2.352068088639784, + "grad_norm": 16.7478084564209, + "learning_rate": 5.7280058386750255e-05, + "loss": 0.2881, + "step": 1418 + }, + { + "epoch": 2.3537287871711037, + "grad_norm": 13.085315704345703, + "learning_rate": 5.722690587326733e-05, + "loss": 0.2618, + "step": 1419 + }, + { + "epoch": 2.355389485702424, + "grad_norm": 10.254049301147461, + "learning_rate": 5.7173745017262336e-05, + "loss": 0.2848, + "step": 1420 + }, + { + "epoch": 2.3570501842337435, + "grad_norm": 26.449800491333008, + "learning_rate": 5.712057588010251e-05, + "loss": 0.2741, + "step": 1421 + }, + { + "epoch": 2.358710882765063, + "grad_norm": 17.35054588317871, + "learning_rate": 5.706739852316475e-05, + "loss": 0.2859, + "step": 1422 + }, + { + "epoch": 2.360371581296383, + "grad_norm": 21.20619773864746, + "learning_rate": 5.701421300783537e-05, + "loss": 0.2767, + "step": 1423 + }, + { + "epoch": 2.3620322798277025, + "grad_norm": 23.19258689880371, + "learning_rate": 5.6961019395510126e-05, + "loss": 0.2727, + "step": 1424 + }, + { + "epoch": 2.363692978359022, + "grad_norm": 27.887035369873047, + "learning_rate": 5.6907817747594116e-05, + "loss": 0.2966, + "step": 1425 + }, + { + "epoch": 2.365353676890342, + "grad_norm": 24.820415496826172, + "learning_rate": 5.685460812550172e-05, + "loss": 0.2643, + "step": 1426 + }, + { + "epoch": 2.367014375421662, + "grad_norm": 172.7886199951172, + "learning_rate": 5.680139059065651e-05, + "loss": 0.2657, + "step": 1427 + }, + { + "epoch": 2.3686750739529816, + "grad_norm": 8.853065490722656, + "learning_rate": 5.6748165204491234e-05, + "loss": 0.2694, + "step": 1428 + }, + { + "epoch": 2.3703357724843013, + "grad_norm": 22.782773971557617, + "learning_rate": 5.6694932028447635e-05, + "loss": 0.2943, + "step": 1429 + }, + { + "epoch": 2.371996471015621, + "grad_norm": 36.88557052612305, + "learning_rate": 5.664169112397651e-05, + "loss": 0.2547, + "step": 1430 + }, + { + "epoch": 2.3736571695469406, + "grad_norm": 51.210880279541016, + "learning_rate": 5.6588442552537545e-05, + "loss": 0.2799, + "step": 1431 + }, + { + "epoch": 2.3753178680782603, + "grad_norm": 20.286684036254883, + "learning_rate": 5.6535186375599266e-05, + "loss": 0.2718, + "step": 1432 + }, + { + "epoch": 2.37697856660958, + "grad_norm": 20.7040958404541, + "learning_rate": 5.648192265463903e-05, + "loss": 0.2652, + "step": 1433 + }, + { + "epoch": 2.3786392651409, + "grad_norm": 11.31206226348877, + "learning_rate": 5.6428651451142856e-05, + "loss": 0.2746, + "step": 1434 + }, + { + "epoch": 2.3802999636722197, + "grad_norm": 12.960429191589355, + "learning_rate": 5.63753728266054e-05, + "loss": 0.2609, + "step": 1435 + }, + { + "epoch": 2.3819606622035394, + "grad_norm": 12.037378311157227, + "learning_rate": 5.6322086842529906e-05, + "loss": 0.2703, + "step": 1436 + }, + { + "epoch": 2.383621360734859, + "grad_norm": 28.145423889160156, + "learning_rate": 5.626879356042812e-05, + "loss": 0.2708, + "step": 1437 + }, + { + "epoch": 2.3852820592661788, + "grad_norm": 18.637296676635742, + "learning_rate": 5.621549304182019e-05, + "loss": 0.2648, + "step": 1438 + }, + { + "epoch": 2.3869427577974984, + "grad_norm": 106.59896850585938, + "learning_rate": 5.616218534823462e-05, + "loss": 0.2895, + "step": 1439 + }, + { + "epoch": 2.388603456328818, + "grad_norm": 38.32217025756836, + "learning_rate": 5.6108870541208224e-05, + "loss": 0.2839, + "step": 1440 + }, + { + "epoch": 2.390264154860138, + "grad_norm": 23.780033111572266, + "learning_rate": 5.6055548682285964e-05, + "loss": 0.2868, + "step": 1441 + }, + { + "epoch": 2.391924853391458, + "grad_norm": 33.757606506347656, + "learning_rate": 5.600221983302103e-05, + "loss": 0.2651, + "step": 1442 + }, + { + "epoch": 2.3935855519227776, + "grad_norm": 16.432411193847656, + "learning_rate": 5.594888405497461e-05, + "loss": 0.2765, + "step": 1443 + }, + { + "epoch": 2.3952462504540972, + "grad_norm": 49.07794952392578, + "learning_rate": 5.589554140971595e-05, + "loss": 0.2951, + "step": 1444 + }, + { + "epoch": 2.396906948985417, + "grad_norm": 173.6234588623047, + "learning_rate": 5.584219195882214e-05, + "loss": 0.2887, + "step": 1445 + }, + { + "epoch": 2.3985676475167366, + "grad_norm": 9.19729995727539, + "learning_rate": 5.578883576387821e-05, + "loss": 0.2747, + "step": 1446 + }, + { + "epoch": 2.4002283460480562, + "grad_norm": 69.4295425415039, + "learning_rate": 5.5735472886476926e-05, + "loss": 0.2787, + "step": 1447 + }, + { + "epoch": 2.4018890445793764, + "grad_norm": 28.011648178100586, + "learning_rate": 5.568210338821881e-05, + "loss": 0.2734, + "step": 1448 + }, + { + "epoch": 2.403549743110696, + "grad_norm": 20.252784729003906, + "learning_rate": 5.562872733071197e-05, + "loss": 0.2965, + "step": 1449 + }, + { + "epoch": 2.4052104416420157, + "grad_norm": 26.35774040222168, + "learning_rate": 5.5575344775572135e-05, + "loss": 0.2766, + "step": 1450 + }, + { + "epoch": 2.4068711401733354, + "grad_norm": 67.3355712890625, + "learning_rate": 5.5521955784422506e-05, + "loss": 0.2714, + "step": 1451 + }, + { + "epoch": 2.408531838704655, + "grad_norm": 13.52342700958252, + "learning_rate": 5.546856041889373e-05, + "loss": 0.2653, + "step": 1452 + }, + { + "epoch": 2.4101925372359747, + "grad_norm": 39.938289642333984, + "learning_rate": 5.54151587406238e-05, + "loss": 0.2657, + "step": 1453 + }, + { + "epoch": 2.4118532357672944, + "grad_norm": 17.8809814453125, + "learning_rate": 5.5361750811257995e-05, + "loss": 0.2527, + "step": 1454 + }, + { + "epoch": 2.4135139342986145, + "grad_norm": 15.756719589233398, + "learning_rate": 5.530833669244885e-05, + "loss": 0.2697, + "step": 1455 + }, + { + "epoch": 2.415174632829934, + "grad_norm": 25.864686965942383, + "learning_rate": 5.5254916445855974e-05, + "loss": 0.2613, + "step": 1456 + }, + { + "epoch": 2.416835331361254, + "grad_norm": 8.709848403930664, + "learning_rate": 5.520149013314611e-05, + "loss": 0.2416, + "step": 1457 + }, + { + "epoch": 2.4184960298925735, + "grad_norm": 10.386701583862305, + "learning_rate": 5.514805781599297e-05, + "loss": 0.2653, + "step": 1458 + }, + { + "epoch": 2.420156728423893, + "grad_norm": 35.915462493896484, + "learning_rate": 5.509461955607721e-05, + "loss": 0.275, + "step": 1459 + }, + { + "epoch": 2.421817426955213, + "grad_norm": 26.91705322265625, + "learning_rate": 5.504117541508636e-05, + "loss": 0.2856, + "step": 1460 + }, + { + "epoch": 2.423478125486533, + "grad_norm": 14.458738327026367, + "learning_rate": 5.4987725454714713e-05, + "loss": 0.267, + "step": 1461 + }, + { + "epoch": 2.4251388240178526, + "grad_norm": 93.97142028808594, + "learning_rate": 5.493426973666328e-05, + "loss": 0.2637, + "step": 1462 + }, + { + "epoch": 2.4267995225491723, + "grad_norm": 17.098514556884766, + "learning_rate": 5.488080832263974e-05, + "loss": 0.2354, + "step": 1463 + }, + { + "epoch": 2.428460221080492, + "grad_norm": 26.992612838745117, + "learning_rate": 5.4827341274358344e-05, + "loss": 0.2657, + "step": 1464 + }, + { + "epoch": 2.4301209196118116, + "grad_norm": 8.289748191833496, + "learning_rate": 5.477386865353982e-05, + "loss": 0.2581, + "step": 1465 + }, + { + "epoch": 2.4317816181431313, + "grad_norm": 44.98151779174805, + "learning_rate": 5.472039052191138e-05, + "loss": 0.2556, + "step": 1466 + }, + { + "epoch": 2.4334423166744514, + "grad_norm": 48.27017593383789, + "learning_rate": 5.466690694120653e-05, + "loss": 0.2626, + "step": 1467 + }, + { + "epoch": 2.435103015205771, + "grad_norm": 164.41455078125, + "learning_rate": 5.4613417973165106e-05, + "loss": 0.2637, + "step": 1468 + }, + { + "epoch": 2.4367637137370908, + "grad_norm": 37.5134391784668, + "learning_rate": 5.4559923679533173e-05, + "loss": 0.2895, + "step": 1469 + }, + { + "epoch": 2.4384244122684104, + "grad_norm": 18.439903259277344, + "learning_rate": 5.450642412206294e-05, + "loss": 0.241, + "step": 1470 + }, + { + "epoch": 2.44008511079973, + "grad_norm": 87.59689331054688, + "learning_rate": 5.4452919362512646e-05, + "loss": 0.2967, + "step": 1471 + }, + { + "epoch": 2.44174580933105, + "grad_norm": 48.56362533569336, + "learning_rate": 5.439940946264662e-05, + "loss": 0.2689, + "step": 1472 + }, + { + "epoch": 2.4434065078623695, + "grad_norm": 26.88450050354004, + "learning_rate": 5.4345894484235035e-05, + "loss": 0.2577, + "step": 1473 + }, + { + "epoch": 2.4450672063936896, + "grad_norm": 691.6940307617188, + "learning_rate": 5.429237448905398e-05, + "loss": 0.2843, + "step": 1474 + }, + { + "epoch": 2.4467279049250092, + "grad_norm": 27.271812438964844, + "learning_rate": 5.423884953888533e-05, + "loss": 0.2851, + "step": 1475 + }, + { + "epoch": 2.448388603456329, + "grad_norm": 37.123023986816406, + "learning_rate": 5.418531969551666e-05, + "loss": 0.2796, + "step": 1476 + }, + { + "epoch": 2.4500493019876486, + "grad_norm": 14.7449369430542, + "learning_rate": 5.413178502074121e-05, + "loss": 0.3011, + "step": 1477 + }, + { + "epoch": 2.4517100005189683, + "grad_norm": 18.41019058227539, + "learning_rate": 5.40782455763578e-05, + "loss": 0.2525, + "step": 1478 + }, + { + "epoch": 2.453370699050288, + "grad_norm": 151.3030242919922, + "learning_rate": 5.4024701424170756e-05, + "loss": 0.2889, + "step": 1479 + }, + { + "epoch": 2.4550313975816076, + "grad_norm": 14.846268653869629, + "learning_rate": 5.39711526259898e-05, + "loss": 0.2623, + "step": 1480 + }, + { + "epoch": 2.4566920961129277, + "grad_norm": 13.3090181350708, + "learning_rate": 5.391759924363008e-05, + "loss": 0.2522, + "step": 1481 + }, + { + "epoch": 2.4583527946442474, + "grad_norm": 354.9453430175781, + "learning_rate": 5.386404133891199e-05, + "loss": 0.2719, + "step": 1482 + }, + { + "epoch": 2.460013493175567, + "grad_norm": 171.53936767578125, + "learning_rate": 5.381047897366117e-05, + "loss": 0.2816, + "step": 1483 + }, + { + "epoch": 2.4616741917068867, + "grad_norm": 31.255678176879883, + "learning_rate": 5.375691220970839e-05, + "loss": 0.268, + "step": 1484 + }, + { + "epoch": 2.4633348902382064, + "grad_norm": 23.45043182373047, + "learning_rate": 5.3703341108889504e-05, + "loss": 0.2582, + "step": 1485 + }, + { + "epoch": 2.464995588769526, + "grad_norm": 11.942648887634277, + "learning_rate": 5.364976573304538e-05, + "loss": 0.2732, + "step": 1486 + }, + { + "epoch": 2.4666562873008457, + "grad_norm": 50.64974594116211, + "learning_rate": 5.359618614402182e-05, + "loss": 0.2684, + "step": 1487 + }, + { + "epoch": 2.468316985832166, + "grad_norm": 30.816364288330078, + "learning_rate": 5.354260240366947e-05, + "loss": 0.2733, + "step": 1488 + }, + { + "epoch": 2.4699776843634855, + "grad_norm": 16.858089447021484, + "learning_rate": 5.3489014573843796e-05, + "loss": 0.2517, + "step": 1489 + }, + { + "epoch": 2.471638382894805, + "grad_norm": 26.83182144165039, + "learning_rate": 5.343542271640498e-05, + "loss": 0.2464, + "step": 1490 + }, + { + "epoch": 2.473299081426125, + "grad_norm": 12.383484840393066, + "learning_rate": 5.338182689321782e-05, + "loss": 0.2754, + "step": 1491 + }, + { + "epoch": 2.4749597799574445, + "grad_norm": 72.79269409179688, + "learning_rate": 5.332822716615172e-05, + "loss": 0.2767, + "step": 1492 + }, + { + "epoch": 2.476620478488764, + "grad_norm": 87.32337188720703, + "learning_rate": 5.327462359708063e-05, + "loss": 0.2674, + "step": 1493 + }, + { + "epoch": 2.478281177020084, + "grad_norm": 33.9007568359375, + "learning_rate": 5.3221016247882825e-05, + "loss": 0.2829, + "step": 1494 + }, + { + "epoch": 2.479941875551404, + "grad_norm": 16.268342971801758, + "learning_rate": 5.316740518044108e-05, + "loss": 0.2663, + "step": 1495 + }, + { + "epoch": 2.4816025740827237, + "grad_norm": 186.22145080566406, + "learning_rate": 5.3113790456642345e-05, + "loss": 0.2949, + "step": 1496 + }, + { + "epoch": 2.4832632726140433, + "grad_norm": 75.62686920166016, + "learning_rate": 5.306017213837784e-05, + "loss": 0.2485, + "step": 1497 + }, + { + "epoch": 2.484923971145363, + "grad_norm": 27.538183212280273, + "learning_rate": 5.300655028754298e-05, + "loss": 0.2595, + "step": 1498 + }, + { + "epoch": 2.4865846696766827, + "grad_norm": 23.336647033691406, + "learning_rate": 5.295292496603718e-05, + "loss": 0.251, + "step": 1499 + }, + { + "epoch": 2.4882453682080024, + "grad_norm": 34.83784103393555, + "learning_rate": 5.289929623576386e-05, + "loss": 0.2562, + "step": 1500 + }, + { + "epoch": 2.489906066739322, + "grad_norm": 10.1558837890625, + "learning_rate": 5.284566415863047e-05, + "loss": 0.2862, + "step": 1501 + }, + { + "epoch": 2.491566765270642, + "grad_norm": 31.47645378112793, + "learning_rate": 5.2792028796548234e-05, + "loss": 0.2696, + "step": 1502 + }, + { + "epoch": 2.493227463801962, + "grad_norm": 96.06140899658203, + "learning_rate": 5.273839021143218e-05, + "loss": 0.2774, + "step": 1503 + }, + { + "epoch": 2.4948881623332815, + "grad_norm": 15.568284034729004, + "learning_rate": 5.268474846520112e-05, + "loss": 0.2745, + "step": 1504 + }, + { + "epoch": 2.496548860864601, + "grad_norm": 42.614139556884766, + "learning_rate": 5.263110361977741e-05, + "loss": 0.268, + "step": 1505 + }, + { + "epoch": 2.498209559395921, + "grad_norm": 29.587797164916992, + "learning_rate": 5.257745573708707e-05, + "loss": 0.2893, + "step": 1506 + }, + { + "epoch": 2.4998702579272405, + "grad_norm": 9.840633392333984, + "learning_rate": 5.252380487905959e-05, + "loss": 0.2578, + "step": 1507 + }, + { + "epoch": 2.50153095645856, + "grad_norm": 18.386051177978516, + "learning_rate": 5.2470151107627907e-05, + "loss": 0.2837, + "step": 1508 + }, + { + "epoch": 2.5031916549898803, + "grad_norm": 55.89763259887695, + "learning_rate": 5.241649448472833e-05, + "loss": 0.2585, + "step": 1509 + }, + { + "epoch": 2.5048523535212, + "grad_norm": 53.75349807739258, + "learning_rate": 5.236283507230042e-05, + "loss": 0.2482, + "step": 1510 + }, + { + "epoch": 2.5065130520525196, + "grad_norm": 12.337052345275879, + "learning_rate": 5.230917293228699e-05, + "loss": 0.2704, + "step": 1511 + }, + { + "epoch": 2.5081737505838393, + "grad_norm": 35.96687698364258, + "learning_rate": 5.225550812663399e-05, + "loss": 0.2798, + "step": 1512 + }, + { + "epoch": 2.509834449115159, + "grad_norm": 22.93730926513672, + "learning_rate": 5.220184071729047e-05, + "loss": 0.251, + "step": 1513 + }, + { + "epoch": 2.511495147646479, + "grad_norm": 14.520767211914062, + "learning_rate": 5.214817076620848e-05, + "loss": 0.2467, + "step": 1514 + }, + { + "epoch": 2.5131558461777983, + "grad_norm": 11.448175430297852, + "learning_rate": 5.209449833534295e-05, + "loss": 0.2612, + "step": 1515 + }, + { + "epoch": 2.5148165447091184, + "grad_norm": 24.107498168945312, + "learning_rate": 5.2040823486651736e-05, + "loss": 0.2512, + "step": 1516 + }, + { + "epoch": 2.516477243240438, + "grad_norm": 22.277687072753906, + "learning_rate": 5.1987146282095475e-05, + "loss": 0.2605, + "step": 1517 + }, + { + "epoch": 2.5181379417717578, + "grad_norm": 25.920631408691406, + "learning_rate": 5.1933466783637485e-05, + "loss": 0.2851, + "step": 1518 + }, + { + "epoch": 2.5197986403030774, + "grad_norm": 13.566516876220703, + "learning_rate": 5.187978505324379e-05, + "loss": 0.2585, + "step": 1519 + }, + { + "epoch": 2.521459338834397, + "grad_norm": 60.2373046875, + "learning_rate": 5.182610115288295e-05, + "loss": 0.287, + "step": 1520 + }, + { + "epoch": 2.523120037365717, + "grad_norm": 44.851165771484375, + "learning_rate": 5.177241514452602e-05, + "loss": 0.264, + "step": 1521 + }, + { + "epoch": 2.5247807358970364, + "grad_norm": 26.135576248168945, + "learning_rate": 5.171872709014655e-05, + "loss": 0.2465, + "step": 1522 + }, + { + "epoch": 2.5264414344283566, + "grad_norm": 63.41693115234375, + "learning_rate": 5.166503705172037e-05, + "loss": 0.2871, + "step": 1523 + }, + { + "epoch": 2.5281021329596762, + "grad_norm": 13.224884033203125, + "learning_rate": 5.161134509122564e-05, + "loss": 0.2576, + "step": 1524 + }, + { + "epoch": 2.529762831490996, + "grad_norm": 163.56439208984375, + "learning_rate": 5.155765127064277e-05, + "loss": 0.2657, + "step": 1525 + }, + { + "epoch": 2.5314235300223156, + "grad_norm": 29.05950164794922, + "learning_rate": 5.1503955651954263e-05, + "loss": 0.2823, + "step": 1526 + }, + { + "epoch": 2.5330842285536352, + "grad_norm": 21.972442626953125, + "learning_rate": 5.145025829714474e-05, + "loss": 0.2774, + "step": 1527 + }, + { + "epoch": 2.5347449270849554, + "grad_norm": 9.311490058898926, + "learning_rate": 5.139655926820078e-05, + "loss": 0.2678, + "step": 1528 + }, + { + "epoch": 2.5364056256162746, + "grad_norm": 93.41901397705078, + "learning_rate": 5.134285862711096e-05, + "loss": 0.2744, + "step": 1529 + }, + { + "epoch": 2.5380663241475947, + "grad_norm": 52.7574577331543, + "learning_rate": 5.1289156435865634e-05, + "loss": 0.2745, + "step": 1530 + }, + { + "epoch": 2.5397270226789144, + "grad_norm": 14.880579948425293, + "learning_rate": 5.123545275645703e-05, + "loss": 0.292, + "step": 1531 + }, + { + "epoch": 2.541387721210234, + "grad_norm": 21.989952087402344, + "learning_rate": 5.118174765087903e-05, + "loss": 0.2719, + "step": 1532 + }, + { + "epoch": 2.5430484197415537, + "grad_norm": 72.66094970703125, + "learning_rate": 5.1128041181127196e-05, + "loss": 0.2891, + "step": 1533 + }, + { + "epoch": 2.5447091182728734, + "grad_norm": 30.119720458984375, + "learning_rate": 5.107433340919867e-05, + "loss": 0.2483, + "step": 1534 + }, + { + "epoch": 2.5463698168041935, + "grad_norm": 9.786760330200195, + "learning_rate": 5.1020624397092064e-05, + "loss": 0.2775, + "step": 1535 + }, + { + "epoch": 2.5480305153355127, + "grad_norm": 16.673677444458008, + "learning_rate": 5.096691420680745e-05, + "loss": 0.2506, + "step": 1536 + }, + { + "epoch": 2.549691213866833, + "grad_norm": 20.44023895263672, + "learning_rate": 5.091320290034625e-05, + "loss": 0.2737, + "step": 1537 + }, + { + "epoch": 2.5513519123981525, + "grad_norm": 20.888185501098633, + "learning_rate": 5.085949053971116e-05, + "loss": 0.2604, + "step": 1538 + }, + { + "epoch": 2.553012610929472, + "grad_norm": 10.797704696655273, + "learning_rate": 5.0805777186906146e-05, + "loss": 0.2487, + "step": 1539 + }, + { + "epoch": 2.554673309460792, + "grad_norm": 19.005231857299805, + "learning_rate": 5.075206290393625e-05, + "loss": 0.2694, + "step": 1540 + }, + { + "epoch": 2.5563340079921115, + "grad_norm": 15.410425186157227, + "learning_rate": 5.069834775280763e-05, + "loss": 0.2336, + "step": 1541 + }, + { + "epoch": 2.5579947065234316, + "grad_norm": 50.80555725097656, + "learning_rate": 5.064463179552743e-05, + "loss": 0.2847, + "step": 1542 + }, + { + "epoch": 2.5596554050547513, + "grad_norm": 569.7647705078125, + "learning_rate": 5.059091509410374e-05, + "loss": 0.272, + "step": 1543 + }, + { + "epoch": 2.561316103586071, + "grad_norm": 18.915205001831055, + "learning_rate": 5.053719771054548e-05, + "loss": 0.2487, + "step": 1544 + }, + { + "epoch": 2.5629768021173907, + "grad_norm": 43.03477478027344, + "learning_rate": 5.04834797068624e-05, + "loss": 0.2483, + "step": 1545 + }, + { + "epoch": 2.5646375006487103, + "grad_norm": 172.41160583496094, + "learning_rate": 5.042976114506496e-05, + "loss": 0.2757, + "step": 1546 + }, + { + "epoch": 2.56629819918003, + "grad_norm": 44.495338439941406, + "learning_rate": 5.0376042087164185e-05, + "loss": 0.2701, + "step": 1547 + }, + { + "epoch": 2.5679588977113497, + "grad_norm": 11.617120742797852, + "learning_rate": 5.032232259517179e-05, + "loss": 0.2719, + "step": 1548 + }, + { + "epoch": 2.56961959624267, + "grad_norm": 66.66431427001953, + "learning_rate": 5.0268602731099914e-05, + "loss": 0.2699, + "step": 1549 + }, + { + "epoch": 2.5712802947739894, + "grad_norm": 16.106897354125977, + "learning_rate": 5.021488255696113e-05, + "loss": 0.278, + "step": 1550 + }, + { + "epoch": 2.572940993305309, + "grad_norm": 18.151145935058594, + "learning_rate": 5.016116213476842e-05, + "loss": 0.2805, + "step": 1551 + }, + { + "epoch": 2.574601691836629, + "grad_norm": 18.945083618164062, + "learning_rate": 5.010744152653501e-05, + "loss": 0.2741, + "step": 1552 + }, + { + "epoch": 2.5762623903679485, + "grad_norm": 31.062414169311523, + "learning_rate": 5.005372079427434e-05, + "loss": 0.2761, + "step": 1553 + }, + { + "epoch": 2.577923088899268, + "grad_norm": 662.2258911132812, + "learning_rate": 5e-05, + "loss": 0.2699, + "step": 1554 + }, + { + "epoch": 2.579583787430588, + "grad_norm": 14.599547386169434, + "learning_rate": 4.994627920572567e-05, + "loss": 0.2539, + "step": 1555 + }, + { + "epoch": 2.581244485961908, + "grad_norm": 11.749177932739258, + "learning_rate": 4.989255847346499e-05, + "loss": 0.2732, + "step": 1556 + }, + { + "epoch": 2.5829051844932276, + "grad_norm": 18.639333724975586, + "learning_rate": 4.983883786523159e-05, + "loss": 0.2463, + "step": 1557 + }, + { + "epoch": 2.5845658830245473, + "grad_norm": 20.219881057739258, + "learning_rate": 4.978511744303887e-05, + "loss": 0.2635, + "step": 1558 + }, + { + "epoch": 2.586226581555867, + "grad_norm": 17.19252586364746, + "learning_rate": 4.9731397268900104e-05, + "loss": 0.2645, + "step": 1559 + }, + { + "epoch": 2.5878872800871866, + "grad_norm": 16.56593132019043, + "learning_rate": 4.967767740482824e-05, + "loss": 0.3083, + "step": 1560 + }, + { + "epoch": 2.5895479786185063, + "grad_norm": 20.651159286499023, + "learning_rate": 4.962395791283582e-05, + "loss": 0.2615, + "step": 1561 + }, + { + "epoch": 2.591208677149826, + "grad_norm": 14.17055606842041, + "learning_rate": 4.9570238854935066e-05, + "loss": 0.2638, + "step": 1562 + }, + { + "epoch": 2.592869375681146, + "grad_norm": 12.651618957519531, + "learning_rate": 4.951652029313759e-05, + "loss": 0.2427, + "step": 1563 + }, + { + "epoch": 2.5945300742124657, + "grad_norm": 115.00933837890625, + "learning_rate": 4.946280228945453e-05, + "loss": 0.2897, + "step": 1564 + }, + { + "epoch": 2.5961907727437854, + "grad_norm": 12.440101623535156, + "learning_rate": 4.940908490589628e-05, + "loss": 0.2688, + "step": 1565 + }, + { + "epoch": 2.597851471275105, + "grad_norm": 14.037446975708008, + "learning_rate": 4.935536820447258e-05, + "loss": 0.2728, + "step": 1566 + }, + { + "epoch": 2.5995121698064247, + "grad_norm": 24.350257873535156, + "learning_rate": 4.930165224719239e-05, + "loss": 0.2755, + "step": 1567 + }, + { + "epoch": 2.6011728683377444, + "grad_norm": 21.01694107055664, + "learning_rate": 4.9247937096063756e-05, + "loss": 0.2794, + "step": 1568 + }, + { + "epoch": 2.602833566869064, + "grad_norm": 27.4760799407959, + "learning_rate": 4.919422281309387e-05, + "loss": 0.272, + "step": 1569 + }, + { + "epoch": 2.604494265400384, + "grad_norm": 25.87298011779785, + "learning_rate": 4.9140509460288836e-05, + "loss": 0.2997, + "step": 1570 + }, + { + "epoch": 2.606154963931704, + "grad_norm": 16.832847595214844, + "learning_rate": 4.9086797099653756e-05, + "loss": 0.2571, + "step": 1571 + }, + { + "epoch": 2.6078156624630235, + "grad_norm": 237.48619079589844, + "learning_rate": 4.9033085793192574e-05, + "loss": 0.2785, + "step": 1572 + }, + { + "epoch": 2.609476360994343, + "grad_norm": 39.28139877319336, + "learning_rate": 4.897937560290795e-05, + "loss": 0.283, + "step": 1573 + }, + { + "epoch": 2.611137059525663, + "grad_norm": 51.18706130981445, + "learning_rate": 4.8925666590801346e-05, + "loss": 0.2576, + "step": 1574 + }, + { + "epoch": 2.6127977580569826, + "grad_norm": 41.859642028808594, + "learning_rate": 4.8871958818872796e-05, + "loss": 0.2568, + "step": 1575 + }, + { + "epoch": 2.6144584565883022, + "grad_norm": 10.604504585266113, + "learning_rate": 4.881825234912098e-05, + "loss": 0.2715, + "step": 1576 + }, + { + "epoch": 2.6161191551196223, + "grad_norm": 19.53549575805664, + "learning_rate": 4.876454724354299e-05, + "loss": 0.2511, + "step": 1577 + }, + { + "epoch": 2.617779853650942, + "grad_norm": 19.50374984741211, + "learning_rate": 4.871084356413437e-05, + "loss": 0.258, + "step": 1578 + }, + { + "epoch": 2.6194405521822617, + "grad_norm": 284.0364074707031, + "learning_rate": 4.865714137288906e-05, + "loss": 0.2631, + "step": 1579 + }, + { + "epoch": 2.6211012507135814, + "grad_norm": 9.268389701843262, + "learning_rate": 4.8603440731799216e-05, + "loss": 0.2578, + "step": 1580 + }, + { + "epoch": 2.622761949244901, + "grad_norm": 12.844319343566895, + "learning_rate": 4.854974170285528e-05, + "loss": 0.2407, + "step": 1581 + }, + { + "epoch": 2.624422647776221, + "grad_norm": 16.292203903198242, + "learning_rate": 4.849604434804575e-05, + "loss": 0.2492, + "step": 1582 + }, + { + "epoch": 2.6260833463075404, + "grad_norm": 37.84053421020508, + "learning_rate": 4.8442348729357235e-05, + "loss": 0.265, + "step": 1583 + }, + { + "epoch": 2.6277440448388605, + "grad_norm": 11.34038257598877, + "learning_rate": 4.838865490877438e-05, + "loss": 0.2671, + "step": 1584 + }, + { + "epoch": 2.62940474337018, + "grad_norm": 13.328808784484863, + "learning_rate": 4.833496294827965e-05, + "loss": 0.2491, + "step": 1585 + }, + { + "epoch": 2.6310654419015, + "grad_norm": 23.189199447631836, + "learning_rate": 4.828127290985347e-05, + "loss": 0.278, + "step": 1586 + }, + { + "epoch": 2.6327261404328195, + "grad_norm": 49.71121597290039, + "learning_rate": 4.8227584855474e-05, + "loss": 0.2646, + "step": 1587 + }, + { + "epoch": 2.634386838964139, + "grad_norm": 24.127573013305664, + "learning_rate": 4.817389884711705e-05, + "loss": 0.2724, + "step": 1588 + }, + { + "epoch": 2.6360475374954593, + "grad_norm": 13.482179641723633, + "learning_rate": 4.812021494675622e-05, + "loss": 0.2741, + "step": 1589 + }, + { + "epoch": 2.6377082360267785, + "grad_norm": 37.96223449707031, + "learning_rate": 4.806653321636251e-05, + "loss": 0.2707, + "step": 1590 + }, + { + "epoch": 2.6393689345580986, + "grad_norm": 49.1181526184082, + "learning_rate": 4.801285371790454e-05, + "loss": 0.2716, + "step": 1591 + }, + { + "epoch": 2.6410296330894183, + "grad_norm": 15.158787727355957, + "learning_rate": 4.795917651334828e-05, + "loss": 0.2557, + "step": 1592 + }, + { + "epoch": 2.642690331620738, + "grad_norm": 37.52794647216797, + "learning_rate": 4.7905501664657064e-05, + "loss": 0.2793, + "step": 1593 + }, + { + "epoch": 2.6443510301520576, + "grad_norm": 14.638628959655762, + "learning_rate": 4.7851829233791536e-05, + "loss": 0.2594, + "step": 1594 + }, + { + "epoch": 2.6460117286833773, + "grad_norm": 1139.1351318359375, + "learning_rate": 4.7798159282709525e-05, + "loss": 0.2827, + "step": 1595 + }, + { + "epoch": 2.6476724272146974, + "grad_norm": 44.98678207397461, + "learning_rate": 4.774449187336602e-05, + "loss": 0.2645, + "step": 1596 + }, + { + "epoch": 2.6493331257460166, + "grad_norm": 35.51799011230469, + "learning_rate": 4.7690827067713035e-05, + "loss": 0.2551, + "step": 1597 + }, + { + "epoch": 2.6509938242773368, + "grad_norm": 46.31774139404297, + "learning_rate": 4.7637164927699595e-05, + "loss": 0.2777, + "step": 1598 + }, + { + "epoch": 2.6526545228086564, + "grad_norm": 10.960673332214355, + "learning_rate": 4.75835055152717e-05, + "loss": 0.2835, + "step": 1599 + }, + { + "epoch": 2.654315221339976, + "grad_norm": 46.794979095458984, + "learning_rate": 4.7529848892372085e-05, + "loss": 0.2513, + "step": 1600 + } + ], + "logging_steps": 1.0, + "max_steps": 3015, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 200, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.1500836381746284e+18, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}