diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,14620 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 500, + "global_step": 2078, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0009624639076034649, + "grad_norm": 8.533728170597175, + "learning_rate": 9.999994285890155e-06, + "loss": 1.0071, + "step": 1 + }, + { + "epoch": 0.0019249278152069298, + "grad_norm": 4.954957226621603, + "learning_rate": 9.999977143573675e-06, + "loss": 1.0433, + "step": 2 + }, + { + "epoch": 0.0028873917228103944, + "grad_norm": 3.4746124747129605, + "learning_rate": 9.999948573089744e-06, + "loss": 0.9501, + "step": 3 + }, + { + "epoch": 0.0038498556304138597, + "grad_norm": 2.3404107720838447, + "learning_rate": 9.999908574503663e-06, + "loss": 0.7827, + "step": 4 + }, + { + "epoch": 0.004812319538017324, + "grad_norm": 2.980095111944915, + "learning_rate": 9.999857147906855e-06, + "loss": 1.0233, + "step": 5 + }, + { + "epoch": 0.005774783445620789, + "grad_norm": 1.8682813579048214, + "learning_rate": 9.999794293416862e-06, + "loss": 1.0142, + "step": 6 + }, + { + "epoch": 0.006737247353224254, + "grad_norm": 2.03308775343469, + "learning_rate": 9.999720011177348e-06, + "loss": 1.0316, + "step": 7 + }, + { + "epoch": 0.007699711260827719, + "grad_norm": 49.604088334711555, + "learning_rate": 9.999634301358096e-06, + "loss": 1.0006, + "step": 8 + }, + { + "epoch": 0.008662175168431183, + "grad_norm": 1.5191776557626657, + "learning_rate": 9.999537164155008e-06, + "loss": 0.7199, + "step": 9 + }, + { + "epoch": 0.009624639076034648, + "grad_norm": 1.6875798836566653, + "learning_rate": 9.999428599790104e-06, + "loss": 0.8919, + "step": 10 + }, + { + "epoch": 0.010587102983638113, + "grad_norm": 1.912551156111478, + "learning_rate": 9.999308608511525e-06, + "loss": 1.0377, + "step": 11 + }, + { + "epoch": 0.011549566891241578, + "grad_norm": 1.751709148375545, + "learning_rate": 9.999177190593526e-06, + "loss": 0.9233, + "step": 12 + }, + { + "epoch": 0.012512030798845043, + "grad_norm": 1.5936571585644788, + "learning_rate": 9.999034346336483e-06, + "loss": 0.8637, + "step": 13 + }, + { + "epoch": 0.013474494706448507, + "grad_norm": 20.02475877380703, + "learning_rate": 9.998880076066888e-06, + "loss": 0.6973, + "step": 14 + }, + { + "epoch": 0.014436958614051972, + "grad_norm": 1.4901915172973594, + "learning_rate": 9.998714380137348e-06, + "loss": 0.8097, + "step": 15 + }, + { + "epoch": 0.015399422521655439, + "grad_norm": 1.2703819184023857, + "learning_rate": 9.998537258926582e-06, + "loss": 0.7067, + "step": 16 + }, + { + "epoch": 0.016361886429258902, + "grad_norm": 1.5768262138840043, + "learning_rate": 9.99834871283943e-06, + "loss": 1.0075, + "step": 17 + }, + { + "epoch": 0.017324350336862367, + "grad_norm": 1.2543174241397173, + "learning_rate": 9.998148742306837e-06, + "loss": 0.8004, + "step": 18 + }, + { + "epoch": 0.01828681424446583, + "grad_norm": 24.722034121776066, + "learning_rate": 9.99793734778587e-06, + "loss": 0.5175, + "step": 19 + }, + { + "epoch": 0.019249278152069296, + "grad_norm": 1.722782776413548, + "learning_rate": 9.997714529759695e-06, + "loss": 0.8788, + "step": 20 + }, + { + "epoch": 0.02021174205967276, + "grad_norm": 1.1478872437988172, + "learning_rate": 9.997480288737598e-06, + "loss": 0.6227, + "step": 21 + }, + { + "epoch": 0.021174205967276226, + "grad_norm": 1.2361089194248105, + "learning_rate": 9.99723462525497e-06, + "loss": 0.7873, + "step": 22 + }, + { + "epoch": 0.02213666987487969, + "grad_norm": 1.4436580539346915, + "learning_rate": 9.996977539873313e-06, + "loss": 0.925, + "step": 23 + }, + { + "epoch": 0.023099133782483156, + "grad_norm": 15.058849614506642, + "learning_rate": 9.99670903318023e-06, + "loss": 0.8977, + "step": 24 + }, + { + "epoch": 0.02406159769008662, + "grad_norm": 1.5583037380818676, + "learning_rate": 9.996429105789432e-06, + "loss": 0.8809, + "step": 25 + }, + { + "epoch": 0.025024061597690085, + "grad_norm": 1.1102696813484056, + "learning_rate": 9.99613775834073e-06, + "loss": 0.6093, + "step": 26 + }, + { + "epoch": 0.02598652550529355, + "grad_norm": 1.4851252316615888, + "learning_rate": 9.995834991500049e-06, + "loss": 0.9919, + "step": 27 + }, + { + "epoch": 0.026948989412897015, + "grad_norm": 1.2381494595494837, + "learning_rate": 9.995520805959397e-06, + "loss": 0.8471, + "step": 28 + }, + { + "epoch": 0.02791145332050048, + "grad_norm": 1.2728197312237655, + "learning_rate": 9.995195202436899e-06, + "loss": 0.823, + "step": 29 + }, + { + "epoch": 0.028873917228103944, + "grad_norm": 1.1986580490449303, + "learning_rate": 9.994858181676761e-06, + "loss": 0.6754, + "step": 30 + }, + { + "epoch": 0.029836381135707413, + "grad_norm": 1.4075477944879133, + "learning_rate": 9.994509744449294e-06, + "loss": 0.8698, + "step": 31 + }, + { + "epoch": 0.030798845043310877, + "grad_norm": 1.36964487264404, + "learning_rate": 9.994149891550906e-06, + "loss": 0.961, + "step": 32 + }, + { + "epoch": 0.03176130895091434, + "grad_norm": 1.2888698866985555, + "learning_rate": 9.99377862380409e-06, + "loss": 0.8475, + "step": 33 + }, + { + "epoch": 0.032723772858517804, + "grad_norm": 1.461883206612379, + "learning_rate": 9.99339594205743e-06, + "loss": 0.9846, + "step": 34 + }, + { + "epoch": 0.03368623676612127, + "grad_norm": 1.3151800847026758, + "learning_rate": 9.993001847185601e-06, + "loss": 0.8573, + "step": 35 + }, + { + "epoch": 0.03464870067372473, + "grad_norm": 1.3400156086679185, + "learning_rate": 9.992596340089365e-06, + "loss": 0.8911, + "step": 36 + }, + { + "epoch": 0.0356111645813282, + "grad_norm": 1.334534537740849, + "learning_rate": 9.992179421695568e-06, + "loss": 0.7738, + "step": 37 + }, + { + "epoch": 0.03657362848893166, + "grad_norm": 1.1668724628780804, + "learning_rate": 9.991751092957132e-06, + "loss": 0.7267, + "step": 38 + }, + { + "epoch": 0.03753609239653513, + "grad_norm": 1.2665323547723972, + "learning_rate": 9.991311354853068e-06, + "loss": 0.831, + "step": 39 + }, + { + "epoch": 0.03849855630413859, + "grad_norm": 9.602648279319467, + "learning_rate": 9.99086020838846e-06, + "loss": 0.7758, + "step": 40 + }, + { + "epoch": 0.03946102021174206, + "grad_norm": 1.3484511148059184, + "learning_rate": 9.99039765459447e-06, + "loss": 0.6557, + "step": 41 + }, + { + "epoch": 0.04042348411934552, + "grad_norm": 1.505406859498839, + "learning_rate": 9.989923694528328e-06, + "loss": 0.8052, + "step": 42 + }, + { + "epoch": 0.04138594802694899, + "grad_norm": 1.3879520087995323, + "learning_rate": 9.989438329273338e-06, + "loss": 0.9107, + "step": 43 + }, + { + "epoch": 0.04234841193455245, + "grad_norm": 1.0195360875502828, + "learning_rate": 9.988941559938874e-06, + "loss": 0.6438, + "step": 44 + }, + { + "epoch": 0.04331087584215592, + "grad_norm": 1.4786548381706306, + "learning_rate": 9.988433387660372e-06, + "loss": 0.9521, + "step": 45 + }, + { + "epoch": 0.04427333974975938, + "grad_norm": 1.3677033498388815, + "learning_rate": 9.987913813599337e-06, + "loss": 0.896, + "step": 46 + }, + { + "epoch": 0.04523580365736285, + "grad_norm": 1.363067326174573, + "learning_rate": 9.987382838943325e-06, + "loss": 0.8834, + "step": 47 + }, + { + "epoch": 0.04619826756496631, + "grad_norm": 1.3710271868832096, + "learning_rate": 9.986840464905958e-06, + "loss": 0.9842, + "step": 48 + }, + { + "epoch": 0.04716073147256978, + "grad_norm": 1.3969352128545256, + "learning_rate": 9.98628669272691e-06, + "loss": 0.9817, + "step": 49 + }, + { + "epoch": 0.04812319538017324, + "grad_norm": 1.499471791926457, + "learning_rate": 9.985721523671905e-06, + "loss": 0.9999, + "step": 50 + }, + { + "epoch": 0.04908565928777671, + "grad_norm": 1.342579125304116, + "learning_rate": 9.985144959032721e-06, + "loss": 0.8592, + "step": 51 + }, + { + "epoch": 0.05004812319538017, + "grad_norm": 1.3826594729661021, + "learning_rate": 9.984557000127177e-06, + "loss": 0.8337, + "step": 52 + }, + { + "epoch": 0.05101058710298364, + "grad_norm": 1.3499536915433716, + "learning_rate": 9.98395764829914e-06, + "loss": 0.983, + "step": 53 + }, + { + "epoch": 0.0519730510105871, + "grad_norm": 1.2257247965511626, + "learning_rate": 9.983346904918514e-06, + "loss": 0.7699, + "step": 54 + }, + { + "epoch": 0.05293551491819057, + "grad_norm": 1.3526971946047717, + "learning_rate": 9.982724771381239e-06, + "loss": 0.9494, + "step": 55 + }, + { + "epoch": 0.05389797882579403, + "grad_norm": 0.916102608506668, + "learning_rate": 9.982091249109292e-06, + "loss": 0.5185, + "step": 56 + }, + { + "epoch": 0.0548604427333975, + "grad_norm": 1.3390374502263394, + "learning_rate": 9.98144633955068e-06, + "loss": 0.7696, + "step": 57 + }, + { + "epoch": 0.05582290664100096, + "grad_norm": 1.1842564005216079, + "learning_rate": 9.980790044179438e-06, + "loss": 0.7482, + "step": 58 + }, + { + "epoch": 0.05678537054860443, + "grad_norm": 1.1836208909246537, + "learning_rate": 9.980122364495619e-06, + "loss": 0.8247, + "step": 59 + }, + { + "epoch": 0.05774783445620789, + "grad_norm": 5.294289684900849, + "learning_rate": 9.979443302025305e-06, + "loss": 0.8636, + "step": 60 + }, + { + "epoch": 0.05871029836381136, + "grad_norm": 1.476897843698247, + "learning_rate": 9.97875285832059e-06, + "loss": 0.9399, + "step": 61 + }, + { + "epoch": 0.059672762271414825, + "grad_norm": 1.272016083565467, + "learning_rate": 9.978051034959584e-06, + "loss": 0.7785, + "step": 62 + }, + { + "epoch": 0.06063522617901829, + "grad_norm": 1.1046200843092002, + "learning_rate": 9.9773378335464e-06, + "loss": 0.7343, + "step": 63 + }, + { + "epoch": 0.061597690086621755, + "grad_norm": 1.2180783625800948, + "learning_rate": 9.976613255711169e-06, + "loss": 0.7389, + "step": 64 + }, + { + "epoch": 0.06256015399422522, + "grad_norm": 1.3618699945182609, + "learning_rate": 9.975877303110014e-06, + "loss": 0.9162, + "step": 65 + }, + { + "epoch": 0.06352261790182868, + "grad_norm": 1.2570116705437782, + "learning_rate": 9.975129977425062e-06, + "loss": 0.7737, + "step": 66 + }, + { + "epoch": 0.06448508180943215, + "grad_norm": 1.3315987008735837, + "learning_rate": 9.974371280364432e-06, + "loss": 0.8996, + "step": 67 + }, + { + "epoch": 0.06544754571703561, + "grad_norm": 1.3813934526565022, + "learning_rate": 9.973601213662238e-06, + "loss": 0.9848, + "step": 68 + }, + { + "epoch": 0.06641000962463908, + "grad_norm": 1.3914242698966028, + "learning_rate": 9.972819779078574e-06, + "loss": 0.9868, + "step": 69 + }, + { + "epoch": 0.06737247353224254, + "grad_norm": 1.1124633547619174, + "learning_rate": 9.972026978399527e-06, + "loss": 0.7582, + "step": 70 + }, + { + "epoch": 0.068334937439846, + "grad_norm": 1.1902117610450478, + "learning_rate": 9.971222813437153e-06, + "loss": 0.7911, + "step": 71 + }, + { + "epoch": 0.06929740134744947, + "grad_norm": 3.5091894777821278, + "learning_rate": 9.970407286029488e-06, + "loss": 0.6935, + "step": 72 + }, + { + "epoch": 0.07025986525505294, + "grad_norm": 1.1765023055655908, + "learning_rate": 9.969580398040535e-06, + "loss": 0.7074, + "step": 73 + }, + { + "epoch": 0.0712223291626564, + "grad_norm": 1.3055343139829314, + "learning_rate": 9.96874215136027e-06, + "loss": 0.8696, + "step": 74 + }, + { + "epoch": 0.07218479307025986, + "grad_norm": 1.279081780028313, + "learning_rate": 9.967892547904624e-06, + "loss": 0.8524, + "step": 75 + }, + { + "epoch": 0.07314725697786333, + "grad_norm": 4.033981508954421, + "learning_rate": 9.967031589615485e-06, + "loss": 0.641, + "step": 76 + }, + { + "epoch": 0.0741097208854668, + "grad_norm": 1.4358808126225298, + "learning_rate": 9.966159278460703e-06, + "loss": 0.8949, + "step": 77 + }, + { + "epoch": 0.07507218479307026, + "grad_norm": 1.2513289962250587, + "learning_rate": 9.965275616434068e-06, + "loss": 0.769, + "step": 78 + }, + { + "epoch": 0.07603464870067372, + "grad_norm": 1.4073841440905945, + "learning_rate": 9.964380605555315e-06, + "loss": 0.9322, + "step": 79 + }, + { + "epoch": 0.07699711260827719, + "grad_norm": 1.1138991921513837, + "learning_rate": 9.963474247870122e-06, + "loss": 0.6755, + "step": 80 + }, + { + "epoch": 0.07795957651588066, + "grad_norm": 1.4677341599481075, + "learning_rate": 9.9625565454501e-06, + "loss": 0.9368, + "step": 81 + }, + { + "epoch": 0.07892204042348412, + "grad_norm": 1.2822819787457242, + "learning_rate": 9.961627500392788e-06, + "loss": 0.8018, + "step": 82 + }, + { + "epoch": 0.07988450433108758, + "grad_norm": 1.181232539241254, + "learning_rate": 9.960687114821656e-06, + "loss": 0.8796, + "step": 83 + }, + { + "epoch": 0.08084696823869104, + "grad_norm": 1.1269481606485832, + "learning_rate": 9.959735390886087e-06, + "loss": 0.809, + "step": 84 + }, + { + "epoch": 0.08180943214629452, + "grad_norm": 1.3090500400395901, + "learning_rate": 9.958772330761384e-06, + "loss": 0.8541, + "step": 85 + }, + { + "epoch": 0.08277189605389798, + "grad_norm": 1.1199123066566952, + "learning_rate": 9.957797936648762e-06, + "loss": 0.7709, + "step": 86 + }, + { + "epoch": 0.08373435996150144, + "grad_norm": 1.4500665857110848, + "learning_rate": 9.956812210775337e-06, + "loss": 0.7571, + "step": 87 + }, + { + "epoch": 0.0846968238691049, + "grad_norm": 1.1528952389698748, + "learning_rate": 9.955815155394127e-06, + "loss": 0.7596, + "step": 88 + }, + { + "epoch": 0.08565928777670838, + "grad_norm": 1.3255172516988105, + "learning_rate": 9.954806772784045e-06, + "loss": 0.9284, + "step": 89 + }, + { + "epoch": 0.08662175168431184, + "grad_norm": 1.3044172411526918, + "learning_rate": 9.953787065249896e-06, + "loss": 0.8847, + "step": 90 + }, + { + "epoch": 0.0875842155919153, + "grad_norm": 1.1922397706148844, + "learning_rate": 9.952756035122365e-06, + "loss": 0.7825, + "step": 91 + }, + { + "epoch": 0.08854667949951876, + "grad_norm": 1.0258504780713016, + "learning_rate": 9.951713684758027e-06, + "loss": 0.6362, + "step": 92 + }, + { + "epoch": 0.08950914340712224, + "grad_norm": 1.2641224689552015, + "learning_rate": 9.950660016539318e-06, + "loss": 0.8677, + "step": 93 + }, + { + "epoch": 0.0904716073147257, + "grad_norm": 1.2697289005356154, + "learning_rate": 9.94959503287455e-06, + "loss": 0.8584, + "step": 94 + }, + { + "epoch": 0.09143407122232916, + "grad_norm": 1.2633411826901944, + "learning_rate": 9.948518736197896e-06, + "loss": 0.8204, + "step": 95 + }, + { + "epoch": 0.09239653512993262, + "grad_norm": 1.309701895776113, + "learning_rate": 9.947431128969388e-06, + "loss": 0.8691, + "step": 96 + }, + { + "epoch": 0.0933589990375361, + "grad_norm": 1.1954503367479907, + "learning_rate": 9.946332213674907e-06, + "loss": 0.7871, + "step": 97 + }, + { + "epoch": 0.09432146294513956, + "grad_norm": 1.228197873018699, + "learning_rate": 9.945221992826184e-06, + "loss": 0.7788, + "step": 98 + }, + { + "epoch": 0.09528392685274302, + "grad_norm": 1.2107544780352317, + "learning_rate": 9.944100468960788e-06, + "loss": 0.8356, + "step": 99 + }, + { + "epoch": 0.09624639076034648, + "grad_norm": 1.1258054975336487, + "learning_rate": 9.942967644642122e-06, + "loss": 0.7023, + "step": 100 + }, + { + "epoch": 0.09720885466794996, + "grad_norm": 1.2187292967778287, + "learning_rate": 9.94182352245942e-06, + "loss": 0.8521, + "step": 101 + }, + { + "epoch": 0.09817131857555342, + "grad_norm": 1.0826033936166397, + "learning_rate": 9.940668105027739e-06, + "loss": 0.7299, + "step": 102 + }, + { + "epoch": 0.09913378248315688, + "grad_norm": 1.0180258303211742, + "learning_rate": 9.93950139498795e-06, + "loss": 0.6511, + "step": 103 + }, + { + "epoch": 0.10009624639076034, + "grad_norm": 1.04076738626505, + "learning_rate": 9.938323395006738e-06, + "loss": 0.7128, + "step": 104 + }, + { + "epoch": 0.10105871029836382, + "grad_norm": 1.3039767793024415, + "learning_rate": 9.937134107776592e-06, + "loss": 0.9069, + "step": 105 + }, + { + "epoch": 0.10202117420596728, + "grad_norm": 1.2645466013391258, + "learning_rate": 9.935933536015796e-06, + "loss": 0.9538, + "step": 106 + }, + { + "epoch": 0.10298363811357074, + "grad_norm": 1.1996169194785464, + "learning_rate": 9.934721682468433e-06, + "loss": 0.8422, + "step": 107 + }, + { + "epoch": 0.1039461020211742, + "grad_norm": 1.2696561124832109, + "learning_rate": 9.933498549904367e-06, + "loss": 0.8862, + "step": 108 + }, + { + "epoch": 0.10490856592877768, + "grad_norm": 1.2168608612323184, + "learning_rate": 9.932264141119244e-06, + "loss": 0.8478, + "step": 109 + }, + { + "epoch": 0.10587102983638114, + "grad_norm": 1.2541011900521846, + "learning_rate": 9.931018458934483e-06, + "loss": 0.9576, + "step": 110 + }, + { + "epoch": 0.1068334937439846, + "grad_norm": 1.225389736782274, + "learning_rate": 9.929761506197272e-06, + "loss": 0.762, + "step": 111 + }, + { + "epoch": 0.10779595765158806, + "grad_norm": 1.3968522355132231, + "learning_rate": 9.928493285780554e-06, + "loss": 0.9657, + "step": 112 + }, + { + "epoch": 0.10875842155919153, + "grad_norm": 1.2619357509756692, + "learning_rate": 9.927213800583029e-06, + "loss": 0.93, + "step": 113 + }, + { + "epoch": 0.109720885466795, + "grad_norm": 0.9600838027462066, + "learning_rate": 9.925923053529147e-06, + "loss": 0.6093, + "step": 114 + }, + { + "epoch": 0.11068334937439846, + "grad_norm": 1.1162627084131445, + "learning_rate": 9.924621047569096e-06, + "loss": 0.6713, + "step": 115 + }, + { + "epoch": 0.11164581328200192, + "grad_norm": 1.125467783012257, + "learning_rate": 9.923307785678798e-06, + "loss": 0.7472, + "step": 116 + }, + { + "epoch": 0.1126082771896054, + "grad_norm": 1.1441710475165192, + "learning_rate": 9.9219832708599e-06, + "loss": 0.7858, + "step": 117 + }, + { + "epoch": 0.11357074109720885, + "grad_norm": 1.2882566376495042, + "learning_rate": 9.920647506139774e-06, + "loss": 0.9245, + "step": 118 + }, + { + "epoch": 0.11453320500481232, + "grad_norm": 1.3106741476050527, + "learning_rate": 9.919300494571501e-06, + "loss": 0.8633, + "step": 119 + }, + { + "epoch": 0.11549566891241578, + "grad_norm": 2.788646151114755, + "learning_rate": 9.917942239233872e-06, + "loss": 0.7264, + "step": 120 + }, + { + "epoch": 0.11645813282001925, + "grad_norm": 1.1400375122089337, + "learning_rate": 9.91657274323137e-06, + "loss": 0.7842, + "step": 121 + }, + { + "epoch": 0.11742059672762271, + "grad_norm": 1.1545635288815685, + "learning_rate": 9.91519200969418e-06, + "loss": 0.7227, + "step": 122 + }, + { + "epoch": 0.11838306063522618, + "grad_norm": 1.1407297642795147, + "learning_rate": 9.913800041778165e-06, + "loss": 0.684, + "step": 123 + }, + { + "epoch": 0.11934552454282965, + "grad_norm": 1.0410951528812284, + "learning_rate": 9.912396842664869e-06, + "loss": 0.5811, + "step": 124 + }, + { + "epoch": 0.12030798845043311, + "grad_norm": 1.0811171884992405, + "learning_rate": 9.910982415561506e-06, + "loss": 0.6455, + "step": 125 + }, + { + "epoch": 0.12127045235803657, + "grad_norm": 1.2912451531379068, + "learning_rate": 9.909556763700952e-06, + "loss": 0.7905, + "step": 126 + }, + { + "epoch": 0.12223291626564003, + "grad_norm": 1.314010162958414, + "learning_rate": 9.908119890341738e-06, + "loss": 0.78, + "step": 127 + }, + { + "epoch": 0.12319538017324351, + "grad_norm": 1.405280033518581, + "learning_rate": 9.906671798768048e-06, + "loss": 0.9553, + "step": 128 + }, + { + "epoch": 0.12415784408084697, + "grad_norm": 1.2962597692378253, + "learning_rate": 9.9052124922897e-06, + "loss": 0.9303, + "step": 129 + }, + { + "epoch": 0.12512030798845045, + "grad_norm": 1.3399390240597338, + "learning_rate": 9.903741974242153e-06, + "loss": 0.8331, + "step": 130 + }, + { + "epoch": 0.1260827718960539, + "grad_norm": 1.4463318357692128, + "learning_rate": 9.902260247986485e-06, + "loss": 0.859, + "step": 131 + }, + { + "epoch": 0.12704523580365737, + "grad_norm": 1.1864503874495682, + "learning_rate": 9.900767316909396e-06, + "loss": 0.7894, + "step": 132 + }, + { + "epoch": 0.12800769971126083, + "grad_norm": 1.1860332148175325, + "learning_rate": 9.899263184423195e-06, + "loss": 0.7595, + "step": 133 + }, + { + "epoch": 0.1289701636188643, + "grad_norm": 63.84109198475924, + "learning_rate": 9.897747853965792e-06, + "loss": 1.4116, + "step": 134 + }, + { + "epoch": 0.12993262752646775, + "grad_norm": 1.6237061712013197, + "learning_rate": 9.896221329000693e-06, + "loss": 0.9142, + "step": 135 + }, + { + "epoch": 0.13089509143407121, + "grad_norm": 0.9623090199238966, + "learning_rate": 9.894683613016992e-06, + "loss": 0.4783, + "step": 136 + }, + { + "epoch": 0.13185755534167468, + "grad_norm": 1.3890587980327496, + "learning_rate": 9.893134709529359e-06, + "loss": 0.8261, + "step": 137 + }, + { + "epoch": 0.13282001924927817, + "grad_norm": 1.3247131944950392, + "learning_rate": 9.891574622078035e-06, + "loss": 0.9349, + "step": 138 + }, + { + "epoch": 0.13378248315688163, + "grad_norm": 1.1815501720091948, + "learning_rate": 9.890003354228828e-06, + "loss": 0.7429, + "step": 139 + }, + { + "epoch": 0.1347449470644851, + "grad_norm": 1.4844000198295566, + "learning_rate": 9.888420909573093e-06, + "loss": 0.8618, + "step": 140 + }, + { + "epoch": 0.13570741097208855, + "grad_norm": 1.3093885534086986, + "learning_rate": 9.886827291727736e-06, + "loss": 0.8573, + "step": 141 + }, + { + "epoch": 0.136669874879692, + "grad_norm": 1.4016223406076316, + "learning_rate": 9.885222504335199e-06, + "loss": 0.9632, + "step": 142 + }, + { + "epoch": 0.13763233878729547, + "grad_norm": 1.4216446890130563, + "learning_rate": 9.883606551063458e-06, + "loss": 0.9697, + "step": 143 + }, + { + "epoch": 0.13859480269489893, + "grad_norm": 1.3560158281329175, + "learning_rate": 9.881979435606004e-06, + "loss": 0.9187, + "step": 144 + }, + { + "epoch": 0.1395572666025024, + "grad_norm": 1.3540995937522022, + "learning_rate": 9.880341161681845e-06, + "loss": 0.8675, + "step": 145 + }, + { + "epoch": 0.14051973051010588, + "grad_norm": 1.4933670817006048, + "learning_rate": 9.87869173303549e-06, + "loss": 0.9233, + "step": 146 + }, + { + "epoch": 0.14148219441770934, + "grad_norm": 1.4357186513453173, + "learning_rate": 9.877031153436949e-06, + "loss": 0.9258, + "step": 147 + }, + { + "epoch": 0.1424446583253128, + "grad_norm": 1.2126822844366585, + "learning_rate": 9.875359426681712e-06, + "loss": 0.8398, + "step": 148 + }, + { + "epoch": 0.14340712223291627, + "grad_norm": 1.1241560347515867, + "learning_rate": 9.873676556590751e-06, + "loss": 0.7622, + "step": 149 + }, + { + "epoch": 0.14436958614051973, + "grad_norm": 1.353357945858469, + "learning_rate": 9.871982547010512e-06, + "loss": 0.9031, + "step": 150 + }, + { + "epoch": 0.1453320500481232, + "grad_norm": 1.117685918251694, + "learning_rate": 9.870277401812896e-06, + "loss": 0.6801, + "step": 151 + }, + { + "epoch": 0.14629451395572665, + "grad_norm": 0.8249826916763036, + "learning_rate": 9.868561124895254e-06, + "loss": 0.4813, + "step": 152 + }, + { + "epoch": 0.1472569778633301, + "grad_norm": 1.2719775934434754, + "learning_rate": 9.866833720180389e-06, + "loss": 0.9139, + "step": 153 + }, + { + "epoch": 0.1482194417709336, + "grad_norm": 1.1489239141759262, + "learning_rate": 9.86509519161653e-06, + "loss": 0.7821, + "step": 154 + }, + { + "epoch": 0.14918190567853706, + "grad_norm": 1.294995239428847, + "learning_rate": 9.863345543177335e-06, + "loss": 0.9164, + "step": 155 + }, + { + "epoch": 0.15014436958614052, + "grad_norm": 1.1452834210779264, + "learning_rate": 9.861584778861879e-06, + "loss": 0.8438, + "step": 156 + }, + { + "epoch": 0.15110683349374399, + "grad_norm": 114.3109437868852, + "learning_rate": 9.85981290269464e-06, + "loss": 0.9679, + "step": 157 + }, + { + "epoch": 0.15206929740134745, + "grad_norm": 1.2660809065810192, + "learning_rate": 9.858029918725496e-06, + "loss": 0.8251, + "step": 158 + }, + { + "epoch": 0.1530317613089509, + "grad_norm": 1.1715622077614172, + "learning_rate": 9.856235831029716e-06, + "loss": 0.7778, + "step": 159 + }, + { + "epoch": 0.15399422521655437, + "grad_norm": 1.1635665603311665, + "learning_rate": 9.854430643707944e-06, + "loss": 0.757, + "step": 160 + }, + { + "epoch": 0.15495668912415783, + "grad_norm": 1.1310616849479287, + "learning_rate": 9.852614360886195e-06, + "loss": 0.8622, + "step": 161 + }, + { + "epoch": 0.15591915303176132, + "grad_norm": 1.2479512654565674, + "learning_rate": 9.850786986715845e-06, + "loss": 0.9214, + "step": 162 + }, + { + "epoch": 0.15688161693936478, + "grad_norm": 1.194648765869352, + "learning_rate": 9.848948525373623e-06, + "loss": 0.8699, + "step": 163 + }, + { + "epoch": 0.15784408084696824, + "grad_norm": 1.202969021045747, + "learning_rate": 9.847098981061593e-06, + "loss": 0.8225, + "step": 164 + }, + { + "epoch": 0.1588065447545717, + "grad_norm": 1.3986972549093302, + "learning_rate": 9.845238358007158e-06, + "loss": 0.9393, + "step": 165 + }, + { + "epoch": 0.15976900866217517, + "grad_norm": 0.9438889810886896, + "learning_rate": 9.84336666046304e-06, + "loss": 0.5813, + "step": 166 + }, + { + "epoch": 0.16073147256977863, + "grad_norm": 1.2443186646471902, + "learning_rate": 9.841483892707267e-06, + "loss": 0.8433, + "step": 167 + }, + { + "epoch": 0.1616939364773821, + "grad_norm": 1.0455934798147948, + "learning_rate": 9.839590059043185e-06, + "loss": 0.7368, + "step": 168 + }, + { + "epoch": 0.16265640038498555, + "grad_norm": 1.2472352194576524, + "learning_rate": 9.837685163799417e-06, + "loss": 0.8551, + "step": 169 + }, + { + "epoch": 0.16361886429258904, + "grad_norm": 1.1519869555998399, + "learning_rate": 9.835769211329877e-06, + "loss": 0.7781, + "step": 170 + }, + { + "epoch": 0.1645813282001925, + "grad_norm": 1.1684689436287756, + "learning_rate": 9.833842206013751e-06, + "loss": 0.7665, + "step": 171 + }, + { + "epoch": 0.16554379210779596, + "grad_norm": 1.206368919675947, + "learning_rate": 9.831904152255488e-06, + "loss": 0.8427, + "step": 172 + }, + { + "epoch": 0.16650625601539942, + "grad_norm": 1.1839324381841279, + "learning_rate": 9.829955054484785e-06, + "loss": 0.6807, + "step": 173 + }, + { + "epoch": 0.16746871992300288, + "grad_norm": 1.198791751028212, + "learning_rate": 9.827994917156587e-06, + "loss": 0.8287, + "step": 174 + }, + { + "epoch": 0.16843118383060635, + "grad_norm": 1.3073382580111497, + "learning_rate": 9.826023744751073e-06, + "loss": 0.9365, + "step": 175 + }, + { + "epoch": 0.1693936477382098, + "grad_norm": 1.6047058449053848, + "learning_rate": 9.824041541773639e-06, + "loss": 0.8147, + "step": 176 + }, + { + "epoch": 0.17035611164581327, + "grad_norm": 1.2997343858917745, + "learning_rate": 9.822048312754894e-06, + "loss": 0.921, + "step": 177 + }, + { + "epoch": 0.17131857555341676, + "grad_norm": 1.2522710324024182, + "learning_rate": 9.820044062250651e-06, + "loss": 0.8083, + "step": 178 + }, + { + "epoch": 0.17228103946102022, + "grad_norm": 1.2521672269027646, + "learning_rate": 9.818028794841914e-06, + "loss": 0.9057, + "step": 179 + }, + { + "epoch": 0.17324350336862368, + "grad_norm": 1.2587551532017769, + "learning_rate": 9.816002515134865e-06, + "loss": 0.9095, + "step": 180 + }, + { + "epoch": 0.17420596727622714, + "grad_norm": 1.3616264853774362, + "learning_rate": 9.813965227760858e-06, + "loss": 0.9131, + "step": 181 + }, + { + "epoch": 0.1751684311838306, + "grad_norm": 1.2399713853023682, + "learning_rate": 9.81191693737641e-06, + "loss": 0.8743, + "step": 182 + }, + { + "epoch": 0.17613089509143406, + "grad_norm": 1.282255664593352, + "learning_rate": 9.809857648663177e-06, + "loss": 0.8093, + "step": 183 + }, + { + "epoch": 0.17709335899903753, + "grad_norm": 1.257054531843103, + "learning_rate": 9.807787366327966e-06, + "loss": 0.7816, + "step": 184 + }, + { + "epoch": 0.17805582290664101, + "grad_norm": 1.2796808194767766, + "learning_rate": 9.805706095102702e-06, + "loss": 0.8743, + "step": 185 + }, + { + "epoch": 0.17901828681424448, + "grad_norm": 1.1748065801807128, + "learning_rate": 9.80361383974443e-06, + "loss": 0.7913, + "step": 186 + }, + { + "epoch": 0.17998075072184794, + "grad_norm": 1.1384844869131308, + "learning_rate": 9.801510605035304e-06, + "loss": 0.6898, + "step": 187 + }, + { + "epoch": 0.1809432146294514, + "grad_norm": 1.3004865175653104, + "learning_rate": 9.799396395782566e-06, + "loss": 0.9216, + "step": 188 + }, + { + "epoch": 0.18190567853705486, + "grad_norm": 1.2229202665965897, + "learning_rate": 9.797271216818546e-06, + "loss": 0.8502, + "step": 189 + }, + { + "epoch": 0.18286814244465832, + "grad_norm": 1.3140239366107738, + "learning_rate": 9.795135073000648e-06, + "loss": 0.9457, + "step": 190 + }, + { + "epoch": 0.18383060635226178, + "grad_norm": 1.1720877758843193, + "learning_rate": 9.792987969211335e-06, + "loss": 0.8419, + "step": 191 + }, + { + "epoch": 0.18479307025986524, + "grad_norm": 0.9748925604289701, + "learning_rate": 9.790829910358122e-06, + "loss": 0.5835, + "step": 192 + }, + { + "epoch": 0.18575553416746873, + "grad_norm": 1.1257694245474839, + "learning_rate": 9.788660901373562e-06, + "loss": 0.6555, + "step": 193 + }, + { + "epoch": 0.1867179980750722, + "grad_norm": 1.2087140063549708, + "learning_rate": 9.78648094721524e-06, + "loss": 0.8571, + "step": 194 + }, + { + "epoch": 0.18768046198267566, + "grad_norm": 1.2553159223807955, + "learning_rate": 9.784290052865754e-06, + "loss": 0.8268, + "step": 195 + }, + { + "epoch": 0.18864292589027912, + "grad_norm": 0.9646555410257132, + "learning_rate": 9.782088223332706e-06, + "loss": 0.5738, + "step": 196 + }, + { + "epoch": 0.18960538979788258, + "grad_norm": 1.0823951006286359, + "learning_rate": 9.779875463648697e-06, + "loss": 0.7418, + "step": 197 + }, + { + "epoch": 0.19056785370548604, + "grad_norm": 1.271151378523244, + "learning_rate": 9.777651778871309e-06, + "loss": 0.8247, + "step": 198 + }, + { + "epoch": 0.1915303176130895, + "grad_norm": 1.2916049633248714, + "learning_rate": 9.775417174083088e-06, + "loss": 0.8132, + "step": 199 + }, + { + "epoch": 0.19249278152069296, + "grad_norm": 1.1765064905030775, + "learning_rate": 9.773171654391552e-06, + "loss": 0.7661, + "step": 200 + }, + { + "epoch": 0.19345524542829645, + "grad_norm": 1.167753829883933, + "learning_rate": 9.770915224929155e-06, + "loss": 0.7915, + "step": 201 + }, + { + "epoch": 0.1944177093358999, + "grad_norm": 1.1173433070450232, + "learning_rate": 9.76864789085329e-06, + "loss": 0.7581, + "step": 202 + }, + { + "epoch": 0.19538017324350337, + "grad_norm": 1.3245986477141838, + "learning_rate": 9.76636965734628e-06, + "loss": 0.8363, + "step": 203 + }, + { + "epoch": 0.19634263715110684, + "grad_norm": 1.2313640175890157, + "learning_rate": 9.764080529615352e-06, + "loss": 0.8465, + "step": 204 + }, + { + "epoch": 0.1973051010587103, + "grad_norm": 1.220075015600601, + "learning_rate": 9.76178051289264e-06, + "loss": 0.837, + "step": 205 + }, + { + "epoch": 0.19826756496631376, + "grad_norm": 1.2978796773234405, + "learning_rate": 9.759469612435159e-06, + "loss": 0.8438, + "step": 206 + }, + { + "epoch": 0.19923002887391722, + "grad_norm": 1.2273253837694793, + "learning_rate": 9.75714783352481e-06, + "loss": 0.7887, + "step": 207 + }, + { + "epoch": 0.20019249278152068, + "grad_norm": 1.2930425780673338, + "learning_rate": 9.754815181468347e-06, + "loss": 0.8933, + "step": 208 + }, + { + "epoch": 0.20115495668912417, + "grad_norm": 1.2248577156817855, + "learning_rate": 9.752471661597387e-06, + "loss": 0.7792, + "step": 209 + }, + { + "epoch": 0.20211742059672763, + "grad_norm": 7.842496938362841, + "learning_rate": 9.750117279268379e-06, + "loss": 0.7748, + "step": 210 + }, + { + "epoch": 0.2030798845043311, + "grad_norm": 0.9608872803428284, + "learning_rate": 9.747752039862604e-06, + "loss": 0.5305, + "step": 211 + }, + { + "epoch": 0.20404234841193455, + "grad_norm": 1.40558476553411, + "learning_rate": 9.745375948786158e-06, + "loss": 0.8367, + "step": 212 + }, + { + "epoch": 0.20500481231953802, + "grad_norm": 1.203341695161007, + "learning_rate": 9.742989011469937e-06, + "loss": 0.7821, + "step": 213 + }, + { + "epoch": 0.20596727622714148, + "grad_norm": 1.1829317945066327, + "learning_rate": 9.74059123336963e-06, + "loss": 0.8275, + "step": 214 + }, + { + "epoch": 0.20692974013474494, + "grad_norm": 2.315815444853789, + "learning_rate": 9.738182619965706e-06, + "loss": 0.7377, + "step": 215 + }, + { + "epoch": 0.2078922040423484, + "grad_norm": 7.220837056921597, + "learning_rate": 9.735763176763395e-06, + "loss": 0.7153, + "step": 216 + }, + { + "epoch": 0.2088546679499519, + "grad_norm": 1.1930786316892121, + "learning_rate": 9.733332909292686e-06, + "loss": 0.6675, + "step": 217 + }, + { + "epoch": 0.20981713185755535, + "grad_norm": 1.1500660146021693, + "learning_rate": 9.730891823108302e-06, + "loss": 0.6612, + "step": 218 + }, + { + "epoch": 0.2107795957651588, + "grad_norm": 1.2513153951104516, + "learning_rate": 9.728439923789697e-06, + "loss": 0.8984, + "step": 219 + }, + { + "epoch": 0.21174205967276227, + "grad_norm": 1.3307091662088406, + "learning_rate": 9.725977216941044e-06, + "loss": 0.8869, + "step": 220 + }, + { + "epoch": 0.21270452358036573, + "grad_norm": 1.2597916397562658, + "learning_rate": 9.723503708191209e-06, + "loss": 0.8977, + "step": 221 + }, + { + "epoch": 0.2136669874879692, + "grad_norm": 1.7955041692343559, + "learning_rate": 9.721019403193754e-06, + "loss": 0.8271, + "step": 222 + }, + { + "epoch": 0.21462945139557266, + "grad_norm": 1.1096459326410524, + "learning_rate": 9.718524307626916e-06, + "loss": 0.6339, + "step": 223 + }, + { + "epoch": 0.21559191530317612, + "grad_norm": 1.1931855075342475, + "learning_rate": 9.716018427193593e-06, + "loss": 0.7054, + "step": 224 + }, + { + "epoch": 0.2165543792107796, + "grad_norm": 1.1197004281387106, + "learning_rate": 9.71350176762134e-06, + "loss": 0.7177, + "step": 225 + }, + { + "epoch": 0.21751684311838307, + "grad_norm": 1.6114292162769497, + "learning_rate": 9.710974334662343e-06, + "loss": 0.8378, + "step": 226 + }, + { + "epoch": 0.21847930702598653, + "grad_norm": 1.1927323194742112, + "learning_rate": 9.708436134093408e-06, + "loss": 0.714, + "step": 227 + }, + { + "epoch": 0.21944177093359, + "grad_norm": 1.1127256517422952, + "learning_rate": 9.705887171715966e-06, + "loss": 0.7263, + "step": 228 + }, + { + "epoch": 0.22040423484119345, + "grad_norm": 1.1897367269753107, + "learning_rate": 9.703327453356034e-06, + "loss": 0.8268, + "step": 229 + }, + { + "epoch": 0.22136669874879691, + "grad_norm": 12.57935555567707, + "learning_rate": 9.700756984864217e-06, + "loss": 0.8225, + "step": 230 + }, + { + "epoch": 0.22232916265640038, + "grad_norm": 1.1505345929974828, + "learning_rate": 9.698175772115688e-06, + "loss": 0.8147, + "step": 231 + }, + { + "epoch": 0.22329162656400384, + "grad_norm": 1.1772400794741453, + "learning_rate": 9.695583821010185e-06, + "loss": 0.7356, + "step": 232 + }, + { + "epoch": 0.22425409047160733, + "grad_norm": 1.1987350009582605, + "learning_rate": 9.692981137471982e-06, + "loss": 0.8142, + "step": 233 + }, + { + "epoch": 0.2252165543792108, + "grad_norm": 1.004979719938657, + "learning_rate": 9.690367727449889e-06, + "loss": 0.6209, + "step": 234 + }, + { + "epoch": 0.22617901828681425, + "grad_norm": 1.0579746880384042, + "learning_rate": 9.687743596917228e-06, + "loss": 0.6729, + "step": 235 + }, + { + "epoch": 0.2271414821944177, + "grad_norm": 1.3435547156712695, + "learning_rate": 9.68510875187183e-06, + "loss": 0.9018, + "step": 236 + }, + { + "epoch": 0.22810394610202117, + "grad_norm": 1.1392221935483398, + "learning_rate": 9.68246319833601e-06, + "loss": 0.7761, + "step": 237 + }, + { + "epoch": 0.22906641000962463, + "grad_norm": 12.523271619337562, + "learning_rate": 9.679806942356564e-06, + "loss": 0.858, + "step": 238 + }, + { + "epoch": 0.2300288739172281, + "grad_norm": 1.1365046487654749, + "learning_rate": 9.677139990004747e-06, + "loss": 0.6733, + "step": 239 + }, + { + "epoch": 0.23099133782483156, + "grad_norm": 1.3855355931558864, + "learning_rate": 9.674462347376258e-06, + "loss": 0.9169, + "step": 240 + }, + { + "epoch": 0.23195380173243504, + "grad_norm": 1.2443517469026233, + "learning_rate": 9.671774020591242e-06, + "loss": 0.8335, + "step": 241 + }, + { + "epoch": 0.2329162656400385, + "grad_norm": 1.1354887313233766, + "learning_rate": 9.669075015794251e-06, + "loss": 0.7717, + "step": 242 + }, + { + "epoch": 0.23387872954764197, + "grad_norm": 5.7682553258913485, + "learning_rate": 9.666365339154252e-06, + "loss": 0.7185, + "step": 243 + }, + { + "epoch": 0.23484119345524543, + "grad_norm": 1.3341366047888865, + "learning_rate": 9.663644996864599e-06, + "loss": 0.8101, + "step": 244 + }, + { + "epoch": 0.2358036573628489, + "grad_norm": 1.2492098404247336, + "learning_rate": 9.66091399514303e-06, + "loss": 0.8566, + "step": 245 + }, + { + "epoch": 0.23676612127045235, + "grad_norm": 1.15685435772373, + "learning_rate": 9.658172340231636e-06, + "loss": 0.8363, + "step": 246 + }, + { + "epoch": 0.2377285851780558, + "grad_norm": 1.1550861179206524, + "learning_rate": 9.655420038396868e-06, + "loss": 0.7258, + "step": 247 + }, + { + "epoch": 0.2386910490856593, + "grad_norm": 1.3620098513221588, + "learning_rate": 9.652657095929508e-06, + "loss": 0.9455, + "step": 248 + }, + { + "epoch": 0.23965351299326276, + "grad_norm": 1.1319567677296811, + "learning_rate": 9.649883519144656e-06, + "loss": 0.7792, + "step": 249 + }, + { + "epoch": 0.24061597690086622, + "grad_norm": 1.2068229040675968, + "learning_rate": 9.647099314381725e-06, + "loss": 0.9408, + "step": 250 + }, + { + "epoch": 0.24157844080846969, + "grad_norm": 1.1859969457940316, + "learning_rate": 9.644304488004412e-06, + "loss": 0.8007, + "step": 251 + }, + { + "epoch": 0.24254090471607315, + "grad_norm": 1.1423726803247123, + "learning_rate": 9.6414990464007e-06, + "loss": 0.7328, + "step": 252 + }, + { + "epoch": 0.2435033686236766, + "grad_norm": 1.2132108303405051, + "learning_rate": 9.638682995982822e-06, + "loss": 0.8748, + "step": 253 + }, + { + "epoch": 0.24446583253128007, + "grad_norm": 1.3138437773635407, + "learning_rate": 9.635856343187274e-06, + "loss": 0.9531, + "step": 254 + }, + { + "epoch": 0.24542829643888353, + "grad_norm": 1.1604864022636086, + "learning_rate": 9.633019094474773e-06, + "loss": 0.7039, + "step": 255 + }, + { + "epoch": 0.24639076034648702, + "grad_norm": 1.123618747124324, + "learning_rate": 9.630171256330263e-06, + "loss": 0.7558, + "step": 256 + }, + { + "epoch": 0.24735322425409048, + "grad_norm": 1.2391819430695652, + "learning_rate": 9.627312835262885e-06, + "loss": 0.8222, + "step": 257 + }, + { + "epoch": 0.24831568816169394, + "grad_norm": 1.2182214119312005, + "learning_rate": 9.624443837805972e-06, + "loss": 0.6436, + "step": 258 + }, + { + "epoch": 0.2492781520692974, + "grad_norm": 7.28838414931088, + "learning_rate": 9.621564270517032e-06, + "loss": 0.8005, + "step": 259 + }, + { + "epoch": 0.2502406159769009, + "grad_norm": 1.1027467795500523, + "learning_rate": 9.618674139977731e-06, + "loss": 0.7268, + "step": 260 + }, + { + "epoch": 0.25120307988450435, + "grad_norm": 1.2110258566012382, + "learning_rate": 9.615773452793876e-06, + "loss": 0.8224, + "step": 261 + }, + { + "epoch": 0.2521655437921078, + "grad_norm": 1.2172625290855361, + "learning_rate": 9.612862215595406e-06, + "loss": 0.8426, + "step": 262 + }, + { + "epoch": 0.2531280076997113, + "grad_norm": 1.1006422759628531, + "learning_rate": 9.609940435036374e-06, + "loss": 0.7103, + "step": 263 + }, + { + "epoch": 0.25409047160731474, + "grad_norm": 1.0162845129383353, + "learning_rate": 9.607008117794929e-06, + "loss": 0.5754, + "step": 264 + }, + { + "epoch": 0.2550529355149182, + "grad_norm": 13.174746678511346, + "learning_rate": 9.604065270573304e-06, + "loss": 0.7302, + "step": 265 + }, + { + "epoch": 0.25601539942252166, + "grad_norm": 1.162287359883351, + "learning_rate": 9.6011119000978e-06, + "loss": 0.7879, + "step": 266 + }, + { + "epoch": 0.2569778633301251, + "grad_norm": 1.2163008865042435, + "learning_rate": 9.59814801311877e-06, + "loss": 0.8482, + "step": 267 + }, + { + "epoch": 0.2579403272377286, + "grad_norm": 1.9816207537679449, + "learning_rate": 9.595173616410606e-06, + "loss": 0.7748, + "step": 268 + }, + { + "epoch": 0.25890279114533205, + "grad_norm": 1.1925643259228529, + "learning_rate": 9.59218871677172e-06, + "loss": 0.7928, + "step": 269 + }, + { + "epoch": 0.2598652550529355, + "grad_norm": 1.3078473008000486, + "learning_rate": 9.589193321024525e-06, + "loss": 0.8503, + "step": 270 + }, + { + "epoch": 0.26082771896053897, + "grad_norm": 1.2234407082336847, + "learning_rate": 9.586187436015433e-06, + "loss": 0.9107, + "step": 271 + }, + { + "epoch": 0.26179018286814243, + "grad_norm": 1.1581967839367917, + "learning_rate": 9.583171068614828e-06, + "loss": 0.8151, + "step": 272 + }, + { + "epoch": 0.2627526467757459, + "grad_norm": 1.333182929405783, + "learning_rate": 9.58014422571705e-06, + "loss": 0.9335, + "step": 273 + }, + { + "epoch": 0.26371511068334935, + "grad_norm": 1.1132884102057583, + "learning_rate": 9.577106914240383e-06, + "loss": 0.712, + "step": 274 + }, + { + "epoch": 0.2646775745909528, + "grad_norm": 2.0794232196047533, + "learning_rate": 9.57405914112704e-06, + "loss": 0.6567, + "step": 275 + }, + { + "epoch": 0.26564003849855633, + "grad_norm": 1.2401285547269145, + "learning_rate": 9.571000913343149e-06, + "loss": 0.8209, + "step": 276 + }, + { + "epoch": 0.2666025024061598, + "grad_norm": 1.1739931736778797, + "learning_rate": 9.567932237878725e-06, + "loss": 0.8019, + "step": 277 + }, + { + "epoch": 0.26756496631376325, + "grad_norm": 1.217185577815541, + "learning_rate": 9.56485312174767e-06, + "loss": 0.8364, + "step": 278 + }, + { + "epoch": 0.2685274302213667, + "grad_norm": 1.1564634898656863, + "learning_rate": 9.561763571987745e-06, + "loss": 0.6822, + "step": 279 + }, + { + "epoch": 0.2694898941289702, + "grad_norm": 1.0323795365294182, + "learning_rate": 9.558663595660563e-06, + "loss": 0.651, + "step": 280 + }, + { + "epoch": 0.27045235803657364, + "grad_norm": 1.1960984562898658, + "learning_rate": 9.555553199851566e-06, + "loss": 0.7415, + "step": 281 + }, + { + "epoch": 0.2714148219441771, + "grad_norm": 1.031153048581712, + "learning_rate": 9.55243239167001e-06, + "loss": 0.6151, + "step": 282 + }, + { + "epoch": 0.27237728585178056, + "grad_norm": 1.2123126058292175, + "learning_rate": 9.54930117824895e-06, + "loss": 0.8621, + "step": 283 + }, + { + "epoch": 0.273339749759384, + "grad_norm": 1.2108259346791428, + "learning_rate": 9.54615956674523e-06, + "loss": 0.7878, + "step": 284 + }, + { + "epoch": 0.2743022136669875, + "grad_norm": 1.1084651827735201, + "learning_rate": 9.54300756433945e-06, + "loss": 0.7114, + "step": 285 + }, + { + "epoch": 0.27526467757459094, + "grad_norm": 1.3184061060889731, + "learning_rate": 9.539845178235967e-06, + "loss": 0.8464, + "step": 286 + }, + { + "epoch": 0.2762271414821944, + "grad_norm": 2.095604159409325, + "learning_rate": 9.53667241566287e-06, + "loss": 0.5897, + "step": 287 + }, + { + "epoch": 0.27718960538979787, + "grad_norm": 1.2324506871308079, + "learning_rate": 9.533489283871967e-06, + "loss": 0.834, + "step": 288 + }, + { + "epoch": 0.2781520692974013, + "grad_norm": 1.2804399267926678, + "learning_rate": 9.53029579013876e-06, + "loss": 0.9153, + "step": 289 + }, + { + "epoch": 0.2791145332050048, + "grad_norm": 1.0020120528216259, + "learning_rate": 9.527091941762438e-06, + "loss": 0.637, + "step": 290 + }, + { + "epoch": 0.28007699711260825, + "grad_norm": 1.232122611016358, + "learning_rate": 9.523877746065862e-06, + "loss": 0.8074, + "step": 291 + }, + { + "epoch": 0.28103946102021177, + "grad_norm": 1.1186940898273359, + "learning_rate": 9.520653210395534e-06, + "loss": 0.7474, + "step": 292 + }, + { + "epoch": 0.28200192492781523, + "grad_norm": 1.1088004710411263, + "learning_rate": 9.517418342121599e-06, + "loss": 0.7354, + "step": 293 + }, + { + "epoch": 0.2829643888354187, + "grad_norm": 1.4582401709899113, + "learning_rate": 9.514173148637811e-06, + "loss": 0.7002, + "step": 294 + }, + { + "epoch": 0.28392685274302215, + "grad_norm": 0.9255468890090981, + "learning_rate": 9.510917637361529e-06, + "loss": 0.5586, + "step": 295 + }, + { + "epoch": 0.2848893166506256, + "grad_norm": 1.270840512305476, + "learning_rate": 9.507651815733689e-06, + "loss": 0.8536, + "step": 296 + }, + { + "epoch": 0.2858517805582291, + "grad_norm": 0.9856654106965975, + "learning_rate": 9.5043756912188e-06, + "loss": 0.6208, + "step": 297 + }, + { + "epoch": 0.28681424446583254, + "grad_norm": 1.2907029948745112, + "learning_rate": 9.501089271304917e-06, + "loss": 0.9414, + "step": 298 + }, + { + "epoch": 0.287776708373436, + "grad_norm": 1.1503562899543514, + "learning_rate": 9.49779256350362e-06, + "loss": 0.7461, + "step": 299 + }, + { + "epoch": 0.28873917228103946, + "grad_norm": 1.154473535497434, + "learning_rate": 9.494485575350015e-06, + "loss": 0.7518, + "step": 300 + }, + { + "epoch": 0.2897016361886429, + "grad_norm": 1.118274261200585, + "learning_rate": 9.491168314402696e-06, + "loss": 0.7295, + "step": 301 + }, + { + "epoch": 0.2906641000962464, + "grad_norm": 1.3696439172121924, + "learning_rate": 9.487840788243743e-06, + "loss": 0.9364, + "step": 302 + }, + { + "epoch": 0.29162656400384984, + "grad_norm": 1.251864770735364, + "learning_rate": 9.484503004478695e-06, + "loss": 0.8512, + "step": 303 + }, + { + "epoch": 0.2925890279114533, + "grad_norm": 1.2878209959697082, + "learning_rate": 9.481154970736537e-06, + "loss": 0.8583, + "step": 304 + }, + { + "epoch": 0.29355149181905676, + "grad_norm": 1.3695890470709353, + "learning_rate": 9.47779669466968e-06, + "loss": 0.9503, + "step": 305 + }, + { + "epoch": 0.2945139557266602, + "grad_norm": 1.1272519750105674, + "learning_rate": 9.474428183953952e-06, + "loss": 0.8224, + "step": 306 + }, + { + "epoch": 0.29547641963426374, + "grad_norm": 1.111823263088695, + "learning_rate": 9.471049446288563e-06, + "loss": 0.6993, + "step": 307 + }, + { + "epoch": 0.2964388835418672, + "grad_norm": 1.2828648321233644, + "learning_rate": 9.467660489396111e-06, + "loss": 0.8506, + "step": 308 + }, + { + "epoch": 0.29740134744947067, + "grad_norm": 1.0203060278507559, + "learning_rate": 9.46426132102254e-06, + "loss": 0.6427, + "step": 309 + }, + { + "epoch": 0.2983638113570741, + "grad_norm": 1.175958598513402, + "learning_rate": 9.46085194893714e-06, + "loss": 0.809, + "step": 310 + }, + { + "epoch": 0.2993262752646776, + "grad_norm": 1.0916281990755645, + "learning_rate": 9.45743238093252e-06, + "loss": 0.6989, + "step": 311 + }, + { + "epoch": 0.30028873917228105, + "grad_norm": 1.2738515157030694, + "learning_rate": 9.454002624824599e-06, + "loss": 0.9062, + "step": 312 + }, + { + "epoch": 0.3012512030798845, + "grad_norm": 0.905151022278796, + "learning_rate": 9.450562688452576e-06, + "loss": 0.5077, + "step": 313 + }, + { + "epoch": 0.30221366698748797, + "grad_norm": 1.2254992604042947, + "learning_rate": 9.44711257967892e-06, + "loss": 0.8369, + "step": 314 + }, + { + "epoch": 0.30317613089509143, + "grad_norm": 1.2184999135447154, + "learning_rate": 9.44365230638935e-06, + "loss": 0.8032, + "step": 315 + }, + { + "epoch": 0.3041385948026949, + "grad_norm": 1.3246102379927964, + "learning_rate": 9.44018187649282e-06, + "loss": 0.9112, + "step": 316 + }, + { + "epoch": 0.30510105871029836, + "grad_norm": 1.1030550832595363, + "learning_rate": 9.4367012979215e-06, + "loss": 0.7319, + "step": 317 + }, + { + "epoch": 0.3060635226179018, + "grad_norm": 1.1119174364306863, + "learning_rate": 9.43321057863075e-06, + "loss": 0.7358, + "step": 318 + }, + { + "epoch": 0.3070259865255053, + "grad_norm": 1.0829816892446624, + "learning_rate": 9.429709726599113e-06, + "loss": 0.7291, + "step": 319 + }, + { + "epoch": 0.30798845043310874, + "grad_norm": 1.094560688066654, + "learning_rate": 9.426198749828287e-06, + "loss": 0.6975, + "step": 320 + }, + { + "epoch": 0.3089509143407122, + "grad_norm": 0.9115627059781278, + "learning_rate": 9.422677656343118e-06, + "loss": 0.5604, + "step": 321 + }, + { + "epoch": 0.30991337824831566, + "grad_norm": 1.0643495431814285, + "learning_rate": 9.419146454191572e-06, + "loss": 0.6531, + "step": 322 + }, + { + "epoch": 0.3108758421559192, + "grad_norm": 1.04559541756201, + "learning_rate": 9.415605151444718e-06, + "loss": 0.7205, + "step": 323 + }, + { + "epoch": 0.31183830606352264, + "grad_norm": 1.0937596037186783, + "learning_rate": 9.412053756196715e-06, + "loss": 0.7147, + "step": 324 + }, + { + "epoch": 0.3128007699711261, + "grad_norm": 1.1392462397625058, + "learning_rate": 9.408492276564786e-06, + "loss": 0.7784, + "step": 325 + }, + { + "epoch": 0.31376323387872956, + "grad_norm": 1.1074013808762415, + "learning_rate": 9.404920720689208e-06, + "loss": 0.7332, + "step": 326 + }, + { + "epoch": 0.314725697786333, + "grad_norm": 1.3153712288542112, + "learning_rate": 9.401339096733285e-06, + "loss": 0.9241, + "step": 327 + }, + { + "epoch": 0.3156881616939365, + "grad_norm": 1.2074733215626232, + "learning_rate": 9.397747412883332e-06, + "loss": 0.8684, + "step": 328 + }, + { + "epoch": 0.31665062560153995, + "grad_norm": 1.2414243467706674, + "learning_rate": 9.39414567734866e-06, + "loss": 0.8366, + "step": 329 + }, + { + "epoch": 0.3176130895091434, + "grad_norm": 1.1096806870294706, + "learning_rate": 9.390533898361556e-06, + "loss": 0.8016, + "step": 330 + }, + { + "epoch": 0.31857555341674687, + "grad_norm": 1.2646629391181283, + "learning_rate": 9.38691208417726e-06, + "loss": 0.833, + "step": 331 + }, + { + "epoch": 0.31953801732435033, + "grad_norm": 1.2103950571346636, + "learning_rate": 9.383280243073949e-06, + "loss": 0.7928, + "step": 332 + }, + { + "epoch": 0.3205004812319538, + "grad_norm": 1.1356145902132118, + "learning_rate": 9.379638383352719e-06, + "loss": 0.7274, + "step": 333 + }, + { + "epoch": 0.32146294513955725, + "grad_norm": 1.2172814754131325, + "learning_rate": 9.375986513337565e-06, + "loss": 0.8084, + "step": 334 + }, + { + "epoch": 0.3224254090471607, + "grad_norm": 1.2369486406709926, + "learning_rate": 9.372324641375361e-06, + "loss": 0.8358, + "step": 335 + }, + { + "epoch": 0.3233878729547642, + "grad_norm": 1.242602576486121, + "learning_rate": 9.368652775835843e-06, + "loss": 0.824, + "step": 336 + }, + { + "epoch": 0.32435033686236764, + "grad_norm": 1.209715018249487, + "learning_rate": 9.364970925111588e-06, + "loss": 0.8128, + "step": 337 + }, + { + "epoch": 0.3253128007699711, + "grad_norm": 1.2102988445322367, + "learning_rate": 9.361279097617993e-06, + "loss": 0.9045, + "step": 338 + }, + { + "epoch": 0.3262752646775746, + "grad_norm": 1.2068565245678293, + "learning_rate": 9.357577301793266e-06, + "loss": 0.8149, + "step": 339 + }, + { + "epoch": 0.3272377285851781, + "grad_norm": 1.1507560506292864, + "learning_rate": 9.353865546098392e-06, + "loss": 0.8255, + "step": 340 + }, + { + "epoch": 0.32820019249278154, + "grad_norm": 1.1297482097311666, + "learning_rate": 9.350143839017124e-06, + "loss": 0.7605, + "step": 341 + }, + { + "epoch": 0.329162656400385, + "grad_norm": 1.2155287689707892, + "learning_rate": 9.346412189055955e-06, + "loss": 0.907, + "step": 342 + }, + { + "epoch": 0.33012512030798846, + "grad_norm": 1.1033447728001136, + "learning_rate": 9.342670604744113e-06, + "loss": 0.7573, + "step": 343 + }, + { + "epoch": 0.3310875842155919, + "grad_norm": 1.0424384449149806, + "learning_rate": 9.338919094633524e-06, + "loss": 0.7313, + "step": 344 + }, + { + "epoch": 0.3320500481231954, + "grad_norm": 1.2077440536067385, + "learning_rate": 9.335157667298808e-06, + "loss": 0.8508, + "step": 345 + }, + { + "epoch": 0.33301251203079885, + "grad_norm": 1.1062372058069239, + "learning_rate": 9.331386331337246e-06, + "loss": 0.8527, + "step": 346 + }, + { + "epoch": 0.3339749759384023, + "grad_norm": 1.0652694278688877, + "learning_rate": 9.327605095368768e-06, + "loss": 0.6945, + "step": 347 + }, + { + "epoch": 0.33493743984600577, + "grad_norm": 1.1522805790761372, + "learning_rate": 9.323813968035938e-06, + "loss": 0.809, + "step": 348 + }, + { + "epoch": 0.33589990375360923, + "grad_norm": 1.0597828783816325, + "learning_rate": 9.32001295800392e-06, + "loss": 0.7359, + "step": 349 + }, + { + "epoch": 0.3368623676612127, + "grad_norm": 1.032248797380791, + "learning_rate": 9.316202073960467e-06, + "loss": 0.7387, + "step": 350 + }, + { + "epoch": 0.33782483156881615, + "grad_norm": 1.1284886575043955, + "learning_rate": 9.312381324615906e-06, + "loss": 0.7339, + "step": 351 + }, + { + "epoch": 0.3387872954764196, + "grad_norm": 1.3623621698353536, + "learning_rate": 9.308550718703111e-06, + "loss": 0.9367, + "step": 352 + }, + { + "epoch": 0.3397497593840231, + "grad_norm": 1.1514840626394809, + "learning_rate": 9.30471026497748e-06, + "loss": 0.8117, + "step": 353 + }, + { + "epoch": 0.34071222329162654, + "grad_norm": 1.189819296890319, + "learning_rate": 9.300859972216925e-06, + "loss": 0.9076, + "step": 354 + }, + { + "epoch": 0.34167468719923005, + "grad_norm": 1.166654268446047, + "learning_rate": 9.296999849221843e-06, + "loss": 0.8414, + "step": 355 + }, + { + "epoch": 0.3426371511068335, + "grad_norm": 1.0763942385974345, + "learning_rate": 9.2931299048151e-06, + "loss": 0.7591, + "step": 356 + }, + { + "epoch": 0.343599615014437, + "grad_norm": 1.062301146276317, + "learning_rate": 9.289250147842014e-06, + "loss": 0.7263, + "step": 357 + }, + { + "epoch": 0.34456207892204044, + "grad_norm": 1.0868004397488311, + "learning_rate": 9.285360587170325e-06, + "loss": 0.7642, + "step": 358 + }, + { + "epoch": 0.3455245428296439, + "grad_norm": 1.2049993917724398, + "learning_rate": 9.281461231690187e-06, + "loss": 0.8982, + "step": 359 + }, + { + "epoch": 0.34648700673724736, + "grad_norm": 1.2044499018530335, + "learning_rate": 9.277552090314135e-06, + "loss": 0.8389, + "step": 360 + }, + { + "epoch": 0.3474494706448508, + "grad_norm": 1.1459934254014361, + "learning_rate": 9.273633171977074e-06, + "loss": 0.811, + "step": 361 + }, + { + "epoch": 0.3484119345524543, + "grad_norm": 1.123363979049624, + "learning_rate": 9.26970448563626e-06, + "loss": 0.7352, + "step": 362 + }, + { + "epoch": 0.34937439846005774, + "grad_norm": 1.092894306912654, + "learning_rate": 9.265766040271267e-06, + "loss": 0.7243, + "step": 363 + }, + { + "epoch": 0.3503368623676612, + "grad_norm": 1.7279743260793086, + "learning_rate": 9.26181784488398e-06, + "loss": 0.7429, + "step": 364 + }, + { + "epoch": 0.35129932627526467, + "grad_norm": 1.1448164757883446, + "learning_rate": 9.257859908498568e-06, + "loss": 0.754, + "step": 365 + }, + { + "epoch": 0.35226179018286813, + "grad_norm": 0.9974664449986839, + "learning_rate": 9.253892240161467e-06, + "loss": 0.6566, + "step": 366 + }, + { + "epoch": 0.3532242540904716, + "grad_norm": 1.2544452995769204, + "learning_rate": 9.249914848941347e-06, + "loss": 0.8526, + "step": 367 + }, + { + "epoch": 0.35418671799807505, + "grad_norm": 1.128503167604882, + "learning_rate": 9.245927743929117e-06, + "loss": 0.8034, + "step": 368 + }, + { + "epoch": 0.3551491819056785, + "grad_norm": 1.2068179044202683, + "learning_rate": 9.241930934237873e-06, + "loss": 0.8228, + "step": 369 + }, + { + "epoch": 0.35611164581328203, + "grad_norm": 1.337510754778371, + "learning_rate": 9.237924429002902e-06, + "loss": 0.9414, + "step": 370 + }, + { + "epoch": 0.3570741097208855, + "grad_norm": 1.1874476843663977, + "learning_rate": 9.233908237381648e-06, + "loss": 0.7774, + "step": 371 + }, + { + "epoch": 0.35803657362848895, + "grad_norm": 1.1078991004955732, + "learning_rate": 9.229882368553692e-06, + "loss": 0.7423, + "step": 372 + }, + { + "epoch": 0.3589990375360924, + "grad_norm": 1.1221560916234117, + "learning_rate": 9.225846831720741e-06, + "loss": 0.7327, + "step": 373 + }, + { + "epoch": 0.3599615014436959, + "grad_norm": 1.2374264173074037, + "learning_rate": 9.221801636106591e-06, + "loss": 0.8216, + "step": 374 + }, + { + "epoch": 0.36092396535129934, + "grad_norm": 1.2068367969717766, + "learning_rate": 9.217746790957124e-06, + "loss": 0.8416, + "step": 375 + }, + { + "epoch": 0.3618864292589028, + "grad_norm": 1.2555475452767837, + "learning_rate": 9.213682305540268e-06, + "loss": 0.8164, + "step": 376 + }, + { + "epoch": 0.36284889316650626, + "grad_norm": 1.0334372620022907, + "learning_rate": 9.20960818914599e-06, + "loss": 0.699, + "step": 377 + }, + { + "epoch": 0.3638113570741097, + "grad_norm": 1.176142066718557, + "learning_rate": 9.205524451086274e-06, + "loss": 0.7904, + "step": 378 + }, + { + "epoch": 0.3647738209817132, + "grad_norm": 1.1485373216411572, + "learning_rate": 9.201431100695085e-06, + "loss": 0.821, + "step": 379 + }, + { + "epoch": 0.36573628488931664, + "grad_norm": 1.0558530530907588, + "learning_rate": 9.197328147328366e-06, + "loss": 0.7396, + "step": 380 + }, + { + "epoch": 0.3666987487969201, + "grad_norm": 1.1737781833101024, + "learning_rate": 9.19321560036401e-06, + "loss": 0.8192, + "step": 381 + }, + { + "epoch": 0.36766121270452357, + "grad_norm": 1.2124835372257132, + "learning_rate": 9.189093469201833e-06, + "loss": 0.7997, + "step": 382 + }, + { + "epoch": 0.368623676612127, + "grad_norm": 1.0903073606135485, + "learning_rate": 9.18496176326356e-06, + "loss": 0.7102, + "step": 383 + }, + { + "epoch": 0.3695861405197305, + "grad_norm": 1.1492313076049732, + "learning_rate": 9.180820491992799e-06, + "loss": 0.7699, + "step": 384 + }, + { + "epoch": 0.37054860442733395, + "grad_norm": 1.1783988057500718, + "learning_rate": 9.176669664855022e-06, + "loss": 0.8433, + "step": 385 + }, + { + "epoch": 0.37151106833493747, + "grad_norm": 1.2470987128515285, + "learning_rate": 9.172509291337542e-06, + "loss": 0.8331, + "step": 386 + }, + { + "epoch": 0.37247353224254093, + "grad_norm": 1.2167367273031047, + "learning_rate": 9.16833938094949e-06, + "loss": 0.8869, + "step": 387 + }, + { + "epoch": 0.3734359961501444, + "grad_norm": 1.1087734222734136, + "learning_rate": 9.1641599432218e-06, + "loss": 0.8155, + "step": 388 + }, + { + "epoch": 0.37439846005774785, + "grad_norm": 1.1666945005738494, + "learning_rate": 9.159970987707175e-06, + "loss": 0.7974, + "step": 389 + }, + { + "epoch": 0.3753609239653513, + "grad_norm": 1.1033510158551292, + "learning_rate": 9.155772523980075e-06, + "loss": 0.8056, + "step": 390 + }, + { + "epoch": 0.3763233878729548, + "grad_norm": 1.1612374423800165, + "learning_rate": 9.151564561636697e-06, + "loss": 0.8057, + "step": 391 + }, + { + "epoch": 0.37728585178055823, + "grad_norm": 1.1703648782991638, + "learning_rate": 9.147347110294942e-06, + "loss": 0.7989, + "step": 392 + }, + { + "epoch": 0.3782483156881617, + "grad_norm": 1.1497463871586988, + "learning_rate": 9.143120179594404e-06, + "loss": 0.7896, + "step": 393 + }, + { + "epoch": 0.37921077959576516, + "grad_norm": 1.1783144243637425, + "learning_rate": 9.138883779196338e-06, + "loss": 0.8588, + "step": 394 + }, + { + "epoch": 0.3801732435033686, + "grad_norm": 1.1059264161157114, + "learning_rate": 9.134637918783653e-06, + "loss": 0.8058, + "step": 395 + }, + { + "epoch": 0.3811357074109721, + "grad_norm": 1.071000551292565, + "learning_rate": 9.130382608060869e-06, + "loss": 0.733, + "step": 396 + }, + { + "epoch": 0.38209817131857554, + "grad_norm": 1.088034581945864, + "learning_rate": 9.126117856754111e-06, + "loss": 0.735, + "step": 397 + }, + { + "epoch": 0.383060635226179, + "grad_norm": 1.1892330454919597, + "learning_rate": 9.121843674611084e-06, + "loss": 0.8721, + "step": 398 + }, + { + "epoch": 0.38402309913378246, + "grad_norm": 1.2635732164646731, + "learning_rate": 9.117560071401047e-06, + "loss": 0.9103, + "step": 399 + }, + { + "epoch": 0.3849855630413859, + "grad_norm": 1.0871560446138382, + "learning_rate": 9.11326705691479e-06, + "loss": 0.7079, + "step": 400 + }, + { + "epoch": 0.3859480269489894, + "grad_norm": 1.0482418328071237, + "learning_rate": 9.108964640964617e-06, + "loss": 0.7075, + "step": 401 + }, + { + "epoch": 0.3869104908565929, + "grad_norm": 1.008789797415242, + "learning_rate": 9.104652833384317e-06, + "loss": 0.671, + "step": 402 + }, + { + "epoch": 0.38787295476419636, + "grad_norm": 1.7538060453256632, + "learning_rate": 9.10033164402915e-06, + "loss": 0.7289, + "step": 403 + }, + { + "epoch": 0.3888354186717998, + "grad_norm": 1.2253210840836668, + "learning_rate": 9.096001082775812e-06, + "loss": 0.8363, + "step": 404 + }, + { + "epoch": 0.3897978825794033, + "grad_norm": 0.9352947079018498, + "learning_rate": 9.091661159522426e-06, + "loss": 0.5481, + "step": 405 + }, + { + "epoch": 0.39076034648700675, + "grad_norm": 0.9740921876528373, + "learning_rate": 9.087311884188514e-06, + "loss": 0.5914, + "step": 406 + }, + { + "epoch": 0.3917228103946102, + "grad_norm": 1.252784611853974, + "learning_rate": 9.082953266714968e-06, + "loss": 0.8096, + "step": 407 + }, + { + "epoch": 0.39268527430221367, + "grad_norm": 1.2318734460370306, + "learning_rate": 9.078585317064036e-06, + "loss": 0.9152, + "step": 408 + }, + { + "epoch": 0.39364773820981713, + "grad_norm": 1.1818343974092023, + "learning_rate": 9.074208045219297e-06, + "loss": 0.8185, + "step": 409 + }, + { + "epoch": 0.3946102021174206, + "grad_norm": 1.1805054664222911, + "learning_rate": 9.069821461185631e-06, + "loss": 0.8256, + "step": 410 + }, + { + "epoch": 0.39557266602502406, + "grad_norm": 1.1517680420104146, + "learning_rate": 9.065425574989214e-06, + "loss": 0.7704, + "step": 411 + }, + { + "epoch": 0.3965351299326275, + "grad_norm": 1.9437620341689106, + "learning_rate": 9.06102039667747e-06, + "loss": 0.7542, + "step": 412 + }, + { + "epoch": 0.397497593840231, + "grad_norm": 1.3327387971750062, + "learning_rate": 9.056605936319076e-06, + "loss": 0.8974, + "step": 413 + }, + { + "epoch": 0.39846005774783444, + "grad_norm": 1.1958224461630134, + "learning_rate": 9.05218220400391e-06, + "loss": 0.9063, + "step": 414 + }, + { + "epoch": 0.3994225216554379, + "grad_norm": 1.0825574246626648, + "learning_rate": 9.04774920984305e-06, + "loss": 0.6978, + "step": 415 + }, + { + "epoch": 0.40038498556304136, + "grad_norm": 1.185973378479154, + "learning_rate": 9.043306963968746e-06, + "loss": 0.803, + "step": 416 + }, + { + "epoch": 0.4013474494706448, + "grad_norm": 1.1931711632846371, + "learning_rate": 9.038855476534386e-06, + "loss": 0.7105, + "step": 417 + }, + { + "epoch": 0.40230991337824834, + "grad_norm": 0.8323149271437723, + "learning_rate": 9.034394757714486e-06, + "loss": 0.4876, + "step": 418 + }, + { + "epoch": 0.4032723772858518, + "grad_norm": 1.1977840576184686, + "learning_rate": 9.029924817704664e-06, + "loss": 0.8285, + "step": 419 + }, + { + "epoch": 0.40423484119345526, + "grad_norm": 1.1064360173802363, + "learning_rate": 9.025445666721608e-06, + "loss": 0.6784, + "step": 420 + }, + { + "epoch": 0.4051973051010587, + "grad_norm": 1.3265930128303254, + "learning_rate": 9.020957315003064e-06, + "loss": 0.879, + "step": 421 + }, + { + "epoch": 0.4061597690086622, + "grad_norm": 1.119298326219782, + "learning_rate": 9.016459772807804e-06, + "loss": 0.6194, + "step": 422 + }, + { + "epoch": 0.40712223291626565, + "grad_norm": 1.144979072901613, + "learning_rate": 9.011953050415612e-06, + "loss": 0.7198, + "step": 423 + }, + { + "epoch": 0.4080846968238691, + "grad_norm": 1.0500961743424124, + "learning_rate": 9.007437158127245e-06, + "loss": 0.7645, + "step": 424 + }, + { + "epoch": 0.40904716073147257, + "grad_norm": 1.236096568997424, + "learning_rate": 9.00291210626443e-06, + "loss": 0.8809, + "step": 425 + }, + { + "epoch": 0.41000962463907603, + "grad_norm": 1.3146348316144556, + "learning_rate": 8.998377905169822e-06, + "loss": 0.9189, + "step": 426 + }, + { + "epoch": 0.4109720885466795, + "grad_norm": 1.0968997382593533, + "learning_rate": 8.99383456520699e-06, + "loss": 0.6967, + "step": 427 + }, + { + "epoch": 0.41193455245428295, + "grad_norm": 1.0896117391896827, + "learning_rate": 8.989282096760392e-06, + "loss": 0.8327, + "step": 428 + }, + { + "epoch": 0.4128970163618864, + "grad_norm": 1.1765939592994745, + "learning_rate": 8.98472051023535e-06, + "loss": 0.8348, + "step": 429 + }, + { + "epoch": 0.4138594802694899, + "grad_norm": 1.2079865654646527, + "learning_rate": 8.980149816058028e-06, + "loss": 0.8961, + "step": 430 + }, + { + "epoch": 0.41482194417709334, + "grad_norm": 1.0408711763881806, + "learning_rate": 8.975570024675402e-06, + "loss": 0.6846, + "step": 431 + }, + { + "epoch": 0.4157844080846968, + "grad_norm": 1.1891847913399132, + "learning_rate": 8.970981146555247e-06, + "loss": 0.8067, + "step": 432 + }, + { + "epoch": 0.4167468719923003, + "grad_norm": 1.111037005916126, + "learning_rate": 8.966383192186105e-06, + "loss": 0.716, + "step": 433 + }, + { + "epoch": 0.4177093358999038, + "grad_norm": 1.0632808745677016, + "learning_rate": 8.96177617207726e-06, + "loss": 0.6741, + "step": 434 + }, + { + "epoch": 0.41867179980750724, + "grad_norm": 1.1980129937653308, + "learning_rate": 8.957160096758722e-06, + "loss": 0.8997, + "step": 435 + }, + { + "epoch": 0.4196342637151107, + "grad_norm": 1.1693802052800903, + "learning_rate": 8.952534976781192e-06, + "loss": 0.8015, + "step": 436 + }, + { + "epoch": 0.42059672762271416, + "grad_norm": 0.9579901355247453, + "learning_rate": 8.947900822716054e-06, + "loss": 0.661, + "step": 437 + }, + { + "epoch": 0.4215591915303176, + "grad_norm": 1.1771599840482465, + "learning_rate": 8.943257645155328e-06, + "loss": 0.9348, + "step": 438 + }, + { + "epoch": 0.4225216554379211, + "grad_norm": 1.0393671063648802, + "learning_rate": 8.938605454711667e-06, + "loss": 0.7245, + "step": 439 + }, + { + "epoch": 0.42348411934552455, + "grad_norm": 0.9989670822204548, + "learning_rate": 8.93394426201832e-06, + "loss": 0.6812, + "step": 440 + }, + { + "epoch": 0.424446583253128, + "grad_norm": 2.2759458497050282, + "learning_rate": 8.929274077729118e-06, + "loss": 0.5887, + "step": 441 + }, + { + "epoch": 0.42540904716073147, + "grad_norm": 1.1557649787951458, + "learning_rate": 8.924594912518434e-06, + "loss": 0.7971, + "step": 442 + }, + { + "epoch": 0.42637151106833493, + "grad_norm": 1.246487821614179, + "learning_rate": 8.919906777081179e-06, + "loss": 0.8396, + "step": 443 + }, + { + "epoch": 0.4273339749759384, + "grad_norm": 1.0858784701129216, + "learning_rate": 8.91520968213276e-06, + "loss": 0.7942, + "step": 444 + }, + { + "epoch": 0.42829643888354185, + "grad_norm": 1.1293024628921333, + "learning_rate": 8.91050363840906e-06, + "loss": 0.8188, + "step": 445 + }, + { + "epoch": 0.4292589027911453, + "grad_norm": 1.1786246158821891, + "learning_rate": 8.905788656666422e-06, + "loss": 0.8621, + "step": 446 + }, + { + "epoch": 0.4302213666987488, + "grad_norm": 1.2692032633455093, + "learning_rate": 8.901064747681615e-06, + "loss": 0.9231, + "step": 447 + }, + { + "epoch": 0.43118383060635224, + "grad_norm": 1.246237957772232, + "learning_rate": 8.896331922251814e-06, + "loss": 0.9086, + "step": 448 + }, + { + "epoch": 0.43214629451395575, + "grad_norm": 1.1631502925728086, + "learning_rate": 8.891590191194573e-06, + "loss": 0.8546, + "step": 449 + }, + { + "epoch": 0.4331087584215592, + "grad_norm": 1.0967974868614545, + "learning_rate": 8.886839565347798e-06, + "loss": 0.7436, + "step": 450 + }, + { + "epoch": 0.4340712223291627, + "grad_norm": 1.1129543679376954, + "learning_rate": 8.88208005556973e-06, + "loss": 0.7307, + "step": 451 + }, + { + "epoch": 0.43503368623676614, + "grad_norm": 1.191838376327943, + "learning_rate": 8.877311672738912e-06, + "loss": 0.8759, + "step": 452 + }, + { + "epoch": 0.4359961501443696, + "grad_norm": 1.2483634461298805, + "learning_rate": 8.872534427754174e-06, + "loss": 0.7996, + "step": 453 + }, + { + "epoch": 0.43695861405197306, + "grad_norm": 1.196644722592687, + "learning_rate": 8.867748331534594e-06, + "loss": 0.8074, + "step": 454 + }, + { + "epoch": 0.4379210779595765, + "grad_norm": 1.1719478060690283, + "learning_rate": 8.862953395019483e-06, + "loss": 0.7347, + "step": 455 + }, + { + "epoch": 0.43888354186718, + "grad_norm": 1.3594434008029819, + "learning_rate": 8.858149629168358e-06, + "loss": 0.893, + "step": 456 + }, + { + "epoch": 0.43984600577478344, + "grad_norm": 1.1130284805740818, + "learning_rate": 8.853337044960919e-06, + "loss": 0.7984, + "step": 457 + }, + { + "epoch": 0.4408084696823869, + "grad_norm": 0.9417120976784563, + "learning_rate": 8.84851565339702e-06, + "loss": 0.6246, + "step": 458 + }, + { + "epoch": 0.44177093358999037, + "grad_norm": 1.241550917700917, + "learning_rate": 8.843685465496645e-06, + "loss": 0.785, + "step": 459 + }, + { + "epoch": 0.44273339749759383, + "grad_norm": 1.2235985658579593, + "learning_rate": 8.838846492299884e-06, + "loss": 0.7719, + "step": 460 + }, + { + "epoch": 0.4436958614051973, + "grad_norm": 1.243186820830219, + "learning_rate": 8.833998744866908e-06, + "loss": 0.8, + "step": 461 + }, + { + "epoch": 0.44465832531280075, + "grad_norm": 0.9798608562596828, + "learning_rate": 8.829142234277937e-06, + "loss": 0.6556, + "step": 462 + }, + { + "epoch": 0.4456207892204042, + "grad_norm": 1.141016709915036, + "learning_rate": 8.82427697163323e-06, + "loss": 0.8164, + "step": 463 + }, + { + "epoch": 0.4465832531280077, + "grad_norm": 1.2991771239719618, + "learning_rate": 8.819402968053042e-06, + "loss": 0.9191, + "step": 464 + }, + { + "epoch": 0.4475457170356112, + "grad_norm": 1.2127668199016561, + "learning_rate": 8.81452023467761e-06, + "loss": 0.7241, + "step": 465 + }, + { + "epoch": 0.44850818094321465, + "grad_norm": 1.0135243938873073, + "learning_rate": 8.809628782667126e-06, + "loss": 0.6956, + "step": 466 + }, + { + "epoch": 0.4494706448508181, + "grad_norm": 1.1776164347577611, + "learning_rate": 8.804728623201706e-06, + "loss": 0.786, + "step": 467 + }, + { + "epoch": 0.4504331087584216, + "grad_norm": 0.9247288642843106, + "learning_rate": 8.79981976748137e-06, + "loss": 0.5342, + "step": 468 + }, + { + "epoch": 0.45139557266602504, + "grad_norm": 1.1757187099786908, + "learning_rate": 8.794902226726014e-06, + "loss": 0.9013, + "step": 469 + }, + { + "epoch": 0.4523580365736285, + "grad_norm": 1.155833084418181, + "learning_rate": 8.789976012175387e-06, + "loss": 0.8216, + "step": 470 + }, + { + "epoch": 0.45332050048123196, + "grad_norm": 2.3087666555627484, + "learning_rate": 8.78504113508906e-06, + "loss": 0.7725, + "step": 471 + }, + { + "epoch": 0.4542829643888354, + "grad_norm": 1.3376876093895327, + "learning_rate": 8.780097606746405e-06, + "loss": 0.8885, + "step": 472 + }, + { + "epoch": 0.4552454282964389, + "grad_norm": 1.096925951725579, + "learning_rate": 8.775145438446566e-06, + "loss": 0.8139, + "step": 473 + }, + { + "epoch": 0.45620789220404234, + "grad_norm": 1.2457522266407655, + "learning_rate": 8.770184641508438e-06, + "loss": 0.8689, + "step": 474 + }, + { + "epoch": 0.4571703561116458, + "grad_norm": 1.035437887467623, + "learning_rate": 8.765215227270639e-06, + "loss": 0.6763, + "step": 475 + }, + { + "epoch": 0.45813282001924927, + "grad_norm": 1.0143876718762967, + "learning_rate": 8.760237207091476e-06, + "loss": 0.6883, + "step": 476 + }, + { + "epoch": 0.4590952839268527, + "grad_norm": 0.9213340805303033, + "learning_rate": 8.755250592348935e-06, + "loss": 0.5825, + "step": 477 + }, + { + "epoch": 0.4600577478344562, + "grad_norm": 0.9576945262548351, + "learning_rate": 8.750255394440636e-06, + "loss": 0.5369, + "step": 478 + }, + { + "epoch": 0.46102021174205965, + "grad_norm": 1.1232240231420247, + "learning_rate": 8.745251624783828e-06, + "loss": 0.7934, + "step": 479 + }, + { + "epoch": 0.4619826756496631, + "grad_norm": 0.9661412051425768, + "learning_rate": 8.740239294815347e-06, + "loss": 0.6256, + "step": 480 + }, + { + "epoch": 0.4629451395572666, + "grad_norm": 1.1162045324221401, + "learning_rate": 8.735218415991589e-06, + "loss": 0.7737, + "step": 481 + }, + { + "epoch": 0.4639076034648701, + "grad_norm": 1.0385725680408904, + "learning_rate": 8.7301889997885e-06, + "loss": 0.788, + "step": 482 + }, + { + "epoch": 0.46487006737247355, + "grad_norm": 1.1066735618775823, + "learning_rate": 8.725151057701534e-06, + "loss": 0.8091, + "step": 483 + }, + { + "epoch": 0.465832531280077, + "grad_norm": 1.0660771650769416, + "learning_rate": 8.720104601245633e-06, + "loss": 0.7727, + "step": 484 + }, + { + "epoch": 0.4667949951876805, + "grad_norm": 1.1069516890371065, + "learning_rate": 8.7150496419552e-06, + "loss": 0.7418, + "step": 485 + }, + { + "epoch": 0.46775745909528393, + "grad_norm": 0.9665869802082596, + "learning_rate": 8.70998619138407e-06, + "loss": 0.5785, + "step": 486 + }, + { + "epoch": 0.4687199230028874, + "grad_norm": 1.3004419825256797, + "learning_rate": 8.704914261105489e-06, + "loss": 0.7281, + "step": 487 + }, + { + "epoch": 0.46968238691049086, + "grad_norm": 1.216053136358682, + "learning_rate": 8.699833862712082e-06, + "loss": 0.8933, + "step": 488 + }, + { + "epoch": 0.4706448508180943, + "grad_norm": 0.9734354214307193, + "learning_rate": 8.694745007815835e-06, + "loss": 0.633, + "step": 489 + }, + { + "epoch": 0.4716073147256978, + "grad_norm": 1.5525178037644163, + "learning_rate": 8.689647708048055e-06, + "loss": 0.8742, + "step": 490 + }, + { + "epoch": 0.47256977863330124, + "grad_norm": 1.2556031655443751, + "learning_rate": 8.684541975059357e-06, + "loss": 0.8532, + "step": 491 + }, + { + "epoch": 0.4735322425409047, + "grad_norm": 1.1602236373184047, + "learning_rate": 8.679427820519624e-06, + "loss": 0.8203, + "step": 492 + }, + { + "epoch": 0.47449470644850816, + "grad_norm": 1.2952430573850275, + "learning_rate": 8.674305256117997e-06, + "loss": 0.8848, + "step": 493 + }, + { + "epoch": 0.4754571703561116, + "grad_norm": 1.170237490007117, + "learning_rate": 8.669174293562833e-06, + "loss": 0.7255, + "step": 494 + }, + { + "epoch": 0.4764196342637151, + "grad_norm": 63.258391898222115, + "learning_rate": 8.664034944581685e-06, + "loss": 0.8649, + "step": 495 + }, + { + "epoch": 0.4773820981713186, + "grad_norm": 1.1023128498287937, + "learning_rate": 8.658887220921274e-06, + "loss": 0.7352, + "step": 496 + }, + { + "epoch": 0.47834456207892206, + "grad_norm": 1.136339596273602, + "learning_rate": 8.653731134347465e-06, + "loss": 0.7246, + "step": 497 + }, + { + "epoch": 0.4793070259865255, + "grad_norm": 3.1613716762830966, + "learning_rate": 8.648566696645235e-06, + "loss": 0.6939, + "step": 498 + }, + { + "epoch": 0.480269489894129, + "grad_norm": 1.19953367336472, + "learning_rate": 8.64339391961865e-06, + "loss": 0.8648, + "step": 499 + }, + { + "epoch": 0.48123195380173245, + "grad_norm": 1.1476741495173035, + "learning_rate": 8.638212815090836e-06, + "loss": 0.8185, + "step": 500 + }, + { + "epoch": 0.48123195380173245, + "eval_loss": 0.8154048919677734, + "eval_runtime": 12.5911, + "eval_samples_per_second": 3.336, + "eval_steps_per_second": 0.874, + "step": 500 + }, + { + "epoch": 0.4821944177093359, + "grad_norm": 1.019029296590487, + "learning_rate": 8.633023394903955e-06, + "loss": 0.6892, + "step": 501 + }, + { + "epoch": 0.48315688161693937, + "grad_norm": 1.2230715404300774, + "learning_rate": 8.62782567091917e-06, + "loss": 0.8902, + "step": 502 + }, + { + "epoch": 0.48411934552454283, + "grad_norm": 0.9348859209674789, + "learning_rate": 8.622619655016632e-06, + "loss": 0.6125, + "step": 503 + }, + { + "epoch": 0.4850818094321463, + "grad_norm": 1.0828730377571367, + "learning_rate": 8.617405359095437e-06, + "loss": 0.797, + "step": 504 + }, + { + "epoch": 0.48604427333974976, + "grad_norm": 1.184259710971126, + "learning_rate": 8.612182795073608e-06, + "loss": 0.8215, + "step": 505 + }, + { + "epoch": 0.4870067372473532, + "grad_norm": 0.94238902295402, + "learning_rate": 8.60695197488807e-06, + "loss": 0.5989, + "step": 506 + }, + { + "epoch": 0.4879692011549567, + "grad_norm": 1.0640090793448387, + "learning_rate": 8.601712910494613e-06, + "loss": 0.7459, + "step": 507 + }, + { + "epoch": 0.48893166506256014, + "grad_norm": 1.1718705734803416, + "learning_rate": 8.596465613867873e-06, + "loss": 0.8094, + "step": 508 + }, + { + "epoch": 0.4898941289701636, + "grad_norm": 1.211895751591689, + "learning_rate": 8.5912100970013e-06, + "loss": 0.8663, + "step": 509 + }, + { + "epoch": 0.49085659287776706, + "grad_norm": 1.1666867286802225, + "learning_rate": 8.585946371907138e-06, + "loss": 0.8789, + "step": 510 + }, + { + "epoch": 0.4918190567853705, + "grad_norm": 1.1534573526845968, + "learning_rate": 8.58067445061639e-06, + "loss": 0.7978, + "step": 511 + }, + { + "epoch": 0.49278152069297404, + "grad_norm": 0.9113408561548744, + "learning_rate": 8.575394345178782e-06, + "loss": 0.5629, + "step": 512 + }, + { + "epoch": 0.4937439846005775, + "grad_norm": 1.2784019231118626, + "learning_rate": 8.570106067662765e-06, + "loss": 0.9127, + "step": 513 + }, + { + "epoch": 0.49470644850818096, + "grad_norm": 1.105880966897377, + "learning_rate": 8.564809630155451e-06, + "loss": 0.7419, + "step": 514 + }, + { + "epoch": 0.4956689124157844, + "grad_norm": 1.1001659730329003, + "learning_rate": 8.559505044762617e-06, + "loss": 0.7097, + "step": 515 + }, + { + "epoch": 0.4966313763233879, + "grad_norm": 0.9367406769695084, + "learning_rate": 8.55419232360865e-06, + "loss": 0.6108, + "step": 516 + }, + { + "epoch": 0.49759384023099135, + "grad_norm": 1.2443157048237867, + "learning_rate": 8.548871478836542e-06, + "loss": 0.8945, + "step": 517 + }, + { + "epoch": 0.4985563041385948, + "grad_norm": 1.1624124876261783, + "learning_rate": 8.54354252260785e-06, + "loss": 0.8038, + "step": 518 + }, + { + "epoch": 0.49951876804619827, + "grad_norm": 1.07267856325009, + "learning_rate": 8.53820546710267e-06, + "loss": 0.7079, + "step": 519 + }, + { + "epoch": 0.5004812319538018, + "grad_norm": 1.2541180138684938, + "learning_rate": 8.532860324519608e-06, + "loss": 0.9083, + "step": 520 + }, + { + "epoch": 0.5014436958614052, + "grad_norm": 1.0731707470939502, + "learning_rate": 8.527507107075761e-06, + "loss": 0.763, + "step": 521 + }, + { + "epoch": 0.5024061597690087, + "grad_norm": 1.059291191696007, + "learning_rate": 8.522145827006677e-06, + "loss": 0.7543, + "step": 522 + }, + { + "epoch": 0.5033686236766122, + "grad_norm": 1.2056204623615607, + "learning_rate": 8.516776496566332e-06, + "loss": 0.8208, + "step": 523 + }, + { + "epoch": 0.5043310875842156, + "grad_norm": 0.9546864721768225, + "learning_rate": 8.511399128027102e-06, + "loss": 0.6125, + "step": 524 + }, + { + "epoch": 0.5052935514918191, + "grad_norm": 1.05362819938232, + "learning_rate": 8.50601373367974e-06, + "loss": 0.723, + "step": 525 + }, + { + "epoch": 0.5062560153994226, + "grad_norm": 1.1051079333407121, + "learning_rate": 8.50062032583334e-06, + "loss": 0.8021, + "step": 526 + }, + { + "epoch": 0.507218479307026, + "grad_norm": 1.1841531707704076, + "learning_rate": 8.49521891681531e-06, + "loss": 0.9021, + "step": 527 + }, + { + "epoch": 0.5081809432146295, + "grad_norm": 0.9410136713074027, + "learning_rate": 8.489809518971348e-06, + "loss": 0.559, + "step": 528 + }, + { + "epoch": 0.5091434071222329, + "grad_norm": 1.2079274015850334, + "learning_rate": 8.484392144665412e-06, + "loss": 0.8818, + "step": 529 + }, + { + "epoch": 0.5101058710298364, + "grad_norm": 1.1260251653966724, + "learning_rate": 8.478966806279692e-06, + "loss": 0.8004, + "step": 530 + }, + { + "epoch": 0.5110683349374399, + "grad_norm": 1.0559068383013595, + "learning_rate": 8.473533516214577e-06, + "loss": 0.6918, + "step": 531 + }, + { + "epoch": 0.5120307988450433, + "grad_norm": 1.128327239042325, + "learning_rate": 8.468092286888635e-06, + "loss": 0.7732, + "step": 532 + }, + { + "epoch": 0.5129932627526468, + "grad_norm": 1.2167057392206737, + "learning_rate": 8.462643130738578e-06, + "loss": 0.7972, + "step": 533 + }, + { + "epoch": 0.5139557266602502, + "grad_norm": 0.9676057575963026, + "learning_rate": 8.45718606021924e-06, + "loss": 0.6399, + "step": 534 + }, + { + "epoch": 0.5149181905678537, + "grad_norm": 1.0656109728412289, + "learning_rate": 8.451721087803537e-06, + "loss": 0.7268, + "step": 535 + }, + { + "epoch": 0.5158806544754572, + "grad_norm": 1.04744669995521, + "learning_rate": 8.446248225982451e-06, + "loss": 0.6369, + "step": 536 + }, + { + "epoch": 0.5168431183830606, + "grad_norm": 1.2493680500392603, + "learning_rate": 8.440767487264999e-06, + "loss": 0.9148, + "step": 537 + }, + { + "epoch": 0.5178055822906641, + "grad_norm": 0.9972893138556869, + "learning_rate": 8.43527888417819e-06, + "loss": 0.6974, + "step": 538 + }, + { + "epoch": 0.5187680461982676, + "grad_norm": 1.2358196181267047, + "learning_rate": 8.429782429267025e-06, + "loss": 0.8736, + "step": 539 + }, + { + "epoch": 0.519730510105871, + "grad_norm": 1.1383002206435378, + "learning_rate": 8.424278135094442e-06, + "loss": 0.7957, + "step": 540 + }, + { + "epoch": 0.5206929740134745, + "grad_norm": 1.033487358485919, + "learning_rate": 8.418766014241294e-06, + "loss": 0.7092, + "step": 541 + }, + { + "epoch": 0.5216554379210779, + "grad_norm": 1.0939489107764842, + "learning_rate": 8.413246079306325e-06, + "loss": 0.7154, + "step": 542 + }, + { + "epoch": 0.5226179018286814, + "grad_norm": 1.1553919640394266, + "learning_rate": 8.407718342906146e-06, + "loss": 0.7958, + "step": 543 + }, + { + "epoch": 0.5235803657362849, + "grad_norm": 1.07080379518568, + "learning_rate": 8.40218281767519e-06, + "loss": 0.6882, + "step": 544 + }, + { + "epoch": 0.5245428296438883, + "grad_norm": 1.0197880786778006, + "learning_rate": 8.3966395162657e-06, + "loss": 0.7186, + "step": 545 + }, + { + "epoch": 0.5255052935514918, + "grad_norm": 1.0867866520181695, + "learning_rate": 8.391088451347688e-06, + "loss": 0.7331, + "step": 546 + }, + { + "epoch": 0.5264677574590952, + "grad_norm": 1.1025485461915157, + "learning_rate": 8.385529635608913e-06, + "loss": 0.6907, + "step": 547 + }, + { + "epoch": 0.5274302213666987, + "grad_norm": 1.1102998894970588, + "learning_rate": 8.379963081754846e-06, + "loss": 0.8117, + "step": 548 + }, + { + "epoch": 0.5283926852743022, + "grad_norm": 1.1908596882995257, + "learning_rate": 8.374388802508648e-06, + "loss": 0.8613, + "step": 549 + }, + { + "epoch": 0.5293551491819056, + "grad_norm": 1.1670397974951192, + "learning_rate": 8.368806810611137e-06, + "loss": 0.7882, + "step": 550 + }, + { + "epoch": 0.5303176130895092, + "grad_norm": 1.2079990600055364, + "learning_rate": 8.363217118820758e-06, + "loss": 0.7979, + "step": 551 + }, + { + "epoch": 0.5312800769971127, + "grad_norm": 1.214023964111433, + "learning_rate": 8.357619739913557e-06, + "loss": 0.8997, + "step": 552 + }, + { + "epoch": 0.5322425409047161, + "grad_norm": 1.2454549509595867, + "learning_rate": 8.35201468668315e-06, + "loss": 0.9049, + "step": 553 + }, + { + "epoch": 0.5332050048123196, + "grad_norm": 1.0559785118298561, + "learning_rate": 8.346401971940693e-06, + "loss": 0.7144, + "step": 554 + }, + { + "epoch": 0.534167468719923, + "grad_norm": 1.0599712709481093, + "learning_rate": 8.34078160851485e-06, + "loss": 0.6191, + "step": 555 + }, + { + "epoch": 0.5351299326275265, + "grad_norm": 1.0349127116375185, + "learning_rate": 8.335153609251775e-06, + "loss": 0.6979, + "step": 556 + }, + { + "epoch": 0.53609239653513, + "grad_norm": 1.185343910806179, + "learning_rate": 8.329517987015068e-06, + "loss": 0.8513, + "step": 557 + }, + { + "epoch": 0.5370548604427334, + "grad_norm": 1.1890546821106835, + "learning_rate": 8.323874754685755e-06, + "loss": 0.8548, + "step": 558 + }, + { + "epoch": 0.5380173243503369, + "grad_norm": 1.1465337719368263, + "learning_rate": 8.318223925162257e-06, + "loss": 0.8645, + "step": 559 + }, + { + "epoch": 0.5389797882579404, + "grad_norm": 1.1362689813480327, + "learning_rate": 8.312565511360358e-06, + "loss": 0.8084, + "step": 560 + }, + { + "epoch": 0.5399422521655438, + "grad_norm": 0.9774293470790113, + "learning_rate": 8.306899526213175e-06, + "loss": 0.6363, + "step": 561 + }, + { + "epoch": 0.5409047160731473, + "grad_norm": 1.2159910317917713, + "learning_rate": 8.301225982671135e-06, + "loss": 0.7016, + "step": 562 + }, + { + "epoch": 0.5418671799807507, + "grad_norm": 1.1364296421230091, + "learning_rate": 8.295544893701936e-06, + "loss": 0.7813, + "step": 563 + }, + { + "epoch": 0.5428296438883542, + "grad_norm": 1.0336434203301383, + "learning_rate": 8.289856272290526e-06, + "loss": 0.7189, + "step": 564 + }, + { + "epoch": 0.5437921077959577, + "grad_norm": 1.1058816588370466, + "learning_rate": 8.284160131439071e-06, + "loss": 0.624, + "step": 565 + }, + { + "epoch": 0.5447545717035611, + "grad_norm": 1.172034803602417, + "learning_rate": 8.278456484166916e-06, + "loss": 0.7992, + "step": 566 + }, + { + "epoch": 0.5457170356111646, + "grad_norm": 1.0492174791783704, + "learning_rate": 8.27274534351057e-06, + "loss": 0.7117, + "step": 567 + }, + { + "epoch": 0.546679499518768, + "grad_norm": 1.0331453990551334, + "learning_rate": 8.267026722523668e-06, + "loss": 0.7053, + "step": 568 + }, + { + "epoch": 0.5476419634263715, + "grad_norm": 1.018576335292646, + "learning_rate": 8.261300634276938e-06, + "loss": 0.6069, + "step": 569 + }, + { + "epoch": 0.548604427333975, + "grad_norm": 1.1922803413272156, + "learning_rate": 8.255567091858183e-06, + "loss": 0.8284, + "step": 570 + }, + { + "epoch": 0.5495668912415784, + "grad_norm": 1.0643346500326671, + "learning_rate": 8.249826108372236e-06, + "loss": 0.6552, + "step": 571 + }, + { + "epoch": 0.5505293551491819, + "grad_norm": 1.220537739418529, + "learning_rate": 8.244077696940943e-06, + "loss": 0.9067, + "step": 572 + }, + { + "epoch": 0.5514918190567853, + "grad_norm": 1.1402337112384084, + "learning_rate": 8.238321870703128e-06, + "loss": 0.7346, + "step": 573 + }, + { + "epoch": 0.5524542829643888, + "grad_norm": 1.2134189953657881, + "learning_rate": 8.232558642814555e-06, + "loss": 0.8531, + "step": 574 + }, + { + "epoch": 0.5534167468719923, + "grad_norm": 1.0783630485115294, + "learning_rate": 8.226788026447915e-06, + "loss": 0.7774, + "step": 575 + }, + { + "epoch": 0.5543792107795957, + "grad_norm": 1.1213374569474097, + "learning_rate": 8.22101003479278e-06, + "loss": 0.7546, + "step": 576 + }, + { + "epoch": 0.5553416746871992, + "grad_norm": 1.227514935837079, + "learning_rate": 8.215224681055585e-06, + "loss": 0.8805, + "step": 577 + }, + { + "epoch": 0.5563041385948027, + "grad_norm": 1.175779766529062, + "learning_rate": 8.209431978459583e-06, + "loss": 0.8511, + "step": 578 + }, + { + "epoch": 0.5572666025024061, + "grad_norm": 1.1124517226221624, + "learning_rate": 8.203631940244836e-06, + "loss": 0.8151, + "step": 579 + }, + { + "epoch": 0.5582290664100096, + "grad_norm": 1.2543278617470568, + "learning_rate": 8.197824579668162e-06, + "loss": 0.824, + "step": 580 + }, + { + "epoch": 0.559191530317613, + "grad_norm": 1.0838725027571627, + "learning_rate": 8.192009910003118e-06, + "loss": 0.7147, + "step": 581 + }, + { + "epoch": 0.5601539942252165, + "grad_norm": 5.733092619837746, + "learning_rate": 8.186187944539974e-06, + "loss": 0.7204, + "step": 582 + }, + { + "epoch": 0.5611164581328201, + "grad_norm": 1.1051737369363126, + "learning_rate": 8.180358696585665e-06, + "loss": 0.7438, + "step": 583 + }, + { + "epoch": 0.5620789220404235, + "grad_norm": 1.0826287960526018, + "learning_rate": 8.174522179463779e-06, + "loss": 0.7158, + "step": 584 + }, + { + "epoch": 0.563041385948027, + "grad_norm": 1.0147481003486847, + "learning_rate": 8.168678406514517e-06, + "loss": 0.716, + "step": 585 + }, + { + "epoch": 0.5640038498556305, + "grad_norm": 1.233086923360333, + "learning_rate": 8.162827391094661e-06, + "loss": 0.8717, + "step": 586 + }, + { + "epoch": 0.5649663137632339, + "grad_norm": 1.1960191755716574, + "learning_rate": 8.156969146577549e-06, + "loss": 0.897, + "step": 587 + }, + { + "epoch": 0.5659287776708374, + "grad_norm": 1.2764824852087142, + "learning_rate": 8.151103686353042e-06, + "loss": 0.9, + "step": 588 + }, + { + "epoch": 0.5668912415784408, + "grad_norm": 1.1354511474811366, + "learning_rate": 8.145231023827495e-06, + "loss": 0.8224, + "step": 589 + }, + { + "epoch": 0.5678537054860443, + "grad_norm": 1.0511211562754392, + "learning_rate": 8.139351172423724e-06, + "loss": 0.726, + "step": 590 + }, + { + "epoch": 0.5688161693936478, + "grad_norm": 0.9793662432299434, + "learning_rate": 8.133464145580973e-06, + "loss": 0.6264, + "step": 591 + }, + { + "epoch": 0.5697786333012512, + "grad_norm": 1.185008219142365, + "learning_rate": 8.12756995675489e-06, + "loss": 0.7675, + "step": 592 + }, + { + "epoch": 0.5707410972088547, + "grad_norm": 0.9757817477046189, + "learning_rate": 8.121668619417493e-06, + "loss": 0.5764, + "step": 593 + }, + { + "epoch": 0.5717035611164581, + "grad_norm": 1.1997715711173011, + "learning_rate": 8.115760147057139e-06, + "loss": 0.9079, + "step": 594 + }, + { + "epoch": 0.5726660250240616, + "grad_norm": 1.2372975606852574, + "learning_rate": 8.109844553178488e-06, + "loss": 0.8824, + "step": 595 + }, + { + "epoch": 0.5736284889316651, + "grad_norm": 1.0861210262548882, + "learning_rate": 8.103921851302486e-06, + "loss": 0.8006, + "step": 596 + }, + { + "epoch": 0.5745909528392685, + "grad_norm": 1.0433397332852363, + "learning_rate": 8.097992054966317e-06, + "loss": 0.6735, + "step": 597 + }, + { + "epoch": 0.575553416746872, + "grad_norm": 1.076765099057458, + "learning_rate": 8.092055177723388e-06, + "loss": 0.7318, + "step": 598 + }, + { + "epoch": 0.5765158806544755, + "grad_norm": 1.1783517719506498, + "learning_rate": 8.086111233143282e-06, + "loss": 0.8476, + "step": 599 + }, + { + "epoch": 0.5774783445620789, + "grad_norm": 1.2298954793681722, + "learning_rate": 8.080160234811743e-06, + "loss": 0.9166, + "step": 600 + }, + { + "epoch": 0.5784408084696824, + "grad_norm": 1.1050293073301662, + "learning_rate": 8.074202196330631e-06, + "loss": 0.8037, + "step": 601 + }, + { + "epoch": 0.5794032723772858, + "grad_norm": 1.1373722081545534, + "learning_rate": 8.068237131317904e-06, + "loss": 0.785, + "step": 602 + }, + { + "epoch": 0.5803657362848893, + "grad_norm": 1.2285788167244456, + "learning_rate": 8.062265053407577e-06, + "loss": 0.797, + "step": 603 + }, + { + "epoch": 0.5813282001924928, + "grad_norm": 1.227040430983201, + "learning_rate": 8.056285976249689e-06, + "loss": 0.9037, + "step": 604 + }, + { + "epoch": 0.5822906641000962, + "grad_norm": 1.1370047693471745, + "learning_rate": 8.050299913510287e-06, + "loss": 0.8098, + "step": 605 + }, + { + "epoch": 0.5832531280076997, + "grad_norm": 1.0964252986323508, + "learning_rate": 8.044306878871376e-06, + "loss": 0.8085, + "step": 606 + }, + { + "epoch": 0.5842155919153031, + "grad_norm": 1.08526302267498, + "learning_rate": 8.0383068860309e-06, + "loss": 0.7858, + "step": 607 + }, + { + "epoch": 0.5851780558229066, + "grad_norm": 1.067325650632938, + "learning_rate": 8.032299948702707e-06, + "loss": 0.7061, + "step": 608 + }, + { + "epoch": 0.5861405197305101, + "grad_norm": 0.7941115645062901, + "learning_rate": 8.026286080616515e-06, + "loss": 0.4688, + "step": 609 + }, + { + "epoch": 0.5871029836381135, + "grad_norm": 1.0433967698767077, + "learning_rate": 8.020265295517885e-06, + "loss": 0.6567, + "step": 610 + }, + { + "epoch": 0.588065447545717, + "grad_norm": 1.1611198306194022, + "learning_rate": 8.014237607168192e-06, + "loss": 0.8179, + "step": 611 + }, + { + "epoch": 0.5890279114533205, + "grad_norm": 1.0372664007195125, + "learning_rate": 8.00820302934458e-06, + "loss": 0.7157, + "step": 612 + }, + { + "epoch": 0.5899903753609239, + "grad_norm": 1.3025918235874918, + "learning_rate": 8.002161575839948e-06, + "loss": 0.8981, + "step": 613 + }, + { + "epoch": 0.5909528392685275, + "grad_norm": 1.0412578758775526, + "learning_rate": 7.996113260462908e-06, + "loss": 0.6882, + "step": 614 + }, + { + "epoch": 0.591915303176131, + "grad_norm": 1.2475179627774737, + "learning_rate": 7.990058097037755e-06, + "loss": 0.8807, + "step": 615 + }, + { + "epoch": 0.5928777670837344, + "grad_norm": 0.9834332581717565, + "learning_rate": 7.983996099404435e-06, + "loss": 0.633, + "step": 616 + }, + { + "epoch": 0.5938402309913379, + "grad_norm": 1.010180428731807, + "learning_rate": 7.977927281418519e-06, + "loss": 0.6242, + "step": 617 + }, + { + "epoch": 0.5948026948989413, + "grad_norm": 1.2622346744553483, + "learning_rate": 7.971851656951161e-06, + "loss": 0.8672, + "step": 618 + }, + { + "epoch": 0.5957651588065448, + "grad_norm": 1.107205753320005, + "learning_rate": 7.965769239889078e-06, + "loss": 0.7514, + "step": 619 + }, + { + "epoch": 0.5967276227141483, + "grad_norm": 1.1790336082104342, + "learning_rate": 7.959680044134507e-06, + "loss": 0.801, + "step": 620 + }, + { + "epoch": 0.5976900866217517, + "grad_norm": 1.0296315313174311, + "learning_rate": 7.953584083605183e-06, + "loss": 0.7212, + "step": 621 + }, + { + "epoch": 0.5986525505293552, + "grad_norm": 1.1101721564101392, + "learning_rate": 7.947481372234303e-06, + "loss": 0.8019, + "step": 622 + }, + { + "epoch": 0.5996150144369586, + "grad_norm": 1.1793316856998952, + "learning_rate": 7.941371923970489e-06, + "loss": 0.8261, + "step": 623 + }, + { + "epoch": 0.6005774783445621, + "grad_norm": 1.2051381616231494, + "learning_rate": 7.935255752777766e-06, + "loss": 0.9117, + "step": 624 + }, + { + "epoch": 0.6015399422521656, + "grad_norm": 1.1107502735868167, + "learning_rate": 7.92913287263552e-06, + "loss": 0.7814, + "step": 625 + }, + { + "epoch": 0.602502406159769, + "grad_norm": 1.0972706875884328, + "learning_rate": 7.923003297538483e-06, + "loss": 0.7914, + "step": 626 + }, + { + "epoch": 0.6034648700673725, + "grad_norm": 1.1236404215014324, + "learning_rate": 7.916867041496674e-06, + "loss": 0.7372, + "step": 627 + }, + { + "epoch": 0.6044273339749759, + "grad_norm": 0.9266691436599598, + "learning_rate": 7.91072411853539e-06, + "loss": 0.581, + "step": 628 + }, + { + "epoch": 0.6053897978825794, + "grad_norm": 1.2786506171328496, + "learning_rate": 7.90457454269517e-06, + "loss": 0.8762, + "step": 629 + }, + { + "epoch": 0.6063522617901829, + "grad_norm": 1.1200595951302723, + "learning_rate": 7.898418328031752e-06, + "loss": 0.7739, + "step": 630 + }, + { + "epoch": 0.6073147256977863, + "grad_norm": 1.0111112321323583, + "learning_rate": 7.892255488616049e-06, + "loss": 0.6371, + "step": 631 + }, + { + "epoch": 0.6082771896053898, + "grad_norm": 1.1804873300392595, + "learning_rate": 7.886086038534121e-06, + "loss": 0.8781, + "step": 632 + }, + { + "epoch": 0.6092396535129933, + "grad_norm": 1.0401014516648124, + "learning_rate": 7.879909991887133e-06, + "loss": 0.6911, + "step": 633 + }, + { + "epoch": 0.6102021174205967, + "grad_norm": 1.162510014603857, + "learning_rate": 7.873727362791327e-06, + "loss": 0.8154, + "step": 634 + }, + { + "epoch": 0.6111645813282002, + "grad_norm": 1.117712520469151, + "learning_rate": 7.867538165377992e-06, + "loss": 0.6942, + "step": 635 + }, + { + "epoch": 0.6121270452358036, + "grad_norm": 1.0313093933854411, + "learning_rate": 7.861342413793434e-06, + "loss": 0.6709, + "step": 636 + }, + { + "epoch": 0.6130895091434071, + "grad_norm": 1.1171972365200822, + "learning_rate": 7.855140122198927e-06, + "loss": 0.7964, + "step": 637 + }, + { + "epoch": 0.6140519730510106, + "grad_norm": 1.0947037270903026, + "learning_rate": 7.848931304770708e-06, + "loss": 0.824, + "step": 638 + }, + { + "epoch": 0.615014436958614, + "grad_norm": 1.03035864935146, + "learning_rate": 7.842715975699921e-06, + "loss": 0.6385, + "step": 639 + }, + { + "epoch": 0.6159769008662175, + "grad_norm": 1.0578703309708393, + "learning_rate": 7.836494149192594e-06, + "loss": 0.7347, + "step": 640 + }, + { + "epoch": 0.6169393647738209, + "grad_norm": 1.265481814826767, + "learning_rate": 7.830265839469608e-06, + "loss": 0.8816, + "step": 641 + }, + { + "epoch": 0.6179018286814244, + "grad_norm": 1.0788924573931662, + "learning_rate": 7.824031060766662e-06, + "loss": 0.7665, + "step": 642 + }, + { + "epoch": 0.6188642925890279, + "grad_norm": 1.0347623704701125, + "learning_rate": 7.817789827334239e-06, + "loss": 0.6824, + "step": 643 + }, + { + "epoch": 0.6198267564966313, + "grad_norm": 1.1400029628961643, + "learning_rate": 7.811542153437575e-06, + "loss": 0.7918, + "step": 644 + }, + { + "epoch": 0.6207892204042348, + "grad_norm": 1.0822139510245041, + "learning_rate": 7.805288053356632e-06, + "loss": 0.7525, + "step": 645 + }, + { + "epoch": 0.6217516843118384, + "grad_norm": 1.0673963310317918, + "learning_rate": 7.799027541386051e-06, + "loss": 0.7295, + "step": 646 + }, + { + "epoch": 0.6227141482194418, + "grad_norm": 1.0167109387451647, + "learning_rate": 7.792760631835138e-06, + "loss": 0.7007, + "step": 647 + }, + { + "epoch": 0.6236766121270453, + "grad_norm": 0.9821223047046962, + "learning_rate": 7.786487339027816e-06, + "loss": 0.6771, + "step": 648 + }, + { + "epoch": 0.6246390760346487, + "grad_norm": 1.0509500639807245, + "learning_rate": 7.780207677302594e-06, + "loss": 0.7183, + "step": 649 + }, + { + "epoch": 0.6256015399422522, + "grad_norm": 1.216969292451248, + "learning_rate": 7.773921661012548e-06, + "loss": 0.803, + "step": 650 + }, + { + "epoch": 0.6265640038498557, + "grad_norm": 1.0391983839914944, + "learning_rate": 7.767629304525271e-06, + "loss": 0.7012, + "step": 651 + }, + { + "epoch": 0.6275264677574591, + "grad_norm": 0.8965253804065677, + "learning_rate": 7.76133062222285e-06, + "loss": 0.567, + "step": 652 + }, + { + "epoch": 0.6284889316650626, + "grad_norm": 1.2361134439942978, + "learning_rate": 7.75502562850183e-06, + "loss": 0.8585, + "step": 653 + }, + { + "epoch": 0.629451395572666, + "grad_norm": 1.1955340524059075, + "learning_rate": 7.748714337773179e-06, + "loss": 0.8171, + "step": 654 + }, + { + "epoch": 0.6304138594802695, + "grad_norm": 1.1467123048535848, + "learning_rate": 7.742396764462266e-06, + "loss": 0.7943, + "step": 655 + }, + { + "epoch": 0.631376323387873, + "grad_norm": 1.138841687022404, + "learning_rate": 7.736072923008808e-06, + "loss": 0.771, + "step": 656 + }, + { + "epoch": 0.6323387872954764, + "grad_norm": 1.1081659415036589, + "learning_rate": 7.72974282786686e-06, + "loss": 0.7661, + "step": 657 + }, + { + "epoch": 0.6333012512030799, + "grad_norm": 1.190126634597938, + "learning_rate": 7.72340649350476e-06, + "loss": 0.8744, + "step": 658 + }, + { + "epoch": 0.6342637151106834, + "grad_norm": 1.135274285667019, + "learning_rate": 7.71706393440512e-06, + "loss": 0.8174, + "step": 659 + }, + { + "epoch": 0.6352261790182868, + "grad_norm": 1.0581110287701117, + "learning_rate": 7.710715165064766e-06, + "loss": 0.7196, + "step": 660 + }, + { + "epoch": 0.6361886429258903, + "grad_norm": 1.1971178999248504, + "learning_rate": 7.704360199994725e-06, + "loss": 0.7948, + "step": 661 + }, + { + "epoch": 0.6371511068334937, + "grad_norm": 1.1610322319344004, + "learning_rate": 7.697999053720185e-06, + "loss": 0.8035, + "step": 662 + }, + { + "epoch": 0.6381135707410972, + "grad_norm": 1.051185822028776, + "learning_rate": 7.691631740780463e-06, + "loss": 0.7412, + "step": 663 + }, + { + "epoch": 0.6390760346487007, + "grad_norm": 0.8236747405869431, + "learning_rate": 7.685258275728964e-06, + "loss": 0.522, + "step": 664 + }, + { + "epoch": 0.6400384985563041, + "grad_norm": 1.125002893290332, + "learning_rate": 7.678878673133165e-06, + "loss": 0.7328, + "step": 665 + }, + { + "epoch": 0.6410009624639076, + "grad_norm": 1.0562142820855736, + "learning_rate": 7.672492947574565e-06, + "loss": 0.6928, + "step": 666 + }, + { + "epoch": 0.641963426371511, + "grad_norm": 1.0949376033945526, + "learning_rate": 7.666101113648659e-06, + "loss": 0.7444, + "step": 667 + }, + { + "epoch": 0.6429258902791145, + "grad_norm": 1.0220539579758094, + "learning_rate": 7.659703185964902e-06, + "loss": 0.6742, + "step": 668 + }, + { + "epoch": 0.643888354186718, + "grad_norm": 1.097560359064998, + "learning_rate": 7.653299179146678e-06, + "loss": 0.784, + "step": 669 + }, + { + "epoch": 0.6448508180943214, + "grad_norm": 1.2313156775678213, + "learning_rate": 7.64688910783127e-06, + "loss": 0.8873, + "step": 670 + }, + { + "epoch": 0.6458132820019249, + "grad_norm": 1.0914884451137015, + "learning_rate": 7.640472986669814e-06, + "loss": 0.6562, + "step": 671 + }, + { + "epoch": 0.6467757459095284, + "grad_norm": 1.078386533815879, + "learning_rate": 7.634050830327283e-06, + "loss": 0.7109, + "step": 672 + }, + { + "epoch": 0.6477382098171318, + "grad_norm": 1.1803213666611212, + "learning_rate": 7.627622653482436e-06, + "loss": 0.8966, + "step": 673 + }, + { + "epoch": 0.6487006737247353, + "grad_norm": 0.9905248348925672, + "learning_rate": 7.621188470827798e-06, + "loss": 0.6034, + "step": 674 + }, + { + "epoch": 0.6496631376323387, + "grad_norm": 1.1195961431003458, + "learning_rate": 7.6147482970696205e-06, + "loss": 0.5948, + "step": 675 + }, + { + "epoch": 0.6506256015399422, + "grad_norm": 1.1459713985503839, + "learning_rate": 7.608302146927846e-06, + "loss": 0.7838, + "step": 676 + }, + { + "epoch": 0.6515880654475458, + "grad_norm": 1.1309075070641392, + "learning_rate": 7.601850035136079e-06, + "loss": 0.783, + "step": 677 + }, + { + "epoch": 0.6525505293551492, + "grad_norm": 1.2712515066290946, + "learning_rate": 7.5953919764415505e-06, + "loss": 0.8269, + "step": 678 + }, + { + "epoch": 0.6535129932627527, + "grad_norm": 1.0813192272817487, + "learning_rate": 7.588927985605083e-06, + "loss": 0.7491, + "step": 679 + }, + { + "epoch": 0.6544754571703562, + "grad_norm": 1.0452690107618676, + "learning_rate": 7.582458077401056e-06, + "loss": 0.664, + "step": 680 + }, + { + "epoch": 0.6554379210779596, + "grad_norm": 1.031685025916053, + "learning_rate": 7.575982266617378e-06, + "loss": 0.7041, + "step": 681 + }, + { + "epoch": 0.6564003849855631, + "grad_norm": 1.1202097381518632, + "learning_rate": 7.569500568055448e-06, + "loss": 0.7866, + "step": 682 + }, + { + "epoch": 0.6573628488931665, + "grad_norm": 1.0892871036612946, + "learning_rate": 7.563012996530117e-06, + "loss": 0.7876, + "step": 683 + }, + { + "epoch": 0.65832531280077, + "grad_norm": 1.0314247119362465, + "learning_rate": 7.556519566869667e-06, + "loss": 0.7003, + "step": 684 + }, + { + "epoch": 0.6592877767083735, + "grad_norm": 1.157299472672375, + "learning_rate": 7.550020293915765e-06, + "loss": 0.7637, + "step": 685 + }, + { + "epoch": 0.6602502406159769, + "grad_norm": 1.2206247439792453, + "learning_rate": 7.543515192523435e-06, + "loss": 0.8931, + "step": 686 + }, + { + "epoch": 0.6612127045235804, + "grad_norm": 1.1680440855656589, + "learning_rate": 7.537004277561022e-06, + "loss": 0.7781, + "step": 687 + }, + { + "epoch": 0.6621751684311838, + "grad_norm": 1.0631959224690661, + "learning_rate": 7.530487563910158e-06, + "loss": 0.6933, + "step": 688 + }, + { + "epoch": 0.6631376323387873, + "grad_norm": 1.1465553790754641, + "learning_rate": 7.523965066465734e-06, + "loss": 0.7909, + "step": 689 + }, + { + "epoch": 0.6641000962463908, + "grad_norm": 1.0127458135689362, + "learning_rate": 7.517436800135853e-06, + "loss": 0.7175, + "step": 690 + }, + { + "epoch": 0.6650625601539942, + "grad_norm": 0.9845663643625207, + "learning_rate": 7.510902779841811e-06, + "loss": 0.7271, + "step": 691 + }, + { + "epoch": 0.6660250240615977, + "grad_norm": 1.1871378197273714, + "learning_rate": 7.504363020518046e-06, + "loss": 0.8644, + "step": 692 + }, + { + "epoch": 0.6669874879692012, + "grad_norm": 1.0289064574977809, + "learning_rate": 7.497817537112127e-06, + "loss": 0.7044, + "step": 693 + }, + { + "epoch": 0.6679499518768046, + "grad_norm": 1.0920460229422444, + "learning_rate": 7.491266344584691e-06, + "loss": 0.7914, + "step": 694 + }, + { + "epoch": 0.6689124157844081, + "grad_norm": 0.9597442400351844, + "learning_rate": 7.484709457909437e-06, + "loss": 0.662, + "step": 695 + }, + { + "epoch": 0.6698748796920115, + "grad_norm": 1.0723924764856265, + "learning_rate": 7.47814689207307e-06, + "loss": 0.7987, + "step": 696 + }, + { + "epoch": 0.670837343599615, + "grad_norm": 1.0993031216666644, + "learning_rate": 7.4715786620752805e-06, + "loss": 0.7369, + "step": 697 + }, + { + "epoch": 0.6717998075072185, + "grad_norm": 1.0757116704616414, + "learning_rate": 7.465004782928704e-06, + "loss": 0.8048, + "step": 698 + }, + { + "epoch": 0.6727622714148219, + "grad_norm": 1.142800376164258, + "learning_rate": 7.4584252696588854e-06, + "loss": 0.8808, + "step": 699 + }, + { + "epoch": 0.6737247353224254, + "grad_norm": 0.9815799768078552, + "learning_rate": 7.4518401373042515e-06, + "loss": 0.587, + "step": 700 + }, + { + "epoch": 0.6746871992300288, + "grad_norm": 1.091152390449687, + "learning_rate": 7.4452494009160695e-06, + "loss": 0.8154, + "step": 701 + }, + { + "epoch": 0.6756496631376323, + "grad_norm": 1.1344266432688512, + "learning_rate": 7.438653075558413e-06, + "loss": 0.8769, + "step": 702 + }, + { + "epoch": 0.6766121270452358, + "grad_norm": 1.0134769707970794, + "learning_rate": 7.432051176308139e-06, + "loss": 0.6953, + "step": 703 + }, + { + "epoch": 0.6775745909528392, + "grad_norm": 1.1188972141341302, + "learning_rate": 7.425443718254835e-06, + "loss": 0.7334, + "step": 704 + }, + { + "epoch": 0.6785370548604427, + "grad_norm": 0.8882475211464808, + "learning_rate": 7.4188307165007964e-06, + "loss": 0.5798, + "step": 705 + }, + { + "epoch": 0.6794995187680462, + "grad_norm": 1.13019391115174, + "learning_rate": 7.412212186160992e-06, + "loss": 0.8305, + "step": 706 + }, + { + "epoch": 0.6804619826756496, + "grad_norm": 1.196934127764919, + "learning_rate": 7.405588142363026e-06, + "loss": 0.8331, + "step": 707 + }, + { + "epoch": 0.6814244465832531, + "grad_norm": 1.1996016225013237, + "learning_rate": 7.398958600247104e-06, + "loss": 0.8354, + "step": 708 + }, + { + "epoch": 0.6823869104908566, + "grad_norm": 0.9531850820305385, + "learning_rate": 7.3923235749659976e-06, + "loss": 0.6353, + "step": 709 + }, + { + "epoch": 0.6833493743984601, + "grad_norm": 1.139373205916164, + "learning_rate": 7.3856830816850135e-06, + "loss": 0.8084, + "step": 710 + }, + { + "epoch": 0.6843118383060636, + "grad_norm": 1.0647084443746593, + "learning_rate": 7.379037135581954e-06, + "loss": 0.7738, + "step": 711 + }, + { + "epoch": 0.685274302213667, + "grad_norm": 1.092946236501217, + "learning_rate": 7.372385751847084e-06, + "loss": 0.767, + "step": 712 + }, + { + "epoch": 0.6862367661212705, + "grad_norm": 1.1293043400537968, + "learning_rate": 7.3657289456831015e-06, + "loss": 0.6868, + "step": 713 + }, + { + "epoch": 0.687199230028874, + "grad_norm": 2.622628949239281, + "learning_rate": 7.359066732305095e-06, + "loss": 0.621, + "step": 714 + }, + { + "epoch": 0.6881616939364774, + "grad_norm": 0.9351405143173539, + "learning_rate": 7.3523991269405095e-06, + "loss": 0.5245, + "step": 715 + }, + { + "epoch": 0.6891241578440809, + "grad_norm": 1.126885969972214, + "learning_rate": 7.345726144829117e-06, + "loss": 0.8192, + "step": 716 + }, + { + "epoch": 0.6900866217516843, + "grad_norm": 0.9862906041841694, + "learning_rate": 7.339047801222982e-06, + "loss": 0.6494, + "step": 717 + }, + { + "epoch": 0.6910490856592878, + "grad_norm": 2.1508550881827215, + "learning_rate": 7.3323641113864175e-06, + "loss": 0.8411, + "step": 718 + }, + { + "epoch": 0.6920115495668913, + "grad_norm": 1.1085985125957545, + "learning_rate": 7.325675090595958e-06, + "loss": 0.7277, + "step": 719 + }, + { + "epoch": 0.6929740134744947, + "grad_norm": 1.0358567007220278, + "learning_rate": 7.318980754140325e-06, + "loss": 0.6287, + "step": 720 + }, + { + "epoch": 0.6939364773820982, + "grad_norm": 0.8317623218488485, + "learning_rate": 7.312281117320388e-06, + "loss": 0.4835, + "step": 721 + }, + { + "epoch": 0.6948989412897016, + "grad_norm": 0.9887328716532409, + "learning_rate": 7.305576195449131e-06, + "loss": 0.7021, + "step": 722 + }, + { + "epoch": 0.6958614051973051, + "grad_norm": 1.08615281329148, + "learning_rate": 7.298866003851617e-06, + "loss": 0.7955, + "step": 723 + }, + { + "epoch": 0.6968238691049086, + "grad_norm": 1.0568700270422278, + "learning_rate": 7.292150557864956e-06, + "loss": 0.6914, + "step": 724 + }, + { + "epoch": 0.697786333012512, + "grad_norm": 1.023150512191821, + "learning_rate": 7.285429872838264e-06, + "loss": 0.6656, + "step": 725 + }, + { + "epoch": 0.6987487969201155, + "grad_norm": 1.0863379852216353, + "learning_rate": 7.2787039641326384e-06, + "loss": 0.7811, + "step": 726 + }, + { + "epoch": 0.699711260827719, + "grad_norm": 1.0614583787363625, + "learning_rate": 7.271972847121108e-06, + "loss": 0.7766, + "step": 727 + }, + { + "epoch": 0.7006737247353224, + "grad_norm": 1.0472609683043481, + "learning_rate": 7.2652365371886116e-06, + "loss": 0.6965, + "step": 728 + }, + { + "epoch": 0.7016361886429259, + "grad_norm": 0.9903897808778249, + "learning_rate": 7.258495049731953e-06, + "loss": 0.6832, + "step": 729 + }, + { + "epoch": 0.7025986525505293, + "grad_norm": 1.1535398642848251, + "learning_rate": 7.251748400159774e-06, + "loss": 0.7913, + "step": 730 + }, + { + "epoch": 0.7035611164581328, + "grad_norm": 1.0730665961088208, + "learning_rate": 7.244996603892513e-06, + "loss": 0.7884, + "step": 731 + }, + { + "epoch": 0.7045235803657363, + "grad_norm": 1.2003423178340302, + "learning_rate": 7.238239676362372e-06, + "loss": 0.8956, + "step": 732 + }, + { + "epoch": 0.7054860442733397, + "grad_norm": 1.0201142546248403, + "learning_rate": 7.231477633013282e-06, + "loss": 0.6997, + "step": 733 + }, + { + "epoch": 0.7064485081809432, + "grad_norm": 1.1120976462321415, + "learning_rate": 7.224710489300866e-06, + "loss": 0.7792, + "step": 734 + }, + { + "epoch": 0.7074109720885466, + "grad_norm": 1.0637967152260106, + "learning_rate": 7.217938260692403e-06, + "loss": 0.712, + "step": 735 + }, + { + "epoch": 0.7083734359961501, + "grad_norm": 1.1387205952002712, + "learning_rate": 7.211160962666799e-06, + "loss": 0.7593, + "step": 736 + }, + { + "epoch": 0.7093358999037536, + "grad_norm": 1.0874835432560361, + "learning_rate": 7.204378610714544e-06, + "loss": 0.6825, + "step": 737 + }, + { + "epoch": 0.710298363811357, + "grad_norm": 1.0052143466819878, + "learning_rate": 7.197591220337678e-06, + "loss": 0.6938, + "step": 738 + }, + { + "epoch": 0.7112608277189605, + "grad_norm": 1.1170713588315895, + "learning_rate": 7.190798807049759e-06, + "loss": 0.7614, + "step": 739 + }, + { + "epoch": 0.7122232916265641, + "grad_norm": 1.0584686071351366, + "learning_rate": 7.1840013863758276e-06, + "loss": 0.6914, + "step": 740 + }, + { + "epoch": 0.7131857555341675, + "grad_norm": 1.142389282814946, + "learning_rate": 7.177198973852364e-06, + "loss": 0.8893, + "step": 741 + }, + { + "epoch": 0.714148219441771, + "grad_norm": 0.9095000217002229, + "learning_rate": 7.170391585027263e-06, + "loss": 0.6309, + "step": 742 + }, + { + "epoch": 0.7151106833493744, + "grad_norm": 0.9972002032627318, + "learning_rate": 7.163579235459792e-06, + "loss": 0.6195, + "step": 743 + }, + { + "epoch": 0.7160731472569779, + "grad_norm": 0.9765098710220961, + "learning_rate": 7.156761940720556e-06, + "loss": 0.5884, + "step": 744 + }, + { + "epoch": 0.7170356111645814, + "grad_norm": 1.121557412728386, + "learning_rate": 7.149939716391461e-06, + "loss": 0.7624, + "step": 745 + }, + { + "epoch": 0.7179980750721848, + "grad_norm": 1.102803684811716, + "learning_rate": 7.1431125780656865e-06, + "loss": 0.8097, + "step": 746 + }, + { + "epoch": 0.7189605389797883, + "grad_norm": 0.9745241918303451, + "learning_rate": 7.136280541347638e-06, + "loss": 0.6092, + "step": 747 + }, + { + "epoch": 0.7199230028873917, + "grad_norm": 0.947986378070234, + "learning_rate": 7.129443621852919e-06, + "loss": 0.6233, + "step": 748 + }, + { + "epoch": 0.7208854667949952, + "grad_norm": 1.0680257451813515, + "learning_rate": 7.122601835208291e-06, + "loss": 0.7322, + "step": 749 + }, + { + "epoch": 0.7218479307025987, + "grad_norm": 1.1024280737056953, + "learning_rate": 7.115755197051645e-06, + "loss": 0.7397, + "step": 750 + }, + { + "epoch": 0.7228103946102021, + "grad_norm": 1.1789414181954732, + "learning_rate": 7.108903723031958e-06, + "loss": 0.8733, + "step": 751 + }, + { + "epoch": 0.7237728585178056, + "grad_norm": 1.0663338013056864, + "learning_rate": 7.102047428809259e-06, + "loss": 0.7455, + "step": 752 + }, + { + "epoch": 0.7247353224254091, + "grad_norm": 1.0650018584553043, + "learning_rate": 7.095186330054595e-06, + "loss": 0.6952, + "step": 753 + }, + { + "epoch": 0.7256977863330125, + "grad_norm": 1.041211569101052, + "learning_rate": 7.088320442449996e-06, + "loss": 0.7389, + "step": 754 + }, + { + "epoch": 0.726660250240616, + "grad_norm": 0.9356910111488846, + "learning_rate": 7.081449781688435e-06, + "loss": 0.6301, + "step": 755 + }, + { + "epoch": 0.7276227141482194, + "grad_norm": 1.0001173440703566, + "learning_rate": 7.074574363473798e-06, + "loss": 0.7108, + "step": 756 + }, + { + "epoch": 0.7285851780558229, + "grad_norm": 1.0989032874575713, + "learning_rate": 7.067694203520842e-06, + "loss": 0.8064, + "step": 757 + }, + { + "epoch": 0.7295476419634264, + "grad_norm": 1.1220729346469287, + "learning_rate": 7.0608093175551615e-06, + "loss": 0.7399, + "step": 758 + }, + { + "epoch": 0.7305101058710298, + "grad_norm": 1.0348117141326727, + "learning_rate": 7.053919721313156e-06, + "loss": 0.7278, + "step": 759 + }, + { + "epoch": 0.7314725697786333, + "grad_norm": 6.0943581639962385, + "learning_rate": 7.047025430541989e-06, + "loss": 0.6855, + "step": 760 + }, + { + "epoch": 0.7324350336862367, + "grad_norm": 1.0888321469600946, + "learning_rate": 7.040126460999554e-06, + "loss": 0.7103, + "step": 761 + }, + { + "epoch": 0.7333974975938402, + "grad_norm": 1.1351105695948875, + "learning_rate": 7.033222828454441e-06, + "loss": 0.7022, + "step": 762 + }, + { + "epoch": 0.7343599615014437, + "grad_norm": 1.1827836325850516, + "learning_rate": 7.0263145486858915e-06, + "loss": 0.8581, + "step": 763 + }, + { + "epoch": 0.7353224254090471, + "grad_norm": 1.22271682908927, + "learning_rate": 7.019401637483778e-06, + "loss": 0.8843, + "step": 764 + }, + { + "epoch": 0.7362848893166506, + "grad_norm": 1.011879462277834, + "learning_rate": 7.012484110648552e-06, + "loss": 0.6979, + "step": 765 + }, + { + "epoch": 0.737247353224254, + "grad_norm": 1.863767354175747, + "learning_rate": 7.005561983991217e-06, + "loss": 0.7974, + "step": 766 + }, + { + "epoch": 0.7382098171318575, + "grad_norm": 1.6701131410869008, + "learning_rate": 6.9986352733332886e-06, + "loss": 0.6678, + "step": 767 + }, + { + "epoch": 0.739172281039461, + "grad_norm": 1.1551685655682593, + "learning_rate": 6.991703994506762e-06, + "loss": 0.7752, + "step": 768 + }, + { + "epoch": 0.7401347449470644, + "grad_norm": 1.1070406789292513, + "learning_rate": 6.9847681633540744e-06, + "loss": 0.7708, + "step": 769 + }, + { + "epoch": 0.7410972088546679, + "grad_norm": 1.203224921735285, + "learning_rate": 6.9778277957280626e-06, + "loss": 0.8553, + "step": 770 + }, + { + "epoch": 0.7420596727622714, + "grad_norm": 1.0943976827918653, + "learning_rate": 6.970882907491939e-06, + "loss": 0.7537, + "step": 771 + }, + { + "epoch": 0.7430221366698749, + "grad_norm": 0.9067338503140381, + "learning_rate": 6.963933514519244e-06, + "loss": 0.5425, + "step": 772 + }, + { + "epoch": 0.7439846005774784, + "grad_norm": 0.9521415789410606, + "learning_rate": 6.956979632693814e-06, + "loss": 0.5809, + "step": 773 + }, + { + "epoch": 0.7449470644850819, + "grad_norm": 1.046742914247594, + "learning_rate": 6.95002127790975e-06, + "loss": 0.5139, + "step": 774 + }, + { + "epoch": 0.7459095283926853, + "grad_norm": 1.1032451906264453, + "learning_rate": 6.943058466071371e-06, + "loss": 0.6949, + "step": 775 + }, + { + "epoch": 0.7468719923002888, + "grad_norm": 1.122725683318, + "learning_rate": 6.936091213093188e-06, + "loss": 0.8494, + "step": 776 + }, + { + "epoch": 0.7478344562078922, + "grad_norm": 1.2027564014738539, + "learning_rate": 6.92911953489986e-06, + "loss": 0.8449, + "step": 777 + }, + { + "epoch": 0.7487969201154957, + "grad_norm": 1.1235559618485227, + "learning_rate": 6.9221434474261574e-06, + "loss": 0.8304, + "step": 778 + }, + { + "epoch": 0.7497593840230992, + "grad_norm": 1.134558652336569, + "learning_rate": 6.915162966616935e-06, + "loss": 0.7743, + "step": 779 + }, + { + "epoch": 0.7507218479307026, + "grad_norm": 1.1107602147628057, + "learning_rate": 6.908178108427088e-06, + "loss": 0.845, + "step": 780 + }, + { + "epoch": 0.7516843118383061, + "grad_norm": 1.1057078653688215, + "learning_rate": 6.901188888821514e-06, + "loss": 0.7844, + "step": 781 + }, + { + "epoch": 0.7526467757459095, + "grad_norm": 1.2823516431823678, + "learning_rate": 6.894195323775078e-06, + "loss": 0.8326, + "step": 782 + }, + { + "epoch": 0.753609239653513, + "grad_norm": 1.1166941902716063, + "learning_rate": 6.887197429272583e-06, + "loss": 0.8086, + "step": 783 + }, + { + "epoch": 0.7545717035611165, + "grad_norm": 1.0188682903992827, + "learning_rate": 6.880195221308723e-06, + "loss": 0.7272, + "step": 784 + }, + { + "epoch": 0.7555341674687199, + "grad_norm": 1.033314641014155, + "learning_rate": 6.873188715888051e-06, + "loss": 0.7212, + "step": 785 + }, + { + "epoch": 0.7564966313763234, + "grad_norm": 0.9939108457046145, + "learning_rate": 6.8661779290249454e-06, + "loss": 0.6677, + "step": 786 + }, + { + "epoch": 0.7574590952839269, + "grad_norm": 1.2042655274173144, + "learning_rate": 6.859162876743566e-06, + "loss": 0.8583, + "step": 787 + }, + { + "epoch": 0.7584215591915303, + "grad_norm": 1.1870136614499838, + "learning_rate": 6.852143575077827e-06, + "loss": 0.8889, + "step": 788 + }, + { + "epoch": 0.7593840230991338, + "grad_norm": 0.8803059951710882, + "learning_rate": 6.845120040071354e-06, + "loss": 0.5733, + "step": 789 + }, + { + "epoch": 0.7603464870067372, + "grad_norm": 1.146532718625088, + "learning_rate": 6.838092287777444e-06, + "loss": 0.8838, + "step": 790 + }, + { + "epoch": 0.7613089509143407, + "grad_norm": 1.028511705298473, + "learning_rate": 6.831060334259038e-06, + "loss": 0.6809, + "step": 791 + }, + { + "epoch": 0.7622714148219442, + "grad_norm": 0.9617534785213463, + "learning_rate": 6.824024195588678e-06, + "loss": 0.5969, + "step": 792 + }, + { + "epoch": 0.7632338787295476, + "grad_norm": 1.0348494762821898, + "learning_rate": 6.816983887848471e-06, + "loss": 0.7204, + "step": 793 + }, + { + "epoch": 0.7641963426371511, + "grad_norm": 1.1275332000045741, + "learning_rate": 6.809939427130053e-06, + "loss": 0.8337, + "step": 794 + }, + { + "epoch": 0.7651588065447545, + "grad_norm": 1.1837092258406767, + "learning_rate": 6.802890829534557e-06, + "loss": 0.8264, + "step": 795 + }, + { + "epoch": 0.766121270452358, + "grad_norm": 1.0405566144515346, + "learning_rate": 6.795838111172561e-06, + "loss": 0.7856, + "step": 796 + }, + { + "epoch": 0.7670837343599615, + "grad_norm": 1.0303534287429046, + "learning_rate": 6.788781288164073e-06, + "loss": 0.6909, + "step": 797 + }, + { + "epoch": 0.7680461982675649, + "grad_norm": 0.9954943551050333, + "learning_rate": 6.7817203766384765e-06, + "loss": 0.7023, + "step": 798 + }, + { + "epoch": 0.7690086621751684, + "grad_norm": 1.07010195230496, + "learning_rate": 6.7746553927345e-06, + "loss": 0.7911, + "step": 799 + }, + { + "epoch": 0.7699711260827719, + "grad_norm": 1.142168538374647, + "learning_rate": 6.767586352600183e-06, + "loss": 0.7485, + "step": 800 + }, + { + "epoch": 0.7709335899903753, + "grad_norm": 1.0995432211838443, + "learning_rate": 6.76051327239283e-06, + "loss": 0.7772, + "step": 801 + }, + { + "epoch": 0.7718960538979788, + "grad_norm": 0.9939745406480389, + "learning_rate": 6.753436168278991e-06, + "loss": 0.6771, + "step": 802 + }, + { + "epoch": 0.7728585178055823, + "grad_norm": 1.0893704800868063, + "learning_rate": 6.746355056434399e-06, + "loss": 0.8199, + "step": 803 + }, + { + "epoch": 0.7738209817131858, + "grad_norm": 3.730166975921003, + "learning_rate": 6.739269953043959e-06, + "loss": 0.7816, + "step": 804 + }, + { + "epoch": 0.7747834456207893, + "grad_norm": 1.0846959857196472, + "learning_rate": 6.732180874301694e-06, + "loss": 0.7876, + "step": 805 + }, + { + "epoch": 0.7757459095283927, + "grad_norm": 1.0743361146124417, + "learning_rate": 6.725087836410711e-06, + "loss": 0.7759, + "step": 806 + }, + { + "epoch": 0.7767083734359962, + "grad_norm": 1.1346855609777062, + "learning_rate": 6.717990855583172e-06, + "loss": 0.8015, + "step": 807 + }, + { + "epoch": 0.7776708373435997, + "grad_norm": 1.1090479815631147, + "learning_rate": 6.7108899480402476e-06, + "loss": 0.7641, + "step": 808 + }, + { + "epoch": 0.7786333012512031, + "grad_norm": 1.0811679191564731, + "learning_rate": 6.703785130012084e-06, + "loss": 0.802, + "step": 809 + }, + { + "epoch": 0.7795957651588066, + "grad_norm": 1.0895556090117984, + "learning_rate": 6.696676417737764e-06, + "loss": 0.7701, + "step": 810 + }, + { + "epoch": 0.78055822906641, + "grad_norm": 1.178635995400632, + "learning_rate": 6.689563827465275e-06, + "loss": 0.8904, + "step": 811 + }, + { + "epoch": 0.7815206929740135, + "grad_norm": 1.024018951620854, + "learning_rate": 6.682447375451464e-06, + "loss": 0.6923, + "step": 812 + }, + { + "epoch": 0.782483156881617, + "grad_norm": 1.2067726157746783, + "learning_rate": 6.6753270779620085e-06, + "loss": 0.8745, + "step": 813 + }, + { + "epoch": 0.7834456207892204, + "grad_norm": 1.0541496397099386, + "learning_rate": 6.668202951271371e-06, + "loss": 0.7204, + "step": 814 + }, + { + "epoch": 0.7844080846968239, + "grad_norm": 1.081421932200572, + "learning_rate": 6.661075011662769e-06, + "loss": 0.7028, + "step": 815 + }, + { + "epoch": 0.7853705486044273, + "grad_norm": 1.522907234647011, + "learning_rate": 6.653943275428135e-06, + "loss": 0.7595, + "step": 816 + }, + { + "epoch": 0.7863330125120308, + "grad_norm": 1.1977817587993846, + "learning_rate": 6.64680775886808e-06, + "loss": 0.8762, + "step": 817 + }, + { + "epoch": 0.7872954764196343, + "grad_norm": 1.0462388963367621, + "learning_rate": 6.6396684782918515e-06, + "loss": 0.7419, + "step": 818 + }, + { + "epoch": 0.7882579403272377, + "grad_norm": 9.926784304826983, + "learning_rate": 6.632525450017305e-06, + "loss": 0.7233, + "step": 819 + }, + { + "epoch": 0.7892204042348412, + "grad_norm": 0.9490859148535667, + "learning_rate": 6.625378690370859e-06, + "loss": 0.6132, + "step": 820 + }, + { + "epoch": 0.7901828681424446, + "grad_norm": 1.2127527386837722, + "learning_rate": 6.61822821568746e-06, + "loss": 0.8236, + "step": 821 + }, + { + "epoch": 0.7911453320500481, + "grad_norm": 1.0950690108095604, + "learning_rate": 6.61107404231055e-06, + "loss": 0.6841, + "step": 822 + }, + { + "epoch": 0.7921077959576516, + "grad_norm": 1.1546314579385493, + "learning_rate": 6.60391618659202e-06, + "loss": 0.8897, + "step": 823 + }, + { + "epoch": 0.793070259865255, + "grad_norm": 1.1745332277770646, + "learning_rate": 6.596754664892179e-06, + "loss": 0.7812, + "step": 824 + }, + { + "epoch": 0.7940327237728585, + "grad_norm": 1.0598561099434312, + "learning_rate": 6.589589493579718e-06, + "loss": 0.7702, + "step": 825 + }, + { + "epoch": 0.794995187680462, + "grad_norm": 1.1979764527887522, + "learning_rate": 6.582420689031665e-06, + "loss": 0.8824, + "step": 826 + }, + { + "epoch": 0.7959576515880654, + "grad_norm": 0.9525390818453482, + "learning_rate": 6.575248267633357e-06, + "loss": 0.6299, + "step": 827 + }, + { + "epoch": 0.7969201154956689, + "grad_norm": 1.0465164299475032, + "learning_rate": 6.568072245778394e-06, + "loss": 0.7027, + "step": 828 + }, + { + "epoch": 0.7978825794032723, + "grad_norm": 1.129628172668072, + "learning_rate": 6.560892639868605e-06, + "loss": 0.7941, + "step": 829 + }, + { + "epoch": 0.7988450433108758, + "grad_norm": 1.3700738609565286, + "learning_rate": 6.553709466314017e-06, + "loss": 0.6013, + "step": 830 + }, + { + "epoch": 0.7998075072184793, + "grad_norm": 1.1414705058247245, + "learning_rate": 6.546522741532803e-06, + "loss": 0.7657, + "step": 831 + }, + { + "epoch": 0.8007699711260827, + "grad_norm": 1.2145710770958322, + "learning_rate": 6.539332481951262e-06, + "loss": 0.8614, + "step": 832 + }, + { + "epoch": 0.8017324350336862, + "grad_norm": 1.14182144918987, + "learning_rate": 6.532138704003762e-06, + "loss": 0.8511, + "step": 833 + }, + { + "epoch": 0.8026948989412896, + "grad_norm": 0.9219002380457963, + "learning_rate": 6.524941424132719e-06, + "loss": 0.5929, + "step": 834 + }, + { + "epoch": 0.8036573628488932, + "grad_norm": 1.2239060896767116, + "learning_rate": 6.517740658788554e-06, + "loss": 0.8445, + "step": 835 + }, + { + "epoch": 0.8046198267564967, + "grad_norm": 1.081240363391187, + "learning_rate": 6.510536424429652e-06, + "loss": 0.7296, + "step": 836 + }, + { + "epoch": 0.8055822906641001, + "grad_norm": 1.0254744164922482, + "learning_rate": 6.5033287375223275e-06, + "loss": 0.6807, + "step": 837 + }, + { + "epoch": 0.8065447545717036, + "grad_norm": 1.1119857511988334, + "learning_rate": 6.496117614540786e-06, + "loss": 0.7843, + "step": 838 + }, + { + "epoch": 0.8075072184793071, + "grad_norm": 1.10090407609044, + "learning_rate": 6.488903071967087e-06, + "loss": 0.7475, + "step": 839 + }, + { + "epoch": 0.8084696823869105, + "grad_norm": 1.1672230033866757, + "learning_rate": 6.481685126291106e-06, + "loss": 0.731, + "step": 840 + }, + { + "epoch": 0.809432146294514, + "grad_norm": 0.9972853606530193, + "learning_rate": 6.474463794010498e-06, + "loss": 0.6517, + "step": 841 + }, + { + "epoch": 0.8103946102021174, + "grad_norm": 1.10038678246642, + "learning_rate": 6.467239091630657e-06, + "loss": 0.7593, + "step": 842 + }, + { + "epoch": 0.8113570741097209, + "grad_norm": 1.064889129954703, + "learning_rate": 6.46001103566468e-06, + "loss": 0.7213, + "step": 843 + }, + { + "epoch": 0.8123195380173244, + "grad_norm": 1.2002246913093193, + "learning_rate": 6.4527796426333265e-06, + "loss": 0.8547, + "step": 844 + }, + { + "epoch": 0.8132820019249278, + "grad_norm": 1.0814836463027706, + "learning_rate": 6.44554492906499e-06, + "loss": 0.6567, + "step": 845 + }, + { + "epoch": 0.8142444658325313, + "grad_norm": 2.6146410481866833, + "learning_rate": 6.438306911495649e-06, + "loss": 0.6941, + "step": 846 + }, + { + "epoch": 0.8152069297401348, + "grad_norm": 0.7658809572734653, + "learning_rate": 6.431065606468833e-06, + "loss": 0.4637, + "step": 847 + }, + { + "epoch": 0.8161693936477382, + "grad_norm": 0.9994472774168446, + "learning_rate": 6.423821030535586e-06, + "loss": 0.641, + "step": 848 + }, + { + "epoch": 0.8171318575553417, + "grad_norm": 1.1489288485851665, + "learning_rate": 6.416573200254433e-06, + "loss": 0.8127, + "step": 849 + }, + { + "epoch": 0.8180943214629451, + "grad_norm": 0.9017408363055436, + "learning_rate": 6.409322132191331e-06, + "loss": 0.612, + "step": 850 + }, + { + "epoch": 0.8190567853705486, + "grad_norm": 1.2224877865697046, + "learning_rate": 6.402067842919641e-06, + "loss": 0.6458, + "step": 851 + }, + { + "epoch": 0.8200192492781521, + "grad_norm": 1.0662165419392151, + "learning_rate": 6.394810349020083e-06, + "loss": 0.7514, + "step": 852 + }, + { + "epoch": 0.8209817131857555, + "grad_norm": 150.13663221544246, + "learning_rate": 6.387549667080704e-06, + "loss": 1.1108, + "step": 853 + }, + { + "epoch": 0.821944177093359, + "grad_norm": 1.0855179457671091, + "learning_rate": 6.38028581369684e-06, + "loss": 0.7755, + "step": 854 + }, + { + "epoch": 0.8229066410009624, + "grad_norm": 0.8746616725970394, + "learning_rate": 6.373018805471073e-06, + "loss": 0.5174, + "step": 855 + }, + { + "epoch": 0.8238691049085659, + "grad_norm": 1.0960881051162532, + "learning_rate": 6.365748659013195e-06, + "loss": 0.7804, + "step": 856 + }, + { + "epoch": 0.8248315688161694, + "grad_norm": 0.9259726682086828, + "learning_rate": 6.358475390940172e-06, + "loss": 0.609, + "step": 857 + }, + { + "epoch": 0.8257940327237728, + "grad_norm": 1.0277136343289939, + "learning_rate": 6.351199017876106e-06, + "loss": 0.7031, + "step": 858 + }, + { + "epoch": 0.8267564966313763, + "grad_norm": 1.0558196787610592, + "learning_rate": 6.343919556452196e-06, + "loss": 0.7833, + "step": 859 + }, + { + "epoch": 0.8277189605389798, + "grad_norm": 1.051412433968949, + "learning_rate": 6.336637023306698e-06, + "loss": 0.6985, + "step": 860 + }, + { + "epoch": 0.8286814244465832, + "grad_norm": 1.0890700317300475, + "learning_rate": 6.32935143508489e-06, + "loss": 0.7618, + "step": 861 + }, + { + "epoch": 0.8296438883541867, + "grad_norm": 1.0800061046847536, + "learning_rate": 6.322062808439029e-06, + "loss": 0.7775, + "step": 862 + }, + { + "epoch": 0.8306063522617901, + "grad_norm": 1.0958234203859165, + "learning_rate": 6.314771160028325e-06, + "loss": 0.753, + "step": 863 + }, + { + "epoch": 0.8315688161693936, + "grad_norm": 1.0729187489262533, + "learning_rate": 6.307476506518889e-06, + "loss": 0.7418, + "step": 864 + }, + { + "epoch": 0.8325312800769971, + "grad_norm": 1.0909453279611865, + "learning_rate": 6.300178864583701e-06, + "loss": 0.7003, + "step": 865 + }, + { + "epoch": 0.8334937439846006, + "grad_norm": 1.2378532899776926, + "learning_rate": 6.2928782509025695e-06, + "loss": 0.86, + "step": 866 + }, + { + "epoch": 0.8344562078922041, + "grad_norm": 1.0571733229226106, + "learning_rate": 6.285574682162103e-06, + "loss": 0.7165, + "step": 867 + }, + { + "epoch": 0.8354186717998076, + "grad_norm": 0.9541466276762053, + "learning_rate": 6.278268175055655e-06, + "loss": 0.7094, + "step": 868 + }, + { + "epoch": 0.836381135707411, + "grad_norm": 1.0043580309456042, + "learning_rate": 6.270958746283302e-06, + "loss": 0.659, + "step": 869 + }, + { + "epoch": 0.8373435996150145, + "grad_norm": 1.1061845842756033, + "learning_rate": 6.2636464125517935e-06, + "loss": 0.7487, + "step": 870 + }, + { + "epoch": 0.8383060635226179, + "grad_norm": 1.1956052803725408, + "learning_rate": 6.2563311905745225e-06, + "loss": 0.8813, + "step": 871 + }, + { + "epoch": 0.8392685274302214, + "grad_norm": 0.9685229133315701, + "learning_rate": 6.2490130970714805e-06, + "loss": 0.6547, + "step": 872 + }, + { + "epoch": 0.8402309913378249, + "grad_norm": 0.9989234046325549, + "learning_rate": 6.241692148769223e-06, + "loss": 0.6786, + "step": 873 + }, + { + "epoch": 0.8411934552454283, + "grad_norm": 1.0202210443925335, + "learning_rate": 6.234368362400833e-06, + "loss": 0.721, + "step": 874 + }, + { + "epoch": 0.8421559191530318, + "grad_norm": 1.0079791790005828, + "learning_rate": 6.227041754705877e-06, + "loss": 0.6858, + "step": 875 + }, + { + "epoch": 0.8431183830606352, + "grad_norm": 1.1517515740640052, + "learning_rate": 6.2197123424303715e-06, + "loss": 0.8363, + "step": 876 + }, + { + "epoch": 0.8440808469682387, + "grad_norm": 1.1037388897988851, + "learning_rate": 6.212380142326744e-06, + "loss": 0.7581, + "step": 877 + }, + { + "epoch": 0.8450433108758422, + "grad_norm": 1.1977925754323524, + "learning_rate": 6.205045171153793e-06, + "loss": 0.8818, + "step": 878 + }, + { + "epoch": 0.8460057747834456, + "grad_norm": 0.8726575572144369, + "learning_rate": 6.19770744567665e-06, + "loss": 0.5384, + "step": 879 + }, + { + "epoch": 0.8469682386910491, + "grad_norm": 5.852798955226513, + "learning_rate": 6.1903669826667445e-06, + "loss": 0.677, + "step": 880 + }, + { + "epoch": 0.8479307025986526, + "grad_norm": 1.0869426728571296, + "learning_rate": 6.18302379890176e-06, + "loss": 0.7366, + "step": 881 + }, + { + "epoch": 0.848893166506256, + "grad_norm": 1.0621046982015252, + "learning_rate": 6.175677911165599e-06, + "loss": 0.6001, + "step": 882 + }, + { + "epoch": 0.8498556304138595, + "grad_norm": 0.9230472407840359, + "learning_rate": 6.168329336248349e-06, + "loss": 0.5791, + "step": 883 + }, + { + "epoch": 0.8508180943214629, + "grad_norm": 1.0946310027756285, + "learning_rate": 6.160978090946231e-06, + "loss": 0.819, + "step": 884 + }, + { + "epoch": 0.8517805582290664, + "grad_norm": 1.162806923469983, + "learning_rate": 6.153624192061578e-06, + "loss": 0.7612, + "step": 885 + }, + { + "epoch": 0.8527430221366699, + "grad_norm": 1.2628266575643763, + "learning_rate": 6.1462676564027814e-06, + "loss": 0.886, + "step": 886 + }, + { + "epoch": 0.8537054860442733, + "grad_norm": 1.0234740885653542, + "learning_rate": 6.138908500784265e-06, + "loss": 0.7264, + "step": 887 + }, + { + "epoch": 0.8546679499518768, + "grad_norm": 1.0937359360858299, + "learning_rate": 6.131546742026438e-06, + "loss": 0.767, + "step": 888 + }, + { + "epoch": 0.8556304138594802, + "grad_norm": 1.0391570116520215, + "learning_rate": 6.124182396955657e-06, + "loss": 0.7489, + "step": 889 + }, + { + "epoch": 0.8565928777670837, + "grad_norm": 1.1233130383911616, + "learning_rate": 6.116815482404194e-06, + "loss": 0.7748, + "step": 890 + }, + { + "epoch": 0.8575553416746872, + "grad_norm": 0.9548797697599618, + "learning_rate": 6.109446015210194e-06, + "loss": 0.6225, + "step": 891 + }, + { + "epoch": 0.8585178055822906, + "grad_norm": 1.317604042235006, + "learning_rate": 6.102074012217635e-06, + "loss": 0.8921, + "step": 892 + }, + { + "epoch": 0.8594802694898941, + "grad_norm": 0.9388934415665238, + "learning_rate": 6.094699490276289e-06, + "loss": 0.6047, + "step": 893 + }, + { + "epoch": 0.8604427333974976, + "grad_norm": 1.0564748710321863, + "learning_rate": 6.08732246624169e-06, + "loss": 0.7283, + "step": 894 + }, + { + "epoch": 0.861405197305101, + "grad_norm": 1.0978368528557816, + "learning_rate": 6.0799429569750836e-06, + "loss": 0.7734, + "step": 895 + }, + { + "epoch": 0.8623676612127045, + "grad_norm": 1.122133441033741, + "learning_rate": 6.072560979343405e-06, + "loss": 0.8263, + "step": 896 + }, + { + "epoch": 0.8633301251203079, + "grad_norm": 1.151407608846726, + "learning_rate": 6.065176550219226e-06, + "loss": 0.8077, + "step": 897 + }, + { + "epoch": 0.8642925890279115, + "grad_norm": 1.0054253622175293, + "learning_rate": 6.05778968648072e-06, + "loss": 0.7072, + "step": 898 + }, + { + "epoch": 0.865255052935515, + "grad_norm": 1.0722892576892684, + "learning_rate": 6.050400405011629e-06, + "loss": 0.7966, + "step": 899 + }, + { + "epoch": 0.8662175168431184, + "grad_norm": 0.8865200058896751, + "learning_rate": 6.04300872270122e-06, + "loss": 0.5508, + "step": 900 + }, + { + "epoch": 0.8671799807507219, + "grad_norm": 1.1285799421305356, + "learning_rate": 6.035614656444244e-06, + "loss": 0.775, + "step": 901 + }, + { + "epoch": 0.8681424446583254, + "grad_norm": 3.122430297179407, + "learning_rate": 6.028218223140908e-06, + "loss": 0.8119, + "step": 902 + }, + { + "epoch": 0.8691049085659288, + "grad_norm": 1.0529374322764498, + "learning_rate": 6.020819439696823e-06, + "loss": 0.7805, + "step": 903 + }, + { + "epoch": 0.8700673724735323, + "grad_norm": 1.1283140964686826, + "learning_rate": 6.013418323022971e-06, + "loss": 0.7925, + "step": 904 + }, + { + "epoch": 0.8710298363811357, + "grad_norm": 0.9635197983076199, + "learning_rate": 6.006014890035672e-06, + "loss": 0.6619, + "step": 905 + }, + { + "epoch": 0.8719923002887392, + "grad_norm": 1.1121202745637881, + "learning_rate": 5.998609157656538e-06, + "loss": 0.7864, + "step": 906 + }, + { + "epoch": 0.8729547641963427, + "grad_norm": 1.2558977049775057, + "learning_rate": 5.991201142812436e-06, + "loss": 0.8709, + "step": 907 + }, + { + "epoch": 0.8739172281039461, + "grad_norm": 1.0242168059499188, + "learning_rate": 5.98379086243545e-06, + "loss": 0.6738, + "step": 908 + }, + { + "epoch": 0.8748796920115496, + "grad_norm": 1.1275265454620331, + "learning_rate": 5.976378333462843e-06, + "loss": 0.7703, + "step": 909 + }, + { + "epoch": 0.875842155919153, + "grad_norm": 0.9987336722083385, + "learning_rate": 5.968963572837015e-06, + "loss": 0.6371, + "step": 910 + }, + { + "epoch": 0.8768046198267565, + "grad_norm": 0.9692257849023777, + "learning_rate": 5.961546597505469e-06, + "loss": 0.5718, + "step": 911 + }, + { + "epoch": 0.87776708373436, + "grad_norm": 1.288051598887658, + "learning_rate": 5.954127424420774e-06, + "loss": 0.8915, + "step": 912 + }, + { + "epoch": 0.8787295476419634, + "grad_norm": 0.9761282715266751, + "learning_rate": 5.9467060705405124e-06, + "loss": 0.6925, + "step": 913 + }, + { + "epoch": 0.8796920115495669, + "grad_norm": 1.1758335711575099, + "learning_rate": 5.9392825528272575e-06, + "loss": 0.8613, + "step": 914 + }, + { + "epoch": 0.8806544754571703, + "grad_norm": 1.000422960240342, + "learning_rate": 5.9318568882485315e-06, + "loss": 0.6755, + "step": 915 + }, + { + "epoch": 0.8816169393647738, + "grad_norm": 1.0000742856261244, + "learning_rate": 5.924429093776757e-06, + "loss": 0.685, + "step": 916 + }, + { + "epoch": 0.8825794032723773, + "grad_norm": 1.1310967496971185, + "learning_rate": 5.916999186389227e-06, + "loss": 0.8011, + "step": 917 + }, + { + "epoch": 0.8835418671799807, + "grad_norm": 4.891808106562346, + "learning_rate": 5.909567183068065e-06, + "loss": 0.776, + "step": 918 + }, + { + "epoch": 0.8845043310875842, + "grad_norm": 1.0842412540487447, + "learning_rate": 5.9021331008001836e-06, + "loss": 0.8108, + "step": 919 + }, + { + "epoch": 0.8854667949951877, + "grad_norm": 1.1824360591562886, + "learning_rate": 5.89469695657725e-06, + "loss": 0.8577, + "step": 920 + }, + { + "epoch": 0.8864292589027911, + "grad_norm": 1.0689413631846516, + "learning_rate": 5.887258767395639e-06, + "loss": 0.79, + "step": 921 + }, + { + "epoch": 0.8873917228103946, + "grad_norm": 0.933489535897218, + "learning_rate": 5.879818550256405e-06, + "loss": 0.5752, + "step": 922 + }, + { + "epoch": 0.888354186717998, + "grad_norm": 1.109133472795779, + "learning_rate": 5.872376322165234e-06, + "loss": 0.8195, + "step": 923 + }, + { + "epoch": 0.8893166506256015, + "grad_norm": 1.072964977713411, + "learning_rate": 5.864932100132411e-06, + "loss": 0.7742, + "step": 924 + }, + { + "epoch": 0.890279114533205, + "grad_norm": 0.9802307787117865, + "learning_rate": 5.857485901172776e-06, + "loss": 0.6619, + "step": 925 + }, + { + "epoch": 0.8912415784408084, + "grad_norm": 0.8534249730549125, + "learning_rate": 5.850037742305687e-06, + "loss": 0.4989, + "step": 926 + }, + { + "epoch": 0.8922040423484119, + "grad_norm": 1.1040173040565147, + "learning_rate": 5.842587640554987e-06, + "loss": 0.7481, + "step": 927 + }, + { + "epoch": 0.8931665062560153, + "grad_norm": 1.1520024366184813, + "learning_rate": 5.835135612948951e-06, + "loss": 0.7867, + "step": 928 + }, + { + "epoch": 0.8941289701636189, + "grad_norm": 1.0143985561774662, + "learning_rate": 5.827681676520264e-06, + "loss": 0.6679, + "step": 929 + }, + { + "epoch": 0.8950914340712224, + "grad_norm": 0.9639400887686717, + "learning_rate": 5.820225848305971e-06, + "loss": 0.6134, + "step": 930 + }, + { + "epoch": 0.8960538979788258, + "grad_norm": 1.1085801375971507, + "learning_rate": 5.812768145347438e-06, + "loss": 0.8267, + "step": 931 + }, + { + "epoch": 0.8970163618864293, + "grad_norm": 1.095199382862551, + "learning_rate": 5.805308584690322e-06, + "loss": 0.6494, + "step": 932 + }, + { + "epoch": 0.8979788257940328, + "grad_norm": 1.075318321110228, + "learning_rate": 5.7978471833845194e-06, + "loss": 0.7691, + "step": 933 + }, + { + "epoch": 0.8989412897016362, + "grad_norm": 1.0182890897617227, + "learning_rate": 5.790383958484138e-06, + "loss": 0.695, + "step": 934 + }, + { + "epoch": 0.8999037536092397, + "grad_norm": 1.1817194448422146, + "learning_rate": 5.782918927047453e-06, + "loss": 0.8624, + "step": 935 + }, + { + "epoch": 0.9008662175168431, + "grad_norm": 1.0111925601099048, + "learning_rate": 5.775452106136868e-06, + "loss": 0.6927, + "step": 936 + }, + { + "epoch": 0.9018286814244466, + "grad_norm": 0.9017415657060819, + "learning_rate": 5.7679835128188776e-06, + "loss": 0.5851, + "step": 937 + }, + { + "epoch": 0.9027911453320501, + "grad_norm": 1.1370134163155934, + "learning_rate": 5.760513164164025e-06, + "loss": 0.8699, + "step": 938 + }, + { + "epoch": 0.9037536092396535, + "grad_norm": 1.1706123066674343, + "learning_rate": 5.753041077246868e-06, + "loss": 0.8652, + "step": 939 + }, + { + "epoch": 0.904716073147257, + "grad_norm": 1.0495491296507502, + "learning_rate": 5.745567269145939e-06, + "loss": 0.6972, + "step": 940 + }, + { + "epoch": 0.9056785370548605, + "grad_norm": 1.2205197152956557, + "learning_rate": 5.738091756943699e-06, + "loss": 0.8547, + "step": 941 + }, + { + "epoch": 0.9066410009624639, + "grad_norm": 1.1024739277306173, + "learning_rate": 5.7306145577265095e-06, + "loss": 0.7352, + "step": 942 + }, + { + "epoch": 0.9076034648700674, + "grad_norm": 0.8472398249109147, + "learning_rate": 5.723135688584583e-06, + "loss": 0.5333, + "step": 943 + }, + { + "epoch": 0.9085659287776708, + "grad_norm": 0.8230389037107257, + "learning_rate": 5.7156551666119554e-06, + "loss": 0.4811, + "step": 944 + }, + { + "epoch": 0.9095283926852743, + "grad_norm": 1.2058104625760522, + "learning_rate": 5.708173008906432e-06, + "loss": 0.8386, + "step": 945 + }, + { + "epoch": 0.9104908565928778, + "grad_norm": 0.8695529912918513, + "learning_rate": 5.700689232569565e-06, + "loss": 0.5283, + "step": 946 + }, + { + "epoch": 0.9114533205004812, + "grad_norm": 1.182407968974184, + "learning_rate": 5.6932038547066e-06, + "loss": 0.8283, + "step": 947 + }, + { + "epoch": 0.9124157844080847, + "grad_norm": 1.0366426897135523, + "learning_rate": 5.685716892426445e-06, + "loss": 0.6874, + "step": 948 + }, + { + "epoch": 0.9133782483156881, + "grad_norm": 1.1256249044714506, + "learning_rate": 5.678228362841633e-06, + "loss": 0.8405, + "step": 949 + }, + { + "epoch": 0.9143407122232916, + "grad_norm": 1.0235837622529895, + "learning_rate": 5.670738283068273e-06, + "loss": 0.6942, + "step": 950 + }, + { + "epoch": 0.9153031761308951, + "grad_norm": 1.068465215307631, + "learning_rate": 5.6632466702260235e-06, + "loss": 0.7795, + "step": 951 + }, + { + "epoch": 0.9162656400384985, + "grad_norm": 1.1609522575742557, + "learning_rate": 5.655753541438041e-06, + "loss": 0.8161, + "step": 952 + }, + { + "epoch": 0.917228103946102, + "grad_norm": 1.140541853977165, + "learning_rate": 5.648258913830952e-06, + "loss": 0.8566, + "step": 953 + }, + { + "epoch": 0.9181905678537055, + "grad_norm": 1.1143439073610513, + "learning_rate": 5.640762804534806e-06, + "loss": 0.7824, + "step": 954 + }, + { + "epoch": 0.9191530317613089, + "grad_norm": 1.1706884952241303, + "learning_rate": 5.633265230683039e-06, + "loss": 0.8485, + "step": 955 + }, + { + "epoch": 0.9201154956689124, + "grad_norm": 1.13378450606679, + "learning_rate": 5.625766209412436e-06, + "loss": 0.8494, + "step": 956 + }, + { + "epoch": 0.9210779595765158, + "grad_norm": 0.9090653181903438, + "learning_rate": 5.6182657578630896e-06, + "loss": 0.5813, + "step": 957 + }, + { + "epoch": 0.9220404234841193, + "grad_norm": 1.020543215261829, + "learning_rate": 5.610763893178361e-06, + "loss": 0.7124, + "step": 958 + }, + { + "epoch": 0.9230028873917228, + "grad_norm": 1.0940185566137863, + "learning_rate": 5.603260632504843e-06, + "loss": 0.7701, + "step": 959 + }, + { + "epoch": 0.9239653512993262, + "grad_norm": 0.9922637920932359, + "learning_rate": 5.595755992992316e-06, + "loss": 0.697, + "step": 960 + }, + { + "epoch": 0.9249278152069298, + "grad_norm": 1.1905745629463385, + "learning_rate": 5.588249991793715e-06, + "loss": 0.8611, + "step": 961 + }, + { + "epoch": 0.9258902791145333, + "grad_norm": 1.123400119366617, + "learning_rate": 5.580742646065085e-06, + "loss": 0.7639, + "step": 962 + }, + { + "epoch": 0.9268527430221367, + "grad_norm": 1.0380832293158944, + "learning_rate": 5.5732339729655475e-06, + "loss": 0.6634, + "step": 963 + }, + { + "epoch": 0.9278152069297402, + "grad_norm": 1.1746830302942537, + "learning_rate": 5.565723989657253e-06, + "loss": 0.763, + "step": 964 + }, + { + "epoch": 0.9287776708373436, + "grad_norm": 1.0312987077954296, + "learning_rate": 5.558212713305352e-06, + "loss": 0.6287, + "step": 965 + }, + { + "epoch": 0.9297401347449471, + "grad_norm": 1.196837724416088, + "learning_rate": 5.550700161077946e-06, + "loss": 0.841, + "step": 966 + }, + { + "epoch": 0.9307025986525506, + "grad_norm": 0.9688396612014245, + "learning_rate": 5.5431863501460534e-06, + "loss": 0.6622, + "step": 967 + }, + { + "epoch": 0.931665062560154, + "grad_norm": 1.0516621228820584, + "learning_rate": 5.535671297683574e-06, + "loss": 0.752, + "step": 968 + }, + { + "epoch": 0.9326275264677575, + "grad_norm": 1.212865489631781, + "learning_rate": 5.528155020867239e-06, + "loss": 0.8576, + "step": 969 + }, + { + "epoch": 0.933589990375361, + "grad_norm": 1.0814761723972997, + "learning_rate": 5.520637536876582e-06, + "loss": 0.7897, + "step": 970 + }, + { + "epoch": 0.9345524542829644, + "grad_norm": 1.2264653522865583, + "learning_rate": 5.513118862893894e-06, + "loss": 0.8069, + "step": 971 + }, + { + "epoch": 0.9355149181905679, + "grad_norm": 1.0765246165880884, + "learning_rate": 5.505599016104187e-06, + "loss": 0.7616, + "step": 972 + }, + { + "epoch": 0.9364773820981713, + "grad_norm": 1.1304694755934617, + "learning_rate": 5.498078013695154e-06, + "loss": 0.8837, + "step": 973 + }, + { + "epoch": 0.9374398460057748, + "grad_norm": 1.0886890744552868, + "learning_rate": 5.490555872857128e-06, + "loss": 0.7549, + "step": 974 + }, + { + "epoch": 0.9384023099133783, + "grad_norm": 1.0175554479049977, + "learning_rate": 5.483032610783046e-06, + "loss": 0.7235, + "step": 975 + }, + { + "epoch": 0.9393647738209817, + "grad_norm": 1.1741933065842058, + "learning_rate": 5.475508244668402e-06, + "loss": 0.8384, + "step": 976 + }, + { + "epoch": 0.9403272377285852, + "grad_norm": 1.0132101044784492, + "learning_rate": 5.467982791711224e-06, + "loss": 0.7199, + "step": 977 + }, + { + "epoch": 0.9412897016361886, + "grad_norm": 1.0793637948230486, + "learning_rate": 5.4604562691120135e-06, + "loss": 0.7979, + "step": 978 + }, + { + "epoch": 0.9422521655437921, + "grad_norm": 1.11455828759669, + "learning_rate": 5.452928694073722e-06, + "loss": 0.7738, + "step": 979 + }, + { + "epoch": 0.9432146294513956, + "grad_norm": 1.097329240851139, + "learning_rate": 5.445400083801705e-06, + "loss": 0.7776, + "step": 980 + }, + { + "epoch": 0.944177093358999, + "grad_norm": 1.1046456622322636, + "learning_rate": 5.437870455503689e-06, + "loss": 0.7952, + "step": 981 + }, + { + "epoch": 0.9451395572666025, + "grad_norm": 1.1504874330174164, + "learning_rate": 5.430339826389719e-06, + "loss": 0.838, + "step": 982 + }, + { + "epoch": 0.9461020211742059, + "grad_norm": 1.1281844801979537, + "learning_rate": 5.422808213672134e-06, + "loss": 0.808, + "step": 983 + }, + { + "epoch": 0.9470644850818094, + "grad_norm": 1.1275785211992444, + "learning_rate": 5.415275634565518e-06, + "loss": 0.8033, + "step": 984 + }, + { + "epoch": 0.9480269489894129, + "grad_norm": 0.9021948595519562, + "learning_rate": 5.407742106286663e-06, + "loss": 0.5947, + "step": 985 + }, + { + "epoch": 0.9489894128970163, + "grad_norm": 0.9923595505627854, + "learning_rate": 5.400207646054536e-06, + "loss": 0.6703, + "step": 986 + }, + { + "epoch": 0.9499518768046198, + "grad_norm": 1.1995385180169738, + "learning_rate": 5.392672271090229e-06, + "loss": 0.8302, + "step": 987 + }, + { + "epoch": 0.9509143407122232, + "grad_norm": 1.0249570180026626, + "learning_rate": 5.385135998616925e-06, + "loss": 0.6866, + "step": 988 + }, + { + "epoch": 0.9518768046198267, + "grad_norm": 1.2041995590759025, + "learning_rate": 5.3775988458598595e-06, + "loss": 0.8, + "step": 989 + }, + { + "epoch": 0.9528392685274302, + "grad_norm": 1.0178864860130796, + "learning_rate": 5.370060830046282e-06, + "loss": 0.6504, + "step": 990 + }, + { + "epoch": 0.9538017324350336, + "grad_norm": 1.0280973932275397, + "learning_rate": 5.362521968405411e-06, + "loss": 0.7139, + "step": 991 + }, + { + "epoch": 0.9547641963426372, + "grad_norm": 1.065107148256385, + "learning_rate": 5.354982278168399e-06, + "loss": 0.753, + "step": 992 + }, + { + "epoch": 0.9557266602502407, + "grad_norm": 0.9244481946046854, + "learning_rate": 5.347441776568296e-06, + "loss": 0.6058, + "step": 993 + }, + { + "epoch": 0.9566891241578441, + "grad_norm": 1.0789135666356284, + "learning_rate": 5.339900480840002e-06, + "loss": 0.7557, + "step": 994 + }, + { + "epoch": 0.9576515880654476, + "grad_norm": 1.0021091177482366, + "learning_rate": 5.332358408220234e-06, + "loss": 0.6821, + "step": 995 + }, + { + "epoch": 0.958614051973051, + "grad_norm": 0.9697566866813903, + "learning_rate": 5.324815575947485e-06, + "loss": 0.6484, + "step": 996 + }, + { + "epoch": 0.9595765158806545, + "grad_norm": 0.9376974758414356, + "learning_rate": 5.317272001261985e-06, + "loss": 0.6125, + "step": 997 + }, + { + "epoch": 0.960538979788258, + "grad_norm": 1.0316097148866545, + "learning_rate": 5.309727701405657e-06, + "loss": 0.6829, + "step": 998 + }, + { + "epoch": 0.9615014436958614, + "grad_norm": 0.9989307097384752, + "learning_rate": 5.302182693622085e-06, + "loss": 0.707, + "step": 999 + }, + { + "epoch": 0.9624639076034649, + "grad_norm": 1.0300657613895878, + "learning_rate": 5.294636995156472e-06, + "loss": 0.781, + "step": 1000 + }, + { + "epoch": 0.9624639076034649, + "eval_loss": 0.7829577922821045, + "eval_runtime": 12.5378, + "eval_samples_per_second": 3.35, + "eval_steps_per_second": 0.877, + "step": 1000 + }, + { + "epoch": 0.9634263715110684, + "grad_norm": 0.9518032836778593, + "learning_rate": 5.287090623255597e-06, + "loss": 0.612, + "step": 1001 + }, + { + "epoch": 0.9643888354186718, + "grad_norm": 1.0946711677877239, + "learning_rate": 5.2795435951677785e-06, + "loss": 0.7611, + "step": 1002 + }, + { + "epoch": 0.9653512993262753, + "grad_norm": 1.0112855205125533, + "learning_rate": 5.271995928142837e-06, + "loss": 0.6795, + "step": 1003 + }, + { + "epoch": 0.9663137632338787, + "grad_norm": 1.0571305636495947, + "learning_rate": 5.26444763943205e-06, + "loss": 0.6803, + "step": 1004 + }, + { + "epoch": 0.9672762271414822, + "grad_norm": 0.9226474810788041, + "learning_rate": 5.256898746288119e-06, + "loss": 0.618, + "step": 1005 + }, + { + "epoch": 0.9682386910490857, + "grad_norm": 1.0531361988343444, + "learning_rate": 5.2493492659651265e-06, + "loss": 0.7164, + "step": 1006 + }, + { + "epoch": 0.9692011549566891, + "grad_norm": 1.1362102562469618, + "learning_rate": 5.241799215718495e-06, + "loss": 0.8673, + "step": 1007 + }, + { + "epoch": 0.9701636188642926, + "grad_norm": 0.9945618469299844, + "learning_rate": 5.234248612804952e-06, + "loss": 0.7004, + "step": 1008 + }, + { + "epoch": 0.971126082771896, + "grad_norm": 1.0256145653600026, + "learning_rate": 5.226697474482486e-06, + "loss": 0.7374, + "step": 1009 + }, + { + "epoch": 0.9720885466794995, + "grad_norm": 1.0623330588642539, + "learning_rate": 5.219145818010312e-06, + "loss": 0.7818, + "step": 1010 + }, + { + "epoch": 0.973051010587103, + "grad_norm": 1.146356274261188, + "learning_rate": 5.211593660648829e-06, + "loss": 0.8201, + "step": 1011 + }, + { + "epoch": 0.9740134744947064, + "grad_norm": 1.16233108197497, + "learning_rate": 5.204041019659577e-06, + "loss": 0.8169, + "step": 1012 + }, + { + "epoch": 0.9749759384023099, + "grad_norm": 1.1496741462052809, + "learning_rate": 5.196487912305203e-06, + "loss": 0.8486, + "step": 1013 + }, + { + "epoch": 0.9759384023099134, + "grad_norm": 1.0002223364620249, + "learning_rate": 5.1889343558494264e-06, + "loss": 0.7049, + "step": 1014 + }, + { + "epoch": 0.9769008662175168, + "grad_norm": 1.1353635927083958, + "learning_rate": 5.181380367556983e-06, + "loss": 0.8204, + "step": 1015 + }, + { + "epoch": 0.9778633301251203, + "grad_norm": 0.9616197248039192, + "learning_rate": 5.173825964693603e-06, + "loss": 0.6843, + "step": 1016 + }, + { + "epoch": 0.9788257940327237, + "grad_norm": 1.0844662730460999, + "learning_rate": 5.16627116452596e-06, + "loss": 0.7619, + "step": 1017 + }, + { + "epoch": 0.9797882579403272, + "grad_norm": 1.1590858810017395, + "learning_rate": 5.158715984321636e-06, + "loss": 0.8485, + "step": 1018 + }, + { + "epoch": 0.9807507218479307, + "grad_norm": 1.1265810300544474, + "learning_rate": 5.151160441349087e-06, + "loss": 0.8306, + "step": 1019 + }, + { + "epoch": 0.9817131857555341, + "grad_norm": 1.0611727384708343, + "learning_rate": 5.14360455287759e-06, + "loss": 0.7574, + "step": 1020 + }, + { + "epoch": 0.9826756496631376, + "grad_norm": 1.789186612657638, + "learning_rate": 5.1360483361772175e-06, + "loss": 0.5385, + "step": 1021 + }, + { + "epoch": 0.983638113570741, + "grad_norm": 1.088901370828122, + "learning_rate": 5.128491808518791e-06, + "loss": 0.775, + "step": 1022 + }, + { + "epoch": 0.9846005774783445, + "grad_norm": 1.0991884057567713, + "learning_rate": 5.120934987173838e-06, + "loss": 0.7887, + "step": 1023 + }, + { + "epoch": 0.9855630413859481, + "grad_norm": 1.0625600712867114, + "learning_rate": 5.113377889414566e-06, + "loss": 0.7716, + "step": 1024 + }, + { + "epoch": 0.9865255052935515, + "grad_norm": 1.0429852549770542, + "learning_rate": 5.10582053251381e-06, + "loss": 0.7811, + "step": 1025 + }, + { + "epoch": 0.987487969201155, + "grad_norm": 1.1387642438636076, + "learning_rate": 5.098262933744994e-06, + "loss": 0.8433, + "step": 1026 + }, + { + "epoch": 0.9884504331087585, + "grad_norm": 1.0840790256911808, + "learning_rate": 5.0907051103820995e-06, + "loss": 0.7676, + "step": 1027 + }, + { + "epoch": 0.9894128970163619, + "grad_norm": 1.0556016488410995, + "learning_rate": 5.083147079699619e-06, + "loss": 0.7205, + "step": 1028 + }, + { + "epoch": 0.9903753609239654, + "grad_norm": 1.1605479400743541, + "learning_rate": 5.07558885897252e-06, + "loss": 0.8702, + "step": 1029 + }, + { + "epoch": 0.9913378248315688, + "grad_norm": 1.08181642856895, + "learning_rate": 5.068030465476204e-06, + "loss": 0.7656, + "step": 1030 + }, + { + "epoch": 0.9923002887391723, + "grad_norm": 1.0553882767425231, + "learning_rate": 5.060471916486466e-06, + "loss": 0.794, + "step": 1031 + }, + { + "epoch": 0.9932627526467758, + "grad_norm": 1.0873792899553378, + "learning_rate": 5.052913229279459e-06, + "loss": 0.8073, + "step": 1032 + }, + { + "epoch": 0.9942252165543792, + "grad_norm": 1.0464742511864424, + "learning_rate": 5.04535442113165e-06, + "loss": 0.7278, + "step": 1033 + }, + { + "epoch": 0.9951876804619827, + "grad_norm": 1.1100060418414928, + "learning_rate": 5.037795509319784e-06, + "loss": 0.7878, + "step": 1034 + }, + { + "epoch": 0.9961501443695862, + "grad_norm": 0.9707144914904634, + "learning_rate": 5.030236511120841e-06, + "loss": 0.6755, + "step": 1035 + }, + { + "epoch": 0.9971126082771896, + "grad_norm": 1.1463635784327013, + "learning_rate": 5.0226774438119995e-06, + "loss": 0.8969, + "step": 1036 + }, + { + "epoch": 0.9980750721847931, + "grad_norm": 1.1005329795349086, + "learning_rate": 5.015118324670595e-06, + "loss": 0.7706, + "step": 1037 + }, + { + "epoch": 0.9990375360923965, + "grad_norm": 1.0646166585936965, + "learning_rate": 5.0075591709740845e-06, + "loss": 0.7728, + "step": 1038 + }, + { + "epoch": 1.0, + "grad_norm": 1.2062428636833173, + "learning_rate": 5e-06, + "loss": 0.628, + "step": 1039 + }, + { + "epoch": 1.0009624639076036, + "grad_norm": 1.168458828248276, + "learning_rate": 4.9924408290259154e-06, + "loss": 0.4843, + "step": 1040 + }, + { + "epoch": 1.001924927815207, + "grad_norm": 1.2339583333436663, + "learning_rate": 4.984881675329407e-06, + "loss": 0.5947, + "step": 1041 + }, + { + "epoch": 1.0028873917228105, + "grad_norm": 1.2709050533429962, + "learning_rate": 4.977322556188001e-06, + "loss": 0.6429, + "step": 1042 + }, + { + "epoch": 1.0038498556304138, + "grad_norm": 1.1738358201598365, + "learning_rate": 4.969763488879161e-06, + "loss": 0.6262, + "step": 1043 + }, + { + "epoch": 1.0048123195380174, + "grad_norm": 1.3902479136392962, + "learning_rate": 4.962204490680216e-06, + "loss": 0.699, + "step": 1044 + }, + { + "epoch": 1.0057747834456208, + "grad_norm": 1.3076267515780795, + "learning_rate": 4.954645578868352e-06, + "loss": 0.6817, + "step": 1045 + }, + { + "epoch": 1.0067372473532243, + "grad_norm": 1.1912371225780893, + "learning_rate": 4.947086770720542e-06, + "loss": 0.4838, + "step": 1046 + }, + { + "epoch": 1.0076997112608277, + "grad_norm": 1.4466592606323148, + "learning_rate": 4.939528083513536e-06, + "loss": 0.6613, + "step": 1047 + }, + { + "epoch": 1.0086621751684313, + "grad_norm": 1.264304252134362, + "learning_rate": 4.931969534523797e-06, + "loss": 0.5688, + "step": 1048 + }, + { + "epoch": 1.0096246390760346, + "grad_norm": 1.2225369570335536, + "learning_rate": 4.924411141027482e-06, + "loss": 0.5924, + "step": 1049 + }, + { + "epoch": 1.0105871029836382, + "grad_norm": 1.3032593292287808, + "learning_rate": 4.916852920300381e-06, + "loss": 0.6887, + "step": 1050 + }, + { + "epoch": 1.0115495668912415, + "grad_norm": 0.8965168997415174, + "learning_rate": 4.909294889617901e-06, + "loss": 0.4067, + "step": 1051 + }, + { + "epoch": 1.012512030798845, + "grad_norm": 1.0273928807269772, + "learning_rate": 4.9017370662550065e-06, + "loss": 0.4736, + "step": 1052 + }, + { + "epoch": 1.0134744947064485, + "grad_norm": 1.079768747813011, + "learning_rate": 4.894179467486192e-06, + "loss": 0.601, + "step": 1053 + }, + { + "epoch": 1.014436958614052, + "grad_norm": 1.170767405145305, + "learning_rate": 4.8866221105854335e-06, + "loss": 0.6271, + "step": 1054 + }, + { + "epoch": 1.0153994225216554, + "grad_norm": 1.1524827791014203, + "learning_rate": 4.879065012826164e-06, + "loss": 0.6641, + "step": 1055 + }, + { + "epoch": 1.016361886429259, + "grad_norm": 1.1815205416759225, + "learning_rate": 4.871508191481211e-06, + "loss": 0.6679, + "step": 1056 + }, + { + "epoch": 1.0173243503368623, + "grad_norm": 1.0179679224734708, + "learning_rate": 4.863951663822785e-06, + "loss": 0.5682, + "step": 1057 + }, + { + "epoch": 1.0182868142444659, + "grad_norm": 0.9954436425568577, + "learning_rate": 4.856395447122411e-06, + "loss": 0.5519, + "step": 1058 + }, + { + "epoch": 1.0192492781520692, + "grad_norm": 1.093253104717174, + "learning_rate": 4.848839558650917e-06, + "loss": 0.6199, + "step": 1059 + }, + { + "epoch": 1.0202117420596728, + "grad_norm": 1.1974656836171778, + "learning_rate": 4.841284015678366e-06, + "loss": 0.6638, + "step": 1060 + }, + { + "epoch": 1.0211742059672761, + "grad_norm": 1.0536449867674684, + "learning_rate": 4.8337288354740425e-06, + "loss": 0.5112, + "step": 1061 + }, + { + "epoch": 1.0221366698748797, + "grad_norm": 1.0769333362662947, + "learning_rate": 4.8261740353063985e-06, + "loss": 0.5489, + "step": 1062 + }, + { + "epoch": 1.023099133782483, + "grad_norm": 1.2268529634292793, + "learning_rate": 4.818619632443019e-06, + "loss": 0.6686, + "step": 1063 + }, + { + "epoch": 1.0240615976900866, + "grad_norm": 1.1673823120294256, + "learning_rate": 4.811065644150575e-06, + "loss": 0.6849, + "step": 1064 + }, + { + "epoch": 1.02502406159769, + "grad_norm": 1.1093872600378536, + "learning_rate": 4.8035120876947976e-06, + "loss": 0.6163, + "step": 1065 + }, + { + "epoch": 1.0259865255052936, + "grad_norm": 1.1681417741991607, + "learning_rate": 4.795958980340425e-06, + "loss": 0.6695, + "step": 1066 + }, + { + "epoch": 1.026948989412897, + "grad_norm": 1.1135595103880664, + "learning_rate": 4.788406339351174e-06, + "loss": 0.6079, + "step": 1067 + }, + { + "epoch": 1.0279114533205005, + "grad_norm": 1.0996327077262242, + "learning_rate": 4.7808541819896885e-06, + "loss": 0.6056, + "step": 1068 + }, + { + "epoch": 1.0288739172281038, + "grad_norm": 1.090883132025077, + "learning_rate": 4.773302525517515e-06, + "loss": 0.577, + "step": 1069 + }, + { + "epoch": 1.0298363811357074, + "grad_norm": 1.1562707421842207, + "learning_rate": 4.76575138719505e-06, + "loss": 0.6893, + "step": 1070 + }, + { + "epoch": 1.030798845043311, + "grad_norm": 1.0923177693567498, + "learning_rate": 4.758200784281505e-06, + "loss": 0.6111, + "step": 1071 + }, + { + "epoch": 1.0317613089509143, + "grad_norm": 1.144616300886079, + "learning_rate": 4.750650734034875e-06, + "loss": 0.6091, + "step": 1072 + }, + { + "epoch": 1.032723772858518, + "grad_norm": 0.9622095435127656, + "learning_rate": 4.743101253711882e-06, + "loss": 0.4765, + "step": 1073 + }, + { + "epoch": 1.0336862367661213, + "grad_norm": 0.9054540932238566, + "learning_rate": 4.735552360567952e-06, + "loss": 0.4074, + "step": 1074 + }, + { + "epoch": 1.0346487006737248, + "grad_norm": 1.1485507603820715, + "learning_rate": 4.728004071857165e-06, + "loss": 0.5994, + "step": 1075 + }, + { + "epoch": 1.0356111645813282, + "grad_norm": 3.5729427136328518, + "learning_rate": 4.720456404832222e-06, + "loss": 0.4662, + "step": 1076 + }, + { + "epoch": 1.0365736284889318, + "grad_norm": 1.1703731213815847, + "learning_rate": 4.712909376744403e-06, + "loss": 0.6983, + "step": 1077 + }, + { + "epoch": 1.037536092396535, + "grad_norm": 1.1969893883650926, + "learning_rate": 4.7053630048435295e-06, + "loss": 0.6893, + "step": 1078 + }, + { + "epoch": 1.0384985563041387, + "grad_norm": 1.0554646034513666, + "learning_rate": 4.697817306377916e-06, + "loss": 0.6255, + "step": 1079 + }, + { + "epoch": 1.039461020211742, + "grad_norm": 0.9650027118290936, + "learning_rate": 4.690272298594344e-06, + "loss": 0.5307, + "step": 1080 + }, + { + "epoch": 1.0404234841193456, + "grad_norm": 31.003953060160413, + "learning_rate": 4.682727998738016e-06, + "loss": 0.7273, + "step": 1081 + }, + { + "epoch": 1.041385948026949, + "grad_norm": 1.0110640574083698, + "learning_rate": 4.675184424052516e-06, + "loss": 0.5662, + "step": 1082 + }, + { + "epoch": 1.0423484119345525, + "grad_norm": 1.190279226795906, + "learning_rate": 4.6676415917797665e-06, + "loss": 0.6756, + "step": 1083 + }, + { + "epoch": 1.0433108758421559, + "grad_norm": 1.0556614018150878, + "learning_rate": 4.660099519159999e-06, + "loss": 0.5417, + "step": 1084 + }, + { + "epoch": 1.0442733397497594, + "grad_norm": 1.030673619822067, + "learning_rate": 4.652558223431704e-06, + "loss": 0.5406, + "step": 1085 + }, + { + "epoch": 1.0452358036573628, + "grad_norm": 1.057222504716238, + "learning_rate": 4.645017721831602e-06, + "loss": 0.5558, + "step": 1086 + }, + { + "epoch": 1.0461982675649664, + "grad_norm": 1.2105710840727517, + "learning_rate": 4.6374780315945906e-06, + "loss": 0.6521, + "step": 1087 + }, + { + "epoch": 1.0471607314725697, + "grad_norm": 1.203746834364745, + "learning_rate": 4.629939169953721e-06, + "loss": 0.67, + "step": 1088 + }, + { + "epoch": 1.0481231953801733, + "grad_norm": 1.1379891582614203, + "learning_rate": 4.622401154140141e-06, + "loss": 0.5897, + "step": 1089 + }, + { + "epoch": 1.0490856592877766, + "grad_norm": 1.1111046726632525, + "learning_rate": 4.614864001383077e-06, + "loss": 0.6261, + "step": 1090 + }, + { + "epoch": 1.0500481231953802, + "grad_norm": 1.2530041826316447, + "learning_rate": 4.607327728909772e-06, + "loss": 0.6691, + "step": 1091 + }, + { + "epoch": 1.0510105871029836, + "grad_norm": 1.1191474255925957, + "learning_rate": 4.599792353945467e-06, + "loss": 0.6222, + "step": 1092 + }, + { + "epoch": 1.0519730510105871, + "grad_norm": 1.1444218652652156, + "learning_rate": 4.592257893713339e-06, + "loss": 0.6201, + "step": 1093 + }, + { + "epoch": 1.0529355149181905, + "grad_norm": 1.2391657641047866, + "learning_rate": 4.584724365434485e-06, + "loss": 0.6706, + "step": 1094 + }, + { + "epoch": 1.053897978825794, + "grad_norm": 1.144618524032234, + "learning_rate": 4.577191786327868e-06, + "loss": 0.5948, + "step": 1095 + }, + { + "epoch": 1.0548604427333974, + "grad_norm": 1.1476454751130094, + "learning_rate": 4.569660173610284e-06, + "loss": 0.6361, + "step": 1096 + }, + { + "epoch": 1.055822906641001, + "grad_norm": 1.119293864109898, + "learning_rate": 4.5621295444963125e-06, + "loss": 0.6138, + "step": 1097 + }, + { + "epoch": 1.0567853705486043, + "grad_norm": 1.209250880119996, + "learning_rate": 4.554599916198296e-06, + "loss": 0.7127, + "step": 1098 + }, + { + "epoch": 1.057747834456208, + "grad_norm": 1.1507451923833092, + "learning_rate": 4.547071305926279e-06, + "loss": 0.661, + "step": 1099 + }, + { + "epoch": 1.0587102983638113, + "grad_norm": 1.0820970031942014, + "learning_rate": 4.53954373088799e-06, + "loss": 0.5538, + "step": 1100 + }, + { + "epoch": 1.0596727622714148, + "grad_norm": 0.9838245949458192, + "learning_rate": 4.532017208288778e-06, + "loss": 0.4834, + "step": 1101 + }, + { + "epoch": 1.0606352261790182, + "grad_norm": 1.070275565651606, + "learning_rate": 4.524491755331598e-06, + "loss": 0.5609, + "step": 1102 + }, + { + "epoch": 1.0615976900866217, + "grad_norm": 1.0476001459770612, + "learning_rate": 4.516967389216956e-06, + "loss": 0.5579, + "step": 1103 + }, + { + "epoch": 1.0625601539942253, + "grad_norm": 1.1206279606104343, + "learning_rate": 4.509444127142872e-06, + "loss": 0.638, + "step": 1104 + }, + { + "epoch": 1.0635226179018287, + "grad_norm": 1.1385946801868907, + "learning_rate": 4.501921986304847e-06, + "loss": 0.5792, + "step": 1105 + }, + { + "epoch": 1.0644850818094322, + "grad_norm": 1.1367297297767334, + "learning_rate": 4.494400983895813e-06, + "loss": 0.5472, + "step": 1106 + }, + { + "epoch": 1.0654475457170356, + "grad_norm": 1.0955400865066158, + "learning_rate": 4.4868811371061085e-06, + "loss": 0.561, + "step": 1107 + }, + { + "epoch": 1.0664100096246392, + "grad_norm": 1.2095735877534055, + "learning_rate": 4.47936246312342e-06, + "loss": 0.641, + "step": 1108 + }, + { + "epoch": 1.0673724735322425, + "grad_norm": 1.064164585475643, + "learning_rate": 4.471844979132763e-06, + "loss": 0.4867, + "step": 1109 + }, + { + "epoch": 1.068334937439846, + "grad_norm": 1.1792291834400577, + "learning_rate": 4.464328702316427e-06, + "loss": 0.6956, + "step": 1110 + }, + { + "epoch": 1.0692974013474494, + "grad_norm": 1.089548224572839, + "learning_rate": 4.456813649853948e-06, + "loss": 0.5869, + "step": 1111 + }, + { + "epoch": 1.070259865255053, + "grad_norm": 1.0190202048827561, + "learning_rate": 4.449299838922055e-06, + "loss": 0.5268, + "step": 1112 + }, + { + "epoch": 1.0712223291626564, + "grad_norm": 1.10136083211822, + "learning_rate": 4.441787286694649e-06, + "loss": 0.6112, + "step": 1113 + }, + { + "epoch": 1.07218479307026, + "grad_norm": 0.9969499802448273, + "learning_rate": 4.434276010342747e-06, + "loss": 0.4697, + "step": 1114 + }, + { + "epoch": 1.0731472569778633, + "grad_norm": 0.9772466969720769, + "learning_rate": 4.426766027034454e-06, + "loss": 0.5122, + "step": 1115 + }, + { + "epoch": 1.0741097208854669, + "grad_norm": 1.1748123362165872, + "learning_rate": 4.419257353934915e-06, + "loss": 0.6804, + "step": 1116 + }, + { + "epoch": 1.0750721847930702, + "grad_norm": 1.1084597815060642, + "learning_rate": 4.411750008206286e-06, + "loss": 0.5644, + "step": 1117 + }, + { + "epoch": 1.0760346487006738, + "grad_norm": 1.1137466757941175, + "learning_rate": 4.4042440070076845e-06, + "loss": 0.6079, + "step": 1118 + }, + { + "epoch": 1.0769971126082771, + "grad_norm": 1.032284500360651, + "learning_rate": 4.396739367495159e-06, + "loss": 0.5265, + "step": 1119 + }, + { + "epoch": 1.0779595765158807, + "grad_norm": 1.1544134842908762, + "learning_rate": 4.3892361068216395e-06, + "loss": 0.667, + "step": 1120 + }, + { + "epoch": 1.078922040423484, + "grad_norm": 1.233134528382199, + "learning_rate": 4.381734242136913e-06, + "loss": 0.6515, + "step": 1121 + }, + { + "epoch": 1.0798845043310876, + "grad_norm": 1.1198815181474924, + "learning_rate": 4.374233790587565e-06, + "loss": 0.6048, + "step": 1122 + }, + { + "epoch": 1.080846968238691, + "grad_norm": 1.203295006798521, + "learning_rate": 4.366734769316963e-06, + "loss": 0.6835, + "step": 1123 + }, + { + "epoch": 1.0818094321462945, + "grad_norm": 0.9398214062147185, + "learning_rate": 4.359237195465196e-06, + "loss": 0.4611, + "step": 1124 + }, + { + "epoch": 1.082771896053898, + "grad_norm": 1.113977469915678, + "learning_rate": 4.35174108616905e-06, + "loss": 0.5474, + "step": 1125 + }, + { + "epoch": 1.0837343599615015, + "grad_norm": 1.0781766967237163, + "learning_rate": 4.34424645856196e-06, + "loss": 0.5726, + "step": 1126 + }, + { + "epoch": 1.0846968238691048, + "grad_norm": 1.0599657123882598, + "learning_rate": 4.336753329773979e-06, + "loss": 0.5355, + "step": 1127 + }, + { + "epoch": 1.0856592877767084, + "grad_norm": 1.1546586677966086, + "learning_rate": 4.329261716931728e-06, + "loss": 0.6082, + "step": 1128 + }, + { + "epoch": 1.0866217516843117, + "grad_norm": 1.089169940221405, + "learning_rate": 4.32177163715837e-06, + "loss": 0.5802, + "step": 1129 + }, + { + "epoch": 1.0875842155919153, + "grad_norm": 2.38061487133468, + "learning_rate": 4.314283107573556e-06, + "loss": 0.4717, + "step": 1130 + }, + { + "epoch": 1.0885466794995187, + "grad_norm": 1.060125797061331, + "learning_rate": 4.306796145293403e-06, + "loss": 0.5594, + "step": 1131 + }, + { + "epoch": 1.0895091434071222, + "grad_norm": 1.1504541571479996, + "learning_rate": 4.299310767430436e-06, + "loss": 0.6184, + "step": 1132 + }, + { + "epoch": 1.0904716073147256, + "grad_norm": 0.9715144913662728, + "learning_rate": 4.29182699109357e-06, + "loss": 0.4822, + "step": 1133 + }, + { + "epoch": 1.0914340712223292, + "grad_norm": 5.4812554274422745, + "learning_rate": 4.284344833388046e-06, + "loss": 0.5919, + "step": 1134 + }, + { + "epoch": 1.0923965351299327, + "grad_norm": 1.297620910928882, + "learning_rate": 4.276864311415417e-06, + "loss": 0.6443, + "step": 1135 + }, + { + "epoch": 1.093358999037536, + "grad_norm": 1.2742676913992474, + "learning_rate": 4.269385442273492e-06, + "loss": 0.6298, + "step": 1136 + }, + { + "epoch": 1.0943214629451397, + "grad_norm": 0.9883799946807241, + "learning_rate": 4.261908243056301e-06, + "loss": 0.488, + "step": 1137 + }, + { + "epoch": 1.095283926852743, + "grad_norm": 1.181787428643211, + "learning_rate": 4.254432730854063e-06, + "loss": 0.6404, + "step": 1138 + }, + { + "epoch": 1.0962463907603466, + "grad_norm": 1.1476000933834622, + "learning_rate": 4.246958922753132e-06, + "loss": 0.6539, + "step": 1139 + }, + { + "epoch": 1.09720885466795, + "grad_norm": 1.1364477774745074, + "learning_rate": 4.239486835835978e-06, + "loss": 0.6105, + "step": 1140 + }, + { + "epoch": 1.0981713185755535, + "grad_norm": 1.0507202171077419, + "learning_rate": 4.232016487181124e-06, + "loss": 0.562, + "step": 1141 + }, + { + "epoch": 1.0991337824831569, + "grad_norm": 1.1573258546470144, + "learning_rate": 4.224547893863133e-06, + "loss": 0.6095, + "step": 1142 + }, + { + "epoch": 1.1000962463907604, + "grad_norm": 1.1340286844079053, + "learning_rate": 4.217081072952548e-06, + "loss": 0.5994, + "step": 1143 + }, + { + "epoch": 1.1010587102983638, + "grad_norm": 1.1995601227651946, + "learning_rate": 4.209616041515864e-06, + "loss": 0.7019, + "step": 1144 + }, + { + "epoch": 1.1020211742059673, + "grad_norm": 1.2288977052413552, + "learning_rate": 4.202152816615482e-06, + "loss": 0.6442, + "step": 1145 + }, + { + "epoch": 1.1029836381135707, + "grad_norm": 1.2243231882782613, + "learning_rate": 4.194691415309679e-06, + "loss": 0.5771, + "step": 1146 + }, + { + "epoch": 1.1039461020211743, + "grad_norm": 2.071692633043303, + "learning_rate": 4.187231854652562e-06, + "loss": 0.6195, + "step": 1147 + }, + { + "epoch": 1.1049085659287776, + "grad_norm": 1.0448761345918396, + "learning_rate": 4.179774151694031e-06, + "loss": 0.5165, + "step": 1148 + }, + { + "epoch": 1.1058710298363812, + "grad_norm": 1.081014156372236, + "learning_rate": 4.172318323479737e-06, + "loss": 0.566, + "step": 1149 + }, + { + "epoch": 1.1068334937439845, + "grad_norm": 1.1332007006029112, + "learning_rate": 4.16486438705105e-06, + "loss": 0.6217, + "step": 1150 + }, + { + "epoch": 1.1077959576515881, + "grad_norm": 1.0214664487791119, + "learning_rate": 4.157412359445015e-06, + "loss": 0.5545, + "step": 1151 + }, + { + "epoch": 1.1087584215591915, + "grad_norm": 36.0572307651311, + "learning_rate": 4.149962257694315e-06, + "loss": 1.7046, + "step": 1152 + }, + { + "epoch": 1.109720885466795, + "grad_norm": 1.0943534465482867, + "learning_rate": 4.142514098827226e-06, + "loss": 0.6075, + "step": 1153 + }, + { + "epoch": 1.1106833493743984, + "grad_norm": 1.7294229073529148, + "learning_rate": 4.135067899867592e-06, + "loss": 0.5345, + "step": 1154 + }, + { + "epoch": 1.111645813282002, + "grad_norm": 1.1677415747340243, + "learning_rate": 4.127623677834767e-06, + "loss": 0.6275, + "step": 1155 + }, + { + "epoch": 1.1126082771896053, + "grad_norm": 0.9377109600604417, + "learning_rate": 4.120181449743597e-06, + "loss": 0.4669, + "step": 1156 + }, + { + "epoch": 1.1135707410972089, + "grad_norm": 1.3942245496041852, + "learning_rate": 4.112741232604362e-06, + "loss": 0.4904, + "step": 1157 + }, + { + "epoch": 1.1145332050048122, + "grad_norm": 1.1048789202299945, + "learning_rate": 4.105303043422753e-06, + "loss": 0.6071, + "step": 1158 + }, + { + "epoch": 1.1154956689124158, + "grad_norm": 1.0574028893646585, + "learning_rate": 4.097866899199818e-06, + "loss": 0.5329, + "step": 1159 + }, + { + "epoch": 1.1164581328200192, + "grad_norm": 1.0129166049417908, + "learning_rate": 4.090432816931937e-06, + "loss": 0.5858, + "step": 1160 + }, + { + "epoch": 1.1174205967276227, + "grad_norm": 1.0668432055537733, + "learning_rate": 4.083000813610774e-06, + "loss": 0.5947, + "step": 1161 + }, + { + "epoch": 1.118383060635226, + "grad_norm": 0.9122204931779232, + "learning_rate": 4.075570906223245e-06, + "loss": 0.4078, + "step": 1162 + }, + { + "epoch": 1.1193455245428297, + "grad_norm": 1.0236601674086812, + "learning_rate": 4.06814311175147e-06, + "loss": 0.5507, + "step": 1163 + }, + { + "epoch": 1.120307988450433, + "grad_norm": 1.0595893816942472, + "learning_rate": 4.060717447172743e-06, + "loss": 0.5451, + "step": 1164 + }, + { + "epoch": 1.1212704523580366, + "grad_norm": 1.0055874421647237, + "learning_rate": 4.053293929459489e-06, + "loss": 0.4994, + "step": 1165 + }, + { + "epoch": 1.1222329162656401, + "grad_norm": 1.143684318012956, + "learning_rate": 4.045872575579227e-06, + "loss": 0.6087, + "step": 1166 + }, + { + "epoch": 1.1231953801732435, + "grad_norm": 1.2100131480293512, + "learning_rate": 4.038453402494531e-06, + "loss": 0.6076, + "step": 1167 + }, + { + "epoch": 1.124157844080847, + "grad_norm": 1.2114748829049706, + "learning_rate": 4.031036427162986e-06, + "loss": 0.6105, + "step": 1168 + }, + { + "epoch": 1.1251203079884504, + "grad_norm": 1.1687208949998036, + "learning_rate": 4.023621666537158e-06, + "loss": 0.6537, + "step": 1169 + }, + { + "epoch": 1.126082771896054, + "grad_norm": 0.9872191798796656, + "learning_rate": 4.016209137564549e-06, + "loss": 0.5097, + "step": 1170 + }, + { + "epoch": 1.1270452358036573, + "grad_norm": 1.102442897190764, + "learning_rate": 4.0087988571875645e-06, + "loss": 0.5888, + "step": 1171 + }, + { + "epoch": 1.128007699711261, + "grad_norm": 1.1438468386718164, + "learning_rate": 4.0013908423434625e-06, + "loss": 0.5954, + "step": 1172 + }, + { + "epoch": 1.1289701636188643, + "grad_norm": 1.1402116813173888, + "learning_rate": 3.993985109964329e-06, + "loss": 0.4848, + "step": 1173 + }, + { + "epoch": 1.1299326275264678, + "grad_norm": 1.0614432599785364, + "learning_rate": 3.986581676977031e-06, + "loss": 0.5671, + "step": 1174 + }, + { + "epoch": 1.1308950914340712, + "grad_norm": 1.1403305191206234, + "learning_rate": 3.97918056030318e-06, + "loss": 0.6057, + "step": 1175 + }, + { + "epoch": 1.1318575553416748, + "grad_norm": 1.0545329815411733, + "learning_rate": 3.971781776859093e-06, + "loss": 0.5641, + "step": 1176 + }, + { + "epoch": 1.132820019249278, + "grad_norm": 0.9974687828457087, + "learning_rate": 3.9643853435557565e-06, + "loss": 0.501, + "step": 1177 + }, + { + "epoch": 1.1337824831568817, + "grad_norm": 1.1926530333096053, + "learning_rate": 3.9569912772987815e-06, + "loss": 0.6422, + "step": 1178 + }, + { + "epoch": 1.134744947064485, + "grad_norm": 1.066707334338742, + "learning_rate": 3.9495995949883725e-06, + "loss": 0.5602, + "step": 1179 + }, + { + "epoch": 1.1357074109720886, + "grad_norm": 1.0806617821776874, + "learning_rate": 3.942210313519281e-06, + "loss": 0.5479, + "step": 1180 + }, + { + "epoch": 1.136669874879692, + "grad_norm": 1.0422199843046083, + "learning_rate": 3.934823449780777e-06, + "loss": 0.5297, + "step": 1181 + }, + { + "epoch": 1.1376323387872955, + "grad_norm": 1.15601560093133, + "learning_rate": 3.927439020656596e-06, + "loss": 0.6526, + "step": 1182 + }, + { + "epoch": 1.1385948026948989, + "grad_norm": 0.9649080280163498, + "learning_rate": 3.920057043024918e-06, + "loss": 0.4671, + "step": 1183 + }, + { + "epoch": 1.1395572666025024, + "grad_norm": 1.0409965342161733, + "learning_rate": 3.9126775337583126e-06, + "loss": 0.5327, + "step": 1184 + }, + { + "epoch": 1.1405197305101058, + "grad_norm": 1.1453963505777232, + "learning_rate": 3.905300509723713e-06, + "loss": 0.612, + "step": 1185 + }, + { + "epoch": 1.1414821944177094, + "grad_norm": 0.9073857962549913, + "learning_rate": 3.897925987782367e-06, + "loss": 0.4372, + "step": 1186 + }, + { + "epoch": 1.1424446583253127, + "grad_norm": 1.0810309965420104, + "learning_rate": 3.890553984789808e-06, + "loss": 0.5533, + "step": 1187 + }, + { + "epoch": 1.1434071222329163, + "grad_norm": 1.1198696950544198, + "learning_rate": 3.883184517595807e-06, + "loss": 0.5434, + "step": 1188 + }, + { + "epoch": 1.1443695861405196, + "grad_norm": 1.0834659140200769, + "learning_rate": 3.875817603044345e-06, + "loss": 0.5908, + "step": 1189 + }, + { + "epoch": 1.1453320500481232, + "grad_norm": 1.1416169717891405, + "learning_rate": 3.868453257973565e-06, + "loss": 0.6541, + "step": 1190 + }, + { + "epoch": 1.1462945139557266, + "grad_norm": 1.064038317522395, + "learning_rate": 3.8610914992157375e-06, + "loss": 0.5569, + "step": 1191 + }, + { + "epoch": 1.1472569778633301, + "grad_norm": 1.1374376576968386, + "learning_rate": 3.85373234359722e-06, + "loss": 0.4584, + "step": 1192 + }, + { + "epoch": 1.1482194417709337, + "grad_norm": 1.2224510992186857, + "learning_rate": 3.8463758079384246e-06, + "loss": 0.6949, + "step": 1193 + }, + { + "epoch": 1.149181905678537, + "grad_norm": 1.090007772328154, + "learning_rate": 3.83902190905377e-06, + "loss": 0.6293, + "step": 1194 + }, + { + "epoch": 1.1501443695861404, + "grad_norm": 1.1591286396591853, + "learning_rate": 3.831670663751655e-06, + "loss": 0.6807, + "step": 1195 + }, + { + "epoch": 1.151106833493744, + "grad_norm": 1.1112368197996103, + "learning_rate": 3.824322088834402e-06, + "loss": 0.5872, + "step": 1196 + }, + { + "epoch": 1.1520692974013476, + "grad_norm": 1.2102347366596435, + "learning_rate": 3.816976201098241e-06, + "loss": 0.6151, + "step": 1197 + }, + { + "epoch": 1.153031761308951, + "grad_norm": 1.1351132829677333, + "learning_rate": 3.8096330173332568e-06, + "loss": 0.6589, + "step": 1198 + }, + { + "epoch": 1.1539942252165543, + "grad_norm": 0.9875555353091208, + "learning_rate": 3.8022925543233496e-06, + "loss": 0.468, + "step": 1199 + }, + { + "epoch": 1.1549566891241578, + "grad_norm": 1.1897318829503403, + "learning_rate": 3.7949548288462085e-06, + "loss": 0.6033, + "step": 1200 + }, + { + "epoch": 1.1559191530317614, + "grad_norm": 1.0775869902386637, + "learning_rate": 3.7876198576732564e-06, + "loss": 0.5514, + "step": 1201 + }, + { + "epoch": 1.1568816169393648, + "grad_norm": 1.035611865527, + "learning_rate": 3.780287657569629e-06, + "loss": 0.5154, + "step": 1202 + }, + { + "epoch": 1.1578440808469683, + "grad_norm": 1.1523336470648686, + "learning_rate": 3.772958245294123e-06, + "loss": 0.6726, + "step": 1203 + }, + { + "epoch": 1.1588065447545717, + "grad_norm": 1.1616789912321952, + "learning_rate": 3.7656316375991685e-06, + "loss": 0.5327, + "step": 1204 + }, + { + "epoch": 1.1597690086621752, + "grad_norm": 0.8842348814792935, + "learning_rate": 3.758307851230777e-06, + "loss": 0.4059, + "step": 1205 + }, + { + "epoch": 1.1607314725697786, + "grad_norm": 1.1911850956840058, + "learning_rate": 3.7509869029285216e-06, + "loss": 0.6839, + "step": 1206 + }, + { + "epoch": 1.1616939364773822, + "grad_norm": 1.1117156061261073, + "learning_rate": 3.743668809425478e-06, + "loss": 0.5309, + "step": 1207 + }, + { + "epoch": 1.1626564003849855, + "grad_norm": 1.2670327087960906, + "learning_rate": 3.7363535874482073e-06, + "loss": 0.6718, + "step": 1208 + }, + { + "epoch": 1.163618864292589, + "grad_norm": 1.0700133561608403, + "learning_rate": 3.729041253716698e-06, + "loss": 0.5716, + "step": 1209 + }, + { + "epoch": 1.1645813282001924, + "grad_norm": 1.1319967489080087, + "learning_rate": 3.7217318249443457e-06, + "loss": 0.614, + "step": 1210 + }, + { + "epoch": 1.165543792107796, + "grad_norm": 1.2316986013577151, + "learning_rate": 3.714425317837898e-06, + "loss": 0.6447, + "step": 1211 + }, + { + "epoch": 1.1665062560153994, + "grad_norm": 1.0902252515299817, + "learning_rate": 3.707121749097431e-06, + "loss": 0.5491, + "step": 1212 + }, + { + "epoch": 1.167468719923003, + "grad_norm": 0.9866907602864131, + "learning_rate": 3.6998211354163004e-06, + "loss": 0.49, + "step": 1213 + }, + { + "epoch": 1.1684311838306063, + "grad_norm": 1.0493971171362688, + "learning_rate": 3.6925234934811132e-06, + "loss": 0.5424, + "step": 1214 + }, + { + "epoch": 1.1693936477382099, + "grad_norm": 0.9526872323288756, + "learning_rate": 3.6852288399716764e-06, + "loss": 0.4506, + "step": 1215 + }, + { + "epoch": 1.1703561116458132, + "grad_norm": 1.267091946192136, + "learning_rate": 3.6779371915609724e-06, + "loss": 0.6788, + "step": 1216 + }, + { + "epoch": 1.1713185755534168, + "grad_norm": 1.0977080020344492, + "learning_rate": 3.6706485649151124e-06, + "loss": 0.5526, + "step": 1217 + }, + { + "epoch": 1.1722810394610201, + "grad_norm": 1.2020345868585731, + "learning_rate": 3.6633629766933044e-06, + "loss": 0.6441, + "step": 1218 + }, + { + "epoch": 1.1732435033686237, + "grad_norm": 1.0436008873936589, + "learning_rate": 3.656080443547805e-06, + "loss": 0.4982, + "step": 1219 + }, + { + "epoch": 1.174205967276227, + "grad_norm": 1.0720923492604524, + "learning_rate": 3.6488009821238956e-06, + "loss": 0.5768, + "step": 1220 + }, + { + "epoch": 1.1751684311838306, + "grad_norm": 1.1935401490562592, + "learning_rate": 3.6415246090598293e-06, + "loss": 0.6747, + "step": 1221 + }, + { + "epoch": 1.176130895091434, + "grad_norm": 1.082963078516439, + "learning_rate": 3.6342513409868075e-06, + "loss": 0.6072, + "step": 1222 + }, + { + "epoch": 1.1770933589990376, + "grad_norm": 1.1500645830134402, + "learning_rate": 3.6269811945289284e-06, + "loss": 0.6315, + "step": 1223 + }, + { + "epoch": 1.1780558229066411, + "grad_norm": 0.9264205247047045, + "learning_rate": 3.6197141863031622e-06, + "loss": 0.4459, + "step": 1224 + }, + { + "epoch": 1.1790182868142445, + "grad_norm": 1.2040023164843086, + "learning_rate": 3.612450332919297e-06, + "loss": 0.6891, + "step": 1225 + }, + { + "epoch": 1.1799807507218478, + "grad_norm": 1.2195760931603659, + "learning_rate": 3.60518965097992e-06, + "loss": 0.5285, + "step": 1226 + }, + { + "epoch": 1.1809432146294514, + "grad_norm": 1.187483180209594, + "learning_rate": 3.597932157080361e-06, + "loss": 0.6594, + "step": 1227 + }, + { + "epoch": 1.181905678537055, + "grad_norm": 1.2133776756842967, + "learning_rate": 3.590677867808669e-06, + "loss": 0.6789, + "step": 1228 + }, + { + "epoch": 1.1828681424446583, + "grad_norm": 1.074259655691273, + "learning_rate": 3.583426799745568e-06, + "loss": 0.5782, + "step": 1229 + }, + { + "epoch": 1.1838306063522617, + "grad_norm": 1.102764364407735, + "learning_rate": 3.5761789694644138e-06, + "loss": 0.6052, + "step": 1230 + }, + { + "epoch": 1.1847930702598652, + "grad_norm": 0.9522855416235475, + "learning_rate": 3.568934393531169e-06, + "loss": 0.4642, + "step": 1231 + }, + { + "epoch": 1.1857555341674688, + "grad_norm": 1.0925432785321687, + "learning_rate": 3.5616930885043523e-06, + "loss": 0.5898, + "step": 1232 + }, + { + "epoch": 1.1867179980750722, + "grad_norm": 1.1664633325568488, + "learning_rate": 3.5544550709350116e-06, + "loss": 0.5853, + "step": 1233 + }, + { + "epoch": 1.1876804619826757, + "grad_norm": 1.0777334134998835, + "learning_rate": 3.547220357366674e-06, + "loss": 0.5639, + "step": 1234 + }, + { + "epoch": 1.188642925890279, + "grad_norm": 1.0827543074302324, + "learning_rate": 3.5399889643353223e-06, + "loss": 0.5539, + "step": 1235 + }, + { + "epoch": 1.1896053897978827, + "grad_norm": 0.9761916277286855, + "learning_rate": 3.5327609083693436e-06, + "loss": 0.4677, + "step": 1236 + }, + { + "epoch": 1.190567853705486, + "grad_norm": 1.186176278045492, + "learning_rate": 3.5255362059895025e-06, + "loss": 0.6634, + "step": 1237 + }, + { + "epoch": 1.1915303176130896, + "grad_norm": 1.1468089597220275, + "learning_rate": 3.518314873708894e-06, + "loss": 0.6244, + "step": 1238 + }, + { + "epoch": 1.192492781520693, + "grad_norm": 1.1172344181289222, + "learning_rate": 3.511096928032915e-06, + "loss": 0.5802, + "step": 1239 + }, + { + "epoch": 1.1934552454282965, + "grad_norm": 1.0704830798117737, + "learning_rate": 3.503882385459215e-06, + "loss": 0.5271, + "step": 1240 + }, + { + "epoch": 1.1944177093358999, + "grad_norm": 1.1258555223163758, + "learning_rate": 3.4966712624776746e-06, + "loss": 0.6214, + "step": 1241 + }, + { + "epoch": 1.1953801732435034, + "grad_norm": 1.2552739766115177, + "learning_rate": 3.4894635755703487e-06, + "loss": 0.6448, + "step": 1242 + }, + { + "epoch": 1.1963426371511068, + "grad_norm": 1.0521148670154066, + "learning_rate": 3.482259341211448e-06, + "loss": 0.5385, + "step": 1243 + }, + { + "epoch": 1.1973051010587104, + "grad_norm": 1.1174127331255759, + "learning_rate": 3.4750585758672827e-06, + "loss": 0.5869, + "step": 1244 + }, + { + "epoch": 1.1982675649663137, + "grad_norm": 1.2324162492708772, + "learning_rate": 3.467861295996241e-06, + "loss": 0.6297, + "step": 1245 + }, + { + "epoch": 1.1992300288739173, + "grad_norm": 1.2056515263017678, + "learning_rate": 3.46066751804874e-06, + "loss": 0.6275, + "step": 1246 + }, + { + "epoch": 1.2001924927815206, + "grad_norm": 1.1221413753950038, + "learning_rate": 3.4534772584671983e-06, + "loss": 0.5223, + "step": 1247 + }, + { + "epoch": 1.2011549566891242, + "grad_norm": 1.5311315340135143, + "learning_rate": 3.4462905336859842e-06, + "loss": 0.5587, + "step": 1248 + }, + { + "epoch": 1.2021174205967275, + "grad_norm": 1.0302998279686535, + "learning_rate": 3.4391073601313962e-06, + "loss": 0.4964, + "step": 1249 + }, + { + "epoch": 1.2030798845043311, + "grad_norm": 1.2058623856946182, + "learning_rate": 3.4319277542216078e-06, + "loss": 0.6163, + "step": 1250 + }, + { + "epoch": 1.2040423484119345, + "grad_norm": 1.2152908096367265, + "learning_rate": 3.424751732366646e-06, + "loss": 0.6698, + "step": 1251 + }, + { + "epoch": 1.205004812319538, + "grad_norm": 1.1048079039603718, + "learning_rate": 3.417579310968336e-06, + "loss": 0.5752, + "step": 1252 + }, + { + "epoch": 1.2059672762271414, + "grad_norm": 1.0962043519276012, + "learning_rate": 3.410410506420285e-06, + "loss": 0.6041, + "step": 1253 + }, + { + "epoch": 1.206929740134745, + "grad_norm": 1.0346302216414853, + "learning_rate": 3.403245335107822e-06, + "loss": 0.4738, + "step": 1254 + }, + { + "epoch": 1.2078922040423483, + "grad_norm": 1.1001481000243336, + "learning_rate": 3.396083813407982e-06, + "loss": 0.6096, + "step": 1255 + }, + { + "epoch": 1.2088546679499519, + "grad_norm": 1.271099992489326, + "learning_rate": 3.3889259576894517e-06, + "loss": 0.599, + "step": 1256 + }, + { + "epoch": 1.2098171318575552, + "grad_norm": 1.0061514155151503, + "learning_rate": 3.381771784312542e-06, + "loss": 0.4768, + "step": 1257 + }, + { + "epoch": 1.2107795957651588, + "grad_norm": 1.119398945291699, + "learning_rate": 3.3746213096291434e-06, + "loss": 0.5913, + "step": 1258 + }, + { + "epoch": 1.2117420596727624, + "grad_norm": 0.9842010438083552, + "learning_rate": 3.3674745499826966e-06, + "loss": 0.4749, + "step": 1259 + }, + { + "epoch": 1.2127045235803657, + "grad_norm": 1.0626913053188742, + "learning_rate": 3.3603315217081493e-06, + "loss": 0.4883, + "step": 1260 + }, + { + "epoch": 1.213666987487969, + "grad_norm": 1.1050388062206837, + "learning_rate": 3.3531922411319208e-06, + "loss": 0.5881, + "step": 1261 + }, + { + "epoch": 1.2146294513955727, + "grad_norm": 1.1124758188825798, + "learning_rate": 3.3460567245718656e-06, + "loss": 0.5999, + "step": 1262 + }, + { + "epoch": 1.2155919153031762, + "grad_norm": 1.7592599035068168, + "learning_rate": 3.3389249883372317e-06, + "loss": 0.5999, + "step": 1263 + }, + { + "epoch": 1.2165543792107796, + "grad_norm": 1.1316758631865946, + "learning_rate": 3.3317970487286303e-06, + "loss": 0.5393, + "step": 1264 + }, + { + "epoch": 1.2175168431183832, + "grad_norm": 1.1331893044275092, + "learning_rate": 3.3246729220379923e-06, + "loss": 0.5854, + "step": 1265 + }, + { + "epoch": 1.2184793070259865, + "grad_norm": 1.237947157899388, + "learning_rate": 3.3175526245485367e-06, + "loss": 0.6714, + "step": 1266 + }, + { + "epoch": 1.21944177093359, + "grad_norm": 1.1006813575728933, + "learning_rate": 3.3104361725347256e-06, + "loss": 0.5964, + "step": 1267 + }, + { + "epoch": 1.2204042348411934, + "grad_norm": 1.11360161507452, + "learning_rate": 3.3033235822622367e-06, + "loss": 0.5955, + "step": 1268 + }, + { + "epoch": 1.221366698748797, + "grad_norm": 1.1294638886197232, + "learning_rate": 3.2962148699879165e-06, + "loss": 0.6305, + "step": 1269 + }, + { + "epoch": 1.2223291626564003, + "grad_norm": 1.1287072122984838, + "learning_rate": 3.289110051959754e-06, + "loss": 0.6058, + "step": 1270 + }, + { + "epoch": 1.223291626564004, + "grad_norm": 1.1192204464765654, + "learning_rate": 3.2820091444168285e-06, + "loss": 0.5853, + "step": 1271 + }, + { + "epoch": 1.2242540904716073, + "grad_norm": 1.1161906783113278, + "learning_rate": 3.274912163589291e-06, + "loss": 0.632, + "step": 1272 + }, + { + "epoch": 1.2252165543792108, + "grad_norm": 1.2001048513935615, + "learning_rate": 3.2678191256983085e-06, + "loss": 0.6209, + "step": 1273 + }, + { + "epoch": 1.2261790182868142, + "grad_norm": 0.9425529602818286, + "learning_rate": 3.260730046956042e-06, + "loss": 0.4676, + "step": 1274 + }, + { + "epoch": 1.2271414821944178, + "grad_norm": 1.0285608323892141, + "learning_rate": 3.2536449435656013e-06, + "loss": 0.5324, + "step": 1275 + }, + { + "epoch": 1.2281039461020211, + "grad_norm": 1.0416979599137386, + "learning_rate": 3.2465638317210124e-06, + "loss": 0.5201, + "step": 1276 + }, + { + "epoch": 1.2290664100096247, + "grad_norm": 1.1234173913352024, + "learning_rate": 3.2394867276071705e-06, + "loss": 0.6123, + "step": 1277 + }, + { + "epoch": 1.230028873917228, + "grad_norm": 1.1538676000140917, + "learning_rate": 3.23241364739982e-06, + "loss": 0.6432, + "step": 1278 + }, + { + "epoch": 1.2309913378248316, + "grad_norm": 0.9551425908067205, + "learning_rate": 3.2253446072655016e-06, + "loss": 0.4714, + "step": 1279 + }, + { + "epoch": 1.231953801732435, + "grad_norm": 1.0991869056022112, + "learning_rate": 3.2182796233615264e-06, + "loss": 0.5617, + "step": 1280 + }, + { + "epoch": 1.2329162656400385, + "grad_norm": 1.0283863672529134, + "learning_rate": 3.2112187118359284e-06, + "loss": 0.5301, + "step": 1281 + }, + { + "epoch": 1.2338787295476419, + "grad_norm": 1.0647703188871558, + "learning_rate": 3.2041618888274406e-06, + "loss": 0.5561, + "step": 1282 + }, + { + "epoch": 1.2348411934552455, + "grad_norm": 1.1693506350182667, + "learning_rate": 3.1971091704654457e-06, + "loss": 0.6502, + "step": 1283 + }, + { + "epoch": 1.2358036573628488, + "grad_norm": 1.1319602727938594, + "learning_rate": 3.1900605728699484e-06, + "loss": 0.6076, + "step": 1284 + }, + { + "epoch": 1.2367661212704524, + "grad_norm": 1.2351480610629155, + "learning_rate": 3.183016112151531e-06, + "loss": 0.656, + "step": 1285 + }, + { + "epoch": 1.2377285851780557, + "grad_norm": 1.1947376082833203, + "learning_rate": 3.175975804411325e-06, + "loss": 0.6756, + "step": 1286 + }, + { + "epoch": 1.2386910490856593, + "grad_norm": 1.1835681766011275, + "learning_rate": 3.1689396657409637e-06, + "loss": 0.6646, + "step": 1287 + }, + { + "epoch": 1.2396535129932627, + "grad_norm": 1.0153332853656383, + "learning_rate": 3.1619077122225585e-06, + "loss": 0.5041, + "step": 1288 + }, + { + "epoch": 1.2406159769008662, + "grad_norm": 1.2056332096429585, + "learning_rate": 3.1548799599286474e-06, + "loss": 0.7011, + "step": 1289 + }, + { + "epoch": 1.2415784408084698, + "grad_norm": 1.2078532173134453, + "learning_rate": 3.1478564249221744e-06, + "loss": 0.6663, + "step": 1290 + }, + { + "epoch": 1.2425409047160731, + "grad_norm": 1.1056142582019668, + "learning_rate": 3.140837123256435e-06, + "loss": 0.5817, + "step": 1291 + }, + { + "epoch": 1.2435033686236765, + "grad_norm": 1.030004816297868, + "learning_rate": 3.1338220709750567e-06, + "loss": 0.5124, + "step": 1292 + }, + { + "epoch": 1.24446583253128, + "grad_norm": 1.1249293479623885, + "learning_rate": 3.1268112841119504e-06, + "loss": 0.6409, + "step": 1293 + }, + { + "epoch": 1.2454282964388836, + "grad_norm": 1.1307698968520499, + "learning_rate": 3.119804778691278e-06, + "loss": 0.6012, + "step": 1294 + }, + { + "epoch": 1.246390760346487, + "grad_norm": 1.0820880399122097, + "learning_rate": 3.112802570727419e-06, + "loss": 0.5765, + "step": 1295 + }, + { + "epoch": 1.2473532242540906, + "grad_norm": 1.1124345399148399, + "learning_rate": 3.105804676224923e-06, + "loss": 0.5955, + "step": 1296 + }, + { + "epoch": 1.248315688161694, + "grad_norm": 1.1477655903029118, + "learning_rate": 3.098811111178488e-06, + "loss": 0.5925, + "step": 1297 + }, + { + "epoch": 1.2492781520692975, + "grad_norm": 1.163891780589192, + "learning_rate": 3.091821891572912e-06, + "loss": 0.5641, + "step": 1298 + }, + { + "epoch": 1.2502406159769008, + "grad_norm": 1.1410300569217475, + "learning_rate": 3.0848370333830656e-06, + "loss": 0.6056, + "step": 1299 + }, + { + "epoch": 1.2512030798845044, + "grad_norm": 1.0981769470155525, + "learning_rate": 3.0778565525738434e-06, + "loss": 0.5868, + "step": 1300 + }, + { + "epoch": 1.2521655437921078, + "grad_norm": 1.1403573025376474, + "learning_rate": 3.0708804651001425e-06, + "loss": 0.6277, + "step": 1301 + }, + { + "epoch": 1.2531280076997113, + "grad_norm": 1.0217702379705367, + "learning_rate": 3.0639087869068118e-06, + "loss": 0.5302, + "step": 1302 + }, + { + "epoch": 1.2540904716073147, + "grad_norm": 1.0893005986989621, + "learning_rate": 3.0569415339286294e-06, + "loss": 0.5452, + "step": 1303 + }, + { + "epoch": 1.2550529355149183, + "grad_norm": 0.9876796568825769, + "learning_rate": 3.0499787220902506e-06, + "loss": 0.4457, + "step": 1304 + }, + { + "epoch": 1.2560153994225216, + "grad_norm": 1.0116563902674036, + "learning_rate": 3.0430203673061874e-06, + "loss": 0.5282, + "step": 1305 + }, + { + "epoch": 1.2569778633301252, + "grad_norm": 0.992996069239815, + "learning_rate": 3.036066485480758e-06, + "loss": 0.4786, + "step": 1306 + }, + { + "epoch": 1.2579403272377285, + "grad_norm": 1.0514326051414788, + "learning_rate": 3.029117092508063e-06, + "loss": 0.5543, + "step": 1307 + }, + { + "epoch": 1.258902791145332, + "grad_norm": 1.020061763470482, + "learning_rate": 3.0221722042719383e-06, + "loss": 0.4678, + "step": 1308 + }, + { + "epoch": 1.2598652550529355, + "grad_norm": 1.0752120350890806, + "learning_rate": 3.015231836645929e-06, + "loss": 0.5436, + "step": 1309 + }, + { + "epoch": 1.260827718960539, + "grad_norm": 1.1917991195324014, + "learning_rate": 3.008296005493239e-06, + "loss": 0.6474, + "step": 1310 + }, + { + "epoch": 1.2617901828681424, + "grad_norm": 1.043217433631883, + "learning_rate": 3.0013647266667135e-06, + "loss": 0.5284, + "step": 1311 + }, + { + "epoch": 1.262752646775746, + "grad_norm": 1.202856012938046, + "learning_rate": 2.994438016008785e-06, + "loss": 0.6723, + "step": 1312 + }, + { + "epoch": 1.2637151106833493, + "grad_norm": 1.1569136406674396, + "learning_rate": 2.98751588935145e-06, + "loss": 0.6207, + "step": 1313 + }, + { + "epoch": 1.2646775745909529, + "grad_norm": 1.0915547341624632, + "learning_rate": 2.9805983625162226e-06, + "loss": 0.5589, + "step": 1314 + }, + { + "epoch": 1.2656400384985562, + "grad_norm": 1.1626508586248792, + "learning_rate": 2.973685451314109e-06, + "loss": 0.6297, + "step": 1315 + }, + { + "epoch": 1.2666025024061598, + "grad_norm": 1.126874942737971, + "learning_rate": 2.966777171545561e-06, + "loss": 0.5777, + "step": 1316 + }, + { + "epoch": 1.2675649663137634, + "grad_norm": 1.2254487767896072, + "learning_rate": 2.9598735390004473e-06, + "loss": 0.6277, + "step": 1317 + }, + { + "epoch": 1.2685274302213667, + "grad_norm": 1.1551335613077314, + "learning_rate": 2.9529745694580126e-06, + "loss": 0.5647, + "step": 1318 + }, + { + "epoch": 1.26948989412897, + "grad_norm": 0.9871416368589551, + "learning_rate": 2.9460802786868466e-06, + "loss": 0.4989, + "step": 1319 + }, + { + "epoch": 1.2704523580365736, + "grad_norm": 1.0641589906808493, + "learning_rate": 2.93919068244484e-06, + "loss": 0.5704, + "step": 1320 + }, + { + "epoch": 1.2714148219441772, + "grad_norm": 1.049993392283906, + "learning_rate": 2.9323057964791603e-06, + "loss": 0.5314, + "step": 1321 + }, + { + "epoch": 1.2723772858517806, + "grad_norm": 0.8356149272703709, + "learning_rate": 2.925425636526203e-06, + "loss": 0.3774, + "step": 1322 + }, + { + "epoch": 1.273339749759384, + "grad_norm": 0.9632346398638886, + "learning_rate": 2.9185502183115645e-06, + "loss": 0.5164, + "step": 1323 + }, + { + "epoch": 1.2743022136669875, + "grad_norm": 1.2364871189869189, + "learning_rate": 2.9116795575500055e-06, + "loss": 0.6898, + "step": 1324 + }, + { + "epoch": 1.275264677574591, + "grad_norm": 0.9952819853674829, + "learning_rate": 2.9048136699454053e-06, + "loss": 0.544, + "step": 1325 + }, + { + "epoch": 1.2762271414821944, + "grad_norm": 1.1140929401700717, + "learning_rate": 2.897952571190743e-06, + "loss": 0.6253, + "step": 1326 + }, + { + "epoch": 1.2771896053897978, + "grad_norm": 1.1678226808053198, + "learning_rate": 2.8910962769680432e-06, + "loss": 0.6759, + "step": 1327 + }, + { + "epoch": 1.2781520692974013, + "grad_norm": 1.1375855482070154, + "learning_rate": 2.884244802948355e-06, + "loss": 0.6195, + "step": 1328 + }, + { + "epoch": 1.279114533205005, + "grad_norm": 1.1055157430973048, + "learning_rate": 2.8773981647917093e-06, + "loss": 0.6251, + "step": 1329 + }, + { + "epoch": 1.2800769971126083, + "grad_norm": 1.0903678076765642, + "learning_rate": 2.8705563781470835e-06, + "loss": 0.5555, + "step": 1330 + }, + { + "epoch": 1.2810394610202118, + "grad_norm": 1.1278622917207997, + "learning_rate": 2.863719458652363e-06, + "loss": 0.6392, + "step": 1331 + }, + { + "epoch": 1.2820019249278152, + "grad_norm": 1.2076955673436631, + "learning_rate": 2.8568874219343156e-06, + "loss": 0.6723, + "step": 1332 + }, + { + "epoch": 1.2829643888354187, + "grad_norm": 0.9797723961478554, + "learning_rate": 2.8500602836085395e-06, + "loss": 0.4512, + "step": 1333 + }, + { + "epoch": 1.283926852743022, + "grad_norm": 1.164190818467124, + "learning_rate": 2.8432380592794466e-06, + "loss": 0.631, + "step": 1334 + }, + { + "epoch": 1.2848893166506257, + "grad_norm": 1.277589966842865, + "learning_rate": 2.836420764540209e-06, + "loss": 0.6729, + "step": 1335 + }, + { + "epoch": 1.285851780558229, + "grad_norm": 1.095272979724643, + "learning_rate": 2.8296084149727388e-06, + "loss": 0.5621, + "step": 1336 + }, + { + "epoch": 1.2868142444658326, + "grad_norm": 1.0996727627751088, + "learning_rate": 2.822801026147637e-06, + "loss": 0.5886, + "step": 1337 + }, + { + "epoch": 1.287776708373436, + "grad_norm": 1.0374087187125662, + "learning_rate": 2.8159986136241737e-06, + "loss": 0.5131, + "step": 1338 + }, + { + "epoch": 1.2887391722810395, + "grad_norm": 0.9968834682780793, + "learning_rate": 2.8092011929502403e-06, + "loss": 0.4554, + "step": 1339 + }, + { + "epoch": 1.2897016361886429, + "grad_norm": 1.1773682437090403, + "learning_rate": 2.802408779662324e-06, + "loss": 0.6706, + "step": 1340 + }, + { + "epoch": 1.2906641000962464, + "grad_norm": 1.168666640577597, + "learning_rate": 2.7956213892854577e-06, + "loss": 0.6253, + "step": 1341 + }, + { + "epoch": 1.2916265640038498, + "grad_norm": 0.9238312510621753, + "learning_rate": 2.788839037333203e-06, + "loss": 0.4479, + "step": 1342 + }, + { + "epoch": 1.2925890279114534, + "grad_norm": 1.0171937246510896, + "learning_rate": 2.782061739307599e-06, + "loss": 0.5087, + "step": 1343 + }, + { + "epoch": 1.2935514918190567, + "grad_norm": 1.0029443094733823, + "learning_rate": 2.7752895106991382e-06, + "loss": 0.5119, + "step": 1344 + }, + { + "epoch": 1.2945139557266603, + "grad_norm": 1.2463546310220817, + "learning_rate": 2.7685223669867204e-06, + "loss": 0.6856, + "step": 1345 + }, + { + "epoch": 1.2954764196342636, + "grad_norm": 1.068581137727694, + "learning_rate": 2.7617603236376288e-06, + "loss": 0.5148, + "step": 1346 + }, + { + "epoch": 1.2964388835418672, + "grad_norm": 1.1128898605844777, + "learning_rate": 2.7550033961074876e-06, + "loss": 0.5735, + "step": 1347 + }, + { + "epoch": 1.2974013474494708, + "grad_norm": 1.1843573933864946, + "learning_rate": 2.7482515998402278e-06, + "loss": 0.613, + "step": 1348 + }, + { + "epoch": 1.2983638113570741, + "grad_norm": 1.0021224220938105, + "learning_rate": 2.741504950268048e-06, + "loss": 0.4809, + "step": 1349 + }, + { + "epoch": 1.2993262752646775, + "grad_norm": 1.1153463662154428, + "learning_rate": 2.7347634628113918e-06, + "loss": 0.5547, + "step": 1350 + }, + { + "epoch": 1.300288739172281, + "grad_norm": 1.0133610327123355, + "learning_rate": 2.728027152878894e-06, + "loss": 0.508, + "step": 1351 + }, + { + "epoch": 1.3012512030798846, + "grad_norm": 1.1155203361538966, + "learning_rate": 2.7212960358673645e-06, + "loss": 0.6009, + "step": 1352 + }, + { + "epoch": 1.302213666987488, + "grad_norm": 1.0196062571491373, + "learning_rate": 2.7145701271617375e-06, + "loss": 0.5331, + "step": 1353 + }, + { + "epoch": 1.3031761308950913, + "grad_norm": 1.063496153812271, + "learning_rate": 2.7078494421350456e-06, + "loss": 0.5604, + "step": 1354 + }, + { + "epoch": 1.304138594802695, + "grad_norm": 1.2557403173481294, + "learning_rate": 2.7011339961483857e-06, + "loss": 0.6859, + "step": 1355 + }, + { + "epoch": 1.3051010587102985, + "grad_norm": 0.9715748138805197, + "learning_rate": 2.6944238045508708e-06, + "loss": 0.451, + "step": 1356 + }, + { + "epoch": 1.3060635226179018, + "grad_norm": 1.150676810999731, + "learning_rate": 2.6877188826796124e-06, + "loss": 0.5831, + "step": 1357 + }, + { + "epoch": 1.3070259865255052, + "grad_norm": 1.18705896419668, + "learning_rate": 2.681019245859674e-06, + "loss": 0.6766, + "step": 1358 + }, + { + "epoch": 1.3079884504331087, + "grad_norm": 1.1461214944464375, + "learning_rate": 2.6743249094040423e-06, + "loss": 0.6385, + "step": 1359 + }, + { + "epoch": 1.3089509143407123, + "grad_norm": 1.193110900383247, + "learning_rate": 2.6676358886135833e-06, + "loss": 0.6509, + "step": 1360 + }, + { + "epoch": 1.3099133782483157, + "grad_norm": 1.1429072954424568, + "learning_rate": 2.6609521987770194e-06, + "loss": 0.6615, + "step": 1361 + }, + { + "epoch": 1.3108758421559192, + "grad_norm": 1.0162181259066074, + "learning_rate": 2.654273855170883e-06, + "loss": 0.5241, + "step": 1362 + }, + { + "epoch": 1.3118383060635226, + "grad_norm": 1.190721431591362, + "learning_rate": 2.647600873059493e-06, + "loss": 0.6602, + "step": 1363 + }, + { + "epoch": 1.3128007699711262, + "grad_norm": 1.1163037118055004, + "learning_rate": 2.640933267694907e-06, + "loss": 0.5768, + "step": 1364 + }, + { + "epoch": 1.3137632338787295, + "grad_norm": 1.167424143392283, + "learning_rate": 2.6342710543168985e-06, + "loss": 0.5976, + "step": 1365 + }, + { + "epoch": 1.314725697786333, + "grad_norm": 1.1749583286215404, + "learning_rate": 2.6276142481529153e-06, + "loss": 0.5881, + "step": 1366 + }, + { + "epoch": 1.3156881616939364, + "grad_norm": 1.2770035260668564, + "learning_rate": 2.6209628644180484e-06, + "loss": 0.6658, + "step": 1367 + }, + { + "epoch": 1.31665062560154, + "grad_norm": 1.1794521821635355, + "learning_rate": 2.6143169183149873e-06, + "loss": 0.6045, + "step": 1368 + }, + { + "epoch": 1.3176130895091434, + "grad_norm": 1.1179461633952203, + "learning_rate": 2.607676425034004e-06, + "loss": 0.55, + "step": 1369 + }, + { + "epoch": 1.318575553416747, + "grad_norm": 1.0747133400010174, + "learning_rate": 2.601041399752897e-06, + "loss": 0.5279, + "step": 1370 + }, + { + "epoch": 1.3195380173243503, + "grad_norm": 1.0340124021367345, + "learning_rate": 2.5944118576369758e-06, + "loss": 0.5455, + "step": 1371 + }, + { + "epoch": 1.3205004812319538, + "grad_norm": 1.0653982229639185, + "learning_rate": 2.5877878138390096e-06, + "loss": 0.5407, + "step": 1372 + }, + { + "epoch": 1.3214629451395572, + "grad_norm": 1.2440859193411689, + "learning_rate": 2.581169283499207e-06, + "loss": 0.7018, + "step": 1373 + }, + { + "epoch": 1.3224254090471608, + "grad_norm": 1.2622503190587873, + "learning_rate": 2.5745562817451688e-06, + "loss": 0.7058, + "step": 1374 + }, + { + "epoch": 1.3233878729547641, + "grad_norm": 1.1776361415436842, + "learning_rate": 2.5679488236918626e-06, + "loss": 0.6201, + "step": 1375 + }, + { + "epoch": 1.3243503368623677, + "grad_norm": 1.110337438407713, + "learning_rate": 2.5613469244415866e-06, + "loss": 0.6444, + "step": 1376 + }, + { + "epoch": 1.325312800769971, + "grad_norm": 0.9736986478635364, + "learning_rate": 2.5547505990839335e-06, + "loss": 0.4616, + "step": 1377 + }, + { + "epoch": 1.3262752646775746, + "grad_norm": 1.0959329010383354, + "learning_rate": 2.54815986269575e-06, + "loss": 0.5326, + "step": 1378 + }, + { + "epoch": 1.3272377285851782, + "grad_norm": 1.0350890641966373, + "learning_rate": 2.5415747303411167e-06, + "loss": 0.5162, + "step": 1379 + }, + { + "epoch": 1.3282001924927815, + "grad_norm": 0.9992021556365457, + "learning_rate": 2.5349952170712975e-06, + "loss": 0.5131, + "step": 1380 + }, + { + "epoch": 1.329162656400385, + "grad_norm": 1.0330778641550553, + "learning_rate": 2.528421337924721e-06, + "loss": 0.5158, + "step": 1381 + }, + { + "epoch": 1.3301251203079885, + "grad_norm": 1.0556635471927618, + "learning_rate": 2.5218531079269316e-06, + "loss": 0.5484, + "step": 1382 + }, + { + "epoch": 1.331087584215592, + "grad_norm": 1.0742290056544697, + "learning_rate": 2.5152905420905636e-06, + "loss": 0.5197, + "step": 1383 + }, + { + "epoch": 1.3320500481231954, + "grad_norm": 1.257305055722386, + "learning_rate": 2.5087336554153087e-06, + "loss": 0.665, + "step": 1384 + }, + { + "epoch": 1.3330125120307987, + "grad_norm": 0.9619162998006614, + "learning_rate": 2.5021824628878757e-06, + "loss": 0.4705, + "step": 1385 + }, + { + "epoch": 1.3339749759384023, + "grad_norm": 1.1828275111875532, + "learning_rate": 2.4956369794819536e-06, + "loss": 0.6738, + "step": 1386 + }, + { + "epoch": 1.3349374398460059, + "grad_norm": 1.1529861905201622, + "learning_rate": 2.48909722015819e-06, + "loss": 0.6437, + "step": 1387 + }, + { + "epoch": 1.3358999037536092, + "grad_norm": 1.1553722848200345, + "learning_rate": 2.4825631998641474e-06, + "loss": 0.59, + "step": 1388 + }, + { + "epoch": 1.3368623676612126, + "grad_norm": 1.1554637372840686, + "learning_rate": 2.4760349335342665e-06, + "loss": 0.6006, + "step": 1389 + }, + { + "epoch": 1.3378248315688162, + "grad_norm": 1.0620056595651133, + "learning_rate": 2.4695124360898427e-06, + "loss": 0.532, + "step": 1390 + }, + { + "epoch": 1.3387872954764197, + "grad_norm": 1.8927207519838543, + "learning_rate": 2.4629957224389794e-06, + "loss": 0.5485, + "step": 1391 + }, + { + "epoch": 1.339749759384023, + "grad_norm": 1.1480391787595678, + "learning_rate": 2.456484807476567e-06, + "loss": 0.6007, + "step": 1392 + }, + { + "epoch": 1.3407122232916264, + "grad_norm": 1.003179408806779, + "learning_rate": 2.4499797060842363e-06, + "loss": 0.526, + "step": 1393 + }, + { + "epoch": 1.34167468719923, + "grad_norm": 1.0767873809110469, + "learning_rate": 2.443480433130333e-06, + "loss": 0.5496, + "step": 1394 + }, + { + "epoch": 1.3426371511068336, + "grad_norm": 1.2115863702852099, + "learning_rate": 2.4369870034698823e-06, + "loss": 0.6383, + "step": 1395 + }, + { + "epoch": 1.343599615014437, + "grad_norm": 0.8847998341859202, + "learning_rate": 2.4304994319445536e-06, + "loss": 0.3888, + "step": 1396 + }, + { + "epoch": 1.3445620789220405, + "grad_norm": 3.5585428626433266, + "learning_rate": 2.4240177333826216e-06, + "loss": 0.5428, + "step": 1397 + }, + { + "epoch": 1.3455245428296438, + "grad_norm": 1.2080937341635347, + "learning_rate": 2.417541922598945e-06, + "loss": 0.5991, + "step": 1398 + }, + { + "epoch": 1.3464870067372474, + "grad_norm": 1.113848598229492, + "learning_rate": 2.4110720143949185e-06, + "loss": 0.5687, + "step": 1399 + }, + { + "epoch": 1.3474494706448508, + "grad_norm": 1.307894185796789, + "learning_rate": 2.404608023558451e-06, + "loss": 0.7197, + "step": 1400 + }, + { + "epoch": 1.3484119345524543, + "grad_norm": 1.0628842297173193, + "learning_rate": 2.3981499648639223e-06, + "loss": 0.5286, + "step": 1401 + }, + { + "epoch": 1.3493743984600577, + "grad_norm": 1.0507924178214165, + "learning_rate": 2.3916978530721567e-06, + "loss": 0.534, + "step": 1402 + }, + { + "epoch": 1.3503368623676613, + "grad_norm": 0.9045114289879654, + "learning_rate": 2.3852517029303816e-06, + "loss": 0.4172, + "step": 1403 + }, + { + "epoch": 1.3512993262752646, + "grad_norm": 1.1300088211815316, + "learning_rate": 2.378811529172203e-06, + "loss": 0.589, + "step": 1404 + }, + { + "epoch": 1.3522617901828682, + "grad_norm": 1.1129973908170323, + "learning_rate": 2.372377346517565e-06, + "loss": 0.6331, + "step": 1405 + }, + { + "epoch": 1.3532242540904715, + "grad_norm": 0.9846628690304731, + "learning_rate": 2.3659491696727195e-06, + "loss": 0.4884, + "step": 1406 + }, + { + "epoch": 1.354186717998075, + "grad_norm": 1.1697652273940031, + "learning_rate": 2.3595270133301866e-06, + "loss": 0.6962, + "step": 1407 + }, + { + "epoch": 1.3551491819056785, + "grad_norm": 1.077889160373736, + "learning_rate": 2.353110892168733e-06, + "loss": 0.589, + "step": 1408 + }, + { + "epoch": 1.356111645813282, + "grad_norm": 1.1023002055561548, + "learning_rate": 2.3467008208533234e-06, + "loss": 0.5523, + "step": 1409 + }, + { + "epoch": 1.3570741097208856, + "grad_norm": 1.2347323615025085, + "learning_rate": 2.3402968140351014e-06, + "loss": 0.6675, + "step": 1410 + }, + { + "epoch": 1.358036573628489, + "grad_norm": 1.1459039134710123, + "learning_rate": 2.3338988863513435e-06, + "loss": 0.6063, + "step": 1411 + }, + { + "epoch": 1.3589990375360923, + "grad_norm": 1.0256720650041222, + "learning_rate": 2.3275070524254358e-06, + "loss": 0.5427, + "step": 1412 + }, + { + "epoch": 1.3599615014436959, + "grad_norm": 1.1437248764081187, + "learning_rate": 2.321121326866834e-06, + "loss": 0.6582, + "step": 1413 + }, + { + "epoch": 1.3609239653512994, + "grad_norm": 1.1361322616640444, + "learning_rate": 2.3147417242710368e-06, + "loss": 0.6582, + "step": 1414 + }, + { + "epoch": 1.3618864292589028, + "grad_norm": 1.110107937915143, + "learning_rate": 2.3083682592195395e-06, + "loss": 0.602, + "step": 1415 + }, + { + "epoch": 1.3628488931665061, + "grad_norm": 1.1787923309385548, + "learning_rate": 2.3020009462798165e-06, + "loss": 0.6304, + "step": 1416 + }, + { + "epoch": 1.3638113570741097, + "grad_norm": 1.0248691212203134, + "learning_rate": 2.295639800005276e-06, + "loss": 0.5233, + "step": 1417 + }, + { + "epoch": 1.3647738209817133, + "grad_norm": 1.1434899305105686, + "learning_rate": 2.289284834935234e-06, + "loss": 0.6762, + "step": 1418 + }, + { + "epoch": 1.3657362848893166, + "grad_norm": 1.1679474447713436, + "learning_rate": 2.282936065594881e-06, + "loss": 0.6534, + "step": 1419 + }, + { + "epoch": 1.36669874879692, + "grad_norm": 1.061339571309613, + "learning_rate": 2.276593506495239e-06, + "loss": 0.5649, + "step": 1420 + }, + { + "epoch": 1.3676612127045236, + "grad_norm": 1.0719490438369284, + "learning_rate": 2.2702571721331423e-06, + "loss": 0.5498, + "step": 1421 + }, + { + "epoch": 1.3686236766121271, + "grad_norm": 1.0288934515348493, + "learning_rate": 2.263927076991193e-06, + "loss": 0.5197, + "step": 1422 + }, + { + "epoch": 1.3695861405197305, + "grad_norm": 1.211013085474167, + "learning_rate": 2.2576032355377357e-06, + "loss": 0.6394, + "step": 1423 + }, + { + "epoch": 1.3705486044273338, + "grad_norm": 1.1903678292268463, + "learning_rate": 2.2512856622268204e-06, + "loss": 0.6214, + "step": 1424 + }, + { + "epoch": 1.3715110683349374, + "grad_norm": 1.1798781677785575, + "learning_rate": 2.2449743714981727e-06, + "loss": 0.6192, + "step": 1425 + }, + { + "epoch": 1.372473532242541, + "grad_norm": 1.1561376126321308, + "learning_rate": 2.238669377777151e-06, + "loss": 0.6565, + "step": 1426 + }, + { + "epoch": 1.3734359961501443, + "grad_norm": 1.0764308020833737, + "learning_rate": 2.232370695474731e-06, + "loss": 0.584, + "step": 1427 + }, + { + "epoch": 1.374398460057748, + "grad_norm": 1.182151694895717, + "learning_rate": 2.226078338987453e-06, + "loss": 0.6827, + "step": 1428 + }, + { + "epoch": 1.3753609239653513, + "grad_norm": 1.1626556115035063, + "learning_rate": 2.219792322697408e-06, + "loss": 0.6739, + "step": 1429 + }, + { + "epoch": 1.3763233878729548, + "grad_norm": 1.0620692532217344, + "learning_rate": 2.213512660972187e-06, + "loss": 0.5513, + "step": 1430 + }, + { + "epoch": 1.3772858517805582, + "grad_norm": 0.9788630867033864, + "learning_rate": 2.2072393681648626e-06, + "loss": 0.4774, + "step": 1431 + }, + { + "epoch": 1.3782483156881618, + "grad_norm": 0.9631380586701618, + "learning_rate": 2.2009724586139486e-06, + "loss": 0.4462, + "step": 1432 + }, + { + "epoch": 1.379210779595765, + "grad_norm": 1.1719965303270405, + "learning_rate": 2.19471194664337e-06, + "loss": 0.6682, + "step": 1433 + }, + { + "epoch": 1.3801732435033687, + "grad_norm": 1.120816838393378, + "learning_rate": 2.1884578465624257e-06, + "loss": 0.6476, + "step": 1434 + }, + { + "epoch": 1.381135707410972, + "grad_norm": 1.2190110890000962, + "learning_rate": 2.182210172665764e-06, + "loss": 0.6856, + "step": 1435 + }, + { + "epoch": 1.3820981713185756, + "grad_norm": 1.2030434039554476, + "learning_rate": 2.17596893923334e-06, + "loss": 0.6674, + "step": 1436 + }, + { + "epoch": 1.383060635226179, + "grad_norm": 1.1557344819370943, + "learning_rate": 2.1697341605303947e-06, + "loss": 0.6193, + "step": 1437 + }, + { + "epoch": 1.3840230991337825, + "grad_norm": 1.188776377357197, + "learning_rate": 2.1635058508074074e-06, + "loss": 0.6395, + "step": 1438 + }, + { + "epoch": 1.3849855630413859, + "grad_norm": 1.1951717910991513, + "learning_rate": 2.157284024300082e-06, + "loss": 0.6573, + "step": 1439 + }, + { + "epoch": 1.3859480269489894, + "grad_norm": 1.1482965789601272, + "learning_rate": 2.1510686952292936e-06, + "loss": 0.6609, + "step": 1440 + }, + { + "epoch": 1.386910490856593, + "grad_norm": 1.022540148620223, + "learning_rate": 2.144859877801074e-06, + "loss": 0.4921, + "step": 1441 + }, + { + "epoch": 1.3878729547641964, + "grad_norm": 1.2528968838920975, + "learning_rate": 2.1386575862065677e-06, + "loss": 0.6629, + "step": 1442 + }, + { + "epoch": 1.3888354186717997, + "grad_norm": 0.9751452995079534, + "learning_rate": 2.1324618346220082e-06, + "loss": 0.4721, + "step": 1443 + }, + { + "epoch": 1.3897978825794033, + "grad_norm": 1.1560737188733896, + "learning_rate": 2.126272637208674e-06, + "loss": 0.6286, + "step": 1444 + }, + { + "epoch": 1.3907603464870069, + "grad_norm": 1.1419080577875769, + "learning_rate": 2.1200900081128693e-06, + "loss": 0.6261, + "step": 1445 + }, + { + "epoch": 1.3917228103946102, + "grad_norm": 1.135942398648416, + "learning_rate": 2.11391396146588e-06, + "loss": 0.5978, + "step": 1446 + }, + { + "epoch": 1.3926852743022136, + "grad_norm": 1.0921799940639119, + "learning_rate": 2.1077445113839524e-06, + "loss": 0.543, + "step": 1447 + }, + { + "epoch": 1.3936477382098171, + "grad_norm": 1.2147542023431532, + "learning_rate": 2.1015816719682497e-06, + "loss": 0.693, + "step": 1448 + }, + { + "epoch": 1.3946102021174207, + "grad_norm": 1.1327518192791948, + "learning_rate": 2.0954254573048304e-06, + "loss": 0.6793, + "step": 1449 + }, + { + "epoch": 1.395572666025024, + "grad_norm": 1.0809905165535008, + "learning_rate": 2.089275881464609e-06, + "loss": 0.5806, + "step": 1450 + }, + { + "epoch": 1.3965351299326274, + "grad_norm": 1.152497610983864, + "learning_rate": 2.083132958503326e-06, + "loss": 0.6112, + "step": 1451 + }, + { + "epoch": 1.397497593840231, + "grad_norm": 0.9444128854791063, + "learning_rate": 2.0769967024615185e-06, + "loss": 0.4616, + "step": 1452 + }, + { + "epoch": 1.3984600577478346, + "grad_norm": 1.1938834102972384, + "learning_rate": 2.0708671273644785e-06, + "loss": 0.6956, + "step": 1453 + }, + { + "epoch": 1.399422521655438, + "grad_norm": 1.098023118424151, + "learning_rate": 2.064744247222236e-06, + "loss": 0.5925, + "step": 1454 + }, + { + "epoch": 1.4003849855630413, + "grad_norm": 0.9656724683894724, + "learning_rate": 2.0586280760295118e-06, + "loss": 0.4812, + "step": 1455 + }, + { + "epoch": 1.4013474494706448, + "grad_norm": 1.2381649928649805, + "learning_rate": 2.052518627765699e-06, + "loss": 0.7033, + "step": 1456 + }, + { + "epoch": 1.4023099133782484, + "grad_norm": 1.130028296182197, + "learning_rate": 2.0464159163948166e-06, + "loss": 0.591, + "step": 1457 + }, + { + "epoch": 1.4032723772858517, + "grad_norm": 1.0829172528559452, + "learning_rate": 2.0403199558654947e-06, + "loss": 0.5447, + "step": 1458 + }, + { + "epoch": 1.4042348411934553, + "grad_norm": 1.217048906965068, + "learning_rate": 2.0342307601109233e-06, + "loss": 0.612, + "step": 1459 + }, + { + "epoch": 1.4051973051010587, + "grad_norm": 1.197083998404027, + "learning_rate": 2.0281483430488393e-06, + "loss": 0.6856, + "step": 1460 + }, + { + "epoch": 1.4061597690086622, + "grad_norm": 1.0814089867790417, + "learning_rate": 2.0220727185814814e-06, + "loss": 0.5685, + "step": 1461 + }, + { + "epoch": 1.4071222329162656, + "grad_norm": 1.0557369053312775, + "learning_rate": 2.016003900595566e-06, + "loss": 0.5395, + "step": 1462 + }, + { + "epoch": 1.4080846968238692, + "grad_norm": 1.153567418947843, + "learning_rate": 2.0099419029622457e-06, + "loss": 0.6448, + "step": 1463 + }, + { + "epoch": 1.4090471607314725, + "grad_norm": 0.7473234236753786, + "learning_rate": 2.0038867395370937e-06, + "loss": 0.2859, + "step": 1464 + }, + { + "epoch": 1.410009624639076, + "grad_norm": 1.2414156680824924, + "learning_rate": 1.997838424160053e-06, + "loss": 0.654, + "step": 1465 + }, + { + "epoch": 1.4109720885466794, + "grad_norm": 1.1290131606069278, + "learning_rate": 1.9917969706554226e-06, + "loss": 0.6223, + "step": 1466 + }, + { + "epoch": 1.411934552454283, + "grad_norm": 1.0884437172395258, + "learning_rate": 1.985762392831811e-06, + "loss": 0.5868, + "step": 1467 + }, + { + "epoch": 1.4128970163618864, + "grad_norm": 1.0672077827013413, + "learning_rate": 1.9797347044821176e-06, + "loss": 0.5911, + "step": 1468 + }, + { + "epoch": 1.41385948026949, + "grad_norm": 1.0369078790594448, + "learning_rate": 1.9737139193834874e-06, + "loss": 0.5236, + "step": 1469 + }, + { + "epoch": 1.4148219441770933, + "grad_norm": 1.0063875878225668, + "learning_rate": 1.967700051297295e-06, + "loss": 0.499, + "step": 1470 + }, + { + "epoch": 1.4157844080846969, + "grad_norm": 1.0945759929178935, + "learning_rate": 1.9616931139691002e-06, + "loss": 0.5849, + "step": 1471 + }, + { + "epoch": 1.4167468719923004, + "grad_norm": 1.022084970971573, + "learning_rate": 1.9556931211286255e-06, + "loss": 0.5572, + "step": 1472 + }, + { + "epoch": 1.4177093358999038, + "grad_norm": 1.00280378384014, + "learning_rate": 1.9497000864897136e-06, + "loss": 0.5126, + "step": 1473 + }, + { + "epoch": 1.4186717998075071, + "grad_norm": 1.1185998490127498, + "learning_rate": 1.9437140237503122e-06, + "loss": 0.6244, + "step": 1474 + }, + { + "epoch": 1.4196342637151107, + "grad_norm": 0.9111750496448102, + "learning_rate": 1.937734946592425e-06, + "loss": 0.4349, + "step": 1475 + }, + { + "epoch": 1.4205967276227143, + "grad_norm": 1.0334129417378861, + "learning_rate": 1.931762868682098e-06, + "loss": 0.4873, + "step": 1476 + }, + { + "epoch": 1.4215591915303176, + "grad_norm": 1.0939173926394468, + "learning_rate": 1.9257978036693704e-06, + "loss": 0.546, + "step": 1477 + }, + { + "epoch": 1.422521655437921, + "grad_norm": 0.9874006797814725, + "learning_rate": 1.9198397651882594e-06, + "loss": 0.4911, + "step": 1478 + }, + { + "epoch": 1.4234841193455245, + "grad_norm": 1.1404238506952826, + "learning_rate": 1.913888766856719e-06, + "loss": 0.5907, + "step": 1479 + }, + { + "epoch": 1.4244465832531281, + "grad_norm": 1.1599154275748549, + "learning_rate": 1.9079448222766127e-06, + "loss": 0.6833, + "step": 1480 + }, + { + "epoch": 1.4254090471607315, + "grad_norm": 1.211735208074726, + "learning_rate": 1.9020079450336832e-06, + "loss": 0.6291, + "step": 1481 + }, + { + "epoch": 1.4263715110683348, + "grad_norm": 1.1649973839782695, + "learning_rate": 1.8960781486975143e-06, + "loss": 0.6231, + "step": 1482 + }, + { + "epoch": 1.4273339749759384, + "grad_norm": 0.9641620493315649, + "learning_rate": 1.8901554468215128e-06, + "loss": 0.4802, + "step": 1483 + }, + { + "epoch": 1.428296438883542, + "grad_norm": 1.000075783641636, + "learning_rate": 1.8842398529428625e-06, + "loss": 0.5109, + "step": 1484 + }, + { + "epoch": 1.4292589027911453, + "grad_norm": 1.1454401271434729, + "learning_rate": 1.8783313805825081e-06, + "loss": 0.6446, + "step": 1485 + }, + { + "epoch": 1.4302213666987487, + "grad_norm": 1.1589908369232695, + "learning_rate": 1.8724300432451109e-06, + "loss": 0.6272, + "step": 1486 + }, + { + "epoch": 1.4311838306063522, + "grad_norm": 1.1408394080219761, + "learning_rate": 1.8665358544190298e-06, + "loss": 0.6341, + "step": 1487 + }, + { + "epoch": 1.4321462945139558, + "grad_norm": 1.0506823822487772, + "learning_rate": 1.860648827576278e-06, + "loss": 0.552, + "step": 1488 + }, + { + "epoch": 1.4331087584215592, + "grad_norm": 0.8972102939056295, + "learning_rate": 1.8547689761725052e-06, + "loss": 0.4147, + "step": 1489 + }, + { + "epoch": 1.4340712223291627, + "grad_norm": 1.1662340018444437, + "learning_rate": 1.8488963136469573e-06, + "loss": 0.6474, + "step": 1490 + }, + { + "epoch": 1.435033686236766, + "grad_norm": 1.0050995463994379, + "learning_rate": 1.8430308534224522e-06, + "loss": 0.4994, + "step": 1491 + }, + { + "epoch": 1.4359961501443697, + "grad_norm": 1.1737440913571355, + "learning_rate": 1.8371726089053399e-06, + "loss": 0.6648, + "step": 1492 + }, + { + "epoch": 1.436958614051973, + "grad_norm": 1.1076326266013927, + "learning_rate": 1.8313215934854844e-06, + "loss": 0.6135, + "step": 1493 + }, + { + "epoch": 1.4379210779595766, + "grad_norm": 1.0610426021689916, + "learning_rate": 1.8254778205362205e-06, + "loss": 0.5657, + "step": 1494 + }, + { + "epoch": 1.43888354186718, + "grad_norm": 1.1075458454384213, + "learning_rate": 1.8196413034143367e-06, + "loss": 0.629, + "step": 1495 + }, + { + "epoch": 1.4398460057747835, + "grad_norm": 1.1755938428036634, + "learning_rate": 1.813812055460028e-06, + "loss": 0.6519, + "step": 1496 + }, + { + "epoch": 1.4408084696823868, + "grad_norm": 1.0553021522071793, + "learning_rate": 1.8079900899968826e-06, + "loss": 0.5901, + "step": 1497 + }, + { + "epoch": 1.4417709335899904, + "grad_norm": 1.0726966031961442, + "learning_rate": 1.8021754203318403e-06, + "loss": 0.6021, + "step": 1498 + }, + { + "epoch": 1.4427333974975938, + "grad_norm": 1.093144676878953, + "learning_rate": 1.7963680597551669e-06, + "loss": 0.5916, + "step": 1499 + }, + { + "epoch": 1.4436958614051973, + "grad_norm": 1.284120058640894, + "learning_rate": 1.7905680215404176e-06, + "loss": 0.6674, + "step": 1500 + }, + { + "epoch": 1.4436958614051973, + "eval_loss": 0.7866150140762329, + "eval_runtime": 12.5301, + "eval_samples_per_second": 3.352, + "eval_steps_per_second": 0.878, + "step": 1500 + }, + { + "epoch": 1.4446583253128007, + "grad_norm": 1.1374995565054642, + "learning_rate": 1.784775318944418e-06, + "loss": 0.6102, + "step": 1501 + }, + { + "epoch": 1.4456207892204043, + "grad_norm": 1.049487103494092, + "learning_rate": 1.7789899652072206e-06, + "loss": 0.5487, + "step": 1502 + }, + { + "epoch": 1.4465832531280076, + "grad_norm": 1.077458993488053, + "learning_rate": 1.773211973552087e-06, + "loss": 0.573, + "step": 1503 + }, + { + "epoch": 1.4475457170356112, + "grad_norm": 1.1303757699116854, + "learning_rate": 1.767441357185446e-06, + "loss": 0.6307, + "step": 1504 + }, + { + "epoch": 1.4485081809432145, + "grad_norm": 1.1511950947945258, + "learning_rate": 1.7616781292968749e-06, + "loss": 0.6255, + "step": 1505 + }, + { + "epoch": 1.4494706448508181, + "grad_norm": 1.0734888948971792, + "learning_rate": 1.7559223030590578e-06, + "loss": 0.574, + "step": 1506 + }, + { + "epoch": 1.4504331087584217, + "grad_norm": 1.1121349298330556, + "learning_rate": 1.7501738916277643e-06, + "loss": 0.5964, + "step": 1507 + }, + { + "epoch": 1.451395572666025, + "grad_norm": 1.1715939819747685, + "learning_rate": 1.7444329081418176e-06, + "loss": 0.6376, + "step": 1508 + }, + { + "epoch": 1.4523580365736284, + "grad_norm": 1.1347134824407419, + "learning_rate": 1.7386993657230632e-06, + "loss": 0.6435, + "step": 1509 + }, + { + "epoch": 1.453320500481232, + "grad_norm": 1.1407489924997054, + "learning_rate": 1.7329732774763342e-06, + "loss": 0.637, + "step": 1510 + }, + { + "epoch": 1.4542829643888355, + "grad_norm": 1.0059138083031913, + "learning_rate": 1.7272546564894322e-06, + "loss": 0.4752, + "step": 1511 + }, + { + "epoch": 1.4552454282964389, + "grad_norm": 1.13982440525219, + "learning_rate": 1.7215435158330856e-06, + "loss": 0.633, + "step": 1512 + }, + { + "epoch": 1.4562078922040422, + "grad_norm": 0.8567908942730987, + "learning_rate": 1.7158398685609301e-06, + "loss": 0.3796, + "step": 1513 + }, + { + "epoch": 1.4571703561116458, + "grad_norm": 1.0992098894135132, + "learning_rate": 1.7101437277094746e-06, + "loss": 0.5819, + "step": 1514 + }, + { + "epoch": 1.4581328200192494, + "grad_norm": 17.741217969662898, + "learning_rate": 1.7044551062980651e-06, + "loss": 0.4576, + "step": 1515 + }, + { + "epoch": 1.4590952839268527, + "grad_norm": 0.9874649206256819, + "learning_rate": 1.698774017328867e-06, + "loss": 0.4409, + "step": 1516 + }, + { + "epoch": 1.460057747834456, + "grad_norm": 0.8874368050520065, + "learning_rate": 1.6931004737868257e-06, + "loss": 0.4017, + "step": 1517 + }, + { + "epoch": 1.4610202117420596, + "grad_norm": 1.1585020079823702, + "learning_rate": 1.6874344886396438e-06, + "loss": 0.59, + "step": 1518 + }, + { + "epoch": 1.4619826756496632, + "grad_norm": 1.110053476674745, + "learning_rate": 1.6817760748377433e-06, + "loss": 0.5817, + "step": 1519 + }, + { + "epoch": 1.4629451395572666, + "grad_norm": 1.2523573814432587, + "learning_rate": 1.6761252453142458e-06, + "loss": 0.5938, + "step": 1520 + }, + { + "epoch": 1.4639076034648701, + "grad_norm": 0.9738314364340155, + "learning_rate": 1.6704820129849326e-06, + "loss": 0.474, + "step": 1521 + }, + { + "epoch": 1.4648700673724735, + "grad_norm": 0.9906534049647755, + "learning_rate": 1.6648463907482264e-06, + "loss": 0.4901, + "step": 1522 + }, + { + "epoch": 1.465832531280077, + "grad_norm": 1.0669950311062673, + "learning_rate": 1.6592183914851507e-06, + "loss": 0.5395, + "step": 1523 + }, + { + "epoch": 1.4667949951876804, + "grad_norm": 1.0642063395753278, + "learning_rate": 1.65359802805931e-06, + "loss": 0.5017, + "step": 1524 + }, + { + "epoch": 1.467757459095284, + "grad_norm": 1.1649518885892851, + "learning_rate": 1.647985313316851e-06, + "loss": 0.6912, + "step": 1525 + }, + { + "epoch": 1.4687199230028873, + "grad_norm": 1.0608582486384517, + "learning_rate": 1.6423802600864436e-06, + "loss": 0.5308, + "step": 1526 + }, + { + "epoch": 1.469682386910491, + "grad_norm": 1.0563797701143594, + "learning_rate": 1.6367828811792425e-06, + "loss": 0.58, + "step": 1527 + }, + { + "epoch": 1.4706448508180943, + "grad_norm": 0.9235532227588705, + "learning_rate": 1.631193189388865e-06, + "loss": 0.4608, + "step": 1528 + }, + { + "epoch": 1.4716073147256978, + "grad_norm": 1.1551028268535126, + "learning_rate": 1.6256111974913534e-06, + "loss": 0.5775, + "step": 1529 + }, + { + "epoch": 1.4725697786333012, + "grad_norm": 1.1247304921149237, + "learning_rate": 1.6200369182451564e-06, + "loss": 0.6154, + "step": 1530 + }, + { + "epoch": 1.4735322425409048, + "grad_norm": 1.0825293946476329, + "learning_rate": 1.6144703643910887e-06, + "loss": 0.5401, + "step": 1531 + }, + { + "epoch": 1.474494706448508, + "grad_norm": 1.0909740934494851, + "learning_rate": 1.6089115486523133e-06, + "loss": 0.5578, + "step": 1532 + }, + { + "epoch": 1.4754571703561117, + "grad_norm": 1.1875706579565797, + "learning_rate": 1.6033604837343004e-06, + "loss": 0.6273, + "step": 1533 + }, + { + "epoch": 1.476419634263715, + "grad_norm": 1.0434265359391421, + "learning_rate": 1.5978171823248111e-06, + "loss": 0.5379, + "step": 1534 + }, + { + "epoch": 1.4773820981713186, + "grad_norm": 1.0457550689799389, + "learning_rate": 1.592281657093856e-06, + "loss": 0.483, + "step": 1535 + }, + { + "epoch": 1.478344562078922, + "grad_norm": 1.1015178554723923, + "learning_rate": 1.586753920693676e-06, + "loss": 0.5594, + "step": 1536 + }, + { + "epoch": 1.4793070259865255, + "grad_norm": 1.1583484202897973, + "learning_rate": 1.5812339857587083e-06, + "loss": 0.6441, + "step": 1537 + }, + { + "epoch": 1.480269489894129, + "grad_norm": 1.1336849378085136, + "learning_rate": 1.5757218649055605e-06, + "loss": 0.6198, + "step": 1538 + }, + { + "epoch": 1.4812319538017324, + "grad_norm": 0.9598019506460577, + "learning_rate": 1.5702175707329747e-06, + "loss": 0.4837, + "step": 1539 + }, + { + "epoch": 1.4821944177093358, + "grad_norm": 1.0954051340395627, + "learning_rate": 1.5647211158218107e-06, + "loss": 0.6022, + "step": 1540 + }, + { + "epoch": 1.4831568816169394, + "grad_norm": 1.2077111276577153, + "learning_rate": 1.5592325127350044e-06, + "loss": 0.6871, + "step": 1541 + }, + { + "epoch": 1.484119345524543, + "grad_norm": 1.1431775535172768, + "learning_rate": 1.5537517740175512e-06, + "loss": 0.599, + "step": 1542 + }, + { + "epoch": 1.4850818094321463, + "grad_norm": 1.0256076920509156, + "learning_rate": 1.5482789121964653e-06, + "loss": 0.5492, + "step": 1543 + }, + { + "epoch": 1.4860442733397496, + "grad_norm": 1.0995306872515014, + "learning_rate": 1.5428139397807619e-06, + "loss": 0.6229, + "step": 1544 + }, + { + "epoch": 1.4870067372473532, + "grad_norm": 1.0430904131537608, + "learning_rate": 1.5373568692614215e-06, + "loss": 0.5457, + "step": 1545 + }, + { + "epoch": 1.4879692011549568, + "grad_norm": 1.1846052709987545, + "learning_rate": 1.5319077131113652e-06, + "loss": 0.647, + "step": 1546 + }, + { + "epoch": 1.4889316650625601, + "grad_norm": 1.2681740534286612, + "learning_rate": 1.5264664837854242e-06, + "loss": 0.6602, + "step": 1547 + }, + { + "epoch": 1.4898941289701635, + "grad_norm": 1.1331295286312706, + "learning_rate": 1.5210331937203087e-06, + "loss": 0.6064, + "step": 1548 + }, + { + "epoch": 1.490856592877767, + "grad_norm": 1.0628158974336184, + "learning_rate": 1.5156078553345888e-06, + "loss": 0.5482, + "step": 1549 + }, + { + "epoch": 1.4918190567853706, + "grad_norm": 1.1810664744533284, + "learning_rate": 1.5101904810286522e-06, + "loss": 0.6602, + "step": 1550 + }, + { + "epoch": 1.492781520692974, + "grad_norm": 1.0496614805138587, + "learning_rate": 1.5047810831846916e-06, + "loss": 0.5632, + "step": 1551 + }, + { + "epoch": 1.4937439846005776, + "grad_norm": 1.5194277601054333, + "learning_rate": 1.499379674166661e-06, + "loss": 0.6357, + "step": 1552 + }, + { + "epoch": 1.494706448508181, + "grad_norm": 1.0974952282210066, + "learning_rate": 1.493986266320261e-06, + "loss": 0.5947, + "step": 1553 + }, + { + "epoch": 1.4956689124157845, + "grad_norm": 0.9834876459839467, + "learning_rate": 1.488600871972899e-06, + "loss": 0.4681, + "step": 1554 + }, + { + "epoch": 1.4966313763233878, + "grad_norm": 1.0859261622598826, + "learning_rate": 1.4832235034336695e-06, + "loss": 0.5447, + "step": 1555 + }, + { + "epoch": 1.4975938402309914, + "grad_norm": 1.153936384743228, + "learning_rate": 1.4778541729933239e-06, + "loss": 0.6396, + "step": 1556 + }, + { + "epoch": 1.4985563041385948, + "grad_norm": 1.1294815241081775, + "learning_rate": 1.4724928929242393e-06, + "loss": 0.6155, + "step": 1557 + }, + { + "epoch": 1.4995187680461983, + "grad_norm": 1.245218636941583, + "learning_rate": 1.467139675480392e-06, + "loss": 0.6246, + "step": 1558 + }, + { + "epoch": 1.5004812319538017, + "grad_norm": 1.110868751907078, + "learning_rate": 1.4617945328973327e-06, + "loss": 0.62, + "step": 1559 + }, + { + "epoch": 1.5014436958614052, + "grad_norm": 0.9360809020374281, + "learning_rate": 1.4564574773921514e-06, + "loss": 0.4326, + "step": 1560 + }, + { + "epoch": 1.5024061597690088, + "grad_norm": 1.15985498594906, + "learning_rate": 1.4511285211634602e-06, + "loss": 0.6078, + "step": 1561 + }, + { + "epoch": 1.5033686236766122, + "grad_norm": 1.1491048973123879, + "learning_rate": 1.445807676391352e-06, + "loss": 0.6135, + "step": 1562 + }, + { + "epoch": 1.5043310875842155, + "grad_norm": 1.1599526923681776, + "learning_rate": 1.4404949552373848e-06, + "loss": 0.6829, + "step": 1563 + }, + { + "epoch": 1.505293551491819, + "grad_norm": 1.1240352730086625, + "learning_rate": 1.435190369844548e-06, + "loss": 0.6023, + "step": 1564 + }, + { + "epoch": 1.5062560153994227, + "grad_norm": 1.1285631227092439, + "learning_rate": 1.4298939323372363e-06, + "loss": 0.6242, + "step": 1565 + }, + { + "epoch": 1.507218479307026, + "grad_norm": 1.0950136497744267, + "learning_rate": 1.4246056548212173e-06, + "loss": 0.5929, + "step": 1566 + }, + { + "epoch": 1.5081809432146294, + "grad_norm": 1.178452247731739, + "learning_rate": 1.4193255493836127e-06, + "loss": 0.6104, + "step": 1567 + }, + { + "epoch": 1.509143407122233, + "grad_norm": 1.1396648333171127, + "learning_rate": 1.4140536280928614e-06, + "loss": 0.5873, + "step": 1568 + }, + { + "epoch": 1.5101058710298365, + "grad_norm": 1.0587356669419736, + "learning_rate": 1.4087899029987007e-06, + "loss": 0.5751, + "step": 1569 + }, + { + "epoch": 1.5110683349374399, + "grad_norm": 1.137912574642229, + "learning_rate": 1.4035343861321293e-06, + "loss": 0.5957, + "step": 1570 + }, + { + "epoch": 1.5120307988450432, + "grad_norm": 1.1731981583755504, + "learning_rate": 1.3982870895053901e-06, + "loss": 0.6695, + "step": 1571 + }, + { + "epoch": 1.5129932627526468, + "grad_norm": 1.1647952828917711, + "learning_rate": 1.393048025111932e-06, + "loss": 0.6581, + "step": 1572 + }, + { + "epoch": 1.5139557266602504, + "grad_norm": 1.0869357423243637, + "learning_rate": 1.3878172049263928e-06, + "loss": 0.577, + "step": 1573 + }, + { + "epoch": 1.5149181905678537, + "grad_norm": 1.065676597310303, + "learning_rate": 1.3825946409045642e-06, + "loss": 0.5712, + "step": 1574 + }, + { + "epoch": 1.515880654475457, + "grad_norm": 1.202063142256274, + "learning_rate": 1.3773803449833684e-06, + "loss": 0.6808, + "step": 1575 + }, + { + "epoch": 1.5168431183830606, + "grad_norm": 1.0358038331268669, + "learning_rate": 1.3721743290808304e-06, + "loss": 0.5189, + "step": 1576 + }, + { + "epoch": 1.5178055822906642, + "grad_norm": 1.1044142865694802, + "learning_rate": 1.3669766050960465e-06, + "loss": 0.6236, + "step": 1577 + }, + { + "epoch": 1.5187680461982676, + "grad_norm": 1.1222750647614421, + "learning_rate": 1.3617871849091657e-06, + "loss": 0.589, + "step": 1578 + }, + { + "epoch": 1.519730510105871, + "grad_norm": 1.1084273892861216, + "learning_rate": 1.3566060803813513e-06, + "loss": 0.6366, + "step": 1579 + }, + { + "epoch": 1.5206929740134745, + "grad_norm": 1.0246320718785609, + "learning_rate": 1.3514333033547671e-06, + "loss": 0.5393, + "step": 1580 + }, + { + "epoch": 1.521655437921078, + "grad_norm": 1.0809937884894891, + "learning_rate": 1.3462688656525364e-06, + "loss": 0.5754, + "step": 1581 + }, + { + "epoch": 1.5226179018286814, + "grad_norm": 0.9410923880306508, + "learning_rate": 1.3411127790787265e-06, + "loss": 0.4424, + "step": 1582 + }, + { + "epoch": 1.5235803657362847, + "grad_norm": 1.1483453726795392, + "learning_rate": 1.3359650554183157e-06, + "loss": 0.5928, + "step": 1583 + }, + { + "epoch": 1.5245428296438883, + "grad_norm": 1.1187051493934994, + "learning_rate": 1.3308257064371678e-06, + "loss": 0.5978, + "step": 1584 + }, + { + "epoch": 1.525505293551492, + "grad_norm": 1.188286882793742, + "learning_rate": 1.3256947438820029e-06, + "loss": 0.6217, + "step": 1585 + }, + { + "epoch": 1.5264677574590952, + "grad_norm": 1.1470760400385733, + "learning_rate": 1.3205721794803766e-06, + "loss": 0.6238, + "step": 1586 + }, + { + "epoch": 1.5274302213666986, + "grad_norm": 1.0289262772822738, + "learning_rate": 1.3154580249406446e-06, + "loss": 0.5327, + "step": 1587 + }, + { + "epoch": 1.5283926852743022, + "grad_norm": 1.1087397735478353, + "learning_rate": 1.3103522919519463e-06, + "loss": 0.599, + "step": 1588 + }, + { + "epoch": 1.5293551491819057, + "grad_norm": 1.0092690613422182, + "learning_rate": 1.3052549921841662e-06, + "loss": 0.529, + "step": 1589 + }, + { + "epoch": 1.530317613089509, + "grad_norm": 1.1219251632833631, + "learning_rate": 1.3001661372879194e-06, + "loss": 0.6466, + "step": 1590 + }, + { + "epoch": 1.5312800769971127, + "grad_norm": 0.9646607928381605, + "learning_rate": 1.2950857388945132e-06, + "loss": 0.476, + "step": 1591 + }, + { + "epoch": 1.5322425409047162, + "grad_norm": 1.1690818920589625, + "learning_rate": 1.2900138086159315e-06, + "loss": 0.6581, + "step": 1592 + }, + { + "epoch": 1.5332050048123196, + "grad_norm": 1.0047486791072155, + "learning_rate": 1.2849503580448007e-06, + "loss": 0.4231, + "step": 1593 + }, + { + "epoch": 1.534167468719923, + "grad_norm": 1.1188824458929916, + "learning_rate": 1.2798953987543672e-06, + "loss": 0.6157, + "step": 1594 + }, + { + "epoch": 1.5351299326275265, + "grad_norm": 1.1409002570451914, + "learning_rate": 1.2748489422984655e-06, + "loss": 0.6177, + "step": 1595 + }, + { + "epoch": 1.53609239653513, + "grad_norm": 1.155723998096243, + "learning_rate": 1.2698110002115005e-06, + "loss": 0.6275, + "step": 1596 + }, + { + "epoch": 1.5370548604427334, + "grad_norm": 1.1398177807497891, + "learning_rate": 1.2647815840084115e-06, + "loss": 0.6256, + "step": 1597 + }, + { + "epoch": 1.5380173243503368, + "grad_norm": 1.1575065107461178, + "learning_rate": 1.259760705184656e-06, + "loss": 0.5966, + "step": 1598 + }, + { + "epoch": 1.5389797882579404, + "grad_norm": 1.1359360478153782, + "learning_rate": 1.2547483752161726e-06, + "loss": 0.6503, + "step": 1599 + }, + { + "epoch": 1.539942252165544, + "grad_norm": 1.0875099067436853, + "learning_rate": 1.249744605559366e-06, + "loss": 0.6016, + "step": 1600 + }, + { + "epoch": 1.5409047160731473, + "grad_norm": 1.0848032360185864, + "learning_rate": 1.244749407651068e-06, + "loss": 0.555, + "step": 1601 + }, + { + "epoch": 1.5418671799807506, + "grad_norm": 0.9442009931213268, + "learning_rate": 1.2397627929085249e-06, + "loss": 0.4613, + "step": 1602 + }, + { + "epoch": 1.5428296438883542, + "grad_norm": 1.115119579852741, + "learning_rate": 1.2347847727293621e-06, + "loss": 0.6026, + "step": 1603 + }, + { + "epoch": 1.5437921077959578, + "grad_norm": 1.0996443367416437, + "learning_rate": 1.2298153584915623e-06, + "loss": 0.5283, + "step": 1604 + }, + { + "epoch": 1.5447545717035611, + "grad_norm": 1.0796557755198135, + "learning_rate": 1.2248545615534352e-06, + "loss": 0.5456, + "step": 1605 + }, + { + "epoch": 1.5457170356111645, + "grad_norm": 0.9434387179571749, + "learning_rate": 1.2199023932535964e-06, + "loss": 0.4344, + "step": 1606 + }, + { + "epoch": 1.546679499518768, + "grad_norm": 1.0738243739262452, + "learning_rate": 1.2149588649109417e-06, + "loss": 0.5528, + "step": 1607 + }, + { + "epoch": 1.5476419634263716, + "grad_norm": 1.0151069604612344, + "learning_rate": 1.2100239878246134e-06, + "loss": 0.5519, + "step": 1608 + }, + { + "epoch": 1.548604427333975, + "grad_norm": 1.1416069329735725, + "learning_rate": 1.2050977732739872e-06, + "loss": 0.5602, + "step": 1609 + }, + { + "epoch": 1.5495668912415783, + "grad_norm": 1.0926423483309256, + "learning_rate": 1.2001802325186307e-06, + "loss": 0.5569, + "step": 1610 + }, + { + "epoch": 1.5505293551491819, + "grad_norm": 126.34868788639317, + "learning_rate": 1.1952713767982948e-06, + "loss": 0.5124, + "step": 1611 + }, + { + "epoch": 1.5514918190567855, + "grad_norm": 1.003172786726852, + "learning_rate": 1.1903712173328742e-06, + "loss": 0.4943, + "step": 1612 + }, + { + "epoch": 1.5524542829643888, + "grad_norm": 99.94991787344195, + "learning_rate": 1.185479765322391e-06, + "loss": 0.7985, + "step": 1613 + }, + { + "epoch": 1.5534167468719922, + "grad_norm": 30.43661604870284, + "learning_rate": 1.180597031946959e-06, + "loss": 0.6979, + "step": 1614 + }, + { + "epoch": 1.5543792107795957, + "grad_norm": 1.1488362626287265, + "learning_rate": 1.175723028366772e-06, + "loss": 0.5787, + "step": 1615 + }, + { + "epoch": 1.5553416746871993, + "grad_norm": 0.9946103830103348, + "learning_rate": 1.1708577657220637e-06, + "loss": 0.4909, + "step": 1616 + }, + { + "epoch": 1.5563041385948027, + "grad_norm": 1.0779800527138699, + "learning_rate": 1.166001255133095e-06, + "loss": 0.5776, + "step": 1617 + }, + { + "epoch": 1.557266602502406, + "grad_norm": 1.1655609456784075, + "learning_rate": 1.1611535077001168e-06, + "loss": 0.6697, + "step": 1618 + }, + { + "epoch": 1.5582290664100096, + "grad_norm": 1.0632766037238461, + "learning_rate": 1.1563145345033566e-06, + "loss": 0.5381, + "step": 1619 + }, + { + "epoch": 1.5591915303176132, + "grad_norm": 1.08296012480813, + "learning_rate": 1.1514843466029806e-06, + "loss": 0.604, + "step": 1620 + }, + { + "epoch": 1.5601539942252165, + "grad_norm": 1.1068222976086823, + "learning_rate": 1.1466629550390818e-06, + "loss": 0.6025, + "step": 1621 + }, + { + "epoch": 1.56111645813282, + "grad_norm": 1.1613548911759504, + "learning_rate": 1.1418503708316426e-06, + "loss": 0.6769, + "step": 1622 + }, + { + "epoch": 1.5620789220404236, + "grad_norm": 1.210438993461058, + "learning_rate": 1.1370466049805196e-06, + "loss": 0.6465, + "step": 1623 + }, + { + "epoch": 1.563041385948027, + "grad_norm": 0.9584535517351803, + "learning_rate": 1.1322516684654073e-06, + "loss": 0.4582, + "step": 1624 + }, + { + "epoch": 1.5640038498556303, + "grad_norm": 0.9945137696157945, + "learning_rate": 1.1274655722458272e-06, + "loss": 0.5031, + "step": 1625 + }, + { + "epoch": 1.564966313763234, + "grad_norm": 0.9856702072549212, + "learning_rate": 1.1226883272610877e-06, + "loss": 0.4939, + "step": 1626 + }, + { + "epoch": 1.5659287776708375, + "grad_norm": 1.010857514066895, + "learning_rate": 1.1179199444302725e-06, + "loss": 0.535, + "step": 1627 + }, + { + "epoch": 1.5668912415784408, + "grad_norm": 1.051196739846987, + "learning_rate": 1.113160434652204e-06, + "loss": 0.5425, + "step": 1628 + }, + { + "epoch": 1.5678537054860442, + "grad_norm": 1.1208680669114157, + "learning_rate": 1.108409808805429e-06, + "loss": 0.6546, + "step": 1629 + }, + { + "epoch": 1.5688161693936478, + "grad_norm": 1.050837150514082, + "learning_rate": 1.1036680777481862e-06, + "loss": 0.5521, + "step": 1630 + }, + { + "epoch": 1.5697786333012513, + "grad_norm": 1.1824815797246138, + "learning_rate": 1.098935252318386e-06, + "loss": 0.6694, + "step": 1631 + }, + { + "epoch": 1.5707410972088547, + "grad_norm": 0.9447309282073376, + "learning_rate": 1.0942113433335787e-06, + "loss": 0.4367, + "step": 1632 + }, + { + "epoch": 1.571703561116458, + "grad_norm": 1.0362072871008892, + "learning_rate": 1.0894963615909426e-06, + "loss": 0.5208, + "step": 1633 + }, + { + "epoch": 1.5726660250240616, + "grad_norm": 1.1026358161245184, + "learning_rate": 1.0847903178672426e-06, + "loss": 0.5976, + "step": 1634 + }, + { + "epoch": 1.5736284889316652, + "grad_norm": 0.9439099003640106, + "learning_rate": 1.0800932229188226e-06, + "loss": 0.4596, + "step": 1635 + }, + { + "epoch": 1.5745909528392685, + "grad_norm": 1.1011234954785174, + "learning_rate": 1.0754050874815664e-06, + "loss": 0.5689, + "step": 1636 + }, + { + "epoch": 1.5755534167468719, + "grad_norm": 1.1040465833508535, + "learning_rate": 1.0707259222708854e-06, + "loss": 0.6039, + "step": 1637 + }, + { + "epoch": 1.5765158806544755, + "grad_norm": 1.0402303281874261, + "learning_rate": 1.0660557379816816e-06, + "loss": 0.5402, + "step": 1638 + }, + { + "epoch": 1.577478344562079, + "grad_norm": 1.1153567695174917, + "learning_rate": 1.0613945452883357e-06, + "loss": 0.5804, + "step": 1639 + }, + { + "epoch": 1.5784408084696824, + "grad_norm": 0.9738564506984727, + "learning_rate": 1.0567423548446736e-06, + "loss": 0.4708, + "step": 1640 + }, + { + "epoch": 1.5794032723772857, + "grad_norm": 1.2017913127971398, + "learning_rate": 1.0520991772839472e-06, + "loss": 0.67, + "step": 1641 + }, + { + "epoch": 1.5803657362848893, + "grad_norm": 1.0843366105158052, + "learning_rate": 1.0474650232188083e-06, + "loss": 0.6071, + "step": 1642 + }, + { + "epoch": 1.5813282001924929, + "grad_norm": 1.0585022999265887, + "learning_rate": 1.0428399032412795e-06, + "loss": 0.5551, + "step": 1643 + }, + { + "epoch": 1.5822906641000962, + "grad_norm": 0.9891780249673107, + "learning_rate": 1.0382238279227418e-06, + "loss": 0.4931, + "step": 1644 + }, + { + "epoch": 1.5832531280076996, + "grad_norm": 1.097975427487035, + "learning_rate": 1.0336168078138964e-06, + "loss": 0.6131, + "step": 1645 + }, + { + "epoch": 1.5842155919153031, + "grad_norm": 0.9613965672871695, + "learning_rate": 1.0290188534447542e-06, + "loss": 0.4325, + "step": 1646 + }, + { + "epoch": 1.5851780558229067, + "grad_norm": 1.0923326978629544, + "learning_rate": 1.024429975324599e-06, + "loss": 0.6082, + "step": 1647 + }, + { + "epoch": 1.58614051973051, + "grad_norm": 1.2081866701737498, + "learning_rate": 1.0198501839419733e-06, + "loss": 0.6857, + "step": 1648 + }, + { + "epoch": 1.5871029836381134, + "grad_norm": 1.0735941681386951, + "learning_rate": 1.0152794897646495e-06, + "loss": 0.542, + "step": 1649 + }, + { + "epoch": 1.588065447545717, + "grad_norm": 1.075349804039979, + "learning_rate": 1.0107179032396086e-06, + "loss": 0.5422, + "step": 1650 + }, + { + "epoch": 1.5890279114533206, + "grad_norm": 1.161451136375637, + "learning_rate": 1.0061654347930106e-06, + "loss": 0.6614, + "step": 1651 + }, + { + "epoch": 1.589990375360924, + "grad_norm": 0.987387891329692, + "learning_rate": 1.0016220948301797e-06, + "loss": 0.4921, + "step": 1652 + }, + { + "epoch": 1.5909528392685275, + "grad_norm": 1.1292985224044991, + "learning_rate": 9.97087893735571e-07, + "loss": 0.5908, + "step": 1653 + }, + { + "epoch": 1.591915303176131, + "grad_norm": 1.0960743317774624, + "learning_rate": 9.925628418727563e-07, + "loss": 0.5835, + "step": 1654 + }, + { + "epoch": 1.5928777670837344, + "grad_norm": 2.1080857210337998, + "learning_rate": 9.880469495843903e-07, + "loss": 0.5276, + "step": 1655 + }, + { + "epoch": 1.5938402309913378, + "grad_norm": 1.2670310610319664, + "learning_rate": 9.835402271921974e-07, + "loss": 0.6652, + "step": 1656 + }, + { + "epoch": 1.5948026948989413, + "grad_norm": 1.01851138834149, + "learning_rate": 9.790426849969382e-07, + "loss": 0.5196, + "step": 1657 + }, + { + "epoch": 1.595765158806545, + "grad_norm": 0.8739524756752932, + "learning_rate": 9.74554333278393e-07, + "loss": 0.3863, + "step": 1658 + }, + { + "epoch": 1.5967276227141483, + "grad_norm": 1.1133729740993883, + "learning_rate": 9.700751822953365e-07, + "loss": 0.5905, + "step": 1659 + }, + { + "epoch": 1.5976900866217516, + "grad_norm": 1.187988485497695, + "learning_rate": 9.656052422855143e-07, + "loss": 0.5361, + "step": 1660 + }, + { + "epoch": 1.5986525505293552, + "grad_norm": 0.9988367545983375, + "learning_rate": 9.611445234656147e-07, + "loss": 0.4698, + "step": 1661 + }, + { + "epoch": 1.5996150144369587, + "grad_norm": 0.955049021645824, + "learning_rate": 9.56693036031256e-07, + "loss": 0.4421, + "step": 1662 + }, + { + "epoch": 1.600577478344562, + "grad_norm": 1.0841536001088459, + "learning_rate": 9.5225079015695e-07, + "loss": 0.5939, + "step": 1663 + }, + { + "epoch": 1.6015399422521654, + "grad_norm": 1.0508327037666645, + "learning_rate": 9.478177959960916e-07, + "loss": 0.5337, + "step": 1664 + }, + { + "epoch": 1.602502406159769, + "grad_norm": 1.1520458341632893, + "learning_rate": 9.433940636809252e-07, + "loss": 0.6592, + "step": 1665 + }, + { + "epoch": 1.6034648700673726, + "grad_norm": 0.9990138080872729, + "learning_rate": 9.389796033225301e-07, + "loss": 0.4602, + "step": 1666 + }, + { + "epoch": 1.604427333974976, + "grad_norm": 1.0483233564890486, + "learning_rate": 9.34574425010788e-07, + "loss": 0.5543, + "step": 1667 + }, + { + "epoch": 1.6053897978825793, + "grad_norm": 1.1033557242890442, + "learning_rate": 9.301785388143697e-07, + "loss": 0.5833, + "step": 1668 + }, + { + "epoch": 1.6063522617901829, + "grad_norm": 1.0773811508399296, + "learning_rate": 9.257919547807054e-07, + "loss": 0.5903, + "step": 1669 + }, + { + "epoch": 1.6073147256977864, + "grad_norm": 1.0404488712960396, + "learning_rate": 9.214146829359644e-07, + "loss": 0.5561, + "step": 1670 + }, + { + "epoch": 1.6082771896053898, + "grad_norm": 0.999317701785381, + "learning_rate": 9.170467332850325e-07, + "loss": 0.481, + "step": 1671 + }, + { + "epoch": 1.6092396535129931, + "grad_norm": 0.9888734405278206, + "learning_rate": 9.126881158114859e-07, + "loss": 0.4499, + "step": 1672 + }, + { + "epoch": 1.6102021174205967, + "grad_norm": 1.081239619097301, + "learning_rate": 9.083388404775739e-07, + "loss": 0.5969, + "step": 1673 + }, + { + "epoch": 1.6111645813282003, + "grad_norm": 1.0139651338907756, + "learning_rate": 9.039989172241887e-07, + "loss": 0.4936, + "step": 1674 + }, + { + "epoch": 1.6121270452358036, + "grad_norm": 1.2805737613103638, + "learning_rate": 8.99668355970853e-07, + "loss": 0.6992, + "step": 1675 + }, + { + "epoch": 1.613089509143407, + "grad_norm": 1.0830122621364466, + "learning_rate": 8.953471666156843e-07, + "loss": 0.5826, + "step": 1676 + }, + { + "epoch": 1.6140519730510106, + "grad_norm": 1.0801384437960286, + "learning_rate": 8.910353590353843e-07, + "loss": 0.5859, + "step": 1677 + }, + { + "epoch": 1.6150144369586141, + "grad_norm": 1.103908874016731, + "learning_rate": 8.867329430852101e-07, + "loss": 0.584, + "step": 1678 + }, + { + "epoch": 1.6159769008662175, + "grad_norm": 1.0999104717638146, + "learning_rate": 8.824399285989544e-07, + "loss": 0.5776, + "step": 1679 + }, + { + "epoch": 1.6169393647738208, + "grad_norm": 1.2332369554992586, + "learning_rate": 8.781563253889164e-07, + "loss": 0.6569, + "step": 1680 + }, + { + "epoch": 1.6179018286814244, + "grad_norm": 1.8718733776618752, + "learning_rate": 8.738821432458905e-07, + "loss": 0.4674, + "step": 1681 + }, + { + "epoch": 1.618864292589028, + "grad_norm": 56.67242099287576, + "learning_rate": 8.696173919391326e-07, + "loss": 0.8678, + "step": 1682 + }, + { + "epoch": 1.6198267564966313, + "grad_norm": 1.1070987382842836, + "learning_rate": 8.653620812163488e-07, + "loss": 0.606, + "step": 1683 + }, + { + "epoch": 1.6207892204042347, + "grad_norm": 1.1888421228842145, + "learning_rate": 8.611162208036617e-07, + "loss": 0.6631, + "step": 1684 + }, + { + "epoch": 1.6217516843118385, + "grad_norm": 1.0015606841738902, + "learning_rate": 8.568798204055978e-07, + "loss": 0.478, + "step": 1685 + }, + { + "epoch": 1.6227141482194418, + "grad_norm": 0.9802682471017025, + "learning_rate": 8.52652889705059e-07, + "loss": 0.4502, + "step": 1686 + }, + { + "epoch": 1.6236766121270452, + "grad_norm": 1.1243841692436451, + "learning_rate": 8.484354383633037e-07, + "loss": 0.6038, + "step": 1687 + }, + { + "epoch": 1.6246390760346487, + "grad_norm": 1.0808749274266665, + "learning_rate": 8.442274760199254e-07, + "loss": 0.5435, + "step": 1688 + }, + { + "epoch": 1.6256015399422523, + "grad_norm": 1.1323107904353367, + "learning_rate": 8.400290122928279e-07, + "loss": 0.6299, + "step": 1689 + }, + { + "epoch": 1.6265640038498557, + "grad_norm": 0.9247545710766062, + "learning_rate": 8.358400567782021e-07, + "loss": 0.4358, + "step": 1690 + }, + { + "epoch": 1.627526467757459, + "grad_norm": 1.0669840963453252, + "learning_rate": 8.316606190505111e-07, + "loss": 0.5552, + "step": 1691 + }, + { + "epoch": 1.6284889316650626, + "grad_norm": 0.9755050925498627, + "learning_rate": 8.274907086624595e-07, + "loss": 0.4827, + "step": 1692 + }, + { + "epoch": 1.6294513955726662, + "grad_norm": 1.1619559662063597, + "learning_rate": 8.233303351449801e-07, + "loss": 0.6393, + "step": 1693 + }, + { + "epoch": 1.6304138594802695, + "grad_norm": 1.0842448084063308, + "learning_rate": 8.191795080072023e-07, + "loss": 0.5893, + "step": 1694 + }, + { + "epoch": 1.6313763233878729, + "grad_norm": 1.073027123486928, + "learning_rate": 8.150382367364412e-07, + "loss": 0.595, + "step": 1695 + }, + { + "epoch": 1.6323387872954764, + "grad_norm": 1.150450890091677, + "learning_rate": 8.109065307981673e-07, + "loss": 0.6181, + "step": 1696 + }, + { + "epoch": 1.63330125120308, + "grad_norm": 1.1640035495804202, + "learning_rate": 8.067843996359914e-07, + "loss": 0.6346, + "step": 1697 + }, + { + "epoch": 1.6342637151106834, + "grad_norm": 1.239717922639515, + "learning_rate": 8.026718526716343e-07, + "loss": 0.7008, + "step": 1698 + }, + { + "epoch": 1.6352261790182867, + "grad_norm": 1.219583154004719, + "learning_rate": 7.985688993049174e-07, + "loss": 0.5656, + "step": 1699 + }, + { + "epoch": 1.6361886429258903, + "grad_norm": 1.0204081188084908, + "learning_rate": 7.944755489137279e-07, + "loss": 0.5221, + "step": 1700 + }, + { + "epoch": 1.6371511068334939, + "grad_norm": 1.1430553941766144, + "learning_rate": 7.903918108540087e-07, + "loss": 0.6307, + "step": 1701 + }, + { + "epoch": 1.6381135707410972, + "grad_norm": 1.0955546083011252, + "learning_rate": 7.863176944597329e-07, + "loss": 0.5784, + "step": 1702 + }, + { + "epoch": 1.6390760346487006, + "grad_norm": 1.1154602292110243, + "learning_rate": 7.822532090428769e-07, + "loss": 0.573, + "step": 1703 + }, + { + "epoch": 1.6400384985563041, + "grad_norm": 1.150370092114749, + "learning_rate": 7.781983638934093e-07, + "loss": 0.6265, + "step": 1704 + }, + { + "epoch": 1.6410009624639077, + "grad_norm": 1.0680964412736094, + "learning_rate": 7.741531682792608e-07, + "loss": 0.5957, + "step": 1705 + }, + { + "epoch": 1.641963426371511, + "grad_norm": 1.035344537635024, + "learning_rate": 7.701176314463088e-07, + "loss": 0.5489, + "step": 1706 + }, + { + "epoch": 1.6429258902791144, + "grad_norm": 1.0288568969212708, + "learning_rate": 7.660917626183534e-07, + "loss": 0.5442, + "step": 1707 + }, + { + "epoch": 1.643888354186718, + "grad_norm": 1.022500864915087, + "learning_rate": 7.62075570997099e-07, + "loss": 0.5164, + "step": 1708 + }, + { + "epoch": 1.6448508180943215, + "grad_norm": 0.9862670999793669, + "learning_rate": 7.58069065762127e-07, + "loss": 0.4928, + "step": 1709 + }, + { + "epoch": 1.645813282001925, + "grad_norm": 0.9703404381143266, + "learning_rate": 7.540722560708841e-07, + "loss": 0.4474, + "step": 1710 + }, + { + "epoch": 1.6467757459095282, + "grad_norm": 2.884570885356802, + "learning_rate": 7.500851510586521e-07, + "loss": 0.5453, + "step": 1711 + }, + { + "epoch": 1.6477382098171318, + "grad_norm": 1.1537143874429638, + "learning_rate": 7.461077598385357e-07, + "loss": 0.6411, + "step": 1712 + }, + { + "epoch": 1.6487006737247354, + "grad_norm": 160.88987114370167, + "learning_rate": 7.421400915014326e-07, + "loss": 1.5663, + "step": 1713 + }, + { + "epoch": 1.6496631376323387, + "grad_norm": 1.0150707920739481, + "learning_rate": 7.381821551160207e-07, + "loss": 0.4802, + "step": 1714 + }, + { + "epoch": 1.650625601539942, + "grad_norm": 0.8556308453602829, + "learning_rate": 7.34233959728734e-07, + "loss": 0.3883, + "step": 1715 + }, + { + "epoch": 1.6515880654475459, + "grad_norm": 1.2030563484617816, + "learning_rate": 7.302955143637419e-07, + "loss": 0.6633, + "step": 1716 + }, + { + "epoch": 1.6525505293551492, + "grad_norm": 1.005278405927379, + "learning_rate": 7.263668280229258e-07, + "loss": 0.5391, + "step": 1717 + }, + { + "epoch": 1.6535129932627526, + "grad_norm": 1.1797250931101002, + "learning_rate": 7.224479096858672e-07, + "loss": 0.6518, + "step": 1718 + }, + { + "epoch": 1.6544754571703562, + "grad_norm": 0.9682150206365097, + "learning_rate": 7.185387683098144e-07, + "loss": 0.4435, + "step": 1719 + }, + { + "epoch": 1.6554379210779597, + "grad_norm": 1.145352860433583, + "learning_rate": 7.146394128296758e-07, + "loss": 0.6129, + "step": 1720 + }, + { + "epoch": 1.656400384985563, + "grad_norm": 1.209732810957726, + "learning_rate": 7.10749852157987e-07, + "loss": 0.6641, + "step": 1721 + }, + { + "epoch": 1.6573628488931664, + "grad_norm": 1.1016059457078504, + "learning_rate": 7.068700951849012e-07, + "loss": 0.5988, + "step": 1722 + }, + { + "epoch": 1.65832531280077, + "grad_norm": 4.540180373897326, + "learning_rate": 7.030001507781592e-07, + "loss": 0.5097, + "step": 1723 + }, + { + "epoch": 1.6592877767083736, + "grad_norm": 1.0851043305040822, + "learning_rate": 6.991400277830768e-07, + "loss": 0.5952, + "step": 1724 + }, + { + "epoch": 1.660250240615977, + "grad_norm": 1.021454267169162, + "learning_rate": 6.952897350225208e-07, + "loss": 0.4951, + "step": 1725 + }, + { + "epoch": 1.6612127045235803, + "grad_norm": 1.2618596779485607, + "learning_rate": 6.91449281296891e-07, + "loss": 0.6752, + "step": 1726 + }, + { + "epoch": 1.6621751684311838, + "grad_norm": 1.0758796983111472, + "learning_rate": 6.876186753840947e-07, + "loss": 0.5804, + "step": 1727 + }, + { + "epoch": 1.6631376323387874, + "grad_norm": 1.2058951827578022, + "learning_rate": 6.83797926039535e-07, + "loss": 0.6551, + "step": 1728 + }, + { + "epoch": 1.6641000962463908, + "grad_norm": 0.9382620002847977, + "learning_rate": 6.79987041996083e-07, + "loss": 0.4301, + "step": 1729 + }, + { + "epoch": 1.6650625601539941, + "grad_norm": 1.1008104083192982, + "learning_rate": 6.761860319640639e-07, + "loss": 0.5911, + "step": 1730 + }, + { + "epoch": 1.6660250240615977, + "grad_norm": 1.1039428757437446, + "learning_rate": 6.723949046312322e-07, + "loss": 0.5971, + "step": 1731 + }, + { + "epoch": 1.6669874879692013, + "grad_norm": 1.1292737661230046, + "learning_rate": 6.686136686627554e-07, + "loss": 0.6493, + "step": 1732 + }, + { + "epoch": 1.6679499518768046, + "grad_norm": 1.2282299850422176, + "learning_rate": 6.648423327011938e-07, + "loss": 0.6966, + "step": 1733 + }, + { + "epoch": 1.668912415784408, + "grad_norm": 1.0857096031114082, + "learning_rate": 6.610809053664768e-07, + "loss": 0.5789, + "step": 1734 + }, + { + "epoch": 1.6698748796920115, + "grad_norm": 1.060303910046835, + "learning_rate": 6.573293952558885e-07, + "loss": 0.4999, + "step": 1735 + }, + { + "epoch": 1.670837343599615, + "grad_norm": 1.0751279779981688, + "learning_rate": 6.535878109440458e-07, + "loss": 0.5781, + "step": 1736 + }, + { + "epoch": 1.6717998075072185, + "grad_norm": 1.067293404240938, + "learning_rate": 6.49856160982879e-07, + "loss": 0.5796, + "step": 1737 + }, + { + "epoch": 1.6727622714148218, + "grad_norm": 0.9605082450675138, + "learning_rate": 6.46134453901609e-07, + "loss": 0.4716, + "step": 1738 + }, + { + "epoch": 1.6737247353224254, + "grad_norm": 0.9991398531200738, + "learning_rate": 6.42422698206735e-07, + "loss": 0.5129, + "step": 1739 + }, + { + "epoch": 1.674687199230029, + "grad_norm": 1.0893530773573885, + "learning_rate": 6.387209023820073e-07, + "loss": 0.5249, + "step": 1740 + }, + { + "epoch": 1.6756496631376323, + "grad_norm": 1.0296913173097668, + "learning_rate": 6.350290748884147e-07, + "loss": 0.5109, + "step": 1741 + }, + { + "epoch": 1.6766121270452357, + "grad_norm": 1.1812742125000781, + "learning_rate": 6.313472241641588e-07, + "loss": 0.6775, + "step": 1742 + }, + { + "epoch": 1.6775745909528392, + "grad_norm": 1.0139138706145896, + "learning_rate": 6.276753586246398e-07, + "loss": 0.5063, + "step": 1743 + }, + { + "epoch": 1.6785370548604428, + "grad_norm": 1.1864454301031742, + "learning_rate": 6.240134866624353e-07, + "loss": 0.7364, + "step": 1744 + }, + { + "epoch": 1.6794995187680462, + "grad_norm": 1.0452147267809815, + "learning_rate": 6.203616166472821e-07, + "loss": 0.521, + "step": 1745 + }, + { + "epoch": 1.6804619826756495, + "grad_norm": 1.0716660166283567, + "learning_rate": 6.16719756926052e-07, + "loss": 0.5577, + "step": 1746 + }, + { + "epoch": 1.681424446583253, + "grad_norm": 1.2077123852739577, + "learning_rate": 6.130879158227415e-07, + "loss": 0.6915, + "step": 1747 + }, + { + "epoch": 1.6823869104908566, + "grad_norm": 1.1920914696790892, + "learning_rate": 6.094661016384451e-07, + "loss": 0.6536, + "step": 1748 + }, + { + "epoch": 1.68334937439846, + "grad_norm": 0.9870337205949054, + "learning_rate": 6.05854322651342e-07, + "loss": 0.485, + "step": 1749 + }, + { + "epoch": 1.6843118383060636, + "grad_norm": 1.154152203115481, + "learning_rate": 6.022525871166706e-07, + "loss": 0.6547, + "step": 1750 + }, + { + "epoch": 1.6852743022136671, + "grad_norm": 1.1416657018216354, + "learning_rate": 5.986609032667184e-07, + "loss": 0.6652, + "step": 1751 + }, + { + "epoch": 1.6862367661212705, + "grad_norm": 1.0608976958921017, + "learning_rate": 5.950792793107935e-07, + "loss": 0.5986, + "step": 1752 + }, + { + "epoch": 1.6871992300288738, + "grad_norm": 1.161464267123805, + "learning_rate": 5.915077234352146e-07, + "loss": 0.5955, + "step": 1753 + }, + { + "epoch": 1.6881616939364774, + "grad_norm": 1.1094804721332379, + "learning_rate": 5.879462438032857e-07, + "loss": 0.5804, + "step": 1754 + }, + { + "epoch": 1.689124157844081, + "grad_norm": 1.156565464254963, + "learning_rate": 5.843948485552831e-07, + "loss": 0.6654, + "step": 1755 + }, + { + "epoch": 1.6900866217516843, + "grad_norm": 1.0505240502756652, + "learning_rate": 5.808535458084292e-07, + "loss": 0.578, + "step": 1756 + }, + { + "epoch": 1.6910490856592877, + "grad_norm": 0.9596350625878437, + "learning_rate": 5.773223436568831e-07, + "loss": 0.453, + "step": 1757 + }, + { + "epoch": 1.6920115495668913, + "grad_norm": 1.1685452139078851, + "learning_rate": 5.738012501717144e-07, + "loss": 0.6635, + "step": 1758 + }, + { + "epoch": 1.6929740134744948, + "grad_norm": 1.0072524856711287, + "learning_rate": 5.702902734008903e-07, + "loss": 0.5104, + "step": 1759 + }, + { + "epoch": 1.6939364773820982, + "grad_norm": 1.0784567757461934, + "learning_rate": 5.667894213692515e-07, + "loss": 0.5722, + "step": 1760 + }, + { + "epoch": 1.6948989412897015, + "grad_norm": 1.0544716031667956, + "learning_rate": 5.632987020785008e-07, + "loss": 0.5381, + "step": 1761 + }, + { + "epoch": 1.695861405197305, + "grad_norm": 1.0425022097550323, + "learning_rate": 5.598181235071798e-07, + "loss": 0.5372, + "step": 1762 + }, + { + "epoch": 1.6968238691049087, + "grad_norm": 1.1105666872634687, + "learning_rate": 5.563476936106515e-07, + "loss": 0.5779, + "step": 1763 + }, + { + "epoch": 1.697786333012512, + "grad_norm": 1.1060362081739836, + "learning_rate": 5.528874203210827e-07, + "loss": 0.6037, + "step": 1764 + }, + { + "epoch": 1.6987487969201154, + "grad_norm": 0.9521782995911832, + "learning_rate": 5.494373115474256e-07, + "loss": 0.4723, + "step": 1765 + }, + { + "epoch": 1.699711260827719, + "grad_norm": 0.9403863314837697, + "learning_rate": 5.459973751754022e-07, + "loss": 0.4395, + "step": 1766 + }, + { + "epoch": 1.7006737247353225, + "grad_norm": 1.0660067364471093, + "learning_rate": 5.4256761906748e-07, + "loss": 0.5517, + "step": 1767 + }, + { + "epoch": 1.7016361886429259, + "grad_norm": 1.060682834760489, + "learning_rate": 5.391480510628622e-07, + "loss": 0.5239, + "step": 1768 + }, + { + "epoch": 1.7025986525505292, + "grad_norm": 1.0868252931765485, + "learning_rate": 5.357386789774615e-07, + "loss": 0.5859, + "step": 1769 + }, + { + "epoch": 1.7035611164581328, + "grad_norm": 0.9495176645012299, + "learning_rate": 5.323395106038909e-07, + "loss": 0.429, + "step": 1770 + }, + { + "epoch": 1.7045235803657364, + "grad_norm": 1.1345498462678634, + "learning_rate": 5.28950553711437e-07, + "loss": 0.6366, + "step": 1771 + }, + { + "epoch": 1.7054860442733397, + "grad_norm": 1.0951137287485226, + "learning_rate": 5.255718160460493e-07, + "loss": 0.5979, + "step": 1772 + }, + { + "epoch": 1.706448508180943, + "grad_norm": 1.036251501144284, + "learning_rate": 5.222033053303194e-07, + "loss": 0.5792, + "step": 1773 + }, + { + "epoch": 1.7074109720885466, + "grad_norm": 1.1464029553047155, + "learning_rate": 5.188450292634645e-07, + "loss": 0.6095, + "step": 1774 + }, + { + "epoch": 1.7083734359961502, + "grad_norm": 0.9356081869277373, + "learning_rate": 5.154969955213051e-07, + "loss": 0.4594, + "step": 1775 + }, + { + "epoch": 1.7093358999037536, + "grad_norm": 1.0118786640793236, + "learning_rate": 5.121592117562574e-07, + "loss": 0.5045, + "step": 1776 + }, + { + "epoch": 1.710298363811357, + "grad_norm": 1.1163503044354839, + "learning_rate": 5.088316855973036e-07, + "loss": 0.5967, + "step": 1777 + }, + { + "epoch": 1.7112608277189605, + "grad_norm": 1.0855003795610316, + "learning_rate": 5.05514424649986e-07, + "loss": 0.5896, + "step": 1778 + }, + { + "epoch": 1.712223291626564, + "grad_norm": 1.150459680831111, + "learning_rate": 5.022074364963803e-07, + "loss": 0.5846, + "step": 1779 + }, + { + "epoch": 1.7131857555341674, + "grad_norm": 0.9521368653276568, + "learning_rate": 4.989107286950845e-07, + "loss": 0.4531, + "step": 1780 + }, + { + "epoch": 1.714148219441771, + "grad_norm": 1.068272771336741, + "learning_rate": 4.956243087811996e-07, + "loss": 0.5626, + "step": 1781 + }, + { + "epoch": 1.7151106833493746, + "grad_norm": 1.0119628556253373, + "learning_rate": 4.923481842663114e-07, + "loss": 0.5098, + "step": 1782 + }, + { + "epoch": 1.716073147256978, + "grad_norm": 1.1426856840293431, + "learning_rate": 4.890823626384722e-07, + "loss": 0.6582, + "step": 1783 + }, + { + "epoch": 1.7170356111645813, + "grad_norm": 1.1683102503109855, + "learning_rate": 4.8582685136219e-07, + "loss": 0.6717, + "step": 1784 + }, + { + "epoch": 1.7179980750721848, + "grad_norm": 1.0786196177678815, + "learning_rate": 4.825816578784015e-07, + "loss": 0.5758, + "step": 1785 + }, + { + "epoch": 1.7189605389797884, + "grad_norm": 1.129502086041976, + "learning_rate": 4.793467896044668e-07, + "loss": 0.6367, + "step": 1786 + }, + { + "epoch": 1.7199230028873917, + "grad_norm": 0.9958485113831057, + "learning_rate": 4.761222539341398e-07, + "loss": 0.5099, + "step": 1787 + }, + { + "epoch": 1.720885466794995, + "grad_norm": 1.0159468090413972, + "learning_rate": 4.729080582375634e-07, + "loss": 0.5221, + "step": 1788 + }, + { + "epoch": 1.7218479307025987, + "grad_norm": 0.8817650611057956, + "learning_rate": 4.6970420986124256e-07, + "loss": 0.4037, + "step": 1789 + }, + { + "epoch": 1.7228103946102022, + "grad_norm": 1.1012731934317603, + "learning_rate": 4.6651071612803466e-07, + "loss": 0.6251, + "step": 1790 + }, + { + "epoch": 1.7237728585178056, + "grad_norm": 0.8899481771992627, + "learning_rate": 4.633275843371293e-07, + "loss": 0.4069, + "step": 1791 + }, + { + "epoch": 1.724735322425409, + "grad_norm": 1.0995080180775918, + "learning_rate": 4.6015482176403337e-07, + "loss": 0.5854, + "step": 1792 + }, + { + "epoch": 1.7256977863330125, + "grad_norm": 1.008765892394184, + "learning_rate": 4.56992435660551e-07, + "loss": 0.5017, + "step": 1793 + }, + { + "epoch": 1.726660250240616, + "grad_norm": 1.171547718993326, + "learning_rate": 4.5384043325477187e-07, + "loss": 0.6641, + "step": 1794 + }, + { + "epoch": 1.7276227141482194, + "grad_norm": 0.9869471757845772, + "learning_rate": 4.5069882175104997e-07, + "loss": 0.5165, + "step": 1795 + }, + { + "epoch": 1.7285851780558228, + "grad_norm": 1.032084912635945, + "learning_rate": 4.4756760832999126e-07, + "loss": 0.5238, + "step": 1796 + }, + { + "epoch": 1.7295476419634264, + "grad_norm": 1.1443191210668662, + "learning_rate": 4.4444680014843535e-07, + "loss": 0.6573, + "step": 1797 + }, + { + "epoch": 1.73051010587103, + "grad_norm": 0.9583815255230036, + "learning_rate": 4.413364043394375e-07, + "loss": 0.4862, + "step": 1798 + }, + { + "epoch": 1.7314725697786333, + "grad_norm": 1.1212653027218114, + "learning_rate": 4.382364280122559e-07, + "loss": 0.6439, + "step": 1799 + }, + { + "epoch": 1.7324350336862366, + "grad_norm": 1.0649320261072825, + "learning_rate": 4.3514687825233157e-07, + "loss": 0.5757, + "step": 1800 + }, + { + "epoch": 1.7333974975938402, + "grad_norm": 1.0795206516722735, + "learning_rate": 4.3206776212127597e-07, + "loss": 0.5876, + "step": 1801 + }, + { + "epoch": 1.7343599615014438, + "grad_norm": 1.089012700351658, + "learning_rate": 4.289990866568516e-07, + "loss": 0.585, + "step": 1802 + }, + { + "epoch": 1.7353224254090471, + "grad_norm": 1.0134454717219266, + "learning_rate": 4.259408588729597e-07, + "loss": 0.5258, + "step": 1803 + }, + { + "epoch": 1.7362848893166505, + "grad_norm": 1.022889852973938, + "learning_rate": 4.2289308575961805e-07, + "loss": 0.5306, + "step": 1804 + }, + { + "epoch": 1.737247353224254, + "grad_norm": 1.1826754649916251, + "learning_rate": 4.1985577428295233e-07, + "loss": 0.6657, + "step": 1805 + }, + { + "epoch": 1.7382098171318576, + "grad_norm": 1.2065060626025967, + "learning_rate": 4.168289313851731e-07, + "loss": 0.6973, + "step": 1806 + }, + { + "epoch": 1.739172281039461, + "grad_norm": 1.0803072884414586, + "learning_rate": 4.138125639845675e-07, + "loss": 0.5903, + "step": 1807 + }, + { + "epoch": 1.7401347449470643, + "grad_norm": 1.0871763033282515, + "learning_rate": 4.108066789754761e-07, + "loss": 0.6183, + "step": 1808 + }, + { + "epoch": 1.741097208854668, + "grad_norm": 1.2033360453889903, + "learning_rate": 4.078112832282821e-07, + "loss": 0.6046, + "step": 1809 + }, + { + "epoch": 1.7420596727622715, + "grad_norm": 1.1484890530660694, + "learning_rate": 4.0482638358939397e-07, + "loss": 0.6288, + "step": 1810 + }, + { + "epoch": 1.7430221366698748, + "grad_norm": 1.0737785816117196, + "learning_rate": 4.0185198688122986e-07, + "loss": 0.582, + "step": 1811 + }, + { + "epoch": 1.7439846005774784, + "grad_norm": 1.2430608700657932, + "learning_rate": 3.988880999021999e-07, + "loss": 0.6549, + "step": 1812 + }, + { + "epoch": 1.744947064485082, + "grad_norm": 1.1167972536776571, + "learning_rate": 3.9593472942669685e-07, + "loss": 0.6018, + "step": 1813 + }, + { + "epoch": 1.7459095283926853, + "grad_norm": 0.9899566692488727, + "learning_rate": 3.9299188220507157e-07, + "loss": 0.4908, + "step": 1814 + }, + { + "epoch": 1.7468719923002887, + "grad_norm": 1.0649618097419264, + "learning_rate": 3.9005956496362687e-07, + "loss": 0.5479, + "step": 1815 + }, + { + "epoch": 1.7478344562078922, + "grad_norm": 1.1072749980008814, + "learning_rate": 3.8713778440459484e-07, + "loss": 0.5904, + "step": 1816 + }, + { + "epoch": 1.7487969201154958, + "grad_norm": 1.08013231917945, + "learning_rate": 3.8422654720612575e-07, + "loss": 0.6049, + "step": 1817 + }, + { + "epoch": 1.7497593840230992, + "grad_norm": 1.0479128868298608, + "learning_rate": 3.813258600222708e-07, + "loss": 0.6006, + "step": 1818 + }, + { + "epoch": 1.7507218479307025, + "grad_norm": 1.071683538991792, + "learning_rate": 3.7843572948296883e-07, + "loss": 0.5608, + "step": 1819 + }, + { + "epoch": 1.751684311838306, + "grad_norm": 1.1524268256689412, + "learning_rate": 3.7555616219402847e-07, + "loss": 0.5718, + "step": 1820 + }, + { + "epoch": 1.7526467757459097, + "grad_norm": 0.8629230658424345, + "learning_rate": 3.726871647371166e-07, + "loss": 0.3818, + "step": 1821 + }, + { + "epoch": 1.753609239653513, + "grad_norm": 1.145130406436901, + "learning_rate": 3.6982874366973785e-07, + "loss": 0.6697, + "step": 1822 + }, + { + "epoch": 1.7545717035611164, + "grad_norm": 1.1652687408013562, + "learning_rate": 3.669809055252277e-07, + "loss": 0.6626, + "step": 1823 + }, + { + "epoch": 1.75553416746872, + "grad_norm": 1.0005295020099618, + "learning_rate": 3.641436568127271e-07, + "loss": 0.4953, + "step": 1824 + }, + { + "epoch": 1.7564966313763235, + "grad_norm": 1.0965859088296352, + "learning_rate": 3.613170040171787e-07, + "loss": 0.6004, + "step": 1825 + }, + { + "epoch": 1.7574590952839269, + "grad_norm": 0.9916026142339988, + "learning_rate": 3.585009535993023e-07, + "loss": 0.5014, + "step": 1826 + }, + { + "epoch": 1.7584215591915302, + "grad_norm": 1.05957821641981, + "learning_rate": 3.556955119955879e-07, + "loss": 0.517, + "step": 1827 + }, + { + "epoch": 1.7593840230991338, + "grad_norm": 1.0629845684144184, + "learning_rate": 3.5290068561827596e-07, + "loss": 0.5399, + "step": 1828 + }, + { + "epoch": 1.7603464870067373, + "grad_norm": 1.1236828196465076, + "learning_rate": 3.501164808553442e-07, + "loss": 0.5875, + "step": 1829 + }, + { + "epoch": 1.7613089509143407, + "grad_norm": 1.0747996839368146, + "learning_rate": 3.47342904070494e-07, + "loss": 0.5515, + "step": 1830 + }, + { + "epoch": 1.762271414821944, + "grad_norm": 1.1805531478132056, + "learning_rate": 3.4457996160313314e-07, + "loss": 0.5638, + "step": 1831 + }, + { + "epoch": 1.7632338787295476, + "grad_norm": 1.151870923362082, + "learning_rate": 3.418276597683662e-07, + "loss": 0.6876, + "step": 1832 + }, + { + "epoch": 1.7641963426371512, + "grad_norm": 0.7660264600249259, + "learning_rate": 3.3908600485697273e-07, + "loss": 0.3172, + "step": 1833 + }, + { + "epoch": 1.7651588065447545, + "grad_norm": 1.175290773078034, + "learning_rate": 3.3635500313540124e-07, + "loss": 0.6826, + "step": 1834 + }, + { + "epoch": 1.766121270452358, + "grad_norm": 1.0178013313365637, + "learning_rate": 3.336346608457486e-07, + "loss": 0.5435, + "step": 1835 + }, + { + "epoch": 1.7670837343599615, + "grad_norm": 0.9918931934509696, + "learning_rate": 3.3092498420574994e-07, + "loss": 0.4709, + "step": 1836 + }, + { + "epoch": 1.768046198267565, + "grad_norm": 1.1194465893675118, + "learning_rate": 3.282259794087589e-07, + "loss": 0.5847, + "step": 1837 + }, + { + "epoch": 1.7690086621751684, + "grad_norm": 1.0515544986091432, + "learning_rate": 3.255376526237414e-07, + "loss": 0.5559, + "step": 1838 + }, + { + "epoch": 1.7699711260827717, + "grad_norm": 1.0098221856147724, + "learning_rate": 3.2286000999525445e-07, + "loss": 0.5295, + "step": 1839 + }, + { + "epoch": 1.7709335899903753, + "grad_norm": 1.0800609396162029, + "learning_rate": 3.2019305764343644e-07, + "loss": 0.6311, + "step": 1840 + }, + { + "epoch": 1.7718960538979789, + "grad_norm": 1.052433686555485, + "learning_rate": 3.1753680166399015e-07, + "loss": 0.5375, + "step": 1841 + }, + { + "epoch": 1.7728585178055822, + "grad_norm": 1.1058337616444847, + "learning_rate": 3.148912481281713e-07, + "loss": 0.6121, + "step": 1842 + }, + { + "epoch": 1.7738209817131858, + "grad_norm": 1.0695409282446608, + "learning_rate": 3.1225640308277237e-07, + "loss": 0.5893, + "step": 1843 + }, + { + "epoch": 1.7747834456207894, + "grad_norm": 1.1316790801662548, + "learning_rate": 3.0963227255011273e-07, + "loss": 0.6382, + "step": 1844 + }, + { + "epoch": 1.7757459095283927, + "grad_norm": 1.1287609161832328, + "learning_rate": 3.0701886252801906e-07, + "loss": 0.6086, + "step": 1845 + }, + { + "epoch": 1.776708373435996, + "grad_norm": 1.1217383260121874, + "learning_rate": 3.044161789898159e-07, + "loss": 0.5625, + "step": 1846 + }, + { + "epoch": 1.7776708373435997, + "grad_norm": 14.056951621726066, + "learning_rate": 3.01824227884312e-07, + "loss": 0.5503, + "step": 1847 + }, + { + "epoch": 1.7786333012512032, + "grad_norm": 1.064722114174048, + "learning_rate": 2.99243015135785e-07, + "loss": 0.5424, + "step": 1848 + }, + { + "epoch": 1.7795957651588066, + "grad_norm": 1.0321096646774204, + "learning_rate": 2.9667254664396685e-07, + "loss": 0.5402, + "step": 1849 + }, + { + "epoch": 1.78055822906641, + "grad_norm": 0.9720488121879376, + "learning_rate": 2.941128282840344e-07, + "loss": 0.4943, + "step": 1850 + }, + { + "epoch": 1.7815206929740135, + "grad_norm": 1.1074132151401392, + "learning_rate": 2.915638659065917e-07, + "loss": 0.5628, + "step": 1851 + }, + { + "epoch": 1.782483156881617, + "grad_norm": 1.0927589977555838, + "learning_rate": 2.8902566533766e-07, + "loss": 0.5426, + "step": 1852 + }, + { + "epoch": 1.7834456207892204, + "grad_norm": 1.0035995705683731, + "learning_rate": 2.864982323786608e-07, + "loss": 0.5077, + "step": 1853 + }, + { + "epoch": 1.7844080846968238, + "grad_norm": 1.0818003735640755, + "learning_rate": 2.839815728064077e-07, + "loss": 0.5993, + "step": 1854 + }, + { + "epoch": 1.7853705486044273, + "grad_norm": 0.9035906890614125, + "learning_rate": 2.8147569237308603e-07, + "loss": 0.4086, + "step": 1855 + }, + { + "epoch": 1.786333012512031, + "grad_norm": 0.9692917008955094, + "learning_rate": 2.7898059680624754e-07, + "loss": 0.4714, + "step": 1856 + }, + { + "epoch": 1.7872954764196343, + "grad_norm": 1.0414623318193708, + "learning_rate": 2.7649629180879265e-07, + "loss": 0.5628, + "step": 1857 + }, + { + "epoch": 1.7882579403272376, + "grad_norm": 1.0917057575257862, + "learning_rate": 2.7402278305895757e-07, + "loss": 0.5742, + "step": 1858 + }, + { + "epoch": 1.7892204042348412, + "grad_norm": 0.9881177127229289, + "learning_rate": 2.7156007621030323e-07, + "loss": 0.4698, + "step": 1859 + }, + { + "epoch": 1.7901828681424448, + "grad_norm": 1.1007928426179594, + "learning_rate": 2.691081768916992e-07, + "loss": 0.6015, + "step": 1860 + }, + { + "epoch": 1.7911453320500481, + "grad_norm": 1.037166884986354, + "learning_rate": 2.66667090707316e-07, + "loss": 0.516, + "step": 1861 + }, + { + "epoch": 1.7921077959576515, + "grad_norm": 1.0179411903413425, + "learning_rate": 2.642368232366055e-07, + "loss": 0.545, + "step": 1862 + }, + { + "epoch": 1.793070259865255, + "grad_norm": 1.1436097663966587, + "learning_rate": 2.61817380034296e-07, + "loss": 0.6554, + "step": 1863 + }, + { + "epoch": 1.7940327237728586, + "grad_norm": 1.1271676799765595, + "learning_rate": 2.594087666303713e-07, + "loss": 0.6074, + "step": 1864 + }, + { + "epoch": 1.794995187680462, + "grad_norm": 0.9772842010410305, + "learning_rate": 2.5701098853006536e-07, + "loss": 0.4929, + "step": 1865 + }, + { + "epoch": 1.7959576515880653, + "grad_norm": 1.0643617532178886, + "learning_rate": 2.546240512138443e-07, + "loss": 0.5757, + "step": 1866 + }, + { + "epoch": 1.7969201154956689, + "grad_norm": 1.0010551255592657, + "learning_rate": 2.522479601373967e-07, + "loss": 0.5166, + "step": 1867 + }, + { + "epoch": 1.7978825794032725, + "grad_norm": 1.0722792283984761, + "learning_rate": 2.498827207316218e-07, + "loss": 0.5518, + "step": 1868 + }, + { + "epoch": 1.7988450433108758, + "grad_norm": 1.1267236743053854, + "learning_rate": 2.4752833840261445e-07, + "loss": 0.6201, + "step": 1869 + }, + { + "epoch": 1.7998075072184792, + "grad_norm": 1.0443376954196628, + "learning_rate": 2.4518481853165355e-07, + "loss": 0.5857, + "step": 1870 + }, + { + "epoch": 1.8007699711260827, + "grad_norm": 1.1688210196390214, + "learning_rate": 2.4285216647519263e-07, + "loss": 0.652, + "step": 1871 + }, + { + "epoch": 1.8017324350336863, + "grad_norm": 1.1500149117852727, + "learning_rate": 2.4053038756484183e-07, + "loss": 0.6517, + "step": 1872 + }, + { + "epoch": 1.8026948989412896, + "grad_norm": 1.1851849480110834, + "learning_rate": 2.3821948710736232e-07, + "loss": 0.6694, + "step": 1873 + }, + { + "epoch": 1.8036573628488932, + "grad_norm": 1.1207853276016735, + "learning_rate": 2.35919470384649e-07, + "loss": 0.6091, + "step": 1874 + }, + { + "epoch": 1.8046198267564968, + "grad_norm": 0.8885840869481474, + "learning_rate": 2.336303426537212e-07, + "loss": 0.3908, + "step": 1875 + }, + { + "epoch": 1.8055822906641001, + "grad_norm": 1.051828816517248, + "learning_rate": 2.3135210914671025e-07, + "loss": 0.5382, + "step": 1876 + }, + { + "epoch": 1.8065447545717035, + "grad_norm": 0.9484947404818663, + "learning_rate": 2.2908477507084692e-07, + "loss": 0.4796, + "step": 1877 + }, + { + "epoch": 1.807507218479307, + "grad_norm": 1.0149377661580217, + "learning_rate": 2.2682834560844913e-07, + "loss": 0.5206, + "step": 1878 + }, + { + "epoch": 1.8084696823869106, + "grad_norm": 1.1311127630503612, + "learning_rate": 2.2458282591691194e-07, + "loss": 0.6198, + "step": 1879 + }, + { + "epoch": 1.809432146294514, + "grad_norm": 2.0323396061894257, + "learning_rate": 2.2234822112869258e-07, + "loss": 0.6764, + "step": 1880 + }, + { + "epoch": 1.8103946102021173, + "grad_norm": 1.1356017286930367, + "learning_rate": 2.2012453635130327e-07, + "loss": 0.6295, + "step": 1881 + }, + { + "epoch": 1.811357074109721, + "grad_norm": 1.0145426810174205, + "learning_rate": 2.1791177666729447e-07, + "loss": 0.5019, + "step": 1882 + }, + { + "epoch": 1.8123195380173245, + "grad_norm": 1.0547585609048697, + "learning_rate": 2.1570994713424774e-07, + "loss": 0.5507, + "step": 1883 + }, + { + "epoch": 1.8132820019249278, + "grad_norm": 1.1321545375891122, + "learning_rate": 2.135190527847608e-07, + "loss": 0.6698, + "step": 1884 + }, + { + "epoch": 1.8142444658325312, + "grad_norm": 1.019613074957215, + "learning_rate": 2.1133909862643854e-07, + "loss": 0.5306, + "step": 1885 + }, + { + "epoch": 1.8152069297401348, + "grad_norm": 1.1279893748997447, + "learning_rate": 2.0917008964187968e-07, + "loss": 0.6438, + "step": 1886 + }, + { + "epoch": 1.8161693936477383, + "grad_norm": 1.1785340803486217, + "learning_rate": 2.0701203078866693e-07, + "loss": 0.6582, + "step": 1887 + }, + { + "epoch": 1.8171318575553417, + "grad_norm": 1.0587744435693698, + "learning_rate": 2.0486492699935356e-07, + "loss": 0.5505, + "step": 1888 + }, + { + "epoch": 1.818094321462945, + "grad_norm": 1.0112476161874004, + "learning_rate": 2.027287831814556e-07, + "loss": 0.527, + "step": 1889 + }, + { + "epoch": 1.8190567853705486, + "grad_norm": 1.1929805412995642, + "learning_rate": 2.0060360421743585e-07, + "loss": 0.6677, + "step": 1890 + }, + { + "epoch": 1.8200192492781522, + "grad_norm": 1.0142179771245567, + "learning_rate": 1.984893949646971e-07, + "loss": 0.5483, + "step": 1891 + }, + { + "epoch": 1.8209817131857555, + "grad_norm": 1.0337850507645243, + "learning_rate": 1.9638616025557e-07, + "loss": 0.5032, + "step": 1892 + }, + { + "epoch": 1.8219441770933589, + "grad_norm": 0.9839816230713564, + "learning_rate": 1.942939048972986e-07, + "loss": 0.4903, + "step": 1893 + }, + { + "epoch": 1.8229066410009624, + "grad_norm": 1.1312010908953731, + "learning_rate": 1.922126336720348e-07, + "loss": 0.6585, + "step": 1894 + }, + { + "epoch": 1.823869104908566, + "grad_norm": 1.1403767587454792, + "learning_rate": 1.901423513368228e-07, + "loss": 0.6804, + "step": 1895 + }, + { + "epoch": 1.8248315688161694, + "grad_norm": 1.1615774330803523, + "learning_rate": 1.8808306262359245e-07, + "loss": 0.6435, + "step": 1896 + }, + { + "epoch": 1.8257940327237727, + "grad_norm": 0.9795064773806081, + "learning_rate": 1.8603477223914202e-07, + "loss": 0.4485, + "step": 1897 + }, + { + "epoch": 1.8267564966313763, + "grad_norm": 1.029293289319028, + "learning_rate": 1.8399748486513657e-07, + "loss": 0.5712, + "step": 1898 + }, + { + "epoch": 1.8277189605389799, + "grad_norm": 1.0613830796689978, + "learning_rate": 1.819712051580874e-07, + "loss": 0.5536, + "step": 1899 + }, + { + "epoch": 1.8286814244465832, + "grad_norm": 1.1321203934444095, + "learning_rate": 1.799559377493504e-07, + "loss": 0.6614, + "step": 1900 + }, + { + "epoch": 1.8296438883541866, + "grad_norm": 1.0764256189472023, + "learning_rate": 1.7795168724510704e-07, + "loss": 0.6001, + "step": 1901 + }, + { + "epoch": 1.8306063522617901, + "grad_norm": 1.1097750227861125, + "learning_rate": 1.7595845822636292e-07, + "loss": 0.6015, + "step": 1902 + }, + { + "epoch": 1.8315688161693937, + "grad_norm": 1.0746360467259104, + "learning_rate": 1.739762552489277e-07, + "loss": 0.5927, + "step": 1903 + }, + { + "epoch": 1.832531280076997, + "grad_norm": 1.012045751524965, + "learning_rate": 1.7200508284341277e-07, + "loss": 0.5448, + "step": 1904 + }, + { + "epoch": 1.8334937439846006, + "grad_norm": 1.111647183111529, + "learning_rate": 1.700449455152159e-07, + "loss": 0.5595, + "step": 1905 + }, + { + "epoch": 1.8344562078922042, + "grad_norm": 1.116794051072117, + "learning_rate": 1.6809584774451382e-07, + "loss": 0.6077, + "step": 1906 + }, + { + "epoch": 1.8354186717998076, + "grad_norm": 1.0263421697017145, + "learning_rate": 1.661577939862491e-07, + "loss": 0.5254, + "step": 1907 + }, + { + "epoch": 1.836381135707411, + "grad_norm": 1.0351904508790186, + "learning_rate": 1.6423078867012388e-07, + "loss": 0.5267, + "step": 1908 + }, + { + "epoch": 1.8373435996150145, + "grad_norm": 1.0385996725197808, + "learning_rate": 1.623148362005844e-07, + "loss": 0.6102, + "step": 1909 + }, + { + "epoch": 1.838306063522618, + "grad_norm": 1.1034223161041052, + "learning_rate": 1.6040994095681716e-07, + "loss": 0.594, + "step": 1910 + }, + { + "epoch": 1.8392685274302214, + "grad_norm": 1.0978617024425779, + "learning_rate": 1.5851610729273327e-07, + "loss": 0.5654, + "step": 1911 + }, + { + "epoch": 1.8402309913378248, + "grad_norm": 1.079095515818294, + "learning_rate": 1.5663333953696291e-07, + "loss": 0.5867, + "step": 1912 + }, + { + "epoch": 1.8411934552454283, + "grad_norm": 1.1594635255201562, + "learning_rate": 1.5476164199284328e-07, + "loss": 0.6736, + "step": 1913 + }, + { + "epoch": 1.842155919153032, + "grad_norm": 1.1182729829928115, + "learning_rate": 1.529010189384078e-07, + "loss": 0.614, + "step": 1914 + }, + { + "epoch": 1.8431183830606352, + "grad_norm": 1.1069389842238648, + "learning_rate": 1.5105147462637858e-07, + "loss": 0.6075, + "step": 1915 + }, + { + "epoch": 1.8440808469682386, + "grad_norm": 1.023983606169366, + "learning_rate": 1.4921301328415573e-07, + "loss": 0.5417, + "step": 1916 + }, + { + "epoch": 1.8450433108758422, + "grad_norm": 1.1010150131688898, + "learning_rate": 1.4738563911380566e-07, + "loss": 0.631, + "step": 1917 + }, + { + "epoch": 1.8460057747834457, + "grad_norm": 1.0839910552113592, + "learning_rate": 1.4556935629205738e-07, + "loss": 0.6124, + "step": 1918 + }, + { + "epoch": 1.846968238691049, + "grad_norm": 1.0958135024212745, + "learning_rate": 1.4376416897028457e-07, + "loss": 0.589, + "step": 1919 + }, + { + "epoch": 1.8479307025986524, + "grad_norm": 0.9434955697510244, + "learning_rate": 1.419700812745045e-07, + "loss": 0.4801, + "step": 1920 + }, + { + "epoch": 1.848893166506256, + "grad_norm": 1.076430585558076, + "learning_rate": 1.4018709730536095e-07, + "loss": 0.597, + "step": 1921 + }, + { + "epoch": 1.8498556304138596, + "grad_norm": 1.0596397827687025, + "learning_rate": 1.3841522113812235e-07, + "loss": 0.5439, + "step": 1922 + }, + { + "epoch": 1.850818094321463, + "grad_norm": 1.0144981458000042, + "learning_rate": 1.366544568226652e-07, + "loss": 0.5172, + "step": 1923 + }, + { + "epoch": 1.8517805582290663, + "grad_norm": 1.154734578727989, + "learning_rate": 1.349048083834703e-07, + "loss": 0.5734, + "step": 1924 + }, + { + "epoch": 1.8527430221366699, + "grad_norm": 1.0689841180251554, + "learning_rate": 1.3316627981961205e-07, + "loss": 0.5324, + "step": 1925 + }, + { + "epoch": 1.8537054860442734, + "grad_norm": 1.1432338179774626, + "learning_rate": 1.3143887510474628e-07, + "loss": 0.624, + "step": 1926 + }, + { + "epoch": 1.8546679499518768, + "grad_norm": 1.1303284470543782, + "learning_rate": 1.2972259818710587e-07, + "loss": 0.6418, + "step": 1927 + }, + { + "epoch": 1.8556304138594801, + "grad_norm": 0.7813615655337669, + "learning_rate": 1.2801745298948787e-07, + "loss": 0.3263, + "step": 1928 + }, + { + "epoch": 1.8565928777670837, + "grad_norm": 1.008997837054251, + "learning_rate": 1.2632344340924863e-07, + "loss": 0.5403, + "step": 1929 + }, + { + "epoch": 1.8575553416746873, + "grad_norm": 1.0725033107092583, + "learning_rate": 1.2464057331828928e-07, + "loss": 0.5749, + "step": 1930 + }, + { + "epoch": 1.8585178055822906, + "grad_norm": 1.057619392616763, + "learning_rate": 1.229688465630524e-07, + "loss": 0.554, + "step": 1931 + }, + { + "epoch": 1.859480269489894, + "grad_norm": 1.0601397761002194, + "learning_rate": 1.2130826696450993e-07, + "loss": 0.5789, + "step": 1932 + }, + { + "epoch": 1.8604427333974976, + "grad_norm": 9.300325473485534, + "learning_rate": 1.1965883831815573e-07, + "loss": 0.7204, + "step": 1933 + }, + { + "epoch": 1.8614051973051011, + "grad_norm": 0.9264113781050619, + "learning_rate": 1.1802056439399634e-07, + "loss": 0.4509, + "step": 1934 + }, + { + "epoch": 1.8623676612127045, + "grad_norm": 1.2009009130795048, + "learning_rate": 1.1639344893654258e-07, + "loss": 0.6698, + "step": 1935 + }, + { + "epoch": 1.8633301251203078, + "grad_norm": 1.165821008976603, + "learning_rate": 1.1477749566480123e-07, + "loss": 0.6928, + "step": 1936 + }, + { + "epoch": 1.8642925890279116, + "grad_norm": 1.0991236514731537, + "learning_rate": 1.1317270827226612e-07, + "loss": 0.6229, + "step": 1937 + }, + { + "epoch": 1.865255052935515, + "grad_norm": 2.877975152777155, + "learning_rate": 1.1157909042690928e-07, + "loss": 0.4583, + "step": 1938 + }, + { + "epoch": 1.8662175168431183, + "grad_norm": 1.0985629781556827, + "learning_rate": 1.099966457711743e-07, + "loss": 0.5903, + "step": 1939 + }, + { + "epoch": 1.867179980750722, + "grad_norm": 0.987934012412203, + "learning_rate": 1.0842537792196517e-07, + "loss": 0.5154, + "step": 1940 + }, + { + "epoch": 1.8681424446583255, + "grad_norm": 1.049603502552239, + "learning_rate": 1.0686529047064187e-07, + "loss": 0.5592, + "step": 1941 + }, + { + "epoch": 1.8691049085659288, + "grad_norm": 1.091457378158241, + "learning_rate": 1.0531638698300872e-07, + "loss": 0.5703, + "step": 1942 + }, + { + "epoch": 1.8700673724735322, + "grad_norm": 0.8212417642232763, + "learning_rate": 1.0377867099930716e-07, + "loss": 0.341, + "step": 1943 + }, + { + "epoch": 1.8710298363811357, + "grad_norm": 1.0133753878506293, + "learning_rate": 1.0225214603420852e-07, + "loss": 0.5363, + "step": 1944 + }, + { + "epoch": 1.8719923002887393, + "grad_norm": 1.1089809764285148, + "learning_rate": 1.0073681557680571e-07, + "loss": 0.6052, + "step": 1945 + }, + { + "epoch": 1.8729547641963427, + "grad_norm": 1.0928656204899123, + "learning_rate": 9.923268309060374e-08, + "loss": 0.5815, + "step": 1946 + }, + { + "epoch": 1.873917228103946, + "grad_norm": 1.1217807609233401, + "learning_rate": 9.773975201351538e-08, + "loss": 0.6705, + "step": 1947 + }, + { + "epoch": 1.8748796920115496, + "grad_norm": 1.2084155703888577, + "learning_rate": 9.625802575784771e-08, + "loss": 0.6643, + "step": 1948 + }, + { + "epoch": 1.8758421559191532, + "grad_norm": 0.9746096485825155, + "learning_rate": 9.478750771030055e-08, + "loss": 0.4679, + "step": 1949 + }, + { + "epoch": 1.8768046198267565, + "grad_norm": 1.0976745582403262, + "learning_rate": 9.332820123195419e-08, + "loss": 0.5989, + "step": 1950 + }, + { + "epoch": 1.8777670837343599, + "grad_norm": 1.129447147482023, + "learning_rate": 9.188010965826333e-08, + "loss": 0.6183, + "step": 1951 + }, + { + "epoch": 1.8787295476419634, + "grad_norm": 0.9849094454897154, + "learning_rate": 9.044323629904983e-08, + "loss": 0.5206, + "step": 1952 + }, + { + "epoch": 1.879692011549567, + "grad_norm": 0.9554432828714342, + "learning_rate": 8.901758443849551e-08, + "loss": 0.4707, + "step": 1953 + }, + { + "epoch": 1.8806544754571703, + "grad_norm": 5.104803163328381, + "learning_rate": 8.760315733513214e-08, + "loss": 0.466, + "step": 1954 + }, + { + "epoch": 1.8816169393647737, + "grad_norm": 1.177213833872404, + "learning_rate": 8.619995822183591e-08, + "loss": 0.6619, + "step": 1955 + }, + { + "epoch": 1.8825794032723773, + "grad_norm": 16.9741052515645, + "learning_rate": 8.480799030582187e-08, + "loss": 0.5268, + "step": 1956 + }, + { + "epoch": 1.8835418671799808, + "grad_norm": 1.1338667689897608, + "learning_rate": 8.342725676863117e-08, + "loss": 0.6215, + "step": 1957 + }, + { + "epoch": 1.8845043310875842, + "grad_norm": 1.041371136669472, + "learning_rate": 8.205776076613048e-08, + "loss": 0.5645, + "step": 1958 + }, + { + "epoch": 1.8854667949951875, + "grad_norm": 1.104440568647923, + "learning_rate": 8.06995054284998e-08, + "loss": 0.5777, + "step": 1959 + }, + { + "epoch": 1.8864292589027911, + "grad_norm": 1.1538485787772546, + "learning_rate": 7.935249386022637e-08, + "loss": 0.6673, + "step": 1960 + }, + { + "epoch": 1.8873917228103947, + "grad_norm": 1.1451297858743938, + "learning_rate": 7.80167291400996e-08, + "loss": 0.611, + "step": 1961 + }, + { + "epoch": 1.888354186717998, + "grad_norm": 1.1777470574582576, + "learning_rate": 7.669221432120288e-08, + "loss": 0.6923, + "step": 1962 + }, + { + "epoch": 1.8893166506256014, + "grad_norm": 1.1542800860864781, + "learning_rate": 7.537895243090398e-08, + "loss": 0.6643, + "step": 1963 + }, + { + "epoch": 1.890279114533205, + "grad_norm": 1.110477964812046, + "learning_rate": 7.40769464708535e-08, + "loss": 0.5954, + "step": 1964 + }, + { + "epoch": 1.8912415784408085, + "grad_norm": 1.0896062821127508, + "learning_rate": 7.278619941697151e-08, + "loss": 0.5634, + "step": 1965 + }, + { + "epoch": 1.8922040423484119, + "grad_norm": 0.971384858855514, + "learning_rate": 7.15067142194481e-08, + "loss": 0.4704, + "step": 1966 + }, + { + "epoch": 1.8931665062560152, + "grad_norm": 1.1205535092161276, + "learning_rate": 7.023849380272951e-08, + "loss": 0.5874, + "step": 1967 + }, + { + "epoch": 1.894128970163619, + "grad_norm": 1.1138816415294943, + "learning_rate": 6.898154106551702e-08, + "loss": 0.5987, + "step": 1968 + }, + { + "epoch": 1.8950914340712224, + "grad_norm": 1.0955951517573894, + "learning_rate": 6.77358588807564e-08, + "loss": 0.6048, + "step": 1969 + }, + { + "epoch": 1.8960538979788257, + "grad_norm": 1.0700444136398994, + "learning_rate": 6.650145009563347e-08, + "loss": 0.5565, + "step": 1970 + }, + { + "epoch": 1.8970163618864293, + "grad_norm": 1.200450258765386, + "learning_rate": 6.527831753156799e-08, + "loss": 0.6806, + "step": 1971 + }, + { + "epoch": 1.8979788257940329, + "grad_norm": 1.1630826308217448, + "learning_rate": 6.40664639842048e-08, + "loss": 0.5997, + "step": 1972 + }, + { + "epoch": 1.8989412897016362, + "grad_norm": 0.9772536821871577, + "learning_rate": 6.286589222340988e-08, + "loss": 0.486, + "step": 1973 + }, + { + "epoch": 1.8999037536092396, + "grad_norm": 1.0268396680215637, + "learning_rate": 6.167660499326322e-08, + "loss": 0.542, + "step": 1974 + }, + { + "epoch": 1.9008662175168431, + "grad_norm": 1.0115253597825027, + "learning_rate": 6.049860501205096e-08, + "loss": 0.5121, + "step": 1975 + }, + { + "epoch": 1.9018286814244467, + "grad_norm": 1.0193333339836759, + "learning_rate": 5.933189497226211e-08, + "loss": 0.4717, + "step": 1976 + }, + { + "epoch": 1.90279114533205, + "grad_norm": 1.0032808061515854, + "learning_rate": 5.8176477540580756e-08, + "loss": 0.4894, + "step": 1977 + }, + { + "epoch": 1.9037536092396534, + "grad_norm": 1.1469721982188856, + "learning_rate": 5.703235535787888e-08, + "loss": 0.6649, + "step": 1978 + }, + { + "epoch": 1.904716073147257, + "grad_norm": 0.8378595722467119, + "learning_rate": 5.5899531039213526e-08, + "loss": 0.3816, + "step": 1979 + }, + { + "epoch": 1.9056785370548606, + "grad_norm": 1.1201160267913006, + "learning_rate": 5.477800717381687e-08, + "loss": 0.6405, + "step": 1980 + }, + { + "epoch": 1.906641000962464, + "grad_norm": 1.1167661283614456, + "learning_rate": 5.366778632509395e-08, + "loss": 0.612, + "step": 1981 + }, + { + "epoch": 1.9076034648700673, + "grad_norm": 1.1450929449182214, + "learning_rate": 5.2568871030613254e-08, + "loss": 0.6481, + "step": 1982 + }, + { + "epoch": 1.9085659287776708, + "grad_norm": 1.0554663409356777, + "learning_rate": 5.14812638021045e-08, + "loss": 0.5711, + "step": 1983 + }, + { + "epoch": 1.9095283926852744, + "grad_norm": 1.0465375830187722, + "learning_rate": 5.040496712545084e-08, + "loss": 0.589, + "step": 1984 + }, + { + "epoch": 1.9104908565928778, + "grad_norm": 0.973970455307527, + "learning_rate": 4.9339983460682805e-08, + "loss": 0.4727, + "step": 1985 + }, + { + "epoch": 1.9114533205004811, + "grad_norm": 1.0915554406507717, + "learning_rate": 4.828631524197325e-08, + "loss": 0.6259, + "step": 1986 + }, + { + "epoch": 1.9124157844080847, + "grad_norm": 0.9128606000927172, + "learning_rate": 4.7243964877634055e-08, + "loss": 0.4801, + "step": 1987 + }, + { + "epoch": 1.9133782483156883, + "grad_norm": 1.0808915454811043, + "learning_rate": 4.621293475010613e-08, + "loss": 0.6086, + "step": 1988 + }, + { + "epoch": 1.9143407122232916, + "grad_norm": 1.03677951994771, + "learning_rate": 4.519322721595665e-08, + "loss": 0.5314, + "step": 1989 + }, + { + "epoch": 1.915303176130895, + "grad_norm": 0.9091410767453204, + "learning_rate": 4.4184844605875135e-08, + "loss": 0.4395, + "step": 1990 + }, + { + "epoch": 1.9162656400384985, + "grad_norm": 1.082183884062068, + "learning_rate": 4.31877892246646e-08, + "loss": 0.6143, + "step": 1991 + }, + { + "epoch": 1.917228103946102, + "grad_norm": 1.0169938746284928, + "learning_rate": 4.2202063351238775e-08, + "loss": 0.5251, + "step": 1992 + }, + { + "epoch": 1.9181905678537055, + "grad_norm": 1.23030946766217, + "learning_rate": 4.1227669238615985e-08, + "loss": 0.695, + "step": 1993 + }, + { + "epoch": 1.9191530317613088, + "grad_norm": 1.1226467833773246, + "learning_rate": 4.026460911391361e-08, + "loss": 0.6126, + "step": 1994 + }, + { + "epoch": 1.9201154956689124, + "grad_norm": 1.2539274363766513, + "learning_rate": 3.9312885178345304e-08, + "loss": 0.7077, + "step": 1995 + }, + { + "epoch": 1.921077959576516, + "grad_norm": 1.0105893907671197, + "learning_rate": 3.837249960721268e-08, + "loss": 0.5177, + "step": 1996 + }, + { + "epoch": 1.9220404234841193, + "grad_norm": 1.189392634131118, + "learning_rate": 3.744345454990195e-08, + "loss": 0.6981, + "step": 1997 + }, + { + "epoch": 1.9230028873917226, + "grad_norm": 1.0891283811380734, + "learning_rate": 3.652575212987952e-08, + "loss": 0.6431, + "step": 1998 + }, + { + "epoch": 1.9239653512993262, + "grad_norm": 1.0298975335708718, + "learning_rate": 3.561939444468643e-08, + "loss": 0.5596, + "step": 1999 + }, + { + "epoch": 1.9249278152069298, + "grad_norm": 1.2083619343338015, + "learning_rate": 3.472438356593388e-08, + "loss": 0.6517, + "step": 2000 + }, + { + "epoch": 1.9249278152069298, + "eval_loss": 0.7821268439292908, + "eval_runtime": 12.5299, + "eval_samples_per_second": 3.352, + "eval_steps_per_second": 0.878, + "step": 2000 + }, + { + "epoch": 1.9258902791145331, + "grad_norm": 1.1463247180045484, + "learning_rate": 3.384072153929774e-08, + "loss": 0.6774, + "step": 2001 + }, + { + "epoch": 1.9268527430221367, + "grad_norm": 0.9738469902700995, + "learning_rate": 3.2968410384514596e-08, + "loss": 0.4776, + "step": 2002 + }, + { + "epoch": 1.9278152069297403, + "grad_norm": 1.0429511258340376, + "learning_rate": 3.210745209537791e-08, + "loss": 0.5792, + "step": 2003 + }, + { + "epoch": 1.9287776708373436, + "grad_norm": 1.094650835172697, + "learning_rate": 3.1257848639730804e-08, + "loss": 0.5901, + "step": 2004 + }, + { + "epoch": 1.929740134744947, + "grad_norm": 1.0370931119909503, + "learning_rate": 3.0419601959464917e-08, + "loss": 0.5473, + "step": 2005 + }, + { + "epoch": 1.9307025986525506, + "grad_norm": 1.14247233015092, + "learning_rate": 2.9592713970513222e-08, + "loss": 0.6417, + "step": 2006 + }, + { + "epoch": 1.9316650625601541, + "grad_norm": 1.1138552158429296, + "learning_rate": 2.8777186562847782e-08, + "loss": 0.6229, + "step": 2007 + }, + { + "epoch": 1.9326275264677575, + "grad_norm": 1.0855296295714827, + "learning_rate": 2.7973021600473106e-08, + "loss": 0.594, + "step": 2008 + }, + { + "epoch": 1.9335899903753608, + "grad_norm": 1.042071633134226, + "learning_rate": 2.7180220921425025e-08, + "loss": 0.5208, + "step": 2009 + }, + { + "epoch": 1.9345524542829644, + "grad_norm": 0.9895143849756135, + "learning_rate": 2.639878633776294e-08, + "loss": 0.4911, + "step": 2010 + }, + { + "epoch": 1.935514918190568, + "grad_norm": 1.040159976538284, + "learning_rate": 2.5628719635568122e-08, + "loss": 0.5301, + "step": 2011 + }, + { + "epoch": 1.9364773820981713, + "grad_norm": 1.1409854934798294, + "learning_rate": 2.4870022574938758e-08, + "loss": 0.6578, + "step": 2012 + }, + { + "epoch": 1.9374398460057747, + "grad_norm": 1.059386252722632, + "learning_rate": 2.4122696889986586e-08, + "loss": 0.5523, + "step": 2013 + }, + { + "epoch": 1.9384023099133783, + "grad_norm": 1.1546246816908556, + "learning_rate": 2.338674428883192e-08, + "loss": 0.6537, + "step": 2014 + }, + { + "epoch": 1.9393647738209818, + "grad_norm": 2.8453433763431484, + "learning_rate": 2.2662166453599753e-08, + "loss": 0.4388, + "step": 2015 + }, + { + "epoch": 1.9403272377285852, + "grad_norm": 1.1705893477518468, + "learning_rate": 2.1948965040417546e-08, + "loss": 0.6601, + "step": 2016 + }, + { + "epoch": 1.9412897016361885, + "grad_norm": 0.9698921565351144, + "learning_rate": 2.1247141679409667e-08, + "loss": 0.4795, + "step": 2017 + }, + { + "epoch": 1.942252165543792, + "grad_norm": 1.593239268404914, + "learning_rate": 2.0556697974695172e-08, + "loss": 0.5554, + "step": 2018 + }, + { + "epoch": 1.9432146294513957, + "grad_norm": 0.9653415303825732, + "learning_rate": 1.9877635504381155e-08, + "loss": 0.4655, + "step": 2019 + }, + { + "epoch": 1.944177093358999, + "grad_norm": 1.1323478784639953, + "learning_rate": 1.9209955820563287e-08, + "loss": 0.6393, + "step": 2020 + }, + { + "epoch": 1.9451395572666024, + "grad_norm": 1.1484761220350397, + "learning_rate": 1.8553660449320276e-08, + "loss": 0.6794, + "step": 2021 + }, + { + "epoch": 1.946102021174206, + "grad_norm": 1.0752054891662248, + "learning_rate": 1.790875089070887e-08, + "loss": 0.575, + "step": 2022 + }, + { + "epoch": 1.9470644850818095, + "grad_norm": 1.0391896619868095, + "learning_rate": 1.7275228618762184e-08, + "loss": 0.5605, + "step": 2023 + }, + { + "epoch": 1.9480269489894129, + "grad_norm": 1.0767748639073598, + "learning_rate": 1.6653095081488048e-08, + "loss": 0.5816, + "step": 2024 + }, + { + "epoch": 1.9489894128970162, + "grad_norm": 1.018071828474781, + "learning_rate": 1.604235170086066e-08, + "loss": 0.5309, + "step": 2025 + }, + { + "epoch": 1.9499518768046198, + "grad_norm": 1.1470093815256814, + "learning_rate": 1.544299987282283e-08, + "loss": 0.6744, + "step": 2026 + }, + { + "epoch": 1.9509143407122234, + "grad_norm": 1.1216190223340368, + "learning_rate": 1.4855040967279299e-08, + "loss": 0.6233, + "step": 2027 + }, + { + "epoch": 1.9518768046198267, + "grad_norm": 1.0960669996704429, + "learning_rate": 1.4278476328095092e-08, + "loss": 0.5381, + "step": 2028 + }, + { + "epoch": 1.95283926852743, + "grad_norm": 0.9416118576764811, + "learning_rate": 1.3713307273091059e-08, + "loss": 0.4748, + "step": 2029 + }, + { + "epoch": 1.9538017324350336, + "grad_norm": 0.9738147761494587, + "learning_rate": 1.3159535094042775e-08, + "loss": 0.4747, + "step": 2030 + }, + { + "epoch": 1.9547641963426372, + "grad_norm": 1.0453212061330224, + "learning_rate": 1.2617161056675542e-08, + "loss": 0.5167, + "step": 2031 + }, + { + "epoch": 1.9557266602502406, + "grad_norm": 0.9596857394008487, + "learning_rate": 1.208618640066439e-08, + "loss": 0.4663, + "step": 2032 + }, + { + "epoch": 1.9566891241578441, + "grad_norm": 1.1433137521942323, + "learning_rate": 1.1566612339627969e-08, + "loss": 0.6742, + "step": 2033 + }, + { + "epoch": 1.9576515880654477, + "grad_norm": 1.0753102966335442, + "learning_rate": 1.105844006112744e-08, + "loss": 0.5862, + "step": 2034 + }, + { + "epoch": 1.958614051973051, + "grad_norm": 1.0586487279818708, + "learning_rate": 1.0561670726663697e-08, + "loss": 0.5874, + "step": 2035 + }, + { + "epoch": 1.9595765158806544, + "grad_norm": 1.0784755904055383, + "learning_rate": 1.0076305471674042e-08, + "loss": 0.5657, + "step": 2036 + }, + { + "epoch": 1.960538979788258, + "grad_norm": 0.9372493454208994, + "learning_rate": 9.602345405531621e-09, + "loss": 0.4591, + "step": 2037 + }, + { + "epoch": 1.9615014436958615, + "grad_norm": 0.9683734397545739, + "learning_rate": 9.139791611540438e-09, + "loss": 0.5144, + "step": 2038 + }, + { + "epoch": 1.962463907603465, + "grad_norm": 1.1689412937304435, + "learning_rate": 8.688645146932572e-09, + "loss": 0.6585, + "step": 2039 + }, + { + "epoch": 1.9634263715110682, + "grad_norm": 1.1411586239429528, + "learning_rate": 8.248907042868736e-09, + "loss": 0.6433, + "step": 2040 + }, + { + "epoch": 1.9643888354186718, + "grad_norm": 1.0781636279742262, + "learning_rate": 7.82057830443439e-09, + "loss": 0.5951, + "step": 2041 + }, + { + "epoch": 1.9653512993262754, + "grad_norm": 1.0133079686883997, + "learning_rate": 7.403659910635852e-09, + "loss": 0.488, + "step": 2042 + }, + { + "epoch": 1.9663137632338787, + "grad_norm": 1.0895582959662615, + "learning_rate": 6.99815281439975e-09, + "loss": 0.5249, + "step": 2043 + }, + { + "epoch": 1.967276227141482, + "grad_norm": 0.916646167072474, + "learning_rate": 6.604057942571907e-09, + "loss": 0.4344, + "step": 2044 + }, + { + "epoch": 1.9682386910490857, + "grad_norm": 1.0958429813197528, + "learning_rate": 6.221376195911788e-09, + "loss": 0.5805, + "step": 2045 + }, + { + "epoch": 1.9692011549566892, + "grad_norm": 1.1273803950101604, + "learning_rate": 5.850108449094727e-09, + "loss": 0.6662, + "step": 2046 + }, + { + "epoch": 1.9701636188642926, + "grad_norm": 0.959688660168769, + "learning_rate": 5.4902555507058144e-09, + "loss": 0.4502, + "step": 2047 + }, + { + "epoch": 1.971126082771896, + "grad_norm": 1.0106722077715165, + "learning_rate": 5.141818323240455e-09, + "loss": 0.5133, + "step": 2048 + }, + { + "epoch": 1.9720885466794995, + "grad_norm": 1.038961320662795, + "learning_rate": 4.804797563103259e-09, + "loss": 0.5423, + "step": 2049 + }, + { + "epoch": 1.973051010587103, + "grad_norm": 1.0233495771947125, + "learning_rate": 4.479194040602486e-09, + "loss": 0.5886, + "step": 2050 + }, + { + "epoch": 1.9740134744947064, + "grad_norm": 1.149187977247883, + "learning_rate": 4.165008499952272e-09, + "loss": 0.6314, + "step": 2051 + }, + { + "epoch": 1.9749759384023098, + "grad_norm": 1.1079270906813738, + "learning_rate": 3.862241659269294e-09, + "loss": 0.6116, + "step": 2052 + }, + { + "epoch": 1.9759384023099134, + "grad_norm": 0.9971315615512484, + "learning_rate": 3.570894210569997e-09, + "loss": 0.4878, + "step": 2053 + }, + { + "epoch": 1.976900866217517, + "grad_norm": 1.9337058828697369, + "learning_rate": 3.2909668197717016e-09, + "loss": 0.5884, + "step": 2054 + }, + { + "epoch": 1.9778633301251203, + "grad_norm": 0.991210754597578, + "learning_rate": 3.0224601266876097e-09, + "loss": 0.5466, + "step": 2055 + }, + { + "epoch": 1.9788257940327236, + "grad_norm": 1.1595718072921928, + "learning_rate": 2.765374745029581e-09, + "loss": 0.6361, + "step": 2056 + }, + { + "epoch": 1.9797882579403272, + "grad_norm": 0.9899135792769228, + "learning_rate": 2.519711262403135e-09, + "loss": 0.5022, + "step": 2057 + }, + { + "epoch": 1.9807507218479308, + "grad_norm": 1.0600315441515689, + "learning_rate": 2.2854702403068975e-09, + "loss": 0.5633, + "step": 2058 + }, + { + "epoch": 1.9817131857555341, + "grad_norm": 1.0669785663094906, + "learning_rate": 2.062652214133154e-09, + "loss": 0.5976, + "step": 2059 + }, + { + "epoch": 1.9826756496631375, + "grad_norm": 1.566722707248118, + "learning_rate": 1.8512576931634107e-09, + "loss": 0.6162, + "step": 2060 + }, + { + "epoch": 1.983638113570741, + "grad_norm": 0.9848495346905081, + "learning_rate": 1.651287160571724e-09, + "loss": 0.5148, + "step": 2061 + }, + { + "epoch": 1.9846005774783446, + "grad_norm": 4.938572347456576, + "learning_rate": 1.4627410734185942e-09, + "loss": 0.5522, + "step": 2062 + }, + { + "epoch": 1.985563041385948, + "grad_norm": 1.047504472763117, + "learning_rate": 1.285619862653187e-09, + "loss": 0.5298, + "step": 2063 + }, + { + "epoch": 1.9865255052935515, + "grad_norm": 0.8921694638675606, + "learning_rate": 1.1199239331122213e-09, + "loss": 0.4221, + "step": 2064 + }, + { + "epoch": 1.9874879692011551, + "grad_norm": 0.9114459679909427, + "learning_rate": 9.656536635171964e-10, + "loss": 0.4315, + "step": 2065 + }, + { + "epoch": 1.9884504331087585, + "grad_norm": 1.1450130796898275, + "learning_rate": 8.228094064749448e-10, + "loss": 0.6241, + "step": 2066 + }, + { + "epoch": 1.9894128970163618, + "grad_norm": 1.161041155479135, + "learning_rate": 6.913914884765227e-10, + "loss": 0.6614, + "step": 2067 + }, + { + "epoch": 1.9903753609239654, + "grad_norm": 0.958827066921697, + "learning_rate": 5.714002098966554e-10, + "loss": 0.5107, + "step": 2068 + }, + { + "epoch": 1.991337824831569, + "grad_norm": 1.1615158147399, + "learning_rate": 4.6283584499262623e-10, + "loss": 0.6529, + "step": 2069 + }, + { + "epoch": 1.9923002887391723, + "grad_norm": 1.0678960932116024, + "learning_rate": 3.656986419042774e-10, + "loss": 0.5613, + "step": 2070 + }, + { + "epoch": 1.9932627526467757, + "grad_norm": 1.0372431418240753, + "learning_rate": 2.7998882265234393e-10, + "loss": 0.5224, + "step": 2071 + }, + { + "epoch": 1.9942252165543792, + "grad_norm": 1.088888802089693, + "learning_rate": 2.0570658313845414e-10, + "loss": 0.5907, + "step": 2072 + }, + { + "epoch": 1.9951876804619828, + "grad_norm": 1.0974927378319705, + "learning_rate": 1.428520931462396e-10, + "loss": 0.6154, + "step": 2073 + }, + { + "epoch": 1.9961501443695862, + "grad_norm": 1.0747871278083982, + "learning_rate": 9.142549633800457e-11, + "loss": 0.5372, + "step": 2074 + }, + { + "epoch": 1.9971126082771895, + "grad_norm": 1.0753061572747393, + "learning_rate": 5.1426910257501574e-11, + "loss": 0.5875, + "step": 2075 + }, + { + "epoch": 1.998075072184793, + "grad_norm": 0.9965369092232225, + "learning_rate": 2.2856426326045566e-11, + "loss": 0.4877, + "step": 2076 + }, + { + "epoch": 1.9990375360923966, + "grad_norm": 0.9921974064481799, + "learning_rate": 5.714109846399751e-12, + "loss": 0.4922, + "step": 2077 + }, + { + "epoch": 2.0, + "grad_norm": 0.9310805535896068, + "learning_rate": 0.0, + "loss": 0.4402, + "step": 2078 + }, + { + "epoch": 2.0, + "step": 2078, + "total_flos": 217757908992000.0, + "train_loss": 0.6756772079412701, + "train_runtime": 15285.4581, + "train_samples_per_second": 1.087, + "train_steps_per_second": 0.136 + } + ], + "logging_steps": 1, + "max_steps": 2078, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 300, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 217757908992000.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}