| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.99328, | |
| "eval_steps": 400, | |
| "global_step": 1170, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00512, | |
| "grad_norm": 0.033956460654735565, | |
| "learning_rate": 0.001998289136013687, | |
| "loss": 1.5175, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01024, | |
| "grad_norm": 0.11115246266126633, | |
| "learning_rate": 0.0019948674080410606, | |
| "loss": 1.5456, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01536, | |
| "grad_norm": 0.31134533882141113, | |
| "learning_rate": 0.001991445680068435, | |
| "loss": 1.4942, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02048, | |
| "grad_norm": 0.314473956823349, | |
| "learning_rate": 0.0019880239520958082, | |
| "loss": 1.4622, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 0.3232758641242981, | |
| "learning_rate": 0.001984602224123182, | |
| "loss": 1.4351, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03072, | |
| "grad_norm": 0.30670273303985596, | |
| "learning_rate": 0.0019811804961505563, | |
| "loss": 1.3924, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03584, | |
| "grad_norm": 0.2764911651611328, | |
| "learning_rate": 0.0019777587681779297, | |
| "loss": 1.3622, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04096, | |
| "grad_norm": 0.24471326172351837, | |
| "learning_rate": 0.001974337040205304, | |
| "loss": 1.386, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04608, | |
| "grad_norm": 0.21196012198925018, | |
| "learning_rate": 0.0019709153122326774, | |
| "loss": 1.3682, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 0.23733310401439667, | |
| "learning_rate": 0.0019674935842600516, | |
| "loss": 1.3512, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05632, | |
| "grad_norm": 0.16206419467926025, | |
| "learning_rate": 0.001964071856287425, | |
| "loss": 1.3428, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06144, | |
| "grad_norm": 0.2223929464817047, | |
| "learning_rate": 0.001960650128314799, | |
| "loss": 1.3706, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06656, | |
| "grad_norm": 0.1612645536661148, | |
| "learning_rate": 0.0019572284003421727, | |
| "loss": 1.3775, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07168, | |
| "grad_norm": 0.16616274416446686, | |
| "learning_rate": 0.0019538066723695465, | |
| "loss": 1.3572, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 0.15922699868679047, | |
| "learning_rate": 0.0019503849443969204, | |
| "loss": 1.3482, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08192, | |
| "grad_norm": 0.14939665794372559, | |
| "learning_rate": 0.0019469632164242944, | |
| "loss": 1.3312, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08704, | |
| "grad_norm": 0.16849732398986816, | |
| "learning_rate": 0.0019435414884516682, | |
| "loss": 1.3394, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09216, | |
| "grad_norm": 0.1453033685684204, | |
| "learning_rate": 0.0019401197604790419, | |
| "loss": 1.2957, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09728, | |
| "grad_norm": 0.14633750915527344, | |
| "learning_rate": 0.001936698032506416, | |
| "loss": 1.334, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 0.12038926780223846, | |
| "learning_rate": 0.0019332763045337895, | |
| "loss": 1.3257, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10752, | |
| "grad_norm": 0.1144009605050087, | |
| "learning_rate": 0.0019298545765611636, | |
| "loss": 1.337, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11264, | |
| "grad_norm": 0.09070790559053421, | |
| "learning_rate": 0.0019264328485885372, | |
| "loss": 1.2969, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11776, | |
| "grad_norm": 0.09154277294874191, | |
| "learning_rate": 0.001923011120615911, | |
| "loss": 1.3233, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.12288, | |
| "grad_norm": 0.09742377698421478, | |
| "learning_rate": 0.0019195893926432848, | |
| "loss": 1.3272, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.10923709720373154, | |
| "learning_rate": 0.0019161676646706587, | |
| "loss": 1.3353, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13312, | |
| "grad_norm": 0.10012141615152359, | |
| "learning_rate": 0.0019127459366980327, | |
| "loss": 1.3011, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.13824, | |
| "grad_norm": 0.10850805044174194, | |
| "learning_rate": 0.0019093242087254063, | |
| "loss": 1.3372, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.14336, | |
| "grad_norm": 0.11083640158176422, | |
| "learning_rate": 0.0019059024807527804, | |
| "loss": 1.3098, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.14848, | |
| "grad_norm": 0.11255177110433578, | |
| "learning_rate": 0.001902480752780154, | |
| "loss": 1.3223, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 0.09206261485815048, | |
| "learning_rate": 0.0018990590248075278, | |
| "loss": 1.2899, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15872, | |
| "grad_norm": 0.09589900076389313, | |
| "learning_rate": 0.0018956372968349016, | |
| "loss": 1.3026, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.16384, | |
| "grad_norm": 0.0895056203007698, | |
| "learning_rate": 0.0018922155688622755, | |
| "loss": 1.2996, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.16896, | |
| "grad_norm": 0.11926258355379105, | |
| "learning_rate": 0.0018887938408896493, | |
| "loss": 1.2952, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.17408, | |
| "grad_norm": 0.11748767644166946, | |
| "learning_rate": 0.0018853721129170231, | |
| "loss": 1.3085, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 0.09052393585443497, | |
| "learning_rate": 0.001881950384944397, | |
| "loss": 1.3111, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18432, | |
| "grad_norm": 0.11488136649131775, | |
| "learning_rate": 0.0018785286569717708, | |
| "loss": 1.2884, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.18944, | |
| "grad_norm": 0.10231968015432358, | |
| "learning_rate": 0.0018751069289991446, | |
| "loss": 1.2991, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.19456, | |
| "grad_norm": 0.17033444344997406, | |
| "learning_rate": 0.0018716852010265185, | |
| "loss": 1.3188, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.19968, | |
| "grad_norm": 0.1066645011305809, | |
| "learning_rate": 0.0018682634730538923, | |
| "loss": 1.2946, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 0.1261938363313675, | |
| "learning_rate": 0.0018648417450812661, | |
| "loss": 1.316, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.20992, | |
| "grad_norm": 0.09089711308479309, | |
| "learning_rate": 0.00186142001710864, | |
| "loss": 1.2927, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.21504, | |
| "grad_norm": 0.16600407660007477, | |
| "learning_rate": 0.0018579982891360136, | |
| "loss": 1.2806, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.22016, | |
| "grad_norm": 0.10440811514854431, | |
| "learning_rate": 0.0018545765611633876, | |
| "loss": 1.2928, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.22528, | |
| "grad_norm": 0.11303785443305969, | |
| "learning_rate": 0.0018511548331907612, | |
| "loss": 1.2999, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.1060706302523613, | |
| "learning_rate": 0.0018477331052181353, | |
| "loss": 1.2887, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.23552, | |
| "grad_norm": 0.11111288517713547, | |
| "learning_rate": 0.001844311377245509, | |
| "loss": 1.2935, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.24064, | |
| "grad_norm": 0.10950679332017899, | |
| "learning_rate": 0.001840889649272883, | |
| "loss": 1.2865, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.24576, | |
| "grad_norm": 0.09789486229419708, | |
| "learning_rate": 0.0018374679213002568, | |
| "loss": 1.3088, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.25088, | |
| "grad_norm": 0.10912415385246277, | |
| "learning_rate": 0.0018340461933276304, | |
| "loss": 1.2811, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.11448093503713608, | |
| "learning_rate": 0.0018306244653550044, | |
| "loss": 1.2707, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26112, | |
| "grad_norm": 0.10610745847225189, | |
| "learning_rate": 0.001827202737382378, | |
| "loss": 1.2994, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.26624, | |
| "grad_norm": 0.10192134976387024, | |
| "learning_rate": 0.001823781009409752, | |
| "loss": 1.2807, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.27136, | |
| "grad_norm": 0.10114825516939163, | |
| "learning_rate": 0.0018203592814371257, | |
| "loss": 1.2582, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.27648, | |
| "grad_norm": 0.09611225128173828, | |
| "learning_rate": 0.0018169375534644997, | |
| "loss": 1.2951, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.08788544684648514, | |
| "learning_rate": 0.0018135158254918733, | |
| "loss": 1.2869, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.28672, | |
| "grad_norm": 0.10186023265123367, | |
| "learning_rate": 0.0018100940975192472, | |
| "loss": 1.2916, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.29184, | |
| "grad_norm": 0.11426091939210892, | |
| "learning_rate": 0.0018066723695466212, | |
| "loss": 1.2705, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.29696, | |
| "grad_norm": 0.0911969244480133, | |
| "learning_rate": 0.0018032506415739948, | |
| "loss": 1.2826, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.30208, | |
| "grad_norm": 0.08570262044668198, | |
| "learning_rate": 0.0017998289136013689, | |
| "loss": 1.2783, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.10600815713405609, | |
| "learning_rate": 0.0017964071856287425, | |
| "loss": 1.2633, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.31232, | |
| "grad_norm": 0.10680408775806427, | |
| "learning_rate": 0.0017929854576561163, | |
| "loss": 1.2751, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.31744, | |
| "grad_norm": 0.10642975568771362, | |
| "learning_rate": 0.0017895637296834902, | |
| "loss": 1.2776, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.32256, | |
| "grad_norm": 0.09877141565084457, | |
| "learning_rate": 0.001786142001710864, | |
| "loss": 1.2868, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.32768, | |
| "grad_norm": 0.12077789753675461, | |
| "learning_rate": 0.0017827202737382378, | |
| "loss": 1.2825, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.1019749641418457, | |
| "learning_rate": 0.0017792985457656116, | |
| "loss": 1.2689, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.33792, | |
| "grad_norm": 0.08969077467918396, | |
| "learning_rate": 0.0017758768177929857, | |
| "loss": 1.2795, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.34304, | |
| "grad_norm": 0.10050085186958313, | |
| "learning_rate": 0.0017724550898203593, | |
| "loss": 1.2748, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.34816, | |
| "grad_norm": 0.08829426020383835, | |
| "learning_rate": 0.0017690333618477331, | |
| "loss": 1.2577, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.35328, | |
| "grad_norm": 0.09129626303911209, | |
| "learning_rate": 0.001765611633875107, | |
| "loss": 1.2703, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.10055152326822281, | |
| "learning_rate": 0.0017621899059024808, | |
| "loss": 1.2618, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.36352, | |
| "grad_norm": 0.09721378982067108, | |
| "learning_rate": 0.0017587681779298546, | |
| "loss": 1.2809, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.36864, | |
| "grad_norm": 0.09917334467172623, | |
| "learning_rate": 0.0017553464499572285, | |
| "loss": 1.2628, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.37376, | |
| "grad_norm": 0.1129026785492897, | |
| "learning_rate": 0.0017519247219846023, | |
| "loss": 1.2684, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.37888, | |
| "grad_norm": 0.08873754739761353, | |
| "learning_rate": 0.0017485029940119761, | |
| "loss": 1.2726, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.10734206438064575, | |
| "learning_rate": 0.0017450812660393497, | |
| "loss": 1.2545, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.38912, | |
| "grad_norm": 0.09691416472196579, | |
| "learning_rate": 0.0017416595380667238, | |
| "loss": 1.2511, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.39424, | |
| "grad_norm": 0.10880187153816223, | |
| "learning_rate": 0.0017382378100940976, | |
| "loss": 1.2656, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.39936, | |
| "grad_norm": 0.10392981767654419, | |
| "learning_rate": 0.0017348160821214714, | |
| "loss": 1.2755, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.40448, | |
| "grad_norm": 0.08561566472053528, | |
| "learning_rate": 0.0017313943541488453, | |
| "loss": 1.2956, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.09822013229131699, | |
| "learning_rate": 0.0017279726261762189, | |
| "loss": 1.2887, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.41472, | |
| "grad_norm": 0.09553670883178711, | |
| "learning_rate": 0.001724550898203593, | |
| "loss": 1.2743, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.41984, | |
| "grad_norm": 0.10794595628976822, | |
| "learning_rate": 0.0017211291702309665, | |
| "loss": 1.2603, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.42496, | |
| "grad_norm": 0.09373841434717178, | |
| "learning_rate": 0.0017177074422583406, | |
| "loss": 1.2441, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.43008, | |
| "grad_norm": 0.10550329089164734, | |
| "learning_rate": 0.0017142857142857142, | |
| "loss": 1.2714, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.08271865546703339, | |
| "learning_rate": 0.0017108639863130882, | |
| "loss": 1.2671, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.44032, | |
| "grad_norm": 0.09635920822620392, | |
| "learning_rate": 0.001707442258340462, | |
| "loss": 1.2861, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.44544, | |
| "grad_norm": 0.0995635837316513, | |
| "learning_rate": 0.0017040205303678357, | |
| "loss": 1.2803, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.45056, | |
| "grad_norm": 0.10621396452188492, | |
| "learning_rate": 0.0017005988023952097, | |
| "loss": 1.2685, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.45568, | |
| "grad_norm": 0.10926424711942673, | |
| "learning_rate": 0.0016971770744225833, | |
| "loss": 1.2551, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.09723444283008575, | |
| "learning_rate": 0.0016937553464499574, | |
| "loss": 1.2547, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.46592, | |
| "grad_norm": 0.09107507020235062, | |
| "learning_rate": 0.001690333618477331, | |
| "loss": 1.2664, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.47104, | |
| "grad_norm": 0.08052769303321838, | |
| "learning_rate": 0.001686911890504705, | |
| "loss": 1.2645, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.47616, | |
| "grad_norm": 0.0871344730257988, | |
| "learning_rate": 0.0016834901625320787, | |
| "loss": 1.2694, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.48128, | |
| "grad_norm": 0.11148510873317719, | |
| "learning_rate": 0.0016800684345594525, | |
| "loss": 1.2697, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.08355887234210968, | |
| "learning_rate": 0.0016766467065868263, | |
| "loss": 1.2458, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.49152, | |
| "grad_norm": 0.08468321710824966, | |
| "learning_rate": 0.0016732249786142002, | |
| "loss": 1.252, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.49664, | |
| "grad_norm": 0.08812184631824493, | |
| "learning_rate": 0.0016698032506415742, | |
| "loss": 1.2746, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.50176, | |
| "grad_norm": 0.09015596657991409, | |
| "learning_rate": 0.0016663815226689478, | |
| "loss": 1.2659, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.50688, | |
| "grad_norm": 0.08708484470844269, | |
| "learning_rate": 0.0016629597946963216, | |
| "loss": 1.2727, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.09585189074277878, | |
| "learning_rate": 0.0016595380667236955, | |
| "loss": 1.2702, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.51712, | |
| "grad_norm": 0.08958299458026886, | |
| "learning_rate": 0.0016561163387510693, | |
| "loss": 1.2605, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.52224, | |
| "grad_norm": 0.0867680162191391, | |
| "learning_rate": 0.0016526946107784431, | |
| "loss": 1.2561, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.52736, | |
| "grad_norm": 0.08700387924909592, | |
| "learning_rate": 0.001649272882805817, | |
| "loss": 1.2481, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.53248, | |
| "grad_norm": 0.08566949516534805, | |
| "learning_rate": 0.0016458511548331908, | |
| "loss": 1.255, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.08990107476711273, | |
| "learning_rate": 0.0016424294268605646, | |
| "loss": 1.2558, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.54272, | |
| "grad_norm": 0.08288553357124329, | |
| "learning_rate": 0.0016390076988879385, | |
| "loss": 1.2507, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.54784, | |
| "grad_norm": 0.08027470856904984, | |
| "learning_rate": 0.0016355859709153123, | |
| "loss": 1.278, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.55296, | |
| "grad_norm": 0.09287162125110626, | |
| "learning_rate": 0.0016321642429426861, | |
| "loss": 1.2626, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.55808, | |
| "grad_norm": 0.09153173863887787, | |
| "learning_rate": 0.00162874251497006, | |
| "loss": 1.2643, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.10922811180353165, | |
| "learning_rate": 0.0016253207869974338, | |
| "loss": 1.2755, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.56832, | |
| "grad_norm": 0.1041250005364418, | |
| "learning_rate": 0.0016218990590248074, | |
| "loss": 1.2493, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.57344, | |
| "grad_norm": 0.08747130632400513, | |
| "learning_rate": 0.0016184773310521814, | |
| "loss": 1.2491, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.57856, | |
| "grad_norm": 0.10124468803405762, | |
| "learning_rate": 0.001615055603079555, | |
| "loss": 1.2387, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.58368, | |
| "grad_norm": 0.09022431075572968, | |
| "learning_rate": 0.001611633875106929, | |
| "loss": 1.2521, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.09623134136199951, | |
| "learning_rate": 0.0016082121471343027, | |
| "loss": 1.2701, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.59392, | |
| "grad_norm": 0.09208202362060547, | |
| "learning_rate": 0.0016047904191616768, | |
| "loss": 1.2413, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.59904, | |
| "grad_norm": 0.10192185640335083, | |
| "learning_rate": 0.0016013686911890506, | |
| "loss": 1.278, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.60416, | |
| "grad_norm": 0.11536680907011032, | |
| "learning_rate": 0.0015979469632164242, | |
| "loss": 1.2729, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.60928, | |
| "grad_norm": 0.11593331396579742, | |
| "learning_rate": 0.0015945252352437982, | |
| "loss": 1.2772, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.10352569818496704, | |
| "learning_rate": 0.0015911035072711719, | |
| "loss": 1.2587, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.61952, | |
| "grad_norm": 0.11213277280330658, | |
| "learning_rate": 0.001587681779298546, | |
| "loss": 1.2626, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.62464, | |
| "grad_norm": 0.11924043297767639, | |
| "learning_rate": 0.0015842600513259195, | |
| "loss": 1.2826, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.62976, | |
| "grad_norm": 0.09039822220802307, | |
| "learning_rate": 0.0015808383233532936, | |
| "loss": 1.2777, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.63488, | |
| "grad_norm": 0.09418819099664688, | |
| "learning_rate": 0.0015774165953806672, | |
| "loss": 1.2765, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.09826017916202545, | |
| "learning_rate": 0.001573994867408041, | |
| "loss": 1.2474, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.64512, | |
| "grad_norm": 0.09216541796922684, | |
| "learning_rate": 0.001570573139435415, | |
| "loss": 1.243, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.65024, | |
| "grad_norm": 0.09861636161804199, | |
| "learning_rate": 0.0015671514114627887, | |
| "loss": 1.239, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.65536, | |
| "grad_norm": 0.1033303365111351, | |
| "learning_rate": 0.0015637296834901627, | |
| "loss": 1.2571, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.66048, | |
| "grad_norm": 0.10304012894630432, | |
| "learning_rate": 0.0015603079555175363, | |
| "loss": 1.2556, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.0865844339132309, | |
| "learning_rate": 0.0015568862275449104, | |
| "loss": 1.2475, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.67072, | |
| "grad_norm": 0.10008803755044937, | |
| "learning_rate": 0.001553464499572284, | |
| "loss": 1.2795, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.67584, | |
| "grad_norm": 0.09249156713485718, | |
| "learning_rate": 0.0015500427715996578, | |
| "loss": 1.263, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.68096, | |
| "grad_norm": 0.09253022074699402, | |
| "learning_rate": 0.0015466210436270317, | |
| "loss": 1.2901, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.68608, | |
| "grad_norm": 0.09615321457386017, | |
| "learning_rate": 0.0015431993156544055, | |
| "loss": 1.2643, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.13639943301677704, | |
| "learning_rate": 0.0015397775876817793, | |
| "loss": 1.2586, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.69632, | |
| "grad_norm": 0.10662351548671722, | |
| "learning_rate": 0.0015363558597091531, | |
| "loss": 1.2414, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.70144, | |
| "grad_norm": 0.08936125040054321, | |
| "learning_rate": 0.001532934131736527, | |
| "loss": 1.2645, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.70656, | |
| "grad_norm": 0.09400724619626999, | |
| "learning_rate": 0.0015295124037639008, | |
| "loss": 1.2274, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.71168, | |
| "grad_norm": 0.10700514912605286, | |
| "learning_rate": 0.0015260906757912746, | |
| "loss": 1.2564, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.09849894791841507, | |
| "learning_rate": 0.0015226689478186485, | |
| "loss": 1.2658, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.72192, | |
| "grad_norm": 0.10522880405187607, | |
| "learning_rate": 0.0015192472198460223, | |
| "loss": 1.2775, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.72704, | |
| "grad_norm": 0.09173361957073212, | |
| "learning_rate": 0.0015158254918733961, | |
| "loss": 1.2738, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.73216, | |
| "grad_norm": 0.09410373121500015, | |
| "learning_rate": 0.00151240376390077, | |
| "loss": 1.2483, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.73728, | |
| "grad_norm": 0.08925613760948181, | |
| "learning_rate": 0.0015089820359281436, | |
| "loss": 1.2554, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.09131377190351486, | |
| "learning_rate": 0.0015055603079555176, | |
| "loss": 1.2496, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.74752, | |
| "grad_norm": 0.10802093148231506, | |
| "learning_rate": 0.0015021385799828914, | |
| "loss": 1.2675, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.75264, | |
| "grad_norm": 0.09466376155614853, | |
| "learning_rate": 0.0014987168520102653, | |
| "loss": 1.2482, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.75776, | |
| "grad_norm": 0.10151738673448563, | |
| "learning_rate": 0.001495295124037639, | |
| "loss": 1.2676, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.76288, | |
| "grad_norm": 0.08664025366306305, | |
| "learning_rate": 0.0014918733960650127, | |
| "loss": 1.2496, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.08441973477602005, | |
| "learning_rate": 0.0014884516680923868, | |
| "loss": 1.254, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.77312, | |
| "grad_norm": 0.08028802275657654, | |
| "learning_rate": 0.0014850299401197604, | |
| "loss": 1.2718, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.77824, | |
| "grad_norm": 0.08772825449705124, | |
| "learning_rate": 0.0014816082121471344, | |
| "loss": 1.2624, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.78336, | |
| "grad_norm": 0.08271320164203644, | |
| "learning_rate": 0.001478186484174508, | |
| "loss": 1.2467, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.78848, | |
| "grad_norm": 0.08346061408519745, | |
| "learning_rate": 0.001474764756201882, | |
| "loss": 1.2782, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.09925299137830734, | |
| "learning_rate": 0.0014713430282292557, | |
| "loss": 1.2489, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.79872, | |
| "grad_norm": 0.08743379265069962, | |
| "learning_rate": 0.0014679213002566295, | |
| "loss": 1.257, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.80384, | |
| "grad_norm": 0.08556243032217026, | |
| "learning_rate": 0.0014644995722840036, | |
| "loss": 1.2507, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.80896, | |
| "grad_norm": 0.0923091247677803, | |
| "learning_rate": 0.0014610778443113772, | |
| "loss": 1.2551, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.81408, | |
| "grad_norm": 0.0928301066160202, | |
| "learning_rate": 0.0014576561163387512, | |
| "loss": 1.2309, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.08794920146465302, | |
| "learning_rate": 0.0014542343883661248, | |
| "loss": 1.2457, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.82432, | |
| "grad_norm": 0.10327792912721634, | |
| "learning_rate": 0.001450812660393499, | |
| "loss": 1.2786, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.82944, | |
| "grad_norm": 0.12187407165765762, | |
| "learning_rate": 0.0014473909324208725, | |
| "loss": 1.2814, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.83456, | |
| "grad_norm": 0.11819777637720108, | |
| "learning_rate": 0.0014439692044482463, | |
| "loss": 1.2499, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.83968, | |
| "grad_norm": 0.09041640162467957, | |
| "learning_rate": 0.0014405474764756202, | |
| "loss": 1.2693, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.09638890624046326, | |
| "learning_rate": 0.001437125748502994, | |
| "loss": 1.2514, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.84992, | |
| "grad_norm": 0.12534624338150024, | |
| "learning_rate": 0.001433704020530368, | |
| "loss": 1.2407, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.85504, | |
| "grad_norm": 0.09927276521921158, | |
| "learning_rate": 0.0014302822925577417, | |
| "loss": 1.2749, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.86016, | |
| "grad_norm": 0.0886382907629013, | |
| "learning_rate": 0.0014268605645851157, | |
| "loss": 1.2478, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.86528, | |
| "grad_norm": 0.09105540066957474, | |
| "learning_rate": 0.0014234388366124893, | |
| "loss": 1.2538, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.08825322240591049, | |
| "learning_rate": 0.0014200171086398631, | |
| "loss": 1.2679, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.87552, | |
| "grad_norm": 0.08396722376346588, | |
| "learning_rate": 0.001416595380667237, | |
| "loss": 1.2091, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.88064, | |
| "grad_norm": 0.08241663128137589, | |
| "learning_rate": 0.0014131736526946108, | |
| "loss": 1.2401, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.88576, | |
| "grad_norm": 0.11285565793514252, | |
| "learning_rate": 0.0014097519247219846, | |
| "loss": 1.2461, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.89088, | |
| "grad_norm": 0.09898606687784195, | |
| "learning_rate": 0.0014063301967493585, | |
| "loss": 1.2587, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.09175729751586914, | |
| "learning_rate": 0.001402908468776732, | |
| "loss": 1.2318, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.90112, | |
| "grad_norm": 0.08974505960941315, | |
| "learning_rate": 0.0013994867408041061, | |
| "loss": 1.265, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.90624, | |
| "grad_norm": 0.10069482773542404, | |
| "learning_rate": 0.00139606501283148, | |
| "loss": 1.2276, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.91136, | |
| "grad_norm": 0.09355876594781876, | |
| "learning_rate": 0.0013926432848588538, | |
| "loss": 1.2474, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.91648, | |
| "grad_norm": 0.09747931361198425, | |
| "learning_rate": 0.0013892215568862276, | |
| "loss": 1.2677, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.09114759415388107, | |
| "learning_rate": 0.0013857998289136014, | |
| "loss": 1.2442, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.92672, | |
| "grad_norm": 0.09683123230934143, | |
| "learning_rate": 0.0013823781009409753, | |
| "loss": 1.2328, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.93184, | |
| "grad_norm": 0.08525967597961426, | |
| "learning_rate": 0.0013789563729683489, | |
| "loss": 1.2301, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.93696, | |
| "grad_norm": 0.09861680120229721, | |
| "learning_rate": 0.001375534644995723, | |
| "loss": 1.2604, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.94208, | |
| "grad_norm": 0.08435367792844772, | |
| "learning_rate": 0.0013721129170230965, | |
| "loss": 1.2605, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.08719425648450851, | |
| "learning_rate": 0.0013686911890504706, | |
| "loss": 1.2412, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.95232, | |
| "grad_norm": 0.10853152722120285, | |
| "learning_rate": 0.0013652694610778444, | |
| "loss": 1.2503, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.95744, | |
| "grad_norm": 0.08797234296798706, | |
| "learning_rate": 0.001361847733105218, | |
| "loss": 1.2323, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.96256, | |
| "grad_norm": 0.09422844648361206, | |
| "learning_rate": 0.001358426005132592, | |
| "loss": 1.2444, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.96768, | |
| "grad_norm": 0.09000077098608017, | |
| "learning_rate": 0.0013550042771599657, | |
| "loss": 1.2573, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.09097360074520111, | |
| "learning_rate": 0.0013515825491873397, | |
| "loss": 1.2463, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.97792, | |
| "grad_norm": 0.08720215409994125, | |
| "learning_rate": 0.0013481608212147134, | |
| "loss": 1.2719, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.98304, | |
| "grad_norm": 0.09287154674530029, | |
| "learning_rate": 0.0013447390932420874, | |
| "loss": 1.2387, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.98816, | |
| "grad_norm": 0.08979474008083344, | |
| "learning_rate": 0.001341317365269461, | |
| "loss": 1.2353, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.99328, | |
| "grad_norm": 0.0891214981675148, | |
| "learning_rate": 0.0013378956372968348, | |
| "loss": 1.2441, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.09961092472076416, | |
| "learning_rate": 0.0013344739093242087, | |
| "loss": 1.2407, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.00256, | |
| "grad_norm": 0.09316655993461609, | |
| "learning_rate": 0.0013310521813515825, | |
| "loss": 1.2589, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.00768, | |
| "grad_norm": 0.6676607131958008, | |
| "learning_rate": 0.0013276304533789566, | |
| "loss": 1.3337, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.0128, | |
| "grad_norm": 0.24017909169197083, | |
| "learning_rate": 0.0013242087254063302, | |
| "loss": 1.2628, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.01792, | |
| "grad_norm": 0.17012369632720947, | |
| "learning_rate": 0.0013207869974337042, | |
| "loss": 1.2607, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.02304, | |
| "grad_norm": 0.18585637211799622, | |
| "learning_rate": 0.0013173652694610778, | |
| "loss": 1.2502, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.02304, | |
| "eval_loss": 1.2686372995376587, | |
| "eval_runtime": 279.6676, | |
| "eval_samples_per_second": 8.939, | |
| "eval_steps_per_second": 1.119, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.02816, | |
| "grad_norm": 0.16060733795166016, | |
| "learning_rate": 0.0013139435414884517, | |
| "loss": 1.2418, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.03328, | |
| "grad_norm": 0.13407620787620544, | |
| "learning_rate": 0.0013105218135158255, | |
| "loss": 1.2749, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.0384, | |
| "grad_norm": 0.12042540311813354, | |
| "learning_rate": 0.0013071000855431993, | |
| "loss": 1.2465, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.04352, | |
| "grad_norm": 0.13832047581672668, | |
| "learning_rate": 0.0013036783575705731, | |
| "loss": 1.2535, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.04864, | |
| "grad_norm": 0.11310556530952454, | |
| "learning_rate": 0.001300256629597947, | |
| "loss": 1.2352, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.05376, | |
| "grad_norm": 0.13382209837436676, | |
| "learning_rate": 0.001296834901625321, | |
| "loss": 1.2507, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.05888, | |
| "grad_norm": 0.13580721616744995, | |
| "learning_rate": 0.0012934131736526946, | |
| "loss": 1.2401, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 0.11162128299474716, | |
| "learning_rate": 0.0012899914456800685, | |
| "loss": 1.2534, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.06912, | |
| "grad_norm": 0.14186108112335205, | |
| "learning_rate": 0.0012865697177074423, | |
| "loss": 1.248, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.07424, | |
| "grad_norm": 0.13840018212795258, | |
| "learning_rate": 0.0012831479897348161, | |
| "loss": 1.2462, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.07936, | |
| "grad_norm": 0.11488790065050125, | |
| "learning_rate": 0.00127972626176219, | |
| "loss": 1.241, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.08448, | |
| "grad_norm": 0.1290818750858307, | |
| "learning_rate": 0.0012763045337895638, | |
| "loss": 1.2275, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.0896, | |
| "grad_norm": 0.10360855609178543, | |
| "learning_rate": 0.0012728828058169374, | |
| "loss": 1.2378, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.09472, | |
| "grad_norm": 0.12253882735967636, | |
| "learning_rate": 0.0012694610778443114, | |
| "loss": 1.2562, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.09984, | |
| "grad_norm": 0.12016449123620987, | |
| "learning_rate": 0.001266039349871685, | |
| "loss": 1.2742, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.10496, | |
| "grad_norm": 0.12187926471233368, | |
| "learning_rate": 0.0012626176218990591, | |
| "loss": 1.2202, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.11008, | |
| "grad_norm": 0.11081688851118088, | |
| "learning_rate": 0.001259195893926433, | |
| "loss": 1.2474, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.1152, | |
| "grad_norm": 0.09387561678886414, | |
| "learning_rate": 0.0012557741659538068, | |
| "loss": 1.2566, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.12032, | |
| "grad_norm": 0.11739682406187057, | |
| "learning_rate": 0.0012523524379811806, | |
| "loss": 1.2421, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.12544, | |
| "grad_norm": 0.11595962196588516, | |
| "learning_rate": 0.0012489307100085542, | |
| "loss": 1.2525, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.13056, | |
| "grad_norm": 0.10769950598478317, | |
| "learning_rate": 0.0012455089820359283, | |
| "loss": 1.2501, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.13568, | |
| "grad_norm": 0.11145374178886414, | |
| "learning_rate": 0.0012420872540633019, | |
| "loss": 1.2649, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.1408, | |
| "grad_norm": 0.09793559461832047, | |
| "learning_rate": 0.001238665526090676, | |
| "loss": 1.244, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.14592, | |
| "grad_norm": 0.16483676433563232, | |
| "learning_rate": 0.0012352437981180495, | |
| "loss": 1.2298, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.15104, | |
| "grad_norm": 0.13711702823638916, | |
| "learning_rate": 0.0012318220701454234, | |
| "loss": 1.2279, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.15616, | |
| "grad_norm": 0.10064009577035904, | |
| "learning_rate": 0.0012284003421727974, | |
| "loss": 1.2381, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.16128, | |
| "grad_norm": 0.09517936408519745, | |
| "learning_rate": 0.001224978614200171, | |
| "loss": 1.231, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.1663999999999999, | |
| "grad_norm": 0.111509308218956, | |
| "learning_rate": 0.001221556886227545, | |
| "loss": 1.25, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.1715200000000001, | |
| "grad_norm": 0.09270152449607849, | |
| "learning_rate": 0.0012181351582549187, | |
| "loss": 1.2478, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.17664, | |
| "grad_norm": 0.10461369901895523, | |
| "learning_rate": 0.0012147134302822927, | |
| "loss": 1.2536, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.18176, | |
| "grad_norm": 0.10189452022314072, | |
| "learning_rate": 0.0012112917023096663, | |
| "loss": 1.2177, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.18688, | |
| "grad_norm": 0.08983030170202255, | |
| "learning_rate": 0.0012078699743370402, | |
| "loss": 1.2117, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.192, | |
| "grad_norm": 0.09105629473924637, | |
| "learning_rate": 0.001204448246364414, | |
| "loss": 1.2302, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.19712, | |
| "grad_norm": 0.11691851168870926, | |
| "learning_rate": 0.0012010265183917878, | |
| "loss": 1.2101, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.20224, | |
| "grad_norm": 0.08926935493946075, | |
| "learning_rate": 0.0011976047904191617, | |
| "loss": 1.2479, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.20736, | |
| "grad_norm": 0.11435071378946304, | |
| "learning_rate": 0.0011941830624465355, | |
| "loss": 1.2593, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.21248, | |
| "grad_norm": 0.10086748749017715, | |
| "learning_rate": 0.0011907613344739095, | |
| "loss": 1.2413, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.2176, | |
| "grad_norm": 0.098397396504879, | |
| "learning_rate": 0.0011873396065012832, | |
| "loss": 1.228, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.22272, | |
| "grad_norm": 0.08532971143722534, | |
| "learning_rate": 0.001183917878528657, | |
| "loss": 1.2309, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.22784, | |
| "grad_norm": 0.10852818191051483, | |
| "learning_rate": 0.0011804961505560308, | |
| "loss": 1.232, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.23296, | |
| "grad_norm": 0.09808767586946487, | |
| "learning_rate": 0.0011770744225834046, | |
| "loss": 1.2415, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.23808, | |
| "grad_norm": 0.11177875101566315, | |
| "learning_rate": 0.0011736526946107785, | |
| "loss": 1.2592, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.2432, | |
| "grad_norm": 0.1047763004899025, | |
| "learning_rate": 0.0011702309666381523, | |
| "loss": 1.2487, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.24832, | |
| "grad_norm": 0.12227226048707962, | |
| "learning_rate": 0.001166809238665526, | |
| "loss": 1.2481, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.2534399999999999, | |
| "grad_norm": 0.11121272295713425, | |
| "learning_rate": 0.0011633875106929, | |
| "loss": 1.2369, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.2585600000000001, | |
| "grad_norm": 0.1087367981672287, | |
| "learning_rate": 0.0011599657827202738, | |
| "loss": 1.2378, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.26368, | |
| "grad_norm": 0.09499981254339218, | |
| "learning_rate": 0.0011565440547476476, | |
| "loss": 1.2327, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.2688, | |
| "grad_norm": 0.12346815317869186, | |
| "learning_rate": 0.0011531223267750215, | |
| "loss": 1.2538, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.27392, | |
| "grad_norm": 0.10443610697984695, | |
| "learning_rate": 0.0011497005988023953, | |
| "loss": 1.2343, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.27904, | |
| "grad_norm": 0.1339293271303177, | |
| "learning_rate": 0.0011462788708297691, | |
| "loss": 1.2198, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.28416, | |
| "grad_norm": 0.08476725220680237, | |
| "learning_rate": 0.0011428571428571427, | |
| "loss": 1.236, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.28928, | |
| "grad_norm": 0.12042795866727829, | |
| "learning_rate": 0.0011394354148845168, | |
| "loss": 1.2357, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.2944, | |
| "grad_norm": 0.08857988566160202, | |
| "learning_rate": 0.0011360136869118904, | |
| "loss": 1.2476, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.29952, | |
| "grad_norm": 0.1092582419514656, | |
| "learning_rate": 0.0011325919589392644, | |
| "loss": 1.2445, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.30464, | |
| "grad_norm": 0.0912066176533699, | |
| "learning_rate": 0.001129170230966638, | |
| "loss": 1.2556, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.30976, | |
| "grad_norm": 0.12163588404655457, | |
| "learning_rate": 0.001125748502994012, | |
| "loss": 1.2427, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.31488, | |
| "grad_norm": 0.0835074707865715, | |
| "learning_rate": 0.001122326775021386, | |
| "loss": 1.2546, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.1106950119137764, | |
| "learning_rate": 0.0011189050470487595, | |
| "loss": 1.2168, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.32512, | |
| "grad_norm": 0.09452968090772629, | |
| "learning_rate": 0.0011154833190761336, | |
| "loss": 1.2314, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.3302399999999999, | |
| "grad_norm": 0.11672431975603104, | |
| "learning_rate": 0.0011120615911035072, | |
| "loss": 1.2455, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.33536, | |
| "grad_norm": 0.10683607310056686, | |
| "learning_rate": 0.0011086398631308812, | |
| "loss": 1.2474, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.34048, | |
| "grad_norm": 0.08974877744913101, | |
| "learning_rate": 0.0011052181351582549, | |
| "loss": 1.2226, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.3456000000000001, | |
| "grad_norm": 0.10645844787359238, | |
| "learning_rate": 0.0011017964071856287, | |
| "loss": 1.2545, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.35072, | |
| "grad_norm": 0.08754228055477142, | |
| "learning_rate": 0.0010983746792130025, | |
| "loss": 1.217, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.35584, | |
| "grad_norm": 0.10209974646568298, | |
| "learning_rate": 0.0010949529512403763, | |
| "loss": 1.2399, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.36096, | |
| "grad_norm": 0.09669913351535797, | |
| "learning_rate": 0.0010915312232677504, | |
| "loss": 1.2081, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.36608, | |
| "grad_norm": 0.10272342711687088, | |
| "learning_rate": 0.001088109495295124, | |
| "loss": 1.2635, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.3712, | |
| "grad_norm": 0.0976710096001625, | |
| "learning_rate": 0.001084687767322498, | |
| "loss": 1.2347, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.37632, | |
| "grad_norm": 0.09784968197345734, | |
| "learning_rate": 0.0010812660393498717, | |
| "loss": 1.2522, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.38144, | |
| "grad_norm": 0.09353113174438477, | |
| "learning_rate": 0.0010778443113772455, | |
| "loss": 1.2326, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.38656, | |
| "grad_norm": 0.10906370729207993, | |
| "learning_rate": 0.0010744225834046193, | |
| "loss": 1.2375, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.39168, | |
| "grad_norm": 0.11907199025154114, | |
| "learning_rate": 0.0010710008554319932, | |
| "loss": 1.2185, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.3968, | |
| "grad_norm": 0.10644809901714325, | |
| "learning_rate": 0.001067579127459367, | |
| "loss": 1.2216, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.40192, | |
| "grad_norm": 0.10000847280025482, | |
| "learning_rate": 0.0010641573994867408, | |
| "loss": 1.2471, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.40704, | |
| "grad_norm": 0.11204187572002411, | |
| "learning_rate": 0.0010607356715141146, | |
| "loss": 1.2693, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.41216, | |
| "grad_norm": 0.0837775245308876, | |
| "learning_rate": 0.0010573139435414885, | |
| "loss": 1.2444, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.4172799999999999, | |
| "grad_norm": 0.09714753180742264, | |
| "learning_rate": 0.0010538922155688623, | |
| "loss": 1.2525, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.4224, | |
| "grad_norm": 0.08763246238231659, | |
| "learning_rate": 0.0010504704875962361, | |
| "loss": 1.2678, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.42752, | |
| "grad_norm": 0.12144036591053009, | |
| "learning_rate": 0.00104704875962361, | |
| "loss": 1.2285, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.4326400000000001, | |
| "grad_norm": 0.11393667757511139, | |
| "learning_rate": 0.0010436270316509838, | |
| "loss": 1.253, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.43776, | |
| "grad_norm": 0.0934453159570694, | |
| "learning_rate": 0.0010402053036783576, | |
| "loss": 1.2388, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.44288, | |
| "grad_norm": 0.1040380522608757, | |
| "learning_rate": 0.0010367835757057312, | |
| "loss": 1.2446, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 0.09509964287281036, | |
| "learning_rate": 0.0010333618477331053, | |
| "loss": 1.2328, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.45312, | |
| "grad_norm": 0.09978800266981125, | |
| "learning_rate": 0.001029940119760479, | |
| "loss": 1.2385, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.45824, | |
| "grad_norm": 0.08431090414524078, | |
| "learning_rate": 0.001026518391787853, | |
| "loss": 1.2389, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.46336, | |
| "grad_norm": 0.09447863698005676, | |
| "learning_rate": 0.0010230966638152268, | |
| "loss": 1.2301, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.46848, | |
| "grad_norm": 0.09212321043014526, | |
| "learning_rate": 0.0010196749358426006, | |
| "loss": 1.2457, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.4736, | |
| "grad_norm": 0.09489751607179642, | |
| "learning_rate": 0.0010162532078699744, | |
| "loss": 1.2229, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.47872, | |
| "grad_norm": 0.08897145092487335, | |
| "learning_rate": 0.001012831479897348, | |
| "loss": 1.2242, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.48384, | |
| "grad_norm": 0.09199076145887375, | |
| "learning_rate": 0.001009409751924722, | |
| "loss": 1.251, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.48896, | |
| "grad_norm": 0.1008097380399704, | |
| "learning_rate": 0.0010059880239520957, | |
| "loss": 1.2633, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.49408, | |
| "grad_norm": 0.0928090438246727, | |
| "learning_rate": 0.0010025662959794698, | |
| "loss": 1.238, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.4992, | |
| "grad_norm": 0.09936055541038513, | |
| "learning_rate": 0.0009991445680068436, | |
| "loss": 1.2502, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.5043199999999999, | |
| "grad_norm": 0.12435046583414078, | |
| "learning_rate": 0.0009957228400342174, | |
| "loss": 1.238, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.5094400000000001, | |
| "grad_norm": 0.09954190254211426, | |
| "learning_rate": 0.000992301112061591, | |
| "loss": 1.2403, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.51456, | |
| "grad_norm": 0.0886056125164032, | |
| "learning_rate": 0.0009888793840889649, | |
| "loss": 1.2413, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.5196800000000001, | |
| "grad_norm": 0.10016464442014694, | |
| "learning_rate": 0.0009854576561163387, | |
| "loss": 1.2272, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.5248, | |
| "grad_norm": 0.08891763538122177, | |
| "learning_rate": 0.0009820359281437125, | |
| "loss": 1.2198, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.52992, | |
| "grad_norm": 0.08042890578508377, | |
| "learning_rate": 0.0009786142001710863, | |
| "loss": 1.2386, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.53504, | |
| "grad_norm": 0.09712400287389755, | |
| "learning_rate": 0.0009751924721984602, | |
| "loss": 1.2418, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.54016, | |
| "grad_norm": 0.09671667218208313, | |
| "learning_rate": 0.0009717707442258341, | |
| "loss": 1.2493, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.54528, | |
| "grad_norm": 0.1355689913034439, | |
| "learning_rate": 0.000968349016253208, | |
| "loss": 1.2529, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.5504, | |
| "grad_norm": 0.08577972650527954, | |
| "learning_rate": 0.0009649272882805818, | |
| "loss": 1.2444, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.55552, | |
| "grad_norm": 0.08881525695323944, | |
| "learning_rate": 0.0009615055603079555, | |
| "loss": 1.2561, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.56064, | |
| "grad_norm": 0.08444136381149292, | |
| "learning_rate": 0.0009580838323353293, | |
| "loss": 1.2378, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.56576, | |
| "grad_norm": 0.08552881330251694, | |
| "learning_rate": 0.0009546621043627032, | |
| "loss": 1.2186, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.57088, | |
| "grad_norm": 0.08705168962478638, | |
| "learning_rate": 0.000951240376390077, | |
| "loss": 1.2451, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.576, | |
| "grad_norm": 0.08805005997419357, | |
| "learning_rate": 0.0009478186484174508, | |
| "loss": 1.225, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.5811199999999999, | |
| "grad_norm": 0.10901911556720734, | |
| "learning_rate": 0.0009443969204448247, | |
| "loss": 1.2351, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.58624, | |
| "grad_norm": 0.08406229317188263, | |
| "learning_rate": 0.0009409751924721985, | |
| "loss": 1.2535, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.5913599999999999, | |
| "grad_norm": 0.10775440186262131, | |
| "learning_rate": 0.0009375534644995723, | |
| "loss": 1.2474, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.5964800000000001, | |
| "grad_norm": 0.09037076681852341, | |
| "learning_rate": 0.0009341317365269461, | |
| "loss": 1.2346, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.6016, | |
| "grad_norm": 0.10725659132003784, | |
| "learning_rate": 0.00093071000855432, | |
| "loss": 1.2156, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.6067200000000001, | |
| "grad_norm": 0.08220596611499786, | |
| "learning_rate": 0.0009272882805816938, | |
| "loss": 1.2335, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.61184, | |
| "grad_norm": 0.08765338361263275, | |
| "learning_rate": 0.0009238665526090676, | |
| "loss": 1.2311, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.61696, | |
| "grad_norm": 0.0885564312338829, | |
| "learning_rate": 0.0009204448246364415, | |
| "loss": 1.2381, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.62208, | |
| "grad_norm": 0.09842797368764877, | |
| "learning_rate": 0.0009170230966638152, | |
| "loss": 1.2362, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.6272, | |
| "grad_norm": 0.08304440975189209, | |
| "learning_rate": 0.000913601368691189, | |
| "loss": 1.2635, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.63232, | |
| "grad_norm": 0.0953385978937149, | |
| "learning_rate": 0.0009101796407185628, | |
| "loss": 1.2221, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.63744, | |
| "grad_norm": 0.09365107119083405, | |
| "learning_rate": 0.0009067579127459367, | |
| "loss": 1.2449, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.64256, | |
| "grad_norm": 0.09064996242523193, | |
| "learning_rate": 0.0009033361847733106, | |
| "loss": 1.2212, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.64768, | |
| "grad_norm": 0.08681759238243103, | |
| "learning_rate": 0.0008999144568006844, | |
| "loss": 1.2581, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.6528, | |
| "grad_norm": 0.08267663419246674, | |
| "learning_rate": 0.0008964927288280582, | |
| "loss": 1.2347, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.6579199999999998, | |
| "grad_norm": 0.08745119720697403, | |
| "learning_rate": 0.000893071000855432, | |
| "loss": 1.2289, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.66304, | |
| "grad_norm": 0.08867384493350983, | |
| "learning_rate": 0.0008896492728828058, | |
| "loss": 1.2598, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.6681599999999999, | |
| "grad_norm": 0.08820460736751556, | |
| "learning_rate": 0.0008862275449101797, | |
| "loss": 1.2458, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.67328, | |
| "grad_norm": 0.1035899817943573, | |
| "learning_rate": 0.0008828058169375535, | |
| "loss": 1.2346, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.6784, | |
| "grad_norm": 0.08878592401742935, | |
| "learning_rate": 0.0008793840889649273, | |
| "loss": 1.2332, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.6835200000000001, | |
| "grad_norm": 0.08506552875041962, | |
| "learning_rate": 0.0008759623609923011, | |
| "loss": 1.2392, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.68864, | |
| "grad_norm": 0.09227900952100754, | |
| "learning_rate": 0.0008725406330196749, | |
| "loss": 1.2552, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.6937600000000002, | |
| "grad_norm": 0.08019377291202545, | |
| "learning_rate": 0.0008691189050470488, | |
| "loss": 1.2208, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.69888, | |
| "grad_norm": 0.09494514763355255, | |
| "learning_rate": 0.0008656971770744226, | |
| "loss": 1.2367, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 0.09500183910131454, | |
| "learning_rate": 0.0008622754491017965, | |
| "loss": 1.1998, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.70912, | |
| "grad_norm": 0.08492112904787064, | |
| "learning_rate": 0.0008588537211291703, | |
| "loss": 1.2489, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.71424, | |
| "grad_norm": 0.0840214341878891, | |
| "learning_rate": 0.0008554319931565441, | |
| "loss": 1.2358, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.71936, | |
| "grad_norm": 0.0879180058836937, | |
| "learning_rate": 0.0008520102651839178, | |
| "loss": 1.2448, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.72448, | |
| "grad_norm": 0.08872208744287491, | |
| "learning_rate": 0.0008485885372112917, | |
| "loss": 1.2432, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.7296, | |
| "grad_norm": 0.09941036254167557, | |
| "learning_rate": 0.0008451668092386655, | |
| "loss": 1.2416, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.73472, | |
| "grad_norm": 0.08928696811199188, | |
| "learning_rate": 0.0008417450812660393, | |
| "loss": 1.2545, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.73984, | |
| "grad_norm": 0.08976240456104279, | |
| "learning_rate": 0.0008383233532934132, | |
| "loss": 1.2596, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.7449599999999998, | |
| "grad_norm": 0.08667703717947006, | |
| "learning_rate": 0.0008349016253207871, | |
| "loss": 1.2499, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.75008, | |
| "grad_norm": 0.08816345036029816, | |
| "learning_rate": 0.0008314798973481608, | |
| "loss": 1.2339, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.7551999999999999, | |
| "grad_norm": 0.08712169528007507, | |
| "learning_rate": 0.0008280581693755347, | |
| "loss": 1.2206, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.76032, | |
| "grad_norm": 0.08284337818622589, | |
| "learning_rate": 0.0008246364414029085, | |
| "loss": 1.2481, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.76544, | |
| "grad_norm": 0.08372201770544052, | |
| "learning_rate": 0.0008212147134302823, | |
| "loss": 1.2258, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.7705600000000001, | |
| "grad_norm": 0.08462055772542953, | |
| "learning_rate": 0.0008177929854576561, | |
| "loss": 1.255, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.77568, | |
| "grad_norm": 0.08806449919939041, | |
| "learning_rate": 0.00081437125748503, | |
| "loss": 1.2395, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.7808000000000002, | |
| "grad_norm": 0.08771070092916489, | |
| "learning_rate": 0.0008109495295124037, | |
| "loss": 1.2165, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.78592, | |
| "grad_norm": 0.10757436603307724, | |
| "learning_rate": 0.0008075278015397775, | |
| "loss": 1.2463, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.79104, | |
| "grad_norm": 0.10293210297822952, | |
| "learning_rate": 0.0008041060735671514, | |
| "loss": 1.2633, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.79616, | |
| "grad_norm": 0.08851849287748337, | |
| "learning_rate": 0.0008006843455945253, | |
| "loss": 1.2306, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.80128, | |
| "grad_norm": 0.08972053974866867, | |
| "learning_rate": 0.0007972626176218991, | |
| "loss": 1.2173, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.8064, | |
| "grad_norm": 0.12318170815706253, | |
| "learning_rate": 0.000793840889649273, | |
| "loss": 1.2528, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.81152, | |
| "grad_norm": 0.08965172618627548, | |
| "learning_rate": 0.0007904191616766468, | |
| "loss": 1.2102, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.81664, | |
| "grad_norm": 0.10421866178512573, | |
| "learning_rate": 0.0007869974337040205, | |
| "loss": 1.2465, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.82176, | |
| "grad_norm": 0.09522471576929092, | |
| "learning_rate": 0.0007835757057313943, | |
| "loss": 1.2386, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.82688, | |
| "grad_norm": 0.10268909484148026, | |
| "learning_rate": 0.0007801539777587682, | |
| "loss": 1.2592, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 0.10645327717065811, | |
| "learning_rate": 0.000776732249786142, | |
| "loss": 1.2109, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.83712, | |
| "grad_norm": 0.11298143118619919, | |
| "learning_rate": 0.0007733105218135158, | |
| "loss": 1.2527, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.8422399999999999, | |
| "grad_norm": 0.09252069890499115, | |
| "learning_rate": 0.0007698887938408897, | |
| "loss": 1.2461, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.8473600000000001, | |
| "grad_norm": 0.1118890643119812, | |
| "learning_rate": 0.0007664670658682635, | |
| "loss": 1.2334, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.85248, | |
| "grad_norm": 0.09874032437801361, | |
| "learning_rate": 0.0007630453378956373, | |
| "loss": 1.2281, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.8576000000000001, | |
| "grad_norm": 0.09806526452302933, | |
| "learning_rate": 0.0007596236099230111, | |
| "loss": 1.2403, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.86272, | |
| "grad_norm": 0.11208830773830414, | |
| "learning_rate": 0.000756201881950385, | |
| "loss": 1.2433, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.86784, | |
| "grad_norm": 0.09039215743541718, | |
| "learning_rate": 0.0007527801539777588, | |
| "loss": 1.2366, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.87296, | |
| "grad_norm": 0.08755876123905182, | |
| "learning_rate": 0.0007493584260051326, | |
| "loss": 1.2419, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.87808, | |
| "grad_norm": 0.0822838544845581, | |
| "learning_rate": 0.0007459366980325064, | |
| "loss": 1.2155, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.8832, | |
| "grad_norm": 0.08834468573331833, | |
| "learning_rate": 0.0007425149700598802, | |
| "loss": 1.2575, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.88832, | |
| "grad_norm": 0.09723412245512009, | |
| "learning_rate": 0.000739093242087254, | |
| "loss": 1.2137, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.89344, | |
| "grad_norm": 0.11283569037914276, | |
| "learning_rate": 0.0007356715141146278, | |
| "loss": 1.2589, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.89856, | |
| "grad_norm": 0.08855767548084259, | |
| "learning_rate": 0.0007322497861420018, | |
| "loss": 1.2095, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.90368, | |
| "grad_norm": 0.09914392232894897, | |
| "learning_rate": 0.0007288280581693756, | |
| "loss": 1.2272, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.9088, | |
| "grad_norm": 0.10668183118104935, | |
| "learning_rate": 0.0007254063301967494, | |
| "loss": 1.2555, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.91392, | |
| "grad_norm": 0.09255476295948029, | |
| "learning_rate": 0.0007219846022241232, | |
| "loss": 1.2248, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.9190399999999999, | |
| "grad_norm": 0.10384318232536316, | |
| "learning_rate": 0.000718562874251497, | |
| "loss": 1.2128, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.92416, | |
| "grad_norm": 0.09946981072425842, | |
| "learning_rate": 0.0007151411462788708, | |
| "loss": 1.2323, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.9292799999999999, | |
| "grad_norm": 0.10720199346542358, | |
| "learning_rate": 0.0007117194183062447, | |
| "loss": 1.2361, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.9344000000000001, | |
| "grad_norm": 0.10026301443576813, | |
| "learning_rate": 0.0007082976903336185, | |
| "loss": 1.2443, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.93952, | |
| "grad_norm": 0.08168992400169373, | |
| "learning_rate": 0.0007048759623609923, | |
| "loss": 1.2297, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.9446400000000001, | |
| "grad_norm": 0.11108248680830002, | |
| "learning_rate": 0.000701454234388366, | |
| "loss": 1.2359, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.94976, | |
| "grad_norm": 0.1133013665676117, | |
| "learning_rate": 0.00069803250641574, | |
| "loss": 1.2569, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.95488, | |
| "grad_norm": 0.0839882493019104, | |
| "learning_rate": 0.0006946107784431138, | |
| "loss": 1.231, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.09552154690027237, | |
| "learning_rate": 0.0006911890504704876, | |
| "loss": 1.2531, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.96512, | |
| "grad_norm": 0.09877890348434448, | |
| "learning_rate": 0.0006877673224978615, | |
| "loss": 1.2795, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.97024, | |
| "grad_norm": 0.08949697017669678, | |
| "learning_rate": 0.0006843455945252353, | |
| "loss": 1.2104, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.97536, | |
| "grad_norm": 0.09640631079673767, | |
| "learning_rate": 0.000680923866552609, | |
| "loss": 1.2552, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.98048, | |
| "grad_norm": 0.0900396853685379, | |
| "learning_rate": 0.0006775021385799828, | |
| "loss": 1.2282, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.9856, | |
| "grad_norm": 0.0989600196480751, | |
| "learning_rate": 0.0006740804106073567, | |
| "loss": 1.2366, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.99072, | |
| "grad_norm": 0.08420181274414062, | |
| "learning_rate": 0.0006706586826347305, | |
| "loss": 1.2309, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.9958399999999998, | |
| "grad_norm": 0.0978875532746315, | |
| "learning_rate": 0.0006672369546621043, | |
| "loss": 1.2241, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.10489444434642792, | |
| "learning_rate": 0.0006638152266894783, | |
| "loss": 1.2438, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 2.00512, | |
| "grad_norm": 0.08600255101919174, | |
| "learning_rate": 0.0006603934987168521, | |
| "loss": 1.2313, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 2.01024, | |
| "grad_norm": 0.08720952272415161, | |
| "learning_rate": 0.0006569717707442258, | |
| "loss": 1.2322, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 2.01536, | |
| "grad_norm": 0.08914855122566223, | |
| "learning_rate": 0.0006535500427715997, | |
| "loss": 1.2284, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 2.02048, | |
| "grad_norm": 0.08840183168649673, | |
| "learning_rate": 0.0006501283147989735, | |
| "loss": 1.2535, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.08714258670806885, | |
| "learning_rate": 0.0006467065868263473, | |
| "loss": 1.2358, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 2.03072, | |
| "grad_norm": 0.08174372464418411, | |
| "learning_rate": 0.0006432848588537211, | |
| "loss": 1.2419, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 2.03584, | |
| "grad_norm": 0.08205553144216537, | |
| "learning_rate": 0.000639863130881095, | |
| "loss": 1.2222, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 2.04096, | |
| "grad_norm": 0.0794735699892044, | |
| "learning_rate": 0.0006364414029084687, | |
| "loss": 1.2092, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 2.04608, | |
| "grad_norm": 0.07929161936044693, | |
| "learning_rate": 0.0006330196749358425, | |
| "loss": 1.2066, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.04608, | |
| "eval_loss": 1.2471669912338257, | |
| "eval_runtime": 280.5686, | |
| "eval_samples_per_second": 8.91, | |
| "eval_steps_per_second": 1.116, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.0512, | |
| "grad_norm": 0.08398760110139847, | |
| "learning_rate": 0.0006295979469632165, | |
| "loss": 1.2302, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 2.05632, | |
| "grad_norm": 0.08723915368318558, | |
| "learning_rate": 0.0006261762189905903, | |
| "loss": 1.2264, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 2.06144, | |
| "grad_norm": 0.0817415714263916, | |
| "learning_rate": 0.0006227544910179641, | |
| "loss": 1.2301, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 2.06656, | |
| "grad_norm": 0.08779250085353851, | |
| "learning_rate": 0.000619332763045338, | |
| "loss": 1.2228, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 2.07168, | |
| "grad_norm": 0.09133391082286835, | |
| "learning_rate": 0.0006159110350727117, | |
| "loss": 1.2415, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.0768, | |
| "grad_norm": 0.08566722273826599, | |
| "learning_rate": 0.0006124893071000855, | |
| "loss": 1.2121, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 2.08192, | |
| "grad_norm": 0.08748096972703934, | |
| "learning_rate": 0.0006090675791274593, | |
| "loss": 1.2413, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 2.08704, | |
| "grad_norm": 0.0854722335934639, | |
| "learning_rate": 0.0006056458511548332, | |
| "loss": 1.2326, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 2.09216, | |
| "grad_norm": 0.08618238568305969, | |
| "learning_rate": 0.000602224123182207, | |
| "loss": 1.2216, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 2.09728, | |
| "grad_norm": 0.08122539520263672, | |
| "learning_rate": 0.0005988023952095808, | |
| "loss": 1.2048, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.08748337626457214, | |
| "learning_rate": 0.0005953806672369548, | |
| "loss": 1.2377, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 2.10752, | |
| "grad_norm": 0.08610134571790695, | |
| "learning_rate": 0.0005919589392643285, | |
| "loss": 1.2312, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 2.11264, | |
| "grad_norm": 0.09176385402679443, | |
| "learning_rate": 0.0005885372112917023, | |
| "loss": 1.234, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 2.11776, | |
| "grad_norm": 0.08381321281194687, | |
| "learning_rate": 0.0005851154833190762, | |
| "loss": 1.2369, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 2.12288, | |
| "grad_norm": 0.08841554820537567, | |
| "learning_rate": 0.00058169375534645, | |
| "loss": 1.2283, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.09985855966806412, | |
| "learning_rate": 0.0005782720273738238, | |
| "loss": 1.2144, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 2.13312, | |
| "grad_norm": 0.09444481134414673, | |
| "learning_rate": 0.0005748502994011976, | |
| "loss": 1.2503, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 2.13824, | |
| "grad_norm": 0.08858395367860794, | |
| "learning_rate": 0.0005714285714285714, | |
| "loss": 1.2497, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 2.14336, | |
| "grad_norm": 0.08610956370830536, | |
| "learning_rate": 0.0005680068434559452, | |
| "loss": 1.2444, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 2.14848, | |
| "grad_norm": 0.08294233679771423, | |
| "learning_rate": 0.000564585115483319, | |
| "loss": 1.2388, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.1536, | |
| "grad_norm": 0.08838624507188797, | |
| "learning_rate": 0.000561163387510693, | |
| "loss": 1.2353, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 2.15872, | |
| "grad_norm": 0.07888966798782349, | |
| "learning_rate": 0.0005577416595380668, | |
| "loss": 1.2085, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 2.16384, | |
| "grad_norm": 0.08408137410879135, | |
| "learning_rate": 0.0005543199315654406, | |
| "loss": 1.1922, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 2.16896, | |
| "grad_norm": 0.08245803415775299, | |
| "learning_rate": 0.0005508982035928143, | |
| "loss": 1.2478, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 2.17408, | |
| "grad_norm": 0.07858633249998093, | |
| "learning_rate": 0.0005474764756201882, | |
| "loss": 1.2203, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.1002994254231453, | |
| "learning_rate": 0.000544054747647562, | |
| "loss": 1.2295, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 2.18432, | |
| "grad_norm": 0.08837361633777618, | |
| "learning_rate": 0.0005406330196749358, | |
| "loss": 1.238, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 2.18944, | |
| "grad_norm": 0.09690374881029129, | |
| "learning_rate": 0.0005372112917023097, | |
| "loss": 1.2301, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 2.19456, | |
| "grad_norm": 0.1030053198337555, | |
| "learning_rate": 0.0005337895637296835, | |
| "loss": 1.2365, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 2.19968, | |
| "grad_norm": 0.08212369680404663, | |
| "learning_rate": 0.0005303678357570573, | |
| "loss": 1.2325, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.2048, | |
| "grad_norm": 0.10352316498756409, | |
| "learning_rate": 0.0005269461077844312, | |
| "loss": 1.2341, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 2.20992, | |
| "grad_norm": 0.08346536755561829, | |
| "learning_rate": 0.000523524379811805, | |
| "loss": 1.2237, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 2.21504, | |
| "grad_norm": 0.11156366020441055, | |
| "learning_rate": 0.0005201026518391788, | |
| "loss": 1.2325, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 2.22016, | |
| "grad_norm": 0.10335463285446167, | |
| "learning_rate": 0.0005166809238665526, | |
| "loss": 1.2258, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 2.22528, | |
| "grad_norm": 0.09020327776670456, | |
| "learning_rate": 0.0005132591958939265, | |
| "loss": 1.2305, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.2304, | |
| "grad_norm": 0.09839983284473419, | |
| "learning_rate": 0.0005098374679213003, | |
| "loss": 1.2529, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 2.23552, | |
| "grad_norm": 0.1528375744819641, | |
| "learning_rate": 0.000506415739948674, | |
| "loss": 1.2365, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 2.24064, | |
| "grad_norm": 0.08589835464954376, | |
| "learning_rate": 0.0005029940119760479, | |
| "loss": 1.2407, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 2.24576, | |
| "grad_norm": 0.09676992893218994, | |
| "learning_rate": 0.0004995722840034218, | |
| "loss": 1.2244, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 2.25088, | |
| "grad_norm": 0.09816568344831467, | |
| "learning_rate": 0.0004961505560307955, | |
| "loss": 1.2439, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.09415734559297562, | |
| "learning_rate": 0.0004927288280581693, | |
| "loss": 1.2329, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 2.26112, | |
| "grad_norm": 0.0947481021285057, | |
| "learning_rate": 0.0004893071000855432, | |
| "loss": 1.2161, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 2.26624, | |
| "grad_norm": 0.09034735709428787, | |
| "learning_rate": 0.00048588537211291706, | |
| "loss": 1.25, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 2.27136, | |
| "grad_norm": 0.09715861082077026, | |
| "learning_rate": 0.0004824636441402909, | |
| "loss": 1.2536, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 2.27648, | |
| "grad_norm": 0.08793841302394867, | |
| "learning_rate": 0.00047904191616766467, | |
| "loss": 1.2425, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.2816, | |
| "grad_norm": 0.08474820852279663, | |
| "learning_rate": 0.0004756201881950385, | |
| "loss": 1.2409, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.28672, | |
| "grad_norm": 0.0954166129231453, | |
| "learning_rate": 0.0004721984602224123, | |
| "loss": 1.2292, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 2.29184, | |
| "grad_norm": 0.08861096948385239, | |
| "learning_rate": 0.00046877673224978616, | |
| "loss": 1.2509, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 2.29696, | |
| "grad_norm": 0.08726037293672562, | |
| "learning_rate": 0.00046535500427716, | |
| "loss": 1.2471, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 2.30208, | |
| "grad_norm": 0.10455521196126938, | |
| "learning_rate": 0.0004619332763045338, | |
| "loss": 1.2025, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.3072, | |
| "grad_norm": 0.09069357812404633, | |
| "learning_rate": 0.0004585115483319076, | |
| "loss": 1.2267, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 2.31232, | |
| "grad_norm": 0.09878482669591904, | |
| "learning_rate": 0.0004550898203592814, | |
| "loss": 1.2276, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 2.31744, | |
| "grad_norm": 0.1100214272737503, | |
| "learning_rate": 0.0004516680923866553, | |
| "loss": 1.2264, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 2.32256, | |
| "grad_norm": 0.09539143741130829, | |
| "learning_rate": 0.0004482463644140291, | |
| "loss": 1.2298, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 2.32768, | |
| "grad_norm": 0.0885070264339447, | |
| "learning_rate": 0.0004448246364414029, | |
| "loss": 1.2108, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.08761118352413177, | |
| "learning_rate": 0.00044140290846877674, | |
| "loss": 1.2393, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 2.33792, | |
| "grad_norm": 0.09102723747491837, | |
| "learning_rate": 0.00043798118049615057, | |
| "loss": 1.2372, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 2.3430400000000002, | |
| "grad_norm": 0.08345375210046768, | |
| "learning_rate": 0.0004345594525235244, | |
| "loss": 1.2152, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 2.34816, | |
| "grad_norm": 0.0871756300330162, | |
| "learning_rate": 0.00043113772455089823, | |
| "loss": 1.2539, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 2.35328, | |
| "grad_norm": 0.08344704657793045, | |
| "learning_rate": 0.00042771599657827206, | |
| "loss": 1.2553, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.3584, | |
| "grad_norm": 0.09925805032253265, | |
| "learning_rate": 0.00042429426860564584, | |
| "loss": 1.2407, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 2.36352, | |
| "grad_norm": 0.08815500140190125, | |
| "learning_rate": 0.00042087254063301967, | |
| "loss": 1.2165, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 2.36864, | |
| "grad_norm": 0.08601918071508408, | |
| "learning_rate": 0.00041745081266039355, | |
| "loss": 1.2335, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 2.37376, | |
| "grad_norm": 0.08933749049901962, | |
| "learning_rate": 0.0004140290846877673, | |
| "loss": 1.2226, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 2.37888, | |
| "grad_norm": 0.09617882966995239, | |
| "learning_rate": 0.00041060735671514116, | |
| "loss": 1.2232, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 0.08219394832849503, | |
| "learning_rate": 0.000407185628742515, | |
| "loss": 1.2143, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 2.38912, | |
| "grad_norm": 0.08431462943553925, | |
| "learning_rate": 0.00040376390076988876, | |
| "loss": 1.2416, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 2.39424, | |
| "grad_norm": 0.09297817200422287, | |
| "learning_rate": 0.00040034217279726265, | |
| "loss": 1.2151, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 2.39936, | |
| "grad_norm": 0.08247403055429459, | |
| "learning_rate": 0.0003969204448246365, | |
| "loss": 1.2291, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 2.40448, | |
| "grad_norm": 0.09852425754070282, | |
| "learning_rate": 0.00039349871685201025, | |
| "loss": 1.2351, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.08745532482862473, | |
| "learning_rate": 0.0003900769888793841, | |
| "loss": 1.2343, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 2.41472, | |
| "grad_norm": 0.0770939365029335, | |
| "learning_rate": 0.0003866552609067579, | |
| "loss": 1.229, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 2.4198399999999998, | |
| "grad_norm": 0.08143922686576843, | |
| "learning_rate": 0.00038323353293413174, | |
| "loss": 1.2168, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 2.42496, | |
| "grad_norm": 0.0852786973118782, | |
| "learning_rate": 0.0003798118049615056, | |
| "loss": 1.237, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 2.4300800000000002, | |
| "grad_norm": 0.09687593579292297, | |
| "learning_rate": 0.0003763900769888794, | |
| "loss": 1.2508, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.4352, | |
| "grad_norm": 0.09721924364566803, | |
| "learning_rate": 0.0003729683490162532, | |
| "loss": 1.2524, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 2.44032, | |
| "grad_norm": 0.09008084982633591, | |
| "learning_rate": 0.000369546621043627, | |
| "loss": 1.244, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 2.44544, | |
| "grad_norm": 0.0971183255314827, | |
| "learning_rate": 0.0003661248930710009, | |
| "loss": 1.2339, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 2.45056, | |
| "grad_norm": 0.09916377067565918, | |
| "learning_rate": 0.0003627031650983747, | |
| "loss": 1.2361, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 2.45568, | |
| "grad_norm": 0.08370574563741684, | |
| "learning_rate": 0.0003592814371257485, | |
| "loss": 1.2155, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.4608, | |
| "grad_norm": 0.08623719960451126, | |
| "learning_rate": 0.00035585970915312233, | |
| "loss": 1.2165, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 2.46592, | |
| "grad_norm": 0.07589856535196304, | |
| "learning_rate": 0.00035243798118049616, | |
| "loss": 1.2243, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 2.47104, | |
| "grad_norm": 0.07783632725477219, | |
| "learning_rate": 0.00034901625320787, | |
| "loss": 1.2296, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 2.47616, | |
| "grad_norm": 0.08424780517816544, | |
| "learning_rate": 0.0003455945252352438, | |
| "loss": 1.2312, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 2.48128, | |
| "grad_norm": 0.07945290952920914, | |
| "learning_rate": 0.00034217279726261765, | |
| "loss": 1.217, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.0799039751291275, | |
| "learning_rate": 0.0003387510692899914, | |
| "loss": 1.2201, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 2.49152, | |
| "grad_norm": 0.09106255322694778, | |
| "learning_rate": 0.00033532934131736525, | |
| "loss": 1.2159, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 2.49664, | |
| "grad_norm": 0.08736411482095718, | |
| "learning_rate": 0.00033190761334473914, | |
| "loss": 1.2298, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 2.50176, | |
| "grad_norm": 0.08567455410957336, | |
| "learning_rate": 0.0003284858853721129, | |
| "loss": 1.221, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 2.5068799999999998, | |
| "grad_norm": 0.10240741819143295, | |
| "learning_rate": 0.00032506415739948674, | |
| "loss": 1.2296, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.0851578339934349, | |
| "learning_rate": 0.0003216424294268606, | |
| "loss": 1.233, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 2.5171200000000002, | |
| "grad_norm": 0.08598732203245163, | |
| "learning_rate": 0.00031822070145423435, | |
| "loss": 1.1937, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 2.52224, | |
| "grad_norm": 0.08547403663396835, | |
| "learning_rate": 0.00031479897348160823, | |
| "loss": 1.2158, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 2.52736, | |
| "grad_norm": 0.08678396046161652, | |
| "learning_rate": 0.00031137724550898206, | |
| "loss": 1.2113, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 2.53248, | |
| "grad_norm": 0.08527970314025879, | |
| "learning_rate": 0.00030795551753635584, | |
| "loss": 1.2508, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.5376, | |
| "grad_norm": 0.08052323758602142, | |
| "learning_rate": 0.00030453378956372967, | |
| "loss": 1.2324, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 2.54272, | |
| "grad_norm": 0.09333521127700806, | |
| "learning_rate": 0.0003011120615911035, | |
| "loss": 1.2381, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 2.54784, | |
| "grad_norm": 0.08944050967693329, | |
| "learning_rate": 0.0002976903336184774, | |
| "loss": 1.2417, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 2.55296, | |
| "grad_norm": 0.0959937795996666, | |
| "learning_rate": 0.00029426860564585116, | |
| "loss": 1.2346, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 2.55808, | |
| "grad_norm": 0.08061899244785309, | |
| "learning_rate": 0.000290846877673225, | |
| "loss": 1.2377, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.08358662575483322, | |
| "learning_rate": 0.0002874251497005988, | |
| "loss": 1.2432, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 2.56832, | |
| "grad_norm": 0.08278947323560715, | |
| "learning_rate": 0.0002840034217279726, | |
| "loss": 1.222, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 2.5734399999999997, | |
| "grad_norm": 0.08546797931194305, | |
| "learning_rate": 0.0002805816937553465, | |
| "loss": 1.2379, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 2.57856, | |
| "grad_norm": 0.08106214553117752, | |
| "learning_rate": 0.0002771599657827203, | |
| "loss": 1.217, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 2.58368, | |
| "grad_norm": 0.08738037943840027, | |
| "learning_rate": 0.0002737382378100941, | |
| "loss": 1.2202, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.5888, | |
| "grad_norm": 0.07763461023569107, | |
| "learning_rate": 0.0002703165098374679, | |
| "loss": 1.2335, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 2.59392, | |
| "grad_norm": 0.09023375809192657, | |
| "learning_rate": 0.00026689478186484175, | |
| "loss": 1.2557, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 2.59904, | |
| "grad_norm": 0.08477311581373215, | |
| "learning_rate": 0.0002634730538922156, | |
| "loss": 1.23, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 2.6041600000000003, | |
| "grad_norm": 0.0802718847990036, | |
| "learning_rate": 0.0002600513259195894, | |
| "loss": 1.2502, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 2.60928, | |
| "grad_norm": 0.08197642862796783, | |
| "learning_rate": 0.00025662959794696324, | |
| "loss": 1.2117, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.6144, | |
| "grad_norm": 0.07977724820375443, | |
| "learning_rate": 0.000253207869974337, | |
| "loss": 1.2223, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 2.61952, | |
| "grad_norm": 0.084455206990242, | |
| "learning_rate": 0.0002497861420017109, | |
| "loss": 1.2296, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 2.62464, | |
| "grad_norm": 0.08407705277204514, | |
| "learning_rate": 0.00024636441402908467, | |
| "loss": 1.2466, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 2.62976, | |
| "grad_norm": 0.09497237205505371, | |
| "learning_rate": 0.00024294268605645853, | |
| "loss": 1.2204, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 2.63488, | |
| "grad_norm": 0.07946959137916565, | |
| "learning_rate": 0.00023952095808383233, | |
| "loss": 1.2293, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.08926168084144592, | |
| "learning_rate": 0.00023609923011120616, | |
| "loss": 1.2303, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 2.64512, | |
| "grad_norm": 0.08781218528747559, | |
| "learning_rate": 0.00023267750213858, | |
| "loss": 1.2215, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 2.65024, | |
| "grad_norm": 0.10436686873435974, | |
| "learning_rate": 0.0002292557741659538, | |
| "loss": 1.2446, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 2.65536, | |
| "grad_norm": 0.08655078709125519, | |
| "learning_rate": 0.00022583404619332765, | |
| "loss": 1.1889, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 2.6604799999999997, | |
| "grad_norm": 0.08525776863098145, | |
| "learning_rate": 0.00022241231822070146, | |
| "loss": 1.2246, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.6656, | |
| "grad_norm": 0.08515007048845291, | |
| "learning_rate": 0.00021899059024807529, | |
| "loss": 1.2083, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 2.67072, | |
| "grad_norm": 0.08326171338558197, | |
| "learning_rate": 0.00021556886227544912, | |
| "loss": 1.2463, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 2.67584, | |
| "grad_norm": 0.07543444633483887, | |
| "learning_rate": 0.00021214713430282292, | |
| "loss": 1.2229, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 2.68096, | |
| "grad_norm": 0.08731929957866669, | |
| "learning_rate": 0.00020872540633019678, | |
| "loss": 1.2431, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 2.68608, | |
| "grad_norm": 0.0846877247095108, | |
| "learning_rate": 0.00020530367835757058, | |
| "loss": 1.2276, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.6912000000000003, | |
| "grad_norm": 0.08184527605772018, | |
| "learning_rate": 0.00020188195038494438, | |
| "loss": 1.2151, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 2.69632, | |
| "grad_norm": 0.0830821543931961, | |
| "learning_rate": 0.00019846022241231824, | |
| "loss": 1.23, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.70144, | |
| "grad_norm": 0.07801831513643265, | |
| "learning_rate": 0.00019503849443969204, | |
| "loss": 1.2237, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 2.70656, | |
| "grad_norm": 0.07970487326383591, | |
| "learning_rate": 0.00019161676646706587, | |
| "loss": 1.2296, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.71168, | |
| "grad_norm": 0.08674521744251251, | |
| "learning_rate": 0.0001881950384944397, | |
| "loss": 1.2416, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.0784011259675026, | |
| "learning_rate": 0.0001847733105218135, | |
| "loss": 1.2174, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 2.72192, | |
| "grad_norm": 0.08390358090400696, | |
| "learning_rate": 0.00018135158254918736, | |
| "loss": 1.2221, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 2.72704, | |
| "grad_norm": 0.08034460991621017, | |
| "learning_rate": 0.00017792985457656116, | |
| "loss": 1.2367, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 2.73216, | |
| "grad_norm": 0.0805404931306839, | |
| "learning_rate": 0.000174508126603935, | |
| "loss": 1.2228, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 2.73728, | |
| "grad_norm": 0.10428917407989502, | |
| "learning_rate": 0.00017108639863130882, | |
| "loss": 1.2331, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.7424, | |
| "grad_norm": 0.07880023866891861, | |
| "learning_rate": 0.00016766467065868263, | |
| "loss": 1.2151, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 2.7475199999999997, | |
| "grad_norm": 0.08330074697732925, | |
| "learning_rate": 0.00016424294268605646, | |
| "loss": 1.201, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 2.75264, | |
| "grad_norm": 0.08316068351268768, | |
| "learning_rate": 0.0001608212147134303, | |
| "loss": 1.2206, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 2.75776, | |
| "grad_norm": 0.08193733543157578, | |
| "learning_rate": 0.00015739948674080412, | |
| "loss": 1.2133, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 2.76288, | |
| "grad_norm": 0.0799107700586319, | |
| "learning_rate": 0.00015397775876817792, | |
| "loss": 1.2202, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 0.07523773610591888, | |
| "learning_rate": 0.00015055603079555175, | |
| "loss": 1.2088, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 2.77312, | |
| "grad_norm": 0.07782167941331863, | |
| "learning_rate": 0.00014713430282292558, | |
| "loss": 1.2191, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 2.7782400000000003, | |
| "grad_norm": 0.0814930647611618, | |
| "learning_rate": 0.0001437125748502994, | |
| "loss": 1.2096, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 2.78336, | |
| "grad_norm": 0.07619259506464005, | |
| "learning_rate": 0.00014029084687767324, | |
| "loss": 1.2297, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 2.78848, | |
| "grad_norm": 0.08152459561824799, | |
| "learning_rate": 0.00013686911890504704, | |
| "loss": 1.2281, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.08513466268777847, | |
| "learning_rate": 0.00013344739093242087, | |
| "loss": 1.2183, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 2.79872, | |
| "grad_norm": 0.07648808509111404, | |
| "learning_rate": 0.0001300256629597947, | |
| "loss": 1.2171, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 2.80384, | |
| "grad_norm": 0.08608166873455048, | |
| "learning_rate": 0.0001266039349871685, | |
| "loss": 1.233, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 2.80896, | |
| "grad_norm": 0.08903096616268158, | |
| "learning_rate": 0.00012318220701454234, | |
| "loss": 1.2329, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 2.81408, | |
| "grad_norm": 0.07557443529367447, | |
| "learning_rate": 0.00011976047904191617, | |
| "loss": 1.2333, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.8192, | |
| "grad_norm": 0.0815119668841362, | |
| "learning_rate": 0.00011633875106929, | |
| "loss": 1.2353, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 2.82432, | |
| "grad_norm": 0.08905310928821564, | |
| "learning_rate": 0.00011291702309666383, | |
| "loss": 1.2255, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 2.82944, | |
| "grad_norm": 0.08789879828691483, | |
| "learning_rate": 0.00010949529512403764, | |
| "loss": 1.2139, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 2.8345599999999997, | |
| "grad_norm": 0.08430198580026627, | |
| "learning_rate": 0.00010607356715141146, | |
| "loss": 1.229, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 2.83968, | |
| "grad_norm": 0.07988926768302917, | |
| "learning_rate": 0.00010265183917878529, | |
| "loss": 1.2316, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.8448, | |
| "grad_norm": 0.08727908134460449, | |
| "learning_rate": 9.923011120615912e-05, | |
| "loss": 1.2237, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 2.84992, | |
| "grad_norm": 0.10279367119073868, | |
| "learning_rate": 9.580838323353294e-05, | |
| "loss": 1.2565, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 2.85504, | |
| "grad_norm": 0.08122528344392776, | |
| "learning_rate": 9.238665526090675e-05, | |
| "loss": 1.2353, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 2.86016, | |
| "grad_norm": 0.09036324173212051, | |
| "learning_rate": 8.896492728828058e-05, | |
| "loss": 1.2337, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 2.8652800000000003, | |
| "grad_norm": 0.07413888722658157, | |
| "learning_rate": 8.554319931565441e-05, | |
| "loss": 1.2367, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.08744188398122787, | |
| "learning_rate": 8.212147134302823e-05, | |
| "loss": 1.234, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 2.87552, | |
| "grad_norm": 0.07657689601182938, | |
| "learning_rate": 7.869974337040206e-05, | |
| "loss": 1.2094, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 2.88064, | |
| "grad_norm": 0.09120559692382812, | |
| "learning_rate": 7.527801539777588e-05, | |
| "loss": 1.2473, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 2.88576, | |
| "grad_norm": 0.08199866861104965, | |
| "learning_rate": 7.18562874251497e-05, | |
| "loss": 1.2384, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 2.89088, | |
| "grad_norm": 0.07703917473554611, | |
| "learning_rate": 6.843455945252352e-05, | |
| "loss": 1.2197, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.08120223879814148, | |
| "learning_rate": 6.501283147989735e-05, | |
| "loss": 1.2343, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 2.90112, | |
| "grad_norm": 0.08173457533121109, | |
| "learning_rate": 6.159110350727117e-05, | |
| "loss": 1.2422, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 2.90624, | |
| "grad_norm": 0.08017323166131973, | |
| "learning_rate": 5.8169375534645e-05, | |
| "loss": 1.1985, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 2.91136, | |
| "grad_norm": 0.09028081595897675, | |
| "learning_rate": 5.474764756201882e-05, | |
| "loss": 1.1902, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 2.91648, | |
| "grad_norm": 0.07673865556716919, | |
| "learning_rate": 5.1325919589392645e-05, | |
| "loss": 1.2292, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.9215999999999998, | |
| "grad_norm": 0.08590974658727646, | |
| "learning_rate": 4.790419161676647e-05, | |
| "loss": 1.2227, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 2.92672, | |
| "grad_norm": 0.07928025722503662, | |
| "learning_rate": 4.448246364414029e-05, | |
| "loss": 1.2426, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 2.9318400000000002, | |
| "grad_norm": 0.07864856719970703, | |
| "learning_rate": 4.1060735671514114e-05, | |
| "loss": 1.2476, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 2.93696, | |
| "grad_norm": 0.08154473453760147, | |
| "learning_rate": 3.763900769888794e-05, | |
| "loss": 1.2081, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 2.94208, | |
| "grad_norm": 0.07814584672451019, | |
| "learning_rate": 3.421727972626176e-05, | |
| "loss": 1.2463, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.07772421091794968, | |
| "learning_rate": 3.0795551753635584e-05, | |
| "loss": 1.2447, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 2.9523200000000003, | |
| "grad_norm": 0.08610265702009201, | |
| "learning_rate": 2.737382378100941e-05, | |
| "loss": 1.226, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 2.95744, | |
| "grad_norm": 0.086619071662426, | |
| "learning_rate": 2.3952095808383234e-05, | |
| "loss": 1.2616, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 2.96256, | |
| "grad_norm": 0.07644681632518768, | |
| "learning_rate": 2.0530367835757057e-05, | |
| "loss": 1.2244, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 2.96768, | |
| "grad_norm": 0.08778993040323257, | |
| "learning_rate": 1.710863986313088e-05, | |
| "loss": 1.2259, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.9728, | |
| "grad_norm": 0.0812673419713974, | |
| "learning_rate": 1.3686911890504705e-05, | |
| "loss": 1.2081, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 2.97792, | |
| "grad_norm": 0.07923007756471634, | |
| "learning_rate": 1.0265183917878529e-05, | |
| "loss": 1.2297, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 2.98304, | |
| "grad_norm": 0.07516805827617645, | |
| "learning_rate": 6.843455945252353e-06, | |
| "loss": 1.2277, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 2.98816, | |
| "grad_norm": 0.08043860644102097, | |
| "learning_rate": 3.4217279726261763e-06, | |
| "loss": 1.2087, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 2.99328, | |
| "grad_norm": 0.07992921024560928, | |
| "learning_rate": 0.0, | |
| "loss": 1.2479, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.99328, | |
| "step": 1170, | |
| "total_flos": 1.3204661001922806e+19, | |
| "train_loss": 1.2500826900840825, | |
| "train_runtime": 40705.2136, | |
| "train_samples_per_second": 7.37, | |
| "train_steps_per_second": 0.029 | |
| }, | |
| { | |
| "epoch": 2.99328, | |
| "eval_loss": 1.2431308031082153, | |
| "eval_runtime": 280.041, | |
| "eval_samples_per_second": 8.927, | |
| "eval_steps_per_second": 1.118, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.99328, | |
| "eval_loss": 1.2251627445220947, | |
| "eval_runtime": 300.2276, | |
| "eval_samples_per_second": 8.327, | |
| "eval_steps_per_second": 1.043, | |
| "step": 1170 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 1170, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3204661001922806e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |