diff --git "a/g2p/trainer_state.json" "b/g2p/trainer_state.json" new file mode 100644--- /dev/null +++ "b/g2p/trainer_state.json" @@ -0,0 +1,33458 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.3056, + "eval_steps": 500, + "global_step": 95500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.4e-05, + "grad_norm": 19.984381915653476, + "learning_rate": 9.999999988770584e-05, + "loss": 11.2743, + "step": 20 + }, + { + "epoch": 0.000128, + "grad_norm": 5.487518887641428, + "learning_rate": 9.999999955082334e-05, + "loss": 3.8589, + "step": 40 + }, + { + "epoch": 0.000192, + "grad_norm": 4.286714311160932, + "learning_rate": 9.999999898935252e-05, + "loss": 3.356, + "step": 60 + }, + { + "epoch": 0.000256, + "grad_norm": 3.3802473703861526, + "learning_rate": 9.999999820329337e-05, + "loss": 3.1579, + "step": 80 + }, + { + "epoch": 0.00032, + "grad_norm": 5.4682032716968445, + "learning_rate": 9.999999719264588e-05, + "loss": 2.9438, + "step": 100 + }, + { + "epoch": 0.000384, + "grad_norm": 4.188895453976329, + "learning_rate": 9.999999595741009e-05, + "loss": 2.5031, + "step": 120 + }, + { + "epoch": 0.000448, + "grad_norm": 5.127885297805202, + "learning_rate": 9.9999994497586e-05, + "loss": 1.9916, + "step": 140 + }, + { + "epoch": 0.000512, + "grad_norm": 8.299321191456816, + "learning_rate": 9.999999281317358e-05, + "loss": 1.7096, + "step": 160 + }, + { + "epoch": 0.000576, + "grad_norm": 9.847377491500083, + "learning_rate": 9.999999090417287e-05, + "loss": 1.1331, + "step": 180 + }, + { + "epoch": 0.00064, + "grad_norm": 2.9451663138487314, + "learning_rate": 9.999998877058386e-05, + "loss": 0.5498, + "step": 200 + }, + { + "epoch": 0.000704, + "grad_norm": 2.6375340984228384, + "learning_rate": 9.999998641240658e-05, + "loss": 0.3297, + "step": 220 + }, + { + "epoch": 0.000768, + "grad_norm": 2.1657603462311803, + "learning_rate": 9.999998382964102e-05, + "loss": 0.2479, + "step": 240 + }, + { + "epoch": 0.000832, + "grad_norm": 1.6084174334554693, + "learning_rate": 9.999998102228721e-05, + "loss": 0.1949, + "step": 260 + }, + { + "epoch": 0.000896, + "grad_norm": 1.3618933389068795, + "learning_rate": 9.999997799034515e-05, + "loss": 0.1743, + "step": 280 + }, + { + "epoch": 0.00096, + "grad_norm": 1.4689364448955022, + "learning_rate": 9.999997473381486e-05, + "loss": 0.1629, + "step": 300 + }, + { + "epoch": 0.001024, + "grad_norm": 1.0551072709393718, + "learning_rate": 9.999997125269635e-05, + "loss": 0.1512, + "step": 320 + }, + { + "epoch": 0.001088, + "grad_norm": 1.303626280533112, + "learning_rate": 9.999996754698965e-05, + "loss": 0.1312, + "step": 340 + }, + { + "epoch": 0.001152, + "grad_norm": 0.9015109235230854, + "learning_rate": 9.999996361669474e-05, + "loss": 0.1236, + "step": 360 + }, + { + "epoch": 0.001216, + "grad_norm": 0.795632156046167, + "learning_rate": 9.999995946181169e-05, + "loss": 0.1114, + "step": 380 + }, + { + "epoch": 0.00128, + "grad_norm": 0.8853577606070548, + "learning_rate": 9.999995508234047e-05, + "loss": 0.1105, + "step": 400 + }, + { + "epoch": 0.001344, + "grad_norm": 0.7217127786071534, + "learning_rate": 9.999995047828114e-05, + "loss": 0.0972, + "step": 420 + }, + { + "epoch": 0.001408, + "grad_norm": 0.7517999410008976, + "learning_rate": 9.999994564963369e-05, + "loss": 0.0997, + "step": 440 + }, + { + "epoch": 0.001472, + "grad_norm": 0.8263726057198254, + "learning_rate": 9.999994059639815e-05, + "loss": 0.0942, + "step": 460 + }, + { + "epoch": 0.001536, + "grad_norm": 0.8537447532924352, + "learning_rate": 9.999993531857455e-05, + "loss": 0.0849, + "step": 480 + }, + { + "epoch": 0.0016, + "grad_norm": 0.7338816974677734, + "learning_rate": 9.999992981616291e-05, + "loss": 0.0811, + "step": 500 + }, + { + "epoch": 0.001664, + "grad_norm": 0.5769006241899075, + "learning_rate": 9.999992408916325e-05, + "loss": 0.0832, + "step": 520 + }, + { + "epoch": 0.001728, + "grad_norm": 0.5731625699804362, + "learning_rate": 9.99999181375756e-05, + "loss": 0.0822, + "step": 540 + }, + { + "epoch": 0.001792, + "grad_norm": 1.4189600258630701, + "learning_rate": 9.99999119614e-05, + "loss": 0.0842, + "step": 560 + }, + { + "epoch": 0.001856, + "grad_norm": 0.6998019471575104, + "learning_rate": 9.999990556063644e-05, + "loss": 0.0753, + "step": 580 + }, + { + "epoch": 0.00192, + "grad_norm": 0.6720273927543539, + "learning_rate": 9.999989893528499e-05, + "loss": 0.07, + "step": 600 + }, + { + "epoch": 0.001984, + "grad_norm": 0.7288495014910453, + "learning_rate": 9.999989208534566e-05, + "loss": 0.0674, + "step": 620 + }, + { + "epoch": 0.002048, + "grad_norm": 0.4649112257401437, + "learning_rate": 9.999988501081847e-05, + "loss": 0.0677, + "step": 640 + }, + { + "epoch": 0.002112, + "grad_norm": 0.46932076212536666, + "learning_rate": 9.999987771170348e-05, + "loss": 0.0697, + "step": 660 + }, + { + "epoch": 0.002176, + "grad_norm": 0.48148538827058973, + "learning_rate": 9.999987018800069e-05, + "loss": 0.0606, + "step": 680 + }, + { + "epoch": 0.00224, + "grad_norm": 0.4858179133601085, + "learning_rate": 9.999986243971019e-05, + "loss": 0.0725, + "step": 700 + }, + { + "epoch": 0.002304, + "grad_norm": 0.38680134681776157, + "learning_rate": 9.999985446683196e-05, + "loss": 0.0681, + "step": 720 + }, + { + "epoch": 0.002368, + "grad_norm": 0.416560992524678, + "learning_rate": 9.999984626936603e-05, + "loss": 0.0586, + "step": 740 + }, + { + "epoch": 0.002432, + "grad_norm": 0.524010587110687, + "learning_rate": 9.999983784731249e-05, + "loss": 0.0689, + "step": 760 + }, + { + "epoch": 0.002496, + "grad_norm": 0.4556443223311926, + "learning_rate": 9.999982920067132e-05, + "loss": 0.0625, + "step": 780 + }, + { + "epoch": 0.00256, + "grad_norm": 0.364406963490877, + "learning_rate": 9.99998203294426e-05, + "loss": 0.0672, + "step": 800 + }, + { + "epoch": 0.002624, + "grad_norm": 0.553135092718259, + "learning_rate": 9.999981123362637e-05, + "loss": 0.0624, + "step": 820 + }, + { + "epoch": 0.002688, + "grad_norm": 0.42640017810938874, + "learning_rate": 9.999980191322262e-05, + "loss": 0.0608, + "step": 840 + }, + { + "epoch": 0.002752, + "grad_norm": 0.5632745358471698, + "learning_rate": 9.999979236823146e-05, + "loss": 0.0583, + "step": 860 + }, + { + "epoch": 0.002816, + "grad_norm": 0.4202387900044088, + "learning_rate": 9.999978259865289e-05, + "loss": 0.0615, + "step": 880 + }, + { + "epoch": 0.00288, + "grad_norm": 0.4700746106046101, + "learning_rate": 9.999977260448697e-05, + "loss": 0.062, + "step": 900 + }, + { + "epoch": 0.002944, + "grad_norm": 0.4710498557018385, + "learning_rate": 9.999976238573373e-05, + "loss": 0.0572, + "step": 920 + }, + { + "epoch": 0.003008, + "grad_norm": 0.45091889977134386, + "learning_rate": 9.999975194239324e-05, + "loss": 0.0614, + "step": 940 + }, + { + "epoch": 0.003072, + "grad_norm": 0.36502040200710034, + "learning_rate": 9.999974127446552e-05, + "loss": 0.0583, + "step": 960 + }, + { + "epoch": 0.003136, + "grad_norm": 0.4407353477570368, + "learning_rate": 9.999973038195064e-05, + "loss": 0.0656, + "step": 980 + }, + { + "epoch": 0.0032, + "grad_norm": 0.42026896152067805, + "learning_rate": 9.999971926484864e-05, + "loss": 0.0577, + "step": 1000 + }, + { + "epoch": 0.003264, + "grad_norm": 0.3680496837271321, + "learning_rate": 9.999970792315956e-05, + "loss": 0.0603, + "step": 1020 + }, + { + "epoch": 0.003328, + "grad_norm": 0.5256310833583023, + "learning_rate": 9.999969635688347e-05, + "loss": 0.0578, + "step": 1040 + }, + { + "epoch": 0.003392, + "grad_norm": 0.4010549915932494, + "learning_rate": 9.999968456602041e-05, + "loss": 0.0535, + "step": 1060 + }, + { + "epoch": 0.003456, + "grad_norm": 0.3569301669948006, + "learning_rate": 9.999967255057043e-05, + "loss": 0.0544, + "step": 1080 + }, + { + "epoch": 0.00352, + "grad_norm": 0.4085529298881325, + "learning_rate": 9.99996603105336e-05, + "loss": 0.0593, + "step": 1100 + }, + { + "epoch": 0.003584, + "grad_norm": 0.5275335628342391, + "learning_rate": 9.999964784590996e-05, + "loss": 0.0578, + "step": 1120 + }, + { + "epoch": 0.003648, + "grad_norm": 0.330644175208153, + "learning_rate": 9.999963515669956e-05, + "loss": 0.0528, + "step": 1140 + }, + { + "epoch": 0.003712, + "grad_norm": 0.341857691501769, + "learning_rate": 9.99996222429025e-05, + "loss": 0.0492, + "step": 1160 + }, + { + "epoch": 0.003776, + "grad_norm": 0.43878658400878173, + "learning_rate": 9.999960910451878e-05, + "loss": 0.0499, + "step": 1180 + }, + { + "epoch": 0.00384, + "grad_norm": 0.4402535076549671, + "learning_rate": 9.999959574154849e-05, + "loss": 0.0532, + "step": 1200 + }, + { + "epoch": 0.003904, + "grad_norm": 0.45547637242325884, + "learning_rate": 9.999958215399168e-05, + "loss": 0.0659, + "step": 1220 + }, + { + "epoch": 0.003968, + "grad_norm": 0.5965254253569031, + "learning_rate": 9.999956834184843e-05, + "loss": 0.0566, + "step": 1240 + }, + { + "epoch": 0.004032, + "grad_norm": 0.4324405925069988, + "learning_rate": 9.999955430511876e-05, + "loss": 0.0538, + "step": 1260 + }, + { + "epoch": 0.004096, + "grad_norm": 0.3827993868853947, + "learning_rate": 9.999954004380278e-05, + "loss": 0.052, + "step": 1280 + }, + { + "epoch": 0.00416, + "grad_norm": 0.41059924493250727, + "learning_rate": 9.999952555790053e-05, + "loss": 0.0528, + "step": 1300 + }, + { + "epoch": 0.004224, + "grad_norm": 0.3697787761089116, + "learning_rate": 9.999951084741208e-05, + "loss": 0.0547, + "step": 1320 + }, + { + "epoch": 0.004288, + "grad_norm": 0.3518249551296642, + "learning_rate": 9.999949591233751e-05, + "loss": 0.0512, + "step": 1340 + }, + { + "epoch": 0.004352, + "grad_norm": 0.3966048202962381, + "learning_rate": 9.999948075267686e-05, + "loss": 0.0483, + "step": 1360 + }, + { + "epoch": 0.004416, + "grad_norm": 0.2992863917125513, + "learning_rate": 9.999946536843021e-05, + "loss": 0.0588, + "step": 1380 + }, + { + "epoch": 0.00448, + "grad_norm": 0.3642430181347924, + "learning_rate": 9.999944975959762e-05, + "loss": 0.0559, + "step": 1400 + }, + { + "epoch": 0.004544, + "grad_norm": 0.28428952786006667, + "learning_rate": 9.99994339261792e-05, + "loss": 0.0505, + "step": 1420 + }, + { + "epoch": 0.004608, + "grad_norm": 0.27542639692431337, + "learning_rate": 9.999941786817497e-05, + "loss": 0.0503, + "step": 1440 + }, + { + "epoch": 0.004672, + "grad_norm": 0.2893247944703916, + "learning_rate": 9.999940158558503e-05, + "loss": 0.0448, + "step": 1460 + }, + { + "epoch": 0.004736, + "grad_norm": 0.3867132858602596, + "learning_rate": 9.999938507840944e-05, + "loss": 0.0531, + "step": 1480 + }, + { + "epoch": 0.0048, + "grad_norm": 0.3123100305966905, + "learning_rate": 9.99993683466483e-05, + "loss": 0.0486, + "step": 1500 + }, + { + "epoch": 0.004864, + "grad_norm": 0.6860642934875139, + "learning_rate": 9.999935139030164e-05, + "loss": 0.0508, + "step": 1520 + }, + { + "epoch": 0.004928, + "grad_norm": 0.41895308691771405, + "learning_rate": 9.999933420936957e-05, + "loss": 0.0507, + "step": 1540 + }, + { + "epoch": 0.004992, + "grad_norm": 0.2621600433718587, + "learning_rate": 9.999931680385217e-05, + "loss": 0.0504, + "step": 1560 + }, + { + "epoch": 0.005056, + "grad_norm": 0.3815573262171931, + "learning_rate": 9.999929917374951e-05, + "loss": 0.05, + "step": 1580 + }, + { + "epoch": 0.00512, + "grad_norm": 0.37620492419561863, + "learning_rate": 9.999928131906165e-05, + "loss": 0.0487, + "step": 1600 + }, + { + "epoch": 0.005184, + "grad_norm": 0.2984053569602002, + "learning_rate": 9.99992632397887e-05, + "loss": 0.048, + "step": 1620 + }, + { + "epoch": 0.005248, + "grad_norm": 0.27743837838217916, + "learning_rate": 9.999924493593071e-05, + "loss": 0.0479, + "step": 1640 + }, + { + "epoch": 0.005312, + "grad_norm": 0.37400919550719863, + "learning_rate": 9.99992264074878e-05, + "loss": 0.0479, + "step": 1660 + }, + { + "epoch": 0.005376, + "grad_norm": 0.3646151423541133, + "learning_rate": 9.999920765446002e-05, + "loss": 0.0478, + "step": 1680 + }, + { + "epoch": 0.00544, + "grad_norm": 0.2530121165143808, + "learning_rate": 9.999918867684749e-05, + "loss": 0.0471, + "step": 1700 + }, + { + "epoch": 0.005504, + "grad_norm": 0.3064041571959589, + "learning_rate": 9.999916947465026e-05, + "loss": 0.0479, + "step": 1720 + }, + { + "epoch": 0.005568, + "grad_norm": 0.29995424831453693, + "learning_rate": 9.999915004786842e-05, + "loss": 0.0545, + "step": 1740 + }, + { + "epoch": 0.005632, + "grad_norm": 0.2579628747140138, + "learning_rate": 9.999913039650208e-05, + "loss": 0.0481, + "step": 1760 + }, + { + "epoch": 0.005696, + "grad_norm": 0.37679754143686134, + "learning_rate": 9.99991105205513e-05, + "loss": 0.0513, + "step": 1780 + }, + { + "epoch": 0.00576, + "grad_norm": 0.3487000068380941, + "learning_rate": 9.99990904200162e-05, + "loss": 0.0485, + "step": 1800 + }, + { + "epoch": 0.005824, + "grad_norm": 0.42128271498962744, + "learning_rate": 9.999907009489684e-05, + "loss": 0.0523, + "step": 1820 + }, + { + "epoch": 0.005888, + "grad_norm": 0.24481172509209137, + "learning_rate": 9.999904954519335e-05, + "loss": 0.0483, + "step": 1840 + }, + { + "epoch": 0.005952, + "grad_norm": 0.35512611115825016, + "learning_rate": 9.999902877090577e-05, + "loss": 0.0478, + "step": 1860 + }, + { + "epoch": 0.006016, + "grad_norm": 0.43287405676025453, + "learning_rate": 9.999900777203424e-05, + "loss": 0.0461, + "step": 1880 + }, + { + "epoch": 0.00608, + "grad_norm": 0.37065789856595455, + "learning_rate": 9.999898654857882e-05, + "loss": 0.0471, + "step": 1900 + }, + { + "epoch": 0.006144, + "grad_norm": 0.6122637202631629, + "learning_rate": 9.999896510053963e-05, + "loss": 0.0477, + "step": 1920 + }, + { + "epoch": 0.006208, + "grad_norm": 0.41343656699008435, + "learning_rate": 9.999894342791676e-05, + "loss": 0.0471, + "step": 1940 + }, + { + "epoch": 0.006272, + "grad_norm": 0.3880497393670191, + "learning_rate": 9.99989215307103e-05, + "loss": 0.0458, + "step": 1960 + }, + { + "epoch": 0.006336, + "grad_norm": 0.413446669357271, + "learning_rate": 9.999889940892036e-05, + "loss": 0.0471, + "step": 1980 + }, + { + "epoch": 0.0064, + "grad_norm": 0.37063142738634886, + "learning_rate": 9.999887706254702e-05, + "loss": 0.0461, + "step": 2000 + }, + { + "epoch": 0.006464, + "grad_norm": 0.3263873803790246, + "learning_rate": 9.999885449159041e-05, + "loss": 0.0433, + "step": 2020 + }, + { + "epoch": 0.006528, + "grad_norm": 0.2751689144898591, + "learning_rate": 9.99988316960506e-05, + "loss": 0.0472, + "step": 2040 + }, + { + "epoch": 0.006592, + "grad_norm": 0.26563699407934044, + "learning_rate": 9.999880867592771e-05, + "loss": 0.0457, + "step": 2060 + }, + { + "epoch": 0.006656, + "grad_norm": 0.3185404492461951, + "learning_rate": 9.999878543122183e-05, + "loss": 0.0491, + "step": 2080 + }, + { + "epoch": 0.00672, + "grad_norm": 0.37960846473683246, + "learning_rate": 9.99987619619331e-05, + "loss": 0.0469, + "step": 2100 + }, + { + "epoch": 0.006784, + "grad_norm": 0.44510790225490077, + "learning_rate": 9.999873826806158e-05, + "loss": 0.0486, + "step": 2120 + }, + { + "epoch": 0.006848, + "grad_norm": 0.30839445947280475, + "learning_rate": 9.99987143496074e-05, + "loss": 0.0485, + "step": 2140 + }, + { + "epoch": 0.006912, + "grad_norm": 0.37056855516657783, + "learning_rate": 9.999869020657066e-05, + "loss": 0.0511, + "step": 2160 + }, + { + "epoch": 0.006976, + "grad_norm": 0.33190594018897246, + "learning_rate": 9.999866583895148e-05, + "loss": 0.0509, + "step": 2180 + }, + { + "epoch": 0.00704, + "grad_norm": 0.34153839602614805, + "learning_rate": 9.999864124674995e-05, + "loss": 0.0488, + "step": 2200 + }, + { + "epoch": 0.007104, + "grad_norm": 0.2342491607350666, + "learning_rate": 9.999861642996621e-05, + "loss": 0.0443, + "step": 2220 + }, + { + "epoch": 0.007168, + "grad_norm": 0.27549764195771853, + "learning_rate": 9.999859138860035e-05, + "loss": 0.0458, + "step": 2240 + }, + { + "epoch": 0.007232, + "grad_norm": 0.2974959541271812, + "learning_rate": 9.999856612265247e-05, + "loss": 0.0473, + "step": 2260 + }, + { + "epoch": 0.007296, + "grad_norm": 0.41245560801263564, + "learning_rate": 9.99985406321227e-05, + "loss": 0.0463, + "step": 2280 + }, + { + "epoch": 0.00736, + "grad_norm": 0.2970723727113926, + "learning_rate": 9.999851491701117e-05, + "loss": 0.046, + "step": 2300 + }, + { + "epoch": 0.007424, + "grad_norm": 0.26443168057948846, + "learning_rate": 9.999848897731799e-05, + "loss": 0.0464, + "step": 2320 + }, + { + "epoch": 0.007488, + "grad_norm": 0.30004642759969996, + "learning_rate": 9.999846281304323e-05, + "loss": 0.0437, + "step": 2340 + }, + { + "epoch": 0.007552, + "grad_norm": 0.2159707528687582, + "learning_rate": 9.999843642418707e-05, + "loss": 0.0434, + "step": 2360 + }, + { + "epoch": 0.007616, + "grad_norm": 0.5966198334346415, + "learning_rate": 9.99984098107496e-05, + "loss": 0.0498, + "step": 2380 + }, + { + "epoch": 0.00768, + "grad_norm": 0.30750128328976645, + "learning_rate": 9.999838297273093e-05, + "loss": 0.0506, + "step": 2400 + }, + { + "epoch": 0.007744, + "grad_norm": 0.3207344606253422, + "learning_rate": 9.99983559101312e-05, + "loss": 0.0451, + "step": 2420 + }, + { + "epoch": 0.007808, + "grad_norm": 0.2525032200615989, + "learning_rate": 9.999832862295052e-05, + "loss": 0.0473, + "step": 2440 + }, + { + "epoch": 0.007872, + "grad_norm": 0.27054214063577636, + "learning_rate": 9.999830111118902e-05, + "loss": 0.0395, + "step": 2460 + }, + { + "epoch": 0.007936, + "grad_norm": 0.3408326046676982, + "learning_rate": 9.999827337484682e-05, + "loss": 0.045, + "step": 2480 + }, + { + "epoch": 0.008, + "grad_norm": 0.6005376063531308, + "learning_rate": 9.999824541392405e-05, + "loss": 0.0461, + "step": 2500 + }, + { + "epoch": 0.008064, + "grad_norm": 0.24918037914823757, + "learning_rate": 9.999821722842081e-05, + "loss": 0.0502, + "step": 2520 + }, + { + "epoch": 0.008128, + "grad_norm": 0.37328523769528615, + "learning_rate": 9.999818881833725e-05, + "loss": 0.0485, + "step": 2540 + }, + { + "epoch": 0.008192, + "grad_norm": 0.28591965288964116, + "learning_rate": 9.999816018367352e-05, + "loss": 0.0493, + "step": 2560 + }, + { + "epoch": 0.008256, + "grad_norm": 0.38472921576940955, + "learning_rate": 9.99981313244297e-05, + "loss": 0.0476, + "step": 2580 + }, + { + "epoch": 0.00832, + "grad_norm": 0.46374225735695795, + "learning_rate": 9.999810224060596e-05, + "loss": 0.0485, + "step": 2600 + }, + { + "epoch": 0.008384, + "grad_norm": 0.328505186320803, + "learning_rate": 9.999807293220238e-05, + "loss": 0.0517, + "step": 2620 + }, + { + "epoch": 0.008448, + "grad_norm": 0.24695111199862108, + "learning_rate": 9.999804339921916e-05, + "loss": 0.0472, + "step": 2640 + }, + { + "epoch": 0.008512, + "grad_norm": 0.46294031072187397, + "learning_rate": 9.999801364165637e-05, + "loss": 0.0449, + "step": 2660 + }, + { + "epoch": 0.008576, + "grad_norm": 0.45946547377431657, + "learning_rate": 9.999798365951419e-05, + "loss": 0.0476, + "step": 2680 + }, + { + "epoch": 0.00864, + "grad_norm": 0.30287823630874633, + "learning_rate": 9.999795345279272e-05, + "loss": 0.0464, + "step": 2700 + }, + { + "epoch": 0.008704, + "grad_norm": 0.2581325811736559, + "learning_rate": 9.999792302149214e-05, + "loss": 0.045, + "step": 2720 + }, + { + "epoch": 0.008768, + "grad_norm": 0.24522591219084197, + "learning_rate": 9.999789236561253e-05, + "loss": 0.046, + "step": 2740 + }, + { + "epoch": 0.008832, + "grad_norm": 0.3112382154972075, + "learning_rate": 9.999786148515406e-05, + "loss": 0.0452, + "step": 2760 + }, + { + "epoch": 0.008896, + "grad_norm": 0.4420397695481255, + "learning_rate": 9.999783038011688e-05, + "loss": 0.0463, + "step": 2780 + }, + { + "epoch": 0.00896, + "grad_norm": 0.2621404198263619, + "learning_rate": 9.999779905050111e-05, + "loss": 0.0461, + "step": 2800 + }, + { + "epoch": 0.009024, + "grad_norm": 0.28164987956647225, + "learning_rate": 9.999776749630688e-05, + "loss": 0.0481, + "step": 2820 + }, + { + "epoch": 0.009088, + "grad_norm": 0.32147247660009226, + "learning_rate": 9.999773571753436e-05, + "loss": 0.0492, + "step": 2840 + }, + { + "epoch": 0.009152, + "grad_norm": 0.35106777153444557, + "learning_rate": 9.999770371418367e-05, + "loss": 0.0459, + "step": 2860 + }, + { + "epoch": 0.009216, + "grad_norm": 0.2577991408051769, + "learning_rate": 9.999767148625497e-05, + "loss": 0.046, + "step": 2880 + }, + { + "epoch": 0.00928, + "grad_norm": 0.2849087970131688, + "learning_rate": 9.99976390337484e-05, + "loss": 0.0453, + "step": 2900 + }, + { + "epoch": 0.009344, + "grad_norm": 0.28871568176698986, + "learning_rate": 9.999760635666411e-05, + "loss": 0.0437, + "step": 2920 + }, + { + "epoch": 0.009408, + "grad_norm": 0.21064694200733436, + "learning_rate": 9.999757345500222e-05, + "loss": 0.0437, + "step": 2940 + }, + { + "epoch": 0.009472, + "grad_norm": 0.2732553089403749, + "learning_rate": 9.999754032876292e-05, + "loss": 0.0406, + "step": 2960 + }, + { + "epoch": 0.009536, + "grad_norm": 0.30320702190166154, + "learning_rate": 9.999750697794632e-05, + "loss": 0.0442, + "step": 2980 + }, + { + "epoch": 0.0096, + "grad_norm": 0.21447400795709393, + "learning_rate": 9.999747340255259e-05, + "loss": 0.0411, + "step": 3000 + }, + { + "epoch": 0.009664, + "grad_norm": 0.20772844528317766, + "learning_rate": 9.999743960258189e-05, + "loss": 0.0401, + "step": 3020 + }, + { + "epoch": 0.009728, + "grad_norm": 0.23471573180575642, + "learning_rate": 9.999740557803434e-05, + "loss": 0.0437, + "step": 3040 + }, + { + "epoch": 0.009792, + "grad_norm": 0.2157948447807665, + "learning_rate": 9.999737132891012e-05, + "loss": 0.0456, + "step": 3060 + }, + { + "epoch": 0.009856, + "grad_norm": 0.18599773670342115, + "learning_rate": 9.999733685520939e-05, + "loss": 0.039, + "step": 3080 + }, + { + "epoch": 0.00992, + "grad_norm": 0.2248349803131493, + "learning_rate": 9.999730215693229e-05, + "loss": 0.0492, + "step": 3100 + }, + { + "epoch": 0.009984, + "grad_norm": 0.20645943875782566, + "learning_rate": 9.999726723407897e-05, + "loss": 0.0457, + "step": 3120 + }, + { + "epoch": 0.010048, + "grad_norm": 0.28197950285780193, + "learning_rate": 9.999723208664957e-05, + "loss": 0.0446, + "step": 3140 + }, + { + "epoch": 0.010112, + "grad_norm": 0.18527584324677912, + "learning_rate": 9.99971967146443e-05, + "loss": 0.0427, + "step": 3160 + }, + { + "epoch": 0.010176, + "grad_norm": 0.208287885693757, + "learning_rate": 9.999716111806328e-05, + "loss": 0.0445, + "step": 3180 + }, + { + "epoch": 0.01024, + "grad_norm": 0.20606070303281376, + "learning_rate": 9.999712529690668e-05, + "loss": 0.0442, + "step": 3200 + }, + { + "epoch": 0.010304, + "grad_norm": 0.3110099728418316, + "learning_rate": 9.999708925117469e-05, + "loss": 0.0451, + "step": 3220 + }, + { + "epoch": 0.010368, + "grad_norm": 0.24973917455019162, + "learning_rate": 9.99970529808674e-05, + "loss": 0.0452, + "step": 3240 + }, + { + "epoch": 0.010432, + "grad_norm": 0.17422485941160815, + "learning_rate": 9.999701648598503e-05, + "loss": 0.0386, + "step": 3260 + }, + { + "epoch": 0.010496, + "grad_norm": 0.19446294414429377, + "learning_rate": 9.999697976652774e-05, + "loss": 0.0408, + "step": 3280 + }, + { + "epoch": 0.01056, + "grad_norm": 0.2628607393100338, + "learning_rate": 9.999694282249568e-05, + "loss": 0.0465, + "step": 3300 + }, + { + "epoch": 0.010624, + "grad_norm": 0.1914174609030993, + "learning_rate": 9.999690565388902e-05, + "loss": 0.0446, + "step": 3320 + }, + { + "epoch": 0.010688, + "grad_norm": 0.15268927179149996, + "learning_rate": 9.999686826070793e-05, + "loss": 0.0419, + "step": 3340 + }, + { + "epoch": 0.010752, + "grad_norm": 0.21281810264609846, + "learning_rate": 9.999683064295258e-05, + "loss": 0.0416, + "step": 3360 + }, + { + "epoch": 0.010816, + "grad_norm": 0.21419375538025656, + "learning_rate": 9.999679280062312e-05, + "loss": 0.0395, + "step": 3380 + }, + { + "epoch": 0.01088, + "grad_norm": 0.2601202748382088, + "learning_rate": 9.999675473371974e-05, + "loss": 0.0368, + "step": 3400 + }, + { + "epoch": 0.010944, + "grad_norm": 0.22383079056064387, + "learning_rate": 9.999671644224262e-05, + "loss": 0.0415, + "step": 3420 + }, + { + "epoch": 0.011008, + "grad_norm": 0.29560811142064425, + "learning_rate": 9.999667792619189e-05, + "loss": 0.0426, + "step": 3440 + }, + { + "epoch": 0.011072, + "grad_norm": 0.3469329488354822, + "learning_rate": 9.999663918556779e-05, + "loss": 0.0436, + "step": 3460 + }, + { + "epoch": 0.011136, + "grad_norm": 0.22744520444727298, + "learning_rate": 9.999660022037043e-05, + "loss": 0.05, + "step": 3480 + }, + { + "epoch": 0.0112, + "grad_norm": 0.20211426841890542, + "learning_rate": 9.99965610306e-05, + "loss": 0.0484, + "step": 3500 + }, + { + "epoch": 0.011264, + "grad_norm": 0.2946741835938952, + "learning_rate": 9.99965216162567e-05, + "loss": 0.046, + "step": 3520 + }, + { + "epoch": 0.011328, + "grad_norm": 0.21850279078131474, + "learning_rate": 9.99964819773407e-05, + "loss": 0.0437, + "step": 3540 + }, + { + "epoch": 0.011392, + "grad_norm": 0.2311073289897065, + "learning_rate": 9.999644211385216e-05, + "loss": 0.0415, + "step": 3560 + }, + { + "epoch": 0.011456, + "grad_norm": 0.2976609274852421, + "learning_rate": 9.999640202579127e-05, + "loss": 0.039, + "step": 3580 + }, + { + "epoch": 0.01152, + "grad_norm": 0.22258823993173257, + "learning_rate": 9.999636171315821e-05, + "loss": 0.041, + "step": 3600 + }, + { + "epoch": 0.011584, + "grad_norm": 0.28317382795178586, + "learning_rate": 9.999632117595317e-05, + "loss": 0.0445, + "step": 3620 + }, + { + "epoch": 0.011648, + "grad_norm": 0.20428252496144694, + "learning_rate": 9.999628041417632e-05, + "loss": 0.0422, + "step": 3640 + }, + { + "epoch": 0.011712, + "grad_norm": 0.21189307987201558, + "learning_rate": 9.999623942782783e-05, + "loss": 0.0394, + "step": 3660 + }, + { + "epoch": 0.011776, + "grad_norm": 0.31346858028822167, + "learning_rate": 9.999619821690792e-05, + "loss": 0.0437, + "step": 3680 + }, + { + "epoch": 0.01184, + "grad_norm": 0.24739476338473063, + "learning_rate": 9.999615678141674e-05, + "loss": 0.0441, + "step": 3700 + }, + { + "epoch": 0.011904, + "grad_norm": 0.26145402735451817, + "learning_rate": 9.999611512135452e-05, + "loss": 0.0419, + "step": 3720 + }, + { + "epoch": 0.011968, + "grad_norm": 0.24413853904942098, + "learning_rate": 9.99960732367214e-05, + "loss": 0.0447, + "step": 3740 + }, + { + "epoch": 0.012032, + "grad_norm": 0.5526540933212379, + "learning_rate": 9.999603112751759e-05, + "loss": 0.0409, + "step": 3760 + }, + { + "epoch": 0.012096, + "grad_norm": 0.23850799794253388, + "learning_rate": 9.999598879374326e-05, + "loss": 0.043, + "step": 3780 + }, + { + "epoch": 0.01216, + "grad_norm": 0.31969676291937266, + "learning_rate": 9.999594623539862e-05, + "loss": 0.0441, + "step": 3800 + }, + { + "epoch": 0.012224, + "grad_norm": 0.20951360956925488, + "learning_rate": 9.999590345248388e-05, + "loss": 0.0463, + "step": 3820 + }, + { + "epoch": 0.012288, + "grad_norm": 0.25793718921030695, + "learning_rate": 9.999586044499921e-05, + "loss": 0.0446, + "step": 3840 + }, + { + "epoch": 0.012352, + "grad_norm": 0.2297692199750354, + "learning_rate": 9.999581721294477e-05, + "loss": 0.0439, + "step": 3860 + }, + { + "epoch": 0.012416, + "grad_norm": 0.19218350487770397, + "learning_rate": 9.999577375632081e-05, + "loss": 0.0449, + "step": 3880 + }, + { + "epoch": 0.01248, + "grad_norm": 0.3033648329528053, + "learning_rate": 9.99957300751275e-05, + "loss": 0.0428, + "step": 3900 + }, + { + "epoch": 0.012544, + "grad_norm": 0.40250617976922864, + "learning_rate": 9.999568616936503e-05, + "loss": 0.0417, + "step": 3920 + }, + { + "epoch": 0.012608, + "grad_norm": 0.24185351566600372, + "learning_rate": 9.999564203903363e-05, + "loss": 0.0452, + "step": 3940 + }, + { + "epoch": 0.012672, + "grad_norm": 0.2276674393893432, + "learning_rate": 9.999559768413345e-05, + "loss": 0.045, + "step": 3960 + }, + { + "epoch": 0.012736, + "grad_norm": 0.22042326158393796, + "learning_rate": 9.999555310466471e-05, + "loss": 0.0388, + "step": 3980 + }, + { + "epoch": 0.0128, + "grad_norm": 0.2160343386870399, + "learning_rate": 9.999550830062763e-05, + "loss": 0.0389, + "step": 4000 + }, + { + "epoch": 0.012864, + "grad_norm": 0.5155755250188268, + "learning_rate": 9.999546327202238e-05, + "loss": 0.0433, + "step": 4020 + }, + { + "epoch": 0.012928, + "grad_norm": 0.2890268759735471, + "learning_rate": 9.999541801884917e-05, + "loss": 0.0445, + "step": 4040 + }, + { + "epoch": 0.012992, + "grad_norm": 0.27385775279312136, + "learning_rate": 9.999537254110824e-05, + "loss": 0.0461, + "step": 4060 + }, + { + "epoch": 0.013056, + "grad_norm": 0.2020938348958419, + "learning_rate": 9.999532683879974e-05, + "loss": 0.0457, + "step": 4080 + }, + { + "epoch": 0.01312, + "grad_norm": 0.26852894636721175, + "learning_rate": 9.999528091192391e-05, + "loss": 0.0497, + "step": 4100 + }, + { + "epoch": 0.013184, + "grad_norm": 0.2261581058171256, + "learning_rate": 9.999523476048093e-05, + "loss": 0.0433, + "step": 4120 + }, + { + "epoch": 0.013248, + "grad_norm": 0.20294697552399188, + "learning_rate": 9.999518838447103e-05, + "loss": 0.045, + "step": 4140 + }, + { + "epoch": 0.013312, + "grad_norm": 0.25503703498212965, + "learning_rate": 9.999514178389443e-05, + "loss": 0.0468, + "step": 4160 + }, + { + "epoch": 0.013376, + "grad_norm": 0.31416615760545596, + "learning_rate": 9.99950949587513e-05, + "loss": 0.044, + "step": 4180 + }, + { + "epoch": 0.01344, + "grad_norm": 0.3127318682513348, + "learning_rate": 9.999504790904188e-05, + "loss": 0.0421, + "step": 4200 + }, + { + "epoch": 0.013504, + "grad_norm": 0.3883954329946083, + "learning_rate": 9.999500063476639e-05, + "loss": 0.0448, + "step": 4220 + }, + { + "epoch": 0.013568, + "grad_norm": 0.3584302284803773, + "learning_rate": 9.9994953135925e-05, + "loss": 0.0436, + "step": 4240 + }, + { + "epoch": 0.013632, + "grad_norm": 0.2432655417897402, + "learning_rate": 9.999490541251794e-05, + "loss": 0.0441, + "step": 4260 + }, + { + "epoch": 0.013696, + "grad_norm": 0.246759048088496, + "learning_rate": 9.999485746454547e-05, + "loss": 0.0452, + "step": 4280 + }, + { + "epoch": 0.01376, + "grad_norm": 0.32387285050358106, + "learning_rate": 9.999480929200774e-05, + "loss": 0.0391, + "step": 4300 + }, + { + "epoch": 0.013824, + "grad_norm": 0.25562150582991233, + "learning_rate": 9.999476089490499e-05, + "loss": 0.0391, + "step": 4320 + }, + { + "epoch": 0.013888, + "grad_norm": 0.2558228907057912, + "learning_rate": 9.999471227323745e-05, + "loss": 0.0449, + "step": 4340 + }, + { + "epoch": 0.013952, + "grad_norm": 0.20002668791572958, + "learning_rate": 9.999466342700534e-05, + "loss": 0.0413, + "step": 4360 + }, + { + "epoch": 0.014016, + "grad_norm": 0.12224090068082925, + "learning_rate": 9.999461435620886e-05, + "loss": 0.0414, + "step": 4380 + }, + { + "epoch": 0.01408, + "grad_norm": 0.24528343133145106, + "learning_rate": 9.999456506084823e-05, + "loss": 0.0382, + "step": 4400 + }, + { + "epoch": 0.014144, + "grad_norm": 0.20452480266090878, + "learning_rate": 9.999451554092369e-05, + "loss": 0.0452, + "step": 4420 + }, + { + "epoch": 0.014208, + "grad_norm": 0.18483919695658882, + "learning_rate": 9.999446579643547e-05, + "loss": 0.0385, + "step": 4440 + }, + { + "epoch": 0.014272, + "grad_norm": 0.2182194295469889, + "learning_rate": 9.999441582738376e-05, + "loss": 0.0393, + "step": 4460 + }, + { + "epoch": 0.014336, + "grad_norm": 0.23876712224949892, + "learning_rate": 9.99943656337688e-05, + "loss": 0.0407, + "step": 4480 + }, + { + "epoch": 0.0144, + "grad_norm": 0.4384447884289934, + "learning_rate": 9.999431521559082e-05, + "loss": 0.043, + "step": 4500 + }, + { + "epoch": 0.014464, + "grad_norm": 0.30307471328960556, + "learning_rate": 9.999426457285004e-05, + "loss": 0.0437, + "step": 4520 + }, + { + "epoch": 0.014528, + "grad_norm": 0.227436400634274, + "learning_rate": 9.99942137055467e-05, + "loss": 0.0392, + "step": 4540 + }, + { + "epoch": 0.014592, + "grad_norm": 0.15097773797353434, + "learning_rate": 9.999416261368102e-05, + "loss": 0.0437, + "step": 4560 + }, + { + "epoch": 0.014656, + "grad_norm": 0.18980501687573087, + "learning_rate": 9.999411129725322e-05, + "loss": 0.0425, + "step": 4580 + }, + { + "epoch": 0.01472, + "grad_norm": 0.3055583272822677, + "learning_rate": 9.999405975626354e-05, + "loss": 0.0441, + "step": 4600 + }, + { + "epoch": 0.014784, + "grad_norm": 0.16265185830882983, + "learning_rate": 9.999400799071222e-05, + "loss": 0.0422, + "step": 4620 + }, + { + "epoch": 0.014848, + "grad_norm": 0.2678225156503953, + "learning_rate": 9.999395600059949e-05, + "loss": 0.0445, + "step": 4640 + }, + { + "epoch": 0.014912, + "grad_norm": 0.3136899388950313, + "learning_rate": 9.999390378592556e-05, + "loss": 0.0423, + "step": 4660 + }, + { + "epoch": 0.014976, + "grad_norm": 0.2240918298251088, + "learning_rate": 9.999385134669068e-05, + "loss": 0.041, + "step": 4680 + }, + { + "epoch": 0.01504, + "grad_norm": 0.21254312631985064, + "learning_rate": 9.999379868289509e-05, + "loss": 0.0436, + "step": 4700 + }, + { + "epoch": 0.015104, + "grad_norm": 0.2231332870847772, + "learning_rate": 9.999374579453904e-05, + "loss": 0.0419, + "step": 4720 + }, + { + "epoch": 0.015168, + "grad_norm": 0.18493137319251832, + "learning_rate": 9.999369268162273e-05, + "loss": 0.0383, + "step": 4740 + }, + { + "epoch": 0.015232, + "grad_norm": 0.27489319131281786, + "learning_rate": 9.999363934414643e-05, + "loss": 0.0423, + "step": 4760 + }, + { + "epoch": 0.015296, + "grad_norm": 0.2075167028312529, + "learning_rate": 9.999358578211038e-05, + "loss": 0.0426, + "step": 4780 + }, + { + "epoch": 0.01536, + "grad_norm": 0.22680209361651998, + "learning_rate": 9.999353199551478e-05, + "loss": 0.0437, + "step": 4800 + }, + { + "epoch": 0.015424, + "grad_norm": 0.18543121483976474, + "learning_rate": 9.999347798435993e-05, + "loss": 0.0389, + "step": 4820 + }, + { + "epoch": 0.015488, + "grad_norm": 0.1615790423989882, + "learning_rate": 9.999342374864604e-05, + "loss": 0.0419, + "step": 4840 + }, + { + "epoch": 0.015552, + "grad_norm": 0.21311733127805854, + "learning_rate": 9.999336928837335e-05, + "loss": 0.0449, + "step": 4860 + }, + { + "epoch": 0.015616, + "grad_norm": 0.2223299603170898, + "learning_rate": 9.999331460354211e-05, + "loss": 0.0441, + "step": 4880 + }, + { + "epoch": 0.01568, + "grad_norm": 0.24234271182173017, + "learning_rate": 9.999325969415258e-05, + "loss": 0.038, + "step": 4900 + }, + { + "epoch": 0.015744, + "grad_norm": 0.24654813714023002, + "learning_rate": 9.999320456020497e-05, + "loss": 0.043, + "step": 4920 + }, + { + "epoch": 0.015808, + "grad_norm": 0.1699244490825696, + "learning_rate": 9.999314920169958e-05, + "loss": 0.0406, + "step": 4940 + }, + { + "epoch": 0.015872, + "grad_norm": 0.18883656468115048, + "learning_rate": 9.999309361863662e-05, + "loss": 0.0432, + "step": 4960 + }, + { + "epoch": 0.015936, + "grad_norm": 0.2272491945599065, + "learning_rate": 9.999303781101635e-05, + "loss": 0.0386, + "step": 4980 + }, + { + "epoch": 0.016, + "grad_norm": 0.24488576388609443, + "learning_rate": 9.999298177883903e-05, + "loss": 0.0462, + "step": 5000 + }, + { + "epoch": 0.016064, + "grad_norm": 0.21481279115518054, + "learning_rate": 9.999292552210488e-05, + "loss": 0.0415, + "step": 5020 + }, + { + "epoch": 0.016128, + "grad_norm": 0.23192172350908144, + "learning_rate": 9.999286904081419e-05, + "loss": 0.0424, + "step": 5040 + }, + { + "epoch": 0.016192, + "grad_norm": 0.2482611258735945, + "learning_rate": 9.999281233496721e-05, + "loss": 0.0408, + "step": 5060 + }, + { + "epoch": 0.016256, + "grad_norm": 0.14798751306640967, + "learning_rate": 9.999275540456419e-05, + "loss": 0.0381, + "step": 5080 + }, + { + "epoch": 0.01632, + "grad_norm": 0.24469681469863944, + "learning_rate": 9.999269824960535e-05, + "loss": 0.0401, + "step": 5100 + }, + { + "epoch": 0.016384, + "grad_norm": 0.1811077009922402, + "learning_rate": 9.999264087009099e-05, + "loss": 0.041, + "step": 5120 + }, + { + "epoch": 0.016448, + "grad_norm": 0.336911717206514, + "learning_rate": 9.999258326602137e-05, + "loss": 0.043, + "step": 5140 + }, + { + "epoch": 0.016512, + "grad_norm": 0.20643988761597026, + "learning_rate": 9.99925254373967e-05, + "loss": 0.0407, + "step": 5160 + }, + { + "epoch": 0.016576, + "grad_norm": 0.31888498684656585, + "learning_rate": 9.999246738421731e-05, + "loss": 0.0393, + "step": 5180 + }, + { + "epoch": 0.01664, + "grad_norm": 0.23181655374071092, + "learning_rate": 9.999240910648339e-05, + "loss": 0.0392, + "step": 5200 + }, + { + "epoch": 0.016704, + "grad_norm": 0.1386213007577926, + "learning_rate": 9.999235060419525e-05, + "loss": 0.0389, + "step": 5220 + }, + { + "epoch": 0.016768, + "grad_norm": 0.18883798238282606, + "learning_rate": 9.999229187735315e-05, + "loss": 0.0423, + "step": 5240 + }, + { + "epoch": 0.016832, + "grad_norm": 0.14006642257018215, + "learning_rate": 9.999223292595732e-05, + "loss": 0.0382, + "step": 5260 + }, + { + "epoch": 0.016896, + "grad_norm": 0.18435441787783818, + "learning_rate": 9.999217375000806e-05, + "loss": 0.042, + "step": 5280 + }, + { + "epoch": 0.01696, + "grad_norm": 0.17959844094692956, + "learning_rate": 9.999211434950561e-05, + "loss": 0.0402, + "step": 5300 + }, + { + "epoch": 0.017024, + "grad_norm": 0.23849786171144446, + "learning_rate": 9.999205472445027e-05, + "loss": 0.0443, + "step": 5320 + }, + { + "epoch": 0.017088, + "grad_norm": 0.21335476597643852, + "learning_rate": 9.999199487484227e-05, + "loss": 0.0415, + "step": 5340 + }, + { + "epoch": 0.017152, + "grad_norm": 0.1803485361980176, + "learning_rate": 9.99919348006819e-05, + "loss": 0.0427, + "step": 5360 + }, + { + "epoch": 0.017216, + "grad_norm": 0.33385981792085784, + "learning_rate": 9.999187450196944e-05, + "loss": 0.0424, + "step": 5380 + }, + { + "epoch": 0.01728, + "grad_norm": 0.1945367708380311, + "learning_rate": 9.999181397870514e-05, + "loss": 0.0428, + "step": 5400 + }, + { + "epoch": 0.017344, + "grad_norm": 0.1349909456906324, + "learning_rate": 9.999175323088927e-05, + "loss": 0.0415, + "step": 5420 + }, + { + "epoch": 0.017408, + "grad_norm": 0.15270263987731508, + "learning_rate": 9.999169225852212e-05, + "loss": 0.0363, + "step": 5440 + }, + { + "epoch": 0.017472, + "grad_norm": 0.20076037788130052, + "learning_rate": 9.999163106160395e-05, + "loss": 0.04, + "step": 5460 + }, + { + "epoch": 0.017536, + "grad_norm": 0.17645278922324295, + "learning_rate": 9.999156964013503e-05, + "loss": 0.0414, + "step": 5480 + }, + { + "epoch": 0.0176, + "grad_norm": 0.2572144123888314, + "learning_rate": 9.999150799411566e-05, + "loss": 0.0425, + "step": 5500 + }, + { + "epoch": 0.017664, + "grad_norm": 0.20447774147657222, + "learning_rate": 9.999144612354609e-05, + "loss": 0.0417, + "step": 5520 + }, + { + "epoch": 0.017728, + "grad_norm": 0.2984887585657629, + "learning_rate": 9.999138402842662e-05, + "loss": 0.0397, + "step": 5540 + }, + { + "epoch": 0.017792, + "grad_norm": 0.27722018367891504, + "learning_rate": 9.999132170875752e-05, + "loss": 0.0443, + "step": 5560 + }, + { + "epoch": 0.017856, + "grad_norm": 0.18656399917273497, + "learning_rate": 9.999125916453907e-05, + "loss": 0.0422, + "step": 5580 + }, + { + "epoch": 0.01792, + "grad_norm": 0.38911082939661396, + "learning_rate": 9.999119639577155e-05, + "loss": 0.0427, + "step": 5600 + }, + { + "epoch": 0.017984, + "grad_norm": 0.2426718821770736, + "learning_rate": 9.999113340245525e-05, + "loss": 0.0409, + "step": 5620 + }, + { + "epoch": 0.018048, + "grad_norm": 0.2142934344961841, + "learning_rate": 9.999107018459044e-05, + "loss": 0.0407, + "step": 5640 + }, + { + "epoch": 0.018112, + "grad_norm": 0.2217788598087543, + "learning_rate": 9.999100674217741e-05, + "loss": 0.0404, + "step": 5660 + }, + { + "epoch": 0.018176, + "grad_norm": 0.18367822859673974, + "learning_rate": 9.999094307521644e-05, + "loss": 0.0403, + "step": 5680 + }, + { + "epoch": 0.01824, + "grad_norm": 0.20063131701247866, + "learning_rate": 9.999087918370782e-05, + "loss": 0.0429, + "step": 5700 + }, + { + "epoch": 0.018304, + "grad_norm": 0.32225956276545603, + "learning_rate": 9.999081506765185e-05, + "loss": 0.041, + "step": 5720 + }, + { + "epoch": 0.018368, + "grad_norm": 0.18808449918030978, + "learning_rate": 9.999075072704877e-05, + "loss": 0.0404, + "step": 5740 + }, + { + "epoch": 0.018432, + "grad_norm": 0.3853663405580428, + "learning_rate": 9.999068616189894e-05, + "loss": 0.0431, + "step": 5760 + }, + { + "epoch": 0.018496, + "grad_norm": 0.1644007944200343, + "learning_rate": 9.99906213722026e-05, + "loss": 0.0431, + "step": 5780 + }, + { + "epoch": 0.01856, + "grad_norm": 0.17517977327002154, + "learning_rate": 9.999055635796006e-05, + "loss": 0.0425, + "step": 5800 + }, + { + "epoch": 0.018624, + "grad_norm": 0.18088428451899596, + "learning_rate": 9.99904911191716e-05, + "loss": 0.0398, + "step": 5820 + }, + { + "epoch": 0.018688, + "grad_norm": 0.17297858743179007, + "learning_rate": 9.999042565583753e-05, + "loss": 0.0422, + "step": 5840 + }, + { + "epoch": 0.018752, + "grad_norm": 0.2286153813558773, + "learning_rate": 9.999035996795813e-05, + "loss": 0.0428, + "step": 5860 + }, + { + "epoch": 0.018816, + "grad_norm": 0.19546619504183593, + "learning_rate": 9.99902940555337e-05, + "loss": 0.0407, + "step": 5880 + }, + { + "epoch": 0.01888, + "grad_norm": 0.14551806226358296, + "learning_rate": 9.999022791856455e-05, + "loss": 0.0387, + "step": 5900 + }, + { + "epoch": 0.018944, + "grad_norm": 0.2711865188746732, + "learning_rate": 9.999016155705094e-05, + "loss": 0.0382, + "step": 5920 + }, + { + "epoch": 0.019008, + "grad_norm": 0.21706897910593828, + "learning_rate": 9.999009497099321e-05, + "loss": 0.0415, + "step": 5940 + }, + { + "epoch": 0.019072, + "grad_norm": 0.20105105793004252, + "learning_rate": 9.999002816039163e-05, + "loss": 0.0385, + "step": 5960 + }, + { + "epoch": 0.019136, + "grad_norm": 0.17709142206799572, + "learning_rate": 9.998996112524651e-05, + "loss": 0.0387, + "step": 5980 + }, + { + "epoch": 0.0192, + "grad_norm": 0.16370566946488016, + "learning_rate": 9.998989386555814e-05, + "loss": 0.0387, + "step": 6000 + }, + { + "epoch": 0.019264, + "grad_norm": 0.2239297193863687, + "learning_rate": 9.998982638132686e-05, + "loss": 0.0413, + "step": 6020 + }, + { + "epoch": 0.019328, + "grad_norm": 0.3014416034611583, + "learning_rate": 9.998975867255293e-05, + "loss": 0.039, + "step": 6040 + }, + { + "epoch": 0.019392, + "grad_norm": 0.24120513811169209, + "learning_rate": 9.998969073923668e-05, + "loss": 0.0397, + "step": 6060 + }, + { + "epoch": 0.019456, + "grad_norm": 0.23249714054344267, + "learning_rate": 9.998962258137839e-05, + "loss": 0.0403, + "step": 6080 + }, + { + "epoch": 0.01952, + "grad_norm": 0.14600511813355524, + "learning_rate": 9.998955419897838e-05, + "loss": 0.042, + "step": 6100 + }, + { + "epoch": 0.019584, + "grad_norm": 0.16695656150129792, + "learning_rate": 9.998948559203698e-05, + "loss": 0.0409, + "step": 6120 + }, + { + "epoch": 0.019648, + "grad_norm": 0.1696923052821074, + "learning_rate": 9.998941676055446e-05, + "loss": 0.0408, + "step": 6140 + }, + { + "epoch": 0.019712, + "grad_norm": 0.2053437439327295, + "learning_rate": 9.998934770453115e-05, + "loss": 0.0412, + "step": 6160 + }, + { + "epoch": 0.019776, + "grad_norm": 0.19747318155213014, + "learning_rate": 9.998927842396736e-05, + "loss": 0.0376, + "step": 6180 + }, + { + "epoch": 0.01984, + "grad_norm": 0.15256686727763652, + "learning_rate": 9.998920891886339e-05, + "loss": 0.0368, + "step": 6200 + }, + { + "epoch": 0.019904, + "grad_norm": 0.3770590943630661, + "learning_rate": 9.998913918921957e-05, + "loss": 0.0416, + "step": 6220 + }, + { + "epoch": 0.019968, + "grad_norm": 0.2532149056971487, + "learning_rate": 9.99890692350362e-05, + "loss": 0.0404, + "step": 6240 + }, + { + "epoch": 0.020032, + "grad_norm": 0.15435449373864438, + "learning_rate": 9.99889990563136e-05, + "loss": 0.0404, + "step": 6260 + }, + { + "epoch": 0.020096, + "grad_norm": 0.147340626498922, + "learning_rate": 9.998892865305207e-05, + "loss": 0.0369, + "step": 6280 + }, + { + "epoch": 0.02016, + "grad_norm": 0.2402419073635215, + "learning_rate": 9.998885802525196e-05, + "loss": 0.0365, + "step": 6300 + }, + { + "epoch": 0.020224, + "grad_norm": 0.1506992405168411, + "learning_rate": 9.998878717291353e-05, + "loss": 0.0382, + "step": 6320 + }, + { + "epoch": 0.020288, + "grad_norm": 0.2410919238356268, + "learning_rate": 9.998871609603717e-05, + "loss": 0.0395, + "step": 6340 + }, + { + "epoch": 0.020352, + "grad_norm": 0.1541154314193387, + "learning_rate": 9.998864479462315e-05, + "loss": 0.0401, + "step": 6360 + }, + { + "epoch": 0.020416, + "grad_norm": 0.20636861820767585, + "learning_rate": 9.998857326867181e-05, + "loss": 0.038, + "step": 6380 + }, + { + "epoch": 0.02048, + "grad_norm": 0.22320932302135035, + "learning_rate": 9.998850151818345e-05, + "loss": 0.0436, + "step": 6400 + }, + { + "epoch": 0.020544, + "grad_norm": 0.16962600707739206, + "learning_rate": 9.998842954315841e-05, + "loss": 0.0469, + "step": 6420 + }, + { + "epoch": 0.020608, + "grad_norm": 0.20530797039322027, + "learning_rate": 9.998835734359703e-05, + "loss": 0.041, + "step": 6440 + }, + { + "epoch": 0.020672, + "grad_norm": 0.2231020758948462, + "learning_rate": 9.998828491949959e-05, + "loss": 0.0467, + "step": 6460 + }, + { + "epoch": 0.020736, + "grad_norm": 0.21324966056479905, + "learning_rate": 9.998821227086646e-05, + "loss": 0.0423, + "step": 6480 + }, + { + "epoch": 0.0208, + "grad_norm": 0.19911824709229392, + "learning_rate": 9.998813939769794e-05, + "loss": 0.0425, + "step": 6500 + }, + { + "epoch": 0.020864, + "grad_norm": 0.2440791218660832, + "learning_rate": 9.998806629999435e-05, + "loss": 0.0428, + "step": 6520 + }, + { + "epoch": 0.020928, + "grad_norm": 0.16981866057097575, + "learning_rate": 9.998799297775605e-05, + "loss": 0.0402, + "step": 6540 + }, + { + "epoch": 0.020992, + "grad_norm": 0.165807532742159, + "learning_rate": 9.998791943098335e-05, + "loss": 0.0402, + "step": 6560 + }, + { + "epoch": 0.021056, + "grad_norm": 0.1952217880524122, + "learning_rate": 9.998784565967658e-05, + "loss": 0.0414, + "step": 6580 + }, + { + "epoch": 0.02112, + "grad_norm": 0.19654614011630828, + "learning_rate": 9.998777166383608e-05, + "loss": 0.0387, + "step": 6600 + }, + { + "epoch": 0.021184, + "grad_norm": 0.2863565740611335, + "learning_rate": 9.998769744346216e-05, + "loss": 0.038, + "step": 6620 + }, + { + "epoch": 0.021248, + "grad_norm": 0.2124879422748257, + "learning_rate": 9.998762299855517e-05, + "loss": 0.0417, + "step": 6640 + }, + { + "epoch": 0.021312, + "grad_norm": 0.1636953639928248, + "learning_rate": 9.998754832911546e-05, + "loss": 0.0422, + "step": 6660 + }, + { + "epoch": 0.021376, + "grad_norm": 0.23601133270974403, + "learning_rate": 9.998747343514335e-05, + "loss": 0.0366, + "step": 6680 + }, + { + "epoch": 0.02144, + "grad_norm": 0.21408800613272436, + "learning_rate": 9.998739831663914e-05, + "loss": 0.0353, + "step": 6700 + }, + { + "epoch": 0.021504, + "grad_norm": 0.18052832043307795, + "learning_rate": 9.998732297360324e-05, + "loss": 0.0399, + "step": 6720 + }, + { + "epoch": 0.021568, + "grad_norm": 0.3525481496934496, + "learning_rate": 9.998724740603593e-05, + "loss": 0.0395, + "step": 6740 + }, + { + "epoch": 0.021632, + "grad_norm": 0.2369571668099414, + "learning_rate": 9.998717161393758e-05, + "loss": 0.0404, + "step": 6760 + }, + { + "epoch": 0.021696, + "grad_norm": 0.16161923330036226, + "learning_rate": 9.998709559730853e-05, + "loss": 0.0374, + "step": 6780 + }, + { + "epoch": 0.02176, + "grad_norm": 0.14137582646219507, + "learning_rate": 9.998701935614909e-05, + "loss": 0.038, + "step": 6800 + }, + { + "epoch": 0.021824, + "grad_norm": 0.1471023915556908, + "learning_rate": 9.998694289045964e-05, + "loss": 0.0364, + "step": 6820 + }, + { + "epoch": 0.021888, + "grad_norm": 0.1766636345498869, + "learning_rate": 9.998686620024051e-05, + "loss": 0.0398, + "step": 6840 + }, + { + "epoch": 0.021952, + "grad_norm": 0.1980008598583829, + "learning_rate": 9.998678928549203e-05, + "loss": 0.0372, + "step": 6860 + }, + { + "epoch": 0.022016, + "grad_norm": 0.1568671064686567, + "learning_rate": 9.998671214621457e-05, + "loss": 0.0443, + "step": 6880 + }, + { + "epoch": 0.02208, + "grad_norm": 0.15536838591377639, + "learning_rate": 9.998663478240846e-05, + "loss": 0.0405, + "step": 6900 + }, + { + "epoch": 0.022144, + "grad_norm": 0.2564998022152435, + "learning_rate": 9.998655719407404e-05, + "loss": 0.0427, + "step": 6920 + }, + { + "epoch": 0.022208, + "grad_norm": 0.29646525478774344, + "learning_rate": 9.998647938121169e-05, + "loss": 0.0374, + "step": 6940 + }, + { + "epoch": 0.022272, + "grad_norm": 0.1508464411429471, + "learning_rate": 9.998640134382173e-05, + "loss": 0.0363, + "step": 6960 + }, + { + "epoch": 0.022336, + "grad_norm": 0.2213965947537837, + "learning_rate": 9.998632308190453e-05, + "loss": 0.035, + "step": 6980 + }, + { + "epoch": 0.0224, + "grad_norm": 0.19188403122898182, + "learning_rate": 9.998624459546044e-05, + "loss": 0.0426, + "step": 7000 + }, + { + "epoch": 0.022464, + "grad_norm": 0.18823547847862185, + "learning_rate": 9.998616588448978e-05, + "loss": 0.0393, + "step": 7020 + }, + { + "epoch": 0.022528, + "grad_norm": 0.16949788210042196, + "learning_rate": 9.998608694899294e-05, + "loss": 0.0398, + "step": 7040 + }, + { + "epoch": 0.022592, + "grad_norm": 0.16212726739622002, + "learning_rate": 9.998600778897028e-05, + "loss": 0.0392, + "step": 7060 + }, + { + "epoch": 0.022656, + "grad_norm": 0.18536407225773302, + "learning_rate": 9.998592840442212e-05, + "loss": 0.038, + "step": 7080 + }, + { + "epoch": 0.02272, + "grad_norm": 0.18130925969004766, + "learning_rate": 9.998584879534883e-05, + "loss": 0.038, + "step": 7100 + }, + { + "epoch": 0.022784, + "grad_norm": 0.17676041070087142, + "learning_rate": 9.998576896175079e-05, + "loss": 0.0399, + "step": 7120 + }, + { + "epoch": 0.022848, + "grad_norm": 0.34645445981393896, + "learning_rate": 9.998568890362832e-05, + "loss": 0.039, + "step": 7140 + }, + { + "epoch": 0.022912, + "grad_norm": 0.1280697447157394, + "learning_rate": 9.99856086209818e-05, + "loss": 0.0428, + "step": 7160 + }, + { + "epoch": 0.022976, + "grad_norm": 0.16093425836224232, + "learning_rate": 9.99855281138116e-05, + "loss": 0.0426, + "step": 7180 + }, + { + "epoch": 0.02304, + "grad_norm": 0.18112548072255152, + "learning_rate": 9.998544738211809e-05, + "loss": 0.0375, + "step": 7200 + }, + { + "epoch": 0.023104, + "grad_norm": 0.2015163130939942, + "learning_rate": 9.99853664259016e-05, + "loss": 0.0409, + "step": 7220 + }, + { + "epoch": 0.023168, + "grad_norm": 0.20587117079531728, + "learning_rate": 9.99852852451625e-05, + "loss": 0.0412, + "step": 7240 + }, + { + "epoch": 0.023232, + "grad_norm": 0.15914815285760722, + "learning_rate": 9.998520383990119e-05, + "loss": 0.0387, + "step": 7260 + }, + { + "epoch": 0.023296, + "grad_norm": 0.22595317914120058, + "learning_rate": 9.9985122210118e-05, + "loss": 0.0377, + "step": 7280 + }, + { + "epoch": 0.02336, + "grad_norm": 0.1786837340367665, + "learning_rate": 9.99850403558133e-05, + "loss": 0.0359, + "step": 7300 + }, + { + "epoch": 0.023424, + "grad_norm": 0.14531005857609933, + "learning_rate": 9.998495827698746e-05, + "loss": 0.04, + "step": 7320 + }, + { + "epoch": 0.023488, + "grad_norm": 0.20487801260748323, + "learning_rate": 9.998487597364086e-05, + "loss": 0.0409, + "step": 7340 + }, + { + "epoch": 0.023552, + "grad_norm": 0.16658937820195704, + "learning_rate": 9.998479344577386e-05, + "loss": 0.0394, + "step": 7360 + }, + { + "epoch": 0.023616, + "grad_norm": 0.19146991618045886, + "learning_rate": 9.998471069338684e-05, + "loss": 0.0369, + "step": 7380 + }, + { + "epoch": 0.02368, + "grad_norm": 0.2147646001103393, + "learning_rate": 9.998462771648016e-05, + "loss": 0.0408, + "step": 7400 + }, + { + "epoch": 0.023744, + "grad_norm": 0.20719117655318012, + "learning_rate": 9.998454451505419e-05, + "loss": 0.0389, + "step": 7420 + }, + { + "epoch": 0.023808, + "grad_norm": 0.2737724250233668, + "learning_rate": 9.998446108910933e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 0.023872, + "grad_norm": 0.21919387559271272, + "learning_rate": 9.998437743864591e-05, + "loss": 0.0394, + "step": 7460 + }, + { + "epoch": 0.023936, + "grad_norm": 0.28979932750223103, + "learning_rate": 9.998429356366434e-05, + "loss": 0.0372, + "step": 7480 + }, + { + "epoch": 0.024, + "grad_norm": 0.18453598226987689, + "learning_rate": 9.9984209464165e-05, + "loss": 0.0351, + "step": 7500 + }, + { + "epoch": 0.024064, + "grad_norm": 0.13414968051956097, + "learning_rate": 9.998412514014825e-05, + "loss": 0.0379, + "step": 7520 + }, + { + "epoch": 0.024128, + "grad_norm": 0.20059131884245548, + "learning_rate": 9.998404059161448e-05, + "loss": 0.0388, + "step": 7540 + }, + { + "epoch": 0.024192, + "grad_norm": 0.149587738216436, + "learning_rate": 9.998395581856406e-05, + "loss": 0.0361, + "step": 7560 + }, + { + "epoch": 0.024256, + "grad_norm": 0.15132694347285827, + "learning_rate": 9.998387082099738e-05, + "loss": 0.0411, + "step": 7580 + }, + { + "epoch": 0.02432, + "grad_norm": 0.13219501095161013, + "learning_rate": 9.998378559891481e-05, + "loss": 0.0384, + "step": 7600 + }, + { + "epoch": 0.024384, + "grad_norm": 0.16080473548995142, + "learning_rate": 9.998370015231673e-05, + "loss": 0.0382, + "step": 7620 + }, + { + "epoch": 0.024448, + "grad_norm": 0.17207286319953086, + "learning_rate": 9.998361448120356e-05, + "loss": 0.0376, + "step": 7640 + }, + { + "epoch": 0.024512, + "grad_norm": 0.14924823509590857, + "learning_rate": 9.998352858557565e-05, + "loss": 0.0401, + "step": 7660 + }, + { + "epoch": 0.024576, + "grad_norm": 0.3419475288653469, + "learning_rate": 9.998344246543339e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 0.02464, + "grad_norm": 0.26280630339510985, + "learning_rate": 9.998335612077716e-05, + "loss": 0.0402, + "step": 7700 + }, + { + "epoch": 0.024704, + "grad_norm": 0.12230540892969835, + "learning_rate": 9.998326955160737e-05, + "loss": 0.0374, + "step": 7720 + }, + { + "epoch": 0.024768, + "grad_norm": 0.17481274082557954, + "learning_rate": 9.998318275792442e-05, + "loss": 0.0376, + "step": 7740 + }, + { + "epoch": 0.024832, + "grad_norm": 0.19630034225196766, + "learning_rate": 9.998309573972864e-05, + "loss": 0.0412, + "step": 7760 + }, + { + "epoch": 0.024896, + "grad_norm": 0.1729501761934857, + "learning_rate": 9.998300849702048e-05, + "loss": 0.0382, + "step": 7780 + }, + { + "epoch": 0.02496, + "grad_norm": 0.27691757705187797, + "learning_rate": 9.998292102980031e-05, + "loss": 0.0379, + "step": 7800 + }, + { + "epoch": 0.025024, + "grad_norm": 0.1384105412534315, + "learning_rate": 9.998283333806852e-05, + "loss": 0.0376, + "step": 7820 + }, + { + "epoch": 0.025088, + "grad_norm": 0.13370795970288152, + "learning_rate": 9.998274542182551e-05, + "loss": 0.0372, + "step": 7840 + }, + { + "epoch": 0.025152, + "grad_norm": 0.18799996409722464, + "learning_rate": 9.998265728107167e-05, + "loss": 0.0376, + "step": 7860 + }, + { + "epoch": 0.025216, + "grad_norm": 0.12437865151919679, + "learning_rate": 9.99825689158074e-05, + "loss": 0.0388, + "step": 7880 + }, + { + "epoch": 0.02528, + "grad_norm": 0.18808500042301343, + "learning_rate": 9.998248032603308e-05, + "loss": 0.0343, + "step": 7900 + }, + { + "epoch": 0.025344, + "grad_norm": 0.16854270089489595, + "learning_rate": 9.998239151174915e-05, + "loss": 0.0381, + "step": 7920 + }, + { + "epoch": 0.025408, + "grad_norm": 0.17016575138075457, + "learning_rate": 9.998230247295596e-05, + "loss": 0.037, + "step": 7940 + }, + { + "epoch": 0.025472, + "grad_norm": 0.204708594855192, + "learning_rate": 9.998221320965394e-05, + "loss": 0.0379, + "step": 7960 + }, + { + "epoch": 0.025536, + "grad_norm": 0.11019640506024206, + "learning_rate": 9.998212372184349e-05, + "loss": 0.0386, + "step": 7980 + }, + { + "epoch": 0.0256, + "grad_norm": 0.15693975940046118, + "learning_rate": 9.998203400952499e-05, + "loss": 0.0366, + "step": 8000 + }, + { + "epoch": 0.025664, + "grad_norm": 0.15951215332821633, + "learning_rate": 9.998194407269887e-05, + "loss": 0.0397, + "step": 8020 + }, + { + "epoch": 0.025728, + "grad_norm": 0.30023143474409775, + "learning_rate": 9.998185391136552e-05, + "loss": 0.0369, + "step": 8040 + }, + { + "epoch": 0.025792, + "grad_norm": 0.21837937223919168, + "learning_rate": 9.998176352552535e-05, + "loss": 0.0396, + "step": 8060 + }, + { + "epoch": 0.025856, + "grad_norm": 0.17674465551606347, + "learning_rate": 9.998167291517874e-05, + "loss": 0.0429, + "step": 8080 + }, + { + "epoch": 0.02592, + "grad_norm": 0.16606144926047411, + "learning_rate": 9.998158208032615e-05, + "loss": 0.0405, + "step": 8100 + }, + { + "epoch": 0.025984, + "grad_norm": 0.18430384947600362, + "learning_rate": 9.998149102096795e-05, + "loss": 0.0405, + "step": 8120 + }, + { + "epoch": 0.026048, + "grad_norm": 0.17739809487671832, + "learning_rate": 9.998139973710455e-05, + "loss": 0.0379, + "step": 8140 + }, + { + "epoch": 0.026112, + "grad_norm": 0.13967700675768496, + "learning_rate": 9.998130822873637e-05, + "loss": 0.0367, + "step": 8160 + }, + { + "epoch": 0.026176, + "grad_norm": 0.15716813865496018, + "learning_rate": 9.998121649586382e-05, + "loss": 0.0392, + "step": 8180 + }, + { + "epoch": 0.02624, + "grad_norm": 0.13644381755385415, + "learning_rate": 9.99811245384873e-05, + "loss": 0.038, + "step": 8200 + }, + { + "epoch": 0.026304, + "grad_norm": 0.2134126230450536, + "learning_rate": 9.998103235660724e-05, + "loss": 0.0367, + "step": 8220 + }, + { + "epoch": 0.026368, + "grad_norm": 0.14877877092020975, + "learning_rate": 9.998093995022403e-05, + "loss": 0.0387, + "step": 8240 + }, + { + "epoch": 0.026432, + "grad_norm": 0.1574770279730747, + "learning_rate": 9.998084731933815e-05, + "loss": 0.039, + "step": 8260 + }, + { + "epoch": 0.026496, + "grad_norm": 0.21415945708222908, + "learning_rate": 9.998075446394992e-05, + "loss": 0.0322, + "step": 8280 + }, + { + "epoch": 0.02656, + "grad_norm": 0.1707549916825457, + "learning_rate": 9.998066138405983e-05, + "loss": 0.0381, + "step": 8300 + }, + { + "epoch": 0.026624, + "grad_norm": 0.18731397436693656, + "learning_rate": 9.998056807966827e-05, + "loss": 0.0437, + "step": 8320 + }, + { + "epoch": 0.026688, + "grad_norm": 0.2451754877072321, + "learning_rate": 9.998047455077564e-05, + "loss": 0.0408, + "step": 8340 + }, + { + "epoch": 0.026752, + "grad_norm": 0.15413952206375178, + "learning_rate": 9.998038079738241e-05, + "loss": 0.0431, + "step": 8360 + }, + { + "epoch": 0.026816, + "grad_norm": 0.1974076754612933, + "learning_rate": 9.998028681948896e-05, + "loss": 0.0394, + "step": 8380 + }, + { + "epoch": 0.02688, + "grad_norm": 0.19210860113879746, + "learning_rate": 9.998019261709574e-05, + "loss": 0.0432, + "step": 8400 + }, + { + "epoch": 0.026944, + "grad_norm": 0.14295884132936892, + "learning_rate": 9.998009819020315e-05, + "loss": 0.0401, + "step": 8420 + }, + { + "epoch": 0.027008, + "grad_norm": 0.17476265676964364, + "learning_rate": 9.998000353881162e-05, + "loss": 0.0371, + "step": 8440 + }, + { + "epoch": 0.027072, + "grad_norm": 0.15915165497965256, + "learning_rate": 9.997990866292157e-05, + "loss": 0.0392, + "step": 8460 + }, + { + "epoch": 0.027136, + "grad_norm": 0.13774363945823467, + "learning_rate": 9.997981356253345e-05, + "loss": 0.039, + "step": 8480 + }, + { + "epoch": 0.0272, + "grad_norm": 0.28410370337918944, + "learning_rate": 9.997971823764767e-05, + "loss": 0.0381, + "step": 8500 + }, + { + "epoch": 0.027264, + "grad_norm": 0.18124112354414013, + "learning_rate": 9.997962268826467e-05, + "loss": 0.0413, + "step": 8520 + }, + { + "epoch": 0.027328, + "grad_norm": 0.17321053945877202, + "learning_rate": 9.997952691438484e-05, + "loss": 0.0402, + "step": 8540 + }, + { + "epoch": 0.027392, + "grad_norm": 0.15767849101385414, + "learning_rate": 9.997943091600866e-05, + "loss": 0.0428, + "step": 8560 + }, + { + "epoch": 0.027456, + "grad_norm": 0.155224464417075, + "learning_rate": 9.997933469313654e-05, + "loss": 0.0383, + "step": 8580 + }, + { + "epoch": 0.02752, + "grad_norm": 0.14158622548456798, + "learning_rate": 9.99792382457689e-05, + "loss": 0.0382, + "step": 8600 + }, + { + "epoch": 0.027584, + "grad_norm": 0.2390044793087199, + "learning_rate": 9.99791415739062e-05, + "loss": 0.0392, + "step": 8620 + }, + { + "epoch": 0.027648, + "grad_norm": 0.21038678550703277, + "learning_rate": 9.997904467754885e-05, + "loss": 0.0394, + "step": 8640 + }, + { + "epoch": 0.027712, + "grad_norm": 0.24385811808529365, + "learning_rate": 9.99789475566973e-05, + "loss": 0.0373, + "step": 8660 + }, + { + "epoch": 0.027776, + "grad_norm": 0.3065415083307644, + "learning_rate": 9.997885021135199e-05, + "loss": 0.0385, + "step": 8680 + }, + { + "epoch": 0.02784, + "grad_norm": 0.2437758277740535, + "learning_rate": 9.997875264151333e-05, + "loss": 0.0375, + "step": 8700 + }, + { + "epoch": 0.027904, + "grad_norm": 0.22502419788610184, + "learning_rate": 9.997865484718178e-05, + "loss": 0.0411, + "step": 8720 + }, + { + "epoch": 0.027968, + "grad_norm": 0.12443278311736895, + "learning_rate": 9.997855682835778e-05, + "loss": 0.0397, + "step": 8740 + }, + { + "epoch": 0.028032, + "grad_norm": 0.21490873610268243, + "learning_rate": 9.997845858504178e-05, + "loss": 0.0383, + "step": 8760 + }, + { + "epoch": 0.028096, + "grad_norm": 0.14231181795440467, + "learning_rate": 9.99783601172342e-05, + "loss": 0.0368, + "step": 8780 + }, + { + "epoch": 0.02816, + "grad_norm": 0.1771891828770871, + "learning_rate": 9.997826142493548e-05, + "loss": 0.0392, + "step": 8800 + }, + { + "epoch": 0.028224, + "grad_norm": 0.18913571762356016, + "learning_rate": 9.997816250814608e-05, + "loss": 0.0362, + "step": 8820 + }, + { + "epoch": 0.028288, + "grad_norm": 0.14155565298737752, + "learning_rate": 9.997806336686644e-05, + "loss": 0.0374, + "step": 8840 + }, + { + "epoch": 0.028352, + "grad_norm": 0.1873384930843841, + "learning_rate": 9.9977964001097e-05, + "loss": 0.0354, + "step": 8860 + }, + { + "epoch": 0.028416, + "grad_norm": 0.1679821623780715, + "learning_rate": 9.99778644108382e-05, + "loss": 0.0419, + "step": 8880 + }, + { + "epoch": 0.02848, + "grad_norm": 0.1289712299068771, + "learning_rate": 9.997776459609051e-05, + "loss": 0.043, + "step": 8900 + }, + { + "epoch": 0.028544, + "grad_norm": 0.2378486193314008, + "learning_rate": 9.997766455685437e-05, + "loss": 0.0395, + "step": 8920 + }, + { + "epoch": 0.028608, + "grad_norm": 0.10613320877752269, + "learning_rate": 9.997756429313022e-05, + "loss": 0.0372, + "step": 8940 + }, + { + "epoch": 0.028672, + "grad_norm": 0.16267468353089712, + "learning_rate": 9.99774638049185e-05, + "loss": 0.0357, + "step": 8960 + }, + { + "epoch": 0.028736, + "grad_norm": 0.15059497685739207, + "learning_rate": 9.997736309221968e-05, + "loss": 0.0381, + "step": 8980 + }, + { + "epoch": 0.0288, + "grad_norm": 0.25015484198531057, + "learning_rate": 9.997726215503422e-05, + "loss": 0.0407, + "step": 9000 + }, + { + "epoch": 0.028864, + "grad_norm": 0.14633043222278574, + "learning_rate": 9.997716099336256e-05, + "loss": 0.0394, + "step": 9020 + }, + { + "epoch": 0.028928, + "grad_norm": 0.18167086780261996, + "learning_rate": 9.997705960720515e-05, + "loss": 0.0351, + "step": 9040 + }, + { + "epoch": 0.028992, + "grad_norm": 0.1350599645487261, + "learning_rate": 9.997695799656246e-05, + "loss": 0.0351, + "step": 9060 + }, + { + "epoch": 0.029056, + "grad_norm": 0.17440248677045417, + "learning_rate": 9.997685616143494e-05, + "loss": 0.0351, + "step": 9080 + }, + { + "epoch": 0.02912, + "grad_norm": 0.28942625686833, + "learning_rate": 9.997675410182304e-05, + "loss": 0.0413, + "step": 9100 + }, + { + "epoch": 0.029184, + "grad_norm": 0.19129688991313012, + "learning_rate": 9.997665181772721e-05, + "loss": 0.0369, + "step": 9120 + }, + { + "epoch": 0.029248, + "grad_norm": 0.15554348801230075, + "learning_rate": 9.997654930914795e-05, + "loss": 0.0358, + "step": 9140 + }, + { + "epoch": 0.029312, + "grad_norm": 0.25689889209454647, + "learning_rate": 9.997644657608567e-05, + "loss": 0.0357, + "step": 9160 + }, + { + "epoch": 0.029376, + "grad_norm": 0.27661745046985725, + "learning_rate": 9.997634361854088e-05, + "loss": 0.0373, + "step": 9180 + }, + { + "epoch": 0.02944, + "grad_norm": 0.2158671648172072, + "learning_rate": 9.997624043651401e-05, + "loss": 0.0347, + "step": 9200 + }, + { + "epoch": 0.029504, + "grad_norm": 0.192431325011243, + "learning_rate": 9.997613703000553e-05, + "loss": 0.0381, + "step": 9220 + }, + { + "epoch": 0.029568, + "grad_norm": 0.13233573902299448, + "learning_rate": 9.997603339901589e-05, + "loss": 0.0375, + "step": 9240 + }, + { + "epoch": 0.029632, + "grad_norm": 0.1359099622895592, + "learning_rate": 9.997592954354557e-05, + "loss": 0.039, + "step": 9260 + }, + { + "epoch": 0.029696, + "grad_norm": 0.1347300367526654, + "learning_rate": 9.997582546359505e-05, + "loss": 0.0369, + "step": 9280 + }, + { + "epoch": 0.02976, + "grad_norm": 0.14671149560533228, + "learning_rate": 9.997572115916479e-05, + "loss": 0.0374, + "step": 9300 + }, + { + "epoch": 0.029824, + "grad_norm": 0.13799985667916329, + "learning_rate": 9.997561663025526e-05, + "loss": 0.0363, + "step": 9320 + }, + { + "epoch": 0.029888, + "grad_norm": 0.10795437865207656, + "learning_rate": 9.99755118768669e-05, + "loss": 0.0353, + "step": 9340 + }, + { + "epoch": 0.029952, + "grad_norm": 0.19104706273275876, + "learning_rate": 9.997540689900023e-05, + "loss": 0.0361, + "step": 9360 + }, + { + "epoch": 0.030016, + "grad_norm": 0.21605051433949538, + "learning_rate": 9.997530169665567e-05, + "loss": 0.0406, + "step": 9380 + }, + { + "epoch": 0.03008, + "grad_norm": 0.19465621931296648, + "learning_rate": 9.997519626983374e-05, + "loss": 0.0371, + "step": 9400 + }, + { + "epoch": 0.030144, + "grad_norm": 0.19409049230766007, + "learning_rate": 9.997509061853487e-05, + "loss": 0.0379, + "step": 9420 + }, + { + "epoch": 0.030208, + "grad_norm": 0.15345945855081358, + "learning_rate": 9.997498474275957e-05, + "loss": 0.0383, + "step": 9440 + }, + { + "epoch": 0.030272, + "grad_norm": 0.16478275561675032, + "learning_rate": 9.99748786425083e-05, + "loss": 0.0383, + "step": 9460 + }, + { + "epoch": 0.030336, + "grad_norm": 0.2534259922530058, + "learning_rate": 9.997477231778153e-05, + "loss": 0.0355, + "step": 9480 + }, + { + "epoch": 0.0304, + "grad_norm": 0.18691581183631234, + "learning_rate": 9.997466576857974e-05, + "loss": 0.0343, + "step": 9500 + }, + { + "epoch": 0.030464, + "grad_norm": 0.1750066759437617, + "learning_rate": 9.997455899490344e-05, + "loss": 0.0365, + "step": 9520 + }, + { + "epoch": 0.030528, + "grad_norm": 0.11360784873379157, + "learning_rate": 9.997445199675306e-05, + "loss": 0.0361, + "step": 9540 + }, + { + "epoch": 0.030592, + "grad_norm": 0.12241917270243875, + "learning_rate": 9.997434477412913e-05, + "loss": 0.0374, + "step": 9560 + }, + { + "epoch": 0.030656, + "grad_norm": 0.14738392487864718, + "learning_rate": 9.997423732703208e-05, + "loss": 0.0338, + "step": 9580 + }, + { + "epoch": 0.03072, + "grad_norm": 0.15717460291283827, + "learning_rate": 9.997412965546242e-05, + "loss": 0.0362, + "step": 9600 + }, + { + "epoch": 0.030784, + "grad_norm": 0.17186483049770637, + "learning_rate": 9.997402175942066e-05, + "loss": 0.0375, + "step": 9620 + }, + { + "epoch": 0.030848, + "grad_norm": 0.17770942497995662, + "learning_rate": 9.997391363890722e-05, + "loss": 0.0386, + "step": 9640 + }, + { + "epoch": 0.030912, + "grad_norm": 0.12487498849791605, + "learning_rate": 9.997380529392265e-05, + "loss": 0.0344, + "step": 9660 + }, + { + "epoch": 0.030976, + "grad_norm": 0.22024664688168222, + "learning_rate": 9.99736967244674e-05, + "loss": 0.0403, + "step": 9680 + }, + { + "epoch": 0.03104, + "grad_norm": 0.1476486142479031, + "learning_rate": 9.997358793054197e-05, + "loss": 0.0387, + "step": 9700 + }, + { + "epoch": 0.031104, + "grad_norm": 0.17321645693606985, + "learning_rate": 9.997347891214684e-05, + "loss": 0.0382, + "step": 9720 + }, + { + "epoch": 0.031168, + "grad_norm": 0.17721735776386274, + "learning_rate": 9.997336966928252e-05, + "loss": 0.0322, + "step": 9740 + }, + { + "epoch": 0.031232, + "grad_norm": 0.1426680554326767, + "learning_rate": 9.997326020194946e-05, + "loss": 0.0344, + "step": 9760 + }, + { + "epoch": 0.031296, + "grad_norm": 0.17827572540853703, + "learning_rate": 9.99731505101482e-05, + "loss": 0.0345, + "step": 9780 + }, + { + "epoch": 0.03136, + "grad_norm": 0.13427295705265216, + "learning_rate": 9.99730405938792e-05, + "loss": 0.0361, + "step": 9800 + }, + { + "epoch": 0.031424, + "grad_norm": 0.14884346388748904, + "learning_rate": 9.997293045314297e-05, + "loss": 0.0384, + "step": 9820 + }, + { + "epoch": 0.031488, + "grad_norm": 0.15837656694355667, + "learning_rate": 9.997282008794e-05, + "loss": 0.0381, + "step": 9840 + }, + { + "epoch": 0.031552, + "grad_norm": 0.19414602443576, + "learning_rate": 9.997270949827077e-05, + "loss": 0.0346, + "step": 9860 + }, + { + "epoch": 0.031616, + "grad_norm": 0.12885342143744724, + "learning_rate": 9.997259868413579e-05, + "loss": 0.0402, + "step": 9880 + }, + { + "epoch": 0.03168, + "grad_norm": 0.12212417344477669, + "learning_rate": 9.997248764553559e-05, + "loss": 0.0395, + "step": 9900 + }, + { + "epoch": 0.031744, + "grad_norm": 0.19405930298485896, + "learning_rate": 9.997237638247061e-05, + "loss": 0.0344, + "step": 9920 + }, + { + "epoch": 0.031808, + "grad_norm": 0.13231332051530995, + "learning_rate": 9.997226489494139e-05, + "loss": 0.0359, + "step": 9940 + }, + { + "epoch": 0.031872, + "grad_norm": 0.27203157110079923, + "learning_rate": 9.997215318294842e-05, + "loss": 0.0393, + "step": 9960 + }, + { + "epoch": 0.031936, + "grad_norm": 0.20637275638754626, + "learning_rate": 9.99720412464922e-05, + "loss": 0.0372, + "step": 9980 + }, + { + "epoch": 0.032, + "grad_norm": 0.2958814261986718, + "learning_rate": 9.997192908557323e-05, + "loss": 0.0351, + "step": 10000 + }, + { + "epoch": 0.032064, + "grad_norm": 0.12364219590025602, + "learning_rate": 9.9971816700192e-05, + "loss": 0.0383, + "step": 10020 + }, + { + "epoch": 0.032128, + "grad_norm": 0.17885229131911337, + "learning_rate": 9.997170409034905e-05, + "loss": 0.0431, + "step": 10040 + }, + { + "epoch": 0.032192, + "grad_norm": 0.13989907836322624, + "learning_rate": 9.997159125604487e-05, + "loss": 0.0429, + "step": 10060 + }, + { + "epoch": 0.032256, + "grad_norm": 0.22538516370013106, + "learning_rate": 9.997147819727996e-05, + "loss": 0.0423, + "step": 10080 + }, + { + "epoch": 0.03232, + "grad_norm": 0.24523823598018699, + "learning_rate": 9.997136491405482e-05, + "loss": 0.0374, + "step": 10100 + }, + { + "epoch": 0.032384, + "grad_norm": 0.16486170543283055, + "learning_rate": 9.997125140636999e-05, + "loss": 0.0371, + "step": 10120 + }, + { + "epoch": 0.032448, + "grad_norm": 0.12265271421309225, + "learning_rate": 9.997113767422595e-05, + "loss": 0.0365, + "step": 10140 + }, + { + "epoch": 0.032512, + "grad_norm": 0.10543952055200029, + "learning_rate": 9.997102371762321e-05, + "loss": 0.0381, + "step": 10160 + }, + { + "epoch": 0.032576, + "grad_norm": 0.13709696412143763, + "learning_rate": 9.997090953656232e-05, + "loss": 0.038, + "step": 10180 + }, + { + "epoch": 0.03264, + "grad_norm": 0.20515581508288788, + "learning_rate": 9.997079513104375e-05, + "loss": 0.0357, + "step": 10200 + }, + { + "epoch": 0.032704, + "grad_norm": 0.11840899918486995, + "learning_rate": 9.997068050106803e-05, + "loss": 0.0386, + "step": 10220 + }, + { + "epoch": 0.032768, + "grad_norm": 0.15197883471104304, + "learning_rate": 9.997056564663568e-05, + "loss": 0.0397, + "step": 10240 + }, + { + "epoch": 0.032832, + "grad_norm": 0.11968339947972821, + "learning_rate": 9.997045056774721e-05, + "loss": 0.0387, + "step": 10260 + }, + { + "epoch": 0.032896, + "grad_norm": 0.11173079165222735, + "learning_rate": 9.997033526440313e-05, + "loss": 0.0366, + "step": 10280 + }, + { + "epoch": 0.03296, + "grad_norm": 0.21754134746478943, + "learning_rate": 9.997021973660399e-05, + "loss": 0.038, + "step": 10300 + }, + { + "epoch": 0.033024, + "grad_norm": 0.11098360085700987, + "learning_rate": 9.997010398435027e-05, + "loss": 0.0349, + "step": 10320 + }, + { + "epoch": 0.033088, + "grad_norm": 0.18338490217913303, + "learning_rate": 9.99699880076425e-05, + "loss": 0.0348, + "step": 10340 + }, + { + "epoch": 0.033152, + "grad_norm": 0.125163022746929, + "learning_rate": 9.99698718064812e-05, + "loss": 0.0373, + "step": 10360 + }, + { + "epoch": 0.033216, + "grad_norm": 0.15270035747162478, + "learning_rate": 9.996975538086692e-05, + "loss": 0.0412, + "step": 10380 + }, + { + "epoch": 0.03328, + "grad_norm": 0.21127911707115204, + "learning_rate": 9.996963873080013e-05, + "loss": 0.0402, + "step": 10400 + }, + { + "epoch": 0.033344, + "grad_norm": 0.20163107835361802, + "learning_rate": 9.996952185628142e-05, + "loss": 0.0404, + "step": 10420 + }, + { + "epoch": 0.033408, + "grad_norm": 0.16634817109058345, + "learning_rate": 9.996940475731126e-05, + "loss": 0.0345, + "step": 10440 + }, + { + "epoch": 0.033472, + "grad_norm": 0.15359154199442276, + "learning_rate": 9.99692874338902e-05, + "loss": 0.0385, + "step": 10460 + }, + { + "epoch": 0.033536, + "grad_norm": 0.13446380160504812, + "learning_rate": 9.996916988601875e-05, + "loss": 0.0394, + "step": 10480 + }, + { + "epoch": 0.0336, + "grad_norm": 0.13905733768668546, + "learning_rate": 9.996905211369748e-05, + "loss": 0.0349, + "step": 10500 + }, + { + "epoch": 0.033664, + "grad_norm": 0.12568947357166596, + "learning_rate": 9.996893411692685e-05, + "loss": 0.0361, + "step": 10520 + }, + { + "epoch": 0.033728, + "grad_norm": 0.14670008087388123, + "learning_rate": 9.996881589570746e-05, + "loss": 0.0361, + "step": 10540 + }, + { + "epoch": 0.033792, + "grad_norm": 0.20401259733301266, + "learning_rate": 9.99686974500398e-05, + "loss": 0.0373, + "step": 10560 + }, + { + "epoch": 0.033856, + "grad_norm": 0.13135706559737265, + "learning_rate": 9.996857877992441e-05, + "loss": 0.0385, + "step": 10580 + }, + { + "epoch": 0.03392, + "grad_norm": 0.1589052289554786, + "learning_rate": 9.996845988536184e-05, + "loss": 0.0368, + "step": 10600 + }, + { + "epoch": 0.033984, + "grad_norm": 0.11622772158064934, + "learning_rate": 9.996834076635258e-05, + "loss": 0.0408, + "step": 10620 + }, + { + "epoch": 0.034048, + "grad_norm": 0.11941090947601589, + "learning_rate": 9.996822142289722e-05, + "loss": 0.0447, + "step": 10640 + }, + { + "epoch": 0.034112, + "grad_norm": 0.15993170096621553, + "learning_rate": 9.996810185499625e-05, + "loss": 0.042, + "step": 10660 + }, + { + "epoch": 0.034176, + "grad_norm": 0.11993579024688374, + "learning_rate": 9.996798206265025e-05, + "loss": 0.04, + "step": 10680 + }, + { + "epoch": 0.03424, + "grad_norm": 0.16345538737576662, + "learning_rate": 9.996786204585973e-05, + "loss": 0.0366, + "step": 10700 + }, + { + "epoch": 0.034304, + "grad_norm": 0.1517473128315017, + "learning_rate": 9.996774180462523e-05, + "loss": 0.038, + "step": 10720 + }, + { + "epoch": 0.034368, + "grad_norm": 0.15938278679162105, + "learning_rate": 9.996762133894731e-05, + "loss": 0.0369, + "step": 10740 + }, + { + "epoch": 0.034432, + "grad_norm": 0.20983278814290815, + "learning_rate": 9.996750064882647e-05, + "loss": 0.0367, + "step": 10760 + }, + { + "epoch": 0.034496, + "grad_norm": 0.17542033624357714, + "learning_rate": 9.99673797342633e-05, + "loss": 0.0394, + "step": 10780 + }, + { + "epoch": 0.03456, + "grad_norm": 0.14427713679071308, + "learning_rate": 9.996725859525831e-05, + "loss": 0.0397, + "step": 10800 + }, + { + "epoch": 0.034624, + "grad_norm": 0.22470113899554028, + "learning_rate": 9.996713723181206e-05, + "loss": 0.0401, + "step": 10820 + }, + { + "epoch": 0.034688, + "grad_norm": 0.1279397128228936, + "learning_rate": 9.996701564392508e-05, + "loss": 0.0344, + "step": 10840 + }, + { + "epoch": 0.034752, + "grad_norm": 0.20466611619925004, + "learning_rate": 9.996689383159794e-05, + "loss": 0.0377, + "step": 10860 + }, + { + "epoch": 0.034816, + "grad_norm": 0.13577572864114776, + "learning_rate": 9.996677179483117e-05, + "loss": 0.0378, + "step": 10880 + }, + { + "epoch": 0.03488, + "grad_norm": 0.15008175303778848, + "learning_rate": 9.996664953362533e-05, + "loss": 0.0376, + "step": 10900 + }, + { + "epoch": 0.034944, + "grad_norm": 0.11314388449011832, + "learning_rate": 9.996652704798095e-05, + "loss": 0.0387, + "step": 10920 + }, + { + "epoch": 0.035008, + "grad_norm": 0.11986157460647344, + "learning_rate": 9.996640433789858e-05, + "loss": 0.0364, + "step": 10940 + }, + { + "epoch": 0.035072, + "grad_norm": 0.1463074937022462, + "learning_rate": 9.99662814033788e-05, + "loss": 0.0367, + "step": 10960 + }, + { + "epoch": 0.035136, + "grad_norm": 0.15043332572236226, + "learning_rate": 9.996615824442215e-05, + "loss": 0.0396, + "step": 10980 + }, + { + "epoch": 0.0352, + "grad_norm": 0.10099443767247628, + "learning_rate": 9.996603486102918e-05, + "loss": 0.0375, + "step": 11000 + }, + { + "epoch": 0.035264, + "grad_norm": 0.15083972003614504, + "learning_rate": 9.996591125320043e-05, + "loss": 0.0351, + "step": 11020 + }, + { + "epoch": 0.035328, + "grad_norm": 0.1644775886015293, + "learning_rate": 9.996578742093646e-05, + "loss": 0.0339, + "step": 11040 + }, + { + "epoch": 0.035392, + "grad_norm": 0.13842930352494692, + "learning_rate": 9.996566336423784e-05, + "loss": 0.0351, + "step": 11060 + }, + { + "epoch": 0.035456, + "grad_norm": 0.2209448933078816, + "learning_rate": 9.996553908310514e-05, + "loss": 0.0369, + "step": 11080 + }, + { + "epoch": 0.03552, + "grad_norm": 0.1155365569576674, + "learning_rate": 9.996541457753888e-05, + "loss": 0.0373, + "step": 11100 + }, + { + "epoch": 0.035584, + "grad_norm": 0.2853021855451896, + "learning_rate": 9.996528984753964e-05, + "loss": 0.0378, + "step": 11120 + }, + { + "epoch": 0.035648, + "grad_norm": 0.20305993618994458, + "learning_rate": 9.9965164893108e-05, + "loss": 0.0354, + "step": 11140 + }, + { + "epoch": 0.035712, + "grad_norm": 0.17255199154270306, + "learning_rate": 9.996503971424446e-05, + "loss": 0.0367, + "step": 11160 + }, + { + "epoch": 0.035776, + "grad_norm": 0.12756373090930814, + "learning_rate": 9.996491431094966e-05, + "loss": 0.0346, + "step": 11180 + }, + { + "epoch": 0.03584, + "grad_norm": 0.15187899484098727, + "learning_rate": 9.99647886832241e-05, + "loss": 0.0375, + "step": 11200 + }, + { + "epoch": 0.035904, + "grad_norm": 0.11290906265087604, + "learning_rate": 9.996466283106837e-05, + "loss": 0.0338, + "step": 11220 + }, + { + "epoch": 0.035968, + "grad_norm": 0.26856536016750016, + "learning_rate": 9.996453675448306e-05, + "loss": 0.0335, + "step": 11240 + }, + { + "epoch": 0.036032, + "grad_norm": 0.18351464787025443, + "learning_rate": 9.996441045346869e-05, + "loss": 0.0353, + "step": 11260 + }, + { + "epoch": 0.036096, + "grad_norm": 0.1398278453650458, + "learning_rate": 9.996428392802586e-05, + "loss": 0.0393, + "step": 11280 + }, + { + "epoch": 0.03616, + "grad_norm": 0.22785679669373496, + "learning_rate": 9.996415717815511e-05, + "loss": 0.0378, + "step": 11300 + }, + { + "epoch": 0.036224, + "grad_norm": 0.1491796734652503, + "learning_rate": 9.996403020385705e-05, + "loss": 0.032, + "step": 11320 + }, + { + "epoch": 0.036288, + "grad_norm": 0.13241352469233025, + "learning_rate": 9.996390300513221e-05, + "loss": 0.036, + "step": 11340 + }, + { + "epoch": 0.036352, + "grad_norm": 0.18554557508019476, + "learning_rate": 9.996377558198118e-05, + "loss": 0.0397, + "step": 11360 + }, + { + "epoch": 0.036416, + "grad_norm": 0.12922688276727173, + "learning_rate": 9.996364793440454e-05, + "loss": 0.0361, + "step": 11380 + }, + { + "epoch": 0.03648, + "grad_norm": 0.1331412041661368, + "learning_rate": 9.996352006240284e-05, + "loss": 0.0406, + "step": 11400 + }, + { + "epoch": 0.036544, + "grad_norm": 0.18744377168059323, + "learning_rate": 9.996339196597668e-05, + "loss": 0.0358, + "step": 11420 + }, + { + "epoch": 0.036608, + "grad_norm": 0.1250934733546199, + "learning_rate": 9.996326364512662e-05, + "loss": 0.0371, + "step": 11440 + }, + { + "epoch": 0.036672, + "grad_norm": 0.17539652306727843, + "learning_rate": 9.996313509985324e-05, + "loss": 0.0354, + "step": 11460 + }, + { + "epoch": 0.036736, + "grad_norm": 0.15043986350528155, + "learning_rate": 9.996300633015712e-05, + "loss": 0.0362, + "step": 11480 + }, + { + "epoch": 0.0368, + "grad_norm": 0.1501145494743168, + "learning_rate": 9.996287733603884e-05, + "loss": 0.0381, + "step": 11500 + }, + { + "epoch": 0.036864, + "grad_norm": 0.14278778458523822, + "learning_rate": 9.996274811749896e-05, + "loss": 0.0368, + "step": 11520 + }, + { + "epoch": 0.036928, + "grad_norm": 0.11566950436731031, + "learning_rate": 9.996261867453808e-05, + "loss": 0.0378, + "step": 11540 + }, + { + "epoch": 0.036992, + "grad_norm": 0.10269585619883936, + "learning_rate": 9.996248900715678e-05, + "loss": 0.0345, + "step": 11560 + }, + { + "epoch": 0.037056, + "grad_norm": 0.1189438585435548, + "learning_rate": 9.996235911535563e-05, + "loss": 0.033, + "step": 11580 + }, + { + "epoch": 0.03712, + "grad_norm": 0.14472258498493745, + "learning_rate": 9.996222899913525e-05, + "loss": 0.0369, + "step": 11600 + }, + { + "epoch": 0.037184, + "grad_norm": 0.14041796430376102, + "learning_rate": 9.996209865849617e-05, + "loss": 0.0376, + "step": 11620 + }, + { + "epoch": 0.037248, + "grad_norm": 0.17879560074533937, + "learning_rate": 9.9961968093439e-05, + "loss": 0.035, + "step": 11640 + }, + { + "epoch": 0.037312, + "grad_norm": 0.13379985074251763, + "learning_rate": 9.996183730396435e-05, + "loss": 0.0378, + "step": 11660 + }, + { + "epoch": 0.037376, + "grad_norm": 0.16341117755702472, + "learning_rate": 9.996170629007277e-05, + "loss": 0.0371, + "step": 11680 + }, + { + "epoch": 0.03744, + "grad_norm": 0.10811802761049413, + "learning_rate": 9.996157505176489e-05, + "loss": 0.0363, + "step": 11700 + }, + { + "epoch": 0.037504, + "grad_norm": 0.19865000377822636, + "learning_rate": 9.996144358904125e-05, + "loss": 0.0415, + "step": 11720 + }, + { + "epoch": 0.037568, + "grad_norm": 0.1743233130132439, + "learning_rate": 9.996131190190247e-05, + "loss": 0.0385, + "step": 11740 + }, + { + "epoch": 0.037632, + "grad_norm": 0.1937270097260595, + "learning_rate": 9.996117999034915e-05, + "loss": 0.0377, + "step": 11760 + }, + { + "epoch": 0.037696, + "grad_norm": 0.1389416843095414, + "learning_rate": 9.996104785438186e-05, + "loss": 0.0411, + "step": 11780 + }, + { + "epoch": 0.03776, + "grad_norm": 0.22105704341503526, + "learning_rate": 9.996091549400119e-05, + "loss": 0.0377, + "step": 11800 + }, + { + "epoch": 0.037824, + "grad_norm": 0.2151287638414484, + "learning_rate": 9.996078290920777e-05, + "loss": 0.0345, + "step": 11820 + }, + { + "epoch": 0.037888, + "grad_norm": 0.29197368667274337, + "learning_rate": 9.996065010000216e-05, + "loss": 0.0387, + "step": 11840 + }, + { + "epoch": 0.037952, + "grad_norm": 0.1858849790645508, + "learning_rate": 9.996051706638499e-05, + "loss": 0.0378, + "step": 11860 + }, + { + "epoch": 0.038016, + "grad_norm": 0.1490594040314182, + "learning_rate": 9.99603838083568e-05, + "loss": 0.0389, + "step": 11880 + }, + { + "epoch": 0.03808, + "grad_norm": 0.12239341154660383, + "learning_rate": 9.996025032591825e-05, + "loss": 0.0394, + "step": 11900 + }, + { + "epoch": 0.038144, + "grad_norm": 0.16957183037525053, + "learning_rate": 9.996011661906991e-05, + "loss": 0.035, + "step": 11920 + }, + { + "epoch": 0.038208, + "grad_norm": 0.2487795404791005, + "learning_rate": 9.99599826878124e-05, + "loss": 0.0347, + "step": 11940 + }, + { + "epoch": 0.038272, + "grad_norm": 0.15786567308373664, + "learning_rate": 9.995984853214628e-05, + "loss": 0.0349, + "step": 11960 + }, + { + "epoch": 0.038336, + "grad_norm": 0.1668357575367027, + "learning_rate": 9.995971415207219e-05, + "loss": 0.0402, + "step": 11980 + }, + { + "epoch": 0.0384, + "grad_norm": 0.09726897130178946, + "learning_rate": 9.995957954759071e-05, + "loss": 0.0381, + "step": 12000 + }, + { + "epoch": 0.038464, + "grad_norm": 0.12913885345677378, + "learning_rate": 9.995944471870249e-05, + "loss": 0.0325, + "step": 12020 + }, + { + "epoch": 0.038528, + "grad_norm": 0.12487863571128327, + "learning_rate": 9.995930966540809e-05, + "loss": 0.0344, + "step": 12040 + }, + { + "epoch": 0.038592, + "grad_norm": 0.10778044087186125, + "learning_rate": 9.995917438770813e-05, + "loss": 0.0385, + "step": 12060 + }, + { + "epoch": 0.038656, + "grad_norm": 0.11646116029395205, + "learning_rate": 9.99590388856032e-05, + "loss": 0.037, + "step": 12080 + }, + { + "epoch": 0.03872, + "grad_norm": 0.15839744786525423, + "learning_rate": 9.995890315909395e-05, + "loss": 0.0409, + "step": 12100 + }, + { + "epoch": 0.038784, + "grad_norm": 0.09747908143672461, + "learning_rate": 9.995876720818096e-05, + "loss": 0.0355, + "step": 12120 + }, + { + "epoch": 0.038848, + "grad_norm": 0.096243765031423, + "learning_rate": 9.995863103286484e-05, + "loss": 0.0362, + "step": 12140 + }, + { + "epoch": 0.038912, + "grad_norm": 0.1607641957298064, + "learning_rate": 9.995849463314622e-05, + "loss": 0.0355, + "step": 12160 + }, + { + "epoch": 0.038976, + "grad_norm": 0.16114503053227267, + "learning_rate": 9.99583580090257e-05, + "loss": 0.0372, + "step": 12180 + }, + { + "epoch": 0.03904, + "grad_norm": 0.10821751829934675, + "learning_rate": 9.995822116050389e-05, + "loss": 0.0353, + "step": 12200 + }, + { + "epoch": 0.039104, + "grad_norm": 0.1446137506165817, + "learning_rate": 9.99580840875814e-05, + "loss": 0.0353, + "step": 12220 + }, + { + "epoch": 0.039168, + "grad_norm": 0.21004650125979177, + "learning_rate": 9.995794679025887e-05, + "loss": 0.0385, + "step": 12240 + }, + { + "epoch": 0.039232, + "grad_norm": 0.13314580921705313, + "learning_rate": 9.99578092685369e-05, + "loss": 0.0368, + "step": 12260 + }, + { + "epoch": 0.039296, + "grad_norm": 0.1784081782963823, + "learning_rate": 9.99576715224161e-05, + "loss": 0.0359, + "step": 12280 + }, + { + "epoch": 0.03936, + "grad_norm": 0.11129020315145771, + "learning_rate": 9.995753355189711e-05, + "loss": 0.0404, + "step": 12300 + }, + { + "epoch": 0.039424, + "grad_norm": 0.130097130773632, + "learning_rate": 9.995739535698054e-05, + "loss": 0.0348, + "step": 12320 + }, + { + "epoch": 0.039488, + "grad_norm": 0.10320651192456066, + "learning_rate": 9.9957256937667e-05, + "loss": 0.0367, + "step": 12340 + }, + { + "epoch": 0.039552, + "grad_norm": 0.15977153292603796, + "learning_rate": 9.995711829395711e-05, + "loss": 0.0397, + "step": 12360 + }, + { + "epoch": 0.039616, + "grad_norm": 0.11581780973697793, + "learning_rate": 9.995697942585153e-05, + "loss": 0.0356, + "step": 12380 + }, + { + "epoch": 0.03968, + "grad_norm": 0.18987360767579936, + "learning_rate": 9.995684033335084e-05, + "loss": 0.0384, + "step": 12400 + }, + { + "epoch": 0.039744, + "grad_norm": 0.35354079442981895, + "learning_rate": 9.995670101645569e-05, + "loss": 0.0357, + "step": 12420 + }, + { + "epoch": 0.039808, + "grad_norm": 0.1314357442217599, + "learning_rate": 9.99565614751667e-05, + "loss": 0.0387, + "step": 12440 + }, + { + "epoch": 0.039872, + "grad_norm": 0.13040224392602923, + "learning_rate": 9.995642170948449e-05, + "loss": 0.0351, + "step": 12460 + }, + { + "epoch": 0.039936, + "grad_norm": 0.11769237969923194, + "learning_rate": 9.995628171940967e-05, + "loss": 0.0405, + "step": 12480 + }, + { + "epoch": 0.04, + "grad_norm": 0.11511343998788336, + "learning_rate": 9.995614150494293e-05, + "loss": 0.0352, + "step": 12500 + }, + { + "epoch": 0.040064, + "grad_norm": 0.18369832965536284, + "learning_rate": 9.995600106608483e-05, + "loss": 0.0339, + "step": 12520 + }, + { + "epoch": 0.040128, + "grad_norm": 0.12883805811225815, + "learning_rate": 9.995586040283604e-05, + "loss": 0.0342, + "step": 12540 + }, + { + "epoch": 0.040192, + "grad_norm": 0.15636426694886893, + "learning_rate": 9.99557195151972e-05, + "loss": 0.0346, + "step": 12560 + }, + { + "epoch": 0.040256, + "grad_norm": 0.18796133373192, + "learning_rate": 9.995557840316891e-05, + "loss": 0.032, + "step": 12580 + }, + { + "epoch": 0.04032, + "grad_norm": 0.0909389337159971, + "learning_rate": 9.995543706675184e-05, + "loss": 0.0371, + "step": 12600 + }, + { + "epoch": 0.040384, + "grad_norm": 0.12759561965612742, + "learning_rate": 9.995529550594658e-05, + "loss": 0.0361, + "step": 12620 + }, + { + "epoch": 0.040448, + "grad_norm": 0.242362791840337, + "learning_rate": 9.995515372075379e-05, + "loss": 0.0353, + "step": 12640 + }, + { + "epoch": 0.040512, + "grad_norm": 0.11302460623556791, + "learning_rate": 9.995501171117412e-05, + "loss": 0.0334, + "step": 12660 + }, + { + "epoch": 0.040576, + "grad_norm": 0.1050608081530142, + "learning_rate": 9.995486947720821e-05, + "loss": 0.036, + "step": 12680 + }, + { + "epoch": 0.04064, + "grad_norm": 0.13333935924497028, + "learning_rate": 9.995472701885667e-05, + "loss": 0.0336, + "step": 12700 + }, + { + "epoch": 0.040704, + "grad_norm": 0.11206509928816853, + "learning_rate": 9.995458433612015e-05, + "loss": 0.0354, + "step": 12720 + }, + { + "epoch": 0.040768, + "grad_norm": 0.12775654590168317, + "learning_rate": 9.99544414289993e-05, + "loss": 0.0367, + "step": 12740 + }, + { + "epoch": 0.040832, + "grad_norm": 0.3598468765640993, + "learning_rate": 9.995429829749474e-05, + "loss": 0.0344, + "step": 12760 + }, + { + "epoch": 0.040896, + "grad_norm": 0.21642176816991943, + "learning_rate": 9.995415494160716e-05, + "loss": 0.0325, + "step": 12780 + }, + { + "epoch": 0.04096, + "grad_norm": 0.10037960341475474, + "learning_rate": 9.995401136133715e-05, + "loss": 0.034, + "step": 12800 + }, + { + "epoch": 0.041024, + "grad_norm": 0.09821299251650809, + "learning_rate": 9.995386755668538e-05, + "loss": 0.0377, + "step": 12820 + }, + { + "epoch": 0.041088, + "grad_norm": 0.10789084474365261, + "learning_rate": 9.995372352765251e-05, + "loss": 0.0329, + "step": 12840 + }, + { + "epoch": 0.041152, + "grad_norm": 0.1374612015625052, + "learning_rate": 9.995357927423917e-05, + "loss": 0.032, + "step": 12860 + }, + { + "epoch": 0.041216, + "grad_norm": 0.17383480275837235, + "learning_rate": 9.995343479644601e-05, + "loss": 0.035, + "step": 12880 + }, + { + "epoch": 0.04128, + "grad_norm": 0.1852959768761994, + "learning_rate": 9.995329009427367e-05, + "loss": 0.0378, + "step": 12900 + }, + { + "epoch": 0.041344, + "grad_norm": 0.09687821646109633, + "learning_rate": 9.995314516772282e-05, + "loss": 0.0354, + "step": 12920 + }, + { + "epoch": 0.041408, + "grad_norm": 0.1267931723534882, + "learning_rate": 9.99530000167941e-05, + "loss": 0.0352, + "step": 12940 + }, + { + "epoch": 0.041472, + "grad_norm": 0.1682029816171222, + "learning_rate": 9.995285464148815e-05, + "loss": 0.036, + "step": 12960 + }, + { + "epoch": 0.041536, + "grad_norm": 0.1579164970140408, + "learning_rate": 9.995270904180564e-05, + "loss": 0.0361, + "step": 12980 + }, + { + "epoch": 0.0416, + "grad_norm": 0.19650454272083956, + "learning_rate": 9.995256321774722e-05, + "loss": 0.035, + "step": 13000 + }, + { + "epoch": 0.041664, + "grad_norm": 0.10202358779641256, + "learning_rate": 9.995241716931353e-05, + "loss": 0.0321, + "step": 13020 + }, + { + "epoch": 0.041728, + "grad_norm": 0.09262351926472484, + "learning_rate": 9.995227089650527e-05, + "loss": 0.0339, + "step": 13040 + }, + { + "epoch": 0.041792, + "grad_norm": 0.12508529077650649, + "learning_rate": 9.995212439932304e-05, + "loss": 0.038, + "step": 13060 + }, + { + "epoch": 0.041856, + "grad_norm": 0.17747845396967746, + "learning_rate": 9.995197767776754e-05, + "loss": 0.0366, + "step": 13080 + }, + { + "epoch": 0.04192, + "grad_norm": 0.13880164195357586, + "learning_rate": 9.99518307318394e-05, + "loss": 0.0364, + "step": 13100 + }, + { + "epoch": 0.041984, + "grad_norm": 0.16069790135376683, + "learning_rate": 9.99516835615393e-05, + "loss": 0.0336, + "step": 13120 + }, + { + "epoch": 0.042048, + "grad_norm": 0.26908374553814524, + "learning_rate": 9.99515361668679e-05, + "loss": 0.0399, + "step": 13140 + }, + { + "epoch": 0.042112, + "grad_norm": 0.09586578069757219, + "learning_rate": 9.995138854782586e-05, + "loss": 0.0378, + "step": 13160 + }, + { + "epoch": 0.042176, + "grad_norm": 0.09558556058525725, + "learning_rate": 9.995124070441384e-05, + "loss": 0.0381, + "step": 13180 + }, + { + "epoch": 0.04224, + "grad_norm": 0.10318675545129466, + "learning_rate": 9.995109263663249e-05, + "loss": 0.0335, + "step": 13200 + }, + { + "epoch": 0.042304, + "grad_norm": 0.12935872362020825, + "learning_rate": 9.99509443444825e-05, + "loss": 0.0334, + "step": 13220 + }, + { + "epoch": 0.042368, + "grad_norm": 0.10901057323532523, + "learning_rate": 9.995079582796451e-05, + "loss": 0.0382, + "step": 13240 + }, + { + "epoch": 0.042432, + "grad_norm": 0.17519822237595992, + "learning_rate": 9.995064708707924e-05, + "loss": 0.0364, + "step": 13260 + }, + { + "epoch": 0.042496, + "grad_norm": 0.10232396388378308, + "learning_rate": 9.995049812182729e-05, + "loss": 0.0334, + "step": 13280 + }, + { + "epoch": 0.04256, + "grad_norm": 0.0868475409499182, + "learning_rate": 9.995034893220936e-05, + "loss": 0.0327, + "step": 13300 + }, + { + "epoch": 0.042624, + "grad_norm": 0.18061253806674193, + "learning_rate": 9.995019951822612e-05, + "loss": 0.0356, + "step": 13320 + }, + { + "epoch": 0.042688, + "grad_norm": 0.13032878041209478, + "learning_rate": 9.995004987987827e-05, + "loss": 0.035, + "step": 13340 + }, + { + "epoch": 0.042752, + "grad_norm": 0.15027911411969905, + "learning_rate": 9.994990001716641e-05, + "loss": 0.0371, + "step": 13360 + }, + { + "epoch": 0.042816, + "grad_norm": 0.191858693639075, + "learning_rate": 9.994974993009128e-05, + "loss": 0.0374, + "step": 13380 + }, + { + "epoch": 0.04288, + "grad_norm": 0.09444734453085897, + "learning_rate": 9.994959961865353e-05, + "loss": 0.0336, + "step": 13400 + }, + { + "epoch": 0.042944, + "grad_norm": 0.14032247873655493, + "learning_rate": 9.994944908285383e-05, + "loss": 0.032, + "step": 13420 + }, + { + "epoch": 0.043008, + "grad_norm": 0.1461852221174766, + "learning_rate": 9.994929832269286e-05, + "loss": 0.0345, + "step": 13440 + }, + { + "epoch": 0.043072, + "grad_norm": 0.12743633095403814, + "learning_rate": 9.994914733817131e-05, + "loss": 0.0381, + "step": 13460 + }, + { + "epoch": 0.043136, + "grad_norm": 0.13410929776721572, + "learning_rate": 9.994899612928983e-05, + "loss": 0.0357, + "step": 13480 + }, + { + "epoch": 0.0432, + "grad_norm": 0.11803197962502458, + "learning_rate": 9.994884469604912e-05, + "loss": 0.032, + "step": 13500 + }, + { + "epoch": 0.043264, + "grad_norm": 0.1413821949222677, + "learning_rate": 9.994869303844987e-05, + "loss": 0.0367, + "step": 13520 + }, + { + "epoch": 0.043328, + "grad_norm": 0.09181712940747241, + "learning_rate": 9.994854115649274e-05, + "loss": 0.035, + "step": 13540 + }, + { + "epoch": 0.043392, + "grad_norm": 0.23561126650673253, + "learning_rate": 9.994838905017842e-05, + "loss": 0.0327, + "step": 13560 + }, + { + "epoch": 0.043456, + "grad_norm": 0.11374922712779337, + "learning_rate": 9.994823671950759e-05, + "loss": 0.0413, + "step": 13580 + }, + { + "epoch": 0.04352, + "grad_norm": 0.10043334848056061, + "learning_rate": 9.994808416448095e-05, + "loss": 0.0376, + "step": 13600 + }, + { + "epoch": 0.043584, + "grad_norm": 0.17734292779521313, + "learning_rate": 9.994793138509917e-05, + "loss": 0.0372, + "step": 13620 + }, + { + "epoch": 0.043648, + "grad_norm": 0.17902198716609471, + "learning_rate": 9.994777838136293e-05, + "loss": 0.0368, + "step": 13640 + }, + { + "epoch": 0.043712, + "grad_norm": 0.14588501975238247, + "learning_rate": 9.994762515327292e-05, + "loss": 0.0366, + "step": 13660 + }, + { + "epoch": 0.043776, + "grad_norm": 0.12233879784232168, + "learning_rate": 9.994747170082986e-05, + "loss": 0.0355, + "step": 13680 + }, + { + "epoch": 0.04384, + "grad_norm": 0.11398598235919205, + "learning_rate": 9.994731802403439e-05, + "loss": 0.0401, + "step": 13700 + }, + { + "epoch": 0.043904, + "grad_norm": 0.1847482394489853, + "learning_rate": 9.994716412288722e-05, + "loss": 0.0397, + "step": 13720 + }, + { + "epoch": 0.043968, + "grad_norm": 0.16329704379135582, + "learning_rate": 9.994700999738906e-05, + "loss": 0.0356, + "step": 13740 + }, + { + "epoch": 0.044032, + "grad_norm": 0.1160686238200498, + "learning_rate": 9.994685564754058e-05, + "loss": 0.0366, + "step": 13760 + }, + { + "epoch": 0.044096, + "grad_norm": 0.1372646910801222, + "learning_rate": 9.994670107334248e-05, + "loss": 0.0342, + "step": 13780 + }, + { + "epoch": 0.04416, + "grad_norm": 0.1310602022417257, + "learning_rate": 9.994654627479547e-05, + "loss": 0.0334, + "step": 13800 + }, + { + "epoch": 0.044224, + "grad_norm": 0.12200065350511755, + "learning_rate": 9.994639125190022e-05, + "loss": 0.0343, + "step": 13820 + }, + { + "epoch": 0.044288, + "grad_norm": 0.11151222748733197, + "learning_rate": 9.994623600465743e-05, + "loss": 0.0349, + "step": 13840 + }, + { + "epoch": 0.044352, + "grad_norm": 0.13968436398067596, + "learning_rate": 9.994608053306783e-05, + "loss": 0.0372, + "step": 13860 + }, + { + "epoch": 0.044416, + "grad_norm": 0.1647327315652881, + "learning_rate": 9.994592483713205e-05, + "loss": 0.0362, + "step": 13880 + }, + { + "epoch": 0.04448, + "grad_norm": 0.12389698187463422, + "learning_rate": 9.994576891685085e-05, + "loss": 0.0373, + "step": 13900 + }, + { + "epoch": 0.044544, + "grad_norm": 0.14478315360450453, + "learning_rate": 9.994561277222494e-05, + "loss": 0.0361, + "step": 13920 + }, + { + "epoch": 0.044608, + "grad_norm": 0.1055836990591201, + "learning_rate": 9.994545640325497e-05, + "loss": 0.0371, + "step": 13940 + }, + { + "epoch": 0.044672, + "grad_norm": 0.1356064229542883, + "learning_rate": 9.994529980994166e-05, + "loss": 0.0399, + "step": 13960 + }, + { + "epoch": 0.044736, + "grad_norm": 0.20308119362177499, + "learning_rate": 9.994514299228573e-05, + "loss": 0.0354, + "step": 13980 + }, + { + "epoch": 0.0448, + "grad_norm": 0.14355315034028093, + "learning_rate": 9.994498595028787e-05, + "loss": 0.0361, + "step": 14000 + }, + { + "epoch": 0.044864, + "grad_norm": 0.17914259544217181, + "learning_rate": 9.994482868394879e-05, + "loss": 0.0369, + "step": 14020 + }, + { + "epoch": 0.044928, + "grad_norm": 0.12291462569602868, + "learning_rate": 9.99446711932692e-05, + "loss": 0.0374, + "step": 14040 + }, + { + "epoch": 0.044992, + "grad_norm": 0.09498172058438609, + "learning_rate": 9.994451347824978e-05, + "loss": 0.0336, + "step": 14060 + }, + { + "epoch": 0.045056, + "grad_norm": 0.12742807010797344, + "learning_rate": 9.994435553889127e-05, + "loss": 0.0335, + "step": 14080 + }, + { + "epoch": 0.04512, + "grad_norm": 0.11235963487548496, + "learning_rate": 9.99441973751944e-05, + "loss": 0.0332, + "step": 14100 + }, + { + "epoch": 0.045184, + "grad_norm": 0.09177863575589583, + "learning_rate": 9.994403898715983e-05, + "loss": 0.0326, + "step": 14120 + }, + { + "epoch": 0.045248, + "grad_norm": 0.10782303549293681, + "learning_rate": 9.994388037478829e-05, + "loss": 0.0315, + "step": 14140 + }, + { + "epoch": 0.045312, + "grad_norm": 0.21523747883605912, + "learning_rate": 9.994372153808049e-05, + "loss": 0.0339, + "step": 14160 + }, + { + "epoch": 0.045376, + "grad_norm": 0.2380971217420063, + "learning_rate": 9.994356247703716e-05, + "loss": 0.0381, + "step": 14180 + }, + { + "epoch": 0.04544, + "grad_norm": 0.12826389598161775, + "learning_rate": 9.9943403191659e-05, + "loss": 0.0322, + "step": 14200 + }, + { + "epoch": 0.045504, + "grad_norm": 0.09826603130800766, + "learning_rate": 9.994324368194673e-05, + "loss": 0.0328, + "step": 14220 + }, + { + "epoch": 0.045568, + "grad_norm": 0.14196179017636382, + "learning_rate": 9.994308394790107e-05, + "loss": 0.0367, + "step": 14240 + }, + { + "epoch": 0.045632, + "grad_norm": 0.1591334904686358, + "learning_rate": 9.994292398952272e-05, + "loss": 0.0385, + "step": 14260 + }, + { + "epoch": 0.045696, + "grad_norm": 0.4321749968113467, + "learning_rate": 9.994276380681242e-05, + "loss": 0.0378, + "step": 14280 + }, + { + "epoch": 0.04576, + "grad_norm": 0.1301194571474574, + "learning_rate": 9.994260339977088e-05, + "loss": 0.0373, + "step": 14300 + }, + { + "epoch": 0.045824, + "grad_norm": 0.3169208377800778, + "learning_rate": 9.994244276839881e-05, + "loss": 0.0385, + "step": 14320 + }, + { + "epoch": 0.045888, + "grad_norm": 0.13982955942076783, + "learning_rate": 9.994228191269696e-05, + "loss": 0.0383, + "step": 14340 + }, + { + "epoch": 0.045952, + "grad_norm": 0.2025238367521379, + "learning_rate": 9.994212083266602e-05, + "loss": 0.0396, + "step": 14360 + }, + { + "epoch": 0.046016, + "grad_norm": 0.20905358691622072, + "learning_rate": 9.994195952830674e-05, + "loss": 0.0365, + "step": 14380 + }, + { + "epoch": 0.04608, + "grad_norm": 0.16460702466994437, + "learning_rate": 9.994179799961982e-05, + "loss": 0.0368, + "step": 14400 + }, + { + "epoch": 0.046144, + "grad_norm": 0.10136251861787189, + "learning_rate": 9.9941636246606e-05, + "loss": 0.0357, + "step": 14420 + }, + { + "epoch": 0.046208, + "grad_norm": 0.10912423146219984, + "learning_rate": 9.994147426926601e-05, + "loss": 0.0352, + "step": 14440 + }, + { + "epoch": 0.046272, + "grad_norm": 0.1518821927874365, + "learning_rate": 9.994131206760057e-05, + "loss": 0.0337, + "step": 14460 + }, + { + "epoch": 0.046336, + "grad_norm": 0.15836923553734494, + "learning_rate": 9.994114964161043e-05, + "loss": 0.0366, + "step": 14480 + }, + { + "epoch": 0.0464, + "grad_norm": 0.12628893714016717, + "learning_rate": 9.994098699129628e-05, + "loss": 0.0374, + "step": 14500 + }, + { + "epoch": 0.046464, + "grad_norm": 0.19217939962580072, + "learning_rate": 9.994082411665888e-05, + "loss": 0.0385, + "step": 14520 + }, + { + "epoch": 0.046528, + "grad_norm": 0.22718867155352931, + "learning_rate": 9.994066101769895e-05, + "loss": 0.036, + "step": 14540 + }, + { + "epoch": 0.046592, + "grad_norm": 0.1308520625037308, + "learning_rate": 9.994049769441722e-05, + "loss": 0.0395, + "step": 14560 + }, + { + "epoch": 0.046656, + "grad_norm": 0.129888659257907, + "learning_rate": 9.994033414681445e-05, + "loss": 0.0365, + "step": 14580 + }, + { + "epoch": 0.04672, + "grad_norm": 0.15055996336200528, + "learning_rate": 9.994017037489134e-05, + "loss": 0.0324, + "step": 14600 + }, + { + "epoch": 0.046784, + "grad_norm": 0.13747723026693087, + "learning_rate": 9.994000637864864e-05, + "loss": 0.0388, + "step": 14620 + }, + { + "epoch": 0.046848, + "grad_norm": 0.09469958422152938, + "learning_rate": 9.993984215808707e-05, + "loss": 0.035, + "step": 14640 + }, + { + "epoch": 0.046912, + "grad_norm": 0.13825783147489862, + "learning_rate": 9.99396777132074e-05, + "loss": 0.0352, + "step": 14660 + }, + { + "epoch": 0.046976, + "grad_norm": 0.17996395216847563, + "learning_rate": 9.993951304401036e-05, + "loss": 0.036, + "step": 14680 + }, + { + "epoch": 0.04704, + "grad_norm": 0.10124427782083134, + "learning_rate": 9.993934815049667e-05, + "loss": 0.036, + "step": 14700 + }, + { + "epoch": 0.047104, + "grad_norm": 0.1360054352481604, + "learning_rate": 9.99391830326671e-05, + "loss": 0.0359, + "step": 14720 + }, + { + "epoch": 0.047168, + "grad_norm": 0.2222718662485625, + "learning_rate": 9.993901769052235e-05, + "loss": 0.0391, + "step": 14740 + }, + { + "epoch": 0.047232, + "grad_norm": 0.1248655666639231, + "learning_rate": 9.99388521240632e-05, + "loss": 0.0342, + "step": 14760 + }, + { + "epoch": 0.047296, + "grad_norm": 0.17962824406238426, + "learning_rate": 9.993868633329038e-05, + "loss": 0.0334, + "step": 14780 + }, + { + "epoch": 0.04736, + "grad_norm": 0.12405559762782221, + "learning_rate": 9.993852031820464e-05, + "loss": 0.0352, + "step": 14800 + }, + { + "epoch": 0.047424, + "grad_norm": 0.19324765245046804, + "learning_rate": 9.993835407880671e-05, + "loss": 0.0365, + "step": 14820 + }, + { + "epoch": 0.047488, + "grad_norm": 0.10180881429208166, + "learning_rate": 9.993818761509734e-05, + "loss": 0.0353, + "step": 14840 + }, + { + "epoch": 0.047552, + "grad_norm": 0.11535669705078191, + "learning_rate": 9.993802092707731e-05, + "loss": 0.0339, + "step": 14860 + }, + { + "epoch": 0.047616, + "grad_norm": 0.12658588356783088, + "learning_rate": 9.993785401474734e-05, + "loss": 0.0323, + "step": 14880 + }, + { + "epoch": 0.04768, + "grad_norm": 0.29463867532835775, + "learning_rate": 9.993768687810817e-05, + "loss": 0.035, + "step": 14900 + }, + { + "epoch": 0.047744, + "grad_norm": 0.11828806598573192, + "learning_rate": 9.993751951716058e-05, + "loss": 0.0342, + "step": 14920 + }, + { + "epoch": 0.047808, + "grad_norm": 0.1434447125662265, + "learning_rate": 9.99373519319053e-05, + "loss": 0.0348, + "step": 14940 + }, + { + "epoch": 0.047872, + "grad_norm": 0.1140664304308181, + "learning_rate": 9.993718412234308e-05, + "loss": 0.0334, + "step": 14960 + }, + { + "epoch": 0.047936, + "grad_norm": 0.16457879081691087, + "learning_rate": 9.993701608847468e-05, + "loss": 0.0348, + "step": 14980 + }, + { + "epoch": 0.048, + "grad_norm": 0.0872050972778628, + "learning_rate": 9.993684783030088e-05, + "loss": 0.0321, + "step": 15000 + }, + { + "epoch": 0.048064, + "grad_norm": 0.10489291283346126, + "learning_rate": 9.993667934782241e-05, + "loss": 0.0355, + "step": 15020 + }, + { + "epoch": 0.048128, + "grad_norm": 0.1549662440470864, + "learning_rate": 9.993651064104002e-05, + "loss": 0.0342, + "step": 15040 + }, + { + "epoch": 0.048192, + "grad_norm": 0.14218942896035872, + "learning_rate": 9.993634170995447e-05, + "loss": 0.0357, + "step": 15060 + }, + { + "epoch": 0.048256, + "grad_norm": 0.17901894353111553, + "learning_rate": 9.993617255456654e-05, + "loss": 0.0377, + "step": 15080 + }, + { + "epoch": 0.04832, + "grad_norm": 0.11490904691496781, + "learning_rate": 9.993600317487697e-05, + "loss": 0.0396, + "step": 15100 + }, + { + "epoch": 0.048384, + "grad_norm": 0.10158856657014537, + "learning_rate": 9.993583357088654e-05, + "loss": 0.0356, + "step": 15120 + }, + { + "epoch": 0.048448, + "grad_norm": 0.14616911007947328, + "learning_rate": 9.9935663742596e-05, + "loss": 0.0344, + "step": 15140 + }, + { + "epoch": 0.048512, + "grad_norm": 0.08066204896655678, + "learning_rate": 9.99354936900061e-05, + "loss": 0.0326, + "step": 15160 + }, + { + "epoch": 0.048576, + "grad_norm": 0.15488480425431767, + "learning_rate": 9.993532341311761e-05, + "loss": 0.0374, + "step": 15180 + }, + { + "epoch": 0.04864, + "grad_norm": 0.08807594781960085, + "learning_rate": 9.993515291193132e-05, + "loss": 0.0343, + "step": 15200 + }, + { + "epoch": 0.048704, + "grad_norm": 0.10621529203093809, + "learning_rate": 9.993498218644797e-05, + "loss": 0.0347, + "step": 15220 + }, + { + "epoch": 0.048768, + "grad_norm": 0.11430639371672982, + "learning_rate": 9.993481123666833e-05, + "loss": 0.0408, + "step": 15240 + }, + { + "epoch": 0.048832, + "grad_norm": 0.36945995391943465, + "learning_rate": 9.993464006259317e-05, + "loss": 0.0348, + "step": 15260 + }, + { + "epoch": 0.048896, + "grad_norm": 0.13269058485384, + "learning_rate": 9.993446866422327e-05, + "loss": 0.0349, + "step": 15280 + }, + { + "epoch": 0.04896, + "grad_norm": 0.10589843117561257, + "learning_rate": 9.993429704155938e-05, + "loss": 0.038, + "step": 15300 + }, + { + "epoch": 0.049024, + "grad_norm": 0.15700902680365109, + "learning_rate": 9.99341251946023e-05, + "loss": 0.0383, + "step": 15320 + }, + { + "epoch": 0.049088, + "grad_norm": 0.15040423058070657, + "learning_rate": 9.993395312335276e-05, + "loss": 0.0376, + "step": 15340 + }, + { + "epoch": 0.049152, + "grad_norm": 0.21553576747824754, + "learning_rate": 9.993378082781157e-05, + "loss": 0.0372, + "step": 15360 + }, + { + "epoch": 0.049216, + "grad_norm": 0.18650799795645834, + "learning_rate": 9.993360830797949e-05, + "loss": 0.0411, + "step": 15380 + }, + { + "epoch": 0.04928, + "grad_norm": 0.30238755134344447, + "learning_rate": 9.993343556385729e-05, + "loss": 0.0372, + "step": 15400 + }, + { + "epoch": 0.049344, + "grad_norm": 0.1107334102287796, + "learning_rate": 9.993326259544573e-05, + "loss": 0.0364, + "step": 15420 + }, + { + "epoch": 0.049408, + "grad_norm": 0.16987525598901113, + "learning_rate": 9.993308940274564e-05, + "loss": 0.0349, + "step": 15440 + }, + { + "epoch": 0.049472, + "grad_norm": 0.14271833030442146, + "learning_rate": 9.993291598575775e-05, + "loss": 0.0355, + "step": 15460 + }, + { + "epoch": 0.049536, + "grad_norm": 0.12424955808241304, + "learning_rate": 9.993274234448286e-05, + "loss": 0.0336, + "step": 15480 + }, + { + "epoch": 0.0496, + "grad_norm": 0.1252983660891284, + "learning_rate": 9.993256847892174e-05, + "loss": 0.0329, + "step": 15500 + }, + { + "epoch": 0.049664, + "grad_norm": 0.1632271914944957, + "learning_rate": 9.993239438907518e-05, + "loss": 0.0364, + "step": 15520 + }, + { + "epoch": 0.049728, + "grad_norm": 0.11146834649657573, + "learning_rate": 9.993222007494396e-05, + "loss": 0.0348, + "step": 15540 + }, + { + "epoch": 0.049792, + "grad_norm": 0.19790605842087128, + "learning_rate": 9.993204553652886e-05, + "loss": 0.0326, + "step": 15560 + }, + { + "epoch": 0.049856, + "grad_norm": 0.15819604089219855, + "learning_rate": 9.993187077383065e-05, + "loss": 0.0361, + "step": 15580 + }, + { + "epoch": 0.04992, + "grad_norm": 0.16831492818404428, + "learning_rate": 9.993169578685014e-05, + "loss": 0.0364, + "step": 15600 + }, + { + "epoch": 0.049984, + "grad_norm": 0.1366440587770829, + "learning_rate": 9.99315205755881e-05, + "loss": 0.033, + "step": 15620 + }, + { + "epoch": 0.050048, + "grad_norm": 0.15495610051348688, + "learning_rate": 9.993134514004533e-05, + "loss": 0.033, + "step": 15640 + }, + { + "epoch": 0.050112, + "grad_norm": 0.18647477227661124, + "learning_rate": 9.993116948022261e-05, + "loss": 0.0348, + "step": 15660 + }, + { + "epoch": 0.050176, + "grad_norm": 0.15840719768365794, + "learning_rate": 9.993099359612073e-05, + "loss": 0.0378, + "step": 15680 + }, + { + "epoch": 0.05024, + "grad_norm": 0.112873839185945, + "learning_rate": 9.993081748774047e-05, + "loss": 0.0341, + "step": 15700 + }, + { + "epoch": 0.050304, + "grad_norm": 0.12090058286256929, + "learning_rate": 9.993064115508264e-05, + "loss": 0.0343, + "step": 15720 + }, + { + "epoch": 0.050368, + "grad_norm": 0.24511854702843727, + "learning_rate": 9.993046459814803e-05, + "loss": 0.037, + "step": 15740 + }, + { + "epoch": 0.050432, + "grad_norm": 0.12455998067408601, + "learning_rate": 9.993028781693741e-05, + "loss": 0.0339, + "step": 15760 + }, + { + "epoch": 0.050496, + "grad_norm": 0.18205914901228507, + "learning_rate": 9.993011081145158e-05, + "loss": 0.0391, + "step": 15780 + }, + { + "epoch": 0.05056, + "grad_norm": 0.11097677551840077, + "learning_rate": 9.992993358169138e-05, + "loss": 0.0338, + "step": 15800 + }, + { + "epoch": 0.050624, + "grad_norm": 0.1132496056670998, + "learning_rate": 9.992975612765754e-05, + "loss": 0.0338, + "step": 15820 + }, + { + "epoch": 0.050688, + "grad_norm": 0.14910594194944238, + "learning_rate": 9.99295784493509e-05, + "loss": 0.0366, + "step": 15840 + }, + { + "epoch": 0.050752, + "grad_norm": 0.1641586497159284, + "learning_rate": 9.992940054677225e-05, + "loss": 0.0334, + "step": 15860 + }, + { + "epoch": 0.050816, + "grad_norm": 0.14373013659326545, + "learning_rate": 9.992922241992238e-05, + "loss": 0.0341, + "step": 15880 + }, + { + "epoch": 0.05088, + "grad_norm": 0.1171140793958473, + "learning_rate": 9.992904406880209e-05, + "loss": 0.0354, + "step": 15900 + }, + { + "epoch": 0.050944, + "grad_norm": 0.13939465906525142, + "learning_rate": 9.992886549341218e-05, + "loss": 0.0388, + "step": 15920 + }, + { + "epoch": 0.051008, + "grad_norm": 0.1190659403202558, + "learning_rate": 9.992868669375348e-05, + "loss": 0.0405, + "step": 15940 + }, + { + "epoch": 0.051072, + "grad_norm": 0.12612388356621773, + "learning_rate": 9.992850766982676e-05, + "loss": 0.036, + "step": 15960 + }, + { + "epoch": 0.051136, + "grad_norm": 0.1762925196097625, + "learning_rate": 9.992832842163285e-05, + "loss": 0.0356, + "step": 15980 + }, + { + "epoch": 0.0512, + "grad_norm": 0.09386019706490482, + "learning_rate": 9.992814894917252e-05, + "loss": 0.0328, + "step": 16000 + }, + { + "epoch": 0.051264, + "grad_norm": 0.09684560935513954, + "learning_rate": 9.992796925244661e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 0.051328, + "grad_norm": 0.14426183253418937, + "learning_rate": 9.99277893314559e-05, + "loss": 0.0347, + "step": 16040 + }, + { + "epoch": 0.051392, + "grad_norm": 0.09442640243966588, + "learning_rate": 9.992760918620122e-05, + "loss": 0.0391, + "step": 16060 + }, + { + "epoch": 0.051456, + "grad_norm": 0.12465464806603783, + "learning_rate": 9.992742881668338e-05, + "loss": 0.036, + "step": 16080 + }, + { + "epoch": 0.05152, + "grad_norm": 0.11096561198850499, + "learning_rate": 9.992724822290316e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 0.051584, + "grad_norm": 0.09714232482847845, + "learning_rate": 9.992706740486142e-05, + "loss": 0.0338, + "step": 16120 + }, + { + "epoch": 0.051648, + "grad_norm": 0.10576988176927238, + "learning_rate": 9.992688636255893e-05, + "loss": 0.0347, + "step": 16140 + }, + { + "epoch": 0.051712, + "grad_norm": 0.10228366839813258, + "learning_rate": 9.992670509599651e-05, + "loss": 0.0342, + "step": 16160 + }, + { + "epoch": 0.051776, + "grad_norm": 0.10277090936468604, + "learning_rate": 9.992652360517501e-05, + "loss": 0.0296, + "step": 16180 + }, + { + "epoch": 0.05184, + "grad_norm": 0.10747004890666984, + "learning_rate": 9.992634189009519e-05, + "loss": 0.0355, + "step": 16200 + }, + { + "epoch": 0.051904, + "grad_norm": 0.12535374949337835, + "learning_rate": 9.99261599507579e-05, + "loss": 0.0338, + "step": 16220 + }, + { + "epoch": 0.051968, + "grad_norm": 0.12490182924855917, + "learning_rate": 9.992597778716396e-05, + "loss": 0.0371, + "step": 16240 + }, + { + "epoch": 0.052032, + "grad_norm": 0.11732838580516043, + "learning_rate": 9.992579539931416e-05, + "loss": 0.0387, + "step": 16260 + }, + { + "epoch": 0.052096, + "grad_norm": 0.11692296773059428, + "learning_rate": 9.992561278720937e-05, + "loss": 0.0392, + "step": 16280 + }, + { + "epoch": 0.05216, + "grad_norm": 0.17919967223252306, + "learning_rate": 9.992542995085036e-05, + "loss": 0.0361, + "step": 16300 + }, + { + "epoch": 0.052224, + "grad_norm": 0.11856582687977317, + "learning_rate": 9.992524689023797e-05, + "loss": 0.0374, + "step": 16320 + }, + { + "epoch": 0.052288, + "grad_norm": 0.14212369362365593, + "learning_rate": 9.992506360537302e-05, + "loss": 0.033, + "step": 16340 + }, + { + "epoch": 0.052352, + "grad_norm": 0.13905784273136748, + "learning_rate": 9.992488009625634e-05, + "loss": 0.0319, + "step": 16360 + }, + { + "epoch": 0.052416, + "grad_norm": 0.11361108413183314, + "learning_rate": 9.992469636288875e-05, + "loss": 0.036, + "step": 16380 + }, + { + "epoch": 0.05248, + "grad_norm": 0.1759092588731618, + "learning_rate": 9.992451240527107e-05, + "loss": 0.04, + "step": 16400 + }, + { + "epoch": 0.052544, + "grad_norm": 0.10997941270962663, + "learning_rate": 9.992432822340414e-05, + "loss": 0.0337, + "step": 16420 + }, + { + "epoch": 0.052608, + "grad_norm": 0.17191687384169788, + "learning_rate": 9.992414381728878e-05, + "loss": 0.0361, + "step": 16440 + }, + { + "epoch": 0.052672, + "grad_norm": 0.11953244747490259, + "learning_rate": 9.992395918692581e-05, + "loss": 0.0355, + "step": 16460 + }, + { + "epoch": 0.052736, + "grad_norm": 0.17495329768503157, + "learning_rate": 9.992377433231608e-05, + "loss": 0.0367, + "step": 16480 + }, + { + "epoch": 0.0528, + "grad_norm": 0.17048154953811825, + "learning_rate": 9.99235892534604e-05, + "loss": 0.0355, + "step": 16500 + }, + { + "epoch": 0.052864, + "grad_norm": 0.10606593110473865, + "learning_rate": 9.99234039503596e-05, + "loss": 0.0354, + "step": 16520 + }, + { + "epoch": 0.052928, + "grad_norm": 0.20442947092222155, + "learning_rate": 9.992321842301452e-05, + "loss": 0.0351, + "step": 16540 + }, + { + "epoch": 0.052992, + "grad_norm": 0.2528353337432819, + "learning_rate": 9.9923032671426e-05, + "loss": 0.0338, + "step": 16560 + }, + { + "epoch": 0.053056, + "grad_norm": 0.10064657000485325, + "learning_rate": 9.992284669559486e-05, + "loss": 0.0378, + "step": 16580 + }, + { + "epoch": 0.05312, + "grad_norm": 0.1044129813357252, + "learning_rate": 9.992266049552195e-05, + "loss": 0.0331, + "step": 16600 + }, + { + "epoch": 0.053184, + "grad_norm": 0.18230274076501582, + "learning_rate": 9.99224740712081e-05, + "loss": 0.0369, + "step": 16620 + }, + { + "epoch": 0.053248, + "grad_norm": 0.1938957673283127, + "learning_rate": 9.992228742265415e-05, + "loss": 0.0378, + "step": 16640 + }, + { + "epoch": 0.053312, + "grad_norm": 0.12146508602103125, + "learning_rate": 9.992210054986093e-05, + "loss": 0.0379, + "step": 16660 + }, + { + "epoch": 0.053376, + "grad_norm": 0.07889922663367899, + "learning_rate": 9.992191345282928e-05, + "loss": 0.0336, + "step": 16680 + }, + { + "epoch": 0.05344, + "grad_norm": 0.1728184818958167, + "learning_rate": 9.992172613156005e-05, + "loss": 0.0354, + "step": 16700 + }, + { + "epoch": 0.053504, + "grad_norm": 0.10213543307742655, + "learning_rate": 9.992153858605409e-05, + "loss": 0.0331, + "step": 16720 + }, + { + "epoch": 0.053568, + "grad_norm": 0.09680245345420041, + "learning_rate": 9.992135081631222e-05, + "loss": 0.0309, + "step": 16740 + }, + { + "epoch": 0.053632, + "grad_norm": 0.16081126956732297, + "learning_rate": 9.99211628223353e-05, + "loss": 0.0332, + "step": 16760 + }, + { + "epoch": 0.053696, + "grad_norm": 0.11728220164945362, + "learning_rate": 9.992097460412415e-05, + "loss": 0.0377, + "step": 16780 + }, + { + "epoch": 0.05376, + "grad_norm": 0.08584639631267706, + "learning_rate": 9.992078616167964e-05, + "loss": 0.0365, + "step": 16800 + }, + { + "epoch": 0.053824, + "grad_norm": 0.20963288379467385, + "learning_rate": 9.99205974950026e-05, + "loss": 0.0354, + "step": 16820 + }, + { + "epoch": 0.053888, + "grad_norm": 0.15742272252431652, + "learning_rate": 9.992040860409391e-05, + "loss": 0.035, + "step": 16840 + }, + { + "epoch": 0.053952, + "grad_norm": 0.1278751078021988, + "learning_rate": 9.992021948895437e-05, + "loss": 0.0357, + "step": 16860 + }, + { + "epoch": 0.054016, + "grad_norm": 0.13241155385690026, + "learning_rate": 9.992003014958486e-05, + "loss": 0.0329, + "step": 16880 + }, + { + "epoch": 0.05408, + "grad_norm": 0.10715627650243452, + "learning_rate": 9.991984058598624e-05, + "loss": 0.0333, + "step": 16900 + }, + { + "epoch": 0.054144, + "grad_norm": 0.10234875719834678, + "learning_rate": 9.991965079815933e-05, + "loss": 0.034, + "step": 16920 + }, + { + "epoch": 0.054208, + "grad_norm": 0.1025761271480278, + "learning_rate": 9.991946078610501e-05, + "loss": 0.0317, + "step": 16940 + }, + { + "epoch": 0.054272, + "grad_norm": 0.13086966810693465, + "learning_rate": 9.991927054982411e-05, + "loss": 0.0375, + "step": 16960 + }, + { + "epoch": 0.054336, + "grad_norm": 0.10365092790857679, + "learning_rate": 9.99190800893175e-05, + "loss": 0.0329, + "step": 16980 + }, + { + "epoch": 0.0544, + "grad_norm": 0.09961402826283518, + "learning_rate": 9.991888940458605e-05, + "loss": 0.0335, + "step": 17000 + }, + { + "epoch": 0.054464, + "grad_norm": 0.10957491947883129, + "learning_rate": 9.991869849563059e-05, + "loss": 0.0332, + "step": 17020 + }, + { + "epoch": 0.054528, + "grad_norm": 0.0927093316773643, + "learning_rate": 9.991850736245197e-05, + "loss": 0.0365, + "step": 17040 + }, + { + "epoch": 0.054592, + "grad_norm": 0.1017137067642453, + "learning_rate": 9.991831600505107e-05, + "loss": 0.0371, + "step": 17060 + }, + { + "epoch": 0.054656, + "grad_norm": 0.13317395720989603, + "learning_rate": 9.991812442342876e-05, + "loss": 0.0343, + "step": 17080 + }, + { + "epoch": 0.05472, + "grad_norm": 0.16406970670368645, + "learning_rate": 9.991793261758586e-05, + "loss": 0.0367, + "step": 17100 + }, + { + "epoch": 0.054784, + "grad_norm": 0.23944718482766253, + "learning_rate": 9.991774058752327e-05, + "loss": 0.0333, + "step": 17120 + }, + { + "epoch": 0.054848, + "grad_norm": 0.12307568774973436, + "learning_rate": 9.991754833324185e-05, + "loss": 0.0375, + "step": 17140 + }, + { + "epoch": 0.054912, + "grad_norm": 0.12020699304756892, + "learning_rate": 9.991735585474244e-05, + "loss": 0.0366, + "step": 17160 + }, + { + "epoch": 0.054976, + "grad_norm": 0.09029017893806349, + "learning_rate": 9.991716315202592e-05, + "loss": 0.0345, + "step": 17180 + }, + { + "epoch": 0.05504, + "grad_norm": 0.10679322352203884, + "learning_rate": 9.991697022509315e-05, + "loss": 0.0342, + "step": 17200 + }, + { + "epoch": 0.055104, + "grad_norm": 0.10696593739887994, + "learning_rate": 9.9916777073945e-05, + "loss": 0.0336, + "step": 17220 + }, + { + "epoch": 0.055168, + "grad_norm": 0.16167680756239386, + "learning_rate": 9.991658369858233e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 0.055232, + "grad_norm": 0.1352538602908927, + "learning_rate": 9.991639009900603e-05, + "loss": 0.0331, + "step": 17260 + }, + { + "epoch": 0.055296, + "grad_norm": 0.09905252698789209, + "learning_rate": 9.991619627521694e-05, + "loss": 0.0321, + "step": 17280 + }, + { + "epoch": 0.05536, + "grad_norm": 0.15155107834389772, + "learning_rate": 9.991600222721595e-05, + "loss": 0.0342, + "step": 17300 + }, + { + "epoch": 0.055424, + "grad_norm": 0.19671934535639352, + "learning_rate": 9.991580795500395e-05, + "loss": 0.0379, + "step": 17320 + }, + { + "epoch": 0.055488, + "grad_norm": 0.10023397181395324, + "learning_rate": 9.991561345858175e-05, + "loss": 0.0336, + "step": 17340 + }, + { + "epoch": 0.055552, + "grad_norm": 0.09636066886480535, + "learning_rate": 9.99154187379503e-05, + "loss": 0.0332, + "step": 17360 + }, + { + "epoch": 0.055616, + "grad_norm": 0.10439982204314885, + "learning_rate": 9.991522379311042e-05, + "loss": 0.0347, + "step": 17380 + }, + { + "epoch": 0.05568, + "grad_norm": 0.11928952316721914, + "learning_rate": 9.9915028624063e-05, + "loss": 0.0316, + "step": 17400 + }, + { + "epoch": 0.055744, + "grad_norm": 0.20672585824959547, + "learning_rate": 9.991483323080893e-05, + "loss": 0.0341, + "step": 17420 + }, + { + "epoch": 0.055808, + "grad_norm": 0.12002024866559705, + "learning_rate": 9.991463761334907e-05, + "loss": 0.0366, + "step": 17440 + }, + { + "epoch": 0.055872, + "grad_norm": 0.1498603835658597, + "learning_rate": 9.991444177168432e-05, + "loss": 0.0348, + "step": 17460 + }, + { + "epoch": 0.055936, + "grad_norm": 0.11949422669049035, + "learning_rate": 9.991424570581554e-05, + "loss": 0.0333, + "step": 17480 + }, + { + "epoch": 0.056, + "grad_norm": 0.15552596340785763, + "learning_rate": 9.991404941574361e-05, + "loss": 0.0365, + "step": 17500 + }, + { + "epoch": 0.056064, + "grad_norm": 0.24390940092927904, + "learning_rate": 9.991385290146942e-05, + "loss": 0.0319, + "step": 17520 + }, + { + "epoch": 0.056128, + "grad_norm": 0.1769937720225227, + "learning_rate": 9.991365616299386e-05, + "loss": 0.0383, + "step": 17540 + }, + { + "epoch": 0.056192, + "grad_norm": 0.12683272355407324, + "learning_rate": 9.991345920031779e-05, + "loss": 0.0329, + "step": 17560 + }, + { + "epoch": 0.056256, + "grad_norm": 0.18849631934236188, + "learning_rate": 9.991326201344213e-05, + "loss": 0.0325, + "step": 17580 + }, + { + "epoch": 0.05632, + "grad_norm": 0.2108928972656276, + "learning_rate": 9.991306460236773e-05, + "loss": 0.0331, + "step": 17600 + }, + { + "epoch": 0.056384, + "grad_norm": 0.1059909125245483, + "learning_rate": 9.99128669670955e-05, + "loss": 0.0365, + "step": 17620 + }, + { + "epoch": 0.056448, + "grad_norm": 0.11824659774716664, + "learning_rate": 9.991266910762631e-05, + "loss": 0.0311, + "step": 17640 + }, + { + "epoch": 0.056512, + "grad_norm": 0.11115965499941091, + "learning_rate": 9.991247102396107e-05, + "loss": 0.032, + "step": 17660 + }, + { + "epoch": 0.056576, + "grad_norm": 0.1358715975746066, + "learning_rate": 9.991227271610067e-05, + "loss": 0.0321, + "step": 17680 + }, + { + "epoch": 0.05664, + "grad_norm": 0.16275223242757147, + "learning_rate": 9.991207418404595e-05, + "loss": 0.0353, + "step": 17700 + }, + { + "epoch": 0.056704, + "grad_norm": 0.09493439795229432, + "learning_rate": 9.991187542779788e-05, + "loss": 0.036, + "step": 17720 + }, + { + "epoch": 0.056768, + "grad_norm": 0.3175580496763981, + "learning_rate": 9.99116764473573e-05, + "loss": 0.0354, + "step": 17740 + }, + { + "epoch": 0.056832, + "grad_norm": 0.09553338760934682, + "learning_rate": 9.991147724272512e-05, + "loss": 0.0352, + "step": 17760 + }, + { + "epoch": 0.056896, + "grad_norm": 0.14651912804465514, + "learning_rate": 9.991127781390223e-05, + "loss": 0.0339, + "step": 17780 + }, + { + "epoch": 0.05696, + "grad_norm": 0.147031438858404, + "learning_rate": 9.991107816088952e-05, + "loss": 0.0329, + "step": 17800 + }, + { + "epoch": 0.057024, + "grad_norm": 0.09185898143616375, + "learning_rate": 9.991087828368791e-05, + "loss": 0.0334, + "step": 17820 + }, + { + "epoch": 0.057088, + "grad_norm": 0.13551425112936327, + "learning_rate": 9.991067818229827e-05, + "loss": 0.0383, + "step": 17840 + }, + { + "epoch": 0.057152, + "grad_norm": 0.2794838092789423, + "learning_rate": 9.991047785672152e-05, + "loss": 0.0371, + "step": 17860 + }, + { + "epoch": 0.057216, + "grad_norm": 0.11449254977615445, + "learning_rate": 9.991027730695855e-05, + "loss": 0.0349, + "step": 17880 + }, + { + "epoch": 0.05728, + "grad_norm": 0.0811229114081402, + "learning_rate": 9.991007653301026e-05, + "loss": 0.0314, + "step": 17900 + }, + { + "epoch": 0.057344, + "grad_norm": 0.10862861432977336, + "learning_rate": 9.990987553487754e-05, + "loss": 0.0331, + "step": 17920 + }, + { + "epoch": 0.057408, + "grad_norm": 0.15745378169901814, + "learning_rate": 9.990967431256134e-05, + "loss": 0.0349, + "step": 17940 + }, + { + "epoch": 0.057472, + "grad_norm": 0.10544788632370294, + "learning_rate": 9.990947286606252e-05, + "loss": 0.0337, + "step": 17960 + }, + { + "epoch": 0.057536, + "grad_norm": 0.13942015689736034, + "learning_rate": 9.990927119538198e-05, + "loss": 0.0332, + "step": 17980 + }, + { + "epoch": 0.0576, + "grad_norm": 0.17505802424697345, + "learning_rate": 9.990906930052064e-05, + "loss": 0.0375, + "step": 18000 + }, + { + "epoch": 0.057664, + "grad_norm": 0.1515367875798829, + "learning_rate": 9.990886718147942e-05, + "loss": 0.0343, + "step": 18020 + }, + { + "epoch": 0.057728, + "grad_norm": 0.12768374745547934, + "learning_rate": 9.990866483825921e-05, + "loss": 0.0324, + "step": 18040 + }, + { + "epoch": 0.057792, + "grad_norm": 0.09434701899970553, + "learning_rate": 9.990846227086093e-05, + "loss": 0.0328, + "step": 18060 + }, + { + "epoch": 0.057856, + "grad_norm": 0.13908542152899867, + "learning_rate": 9.990825947928548e-05, + "loss": 0.0364, + "step": 18080 + }, + { + "epoch": 0.05792, + "grad_norm": 0.11917830543521794, + "learning_rate": 9.990805646353378e-05, + "loss": 0.0327, + "step": 18100 + }, + { + "epoch": 0.057984, + "grad_norm": 0.15256780533481193, + "learning_rate": 9.990785322360673e-05, + "loss": 0.0355, + "step": 18120 + }, + { + "epoch": 0.058048, + "grad_norm": 0.12672488473883228, + "learning_rate": 9.990764975950528e-05, + "loss": 0.0338, + "step": 18140 + }, + { + "epoch": 0.058112, + "grad_norm": 0.12297585140230674, + "learning_rate": 9.990744607123028e-05, + "loss": 0.0329, + "step": 18160 + }, + { + "epoch": 0.058176, + "grad_norm": 0.14255308838932138, + "learning_rate": 9.99072421587827e-05, + "loss": 0.0355, + "step": 18180 + }, + { + "epoch": 0.05824, + "grad_norm": 0.11200753000960445, + "learning_rate": 9.990703802216342e-05, + "loss": 0.0386, + "step": 18200 + }, + { + "epoch": 0.058304, + "grad_norm": 0.11355356355874739, + "learning_rate": 9.990683366137339e-05, + "loss": 0.0323, + "step": 18220 + }, + { + "epoch": 0.058368, + "grad_norm": 0.12034569126015282, + "learning_rate": 9.99066290764135e-05, + "loss": 0.0342, + "step": 18240 + }, + { + "epoch": 0.058432, + "grad_norm": 0.1285653876365367, + "learning_rate": 9.990642426728468e-05, + "loss": 0.0359, + "step": 18260 + }, + { + "epoch": 0.058496, + "grad_norm": 0.11112692004872396, + "learning_rate": 9.990621923398785e-05, + "loss": 0.0339, + "step": 18280 + }, + { + "epoch": 0.05856, + "grad_norm": 0.1563565342642215, + "learning_rate": 9.990601397652393e-05, + "loss": 0.0354, + "step": 18300 + }, + { + "epoch": 0.058624, + "grad_norm": 0.14043547131885564, + "learning_rate": 9.990580849489384e-05, + "loss": 0.0313, + "step": 18320 + }, + { + "epoch": 0.058688, + "grad_norm": 0.10193902540028733, + "learning_rate": 9.99056027890985e-05, + "loss": 0.0335, + "step": 18340 + }, + { + "epoch": 0.058752, + "grad_norm": 0.12202435610224495, + "learning_rate": 9.990539685913884e-05, + "loss": 0.0311, + "step": 18360 + }, + { + "epoch": 0.058816, + "grad_norm": 0.10023472155745906, + "learning_rate": 9.99051907050158e-05, + "loss": 0.0324, + "step": 18380 + }, + { + "epoch": 0.05888, + "grad_norm": 0.11928000676105618, + "learning_rate": 9.990498432673028e-05, + "loss": 0.0335, + "step": 18400 + }, + { + "epoch": 0.058944, + "grad_norm": 0.12134339926636085, + "learning_rate": 9.990477772428322e-05, + "loss": 0.0349, + "step": 18420 + }, + { + "epoch": 0.059008, + "grad_norm": 0.09979515553952921, + "learning_rate": 9.990457089767553e-05, + "loss": 0.0325, + "step": 18440 + }, + { + "epoch": 0.059072, + "grad_norm": 0.11769755778188208, + "learning_rate": 9.990436384690817e-05, + "loss": 0.0364, + "step": 18460 + }, + { + "epoch": 0.059136, + "grad_norm": 0.13767932583506837, + "learning_rate": 9.990415657198205e-05, + "loss": 0.036, + "step": 18480 + }, + { + "epoch": 0.0592, + "grad_norm": 0.11612865143666769, + "learning_rate": 9.99039490728981e-05, + "loss": 0.0339, + "step": 18500 + }, + { + "epoch": 0.059264, + "grad_norm": 0.08181794200159138, + "learning_rate": 9.990374134965725e-05, + "loss": 0.0312, + "step": 18520 + }, + { + "epoch": 0.059328, + "grad_norm": 0.12236719829746981, + "learning_rate": 9.990353340226047e-05, + "loss": 0.0331, + "step": 18540 + }, + { + "epoch": 0.059392, + "grad_norm": 0.10898679635604598, + "learning_rate": 9.990332523070864e-05, + "loss": 0.0406, + "step": 18560 + }, + { + "epoch": 0.059456, + "grad_norm": 0.12233154935507864, + "learning_rate": 9.990311683500273e-05, + "loss": 0.0366, + "step": 18580 + }, + { + "epoch": 0.05952, + "grad_norm": 0.2600631094332187, + "learning_rate": 9.990290821514366e-05, + "loss": 0.0387, + "step": 18600 + }, + { + "epoch": 0.059584, + "grad_norm": 0.15469006122046597, + "learning_rate": 9.990269937113237e-05, + "loss": 0.0361, + "step": 18620 + }, + { + "epoch": 0.059648, + "grad_norm": 0.08252625811538122, + "learning_rate": 9.990249030296981e-05, + "loss": 0.031, + "step": 18640 + }, + { + "epoch": 0.059712, + "grad_norm": 0.1414663873773736, + "learning_rate": 9.990228101065689e-05, + "loss": 0.0339, + "step": 18660 + }, + { + "epoch": 0.059776, + "grad_norm": 0.1465211520646684, + "learning_rate": 9.990207149419459e-05, + "loss": 0.0346, + "step": 18680 + }, + { + "epoch": 0.05984, + "grad_norm": 0.12121307661126639, + "learning_rate": 9.990186175358382e-05, + "loss": 0.0348, + "step": 18700 + }, + { + "epoch": 0.059904, + "grad_norm": 0.08996386444771336, + "learning_rate": 9.990165178882554e-05, + "loss": 0.0341, + "step": 18720 + }, + { + "epoch": 0.059968, + "grad_norm": 0.15569067738532616, + "learning_rate": 9.990144159992067e-05, + "loss": 0.031, + "step": 18740 + }, + { + "epoch": 0.060032, + "grad_norm": 0.12224068009247853, + "learning_rate": 9.990123118687019e-05, + "loss": 0.0342, + "step": 18760 + }, + { + "epoch": 0.060096, + "grad_norm": 0.09792409711888014, + "learning_rate": 9.990102054967503e-05, + "loss": 0.0355, + "step": 18780 + }, + { + "epoch": 0.06016, + "grad_norm": 0.11012316352934545, + "learning_rate": 9.99008096883361e-05, + "loss": 0.035, + "step": 18800 + }, + { + "epoch": 0.060224, + "grad_norm": 0.20201513967656096, + "learning_rate": 9.990059860285441e-05, + "loss": 0.0344, + "step": 18820 + }, + { + "epoch": 0.060288, + "grad_norm": 0.1285491299987621, + "learning_rate": 9.990038729323087e-05, + "loss": 0.033, + "step": 18840 + }, + { + "epoch": 0.060352, + "grad_norm": 0.1357075823107087, + "learning_rate": 9.990017575946643e-05, + "loss": 0.0396, + "step": 18860 + }, + { + "epoch": 0.060416, + "grad_norm": 0.10438735211225703, + "learning_rate": 9.989996400156205e-05, + "loss": 0.0364, + "step": 18880 + }, + { + "epoch": 0.06048, + "grad_norm": 0.10779371945686081, + "learning_rate": 9.989975201951867e-05, + "loss": 0.0358, + "step": 18900 + }, + { + "epoch": 0.060544, + "grad_norm": 0.10849754518475155, + "learning_rate": 9.989953981333725e-05, + "loss": 0.0323, + "step": 18920 + }, + { + "epoch": 0.060608, + "grad_norm": 0.13785281612711664, + "learning_rate": 9.989932738301875e-05, + "loss": 0.0357, + "step": 18940 + }, + { + "epoch": 0.060672, + "grad_norm": 0.12359944220096011, + "learning_rate": 9.98991147285641e-05, + "loss": 0.034, + "step": 18960 + }, + { + "epoch": 0.060736, + "grad_norm": 0.2119481391949673, + "learning_rate": 9.98989018499743e-05, + "loss": 0.0348, + "step": 18980 + }, + { + "epoch": 0.0608, + "grad_norm": 0.1515189682298724, + "learning_rate": 9.989868874725026e-05, + "loss": 0.0379, + "step": 19000 + }, + { + "epoch": 0.060864, + "grad_norm": 0.18079383528520512, + "learning_rate": 9.989847542039295e-05, + "loss": 0.0375, + "step": 19020 + }, + { + "epoch": 0.060928, + "grad_norm": 0.14575406379674052, + "learning_rate": 9.989826186940334e-05, + "loss": 0.0344, + "step": 19040 + }, + { + "epoch": 0.060992, + "grad_norm": 0.11003266959984635, + "learning_rate": 9.989804809428239e-05, + "loss": 0.0364, + "step": 19060 + }, + { + "epoch": 0.061056, + "grad_norm": 0.11052559837587579, + "learning_rate": 9.989783409503105e-05, + "loss": 0.0335, + "step": 19080 + }, + { + "epoch": 0.06112, + "grad_norm": 0.09700513267560862, + "learning_rate": 9.989761987165029e-05, + "loss": 0.0301, + "step": 19100 + }, + { + "epoch": 0.061184, + "grad_norm": 0.14182378064807033, + "learning_rate": 9.989740542414106e-05, + "loss": 0.0341, + "step": 19120 + }, + { + "epoch": 0.061248, + "grad_norm": 0.15805631954007268, + "learning_rate": 9.989719075250432e-05, + "loss": 0.0335, + "step": 19140 + }, + { + "epoch": 0.061312, + "grad_norm": 0.10875748968299204, + "learning_rate": 9.989697585674104e-05, + "loss": 0.0346, + "step": 19160 + }, + { + "epoch": 0.061376, + "grad_norm": 0.1274906198316659, + "learning_rate": 9.989676073685221e-05, + "loss": 0.0339, + "step": 19180 + }, + { + "epoch": 0.06144, + "grad_norm": 0.12903792917248563, + "learning_rate": 9.989654539283877e-05, + "loss": 0.0333, + "step": 19200 + }, + { + "epoch": 0.061504, + "grad_norm": 0.09074888178912414, + "learning_rate": 9.989632982470168e-05, + "loss": 0.033, + "step": 19220 + }, + { + "epoch": 0.061568, + "grad_norm": 0.09875004952977967, + "learning_rate": 9.989611403244193e-05, + "loss": 0.0305, + "step": 19240 + }, + { + "epoch": 0.061632, + "grad_norm": 0.20204301572168246, + "learning_rate": 9.989589801606048e-05, + "loss": 0.0329, + "step": 19260 + }, + { + "epoch": 0.061696, + "grad_norm": 0.11847988152062885, + "learning_rate": 9.989568177555831e-05, + "loss": 0.0348, + "step": 19280 + }, + { + "epoch": 0.06176, + "grad_norm": 0.16655053475710985, + "learning_rate": 9.989546531093637e-05, + "loss": 0.0383, + "step": 19300 + }, + { + "epoch": 0.061824, + "grad_norm": 0.10908687928662378, + "learning_rate": 9.989524862219564e-05, + "loss": 0.0358, + "step": 19320 + }, + { + "epoch": 0.061888, + "grad_norm": 0.09543858236886209, + "learning_rate": 9.989503170933712e-05, + "loss": 0.0337, + "step": 19340 + }, + { + "epoch": 0.061952, + "grad_norm": 0.11407101112427125, + "learning_rate": 9.989481457236174e-05, + "loss": 0.0332, + "step": 19360 + }, + { + "epoch": 0.062016, + "grad_norm": 0.09410146795718764, + "learning_rate": 9.98945972112705e-05, + "loss": 0.0362, + "step": 19380 + }, + { + "epoch": 0.06208, + "grad_norm": 0.07725218433624248, + "learning_rate": 9.989437962606438e-05, + "loss": 0.0302, + "step": 19400 + }, + { + "epoch": 0.062144, + "grad_norm": 0.18434246358118317, + "learning_rate": 9.989416181674436e-05, + "loss": 0.0354, + "step": 19420 + }, + { + "epoch": 0.062208, + "grad_norm": 0.17122153378559052, + "learning_rate": 9.989394378331138e-05, + "loss": 0.0385, + "step": 19440 + }, + { + "epoch": 0.062272, + "grad_norm": 0.10789202301142051, + "learning_rate": 9.989372552576647e-05, + "loss": 0.0361, + "step": 19460 + }, + { + "epoch": 0.062336, + "grad_norm": 0.11885968726974544, + "learning_rate": 9.98935070441106e-05, + "loss": 0.033, + "step": 19480 + }, + { + "epoch": 0.0624, + "grad_norm": 0.10569790900373464, + "learning_rate": 9.989328833834471e-05, + "loss": 0.0317, + "step": 19500 + }, + { + "epoch": 0.062464, + "grad_norm": 0.10959879014067173, + "learning_rate": 9.989306940846983e-05, + "loss": 0.0344, + "step": 19520 + }, + { + "epoch": 0.062528, + "grad_norm": 0.10323561374459209, + "learning_rate": 9.989285025448693e-05, + "loss": 0.0361, + "step": 19540 + }, + { + "epoch": 0.062592, + "grad_norm": 0.2907752444232, + "learning_rate": 9.989263087639699e-05, + "loss": 0.0357, + "step": 19560 + }, + { + "epoch": 0.062656, + "grad_norm": 0.1542904560576345, + "learning_rate": 9.989241127420099e-05, + "loss": 0.0376, + "step": 19580 + }, + { + "epoch": 0.06272, + "grad_norm": 0.1271029489308482, + "learning_rate": 9.989219144789992e-05, + "loss": 0.0311, + "step": 19600 + }, + { + "epoch": 0.062784, + "grad_norm": 0.08087576796257898, + "learning_rate": 9.989197139749477e-05, + "loss": 0.0337, + "step": 19620 + }, + { + "epoch": 0.062848, + "grad_norm": 0.10085037238919506, + "learning_rate": 9.989175112298653e-05, + "loss": 0.0346, + "step": 19640 + }, + { + "epoch": 0.062912, + "grad_norm": 0.09812021730250585, + "learning_rate": 9.98915306243762e-05, + "loss": 0.0321, + "step": 19660 + }, + { + "epoch": 0.062976, + "grad_norm": 0.09159899537227678, + "learning_rate": 9.989130990166474e-05, + "loss": 0.0332, + "step": 19680 + }, + { + "epoch": 0.06304, + "grad_norm": 0.10701633815863598, + "learning_rate": 9.989108895485316e-05, + "loss": 0.034, + "step": 19700 + }, + { + "epoch": 0.063104, + "grad_norm": 0.09230289785687283, + "learning_rate": 9.989086778394245e-05, + "loss": 0.0309, + "step": 19720 + }, + { + "epoch": 0.063168, + "grad_norm": 0.2073410268663692, + "learning_rate": 9.989064638893362e-05, + "loss": 0.0344, + "step": 19740 + }, + { + "epoch": 0.063232, + "grad_norm": 0.10908774468773007, + "learning_rate": 9.989042476982764e-05, + "loss": 0.0362, + "step": 19760 + }, + { + "epoch": 0.063296, + "grad_norm": 0.10169583237968459, + "learning_rate": 9.989020292662552e-05, + "loss": 0.0369, + "step": 19780 + }, + { + "epoch": 0.06336, + "grad_norm": 0.10489209772688647, + "learning_rate": 9.988998085932825e-05, + "loss": 0.0334, + "step": 19800 + }, + { + "epoch": 0.063424, + "grad_norm": 0.10941213416882212, + "learning_rate": 9.988975856793684e-05, + "loss": 0.0333, + "step": 19820 + }, + { + "epoch": 0.063488, + "grad_norm": 0.1259127245349666, + "learning_rate": 9.988953605245226e-05, + "loss": 0.0362, + "step": 19840 + }, + { + "epoch": 0.063552, + "grad_norm": 0.1029882494876567, + "learning_rate": 9.988931331287553e-05, + "loss": 0.0351, + "step": 19860 + }, + { + "epoch": 0.063616, + "grad_norm": 0.1196303714125552, + "learning_rate": 9.988909034920765e-05, + "loss": 0.037, + "step": 19880 + }, + { + "epoch": 0.06368, + "grad_norm": 0.09711534888553547, + "learning_rate": 9.988886716144962e-05, + "loss": 0.0348, + "step": 19900 + }, + { + "epoch": 0.063744, + "grad_norm": 0.08308703847779922, + "learning_rate": 9.988864374960246e-05, + "loss": 0.0312, + "step": 19920 + }, + { + "epoch": 0.063808, + "grad_norm": 0.09251636468438884, + "learning_rate": 9.988842011366712e-05, + "loss": 0.0318, + "step": 19940 + }, + { + "epoch": 0.063872, + "grad_norm": 0.111629445438068, + "learning_rate": 9.988819625364465e-05, + "loss": 0.032, + "step": 19960 + }, + { + "epoch": 0.063936, + "grad_norm": 0.09312178642101106, + "learning_rate": 9.988797216953606e-05, + "loss": 0.0338, + "step": 19980 + }, + { + "epoch": 0.064, + "grad_norm": 0.13281444009553867, + "learning_rate": 9.988774786134234e-05, + "loss": 0.0342, + "step": 20000 + }, + { + "epoch": 0.064064, + "grad_norm": 0.1946919488956573, + "learning_rate": 9.988752332906448e-05, + "loss": 0.035, + "step": 20020 + }, + { + "epoch": 0.064128, + "grad_norm": 0.10905724480159042, + "learning_rate": 9.988729857270352e-05, + "loss": 0.0373, + "step": 20040 + }, + { + "epoch": 0.064192, + "grad_norm": 0.10228399147734746, + "learning_rate": 9.988707359226046e-05, + "loss": 0.0352, + "step": 20060 + }, + { + "epoch": 0.064256, + "grad_norm": 0.1613613337891287, + "learning_rate": 9.988684838773631e-05, + "loss": 0.036, + "step": 20080 + }, + { + "epoch": 0.06432, + "grad_norm": 0.1442467397823259, + "learning_rate": 9.988662295913207e-05, + "loss": 0.0348, + "step": 20100 + }, + { + "epoch": 0.064384, + "grad_norm": 0.11227551650506154, + "learning_rate": 9.988639730644878e-05, + "loss": 0.038, + "step": 20120 + }, + { + "epoch": 0.064448, + "grad_norm": 0.09236238289097193, + "learning_rate": 9.988617142968742e-05, + "loss": 0.0374, + "step": 20140 + }, + { + "epoch": 0.064512, + "grad_norm": 0.11519695919059426, + "learning_rate": 9.988594532884901e-05, + "loss": 0.0357, + "step": 20160 + }, + { + "epoch": 0.064576, + "grad_norm": 0.14688764367745852, + "learning_rate": 9.98857190039346e-05, + "loss": 0.0339, + "step": 20180 + }, + { + "epoch": 0.06464, + "grad_norm": 0.13612490655294107, + "learning_rate": 9.988549245494516e-05, + "loss": 0.0305, + "step": 20200 + }, + { + "epoch": 0.064704, + "grad_norm": 0.0881082733425425, + "learning_rate": 9.988526568188175e-05, + "loss": 0.0342, + "step": 20220 + }, + { + "epoch": 0.064768, + "grad_norm": 0.09019350939902039, + "learning_rate": 9.988503868474535e-05, + "loss": 0.0347, + "step": 20240 + }, + { + "epoch": 0.064832, + "grad_norm": 0.0672491035083009, + "learning_rate": 9.988481146353701e-05, + "loss": 0.0311, + "step": 20260 + }, + { + "epoch": 0.064896, + "grad_norm": 0.08533985351303365, + "learning_rate": 9.988458401825773e-05, + "loss": 0.0294, + "step": 20280 + }, + { + "epoch": 0.06496, + "grad_norm": 0.1412247276950405, + "learning_rate": 9.988435634890856e-05, + "loss": 0.0322, + "step": 20300 + }, + { + "epoch": 0.065024, + "grad_norm": 0.11688902620279017, + "learning_rate": 9.988412845549049e-05, + "loss": 0.0338, + "step": 20320 + }, + { + "epoch": 0.065088, + "grad_norm": 0.12587925805163994, + "learning_rate": 9.988390033800455e-05, + "loss": 0.034, + "step": 20340 + }, + { + "epoch": 0.065152, + "grad_norm": 0.09200386765094475, + "learning_rate": 9.988367199645179e-05, + "loss": 0.0376, + "step": 20360 + }, + { + "epoch": 0.065216, + "grad_norm": 0.08709078078675637, + "learning_rate": 9.98834434308332e-05, + "loss": 0.0339, + "step": 20380 + }, + { + "epoch": 0.06528, + "grad_norm": 0.10377614256196271, + "learning_rate": 9.988321464114983e-05, + "loss": 0.0337, + "step": 20400 + }, + { + "epoch": 0.065344, + "grad_norm": 0.1102509101769412, + "learning_rate": 9.988298562740271e-05, + "loss": 0.0347, + "step": 20420 + }, + { + "epoch": 0.065408, + "grad_norm": 0.12764401531714004, + "learning_rate": 9.988275638959284e-05, + "loss": 0.0344, + "step": 20440 + }, + { + "epoch": 0.065472, + "grad_norm": 0.12437730988637337, + "learning_rate": 9.988252692772128e-05, + "loss": 0.0368, + "step": 20460 + }, + { + "epoch": 0.065536, + "grad_norm": 0.12975743178571492, + "learning_rate": 9.988229724178905e-05, + "loss": 0.0351, + "step": 20480 + }, + { + "epoch": 0.0656, + "grad_norm": 0.09155438986807551, + "learning_rate": 9.988206733179718e-05, + "loss": 0.034, + "step": 20500 + }, + { + "epoch": 0.065664, + "grad_norm": 0.1349778081408989, + "learning_rate": 9.98818371977467e-05, + "loss": 0.0346, + "step": 20520 + }, + { + "epoch": 0.065728, + "grad_norm": 0.10545188394964498, + "learning_rate": 9.988160683963867e-05, + "loss": 0.0365, + "step": 20540 + }, + { + "epoch": 0.065792, + "grad_norm": 0.10622394849619642, + "learning_rate": 9.988137625747409e-05, + "loss": 0.0364, + "step": 20560 + }, + { + "epoch": 0.065856, + "grad_norm": 0.15379194517094838, + "learning_rate": 9.9881145451254e-05, + "loss": 0.0346, + "step": 20580 + }, + { + "epoch": 0.06592, + "grad_norm": 0.11170316528472428, + "learning_rate": 9.988091442097946e-05, + "loss": 0.0366, + "step": 20600 + }, + { + "epoch": 0.065984, + "grad_norm": 0.09986089160740233, + "learning_rate": 9.988068316665149e-05, + "loss": 0.0315, + "step": 20620 + }, + { + "epoch": 0.066048, + "grad_norm": 0.0960743040914773, + "learning_rate": 9.988045168827112e-05, + "loss": 0.0341, + "step": 20640 + }, + { + "epoch": 0.066112, + "grad_norm": 0.08902895311018785, + "learning_rate": 9.988021998583941e-05, + "loss": 0.035, + "step": 20660 + }, + { + "epoch": 0.066176, + "grad_norm": 0.12481248581810564, + "learning_rate": 9.98799880593574e-05, + "loss": 0.0312, + "step": 20680 + }, + { + "epoch": 0.06624, + "grad_norm": 0.1311086498766472, + "learning_rate": 9.987975590882612e-05, + "loss": 0.0312, + "step": 20700 + }, + { + "epoch": 0.066304, + "grad_norm": 0.11081441380894388, + "learning_rate": 9.98795235342466e-05, + "loss": 0.0345, + "step": 20720 + }, + { + "epoch": 0.066368, + "grad_norm": 0.12249528680320658, + "learning_rate": 9.987929093561994e-05, + "loss": 0.0322, + "step": 20740 + }, + { + "epoch": 0.066432, + "grad_norm": 0.08542131645152584, + "learning_rate": 9.987905811294711e-05, + "loss": 0.0322, + "step": 20760 + }, + { + "epoch": 0.066496, + "grad_norm": 0.1019385843912144, + "learning_rate": 9.987882506622921e-05, + "loss": 0.0304, + "step": 20780 + }, + { + "epoch": 0.06656, + "grad_norm": 0.08357225682733553, + "learning_rate": 9.987859179546725e-05, + "loss": 0.0301, + "step": 20800 + }, + { + "epoch": 0.066624, + "grad_norm": 0.09715634803833328, + "learning_rate": 9.987835830066233e-05, + "loss": 0.0345, + "step": 20820 + }, + { + "epoch": 0.066688, + "grad_norm": 0.11742762304027937, + "learning_rate": 9.987812458181544e-05, + "loss": 0.0366, + "step": 20840 + }, + { + "epoch": 0.066752, + "grad_norm": 0.13633109161076418, + "learning_rate": 9.987789063892765e-05, + "loss": 0.0375, + "step": 20860 + }, + { + "epoch": 0.066816, + "grad_norm": 0.0905279376663852, + "learning_rate": 9.987765647200003e-05, + "loss": 0.032, + "step": 20880 + }, + { + "epoch": 0.06688, + "grad_norm": 0.09638298452586012, + "learning_rate": 9.987742208103361e-05, + "loss": 0.0337, + "step": 20900 + }, + { + "epoch": 0.066944, + "grad_norm": 0.10902499220861699, + "learning_rate": 9.987718746602945e-05, + "loss": 0.0367, + "step": 20920 + }, + { + "epoch": 0.067008, + "grad_norm": 0.17972055697479955, + "learning_rate": 9.987695262698862e-05, + "loss": 0.0325, + "step": 20940 + }, + { + "epoch": 0.067072, + "grad_norm": 0.0934453518671778, + "learning_rate": 9.987671756391214e-05, + "loss": 0.0358, + "step": 20960 + }, + { + "epoch": 0.067136, + "grad_norm": 0.10079677565414472, + "learning_rate": 9.987648227680109e-05, + "loss": 0.0353, + "step": 20980 + }, + { + "epoch": 0.0672, + "grad_norm": 0.1456544004330609, + "learning_rate": 9.987624676565652e-05, + "loss": 0.0349, + "step": 21000 + }, + { + "epoch": 0.067264, + "grad_norm": 0.13223111664664167, + "learning_rate": 9.987601103047949e-05, + "loss": 0.0337, + "step": 21020 + }, + { + "epoch": 0.067328, + "grad_norm": 0.11760018238690675, + "learning_rate": 9.987577507127105e-05, + "loss": 0.0349, + "step": 21040 + }, + { + "epoch": 0.067392, + "grad_norm": 0.09636714559316599, + "learning_rate": 9.987553888803228e-05, + "loss": 0.0346, + "step": 21060 + }, + { + "epoch": 0.067456, + "grad_norm": 0.12435365321895789, + "learning_rate": 9.987530248076423e-05, + "loss": 0.0309, + "step": 21080 + }, + { + "epoch": 0.06752, + "grad_norm": 0.08448745877960774, + "learning_rate": 9.987506584946796e-05, + "loss": 0.0326, + "step": 21100 + }, + { + "epoch": 0.067584, + "grad_norm": 0.07675594074264326, + "learning_rate": 9.987482899414453e-05, + "loss": 0.0304, + "step": 21120 + }, + { + "epoch": 0.067648, + "grad_norm": 0.2142432419660501, + "learning_rate": 9.9874591914795e-05, + "loss": 0.0349, + "step": 21140 + }, + { + "epoch": 0.067712, + "grad_norm": 0.15751114395153415, + "learning_rate": 9.987435461142047e-05, + "loss": 0.0359, + "step": 21160 + }, + { + "epoch": 0.067776, + "grad_norm": 0.08525954219411724, + "learning_rate": 9.987411708402195e-05, + "loss": 0.0285, + "step": 21180 + }, + { + "epoch": 0.06784, + "grad_norm": 0.09373727876751091, + "learning_rate": 9.987387933260055e-05, + "loss": 0.0312, + "step": 21200 + }, + { + "epoch": 0.067904, + "grad_norm": 0.16463788252822245, + "learning_rate": 9.987364135715731e-05, + "loss": 0.0358, + "step": 21220 + }, + { + "epoch": 0.067968, + "grad_norm": 0.1940625663842927, + "learning_rate": 9.987340315769333e-05, + "loss": 0.0353, + "step": 21240 + }, + { + "epoch": 0.068032, + "grad_norm": 0.2866789336405311, + "learning_rate": 9.987316473420964e-05, + "loss": 0.0351, + "step": 21260 + }, + { + "epoch": 0.068096, + "grad_norm": 0.09824261910585896, + "learning_rate": 9.987292608670736e-05, + "loss": 0.0354, + "step": 21280 + }, + { + "epoch": 0.06816, + "grad_norm": 0.16362017010972113, + "learning_rate": 9.987268721518752e-05, + "loss": 0.0347, + "step": 21300 + }, + { + "epoch": 0.068224, + "grad_norm": 0.11255524733216665, + "learning_rate": 9.987244811965122e-05, + "loss": 0.0337, + "step": 21320 + }, + { + "epoch": 0.068288, + "grad_norm": 0.1727631612070328, + "learning_rate": 9.98722088000995e-05, + "loss": 0.0369, + "step": 21340 + }, + { + "epoch": 0.068352, + "grad_norm": 0.1467422232451959, + "learning_rate": 9.987196925653347e-05, + "loss": 0.0323, + "step": 21360 + }, + { + "epoch": 0.068416, + "grad_norm": 0.08780110882080794, + "learning_rate": 9.987172948895421e-05, + "loss": 0.0331, + "step": 21380 + }, + { + "epoch": 0.06848, + "grad_norm": 0.07612285922676479, + "learning_rate": 9.987148949736275e-05, + "loss": 0.0335, + "step": 21400 + }, + { + "epoch": 0.068544, + "grad_norm": 0.08400785906600124, + "learning_rate": 9.987124928176021e-05, + "loss": 0.0326, + "step": 21420 + }, + { + "epoch": 0.068608, + "grad_norm": 0.10804813391983129, + "learning_rate": 9.987100884214766e-05, + "loss": 0.0325, + "step": 21440 + }, + { + "epoch": 0.068672, + "grad_norm": 0.10586249085626774, + "learning_rate": 9.987076817852616e-05, + "loss": 0.0331, + "step": 21460 + }, + { + "epoch": 0.068736, + "grad_norm": 0.09005642456938082, + "learning_rate": 9.987052729089684e-05, + "loss": 0.0309, + "step": 21480 + }, + { + "epoch": 0.0688, + "grad_norm": 0.13534579842689604, + "learning_rate": 9.987028617926073e-05, + "loss": 0.0359, + "step": 21500 + }, + { + "epoch": 0.068864, + "grad_norm": 0.13799601442537843, + "learning_rate": 9.987004484361892e-05, + "loss": 0.0366, + "step": 21520 + }, + { + "epoch": 0.068928, + "grad_norm": 0.11352284235976513, + "learning_rate": 9.986980328397253e-05, + "loss": 0.0344, + "step": 21540 + }, + { + "epoch": 0.068992, + "grad_norm": 0.13959411558621687, + "learning_rate": 9.986956150032263e-05, + "loss": 0.0364, + "step": 21560 + }, + { + "epoch": 0.069056, + "grad_norm": 0.1580014252229343, + "learning_rate": 9.986931949267026e-05, + "loss": 0.0348, + "step": 21580 + }, + { + "epoch": 0.06912, + "grad_norm": 0.08915989419157275, + "learning_rate": 9.986907726101657e-05, + "loss": 0.0316, + "step": 21600 + }, + { + "epoch": 0.069184, + "grad_norm": 0.14700341291082408, + "learning_rate": 9.986883480536263e-05, + "loss": 0.0298, + "step": 21620 + }, + { + "epoch": 0.069248, + "grad_norm": 0.168560323619228, + "learning_rate": 9.986859212570952e-05, + "loss": 0.0346, + "step": 21640 + }, + { + "epoch": 0.069312, + "grad_norm": 0.17136835262565733, + "learning_rate": 9.986834922205832e-05, + "loss": 0.0349, + "step": 21660 + }, + { + "epoch": 0.069376, + "grad_norm": 0.1172351523982956, + "learning_rate": 9.986810609441014e-05, + "loss": 0.0374, + "step": 21680 + }, + { + "epoch": 0.06944, + "grad_norm": 0.17154488357160136, + "learning_rate": 9.986786274276607e-05, + "loss": 0.0351, + "step": 21700 + }, + { + "epoch": 0.069504, + "grad_norm": 0.11076887481087162, + "learning_rate": 9.98676191671272e-05, + "loss": 0.0391, + "step": 21720 + }, + { + "epoch": 0.069568, + "grad_norm": 0.11195043193378462, + "learning_rate": 9.986737536749461e-05, + "loss": 0.0312, + "step": 21740 + }, + { + "epoch": 0.069632, + "grad_norm": 0.10691111031399213, + "learning_rate": 9.98671313438694e-05, + "loss": 0.035, + "step": 21760 + }, + { + "epoch": 0.069696, + "grad_norm": 0.17322516095407947, + "learning_rate": 9.986688709625269e-05, + "loss": 0.0337, + "step": 21780 + }, + { + "epoch": 0.06976, + "grad_norm": 0.08928676928025138, + "learning_rate": 9.986664262464556e-05, + "loss": 0.0312, + "step": 21800 + }, + { + "epoch": 0.069824, + "grad_norm": 0.10028794145740745, + "learning_rate": 9.986639792904911e-05, + "loss": 0.0313, + "step": 21820 + }, + { + "epoch": 0.069888, + "grad_norm": 0.07475651342883703, + "learning_rate": 9.986615300946444e-05, + "loss": 0.0324, + "step": 21840 + }, + { + "epoch": 0.069952, + "grad_norm": 0.1791172822011138, + "learning_rate": 9.986590786589265e-05, + "loss": 0.0333, + "step": 21860 + }, + { + "epoch": 0.070016, + "grad_norm": 0.17374840945085895, + "learning_rate": 9.986566249833482e-05, + "loss": 0.0374, + "step": 21880 + }, + { + "epoch": 0.07008, + "grad_norm": 0.11277387828941446, + "learning_rate": 9.98654169067921e-05, + "loss": 0.0366, + "step": 21900 + }, + { + "epoch": 0.070144, + "grad_norm": 0.09136608426004007, + "learning_rate": 9.986517109126554e-05, + "loss": 0.034, + "step": 21920 + }, + { + "epoch": 0.070208, + "grad_norm": 0.07601115825397944, + "learning_rate": 9.986492505175628e-05, + "loss": 0.0329, + "step": 21940 + }, + { + "epoch": 0.070272, + "grad_norm": 0.11780046182188607, + "learning_rate": 9.986467878826541e-05, + "loss": 0.0336, + "step": 21960 + }, + { + "epoch": 0.070336, + "grad_norm": 0.10988987242826376, + "learning_rate": 9.986443230079404e-05, + "loss": 0.0338, + "step": 21980 + }, + { + "epoch": 0.0704, + "grad_norm": 0.09965298417804483, + "learning_rate": 9.986418558934328e-05, + "loss": 0.0322, + "step": 22000 + }, + { + "epoch": 0.070464, + "grad_norm": 0.15290961064091546, + "learning_rate": 9.986393865391424e-05, + "loss": 0.0338, + "step": 22020 + }, + { + "epoch": 0.070528, + "grad_norm": 0.11983624100752203, + "learning_rate": 9.986369149450803e-05, + "loss": 0.0328, + "step": 22040 + }, + { + "epoch": 0.070592, + "grad_norm": 0.10846236021220694, + "learning_rate": 9.986344411112574e-05, + "loss": 0.035, + "step": 22060 + }, + { + "epoch": 0.070656, + "grad_norm": 0.09480335695068524, + "learning_rate": 9.98631965037685e-05, + "loss": 0.0356, + "step": 22080 + }, + { + "epoch": 0.07072, + "grad_norm": 0.08494547047690275, + "learning_rate": 9.986294867243741e-05, + "loss": 0.0328, + "step": 22100 + }, + { + "epoch": 0.070784, + "grad_norm": 0.08927285255544838, + "learning_rate": 9.98627006171336e-05, + "loss": 0.0358, + "step": 22120 + }, + { + "epoch": 0.070848, + "grad_norm": 0.1118799919087846, + "learning_rate": 9.986245233785818e-05, + "loss": 0.0352, + "step": 22140 + }, + { + "epoch": 0.070912, + "grad_norm": 0.10905824083152592, + "learning_rate": 9.986220383461226e-05, + "loss": 0.035, + "step": 22160 + }, + { + "epoch": 0.070976, + "grad_norm": 0.08710638804272211, + "learning_rate": 9.986195510739695e-05, + "loss": 0.0337, + "step": 22180 + }, + { + "epoch": 0.07104, + "grad_norm": 0.25457471012115274, + "learning_rate": 9.986170615621338e-05, + "loss": 0.0336, + "step": 22200 + }, + { + "epoch": 0.071104, + "grad_norm": 0.12381429533408675, + "learning_rate": 9.986145698106266e-05, + "loss": 0.0356, + "step": 22220 + }, + { + "epoch": 0.071168, + "grad_norm": 0.11927002446767047, + "learning_rate": 9.986120758194591e-05, + "loss": 0.0363, + "step": 22240 + }, + { + "epoch": 0.071232, + "grad_norm": 0.15291653038749115, + "learning_rate": 9.986095795886425e-05, + "loss": 0.0357, + "step": 22260 + }, + { + "epoch": 0.071296, + "grad_norm": 0.1620245988658392, + "learning_rate": 9.986070811181881e-05, + "loss": 0.0355, + "step": 22280 + }, + { + "epoch": 0.07136, + "grad_norm": 0.11677865005241177, + "learning_rate": 9.986045804081071e-05, + "loss": 0.0329, + "step": 22300 + }, + { + "epoch": 0.071424, + "grad_norm": 0.14375380888355657, + "learning_rate": 9.986020774584106e-05, + "loss": 0.0336, + "step": 22320 + }, + { + "epoch": 0.071488, + "grad_norm": 0.09387330341130333, + "learning_rate": 9.9859957226911e-05, + "loss": 0.0306, + "step": 22340 + }, + { + "epoch": 0.071552, + "grad_norm": 0.10192429603829761, + "learning_rate": 9.985970648402163e-05, + "loss": 0.0345, + "step": 22360 + }, + { + "epoch": 0.071616, + "grad_norm": 0.16854455715990768, + "learning_rate": 9.98594555171741e-05, + "loss": 0.0379, + "step": 22380 + }, + { + "epoch": 0.07168, + "grad_norm": 0.08577447965454207, + "learning_rate": 9.985920432636955e-05, + "loss": 0.0341, + "step": 22400 + }, + { + "epoch": 0.071744, + "grad_norm": 0.12286390524066537, + "learning_rate": 9.985895291160908e-05, + "loss": 0.0359, + "step": 22420 + }, + { + "epoch": 0.071808, + "grad_norm": 0.14869475655142642, + "learning_rate": 9.985870127289383e-05, + "loss": 0.0354, + "step": 22440 + }, + { + "epoch": 0.071872, + "grad_norm": 0.1042488730088998, + "learning_rate": 9.985844941022493e-05, + "loss": 0.0364, + "step": 22460 + }, + { + "epoch": 0.071936, + "grad_norm": 0.10789509209694285, + "learning_rate": 9.985819732360351e-05, + "loss": 0.0344, + "step": 22480 + }, + { + "epoch": 0.072, + "grad_norm": 0.10894234387116458, + "learning_rate": 9.98579450130307e-05, + "loss": 0.0349, + "step": 22500 + }, + { + "epoch": 0.072064, + "grad_norm": 0.11867946751278823, + "learning_rate": 9.985769247850764e-05, + "loss": 0.0291, + "step": 22520 + }, + { + "epoch": 0.072128, + "grad_norm": 0.09705695618330211, + "learning_rate": 9.985743972003546e-05, + "loss": 0.0344, + "step": 22540 + }, + { + "epoch": 0.072192, + "grad_norm": 0.10786022442059426, + "learning_rate": 9.985718673761531e-05, + "loss": 0.0346, + "step": 22560 + }, + { + "epoch": 0.072256, + "grad_norm": 0.11144097486663002, + "learning_rate": 9.985693353124829e-05, + "loss": 0.0331, + "step": 22580 + }, + { + "epoch": 0.07232, + "grad_norm": 0.08289402728592855, + "learning_rate": 9.985668010093556e-05, + "loss": 0.0314, + "step": 22600 + }, + { + "epoch": 0.072384, + "grad_norm": 0.0822978678733783, + "learning_rate": 9.985642644667828e-05, + "loss": 0.0328, + "step": 22620 + }, + { + "epoch": 0.072448, + "grad_norm": 0.1543036128413852, + "learning_rate": 9.985617256847755e-05, + "loss": 0.0352, + "step": 22640 + }, + { + "epoch": 0.072512, + "grad_norm": 0.14063184336609527, + "learning_rate": 9.985591846633453e-05, + "loss": 0.0342, + "step": 22660 + }, + { + "epoch": 0.072576, + "grad_norm": 0.1495811644007825, + "learning_rate": 9.985566414025038e-05, + "loss": 0.0379, + "step": 22680 + }, + { + "epoch": 0.07264, + "grad_norm": 0.106790204224623, + "learning_rate": 9.98554095902262e-05, + "loss": 0.0374, + "step": 22700 + }, + { + "epoch": 0.072704, + "grad_norm": 0.111100937567488, + "learning_rate": 9.985515481626315e-05, + "loss": 0.0346, + "step": 22720 + }, + { + "epoch": 0.072768, + "grad_norm": 0.13819359649565374, + "learning_rate": 9.98548998183624e-05, + "loss": 0.0337, + "step": 22740 + }, + { + "epoch": 0.072832, + "grad_norm": 0.07944071253155681, + "learning_rate": 9.985464459652507e-05, + "loss": 0.0312, + "step": 22760 + }, + { + "epoch": 0.072896, + "grad_norm": 0.1096125354403516, + "learning_rate": 9.98543891507523e-05, + "loss": 0.033, + "step": 22780 + }, + { + "epoch": 0.07296, + "grad_norm": 0.13947531737669475, + "learning_rate": 9.985413348104525e-05, + "loss": 0.0368, + "step": 22800 + }, + { + "epoch": 0.073024, + "grad_norm": 0.15737557657548285, + "learning_rate": 9.985387758740507e-05, + "loss": 0.0369, + "step": 22820 + }, + { + "epoch": 0.073088, + "grad_norm": 0.1301242895645652, + "learning_rate": 9.985362146983292e-05, + "loss": 0.0356, + "step": 22840 + }, + { + "epoch": 0.073152, + "grad_norm": 0.11475803176252446, + "learning_rate": 9.985336512832992e-05, + "loss": 0.0334, + "step": 22860 + }, + { + "epoch": 0.073216, + "grad_norm": 0.12927507541802974, + "learning_rate": 9.985310856289725e-05, + "loss": 0.0332, + "step": 22880 + }, + { + "epoch": 0.07328, + "grad_norm": 0.11828562186544112, + "learning_rate": 9.985285177353604e-05, + "loss": 0.038, + "step": 22900 + }, + { + "epoch": 0.073344, + "grad_norm": 0.20484958284916652, + "learning_rate": 9.985259476024746e-05, + "loss": 0.0354, + "step": 22920 + }, + { + "epoch": 0.073408, + "grad_norm": 0.1352763810071271, + "learning_rate": 9.985233752303265e-05, + "loss": 0.0368, + "step": 22940 + }, + { + "epoch": 0.073472, + "grad_norm": 0.09721109652072536, + "learning_rate": 9.985208006189279e-05, + "loss": 0.0379, + "step": 22960 + }, + { + "epoch": 0.073536, + "grad_norm": 0.1280728472700533, + "learning_rate": 9.985182237682902e-05, + "loss": 0.0354, + "step": 22980 + }, + { + "epoch": 0.0736, + "grad_norm": 0.17153856382797025, + "learning_rate": 9.985156446784248e-05, + "loss": 0.0345, + "step": 23000 + }, + { + "epoch": 0.073664, + "grad_norm": 0.16049632666044922, + "learning_rate": 9.985130633493436e-05, + "loss": 0.0367, + "step": 23020 + }, + { + "epoch": 0.073728, + "grad_norm": 0.10668301542994196, + "learning_rate": 9.98510479781058e-05, + "loss": 0.0353, + "step": 23040 + }, + { + "epoch": 0.073792, + "grad_norm": 0.12396053721861013, + "learning_rate": 9.985078939735797e-05, + "loss": 0.0372, + "step": 23060 + }, + { + "epoch": 0.073856, + "grad_norm": 0.1559179467497307, + "learning_rate": 9.985053059269203e-05, + "loss": 0.037, + "step": 23080 + }, + { + "epoch": 0.07392, + "grad_norm": 0.13852000977347076, + "learning_rate": 9.985027156410915e-05, + "loss": 0.0361, + "step": 23100 + }, + { + "epoch": 0.073984, + "grad_norm": 0.1383224024822978, + "learning_rate": 9.985001231161046e-05, + "loss": 0.0334, + "step": 23120 + }, + { + "epoch": 0.074048, + "grad_norm": 0.12230929256897073, + "learning_rate": 9.984975283519718e-05, + "loss": 0.0357, + "step": 23140 + }, + { + "epoch": 0.074112, + "grad_norm": 0.13142021168582768, + "learning_rate": 9.984949313487042e-05, + "loss": 0.0324, + "step": 23160 + }, + { + "epoch": 0.074176, + "grad_norm": 0.10858457301389712, + "learning_rate": 9.984923321063136e-05, + "loss": 0.0347, + "step": 23180 + }, + { + "epoch": 0.07424, + "grad_norm": 0.1657101590786394, + "learning_rate": 9.98489730624812e-05, + "loss": 0.0311, + "step": 23200 + }, + { + "epoch": 0.074304, + "grad_norm": 0.11723274373611986, + "learning_rate": 9.984871269042109e-05, + "loss": 0.0306, + "step": 23220 + }, + { + "epoch": 0.074368, + "grad_norm": 0.11639365922378339, + "learning_rate": 9.984845209445218e-05, + "loss": 0.0322, + "step": 23240 + }, + { + "epoch": 0.074432, + "grad_norm": 0.15288932894379312, + "learning_rate": 9.984819127457567e-05, + "loss": 0.0327, + "step": 23260 + }, + { + "epoch": 0.074496, + "grad_norm": 0.07563649275887233, + "learning_rate": 9.984793023079271e-05, + "loss": 0.0284, + "step": 23280 + }, + { + "epoch": 0.07456, + "grad_norm": 0.0809438123424736, + "learning_rate": 9.984766896310448e-05, + "loss": 0.0308, + "step": 23300 + }, + { + "epoch": 0.074624, + "grad_norm": 0.1350131396258057, + "learning_rate": 9.984740747151216e-05, + "loss": 0.0383, + "step": 23320 + }, + { + "epoch": 0.074688, + "grad_norm": 0.1860276689511777, + "learning_rate": 9.98471457560169e-05, + "loss": 0.031, + "step": 23340 + }, + { + "epoch": 0.074752, + "grad_norm": 0.11324969659650413, + "learning_rate": 9.984688381661991e-05, + "loss": 0.0334, + "step": 23360 + }, + { + "epoch": 0.074816, + "grad_norm": 0.11570226230839156, + "learning_rate": 9.984662165332236e-05, + "loss": 0.0355, + "step": 23380 + }, + { + "epoch": 0.07488, + "grad_norm": 0.12287423140811278, + "learning_rate": 9.984635926612541e-05, + "loss": 0.0315, + "step": 23400 + }, + { + "epoch": 0.074944, + "grad_norm": 0.1085045436875802, + "learning_rate": 9.984609665503024e-05, + "loss": 0.0347, + "step": 23420 + }, + { + "epoch": 0.075008, + "grad_norm": 0.15013649434954515, + "learning_rate": 9.984583382003804e-05, + "loss": 0.0354, + "step": 23440 + }, + { + "epoch": 0.075072, + "grad_norm": 0.10011970851404092, + "learning_rate": 9.984557076115e-05, + "loss": 0.0349, + "step": 23460 + }, + { + "epoch": 0.075136, + "grad_norm": 0.11176852377881148, + "learning_rate": 9.984530747836727e-05, + "loss": 0.0318, + "step": 23480 + }, + { + "epoch": 0.0752, + "grad_norm": 0.10452527078675526, + "learning_rate": 9.984504397169108e-05, + "loss": 0.0333, + "step": 23500 + }, + { + "epoch": 0.075264, + "grad_norm": 0.1575744902047966, + "learning_rate": 9.984478024112255e-05, + "loss": 0.0334, + "step": 23520 + }, + { + "epoch": 0.075328, + "grad_norm": 0.090154081985554, + "learning_rate": 9.984451628666293e-05, + "loss": 0.0335, + "step": 23540 + }, + { + "epoch": 0.075392, + "grad_norm": 0.10652200170072235, + "learning_rate": 9.984425210831337e-05, + "loss": 0.0331, + "step": 23560 + }, + { + "epoch": 0.075456, + "grad_norm": 0.1433303180431082, + "learning_rate": 9.984398770607504e-05, + "loss": 0.0388, + "step": 23580 + }, + { + "epoch": 0.07552, + "grad_norm": 0.09450120284550108, + "learning_rate": 9.984372307994916e-05, + "loss": 0.0345, + "step": 23600 + }, + { + "epoch": 0.075584, + "grad_norm": 0.07776704489299817, + "learning_rate": 9.984345822993692e-05, + "loss": 0.0358, + "step": 23620 + }, + { + "epoch": 0.075648, + "grad_norm": 0.09212792972449117, + "learning_rate": 9.984319315603949e-05, + "loss": 0.0333, + "step": 23640 + }, + { + "epoch": 0.075712, + "grad_norm": 0.11412755202708111, + "learning_rate": 9.984292785825806e-05, + "loss": 0.0348, + "step": 23660 + }, + { + "epoch": 0.075776, + "grad_norm": 0.0855586312714624, + "learning_rate": 9.984266233659384e-05, + "loss": 0.0354, + "step": 23680 + }, + { + "epoch": 0.07584, + "grad_norm": 0.10354229239345891, + "learning_rate": 9.984239659104802e-05, + "loss": 0.0336, + "step": 23700 + }, + { + "epoch": 0.075904, + "grad_norm": 0.1227634082449925, + "learning_rate": 9.984213062162177e-05, + "loss": 0.0318, + "step": 23720 + }, + { + "epoch": 0.075968, + "grad_norm": 0.07977200879308788, + "learning_rate": 9.984186442831632e-05, + "loss": 0.0332, + "step": 23740 + }, + { + "epoch": 0.076032, + "grad_norm": 0.09641596188427982, + "learning_rate": 9.984159801113282e-05, + "loss": 0.0366, + "step": 23760 + }, + { + "epoch": 0.076096, + "grad_norm": 0.09422931252018964, + "learning_rate": 9.984133137007252e-05, + "loss": 0.0326, + "step": 23780 + }, + { + "epoch": 0.07616, + "grad_norm": 0.09866807114573506, + "learning_rate": 9.984106450513658e-05, + "loss": 0.0323, + "step": 23800 + }, + { + "epoch": 0.076224, + "grad_norm": 0.08813252213801066, + "learning_rate": 9.98407974163262e-05, + "loss": 0.0346, + "step": 23820 + }, + { + "epoch": 0.076288, + "grad_norm": 0.08657876504824362, + "learning_rate": 9.98405301036426e-05, + "loss": 0.0297, + "step": 23840 + }, + { + "epoch": 0.076352, + "grad_norm": 0.11692729636424436, + "learning_rate": 9.984026256708698e-05, + "loss": 0.0301, + "step": 23860 + }, + { + "epoch": 0.076416, + "grad_norm": 0.12023463530061697, + "learning_rate": 9.983999480666051e-05, + "loss": 0.033, + "step": 23880 + }, + { + "epoch": 0.07648, + "grad_norm": 0.09727304732318916, + "learning_rate": 9.983972682236444e-05, + "loss": 0.0303, + "step": 23900 + }, + { + "epoch": 0.076544, + "grad_norm": 0.0881919324293794, + "learning_rate": 9.983945861419992e-05, + "loss": 0.0337, + "step": 23920 + }, + { + "epoch": 0.076608, + "grad_norm": 0.08178845006946162, + "learning_rate": 9.983919018216821e-05, + "loss": 0.0403, + "step": 23940 + }, + { + "epoch": 0.076672, + "grad_norm": 0.09172218751808453, + "learning_rate": 9.983892152627048e-05, + "loss": 0.0341, + "step": 23960 + }, + { + "epoch": 0.076736, + "grad_norm": 0.09422215462398728, + "learning_rate": 9.983865264650794e-05, + "loss": 0.0345, + "step": 23980 + }, + { + "epoch": 0.0768, + "grad_norm": 0.09164471308796546, + "learning_rate": 9.983838354288181e-05, + "loss": 0.0335, + "step": 24000 + }, + { + "epoch": 0.076864, + "grad_norm": 0.09370380265865201, + "learning_rate": 9.983811421539329e-05, + "loss": 0.0337, + "step": 24020 + }, + { + "epoch": 0.076928, + "grad_norm": 0.08397203228215365, + "learning_rate": 9.983784466404358e-05, + "loss": 0.0313, + "step": 24040 + }, + { + "epoch": 0.076992, + "grad_norm": 0.16319011558463653, + "learning_rate": 9.983757488883391e-05, + "loss": 0.0298, + "step": 24060 + }, + { + "epoch": 0.077056, + "grad_norm": 0.13577492284955922, + "learning_rate": 9.98373048897655e-05, + "loss": 0.0326, + "step": 24080 + }, + { + "epoch": 0.07712, + "grad_norm": 0.2080289049387854, + "learning_rate": 9.983703466683953e-05, + "loss": 0.0321, + "step": 24100 + }, + { + "epoch": 0.077184, + "grad_norm": 0.10349573589620358, + "learning_rate": 9.983676422005726e-05, + "loss": 0.0381, + "step": 24120 + }, + { + "epoch": 0.077248, + "grad_norm": 0.09485697956727139, + "learning_rate": 9.983649354941984e-05, + "loss": 0.0329, + "step": 24140 + }, + { + "epoch": 0.077312, + "grad_norm": 0.14470244111587938, + "learning_rate": 9.983622265492854e-05, + "loss": 0.0346, + "step": 24160 + }, + { + "epoch": 0.077376, + "grad_norm": 0.247931010251161, + "learning_rate": 9.983595153658455e-05, + "loss": 0.0325, + "step": 24180 + }, + { + "epoch": 0.07744, + "grad_norm": 0.1160921992607393, + "learning_rate": 9.983568019438909e-05, + "loss": 0.0336, + "step": 24200 + }, + { + "epoch": 0.077504, + "grad_norm": 0.12236922485558886, + "learning_rate": 9.98354086283434e-05, + "loss": 0.0305, + "step": 24220 + }, + { + "epoch": 0.077568, + "grad_norm": 0.10852443449328167, + "learning_rate": 9.983513683844869e-05, + "loss": 0.0339, + "step": 24240 + }, + { + "epoch": 0.077632, + "grad_norm": 0.11681069345985914, + "learning_rate": 9.983486482470616e-05, + "loss": 0.0341, + "step": 24260 + }, + { + "epoch": 0.077696, + "grad_norm": 0.12221354147434337, + "learning_rate": 9.983459258711705e-05, + "loss": 0.036, + "step": 24280 + }, + { + "epoch": 0.07776, + "grad_norm": 0.08925623808077693, + "learning_rate": 9.983432012568259e-05, + "loss": 0.0345, + "step": 24300 + }, + { + "epoch": 0.077824, + "grad_norm": 0.15232534757616956, + "learning_rate": 9.983404744040397e-05, + "loss": 0.0346, + "step": 24320 + }, + { + "epoch": 0.077888, + "grad_norm": 0.1153236023925341, + "learning_rate": 9.983377453128246e-05, + "loss": 0.0348, + "step": 24340 + }, + { + "epoch": 0.077952, + "grad_norm": 0.11645857817320884, + "learning_rate": 9.983350139831926e-05, + "loss": 0.0332, + "step": 24360 + }, + { + "epoch": 0.078016, + "grad_norm": 0.15716213539521012, + "learning_rate": 9.983322804151561e-05, + "loss": 0.0308, + "step": 24380 + }, + { + "epoch": 0.07808, + "grad_norm": 0.09582219389460982, + "learning_rate": 9.983295446087271e-05, + "loss": 0.0389, + "step": 24400 + }, + { + "epoch": 0.078144, + "grad_norm": 0.18008561224826747, + "learning_rate": 9.983268065639183e-05, + "loss": 0.0374, + "step": 24420 + }, + { + "epoch": 0.078208, + "grad_norm": 0.09700448424048778, + "learning_rate": 9.983240662807416e-05, + "loss": 0.0373, + "step": 24440 + }, + { + "epoch": 0.078272, + "grad_norm": 0.09433722466504182, + "learning_rate": 9.983213237592098e-05, + "loss": 0.0362, + "step": 24460 + }, + { + "epoch": 0.078336, + "grad_norm": 0.09844078926611434, + "learning_rate": 9.983185789993345e-05, + "loss": 0.0362, + "step": 24480 + }, + { + "epoch": 0.0784, + "grad_norm": 0.12133351355001834, + "learning_rate": 9.983158320011288e-05, + "loss": 0.0347, + "step": 24500 + }, + { + "epoch": 0.078464, + "grad_norm": 0.10226172322273738, + "learning_rate": 9.983130827646043e-05, + "loss": 0.0318, + "step": 24520 + }, + { + "epoch": 0.078528, + "grad_norm": 0.13081582584461734, + "learning_rate": 9.98310331289774e-05, + "loss": 0.0346, + "step": 24540 + }, + { + "epoch": 0.078592, + "grad_norm": 0.11509693404078751, + "learning_rate": 9.983075775766499e-05, + "loss": 0.0335, + "step": 24560 + }, + { + "epoch": 0.078656, + "grad_norm": 0.10002770801727752, + "learning_rate": 9.983048216252445e-05, + "loss": 0.032, + "step": 24580 + }, + { + "epoch": 0.07872, + "grad_norm": 0.097801171498897, + "learning_rate": 9.983020634355703e-05, + "loss": 0.033, + "step": 24600 + }, + { + "epoch": 0.078784, + "grad_norm": 0.15618786819963487, + "learning_rate": 9.982993030076392e-05, + "loss": 0.0343, + "step": 24620 + }, + { + "epoch": 0.078848, + "grad_norm": 0.1439849322435853, + "learning_rate": 9.98296540341464e-05, + "loss": 0.0333, + "step": 24640 + }, + { + "epoch": 0.078912, + "grad_norm": 0.0841734480477287, + "learning_rate": 9.982937754370571e-05, + "loss": 0.0343, + "step": 24660 + }, + { + "epoch": 0.078976, + "grad_norm": 0.0964035108175424, + "learning_rate": 9.982910082944308e-05, + "loss": 0.0329, + "step": 24680 + }, + { + "epoch": 0.07904, + "grad_norm": 0.12902371162110443, + "learning_rate": 9.982882389135977e-05, + "loss": 0.032, + "step": 24700 + }, + { + "epoch": 0.079104, + "grad_norm": 0.13634758754882226, + "learning_rate": 9.9828546729457e-05, + "loss": 0.0321, + "step": 24720 + }, + { + "epoch": 0.079168, + "grad_norm": 0.11364374593349984, + "learning_rate": 9.982826934373604e-05, + "loss": 0.0356, + "step": 24740 + }, + { + "epoch": 0.079232, + "grad_norm": 0.11119432247672688, + "learning_rate": 9.98279917341981e-05, + "loss": 0.0382, + "step": 24760 + }, + { + "epoch": 0.079296, + "grad_norm": 0.10298859948437575, + "learning_rate": 9.982771390084446e-05, + "loss": 0.0333, + "step": 24780 + }, + { + "epoch": 0.07936, + "grad_norm": 0.12396091007150137, + "learning_rate": 9.982743584367637e-05, + "loss": 0.0332, + "step": 24800 + }, + { + "epoch": 0.079424, + "grad_norm": 0.09996606536090746, + "learning_rate": 9.982715756269505e-05, + "loss": 0.0332, + "step": 24820 + }, + { + "epoch": 0.079488, + "grad_norm": 0.13933585614323685, + "learning_rate": 9.982687905790178e-05, + "loss": 0.0329, + "step": 24840 + }, + { + "epoch": 0.079552, + "grad_norm": 0.09471685794063534, + "learning_rate": 9.982660032929778e-05, + "loss": 0.0298, + "step": 24860 + }, + { + "epoch": 0.079616, + "grad_norm": 0.1374380867744802, + "learning_rate": 9.982632137688435e-05, + "loss": 0.0344, + "step": 24880 + }, + { + "epoch": 0.07968, + "grad_norm": 0.08803363447017606, + "learning_rate": 9.982604220066269e-05, + "loss": 0.0364, + "step": 24900 + }, + { + "epoch": 0.079744, + "grad_norm": 0.1386970521201084, + "learning_rate": 9.982576280063407e-05, + "loss": 0.0335, + "step": 24920 + }, + { + "epoch": 0.079808, + "grad_norm": 0.14197989770467562, + "learning_rate": 9.982548317679976e-05, + "loss": 0.0315, + "step": 24940 + }, + { + "epoch": 0.079872, + "grad_norm": 0.09712238102162671, + "learning_rate": 9.982520332916102e-05, + "loss": 0.0321, + "step": 24960 + }, + { + "epoch": 0.079936, + "grad_norm": 0.09227798757015492, + "learning_rate": 9.982492325771908e-05, + "loss": 0.0303, + "step": 24980 + }, + { + "epoch": 0.08, + "grad_norm": 0.10211406590987661, + "learning_rate": 9.982464296247522e-05, + "loss": 0.0342, + "step": 25000 + }, + { + "epoch": 0.080064, + "grad_norm": 0.18082385185905728, + "learning_rate": 9.98243624434307e-05, + "loss": 0.0318, + "step": 25020 + }, + { + "epoch": 0.080128, + "grad_norm": 0.07103310374430172, + "learning_rate": 9.982408170058675e-05, + "loss": 0.0356, + "step": 25040 + }, + { + "epoch": 0.080192, + "grad_norm": 0.10584007822092685, + "learning_rate": 9.982380073394465e-05, + "loss": 0.0363, + "step": 25060 + }, + { + "epoch": 0.080256, + "grad_norm": 0.09522641144881656, + "learning_rate": 9.982351954350569e-05, + "loss": 0.0351, + "step": 25080 + }, + { + "epoch": 0.08032, + "grad_norm": 0.1425724898902192, + "learning_rate": 9.98232381292711e-05, + "loss": 0.0364, + "step": 25100 + }, + { + "epoch": 0.080384, + "grad_norm": 0.2089148016770719, + "learning_rate": 9.982295649124215e-05, + "loss": 0.0365, + "step": 25120 + }, + { + "epoch": 0.080448, + "grad_norm": 0.2186503553193735, + "learning_rate": 9.98226746294201e-05, + "loss": 0.0312, + "step": 25140 + }, + { + "epoch": 0.080512, + "grad_norm": 0.13424573188961297, + "learning_rate": 9.982239254380624e-05, + "loss": 0.0315, + "step": 25160 + }, + { + "epoch": 0.080576, + "grad_norm": 0.15442275500620958, + "learning_rate": 9.98221102344018e-05, + "loss": 0.0368, + "step": 25180 + }, + { + "epoch": 0.08064, + "grad_norm": 0.1139235449092795, + "learning_rate": 9.982182770120807e-05, + "loss": 0.0365, + "step": 25200 + }, + { + "epoch": 0.080704, + "grad_norm": 0.07326654247311916, + "learning_rate": 9.982154494422633e-05, + "loss": 0.0343, + "step": 25220 + }, + { + "epoch": 0.080768, + "grad_norm": 0.11521027938534616, + "learning_rate": 9.982126196345784e-05, + "loss": 0.0329, + "step": 25240 + }, + { + "epoch": 0.080832, + "grad_norm": 0.09615136105389115, + "learning_rate": 9.982097875890386e-05, + "loss": 0.036, + "step": 25260 + }, + { + "epoch": 0.080896, + "grad_norm": 0.1091019555906166, + "learning_rate": 9.982069533056566e-05, + "loss": 0.0374, + "step": 25280 + }, + { + "epoch": 0.08096, + "grad_norm": 0.10089828094391107, + "learning_rate": 9.982041167844454e-05, + "loss": 0.033, + "step": 25300 + }, + { + "epoch": 0.081024, + "grad_norm": 0.0666994908594565, + "learning_rate": 9.982012780254175e-05, + "loss": 0.0305, + "step": 25320 + }, + { + "epoch": 0.081088, + "grad_norm": 0.10244077735697388, + "learning_rate": 9.981984370285857e-05, + "loss": 0.0356, + "step": 25340 + }, + { + "epoch": 0.081152, + "grad_norm": 0.09437752237104859, + "learning_rate": 9.981955937939629e-05, + "loss": 0.0352, + "step": 25360 + }, + { + "epoch": 0.081216, + "grad_norm": 0.11200821455002646, + "learning_rate": 9.981927483215617e-05, + "loss": 0.0309, + "step": 25380 + }, + { + "epoch": 0.08128, + "grad_norm": 0.08896521758843787, + "learning_rate": 9.981899006113949e-05, + "loss": 0.0335, + "step": 25400 + }, + { + "epoch": 0.081344, + "grad_norm": 0.1644478330031223, + "learning_rate": 9.981870506634752e-05, + "loss": 0.0373, + "step": 25420 + }, + { + "epoch": 0.081408, + "grad_norm": 0.07971674925906712, + "learning_rate": 9.981841984778158e-05, + "loss": 0.0343, + "step": 25440 + }, + { + "epoch": 0.081472, + "grad_norm": 0.10442652834495857, + "learning_rate": 9.981813440544292e-05, + "loss": 0.0317, + "step": 25460 + }, + { + "epoch": 0.081536, + "grad_norm": 0.07785670732939133, + "learning_rate": 9.981784873933283e-05, + "loss": 0.0339, + "step": 25480 + }, + { + "epoch": 0.0816, + "grad_norm": 0.07163519821657759, + "learning_rate": 9.981756284945256e-05, + "loss": 0.0343, + "step": 25500 + }, + { + "epoch": 0.081664, + "grad_norm": 0.3101263292137316, + "learning_rate": 9.981727673580346e-05, + "loss": 0.037, + "step": 25520 + }, + { + "epoch": 0.081728, + "grad_norm": 0.12408034693958458, + "learning_rate": 9.981699039838675e-05, + "loss": 0.0341, + "step": 25540 + }, + { + "epoch": 0.081792, + "grad_norm": 0.10668163958916406, + "learning_rate": 9.981670383720376e-05, + "loss": 0.0354, + "step": 25560 + }, + { + "epoch": 0.081856, + "grad_norm": 0.11666162452583224, + "learning_rate": 9.981641705225576e-05, + "loss": 0.0337, + "step": 25580 + }, + { + "epoch": 0.08192, + "grad_norm": 0.1266754357784379, + "learning_rate": 9.981613004354404e-05, + "loss": 0.0324, + "step": 25600 + }, + { + "epoch": 0.081984, + "grad_norm": 0.13636621417411812, + "learning_rate": 9.98158428110699e-05, + "loss": 0.0309, + "step": 25620 + }, + { + "epoch": 0.082048, + "grad_norm": 0.07406653052635553, + "learning_rate": 9.98155553548346e-05, + "loss": 0.0353, + "step": 25640 + }, + { + "epoch": 0.082112, + "grad_norm": 0.11723635656900046, + "learning_rate": 9.981526767483947e-05, + "loss": 0.0363, + "step": 25660 + }, + { + "epoch": 0.082176, + "grad_norm": 0.09303343243487204, + "learning_rate": 9.981497977108579e-05, + "loss": 0.0337, + "step": 25680 + }, + { + "epoch": 0.08224, + "grad_norm": 0.11279445881566842, + "learning_rate": 9.981469164357481e-05, + "loss": 0.034, + "step": 25700 + }, + { + "epoch": 0.082304, + "grad_norm": 0.13762903968721363, + "learning_rate": 9.981440329230789e-05, + "loss": 0.0343, + "step": 25720 + }, + { + "epoch": 0.082368, + "grad_norm": 0.09792825509024976, + "learning_rate": 9.981411471728629e-05, + "loss": 0.0353, + "step": 25740 + }, + { + "epoch": 0.082432, + "grad_norm": 0.12201563666901388, + "learning_rate": 9.98138259185113e-05, + "loss": 0.0362, + "step": 25760 + }, + { + "epoch": 0.082496, + "grad_norm": 0.09758342960261081, + "learning_rate": 9.981353689598426e-05, + "loss": 0.0345, + "step": 25780 + }, + { + "epoch": 0.08256, + "grad_norm": 0.08516873877131799, + "learning_rate": 9.981324764970641e-05, + "loss": 0.0341, + "step": 25800 + }, + { + "epoch": 0.082624, + "grad_norm": 0.08884866170475066, + "learning_rate": 9.981295817967909e-05, + "loss": 0.0377, + "step": 25820 + }, + { + "epoch": 0.082688, + "grad_norm": 0.17891002467190936, + "learning_rate": 9.981266848590357e-05, + "loss": 0.0328, + "step": 25840 + }, + { + "epoch": 0.082752, + "grad_norm": 0.16415233599777201, + "learning_rate": 9.981237856838118e-05, + "loss": 0.0349, + "step": 25860 + }, + { + "epoch": 0.082816, + "grad_norm": 0.15304691700742276, + "learning_rate": 9.981208842711322e-05, + "loss": 0.0326, + "step": 25880 + }, + { + "epoch": 0.08288, + "grad_norm": 0.098739681703439, + "learning_rate": 9.981179806210097e-05, + "loss": 0.0303, + "step": 25900 + }, + { + "epoch": 0.082944, + "grad_norm": 0.10394103448642981, + "learning_rate": 9.981150747334576e-05, + "loss": 0.0325, + "step": 25920 + }, + { + "epoch": 0.083008, + "grad_norm": 0.09534363779630273, + "learning_rate": 9.981121666084887e-05, + "loss": 0.0343, + "step": 25940 + }, + { + "epoch": 0.083072, + "grad_norm": 0.09794184663173701, + "learning_rate": 9.981092562461162e-05, + "loss": 0.0316, + "step": 25960 + }, + { + "epoch": 0.083136, + "grad_norm": 0.12107921328403581, + "learning_rate": 9.981063436463532e-05, + "loss": 0.0333, + "step": 25980 + }, + { + "epoch": 0.0832, + "grad_norm": 0.12276145898120404, + "learning_rate": 9.981034288092128e-05, + "loss": 0.0354, + "step": 26000 + }, + { + "epoch": 0.083264, + "grad_norm": 0.0728990808026228, + "learning_rate": 9.98100511734708e-05, + "loss": 0.0344, + "step": 26020 + }, + { + "epoch": 0.083328, + "grad_norm": 0.09309040905069606, + "learning_rate": 9.980975924228519e-05, + "loss": 0.03, + "step": 26040 + }, + { + "epoch": 0.083392, + "grad_norm": 0.09601135061338163, + "learning_rate": 9.980946708736578e-05, + "loss": 0.0343, + "step": 26060 + }, + { + "epoch": 0.083456, + "grad_norm": 0.11404231901998185, + "learning_rate": 9.980917470871386e-05, + "loss": 0.0321, + "step": 26080 + }, + { + "epoch": 0.08352, + "grad_norm": 0.14810312951620502, + "learning_rate": 9.980888210633075e-05, + "loss": 0.0358, + "step": 26100 + }, + { + "epoch": 0.083584, + "grad_norm": 0.07238199047553598, + "learning_rate": 9.980858928021777e-05, + "loss": 0.0373, + "step": 26120 + }, + { + "epoch": 0.083648, + "grad_norm": 0.11099168381219922, + "learning_rate": 9.980829623037622e-05, + "loss": 0.0337, + "step": 26140 + }, + { + "epoch": 0.083712, + "grad_norm": 0.09036704843816833, + "learning_rate": 9.980800295680744e-05, + "loss": 0.0324, + "step": 26160 + }, + { + "epoch": 0.083776, + "grad_norm": 0.09777126950432659, + "learning_rate": 9.980770945951273e-05, + "loss": 0.037, + "step": 26180 + }, + { + "epoch": 0.08384, + "grad_norm": 0.13363436890903405, + "learning_rate": 9.980741573849341e-05, + "loss": 0.0349, + "step": 26200 + }, + { + "epoch": 0.083904, + "grad_norm": 0.09468560758356773, + "learning_rate": 9.98071217937508e-05, + "loss": 0.0334, + "step": 26220 + }, + { + "epoch": 0.083968, + "grad_norm": 0.1594617038830298, + "learning_rate": 9.980682762528624e-05, + "loss": 0.0329, + "step": 26240 + }, + { + "epoch": 0.084032, + "grad_norm": 0.09945591035403296, + "learning_rate": 9.980653323310103e-05, + "loss": 0.0336, + "step": 26260 + }, + { + "epoch": 0.084096, + "grad_norm": 0.1364787574294266, + "learning_rate": 9.980623861719648e-05, + "loss": 0.0352, + "step": 26280 + }, + { + "epoch": 0.08416, + "grad_norm": 0.07860179760009363, + "learning_rate": 9.980594377757394e-05, + "loss": 0.0373, + "step": 26300 + }, + { + "epoch": 0.084224, + "grad_norm": 0.08982282756780344, + "learning_rate": 9.980564871423473e-05, + "loss": 0.032, + "step": 26320 + }, + { + "epoch": 0.084288, + "grad_norm": 0.0750707072305192, + "learning_rate": 9.980535342718017e-05, + "loss": 0.034, + "step": 26340 + }, + { + "epoch": 0.084352, + "grad_norm": 0.06625939120329474, + "learning_rate": 9.980505791641157e-05, + "loss": 0.0349, + "step": 26360 + }, + { + "epoch": 0.084416, + "grad_norm": 0.16785140802536722, + "learning_rate": 9.980476218193026e-05, + "loss": 0.0349, + "step": 26380 + }, + { + "epoch": 0.08448, + "grad_norm": 0.2214128378608345, + "learning_rate": 9.980446622373763e-05, + "loss": 0.0341, + "step": 26400 + }, + { + "epoch": 0.084544, + "grad_norm": 0.248307276431833, + "learning_rate": 9.980417004183492e-05, + "loss": 0.0318, + "step": 26420 + }, + { + "epoch": 0.084608, + "grad_norm": 0.18916924104820848, + "learning_rate": 9.980387363622352e-05, + "loss": 0.0353, + "step": 26440 + }, + { + "epoch": 0.084672, + "grad_norm": 0.13319409896760004, + "learning_rate": 9.980357700690473e-05, + "loss": 0.0353, + "step": 26460 + }, + { + "epoch": 0.084736, + "grad_norm": 0.1784927564508636, + "learning_rate": 9.98032801538799e-05, + "loss": 0.034, + "step": 26480 + }, + { + "epoch": 0.0848, + "grad_norm": 0.09985811554126596, + "learning_rate": 9.980298307715037e-05, + "loss": 0.029, + "step": 26500 + }, + { + "epoch": 0.084864, + "grad_norm": 0.1042248250719875, + "learning_rate": 9.980268577671745e-05, + "loss": 0.031, + "step": 26520 + }, + { + "epoch": 0.084928, + "grad_norm": 0.12675606112647853, + "learning_rate": 9.98023882525825e-05, + "loss": 0.0347, + "step": 26540 + }, + { + "epoch": 0.084992, + "grad_norm": 0.1555947064913012, + "learning_rate": 9.980209050474682e-05, + "loss": 0.0338, + "step": 26560 + }, + { + "epoch": 0.085056, + "grad_norm": 0.2002696071643262, + "learning_rate": 9.98017925332118e-05, + "loss": 0.0334, + "step": 26580 + }, + { + "epoch": 0.08512, + "grad_norm": 0.16145064552152683, + "learning_rate": 9.980149433797874e-05, + "loss": 0.0359, + "step": 26600 + }, + { + "epoch": 0.085184, + "grad_norm": 0.12937012871863116, + "learning_rate": 9.9801195919049e-05, + "loss": 0.0343, + "step": 26620 + }, + { + "epoch": 0.085248, + "grad_norm": 0.125592760727966, + "learning_rate": 9.980089727642388e-05, + "loss": 0.0342, + "step": 26640 + }, + { + "epoch": 0.085312, + "grad_norm": 0.10607587741499251, + "learning_rate": 9.980059841010478e-05, + "loss": 0.0317, + "step": 26660 + }, + { + "epoch": 0.085376, + "grad_norm": 0.09229833577966556, + "learning_rate": 9.9800299320093e-05, + "loss": 0.0346, + "step": 26680 + }, + { + "epoch": 0.08544, + "grad_norm": 0.121626444886051, + "learning_rate": 9.980000000638992e-05, + "loss": 0.0316, + "step": 26700 + }, + { + "epoch": 0.085504, + "grad_norm": 0.14013946994825283, + "learning_rate": 9.979970046899684e-05, + "loss": 0.0325, + "step": 26720 + }, + { + "epoch": 0.085568, + "grad_norm": 0.0898637998612099, + "learning_rate": 9.979940070791513e-05, + "loss": 0.0354, + "step": 26740 + }, + { + "epoch": 0.085632, + "grad_norm": 0.17266235695287616, + "learning_rate": 9.979910072314613e-05, + "loss": 0.0348, + "step": 26760 + }, + { + "epoch": 0.085696, + "grad_norm": 0.18300040009353344, + "learning_rate": 9.97988005146912e-05, + "loss": 0.0368, + "step": 26780 + }, + { + "epoch": 0.08576, + "grad_norm": 0.09865968202329661, + "learning_rate": 9.979850008255169e-05, + "loss": 0.0373, + "step": 26800 + }, + { + "epoch": 0.085824, + "grad_norm": 0.1192016983445076, + "learning_rate": 9.979819942672892e-05, + "loss": 0.0342, + "step": 26820 + }, + { + "epoch": 0.085888, + "grad_norm": 0.07038640132499667, + "learning_rate": 9.979789854722428e-05, + "loss": 0.0323, + "step": 26840 + }, + { + "epoch": 0.085952, + "grad_norm": 0.09984638227835767, + "learning_rate": 9.979759744403909e-05, + "loss": 0.0328, + "step": 26860 + }, + { + "epoch": 0.086016, + "grad_norm": 0.08778506527520731, + "learning_rate": 9.979729611717469e-05, + "loss": 0.0343, + "step": 26880 + }, + { + "epoch": 0.08608, + "grad_norm": 0.16102979882329513, + "learning_rate": 9.979699456663249e-05, + "loss": 0.0355, + "step": 26900 + }, + { + "epoch": 0.086144, + "grad_norm": 0.10599790081675076, + "learning_rate": 9.979669279241381e-05, + "loss": 0.0323, + "step": 26920 + }, + { + "epoch": 0.086208, + "grad_norm": 0.17778375969959734, + "learning_rate": 9.979639079451999e-05, + "loss": 0.0324, + "step": 26940 + }, + { + "epoch": 0.086272, + "grad_norm": 0.13130108629491347, + "learning_rate": 9.979608857295242e-05, + "loss": 0.0333, + "step": 26960 + }, + { + "epoch": 0.086336, + "grad_norm": 0.1332503811098483, + "learning_rate": 9.979578612771244e-05, + "loss": 0.0289, + "step": 26980 + }, + { + "epoch": 0.0864, + "grad_norm": 0.10135866634209033, + "learning_rate": 9.979548345880141e-05, + "loss": 0.0355, + "step": 27000 + }, + { + "epoch": 0.086464, + "grad_norm": 0.08881258103630237, + "learning_rate": 9.979518056622068e-05, + "loss": 0.0361, + "step": 27020 + }, + { + "epoch": 0.086528, + "grad_norm": 0.10500893021948983, + "learning_rate": 9.979487744997162e-05, + "loss": 0.0311, + "step": 27040 + }, + { + "epoch": 0.086592, + "grad_norm": 0.08634608811932463, + "learning_rate": 9.97945741100556e-05, + "loss": 0.0325, + "step": 27060 + }, + { + "epoch": 0.086656, + "grad_norm": 0.11201637936309917, + "learning_rate": 9.979427054647396e-05, + "loss": 0.0338, + "step": 27080 + }, + { + "epoch": 0.08672, + "grad_norm": 0.09044858487049827, + "learning_rate": 9.97939667592281e-05, + "loss": 0.0311, + "step": 27100 + }, + { + "epoch": 0.086784, + "grad_norm": 0.1775563615034672, + "learning_rate": 9.979366274831934e-05, + "loss": 0.0317, + "step": 27120 + }, + { + "epoch": 0.086848, + "grad_norm": 0.09649282813404556, + "learning_rate": 9.979335851374908e-05, + "loss": 0.0325, + "step": 27140 + }, + { + "epoch": 0.086912, + "grad_norm": 0.08654850027011385, + "learning_rate": 9.979305405551866e-05, + "loss": 0.033, + "step": 27160 + }, + { + "epoch": 0.086976, + "grad_norm": 0.21517557322627548, + "learning_rate": 9.979274937362948e-05, + "loss": 0.0328, + "step": 27180 + }, + { + "epoch": 0.08704, + "grad_norm": 0.08473273610198778, + "learning_rate": 9.979244446808289e-05, + "loss": 0.0329, + "step": 27200 + }, + { + "epoch": 0.087104, + "grad_norm": 0.1397192087903003, + "learning_rate": 9.979213933888025e-05, + "loss": 0.0319, + "step": 27220 + }, + { + "epoch": 0.087168, + "grad_norm": 0.08650307730105586, + "learning_rate": 9.979183398602295e-05, + "loss": 0.0345, + "step": 27240 + }, + { + "epoch": 0.087232, + "grad_norm": 0.16456251715354028, + "learning_rate": 9.979152840951233e-05, + "loss": 0.0313, + "step": 27260 + }, + { + "epoch": 0.087296, + "grad_norm": 0.13599818947399986, + "learning_rate": 9.979122260934981e-05, + "loss": 0.0337, + "step": 27280 + }, + { + "epoch": 0.08736, + "grad_norm": 0.10749821023105077, + "learning_rate": 9.979091658553672e-05, + "loss": 0.0322, + "step": 27300 + }, + { + "epoch": 0.087424, + "grad_norm": 0.0865907975129192, + "learning_rate": 9.979061033807446e-05, + "loss": 0.0346, + "step": 27320 + }, + { + "epoch": 0.087488, + "grad_norm": 0.11873638650971817, + "learning_rate": 9.97903038669644e-05, + "loss": 0.0345, + "step": 27340 + }, + { + "epoch": 0.087552, + "grad_norm": 0.14414853511156653, + "learning_rate": 9.978999717220791e-05, + "loss": 0.0302, + "step": 27360 + }, + { + "epoch": 0.087616, + "grad_norm": 0.10027434094727658, + "learning_rate": 9.978969025380639e-05, + "loss": 0.0335, + "step": 27380 + }, + { + "epoch": 0.08768, + "grad_norm": 0.17439373895053134, + "learning_rate": 9.978938311176118e-05, + "loss": 0.0356, + "step": 27400 + }, + { + "epoch": 0.087744, + "grad_norm": 0.0952652862053885, + "learning_rate": 9.97890757460737e-05, + "loss": 0.0342, + "step": 27420 + }, + { + "epoch": 0.087808, + "grad_norm": 0.12062573839458407, + "learning_rate": 9.97887681567453e-05, + "loss": 0.0347, + "step": 27440 + }, + { + "epoch": 0.087872, + "grad_norm": 0.08271422349877665, + "learning_rate": 9.978846034377737e-05, + "loss": 0.0326, + "step": 27460 + }, + { + "epoch": 0.087936, + "grad_norm": 0.09099238308057277, + "learning_rate": 9.978815230717128e-05, + "loss": 0.0345, + "step": 27480 + }, + { + "epoch": 0.088, + "grad_norm": 0.08389020146211872, + "learning_rate": 9.978784404692847e-05, + "loss": 0.0323, + "step": 27500 + }, + { + "epoch": 0.088064, + "grad_norm": 0.09318687815543503, + "learning_rate": 9.978753556305025e-05, + "loss": 0.0303, + "step": 27520 + }, + { + "epoch": 0.088128, + "grad_norm": 0.1736526090809674, + "learning_rate": 9.978722685553806e-05, + "loss": 0.0342, + "step": 27540 + }, + { + "epoch": 0.088192, + "grad_norm": 0.06556682594955504, + "learning_rate": 9.978691792439326e-05, + "loss": 0.0315, + "step": 27560 + }, + { + "epoch": 0.088256, + "grad_norm": 0.16506186730723346, + "learning_rate": 9.978660876961723e-05, + "loss": 0.0334, + "step": 27580 + }, + { + "epoch": 0.08832, + "grad_norm": 0.17320003960370256, + "learning_rate": 9.978629939121138e-05, + "loss": 0.0346, + "step": 27600 + }, + { + "epoch": 0.088384, + "grad_norm": 0.06836463647767692, + "learning_rate": 9.97859897891771e-05, + "loss": 0.0312, + "step": 27620 + }, + { + "epoch": 0.088448, + "grad_norm": 0.14102166469649105, + "learning_rate": 9.978567996351577e-05, + "loss": 0.0342, + "step": 27640 + }, + { + "epoch": 0.088512, + "grad_norm": 0.11282781762951137, + "learning_rate": 9.978536991422877e-05, + "loss": 0.0297, + "step": 27660 + }, + { + "epoch": 0.088576, + "grad_norm": 0.09442820131078558, + "learning_rate": 9.978505964131753e-05, + "loss": 0.0308, + "step": 27680 + }, + { + "epoch": 0.08864, + "grad_norm": 0.06881502349206892, + "learning_rate": 9.97847491447834e-05, + "loss": 0.0338, + "step": 27700 + }, + { + "epoch": 0.088704, + "grad_norm": 0.1661319892102743, + "learning_rate": 9.978443842462781e-05, + "loss": 0.0323, + "step": 27720 + }, + { + "epoch": 0.088768, + "grad_norm": 0.12367270917427937, + "learning_rate": 9.978412748085215e-05, + "loss": 0.0317, + "step": 27740 + }, + { + "epoch": 0.088832, + "grad_norm": 0.17380886112768024, + "learning_rate": 9.978381631345779e-05, + "loss": 0.0339, + "step": 27760 + }, + { + "epoch": 0.088896, + "grad_norm": 0.09631126346224517, + "learning_rate": 9.978350492244614e-05, + "loss": 0.0309, + "step": 27780 + }, + { + "epoch": 0.08896, + "grad_norm": 0.11097409012323432, + "learning_rate": 9.978319330781862e-05, + "loss": 0.0341, + "step": 27800 + }, + { + "epoch": 0.089024, + "grad_norm": 0.10458043069137153, + "learning_rate": 9.978288146957661e-05, + "loss": 0.0357, + "step": 27820 + }, + { + "epoch": 0.089088, + "grad_norm": 0.07451959272347594, + "learning_rate": 9.978256940772151e-05, + "loss": 0.0308, + "step": 27840 + }, + { + "epoch": 0.089152, + "grad_norm": 0.07010272719535834, + "learning_rate": 9.978225712225474e-05, + "loss": 0.0342, + "step": 27860 + }, + { + "epoch": 0.089216, + "grad_norm": 0.08730630513972729, + "learning_rate": 9.978194461317768e-05, + "loss": 0.0318, + "step": 27880 + }, + { + "epoch": 0.08928, + "grad_norm": 0.11438032722933647, + "learning_rate": 9.978163188049172e-05, + "loss": 0.0346, + "step": 27900 + }, + { + "epoch": 0.089344, + "grad_norm": 0.18069230727615407, + "learning_rate": 9.978131892419832e-05, + "loss": 0.033, + "step": 27920 + }, + { + "epoch": 0.089408, + "grad_norm": 0.13817597701472376, + "learning_rate": 9.978100574429883e-05, + "loss": 0.0351, + "step": 27940 + }, + { + "epoch": 0.089472, + "grad_norm": 0.16849240827331133, + "learning_rate": 9.97806923407947e-05, + "loss": 0.0349, + "step": 27960 + }, + { + "epoch": 0.089536, + "grad_norm": 0.13684633978342267, + "learning_rate": 9.97803787136873e-05, + "loss": 0.0365, + "step": 27980 + }, + { + "epoch": 0.0896, + "grad_norm": 0.12233710523359623, + "learning_rate": 9.978006486297808e-05, + "loss": 0.0379, + "step": 28000 + }, + { + "epoch": 0.089664, + "grad_norm": 0.12054786214814472, + "learning_rate": 9.977975078866841e-05, + "loss": 0.033, + "step": 28020 + }, + { + "epoch": 0.089728, + "grad_norm": 0.12509239778608508, + "learning_rate": 9.977943649075971e-05, + "loss": 0.0317, + "step": 28040 + }, + { + "epoch": 0.089792, + "grad_norm": 0.12920034013048778, + "learning_rate": 9.97791219692534e-05, + "loss": 0.0313, + "step": 28060 + }, + { + "epoch": 0.089856, + "grad_norm": 0.08254257453077578, + "learning_rate": 9.977880722415092e-05, + "loss": 0.0286, + "step": 28080 + }, + { + "epoch": 0.08992, + "grad_norm": 0.07514165766367642, + "learning_rate": 9.977849225545362e-05, + "loss": 0.0316, + "step": 28100 + }, + { + "epoch": 0.089984, + "grad_norm": 0.17487402618519027, + "learning_rate": 9.977817706316297e-05, + "loss": 0.0377, + "step": 28120 + }, + { + "epoch": 0.090048, + "grad_norm": 0.15835032516827238, + "learning_rate": 9.977786164728036e-05, + "loss": 0.035, + "step": 28140 + }, + { + "epoch": 0.090112, + "grad_norm": 0.07948123167654329, + "learning_rate": 9.977754600780721e-05, + "loss": 0.037, + "step": 28160 + }, + { + "epoch": 0.090176, + "grad_norm": 0.16162832387698114, + "learning_rate": 9.977723014474494e-05, + "loss": 0.0297, + "step": 28180 + }, + { + "epoch": 0.09024, + "grad_norm": 0.1089538191220854, + "learning_rate": 9.977691405809498e-05, + "loss": 0.04, + "step": 28200 + }, + { + "epoch": 0.090304, + "grad_norm": 0.10656254509044752, + "learning_rate": 9.977659774785873e-05, + "loss": 0.0404, + "step": 28220 + }, + { + "epoch": 0.090368, + "grad_norm": 0.10815279926617702, + "learning_rate": 9.977628121403761e-05, + "loss": 0.0343, + "step": 28240 + }, + { + "epoch": 0.090432, + "grad_norm": 0.2002619528517407, + "learning_rate": 9.977596445663306e-05, + "loss": 0.032, + "step": 28260 + }, + { + "epoch": 0.090496, + "grad_norm": 0.13876161554494884, + "learning_rate": 9.977564747564649e-05, + "loss": 0.0346, + "step": 28280 + }, + { + "epoch": 0.09056, + "grad_norm": 0.11300051995159172, + "learning_rate": 9.977533027107934e-05, + "loss": 0.0342, + "step": 28300 + }, + { + "epoch": 0.090624, + "grad_norm": 0.10960332432074564, + "learning_rate": 9.977501284293302e-05, + "loss": 0.036, + "step": 28320 + }, + { + "epoch": 0.090688, + "grad_norm": 0.08914569078037383, + "learning_rate": 9.977469519120894e-05, + "loss": 0.0327, + "step": 28340 + }, + { + "epoch": 0.090752, + "grad_norm": 0.11508348841481472, + "learning_rate": 9.977437731590858e-05, + "loss": 0.0306, + "step": 28360 + }, + { + "epoch": 0.090816, + "grad_norm": 0.1742130513040511, + "learning_rate": 9.97740592170333e-05, + "loss": 0.0325, + "step": 28380 + }, + { + "epoch": 0.09088, + "grad_norm": 0.2091599445411073, + "learning_rate": 9.977374089458458e-05, + "loss": 0.0332, + "step": 28400 + }, + { + "epoch": 0.090944, + "grad_norm": 0.18037912833784317, + "learning_rate": 9.977342234856382e-05, + "loss": 0.0309, + "step": 28420 + }, + { + "epoch": 0.091008, + "grad_norm": 0.0749426721472801, + "learning_rate": 9.977310357897247e-05, + "loss": 0.0307, + "step": 28440 + }, + { + "epoch": 0.091072, + "grad_norm": 0.08944903794497892, + "learning_rate": 9.977278458581196e-05, + "loss": 0.0342, + "step": 28460 + }, + { + "epoch": 0.091136, + "grad_norm": 0.10388976947840407, + "learning_rate": 9.977246536908372e-05, + "loss": 0.0321, + "step": 28480 + }, + { + "epoch": 0.0912, + "grad_norm": 0.07075066927257913, + "learning_rate": 9.977214592878916e-05, + "loss": 0.0342, + "step": 28500 + }, + { + "epoch": 0.091264, + "grad_norm": 0.12765818173255997, + "learning_rate": 9.977182626492976e-05, + "loss": 0.0339, + "step": 28520 + }, + { + "epoch": 0.091328, + "grad_norm": 0.15251325618986036, + "learning_rate": 9.977150637750692e-05, + "loss": 0.0358, + "step": 28540 + }, + { + "epoch": 0.091392, + "grad_norm": 0.09663667398433569, + "learning_rate": 9.977118626652208e-05, + "loss": 0.0368, + "step": 28560 + }, + { + "epoch": 0.091456, + "grad_norm": 0.1051818396380922, + "learning_rate": 9.97708659319767e-05, + "loss": 0.0336, + "step": 28580 + }, + { + "epoch": 0.09152, + "grad_norm": 0.07911653600016368, + "learning_rate": 9.97705453738722e-05, + "loss": 0.0336, + "step": 28600 + }, + { + "epoch": 0.091584, + "grad_norm": 0.11767998133691641, + "learning_rate": 9.977022459221004e-05, + "loss": 0.0337, + "step": 28620 + }, + { + "epoch": 0.091648, + "grad_norm": 0.11756485292877743, + "learning_rate": 9.976990358699163e-05, + "loss": 0.0335, + "step": 28640 + }, + { + "epoch": 0.091712, + "grad_norm": 0.08021285901175416, + "learning_rate": 9.976958235821842e-05, + "loss": 0.0321, + "step": 28660 + }, + { + "epoch": 0.091776, + "grad_norm": 0.0961326009363166, + "learning_rate": 9.976926090589188e-05, + "loss": 0.0307, + "step": 28680 + }, + { + "epoch": 0.09184, + "grad_norm": 0.09043633771366169, + "learning_rate": 9.976893923001342e-05, + "loss": 0.0307, + "step": 28700 + }, + { + "epoch": 0.091904, + "grad_norm": 0.09129606158111475, + "learning_rate": 9.976861733058452e-05, + "loss": 0.0321, + "step": 28720 + }, + { + "epoch": 0.091968, + "grad_norm": 0.07641090636134419, + "learning_rate": 9.97682952076066e-05, + "loss": 0.0298, + "step": 28740 + }, + { + "epoch": 0.092032, + "grad_norm": 0.16801390969116842, + "learning_rate": 9.97679728610811e-05, + "loss": 0.0309, + "step": 28760 + }, + { + "epoch": 0.092096, + "grad_norm": 0.1418208675410538, + "learning_rate": 9.976765029100949e-05, + "loss": 0.0352, + "step": 28780 + }, + { + "epoch": 0.09216, + "grad_norm": 0.11327348210813974, + "learning_rate": 9.97673274973932e-05, + "loss": 0.0343, + "step": 28800 + }, + { + "epoch": 0.092224, + "grad_norm": 0.1771274889364176, + "learning_rate": 9.976700448023371e-05, + "loss": 0.0343, + "step": 28820 + }, + { + "epoch": 0.092288, + "grad_norm": 0.08791349862441265, + "learning_rate": 9.976668123953243e-05, + "loss": 0.0313, + "step": 28840 + }, + { + "epoch": 0.092352, + "grad_norm": 0.0966143717904765, + "learning_rate": 9.976635777529085e-05, + "loss": 0.0337, + "step": 28860 + }, + { + "epoch": 0.092416, + "grad_norm": 0.11653646483985518, + "learning_rate": 9.976603408751037e-05, + "loss": 0.0309, + "step": 28880 + }, + { + "epoch": 0.09248, + "grad_norm": 0.08189381578065129, + "learning_rate": 9.976571017619252e-05, + "loss": 0.0312, + "step": 28900 + }, + { + "epoch": 0.092544, + "grad_norm": 0.08897685932052636, + "learning_rate": 9.97653860413387e-05, + "loss": 0.0306, + "step": 28920 + }, + { + "epoch": 0.092608, + "grad_norm": 0.09438710625565576, + "learning_rate": 9.976506168295036e-05, + "loss": 0.0325, + "step": 28940 + }, + { + "epoch": 0.092672, + "grad_norm": 0.10762498007009522, + "learning_rate": 9.9764737101029e-05, + "loss": 0.0328, + "step": 28960 + }, + { + "epoch": 0.092736, + "grad_norm": 0.08529701631277038, + "learning_rate": 9.976441229557605e-05, + "loss": 0.0308, + "step": 28980 + }, + { + "epoch": 0.0928, + "grad_norm": 0.19667944233611961, + "learning_rate": 9.976408726659296e-05, + "loss": 0.0319, + "step": 29000 + }, + { + "epoch": 0.092864, + "grad_norm": 0.12465251366465091, + "learning_rate": 9.976376201408123e-05, + "loss": 0.0329, + "step": 29020 + }, + { + "epoch": 0.092928, + "grad_norm": 0.08833566639217344, + "learning_rate": 9.976343653804227e-05, + "loss": 0.0327, + "step": 29040 + }, + { + "epoch": 0.092992, + "grad_norm": 0.07028769280427428, + "learning_rate": 9.97631108384776e-05, + "loss": 0.0313, + "step": 29060 + }, + { + "epoch": 0.093056, + "grad_norm": 0.1633563968048828, + "learning_rate": 9.976278491538862e-05, + "loss": 0.0311, + "step": 29080 + }, + { + "epoch": 0.09312, + "grad_norm": 0.10146962029610429, + "learning_rate": 9.976245876877683e-05, + "loss": 0.0324, + "step": 29100 + }, + { + "epoch": 0.093184, + "grad_norm": 0.18894036340895595, + "learning_rate": 9.97621323986437e-05, + "loss": 0.0338, + "step": 29120 + }, + { + "epoch": 0.093248, + "grad_norm": 0.11496772381155952, + "learning_rate": 9.976180580499065e-05, + "loss": 0.0323, + "step": 29140 + }, + { + "epoch": 0.093312, + "grad_norm": 0.1113054493261797, + "learning_rate": 9.976147898781923e-05, + "loss": 0.0302, + "step": 29160 + }, + { + "epoch": 0.093376, + "grad_norm": 0.10759702933483582, + "learning_rate": 9.976115194713083e-05, + "loss": 0.029, + "step": 29180 + }, + { + "epoch": 0.09344, + "grad_norm": 0.15409999986293257, + "learning_rate": 9.976082468292695e-05, + "loss": 0.0309, + "step": 29200 + }, + { + "epoch": 0.093504, + "grad_norm": 0.16466247595901332, + "learning_rate": 9.976049719520906e-05, + "loss": 0.0328, + "step": 29220 + }, + { + "epoch": 0.093568, + "grad_norm": 0.10255773618426264, + "learning_rate": 9.976016948397864e-05, + "loss": 0.0329, + "step": 29240 + }, + { + "epoch": 0.093632, + "grad_norm": 0.08808547177020168, + "learning_rate": 9.975984154923715e-05, + "loss": 0.0322, + "step": 29260 + }, + { + "epoch": 0.093696, + "grad_norm": 0.07704467978912141, + "learning_rate": 9.975951339098603e-05, + "loss": 0.031, + "step": 29280 + }, + { + "epoch": 0.09376, + "grad_norm": 0.14188235382491463, + "learning_rate": 9.975918500922683e-05, + "loss": 0.0345, + "step": 29300 + }, + { + "epoch": 0.093824, + "grad_norm": 0.08098600184545063, + "learning_rate": 9.975885640396096e-05, + "loss": 0.0316, + "step": 29320 + }, + { + "epoch": 0.093888, + "grad_norm": 0.087989729625399, + "learning_rate": 9.975852757518994e-05, + "loss": 0.0347, + "step": 29340 + }, + { + "epoch": 0.093952, + "grad_norm": 0.06656259638325708, + "learning_rate": 9.975819852291522e-05, + "loss": 0.0313, + "step": 29360 + }, + { + "epoch": 0.094016, + "grad_norm": 0.10418907382044858, + "learning_rate": 9.975786924713829e-05, + "loss": 0.0341, + "step": 29380 + }, + { + "epoch": 0.09408, + "grad_norm": 0.08513960067492725, + "learning_rate": 9.97575397478606e-05, + "loss": 0.0326, + "step": 29400 + }, + { + "epoch": 0.094144, + "grad_norm": 0.14662119241223565, + "learning_rate": 9.975721002508367e-05, + "loss": 0.034, + "step": 29420 + }, + { + "epoch": 0.094208, + "grad_norm": 0.08161096131364243, + "learning_rate": 9.975688007880897e-05, + "loss": 0.0335, + "step": 29440 + }, + { + "epoch": 0.094272, + "grad_norm": 0.09768885462538904, + "learning_rate": 9.975654990903797e-05, + "loss": 0.0318, + "step": 29460 + }, + { + "epoch": 0.094336, + "grad_norm": 0.10013869506692491, + "learning_rate": 9.975621951577215e-05, + "loss": 0.0335, + "step": 29480 + }, + { + "epoch": 0.0944, + "grad_norm": 0.18187472433525523, + "learning_rate": 9.975588889901301e-05, + "loss": 0.0356, + "step": 29500 + }, + { + "epoch": 0.094464, + "grad_norm": 0.18013832280643413, + "learning_rate": 9.975555805876204e-05, + "loss": 0.0328, + "step": 29520 + }, + { + "epoch": 0.094528, + "grad_norm": 0.22365722801718557, + "learning_rate": 9.975522699502072e-05, + "loss": 0.0329, + "step": 29540 + }, + { + "epoch": 0.094592, + "grad_norm": 0.16240296267371354, + "learning_rate": 9.975489570779051e-05, + "loss": 0.0331, + "step": 29560 + }, + { + "epoch": 0.094656, + "grad_norm": 0.06403039487609805, + "learning_rate": 9.975456419707293e-05, + "loss": 0.0312, + "step": 29580 + }, + { + "epoch": 0.09472, + "grad_norm": 0.08158239065980073, + "learning_rate": 9.975423246286945e-05, + "loss": 0.0309, + "step": 29600 + }, + { + "epoch": 0.094784, + "grad_norm": 0.08281976552953248, + "learning_rate": 9.975390050518159e-05, + "loss": 0.0329, + "step": 29620 + }, + { + "epoch": 0.094848, + "grad_norm": 0.07309871563540384, + "learning_rate": 9.975356832401081e-05, + "loss": 0.0343, + "step": 29640 + }, + { + "epoch": 0.094912, + "grad_norm": 0.06656807189282106, + "learning_rate": 9.97532359193586e-05, + "loss": 0.0298, + "step": 29660 + }, + { + "epoch": 0.094976, + "grad_norm": 0.07547274121983936, + "learning_rate": 9.975290329122649e-05, + "loss": 0.0357, + "step": 29680 + }, + { + "epoch": 0.09504, + "grad_norm": 0.07419065009560412, + "learning_rate": 9.975257043961592e-05, + "loss": 0.0316, + "step": 29700 + }, + { + "epoch": 0.095104, + "grad_norm": 0.11674756706305164, + "learning_rate": 9.975223736452844e-05, + "loss": 0.0321, + "step": 29720 + }, + { + "epoch": 0.095168, + "grad_norm": 0.10451441946586823, + "learning_rate": 9.975190406596551e-05, + "loss": 0.0307, + "step": 29740 + }, + { + "epoch": 0.095232, + "grad_norm": 0.11090248530285751, + "learning_rate": 9.975157054392863e-05, + "loss": 0.0346, + "step": 29760 + }, + { + "epoch": 0.095296, + "grad_norm": 0.13180218896846468, + "learning_rate": 9.975123679841932e-05, + "loss": 0.0302, + "step": 29780 + }, + { + "epoch": 0.09536, + "grad_norm": 0.08472926722005049, + "learning_rate": 9.975090282943907e-05, + "loss": 0.0307, + "step": 29800 + }, + { + "epoch": 0.095424, + "grad_norm": 0.0985938092892877, + "learning_rate": 9.975056863698935e-05, + "loss": 0.0326, + "step": 29820 + }, + { + "epoch": 0.095488, + "grad_norm": 0.07954722469561218, + "learning_rate": 9.975023422107171e-05, + "loss": 0.0335, + "step": 29840 + }, + { + "epoch": 0.095552, + "grad_norm": 0.16150195929582079, + "learning_rate": 9.974989958168761e-05, + "loss": 0.0324, + "step": 29860 + }, + { + "epoch": 0.095616, + "grad_norm": 0.07359550805670287, + "learning_rate": 9.974956471883858e-05, + "loss": 0.0338, + "step": 29880 + }, + { + "epoch": 0.09568, + "grad_norm": 0.07277432181346337, + "learning_rate": 9.974922963252613e-05, + "loss": 0.0333, + "step": 29900 + }, + { + "epoch": 0.095744, + "grad_norm": 0.09497379187279864, + "learning_rate": 9.974889432275174e-05, + "loss": 0.0308, + "step": 29920 + }, + { + "epoch": 0.095808, + "grad_norm": 0.1016216923374487, + "learning_rate": 9.974855878951692e-05, + "loss": 0.0314, + "step": 29940 + }, + { + "epoch": 0.095872, + "grad_norm": 0.08006787934574093, + "learning_rate": 9.974822303282319e-05, + "loss": 0.033, + "step": 29960 + }, + { + "epoch": 0.095936, + "grad_norm": 0.09841682277856208, + "learning_rate": 9.974788705267204e-05, + "loss": 0.03, + "step": 29980 + }, + { + "epoch": 0.096, + "grad_norm": 0.1299511691686431, + "learning_rate": 9.974755084906502e-05, + "loss": 0.0362, + "step": 30000 + }, + { + "epoch": 0.096064, + "grad_norm": 0.11063175251849341, + "learning_rate": 9.974721442200358e-05, + "loss": 0.035, + "step": 30020 + }, + { + "epoch": 0.096128, + "grad_norm": 0.12220240545419464, + "learning_rate": 9.974687777148927e-05, + "loss": 0.0366, + "step": 30040 + }, + { + "epoch": 0.096192, + "grad_norm": 0.14505486923419622, + "learning_rate": 9.974654089752362e-05, + "loss": 0.0391, + "step": 30060 + }, + { + "epoch": 0.096256, + "grad_norm": 0.1408034557683459, + "learning_rate": 9.97462038001081e-05, + "loss": 0.0327, + "step": 30080 + }, + { + "epoch": 0.09632, + "grad_norm": 0.06926726796572201, + "learning_rate": 9.974586647924425e-05, + "loss": 0.0325, + "step": 30100 + }, + { + "epoch": 0.096384, + "grad_norm": 0.09017995341866855, + "learning_rate": 9.974552893493356e-05, + "loss": 0.0328, + "step": 30120 + }, + { + "epoch": 0.096448, + "grad_norm": 0.1119004076803094, + "learning_rate": 9.974519116717756e-05, + "loss": 0.0297, + "step": 30140 + }, + { + "epoch": 0.096512, + "grad_norm": 0.11721323960858096, + "learning_rate": 9.97448531759778e-05, + "loss": 0.0321, + "step": 30160 + }, + { + "epoch": 0.096576, + "grad_norm": 0.10441897356059847, + "learning_rate": 9.974451496133574e-05, + "loss": 0.0406, + "step": 30180 + }, + { + "epoch": 0.09664, + "grad_norm": 0.10254723092542828, + "learning_rate": 9.974417652325294e-05, + "loss": 0.0332, + "step": 30200 + }, + { + "epoch": 0.096704, + "grad_norm": 0.09811760151465074, + "learning_rate": 9.974383786173092e-05, + "loss": 0.0308, + "step": 30220 + }, + { + "epoch": 0.096768, + "grad_norm": 0.07097362160094467, + "learning_rate": 9.974349897677117e-05, + "loss": 0.0302, + "step": 30240 + }, + { + "epoch": 0.096832, + "grad_norm": 0.07722830033705815, + "learning_rate": 9.974315986837524e-05, + "loss": 0.0343, + "step": 30260 + }, + { + "epoch": 0.096896, + "grad_norm": 0.12609632338274587, + "learning_rate": 9.974282053654464e-05, + "loss": 0.0351, + "step": 30280 + }, + { + "epoch": 0.09696, + "grad_norm": 0.09784417036985352, + "learning_rate": 9.97424809812809e-05, + "loss": 0.0343, + "step": 30300 + }, + { + "epoch": 0.097024, + "grad_norm": 0.09960117309438471, + "learning_rate": 9.974214120258555e-05, + "loss": 0.0339, + "step": 30320 + }, + { + "epoch": 0.097088, + "grad_norm": 0.08629327143705236, + "learning_rate": 9.974180120046011e-05, + "loss": 0.0336, + "step": 30340 + }, + { + "epoch": 0.097152, + "grad_norm": 0.0752869721345672, + "learning_rate": 9.97414609749061e-05, + "loss": 0.0324, + "step": 30360 + }, + { + "epoch": 0.097216, + "grad_norm": 0.08101292556319897, + "learning_rate": 9.974112052592506e-05, + "loss": 0.0342, + "step": 30380 + }, + { + "epoch": 0.09728, + "grad_norm": 0.08173805715684286, + "learning_rate": 9.974077985351852e-05, + "loss": 0.0344, + "step": 30400 + }, + { + "epoch": 0.097344, + "grad_norm": 0.08606590612359086, + "learning_rate": 9.974043895768799e-05, + "loss": 0.033, + "step": 30420 + }, + { + "epoch": 0.097408, + "grad_norm": 0.06523703658176408, + "learning_rate": 9.974009783843503e-05, + "loss": 0.0345, + "step": 30440 + }, + { + "epoch": 0.097472, + "grad_norm": 0.12499640988195328, + "learning_rate": 9.973975649576116e-05, + "loss": 0.0298, + "step": 30460 + }, + { + "epoch": 0.097536, + "grad_norm": 0.09146758132871383, + "learning_rate": 9.973941492966791e-05, + "loss": 0.033, + "step": 30480 + }, + { + "epoch": 0.0976, + "grad_norm": 0.16156016613907717, + "learning_rate": 9.973907314015682e-05, + "loss": 0.0342, + "step": 30500 + }, + { + "epoch": 0.097664, + "grad_norm": 0.08884177256895638, + "learning_rate": 9.97387311272294e-05, + "loss": 0.0313, + "step": 30520 + }, + { + "epoch": 0.097728, + "grad_norm": 0.13995932930251273, + "learning_rate": 9.973838889088723e-05, + "loss": 0.034, + "step": 30540 + }, + { + "epoch": 0.097792, + "grad_norm": 0.10005411636876668, + "learning_rate": 9.973804643113182e-05, + "loss": 0.0357, + "step": 30560 + }, + { + "epoch": 0.097856, + "grad_norm": 0.08642501294292054, + "learning_rate": 9.973770374796471e-05, + "loss": 0.033, + "step": 30580 + }, + { + "epoch": 0.09792, + "grad_norm": 0.12865147405745314, + "learning_rate": 9.973736084138743e-05, + "loss": 0.0335, + "step": 30600 + }, + { + "epoch": 0.097984, + "grad_norm": 0.06826955715166223, + "learning_rate": 9.973701771140156e-05, + "loss": 0.0376, + "step": 30620 + }, + { + "epoch": 0.098048, + "grad_norm": 0.09868598194123517, + "learning_rate": 9.97366743580086e-05, + "loss": 0.036, + "step": 30640 + }, + { + "epoch": 0.098112, + "grad_norm": 0.08439344263422871, + "learning_rate": 9.97363307812101e-05, + "loss": 0.0337, + "step": 30660 + }, + { + "epoch": 0.098176, + "grad_norm": 0.07391584859175118, + "learning_rate": 9.973598698100762e-05, + "loss": 0.0306, + "step": 30680 + }, + { + "epoch": 0.09824, + "grad_norm": 0.09012230273821684, + "learning_rate": 9.973564295740268e-05, + "loss": 0.0332, + "step": 30700 + }, + { + "epoch": 0.098304, + "grad_norm": 0.12631475678175286, + "learning_rate": 9.973529871039686e-05, + "loss": 0.0294, + "step": 30720 + }, + { + "epoch": 0.098368, + "grad_norm": 0.10932120299055234, + "learning_rate": 9.973495423999166e-05, + "loss": 0.0317, + "step": 30740 + }, + { + "epoch": 0.098432, + "grad_norm": 0.11454530623070215, + "learning_rate": 9.973460954618867e-05, + "loss": 0.033, + "step": 30760 + }, + { + "epoch": 0.098496, + "grad_norm": 0.06902786828537001, + "learning_rate": 9.973426462898942e-05, + "loss": 0.0302, + "step": 30780 + }, + { + "epoch": 0.09856, + "grad_norm": 0.10871924397887624, + "learning_rate": 9.973391948839546e-05, + "loss": 0.0332, + "step": 30800 + }, + { + "epoch": 0.098624, + "grad_norm": 0.07398374579414338, + "learning_rate": 9.973357412440834e-05, + "loss": 0.0333, + "step": 30820 + }, + { + "epoch": 0.098688, + "grad_norm": 0.09719029350221942, + "learning_rate": 9.97332285370296e-05, + "loss": 0.0325, + "step": 30840 + }, + { + "epoch": 0.098752, + "grad_norm": 0.14594596654596734, + "learning_rate": 9.973288272626082e-05, + "loss": 0.0338, + "step": 30860 + }, + { + "epoch": 0.098816, + "grad_norm": 0.09268680677967088, + "learning_rate": 9.973253669210353e-05, + "loss": 0.0325, + "step": 30880 + }, + { + "epoch": 0.09888, + "grad_norm": 0.09292739706459761, + "learning_rate": 9.973219043455928e-05, + "loss": 0.0352, + "step": 30900 + }, + { + "epoch": 0.098944, + "grad_norm": 0.12021937902026852, + "learning_rate": 9.973184395362966e-05, + "loss": 0.0321, + "step": 30920 + }, + { + "epoch": 0.099008, + "grad_norm": 0.09814273005889969, + "learning_rate": 9.973149724931618e-05, + "loss": 0.0319, + "step": 30940 + }, + { + "epoch": 0.099072, + "grad_norm": 0.09078729836935798, + "learning_rate": 9.973115032162044e-05, + "loss": 0.0329, + "step": 30960 + }, + { + "epoch": 0.099136, + "grad_norm": 0.10244850290445295, + "learning_rate": 9.973080317054397e-05, + "loss": 0.0309, + "step": 30980 + }, + { + "epoch": 0.0992, + "grad_norm": 0.08346334696045318, + "learning_rate": 9.973045579608834e-05, + "loss": 0.0331, + "step": 31000 + }, + { + "epoch": 0.099264, + "grad_norm": 0.18701047144730432, + "learning_rate": 9.97301081982551e-05, + "loss": 0.0326, + "step": 31020 + }, + { + "epoch": 0.099328, + "grad_norm": 0.1201759892128287, + "learning_rate": 9.972976037704584e-05, + "loss": 0.0359, + "step": 31040 + }, + { + "epoch": 0.099392, + "grad_norm": 0.10557937356352676, + "learning_rate": 9.972941233246209e-05, + "loss": 0.0393, + "step": 31060 + }, + { + "epoch": 0.099456, + "grad_norm": 0.1743182975842921, + "learning_rate": 9.972906406450542e-05, + "loss": 0.0401, + "step": 31080 + }, + { + "epoch": 0.09952, + "grad_norm": 0.16651371568548004, + "learning_rate": 9.97287155731774e-05, + "loss": 0.032, + "step": 31100 + }, + { + "epoch": 0.099584, + "grad_norm": 0.1029534297549954, + "learning_rate": 9.972836685847961e-05, + "loss": 0.0335, + "step": 31120 + }, + { + "epoch": 0.099648, + "grad_norm": 0.09420021180076553, + "learning_rate": 9.972801792041358e-05, + "loss": 0.0345, + "step": 31140 + }, + { + "epoch": 0.099712, + "grad_norm": 0.29501769793216276, + "learning_rate": 9.972766875898092e-05, + "loss": 0.036, + "step": 31160 + }, + { + "epoch": 0.099776, + "grad_norm": 0.11711789522606665, + "learning_rate": 9.972731937418316e-05, + "loss": 0.0341, + "step": 31180 + }, + { + "epoch": 0.09984, + "grad_norm": 0.062420324862498024, + "learning_rate": 9.97269697660219e-05, + "loss": 0.0298, + "step": 31200 + }, + { + "epoch": 0.099904, + "grad_norm": 0.12863230721467506, + "learning_rate": 9.972661993449869e-05, + "loss": 0.034, + "step": 31220 + }, + { + "epoch": 0.099968, + "grad_norm": 0.10616451124060705, + "learning_rate": 9.97262698796151e-05, + "loss": 0.0337, + "step": 31240 + }, + { + "epoch": 0.100032, + "grad_norm": 0.13852034714616301, + "learning_rate": 9.972591960137272e-05, + "loss": 0.0344, + "step": 31260 + }, + { + "epoch": 0.100096, + "grad_norm": 0.10257609212541682, + "learning_rate": 9.972556909977314e-05, + "loss": 0.0338, + "step": 31280 + }, + { + "epoch": 0.10016, + "grad_norm": 0.07845978051101761, + "learning_rate": 9.972521837481786e-05, + "loss": 0.0296, + "step": 31300 + }, + { + "epoch": 0.100224, + "grad_norm": 0.1295534128482442, + "learning_rate": 9.972486742650853e-05, + "loss": 0.0329, + "step": 31320 + }, + { + "epoch": 0.100288, + "grad_norm": 0.17114741558311236, + "learning_rate": 9.97245162548467e-05, + "loss": 0.0281, + "step": 31340 + }, + { + "epoch": 0.100352, + "grad_norm": 0.15604593096718689, + "learning_rate": 9.972416485983394e-05, + "loss": 0.0315, + "step": 31360 + }, + { + "epoch": 0.100416, + "grad_norm": 0.11453687445565204, + "learning_rate": 9.972381324147184e-05, + "loss": 0.0333, + "step": 31380 + }, + { + "epoch": 0.10048, + "grad_norm": 0.19813852631502296, + "learning_rate": 9.972346139976198e-05, + "loss": 0.0321, + "step": 31400 + }, + { + "epoch": 0.100544, + "grad_norm": 0.07619159093953906, + "learning_rate": 9.972310933470593e-05, + "loss": 0.0319, + "step": 31420 + }, + { + "epoch": 0.100608, + "grad_norm": 0.11120032260155135, + "learning_rate": 9.972275704630527e-05, + "loss": 0.0318, + "step": 31440 + }, + { + "epoch": 0.100672, + "grad_norm": 0.07653064989258833, + "learning_rate": 9.97224045345616e-05, + "loss": 0.0316, + "step": 31460 + }, + { + "epoch": 0.100736, + "grad_norm": 0.20396186435101749, + "learning_rate": 9.972205179947648e-05, + "loss": 0.0306, + "step": 31480 + }, + { + "epoch": 0.1008, + "grad_norm": 0.14984403938835592, + "learning_rate": 9.972169884105153e-05, + "loss": 0.0352, + "step": 31500 + }, + { + "epoch": 0.100864, + "grad_norm": 0.20142167039486347, + "learning_rate": 9.97213456592883e-05, + "loss": 0.0332, + "step": 31520 + }, + { + "epoch": 0.100928, + "grad_norm": 0.07890510145428499, + "learning_rate": 9.972099225418839e-05, + "loss": 0.0362, + "step": 31540 + }, + { + "epoch": 0.100992, + "grad_norm": 0.10478354131211647, + "learning_rate": 9.972063862575338e-05, + "loss": 0.0337, + "step": 31560 + }, + { + "epoch": 0.101056, + "grad_norm": 0.1295243565923376, + "learning_rate": 9.972028477398487e-05, + "loss": 0.0364, + "step": 31580 + }, + { + "epoch": 0.10112, + "grad_norm": 0.08742814949909723, + "learning_rate": 9.971993069888445e-05, + "loss": 0.0323, + "step": 31600 + }, + { + "epoch": 0.101184, + "grad_norm": 0.1428753093140812, + "learning_rate": 9.97195764004537e-05, + "loss": 0.0323, + "step": 31620 + }, + { + "epoch": 0.101248, + "grad_norm": 0.09135146206237223, + "learning_rate": 9.971922187869422e-05, + "loss": 0.0306, + "step": 31640 + }, + { + "epoch": 0.101312, + "grad_norm": 0.07758556572569282, + "learning_rate": 9.971886713360759e-05, + "loss": 0.0334, + "step": 31660 + }, + { + "epoch": 0.101376, + "grad_norm": 0.10087444494625634, + "learning_rate": 9.971851216519541e-05, + "loss": 0.0331, + "step": 31680 + }, + { + "epoch": 0.10144, + "grad_norm": 0.10065682086312257, + "learning_rate": 9.971815697345929e-05, + "loss": 0.0341, + "step": 31700 + }, + { + "epoch": 0.101504, + "grad_norm": 0.07931571418665577, + "learning_rate": 9.97178015584008e-05, + "loss": 0.0339, + "step": 31720 + }, + { + "epoch": 0.101568, + "grad_norm": 0.10192901342586078, + "learning_rate": 9.971744592002155e-05, + "loss": 0.0344, + "step": 31740 + }, + { + "epoch": 0.101632, + "grad_norm": 0.1485328699566904, + "learning_rate": 9.971709005832315e-05, + "loss": 0.0337, + "step": 31760 + }, + { + "epoch": 0.101696, + "grad_norm": 0.08277557937086895, + "learning_rate": 9.971673397330716e-05, + "loss": 0.036, + "step": 31780 + }, + { + "epoch": 0.10176, + "grad_norm": 0.1114683699208906, + "learning_rate": 9.971637766497524e-05, + "loss": 0.0325, + "step": 31800 + }, + { + "epoch": 0.101824, + "grad_norm": 0.14250413647309068, + "learning_rate": 9.971602113332892e-05, + "loss": 0.0305, + "step": 31820 + }, + { + "epoch": 0.101888, + "grad_norm": 0.17243724950698003, + "learning_rate": 9.971566437836984e-05, + "loss": 0.0338, + "step": 31840 + }, + { + "epoch": 0.101952, + "grad_norm": 0.15670319579651057, + "learning_rate": 9.971530740009961e-05, + "loss": 0.0344, + "step": 31860 + }, + { + "epoch": 0.102016, + "grad_norm": 0.0734786091536894, + "learning_rate": 9.971495019851982e-05, + "loss": 0.0368, + "step": 31880 + }, + { + "epoch": 0.10208, + "grad_norm": 0.08802697101686133, + "learning_rate": 9.971459277363206e-05, + "loss": 0.0306, + "step": 31900 + }, + { + "epoch": 0.102144, + "grad_norm": 0.09228599436442161, + "learning_rate": 9.971423512543797e-05, + "loss": 0.0373, + "step": 31920 + }, + { + "epoch": 0.102208, + "grad_norm": 0.12508734385439804, + "learning_rate": 9.971387725393914e-05, + "loss": 0.0361, + "step": 31940 + }, + { + "epoch": 0.102272, + "grad_norm": 0.08632683525165707, + "learning_rate": 9.971351915913716e-05, + "loss": 0.0353, + "step": 31960 + }, + { + "epoch": 0.102336, + "grad_norm": 0.16498754950621342, + "learning_rate": 9.971316084103366e-05, + "loss": 0.0326, + "step": 31980 + }, + { + "epoch": 0.1024, + "grad_norm": 0.06996310024694859, + "learning_rate": 9.971280229963025e-05, + "loss": 0.0286, + "step": 32000 + }, + { + "epoch": 0.102464, + "grad_norm": 0.1959545669771389, + "learning_rate": 9.971244353492852e-05, + "loss": 0.028, + "step": 32020 + }, + { + "epoch": 0.102528, + "grad_norm": 0.09325637838238043, + "learning_rate": 9.971208454693011e-05, + "loss": 0.0339, + "step": 32040 + }, + { + "epoch": 0.102592, + "grad_norm": 0.11300540522174304, + "learning_rate": 9.971172533563661e-05, + "loss": 0.0322, + "step": 32060 + }, + { + "epoch": 0.102656, + "grad_norm": 0.0824585738862516, + "learning_rate": 9.971136590104965e-05, + "loss": 0.0342, + "step": 32080 + }, + { + "epoch": 0.10272, + "grad_norm": 0.0827516492888346, + "learning_rate": 9.971100624317082e-05, + "loss": 0.0355, + "step": 32100 + }, + { + "epoch": 0.102784, + "grad_norm": 0.10308376729648609, + "learning_rate": 9.971064636200176e-05, + "loss": 0.0303, + "step": 32120 + }, + { + "epoch": 0.102848, + "grad_norm": 0.09506547669565192, + "learning_rate": 9.971028625754408e-05, + "loss": 0.0294, + "step": 32140 + }, + { + "epoch": 0.102912, + "grad_norm": 0.09101885157069524, + "learning_rate": 9.97099259297994e-05, + "loss": 0.0322, + "step": 32160 + }, + { + "epoch": 0.102976, + "grad_norm": 0.06872243741846114, + "learning_rate": 9.970956537876932e-05, + "loss": 0.0296, + "step": 32180 + }, + { + "epoch": 0.10304, + "grad_norm": 0.1887249575493209, + "learning_rate": 9.970920460445549e-05, + "loss": 0.0297, + "step": 32200 + }, + { + "epoch": 0.103104, + "grad_norm": 0.1839284389675204, + "learning_rate": 9.97088436068595e-05, + "loss": 0.0323, + "step": 32220 + }, + { + "epoch": 0.103168, + "grad_norm": 0.10291338900401395, + "learning_rate": 9.970848238598298e-05, + "loss": 0.0334, + "step": 32240 + }, + { + "epoch": 0.103232, + "grad_norm": 0.14004646147309438, + "learning_rate": 9.970812094182757e-05, + "loss": 0.0361, + "step": 32260 + }, + { + "epoch": 0.103296, + "grad_norm": 0.1030451857376463, + "learning_rate": 9.970775927439489e-05, + "loss": 0.0327, + "step": 32280 + }, + { + "epoch": 0.10336, + "grad_norm": 0.07923788655883904, + "learning_rate": 9.970739738368654e-05, + "loss": 0.0342, + "step": 32300 + }, + { + "epoch": 0.103424, + "grad_norm": 0.08988754999390206, + "learning_rate": 9.970703526970418e-05, + "loss": 0.0328, + "step": 32320 + }, + { + "epoch": 0.103488, + "grad_norm": 0.08908146425473301, + "learning_rate": 9.97066729324494e-05, + "loss": 0.0316, + "step": 32340 + }, + { + "epoch": 0.103552, + "grad_norm": 0.08820222455230466, + "learning_rate": 9.970631037192384e-05, + "loss": 0.0348, + "step": 32360 + }, + { + "epoch": 0.103616, + "grad_norm": 0.11135945319607804, + "learning_rate": 9.970594758812915e-05, + "loss": 0.0321, + "step": 32380 + }, + { + "epoch": 0.10368, + "grad_norm": 0.07987088362927129, + "learning_rate": 9.970558458106694e-05, + "loss": 0.034, + "step": 32400 + }, + { + "epoch": 0.103744, + "grad_norm": 0.13966720658590714, + "learning_rate": 9.970522135073883e-05, + "loss": 0.0332, + "step": 32420 + }, + { + "epoch": 0.103808, + "grad_norm": 0.1688180175651959, + "learning_rate": 9.970485789714649e-05, + "loss": 0.0366, + "step": 32440 + }, + { + "epoch": 0.103872, + "grad_norm": 0.08091115244919901, + "learning_rate": 9.970449422029152e-05, + "loss": 0.0321, + "step": 32460 + }, + { + "epoch": 0.103936, + "grad_norm": 0.15209187212438552, + "learning_rate": 9.970413032017554e-05, + "loss": 0.0301, + "step": 32480 + }, + { + "epoch": 0.104, + "grad_norm": 0.11715534695341157, + "learning_rate": 9.970376619680024e-05, + "loss": 0.0299, + "step": 32500 + }, + { + "epoch": 0.104064, + "grad_norm": 0.08962885447167879, + "learning_rate": 9.97034018501672e-05, + "loss": 0.033, + "step": 32520 + }, + { + "epoch": 0.104128, + "grad_norm": 0.1755333415441921, + "learning_rate": 9.970303728027807e-05, + "loss": 0.0373, + "step": 32540 + }, + { + "epoch": 0.104192, + "grad_norm": 0.10250655731561499, + "learning_rate": 9.970267248713452e-05, + "loss": 0.0354, + "step": 32560 + }, + { + "epoch": 0.104256, + "grad_norm": 0.10846775825182003, + "learning_rate": 9.970230747073816e-05, + "loss": 0.0321, + "step": 32580 + }, + { + "epoch": 0.10432, + "grad_norm": 0.08995467294873673, + "learning_rate": 9.970194223109062e-05, + "loss": 0.0349, + "step": 32600 + }, + { + "epoch": 0.104384, + "grad_norm": 0.062579813936164, + "learning_rate": 9.970157676819359e-05, + "loss": 0.029, + "step": 32620 + }, + { + "epoch": 0.104448, + "grad_norm": 0.12053434188012464, + "learning_rate": 9.970121108204864e-05, + "loss": 0.035, + "step": 32640 + }, + { + "epoch": 0.104512, + "grad_norm": 0.1887429841241576, + "learning_rate": 9.970084517265746e-05, + "loss": 0.0337, + "step": 32660 + }, + { + "epoch": 0.104576, + "grad_norm": 0.12505366206441224, + "learning_rate": 9.970047904002168e-05, + "loss": 0.0329, + "step": 32680 + }, + { + "epoch": 0.10464, + "grad_norm": 0.07293221784420005, + "learning_rate": 9.970011268414294e-05, + "loss": 0.0305, + "step": 32700 + }, + { + "epoch": 0.104704, + "grad_norm": 0.1056085582227811, + "learning_rate": 9.96997461050229e-05, + "loss": 0.0295, + "step": 32720 + }, + { + "epoch": 0.104768, + "grad_norm": 0.07054947819850357, + "learning_rate": 9.969937930266321e-05, + "loss": 0.0322, + "step": 32740 + }, + { + "epoch": 0.104832, + "grad_norm": 0.14336346691534652, + "learning_rate": 9.96990122770655e-05, + "loss": 0.0331, + "step": 32760 + }, + { + "epoch": 0.104896, + "grad_norm": 0.08325060175072148, + "learning_rate": 9.969864502823142e-05, + "loss": 0.032, + "step": 32780 + }, + { + "epoch": 0.10496, + "grad_norm": 0.09360226215051706, + "learning_rate": 9.969827755616263e-05, + "loss": 0.0332, + "step": 32800 + }, + { + "epoch": 0.105024, + "grad_norm": 0.11962650958228059, + "learning_rate": 9.969790986086077e-05, + "loss": 0.0351, + "step": 32820 + }, + { + "epoch": 0.105088, + "grad_norm": 0.08627066880413202, + "learning_rate": 9.96975419423275e-05, + "loss": 0.0333, + "step": 32840 + }, + { + "epoch": 0.105152, + "grad_norm": 0.08027216234572701, + "learning_rate": 9.969717380056447e-05, + "loss": 0.034, + "step": 32860 + }, + { + "epoch": 0.105216, + "grad_norm": 0.08070882854077836, + "learning_rate": 9.969680543557334e-05, + "loss": 0.0328, + "step": 32880 + }, + { + "epoch": 0.10528, + "grad_norm": 0.10600425820625187, + "learning_rate": 9.969643684735575e-05, + "loss": 0.0323, + "step": 32900 + }, + { + "epoch": 0.105344, + "grad_norm": 0.16128923418380592, + "learning_rate": 9.969606803591337e-05, + "loss": 0.0327, + "step": 32920 + }, + { + "epoch": 0.105408, + "grad_norm": 0.07836433199847413, + "learning_rate": 9.969569900124785e-05, + "loss": 0.0333, + "step": 32940 + }, + { + "epoch": 0.105472, + "grad_norm": 0.10380878829516048, + "learning_rate": 9.969532974336084e-05, + "loss": 0.0279, + "step": 32960 + }, + { + "epoch": 0.105536, + "grad_norm": 0.12058564455509305, + "learning_rate": 9.969496026225402e-05, + "loss": 0.0359, + "step": 32980 + }, + { + "epoch": 0.1056, + "grad_norm": 0.10336570306234161, + "learning_rate": 9.969459055792903e-05, + "loss": 0.032, + "step": 33000 + }, + { + "epoch": 0.105664, + "grad_norm": 0.12713276969389362, + "learning_rate": 9.969422063038752e-05, + "loss": 0.0358, + "step": 33020 + }, + { + "epoch": 0.105728, + "grad_norm": 0.08507008241790076, + "learning_rate": 9.969385047963117e-05, + "loss": 0.0345, + "step": 33040 + }, + { + "epoch": 0.105792, + "grad_norm": 0.11181059018180205, + "learning_rate": 9.969348010566166e-05, + "loss": 0.0315, + "step": 33060 + }, + { + "epoch": 0.105856, + "grad_norm": 0.0773846683008081, + "learning_rate": 9.969310950848063e-05, + "loss": 0.0352, + "step": 33080 + }, + { + "epoch": 0.10592, + "grad_norm": 0.1481978459435784, + "learning_rate": 9.969273868808975e-05, + "loss": 0.0294, + "step": 33100 + }, + { + "epoch": 0.105984, + "grad_norm": 0.13525502805229458, + "learning_rate": 9.969236764449068e-05, + "loss": 0.0297, + "step": 33120 + }, + { + "epoch": 0.106048, + "grad_norm": 0.06224866886131812, + "learning_rate": 9.969199637768509e-05, + "loss": 0.0324, + "step": 33140 + }, + { + "epoch": 0.106112, + "grad_norm": 0.07613885953979165, + "learning_rate": 9.969162488767466e-05, + "loss": 0.0355, + "step": 33160 + }, + { + "epoch": 0.106176, + "grad_norm": 0.1726782339167767, + "learning_rate": 9.969125317446104e-05, + "loss": 0.0356, + "step": 33180 + }, + { + "epoch": 0.10624, + "grad_norm": 0.11051892150508513, + "learning_rate": 9.96908812380459e-05, + "loss": 0.0326, + "step": 33200 + }, + { + "epoch": 0.106304, + "grad_norm": 0.10447556693094644, + "learning_rate": 9.969050907843094e-05, + "loss": 0.0319, + "step": 33220 + }, + { + "epoch": 0.106368, + "grad_norm": 0.06690868792437975, + "learning_rate": 9.96901366956178e-05, + "loss": 0.0329, + "step": 33240 + }, + { + "epoch": 0.106432, + "grad_norm": 0.07707526650663946, + "learning_rate": 9.968976408960815e-05, + "loss": 0.0307, + "step": 33260 + }, + { + "epoch": 0.106496, + "grad_norm": 0.11737144356253157, + "learning_rate": 9.968939126040369e-05, + "loss": 0.0304, + "step": 33280 + }, + { + "epoch": 0.10656, + "grad_norm": 0.05806661669138139, + "learning_rate": 9.968901820800607e-05, + "loss": 0.0336, + "step": 33300 + }, + { + "epoch": 0.106624, + "grad_norm": 0.10013813933793975, + "learning_rate": 9.968864493241699e-05, + "loss": 0.0303, + "step": 33320 + }, + { + "epoch": 0.106688, + "grad_norm": 0.14041833909154006, + "learning_rate": 9.96882714336381e-05, + "loss": 0.0292, + "step": 33340 + }, + { + "epoch": 0.106752, + "grad_norm": 0.08283870080160281, + "learning_rate": 9.96878977116711e-05, + "loss": 0.0294, + "step": 33360 + }, + { + "epoch": 0.106816, + "grad_norm": 0.08957349081639356, + "learning_rate": 9.968752376651765e-05, + "loss": 0.033, + "step": 33380 + }, + { + "epoch": 0.10688, + "grad_norm": 0.08710406866776232, + "learning_rate": 9.968714959817943e-05, + "loss": 0.0324, + "step": 33400 + }, + { + "epoch": 0.106944, + "grad_norm": 0.16171824124635029, + "learning_rate": 9.968677520665816e-05, + "loss": 0.0302, + "step": 33420 + }, + { + "epoch": 0.107008, + "grad_norm": 0.24924451076587933, + "learning_rate": 9.968640059195548e-05, + "loss": 0.0314, + "step": 33440 + }, + { + "epoch": 0.107072, + "grad_norm": 0.07017396582488891, + "learning_rate": 9.968602575407308e-05, + "loss": 0.0325, + "step": 33460 + }, + { + "epoch": 0.107136, + "grad_norm": 0.09703947429741804, + "learning_rate": 9.968565069301265e-05, + "loss": 0.0339, + "step": 33480 + }, + { + "epoch": 0.1072, + "grad_norm": 0.10465812030668217, + "learning_rate": 9.968527540877585e-05, + "loss": 0.0331, + "step": 33500 + }, + { + "epoch": 0.107264, + "grad_norm": 0.07203297551131219, + "learning_rate": 9.968489990136442e-05, + "loss": 0.0316, + "step": 33520 + }, + { + "epoch": 0.107328, + "grad_norm": 0.09667343105003309, + "learning_rate": 9.968452417078e-05, + "loss": 0.0339, + "step": 33540 + }, + { + "epoch": 0.107392, + "grad_norm": 0.1720302227945823, + "learning_rate": 9.96841482170243e-05, + "loss": 0.0316, + "step": 33560 + }, + { + "epoch": 0.107456, + "grad_norm": 0.10641947252003983, + "learning_rate": 9.968377204009899e-05, + "loss": 0.0313, + "step": 33580 + }, + { + "epoch": 0.10752, + "grad_norm": 0.09794552682883996, + "learning_rate": 9.96833956400058e-05, + "loss": 0.0318, + "step": 33600 + }, + { + "epoch": 0.107584, + "grad_norm": 0.07959330488946803, + "learning_rate": 9.968301901674636e-05, + "loss": 0.0312, + "step": 33620 + }, + { + "epoch": 0.107648, + "grad_norm": 0.07693175134815734, + "learning_rate": 9.968264217032241e-05, + "loss": 0.034, + "step": 33640 + }, + { + "epoch": 0.107712, + "grad_norm": 0.0929141559379662, + "learning_rate": 9.968226510073562e-05, + "loss": 0.0315, + "step": 33660 + }, + { + "epoch": 0.107776, + "grad_norm": 0.09129441259114457, + "learning_rate": 9.968188780798769e-05, + "loss": 0.0319, + "step": 33680 + }, + { + "epoch": 0.10784, + "grad_norm": 0.09072830284950234, + "learning_rate": 9.96815102920803e-05, + "loss": 0.0338, + "step": 33700 + }, + { + "epoch": 0.107904, + "grad_norm": 0.07342049888076141, + "learning_rate": 9.968113255301519e-05, + "loss": 0.0326, + "step": 33720 + }, + { + "epoch": 0.107968, + "grad_norm": 0.07409497555249488, + "learning_rate": 9.9680754590794e-05, + "loss": 0.0298, + "step": 33740 + }, + { + "epoch": 0.108032, + "grad_norm": 0.0978328077018129, + "learning_rate": 9.968037640541848e-05, + "loss": 0.0262, + "step": 33760 + }, + { + "epoch": 0.108096, + "grad_norm": 0.11128091416467521, + "learning_rate": 9.967999799689027e-05, + "loss": 0.0312, + "step": 33780 + }, + { + "epoch": 0.10816, + "grad_norm": 0.10022385382023519, + "learning_rate": 9.967961936521113e-05, + "loss": 0.0327, + "step": 33800 + }, + { + "epoch": 0.108224, + "grad_norm": 0.07302531818086366, + "learning_rate": 9.967924051038273e-05, + "loss": 0.0332, + "step": 33820 + }, + { + "epoch": 0.108288, + "grad_norm": 0.12851180443484947, + "learning_rate": 9.967886143240676e-05, + "loss": 0.0292, + "step": 33840 + }, + { + "epoch": 0.108352, + "grad_norm": 0.07882324357049737, + "learning_rate": 9.967848213128497e-05, + "loss": 0.0349, + "step": 33860 + }, + { + "epoch": 0.108416, + "grad_norm": 0.13464494201301158, + "learning_rate": 9.9678102607019e-05, + "loss": 0.0338, + "step": 33880 + }, + { + "epoch": 0.10848, + "grad_norm": 0.12774870408321368, + "learning_rate": 9.967772285961062e-05, + "loss": 0.031, + "step": 33900 + }, + { + "epoch": 0.108544, + "grad_norm": 0.09598308730747136, + "learning_rate": 9.967734288906147e-05, + "loss": 0.0344, + "step": 33920 + }, + { + "epoch": 0.108608, + "grad_norm": 0.12656168429502637, + "learning_rate": 9.96769626953733e-05, + "loss": 0.0371, + "step": 33940 + }, + { + "epoch": 0.108672, + "grad_norm": 0.130946236455287, + "learning_rate": 9.96765822785478e-05, + "loss": 0.0344, + "step": 33960 + }, + { + "epoch": 0.108736, + "grad_norm": 0.12366195158099676, + "learning_rate": 9.967620163858671e-05, + "loss": 0.0318, + "step": 33980 + }, + { + "epoch": 0.1088, + "grad_norm": 0.07279125101927159, + "learning_rate": 9.967582077549169e-05, + "loss": 0.0286, + "step": 34000 + }, + { + "epoch": 0.108864, + "grad_norm": 0.1003430056347591, + "learning_rate": 9.967543968926449e-05, + "loss": 0.0317, + "step": 34020 + }, + { + "epoch": 0.108928, + "grad_norm": 0.09260375306449617, + "learning_rate": 9.967505837990681e-05, + "loss": 0.0347, + "step": 34040 + }, + { + "epoch": 0.108992, + "grad_norm": 0.0923801076434255, + "learning_rate": 9.967467684742035e-05, + "loss": 0.0325, + "step": 34060 + }, + { + "epoch": 0.109056, + "grad_norm": 0.07379210551677713, + "learning_rate": 9.967429509180683e-05, + "loss": 0.0336, + "step": 34080 + }, + { + "epoch": 0.10912, + "grad_norm": 0.09223495729107571, + "learning_rate": 9.967391311306798e-05, + "loss": 0.034, + "step": 34100 + }, + { + "epoch": 0.109184, + "grad_norm": 0.10718609943583018, + "learning_rate": 9.967353091120551e-05, + "loss": 0.0366, + "step": 34120 + }, + { + "epoch": 0.109248, + "grad_norm": 0.09348893997420958, + "learning_rate": 9.967314848622111e-05, + "loss": 0.0325, + "step": 34140 + }, + { + "epoch": 0.109312, + "grad_norm": 0.08353324307011697, + "learning_rate": 9.967276583811655e-05, + "loss": 0.0352, + "step": 34160 + }, + { + "epoch": 0.109376, + "grad_norm": 0.1119537263468282, + "learning_rate": 9.96723829668935e-05, + "loss": 0.033, + "step": 34180 + }, + { + "epoch": 0.10944, + "grad_norm": 0.12949264012208023, + "learning_rate": 9.967199987255369e-05, + "loss": 0.0337, + "step": 34200 + }, + { + "epoch": 0.109504, + "grad_norm": 0.09209263345116006, + "learning_rate": 9.967161655509887e-05, + "loss": 0.0355, + "step": 34220 + }, + { + "epoch": 0.109568, + "grad_norm": 0.07456194193490664, + "learning_rate": 9.967123301453073e-05, + "loss": 0.0304, + "step": 34240 + }, + { + "epoch": 0.109632, + "grad_norm": 0.17750935492475595, + "learning_rate": 9.9670849250851e-05, + "loss": 0.0297, + "step": 34260 + }, + { + "epoch": 0.109696, + "grad_norm": 0.12496227220709985, + "learning_rate": 9.967046526406141e-05, + "loss": 0.0332, + "step": 34280 + }, + { + "epoch": 0.10976, + "grad_norm": 0.0818117531943462, + "learning_rate": 9.967008105416369e-05, + "loss": 0.0311, + "step": 34300 + }, + { + "epoch": 0.109824, + "grad_norm": 0.07763838151615514, + "learning_rate": 9.966969662115954e-05, + "loss": 0.0326, + "step": 34320 + }, + { + "epoch": 0.109888, + "grad_norm": 0.13172104271306334, + "learning_rate": 9.966931196505071e-05, + "loss": 0.0326, + "step": 34340 + }, + { + "epoch": 0.109952, + "grad_norm": 0.21239926546746965, + "learning_rate": 9.966892708583893e-05, + "loss": 0.0306, + "step": 34360 + }, + { + "epoch": 0.110016, + "grad_norm": 0.16647877176508197, + "learning_rate": 9.966854198352593e-05, + "loss": 0.0323, + "step": 34380 + }, + { + "epoch": 0.11008, + "grad_norm": 0.07954586211745605, + "learning_rate": 9.966815665811342e-05, + "loss": 0.0333, + "step": 34400 + }, + { + "epoch": 0.110144, + "grad_norm": 0.098133363792694, + "learning_rate": 9.966777110960314e-05, + "loss": 0.0324, + "step": 34420 + }, + { + "epoch": 0.110208, + "grad_norm": 0.12828955538729345, + "learning_rate": 9.966738533799681e-05, + "loss": 0.0324, + "step": 34440 + }, + { + "epoch": 0.110272, + "grad_norm": 0.11635381428101654, + "learning_rate": 9.96669993432962e-05, + "loss": 0.033, + "step": 34460 + }, + { + "epoch": 0.110336, + "grad_norm": 0.09765150742591423, + "learning_rate": 9.966661312550302e-05, + "loss": 0.036, + "step": 34480 + }, + { + "epoch": 0.1104, + "grad_norm": 0.15987883697666064, + "learning_rate": 9.9666226684619e-05, + "loss": 0.0341, + "step": 34500 + }, + { + "epoch": 0.110464, + "grad_norm": 0.20526422904389086, + "learning_rate": 9.966584002064587e-05, + "loss": 0.0351, + "step": 34520 + }, + { + "epoch": 0.110528, + "grad_norm": 0.15687462622152515, + "learning_rate": 9.966545313358537e-05, + "loss": 0.0322, + "step": 34540 + }, + { + "epoch": 0.110592, + "grad_norm": 0.06864821071203991, + "learning_rate": 9.966506602343927e-05, + "loss": 0.0307, + "step": 34560 + }, + { + "epoch": 0.110656, + "grad_norm": 0.07954880762676457, + "learning_rate": 9.966467869020928e-05, + "loss": 0.035, + "step": 34580 + }, + { + "epoch": 0.11072, + "grad_norm": 0.10272942143168402, + "learning_rate": 9.966429113389713e-05, + "loss": 0.0321, + "step": 34600 + }, + { + "epoch": 0.110784, + "grad_norm": 0.10518813600091349, + "learning_rate": 9.966390335450459e-05, + "loss": 0.0334, + "step": 34620 + }, + { + "epoch": 0.110848, + "grad_norm": 0.06577893727131802, + "learning_rate": 9.966351535203337e-05, + "loss": 0.0319, + "step": 34640 + }, + { + "epoch": 0.110912, + "grad_norm": 0.11077496039335785, + "learning_rate": 9.966312712648524e-05, + "loss": 0.0347, + "step": 34660 + }, + { + "epoch": 0.110976, + "grad_norm": 0.07653849681518922, + "learning_rate": 9.966273867786193e-05, + "loss": 0.0322, + "step": 34680 + }, + { + "epoch": 0.11104, + "grad_norm": 0.19128802614444906, + "learning_rate": 9.96623500061652e-05, + "loss": 0.0343, + "step": 34700 + }, + { + "epoch": 0.111104, + "grad_norm": 0.08112054172444903, + "learning_rate": 9.966196111139676e-05, + "loss": 0.0339, + "step": 34720 + }, + { + "epoch": 0.111168, + "grad_norm": 0.06105611108146409, + "learning_rate": 9.966157199355839e-05, + "loss": 0.0309, + "step": 34740 + }, + { + "epoch": 0.111232, + "grad_norm": 0.26120311270766583, + "learning_rate": 9.966118265265183e-05, + "loss": 0.0376, + "step": 34760 + }, + { + "epoch": 0.111296, + "grad_norm": 0.10895669571422546, + "learning_rate": 9.966079308867884e-05, + "loss": 0.0327, + "step": 34780 + }, + { + "epoch": 0.11136, + "grad_norm": 0.06621132672225348, + "learning_rate": 9.966040330164115e-05, + "loss": 0.0317, + "step": 34800 + }, + { + "epoch": 0.111424, + "grad_norm": 0.2305520127981935, + "learning_rate": 9.966001329154051e-05, + "loss": 0.0326, + "step": 34820 + }, + { + "epoch": 0.111488, + "grad_norm": 0.11599012336005006, + "learning_rate": 9.965962305837868e-05, + "loss": 0.0341, + "step": 34840 + }, + { + "epoch": 0.111552, + "grad_norm": 0.21788500157529989, + "learning_rate": 9.965923260215741e-05, + "loss": 0.0305, + "step": 34860 + }, + { + "epoch": 0.111616, + "grad_norm": 0.10354581493204382, + "learning_rate": 9.965884192287846e-05, + "loss": 0.0318, + "step": 34880 + }, + { + "epoch": 0.11168, + "grad_norm": 0.07515625422843464, + "learning_rate": 9.965845102054359e-05, + "loss": 0.0338, + "step": 34900 + }, + { + "epoch": 0.111744, + "grad_norm": 0.11032975800640689, + "learning_rate": 9.965805989515454e-05, + "loss": 0.0347, + "step": 34920 + }, + { + "epoch": 0.111808, + "grad_norm": 0.13430200485598753, + "learning_rate": 9.965766854671308e-05, + "loss": 0.0303, + "step": 34940 + }, + { + "epoch": 0.111872, + "grad_norm": 0.06716717698722977, + "learning_rate": 9.965727697522095e-05, + "loss": 0.0312, + "step": 34960 + }, + { + "epoch": 0.111936, + "grad_norm": 0.14114153614142033, + "learning_rate": 9.965688518067993e-05, + "loss": 0.0327, + "step": 34980 + }, + { + "epoch": 0.112, + "grad_norm": 0.07414635211100958, + "learning_rate": 9.965649316309178e-05, + "loss": 0.0301, + "step": 35000 + }, + { + "epoch": 0.112064, + "grad_norm": 0.07118471874140454, + "learning_rate": 9.965610092245823e-05, + "loss": 0.0345, + "step": 35020 + }, + { + "epoch": 0.112128, + "grad_norm": 0.15535855751004604, + "learning_rate": 9.965570845878107e-05, + "loss": 0.0327, + "step": 35040 + }, + { + "epoch": 0.112192, + "grad_norm": 0.08840878538254744, + "learning_rate": 9.965531577206207e-05, + "loss": 0.0338, + "step": 35060 + }, + { + "epoch": 0.112256, + "grad_norm": 0.08740748618307703, + "learning_rate": 9.965492286230295e-05, + "loss": 0.0329, + "step": 35080 + }, + { + "epoch": 0.11232, + "grad_norm": 0.09877432522625522, + "learning_rate": 9.965452972950553e-05, + "loss": 0.0315, + "step": 35100 + }, + { + "epoch": 0.112384, + "grad_norm": 0.11235007788940649, + "learning_rate": 9.965413637367156e-05, + "loss": 0.0339, + "step": 35120 + }, + { + "epoch": 0.112448, + "grad_norm": 0.09483559661774024, + "learning_rate": 9.965374279480277e-05, + "loss": 0.0285, + "step": 35140 + }, + { + "epoch": 0.112512, + "grad_norm": 0.07621990173282894, + "learning_rate": 9.965334899290097e-05, + "loss": 0.0326, + "step": 35160 + }, + { + "epoch": 0.112576, + "grad_norm": 0.11165487239161613, + "learning_rate": 9.965295496796792e-05, + "loss": 0.0329, + "step": 35180 + }, + { + "epoch": 0.11264, + "grad_norm": 0.10867946658593554, + "learning_rate": 9.965256072000537e-05, + "loss": 0.0322, + "step": 35200 + }, + { + "epoch": 0.112704, + "grad_norm": 0.09039177422645221, + "learning_rate": 9.96521662490151e-05, + "loss": 0.0323, + "step": 35220 + }, + { + "epoch": 0.112768, + "grad_norm": 0.07797394265195395, + "learning_rate": 9.96517715549989e-05, + "loss": 0.0293, + "step": 35240 + }, + { + "epoch": 0.112832, + "grad_norm": 0.08105110731593314, + "learning_rate": 9.965137663795853e-05, + "loss": 0.0305, + "step": 35260 + }, + { + "epoch": 0.112896, + "grad_norm": 0.08529639252197711, + "learning_rate": 9.965098149789576e-05, + "loss": 0.0324, + "step": 35280 + }, + { + "epoch": 0.11296, + "grad_norm": 0.09687399142071647, + "learning_rate": 9.965058613481237e-05, + "loss": 0.0346, + "step": 35300 + }, + { + "epoch": 0.113024, + "grad_norm": 0.07898446868648812, + "learning_rate": 9.965019054871015e-05, + "loss": 0.0356, + "step": 35320 + }, + { + "epoch": 0.113088, + "grad_norm": 0.11245333429759206, + "learning_rate": 9.964979473959084e-05, + "loss": 0.0311, + "step": 35340 + }, + { + "epoch": 0.113152, + "grad_norm": 0.07992269739008, + "learning_rate": 9.964939870745623e-05, + "loss": 0.0301, + "step": 35360 + }, + { + "epoch": 0.113216, + "grad_norm": 0.07071795690274804, + "learning_rate": 9.964900245230813e-05, + "loss": 0.032, + "step": 35380 + }, + { + "epoch": 0.11328, + "grad_norm": 0.08593089022439952, + "learning_rate": 9.96486059741483e-05, + "loss": 0.0324, + "step": 35400 + }, + { + "epoch": 0.113344, + "grad_norm": 0.07448160066411781, + "learning_rate": 9.964820927297848e-05, + "loss": 0.0298, + "step": 35420 + }, + { + "epoch": 0.113408, + "grad_norm": 0.066267619079859, + "learning_rate": 9.964781234880054e-05, + "loss": 0.0307, + "step": 35440 + }, + { + "epoch": 0.113472, + "grad_norm": 0.09633084242587177, + "learning_rate": 9.964741520161617e-05, + "loss": 0.0324, + "step": 35460 + }, + { + "epoch": 0.113536, + "grad_norm": 0.12775898782951386, + "learning_rate": 9.964701783142724e-05, + "loss": 0.0369, + "step": 35480 + }, + { + "epoch": 0.1136, + "grad_norm": 0.21781389689751313, + "learning_rate": 9.964662023823547e-05, + "loss": 0.0351, + "step": 35500 + }, + { + "epoch": 0.113664, + "grad_norm": 0.0905405485420799, + "learning_rate": 9.964622242204267e-05, + "loss": 0.0342, + "step": 35520 + }, + { + "epoch": 0.113728, + "grad_norm": 0.1050233415780439, + "learning_rate": 9.964582438285062e-05, + "loss": 0.0326, + "step": 35540 + }, + { + "epoch": 0.113792, + "grad_norm": 0.21260365424920846, + "learning_rate": 9.964542612066112e-05, + "loss": 0.0313, + "step": 35560 + }, + { + "epoch": 0.113856, + "grad_norm": 0.07598314771829522, + "learning_rate": 9.964502763547596e-05, + "loss": 0.0325, + "step": 35580 + }, + { + "epoch": 0.11392, + "grad_norm": 0.08283825834079331, + "learning_rate": 9.964462892729692e-05, + "loss": 0.0295, + "step": 35600 + }, + { + "epoch": 0.113984, + "grad_norm": 0.08606929264449002, + "learning_rate": 9.964422999612577e-05, + "loss": 0.0321, + "step": 35620 + }, + { + "epoch": 0.114048, + "grad_norm": 0.08219155614431484, + "learning_rate": 9.964383084196434e-05, + "loss": 0.0334, + "step": 35640 + }, + { + "epoch": 0.114112, + "grad_norm": 0.06889528570284775, + "learning_rate": 9.96434314648144e-05, + "loss": 0.0348, + "step": 35660 + }, + { + "epoch": 0.114176, + "grad_norm": 0.10682171940961276, + "learning_rate": 9.964303186467777e-05, + "loss": 0.0322, + "step": 35680 + }, + { + "epoch": 0.11424, + "grad_norm": 0.1188302082481202, + "learning_rate": 9.964263204155621e-05, + "loss": 0.0328, + "step": 35700 + }, + { + "epoch": 0.114304, + "grad_norm": 0.1212317443299198, + "learning_rate": 9.964223199545156e-05, + "loss": 0.0324, + "step": 35720 + }, + { + "epoch": 0.114368, + "grad_norm": 0.07206975891981421, + "learning_rate": 9.964183172636556e-05, + "loss": 0.0316, + "step": 35740 + }, + { + "epoch": 0.114432, + "grad_norm": 0.10371238990275178, + "learning_rate": 9.964143123430004e-05, + "loss": 0.0326, + "step": 35760 + }, + { + "epoch": 0.114496, + "grad_norm": 0.25222712982183304, + "learning_rate": 9.964103051925679e-05, + "loss": 0.0357, + "step": 35780 + }, + { + "epoch": 0.11456, + "grad_norm": 0.08709101829855079, + "learning_rate": 9.964062958123764e-05, + "loss": 0.0321, + "step": 35800 + }, + { + "epoch": 0.114624, + "grad_norm": 0.14738196313053575, + "learning_rate": 9.964022842024434e-05, + "loss": 0.0322, + "step": 35820 + }, + { + "epoch": 0.114688, + "grad_norm": 0.0869109792239195, + "learning_rate": 9.963982703627875e-05, + "loss": 0.0318, + "step": 35840 + }, + { + "epoch": 0.114752, + "grad_norm": 0.09289650622094349, + "learning_rate": 9.963942542934262e-05, + "loss": 0.032, + "step": 35860 + }, + { + "epoch": 0.114816, + "grad_norm": 0.14603235538201367, + "learning_rate": 9.963902359943778e-05, + "loss": 0.0309, + "step": 35880 + }, + { + "epoch": 0.11488, + "grad_norm": 0.11181447536126167, + "learning_rate": 9.963862154656603e-05, + "loss": 0.0304, + "step": 35900 + }, + { + "epoch": 0.114944, + "grad_norm": 0.1201507731219381, + "learning_rate": 9.963821927072917e-05, + "loss": 0.0327, + "step": 35920 + }, + { + "epoch": 0.115008, + "grad_norm": 0.11163069639406126, + "learning_rate": 9.963781677192903e-05, + "loss": 0.0349, + "step": 35940 + }, + { + "epoch": 0.115072, + "grad_norm": 0.07784624014029493, + "learning_rate": 9.963741405016737e-05, + "loss": 0.0355, + "step": 35960 + }, + { + "epoch": 0.115136, + "grad_norm": 0.13112807051339812, + "learning_rate": 9.963701110544605e-05, + "loss": 0.0341, + "step": 35980 + }, + { + "epoch": 0.1152, + "grad_norm": 0.14163617665355058, + "learning_rate": 9.963660793776688e-05, + "loss": 0.0352, + "step": 36000 + }, + { + "epoch": 0.115264, + "grad_norm": 0.08301227358743235, + "learning_rate": 9.963620454713163e-05, + "loss": 0.0299, + "step": 36020 + }, + { + "epoch": 0.115328, + "grad_norm": 0.12552668809291267, + "learning_rate": 9.963580093354214e-05, + "loss": 0.0323, + "step": 36040 + }, + { + "epoch": 0.115392, + "grad_norm": 0.11192323800132616, + "learning_rate": 9.963539709700021e-05, + "loss": 0.0292, + "step": 36060 + }, + { + "epoch": 0.115456, + "grad_norm": 0.07689316691403217, + "learning_rate": 9.963499303750765e-05, + "loss": 0.0333, + "step": 36080 + }, + { + "epoch": 0.11552, + "grad_norm": 0.12870685004086013, + "learning_rate": 9.963458875506631e-05, + "loss": 0.0351, + "step": 36100 + }, + { + "epoch": 0.115584, + "grad_norm": 0.10286516033856004, + "learning_rate": 9.963418424967796e-05, + "loss": 0.0315, + "step": 36120 + }, + { + "epoch": 0.115648, + "grad_norm": 0.1853140649213712, + "learning_rate": 9.963377952134445e-05, + "loss": 0.0334, + "step": 36140 + }, + { + "epoch": 0.115712, + "grad_norm": 0.0816907492207387, + "learning_rate": 9.963337457006759e-05, + "loss": 0.0331, + "step": 36160 + }, + { + "epoch": 0.115776, + "grad_norm": 0.12435708944252365, + "learning_rate": 9.963296939584917e-05, + "loss": 0.0332, + "step": 36180 + }, + { + "epoch": 0.11584, + "grad_norm": 0.24126706543084164, + "learning_rate": 9.963256399869105e-05, + "loss": 0.0346, + "step": 36200 + }, + { + "epoch": 0.115904, + "grad_norm": 0.09324461608543054, + "learning_rate": 9.963215837859504e-05, + "loss": 0.032, + "step": 36220 + }, + { + "epoch": 0.115968, + "grad_norm": 0.07005958573983423, + "learning_rate": 9.963175253556297e-05, + "loss": 0.0326, + "step": 36240 + }, + { + "epoch": 0.116032, + "grad_norm": 0.08482601348167765, + "learning_rate": 9.963134646959662e-05, + "loss": 0.0307, + "step": 36260 + }, + { + "epoch": 0.116096, + "grad_norm": 0.09332993998237313, + "learning_rate": 9.963094018069785e-05, + "loss": 0.0335, + "step": 36280 + }, + { + "epoch": 0.11616, + "grad_norm": 0.08102208392039635, + "learning_rate": 9.96305336688685e-05, + "loss": 0.0294, + "step": 36300 + }, + { + "epoch": 0.116224, + "grad_norm": 0.07889400053020895, + "learning_rate": 9.963012693411035e-05, + "loss": 0.0315, + "step": 36320 + }, + { + "epoch": 0.116288, + "grad_norm": 0.08290655900922446, + "learning_rate": 9.962971997642527e-05, + "loss": 0.0332, + "step": 36340 + }, + { + "epoch": 0.116352, + "grad_norm": 0.16639584641596317, + "learning_rate": 9.962931279581507e-05, + "loss": 0.0323, + "step": 36360 + }, + { + "epoch": 0.116416, + "grad_norm": 0.08624009025538779, + "learning_rate": 9.962890539228158e-05, + "loss": 0.0319, + "step": 36380 + }, + { + "epoch": 0.11648, + "grad_norm": 0.09049129967569076, + "learning_rate": 9.962849776582662e-05, + "loss": 0.0333, + "step": 36400 + }, + { + "epoch": 0.116544, + "grad_norm": 0.17054767438699914, + "learning_rate": 9.962808991645203e-05, + "loss": 0.0269, + "step": 36420 + }, + { + "epoch": 0.116608, + "grad_norm": 0.10314869197136042, + "learning_rate": 9.962768184415965e-05, + "loss": 0.0348, + "step": 36440 + }, + { + "epoch": 0.116672, + "grad_norm": 0.07988447338129137, + "learning_rate": 9.96272735489513e-05, + "loss": 0.0302, + "step": 36460 + }, + { + "epoch": 0.116736, + "grad_norm": 0.10901812093562432, + "learning_rate": 9.962686503082883e-05, + "loss": 0.0316, + "step": 36480 + }, + { + "epoch": 0.1168, + "grad_norm": 0.08448437307653514, + "learning_rate": 9.962645628979406e-05, + "loss": 0.0337, + "step": 36500 + }, + { + "epoch": 0.116864, + "grad_norm": 0.09970780438779868, + "learning_rate": 9.962604732584882e-05, + "loss": 0.034, + "step": 36520 + }, + { + "epoch": 0.116928, + "grad_norm": 0.08370613425580523, + "learning_rate": 9.962563813899496e-05, + "loss": 0.0285, + "step": 36540 + }, + { + "epoch": 0.116992, + "grad_norm": 0.11498425229332991, + "learning_rate": 9.962522872923432e-05, + "loss": 0.0345, + "step": 36560 + }, + { + "epoch": 0.117056, + "grad_norm": 0.07347575431659874, + "learning_rate": 9.962481909656872e-05, + "loss": 0.0334, + "step": 36580 + }, + { + "epoch": 0.11712, + "grad_norm": 0.09847315221511913, + "learning_rate": 9.962440924100002e-05, + "loss": 0.0309, + "step": 36600 + }, + { + "epoch": 0.117184, + "grad_norm": 0.11025392463376181, + "learning_rate": 9.962399916253006e-05, + "loss": 0.0333, + "step": 36620 + }, + { + "epoch": 0.117248, + "grad_norm": 0.13344232688950747, + "learning_rate": 9.962358886116067e-05, + "loss": 0.0313, + "step": 36640 + }, + { + "epoch": 0.117312, + "grad_norm": 0.1275129167201773, + "learning_rate": 9.962317833689372e-05, + "loss": 0.0294, + "step": 36660 + }, + { + "epoch": 0.117376, + "grad_norm": 0.07249345607926054, + "learning_rate": 9.962276758973101e-05, + "loss": 0.0345, + "step": 36680 + }, + { + "epoch": 0.11744, + "grad_norm": 0.09248714158655268, + "learning_rate": 9.962235661967443e-05, + "loss": 0.0336, + "step": 36700 + }, + { + "epoch": 0.117504, + "grad_norm": 0.12860822382391035, + "learning_rate": 9.96219454267258e-05, + "loss": 0.0296, + "step": 36720 + }, + { + "epoch": 0.117568, + "grad_norm": 0.07789809871865498, + "learning_rate": 9.962153401088697e-05, + "loss": 0.0328, + "step": 36740 + }, + { + "epoch": 0.117632, + "grad_norm": 0.07069656804970778, + "learning_rate": 9.962112237215977e-05, + "loss": 0.0324, + "step": 36760 + }, + { + "epoch": 0.117696, + "grad_norm": 0.09917526205337666, + "learning_rate": 9.96207105105461e-05, + "loss": 0.0342, + "step": 36780 + }, + { + "epoch": 0.11776, + "grad_norm": 0.08874343309573036, + "learning_rate": 9.962029842604775e-05, + "loss": 0.0316, + "step": 36800 + }, + { + "epoch": 0.117824, + "grad_norm": 0.1334900106199603, + "learning_rate": 9.961988611866661e-05, + "loss": 0.034, + "step": 36820 + }, + { + "epoch": 0.117888, + "grad_norm": 0.08769284167520282, + "learning_rate": 9.961947358840453e-05, + "loss": 0.0318, + "step": 36840 + }, + { + "epoch": 0.117952, + "grad_norm": 0.13450040532854868, + "learning_rate": 9.961906083526335e-05, + "loss": 0.0315, + "step": 36860 + }, + { + "epoch": 0.118016, + "grad_norm": 0.1099362879940964, + "learning_rate": 9.961864785924492e-05, + "loss": 0.0328, + "step": 36880 + }, + { + "epoch": 0.11808, + "grad_norm": 0.14485320645545152, + "learning_rate": 9.961823466035112e-05, + "loss": 0.0347, + "step": 36900 + }, + { + "epoch": 0.118144, + "grad_norm": 0.116712882415643, + "learning_rate": 9.961782123858376e-05, + "loss": 0.0319, + "step": 36920 + }, + { + "epoch": 0.118208, + "grad_norm": 0.08472762090815263, + "learning_rate": 9.961740759394475e-05, + "loss": 0.0326, + "step": 36940 + }, + { + "epoch": 0.118272, + "grad_norm": 0.21433302859337203, + "learning_rate": 9.961699372643594e-05, + "loss": 0.0319, + "step": 36960 + }, + { + "epoch": 0.118336, + "grad_norm": 0.15915141512032954, + "learning_rate": 9.961657963605915e-05, + "loss": 0.0319, + "step": 36980 + }, + { + "epoch": 0.1184, + "grad_norm": 0.08825772952212092, + "learning_rate": 9.961616532281626e-05, + "loss": 0.0322, + "step": 37000 + }, + { + "epoch": 0.118464, + "grad_norm": 0.11472558796082853, + "learning_rate": 9.961575078670914e-05, + "loss": 0.0357, + "step": 37020 + }, + { + "epoch": 0.118528, + "grad_norm": 0.12525367084922528, + "learning_rate": 9.961533602773966e-05, + "loss": 0.0299, + "step": 37040 + }, + { + "epoch": 0.118592, + "grad_norm": 0.07667885639208896, + "learning_rate": 9.961492104590966e-05, + "loss": 0.0313, + "step": 37060 + }, + { + "epoch": 0.118656, + "grad_norm": 0.06394329628703295, + "learning_rate": 9.9614505841221e-05, + "loss": 0.0319, + "step": 37080 + }, + { + "epoch": 0.11872, + "grad_norm": 0.08889329997105967, + "learning_rate": 9.961409041367558e-05, + "loss": 0.0324, + "step": 37100 + }, + { + "epoch": 0.118784, + "grad_norm": 0.16448788335958803, + "learning_rate": 9.961367476327524e-05, + "loss": 0.0306, + "step": 37120 + }, + { + "epoch": 0.118848, + "grad_norm": 0.07129643674223644, + "learning_rate": 9.961325889002184e-05, + "loss": 0.0314, + "step": 37140 + }, + { + "epoch": 0.118912, + "grad_norm": 0.17996250742373984, + "learning_rate": 9.961284279391728e-05, + "loss": 0.0322, + "step": 37160 + }, + { + "epoch": 0.118976, + "grad_norm": 0.08283863768422912, + "learning_rate": 9.961242647496338e-05, + "loss": 0.0336, + "step": 37180 + }, + { + "epoch": 0.11904, + "grad_norm": 0.05928128878429634, + "learning_rate": 9.961200993316206e-05, + "loss": 0.0334, + "step": 37200 + }, + { + "epoch": 0.119104, + "grad_norm": 0.0921784566230797, + "learning_rate": 9.961159316851517e-05, + "loss": 0.0289, + "step": 37220 + }, + { + "epoch": 0.119168, + "grad_norm": 0.09774449373308244, + "learning_rate": 9.961117618102458e-05, + "loss": 0.0305, + "step": 37240 + }, + { + "epoch": 0.119232, + "grad_norm": 0.07920167374695274, + "learning_rate": 9.961075897069216e-05, + "loss": 0.0296, + "step": 37260 + }, + { + "epoch": 0.119296, + "grad_norm": 0.06814190272390024, + "learning_rate": 9.961034153751979e-05, + "loss": 0.03, + "step": 37280 + }, + { + "epoch": 0.11936, + "grad_norm": 0.10863133391098156, + "learning_rate": 9.960992388150934e-05, + "loss": 0.0306, + "step": 37300 + }, + { + "epoch": 0.119424, + "grad_norm": 0.10003355823270552, + "learning_rate": 9.96095060026627e-05, + "loss": 0.032, + "step": 37320 + }, + { + "epoch": 0.119488, + "grad_norm": 0.07343144935971223, + "learning_rate": 9.960908790098173e-05, + "loss": 0.0345, + "step": 37340 + }, + { + "epoch": 0.119552, + "grad_norm": 0.199754190942915, + "learning_rate": 9.96086695764683e-05, + "loss": 0.0362, + "step": 37360 + }, + { + "epoch": 0.119616, + "grad_norm": 0.09478636483968902, + "learning_rate": 9.960825102912433e-05, + "loss": 0.0325, + "step": 37380 + }, + { + "epoch": 0.11968, + "grad_norm": 0.08608315181705586, + "learning_rate": 9.960783225895167e-05, + "loss": 0.0311, + "step": 37400 + }, + { + "epoch": 0.119744, + "grad_norm": 0.14099814528510562, + "learning_rate": 9.960741326595219e-05, + "loss": 0.03, + "step": 37420 + }, + { + "epoch": 0.119808, + "grad_norm": 0.10023477835641532, + "learning_rate": 9.960699405012778e-05, + "loss": 0.0309, + "step": 37440 + }, + { + "epoch": 0.119872, + "grad_norm": 0.08632691022757225, + "learning_rate": 9.960657461148035e-05, + "loss": 0.0314, + "step": 37460 + }, + { + "epoch": 0.119936, + "grad_norm": 0.09088508742133818, + "learning_rate": 9.960615495001176e-05, + "loss": 0.0326, + "step": 37480 + }, + { + "epoch": 0.12, + "grad_norm": 0.08246502504082026, + "learning_rate": 9.96057350657239e-05, + "loss": 0.0327, + "step": 37500 + }, + { + "epoch": 0.120064, + "grad_norm": 0.08889452829563031, + "learning_rate": 9.960531495861864e-05, + "loss": 0.0327, + "step": 37520 + }, + { + "epoch": 0.120128, + "grad_norm": 0.1321810821875133, + "learning_rate": 9.960489462869791e-05, + "loss": 0.0317, + "step": 37540 + }, + { + "epoch": 0.120192, + "grad_norm": 0.07901498676306969, + "learning_rate": 9.960447407596353e-05, + "loss": 0.0339, + "step": 37560 + }, + { + "epoch": 0.120256, + "grad_norm": 0.11663780289091064, + "learning_rate": 9.960405330041747e-05, + "loss": 0.0329, + "step": 37580 + }, + { + "epoch": 0.12032, + "grad_norm": 0.08682732591404133, + "learning_rate": 9.960363230206155e-05, + "loss": 0.0315, + "step": 37600 + }, + { + "epoch": 0.120384, + "grad_norm": 0.0685551512663347, + "learning_rate": 9.96032110808977e-05, + "loss": 0.0327, + "step": 37620 + }, + { + "epoch": 0.120448, + "grad_norm": 0.08001224602140601, + "learning_rate": 9.960278963692781e-05, + "loss": 0.0317, + "step": 37640 + }, + { + "epoch": 0.120512, + "grad_norm": 0.11604695589436578, + "learning_rate": 9.960236797015376e-05, + "loss": 0.0312, + "step": 37660 + }, + { + "epoch": 0.120576, + "grad_norm": 0.07781578153568748, + "learning_rate": 9.960194608057744e-05, + "loss": 0.03, + "step": 37680 + }, + { + "epoch": 0.12064, + "grad_norm": 0.07393264932850753, + "learning_rate": 9.960152396820077e-05, + "loss": 0.0302, + "step": 37700 + }, + { + "epoch": 0.120704, + "grad_norm": 0.09496256664945173, + "learning_rate": 9.960110163302561e-05, + "loss": 0.0321, + "step": 37720 + }, + { + "epoch": 0.120768, + "grad_norm": 0.07869260080496425, + "learning_rate": 9.96006790750539e-05, + "loss": 0.0333, + "step": 37740 + }, + { + "epoch": 0.120832, + "grad_norm": 0.08479614574939452, + "learning_rate": 9.96002562942875e-05, + "loss": 0.0312, + "step": 37760 + }, + { + "epoch": 0.120896, + "grad_norm": 0.09627047464540824, + "learning_rate": 9.959983329072832e-05, + "loss": 0.0328, + "step": 37780 + }, + { + "epoch": 0.12096, + "grad_norm": 0.07777341070169186, + "learning_rate": 9.959941006437828e-05, + "loss": 0.03, + "step": 37800 + }, + { + "epoch": 0.121024, + "grad_norm": 0.08964891133063992, + "learning_rate": 9.959898661523926e-05, + "loss": 0.0302, + "step": 37820 + }, + { + "epoch": 0.121088, + "grad_norm": 0.096323282291661, + "learning_rate": 9.959856294331317e-05, + "loss": 0.0304, + "step": 37840 + }, + { + "epoch": 0.121152, + "grad_norm": 0.059413719899620195, + "learning_rate": 9.95981390486019e-05, + "loss": 0.0339, + "step": 37860 + }, + { + "epoch": 0.121216, + "grad_norm": 0.18543590326411122, + "learning_rate": 9.959771493110737e-05, + "loss": 0.0335, + "step": 37880 + }, + { + "epoch": 0.12128, + "grad_norm": 0.16241868184193745, + "learning_rate": 9.959729059083148e-05, + "loss": 0.0334, + "step": 37900 + }, + { + "epoch": 0.121344, + "grad_norm": 0.07762208489545298, + "learning_rate": 9.959686602777613e-05, + "loss": 0.0327, + "step": 37920 + }, + { + "epoch": 0.121408, + "grad_norm": 0.06391938325212523, + "learning_rate": 9.959644124194325e-05, + "loss": 0.0311, + "step": 37940 + }, + { + "epoch": 0.121472, + "grad_norm": 0.08954332437290162, + "learning_rate": 9.95960162333347e-05, + "loss": 0.0306, + "step": 37960 + }, + { + "epoch": 0.121536, + "grad_norm": 0.148763576552993, + "learning_rate": 9.959559100195245e-05, + "loss": 0.0319, + "step": 37980 + }, + { + "epoch": 0.1216, + "grad_norm": 0.1455620143215255, + "learning_rate": 9.959516554779838e-05, + "loss": 0.0312, + "step": 38000 + }, + { + "epoch": 0.121664, + "grad_norm": 0.09155855801499024, + "learning_rate": 9.959473987087437e-05, + "loss": 0.0342, + "step": 38020 + }, + { + "epoch": 0.121728, + "grad_norm": 0.09010646402640614, + "learning_rate": 9.959431397118238e-05, + "loss": 0.0352, + "step": 38040 + }, + { + "epoch": 0.121792, + "grad_norm": 0.11263007761936172, + "learning_rate": 9.959388784872432e-05, + "loss": 0.0341, + "step": 38060 + }, + { + "epoch": 0.121856, + "grad_norm": 0.09073746559180812, + "learning_rate": 9.959346150350206e-05, + "loss": 0.0308, + "step": 38080 + }, + { + "epoch": 0.12192, + "grad_norm": 0.0742683933049584, + "learning_rate": 9.959303493551757e-05, + "loss": 0.0299, + "step": 38100 + }, + { + "epoch": 0.121984, + "grad_norm": 0.12539145298724186, + "learning_rate": 9.959260814477273e-05, + "loss": 0.029, + "step": 38120 + }, + { + "epoch": 0.122048, + "grad_norm": 0.06248153998868201, + "learning_rate": 9.959218113126947e-05, + "loss": 0.0318, + "step": 38140 + }, + { + "epoch": 0.122112, + "grad_norm": 0.11062921014169383, + "learning_rate": 9.959175389500971e-05, + "loss": 0.0325, + "step": 38160 + }, + { + "epoch": 0.122176, + "grad_norm": 0.1320685806951766, + "learning_rate": 9.959132643599536e-05, + "loss": 0.0314, + "step": 38180 + }, + { + "epoch": 0.12224, + "grad_norm": 0.14569165704338244, + "learning_rate": 9.959089875422835e-05, + "loss": 0.0323, + "step": 38200 + }, + { + "epoch": 0.122304, + "grad_norm": 0.06420857359762702, + "learning_rate": 9.959047084971058e-05, + "loss": 0.0314, + "step": 38220 + }, + { + "epoch": 0.122368, + "grad_norm": 0.13504041633184086, + "learning_rate": 9.9590042722444e-05, + "loss": 0.03, + "step": 38240 + }, + { + "epoch": 0.122432, + "grad_norm": 0.15690759686030945, + "learning_rate": 9.958961437243053e-05, + "loss": 0.0337, + "step": 38260 + }, + { + "epoch": 0.122496, + "grad_norm": 0.0866867000474349, + "learning_rate": 9.958918579967209e-05, + "loss": 0.0315, + "step": 38280 + }, + { + "epoch": 0.12256, + "grad_norm": 0.10080800024064544, + "learning_rate": 9.958875700417058e-05, + "loss": 0.0333, + "step": 38300 + }, + { + "epoch": 0.122624, + "grad_norm": 0.07058440310887037, + "learning_rate": 9.958832798592794e-05, + "loss": 0.0358, + "step": 38320 + }, + { + "epoch": 0.122688, + "grad_norm": 0.0872821702594187, + "learning_rate": 9.958789874494613e-05, + "loss": 0.0305, + "step": 38340 + }, + { + "epoch": 0.122752, + "grad_norm": 0.07691606593550346, + "learning_rate": 9.958746928122704e-05, + "loss": 0.034, + "step": 38360 + }, + { + "epoch": 0.122816, + "grad_norm": 0.06526085192180134, + "learning_rate": 9.95870395947726e-05, + "loss": 0.0338, + "step": 38380 + }, + { + "epoch": 0.12288, + "grad_norm": 0.07692751393944329, + "learning_rate": 9.958660968558477e-05, + "loss": 0.031, + "step": 38400 + }, + { + "epoch": 0.122944, + "grad_norm": 0.07056905252964311, + "learning_rate": 9.958617955366545e-05, + "loss": 0.0332, + "step": 38420 + }, + { + "epoch": 0.123008, + "grad_norm": 0.08106161963777715, + "learning_rate": 9.958574919901658e-05, + "loss": 0.0301, + "step": 38440 + }, + { + "epoch": 0.123072, + "grad_norm": 0.10813464569357546, + "learning_rate": 9.958531862164012e-05, + "loss": 0.031, + "step": 38460 + }, + { + "epoch": 0.123136, + "grad_norm": 0.09981836188862685, + "learning_rate": 9.958488782153794e-05, + "loss": 0.0313, + "step": 38480 + }, + { + "epoch": 0.1232, + "grad_norm": 0.06453276326648512, + "learning_rate": 9.958445679871204e-05, + "loss": 0.0322, + "step": 38500 + }, + { + "epoch": 0.123264, + "grad_norm": 0.1263919886806742, + "learning_rate": 9.958402555316434e-05, + "loss": 0.0277, + "step": 38520 + }, + { + "epoch": 0.123328, + "grad_norm": 0.13657751013926692, + "learning_rate": 9.958359408489675e-05, + "loss": 0.0305, + "step": 38540 + }, + { + "epoch": 0.123392, + "grad_norm": 0.07438643098926981, + "learning_rate": 9.958316239391125e-05, + "loss": 0.0293, + "step": 38560 + }, + { + "epoch": 0.123456, + "grad_norm": 0.06962060373831257, + "learning_rate": 9.958273048020974e-05, + "loss": 0.027, + "step": 38580 + }, + { + "epoch": 0.12352, + "grad_norm": 0.12479607813053446, + "learning_rate": 9.958229834379418e-05, + "loss": 0.0323, + "step": 38600 + }, + { + "epoch": 0.123584, + "grad_norm": 0.08314557303219845, + "learning_rate": 9.95818659846665e-05, + "loss": 0.0328, + "step": 38620 + }, + { + "epoch": 0.123648, + "grad_norm": 0.1578443272812555, + "learning_rate": 9.958143340282866e-05, + "loss": 0.0368, + "step": 38640 + }, + { + "epoch": 0.123712, + "grad_norm": 0.08696029487331432, + "learning_rate": 9.95810005982826e-05, + "loss": 0.0305, + "step": 38660 + }, + { + "epoch": 0.123776, + "grad_norm": 0.1479389415135321, + "learning_rate": 9.958056757103022e-05, + "loss": 0.0331, + "step": 38680 + }, + { + "epoch": 0.12384, + "grad_norm": 0.08486353588233876, + "learning_rate": 9.958013432107354e-05, + "loss": 0.0361, + "step": 38700 + }, + { + "epoch": 0.123904, + "grad_norm": 0.09787845679421414, + "learning_rate": 9.957970084841445e-05, + "loss": 0.0303, + "step": 38720 + }, + { + "epoch": 0.123968, + "grad_norm": 0.07705037361936531, + "learning_rate": 9.957926715305492e-05, + "loss": 0.0324, + "step": 38740 + }, + { + "epoch": 0.124032, + "grad_norm": 0.08841673125882069, + "learning_rate": 9.957883323499688e-05, + "loss": 0.0321, + "step": 38760 + }, + { + "epoch": 0.124096, + "grad_norm": 0.13983398233631755, + "learning_rate": 9.957839909424231e-05, + "loss": 0.0297, + "step": 38780 + }, + { + "epoch": 0.12416, + "grad_norm": 0.10339239535259723, + "learning_rate": 9.957796473079312e-05, + "loss": 0.0293, + "step": 38800 + }, + { + "epoch": 0.124224, + "grad_norm": 0.0573774100847272, + "learning_rate": 9.95775301446513e-05, + "loss": 0.0341, + "step": 38820 + }, + { + "epoch": 0.124288, + "grad_norm": 0.10976689924127478, + "learning_rate": 9.957709533581879e-05, + "loss": 0.0294, + "step": 38840 + }, + { + "epoch": 0.124352, + "grad_norm": 0.09205294342991327, + "learning_rate": 9.957666030429754e-05, + "loss": 0.0324, + "step": 38860 + }, + { + "epoch": 0.124416, + "grad_norm": 0.0951765953928565, + "learning_rate": 9.957622505008948e-05, + "loss": 0.0347, + "step": 38880 + }, + { + "epoch": 0.12448, + "grad_norm": 0.13906652155605545, + "learning_rate": 9.95757895731966e-05, + "loss": 0.0319, + "step": 38900 + }, + { + "epoch": 0.124544, + "grad_norm": 0.10560021604480169, + "learning_rate": 9.957535387362083e-05, + "loss": 0.0305, + "step": 38920 + }, + { + "epoch": 0.124608, + "grad_norm": 0.16123240415331652, + "learning_rate": 9.957491795136415e-05, + "loss": 0.0338, + "step": 38940 + }, + { + "epoch": 0.124672, + "grad_norm": 0.22815574383559675, + "learning_rate": 9.957448180642853e-05, + "loss": 0.033, + "step": 38960 + }, + { + "epoch": 0.124736, + "grad_norm": 0.11937765593408194, + "learning_rate": 9.957404543881588e-05, + "loss": 0.0342, + "step": 38980 + }, + { + "epoch": 0.1248, + "grad_norm": 0.13034239211092538, + "learning_rate": 9.957360884852817e-05, + "loss": 0.0339, + "step": 39000 + }, + { + "epoch": 0.124864, + "grad_norm": 0.09409582360001219, + "learning_rate": 9.957317203556742e-05, + "loss": 0.0309, + "step": 39020 + }, + { + "epoch": 0.124928, + "grad_norm": 0.11407115901027594, + "learning_rate": 9.957273499993554e-05, + "loss": 0.0325, + "step": 39040 + }, + { + "epoch": 0.124992, + "grad_norm": 0.09297186598463913, + "learning_rate": 9.957229774163448e-05, + "loss": 0.0352, + "step": 39060 + }, + { + "epoch": 0.125056, + "grad_norm": 0.07681236665967694, + "learning_rate": 9.957186026066625e-05, + "loss": 0.032, + "step": 39080 + }, + { + "epoch": 0.12512, + "grad_norm": 0.07652828196392733, + "learning_rate": 9.957142255703278e-05, + "loss": 0.0324, + "step": 39100 + }, + { + "epoch": 0.125184, + "grad_norm": 0.10642717663941928, + "learning_rate": 9.957098463073606e-05, + "loss": 0.0321, + "step": 39120 + }, + { + "epoch": 0.125248, + "grad_norm": 0.09795610309229415, + "learning_rate": 9.957054648177804e-05, + "loss": 0.0339, + "step": 39140 + }, + { + "epoch": 0.125312, + "grad_norm": 0.09449407885671815, + "learning_rate": 9.95701081101607e-05, + "loss": 0.0313, + "step": 39160 + }, + { + "epoch": 0.125376, + "grad_norm": 0.05662477000546974, + "learning_rate": 9.956966951588599e-05, + "loss": 0.0316, + "step": 39180 + }, + { + "epoch": 0.12544, + "grad_norm": 0.1073213635236539, + "learning_rate": 9.956923069895589e-05, + "loss": 0.0313, + "step": 39200 + }, + { + "epoch": 0.125504, + "grad_norm": 0.15738653049903847, + "learning_rate": 9.956879165937238e-05, + "loss": 0.032, + "step": 39220 + }, + { + "epoch": 0.125568, + "grad_norm": 0.10696302004775364, + "learning_rate": 9.956835239713744e-05, + "loss": 0.0352, + "step": 39240 + }, + { + "epoch": 0.125632, + "grad_norm": 0.20304967423644743, + "learning_rate": 9.956791291225303e-05, + "loss": 0.0356, + "step": 39260 + }, + { + "epoch": 0.125696, + "grad_norm": 0.1210661986368306, + "learning_rate": 9.956747320472111e-05, + "loss": 0.0318, + "step": 39280 + }, + { + "epoch": 0.12576, + "grad_norm": 0.14886504929292205, + "learning_rate": 9.956703327454366e-05, + "loss": 0.0346, + "step": 39300 + }, + { + "epoch": 0.125824, + "grad_norm": 0.08389351362792408, + "learning_rate": 9.956659312172268e-05, + "loss": 0.0321, + "step": 39320 + }, + { + "epoch": 0.125888, + "grad_norm": 0.11878246431183871, + "learning_rate": 9.956615274626013e-05, + "loss": 0.0347, + "step": 39340 + }, + { + "epoch": 0.125952, + "grad_norm": 0.07202256898378927, + "learning_rate": 9.956571214815798e-05, + "loss": 0.0282, + "step": 39360 + }, + { + "epoch": 0.126016, + "grad_norm": 0.07720423852241791, + "learning_rate": 9.956527132741822e-05, + "loss": 0.0329, + "step": 39380 + }, + { + "epoch": 0.12608, + "grad_norm": 0.14900831389737892, + "learning_rate": 9.956483028404284e-05, + "loss": 0.0314, + "step": 39400 + }, + { + "epoch": 0.126144, + "grad_norm": 0.20536609035846695, + "learning_rate": 9.95643890180338e-05, + "loss": 0.0355, + "step": 39420 + }, + { + "epoch": 0.126208, + "grad_norm": 0.0813143555221332, + "learning_rate": 9.95639475293931e-05, + "loss": 0.0309, + "step": 39440 + }, + { + "epoch": 0.126272, + "grad_norm": 0.0764856648150978, + "learning_rate": 9.95635058181227e-05, + "loss": 0.0322, + "step": 39460 + }, + { + "epoch": 0.126336, + "grad_norm": 0.07504408267757684, + "learning_rate": 9.95630638842246e-05, + "loss": 0.0317, + "step": 39480 + }, + { + "epoch": 0.1264, + "grad_norm": 0.11457636337697667, + "learning_rate": 9.956262172770081e-05, + "loss": 0.0334, + "step": 39500 + }, + { + "epoch": 0.126464, + "grad_norm": 0.08060679317289925, + "learning_rate": 9.956217934855327e-05, + "loss": 0.0326, + "step": 39520 + }, + { + "epoch": 0.126528, + "grad_norm": 0.09030527663192844, + "learning_rate": 9.956173674678399e-05, + "loss": 0.0324, + "step": 39540 + }, + { + "epoch": 0.126592, + "grad_norm": 0.10674975014923327, + "learning_rate": 9.956129392239494e-05, + "loss": 0.0322, + "step": 39560 + }, + { + "epoch": 0.126656, + "grad_norm": 0.10411163035870083, + "learning_rate": 9.956085087538815e-05, + "loss": 0.0338, + "step": 39580 + }, + { + "epoch": 0.12672, + "grad_norm": 0.09587562018185987, + "learning_rate": 9.956040760576558e-05, + "loss": 0.0365, + "step": 39600 + }, + { + "epoch": 0.126784, + "grad_norm": 0.09678445233154945, + "learning_rate": 9.95599641135292e-05, + "loss": 0.0352, + "step": 39620 + }, + { + "epoch": 0.126848, + "grad_norm": 0.06941188138290474, + "learning_rate": 9.955952039868104e-05, + "loss": 0.0355, + "step": 39640 + }, + { + "epoch": 0.126912, + "grad_norm": 0.08906280805652561, + "learning_rate": 9.95590764612231e-05, + "loss": 0.0329, + "step": 39660 + }, + { + "epoch": 0.126976, + "grad_norm": 0.12063572600007588, + "learning_rate": 9.955863230115734e-05, + "loss": 0.0296, + "step": 39680 + }, + { + "epoch": 0.12704, + "grad_norm": 0.09593095220152408, + "learning_rate": 9.955818791848577e-05, + "loss": 0.0311, + "step": 39700 + }, + { + "epoch": 0.127104, + "grad_norm": 0.06888410724636199, + "learning_rate": 9.955774331321038e-05, + "loss": 0.0352, + "step": 39720 + }, + { + "epoch": 0.127168, + "grad_norm": 0.08267891878669127, + "learning_rate": 9.955729848533317e-05, + "loss": 0.0328, + "step": 39740 + }, + { + "epoch": 0.127232, + "grad_norm": 0.05268751042868993, + "learning_rate": 9.955685343485614e-05, + "loss": 0.0304, + "step": 39760 + }, + { + "epoch": 0.127296, + "grad_norm": 0.07457312714482409, + "learning_rate": 9.955640816178132e-05, + "loss": 0.0275, + "step": 39780 + }, + { + "epoch": 0.12736, + "grad_norm": 0.09277668520421922, + "learning_rate": 9.955596266611065e-05, + "loss": 0.0308, + "step": 39800 + }, + { + "epoch": 0.127424, + "grad_norm": 0.10209390127341922, + "learning_rate": 9.955551694784617e-05, + "loss": 0.0316, + "step": 39820 + }, + { + "epoch": 0.127488, + "grad_norm": 0.08738138992487515, + "learning_rate": 9.955507100698987e-05, + "loss": 0.0337, + "step": 39840 + }, + { + "epoch": 0.127552, + "grad_norm": 0.09458218213787102, + "learning_rate": 9.955462484354376e-05, + "loss": 0.0322, + "step": 39860 + }, + { + "epoch": 0.127616, + "grad_norm": 0.10943637967201195, + "learning_rate": 9.955417845750983e-05, + "loss": 0.0323, + "step": 39880 + }, + { + "epoch": 0.12768, + "grad_norm": 0.08951642128603933, + "learning_rate": 9.95537318488901e-05, + "loss": 0.0313, + "step": 39900 + }, + { + "epoch": 0.127744, + "grad_norm": 0.09017747289827038, + "learning_rate": 9.955328501768658e-05, + "loss": 0.034, + "step": 39920 + }, + { + "epoch": 0.127808, + "grad_norm": 0.0791438670043615, + "learning_rate": 9.955283796390127e-05, + "loss": 0.0308, + "step": 39940 + }, + { + "epoch": 0.127872, + "grad_norm": 0.10182932469577603, + "learning_rate": 9.955239068753615e-05, + "loss": 0.0357, + "step": 39960 + }, + { + "epoch": 0.127936, + "grad_norm": 0.10794714253061603, + "learning_rate": 9.955194318859329e-05, + "loss": 0.0334, + "step": 39980 + }, + { + "epoch": 0.128, + "grad_norm": 0.07775143816921318, + "learning_rate": 9.955149546707465e-05, + "loss": 0.0313, + "step": 40000 + }, + { + "epoch": 0.128064, + "grad_norm": 0.12111415947423365, + "learning_rate": 9.955104752298225e-05, + "loss": 0.0326, + "step": 40020 + }, + { + "epoch": 0.128128, + "grad_norm": 0.08521910863822091, + "learning_rate": 9.955059935631812e-05, + "loss": 0.0321, + "step": 40040 + }, + { + "epoch": 0.128192, + "grad_norm": 0.22515017294982573, + "learning_rate": 9.955015096708425e-05, + "loss": 0.0333, + "step": 40060 + }, + { + "epoch": 0.128256, + "grad_norm": 0.08638574572649986, + "learning_rate": 9.954970235528268e-05, + "loss": 0.0323, + "step": 40080 + }, + { + "epoch": 0.12832, + "grad_norm": 0.08067480597548514, + "learning_rate": 9.95492535209154e-05, + "loss": 0.0338, + "step": 40100 + }, + { + "epoch": 0.128384, + "grad_norm": 0.13912843348875104, + "learning_rate": 9.954880446398443e-05, + "loss": 0.0337, + "step": 40120 + }, + { + "epoch": 0.128448, + "grad_norm": 0.19839692418848523, + "learning_rate": 9.95483551844918e-05, + "loss": 0.0311, + "step": 40140 + }, + { + "epoch": 0.128512, + "grad_norm": 0.09460287034690226, + "learning_rate": 9.954790568243953e-05, + "loss": 0.0335, + "step": 40160 + }, + { + "epoch": 0.128576, + "grad_norm": 0.08422258199923877, + "learning_rate": 9.954745595782963e-05, + "loss": 0.0346, + "step": 40180 + }, + { + "epoch": 0.12864, + "grad_norm": 0.08237558677822139, + "learning_rate": 9.954700601066411e-05, + "loss": 0.0313, + "step": 40200 + }, + { + "epoch": 0.128704, + "grad_norm": 0.11919659778323012, + "learning_rate": 9.9546555840945e-05, + "loss": 0.0299, + "step": 40220 + }, + { + "epoch": 0.128768, + "grad_norm": 0.115897405389557, + "learning_rate": 9.954610544867435e-05, + "loss": 0.0319, + "step": 40240 + }, + { + "epoch": 0.128832, + "grad_norm": 0.12702723257618226, + "learning_rate": 9.954565483385414e-05, + "loss": 0.0331, + "step": 40260 + }, + { + "epoch": 0.128896, + "grad_norm": 0.07355593359197983, + "learning_rate": 9.954520399648641e-05, + "loss": 0.0333, + "step": 40280 + }, + { + "epoch": 0.12896, + "grad_norm": 0.08201746668289486, + "learning_rate": 9.954475293657319e-05, + "loss": 0.0303, + "step": 40300 + }, + { + "epoch": 0.129024, + "grad_norm": 0.09889558838267845, + "learning_rate": 9.95443016541165e-05, + "loss": 0.029, + "step": 40320 + }, + { + "epoch": 0.129088, + "grad_norm": 0.10763538244875848, + "learning_rate": 9.954385014911837e-05, + "loss": 0.0293, + "step": 40340 + }, + { + "epoch": 0.129152, + "grad_norm": 0.11300801484558425, + "learning_rate": 9.954339842158082e-05, + "loss": 0.035, + "step": 40360 + }, + { + "epoch": 0.129216, + "grad_norm": 0.07770659894096697, + "learning_rate": 9.954294647150591e-05, + "loss": 0.0312, + "step": 40380 + }, + { + "epoch": 0.12928, + "grad_norm": 0.11459264621942214, + "learning_rate": 9.954249429889561e-05, + "loss": 0.0297, + "step": 40400 + }, + { + "epoch": 0.129344, + "grad_norm": 0.06555968557994865, + "learning_rate": 9.954204190375202e-05, + "loss": 0.027, + "step": 40420 + }, + { + "epoch": 0.129408, + "grad_norm": 0.13224845796507306, + "learning_rate": 9.954158928607713e-05, + "loss": 0.0298, + "step": 40440 + }, + { + "epoch": 0.129472, + "grad_norm": 0.08742638659981623, + "learning_rate": 9.954113644587298e-05, + "loss": 0.0323, + "step": 40460 + }, + { + "epoch": 0.129536, + "grad_norm": 0.12258153162505332, + "learning_rate": 9.954068338314159e-05, + "loss": 0.0342, + "step": 40480 + }, + { + "epoch": 0.1296, + "grad_norm": 0.08355400628367804, + "learning_rate": 9.954023009788504e-05, + "loss": 0.0297, + "step": 40500 + }, + { + "epoch": 0.129664, + "grad_norm": 0.09128510119969128, + "learning_rate": 9.953977659010534e-05, + "loss": 0.0337, + "step": 40520 + }, + { + "epoch": 0.129728, + "grad_norm": 0.0651543225995682, + "learning_rate": 9.953932285980451e-05, + "loss": 0.0331, + "step": 40540 + }, + { + "epoch": 0.129792, + "grad_norm": 0.07929497868530182, + "learning_rate": 9.953886890698461e-05, + "loss": 0.0338, + "step": 40560 + }, + { + "epoch": 0.129856, + "grad_norm": 0.111920551861392, + "learning_rate": 9.953841473164766e-05, + "loss": 0.0341, + "step": 40580 + }, + { + "epoch": 0.12992, + "grad_norm": 0.09707478583199529, + "learning_rate": 9.953796033379573e-05, + "loss": 0.0331, + "step": 40600 + }, + { + "epoch": 0.129984, + "grad_norm": 0.14714173343393702, + "learning_rate": 9.953750571343081e-05, + "loss": 0.0308, + "step": 40620 + }, + { + "epoch": 0.130048, + "grad_norm": 0.0727203321767133, + "learning_rate": 9.953705087055501e-05, + "loss": 0.0311, + "step": 40640 + }, + { + "epoch": 0.130112, + "grad_norm": 0.10667970624858604, + "learning_rate": 9.953659580517031e-05, + "loss": 0.0318, + "step": 40660 + }, + { + "epoch": 0.130176, + "grad_norm": 0.20298499051017452, + "learning_rate": 9.953614051727882e-05, + "loss": 0.0373, + "step": 40680 + }, + { + "epoch": 0.13024, + "grad_norm": 0.07999839292014595, + "learning_rate": 9.953568500688252e-05, + "loss": 0.0355, + "step": 40700 + }, + { + "epoch": 0.130304, + "grad_norm": 0.14012249431879975, + "learning_rate": 9.95352292739835e-05, + "loss": 0.0317, + "step": 40720 + }, + { + "epoch": 0.130368, + "grad_norm": 0.06263024342071373, + "learning_rate": 9.953477331858377e-05, + "loss": 0.0332, + "step": 40740 + }, + { + "epoch": 0.130432, + "grad_norm": 0.10297498151540073, + "learning_rate": 9.95343171406854e-05, + "loss": 0.0306, + "step": 40760 + }, + { + "epoch": 0.130496, + "grad_norm": 0.07485890939866761, + "learning_rate": 9.953386074029045e-05, + "loss": 0.0274, + "step": 40780 + }, + { + "epoch": 0.13056, + "grad_norm": 0.07429640946719845, + "learning_rate": 9.953340411740096e-05, + "loss": 0.0275, + "step": 40800 + }, + { + "epoch": 0.130624, + "grad_norm": 0.23583852939739408, + "learning_rate": 9.953294727201896e-05, + "loss": 0.0333, + "step": 40820 + }, + { + "epoch": 0.130688, + "grad_norm": 0.08508782641201994, + "learning_rate": 9.953249020414654e-05, + "loss": 0.033, + "step": 40840 + }, + { + "epoch": 0.130752, + "grad_norm": 0.17440456206079435, + "learning_rate": 9.953203291378572e-05, + "loss": 0.0336, + "step": 40860 + }, + { + "epoch": 0.130816, + "grad_norm": 0.06961395672795713, + "learning_rate": 9.953157540093859e-05, + "loss": 0.0334, + "step": 40880 + }, + { + "epoch": 0.13088, + "grad_norm": 0.05950703055074971, + "learning_rate": 9.953111766560716e-05, + "loss": 0.0335, + "step": 40900 + }, + { + "epoch": 0.130944, + "grad_norm": 0.18999023219790784, + "learning_rate": 9.953065970779351e-05, + "loss": 0.0322, + "step": 40920 + }, + { + "epoch": 0.131008, + "grad_norm": 0.06639355533746925, + "learning_rate": 9.95302015274997e-05, + "loss": 0.0291, + "step": 40940 + }, + { + "epoch": 0.131072, + "grad_norm": 0.08358978434273276, + "learning_rate": 9.95297431247278e-05, + "loss": 0.032, + "step": 40960 + }, + { + "epoch": 0.131136, + "grad_norm": 0.07943379235271977, + "learning_rate": 9.952928449947983e-05, + "loss": 0.0334, + "step": 40980 + }, + { + "epoch": 0.1312, + "grad_norm": 0.06806733497470308, + "learning_rate": 9.952882565175789e-05, + "loss": 0.0322, + "step": 41000 + }, + { + "epoch": 0.131264, + "grad_norm": 0.0648665308452779, + "learning_rate": 9.952836658156401e-05, + "loss": 0.0324, + "step": 41020 + }, + { + "epoch": 0.131328, + "grad_norm": 0.07002203223405568, + "learning_rate": 9.952790728890027e-05, + "loss": 0.029, + "step": 41040 + }, + { + "epoch": 0.131392, + "grad_norm": 0.0748918539643221, + "learning_rate": 9.952744777376873e-05, + "loss": 0.032, + "step": 41060 + }, + { + "epoch": 0.131456, + "grad_norm": 0.09238244641923951, + "learning_rate": 9.952698803617147e-05, + "loss": 0.0323, + "step": 41080 + }, + { + "epoch": 0.13152, + "grad_norm": 0.062078471827953834, + "learning_rate": 9.952652807611052e-05, + "loss": 0.0312, + "step": 41100 + }, + { + "epoch": 0.131584, + "grad_norm": 0.08770868833681225, + "learning_rate": 9.952606789358796e-05, + "loss": 0.0304, + "step": 41120 + }, + { + "epoch": 0.131648, + "grad_norm": 0.10807909703136269, + "learning_rate": 9.952560748860587e-05, + "loss": 0.031, + "step": 41140 + }, + { + "epoch": 0.131712, + "grad_norm": 0.06724600260043818, + "learning_rate": 9.95251468611663e-05, + "loss": 0.0326, + "step": 41160 + }, + { + "epoch": 0.131776, + "grad_norm": 0.08158097290400615, + "learning_rate": 9.952468601127134e-05, + "loss": 0.0322, + "step": 41180 + }, + { + "epoch": 0.13184, + "grad_norm": 0.07752336067211266, + "learning_rate": 9.952422493892304e-05, + "loss": 0.0324, + "step": 41200 + }, + { + "epoch": 0.131904, + "grad_norm": 0.11111410488218725, + "learning_rate": 9.952376364412348e-05, + "loss": 0.0326, + "step": 41220 + }, + { + "epoch": 0.131968, + "grad_norm": 0.07944285509125448, + "learning_rate": 9.952330212687472e-05, + "loss": 0.0308, + "step": 41240 + }, + { + "epoch": 0.132032, + "grad_norm": 0.08070038973666847, + "learning_rate": 9.952284038717887e-05, + "loss": 0.0303, + "step": 41260 + }, + { + "epoch": 0.132096, + "grad_norm": 0.131691533058166, + "learning_rate": 9.952237842503798e-05, + "loss": 0.0331, + "step": 41280 + }, + { + "epoch": 0.13216, + "grad_norm": 0.22978648246081593, + "learning_rate": 9.952191624045411e-05, + "loss": 0.032, + "step": 41300 + }, + { + "epoch": 0.132224, + "grad_norm": 0.0861904800274661, + "learning_rate": 9.952145383342934e-05, + "loss": 0.032, + "step": 41320 + }, + { + "epoch": 0.132288, + "grad_norm": 0.07280841264117215, + "learning_rate": 9.952099120396576e-05, + "loss": 0.0322, + "step": 41340 + }, + { + "epoch": 0.132352, + "grad_norm": 0.07048205299628756, + "learning_rate": 9.952052835206546e-05, + "loss": 0.0294, + "step": 41360 + }, + { + "epoch": 0.132416, + "grad_norm": 0.0997919809083768, + "learning_rate": 9.95200652777305e-05, + "loss": 0.0334, + "step": 41380 + }, + { + "epoch": 0.13248, + "grad_norm": 0.07144235842499633, + "learning_rate": 9.951960198096296e-05, + "loss": 0.0313, + "step": 41400 + }, + { + "epoch": 0.132544, + "grad_norm": 0.08824468736820908, + "learning_rate": 9.951913846176492e-05, + "loss": 0.0342, + "step": 41420 + }, + { + "epoch": 0.132608, + "grad_norm": 0.060014634858746425, + "learning_rate": 9.951867472013848e-05, + "loss": 0.0321, + "step": 41440 + }, + { + "epoch": 0.132672, + "grad_norm": 0.12746693409814322, + "learning_rate": 9.95182107560857e-05, + "loss": 0.0303, + "step": 41460 + }, + { + "epoch": 0.132736, + "grad_norm": 0.1538932321892048, + "learning_rate": 9.951774656960865e-05, + "loss": 0.0298, + "step": 41480 + }, + { + "epoch": 0.1328, + "grad_norm": 0.12397825274647074, + "learning_rate": 9.951728216070949e-05, + "loss": 0.0329, + "step": 41500 + }, + { + "epoch": 0.132864, + "grad_norm": 0.06783913794611994, + "learning_rate": 9.951681752939021e-05, + "loss": 0.035, + "step": 41520 + }, + { + "epoch": 0.132928, + "grad_norm": 0.0702825387893834, + "learning_rate": 9.951635267565295e-05, + "loss": 0.029, + "step": 41540 + }, + { + "epoch": 0.132992, + "grad_norm": 0.07343306539410793, + "learning_rate": 9.95158875994998e-05, + "loss": 0.0308, + "step": 41560 + }, + { + "epoch": 0.133056, + "grad_norm": 0.07584369111465726, + "learning_rate": 9.951542230093284e-05, + "loss": 0.032, + "step": 41580 + }, + { + "epoch": 0.13312, + "grad_norm": 0.08880768066946325, + "learning_rate": 9.951495677995414e-05, + "loss": 0.0314, + "step": 41600 + }, + { + "epoch": 0.133184, + "grad_norm": 0.16975034326910918, + "learning_rate": 9.951449103656584e-05, + "loss": 0.0346, + "step": 41620 + }, + { + "epoch": 0.133248, + "grad_norm": 0.0577467205446295, + "learning_rate": 9.951402507076999e-05, + "loss": 0.0317, + "step": 41640 + }, + { + "epoch": 0.133312, + "grad_norm": 0.07589329179612517, + "learning_rate": 9.951355888256869e-05, + "loss": 0.0296, + "step": 41660 + }, + { + "epoch": 0.133376, + "grad_norm": 0.06336742000894736, + "learning_rate": 9.951309247196404e-05, + "loss": 0.0281, + "step": 41680 + }, + { + "epoch": 0.13344, + "grad_norm": 0.09963940332126157, + "learning_rate": 9.951262583895812e-05, + "loss": 0.033, + "step": 41700 + }, + { + "epoch": 0.133504, + "grad_norm": 0.06087502563937401, + "learning_rate": 9.951215898355307e-05, + "loss": 0.0314, + "step": 41720 + }, + { + "epoch": 0.133568, + "grad_norm": 0.09783276673902244, + "learning_rate": 9.951169190575093e-05, + "loss": 0.0288, + "step": 41740 + }, + { + "epoch": 0.133632, + "grad_norm": 0.05466824878409184, + "learning_rate": 9.951122460555384e-05, + "loss": 0.0318, + "step": 41760 + }, + { + "epoch": 0.133696, + "grad_norm": 0.10075859099411093, + "learning_rate": 9.95107570829639e-05, + "loss": 0.0316, + "step": 41780 + }, + { + "epoch": 0.13376, + "grad_norm": 0.09833902692780028, + "learning_rate": 9.951028933798317e-05, + "loss": 0.0333, + "step": 41800 + }, + { + "epoch": 0.133824, + "grad_norm": 0.22321415562834282, + "learning_rate": 9.950982137061376e-05, + "loss": 0.0339, + "step": 41820 + }, + { + "epoch": 0.133888, + "grad_norm": 0.16990104992311894, + "learning_rate": 9.950935318085781e-05, + "loss": 0.0309, + "step": 41840 + }, + { + "epoch": 0.133952, + "grad_norm": 0.07855564849265495, + "learning_rate": 9.950888476871741e-05, + "loss": 0.0312, + "step": 41860 + }, + { + "epoch": 0.134016, + "grad_norm": 0.07613221369596639, + "learning_rate": 9.950841613419463e-05, + "loss": 0.0296, + "step": 41880 + }, + { + "epoch": 0.13408, + "grad_norm": 0.07471296071565553, + "learning_rate": 9.95079472772916e-05, + "loss": 0.034, + "step": 41900 + }, + { + "epoch": 0.134144, + "grad_norm": 0.1902086616829644, + "learning_rate": 9.950747819801043e-05, + "loss": 0.0325, + "step": 41920 + }, + { + "epoch": 0.134208, + "grad_norm": 0.08673947918868105, + "learning_rate": 9.950700889635323e-05, + "loss": 0.032, + "step": 41940 + }, + { + "epoch": 0.134272, + "grad_norm": 0.0768414924928411, + "learning_rate": 9.95065393723221e-05, + "loss": 0.0282, + "step": 41960 + }, + { + "epoch": 0.134336, + "grad_norm": 0.12722566753794942, + "learning_rate": 9.950606962591913e-05, + "loss": 0.0308, + "step": 41980 + }, + { + "epoch": 0.1344, + "grad_norm": 0.07969353225275928, + "learning_rate": 9.950559965714648e-05, + "loss": 0.033, + "step": 42000 + }, + { + "epoch": 0.134464, + "grad_norm": 0.06512861153706287, + "learning_rate": 9.950512946600621e-05, + "loss": 0.0318, + "step": 42020 + }, + { + "epoch": 0.134528, + "grad_norm": 0.08651996289082056, + "learning_rate": 9.950465905250044e-05, + "loss": 0.0316, + "step": 42040 + }, + { + "epoch": 0.134592, + "grad_norm": 0.06852091455826752, + "learning_rate": 9.950418841663131e-05, + "loss": 0.0315, + "step": 42060 + }, + { + "epoch": 0.134656, + "grad_norm": 0.07486769850043155, + "learning_rate": 9.950371755840091e-05, + "loss": 0.0326, + "step": 42080 + }, + { + "epoch": 0.13472, + "grad_norm": 0.10014412546859712, + "learning_rate": 9.950324647781137e-05, + "loss": 0.0268, + "step": 42100 + }, + { + "epoch": 0.134784, + "grad_norm": 0.07264300109446775, + "learning_rate": 9.950277517486482e-05, + "loss": 0.0302, + "step": 42120 + }, + { + "epoch": 0.134848, + "grad_norm": 0.0937866211448371, + "learning_rate": 9.950230364956332e-05, + "loss": 0.0349, + "step": 42140 + }, + { + "epoch": 0.134912, + "grad_norm": 0.13432973486935076, + "learning_rate": 9.950183190190905e-05, + "loss": 0.0311, + "step": 42160 + }, + { + "epoch": 0.134976, + "grad_norm": 0.08107620678853775, + "learning_rate": 9.950135993190409e-05, + "loss": 0.0327, + "step": 42180 + }, + { + "epoch": 0.13504, + "grad_norm": 0.06787385751745507, + "learning_rate": 9.950088773955058e-05, + "loss": 0.0295, + "step": 42200 + }, + { + "epoch": 0.135104, + "grad_norm": 0.12238813827692771, + "learning_rate": 9.950041532485063e-05, + "loss": 0.0313, + "step": 42220 + }, + { + "epoch": 0.135168, + "grad_norm": 0.20826547034533952, + "learning_rate": 9.949994268780639e-05, + "loss": 0.0315, + "step": 42240 + }, + { + "epoch": 0.135232, + "grad_norm": 0.07914749245068702, + "learning_rate": 9.949946982841992e-05, + "loss": 0.0345, + "step": 42260 + }, + { + "epoch": 0.135296, + "grad_norm": 0.1257478787765574, + "learning_rate": 9.94989967466934e-05, + "loss": 0.0362, + "step": 42280 + }, + { + "epoch": 0.13536, + "grad_norm": 0.08219629694476246, + "learning_rate": 9.949852344262895e-05, + "loss": 0.0355, + "step": 42300 + }, + { + "epoch": 0.135424, + "grad_norm": 0.13584094464604132, + "learning_rate": 9.949804991622868e-05, + "loss": 0.0314, + "step": 42320 + }, + { + "epoch": 0.135488, + "grad_norm": 0.22936894439439323, + "learning_rate": 9.949757616749471e-05, + "loss": 0.0306, + "step": 42340 + }, + { + "epoch": 0.135552, + "grad_norm": 0.12505019890260918, + "learning_rate": 9.949710219642918e-05, + "loss": 0.0305, + "step": 42360 + }, + { + "epoch": 0.135616, + "grad_norm": 0.08010021115828088, + "learning_rate": 9.949662800303423e-05, + "loss": 0.0325, + "step": 42380 + }, + { + "epoch": 0.13568, + "grad_norm": 0.07843595402995292, + "learning_rate": 9.949615358731197e-05, + "loss": 0.0373, + "step": 42400 + }, + { + "epoch": 0.135744, + "grad_norm": 0.11402927426508497, + "learning_rate": 9.949567894926455e-05, + "loss": 0.0363, + "step": 42420 + }, + { + "epoch": 0.135808, + "grad_norm": 0.2322388164179178, + "learning_rate": 9.949520408889408e-05, + "loss": 0.0353, + "step": 42440 + }, + { + "epoch": 0.135872, + "grad_norm": 0.12328157280407359, + "learning_rate": 9.949472900620271e-05, + "loss": 0.0331, + "step": 42460 + }, + { + "epoch": 0.135936, + "grad_norm": 0.07500274102638292, + "learning_rate": 9.949425370119256e-05, + "loss": 0.032, + "step": 42480 + }, + { + "epoch": 0.136, + "grad_norm": 0.09184218304185483, + "learning_rate": 9.949377817386579e-05, + "loss": 0.031, + "step": 42500 + }, + { + "epoch": 0.136064, + "grad_norm": 0.12437866726968963, + "learning_rate": 9.949330242422452e-05, + "loss": 0.0295, + "step": 42520 + }, + { + "epoch": 0.136128, + "grad_norm": 0.07051187846792681, + "learning_rate": 9.949282645227087e-05, + "loss": 0.0315, + "step": 42540 + }, + { + "epoch": 0.136192, + "grad_norm": 0.07817752222580296, + "learning_rate": 9.9492350258007e-05, + "loss": 0.0315, + "step": 42560 + }, + { + "epoch": 0.136256, + "grad_norm": 0.0769598069881319, + "learning_rate": 9.949187384143503e-05, + "loss": 0.0347, + "step": 42580 + }, + { + "epoch": 0.13632, + "grad_norm": 0.07170667651901967, + "learning_rate": 9.949139720255714e-05, + "loss": 0.0329, + "step": 42600 + }, + { + "epoch": 0.136384, + "grad_norm": 0.10794071680158461, + "learning_rate": 9.949092034137541e-05, + "loss": 0.0347, + "step": 42620 + }, + { + "epoch": 0.136448, + "grad_norm": 0.07681502070751192, + "learning_rate": 9.949044325789206e-05, + "loss": 0.031, + "step": 42640 + }, + { + "epoch": 0.136512, + "grad_norm": 0.09217239304777364, + "learning_rate": 9.948996595210915e-05, + "loss": 0.0312, + "step": 42660 + }, + { + "epoch": 0.136576, + "grad_norm": 0.10853603947692773, + "learning_rate": 9.948948842402888e-05, + "loss": 0.0313, + "step": 42680 + }, + { + "epoch": 0.13664, + "grad_norm": 0.08103549644473462, + "learning_rate": 9.948901067365338e-05, + "loss": 0.0316, + "step": 42700 + }, + { + "epoch": 0.136704, + "grad_norm": 0.12297280074658411, + "learning_rate": 9.948853270098477e-05, + "loss": 0.0323, + "step": 42720 + }, + { + "epoch": 0.136768, + "grad_norm": 0.061149836124419606, + "learning_rate": 9.948805450602526e-05, + "loss": 0.0303, + "step": 42740 + }, + { + "epoch": 0.136832, + "grad_norm": 0.1085419150241325, + "learning_rate": 9.948757608877692e-05, + "loss": 0.031, + "step": 42760 + }, + { + "epoch": 0.136896, + "grad_norm": 0.06263062224352771, + "learning_rate": 9.948709744924196e-05, + "loss": 0.0307, + "step": 42780 + }, + { + "epoch": 0.13696, + "grad_norm": 0.10894210343509753, + "learning_rate": 9.94866185874225e-05, + "loss": 0.0284, + "step": 42800 + }, + { + "epoch": 0.137024, + "grad_norm": 0.1591199729920336, + "learning_rate": 9.948613950332071e-05, + "loss": 0.0345, + "step": 42820 + }, + { + "epoch": 0.137088, + "grad_norm": 0.07291582529499462, + "learning_rate": 9.948566019693872e-05, + "loss": 0.0288, + "step": 42840 + }, + { + "epoch": 0.137152, + "grad_norm": 0.14259247956246893, + "learning_rate": 9.948518066827868e-05, + "loss": 0.0294, + "step": 42860 + }, + { + "epoch": 0.137216, + "grad_norm": 0.08526659426247758, + "learning_rate": 9.948470091734277e-05, + "loss": 0.0312, + "step": 42880 + }, + { + "epoch": 0.13728, + "grad_norm": 0.08398769815070868, + "learning_rate": 9.948422094413315e-05, + "loss": 0.0322, + "step": 42900 + }, + { + "epoch": 0.137344, + "grad_norm": 0.1005253738831398, + "learning_rate": 9.948374074865192e-05, + "loss": 0.0321, + "step": 42920 + }, + { + "epoch": 0.137408, + "grad_norm": 0.13419342365947656, + "learning_rate": 9.94832603309013e-05, + "loss": 0.0324, + "step": 42940 + }, + { + "epoch": 0.137472, + "grad_norm": 0.44754006021674575, + "learning_rate": 9.948277969088342e-05, + "loss": 0.0335, + "step": 42960 + }, + { + "epoch": 0.137536, + "grad_norm": 0.07416001988868091, + "learning_rate": 9.948229882860044e-05, + "loss": 0.0321, + "step": 42980 + }, + { + "epoch": 0.1376, + "grad_norm": 0.09575406426181156, + "learning_rate": 9.948181774405452e-05, + "loss": 0.0328, + "step": 43000 + }, + { + "epoch": 0.137664, + "grad_norm": 0.10080818572541332, + "learning_rate": 9.948133643724782e-05, + "loss": 0.0317, + "step": 43020 + }, + { + "epoch": 0.137728, + "grad_norm": 0.085039641152275, + "learning_rate": 9.948085490818251e-05, + "loss": 0.0322, + "step": 43040 + }, + { + "epoch": 0.137792, + "grad_norm": 0.07537975820334342, + "learning_rate": 9.948037315686075e-05, + "loss": 0.0306, + "step": 43060 + }, + { + "epoch": 0.137856, + "grad_norm": 0.08469853162432425, + "learning_rate": 9.947989118328471e-05, + "loss": 0.0339, + "step": 43080 + }, + { + "epoch": 0.13792, + "grad_norm": 0.13627252526213693, + "learning_rate": 9.947940898745652e-05, + "loss": 0.0287, + "step": 43100 + }, + { + "epoch": 0.137984, + "grad_norm": 0.0935428958130107, + "learning_rate": 9.947892656937838e-05, + "loss": 0.0311, + "step": 43120 + }, + { + "epoch": 0.138048, + "grad_norm": 0.16201928296848297, + "learning_rate": 9.947844392905245e-05, + "loss": 0.0304, + "step": 43140 + }, + { + "epoch": 0.138112, + "grad_norm": 0.07634665705009923, + "learning_rate": 9.947796106648092e-05, + "loss": 0.031, + "step": 43160 + }, + { + "epoch": 0.138176, + "grad_norm": 0.18404840923906227, + "learning_rate": 9.947747798166591e-05, + "loss": 0.0334, + "step": 43180 + }, + { + "epoch": 0.13824, + "grad_norm": 0.06627281063116525, + "learning_rate": 9.947699467460963e-05, + "loss": 0.0324, + "step": 43200 + }, + { + "epoch": 0.138304, + "grad_norm": 0.07097173157617676, + "learning_rate": 9.947651114531422e-05, + "loss": 0.0324, + "step": 43220 + }, + { + "epoch": 0.138368, + "grad_norm": 0.07695453020839546, + "learning_rate": 9.947602739378188e-05, + "loss": 0.0347, + "step": 43240 + }, + { + "epoch": 0.138432, + "grad_norm": 0.07271507658249661, + "learning_rate": 9.947554342001478e-05, + "loss": 0.0326, + "step": 43260 + }, + { + "epoch": 0.138496, + "grad_norm": 0.0907294362769514, + "learning_rate": 9.947505922401507e-05, + "loss": 0.0317, + "step": 43280 + }, + { + "epoch": 0.13856, + "grad_norm": 0.11702148868502561, + "learning_rate": 9.947457480578495e-05, + "loss": 0.0341, + "step": 43300 + }, + { + "epoch": 0.138624, + "grad_norm": 0.07092002802625459, + "learning_rate": 9.947409016532658e-05, + "loss": 0.0343, + "step": 43320 + }, + { + "epoch": 0.138688, + "grad_norm": 0.09559150188499406, + "learning_rate": 9.947360530264214e-05, + "loss": 0.0307, + "step": 43340 + }, + { + "epoch": 0.138752, + "grad_norm": 0.07575799077733586, + "learning_rate": 9.947312021773383e-05, + "loss": 0.0333, + "step": 43360 + }, + { + "epoch": 0.138816, + "grad_norm": 0.07991725391255566, + "learning_rate": 9.947263491060379e-05, + "loss": 0.0321, + "step": 43380 + }, + { + "epoch": 0.13888, + "grad_norm": 0.08392834897280924, + "learning_rate": 9.947214938125422e-05, + "loss": 0.0293, + "step": 43400 + }, + { + "epoch": 0.138944, + "grad_norm": 0.07143880189504376, + "learning_rate": 9.947166362968732e-05, + "loss": 0.031, + "step": 43420 + }, + { + "epoch": 0.139008, + "grad_norm": 0.05710903095031855, + "learning_rate": 9.947117765590523e-05, + "loss": 0.0285, + "step": 43440 + }, + { + "epoch": 0.139072, + "grad_norm": 0.11352949324055561, + "learning_rate": 9.947069145991017e-05, + "loss": 0.0311, + "step": 43460 + }, + { + "epoch": 0.139136, + "grad_norm": 0.13588474411142884, + "learning_rate": 9.94702050417043e-05, + "loss": 0.0306, + "step": 43480 + }, + { + "epoch": 0.1392, + "grad_norm": 0.1054201025936334, + "learning_rate": 9.946971840128981e-05, + "loss": 0.0321, + "step": 43500 + }, + { + "epoch": 0.139264, + "grad_norm": 0.08933119637963963, + "learning_rate": 9.94692315386689e-05, + "loss": 0.0341, + "step": 43520 + }, + { + "epoch": 0.139328, + "grad_norm": 0.09625817735827619, + "learning_rate": 9.946874445384375e-05, + "loss": 0.0332, + "step": 43540 + }, + { + "epoch": 0.139392, + "grad_norm": 0.14484393338434143, + "learning_rate": 9.946825714681654e-05, + "loss": 0.0325, + "step": 43560 + }, + { + "epoch": 0.139456, + "grad_norm": 0.0683339596166035, + "learning_rate": 9.946776961758945e-05, + "loss": 0.03, + "step": 43580 + }, + { + "epoch": 0.13952, + "grad_norm": 0.10177336821474872, + "learning_rate": 9.94672818661647e-05, + "loss": 0.0306, + "step": 43600 + }, + { + "epoch": 0.139584, + "grad_norm": 0.07553564513646266, + "learning_rate": 9.946679389254448e-05, + "loss": 0.0333, + "step": 43620 + }, + { + "epoch": 0.139648, + "grad_norm": 0.06650640996577269, + "learning_rate": 9.946630569673094e-05, + "loss": 0.0302, + "step": 43640 + }, + { + "epoch": 0.139712, + "grad_norm": 0.10896985518469338, + "learning_rate": 9.94658172787263e-05, + "loss": 0.0286, + "step": 43660 + }, + { + "epoch": 0.139776, + "grad_norm": 0.07792751713105323, + "learning_rate": 9.946532863853277e-05, + "loss": 0.0304, + "step": 43680 + }, + { + "epoch": 0.13984, + "grad_norm": 0.08708487014111582, + "learning_rate": 9.94648397761525e-05, + "loss": 0.0311, + "step": 43700 + }, + { + "epoch": 0.139904, + "grad_norm": 0.14665370314964687, + "learning_rate": 9.946435069158774e-05, + "loss": 0.0284, + "step": 43720 + }, + { + "epoch": 0.139968, + "grad_norm": 0.13126047953768288, + "learning_rate": 9.946386138484065e-05, + "loss": 0.0326, + "step": 43740 + }, + { + "epoch": 0.140032, + "grad_norm": 0.1667786189845606, + "learning_rate": 9.946337185591343e-05, + "loss": 0.0319, + "step": 43760 + }, + { + "epoch": 0.140096, + "grad_norm": 0.08203177439237458, + "learning_rate": 9.94628821048083e-05, + "loss": 0.0319, + "step": 43780 + }, + { + "epoch": 0.14016, + "grad_norm": 0.0840635672034877, + "learning_rate": 9.946239213152745e-05, + "loss": 0.0343, + "step": 43800 + }, + { + "epoch": 0.140224, + "grad_norm": 0.19184332980124338, + "learning_rate": 9.946190193607307e-05, + "loss": 0.0359, + "step": 43820 + }, + { + "epoch": 0.140288, + "grad_norm": 0.08765050509586234, + "learning_rate": 9.946141151844737e-05, + "loss": 0.0325, + "step": 43840 + }, + { + "epoch": 0.140352, + "grad_norm": 0.09556817931819744, + "learning_rate": 9.946092087865255e-05, + "loss": 0.0296, + "step": 43860 + }, + { + "epoch": 0.140416, + "grad_norm": 0.08672951031920267, + "learning_rate": 9.946043001669083e-05, + "loss": 0.0324, + "step": 43880 + }, + { + "epoch": 0.14048, + "grad_norm": 0.07298977539915834, + "learning_rate": 9.945993893256439e-05, + "loss": 0.0338, + "step": 43900 + }, + { + "epoch": 0.140544, + "grad_norm": 0.17233156618349463, + "learning_rate": 9.945944762627545e-05, + "loss": 0.034, + "step": 43920 + }, + { + "epoch": 0.140608, + "grad_norm": 0.06207802320170917, + "learning_rate": 9.94589560978262e-05, + "loss": 0.0327, + "step": 43940 + }, + { + "epoch": 0.140672, + "grad_norm": 0.09764171518154835, + "learning_rate": 9.945846434721889e-05, + "loss": 0.031, + "step": 43960 + }, + { + "epoch": 0.140736, + "grad_norm": 0.10636364453644002, + "learning_rate": 9.945797237445567e-05, + "loss": 0.0342, + "step": 43980 + }, + { + "epoch": 0.1408, + "grad_norm": 0.06716676468685795, + "learning_rate": 9.94574801795388e-05, + "loss": 0.0356, + "step": 44000 + }, + { + "epoch": 0.140864, + "grad_norm": 0.1122382114016688, + "learning_rate": 9.945698776247048e-05, + "loss": 0.034, + "step": 44020 + }, + { + "epoch": 0.140928, + "grad_norm": 0.0839185343203512, + "learning_rate": 9.94564951232529e-05, + "loss": 0.0321, + "step": 44040 + }, + { + "epoch": 0.140992, + "grad_norm": 0.06913776514673886, + "learning_rate": 9.945600226188827e-05, + "loss": 0.0299, + "step": 44060 + }, + { + "epoch": 0.141056, + "grad_norm": 0.13219850310075404, + "learning_rate": 9.945550917837885e-05, + "loss": 0.0307, + "step": 44080 + }, + { + "epoch": 0.14112, + "grad_norm": 0.11009348776955531, + "learning_rate": 9.94550158727268e-05, + "loss": 0.0329, + "step": 44100 + }, + { + "epoch": 0.141184, + "grad_norm": 0.11465943285226894, + "learning_rate": 9.945452234493437e-05, + "loss": 0.0289, + "step": 44120 + }, + { + "epoch": 0.141248, + "grad_norm": 0.08245215150787376, + "learning_rate": 9.945402859500377e-05, + "loss": 0.0292, + "step": 44140 + }, + { + "epoch": 0.141312, + "grad_norm": 0.061509068599355635, + "learning_rate": 9.94535346229372e-05, + "loss": 0.0329, + "step": 44160 + }, + { + "epoch": 0.141376, + "grad_norm": 0.05339866181914807, + "learning_rate": 9.945304042873691e-05, + "loss": 0.0308, + "step": 44180 + }, + { + "epoch": 0.14144, + "grad_norm": 0.06868911322842283, + "learning_rate": 9.94525460124051e-05, + "loss": 0.0311, + "step": 44200 + }, + { + "epoch": 0.141504, + "grad_norm": 0.0676701198001911, + "learning_rate": 9.9452051373944e-05, + "loss": 0.0325, + "step": 44220 + }, + { + "epoch": 0.141568, + "grad_norm": 0.09804076495766727, + "learning_rate": 9.945155651335583e-05, + "loss": 0.0295, + "step": 44240 + }, + { + "epoch": 0.141632, + "grad_norm": 0.11437173867529303, + "learning_rate": 9.945106143064279e-05, + "loss": 0.0335, + "step": 44260 + }, + { + "epoch": 0.141696, + "grad_norm": 0.07728718845937241, + "learning_rate": 9.945056612580713e-05, + "loss": 0.0315, + "step": 44280 + }, + { + "epoch": 0.14176, + "grad_norm": 0.08130251469449006, + "learning_rate": 9.945007059885107e-05, + "loss": 0.0334, + "step": 44300 + }, + { + "epoch": 0.141824, + "grad_norm": 0.08075980284946264, + "learning_rate": 9.944957484977683e-05, + "loss": 0.031, + "step": 44320 + }, + { + "epoch": 0.141888, + "grad_norm": 0.23518714080263184, + "learning_rate": 9.944907887858665e-05, + "loss": 0.0319, + "step": 44340 + }, + { + "epoch": 0.141952, + "grad_norm": 0.11053678618225929, + "learning_rate": 9.944858268528274e-05, + "loss": 0.0365, + "step": 44360 + }, + { + "epoch": 0.142016, + "grad_norm": 0.08609581400036871, + "learning_rate": 9.944808626986734e-05, + "loss": 0.032, + "step": 44380 + }, + { + "epoch": 0.14208, + "grad_norm": 0.09896133451777779, + "learning_rate": 9.944758963234266e-05, + "loss": 0.0319, + "step": 44400 + }, + { + "epoch": 0.142144, + "grad_norm": 0.11283720492191336, + "learning_rate": 9.944709277271097e-05, + "loss": 0.0344, + "step": 44420 + }, + { + "epoch": 0.142208, + "grad_norm": 0.22978125442311853, + "learning_rate": 9.944659569097448e-05, + "loss": 0.032, + "step": 44440 + }, + { + "epoch": 0.142272, + "grad_norm": 0.1315335767759104, + "learning_rate": 9.944609838713542e-05, + "loss": 0.0308, + "step": 44460 + }, + { + "epoch": 0.142336, + "grad_norm": 0.09714272752372957, + "learning_rate": 9.944560086119601e-05, + "loss": 0.0363, + "step": 44480 + }, + { + "epoch": 0.1424, + "grad_norm": 0.09227809766663257, + "learning_rate": 9.94451031131585e-05, + "loss": 0.0347, + "step": 44500 + }, + { + "epoch": 0.142464, + "grad_norm": 0.10262969948075865, + "learning_rate": 9.944460514302513e-05, + "loss": 0.0298, + "step": 44520 + }, + { + "epoch": 0.142528, + "grad_norm": 0.07339862831550231, + "learning_rate": 9.944410695079814e-05, + "loss": 0.0311, + "step": 44540 + }, + { + "epoch": 0.142592, + "grad_norm": 0.09771678754754455, + "learning_rate": 9.944360853647976e-05, + "loss": 0.0303, + "step": 44560 + }, + { + "epoch": 0.142656, + "grad_norm": 0.06852349464050257, + "learning_rate": 9.944310990007223e-05, + "loss": 0.0316, + "step": 44580 + }, + { + "epoch": 0.14272, + "grad_norm": 0.13934347155132817, + "learning_rate": 9.944261104157778e-05, + "loss": 0.0296, + "step": 44600 + }, + { + "epoch": 0.142784, + "grad_norm": 0.13306207936240774, + "learning_rate": 9.944211196099866e-05, + "loss": 0.03, + "step": 44620 + }, + { + "epoch": 0.142848, + "grad_norm": 0.12148831245370686, + "learning_rate": 9.944161265833713e-05, + "loss": 0.0296, + "step": 44640 + }, + { + "epoch": 0.142912, + "grad_norm": 0.09935488097025678, + "learning_rate": 9.94411131335954e-05, + "loss": 0.0305, + "step": 44660 + }, + { + "epoch": 0.142976, + "grad_norm": 0.0839528953603813, + "learning_rate": 9.944061338677572e-05, + "loss": 0.0319, + "step": 44680 + }, + { + "epoch": 0.14304, + "grad_norm": 0.07255629713976992, + "learning_rate": 9.944011341788036e-05, + "loss": 0.0332, + "step": 44700 + }, + { + "epoch": 0.143104, + "grad_norm": 0.08208459227085675, + "learning_rate": 9.943961322691156e-05, + "loss": 0.0305, + "step": 44720 + }, + { + "epoch": 0.143168, + "grad_norm": 0.1130568637514043, + "learning_rate": 9.943911281387152e-05, + "loss": 0.0381, + "step": 44740 + }, + { + "epoch": 0.143232, + "grad_norm": 0.07918837464120848, + "learning_rate": 9.943861217876255e-05, + "loss": 0.0317, + "step": 44760 + }, + { + "epoch": 0.143296, + "grad_norm": 0.07962213892261462, + "learning_rate": 9.943811132158686e-05, + "loss": 0.032, + "step": 44780 + }, + { + "epoch": 0.14336, + "grad_norm": 0.07810614497076983, + "learning_rate": 9.943761024234672e-05, + "loss": 0.0308, + "step": 44800 + }, + { + "epoch": 0.143424, + "grad_norm": 0.08074008935520487, + "learning_rate": 9.943710894104437e-05, + "loss": 0.0277, + "step": 44820 + }, + { + "epoch": 0.143488, + "grad_norm": 0.0784583075106754, + "learning_rate": 9.943660741768206e-05, + "loss": 0.0287, + "step": 44840 + }, + { + "epoch": 0.143552, + "grad_norm": 0.07116579090110624, + "learning_rate": 9.943610567226205e-05, + "loss": 0.0329, + "step": 44860 + }, + { + "epoch": 0.143616, + "grad_norm": 0.06686010862961496, + "learning_rate": 9.943560370478657e-05, + "loss": 0.0317, + "step": 44880 + }, + { + "epoch": 0.14368, + "grad_norm": 0.07582682979300236, + "learning_rate": 9.943510151525792e-05, + "loss": 0.0292, + "step": 44900 + }, + { + "epoch": 0.143744, + "grad_norm": 0.06244356411598463, + "learning_rate": 9.943459910367833e-05, + "loss": 0.0328, + "step": 44920 + }, + { + "epoch": 0.143808, + "grad_norm": 0.09932211287497399, + "learning_rate": 9.943409647005004e-05, + "loss": 0.0322, + "step": 44940 + }, + { + "epoch": 0.143872, + "grad_norm": 0.13436156203484195, + "learning_rate": 9.943359361437534e-05, + "loss": 0.0355, + "step": 44960 + }, + { + "epoch": 0.143936, + "grad_norm": 0.060035223037218664, + "learning_rate": 9.943309053665648e-05, + "loss": 0.0341, + "step": 44980 + }, + { + "epoch": 0.144, + "grad_norm": 0.0735895460859978, + "learning_rate": 9.94325872368957e-05, + "loss": 0.0303, + "step": 45000 + }, + { + "epoch": 0.144064, + "grad_norm": 0.09283087910635777, + "learning_rate": 9.943208371509529e-05, + "loss": 0.0353, + "step": 45020 + }, + { + "epoch": 0.144128, + "grad_norm": 0.06370576805589122, + "learning_rate": 9.943157997125748e-05, + "loss": 0.0329, + "step": 45040 + }, + { + "epoch": 0.144192, + "grad_norm": 0.11249501379345124, + "learning_rate": 9.943107600538457e-05, + "loss": 0.0357, + "step": 45060 + }, + { + "epoch": 0.144256, + "grad_norm": 0.1743569865292881, + "learning_rate": 9.943057181747877e-05, + "loss": 0.0334, + "step": 45080 + }, + { + "epoch": 0.14432, + "grad_norm": 0.15340215685007777, + "learning_rate": 9.94300674075424e-05, + "loss": 0.0314, + "step": 45100 + }, + { + "epoch": 0.144384, + "grad_norm": 0.07734797801592953, + "learning_rate": 9.94295627755777e-05, + "loss": 0.0361, + "step": 45120 + }, + { + "epoch": 0.144448, + "grad_norm": 0.0944424417827727, + "learning_rate": 9.942905792158695e-05, + "loss": 0.0329, + "step": 45140 + }, + { + "epoch": 0.144512, + "grad_norm": 0.18694625052907426, + "learning_rate": 9.942855284557238e-05, + "loss": 0.0329, + "step": 45160 + }, + { + "epoch": 0.144576, + "grad_norm": 0.17912878062783963, + "learning_rate": 9.942804754753631e-05, + "loss": 0.0336, + "step": 45180 + }, + { + "epoch": 0.14464, + "grad_norm": 0.0822804187749451, + "learning_rate": 9.942754202748097e-05, + "loss": 0.033, + "step": 45200 + }, + { + "epoch": 0.144704, + "grad_norm": 0.07199907837023084, + "learning_rate": 9.942703628540866e-05, + "loss": 0.0341, + "step": 45220 + }, + { + "epoch": 0.144768, + "grad_norm": 0.06595484442305903, + "learning_rate": 9.942653032132163e-05, + "loss": 0.0314, + "step": 45240 + }, + { + "epoch": 0.144832, + "grad_norm": 0.21101409344193045, + "learning_rate": 9.942602413522216e-05, + "loss": 0.031, + "step": 45260 + }, + { + "epoch": 0.144896, + "grad_norm": 0.27433541153479357, + "learning_rate": 9.942551772711253e-05, + "loss": 0.0296, + "step": 45280 + }, + { + "epoch": 0.14496, + "grad_norm": 0.10164290483102832, + "learning_rate": 9.9425011096995e-05, + "loss": 0.0302, + "step": 45300 + }, + { + "epoch": 0.145024, + "grad_norm": 0.07756871470530831, + "learning_rate": 9.942450424487187e-05, + "loss": 0.0321, + "step": 45320 + }, + { + "epoch": 0.145088, + "grad_norm": 0.1129968037904023, + "learning_rate": 9.942399717074539e-05, + "loss": 0.0306, + "step": 45340 + }, + { + "epoch": 0.145152, + "grad_norm": 0.09327859722297321, + "learning_rate": 9.942348987461785e-05, + "loss": 0.0305, + "step": 45360 + }, + { + "epoch": 0.145216, + "grad_norm": 0.1590969937858857, + "learning_rate": 9.942298235649152e-05, + "loss": 0.0329, + "step": 45380 + }, + { + "epoch": 0.14528, + "grad_norm": 0.06630503445706937, + "learning_rate": 9.94224746163687e-05, + "loss": 0.0305, + "step": 45400 + }, + { + "epoch": 0.145344, + "grad_norm": 0.11045453900590403, + "learning_rate": 9.942196665425165e-05, + "loss": 0.0317, + "step": 45420 + }, + { + "epoch": 0.145408, + "grad_norm": 0.07262598243246876, + "learning_rate": 9.942145847014268e-05, + "loss": 0.0345, + "step": 45440 + }, + { + "epoch": 0.145472, + "grad_norm": 0.07834783262286689, + "learning_rate": 9.942095006404404e-05, + "loss": 0.0317, + "step": 45460 + }, + { + "epoch": 0.145536, + "grad_norm": 0.07191157466035178, + "learning_rate": 9.942044143595801e-05, + "loss": 0.0346, + "step": 45480 + }, + { + "epoch": 0.1456, + "grad_norm": 0.07668680647600666, + "learning_rate": 9.941993258588691e-05, + "loss": 0.0309, + "step": 45500 + }, + { + "epoch": 0.145664, + "grad_norm": 0.07041920888807483, + "learning_rate": 9.9419423513833e-05, + "loss": 0.0287, + "step": 45520 + }, + { + "epoch": 0.145728, + "grad_norm": 0.11835562972477168, + "learning_rate": 9.941891421979857e-05, + "loss": 0.0318, + "step": 45540 + }, + { + "epoch": 0.145792, + "grad_norm": 0.0937626034968694, + "learning_rate": 9.941840470378592e-05, + "loss": 0.033, + "step": 45560 + }, + { + "epoch": 0.145856, + "grad_norm": 0.09318282553620828, + "learning_rate": 9.941789496579733e-05, + "loss": 0.0315, + "step": 45580 + }, + { + "epoch": 0.14592, + "grad_norm": 0.10281820061906448, + "learning_rate": 9.941738500583506e-05, + "loss": 0.0292, + "step": 45600 + }, + { + "epoch": 0.145984, + "grad_norm": 0.1100995093191855, + "learning_rate": 9.941687482390146e-05, + "loss": 0.0315, + "step": 45620 + }, + { + "epoch": 0.146048, + "grad_norm": 0.08801731855835009, + "learning_rate": 9.941636441999877e-05, + "loss": 0.0296, + "step": 45640 + }, + { + "epoch": 0.146112, + "grad_norm": 0.07973052830645552, + "learning_rate": 9.941585379412932e-05, + "loss": 0.0273, + "step": 45660 + }, + { + "epoch": 0.146176, + "grad_norm": 0.06405342410231209, + "learning_rate": 9.941534294629538e-05, + "loss": 0.0329, + "step": 45680 + }, + { + "epoch": 0.14624, + "grad_norm": 0.07661828366254775, + "learning_rate": 9.941483187649927e-05, + "loss": 0.0302, + "step": 45700 + }, + { + "epoch": 0.146304, + "grad_norm": 0.09846929536015242, + "learning_rate": 9.941432058474325e-05, + "loss": 0.0279, + "step": 45720 + }, + { + "epoch": 0.146368, + "grad_norm": 0.06676787830470808, + "learning_rate": 9.941380907102965e-05, + "loss": 0.0287, + "step": 45740 + }, + { + "epoch": 0.146432, + "grad_norm": 0.10850745215323246, + "learning_rate": 9.941329733536071e-05, + "loss": 0.0298, + "step": 45760 + }, + { + "epoch": 0.146496, + "grad_norm": 0.06857623117633072, + "learning_rate": 9.941278537773881e-05, + "loss": 0.0276, + "step": 45780 + }, + { + "epoch": 0.14656, + "grad_norm": 0.08976718895389513, + "learning_rate": 9.94122731981662e-05, + "loss": 0.0317, + "step": 45800 + }, + { + "epoch": 0.146624, + "grad_norm": 0.08460687992055017, + "learning_rate": 9.941176079664518e-05, + "loss": 0.031, + "step": 45820 + }, + { + "epoch": 0.146688, + "grad_norm": 0.073861298680393, + "learning_rate": 9.941124817317807e-05, + "loss": 0.0287, + "step": 45840 + }, + { + "epoch": 0.146752, + "grad_norm": 0.07962980604798256, + "learning_rate": 9.941073532776717e-05, + "loss": 0.0345, + "step": 45860 + }, + { + "epoch": 0.146816, + "grad_norm": 0.08888390369116214, + "learning_rate": 9.941022226041477e-05, + "loss": 0.0324, + "step": 45880 + }, + { + "epoch": 0.14688, + "grad_norm": 0.06465314487516263, + "learning_rate": 9.940970897112318e-05, + "loss": 0.0302, + "step": 45900 + }, + { + "epoch": 0.146944, + "grad_norm": 0.0852821757519658, + "learning_rate": 9.940919545989473e-05, + "loss": 0.031, + "step": 45920 + }, + { + "epoch": 0.147008, + "grad_norm": 0.06472151014883815, + "learning_rate": 9.940868172673167e-05, + "loss": 0.0321, + "step": 45940 + }, + { + "epoch": 0.147072, + "grad_norm": 0.07322408202590953, + "learning_rate": 9.940816777163637e-05, + "loss": 0.0307, + "step": 45960 + }, + { + "epoch": 0.147136, + "grad_norm": 0.0763459418667312, + "learning_rate": 9.940765359461111e-05, + "loss": 0.0339, + "step": 45980 + }, + { + "epoch": 0.1472, + "grad_norm": 0.0713914196685617, + "learning_rate": 9.940713919565818e-05, + "loss": 0.0345, + "step": 46000 + }, + { + "epoch": 0.147264, + "grad_norm": 0.11324428818708661, + "learning_rate": 9.940662457477993e-05, + "loss": 0.0323, + "step": 46020 + }, + { + "epoch": 0.147328, + "grad_norm": 0.06586571217892605, + "learning_rate": 9.940610973197866e-05, + "loss": 0.0301, + "step": 46040 + }, + { + "epoch": 0.147392, + "grad_norm": 0.1456179123267802, + "learning_rate": 9.940559466725665e-05, + "loss": 0.0333, + "step": 46060 + }, + { + "epoch": 0.147456, + "grad_norm": 0.07725844143023942, + "learning_rate": 9.940507938061624e-05, + "loss": 0.0328, + "step": 46080 + }, + { + "epoch": 0.14752, + "grad_norm": 0.06554538602498296, + "learning_rate": 9.940456387205976e-05, + "loss": 0.0299, + "step": 46100 + }, + { + "epoch": 0.147584, + "grad_norm": 0.15604864495179765, + "learning_rate": 9.940404814158952e-05, + "loss": 0.0318, + "step": 46120 + }, + { + "epoch": 0.147648, + "grad_norm": 0.06390259851170609, + "learning_rate": 9.94035321892078e-05, + "loss": 0.0306, + "step": 46140 + }, + { + "epoch": 0.147712, + "grad_norm": 0.06377724964397978, + "learning_rate": 9.940301601491696e-05, + "loss": 0.0281, + "step": 46160 + }, + { + "epoch": 0.147776, + "grad_norm": 0.14040731774437945, + "learning_rate": 9.940249961871929e-05, + "loss": 0.0311, + "step": 46180 + }, + { + "epoch": 0.14784, + "grad_norm": 0.13744763617337144, + "learning_rate": 9.940198300061712e-05, + "loss": 0.0316, + "step": 46200 + }, + { + "epoch": 0.147904, + "grad_norm": 0.08924213792743513, + "learning_rate": 9.940146616061279e-05, + "loss": 0.034, + "step": 46220 + }, + { + "epoch": 0.147968, + "grad_norm": 0.09632146089942842, + "learning_rate": 9.940094909870859e-05, + "loss": 0.0301, + "step": 46240 + }, + { + "epoch": 0.148032, + "grad_norm": 0.07140229473390827, + "learning_rate": 9.940043181490685e-05, + "loss": 0.032, + "step": 46260 + }, + { + "epoch": 0.148096, + "grad_norm": 0.10996114684599252, + "learning_rate": 9.93999143092099e-05, + "loss": 0.0294, + "step": 46280 + }, + { + "epoch": 0.14816, + "grad_norm": 0.08984108632433775, + "learning_rate": 9.939939658162007e-05, + "loss": 0.0348, + "step": 46300 + }, + { + "epoch": 0.148224, + "grad_norm": 0.07254968355836286, + "learning_rate": 9.939887863213969e-05, + "loss": 0.0356, + "step": 46320 + }, + { + "epoch": 0.148288, + "grad_norm": 0.12003039044495054, + "learning_rate": 9.939836046077104e-05, + "loss": 0.0344, + "step": 46340 + }, + { + "epoch": 0.148352, + "grad_norm": 0.06813999856464828, + "learning_rate": 9.939784206751652e-05, + "loss": 0.0327, + "step": 46360 + }, + { + "epoch": 0.148416, + "grad_norm": 0.16690763985110113, + "learning_rate": 9.93973234523784e-05, + "loss": 0.0322, + "step": 46380 + }, + { + "epoch": 0.14848, + "grad_norm": 0.09827324829997114, + "learning_rate": 9.939680461535905e-05, + "loss": 0.0312, + "step": 46400 + }, + { + "epoch": 0.148544, + "grad_norm": 0.07838333271046842, + "learning_rate": 9.939628555646077e-05, + "loss": 0.0282, + "step": 46420 + }, + { + "epoch": 0.148608, + "grad_norm": 0.09914416861598797, + "learning_rate": 9.93957662756859e-05, + "loss": 0.0363, + "step": 46440 + }, + { + "epoch": 0.148672, + "grad_norm": 0.09648094210648592, + "learning_rate": 9.939524677303677e-05, + "loss": 0.0331, + "step": 46460 + }, + { + "epoch": 0.148736, + "grad_norm": 0.06205894600673664, + "learning_rate": 9.939472704851573e-05, + "loss": 0.0274, + "step": 46480 + }, + { + "epoch": 0.1488, + "grad_norm": 0.13086498230804006, + "learning_rate": 9.939420710212511e-05, + "loss": 0.0273, + "step": 46500 + }, + { + "epoch": 0.148864, + "grad_norm": 0.0936157924584635, + "learning_rate": 9.939368693386724e-05, + "loss": 0.0289, + "step": 46520 + }, + { + "epoch": 0.148928, + "grad_norm": 0.10357413821536994, + "learning_rate": 9.939316654374444e-05, + "loss": 0.032, + "step": 46540 + }, + { + "epoch": 0.148992, + "grad_norm": 0.09739226154550425, + "learning_rate": 9.939264593175908e-05, + "loss": 0.0302, + "step": 46560 + }, + { + "epoch": 0.149056, + "grad_norm": 0.10253538558746574, + "learning_rate": 9.939212509791347e-05, + "loss": 0.03, + "step": 46580 + }, + { + "epoch": 0.14912, + "grad_norm": 0.08714046655221884, + "learning_rate": 9.939160404220997e-05, + "loss": 0.032, + "step": 46600 + }, + { + "epoch": 0.149184, + "grad_norm": 0.07105167641551288, + "learning_rate": 9.93910827646509e-05, + "loss": 0.0313, + "step": 46620 + }, + { + "epoch": 0.149248, + "grad_norm": 0.07769119862566892, + "learning_rate": 9.939056126523861e-05, + "loss": 0.0315, + "step": 46640 + }, + { + "epoch": 0.149312, + "grad_norm": 0.07658669882878325, + "learning_rate": 9.939003954397546e-05, + "loss": 0.0332, + "step": 46660 + }, + { + "epoch": 0.149376, + "grad_norm": 0.06952651886100032, + "learning_rate": 9.938951760086377e-05, + "loss": 0.03, + "step": 46680 + }, + { + "epoch": 0.14944, + "grad_norm": 0.05657640155364597, + "learning_rate": 9.93889954359059e-05, + "loss": 0.0307, + "step": 46700 + }, + { + "epoch": 0.149504, + "grad_norm": 0.060772649234232846, + "learning_rate": 9.938847304910418e-05, + "loss": 0.0313, + "step": 46720 + }, + { + "epoch": 0.149568, + "grad_norm": 0.08657837054906613, + "learning_rate": 9.938795044046097e-05, + "loss": 0.0307, + "step": 46740 + }, + { + "epoch": 0.149632, + "grad_norm": 0.08027933804587406, + "learning_rate": 9.938742760997861e-05, + "loss": 0.0305, + "step": 46760 + }, + { + "epoch": 0.149696, + "grad_norm": 0.09409280808655238, + "learning_rate": 9.938690455765945e-05, + "loss": 0.0313, + "step": 46780 + }, + { + "epoch": 0.14976, + "grad_norm": 0.08509766405976296, + "learning_rate": 9.938638128350585e-05, + "loss": 0.032, + "step": 46800 + }, + { + "epoch": 0.149824, + "grad_norm": 0.05583496773047922, + "learning_rate": 9.938585778752014e-05, + "loss": 0.0312, + "step": 46820 + }, + { + "epoch": 0.149888, + "grad_norm": 0.10010265872088056, + "learning_rate": 9.938533406970469e-05, + "loss": 0.0333, + "step": 46840 + }, + { + "epoch": 0.149952, + "grad_norm": 0.08872981950016327, + "learning_rate": 9.938481013006184e-05, + "loss": 0.0303, + "step": 46860 + }, + { + "epoch": 0.150016, + "grad_norm": 0.10262009964284717, + "learning_rate": 9.938428596859395e-05, + "loss": 0.0335, + "step": 46880 + }, + { + "epoch": 0.15008, + "grad_norm": 0.06717229212581256, + "learning_rate": 9.938376158530337e-05, + "loss": 0.0294, + "step": 46900 + }, + { + "epoch": 0.150144, + "grad_norm": 0.10930912100750657, + "learning_rate": 9.938323698019246e-05, + "loss": 0.0292, + "step": 46920 + }, + { + "epoch": 0.150208, + "grad_norm": 0.12899339472913562, + "learning_rate": 9.938271215326358e-05, + "loss": 0.0316, + "step": 46940 + }, + { + "epoch": 0.150272, + "grad_norm": 0.14791430349800785, + "learning_rate": 9.938218710451907e-05, + "loss": 0.0341, + "step": 46960 + }, + { + "epoch": 0.150336, + "grad_norm": 0.12695112518559196, + "learning_rate": 9.938166183396131e-05, + "loss": 0.0299, + "step": 46980 + }, + { + "epoch": 0.1504, + "grad_norm": 0.10449814218920127, + "learning_rate": 9.938113634159265e-05, + "loss": 0.0297, + "step": 47000 + }, + { + "epoch": 0.150464, + "grad_norm": 0.06885648507316686, + "learning_rate": 9.938061062741544e-05, + "loss": 0.0315, + "step": 47020 + }, + { + "epoch": 0.150528, + "grad_norm": 0.08223492148444912, + "learning_rate": 9.938008469143205e-05, + "loss": 0.0326, + "step": 47040 + }, + { + "epoch": 0.150592, + "grad_norm": 0.14005598580387021, + "learning_rate": 9.937955853364486e-05, + "loss": 0.0328, + "step": 47060 + }, + { + "epoch": 0.150656, + "grad_norm": 0.09102648426699513, + "learning_rate": 9.93790321540562e-05, + "loss": 0.0316, + "step": 47080 + }, + { + "epoch": 0.15072, + "grad_norm": 0.1161245812908459, + "learning_rate": 9.937850555266846e-05, + "loss": 0.031, + "step": 47100 + }, + { + "epoch": 0.150784, + "grad_norm": 0.09258502563134016, + "learning_rate": 9.9377978729484e-05, + "loss": 0.03, + "step": 47120 + }, + { + "epoch": 0.150848, + "grad_norm": 0.09455246032115502, + "learning_rate": 9.937745168450518e-05, + "loss": 0.0297, + "step": 47140 + }, + { + "epoch": 0.150912, + "grad_norm": 0.11034507521592121, + "learning_rate": 9.937692441773437e-05, + "loss": 0.0331, + "step": 47160 + }, + { + "epoch": 0.150976, + "grad_norm": 0.1563965361057462, + "learning_rate": 9.937639692917394e-05, + "loss": 0.034, + "step": 47180 + }, + { + "epoch": 0.15104, + "grad_norm": 0.11750626625630985, + "learning_rate": 9.937586921882626e-05, + "loss": 0.0296, + "step": 47200 + }, + { + "epoch": 0.151104, + "grad_norm": 0.0991318193920632, + "learning_rate": 9.93753412866937e-05, + "loss": 0.0345, + "step": 47220 + }, + { + "epoch": 0.151168, + "grad_norm": 0.10708570931373076, + "learning_rate": 9.937481313277863e-05, + "loss": 0.0313, + "step": 47240 + }, + { + "epoch": 0.151232, + "grad_norm": 0.07659709455960267, + "learning_rate": 9.937428475708342e-05, + "loss": 0.0294, + "step": 47260 + }, + { + "epoch": 0.151296, + "grad_norm": 0.09766081678448985, + "learning_rate": 9.937375615961043e-05, + "loss": 0.0326, + "step": 47280 + }, + { + "epoch": 0.15136, + "grad_norm": 0.05865594002239727, + "learning_rate": 9.937322734036207e-05, + "loss": 0.0316, + "step": 47300 + }, + { + "epoch": 0.151424, + "grad_norm": 0.11250546550197693, + "learning_rate": 9.937269829934068e-05, + "loss": 0.0313, + "step": 47320 + }, + { + "epoch": 0.151488, + "grad_norm": 0.07881982644198857, + "learning_rate": 9.937216903654867e-05, + "loss": 0.0314, + "step": 47340 + }, + { + "epoch": 0.151552, + "grad_norm": 0.06792830991567207, + "learning_rate": 9.937163955198839e-05, + "loss": 0.0336, + "step": 47360 + }, + { + "epoch": 0.151616, + "grad_norm": 0.1046935176801294, + "learning_rate": 9.937110984566222e-05, + "loss": 0.0303, + "step": 47380 + }, + { + "epoch": 0.15168, + "grad_norm": 0.09471817929841085, + "learning_rate": 9.937057991757256e-05, + "loss": 0.0294, + "step": 47400 + }, + { + "epoch": 0.151744, + "grad_norm": 0.06519829639884701, + "learning_rate": 9.937004976772177e-05, + "loss": 0.0296, + "step": 47420 + }, + { + "epoch": 0.151808, + "grad_norm": 0.0914108037579146, + "learning_rate": 9.936951939611224e-05, + "loss": 0.0333, + "step": 47440 + }, + { + "epoch": 0.151872, + "grad_norm": 0.1095163164154996, + "learning_rate": 9.936898880274635e-05, + "loss": 0.0349, + "step": 47460 + }, + { + "epoch": 0.151936, + "grad_norm": 0.0872318682174451, + "learning_rate": 9.936845798762649e-05, + "loss": 0.0314, + "step": 47480 + }, + { + "epoch": 0.152, + "grad_norm": 0.0894511986316892, + "learning_rate": 9.936792695075502e-05, + "loss": 0.0313, + "step": 47500 + }, + { + "epoch": 0.152064, + "grad_norm": 0.11465131367242962, + "learning_rate": 9.936739569213436e-05, + "loss": 0.0298, + "step": 47520 + }, + { + "epoch": 0.152128, + "grad_norm": 0.11415187623932345, + "learning_rate": 9.936686421176687e-05, + "loss": 0.0321, + "step": 47540 + }, + { + "epoch": 0.152192, + "grad_norm": 0.05614353031457746, + "learning_rate": 9.936633250965495e-05, + "loss": 0.0314, + "step": 47560 + }, + { + "epoch": 0.152256, + "grad_norm": 0.11187349813578808, + "learning_rate": 9.936580058580099e-05, + "loss": 0.0321, + "step": 47580 + }, + { + "epoch": 0.15232, + "grad_norm": 0.095249704237084, + "learning_rate": 9.936526844020737e-05, + "loss": 0.0347, + "step": 47600 + }, + { + "epoch": 0.152384, + "grad_norm": 0.06878696706060054, + "learning_rate": 9.936473607287648e-05, + "loss": 0.0321, + "step": 47620 + }, + { + "epoch": 0.152448, + "grad_norm": 0.09657385405089218, + "learning_rate": 9.936420348381073e-05, + "loss": 0.0342, + "step": 47640 + }, + { + "epoch": 0.152512, + "grad_norm": 0.06523165251001424, + "learning_rate": 9.936367067301249e-05, + "loss": 0.0334, + "step": 47660 + }, + { + "epoch": 0.152576, + "grad_norm": 0.08361358394777578, + "learning_rate": 9.936313764048418e-05, + "loss": 0.0368, + "step": 47680 + }, + { + "epoch": 0.15264, + "grad_norm": 0.11105453719160574, + "learning_rate": 9.936260438622815e-05, + "loss": 0.031, + "step": 47700 + }, + { + "epoch": 0.152704, + "grad_norm": 0.049282486443328076, + "learning_rate": 9.936207091024683e-05, + "loss": 0.0305, + "step": 47720 + }, + { + "epoch": 0.152768, + "grad_norm": 0.05392642900739235, + "learning_rate": 9.93615372125426e-05, + "loss": 0.0295, + "step": 47740 + }, + { + "epoch": 0.152832, + "grad_norm": 0.08171266608289282, + "learning_rate": 9.936100329311787e-05, + "loss": 0.0294, + "step": 47760 + }, + { + "epoch": 0.152896, + "grad_norm": 0.21723179197645665, + "learning_rate": 9.936046915197504e-05, + "loss": 0.0334, + "step": 47780 + }, + { + "epoch": 0.15296, + "grad_norm": 0.06804340390934793, + "learning_rate": 9.93599347891165e-05, + "loss": 0.0336, + "step": 47800 + }, + { + "epoch": 0.153024, + "grad_norm": 0.10602678492051296, + "learning_rate": 9.935940020454464e-05, + "loss": 0.0294, + "step": 47820 + }, + { + "epoch": 0.153088, + "grad_norm": 0.09373669930158983, + "learning_rate": 9.935886539826189e-05, + "loss": 0.0316, + "step": 47840 + }, + { + "epoch": 0.153152, + "grad_norm": 0.06574231525928272, + "learning_rate": 9.935833037027063e-05, + "loss": 0.0317, + "step": 47860 + }, + { + "epoch": 0.153216, + "grad_norm": 0.05741591396011395, + "learning_rate": 9.935779512057326e-05, + "loss": 0.031, + "step": 47880 + }, + { + "epoch": 0.15328, + "grad_norm": 0.08651174210177837, + "learning_rate": 9.93572596491722e-05, + "loss": 0.0319, + "step": 47900 + }, + { + "epoch": 0.153344, + "grad_norm": 0.07670880617933137, + "learning_rate": 9.935672395606986e-05, + "loss": 0.029, + "step": 47920 + }, + { + "epoch": 0.153408, + "grad_norm": 0.11826103380508518, + "learning_rate": 9.935618804126862e-05, + "loss": 0.0325, + "step": 47940 + }, + { + "epoch": 0.153472, + "grad_norm": 0.0951924350658214, + "learning_rate": 9.93556519047709e-05, + "loss": 0.0317, + "step": 47960 + }, + { + "epoch": 0.153536, + "grad_norm": 0.12285451581924553, + "learning_rate": 9.935511554657913e-05, + "loss": 0.03, + "step": 47980 + }, + { + "epoch": 0.1536, + "grad_norm": 0.09295613763446567, + "learning_rate": 9.935457896669568e-05, + "loss": 0.0337, + "step": 48000 + }, + { + "epoch": 0.153664, + "grad_norm": 0.07649830183179868, + "learning_rate": 9.935404216512298e-05, + "loss": 0.0329, + "step": 48020 + }, + { + "epoch": 0.153728, + "grad_norm": 0.12308444877669553, + "learning_rate": 9.935350514186345e-05, + "loss": 0.0298, + "step": 48040 + }, + { + "epoch": 0.153792, + "grad_norm": 0.073565302633695, + "learning_rate": 9.935296789691948e-05, + "loss": 0.0347, + "step": 48060 + }, + { + "epoch": 0.153856, + "grad_norm": 0.16435234169562224, + "learning_rate": 9.93524304302935e-05, + "loss": 0.0318, + "step": 48080 + }, + { + "epoch": 0.15392, + "grad_norm": 0.12342510180018815, + "learning_rate": 9.935189274198792e-05, + "loss": 0.0319, + "step": 48100 + }, + { + "epoch": 0.153984, + "grad_norm": 0.15711880367162234, + "learning_rate": 9.935135483200517e-05, + "loss": 0.0317, + "step": 48120 + }, + { + "epoch": 0.154048, + "grad_norm": 0.13183904430458246, + "learning_rate": 9.935081670034765e-05, + "loss": 0.0326, + "step": 48140 + }, + { + "epoch": 0.154112, + "grad_norm": 0.1603813572220954, + "learning_rate": 9.935027834701777e-05, + "loss": 0.0298, + "step": 48160 + }, + { + "epoch": 0.154176, + "grad_norm": 0.05849572884551675, + "learning_rate": 9.934973977201796e-05, + "loss": 0.0307, + "step": 48180 + }, + { + "epoch": 0.15424, + "grad_norm": 0.07607255537902217, + "learning_rate": 9.934920097535063e-05, + "loss": 0.029, + "step": 48200 + }, + { + "epoch": 0.154304, + "grad_norm": 0.10129880444059655, + "learning_rate": 9.934866195701821e-05, + "loss": 0.0319, + "step": 48220 + }, + { + "epoch": 0.154368, + "grad_norm": 0.07338006479938176, + "learning_rate": 9.934812271702312e-05, + "loss": 0.0297, + "step": 48240 + }, + { + "epoch": 0.154432, + "grad_norm": 0.07899350096244173, + "learning_rate": 9.934758325536777e-05, + "loss": 0.0279, + "step": 48260 + }, + { + "epoch": 0.154496, + "grad_norm": 0.07471668688072794, + "learning_rate": 9.93470435720546e-05, + "loss": 0.0285, + "step": 48280 + }, + { + "epoch": 0.15456, + "grad_norm": 0.0606879441265874, + "learning_rate": 9.934650366708602e-05, + "loss": 0.0315, + "step": 48300 + }, + { + "epoch": 0.154624, + "grad_norm": 0.06056321723365228, + "learning_rate": 9.934596354046448e-05, + "loss": 0.0298, + "step": 48320 + }, + { + "epoch": 0.154688, + "grad_norm": 0.061812749033990345, + "learning_rate": 9.934542319219237e-05, + "loss": 0.0311, + "step": 48340 + }, + { + "epoch": 0.154752, + "grad_norm": 0.18511088830744588, + "learning_rate": 9.934488262227213e-05, + "loss": 0.0365, + "step": 48360 + }, + { + "epoch": 0.154816, + "grad_norm": 0.12615828983360142, + "learning_rate": 9.934434183070621e-05, + "loss": 0.0337, + "step": 48380 + }, + { + "epoch": 0.15488, + "grad_norm": 0.06314131120509267, + "learning_rate": 9.9343800817497e-05, + "loss": 0.0324, + "step": 48400 + }, + { + "epoch": 0.154944, + "grad_norm": 0.11805073927804464, + "learning_rate": 9.934325958264699e-05, + "loss": 0.032, + "step": 48420 + }, + { + "epoch": 0.155008, + "grad_norm": 0.13177410375184753, + "learning_rate": 9.934271812615854e-05, + "loss": 0.0293, + "step": 48440 + }, + { + "epoch": 0.155072, + "grad_norm": 0.11968151949427579, + "learning_rate": 9.934217644803415e-05, + "loss": 0.0314, + "step": 48460 + }, + { + "epoch": 0.155136, + "grad_norm": 0.10424935559330906, + "learning_rate": 9.934163454827618e-05, + "loss": 0.0344, + "step": 48480 + }, + { + "epoch": 0.1552, + "grad_norm": 0.06268611257712212, + "learning_rate": 9.934109242688712e-05, + "loss": 0.0325, + "step": 48500 + }, + { + "epoch": 0.155264, + "grad_norm": 0.10164583215207992, + "learning_rate": 9.934055008386939e-05, + "loss": 0.0309, + "step": 48520 + }, + { + "epoch": 0.155328, + "grad_norm": 0.0466283264623726, + "learning_rate": 9.93400075192254e-05, + "loss": 0.029, + "step": 48540 + }, + { + "epoch": 0.155392, + "grad_norm": 0.11685403697332548, + "learning_rate": 9.933946473295765e-05, + "loss": 0.0308, + "step": 48560 + }, + { + "epoch": 0.155456, + "grad_norm": 0.07968094236867744, + "learning_rate": 9.93389217250685e-05, + "loss": 0.0288, + "step": 48580 + }, + { + "epoch": 0.15552, + "grad_norm": 0.0871747731854775, + "learning_rate": 9.933837849556047e-05, + "loss": 0.0309, + "step": 48600 + }, + { + "epoch": 0.155584, + "grad_norm": 0.08585406210884133, + "learning_rate": 9.933783504443593e-05, + "loss": 0.0296, + "step": 48620 + }, + { + "epoch": 0.155648, + "grad_norm": 0.07902233622185316, + "learning_rate": 9.933729137169736e-05, + "loss": 0.0316, + "step": 48640 + }, + { + "epoch": 0.155712, + "grad_norm": 0.05141512409029975, + "learning_rate": 9.933674747734719e-05, + "loss": 0.0292, + "step": 48660 + }, + { + "epoch": 0.155776, + "grad_norm": 0.08580218788197641, + "learning_rate": 9.933620336138787e-05, + "loss": 0.0312, + "step": 48680 + }, + { + "epoch": 0.15584, + "grad_norm": 0.1031184942870981, + "learning_rate": 9.933565902382183e-05, + "loss": 0.0337, + "step": 48700 + }, + { + "epoch": 0.155904, + "grad_norm": 0.195970797921049, + "learning_rate": 9.933511446465151e-05, + "loss": 0.0337, + "step": 48720 + }, + { + "epoch": 0.155968, + "grad_norm": 0.17182881711200834, + "learning_rate": 9.93345696838794e-05, + "loss": 0.0313, + "step": 48740 + }, + { + "epoch": 0.156032, + "grad_norm": 0.11188222083980782, + "learning_rate": 9.93340246815079e-05, + "loss": 0.0306, + "step": 48760 + }, + { + "epoch": 0.156096, + "grad_norm": 0.08352933439259387, + "learning_rate": 9.933347945753946e-05, + "loss": 0.0321, + "step": 48780 + }, + { + "epoch": 0.15616, + "grad_norm": 0.07587366896792365, + "learning_rate": 9.933293401197657e-05, + "loss": 0.031, + "step": 48800 + }, + { + "epoch": 0.156224, + "grad_norm": 0.10261290966573956, + "learning_rate": 9.933238834482164e-05, + "loss": 0.0318, + "step": 48820 + }, + { + "epoch": 0.156288, + "grad_norm": 0.05857025332956055, + "learning_rate": 9.933184245607712e-05, + "loss": 0.0318, + "step": 48840 + }, + { + "epoch": 0.156352, + "grad_norm": 0.08511855904508554, + "learning_rate": 9.93312963457455e-05, + "loss": 0.0307, + "step": 48860 + }, + { + "epoch": 0.156416, + "grad_norm": 0.06230406833463281, + "learning_rate": 9.933075001382921e-05, + "loss": 0.032, + "step": 48880 + }, + { + "epoch": 0.15648, + "grad_norm": 0.07172964512517768, + "learning_rate": 9.933020346033068e-05, + "loss": 0.0317, + "step": 48900 + }, + { + "epoch": 0.156544, + "grad_norm": 0.04747272626907838, + "learning_rate": 9.932965668525241e-05, + "loss": 0.0324, + "step": 48920 + }, + { + "epoch": 0.156608, + "grad_norm": 0.10074295129664218, + "learning_rate": 9.932910968859684e-05, + "loss": 0.033, + "step": 48940 + }, + { + "epoch": 0.156672, + "grad_norm": 0.06433961268710074, + "learning_rate": 9.932856247036641e-05, + "loss": 0.0318, + "step": 48960 + }, + { + "epoch": 0.156736, + "grad_norm": 0.05786985665701216, + "learning_rate": 9.93280150305636e-05, + "loss": 0.0318, + "step": 48980 + }, + { + "epoch": 0.1568, + "grad_norm": 0.16628679493344883, + "learning_rate": 9.932746736919083e-05, + "loss": 0.0346, + "step": 49000 + }, + { + "epoch": 0.156864, + "grad_norm": 0.08000269878384533, + "learning_rate": 9.932691948625061e-05, + "loss": 0.0312, + "step": 49020 + }, + { + "epoch": 0.156928, + "grad_norm": 0.13121743849844697, + "learning_rate": 9.932637138174537e-05, + "loss": 0.0317, + "step": 49040 + }, + { + "epoch": 0.156992, + "grad_norm": 0.12361098048767018, + "learning_rate": 9.93258230556776e-05, + "loss": 0.033, + "step": 49060 + }, + { + "epoch": 0.157056, + "grad_norm": 0.10429858292176185, + "learning_rate": 9.932527450804973e-05, + "loss": 0.0337, + "step": 49080 + }, + { + "epoch": 0.15712, + "grad_norm": 0.0647764292036461, + "learning_rate": 9.932472573886424e-05, + "loss": 0.0319, + "step": 49100 + }, + { + "epoch": 0.157184, + "grad_norm": 0.09384016399478, + "learning_rate": 9.93241767481236e-05, + "loss": 0.0317, + "step": 49120 + }, + { + "epoch": 0.157248, + "grad_norm": 0.12659843442396596, + "learning_rate": 9.932362753583025e-05, + "loss": 0.0327, + "step": 49140 + }, + { + "epoch": 0.157312, + "grad_norm": 0.07756448049192011, + "learning_rate": 9.93230781019867e-05, + "loss": 0.0296, + "step": 49160 + }, + { + "epoch": 0.157376, + "grad_norm": 0.12804142641185356, + "learning_rate": 9.932252844659537e-05, + "loss": 0.0273, + "step": 49180 + }, + { + "epoch": 0.15744, + "grad_norm": 0.06399469639064538, + "learning_rate": 9.932197856965877e-05, + "loss": 0.0311, + "step": 49200 + }, + { + "epoch": 0.157504, + "grad_norm": 0.07353854042600966, + "learning_rate": 9.932142847117936e-05, + "loss": 0.0318, + "step": 49220 + }, + { + "epoch": 0.157568, + "grad_norm": 0.09335556916903189, + "learning_rate": 9.932087815115958e-05, + "loss": 0.0322, + "step": 49240 + }, + { + "epoch": 0.157632, + "grad_norm": 0.07250359823535418, + "learning_rate": 9.932032760960194e-05, + "loss": 0.0295, + "step": 49260 + }, + { + "epoch": 0.157696, + "grad_norm": 0.09539354157045841, + "learning_rate": 9.931977684650889e-05, + "loss": 0.0329, + "step": 49280 + }, + { + "epoch": 0.15776, + "grad_norm": 0.12241863334999789, + "learning_rate": 9.931922586188292e-05, + "loss": 0.0332, + "step": 49300 + }, + { + "epoch": 0.157824, + "grad_norm": 0.09544072927237139, + "learning_rate": 9.931867465572651e-05, + "loss": 0.0337, + "step": 49320 + }, + { + "epoch": 0.157888, + "grad_norm": 0.17981690117267587, + "learning_rate": 9.931812322804211e-05, + "loss": 0.0313, + "step": 49340 + }, + { + "epoch": 0.157952, + "grad_norm": 0.12717263082414357, + "learning_rate": 9.931757157883221e-05, + "loss": 0.0286, + "step": 49360 + }, + { + "epoch": 0.158016, + "grad_norm": 0.08473942921715966, + "learning_rate": 9.931701970809927e-05, + "loss": 0.0267, + "step": 49380 + }, + { + "epoch": 0.15808, + "grad_norm": 0.055487440764002215, + "learning_rate": 9.931646761584581e-05, + "loss": 0.0287, + "step": 49400 + }, + { + "epoch": 0.158144, + "grad_norm": 0.08651242874729155, + "learning_rate": 9.93159153020743e-05, + "loss": 0.0345, + "step": 49420 + }, + { + "epoch": 0.158208, + "grad_norm": 0.0763901865694479, + "learning_rate": 9.931536276678717e-05, + "loss": 0.0336, + "step": 49440 + }, + { + "epoch": 0.158272, + "grad_norm": 0.0926115735063431, + "learning_rate": 9.931481000998697e-05, + "loss": 0.0343, + "step": 49460 + }, + { + "epoch": 0.158336, + "grad_norm": 0.09649007467061342, + "learning_rate": 9.931425703167614e-05, + "loss": 0.0289, + "step": 49480 + }, + { + "epoch": 0.1584, + "grad_norm": 0.055147182717572775, + "learning_rate": 9.931370383185718e-05, + "loss": 0.0276, + "step": 49500 + }, + { + "epoch": 0.158464, + "grad_norm": 0.08564565453956721, + "learning_rate": 9.931315041053256e-05, + "loss": 0.0318, + "step": 49520 + }, + { + "epoch": 0.158528, + "grad_norm": 0.07635713954087532, + "learning_rate": 9.931259676770479e-05, + "loss": 0.0323, + "step": 49540 + }, + { + "epoch": 0.158592, + "grad_norm": 0.052007598933735785, + "learning_rate": 9.931204290337634e-05, + "loss": 0.0323, + "step": 49560 + }, + { + "epoch": 0.158656, + "grad_norm": 0.07438361683379209, + "learning_rate": 9.931148881754971e-05, + "loss": 0.0332, + "step": 49580 + }, + { + "epoch": 0.15872, + "grad_norm": 0.1556976763584939, + "learning_rate": 9.931093451022735e-05, + "loss": 0.0338, + "step": 49600 + }, + { + "epoch": 0.158784, + "grad_norm": 0.07751104895192198, + "learning_rate": 9.931037998141181e-05, + "loss": 0.0283, + "step": 49620 + }, + { + "epoch": 0.158848, + "grad_norm": 0.11637964142966958, + "learning_rate": 9.930982523110554e-05, + "loss": 0.0325, + "step": 49640 + }, + { + "epoch": 0.158912, + "grad_norm": 0.11052488360844216, + "learning_rate": 9.930927025931105e-05, + "loss": 0.0321, + "step": 49660 + }, + { + "epoch": 0.158976, + "grad_norm": 0.07835323101756353, + "learning_rate": 9.930871506603081e-05, + "loss": 0.0306, + "step": 49680 + }, + { + "epoch": 0.15904, + "grad_norm": 0.0943194225294084, + "learning_rate": 9.930815965126734e-05, + "loss": 0.0353, + "step": 49700 + }, + { + "epoch": 0.159104, + "grad_norm": 0.059596079934019, + "learning_rate": 9.930760401502313e-05, + "loss": 0.0334, + "step": 49720 + }, + { + "epoch": 0.159168, + "grad_norm": 0.11983798600383945, + "learning_rate": 9.930704815730066e-05, + "loss": 0.0329, + "step": 49740 + }, + { + "epoch": 0.159232, + "grad_norm": 0.08806423718862402, + "learning_rate": 9.930649207810244e-05, + "loss": 0.0347, + "step": 49760 + }, + { + "epoch": 0.159296, + "grad_norm": 0.12227629251422667, + "learning_rate": 9.930593577743097e-05, + "loss": 0.0331, + "step": 49780 + }, + { + "epoch": 0.15936, + "grad_norm": 0.08686716783636916, + "learning_rate": 9.930537925528874e-05, + "loss": 0.0305, + "step": 49800 + }, + { + "epoch": 0.159424, + "grad_norm": 0.06346037990121205, + "learning_rate": 9.930482251167826e-05, + "loss": 0.0298, + "step": 49820 + }, + { + "epoch": 0.159488, + "grad_norm": 0.08159099411169625, + "learning_rate": 9.930426554660201e-05, + "loss": 0.03, + "step": 49840 + }, + { + "epoch": 0.159552, + "grad_norm": 0.09955861551417573, + "learning_rate": 9.930370836006251e-05, + "loss": 0.0271, + "step": 49860 + }, + { + "epoch": 0.159616, + "grad_norm": 0.09261095927437955, + "learning_rate": 9.930315095206226e-05, + "loss": 0.0326, + "step": 49880 + }, + { + "epoch": 0.15968, + "grad_norm": 0.07297339814073228, + "learning_rate": 9.930259332260377e-05, + "loss": 0.0309, + "step": 49900 + }, + { + "epoch": 0.159744, + "grad_norm": 0.06924682230448155, + "learning_rate": 9.930203547168953e-05, + "loss": 0.0308, + "step": 49920 + }, + { + "epoch": 0.159808, + "grad_norm": 0.07797906326662384, + "learning_rate": 9.930147739932204e-05, + "loss": 0.031, + "step": 49940 + }, + { + "epoch": 0.159872, + "grad_norm": 0.09850170486500891, + "learning_rate": 9.930091910550384e-05, + "loss": 0.034, + "step": 49960 + }, + { + "epoch": 0.159936, + "grad_norm": 0.12668685590072776, + "learning_rate": 9.93003605902374e-05, + "loss": 0.0304, + "step": 49980 + }, + { + "epoch": 0.16, + "grad_norm": 0.08995334510165826, + "learning_rate": 9.929980185352526e-05, + "loss": 0.0317, + "step": 50000 + }, + { + "epoch": 0.160064, + "grad_norm": 0.10484565801052002, + "learning_rate": 9.929924289536989e-05, + "loss": 0.0289, + "step": 50020 + }, + { + "epoch": 0.160128, + "grad_norm": 0.06870780139721683, + "learning_rate": 9.929868371577385e-05, + "loss": 0.0309, + "step": 50040 + }, + { + "epoch": 0.160192, + "grad_norm": 0.07729315804348545, + "learning_rate": 9.929812431473962e-05, + "loss": 0.0315, + "step": 50060 + }, + { + "epoch": 0.160256, + "grad_norm": 0.06636118050927628, + "learning_rate": 9.929756469226975e-05, + "loss": 0.0303, + "step": 50080 + }, + { + "epoch": 0.16032, + "grad_norm": 0.06343752170211274, + "learning_rate": 9.929700484836668e-05, + "loss": 0.0286, + "step": 50100 + }, + { + "epoch": 0.160384, + "grad_norm": 0.07586865005024584, + "learning_rate": 9.929644478303299e-05, + "loss": 0.0316, + "step": 50120 + }, + { + "epoch": 0.160448, + "grad_norm": 0.07388963618816875, + "learning_rate": 9.929588449627116e-05, + "loss": 0.0304, + "step": 50140 + }, + { + "epoch": 0.160512, + "grad_norm": 0.15512214275072297, + "learning_rate": 9.929532398808375e-05, + "loss": 0.0329, + "step": 50160 + }, + { + "epoch": 0.160576, + "grad_norm": 0.10509976931514123, + "learning_rate": 9.929476325847322e-05, + "loss": 0.031, + "step": 50180 + }, + { + "epoch": 0.16064, + "grad_norm": 0.10879812237408239, + "learning_rate": 9.929420230744214e-05, + "loss": 0.0309, + "step": 50200 + }, + { + "epoch": 0.160704, + "grad_norm": 0.06640496434008838, + "learning_rate": 9.9293641134993e-05, + "loss": 0.0335, + "step": 50220 + }, + { + "epoch": 0.160768, + "grad_norm": 0.0827563420078682, + "learning_rate": 9.929307974112832e-05, + "loss": 0.0323, + "step": 50240 + }, + { + "epoch": 0.160832, + "grad_norm": 0.11127909388966875, + "learning_rate": 9.929251812585062e-05, + "loss": 0.029, + "step": 50260 + }, + { + "epoch": 0.160896, + "grad_norm": 0.09713906597783596, + "learning_rate": 9.929195628916246e-05, + "loss": 0.0302, + "step": 50280 + }, + { + "epoch": 0.16096, + "grad_norm": 0.0723694695822792, + "learning_rate": 9.929139423106633e-05, + "loss": 0.031, + "step": 50300 + }, + { + "epoch": 0.161024, + "grad_norm": 0.0964101822993798, + "learning_rate": 9.929083195156475e-05, + "loss": 0.0312, + "step": 50320 + }, + { + "epoch": 0.161088, + "grad_norm": 0.07051873461660252, + "learning_rate": 9.929026945066025e-05, + "loss": 0.0332, + "step": 50340 + }, + { + "epoch": 0.161152, + "grad_norm": 0.09414106429200186, + "learning_rate": 9.928970672835538e-05, + "loss": 0.0313, + "step": 50360 + }, + { + "epoch": 0.161216, + "grad_norm": 0.11772716243594701, + "learning_rate": 9.928914378465263e-05, + "loss": 0.0345, + "step": 50380 + }, + { + "epoch": 0.16128, + "grad_norm": 0.07461705641114656, + "learning_rate": 9.928858061955457e-05, + "loss": 0.0306, + "step": 50400 + }, + { + "epoch": 0.161344, + "grad_norm": 0.16689222510905194, + "learning_rate": 9.92880172330637e-05, + "loss": 0.03, + "step": 50420 + }, + { + "epoch": 0.161408, + "grad_norm": 0.07913562416839212, + "learning_rate": 9.928745362518256e-05, + "loss": 0.0301, + "step": 50440 + }, + { + "epoch": 0.161472, + "grad_norm": 0.050300774448710464, + "learning_rate": 9.928688979591368e-05, + "loss": 0.029, + "step": 50460 + }, + { + "epoch": 0.161536, + "grad_norm": 0.08429781116475236, + "learning_rate": 9.92863257452596e-05, + "loss": 0.0276, + "step": 50480 + }, + { + "epoch": 0.1616, + "grad_norm": 0.11099378024299517, + "learning_rate": 9.928576147322284e-05, + "loss": 0.031, + "step": 50500 + }, + { + "epoch": 0.161664, + "grad_norm": 0.10813686447975093, + "learning_rate": 9.928519697980594e-05, + "loss": 0.031, + "step": 50520 + }, + { + "epoch": 0.161728, + "grad_norm": 0.0685192254157951, + "learning_rate": 9.928463226501143e-05, + "loss": 0.0266, + "step": 50540 + }, + { + "epoch": 0.161792, + "grad_norm": 0.06863719815742642, + "learning_rate": 9.928406732884188e-05, + "loss": 0.0342, + "step": 50560 + }, + { + "epoch": 0.161856, + "grad_norm": 0.08230820340311004, + "learning_rate": 9.928350217129977e-05, + "loss": 0.0326, + "step": 50580 + }, + { + "epoch": 0.16192, + "grad_norm": 0.06509154971910303, + "learning_rate": 9.92829367923877e-05, + "loss": 0.0312, + "step": 50600 + }, + { + "epoch": 0.161984, + "grad_norm": 0.11518141006203565, + "learning_rate": 9.928237119210815e-05, + "loss": 0.0314, + "step": 50620 + }, + { + "epoch": 0.162048, + "grad_norm": 0.09876785991501875, + "learning_rate": 9.92818053704637e-05, + "loss": 0.0324, + "step": 50640 + }, + { + "epoch": 0.162112, + "grad_norm": 0.0892518642207428, + "learning_rate": 9.92812393274569e-05, + "loss": 0.0347, + "step": 50660 + }, + { + "epoch": 0.162176, + "grad_norm": 0.07550052485582388, + "learning_rate": 9.928067306309026e-05, + "loss": 0.034, + "step": 50680 + }, + { + "epoch": 0.16224, + "grad_norm": 0.06482135585403946, + "learning_rate": 9.928010657736633e-05, + "loss": 0.0366, + "step": 50700 + }, + { + "epoch": 0.162304, + "grad_norm": 0.16458258672769438, + "learning_rate": 9.927953987028767e-05, + "loss": 0.0323, + "step": 50720 + }, + { + "epoch": 0.162368, + "grad_norm": 0.10358684245956191, + "learning_rate": 9.927897294185682e-05, + "loss": 0.0317, + "step": 50740 + }, + { + "epoch": 0.162432, + "grad_norm": 0.14390584348464086, + "learning_rate": 9.927840579207633e-05, + "loss": 0.0291, + "step": 50760 + }, + { + "epoch": 0.162496, + "grad_norm": 0.06543715676469103, + "learning_rate": 9.927783842094872e-05, + "loss": 0.0295, + "step": 50780 + }, + { + "epoch": 0.16256, + "grad_norm": 0.07231833572505786, + "learning_rate": 9.927727082847658e-05, + "loss": 0.0318, + "step": 50800 + }, + { + "epoch": 0.162624, + "grad_norm": 0.09524849656835833, + "learning_rate": 9.927670301466244e-05, + "loss": 0.0306, + "step": 50820 + }, + { + "epoch": 0.162688, + "grad_norm": 0.19935645305316482, + "learning_rate": 9.927613497950883e-05, + "loss": 0.0326, + "step": 50840 + }, + { + "epoch": 0.162752, + "grad_norm": 0.0680535305092903, + "learning_rate": 9.927556672301834e-05, + "loss": 0.0312, + "step": 50860 + }, + { + "epoch": 0.162816, + "grad_norm": 0.08080350899695349, + "learning_rate": 9.927499824519348e-05, + "loss": 0.0363, + "step": 50880 + }, + { + "epoch": 0.16288, + "grad_norm": 0.14849199161636034, + "learning_rate": 9.927442954603685e-05, + "loss": 0.0393, + "step": 50900 + }, + { + "epoch": 0.162944, + "grad_norm": 0.14972292155397732, + "learning_rate": 9.927386062555099e-05, + "loss": 0.0321, + "step": 50920 + }, + { + "epoch": 0.163008, + "grad_norm": 0.10823737414139363, + "learning_rate": 9.927329148373843e-05, + "loss": 0.0292, + "step": 50940 + }, + { + "epoch": 0.163072, + "grad_norm": 0.09126033619025652, + "learning_rate": 9.927272212060175e-05, + "loss": 0.0344, + "step": 50960 + }, + { + "epoch": 0.163136, + "grad_norm": 0.07754175292339982, + "learning_rate": 9.92721525361435e-05, + "loss": 0.0334, + "step": 50980 + }, + { + "epoch": 0.1632, + "grad_norm": 0.10516637252121326, + "learning_rate": 9.927158273036625e-05, + "loss": 0.0327, + "step": 51000 + }, + { + "epoch": 0.163264, + "grad_norm": 0.13199644513495773, + "learning_rate": 9.927101270327252e-05, + "loss": 0.0335, + "step": 51020 + }, + { + "epoch": 0.163328, + "grad_norm": 0.06694751698605156, + "learning_rate": 9.927044245486492e-05, + "loss": 0.032, + "step": 51040 + }, + { + "epoch": 0.163392, + "grad_norm": 0.08706794643911839, + "learning_rate": 9.9269871985146e-05, + "loss": 0.0296, + "step": 51060 + }, + { + "epoch": 0.163456, + "grad_norm": 0.06883197814459027, + "learning_rate": 9.92693012941183e-05, + "loss": 0.031, + "step": 51080 + }, + { + "epoch": 0.16352, + "grad_norm": 0.14265238195938584, + "learning_rate": 9.926873038178439e-05, + "loss": 0.0312, + "step": 51100 + }, + { + "epoch": 0.163584, + "grad_norm": 0.08199550221964322, + "learning_rate": 9.926815924814687e-05, + "loss": 0.032, + "step": 51120 + }, + { + "epoch": 0.163648, + "grad_norm": 0.1286889855575657, + "learning_rate": 9.926758789320825e-05, + "loss": 0.0325, + "step": 51140 + }, + { + "epoch": 0.163712, + "grad_norm": 0.17403492446015967, + "learning_rate": 9.926701631697113e-05, + "loss": 0.0319, + "step": 51160 + }, + { + "epoch": 0.163776, + "grad_norm": 0.07470681568472168, + "learning_rate": 9.926644451943807e-05, + "loss": 0.0282, + "step": 51180 + }, + { + "epoch": 0.16384, + "grad_norm": 0.09782699404728386, + "learning_rate": 9.926587250061164e-05, + "loss": 0.0283, + "step": 51200 + }, + { + "epoch": 0.163904, + "grad_norm": 0.06487118520912383, + "learning_rate": 9.926530026049441e-05, + "loss": 0.029, + "step": 51220 + }, + { + "epoch": 0.163968, + "grad_norm": 0.08146092177938395, + "learning_rate": 9.926472779908896e-05, + "loss": 0.0314, + "step": 51240 + }, + { + "epoch": 0.164032, + "grad_norm": 0.10401449778480971, + "learning_rate": 9.926415511639783e-05, + "loss": 0.0321, + "step": 51260 + }, + { + "epoch": 0.164096, + "grad_norm": 0.06450719894636027, + "learning_rate": 9.926358221242362e-05, + "loss": 0.0295, + "step": 51280 + }, + { + "epoch": 0.16416, + "grad_norm": 0.07639253295228808, + "learning_rate": 9.926300908716889e-05, + "loss": 0.0326, + "step": 51300 + }, + { + "epoch": 0.164224, + "grad_norm": 0.07906989813952332, + "learning_rate": 9.926243574063623e-05, + "loss": 0.033, + "step": 51320 + }, + { + "epoch": 0.164288, + "grad_norm": 0.1218135693052076, + "learning_rate": 9.92618621728282e-05, + "loss": 0.0332, + "step": 51340 + }, + { + "epoch": 0.164352, + "grad_norm": 0.0648113099023763, + "learning_rate": 9.926128838374738e-05, + "loss": 0.0295, + "step": 51360 + }, + { + "epoch": 0.164416, + "grad_norm": 0.07123555000472492, + "learning_rate": 9.926071437339634e-05, + "loss": 0.0271, + "step": 51380 + }, + { + "epoch": 0.16448, + "grad_norm": 0.0696346716827423, + "learning_rate": 9.92601401417777e-05, + "loss": 0.0325, + "step": 51400 + }, + { + "epoch": 0.164544, + "grad_norm": 0.10624430299885847, + "learning_rate": 9.925956568889397e-05, + "loss": 0.0362, + "step": 51420 + }, + { + "epoch": 0.164608, + "grad_norm": 0.11595614242976489, + "learning_rate": 9.925899101474778e-05, + "loss": 0.0302, + "step": 51440 + }, + { + "epoch": 0.164672, + "grad_norm": 0.08169263192304588, + "learning_rate": 9.92584161193417e-05, + "loss": 0.0311, + "step": 51460 + }, + { + "epoch": 0.164736, + "grad_norm": 0.10775947598176998, + "learning_rate": 9.925784100267832e-05, + "loss": 0.0322, + "step": 51480 + }, + { + "epoch": 0.1648, + "grad_norm": 0.07657609166749417, + "learning_rate": 9.92572656647602e-05, + "loss": 0.0308, + "step": 51500 + }, + { + "epoch": 0.164864, + "grad_norm": 0.09146354766575497, + "learning_rate": 9.925669010558994e-05, + "loss": 0.0332, + "step": 51520 + }, + { + "epoch": 0.164928, + "grad_norm": 0.10038750166039168, + "learning_rate": 9.925611432517013e-05, + "loss": 0.0312, + "step": 51540 + }, + { + "epoch": 0.164992, + "grad_norm": 0.13152276685572026, + "learning_rate": 9.925553832350335e-05, + "loss": 0.0336, + "step": 51560 + }, + { + "epoch": 0.165056, + "grad_norm": 0.07257892834662695, + "learning_rate": 9.925496210059217e-05, + "loss": 0.034, + "step": 51580 + }, + { + "epoch": 0.16512, + "grad_norm": 0.1368160087000327, + "learning_rate": 9.925438565643922e-05, + "loss": 0.0318, + "step": 51600 + }, + { + "epoch": 0.165184, + "grad_norm": 0.10509608417678797, + "learning_rate": 9.925380899104706e-05, + "loss": 0.0317, + "step": 51620 + }, + { + "epoch": 0.165248, + "grad_norm": 0.10170376891373777, + "learning_rate": 9.925323210441827e-05, + "loss": 0.0309, + "step": 51640 + }, + { + "epoch": 0.165312, + "grad_norm": 0.06544655377744597, + "learning_rate": 9.925265499655547e-05, + "loss": 0.0334, + "step": 51660 + }, + { + "epoch": 0.165376, + "grad_norm": 0.06503813536977048, + "learning_rate": 9.925207766746123e-05, + "loss": 0.032, + "step": 51680 + }, + { + "epoch": 0.16544, + "grad_norm": 0.05718428299776666, + "learning_rate": 9.925150011713815e-05, + "loss": 0.0308, + "step": 51700 + }, + { + "epoch": 0.165504, + "grad_norm": 0.08339893608014584, + "learning_rate": 9.925092234558882e-05, + "loss": 0.0307, + "step": 51720 + }, + { + "epoch": 0.165568, + "grad_norm": 0.10773249539027138, + "learning_rate": 9.925034435281586e-05, + "loss": 0.032, + "step": 51740 + }, + { + "epoch": 0.165632, + "grad_norm": 0.07801557840251831, + "learning_rate": 9.924976613882183e-05, + "loss": 0.0285, + "step": 51760 + }, + { + "epoch": 0.165696, + "grad_norm": 0.05561458972447706, + "learning_rate": 9.924918770360935e-05, + "loss": 0.0293, + "step": 51780 + }, + { + "epoch": 0.16576, + "grad_norm": 0.1022244361708253, + "learning_rate": 9.924860904718101e-05, + "loss": 0.0271, + "step": 51800 + }, + { + "epoch": 0.165824, + "grad_norm": 0.07001538775612266, + "learning_rate": 9.924803016953942e-05, + "loss": 0.0294, + "step": 51820 + }, + { + "epoch": 0.165888, + "grad_norm": 0.10252383668519809, + "learning_rate": 9.924745107068716e-05, + "loss": 0.0333, + "step": 51840 + }, + { + "epoch": 0.165952, + "grad_norm": 0.10561463285844534, + "learning_rate": 9.924687175062686e-05, + "loss": 0.0376, + "step": 51860 + }, + { + "epoch": 0.166016, + "grad_norm": 0.07482436166623027, + "learning_rate": 9.924629220936108e-05, + "loss": 0.0327, + "step": 51880 + }, + { + "epoch": 0.16608, + "grad_norm": 0.06785578315135021, + "learning_rate": 9.924571244689247e-05, + "loss": 0.0278, + "step": 51900 + }, + { + "epoch": 0.166144, + "grad_norm": 0.08702415576792027, + "learning_rate": 9.92451324632236e-05, + "loss": 0.0308, + "step": 51920 + }, + { + "epoch": 0.166208, + "grad_norm": 0.06905621349735323, + "learning_rate": 9.924455225835709e-05, + "loss": 0.0346, + "step": 51940 + }, + { + "epoch": 0.166272, + "grad_norm": 0.08197492519966695, + "learning_rate": 9.924397183229556e-05, + "loss": 0.0332, + "step": 51960 + }, + { + "epoch": 0.166336, + "grad_norm": 0.21338512474612942, + "learning_rate": 9.924339118504158e-05, + "loss": 0.0321, + "step": 51980 + }, + { + "epoch": 0.1664, + "grad_norm": 0.14102681187289734, + "learning_rate": 9.92428103165978e-05, + "loss": 0.0356, + "step": 52000 + }, + { + "epoch": 0.166464, + "grad_norm": 0.10183137664835576, + "learning_rate": 9.924222922696678e-05, + "loss": 0.0331, + "step": 52020 + }, + { + "epoch": 0.166528, + "grad_norm": 0.13725531938837807, + "learning_rate": 9.924164791615119e-05, + "loss": 0.0333, + "step": 52040 + }, + { + "epoch": 0.166592, + "grad_norm": 0.08604272771775957, + "learning_rate": 9.924106638415358e-05, + "loss": 0.0297, + "step": 52060 + }, + { + "epoch": 0.166656, + "grad_norm": 0.12455713972365082, + "learning_rate": 9.92404846309766e-05, + "loss": 0.0306, + "step": 52080 + }, + { + "epoch": 0.16672, + "grad_norm": 0.10586638057793347, + "learning_rate": 9.923990265662286e-05, + "loss": 0.0299, + "step": 52100 + }, + { + "epoch": 0.166784, + "grad_norm": 0.09354331096569239, + "learning_rate": 9.923932046109496e-05, + "loss": 0.0292, + "step": 52120 + }, + { + "epoch": 0.166848, + "grad_norm": 0.0725808832422317, + "learning_rate": 9.923873804439554e-05, + "loss": 0.0279, + "step": 52140 + }, + { + "epoch": 0.166912, + "grad_norm": 0.0930606813308102, + "learning_rate": 9.923815540652719e-05, + "loss": 0.0322, + "step": 52160 + }, + { + "epoch": 0.166976, + "grad_norm": 0.07966735750716024, + "learning_rate": 9.923757254749254e-05, + "loss": 0.0336, + "step": 52180 + }, + { + "epoch": 0.16704, + "grad_norm": 0.11952993988819609, + "learning_rate": 9.92369894672942e-05, + "loss": 0.0307, + "step": 52200 + }, + { + "epoch": 0.167104, + "grad_norm": 0.08336947137503813, + "learning_rate": 9.92364061659348e-05, + "loss": 0.0336, + "step": 52220 + }, + { + "epoch": 0.167168, + "grad_norm": 0.06809469975476605, + "learning_rate": 9.923582264341695e-05, + "loss": 0.0303, + "step": 52240 + }, + { + "epoch": 0.167232, + "grad_norm": 0.07866981825638819, + "learning_rate": 9.923523889974327e-05, + "loss": 0.0309, + "step": 52260 + }, + { + "epoch": 0.167296, + "grad_norm": 0.10143591042860282, + "learning_rate": 9.923465493491639e-05, + "loss": 0.0334, + "step": 52280 + }, + { + "epoch": 0.16736, + "grad_norm": 0.09757361316801456, + "learning_rate": 9.923407074893894e-05, + "loss": 0.0293, + "step": 52300 + }, + { + "epoch": 0.167424, + "grad_norm": 0.10434374813917567, + "learning_rate": 9.923348634181353e-05, + "loss": 0.0264, + "step": 52320 + }, + { + "epoch": 0.167488, + "grad_norm": 0.06820094614970226, + "learning_rate": 9.92329017135428e-05, + "loss": 0.0313, + "step": 52340 + }, + { + "epoch": 0.167552, + "grad_norm": 0.09077555361309897, + "learning_rate": 9.923231686412935e-05, + "loss": 0.0289, + "step": 52360 + }, + { + "epoch": 0.167616, + "grad_norm": 0.055018866253847214, + "learning_rate": 9.923173179357581e-05, + "loss": 0.0298, + "step": 52380 + }, + { + "epoch": 0.16768, + "grad_norm": 0.11484318921072584, + "learning_rate": 9.923114650188484e-05, + "loss": 0.0339, + "step": 52400 + }, + { + "epoch": 0.167744, + "grad_norm": 0.08856312616907716, + "learning_rate": 9.923056098905905e-05, + "loss": 0.0313, + "step": 52420 + }, + { + "epoch": 0.167808, + "grad_norm": 0.08663006417387209, + "learning_rate": 9.922997525510106e-05, + "loss": 0.0304, + "step": 52440 + }, + { + "epoch": 0.167872, + "grad_norm": 0.1077835338482995, + "learning_rate": 9.922938930001351e-05, + "loss": 0.0336, + "step": 52460 + }, + { + "epoch": 0.167936, + "grad_norm": 0.08136703202548481, + "learning_rate": 9.922880312379904e-05, + "loss": 0.0323, + "step": 52480 + }, + { + "epoch": 0.168, + "grad_norm": 0.058970828083442985, + "learning_rate": 9.922821672646027e-05, + "loss": 0.0314, + "step": 52500 + }, + { + "epoch": 0.168064, + "grad_norm": 0.08772403564825054, + "learning_rate": 9.922763010799985e-05, + "loss": 0.0306, + "step": 52520 + }, + { + "epoch": 0.168128, + "grad_norm": 0.07579732349258023, + "learning_rate": 9.922704326842039e-05, + "loss": 0.0322, + "step": 52540 + }, + { + "epoch": 0.168192, + "grad_norm": 0.05756871733740165, + "learning_rate": 9.922645620772455e-05, + "loss": 0.0353, + "step": 52560 + }, + { + "epoch": 0.168256, + "grad_norm": 0.08309827407855326, + "learning_rate": 9.922586892591494e-05, + "loss": 0.0303, + "step": 52580 + }, + { + "epoch": 0.16832, + "grad_norm": 0.1815933671476109, + "learning_rate": 9.922528142299423e-05, + "loss": 0.0326, + "step": 52600 + }, + { + "epoch": 0.168384, + "grad_norm": 0.06510824172518967, + "learning_rate": 9.922469369896505e-05, + "loss": 0.0308, + "step": 52620 + }, + { + "epoch": 0.168448, + "grad_norm": 0.07414921551709572, + "learning_rate": 9.922410575383002e-05, + "loss": 0.0303, + "step": 52640 + }, + { + "epoch": 0.168512, + "grad_norm": 0.07015495544418383, + "learning_rate": 9.92235175875918e-05, + "loss": 0.0322, + "step": 52660 + }, + { + "epoch": 0.168576, + "grad_norm": 0.0687650193427961, + "learning_rate": 9.922292920025302e-05, + "loss": 0.0333, + "step": 52680 + }, + { + "epoch": 0.16864, + "grad_norm": 0.09827637358485757, + "learning_rate": 9.922234059181634e-05, + "loss": 0.0304, + "step": 52700 + }, + { + "epoch": 0.168704, + "grad_norm": 0.12755071227545084, + "learning_rate": 9.922175176228439e-05, + "loss": 0.032, + "step": 52720 + }, + { + "epoch": 0.168768, + "grad_norm": 0.08182040432494203, + "learning_rate": 9.922116271165982e-05, + "loss": 0.0287, + "step": 52740 + }, + { + "epoch": 0.168832, + "grad_norm": 0.06708263020012792, + "learning_rate": 9.922057343994527e-05, + "loss": 0.0298, + "step": 52760 + }, + { + "epoch": 0.168896, + "grad_norm": 0.10054394998860977, + "learning_rate": 9.92199839471434e-05, + "loss": 0.0345, + "step": 52780 + }, + { + "epoch": 0.16896, + "grad_norm": 0.10441608313800908, + "learning_rate": 9.921939423325684e-05, + "loss": 0.0326, + "step": 52800 + }, + { + "epoch": 0.169024, + "grad_norm": 0.08138901035339302, + "learning_rate": 9.921880429828823e-05, + "loss": 0.0289, + "step": 52820 + }, + { + "epoch": 0.169088, + "grad_norm": 0.0789155974650951, + "learning_rate": 9.921821414224026e-05, + "loss": 0.0309, + "step": 52840 + }, + { + "epoch": 0.169152, + "grad_norm": 0.07992279146282948, + "learning_rate": 9.921762376511556e-05, + "loss": 0.0286, + "step": 52860 + }, + { + "epoch": 0.169216, + "grad_norm": 0.1701725206408709, + "learning_rate": 9.921703316691677e-05, + "loss": 0.03, + "step": 52880 + }, + { + "epoch": 0.16928, + "grad_norm": 0.06538783949958216, + "learning_rate": 9.921644234764657e-05, + "loss": 0.0314, + "step": 52900 + }, + { + "epoch": 0.169344, + "grad_norm": 0.07255474742041096, + "learning_rate": 9.921585130730757e-05, + "loss": 0.0274, + "step": 52920 + }, + { + "epoch": 0.169408, + "grad_norm": 0.06842028176915492, + "learning_rate": 9.921526004590247e-05, + "loss": 0.03, + "step": 52940 + }, + { + "epoch": 0.169472, + "grad_norm": 0.06264157629722407, + "learning_rate": 9.92146685634339e-05, + "loss": 0.0296, + "step": 52960 + }, + { + "epoch": 0.169536, + "grad_norm": 0.06578209545573395, + "learning_rate": 9.921407685990454e-05, + "loss": 0.0301, + "step": 52980 + }, + { + "epoch": 0.1696, + "grad_norm": 0.06465399115817935, + "learning_rate": 9.921348493531701e-05, + "loss": 0.0337, + "step": 53000 + }, + { + "epoch": 0.169664, + "grad_norm": 0.13783237396821904, + "learning_rate": 9.9212892789674e-05, + "loss": 0.0294, + "step": 53020 + }, + { + "epoch": 0.169728, + "grad_norm": 0.08357696760383225, + "learning_rate": 9.921230042297817e-05, + "loss": 0.0301, + "step": 53040 + }, + { + "epoch": 0.169792, + "grad_norm": 0.08087643276662872, + "learning_rate": 9.921170783523217e-05, + "loss": 0.0327, + "step": 53060 + }, + { + "epoch": 0.169856, + "grad_norm": 0.09223140281830625, + "learning_rate": 9.921111502643865e-05, + "loss": 0.0296, + "step": 53080 + }, + { + "epoch": 0.16992, + "grad_norm": 0.05906062928301391, + "learning_rate": 9.92105219966003e-05, + "loss": 0.0314, + "step": 53100 + }, + { + "epoch": 0.169984, + "grad_norm": 0.12985854919588832, + "learning_rate": 9.920992874571977e-05, + "loss": 0.0325, + "step": 53120 + }, + { + "epoch": 0.170048, + "grad_norm": 0.08750609966692707, + "learning_rate": 9.920933527379972e-05, + "loss": 0.0279, + "step": 53140 + }, + { + "epoch": 0.170112, + "grad_norm": 0.07849605157123798, + "learning_rate": 9.920874158084281e-05, + "loss": 0.0326, + "step": 53160 + }, + { + "epoch": 0.170176, + "grad_norm": 0.09121358165961127, + "learning_rate": 9.920814766685173e-05, + "loss": 0.0315, + "step": 53180 + }, + { + "epoch": 0.17024, + "grad_norm": 0.10040733623485716, + "learning_rate": 9.920755353182912e-05, + "loss": 0.0315, + "step": 53200 + }, + { + "epoch": 0.170304, + "grad_norm": 0.06949009751419008, + "learning_rate": 9.920695917577768e-05, + "loss": 0.0312, + "step": 53220 + }, + { + "epoch": 0.170368, + "grad_norm": 0.05668613443243568, + "learning_rate": 9.920636459870005e-05, + "loss": 0.0286, + "step": 53240 + }, + { + "epoch": 0.170432, + "grad_norm": 0.118834320576583, + "learning_rate": 9.920576980059893e-05, + "loss": 0.0321, + "step": 53260 + }, + { + "epoch": 0.170496, + "grad_norm": 0.11491425204814783, + "learning_rate": 9.920517478147697e-05, + "loss": 0.0318, + "step": 53280 + }, + { + "epoch": 0.17056, + "grad_norm": 0.10130748213778534, + "learning_rate": 9.920457954133684e-05, + "loss": 0.0297, + "step": 53300 + }, + { + "epoch": 0.170624, + "grad_norm": 0.0788841689387144, + "learning_rate": 9.920398408018121e-05, + "loss": 0.0307, + "step": 53320 + }, + { + "epoch": 0.170688, + "grad_norm": 0.0750577109503732, + "learning_rate": 9.92033883980128e-05, + "loss": 0.0292, + "step": 53340 + }, + { + "epoch": 0.170752, + "grad_norm": 0.0798979323047073, + "learning_rate": 9.920279249483423e-05, + "loss": 0.0326, + "step": 53360 + }, + { + "epoch": 0.170816, + "grad_norm": 0.11861375265083017, + "learning_rate": 9.920219637064821e-05, + "loss": 0.0347, + "step": 53380 + }, + { + "epoch": 0.17088, + "grad_norm": 0.07042137370283622, + "learning_rate": 9.92016000254574e-05, + "loss": 0.0332, + "step": 53400 + }, + { + "epoch": 0.170944, + "grad_norm": 0.06976735506517781, + "learning_rate": 9.920100345926448e-05, + "loss": 0.0305, + "step": 53420 + }, + { + "epoch": 0.171008, + "grad_norm": 0.09884066168152919, + "learning_rate": 9.920040667207213e-05, + "loss": 0.0311, + "step": 53440 + }, + { + "epoch": 0.171072, + "grad_norm": 0.08957776902647491, + "learning_rate": 9.919980966388304e-05, + "loss": 0.0307, + "step": 53460 + }, + { + "epoch": 0.171136, + "grad_norm": 0.14311288846498768, + "learning_rate": 9.91992124346999e-05, + "loss": 0.035, + "step": 53480 + }, + { + "epoch": 0.1712, + "grad_norm": 0.07460503999188918, + "learning_rate": 9.919861498452538e-05, + "loss": 0.0306, + "step": 53500 + }, + { + "epoch": 0.171264, + "grad_norm": 0.06371043305198766, + "learning_rate": 9.919801731336215e-05, + "loss": 0.0308, + "step": 53520 + }, + { + "epoch": 0.171328, + "grad_norm": 0.0604285323046146, + "learning_rate": 9.919741942121291e-05, + "loss": 0.0312, + "step": 53540 + }, + { + "epoch": 0.171392, + "grad_norm": 0.15143559458314848, + "learning_rate": 9.919682130808035e-05, + "loss": 0.0368, + "step": 53560 + }, + { + "epoch": 0.171456, + "grad_norm": 0.06755668370527207, + "learning_rate": 9.919622297396715e-05, + "loss": 0.0347, + "step": 53580 + }, + { + "epoch": 0.17152, + "grad_norm": 0.09883448013475127, + "learning_rate": 9.9195624418876e-05, + "loss": 0.0348, + "step": 53600 + }, + { + "epoch": 0.171584, + "grad_norm": 0.06019099759944975, + "learning_rate": 9.919502564280959e-05, + "loss": 0.0335, + "step": 53620 + }, + { + "epoch": 0.171648, + "grad_norm": 0.06525881854863104, + "learning_rate": 9.919442664577058e-05, + "loss": 0.0312, + "step": 53640 + }, + { + "epoch": 0.171712, + "grad_norm": 0.08938804091826309, + "learning_rate": 9.919382742776171e-05, + "loss": 0.0292, + "step": 53660 + }, + { + "epoch": 0.171776, + "grad_norm": 0.06703141282332108, + "learning_rate": 9.919322798878563e-05, + "loss": 0.0296, + "step": 53680 + }, + { + "epoch": 0.17184, + "grad_norm": 0.05385434367129899, + "learning_rate": 9.919262832884506e-05, + "loss": 0.0275, + "step": 53700 + }, + { + "epoch": 0.171904, + "grad_norm": 0.07286080154617536, + "learning_rate": 9.919202844794269e-05, + "loss": 0.0325, + "step": 53720 + }, + { + "epoch": 0.171968, + "grad_norm": 0.13276363650132328, + "learning_rate": 9.91914283460812e-05, + "loss": 0.0307, + "step": 53740 + }, + { + "epoch": 0.172032, + "grad_norm": 0.08132489743893076, + "learning_rate": 9.91908280232633e-05, + "loss": 0.0324, + "step": 53760 + }, + { + "epoch": 0.172096, + "grad_norm": 0.12442817663886427, + "learning_rate": 9.919022747949167e-05, + "loss": 0.0305, + "step": 53780 + }, + { + "epoch": 0.17216, + "grad_norm": 0.08397710068764623, + "learning_rate": 9.918962671476903e-05, + "loss": 0.031, + "step": 53800 + }, + { + "epoch": 0.172224, + "grad_norm": 0.11539759790231466, + "learning_rate": 9.918902572909807e-05, + "loss": 0.0339, + "step": 53820 + }, + { + "epoch": 0.172288, + "grad_norm": 0.05674429708565355, + "learning_rate": 9.918842452248147e-05, + "loss": 0.033, + "step": 53840 + }, + { + "epoch": 0.172352, + "grad_norm": 0.11328900164746662, + "learning_rate": 9.918782309492196e-05, + "loss": 0.0298, + "step": 53860 + }, + { + "epoch": 0.172416, + "grad_norm": 0.09390005756066026, + "learning_rate": 9.918722144642223e-05, + "loss": 0.0315, + "step": 53880 + }, + { + "epoch": 0.17248, + "grad_norm": 0.06626530414951148, + "learning_rate": 9.918661957698497e-05, + "loss": 0.0327, + "step": 53900 + }, + { + "epoch": 0.172544, + "grad_norm": 0.1860125245592864, + "learning_rate": 9.918601748661289e-05, + "loss": 0.0335, + "step": 53920 + }, + { + "epoch": 0.172608, + "grad_norm": 0.0771748055697441, + "learning_rate": 9.91854151753087e-05, + "loss": 0.031, + "step": 53940 + }, + { + "epoch": 0.172672, + "grad_norm": 0.10082282265842987, + "learning_rate": 9.918481264307512e-05, + "loss": 0.0327, + "step": 53960 + }, + { + "epoch": 0.172736, + "grad_norm": 0.11034141211176166, + "learning_rate": 9.918420988991483e-05, + "loss": 0.0328, + "step": 53980 + }, + { + "epoch": 0.1728, + "grad_norm": 0.09128214605362321, + "learning_rate": 9.918360691583056e-05, + "loss": 0.0325, + "step": 54000 + }, + { + "epoch": 0.172864, + "grad_norm": 0.09015722078925593, + "learning_rate": 9.918300372082499e-05, + "loss": 0.0297, + "step": 54020 + }, + { + "epoch": 0.172928, + "grad_norm": 0.0710194023491567, + "learning_rate": 9.918240030490084e-05, + "loss": 0.0306, + "step": 54040 + }, + { + "epoch": 0.172992, + "grad_norm": 0.12552794552149327, + "learning_rate": 9.918179666806084e-05, + "loss": 0.0332, + "step": 54060 + }, + { + "epoch": 0.173056, + "grad_norm": 0.09017113151744216, + "learning_rate": 9.918119281030766e-05, + "loss": 0.0357, + "step": 54080 + }, + { + "epoch": 0.17312, + "grad_norm": 0.10253985128926699, + "learning_rate": 9.918058873164408e-05, + "loss": 0.0327, + "step": 54100 + }, + { + "epoch": 0.173184, + "grad_norm": 0.09813362396155446, + "learning_rate": 9.917998443207275e-05, + "loss": 0.0302, + "step": 54120 + }, + { + "epoch": 0.173248, + "grad_norm": 0.08933503040913926, + "learning_rate": 9.91793799115964e-05, + "loss": 0.0276, + "step": 54140 + }, + { + "epoch": 0.173312, + "grad_norm": 0.19420673151881498, + "learning_rate": 9.917877517021777e-05, + "loss": 0.03, + "step": 54160 + }, + { + "epoch": 0.173376, + "grad_norm": 0.09490862421460453, + "learning_rate": 9.917817020793956e-05, + "loss": 0.0318, + "step": 54180 + }, + { + "epoch": 0.17344, + "grad_norm": 0.05281978002721933, + "learning_rate": 9.917756502476447e-05, + "loss": 0.0338, + "step": 54200 + }, + { + "epoch": 0.173504, + "grad_norm": 0.1346937942471137, + "learning_rate": 9.917695962069526e-05, + "loss": 0.0288, + "step": 54220 + }, + { + "epoch": 0.173568, + "grad_norm": 0.10340093133273946, + "learning_rate": 9.91763539957346e-05, + "loss": 0.0302, + "step": 54240 + }, + { + "epoch": 0.173632, + "grad_norm": 0.12945925439800732, + "learning_rate": 9.917574814988524e-05, + "loss": 0.0313, + "step": 54260 + }, + { + "epoch": 0.173696, + "grad_norm": 0.07218867013486345, + "learning_rate": 9.91751420831499e-05, + "loss": 0.03, + "step": 54280 + }, + { + "epoch": 0.17376, + "grad_norm": 0.09811110918316432, + "learning_rate": 9.917453579553131e-05, + "loss": 0.0314, + "step": 54300 + }, + { + "epoch": 0.173824, + "grad_norm": 0.09266532423955416, + "learning_rate": 9.917392928703215e-05, + "loss": 0.0331, + "step": 54320 + }, + { + "epoch": 0.173888, + "grad_norm": 0.08765086125126974, + "learning_rate": 9.917332255765521e-05, + "loss": 0.032, + "step": 54340 + }, + { + "epoch": 0.173952, + "grad_norm": 0.12154923514913077, + "learning_rate": 9.917271560740316e-05, + "loss": 0.0319, + "step": 54360 + }, + { + "epoch": 0.174016, + "grad_norm": 0.07059666902726938, + "learning_rate": 9.917210843627876e-05, + "loss": 0.0332, + "step": 54380 + }, + { + "epoch": 0.17408, + "grad_norm": 0.09817166848630199, + "learning_rate": 9.91715010442847e-05, + "loss": 0.0319, + "step": 54400 + }, + { + "epoch": 0.174144, + "grad_norm": 0.08566627308306625, + "learning_rate": 9.917089343142376e-05, + "loss": 0.0287, + "step": 54420 + }, + { + "epoch": 0.174208, + "grad_norm": 0.05658908047202619, + "learning_rate": 9.917028559769864e-05, + "loss": 0.031, + "step": 54440 + }, + { + "epoch": 0.174272, + "grad_norm": 0.17471714735438232, + "learning_rate": 9.916967754311205e-05, + "loss": 0.0303, + "step": 54460 + }, + { + "epoch": 0.174336, + "grad_norm": 0.06514398790138742, + "learning_rate": 9.916906926766676e-05, + "loss": 0.0294, + "step": 54480 + }, + { + "epoch": 0.1744, + "grad_norm": 0.0981508353853162, + "learning_rate": 9.916846077136548e-05, + "loss": 0.0314, + "step": 54500 + }, + { + "epoch": 0.174464, + "grad_norm": 0.10673193529676554, + "learning_rate": 9.916785205421095e-05, + "loss": 0.0316, + "step": 54520 + }, + { + "epoch": 0.174528, + "grad_norm": 0.08738474398218177, + "learning_rate": 9.916724311620591e-05, + "loss": 0.0287, + "step": 54540 + }, + { + "epoch": 0.174592, + "grad_norm": 0.08618553136910817, + "learning_rate": 9.916663395735308e-05, + "loss": 0.0306, + "step": 54560 + }, + { + "epoch": 0.174656, + "grad_norm": 0.12411175415004129, + "learning_rate": 9.916602457765521e-05, + "loss": 0.0267, + "step": 54580 + }, + { + "epoch": 0.17472, + "grad_norm": 0.07384505602743656, + "learning_rate": 9.916541497711504e-05, + "loss": 0.0268, + "step": 54600 + }, + { + "epoch": 0.174784, + "grad_norm": 0.07063389918742245, + "learning_rate": 9.916480515573528e-05, + "loss": 0.0312, + "step": 54620 + }, + { + "epoch": 0.174848, + "grad_norm": 0.15071338236700155, + "learning_rate": 9.916419511351871e-05, + "loss": 0.032, + "step": 54640 + }, + { + "epoch": 0.174912, + "grad_norm": 0.07087289349818059, + "learning_rate": 9.916358485046804e-05, + "loss": 0.0344, + "step": 54660 + }, + { + "epoch": 0.174976, + "grad_norm": 0.07549898343791041, + "learning_rate": 9.916297436658601e-05, + "loss": 0.0352, + "step": 54680 + }, + { + "epoch": 0.17504, + "grad_norm": 0.09875310474321589, + "learning_rate": 9.916236366187539e-05, + "loss": 0.032, + "step": 54700 + }, + { + "epoch": 0.175104, + "grad_norm": 0.0821898938901705, + "learning_rate": 9.91617527363389e-05, + "loss": 0.0336, + "step": 54720 + }, + { + "epoch": 0.175168, + "grad_norm": 0.07857242068524695, + "learning_rate": 9.916114158997929e-05, + "loss": 0.0299, + "step": 54740 + }, + { + "epoch": 0.175232, + "grad_norm": 0.07516806115007266, + "learning_rate": 9.91605302227993e-05, + "loss": 0.0338, + "step": 54760 + }, + { + "epoch": 0.175296, + "grad_norm": 0.059130966495835394, + "learning_rate": 9.915991863480167e-05, + "loss": 0.0327, + "step": 54780 + }, + { + "epoch": 0.17536, + "grad_norm": 0.07625141557380999, + "learning_rate": 9.915930682598917e-05, + "loss": 0.0362, + "step": 54800 + }, + { + "epoch": 0.175424, + "grad_norm": 0.09590426684034094, + "learning_rate": 9.915869479636455e-05, + "loss": 0.0337, + "step": 54820 + }, + { + "epoch": 0.175488, + "grad_norm": 0.08848114486439423, + "learning_rate": 9.915808254593053e-05, + "loss": 0.0317, + "step": 54840 + }, + { + "epoch": 0.175552, + "grad_norm": 0.06969391325088582, + "learning_rate": 9.915747007468988e-05, + "loss": 0.031, + "step": 54860 + }, + { + "epoch": 0.175616, + "grad_norm": 0.07948527225656765, + "learning_rate": 9.915685738264535e-05, + "loss": 0.0303, + "step": 54880 + }, + { + "epoch": 0.17568, + "grad_norm": 0.06635544703537367, + "learning_rate": 9.915624446979968e-05, + "loss": 0.0314, + "step": 54900 + }, + { + "epoch": 0.175744, + "grad_norm": 0.09018055230969806, + "learning_rate": 9.915563133615564e-05, + "loss": 0.0317, + "step": 54920 + }, + { + "epoch": 0.175808, + "grad_norm": 0.0908287043778242, + "learning_rate": 9.915501798171596e-05, + "loss": 0.0274, + "step": 54940 + }, + { + "epoch": 0.175872, + "grad_norm": 0.09102187716540129, + "learning_rate": 9.915440440648343e-05, + "loss": 0.0299, + "step": 54960 + }, + { + "epoch": 0.175936, + "grad_norm": 0.10687802315299476, + "learning_rate": 9.915379061046078e-05, + "loss": 0.0291, + "step": 54980 + }, + { + "epoch": 0.176, + "grad_norm": 0.06500928746817333, + "learning_rate": 9.915317659365077e-05, + "loss": 0.0304, + "step": 55000 + }, + { + "epoch": 0.176064, + "grad_norm": 0.07576406356013711, + "learning_rate": 9.915256235605616e-05, + "loss": 0.0282, + "step": 55020 + }, + { + "epoch": 0.176128, + "grad_norm": 0.06007409050935332, + "learning_rate": 9.915194789767973e-05, + "loss": 0.034, + "step": 55040 + }, + { + "epoch": 0.176192, + "grad_norm": 0.07412818693153386, + "learning_rate": 9.915133321852421e-05, + "loss": 0.0322, + "step": 55060 + }, + { + "epoch": 0.176256, + "grad_norm": 0.08356303685786468, + "learning_rate": 9.915071831859238e-05, + "loss": 0.0311, + "step": 55080 + }, + { + "epoch": 0.17632, + "grad_norm": 0.07137296619871143, + "learning_rate": 9.915010319788697e-05, + "loss": 0.0312, + "step": 55100 + }, + { + "epoch": 0.176384, + "grad_norm": 0.1060604576416164, + "learning_rate": 9.914948785641078e-05, + "loss": 0.029, + "step": 55120 + }, + { + "epoch": 0.176448, + "grad_norm": 0.06578685231982229, + "learning_rate": 9.914887229416657e-05, + "loss": 0.0287, + "step": 55140 + }, + { + "epoch": 0.176512, + "grad_norm": 0.07273049488463115, + "learning_rate": 9.914825651115708e-05, + "loss": 0.0317, + "step": 55160 + }, + { + "epoch": 0.176576, + "grad_norm": 0.0642781572805071, + "learning_rate": 9.91476405073851e-05, + "loss": 0.0289, + "step": 55180 + }, + { + "epoch": 0.17664, + "grad_norm": 0.07139156284996613, + "learning_rate": 9.914702428285339e-05, + "loss": 0.0325, + "step": 55200 + }, + { + "epoch": 0.176704, + "grad_norm": 0.06074867582628363, + "learning_rate": 9.914640783756472e-05, + "loss": 0.0319, + "step": 55220 + }, + { + "epoch": 0.176768, + "grad_norm": 0.07114683341543115, + "learning_rate": 9.914579117152184e-05, + "loss": 0.0294, + "step": 55240 + }, + { + "epoch": 0.176832, + "grad_norm": 0.10528300027129514, + "learning_rate": 9.914517428472755e-05, + "loss": 0.0325, + "step": 55260 + }, + { + "epoch": 0.176896, + "grad_norm": 0.06063191410766032, + "learning_rate": 9.914455717718461e-05, + "loss": 0.0293, + "step": 55280 + }, + { + "epoch": 0.17696, + "grad_norm": 0.12330157857593224, + "learning_rate": 9.914393984889577e-05, + "loss": 0.0295, + "step": 55300 + }, + { + "epoch": 0.177024, + "grad_norm": 0.07862610818068834, + "learning_rate": 9.914332229986382e-05, + "loss": 0.0329, + "step": 55320 + }, + { + "epoch": 0.177088, + "grad_norm": 0.07294536494491552, + "learning_rate": 9.914270453009155e-05, + "loss": 0.028, + "step": 55340 + }, + { + "epoch": 0.177152, + "grad_norm": 0.07598019375089324, + "learning_rate": 9.914208653958173e-05, + "loss": 0.03, + "step": 55360 + }, + { + "epoch": 0.177216, + "grad_norm": 0.07623797873976122, + "learning_rate": 9.914146832833711e-05, + "loss": 0.0287, + "step": 55380 + }, + { + "epoch": 0.17728, + "grad_norm": 0.054833091014180844, + "learning_rate": 9.914084989636047e-05, + "loss": 0.0322, + "step": 55400 + }, + { + "epoch": 0.177344, + "grad_norm": 0.05279147084374848, + "learning_rate": 9.914023124365464e-05, + "loss": 0.0285, + "step": 55420 + }, + { + "epoch": 0.177408, + "grad_norm": 0.05924931725606422, + "learning_rate": 9.913961237022232e-05, + "loss": 0.0298, + "step": 55440 + }, + { + "epoch": 0.177472, + "grad_norm": 0.1023072094207989, + "learning_rate": 9.913899327606635e-05, + "loss": 0.0306, + "step": 55460 + }, + { + "epoch": 0.177536, + "grad_norm": 0.12688561859100922, + "learning_rate": 9.913837396118947e-05, + "loss": 0.0303, + "step": 55480 + }, + { + "epoch": 0.1776, + "grad_norm": 0.10453511488677909, + "learning_rate": 9.913775442559452e-05, + "loss": 0.0304, + "step": 55500 + }, + { + "epoch": 0.177664, + "grad_norm": 0.1981834741506519, + "learning_rate": 9.91371346692842e-05, + "loss": 0.0314, + "step": 55520 + }, + { + "epoch": 0.177728, + "grad_norm": 0.10989328040554035, + "learning_rate": 9.913651469226138e-05, + "loss": 0.0352, + "step": 55540 + }, + { + "epoch": 0.177792, + "grad_norm": 0.1289524936130172, + "learning_rate": 9.913589449452879e-05, + "loss": 0.0335, + "step": 55560 + }, + { + "epoch": 0.177856, + "grad_norm": 0.07248589788477582, + "learning_rate": 9.913527407608921e-05, + "loss": 0.0329, + "step": 55580 + }, + { + "epoch": 0.17792, + "grad_norm": 0.05135868207397774, + "learning_rate": 9.913465343694545e-05, + "loss": 0.0293, + "step": 55600 + }, + { + "epoch": 0.177984, + "grad_norm": 0.10482562660220372, + "learning_rate": 9.91340325771003e-05, + "loss": 0.0333, + "step": 55620 + }, + { + "epoch": 0.178048, + "grad_norm": 0.08948631843297361, + "learning_rate": 9.913341149655654e-05, + "loss": 0.0288, + "step": 55640 + }, + { + "epoch": 0.178112, + "grad_norm": 0.05271922212389456, + "learning_rate": 9.913279019531699e-05, + "loss": 0.0311, + "step": 55660 + }, + { + "epoch": 0.178176, + "grad_norm": 0.06146062059818811, + "learning_rate": 9.913216867338437e-05, + "loss": 0.0289, + "step": 55680 + }, + { + "epoch": 0.17824, + "grad_norm": 0.13385441041291413, + "learning_rate": 9.913154693076154e-05, + "loss": 0.03, + "step": 55700 + }, + { + "epoch": 0.178304, + "grad_norm": 0.09013751884894944, + "learning_rate": 9.913092496745126e-05, + "loss": 0.0324, + "step": 55720 + }, + { + "epoch": 0.178368, + "grad_norm": 0.07695687654990913, + "learning_rate": 9.913030278345634e-05, + "loss": 0.0316, + "step": 55740 + }, + { + "epoch": 0.178432, + "grad_norm": 0.10432404450418784, + "learning_rate": 9.912968037877954e-05, + "loss": 0.029, + "step": 55760 + }, + { + "epoch": 0.178496, + "grad_norm": 0.15972553269282552, + "learning_rate": 9.912905775342371e-05, + "loss": 0.0312, + "step": 55780 + }, + { + "epoch": 0.17856, + "grad_norm": 0.07375013758832, + "learning_rate": 9.912843490739161e-05, + "loss": 0.0324, + "step": 55800 + }, + { + "epoch": 0.178624, + "grad_norm": 0.07781443787466266, + "learning_rate": 9.912781184068604e-05, + "loss": 0.0272, + "step": 55820 + }, + { + "epoch": 0.178688, + "grad_norm": 0.07564892050891706, + "learning_rate": 9.912718855330981e-05, + "loss": 0.0331, + "step": 55840 + }, + { + "epoch": 0.178752, + "grad_norm": 0.10414462047560454, + "learning_rate": 9.912656504526572e-05, + "loss": 0.0366, + "step": 55860 + }, + { + "epoch": 0.178816, + "grad_norm": 0.13701186428373793, + "learning_rate": 9.912594131655655e-05, + "loss": 0.0315, + "step": 55880 + }, + { + "epoch": 0.17888, + "grad_norm": 0.0748067491132546, + "learning_rate": 9.912531736718513e-05, + "loss": 0.0272, + "step": 55900 + }, + { + "epoch": 0.178944, + "grad_norm": 0.12545775741132228, + "learning_rate": 9.912469319715423e-05, + "loss": 0.0311, + "step": 55920 + }, + { + "epoch": 0.179008, + "grad_norm": 0.061478823344628876, + "learning_rate": 9.912406880646669e-05, + "loss": 0.0305, + "step": 55940 + }, + { + "epoch": 0.179072, + "grad_norm": 0.14728005651704806, + "learning_rate": 9.91234441951253e-05, + "loss": 0.0312, + "step": 55960 + }, + { + "epoch": 0.179136, + "grad_norm": 0.07069137787231027, + "learning_rate": 9.912281936313285e-05, + "loss": 0.0314, + "step": 55980 + }, + { + "epoch": 0.1792, + "grad_norm": 0.060715287830979604, + "learning_rate": 9.912219431049217e-05, + "loss": 0.0302, + "step": 56000 + }, + { + "epoch": 0.179264, + "grad_norm": 0.12350719048450609, + "learning_rate": 9.912156903720604e-05, + "loss": 0.0301, + "step": 56020 + }, + { + "epoch": 0.179328, + "grad_norm": 0.10273929707499123, + "learning_rate": 9.91209435432773e-05, + "loss": 0.0298, + "step": 56040 + }, + { + "epoch": 0.179392, + "grad_norm": 0.09410861765919187, + "learning_rate": 9.912031782870874e-05, + "loss": 0.0307, + "step": 56060 + }, + { + "epoch": 0.179456, + "grad_norm": 0.08169279470106103, + "learning_rate": 9.911969189350318e-05, + "loss": 0.0295, + "step": 56080 + }, + { + "epoch": 0.17952, + "grad_norm": 0.05309547248520194, + "learning_rate": 9.911906573766342e-05, + "loss": 0.0297, + "step": 56100 + }, + { + "epoch": 0.179584, + "grad_norm": 0.11849248391154033, + "learning_rate": 9.91184393611923e-05, + "loss": 0.0353, + "step": 56120 + }, + { + "epoch": 0.179648, + "grad_norm": 0.07189183414839781, + "learning_rate": 9.91178127640926e-05, + "loss": 0.0324, + "step": 56140 + }, + { + "epoch": 0.179712, + "grad_norm": 0.1699438769631181, + "learning_rate": 9.911718594636714e-05, + "loss": 0.0323, + "step": 56160 + }, + { + "epoch": 0.179776, + "grad_norm": 0.07653701406980844, + "learning_rate": 9.911655890801875e-05, + "loss": 0.0272, + "step": 56180 + }, + { + "epoch": 0.17984, + "grad_norm": 0.07536764145566109, + "learning_rate": 9.911593164905024e-05, + "loss": 0.0317, + "step": 56200 + }, + { + "epoch": 0.179904, + "grad_norm": 0.11038785629926327, + "learning_rate": 9.911530416946444e-05, + "loss": 0.0303, + "step": 56220 + }, + { + "epoch": 0.179968, + "grad_norm": 0.12312996283196055, + "learning_rate": 9.911467646926414e-05, + "loss": 0.0375, + "step": 56240 + }, + { + "epoch": 0.180032, + "grad_norm": 0.0702584870158073, + "learning_rate": 9.911404854845218e-05, + "loss": 0.0311, + "step": 56260 + }, + { + "epoch": 0.180096, + "grad_norm": 0.07530939726463182, + "learning_rate": 9.911342040703138e-05, + "loss": 0.0334, + "step": 56280 + }, + { + "epoch": 0.18016, + "grad_norm": 0.07459130549840669, + "learning_rate": 9.911279204500454e-05, + "loss": 0.0307, + "step": 56300 + }, + { + "epoch": 0.180224, + "grad_norm": 0.05065137758340631, + "learning_rate": 9.911216346237452e-05, + "loss": 0.0301, + "step": 56320 + }, + { + "epoch": 0.180288, + "grad_norm": 0.14169477434264507, + "learning_rate": 9.911153465914411e-05, + "loss": 0.03, + "step": 56340 + }, + { + "epoch": 0.180352, + "grad_norm": 0.0648182406956438, + "learning_rate": 9.911090563531615e-05, + "loss": 0.0294, + "step": 56360 + }, + { + "epoch": 0.180416, + "grad_norm": 0.06690949858181038, + "learning_rate": 9.911027639089348e-05, + "loss": 0.0299, + "step": 56380 + }, + { + "epoch": 0.18048, + "grad_norm": 0.06859594889185171, + "learning_rate": 9.910964692587889e-05, + "loss": 0.033, + "step": 56400 + }, + { + "epoch": 0.180544, + "grad_norm": 0.08307587472699843, + "learning_rate": 9.910901724027524e-05, + "loss": 0.0302, + "step": 56420 + }, + { + "epoch": 0.180608, + "grad_norm": 0.08209489852014086, + "learning_rate": 9.910838733408534e-05, + "loss": 0.0301, + "step": 56440 + }, + { + "epoch": 0.180672, + "grad_norm": 0.058682850242809934, + "learning_rate": 9.910775720731203e-05, + "loss": 0.0323, + "step": 56460 + }, + { + "epoch": 0.180736, + "grad_norm": 0.09752275319418642, + "learning_rate": 9.910712685995814e-05, + "loss": 0.0335, + "step": 56480 + }, + { + "epoch": 0.1808, + "grad_norm": 0.08555267136679671, + "learning_rate": 9.910649629202647e-05, + "loss": 0.0304, + "step": 56500 + }, + { + "epoch": 0.180864, + "grad_norm": 0.07368800937799537, + "learning_rate": 9.910586550351991e-05, + "loss": 0.0285, + "step": 56520 + }, + { + "epoch": 0.180928, + "grad_norm": 0.07019543441877546, + "learning_rate": 9.910523449444123e-05, + "loss": 0.0349, + "step": 56540 + }, + { + "epoch": 0.180992, + "grad_norm": 0.08171540310539172, + "learning_rate": 9.910460326479332e-05, + "loss": 0.0316, + "step": 56560 + }, + { + "epoch": 0.181056, + "grad_norm": 0.05815060868856384, + "learning_rate": 9.9103971814579e-05, + "loss": 0.0313, + "step": 56580 + }, + { + "epoch": 0.18112, + "grad_norm": 0.0737736967418381, + "learning_rate": 9.910334014380107e-05, + "loss": 0.0344, + "step": 56600 + }, + { + "epoch": 0.181184, + "grad_norm": 0.09170896839940221, + "learning_rate": 9.910270825246242e-05, + "loss": 0.0278, + "step": 56620 + }, + { + "epoch": 0.181248, + "grad_norm": 0.06760672076632138, + "learning_rate": 9.910207614056584e-05, + "loss": 0.0298, + "step": 56640 + }, + { + "epoch": 0.181312, + "grad_norm": 0.08431859265369361, + "learning_rate": 9.910144380811421e-05, + "loss": 0.0313, + "step": 56660 + }, + { + "epoch": 0.181376, + "grad_norm": 0.10711550949603708, + "learning_rate": 9.910081125511035e-05, + "loss": 0.031, + "step": 56680 + }, + { + "epoch": 0.18144, + "grad_norm": 0.08293984111946934, + "learning_rate": 9.910017848155711e-05, + "loss": 0.0334, + "step": 56700 + }, + { + "epoch": 0.181504, + "grad_norm": 0.083191835354564, + "learning_rate": 9.909954548745732e-05, + "loss": 0.0297, + "step": 56720 + }, + { + "epoch": 0.181568, + "grad_norm": 0.07573444360745306, + "learning_rate": 9.909891227281384e-05, + "loss": 0.0301, + "step": 56740 + }, + { + "epoch": 0.181632, + "grad_norm": 0.05596412601630615, + "learning_rate": 9.90982788376295e-05, + "loss": 0.0312, + "step": 56760 + }, + { + "epoch": 0.181696, + "grad_norm": 0.08605031321703613, + "learning_rate": 9.909764518190714e-05, + "loss": 0.0323, + "step": 56780 + }, + { + "epoch": 0.18176, + "grad_norm": 0.1015272472593763, + "learning_rate": 9.909701130564963e-05, + "loss": 0.0301, + "step": 56800 + }, + { + "epoch": 0.181824, + "grad_norm": 0.10415556897657033, + "learning_rate": 9.90963772088598e-05, + "loss": 0.0321, + "step": 56820 + }, + { + "epoch": 0.181888, + "grad_norm": 0.12408160262975111, + "learning_rate": 9.909574289154049e-05, + "loss": 0.0304, + "step": 56840 + }, + { + "epoch": 0.181952, + "grad_norm": 0.09297772925523458, + "learning_rate": 9.909510835369457e-05, + "loss": 0.0295, + "step": 56860 + }, + { + "epoch": 0.182016, + "grad_norm": 0.05305486209263366, + "learning_rate": 9.909447359532489e-05, + "loss": 0.0328, + "step": 56880 + }, + { + "epoch": 0.18208, + "grad_norm": 0.10102566198341581, + "learning_rate": 9.909383861643428e-05, + "loss": 0.0314, + "step": 56900 + }, + { + "epoch": 0.182144, + "grad_norm": 0.07893723495202629, + "learning_rate": 9.909320341702561e-05, + "loss": 0.0316, + "step": 56920 + }, + { + "epoch": 0.182208, + "grad_norm": 0.05664638759214131, + "learning_rate": 9.909256799710171e-05, + "loss": 0.0321, + "step": 56940 + }, + { + "epoch": 0.182272, + "grad_norm": 0.1125596071707452, + "learning_rate": 9.909193235666547e-05, + "loss": 0.033, + "step": 56960 + }, + { + "epoch": 0.182336, + "grad_norm": 0.1737896772541999, + "learning_rate": 9.909129649571973e-05, + "loss": 0.0335, + "step": 56980 + }, + { + "epoch": 0.1824, + "grad_norm": 0.09583169543818493, + "learning_rate": 9.909066041426733e-05, + "loss": 0.0302, + "step": 57000 + }, + { + "epoch": 0.182464, + "grad_norm": 0.07532716741546423, + "learning_rate": 9.909002411231115e-05, + "loss": 0.0331, + "step": 57020 + }, + { + "epoch": 0.182528, + "grad_norm": 0.08963896438603018, + "learning_rate": 9.908938758985404e-05, + "loss": 0.0293, + "step": 57040 + }, + { + "epoch": 0.182592, + "grad_norm": 0.08439617570155365, + "learning_rate": 9.908875084689884e-05, + "loss": 0.0317, + "step": 57060 + }, + { + "epoch": 0.182656, + "grad_norm": 0.0648892001702721, + "learning_rate": 9.908811388344845e-05, + "loss": 0.0354, + "step": 57080 + }, + { + "epoch": 0.18272, + "grad_norm": 0.1278686454188513, + "learning_rate": 9.908747669950568e-05, + "loss": 0.0311, + "step": 57100 + }, + { + "epoch": 0.182784, + "grad_norm": 0.06124289491049138, + "learning_rate": 9.908683929507346e-05, + "loss": 0.031, + "step": 57120 + }, + { + "epoch": 0.182848, + "grad_norm": 0.05422858781977851, + "learning_rate": 9.908620167015458e-05, + "loss": 0.0306, + "step": 57140 + }, + { + "epoch": 0.182912, + "grad_norm": 0.1723098831529928, + "learning_rate": 9.908556382475195e-05, + "loss": 0.0318, + "step": 57160 + }, + { + "epoch": 0.182976, + "grad_norm": 0.08604328820402815, + "learning_rate": 9.908492575886842e-05, + "loss": 0.0306, + "step": 57180 + }, + { + "epoch": 0.18304, + "grad_norm": 0.07506632504164326, + "learning_rate": 9.908428747250686e-05, + "loss": 0.032, + "step": 57200 + }, + { + "epoch": 0.183104, + "grad_norm": 0.06073716602567336, + "learning_rate": 9.908364896567013e-05, + "loss": 0.0291, + "step": 57220 + }, + { + "epoch": 0.183168, + "grad_norm": 0.06555171283517904, + "learning_rate": 9.90830102383611e-05, + "loss": 0.0305, + "step": 57240 + }, + { + "epoch": 0.183232, + "grad_norm": 0.06347101171198498, + "learning_rate": 9.908237129058267e-05, + "loss": 0.0297, + "step": 57260 + }, + { + "epoch": 0.183296, + "grad_norm": 0.11978616834427984, + "learning_rate": 9.908173212233767e-05, + "loss": 0.0331, + "step": 57280 + }, + { + "epoch": 0.18336, + "grad_norm": 0.09550141819550205, + "learning_rate": 9.908109273362897e-05, + "loss": 0.0296, + "step": 57300 + }, + { + "epoch": 0.183424, + "grad_norm": 0.06810663842852016, + "learning_rate": 9.908045312445946e-05, + "loss": 0.0304, + "step": 57320 + }, + { + "epoch": 0.183488, + "grad_norm": 0.14501367504534543, + "learning_rate": 9.907981329483202e-05, + "loss": 0.0304, + "step": 57340 + }, + { + "epoch": 0.183552, + "grad_norm": 0.0554265901895841, + "learning_rate": 9.907917324474951e-05, + "loss": 0.0297, + "step": 57360 + }, + { + "epoch": 0.183616, + "grad_norm": 0.08710988017084123, + "learning_rate": 9.90785329742148e-05, + "loss": 0.0329, + "step": 57380 + }, + { + "epoch": 0.18368, + "grad_norm": 0.05330987866371492, + "learning_rate": 9.907789248323078e-05, + "loss": 0.0312, + "step": 57400 + }, + { + "epoch": 0.183744, + "grad_norm": 0.07772023653561798, + "learning_rate": 9.907725177180032e-05, + "loss": 0.0262, + "step": 57420 + }, + { + "epoch": 0.183808, + "grad_norm": 0.09233791667003935, + "learning_rate": 9.90766108399263e-05, + "loss": 0.0315, + "step": 57440 + }, + { + "epoch": 0.183872, + "grad_norm": 0.19729096416239655, + "learning_rate": 9.907596968761159e-05, + "loss": 0.0326, + "step": 57460 + }, + { + "epoch": 0.183936, + "grad_norm": 0.06460342799204342, + "learning_rate": 9.907532831485909e-05, + "loss": 0.0338, + "step": 57480 + }, + { + "epoch": 0.184, + "grad_norm": 0.1342037208495233, + "learning_rate": 9.907468672167165e-05, + "loss": 0.0316, + "step": 57500 + }, + { + "epoch": 0.184064, + "grad_norm": 0.16442538588831396, + "learning_rate": 9.907404490805218e-05, + "loss": 0.0325, + "step": 57520 + }, + { + "epoch": 0.184128, + "grad_norm": 0.08156563622311923, + "learning_rate": 9.907340287400354e-05, + "loss": 0.0317, + "step": 57540 + }, + { + "epoch": 0.184192, + "grad_norm": 0.08684543433942785, + "learning_rate": 9.907276061952865e-05, + "loss": 0.0337, + "step": 57560 + }, + { + "epoch": 0.184256, + "grad_norm": 0.07410014110286034, + "learning_rate": 9.907211814463037e-05, + "loss": 0.0342, + "step": 57580 + }, + { + "epoch": 0.18432, + "grad_norm": 0.05480608773783919, + "learning_rate": 9.907147544931157e-05, + "loss": 0.0345, + "step": 57600 + }, + { + "epoch": 0.184384, + "grad_norm": 0.05906528458308589, + "learning_rate": 9.907083253357516e-05, + "loss": 0.0326, + "step": 57620 + }, + { + "epoch": 0.184448, + "grad_norm": 0.08903951766926992, + "learning_rate": 9.907018939742403e-05, + "loss": 0.0323, + "step": 57640 + }, + { + "epoch": 0.184512, + "grad_norm": 0.07081410785107872, + "learning_rate": 9.906954604086106e-05, + "loss": 0.0292, + "step": 57660 + }, + { + "epoch": 0.184576, + "grad_norm": 0.07586256810217905, + "learning_rate": 9.906890246388914e-05, + "loss": 0.0292, + "step": 57680 + }, + { + "epoch": 0.18464, + "grad_norm": 0.0744451499589352, + "learning_rate": 9.906825866651116e-05, + "loss": 0.0334, + "step": 57700 + }, + { + "epoch": 0.184704, + "grad_norm": 0.12983934858093307, + "learning_rate": 9.906761464873002e-05, + "loss": 0.0325, + "step": 57720 + }, + { + "epoch": 0.184768, + "grad_norm": 0.08830877520963591, + "learning_rate": 9.906697041054859e-05, + "loss": 0.031, + "step": 57740 + }, + { + "epoch": 0.184832, + "grad_norm": 0.11877272113202703, + "learning_rate": 9.906632595196979e-05, + "loss": 0.0286, + "step": 57760 + }, + { + "epoch": 0.184896, + "grad_norm": 0.04945080611780583, + "learning_rate": 9.90656812729965e-05, + "loss": 0.0331, + "step": 57780 + }, + { + "epoch": 0.18496, + "grad_norm": 0.07139052051405267, + "learning_rate": 9.906503637363162e-05, + "loss": 0.0323, + "step": 57800 + }, + { + "epoch": 0.185024, + "grad_norm": 0.06954108646148342, + "learning_rate": 9.906439125387805e-05, + "loss": 0.0307, + "step": 57820 + }, + { + "epoch": 0.185088, + "grad_norm": 0.14346285593146796, + "learning_rate": 9.90637459137387e-05, + "loss": 0.0295, + "step": 57840 + }, + { + "epoch": 0.185152, + "grad_norm": 0.11312972083641779, + "learning_rate": 9.906310035321644e-05, + "loss": 0.0298, + "step": 57860 + }, + { + "epoch": 0.185216, + "grad_norm": 0.05271038304640455, + "learning_rate": 9.906245457231418e-05, + "loss": 0.031, + "step": 57880 + }, + { + "epoch": 0.18528, + "grad_norm": 0.12532956372097634, + "learning_rate": 9.906180857103483e-05, + "loss": 0.0335, + "step": 57900 + }, + { + "epoch": 0.185344, + "grad_norm": 0.12245001381854656, + "learning_rate": 9.906116234938127e-05, + "loss": 0.0327, + "step": 57920 + }, + { + "epoch": 0.185408, + "grad_norm": 0.0937271583837394, + "learning_rate": 9.906051590735644e-05, + "loss": 0.0303, + "step": 57940 + }, + { + "epoch": 0.185472, + "grad_norm": 0.14448225250330357, + "learning_rate": 9.905986924496323e-05, + "loss": 0.0289, + "step": 57960 + }, + { + "epoch": 0.185536, + "grad_norm": 0.060081288924044005, + "learning_rate": 9.905922236220451e-05, + "loss": 0.0331, + "step": 57980 + }, + { + "epoch": 0.1856, + "grad_norm": 0.07078324606397143, + "learning_rate": 9.905857525908322e-05, + "loss": 0.0358, + "step": 58000 + }, + { + "epoch": 0.185664, + "grad_norm": 0.05982410503298079, + "learning_rate": 9.905792793560227e-05, + "loss": 0.0292, + "step": 58020 + }, + { + "epoch": 0.185728, + "grad_norm": 0.07724324963379274, + "learning_rate": 9.905728039176455e-05, + "loss": 0.0305, + "step": 58040 + }, + { + "epoch": 0.185792, + "grad_norm": 0.1685445141227329, + "learning_rate": 9.905663262757296e-05, + "loss": 0.0323, + "step": 58060 + }, + { + "epoch": 0.185856, + "grad_norm": 0.06875668170257895, + "learning_rate": 9.905598464303045e-05, + "loss": 0.0337, + "step": 58080 + }, + { + "epoch": 0.18592, + "grad_norm": 0.05158441186128389, + "learning_rate": 9.905533643813989e-05, + "loss": 0.03, + "step": 58100 + }, + { + "epoch": 0.185984, + "grad_norm": 0.09123109076352655, + "learning_rate": 9.905468801290421e-05, + "loss": 0.0297, + "step": 58120 + }, + { + "epoch": 0.186048, + "grad_norm": 0.08462838839491714, + "learning_rate": 9.905403936732632e-05, + "loss": 0.0313, + "step": 58140 + }, + { + "epoch": 0.186112, + "grad_norm": 0.09346657234491286, + "learning_rate": 9.905339050140914e-05, + "loss": 0.0337, + "step": 58160 + }, + { + "epoch": 0.186176, + "grad_norm": 0.0768935643514365, + "learning_rate": 9.905274141515557e-05, + "loss": 0.0339, + "step": 58180 + }, + { + "epoch": 0.18624, + "grad_norm": 0.09331706088229727, + "learning_rate": 9.905209210856853e-05, + "loss": 0.0315, + "step": 58200 + }, + { + "epoch": 0.186304, + "grad_norm": 0.07275419801303047, + "learning_rate": 9.905144258165096e-05, + "loss": 0.0328, + "step": 58220 + }, + { + "epoch": 0.186368, + "grad_norm": 0.0757349027872123, + "learning_rate": 9.905079283440573e-05, + "loss": 0.029, + "step": 58240 + }, + { + "epoch": 0.186432, + "grad_norm": 0.0669475304033215, + "learning_rate": 9.90501428668358e-05, + "loss": 0.0322, + "step": 58260 + }, + { + "epoch": 0.186496, + "grad_norm": 0.08847028405017106, + "learning_rate": 9.904949267894408e-05, + "loss": 0.0283, + "step": 58280 + }, + { + "epoch": 0.18656, + "grad_norm": 0.06909242038783779, + "learning_rate": 9.904884227073348e-05, + "loss": 0.0325, + "step": 58300 + }, + { + "epoch": 0.186624, + "grad_norm": 0.19995549201112395, + "learning_rate": 9.904819164220694e-05, + "loss": 0.0321, + "step": 58320 + }, + { + "epoch": 0.186688, + "grad_norm": 0.0665049616967318, + "learning_rate": 9.904754079336734e-05, + "loss": 0.0314, + "step": 58340 + }, + { + "epoch": 0.186752, + "grad_norm": 0.10241153882803158, + "learning_rate": 9.904688972421765e-05, + "loss": 0.0338, + "step": 58360 + }, + { + "epoch": 0.186816, + "grad_norm": 0.13692861313251287, + "learning_rate": 9.904623843476079e-05, + "loss": 0.03, + "step": 58380 + }, + { + "epoch": 0.18688, + "grad_norm": 0.07898760181456554, + "learning_rate": 9.904558692499966e-05, + "loss": 0.034, + "step": 58400 + }, + { + "epoch": 0.186944, + "grad_norm": 0.05055437891294893, + "learning_rate": 9.904493519493719e-05, + "loss": 0.0313, + "step": 58420 + }, + { + "epoch": 0.187008, + "grad_norm": 0.07358526720126503, + "learning_rate": 9.904428324457634e-05, + "loss": 0.0289, + "step": 58440 + }, + { + "epoch": 0.187072, + "grad_norm": 0.08088897128388826, + "learning_rate": 9.904363107391999e-05, + "loss": 0.0322, + "step": 58460 + }, + { + "epoch": 0.187136, + "grad_norm": 0.08869791717377383, + "learning_rate": 9.904297868297112e-05, + "loss": 0.0307, + "step": 58480 + }, + { + "epoch": 0.1872, + "grad_norm": 0.147415059764648, + "learning_rate": 9.904232607173262e-05, + "loss": 0.0308, + "step": 58500 + }, + { + "epoch": 0.187264, + "grad_norm": 0.09008447669058428, + "learning_rate": 9.904167324020743e-05, + "loss": 0.0301, + "step": 58520 + }, + { + "epoch": 0.187328, + "grad_norm": 0.06257975273729069, + "learning_rate": 9.904102018839851e-05, + "loss": 0.0285, + "step": 58540 + }, + { + "epoch": 0.187392, + "grad_norm": 0.09236922774292053, + "learning_rate": 9.904036691630875e-05, + "loss": 0.0283, + "step": 58560 + }, + { + "epoch": 0.187456, + "grad_norm": 0.05783479242944942, + "learning_rate": 9.903971342394114e-05, + "loss": 0.0327, + "step": 58580 + }, + { + "epoch": 0.18752, + "grad_norm": 0.0642322331243713, + "learning_rate": 9.903905971129854e-05, + "loss": 0.0332, + "step": 58600 + }, + { + "epoch": 0.187584, + "grad_norm": 0.054949868027486734, + "learning_rate": 9.903840577838395e-05, + "loss": 0.0289, + "step": 58620 + }, + { + "epoch": 0.187648, + "grad_norm": 0.059549791903426685, + "learning_rate": 9.903775162520029e-05, + "loss": 0.0316, + "step": 58640 + }, + { + "epoch": 0.187712, + "grad_norm": 0.09237456590644062, + "learning_rate": 9.90370972517505e-05, + "loss": 0.0333, + "step": 58660 + }, + { + "epoch": 0.187776, + "grad_norm": 0.09398873717077447, + "learning_rate": 9.903644265803748e-05, + "loss": 0.0336, + "step": 58680 + }, + { + "epoch": 0.18784, + "grad_norm": 0.1218349700234445, + "learning_rate": 9.903578784406424e-05, + "loss": 0.037, + "step": 58700 + }, + { + "epoch": 0.187904, + "grad_norm": 0.09955248436369646, + "learning_rate": 9.903513280983368e-05, + "loss": 0.0295, + "step": 58720 + }, + { + "epoch": 0.187968, + "grad_norm": 0.05870748690406761, + "learning_rate": 9.903447755534873e-05, + "loss": 0.0292, + "step": 58740 + }, + { + "epoch": 0.188032, + "grad_norm": 0.11603883656493827, + "learning_rate": 9.903382208061236e-05, + "loss": 0.0317, + "step": 58760 + }, + { + "epoch": 0.188096, + "grad_norm": 0.06554559506074953, + "learning_rate": 9.90331663856275e-05, + "loss": 0.0286, + "step": 58780 + }, + { + "epoch": 0.18816, + "grad_norm": 0.07941833638504525, + "learning_rate": 9.90325104703971e-05, + "loss": 0.0317, + "step": 58800 + }, + { + "epoch": 0.188224, + "grad_norm": 0.09202662340089272, + "learning_rate": 9.90318543349241e-05, + "loss": 0.031, + "step": 58820 + }, + { + "epoch": 0.188288, + "grad_norm": 0.07856832407519056, + "learning_rate": 9.903119797921147e-05, + "loss": 0.0288, + "step": 58840 + }, + { + "epoch": 0.188352, + "grad_norm": 0.07439075329634934, + "learning_rate": 9.903054140326214e-05, + "loss": 0.0307, + "step": 58860 + }, + { + "epoch": 0.188416, + "grad_norm": 0.06984042006854699, + "learning_rate": 9.902988460707904e-05, + "loss": 0.0294, + "step": 58880 + }, + { + "epoch": 0.18848, + "grad_norm": 0.09006171841011561, + "learning_rate": 9.902922759066515e-05, + "loss": 0.0304, + "step": 58900 + }, + { + "epoch": 0.188544, + "grad_norm": 0.09392771606545723, + "learning_rate": 9.902857035402342e-05, + "loss": 0.0295, + "step": 58920 + }, + { + "epoch": 0.188608, + "grad_norm": 0.08471208694136559, + "learning_rate": 9.902791289715678e-05, + "loss": 0.0292, + "step": 58940 + }, + { + "epoch": 0.188672, + "grad_norm": 0.07998655826277877, + "learning_rate": 9.902725522006821e-05, + "loss": 0.0341, + "step": 58960 + }, + { + "epoch": 0.188736, + "grad_norm": 0.06395319894806595, + "learning_rate": 9.902659732276065e-05, + "loss": 0.0303, + "step": 58980 + }, + { + "epoch": 0.1888, + "grad_norm": 0.16468244870864002, + "learning_rate": 9.902593920523706e-05, + "loss": 0.0344, + "step": 59000 + }, + { + "epoch": 0.188864, + "grad_norm": 0.08946765415230847, + "learning_rate": 9.902528086750038e-05, + "loss": 0.0313, + "step": 59020 + }, + { + "epoch": 0.188928, + "grad_norm": 0.06601339592631864, + "learning_rate": 9.90246223095536e-05, + "loss": 0.0291, + "step": 59040 + }, + { + "epoch": 0.188992, + "grad_norm": 0.07364500215047344, + "learning_rate": 9.902396353139965e-05, + "loss": 0.0297, + "step": 59060 + }, + { + "epoch": 0.189056, + "grad_norm": 0.1484921011831819, + "learning_rate": 9.902330453304147e-05, + "loss": 0.0316, + "step": 59080 + }, + { + "epoch": 0.18912, + "grad_norm": 0.12532553248299033, + "learning_rate": 9.902264531448208e-05, + "loss": 0.0284, + "step": 59100 + }, + { + "epoch": 0.189184, + "grad_norm": 0.12709477430716265, + "learning_rate": 9.90219858757244e-05, + "loss": 0.0276, + "step": 59120 + }, + { + "epoch": 0.189248, + "grad_norm": 0.09190557956496918, + "learning_rate": 9.90213262167714e-05, + "loss": 0.0337, + "step": 59140 + }, + { + "epoch": 0.189312, + "grad_norm": 0.09516964950831684, + "learning_rate": 9.902066633762604e-05, + "loss": 0.034, + "step": 59160 + }, + { + "epoch": 0.189376, + "grad_norm": 0.06866381880927556, + "learning_rate": 9.902000623829128e-05, + "loss": 0.0321, + "step": 59180 + }, + { + "epoch": 0.18944, + "grad_norm": 0.08075498435596498, + "learning_rate": 9.901934591877009e-05, + "loss": 0.0311, + "step": 59200 + }, + { + "epoch": 0.189504, + "grad_norm": 0.06443910038982426, + "learning_rate": 9.901868537906545e-05, + "loss": 0.0337, + "step": 59220 + }, + { + "epoch": 0.189568, + "grad_norm": 0.0794929062573456, + "learning_rate": 9.901802461918032e-05, + "loss": 0.0303, + "step": 59240 + }, + { + "epoch": 0.189632, + "grad_norm": 0.06745287089125945, + "learning_rate": 9.901736363911764e-05, + "loss": 0.0305, + "step": 59260 + }, + { + "epoch": 0.189696, + "grad_norm": 0.08099701285275905, + "learning_rate": 9.901670243888042e-05, + "loss": 0.0311, + "step": 59280 + }, + { + "epoch": 0.18976, + "grad_norm": 0.07965720398341562, + "learning_rate": 9.901604101847162e-05, + "loss": 0.0315, + "step": 59300 + }, + { + "epoch": 0.189824, + "grad_norm": 0.07593724989335497, + "learning_rate": 9.901537937789418e-05, + "loss": 0.0319, + "step": 59320 + }, + { + "epoch": 0.189888, + "grad_norm": 0.09317259336978356, + "learning_rate": 9.901471751715112e-05, + "loss": 0.0293, + "step": 59340 + }, + { + "epoch": 0.189952, + "grad_norm": 0.05905459412519879, + "learning_rate": 9.901405543624537e-05, + "loss": 0.0281, + "step": 59360 + }, + { + "epoch": 0.190016, + "grad_norm": 0.07968501745742275, + "learning_rate": 9.901339313517992e-05, + "loss": 0.0324, + "step": 59380 + }, + { + "epoch": 0.19008, + "grad_norm": 0.12536871771083333, + "learning_rate": 9.901273061395775e-05, + "loss": 0.029, + "step": 59400 + }, + { + "epoch": 0.190144, + "grad_norm": 0.2147906065747188, + "learning_rate": 9.901206787258184e-05, + "loss": 0.0305, + "step": 59420 + }, + { + "epoch": 0.190208, + "grad_norm": 0.07182313068834152, + "learning_rate": 9.901140491105516e-05, + "loss": 0.0328, + "step": 59440 + }, + { + "epoch": 0.190272, + "grad_norm": 0.10155443638246053, + "learning_rate": 9.901074172938069e-05, + "loss": 0.0308, + "step": 59460 + }, + { + "epoch": 0.190336, + "grad_norm": 0.1082382477169541, + "learning_rate": 9.90100783275614e-05, + "loss": 0.033, + "step": 59480 + }, + { + "epoch": 0.1904, + "grad_norm": 0.09647242298128025, + "learning_rate": 9.900941470560025e-05, + "loss": 0.0312, + "step": 59500 + }, + { + "epoch": 0.190464, + "grad_norm": 0.07094178732139758, + "learning_rate": 9.90087508635003e-05, + "loss": 0.0286, + "step": 59520 + }, + { + "epoch": 0.190528, + "grad_norm": 0.06474229007468127, + "learning_rate": 9.900808680126443e-05, + "loss": 0.0284, + "step": 59540 + }, + { + "epoch": 0.190592, + "grad_norm": 0.10677293858882929, + "learning_rate": 9.90074225188957e-05, + "loss": 0.0307, + "step": 59560 + }, + { + "epoch": 0.190656, + "grad_norm": 0.0537825338301271, + "learning_rate": 9.900675801639705e-05, + "loss": 0.0295, + "step": 59580 + }, + { + "epoch": 0.19072, + "grad_norm": 0.09617338246068792, + "learning_rate": 9.900609329377148e-05, + "loss": 0.0298, + "step": 59600 + }, + { + "epoch": 0.190784, + "grad_norm": 0.08111650882835705, + "learning_rate": 9.900542835102199e-05, + "loss": 0.0331, + "step": 59620 + }, + { + "epoch": 0.190848, + "grad_norm": 0.08847259118147248, + "learning_rate": 9.900476318815155e-05, + "loss": 0.0338, + "step": 59640 + }, + { + "epoch": 0.190912, + "grad_norm": 0.06817495673650092, + "learning_rate": 9.900409780516313e-05, + "loss": 0.0303, + "step": 59660 + }, + { + "epoch": 0.190976, + "grad_norm": 0.12673790494706766, + "learning_rate": 9.900343220205976e-05, + "loss": 0.0328, + "step": 59680 + }, + { + "epoch": 0.19104, + "grad_norm": 0.07623826472990251, + "learning_rate": 9.90027663788444e-05, + "loss": 0.0328, + "step": 59700 + }, + { + "epoch": 0.191104, + "grad_norm": 0.10876096146862675, + "learning_rate": 9.900210033552007e-05, + "loss": 0.0339, + "step": 59720 + }, + { + "epoch": 0.191168, + "grad_norm": 0.10399991179139303, + "learning_rate": 9.900143407208971e-05, + "loss": 0.0313, + "step": 59740 + }, + { + "epoch": 0.191232, + "grad_norm": 0.11294972424734066, + "learning_rate": 9.900076758855636e-05, + "loss": 0.0297, + "step": 59760 + }, + { + "epoch": 0.191296, + "grad_norm": 0.055402407188321415, + "learning_rate": 9.9000100884923e-05, + "loss": 0.0318, + "step": 59780 + }, + { + "epoch": 0.19136, + "grad_norm": 0.05761837819705909, + "learning_rate": 9.899943396119262e-05, + "loss": 0.0269, + "step": 59800 + }, + { + "epoch": 0.191424, + "grad_norm": 0.08164212915303873, + "learning_rate": 9.89987668173682e-05, + "loss": 0.0298, + "step": 59820 + }, + { + "epoch": 0.191488, + "grad_norm": 0.2302338665536421, + "learning_rate": 9.899809945345277e-05, + "loss": 0.0325, + "step": 59840 + }, + { + "epoch": 0.191552, + "grad_norm": 0.04968183583372562, + "learning_rate": 9.899743186944932e-05, + "loss": 0.0285, + "step": 59860 + }, + { + "epoch": 0.191616, + "grad_norm": 0.17335388495069726, + "learning_rate": 9.899676406536083e-05, + "loss": 0.0333, + "step": 59880 + }, + { + "epoch": 0.19168, + "grad_norm": 0.07784064472922508, + "learning_rate": 9.899609604119031e-05, + "loss": 0.0289, + "step": 59900 + }, + { + "epoch": 0.191744, + "grad_norm": 0.09696709547793268, + "learning_rate": 9.899542779694077e-05, + "loss": 0.032, + "step": 59920 + }, + { + "epoch": 0.191808, + "grad_norm": 0.06729906168771423, + "learning_rate": 9.89947593326152e-05, + "loss": 0.0293, + "step": 59940 + }, + { + "epoch": 0.191872, + "grad_norm": 0.06021421677168573, + "learning_rate": 9.89940906482166e-05, + "loss": 0.0329, + "step": 59960 + }, + { + "epoch": 0.191936, + "grad_norm": 0.04960356883500862, + "learning_rate": 9.899342174374798e-05, + "loss": 0.0302, + "step": 59980 + }, + { + "epoch": 0.192, + "grad_norm": 0.08190764611894477, + "learning_rate": 9.899275261921234e-05, + "loss": 0.0338, + "step": 60000 + }, + { + "epoch": 0.192064, + "grad_norm": 0.07750651514111975, + "learning_rate": 9.89920832746127e-05, + "loss": 0.0296, + "step": 60020 + }, + { + "epoch": 0.192128, + "grad_norm": 0.12068706405558143, + "learning_rate": 9.899141370995205e-05, + "loss": 0.0288, + "step": 60040 + }, + { + "epoch": 0.192192, + "grad_norm": 0.06030187931123488, + "learning_rate": 9.89907439252334e-05, + "loss": 0.0303, + "step": 60060 + }, + { + "epoch": 0.192256, + "grad_norm": 0.06329086053099804, + "learning_rate": 9.899007392045977e-05, + "loss": 0.0306, + "step": 60080 + }, + { + "epoch": 0.19232, + "grad_norm": 0.09425024788592297, + "learning_rate": 9.898940369563416e-05, + "loss": 0.031, + "step": 60100 + }, + { + "epoch": 0.192384, + "grad_norm": 0.12986101014598134, + "learning_rate": 9.898873325075957e-05, + "loss": 0.0319, + "step": 60120 + }, + { + "epoch": 0.192448, + "grad_norm": 0.09390149694711658, + "learning_rate": 9.898806258583904e-05, + "loss": 0.0309, + "step": 60140 + }, + { + "epoch": 0.192512, + "grad_norm": 0.08418573659121127, + "learning_rate": 9.898739170087554e-05, + "loss": 0.0297, + "step": 60160 + }, + { + "epoch": 0.192576, + "grad_norm": 0.0704679585610366, + "learning_rate": 9.898672059587213e-05, + "loss": 0.0306, + "step": 60180 + }, + { + "epoch": 0.19264, + "grad_norm": 0.08951064028171715, + "learning_rate": 9.898604927083179e-05, + "loss": 0.033, + "step": 60200 + }, + { + "epoch": 0.192704, + "grad_norm": 0.13454959719514914, + "learning_rate": 9.898537772575756e-05, + "loss": 0.0322, + "step": 60220 + }, + { + "epoch": 0.192768, + "grad_norm": 0.11047458939317589, + "learning_rate": 9.898470596065244e-05, + "loss": 0.0289, + "step": 60240 + }, + { + "epoch": 0.192832, + "grad_norm": 0.12751289374018013, + "learning_rate": 9.898403397551946e-05, + "loss": 0.0313, + "step": 60260 + }, + { + "epoch": 0.192896, + "grad_norm": 0.10434423526967476, + "learning_rate": 9.898336177036163e-05, + "loss": 0.0319, + "step": 60280 + }, + { + "epoch": 0.19296, + "grad_norm": 0.07071227445770183, + "learning_rate": 9.898268934518195e-05, + "loss": 0.0301, + "step": 60300 + }, + { + "epoch": 0.193024, + "grad_norm": 0.05139802388862973, + "learning_rate": 9.898201669998349e-05, + "loss": 0.0281, + "step": 60320 + }, + { + "epoch": 0.193088, + "grad_norm": 0.07961721565166838, + "learning_rate": 9.898134383476923e-05, + "loss": 0.0331, + "step": 60340 + }, + { + "epoch": 0.193152, + "grad_norm": 0.12446669206563656, + "learning_rate": 9.89806707495422e-05, + "loss": 0.0298, + "step": 60360 + }, + { + "epoch": 0.193216, + "grad_norm": 0.06648989390945005, + "learning_rate": 9.897999744430544e-05, + "loss": 0.0296, + "step": 60380 + }, + { + "epoch": 0.19328, + "grad_norm": 0.10789188351453063, + "learning_rate": 9.897932391906194e-05, + "loss": 0.0321, + "step": 60400 + }, + { + "epoch": 0.193344, + "grad_norm": 0.06158966927071713, + "learning_rate": 9.897865017381478e-05, + "loss": 0.0315, + "step": 60420 + }, + { + "epoch": 0.193408, + "grad_norm": 0.11237228045891828, + "learning_rate": 9.897797620856693e-05, + "loss": 0.0267, + "step": 60440 + }, + { + "epoch": 0.193472, + "grad_norm": 0.06100002967528478, + "learning_rate": 9.897730202332144e-05, + "loss": 0.0291, + "step": 60460 + }, + { + "epoch": 0.193536, + "grad_norm": 0.08474741080533928, + "learning_rate": 9.897662761808135e-05, + "loss": 0.0299, + "step": 60480 + }, + { + "epoch": 0.1936, + "grad_norm": 0.096926981113508, + "learning_rate": 9.897595299284968e-05, + "loss": 0.0331, + "step": 60500 + }, + { + "epoch": 0.193664, + "grad_norm": 0.07855185510675963, + "learning_rate": 9.897527814762946e-05, + "loss": 0.0333, + "step": 60520 + }, + { + "epoch": 0.193728, + "grad_norm": 0.05561983370065681, + "learning_rate": 9.89746030824237e-05, + "loss": 0.0318, + "step": 60540 + }, + { + "epoch": 0.193792, + "grad_norm": 0.0878095607355052, + "learning_rate": 9.897392779723547e-05, + "loss": 0.0337, + "step": 60560 + }, + { + "epoch": 0.193856, + "grad_norm": 0.06959904863320115, + "learning_rate": 9.897325229206777e-05, + "loss": 0.034, + "step": 60580 + }, + { + "epoch": 0.19392, + "grad_norm": 0.10122390124259824, + "learning_rate": 9.897257656692367e-05, + "loss": 0.0333, + "step": 60600 + }, + { + "epoch": 0.193984, + "grad_norm": 0.07790775664594773, + "learning_rate": 9.897190062180619e-05, + "loss": 0.0342, + "step": 60620 + }, + { + "epoch": 0.194048, + "grad_norm": 0.1409200674558783, + "learning_rate": 9.897122445671833e-05, + "loss": 0.0304, + "step": 60640 + }, + { + "epoch": 0.194112, + "grad_norm": 0.07186219182819582, + "learning_rate": 9.897054807166319e-05, + "loss": 0.0271, + "step": 60660 + }, + { + "epoch": 0.194176, + "grad_norm": 0.07730216261491468, + "learning_rate": 9.896987146664376e-05, + "loss": 0.0284, + "step": 60680 + }, + { + "epoch": 0.19424, + "grad_norm": 0.07092746616956244, + "learning_rate": 9.89691946416631e-05, + "loss": 0.0308, + "step": 60700 + }, + { + "epoch": 0.194304, + "grad_norm": 0.09672085613502611, + "learning_rate": 9.896851759672425e-05, + "loss": 0.0286, + "step": 60720 + }, + { + "epoch": 0.194368, + "grad_norm": 0.0686086032560444, + "learning_rate": 9.896784033183025e-05, + "loss": 0.0318, + "step": 60740 + }, + { + "epoch": 0.194432, + "grad_norm": 0.1126728836128364, + "learning_rate": 9.896716284698412e-05, + "loss": 0.0299, + "step": 60760 + }, + { + "epoch": 0.194496, + "grad_norm": 0.16872112576819884, + "learning_rate": 9.896648514218894e-05, + "loss": 0.0333, + "step": 60780 + }, + { + "epoch": 0.19456, + "grad_norm": 0.08001816011286227, + "learning_rate": 9.896580721744772e-05, + "loss": 0.0319, + "step": 60800 + }, + { + "epoch": 0.194624, + "grad_norm": 0.11760484434579395, + "learning_rate": 9.896512907276354e-05, + "loss": 0.0327, + "step": 60820 + }, + { + "epoch": 0.194688, + "grad_norm": 0.0802170357540477, + "learning_rate": 9.896445070813943e-05, + "loss": 0.0344, + "step": 60840 + }, + { + "epoch": 0.194752, + "grad_norm": 0.0736701887386983, + "learning_rate": 9.89637721235784e-05, + "loss": 0.0291, + "step": 60860 + }, + { + "epoch": 0.194816, + "grad_norm": 0.05677184268701295, + "learning_rate": 9.896309331908356e-05, + "loss": 0.0308, + "step": 60880 + }, + { + "epoch": 0.19488, + "grad_norm": 0.11042946171951377, + "learning_rate": 9.896241429465794e-05, + "loss": 0.0321, + "step": 60900 + }, + { + "epoch": 0.194944, + "grad_norm": 0.0678211761816004, + "learning_rate": 9.896173505030458e-05, + "loss": 0.0295, + "step": 60920 + }, + { + "epoch": 0.195008, + "grad_norm": 0.10701547818281253, + "learning_rate": 9.896105558602651e-05, + "loss": 0.0318, + "step": 60940 + }, + { + "epoch": 0.195072, + "grad_norm": 0.0640714627087837, + "learning_rate": 9.896037590182682e-05, + "loss": 0.0321, + "step": 60960 + }, + { + "epoch": 0.195136, + "grad_norm": 0.062094725726349956, + "learning_rate": 9.895969599770853e-05, + "loss": 0.0291, + "step": 60980 + }, + { + "epoch": 0.1952, + "grad_norm": 0.06687166013531315, + "learning_rate": 9.895901587367473e-05, + "loss": 0.032, + "step": 61000 + }, + { + "epoch": 0.195264, + "grad_norm": 0.09537621545359926, + "learning_rate": 9.895833552972847e-05, + "loss": 0.0288, + "step": 61020 + }, + { + "epoch": 0.195328, + "grad_norm": 0.06471981423394278, + "learning_rate": 9.895765496587276e-05, + "loss": 0.031, + "step": 61040 + }, + { + "epoch": 0.195392, + "grad_norm": 0.0819355863369456, + "learning_rate": 9.895697418211072e-05, + "loss": 0.0294, + "step": 61060 + }, + { + "epoch": 0.195456, + "grad_norm": 0.05866861005891022, + "learning_rate": 9.895629317844535e-05, + "loss": 0.0322, + "step": 61080 + }, + { + "epoch": 0.19552, + "grad_norm": 0.10875114650799464, + "learning_rate": 9.895561195487975e-05, + "loss": 0.0321, + "step": 61100 + }, + { + "epoch": 0.195584, + "grad_norm": 0.07623822675649831, + "learning_rate": 9.895493051141697e-05, + "loss": 0.0301, + "step": 61120 + }, + { + "epoch": 0.195648, + "grad_norm": 0.09785094467827996, + "learning_rate": 9.895424884806007e-05, + "loss": 0.0286, + "step": 61140 + }, + { + "epoch": 0.195712, + "grad_norm": 0.06931473689650194, + "learning_rate": 9.895356696481208e-05, + "loss": 0.0297, + "step": 61160 + }, + { + "epoch": 0.195776, + "grad_norm": 0.05549493272907254, + "learning_rate": 9.895288486167613e-05, + "loss": 0.0283, + "step": 61180 + }, + { + "epoch": 0.19584, + "grad_norm": 0.07784006463678438, + "learning_rate": 9.895220253865522e-05, + "loss": 0.0312, + "step": 61200 + }, + { + "epoch": 0.195904, + "grad_norm": 0.12533703760342307, + "learning_rate": 9.895151999575245e-05, + "loss": 0.0346, + "step": 61220 + }, + { + "epoch": 0.195968, + "grad_norm": 0.07205968559113399, + "learning_rate": 9.895083723297088e-05, + "loss": 0.0329, + "step": 61240 + }, + { + "epoch": 0.196032, + "grad_norm": 0.06113169352554559, + "learning_rate": 9.895015425031356e-05, + "loss": 0.0293, + "step": 61260 + }, + { + "epoch": 0.196096, + "grad_norm": 0.1303784904252652, + "learning_rate": 9.894947104778359e-05, + "loss": 0.032, + "step": 61280 + }, + { + "epoch": 0.19616, + "grad_norm": 0.14684742797318978, + "learning_rate": 9.8948787625384e-05, + "loss": 0.0333, + "step": 61300 + }, + { + "epoch": 0.196224, + "grad_norm": 0.07605924929327698, + "learning_rate": 9.89481039831179e-05, + "loss": 0.0319, + "step": 61320 + }, + { + "epoch": 0.196288, + "grad_norm": 0.10448186062506538, + "learning_rate": 9.894742012098832e-05, + "loss": 0.0312, + "step": 61340 + }, + { + "epoch": 0.196352, + "grad_norm": 0.0698382859985516, + "learning_rate": 9.894673603899836e-05, + "loss": 0.0312, + "step": 61360 + }, + { + "epoch": 0.196416, + "grad_norm": 0.08206932477813039, + "learning_rate": 9.894605173715107e-05, + "loss": 0.0303, + "step": 61380 + }, + { + "epoch": 0.19648, + "grad_norm": 0.07141172314713462, + "learning_rate": 9.894536721544957e-05, + "loss": 0.031, + "step": 61400 + }, + { + "epoch": 0.196544, + "grad_norm": 0.06709066594101129, + "learning_rate": 9.894468247389689e-05, + "loss": 0.0305, + "step": 61420 + }, + { + "epoch": 0.196608, + "grad_norm": 0.12169138506048809, + "learning_rate": 9.894399751249611e-05, + "loss": 0.0326, + "step": 61440 + }, + { + "epoch": 0.196672, + "grad_norm": 0.09613688788499254, + "learning_rate": 9.894331233125033e-05, + "loss": 0.0302, + "step": 61460 + }, + { + "epoch": 0.196736, + "grad_norm": 0.05278142388055059, + "learning_rate": 9.89426269301626e-05, + "loss": 0.0284, + "step": 61480 + }, + { + "epoch": 0.1968, + "grad_norm": 0.13578121101185436, + "learning_rate": 9.894194130923602e-05, + "loss": 0.0309, + "step": 61500 + }, + { + "epoch": 0.196864, + "grad_norm": 0.06637212701748277, + "learning_rate": 9.894125546847366e-05, + "loss": 0.0324, + "step": 61520 + }, + { + "epoch": 0.196928, + "grad_norm": 0.05240225896770924, + "learning_rate": 9.89405694078786e-05, + "loss": 0.0298, + "step": 61540 + }, + { + "epoch": 0.196992, + "grad_norm": 0.05487980617276184, + "learning_rate": 9.893988312745392e-05, + "loss": 0.0283, + "step": 61560 + }, + { + "epoch": 0.197056, + "grad_norm": 0.09063662250011725, + "learning_rate": 9.89391966272027e-05, + "loss": 0.0318, + "step": 61580 + }, + { + "epoch": 0.19712, + "grad_norm": 0.11604046588493298, + "learning_rate": 9.893850990712806e-05, + "loss": 0.0308, + "step": 61600 + }, + { + "epoch": 0.197184, + "grad_norm": 0.07700860958460015, + "learning_rate": 9.893782296723302e-05, + "loss": 0.0305, + "step": 61620 + }, + { + "epoch": 0.197248, + "grad_norm": 0.07933001334214801, + "learning_rate": 9.893713580752072e-05, + "loss": 0.0293, + "step": 61640 + }, + { + "epoch": 0.197312, + "grad_norm": 0.09772465961427232, + "learning_rate": 9.893644842799422e-05, + "loss": 0.0293, + "step": 61660 + }, + { + "epoch": 0.197376, + "grad_norm": 0.10204794159022518, + "learning_rate": 9.893576082865661e-05, + "loss": 0.0299, + "step": 61680 + }, + { + "epoch": 0.19744, + "grad_norm": 0.05855340392624238, + "learning_rate": 9.8935073009511e-05, + "loss": 0.0316, + "step": 61700 + }, + { + "epoch": 0.197504, + "grad_norm": 0.08714582062645373, + "learning_rate": 9.893438497056045e-05, + "loss": 0.0293, + "step": 61720 + }, + { + "epoch": 0.197568, + "grad_norm": 0.08166689874857597, + "learning_rate": 9.893369671180806e-05, + "loss": 0.0292, + "step": 61740 + }, + { + "epoch": 0.197632, + "grad_norm": 0.07942049446652197, + "learning_rate": 9.893300823325693e-05, + "loss": 0.0315, + "step": 61760 + }, + { + "epoch": 0.197696, + "grad_norm": 0.05679037065786764, + "learning_rate": 9.893231953491015e-05, + "loss": 0.0302, + "step": 61780 + }, + { + "epoch": 0.19776, + "grad_norm": 0.0973253556157589, + "learning_rate": 9.89316306167708e-05, + "loss": 0.0343, + "step": 61800 + }, + { + "epoch": 0.197824, + "grad_norm": 0.051617282043366325, + "learning_rate": 9.893094147884199e-05, + "loss": 0.0327, + "step": 61820 + }, + { + "epoch": 0.197888, + "grad_norm": 0.06490598622432976, + "learning_rate": 9.893025212112681e-05, + "loss": 0.0277, + "step": 61840 + }, + { + "epoch": 0.197952, + "grad_norm": 0.06926496415549095, + "learning_rate": 9.892956254362835e-05, + "loss": 0.0286, + "step": 61860 + }, + { + "epoch": 0.198016, + "grad_norm": 0.06583024638132584, + "learning_rate": 9.892887274634973e-05, + "loss": 0.0309, + "step": 61880 + }, + { + "epoch": 0.19808, + "grad_norm": 0.11182811176977141, + "learning_rate": 9.892818272929403e-05, + "loss": 0.0271, + "step": 61900 + }, + { + "epoch": 0.198144, + "grad_norm": 0.05014027710008368, + "learning_rate": 9.892749249246434e-05, + "loss": 0.0298, + "step": 61920 + }, + { + "epoch": 0.198208, + "grad_norm": 0.06955886565878724, + "learning_rate": 9.892680203586377e-05, + "loss": 0.0286, + "step": 61940 + }, + { + "epoch": 0.198272, + "grad_norm": 0.10697841313446631, + "learning_rate": 9.892611135949543e-05, + "loss": 0.0302, + "step": 61960 + }, + { + "epoch": 0.198336, + "grad_norm": 0.06625891120506502, + "learning_rate": 9.892542046336243e-05, + "loss": 0.0305, + "step": 61980 + }, + { + "epoch": 0.1984, + "grad_norm": 0.15507989222184676, + "learning_rate": 9.892472934746784e-05, + "loss": 0.0326, + "step": 62000 + }, + { + "epoch": 0.198464, + "grad_norm": 0.18375831456167252, + "learning_rate": 9.89240380118148e-05, + "loss": 0.0317, + "step": 62020 + }, + { + "epoch": 0.198528, + "grad_norm": 0.07149617138962547, + "learning_rate": 9.892334645640638e-05, + "loss": 0.0351, + "step": 62040 + }, + { + "epoch": 0.198592, + "grad_norm": 0.04779945142917935, + "learning_rate": 9.892265468124572e-05, + "loss": 0.0308, + "step": 62060 + }, + { + "epoch": 0.198656, + "grad_norm": 0.10284807613957302, + "learning_rate": 9.892196268633592e-05, + "loss": 0.0317, + "step": 62080 + }, + { + "epoch": 0.19872, + "grad_norm": 0.06000234518312874, + "learning_rate": 9.892127047168004e-05, + "loss": 0.0344, + "step": 62100 + }, + { + "epoch": 0.198784, + "grad_norm": 0.06429949449582269, + "learning_rate": 9.892057803728127e-05, + "loss": 0.0302, + "step": 62120 + }, + { + "epoch": 0.198848, + "grad_norm": 0.10842053577962579, + "learning_rate": 9.891988538314268e-05, + "loss": 0.0321, + "step": 62140 + }, + { + "epoch": 0.198912, + "grad_norm": 0.13938717345038126, + "learning_rate": 9.891919250926735e-05, + "loss": 0.0322, + "step": 62160 + }, + { + "epoch": 0.198976, + "grad_norm": 0.057713137488330776, + "learning_rate": 9.891849941565845e-05, + "loss": 0.0349, + "step": 62180 + }, + { + "epoch": 0.19904, + "grad_norm": 0.0768692173925232, + "learning_rate": 9.891780610231907e-05, + "loss": 0.0317, + "step": 62200 + }, + { + "epoch": 0.199104, + "grad_norm": 0.09271261722337276, + "learning_rate": 9.891711256925232e-05, + "loss": 0.0285, + "step": 62220 + }, + { + "epoch": 0.199168, + "grad_norm": 0.0898927603500762, + "learning_rate": 9.89164188164613e-05, + "loss": 0.0315, + "step": 62240 + }, + { + "epoch": 0.199232, + "grad_norm": 0.07646915821152574, + "learning_rate": 9.891572484394916e-05, + "loss": 0.0327, + "step": 62260 + }, + { + "epoch": 0.199296, + "grad_norm": 0.07865861512866454, + "learning_rate": 9.8915030651719e-05, + "loss": 0.0302, + "step": 62280 + }, + { + "epoch": 0.19936, + "grad_norm": 0.08436364241977502, + "learning_rate": 9.891433623977391e-05, + "loss": 0.0318, + "step": 62300 + }, + { + "epoch": 0.199424, + "grad_norm": 0.062311075034575523, + "learning_rate": 9.891364160811705e-05, + "loss": 0.0326, + "step": 62320 + }, + { + "epoch": 0.199488, + "grad_norm": 0.08078036865067863, + "learning_rate": 9.891294675675153e-05, + "loss": 0.0298, + "step": 62340 + }, + { + "epoch": 0.199552, + "grad_norm": 0.11242842084438706, + "learning_rate": 9.891225168568048e-05, + "loss": 0.0345, + "step": 62360 + }, + { + "epoch": 0.199616, + "grad_norm": 0.0863466846534019, + "learning_rate": 9.8911556394907e-05, + "loss": 0.0285, + "step": 62380 + }, + { + "epoch": 0.19968, + "grad_norm": 0.05808027699246877, + "learning_rate": 9.891086088443423e-05, + "loss": 0.0264, + "step": 62400 + }, + { + "epoch": 0.199744, + "grad_norm": 0.07995310188949552, + "learning_rate": 9.891016515426528e-05, + "loss": 0.0301, + "step": 62420 + }, + { + "epoch": 0.199808, + "grad_norm": 0.08107171739928222, + "learning_rate": 9.890946920440329e-05, + "loss": 0.0323, + "step": 62440 + }, + { + "epoch": 0.199872, + "grad_norm": 0.06297589335834405, + "learning_rate": 9.890877303485138e-05, + "loss": 0.0317, + "step": 62460 + }, + { + "epoch": 0.199936, + "grad_norm": 0.07762944076852882, + "learning_rate": 9.890807664561266e-05, + "loss": 0.0288, + "step": 62480 + }, + { + "epoch": 0.2, + "grad_norm": 0.12186999436172585, + "learning_rate": 9.890738003669029e-05, + "loss": 0.0327, + "step": 62500 + }, + { + "epoch": 0.200064, + "grad_norm": 0.06563098417455279, + "learning_rate": 9.890668320808739e-05, + "loss": 0.033, + "step": 62520 + }, + { + "epoch": 0.200128, + "grad_norm": 0.05871107683333555, + "learning_rate": 9.890598615980706e-05, + "loss": 0.0283, + "step": 62540 + }, + { + "epoch": 0.200192, + "grad_norm": 0.08004769290431618, + "learning_rate": 9.890528889185246e-05, + "loss": 0.0303, + "step": 62560 + }, + { + "epoch": 0.200256, + "grad_norm": 0.07938834792632665, + "learning_rate": 9.890459140422673e-05, + "loss": 0.0278, + "step": 62580 + }, + { + "epoch": 0.20032, + "grad_norm": 0.08603775407234336, + "learning_rate": 9.890389369693296e-05, + "loss": 0.0308, + "step": 62600 + }, + { + "epoch": 0.200384, + "grad_norm": 0.11055551034732886, + "learning_rate": 9.890319576997433e-05, + "loss": 0.0303, + "step": 62620 + }, + { + "epoch": 0.200448, + "grad_norm": 0.06123784485716809, + "learning_rate": 9.890249762335395e-05, + "loss": 0.0325, + "step": 62640 + }, + { + "epoch": 0.200512, + "grad_norm": 0.07983469946844252, + "learning_rate": 9.890179925707497e-05, + "loss": 0.0297, + "step": 62660 + }, + { + "epoch": 0.200576, + "grad_norm": 0.05206335144078582, + "learning_rate": 9.890110067114054e-05, + "loss": 0.0307, + "step": 62680 + }, + { + "epoch": 0.20064, + "grad_norm": 0.09512223442399825, + "learning_rate": 9.890040186555375e-05, + "loss": 0.0291, + "step": 62700 + }, + { + "epoch": 0.200704, + "grad_norm": 0.06451238270080743, + "learning_rate": 9.889970284031778e-05, + "loss": 0.0281, + "step": 62720 + }, + { + "epoch": 0.200768, + "grad_norm": 0.08550344291490274, + "learning_rate": 9.889900359543575e-05, + "loss": 0.0307, + "step": 62740 + }, + { + "epoch": 0.200832, + "grad_norm": 0.07270337059535208, + "learning_rate": 9.889830413091082e-05, + "loss": 0.0338, + "step": 62760 + }, + { + "epoch": 0.200896, + "grad_norm": 0.09512729992735945, + "learning_rate": 9.88976044467461e-05, + "loss": 0.034, + "step": 62780 + }, + { + "epoch": 0.20096, + "grad_norm": 0.06615045591555849, + "learning_rate": 9.889690454294476e-05, + "loss": 0.0292, + "step": 62800 + }, + { + "epoch": 0.201024, + "grad_norm": 0.08910879008877794, + "learning_rate": 9.889620441950995e-05, + "loss": 0.0305, + "step": 62820 + }, + { + "epoch": 0.201088, + "grad_norm": 0.0605152520287331, + "learning_rate": 9.889550407644478e-05, + "loss": 0.0348, + "step": 62840 + }, + { + "epoch": 0.201152, + "grad_norm": 0.07728539640240824, + "learning_rate": 9.889480351375242e-05, + "loss": 0.0334, + "step": 62860 + }, + { + "epoch": 0.201216, + "grad_norm": 0.06152172644097212, + "learning_rate": 9.889410273143602e-05, + "loss": 0.0317, + "step": 62880 + }, + { + "epoch": 0.20128, + "grad_norm": 0.05386747881846268, + "learning_rate": 9.889340172949872e-05, + "loss": 0.031, + "step": 62900 + }, + { + "epoch": 0.201344, + "grad_norm": 0.07742339573750492, + "learning_rate": 9.889270050794369e-05, + "loss": 0.0302, + "step": 62920 + }, + { + "epoch": 0.201408, + "grad_norm": 0.11064297781592901, + "learning_rate": 9.889199906677402e-05, + "loss": 0.033, + "step": 62940 + }, + { + "epoch": 0.201472, + "grad_norm": 0.059745369438746164, + "learning_rate": 9.889129740599293e-05, + "loss": 0.0286, + "step": 62960 + }, + { + "epoch": 0.201536, + "grad_norm": 0.06454175695880615, + "learning_rate": 9.889059552560353e-05, + "loss": 0.0311, + "step": 62980 + }, + { + "epoch": 0.2016, + "grad_norm": 0.056593599562569506, + "learning_rate": 9.888989342560899e-05, + "loss": 0.0307, + "step": 63000 + }, + { + "epoch": 0.201664, + "grad_norm": 0.06417742242848436, + "learning_rate": 9.888919110601245e-05, + "loss": 0.0307, + "step": 63020 + }, + { + "epoch": 0.201728, + "grad_norm": 0.1268828004041676, + "learning_rate": 9.88884885668171e-05, + "loss": 0.0271, + "step": 63040 + }, + { + "epoch": 0.201792, + "grad_norm": 0.11075245721018577, + "learning_rate": 9.888778580802603e-05, + "loss": 0.0276, + "step": 63060 + }, + { + "epoch": 0.201856, + "grad_norm": 0.12940877064919387, + "learning_rate": 9.888708282964245e-05, + "loss": 0.0326, + "step": 63080 + }, + { + "epoch": 0.20192, + "grad_norm": 0.051846573628494465, + "learning_rate": 9.888637963166951e-05, + "loss": 0.0311, + "step": 63100 + }, + { + "epoch": 0.201984, + "grad_norm": 0.0967980313640854, + "learning_rate": 9.888567621411037e-05, + "loss": 0.0305, + "step": 63120 + }, + { + "epoch": 0.202048, + "grad_norm": 0.14683090184630224, + "learning_rate": 9.888497257696816e-05, + "loss": 0.0322, + "step": 63140 + }, + { + "epoch": 0.202112, + "grad_norm": 0.05832873128240348, + "learning_rate": 9.888426872024607e-05, + "loss": 0.0315, + "step": 63160 + }, + { + "epoch": 0.202176, + "grad_norm": 0.07444137981373461, + "learning_rate": 9.888356464394726e-05, + "loss": 0.0317, + "step": 63180 + }, + { + "epoch": 0.20224, + "grad_norm": 0.060514928654942284, + "learning_rate": 9.888286034807488e-05, + "loss": 0.0301, + "step": 63200 + }, + { + "epoch": 0.202304, + "grad_norm": 0.0838569680536182, + "learning_rate": 9.888215583263211e-05, + "loss": 0.0281, + "step": 63220 + }, + { + "epoch": 0.202368, + "grad_norm": 0.08829041062327733, + "learning_rate": 9.88814510976221e-05, + "loss": 0.0308, + "step": 63240 + }, + { + "epoch": 0.202432, + "grad_norm": 0.06215733658389588, + "learning_rate": 9.888074614304801e-05, + "loss": 0.0294, + "step": 63260 + }, + { + "epoch": 0.202496, + "grad_norm": 0.06359654917293786, + "learning_rate": 9.888004096891301e-05, + "loss": 0.0341, + "step": 63280 + }, + { + "epoch": 0.20256, + "grad_norm": 0.17896120220466766, + "learning_rate": 9.88793355752203e-05, + "loss": 0.0311, + "step": 63300 + }, + { + "epoch": 0.202624, + "grad_norm": 0.05048565255195389, + "learning_rate": 9.887862996197299e-05, + "loss": 0.0318, + "step": 63320 + }, + { + "epoch": 0.202688, + "grad_norm": 0.0646998510329791, + "learning_rate": 9.887792412917429e-05, + "loss": 0.0302, + "step": 63340 + }, + { + "epoch": 0.202752, + "grad_norm": 0.0562610591555951, + "learning_rate": 9.887721807682737e-05, + "loss": 0.0307, + "step": 63360 + }, + { + "epoch": 0.202816, + "grad_norm": 0.1477430376627839, + "learning_rate": 9.887651180493541e-05, + "loss": 0.0305, + "step": 63380 + }, + { + "epoch": 0.20288, + "grad_norm": 0.0674345676720661, + "learning_rate": 9.887580531350153e-05, + "loss": 0.0317, + "step": 63400 + }, + { + "epoch": 0.202944, + "grad_norm": 0.07734606496544501, + "learning_rate": 9.887509860252896e-05, + "loss": 0.033, + "step": 63420 + }, + { + "epoch": 0.203008, + "grad_norm": 0.12056069746244899, + "learning_rate": 9.887439167202085e-05, + "loss": 0.0342, + "step": 63440 + }, + { + "epoch": 0.203072, + "grad_norm": 0.119200794918838, + "learning_rate": 9.887368452198039e-05, + "loss": 0.032, + "step": 63460 + }, + { + "epoch": 0.203136, + "grad_norm": 0.10154048541431089, + "learning_rate": 9.887297715241071e-05, + "loss": 0.0302, + "step": 63480 + }, + { + "epoch": 0.2032, + "grad_norm": 0.0672682043952556, + "learning_rate": 9.887226956331505e-05, + "loss": 0.0317, + "step": 63500 + }, + { + "epoch": 0.203264, + "grad_norm": 0.06015799237233984, + "learning_rate": 9.887156175469656e-05, + "loss": 0.0297, + "step": 63520 + }, + { + "epoch": 0.203328, + "grad_norm": 0.08212798306311808, + "learning_rate": 9.887085372655842e-05, + "loss": 0.029, + "step": 63540 + }, + { + "epoch": 0.203392, + "grad_norm": 0.08341155068462985, + "learning_rate": 9.88701454789038e-05, + "loss": 0.0276, + "step": 63560 + }, + { + "epoch": 0.203456, + "grad_norm": 0.10931404313253278, + "learning_rate": 9.88694370117359e-05, + "loss": 0.0332, + "step": 63580 + }, + { + "epoch": 0.20352, + "grad_norm": 0.11037579197073154, + "learning_rate": 9.886872832505789e-05, + "loss": 0.0348, + "step": 63600 + }, + { + "epoch": 0.203584, + "grad_norm": 0.06616123590128432, + "learning_rate": 9.886801941887297e-05, + "loss": 0.0324, + "step": 63620 + }, + { + "epoch": 0.203648, + "grad_norm": 0.0657499508743157, + "learning_rate": 9.88673102931843e-05, + "loss": 0.0297, + "step": 63640 + }, + { + "epoch": 0.203712, + "grad_norm": 0.1231905127093973, + "learning_rate": 9.886660094799507e-05, + "loss": 0.0304, + "step": 63660 + }, + { + "epoch": 0.203776, + "grad_norm": 0.08221796481018362, + "learning_rate": 9.886589138330848e-05, + "loss": 0.0309, + "step": 63680 + }, + { + "epoch": 0.20384, + "grad_norm": 0.09298171826584051, + "learning_rate": 9.886518159912773e-05, + "loss": 0.0348, + "step": 63700 + }, + { + "epoch": 0.203904, + "grad_norm": 0.0687283839880023, + "learning_rate": 9.886447159545596e-05, + "loss": 0.0305, + "step": 63720 + }, + { + "epoch": 0.203968, + "grad_norm": 0.06302537301108221, + "learning_rate": 9.886376137229638e-05, + "loss": 0.0305, + "step": 63740 + }, + { + "epoch": 0.204032, + "grad_norm": 0.09716567822536326, + "learning_rate": 9.886305092965221e-05, + "loss": 0.0257, + "step": 63760 + }, + { + "epoch": 0.204096, + "grad_norm": 0.10773349544042748, + "learning_rate": 9.886234026752661e-05, + "loss": 0.0285, + "step": 63780 + }, + { + "epoch": 0.20416, + "grad_norm": 0.07901067280872685, + "learning_rate": 9.886162938592279e-05, + "loss": 0.032, + "step": 63800 + }, + { + "epoch": 0.204224, + "grad_norm": 0.07664776168887391, + "learning_rate": 9.886091828484392e-05, + "loss": 0.0299, + "step": 63820 + }, + { + "epoch": 0.204288, + "grad_norm": 0.05456249934455802, + "learning_rate": 9.886020696429322e-05, + "loss": 0.03, + "step": 63840 + }, + { + "epoch": 0.204352, + "grad_norm": 0.06837568986442204, + "learning_rate": 9.885949542427387e-05, + "loss": 0.0301, + "step": 63860 + }, + { + "epoch": 0.204416, + "grad_norm": 0.061365967702563716, + "learning_rate": 9.885878366478906e-05, + "loss": 0.0308, + "step": 63880 + }, + { + "epoch": 0.20448, + "grad_norm": 0.09146446078901226, + "learning_rate": 9.885807168584202e-05, + "loss": 0.0334, + "step": 63900 + }, + { + "epoch": 0.204544, + "grad_norm": 0.10042686227990938, + "learning_rate": 9.885735948743589e-05, + "loss": 0.0309, + "step": 63920 + }, + { + "epoch": 0.204608, + "grad_norm": 0.06603926766223514, + "learning_rate": 9.885664706957394e-05, + "loss": 0.0308, + "step": 63940 + }, + { + "epoch": 0.204672, + "grad_norm": 0.06374318497870726, + "learning_rate": 9.885593443225931e-05, + "loss": 0.0321, + "step": 63960 + }, + { + "epoch": 0.204736, + "grad_norm": 0.07408396367228946, + "learning_rate": 9.885522157549523e-05, + "loss": 0.0292, + "step": 63980 + }, + { + "epoch": 0.2048, + "grad_norm": 0.06248315119423602, + "learning_rate": 9.885450849928489e-05, + "loss": 0.0287, + "step": 64000 + }, + { + "epoch": 0.204864, + "grad_norm": 0.11217650994761383, + "learning_rate": 9.885379520363151e-05, + "loss": 0.0309, + "step": 64020 + }, + { + "epoch": 0.204928, + "grad_norm": 0.07386765104562358, + "learning_rate": 9.885308168853827e-05, + "loss": 0.0349, + "step": 64040 + }, + { + "epoch": 0.204992, + "grad_norm": 0.11135194803739466, + "learning_rate": 9.88523679540084e-05, + "loss": 0.0309, + "step": 64060 + }, + { + "epoch": 0.205056, + "grad_norm": 0.058430369177536384, + "learning_rate": 9.885165400004508e-05, + "loss": 0.0281, + "step": 64080 + }, + { + "epoch": 0.20512, + "grad_norm": 0.08220734841135377, + "learning_rate": 9.885093982665155e-05, + "loss": 0.0303, + "step": 64100 + }, + { + "epoch": 0.205184, + "grad_norm": 0.060990656579194336, + "learning_rate": 9.885022543383099e-05, + "loss": 0.0312, + "step": 64120 + }, + { + "epoch": 0.205248, + "grad_norm": 0.09034580299613967, + "learning_rate": 9.884951082158662e-05, + "loss": 0.03, + "step": 64140 + }, + { + "epoch": 0.205312, + "grad_norm": 0.07721098555870005, + "learning_rate": 9.884879598992165e-05, + "loss": 0.03, + "step": 64160 + }, + { + "epoch": 0.205376, + "grad_norm": 0.05245287110735163, + "learning_rate": 9.884808093883929e-05, + "loss": 0.0293, + "step": 64180 + }, + { + "epoch": 0.20544, + "grad_norm": 0.05473640872983397, + "learning_rate": 9.884736566834274e-05, + "loss": 0.0306, + "step": 64200 + }, + { + "epoch": 0.205504, + "grad_norm": 0.09700912482389562, + "learning_rate": 9.884665017843522e-05, + "loss": 0.0317, + "step": 64220 + }, + { + "epoch": 0.205568, + "grad_norm": 0.08198653204921219, + "learning_rate": 9.884593446911996e-05, + "loss": 0.0297, + "step": 64240 + }, + { + "epoch": 0.205632, + "grad_norm": 0.12875403382407916, + "learning_rate": 9.884521854040017e-05, + "loss": 0.0312, + "step": 64260 + }, + { + "epoch": 0.205696, + "grad_norm": 0.0625957501679003, + "learning_rate": 9.884450239227903e-05, + "loss": 0.027, + "step": 64280 + }, + { + "epoch": 0.20576, + "grad_norm": 0.0748896994819878, + "learning_rate": 9.88437860247598e-05, + "loss": 0.0304, + "step": 64300 + }, + { + "epoch": 0.205824, + "grad_norm": 0.07691370159412471, + "learning_rate": 9.884306943784568e-05, + "loss": 0.0287, + "step": 64320 + }, + { + "epoch": 0.205888, + "grad_norm": 0.09988206869665896, + "learning_rate": 9.88423526315399e-05, + "loss": 0.0313, + "step": 64340 + }, + { + "epoch": 0.205952, + "grad_norm": 0.06096291857040583, + "learning_rate": 9.884163560584565e-05, + "loss": 0.03, + "step": 64360 + }, + { + "epoch": 0.206016, + "grad_norm": 0.05882499625324927, + "learning_rate": 9.884091836076619e-05, + "loss": 0.0283, + "step": 64380 + }, + { + "epoch": 0.20608, + "grad_norm": 0.07311449352289706, + "learning_rate": 9.884020089630473e-05, + "loss": 0.0262, + "step": 64400 + }, + { + "epoch": 0.206144, + "grad_norm": 0.062244495478648035, + "learning_rate": 9.883948321246446e-05, + "loss": 0.0294, + "step": 64420 + }, + { + "epoch": 0.206208, + "grad_norm": 0.10751973944075745, + "learning_rate": 9.883876530924865e-05, + "loss": 0.0296, + "step": 64440 + }, + { + "epoch": 0.206272, + "grad_norm": 0.08757285110306214, + "learning_rate": 9.88380471866605e-05, + "loss": 0.0303, + "step": 64460 + }, + { + "epoch": 0.206336, + "grad_norm": 0.11072280387193506, + "learning_rate": 9.883732884470322e-05, + "loss": 0.0294, + "step": 64480 + }, + { + "epoch": 0.2064, + "grad_norm": 0.07452427659408904, + "learning_rate": 9.883661028338008e-05, + "loss": 0.031, + "step": 64500 + }, + { + "epoch": 0.206464, + "grad_norm": 0.10289466290419874, + "learning_rate": 9.883589150269428e-05, + "loss": 0.0324, + "step": 64520 + }, + { + "epoch": 0.206528, + "grad_norm": 0.09347073594765434, + "learning_rate": 9.883517250264904e-05, + "loss": 0.0324, + "step": 64540 + }, + { + "epoch": 0.206592, + "grad_norm": 0.06950875889445961, + "learning_rate": 9.883445328324762e-05, + "loss": 0.0315, + "step": 64560 + }, + { + "epoch": 0.206656, + "grad_norm": 0.16004413109609228, + "learning_rate": 9.883373384449321e-05, + "loss": 0.0318, + "step": 64580 + }, + { + "epoch": 0.20672, + "grad_norm": 0.15477221883523498, + "learning_rate": 9.883301418638908e-05, + "loss": 0.0296, + "step": 64600 + }, + { + "epoch": 0.206784, + "grad_norm": 0.08781466540967116, + "learning_rate": 9.883229430893846e-05, + "loss": 0.028, + "step": 64620 + }, + { + "epoch": 0.206848, + "grad_norm": 0.09294512414526, + "learning_rate": 9.883157421214455e-05, + "loss": 0.0313, + "step": 64640 + }, + { + "epoch": 0.206912, + "grad_norm": 0.08566778758251467, + "learning_rate": 9.883085389601061e-05, + "loss": 0.0329, + "step": 64660 + }, + { + "epoch": 0.206976, + "grad_norm": 0.08230644110344161, + "learning_rate": 9.883013336053987e-05, + "loss": 0.0309, + "step": 64680 + }, + { + "epoch": 0.20704, + "grad_norm": 0.0690084071551508, + "learning_rate": 9.882941260573557e-05, + "loss": 0.0307, + "step": 64700 + }, + { + "epoch": 0.207104, + "grad_norm": 0.09759997168483174, + "learning_rate": 9.882869163160093e-05, + "loss": 0.0284, + "step": 64720 + }, + { + "epoch": 0.207168, + "grad_norm": 0.07242837389111798, + "learning_rate": 9.882797043813922e-05, + "loss": 0.0285, + "step": 64740 + }, + { + "epoch": 0.207232, + "grad_norm": 0.07482801742379408, + "learning_rate": 9.882724902535365e-05, + "loss": 0.0337, + "step": 64760 + }, + { + "epoch": 0.207296, + "grad_norm": 0.07163112894765901, + "learning_rate": 9.882652739324749e-05, + "loss": 0.0325, + "step": 64780 + }, + { + "epoch": 0.20736, + "grad_norm": 0.06799728310017482, + "learning_rate": 9.882580554182395e-05, + "loss": 0.0295, + "step": 64800 + }, + { + "epoch": 0.207424, + "grad_norm": 0.05140417337600272, + "learning_rate": 9.882508347108629e-05, + "loss": 0.0313, + "step": 64820 + }, + { + "epoch": 0.207488, + "grad_norm": 0.0844284109742039, + "learning_rate": 9.882436118103776e-05, + "loss": 0.0325, + "step": 64840 + }, + { + "epoch": 0.207552, + "grad_norm": 0.1082347599460806, + "learning_rate": 9.882363867168157e-05, + "loss": 0.0286, + "step": 64860 + }, + { + "epoch": 0.207616, + "grad_norm": 0.1321313726237459, + "learning_rate": 9.8822915943021e-05, + "loss": 0.0294, + "step": 64880 + }, + { + "epoch": 0.20768, + "grad_norm": 0.053184893732247385, + "learning_rate": 9.882219299505928e-05, + "loss": 0.0312, + "step": 64900 + }, + { + "epoch": 0.207744, + "grad_norm": 0.07928081127986399, + "learning_rate": 9.882146982779968e-05, + "loss": 0.0285, + "step": 64920 + }, + { + "epoch": 0.207808, + "grad_norm": 0.07231535765809001, + "learning_rate": 9.882074644124542e-05, + "loss": 0.0317, + "step": 64940 + }, + { + "epoch": 0.207872, + "grad_norm": 0.06334499974691624, + "learning_rate": 9.882002283539976e-05, + "loss": 0.0282, + "step": 64960 + }, + { + "epoch": 0.207936, + "grad_norm": 0.07207185842086515, + "learning_rate": 9.881929901026596e-05, + "loss": 0.0307, + "step": 64980 + }, + { + "epoch": 0.208, + "grad_norm": 0.1721508134809888, + "learning_rate": 9.881857496584726e-05, + "loss": 0.0289, + "step": 65000 + }, + { + "epoch": 0.208064, + "grad_norm": 0.08803090965037179, + "learning_rate": 9.88178507021469e-05, + "loss": 0.0278, + "step": 65020 + }, + { + "epoch": 0.208128, + "grad_norm": 0.06328641056238643, + "learning_rate": 9.881712621916816e-05, + "loss": 0.0319, + "step": 65040 + }, + { + "epoch": 0.208192, + "grad_norm": 0.06464172430577718, + "learning_rate": 9.881640151691426e-05, + "loss": 0.0291, + "step": 65060 + }, + { + "epoch": 0.208256, + "grad_norm": 0.0811637248113288, + "learning_rate": 9.88156765953885e-05, + "loss": 0.0305, + "step": 65080 + }, + { + "epoch": 0.20832, + "grad_norm": 0.10973631260988599, + "learning_rate": 9.881495145459411e-05, + "loss": 0.0287, + "step": 65100 + }, + { + "epoch": 0.208384, + "grad_norm": 0.08359290176455803, + "learning_rate": 9.881422609453436e-05, + "loss": 0.0293, + "step": 65120 + }, + { + "epoch": 0.208448, + "grad_norm": 0.0646451899461119, + "learning_rate": 9.881350051521249e-05, + "loss": 0.0309, + "step": 65140 + }, + { + "epoch": 0.208512, + "grad_norm": 0.057568645461188096, + "learning_rate": 9.881277471663176e-05, + "loss": 0.0311, + "step": 65160 + }, + { + "epoch": 0.208576, + "grad_norm": 0.06858471435822595, + "learning_rate": 9.881204869879545e-05, + "loss": 0.0284, + "step": 65180 + }, + { + "epoch": 0.20864, + "grad_norm": 0.13768895249512972, + "learning_rate": 9.881132246170679e-05, + "loss": 0.0312, + "step": 65200 + }, + { + "epoch": 0.208704, + "grad_norm": 0.06067100071369753, + "learning_rate": 9.881059600536908e-05, + "loss": 0.0281, + "step": 65220 + }, + { + "epoch": 0.208768, + "grad_norm": 0.16741521431349984, + "learning_rate": 9.880986932978555e-05, + "loss": 0.0324, + "step": 65240 + }, + { + "epoch": 0.208832, + "grad_norm": 0.05726575772112028, + "learning_rate": 9.880914243495949e-05, + "loss": 0.0289, + "step": 65260 + }, + { + "epoch": 0.208896, + "grad_norm": 0.10027081508957099, + "learning_rate": 9.880841532089415e-05, + "loss": 0.0301, + "step": 65280 + }, + { + "epoch": 0.20896, + "grad_norm": 0.08863928465734576, + "learning_rate": 9.880768798759279e-05, + "loss": 0.0322, + "step": 65300 + }, + { + "epoch": 0.209024, + "grad_norm": 0.05958664703588193, + "learning_rate": 9.88069604350587e-05, + "loss": 0.0328, + "step": 65320 + }, + { + "epoch": 0.209088, + "grad_norm": 0.10350068693876718, + "learning_rate": 9.880623266329513e-05, + "loss": 0.0307, + "step": 65340 + }, + { + "epoch": 0.209152, + "grad_norm": 0.0562583856087072, + "learning_rate": 9.880550467230534e-05, + "loss": 0.0284, + "step": 65360 + }, + { + "epoch": 0.209216, + "grad_norm": 0.05908125080337446, + "learning_rate": 9.880477646209263e-05, + "loss": 0.0306, + "step": 65380 + }, + { + "epoch": 0.20928, + "grad_norm": 0.06332803717872607, + "learning_rate": 9.880404803266026e-05, + "loss": 0.0308, + "step": 65400 + }, + { + "epoch": 0.209344, + "grad_norm": 0.06042723871723859, + "learning_rate": 9.880331938401148e-05, + "loss": 0.0316, + "step": 65420 + }, + { + "epoch": 0.209408, + "grad_norm": 0.06602475321586529, + "learning_rate": 9.880259051614958e-05, + "loss": 0.0267, + "step": 65440 + }, + { + "epoch": 0.209472, + "grad_norm": 0.13518117720918105, + "learning_rate": 9.880186142907784e-05, + "loss": 0.0323, + "step": 65460 + }, + { + "epoch": 0.209536, + "grad_norm": 0.12275732473190211, + "learning_rate": 9.880113212279951e-05, + "loss": 0.0294, + "step": 65480 + }, + { + "epoch": 0.2096, + "grad_norm": 0.12786605206225263, + "learning_rate": 9.88004025973179e-05, + "loss": 0.0317, + "step": 65500 + }, + { + "epoch": 0.209664, + "grad_norm": 0.09416366518834257, + "learning_rate": 9.879967285263627e-05, + "loss": 0.0295, + "step": 65520 + }, + { + "epoch": 0.209728, + "grad_norm": 0.06578730232202905, + "learning_rate": 9.87989428887579e-05, + "loss": 0.031, + "step": 65540 + }, + { + "epoch": 0.209792, + "grad_norm": 0.09013511319645717, + "learning_rate": 9.879821270568606e-05, + "loss": 0.0305, + "step": 65560 + }, + { + "epoch": 0.209856, + "grad_norm": 0.07944115797382639, + "learning_rate": 9.879748230342403e-05, + "loss": 0.0284, + "step": 65580 + }, + { + "epoch": 0.20992, + "grad_norm": 0.08855734107984915, + "learning_rate": 9.879675168197511e-05, + "loss": 0.0301, + "step": 65600 + }, + { + "epoch": 0.209984, + "grad_norm": 0.12427098183292717, + "learning_rate": 9.879602084134256e-05, + "loss": 0.0313, + "step": 65620 + }, + { + "epoch": 0.210048, + "grad_norm": 0.05430222823579225, + "learning_rate": 9.879528978152968e-05, + "loss": 0.0289, + "step": 65640 + }, + { + "epoch": 0.210112, + "grad_norm": 0.06137296351349123, + "learning_rate": 9.879455850253975e-05, + "loss": 0.028, + "step": 65660 + }, + { + "epoch": 0.210176, + "grad_norm": 0.09179284624749623, + "learning_rate": 9.879382700437605e-05, + "loss": 0.0296, + "step": 65680 + }, + { + "epoch": 0.21024, + "grad_norm": 0.06480670108087928, + "learning_rate": 9.879309528704186e-05, + "loss": 0.0296, + "step": 65700 + }, + { + "epoch": 0.210304, + "grad_norm": 0.058823207860685435, + "learning_rate": 9.879236335054048e-05, + "loss": 0.0277, + "step": 65720 + }, + { + "epoch": 0.210368, + "grad_norm": 0.07097876662328613, + "learning_rate": 9.879163119487517e-05, + "loss": 0.0322, + "step": 65740 + }, + { + "epoch": 0.210432, + "grad_norm": 0.05602535705272238, + "learning_rate": 9.879089882004927e-05, + "loss": 0.0265, + "step": 65760 + }, + { + "epoch": 0.210496, + "grad_norm": 0.16553919162885017, + "learning_rate": 9.879016622606601e-05, + "loss": 0.031, + "step": 65780 + }, + { + "epoch": 0.21056, + "grad_norm": 0.05712313358852256, + "learning_rate": 9.878943341292872e-05, + "loss": 0.0331, + "step": 65800 + }, + { + "epoch": 0.210624, + "grad_norm": 0.06393461887123342, + "learning_rate": 9.87887003806407e-05, + "loss": 0.0319, + "step": 65820 + }, + { + "epoch": 0.210688, + "grad_norm": 0.06304808254808351, + "learning_rate": 9.87879671292052e-05, + "loss": 0.0295, + "step": 65840 + }, + { + "epoch": 0.210752, + "grad_norm": 0.11059108773570893, + "learning_rate": 9.878723365862554e-05, + "loss": 0.032, + "step": 65860 + }, + { + "epoch": 0.210816, + "grad_norm": 0.10744175638042021, + "learning_rate": 9.878649996890503e-05, + "loss": 0.0321, + "step": 65880 + }, + { + "epoch": 0.21088, + "grad_norm": 0.09046224499552181, + "learning_rate": 9.878576606004694e-05, + "loss": 0.0339, + "step": 65900 + }, + { + "epoch": 0.210944, + "grad_norm": 0.07676937107170896, + "learning_rate": 9.878503193205457e-05, + "loss": 0.0331, + "step": 65920 + }, + { + "epoch": 0.211008, + "grad_norm": 0.07090799273134167, + "learning_rate": 9.878429758493122e-05, + "loss": 0.0309, + "step": 65940 + }, + { + "epoch": 0.211072, + "grad_norm": 0.04585085214838297, + "learning_rate": 9.878356301868019e-05, + "loss": 0.0279, + "step": 65960 + }, + { + "epoch": 0.211136, + "grad_norm": 0.07262035826014616, + "learning_rate": 9.878282823330478e-05, + "loss": 0.0309, + "step": 65980 + }, + { + "epoch": 0.2112, + "grad_norm": 0.07444829110114114, + "learning_rate": 9.87820932288083e-05, + "loss": 0.0294, + "step": 66000 + }, + { + "epoch": 0.211264, + "grad_norm": 0.08314658382986373, + "learning_rate": 9.878135800519404e-05, + "loss": 0.03, + "step": 66020 + }, + { + "epoch": 0.211328, + "grad_norm": 0.10260505758260019, + "learning_rate": 9.87806225624653e-05, + "loss": 0.0298, + "step": 66040 + }, + { + "epoch": 0.211392, + "grad_norm": 0.11234865583599674, + "learning_rate": 9.877988690062538e-05, + "loss": 0.0303, + "step": 66060 + }, + { + "epoch": 0.211456, + "grad_norm": 0.061335710646279885, + "learning_rate": 9.877915101967761e-05, + "loss": 0.0326, + "step": 66080 + }, + { + "epoch": 0.21152, + "grad_norm": 0.07086046107088644, + "learning_rate": 9.877841491962526e-05, + "loss": 0.0255, + "step": 66100 + }, + { + "epoch": 0.211584, + "grad_norm": 0.06304385096320285, + "learning_rate": 9.877767860047167e-05, + "loss": 0.028, + "step": 66120 + }, + { + "epoch": 0.211648, + "grad_norm": 0.09172148196521665, + "learning_rate": 9.877694206222011e-05, + "loss": 0.0319, + "step": 66140 + }, + { + "epoch": 0.211712, + "grad_norm": 0.11293899674061732, + "learning_rate": 9.877620530487392e-05, + "loss": 0.0306, + "step": 66160 + }, + { + "epoch": 0.211776, + "grad_norm": 0.18020814284526251, + "learning_rate": 9.87754683284364e-05, + "loss": 0.0316, + "step": 66180 + }, + { + "epoch": 0.21184, + "grad_norm": 0.05393480204452683, + "learning_rate": 9.877473113291087e-05, + "loss": 0.0317, + "step": 66200 + }, + { + "epoch": 0.211904, + "grad_norm": 0.0962395034930969, + "learning_rate": 9.877399371830062e-05, + "loss": 0.0313, + "step": 66220 + }, + { + "epoch": 0.211968, + "grad_norm": 0.06600540584477375, + "learning_rate": 9.877325608460897e-05, + "loss": 0.0301, + "step": 66240 + }, + { + "epoch": 0.212032, + "grad_norm": 0.06950117837027779, + "learning_rate": 9.877251823183925e-05, + "loss": 0.0302, + "step": 66260 + }, + { + "epoch": 0.212096, + "grad_norm": 0.08509827406158259, + "learning_rate": 9.877178015999475e-05, + "loss": 0.0298, + "step": 66280 + }, + { + "epoch": 0.21216, + "grad_norm": 0.13662299728142893, + "learning_rate": 9.87710418690788e-05, + "loss": 0.0322, + "step": 66300 + }, + { + "epoch": 0.212224, + "grad_norm": 0.07606081170277909, + "learning_rate": 9.87703033590947e-05, + "loss": 0.0307, + "step": 66320 + }, + { + "epoch": 0.212288, + "grad_norm": 0.10297451755167362, + "learning_rate": 9.876956463004579e-05, + "loss": 0.0284, + "step": 66340 + }, + { + "epoch": 0.212352, + "grad_norm": 0.07189507979759573, + "learning_rate": 9.876882568193537e-05, + "loss": 0.0312, + "step": 66360 + }, + { + "epoch": 0.212416, + "grad_norm": 0.07516223002121793, + "learning_rate": 9.876808651476677e-05, + "loss": 0.0307, + "step": 66380 + }, + { + "epoch": 0.21248, + "grad_norm": 0.0726003966033181, + "learning_rate": 9.87673471285433e-05, + "loss": 0.0347, + "step": 66400 + }, + { + "epoch": 0.212544, + "grad_norm": 0.07778501977207862, + "learning_rate": 9.87666075232683e-05, + "loss": 0.031, + "step": 66420 + }, + { + "epoch": 0.212608, + "grad_norm": 0.08637884442577837, + "learning_rate": 9.876586769894508e-05, + "loss": 0.0299, + "step": 66440 + }, + { + "epoch": 0.212672, + "grad_norm": 0.08674391123902865, + "learning_rate": 9.876512765557695e-05, + "loss": 0.0303, + "step": 66460 + }, + { + "epoch": 0.212736, + "grad_norm": 0.07016449148306768, + "learning_rate": 9.876438739316727e-05, + "loss": 0.0286, + "step": 66480 + }, + { + "epoch": 0.2128, + "grad_norm": 0.07010860547956146, + "learning_rate": 9.876364691171932e-05, + "loss": 0.0295, + "step": 66500 + }, + { + "epoch": 0.212864, + "grad_norm": 0.08670551723411213, + "learning_rate": 9.876290621123645e-05, + "loss": 0.0357, + "step": 66520 + }, + { + "epoch": 0.212928, + "grad_norm": 0.07801653715471453, + "learning_rate": 9.8762165291722e-05, + "loss": 0.032, + "step": 66540 + }, + { + "epoch": 0.212992, + "grad_norm": 0.05904062857960497, + "learning_rate": 9.876142415317928e-05, + "loss": 0.031, + "step": 66560 + }, + { + "epoch": 0.213056, + "grad_norm": 0.07548206637336446, + "learning_rate": 9.876068279561162e-05, + "loss": 0.0317, + "step": 66580 + }, + { + "epoch": 0.21312, + "grad_norm": 0.06260345583574385, + "learning_rate": 9.875994121902234e-05, + "loss": 0.0301, + "step": 66600 + }, + { + "epoch": 0.213184, + "grad_norm": 0.07490355158714451, + "learning_rate": 9.87591994234148e-05, + "loss": 0.029, + "step": 66620 + }, + { + "epoch": 0.213248, + "grad_norm": 0.0672918655386593, + "learning_rate": 9.87584574087923e-05, + "loss": 0.0288, + "step": 66640 + }, + { + "epoch": 0.213312, + "grad_norm": 0.13492948460939544, + "learning_rate": 9.87577151751582e-05, + "loss": 0.0298, + "step": 66660 + }, + { + "epoch": 0.213376, + "grad_norm": 0.06387780273175377, + "learning_rate": 9.875697272251583e-05, + "loss": 0.0306, + "step": 66680 + }, + { + "epoch": 0.21344, + "grad_norm": 0.059812147910202916, + "learning_rate": 9.87562300508685e-05, + "loss": 0.0314, + "step": 66700 + }, + { + "epoch": 0.213504, + "grad_norm": 0.07660155427685679, + "learning_rate": 9.875548716021958e-05, + "loss": 0.0294, + "step": 66720 + }, + { + "epoch": 0.213568, + "grad_norm": 0.12606306519578506, + "learning_rate": 9.875474405057237e-05, + "loss": 0.0316, + "step": 66740 + }, + { + "epoch": 0.213632, + "grad_norm": 0.06725353532980463, + "learning_rate": 9.875400072193024e-05, + "loss": 0.0323, + "step": 66760 + }, + { + "epoch": 0.213696, + "grad_norm": 0.06727550112647027, + "learning_rate": 9.875325717429652e-05, + "loss": 0.0354, + "step": 66780 + }, + { + "epoch": 0.21376, + "grad_norm": 0.07780889077323642, + "learning_rate": 9.875251340767453e-05, + "loss": 0.0325, + "step": 66800 + }, + { + "epoch": 0.213824, + "grad_norm": 0.10243582524410631, + "learning_rate": 9.875176942206765e-05, + "loss": 0.0287, + "step": 66820 + }, + { + "epoch": 0.213888, + "grad_norm": 0.05662151695491244, + "learning_rate": 9.875102521747919e-05, + "loss": 0.0319, + "step": 66840 + }, + { + "epoch": 0.213952, + "grad_norm": 0.08143249440846456, + "learning_rate": 9.87502807939125e-05, + "loss": 0.0283, + "step": 66860 + }, + { + "epoch": 0.214016, + "grad_norm": 0.07906029748393649, + "learning_rate": 9.874953615137093e-05, + "loss": 0.0325, + "step": 66880 + }, + { + "epoch": 0.21408, + "grad_norm": 0.05623292283782337, + "learning_rate": 9.874879128985781e-05, + "loss": 0.0281, + "step": 66900 + }, + { + "epoch": 0.214144, + "grad_norm": 0.03984736499432461, + "learning_rate": 9.87480462093765e-05, + "loss": 0.0265, + "step": 66920 + }, + { + "epoch": 0.214208, + "grad_norm": 0.08704841714914519, + "learning_rate": 9.874730090993035e-05, + "loss": 0.0285, + "step": 66940 + }, + { + "epoch": 0.214272, + "grad_norm": 0.12582909455596808, + "learning_rate": 9.87465553915227e-05, + "loss": 0.0347, + "step": 66960 + }, + { + "epoch": 0.214336, + "grad_norm": 0.09271212086836707, + "learning_rate": 9.87458096541569e-05, + "loss": 0.0321, + "step": 66980 + }, + { + "epoch": 0.2144, + "grad_norm": 0.06118585193056204, + "learning_rate": 9.874506369783629e-05, + "loss": 0.0291, + "step": 67000 + }, + { + "epoch": 0.214464, + "grad_norm": 0.07639850503541895, + "learning_rate": 9.874431752256423e-05, + "loss": 0.0316, + "step": 67020 + }, + { + "epoch": 0.214528, + "grad_norm": 0.11287480466422598, + "learning_rate": 9.87435711283441e-05, + "loss": 0.033, + "step": 67040 + }, + { + "epoch": 0.214592, + "grad_norm": 0.12840883531753997, + "learning_rate": 9.874282451517918e-05, + "loss": 0.0323, + "step": 67060 + }, + { + "epoch": 0.214656, + "grad_norm": 0.05725246818422788, + "learning_rate": 9.874207768307289e-05, + "loss": 0.0325, + "step": 67080 + }, + { + "epoch": 0.21472, + "grad_norm": 0.15343184432184542, + "learning_rate": 9.874133063202857e-05, + "loss": 0.0273, + "step": 67100 + }, + { + "epoch": 0.214784, + "grad_norm": 0.07523781487558626, + "learning_rate": 9.874058336204955e-05, + "loss": 0.0283, + "step": 67120 + }, + { + "epoch": 0.214848, + "grad_norm": 0.07845050406310958, + "learning_rate": 9.873983587313922e-05, + "loss": 0.0315, + "step": 67140 + }, + { + "epoch": 0.214912, + "grad_norm": 0.10556248499595557, + "learning_rate": 9.873908816530091e-05, + "loss": 0.0314, + "step": 67160 + }, + { + "epoch": 0.214976, + "grad_norm": 0.07458678252752649, + "learning_rate": 9.8738340238538e-05, + "loss": 0.0329, + "step": 67180 + }, + { + "epoch": 0.21504, + "grad_norm": 0.13480286254807902, + "learning_rate": 9.873759209285383e-05, + "loss": 0.0333, + "step": 67200 + }, + { + "epoch": 0.215104, + "grad_norm": 0.12941368012492618, + "learning_rate": 9.873684372825179e-05, + "loss": 0.0303, + "step": 67220 + }, + { + "epoch": 0.215168, + "grad_norm": 0.11877303412620768, + "learning_rate": 9.873609514473521e-05, + "loss": 0.0302, + "step": 67240 + }, + { + "epoch": 0.215232, + "grad_norm": 0.08249525680924728, + "learning_rate": 9.873534634230745e-05, + "loss": 0.0324, + "step": 67260 + }, + { + "epoch": 0.215296, + "grad_norm": 0.04885210667912073, + "learning_rate": 9.87345973209719e-05, + "loss": 0.0304, + "step": 67280 + }, + { + "epoch": 0.21536, + "grad_norm": 0.08077339514755692, + "learning_rate": 9.873384808073192e-05, + "loss": 0.027, + "step": 67300 + }, + { + "epoch": 0.215424, + "grad_norm": 0.08209176980644767, + "learning_rate": 9.873309862159085e-05, + "loss": 0.0287, + "step": 67320 + }, + { + "epoch": 0.215488, + "grad_norm": 0.08186694964857857, + "learning_rate": 9.87323489435521e-05, + "loss": 0.0308, + "step": 67340 + }, + { + "epoch": 0.215552, + "grad_norm": 0.04048221358208875, + "learning_rate": 9.873159904661899e-05, + "loss": 0.0286, + "step": 67360 + }, + { + "epoch": 0.215616, + "grad_norm": 0.1327438894455643, + "learning_rate": 9.873084893079492e-05, + "loss": 0.0285, + "step": 67380 + }, + { + "epoch": 0.21568, + "grad_norm": 0.0867618842218541, + "learning_rate": 9.873009859608324e-05, + "loss": 0.0275, + "step": 67400 + }, + { + "epoch": 0.215744, + "grad_norm": 0.06559333893994203, + "learning_rate": 9.872934804248735e-05, + "loss": 0.0314, + "step": 67420 + }, + { + "epoch": 0.215808, + "grad_norm": 0.11646669475001971, + "learning_rate": 9.872859727001058e-05, + "loss": 0.0292, + "step": 67440 + }, + { + "epoch": 0.215872, + "grad_norm": 0.07849231516675449, + "learning_rate": 9.872784627865632e-05, + "loss": 0.0273, + "step": 67460 + }, + { + "epoch": 0.215936, + "grad_norm": 0.13965998539438756, + "learning_rate": 9.872709506842796e-05, + "loss": 0.0292, + "step": 67480 + }, + { + "epoch": 0.216, + "grad_norm": 0.08488628668875155, + "learning_rate": 9.872634363932887e-05, + "loss": 0.0268, + "step": 67500 + }, + { + "epoch": 0.216064, + "grad_norm": 0.07377945229537208, + "learning_rate": 9.87255919913624e-05, + "loss": 0.0297, + "step": 67520 + }, + { + "epoch": 0.216128, + "grad_norm": 0.09933591969212001, + "learning_rate": 9.872484012453194e-05, + "loss": 0.029, + "step": 67540 + }, + { + "epoch": 0.216192, + "grad_norm": 0.04267748442442709, + "learning_rate": 9.872408803884088e-05, + "loss": 0.0332, + "step": 67560 + }, + { + "epoch": 0.216256, + "grad_norm": 0.07684293202764399, + "learning_rate": 9.872333573429258e-05, + "loss": 0.0264, + "step": 67580 + }, + { + "epoch": 0.21632, + "grad_norm": 0.05810291490448085, + "learning_rate": 9.872258321089042e-05, + "loss": 0.0275, + "step": 67600 + }, + { + "epoch": 0.216384, + "grad_norm": 0.065205498864142, + "learning_rate": 9.872183046863779e-05, + "loss": 0.0303, + "step": 67620 + }, + { + "epoch": 0.216448, + "grad_norm": 0.08960050712340008, + "learning_rate": 9.872107750753809e-05, + "loss": 0.0305, + "step": 67640 + }, + { + "epoch": 0.216512, + "grad_norm": 0.08923840358438835, + "learning_rate": 9.872032432759466e-05, + "loss": 0.0325, + "step": 67660 + }, + { + "epoch": 0.216576, + "grad_norm": 0.06798349729989919, + "learning_rate": 9.87195709288109e-05, + "loss": 0.0302, + "step": 67680 + }, + { + "epoch": 0.21664, + "grad_norm": 0.11214714616809551, + "learning_rate": 9.871881731119022e-05, + "loss": 0.0287, + "step": 67700 + }, + { + "epoch": 0.216704, + "grad_norm": 0.1088425137283061, + "learning_rate": 9.871806347473595e-05, + "loss": 0.0317, + "step": 67720 + }, + { + "epoch": 0.216768, + "grad_norm": 0.07056837820921764, + "learning_rate": 9.871730941945154e-05, + "loss": 0.0347, + "step": 67740 + }, + { + "epoch": 0.216832, + "grad_norm": 0.06801061767057032, + "learning_rate": 9.871655514534032e-05, + "loss": 0.0329, + "step": 67760 + }, + { + "epoch": 0.216896, + "grad_norm": 0.13493563047956883, + "learning_rate": 9.871580065240572e-05, + "loss": 0.0309, + "step": 67780 + }, + { + "epoch": 0.21696, + "grad_norm": 0.09181917470371516, + "learning_rate": 9.871504594065111e-05, + "loss": 0.0322, + "step": 67800 + }, + { + "epoch": 0.217024, + "grad_norm": 0.07273869132462686, + "learning_rate": 9.871429101007987e-05, + "loss": 0.0343, + "step": 67820 + }, + { + "epoch": 0.217088, + "grad_norm": 0.12655370201647823, + "learning_rate": 9.871353586069542e-05, + "loss": 0.034, + "step": 67840 + }, + { + "epoch": 0.217152, + "grad_norm": 0.07248954400355188, + "learning_rate": 9.871278049250113e-05, + "loss": 0.0331, + "step": 67860 + }, + { + "epoch": 0.217216, + "grad_norm": 0.07201980162607466, + "learning_rate": 9.87120249055004e-05, + "loss": 0.0305, + "step": 67880 + }, + { + "epoch": 0.21728, + "grad_norm": 0.06672467850491343, + "learning_rate": 9.871126909969661e-05, + "loss": 0.0309, + "step": 67900 + }, + { + "epoch": 0.217344, + "grad_norm": 0.08097269494293267, + "learning_rate": 9.871051307509318e-05, + "loss": 0.0296, + "step": 67920 + }, + { + "epoch": 0.217408, + "grad_norm": 0.08149620038151364, + "learning_rate": 9.87097568316935e-05, + "loss": 0.0317, + "step": 67940 + }, + { + "epoch": 0.217472, + "grad_norm": 0.14182599450038555, + "learning_rate": 9.870900036950095e-05, + "loss": 0.0326, + "step": 67960 + }, + { + "epoch": 0.217536, + "grad_norm": 0.06905243954447576, + "learning_rate": 9.870824368851893e-05, + "loss": 0.0294, + "step": 67980 + }, + { + "epoch": 0.2176, + "grad_norm": 0.051379342542332875, + "learning_rate": 9.870748678875086e-05, + "loss": 0.0279, + "step": 68000 + }, + { + "epoch": 0.217664, + "grad_norm": 0.11714916526157418, + "learning_rate": 9.870672967020012e-05, + "loss": 0.029, + "step": 68020 + }, + { + "epoch": 0.217728, + "grad_norm": 0.05876549963826233, + "learning_rate": 9.870597233287011e-05, + "loss": 0.0296, + "step": 68040 + }, + { + "epoch": 0.217792, + "grad_norm": 0.1054752551697724, + "learning_rate": 9.870521477676426e-05, + "loss": 0.0317, + "step": 68060 + }, + { + "epoch": 0.217856, + "grad_norm": 0.062091893914243135, + "learning_rate": 9.870445700188593e-05, + "loss": 0.0319, + "step": 68080 + }, + { + "epoch": 0.21792, + "grad_norm": 0.059789747971987495, + "learning_rate": 9.870369900823856e-05, + "loss": 0.0293, + "step": 68100 + }, + { + "epoch": 0.217984, + "grad_norm": 0.04791912422638542, + "learning_rate": 9.870294079582554e-05, + "loss": 0.0303, + "step": 68120 + }, + { + "epoch": 0.218048, + "grad_norm": 0.09599862325507434, + "learning_rate": 9.870218236465027e-05, + "loss": 0.0345, + "step": 68140 + }, + { + "epoch": 0.218112, + "grad_norm": 0.08339019857539463, + "learning_rate": 9.870142371471616e-05, + "loss": 0.0312, + "step": 68160 + }, + { + "epoch": 0.218176, + "grad_norm": 0.0761389638286953, + "learning_rate": 9.870066484602662e-05, + "loss": 0.0326, + "step": 68180 + }, + { + "epoch": 0.21824, + "grad_norm": 0.11389618608868135, + "learning_rate": 9.869990575858508e-05, + "loss": 0.0296, + "step": 68200 + }, + { + "epoch": 0.218304, + "grad_norm": 0.15392625212120667, + "learning_rate": 9.869914645239491e-05, + "loss": 0.0311, + "step": 68220 + }, + { + "epoch": 0.218368, + "grad_norm": 0.10050012147292824, + "learning_rate": 9.869838692745954e-05, + "loss": 0.0349, + "step": 68240 + }, + { + "epoch": 0.218432, + "grad_norm": 0.06612473573334282, + "learning_rate": 9.869762718378237e-05, + "loss": 0.0281, + "step": 68260 + }, + { + "epoch": 0.218496, + "grad_norm": 0.04140456027936097, + "learning_rate": 9.869686722136684e-05, + "loss": 0.0309, + "step": 68280 + }, + { + "epoch": 0.21856, + "grad_norm": 0.09175069374322363, + "learning_rate": 9.869610704021635e-05, + "loss": 0.0332, + "step": 68300 + }, + { + "epoch": 0.218624, + "grad_norm": 0.14221275023779972, + "learning_rate": 9.86953466403343e-05, + "loss": 0.0338, + "step": 68320 + }, + { + "epoch": 0.218688, + "grad_norm": 0.06957420834007481, + "learning_rate": 9.869458602172412e-05, + "loss": 0.0324, + "step": 68340 + }, + { + "epoch": 0.218752, + "grad_norm": 0.09575596165704836, + "learning_rate": 9.869382518438923e-05, + "loss": 0.0301, + "step": 68360 + }, + { + "epoch": 0.218816, + "grad_norm": 0.06792052766037862, + "learning_rate": 9.869306412833304e-05, + "loss": 0.032, + "step": 68380 + }, + { + "epoch": 0.21888, + "grad_norm": 0.11393765134938433, + "learning_rate": 9.869230285355897e-05, + "loss": 0.0318, + "step": 68400 + }, + { + "epoch": 0.218944, + "grad_norm": 0.09251537765404758, + "learning_rate": 9.869154136007044e-05, + "loss": 0.0329, + "step": 68420 + }, + { + "epoch": 0.219008, + "grad_norm": 0.057662051956436707, + "learning_rate": 9.869077964787086e-05, + "loss": 0.0322, + "step": 68440 + }, + { + "epoch": 0.219072, + "grad_norm": 0.09898662378028712, + "learning_rate": 9.869001771696367e-05, + "loss": 0.0316, + "step": 68460 + }, + { + "epoch": 0.219136, + "grad_norm": 0.05781578600547082, + "learning_rate": 9.868925556735228e-05, + "loss": 0.0324, + "step": 68480 + }, + { + "epoch": 0.2192, + "grad_norm": 0.07605016040599795, + "learning_rate": 9.868849319904012e-05, + "loss": 0.0305, + "step": 68500 + }, + { + "epoch": 0.219264, + "grad_norm": 0.08378701150905193, + "learning_rate": 9.86877306120306e-05, + "loss": 0.0317, + "step": 68520 + }, + { + "epoch": 0.219328, + "grad_norm": 0.1001351306044324, + "learning_rate": 9.868696780632717e-05, + "loss": 0.0299, + "step": 68540 + }, + { + "epoch": 0.219392, + "grad_norm": 0.05959221521103893, + "learning_rate": 9.868620478193324e-05, + "loss": 0.0322, + "step": 68560 + }, + { + "epoch": 0.219456, + "grad_norm": 0.09052356298479967, + "learning_rate": 9.868544153885224e-05, + "loss": 0.0311, + "step": 68580 + }, + { + "epoch": 0.21952, + "grad_norm": 0.12104670831758771, + "learning_rate": 9.86846780770876e-05, + "loss": 0.0319, + "step": 68600 + }, + { + "epoch": 0.219584, + "grad_norm": 0.12493117494130951, + "learning_rate": 9.868391439664274e-05, + "loss": 0.0282, + "step": 68620 + }, + { + "epoch": 0.219648, + "grad_norm": 0.057792980021449156, + "learning_rate": 9.86831504975211e-05, + "loss": 0.0303, + "step": 68640 + }, + { + "epoch": 0.219712, + "grad_norm": 0.08721662730812321, + "learning_rate": 9.868238637972612e-05, + "loss": 0.0289, + "step": 68660 + }, + { + "epoch": 0.219776, + "grad_norm": 0.07111036225678337, + "learning_rate": 9.868162204326122e-05, + "loss": 0.0323, + "step": 68680 + }, + { + "epoch": 0.21984, + "grad_norm": 0.061945063042386415, + "learning_rate": 9.868085748812983e-05, + "loss": 0.0277, + "step": 68700 + }, + { + "epoch": 0.219904, + "grad_norm": 0.06238321433134148, + "learning_rate": 9.868009271433538e-05, + "loss": 0.031, + "step": 68720 + }, + { + "epoch": 0.219968, + "grad_norm": 0.10073922845660825, + "learning_rate": 9.867932772188132e-05, + "loss": 0.0293, + "step": 68740 + }, + { + "epoch": 0.220032, + "grad_norm": 0.07992004236441834, + "learning_rate": 9.86785625107711e-05, + "loss": 0.0293, + "step": 68760 + }, + { + "epoch": 0.220096, + "grad_norm": 0.09691541437464792, + "learning_rate": 9.867779708100812e-05, + "loss": 0.0316, + "step": 68780 + }, + { + "epoch": 0.22016, + "grad_norm": 0.10342710995607346, + "learning_rate": 9.867703143259583e-05, + "loss": 0.0277, + "step": 68800 + }, + { + "epoch": 0.220224, + "grad_norm": 0.053308622387769435, + "learning_rate": 9.867626556553769e-05, + "loss": 0.0322, + "step": 68820 + }, + { + "epoch": 0.220288, + "grad_norm": 0.1079514400244129, + "learning_rate": 9.867549947983712e-05, + "loss": 0.0289, + "step": 68840 + }, + { + "epoch": 0.220352, + "grad_norm": 0.06950479003268839, + "learning_rate": 9.867473317549755e-05, + "loss": 0.0283, + "step": 68860 + }, + { + "epoch": 0.220416, + "grad_norm": 0.11259269088913085, + "learning_rate": 9.867396665252245e-05, + "loss": 0.0311, + "step": 68880 + }, + { + "epoch": 0.22048, + "grad_norm": 0.08215620720888792, + "learning_rate": 9.867319991091526e-05, + "loss": 0.0305, + "step": 68900 + }, + { + "epoch": 0.220544, + "grad_norm": 0.05771123855221282, + "learning_rate": 9.867243295067941e-05, + "loss": 0.0274, + "step": 68920 + }, + { + "epoch": 0.220608, + "grad_norm": 0.06628732950144485, + "learning_rate": 9.867166577181835e-05, + "loss": 0.0289, + "step": 68940 + }, + { + "epoch": 0.220672, + "grad_norm": 0.07917533160137588, + "learning_rate": 9.867089837433552e-05, + "loss": 0.0339, + "step": 68960 + }, + { + "epoch": 0.220736, + "grad_norm": 0.11606543283484985, + "learning_rate": 9.867013075823437e-05, + "loss": 0.0306, + "step": 68980 + }, + { + "epoch": 0.2208, + "grad_norm": 0.08363859604261542, + "learning_rate": 9.866936292351836e-05, + "loss": 0.0306, + "step": 69000 + }, + { + "epoch": 0.220864, + "grad_norm": 0.07062203276195746, + "learning_rate": 9.866859487019093e-05, + "loss": 0.0283, + "step": 69020 + }, + { + "epoch": 0.220928, + "grad_norm": 0.07071115083141023, + "learning_rate": 9.866782659825554e-05, + "loss": 0.03, + "step": 69040 + }, + { + "epoch": 0.220992, + "grad_norm": 0.09655140672597251, + "learning_rate": 9.866705810771562e-05, + "loss": 0.0305, + "step": 69060 + }, + { + "epoch": 0.221056, + "grad_norm": 0.05046077946477986, + "learning_rate": 9.866628939857462e-05, + "loss": 0.0313, + "step": 69080 + }, + { + "epoch": 0.22112, + "grad_norm": 0.060558568359093075, + "learning_rate": 9.866552047083602e-05, + "loss": 0.0292, + "step": 69100 + }, + { + "epoch": 0.221184, + "grad_norm": 0.08303770828837509, + "learning_rate": 9.866475132450325e-05, + "loss": 0.0311, + "step": 69120 + }, + { + "epoch": 0.221248, + "grad_norm": 0.1119541878616985, + "learning_rate": 9.866398195957978e-05, + "loss": 0.0317, + "step": 69140 + }, + { + "epoch": 0.221312, + "grad_norm": 0.07490898762209262, + "learning_rate": 9.866321237606907e-05, + "loss": 0.0281, + "step": 69160 + }, + { + "epoch": 0.221376, + "grad_norm": 0.05855941665835596, + "learning_rate": 9.866244257397455e-05, + "loss": 0.0293, + "step": 69180 + }, + { + "epoch": 0.22144, + "grad_norm": 0.0978977459350901, + "learning_rate": 9.866167255329971e-05, + "loss": 0.0286, + "step": 69200 + }, + { + "epoch": 0.221504, + "grad_norm": 0.06373255795105619, + "learning_rate": 9.866090231404799e-05, + "loss": 0.029, + "step": 69220 + }, + { + "epoch": 0.221568, + "grad_norm": 0.05294443695661544, + "learning_rate": 9.866013185622285e-05, + "loss": 0.0353, + "step": 69240 + }, + { + "epoch": 0.221632, + "grad_norm": 0.11510963286014482, + "learning_rate": 9.865936117982775e-05, + "loss": 0.0298, + "step": 69260 + }, + { + "epoch": 0.221696, + "grad_norm": 0.08900469947997638, + "learning_rate": 9.865859028486616e-05, + "loss": 0.0299, + "step": 69280 + }, + { + "epoch": 0.22176, + "grad_norm": 0.0943613027735701, + "learning_rate": 9.865781917134153e-05, + "loss": 0.0297, + "step": 69300 + }, + { + "epoch": 0.221824, + "grad_norm": 0.05981912376357919, + "learning_rate": 9.865704783925735e-05, + "loss": 0.0324, + "step": 69320 + }, + { + "epoch": 0.221888, + "grad_norm": 0.05941475582224709, + "learning_rate": 9.865627628861704e-05, + "loss": 0.027, + "step": 69340 + }, + { + "epoch": 0.221952, + "grad_norm": 0.058893571788512, + "learning_rate": 9.865550451942412e-05, + "loss": 0.0292, + "step": 69360 + }, + { + "epoch": 0.222016, + "grad_norm": 0.08455912033762589, + "learning_rate": 9.865473253168201e-05, + "loss": 0.0296, + "step": 69380 + }, + { + "epoch": 0.22208, + "grad_norm": 0.08585004571001624, + "learning_rate": 9.86539603253942e-05, + "loss": 0.0314, + "step": 69400 + }, + { + "epoch": 0.222144, + "grad_norm": 0.08374023805836503, + "learning_rate": 9.865318790056416e-05, + "loss": 0.0308, + "step": 69420 + }, + { + "epoch": 0.222208, + "grad_norm": 0.05686443997755606, + "learning_rate": 9.865241525719535e-05, + "loss": 0.0322, + "step": 69440 + }, + { + "epoch": 0.222272, + "grad_norm": 0.09342012286271038, + "learning_rate": 9.865164239529126e-05, + "loss": 0.0317, + "step": 69460 + }, + { + "epoch": 0.222336, + "grad_norm": 0.08434589662213086, + "learning_rate": 9.865086931485532e-05, + "loss": 0.0295, + "step": 69480 + }, + { + "epoch": 0.2224, + "grad_norm": 0.09588798702581318, + "learning_rate": 9.865009601589105e-05, + "loss": 0.0295, + "step": 69500 + }, + { + "epoch": 0.222464, + "grad_norm": 0.1593998903373086, + "learning_rate": 9.864932249840188e-05, + "loss": 0.0317, + "step": 69520 + }, + { + "epoch": 0.222528, + "grad_norm": 0.06020603953728634, + "learning_rate": 9.864854876239133e-05, + "loss": 0.0303, + "step": 69540 + }, + { + "epoch": 0.222592, + "grad_norm": 0.08129111370164327, + "learning_rate": 9.864777480786285e-05, + "loss": 0.0331, + "step": 69560 + }, + { + "epoch": 0.222656, + "grad_norm": 0.13440618136206764, + "learning_rate": 9.864700063481991e-05, + "loss": 0.0308, + "step": 69580 + }, + { + "epoch": 0.22272, + "grad_norm": 0.11914988104000453, + "learning_rate": 9.8646226243266e-05, + "loss": 0.0298, + "step": 69600 + }, + { + "epoch": 0.222784, + "grad_norm": 0.06866231156710112, + "learning_rate": 9.86454516332046e-05, + "loss": 0.031, + "step": 69620 + }, + { + "epoch": 0.222848, + "grad_norm": 0.08603332176271056, + "learning_rate": 9.864467680463917e-05, + "loss": 0.0277, + "step": 69640 + }, + { + "epoch": 0.222912, + "grad_norm": 0.06213886669787686, + "learning_rate": 9.86439017575732e-05, + "loss": 0.0286, + "step": 69660 + }, + { + "epoch": 0.222976, + "grad_norm": 0.13247532075528468, + "learning_rate": 9.86431264920102e-05, + "loss": 0.031, + "step": 69680 + }, + { + "epoch": 0.22304, + "grad_norm": 0.09652823205146623, + "learning_rate": 9.864235100795361e-05, + "loss": 0.0272, + "step": 69700 + }, + { + "epoch": 0.223104, + "grad_norm": 0.07384354115863988, + "learning_rate": 9.864157530540694e-05, + "loss": 0.0314, + "step": 69720 + }, + { + "epoch": 0.223168, + "grad_norm": 0.05372201896788973, + "learning_rate": 9.864079938437367e-05, + "loss": 0.029, + "step": 69740 + }, + { + "epoch": 0.223232, + "grad_norm": 0.0659595111517275, + "learning_rate": 9.864002324485726e-05, + "loss": 0.0302, + "step": 69760 + }, + { + "epoch": 0.223296, + "grad_norm": 0.10875957829049508, + "learning_rate": 9.863924688686123e-05, + "loss": 0.0331, + "step": 69780 + }, + { + "epoch": 0.22336, + "grad_norm": 0.0681331014402996, + "learning_rate": 9.863847031038906e-05, + "loss": 0.0304, + "step": 69800 + }, + { + "epoch": 0.223424, + "grad_norm": 0.08115272395146839, + "learning_rate": 9.863769351544423e-05, + "loss": 0.032, + "step": 69820 + }, + { + "epoch": 0.223488, + "grad_norm": 0.08545763166007583, + "learning_rate": 9.863691650203023e-05, + "loss": 0.0309, + "step": 69840 + }, + { + "epoch": 0.223552, + "grad_norm": 0.04510506112034373, + "learning_rate": 9.863613927015054e-05, + "loss": 0.0299, + "step": 69860 + }, + { + "epoch": 0.223616, + "grad_norm": 0.15148366943931146, + "learning_rate": 9.863536181980869e-05, + "loss": 0.0319, + "step": 69880 + }, + { + "epoch": 0.22368, + "grad_norm": 0.08736265223527875, + "learning_rate": 9.863458415100813e-05, + "loss": 0.032, + "step": 69900 + }, + { + "epoch": 0.223744, + "grad_norm": 0.0995569055451456, + "learning_rate": 9.863380626375237e-05, + "loss": 0.028, + "step": 69920 + }, + { + "epoch": 0.223808, + "grad_norm": 0.0745219615332586, + "learning_rate": 9.86330281580449e-05, + "loss": 0.0282, + "step": 69940 + }, + { + "epoch": 0.223872, + "grad_norm": 0.05943074838776341, + "learning_rate": 9.86322498338892e-05, + "loss": 0.0323, + "step": 69960 + }, + { + "epoch": 0.223936, + "grad_norm": 0.0648720037921864, + "learning_rate": 9.86314712912888e-05, + "loss": 0.0341, + "step": 69980 + }, + { + "epoch": 0.224, + "grad_norm": 0.06541090701503417, + "learning_rate": 9.863069253024719e-05, + "loss": 0.0269, + "step": 70000 + }, + { + "epoch": 0.224064, + "grad_norm": 0.06886532306066297, + "learning_rate": 9.862991355076784e-05, + "loss": 0.0275, + "step": 70020 + }, + { + "epoch": 0.224128, + "grad_norm": 0.051552526477362436, + "learning_rate": 9.862913435285428e-05, + "loss": 0.0305, + "step": 70040 + }, + { + "epoch": 0.224192, + "grad_norm": 0.05431933336465912, + "learning_rate": 9.862835493651e-05, + "loss": 0.0296, + "step": 70060 + }, + { + "epoch": 0.224256, + "grad_norm": 0.06423449453176712, + "learning_rate": 9.862757530173847e-05, + "loss": 0.0275, + "step": 70080 + }, + { + "epoch": 0.22432, + "grad_norm": 0.055022307984107864, + "learning_rate": 9.862679544854326e-05, + "loss": 0.0293, + "step": 70100 + }, + { + "epoch": 0.224384, + "grad_norm": 0.06627376218236176, + "learning_rate": 9.86260153769278e-05, + "loss": 0.0296, + "step": 70120 + }, + { + "epoch": 0.224448, + "grad_norm": 0.04935847000620073, + "learning_rate": 9.862523508689565e-05, + "loss": 0.0301, + "step": 70140 + }, + { + "epoch": 0.224512, + "grad_norm": 0.05689089048743246, + "learning_rate": 9.862445457845029e-05, + "loss": 0.0289, + "step": 70160 + }, + { + "epoch": 0.224576, + "grad_norm": 0.06719833616918527, + "learning_rate": 9.86236738515952e-05, + "loss": 0.0305, + "step": 70180 + }, + { + "epoch": 0.22464, + "grad_norm": 0.07241004830050171, + "learning_rate": 9.862289290633395e-05, + "loss": 0.0344, + "step": 70200 + }, + { + "epoch": 0.224704, + "grad_norm": 0.0564231795951191, + "learning_rate": 9.862211174267e-05, + "loss": 0.0317, + "step": 70220 + }, + { + "epoch": 0.224768, + "grad_norm": 0.07766551135367324, + "learning_rate": 9.862133036060686e-05, + "loss": 0.033, + "step": 70240 + }, + { + "epoch": 0.224832, + "grad_norm": 0.06802968188954439, + "learning_rate": 9.862054876014807e-05, + "loss": 0.0329, + "step": 70260 + }, + { + "epoch": 0.224896, + "grad_norm": 0.11848079958814707, + "learning_rate": 9.86197669412971e-05, + "loss": 0.0292, + "step": 70280 + }, + { + "epoch": 0.22496, + "grad_norm": 0.07979101928775165, + "learning_rate": 9.86189849040575e-05, + "loss": 0.0294, + "step": 70300 + }, + { + "epoch": 0.225024, + "grad_norm": 0.06788643323059138, + "learning_rate": 9.861820264843278e-05, + "loss": 0.0287, + "step": 70320 + }, + { + "epoch": 0.225088, + "grad_norm": 0.11808172951143883, + "learning_rate": 9.861742017442642e-05, + "loss": 0.0294, + "step": 70340 + }, + { + "epoch": 0.225152, + "grad_norm": 0.09142622314927878, + "learning_rate": 9.861663748204195e-05, + "loss": 0.0331, + "step": 70360 + }, + { + "epoch": 0.225216, + "grad_norm": 0.07583686174845784, + "learning_rate": 9.861585457128291e-05, + "loss": 0.0308, + "step": 70380 + }, + { + "epoch": 0.22528, + "grad_norm": 0.13242172687521286, + "learning_rate": 9.861507144215279e-05, + "loss": 0.0316, + "step": 70400 + }, + { + "epoch": 0.225344, + "grad_norm": 0.09477221800311485, + "learning_rate": 9.86142880946551e-05, + "loss": 0.03, + "step": 70420 + }, + { + "epoch": 0.225408, + "grad_norm": 0.06637270182019295, + "learning_rate": 9.861350452879339e-05, + "loss": 0.0276, + "step": 70440 + }, + { + "epoch": 0.225472, + "grad_norm": 0.09414619956863823, + "learning_rate": 9.861272074457115e-05, + "loss": 0.0271, + "step": 70460 + }, + { + "epoch": 0.225536, + "grad_norm": 0.0582309482239349, + "learning_rate": 9.861193674199191e-05, + "loss": 0.0316, + "step": 70480 + }, + { + "epoch": 0.2256, + "grad_norm": 0.0757224232444805, + "learning_rate": 9.861115252105921e-05, + "loss": 0.0335, + "step": 70500 + }, + { + "epoch": 0.225664, + "grad_norm": 0.09264569733967844, + "learning_rate": 9.861036808177654e-05, + "loss": 0.0301, + "step": 70520 + }, + { + "epoch": 0.225728, + "grad_norm": 0.06370993187085014, + "learning_rate": 9.860958342414747e-05, + "loss": 0.0292, + "step": 70540 + }, + { + "epoch": 0.225792, + "grad_norm": 0.04199171767832429, + "learning_rate": 9.860879854817549e-05, + "loss": 0.0298, + "step": 70560 + }, + { + "epoch": 0.225856, + "grad_norm": 0.06694918047491029, + "learning_rate": 9.86080134538641e-05, + "loss": 0.0316, + "step": 70580 + }, + { + "epoch": 0.22592, + "grad_norm": 0.10237383746198378, + "learning_rate": 9.860722814121688e-05, + "loss": 0.033, + "step": 70600 + }, + { + "epoch": 0.225984, + "grad_norm": 0.10330753876007515, + "learning_rate": 9.860644261023733e-05, + "loss": 0.027, + "step": 70620 + }, + { + "epoch": 0.226048, + "grad_norm": 0.051433865369725884, + "learning_rate": 9.8605656860929e-05, + "loss": 0.03, + "step": 70640 + }, + { + "epoch": 0.226112, + "grad_norm": 0.11426363385974622, + "learning_rate": 9.860487089329538e-05, + "loss": 0.0337, + "step": 70660 + }, + { + "epoch": 0.226176, + "grad_norm": 0.1274024731509269, + "learning_rate": 9.860408470734004e-05, + "loss": 0.0293, + "step": 70680 + }, + { + "epoch": 0.22624, + "grad_norm": 0.0709922635448835, + "learning_rate": 9.860329830306648e-05, + "loss": 0.0302, + "step": 70700 + }, + { + "epoch": 0.226304, + "grad_norm": 0.11089704072971417, + "learning_rate": 9.860251168047825e-05, + "loss": 0.0316, + "step": 70720 + }, + { + "epoch": 0.226368, + "grad_norm": 0.05630543056212624, + "learning_rate": 9.860172483957889e-05, + "loss": 0.0334, + "step": 70740 + }, + { + "epoch": 0.226432, + "grad_norm": 0.12199056615744541, + "learning_rate": 9.860093778037191e-05, + "loss": 0.0323, + "step": 70760 + }, + { + "epoch": 0.226496, + "grad_norm": 0.05680338126731628, + "learning_rate": 9.860015050286088e-05, + "loss": 0.03, + "step": 70780 + }, + { + "epoch": 0.22656, + "grad_norm": 0.06036359703963939, + "learning_rate": 9.85993630070493e-05, + "loss": 0.0303, + "step": 70800 + }, + { + "epoch": 0.226624, + "grad_norm": 0.0681112874786638, + "learning_rate": 9.859857529294072e-05, + "loss": 0.031, + "step": 70820 + }, + { + "epoch": 0.226688, + "grad_norm": 0.057277540092733795, + "learning_rate": 9.85977873605387e-05, + "loss": 0.0286, + "step": 70840 + }, + { + "epoch": 0.226752, + "grad_norm": 0.0604675958336251, + "learning_rate": 9.859699920984676e-05, + "loss": 0.0317, + "step": 70860 + }, + { + "epoch": 0.226816, + "grad_norm": 0.07250683647354315, + "learning_rate": 9.859621084086842e-05, + "loss": 0.0319, + "step": 70880 + }, + { + "epoch": 0.22688, + "grad_norm": 0.044788326843599846, + "learning_rate": 9.859542225360726e-05, + "loss": 0.0292, + "step": 70900 + }, + { + "epoch": 0.226944, + "grad_norm": 0.11287796680973239, + "learning_rate": 9.859463344806678e-05, + "loss": 0.031, + "step": 70920 + }, + { + "epoch": 0.227008, + "grad_norm": 0.07257527907318567, + "learning_rate": 9.859384442425057e-05, + "loss": 0.0307, + "step": 70940 + }, + { + "epoch": 0.227072, + "grad_norm": 0.054857796379834636, + "learning_rate": 9.859305518216215e-05, + "loss": 0.0328, + "step": 70960 + }, + { + "epoch": 0.227136, + "grad_norm": 0.054961980582535576, + "learning_rate": 9.859226572180507e-05, + "loss": 0.0344, + "step": 70980 + }, + { + "epoch": 0.2272, + "grad_norm": 0.05044270101863525, + "learning_rate": 9.859147604318287e-05, + "loss": 0.029, + "step": 71000 + }, + { + "epoch": 0.227264, + "grad_norm": 0.10841430005961779, + "learning_rate": 9.859068614629907e-05, + "loss": 0.0319, + "step": 71020 + }, + { + "epoch": 0.227328, + "grad_norm": 0.08221273153549728, + "learning_rate": 9.858989603115729e-05, + "loss": 0.0288, + "step": 71040 + }, + { + "epoch": 0.227392, + "grad_norm": 0.09097897179277246, + "learning_rate": 9.858910569776101e-05, + "loss": 0.0335, + "step": 71060 + }, + { + "epoch": 0.227456, + "grad_norm": 0.055297123314702185, + "learning_rate": 9.858831514611382e-05, + "loss": 0.0282, + "step": 71080 + }, + { + "epoch": 0.22752, + "grad_norm": 0.07608198757606531, + "learning_rate": 9.858752437621925e-05, + "loss": 0.0307, + "step": 71100 + }, + { + "epoch": 0.227584, + "grad_norm": 0.0599109050514312, + "learning_rate": 9.858673338808086e-05, + "loss": 0.0312, + "step": 71120 + }, + { + "epoch": 0.227648, + "grad_norm": 0.03942068634870746, + "learning_rate": 9.858594218170221e-05, + "loss": 0.0252, + "step": 71140 + }, + { + "epoch": 0.227712, + "grad_norm": 0.05416762234939015, + "learning_rate": 9.858515075708685e-05, + "loss": 0.0282, + "step": 71160 + }, + { + "epoch": 0.227776, + "grad_norm": 0.12121948936915276, + "learning_rate": 9.858435911423831e-05, + "loss": 0.0309, + "step": 71180 + }, + { + "epoch": 0.22784, + "grad_norm": 0.041868368960762514, + "learning_rate": 9.858356725316019e-05, + "loss": 0.0299, + "step": 71200 + }, + { + "epoch": 0.227904, + "grad_norm": 0.058200155555062395, + "learning_rate": 9.858277517385601e-05, + "loss": 0.0307, + "step": 71220 + }, + { + "epoch": 0.227968, + "grad_norm": 0.06608379822285913, + "learning_rate": 9.858198287632935e-05, + "loss": 0.0307, + "step": 71240 + }, + { + "epoch": 0.228032, + "grad_norm": 0.05708604779548623, + "learning_rate": 9.858119036058377e-05, + "loss": 0.0345, + "step": 71260 + }, + { + "epoch": 0.228096, + "grad_norm": 0.08899591988316054, + "learning_rate": 9.85803976266228e-05, + "loss": 0.0348, + "step": 71280 + }, + { + "epoch": 0.22816, + "grad_norm": 0.08169544353497313, + "learning_rate": 9.857960467445003e-05, + "loss": 0.0321, + "step": 71300 + }, + { + "epoch": 0.228224, + "grad_norm": 0.1348740975243117, + "learning_rate": 9.857881150406901e-05, + "loss": 0.029, + "step": 71320 + }, + { + "epoch": 0.228288, + "grad_norm": 0.06257130258345407, + "learning_rate": 9.857801811548332e-05, + "loss": 0.0306, + "step": 71340 + }, + { + "epoch": 0.228352, + "grad_norm": 0.08424239104082941, + "learning_rate": 9.857722450869648e-05, + "loss": 0.0263, + "step": 71360 + }, + { + "epoch": 0.228416, + "grad_norm": 0.0699461452147675, + "learning_rate": 9.857643068371212e-05, + "loss": 0.0336, + "step": 71380 + }, + { + "epoch": 0.22848, + "grad_norm": 0.07777112326677162, + "learning_rate": 9.857563664053376e-05, + "loss": 0.0332, + "step": 71400 + }, + { + "epoch": 0.228544, + "grad_norm": 0.08098745005590956, + "learning_rate": 9.857484237916497e-05, + "loss": 0.0284, + "step": 71420 + }, + { + "epoch": 0.228608, + "grad_norm": 0.07398796965985235, + "learning_rate": 9.857404789960931e-05, + "loss": 0.0315, + "step": 71440 + }, + { + "epoch": 0.228672, + "grad_norm": 0.07373104298183948, + "learning_rate": 9.857325320187038e-05, + "loss": 0.0299, + "step": 71460 + }, + { + "epoch": 0.228736, + "grad_norm": 0.12303508748213184, + "learning_rate": 9.857245828595173e-05, + "loss": 0.0327, + "step": 71480 + }, + { + "epoch": 0.2288, + "grad_norm": 0.057822655736534175, + "learning_rate": 9.857166315185693e-05, + "loss": 0.0337, + "step": 71500 + }, + { + "epoch": 0.228864, + "grad_norm": 0.05462273842366747, + "learning_rate": 9.857086779958957e-05, + "loss": 0.033, + "step": 71520 + }, + { + "epoch": 0.228928, + "grad_norm": 0.05720386944429372, + "learning_rate": 9.857007222915319e-05, + "loss": 0.0306, + "step": 71540 + }, + { + "epoch": 0.228992, + "grad_norm": 0.07307075942986503, + "learning_rate": 9.856927644055138e-05, + "loss": 0.0276, + "step": 71560 + }, + { + "epoch": 0.229056, + "grad_norm": 0.06341500489193089, + "learning_rate": 9.856848043378772e-05, + "loss": 0.0278, + "step": 71580 + }, + { + "epoch": 0.22912, + "grad_norm": 0.06210893774492558, + "learning_rate": 9.856768420886577e-05, + "loss": 0.0302, + "step": 71600 + }, + { + "epoch": 0.229184, + "grad_norm": 0.10237382693517903, + "learning_rate": 9.856688776578914e-05, + "loss": 0.0297, + "step": 71620 + }, + { + "epoch": 0.229248, + "grad_norm": 0.05943778262486167, + "learning_rate": 9.856609110456136e-05, + "loss": 0.0313, + "step": 71640 + }, + { + "epoch": 0.229312, + "grad_norm": 0.06271999406604178, + "learning_rate": 9.856529422518606e-05, + "loss": 0.0325, + "step": 71660 + }, + { + "epoch": 0.229376, + "grad_norm": 0.06706829684427264, + "learning_rate": 9.856449712766678e-05, + "loss": 0.0277, + "step": 71680 + }, + { + "epoch": 0.22944, + "grad_norm": 0.06314365798703563, + "learning_rate": 9.85636998120071e-05, + "loss": 0.0296, + "step": 71700 + }, + { + "epoch": 0.229504, + "grad_norm": 0.052483717606852866, + "learning_rate": 9.856290227821064e-05, + "loss": 0.0293, + "step": 71720 + }, + { + "epoch": 0.229568, + "grad_norm": 0.08172471901602323, + "learning_rate": 9.856210452628092e-05, + "loss": 0.0282, + "step": 71740 + }, + { + "epoch": 0.229632, + "grad_norm": 0.11113952485063047, + "learning_rate": 9.85613065562216e-05, + "loss": 0.0327, + "step": 71760 + }, + { + "epoch": 0.229696, + "grad_norm": 0.06440926505141023, + "learning_rate": 9.85605083680362e-05, + "loss": 0.0297, + "step": 71780 + }, + { + "epoch": 0.22976, + "grad_norm": 0.05531329358997412, + "learning_rate": 9.855970996172833e-05, + "loss": 0.0308, + "step": 71800 + }, + { + "epoch": 0.229824, + "grad_norm": 0.06196579071209111, + "learning_rate": 9.855891133730158e-05, + "loss": 0.0334, + "step": 71820 + }, + { + "epoch": 0.229888, + "grad_norm": 0.10302780948203324, + "learning_rate": 9.855811249475952e-05, + "loss": 0.0308, + "step": 71840 + }, + { + "epoch": 0.229952, + "grad_norm": 0.11494051287857439, + "learning_rate": 9.855731343410576e-05, + "loss": 0.0278, + "step": 71860 + }, + { + "epoch": 0.230016, + "grad_norm": 0.05434459077806702, + "learning_rate": 9.855651415534391e-05, + "loss": 0.0299, + "step": 71880 + }, + { + "epoch": 0.23008, + "grad_norm": 0.07056731526516997, + "learning_rate": 9.855571465847749e-05, + "loss": 0.0277, + "step": 71900 + }, + { + "epoch": 0.230144, + "grad_norm": 0.0903388774666306, + "learning_rate": 9.855491494351015e-05, + "loss": 0.0293, + "step": 71920 + }, + { + "epoch": 0.230208, + "grad_norm": 0.04642249806281933, + "learning_rate": 9.855411501044545e-05, + "loss": 0.0267, + "step": 71940 + }, + { + "epoch": 0.230272, + "grad_norm": 0.08195177250898336, + "learning_rate": 9.8553314859287e-05, + "loss": 0.0305, + "step": 71960 + }, + { + "epoch": 0.230336, + "grad_norm": 0.045829004867807194, + "learning_rate": 9.855251449003842e-05, + "loss": 0.031, + "step": 71980 + }, + { + "epoch": 0.2304, + "grad_norm": 0.1869064642574679, + "learning_rate": 9.855171390270324e-05, + "loss": 0.0309, + "step": 72000 + }, + { + "epoch": 0.230464, + "grad_norm": 0.0764603708826692, + "learning_rate": 9.85509130972851e-05, + "loss": 0.0311, + "step": 72020 + }, + { + "epoch": 0.230528, + "grad_norm": 0.15592349234233618, + "learning_rate": 9.85501120737876e-05, + "loss": 0.0314, + "step": 72040 + }, + { + "epoch": 0.230592, + "grad_norm": 0.10552664150240071, + "learning_rate": 9.854931083221433e-05, + "loss": 0.0318, + "step": 72060 + }, + { + "epoch": 0.230656, + "grad_norm": 0.07495180337756942, + "learning_rate": 9.854850937256887e-05, + "loss": 0.0319, + "step": 72080 + }, + { + "epoch": 0.23072, + "grad_norm": 0.05206205638687262, + "learning_rate": 9.854770769485484e-05, + "loss": 0.0303, + "step": 72100 + }, + { + "epoch": 0.230784, + "grad_norm": 0.0861631549347789, + "learning_rate": 9.854690579907584e-05, + "loss": 0.0298, + "step": 72120 + }, + { + "epoch": 0.230848, + "grad_norm": 0.09837617101043866, + "learning_rate": 9.854610368523547e-05, + "loss": 0.0284, + "step": 72140 + }, + { + "epoch": 0.230912, + "grad_norm": 0.13567813095528705, + "learning_rate": 9.854530135333733e-05, + "loss": 0.0308, + "step": 72160 + }, + { + "epoch": 0.230976, + "grad_norm": 0.11513132331976397, + "learning_rate": 9.854449880338503e-05, + "loss": 0.0315, + "step": 72180 + }, + { + "epoch": 0.23104, + "grad_norm": 0.16515107381220356, + "learning_rate": 9.854369603538216e-05, + "loss": 0.0355, + "step": 72200 + }, + { + "epoch": 0.231104, + "grad_norm": 0.05606183173010271, + "learning_rate": 9.854289304933234e-05, + "loss": 0.0293, + "step": 72220 + }, + { + "epoch": 0.231168, + "grad_norm": 0.0849731232024897, + "learning_rate": 9.854208984523918e-05, + "loss": 0.0301, + "step": 72240 + }, + { + "epoch": 0.231232, + "grad_norm": 0.06643895312220752, + "learning_rate": 9.854128642310627e-05, + "loss": 0.0303, + "step": 72260 + }, + { + "epoch": 0.231296, + "grad_norm": 0.058753004920061226, + "learning_rate": 9.854048278293723e-05, + "loss": 0.0294, + "step": 72280 + }, + { + "epoch": 0.23136, + "grad_norm": 0.24123233547069345, + "learning_rate": 9.853967892473569e-05, + "loss": 0.0291, + "step": 72300 + }, + { + "epoch": 0.231424, + "grad_norm": 0.04594092651383566, + "learning_rate": 9.853887484850523e-05, + "loss": 0.0288, + "step": 72320 + }, + { + "epoch": 0.231488, + "grad_norm": 0.1347517942808203, + "learning_rate": 9.853807055424947e-05, + "loss": 0.0308, + "step": 72340 + }, + { + "epoch": 0.231552, + "grad_norm": 0.09659551760733022, + "learning_rate": 9.853726604197202e-05, + "loss": 0.0285, + "step": 72360 + }, + { + "epoch": 0.231616, + "grad_norm": 0.05950393429453599, + "learning_rate": 9.85364613116765e-05, + "loss": 0.0291, + "step": 72380 + }, + { + "epoch": 0.23168, + "grad_norm": 0.07527490142209554, + "learning_rate": 9.853565636336652e-05, + "loss": 0.0313, + "step": 72400 + }, + { + "epoch": 0.231744, + "grad_norm": 0.08145133182154145, + "learning_rate": 9.853485119704571e-05, + "loss": 0.0302, + "step": 72420 + }, + { + "epoch": 0.231808, + "grad_norm": 0.05019215430570018, + "learning_rate": 9.853404581271766e-05, + "loss": 0.0287, + "step": 72440 + }, + { + "epoch": 0.231872, + "grad_norm": 0.060856781847081196, + "learning_rate": 9.853324021038602e-05, + "loss": 0.0318, + "step": 72460 + }, + { + "epoch": 0.231936, + "grad_norm": 0.05248657855799008, + "learning_rate": 9.853243439005437e-05, + "loss": 0.029, + "step": 72480 + }, + { + "epoch": 0.232, + "grad_norm": 0.08487587807572816, + "learning_rate": 9.853162835172637e-05, + "loss": 0.0312, + "step": 72500 + }, + { + "epoch": 0.232064, + "grad_norm": 0.06194473556295857, + "learning_rate": 9.853082209540562e-05, + "loss": 0.0273, + "step": 72520 + }, + { + "epoch": 0.232128, + "grad_norm": 0.058843435829742614, + "learning_rate": 9.853001562109573e-05, + "loss": 0.0322, + "step": 72540 + }, + { + "epoch": 0.232192, + "grad_norm": 0.13416378690502218, + "learning_rate": 9.852920892880036e-05, + "loss": 0.0318, + "step": 72560 + }, + { + "epoch": 0.232256, + "grad_norm": 0.056075980759534405, + "learning_rate": 9.852840201852307e-05, + "loss": 0.0323, + "step": 72580 + }, + { + "epoch": 0.23232, + "grad_norm": 0.08170905644250934, + "learning_rate": 9.852759489026755e-05, + "loss": 0.0316, + "step": 72600 + }, + { + "epoch": 0.232384, + "grad_norm": 0.08291162617534513, + "learning_rate": 9.852678754403741e-05, + "loss": 0.0299, + "step": 72620 + }, + { + "epoch": 0.232448, + "grad_norm": 0.07373489452897254, + "learning_rate": 9.852597997983625e-05, + "loss": 0.0277, + "step": 72640 + }, + { + "epoch": 0.232512, + "grad_norm": 0.06618492591109143, + "learning_rate": 9.85251721976677e-05, + "loss": 0.0321, + "step": 72660 + }, + { + "epoch": 0.232576, + "grad_norm": 0.11748038921821068, + "learning_rate": 9.852436419753542e-05, + "loss": 0.0331, + "step": 72680 + }, + { + "epoch": 0.23264, + "grad_norm": 0.1160299393632872, + "learning_rate": 9.852355597944303e-05, + "loss": 0.0311, + "step": 72700 + }, + { + "epoch": 0.232704, + "grad_norm": 0.05902344545507101, + "learning_rate": 9.852274754339412e-05, + "loss": 0.0298, + "step": 72720 + }, + { + "epoch": 0.232768, + "grad_norm": 0.05753557154074273, + "learning_rate": 9.852193888939238e-05, + "loss": 0.0302, + "step": 72740 + }, + { + "epoch": 0.232832, + "grad_norm": 0.056919126024623456, + "learning_rate": 9.85211300174414e-05, + "loss": 0.034, + "step": 72760 + }, + { + "epoch": 0.232896, + "grad_norm": 0.050485798781149886, + "learning_rate": 9.852032092754483e-05, + "loss": 0.0328, + "step": 72780 + }, + { + "epoch": 0.23296, + "grad_norm": 0.06724347152252266, + "learning_rate": 9.85195116197063e-05, + "loss": 0.0313, + "step": 72800 + }, + { + "epoch": 0.233024, + "grad_norm": 0.0749119493276647, + "learning_rate": 9.851870209392944e-05, + "loss": 0.0285, + "step": 72820 + }, + { + "epoch": 0.233088, + "grad_norm": 0.0747565212144267, + "learning_rate": 9.851789235021791e-05, + "loss": 0.033, + "step": 72840 + }, + { + "epoch": 0.233152, + "grad_norm": 0.14922068837177918, + "learning_rate": 9.851708238857534e-05, + "loss": 0.0269, + "step": 72860 + }, + { + "epoch": 0.233216, + "grad_norm": 0.130045536240356, + "learning_rate": 9.851627220900533e-05, + "loss": 0.031, + "step": 72880 + }, + { + "epoch": 0.23328, + "grad_norm": 0.08096347998963688, + "learning_rate": 9.851546181151157e-05, + "loss": 0.0278, + "step": 72900 + }, + { + "epoch": 0.233344, + "grad_norm": 0.13871661455987636, + "learning_rate": 9.851465119609767e-05, + "loss": 0.0308, + "step": 72920 + }, + { + "epoch": 0.233408, + "grad_norm": 0.06201813537612768, + "learning_rate": 9.851384036276726e-05, + "loss": 0.0304, + "step": 72940 + }, + { + "epoch": 0.233472, + "grad_norm": 0.08931282117946661, + "learning_rate": 9.851302931152402e-05, + "loss": 0.0313, + "step": 72960 + }, + { + "epoch": 0.233536, + "grad_norm": 0.09496835265923786, + "learning_rate": 9.851221804237157e-05, + "loss": 0.0264, + "step": 72980 + }, + { + "epoch": 0.2336, + "grad_norm": 0.056307204672306474, + "learning_rate": 9.851140655531356e-05, + "loss": 0.0301, + "step": 73000 + }, + { + "epoch": 0.233664, + "grad_norm": 0.06429057948898363, + "learning_rate": 9.851059485035363e-05, + "loss": 0.0319, + "step": 73020 + }, + { + "epoch": 0.233728, + "grad_norm": 0.08041716825232202, + "learning_rate": 9.850978292749545e-05, + "loss": 0.0307, + "step": 73040 + }, + { + "epoch": 0.233792, + "grad_norm": 0.06610063827515962, + "learning_rate": 9.850897078674262e-05, + "loss": 0.03, + "step": 73060 + }, + { + "epoch": 0.233856, + "grad_norm": 0.06478378448968702, + "learning_rate": 9.850815842809882e-05, + "loss": 0.0318, + "step": 73080 + }, + { + "epoch": 0.23392, + "grad_norm": 0.08875153509941616, + "learning_rate": 9.850734585156769e-05, + "loss": 0.0323, + "step": 73100 + }, + { + "epoch": 0.233984, + "grad_norm": 0.1461013541162037, + "learning_rate": 9.85065330571529e-05, + "loss": 0.0327, + "step": 73120 + }, + { + "epoch": 0.234048, + "grad_norm": 0.12015203748012013, + "learning_rate": 9.850572004485807e-05, + "loss": 0.0334, + "step": 73140 + }, + { + "epoch": 0.234112, + "grad_norm": 0.07319280325016944, + "learning_rate": 9.850490681468688e-05, + "loss": 0.0301, + "step": 73160 + }, + { + "epoch": 0.234176, + "grad_norm": 0.08537072315072389, + "learning_rate": 9.850409336664295e-05, + "loss": 0.0324, + "step": 73180 + }, + { + "epoch": 0.23424, + "grad_norm": 0.06933867739439176, + "learning_rate": 9.850327970072995e-05, + "loss": 0.0292, + "step": 73200 + }, + { + "epoch": 0.234304, + "grad_norm": 0.10618527187894503, + "learning_rate": 9.850246581695155e-05, + "loss": 0.0299, + "step": 73220 + }, + { + "epoch": 0.234368, + "grad_norm": 0.0592141978464065, + "learning_rate": 9.850165171531138e-05, + "loss": 0.0316, + "step": 73240 + }, + { + "epoch": 0.234432, + "grad_norm": 0.11331446876855693, + "learning_rate": 9.850083739581313e-05, + "loss": 0.0282, + "step": 73260 + }, + { + "epoch": 0.234496, + "grad_norm": 0.07214465406807367, + "learning_rate": 9.850002285846041e-05, + "loss": 0.0307, + "step": 73280 + }, + { + "epoch": 0.23456, + "grad_norm": 0.09214934523273946, + "learning_rate": 9.849920810325692e-05, + "loss": 0.0323, + "step": 73300 + }, + { + "epoch": 0.234624, + "grad_norm": 0.06662451040451721, + "learning_rate": 9.849839313020632e-05, + "loss": 0.0321, + "step": 73320 + }, + { + "epoch": 0.234688, + "grad_norm": 0.05681010824671877, + "learning_rate": 9.849757793931224e-05, + "loss": 0.0305, + "step": 73340 + }, + { + "epoch": 0.234752, + "grad_norm": 0.08932626849913457, + "learning_rate": 9.849676253057835e-05, + "loss": 0.0307, + "step": 73360 + }, + { + "epoch": 0.234816, + "grad_norm": 0.053972331227776525, + "learning_rate": 9.849594690400834e-05, + "loss": 0.0272, + "step": 73380 + }, + { + "epoch": 0.23488, + "grad_norm": 0.06872158527024105, + "learning_rate": 9.849513105960586e-05, + "loss": 0.0298, + "step": 73400 + }, + { + "epoch": 0.234944, + "grad_norm": 0.05752597820959559, + "learning_rate": 9.849431499737455e-05, + "loss": 0.029, + "step": 73420 + }, + { + "epoch": 0.235008, + "grad_norm": 0.07124086533570095, + "learning_rate": 9.84934987173181e-05, + "loss": 0.0298, + "step": 73440 + }, + { + "epoch": 0.235072, + "grad_norm": 0.0559715033058606, + "learning_rate": 9.849268221944017e-05, + "loss": 0.0291, + "step": 73460 + }, + { + "epoch": 0.235136, + "grad_norm": 0.08191126959248517, + "learning_rate": 9.849186550374443e-05, + "loss": 0.0278, + "step": 73480 + }, + { + "epoch": 0.2352, + "grad_norm": 0.08273739800958059, + "learning_rate": 9.849104857023455e-05, + "loss": 0.0297, + "step": 73500 + }, + { + "epoch": 0.235264, + "grad_norm": 0.07598945937019032, + "learning_rate": 9.84902314189142e-05, + "loss": 0.0301, + "step": 73520 + }, + { + "epoch": 0.235328, + "grad_norm": 0.11996537515976421, + "learning_rate": 9.848941404978704e-05, + "loss": 0.0289, + "step": 73540 + }, + { + "epoch": 0.235392, + "grad_norm": 0.06735558748171384, + "learning_rate": 9.848859646285674e-05, + "loss": 0.0314, + "step": 73560 + }, + { + "epoch": 0.235456, + "grad_norm": 0.1376354388182584, + "learning_rate": 9.8487778658127e-05, + "loss": 0.0341, + "step": 73580 + }, + { + "epoch": 0.23552, + "grad_norm": 0.08248441431155565, + "learning_rate": 9.848696063560145e-05, + "loss": 0.0335, + "step": 73600 + }, + { + "epoch": 0.235584, + "grad_norm": 0.17015547019203925, + "learning_rate": 9.848614239528382e-05, + "loss": 0.0308, + "step": 73620 + }, + { + "epoch": 0.235648, + "grad_norm": 0.08528335001360855, + "learning_rate": 9.848532393717773e-05, + "loss": 0.0299, + "step": 73640 + }, + { + "epoch": 0.235712, + "grad_norm": 0.11351969947048708, + "learning_rate": 9.848450526128687e-05, + "loss": 0.0362, + "step": 73660 + }, + { + "epoch": 0.235776, + "grad_norm": 0.17909695871508194, + "learning_rate": 9.848368636761496e-05, + "loss": 0.0305, + "step": 73680 + }, + { + "epoch": 0.23584, + "grad_norm": 0.060247263401596564, + "learning_rate": 9.848286725616563e-05, + "loss": 0.0293, + "step": 73700 + }, + { + "epoch": 0.235904, + "grad_norm": 0.1254765242542778, + "learning_rate": 9.848204792694258e-05, + "loss": 0.0282, + "step": 73720 + }, + { + "epoch": 0.235968, + "grad_norm": 0.05062419509293313, + "learning_rate": 9.848122837994947e-05, + "loss": 0.0303, + "step": 73740 + }, + { + "epoch": 0.236032, + "grad_norm": 0.06058153936585436, + "learning_rate": 9.848040861519002e-05, + "loss": 0.026, + "step": 73760 + }, + { + "epoch": 0.236096, + "grad_norm": 0.08229034278063106, + "learning_rate": 9.847958863266787e-05, + "loss": 0.0274, + "step": 73780 + }, + { + "epoch": 0.23616, + "grad_norm": 0.06160442766718873, + "learning_rate": 9.847876843238672e-05, + "loss": 0.0297, + "step": 73800 + }, + { + "epoch": 0.236224, + "grad_norm": 0.07373141586076273, + "learning_rate": 9.847794801435026e-05, + "loss": 0.0308, + "step": 73820 + }, + { + "epoch": 0.236288, + "grad_norm": 0.04087093346366814, + "learning_rate": 9.847712737856218e-05, + "loss": 0.0316, + "step": 73840 + }, + { + "epoch": 0.236352, + "grad_norm": 0.10803671903985528, + "learning_rate": 9.847630652502616e-05, + "loss": 0.0307, + "step": 73860 + }, + { + "epoch": 0.236416, + "grad_norm": 0.05267634464677613, + "learning_rate": 9.847548545374587e-05, + "loss": 0.0312, + "step": 73880 + }, + { + "epoch": 0.23648, + "grad_norm": 0.08816769588836898, + "learning_rate": 9.847466416472501e-05, + "loss": 0.0292, + "step": 73900 + }, + { + "epoch": 0.236544, + "grad_norm": 0.06608869805524964, + "learning_rate": 9.847384265796728e-05, + "loss": 0.0271, + "step": 73920 + }, + { + "epoch": 0.236608, + "grad_norm": 0.11830898343450925, + "learning_rate": 9.847302093347636e-05, + "loss": 0.0304, + "step": 73940 + }, + { + "epoch": 0.236672, + "grad_norm": 0.11833431684363141, + "learning_rate": 9.847219899125594e-05, + "loss": 0.0309, + "step": 73960 + }, + { + "epoch": 0.236736, + "grad_norm": 0.10763763187066101, + "learning_rate": 9.847137683130973e-05, + "loss": 0.0306, + "step": 73980 + }, + { + "epoch": 0.2368, + "grad_norm": 0.11666805473157174, + "learning_rate": 9.847055445364139e-05, + "loss": 0.0297, + "step": 74000 + }, + { + "epoch": 0.236864, + "grad_norm": 0.06455038978450987, + "learning_rate": 9.846973185825463e-05, + "loss": 0.0338, + "step": 74020 + }, + { + "epoch": 0.236928, + "grad_norm": 0.06422816194578176, + "learning_rate": 9.846890904515315e-05, + "loss": 0.0336, + "step": 74040 + }, + { + "epoch": 0.236992, + "grad_norm": 0.06974785447800486, + "learning_rate": 9.846808601434065e-05, + "loss": 0.0318, + "step": 74060 + }, + { + "epoch": 0.237056, + "grad_norm": 0.0621248139391275, + "learning_rate": 9.846726276582081e-05, + "loss": 0.0277, + "step": 74080 + }, + { + "epoch": 0.23712, + "grad_norm": 0.0576323594134353, + "learning_rate": 9.846643929959735e-05, + "loss": 0.028, + "step": 74100 + }, + { + "epoch": 0.237184, + "grad_norm": 0.07056610832566135, + "learning_rate": 9.846561561567394e-05, + "loss": 0.0311, + "step": 74120 + }, + { + "epoch": 0.237248, + "grad_norm": 0.07671404815294516, + "learning_rate": 9.84647917140543e-05, + "loss": 0.0309, + "step": 74140 + }, + { + "epoch": 0.237312, + "grad_norm": 0.09053077568032004, + "learning_rate": 9.846396759474213e-05, + "loss": 0.028, + "step": 74160 + }, + { + "epoch": 0.237376, + "grad_norm": 0.08957798868528549, + "learning_rate": 9.846314325774113e-05, + "loss": 0.03, + "step": 74180 + }, + { + "epoch": 0.23744, + "grad_norm": 0.05337466435179542, + "learning_rate": 9.8462318703055e-05, + "loss": 0.0315, + "step": 74200 + }, + { + "epoch": 0.237504, + "grad_norm": 0.06968516692025764, + "learning_rate": 9.846149393068744e-05, + "loss": 0.0284, + "step": 74220 + }, + { + "epoch": 0.237568, + "grad_norm": 0.05058248393066313, + "learning_rate": 9.846066894064217e-05, + "loss": 0.0279, + "step": 74240 + }, + { + "epoch": 0.237632, + "grad_norm": 0.05316381100548303, + "learning_rate": 9.845984373292286e-05, + "loss": 0.0274, + "step": 74260 + }, + { + "epoch": 0.237696, + "grad_norm": 0.10876095392449166, + "learning_rate": 9.845901830753325e-05, + "loss": 0.0255, + "step": 74280 + }, + { + "epoch": 0.23776, + "grad_norm": 0.10985201297319612, + "learning_rate": 9.845819266447705e-05, + "loss": 0.026, + "step": 74300 + }, + { + "epoch": 0.237824, + "grad_norm": 0.0937930379900109, + "learning_rate": 9.845736680375795e-05, + "loss": 0.0311, + "step": 74320 + }, + { + "epoch": 0.237888, + "grad_norm": 0.091320114166907, + "learning_rate": 9.845654072537967e-05, + "loss": 0.0294, + "step": 74340 + }, + { + "epoch": 0.237952, + "grad_norm": 0.10419620454802064, + "learning_rate": 9.845571442934592e-05, + "loss": 0.0322, + "step": 74360 + }, + { + "epoch": 0.238016, + "grad_norm": 0.11630033619213617, + "learning_rate": 9.845488791566039e-05, + "loss": 0.035, + "step": 74380 + }, + { + "epoch": 0.23808, + "grad_norm": 0.06448702426022326, + "learning_rate": 9.845406118432685e-05, + "loss": 0.031, + "step": 74400 + }, + { + "epoch": 0.238144, + "grad_norm": 0.07934487373129798, + "learning_rate": 9.845323423534895e-05, + "loss": 0.031, + "step": 74420 + }, + { + "epoch": 0.238208, + "grad_norm": 0.057226907174601424, + "learning_rate": 9.845240706873042e-05, + "loss": 0.0288, + "step": 74440 + }, + { + "epoch": 0.238272, + "grad_norm": 0.05862630157613902, + "learning_rate": 9.8451579684475e-05, + "loss": 0.0272, + "step": 74460 + }, + { + "epoch": 0.238336, + "grad_norm": 0.049791687938707814, + "learning_rate": 9.84507520825864e-05, + "loss": 0.0313, + "step": 74480 + }, + { + "epoch": 0.2384, + "grad_norm": 0.10189010612878872, + "learning_rate": 9.844992426306832e-05, + "loss": 0.0279, + "step": 74500 + }, + { + "epoch": 0.238464, + "grad_norm": 0.05469239053897695, + "learning_rate": 9.844909622592449e-05, + "loss": 0.0325, + "step": 74520 + }, + { + "epoch": 0.238528, + "grad_norm": 0.1497191055849579, + "learning_rate": 9.844826797115861e-05, + "loss": 0.0316, + "step": 74540 + }, + { + "epoch": 0.238592, + "grad_norm": 0.05846861111562548, + "learning_rate": 9.844743949877442e-05, + "loss": 0.0275, + "step": 74560 + }, + { + "epoch": 0.238656, + "grad_norm": 0.051167784732405894, + "learning_rate": 9.844661080877567e-05, + "loss": 0.0285, + "step": 74580 + }, + { + "epoch": 0.23872, + "grad_norm": 0.06046727060241974, + "learning_rate": 9.844578190116602e-05, + "loss": 0.0281, + "step": 74600 + }, + { + "epoch": 0.238784, + "grad_norm": 0.06192622784986801, + "learning_rate": 9.844495277594922e-05, + "loss": 0.0267, + "step": 74620 + }, + { + "epoch": 0.238848, + "grad_norm": 0.07386421502205073, + "learning_rate": 9.844412343312901e-05, + "loss": 0.0288, + "step": 74640 + }, + { + "epoch": 0.238912, + "grad_norm": 0.1034401418418549, + "learning_rate": 9.844329387270912e-05, + "loss": 0.031, + "step": 74660 + }, + { + "epoch": 0.238976, + "grad_norm": 0.05847352846245029, + "learning_rate": 9.844246409469323e-05, + "loss": 0.0284, + "step": 74680 + }, + { + "epoch": 0.23904, + "grad_norm": 0.05030969306590917, + "learning_rate": 9.844163409908512e-05, + "loss": 0.028, + "step": 74700 + }, + { + "epoch": 0.239104, + "grad_norm": 0.041788225454165134, + "learning_rate": 9.844080388588848e-05, + "loss": 0.0298, + "step": 74720 + }, + { + "epoch": 0.239168, + "grad_norm": 0.06180616033186569, + "learning_rate": 9.843997345510705e-05, + "loss": 0.0313, + "step": 74740 + }, + { + "epoch": 0.239232, + "grad_norm": 0.0632717980652912, + "learning_rate": 9.843914280674458e-05, + "loss": 0.0277, + "step": 74760 + }, + { + "epoch": 0.239296, + "grad_norm": 0.05911849306000406, + "learning_rate": 9.843831194080478e-05, + "loss": 0.032, + "step": 74780 + }, + { + "epoch": 0.23936, + "grad_norm": 0.1151497719910364, + "learning_rate": 9.843748085729137e-05, + "loss": 0.0288, + "step": 74800 + }, + { + "epoch": 0.239424, + "grad_norm": 0.11722802954081228, + "learning_rate": 9.843664955620813e-05, + "loss": 0.0279, + "step": 74820 + }, + { + "epoch": 0.239488, + "grad_norm": 0.05516803285041774, + "learning_rate": 9.843581803755874e-05, + "loss": 0.0303, + "step": 74840 + }, + { + "epoch": 0.239552, + "grad_norm": 0.06453990585011933, + "learning_rate": 9.843498630134698e-05, + "loss": 0.0331, + "step": 74860 + }, + { + "epoch": 0.239616, + "grad_norm": 0.07541387803486191, + "learning_rate": 9.843415434757655e-05, + "loss": 0.0326, + "step": 74880 + }, + { + "epoch": 0.23968, + "grad_norm": 0.06242256556353007, + "learning_rate": 9.84333221762512e-05, + "loss": 0.0285, + "step": 74900 + }, + { + "epoch": 0.239744, + "grad_norm": 0.0664522994274791, + "learning_rate": 9.843248978737468e-05, + "loss": 0.0314, + "step": 74920 + }, + { + "epoch": 0.239808, + "grad_norm": 0.10259221337982401, + "learning_rate": 9.843165718095071e-05, + "loss": 0.0303, + "step": 74940 + }, + { + "epoch": 0.239872, + "grad_norm": 0.06844757421631792, + "learning_rate": 9.843082435698303e-05, + "loss": 0.0283, + "step": 74960 + }, + { + "epoch": 0.239936, + "grad_norm": 0.1294513965685381, + "learning_rate": 9.842999131547542e-05, + "loss": 0.029, + "step": 74980 + }, + { + "epoch": 0.24, + "grad_norm": 0.09107291860812793, + "learning_rate": 9.842915805643155e-05, + "loss": 0.0298, + "step": 75000 + }, + { + "epoch": 0.240064, + "grad_norm": 0.065584023919824, + "learning_rate": 9.842832457985524e-05, + "loss": 0.0306, + "step": 75020 + }, + { + "epoch": 0.240128, + "grad_norm": 0.09560901023483112, + "learning_rate": 9.842749088575018e-05, + "loss": 0.0325, + "step": 75040 + }, + { + "epoch": 0.240192, + "grad_norm": 0.08385320206113882, + "learning_rate": 9.842665697412015e-05, + "loss": 0.0313, + "step": 75060 + }, + { + "epoch": 0.240256, + "grad_norm": 0.11271104262567187, + "learning_rate": 9.842582284496886e-05, + "loss": 0.0326, + "step": 75080 + }, + { + "epoch": 0.24032, + "grad_norm": 0.1306758567056416, + "learning_rate": 9.842498849830009e-05, + "loss": 0.0296, + "step": 75100 + }, + { + "epoch": 0.240384, + "grad_norm": 0.08729741724764371, + "learning_rate": 9.842415393411756e-05, + "loss": 0.0319, + "step": 75120 + }, + { + "epoch": 0.240448, + "grad_norm": 0.09138074500122444, + "learning_rate": 9.842331915242503e-05, + "loss": 0.0327, + "step": 75140 + }, + { + "epoch": 0.240512, + "grad_norm": 0.07223615507656593, + "learning_rate": 9.842248415322626e-05, + "loss": 0.0312, + "step": 75160 + }, + { + "epoch": 0.240576, + "grad_norm": 0.09092386466451736, + "learning_rate": 9.842164893652498e-05, + "loss": 0.0304, + "step": 75180 + }, + { + "epoch": 0.24064, + "grad_norm": 0.1060830872137615, + "learning_rate": 9.842081350232497e-05, + "loss": 0.0295, + "step": 75200 + }, + { + "epoch": 0.240704, + "grad_norm": 0.07253465442860489, + "learning_rate": 9.841997785062996e-05, + "loss": 0.0301, + "step": 75220 + }, + { + "epoch": 0.240768, + "grad_norm": 0.12029569981983633, + "learning_rate": 9.841914198144371e-05, + "loss": 0.0305, + "step": 75240 + }, + { + "epoch": 0.240832, + "grad_norm": 0.12878346882860503, + "learning_rate": 9.841830589476997e-05, + "loss": 0.0293, + "step": 75260 + }, + { + "epoch": 0.240896, + "grad_norm": 0.11208580219226276, + "learning_rate": 9.84174695906125e-05, + "loss": 0.0318, + "step": 75280 + }, + { + "epoch": 0.24096, + "grad_norm": 0.07272513327685043, + "learning_rate": 9.841663306897506e-05, + "loss": 0.0283, + "step": 75300 + }, + { + "epoch": 0.241024, + "grad_norm": 0.08059159786577916, + "learning_rate": 9.841579632986141e-05, + "loss": 0.0321, + "step": 75320 + }, + { + "epoch": 0.241088, + "grad_norm": 0.041763518377769, + "learning_rate": 9.841495937327529e-05, + "loss": 0.03, + "step": 75340 + }, + { + "epoch": 0.241152, + "grad_norm": 0.12265735614200002, + "learning_rate": 9.841412219922048e-05, + "loss": 0.0326, + "step": 75360 + }, + { + "epoch": 0.241216, + "grad_norm": 0.08370074327224666, + "learning_rate": 9.841328480770072e-05, + "loss": 0.0342, + "step": 75380 + }, + { + "epoch": 0.24128, + "grad_norm": 0.07261776174199651, + "learning_rate": 9.841244719871979e-05, + "loss": 0.0337, + "step": 75400 + }, + { + "epoch": 0.241344, + "grad_norm": 0.14367652055605545, + "learning_rate": 9.841160937228144e-05, + "loss": 0.0317, + "step": 75420 + }, + { + "epoch": 0.241408, + "grad_norm": 0.06795748208057648, + "learning_rate": 9.841077132838945e-05, + "loss": 0.0336, + "step": 75440 + }, + { + "epoch": 0.241472, + "grad_norm": 0.08862408875577264, + "learning_rate": 9.840993306704758e-05, + "loss": 0.0276, + "step": 75460 + }, + { + "epoch": 0.241536, + "grad_norm": 0.100862640022274, + "learning_rate": 9.840909458825957e-05, + "loss": 0.0327, + "step": 75480 + }, + { + "epoch": 0.2416, + "grad_norm": 0.06623897632055767, + "learning_rate": 9.840825589202922e-05, + "loss": 0.0338, + "step": 75500 + }, + { + "epoch": 0.241664, + "grad_norm": 0.06960896485864733, + "learning_rate": 9.840741697836027e-05, + "loss": 0.0339, + "step": 75520 + }, + { + "epoch": 0.241728, + "grad_norm": 0.055396091183682286, + "learning_rate": 9.84065778472565e-05, + "loss": 0.0333, + "step": 75540 + }, + { + "epoch": 0.241792, + "grad_norm": 0.07051217244228872, + "learning_rate": 9.840573849872168e-05, + "loss": 0.0305, + "step": 75560 + }, + { + "epoch": 0.241856, + "grad_norm": 0.08109862932951696, + "learning_rate": 9.840489893275958e-05, + "loss": 0.0309, + "step": 75580 + }, + { + "epoch": 0.24192, + "grad_norm": 0.082116669636281, + "learning_rate": 9.840405914937397e-05, + "loss": 0.0303, + "step": 75600 + }, + { + "epoch": 0.241984, + "grad_norm": 0.11144333995349691, + "learning_rate": 9.840321914856862e-05, + "loss": 0.0287, + "step": 75620 + }, + { + "epoch": 0.242048, + "grad_norm": 0.07252721450558351, + "learning_rate": 9.840237893034732e-05, + "loss": 0.0319, + "step": 75640 + }, + { + "epoch": 0.242112, + "grad_norm": 0.06338077105411824, + "learning_rate": 9.840153849471382e-05, + "loss": 0.0327, + "step": 75660 + }, + { + "epoch": 0.242176, + "grad_norm": 0.0823634023370179, + "learning_rate": 9.840069784167189e-05, + "loss": 0.0279, + "step": 75680 + }, + { + "epoch": 0.24224, + "grad_norm": 0.05633925323777528, + "learning_rate": 9.839985697122535e-05, + "loss": 0.0279, + "step": 75700 + }, + { + "epoch": 0.242304, + "grad_norm": 0.11868043095235185, + "learning_rate": 9.839901588337794e-05, + "loss": 0.0332, + "step": 75720 + }, + { + "epoch": 0.242368, + "grad_norm": 0.05167566744116622, + "learning_rate": 9.839817457813343e-05, + "loss": 0.0312, + "step": 75740 + }, + { + "epoch": 0.242432, + "grad_norm": 0.07965503603273497, + "learning_rate": 9.839733305549563e-05, + "loss": 0.0311, + "step": 75760 + }, + { + "epoch": 0.242496, + "grad_norm": 0.051797361580601234, + "learning_rate": 9.83964913154683e-05, + "loss": 0.0293, + "step": 75780 + }, + { + "epoch": 0.24256, + "grad_norm": 0.08194270094632075, + "learning_rate": 9.839564935805522e-05, + "loss": 0.0303, + "step": 75800 + }, + { + "epoch": 0.242624, + "grad_norm": 0.11614724221854153, + "learning_rate": 9.839480718326018e-05, + "loss": 0.0293, + "step": 75820 + }, + { + "epoch": 0.242688, + "grad_norm": 0.10095941696256949, + "learning_rate": 9.839396479108696e-05, + "loss": 0.0274, + "step": 75840 + }, + { + "epoch": 0.242752, + "grad_norm": 0.08656951949934724, + "learning_rate": 9.839312218153936e-05, + "loss": 0.0292, + "step": 75860 + }, + { + "epoch": 0.242816, + "grad_norm": 0.05363075283005728, + "learning_rate": 9.839227935462113e-05, + "loss": 0.0263, + "step": 75880 + }, + { + "epoch": 0.24288, + "grad_norm": 0.08136538396257342, + "learning_rate": 9.839143631033609e-05, + "loss": 0.0257, + "step": 75900 + }, + { + "epoch": 0.242944, + "grad_norm": 0.11441507471062821, + "learning_rate": 9.8390593048688e-05, + "loss": 0.0282, + "step": 75920 + }, + { + "epoch": 0.243008, + "grad_norm": 0.053824921405020966, + "learning_rate": 9.838974956968065e-05, + "loss": 0.031, + "step": 75940 + }, + { + "epoch": 0.243072, + "grad_norm": 0.11311796096233652, + "learning_rate": 9.838890587331787e-05, + "loss": 0.033, + "step": 75960 + }, + { + "epoch": 0.243136, + "grad_norm": 0.07760726689821014, + "learning_rate": 9.838806195960339e-05, + "loss": 0.031, + "step": 75980 + }, + { + "epoch": 0.2432, + "grad_norm": 0.054907806782401654, + "learning_rate": 9.838721782854103e-05, + "loss": 0.0321, + "step": 76000 + }, + { + "epoch": 0.243264, + "grad_norm": 0.05421980681761685, + "learning_rate": 9.838637348013459e-05, + "loss": 0.0284, + "step": 76020 + }, + { + "epoch": 0.243328, + "grad_norm": 0.17211193283158316, + "learning_rate": 9.838552891438783e-05, + "loss": 0.0323, + "step": 76040 + }, + { + "epoch": 0.243392, + "grad_norm": 0.12943102314118832, + "learning_rate": 9.83846841313046e-05, + "loss": 0.0331, + "step": 76060 + }, + { + "epoch": 0.243456, + "grad_norm": 0.07306620831294458, + "learning_rate": 9.838383913088865e-05, + "loss": 0.0301, + "step": 76080 + }, + { + "epoch": 0.24352, + "grad_norm": 0.09334050633062034, + "learning_rate": 9.838299391314377e-05, + "loss": 0.0303, + "step": 76100 + }, + { + "epoch": 0.243584, + "grad_norm": 0.11094428704503621, + "learning_rate": 9.838214847807378e-05, + "loss": 0.0275, + "step": 76120 + }, + { + "epoch": 0.243648, + "grad_norm": 0.06014980405667603, + "learning_rate": 9.838130282568248e-05, + "loss": 0.0289, + "step": 76140 + }, + { + "epoch": 0.243712, + "grad_norm": 0.06511569587223197, + "learning_rate": 9.838045695597366e-05, + "loss": 0.0278, + "step": 76160 + }, + { + "epoch": 0.243776, + "grad_norm": 0.12912756406288878, + "learning_rate": 9.837961086895111e-05, + "loss": 0.0279, + "step": 76180 + }, + { + "epoch": 0.24384, + "grad_norm": 0.21140688177877892, + "learning_rate": 9.837876456461863e-05, + "loss": 0.0303, + "step": 76200 + }, + { + "epoch": 0.243904, + "grad_norm": 0.11682353169127459, + "learning_rate": 9.837791804298004e-05, + "loss": 0.0303, + "step": 76220 + }, + { + "epoch": 0.243968, + "grad_norm": 0.04890235485560692, + "learning_rate": 9.837707130403914e-05, + "loss": 0.0284, + "step": 76240 + }, + { + "epoch": 0.244032, + "grad_norm": 0.1443420489632538, + "learning_rate": 9.837622434779972e-05, + "loss": 0.0321, + "step": 76260 + }, + { + "epoch": 0.244096, + "grad_norm": 0.05551651036284427, + "learning_rate": 9.837537717426558e-05, + "loss": 0.0275, + "step": 76280 + }, + { + "epoch": 0.24416, + "grad_norm": 0.05469018564943794, + "learning_rate": 9.837452978344055e-05, + "loss": 0.032, + "step": 76300 + }, + { + "epoch": 0.244224, + "grad_norm": 0.14420695280174145, + "learning_rate": 9.837368217532841e-05, + "loss": 0.0321, + "step": 76320 + }, + { + "epoch": 0.244288, + "grad_norm": 0.08072035424872051, + "learning_rate": 9.837283434993299e-05, + "loss": 0.0328, + "step": 76340 + }, + { + "epoch": 0.244352, + "grad_norm": 0.08134410699162518, + "learning_rate": 9.837198630725809e-05, + "loss": 0.0304, + "step": 76360 + }, + { + "epoch": 0.244416, + "grad_norm": 0.20766627533964196, + "learning_rate": 9.837113804730748e-05, + "loss": 0.0334, + "step": 76380 + }, + { + "epoch": 0.24448, + "grad_norm": 0.07761996894839283, + "learning_rate": 9.837028957008505e-05, + "loss": 0.0304, + "step": 76400 + }, + { + "epoch": 0.244544, + "grad_norm": 0.0718093416297512, + "learning_rate": 9.836944087559454e-05, + "loss": 0.0337, + "step": 76420 + }, + { + "epoch": 0.244608, + "grad_norm": 0.09542797531609302, + "learning_rate": 9.836859196383981e-05, + "loss": 0.0328, + "step": 76440 + }, + { + "epoch": 0.244672, + "grad_norm": 0.07118146419230027, + "learning_rate": 9.836774283482465e-05, + "loss": 0.0301, + "step": 76460 + }, + { + "epoch": 0.244736, + "grad_norm": 0.07169651634961503, + "learning_rate": 9.836689348855287e-05, + "loss": 0.0327, + "step": 76480 + }, + { + "epoch": 0.2448, + "grad_norm": 0.10067836871486986, + "learning_rate": 9.83660439250283e-05, + "loss": 0.0299, + "step": 76500 + }, + { + "epoch": 0.244864, + "grad_norm": 0.0612769273747591, + "learning_rate": 9.836519414425473e-05, + "loss": 0.03, + "step": 76520 + }, + { + "epoch": 0.244928, + "grad_norm": 0.05323430100573117, + "learning_rate": 9.836434414623601e-05, + "loss": 0.0338, + "step": 76540 + }, + { + "epoch": 0.244992, + "grad_norm": 0.060166023215523064, + "learning_rate": 9.836349393097595e-05, + "loss": 0.0303, + "step": 76560 + }, + { + "epoch": 0.245056, + "grad_norm": 0.06670798865807341, + "learning_rate": 9.836264349847834e-05, + "loss": 0.0276, + "step": 76580 + }, + { + "epoch": 0.24512, + "grad_norm": 0.0924887380559937, + "learning_rate": 9.836179284874705e-05, + "loss": 0.0303, + "step": 76600 + }, + { + "epoch": 0.245184, + "grad_norm": 0.1748014715501169, + "learning_rate": 9.836094198178585e-05, + "loss": 0.0306, + "step": 76620 + }, + { + "epoch": 0.245248, + "grad_norm": 0.06334115605273537, + "learning_rate": 9.836009089759858e-05, + "loss": 0.0314, + "step": 76640 + }, + { + "epoch": 0.245312, + "grad_norm": 0.07557922164391759, + "learning_rate": 9.835923959618909e-05, + "loss": 0.0305, + "step": 76660 + }, + { + "epoch": 0.245376, + "grad_norm": 0.07202051052605302, + "learning_rate": 9.835838807756117e-05, + "loss": 0.0313, + "step": 76680 + }, + { + "epoch": 0.24544, + "grad_norm": 0.057647582387074243, + "learning_rate": 9.835753634171867e-05, + "loss": 0.0305, + "step": 76700 + }, + { + "epoch": 0.245504, + "grad_norm": 0.07739491268279507, + "learning_rate": 9.835668438866539e-05, + "loss": 0.0294, + "step": 76720 + }, + { + "epoch": 0.245568, + "grad_norm": 0.055959057245121076, + "learning_rate": 9.835583221840517e-05, + "loss": 0.0311, + "step": 76740 + }, + { + "epoch": 0.245632, + "grad_norm": 0.07899574307674032, + "learning_rate": 9.835497983094183e-05, + "loss": 0.0299, + "step": 76760 + }, + { + "epoch": 0.245696, + "grad_norm": 0.057799453819936876, + "learning_rate": 9.835412722627921e-05, + "loss": 0.0321, + "step": 76780 + }, + { + "epoch": 0.24576, + "grad_norm": 0.11643829837244915, + "learning_rate": 9.835327440442115e-05, + "loss": 0.0333, + "step": 76800 + }, + { + "epoch": 0.245824, + "grad_norm": 0.08624893771687397, + "learning_rate": 9.835242136537145e-05, + "loss": 0.0328, + "step": 76820 + }, + { + "epoch": 0.245888, + "grad_norm": 0.04596068917249059, + "learning_rate": 9.835156810913396e-05, + "loss": 0.0306, + "step": 76840 + }, + { + "epoch": 0.245952, + "grad_norm": 0.07869583556877166, + "learning_rate": 9.835071463571251e-05, + "loss": 0.0312, + "step": 76860 + }, + { + "epoch": 0.246016, + "grad_norm": 0.13684626957591095, + "learning_rate": 9.834986094511094e-05, + "loss": 0.0303, + "step": 76880 + }, + { + "epoch": 0.24608, + "grad_norm": 0.1334043304577334, + "learning_rate": 9.834900703733307e-05, + "loss": 0.0313, + "step": 76900 + }, + { + "epoch": 0.246144, + "grad_norm": 0.06703308575094528, + "learning_rate": 9.834815291238275e-05, + "loss": 0.0311, + "step": 76920 + }, + { + "epoch": 0.246208, + "grad_norm": 0.06881031254180857, + "learning_rate": 9.834729857026381e-05, + "loss": 0.0273, + "step": 76940 + }, + { + "epoch": 0.246272, + "grad_norm": 0.04849289519230238, + "learning_rate": 9.834644401098008e-05, + "loss": 0.0304, + "step": 76960 + }, + { + "epoch": 0.246336, + "grad_norm": 0.05459818641608831, + "learning_rate": 9.834558923453541e-05, + "loss": 0.0285, + "step": 76980 + }, + { + "epoch": 0.2464, + "grad_norm": 0.11798483005671742, + "learning_rate": 9.834473424093364e-05, + "loss": 0.0297, + "step": 77000 + }, + { + "epoch": 0.246464, + "grad_norm": 0.12591111899375504, + "learning_rate": 9.83438790301786e-05, + "loss": 0.0305, + "step": 77020 + }, + { + "epoch": 0.246528, + "grad_norm": 0.10548308955584164, + "learning_rate": 9.834302360227414e-05, + "loss": 0.0315, + "step": 77040 + }, + { + "epoch": 0.246592, + "grad_norm": 0.060514481728976006, + "learning_rate": 9.834216795722411e-05, + "loss": 0.0331, + "step": 77060 + }, + { + "epoch": 0.246656, + "grad_norm": 0.09798948715073799, + "learning_rate": 9.834131209503234e-05, + "loss": 0.0276, + "step": 77080 + }, + { + "epoch": 0.24672, + "grad_norm": 0.1226635075811686, + "learning_rate": 9.834045601570268e-05, + "loss": 0.0283, + "step": 77100 + }, + { + "epoch": 0.246784, + "grad_norm": 0.1133288002661007, + "learning_rate": 9.833959971923896e-05, + "loss": 0.0294, + "step": 77120 + }, + { + "epoch": 0.246848, + "grad_norm": 0.08880995574568823, + "learning_rate": 9.833874320564505e-05, + "loss": 0.0303, + "step": 77140 + }, + { + "epoch": 0.246912, + "grad_norm": 0.0680191401890256, + "learning_rate": 9.833788647492476e-05, + "loss": 0.0279, + "step": 77160 + }, + { + "epoch": 0.246976, + "grad_norm": 0.11108621594750857, + "learning_rate": 9.833702952708201e-05, + "loss": 0.0321, + "step": 77180 + }, + { + "epoch": 0.24704, + "grad_norm": 0.10269296161350197, + "learning_rate": 9.833617236212058e-05, + "loss": 0.0282, + "step": 77200 + }, + { + "epoch": 0.247104, + "grad_norm": 0.06686281183308215, + "learning_rate": 9.833531498004434e-05, + "loss": 0.03, + "step": 77220 + }, + { + "epoch": 0.247168, + "grad_norm": 0.06545199033933594, + "learning_rate": 9.833445738085716e-05, + "loss": 0.0302, + "step": 77240 + }, + { + "epoch": 0.247232, + "grad_norm": 0.1063622710446205, + "learning_rate": 9.833359956456286e-05, + "loss": 0.0288, + "step": 77260 + }, + { + "epoch": 0.247296, + "grad_norm": 0.06616559277544147, + "learning_rate": 9.833274153116532e-05, + "loss": 0.0305, + "step": 77280 + }, + { + "epoch": 0.24736, + "grad_norm": 0.04785651257103933, + "learning_rate": 9.833188328066838e-05, + "loss": 0.0305, + "step": 77300 + }, + { + "epoch": 0.247424, + "grad_norm": 0.12466527698517027, + "learning_rate": 9.83310248130759e-05, + "loss": 0.0295, + "step": 77320 + }, + { + "epoch": 0.247488, + "grad_norm": 0.07278809306280792, + "learning_rate": 9.833016612839174e-05, + "loss": 0.0284, + "step": 77340 + }, + { + "epoch": 0.247552, + "grad_norm": 0.09423818905593444, + "learning_rate": 9.832930722661975e-05, + "loss": 0.0312, + "step": 77360 + }, + { + "epoch": 0.247616, + "grad_norm": 0.05177919790010331, + "learning_rate": 9.832844810776381e-05, + "loss": 0.0291, + "step": 77380 + }, + { + "epoch": 0.24768, + "grad_norm": 0.10172456377581898, + "learning_rate": 9.832758877182774e-05, + "loss": 0.032, + "step": 77400 + }, + { + "epoch": 0.247744, + "grad_norm": 0.048264535962026144, + "learning_rate": 9.832672921881542e-05, + "loss": 0.0305, + "step": 77420 + }, + { + "epoch": 0.247808, + "grad_norm": 0.11135292699733396, + "learning_rate": 9.832586944873071e-05, + "loss": 0.0251, + "step": 77440 + }, + { + "epoch": 0.247872, + "grad_norm": 0.10873800990550316, + "learning_rate": 9.832500946157748e-05, + "loss": 0.0278, + "step": 77460 + }, + { + "epoch": 0.247936, + "grad_norm": 0.07030144709651247, + "learning_rate": 9.832414925735958e-05, + "loss": 0.0291, + "step": 77480 + }, + { + "epoch": 0.248, + "grad_norm": 0.12103864638116191, + "learning_rate": 9.832328883608088e-05, + "loss": 0.0313, + "step": 77500 + }, + { + "epoch": 0.248064, + "grad_norm": 0.05490590115872208, + "learning_rate": 9.832242819774524e-05, + "loss": 0.029, + "step": 77520 + }, + { + "epoch": 0.248128, + "grad_norm": 0.10682516899331987, + "learning_rate": 9.832156734235655e-05, + "loss": 0.0308, + "step": 77540 + }, + { + "epoch": 0.248192, + "grad_norm": 0.08565029298773781, + "learning_rate": 9.832070626991864e-05, + "loss": 0.0321, + "step": 77560 + }, + { + "epoch": 0.248256, + "grad_norm": 0.06830226989994787, + "learning_rate": 9.83198449804354e-05, + "loss": 0.0303, + "step": 77580 + }, + { + "epoch": 0.24832, + "grad_norm": 0.06400127148481424, + "learning_rate": 9.831898347391068e-05, + "loss": 0.0304, + "step": 77600 + }, + { + "epoch": 0.248384, + "grad_norm": 0.1289076541918976, + "learning_rate": 9.831812175034839e-05, + "loss": 0.0309, + "step": 77620 + }, + { + "epoch": 0.248448, + "grad_norm": 0.10268903902974988, + "learning_rate": 9.831725980975234e-05, + "loss": 0.0292, + "step": 77640 + }, + { + "epoch": 0.248512, + "grad_norm": 0.08131089624304555, + "learning_rate": 9.831639765212646e-05, + "loss": 0.0279, + "step": 77660 + }, + { + "epoch": 0.248576, + "grad_norm": 0.06727296576277478, + "learning_rate": 9.831553527747458e-05, + "loss": 0.0302, + "step": 77680 + }, + { + "epoch": 0.24864, + "grad_norm": 0.07896931175174525, + "learning_rate": 9.831467268580058e-05, + "loss": 0.0347, + "step": 77700 + }, + { + "epoch": 0.248704, + "grad_norm": 0.07543986966319662, + "learning_rate": 9.831380987710839e-05, + "loss": 0.0276, + "step": 77720 + }, + { + "epoch": 0.248768, + "grad_norm": 0.07853862265100248, + "learning_rate": 9.831294685140181e-05, + "loss": 0.0292, + "step": 77740 + }, + { + "epoch": 0.248832, + "grad_norm": 0.07548178472344115, + "learning_rate": 9.831208360868476e-05, + "loss": 0.0313, + "step": 77760 + }, + { + "epoch": 0.248896, + "grad_norm": 0.04356577268160738, + "learning_rate": 9.831122014896109e-05, + "loss": 0.0274, + "step": 77780 + }, + { + "epoch": 0.24896, + "grad_norm": 0.07696284344124553, + "learning_rate": 9.83103564722347e-05, + "loss": 0.0315, + "step": 77800 + }, + { + "epoch": 0.249024, + "grad_norm": 0.06451635893351533, + "learning_rate": 9.830949257850947e-05, + "loss": 0.0292, + "step": 77820 + }, + { + "epoch": 0.249088, + "grad_norm": 0.06483559410974471, + "learning_rate": 9.830862846778928e-05, + "loss": 0.0275, + "step": 77840 + }, + { + "epoch": 0.249152, + "grad_norm": 0.06257365637286286, + "learning_rate": 9.830776414007799e-05, + "loss": 0.0261, + "step": 77860 + }, + { + "epoch": 0.249216, + "grad_norm": 0.06652426972771643, + "learning_rate": 9.830689959537952e-05, + "loss": 0.0284, + "step": 77880 + }, + { + "epoch": 0.24928, + "grad_norm": 0.09662805890308412, + "learning_rate": 9.83060348336977e-05, + "loss": 0.0315, + "step": 77900 + }, + { + "epoch": 0.249344, + "grad_norm": 0.049093729443121015, + "learning_rate": 9.830516985503645e-05, + "loss": 0.0307, + "step": 77920 + }, + { + "epoch": 0.249408, + "grad_norm": 0.05306713163586091, + "learning_rate": 9.830430465939966e-05, + "loss": 0.032, + "step": 77940 + }, + { + "epoch": 0.249472, + "grad_norm": 0.12642777448502335, + "learning_rate": 9.830343924679121e-05, + "loss": 0.0306, + "step": 77960 + }, + { + "epoch": 0.249536, + "grad_norm": 0.07175161103336694, + "learning_rate": 9.830257361721497e-05, + "loss": 0.0323, + "step": 77980 + }, + { + "epoch": 0.2496, + "grad_norm": 0.08937313958416067, + "learning_rate": 9.830170777067485e-05, + "loss": 0.0313, + "step": 78000 + }, + { + "epoch": 0.249664, + "grad_norm": 0.10213694889388962, + "learning_rate": 9.830084170717473e-05, + "loss": 0.0321, + "step": 78020 + }, + { + "epoch": 0.249728, + "grad_norm": 0.0891454155729414, + "learning_rate": 9.829997542671851e-05, + "loss": 0.0279, + "step": 78040 + }, + { + "epoch": 0.249792, + "grad_norm": 0.08296183027519251, + "learning_rate": 9.829910892931007e-05, + "loss": 0.0303, + "step": 78060 + }, + { + "epoch": 0.249856, + "grad_norm": 0.06329615025836577, + "learning_rate": 9.82982422149533e-05, + "loss": 0.033, + "step": 78080 + }, + { + "epoch": 0.24992, + "grad_norm": 0.050042055448042175, + "learning_rate": 9.82973752836521e-05, + "loss": 0.0317, + "step": 78100 + }, + { + "epoch": 0.249984, + "grad_norm": 0.09103175259121889, + "learning_rate": 9.829650813541036e-05, + "loss": 0.0291, + "step": 78120 + }, + { + "epoch": 0.250048, + "grad_norm": 0.11344347331378879, + "learning_rate": 9.829564077023197e-05, + "loss": 0.0313, + "step": 78140 + }, + { + "epoch": 0.250112, + "grad_norm": 0.05607198401695192, + "learning_rate": 9.829477318812083e-05, + "loss": 0.0305, + "step": 78160 + }, + { + "epoch": 0.250176, + "grad_norm": 0.057861448233872556, + "learning_rate": 9.829390538908086e-05, + "loss": 0.0294, + "step": 78180 + }, + { + "epoch": 0.25024, + "grad_norm": 0.056910754285138004, + "learning_rate": 9.829303737311593e-05, + "loss": 0.0364, + "step": 78200 + }, + { + "epoch": 0.250304, + "grad_norm": 0.05110854915484866, + "learning_rate": 9.829216914022994e-05, + "loss": 0.0303, + "step": 78220 + }, + { + "epoch": 0.250368, + "grad_norm": 0.06340253366776188, + "learning_rate": 9.82913006904268e-05, + "loss": 0.0303, + "step": 78240 + }, + { + "epoch": 0.250432, + "grad_norm": 0.07060800008782184, + "learning_rate": 9.829043202371041e-05, + "loss": 0.0302, + "step": 78260 + }, + { + "epoch": 0.250496, + "grad_norm": 0.058812205786097645, + "learning_rate": 9.828956314008466e-05, + "loss": 0.0317, + "step": 78280 + }, + { + "epoch": 0.25056, + "grad_norm": 0.05669931385874617, + "learning_rate": 9.828869403955348e-05, + "loss": 0.0301, + "step": 78300 + }, + { + "epoch": 0.250624, + "grad_norm": 0.24026646329345613, + "learning_rate": 9.828782472212073e-05, + "loss": 0.0302, + "step": 78320 + }, + { + "epoch": 0.250688, + "grad_norm": 0.04951313476192179, + "learning_rate": 9.828695518779037e-05, + "loss": 0.0279, + "step": 78340 + }, + { + "epoch": 0.250752, + "grad_norm": 0.0805412200847954, + "learning_rate": 9.828608543656626e-05, + "loss": 0.0285, + "step": 78360 + }, + { + "epoch": 0.250816, + "grad_norm": 0.08549113639175138, + "learning_rate": 9.828521546845232e-05, + "loss": 0.031, + "step": 78380 + }, + { + "epoch": 0.25088, + "grad_norm": 0.08029612634340495, + "learning_rate": 9.828434528345246e-05, + "loss": 0.0297, + "step": 78400 + }, + { + "epoch": 0.250944, + "grad_norm": 0.06068178567996576, + "learning_rate": 9.82834748815706e-05, + "loss": 0.0314, + "step": 78420 + }, + { + "epoch": 0.251008, + "grad_norm": 0.06426209848275896, + "learning_rate": 9.828260426281062e-05, + "loss": 0.0277, + "step": 78440 + }, + { + "epoch": 0.251072, + "grad_norm": 0.050895077060587296, + "learning_rate": 9.828173342717648e-05, + "loss": 0.0271, + "step": 78460 + }, + { + "epoch": 0.251136, + "grad_norm": 0.13159640576956252, + "learning_rate": 9.828086237467204e-05, + "loss": 0.0314, + "step": 78480 + }, + { + "epoch": 0.2512, + "grad_norm": 0.06978428213278695, + "learning_rate": 9.827999110530123e-05, + "loss": 0.033, + "step": 78500 + }, + { + "epoch": 0.251264, + "grad_norm": 0.06668744407271042, + "learning_rate": 9.827911961906799e-05, + "loss": 0.0319, + "step": 78520 + }, + { + "epoch": 0.251328, + "grad_norm": 0.07810984220809664, + "learning_rate": 9.82782479159762e-05, + "loss": 0.0304, + "step": 78540 + }, + { + "epoch": 0.251392, + "grad_norm": 0.07176983446722066, + "learning_rate": 9.827737599602977e-05, + "loss": 0.0309, + "step": 78560 + }, + { + "epoch": 0.251456, + "grad_norm": 0.13246234366306997, + "learning_rate": 9.827650385923266e-05, + "loss": 0.0342, + "step": 78580 + }, + { + "epoch": 0.25152, + "grad_norm": 0.06782392818550931, + "learning_rate": 9.827563150558874e-05, + "loss": 0.0286, + "step": 78600 + }, + { + "epoch": 0.251584, + "grad_norm": 0.15766432061444388, + "learning_rate": 9.827475893510197e-05, + "loss": 0.0272, + "step": 78620 + }, + { + "epoch": 0.251648, + "grad_norm": 0.06517815181701653, + "learning_rate": 9.827388614777622e-05, + "loss": 0.0284, + "step": 78640 + }, + { + "epoch": 0.251712, + "grad_norm": 0.04449363377826685, + "learning_rate": 9.827301314361547e-05, + "loss": 0.0281, + "step": 78660 + }, + { + "epoch": 0.251776, + "grad_norm": 0.05972258737737138, + "learning_rate": 9.82721399226236e-05, + "loss": 0.0321, + "step": 78680 + }, + { + "epoch": 0.25184, + "grad_norm": 0.04642535453077361, + "learning_rate": 9.827126648480454e-05, + "loss": 0.0302, + "step": 78700 + }, + { + "epoch": 0.251904, + "grad_norm": 0.07230552143316371, + "learning_rate": 9.827039283016222e-05, + "loss": 0.03, + "step": 78720 + }, + { + "epoch": 0.251968, + "grad_norm": 0.04400990805093012, + "learning_rate": 9.826951895870056e-05, + "loss": 0.0321, + "step": 78740 + }, + { + "epoch": 0.252032, + "grad_norm": 0.05823823757675434, + "learning_rate": 9.826864487042349e-05, + "loss": 0.0322, + "step": 78760 + }, + { + "epoch": 0.252096, + "grad_norm": 0.13236327909933257, + "learning_rate": 9.826777056533491e-05, + "loss": 0.0336, + "step": 78780 + }, + { + "epoch": 0.25216, + "grad_norm": 0.04819873422211765, + "learning_rate": 9.82668960434388e-05, + "loss": 0.031, + "step": 78800 + }, + { + "epoch": 0.252224, + "grad_norm": 0.08645905562891312, + "learning_rate": 9.826602130473903e-05, + "loss": 0.0308, + "step": 78820 + }, + { + "epoch": 0.252288, + "grad_norm": 0.07233498534487104, + "learning_rate": 9.826514634923958e-05, + "loss": 0.0298, + "step": 78840 + }, + { + "epoch": 0.252352, + "grad_norm": 0.0744801491601935, + "learning_rate": 9.826427117694435e-05, + "loss": 0.0274, + "step": 78860 + }, + { + "epoch": 0.252416, + "grad_norm": 0.10590792200145402, + "learning_rate": 9.826339578785726e-05, + "loss": 0.0317, + "step": 78880 + }, + { + "epoch": 0.25248, + "grad_norm": 0.06164106718259327, + "learning_rate": 9.826252018198228e-05, + "loss": 0.0321, + "step": 78900 + }, + { + "epoch": 0.252544, + "grad_norm": 0.06458212225610689, + "learning_rate": 9.826164435932332e-05, + "loss": 0.0342, + "step": 78920 + }, + { + "epoch": 0.252608, + "grad_norm": 0.05582948480120334, + "learning_rate": 9.826076831988432e-05, + "loss": 0.0294, + "step": 78940 + }, + { + "epoch": 0.252672, + "grad_norm": 0.10019100228415243, + "learning_rate": 9.825989206366919e-05, + "loss": 0.0278, + "step": 78960 + }, + { + "epoch": 0.252736, + "grad_norm": 0.06861413856339732, + "learning_rate": 9.82590155906819e-05, + "loss": 0.0285, + "step": 78980 + }, + { + "epoch": 0.2528, + "grad_norm": 0.07821143363812459, + "learning_rate": 9.825813890092637e-05, + "loss": 0.0327, + "step": 79000 + }, + { + "epoch": 0.252864, + "grad_norm": 0.058331578974219166, + "learning_rate": 9.825726199440656e-05, + "loss": 0.0298, + "step": 79020 + }, + { + "epoch": 0.252928, + "grad_norm": 0.09366912172003448, + "learning_rate": 9.825638487112639e-05, + "loss": 0.0266, + "step": 79040 + }, + { + "epoch": 0.252992, + "grad_norm": 0.06255353050511842, + "learning_rate": 9.825550753108979e-05, + "loss": 0.0291, + "step": 79060 + }, + { + "epoch": 0.253056, + "grad_norm": 0.0688141245874151, + "learning_rate": 9.825462997430071e-05, + "loss": 0.0339, + "step": 79080 + }, + { + "epoch": 0.25312, + "grad_norm": 0.05495070999806289, + "learning_rate": 9.82537522007631e-05, + "loss": 0.0263, + "step": 79100 + }, + { + "epoch": 0.253184, + "grad_norm": 0.08118166147858945, + "learning_rate": 9.825287421048088e-05, + "loss": 0.0285, + "step": 79120 + }, + { + "epoch": 0.253248, + "grad_norm": 0.06552850220555453, + "learning_rate": 9.825199600345802e-05, + "loss": 0.0299, + "step": 79140 + }, + { + "epoch": 0.253312, + "grad_norm": 0.13817780316877504, + "learning_rate": 9.825111757969846e-05, + "loss": 0.0314, + "step": 79160 + }, + { + "epoch": 0.253376, + "grad_norm": 0.07176870823192916, + "learning_rate": 9.825023893920614e-05, + "loss": 0.0286, + "step": 79180 + }, + { + "epoch": 0.25344, + "grad_norm": 0.056512643913864444, + "learning_rate": 9.8249360081985e-05, + "loss": 0.0288, + "step": 79200 + }, + { + "epoch": 0.253504, + "grad_norm": 0.06100558479497491, + "learning_rate": 9.824848100803901e-05, + "loss": 0.0325, + "step": 79220 + }, + { + "epoch": 0.253568, + "grad_norm": 0.05868749562230104, + "learning_rate": 9.824760171737208e-05, + "loss": 0.0302, + "step": 79240 + }, + { + "epoch": 0.253632, + "grad_norm": 0.05669320044494591, + "learning_rate": 9.82467222099882e-05, + "loss": 0.0266, + "step": 79260 + }, + { + "epoch": 0.253696, + "grad_norm": 0.056992850294694815, + "learning_rate": 9.82458424858913e-05, + "loss": 0.029, + "step": 79280 + }, + { + "epoch": 0.25376, + "grad_norm": 0.05279529883271108, + "learning_rate": 9.824496254508533e-05, + "loss": 0.0326, + "step": 79300 + }, + { + "epoch": 0.253824, + "grad_norm": 0.0681924422686678, + "learning_rate": 9.824408238757426e-05, + "loss": 0.0318, + "step": 79320 + }, + { + "epoch": 0.253888, + "grad_norm": 0.11506294693731017, + "learning_rate": 9.824320201336204e-05, + "loss": 0.0312, + "step": 79340 + }, + { + "epoch": 0.253952, + "grad_norm": 0.12897682111765474, + "learning_rate": 9.824232142245259e-05, + "loss": 0.0284, + "step": 79360 + }, + { + "epoch": 0.254016, + "grad_norm": 0.06372258651627359, + "learning_rate": 9.82414406148499e-05, + "loss": 0.0303, + "step": 79380 + }, + { + "epoch": 0.25408, + "grad_norm": 0.06616387326001123, + "learning_rate": 9.824055959055791e-05, + "loss": 0.0329, + "step": 79400 + }, + { + "epoch": 0.254144, + "grad_norm": 0.07605244762096397, + "learning_rate": 9.823967834958059e-05, + "loss": 0.0306, + "step": 79420 + }, + { + "epoch": 0.254208, + "grad_norm": 0.058875584959462234, + "learning_rate": 9.82387968919219e-05, + "loss": 0.0336, + "step": 79440 + }, + { + "epoch": 0.254272, + "grad_norm": 0.07709415911169192, + "learning_rate": 9.823791521758579e-05, + "loss": 0.0306, + "step": 79460 + }, + { + "epoch": 0.254336, + "grad_norm": 0.13684115385125947, + "learning_rate": 9.823703332657621e-05, + "loss": 0.0315, + "step": 79480 + }, + { + "epoch": 0.2544, + "grad_norm": 0.10821203828357712, + "learning_rate": 9.823615121889716e-05, + "loss": 0.0296, + "step": 79500 + }, + { + "epoch": 0.254464, + "grad_norm": 0.08244829791429087, + "learning_rate": 9.823526889455255e-05, + "loss": 0.0281, + "step": 79520 + }, + { + "epoch": 0.254528, + "grad_norm": 0.07710309520883259, + "learning_rate": 9.823438635354639e-05, + "loss": 0.0303, + "step": 79540 + }, + { + "epoch": 0.254592, + "grad_norm": 0.05271271298933364, + "learning_rate": 9.82335035958826e-05, + "loss": 0.0315, + "step": 79560 + }, + { + "epoch": 0.254656, + "grad_norm": 0.05915432849330216, + "learning_rate": 9.823262062156517e-05, + "loss": 0.0322, + "step": 79580 + }, + { + "epoch": 0.25472, + "grad_norm": 0.07801281949559437, + "learning_rate": 9.823173743059809e-05, + "loss": 0.0346, + "step": 79600 + }, + { + "epoch": 0.254784, + "grad_norm": 0.07361694461158655, + "learning_rate": 9.823085402298528e-05, + "loss": 0.032, + "step": 79620 + }, + { + "epoch": 0.254848, + "grad_norm": 0.0800530107945764, + "learning_rate": 9.822997039873073e-05, + "loss": 0.0332, + "step": 79640 + }, + { + "epoch": 0.254912, + "grad_norm": 0.05549395167229514, + "learning_rate": 9.822908655783841e-05, + "loss": 0.0284, + "step": 79660 + }, + { + "epoch": 0.254976, + "grad_norm": 0.054152509955608365, + "learning_rate": 9.822820250031229e-05, + "loss": 0.0314, + "step": 79680 + }, + { + "epoch": 0.25504, + "grad_norm": 0.05952221985471428, + "learning_rate": 9.822731822615634e-05, + "loss": 0.0287, + "step": 79700 + }, + { + "epoch": 0.255104, + "grad_norm": 0.059836099855398575, + "learning_rate": 9.822643373537454e-05, + "loss": 0.0295, + "step": 79720 + }, + { + "epoch": 0.255168, + "grad_norm": 0.06715782651072263, + "learning_rate": 9.822554902797085e-05, + "loss": 0.0302, + "step": 79740 + }, + { + "epoch": 0.255232, + "grad_norm": 0.07561743503481089, + "learning_rate": 9.822466410394926e-05, + "loss": 0.0292, + "step": 79760 + }, + { + "epoch": 0.255296, + "grad_norm": 0.05845715434568501, + "learning_rate": 9.82237789633137e-05, + "loss": 0.0313, + "step": 79780 + }, + { + "epoch": 0.25536, + "grad_norm": 0.13798946482568997, + "learning_rate": 9.82228936060682e-05, + "loss": 0.0304, + "step": 79800 + }, + { + "epoch": 0.255424, + "grad_norm": 0.07440087246092897, + "learning_rate": 9.822200803221673e-05, + "loss": 0.0316, + "step": 79820 + }, + { + "epoch": 0.255488, + "grad_norm": 0.058707283427543944, + "learning_rate": 9.822112224176323e-05, + "loss": 0.0311, + "step": 79840 + }, + { + "epoch": 0.255552, + "grad_norm": 0.06006530261997314, + "learning_rate": 9.822023623471172e-05, + "loss": 0.0299, + "step": 79860 + }, + { + "epoch": 0.255616, + "grad_norm": 0.10956573571804336, + "learning_rate": 9.821935001106615e-05, + "loss": 0.0293, + "step": 79880 + }, + { + "epoch": 0.25568, + "grad_norm": 0.06294482674982198, + "learning_rate": 9.821846357083051e-05, + "loss": 0.034, + "step": 79900 + }, + { + "epoch": 0.255744, + "grad_norm": 0.09169174510257477, + "learning_rate": 9.821757691400879e-05, + "loss": 0.0305, + "step": 79920 + }, + { + "epoch": 0.255808, + "grad_norm": 0.0653400476858161, + "learning_rate": 9.821669004060497e-05, + "loss": 0.0323, + "step": 79940 + }, + { + "epoch": 0.255872, + "grad_norm": 0.057565416350614775, + "learning_rate": 9.821580295062303e-05, + "loss": 0.0337, + "step": 79960 + }, + { + "epoch": 0.255936, + "grad_norm": 0.04684698250802921, + "learning_rate": 9.821491564406696e-05, + "loss": 0.0275, + "step": 79980 + }, + { + "epoch": 0.256, + "grad_norm": 0.051481881448666805, + "learning_rate": 9.821402812094073e-05, + "loss": 0.0301, + "step": 80000 + }, + { + "epoch": 0.256064, + "grad_norm": 0.10348352407832796, + "learning_rate": 9.821314038124836e-05, + "loss": 0.0303, + "step": 80020 + }, + { + "epoch": 0.256128, + "grad_norm": 0.07998212865699095, + "learning_rate": 9.821225242499379e-05, + "loss": 0.0343, + "step": 80040 + }, + { + "epoch": 0.256192, + "grad_norm": 0.05699575727399758, + "learning_rate": 9.821136425218104e-05, + "loss": 0.0317, + "step": 80060 + }, + { + "epoch": 0.256256, + "grad_norm": 0.09179057759407547, + "learning_rate": 9.821047586281412e-05, + "loss": 0.0298, + "step": 80080 + }, + { + "epoch": 0.25632, + "grad_norm": 0.0404718997503838, + "learning_rate": 9.820958725689695e-05, + "loss": 0.0292, + "step": 80100 + }, + { + "epoch": 0.256384, + "grad_norm": 0.05867817751504825, + "learning_rate": 9.820869843443358e-05, + "loss": 0.0266, + "step": 80120 + }, + { + "epoch": 0.256448, + "grad_norm": 0.05012720250689934, + "learning_rate": 9.820780939542799e-05, + "loss": 0.0293, + "step": 80140 + }, + { + "epoch": 0.256512, + "grad_norm": 0.050060013154556236, + "learning_rate": 9.820692013988418e-05, + "loss": 0.032, + "step": 80160 + }, + { + "epoch": 0.256576, + "grad_norm": 0.09957734596690702, + "learning_rate": 9.820603066780613e-05, + "loss": 0.028, + "step": 80180 + }, + { + "epoch": 0.25664, + "grad_norm": 0.06763112401289502, + "learning_rate": 9.820514097919784e-05, + "loss": 0.0348, + "step": 80200 + }, + { + "epoch": 0.256704, + "grad_norm": 0.06737574112005623, + "learning_rate": 9.82042510740633e-05, + "loss": 0.0332, + "step": 80220 + }, + { + "epoch": 0.256768, + "grad_norm": 0.06701179871594416, + "learning_rate": 9.820336095240652e-05, + "loss": 0.0277, + "step": 80240 + }, + { + "epoch": 0.256832, + "grad_norm": 0.14272412703173623, + "learning_rate": 9.82024706142315e-05, + "loss": 0.0295, + "step": 80260 + }, + { + "epoch": 0.256896, + "grad_norm": 0.06712952188686684, + "learning_rate": 9.820158005954223e-05, + "loss": 0.029, + "step": 80280 + }, + { + "epoch": 0.25696, + "grad_norm": 0.06127182934757832, + "learning_rate": 9.82006892883427e-05, + "loss": 0.0357, + "step": 80300 + }, + { + "epoch": 0.257024, + "grad_norm": 0.07637282180157111, + "learning_rate": 9.819979830063693e-05, + "loss": 0.0306, + "step": 80320 + }, + { + "epoch": 0.257088, + "grad_norm": 0.06050276703049385, + "learning_rate": 9.81989070964289e-05, + "loss": 0.0308, + "step": 80340 + }, + { + "epoch": 0.257152, + "grad_norm": 0.16482091044791472, + "learning_rate": 9.819801567572264e-05, + "loss": 0.0319, + "step": 80360 + }, + { + "epoch": 0.257216, + "grad_norm": 0.11106276428262839, + "learning_rate": 9.819712403852215e-05, + "loss": 0.0314, + "step": 80380 + }, + { + "epoch": 0.25728, + "grad_norm": 0.07282805275781835, + "learning_rate": 9.819623218483141e-05, + "loss": 0.034, + "step": 80400 + }, + { + "epoch": 0.257344, + "grad_norm": 0.05225038639125422, + "learning_rate": 9.819534011465445e-05, + "loss": 0.0276, + "step": 80420 + }, + { + "epoch": 0.257408, + "grad_norm": 0.1945117131428148, + "learning_rate": 9.819444782799527e-05, + "loss": 0.0337, + "step": 80440 + }, + { + "epoch": 0.257472, + "grad_norm": 0.11048494777938277, + "learning_rate": 9.819355532485789e-05, + "loss": 0.0332, + "step": 80460 + }, + { + "epoch": 0.257536, + "grad_norm": 0.07946331243942782, + "learning_rate": 9.819266260524629e-05, + "loss": 0.0321, + "step": 80480 + }, + { + "epoch": 0.2576, + "grad_norm": 0.04945339282303996, + "learning_rate": 9.81917696691645e-05, + "loss": 0.03, + "step": 80500 + }, + { + "epoch": 0.257664, + "grad_norm": 0.05342191180754064, + "learning_rate": 9.819087651661654e-05, + "loss": 0.0345, + "step": 80520 + }, + { + "epoch": 0.257728, + "grad_norm": 0.0857724606332447, + "learning_rate": 9.81899831476064e-05, + "loss": 0.0301, + "step": 80540 + }, + { + "epoch": 0.257792, + "grad_norm": 0.04679502073647181, + "learning_rate": 9.81890895621381e-05, + "loss": 0.0281, + "step": 80560 + }, + { + "epoch": 0.257856, + "grad_norm": 0.05525005697882699, + "learning_rate": 9.818819576021566e-05, + "loss": 0.0298, + "step": 80580 + }, + { + "epoch": 0.25792, + "grad_norm": 0.0663905514849343, + "learning_rate": 9.818730174184307e-05, + "loss": 0.0329, + "step": 80600 + }, + { + "epoch": 0.257984, + "grad_norm": 0.12153188529194182, + "learning_rate": 9.818640750702439e-05, + "loss": 0.0323, + "step": 80620 + }, + { + "epoch": 0.258048, + "grad_norm": 0.0863666060499636, + "learning_rate": 9.818551305576361e-05, + "loss": 0.0292, + "step": 80640 + }, + { + "epoch": 0.258112, + "grad_norm": 0.06580074043203675, + "learning_rate": 9.818461838806475e-05, + "loss": 0.0274, + "step": 80660 + }, + { + "epoch": 0.258176, + "grad_norm": 0.08597577257820976, + "learning_rate": 9.818372350393183e-05, + "loss": 0.0286, + "step": 80680 + }, + { + "epoch": 0.25824, + "grad_norm": 0.07692577917454702, + "learning_rate": 9.818282840336887e-05, + "loss": 0.0299, + "step": 80700 + }, + { + "epoch": 0.258304, + "grad_norm": 0.051625804774256365, + "learning_rate": 9.818193308637989e-05, + "loss": 0.0307, + "step": 80720 + }, + { + "epoch": 0.258368, + "grad_norm": 0.0941947497599387, + "learning_rate": 9.81810375529689e-05, + "loss": 0.0331, + "step": 80740 + }, + { + "epoch": 0.258432, + "grad_norm": 0.062537354762688, + "learning_rate": 9.818014180313995e-05, + "loss": 0.0287, + "step": 80760 + }, + { + "epoch": 0.258496, + "grad_norm": 0.10654082457070405, + "learning_rate": 9.817924583689703e-05, + "loss": 0.0321, + "step": 80780 + }, + { + "epoch": 0.25856, + "grad_norm": 0.06335092396578564, + "learning_rate": 9.817834965424418e-05, + "loss": 0.0311, + "step": 80800 + }, + { + "epoch": 0.258624, + "grad_norm": 0.0510586329974211, + "learning_rate": 9.817745325518544e-05, + "loss": 0.0304, + "step": 80820 + }, + { + "epoch": 0.258688, + "grad_norm": 0.054574368054886954, + "learning_rate": 9.817655663972482e-05, + "loss": 0.0294, + "step": 80840 + }, + { + "epoch": 0.258752, + "grad_norm": 0.053816337085745486, + "learning_rate": 9.817565980786634e-05, + "loss": 0.0287, + "step": 80860 + }, + { + "epoch": 0.258816, + "grad_norm": 0.07031536363302589, + "learning_rate": 9.817476275961405e-05, + "loss": 0.0316, + "step": 80880 + }, + { + "epoch": 0.25888, + "grad_norm": 0.056513539865396076, + "learning_rate": 9.817386549497198e-05, + "loss": 0.0316, + "step": 80900 + }, + { + "epoch": 0.258944, + "grad_norm": 0.08594161849582678, + "learning_rate": 9.817296801394413e-05, + "loss": 0.0277, + "step": 80920 + }, + { + "epoch": 0.259008, + "grad_norm": 0.06124775758700676, + "learning_rate": 9.817207031653457e-05, + "loss": 0.0298, + "step": 80940 + }, + { + "epoch": 0.259072, + "grad_norm": 0.09469645615054358, + "learning_rate": 9.817117240274729e-05, + "loss": 0.0277, + "step": 80960 + }, + { + "epoch": 0.259136, + "grad_norm": 0.05574095694641624, + "learning_rate": 9.817027427258636e-05, + "loss": 0.0298, + "step": 80980 + }, + { + "epoch": 0.2592, + "grad_norm": 0.059973070323330366, + "learning_rate": 9.816937592605579e-05, + "loss": 0.0281, + "step": 81000 + }, + { + "epoch": 0.259264, + "grad_norm": 0.10666801523911035, + "learning_rate": 9.816847736315963e-05, + "loss": 0.029, + "step": 81020 + }, + { + "epoch": 0.259328, + "grad_norm": 0.06975186943306917, + "learning_rate": 9.816757858390191e-05, + "loss": 0.0316, + "step": 81040 + }, + { + "epoch": 0.259392, + "grad_norm": 0.08111864832528527, + "learning_rate": 9.816667958828667e-05, + "loss": 0.0331, + "step": 81060 + }, + { + "epoch": 0.259456, + "grad_norm": 0.04915575394458525, + "learning_rate": 9.816578037631795e-05, + "loss": 0.0308, + "step": 81080 + }, + { + "epoch": 0.25952, + "grad_norm": 0.0731447427936237, + "learning_rate": 9.816488094799979e-05, + "loss": 0.0289, + "step": 81100 + }, + { + "epoch": 0.259584, + "grad_norm": 0.08194905614355694, + "learning_rate": 9.81639813033362e-05, + "loss": 0.0286, + "step": 81120 + }, + { + "epoch": 0.259648, + "grad_norm": 0.09500970904970563, + "learning_rate": 9.816308144233127e-05, + "loss": 0.0286, + "step": 81140 + }, + { + "epoch": 0.259712, + "grad_norm": 0.06518560360686543, + "learning_rate": 9.8162181364989e-05, + "loss": 0.0285, + "step": 81160 + }, + { + "epoch": 0.259776, + "grad_norm": 0.09664734863207758, + "learning_rate": 9.816128107131347e-05, + "loss": 0.029, + "step": 81180 + }, + { + "epoch": 0.25984, + "grad_norm": 0.09088977944237343, + "learning_rate": 9.81603805613087e-05, + "loss": 0.0286, + "step": 81200 + }, + { + "epoch": 0.259904, + "grad_norm": 0.11827952083425156, + "learning_rate": 9.815947983497872e-05, + "loss": 0.0305, + "step": 81220 + }, + { + "epoch": 0.259968, + "grad_norm": 0.06716928447437376, + "learning_rate": 9.815857889232762e-05, + "loss": 0.0325, + "step": 81240 + }, + { + "epoch": 0.260032, + "grad_norm": 0.15848059358519304, + "learning_rate": 9.815767773335943e-05, + "loss": 0.0311, + "step": 81260 + }, + { + "epoch": 0.260096, + "grad_norm": 0.08449594589566033, + "learning_rate": 9.815677635807817e-05, + "loss": 0.0305, + "step": 81280 + }, + { + "epoch": 0.26016, + "grad_norm": 0.08708967121740034, + "learning_rate": 9.815587476648791e-05, + "loss": 0.0298, + "step": 81300 + }, + { + "epoch": 0.260224, + "grad_norm": 0.11327293866982925, + "learning_rate": 9.81549729585927e-05, + "loss": 0.0318, + "step": 81320 + }, + { + "epoch": 0.260288, + "grad_norm": 0.07392135189078142, + "learning_rate": 9.815407093439662e-05, + "loss": 0.0297, + "step": 81340 + }, + { + "epoch": 0.260352, + "grad_norm": 0.06756293763526075, + "learning_rate": 9.815316869390365e-05, + "loss": 0.0299, + "step": 81360 + }, + { + "epoch": 0.260416, + "grad_norm": 0.07012657181568799, + "learning_rate": 9.81522662371179e-05, + "loss": 0.0277, + "step": 81380 + }, + { + "epoch": 0.26048, + "grad_norm": 0.08621584889473753, + "learning_rate": 9.81513635640434e-05, + "loss": 0.03, + "step": 81400 + }, + { + "epoch": 0.260544, + "grad_norm": 0.05720849290507598, + "learning_rate": 9.815046067468423e-05, + "loss": 0.0306, + "step": 81420 + }, + { + "epoch": 0.260608, + "grad_norm": 0.10446105306100466, + "learning_rate": 9.814955756904441e-05, + "loss": 0.0301, + "step": 81440 + }, + { + "epoch": 0.260672, + "grad_norm": 0.2408105831405315, + "learning_rate": 9.814865424712803e-05, + "loss": 0.0261, + "step": 81460 + }, + { + "epoch": 0.260736, + "grad_norm": 0.09839830869178157, + "learning_rate": 9.814775070893913e-05, + "loss": 0.0286, + "step": 81480 + }, + { + "epoch": 0.2608, + "grad_norm": 0.0697326903528802, + "learning_rate": 9.814684695448175e-05, + "loss": 0.0286, + "step": 81500 + }, + { + "epoch": 0.260864, + "grad_norm": 0.09794311869679777, + "learning_rate": 9.814594298375998e-05, + "loss": 0.0276, + "step": 81520 + }, + { + "epoch": 0.260928, + "grad_norm": 0.09780757500744494, + "learning_rate": 9.814503879677787e-05, + "loss": 0.0305, + "step": 81540 + }, + { + "epoch": 0.260992, + "grad_norm": 0.1197732211908016, + "learning_rate": 9.81441343935395e-05, + "loss": 0.028, + "step": 81560 + }, + { + "epoch": 0.261056, + "grad_norm": 0.081949988141081, + "learning_rate": 9.814322977404891e-05, + "loss": 0.0313, + "step": 81580 + }, + { + "epoch": 0.26112, + "grad_norm": 0.06800502240732292, + "learning_rate": 9.814232493831014e-05, + "loss": 0.0304, + "step": 81600 + }, + { + "epoch": 0.261184, + "grad_norm": 0.06035202843635463, + "learning_rate": 9.814141988632731e-05, + "loss": 0.0312, + "step": 81620 + }, + { + "epoch": 0.261248, + "grad_norm": 0.060216976246327776, + "learning_rate": 9.814051461810444e-05, + "loss": 0.0312, + "step": 81640 + }, + { + "epoch": 0.261312, + "grad_norm": 0.06026072292505581, + "learning_rate": 9.813960913364561e-05, + "loss": 0.0342, + "step": 81660 + }, + { + "epoch": 0.261376, + "grad_norm": 0.06469011581823055, + "learning_rate": 9.81387034329549e-05, + "loss": 0.0336, + "step": 81680 + }, + { + "epoch": 0.26144, + "grad_norm": 0.05568897461195142, + "learning_rate": 9.813779751603638e-05, + "loss": 0.0323, + "step": 81700 + }, + { + "epoch": 0.261504, + "grad_norm": 0.09271259854842587, + "learning_rate": 9.813689138289409e-05, + "loss": 0.0292, + "step": 81720 + }, + { + "epoch": 0.261568, + "grad_norm": 0.04448803758439845, + "learning_rate": 9.813598503353213e-05, + "loss": 0.0296, + "step": 81740 + }, + { + "epoch": 0.261632, + "grad_norm": 0.0700163238975115, + "learning_rate": 9.813507846795455e-05, + "loss": 0.0313, + "step": 81760 + }, + { + "epoch": 0.261696, + "grad_norm": 0.10372909526588295, + "learning_rate": 9.813417168616543e-05, + "loss": 0.0288, + "step": 81780 + }, + { + "epoch": 0.26176, + "grad_norm": 0.0689318972573089, + "learning_rate": 9.813326468816886e-05, + "loss": 0.0283, + "step": 81800 + }, + { + "epoch": 0.261824, + "grad_norm": 0.05376779469467516, + "learning_rate": 9.813235747396887e-05, + "loss": 0.0293, + "step": 81820 + }, + { + "epoch": 0.261888, + "grad_norm": 0.04745857654412596, + "learning_rate": 9.813145004356959e-05, + "loss": 0.0289, + "step": 81840 + }, + { + "epoch": 0.261952, + "grad_norm": 0.10358046582279584, + "learning_rate": 9.813054239697506e-05, + "loss": 0.0279, + "step": 81860 + }, + { + "epoch": 0.262016, + "grad_norm": 0.11327081371182941, + "learning_rate": 9.812963453418935e-05, + "loss": 0.0314, + "step": 81880 + }, + { + "epoch": 0.26208, + "grad_norm": 0.06805392779208723, + "learning_rate": 9.812872645521659e-05, + "loss": 0.0318, + "step": 81900 + }, + { + "epoch": 0.262144, + "grad_norm": 0.08826098576115303, + "learning_rate": 9.81278181600608e-05, + "loss": 0.0298, + "step": 81920 + }, + { + "epoch": 0.262208, + "grad_norm": 0.0743733919726987, + "learning_rate": 9.812690964872607e-05, + "loss": 0.0313, + "step": 81940 + }, + { + "epoch": 0.262272, + "grad_norm": 0.08141661011693588, + "learning_rate": 9.812600092121651e-05, + "loss": 0.0267, + "step": 81960 + }, + { + "epoch": 0.262336, + "grad_norm": 0.05339682099320481, + "learning_rate": 9.812509197753618e-05, + "loss": 0.029, + "step": 81980 + }, + { + "epoch": 0.2624, + "grad_norm": 0.06527011641043304, + "learning_rate": 9.812418281768918e-05, + "loss": 0.0291, + "step": 82000 + }, + { + "epoch": 0.262464, + "grad_norm": 0.10724814188083759, + "learning_rate": 9.812327344167957e-05, + "loss": 0.0329, + "step": 82020 + }, + { + "epoch": 0.262528, + "grad_norm": 0.09267235220214855, + "learning_rate": 9.812236384951145e-05, + "loss": 0.0309, + "step": 82040 + }, + { + "epoch": 0.262592, + "grad_norm": 0.0787347844346524, + "learning_rate": 9.81214540411889e-05, + "loss": 0.0335, + "step": 82060 + }, + { + "epoch": 0.262656, + "grad_norm": 0.06652226314305634, + "learning_rate": 9.812054401671601e-05, + "loss": 0.0306, + "step": 82080 + }, + { + "epoch": 0.26272, + "grad_norm": 0.11702629384103384, + "learning_rate": 9.811963377609687e-05, + "loss": 0.0312, + "step": 82100 + }, + { + "epoch": 0.262784, + "grad_norm": 0.09740386779261498, + "learning_rate": 9.811872331933556e-05, + "loss": 0.0316, + "step": 82120 + }, + { + "epoch": 0.262848, + "grad_norm": 0.08625983478789148, + "learning_rate": 9.811781264643618e-05, + "loss": 0.031, + "step": 82140 + }, + { + "epoch": 0.262912, + "grad_norm": 0.08435834094911722, + "learning_rate": 9.81169017574028e-05, + "loss": 0.0293, + "step": 82160 + }, + { + "epoch": 0.262976, + "grad_norm": 0.054182967548826455, + "learning_rate": 9.811599065223955e-05, + "loss": 0.0309, + "step": 82180 + }, + { + "epoch": 0.26304, + "grad_norm": 0.09404408733885183, + "learning_rate": 9.811507933095049e-05, + "loss": 0.0297, + "step": 82200 + }, + { + "epoch": 0.263104, + "grad_norm": 0.07798906024146565, + "learning_rate": 9.811416779353973e-05, + "loss": 0.0282, + "step": 82220 + }, + { + "epoch": 0.263168, + "grad_norm": 0.05644303330300428, + "learning_rate": 9.811325604001134e-05, + "loss": 0.0295, + "step": 82240 + }, + { + "epoch": 0.263232, + "grad_norm": 0.08238391090596367, + "learning_rate": 9.811234407036943e-05, + "loss": 0.0293, + "step": 82260 + }, + { + "epoch": 0.263296, + "grad_norm": 0.1043279599805744, + "learning_rate": 9.811143188461811e-05, + "loss": 0.0287, + "step": 82280 + }, + { + "epoch": 0.26336, + "grad_norm": 0.12285385131147136, + "learning_rate": 9.811051948276146e-05, + "loss": 0.0325, + "step": 82300 + }, + { + "epoch": 0.263424, + "grad_norm": 0.05591431016466163, + "learning_rate": 9.81096068648036e-05, + "loss": 0.0352, + "step": 82320 + }, + { + "epoch": 0.263488, + "grad_norm": 0.15812167649990463, + "learning_rate": 9.81086940307486e-05, + "loss": 0.0331, + "step": 82340 + }, + { + "epoch": 0.263552, + "grad_norm": 0.0860341796119538, + "learning_rate": 9.810778098060057e-05, + "loss": 0.0315, + "step": 82360 + }, + { + "epoch": 0.263616, + "grad_norm": 0.04860840060452707, + "learning_rate": 9.810686771436363e-05, + "loss": 0.0298, + "step": 82380 + }, + { + "epoch": 0.26368, + "grad_norm": 0.08782150963846318, + "learning_rate": 9.810595423204185e-05, + "loss": 0.0299, + "step": 82400 + }, + { + "epoch": 0.263744, + "grad_norm": 0.06470179903630253, + "learning_rate": 9.810504053363936e-05, + "loss": 0.0316, + "step": 82420 + }, + { + "epoch": 0.263808, + "grad_norm": 0.04362726244937285, + "learning_rate": 9.810412661916026e-05, + "loss": 0.0296, + "step": 82440 + }, + { + "epoch": 0.263872, + "grad_norm": 0.05823542292676939, + "learning_rate": 9.810321248860864e-05, + "loss": 0.0286, + "step": 82460 + }, + { + "epoch": 0.263936, + "grad_norm": 0.052556805970796185, + "learning_rate": 9.810229814198861e-05, + "loss": 0.0287, + "step": 82480 + }, + { + "epoch": 0.264, + "grad_norm": 0.05246565378324967, + "learning_rate": 9.81013835793043e-05, + "loss": 0.0278, + "step": 82500 + }, + { + "epoch": 0.264064, + "grad_norm": 0.05098062108331092, + "learning_rate": 9.810046880055979e-05, + "loss": 0.0281, + "step": 82520 + }, + { + "epoch": 0.264128, + "grad_norm": 0.06358754218972872, + "learning_rate": 9.809955380575921e-05, + "loss": 0.0298, + "step": 82540 + }, + { + "epoch": 0.264192, + "grad_norm": 0.053068448530542565, + "learning_rate": 9.809863859490665e-05, + "loss": 0.0314, + "step": 82560 + }, + { + "epoch": 0.264256, + "grad_norm": 0.08909274805819939, + "learning_rate": 9.809772316800622e-05, + "loss": 0.0282, + "step": 82580 + }, + { + "epoch": 0.26432, + "grad_norm": 0.0533745311320274, + "learning_rate": 9.809680752506207e-05, + "loss": 0.031, + "step": 82600 + }, + { + "epoch": 0.264384, + "grad_norm": 0.05552295021353528, + "learning_rate": 9.809589166607827e-05, + "loss": 0.0303, + "step": 82620 + }, + { + "epoch": 0.264448, + "grad_norm": 0.07063557717636143, + "learning_rate": 9.809497559105893e-05, + "loss": 0.0307, + "step": 82640 + }, + { + "epoch": 0.264512, + "grad_norm": 0.08525692346644853, + "learning_rate": 9.809405930000821e-05, + "loss": 0.0302, + "step": 82660 + }, + { + "epoch": 0.264576, + "grad_norm": 0.06904069494601878, + "learning_rate": 9.809314279293019e-05, + "loss": 0.0299, + "step": 82680 + }, + { + "epoch": 0.26464, + "grad_norm": 0.07834860143495333, + "learning_rate": 9.809222606982899e-05, + "loss": 0.0327, + "step": 82700 + }, + { + "epoch": 0.264704, + "grad_norm": 0.048476885117785806, + "learning_rate": 9.809130913070875e-05, + "loss": 0.0287, + "step": 82720 + }, + { + "epoch": 0.264768, + "grad_norm": 0.07594934709559249, + "learning_rate": 9.809039197557357e-05, + "loss": 0.0328, + "step": 82740 + }, + { + "epoch": 0.264832, + "grad_norm": 0.05329317919173235, + "learning_rate": 9.808947460442757e-05, + "loss": 0.0275, + "step": 82760 + }, + { + "epoch": 0.264896, + "grad_norm": 0.048546923206690254, + "learning_rate": 9.808855701727486e-05, + "loss": 0.026, + "step": 82780 + }, + { + "epoch": 0.26496, + "grad_norm": 0.08896429148318381, + "learning_rate": 9.80876392141196e-05, + "loss": 0.0332, + "step": 82800 + }, + { + "epoch": 0.265024, + "grad_norm": 0.09207320437853551, + "learning_rate": 9.808672119496587e-05, + "loss": 0.0324, + "step": 82820 + }, + { + "epoch": 0.265088, + "grad_norm": 0.07183155953399545, + "learning_rate": 9.808580295981782e-05, + "loss": 0.0278, + "step": 82840 + }, + { + "epoch": 0.265152, + "grad_norm": 0.054770862998715815, + "learning_rate": 9.808488450867955e-05, + "loss": 0.0286, + "step": 82860 + }, + { + "epoch": 0.265216, + "grad_norm": 0.10969675561424654, + "learning_rate": 9.808396584155521e-05, + "loss": 0.0279, + "step": 82880 + }, + { + "epoch": 0.26528, + "grad_norm": 0.044993427871510186, + "learning_rate": 9.808304695844892e-05, + "loss": 0.0261, + "step": 82900 + }, + { + "epoch": 0.265344, + "grad_norm": 0.13350018965163324, + "learning_rate": 9.808212785936479e-05, + "loss": 0.0278, + "step": 82920 + }, + { + "epoch": 0.265408, + "grad_norm": 0.0962440979448633, + "learning_rate": 9.808120854430698e-05, + "loss": 0.0303, + "step": 82940 + }, + { + "epoch": 0.265472, + "grad_norm": 0.07901939686419454, + "learning_rate": 9.808028901327959e-05, + "loss": 0.0298, + "step": 82960 + }, + { + "epoch": 0.265536, + "grad_norm": 0.056455710362152016, + "learning_rate": 9.807936926628679e-05, + "loss": 0.0356, + "step": 82980 + }, + { + "epoch": 0.2656, + "grad_norm": 0.0799307569450036, + "learning_rate": 9.807844930333265e-05, + "loss": 0.0334, + "step": 83000 + }, + { + "epoch": 0.265664, + "grad_norm": 0.08242803749427387, + "learning_rate": 9.807752912442134e-05, + "loss": 0.0314, + "step": 83020 + }, + { + "epoch": 0.265728, + "grad_norm": 0.20927871116377614, + "learning_rate": 9.8076608729557e-05, + "loss": 0.0292, + "step": 83040 + }, + { + "epoch": 0.265792, + "grad_norm": 0.0539293873528177, + "learning_rate": 9.807568811874375e-05, + "loss": 0.0295, + "step": 83060 + }, + { + "epoch": 0.265856, + "grad_norm": 0.08907445720381324, + "learning_rate": 9.807476729198574e-05, + "loss": 0.026, + "step": 83080 + }, + { + "epoch": 0.26592, + "grad_norm": 0.05382542527044014, + "learning_rate": 9.807384624928708e-05, + "loss": 0.0289, + "step": 83100 + }, + { + "epoch": 0.265984, + "grad_norm": 0.061018822787795975, + "learning_rate": 9.807292499065192e-05, + "loss": 0.0307, + "step": 83120 + }, + { + "epoch": 0.266048, + "grad_norm": 0.05502075686613906, + "learning_rate": 9.807200351608442e-05, + "loss": 0.0289, + "step": 83140 + }, + { + "epoch": 0.266112, + "grad_norm": 0.059400822046276024, + "learning_rate": 9.807108182558868e-05, + "loss": 0.0264, + "step": 83160 + }, + { + "epoch": 0.266176, + "grad_norm": 0.056793519015407815, + "learning_rate": 9.807015991916886e-05, + "loss": 0.0287, + "step": 83180 + }, + { + "epoch": 0.26624, + "grad_norm": 0.08116013394870043, + "learning_rate": 9.80692377968291e-05, + "loss": 0.0292, + "step": 83200 + }, + { + "epoch": 0.266304, + "grad_norm": 0.12305114272001146, + "learning_rate": 9.806831545857355e-05, + "loss": 0.0306, + "step": 83220 + }, + { + "epoch": 0.266368, + "grad_norm": 0.08576420917547242, + "learning_rate": 9.806739290440634e-05, + "loss": 0.0311, + "step": 83240 + }, + { + "epoch": 0.266432, + "grad_norm": 0.06050547962807207, + "learning_rate": 9.806647013433163e-05, + "loss": 0.0287, + "step": 83260 + }, + { + "epoch": 0.266496, + "grad_norm": 0.06380217200415908, + "learning_rate": 9.806554714835354e-05, + "loss": 0.0303, + "step": 83280 + }, + { + "epoch": 0.26656, + "grad_norm": 0.05784158302046477, + "learning_rate": 9.806462394647625e-05, + "loss": 0.0306, + "step": 83300 + }, + { + "epoch": 0.266624, + "grad_norm": 0.0821086176670607, + "learning_rate": 9.806370052870387e-05, + "loss": 0.0322, + "step": 83320 + }, + { + "epoch": 0.266688, + "grad_norm": 0.12418416797770163, + "learning_rate": 9.806277689504055e-05, + "loss": 0.0325, + "step": 83340 + }, + { + "epoch": 0.266752, + "grad_norm": 0.0795882345431432, + "learning_rate": 9.806185304549047e-05, + "loss": 0.0298, + "step": 83360 + }, + { + "epoch": 0.266816, + "grad_norm": 0.05904477827471279, + "learning_rate": 9.806092898005776e-05, + "loss": 0.0315, + "step": 83380 + }, + { + "epoch": 0.26688, + "grad_norm": 0.05886312890314173, + "learning_rate": 9.806000469874658e-05, + "loss": 0.0297, + "step": 83400 + }, + { + "epoch": 0.266944, + "grad_norm": 0.11337657251818192, + "learning_rate": 9.805908020156107e-05, + "loss": 0.03, + "step": 83420 + }, + { + "epoch": 0.267008, + "grad_norm": 0.13213416437650924, + "learning_rate": 9.805815548850539e-05, + "loss": 0.031, + "step": 83440 + }, + { + "epoch": 0.267072, + "grad_norm": 0.14603095590654483, + "learning_rate": 9.805723055958369e-05, + "loss": 0.0318, + "step": 83460 + }, + { + "epoch": 0.267136, + "grad_norm": 0.05654691084481549, + "learning_rate": 9.805630541480011e-05, + "loss": 0.0265, + "step": 83480 + }, + { + "epoch": 0.2672, + "grad_norm": 0.05286593249076137, + "learning_rate": 9.805538005415883e-05, + "loss": 0.0272, + "step": 83500 + }, + { + "epoch": 0.267264, + "grad_norm": 0.11100359696754704, + "learning_rate": 9.805445447766402e-05, + "loss": 0.0305, + "step": 83520 + }, + { + "epoch": 0.267328, + "grad_norm": 0.1095437781787013, + "learning_rate": 9.805352868531978e-05, + "loss": 0.0317, + "step": 83540 + }, + { + "epoch": 0.267392, + "grad_norm": 0.11758156960394733, + "learning_rate": 9.805260267713032e-05, + "loss": 0.0327, + "step": 83560 + }, + { + "epoch": 0.267456, + "grad_norm": 0.06142819496947849, + "learning_rate": 9.80516764530998e-05, + "loss": 0.0306, + "step": 83580 + }, + { + "epoch": 0.26752, + "grad_norm": 0.08461361778639025, + "learning_rate": 9.805075001323235e-05, + "loss": 0.0337, + "step": 83600 + }, + { + "epoch": 0.267584, + "grad_norm": 0.11341523994067702, + "learning_rate": 9.804982335753213e-05, + "loss": 0.0318, + "step": 83620 + }, + { + "epoch": 0.267648, + "grad_norm": 0.17135941316363298, + "learning_rate": 9.804889648600333e-05, + "loss": 0.0315, + "step": 83640 + }, + { + "epoch": 0.267712, + "grad_norm": 0.05719203022538088, + "learning_rate": 9.80479693986501e-05, + "loss": 0.0319, + "step": 83660 + }, + { + "epoch": 0.267776, + "grad_norm": 0.06846239563936252, + "learning_rate": 9.804704209547662e-05, + "loss": 0.0302, + "step": 83680 + }, + { + "epoch": 0.26784, + "grad_norm": 0.060188672940843686, + "learning_rate": 9.804611457648701e-05, + "loss": 0.0321, + "step": 83700 + }, + { + "epoch": 0.267904, + "grad_norm": 0.07048526257122847, + "learning_rate": 9.804518684168549e-05, + "loss": 0.0296, + "step": 83720 + }, + { + "epoch": 0.267968, + "grad_norm": 0.08617592330735553, + "learning_rate": 9.804425889107619e-05, + "loss": 0.0302, + "step": 83740 + }, + { + "epoch": 0.268032, + "grad_norm": 0.10785136787932685, + "learning_rate": 9.80433307246633e-05, + "loss": 0.03, + "step": 83760 + }, + { + "epoch": 0.268096, + "grad_norm": 0.04709386912577561, + "learning_rate": 9.804240234245096e-05, + "loss": 0.0311, + "step": 83780 + }, + { + "epoch": 0.26816, + "grad_norm": 0.07624198765432086, + "learning_rate": 9.804147374444339e-05, + "loss": 0.027, + "step": 83800 + }, + { + "epoch": 0.268224, + "grad_norm": 0.05659355702614735, + "learning_rate": 9.804054493064472e-05, + "loss": 0.0285, + "step": 83820 + }, + { + "epoch": 0.268288, + "grad_norm": 0.09442541728789391, + "learning_rate": 9.803961590105913e-05, + "loss": 0.0326, + "step": 83840 + }, + { + "epoch": 0.268352, + "grad_norm": 0.08145683024820768, + "learning_rate": 9.80386866556908e-05, + "loss": 0.0288, + "step": 83860 + }, + { + "epoch": 0.268416, + "grad_norm": 0.06367600734575908, + "learning_rate": 9.80377571945439e-05, + "loss": 0.0273, + "step": 83880 + }, + { + "epoch": 0.26848, + "grad_norm": 0.14422940249471647, + "learning_rate": 9.80368275176226e-05, + "loss": 0.0292, + "step": 83900 + }, + { + "epoch": 0.268544, + "grad_norm": 0.07948842822171882, + "learning_rate": 9.80358976249311e-05, + "loss": 0.0315, + "step": 83920 + }, + { + "epoch": 0.268608, + "grad_norm": 0.04827399032183594, + "learning_rate": 9.803496751647354e-05, + "loss": 0.0299, + "step": 83940 + }, + { + "epoch": 0.268672, + "grad_norm": 0.07905905307559835, + "learning_rate": 9.803403719225412e-05, + "loss": 0.028, + "step": 83960 + }, + { + "epoch": 0.268736, + "grad_norm": 0.06400415304780636, + "learning_rate": 9.803310665227702e-05, + "loss": 0.0273, + "step": 83980 + }, + { + "epoch": 0.2688, + "grad_norm": 0.08879650397370734, + "learning_rate": 9.80321758965464e-05, + "loss": 0.0301, + "step": 84000 + }, + { + "epoch": 0.268864, + "grad_norm": 0.05051466954464443, + "learning_rate": 9.803124492506649e-05, + "loss": 0.0297, + "step": 84020 + }, + { + "epoch": 0.268928, + "grad_norm": 0.08875681379309433, + "learning_rate": 9.803031373784139e-05, + "loss": 0.0312, + "step": 84040 + }, + { + "epoch": 0.268992, + "grad_norm": 0.04905621779177208, + "learning_rate": 9.802938233487537e-05, + "loss": 0.0312, + "step": 84060 + }, + { + "epoch": 0.269056, + "grad_norm": 0.061605606064430204, + "learning_rate": 9.802845071617255e-05, + "loss": 0.0323, + "step": 84080 + }, + { + "epoch": 0.26912, + "grad_norm": 0.13002920965100176, + "learning_rate": 9.802751888173714e-05, + "loss": 0.0323, + "step": 84100 + }, + { + "epoch": 0.269184, + "grad_norm": 0.12527330086309482, + "learning_rate": 9.802658683157332e-05, + "loss": 0.0342, + "step": 84120 + }, + { + "epoch": 0.269248, + "grad_norm": 0.05306600849477869, + "learning_rate": 9.802565456568532e-05, + "loss": 0.0288, + "step": 84140 + }, + { + "epoch": 0.269312, + "grad_norm": 0.08326756145931723, + "learning_rate": 9.802472208407725e-05, + "loss": 0.0317, + "step": 84160 + }, + { + "epoch": 0.269376, + "grad_norm": 0.05991642605912518, + "learning_rate": 9.802378938675334e-05, + "loss": 0.029, + "step": 84180 + }, + { + "epoch": 0.26944, + "grad_norm": 0.06939870452093788, + "learning_rate": 9.802285647371777e-05, + "loss": 0.0318, + "step": 84200 + }, + { + "epoch": 0.269504, + "grad_norm": 0.08732350545099776, + "learning_rate": 9.802192334497475e-05, + "loss": 0.0313, + "step": 84220 + }, + { + "epoch": 0.269568, + "grad_norm": 0.06525631033888081, + "learning_rate": 9.802099000052844e-05, + "loss": 0.0292, + "step": 84240 + }, + { + "epoch": 0.269632, + "grad_norm": 0.05811240368582541, + "learning_rate": 9.802005644038307e-05, + "loss": 0.0295, + "step": 84260 + }, + { + "epoch": 0.269696, + "grad_norm": 0.05498861911976698, + "learning_rate": 9.80191226645428e-05, + "loss": 0.0297, + "step": 84280 + }, + { + "epoch": 0.26976, + "grad_norm": 0.05978738507642877, + "learning_rate": 9.801818867301185e-05, + "loss": 0.0323, + "step": 84300 + }, + { + "epoch": 0.269824, + "grad_norm": 0.06846160899227903, + "learning_rate": 9.801725446579439e-05, + "loss": 0.0285, + "step": 84320 + }, + { + "epoch": 0.269888, + "grad_norm": 0.07560469691925452, + "learning_rate": 9.801632004289464e-05, + "loss": 0.0323, + "step": 84340 + }, + { + "epoch": 0.269952, + "grad_norm": 0.05342238755106254, + "learning_rate": 9.801538540431678e-05, + "loss": 0.0333, + "step": 84360 + }, + { + "epoch": 0.270016, + "grad_norm": 0.07723935859970293, + "learning_rate": 9.801445055006502e-05, + "loss": 0.0296, + "step": 84380 + }, + { + "epoch": 0.27008, + "grad_norm": 0.08018040919385412, + "learning_rate": 9.801351548014355e-05, + "loss": 0.0298, + "step": 84400 + }, + { + "epoch": 0.270144, + "grad_norm": 0.24518639464800812, + "learning_rate": 9.801258019455657e-05, + "loss": 0.0303, + "step": 84420 + }, + { + "epoch": 0.270208, + "grad_norm": 0.07175643329712977, + "learning_rate": 9.801164469330828e-05, + "loss": 0.0307, + "step": 84440 + }, + { + "epoch": 0.270272, + "grad_norm": 0.06497103123223065, + "learning_rate": 9.80107089764029e-05, + "loss": 0.0309, + "step": 84460 + }, + { + "epoch": 0.270336, + "grad_norm": 0.07397703204474573, + "learning_rate": 9.800977304384463e-05, + "loss": 0.0291, + "step": 84480 + }, + { + "epoch": 0.2704, + "grad_norm": 0.08838116531025192, + "learning_rate": 9.800883689563763e-05, + "loss": 0.0324, + "step": 84500 + }, + { + "epoch": 0.270464, + "grad_norm": 0.0878010237333341, + "learning_rate": 9.800790053178617e-05, + "loss": 0.0304, + "step": 84520 + }, + { + "epoch": 0.270528, + "grad_norm": 0.060596572739442287, + "learning_rate": 9.80069639522944e-05, + "loss": 0.031, + "step": 84540 + }, + { + "epoch": 0.270592, + "grad_norm": 0.08069378972392388, + "learning_rate": 9.800602715716657e-05, + "loss": 0.0318, + "step": 84560 + }, + { + "epoch": 0.270656, + "grad_norm": 0.09604190548341528, + "learning_rate": 9.800509014640687e-05, + "loss": 0.0278, + "step": 84580 + }, + { + "epoch": 0.27072, + "grad_norm": 0.05566169449220183, + "learning_rate": 9.80041529200195e-05, + "loss": 0.0268, + "step": 84600 + }, + { + "epoch": 0.270784, + "grad_norm": 0.05497926504454757, + "learning_rate": 9.800321547800868e-05, + "loss": 0.0315, + "step": 84620 + }, + { + "epoch": 0.270848, + "grad_norm": 0.09314957843132894, + "learning_rate": 9.80022778203786e-05, + "loss": 0.0323, + "step": 84640 + }, + { + "epoch": 0.270912, + "grad_norm": 0.041308005308699346, + "learning_rate": 9.800133994713353e-05, + "loss": 0.0274, + "step": 84660 + }, + { + "epoch": 0.270976, + "grad_norm": 0.1677917064111496, + "learning_rate": 9.800040185827761e-05, + "loss": 0.0324, + "step": 84680 + }, + { + "epoch": 0.27104, + "grad_norm": 0.06273966244738766, + "learning_rate": 9.799946355381509e-05, + "loss": 0.0302, + "step": 84700 + }, + { + "epoch": 0.271104, + "grad_norm": 0.06352611855355872, + "learning_rate": 9.799852503375019e-05, + "loss": 0.0273, + "step": 84720 + }, + { + "epoch": 0.271168, + "grad_norm": 0.09325996628082922, + "learning_rate": 9.799758629808712e-05, + "loss": 0.0297, + "step": 84740 + }, + { + "epoch": 0.271232, + "grad_norm": 0.0786642599073224, + "learning_rate": 9.79966473468301e-05, + "loss": 0.0292, + "step": 84760 + }, + { + "epoch": 0.271296, + "grad_norm": 0.06626413740482451, + "learning_rate": 9.799570817998334e-05, + "loss": 0.0304, + "step": 84780 + }, + { + "epoch": 0.27136, + "grad_norm": 0.08519917166108687, + "learning_rate": 9.799476879755105e-05, + "loss": 0.0293, + "step": 84800 + }, + { + "epoch": 0.271424, + "grad_norm": 0.06089092400708726, + "learning_rate": 9.799382919953745e-05, + "loss": 0.0338, + "step": 84820 + }, + { + "epoch": 0.271488, + "grad_norm": 0.11995367248914196, + "learning_rate": 9.79928893859468e-05, + "loss": 0.0309, + "step": 84840 + }, + { + "epoch": 0.271552, + "grad_norm": 0.0681922965306965, + "learning_rate": 9.799194935678327e-05, + "loss": 0.0304, + "step": 84860 + }, + { + "epoch": 0.271616, + "grad_norm": 0.09819432056694928, + "learning_rate": 9.79910091120511e-05, + "loss": 0.0318, + "step": 84880 + }, + { + "epoch": 0.27168, + "grad_norm": 0.12725861957601556, + "learning_rate": 9.799006865175453e-05, + "loss": 0.0331, + "step": 84900 + }, + { + "epoch": 0.271744, + "grad_norm": 0.08678271465884582, + "learning_rate": 9.798912797589777e-05, + "loss": 0.0301, + "step": 84920 + }, + { + "epoch": 0.271808, + "grad_norm": 0.0801701541304339, + "learning_rate": 9.798818708448503e-05, + "loss": 0.0285, + "step": 84940 + }, + { + "epoch": 0.271872, + "grad_norm": 0.05660896989207953, + "learning_rate": 9.798724597752057e-05, + "loss": 0.0292, + "step": 84960 + }, + { + "epoch": 0.271936, + "grad_norm": 0.09016039784040258, + "learning_rate": 9.79863046550086e-05, + "loss": 0.0299, + "step": 84980 + }, + { + "epoch": 0.272, + "grad_norm": 0.08174356912198062, + "learning_rate": 9.798536311695334e-05, + "loss": 0.0312, + "step": 85000 + }, + { + "epoch": 0.272064, + "grad_norm": 0.12707304076724268, + "learning_rate": 9.798442136335904e-05, + "loss": 0.0293, + "step": 85020 + }, + { + "epoch": 0.272128, + "grad_norm": 0.12311919714449544, + "learning_rate": 9.79834793942299e-05, + "loss": 0.03, + "step": 85040 + }, + { + "epoch": 0.272192, + "grad_norm": 0.10883787152805918, + "learning_rate": 9.798253720957017e-05, + "loss": 0.0338, + "step": 85060 + }, + { + "epoch": 0.272256, + "grad_norm": 0.054824874470380695, + "learning_rate": 9.798159480938409e-05, + "loss": 0.031, + "step": 85080 + }, + { + "epoch": 0.27232, + "grad_norm": 0.06097059615035825, + "learning_rate": 9.798065219367589e-05, + "loss": 0.0288, + "step": 85100 + }, + { + "epoch": 0.272384, + "grad_norm": 0.05836905962488298, + "learning_rate": 9.79797093624498e-05, + "loss": 0.0297, + "step": 85120 + }, + { + "epoch": 0.272448, + "grad_norm": 0.08518501764553253, + "learning_rate": 9.797876631571004e-05, + "loss": 0.0306, + "step": 85140 + }, + { + "epoch": 0.272512, + "grad_norm": 0.05333231558663774, + "learning_rate": 9.797782305346086e-05, + "loss": 0.0296, + "step": 85160 + }, + { + "epoch": 0.272576, + "grad_norm": 0.05500228079502022, + "learning_rate": 9.79768795757065e-05, + "loss": 0.03, + "step": 85180 + }, + { + "epoch": 0.27264, + "grad_norm": 0.058086025985596784, + "learning_rate": 9.797593588245119e-05, + "loss": 0.0311, + "step": 85200 + }, + { + "epoch": 0.272704, + "grad_norm": 0.056115054169460964, + "learning_rate": 9.797499197369917e-05, + "loss": 0.0316, + "step": 85220 + }, + { + "epoch": 0.272768, + "grad_norm": 0.08240041059262236, + "learning_rate": 9.797404784945468e-05, + "loss": 0.0309, + "step": 85240 + }, + { + "epoch": 0.272832, + "grad_norm": 0.11274368765566864, + "learning_rate": 9.797310350972197e-05, + "loss": 0.0317, + "step": 85260 + }, + { + "epoch": 0.272896, + "grad_norm": 0.06876873384694675, + "learning_rate": 9.797215895450528e-05, + "loss": 0.0322, + "step": 85280 + }, + { + "epoch": 0.27296, + "grad_norm": 0.052225760597196046, + "learning_rate": 9.797121418380883e-05, + "loss": 0.0323, + "step": 85300 + }, + { + "epoch": 0.273024, + "grad_norm": 0.06015037346163621, + "learning_rate": 9.79702691976369e-05, + "loss": 0.0296, + "step": 85320 + }, + { + "epoch": 0.273088, + "grad_norm": 0.05781054238714294, + "learning_rate": 9.796932399599371e-05, + "loss": 0.032, + "step": 85340 + }, + { + "epoch": 0.273152, + "grad_norm": 0.06317369040919812, + "learning_rate": 9.796837857888351e-05, + "loss": 0.0316, + "step": 85360 + }, + { + "epoch": 0.273216, + "grad_norm": 0.0832702442050206, + "learning_rate": 9.796743294631055e-05, + "loss": 0.0315, + "step": 85380 + }, + { + "epoch": 0.27328, + "grad_norm": 0.06266493999264361, + "learning_rate": 9.796648709827907e-05, + "loss": 0.0285, + "step": 85400 + }, + { + "epoch": 0.273344, + "grad_norm": 0.05567813190490739, + "learning_rate": 9.796554103479334e-05, + "loss": 0.0262, + "step": 85420 + }, + { + "epoch": 0.273408, + "grad_norm": 0.055038755296454446, + "learning_rate": 9.796459475585757e-05, + "loss": 0.0306, + "step": 85440 + }, + { + "epoch": 0.273472, + "grad_norm": 0.04596768925943805, + "learning_rate": 9.796364826147606e-05, + "loss": 0.0291, + "step": 85460 + }, + { + "epoch": 0.273536, + "grad_norm": 0.05376378355457439, + "learning_rate": 9.796270155165301e-05, + "loss": 0.03, + "step": 85480 + }, + { + "epoch": 0.2736, + "grad_norm": 0.09836295804457233, + "learning_rate": 9.796175462639272e-05, + "loss": 0.0302, + "step": 85500 + }, + { + "epoch": 0.273664, + "grad_norm": 0.038182229175833506, + "learning_rate": 9.79608074856994e-05, + "loss": 0.031, + "step": 85520 + }, + { + "epoch": 0.273728, + "grad_norm": 0.0825642382254262, + "learning_rate": 9.795986012957735e-05, + "loss": 0.0299, + "step": 85540 + }, + { + "epoch": 0.273792, + "grad_norm": 0.07299668852744007, + "learning_rate": 9.795891255803078e-05, + "loss": 0.0291, + "step": 85560 + }, + { + "epoch": 0.273856, + "grad_norm": 0.08299507101785107, + "learning_rate": 9.795796477106398e-05, + "loss": 0.0305, + "step": 85580 + }, + { + "epoch": 0.27392, + "grad_norm": 0.06617500607608204, + "learning_rate": 9.795701676868118e-05, + "loss": 0.0299, + "step": 85600 + }, + { + "epoch": 0.273984, + "grad_norm": 0.06268138452121712, + "learning_rate": 9.795606855088667e-05, + "loss": 0.0297, + "step": 85620 + }, + { + "epoch": 0.274048, + "grad_norm": 0.07579649670096585, + "learning_rate": 9.795512011768469e-05, + "loss": 0.031, + "step": 85640 + }, + { + "epoch": 0.274112, + "grad_norm": 0.06705169163430981, + "learning_rate": 9.795417146907949e-05, + "loss": 0.0309, + "step": 85660 + }, + { + "epoch": 0.274176, + "grad_norm": 0.047804243192109644, + "learning_rate": 9.795322260507535e-05, + "loss": 0.0279, + "step": 85680 + }, + { + "epoch": 0.27424, + "grad_norm": 0.14898783175267707, + "learning_rate": 9.795227352567653e-05, + "loss": 0.0281, + "step": 85700 + }, + { + "epoch": 0.274304, + "grad_norm": 0.06547951319468213, + "learning_rate": 9.795132423088728e-05, + "loss": 0.0301, + "step": 85720 + }, + { + "epoch": 0.274368, + "grad_norm": 0.061865644974783925, + "learning_rate": 9.795037472071187e-05, + "loss": 0.0301, + "step": 85740 + }, + { + "epoch": 0.274432, + "grad_norm": 0.10018796097965492, + "learning_rate": 9.794942499515458e-05, + "loss": 0.031, + "step": 85760 + }, + { + "epoch": 0.274496, + "grad_norm": 0.05287937747053099, + "learning_rate": 9.794847505421964e-05, + "loss": 0.0283, + "step": 85780 + }, + { + "epoch": 0.27456, + "grad_norm": 0.06792134738717569, + "learning_rate": 9.794752489791137e-05, + "loss": 0.0278, + "step": 85800 + }, + { + "epoch": 0.274624, + "grad_norm": 0.0657432201599237, + "learning_rate": 9.794657452623399e-05, + "loss": 0.0296, + "step": 85820 + }, + { + "epoch": 0.274688, + "grad_norm": 0.130538813508102, + "learning_rate": 9.79456239391918e-05, + "loss": 0.0297, + "step": 85840 + }, + { + "epoch": 0.274752, + "grad_norm": 0.09722925913876694, + "learning_rate": 9.794467313678903e-05, + "loss": 0.0307, + "step": 85860 + }, + { + "epoch": 0.274816, + "grad_norm": 0.06820036369345822, + "learning_rate": 9.794372211903e-05, + "loss": 0.0275, + "step": 85880 + }, + { + "epoch": 0.27488, + "grad_norm": 0.0648979535393599, + "learning_rate": 9.794277088591896e-05, + "loss": 0.0326, + "step": 85900 + }, + { + "epoch": 0.274944, + "grad_norm": 0.07450109670677589, + "learning_rate": 9.794181943746016e-05, + "loss": 0.0335, + "step": 85920 + }, + { + "epoch": 0.275008, + "grad_norm": 0.05501593972194541, + "learning_rate": 9.794086777365792e-05, + "loss": 0.0312, + "step": 85940 + }, + { + "epoch": 0.275072, + "grad_norm": 0.09658042504712852, + "learning_rate": 9.793991589451649e-05, + "loss": 0.0341, + "step": 85960 + }, + { + "epoch": 0.275136, + "grad_norm": 0.06867279535277838, + "learning_rate": 9.793896380004013e-05, + "loss": 0.0289, + "step": 85980 + }, + { + "epoch": 0.2752, + "grad_norm": 0.08318097974867286, + "learning_rate": 9.793801149023315e-05, + "loss": 0.0346, + "step": 86000 + }, + { + "epoch": 0.275264, + "grad_norm": 0.0643437635406813, + "learning_rate": 9.793705896509979e-05, + "loss": 0.0327, + "step": 86020 + }, + { + "epoch": 0.275328, + "grad_norm": 0.10777496131867699, + "learning_rate": 9.793610622464436e-05, + "loss": 0.0293, + "step": 86040 + }, + { + "epoch": 0.275392, + "grad_norm": 0.1042368934969462, + "learning_rate": 9.793515326887112e-05, + "loss": 0.0312, + "step": 86060 + }, + { + "epoch": 0.275456, + "grad_norm": 0.05918410850329582, + "learning_rate": 9.793420009778439e-05, + "loss": 0.0291, + "step": 86080 + }, + { + "epoch": 0.27552, + "grad_norm": 0.07314888757026075, + "learning_rate": 9.793324671138838e-05, + "loss": 0.0294, + "step": 86100 + }, + { + "epoch": 0.275584, + "grad_norm": 0.11125913626107348, + "learning_rate": 9.793229310968744e-05, + "loss": 0.0314, + "step": 86120 + }, + { + "epoch": 0.275648, + "grad_norm": 0.06474536870651333, + "learning_rate": 9.793133929268579e-05, + "loss": 0.0306, + "step": 86140 + }, + { + "epoch": 0.275712, + "grad_norm": 0.1269847018114356, + "learning_rate": 9.793038526038778e-05, + "loss": 0.0319, + "step": 86160 + }, + { + "epoch": 0.275776, + "grad_norm": 0.06180642748263615, + "learning_rate": 9.792943101279765e-05, + "loss": 0.027, + "step": 86180 + }, + { + "epoch": 0.27584, + "grad_norm": 0.0540963517270991, + "learning_rate": 9.792847654991973e-05, + "loss": 0.0264, + "step": 86200 + }, + { + "epoch": 0.275904, + "grad_norm": 0.050657832243577366, + "learning_rate": 9.792752187175824e-05, + "loss": 0.0304, + "step": 86220 + }, + { + "epoch": 0.275968, + "grad_norm": 0.0594789426264427, + "learning_rate": 9.792656697831752e-05, + "loss": 0.0304, + "step": 86240 + }, + { + "epoch": 0.276032, + "grad_norm": 0.10940930600609011, + "learning_rate": 9.792561186960187e-05, + "loss": 0.0333, + "step": 86260 + }, + { + "epoch": 0.276096, + "grad_norm": 0.07878726246716143, + "learning_rate": 9.792465654561553e-05, + "loss": 0.0264, + "step": 86280 + }, + { + "epoch": 0.27616, + "grad_norm": 0.049459638481452045, + "learning_rate": 9.792370100636284e-05, + "loss": 0.0295, + "step": 86300 + }, + { + "epoch": 0.276224, + "grad_norm": 0.1534288336008078, + "learning_rate": 9.792274525184804e-05, + "loss": 0.0277, + "step": 86320 + }, + { + "epoch": 0.276288, + "grad_norm": 0.06056755514828449, + "learning_rate": 9.792178928207547e-05, + "loss": 0.0319, + "step": 86340 + }, + { + "epoch": 0.276352, + "grad_norm": 0.05531651290280109, + "learning_rate": 9.79208330970494e-05, + "loss": 0.0283, + "step": 86360 + }, + { + "epoch": 0.276416, + "grad_norm": 0.09022758450309691, + "learning_rate": 9.791987669677413e-05, + "loss": 0.0322, + "step": 86380 + }, + { + "epoch": 0.27648, + "grad_norm": 0.10694658017729007, + "learning_rate": 9.791892008125396e-05, + "loss": 0.0322, + "step": 86400 + }, + { + "epoch": 0.276544, + "grad_norm": 0.06339716168520947, + "learning_rate": 9.791796325049318e-05, + "loss": 0.0314, + "step": 86420 + }, + { + "epoch": 0.276608, + "grad_norm": 0.07350140643381871, + "learning_rate": 9.791700620449611e-05, + "loss": 0.0294, + "step": 86440 + }, + { + "epoch": 0.276672, + "grad_norm": 0.0728987286265031, + "learning_rate": 9.791604894326702e-05, + "loss": 0.0268, + "step": 86460 + }, + { + "epoch": 0.276736, + "grad_norm": 0.060632471717695625, + "learning_rate": 9.791509146681021e-05, + "loss": 0.029, + "step": 86480 + }, + { + "epoch": 0.2768, + "grad_norm": 0.05662697976667785, + "learning_rate": 9.791413377513e-05, + "loss": 0.0318, + "step": 86500 + }, + { + "epoch": 0.276864, + "grad_norm": 0.09638368246998033, + "learning_rate": 9.791317586823069e-05, + "loss": 0.0281, + "step": 86520 + }, + { + "epoch": 0.276928, + "grad_norm": 0.05952625489701873, + "learning_rate": 9.791221774611657e-05, + "loss": 0.0312, + "step": 86540 + }, + { + "epoch": 0.276992, + "grad_norm": 0.051886649453004, + "learning_rate": 9.791125940879195e-05, + "loss": 0.0289, + "step": 86560 + }, + { + "epoch": 0.277056, + "grad_norm": 0.09019776516560665, + "learning_rate": 9.791030085626113e-05, + "loss": 0.0288, + "step": 86580 + }, + { + "epoch": 0.27712, + "grad_norm": 0.07869405139886636, + "learning_rate": 9.790934208852842e-05, + "loss": 0.0308, + "step": 86600 + }, + { + "epoch": 0.277184, + "grad_norm": 0.041619912820924906, + "learning_rate": 9.790838310559813e-05, + "loss": 0.0317, + "step": 86620 + }, + { + "epoch": 0.277248, + "grad_norm": 0.0919733048092292, + "learning_rate": 9.790742390747456e-05, + "loss": 0.0308, + "step": 86640 + }, + { + "epoch": 0.277312, + "grad_norm": 0.05977085425331287, + "learning_rate": 9.790646449416203e-05, + "loss": 0.0305, + "step": 86660 + }, + { + "epoch": 0.277376, + "grad_norm": 0.053444572976496886, + "learning_rate": 9.790550486566484e-05, + "loss": 0.0344, + "step": 86680 + }, + { + "epoch": 0.27744, + "grad_norm": 0.04987945064223151, + "learning_rate": 9.790454502198729e-05, + "loss": 0.029, + "step": 86700 + }, + { + "epoch": 0.277504, + "grad_norm": 0.14874682594634941, + "learning_rate": 9.79035849631337e-05, + "loss": 0.0328, + "step": 86720 + }, + { + "epoch": 0.277568, + "grad_norm": 0.08329816196199069, + "learning_rate": 9.79026246891084e-05, + "loss": 0.0318, + "step": 86740 + }, + { + "epoch": 0.277632, + "grad_norm": 0.10139743005476297, + "learning_rate": 9.790166419991569e-05, + "loss": 0.0283, + "step": 86760 + }, + { + "epoch": 0.277696, + "grad_norm": 0.058987790897106034, + "learning_rate": 9.790070349555988e-05, + "loss": 0.029, + "step": 86780 + }, + { + "epoch": 0.27776, + "grad_norm": 0.04690977078573306, + "learning_rate": 9.78997425760453e-05, + "loss": 0.0266, + "step": 86800 + }, + { + "epoch": 0.277824, + "grad_norm": 0.10533020413123005, + "learning_rate": 9.789878144137624e-05, + "loss": 0.0304, + "step": 86820 + }, + { + "epoch": 0.277888, + "grad_norm": 0.06928994150015012, + "learning_rate": 9.789782009155703e-05, + "loss": 0.0323, + "step": 86840 + }, + { + "epoch": 0.277952, + "grad_norm": 0.05024198716775345, + "learning_rate": 9.789685852659201e-05, + "loss": 0.0286, + "step": 86860 + }, + { + "epoch": 0.278016, + "grad_norm": 0.05769235583202409, + "learning_rate": 9.789589674648546e-05, + "loss": 0.03, + "step": 86880 + }, + { + "epoch": 0.27808, + "grad_norm": 0.06799772548319964, + "learning_rate": 9.789493475124172e-05, + "loss": 0.0295, + "step": 86900 + }, + { + "epoch": 0.278144, + "grad_norm": 0.05499796267403035, + "learning_rate": 9.789397254086512e-05, + "loss": 0.0307, + "step": 86920 + }, + { + "epoch": 0.278208, + "grad_norm": 0.056765407475772786, + "learning_rate": 9.789301011535997e-05, + "loss": 0.0342, + "step": 86940 + }, + { + "epoch": 0.278272, + "grad_norm": 0.10057380979106006, + "learning_rate": 9.78920474747306e-05, + "loss": 0.0291, + "step": 86960 + }, + { + "epoch": 0.278336, + "grad_norm": 0.044651434140292524, + "learning_rate": 9.789108461898134e-05, + "loss": 0.0286, + "step": 86980 + }, + { + "epoch": 0.2784, + "grad_norm": 0.08231344306035984, + "learning_rate": 9.789012154811647e-05, + "loss": 0.0276, + "step": 87000 + }, + { + "epoch": 0.278464, + "grad_norm": 0.08359464122419687, + "learning_rate": 9.788915826214038e-05, + "loss": 0.0268, + "step": 87020 + }, + { + "epoch": 0.278528, + "grad_norm": 0.05069940259484946, + "learning_rate": 9.788819476105735e-05, + "loss": 0.0322, + "step": 87040 + }, + { + "epoch": 0.278592, + "grad_norm": 0.04606302440688619, + "learning_rate": 9.788723104487173e-05, + "loss": 0.0275, + "step": 87060 + }, + { + "epoch": 0.278656, + "grad_norm": 0.10052722766795257, + "learning_rate": 9.788626711358784e-05, + "loss": 0.0315, + "step": 87080 + }, + { + "epoch": 0.27872, + "grad_norm": 0.046607637651675256, + "learning_rate": 9.788530296721004e-05, + "loss": 0.0299, + "step": 87100 + }, + { + "epoch": 0.278784, + "grad_norm": 0.085825876056137, + "learning_rate": 9.788433860574259e-05, + "loss": 0.0284, + "step": 87120 + }, + { + "epoch": 0.278848, + "grad_norm": 0.0483105038535408, + "learning_rate": 9.78833740291899e-05, + "loss": 0.0292, + "step": 87140 + }, + { + "epoch": 0.278912, + "grad_norm": 0.08396270922012711, + "learning_rate": 9.788240923755626e-05, + "loss": 0.0314, + "step": 87160 + }, + { + "epoch": 0.278976, + "grad_norm": 0.06600495869368793, + "learning_rate": 9.788144423084602e-05, + "loss": 0.0318, + "step": 87180 + }, + { + "epoch": 0.27904, + "grad_norm": 0.07039616227869977, + "learning_rate": 9.78804790090635e-05, + "loss": 0.0289, + "step": 87200 + }, + { + "epoch": 0.279104, + "grad_norm": 0.05084539090153172, + "learning_rate": 9.787951357221304e-05, + "loss": 0.0295, + "step": 87220 + }, + { + "epoch": 0.279168, + "grad_norm": 0.05869937692400422, + "learning_rate": 9.787854792029898e-05, + "loss": 0.029, + "step": 87240 + }, + { + "epoch": 0.279232, + "grad_norm": 0.058435245175246585, + "learning_rate": 9.787758205332565e-05, + "loss": 0.0313, + "step": 87260 + }, + { + "epoch": 0.279296, + "grad_norm": 0.10909701838162907, + "learning_rate": 9.78766159712974e-05, + "loss": 0.0296, + "step": 87280 + }, + { + "epoch": 0.27936, + "grad_norm": 0.07962040098009483, + "learning_rate": 9.787564967421856e-05, + "loss": 0.0302, + "step": 87300 + }, + { + "epoch": 0.279424, + "grad_norm": 0.05777992700468173, + "learning_rate": 9.787468316209348e-05, + "loss": 0.03, + "step": 87320 + }, + { + "epoch": 0.279488, + "grad_norm": 0.10068539971746267, + "learning_rate": 9.78737164349265e-05, + "loss": 0.0294, + "step": 87340 + }, + { + "epoch": 0.279552, + "grad_norm": 0.12867784239491714, + "learning_rate": 9.787274949272196e-05, + "loss": 0.0296, + "step": 87360 + }, + { + "epoch": 0.279616, + "grad_norm": 0.05918472376299408, + "learning_rate": 9.78717823354842e-05, + "loss": 0.0296, + "step": 87380 + }, + { + "epoch": 0.27968, + "grad_norm": 0.11037426451433513, + "learning_rate": 9.787081496321755e-05, + "loss": 0.0268, + "step": 87400 + }, + { + "epoch": 0.279744, + "grad_norm": 0.07750467271927361, + "learning_rate": 9.786984737592639e-05, + "loss": 0.0285, + "step": 87420 + }, + { + "epoch": 0.279808, + "grad_norm": 0.07153689931213025, + "learning_rate": 9.786887957361504e-05, + "loss": 0.0298, + "step": 87440 + }, + { + "epoch": 0.279872, + "grad_norm": 0.05622509774366207, + "learning_rate": 9.786791155628785e-05, + "loss": 0.0304, + "step": 87460 + }, + { + "epoch": 0.279936, + "grad_norm": 0.06928629182959103, + "learning_rate": 9.786694332394917e-05, + "loss": 0.0297, + "step": 87480 + }, + { + "epoch": 0.28, + "grad_norm": 0.07343317426812289, + "learning_rate": 9.786597487660337e-05, + "loss": 0.0274, + "step": 87500 + }, + { + "epoch": 0.280064, + "grad_norm": 0.05053321047334571, + "learning_rate": 9.786500621425476e-05, + "loss": 0.0305, + "step": 87520 + }, + { + "epoch": 0.280128, + "grad_norm": 0.06271978577819856, + "learning_rate": 9.786403733690773e-05, + "loss": 0.0283, + "step": 87540 + }, + { + "epoch": 0.280192, + "grad_norm": 0.07466438604798549, + "learning_rate": 9.786306824456661e-05, + "loss": 0.0315, + "step": 87560 + }, + { + "epoch": 0.280256, + "grad_norm": 0.04870267024623838, + "learning_rate": 9.786209893723576e-05, + "loss": 0.0328, + "step": 87580 + }, + { + "epoch": 0.28032, + "grad_norm": 0.07132627744862043, + "learning_rate": 9.786112941491951e-05, + "loss": 0.0321, + "step": 87600 + }, + { + "epoch": 0.280384, + "grad_norm": 0.054736281379201224, + "learning_rate": 9.786015967762227e-05, + "loss": 0.0304, + "step": 87620 + }, + { + "epoch": 0.280448, + "grad_norm": 0.13911064240644316, + "learning_rate": 9.785918972534835e-05, + "loss": 0.0299, + "step": 87640 + }, + { + "epoch": 0.280512, + "grad_norm": 0.06041006297236142, + "learning_rate": 9.785821955810211e-05, + "loss": 0.0277, + "step": 87660 + }, + { + "epoch": 0.280576, + "grad_norm": 0.08371421081887791, + "learning_rate": 9.785724917588793e-05, + "loss": 0.0312, + "step": 87680 + }, + { + "epoch": 0.28064, + "grad_norm": 0.05380178955558852, + "learning_rate": 9.785627857871014e-05, + "loss": 0.0288, + "step": 87700 + }, + { + "epoch": 0.280704, + "grad_norm": 0.0633165202959439, + "learning_rate": 9.785530776657311e-05, + "loss": 0.0293, + "step": 87720 + }, + { + "epoch": 0.280768, + "grad_norm": 0.10383334516992913, + "learning_rate": 9.785433673948122e-05, + "loss": 0.0302, + "step": 87740 + }, + { + "epoch": 0.280832, + "grad_norm": 0.06836653180060898, + "learning_rate": 9.785336549743882e-05, + "loss": 0.0293, + "step": 87760 + }, + { + "epoch": 0.280896, + "grad_norm": 0.08977274507134558, + "learning_rate": 9.785239404045026e-05, + "loss": 0.026, + "step": 87780 + }, + { + "epoch": 0.28096, + "grad_norm": 0.07882766459491623, + "learning_rate": 9.785142236851992e-05, + "loss": 0.028, + "step": 87800 + }, + { + "epoch": 0.281024, + "grad_norm": 0.0473756964153374, + "learning_rate": 9.785045048165215e-05, + "loss": 0.0294, + "step": 87820 + }, + { + "epoch": 0.281088, + "grad_norm": 0.08741708088111487, + "learning_rate": 9.784947837985135e-05, + "loss": 0.03, + "step": 87840 + }, + { + "epoch": 0.281152, + "grad_norm": 0.08896864533232335, + "learning_rate": 9.784850606312183e-05, + "loss": 0.0325, + "step": 87860 + }, + { + "epoch": 0.281216, + "grad_norm": 0.07191041666580815, + "learning_rate": 9.784753353146799e-05, + "loss": 0.03, + "step": 87880 + }, + { + "epoch": 0.28128, + "grad_norm": 0.06807424602043183, + "learning_rate": 9.78465607848942e-05, + "loss": 0.0333, + "step": 87900 + }, + { + "epoch": 0.281344, + "grad_norm": 0.059038763684752214, + "learning_rate": 9.784558782340484e-05, + "loss": 0.03, + "step": 87920 + }, + { + "epoch": 0.281408, + "grad_norm": 0.052960339839825524, + "learning_rate": 9.784461464700425e-05, + "loss": 0.0306, + "step": 87940 + }, + { + "epoch": 0.281472, + "grad_norm": 0.07398003151996829, + "learning_rate": 9.784364125569682e-05, + "loss": 0.0315, + "step": 87960 + }, + { + "epoch": 0.281536, + "grad_norm": 0.13746150670390714, + "learning_rate": 9.784266764948693e-05, + "loss": 0.0303, + "step": 87980 + }, + { + "epoch": 0.2816, + "grad_norm": 0.06126668560269124, + "learning_rate": 9.784169382837893e-05, + "loss": 0.0303, + "step": 88000 + }, + { + "epoch": 0.281664, + "grad_norm": 0.12349929981244709, + "learning_rate": 9.784071979237722e-05, + "loss": 0.0322, + "step": 88020 + }, + { + "epoch": 0.281728, + "grad_norm": 0.049660911312167676, + "learning_rate": 9.783974554148614e-05, + "loss": 0.0321, + "step": 88040 + }, + { + "epoch": 0.281792, + "grad_norm": 0.05332389690175779, + "learning_rate": 9.783877107571008e-05, + "loss": 0.0324, + "step": 88060 + }, + { + "epoch": 0.281856, + "grad_norm": 0.061465295396124386, + "learning_rate": 9.783779639505347e-05, + "loss": 0.03, + "step": 88080 + }, + { + "epoch": 0.28192, + "grad_norm": 0.05223477104142059, + "learning_rate": 9.78368214995206e-05, + "loss": 0.0308, + "step": 88100 + }, + { + "epoch": 0.281984, + "grad_norm": 0.07960575393553908, + "learning_rate": 9.783584638911591e-05, + "loss": 0.0307, + "step": 88120 + }, + { + "epoch": 0.282048, + "grad_norm": 0.09587232676052003, + "learning_rate": 9.783487106384377e-05, + "loss": 0.0313, + "step": 88140 + }, + { + "epoch": 0.282112, + "grad_norm": 0.05224901400640773, + "learning_rate": 9.783389552370854e-05, + "loss": 0.0263, + "step": 88160 + }, + { + "epoch": 0.282176, + "grad_norm": 0.06589659869829143, + "learning_rate": 9.783291976871461e-05, + "loss": 0.0302, + "step": 88180 + }, + { + "epoch": 0.28224, + "grad_norm": 0.09856512467752102, + "learning_rate": 9.783194379886638e-05, + "loss": 0.0297, + "step": 88200 + }, + { + "epoch": 0.282304, + "grad_norm": 0.06724834841550711, + "learning_rate": 9.783096761416821e-05, + "loss": 0.0284, + "step": 88220 + }, + { + "epoch": 0.282368, + "grad_norm": 0.13464014024573914, + "learning_rate": 9.78299912146245e-05, + "loss": 0.0313, + "step": 88240 + }, + { + "epoch": 0.282432, + "grad_norm": 0.0751026318896103, + "learning_rate": 9.782901460023964e-05, + "loss": 0.0297, + "step": 88260 + }, + { + "epoch": 0.282496, + "grad_norm": 0.049440990259487685, + "learning_rate": 9.7828037771018e-05, + "loss": 0.0286, + "step": 88280 + }, + { + "epoch": 0.28256, + "grad_norm": 0.06006985393059307, + "learning_rate": 9.782706072696397e-05, + "loss": 0.0314, + "step": 88300 + }, + { + "epoch": 0.282624, + "grad_norm": 0.04658429990932494, + "learning_rate": 9.782608346808197e-05, + "loss": 0.028, + "step": 88320 + }, + { + "epoch": 0.282688, + "grad_norm": 0.05097101886833978, + "learning_rate": 9.782510599437634e-05, + "loss": 0.0282, + "step": 88340 + }, + { + "epoch": 0.282752, + "grad_norm": 0.04749484404470977, + "learning_rate": 9.78241283058515e-05, + "loss": 0.03, + "step": 88360 + }, + { + "epoch": 0.282816, + "grad_norm": 0.05305739924693055, + "learning_rate": 9.782315040251184e-05, + "loss": 0.0324, + "step": 88380 + }, + { + "epoch": 0.28288, + "grad_norm": 0.06833958649238049, + "learning_rate": 9.782217228436174e-05, + "loss": 0.0329, + "step": 88400 + }, + { + "epoch": 0.282944, + "grad_norm": 0.07680255434765057, + "learning_rate": 9.782119395140562e-05, + "loss": 0.0318, + "step": 88420 + }, + { + "epoch": 0.283008, + "grad_norm": 0.0696481173159462, + "learning_rate": 9.782021540364785e-05, + "loss": 0.0293, + "step": 88440 + }, + { + "epoch": 0.283072, + "grad_norm": 0.055033808746435, + "learning_rate": 9.781923664109282e-05, + "loss": 0.0285, + "step": 88460 + }, + { + "epoch": 0.283136, + "grad_norm": 0.0548995321521546, + "learning_rate": 9.781825766374496e-05, + "loss": 0.03, + "step": 88480 + }, + { + "epoch": 0.2832, + "grad_norm": 0.059826105208188235, + "learning_rate": 9.781727847160865e-05, + "loss": 0.0337, + "step": 88500 + }, + { + "epoch": 0.283264, + "grad_norm": 0.06404081555250492, + "learning_rate": 9.781629906468826e-05, + "loss": 0.0302, + "step": 88520 + }, + { + "epoch": 0.283328, + "grad_norm": 0.09659911763500836, + "learning_rate": 9.781531944298823e-05, + "loss": 0.0272, + "step": 88540 + }, + { + "epoch": 0.283392, + "grad_norm": 0.06928993105979961, + "learning_rate": 9.781433960651294e-05, + "loss": 0.0279, + "step": 88560 + }, + { + "epoch": 0.283456, + "grad_norm": 0.2261126099663581, + "learning_rate": 9.781335955526678e-05, + "loss": 0.0286, + "step": 88580 + }, + { + "epoch": 0.28352, + "grad_norm": 0.05803965003481927, + "learning_rate": 9.78123792892542e-05, + "loss": 0.0326, + "step": 88600 + }, + { + "epoch": 0.283584, + "grad_norm": 0.05696573251131896, + "learning_rate": 9.781139880847954e-05, + "loss": 0.0276, + "step": 88620 + }, + { + "epoch": 0.283648, + "grad_norm": 0.051599573783895, + "learning_rate": 9.781041811294725e-05, + "loss": 0.0286, + "step": 88640 + }, + { + "epoch": 0.283712, + "grad_norm": 0.12841100196525285, + "learning_rate": 9.780943720266171e-05, + "loss": 0.0283, + "step": 88660 + }, + { + "epoch": 0.283776, + "grad_norm": 0.11013377575501265, + "learning_rate": 9.780845607762735e-05, + "loss": 0.0284, + "step": 88680 + }, + { + "epoch": 0.28384, + "grad_norm": 0.14117356580801682, + "learning_rate": 9.780747473784855e-05, + "loss": 0.0302, + "step": 88700 + }, + { + "epoch": 0.283904, + "grad_norm": 0.04854178685906144, + "learning_rate": 9.780649318332972e-05, + "loss": 0.0301, + "step": 88720 + }, + { + "epoch": 0.283968, + "grad_norm": 0.057751928754091617, + "learning_rate": 9.780551141407532e-05, + "loss": 0.0296, + "step": 88740 + }, + { + "epoch": 0.284032, + "grad_norm": 0.05281744898626129, + "learning_rate": 9.780452943008968e-05, + "loss": 0.0307, + "step": 88760 + }, + { + "epoch": 0.284096, + "grad_norm": 0.09741459905430251, + "learning_rate": 9.780354723137726e-05, + "loss": 0.0302, + "step": 88780 + }, + { + "epoch": 0.28416, + "grad_norm": 0.06582916777482836, + "learning_rate": 9.780256481794248e-05, + "loss": 0.029, + "step": 88800 + }, + { + "epoch": 0.284224, + "grad_norm": 0.046021894485924426, + "learning_rate": 9.780158218978973e-05, + "loss": 0.0281, + "step": 88820 + }, + { + "epoch": 0.284288, + "grad_norm": 0.054715010439494824, + "learning_rate": 9.78005993469234e-05, + "loss": 0.0281, + "step": 88840 + }, + { + "epoch": 0.284352, + "grad_norm": 0.06949680922523226, + "learning_rate": 9.779961628934797e-05, + "loss": 0.0291, + "step": 88860 + }, + { + "epoch": 0.284416, + "grad_norm": 0.0702222477731458, + "learning_rate": 9.77986330170678e-05, + "loss": 0.0322, + "step": 88880 + }, + { + "epoch": 0.28448, + "grad_norm": 0.15491450208978916, + "learning_rate": 9.779764953008732e-05, + "loss": 0.0284, + "step": 88900 + }, + { + "epoch": 0.284544, + "grad_norm": 0.05502071593515309, + "learning_rate": 9.779666582841096e-05, + "loss": 0.0304, + "step": 88920 + }, + { + "epoch": 0.284608, + "grad_norm": 0.09695113134036062, + "learning_rate": 9.779568191204313e-05, + "loss": 0.0311, + "step": 88940 + }, + { + "epoch": 0.284672, + "grad_norm": 0.056357622199396516, + "learning_rate": 9.779469778098826e-05, + "loss": 0.0303, + "step": 88960 + }, + { + "epoch": 0.284736, + "grad_norm": 0.07533571126966745, + "learning_rate": 9.779371343525074e-05, + "loss": 0.0274, + "step": 88980 + }, + { + "epoch": 0.2848, + "grad_norm": 0.05293558131064359, + "learning_rate": 9.779272887483504e-05, + "loss": 0.0276, + "step": 89000 + }, + { + "epoch": 0.284864, + "grad_norm": 0.06895046497164596, + "learning_rate": 9.779174409974554e-05, + "loss": 0.028, + "step": 89020 + }, + { + "epoch": 0.284928, + "grad_norm": 0.08111736348789243, + "learning_rate": 9.779075910998669e-05, + "loss": 0.0301, + "step": 89040 + }, + { + "epoch": 0.284992, + "grad_norm": 0.07301832095757135, + "learning_rate": 9.778977390556289e-05, + "loss": 0.031, + "step": 89060 + }, + { + "epoch": 0.285056, + "grad_norm": 0.08312772368567646, + "learning_rate": 9.778878848647859e-05, + "loss": 0.0269, + "step": 89080 + }, + { + "epoch": 0.28512, + "grad_norm": 0.07371294642975473, + "learning_rate": 9.778780285273819e-05, + "loss": 0.0285, + "step": 89100 + }, + { + "epoch": 0.285184, + "grad_norm": 0.044083604692719096, + "learning_rate": 9.778681700434615e-05, + "loss": 0.0301, + "step": 89120 + }, + { + "epoch": 0.285248, + "grad_norm": 0.05618575810808776, + "learning_rate": 9.778583094130686e-05, + "loss": 0.0304, + "step": 89140 + }, + { + "epoch": 0.285312, + "grad_norm": 0.05329057691846767, + "learning_rate": 9.77848446636248e-05, + "loss": 0.0276, + "step": 89160 + }, + { + "epoch": 0.285376, + "grad_norm": 0.04884853007009725, + "learning_rate": 9.778385817130435e-05, + "loss": 0.0326, + "step": 89180 + }, + { + "epoch": 0.28544, + "grad_norm": 0.06506324466841085, + "learning_rate": 9.778287146434996e-05, + "loss": 0.0312, + "step": 89200 + }, + { + "epoch": 0.285504, + "grad_norm": 0.055351407225570345, + "learning_rate": 9.778188454276606e-05, + "loss": 0.0302, + "step": 89220 + }, + { + "epoch": 0.285568, + "grad_norm": 0.09047239717841353, + "learning_rate": 9.77808974065571e-05, + "loss": 0.0264, + "step": 89240 + }, + { + "epoch": 0.285632, + "grad_norm": 0.05122396788105196, + "learning_rate": 9.777991005572749e-05, + "loss": 0.0264, + "step": 89260 + }, + { + "epoch": 0.285696, + "grad_norm": 0.05358990156616446, + "learning_rate": 9.777892249028167e-05, + "loss": 0.0284, + "step": 89280 + }, + { + "epoch": 0.28576, + "grad_norm": 0.06251029989484211, + "learning_rate": 9.777793471022409e-05, + "loss": 0.027, + "step": 89300 + }, + { + "epoch": 0.285824, + "grad_norm": 0.08001392537360158, + "learning_rate": 9.777694671555917e-05, + "loss": 0.0328, + "step": 89320 + }, + { + "epoch": 0.285888, + "grad_norm": 0.06365740828699938, + "learning_rate": 9.777595850629135e-05, + "loss": 0.0289, + "step": 89340 + }, + { + "epoch": 0.285952, + "grad_norm": 0.06501786766443653, + "learning_rate": 9.777497008242509e-05, + "loss": 0.029, + "step": 89360 + }, + { + "epoch": 0.286016, + "grad_norm": 0.05542452668656113, + "learning_rate": 9.777398144396481e-05, + "loss": 0.0333, + "step": 89380 + }, + { + "epoch": 0.28608, + "grad_norm": 0.08471202894154078, + "learning_rate": 9.777299259091495e-05, + "loss": 0.0284, + "step": 89400 + }, + { + "epoch": 0.286144, + "grad_norm": 0.058125520702654525, + "learning_rate": 9.777200352327996e-05, + "loss": 0.028, + "step": 89420 + }, + { + "epoch": 0.286208, + "grad_norm": 0.04964450714328324, + "learning_rate": 9.777101424106427e-05, + "loss": 0.0314, + "step": 89440 + }, + { + "epoch": 0.286272, + "grad_norm": 0.12373044477651093, + "learning_rate": 9.777002474427235e-05, + "loss": 0.0314, + "step": 89460 + }, + { + "epoch": 0.286336, + "grad_norm": 0.051312192267298405, + "learning_rate": 9.776903503290861e-05, + "loss": 0.03, + "step": 89480 + }, + { + "epoch": 0.2864, + "grad_norm": 0.06508992516174161, + "learning_rate": 9.776804510697752e-05, + "loss": 0.0292, + "step": 89500 + }, + { + "epoch": 0.286464, + "grad_norm": 0.08404045329784773, + "learning_rate": 9.776705496648353e-05, + "loss": 0.0297, + "step": 89520 + }, + { + "epoch": 0.286528, + "grad_norm": 0.15481356869500665, + "learning_rate": 9.776606461143105e-05, + "loss": 0.035, + "step": 89540 + }, + { + "epoch": 0.286592, + "grad_norm": 0.11139813415110017, + "learning_rate": 9.776507404182458e-05, + "loss": 0.0341, + "step": 89560 + }, + { + "epoch": 0.286656, + "grad_norm": 0.06342343650175972, + "learning_rate": 9.776408325766854e-05, + "loss": 0.0318, + "step": 89580 + }, + { + "epoch": 0.28672, + "grad_norm": 0.062287289389653715, + "learning_rate": 9.776309225896738e-05, + "loss": 0.029, + "step": 89600 + }, + { + "epoch": 0.286784, + "grad_norm": 0.05686312506214288, + "learning_rate": 9.776210104572556e-05, + "loss": 0.0303, + "step": 89620 + }, + { + "epoch": 0.286848, + "grad_norm": 0.10491833774174011, + "learning_rate": 9.776110961794752e-05, + "loss": 0.0296, + "step": 89640 + }, + { + "epoch": 0.286912, + "grad_norm": 0.12425883020021465, + "learning_rate": 9.776011797563774e-05, + "loss": 0.0315, + "step": 89660 + }, + { + "epoch": 0.286976, + "grad_norm": 0.08080751842535337, + "learning_rate": 9.775912611880062e-05, + "loss": 0.029, + "step": 89680 + }, + { + "epoch": 0.28704, + "grad_norm": 0.05938281173314217, + "learning_rate": 9.775813404744069e-05, + "loss": 0.031, + "step": 89700 + }, + { + "epoch": 0.287104, + "grad_norm": 0.0774763110433501, + "learning_rate": 9.775714176156235e-05, + "loss": 0.0298, + "step": 89720 + }, + { + "epoch": 0.287168, + "grad_norm": 0.07109720900874707, + "learning_rate": 9.775614926117008e-05, + "loss": 0.0306, + "step": 89740 + }, + { + "epoch": 0.287232, + "grad_norm": 0.07870570994947514, + "learning_rate": 9.775515654626833e-05, + "loss": 0.0322, + "step": 89760 + }, + { + "epoch": 0.287296, + "grad_norm": 0.0938561685924632, + "learning_rate": 9.775416361686155e-05, + "loss": 0.0302, + "step": 89780 + }, + { + "epoch": 0.28736, + "grad_norm": 0.055573550608340246, + "learning_rate": 9.775317047295423e-05, + "loss": 0.0296, + "step": 89800 + }, + { + "epoch": 0.287424, + "grad_norm": 0.055709014989606516, + "learning_rate": 9.77521771145508e-05, + "loss": 0.0282, + "step": 89820 + }, + { + "epoch": 0.287488, + "grad_norm": 0.05250988161200879, + "learning_rate": 9.775118354165576e-05, + "loss": 0.0282, + "step": 89840 + }, + { + "epoch": 0.287552, + "grad_norm": 0.0675032718202841, + "learning_rate": 9.775018975427352e-05, + "loss": 0.0296, + "step": 89860 + }, + { + "epoch": 0.287616, + "grad_norm": 0.04785461887105399, + "learning_rate": 9.774919575240857e-05, + "loss": 0.03, + "step": 89880 + }, + { + "epoch": 0.28768, + "grad_norm": 0.06139017218692497, + "learning_rate": 9.774820153606538e-05, + "loss": 0.0328, + "step": 89900 + }, + { + "epoch": 0.287744, + "grad_norm": 0.04763281257710071, + "learning_rate": 9.774720710524842e-05, + "loss": 0.0317, + "step": 89920 + }, + { + "epoch": 0.287808, + "grad_norm": 0.052932917411860646, + "learning_rate": 9.774621245996216e-05, + "loss": 0.0269, + "step": 89940 + }, + { + "epoch": 0.287872, + "grad_norm": 0.10611065462391918, + "learning_rate": 9.774521760021105e-05, + "loss": 0.0303, + "step": 89960 + }, + { + "epoch": 0.287936, + "grad_norm": 0.09951662292814249, + "learning_rate": 9.774422252599956e-05, + "loss": 0.0322, + "step": 89980 + }, + { + "epoch": 0.288, + "grad_norm": 0.04299500614826532, + "learning_rate": 9.774322723733216e-05, + "loss": 0.0284, + "step": 90000 + }, + { + "epoch": 0.288064, + "grad_norm": 0.04282735269483895, + "learning_rate": 9.774223173421332e-05, + "loss": 0.0271, + "step": 90020 + }, + { + "epoch": 0.288128, + "grad_norm": 0.08442666608917483, + "learning_rate": 9.774123601664754e-05, + "loss": 0.0296, + "step": 90040 + }, + { + "epoch": 0.288192, + "grad_norm": 0.0583383875958434, + "learning_rate": 9.774024008463924e-05, + "loss": 0.0303, + "step": 90060 + }, + { + "epoch": 0.288256, + "grad_norm": 0.06620937697411836, + "learning_rate": 9.773924393819294e-05, + "loss": 0.0298, + "step": 90080 + }, + { + "epoch": 0.28832, + "grad_norm": 0.07063913117986942, + "learning_rate": 9.77382475773131e-05, + "loss": 0.0299, + "step": 90100 + }, + { + "epoch": 0.288384, + "grad_norm": 0.0563817252766543, + "learning_rate": 9.773725100200418e-05, + "loss": 0.0361, + "step": 90120 + }, + { + "epoch": 0.288448, + "grad_norm": 0.0723818380257318, + "learning_rate": 9.773625421227069e-05, + "loss": 0.035, + "step": 90140 + }, + { + "epoch": 0.288512, + "grad_norm": 0.07440589480646423, + "learning_rate": 9.773525720811707e-05, + "loss": 0.0333, + "step": 90160 + }, + { + "epoch": 0.288576, + "grad_norm": 0.0832625335975234, + "learning_rate": 9.773425998954782e-05, + "loss": 0.0335, + "step": 90180 + }, + { + "epoch": 0.28864, + "grad_norm": 0.10574999276323388, + "learning_rate": 9.773326255656741e-05, + "loss": 0.0293, + "step": 90200 + }, + { + "epoch": 0.288704, + "grad_norm": 0.10007504769718206, + "learning_rate": 9.773226490918033e-05, + "loss": 0.0295, + "step": 90220 + }, + { + "epoch": 0.288768, + "grad_norm": 0.05524518377736703, + "learning_rate": 9.773126704739105e-05, + "loss": 0.0301, + "step": 90240 + }, + { + "epoch": 0.288832, + "grad_norm": 0.07815601214509443, + "learning_rate": 9.773026897120405e-05, + "loss": 0.0281, + "step": 90260 + }, + { + "epoch": 0.288896, + "grad_norm": 0.04091269842335948, + "learning_rate": 9.772927068062384e-05, + "loss": 0.0323, + "step": 90280 + }, + { + "epoch": 0.28896, + "grad_norm": 0.08943291223939251, + "learning_rate": 9.772827217565488e-05, + "loss": 0.0308, + "step": 90300 + }, + { + "epoch": 0.289024, + "grad_norm": 0.09665757665974707, + "learning_rate": 9.772727345630166e-05, + "loss": 0.0308, + "step": 90320 + }, + { + "epoch": 0.289088, + "grad_norm": 0.07256219218147544, + "learning_rate": 9.772627452256866e-05, + "loss": 0.0301, + "step": 90340 + }, + { + "epoch": 0.289152, + "grad_norm": 0.05793711996528834, + "learning_rate": 9.772527537446037e-05, + "loss": 0.0281, + "step": 90360 + }, + { + "epoch": 0.289216, + "grad_norm": 0.06603996435925406, + "learning_rate": 9.772427601198128e-05, + "loss": 0.0306, + "step": 90380 + }, + { + "epoch": 0.28928, + "grad_norm": 0.06249210786793127, + "learning_rate": 9.77232764351359e-05, + "loss": 0.0311, + "step": 90400 + }, + { + "epoch": 0.289344, + "grad_norm": 0.04778339731662798, + "learning_rate": 9.772227664392868e-05, + "loss": 0.0315, + "step": 90420 + }, + { + "epoch": 0.289408, + "grad_norm": 0.09962183509833568, + "learning_rate": 9.772127663836415e-05, + "loss": 0.0304, + "step": 90440 + }, + { + "epoch": 0.289472, + "grad_norm": 0.13778806039425548, + "learning_rate": 9.772027641844676e-05, + "loss": 0.0313, + "step": 90460 + }, + { + "epoch": 0.289536, + "grad_norm": 0.05904370517143128, + "learning_rate": 9.771927598418104e-05, + "loss": 0.0309, + "step": 90480 + }, + { + "epoch": 0.2896, + "grad_norm": 0.06905550334897796, + "learning_rate": 9.771827533557147e-05, + "loss": 0.0306, + "step": 90500 + }, + { + "epoch": 0.289664, + "grad_norm": 0.060002498503972074, + "learning_rate": 9.771727447262253e-05, + "loss": 0.0281, + "step": 90520 + }, + { + "epoch": 0.289728, + "grad_norm": 0.05344135473375168, + "learning_rate": 9.771627339533873e-05, + "loss": 0.027, + "step": 90540 + }, + { + "epoch": 0.289792, + "grad_norm": 0.16661956274098594, + "learning_rate": 9.771527210372457e-05, + "loss": 0.0293, + "step": 90560 + }, + { + "epoch": 0.289856, + "grad_norm": 0.08545591020838353, + "learning_rate": 9.771427059778456e-05, + "loss": 0.0291, + "step": 90580 + }, + { + "epoch": 0.28992, + "grad_norm": 0.07756818908169856, + "learning_rate": 9.771326887752316e-05, + "loss": 0.0302, + "step": 90600 + }, + { + "epoch": 0.289984, + "grad_norm": 0.09640814905473873, + "learning_rate": 9.77122669429449e-05, + "loss": 0.03, + "step": 90620 + }, + { + "epoch": 0.290048, + "grad_norm": 0.049310160843579275, + "learning_rate": 9.771126479405426e-05, + "loss": 0.0276, + "step": 90640 + }, + { + "epoch": 0.290112, + "grad_norm": 0.10786784429403422, + "learning_rate": 9.771026243085577e-05, + "loss": 0.0313, + "step": 90660 + }, + { + "epoch": 0.290176, + "grad_norm": 0.04125065896468611, + "learning_rate": 9.77092598533539e-05, + "loss": 0.0257, + "step": 90680 + }, + { + "epoch": 0.29024, + "grad_norm": 0.11787465830200179, + "learning_rate": 9.770825706155319e-05, + "loss": 0.0304, + "step": 90700 + }, + { + "epoch": 0.290304, + "grad_norm": 0.08234966042402382, + "learning_rate": 9.770725405545811e-05, + "loss": 0.0279, + "step": 90720 + }, + { + "epoch": 0.290368, + "grad_norm": 0.05925705715683261, + "learning_rate": 9.770625083507319e-05, + "loss": 0.0282, + "step": 90740 + }, + { + "epoch": 0.290432, + "grad_norm": 0.10707354793322338, + "learning_rate": 9.77052474004029e-05, + "loss": 0.0294, + "step": 90760 + }, + { + "epoch": 0.290496, + "grad_norm": 0.052827395927943194, + "learning_rate": 9.77042437514518e-05, + "loss": 0.031, + "step": 90780 + }, + { + "epoch": 0.29056, + "grad_norm": 0.1228298536754665, + "learning_rate": 9.770323988822435e-05, + "loss": 0.0293, + "step": 90800 + }, + { + "epoch": 0.290624, + "grad_norm": 0.11183444491597545, + "learning_rate": 9.770223581072508e-05, + "loss": 0.0256, + "step": 90820 + }, + { + "epoch": 0.290688, + "grad_norm": 0.11024128961720699, + "learning_rate": 9.770123151895851e-05, + "loss": 0.0297, + "step": 90840 + }, + { + "epoch": 0.290752, + "grad_norm": 0.07327689541635074, + "learning_rate": 9.770022701292913e-05, + "loss": 0.0297, + "step": 90860 + }, + { + "epoch": 0.290816, + "grad_norm": 0.11274652585358554, + "learning_rate": 9.769922229264148e-05, + "loss": 0.0314, + "step": 90880 + }, + { + "epoch": 0.29088, + "grad_norm": 0.15273051205803598, + "learning_rate": 9.769821735810004e-05, + "loss": 0.0303, + "step": 90900 + }, + { + "epoch": 0.290944, + "grad_norm": 0.08526491956939158, + "learning_rate": 9.769721220930934e-05, + "loss": 0.0297, + "step": 90920 + }, + { + "epoch": 0.291008, + "grad_norm": 0.11502833731827337, + "learning_rate": 9.76962068462739e-05, + "loss": 0.0302, + "step": 90940 + }, + { + "epoch": 0.291072, + "grad_norm": 0.04695453241942728, + "learning_rate": 9.769520126899823e-05, + "loss": 0.028, + "step": 90960 + }, + { + "epoch": 0.291136, + "grad_norm": 0.09808236898860505, + "learning_rate": 9.769419547748683e-05, + "loss": 0.0292, + "step": 90980 + }, + { + "epoch": 0.2912, + "grad_norm": 0.14480822981144173, + "learning_rate": 9.769318947174426e-05, + "loss": 0.031, + "step": 91000 + }, + { + "epoch": 0.291264, + "grad_norm": 0.07866717198496924, + "learning_rate": 9.7692183251775e-05, + "loss": 0.032, + "step": 91020 + }, + { + "epoch": 0.291328, + "grad_norm": 0.049650311672941816, + "learning_rate": 9.769117681758358e-05, + "loss": 0.0302, + "step": 91040 + }, + { + "epoch": 0.291392, + "grad_norm": 0.06965456403528936, + "learning_rate": 9.769017016917454e-05, + "loss": 0.028, + "step": 91060 + }, + { + "epoch": 0.291456, + "grad_norm": 0.0572071517883108, + "learning_rate": 9.768916330655237e-05, + "loss": 0.0273, + "step": 91080 + }, + { + "epoch": 0.29152, + "grad_norm": 0.06540073909124824, + "learning_rate": 9.76881562297216e-05, + "loss": 0.0308, + "step": 91100 + }, + { + "epoch": 0.291584, + "grad_norm": 0.10395635642015878, + "learning_rate": 9.768714893868678e-05, + "loss": 0.0312, + "step": 91120 + }, + { + "epoch": 0.291648, + "grad_norm": 0.10317713023235398, + "learning_rate": 9.76861414334524e-05, + "loss": 0.0291, + "step": 91140 + }, + { + "epoch": 0.291712, + "grad_norm": 0.08379218884760899, + "learning_rate": 9.768513371402302e-05, + "loss": 0.0296, + "step": 91160 + }, + { + "epoch": 0.291776, + "grad_norm": 0.06275444638801035, + "learning_rate": 9.768412578040314e-05, + "loss": 0.0304, + "step": 91180 + }, + { + "epoch": 0.29184, + "grad_norm": 0.1441161861802795, + "learning_rate": 9.76831176325973e-05, + "loss": 0.0317, + "step": 91200 + }, + { + "epoch": 0.291904, + "grad_norm": 0.07582917438235694, + "learning_rate": 9.768210927061001e-05, + "loss": 0.0319, + "step": 91220 + }, + { + "epoch": 0.291968, + "grad_norm": 0.06907292510389078, + "learning_rate": 9.768110069444583e-05, + "loss": 0.0326, + "step": 91240 + }, + { + "epoch": 0.292032, + "grad_norm": 0.056168450011103786, + "learning_rate": 9.768009190410925e-05, + "loss": 0.0309, + "step": 91260 + }, + { + "epoch": 0.292096, + "grad_norm": 0.059905400701916564, + "learning_rate": 9.767908289960485e-05, + "loss": 0.0336, + "step": 91280 + }, + { + "epoch": 0.29216, + "grad_norm": 0.068418325137712, + "learning_rate": 9.767807368093712e-05, + "loss": 0.0293, + "step": 91300 + }, + { + "epoch": 0.292224, + "grad_norm": 0.12201716790175107, + "learning_rate": 9.767706424811062e-05, + "loss": 0.0294, + "step": 91320 + }, + { + "epoch": 0.292288, + "grad_norm": 0.0656391123052192, + "learning_rate": 9.767605460112987e-05, + "loss": 0.0312, + "step": 91340 + }, + { + "epoch": 0.292352, + "grad_norm": 0.07313453415323429, + "learning_rate": 9.76750447399994e-05, + "loss": 0.0321, + "step": 91360 + }, + { + "epoch": 0.292416, + "grad_norm": 0.05311425433808709, + "learning_rate": 9.767403466472377e-05, + "loss": 0.0312, + "step": 91380 + }, + { + "epoch": 0.29248, + "grad_norm": 0.05100215923486794, + "learning_rate": 9.76730243753075e-05, + "loss": 0.0319, + "step": 91400 + }, + { + "epoch": 0.292544, + "grad_norm": 0.058368682684162135, + "learning_rate": 9.767201387175513e-05, + "loss": 0.0289, + "step": 91420 + }, + { + "epoch": 0.292608, + "grad_norm": 0.05295764885521145, + "learning_rate": 9.767100315407118e-05, + "loss": 0.0276, + "step": 91440 + }, + { + "epoch": 0.292672, + "grad_norm": 0.11181656899197412, + "learning_rate": 9.766999222226024e-05, + "loss": 0.0303, + "step": 91460 + }, + { + "epoch": 0.292736, + "grad_norm": 0.05581452207360544, + "learning_rate": 9.76689810763268e-05, + "loss": 0.0336, + "step": 91480 + }, + { + "epoch": 0.2928, + "grad_norm": 0.07865797959959958, + "learning_rate": 9.766796971627543e-05, + "loss": 0.031, + "step": 91500 + }, + { + "epoch": 0.292864, + "grad_norm": 0.07081477593630005, + "learning_rate": 9.766695814211067e-05, + "loss": 0.0259, + "step": 91520 + }, + { + "epoch": 0.292928, + "grad_norm": 0.08020682934050305, + "learning_rate": 9.766594635383703e-05, + "loss": 0.0277, + "step": 91540 + }, + { + "epoch": 0.292992, + "grad_norm": 0.09197806165680074, + "learning_rate": 9.766493435145911e-05, + "loss": 0.0323, + "step": 91560 + }, + { + "epoch": 0.293056, + "grad_norm": 0.0715514169888002, + "learning_rate": 9.766392213498143e-05, + "loss": 0.0324, + "step": 91580 + }, + { + "epoch": 0.29312, + "grad_norm": 0.06449296342457878, + "learning_rate": 9.766290970440852e-05, + "loss": 0.033, + "step": 91600 + }, + { + "epoch": 0.293184, + "grad_norm": 0.04727015296580667, + "learning_rate": 9.766189705974495e-05, + "loss": 0.0284, + "step": 91620 + }, + { + "epoch": 0.293248, + "grad_norm": 0.06958421543175881, + "learning_rate": 9.766088420099526e-05, + "loss": 0.0299, + "step": 91640 + }, + { + "epoch": 0.293312, + "grad_norm": 0.06023564953660223, + "learning_rate": 9.765987112816401e-05, + "loss": 0.0294, + "step": 91660 + }, + { + "epoch": 0.293376, + "grad_norm": 0.0628297333038737, + "learning_rate": 9.765885784125572e-05, + "loss": 0.0307, + "step": 91680 + }, + { + "epoch": 0.29344, + "grad_norm": 0.09645617750849343, + "learning_rate": 9.765784434027499e-05, + "loss": 0.0291, + "step": 91700 + }, + { + "epoch": 0.293504, + "grad_norm": 0.11391629473489072, + "learning_rate": 9.765683062522633e-05, + "loss": 0.0293, + "step": 91720 + }, + { + "epoch": 0.293568, + "grad_norm": 0.08001957907025346, + "learning_rate": 9.765581669611431e-05, + "loss": 0.0285, + "step": 91740 + }, + { + "epoch": 0.293632, + "grad_norm": 0.04834159808894664, + "learning_rate": 9.765480255294348e-05, + "loss": 0.0306, + "step": 91760 + }, + { + "epoch": 0.293696, + "grad_norm": 0.08007300543894137, + "learning_rate": 9.76537881957184e-05, + "loss": 0.0301, + "step": 91780 + }, + { + "epoch": 0.29376, + "grad_norm": 0.04759108679956382, + "learning_rate": 9.765277362444363e-05, + "loss": 0.0321, + "step": 91800 + }, + { + "epoch": 0.293824, + "grad_norm": 0.07865830657976951, + "learning_rate": 9.765175883912372e-05, + "loss": 0.0276, + "step": 91820 + }, + { + "epoch": 0.293888, + "grad_norm": 0.06130030804815428, + "learning_rate": 9.765074383976324e-05, + "loss": 0.0305, + "step": 91840 + }, + { + "epoch": 0.293952, + "grad_norm": 0.06814356764303563, + "learning_rate": 9.764972862636672e-05, + "loss": 0.0318, + "step": 91860 + }, + { + "epoch": 0.294016, + "grad_norm": 0.0650452328637573, + "learning_rate": 9.764871319893876e-05, + "loss": 0.029, + "step": 91880 + }, + { + "epoch": 0.29408, + "grad_norm": 0.06966640918520409, + "learning_rate": 9.764769755748388e-05, + "loss": 0.0314, + "step": 91900 + }, + { + "epoch": 0.294144, + "grad_norm": 0.04622342552874327, + "learning_rate": 9.764668170200667e-05, + "loss": 0.0285, + "step": 91920 + }, + { + "epoch": 0.294208, + "grad_norm": 0.08146967694424521, + "learning_rate": 9.764566563251169e-05, + "loss": 0.0316, + "step": 91940 + }, + { + "epoch": 0.294272, + "grad_norm": 0.05955156799717587, + "learning_rate": 9.76446493490035e-05, + "loss": 0.0295, + "step": 91960 + }, + { + "epoch": 0.294336, + "grad_norm": 0.13353790737355584, + "learning_rate": 9.764363285148666e-05, + "loss": 0.0331, + "step": 91980 + }, + { + "epoch": 0.2944, + "grad_norm": 0.04756594618684242, + "learning_rate": 9.764261613996573e-05, + "loss": 0.0278, + "step": 92000 + }, + { + "epoch": 0.294464, + "grad_norm": 0.07477437839456276, + "learning_rate": 9.764159921444531e-05, + "loss": 0.0268, + "step": 92020 + }, + { + "epoch": 0.294528, + "grad_norm": 0.07978564942842624, + "learning_rate": 9.764058207492992e-05, + "loss": 0.0289, + "step": 92040 + }, + { + "epoch": 0.294592, + "grad_norm": 0.07627036368404372, + "learning_rate": 9.763956472142417e-05, + "loss": 0.0276, + "step": 92060 + }, + { + "epoch": 0.294656, + "grad_norm": 0.09178556239511186, + "learning_rate": 9.763854715393261e-05, + "loss": 0.0264, + "step": 92080 + }, + { + "epoch": 0.29472, + "grad_norm": 0.13607374133071756, + "learning_rate": 9.763752937245981e-05, + "loss": 0.0303, + "step": 92100 + }, + { + "epoch": 0.294784, + "grad_norm": 0.0664654959912239, + "learning_rate": 9.763651137701036e-05, + "loss": 0.0295, + "step": 92120 + }, + { + "epoch": 0.294848, + "grad_norm": 0.06940529587042338, + "learning_rate": 9.76354931675888e-05, + "loss": 0.0279, + "step": 92140 + }, + { + "epoch": 0.294912, + "grad_norm": 0.04316829998102194, + "learning_rate": 9.763447474419973e-05, + "loss": 0.028, + "step": 92160 + }, + { + "epoch": 0.294976, + "grad_norm": 0.07582613427334536, + "learning_rate": 9.76334561068477e-05, + "loss": 0.0292, + "step": 92180 + }, + { + "epoch": 0.29504, + "grad_norm": 0.04694063326748292, + "learning_rate": 9.763243725553732e-05, + "loss": 0.0273, + "step": 92200 + }, + { + "epoch": 0.295104, + "grad_norm": 0.08714637332164257, + "learning_rate": 9.763141819027316e-05, + "loss": 0.0287, + "step": 92220 + }, + { + "epoch": 0.295168, + "grad_norm": 0.07274383846249166, + "learning_rate": 9.763039891105976e-05, + "loss": 0.0296, + "step": 92240 + }, + { + "epoch": 0.295232, + "grad_norm": 0.04738449931225696, + "learning_rate": 9.762937941790173e-05, + "loss": 0.0299, + "step": 92260 + }, + { + "epoch": 0.295296, + "grad_norm": 0.03534793937859938, + "learning_rate": 9.762835971080365e-05, + "loss": 0.0249, + "step": 92280 + }, + { + "epoch": 0.29536, + "grad_norm": 0.05646974144703749, + "learning_rate": 9.76273397897701e-05, + "loss": 0.0304, + "step": 92300 + }, + { + "epoch": 0.295424, + "grad_norm": 0.06437554700087936, + "learning_rate": 9.762631965480564e-05, + "loss": 0.0305, + "step": 92320 + }, + { + "epoch": 0.295488, + "grad_norm": 0.06365464219128028, + "learning_rate": 9.762529930591487e-05, + "loss": 0.0261, + "step": 92340 + }, + { + "epoch": 0.295552, + "grad_norm": 0.06366934408995589, + "learning_rate": 9.762427874310237e-05, + "loss": 0.0267, + "step": 92360 + }, + { + "epoch": 0.295616, + "grad_norm": 0.041541166168540145, + "learning_rate": 9.762325796637273e-05, + "loss": 0.0279, + "step": 92380 + }, + { + "epoch": 0.29568, + "grad_norm": 0.07629269445902416, + "learning_rate": 9.762223697573052e-05, + "loss": 0.0307, + "step": 92400 + }, + { + "epoch": 0.295744, + "grad_norm": 0.06917241613386792, + "learning_rate": 9.762121577118035e-05, + "loss": 0.0286, + "step": 92420 + }, + { + "epoch": 0.295808, + "grad_norm": 0.04713584219006878, + "learning_rate": 9.762019435272677e-05, + "loss": 0.0293, + "step": 92440 + }, + { + "epoch": 0.295872, + "grad_norm": 0.07325613370387413, + "learning_rate": 9.761917272037442e-05, + "loss": 0.0281, + "step": 92460 + }, + { + "epoch": 0.295936, + "grad_norm": 0.05508016049038129, + "learning_rate": 9.761815087412785e-05, + "loss": 0.0294, + "step": 92480 + }, + { + "epoch": 0.296, + "grad_norm": 0.07920364568934207, + "learning_rate": 9.761712881399164e-05, + "loss": 0.0281, + "step": 92500 + }, + { + "epoch": 0.296064, + "grad_norm": 0.04133992778118721, + "learning_rate": 9.761610653997042e-05, + "loss": 0.0267, + "step": 92520 + }, + { + "epoch": 0.296128, + "grad_norm": 0.04760126596766222, + "learning_rate": 9.761508405206875e-05, + "loss": 0.0303, + "step": 92540 + }, + { + "epoch": 0.296192, + "grad_norm": 0.06717560025692335, + "learning_rate": 9.761406135029124e-05, + "loss": 0.0316, + "step": 92560 + }, + { + "epoch": 0.296256, + "grad_norm": 0.050695331222928505, + "learning_rate": 9.761303843464247e-05, + "loss": 0.0292, + "step": 92580 + }, + { + "epoch": 0.29632, + "grad_norm": 0.07452957385808104, + "learning_rate": 9.761201530512705e-05, + "loss": 0.0291, + "step": 92600 + }, + { + "epoch": 0.296384, + "grad_norm": 0.08772283476796751, + "learning_rate": 9.761099196174958e-05, + "loss": 0.0272, + "step": 92620 + }, + { + "epoch": 0.296448, + "grad_norm": 0.05481834453119384, + "learning_rate": 9.760996840451464e-05, + "loss": 0.0332, + "step": 92640 + }, + { + "epoch": 0.296512, + "grad_norm": 0.06807851974280342, + "learning_rate": 9.760894463342682e-05, + "loss": 0.0304, + "step": 92660 + }, + { + "epoch": 0.296576, + "grad_norm": 0.08789378473903786, + "learning_rate": 9.760792064849074e-05, + "loss": 0.0287, + "step": 92680 + }, + { + "epoch": 0.29664, + "grad_norm": 0.0620620489546058, + "learning_rate": 9.7606896449711e-05, + "loss": 0.0286, + "step": 92700 + }, + { + "epoch": 0.296704, + "grad_norm": 0.05037463710001375, + "learning_rate": 9.760587203709217e-05, + "loss": 0.0282, + "step": 92720 + }, + { + "epoch": 0.296768, + "grad_norm": 0.05326190850822441, + "learning_rate": 9.760484741063889e-05, + "loss": 0.0294, + "step": 92740 + }, + { + "epoch": 0.296832, + "grad_norm": 0.06612417240506965, + "learning_rate": 9.760382257035574e-05, + "loss": 0.0305, + "step": 92760 + }, + { + "epoch": 0.296896, + "grad_norm": 0.03826444620431855, + "learning_rate": 9.760279751624733e-05, + "loss": 0.0262, + "step": 92780 + }, + { + "epoch": 0.29696, + "grad_norm": 0.05057680813634849, + "learning_rate": 9.760177224831826e-05, + "loss": 0.0283, + "step": 92800 + }, + { + "epoch": 0.297024, + "grad_norm": 0.09875891620513237, + "learning_rate": 9.760074676657314e-05, + "loss": 0.0292, + "step": 92820 + }, + { + "epoch": 0.297088, + "grad_norm": 0.050771519436738806, + "learning_rate": 9.759972107101657e-05, + "loss": 0.0301, + "step": 92840 + }, + { + "epoch": 0.297152, + "grad_norm": 0.053507149160257624, + "learning_rate": 9.759869516165316e-05, + "loss": 0.0291, + "step": 92860 + }, + { + "epoch": 0.297216, + "grad_norm": 0.06966717668877542, + "learning_rate": 9.759766903848754e-05, + "loss": 0.0304, + "step": 92880 + }, + { + "epoch": 0.29728, + "grad_norm": 0.050124412811470107, + "learning_rate": 9.759664270152429e-05, + "loss": 0.0285, + "step": 92900 + }, + { + "epoch": 0.297344, + "grad_norm": 0.043830289315342955, + "learning_rate": 9.759561615076802e-05, + "loss": 0.0266, + "step": 92920 + }, + { + "epoch": 0.297408, + "grad_norm": 0.03874883852979017, + "learning_rate": 9.759458938622334e-05, + "loss": 0.0273, + "step": 92940 + }, + { + "epoch": 0.297472, + "grad_norm": 0.06363942452364184, + "learning_rate": 9.75935624078949e-05, + "loss": 0.0339, + "step": 92960 + }, + { + "epoch": 0.297536, + "grad_norm": 0.05080533622856413, + "learning_rate": 9.759253521578727e-05, + "loss": 0.0278, + "step": 92980 + }, + { + "epoch": 0.2976, + "grad_norm": 0.04644300734170224, + "learning_rate": 9.759150780990507e-05, + "loss": 0.0271, + "step": 93000 + }, + { + "epoch": 0.297664, + "grad_norm": 0.04911397203948263, + "learning_rate": 9.759048019025294e-05, + "loss": 0.0291, + "step": 93020 + }, + { + "epoch": 0.297728, + "grad_norm": 0.04211893247353645, + "learning_rate": 9.758945235683548e-05, + "loss": 0.0283, + "step": 93040 + }, + { + "epoch": 0.297792, + "grad_norm": 0.04748009191090313, + "learning_rate": 9.758842430965729e-05, + "loss": 0.0294, + "step": 93060 + }, + { + "epoch": 0.297856, + "grad_norm": 0.07385990905838628, + "learning_rate": 9.758739604872302e-05, + "loss": 0.0338, + "step": 93080 + }, + { + "epoch": 0.29792, + "grad_norm": 0.10180001775422177, + "learning_rate": 9.758636757403726e-05, + "loss": 0.0313, + "step": 93100 + }, + { + "epoch": 0.297984, + "grad_norm": 0.10042417589797091, + "learning_rate": 9.758533888560463e-05, + "loss": 0.0308, + "step": 93120 + }, + { + "epoch": 0.298048, + "grad_norm": 0.09651787509293541, + "learning_rate": 9.75843099834298e-05, + "loss": 0.0304, + "step": 93140 + }, + { + "epoch": 0.298112, + "grad_norm": 0.06690495891199258, + "learning_rate": 9.758328086751733e-05, + "loss": 0.0296, + "step": 93160 + }, + { + "epoch": 0.298176, + "grad_norm": 0.0631208581702279, + "learning_rate": 9.758225153787187e-05, + "loss": 0.0305, + "step": 93180 + }, + { + "epoch": 0.29824, + "grad_norm": 0.10269969248795853, + "learning_rate": 9.758122199449803e-05, + "loss": 0.0314, + "step": 93200 + }, + { + "epoch": 0.298304, + "grad_norm": 0.059109484347261425, + "learning_rate": 9.758019223740047e-05, + "loss": 0.0336, + "step": 93220 + }, + { + "epoch": 0.298368, + "grad_norm": 0.06054948833443805, + "learning_rate": 9.757916226658379e-05, + "loss": 0.0322, + "step": 93240 + }, + { + "epoch": 0.298432, + "grad_norm": 0.05096099556971214, + "learning_rate": 9.75781320820526e-05, + "loss": 0.03, + "step": 93260 + }, + { + "epoch": 0.298496, + "grad_norm": 0.04648841938022175, + "learning_rate": 9.757710168381156e-05, + "loss": 0.0257, + "step": 93280 + }, + { + "epoch": 0.29856, + "grad_norm": 0.05743972877335001, + "learning_rate": 9.757607107186526e-05, + "loss": 0.0291, + "step": 93300 + }, + { + "epoch": 0.298624, + "grad_norm": 0.07420712486854056, + "learning_rate": 9.757504024621837e-05, + "loss": 0.0294, + "step": 93320 + }, + { + "epoch": 0.298688, + "grad_norm": 0.04184720958373901, + "learning_rate": 9.75740092068755e-05, + "loss": 0.0277, + "step": 93340 + }, + { + "epoch": 0.298752, + "grad_norm": 0.04987237949332882, + "learning_rate": 9.757297795384127e-05, + "loss": 0.0322, + "step": 93360 + }, + { + "epoch": 0.298816, + "grad_norm": 0.05930230311028524, + "learning_rate": 9.757194648712033e-05, + "loss": 0.032, + "step": 93380 + }, + { + "epoch": 0.29888, + "grad_norm": 0.0533940352684878, + "learning_rate": 9.757091480671732e-05, + "loss": 0.0299, + "step": 93400 + }, + { + "epoch": 0.298944, + "grad_norm": 0.06024413847768373, + "learning_rate": 9.756988291263687e-05, + "loss": 0.0298, + "step": 93420 + }, + { + "epoch": 0.299008, + "grad_norm": 0.0473554272099614, + "learning_rate": 9.756885080488359e-05, + "loss": 0.0324, + "step": 93440 + }, + { + "epoch": 0.299072, + "grad_norm": 0.06022235069705834, + "learning_rate": 9.756781848346213e-05, + "loss": 0.0324, + "step": 93460 + }, + { + "epoch": 0.299136, + "grad_norm": 0.04044960357924723, + "learning_rate": 9.756678594837713e-05, + "loss": 0.0317, + "step": 93480 + }, + { + "epoch": 0.2992, + "grad_norm": 0.09569020584469361, + "learning_rate": 9.756575319963324e-05, + "loss": 0.028, + "step": 93500 + }, + { + "epoch": 0.299264, + "grad_norm": 0.10281577721148137, + "learning_rate": 9.756472023723509e-05, + "loss": 0.031, + "step": 93520 + }, + { + "epoch": 0.299328, + "grad_norm": 0.10847561896768765, + "learning_rate": 9.756368706118731e-05, + "loss": 0.0302, + "step": 93540 + }, + { + "epoch": 0.299392, + "grad_norm": 0.06936352661580504, + "learning_rate": 9.756265367149455e-05, + "loss": 0.0289, + "step": 93560 + }, + { + "epoch": 0.299456, + "grad_norm": 0.08971338998454124, + "learning_rate": 9.756162006816146e-05, + "loss": 0.026, + "step": 93580 + }, + { + "epoch": 0.29952, + "grad_norm": 0.10501111062829009, + "learning_rate": 9.756058625119266e-05, + "loss": 0.0278, + "step": 93600 + }, + { + "epoch": 0.299584, + "grad_norm": 0.05390507571589922, + "learning_rate": 9.755955222059281e-05, + "loss": 0.0271, + "step": 93620 + }, + { + "epoch": 0.299648, + "grad_norm": 0.08296315077328113, + "learning_rate": 9.755851797636655e-05, + "loss": 0.0262, + "step": 93640 + }, + { + "epoch": 0.299712, + "grad_norm": 0.05938851280829022, + "learning_rate": 9.755748351851852e-05, + "loss": 0.0291, + "step": 93660 + }, + { + "epoch": 0.299776, + "grad_norm": 0.0900857015641943, + "learning_rate": 9.755644884705337e-05, + "loss": 0.0315, + "step": 93680 + }, + { + "epoch": 0.29984, + "grad_norm": 0.07835228758017287, + "learning_rate": 9.755541396197577e-05, + "loss": 0.0257, + "step": 93700 + }, + { + "epoch": 0.299904, + "grad_norm": 0.0537520764195435, + "learning_rate": 9.755437886329033e-05, + "loss": 0.0285, + "step": 93720 + }, + { + "epoch": 0.299968, + "grad_norm": 0.057121313293937576, + "learning_rate": 9.755334355100171e-05, + "loss": 0.0297, + "step": 93740 + }, + { + "epoch": 0.300032, + "grad_norm": 0.07360851673712984, + "learning_rate": 9.755230802511459e-05, + "loss": 0.0323, + "step": 93760 + }, + { + "epoch": 0.300096, + "grad_norm": 0.08795637001847535, + "learning_rate": 9.75512722856336e-05, + "loss": 0.0315, + "step": 93780 + }, + { + "epoch": 0.30016, + "grad_norm": 0.056263459602137085, + "learning_rate": 9.755023633256338e-05, + "loss": 0.0274, + "step": 93800 + }, + { + "epoch": 0.300224, + "grad_norm": 0.1397017486075912, + "learning_rate": 9.75492001659086e-05, + "loss": 0.032, + "step": 93820 + }, + { + "epoch": 0.300288, + "grad_norm": 0.10216143165055774, + "learning_rate": 9.754816378567392e-05, + "loss": 0.0319, + "step": 93840 + }, + { + "epoch": 0.300352, + "grad_norm": 0.06788687760245449, + "learning_rate": 9.754712719186397e-05, + "loss": 0.0322, + "step": 93860 + }, + { + "epoch": 0.300416, + "grad_norm": 0.06559939588754048, + "learning_rate": 9.754609038448343e-05, + "loss": 0.0303, + "step": 93880 + }, + { + "epoch": 0.30048, + "grad_norm": 0.06814221975988273, + "learning_rate": 9.754505336353694e-05, + "loss": 0.0299, + "step": 93900 + }, + { + "epoch": 0.300544, + "grad_norm": 0.04963864249364729, + "learning_rate": 9.754401612902917e-05, + "loss": 0.0304, + "step": 93920 + }, + { + "epoch": 0.300608, + "grad_norm": 0.04562470172151278, + "learning_rate": 9.754297868096478e-05, + "loss": 0.0279, + "step": 93940 + }, + { + "epoch": 0.300672, + "grad_norm": 0.05252628317861611, + "learning_rate": 9.754194101934843e-05, + "loss": 0.0287, + "step": 93960 + }, + { + "epoch": 0.300736, + "grad_norm": 0.051878531476474435, + "learning_rate": 9.754090314418476e-05, + "loss": 0.0311, + "step": 93980 + }, + { + "epoch": 0.3008, + "grad_norm": 0.052972617369619106, + "learning_rate": 9.753986505547845e-05, + "loss": 0.0303, + "step": 94000 + }, + { + "epoch": 0.300864, + "grad_norm": 0.05312557816667071, + "learning_rate": 9.753882675323418e-05, + "loss": 0.0316, + "step": 94020 + }, + { + "epoch": 0.300928, + "grad_norm": 0.11002007188600196, + "learning_rate": 9.753778823745657e-05, + "loss": 0.031, + "step": 94040 + }, + { + "epoch": 0.300992, + "grad_norm": 0.1029225368164262, + "learning_rate": 9.753674950815032e-05, + "loss": 0.034, + "step": 94060 + }, + { + "epoch": 0.301056, + "grad_norm": 0.06604733607269084, + "learning_rate": 9.753571056532009e-05, + "loss": 0.0287, + "step": 94080 + }, + { + "epoch": 0.30112, + "grad_norm": 0.07882769172343433, + "learning_rate": 9.753467140897054e-05, + "loss": 0.0293, + "step": 94100 + }, + { + "epoch": 0.301184, + "grad_norm": 0.08866181267311185, + "learning_rate": 9.753363203910634e-05, + "loss": 0.0311, + "step": 94120 + }, + { + "epoch": 0.301248, + "grad_norm": 0.05555129589892021, + "learning_rate": 9.753259245573215e-05, + "loss": 0.0293, + "step": 94140 + }, + { + "epoch": 0.301312, + "grad_norm": 0.07447333259261675, + "learning_rate": 9.753155265885265e-05, + "loss": 0.0291, + "step": 94160 + }, + { + "epoch": 0.301376, + "grad_norm": 0.055388936824571765, + "learning_rate": 9.753051264847251e-05, + "loss": 0.0331, + "step": 94180 + }, + { + "epoch": 0.30144, + "grad_norm": 0.057413310152814374, + "learning_rate": 9.75294724245964e-05, + "loss": 0.0277, + "step": 94200 + }, + { + "epoch": 0.301504, + "grad_norm": 0.08903227362528963, + "learning_rate": 9.752843198722901e-05, + "loss": 0.0328, + "step": 94220 + }, + { + "epoch": 0.301568, + "grad_norm": 0.07969510804598866, + "learning_rate": 9.752739133637496e-05, + "loss": 0.0316, + "step": 94240 + }, + { + "epoch": 0.301632, + "grad_norm": 0.08473942158004477, + "learning_rate": 9.752635047203899e-05, + "loss": 0.0287, + "step": 94260 + }, + { + "epoch": 0.301696, + "grad_norm": 0.05706151591545048, + "learning_rate": 9.752530939422573e-05, + "loss": 0.0268, + "step": 94280 + }, + { + "epoch": 0.30176, + "grad_norm": 0.06707821647345207, + "learning_rate": 9.752426810293987e-05, + "loss": 0.0301, + "step": 94300 + }, + { + "epoch": 0.301824, + "grad_norm": 0.059785595828966934, + "learning_rate": 9.75232265981861e-05, + "loss": 0.032, + "step": 94320 + }, + { + "epoch": 0.301888, + "grad_norm": 0.0512415827104712, + "learning_rate": 9.752218487996909e-05, + "loss": 0.0288, + "step": 94340 + }, + { + "epoch": 0.301952, + "grad_norm": 0.06911418771091334, + "learning_rate": 9.752114294829349e-05, + "loss": 0.0275, + "step": 94360 + }, + { + "epoch": 0.302016, + "grad_norm": 0.08650974095761418, + "learning_rate": 9.752010080316402e-05, + "loss": 0.0278, + "step": 94380 + }, + { + "epoch": 0.30208, + "grad_norm": 0.09545349751644608, + "learning_rate": 9.751905844458535e-05, + "loss": 0.0277, + "step": 94400 + }, + { + "epoch": 0.302144, + "grad_norm": 0.08237602237927703, + "learning_rate": 9.751801587256217e-05, + "loss": 0.0286, + "step": 94420 + }, + { + "epoch": 0.302208, + "grad_norm": 0.06379308069638082, + "learning_rate": 9.751697308709915e-05, + "loss": 0.0283, + "step": 94440 + }, + { + "epoch": 0.302272, + "grad_norm": 0.055015364650830116, + "learning_rate": 9.751593008820097e-05, + "loss": 0.0294, + "step": 94460 + }, + { + "epoch": 0.302336, + "grad_norm": 0.07225704636302945, + "learning_rate": 9.751488687587232e-05, + "loss": 0.0293, + "step": 94480 + }, + { + "epoch": 0.3024, + "grad_norm": 0.061038777135115484, + "learning_rate": 9.751384345011787e-05, + "loss": 0.0283, + "step": 94500 + }, + { + "epoch": 0.302464, + "grad_norm": 0.04907033046012558, + "learning_rate": 9.751279981094234e-05, + "loss": 0.0283, + "step": 94520 + }, + { + "epoch": 0.302528, + "grad_norm": 0.0680471578699037, + "learning_rate": 9.75117559583504e-05, + "loss": 0.0303, + "step": 94540 + }, + { + "epoch": 0.302592, + "grad_norm": 0.07070786385281531, + "learning_rate": 9.751071189234675e-05, + "loss": 0.0296, + "step": 94560 + }, + { + "epoch": 0.302656, + "grad_norm": 0.059115955144646536, + "learning_rate": 9.750966761293606e-05, + "loss": 0.0327, + "step": 94580 + }, + { + "epoch": 0.30272, + "grad_norm": 0.05771918871239152, + "learning_rate": 9.750862312012303e-05, + "loss": 0.0323, + "step": 94600 + }, + { + "epoch": 0.302784, + "grad_norm": 0.04967366675824461, + "learning_rate": 9.750757841391236e-05, + "loss": 0.03, + "step": 94620 + }, + { + "epoch": 0.302848, + "grad_norm": 0.06555586395922491, + "learning_rate": 9.750653349430873e-05, + "loss": 0.0277, + "step": 94640 + }, + { + "epoch": 0.302912, + "grad_norm": 0.058920360079252825, + "learning_rate": 9.750548836131684e-05, + "loss": 0.0285, + "step": 94660 + }, + { + "epoch": 0.302976, + "grad_norm": 0.0460230943989658, + "learning_rate": 9.750444301494139e-05, + "loss": 0.0308, + "step": 94680 + }, + { + "epoch": 0.30304, + "grad_norm": 0.04231715216312257, + "learning_rate": 9.750339745518705e-05, + "loss": 0.0286, + "step": 94700 + }, + { + "epoch": 0.303104, + "grad_norm": 0.07388294955860458, + "learning_rate": 9.750235168205855e-05, + "loss": 0.0289, + "step": 94720 + }, + { + "epoch": 0.303168, + "grad_norm": 0.08791046597744546, + "learning_rate": 9.750130569556057e-05, + "loss": 0.0318, + "step": 94740 + }, + { + "epoch": 0.303232, + "grad_norm": 0.06405538489248172, + "learning_rate": 9.75002594956978e-05, + "loss": 0.0262, + "step": 94760 + }, + { + "epoch": 0.303296, + "grad_norm": 0.054764885265559456, + "learning_rate": 9.749921308247496e-05, + "loss": 0.0298, + "step": 94780 + }, + { + "epoch": 0.30336, + "grad_norm": 0.07739645263379448, + "learning_rate": 9.749816645589674e-05, + "loss": 0.0314, + "step": 94800 + }, + { + "epoch": 0.303424, + "grad_norm": 0.05386418854229771, + "learning_rate": 9.749711961596784e-05, + "loss": 0.0324, + "step": 94820 + }, + { + "epoch": 0.303488, + "grad_norm": 0.12389060784160438, + "learning_rate": 9.749607256269296e-05, + "loss": 0.0303, + "step": 94840 + }, + { + "epoch": 0.303552, + "grad_norm": 0.09090788685668025, + "learning_rate": 9.74950252960768e-05, + "loss": 0.0319, + "step": 94860 + }, + { + "epoch": 0.303616, + "grad_norm": 0.1360376053731009, + "learning_rate": 9.749397781612409e-05, + "loss": 0.0323, + "step": 94880 + }, + { + "epoch": 0.30368, + "grad_norm": 0.09341251073244318, + "learning_rate": 9.74929301228395e-05, + "loss": 0.0303, + "step": 94900 + }, + { + "epoch": 0.303744, + "grad_norm": 0.055143071529215694, + "learning_rate": 9.749188221622776e-05, + "loss": 0.0301, + "step": 94920 + }, + { + "epoch": 0.303808, + "grad_norm": 0.08780404272319904, + "learning_rate": 9.749083409629356e-05, + "loss": 0.0289, + "step": 94940 + }, + { + "epoch": 0.303872, + "grad_norm": 0.06374780153433633, + "learning_rate": 9.748978576304162e-05, + "loss": 0.0312, + "step": 94960 + }, + { + "epoch": 0.303936, + "grad_norm": 0.0760821747070278, + "learning_rate": 9.748873721647664e-05, + "loss": 0.0256, + "step": 94980 + }, + { + "epoch": 0.304, + "grad_norm": 0.11453883299068271, + "learning_rate": 9.748768845660334e-05, + "loss": 0.0282, + "step": 95000 + }, + { + "epoch": 0.304064, + "grad_norm": 0.07998729821796635, + "learning_rate": 9.748663948342644e-05, + "loss": 0.031, + "step": 95020 + }, + { + "epoch": 0.304128, + "grad_norm": 0.1033771312945872, + "learning_rate": 9.748559029695061e-05, + "loss": 0.0284, + "step": 95040 + }, + { + "epoch": 0.304192, + "grad_norm": 0.044207039882107196, + "learning_rate": 9.748454089718062e-05, + "loss": 0.0266, + "step": 95060 + }, + { + "epoch": 0.304256, + "grad_norm": 0.08766516195245479, + "learning_rate": 9.748349128412113e-05, + "loss": 0.0308, + "step": 95080 + }, + { + "epoch": 0.30432, + "grad_norm": 0.06853870518347974, + "learning_rate": 9.74824414577769e-05, + "loss": 0.0313, + "step": 95100 + }, + { + "epoch": 0.304384, + "grad_norm": 0.10316892233800863, + "learning_rate": 9.748139141815261e-05, + "loss": 0.0296, + "step": 95120 + }, + { + "epoch": 0.304448, + "grad_norm": 0.054818415540439924, + "learning_rate": 9.7480341165253e-05, + "loss": 0.0264, + "step": 95140 + }, + { + "epoch": 0.304512, + "grad_norm": 0.05612375000815914, + "learning_rate": 9.747929069908277e-05, + "loss": 0.0291, + "step": 95160 + }, + { + "epoch": 0.304576, + "grad_norm": 0.07442466363001046, + "learning_rate": 9.747824001964665e-05, + "loss": 0.0335, + "step": 95180 + }, + { + "epoch": 0.30464, + "grad_norm": 0.048591995862759696, + "learning_rate": 9.747718912694935e-05, + "loss": 0.029, + "step": 95200 + }, + { + "epoch": 0.304704, + "grad_norm": 0.06145006317758279, + "learning_rate": 9.74761380209956e-05, + "loss": 0.0273, + "step": 95220 + }, + { + "epoch": 0.304768, + "grad_norm": 0.10021955076680264, + "learning_rate": 9.747508670179011e-05, + "loss": 0.0332, + "step": 95240 + }, + { + "epoch": 0.304832, + "grad_norm": 0.06274187059524505, + "learning_rate": 9.747403516933763e-05, + "loss": 0.0303, + "step": 95260 + }, + { + "epoch": 0.304896, + "grad_norm": 0.07070028183757589, + "learning_rate": 9.747298342364284e-05, + "loss": 0.0327, + "step": 95280 + }, + { + "epoch": 0.30496, + "grad_norm": 0.03639384910484367, + "learning_rate": 9.747193146471051e-05, + "loss": 0.028, + "step": 95300 + }, + { + "epoch": 0.305024, + "grad_norm": 0.055260214284431855, + "learning_rate": 9.747087929254532e-05, + "loss": 0.027, + "step": 95320 + }, + { + "epoch": 0.305088, + "grad_norm": 0.08552672212087975, + "learning_rate": 9.746982690715206e-05, + "loss": 0.031, + "step": 95340 + }, + { + "epoch": 0.305152, + "grad_norm": 0.08830685513841095, + "learning_rate": 9.746877430853538e-05, + "loss": 0.0278, + "step": 95360 + }, + { + "epoch": 0.305216, + "grad_norm": 0.08341111441675421, + "learning_rate": 9.746772149670005e-05, + "loss": 0.0294, + "step": 95380 + }, + { + "epoch": 0.30528, + "grad_norm": 0.05592153437164781, + "learning_rate": 9.746666847165077e-05, + "loss": 0.0299, + "step": 95400 + }, + { + "epoch": 0.305344, + "grad_norm": 0.050315700205101975, + "learning_rate": 9.746561523339232e-05, + "loss": 0.0332, + "step": 95420 + }, + { + "epoch": 0.305408, + "grad_norm": 0.06599171782823489, + "learning_rate": 9.746456178192942e-05, + "loss": 0.0332, + "step": 95440 + }, + { + "epoch": 0.305472, + "grad_norm": 0.13047133943811084, + "learning_rate": 9.746350811726677e-05, + "loss": 0.031, + "step": 95460 + }, + { + "epoch": 0.305536, + "grad_norm": 0.12478517770615456, + "learning_rate": 9.74624542394091e-05, + "loss": 0.0292, + "step": 95480 + }, + { + "epoch": 0.3056, + "grad_norm": 0.049734732453163684, + "learning_rate": 9.746140014836118e-05, + "loss": 0.0279, + "step": 95500 + } + ], + "logging_steps": 20, + "max_steps": 937500, + "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": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}