{ "best_global_step": 11148, "best_metric": 1.9240386486053467, "best_model_checkpoint": "./best_mcqa_model/checkpoint-11148", "epoch": 2.0, "eval_steps": 500, "global_step": 11148, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0001794043774668102, "grad_norm": Infinity, "learning_rate": 5e-05, "loss": 2.0537, "step": 1 }, { "epoch": 0.001794043774668102, "grad_norm": 50.121803283691406, "learning_rate": 4.9994617868676e-05, "loss": 2.3674, "step": 10 }, { "epoch": 0.003588087549336204, "grad_norm": 25.215618133544922, "learning_rate": 4.997667743092932e-05, "loss": 2.2324, "step": 20 }, { "epoch": 0.005382131324004306, "grad_norm": 17.250703811645508, "learning_rate": 4.995873699318264e-05, "loss": 2.1936, "step": 30 }, { "epoch": 0.007176175098672408, "grad_norm": 16.687049865722656, "learning_rate": 4.994079655543596e-05, "loss": 2.1778, "step": 40 }, { "epoch": 0.00897021887334051, "grad_norm": 18.19718360900879, "learning_rate": 4.9922856117689274e-05, "loss": 1.9997, "step": 50 }, { "epoch": 0.010764262648008612, "grad_norm": 18.301307678222656, "learning_rate": 4.990491567994259e-05, "loss": 2.1654, "step": 60 }, { "epoch": 0.012558306422676713, "grad_norm": 18.938064575195312, "learning_rate": 4.988697524219591e-05, "loss": 2.1311, "step": 70 }, { "epoch": 0.014352350197344816, "grad_norm": 24.058290481567383, "learning_rate": 4.986903480444923e-05, "loss": 2.1101, "step": 80 }, { "epoch": 0.016146393972012917, "grad_norm": 20.780832290649414, "learning_rate": 4.985109436670255e-05, "loss": 2.043, "step": 90 }, { "epoch": 0.01794043774668102, "grad_norm": 18.546377182006836, "learning_rate": 4.983315392895587e-05, "loss": 2.1232, "step": 100 }, { "epoch": 0.019734481521349122, "grad_norm": 15.411616325378418, "learning_rate": 4.981521349120919e-05, "loss": 2.1469, "step": 110 }, { "epoch": 0.021528525296017224, "grad_norm": 20.646089553833008, "learning_rate": 4.979727305346251e-05, "loss": 2.2586, "step": 120 }, { "epoch": 0.023322569070685324, "grad_norm": 17.679092407226562, "learning_rate": 4.9779332615715826e-05, "loss": 2.0833, "step": 130 }, { "epoch": 0.025116612845353426, "grad_norm": 16.85576629638672, "learning_rate": 4.976139217796914e-05, "loss": 2.0967, "step": 140 }, { "epoch": 0.02691065662002153, "grad_norm": 13.570154190063477, "learning_rate": 4.974345174022246e-05, "loss": 2.1794, "step": 150 }, { "epoch": 0.02870470039468963, "grad_norm": 20.535985946655273, "learning_rate": 4.972551130247578e-05, "loss": 2.1686, "step": 160 }, { "epoch": 0.030498744169357734, "grad_norm": 16.294347763061523, "learning_rate": 4.97075708647291e-05, "loss": 2.0392, "step": 170 }, { "epoch": 0.03229278794402583, "grad_norm": 12.76968765258789, "learning_rate": 4.968963042698242e-05, "loss": 2.1039, "step": 180 }, { "epoch": 0.034086831718693936, "grad_norm": 14.646772384643555, "learning_rate": 4.967168998923574e-05, "loss": 2.0321, "step": 190 }, { "epoch": 0.03588087549336204, "grad_norm": 13.305708885192871, "learning_rate": 4.965374955148906e-05, "loss": 2.1168, "step": 200 }, { "epoch": 0.03767491926803014, "grad_norm": 14.862943649291992, "learning_rate": 4.963580911374238e-05, "loss": 2.0901, "step": 210 }, { "epoch": 0.039468963042698244, "grad_norm": 15.443259239196777, "learning_rate": 4.96178686759957e-05, "loss": 2.0921, "step": 220 }, { "epoch": 0.041263006817366346, "grad_norm": 15.883851051330566, "learning_rate": 4.959992823824902e-05, "loss": 2.2321, "step": 230 }, { "epoch": 0.04305705059203445, "grad_norm": 12.791455268859863, "learning_rate": 4.9581987800502336e-05, "loss": 2.0071, "step": 240 }, { "epoch": 0.044851094366702544, "grad_norm": 15.706707954406738, "learning_rate": 4.956404736275565e-05, "loss": 2.1264, "step": 250 }, { "epoch": 0.04664513814137065, "grad_norm": 18.551183700561523, "learning_rate": 4.954610692500897e-05, "loss": 1.8886, "step": 260 }, { "epoch": 0.04843918191603875, "grad_norm": 24.170883178710938, "learning_rate": 4.952816648726229e-05, "loss": 2.2594, "step": 270 }, { "epoch": 0.05023322569070685, "grad_norm": 12.178893089294434, "learning_rate": 4.951022604951561e-05, "loss": 2.0534, "step": 280 }, { "epoch": 0.052027269465374955, "grad_norm": 14.537866592407227, "learning_rate": 4.949228561176893e-05, "loss": 1.9462, "step": 290 }, { "epoch": 0.05382131324004306, "grad_norm": 15.807729721069336, "learning_rate": 4.9474345174022246e-05, "loss": 2.0243, "step": 300 }, { "epoch": 0.05561535701471116, "grad_norm": 14.98117446899414, "learning_rate": 4.945640473627557e-05, "loss": 2.0688, "step": 310 }, { "epoch": 0.05740940078937926, "grad_norm": 17.541954040527344, "learning_rate": 4.943846429852889e-05, "loss": 1.9925, "step": 320 }, { "epoch": 0.059203444564047365, "grad_norm": 9.95796012878418, "learning_rate": 4.9420523860782205e-05, "loss": 1.9715, "step": 330 }, { "epoch": 0.06099748833871547, "grad_norm": 12.977546691894531, "learning_rate": 4.940258342303553e-05, "loss": 2.0491, "step": 340 }, { "epoch": 0.06279153211338356, "grad_norm": 15.887458801269531, "learning_rate": 4.938464298528884e-05, "loss": 2.1484, "step": 350 }, { "epoch": 0.06458557588805167, "grad_norm": 11.04765510559082, "learning_rate": 4.9366702547542157e-05, "loss": 2.1138, "step": 360 }, { "epoch": 0.06637961966271977, "grad_norm": 13.182682991027832, "learning_rate": 4.934876210979548e-05, "loss": 2.1559, "step": 370 }, { "epoch": 0.06817366343738787, "grad_norm": 11.596452713012695, "learning_rate": 4.93308216720488e-05, "loss": 2.0895, "step": 380 }, { "epoch": 0.06996770721205597, "grad_norm": 13.062545776367188, "learning_rate": 4.931288123430212e-05, "loss": 2.1482, "step": 390 }, { "epoch": 0.07176175098672408, "grad_norm": 12.839446067810059, "learning_rate": 4.929494079655544e-05, "loss": 2.0827, "step": 400 }, { "epoch": 0.07355579476139218, "grad_norm": 11.77519702911377, "learning_rate": 4.9277000358808756e-05, "loss": 2.1411, "step": 410 }, { "epoch": 0.07534983853606028, "grad_norm": 10.112687110900879, "learning_rate": 4.925905992106208e-05, "loss": 2.0311, "step": 420 }, { "epoch": 0.07714388231072838, "grad_norm": 13.721075057983398, "learning_rate": 4.92411194833154e-05, "loss": 1.9533, "step": 430 }, { "epoch": 0.07893792608539649, "grad_norm": 13.185443878173828, "learning_rate": 4.9223179045568715e-05, "loss": 2.1598, "step": 440 }, { "epoch": 0.08073196986006459, "grad_norm": 11.687894821166992, "learning_rate": 4.920523860782203e-05, "loss": 2.2673, "step": 450 }, { "epoch": 0.08252601363473269, "grad_norm": 17.916851043701172, "learning_rate": 4.918729817007535e-05, "loss": 2.1737, "step": 460 }, { "epoch": 0.0843200574094008, "grad_norm": 11.431639671325684, "learning_rate": 4.916935773232867e-05, "loss": 2.1655, "step": 470 }, { "epoch": 0.0861141011840689, "grad_norm": 14.9763822555542, "learning_rate": 4.915141729458199e-05, "loss": 2.0351, "step": 480 }, { "epoch": 0.087908144958737, "grad_norm": 12.372869491577148, "learning_rate": 4.913347685683531e-05, "loss": 2.2227, "step": 490 }, { "epoch": 0.08970218873340509, "grad_norm": 11.000882148742676, "learning_rate": 4.9115536419088625e-05, "loss": 2.0837, "step": 500 }, { "epoch": 0.09149623250807319, "grad_norm": 22.3309326171875, "learning_rate": 4.909759598134195e-05, "loss": 2.1196, "step": 510 }, { "epoch": 0.0932902762827413, "grad_norm": 15.499113082885742, "learning_rate": 4.9079655543595267e-05, "loss": 2.2164, "step": 520 }, { "epoch": 0.0950843200574094, "grad_norm": 9.863581657409668, "learning_rate": 4.9061715105848584e-05, "loss": 2.1818, "step": 530 }, { "epoch": 0.0968783638320775, "grad_norm": 14.396673202514648, "learning_rate": 4.904377466810191e-05, "loss": 2.0059, "step": 540 }, { "epoch": 0.0986724076067456, "grad_norm": 10.207282066345215, "learning_rate": 4.9025834230355225e-05, "loss": 2.1265, "step": 550 }, { "epoch": 0.1004664513814137, "grad_norm": 11.846447944641113, "learning_rate": 4.900789379260854e-05, "loss": 2.2348, "step": 560 }, { "epoch": 0.10226049515608181, "grad_norm": 15.00061321258545, "learning_rate": 4.898995335486186e-05, "loss": 2.0822, "step": 570 }, { "epoch": 0.10405453893074991, "grad_norm": 19.457279205322266, "learning_rate": 4.897201291711518e-05, "loss": 2.1838, "step": 580 }, { "epoch": 0.10584858270541801, "grad_norm": 9.659820556640625, "learning_rate": 4.89540724793685e-05, "loss": 1.9626, "step": 590 }, { "epoch": 0.10764262648008611, "grad_norm": 14.713663101196289, "learning_rate": 4.893613204162182e-05, "loss": 2.2248, "step": 600 }, { "epoch": 0.10943667025475422, "grad_norm": 10.287516593933105, "learning_rate": 4.8918191603875135e-05, "loss": 2.1525, "step": 610 }, { "epoch": 0.11123071402942232, "grad_norm": 10.284120559692383, "learning_rate": 4.890025116612846e-05, "loss": 2.1112, "step": 620 }, { "epoch": 0.11302475780409042, "grad_norm": 9.741727828979492, "learning_rate": 4.888231072838178e-05, "loss": 2.0442, "step": 630 }, { "epoch": 0.11481880157875853, "grad_norm": 8.978920936584473, "learning_rate": 4.8864370290635094e-05, "loss": 2.0471, "step": 640 }, { "epoch": 0.11661284535342663, "grad_norm": 12.450641632080078, "learning_rate": 4.884642985288841e-05, "loss": 2.1083, "step": 650 }, { "epoch": 0.11840688912809473, "grad_norm": 11.721051216125488, "learning_rate": 4.882848941514173e-05, "loss": 2.2182, "step": 660 }, { "epoch": 0.12020093290276283, "grad_norm": 9.241230964660645, "learning_rate": 4.8810548977395046e-05, "loss": 2.1022, "step": 670 }, { "epoch": 0.12199497667743094, "grad_norm": 9.87799072265625, "learning_rate": 4.879260853964837e-05, "loss": 1.9777, "step": 680 }, { "epoch": 0.12378902045209902, "grad_norm": 13.317924499511719, "learning_rate": 4.877466810190169e-05, "loss": 2.0508, "step": 690 }, { "epoch": 0.12558306422676713, "grad_norm": 11.3270845413208, "learning_rate": 4.8756727664155004e-05, "loss": 2.0391, "step": 700 }, { "epoch": 0.12737710800143523, "grad_norm": 10.925955772399902, "learning_rate": 4.873878722640833e-05, "loss": 2.1051, "step": 710 }, { "epoch": 0.12917115177610333, "grad_norm": 11.26830768585205, "learning_rate": 4.8720846788661646e-05, "loss": 1.994, "step": 720 }, { "epoch": 0.13096519555077143, "grad_norm": 16.53739356994629, "learning_rate": 4.870290635091496e-05, "loss": 2.1491, "step": 730 }, { "epoch": 0.13275923932543954, "grad_norm": 11.774067878723145, "learning_rate": 4.868496591316829e-05, "loss": 2.1163, "step": 740 }, { "epoch": 0.13455328310010764, "grad_norm": 11.250940322875977, "learning_rate": 4.8667025475421604e-05, "loss": 2.1339, "step": 750 }, { "epoch": 0.13634732687477574, "grad_norm": 12.242491722106934, "learning_rate": 4.864908503767492e-05, "loss": 1.8935, "step": 760 }, { "epoch": 0.13814137064944385, "grad_norm": 9.951498985290527, "learning_rate": 4.863114459992824e-05, "loss": 1.9869, "step": 770 }, { "epoch": 0.13993541442411195, "grad_norm": 8.47032356262207, "learning_rate": 4.8613204162181556e-05, "loss": 2.008, "step": 780 }, { "epoch": 0.14172945819878005, "grad_norm": 10.949905395507812, "learning_rate": 4.859526372443488e-05, "loss": 2.1019, "step": 790 }, { "epoch": 0.14352350197344815, "grad_norm": 11.062277793884277, "learning_rate": 4.85773232866882e-05, "loss": 2.1903, "step": 800 }, { "epoch": 0.14531754574811626, "grad_norm": 16.415149688720703, "learning_rate": 4.8559382848941514e-05, "loss": 2.1214, "step": 810 }, { "epoch": 0.14711158952278436, "grad_norm": 10.11135196685791, "learning_rate": 4.854144241119484e-05, "loss": 1.8267, "step": 820 }, { "epoch": 0.14890563329745246, "grad_norm": 10.645978927612305, "learning_rate": 4.8523501973448156e-05, "loss": 2.0422, "step": 830 }, { "epoch": 0.15069967707212056, "grad_norm": 8.619297981262207, "learning_rate": 4.850556153570147e-05, "loss": 2.2584, "step": 840 }, { "epoch": 0.15249372084678867, "grad_norm": 9.050743103027344, "learning_rate": 4.84876210979548e-05, "loss": 2.0256, "step": 850 }, { "epoch": 0.15428776462145677, "grad_norm": 10.283220291137695, "learning_rate": 4.8469680660208114e-05, "loss": 2.0351, "step": 860 }, { "epoch": 0.15608180839612487, "grad_norm": 10.877766609191895, "learning_rate": 4.8451740222461425e-05, "loss": 2.0931, "step": 870 }, { "epoch": 0.15787585217079297, "grad_norm": 13.12762451171875, "learning_rate": 4.843379978471475e-05, "loss": 2.1018, "step": 880 }, { "epoch": 0.15966989594546108, "grad_norm": 9.463711738586426, "learning_rate": 4.8415859346968066e-05, "loss": 2.0868, "step": 890 }, { "epoch": 0.16146393972012918, "grad_norm": 7.113370418548584, "learning_rate": 4.839791890922138e-05, "loss": 2.0611, "step": 900 }, { "epoch": 0.16325798349479728, "grad_norm": 11.066067695617676, "learning_rate": 4.837997847147471e-05, "loss": 2.0621, "step": 910 }, { "epoch": 0.16505202726946538, "grad_norm": 12.794750213623047, "learning_rate": 4.8362038033728025e-05, "loss": 2.0361, "step": 920 }, { "epoch": 0.1668460710441335, "grad_norm": 10.758418083190918, "learning_rate": 4.834409759598135e-05, "loss": 1.9764, "step": 930 }, { "epoch": 0.1686401148188016, "grad_norm": 13.04681396484375, "learning_rate": 4.8326157158234666e-05, "loss": 2.0678, "step": 940 }, { "epoch": 0.1704341585934697, "grad_norm": 12.170126914978027, "learning_rate": 4.830821672048798e-05, "loss": 2.1812, "step": 950 }, { "epoch": 0.1722282023681378, "grad_norm": 9.722840309143066, "learning_rate": 4.82902762827413e-05, "loss": 2.0603, "step": 960 }, { "epoch": 0.1740222461428059, "grad_norm": 11.84780216217041, "learning_rate": 4.827233584499462e-05, "loss": 1.9741, "step": 970 }, { "epoch": 0.175816289917474, "grad_norm": 11.813640594482422, "learning_rate": 4.8254395407247935e-05, "loss": 2.0533, "step": 980 }, { "epoch": 0.17761033369214208, "grad_norm": 14.080793380737305, "learning_rate": 4.823645496950126e-05, "loss": 2.0094, "step": 990 }, { "epoch": 0.17940437746681018, "grad_norm": 10.01248550415039, "learning_rate": 4.8218514531754576e-05, "loss": 2.1019, "step": 1000 }, { "epoch": 0.18119842124147828, "grad_norm": 10.648508071899414, "learning_rate": 4.8200574094007894e-05, "loss": 2.018, "step": 1010 }, { "epoch": 0.18299246501614638, "grad_norm": 12.019718170166016, "learning_rate": 4.818263365626122e-05, "loss": 1.9094, "step": 1020 }, { "epoch": 0.18478650879081449, "grad_norm": 8.853078842163086, "learning_rate": 4.8164693218514535e-05, "loss": 2.1722, "step": 1030 }, { "epoch": 0.1865805525654826, "grad_norm": 8.063078880310059, "learning_rate": 4.814675278076785e-05, "loss": 2.0377, "step": 1040 }, { "epoch": 0.1883745963401507, "grad_norm": 9.952075958251953, "learning_rate": 4.8128812343021176e-05, "loss": 2.1291, "step": 1050 }, { "epoch": 0.1901686401148188, "grad_norm": 10.393296241760254, "learning_rate": 4.811087190527449e-05, "loss": 2.0707, "step": 1060 }, { "epoch": 0.1919626838894869, "grad_norm": 10.250103950500488, "learning_rate": 4.809293146752781e-05, "loss": 2.0411, "step": 1070 }, { "epoch": 0.193756727664155, "grad_norm": 10.035394668579102, "learning_rate": 4.807499102978113e-05, "loss": 1.9696, "step": 1080 }, { "epoch": 0.1955507714388231, "grad_norm": 12.22006893157959, "learning_rate": 4.8057050592034445e-05, "loss": 2.0681, "step": 1090 }, { "epoch": 0.1973448152134912, "grad_norm": 9.505834579467773, "learning_rate": 4.803911015428776e-05, "loss": 2.0469, "step": 1100 }, { "epoch": 0.1991388589881593, "grad_norm": 8.87895393371582, "learning_rate": 4.8021169716541086e-05, "loss": 2.1125, "step": 1110 }, { "epoch": 0.2009329027628274, "grad_norm": 8.545978546142578, "learning_rate": 4.8003229278794404e-05, "loss": 2.008, "step": 1120 }, { "epoch": 0.2027269465374955, "grad_norm": 9.056685447692871, "learning_rate": 4.798528884104773e-05, "loss": 1.9602, "step": 1130 }, { "epoch": 0.20452099031216361, "grad_norm": 15.332698822021484, "learning_rate": 4.7967348403301045e-05, "loss": 1.8802, "step": 1140 }, { "epoch": 0.20631503408683172, "grad_norm": 12.204120635986328, "learning_rate": 4.794940796555436e-05, "loss": 1.9994, "step": 1150 }, { "epoch": 0.20810907786149982, "grad_norm": 10.399602890014648, "learning_rate": 4.7931467527807686e-05, "loss": 2.1549, "step": 1160 }, { "epoch": 0.20990312163616792, "grad_norm": 9.073091506958008, "learning_rate": 4.7913527090061e-05, "loss": 2.1483, "step": 1170 }, { "epoch": 0.21169716541083602, "grad_norm": 11.733457565307617, "learning_rate": 4.7895586652314314e-05, "loss": 2.1155, "step": 1180 }, { "epoch": 0.21349120918550413, "grad_norm": 7.083014011383057, "learning_rate": 4.787764621456764e-05, "loss": 2.0738, "step": 1190 }, { "epoch": 0.21528525296017223, "grad_norm": 11.031323432922363, "learning_rate": 4.7859705776820955e-05, "loss": 2.1221, "step": 1200 }, { "epoch": 0.21707929673484033, "grad_norm": 9.98110580444336, "learning_rate": 4.784176533907427e-05, "loss": 2.0356, "step": 1210 }, { "epoch": 0.21887334050950844, "grad_norm": 13.873873710632324, "learning_rate": 4.78238249013276e-05, "loss": 2.0523, "step": 1220 }, { "epoch": 0.22066738428417654, "grad_norm": 10.365226745605469, "learning_rate": 4.7805884463580914e-05, "loss": 2.0989, "step": 1230 }, { "epoch": 0.22246142805884464, "grad_norm": 9.975032806396484, "learning_rate": 4.778794402583423e-05, "loss": 2.0278, "step": 1240 }, { "epoch": 0.22425547183351274, "grad_norm": 10.711677551269531, "learning_rate": 4.7770003588087555e-05, "loss": 2.1347, "step": 1250 }, { "epoch": 0.22604951560818085, "grad_norm": 8.741484642028809, "learning_rate": 4.775206315034087e-05, "loss": 2.0043, "step": 1260 }, { "epoch": 0.22784355938284895, "grad_norm": 11.853199005126953, "learning_rate": 4.773412271259419e-05, "loss": 2.1624, "step": 1270 }, { "epoch": 0.22963760315751705, "grad_norm": 8.458366394042969, "learning_rate": 4.771618227484751e-05, "loss": 2.0635, "step": 1280 }, { "epoch": 0.23143164693218515, "grad_norm": 9.640998840332031, "learning_rate": 4.7698241837100824e-05, "loss": 1.9351, "step": 1290 }, { "epoch": 0.23322569070685326, "grad_norm": 11.35208511352539, "learning_rate": 4.768030139935415e-05, "loss": 2.0016, "step": 1300 }, { "epoch": 0.23501973448152136, "grad_norm": 12.047490119934082, "learning_rate": 4.7662360961607465e-05, "loss": 1.9982, "step": 1310 }, { "epoch": 0.23681377825618946, "grad_norm": 7.340638637542725, "learning_rate": 4.764442052386078e-05, "loss": 1.9317, "step": 1320 }, { "epoch": 0.23860782203085756, "grad_norm": 9.259571075439453, "learning_rate": 4.762648008611411e-05, "loss": 1.867, "step": 1330 }, { "epoch": 0.24040186580552567, "grad_norm": 10.415928840637207, "learning_rate": 4.7608539648367424e-05, "loss": 1.8926, "step": 1340 }, { "epoch": 0.24219590958019377, "grad_norm": 10.523111343383789, "learning_rate": 4.759059921062074e-05, "loss": 1.9432, "step": 1350 }, { "epoch": 0.24398995335486187, "grad_norm": 10.31205940246582, "learning_rate": 4.7572658772874065e-05, "loss": 2.0679, "step": 1360 }, { "epoch": 0.24578399712952997, "grad_norm": 11.592795372009277, "learning_rate": 4.755471833512738e-05, "loss": 1.8786, "step": 1370 }, { "epoch": 0.24757804090419805, "grad_norm": 13.131718635559082, "learning_rate": 4.75367778973807e-05, "loss": 2.1331, "step": 1380 }, { "epoch": 0.24937208467886615, "grad_norm": 8.448001861572266, "learning_rate": 4.751883745963402e-05, "loss": 2.2519, "step": 1390 }, { "epoch": 0.25116612845353425, "grad_norm": 9.400849342346191, "learning_rate": 4.7500897021887334e-05, "loss": 2.162, "step": 1400 }, { "epoch": 0.2529601722282024, "grad_norm": 9.269533157348633, "learning_rate": 4.748295658414065e-05, "loss": 2.0009, "step": 1410 }, { "epoch": 0.25475421600287046, "grad_norm": 12.824959754943848, "learning_rate": 4.7465016146393976e-05, "loss": 2.1757, "step": 1420 }, { "epoch": 0.2565482597775386, "grad_norm": 6.765061378479004, "learning_rate": 4.744707570864729e-05, "loss": 1.9529, "step": 1430 }, { "epoch": 0.25834230355220666, "grad_norm": 8.876534461975098, "learning_rate": 4.742913527090061e-05, "loss": 1.9964, "step": 1440 }, { "epoch": 0.2601363473268748, "grad_norm": 13.438302040100098, "learning_rate": 4.7411194833153934e-05, "loss": 2.0535, "step": 1450 }, { "epoch": 0.26193039110154287, "grad_norm": 11.55459213256836, "learning_rate": 4.739325439540725e-05, "loss": 2.0917, "step": 1460 }, { "epoch": 0.263724434876211, "grad_norm": 13.218855857849121, "learning_rate": 4.737531395766057e-05, "loss": 1.9388, "step": 1470 }, { "epoch": 0.2655184786508791, "grad_norm": 8.17874526977539, "learning_rate": 4.7357373519913886e-05, "loss": 1.9581, "step": 1480 }, { "epoch": 0.2673125224255472, "grad_norm": 9.68444538116455, "learning_rate": 4.73394330821672e-05, "loss": 2.0179, "step": 1490 }, { "epoch": 0.2691065662002153, "grad_norm": 10.616148948669434, "learning_rate": 4.732149264442053e-05, "loss": 2.024, "step": 1500 }, { "epoch": 0.2709006099748834, "grad_norm": 11.970083236694336, "learning_rate": 4.7303552206673845e-05, "loss": 2.1263, "step": 1510 }, { "epoch": 0.2726946537495515, "grad_norm": 9.173564910888672, "learning_rate": 4.728561176892716e-05, "loss": 1.9028, "step": 1520 }, { "epoch": 0.2744886975242196, "grad_norm": 12.469501495361328, "learning_rate": 4.7267671331180486e-05, "loss": 2.1234, "step": 1530 }, { "epoch": 0.2762827412988877, "grad_norm": 7.576513290405273, "learning_rate": 4.72497308934338e-05, "loss": 1.9832, "step": 1540 }, { "epoch": 0.2780767850735558, "grad_norm": 7.813937187194824, "learning_rate": 4.723179045568712e-05, "loss": 2.0337, "step": 1550 }, { "epoch": 0.2798708288482239, "grad_norm": 8.702999114990234, "learning_rate": 4.7213850017940444e-05, "loss": 1.8968, "step": 1560 }, { "epoch": 0.281664872622892, "grad_norm": 9.221305847167969, "learning_rate": 4.719590958019376e-05, "loss": 1.8972, "step": 1570 }, { "epoch": 0.2834589163975601, "grad_norm": 7.7957892417907715, "learning_rate": 4.717796914244708e-05, "loss": 1.985, "step": 1580 }, { "epoch": 0.2852529601722282, "grad_norm": 15.21849250793457, "learning_rate": 4.7160028704700396e-05, "loss": 1.8525, "step": 1590 }, { "epoch": 0.2870470039468963, "grad_norm": 11.758828163146973, "learning_rate": 4.7142088266953713e-05, "loss": 2.039, "step": 1600 }, { "epoch": 0.2888410477215644, "grad_norm": 9.403512001037598, "learning_rate": 4.712414782920703e-05, "loss": 2.0726, "step": 1610 }, { "epoch": 0.2906350914962325, "grad_norm": 10.458137512207031, "learning_rate": 4.7106207391460355e-05, "loss": 2.1689, "step": 1620 }, { "epoch": 0.2924291352709006, "grad_norm": 7.665112495422363, "learning_rate": 4.708826695371367e-05, "loss": 1.9077, "step": 1630 }, { "epoch": 0.2942231790455687, "grad_norm": 11.904746055603027, "learning_rate": 4.707032651596699e-05, "loss": 2.0425, "step": 1640 }, { "epoch": 0.2960172228202368, "grad_norm": 8.795477867126465, "learning_rate": 4.705238607822031e-05, "loss": 1.9992, "step": 1650 }, { "epoch": 0.2978112665949049, "grad_norm": 7.0219550132751465, "learning_rate": 4.703444564047363e-05, "loss": 1.9143, "step": 1660 }, { "epoch": 0.299605310369573, "grad_norm": 7.745930194854736, "learning_rate": 4.7016505202726955e-05, "loss": 1.8667, "step": 1670 }, { "epoch": 0.3013993541442411, "grad_norm": 7.579682350158691, "learning_rate": 4.699856476498027e-05, "loss": 2.0123, "step": 1680 }, { "epoch": 0.3031933979189092, "grad_norm": 7.5498199462890625, "learning_rate": 4.698062432723358e-05, "loss": 2.1165, "step": 1690 }, { "epoch": 0.30498744169357733, "grad_norm": 8.588129043579102, "learning_rate": 4.6962683889486906e-05, "loss": 2.0662, "step": 1700 }, { "epoch": 0.3067814854682454, "grad_norm": 6.789324760437012, "learning_rate": 4.6944743451740224e-05, "loss": 1.9604, "step": 1710 }, { "epoch": 0.30857552924291354, "grad_norm": 8.46413803100586, "learning_rate": 4.692680301399354e-05, "loss": 2.0078, "step": 1720 }, { "epoch": 0.3103695730175816, "grad_norm": 9.2481689453125, "learning_rate": 4.6908862576246865e-05, "loss": 2.0001, "step": 1730 }, { "epoch": 0.31216361679224974, "grad_norm": 9.034040451049805, "learning_rate": 4.689092213850018e-05, "loss": 2.1038, "step": 1740 }, { "epoch": 0.3139576605669178, "grad_norm": 9.660133361816406, "learning_rate": 4.68729817007535e-05, "loss": 2.0491, "step": 1750 }, { "epoch": 0.31575170434158595, "grad_norm": 8.319841384887695, "learning_rate": 4.6855041263006823e-05, "loss": 1.9269, "step": 1760 }, { "epoch": 0.317545748116254, "grad_norm": 12.393325805664062, "learning_rate": 4.683710082526014e-05, "loss": 2.0983, "step": 1770 }, { "epoch": 0.31933979189092215, "grad_norm": 9.26393985748291, "learning_rate": 4.681916038751346e-05, "loss": 1.8658, "step": 1780 }, { "epoch": 0.32113383566559023, "grad_norm": 10.711555480957031, "learning_rate": 4.6801219949766775e-05, "loss": 2.0091, "step": 1790 }, { "epoch": 0.32292787944025836, "grad_norm": 9.118010520935059, "learning_rate": 4.678327951202009e-05, "loss": 2.119, "step": 1800 }, { "epoch": 0.32472192321492643, "grad_norm": 10.101414680480957, "learning_rate": 4.676533907427341e-05, "loss": 2.1448, "step": 1810 }, { "epoch": 0.32651596698959456, "grad_norm": 11.349935531616211, "learning_rate": 4.6747398636526734e-05, "loss": 2.0042, "step": 1820 }, { "epoch": 0.32831001076426264, "grad_norm": 8.576199531555176, "learning_rate": 4.672945819878005e-05, "loss": 2.0786, "step": 1830 }, { "epoch": 0.33010405453893077, "grad_norm": 8.958368301391602, "learning_rate": 4.671151776103337e-05, "loss": 1.9364, "step": 1840 }, { "epoch": 0.33189809831359884, "grad_norm": 10.209322929382324, "learning_rate": 4.669357732328669e-05, "loss": 2.1192, "step": 1850 }, { "epoch": 0.333692142088267, "grad_norm": 10.517390251159668, "learning_rate": 4.667563688554001e-05, "loss": 1.8833, "step": 1860 }, { "epoch": 0.33548618586293505, "grad_norm": 11.347796440124512, "learning_rate": 4.6657696447793334e-05, "loss": 2.0498, "step": 1870 }, { "epoch": 0.3372802296376032, "grad_norm": 9.35456371307373, "learning_rate": 4.663975601004665e-05, "loss": 2.0158, "step": 1880 }, { "epoch": 0.33907427341227125, "grad_norm": 7.018476486206055, "learning_rate": 4.662181557229997e-05, "loss": 1.9527, "step": 1890 }, { "epoch": 0.3408683171869394, "grad_norm": 13.440537452697754, "learning_rate": 4.6603875134553285e-05, "loss": 1.9492, "step": 1900 }, { "epoch": 0.34266236096160746, "grad_norm": 7.040363788604736, "learning_rate": 4.65859346968066e-05, "loss": 2.085, "step": 1910 }, { "epoch": 0.3444564047362756, "grad_norm": 8.57967472076416, "learning_rate": 4.656799425905992e-05, "loss": 1.9639, "step": 1920 }, { "epoch": 0.34625044851094366, "grad_norm": 7.877899169921875, "learning_rate": 4.6550053821313244e-05, "loss": 1.9315, "step": 1930 }, { "epoch": 0.3480444922856118, "grad_norm": 10.133997917175293, "learning_rate": 4.653211338356656e-05, "loss": 1.9489, "step": 1940 }, { "epoch": 0.34983853606027987, "grad_norm": 10.46199893951416, "learning_rate": 4.651417294581988e-05, "loss": 2.1294, "step": 1950 }, { "epoch": 0.351632579834948, "grad_norm": 10.541289329528809, "learning_rate": 4.64962325080732e-05, "loss": 2.0334, "step": 1960 }, { "epoch": 0.3534266236096161, "grad_norm": 10.026679039001465, "learning_rate": 4.647829207032652e-05, "loss": 1.9669, "step": 1970 }, { "epoch": 0.35522066738428415, "grad_norm": 9.2806978225708, "learning_rate": 4.646035163257984e-05, "loss": 2.0065, "step": 1980 }, { "epoch": 0.3570147111589523, "grad_norm": 11.332128524780273, "learning_rate": 4.6442411194833154e-05, "loss": 2.1402, "step": 1990 }, { "epoch": 0.35880875493362036, "grad_norm": 9.25808048248291, "learning_rate": 4.642447075708647e-05, "loss": 2.1242, "step": 2000 }, { "epoch": 0.3606027987082885, "grad_norm": 9.891722679138184, "learning_rate": 4.640653031933979e-05, "loss": 2.0954, "step": 2010 }, { "epoch": 0.36239684248295656, "grad_norm": 8.027229309082031, "learning_rate": 4.638858988159311e-05, "loss": 1.94, "step": 2020 }, { "epoch": 0.3641908862576247, "grad_norm": 7.155013084411621, "learning_rate": 4.637064944384643e-05, "loss": 2.0099, "step": 2030 }, { "epoch": 0.36598493003229277, "grad_norm": 7.851609230041504, "learning_rate": 4.6352709006099754e-05, "loss": 2.0819, "step": 2040 }, { "epoch": 0.3677789738069609, "grad_norm": 9.718694686889648, "learning_rate": 4.633476856835307e-05, "loss": 2.0265, "step": 2050 }, { "epoch": 0.36957301758162897, "grad_norm": 7.794171333312988, "learning_rate": 4.631682813060639e-05, "loss": 1.9589, "step": 2060 }, { "epoch": 0.3713670613562971, "grad_norm": 9.896645545959473, "learning_rate": 4.629888769285971e-05, "loss": 1.8671, "step": 2070 }, { "epoch": 0.3731611051309652, "grad_norm": 8.257375717163086, "learning_rate": 4.628094725511303e-05, "loss": 2.0167, "step": 2080 }, { "epoch": 0.3749551489056333, "grad_norm": 9.882346153259277, "learning_rate": 4.626300681736635e-05, "loss": 2.079, "step": 2090 }, { "epoch": 0.3767491926803014, "grad_norm": 8.422935485839844, "learning_rate": 4.6245066379619664e-05, "loss": 1.998, "step": 2100 }, { "epoch": 0.3785432364549695, "grad_norm": 9.090377807617188, "learning_rate": 4.622712594187298e-05, "loss": 1.8588, "step": 2110 }, { "epoch": 0.3803372802296376, "grad_norm": 9.11992359161377, "learning_rate": 4.62091855041263e-05, "loss": 1.9268, "step": 2120 }, { "epoch": 0.3821313240043057, "grad_norm": 7.935147285461426, "learning_rate": 4.619124506637962e-05, "loss": 2.1426, "step": 2130 }, { "epoch": 0.3839253677789738, "grad_norm": 6.065644264221191, "learning_rate": 4.617330462863294e-05, "loss": 1.9426, "step": 2140 }, { "epoch": 0.3857194115536419, "grad_norm": 7.789048194885254, "learning_rate": 4.615536419088626e-05, "loss": 1.9607, "step": 2150 }, { "epoch": 0.38751345532831, "grad_norm": 10.809876441955566, "learning_rate": 4.613742375313958e-05, "loss": 2.0248, "step": 2160 }, { "epoch": 0.3893074991029781, "grad_norm": 7.871123790740967, "learning_rate": 4.61194833153929e-05, "loss": 2.0491, "step": 2170 }, { "epoch": 0.3911015428776462, "grad_norm": 10.08164119720459, "learning_rate": 4.6101542877646216e-05, "loss": 1.8583, "step": 2180 }, { "epoch": 0.39289558665231433, "grad_norm": 8.326801300048828, "learning_rate": 4.608360243989954e-05, "loss": 2.0879, "step": 2190 }, { "epoch": 0.3946896304269824, "grad_norm": 9.958331108093262, "learning_rate": 4.606566200215286e-05, "loss": 1.9132, "step": 2200 }, { "epoch": 0.39648367420165054, "grad_norm": 8.36109733581543, "learning_rate": 4.604772156440617e-05, "loss": 1.9548, "step": 2210 }, { "epoch": 0.3982777179763186, "grad_norm": 8.014139175415039, "learning_rate": 4.602978112665949e-05, "loss": 1.8601, "step": 2220 }, { "epoch": 0.40007176175098674, "grad_norm": 8.810429573059082, "learning_rate": 4.601184068891281e-05, "loss": 2.0101, "step": 2230 }, { "epoch": 0.4018658055256548, "grad_norm": 9.849298477172852, "learning_rate": 4.599390025116613e-05, "loss": 1.9178, "step": 2240 }, { "epoch": 0.40365984930032295, "grad_norm": 7.431755065917969, "learning_rate": 4.597595981341945e-05, "loss": 1.9154, "step": 2250 }, { "epoch": 0.405453893074991, "grad_norm": 10.968128204345703, "learning_rate": 4.595801937567277e-05, "loss": 2.0084, "step": 2260 }, { "epoch": 0.40724793684965915, "grad_norm": 8.786876678466797, "learning_rate": 4.594007893792609e-05, "loss": 2.1216, "step": 2270 }, { "epoch": 0.40904198062432723, "grad_norm": 8.992511749267578, "learning_rate": 4.592213850017941e-05, "loss": 1.9462, "step": 2280 }, { "epoch": 0.41083602439899536, "grad_norm": 8.601082801818848, "learning_rate": 4.5904198062432726e-05, "loss": 1.988, "step": 2290 }, { "epoch": 0.41263006817366343, "grad_norm": 9.053206443786621, "learning_rate": 4.5886257624686044e-05, "loss": 1.8687, "step": 2300 }, { "epoch": 0.41442411194833156, "grad_norm": 10.55015754699707, "learning_rate": 4.586831718693936e-05, "loss": 1.9853, "step": 2310 }, { "epoch": 0.41621815572299964, "grad_norm": 7.392729759216309, "learning_rate": 4.585037674919268e-05, "loss": 2.0305, "step": 2320 }, { "epoch": 0.41801219949766777, "grad_norm": 8.890578269958496, "learning_rate": 4.5832436311446e-05, "loss": 2.1245, "step": 2330 }, { "epoch": 0.41980624327233584, "grad_norm": 8.476337432861328, "learning_rate": 4.581449587369932e-05, "loss": 1.8404, "step": 2340 }, { "epoch": 0.421600287047004, "grad_norm": 11.671123504638672, "learning_rate": 4.5796555435952637e-05, "loss": 2.0296, "step": 2350 }, { "epoch": 0.42339433082167205, "grad_norm": 9.152827262878418, "learning_rate": 4.577861499820596e-05, "loss": 1.9146, "step": 2360 }, { "epoch": 0.4251883745963401, "grad_norm": 7.549522876739502, "learning_rate": 4.576067456045928e-05, "loss": 2.0178, "step": 2370 }, { "epoch": 0.42698241837100825, "grad_norm": 11.978499412536621, "learning_rate": 4.5742734122712595e-05, "loss": 1.7915, "step": 2380 }, { "epoch": 0.42877646214567633, "grad_norm": 10.303351402282715, "learning_rate": 4.572479368496592e-05, "loss": 1.8197, "step": 2390 }, { "epoch": 0.43057050592034446, "grad_norm": 9.132424354553223, "learning_rate": 4.5706853247219236e-05, "loss": 2.1076, "step": 2400 }, { "epoch": 0.43236454969501253, "grad_norm": 9.768096923828125, "learning_rate": 4.5688912809472554e-05, "loss": 1.8818, "step": 2410 }, { "epoch": 0.43415859346968066, "grad_norm": 7.977086067199707, "learning_rate": 4.567097237172587e-05, "loss": 1.9721, "step": 2420 }, { "epoch": 0.43595263724434874, "grad_norm": 10.111530303955078, "learning_rate": 4.565303193397919e-05, "loss": 1.9108, "step": 2430 }, { "epoch": 0.43774668101901687, "grad_norm": 8.572367668151855, "learning_rate": 4.563509149623251e-05, "loss": 1.8878, "step": 2440 }, { "epoch": 0.43954072479368494, "grad_norm": 15.499300956726074, "learning_rate": 4.561715105848583e-05, "loss": 2.0646, "step": 2450 }, { "epoch": 0.4413347685683531, "grad_norm": 8.507719993591309, "learning_rate": 4.559921062073915e-05, "loss": 2.0785, "step": 2460 }, { "epoch": 0.44312881234302115, "grad_norm": 8.560482025146484, "learning_rate": 4.558127018299247e-05, "loss": 1.919, "step": 2470 }, { "epoch": 0.4449228561176893, "grad_norm": 8.793256759643555, "learning_rate": 4.556332974524579e-05, "loss": 2.0275, "step": 2480 }, { "epoch": 0.44671689989235736, "grad_norm": 6.947265625, "learning_rate": 4.5545389307499105e-05, "loss": 1.8708, "step": 2490 }, { "epoch": 0.4485109436670255, "grad_norm": 8.579648971557617, "learning_rate": 4.552744886975242e-05, "loss": 2.0002, "step": 2500 }, { "epoch": 0.45030498744169356, "grad_norm": 7.623202323913574, "learning_rate": 4.550950843200574e-05, "loss": 1.9444, "step": 2510 }, { "epoch": 0.4520990312163617, "grad_norm": 8.441667556762695, "learning_rate": 4.549156799425906e-05, "loss": 2.0275, "step": 2520 }, { "epoch": 0.45389307499102977, "grad_norm": 9.18626594543457, "learning_rate": 4.547362755651238e-05, "loss": 1.9512, "step": 2530 }, { "epoch": 0.4556871187656979, "grad_norm": 8.11828899383545, "learning_rate": 4.54556871187657e-05, "loss": 1.9609, "step": 2540 }, { "epoch": 0.45748116254036597, "grad_norm": 7.3665852546691895, "learning_rate": 4.5437746681019016e-05, "loss": 1.9721, "step": 2550 }, { "epoch": 0.4592752063150341, "grad_norm": 8.686524391174316, "learning_rate": 4.541980624327234e-05, "loss": 2.0039, "step": 2560 }, { "epoch": 0.4610692500897022, "grad_norm": 8.56877326965332, "learning_rate": 4.540186580552566e-05, "loss": 2.0762, "step": 2570 }, { "epoch": 0.4628632938643703, "grad_norm": 9.964433670043945, "learning_rate": 4.5383925367778974e-05, "loss": 1.9773, "step": 2580 }, { "epoch": 0.4646573376390384, "grad_norm": 8.93453311920166, "learning_rate": 4.53659849300323e-05, "loss": 1.9763, "step": 2590 }, { "epoch": 0.4664513814137065, "grad_norm": 6.911281108856201, "learning_rate": 4.5348044492285615e-05, "loss": 1.8979, "step": 2600 }, { "epoch": 0.4682454251883746, "grad_norm": 6.9323410987854, "learning_rate": 4.533010405453893e-05, "loss": 1.9935, "step": 2610 }, { "epoch": 0.4700394689630427, "grad_norm": 7.738839626312256, "learning_rate": 4.531216361679225e-05, "loss": 1.9254, "step": 2620 }, { "epoch": 0.4718335127377108, "grad_norm": 6.664251804351807, "learning_rate": 4.529422317904557e-05, "loss": 1.9573, "step": 2630 }, { "epoch": 0.4736275565123789, "grad_norm": 8.574616432189941, "learning_rate": 4.527628274129889e-05, "loss": 1.9251, "step": 2640 }, { "epoch": 0.475421600287047, "grad_norm": 7.2672834396362305, "learning_rate": 4.525834230355221e-05, "loss": 1.8782, "step": 2650 }, { "epoch": 0.4772156440617151, "grad_norm": 6.810856342315674, "learning_rate": 4.5240401865805526e-05, "loss": 1.9968, "step": 2660 }, { "epoch": 0.4790096878363832, "grad_norm": 16.551050186157227, "learning_rate": 4.522246142805885e-05, "loss": 1.9016, "step": 2670 }, { "epoch": 0.48080373161105133, "grad_norm": 9.01229190826416, "learning_rate": 4.520452099031217e-05, "loss": 1.9715, "step": 2680 }, { "epoch": 0.4825977753857194, "grad_norm": 9.2802152633667, "learning_rate": 4.5186580552565484e-05, "loss": 2.0513, "step": 2690 }, { "epoch": 0.48439181916038754, "grad_norm": 7.945450782775879, "learning_rate": 4.516864011481881e-05, "loss": 1.849, "step": 2700 }, { "epoch": 0.4861858629350556, "grad_norm": 9.4662446975708, "learning_rate": 4.5150699677072126e-05, "loss": 1.9349, "step": 2710 }, { "epoch": 0.48797990670972374, "grad_norm": 7.22721004486084, "learning_rate": 4.5132759239325436e-05, "loss": 2.0256, "step": 2720 }, { "epoch": 0.4897739504843918, "grad_norm": 8.505002975463867, "learning_rate": 4.511481880157876e-05, "loss": 2.0187, "step": 2730 }, { "epoch": 0.49156799425905995, "grad_norm": 8.256681442260742, "learning_rate": 4.509687836383208e-05, "loss": 1.8934, "step": 2740 }, { "epoch": 0.493362038033728, "grad_norm": 13.187761306762695, "learning_rate": 4.5078937926085395e-05, "loss": 1.9564, "step": 2750 }, { "epoch": 0.4951560818083961, "grad_norm": 9.577531814575195, "learning_rate": 4.506099748833872e-05, "loss": 1.9369, "step": 2760 }, { "epoch": 0.49695012558306423, "grad_norm": 8.296804428100586, "learning_rate": 4.5043057050592036e-05, "loss": 2.043, "step": 2770 }, { "epoch": 0.4987441693577323, "grad_norm": 7.145248889923096, "learning_rate": 4.502511661284536e-05, "loss": 1.7847, "step": 2780 }, { "epoch": 0.5005382131324004, "grad_norm": 10.56977367401123, "learning_rate": 4.500717617509868e-05, "loss": 1.8364, "step": 2790 }, { "epoch": 0.5023322569070685, "grad_norm": 9.34067440032959, "learning_rate": 4.4989235737351995e-05, "loss": 1.9133, "step": 2800 }, { "epoch": 0.5041263006817366, "grad_norm": 8.067873001098633, "learning_rate": 4.497129529960531e-05, "loss": 1.8836, "step": 2810 }, { "epoch": 0.5059203444564048, "grad_norm": 8.651354789733887, "learning_rate": 4.495335486185863e-05, "loss": 1.8979, "step": 2820 }, { "epoch": 0.5077143882310728, "grad_norm": 6.594571590423584, "learning_rate": 4.4935414424111946e-05, "loss": 1.9502, "step": 2830 }, { "epoch": 0.5095084320057409, "grad_norm": 10.328874588012695, "learning_rate": 4.491747398636527e-05, "loss": 2.0826, "step": 2840 }, { "epoch": 0.511302475780409, "grad_norm": 6.956106185913086, "learning_rate": 4.489953354861859e-05, "loss": 1.9671, "step": 2850 }, { "epoch": 0.5130965195550772, "grad_norm": 7.3767313957214355, "learning_rate": 4.4881593110871905e-05, "loss": 1.902, "step": 2860 }, { "epoch": 0.5148905633297453, "grad_norm": 11.275064468383789, "learning_rate": 4.486365267312523e-05, "loss": 2.0932, "step": 2870 }, { "epoch": 0.5166846071044133, "grad_norm": 7.849559783935547, "learning_rate": 4.4845712235378546e-05, "loss": 1.9519, "step": 2880 }, { "epoch": 0.5184786508790814, "grad_norm": 7.759171962738037, "learning_rate": 4.4827771797631863e-05, "loss": 1.853, "step": 2890 }, { "epoch": 0.5202726946537496, "grad_norm": 12.3310546875, "learning_rate": 4.480983135988519e-05, "loss": 2.017, "step": 2900 }, { "epoch": 0.5220667384284177, "grad_norm": 7.477339267730713, "learning_rate": 4.4791890922138505e-05, "loss": 1.8495, "step": 2910 }, { "epoch": 0.5238607822030857, "grad_norm": 8.375692367553711, "learning_rate": 4.477395048439182e-05, "loss": 1.9654, "step": 2920 }, { "epoch": 0.5256548259777538, "grad_norm": 7.368261337280273, "learning_rate": 4.475601004664514e-05, "loss": 1.8585, "step": 2930 }, { "epoch": 0.527448869752422, "grad_norm": 7.570740699768066, "learning_rate": 4.4738069608898456e-05, "loss": 1.965, "step": 2940 }, { "epoch": 0.5292429135270901, "grad_norm": 7.81984281539917, "learning_rate": 4.4720129171151774e-05, "loss": 2.0262, "step": 2950 }, { "epoch": 0.5310369573017582, "grad_norm": 7.456051349639893, "learning_rate": 4.47021887334051e-05, "loss": 1.8934, "step": 2960 }, { "epoch": 0.5328310010764262, "grad_norm": 8.479120254516602, "learning_rate": 4.4684248295658415e-05, "loss": 1.9117, "step": 2970 }, { "epoch": 0.5346250448510944, "grad_norm": 8.73019027709961, "learning_rate": 4.466630785791174e-05, "loss": 2.0238, "step": 2980 }, { "epoch": 0.5364190886257625, "grad_norm": 8.31013298034668, "learning_rate": 4.4648367420165056e-05, "loss": 1.9404, "step": 2990 }, { "epoch": 0.5382131324004306, "grad_norm": 6.616645812988281, "learning_rate": 4.4630426982418374e-05, "loss": 1.9167, "step": 3000 }, { "epoch": 0.5400071761750986, "grad_norm": 7.4681172370910645, "learning_rate": 4.46124865446717e-05, "loss": 2.0548, "step": 3010 }, { "epoch": 0.5418012199497668, "grad_norm": 11.563379287719727, "learning_rate": 4.459454610692501e-05, "loss": 2.0929, "step": 3020 }, { "epoch": 0.5435952637244349, "grad_norm": 7.389353275299072, "learning_rate": 4.4576605669178325e-05, "loss": 1.7729, "step": 3030 }, { "epoch": 0.545389307499103, "grad_norm": 8.560032844543457, "learning_rate": 4.455866523143165e-05, "loss": 1.8803, "step": 3040 }, { "epoch": 0.547183351273771, "grad_norm": 8.208505630493164, "learning_rate": 4.454072479368497e-05, "loss": 1.8715, "step": 3050 }, { "epoch": 0.5489773950484392, "grad_norm": 8.2820463180542, "learning_rate": 4.4522784355938284e-05, "loss": 2.0147, "step": 3060 }, { "epoch": 0.5507714388231073, "grad_norm": 7.4810285568237305, "learning_rate": 4.450484391819161e-05, "loss": 1.9301, "step": 3070 }, { "epoch": 0.5525654825977754, "grad_norm": 7.442780017852783, "learning_rate": 4.4486903480444925e-05, "loss": 2.0132, "step": 3080 }, { "epoch": 0.5543595263724435, "grad_norm": 7.857019424438477, "learning_rate": 4.446896304269824e-05, "loss": 2.1327, "step": 3090 }, { "epoch": 0.5561535701471116, "grad_norm": 8.195714950561523, "learning_rate": 4.4451022604951567e-05, "loss": 1.9133, "step": 3100 }, { "epoch": 0.5579476139217797, "grad_norm": 9.424678802490234, "learning_rate": 4.4433082167204884e-05, "loss": 1.8079, "step": 3110 }, { "epoch": 0.5597416576964478, "grad_norm": 7.7518768310546875, "learning_rate": 4.44151417294582e-05, "loss": 2.0345, "step": 3120 }, { "epoch": 0.5615357014711159, "grad_norm": 8.170982360839844, "learning_rate": 4.439720129171152e-05, "loss": 2.089, "step": 3130 }, { "epoch": 0.563329745245784, "grad_norm": 7.443562030792236, "learning_rate": 4.4379260853964836e-05, "loss": 1.805, "step": 3140 }, { "epoch": 0.5651237890204521, "grad_norm": 8.59354019165039, "learning_rate": 4.436132041621816e-05, "loss": 1.9759, "step": 3150 }, { "epoch": 0.5669178327951202, "grad_norm": 13.471951484680176, "learning_rate": 4.434337997847148e-05, "loss": 1.7935, "step": 3160 }, { "epoch": 0.5687118765697883, "grad_norm": 9.660313606262207, "learning_rate": 4.4325439540724794e-05, "loss": 1.9807, "step": 3170 }, { "epoch": 0.5705059203444564, "grad_norm": 6.917825222015381, "learning_rate": 4.430749910297812e-05, "loss": 2.0491, "step": 3180 }, { "epoch": 0.5722999641191245, "grad_norm": 9.734037399291992, "learning_rate": 4.4289558665231435e-05, "loss": 1.8836, "step": 3190 }, { "epoch": 0.5740940078937926, "grad_norm": 7.515590190887451, "learning_rate": 4.427161822748475e-05, "loss": 1.9391, "step": 3200 }, { "epoch": 0.5758880516684607, "grad_norm": 7.2530670166015625, "learning_rate": 4.425367778973808e-05, "loss": 1.8615, "step": 3210 }, { "epoch": 0.5776820954431288, "grad_norm": 8.29240894317627, "learning_rate": 4.4235737351991394e-05, "loss": 2.0954, "step": 3220 }, { "epoch": 0.579476139217797, "grad_norm": 8.519623756408691, "learning_rate": 4.421779691424471e-05, "loss": 1.983, "step": 3230 }, { "epoch": 0.581270182992465, "grad_norm": 7.667181015014648, "learning_rate": 4.419985647649803e-05, "loss": 1.9309, "step": 3240 }, { "epoch": 0.5830642267671331, "grad_norm": 9.281998634338379, "learning_rate": 4.4181916038751346e-05, "loss": 1.9057, "step": 3250 }, { "epoch": 0.5848582705418012, "grad_norm": 7.8026123046875, "learning_rate": 4.416397560100466e-05, "loss": 1.8834, "step": 3260 }, { "epoch": 0.5866523143164694, "grad_norm": 7.448451042175293, "learning_rate": 4.414603516325799e-05, "loss": 1.8942, "step": 3270 }, { "epoch": 0.5884463580911374, "grad_norm": 8.204652786254883, "learning_rate": 4.4128094725511304e-05, "loss": 2.0565, "step": 3280 }, { "epoch": 0.5902404018658055, "grad_norm": 8.598773002624512, "learning_rate": 4.411015428776462e-05, "loss": 1.9708, "step": 3290 }, { "epoch": 0.5920344456404736, "grad_norm": 7.167200088500977, "learning_rate": 4.4092213850017946e-05, "loss": 1.8534, "step": 3300 }, { "epoch": 0.5938284894151418, "grad_norm": 11.437542915344238, "learning_rate": 4.407427341227126e-05, "loss": 2.0349, "step": 3310 }, { "epoch": 0.5956225331898098, "grad_norm": 8.82004451751709, "learning_rate": 4.405633297452458e-05, "loss": 1.9427, "step": 3320 }, { "epoch": 0.5974165769644779, "grad_norm": 6.192866802215576, "learning_rate": 4.40383925367779e-05, "loss": 1.9727, "step": 3330 }, { "epoch": 0.599210620739146, "grad_norm": 7.089540481567383, "learning_rate": 4.4020452099031215e-05, "loss": 1.9497, "step": 3340 }, { "epoch": 0.6010046645138142, "grad_norm": 8.135286331176758, "learning_rate": 4.400251166128454e-05, "loss": 1.8405, "step": 3350 }, { "epoch": 0.6027987082884823, "grad_norm": 8.762147903442383, "learning_rate": 4.3984571223537856e-05, "loss": 1.9011, "step": 3360 }, { "epoch": 0.6045927520631503, "grad_norm": 9.658001899719238, "learning_rate": 4.396663078579117e-05, "loss": 1.8995, "step": 3370 }, { "epoch": 0.6063867958378184, "grad_norm": 7.004101276397705, "learning_rate": 4.39486903480445e-05, "loss": 1.8548, "step": 3380 }, { "epoch": 0.6081808396124866, "grad_norm": 7.552975177764893, "learning_rate": 4.3930749910297814e-05, "loss": 1.9911, "step": 3390 }, { "epoch": 0.6099748833871547, "grad_norm": 7.0711774826049805, "learning_rate": 4.391280947255113e-05, "loss": 1.8611, "step": 3400 }, { "epoch": 0.6117689271618227, "grad_norm": 9.78061580657959, "learning_rate": 4.3894869034804456e-05, "loss": 1.9661, "step": 3410 }, { "epoch": 0.6135629709364908, "grad_norm": 8.065385818481445, "learning_rate": 4.387692859705777e-05, "loss": 1.975, "step": 3420 }, { "epoch": 0.615357014711159, "grad_norm": 9.28231430053711, "learning_rate": 4.385898815931109e-05, "loss": 1.9083, "step": 3430 }, { "epoch": 0.6171510584858271, "grad_norm": 8.710234642028809, "learning_rate": 4.384104772156441e-05, "loss": 2.0402, "step": 3440 }, { "epoch": 0.6189451022604952, "grad_norm": 8.679749488830566, "learning_rate": 4.3823107283817725e-05, "loss": 1.9162, "step": 3450 }, { "epoch": 0.6207391460351632, "grad_norm": 7.5028228759765625, "learning_rate": 4.380516684607104e-05, "loss": 2.0918, "step": 3460 }, { "epoch": 0.6225331898098314, "grad_norm": 8.817152976989746, "learning_rate": 4.3787226408324366e-05, "loss": 1.9089, "step": 3470 }, { "epoch": 0.6243272335844995, "grad_norm": 7.720176696777344, "learning_rate": 4.376928597057768e-05, "loss": 1.8395, "step": 3480 }, { "epoch": 0.6261212773591676, "grad_norm": 8.804288864135742, "learning_rate": 4.3751345532831e-05, "loss": 2.0384, "step": 3490 }, { "epoch": 0.6279153211338356, "grad_norm": 6.383409023284912, "learning_rate": 4.3733405095084325e-05, "loss": 1.917, "step": 3500 }, { "epoch": 0.6297093649085038, "grad_norm": 7.6271233558654785, "learning_rate": 4.371546465733764e-05, "loss": 1.884, "step": 3510 }, { "epoch": 0.6315034086831719, "grad_norm": 7.319206237792969, "learning_rate": 4.3697524219590966e-05, "loss": 2.0081, "step": 3520 }, { "epoch": 0.63329745245784, "grad_norm": 7.624337196350098, "learning_rate": 4.367958378184428e-05, "loss": 1.9573, "step": 3530 }, { "epoch": 0.635091496232508, "grad_norm": 8.105642318725586, "learning_rate": 4.3661643344097594e-05, "loss": 1.9264, "step": 3540 }, { "epoch": 0.6368855400071761, "grad_norm": 6.624615669250488, "learning_rate": 4.364370290635092e-05, "loss": 1.9273, "step": 3550 }, { "epoch": 0.6386795837818443, "grad_norm": 8.527731895446777, "learning_rate": 4.3625762468604235e-05, "loss": 1.8837, "step": 3560 }, { "epoch": 0.6404736275565124, "grad_norm": 8.374712944030762, "learning_rate": 4.360782203085755e-05, "loss": 1.9182, "step": 3570 }, { "epoch": 0.6422676713311805, "grad_norm": 6.726866245269775, "learning_rate": 4.3589881593110876e-05, "loss": 1.7699, "step": 3580 }, { "epoch": 0.6440617151058485, "grad_norm": 10.059771537780762, "learning_rate": 4.3571941155364194e-05, "loss": 1.8803, "step": 3590 }, { "epoch": 0.6458557588805167, "grad_norm": 6.3898024559021, "learning_rate": 4.355400071761751e-05, "loss": 1.9215, "step": 3600 }, { "epoch": 0.6476498026551848, "grad_norm": 9.93798542022705, "learning_rate": 4.3536060279870835e-05, "loss": 1.9015, "step": 3610 }, { "epoch": 0.6494438464298529, "grad_norm": 8.230923652648926, "learning_rate": 4.351811984212415e-05, "loss": 1.8842, "step": 3620 }, { "epoch": 0.6512378902045209, "grad_norm": 8.391541481018066, "learning_rate": 4.350017940437747e-05, "loss": 2.0254, "step": 3630 }, { "epoch": 0.6530319339791891, "grad_norm": 8.703720092773438, "learning_rate": 4.3482238966630787e-05, "loss": 1.9441, "step": 3640 }, { "epoch": 0.6548259777538572, "grad_norm": 8.434385299682617, "learning_rate": 4.3464298528884104e-05, "loss": 2.1388, "step": 3650 }, { "epoch": 0.6566200215285253, "grad_norm": 9.658421516418457, "learning_rate": 4.344635809113742e-05, "loss": 2.1284, "step": 3660 }, { "epoch": 0.6584140653031934, "grad_norm": 9.099438667297363, "learning_rate": 4.3428417653390745e-05, "loss": 1.9191, "step": 3670 }, { "epoch": 0.6602081090778615, "grad_norm": 5.330417633056641, "learning_rate": 4.341047721564406e-05, "loss": 1.7558, "step": 3680 }, { "epoch": 0.6620021528525296, "grad_norm": 8.256141662597656, "learning_rate": 4.339253677789738e-05, "loss": 1.904, "step": 3690 }, { "epoch": 0.6637961966271977, "grad_norm": 7.978524208068848, "learning_rate": 4.3374596340150704e-05, "loss": 1.9706, "step": 3700 }, { "epoch": 0.6655902404018658, "grad_norm": 7.254574298858643, "learning_rate": 4.335665590240402e-05, "loss": 1.8038, "step": 3710 }, { "epoch": 0.667384284176534, "grad_norm": 6.119344711303711, "learning_rate": 4.3338715464657345e-05, "loss": 2.0935, "step": 3720 }, { "epoch": 0.669178327951202, "grad_norm": 10.237481117248535, "learning_rate": 4.332077502691066e-05, "loss": 1.9197, "step": 3730 }, { "epoch": 0.6709723717258701, "grad_norm": 8.238425254821777, "learning_rate": 4.330283458916398e-05, "loss": 1.8335, "step": 3740 }, { "epoch": 0.6727664155005382, "grad_norm": 10.77346420288086, "learning_rate": 4.32848941514173e-05, "loss": 1.9849, "step": 3750 }, { "epoch": 0.6745604592752064, "grad_norm": 5.878365993499756, "learning_rate": 4.3266953713670614e-05, "loss": 1.8184, "step": 3760 }, { "epoch": 0.6763545030498744, "grad_norm": 5.962438583374023, "learning_rate": 4.324901327592393e-05, "loss": 1.942, "step": 3770 }, { "epoch": 0.6781485468245425, "grad_norm": 7.118070602416992, "learning_rate": 4.3231072838177255e-05, "loss": 1.8888, "step": 3780 }, { "epoch": 0.6799425905992106, "grad_norm": 7.19279146194458, "learning_rate": 4.321313240043057e-05, "loss": 1.9875, "step": 3790 }, { "epoch": 0.6817366343738788, "grad_norm": 9.004045486450195, "learning_rate": 4.319519196268389e-05, "loss": 1.8713, "step": 3800 }, { "epoch": 0.6835306781485468, "grad_norm": 6.644644737243652, "learning_rate": 4.3177251524937214e-05, "loss": 1.8883, "step": 3810 }, { "epoch": 0.6853247219232149, "grad_norm": 7.53091287612915, "learning_rate": 4.315931108719053e-05, "loss": 1.9133, "step": 3820 }, { "epoch": 0.687118765697883, "grad_norm": 7.283283710479736, "learning_rate": 4.314137064944385e-05, "loss": 1.8786, "step": 3830 }, { "epoch": 0.6889128094725512, "grad_norm": 8.955952644348145, "learning_rate": 4.3123430211697166e-05, "loss": 1.9249, "step": 3840 }, { "epoch": 0.6907068532472193, "grad_norm": 6.595334053039551, "learning_rate": 4.310548977395048e-05, "loss": 1.8995, "step": 3850 }, { "epoch": 0.6925008970218873, "grad_norm": 6.214903831481934, "learning_rate": 4.30875493362038e-05, "loss": 1.8988, "step": 3860 }, { "epoch": 0.6942949407965554, "grad_norm": 5.9941182136535645, "learning_rate": 4.3069608898457124e-05, "loss": 1.8008, "step": 3870 }, { "epoch": 0.6960889845712236, "grad_norm": 9.003954887390137, "learning_rate": 4.305166846071044e-05, "loss": 2.0354, "step": 3880 }, { "epoch": 0.6978830283458917, "grad_norm": 7.329159259796143, "learning_rate": 4.303372802296376e-05, "loss": 2.0086, "step": 3890 }, { "epoch": 0.6996770721205597, "grad_norm": 7.872637748718262, "learning_rate": 4.301578758521708e-05, "loss": 1.9595, "step": 3900 }, { "epoch": 0.7014711158952278, "grad_norm": 10.439992904663086, "learning_rate": 4.29978471474704e-05, "loss": 1.9187, "step": 3910 }, { "epoch": 0.703265159669896, "grad_norm": 10.078546524047852, "learning_rate": 4.2979906709723724e-05, "loss": 1.9979, "step": 3920 }, { "epoch": 0.7050592034445641, "grad_norm": 8.9290771484375, "learning_rate": 4.296196627197704e-05, "loss": 1.9665, "step": 3930 }, { "epoch": 0.7068532472192322, "grad_norm": 8.043295860290527, "learning_rate": 4.294402583423036e-05, "loss": 2.0803, "step": 3940 }, { "epoch": 0.7086472909939002, "grad_norm": 7.653200149536133, "learning_rate": 4.2926085396483676e-05, "loss": 2.1244, "step": 3950 }, { "epoch": 0.7104413347685683, "grad_norm": 12.027050971984863, "learning_rate": 4.290814495873699e-05, "loss": 2.0046, "step": 3960 }, { "epoch": 0.7122353785432365, "grad_norm": 6.169637680053711, "learning_rate": 4.289020452099031e-05, "loss": 1.9957, "step": 3970 }, { "epoch": 0.7140294223179046, "grad_norm": 8.492105484008789, "learning_rate": 4.2872264083243634e-05, "loss": 1.9959, "step": 3980 }, { "epoch": 0.7158234660925726, "grad_norm": 7.080846309661865, "learning_rate": 4.285432364549695e-05, "loss": 1.9098, "step": 3990 }, { "epoch": 0.7176175098672407, "grad_norm": 7.8670454025268555, "learning_rate": 4.283638320775027e-05, "loss": 1.8557, "step": 4000 }, { "epoch": 0.7194115536419089, "grad_norm": 8.910687446594238, "learning_rate": 4.281844277000359e-05, "loss": 2.0108, "step": 4010 }, { "epoch": 0.721205597416577, "grad_norm": 7.866197109222412, "learning_rate": 4.280050233225691e-05, "loss": 1.9711, "step": 4020 }, { "epoch": 0.722999641191245, "grad_norm": 7.0763044357299805, "learning_rate": 4.278256189451023e-05, "loss": 1.9565, "step": 4030 }, { "epoch": 0.7247936849659131, "grad_norm": 10.510825157165527, "learning_rate": 4.276462145676355e-05, "loss": 1.9513, "step": 4040 }, { "epoch": 0.7265877287405813, "grad_norm": 6.813848495483398, "learning_rate": 4.274668101901687e-05, "loss": 1.9841, "step": 4050 }, { "epoch": 0.7283817725152494, "grad_norm": 8.661791801452637, "learning_rate": 4.272874058127018e-05, "loss": 1.9382, "step": 4060 }, { "epoch": 0.7301758162899175, "grad_norm": 8.576088905334473, "learning_rate": 4.27108001435235e-05, "loss": 1.7463, "step": 4070 }, { "epoch": 0.7319698600645855, "grad_norm": 9.310657501220703, "learning_rate": 4.269285970577682e-05, "loss": 1.9832, "step": 4080 }, { "epoch": 0.7337639038392537, "grad_norm": 8.533422470092773, "learning_rate": 4.2674919268030145e-05, "loss": 2.0184, "step": 4090 }, { "epoch": 0.7355579476139218, "grad_norm": 6.781728267669678, "learning_rate": 4.265697883028346e-05, "loss": 1.9384, "step": 4100 }, { "epoch": 0.7373519913885899, "grad_norm": 7.908256530761719, "learning_rate": 4.263903839253678e-05, "loss": 1.9602, "step": 4110 }, { "epoch": 0.7391460351632579, "grad_norm": 6.28724479675293, "learning_rate": 4.26210979547901e-05, "loss": 2.0066, "step": 4120 }, { "epoch": 0.7409400789379261, "grad_norm": 6.198331832885742, "learning_rate": 4.260315751704342e-05, "loss": 2.132, "step": 4130 }, { "epoch": 0.7427341227125942, "grad_norm": 7.816977024078369, "learning_rate": 4.258521707929674e-05, "loss": 1.9536, "step": 4140 }, { "epoch": 0.7445281664872623, "grad_norm": 7.433613300323486, "learning_rate": 4.2567276641550055e-05, "loss": 1.885, "step": 4150 }, { "epoch": 0.7463222102619304, "grad_norm": 6.046330451965332, "learning_rate": 4.254933620380337e-05, "loss": 1.7975, "step": 4160 }, { "epoch": 0.7481162540365985, "grad_norm": 8.465211868286133, "learning_rate": 4.253139576605669e-05, "loss": 2.0454, "step": 4170 }, { "epoch": 0.7499102978112666, "grad_norm": 5.859003067016602, "learning_rate": 4.2513455328310013e-05, "loss": 2.0586, "step": 4180 }, { "epoch": 0.7517043415859347, "grad_norm": 8.045632362365723, "learning_rate": 4.249551489056333e-05, "loss": 1.9465, "step": 4190 }, { "epoch": 0.7534983853606028, "grad_norm": 6.814916133880615, "learning_rate": 4.247757445281665e-05, "loss": 1.8779, "step": 4200 }, { "epoch": 0.755292429135271, "grad_norm": 7.628875732421875, "learning_rate": 4.245963401506997e-05, "loss": 2.1836, "step": 4210 }, { "epoch": 0.757086472909939, "grad_norm": 6.975657939910889, "learning_rate": 4.244169357732329e-05, "loss": 1.7787, "step": 4220 }, { "epoch": 0.7588805166846071, "grad_norm": 7.706836700439453, "learning_rate": 4.2423753139576606e-05, "loss": 1.9749, "step": 4230 }, { "epoch": 0.7606745604592752, "grad_norm": 8.721719741821289, "learning_rate": 4.240581270182993e-05, "loss": 1.9422, "step": 4240 }, { "epoch": 0.7624686042339434, "grad_norm": 8.833395957946777, "learning_rate": 4.238787226408325e-05, "loss": 2.0886, "step": 4250 }, { "epoch": 0.7642626480086114, "grad_norm": 8.308511734008789, "learning_rate": 4.2369931826336565e-05, "loss": 1.9319, "step": 4260 }, { "epoch": 0.7660566917832795, "grad_norm": 7.8894476890563965, "learning_rate": 4.235199138858988e-05, "loss": 1.9378, "step": 4270 }, { "epoch": 0.7678507355579476, "grad_norm": 6.138456344604492, "learning_rate": 4.23340509508432e-05, "loss": 1.9009, "step": 4280 }, { "epoch": 0.7696447793326158, "grad_norm": 7.507815837860107, "learning_rate": 4.2316110513096524e-05, "loss": 1.9461, "step": 4290 }, { "epoch": 0.7714388231072838, "grad_norm": 7.388694763183594, "learning_rate": 4.229817007534984e-05, "loss": 1.7823, "step": 4300 }, { "epoch": 0.7732328668819519, "grad_norm": 7.483008861541748, "learning_rate": 4.228022963760316e-05, "loss": 2.0558, "step": 4310 }, { "epoch": 0.77502691065662, "grad_norm": 8.476717948913574, "learning_rate": 4.226228919985648e-05, "loss": 1.97, "step": 4320 }, { "epoch": 0.7768209544312881, "grad_norm": 8.137911796569824, "learning_rate": 4.22443487621098e-05, "loss": 2.0543, "step": 4330 }, { "epoch": 0.7786149982059563, "grad_norm": 7.722476482391357, "learning_rate": 4.222640832436312e-05, "loss": 1.8836, "step": 4340 }, { "epoch": 0.7804090419806243, "grad_norm": 7.449862003326416, "learning_rate": 4.220846788661644e-05, "loss": 1.9243, "step": 4350 }, { "epoch": 0.7822030857552924, "grad_norm": 6.4095563888549805, "learning_rate": 4.219052744886975e-05, "loss": 1.8377, "step": 4360 }, { "epoch": 0.7839971295299605, "grad_norm": 6.866125106811523, "learning_rate": 4.217258701112307e-05, "loss": 1.8246, "step": 4370 }, { "epoch": 0.7857911733046287, "grad_norm": 9.195448875427246, "learning_rate": 4.215464657337639e-05, "loss": 1.8019, "step": 4380 }, { "epoch": 0.7875852170792967, "grad_norm": 6.7557196617126465, "learning_rate": 4.213670613562971e-05, "loss": 1.8733, "step": 4390 }, { "epoch": 0.7893792608539648, "grad_norm": 7.832233905792236, "learning_rate": 4.211876569788303e-05, "loss": 1.727, "step": 4400 }, { "epoch": 0.7911733046286329, "grad_norm": 6.849626541137695, "learning_rate": 4.210082526013635e-05, "loss": 1.8051, "step": 4410 }, { "epoch": 0.7929673484033011, "grad_norm": 7.668883323669434, "learning_rate": 4.208288482238967e-05, "loss": 1.9021, "step": 4420 }, { "epoch": 0.7947613921779692, "grad_norm": 6.98665714263916, "learning_rate": 4.2064944384642986e-05, "loss": 1.9314, "step": 4430 }, { "epoch": 0.7965554359526372, "grad_norm": 7.1887383460998535, "learning_rate": 4.204700394689631e-05, "loss": 1.8046, "step": 4440 }, { "epoch": 0.7983494797273053, "grad_norm": 8.068037033081055, "learning_rate": 4.202906350914963e-05, "loss": 2.0018, "step": 4450 }, { "epoch": 0.8001435235019735, "grad_norm": 8.662615776062012, "learning_rate": 4.2011123071402944e-05, "loss": 1.9545, "step": 4460 }, { "epoch": 0.8019375672766416, "grad_norm": 6.354881286621094, "learning_rate": 4.199318263365626e-05, "loss": 1.699, "step": 4470 }, { "epoch": 0.8037316110513096, "grad_norm": 7.5929059982299805, "learning_rate": 4.197524219590958e-05, "loss": 1.7685, "step": 4480 }, { "epoch": 0.8055256548259777, "grad_norm": 7.677204608917236, "learning_rate": 4.19573017581629e-05, "loss": 1.876, "step": 4490 }, { "epoch": 0.8073196986006459, "grad_norm": 6.630999565124512, "learning_rate": 4.193936132041622e-05, "loss": 1.6986, "step": 4500 }, { "epoch": 0.809113742375314, "grad_norm": 6.430192947387695, "learning_rate": 4.192142088266954e-05, "loss": 1.9533, "step": 4510 }, { "epoch": 0.810907786149982, "grad_norm": 8.051733016967773, "learning_rate": 4.190348044492286e-05, "loss": 1.8555, "step": 4520 }, { "epoch": 0.8127018299246501, "grad_norm": 10.02955150604248, "learning_rate": 4.188554000717618e-05, "loss": 1.8035, "step": 4530 }, { "epoch": 0.8144958736993183, "grad_norm": 9.238576889038086, "learning_rate": 4.1867599569429496e-05, "loss": 1.9818, "step": 4540 }, { "epoch": 0.8162899174739864, "grad_norm": 6.678407192230225, "learning_rate": 4.184965913168282e-05, "loss": 1.808, "step": 4550 }, { "epoch": 0.8180839612486545, "grad_norm": 8.583653450012207, "learning_rate": 4.183171869393614e-05, "loss": 1.7423, "step": 4560 }, { "epoch": 0.8198780050233225, "grad_norm": 7.882835865020752, "learning_rate": 4.1813778256189454e-05, "loss": 1.9654, "step": 4570 }, { "epoch": 0.8216720487979907, "grad_norm": 7.367639541625977, "learning_rate": 4.179583781844277e-05, "loss": 1.8759, "step": 4580 }, { "epoch": 0.8234660925726588, "grad_norm": 5.808956146240234, "learning_rate": 4.177789738069609e-05, "loss": 1.8092, "step": 4590 }, { "epoch": 0.8252601363473269, "grad_norm": 8.538715362548828, "learning_rate": 4.1759956942949406e-05, "loss": 1.9468, "step": 4600 }, { "epoch": 0.8270541801219949, "grad_norm": 7.508995532989502, "learning_rate": 4.174201650520273e-05, "loss": 1.9887, "step": 4610 }, { "epoch": 0.8288482238966631, "grad_norm": 7.2825446128845215, "learning_rate": 4.172407606745605e-05, "loss": 1.9767, "step": 4620 }, { "epoch": 0.8306422676713312, "grad_norm": 7.998370170593262, "learning_rate": 4.1706135629709365e-05, "loss": 1.9921, "step": 4630 }, { "epoch": 0.8324363114459993, "grad_norm": 6.514996528625488, "learning_rate": 4.168819519196269e-05, "loss": 1.8236, "step": 4640 }, { "epoch": 0.8342303552206674, "grad_norm": 11.413960456848145, "learning_rate": 4.1670254754216006e-05, "loss": 1.9107, "step": 4650 }, { "epoch": 0.8360243989953355, "grad_norm": 10.405426025390625, "learning_rate": 4.165231431646932e-05, "loss": 1.936, "step": 4660 }, { "epoch": 0.8378184427700036, "grad_norm": 7.796232223510742, "learning_rate": 4.163437387872264e-05, "loss": 1.9083, "step": 4670 }, { "epoch": 0.8396124865446717, "grad_norm": 7.413230895996094, "learning_rate": 4.161643344097596e-05, "loss": 1.9956, "step": 4680 }, { "epoch": 0.8414065303193398, "grad_norm": 6.458949565887451, "learning_rate": 4.159849300322928e-05, "loss": 1.8525, "step": 4690 }, { "epoch": 0.843200574094008, "grad_norm": 7.839207172393799, "learning_rate": 4.15805525654826e-05, "loss": 1.9213, "step": 4700 }, { "epoch": 0.844994617868676, "grad_norm": 7.596676349639893, "learning_rate": 4.1562612127735916e-05, "loss": 1.7207, "step": 4710 }, { "epoch": 0.8467886616433441, "grad_norm": 8.215608596801758, "learning_rate": 4.154467168998924e-05, "loss": 1.8861, "step": 4720 }, { "epoch": 0.8485827054180122, "grad_norm": 8.735016822814941, "learning_rate": 4.152673125224256e-05, "loss": 1.7922, "step": 4730 }, { "epoch": 0.8503767491926802, "grad_norm": 7.5839667320251465, "learning_rate": 4.1508790814495875e-05, "loss": 1.8327, "step": 4740 }, { "epoch": 0.8521707929673484, "grad_norm": 8.134754180908203, "learning_rate": 4.14908503767492e-05, "loss": 2.0809, "step": 4750 }, { "epoch": 0.8539648367420165, "grad_norm": 6.481571674346924, "learning_rate": 4.1472909939002516e-05, "loss": 1.8645, "step": 4760 }, { "epoch": 0.8557588805166846, "grad_norm": 6.660789489746094, "learning_rate": 4.145496950125583e-05, "loss": 1.8249, "step": 4770 }, { "epoch": 0.8575529242913527, "grad_norm": 9.439289093017578, "learning_rate": 4.143702906350915e-05, "loss": 1.9101, "step": 4780 }, { "epoch": 0.8593469680660208, "grad_norm": 7.637185096740723, "learning_rate": 4.141908862576247e-05, "loss": 1.8607, "step": 4790 }, { "epoch": 0.8611410118406889, "grad_norm": 6.203429698944092, "learning_rate": 4.1401148188015785e-05, "loss": 1.6808, "step": 4800 }, { "epoch": 0.862935055615357, "grad_norm": 7.630656719207764, "learning_rate": 4.138320775026911e-05, "loss": 1.8628, "step": 4810 }, { "epoch": 0.8647290993900251, "grad_norm": 7.22261905670166, "learning_rate": 4.1365267312522426e-05, "loss": 1.8176, "step": 4820 }, { "epoch": 0.8665231431646933, "grad_norm": 7.170011520385742, "learning_rate": 4.134732687477575e-05, "loss": 1.9082, "step": 4830 }, { "epoch": 0.8683171869393613, "grad_norm": 6.344324588775635, "learning_rate": 4.132938643702907e-05, "loss": 1.8772, "step": 4840 }, { "epoch": 0.8701112307140294, "grad_norm": 8.296944618225098, "learning_rate": 4.1311445999282385e-05, "loss": 1.7995, "step": 4850 }, { "epoch": 0.8719052744886975, "grad_norm": 10.2393159866333, "learning_rate": 4.129350556153571e-05, "loss": 1.7471, "step": 4860 }, { "epoch": 0.8736993182633657, "grad_norm": 6.657433032989502, "learning_rate": 4.127556512378902e-05, "loss": 1.8175, "step": 4870 }, { "epoch": 0.8754933620380337, "grad_norm": 9.101529121398926, "learning_rate": 4.125762468604234e-05, "loss": 1.8132, "step": 4880 }, { "epoch": 0.8772874058127018, "grad_norm": 8.015061378479004, "learning_rate": 4.123968424829566e-05, "loss": 1.7798, "step": 4890 }, { "epoch": 0.8790814495873699, "grad_norm": 8.774176597595215, "learning_rate": 4.122174381054898e-05, "loss": 1.823, "step": 4900 }, { "epoch": 0.8808754933620381, "grad_norm": 8.144107818603516, "learning_rate": 4.1203803372802295e-05, "loss": 1.8932, "step": 4910 }, { "epoch": 0.8826695371367062, "grad_norm": 9.238556861877441, "learning_rate": 4.118586293505562e-05, "loss": 1.9033, "step": 4920 }, { "epoch": 0.8844635809113742, "grad_norm": 7.814840316772461, "learning_rate": 4.1167922497308937e-05, "loss": 1.9437, "step": 4930 }, { "epoch": 0.8862576246860423, "grad_norm": 8.720809936523438, "learning_rate": 4.1149982059562254e-05, "loss": 2.0195, "step": 4940 }, { "epoch": 0.8880516684607105, "grad_norm": 8.207976341247559, "learning_rate": 4.113204162181558e-05, "loss": 1.6992, "step": 4950 }, { "epoch": 0.8898457122353786, "grad_norm": 6.2852301597595215, "learning_rate": 4.1114101184068895e-05, "loss": 1.9237, "step": 4960 }, { "epoch": 0.8916397560100466, "grad_norm": 7.93939208984375, "learning_rate": 4.109616074632221e-05, "loss": 1.8576, "step": 4970 }, { "epoch": 0.8934337997847147, "grad_norm": 7.219597339630127, "learning_rate": 4.107822030857553e-05, "loss": 1.9087, "step": 4980 }, { "epoch": 0.8952278435593829, "grad_norm": 6.388714790344238, "learning_rate": 4.106027987082885e-05, "loss": 1.8785, "step": 4990 }, { "epoch": 0.897021887334051, "grad_norm": 7.048055648803711, "learning_rate": 4.1042339433082164e-05, "loss": 1.7598, "step": 5000 }, { "epoch": 0.898815931108719, "grad_norm": 9.254620552062988, "learning_rate": 4.102439899533549e-05, "loss": 1.8903, "step": 5010 }, { "epoch": 0.9006099748833871, "grad_norm": 7.499938488006592, "learning_rate": 4.1006458557588805e-05, "loss": 2.0262, "step": 5020 }, { "epoch": 0.9024040186580553, "grad_norm": 7.229835033416748, "learning_rate": 4.098851811984213e-05, "loss": 1.7852, "step": 5030 }, { "epoch": 0.9041980624327234, "grad_norm": 7.887816429138184, "learning_rate": 4.097057768209545e-05, "loss": 1.8154, "step": 5040 }, { "epoch": 0.9059921062073915, "grad_norm": 11.909371376037598, "learning_rate": 4.0952637244348764e-05, "loss": 1.8699, "step": 5050 }, { "epoch": 0.9077861499820595, "grad_norm": 9.439387321472168, "learning_rate": 4.093469680660209e-05, "loss": 2.0234, "step": 5060 }, { "epoch": 0.9095801937567277, "grad_norm": 7.028390884399414, "learning_rate": 4.0916756368855405e-05, "loss": 1.8577, "step": 5070 }, { "epoch": 0.9113742375313958, "grad_norm": 8.708728790283203, "learning_rate": 4.089881593110872e-05, "loss": 1.8517, "step": 5080 }, { "epoch": 0.9131682813060639, "grad_norm": 7.153110027313232, "learning_rate": 4.088087549336204e-05, "loss": 1.8174, "step": 5090 }, { "epoch": 0.9149623250807319, "grad_norm": 10.017348289489746, "learning_rate": 4.086293505561536e-05, "loss": 1.7079, "step": 5100 }, { "epoch": 0.9167563688554, "grad_norm": 8.1173734664917, "learning_rate": 4.0844994617868674e-05, "loss": 1.914, "step": 5110 }, { "epoch": 0.9185504126300682, "grad_norm": 7.267770290374756, "learning_rate": 4.0827054180122e-05, "loss": 1.8668, "step": 5120 }, { "epoch": 0.9203444564047363, "grad_norm": 8.86645793914795, "learning_rate": 4.0809113742375316e-05, "loss": 1.8357, "step": 5130 }, { "epoch": 0.9221385001794044, "grad_norm": 8.062966346740723, "learning_rate": 4.079117330462863e-05, "loss": 1.8604, "step": 5140 }, { "epoch": 0.9239325439540724, "grad_norm": 7.72880220413208, "learning_rate": 4.077323286688196e-05, "loss": 1.9482, "step": 5150 }, { "epoch": 0.9257265877287406, "grad_norm": 8.267526626586914, "learning_rate": 4.0755292429135274e-05, "loss": 1.6839, "step": 5160 }, { "epoch": 0.9275206315034087, "grad_norm": 10.039454460144043, "learning_rate": 4.073735199138859e-05, "loss": 1.9365, "step": 5170 }, { "epoch": 0.9293146752780768, "grad_norm": 9.875499725341797, "learning_rate": 4.071941155364191e-05, "loss": 1.8115, "step": 5180 }, { "epoch": 0.9311087190527448, "grad_norm": 7.694703102111816, "learning_rate": 4.0701471115895226e-05, "loss": 1.8054, "step": 5190 }, { "epoch": 0.932902762827413, "grad_norm": 6.072929382324219, "learning_rate": 4.068353067814855e-05, "loss": 1.9556, "step": 5200 }, { "epoch": 0.9346968066020811, "grad_norm": 9.701952934265137, "learning_rate": 4.066559024040187e-05, "loss": 1.9576, "step": 5210 }, { "epoch": 0.9364908503767492, "grad_norm": 8.353795051574707, "learning_rate": 4.0647649802655185e-05, "loss": 1.8579, "step": 5220 }, { "epoch": 0.9382848941514172, "grad_norm": 7.89420747756958, "learning_rate": 4.062970936490851e-05, "loss": 1.8866, "step": 5230 }, { "epoch": 0.9400789379260854, "grad_norm": 5.647555351257324, "learning_rate": 4.0611768927161826e-05, "loss": 1.742, "step": 5240 }, { "epoch": 0.9418729817007535, "grad_norm": 8.360129356384277, "learning_rate": 4.059382848941514e-05, "loss": 1.9297, "step": 5250 }, { "epoch": 0.9436670254754216, "grad_norm": 10.252833366394043, "learning_rate": 4.057588805166847e-05, "loss": 1.9036, "step": 5260 }, { "epoch": 0.9454610692500897, "grad_norm": 8.29806900024414, "learning_rate": 4.0557947613921784e-05, "loss": 1.7497, "step": 5270 }, { "epoch": 0.9472551130247578, "grad_norm": 6.64280891418457, "learning_rate": 4.05400071761751e-05, "loss": 1.9239, "step": 5280 }, { "epoch": 0.9490491567994259, "grad_norm": 7.992321014404297, "learning_rate": 4.052206673842842e-05, "loss": 1.8058, "step": 5290 }, { "epoch": 0.950843200574094, "grad_norm": 9.094325065612793, "learning_rate": 4.0504126300681736e-05, "loss": 1.9108, "step": 5300 }, { "epoch": 0.9526372443487621, "grad_norm": 8.64013385772705, "learning_rate": 4.048618586293505e-05, "loss": 2.1198, "step": 5310 }, { "epoch": 0.9544312881234303, "grad_norm": 6.910890579223633, "learning_rate": 4.046824542518838e-05, "loss": 1.9502, "step": 5320 }, { "epoch": 0.9562253318980983, "grad_norm": 7.383171081542969, "learning_rate": 4.0450304987441695e-05, "loss": 1.7905, "step": 5330 }, { "epoch": 0.9580193756727664, "grad_norm": 8.704651832580566, "learning_rate": 4.043236454969501e-05, "loss": 1.9428, "step": 5340 }, { "epoch": 0.9598134194474345, "grad_norm": 8.296703338623047, "learning_rate": 4.0414424111948336e-05, "loss": 1.8346, "step": 5350 }, { "epoch": 0.9616074632221027, "grad_norm": 7.613495349884033, "learning_rate": 4.039648367420165e-05, "loss": 1.9988, "step": 5360 }, { "epoch": 0.9634015069967707, "grad_norm": 8.145048141479492, "learning_rate": 4.037854323645497e-05, "loss": 1.8056, "step": 5370 }, { "epoch": 0.9651955507714388, "grad_norm": 9.955933570861816, "learning_rate": 4.0360602798708295e-05, "loss": 1.87, "step": 5380 }, { "epoch": 0.9669895945461069, "grad_norm": 5.713760852813721, "learning_rate": 4.0342662360961605e-05, "loss": 1.8019, "step": 5390 }, { "epoch": 0.9687836383207751, "grad_norm": 9.013385772705078, "learning_rate": 4.032472192321493e-05, "loss": 1.922, "step": 5400 }, { "epoch": 0.9705776820954432, "grad_norm": 8.361518859863281, "learning_rate": 4.0306781485468246e-05, "loss": 1.8847, "step": 5410 }, { "epoch": 0.9723717258701112, "grad_norm": 8.547319412231445, "learning_rate": 4.0288841047721564e-05, "loss": 1.9232, "step": 5420 }, { "epoch": 0.9741657696447793, "grad_norm": 8.168745994567871, "learning_rate": 4.027090060997489e-05, "loss": 1.834, "step": 5430 }, { "epoch": 0.9759598134194475, "grad_norm": 7.9649810791015625, "learning_rate": 4.0252960172228205e-05, "loss": 1.8909, "step": 5440 }, { "epoch": 0.9777538571941156, "grad_norm": 7.361401557922363, "learning_rate": 4.023501973448152e-05, "loss": 1.8859, "step": 5450 }, { "epoch": 0.9795479009687836, "grad_norm": 6.616429328918457, "learning_rate": 4.0217079296734846e-05, "loss": 1.8936, "step": 5460 }, { "epoch": 0.9813419447434517, "grad_norm": 8.482680320739746, "learning_rate": 4.0199138858988163e-05, "loss": 1.8714, "step": 5470 }, { "epoch": 0.9831359885181199, "grad_norm": 7.886808395385742, "learning_rate": 4.018119842124148e-05, "loss": 1.8772, "step": 5480 }, { "epoch": 0.984930032292788, "grad_norm": 8.63215446472168, "learning_rate": 4.01632579834948e-05, "loss": 1.8233, "step": 5490 }, { "epoch": 0.986724076067456, "grad_norm": 7.060615539550781, "learning_rate": 4.0145317545748115e-05, "loss": 1.7991, "step": 5500 }, { "epoch": 0.9885181198421241, "grad_norm": 7.455167293548584, "learning_rate": 4.012737710800143e-05, "loss": 1.8871, "step": 5510 }, { "epoch": 0.9903121636167922, "grad_norm": 8.120213508605957, "learning_rate": 4.0109436670254756e-05, "loss": 1.7909, "step": 5520 }, { "epoch": 0.9921062073914604, "grad_norm": 8.847984313964844, "learning_rate": 4.0091496232508074e-05, "loss": 1.9595, "step": 5530 }, { "epoch": 0.9939002511661285, "grad_norm": 7.193493366241455, "learning_rate": 4.007355579476139e-05, "loss": 1.7494, "step": 5540 }, { "epoch": 0.9956942949407965, "grad_norm": 9.286087989807129, "learning_rate": 4.0055615357014715e-05, "loss": 1.7659, "step": 5550 }, { "epoch": 0.9974883387154646, "grad_norm": 7.888460159301758, "learning_rate": 4.003767491926803e-05, "loss": 1.9127, "step": 5560 }, { "epoch": 0.9992823824901328, "grad_norm": 9.469123840332031, "learning_rate": 4.0019734481521356e-05, "loss": 1.9231, "step": 5570 }, { "epoch": 1.0, "eval_loss": 1.932178020477295, "eval_runtime": 189.8033, "eval_samples_per_second": 13.05, "eval_steps_per_second": 13.05, "step": 5574 }, { "epoch": 1.0010764262648009, "grad_norm": 8.350410461425781, "learning_rate": 4.0001794043774674e-05, "loss": 1.593, "step": 5580 }, { "epoch": 1.002870470039469, "grad_norm": 6.830000877380371, "learning_rate": 3.998385360602799e-05, "loss": 1.5654, "step": 5590 }, { "epoch": 1.004664513814137, "grad_norm": 7.733310222625732, "learning_rate": 3.996591316828131e-05, "loss": 1.3989, "step": 5600 }, { "epoch": 1.0064585575888052, "grad_norm": 8.72770881652832, "learning_rate": 3.9947972730534625e-05, "loss": 1.3938, "step": 5610 }, { "epoch": 1.0082526013634732, "grad_norm": 8.793525695800781, "learning_rate": 3.993003229278794e-05, "loss": 1.3272, "step": 5620 }, { "epoch": 1.0100466451381414, "grad_norm": 7.241087436676025, "learning_rate": 3.991209185504127e-05, "loss": 1.2756, "step": 5630 }, { "epoch": 1.0118406889128095, "grad_norm": 5.840522289276123, "learning_rate": 3.9894151417294584e-05, "loss": 1.3777, "step": 5640 }, { "epoch": 1.0136347326874775, "grad_norm": 8.860857009887695, "learning_rate": 3.98762109795479e-05, "loss": 1.3392, "step": 5650 }, { "epoch": 1.0154287764621457, "grad_norm": 7.003204822540283, "learning_rate": 3.9858270541801225e-05, "loss": 1.4117, "step": 5660 }, { "epoch": 1.0172228202368139, "grad_norm": 10.400496482849121, "learning_rate": 3.984033010405454e-05, "loss": 1.3564, "step": 5670 }, { "epoch": 1.0190168640114818, "grad_norm": 11.60659122467041, "learning_rate": 3.982238966630786e-05, "loss": 1.4216, "step": 5680 }, { "epoch": 1.02081090778615, "grad_norm": 8.097295761108398, "learning_rate": 3.980444922856118e-05, "loss": 1.4146, "step": 5690 }, { "epoch": 1.022604951560818, "grad_norm": 6.870365142822266, "learning_rate": 3.9786508790814494e-05, "loss": 1.4028, "step": 5700 }, { "epoch": 1.0243989953354862, "grad_norm": 8.893730163574219, "learning_rate": 3.976856835306781e-05, "loss": 1.4853, "step": 5710 }, { "epoch": 1.0261930391101544, "grad_norm": 9.703378677368164, "learning_rate": 3.9750627915321136e-05, "loss": 1.3958, "step": 5720 }, { "epoch": 1.0279870828848223, "grad_norm": 7.569676399230957, "learning_rate": 3.973268747757445e-05, "loss": 1.3828, "step": 5730 }, { "epoch": 1.0297811266594905, "grad_norm": 9.702757835388184, "learning_rate": 3.971474703982777e-05, "loss": 1.4878, "step": 5740 }, { "epoch": 1.0315751704341587, "grad_norm": 8.123019218444824, "learning_rate": 3.9696806602081094e-05, "loss": 1.3594, "step": 5750 }, { "epoch": 1.0333692142088267, "grad_norm": 7.301413059234619, "learning_rate": 3.967886616433441e-05, "loss": 1.4338, "step": 5760 }, { "epoch": 1.0351632579834948, "grad_norm": 9.144363403320312, "learning_rate": 3.9660925726587735e-05, "loss": 1.3546, "step": 5770 }, { "epoch": 1.0369573017581628, "grad_norm": 10.08015251159668, "learning_rate": 3.964298528884105e-05, "loss": 1.3771, "step": 5780 }, { "epoch": 1.038751345532831, "grad_norm": 7.706661701202393, "learning_rate": 3.962504485109437e-05, "loss": 1.3893, "step": 5790 }, { "epoch": 1.0405453893074992, "grad_norm": 7.710210800170898, "learning_rate": 3.960710441334769e-05, "loss": 1.4166, "step": 5800 }, { "epoch": 1.0423394330821671, "grad_norm": 9.725653648376465, "learning_rate": 3.9589163975601004e-05, "loss": 1.487, "step": 5810 }, { "epoch": 1.0441334768568353, "grad_norm": 6.656285762786865, "learning_rate": 3.957122353785432e-05, "loss": 1.2852, "step": 5820 }, { "epoch": 1.0459275206315035, "grad_norm": 11.030705451965332, "learning_rate": 3.9553283100107646e-05, "loss": 1.3853, "step": 5830 }, { "epoch": 1.0477215644061715, "grad_norm": 6.811320781707764, "learning_rate": 3.953534266236096e-05, "loss": 1.4783, "step": 5840 }, { "epoch": 1.0495156081808397, "grad_norm": 7.825105667114258, "learning_rate": 3.951740222461428e-05, "loss": 1.3688, "step": 5850 }, { "epoch": 1.0513096519555076, "grad_norm": 8.75365161895752, "learning_rate": 3.9499461786867604e-05, "loss": 1.3969, "step": 5860 }, { "epoch": 1.0531036957301758, "grad_norm": 9.158076286315918, "learning_rate": 3.948152134912092e-05, "loss": 1.4, "step": 5870 }, { "epoch": 1.054897739504844, "grad_norm": 9.204401016235352, "learning_rate": 3.946358091137424e-05, "loss": 1.3364, "step": 5880 }, { "epoch": 1.056691783279512, "grad_norm": 6.7577223777771, "learning_rate": 3.944564047362756e-05, "loss": 1.34, "step": 5890 }, { "epoch": 1.0584858270541802, "grad_norm": 7.4541497230529785, "learning_rate": 3.942770003588088e-05, "loss": 1.339, "step": 5900 }, { "epoch": 1.0602798708288481, "grad_norm": 8.368121147155762, "learning_rate": 3.940975959813419e-05, "loss": 1.453, "step": 5910 }, { "epoch": 1.0620739146035163, "grad_norm": 9.812310218811035, "learning_rate": 3.9391819160387515e-05, "loss": 1.4421, "step": 5920 }, { "epoch": 1.0638679583781845, "grad_norm": 8.019757270812988, "learning_rate": 3.937387872264083e-05, "loss": 1.3697, "step": 5930 }, { "epoch": 1.0656620021528525, "grad_norm": 7.330229759216309, "learning_rate": 3.9355938284894156e-05, "loss": 1.3469, "step": 5940 }, { "epoch": 1.0674560459275206, "grad_norm": 9.923416137695312, "learning_rate": 3.933799784714747e-05, "loss": 1.3891, "step": 5950 }, { "epoch": 1.0692500897021888, "grad_norm": 7.279324531555176, "learning_rate": 3.932005740940079e-05, "loss": 1.4288, "step": 5960 }, { "epoch": 1.0710441334768568, "grad_norm": 7.16546106338501, "learning_rate": 3.9302116971654114e-05, "loss": 1.347, "step": 5970 }, { "epoch": 1.072838177251525, "grad_norm": 8.2477445602417, "learning_rate": 3.928417653390743e-05, "loss": 1.3416, "step": 5980 }, { "epoch": 1.074632221026193, "grad_norm": 8.947925567626953, "learning_rate": 3.926623609616075e-05, "loss": 1.325, "step": 5990 }, { "epoch": 1.0764262648008611, "grad_norm": 10.749951362609863, "learning_rate": 3.9248295658414066e-05, "loss": 1.4012, "step": 6000 }, { "epoch": 1.0782203085755293, "grad_norm": 8.726066589355469, "learning_rate": 3.9230355220667383e-05, "loss": 1.3915, "step": 6010 }, { "epoch": 1.0800143523501973, "grad_norm": 8.84666919708252, "learning_rate": 3.92124147829207e-05, "loss": 1.4247, "step": 6020 }, { "epoch": 1.0818083961248655, "grad_norm": 8.546545028686523, "learning_rate": 3.9194474345174025e-05, "loss": 1.3545, "step": 6030 }, { "epoch": 1.0836024398995336, "grad_norm": 7.942462921142578, "learning_rate": 3.917653390742734e-05, "loss": 1.3572, "step": 6040 }, { "epoch": 1.0853964836742016, "grad_norm": 7.295432090759277, "learning_rate": 3.915859346968066e-05, "loss": 1.3773, "step": 6050 }, { "epoch": 1.0871905274488698, "grad_norm": 5.990417003631592, "learning_rate": 3.914065303193398e-05, "loss": 1.3617, "step": 6060 }, { "epoch": 1.0889845712235378, "grad_norm": 7.127756595611572, "learning_rate": 3.91227125941873e-05, "loss": 1.3374, "step": 6070 }, { "epoch": 1.090778614998206, "grad_norm": 9.005511283874512, "learning_rate": 3.910477215644062e-05, "loss": 1.4517, "step": 6080 }, { "epoch": 1.0925726587728741, "grad_norm": 7.71065616607666, "learning_rate": 3.908683171869394e-05, "loss": 1.4037, "step": 6090 }, { "epoch": 1.094366702547542, "grad_norm": 9.480076789855957, "learning_rate": 3.906889128094726e-05, "loss": 1.4425, "step": 6100 }, { "epoch": 1.0961607463222103, "grad_norm": 6.972677707672119, "learning_rate": 3.9050950843200576e-05, "loss": 1.356, "step": 6110 }, { "epoch": 1.0979547900968785, "grad_norm": 7.207665920257568, "learning_rate": 3.9033010405453894e-05, "loss": 1.3262, "step": 6120 }, { "epoch": 1.0997488338715464, "grad_norm": 6.460121154785156, "learning_rate": 3.901506996770721e-05, "loss": 1.3096, "step": 6130 }, { "epoch": 1.1015428776462146, "grad_norm": 10.60547924041748, "learning_rate": 3.8997129529960535e-05, "loss": 1.4342, "step": 6140 }, { "epoch": 1.1033369214208826, "grad_norm": 7.947498798370361, "learning_rate": 3.897918909221385e-05, "loss": 1.5028, "step": 6150 }, { "epoch": 1.1051309651955508, "grad_norm": 6.16427755355835, "learning_rate": 3.896124865446717e-05, "loss": 1.4129, "step": 6160 }, { "epoch": 1.106925008970219, "grad_norm": 6.936514377593994, "learning_rate": 3.8943308216720493e-05, "loss": 1.3718, "step": 6170 }, { "epoch": 1.108719052744887, "grad_norm": 6.0827555656433105, "learning_rate": 3.892536777897381e-05, "loss": 1.4196, "step": 6180 }, { "epoch": 1.110513096519555, "grad_norm": 9.733860969543457, "learning_rate": 3.890742734122713e-05, "loss": 1.302, "step": 6190 }, { "epoch": 1.1123071402942233, "grad_norm": 7.239533424377441, "learning_rate": 3.888948690348045e-05, "loss": 1.4005, "step": 6200 }, { "epoch": 1.1141011840688912, "grad_norm": 8.426368713378906, "learning_rate": 3.887154646573376e-05, "loss": 1.3118, "step": 6210 }, { "epoch": 1.1158952278435594, "grad_norm": 7.965389251708984, "learning_rate": 3.885360602798708e-05, "loss": 1.4111, "step": 6220 }, { "epoch": 1.1176892716182274, "grad_norm": 8.81957721710205, "learning_rate": 3.8835665590240404e-05, "loss": 1.4134, "step": 6230 }, { "epoch": 1.1194833153928956, "grad_norm": 7.341325283050537, "learning_rate": 3.881772515249372e-05, "loss": 1.2605, "step": 6240 }, { "epoch": 1.1212773591675638, "grad_norm": 8.882126808166504, "learning_rate": 3.879978471474704e-05, "loss": 1.4756, "step": 6250 }, { "epoch": 1.1230714029422317, "grad_norm": 7.364952564239502, "learning_rate": 3.878184427700036e-05, "loss": 1.3999, "step": 6260 }, { "epoch": 1.1248654467169, "grad_norm": 8.056137084960938, "learning_rate": 3.876390383925368e-05, "loss": 1.3775, "step": 6270 }, { "epoch": 1.1266594904915679, "grad_norm": 8.733875274658203, "learning_rate": 3.8745963401507e-05, "loss": 1.378, "step": 6280 }, { "epoch": 1.128453534266236, "grad_norm": 7.141792297363281, "learning_rate": 3.872802296376032e-05, "loss": 1.3789, "step": 6290 }, { "epoch": 1.1302475780409043, "grad_norm": 6.646508693695068, "learning_rate": 3.871008252601364e-05, "loss": 1.3432, "step": 6300 }, { "epoch": 1.1320416218155722, "grad_norm": 8.025551795959473, "learning_rate": 3.8692142088266955e-05, "loss": 1.4156, "step": 6310 }, { "epoch": 1.1338356655902404, "grad_norm": 9.390379905700684, "learning_rate": 3.867420165052027e-05, "loss": 1.3846, "step": 6320 }, { "epoch": 1.1356297093649086, "grad_norm": 7.891889572143555, "learning_rate": 3.865626121277359e-05, "loss": 1.4263, "step": 6330 }, { "epoch": 1.1374237531395766, "grad_norm": 7.289933681488037, "learning_rate": 3.8638320775026914e-05, "loss": 1.4822, "step": 6340 }, { "epoch": 1.1392177969142447, "grad_norm": 9.825151443481445, "learning_rate": 3.862038033728023e-05, "loss": 1.4937, "step": 6350 }, { "epoch": 1.141011840688913, "grad_norm": 7.730086803436279, "learning_rate": 3.860243989953355e-05, "loss": 1.3898, "step": 6360 }, { "epoch": 1.142805884463581, "grad_norm": 8.31853199005127, "learning_rate": 3.858449946178687e-05, "loss": 1.4197, "step": 6370 }, { "epoch": 1.144599928238249, "grad_norm": 9.047391891479492, "learning_rate": 3.856655902404019e-05, "loss": 1.3876, "step": 6380 }, { "epoch": 1.146393972012917, "grad_norm": 7.734961032867432, "learning_rate": 3.854861858629351e-05, "loss": 1.3464, "step": 6390 }, { "epoch": 1.1481880157875852, "grad_norm": 6.981812953948975, "learning_rate": 3.853067814854683e-05, "loss": 1.4332, "step": 6400 }, { "epoch": 1.1499820595622534, "grad_norm": 11.00357437133789, "learning_rate": 3.851273771080015e-05, "loss": 1.4755, "step": 6410 }, { "epoch": 1.1517761033369214, "grad_norm": 11.093791007995605, "learning_rate": 3.8494797273053466e-05, "loss": 1.3979, "step": 6420 }, { "epoch": 1.1535701471115896, "grad_norm": 6.121286392211914, "learning_rate": 3.847685683530678e-05, "loss": 1.3293, "step": 6430 }, { "epoch": 1.1553641908862575, "grad_norm": 8.417799949645996, "learning_rate": 3.84589163975601e-05, "loss": 1.4656, "step": 6440 }, { "epoch": 1.1571582346609257, "grad_norm": 7.4152655601501465, "learning_rate": 3.844097595981342e-05, "loss": 1.3806, "step": 6450 }, { "epoch": 1.158952278435594, "grad_norm": 6.638498783111572, "learning_rate": 3.842303552206674e-05, "loss": 1.4396, "step": 6460 }, { "epoch": 1.1607463222102619, "grad_norm": 12.287036895751953, "learning_rate": 3.840509508432006e-05, "loss": 1.3652, "step": 6470 }, { "epoch": 1.16254036598493, "grad_norm": 7.202840805053711, "learning_rate": 3.8387154646573376e-05, "loss": 1.5139, "step": 6480 }, { "epoch": 1.164334409759598, "grad_norm": 9.261022567749023, "learning_rate": 3.83692142088267e-05, "loss": 1.4731, "step": 6490 }, { "epoch": 1.1661284535342662, "grad_norm": 7.603264331817627, "learning_rate": 3.835127377108002e-05, "loss": 1.446, "step": 6500 }, { "epoch": 1.1679224973089344, "grad_norm": 7.141061305999756, "learning_rate": 3.8333333333333334e-05, "loss": 1.4075, "step": 6510 }, { "epoch": 1.1697165410836023, "grad_norm": 5.73789119720459, "learning_rate": 3.831539289558665e-05, "loss": 1.3154, "step": 6520 }, { "epoch": 1.1715105848582705, "grad_norm": 6.780825138092041, "learning_rate": 3.829745245783997e-05, "loss": 1.3756, "step": 6530 }, { "epoch": 1.1733046286329387, "grad_norm": 6.844207286834717, "learning_rate": 3.827951202009329e-05, "loss": 1.4014, "step": 6540 }, { "epoch": 1.1750986724076067, "grad_norm": 5.899712085723877, "learning_rate": 3.826157158234661e-05, "loss": 1.43, "step": 6550 }, { "epoch": 1.1768927161822749, "grad_norm": 7.629055976867676, "learning_rate": 3.824363114459993e-05, "loss": 1.369, "step": 6560 }, { "epoch": 1.178686759956943, "grad_norm": 7.492854118347168, "learning_rate": 3.822569070685325e-05, "loss": 1.3977, "step": 6570 }, { "epoch": 1.180480803731611, "grad_norm": 6.248719215393066, "learning_rate": 3.820775026910657e-05, "loss": 1.3657, "step": 6580 }, { "epoch": 1.1822748475062792, "grad_norm": 7.622773170471191, "learning_rate": 3.8189809831359886e-05, "loss": 1.3337, "step": 6590 }, { "epoch": 1.1840688912809472, "grad_norm": 6.013118267059326, "learning_rate": 3.817186939361321e-05, "loss": 1.3953, "step": 6600 }, { "epoch": 1.1858629350556154, "grad_norm": 7.73339319229126, "learning_rate": 3.815392895586653e-05, "loss": 1.389, "step": 6610 }, { "epoch": 1.1876569788302835, "grad_norm": 7.737553119659424, "learning_rate": 3.8135988518119845e-05, "loss": 1.4428, "step": 6620 }, { "epoch": 1.1894510226049515, "grad_norm": 8.525728225708008, "learning_rate": 3.811804808037316e-05, "loss": 1.4168, "step": 6630 }, { "epoch": 1.1912450663796197, "grad_norm": 8.121135711669922, "learning_rate": 3.810010764262648e-05, "loss": 1.4458, "step": 6640 }, { "epoch": 1.1930391101542877, "grad_norm": 8.677117347717285, "learning_rate": 3.8082167204879796e-05, "loss": 1.4039, "step": 6650 }, { "epoch": 1.1948331539289558, "grad_norm": 7.736855506896973, "learning_rate": 3.806422676713312e-05, "loss": 1.5071, "step": 6660 }, { "epoch": 1.196627197703624, "grad_norm": 8.960105895996094, "learning_rate": 3.804628632938644e-05, "loss": 1.5036, "step": 6670 }, { "epoch": 1.198421241478292, "grad_norm": 8.783327102661133, "learning_rate": 3.802834589163976e-05, "loss": 1.4161, "step": 6680 }, { "epoch": 1.2002152852529602, "grad_norm": 7.308254241943359, "learning_rate": 3.801040545389308e-05, "loss": 1.4173, "step": 6690 }, { "epoch": 1.2020093290276284, "grad_norm": 7.417787075042725, "learning_rate": 3.7992465016146396e-05, "loss": 1.3161, "step": 6700 }, { "epoch": 1.2038033728022963, "grad_norm": 7.6823320388793945, "learning_rate": 3.797452457839972e-05, "loss": 1.419, "step": 6710 }, { "epoch": 1.2055974165769645, "grad_norm": 9.797237396240234, "learning_rate": 3.795658414065304e-05, "loss": 1.3969, "step": 6720 }, { "epoch": 1.2073914603516327, "grad_norm": 7.217626094818115, "learning_rate": 3.793864370290635e-05, "loss": 1.3485, "step": 6730 }, { "epoch": 1.2091855041263007, "grad_norm": 7.798620223999023, "learning_rate": 3.792070326515967e-05, "loss": 1.2904, "step": 6740 }, { "epoch": 1.2109795479009688, "grad_norm": 9.841931343078613, "learning_rate": 3.790276282741299e-05, "loss": 1.4078, "step": 6750 }, { "epoch": 1.2127735916756368, "grad_norm": 8.47020435333252, "learning_rate": 3.788482238966631e-05, "loss": 1.4216, "step": 6760 }, { "epoch": 1.214567635450305, "grad_norm": 7.135184288024902, "learning_rate": 3.786688195191963e-05, "loss": 1.4159, "step": 6770 }, { "epoch": 1.2163616792249732, "grad_norm": 7.083540916442871, "learning_rate": 3.784894151417295e-05, "loss": 1.4004, "step": 6780 }, { "epoch": 1.2181557229996411, "grad_norm": 9.321927070617676, "learning_rate": 3.7831001076426265e-05, "loss": 1.4452, "step": 6790 }, { "epoch": 1.2199497667743093, "grad_norm": 8.969941139221191, "learning_rate": 3.781306063867959e-05, "loss": 1.4588, "step": 6800 }, { "epoch": 1.2217438105489773, "grad_norm": 7.412861347198486, "learning_rate": 3.7795120200932906e-05, "loss": 1.3526, "step": 6810 }, { "epoch": 1.2235378543236455, "grad_norm": 7.843968868255615, "learning_rate": 3.7777179763186224e-05, "loss": 1.3309, "step": 6820 }, { "epoch": 1.2253318980983137, "grad_norm": 6.551264762878418, "learning_rate": 3.775923932543954e-05, "loss": 1.4204, "step": 6830 }, { "epoch": 1.2271259418729816, "grad_norm": 8.688389778137207, "learning_rate": 3.774129888769286e-05, "loss": 1.4399, "step": 6840 }, { "epoch": 1.2289199856476498, "grad_norm": 8.688478469848633, "learning_rate": 3.7723358449946175e-05, "loss": 1.3618, "step": 6850 }, { "epoch": 1.230714029422318, "grad_norm": 8.20807933807373, "learning_rate": 3.77054180121995e-05, "loss": 1.4266, "step": 6860 }, { "epoch": 1.232508073196986, "grad_norm": 8.579695701599121, "learning_rate": 3.768747757445282e-05, "loss": 1.4258, "step": 6870 }, { "epoch": 1.2343021169716542, "grad_norm": 6.802687168121338, "learning_rate": 3.766953713670614e-05, "loss": 1.324, "step": 6880 }, { "epoch": 1.2360961607463221, "grad_norm": 6.273397445678711, "learning_rate": 3.765159669895946e-05, "loss": 1.375, "step": 6890 }, { "epoch": 1.2378902045209903, "grad_norm": 7.909563064575195, "learning_rate": 3.7633656261212775e-05, "loss": 1.297, "step": 6900 }, { "epoch": 1.2396842482956585, "grad_norm": 6.895613193511963, "learning_rate": 3.76157158234661e-05, "loss": 1.4579, "step": 6910 }, { "epoch": 1.2414782920703265, "grad_norm": 7.742472171783447, "learning_rate": 3.759777538571942e-05, "loss": 1.4445, "step": 6920 }, { "epoch": 1.2432723358449946, "grad_norm": 8.930936813354492, "learning_rate": 3.7579834947972734e-05, "loss": 1.2682, "step": 6930 }, { "epoch": 1.2450663796196628, "grad_norm": 6.448217391967773, "learning_rate": 3.756189451022605e-05, "loss": 1.4701, "step": 6940 }, { "epoch": 1.2468604233943308, "grad_norm": 6.52114200592041, "learning_rate": 3.754395407247937e-05, "loss": 1.4056, "step": 6950 }, { "epoch": 1.248654467168999, "grad_norm": 6.87349271774292, "learning_rate": 3.7526013634732686e-05, "loss": 1.3001, "step": 6960 }, { "epoch": 1.250448510943667, "grad_norm": 7.917197227478027, "learning_rate": 3.750807319698601e-05, "loss": 1.3976, "step": 6970 }, { "epoch": 1.2522425547183351, "grad_norm": 7.251535892486572, "learning_rate": 3.749013275923933e-05, "loss": 1.3153, "step": 6980 }, { "epoch": 1.2540365984930033, "grad_norm": 7.524764537811279, "learning_rate": 3.7472192321492644e-05, "loss": 1.4582, "step": 6990 }, { "epoch": 1.2558306422676713, "grad_norm": 8.992074012756348, "learning_rate": 3.745425188374597e-05, "loss": 1.3374, "step": 7000 }, { "epoch": 1.2576246860423395, "grad_norm": 10.817506790161133, "learning_rate": 3.7436311445999286e-05, "loss": 1.3457, "step": 7010 }, { "epoch": 1.2594187298170074, "grad_norm": 7.901443004608154, "learning_rate": 3.74183710082526e-05, "loss": 1.4439, "step": 7020 }, { "epoch": 1.2612127735916756, "grad_norm": 7.716637134552002, "learning_rate": 3.740043057050592e-05, "loss": 1.382, "step": 7030 }, { "epoch": 1.2630068173663438, "grad_norm": 7.97553014755249, "learning_rate": 3.738249013275924e-05, "loss": 1.3515, "step": 7040 }, { "epoch": 1.264800861141012, "grad_norm": 9.756081581115723, "learning_rate": 3.736454969501256e-05, "loss": 1.3815, "step": 7050 }, { "epoch": 1.26659490491568, "grad_norm": 7.838413238525391, "learning_rate": 3.734660925726588e-05, "loss": 1.4392, "step": 7060 }, { "epoch": 1.268388948690348, "grad_norm": 6.7297468185424805, "learning_rate": 3.7328668819519196e-05, "loss": 1.3719, "step": 7070 }, { "epoch": 1.270182992465016, "grad_norm": 7.776455879211426, "learning_rate": 3.731072838177252e-05, "loss": 1.3124, "step": 7080 }, { "epoch": 1.2719770362396843, "grad_norm": 8.554944038391113, "learning_rate": 3.729278794402584e-05, "loss": 1.3695, "step": 7090 }, { "epoch": 1.2737710800143525, "grad_norm": 6.043045997619629, "learning_rate": 3.7274847506279154e-05, "loss": 1.4149, "step": 7100 }, { "epoch": 1.2755651237890204, "grad_norm": 8.197552680969238, "learning_rate": 3.725690706853248e-05, "loss": 1.3095, "step": 7110 }, { "epoch": 1.2773591675636886, "grad_norm": 8.089428901672363, "learning_rate": 3.7238966630785796e-05, "loss": 1.4164, "step": 7120 }, { "epoch": 1.2791532113383566, "grad_norm": 12.543819427490234, "learning_rate": 3.722102619303911e-05, "loss": 1.2828, "step": 7130 }, { "epoch": 1.2809472551130248, "grad_norm": 7.321937561035156, "learning_rate": 3.720308575529243e-05, "loss": 1.2754, "step": 7140 }, { "epoch": 1.282741298887693, "grad_norm": 8.445917129516602, "learning_rate": 3.718514531754575e-05, "loss": 1.4625, "step": 7150 }, { "epoch": 1.284535342662361, "grad_norm": 5.982776165008545, "learning_rate": 3.7167204879799065e-05, "loss": 1.4169, "step": 7160 }, { "epoch": 1.286329386437029, "grad_norm": 7.333141803741455, "learning_rate": 3.714926444205239e-05, "loss": 1.466, "step": 7170 }, { "epoch": 1.288123430211697, "grad_norm": 7.591367244720459, "learning_rate": 3.7131324004305706e-05, "loss": 1.4827, "step": 7180 }, { "epoch": 1.2899174739863652, "grad_norm": 8.127729415893555, "learning_rate": 3.711338356655902e-05, "loss": 1.4232, "step": 7190 }, { "epoch": 1.2917115177610334, "grad_norm": 6.65064001083374, "learning_rate": 3.709544312881235e-05, "loss": 1.5094, "step": 7200 }, { "epoch": 1.2935055615357014, "grad_norm": 6.2712202072143555, "learning_rate": 3.7077502691065665e-05, "loss": 1.3632, "step": 7210 }, { "epoch": 1.2952996053103696, "grad_norm": 5.723544120788574, "learning_rate": 3.705956225331898e-05, "loss": 1.3911, "step": 7220 }, { "epoch": 1.2970936490850375, "grad_norm": 8.133112907409668, "learning_rate": 3.7041621815572306e-05, "loss": 1.3334, "step": 7230 }, { "epoch": 1.2988876928597057, "grad_norm": 9.744547843933105, "learning_rate": 3.702368137782562e-05, "loss": 1.4248, "step": 7240 }, { "epoch": 1.300681736634374, "grad_norm": 8.395071983337402, "learning_rate": 3.700574094007894e-05, "loss": 1.5036, "step": 7250 }, { "epoch": 1.302475780409042, "grad_norm": 8.343621253967285, "learning_rate": 3.698780050233226e-05, "loss": 1.3782, "step": 7260 }, { "epoch": 1.30426982418371, "grad_norm": 10.779861450195312, "learning_rate": 3.6969860064585575e-05, "loss": 1.4092, "step": 7270 }, { "epoch": 1.3060638679583783, "grad_norm": 12.210502624511719, "learning_rate": 3.69519196268389e-05, "loss": 1.3272, "step": 7280 }, { "epoch": 1.3078579117330462, "grad_norm": 8.902063369750977, "learning_rate": 3.6933979189092216e-05, "loss": 1.5054, "step": 7290 }, { "epoch": 1.3096519555077144, "grad_norm": 7.104801177978516, "learning_rate": 3.6916038751345533e-05, "loss": 1.4203, "step": 7300 }, { "epoch": 1.3114459992823826, "grad_norm": 5.748100757598877, "learning_rate": 3.689809831359886e-05, "loss": 1.4724, "step": 7310 }, { "epoch": 1.3132400430570506, "grad_norm": 8.453279495239258, "learning_rate": 3.6880157875852175e-05, "loss": 1.3665, "step": 7320 }, { "epoch": 1.3150340868317187, "grad_norm": 7.534119606018066, "learning_rate": 3.686221743810549e-05, "loss": 1.3748, "step": 7330 }, { "epoch": 1.3168281306063867, "grad_norm": 7.105281829833984, "learning_rate": 3.684427700035881e-05, "loss": 1.4103, "step": 7340 }, { "epoch": 1.318622174381055, "grad_norm": 7.355068683624268, "learning_rate": 3.6826336562612127e-05, "loss": 1.4492, "step": 7350 }, { "epoch": 1.320416218155723, "grad_norm": 6.437581539154053, "learning_rate": 3.6808396124865444e-05, "loss": 1.4176, "step": 7360 }, { "epoch": 1.322210261930391, "grad_norm": 8.563718795776367, "learning_rate": 3.679045568711877e-05, "loss": 1.3382, "step": 7370 }, { "epoch": 1.3240043057050592, "grad_norm": 7.735002040863037, "learning_rate": 3.6772515249372085e-05, "loss": 1.429, "step": 7380 }, { "epoch": 1.3257983494797272, "grad_norm": 7.190489768981934, "learning_rate": 3.67545748116254e-05, "loss": 1.3862, "step": 7390 }, { "epoch": 1.3275923932543954, "grad_norm": 8.7313814163208, "learning_rate": 3.6736634373878726e-05, "loss": 1.4154, "step": 7400 }, { "epoch": 1.3293864370290636, "grad_norm": 7.9539313316345215, "learning_rate": 3.6718693936132044e-05, "loss": 1.4132, "step": 7410 }, { "epoch": 1.3311804808037317, "grad_norm": 7.147149562835693, "learning_rate": 3.670075349838536e-05, "loss": 1.4062, "step": 7420 }, { "epoch": 1.3329745245783997, "grad_norm": 10.268470764160156, "learning_rate": 3.6682813060638685e-05, "loss": 1.4003, "step": 7430 }, { "epoch": 1.334768568353068, "grad_norm": 8.228311538696289, "learning_rate": 3.6664872622892e-05, "loss": 1.4234, "step": 7440 }, { "epoch": 1.3365626121277359, "grad_norm": 7.816580772399902, "learning_rate": 3.664693218514532e-05, "loss": 1.3835, "step": 7450 }, { "epoch": 1.338356655902404, "grad_norm": 5.961045742034912, "learning_rate": 3.662899174739864e-05, "loss": 1.4774, "step": 7460 }, { "epoch": 1.3401506996770722, "grad_norm": 9.59859848022461, "learning_rate": 3.6611051309651954e-05, "loss": 1.4122, "step": 7470 }, { "epoch": 1.3419447434517402, "grad_norm": 7.651191234588623, "learning_rate": 3.659311087190528e-05, "loss": 1.4664, "step": 7480 }, { "epoch": 1.3437387872264084, "grad_norm": 6.948186874389648, "learning_rate": 3.6575170434158595e-05, "loss": 1.3762, "step": 7490 }, { "epoch": 1.3455328310010763, "grad_norm": 5.879620552062988, "learning_rate": 3.655722999641191e-05, "loss": 1.3472, "step": 7500 }, { "epoch": 1.3473268747757445, "grad_norm": 10.246485710144043, "learning_rate": 3.6539289558665237e-05, "loss": 1.4756, "step": 7510 }, { "epoch": 1.3491209185504127, "grad_norm": 6.890244007110596, "learning_rate": 3.6521349120918554e-05, "loss": 1.3634, "step": 7520 }, { "epoch": 1.3509149623250807, "grad_norm": 9.019742012023926, "learning_rate": 3.650340868317187e-05, "loss": 1.5307, "step": 7530 }, { "epoch": 1.3527090060997489, "grad_norm": 7.515359401702881, "learning_rate": 3.648546824542519e-05, "loss": 1.2479, "step": 7540 }, { "epoch": 1.3545030498744168, "grad_norm": 7.437852382659912, "learning_rate": 3.6467527807678506e-05, "loss": 1.372, "step": 7550 }, { "epoch": 1.356297093649085, "grad_norm": 7.529526233673096, "learning_rate": 3.644958736993182e-05, "loss": 1.3672, "step": 7560 }, { "epoch": 1.3580911374237532, "grad_norm": 9.4348726272583, "learning_rate": 3.643164693218515e-05, "loss": 1.438, "step": 7570 }, { "epoch": 1.3598851811984212, "grad_norm": 8.870586395263672, "learning_rate": 3.6413706494438464e-05, "loss": 1.4271, "step": 7580 }, { "epoch": 1.3616792249730894, "grad_norm": 7.35816764831543, "learning_rate": 3.639576605669178e-05, "loss": 1.4001, "step": 7590 }, { "epoch": 1.3634732687477573, "grad_norm": 7.633657455444336, "learning_rate": 3.6377825618945105e-05, "loss": 1.4443, "step": 7600 }, { "epoch": 1.3652673125224255, "grad_norm": 8.26876163482666, "learning_rate": 3.635988518119842e-05, "loss": 1.3539, "step": 7610 }, { "epoch": 1.3670613562970937, "grad_norm": 9.011847496032715, "learning_rate": 3.634194474345175e-05, "loss": 1.3701, "step": 7620 }, { "epoch": 1.3688554000717619, "grad_norm": 8.344417572021484, "learning_rate": 3.6324004305705064e-05, "loss": 1.4872, "step": 7630 }, { "epoch": 1.3706494438464298, "grad_norm": 6.184938907623291, "learning_rate": 3.630606386795838e-05, "loss": 1.284, "step": 7640 }, { "epoch": 1.372443487621098, "grad_norm": 8.20331859588623, "learning_rate": 3.62881234302117e-05, "loss": 1.363, "step": 7650 }, { "epoch": 1.374237531395766, "grad_norm": 6.358704090118408, "learning_rate": 3.6270182992465016e-05, "loss": 1.3962, "step": 7660 }, { "epoch": 1.3760315751704342, "grad_norm": 7.734986305236816, "learning_rate": 3.625224255471833e-05, "loss": 1.3673, "step": 7670 }, { "epoch": 1.3778256189451024, "grad_norm": 8.597131729125977, "learning_rate": 3.623430211697166e-05, "loss": 1.372, "step": 7680 }, { "epoch": 1.3796196627197703, "grad_norm": 9.663629531860352, "learning_rate": 3.6216361679224974e-05, "loss": 1.3264, "step": 7690 }, { "epoch": 1.3814137064944385, "grad_norm": 8.822710990905762, "learning_rate": 3.619842124147829e-05, "loss": 1.4173, "step": 7700 }, { "epoch": 1.3832077502691065, "grad_norm": 6.7463459968566895, "learning_rate": 3.6180480803731616e-05, "loss": 1.3586, "step": 7710 }, { "epoch": 1.3850017940437747, "grad_norm": 7.816486835479736, "learning_rate": 3.616254036598493e-05, "loss": 1.4229, "step": 7720 }, { "epoch": 1.3867958378184428, "grad_norm": 8.482382774353027, "learning_rate": 3.614459992823825e-05, "loss": 1.4222, "step": 7730 }, { "epoch": 1.3885898815931108, "grad_norm": 8.289495468139648, "learning_rate": 3.6126659490491574e-05, "loss": 1.387, "step": 7740 }, { "epoch": 1.390383925367779, "grad_norm": 7.847174167633057, "learning_rate": 3.610871905274489e-05, "loss": 1.3878, "step": 7750 }, { "epoch": 1.392177969142447, "grad_norm": 7.200404167175293, "learning_rate": 3.60907786149982e-05, "loss": 1.4336, "step": 7760 }, { "epoch": 1.3939720129171151, "grad_norm": 7.738436698913574, "learning_rate": 3.6072838177251526e-05, "loss": 1.3772, "step": 7770 }, { "epoch": 1.3957660566917833, "grad_norm": 6.875908374786377, "learning_rate": 3.605489773950484e-05, "loss": 1.4554, "step": 7780 }, { "epoch": 1.3975601004664515, "grad_norm": 9.192317008972168, "learning_rate": 3.603695730175816e-05, "loss": 1.3069, "step": 7790 }, { "epoch": 1.3993541442411195, "grad_norm": 10.130300521850586, "learning_rate": 3.6019016864011484e-05, "loss": 1.3736, "step": 7800 }, { "epoch": 1.4011481880157877, "grad_norm": 7.170609951019287, "learning_rate": 3.60010764262648e-05, "loss": 1.3729, "step": 7810 }, { "epoch": 1.4029422317904556, "grad_norm": 9.054718971252441, "learning_rate": 3.5983135988518126e-05, "loss": 1.43, "step": 7820 }, { "epoch": 1.4047362755651238, "grad_norm": 7.545653820037842, "learning_rate": 3.596519555077144e-05, "loss": 1.3767, "step": 7830 }, { "epoch": 1.406530319339792, "grad_norm": 8.128164291381836, "learning_rate": 3.594725511302476e-05, "loss": 1.3917, "step": 7840 }, { "epoch": 1.40832436311446, "grad_norm": 8.18182373046875, "learning_rate": 3.592931467527808e-05, "loss": 1.4083, "step": 7850 }, { "epoch": 1.4101184068891282, "grad_norm": 9.797805786132812, "learning_rate": 3.5911374237531395e-05, "loss": 1.4863, "step": 7860 }, { "epoch": 1.4119124506637961, "grad_norm": 25.17400550842285, "learning_rate": 3.589343379978471e-05, "loss": 1.5191, "step": 7870 }, { "epoch": 1.4137064944384643, "grad_norm": 7.167262077331543, "learning_rate": 3.5875493362038036e-05, "loss": 1.4151, "step": 7880 }, { "epoch": 1.4155005382131325, "grad_norm": 6.82859992980957, "learning_rate": 3.585755292429135e-05, "loss": 1.374, "step": 7890 }, { "epoch": 1.4172945819878005, "grad_norm": 8.16007137298584, "learning_rate": 3.583961248654467e-05, "loss": 1.3523, "step": 7900 }, { "epoch": 1.4190886257624686, "grad_norm": 7.259387969970703, "learning_rate": 3.5821672048797995e-05, "loss": 1.3811, "step": 7910 }, { "epoch": 1.4208826695371366, "grad_norm": 7.798694610595703, "learning_rate": 3.580373161105131e-05, "loss": 1.4331, "step": 7920 }, { "epoch": 1.4226767133118048, "grad_norm": 6.832756519317627, "learning_rate": 3.578579117330463e-05, "loss": 1.3798, "step": 7930 }, { "epoch": 1.424470757086473, "grad_norm": 8.465885162353516, "learning_rate": 3.576785073555795e-05, "loss": 1.368, "step": 7940 }, { "epoch": 1.426264800861141, "grad_norm": 7.031029224395752, "learning_rate": 3.574991029781127e-05, "loss": 1.2938, "step": 7950 }, { "epoch": 1.4280588446358091, "grad_norm": 8.129319190979004, "learning_rate": 3.573196986006459e-05, "loss": 1.3336, "step": 7960 }, { "epoch": 1.429852888410477, "grad_norm": 7.408348560333252, "learning_rate": 3.5714029422317905e-05, "loss": 1.2998, "step": 7970 }, { "epoch": 1.4316469321851453, "grad_norm": 7.85107421875, "learning_rate": 3.569608898457122e-05, "loss": 1.483, "step": 7980 }, { "epoch": 1.4334409759598135, "grad_norm": 7.82799768447876, "learning_rate": 3.5678148546824546e-05, "loss": 1.4505, "step": 7990 }, { "epoch": 1.4352350197344816, "grad_norm": 10.371716499328613, "learning_rate": 3.5660208109077864e-05, "loss": 1.3282, "step": 8000 }, { "epoch": 1.4370290635091496, "grad_norm": 9.943699836730957, "learning_rate": 3.564226767133118e-05, "loss": 1.4015, "step": 8010 }, { "epoch": 1.4388231072838178, "grad_norm": 6.916280746459961, "learning_rate": 3.5624327233584505e-05, "loss": 1.4301, "step": 8020 }, { "epoch": 1.4406171510584858, "grad_norm": 8.76120662689209, "learning_rate": 3.560638679583782e-05, "loss": 1.386, "step": 8030 }, { "epoch": 1.442411194833154, "grad_norm": 8.706786155700684, "learning_rate": 3.558844635809114e-05, "loss": 1.4428, "step": 8040 }, { "epoch": 1.4442052386078221, "grad_norm": 6.031268119812012, "learning_rate": 3.557050592034446e-05, "loss": 1.347, "step": 8050 }, { "epoch": 1.44599928238249, "grad_norm": 9.217594146728516, "learning_rate": 3.5552565482597774e-05, "loss": 1.5703, "step": 8060 }, { "epoch": 1.4477933261571583, "grad_norm": 8.151002883911133, "learning_rate": 3.553462504485109e-05, "loss": 1.3317, "step": 8070 }, { "epoch": 1.4495873699318262, "grad_norm": 8.76799488067627, "learning_rate": 3.5516684607104415e-05, "loss": 1.4876, "step": 8080 }, { "epoch": 1.4513814137064944, "grad_norm": 11.447850227355957, "learning_rate": 3.549874416935773e-05, "loss": 1.3937, "step": 8090 }, { "epoch": 1.4531754574811626, "grad_norm": 7.755031585693359, "learning_rate": 3.548080373161105e-05, "loss": 1.4311, "step": 8100 }, { "epoch": 1.4549695012558306, "grad_norm": 5.995707988739014, "learning_rate": 3.5462863293864374e-05, "loss": 1.4221, "step": 8110 }, { "epoch": 1.4567635450304988, "grad_norm": 6.535067558288574, "learning_rate": 3.544492285611769e-05, "loss": 1.3953, "step": 8120 }, { "epoch": 1.4585575888051667, "grad_norm": 14.317142486572266, "learning_rate": 3.542698241837101e-05, "loss": 1.4777, "step": 8130 }, { "epoch": 1.460351632579835, "grad_norm": 9.180553436279297, "learning_rate": 3.540904198062433e-05, "loss": 1.3552, "step": 8140 }, { "epoch": 1.462145676354503, "grad_norm": 7.935247898101807, "learning_rate": 3.539110154287765e-05, "loss": 1.296, "step": 8150 }, { "epoch": 1.4639397201291713, "grad_norm": 9.270010948181152, "learning_rate": 3.537316110513097e-05, "loss": 1.3484, "step": 8160 }, { "epoch": 1.4657337639038392, "grad_norm": 7.344752311706543, "learning_rate": 3.5355220667384284e-05, "loss": 1.3878, "step": 8170 }, { "epoch": 1.4675278076785074, "grad_norm": 5.794569492340088, "learning_rate": 3.53372802296376e-05, "loss": 1.2688, "step": 8180 }, { "epoch": 1.4693218514531754, "grad_norm": 8.966668128967285, "learning_rate": 3.5319339791890925e-05, "loss": 1.3682, "step": 8190 }, { "epoch": 1.4711158952278436, "grad_norm": 8.947368621826172, "learning_rate": 3.530139935414424e-05, "loss": 1.3869, "step": 8200 }, { "epoch": 1.4729099390025118, "grad_norm": 8.610750198364258, "learning_rate": 3.528345891639756e-05, "loss": 1.3788, "step": 8210 }, { "epoch": 1.4747039827771797, "grad_norm": 8.098158836364746, "learning_rate": 3.5265518478650884e-05, "loss": 1.364, "step": 8220 }, { "epoch": 1.476498026551848, "grad_norm": 5.487945079803467, "learning_rate": 3.52475780409042e-05, "loss": 1.4057, "step": 8230 }, { "epoch": 1.4782920703265159, "grad_norm": 7.017474174499512, "learning_rate": 3.522963760315752e-05, "loss": 1.4042, "step": 8240 }, { "epoch": 1.480086114101184, "grad_norm": 7.3450469970703125, "learning_rate": 3.521169716541084e-05, "loss": 1.3999, "step": 8250 }, { "epoch": 1.4818801578758523, "grad_norm": 7.210866928100586, "learning_rate": 3.519375672766416e-05, "loss": 1.3868, "step": 8260 }, { "epoch": 1.4836742016505202, "grad_norm": 8.641252517700195, "learning_rate": 3.517581628991748e-05, "loss": 1.374, "step": 8270 }, { "epoch": 1.4854682454251884, "grad_norm": 8.86153507232666, "learning_rate": 3.5157875852170794e-05, "loss": 1.4113, "step": 8280 }, { "epoch": 1.4872622891998564, "grad_norm": 8.298069953918457, "learning_rate": 3.513993541442411e-05, "loss": 1.3787, "step": 8290 }, { "epoch": 1.4890563329745246, "grad_norm": 7.584449291229248, "learning_rate": 3.512199497667743e-05, "loss": 1.4623, "step": 8300 }, { "epoch": 1.4908503767491927, "grad_norm": 9.79563045501709, "learning_rate": 3.510405453893075e-05, "loss": 1.3634, "step": 8310 }, { "epoch": 1.4926444205238607, "grad_norm": 7.416035175323486, "learning_rate": 3.508611410118407e-05, "loss": 1.3367, "step": 8320 }, { "epoch": 1.494438464298529, "grad_norm": 7.025708198547363, "learning_rate": 3.506817366343739e-05, "loss": 1.4087, "step": 8330 }, { "epoch": 1.4962325080731969, "grad_norm": 6.349498271942139, "learning_rate": 3.505023322569071e-05, "loss": 1.2794, "step": 8340 }, { "epoch": 1.498026551847865, "grad_norm": 9.222763061523438, "learning_rate": 3.503229278794403e-05, "loss": 1.4334, "step": 8350 }, { "epoch": 1.4998205956225332, "grad_norm": 7.805530548095703, "learning_rate": 3.5014352350197346e-05, "loss": 1.5309, "step": 8360 }, { "epoch": 1.5016146393972014, "grad_norm": 7.701581954956055, "learning_rate": 3.499641191245066e-05, "loss": 1.4599, "step": 8370 }, { "epoch": 1.5034086831718694, "grad_norm": 7.224992275238037, "learning_rate": 3.497847147470398e-05, "loss": 1.3609, "step": 8380 }, { "epoch": 1.5052027269465373, "grad_norm": 6.719163417816162, "learning_rate": 3.4960531036957304e-05, "loss": 1.3586, "step": 8390 }, { "epoch": 1.5069967707212055, "grad_norm": 6.938652038574219, "learning_rate": 3.494259059921062e-05, "loss": 1.455, "step": 8400 }, { "epoch": 1.5087908144958737, "grad_norm": 7.92690896987915, "learning_rate": 3.492465016146394e-05, "loss": 1.4174, "step": 8410 }, { "epoch": 1.510584858270542, "grad_norm": 6.6979660987854, "learning_rate": 3.490670972371726e-05, "loss": 1.337, "step": 8420 }, { "epoch": 1.5123789020452099, "grad_norm": 7.451391220092773, "learning_rate": 3.488876928597058e-05, "loss": 1.3879, "step": 8430 }, { "epoch": 1.514172945819878, "grad_norm": 8.262466430664062, "learning_rate": 3.48708288482239e-05, "loss": 1.3828, "step": 8440 }, { "epoch": 1.515966989594546, "grad_norm": 8.4102201461792, "learning_rate": 3.485288841047722e-05, "loss": 1.4623, "step": 8450 }, { "epoch": 1.5177610333692142, "grad_norm": 7.731327533721924, "learning_rate": 3.483494797273054e-05, "loss": 1.381, "step": 8460 }, { "epoch": 1.5195550771438824, "grad_norm": 8.887019157409668, "learning_rate": 3.4817007534983856e-05, "loss": 1.3625, "step": 8470 }, { "epoch": 1.5213491209185506, "grad_norm": 7.205885410308838, "learning_rate": 3.479906709723717e-05, "loss": 1.3577, "step": 8480 }, { "epoch": 1.5231431646932185, "grad_norm": 12.000414848327637, "learning_rate": 3.478112665949049e-05, "loss": 1.4075, "step": 8490 }, { "epoch": 1.5249372084678865, "grad_norm": 7.290775775909424, "learning_rate": 3.476318622174381e-05, "loss": 1.3658, "step": 8500 }, { "epoch": 1.5267312522425547, "grad_norm": 7.6026225090026855, "learning_rate": 3.474524578399713e-05, "loss": 1.3091, "step": 8510 }, { "epoch": 1.5285252960172229, "grad_norm": 8.19789981842041, "learning_rate": 3.472730534625045e-05, "loss": 1.4242, "step": 8520 }, { "epoch": 1.530319339791891, "grad_norm": 9.370061874389648, "learning_rate": 3.4709364908503766e-05, "loss": 1.4437, "step": 8530 }, { "epoch": 1.532113383566559, "grad_norm": 7.536168575286865, "learning_rate": 3.469142447075709e-05, "loss": 1.3764, "step": 8540 }, { "epoch": 1.533907427341227, "grad_norm": 8.509814262390137, "learning_rate": 3.467348403301041e-05, "loss": 1.3138, "step": 8550 }, { "epoch": 1.5357014711158952, "grad_norm": 7.023370265960693, "learning_rate": 3.465554359526373e-05, "loss": 1.43, "step": 8560 }, { "epoch": 1.5374955148905634, "grad_norm": 8.391500473022461, "learning_rate": 3.463760315751705e-05, "loss": 1.3419, "step": 8570 }, { "epoch": 1.5392895586652315, "grad_norm": 8.216898918151855, "learning_rate": 3.461966271977036e-05, "loss": 1.3849, "step": 8580 }, { "epoch": 1.5410836024398995, "grad_norm": 7.629865646362305, "learning_rate": 3.4601722282023683e-05, "loss": 1.4518, "step": 8590 }, { "epoch": 1.5428776462145677, "grad_norm": 7.480790138244629, "learning_rate": 3.4583781844277e-05, "loss": 1.3378, "step": 8600 }, { "epoch": 1.5446716899892357, "grad_norm": 7.571674346923828, "learning_rate": 3.456584140653032e-05, "loss": 1.4035, "step": 8610 }, { "epoch": 1.5464657337639038, "grad_norm": 8.145289421081543, "learning_rate": 3.454790096878364e-05, "loss": 1.3878, "step": 8620 }, { "epoch": 1.548259777538572, "grad_norm": 8.560420036315918, "learning_rate": 3.452996053103696e-05, "loss": 1.4033, "step": 8630 }, { "epoch": 1.55005382131324, "grad_norm": 8.526228904724121, "learning_rate": 3.4512020093290277e-05, "loss": 1.3257, "step": 8640 }, { "epoch": 1.5518478650879082, "grad_norm": 7.641489505767822, "learning_rate": 3.44940796555436e-05, "loss": 1.37, "step": 8650 }, { "epoch": 1.5536419088625761, "grad_norm": 7.322695732116699, "learning_rate": 3.447613921779692e-05, "loss": 1.3798, "step": 8660 }, { "epoch": 1.5554359526372443, "grad_norm": 9.153738975524902, "learning_rate": 3.4458198780050235e-05, "loss": 1.5305, "step": 8670 }, { "epoch": 1.5572299964119125, "grad_norm": 6.289071559906006, "learning_rate": 3.444025834230355e-05, "loss": 1.213, "step": 8680 }, { "epoch": 1.5590240401865807, "grad_norm": 8.083537101745605, "learning_rate": 3.442231790455687e-05, "loss": 1.3258, "step": 8690 }, { "epoch": 1.5608180839612487, "grad_norm": 8.730308532714844, "learning_rate": 3.440437746681019e-05, "loss": 1.4032, "step": 8700 }, { "epoch": 1.5626121277359166, "grad_norm": 8.361651420593262, "learning_rate": 3.438643702906351e-05, "loss": 1.3918, "step": 8710 }, { "epoch": 1.5644061715105848, "grad_norm": 7.762195110321045, "learning_rate": 3.436849659131683e-05, "loss": 1.4657, "step": 8720 }, { "epoch": 1.566200215285253, "grad_norm": 8.610489845275879, "learning_rate": 3.435055615357015e-05, "loss": 1.4472, "step": 8730 }, { "epoch": 1.5679942590599212, "grad_norm": 6.9795098304748535, "learning_rate": 3.433261571582347e-05, "loss": 1.5167, "step": 8740 }, { "epoch": 1.5697883028345891, "grad_norm": 6.0824408531188965, "learning_rate": 3.431467527807679e-05, "loss": 1.446, "step": 8750 }, { "epoch": 1.571582346609257, "grad_norm": 6.787125587463379, "learning_rate": 3.429673484033011e-05, "loss": 1.5047, "step": 8760 }, { "epoch": 1.5733763903839253, "grad_norm": 8.646109580993652, "learning_rate": 3.427879440258343e-05, "loss": 1.3271, "step": 8770 }, { "epoch": 1.5751704341585935, "grad_norm": 7.65358304977417, "learning_rate": 3.4260853964836745e-05, "loss": 1.3963, "step": 8780 }, { "epoch": 1.5769644779332617, "grad_norm": 9.546274185180664, "learning_rate": 3.424291352709006e-05, "loss": 1.4044, "step": 8790 }, { "epoch": 1.5787585217079296, "grad_norm": 8.250378608703613, "learning_rate": 3.422497308934338e-05, "loss": 1.3717, "step": 8800 }, { "epoch": 1.5805525654825978, "grad_norm": 8.04243278503418, "learning_rate": 3.42070326515967e-05, "loss": 1.3954, "step": 8810 }, { "epoch": 1.5823466092572658, "grad_norm": 7.013770580291748, "learning_rate": 3.418909221385002e-05, "loss": 1.3597, "step": 8820 }, { "epoch": 1.584140653031934, "grad_norm": 7.129620552062988, "learning_rate": 3.417115177610334e-05, "loss": 1.3492, "step": 8830 }, { "epoch": 1.5859346968066022, "grad_norm": 7.23296594619751, "learning_rate": 3.4153211338356656e-05, "loss": 1.4067, "step": 8840 }, { "epoch": 1.5877287405812703, "grad_norm": 6.118582248687744, "learning_rate": 3.413527090060998e-05, "loss": 1.3626, "step": 8850 }, { "epoch": 1.5895227843559383, "grad_norm": 8.783863067626953, "learning_rate": 3.41173304628633e-05, "loss": 1.4064, "step": 8860 }, { "epoch": 1.5913168281306063, "grad_norm": 7.83473014831543, "learning_rate": 3.4099390025116614e-05, "loss": 1.391, "step": 8870 }, { "epoch": 1.5931108719052745, "grad_norm": 7.351261138916016, "learning_rate": 3.408144958736993e-05, "loss": 1.3933, "step": 8880 }, { "epoch": 1.5949049156799426, "grad_norm": 7.231532096862793, "learning_rate": 3.406350914962325e-05, "loss": 1.4398, "step": 8890 }, { "epoch": 1.5966989594546108, "grad_norm": 7.532142639160156, "learning_rate": 3.4045568711876566e-05, "loss": 1.3316, "step": 8900 }, { "epoch": 1.5984930032292788, "grad_norm": 6.363405227661133, "learning_rate": 3.402762827412989e-05, "loss": 1.4548, "step": 8910 }, { "epoch": 1.6002870470039467, "grad_norm": 8.271924018859863, "learning_rate": 3.400968783638321e-05, "loss": 1.2696, "step": 8920 }, { "epoch": 1.602081090778615, "grad_norm": 6.662013053894043, "learning_rate": 3.399174739863653e-05, "loss": 1.3041, "step": 8930 }, { "epoch": 1.6038751345532831, "grad_norm": 9.983985900878906, "learning_rate": 3.397380696088985e-05, "loss": 1.3968, "step": 8940 }, { "epoch": 1.6056691783279513, "grad_norm": 7.235292911529541, "learning_rate": 3.3955866523143166e-05, "loss": 1.3696, "step": 8950 }, { "epoch": 1.6074632221026193, "grad_norm": 10.884222984313965, "learning_rate": 3.393792608539649e-05, "loss": 1.5237, "step": 8960 }, { "epoch": 1.6092572658772875, "grad_norm": 7.906916618347168, "learning_rate": 3.391998564764981e-05, "loss": 1.5189, "step": 8970 }, { "epoch": 1.6110513096519554, "grad_norm": 6.324611186981201, "learning_rate": 3.3902045209903124e-05, "loss": 1.4566, "step": 8980 }, { "epoch": 1.6128453534266236, "grad_norm": 9.540352821350098, "learning_rate": 3.388410477215644e-05, "loss": 1.3418, "step": 8990 }, { "epoch": 1.6146393972012918, "grad_norm": 9.794340133666992, "learning_rate": 3.386616433440976e-05, "loss": 1.532, "step": 9000 }, { "epoch": 1.61643344097596, "grad_norm": 7.887220859527588, "learning_rate": 3.3848223896663076e-05, "loss": 1.3238, "step": 9010 }, { "epoch": 1.618227484750628, "grad_norm": 8.027101516723633, "learning_rate": 3.38302834589164e-05, "loss": 1.255, "step": 9020 }, { "epoch": 1.620021528525296, "grad_norm": 13.341578483581543, "learning_rate": 3.381234302116972e-05, "loss": 1.4349, "step": 9030 }, { "epoch": 1.621815572299964, "grad_norm": 8.621784210205078, "learning_rate": 3.3794402583423035e-05, "loss": 1.3003, "step": 9040 }, { "epoch": 1.6236096160746323, "grad_norm": 6.994039058685303, "learning_rate": 3.377646214567636e-05, "loss": 1.4052, "step": 9050 }, { "epoch": 1.6254036598493005, "grad_norm": 7.07314920425415, "learning_rate": 3.3758521707929676e-05, "loss": 1.4283, "step": 9060 }, { "epoch": 1.6271977036239684, "grad_norm": 6.580167293548584, "learning_rate": 3.374058127018299e-05, "loss": 1.3234, "step": 9070 }, { "epoch": 1.6289917473986364, "grad_norm": 9.062819480895996, "learning_rate": 3.372264083243632e-05, "loss": 1.4362, "step": 9080 }, { "epoch": 1.6307857911733046, "grad_norm": 7.770894527435303, "learning_rate": 3.3704700394689634e-05, "loss": 1.3414, "step": 9090 }, { "epoch": 1.6325798349479728, "grad_norm": 6.877737522125244, "learning_rate": 3.368675995694295e-05, "loss": 1.2824, "step": 9100 }, { "epoch": 1.634373878722641, "grad_norm": 8.166614532470703, "learning_rate": 3.366881951919627e-05, "loss": 1.4358, "step": 9110 }, { "epoch": 1.636167922497309, "grad_norm": 8.125406265258789, "learning_rate": 3.3650879081449586e-05, "loss": 1.4451, "step": 9120 }, { "epoch": 1.6379619662719769, "grad_norm": 6.3429412841796875, "learning_rate": 3.363293864370291e-05, "loss": 1.3547, "step": 9130 }, { "epoch": 1.639756010046645, "grad_norm": 9.225153923034668, "learning_rate": 3.361499820595623e-05, "loss": 1.3582, "step": 9140 }, { "epoch": 1.6415500538213132, "grad_norm": 6.7825117111206055, "learning_rate": 3.3597057768209545e-05, "loss": 1.3817, "step": 9150 }, { "epoch": 1.6433440975959814, "grad_norm": 7.816966533660889, "learning_rate": 3.357911733046287e-05, "loss": 1.2981, "step": 9160 }, { "epoch": 1.6451381413706494, "grad_norm": 8.143369674682617, "learning_rate": 3.3561176892716186e-05, "loss": 1.3185, "step": 9170 }, { "epoch": 1.6469321851453176, "grad_norm": 9.256525993347168, "learning_rate": 3.35432364549695e-05, "loss": 1.4458, "step": 9180 }, { "epoch": 1.6487262289199855, "grad_norm": 8.836395263671875, "learning_rate": 3.352529601722282e-05, "loss": 1.3648, "step": 9190 }, { "epoch": 1.6505202726946537, "grad_norm": 10.375576972961426, "learning_rate": 3.350735557947614e-05, "loss": 1.4796, "step": 9200 }, { "epoch": 1.652314316469322, "grad_norm": 8.008833885192871, "learning_rate": 3.3489415141729455e-05, "loss": 1.3293, "step": 9210 }, { "epoch": 1.65410836024399, "grad_norm": 7.188244819641113, "learning_rate": 3.347147470398278e-05, "loss": 1.4948, "step": 9220 }, { "epoch": 1.655902404018658, "grad_norm": 8.808504104614258, "learning_rate": 3.3453534266236096e-05, "loss": 1.4974, "step": 9230 }, { "epoch": 1.657696447793326, "grad_norm": 9.407508850097656, "learning_rate": 3.3435593828489414e-05, "loss": 1.4261, "step": 9240 }, { "epoch": 1.6594904915679942, "grad_norm": 7.234744548797607, "learning_rate": 3.341765339074274e-05, "loss": 1.3784, "step": 9250 }, { "epoch": 1.6612845353426624, "grad_norm": 10.013113021850586, "learning_rate": 3.3399712952996055e-05, "loss": 1.4982, "step": 9260 }, { "epoch": 1.6630785791173306, "grad_norm": 7.584069728851318, "learning_rate": 3.338177251524937e-05, "loss": 1.3648, "step": 9270 }, { "epoch": 1.6648726228919986, "grad_norm": 8.876629829406738, "learning_rate": 3.3363832077502696e-05, "loss": 1.416, "step": 9280 }, { "epoch": 1.6666666666666665, "grad_norm": 6.557656288146973, "learning_rate": 3.3345891639756014e-05, "loss": 1.37, "step": 9290 }, { "epoch": 1.6684607104413347, "grad_norm": 6.436371803283691, "learning_rate": 3.332795120200933e-05, "loss": 1.3965, "step": 9300 }, { "epoch": 1.670254754216003, "grad_norm": 11.4998197555542, "learning_rate": 3.331001076426265e-05, "loss": 1.5301, "step": 9310 }, { "epoch": 1.672048797990671, "grad_norm": 8.139901161193848, "learning_rate": 3.3292070326515965e-05, "loss": 1.4604, "step": 9320 }, { "epoch": 1.673842841765339, "grad_norm": 6.9658074378967285, "learning_rate": 3.327412988876929e-05, "loss": 1.3549, "step": 9330 }, { "epoch": 1.6756368855400072, "grad_norm": 7.35159158706665, "learning_rate": 3.3256189451022607e-05, "loss": 1.3491, "step": 9340 }, { "epoch": 1.6774309293146752, "grad_norm": 8.391031265258789, "learning_rate": 3.3238249013275924e-05, "loss": 1.4108, "step": 9350 }, { "epoch": 1.6792249730893434, "grad_norm": 9.44320011138916, "learning_rate": 3.322030857552925e-05, "loss": 1.46, "step": 9360 }, { "epoch": 1.6810190168640116, "grad_norm": 7.805198669433594, "learning_rate": 3.3202368137782565e-05, "loss": 1.4888, "step": 9370 }, { "epoch": 1.6828130606386797, "grad_norm": 8.305425643920898, "learning_rate": 3.318442770003588e-05, "loss": 1.4868, "step": 9380 }, { "epoch": 1.6846071044133477, "grad_norm": 7.926264762878418, "learning_rate": 3.3166487262289206e-05, "loss": 1.4545, "step": 9390 }, { "epoch": 1.6864011481880157, "grad_norm": 7.380917549133301, "learning_rate": 3.314854682454252e-05, "loss": 1.349, "step": 9400 }, { "epoch": 1.6881951919626839, "grad_norm": 8.860836029052734, "learning_rate": 3.3130606386795834e-05, "loss": 1.3751, "step": 9410 }, { "epoch": 1.689989235737352, "grad_norm": 9.027103424072266, "learning_rate": 3.311266594904916e-05, "loss": 1.4403, "step": 9420 }, { "epoch": 1.6917832795120202, "grad_norm": 7.279397010803223, "learning_rate": 3.3094725511302475e-05, "loss": 1.3976, "step": 9430 }, { "epoch": 1.6935773232866882, "grad_norm": 8.56206226348877, "learning_rate": 3.307678507355579e-05, "loss": 1.3561, "step": 9440 }, { "epoch": 1.6953713670613562, "grad_norm": 7.4955925941467285, "learning_rate": 3.305884463580912e-05, "loss": 1.4076, "step": 9450 }, { "epoch": 1.6971654108360243, "grad_norm": 7.064661026000977, "learning_rate": 3.3040904198062434e-05, "loss": 1.3777, "step": 9460 }, { "epoch": 1.6989594546106925, "grad_norm": 6.131346702575684, "learning_rate": 3.302296376031576e-05, "loss": 1.3927, "step": 9470 }, { "epoch": 1.7007534983853607, "grad_norm": 7.0382866859436035, "learning_rate": 3.3005023322569075e-05, "loss": 1.3617, "step": 9480 }, { "epoch": 1.7025475421600287, "grad_norm": 6.852908134460449, "learning_rate": 3.298708288482239e-05, "loss": 1.3312, "step": 9490 }, { "epoch": 1.7043415859346966, "grad_norm": 8.102025985717773, "learning_rate": 3.296914244707571e-05, "loss": 1.4041, "step": 9500 }, { "epoch": 1.7061356297093648, "grad_norm": 6.2295403480529785, "learning_rate": 3.295120200932903e-05, "loss": 1.4131, "step": 9510 }, { "epoch": 1.707929673484033, "grad_norm": 6.514602184295654, "learning_rate": 3.2933261571582344e-05, "loss": 1.3411, "step": 9520 }, { "epoch": 1.7097237172587012, "grad_norm": 8.146480560302734, "learning_rate": 3.291532113383567e-05, "loss": 1.4027, "step": 9530 }, { "epoch": 1.7115177610333692, "grad_norm": 7.84113883972168, "learning_rate": 3.2897380696088986e-05, "loss": 1.4448, "step": 9540 }, { "epoch": 1.7133118048080374, "grad_norm": 7.640854835510254, "learning_rate": 3.28794402583423e-05, "loss": 1.4225, "step": 9550 }, { "epoch": 1.7151058485827053, "grad_norm": 8.230134963989258, "learning_rate": 3.286149982059563e-05, "loss": 1.4214, "step": 9560 }, { "epoch": 1.7168998923573735, "grad_norm": 9.04704761505127, "learning_rate": 3.2843559382848944e-05, "loss": 1.3879, "step": 9570 }, { "epoch": 1.7186939361320417, "grad_norm": 7.6292619705200195, "learning_rate": 3.282561894510226e-05, "loss": 1.4369, "step": 9580 }, { "epoch": 1.7204879799067099, "grad_norm": 7.765014171600342, "learning_rate": 3.2807678507355585e-05, "loss": 1.4207, "step": 9590 }, { "epoch": 1.7222820236813778, "grad_norm": 8.163656234741211, "learning_rate": 3.27897380696089e-05, "loss": 1.5078, "step": 9600 }, { "epoch": 1.7240760674560458, "grad_norm": 9.697617530822754, "learning_rate": 3.277179763186222e-05, "loss": 1.4383, "step": 9610 }, { "epoch": 1.725870111230714, "grad_norm": 6.583641052246094, "learning_rate": 3.275385719411554e-05, "loss": 1.4345, "step": 9620 }, { "epoch": 1.7276641550053822, "grad_norm": 7.1716437339782715, "learning_rate": 3.2735916756368855e-05, "loss": 1.4338, "step": 9630 }, { "epoch": 1.7294581987800504, "grad_norm": 8.799662590026855, "learning_rate": 3.271797631862217e-05, "loss": 1.4069, "step": 9640 }, { "epoch": 1.7312522425547183, "grad_norm": 8.040757179260254, "learning_rate": 3.2700035880875496e-05, "loss": 1.5181, "step": 9650 }, { "epoch": 1.7330462863293863, "grad_norm": 8.507600784301758, "learning_rate": 3.268209544312881e-05, "loss": 1.4489, "step": 9660 }, { "epoch": 1.7348403301040545, "grad_norm": 7.433869361877441, "learning_rate": 3.266415500538214e-05, "loss": 1.3749, "step": 9670 }, { "epoch": 1.7366343738787227, "grad_norm": 9.173955917358398, "learning_rate": 3.2646214567635454e-05, "loss": 1.325, "step": 9680 }, { "epoch": 1.7384284176533908, "grad_norm": 8.32906436920166, "learning_rate": 3.262827412988877e-05, "loss": 1.4502, "step": 9690 }, { "epoch": 1.7402224614280588, "grad_norm": 7.358117580413818, "learning_rate": 3.261033369214209e-05, "loss": 1.3309, "step": 9700 }, { "epoch": 1.742016505202727, "grad_norm": 6.05413293838501, "learning_rate": 3.2592393254395406e-05, "loss": 1.4092, "step": 9710 }, { "epoch": 1.743810548977395, "grad_norm": 6.9378132820129395, "learning_rate": 3.2574452816648723e-05, "loss": 1.3743, "step": 9720 }, { "epoch": 1.7456045927520631, "grad_norm": 8.749573707580566, "learning_rate": 3.255830642267672e-05, "loss": 1.3122, "step": 9730 }, { "epoch": 1.7473986365267313, "grad_norm": 10.618117332458496, "learning_rate": 3.2540365984930035e-05, "loss": 1.4237, "step": 9740 }, { "epoch": 1.7491926803013995, "grad_norm": 7.997283935546875, "learning_rate": 3.252242554718335e-05, "loss": 1.3925, "step": 9750 }, { "epoch": 1.7509867240760675, "grad_norm": 8.978703498840332, "learning_rate": 3.250448510943667e-05, "loss": 1.3118, "step": 9760 }, { "epoch": 1.7527807678507354, "grad_norm": 7.240395545959473, "learning_rate": 3.248654467168999e-05, "loss": 1.3387, "step": 9770 }, { "epoch": 1.7545748116254036, "grad_norm": 8.210497856140137, "learning_rate": 3.246860423394331e-05, "loss": 1.3689, "step": 9780 }, { "epoch": 1.7563688554000718, "grad_norm": 7.8586344718933105, "learning_rate": 3.245066379619663e-05, "loss": 1.383, "step": 9790 }, { "epoch": 1.75816289917474, "grad_norm": 10.08283805847168, "learning_rate": 3.2432723358449945e-05, "loss": 1.4394, "step": 9800 }, { "epoch": 1.759956942949408, "grad_norm": 9.116758346557617, "learning_rate": 3.241478292070327e-05, "loss": 1.423, "step": 9810 }, { "epoch": 1.761750986724076, "grad_norm": 7.434940814971924, "learning_rate": 3.239684248295659e-05, "loss": 1.3744, "step": 9820 }, { "epoch": 1.7635450304987441, "grad_norm": 10.682246208190918, "learning_rate": 3.2378902045209904e-05, "loss": 1.3878, "step": 9830 }, { "epoch": 1.7653390742734123, "grad_norm": 6.566300868988037, "learning_rate": 3.236096160746323e-05, "loss": 1.431, "step": 9840 }, { "epoch": 1.7671331180480805, "grad_norm": 6.842972278594971, "learning_rate": 3.2343021169716545e-05, "loss": 1.4323, "step": 9850 }, { "epoch": 1.7689271618227485, "grad_norm": 6.9640374183654785, "learning_rate": 3.2325080731969856e-05, "loss": 1.3759, "step": 9860 }, { "epoch": 1.7707212055974164, "grad_norm": 7.9980363845825195, "learning_rate": 3.230714029422318e-05, "loss": 1.3429, "step": 9870 }, { "epoch": 1.7725152493720846, "grad_norm": 7.928410530090332, "learning_rate": 3.22891998564765e-05, "loss": 1.3738, "step": 9880 }, { "epoch": 1.7743092931467528, "grad_norm": 8.214715003967285, "learning_rate": 3.227125941872982e-05, "loss": 1.4069, "step": 9890 }, { "epoch": 1.776103336921421, "grad_norm": 7.336577415466309, "learning_rate": 3.225331898098314e-05, "loss": 1.4363, "step": 9900 }, { "epoch": 1.777897380696089, "grad_norm": 13.41515064239502, "learning_rate": 3.2235378543236456e-05, "loss": 1.4401, "step": 9910 }, { "epoch": 1.7796914244707571, "grad_norm": 8.566363334655762, "learning_rate": 3.221743810548978e-05, "loss": 1.4079, "step": 9920 }, { "epoch": 1.781485468245425, "grad_norm": 8.366653442382812, "learning_rate": 3.21994976677431e-05, "loss": 1.3635, "step": 9930 }, { "epoch": 1.7832795120200933, "grad_norm": 8.311139106750488, "learning_rate": 3.2181557229996414e-05, "loss": 1.4176, "step": 9940 }, { "epoch": 1.7850735557947615, "grad_norm": 9.073042869567871, "learning_rate": 3.216361679224973e-05, "loss": 1.4217, "step": 9950 }, { "epoch": 1.7868675995694296, "grad_norm": 8.806724548339844, "learning_rate": 3.214567635450305e-05, "loss": 1.3746, "step": 9960 }, { "epoch": 1.7886616433440976, "grad_norm": 6.92349910736084, "learning_rate": 3.2127735916756366e-05, "loss": 1.3341, "step": 9970 }, { "epoch": 1.7904556871187656, "grad_norm": 6.854549884796143, "learning_rate": 3.210979547900969e-05, "loss": 1.3557, "step": 9980 }, { "epoch": 1.7922497308934338, "grad_norm": 6.593012809753418, "learning_rate": 3.209185504126301e-05, "loss": 1.4896, "step": 9990 }, { "epoch": 1.794043774668102, "grad_norm": 7.10095739364624, "learning_rate": 3.2073914603516324e-05, "loss": 1.3983, "step": 10000 }, { "epoch": 1.7958378184427701, "grad_norm": 7.834686756134033, "learning_rate": 3.205597416576965e-05, "loss": 1.3975, "step": 10010 }, { "epoch": 1.797631862217438, "grad_norm": 6.41478157043457, "learning_rate": 3.2038033728022966e-05, "loss": 1.3854, "step": 10020 }, { "epoch": 1.799425905992106, "grad_norm": 6.298712253570557, "learning_rate": 3.202009329027628e-05, "loss": 1.3197, "step": 10030 }, { "epoch": 1.8012199497667742, "grad_norm": 8.800332069396973, "learning_rate": 3.200215285252961e-05, "loss": 1.3579, "step": 10040 }, { "epoch": 1.8030139935414424, "grad_norm": 7.364461898803711, "learning_rate": 3.1984212414782924e-05, "loss": 1.3631, "step": 10050 }, { "epoch": 1.8048080373161106, "grad_norm": 6.772685527801514, "learning_rate": 3.196627197703624e-05, "loss": 1.2572, "step": 10060 }, { "epoch": 1.8066020810907786, "grad_norm": 7.126546382904053, "learning_rate": 3.194833153928956e-05, "loss": 1.3785, "step": 10070 }, { "epoch": 1.8083961248654468, "grad_norm": 8.27735424041748, "learning_rate": 3.1930391101542876e-05, "loss": 1.3255, "step": 10080 }, { "epoch": 1.8101901686401147, "grad_norm": 9.396734237670898, "learning_rate": 3.19124506637962e-05, "loss": 1.3753, "step": 10090 }, { "epoch": 1.811984212414783, "grad_norm": 6.9930267333984375, "learning_rate": 3.189451022604952e-05, "loss": 1.4313, "step": 10100 }, { "epoch": 1.813778256189451, "grad_norm": 8.459685325622559, "learning_rate": 3.1876569788302835e-05, "loss": 1.3857, "step": 10110 }, { "epoch": 1.8155722999641193, "grad_norm": 9.824789047241211, "learning_rate": 3.185862935055616e-05, "loss": 1.4138, "step": 10120 }, { "epoch": 1.8173663437387872, "grad_norm": 6.947724342346191, "learning_rate": 3.1840688912809476e-05, "loss": 1.3951, "step": 10130 }, { "epoch": 1.8191603875134552, "grad_norm": 7.63868522644043, "learning_rate": 3.182274847506279e-05, "loss": 1.3397, "step": 10140 }, { "epoch": 1.8209544312881234, "grad_norm": 7.462377071380615, "learning_rate": 3.180480803731612e-05, "loss": 1.3776, "step": 10150 }, { "epoch": 1.8227484750627916, "grad_norm": 8.488241195678711, "learning_rate": 3.178686759956943e-05, "loss": 1.3902, "step": 10160 }, { "epoch": 1.8245425188374598, "grad_norm": 7.8031158447265625, "learning_rate": 3.1768927161822745e-05, "loss": 1.4022, "step": 10170 }, { "epoch": 1.8263365626121277, "grad_norm": 8.310538291931152, "learning_rate": 3.175098672407607e-05, "loss": 1.4432, "step": 10180 }, { "epoch": 1.8281306063867957, "grad_norm": 5.958616733551025, "learning_rate": 3.1733046286329386e-05, "loss": 1.4089, "step": 10190 }, { "epoch": 1.8299246501614639, "grad_norm": 8.99071216583252, "learning_rate": 3.1715105848582703e-05, "loss": 1.4212, "step": 10200 }, { "epoch": 1.831718693936132, "grad_norm": 9.389041900634766, "learning_rate": 3.169716541083603e-05, "loss": 1.3511, "step": 10210 }, { "epoch": 1.8335127377108003, "grad_norm": 9.519119262695312, "learning_rate": 3.1679224973089345e-05, "loss": 1.4521, "step": 10220 }, { "epoch": 1.8353067814854682, "grad_norm": 8.499699592590332, "learning_rate": 3.166128453534266e-05, "loss": 1.2829, "step": 10230 }, { "epoch": 1.8371008252601362, "grad_norm": 8.038819313049316, "learning_rate": 3.1643344097595986e-05, "loss": 1.4808, "step": 10240 }, { "epoch": 1.8388948690348044, "grad_norm": 9.12056827545166, "learning_rate": 3.16254036598493e-05, "loss": 1.4188, "step": 10250 }, { "epoch": 1.8406889128094726, "grad_norm": 9.239951133728027, "learning_rate": 3.160746322210262e-05, "loss": 1.3091, "step": 10260 }, { "epoch": 1.8424829565841407, "grad_norm": 7.837065696716309, "learning_rate": 3.158952278435594e-05, "loss": 1.4372, "step": 10270 }, { "epoch": 1.8442770003588087, "grad_norm": 8.48546314239502, "learning_rate": 3.1571582346609255e-05, "loss": 1.4756, "step": 10280 }, { "epoch": 1.846071044133477, "grad_norm": 7.222843647003174, "learning_rate": 3.155364190886258e-05, "loss": 1.3421, "step": 10290 }, { "epoch": 1.8478650879081449, "grad_norm": 6.328611373901367, "learning_rate": 3.1535701471115896e-05, "loss": 1.3897, "step": 10300 }, { "epoch": 1.849659131682813, "grad_norm": 7.180152416229248, "learning_rate": 3.1517761033369214e-05, "loss": 1.4222, "step": 10310 }, { "epoch": 1.8514531754574812, "grad_norm": 7.584138870239258, "learning_rate": 3.149982059562254e-05, "loss": 1.4284, "step": 10320 }, { "epoch": 1.8532472192321494, "grad_norm": 6.921727180480957, "learning_rate": 3.1481880157875855e-05, "loss": 1.3764, "step": 10330 }, { "epoch": 1.8550412630068174, "grad_norm": 7.458225250244141, "learning_rate": 3.146393972012917e-05, "loss": 1.3253, "step": 10340 }, { "epoch": 1.8568353067814853, "grad_norm": 8.784356117248535, "learning_rate": 3.1445999282382496e-05, "loss": 1.3682, "step": 10350 }, { "epoch": 1.8586293505561535, "grad_norm": 7.8843674659729, "learning_rate": 3.1428058844635813e-05, "loss": 1.3611, "step": 10360 }, { "epoch": 1.8604233943308217, "grad_norm": 6.8401198387146, "learning_rate": 3.141011840688913e-05, "loss": 1.3852, "step": 10370 }, { "epoch": 1.86221743810549, "grad_norm": 7.674398899078369, "learning_rate": 3.139217796914245e-05, "loss": 1.3522, "step": 10380 }, { "epoch": 1.8640114818801579, "grad_norm": 6.873990535736084, "learning_rate": 3.1374237531395765e-05, "loss": 1.3986, "step": 10390 }, { "epoch": 1.8658055256548258, "grad_norm": 6.411556243896484, "learning_rate": 3.135629709364908e-05, "loss": 1.3846, "step": 10400 }, { "epoch": 1.867599569429494, "grad_norm": 7.64192533493042, "learning_rate": 3.1338356655902407e-05, "loss": 1.3235, "step": 10410 }, { "epoch": 1.8693936132041622, "grad_norm": 8.541569709777832, "learning_rate": 3.1320416218155724e-05, "loss": 1.3913, "step": 10420 }, { "epoch": 1.8711876569788304, "grad_norm": 10.887429237365723, "learning_rate": 3.130247578040904e-05, "loss": 1.4691, "step": 10430 }, { "epoch": 1.8729817007534983, "grad_norm": 5.951722145080566, "learning_rate": 3.1284535342662365e-05, "loss": 1.3866, "step": 10440 }, { "epoch": 1.8747757445281665, "grad_norm": 9.592148780822754, "learning_rate": 3.126659490491568e-05, "loss": 1.4277, "step": 10450 }, { "epoch": 1.8765697883028345, "grad_norm": 5.6519060134887695, "learning_rate": 3.1248654467169e-05, "loss": 1.3481, "step": 10460 }, { "epoch": 1.8783638320775027, "grad_norm": 6.1073408126831055, "learning_rate": 3.123071402942232e-05, "loss": 1.4777, "step": 10470 }, { "epoch": 1.8801578758521709, "grad_norm": 6.2250871658325195, "learning_rate": 3.1212773591675634e-05, "loss": 1.5386, "step": 10480 }, { "epoch": 1.881951919626839, "grad_norm": 7.318933486938477, "learning_rate": 3.119483315392896e-05, "loss": 1.3898, "step": 10490 }, { "epoch": 1.883745963401507, "grad_norm": 7.27461576461792, "learning_rate": 3.1176892716182275e-05, "loss": 1.4696, "step": 10500 }, { "epoch": 1.885540007176175, "grad_norm": 7.856091022491455, "learning_rate": 3.115895227843559e-05, "loss": 1.3571, "step": 10510 }, { "epoch": 1.8873340509508432, "grad_norm": 6.904455661773682, "learning_rate": 3.114101184068892e-05, "loss": 1.386, "step": 10520 }, { "epoch": 1.8891280947255114, "grad_norm": 7.072729587554932, "learning_rate": 3.1123071402942234e-05, "loss": 1.3334, "step": 10530 }, { "epoch": 1.8909221385001795, "grad_norm": 7.027041912078857, "learning_rate": 3.110513096519555e-05, "loss": 1.3813, "step": 10540 }, { "epoch": 1.8927161822748475, "grad_norm": 8.132376670837402, "learning_rate": 3.1087190527448875e-05, "loss": 1.3923, "step": 10550 }, { "epoch": 1.8945102260495155, "grad_norm": 7.2380146980285645, "learning_rate": 3.106925008970219e-05, "loss": 1.4301, "step": 10560 }, { "epoch": 1.8963042698241837, "grad_norm": 7.411573886871338, "learning_rate": 3.105130965195551e-05, "loss": 1.4376, "step": 10570 }, { "epoch": 1.8980983135988518, "grad_norm": 8.479894638061523, "learning_rate": 3.103336921420883e-05, "loss": 1.3932, "step": 10580 }, { "epoch": 1.89989235737352, "grad_norm": 5.64423942565918, "learning_rate": 3.1015428776462144e-05, "loss": 1.4356, "step": 10590 }, { "epoch": 1.901686401148188, "grad_norm": 7.496821403503418, "learning_rate": 3.099748833871546e-05, "loss": 1.3875, "step": 10600 }, { "epoch": 1.903480444922856, "grad_norm": 8.228056907653809, "learning_rate": 3.0979547900968786e-05, "loss": 1.4076, "step": 10610 }, { "epoch": 1.9052744886975241, "grad_norm": 6.483157634735107, "learning_rate": 3.09616074632221e-05, "loss": 1.3674, "step": 10620 }, { "epoch": 1.9070685324721923, "grad_norm": 7.570273399353027, "learning_rate": 3.094366702547543e-05, "loss": 1.3251, "step": 10630 }, { "epoch": 1.9088625762468605, "grad_norm": 9.55526065826416, "learning_rate": 3.0925726587728744e-05, "loss": 1.4677, "step": 10640 }, { "epoch": 1.9106566200215285, "grad_norm": 7.437054634094238, "learning_rate": 3.090778614998206e-05, "loss": 1.3634, "step": 10650 }, { "epoch": 1.9124506637961967, "grad_norm": 8.50203800201416, "learning_rate": 3.0889845712235385e-05, "loss": 1.406, "step": 10660 }, { "epoch": 1.9142447075708646, "grad_norm": 7.820067405700684, "learning_rate": 3.08719052744887e-05, "loss": 1.4127, "step": 10670 }, { "epoch": 1.9160387513455328, "grad_norm": 8.92634105682373, "learning_rate": 3.085396483674201e-05, "loss": 1.3228, "step": 10680 }, { "epoch": 1.917832795120201, "grad_norm": 7.616683006286621, "learning_rate": 3.083602439899534e-05, "loss": 1.4236, "step": 10690 }, { "epoch": 1.9196268388948692, "grad_norm": 7.309317588806152, "learning_rate": 3.0818083961248654e-05, "loss": 1.3839, "step": 10700 }, { "epoch": 1.9214208826695371, "grad_norm": 9.383467674255371, "learning_rate": 3.080014352350197e-05, "loss": 1.4143, "step": 10710 }, { "epoch": 1.923214926444205, "grad_norm": 7.853944778442383, "learning_rate": 3.0782203085755296e-05, "loss": 1.4278, "step": 10720 }, { "epoch": 1.9250089702188733, "grad_norm": 7.185943603515625, "learning_rate": 3.076426264800861e-05, "loss": 1.3744, "step": 10730 }, { "epoch": 1.9268030139935415, "grad_norm": 6.492997646331787, "learning_rate": 3.074632221026193e-05, "loss": 1.3643, "step": 10740 }, { "epoch": 1.9285970577682097, "grad_norm": 11.078304290771484, "learning_rate": 3.0728381772515254e-05, "loss": 1.3984, "step": 10750 }, { "epoch": 1.9303911015428776, "grad_norm": 7.419142246246338, "learning_rate": 3.071044133476857e-05, "loss": 1.4019, "step": 10760 }, { "epoch": 1.9321851453175456, "grad_norm": 6.431114673614502, "learning_rate": 3.069250089702189e-05, "loss": 1.3596, "step": 10770 }, { "epoch": 1.9339791890922138, "grad_norm": 7.534803867340088, "learning_rate": 3.0674560459275206e-05, "loss": 1.3333, "step": 10780 }, { "epoch": 1.935773232866882, "grad_norm": 9.282038688659668, "learning_rate": 3.065662002152852e-05, "loss": 1.4125, "step": 10790 }, { "epoch": 1.9375672766415502, "grad_norm": 7.056236267089844, "learning_rate": 3.063867958378184e-05, "loss": 1.4697, "step": 10800 }, { "epoch": 1.9393613204162181, "grad_norm": 7.336516380310059, "learning_rate": 3.0620739146035165e-05, "loss": 1.4819, "step": 10810 }, { "epoch": 1.9411553641908863, "grad_norm": 7.675684928894043, "learning_rate": 3.060279870828848e-05, "loss": 1.2941, "step": 10820 }, { "epoch": 1.9429494079655543, "grad_norm": 6.298435211181641, "learning_rate": 3.0584858270541806e-05, "loss": 1.3083, "step": 10830 }, { "epoch": 1.9447434517402225, "grad_norm": 8.348226547241211, "learning_rate": 3.056691783279512e-05, "loss": 1.4821, "step": 10840 }, { "epoch": 1.9465374955148906, "grad_norm": 7.188017845153809, "learning_rate": 3.054897739504844e-05, "loss": 1.4011, "step": 10850 }, { "epoch": 1.9483315392895588, "grad_norm": 7.320749759674072, "learning_rate": 3.0531036957301765e-05, "loss": 1.4383, "step": 10860 }, { "epoch": 1.9501255830642268, "grad_norm": 8.11515998840332, "learning_rate": 3.051309651955508e-05, "loss": 1.4512, "step": 10870 }, { "epoch": 1.9519196268388948, "grad_norm": 5.978071689605713, "learning_rate": 3.0495156081808396e-05, "loss": 1.3383, "step": 10880 }, { "epoch": 1.953713670613563, "grad_norm": 8.691400527954102, "learning_rate": 3.047721564406172e-05, "loss": 1.3759, "step": 10890 }, { "epoch": 1.9555077143882311, "grad_norm": 7.324152946472168, "learning_rate": 3.0459275206315037e-05, "loss": 1.4581, "step": 10900 }, { "epoch": 1.9573017581628993, "grad_norm": 7.364433288574219, "learning_rate": 3.0441334768568354e-05, "loss": 1.356, "step": 10910 }, { "epoch": 1.9590958019375673, "grad_norm": 8.051295280456543, "learning_rate": 3.0423394330821675e-05, "loss": 1.3423, "step": 10920 }, { "epoch": 1.9608898457122352, "grad_norm": 7.994751930236816, "learning_rate": 3.0405453893074992e-05, "loss": 1.4558, "step": 10930 }, { "epoch": 1.9626838894869034, "grad_norm": 7.786356449127197, "learning_rate": 3.038751345532831e-05, "loss": 1.3661, "step": 10940 }, { "epoch": 1.9644779332615716, "grad_norm": 7.44194221496582, "learning_rate": 3.0369573017581633e-05, "loss": 1.3659, "step": 10950 }, { "epoch": 1.9662719770362398, "grad_norm": 9.217162132263184, "learning_rate": 3.0351632579834947e-05, "loss": 1.4716, "step": 10960 }, { "epoch": 1.9680660208109078, "grad_norm": 7.545557022094727, "learning_rate": 3.0333692142088265e-05, "loss": 1.4073, "step": 10970 }, { "epoch": 1.969860064585576, "grad_norm": 7.098927021026611, "learning_rate": 3.031575170434159e-05, "loss": 1.3973, "step": 10980 }, { "epoch": 1.971654108360244, "grad_norm": 8.295470237731934, "learning_rate": 3.0297811266594906e-05, "loss": 1.3848, "step": 10990 }, { "epoch": 1.973448152134912, "grad_norm": 5.602086544036865, "learning_rate": 3.0279870828848226e-05, "loss": 1.3453, "step": 11000 }, { "epoch": 1.9752421959095803, "grad_norm": 7.365663051605225, "learning_rate": 3.0261930391101544e-05, "loss": 1.3177, "step": 11010 }, { "epoch": 1.9770362396842482, "grad_norm": 9.722752571105957, "learning_rate": 3.024398995335486e-05, "loss": 1.4265, "step": 11020 }, { "epoch": 1.9788302834589164, "grad_norm": 7.279769420623779, "learning_rate": 3.0226049515608185e-05, "loss": 1.33, "step": 11030 }, { "epoch": 1.9806243272335844, "grad_norm": 7.57685661315918, "learning_rate": 3.0208109077861502e-05, "loss": 1.4305, "step": 11040 }, { "epoch": 1.9824183710082526, "grad_norm": 8.122096061706543, "learning_rate": 3.019016864011482e-05, "loss": 1.3327, "step": 11050 }, { "epoch": 1.9842124147829208, "grad_norm": 9.706563949584961, "learning_rate": 3.017222820236814e-05, "loss": 1.4532, "step": 11060 }, { "epoch": 1.986006458557589, "grad_norm": 9.653244972229004, "learning_rate": 3.0154287764621457e-05, "loss": 1.4331, "step": 11070 }, { "epoch": 1.987800502332257, "grad_norm": 6.598245620727539, "learning_rate": 3.0136347326874775e-05, "loss": 1.3661, "step": 11080 }, { "epoch": 1.9895945461069249, "grad_norm": 8.551617622375488, "learning_rate": 3.01184068891281e-05, "loss": 1.3247, "step": 11090 }, { "epoch": 1.991388589881593, "grad_norm": 6.94644021987915, "learning_rate": 3.0100466451381416e-05, "loss": 1.376, "step": 11100 }, { "epoch": 1.9931826336562612, "grad_norm": 8.096078872680664, "learning_rate": 3.0082526013634733e-05, "loss": 1.4114, "step": 11110 }, { "epoch": 1.9949766774309294, "grad_norm": 9.47823429107666, "learning_rate": 3.0064585575888054e-05, "loss": 1.4147, "step": 11120 }, { "epoch": 1.9967707212055974, "grad_norm": 6.910612106323242, "learning_rate": 3.004664513814137e-05, "loss": 1.365, "step": 11130 }, { "epoch": 1.9985647649802654, "grad_norm": 9.393196105957031, "learning_rate": 3.002870470039469e-05, "loss": 1.4358, "step": 11140 }, { "epoch": 2.0, "eval_loss": 1.9240386486053467, "eval_runtime": 199.4981, "eval_samples_per_second": 12.416, "eval_steps_per_second": 12.416, "step": 11148 } ], "logging_steps": 10, "max_steps": 27870, "num_input_tokens_seen": 0, "num_train_epochs": 5, "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": 6.033003108315955e+16, "train_batch_size": 4, "trial_name": null, "trial_params": null }