{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.0, "eval_steps": 500, "global_step": 932, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.004308023694130318, "grad_norm": 36.38650131225586, "learning_rate": 9.9999954157983e-06, "loss": 2.3411, "step": 1 }, { "epoch": 0.008616047388260635, "grad_norm": 24.211559295654297, "learning_rate": 9.999981663201606e-06, "loss": 1.9732, "step": 2 }, { "epoch": 0.012924071082390954, "grad_norm": 15.806609153747559, "learning_rate": 9.999958742235133e-06, "loss": 1.6866, "step": 3 }, { "epoch": 0.01723209477652127, "grad_norm": 11.639154434204102, "learning_rate": 9.999926652940914e-06, "loss": 1.5735, "step": 4 }, { "epoch": 0.021540118470651588, "grad_norm": 13.66666316986084, "learning_rate": 9.999885395377788e-06, "loss": 1.3392, "step": 5 }, { "epoch": 0.025848142164781908, "grad_norm": 6.430898189544678, "learning_rate": 9.999834969621408e-06, "loss": 1.2447, "step": 6 }, { "epoch": 0.030156165858912225, "grad_norm": 5.439924240112305, "learning_rate": 9.999775375764244e-06, "loss": 1.1359, "step": 7 }, { "epoch": 0.03446418955304254, "grad_norm": 5.654125213623047, "learning_rate": 9.999706613915567e-06, "loss": 1.059, "step": 8 }, { "epoch": 0.03877221324717286, "grad_norm": 5.610560417175293, "learning_rate": 9.999628684201464e-06, "loss": 0.9955, "step": 9 }, { "epoch": 0.043080236941303175, "grad_norm": 12.181400299072266, "learning_rate": 9.999541586764836e-06, "loss": 1.0143, "step": 10 }, { "epoch": 0.04738826063543349, "grad_norm": 6.2043657302856445, "learning_rate": 9.999445321765392e-06, "loss": 0.9816, "step": 11 }, { "epoch": 0.051696284329563816, "grad_norm": 4.889285564422607, "learning_rate": 9.999339889379647e-06, "loss": 0.911, "step": 12 }, { "epoch": 0.05600430802369413, "grad_norm": 6.038234710693359, "learning_rate": 9.999225289800935e-06, "loss": 0.878, "step": 13 }, { "epoch": 0.06031233171782445, "grad_norm": 6.304042339324951, "learning_rate": 9.999101523239392e-06, "loss": 1.0026, "step": 14 }, { "epoch": 0.06462035541195477, "grad_norm": 4.621156692504883, "learning_rate": 9.998968589921969e-06, "loss": 0.8617, "step": 15 }, { "epoch": 0.06892837910608508, "grad_norm": 5.013041019439697, "learning_rate": 9.99882649009242e-06, "loss": 0.8654, "step": 16 }, { "epoch": 0.0732364028002154, "grad_norm": 4.962759971618652, "learning_rate": 9.998675224011317e-06, "loss": 0.8501, "step": 17 }, { "epoch": 0.07754442649434572, "grad_norm": 5.091447353363037, "learning_rate": 9.998514791956025e-06, "loss": 0.8793, "step": 18 }, { "epoch": 0.08185245018847603, "grad_norm": 5.140117645263672, "learning_rate": 9.998345194220732e-06, "loss": 0.8089, "step": 19 }, { "epoch": 0.08616047388260635, "grad_norm": 5.672287464141846, "learning_rate": 9.998166431116421e-06, "loss": 0.7688, "step": 20 }, { "epoch": 0.09046849757673667, "grad_norm": 5.412152290344238, "learning_rate": 9.99797850297089e-06, "loss": 0.8008, "step": 21 }, { "epoch": 0.09477652127086698, "grad_norm": 5.4089531898498535, "learning_rate": 9.997781410128737e-06, "loss": 0.7819, "step": 22 }, { "epoch": 0.0990845449649973, "grad_norm": 5.294145584106445, "learning_rate": 9.99757515295137e-06, "loss": 0.8042, "step": 23 }, { "epoch": 0.10339256865912763, "grad_norm": 5.16695499420166, "learning_rate": 9.997359731816998e-06, "loss": 0.8092, "step": 24 }, { "epoch": 0.10770059235325795, "grad_norm": 5.167812824249268, "learning_rate": 9.997135147120633e-06, "loss": 0.7504, "step": 25 }, { "epoch": 0.11200861604738827, "grad_norm": 5.223298072814941, "learning_rate": 9.996901399274093e-06, "loss": 0.7643, "step": 26 }, { "epoch": 0.11631663974151858, "grad_norm": 5.137231349945068, "learning_rate": 9.996658488705997e-06, "loss": 0.7371, "step": 27 }, { "epoch": 0.1206246634356489, "grad_norm": 5.101226329803467, "learning_rate": 9.996406415861763e-06, "loss": 0.725, "step": 28 }, { "epoch": 0.12493268712977922, "grad_norm": 5.096724510192871, "learning_rate": 9.996145181203616e-06, "loss": 0.6918, "step": 29 }, { "epoch": 0.12924071082390953, "grad_norm": 5.091918468475342, "learning_rate": 9.995874785210573e-06, "loss": 0.7093, "step": 30 }, { "epoch": 0.13354873451803984, "grad_norm": 5.17004919052124, "learning_rate": 9.995595228378456e-06, "loss": 0.6821, "step": 31 }, { "epoch": 0.13785675821217017, "grad_norm": 5.189696788787842, "learning_rate": 9.995306511219885e-06, "loss": 0.7004, "step": 32 }, { "epoch": 0.1421647819063005, "grad_norm": 5.040393829345703, "learning_rate": 9.995008634264272e-06, "loss": 0.69, "step": 33 }, { "epoch": 0.1464728056004308, "grad_norm": 4.890658378601074, "learning_rate": 9.994701598057828e-06, "loss": 0.6587, "step": 34 }, { "epoch": 0.15078082929456113, "grad_norm": 5.024774074554443, "learning_rate": 9.99438540316356e-06, "loss": 0.6554, "step": 35 }, { "epoch": 0.15508885298869143, "grad_norm": 4.879336833953857, "learning_rate": 9.99406005016127e-06, "loss": 0.6545, "step": 36 }, { "epoch": 0.15939687668282176, "grad_norm": 4.925673484802246, "learning_rate": 9.99372553964755e-06, "loss": 0.6253, "step": 37 }, { "epoch": 0.16370490037695207, "grad_norm": 4.918234825134277, "learning_rate": 9.993381872235785e-06, "loss": 0.6022, "step": 38 }, { "epoch": 0.1680129240710824, "grad_norm": 4.864017009735107, "learning_rate": 9.993029048556154e-06, "loss": 0.6211, "step": 39 }, { "epoch": 0.1723209477652127, "grad_norm": 4.827836990356445, "learning_rate": 9.99266706925562e-06, "loss": 0.6532, "step": 40 }, { "epoch": 0.17662897145934303, "grad_norm": 4.730919361114502, "learning_rate": 9.99229593499794e-06, "loss": 0.6282, "step": 41 }, { "epoch": 0.18093699515347333, "grad_norm": 4.74766731262207, "learning_rate": 9.991915646463652e-06, "loss": 0.5985, "step": 42 }, { "epoch": 0.18524501884760367, "grad_norm": 4.6774001121521, "learning_rate": 9.991526204350087e-06, "loss": 0.6114, "step": 43 }, { "epoch": 0.18955304254173397, "grad_norm": 4.586522579193115, "learning_rate": 9.991127609371357e-06, "loss": 0.5937, "step": 44 }, { "epoch": 0.1938610662358643, "grad_norm": 4.603926658630371, "learning_rate": 9.990719862258357e-06, "loss": 0.5893, "step": 45 }, { "epoch": 0.1981690899299946, "grad_norm": 4.538887023925781, "learning_rate": 9.990302963758765e-06, "loss": 0.5772, "step": 46 }, { "epoch": 0.20247711362412493, "grad_norm": 4.549485206604004, "learning_rate": 9.989876914637042e-06, "loss": 0.5919, "step": 47 }, { "epoch": 0.20678513731825526, "grad_norm": 4.3624186515808105, "learning_rate": 9.989441715674422e-06, "loss": 0.5612, "step": 48 }, { "epoch": 0.21109316101238557, "grad_norm": 4.213372707366943, "learning_rate": 9.988997367668924e-06, "loss": 0.5322, "step": 49 }, { "epoch": 0.2154011847065159, "grad_norm": 4.253533363342285, "learning_rate": 9.988543871435342e-06, "loss": 0.5735, "step": 50 }, { "epoch": 0.2197092084006462, "grad_norm": 4.111804962158203, "learning_rate": 9.988081227805237e-06, "loss": 0.5542, "step": 51 }, { "epoch": 0.22401723209477653, "grad_norm": 4.026226997375488, "learning_rate": 9.987609437626955e-06, "loss": 0.5325, "step": 52 }, { "epoch": 0.22832525578890683, "grad_norm": 3.915330171585083, "learning_rate": 9.987128501765606e-06, "loss": 0.5097, "step": 53 }, { "epoch": 0.23263327948303716, "grad_norm": 3.7279622554779053, "learning_rate": 9.986638421103074e-06, "loss": 0.4823, "step": 54 }, { "epoch": 0.23694130317716747, "grad_norm": 3.59734845161438, "learning_rate": 9.986139196538011e-06, "loss": 0.5023, "step": 55 }, { "epoch": 0.2412493268712978, "grad_norm": 3.560422658920288, "learning_rate": 9.985630828985835e-06, "loss": 0.5042, "step": 56 }, { "epoch": 0.2455573505654281, "grad_norm": 3.475142478942871, "learning_rate": 9.98511331937873e-06, "loss": 0.484, "step": 57 }, { "epoch": 0.24986537425955843, "grad_norm": 3.38651704788208, "learning_rate": 9.984586668665641e-06, "loss": 0.5207, "step": 58 }, { "epoch": 0.25417339795368876, "grad_norm": 3.1024978160858154, "learning_rate": 9.98405087781228e-06, "loss": 0.4747, "step": 59 }, { "epoch": 0.25848142164781907, "grad_norm": 3.198146343231201, "learning_rate": 9.983505947801115e-06, "loss": 0.5192, "step": 60 }, { "epoch": 0.26278944534194937, "grad_norm": 2.8837313652038574, "learning_rate": 9.982951879631373e-06, "loss": 0.4547, "step": 61 }, { "epoch": 0.26709746903607967, "grad_norm": 2.832683563232422, "learning_rate": 9.982388674319041e-06, "loss": 0.4637, "step": 62 }, { "epoch": 0.27140549273021003, "grad_norm": 2.7181856632232666, "learning_rate": 9.981816332896854e-06, "loss": 0.5033, "step": 63 }, { "epoch": 0.27571351642434033, "grad_norm": 2.530658483505249, "learning_rate": 9.981234856414306e-06, "loss": 0.4659, "step": 64 }, { "epoch": 0.28002154011847064, "grad_norm": 2.3317317962646484, "learning_rate": 9.98064424593764e-06, "loss": 0.4334, "step": 65 }, { "epoch": 0.284329563812601, "grad_norm": 2.312959909439087, "learning_rate": 9.980044502549843e-06, "loss": 0.4418, "step": 66 }, { "epoch": 0.2886375875067313, "grad_norm": 2.3766074180603027, "learning_rate": 9.979435627350658e-06, "loss": 0.5132, "step": 67 }, { "epoch": 0.2929456112008616, "grad_norm": 2.692957639694214, "learning_rate": 9.978817621456562e-06, "loss": 0.5016, "step": 68 }, { "epoch": 0.2972536348949919, "grad_norm": 2.4902384281158447, "learning_rate": 9.978190486000784e-06, "loss": 0.4641, "step": 69 }, { "epoch": 0.30156165858912226, "grad_norm": 2.1922872066497803, "learning_rate": 9.977554222133293e-06, "loss": 0.4562, "step": 70 }, { "epoch": 0.30586968228325256, "grad_norm": 2.310131549835205, "learning_rate": 9.976908831020787e-06, "loss": 0.4693, "step": 71 }, { "epoch": 0.31017770597738287, "grad_norm": 2.3880672454833984, "learning_rate": 9.97625431384671e-06, "loss": 0.5, "step": 72 }, { "epoch": 0.31448572967151317, "grad_norm": 2.156353235244751, "learning_rate": 9.975590671811239e-06, "loss": 0.4648, "step": 73 }, { "epoch": 0.31879375336564353, "grad_norm": 2.0981221199035645, "learning_rate": 9.974917906131283e-06, "loss": 0.4619, "step": 74 }, { "epoch": 0.32310177705977383, "grad_norm": 2.058878183364868, "learning_rate": 9.974236018040476e-06, "loss": 0.4485, "step": 75 }, { "epoch": 0.32740980075390413, "grad_norm": 1.9989655017852783, "learning_rate": 9.973545008789182e-06, "loss": 0.4616, "step": 76 }, { "epoch": 0.33171782444803444, "grad_norm": 2.109799385070801, "learning_rate": 9.972844879644494e-06, "loss": 0.4556, "step": 77 }, { "epoch": 0.3360258481421648, "grad_norm": 1.9261740446090698, "learning_rate": 9.972135631890226e-06, "loss": 0.4614, "step": 78 }, { "epoch": 0.3403338718362951, "grad_norm": 2.0897579193115234, "learning_rate": 9.97141726682691e-06, "loss": 0.4745, "step": 79 }, { "epoch": 0.3446418955304254, "grad_norm": 1.7466074228286743, "learning_rate": 9.970689785771798e-06, "loss": 0.435, "step": 80 }, { "epoch": 0.34894991922455576, "grad_norm": 2.041557550430298, "learning_rate": 9.969953190058861e-06, "loss": 0.4867, "step": 81 }, { "epoch": 0.35325794291868606, "grad_norm": 2.0437612533569336, "learning_rate": 9.969207481038776e-06, "loss": 0.4956, "step": 82 }, { "epoch": 0.35756596661281637, "grad_norm": 1.7604644298553467, "learning_rate": 9.968452660078939e-06, "loss": 0.4568, "step": 83 }, { "epoch": 0.36187399030694667, "grad_norm": 2.0843091011047363, "learning_rate": 9.967688728563446e-06, "loss": 0.4708, "step": 84 }, { "epoch": 0.366182014001077, "grad_norm": 2.1242294311523438, "learning_rate": 9.966915687893109e-06, "loss": 0.4753, "step": 85 }, { "epoch": 0.37049003769520733, "grad_norm": 1.9133800268173218, "learning_rate": 9.966133539485435e-06, "loss": 0.4614, "step": 86 }, { "epoch": 0.37479806138933763, "grad_norm": 1.963077187538147, "learning_rate": 9.965342284774633e-06, "loss": 0.4293, "step": 87 }, { "epoch": 0.37910608508346794, "grad_norm": 2.119835615158081, "learning_rate": 9.964541925211613e-06, "loss": 0.4301, "step": 88 }, { "epoch": 0.3834141087775983, "grad_norm": 2.19815993309021, "learning_rate": 9.963732462263979e-06, "loss": 0.4519, "step": 89 }, { "epoch": 0.3877221324717286, "grad_norm": 2.0409700870513916, "learning_rate": 9.962913897416029e-06, "loss": 0.4018, "step": 90 }, { "epoch": 0.3920301561658589, "grad_norm": 2.0001046657562256, "learning_rate": 9.962086232168747e-06, "loss": 0.4075, "step": 91 }, { "epoch": 0.3963381798599892, "grad_norm": 2.015749216079712, "learning_rate": 9.961249468039806e-06, "loss": 0.3956, "step": 92 }, { "epoch": 0.40064620355411956, "grad_norm": 2.2066595554351807, "learning_rate": 9.960403606563568e-06, "loss": 0.4533, "step": 93 }, { "epoch": 0.40495422724824987, "grad_norm": 2.2596445083618164, "learning_rate": 9.959548649291071e-06, "loss": 0.4448, "step": 94 }, { "epoch": 0.40926225094238017, "grad_norm": 2.2961678504943848, "learning_rate": 9.958684597790031e-06, "loss": 0.4805, "step": 95 }, { "epoch": 0.4135702746365105, "grad_norm": 2.1837716102600098, "learning_rate": 9.957811453644848e-06, "loss": 0.4544, "step": 96 }, { "epoch": 0.41787829833064083, "grad_norm": 2.129624128341675, "learning_rate": 9.956929218456586e-06, "loss": 0.4367, "step": 97 }, { "epoch": 0.42218632202477113, "grad_norm": 2.169912099838257, "learning_rate": 9.956037893842982e-06, "loss": 0.4436, "step": 98 }, { "epoch": 0.42649434571890144, "grad_norm": 2.217552423477173, "learning_rate": 9.955137481438442e-06, "loss": 0.4619, "step": 99 }, { "epoch": 0.4308023694130318, "grad_norm": 2.1200716495513916, "learning_rate": 9.954227982894034e-06, "loss": 0.397, "step": 100 }, { "epoch": 0.4351103931071621, "grad_norm": 2.134117603302002, "learning_rate": 9.953309399877491e-06, "loss": 0.4445, "step": 101 }, { "epoch": 0.4394184168012924, "grad_norm": 1.9783546924591064, "learning_rate": 9.952381734073197e-06, "loss": 0.4309, "step": 102 }, { "epoch": 0.4437264404954227, "grad_norm": 2.1107230186462402, "learning_rate": 9.951444987182195e-06, "loss": 0.4189, "step": 103 }, { "epoch": 0.44803446418955306, "grad_norm": 1.987900972366333, "learning_rate": 9.950499160922184e-06, "loss": 0.3838, "step": 104 }, { "epoch": 0.45234248788368336, "grad_norm": 2.0224177837371826, "learning_rate": 9.949544257027503e-06, "loss": 0.4226, "step": 105 }, { "epoch": 0.45665051157781367, "grad_norm": 2.0050714015960693, "learning_rate": 9.948580277249142e-06, "loss": 0.3746, "step": 106 }, { "epoch": 0.46095853527194397, "grad_norm": 2.179802894592285, "learning_rate": 9.947607223354731e-06, "loss": 0.4444, "step": 107 }, { "epoch": 0.46526655896607433, "grad_norm": 1.9253278970718384, "learning_rate": 9.946625097128544e-06, "loss": 0.42, "step": 108 }, { "epoch": 0.46957458266020463, "grad_norm": 2.0330090522766113, "learning_rate": 9.945633900371483e-06, "loss": 0.4054, "step": 109 }, { "epoch": 0.47388260635433493, "grad_norm": 2.1246986389160156, "learning_rate": 9.94463363490109e-06, "loss": 0.4037, "step": 110 }, { "epoch": 0.4781906300484653, "grad_norm": 1.877738356590271, "learning_rate": 9.943624302551527e-06, "loss": 0.3897, "step": 111 }, { "epoch": 0.4824986537425956, "grad_norm": 2.0079073905944824, "learning_rate": 9.942605905173593e-06, "loss": 0.4023, "step": 112 }, { "epoch": 0.4868066774367259, "grad_norm": 1.9090015888214111, "learning_rate": 9.941578444634699e-06, "loss": 0.3678, "step": 113 }, { "epoch": 0.4911147011308562, "grad_norm": 1.9314876794815063, "learning_rate": 9.940541922818882e-06, "loss": 0.3823, "step": 114 }, { "epoch": 0.49542272482498656, "grad_norm": 1.9568986892700195, "learning_rate": 9.939496341626791e-06, "loss": 0.4022, "step": 115 }, { "epoch": 0.49973074851911686, "grad_norm": 2.0613176822662354, "learning_rate": 9.938441702975689e-06, "loss": 0.4223, "step": 116 }, { "epoch": 0.5040387722132472, "grad_norm": 1.795121669769287, "learning_rate": 9.937378008799448e-06, "loss": 0.3711, "step": 117 }, { "epoch": 0.5083467959073775, "grad_norm": 1.8435457944869995, "learning_rate": 9.93630526104854e-06, "loss": 0.3736, "step": 118 }, { "epoch": 0.5126548196015078, "grad_norm": 1.8587696552276611, "learning_rate": 9.935223461690042e-06, "loss": 0.3741, "step": 119 }, { "epoch": 0.5169628432956381, "grad_norm": 1.8080061674118042, "learning_rate": 9.934132612707631e-06, "loss": 0.3473, "step": 120 }, { "epoch": 0.5212708669897684, "grad_norm": 1.8691555261611938, "learning_rate": 9.933032716101576e-06, "loss": 0.4004, "step": 121 }, { "epoch": 0.5255788906838987, "grad_norm": 1.8062400817871094, "learning_rate": 9.931923773888734e-06, "loss": 0.3545, "step": 122 }, { "epoch": 0.529886914378029, "grad_norm": 1.9322127103805542, "learning_rate": 9.930805788102551e-06, "loss": 0.4011, "step": 123 }, { "epoch": 0.5341949380721593, "grad_norm": 1.6584973335266113, "learning_rate": 9.929678760793057e-06, "loss": 0.3217, "step": 124 }, { "epoch": 0.5385029617662898, "grad_norm": 1.9472841024398804, "learning_rate": 9.928542694026862e-06, "loss": 0.4014, "step": 125 }, { "epoch": 0.5428109854604201, "grad_norm": 1.9014495611190796, "learning_rate": 9.927397589887144e-06, "loss": 0.3769, "step": 126 }, { "epoch": 0.5471190091545504, "grad_norm": 1.8707023859024048, "learning_rate": 9.926243450473664e-06, "loss": 0.3558, "step": 127 }, { "epoch": 0.5514270328486807, "grad_norm": 1.8487975597381592, "learning_rate": 9.925080277902743e-06, "loss": 0.371, "step": 128 }, { "epoch": 0.555735056542811, "grad_norm": 2.04178786277771, "learning_rate": 9.923908074307267e-06, "loss": 0.3681, "step": 129 }, { "epoch": 0.5600430802369413, "grad_norm": 1.6615709066390991, "learning_rate": 9.922726841836685e-06, "loss": 0.3386, "step": 130 }, { "epoch": 0.5643511039310716, "grad_norm": 1.7785009145736694, "learning_rate": 9.921536582657002e-06, "loss": 0.3709, "step": 131 }, { "epoch": 0.568659127625202, "grad_norm": 1.714189052581787, "learning_rate": 9.920337298950767e-06, "loss": 0.3761, "step": 132 }, { "epoch": 0.5729671513193323, "grad_norm": 1.8931145668029785, "learning_rate": 9.919128992917086e-06, "loss": 0.3982, "step": 133 }, { "epoch": 0.5772751750134626, "grad_norm": 1.7474888563156128, "learning_rate": 9.91791166677161e-06, "loss": 0.3756, "step": 134 }, { "epoch": 0.5815831987075929, "grad_norm": 2.088261127471924, "learning_rate": 9.916685322746524e-06, "loss": 0.4246, "step": 135 }, { "epoch": 0.5858912224017232, "grad_norm": 1.5932164192199707, "learning_rate": 9.915449963090551e-06, "loss": 0.3746, "step": 136 }, { "epoch": 0.5901992460958535, "grad_norm": 1.5779109001159668, "learning_rate": 9.914205590068946e-06, "loss": 0.3605, "step": 137 }, { "epoch": 0.5945072697899838, "grad_norm": 1.5977842807769775, "learning_rate": 9.912952205963491e-06, "loss": 0.3507, "step": 138 }, { "epoch": 0.5988152934841141, "grad_norm": 1.6703522205352783, "learning_rate": 9.911689813072495e-06, "loss": 0.367, "step": 139 }, { "epoch": 0.6031233171782445, "grad_norm": 1.654060959815979, "learning_rate": 9.91041841371078e-06, "loss": 0.3543, "step": 140 }, { "epoch": 0.6074313408723748, "grad_norm": 1.6373342275619507, "learning_rate": 9.90913801020969e-06, "loss": 0.3789, "step": 141 }, { "epoch": 0.6117393645665051, "grad_norm": 1.8475476503372192, "learning_rate": 9.907848604917075e-06, "loss": 0.4124, "step": 142 }, { "epoch": 0.6160473882606354, "grad_norm": 1.4528111219406128, "learning_rate": 9.906550200197288e-06, "loss": 0.3446, "step": 143 }, { "epoch": 0.6203554119547657, "grad_norm": 1.4967495203018188, "learning_rate": 9.905242798431196e-06, "loss": 0.3558, "step": 144 }, { "epoch": 0.624663435648896, "grad_norm": 1.6655304431915283, "learning_rate": 9.903926402016153e-06, "loss": 0.3542, "step": 145 }, { "epoch": 0.6289714593430263, "grad_norm": 1.6033902168273926, "learning_rate": 9.902601013366009e-06, "loss": 0.3899, "step": 146 }, { "epoch": 0.6332794830371568, "grad_norm": 1.4599045515060425, "learning_rate": 9.901266634911104e-06, "loss": 0.3521, "step": 147 }, { "epoch": 0.6375875067312871, "grad_norm": 1.4676603078842163, "learning_rate": 9.899923269098262e-06, "loss": 0.3752, "step": 148 }, { "epoch": 0.6418955304254174, "grad_norm": 1.7495596408843994, "learning_rate": 9.898570918390789e-06, "loss": 0.4108, "step": 149 }, { "epoch": 0.6462035541195477, "grad_norm": 1.5633522272109985, "learning_rate": 9.897209585268459e-06, "loss": 0.3587, "step": 150 }, { "epoch": 0.650511577813678, "grad_norm": 1.5356929302215576, "learning_rate": 9.895839272227529e-06, "loss": 0.3665, "step": 151 }, { "epoch": 0.6548196015078083, "grad_norm": 1.6182607412338257, "learning_rate": 9.894459981780711e-06, "loss": 0.3448, "step": 152 }, { "epoch": 0.6591276252019386, "grad_norm": 1.9151091575622559, "learning_rate": 9.893071716457183e-06, "loss": 0.4067, "step": 153 }, { "epoch": 0.6634356488960689, "grad_norm": 1.4378305673599243, "learning_rate": 9.891674478802585e-06, "loss": 0.3243, "step": 154 }, { "epoch": 0.6677436725901993, "grad_norm": 1.545253038406372, "learning_rate": 9.890268271379e-06, "loss": 0.3744, "step": 155 }, { "epoch": 0.6720516962843296, "grad_norm": 1.5699349641799927, "learning_rate": 9.888853096764963e-06, "loss": 0.3479, "step": 156 }, { "epoch": 0.6763597199784599, "grad_norm": 1.8264131546020508, "learning_rate": 9.887428957555457e-06, "loss": 0.3932, "step": 157 }, { "epoch": 0.6806677436725902, "grad_norm": 1.528017520904541, "learning_rate": 9.885995856361895e-06, "loss": 0.3508, "step": 158 }, { "epoch": 0.6849757673667205, "grad_norm": 1.4379171133041382, "learning_rate": 9.884553795812128e-06, "loss": 0.3343, "step": 159 }, { "epoch": 0.6892837910608508, "grad_norm": 1.3340097665786743, "learning_rate": 9.883102778550434e-06, "loss": 0.3292, "step": 160 }, { "epoch": 0.6935918147549811, "grad_norm": 1.4595935344696045, "learning_rate": 9.881642807237515e-06, "loss": 0.3615, "step": 161 }, { "epoch": 0.6978998384491115, "grad_norm": 1.3338977098464966, "learning_rate": 9.880173884550495e-06, "loss": 0.3546, "step": 162 }, { "epoch": 0.7022078621432418, "grad_norm": 1.4592150449752808, "learning_rate": 9.878696013182906e-06, "loss": 0.3754, "step": 163 }, { "epoch": 0.7065158858373721, "grad_norm": 1.340230941772461, "learning_rate": 9.877209195844692e-06, "loss": 0.3205, "step": 164 }, { "epoch": 0.7108239095315024, "grad_norm": 1.5795950889587402, "learning_rate": 9.875713435262205e-06, "loss": 0.3636, "step": 165 }, { "epoch": 0.7151319332256327, "grad_norm": 1.2886184453964233, "learning_rate": 9.874208734178187e-06, "loss": 0.3273, "step": 166 }, { "epoch": 0.719439956919763, "grad_norm": 1.4012725353240967, "learning_rate": 9.872695095351784e-06, "loss": 0.3594, "step": 167 }, { "epoch": 0.7237479806138933, "grad_norm": 1.4970510005950928, "learning_rate": 9.871172521558523e-06, "loss": 0.3763, "step": 168 }, { "epoch": 0.7280560043080236, "grad_norm": 1.3455449342727661, "learning_rate": 9.869641015590319e-06, "loss": 0.3359, "step": 169 }, { "epoch": 0.732364028002154, "grad_norm": 1.500728964805603, "learning_rate": 9.868100580255466e-06, "loss": 0.3386, "step": 170 }, { "epoch": 0.7366720516962844, "grad_norm": 1.409002423286438, "learning_rate": 9.86655121837863e-06, "loss": 0.3548, "step": 171 }, { "epoch": 0.7409800753904147, "grad_norm": 1.3969871997833252, "learning_rate": 9.864992932800845e-06, "loss": 0.3649, "step": 172 }, { "epoch": 0.745288099084545, "grad_norm": 1.371010184288025, "learning_rate": 9.863425726379512e-06, "loss": 0.3197, "step": 173 }, { "epoch": 0.7495961227786753, "grad_norm": 1.590254545211792, "learning_rate": 9.861849601988384e-06, "loss": 0.4077, "step": 174 }, { "epoch": 0.7539041464728056, "grad_norm": 1.4632492065429688, "learning_rate": 9.860264562517571e-06, "loss": 0.369, "step": 175 }, { "epoch": 0.7582121701669359, "grad_norm": 1.320504069328308, "learning_rate": 9.858670610873528e-06, "loss": 0.3583, "step": 176 }, { "epoch": 0.7625201938610663, "grad_norm": 1.3206356763839722, "learning_rate": 9.857067749979057e-06, "loss": 0.334, "step": 177 }, { "epoch": 0.7668282175551966, "grad_norm": 1.314485788345337, "learning_rate": 9.855455982773288e-06, "loss": 0.3694, "step": 178 }, { "epoch": 0.7711362412493269, "grad_norm": 1.317907452583313, "learning_rate": 9.853835312211692e-06, "loss": 0.3561, "step": 179 }, { "epoch": 0.7754442649434572, "grad_norm": 1.5058598518371582, "learning_rate": 9.852205741266058e-06, "loss": 0.3605, "step": 180 }, { "epoch": 0.7797522886375875, "grad_norm": 1.6332582235336304, "learning_rate": 9.8505672729245e-06, "loss": 0.377, "step": 181 }, { "epoch": 0.7840603123317178, "grad_norm": 1.2938036918640137, "learning_rate": 9.848919910191446e-06, "loss": 0.3663, "step": 182 }, { "epoch": 0.7883683360258481, "grad_norm": 1.4130111932754517, "learning_rate": 9.847263656087633e-06, "loss": 0.3211, "step": 183 }, { "epoch": 0.7926763597199784, "grad_norm": 1.5852794647216797, "learning_rate": 9.845598513650104e-06, "loss": 0.3689, "step": 184 }, { "epoch": 0.7969843834141088, "grad_norm": 1.4443858861923218, "learning_rate": 9.843924485932195e-06, "loss": 0.3746, "step": 185 }, { "epoch": 0.8012924071082391, "grad_norm": 1.2581782341003418, "learning_rate": 9.84224157600354e-06, "loss": 0.3314, "step": 186 }, { "epoch": 0.8056004308023694, "grad_norm": 1.3254746198654175, "learning_rate": 9.840549786950058e-06, "loss": 0.3416, "step": 187 }, { "epoch": 0.8099084544964997, "grad_norm": 1.3402105569839478, "learning_rate": 9.83884912187395e-06, "loss": 0.3282, "step": 188 }, { "epoch": 0.81421647819063, "grad_norm": 1.3537483215332031, "learning_rate": 9.837139583893693e-06, "loss": 0.3221, "step": 189 }, { "epoch": 0.8185245018847603, "grad_norm": 1.259764313697815, "learning_rate": 9.835421176144035e-06, "loss": 0.3428, "step": 190 }, { "epoch": 0.8228325255788906, "grad_norm": 1.2526212930679321, "learning_rate": 9.833693901775985e-06, "loss": 0.3511, "step": 191 }, { "epoch": 0.827140549273021, "grad_norm": 1.4188737869262695, "learning_rate": 9.831957763956814e-06, "loss": 0.3532, "step": 192 }, { "epoch": 0.8314485729671514, "grad_norm": 1.4591296911239624, "learning_rate": 9.830212765870043e-06, "loss": 0.3544, "step": 193 }, { "epoch": 0.8357565966612817, "grad_norm": 1.7813855409622192, "learning_rate": 9.828458910715442e-06, "loss": 0.4027, "step": 194 }, { "epoch": 0.840064620355412, "grad_norm": 1.5284059047698975, "learning_rate": 9.826696201709022e-06, "loss": 0.3897, "step": 195 }, { "epoch": 0.8443726440495423, "grad_norm": 1.4871803522109985, "learning_rate": 9.824924642083026e-06, "loss": 0.3871, "step": 196 }, { "epoch": 0.8486806677436726, "grad_norm": 1.270748257637024, "learning_rate": 9.823144235085934e-06, "loss": 0.3148, "step": 197 }, { "epoch": 0.8529886914378029, "grad_norm": 1.4008361101150513, "learning_rate": 9.821354983982438e-06, "loss": 0.3529, "step": 198 }, { "epoch": 0.8572967151319332, "grad_norm": 1.178896188735962, "learning_rate": 9.819556892053456e-06, "loss": 0.2824, "step": 199 }, { "epoch": 0.8616047388260636, "grad_norm": 1.3308268785476685, "learning_rate": 9.817749962596115e-06, "loss": 0.3315, "step": 200 }, { "epoch": 0.8659127625201939, "grad_norm": 1.3277922868728638, "learning_rate": 9.815934198923746e-06, "loss": 0.3429, "step": 201 }, { "epoch": 0.8702207862143242, "grad_norm": 1.2668861150741577, "learning_rate": 9.814109604365878e-06, "loss": 0.3422, "step": 202 }, { "epoch": 0.8745288099084545, "grad_norm": 1.2362356185913086, "learning_rate": 9.812276182268236e-06, "loss": 0.3175, "step": 203 }, { "epoch": 0.8788368336025848, "grad_norm": 1.5288373231887817, "learning_rate": 9.810433935992734e-06, "loss": 0.3703, "step": 204 }, { "epoch": 0.8831448572967151, "grad_norm": 1.5643410682678223, "learning_rate": 9.808582868917458e-06, "loss": 0.353, "step": 205 }, { "epoch": 0.8874528809908454, "grad_norm": 1.3788419961929321, "learning_rate": 9.806722984436676e-06, "loss": 0.3355, "step": 206 }, { "epoch": 0.8917609046849758, "grad_norm": 1.4801995754241943, "learning_rate": 9.804854285960823e-06, "loss": 0.3342, "step": 207 }, { "epoch": 0.8960689283791061, "grad_norm": 1.2330700159072876, "learning_rate": 9.802976776916493e-06, "loss": 0.3127, "step": 208 }, { "epoch": 0.9003769520732364, "grad_norm": 1.384192943572998, "learning_rate": 9.801090460746442e-06, "loss": 0.3564, "step": 209 }, { "epoch": 0.9046849757673667, "grad_norm": 1.2636990547180176, "learning_rate": 9.799195340909569e-06, "loss": 0.3284, "step": 210 }, { "epoch": 0.908992999461497, "grad_norm": 1.486065149307251, "learning_rate": 9.79729142088092e-06, "loss": 0.3645, "step": 211 }, { "epoch": 0.9133010231556273, "grad_norm": 1.233577847480774, "learning_rate": 9.795378704151675e-06, "loss": 0.3468, "step": 212 }, { "epoch": 0.9176090468497576, "grad_norm": 1.2524428367614746, "learning_rate": 9.793457194229145e-06, "loss": 0.3252, "step": 213 }, { "epoch": 0.9219170705438879, "grad_norm": 1.1312000751495361, "learning_rate": 9.791526894636767e-06, "loss": 0.3255, "step": 214 }, { "epoch": 0.9262250942380184, "grad_norm": 1.2899000644683838, "learning_rate": 9.789587808914094e-06, "loss": 0.3388, "step": 215 }, { "epoch": 0.9305331179321487, "grad_norm": 1.3396518230438232, "learning_rate": 9.787639940616789e-06, "loss": 0.3757, "step": 216 }, { "epoch": 0.934841141626279, "grad_norm": 1.3656538724899292, "learning_rate": 9.785683293316622e-06, "loss": 0.3497, "step": 217 }, { "epoch": 0.9391491653204093, "grad_norm": 1.1662282943725586, "learning_rate": 9.783717870601458e-06, "loss": 0.307, "step": 218 }, { "epoch": 0.9434571890145396, "grad_norm": 1.3306939601898193, "learning_rate": 9.781743676075257e-06, "loss": 0.3609, "step": 219 }, { "epoch": 0.9477652127086699, "grad_norm": 1.3107666969299316, "learning_rate": 9.77976071335806e-06, "loss": 0.319, "step": 220 }, { "epoch": 0.9520732364028002, "grad_norm": 1.2948092222213745, "learning_rate": 9.777768986085985e-06, "loss": 0.3615, "step": 221 }, { "epoch": 0.9563812600969306, "grad_norm": 1.3614375591278076, "learning_rate": 9.775768497911226e-06, "loss": 0.3563, "step": 222 }, { "epoch": 0.9606892837910609, "grad_norm": 1.3317749500274658, "learning_rate": 9.77375925250204e-06, "loss": 0.3564, "step": 223 }, { "epoch": 0.9649973074851912, "grad_norm": 1.2995858192443848, "learning_rate": 9.771741253542742e-06, "loss": 0.3432, "step": 224 }, { "epoch": 0.9693053311793215, "grad_norm": 1.512982964515686, "learning_rate": 9.769714504733695e-06, "loss": 0.3616, "step": 225 }, { "epoch": 0.9736133548734518, "grad_norm": 1.3074134588241577, "learning_rate": 9.767679009791312e-06, "loss": 0.3527, "step": 226 }, { "epoch": 0.9779213785675821, "grad_norm": 1.2712088823318481, "learning_rate": 9.765634772448039e-06, "loss": 0.3368, "step": 227 }, { "epoch": 0.9822294022617124, "grad_norm": 1.464223861694336, "learning_rate": 9.763581796452353e-06, "loss": 0.311, "step": 228 }, { "epoch": 0.9865374259558427, "grad_norm": 1.179933786392212, "learning_rate": 9.76152008556876e-06, "loss": 0.3592, "step": 229 }, { "epoch": 0.9908454496499731, "grad_norm": 1.3902326822280884, "learning_rate": 9.759449643577779e-06, "loss": 0.3597, "step": 230 }, { "epoch": 0.9951534733441034, "grad_norm": 1.6388298273086548, "learning_rate": 9.757370474275938e-06, "loss": 0.3764, "step": 231 }, { "epoch": 0.9994614970382337, "grad_norm": 1.1966930627822876, "learning_rate": 9.755282581475769e-06, "loss": 0.3475, "step": 232 }, { "epoch": 1.0, "grad_norm": 1.1966930627822876, "learning_rate": 9.753185969005802e-06, "loss": 0.4553, "step": 233 }, { "epoch": 1.0043080236941304, "grad_norm": 5.072446823120117, "learning_rate": 9.751080640710554e-06, "loss": 0.3065, "step": 234 }, { "epoch": 1.0086160473882606, "grad_norm": 1.2046287059783936, "learning_rate": 9.748966600450526e-06, "loss": 0.3106, "step": 235 }, { "epoch": 1.012924071082391, "grad_norm": 1.285788655281067, "learning_rate": 9.746843852102191e-06, "loss": 0.305, "step": 236 }, { "epoch": 1.0172320947765212, "grad_norm": 1.2670906782150269, "learning_rate": 9.744712399557992e-06, "loss": 0.2838, "step": 237 }, { "epoch": 1.0215401184706516, "grad_norm": 1.3093379735946655, "learning_rate": 9.742572246726336e-06, "loss": 0.3232, "step": 238 }, { "epoch": 1.0258481421647818, "grad_norm": 1.0496468544006348, "learning_rate": 9.740423397531573e-06, "loss": 0.2989, "step": 239 }, { "epoch": 1.0301561658589122, "grad_norm": 1.095651388168335, "learning_rate": 9.738265855914014e-06, "loss": 0.3003, "step": 240 }, { "epoch": 1.0344641895530426, "grad_norm": 1.1687029600143433, "learning_rate": 9.736099625829894e-06, "loss": 0.3208, "step": 241 }, { "epoch": 1.0387722132471728, "grad_norm": 1.6566765308380127, "learning_rate": 9.733924711251393e-06, "loss": 0.3106, "step": 242 }, { "epoch": 1.0430802369413033, "grad_norm": 1.4189642667770386, "learning_rate": 9.731741116166607e-06, "loss": 0.3062, "step": 243 }, { "epoch": 1.0473882606354334, "grad_norm": 1.2597739696502686, "learning_rate": 9.729548844579552e-06, "loss": 0.3045, "step": 244 }, { "epoch": 1.0516962843295639, "grad_norm": 1.799160122871399, "learning_rate": 9.727347900510155e-06, "loss": 0.3076, "step": 245 }, { "epoch": 1.056004308023694, "grad_norm": 3.2403290271759033, "learning_rate": 9.725138287994246e-06, "loss": 0.2985, "step": 246 }, { "epoch": 1.0603123317178245, "grad_norm": 1.3010872602462769, "learning_rate": 9.722920011083546e-06, "loss": 0.3062, "step": 247 }, { "epoch": 1.0646203554119547, "grad_norm": 1.3619431257247925, "learning_rate": 9.720693073845668e-06, "loss": 0.3137, "step": 248 }, { "epoch": 1.068928379106085, "grad_norm": 1.2294771671295166, "learning_rate": 9.718457480364103e-06, "loss": 0.3264, "step": 249 }, { "epoch": 1.0732364028002155, "grad_norm": 1.2018005847930908, "learning_rate": 9.716213234738216e-06, "loss": 0.3027, "step": 250 }, { "epoch": 1.0775444264943457, "grad_norm": 1.1431026458740234, "learning_rate": 9.713960341083237e-06, "loss": 0.2792, "step": 251 }, { "epoch": 1.081852450188476, "grad_norm": 1.3094645738601685, "learning_rate": 9.711698803530253e-06, "loss": 0.3126, "step": 252 }, { "epoch": 1.0861604738826063, "grad_norm": 1.0354987382888794, "learning_rate": 9.709428626226204e-06, "loss": 0.2935, "step": 253 }, { "epoch": 1.0904684975767367, "grad_norm": 1.1405781507492065, "learning_rate": 9.707149813333866e-06, "loss": 0.2886, "step": 254 }, { "epoch": 1.0947765212708669, "grad_norm": 1.3035577535629272, "learning_rate": 9.704862369031857e-06, "loss": 0.3231, "step": 255 }, { "epoch": 1.0990845449649973, "grad_norm": 1.3560417890548706, "learning_rate": 9.70256629751462e-06, "loss": 0.3304, "step": 256 }, { "epoch": 1.1033925686591277, "grad_norm": 1.3723593950271606, "learning_rate": 9.700261602992417e-06, "loss": 0.3418, "step": 257 }, { "epoch": 1.107700592353258, "grad_norm": 1.1065815687179565, "learning_rate": 9.69794828969132e-06, "loss": 0.287, "step": 258 }, { "epoch": 1.1120086160473883, "grad_norm": 1.5255553722381592, "learning_rate": 9.695626361853207e-06, "loss": 0.3027, "step": 259 }, { "epoch": 1.1163166397415185, "grad_norm": 1.1859606504440308, "learning_rate": 9.693295823735754e-06, "loss": 0.3003, "step": 260 }, { "epoch": 1.120624663435649, "grad_norm": 1.1855946779251099, "learning_rate": 9.690956679612422e-06, "loss": 0.2973, "step": 261 }, { "epoch": 1.1249326871297791, "grad_norm": 1.1954553127288818, "learning_rate": 9.688608933772454e-06, "loss": 0.2837, "step": 262 }, { "epoch": 1.1292407108239095, "grad_norm": 1.1184488534927368, "learning_rate": 9.686252590520869e-06, "loss": 0.2915, "step": 263 }, { "epoch": 1.13354873451804, "grad_norm": 1.190000057220459, "learning_rate": 9.683887654178446e-06, "loss": 0.3054, "step": 264 }, { "epoch": 1.1378567582121701, "grad_norm": 1.272902011871338, "learning_rate": 9.681514129081725e-06, "loss": 0.3111, "step": 265 }, { "epoch": 1.1421647819063006, "grad_norm": 1.0627204179763794, "learning_rate": 9.679132019582988e-06, "loss": 0.2828, "step": 266 }, { "epoch": 1.1464728056004307, "grad_norm": 1.1977815628051758, "learning_rate": 9.67674133005027e-06, "loss": 0.2952, "step": 267 }, { "epoch": 1.1507808292945612, "grad_norm": 1.3312697410583496, "learning_rate": 9.674342064867326e-06, "loss": 0.3163, "step": 268 }, { "epoch": 1.1550888529886914, "grad_norm": 1.3952230215072632, "learning_rate": 9.671934228433647e-06, "loss": 0.3015, "step": 269 }, { "epoch": 1.1593968766828218, "grad_norm": 1.2030738592147827, "learning_rate": 9.669517825164435e-06, "loss": 0.2918, "step": 270 }, { "epoch": 1.1637049003769522, "grad_norm": 1.1573476791381836, "learning_rate": 9.667092859490599e-06, "loss": 0.2963, "step": 271 }, { "epoch": 1.1680129240710824, "grad_norm": 1.2124096155166626, "learning_rate": 9.664659335858755e-06, "loss": 0.2825, "step": 272 }, { "epoch": 1.1723209477652128, "grad_norm": 1.1482504606246948, "learning_rate": 9.662217258731208e-06, "loss": 0.2895, "step": 273 }, { "epoch": 1.176628971459343, "grad_norm": 1.2455042600631714, "learning_rate": 9.659766632585946e-06, "loss": 0.3021, "step": 274 }, { "epoch": 1.1809369951534734, "grad_norm": 1.208392858505249, "learning_rate": 9.657307461916637e-06, "loss": 0.298, "step": 275 }, { "epoch": 1.1852450188476036, "grad_norm": 1.2410141229629517, "learning_rate": 9.654839751232612e-06, "loss": 0.3029, "step": 276 }, { "epoch": 1.189553042541734, "grad_norm": 1.4115912914276123, "learning_rate": 9.652363505058866e-06, "loss": 0.3306, "step": 277 }, { "epoch": 1.1938610662358644, "grad_norm": 1.3072352409362793, "learning_rate": 9.649878727936044e-06, "loss": 0.3226, "step": 278 }, { "epoch": 1.1981690899299946, "grad_norm": 1.3294739723205566, "learning_rate": 9.647385424420435e-06, "loss": 0.2821, "step": 279 }, { "epoch": 1.202477113624125, "grad_norm": 1.4789550304412842, "learning_rate": 9.644883599083959e-06, "loss": 0.322, "step": 280 }, { "epoch": 1.2067851373182552, "grad_norm": 1.3207513093948364, "learning_rate": 9.642373256514164e-06, "loss": 0.3128, "step": 281 }, { "epoch": 1.2110931610123856, "grad_norm": 1.1648540496826172, "learning_rate": 9.639854401314219e-06, "loss": 0.3031, "step": 282 }, { "epoch": 1.2154011847065158, "grad_norm": 1.3340126276016235, "learning_rate": 9.637327038102902e-06, "loss": 0.3149, "step": 283 }, { "epoch": 1.2197092084006462, "grad_norm": 1.2494897842407227, "learning_rate": 9.634791171514585e-06, "loss": 0.3246, "step": 284 }, { "epoch": 1.2240172320947766, "grad_norm": 1.4263726472854614, "learning_rate": 9.632246806199242e-06, "loss": 0.3244, "step": 285 }, { "epoch": 1.2283252557889068, "grad_norm": 1.4411340951919556, "learning_rate": 9.629693946822423e-06, "loss": 0.3404, "step": 286 }, { "epoch": 1.2326332794830372, "grad_norm": 1.2895034551620483, "learning_rate": 9.627132598065258e-06, "loss": 0.3024, "step": 287 }, { "epoch": 1.2369413031771674, "grad_norm": 1.2556586265563965, "learning_rate": 9.624562764624445e-06, "loss": 0.3192, "step": 288 }, { "epoch": 1.2412493268712979, "grad_norm": 1.2426923513412476, "learning_rate": 9.621984451212237e-06, "loss": 0.3025, "step": 289 }, { "epoch": 1.245557350565428, "grad_norm": 1.331985592842102, "learning_rate": 9.619397662556434e-06, "loss": 0.3184, "step": 290 }, { "epoch": 1.2498653742595585, "grad_norm": 1.3090969324111938, "learning_rate": 9.616802403400384e-06, "loss": 0.3122, "step": 291 }, { "epoch": 1.2541733979536889, "grad_norm": 1.2021125555038452, "learning_rate": 9.614198678502965e-06, "loss": 0.3116, "step": 292 }, { "epoch": 1.258481421647819, "grad_norm": 1.3388304710388184, "learning_rate": 9.611586492638573e-06, "loss": 0.3115, "step": 293 }, { "epoch": 1.2627894453419493, "grad_norm": 1.0857573747634888, "learning_rate": 9.608965850597125e-06, "loss": 0.3155, "step": 294 }, { "epoch": 1.2670974690360797, "grad_norm": 1.1337326765060425, "learning_rate": 9.606336757184041e-06, "loss": 0.3091, "step": 295 }, { "epoch": 1.27140549273021, "grad_norm": 1.2282270193099976, "learning_rate": 9.603699217220239e-06, "loss": 0.3015, "step": 296 }, { "epoch": 1.2757135164243403, "grad_norm": 1.386949062347412, "learning_rate": 9.601053235542124e-06, "loss": 0.31, "step": 297 }, { "epoch": 1.2800215401184707, "grad_norm": 1.1345770359039307, "learning_rate": 9.598398817001585e-06, "loss": 0.2828, "step": 298 }, { "epoch": 1.284329563812601, "grad_norm": 1.24062979221344, "learning_rate": 9.595735966465973e-06, "loss": 0.3233, "step": 299 }, { "epoch": 1.2886375875067313, "grad_norm": 1.2808305025100708, "learning_rate": 9.59306468881811e-06, "loss": 0.3158, "step": 300 }, { "epoch": 1.2929456112008615, "grad_norm": 1.2213916778564453, "learning_rate": 9.590384988956264e-06, "loss": 0.2939, "step": 301 }, { "epoch": 1.297253634894992, "grad_norm": 1.2903624773025513, "learning_rate": 9.587696871794148e-06, "loss": 0.3257, "step": 302 }, { "epoch": 1.3015616585891223, "grad_norm": 1.2106595039367676, "learning_rate": 9.585000342260914e-06, "loss": 0.3037, "step": 303 }, { "epoch": 1.3058696822832525, "grad_norm": 1.446324348449707, "learning_rate": 9.582295405301131e-06, "loss": 0.3514, "step": 304 }, { "epoch": 1.310177705977383, "grad_norm": 1.2530688047409058, "learning_rate": 9.579582065874794e-06, "loss": 0.3084, "step": 305 }, { "epoch": 1.3144857296715131, "grad_norm": 1.36234450340271, "learning_rate": 9.576860328957299e-06, "loss": 0.3205, "step": 306 }, { "epoch": 1.3187937533656435, "grad_norm": 1.2496092319488525, "learning_rate": 9.574130199539443e-06, "loss": 0.276, "step": 307 }, { "epoch": 1.3231017770597737, "grad_norm": 1.0486127138137817, "learning_rate": 9.571391682627413e-06, "loss": 0.2733, "step": 308 }, { "epoch": 1.3274098007539041, "grad_norm": 1.1949307918548584, "learning_rate": 9.568644783242771e-06, "loss": 0.3207, "step": 309 }, { "epoch": 1.3317178244480345, "grad_norm": 1.0762794017791748, "learning_rate": 9.565889506422457e-06, "loss": 0.2798, "step": 310 }, { "epoch": 1.3360258481421647, "grad_norm": 1.0935207605361938, "learning_rate": 9.563125857218766e-06, "loss": 0.3115, "step": 311 }, { "epoch": 1.3403338718362952, "grad_norm": 1.2250378131866455, "learning_rate": 9.56035384069935e-06, "loss": 0.3328, "step": 312 }, { "epoch": 1.3446418955304253, "grad_norm": 1.2242401838302612, "learning_rate": 9.557573461947201e-06, "loss": 0.2823, "step": 313 }, { "epoch": 1.3489499192245558, "grad_norm": 1.2449086904525757, "learning_rate": 9.554784726060647e-06, "loss": 0.3221, "step": 314 }, { "epoch": 1.353257942918686, "grad_norm": 1.1806416511535645, "learning_rate": 9.551987638153339e-06, "loss": 0.2982, "step": 315 }, { "epoch": 1.3575659666128164, "grad_norm": 1.3118568658828735, "learning_rate": 9.549182203354241e-06, "loss": 0.3232, "step": 316 }, { "epoch": 1.3618739903069468, "grad_norm": 1.2797043323516846, "learning_rate": 9.546368426807628e-06, "loss": 0.3257, "step": 317 }, { "epoch": 1.366182014001077, "grad_norm": 1.2776271104812622, "learning_rate": 9.543546313673065e-06, "loss": 0.3122, "step": 318 }, { "epoch": 1.3704900376952074, "grad_norm": 1.2693960666656494, "learning_rate": 9.540715869125407e-06, "loss": 0.3084, "step": 319 }, { "epoch": 1.3747980613893376, "grad_norm": 1.1827868223190308, "learning_rate": 9.537877098354787e-06, "loss": 0.314, "step": 320 }, { "epoch": 1.379106085083468, "grad_norm": 1.1357877254486084, "learning_rate": 9.5350300065666e-06, "loss": 0.3171, "step": 321 }, { "epoch": 1.3834141087775982, "grad_norm": 1.1047065258026123, "learning_rate": 9.532174598981507e-06, "loss": 0.3039, "step": 322 }, { "epoch": 1.3877221324717286, "grad_norm": 1.131152868270874, "learning_rate": 9.529310880835414e-06, "loss": 0.2906, "step": 323 }, { "epoch": 1.392030156165859, "grad_norm": 1.1229904890060425, "learning_rate": 9.526438857379463e-06, "loss": 0.2913, "step": 324 }, { "epoch": 1.3963381798599892, "grad_norm": 1.3774073123931885, "learning_rate": 9.52355853388003e-06, "loss": 0.3216, "step": 325 }, { "epoch": 1.4006462035541196, "grad_norm": 1.0984300374984741, "learning_rate": 9.520669915618708e-06, "loss": 0.2925, "step": 326 }, { "epoch": 1.4049542272482498, "grad_norm": 1.2843397855758667, "learning_rate": 9.5177730078923e-06, "loss": 0.29, "step": 327 }, { "epoch": 1.4092622509423802, "grad_norm": 1.1826907396316528, "learning_rate": 9.514867816012809e-06, "loss": 0.3018, "step": 328 }, { "epoch": 1.4135702746365104, "grad_norm": 1.4079996347427368, "learning_rate": 9.511954345307432e-06, "loss": 0.3295, "step": 329 }, { "epoch": 1.4178782983306408, "grad_norm": 1.1314058303833008, "learning_rate": 9.509032601118541e-06, "loss": 0.3052, "step": 330 }, { "epoch": 1.4221863220247712, "grad_norm": 1.1717647314071655, "learning_rate": 9.506102588803683e-06, "loss": 0.3303, "step": 331 }, { "epoch": 1.4264943457189014, "grad_norm": 1.0674874782562256, "learning_rate": 9.503164313735566e-06, "loss": 0.2946, "step": 332 }, { "epoch": 1.4308023694130318, "grad_norm": 1.188008189201355, "learning_rate": 9.500217781302048e-06, "loss": 0.278, "step": 333 }, { "epoch": 1.435110393107162, "grad_norm": 1.123025894165039, "learning_rate": 9.497262996906126e-06, "loss": 0.3148, "step": 334 }, { "epoch": 1.4394184168012925, "grad_norm": 1.3759355545043945, "learning_rate": 9.494299965965935e-06, "loss": 0.3285, "step": 335 }, { "epoch": 1.4437264404954226, "grad_norm": 1.3646365404129028, "learning_rate": 9.491328693914723e-06, "loss": 0.3125, "step": 336 }, { "epoch": 1.448034464189553, "grad_norm": 1.066312313079834, "learning_rate": 9.488349186200858e-06, "loss": 0.2857, "step": 337 }, { "epoch": 1.4523424878836835, "grad_norm": 1.2194416522979736, "learning_rate": 9.485361448287804e-06, "loss": 0.2827, "step": 338 }, { "epoch": 1.4566505115778137, "grad_norm": 1.3936829566955566, "learning_rate": 9.482365485654118e-06, "loss": 0.3488, "step": 339 }, { "epoch": 1.4609585352719439, "grad_norm": 1.1621965169906616, "learning_rate": 9.479361303793441e-06, "loss": 0.286, "step": 340 }, { "epoch": 1.4652665589660743, "grad_norm": 1.1074098348617554, "learning_rate": 9.476348908214482e-06, "loss": 0.3226, "step": 341 }, { "epoch": 1.4695745826602047, "grad_norm": 1.192676305770874, "learning_rate": 9.47332830444101e-06, "loss": 0.2962, "step": 342 }, { "epoch": 1.4738826063543349, "grad_norm": 1.277662992477417, "learning_rate": 9.470299498011851e-06, "loss": 0.3214, "step": 343 }, { "epoch": 1.4781906300484653, "grad_norm": 1.1942297220230103, "learning_rate": 9.46726249448087e-06, "loss": 0.3061, "step": 344 }, { "epoch": 1.4824986537425957, "grad_norm": 1.2536174058914185, "learning_rate": 9.464217299416956e-06, "loss": 0.3018, "step": 345 }, { "epoch": 1.486806677436726, "grad_norm": 1.1480363607406616, "learning_rate": 9.46116391840403e-06, "loss": 0.2864, "step": 346 }, { "epoch": 1.491114701130856, "grad_norm": 1.133062481880188, "learning_rate": 9.458102357041017e-06, "loss": 0.3061, "step": 347 }, { "epoch": 1.4954227248249865, "grad_norm": 1.2819349765777588, "learning_rate": 9.45503262094184e-06, "loss": 0.3414, "step": 348 }, { "epoch": 1.499730748519117, "grad_norm": 1.1016157865524292, "learning_rate": 9.451954715735416e-06, "loss": 0.2971, "step": 349 }, { "epoch": 1.504038772213247, "grad_norm": 1.2356181144714355, "learning_rate": 9.448868647065644e-06, "loss": 0.3496, "step": 350 }, { "epoch": 1.5083467959073775, "grad_norm": 1.4491801261901855, "learning_rate": 9.445774420591382e-06, "loss": 0.3241, "step": 351 }, { "epoch": 1.512654819601508, "grad_norm": 1.2295855283737183, "learning_rate": 9.442672041986456e-06, "loss": 0.2975, "step": 352 }, { "epoch": 1.5169628432956381, "grad_norm": 1.4450600147247314, "learning_rate": 9.43956151693964e-06, "loss": 0.3366, "step": 353 }, { "epoch": 1.5212708669897683, "grad_norm": 1.1162477731704712, "learning_rate": 9.436442851154642e-06, "loss": 0.3131, "step": 354 }, { "epoch": 1.5255788906838987, "grad_norm": 1.1916290521621704, "learning_rate": 9.433316050350099e-06, "loss": 0.2819, "step": 355 }, { "epoch": 1.5298869143780292, "grad_norm": 1.2616592645645142, "learning_rate": 9.430181120259566e-06, "loss": 0.3129, "step": 356 }, { "epoch": 1.5341949380721593, "grad_norm": 1.18837308883667, "learning_rate": 9.427038066631502e-06, "loss": 0.2751, "step": 357 }, { "epoch": 1.5385029617662898, "grad_norm": 1.3436187505722046, "learning_rate": 9.423886895229266e-06, "loss": 0.3213, "step": 358 }, { "epoch": 1.5428109854604202, "grad_norm": 1.1284348964691162, "learning_rate": 9.420727611831098e-06, "loss": 0.3002, "step": 359 }, { "epoch": 1.5471190091545504, "grad_norm": 1.1525483131408691, "learning_rate": 9.417560222230115e-06, "loss": 0.2808, "step": 360 }, { "epoch": 1.5514270328486806, "grad_norm": 1.280132532119751, "learning_rate": 9.414384732234301e-06, "loss": 0.3314, "step": 361 }, { "epoch": 1.555735056542811, "grad_norm": 1.1586127281188965, "learning_rate": 9.411201147666486e-06, "loss": 0.3075, "step": 362 }, { "epoch": 1.5600430802369414, "grad_norm": 1.055859923362732, "learning_rate": 9.408009474364353e-06, "loss": 0.2944, "step": 363 }, { "epoch": 1.5643511039310716, "grad_norm": 1.2206306457519531, "learning_rate": 9.404809718180408e-06, "loss": 0.31, "step": 364 }, { "epoch": 1.568659127625202, "grad_norm": 1.0499204397201538, "learning_rate": 9.401601884981983e-06, "loss": 0.3095, "step": 365 }, { "epoch": 1.5729671513193324, "grad_norm": 1.1521718502044678, "learning_rate": 9.39838598065122e-06, "loss": 0.2958, "step": 366 }, { "epoch": 1.5772751750134626, "grad_norm": 1.2289202213287354, "learning_rate": 9.39516201108506e-06, "loss": 0.3336, "step": 367 }, { "epoch": 1.5815831987075928, "grad_norm": 1.2974143028259277, "learning_rate": 9.391929982195233e-06, "loss": 0.3301, "step": 368 }, { "epoch": 1.5858912224017232, "grad_norm": 1.2205924987792969, "learning_rate": 9.38868989990825e-06, "loss": 0.3009, "step": 369 }, { "epoch": 1.5901992460958536, "grad_norm": 1.0817668437957764, "learning_rate": 9.385441770165385e-06, "loss": 0.3144, "step": 370 }, { "epoch": 1.5945072697899838, "grad_norm": 1.0423145294189453, "learning_rate": 9.382185598922674e-06, "loss": 0.2883, "step": 371 }, { "epoch": 1.598815293484114, "grad_norm": 1.2427335977554321, "learning_rate": 9.378921392150893e-06, "loss": 0.3424, "step": 372 }, { "epoch": 1.6031233171782446, "grad_norm": 1.1313612461090088, "learning_rate": 9.375649155835554e-06, "loss": 0.2856, "step": 373 }, { "epoch": 1.6074313408723748, "grad_norm": 1.2042919397354126, "learning_rate": 9.372368895976896e-06, "loss": 0.3176, "step": 374 }, { "epoch": 1.611739364566505, "grad_norm": 1.1364398002624512, "learning_rate": 9.369080618589866e-06, "loss": 0.3087, "step": 375 }, { "epoch": 1.6160473882606354, "grad_norm": 1.328291893005371, "learning_rate": 9.365784329704114e-06, "loss": 0.3349, "step": 376 }, { "epoch": 1.6203554119547658, "grad_norm": 1.3974817991256714, "learning_rate": 9.362480035363987e-06, "loss": 0.3198, "step": 377 }, { "epoch": 1.624663435648896, "grad_norm": 1.1749645471572876, "learning_rate": 9.3591677416285e-06, "loss": 0.2795, "step": 378 }, { "epoch": 1.6289714593430262, "grad_norm": 1.1566781997680664, "learning_rate": 9.35584745457134e-06, "loss": 0.315, "step": 379 }, { "epoch": 1.6332794830371569, "grad_norm": 1.1803126335144043, "learning_rate": 9.352519180280862e-06, "loss": 0.2936, "step": 380 }, { "epoch": 1.637587506731287, "grad_norm": 1.0722548961639404, "learning_rate": 9.34918292486005e-06, "loss": 0.322, "step": 381 }, { "epoch": 1.6418955304254172, "grad_norm": 1.2519919872283936, "learning_rate": 9.345838694426535e-06, "loss": 0.341, "step": 382 }, { "epoch": 1.6462035541195477, "grad_norm": 1.0761858224868774, "learning_rate": 9.342486495112566e-06, "loss": 0.2995, "step": 383 }, { "epoch": 1.650511577813678, "grad_norm": 1.28709876537323, "learning_rate": 9.339126333065008e-06, "loss": 0.3171, "step": 384 }, { "epoch": 1.6548196015078083, "grad_norm": 1.0367558002471924, "learning_rate": 9.335758214445323e-06, "loss": 0.305, "step": 385 }, { "epoch": 1.6591276252019385, "grad_norm": 1.0369017124176025, "learning_rate": 9.332382145429568e-06, "loss": 0.3079, "step": 386 }, { "epoch": 1.6634356488960689, "grad_norm": 1.1736524105072021, "learning_rate": 9.328998132208373e-06, "loss": 0.3027, "step": 387 }, { "epoch": 1.6677436725901993, "grad_norm": 1.1699035167694092, "learning_rate": 9.325606180986938e-06, "loss": 0.3022, "step": 388 }, { "epoch": 1.6720516962843295, "grad_norm": 1.1353636980056763, "learning_rate": 9.32220629798502e-06, "loss": 0.3001, "step": 389 }, { "epoch": 1.67635971997846, "grad_norm": 1.1523483991622925, "learning_rate": 9.318798489436917e-06, "loss": 0.2929, "step": 390 }, { "epoch": 1.6806677436725903, "grad_norm": 1.2557148933410645, "learning_rate": 9.315382761591463e-06, "loss": 0.2971, "step": 391 }, { "epoch": 1.6849757673667205, "grad_norm": 1.162603735923767, "learning_rate": 9.311959120712012e-06, "loss": 0.2927, "step": 392 }, { "epoch": 1.6892837910608507, "grad_norm": 1.1214804649353027, "learning_rate": 9.308527573076425e-06, "loss": 0.3175, "step": 393 }, { "epoch": 1.693591814754981, "grad_norm": 1.1605451107025146, "learning_rate": 9.30508812497707e-06, "loss": 0.3045, "step": 394 }, { "epoch": 1.6978998384491115, "grad_norm": 1.1691983938217163, "learning_rate": 9.301640782720792e-06, "loss": 0.306, "step": 395 }, { "epoch": 1.7022078621432417, "grad_norm": 1.150133490562439, "learning_rate": 9.298185552628917e-06, "loss": 0.3005, "step": 396 }, { "epoch": 1.7065158858373721, "grad_norm": 1.1442017555236816, "learning_rate": 9.294722441037238e-06, "loss": 0.304, "step": 397 }, { "epoch": 1.7108239095315025, "grad_norm": 1.1127568483352661, "learning_rate": 9.291251454295989e-06, "loss": 0.2824, "step": 398 }, { "epoch": 1.7151319332256327, "grad_norm": 1.1057833433151245, "learning_rate": 9.287772598769855e-06, "loss": 0.2963, "step": 399 }, { "epoch": 1.719439956919763, "grad_norm": 1.0729984045028687, "learning_rate": 9.284285880837947e-06, "loss": 0.2854, "step": 400 }, { "epoch": 1.7237479806138933, "grad_norm": 1.2790180444717407, "learning_rate": 9.28079130689379e-06, "loss": 0.3346, "step": 401 }, { "epoch": 1.7280560043080238, "grad_norm": 1.1950209140777588, "learning_rate": 9.277288883345318e-06, "loss": 0.3018, "step": 402 }, { "epoch": 1.732364028002154, "grad_norm": 1.1178114414215088, "learning_rate": 9.273778616614857e-06, "loss": 0.2919, "step": 403 }, { "epoch": 1.7366720516962844, "grad_norm": 1.1616685390472412, "learning_rate": 9.270260513139116e-06, "loss": 0.2988, "step": 404 }, { "epoch": 1.7409800753904148, "grad_norm": 0.9974607825279236, "learning_rate": 9.266734579369172e-06, "loss": 0.2944, "step": 405 }, { "epoch": 1.745288099084545, "grad_norm": 1.1959946155548096, "learning_rate": 9.263200821770462e-06, "loss": 0.313, "step": 406 }, { "epoch": 1.7495961227786752, "grad_norm": 1.2752206325531006, "learning_rate": 9.25965924682277e-06, "loss": 0.3159, "step": 407 }, { "epoch": 1.7539041464728056, "grad_norm": 1.4069592952728271, "learning_rate": 9.256109861020213e-06, "loss": 0.3239, "step": 408 }, { "epoch": 1.758212170166936, "grad_norm": 1.1593458652496338, "learning_rate": 9.252552670871232e-06, "loss": 0.308, "step": 409 }, { "epoch": 1.7625201938610662, "grad_norm": 1.0342111587524414, "learning_rate": 9.248987682898576e-06, "loss": 0.306, "step": 410 }, { "epoch": 1.7668282175551966, "grad_norm": 1.130953311920166, "learning_rate": 9.245414903639295e-06, "loss": 0.2992, "step": 411 }, { "epoch": 1.771136241249327, "grad_norm": 1.1652777194976807, "learning_rate": 9.241834339644726e-06, "loss": 0.3116, "step": 412 }, { "epoch": 1.7754442649434572, "grad_norm": 1.0549882650375366, "learning_rate": 9.23824599748048e-06, "loss": 0.2886, "step": 413 }, { "epoch": 1.7797522886375874, "grad_norm": 1.2357959747314453, "learning_rate": 9.234649883726432e-06, "loss": 0.3175, "step": 414 }, { "epoch": 1.7840603123317178, "grad_norm": 1.2025140523910522, "learning_rate": 9.231046004976704e-06, "loss": 0.3122, "step": 415 }, { "epoch": 1.7883683360258482, "grad_norm": 1.1126983165740967, "learning_rate": 9.22743436783966e-06, "loss": 0.32, "step": 416 }, { "epoch": 1.7926763597199784, "grad_norm": 1.2437725067138672, "learning_rate": 9.223814978937888e-06, "loss": 0.344, "step": 417 }, { "epoch": 1.7969843834141088, "grad_norm": 1.0120338201522827, "learning_rate": 9.220187844908194e-06, "loss": 0.2985, "step": 418 }, { "epoch": 1.8012924071082392, "grad_norm": 1.1840782165527344, "learning_rate": 9.216552972401582e-06, "loss": 0.3193, "step": 419 }, { "epoch": 1.8056004308023694, "grad_norm": 1.0429607629776, "learning_rate": 9.212910368083246e-06, "loss": 0.3068, "step": 420 }, { "epoch": 1.8099084544964996, "grad_norm": 1.1996227502822876, "learning_rate": 9.209260038632562e-06, "loss": 0.3108, "step": 421 }, { "epoch": 1.81421647819063, "grad_norm": 1.2330989837646484, "learning_rate": 9.205601990743068e-06, "loss": 0.2991, "step": 422 }, { "epoch": 1.8185245018847604, "grad_norm": 1.0504173040390015, "learning_rate": 9.201936231122453e-06, "loss": 0.3006, "step": 423 }, { "epoch": 1.8228325255788906, "grad_norm": 1.109567642211914, "learning_rate": 9.198262766492554e-06, "loss": 0.3217, "step": 424 }, { "epoch": 1.827140549273021, "grad_norm": 1.0737537145614624, "learning_rate": 9.194581603589327e-06, "loss": 0.3186, "step": 425 }, { "epoch": 1.8314485729671515, "grad_norm": 1.0351943969726562, "learning_rate": 9.190892749162854e-06, "loss": 0.3022, "step": 426 }, { "epoch": 1.8357565966612817, "grad_norm": 1.0790618658065796, "learning_rate": 9.187196209977314e-06, "loss": 0.3004, "step": 427 }, { "epoch": 1.8400646203554119, "grad_norm": 1.111277461051941, "learning_rate": 9.18349199281098e-06, "loss": 0.3033, "step": 428 }, { "epoch": 1.8443726440495423, "grad_norm": 1.2580366134643555, "learning_rate": 9.179780104456205e-06, "loss": 0.3066, "step": 429 }, { "epoch": 1.8486806677436727, "grad_norm": 1.1649465560913086, "learning_rate": 9.176060551719402e-06, "loss": 0.3316, "step": 430 }, { "epoch": 1.8529886914378029, "grad_norm": 1.2711234092712402, "learning_rate": 9.172333341421046e-06, "loss": 0.3419, "step": 431 }, { "epoch": 1.857296715131933, "grad_norm": 1.129465103149414, "learning_rate": 9.168598480395653e-06, "loss": 0.3054, "step": 432 }, { "epoch": 1.8616047388260637, "grad_norm": 1.1872928142547607, "learning_rate": 9.16485597549176e-06, "loss": 0.3148, "step": 433 }, { "epoch": 1.865912762520194, "grad_norm": 1.0811914205551147, "learning_rate": 9.16110583357193e-06, "loss": 0.3106, "step": 434 }, { "epoch": 1.870220786214324, "grad_norm": 1.087684154510498, "learning_rate": 9.157348061512728e-06, "loss": 0.2755, "step": 435 }, { "epoch": 1.8745288099084545, "grad_norm": 1.2518579959869385, "learning_rate": 9.153582666204702e-06, "loss": 0.3113, "step": 436 }, { "epoch": 1.878836833602585, "grad_norm": 1.1797399520874023, "learning_rate": 9.149809654552387e-06, "loss": 0.2762, "step": 437 }, { "epoch": 1.883144857296715, "grad_norm": 1.2617847919464111, "learning_rate": 9.146029033474284e-06, "loss": 0.3151, "step": 438 }, { "epoch": 1.8874528809908453, "grad_norm": 1.2301677465438843, "learning_rate": 9.142240809902841e-06, "loss": 0.3084, "step": 439 }, { "epoch": 1.891760904684976, "grad_norm": 1.0256375074386597, "learning_rate": 9.138444990784455e-06, "loss": 0.2933, "step": 440 }, { "epoch": 1.8960689283791061, "grad_norm": 1.1952866315841675, "learning_rate": 9.13464158307944e-06, "loss": 0.3283, "step": 441 }, { "epoch": 1.9003769520732363, "grad_norm": 1.086095929145813, "learning_rate": 9.130830593762037e-06, "loss": 0.2984, "step": 442 }, { "epoch": 1.9046849757673667, "grad_norm": 1.104231834411621, "learning_rate": 9.12701202982038e-06, "loss": 0.29, "step": 443 }, { "epoch": 1.9089929994614971, "grad_norm": 1.2137507200241089, "learning_rate": 9.123185898256497e-06, "loss": 0.3393, "step": 444 }, { "epoch": 1.9133010231556273, "grad_norm": 1.1142717599868774, "learning_rate": 9.119352206086292e-06, "loss": 0.3004, "step": 445 }, { "epoch": 1.9176090468497575, "grad_norm": 1.1249945163726807, "learning_rate": 9.115510960339533e-06, "loss": 0.3208, "step": 446 }, { "epoch": 1.921917070543888, "grad_norm": 1.124248743057251, "learning_rate": 9.111662168059836e-06, "loss": 0.3282, "step": 447 }, { "epoch": 1.9262250942380184, "grad_norm": 1.1833760738372803, "learning_rate": 9.107805836304658e-06, "loss": 0.3169, "step": 448 }, { "epoch": 1.9305331179321485, "grad_norm": 1.2385332584381104, "learning_rate": 9.10394197214528e-06, "loss": 0.2985, "step": 449 }, { "epoch": 1.934841141626279, "grad_norm": 1.1504766941070557, "learning_rate": 9.100070582666796e-06, "loss": 0.276, "step": 450 }, { "epoch": 1.9391491653204094, "grad_norm": 1.0979230403900146, "learning_rate": 9.096191674968095e-06, "loss": 0.2858, "step": 451 }, { "epoch": 1.9434571890145396, "grad_norm": 1.348646640777588, "learning_rate": 9.09230525616186e-06, "loss": 0.2994, "step": 452 }, { "epoch": 1.9477652127086698, "grad_norm": 0.9811099171638489, "learning_rate": 9.088411333374539e-06, "loss": 0.2648, "step": 453 }, { "epoch": 1.9520732364028002, "grad_norm": 1.2280666828155518, "learning_rate": 9.084509913746342e-06, "loss": 0.3234, "step": 454 }, { "epoch": 1.9563812600969306, "grad_norm": 1.1723735332489014, "learning_rate": 9.08060100443123e-06, "loss": 0.3148, "step": 455 }, { "epoch": 1.9606892837910608, "grad_norm": 1.0642163753509521, "learning_rate": 9.076684612596891e-06, "loss": 0.2749, "step": 456 }, { "epoch": 1.9649973074851912, "grad_norm": 1.2091552019119263, "learning_rate": 9.07276074542474e-06, "loss": 0.3087, "step": 457 }, { "epoch": 1.9693053311793216, "grad_norm": 1.1150028705596924, "learning_rate": 9.068829410109893e-06, "loss": 0.2935, "step": 458 }, { "epoch": 1.9736133548734518, "grad_norm": 1.1689749956130981, "learning_rate": 9.064890613861168e-06, "loss": 0.3, "step": 459 }, { "epoch": 1.977921378567582, "grad_norm": 1.2788281440734863, "learning_rate": 9.060944363901057e-06, "loss": 0.3223, "step": 460 }, { "epoch": 1.9822294022617124, "grad_norm": 1.1005408763885498, "learning_rate": 9.05699066746572e-06, "loss": 0.2945, "step": 461 }, { "epoch": 1.9865374259558428, "grad_norm": 1.036290168762207, "learning_rate": 9.05302953180498e-06, "loss": 0.2865, "step": 462 }, { "epoch": 1.990845449649973, "grad_norm": 1.2217515707015991, "learning_rate": 9.04906096418229e-06, "loss": 0.2861, "step": 463 }, { "epoch": 1.9951534733441034, "grad_norm": 1.3134533166885376, "learning_rate": 9.045084971874738e-06, "loss": 0.3219, "step": 464 }, { "epoch": 1.9994614970382338, "grad_norm": 1.2293431758880615, "learning_rate": 9.041101562173023e-06, "loss": 0.3182, "step": 465 }, { "epoch": 2.0, "grad_norm": 1.2293431758880615, "learning_rate": 9.037110742381445e-06, "loss": 0.2822, "step": 466 }, { "epoch": 2.00430802369413, "grad_norm": 3.1546876430511475, "learning_rate": 9.033112519817897e-06, "loss": 0.2881, "step": 467 }, { "epoch": 2.008616047388261, "grad_norm": 1.0440473556518555, "learning_rate": 9.02910690181384e-06, "loss": 0.2657, "step": 468 }, { "epoch": 2.012924071082391, "grad_norm": 1.118409514427185, "learning_rate": 9.0250938957143e-06, "loss": 0.2873, "step": 469 }, { "epoch": 2.017232094776521, "grad_norm": 1.241923451423645, "learning_rate": 9.021073508877845e-06, "loss": 0.2815, "step": 470 }, { "epoch": 2.0215401184706514, "grad_norm": 1.061378836631775, "learning_rate": 9.017045748676584e-06, "loss": 0.2414, "step": 471 }, { "epoch": 2.025848142164782, "grad_norm": 1.0081288814544678, "learning_rate": 9.013010622496145e-06, "loss": 0.262, "step": 472 }, { "epoch": 2.0301561658589122, "grad_norm": 1.1443556547164917, "learning_rate": 9.008968137735655e-06, "loss": 0.242, "step": 473 }, { "epoch": 2.0344641895530424, "grad_norm": 1.1716876029968262, "learning_rate": 9.004918301807746e-06, "loss": 0.2625, "step": 474 }, { "epoch": 2.038772213247173, "grad_norm": 1.0461676120758057, "learning_rate": 9.000861122138518e-06, "loss": 0.2781, "step": 475 }, { "epoch": 2.0430802369413033, "grad_norm": 1.1435906887054443, "learning_rate": 8.996796606167549e-06, "loss": 0.247, "step": 476 }, { "epoch": 2.0473882606354334, "grad_norm": 1.251587152481079, "learning_rate": 8.99272476134786e-06, "loss": 0.3036, "step": 477 }, { "epoch": 2.0516962843295636, "grad_norm": 1.122776746749878, "learning_rate": 8.988645595145913e-06, "loss": 0.2728, "step": 478 }, { "epoch": 2.0560043080236943, "grad_norm": 1.180074691772461, "learning_rate": 8.9845591150416e-06, "loss": 0.2753, "step": 479 }, { "epoch": 2.0603123317178245, "grad_norm": 1.2057586908340454, "learning_rate": 8.98046532852822e-06, "loss": 0.2843, "step": 480 }, { "epoch": 2.0646203554119547, "grad_norm": 1.2484440803527832, "learning_rate": 8.976364243112468e-06, "loss": 0.2795, "step": 481 }, { "epoch": 2.0689283791060853, "grad_norm": 1.1044594049453735, "learning_rate": 8.972255866314425e-06, "loss": 0.2648, "step": 482 }, { "epoch": 2.0732364028002155, "grad_norm": 1.1577068567276, "learning_rate": 8.968140205667544e-06, "loss": 0.2732, "step": 483 }, { "epoch": 2.0775444264943457, "grad_norm": 1.246111512184143, "learning_rate": 8.964017268718632e-06, "loss": 0.2846, "step": 484 }, { "epoch": 2.081852450188476, "grad_norm": 1.1163862943649292, "learning_rate": 8.959887063027837e-06, "loss": 0.2553, "step": 485 }, { "epoch": 2.0861604738826065, "grad_norm": 1.1406306028366089, "learning_rate": 8.95574959616864e-06, "loss": 0.2884, "step": 486 }, { "epoch": 2.0904684975767367, "grad_norm": 1.1482287645339966, "learning_rate": 8.951604875727833e-06, "loss": 0.2686, "step": 487 }, { "epoch": 2.094776521270867, "grad_norm": 1.154695987701416, "learning_rate": 8.94745290930551e-06, "loss": 0.2716, "step": 488 }, { "epoch": 2.099084544964997, "grad_norm": 1.103218674659729, "learning_rate": 8.94329370451505e-06, "loss": 0.2733, "step": 489 }, { "epoch": 2.1033925686591277, "grad_norm": 0.9970069527626038, "learning_rate": 8.93912726898311e-06, "loss": 0.2646, "step": 490 }, { "epoch": 2.107700592353258, "grad_norm": 1.1556833982467651, "learning_rate": 8.934953610349599e-06, "loss": 0.2726, "step": 491 }, { "epoch": 2.112008616047388, "grad_norm": 1.0892318487167358, "learning_rate": 8.930772736267675e-06, "loss": 0.2519, "step": 492 }, { "epoch": 2.1163166397415187, "grad_norm": 1.1029231548309326, "learning_rate": 8.926584654403725e-06, "loss": 0.2669, "step": 493 }, { "epoch": 2.120624663435649, "grad_norm": 1.066351294517517, "learning_rate": 8.922389372437357e-06, "loss": 0.2637, "step": 494 }, { "epoch": 2.124932687129779, "grad_norm": 1.2578282356262207, "learning_rate": 8.918186898061377e-06, "loss": 0.2597, "step": 495 }, { "epoch": 2.1292407108239093, "grad_norm": 1.187612533569336, "learning_rate": 8.91397723898178e-06, "loss": 0.2888, "step": 496 }, { "epoch": 2.13354873451804, "grad_norm": 1.1878846883773804, "learning_rate": 8.909760402917738e-06, "loss": 0.2698, "step": 497 }, { "epoch": 2.13785675821217, "grad_norm": 1.0593281984329224, "learning_rate": 8.90553639760158e-06, "loss": 0.254, "step": 498 }, { "epoch": 2.1421647819063003, "grad_norm": 1.052610993385315, "learning_rate": 8.901305230778783e-06, "loss": 0.2531, "step": 499 }, { "epoch": 2.146472805600431, "grad_norm": 1.1213619709014893, "learning_rate": 8.897066910207958e-06, "loss": 0.258, "step": 500 }, { "epoch": 2.150780829294561, "grad_norm": 1.1232720613479614, "learning_rate": 8.892821443660831e-06, "loss": 0.2612, "step": 501 }, { "epoch": 2.1550888529886914, "grad_norm": 1.1715662479400635, "learning_rate": 8.888568838922231e-06, "loss": 0.2534, "step": 502 }, { "epoch": 2.1593968766828215, "grad_norm": 1.1247590780258179, "learning_rate": 8.884309103790078e-06, "loss": 0.2638, "step": 503 }, { "epoch": 2.163704900376952, "grad_norm": 1.2371364831924438, "learning_rate": 8.880042246075366e-06, "loss": 0.2738, "step": 504 }, { "epoch": 2.1680129240710824, "grad_norm": 1.1520922183990479, "learning_rate": 8.875768273602148e-06, "loss": 0.2748, "step": 505 }, { "epoch": 2.1723209477652126, "grad_norm": 1.0599032640457153, "learning_rate": 8.871487194207527e-06, "loss": 0.2531, "step": 506 }, { "epoch": 2.176628971459343, "grad_norm": 1.0814327001571655, "learning_rate": 8.867199015741632e-06, "loss": 0.255, "step": 507 }, { "epoch": 2.1809369951534734, "grad_norm": 1.2039546966552734, "learning_rate": 8.862903746067619e-06, "loss": 0.2746, "step": 508 }, { "epoch": 2.1852450188476036, "grad_norm": 1.31549870967865, "learning_rate": 8.858601393061634e-06, "loss": 0.3024, "step": 509 }, { "epoch": 2.1895530425417338, "grad_norm": 1.1667834520339966, "learning_rate": 8.854291964612824e-06, "loss": 0.258, "step": 510 }, { "epoch": 2.1938610662358644, "grad_norm": 1.1152901649475098, "learning_rate": 8.849975468623302e-06, "loss": 0.2649, "step": 511 }, { "epoch": 2.1981690899299946, "grad_norm": 1.2007540464401245, "learning_rate": 8.845651913008145e-06, "loss": 0.2783, "step": 512 }, { "epoch": 2.202477113624125, "grad_norm": 1.0194883346557617, "learning_rate": 8.841321305695372e-06, "loss": 0.2759, "step": 513 }, { "epoch": 2.2067851373182554, "grad_norm": 1.182667851448059, "learning_rate": 8.836983654625934e-06, "loss": 0.2834, "step": 514 }, { "epoch": 2.2110931610123856, "grad_norm": 1.1789295673370361, "learning_rate": 8.832638967753699e-06, "loss": 0.2656, "step": 515 }, { "epoch": 2.215401184706516, "grad_norm": 1.1069103479385376, "learning_rate": 8.828287253045436e-06, "loss": 0.267, "step": 516 }, { "epoch": 2.219709208400646, "grad_norm": 1.0038608312606812, "learning_rate": 8.823928518480797e-06, "loss": 0.2508, "step": 517 }, { "epoch": 2.2240172320947766, "grad_norm": 1.0850584506988525, "learning_rate": 8.819562772052312e-06, "loss": 0.2767, "step": 518 }, { "epoch": 2.228325255788907, "grad_norm": 1.3206738233566284, "learning_rate": 8.815190021765365e-06, "loss": 0.2784, "step": 519 }, { "epoch": 2.232633279483037, "grad_norm": 1.1923176050186157, "learning_rate": 8.810810275638183e-06, "loss": 0.2946, "step": 520 }, { "epoch": 2.2369413031771677, "grad_norm": 1.2443435192108154, "learning_rate": 8.806423541701824e-06, "loss": 0.2754, "step": 521 }, { "epoch": 2.241249326871298, "grad_norm": 1.262756109237671, "learning_rate": 8.802029828000157e-06, "loss": 0.3116, "step": 522 }, { "epoch": 2.245557350565428, "grad_norm": 1.215442419052124, "learning_rate": 8.797629142589846e-06, "loss": 0.27, "step": 523 }, { "epoch": 2.2498653742595582, "grad_norm": 1.295694351196289, "learning_rate": 8.793221493540347e-06, "loss": 0.2929, "step": 524 }, { "epoch": 2.254173397953689, "grad_norm": 1.217462182044983, "learning_rate": 8.788806888933881e-06, "loss": 0.2704, "step": 525 }, { "epoch": 2.258481421647819, "grad_norm": 1.215073823928833, "learning_rate": 8.784385336865419e-06, "loss": 0.2704, "step": 526 }, { "epoch": 2.2627894453419493, "grad_norm": 1.012057900428772, "learning_rate": 8.779956845442682e-06, "loss": 0.268, "step": 527 }, { "epoch": 2.26709746903608, "grad_norm": 1.0462805032730103, "learning_rate": 8.775521422786104e-06, "loss": 0.2601, "step": 528 }, { "epoch": 2.27140549273021, "grad_norm": 0.9762120246887207, "learning_rate": 8.771079077028836e-06, "loss": 0.2591, "step": 529 }, { "epoch": 2.2757135164243403, "grad_norm": 1.0855119228363037, "learning_rate": 8.766629816316722e-06, "loss": 0.2587, "step": 530 }, { "epoch": 2.2800215401184705, "grad_norm": 1.0847527980804443, "learning_rate": 8.762173648808283e-06, "loss": 0.2675, "step": 531 }, { "epoch": 2.284329563812601, "grad_norm": 1.3327538967132568, "learning_rate": 8.757710582674708e-06, "loss": 0.2818, "step": 532 }, { "epoch": 2.2886375875067313, "grad_norm": 1.25534987449646, "learning_rate": 8.753240626099836e-06, "loss": 0.2814, "step": 533 }, { "epoch": 2.2929456112008615, "grad_norm": 1.1547666788101196, "learning_rate": 8.748763787280142e-06, "loss": 0.2707, "step": 534 }, { "epoch": 2.297253634894992, "grad_norm": 1.1245348453521729, "learning_rate": 8.744280074424713e-06, "loss": 0.2727, "step": 535 }, { "epoch": 2.3015616585891223, "grad_norm": 1.1245784759521484, "learning_rate": 8.739789495755254e-06, "loss": 0.2665, "step": 536 }, { "epoch": 2.3058696822832525, "grad_norm": 1.0653190612792969, "learning_rate": 8.735292059506047e-06, "loss": 0.2659, "step": 537 }, { "epoch": 2.3101777059773827, "grad_norm": 1.1056517362594604, "learning_rate": 8.730787773923957e-06, "loss": 0.2779, "step": 538 }, { "epoch": 2.3144857296715133, "grad_norm": 1.0632582902908325, "learning_rate": 8.726276647268403e-06, "loss": 0.2663, "step": 539 }, { "epoch": 2.3187937533656435, "grad_norm": 1.2258599996566772, "learning_rate": 8.721758687811353e-06, "loss": 0.2766, "step": 540 }, { "epoch": 2.3231017770597737, "grad_norm": 1.1886253356933594, "learning_rate": 8.717233903837298e-06, "loss": 0.2752, "step": 541 }, { "epoch": 2.3274098007539044, "grad_norm": 1.1761081218719482, "learning_rate": 8.712702303643254e-06, "loss": 0.2943, "step": 542 }, { "epoch": 2.3317178244480345, "grad_norm": 1.162219762802124, "learning_rate": 8.708163895538722e-06, "loss": 0.285, "step": 543 }, { "epoch": 2.3360258481421647, "grad_norm": 1.1694642305374146, "learning_rate": 8.703618687845697e-06, "loss": 0.2745, "step": 544 }, { "epoch": 2.340333871836295, "grad_norm": 1.2582859992980957, "learning_rate": 8.699066688898636e-06, "loss": 0.2848, "step": 545 }, { "epoch": 2.3446418955304256, "grad_norm": 1.1336910724639893, "learning_rate": 8.694507907044454e-06, "loss": 0.2534, "step": 546 }, { "epoch": 2.3489499192245558, "grad_norm": 1.0225831270217896, "learning_rate": 8.6899423506425e-06, "loss": 0.2678, "step": 547 }, { "epoch": 2.353257942918686, "grad_norm": 1.2403515577316284, "learning_rate": 8.685370028064546e-06, "loss": 0.2958, "step": 548 }, { "epoch": 2.3575659666128166, "grad_norm": 1.2434167861938477, "learning_rate": 8.680790947694772e-06, "loss": 0.2834, "step": 549 }, { "epoch": 2.361873990306947, "grad_norm": 1.179931879043579, "learning_rate": 8.676205117929752e-06, "loss": 0.2843, "step": 550 }, { "epoch": 2.366182014001077, "grad_norm": 1.3219486474990845, "learning_rate": 8.671612547178428e-06, "loss": 0.2994, "step": 551 }, { "epoch": 2.370490037695207, "grad_norm": 1.206898808479309, "learning_rate": 8.667013243862113e-06, "loss": 0.2638, "step": 552 }, { "epoch": 2.374798061389338, "grad_norm": 1.1221058368682861, "learning_rate": 8.66240721641446e-06, "loss": 0.268, "step": 553 }, { "epoch": 2.379106085083468, "grad_norm": 1.2333509922027588, "learning_rate": 8.657794473281447e-06, "loss": 0.2785, "step": 554 }, { "epoch": 2.383414108777598, "grad_norm": 1.1744508743286133, "learning_rate": 8.65317502292138e-06, "loss": 0.2484, "step": 555 }, { "epoch": 2.387722132471729, "grad_norm": 1.1270567178726196, "learning_rate": 8.64854887380485e-06, "loss": 0.2701, "step": 556 }, { "epoch": 2.392030156165859, "grad_norm": 1.143323302268982, "learning_rate": 8.643916034414741e-06, "loss": 0.2557, "step": 557 }, { "epoch": 2.396338179859989, "grad_norm": 1.044955849647522, "learning_rate": 8.639276513246199e-06, "loss": 0.2598, "step": 558 }, { "epoch": 2.4006462035541194, "grad_norm": 1.4415291547775269, "learning_rate": 8.634630318806626e-06, "loss": 0.3084, "step": 559 }, { "epoch": 2.40495422724825, "grad_norm": 1.0152852535247803, "learning_rate": 8.629977459615655e-06, "loss": 0.2794, "step": 560 }, { "epoch": 2.4092622509423802, "grad_norm": 1.1669901609420776, "learning_rate": 8.62531794420515e-06, "loss": 0.266, "step": 561 }, { "epoch": 2.4135702746365104, "grad_norm": 1.144785761833191, "learning_rate": 8.620651781119169e-06, "loss": 0.2543, "step": 562 }, { "epoch": 2.417878298330641, "grad_norm": 1.259813904762268, "learning_rate": 8.615978978913968e-06, "loss": 0.2938, "step": 563 }, { "epoch": 2.4221863220247712, "grad_norm": 1.23562490940094, "learning_rate": 8.611299546157973e-06, "loss": 0.2729, "step": 564 }, { "epoch": 2.4264943457189014, "grad_norm": 0.960300087928772, "learning_rate": 8.60661349143177e-06, "loss": 0.2587, "step": 565 }, { "epoch": 2.4308023694130316, "grad_norm": 1.2794194221496582, "learning_rate": 8.601920823328088e-06, "loss": 0.2587, "step": 566 }, { "epoch": 2.4351103931071623, "grad_norm": 1.182953119277954, "learning_rate": 8.59722155045178e-06, "loss": 0.2771, "step": 567 }, { "epoch": 2.4394184168012925, "grad_norm": 1.170477032661438, "learning_rate": 8.592515681419812e-06, "loss": 0.2501, "step": 568 }, { "epoch": 2.4437264404954226, "grad_norm": 1.3360447883605957, "learning_rate": 8.587803224861248e-06, "loss": 0.2796, "step": 569 }, { "epoch": 2.4480344641895533, "grad_norm": 1.1564768552780151, "learning_rate": 8.583084189417225e-06, "loss": 0.2855, "step": 570 }, { "epoch": 2.4523424878836835, "grad_norm": 1.155604600906372, "learning_rate": 8.578358583740947e-06, "loss": 0.2739, "step": 571 }, { "epoch": 2.4566505115778137, "grad_norm": 1.1428308486938477, "learning_rate": 8.573626416497669e-06, "loss": 0.2809, "step": 572 }, { "epoch": 2.460958535271944, "grad_norm": 1.1681113243103027, "learning_rate": 8.568887696364673e-06, "loss": 0.2713, "step": 573 }, { "epoch": 2.4652665589660745, "grad_norm": 1.1967153549194336, "learning_rate": 8.564142432031257e-06, "loss": 0.2722, "step": 574 }, { "epoch": 2.4695745826602047, "grad_norm": 1.1002535820007324, "learning_rate": 8.559390632198723e-06, "loss": 0.2521, "step": 575 }, { "epoch": 2.473882606354335, "grad_norm": 1.262689232826233, "learning_rate": 8.554632305580355e-06, "loss": 0.2826, "step": 576 }, { "epoch": 2.4781906300484655, "grad_norm": 1.1231080293655396, "learning_rate": 8.549867460901402e-06, "loss": 0.265, "step": 577 }, { "epoch": 2.4824986537425957, "grad_norm": 1.1673024892807007, "learning_rate": 8.545096106899068e-06, "loss": 0.2862, "step": 578 }, { "epoch": 2.486806677436726, "grad_norm": 1.1340245008468628, "learning_rate": 8.540318252322493e-06, "loss": 0.2839, "step": 579 }, { "epoch": 2.491114701130856, "grad_norm": 1.0350943803787231, "learning_rate": 8.535533905932739e-06, "loss": 0.2645, "step": 580 }, { "epoch": 2.4954227248249867, "grad_norm": 1.1802111864089966, "learning_rate": 8.530743076502766e-06, "loss": 0.2818, "step": 581 }, { "epoch": 2.499730748519117, "grad_norm": 1.2306886911392212, "learning_rate": 8.525945772817427e-06, "loss": 0.2845, "step": 582 }, { "epoch": 2.504038772213247, "grad_norm": 1.173303484916687, "learning_rate": 8.521142003673447e-06, "loss": 0.2573, "step": 583 }, { "epoch": 2.5083467959073777, "grad_norm": 1.0806212425231934, "learning_rate": 8.5163317778794e-06, "loss": 0.2879, "step": 584 }, { "epoch": 2.512654819601508, "grad_norm": 1.1595402956008911, "learning_rate": 8.51151510425571e-06, "loss": 0.2919, "step": 585 }, { "epoch": 2.516962843295638, "grad_norm": 1.1446491479873657, "learning_rate": 8.506691991634612e-06, "loss": 0.2732, "step": 586 }, { "epoch": 2.5212708669897683, "grad_norm": 1.172248363494873, "learning_rate": 8.501862448860159e-06, "loss": 0.272, "step": 587 }, { "epoch": 2.5255788906838985, "grad_norm": 1.0578157901763916, "learning_rate": 8.497026484788189e-06, "loss": 0.2757, "step": 588 }, { "epoch": 2.529886914378029, "grad_norm": 1.0634915828704834, "learning_rate": 8.492184108286316e-06, "loss": 0.2678, "step": 589 }, { "epoch": 2.5341949380721593, "grad_norm": 1.2673629522323608, "learning_rate": 8.487335328233912e-06, "loss": 0.2976, "step": 590 }, { "epoch": 2.53850296176629, "grad_norm": 1.1585445404052734, "learning_rate": 8.48248015352209e-06, "loss": 0.2761, "step": 591 }, { "epoch": 2.54281098546042, "grad_norm": 0.9961324334144592, "learning_rate": 8.477618593053693e-06, "loss": 0.2643, "step": 592 }, { "epoch": 2.5471190091545504, "grad_norm": 1.0971559286117554, "learning_rate": 8.47275065574327e-06, "loss": 0.2787, "step": 593 }, { "epoch": 2.5514270328486806, "grad_norm": 1.1897369623184204, "learning_rate": 8.46787635051706e-06, "loss": 0.2802, "step": 594 }, { "epoch": 2.5557350565428107, "grad_norm": 1.2767349481582642, "learning_rate": 8.462995686312985e-06, "loss": 0.2802, "step": 595 }, { "epoch": 2.5600430802369414, "grad_norm": 1.2634470462799072, "learning_rate": 8.458108672080624e-06, "loss": 0.2804, "step": 596 }, { "epoch": 2.5643511039310716, "grad_norm": 0.9885671138763428, "learning_rate": 8.453215316781205e-06, "loss": 0.2496, "step": 597 }, { "epoch": 2.568659127625202, "grad_norm": 1.0467156171798706, "learning_rate": 8.448315629387572e-06, "loss": 0.2893, "step": 598 }, { "epoch": 2.5729671513193324, "grad_norm": 1.1877309083938599, "learning_rate": 8.44340961888419e-06, "loss": 0.2764, "step": 599 }, { "epoch": 2.5772751750134626, "grad_norm": 1.0868911743164062, "learning_rate": 8.438497294267117e-06, "loss": 0.2768, "step": 600 }, { "epoch": 2.581583198707593, "grad_norm": 1.1395546197891235, "learning_rate": 8.433578664543986e-06, "loss": 0.2741, "step": 601 }, { "epoch": 2.585891222401723, "grad_norm": 1.1566461324691772, "learning_rate": 8.428653738733996e-06, "loss": 0.2667, "step": 602 }, { "epoch": 2.5901992460958536, "grad_norm": 1.0726302862167358, "learning_rate": 8.423722525867883e-06, "loss": 0.26, "step": 603 }, { "epoch": 2.594507269789984, "grad_norm": 1.067800521850586, "learning_rate": 8.418785034987921e-06, "loss": 0.2638, "step": 604 }, { "epoch": 2.598815293484114, "grad_norm": 1.1685556173324585, "learning_rate": 8.413841275147893e-06, "loss": 0.268, "step": 605 }, { "epoch": 2.6031233171782446, "grad_norm": 1.2059850692749023, "learning_rate": 8.408891255413072e-06, "loss": 0.2786, "step": 606 }, { "epoch": 2.607431340872375, "grad_norm": 1.1074217557907104, "learning_rate": 8.403934984860216e-06, "loss": 0.2858, "step": 607 }, { "epoch": 2.611739364566505, "grad_norm": 1.3857635259628296, "learning_rate": 8.39897247257754e-06, "loss": 0.2773, "step": 608 }, { "epoch": 2.616047388260635, "grad_norm": 1.0125895738601685, "learning_rate": 8.39400372766471e-06, "loss": 0.2692, "step": 609 }, { "epoch": 2.620355411954766, "grad_norm": 1.1065483093261719, "learning_rate": 8.389028759232816e-06, "loss": 0.2769, "step": 610 }, { "epoch": 2.624663435648896, "grad_norm": 1.1983776092529297, "learning_rate": 8.38404757640436e-06, "loss": 0.2938, "step": 611 }, { "epoch": 2.6289714593430262, "grad_norm": 1.1349999904632568, "learning_rate": 8.379060188313244e-06, "loss": 0.2734, "step": 612 }, { "epoch": 2.633279483037157, "grad_norm": 1.299493432044983, "learning_rate": 8.374066604104742e-06, "loss": 0.3074, "step": 613 }, { "epoch": 2.637587506731287, "grad_norm": 1.1134788990020752, "learning_rate": 8.369066832935498e-06, "loss": 0.2749, "step": 614 }, { "epoch": 2.6418955304254172, "grad_norm": 1.025368332862854, "learning_rate": 8.364060883973488e-06, "loss": 0.2579, "step": 615 }, { "epoch": 2.6462035541195474, "grad_norm": 1.0931003093719482, "learning_rate": 8.359048766398032e-06, "loss": 0.2554, "step": 616 }, { "epoch": 2.650511577813678, "grad_norm": 1.0344171524047852, "learning_rate": 8.354030489399747e-06, "loss": 0.258, "step": 617 }, { "epoch": 2.6548196015078083, "grad_norm": 1.1382379531860352, "learning_rate": 8.349006062180552e-06, "loss": 0.2822, "step": 618 }, { "epoch": 2.6591276252019385, "grad_norm": 1.2358250617980957, "learning_rate": 8.343975493953645e-06, "loss": 0.2771, "step": 619 }, { "epoch": 2.663435648896069, "grad_norm": 1.0387742519378662, "learning_rate": 8.338938793943478e-06, "loss": 0.2694, "step": 620 }, { "epoch": 2.6677436725901993, "grad_norm": 1.1414376497268677, "learning_rate": 8.333895971385754e-06, "loss": 0.2515, "step": 621 }, { "epoch": 2.6720516962843295, "grad_norm": 1.1614770889282227, "learning_rate": 8.328847035527397e-06, "loss": 0.2757, "step": 622 }, { "epoch": 2.6763597199784597, "grad_norm": 1.193450689315796, "learning_rate": 8.323791995626543e-06, "loss": 0.267, "step": 623 }, { "epoch": 2.6806677436725903, "grad_norm": 1.1697204113006592, "learning_rate": 8.318730860952523e-06, "loss": 0.2941, "step": 624 }, { "epoch": 2.6849757673667205, "grad_norm": 1.1913179159164429, "learning_rate": 8.313663640785839e-06, "loss": 0.2795, "step": 625 }, { "epoch": 2.6892837910608507, "grad_norm": 1.208177089691162, "learning_rate": 8.308590344418158e-06, "loss": 0.2631, "step": 626 }, { "epoch": 2.6935918147549813, "grad_norm": 1.1633046865463257, "learning_rate": 8.303510981152283e-06, "loss": 0.2899, "step": 627 }, { "epoch": 2.6978998384491115, "grad_norm": 1.2242285013198853, "learning_rate": 8.298425560302146e-06, "loss": 0.291, "step": 628 }, { "epoch": 2.7022078621432417, "grad_norm": 1.1298656463623047, "learning_rate": 8.293334091192782e-06, "loss": 0.2581, "step": 629 }, { "epoch": 2.706515885837372, "grad_norm": 1.1022253036499023, "learning_rate": 8.288236583160322e-06, "loss": 0.2908, "step": 630 }, { "epoch": 2.7108239095315025, "grad_norm": 1.06134033203125, "learning_rate": 8.28313304555197e-06, "loss": 0.2556, "step": 631 }, { "epoch": 2.7151319332256327, "grad_norm": 1.176902413368225, "learning_rate": 8.278023487725981e-06, "loss": 0.2779, "step": 632 }, { "epoch": 2.719439956919763, "grad_norm": 1.2088485956192017, "learning_rate": 8.272907919051653e-06, "loss": 0.2868, "step": 633 }, { "epoch": 2.7237479806138936, "grad_norm": 1.0156350135803223, "learning_rate": 8.267786348909306e-06, "loss": 0.2657, "step": 634 }, { "epoch": 2.7280560043080238, "grad_norm": 1.1520148515701294, "learning_rate": 8.262658786690262e-06, "loss": 0.2854, "step": 635 }, { "epoch": 2.732364028002154, "grad_norm": 1.0428128242492676, "learning_rate": 8.257525241796837e-06, "loss": 0.2523, "step": 636 }, { "epoch": 2.736672051696284, "grad_norm": 1.107344627380371, "learning_rate": 8.252385723642312e-06, "loss": 0.2534, "step": 637 }, { "epoch": 2.7409800753904148, "grad_norm": 1.1237647533416748, "learning_rate": 8.247240241650918e-06, "loss": 0.2778, "step": 638 }, { "epoch": 2.745288099084545, "grad_norm": 1.1007157564163208, "learning_rate": 8.242088805257832e-06, "loss": 0.2587, "step": 639 }, { "epoch": 2.749596122778675, "grad_norm": 1.0266497135162354, "learning_rate": 8.23693142390914e-06, "loss": 0.2673, "step": 640 }, { "epoch": 2.753904146472806, "grad_norm": 1.1474441289901733, "learning_rate": 8.231768107061831e-06, "loss": 0.2715, "step": 641 }, { "epoch": 2.758212170166936, "grad_norm": 1.077664852142334, "learning_rate": 8.226598864183782e-06, "loss": 0.2775, "step": 642 }, { "epoch": 2.762520193861066, "grad_norm": 1.1297303438186646, "learning_rate": 8.221423704753733e-06, "loss": 0.2768, "step": 643 }, { "epoch": 2.7668282175551964, "grad_norm": 1.0387330055236816, "learning_rate": 8.216242638261277e-06, "loss": 0.2711, "step": 644 }, { "epoch": 2.771136241249327, "grad_norm": 1.1765904426574707, "learning_rate": 8.211055674206828e-06, "loss": 0.2531, "step": 645 }, { "epoch": 2.775444264943457, "grad_norm": 1.136173963546753, "learning_rate": 8.205862822101628e-06, "loss": 0.282, "step": 646 }, { "epoch": 2.7797522886375874, "grad_norm": 1.167690396308899, "learning_rate": 8.200664091467707e-06, "loss": 0.2864, "step": 647 }, { "epoch": 2.784060312331718, "grad_norm": 1.422500491142273, "learning_rate": 8.195459491837881e-06, "loss": 0.3164, "step": 648 }, { "epoch": 2.788368336025848, "grad_norm": 1.191558599472046, "learning_rate": 8.190249032755717e-06, "loss": 0.2878, "step": 649 }, { "epoch": 2.7926763597199784, "grad_norm": 1.290753960609436, "learning_rate": 8.18503272377554e-06, "loss": 0.2683, "step": 650 }, { "epoch": 2.7969843834141086, "grad_norm": 1.353398084640503, "learning_rate": 8.179810574462388e-06, "loss": 0.2662, "step": 651 }, { "epoch": 2.8012924071082392, "grad_norm": 1.180108666419983, "learning_rate": 8.17458259439202e-06, "loss": 0.2797, "step": 652 }, { "epoch": 2.8056004308023694, "grad_norm": 1.1766334772109985, "learning_rate": 8.169348793150884e-06, "loss": 0.2729, "step": 653 }, { "epoch": 2.8099084544964996, "grad_norm": 1.0973520278930664, "learning_rate": 8.164109180336094e-06, "loss": 0.2783, "step": 654 }, { "epoch": 2.8142164781906303, "grad_norm": 1.0976502895355225, "learning_rate": 8.15886376555543e-06, "loss": 0.2709, "step": 655 }, { "epoch": 2.8185245018847604, "grad_norm": 1.1745898723602295, "learning_rate": 8.153612558427311e-06, "loss": 0.2787, "step": 656 }, { "epoch": 2.8228325255788906, "grad_norm": 1.0667402744293213, "learning_rate": 8.148355568580768e-06, "loss": 0.268, "step": 657 }, { "epoch": 2.827140549273021, "grad_norm": 1.1354405879974365, "learning_rate": 8.143092805655445e-06, "loss": 0.2733, "step": 658 }, { "epoch": 2.8314485729671515, "grad_norm": 1.1692426204681396, "learning_rate": 8.13782427930157e-06, "loss": 0.2954, "step": 659 }, { "epoch": 2.8357565966612817, "grad_norm": 1.1983294486999512, "learning_rate": 8.132549999179934e-06, "loss": 0.2625, "step": 660 }, { "epoch": 2.840064620355412, "grad_norm": 1.0363659858703613, "learning_rate": 8.127269974961886e-06, "loss": 0.275, "step": 661 }, { "epoch": 2.8443726440495425, "grad_norm": 0.984197735786438, "learning_rate": 8.121984216329303e-06, "loss": 0.2625, "step": 662 }, { "epoch": 2.8486806677436727, "grad_norm": 1.1481293439865112, "learning_rate": 8.116692732974578e-06, "loss": 0.2931, "step": 663 }, { "epoch": 2.852988691437803, "grad_norm": 1.1083085536956787, "learning_rate": 8.111395534600604e-06, "loss": 0.2752, "step": 664 }, { "epoch": 2.857296715131933, "grad_norm": 1.171528935432434, "learning_rate": 8.10609263092075e-06, "loss": 0.2771, "step": 665 }, { "epoch": 2.8616047388260637, "grad_norm": 1.0989707708358765, "learning_rate": 8.100784031658846e-06, "loss": 0.2921, "step": 666 }, { "epoch": 2.865912762520194, "grad_norm": 1.1379913091659546, "learning_rate": 8.095469746549172e-06, "loss": 0.2938, "step": 667 }, { "epoch": 2.870220786214324, "grad_norm": 1.2145065069198608, "learning_rate": 8.090149785336426e-06, "loss": 0.2584, "step": 668 }, { "epoch": 2.8745288099084547, "grad_norm": 1.1817004680633545, "learning_rate": 8.084824157775719e-06, "loss": 0.256, "step": 669 }, { "epoch": 2.878836833602585, "grad_norm": 1.1178160905838013, "learning_rate": 8.079492873632554e-06, "loss": 0.2868, "step": 670 }, { "epoch": 2.883144857296715, "grad_norm": 1.0334300994873047, "learning_rate": 8.074155942682803e-06, "loss": 0.2779, "step": 671 }, { "epoch": 2.8874528809908453, "grad_norm": 1.0395854711532593, "learning_rate": 8.068813374712689e-06, "loss": 0.2841, "step": 672 }, { "epoch": 2.891760904684976, "grad_norm": 1.0593595504760742, "learning_rate": 8.06346517951878e-06, "loss": 0.262, "step": 673 }, { "epoch": 2.896068928379106, "grad_norm": 1.092256784439087, "learning_rate": 8.058111366907957e-06, "loss": 0.2724, "step": 674 }, { "epoch": 2.9003769520732363, "grad_norm": 1.1988204717636108, "learning_rate": 8.052751946697403e-06, "loss": 0.2866, "step": 675 }, { "epoch": 2.904684975767367, "grad_norm": 1.2021102905273438, "learning_rate": 8.047386928714583e-06, "loss": 0.2893, "step": 676 }, { "epoch": 2.908992999461497, "grad_norm": 1.1751456260681152, "learning_rate": 8.042016322797227e-06, "loss": 0.2708, "step": 677 }, { "epoch": 2.9133010231556273, "grad_norm": 1.0112767219543457, "learning_rate": 8.03664013879331e-06, "loss": 0.2577, "step": 678 }, { "epoch": 2.9176090468497575, "grad_norm": 1.0113451480865479, "learning_rate": 8.031258386561038e-06, "loss": 0.2815, "step": 679 }, { "epoch": 2.9219170705438877, "grad_norm": 1.1114753484725952, "learning_rate": 8.025871075968828e-06, "loss": 0.2618, "step": 680 }, { "epoch": 2.9262250942380184, "grad_norm": 1.0998746156692505, "learning_rate": 8.020478216895282e-06, "loss": 0.2721, "step": 681 }, { "epoch": 2.9305331179321485, "grad_norm": 1.0433932542800903, "learning_rate": 8.015079819229187e-06, "loss": 0.2852, "step": 682 }, { "epoch": 2.934841141626279, "grad_norm": 1.1116548776626587, "learning_rate": 8.009675892869478e-06, "loss": 0.2857, "step": 683 }, { "epoch": 2.9391491653204094, "grad_norm": 1.0505714416503906, "learning_rate": 8.00426644772523e-06, "loss": 0.2739, "step": 684 }, { "epoch": 2.9434571890145396, "grad_norm": 1.1751623153686523, "learning_rate": 7.99885149371564e-06, "loss": 0.2753, "step": 685 }, { "epoch": 2.9477652127086698, "grad_norm": 1.050262451171875, "learning_rate": 7.993431040770002e-06, "loss": 0.2772, "step": 686 }, { "epoch": 2.9520732364028, "grad_norm": 1.1920961141586304, "learning_rate": 7.988005098827699e-06, "loss": 0.2843, "step": 687 }, { "epoch": 2.9563812600969306, "grad_norm": 0.9941353797912598, "learning_rate": 7.982573677838172e-06, "loss": 0.2705, "step": 688 }, { "epoch": 2.9606892837910608, "grad_norm": 1.1231780052185059, "learning_rate": 7.977136787760916e-06, "loss": 0.2471, "step": 689 }, { "epoch": 2.9649973074851914, "grad_norm": 1.1649749279022217, "learning_rate": 7.97169443856545e-06, "loss": 0.2815, "step": 690 }, { "epoch": 2.9693053311793216, "grad_norm": 1.0866994857788086, "learning_rate": 7.966246640231303e-06, "loss": 0.2627, "step": 691 }, { "epoch": 2.973613354873452, "grad_norm": 1.1334017515182495, "learning_rate": 7.960793402748001e-06, "loss": 0.2622, "step": 692 }, { "epoch": 2.977921378567582, "grad_norm": 1.1142147779464722, "learning_rate": 7.955334736115038e-06, "loss": 0.2837, "step": 693 }, { "epoch": 2.982229402261712, "grad_norm": 1.1462668180465698, "learning_rate": 7.949870650341864e-06, "loss": 0.292, "step": 694 }, { "epoch": 2.986537425955843, "grad_norm": 1.1187039613723755, "learning_rate": 7.944401155447872e-06, "loss": 0.2695, "step": 695 }, { "epoch": 2.990845449649973, "grad_norm": 1.1680221557617188, "learning_rate": 7.938926261462366e-06, "loss": 0.2921, "step": 696 }, { "epoch": 2.9951534733441036, "grad_norm": 1.2197476625442505, "learning_rate": 7.933445978424555e-06, "loss": 0.281, "step": 697 }, { "epoch": 2.999461497038234, "grad_norm": 0.993202269077301, "learning_rate": 7.927960316383524e-06, "loss": 0.2581, "step": 698 }, { "epoch": 3.0, "grad_norm": 0.993202269077301, "learning_rate": 7.92246928539823e-06, "loss": 0.2727, "step": 699 }, { "epoch": 3.00430802369413, "grad_norm": 3.982164144515991, "learning_rate": 7.916972895537471e-06, "loss": 0.2609, "step": 700 }, { "epoch": 3.008616047388261, "grad_norm": 1.0742263793945312, "learning_rate": 7.911471156879866e-06, "loss": 0.2245, "step": 701 }, { "epoch": 3.012924071082391, "grad_norm": 1.0976648330688477, "learning_rate": 7.905964079513851e-06, "loss": 0.246, "step": 702 }, { "epoch": 3.017232094776521, "grad_norm": 0.9185388088226318, "learning_rate": 7.900451673537646e-06, "loss": 0.2439, "step": 703 }, { "epoch": 3.0215401184706514, "grad_norm": 1.2105441093444824, "learning_rate": 7.894933949059245e-06, "loss": 0.2327, "step": 704 }, { "epoch": 3.025848142164782, "grad_norm": 0.9731848239898682, "learning_rate": 7.88941091619639e-06, "loss": 0.2231, "step": 705 }, { "epoch": 3.0301561658589122, "grad_norm": 1.1078482866287231, "learning_rate": 7.883882585076558e-06, "loss": 0.2415, "step": 706 }, { "epoch": 3.0344641895530424, "grad_norm": 1.069043517112732, "learning_rate": 7.87834896583695e-06, "loss": 0.2229, "step": 707 }, { "epoch": 3.038772213247173, "grad_norm": 1.197335124015808, "learning_rate": 7.872810068624452e-06, "loss": 0.2466, "step": 708 }, { "epoch": 3.0430802369413033, "grad_norm": 1.0125994682312012, "learning_rate": 7.867265903595632e-06, "loss": 0.2238, "step": 709 }, { "epoch": 3.0473882606354334, "grad_norm": 1.086272120475769, "learning_rate": 7.86171648091672e-06, "loss": 0.213, "step": 710 }, { "epoch": 3.0516962843295636, "grad_norm": 1.0315756797790527, "learning_rate": 7.856161810763584e-06, "loss": 0.2318, "step": 711 }, { "epoch": 3.0560043080236943, "grad_norm": 1.1444330215454102, "learning_rate": 7.850601903321717e-06, "loss": 0.2378, "step": 712 }, { "epoch": 3.0603123317178245, "grad_norm": 1.039595365524292, "learning_rate": 7.845036768786214e-06, "loss": 0.2302, "step": 713 }, { "epoch": 3.0646203554119547, "grad_norm": 1.1095147132873535, "learning_rate": 7.839466417361753e-06, "loss": 0.2143, "step": 714 }, { "epoch": 3.0689283791060853, "grad_norm": 1.1502745151519775, "learning_rate": 7.833890859262579e-06, "loss": 0.2382, "step": 715 }, { "epoch": 3.0732364028002155, "grad_norm": 1.2574323415756226, "learning_rate": 7.828310104712488e-06, "loss": 0.244, "step": 716 }, { "epoch": 3.0775444264943457, "grad_norm": 1.0791950225830078, "learning_rate": 7.822724163944802e-06, "loss": 0.2397, "step": 717 }, { "epoch": 3.081852450188476, "grad_norm": 1.1324896812438965, "learning_rate": 7.81713304720235e-06, "loss": 0.222, "step": 718 }, { "epoch": 3.0861604738826065, "grad_norm": 1.1850955486297607, "learning_rate": 7.811536764737454e-06, "loss": 0.2423, "step": 719 }, { "epoch": 3.0904684975767367, "grad_norm": 1.2072985172271729, "learning_rate": 7.805935326811913e-06, "loss": 0.2454, "step": 720 }, { "epoch": 3.094776521270867, "grad_norm": 1.1842292547225952, "learning_rate": 7.800328743696973e-06, "loss": 0.2334, "step": 721 }, { "epoch": 3.099084544964997, "grad_norm": 1.2322347164154053, "learning_rate": 7.794717025673318e-06, "loss": 0.2215, "step": 722 }, { "epoch": 3.1033925686591277, "grad_norm": 1.1397418975830078, "learning_rate": 7.789100183031045e-06, "loss": 0.2492, "step": 723 }, { "epoch": 3.107700592353258, "grad_norm": 1.1660665273666382, "learning_rate": 7.783478226069652e-06, "loss": 0.2473, "step": 724 }, { "epoch": 3.112008616047388, "grad_norm": 1.162108063697815, "learning_rate": 7.777851165098012e-06, "loss": 0.2259, "step": 725 }, { "epoch": 3.1163166397415187, "grad_norm": 1.1074467897415161, "learning_rate": 7.772219010434359e-06, "loss": 0.2336, "step": 726 }, { "epoch": 3.120624663435649, "grad_norm": 1.180801272392273, "learning_rate": 7.766581772406266e-06, "loss": 0.2272, "step": 727 }, { "epoch": 3.124932687129779, "grad_norm": 0.9812674522399902, "learning_rate": 7.760939461350622e-06, "loss": 0.2382, "step": 728 }, { "epoch": 3.1292407108239093, "grad_norm": 1.1671547889709473, "learning_rate": 7.755292087613635e-06, "loss": 0.2206, "step": 729 }, { "epoch": 3.13354873451804, "grad_norm": 1.2392609119415283, "learning_rate": 7.749639661550775e-06, "loss": 0.2488, "step": 730 }, { "epoch": 3.13785675821217, "grad_norm": 1.2116121053695679, "learning_rate": 7.743982193526791e-06, "loss": 0.2501, "step": 731 }, { "epoch": 3.1421647819063003, "grad_norm": 1.0735809803009033, "learning_rate": 7.738319693915673e-06, "loss": 0.2338, "step": 732 }, { "epoch": 3.146472805600431, "grad_norm": 1.0651217699050903, "learning_rate": 7.732652173100634e-06, "loss": 0.2228, "step": 733 }, { "epoch": 3.150780829294561, "grad_norm": 1.3095252513885498, "learning_rate": 7.726979641474102e-06, "loss": 0.242, "step": 734 }, { "epoch": 3.1550888529886914, "grad_norm": 1.1507078409194946, "learning_rate": 7.721302109437686e-06, "loss": 0.2373, "step": 735 }, { "epoch": 3.1593968766828215, "grad_norm": 1.253400444984436, "learning_rate": 7.715619587402165e-06, "loss": 0.2334, "step": 736 }, { "epoch": 3.163704900376952, "grad_norm": 1.3200879096984863, "learning_rate": 7.709932085787473e-06, "loss": 0.2309, "step": 737 }, { "epoch": 3.1680129240710824, "grad_norm": 1.044729471206665, "learning_rate": 7.704239615022671e-06, "loss": 0.2303, "step": 738 }, { "epoch": 3.1723209477652126, "grad_norm": 1.179209589958191, "learning_rate": 7.698542185545932e-06, "loss": 0.2556, "step": 739 }, { "epoch": 3.176628971459343, "grad_norm": 1.2791534662246704, "learning_rate": 7.692839807804522e-06, "loss": 0.2395, "step": 740 }, { "epoch": 3.1809369951534734, "grad_norm": 1.1774635314941406, "learning_rate": 7.687132492254783e-06, "loss": 0.2371, "step": 741 }, { "epoch": 3.1852450188476036, "grad_norm": 1.0789783000946045, "learning_rate": 7.681420249362107e-06, "loss": 0.21, "step": 742 }, { "epoch": 3.1895530425417338, "grad_norm": 1.0741312503814697, "learning_rate": 7.675703089600926e-06, "loss": 0.2344, "step": 743 }, { "epoch": 3.1938610662358644, "grad_norm": 1.3222390413284302, "learning_rate": 7.669981023454682e-06, "loss": 0.2551, "step": 744 }, { "epoch": 3.1981690899299946, "grad_norm": 1.2439368963241577, "learning_rate": 7.664254061415818e-06, "loss": 0.228, "step": 745 }, { "epoch": 3.202477113624125, "grad_norm": 1.087741732597351, "learning_rate": 7.658522213985757e-06, "loss": 0.2405, "step": 746 }, { "epoch": 3.2067851373182554, "grad_norm": 1.1094344854354858, "learning_rate": 7.652785491674872e-06, "loss": 0.2201, "step": 747 }, { "epoch": 3.2110931610123856, "grad_norm": 1.1255855560302734, "learning_rate": 7.647043905002485e-06, "loss": 0.2416, "step": 748 }, { "epoch": 3.215401184706516, "grad_norm": 1.2410178184509277, "learning_rate": 7.641297464496828e-06, "loss": 0.2329, "step": 749 }, { "epoch": 3.219709208400646, "grad_norm": 0.9963125586509705, "learning_rate": 7.635546180695039e-06, "loss": 0.2226, "step": 750 }, { "epoch": 3.2240172320947766, "grad_norm": 1.0446877479553223, "learning_rate": 7.629790064143139e-06, "loss": 0.2307, "step": 751 }, { "epoch": 3.228325255788907, "grad_norm": 1.1236238479614258, "learning_rate": 7.624029125396004e-06, "loss": 0.2323, "step": 752 }, { "epoch": 3.232633279483037, "grad_norm": 1.1273385286331177, "learning_rate": 7.618263375017358e-06, "loss": 0.2292, "step": 753 }, { "epoch": 3.2369413031771677, "grad_norm": 1.2274240255355835, "learning_rate": 7.612492823579744e-06, "loss": 0.2217, "step": 754 }, { "epoch": 3.241249326871298, "grad_norm": 1.1309409141540527, "learning_rate": 7.606717481664515e-06, "loss": 0.244, "step": 755 }, { "epoch": 3.245557350565428, "grad_norm": 1.2856779098510742, "learning_rate": 7.600937359861799e-06, "loss": 0.2334, "step": 756 }, { "epoch": 3.2498653742595582, "grad_norm": 1.4764436483383179, "learning_rate": 7.595152468770497e-06, "loss": 0.2221, "step": 757 }, { "epoch": 3.254173397953689, "grad_norm": 1.205610752105713, "learning_rate": 7.589362818998251e-06, "loss": 0.2519, "step": 758 }, { "epoch": 3.258481421647819, "grad_norm": 1.2047350406646729, "learning_rate": 7.58356842116143e-06, "loss": 0.2378, "step": 759 }, { "epoch": 3.2627894453419493, "grad_norm": 1.2798528671264648, "learning_rate": 7.57776928588511e-06, "loss": 0.2558, "step": 760 }, { "epoch": 3.26709746903608, "grad_norm": 1.215135097503662, "learning_rate": 7.571965423803052e-06, "loss": 0.2392, "step": 761 }, { "epoch": 3.27140549273021, "grad_norm": 1.2784080505371094, "learning_rate": 7.566156845557684e-06, "loss": 0.2295, "step": 762 }, { "epoch": 3.2757135164243403, "grad_norm": 1.2132107019424438, "learning_rate": 7.560343561800087e-06, "loss": 0.2512, "step": 763 }, { "epoch": 3.2800215401184705, "grad_norm": 1.1378742456436157, "learning_rate": 7.554525583189969e-06, "loss": 0.2488, "step": 764 }, { "epoch": 3.284329563812601, "grad_norm": 1.1240780353546143, "learning_rate": 7.548702920395639e-06, "loss": 0.2278, "step": 765 }, { "epoch": 3.2886375875067313, "grad_norm": 1.1924443244934082, "learning_rate": 7.542875584094006e-06, "loss": 0.2479, "step": 766 }, { "epoch": 3.2929456112008615, "grad_norm": 1.1710002422332764, "learning_rate": 7.537043584970543e-06, "loss": 0.2476, "step": 767 }, { "epoch": 3.297253634894992, "grad_norm": 1.15700364112854, "learning_rate": 7.53120693371927e-06, "loss": 0.2281, "step": 768 }, { "epoch": 3.3015616585891223, "grad_norm": 1.0568920373916626, "learning_rate": 7.525365641042749e-06, "loss": 0.2426, "step": 769 }, { "epoch": 3.3058696822832525, "grad_norm": 1.065804123878479, "learning_rate": 7.519519717652039e-06, "loss": 0.2328, "step": 770 }, { "epoch": 3.3101777059773827, "grad_norm": 1.2565348148345947, "learning_rate": 7.5136691742667e-06, "loss": 0.2393, "step": 771 }, { "epoch": 3.3144857296715133, "grad_norm": 1.1730437278747559, "learning_rate": 7.507814021614761e-06, "loss": 0.2447, "step": 772 }, { "epoch": 3.3187937533656435, "grad_norm": 1.1181384325027466, "learning_rate": 7.501954270432701e-06, "loss": 0.2368, "step": 773 }, { "epoch": 3.3231017770597737, "grad_norm": 1.0160866975784302, "learning_rate": 7.496089931465432e-06, "loss": 0.2441, "step": 774 }, { "epoch": 3.3274098007539044, "grad_norm": 1.2483638525009155, "learning_rate": 7.490221015466279e-06, "loss": 0.2432, "step": 775 }, { "epoch": 3.3317178244480345, "grad_norm": 1.3620024919509888, "learning_rate": 7.4843475331969614e-06, "loss": 0.2329, "step": 776 }, { "epoch": 3.3360258481421647, "grad_norm": 1.0547987222671509, "learning_rate": 7.478469495427569e-06, "loss": 0.2534, "step": 777 }, { "epoch": 3.340333871836295, "grad_norm": 1.240760326385498, "learning_rate": 7.4725869129365484e-06, "loss": 0.2398, "step": 778 }, { "epoch": 3.3446418955304256, "grad_norm": 1.1723148822784424, "learning_rate": 7.4666997965106725e-06, "loss": 0.2405, "step": 779 }, { "epoch": 3.3489499192245558, "grad_norm": 1.0330522060394287, "learning_rate": 7.4608081569450365e-06, "loss": 0.2223, "step": 780 }, { "epoch": 3.353257942918686, "grad_norm": 1.2223782539367676, "learning_rate": 7.4549120050430265e-06, "loss": 0.2474, "step": 781 }, { "epoch": 3.3575659666128166, "grad_norm": 1.1876544952392578, "learning_rate": 7.449011351616302e-06, "loss": 0.2307, "step": 782 }, { "epoch": 3.361873990306947, "grad_norm": 1.106420636177063, "learning_rate": 7.443106207484776e-06, "loss": 0.2428, "step": 783 }, { "epoch": 3.366182014001077, "grad_norm": 1.1982386112213135, "learning_rate": 7.437196583476597e-06, "loss": 0.2221, "step": 784 }, { "epoch": 3.370490037695207, "grad_norm": 1.1878842115402222, "learning_rate": 7.43128249042813e-06, "loss": 0.2381, "step": 785 }, { "epoch": 3.374798061389338, "grad_norm": 1.3298442363739014, "learning_rate": 7.425363939183931e-06, "loss": 0.2525, "step": 786 }, { "epoch": 3.379106085083468, "grad_norm": 1.217363953590393, "learning_rate": 7.419440940596735e-06, "loss": 0.2521, "step": 787 }, { "epoch": 3.383414108777598, "grad_norm": 1.2789652347564697, "learning_rate": 7.41351350552743e-06, "loss": 0.2458, "step": 788 }, { "epoch": 3.387722132471729, "grad_norm": 1.1601816415786743, "learning_rate": 7.407581644845038e-06, "loss": 0.2478, "step": 789 }, { "epoch": 3.392030156165859, "grad_norm": 1.2335379123687744, "learning_rate": 7.401645369426697e-06, "loss": 0.2388, "step": 790 }, { "epoch": 3.396338179859989, "grad_norm": 1.083685040473938, "learning_rate": 7.395704690157644e-06, "loss": 0.2375, "step": 791 }, { "epoch": 3.4006462035541194, "grad_norm": 1.2122055292129517, "learning_rate": 7.389759617931183e-06, "loss": 0.2499, "step": 792 }, { "epoch": 3.40495422724825, "grad_norm": 1.1563218832015991, "learning_rate": 7.383810163648682e-06, "loss": 0.2216, "step": 793 }, { "epoch": 3.4092622509423802, "grad_norm": 1.080259919166565, "learning_rate": 7.3778563382195365e-06, "loss": 0.2356, "step": 794 }, { "epoch": 3.4135702746365104, "grad_norm": 1.132384181022644, "learning_rate": 7.371898152561166e-06, "loss": 0.2317, "step": 795 }, { "epoch": 3.417878298330641, "grad_norm": 1.1473846435546875, "learning_rate": 7.365935617598975e-06, "loss": 0.2317, "step": 796 }, { "epoch": 3.4221863220247712, "grad_norm": 1.137619972229004, "learning_rate": 7.359968744266353e-06, "loss": 0.2546, "step": 797 }, { "epoch": 3.4264943457189014, "grad_norm": 1.105951189994812, "learning_rate": 7.35399754350464e-06, "loss": 0.2373, "step": 798 }, { "epoch": 3.4308023694130316, "grad_norm": 1.1990385055541992, "learning_rate": 7.3480220262631095e-06, "loss": 0.241, "step": 799 }, { "epoch": 3.4351103931071623, "grad_norm": 1.0479223728179932, "learning_rate": 7.342042203498952e-06, "loss": 0.2284, "step": 800 }, { "epoch": 3.4394184168012925, "grad_norm": 1.3471266031265259, "learning_rate": 7.336058086177253e-06, "loss": 0.2523, "step": 801 }, { "epoch": 3.4437264404954226, "grad_norm": 1.2163012027740479, "learning_rate": 7.330069685270976e-06, "loss": 0.2343, "step": 802 }, { "epoch": 3.4480344641895533, "grad_norm": 1.178816556930542, "learning_rate": 7.3240770117609325e-06, "loss": 0.2527, "step": 803 }, { "epoch": 3.4523424878836835, "grad_norm": 1.1871490478515625, "learning_rate": 7.318080076635773e-06, "loss": 0.2407, "step": 804 }, { "epoch": 3.4566505115778137, "grad_norm": 1.0711443424224854, "learning_rate": 7.312078890891962e-06, "loss": 0.2255, "step": 805 }, { "epoch": 3.460958535271944, "grad_norm": 1.1039179563522339, "learning_rate": 7.306073465533759e-06, "loss": 0.2482, "step": 806 }, { "epoch": 3.4652665589660745, "grad_norm": 1.336172342300415, "learning_rate": 7.300063811573194e-06, "loss": 0.2584, "step": 807 }, { "epoch": 3.4695745826602047, "grad_norm": 1.0815609693527222, "learning_rate": 7.294049940030055e-06, "loss": 0.2431, "step": 808 }, { "epoch": 3.473882606354335, "grad_norm": 1.1772844791412354, "learning_rate": 7.2880318619318605e-06, "loss": 0.256, "step": 809 }, { "epoch": 3.4781906300484655, "grad_norm": 1.0280383825302124, "learning_rate": 7.2820095883138456e-06, "loss": 0.2325, "step": 810 }, { "epoch": 3.4824986537425957, "grad_norm": 1.1324665546417236, "learning_rate": 7.2759831302189376e-06, "loss": 0.258, "step": 811 }, { "epoch": 3.486806677436726, "grad_norm": 1.217647671699524, "learning_rate": 7.269952498697734e-06, "loss": 0.2478, "step": 812 }, { "epoch": 3.491114701130856, "grad_norm": 1.1838194131851196, "learning_rate": 7.2639177048084894e-06, "loss": 0.2491, "step": 813 }, { "epoch": 3.4954227248249867, "grad_norm": 1.1052782535552979, "learning_rate": 7.25787875961709e-06, "loss": 0.2339, "step": 814 }, { "epoch": 3.499730748519117, "grad_norm": 1.1642377376556396, "learning_rate": 7.2518356741970285e-06, "loss": 0.231, "step": 815 }, { "epoch": 3.504038772213247, "grad_norm": 1.1658052206039429, "learning_rate": 7.245788459629397e-06, "loss": 0.2361, "step": 816 }, { "epoch": 3.5083467959073777, "grad_norm": 1.129423975944519, "learning_rate": 7.239737127002854e-06, "loss": 0.2261, "step": 817 }, { "epoch": 3.512654819601508, "grad_norm": 1.0502065420150757, "learning_rate": 7.233681687413614e-06, "loss": 0.2351, "step": 818 }, { "epoch": 3.516962843295638, "grad_norm": 1.1234469413757324, "learning_rate": 7.227622151965418e-06, "loss": 0.2404, "step": 819 }, { "epoch": 3.5212708669897683, "grad_norm": 1.1705223321914673, "learning_rate": 7.221558531769519e-06, "loss": 0.2525, "step": 820 }, { "epoch": 3.5255788906838985, "grad_norm": 1.1959092617034912, "learning_rate": 7.21549083794466e-06, "loss": 0.2333, "step": 821 }, { "epoch": 3.529886914378029, "grad_norm": 1.15984308719635, "learning_rate": 7.209419081617055e-06, "loss": 0.2368, "step": 822 }, { "epoch": 3.5341949380721593, "grad_norm": 1.1602686643600464, "learning_rate": 7.203343273920365e-06, "loss": 0.2366, "step": 823 }, { "epoch": 3.53850296176629, "grad_norm": 1.048525094985962, "learning_rate": 7.197263425995682e-06, "loss": 0.2278, "step": 824 }, { "epoch": 3.54281098546042, "grad_norm": 1.1909091472625732, "learning_rate": 7.191179548991507e-06, "loss": 0.2439, "step": 825 }, { "epoch": 3.5471190091545504, "grad_norm": 1.1533970832824707, "learning_rate": 7.185091654063724e-06, "loss": 0.2604, "step": 826 }, { "epoch": 3.5514270328486806, "grad_norm": 1.1275250911712646, "learning_rate": 7.1789997523755915e-06, "loss": 0.2503, "step": 827 }, { "epoch": 3.5557350565428107, "grad_norm": 1.1212456226348877, "learning_rate": 7.172903855097712e-06, "loss": 0.2533, "step": 828 }, { "epoch": 3.5600430802369414, "grad_norm": 1.2368807792663574, "learning_rate": 7.166803973408012e-06, "loss": 0.2428, "step": 829 }, { "epoch": 3.5643511039310716, "grad_norm": 1.1801507472991943, "learning_rate": 7.160700118491729e-06, "loss": 0.25, "step": 830 }, { "epoch": 3.568659127625202, "grad_norm": 1.083946704864502, "learning_rate": 7.154592301541383e-06, "loss": 0.2428, "step": 831 }, { "epoch": 3.5729671513193324, "grad_norm": 1.1042693853378296, "learning_rate": 7.148480533756759e-06, "loss": 0.2329, "step": 832 }, { "epoch": 3.5772751750134626, "grad_norm": 1.1040000915527344, "learning_rate": 7.142364826344891e-06, "loss": 0.2472, "step": 833 }, { "epoch": 3.581583198707593, "grad_norm": 1.2016639709472656, "learning_rate": 7.1362451905200285e-06, "loss": 0.2292, "step": 834 }, { "epoch": 3.585891222401723, "grad_norm": 1.340988278388977, "learning_rate": 7.130121637503633e-06, "loss": 0.2513, "step": 835 }, { "epoch": 3.5901992460958536, "grad_norm": 1.1760993003845215, "learning_rate": 7.123994178524345e-06, "loss": 0.2509, "step": 836 }, { "epoch": 3.594507269789984, "grad_norm": 1.310850977897644, "learning_rate": 7.117862824817966e-06, "loss": 0.2448, "step": 837 }, { "epoch": 3.598815293484114, "grad_norm": 1.2277843952178955, "learning_rate": 7.1117275876274425e-06, "loss": 0.2298, "step": 838 }, { "epoch": 3.6031233171782446, "grad_norm": 1.2455551624298096, "learning_rate": 7.105588478202838e-06, "loss": 0.2282, "step": 839 }, { "epoch": 3.607431340872375, "grad_norm": 1.1012552976608276, "learning_rate": 7.099445507801324e-06, "loss": 0.2314, "step": 840 }, { "epoch": 3.611739364566505, "grad_norm": 1.203344464302063, "learning_rate": 7.093298687687141e-06, "loss": 0.242, "step": 841 }, { "epoch": 3.616047388260635, "grad_norm": 1.2591255903244019, "learning_rate": 7.0871480291315975e-06, "loss": 0.2366, "step": 842 }, { "epoch": 3.620355411954766, "grad_norm": 1.166508436203003, "learning_rate": 7.080993543413035e-06, "loss": 0.2504, "step": 843 }, { "epoch": 3.624663435648896, "grad_norm": 1.1004966497421265, "learning_rate": 7.0748352418168174e-06, "loss": 0.2373, "step": 844 }, { "epoch": 3.6289714593430262, "grad_norm": 1.08232581615448, "learning_rate": 7.068673135635302e-06, "loss": 0.2269, "step": 845 }, { "epoch": 3.633279483037157, "grad_norm": 1.1803537607192993, "learning_rate": 7.062507236167826e-06, "loss": 0.244, "step": 846 }, { "epoch": 3.637587506731287, "grad_norm": 1.169716238975525, "learning_rate": 7.056337554720676e-06, "loss": 0.2486, "step": 847 }, { "epoch": 3.6418955304254172, "grad_norm": 1.4283939599990845, "learning_rate": 7.050164102607081e-06, "loss": 0.2399, "step": 848 }, { "epoch": 3.6462035541195474, "grad_norm": 1.3042423725128174, "learning_rate": 7.043986891147179e-06, "loss": 0.2496, "step": 849 }, { "epoch": 3.650511577813678, "grad_norm": 1.0485799312591553, "learning_rate": 7.037805931668006e-06, "loss": 0.2347, "step": 850 }, { "epoch": 3.6548196015078083, "grad_norm": 1.167236566543579, "learning_rate": 7.031621235503464e-06, "loss": 0.2446, "step": 851 }, { "epoch": 3.6591276252019385, "grad_norm": 1.1593332290649414, "learning_rate": 7.025432813994315e-06, "loss": 0.2425, "step": 852 }, { "epoch": 3.663435648896069, "grad_norm": 1.0361429452896118, "learning_rate": 7.019240678488145e-06, "loss": 0.2474, "step": 853 }, { "epoch": 3.6677436725901993, "grad_norm": 1.1621315479278564, "learning_rate": 7.013044840339353e-06, "loss": 0.2375, "step": 854 }, { "epoch": 3.6720516962843295, "grad_norm": 1.1319594383239746, "learning_rate": 7.006845310909131e-06, "loss": 0.2291, "step": 855 }, { "epoch": 3.6763597199784597, "grad_norm": 1.2212448120117188, "learning_rate": 7.000642101565434e-06, "loss": 0.2389, "step": 856 }, { "epoch": 3.6806677436725903, "grad_norm": 1.0798578262329102, "learning_rate": 6.994435223682966e-06, "loss": 0.2367, "step": 857 }, { "epoch": 3.6849757673667205, "grad_norm": 1.1603275537490845, "learning_rate": 6.9882246886431615e-06, "loss": 0.228, "step": 858 }, { "epoch": 3.6892837910608507, "grad_norm": 1.018467903137207, "learning_rate": 6.982010507834158e-06, "loss": 0.2452, "step": 859 }, { "epoch": 3.6935918147549813, "grad_norm": 1.1213793754577637, "learning_rate": 6.975792692650778e-06, "loss": 0.2413, "step": 860 }, { "epoch": 3.6978998384491115, "grad_norm": 1.2307703495025635, "learning_rate": 6.969571254494509e-06, "loss": 0.2426, "step": 861 }, { "epoch": 3.7022078621432417, "grad_norm": 1.1938848495483398, "learning_rate": 6.963346204773483e-06, "loss": 0.2289, "step": 862 }, { "epoch": 3.706515885837372, "grad_norm": 0.944646418094635, "learning_rate": 6.957117554902452e-06, "loss": 0.2169, "step": 863 }, { "epoch": 3.7108239095315025, "grad_norm": 1.1695207357406616, "learning_rate": 6.950885316302773e-06, "loss": 0.2435, "step": 864 }, { "epoch": 3.7151319332256327, "grad_norm": 1.3754353523254395, "learning_rate": 6.94464950040238e-06, "loss": 0.2385, "step": 865 }, { "epoch": 3.719439956919763, "grad_norm": 1.2150930166244507, "learning_rate": 6.938410118635768e-06, "loss": 0.2401, "step": 866 }, { "epoch": 3.7237479806138936, "grad_norm": 1.150939702987671, "learning_rate": 6.9321671824439715e-06, "loss": 0.2434, "step": 867 }, { "epoch": 3.7280560043080238, "grad_norm": 1.0390138626098633, "learning_rate": 6.9259207032745415e-06, "loss": 0.2236, "step": 868 }, { "epoch": 3.732364028002154, "grad_norm": 1.2754530906677246, "learning_rate": 6.919670692581526e-06, "loss": 0.2503, "step": 869 }, { "epoch": 3.736672051696284, "grad_norm": 1.1730183362960815, "learning_rate": 6.913417161825449e-06, "loss": 0.2368, "step": 870 }, { "epoch": 3.7409800753904148, "grad_norm": 1.0978236198425293, "learning_rate": 6.907160122473291e-06, "loss": 0.2354, "step": 871 }, { "epoch": 3.745288099084545, "grad_norm": 1.2562264204025269, "learning_rate": 6.90089958599846e-06, "loss": 0.2417, "step": 872 }, { "epoch": 3.749596122778675, "grad_norm": 1.0816057920455933, "learning_rate": 6.894635563880785e-06, "loss": 0.242, "step": 873 }, { "epoch": 3.753904146472806, "grad_norm": 1.3008449077606201, "learning_rate": 6.88836806760648e-06, "loss": 0.2343, "step": 874 }, { "epoch": 3.758212170166936, "grad_norm": 1.0396487712860107, "learning_rate": 6.882097108668132e-06, "loss": 0.2319, "step": 875 }, { "epoch": 3.762520193861066, "grad_norm": 1.1531552076339722, "learning_rate": 6.875822698564678e-06, "loss": 0.2262, "step": 876 }, { "epoch": 3.7668282175551964, "grad_norm": 1.1645442247390747, "learning_rate": 6.869544848801383e-06, "loss": 0.2507, "step": 877 }, { "epoch": 3.771136241249327, "grad_norm": 1.1769391298294067, "learning_rate": 6.863263570889818e-06, "loss": 0.2431, "step": 878 }, { "epoch": 3.775444264943457, "grad_norm": 1.1122496128082275, "learning_rate": 6.85697887634784e-06, "loss": 0.2398, "step": 879 }, { "epoch": 3.7797522886375874, "grad_norm": 1.1032952070236206, "learning_rate": 6.850690776699574e-06, "loss": 0.239, "step": 880 }, { "epoch": 3.784060312331718, "grad_norm": 1.050400733947754, "learning_rate": 6.844399283475384e-06, "loss": 0.242, "step": 881 }, { "epoch": 3.788368336025848, "grad_norm": 1.0650246143341064, "learning_rate": 6.838104408211862e-06, "loss": 0.2389, "step": 882 }, { "epoch": 3.7926763597199784, "grad_norm": 1.082397699356079, "learning_rate": 6.831806162451799e-06, "loss": 0.2345, "step": 883 }, { "epoch": 3.7969843834141086, "grad_norm": 1.2133508920669556, "learning_rate": 6.825504557744167e-06, "loss": 0.2384, "step": 884 }, { "epoch": 3.8012924071082392, "grad_norm": 1.1948316097259521, "learning_rate": 6.819199605644093e-06, "loss": 0.2448, "step": 885 }, { "epoch": 3.8056004308023694, "grad_norm": 1.2425826787948608, "learning_rate": 6.812891317712851e-06, "loss": 0.2507, "step": 886 }, { "epoch": 3.8099084544964996, "grad_norm": 1.1594274044036865, "learning_rate": 6.806579705517824e-06, "loss": 0.2352, "step": 887 }, { "epoch": 3.8142164781906303, "grad_norm": 1.09205162525177, "learning_rate": 6.800264780632495e-06, "loss": 0.2441, "step": 888 }, { "epoch": 3.8185245018847604, "grad_norm": 1.052693247795105, "learning_rate": 6.793946554636417e-06, "loss": 0.2485, "step": 889 }, { "epoch": 3.8228325255788906, "grad_norm": 1.193164587020874, "learning_rate": 6.7876250391152e-06, "loss": 0.2316, "step": 890 }, { "epoch": 3.827140549273021, "grad_norm": 1.0423582792282104, "learning_rate": 6.781300245660487e-06, "loss": 0.2294, "step": 891 }, { "epoch": 3.8314485729671515, "grad_norm": 1.1902750730514526, "learning_rate": 6.774972185869928e-06, "loss": 0.2553, "step": 892 }, { "epoch": 3.8357565966612817, "grad_norm": 1.131846308708191, "learning_rate": 6.768640871347163e-06, "loss": 0.2445, "step": 893 }, { "epoch": 3.840064620355412, "grad_norm": 1.1023160219192505, "learning_rate": 6.762306313701803e-06, "loss": 0.2397, "step": 894 }, { "epoch": 3.8443726440495425, "grad_norm": 1.0908433198928833, "learning_rate": 6.7559685245494025e-06, "loss": 0.2295, "step": 895 }, { "epoch": 3.8486806677436727, "grad_norm": 1.2091001272201538, "learning_rate": 6.749627515511443e-06, "loss": 0.2337, "step": 896 }, { "epoch": 3.852988691437803, "grad_norm": 1.0869569778442383, "learning_rate": 6.743283298215312e-06, "loss": 0.2404, "step": 897 }, { "epoch": 3.857296715131933, "grad_norm": 1.2445443868637085, "learning_rate": 6.736935884294275e-06, "loss": 0.2402, "step": 898 }, { "epoch": 3.8616047388260637, "grad_norm": 1.1659339666366577, "learning_rate": 6.730585285387465e-06, "loss": 0.2344, "step": 899 }, { "epoch": 3.865912762520194, "grad_norm": 1.1930010318756104, "learning_rate": 6.724231513139853e-06, "loss": 0.2344, "step": 900 }, { "epoch": 3.870220786214324, "grad_norm": 1.1425530910491943, "learning_rate": 6.717874579202227e-06, "loss": 0.2338, "step": 901 }, { "epoch": 3.8745288099084547, "grad_norm": 1.143779993057251, "learning_rate": 6.711514495231173e-06, "loss": 0.2344, "step": 902 }, { "epoch": 3.878836833602585, "grad_norm": 1.1523170471191406, "learning_rate": 6.705151272889055e-06, "loss": 0.2383, "step": 903 }, { "epoch": 3.883144857296715, "grad_norm": 1.0854690074920654, "learning_rate": 6.698784923843993e-06, "loss": 0.2273, "step": 904 }, { "epoch": 3.8874528809908453, "grad_norm": 1.058034896850586, "learning_rate": 6.692415459769835e-06, "loss": 0.2137, "step": 905 }, { "epoch": 3.891760904684976, "grad_norm": 1.1636855602264404, "learning_rate": 6.686042892346147e-06, "loss": 0.24, "step": 906 }, { "epoch": 3.896068928379106, "grad_norm": 1.1975581645965576, "learning_rate": 6.679667233258179e-06, "loss": 0.2395, "step": 907 }, { "epoch": 3.9003769520732363, "grad_norm": 1.0788935422897339, "learning_rate": 6.673288494196858e-06, "loss": 0.2206, "step": 908 }, { "epoch": 3.904684975767367, "grad_norm": 1.0104889869689941, "learning_rate": 6.666906686858753e-06, "loss": 0.2392, "step": 909 }, { "epoch": 3.908992999461497, "grad_norm": 1.2003514766693115, "learning_rate": 6.66052182294606e-06, "loss": 0.2377, "step": 910 }, { "epoch": 3.9133010231556273, "grad_norm": 1.0786452293395996, "learning_rate": 6.654133914166582e-06, "loss": 0.2339, "step": 911 }, { "epoch": 3.9176090468497575, "grad_norm": 1.4070574045181274, "learning_rate": 6.647742972233703e-06, "loss": 0.2439, "step": 912 }, { "epoch": 3.9219170705438877, "grad_norm": 1.221396565437317, "learning_rate": 6.641349008866369e-06, "loss": 0.2442, "step": 913 }, { "epoch": 3.9262250942380184, "grad_norm": 1.2268798351287842, "learning_rate": 6.634952035789069e-06, "loss": 0.2314, "step": 914 }, { "epoch": 3.9305331179321485, "grad_norm": 1.2313032150268555, "learning_rate": 6.628552064731807e-06, "loss": 0.2422, "step": 915 }, { "epoch": 3.934841141626279, "grad_norm": 1.0997172594070435, "learning_rate": 6.622149107430088e-06, "loss": 0.2292, "step": 916 }, { "epoch": 3.9391491653204094, "grad_norm": 1.1654084920883179, "learning_rate": 6.6157431756248906e-06, "loss": 0.2372, "step": 917 }, { "epoch": 3.9434571890145396, "grad_norm": 1.4347691535949707, "learning_rate": 6.609334281062647e-06, "loss": 0.2603, "step": 918 }, { "epoch": 3.9477652127086698, "grad_norm": 1.3016366958618164, "learning_rate": 6.602922435495225e-06, "loss": 0.2424, "step": 919 }, { "epoch": 3.9520732364028, "grad_norm": 1.158453345298767, "learning_rate": 6.5965076506799e-06, "loss": 0.2233, "step": 920 }, { "epoch": 3.9563812600969306, "grad_norm": 1.206938624382019, "learning_rate": 6.5900899383793415e-06, "loss": 0.2398, "step": 921 }, { "epoch": 3.9606892837910608, "grad_norm": 1.2216382026672363, "learning_rate": 6.583669310361583e-06, "loss": 0.2477, "step": 922 }, { "epoch": 3.9649973074851914, "grad_norm": 1.1788541078567505, "learning_rate": 6.577245778400006e-06, "loss": 0.2415, "step": 923 }, { "epoch": 3.9693053311793216, "grad_norm": 1.2634609937667847, "learning_rate": 6.570819354273317e-06, "loss": 0.2337, "step": 924 }, { "epoch": 3.973613354873452, "grad_norm": 1.0313800573349, "learning_rate": 6.564390049765528e-06, "loss": 0.2406, "step": 925 }, { "epoch": 3.977921378567582, "grad_norm": 1.1443864107131958, "learning_rate": 6.557957876665926e-06, "loss": 0.2399, "step": 926 }, { "epoch": 3.982229402261712, "grad_norm": 1.0647178888320923, "learning_rate": 6.551522846769067e-06, "loss": 0.235, "step": 927 }, { "epoch": 3.986537425955843, "grad_norm": 1.227658748626709, "learning_rate": 6.545084971874738e-06, "loss": 0.2444, "step": 928 }, { "epoch": 3.990845449649973, "grad_norm": 1.2170000076293945, "learning_rate": 6.538644263787948e-06, "loss": 0.244, "step": 929 }, { "epoch": 3.9951534733441036, "grad_norm": 1.095686435699463, "learning_rate": 6.532200734318896e-06, "loss": 0.2366, "step": 930 }, { "epoch": 3.999461497038234, "grad_norm": 1.046686053276062, "learning_rate": 6.525754395282961e-06, "loss": 0.2364, "step": 931 }, { "epoch": 4.0, "grad_norm": 1.046686053276062, "learning_rate": 6.5193052585006666e-06, "loss": 0.2523, "step": 932 } ], "logging_steps": 1, "max_steps": 2320, "num_input_tokens_seen": 0, "num_train_epochs": 10, "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": 10736180330496.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }