{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 8661, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0034638032559750607, "grad_norm": 0.0, "learning_rate": 1.1534025374855826e-06, "loss": 0.5168, "step": 10 }, { "epoch": 0.006927606511950121, "grad_norm": 0.0, "learning_rate": 2.3068050749711653e-06, "loss": 0.4816, "step": 20 }, { "epoch": 0.010391409767925183, "grad_norm": 0.0, "learning_rate": 3.4602076124567477e-06, "loss": 0.5447, "step": 30 }, { "epoch": 0.013855213023900243, "grad_norm": 0.0, "learning_rate": 4.6136101499423305e-06, "loss": 0.4825, "step": 40 }, { "epoch": 0.017319016279875303, "grad_norm": 0.0, "learning_rate": 5.7670126874279126e-06, "loss": 0.5454, "step": 50 }, { "epoch": 0.020782819535850365, "grad_norm": 0.0, "learning_rate": 6.920415224913495e-06, "loss": 0.486, "step": 60 }, { "epoch": 0.024246622791825424, "grad_norm": 0.0, "learning_rate": 8.073817762399077e-06, "loss": 0.4474, "step": 70 }, { "epoch": 0.027710426047800486, "grad_norm": 0.0, "learning_rate": 9.227220299884661e-06, "loss": 0.4528, "step": 80 }, { "epoch": 0.031174229303775544, "grad_norm": 0.0, "learning_rate": 1.0380622837370241e-05, "loss": 0.5972, "step": 90 }, { "epoch": 0.034638032559750606, "grad_norm": 0.0, "learning_rate": 1.1534025374855825e-05, "loss": 0.4725, "step": 100 }, { "epoch": 0.03810183581572567, "grad_norm": 0.0, "learning_rate": 1.2687427912341407e-05, "loss": 0.4798, "step": 110 }, { "epoch": 0.04156563907170073, "grad_norm": 0.0, "learning_rate": 1.384083044982699e-05, "loss": 0.5458, "step": 120 }, { "epoch": 0.045029442327675785, "grad_norm": 0.0, "learning_rate": 1.4994232987312573e-05, "loss": 0.5191, "step": 130 }, { "epoch": 0.04849324558365085, "grad_norm": 0.0, "learning_rate": 1.6147635524798155e-05, "loss": 0.5256, "step": 140 }, { "epoch": 0.05195704883962591, "grad_norm": 0.0, "learning_rate": 1.7301038062283735e-05, "loss": 0.6172, "step": 150 }, { "epoch": 0.05542085209560097, "grad_norm": 0.0, "learning_rate": 1.8454440599769322e-05, "loss": 0.4511, "step": 160 }, { "epoch": 0.05888465535157603, "grad_norm": 0.0, "learning_rate": 1.9607843137254903e-05, "loss": 0.5411, "step": 170 }, { "epoch": 0.06234845860755109, "grad_norm": 0.0, "learning_rate": 2.0761245674740483e-05, "loss": 0.55, "step": 180 }, { "epoch": 0.06581226186352615, "grad_norm": 0.0, "learning_rate": 2.191464821222607e-05, "loss": 0.4638, "step": 190 }, { "epoch": 0.06927606511950121, "grad_norm": 0.0, "learning_rate": 2.306805074971165e-05, "loss": 0.5789, "step": 200 }, { "epoch": 0.07273986837547627, "grad_norm": 0.0, "learning_rate": 2.422145328719723e-05, "loss": 0.454, "step": 210 }, { "epoch": 0.07620367163145134, "grad_norm": 0.0, "learning_rate": 2.5374855824682814e-05, "loss": 0.4422, "step": 220 }, { "epoch": 0.0796674748874264, "grad_norm": 0.0, "learning_rate": 2.6528258362168395e-05, "loss": 0.5477, "step": 230 }, { "epoch": 0.08313127814340146, "grad_norm": 0.0, "learning_rate": 2.768166089965398e-05, "loss": 0.5663, "step": 240 }, { "epoch": 0.08659508139937652, "grad_norm": 0.0, "learning_rate": 2.8835063437139565e-05, "loss": 0.6, "step": 250 }, { "epoch": 0.09005888465535157, "grad_norm": 0.0, "learning_rate": 2.9988465974625146e-05, "loss": 0.498, "step": 260 }, { "epoch": 0.09352268791132663, "grad_norm": 0.0, "learning_rate": 3.1141868512110726e-05, "loss": 0.4921, "step": 270 }, { "epoch": 0.0969864911673017, "grad_norm": 0.0, "learning_rate": 3.229527104959631e-05, "loss": 0.4737, "step": 280 }, { "epoch": 0.10045029442327676, "grad_norm": 0.0, "learning_rate": 3.344867358708189e-05, "loss": 0.6076, "step": 290 }, { "epoch": 0.10391409767925182, "grad_norm": 0.0, "learning_rate": 3.460207612456747e-05, "loss": 0.4479, "step": 300 }, { "epoch": 0.10737790093522688, "grad_norm": 0.0, "learning_rate": 3.575547866205306e-05, "loss": 0.5158, "step": 310 }, { "epoch": 0.11084170419120194, "grad_norm": 0.0, "learning_rate": 3.6908881199538644e-05, "loss": 0.6367, "step": 320 }, { "epoch": 0.114305507447177, "grad_norm": 0.0, "learning_rate": 3.806228373702422e-05, "loss": 0.5023, "step": 330 }, { "epoch": 0.11776931070315207, "grad_norm": 0.0, "learning_rate": 3.9215686274509805e-05, "loss": 0.5007, "step": 340 }, { "epoch": 0.12123311395912713, "grad_norm": 0.0, "learning_rate": 4.036908881199539e-05, "loss": 0.5567, "step": 350 }, { "epoch": 0.12469691721510218, "grad_norm": 0.0, "learning_rate": 4.1522491349480966e-05, "loss": 0.5175, "step": 360 }, { "epoch": 0.12816072047107724, "grad_norm": 0.0, "learning_rate": 4.2675893886966556e-05, "loss": 0.4519, "step": 370 }, { "epoch": 0.1316245237270523, "grad_norm": 0.0, "learning_rate": 4.382929642445214e-05, "loss": 0.7048, "step": 380 }, { "epoch": 0.13508832698302736, "grad_norm": 0.0, "learning_rate": 4.498269896193772e-05, "loss": 0.6115, "step": 390 }, { "epoch": 0.13855213023900242, "grad_norm": 0.0, "learning_rate": 4.61361014994233e-05, "loss": 0.5099, "step": 400 }, { "epoch": 0.1420159334949775, "grad_norm": 0.0, "learning_rate": 4.7289504036908884e-05, "loss": 0.4645, "step": 410 }, { "epoch": 0.14547973675095255, "grad_norm": 0.0, "learning_rate": 4.844290657439446e-05, "loss": 0.4844, "step": 420 }, { "epoch": 0.1489435400069276, "grad_norm": 0.0, "learning_rate": 4.9596309111880045e-05, "loss": 0.5029, "step": 430 }, { "epoch": 0.15240734326290267, "grad_norm": 0.0, "learning_rate": 5.074971164936563e-05, "loss": 0.41, "step": 440 }, { "epoch": 0.15587114651887773, "grad_norm": 0.0, "learning_rate": 5.190311418685121e-05, "loss": 0.4924, "step": 450 }, { "epoch": 0.1593349497748528, "grad_norm": 0.0, "learning_rate": 5.305651672433679e-05, "loss": 0.5781, "step": 460 }, { "epoch": 0.16279875303082786, "grad_norm": 0.0, "learning_rate": 5.4209919261822386e-05, "loss": 0.5311, "step": 470 }, { "epoch": 0.16626255628680292, "grad_norm": 0.0, "learning_rate": 5.536332179930796e-05, "loss": 0.5368, "step": 480 }, { "epoch": 0.16972635954277798, "grad_norm": 0.0, "learning_rate": 5.651672433679355e-05, "loss": 0.5397, "step": 490 }, { "epoch": 0.17319016279875304, "grad_norm": 0.0, "learning_rate": 5.767012687427913e-05, "loss": 0.5694, "step": 500 }, { "epoch": 0.17665396605472808, "grad_norm": 0.0, "learning_rate": 5.882352941176471e-05, "loss": 0.5732, "step": 510 }, { "epoch": 0.18011776931070314, "grad_norm": 0.0, "learning_rate": 5.997693194925029e-05, "loss": 0.4881, "step": 520 }, { "epoch": 0.1835815725666782, "grad_norm": 0.0, "learning_rate": 6.113033448673587e-05, "loss": 0.5079, "step": 530 }, { "epoch": 0.18704537582265326, "grad_norm": 0.0, "learning_rate": 6.228373702422145e-05, "loss": 0.4964, "step": 540 }, { "epoch": 0.19050917907862833, "grad_norm": 0.0, "learning_rate": 6.343713956170704e-05, "loss": 0.4942, "step": 550 }, { "epoch": 0.1939729823346034, "grad_norm": 0.0, "learning_rate": 6.459054209919262e-05, "loss": 0.4802, "step": 560 }, { "epoch": 0.19743678559057845, "grad_norm": 0.0, "learning_rate": 6.57439446366782e-05, "loss": 0.5472, "step": 570 }, { "epoch": 0.2009005888465535, "grad_norm": 0.0, "learning_rate": 6.689734717416379e-05, "loss": 0.5019, "step": 580 }, { "epoch": 0.20436439210252857, "grad_norm": 0.0, "learning_rate": 6.805074971164937e-05, "loss": 0.5509, "step": 590 }, { "epoch": 0.20782819535850364, "grad_norm": 0.0, "learning_rate": 6.920415224913494e-05, "loss": 0.4571, "step": 600 }, { "epoch": 0.2112919986144787, "grad_norm": 0.0, "learning_rate": 7.035755478662054e-05, "loss": 0.5818, "step": 610 }, { "epoch": 0.21475580187045376, "grad_norm": 0.0, "learning_rate": 7.151095732410612e-05, "loss": 0.5283, "step": 620 }, { "epoch": 0.21821960512642882, "grad_norm": 0.0, "learning_rate": 7.26643598615917e-05, "loss": 0.5282, "step": 630 }, { "epoch": 0.22168340838240388, "grad_norm": 0.0, "learning_rate": 7.381776239907729e-05, "loss": 0.5128, "step": 640 }, { "epoch": 0.22514721163837895, "grad_norm": 0.0, "learning_rate": 7.497116493656286e-05, "loss": 0.5452, "step": 650 }, { "epoch": 0.228611014894354, "grad_norm": 0.0, "learning_rate": 7.612456747404844e-05, "loss": 0.6405, "step": 660 }, { "epoch": 0.23207481815032907, "grad_norm": 0.0, "learning_rate": 7.727797001153403e-05, "loss": 0.5584, "step": 670 }, { "epoch": 0.23553862140630413, "grad_norm": 0.0, "learning_rate": 7.843137254901961e-05, "loss": 0.4541, "step": 680 }, { "epoch": 0.2390024246622792, "grad_norm": 0.0, "learning_rate": 7.95847750865052e-05, "loss": 0.5287, "step": 690 }, { "epoch": 0.24246622791825426, "grad_norm": 0.0, "learning_rate": 8.073817762399078e-05, "loss": 0.6159, "step": 700 }, { "epoch": 0.2459300311742293, "grad_norm": 0.0, "learning_rate": 8.189158016147636e-05, "loss": 0.594, "step": 710 }, { "epoch": 0.24939383443020435, "grad_norm": 0.0, "learning_rate": 8.304498269896193e-05, "loss": 0.5769, "step": 720 }, { "epoch": 0.2528576376861794, "grad_norm": 0.0, "learning_rate": 8.419838523644751e-05, "loss": 0.6014, "step": 730 }, { "epoch": 0.2563214409421545, "grad_norm": 0.0, "learning_rate": 8.535178777393311e-05, "loss": 0.451, "step": 740 }, { "epoch": 0.25978524419812954, "grad_norm": 0.0, "learning_rate": 8.65051903114187e-05, "loss": 0.633, "step": 750 }, { "epoch": 0.2632490474541046, "grad_norm": 0.0, "learning_rate": 8.765859284890428e-05, "loss": 0.6181, "step": 760 }, { "epoch": 0.26671285071007966, "grad_norm": 0.0, "learning_rate": 8.881199538638986e-05, "loss": 0.652, "step": 770 }, { "epoch": 0.2701766539660547, "grad_norm": 0.0, "learning_rate": 8.996539792387543e-05, "loss": 0.5852, "step": 780 }, { "epoch": 0.2736404572220298, "grad_norm": 0.0, "learning_rate": 9.111880046136102e-05, "loss": 0.6068, "step": 790 }, { "epoch": 0.27710426047800485, "grad_norm": 0.0, "learning_rate": 9.22722029988466e-05, "loss": 0.5957, "step": 800 }, { "epoch": 0.2805680637339799, "grad_norm": 0.0, "learning_rate": 9.342560553633218e-05, "loss": 0.5381, "step": 810 }, { "epoch": 0.284031866989955, "grad_norm": 0.0, "learning_rate": 9.457900807381777e-05, "loss": 0.6142, "step": 820 }, { "epoch": 0.28749567024593004, "grad_norm": 0.0, "learning_rate": 9.573241061130335e-05, "loss": 0.4412, "step": 830 }, { "epoch": 0.2909594735019051, "grad_norm": 0.0, "learning_rate": 9.688581314878892e-05, "loss": 0.4881, "step": 840 }, { "epoch": 0.29442327675788016, "grad_norm": 0.0, "learning_rate": 9.80392156862745e-05, "loss": 0.6448, "step": 850 }, { "epoch": 0.2978870800138552, "grad_norm": 0.0, "learning_rate": 9.919261822376009e-05, "loss": 0.5194, "step": 860 }, { "epoch": 0.3013508832698303, "grad_norm": 0.0, "learning_rate": 9.999996344376946e-05, "loss": 0.47, "step": 870 }, { "epoch": 0.30481468652580535, "grad_norm": 0.0, "learning_rate": 9.999931355671335e-05, "loss": 0.4842, "step": 880 }, { "epoch": 0.3082784897817804, "grad_norm": 0.0, "learning_rate": 9.999785132113196e-05, "loss": 0.6713, "step": 890 }, { "epoch": 0.31174229303775547, "grad_norm": 0.0, "learning_rate": 9.999557676078253e-05, "loss": 0.5278, "step": 900 }, { "epoch": 0.31520609629373053, "grad_norm": 0.0, "learning_rate": 9.999248991262028e-05, "loss": 0.5635, "step": 910 }, { "epoch": 0.3186698995497056, "grad_norm": 0.0, "learning_rate": 9.998859082679782e-05, "loss": 0.5359, "step": 920 }, { "epoch": 0.32213370280568066, "grad_norm": 0.0, "learning_rate": 9.998387956666437e-05, "loss": 0.5917, "step": 930 }, { "epoch": 0.3255975060616557, "grad_norm": 0.0, "learning_rate": 9.997835620876466e-05, "loss": 0.61, "step": 940 }, { "epoch": 0.3290613093176308, "grad_norm": 0.0, "learning_rate": 9.997202084283778e-05, "loss": 0.5755, "step": 950 }, { "epoch": 0.33252511257360584, "grad_norm": 0.0, "learning_rate": 9.996487357181561e-05, "loss": 0.6457, "step": 960 }, { "epoch": 0.3359889158295809, "grad_norm": 0.0, "learning_rate": 9.99569145118213e-05, "loss": 0.5239, "step": 970 }, { "epoch": 0.33945271908555597, "grad_norm": 0.0, "learning_rate": 9.994814379216719e-05, "loss": 0.5372, "step": 980 }, { "epoch": 0.342916522341531, "grad_norm": 0.0, "learning_rate": 9.99385615553529e-05, "loss": 0.6292, "step": 990 }, { "epoch": 0.3463803255975061, "grad_norm": 0.0, "learning_rate": 9.992816795706285e-05, "loss": 0.6101, "step": 1000 }, { "epoch": 0.3498441288534811, "grad_norm": 0.0, "learning_rate": 9.991696316616386e-05, "loss": 0.5718, "step": 1010 }, { "epoch": 0.35330793210945616, "grad_norm": 0.0, "learning_rate": 9.990494736470232e-05, "loss": 0.569, "step": 1020 }, { "epoch": 0.3567717353654312, "grad_norm": 0.0, "learning_rate": 9.989212074790131e-05, "loss": 0.5238, "step": 1030 }, { "epoch": 0.3602355386214063, "grad_norm": 0.0, "learning_rate": 9.987848352415735e-05, "loss": 0.5855, "step": 1040 }, { "epoch": 0.36369934187738134, "grad_norm": 0.0, "learning_rate": 9.986403591503704e-05, "loss": 0.5268, "step": 1050 }, { "epoch": 0.3671631451333564, "grad_norm": 0.0, "learning_rate": 9.984877815527349e-05, "loss": 0.6226, "step": 1060 }, { "epoch": 0.37062694838933147, "grad_norm": 0.0, "learning_rate": 9.983271049276247e-05, "loss": 0.5789, "step": 1070 }, { "epoch": 0.37409075164530653, "grad_norm": 0.0, "learning_rate": 9.981583318855841e-05, "loss": 0.5208, "step": 1080 }, { "epoch": 0.3775545549012816, "grad_norm": 0.0, "learning_rate": 9.979814651687014e-05, "loss": 0.5148, "step": 1090 }, { "epoch": 0.38101835815725665, "grad_norm": 0.0, "learning_rate": 9.977965076505642e-05, "loss": 0.5665, "step": 1100 }, { "epoch": 0.3844821614132317, "grad_norm": 0.0, "learning_rate": 9.976034623362134e-05, "loss": 0.4862, "step": 1110 }, { "epoch": 0.3879459646692068, "grad_norm": 0.0, "learning_rate": 9.974023323620933e-05, "loss": 0.5117, "step": 1120 }, { "epoch": 0.39140976792518184, "grad_norm": 0.0, "learning_rate": 9.971931209960018e-05, "loss": 0.6368, "step": 1130 }, { "epoch": 0.3948735711811569, "grad_norm": 0.0, "learning_rate": 9.969758316370361e-05, "loss": 0.4832, "step": 1140 }, { "epoch": 0.39833737443713196, "grad_norm": 0.0, "learning_rate": 9.967504678155389e-05, "loss": 0.5465, "step": 1150 }, { "epoch": 0.401801177693107, "grad_norm": 0.0, "learning_rate": 9.965170331930394e-05, "loss": 0.5252, "step": 1160 }, { "epoch": 0.4052649809490821, "grad_norm": 0.0, "learning_rate": 9.962755315621955e-05, "loss": 0.4856, "step": 1170 }, { "epoch": 0.40872878420505715, "grad_norm": 0.0, "learning_rate": 9.960259668467307e-05, "loss": 0.4933, "step": 1180 }, { "epoch": 0.4121925874610322, "grad_norm": 0.0, "learning_rate": 9.957683431013711e-05, "loss": 0.5076, "step": 1190 }, { "epoch": 0.4156563907170073, "grad_norm": 0.0, "learning_rate": 9.9550266451178e-05, "loss": 0.5796, "step": 1200 }, { "epoch": 0.41912019397298234, "grad_norm": 0.0, "learning_rate": 9.952289353944884e-05, "loss": 0.545, "step": 1210 }, { "epoch": 0.4225839972289574, "grad_norm": 0.0, "learning_rate": 9.949471601968265e-05, "loss": 0.6556, "step": 1220 }, { "epoch": 0.42604780048493246, "grad_norm": 0.0, "learning_rate": 9.946573434968505e-05, "loss": 0.559, "step": 1230 }, { "epoch": 0.4295116037409075, "grad_norm": 0.0, "learning_rate": 9.943594900032685e-05, "loss": 0.5644, "step": 1240 }, { "epoch": 0.4329754069968826, "grad_norm": 0.0, "learning_rate": 9.940536045553635e-05, "loss": 0.5346, "step": 1250 }, { "epoch": 0.43643921025285765, "grad_norm": 0.0, "learning_rate": 9.937396921229159e-05, "loss": 0.5685, "step": 1260 }, { "epoch": 0.4399030135088327, "grad_norm": 0.0, "learning_rate": 9.934177578061216e-05, "loss": 0.5408, "step": 1270 }, { "epoch": 0.44336681676480777, "grad_norm": 0.0, "learning_rate": 9.930878068355099e-05, "loss": 0.5042, "step": 1280 }, { "epoch": 0.44683062002078283, "grad_norm": 0.0, "learning_rate": 9.92749844571858e-05, "loss": 0.6322, "step": 1290 }, { "epoch": 0.4502944232767579, "grad_norm": 0.0, "learning_rate": 9.924038765061042e-05, "loss": 0.5287, "step": 1300 }, { "epoch": 0.45375822653273296, "grad_norm": 0.0, "learning_rate": 9.920499082592584e-05, "loss": 0.5062, "step": 1310 }, { "epoch": 0.457222029788708, "grad_norm": 0.0, "learning_rate": 9.916879455823118e-05, "loss": 0.5699, "step": 1320 }, { "epoch": 0.4606858330446831, "grad_norm": 0.0, "learning_rate": 9.913179943561415e-05, "loss": 0.5054, "step": 1330 }, { "epoch": 0.46414963630065814, "grad_norm": 0.0, "learning_rate": 9.909400605914172e-05, "loss": 0.5799, "step": 1340 }, { "epoch": 0.4676134395566332, "grad_norm": 0.0, "learning_rate": 9.905541504285014e-05, "loss": 0.6171, "step": 1350 }, { "epoch": 0.47107724281260827, "grad_norm": 0.0, "learning_rate": 9.901602701373516e-05, "loss": 0.4705, "step": 1360 }, { "epoch": 0.4745410460685833, "grad_norm": 0.0, "learning_rate": 9.897584261174169e-05, "loss": 0.6618, "step": 1370 }, { "epoch": 0.4780048493245584, "grad_norm": 0.0, "learning_rate": 9.89348624897535e-05, "loss": 0.4796, "step": 1380 }, { "epoch": 0.48146865258053345, "grad_norm": 0.0, "learning_rate": 9.889308731358256e-05, "loss": 0.6067, "step": 1390 }, { "epoch": 0.4849324558365085, "grad_norm": 0.0, "learning_rate": 9.885051776195824e-05, "loss": 0.5311, "step": 1400 }, { "epoch": 0.4883962590924835, "grad_norm": 0.0, "learning_rate": 9.88071545265163e-05, "loss": 0.5435, "step": 1410 }, { "epoch": 0.4918600623484586, "grad_norm": 0.0, "learning_rate": 9.876299831178761e-05, "loss": 0.5444, "step": 1420 }, { "epoch": 0.49532386560443364, "grad_norm": 0.0, "learning_rate": 9.871804983518672e-05, "loss": 0.5188, "step": 1430 }, { "epoch": 0.4987876688604087, "grad_norm": 0.0, "learning_rate": 9.867230982700026e-05, "loss": 0.5816, "step": 1440 }, { "epoch": 0.5022514721163838, "grad_norm": 0.0, "learning_rate": 9.862577903037497e-05, "loss": 0.483, "step": 1450 }, { "epoch": 0.5057152753723588, "grad_norm": 0.0, "learning_rate": 9.857845820130573e-05, "loss": 0.6083, "step": 1460 }, { "epoch": 0.5091790786283339, "grad_norm": 0.0, "learning_rate": 9.85303481086232e-05, "loss": 0.6111, "step": 1470 }, { "epoch": 0.512642881884309, "grad_norm": 0.0, "learning_rate": 9.848144953398138e-05, "loss": 0.6105, "step": 1480 }, { "epoch": 0.516106685140284, "grad_norm": 0.0, "learning_rate": 9.843176327184485e-05, "loss": 0.5168, "step": 1490 }, { "epoch": 0.5195704883962591, "grad_norm": 0.0, "learning_rate": 9.838129012947595e-05, "loss": 0.5795, "step": 1500 }, { "epoch": 0.5230342916522341, "grad_norm": 0.0, "learning_rate": 9.833003092692158e-05, "loss": 0.5429, "step": 1510 }, { "epoch": 0.5264980949082092, "grad_norm": 0.0, "learning_rate": 9.827798649699993e-05, "loss": 0.5389, "step": 1520 }, { "epoch": 0.5299618981641843, "grad_norm": 0.0, "learning_rate": 9.822515768528695e-05, "loss": 0.5002, "step": 1530 }, { "epoch": 0.5334257014201593, "grad_norm": 0.0, "learning_rate": 9.817154535010254e-05, "loss": 0.5799, "step": 1540 }, { "epoch": 0.5368895046761344, "grad_norm": 0.0, "learning_rate": 9.811715036249669e-05, "loss": 0.6281, "step": 1550 }, { "epoch": 0.5403533079321095, "grad_norm": 0.0, "learning_rate": 9.806197360623531e-05, "loss": 0.554, "step": 1560 }, { "epoch": 0.5438171111880845, "grad_norm": 0.0, "learning_rate": 9.800601597778582e-05, "loss": 0.5528, "step": 1570 }, { "epoch": 0.5472809144440596, "grad_norm": 0.0, "learning_rate": 9.794927838630265e-05, "loss": 0.4758, "step": 1580 }, { "epoch": 0.5507447177000346, "grad_norm": 0.0, "learning_rate": 9.789176175361245e-05, "loss": 0.5403, "step": 1590 }, { "epoch": 0.5542085209560097, "grad_norm": 0.0, "learning_rate": 9.783346701419904e-05, "loss": 0.5113, "step": 1600 }, { "epoch": 0.5576723242119848, "grad_norm": 0.0, "learning_rate": 9.777439511518837e-05, "loss": 0.6052, "step": 1610 }, { "epoch": 0.5611361274679598, "grad_norm": 0.0, "learning_rate": 9.771454701633301e-05, "loss": 0.4944, "step": 1620 }, { "epoch": 0.5645999307239349, "grad_norm": 0.0, "learning_rate": 9.76539236899966e-05, "loss": 0.5814, "step": 1630 }, { "epoch": 0.56806373397991, "grad_norm": 0.0, "learning_rate": 9.759252612113805e-05, "loss": 1.2429, "step": 1640 }, { "epoch": 0.571527537235885, "grad_norm": 0.0, "learning_rate": 9.753035530729556e-05, "loss": 0.5693, "step": 1650 }, { "epoch": 0.5749913404918601, "grad_norm": 0.0, "learning_rate": 9.746741225857037e-05, "loss": 0.5122, "step": 1660 }, { "epoch": 0.5784551437478351, "grad_norm": 0.0, "learning_rate": 9.740369799761038e-05, "loss": 0.5647, "step": 1670 }, { "epoch": 0.5819189470038102, "grad_norm": 0.0, "learning_rate": 9.733921355959352e-05, "loss": 0.4781, "step": 1680 }, { "epoch": 0.5853827502597853, "grad_norm": 0.0, "learning_rate": 9.727395999221094e-05, "loss": 0.6351, "step": 1690 }, { "epoch": 0.5888465535157603, "grad_norm": 0.0, "learning_rate": 9.720793835564997e-05, "loss": 0.4266, "step": 1700 }, { "epoch": 0.5923103567717354, "grad_norm": 0.0, "learning_rate": 9.71411497225769e-05, "loss": 0.4946, "step": 1710 }, { "epoch": 0.5957741600277104, "grad_norm": 0.0, "learning_rate": 9.707359517811964e-05, "loss": 0.6085, "step": 1720 }, { "epoch": 0.5992379632836855, "grad_norm": 0.0, "learning_rate": 9.700527581984988e-05, "loss": 0.5944, "step": 1730 }, { "epoch": 0.6027017665396606, "grad_norm": 0.0, "learning_rate": 9.693619275776551e-05, "loss": 0.5241, "step": 1740 }, { "epoch": 0.6061655697956356, "grad_norm": 0.0, "learning_rate": 9.686634711427239e-05, "loss": 0.4669, "step": 1750 }, { "epoch": 0.6096293730516107, "grad_norm": 0.0, "learning_rate": 9.67957400241662e-05, "loss": 0.5928, "step": 1760 }, { "epoch": 0.6130931763075858, "grad_norm": 0.0, "learning_rate": 9.6724372634614e-05, "loss": 0.5777, "step": 1770 }, { "epoch": 0.6165569795635608, "grad_norm": 0.0, "learning_rate": 9.665224610513562e-05, "loss": 0.4716, "step": 1780 }, { "epoch": 0.6200207828195359, "grad_norm": 0.0, "learning_rate": 9.65793616075847e-05, "loss": 0.5688, "step": 1790 }, { "epoch": 0.6234845860755109, "grad_norm": 0.0, "learning_rate": 9.650572032612978e-05, "loss": 0.5226, "step": 1800 }, { "epoch": 0.626948389331486, "grad_norm": 0.0, "learning_rate": 9.643132345723506e-05, "loss": 0.4624, "step": 1810 }, { "epoch": 0.6304121925874611, "grad_norm": 0.0, "learning_rate": 9.635617220964082e-05, "loss": 0.609, "step": 1820 }, { "epoch": 0.6338759958434361, "grad_norm": 0.0, "learning_rate": 9.628026780434396e-05, "loss": 0.5873, "step": 1830 }, { "epoch": 0.6373397990994112, "grad_norm": 0.0, "learning_rate": 9.620361147457802e-05, "loss": 0.6371, "step": 1840 }, { "epoch": 0.6408036023553862, "grad_norm": 0.0, "learning_rate": 9.612620446579324e-05, "loss": 0.8071, "step": 1850 }, { "epoch": 0.6442674056113613, "grad_norm": 0.0, "learning_rate": 9.604804803563627e-05, "loss": 0.5144, "step": 1860 }, { "epoch": 0.6477312088673364, "grad_norm": 0.0, "learning_rate": 9.596914345392975e-05, "loss": 0.516, "step": 1870 }, { "epoch": 0.6511950121233114, "grad_norm": 0.0, "learning_rate": 9.588949200265173e-05, "loss": 0.5052, "step": 1880 }, { "epoch": 0.6546588153792865, "grad_norm": 0.0, "learning_rate": 9.580909497591471e-05, "loss": 0.5024, "step": 1890 }, { "epoch": 0.6581226186352616, "grad_norm": 0.0, "learning_rate": 9.572795367994478e-05, "loss": 0.5621, "step": 1900 }, { "epoch": 0.6615864218912366, "grad_norm": 0.0, "learning_rate": 9.564606943306025e-05, "loss": 0.5361, "step": 1910 }, { "epoch": 0.6650502251472117, "grad_norm": 0.0, "learning_rate": 9.556344356565035e-05, "loss": 0.5122, "step": 1920 }, { "epoch": 0.6685140284031867, "grad_norm": 0.0, "learning_rate": 9.548007742015354e-05, "loss": 0.5998, "step": 1930 }, { "epoch": 0.6719778316591618, "grad_norm": 0.0, "learning_rate": 9.539597235103571e-05, "loss": 0.5333, "step": 1940 }, { "epoch": 0.6754416349151369, "grad_norm": 0.0, "learning_rate": 9.531112972476821e-05, "loss": 0.7365, "step": 1950 }, { "epoch": 0.6789054381711119, "grad_norm": 0.0, "learning_rate": 9.522555091980561e-05, "loss": 0.5839, "step": 1960 }, { "epoch": 0.682369241427087, "grad_norm": 0.0, "learning_rate": 9.513923732656331e-05, "loss": 0.5188, "step": 1970 }, { "epoch": 0.685833044683062, "grad_norm": 0.0, "learning_rate": 9.505219034739496e-05, "loss": 0.5624, "step": 1980 }, { "epoch": 0.6892968479390371, "grad_norm": 0.0, "learning_rate": 9.49644113965697e-05, "loss": 0.6018, "step": 1990 }, { "epoch": 0.6927606511950122, "grad_norm": 0.0, "learning_rate": 9.487590190024915e-05, "loss": 0.6334, "step": 2000 }, { "epoch": 0.6962244544509871, "grad_norm": 0.0, "learning_rate": 9.478666329646418e-05, "loss": 0.473, "step": 2010 }, { "epoch": 0.6996882577069622, "grad_norm": 0.0, "learning_rate": 9.469669703509169e-05, "loss": 0.5676, "step": 2020 }, { "epoch": 0.7031520609629373, "grad_norm": 0.0, "learning_rate": 9.460600457783093e-05, "loss": 0.5937, "step": 2030 }, { "epoch": 0.7066158642189123, "grad_norm": 0.0, "learning_rate": 9.451458739817982e-05, "loss": 0.5505, "step": 2040 }, { "epoch": 0.7100796674748874, "grad_norm": 0.0, "learning_rate": 9.442244698141096e-05, "loss": 0.5036, "step": 2050 }, { "epoch": 0.7135434707308624, "grad_norm": 0.0, "learning_rate": 9.432958482454755e-05, "loss": 0.5097, "step": 2060 }, { "epoch": 0.7170072739868375, "grad_norm": 0.0, "learning_rate": 9.423600243633901e-05, "loss": 0.5082, "step": 2070 }, { "epoch": 0.7204710772428126, "grad_norm": 0.0, "learning_rate": 9.414170133723651e-05, "loss": 0.4633, "step": 2080 }, { "epoch": 0.7239348804987876, "grad_norm": 0.0, "learning_rate": 9.40466830593683e-05, "loss": 0.5377, "step": 2090 }, { "epoch": 0.7273986837547627, "grad_norm": 0.0, "learning_rate": 9.395094914651471e-05, "loss": 0.5379, "step": 2100 }, { "epoch": 0.7308624870107377, "grad_norm": 0.0, "learning_rate": 9.385450115408314e-05, "loss": 0.5793, "step": 2110 }, { "epoch": 0.7343262902667128, "grad_norm": 0.0, "learning_rate": 9.37573406490828e-05, "loss": 0.6478, "step": 2120 }, { "epoch": 0.7377900935226879, "grad_norm": 0.0, "learning_rate": 9.365946921009922e-05, "loss": 0.6453, "step": 2130 }, { "epoch": 0.7412538967786629, "grad_norm": 0.0, "learning_rate": 9.356088842726861e-05, "loss": 0.5522, "step": 2140 }, { "epoch": 0.744717700034638, "grad_norm": 0.0, "learning_rate": 9.346159990225205e-05, "loss": 0.5502, "step": 2150 }, { "epoch": 0.7481815032906131, "grad_norm": 0.0, "learning_rate": 9.33616052482094e-05, "loss": 0.4901, "step": 2160 }, { "epoch": 0.7516453065465881, "grad_norm": 0.0, "learning_rate": 9.326090608977317e-05, "loss": 0.6472, "step": 2170 }, { "epoch": 0.7551091098025632, "grad_norm": 0.0, "learning_rate": 9.315950406302207e-05, "loss": 0.538, "step": 2180 }, { "epoch": 0.7585729130585382, "grad_norm": 0.0, "learning_rate": 9.305740081545448e-05, "loss": 0.5139, "step": 2190 }, { "epoch": 0.7620367163145133, "grad_norm": 0.0, "learning_rate": 9.295459800596161e-05, "loss": 0.7008, "step": 2200 }, { "epoch": 0.7655005195704884, "grad_norm": 0.0, "learning_rate": 9.285109730480058e-05, "loss": 0.5551, "step": 2210 }, { "epoch": 0.7689643228264634, "grad_norm": 0.0, "learning_rate": 9.274690039356736e-05, "loss": 0.5626, "step": 2220 }, { "epoch": 0.7724281260824385, "grad_norm": 0.0, "learning_rate": 9.264200896516932e-05, "loss": 0.6539, "step": 2230 }, { "epoch": 0.7758919293384136, "grad_norm": 0.0, "learning_rate": 9.253642472379778e-05, "loss": 0.457, "step": 2240 }, { "epoch": 0.7793557325943886, "grad_norm": 0.0, "learning_rate": 9.243014938490037e-05, "loss": 0.515, "step": 2250 }, { "epoch": 0.7828195358503637, "grad_norm": 0.0, "learning_rate": 9.232318467515309e-05, "loss": 0.5014, "step": 2260 }, { "epoch": 0.7862833391063387, "grad_norm": 0.0, "learning_rate": 9.221553233243226e-05, "loss": 0.6387, "step": 2270 }, { "epoch": 0.7897471423623138, "grad_norm": 0.0, "learning_rate": 9.210719410578634e-05, "loss": 0.5318, "step": 2280 }, { "epoch": 0.7932109456182889, "grad_norm": 0.0, "learning_rate": 9.19981717554075e-05, "loss": 0.5713, "step": 2290 }, { "epoch": 0.7966747488742639, "grad_norm": 0.0, "learning_rate": 9.188846705260293e-05, "loss": 0.5117, "step": 2300 }, { "epoch": 0.800138552130239, "grad_norm": 0.0, "learning_rate": 9.177808177976619e-05, "loss": 0.5321, "step": 2310 }, { "epoch": 0.803602355386214, "grad_norm": 0.0, "learning_rate": 9.166701773034818e-05, "loss": 0.5077, "step": 2320 }, { "epoch": 0.8070661586421891, "grad_norm": 0.0, "learning_rate": 9.155527670882797e-05, "loss": 0.6253, "step": 2330 }, { "epoch": 0.8105299618981642, "grad_norm": 0.0, "learning_rate": 9.144286053068361e-05, "loss": 0.5674, "step": 2340 }, { "epoch": 0.8139937651541392, "grad_norm": 0.0, "learning_rate": 9.132977102236248e-05, "loss": 0.5612, "step": 2350 }, { "epoch": 0.8174575684101143, "grad_norm": 0.0, "learning_rate": 9.121601002125173e-05, "loss": 0.5145, "step": 2360 }, { "epoch": 0.8209213716660894, "grad_norm": 0.0, "learning_rate": 9.110157937564836e-05, "loss": 0.6443, "step": 2370 }, { "epoch": 0.8243851749220644, "grad_norm": 0.0, "learning_rate": 9.098648094472922e-05, "loss": 0.5034, "step": 2380 }, { "epoch": 0.8278489781780395, "grad_norm": 0.0, "learning_rate": 9.087071659852076e-05, "loss": 0.6314, "step": 2390 }, { "epoch": 0.8313127814340145, "grad_norm": 0.0, "learning_rate": 9.075428821786875e-05, "loss": 0.6946, "step": 2400 }, { "epoch": 0.8347765846899896, "grad_norm": 0.0, "learning_rate": 9.063719769440766e-05, "loss": 0.5403, "step": 2410 }, { "epoch": 0.8382403879459647, "grad_norm": 0.0, "learning_rate": 9.051944693052982e-05, "loss": 0.6203, "step": 2420 }, { "epoch": 0.8417041912019397, "grad_norm": 0.0, "learning_rate": 9.040103783935474e-05, "loss": 0.5899, "step": 2430 }, { "epoch": 0.8451679944579148, "grad_norm": 0.0, "learning_rate": 9.028197234469781e-05, "loss": 0.5086, "step": 2440 }, { "epoch": 0.8486317977138899, "grad_norm": 0.0, "learning_rate": 9.016225238103917e-05, "loss": 0.544, "step": 2450 }, { "epoch": 0.8520956009698649, "grad_norm": 0.0, "learning_rate": 9.004187989349227e-05, "loss": 0.7209, "step": 2460 }, { "epoch": 0.85555940422584, "grad_norm": 0.0, "learning_rate": 8.992085683777218e-05, "loss": 0.5105, "step": 2470 }, { "epoch": 0.859023207481815, "grad_norm": 0.0, "learning_rate": 8.979918518016395e-05, "loss": 0.4995, "step": 2480 }, { "epoch": 0.8624870107377901, "grad_norm": 0.0, "learning_rate": 8.96768668974905e-05, "loss": 0.5792, "step": 2490 }, { "epoch": 0.8659508139937652, "grad_norm": 0.0, "learning_rate": 8.95539039770807e-05, "loss": 0.6432, "step": 2500 }, { "epoch": 0.8694146172497402, "grad_norm": 0.0, "learning_rate": 8.943029841673688e-05, "loss": 0.5705, "step": 2510 }, { "epoch": 0.8728784205057153, "grad_norm": 0.0, "learning_rate": 8.930605222470252e-05, "loss": 0.4919, "step": 2520 }, { "epoch": 0.8763422237616904, "grad_norm": 0.0, "learning_rate": 8.918116741962955e-05, "loss": 0.4886, "step": 2530 }, { "epoch": 0.8798060270176654, "grad_norm": 0.0, "learning_rate": 8.905564603054557e-05, "loss": 0.5825, "step": 2540 }, { "epoch": 0.8832698302736405, "grad_norm": 0.0, "learning_rate": 8.892949009682088e-05, "loss": 0.549, "step": 2550 }, { "epoch": 0.8867336335296155, "grad_norm": 0.0, "learning_rate": 8.880270166813536e-05, "loss": 0.5223, "step": 2560 }, { "epoch": 0.8901974367855906, "grad_norm": 0.0, "learning_rate": 8.867528280444515e-05, "loss": 0.5756, "step": 2570 }, { "epoch": 0.8936612400415657, "grad_norm": 0.0, "learning_rate": 8.85472355759492e-05, "loss": 0.4789, "step": 2580 }, { "epoch": 0.8971250432975407, "grad_norm": 0.0, "learning_rate": 8.841856206305558e-05, "loss": 0.452, "step": 2590 }, { "epoch": 0.9005888465535158, "grad_norm": 0.0, "learning_rate": 8.828926435634783e-05, "loss": 0.557, "step": 2600 }, { "epoch": 0.9040526498094908, "grad_norm": 0.0, "learning_rate": 8.815934455655075e-05, "loss": 0.4234, "step": 2610 }, { "epoch": 0.9075164530654659, "grad_norm": 0.0, "learning_rate": 8.80288047744965e-05, "loss": 0.6225, "step": 2620 }, { "epoch": 0.910980256321441, "grad_norm": 0.0, "learning_rate": 8.78976471310902e-05, "loss": 0.5944, "step": 2630 }, { "epoch": 0.914444059577416, "grad_norm": 0.0, "learning_rate": 8.776587375727545e-05, "loss": 0.6333, "step": 2640 }, { "epoch": 0.9179078628333911, "grad_norm": 0.0, "learning_rate": 8.763348679399977e-05, "loss": 0.5074, "step": 2650 }, { "epoch": 0.9213716660893662, "grad_norm": 0.0, "learning_rate": 8.750048839217977e-05, "loss": 0.5312, "step": 2660 }, { "epoch": 0.9248354693453412, "grad_norm": 0.0, "learning_rate": 8.736688071266624e-05, "loss": 0.5147, "step": 2670 }, { "epoch": 0.9282992726013163, "grad_norm": 0.0, "learning_rate": 8.723266592620897e-05, "loss": 0.6035, "step": 2680 }, { "epoch": 0.9317630758572913, "grad_norm": 0.0, "learning_rate": 8.709784621342164e-05, "loss": 0.6515, "step": 2690 }, { "epoch": 0.9352268791132664, "grad_norm": 0.0, "learning_rate": 8.696242376474618e-05, "loss": 0.5814, "step": 2700 }, { "epoch": 0.9386906823692415, "grad_norm": 0.0, "learning_rate": 8.682640078041733e-05, "loss": 0.5382, "step": 2710 }, { "epoch": 0.9421544856252165, "grad_norm": 0.0, "learning_rate": 8.668977947042685e-05, "loss": 0.5345, "step": 2720 }, { "epoch": 0.9456182888811916, "grad_norm": 0.0, "learning_rate": 8.655256205448761e-05, "loss": 0.6344, "step": 2730 }, { "epoch": 0.9490820921371667, "grad_norm": 0.0, "learning_rate": 8.641475076199751e-05, "loss": 0.6546, "step": 2740 }, { "epoch": 0.9525458953931417, "grad_norm": 0.0, "learning_rate": 8.627634783200331e-05, "loss": 0.5743, "step": 2750 }, { "epoch": 0.9560096986491168, "grad_norm": 0.0, "learning_rate": 8.61373555131642e-05, "loss": 0.5926, "step": 2760 }, { "epoch": 0.9594735019050918, "grad_norm": 0.0, "learning_rate": 8.599777606371528e-05, "loss": 0.643, "step": 2770 }, { "epoch": 0.9629373051610669, "grad_norm": 0.0, "learning_rate": 8.585761175143091e-05, "loss": 0.507, "step": 2780 }, { "epoch": 0.966401108417042, "grad_norm": 0.0, "learning_rate": 8.571686485358777e-05, "loss": 0.5488, "step": 2790 }, { "epoch": 0.969864911673017, "grad_norm": 0.0, "learning_rate": 8.557553765692796e-05, "loss": 0.4403, "step": 2800 }, { "epoch": 0.973328714928992, "grad_norm": 0.0, "learning_rate": 8.543363245762182e-05, "loss": 0.5095, "step": 2810 }, { "epoch": 0.976792518184967, "grad_norm": 0.0, "learning_rate": 8.529115156123055e-05, "loss": 0.6275, "step": 2820 }, { "epoch": 0.9802563214409421, "grad_norm": 0.0, "learning_rate": 8.514809728266888e-05, "loss": 0.5198, "step": 2830 }, { "epoch": 0.9837201246969172, "grad_norm": 0.0, "learning_rate": 8.500447194616737e-05, "loss": 0.6168, "step": 2840 }, { "epoch": 0.9871839279528922, "grad_norm": 0.0, "learning_rate": 8.486027788523464e-05, "loss": 0.5684, "step": 2850 }, { "epoch": 0.9906477312088673, "grad_norm": 0.0, "learning_rate": 8.471551744261952e-05, "loss": 0.5247, "step": 2860 }, { "epoch": 0.9941115344648424, "grad_norm": 0.0, "learning_rate": 8.457019297027295e-05, "loss": 0.533, "step": 2870 }, { "epoch": 0.9975753377208174, "grad_norm": 0.0, "learning_rate": 8.442430682930975e-05, "loss": 0.5764, "step": 2880 }, { "epoch": 1.0010391409767925, "grad_norm": 0.0, "learning_rate": 8.427786138997029e-05, "loss": 0.4921, "step": 2890 }, { "epoch": 1.0045029442327675, "grad_norm": 0.0, "learning_rate": 8.413085903158197e-05, "loss": 0.3668, "step": 2900 }, { "epoch": 1.0079667474887426, "grad_norm": 0.0, "learning_rate": 8.398330214252057e-05, "loss": 0.3174, "step": 2910 }, { "epoch": 1.0114305507447177, "grad_norm": 0.0, "learning_rate": 8.383519312017142e-05, "loss": 0.3762, "step": 2920 }, { "epoch": 1.0148943540006927, "grad_norm": 0.0, "learning_rate": 8.368653437089048e-05, "loss": 0.387, "step": 2930 }, { "epoch": 1.0183581572566678, "grad_norm": 0.0, "learning_rate": 8.353732830996522e-05, "loss": 0.4243, "step": 2940 }, { "epoch": 1.0218219605126428, "grad_norm": 0.0, "learning_rate": 8.338757736157536e-05, "loss": 0.3473, "step": 2950 }, { "epoch": 1.025285763768618, "grad_norm": 0.0, "learning_rate": 8.323728395875361e-05, "loss": 0.3895, "step": 2960 }, { "epoch": 1.028749567024593, "grad_norm": 0.0, "learning_rate": 8.308645054334593e-05, "loss": 0.3705, "step": 2970 }, { "epoch": 1.032213370280568, "grad_norm": 0.0, "learning_rate": 8.293507956597202e-05, "loss": 0.4331, "step": 2980 }, { "epoch": 1.035677173536543, "grad_norm": 0.0, "learning_rate": 8.278317348598546e-05, "loss": 0.4036, "step": 2990 }, { "epoch": 1.0391409767925182, "grad_norm": 0.0, "learning_rate": 8.263073477143373e-05, "loss": 0.3403, "step": 3000 }, { "epoch": 1.0426047800484932, "grad_norm": 0.0, "learning_rate": 8.247776589901815e-05, "loss": 0.3748, "step": 3010 }, { "epoch": 1.0460685833044683, "grad_norm": 0.0, "learning_rate": 8.232426935405357e-05, "loss": 0.4261, "step": 3020 }, { "epoch": 1.0495323865604433, "grad_norm": 0.0, "learning_rate": 8.217024763042807e-05, "loss": 0.4723, "step": 3030 }, { "epoch": 1.0529961898164184, "grad_norm": 0.0, "learning_rate": 8.201570323056243e-05, "loss": 0.4268, "step": 3040 }, { "epoch": 1.0564599930723935, "grad_norm": 0.0, "learning_rate": 8.186063866536939e-05, "loss": 0.3854, "step": 3050 }, { "epoch": 1.0599237963283685, "grad_norm": 0.0, "learning_rate": 8.1705056454213e-05, "loss": 0.4284, "step": 3060 }, { "epoch": 1.0633875995843436, "grad_norm": 0.0, "learning_rate": 8.154895912486753e-05, "loss": 0.4008, "step": 3070 }, { "epoch": 1.0668514028403187, "grad_norm": 0.0, "learning_rate": 8.139234921347652e-05, "loss": 0.3511, "step": 3080 }, { "epoch": 1.0703152060962937, "grad_norm": 0.0, "learning_rate": 8.123522926451149e-05, "loss": 0.4124, "step": 3090 }, { "epoch": 1.0737790093522688, "grad_norm": 0.0, "learning_rate": 8.107760183073069e-05, "loss": 0.374, "step": 3100 }, { "epoch": 1.0772428126082438, "grad_norm": 0.0, "learning_rate": 8.091946947313753e-05, "loss": 0.4035, "step": 3110 }, { "epoch": 1.080706615864219, "grad_norm": 0.0, "learning_rate": 8.076083476093904e-05, "loss": 0.4288, "step": 3120 }, { "epoch": 1.084170419120194, "grad_norm": 0.0, "learning_rate": 8.060170027150408e-05, "loss": 0.4528, "step": 3130 }, { "epoch": 1.087634222376169, "grad_norm": 0.0, "learning_rate": 8.044206859032155e-05, "loss": 0.3165, "step": 3140 }, { "epoch": 1.091098025632144, "grad_norm": 0.0, "learning_rate": 8.02819423109582e-05, "loss": 0.418, "step": 3150 }, { "epoch": 1.0945618288881191, "grad_norm": 0.0, "learning_rate": 8.012132403501674e-05, "loss": 0.3858, "step": 3160 }, { "epoch": 1.0980256321440942, "grad_norm": 0.0, "learning_rate": 7.996021637209337e-05, "loss": 0.3647, "step": 3170 }, { "epoch": 1.1014894354000693, "grad_norm": 0.0, "learning_rate": 7.979862193973548e-05, "loss": 0.4054, "step": 3180 }, { "epoch": 1.1049532386560443, "grad_norm": 0.0, "learning_rate": 7.963654336339907e-05, "loss": 0.3867, "step": 3190 }, { "epoch": 1.1084170419120194, "grad_norm": 0.0, "learning_rate": 7.947398327640618e-05, "loss": 0.4027, "step": 3200 }, { "epoch": 1.1118808451679945, "grad_norm": 0.0, "learning_rate": 7.931094431990198e-05, "loss": 0.4349, "step": 3210 }, { "epoch": 1.1153446484239695, "grad_norm": 0.0, "learning_rate": 7.914742914281195e-05, "loss": 0.4888, "step": 3220 }, { "epoch": 1.1188084516799446, "grad_norm": 0.0, "learning_rate": 7.898344040179884e-05, "loss": 0.3822, "step": 3230 }, { "epoch": 1.1222722549359196, "grad_norm": 0.0, "learning_rate": 7.881898076121945e-05, "loss": 0.4214, "step": 3240 }, { "epoch": 1.1257360581918947, "grad_norm": 0.0, "learning_rate": 7.86540528930814e-05, "loss": 0.3544, "step": 3250 }, { "epoch": 1.1291998614478698, "grad_norm": 0.0, "learning_rate": 7.84886594769997e-05, "loss": 0.4062, "step": 3260 }, { "epoch": 1.1326636647038448, "grad_norm": 0.0, "learning_rate": 7.83228032001532e-05, "loss": 0.4396, "step": 3270 }, { "epoch": 1.13612746795982, "grad_norm": 0.0, "learning_rate": 7.81564867572409e-05, "loss": 0.4865, "step": 3280 }, { "epoch": 1.139591271215795, "grad_norm": 0.0, "learning_rate": 7.798971285043825e-05, "loss": 0.4124, "step": 3290 }, { "epoch": 1.14305507447177, "grad_norm": 0.0, "learning_rate": 7.782248418935322e-05, "loss": 0.3888, "step": 3300 }, { "epoch": 1.146518877727745, "grad_norm": 0.0, "learning_rate": 7.765480349098222e-05, "loss": 0.3717, "step": 3310 }, { "epoch": 1.1499826809837201, "grad_norm": 0.0, "learning_rate": 7.748667347966601e-05, "loss": 0.3894, "step": 3320 }, { "epoch": 1.1534464842396952, "grad_norm": 0.0, "learning_rate": 7.73180968870454e-05, "loss": 0.3275, "step": 3330 }, { "epoch": 1.1569102874956703, "grad_norm": 0.0, "learning_rate": 7.714907645201699e-05, "loss": 0.5511, "step": 3340 }, { "epoch": 1.1603740907516453, "grad_norm": 0.0, "learning_rate": 7.697961492068846e-05, "loss": 0.4086, "step": 3350 }, { "epoch": 1.1638378940076204, "grad_norm": 0.0, "learning_rate": 7.680971504633409e-05, "loss": 0.3803, "step": 3360 }, { "epoch": 1.1673016972635955, "grad_norm": 0.0, "learning_rate": 7.663937958935007e-05, "loss": 0.3686, "step": 3370 }, { "epoch": 1.1707655005195705, "grad_norm": 0.0, "learning_rate": 7.64686113172095e-05, "loss": 0.4075, "step": 3380 }, { "epoch": 1.1742293037755456, "grad_norm": 0.0, "learning_rate": 7.629741300441758e-05, "loss": 0.3727, "step": 3390 }, { "epoch": 1.1776931070315206, "grad_norm": 0.0, "learning_rate": 7.612578743246643e-05, "loss": 0.4295, "step": 3400 }, { "epoch": 1.1811569102874957, "grad_norm": 0.0, "learning_rate": 7.595373738978993e-05, "loss": 0.3715, "step": 3410 }, { "epoch": 1.1846207135434708, "grad_norm": 0.0, "learning_rate": 7.578126567171841e-05, "loss": 0.4631, "step": 3420 }, { "epoch": 1.1880845167994458, "grad_norm": 0.0, "learning_rate": 7.56083750804333e-05, "loss": 0.3857, "step": 3430 }, { "epoch": 1.1915483200554209, "grad_norm": 0.0, "learning_rate": 7.543506842492146e-05, "loss": 0.4074, "step": 3440 }, { "epoch": 1.195012123311396, "grad_norm": 0.0, "learning_rate": 7.52613485209297e-05, "loss": 0.3842, "step": 3450 }, { "epoch": 1.198475926567371, "grad_norm": 0.0, "learning_rate": 7.50872181909189e-05, "loss": 0.4116, "step": 3460 }, { "epoch": 1.201939729823346, "grad_norm": 0.0, "learning_rate": 7.491268026401824e-05, "loss": 0.3753, "step": 3470 }, { "epoch": 1.2054035330793211, "grad_norm": 0.0, "learning_rate": 7.473773757597923e-05, "loss": 0.4342, "step": 3480 }, { "epoch": 1.2088673363352962, "grad_norm": 0.0, "learning_rate": 7.456239296912955e-05, "loss": 0.4531, "step": 3490 }, { "epoch": 1.2123311395912713, "grad_norm": 0.0, "learning_rate": 7.438664929232696e-05, "loss": 0.3419, "step": 3500 }, { "epoch": 1.2157949428472463, "grad_norm": 0.0, "learning_rate": 7.421050940091304e-05, "loss": 0.4202, "step": 3510 }, { "epoch": 1.2192587461032214, "grad_norm": 0.0, "learning_rate": 7.403397615666666e-05, "loss": 0.4131, "step": 3520 }, { "epoch": 1.2227225493591964, "grad_norm": 0.0, "learning_rate": 7.385705242775765e-05, "loss": 0.4758, "step": 3530 }, { "epoch": 1.2261863526151715, "grad_norm": 0.0, "learning_rate": 7.367974108870002e-05, "loss": 0.4329, "step": 3540 }, { "epoch": 1.2296501558711466, "grad_norm": 0.0, "learning_rate": 7.350204502030548e-05, "loss": 0.3535, "step": 3550 }, { "epoch": 1.2331139591271216, "grad_norm": 0.0, "learning_rate": 7.33239671096364e-05, "loss": 0.5154, "step": 3560 }, { "epoch": 1.2365777623830967, "grad_norm": 0.0, "learning_rate": 7.314551024995907e-05, "loss": 0.4163, "step": 3570 }, { "epoch": 1.2400415656390718, "grad_norm": 0.0, "learning_rate": 7.296667734069665e-05, "loss": 0.3093, "step": 3580 }, { "epoch": 1.2435053688950468, "grad_norm": 0.0, "learning_rate": 7.278747128738204e-05, "loss": 0.713, "step": 3590 }, { "epoch": 1.2469691721510219, "grad_norm": 0.0, "learning_rate": 7.260789500161064e-05, "loss": 0.3616, "step": 3600 }, { "epoch": 1.250432975406997, "grad_norm": 0.0, "learning_rate": 7.242795140099315e-05, "loss": 0.6048, "step": 3610 }, { "epoch": 1.253896778662972, "grad_norm": 0.0, "learning_rate": 7.224764340910807e-05, "loss": 0.4443, "step": 3620 }, { "epoch": 1.257360581918947, "grad_norm": 0.0, "learning_rate": 7.206697395545425e-05, "loss": 0.4197, "step": 3630 }, { "epoch": 1.2608243851749221, "grad_norm": 0.0, "learning_rate": 7.188594597540326e-05, "loss": 0.4118, "step": 3640 }, { "epoch": 1.2642881884308972, "grad_norm": 0.0, "learning_rate": 7.170456241015175e-05, "loss": 0.3691, "step": 3650 }, { "epoch": 1.2677519916868722, "grad_norm": 0.0, "learning_rate": 7.152282620667362e-05, "loss": 0.3787, "step": 3660 }, { "epoch": 1.2712157949428473, "grad_norm": 0.0, "learning_rate": 7.134074031767213e-05, "loss": 0.4951, "step": 3670 }, { "epoch": 1.2746795981988224, "grad_norm": 0.0, "learning_rate": 7.115830770153194e-05, "loss": 0.3512, "step": 3680 }, { "epoch": 1.2781434014547974, "grad_norm": 0.0, "learning_rate": 7.097553132227111e-05, "loss": 0.4241, "step": 3690 }, { "epoch": 1.2816072047107725, "grad_norm": 0.0, "learning_rate": 7.079241414949287e-05, "loss": 0.5361, "step": 3700 }, { "epoch": 1.2850710079667476, "grad_norm": 0.0, "learning_rate": 7.060895915833732e-05, "loss": 0.4206, "step": 3710 }, { "epoch": 1.2885348112227226, "grad_norm": 0.0, "learning_rate": 7.042516932943324e-05, "loss": 0.4336, "step": 3720 }, { "epoch": 1.2919986144786977, "grad_norm": 0.0, "learning_rate": 7.024104764884954e-05, "loss": 0.4611, "step": 3730 }, { "epoch": 1.2954624177346727, "grad_norm": 0.0, "learning_rate": 7.005659710804683e-05, "loss": 0.4391, "step": 3740 }, { "epoch": 1.2989262209906478, "grad_norm": 0.0, "learning_rate": 6.987182070382872e-05, "loss": 0.4216, "step": 3750 }, { "epoch": 1.3023900242466229, "grad_norm": 0.0, "learning_rate": 6.968672143829323e-05, "loss": 0.414, "step": 3760 }, { "epoch": 1.305853827502598, "grad_norm": 0.0, "learning_rate": 6.950130231878395e-05, "loss": 0.4694, "step": 3770 }, { "epoch": 1.309317630758573, "grad_norm": 0.0, "learning_rate": 6.931556635784121e-05, "loss": 0.5899, "step": 3780 }, { "epoch": 1.312781434014548, "grad_norm": 0.0, "learning_rate": 6.912951657315312e-05, "loss": 0.3917, "step": 3790 }, { "epoch": 1.3162452372705231, "grad_norm": 0.0, "learning_rate": 6.894315598750653e-05, "loss": 0.4419, "step": 3800 }, { "epoch": 1.3197090405264982, "grad_norm": 0.0, "learning_rate": 6.875648762873794e-05, "loss": 0.4226, "step": 3810 }, { "epoch": 1.3231728437824732, "grad_norm": 0.0, "learning_rate": 6.856951452968435e-05, "loss": 0.5802, "step": 3820 }, { "epoch": 1.3266366470384483, "grad_norm": 0.0, "learning_rate": 6.838223972813385e-05, "loss": 0.4646, "step": 3830 }, { "epoch": 1.3301004502944234, "grad_norm": 0.0, "learning_rate": 6.819466626677641e-05, "loss": 0.4044, "step": 3840 }, { "epoch": 1.3335642535503984, "grad_norm": 0.0, "learning_rate": 6.800679719315434e-05, "loss": 0.3212, "step": 3850 }, { "epoch": 1.3370280568063735, "grad_norm": 0.0, "learning_rate": 6.781863555961288e-05, "loss": 0.4491, "step": 3860 }, { "epoch": 1.3404918600623486, "grad_norm": 0.0, "learning_rate": 6.763018442325048e-05, "loss": 0.465, "step": 3870 }, { "epoch": 1.3439556633183236, "grad_norm": 0.0, "learning_rate": 6.744144684586925e-05, "loss": 0.4373, "step": 3880 }, { "epoch": 1.3474194665742987, "grad_norm": 0.0, "learning_rate": 6.725242589392513e-05, "loss": 0.4165, "step": 3890 }, { "epoch": 1.3508832698302737, "grad_norm": 0.0, "learning_rate": 6.70631246384781e-05, "loss": 0.3812, "step": 3900 }, { "epoch": 1.3543470730862488, "grad_norm": 0.0, "learning_rate": 6.687354615514233e-05, "loss": 0.3743, "step": 3910 }, { "epoch": 1.3578108763422239, "grad_norm": 0.0, "learning_rate": 6.668369352403612e-05, "loss": 0.4039, "step": 3920 }, { "epoch": 1.361274679598199, "grad_norm": 0.0, "learning_rate": 6.64935698297319e-05, "loss": 0.4664, "step": 3930 }, { "epoch": 1.364738482854174, "grad_norm": 0.0, "learning_rate": 6.630317816120614e-05, "loss": 0.5085, "step": 3940 }, { "epoch": 1.368202286110149, "grad_norm": 0.0, "learning_rate": 6.611252161178916e-05, "loss": 0.4145, "step": 3950 }, { "epoch": 1.371666089366124, "grad_norm": 0.0, "learning_rate": 6.59216032791148e-05, "loss": 0.4935, "step": 3960 }, { "epoch": 1.3751298926220992, "grad_norm": 0.0, "learning_rate": 6.573042626507014e-05, "loss": 0.4667, "step": 3970 }, { "epoch": 1.3785936958780742, "grad_norm": 0.0, "learning_rate": 6.553899367574517e-05, "loss": 0.3932, "step": 3980 }, { "epoch": 1.3820574991340493, "grad_norm": 0.0, "learning_rate": 6.53473086213822e-05, "loss": 0.2886, "step": 3990 }, { "epoch": 1.3855213023900244, "grad_norm": 0.0, "learning_rate": 6.515537421632542e-05, "loss": 0.4391, "step": 4000 }, { "epoch": 1.3889851056459994, "grad_norm": 0.0, "learning_rate": 6.496319357897027e-05, "loss": 0.3598, "step": 4010 }, { "epoch": 1.3924489089019745, "grad_norm": 0.0, "learning_rate": 6.477076983171275e-05, "loss": 0.3555, "step": 4020 }, { "epoch": 1.3959127121579495, "grad_norm": 0.0, "learning_rate": 6.457810610089871e-05, "loss": 0.4305, "step": 4030 }, { "epoch": 1.3993765154139246, "grad_norm": 0.0, "learning_rate": 6.438520551677312e-05, "loss": 0.3731, "step": 4040 }, { "epoch": 1.4028403186698997, "grad_norm": 0.0, "learning_rate": 6.419207121342907e-05, "loss": 0.4329, "step": 4050 }, { "epoch": 1.4063041219258747, "grad_norm": 0.0, "learning_rate": 6.399870632875699e-05, "loss": 0.4039, "step": 4060 }, { "epoch": 1.4097679251818498, "grad_norm": 0.0, "learning_rate": 6.380511400439358e-05, "loss": 0.4267, "step": 4070 }, { "epoch": 1.4132317284378249, "grad_norm": 0.0, "learning_rate": 6.361129738567081e-05, "loss": 0.4556, "step": 4080 }, { "epoch": 1.4166955316938, "grad_norm": 0.0, "learning_rate": 6.341725962156481e-05, "loss": 0.3885, "step": 4090 }, { "epoch": 1.420159334949775, "grad_norm": 0.0, "learning_rate": 6.322300386464466e-05, "loss": 0.3971, "step": 4100 }, { "epoch": 1.42362313820575, "grad_norm": 0.0, "learning_rate": 6.302853327102129e-05, "loss": 0.4238, "step": 4110 }, { "epoch": 1.427086941461725, "grad_norm": 0.0, "learning_rate": 6.283385100029603e-05, "loss": 0.4204, "step": 4120 }, { "epoch": 1.4305507447177002, "grad_norm": 0.0, "learning_rate": 6.263896021550945e-05, "loss": 0.4143, "step": 4130 }, { "epoch": 1.4340145479736752, "grad_norm": 0.0, "learning_rate": 6.244386408308983e-05, "loss": 0.3515, "step": 4140 }, { "epoch": 1.4374783512296503, "grad_norm": 0.0, "learning_rate": 6.224856577280181e-05, "loss": 0.3727, "step": 4150 }, { "epoch": 1.4409421544856253, "grad_norm": 0.0, "learning_rate": 6.205306845769485e-05, "loss": 0.4541, "step": 4160 }, { "epoch": 1.4444059577416004, "grad_norm": 0.0, "learning_rate": 6.185737531405161e-05, "loss": 0.4538, "step": 4170 }, { "epoch": 1.4478697609975755, "grad_norm": 0.0, "learning_rate": 6.166148952133656e-05, "loss": 0.343, "step": 4180 }, { "epoch": 1.4513335642535505, "grad_norm": 0.0, "learning_rate": 6.146541426214403e-05, "loss": 0.4383, "step": 4190 }, { "epoch": 1.4547973675095256, "grad_norm": 0.0, "learning_rate": 6.126915272214674e-05, "loss": 0.3636, "step": 4200 }, { "epoch": 1.4582611707655007, "grad_norm": 0.0, "learning_rate": 6.107270809004389e-05, "loss": 0.4339, "step": 4210 }, { "epoch": 1.4617249740214755, "grad_norm": 0.0, "learning_rate": 6.087608355750947e-05, "loss": 0.3828, "step": 4220 }, { "epoch": 1.4651887772774506, "grad_norm": 0.0, "learning_rate": 6.0679282319140294e-05, "loss": 0.4197, "step": 4230 }, { "epoch": 1.4686525805334256, "grad_norm": 0.0, "learning_rate": 6.048230757240419e-05, "loss": 0.343, "step": 4240 }, { "epoch": 1.4721163837894007, "grad_norm": 0.0, "learning_rate": 6.0285162517588e-05, "loss": 0.4459, "step": 4250 }, { "epoch": 1.4755801870453757, "grad_norm": 0.0, "learning_rate": 6.0087850357745566e-05, "loss": 0.351, "step": 4260 }, { "epoch": 1.4790439903013508, "grad_norm": 0.0, "learning_rate": 5.989037429864578e-05, "loss": 0.3419, "step": 4270 }, { "epoch": 1.4825077935573259, "grad_norm": 0.0, "learning_rate": 5.969273754872036e-05, "loss": 0.5075, "step": 4280 }, { "epoch": 1.485971596813301, "grad_norm": 0.0, "learning_rate": 5.949494331901183e-05, "loss": 0.4529, "step": 4290 }, { "epoch": 1.489435400069276, "grad_norm": 0.0, "learning_rate": 5.9296994823121365e-05, "loss": 0.4633, "step": 4300 }, { "epoch": 1.492899203325251, "grad_norm": 0.0, "learning_rate": 5.909889527715648e-05, "loss": 0.4304, "step": 4310 }, { "epoch": 1.4963630065812261, "grad_norm": 0.0, "learning_rate": 5.890064789967884e-05, "loss": 0.3747, "step": 4320 }, { "epoch": 1.4998268098372012, "grad_norm": 0.0, "learning_rate": 5.870225591165195e-05, "loss": 0.3392, "step": 4330 }, { "epoch": 1.5032906130931765, "grad_norm": 0.0, "learning_rate": 5.850372253638884e-05, "loss": 0.3755, "step": 4340 }, { "epoch": 1.5067544163491515, "grad_norm": 0.0, "learning_rate": 5.8305050999499686e-05, "loss": 0.471, "step": 4350 }, { "epoch": 1.5102182196051266, "grad_norm": 0.0, "learning_rate": 5.810624452883941e-05, "loss": 0.4076, "step": 4360 }, { "epoch": 1.5136820228611017, "grad_norm": 0.0, "learning_rate": 5.790730635445524e-05, "loss": 0.3986, "step": 4370 }, { "epoch": 1.5171458261170767, "grad_norm": 0.0, "learning_rate": 5.7708239708534174e-05, "loss": 0.4455, "step": 4380 }, { "epoch": 1.5206096293730518, "grad_norm": 0.0, "learning_rate": 5.750904782535055e-05, "loss": 0.4939, "step": 4390 }, { "epoch": 1.5240734326290268, "grad_norm": 0.0, "learning_rate": 5.7309733941213465e-05, "loss": 0.4361, "step": 4400 }, { "epoch": 1.527537235885002, "grad_norm": 0.0, "learning_rate": 5.711030129441413e-05, "loss": 0.4418, "step": 4410 }, { "epoch": 1.5310010391409767, "grad_norm": 0.0, "learning_rate": 5.6910753125173386e-05, "loss": 0.4133, "step": 4420 }, { "epoch": 1.5344648423969518, "grad_norm": 0.0, "learning_rate": 5.6711092675588984e-05, "loss": 0.3846, "step": 4430 }, { "epoch": 1.5379286456529269, "grad_norm": 0.0, "learning_rate": 5.651132318958288e-05, "loss": 0.3305, "step": 4440 }, { "epoch": 1.541392448908902, "grad_norm": 0.0, "learning_rate": 5.631144791284857e-05, "loss": 0.3696, "step": 4450 }, { "epoch": 1.544856252164877, "grad_norm": 0.0, "learning_rate": 5.6111470092798366e-05, "loss": 0.4406, "step": 4460 }, { "epoch": 1.548320055420852, "grad_norm": 0.0, "learning_rate": 5.5911392978510614e-05, "loss": 0.4038, "step": 4470 }, { "epoch": 1.551783858676827, "grad_norm": 0.0, "learning_rate": 5.571121982067693e-05, "loss": 0.5092, "step": 4480 }, { "epoch": 1.5552476619328022, "grad_norm": 0.0, "learning_rate": 5.551095387154936e-05, "loss": 0.4212, "step": 4490 }, { "epoch": 1.5587114651887772, "grad_norm": 0.0, "learning_rate": 5.5310598384887536e-05, "loss": 0.5457, "step": 4500 }, { "epoch": 1.5621752684447523, "grad_norm": 0.0, "learning_rate": 5.511015661590582e-05, "loss": 0.4103, "step": 4510 }, { "epoch": 1.5656390717007274, "grad_norm": 0.0, "learning_rate": 5.490963182122044e-05, "loss": 0.4087, "step": 4520 }, { "epoch": 1.5691028749567024, "grad_norm": 0.0, "learning_rate": 5.470902725879655e-05, "loss": 0.4258, "step": 4530 }, { "epoch": 1.5725666782126775, "grad_norm": 0.0, "learning_rate": 5.45083461878953e-05, "loss": 0.396, "step": 4540 }, { "epoch": 1.5760304814686525, "grad_norm": 0.0, "learning_rate": 5.430759186902089e-05, "loss": 0.3761, "step": 4550 }, { "epoch": 1.5794942847246276, "grad_norm": 0.0, "learning_rate": 5.410676756386761e-05, "loss": 0.4523, "step": 4560 }, { "epoch": 1.5829580879806027, "grad_norm": 0.0, "learning_rate": 5.390587653526682e-05, "loss": 0.3949, "step": 4570 }, { "epoch": 1.5864218912365777, "grad_norm": 0.0, "learning_rate": 5.370492204713392e-05, "loss": 0.4029, "step": 4580 }, { "epoch": 1.5898856944925528, "grad_norm": 0.0, "learning_rate": 5.350390736441539e-05, "loss": 0.439, "step": 4590 }, { "epoch": 1.5933494977485279, "grad_norm": 0.0, "learning_rate": 5.33028357530357e-05, "loss": 0.4381, "step": 4600 }, { "epoch": 1.596813301004503, "grad_norm": 0.0, "learning_rate": 5.3101710479844226e-05, "loss": 0.4609, "step": 4610 }, { "epoch": 1.600277104260478, "grad_norm": 0.0, "learning_rate": 5.290053481256223e-05, "loss": 0.3574, "step": 4620 }, { "epoch": 1.603740907516453, "grad_norm": 0.0, "learning_rate": 5.2699312019729686e-05, "loss": 0.5029, "step": 4630 }, { "epoch": 1.607204710772428, "grad_norm": 0.0, "learning_rate": 5.249804537065228e-05, "loss": 0.5594, "step": 4640 }, { "epoch": 1.6106685140284032, "grad_norm": 0.0, "learning_rate": 5.229673813534821e-05, "loss": 0.3988, "step": 4650 }, { "epoch": 1.6141323172843782, "grad_norm": 0.0, "learning_rate": 5.209539358449511e-05, "loss": 0.4262, "step": 4660 }, { "epoch": 1.6175961205403533, "grad_norm": 0.0, "learning_rate": 5.189401498937685e-05, "loss": 0.3805, "step": 4670 }, { "epoch": 1.6210599237963284, "grad_norm": 0.0, "learning_rate": 5.1692605621830435e-05, "loss": 0.3792, "step": 4680 }, { "epoch": 1.6245237270523034, "grad_norm": 0.0, "learning_rate": 5.1491168754192876e-05, "loss": 0.3725, "step": 4690 }, { "epoch": 1.6279875303082785, "grad_norm": 0.0, "learning_rate": 5.128970765924794e-05, "loss": 0.3389, "step": 4700 }, { "epoch": 1.6314513335642535, "grad_norm": 0.0, "learning_rate": 5.1088225610173016e-05, "loss": 0.3982, "step": 4710 }, { "epoch": 1.6349151368202286, "grad_norm": 0.0, "learning_rate": 5.088672588048596e-05, "loss": 0.3351, "step": 4720 }, { "epoch": 1.6383789400762037, "grad_norm": 0.0, "learning_rate": 5.06852117439919e-05, "loss": 0.4089, "step": 4730 }, { "epoch": 1.6418427433321787, "grad_norm": 0.0, "learning_rate": 5.048368647473e-05, "loss": 0.3532, "step": 4740 }, { "epoch": 1.6453065465881538, "grad_norm": 0.0, "learning_rate": 5.028215334692034e-05, "loss": 0.3791, "step": 4750 }, { "epoch": 1.6487703498441288, "grad_norm": 0.0, "learning_rate": 5.008061563491062e-05, "loss": 0.3813, "step": 4760 }, { "epoch": 1.652234153100104, "grad_norm": 0.0, "learning_rate": 4.9879076613123074e-05, "loss": 0.3775, "step": 4770 }, { "epoch": 1.655697956356079, "grad_norm": 0.0, "learning_rate": 4.9677539556001195e-05, "loss": 0.3367, "step": 4780 }, { "epoch": 1.659161759612054, "grad_norm": 0.0, "learning_rate": 4.947600773795659e-05, "loss": 0.3025, "step": 4790 }, { "epoch": 1.662625562868029, "grad_norm": 0.0, "learning_rate": 4.927448443331566e-05, "loss": 0.3689, "step": 4800 }, { "epoch": 1.6660893661240042, "grad_norm": 0.0, "learning_rate": 4.90729729162666e-05, "loss": 0.451, "step": 4810 }, { "epoch": 1.6695531693799792, "grad_norm": 0.0, "learning_rate": 4.887147646080601e-05, "loss": 0.3949, "step": 4820 }, { "epoch": 1.6730169726359543, "grad_norm": 0.0, "learning_rate": 4.866999834068577e-05, "loss": 0.3868, "step": 4830 }, { "epoch": 1.6764807758919293, "grad_norm": 0.0, "learning_rate": 4.846854182935994e-05, "loss": 0.3887, "step": 4840 }, { "epoch": 1.6799445791479044, "grad_norm": 0.0, "learning_rate": 4.826711019993143e-05, "loss": 0.3507, "step": 4850 }, { "epoch": 1.6834083824038795, "grad_norm": 0.0, "learning_rate": 4.806570672509894e-05, "loss": 0.4483, "step": 4860 }, { "epoch": 1.6868721856598545, "grad_norm": 0.0, "learning_rate": 4.786433467710369e-05, "loss": 0.3929, "step": 4870 }, { "epoch": 1.6903359889158296, "grad_norm": 0.0, "learning_rate": 4.766299732767631e-05, "loss": 0.3771, "step": 4880 }, { "epoch": 1.6937997921718047, "grad_norm": 0.0, "learning_rate": 4.746169794798371e-05, "loss": 0.4597, "step": 4890 }, { "epoch": 1.6972635954277797, "grad_norm": 0.0, "learning_rate": 4.7260439808575855e-05, "loss": 0.3098, "step": 4900 }, { "epoch": 1.7007273986837548, "grad_norm": 0.0, "learning_rate": 4.70592261793327e-05, "loss": 0.3623, "step": 4910 }, { "epoch": 1.7041912019397298, "grad_norm": 0.0, "learning_rate": 4.685806032941104e-05, "loss": 0.4305, "step": 4920 }, { "epoch": 1.707655005195705, "grad_norm": 0.0, "learning_rate": 4.665694552719135e-05, "loss": 0.3996, "step": 4930 }, { "epoch": 1.71111880845168, "grad_norm": 0.0, "learning_rate": 4.6455885040224784e-05, "loss": 0.4478, "step": 4940 }, { "epoch": 1.714582611707655, "grad_norm": 0.0, "learning_rate": 4.625488213517996e-05, "loss": 0.458, "step": 4950 }, { "epoch": 1.71804641496363, "grad_norm": 0.0, "learning_rate": 4.605394007778999e-05, "loss": 0.4088, "step": 4960 }, { "epoch": 1.7215102182196051, "grad_norm": 0.0, "learning_rate": 4.585306213279939e-05, "loss": 0.4371, "step": 4970 }, { "epoch": 1.7249740214755802, "grad_norm": 0.0, "learning_rate": 4.565225156391099e-05, "loss": 0.3715, "step": 4980 }, { "epoch": 1.7284378247315553, "grad_norm": 0.0, "learning_rate": 4.5451511633733e-05, "loss": 0.4433, "step": 4990 }, { "epoch": 1.7319016279875303, "grad_norm": 0.0, "learning_rate": 4.525084560372591e-05, "loss": 0.421, "step": 5000 }, { "epoch": 1.7353654312435054, "grad_norm": 0.0, "learning_rate": 4.505025673414953e-05, "loss": 0.4111, "step": 5010 }, { "epoch": 1.7388292344994805, "grad_norm": 0.0, "learning_rate": 4.4849748284010076e-05, "loss": 0.4606, "step": 5020 }, { "epoch": 1.7422930377554555, "grad_norm": 0.0, "learning_rate": 4.4649323511007146e-05, "loss": 0.385, "step": 5030 }, { "epoch": 1.7457568410114306, "grad_norm": 0.0, "learning_rate": 4.4448985671480795e-05, "loss": 0.4407, "step": 5040 }, { "epoch": 1.7492206442674056, "grad_norm": 0.0, "learning_rate": 4.424873802035872e-05, "loss": 0.3384, "step": 5050 }, { "epoch": 1.7526844475233807, "grad_norm": 0.0, "learning_rate": 4.404858381110323e-05, "loss": 0.424, "step": 5060 }, { "epoch": 1.7561482507793558, "grad_norm": 0.0, "learning_rate": 4.384852629565854e-05, "loss": 0.4366, "step": 5070 }, { "epoch": 1.7596120540353308, "grad_norm": 0.0, "learning_rate": 4.3648568724397796e-05, "loss": 0.3986, "step": 5080 }, { "epoch": 1.763075857291306, "grad_norm": 0.0, "learning_rate": 4.3448714346070375e-05, "loss": 0.371, "step": 5090 }, { "epoch": 1.766539660547281, "grad_norm": 0.0, "learning_rate": 4.3248966407749056e-05, "loss": 0.3735, "step": 5100 }, { "epoch": 1.770003463803256, "grad_norm": 0.0, "learning_rate": 4.304932815477724e-05, "loss": 0.4722, "step": 5110 }, { "epoch": 1.773467267059231, "grad_norm": 0.0, "learning_rate": 4.284980283071628e-05, "loss": 0.4549, "step": 5120 }, { "epoch": 1.7769310703152061, "grad_norm": 0.0, "learning_rate": 4.265039367729273e-05, "loss": 0.3256, "step": 5130 }, { "epoch": 1.7803948735711812, "grad_norm": 0.0, "learning_rate": 4.245110393434569e-05, "loss": 0.4007, "step": 5140 }, { "epoch": 1.7838586768271563, "grad_norm": 0.0, "learning_rate": 4.2251936839774207e-05, "loss": 0.3073, "step": 5150 }, { "epoch": 1.7873224800831313, "grad_norm": 0.0, "learning_rate": 4.2052895629484615e-05, "loss": 0.388, "step": 5160 }, { "epoch": 1.7907862833391064, "grad_norm": 0.0, "learning_rate": 4.1853983537337965e-05, "loss": 0.517, "step": 5170 }, { "epoch": 1.7942500865950815, "grad_norm": 0.0, "learning_rate": 4.165520379509755e-05, "loss": 0.4707, "step": 5180 }, { "epoch": 1.7977138898510565, "grad_norm": 0.0, "learning_rate": 4.145655963237629e-05, "loss": 0.3916, "step": 5190 }, { "epoch": 1.8011776931070316, "grad_norm": 0.0, "learning_rate": 4.125805427658436e-05, "loss": 0.4286, "step": 5200 }, { "epoch": 1.8046414963630066, "grad_norm": 0.0, "learning_rate": 4.1059690952876716e-05, "loss": 0.405, "step": 5210 }, { "epoch": 1.8081052996189815, "grad_norm": 0.0, "learning_rate": 4.0861472884100625e-05, "loss": 0.5129, "step": 5220 }, { "epoch": 1.8115691028749565, "grad_norm": 0.0, "learning_rate": 4.066340329074348e-05, "loss": 0.4715, "step": 5230 }, { "epoch": 1.8150329061309316, "grad_norm": 0.0, "learning_rate": 4.046548539088024e-05, "loss": 0.4478, "step": 5240 }, { "epoch": 1.8184967093869067, "grad_norm": 0.0, "learning_rate": 4.026772240012139e-05, "loss": 0.5157, "step": 5250 }, { "epoch": 1.8219605126428817, "grad_norm": 0.0, "learning_rate": 4.0070117531560494e-05, "loss": 0.497, "step": 5260 }, { "epoch": 1.8254243158988568, "grad_norm": 0.0, "learning_rate": 3.9872673995722074e-05, "loss": 0.4075, "step": 5270 }, { "epoch": 1.8288881191548318, "grad_norm": 0.0, "learning_rate": 3.967539500050953e-05, "loss": 0.3514, "step": 5280 }, { "epoch": 1.832351922410807, "grad_norm": 0.0, "learning_rate": 3.947828375115285e-05, "loss": 0.4047, "step": 5290 }, { "epoch": 1.835815725666782, "grad_norm": 0.0, "learning_rate": 3.928134345015663e-05, "loss": 0.3344, "step": 5300 }, { "epoch": 1.839279528922757, "grad_norm": 0.0, "learning_rate": 3.908457729724812e-05, "loss": 0.3683, "step": 5310 }, { "epoch": 1.842743332178732, "grad_norm": 0.0, "learning_rate": 3.8887988489325036e-05, "loss": 0.3501, "step": 5320 }, { "epoch": 1.8462071354347072, "grad_norm": 0.0, "learning_rate": 3.869158022040383e-05, "loss": 0.4252, "step": 5330 }, { "epoch": 1.8496709386906822, "grad_norm": 0.0, "learning_rate": 3.849535568156766e-05, "loss": 0.3855, "step": 5340 }, { "epoch": 1.8531347419466573, "grad_norm": 0.0, "learning_rate": 3.829931806091456e-05, "loss": 0.3744, "step": 5350 }, { "epoch": 1.8565985452026323, "grad_norm": 0.0, "learning_rate": 3.810347054350574e-05, "loss": 0.4042, "step": 5360 }, { "epoch": 1.8600623484586074, "grad_norm": 0.0, "learning_rate": 3.7907816311313685e-05, "loss": 0.3851, "step": 5370 }, { "epoch": 1.8635261517145825, "grad_norm": 0.0, "learning_rate": 3.7712358543170604e-05, "loss": 0.4728, "step": 5380 }, { "epoch": 1.8669899549705575, "grad_norm": 0.0, "learning_rate": 3.7517100414716696e-05, "loss": 0.5332, "step": 5390 }, { "epoch": 1.8704537582265326, "grad_norm": 0.0, "learning_rate": 3.7322045098348524e-05, "loss": 0.4093, "step": 5400 }, { "epoch": 1.8739175614825077, "grad_norm": 0.0, "learning_rate": 3.712719576316762e-05, "loss": 0.3881, "step": 5410 }, { "epoch": 1.8773813647384827, "grad_norm": 0.0, "learning_rate": 3.693255557492882e-05, "loss": 0.4479, "step": 5420 }, { "epoch": 1.8808451679944578, "grad_norm": 0.0, "learning_rate": 3.673812769598892e-05, "loss": 0.5078, "step": 5430 }, { "epoch": 1.8843089712504328, "grad_norm": 0.0, "learning_rate": 3.6543915285255335e-05, "loss": 0.4404, "step": 5440 }, { "epoch": 1.887772774506408, "grad_norm": 0.0, "learning_rate": 3.6349921498134675e-05, "loss": 0.4353, "step": 5450 }, { "epoch": 1.891236577762383, "grad_norm": 0.0, "learning_rate": 3.615614948648158e-05, "loss": 0.4066, "step": 5460 }, { "epoch": 1.894700381018358, "grad_norm": 0.0, "learning_rate": 3.59626023985474e-05, "loss": 0.4357, "step": 5470 }, { "epoch": 1.898164184274333, "grad_norm": 0.0, "learning_rate": 3.576928337892913e-05, "loss": 0.4371, "step": 5480 }, { "epoch": 1.9016279875303082, "grad_norm": 0.0, "learning_rate": 3.557619556851833e-05, "loss": 0.4576, "step": 5490 }, { "epoch": 1.9050917907862832, "grad_norm": 0.0, "learning_rate": 3.538334210444999e-05, "loss": 0.3742, "step": 5500 }, { "epoch": 1.9085555940422583, "grad_norm": 0.0, "learning_rate": 3.519072612005171e-05, "loss": 0.4015, "step": 5510 }, { "epoch": 1.9120193972982333, "grad_norm": 0.0, "learning_rate": 3.4998350744792666e-05, "loss": 0.371, "step": 5520 }, { "epoch": 1.9154832005542084, "grad_norm": 0.0, "learning_rate": 3.4806219104232775e-05, "loss": 0.4006, "step": 5530 }, { "epoch": 1.9189470038101835, "grad_norm": 0.0, "learning_rate": 3.461433431997204e-05, "loss": 0.4639, "step": 5540 }, { "epoch": 1.9224108070661585, "grad_norm": 0.0, "learning_rate": 3.442269950959967e-05, "loss": 0.3715, "step": 5550 }, { "epoch": 1.9258746103221336, "grad_norm": 0.0, "learning_rate": 3.423131778664352e-05, "loss": 0.3253, "step": 5560 }, { "epoch": 1.9293384135781086, "grad_norm": 0.0, "learning_rate": 3.4040192260519496e-05, "loss": 0.3723, "step": 5570 }, { "epoch": 1.9328022168340837, "grad_norm": 0.0, "learning_rate": 3.3849326036480997e-05, "loss": 0.4015, "step": 5580 }, { "epoch": 1.9362660200900588, "grad_norm": 0.0, "learning_rate": 3.365872221556855e-05, "loss": 0.4422, "step": 5590 }, { "epoch": 1.9397298233460338, "grad_norm": 0.0, "learning_rate": 3.346838389455929e-05, "loss": 0.404, "step": 5600 }, { "epoch": 1.943193626602009, "grad_norm": 0.0, "learning_rate": 3.327831416591676e-05, "loss": 0.3971, "step": 5610 }, { "epoch": 1.946657429857984, "grad_norm": 0.0, "learning_rate": 3.308851611774064e-05, "loss": 0.4884, "step": 5620 }, { "epoch": 1.950121233113959, "grad_norm": 0.0, "learning_rate": 3.289899283371657e-05, "loss": 0.4646, "step": 5630 }, { "epoch": 1.953585036369934, "grad_norm": 0.0, "learning_rate": 3.270974739306601e-05, "loss": 0.3547, "step": 5640 }, { "epoch": 1.9570488396259091, "grad_norm": 0.0, "learning_rate": 3.252078287049628e-05, "loss": 0.5437, "step": 5650 }, { "epoch": 1.9605126428818842, "grad_norm": 0.0, "learning_rate": 3.233210233615054e-05, "loss": 0.4087, "step": 5660 }, { "epoch": 1.9639764461378593, "grad_norm": 0.0, "learning_rate": 3.2143708855557965e-05, "loss": 0.4111, "step": 5670 }, { "epoch": 1.9674402493938343, "grad_norm": 0.0, "learning_rate": 3.1955605489583895e-05, "loss": 0.3618, "step": 5680 }, { "epoch": 1.9709040526498094, "grad_norm": 0.0, "learning_rate": 3.176779529438011e-05, "loss": 0.4344, "step": 5690 }, { "epoch": 1.9743678559057845, "grad_norm": 0.0, "learning_rate": 3.158028132133524e-05, "loss": 0.4019, "step": 5700 }, { "epoch": 1.9778316591617595, "grad_norm": 0.0, "learning_rate": 3.1393066617025057e-05, "loss": 0.4044, "step": 5710 }, { "epoch": 1.9812954624177346, "grad_norm": 0.0, "learning_rate": 3.1206154223163156e-05, "loss": 0.4141, "step": 5720 }, { "epoch": 1.9847592656737096, "grad_norm": 0.0, "learning_rate": 3.101954717655133e-05, "loss": 0.409, "step": 5730 }, { "epoch": 1.9882230689296847, "grad_norm": 0.0, "learning_rate": 3.083324850903039e-05, "loss": 0.4244, "step": 5740 }, { "epoch": 1.9916868721856598, "grad_norm": 0.0, "learning_rate": 3.064726124743087e-05, "loss": 0.4055, "step": 5750 }, { "epoch": 1.9951506754416348, "grad_norm": 0.0, "learning_rate": 3.0461588413523782e-05, "loss": 0.4607, "step": 5760 }, { "epoch": 1.9986144786976099, "grad_norm": 0.0, "learning_rate": 3.0276233023971636e-05, "loss": 0.4081, "step": 5770 }, { "epoch": 2.002078281953585, "grad_norm": 0.0, "learning_rate": 3.0091198090279305e-05, "loss": 0.3135, "step": 5780 }, { "epoch": 2.00554208520956, "grad_norm": 0.0, "learning_rate": 2.9906486618745138e-05, "loss": 0.2966, "step": 5790 }, { "epoch": 2.009005888465535, "grad_norm": 0.0, "learning_rate": 2.972210161041221e-05, "loss": 0.2621, "step": 5800 }, { "epoch": 2.01246969172151, "grad_norm": 0.0, "learning_rate": 2.9538046061019444e-05, "loss": 0.255, "step": 5810 }, { "epoch": 2.015933494977485, "grad_norm": 0.0, "learning_rate": 2.935432296095293e-05, "loss": 0.2559, "step": 5820 }, { "epoch": 2.0193972982334603, "grad_norm": 0.0, "learning_rate": 2.917093529519749e-05, "loss": 0.2914, "step": 5830 }, { "epoch": 2.0228611014894353, "grad_norm": 0.0, "learning_rate": 2.898788604328798e-05, "loss": 0.4122, "step": 5840 }, { "epoch": 2.0263249047454104, "grad_norm": 0.0, "learning_rate": 2.8805178179261072e-05, "loss": 0.3004, "step": 5850 }, { "epoch": 2.0297887080013854, "grad_norm": 0.0, "learning_rate": 2.8622814671606774e-05, "loss": 0.2925, "step": 5860 }, { "epoch": 2.0332525112573605, "grad_norm": 0.0, "learning_rate": 2.8440798483220277e-05, "loss": 0.2903, "step": 5870 }, { "epoch": 2.0367163145133356, "grad_norm": 0.0, "learning_rate": 2.8259132571353863e-05, "loss": 0.2539, "step": 5880 }, { "epoch": 2.0401801177693106, "grad_norm": 0.0, "learning_rate": 2.8077819887568725e-05, "loss": 0.2658, "step": 5890 }, { "epoch": 2.0436439210252857, "grad_norm": 0.0, "learning_rate": 2.7896863377687142e-05, "loss": 0.3567, "step": 5900 }, { "epoch": 2.0471077242812608, "grad_norm": 0.0, "learning_rate": 2.7716265981744595e-05, "loss": 0.3112, "step": 5910 }, { "epoch": 2.050571527537236, "grad_norm": 0.0, "learning_rate": 2.7536030633941884e-05, "loss": 0.3093, "step": 5920 }, { "epoch": 2.054035330793211, "grad_norm": 0.0, "learning_rate": 2.735616026259763e-05, "loss": 0.2606, "step": 5930 }, { "epoch": 2.057499134049186, "grad_norm": 0.0, "learning_rate": 2.7176657790100545e-05, "loss": 0.2554, "step": 5940 }, { "epoch": 2.060962937305161, "grad_norm": 0.0, "learning_rate": 2.699752613286204e-05, "loss": 0.3455, "step": 5950 }, { "epoch": 2.064426740561136, "grad_norm": 0.0, "learning_rate": 2.6818768201268852e-05, "loss": 0.2383, "step": 5960 }, { "epoch": 2.067890543817111, "grad_norm": 0.0, "learning_rate": 2.664038689963566e-05, "loss": 0.2762, "step": 5970 }, { "epoch": 2.071354347073086, "grad_norm": 0.0, "learning_rate": 2.646238512615804e-05, "loss": 0.2606, "step": 5980 }, { "epoch": 2.0748181503290613, "grad_norm": 0.0, "learning_rate": 2.628476577286524e-05, "loss": 0.2504, "step": 5990 }, { "epoch": 2.0782819535850363, "grad_norm": 0.0, "learning_rate": 2.610753172557325e-05, "loss": 0.3267, "step": 6000 }, { "epoch": 2.0817457568410114, "grad_norm": 0.0, "learning_rate": 2.5930685863837957e-05, "loss": 0.2418, "step": 6010 }, { "epoch": 2.0852095600969864, "grad_norm": 0.0, "learning_rate": 2.575423106090834e-05, "loss": 0.2235, "step": 6020 }, { "epoch": 2.0886733633529615, "grad_norm": 0.0, "learning_rate": 2.557817018367969e-05, "loss": 0.2898, "step": 6030 }, { "epoch": 2.0921371666089366, "grad_norm": 0.0, "learning_rate": 2.5402506092647192e-05, "loss": 0.6742, "step": 6040 }, { "epoch": 2.0956009698649116, "grad_norm": 0.0, "learning_rate": 2.522724164185931e-05, "loss": 0.3034, "step": 6050 }, { "epoch": 2.0990647731208867, "grad_norm": 0.0, "learning_rate": 2.505237967887153e-05, "loss": 0.2349, "step": 6060 }, { "epoch": 2.1025285763768617, "grad_norm": 0.0, "learning_rate": 2.487792304469997e-05, "loss": 0.2861, "step": 6070 }, { "epoch": 2.105992379632837, "grad_norm": 0.0, "learning_rate": 2.4703874573775376e-05, "loss": 0.286, "step": 6080 }, { "epoch": 2.109456182888812, "grad_norm": 0.0, "learning_rate": 2.4530237093896918e-05, "loss": 0.2599, "step": 6090 }, { "epoch": 2.112919986144787, "grad_norm": 0.0, "learning_rate": 2.4357013426186338e-05, "loss": 0.2415, "step": 6100 }, { "epoch": 2.116383789400762, "grad_norm": 0.0, "learning_rate": 2.4184206385042114e-05, "loss": 0.2995, "step": 6110 }, { "epoch": 2.119847592656737, "grad_norm": 0.0, "learning_rate": 2.4011818778093686e-05, "loss": 0.291, "step": 6120 }, { "epoch": 2.123311395912712, "grad_norm": 0.0, "learning_rate": 2.383985340615585e-05, "loss": 0.2613, "step": 6130 }, { "epoch": 2.126775199168687, "grad_norm": 0.0, "learning_rate": 2.3668313063183302e-05, "loss": 0.2798, "step": 6140 }, { "epoch": 2.1302390024246622, "grad_norm": 0.0, "learning_rate": 2.3497200536225204e-05, "loss": 0.3092, "step": 6150 }, { "epoch": 2.1337028056806373, "grad_norm": 0.0, "learning_rate": 2.332651860537992e-05, "loss": 0.3397, "step": 6160 }, { "epoch": 2.1371666089366124, "grad_norm": 0.0, "learning_rate": 2.3156270043749806e-05, "loss": 0.2723, "step": 6170 }, { "epoch": 2.1406304121925874, "grad_norm": 0.0, "learning_rate": 2.2986457617396162e-05, "loss": 0.2671, "step": 6180 }, { "epoch": 2.1440942154485625, "grad_norm": 0.0, "learning_rate": 2.2817084085294393e-05, "loss": 0.3068, "step": 6190 }, { "epoch": 2.1475580187045376, "grad_norm": 0.0, "learning_rate": 2.264815219928903e-05, "loss": 0.3149, "step": 6200 }, { "epoch": 2.1510218219605126, "grad_norm": 0.0, "learning_rate": 2.2479664704049147e-05, "loss": 0.2734, "step": 6210 }, { "epoch": 2.1544856252164877, "grad_norm": 0.0, "learning_rate": 2.231162433702368e-05, "loss": 0.2347, "step": 6220 }, { "epoch": 2.1579494284724627, "grad_norm": 0.0, "learning_rate": 2.2144033828396983e-05, "loss": 0.2476, "step": 6230 }, { "epoch": 2.161413231728438, "grad_norm": 0.0, "learning_rate": 2.197689590104452e-05, "loss": 0.2682, "step": 6240 }, { "epoch": 2.164877034984413, "grad_norm": 0.0, "learning_rate": 2.1810213270488512e-05, "loss": 0.2644, "step": 6250 }, { "epoch": 2.168340838240388, "grad_norm": 0.0, "learning_rate": 2.1643988644853957e-05, "loss": 0.3792, "step": 6260 }, { "epoch": 2.171804641496363, "grad_norm": 0.0, "learning_rate": 2.1478224724824487e-05, "loss": 0.331, "step": 6270 }, { "epoch": 2.175268444752338, "grad_norm": 0.0, "learning_rate": 2.1312924203598615e-05, "loss": 0.316, "step": 6280 }, { "epoch": 2.178732248008313, "grad_norm": 0.0, "learning_rate": 2.1148089766845925e-05, "loss": 0.2516, "step": 6290 }, { "epoch": 2.182196051264288, "grad_norm": 0.0, "learning_rate": 2.0983724092663398e-05, "loss": 0.3641, "step": 6300 }, { "epoch": 2.1856598545202632, "grad_norm": 0.0, "learning_rate": 2.0819829851531935e-05, "loss": 0.2633, "step": 6310 }, { "epoch": 2.1891236577762383, "grad_norm": 0.0, "learning_rate": 2.0656409706273035e-05, "loss": 0.3458, "step": 6320 }, { "epoch": 2.1925874610322134, "grad_norm": 0.0, "learning_rate": 2.049346631200539e-05, "loss": 0.3168, "step": 6330 }, { "epoch": 2.1960512642881884, "grad_norm": 0.0, "learning_rate": 2.0331002316101906e-05, "loss": 0.2818, "step": 6340 }, { "epoch": 2.1995150675441635, "grad_norm": 0.0, "learning_rate": 2.0169020358146544e-05, "loss": 0.3056, "step": 6350 }, { "epoch": 2.2029788708001385, "grad_norm": 0.0, "learning_rate": 2.0007523069891525e-05, "loss": 0.2482, "step": 6360 }, { "epoch": 2.2064426740561136, "grad_norm": 0.0, "learning_rate": 1.9846513075214584e-05, "loss": 0.2682, "step": 6370 }, { "epoch": 2.2099064773120887, "grad_norm": 0.0, "learning_rate": 1.9685992990076225e-05, "loss": 0.2687, "step": 6380 }, { "epoch": 2.2133702805680637, "grad_norm": 0.0, "learning_rate": 1.9525965422477367e-05, "loss": 0.3366, "step": 6390 }, { "epoch": 2.216834083824039, "grad_norm": 0.0, "learning_rate": 1.9366432972416882e-05, "loss": 0.2806, "step": 6400 }, { "epoch": 2.220297887080014, "grad_norm": 0.0, "learning_rate": 1.920739823184935e-05, "loss": 0.2216, "step": 6410 }, { "epoch": 2.223761690335989, "grad_norm": 0.0, "learning_rate": 1.9048863784643e-05, "loss": 0.3032, "step": 6420 }, { "epoch": 2.227225493591964, "grad_norm": 0.0, "learning_rate": 1.8890832206537674e-05, "loss": 0.3079, "step": 6430 }, { "epoch": 2.230689296847939, "grad_norm": 0.0, "learning_rate": 1.8733306065102992e-05, "loss": 0.2703, "step": 6440 }, { "epoch": 2.234153100103914, "grad_norm": 0.0, "learning_rate": 1.85762879196967e-05, "loss": 0.3167, "step": 6450 }, { "epoch": 2.237616903359889, "grad_norm": 0.0, "learning_rate": 1.8419780321422957e-05, "loss": 0.3802, "step": 6460 }, { "epoch": 2.2410807066158642, "grad_norm": 0.0, "learning_rate": 1.8263785813091055e-05, "loss": 0.343, "step": 6470 }, { "epoch": 2.2445445098718393, "grad_norm": 0.0, "learning_rate": 1.8108306929173934e-05, "loss": 0.276, "step": 6480 }, { "epoch": 2.2480083131278144, "grad_norm": 0.0, "learning_rate": 1.79533461957671e-05, "loss": 0.2977, "step": 6490 }, { "epoch": 2.2514721163837894, "grad_norm": 0.0, "learning_rate": 1.779890613054761e-05, "loss": 0.2522, "step": 6500 }, { "epoch": 2.2549359196397645, "grad_norm": 0.0, "learning_rate": 1.764498924273311e-05, "loss": 0.2692, "step": 6510 }, { "epoch": 2.2583997228957395, "grad_norm": 0.0, "learning_rate": 1.749159803304103e-05, "loss": 0.2513, "step": 6520 }, { "epoch": 2.2618635261517146, "grad_norm": 0.0, "learning_rate": 1.7338734993648083e-05, "loss": 0.2398, "step": 6530 }, { "epoch": 2.2653273294076897, "grad_norm": 0.0, "learning_rate": 1.718640260814962e-05, "loss": 0.2691, "step": 6540 }, { "epoch": 2.2687911326636647, "grad_norm": 0.0, "learning_rate": 1.7034603351519428e-05, "loss": 0.364, "step": 6550 }, { "epoch": 2.27225493591964, "grad_norm": 0.0, "learning_rate": 1.6883339690069405e-05, "loss": 0.2313, "step": 6560 }, { "epoch": 2.275718739175615, "grad_norm": 0.0, "learning_rate": 1.673261408140951e-05, "loss": 0.2719, "step": 6570 }, { "epoch": 2.27918254243159, "grad_norm": 0.0, "learning_rate": 1.6582428974407914e-05, "loss": 0.2786, "step": 6580 }, { "epoch": 2.282646345687565, "grad_norm": 0.0, "learning_rate": 1.643278680915109e-05, "loss": 0.2373, "step": 6590 }, { "epoch": 2.28611014894354, "grad_norm": 0.0, "learning_rate": 1.62836900169043e-05, "loss": 0.285, "step": 6600 }, { "epoch": 2.289573952199515, "grad_norm": 0.0, "learning_rate": 1.6135141020071963e-05, "loss": 0.2329, "step": 6610 }, { "epoch": 2.29303775545549, "grad_norm": 0.0, "learning_rate": 1.5987142232158375e-05, "loss": 0.3377, "step": 6620 }, { "epoch": 2.296501558711465, "grad_norm": 0.0, "learning_rate": 1.5839696057728503e-05, "loss": 0.326, "step": 6630 }, { "epoch": 2.2999653619674403, "grad_norm": 0.0, "learning_rate": 1.5692804892368896e-05, "loss": 0.3309, "step": 6640 }, { "epoch": 2.3034291652234153, "grad_norm": 0.0, "learning_rate": 1.5546471122648716e-05, "loss": 0.34, "step": 6650 }, { "epoch": 2.3068929684793904, "grad_norm": 0.0, "learning_rate": 1.5400697126081083e-05, "loss": 0.3019, "step": 6660 }, { "epoch": 2.3103567717353655, "grad_norm": 0.0, "learning_rate": 1.5255485271084301e-05, "loss": 0.2761, "step": 6670 }, { "epoch": 2.3138205749913405, "grad_norm": 0.0, "learning_rate": 1.5110837916943515e-05, "loss": 0.2308, "step": 6680 }, { "epoch": 2.3172843782473156, "grad_norm": 0.0, "learning_rate": 1.4966757413772276e-05, "loss": 0.2906, "step": 6690 }, { "epoch": 2.3207481815032907, "grad_norm": 0.0, "learning_rate": 1.4823246102474386e-05, "loss": 0.3039, "step": 6700 }, { "epoch": 2.3242119847592657, "grad_norm": 0.0, "learning_rate": 1.4680306314705927e-05, "loss": 0.211, "step": 6710 }, { "epoch": 2.3276757880152408, "grad_norm": 0.0, "learning_rate": 1.4537940372837272e-05, "loss": 0.321, "step": 6720 }, { "epoch": 2.331139591271216, "grad_norm": 0.0, "learning_rate": 1.4396150589915469e-05, "loss": 0.2624, "step": 6730 }, { "epoch": 2.334603394527191, "grad_norm": 0.0, "learning_rate": 1.4254939269626528e-05, "loss": 0.2671, "step": 6740 }, { "epoch": 2.338067197783166, "grad_norm": 0.0, "learning_rate": 1.4114308706258133e-05, "loss": 0.292, "step": 6750 }, { "epoch": 2.341531001039141, "grad_norm": 0.0, "learning_rate": 1.3974261184662247e-05, "loss": 0.2576, "step": 6760 }, { "epoch": 2.344994804295116, "grad_norm": 0.0, "learning_rate": 1.3834798980218078e-05, "loss": 0.3099, "step": 6770 }, { "epoch": 2.348458607551091, "grad_norm": 0.0, "learning_rate": 1.3695924358795036e-05, "loss": 0.2718, "step": 6780 }, { "epoch": 2.351922410807066, "grad_norm": 0.0, "learning_rate": 1.3557639576716008e-05, "loss": 0.2976, "step": 6790 }, { "epoch": 2.3553862140630413, "grad_norm": 0.0, "learning_rate": 1.3419946880720573e-05, "loss": 0.3244, "step": 6800 }, { "epoch": 2.3588500173190163, "grad_norm": 0.0, "learning_rate": 1.3282848507928657e-05, "loss": 0.3077, "step": 6810 }, { "epoch": 2.3623138205749914, "grad_norm": 0.0, "learning_rate": 1.3146346685804035e-05, "loss": 0.2076, "step": 6820 }, { "epoch": 2.3657776238309665, "grad_norm": 0.0, "learning_rate": 1.3010443632118224e-05, "loss": 0.3079, "step": 6830 }, { "epoch": 2.3692414270869415, "grad_norm": 0.0, "learning_rate": 1.2875141554914472e-05, "loss": 0.3037, "step": 6840 }, { "epoch": 2.3727052303429166, "grad_norm": 0.0, "learning_rate": 1.2740442652471784e-05, "loss": 0.2709, "step": 6850 }, { "epoch": 2.3761690335988916, "grad_norm": 0.0, "learning_rate": 1.2606349113269327e-05, "loss": 0.2395, "step": 6860 }, { "epoch": 2.3796328368548667, "grad_norm": 0.0, "learning_rate": 1.2472863115950766e-05, "loss": 0.3051, "step": 6870 }, { "epoch": 2.3830966401108418, "grad_norm": 0.0, "learning_rate": 1.2339986829288953e-05, "loss": 0.3011, "step": 6880 }, { "epoch": 2.386560443366817, "grad_norm": 0.0, "learning_rate": 1.220772241215064e-05, "loss": 0.2533, "step": 6890 }, { "epoch": 2.390024246622792, "grad_norm": 0.0, "learning_rate": 1.207607201346141e-05, "loss": 0.3088, "step": 6900 }, { "epoch": 2.393488049878767, "grad_norm": 0.0, "learning_rate": 1.1945037772170754e-05, "loss": 0.2416, "step": 6910 }, { "epoch": 2.396951853134742, "grad_norm": 0.0, "learning_rate": 1.181462181721738e-05, "loss": 0.2488, "step": 6920 }, { "epoch": 2.400415656390717, "grad_norm": 0.0, "learning_rate": 1.1684826267494526e-05, "loss": 0.2502, "step": 6930 }, { "epoch": 2.403879459646692, "grad_norm": 0.0, "learning_rate": 1.1555653231815638e-05, "loss": 0.2367, "step": 6940 }, { "epoch": 2.407343262902667, "grad_norm": 0.0, "learning_rate": 1.1427104808880024e-05, "loss": 0.204, "step": 6950 }, { "epoch": 2.4108070661586423, "grad_norm": 0.0, "learning_rate": 1.1299183087238763e-05, "loss": 0.2504, "step": 6960 }, { "epoch": 2.4142708694146173, "grad_norm": 0.0, "learning_rate": 1.1171890145260872e-05, "loss": 0.2303, "step": 6970 }, { "epoch": 2.4177346726705924, "grad_norm": 0.0, "learning_rate": 1.1045228051099377e-05, "loss": 0.4305, "step": 6980 }, { "epoch": 2.4211984759265675, "grad_norm": 0.0, "learning_rate": 1.0919198862657864e-05, "loss": 0.2932, "step": 6990 }, { "epoch": 2.4246622791825425, "grad_norm": 0.0, "learning_rate": 1.0793804627556952e-05, "loss": 0.2584, "step": 7000 }, { "epoch": 2.4281260824385176, "grad_norm": 0.0, "learning_rate": 1.0669047383101038e-05, "loss": 0.3019, "step": 7010 }, { "epoch": 2.4315898856944926, "grad_norm": 0.0, "learning_rate": 1.0544929156245236e-05, "loss": 0.263, "step": 7020 }, { "epoch": 2.4350536889504677, "grad_norm": 0.0, "learning_rate": 1.0421451963562402e-05, "loss": 0.3046, "step": 7030 }, { "epoch": 2.4385174922064428, "grad_norm": 0.0, "learning_rate": 1.0298617811210353e-05, "loss": 0.3564, "step": 7040 }, { "epoch": 2.441981295462418, "grad_norm": 0.0, "learning_rate": 1.0176428694899382e-05, "loss": 0.2351, "step": 7050 }, { "epoch": 2.445445098718393, "grad_norm": 0.0, "learning_rate": 1.0054886599859681e-05, "loss": 0.3119, "step": 7060 }, { "epoch": 2.448908901974368, "grad_norm": 0.0, "learning_rate": 9.933993500809214e-06, "loss": 0.2378, "step": 7070 }, { "epoch": 2.452372705230343, "grad_norm": 0.0, "learning_rate": 9.813751361921557e-06, "loss": 0.2549, "step": 7080 }, { "epoch": 2.455836508486318, "grad_norm": 0.0, "learning_rate": 9.694162136793999e-06, "loss": 0.2446, "step": 7090 }, { "epoch": 2.459300311742293, "grad_norm": 0.0, "learning_rate": 9.57522776841585e-06, "loss": 0.2733, "step": 7100 }, { "epoch": 2.462764114998268, "grad_norm": 0.0, "learning_rate": 9.456950189136793e-06, "loss": 0.2452, "step": 7110 }, { "epoch": 2.4662279182542433, "grad_norm": 0.0, "learning_rate": 9.33933132063557e-06, "loss": 0.1984, "step": 7120 }, { "epoch": 2.4696917215102183, "grad_norm": 0.0, "learning_rate": 9.222373073888708e-06, "loss": 0.2598, "step": 7130 }, { "epoch": 2.4731555247661934, "grad_norm": 0.0, "learning_rate": 9.106077349139452e-06, "loss": 0.3156, "step": 7140 }, { "epoch": 2.4766193280221684, "grad_norm": 0.0, "learning_rate": 8.990446035866989e-06, "loss": 0.316, "step": 7150 }, { "epoch": 2.4800831312781435, "grad_norm": 0.0, "learning_rate": 8.875481012755626e-06, "loss": 0.2212, "step": 7160 }, { "epoch": 2.4835469345341186, "grad_norm": 0.0, "learning_rate": 8.761184147664343e-06, "loss": 0.2204, "step": 7170 }, { "epoch": 2.4870107377900936, "grad_norm": 0.0, "learning_rate": 8.647557297596443e-06, "loss": 0.3107, "step": 7180 }, { "epoch": 2.4904745410460687, "grad_norm": 0.0, "learning_rate": 8.534602308669338e-06, "loss": 0.2735, "step": 7190 }, { "epoch": 2.4939383443020438, "grad_norm": 0.0, "learning_rate": 8.422321016084617e-06, "loss": 0.2839, "step": 7200 }, { "epoch": 2.497402147558019, "grad_norm": 0.0, "learning_rate": 8.310715244098166e-06, "loss": 0.2642, "step": 7210 }, { "epoch": 2.500865950813994, "grad_norm": 0.0, "learning_rate": 8.199786805990544e-06, "loss": 0.2497, "step": 7220 }, { "epoch": 2.504329754069969, "grad_norm": 0.0, "learning_rate": 8.089537504037581e-06, "loss": 0.2627, "step": 7230 }, { "epoch": 2.507793557325944, "grad_norm": 0.0, "learning_rate": 7.979969129481014e-06, "loss": 0.2869, "step": 7240 }, { "epoch": 2.511257360581919, "grad_norm": 0.0, "learning_rate": 7.87108346249945e-06, "loss": 0.2666, "step": 7250 }, { "epoch": 2.514721163837894, "grad_norm": 0.0, "learning_rate": 7.762882272179385e-06, "loss": 0.2493, "step": 7260 }, { "epoch": 2.518184967093869, "grad_norm": 0.0, "learning_rate": 7.655367316486484e-06, "loss": 0.2073, "step": 7270 }, { "epoch": 2.5216487703498442, "grad_norm": 0.0, "learning_rate": 7.548540342237076e-06, "loss": 0.2418, "step": 7280 }, { "epoch": 2.5251125736058193, "grad_norm": 0.0, "learning_rate": 7.442403085069671e-06, "loss": 0.2827, "step": 7290 }, { "epoch": 2.5285763768617944, "grad_norm": 0.0, "learning_rate": 7.3369572694168166e-06, "loss": 0.2811, "step": 7300 }, { "epoch": 2.5320401801177694, "grad_norm": 0.0, "learning_rate": 7.232204608477122e-06, "loss": 0.227, "step": 7310 }, { "epoch": 2.5355039833737445, "grad_norm": 0.0, "learning_rate": 7.128146804187313e-06, "loss": 0.2362, "step": 7320 }, { "epoch": 2.5389677866297196, "grad_norm": 0.0, "learning_rate": 7.024785547194707e-06, "loss": 0.2706, "step": 7330 }, { "epoch": 2.5424315898856946, "grad_norm": 0.0, "learning_rate": 6.9221225168296265e-06, "loss": 0.2889, "step": 7340 }, { "epoch": 2.5458953931416697, "grad_norm": 0.0, "learning_rate": 6.820159381078223e-06, "loss": 0.255, "step": 7350 }, { "epoch": 2.5493591963976447, "grad_norm": 0.0, "learning_rate": 6.7188977965552735e-06, "loss": 0.3005, "step": 7360 }, { "epoch": 2.55282299965362, "grad_norm": 0.0, "learning_rate": 6.6183394084773535e-06, "loss": 0.2226, "step": 7370 }, { "epoch": 2.556286802909595, "grad_norm": 0.0, "learning_rate": 6.518485850636069e-06, "loss": 0.2919, "step": 7380 }, { "epoch": 2.55975060616557, "grad_norm": 0.0, "learning_rate": 6.419338745371495e-06, "loss": 0.2506, "step": 7390 }, { "epoch": 2.563214409421545, "grad_norm": 0.0, "learning_rate": 6.320899703545829e-06, "loss": 0.289, "step": 7400 }, { "epoch": 2.56667821267752, "grad_norm": 0.0, "learning_rate": 6.223170324517258e-06, "loss": 0.2467, "step": 7410 }, { "epoch": 2.570142015933495, "grad_norm": 0.0, "learning_rate": 6.126152196113899e-06, "loss": 0.3024, "step": 7420 }, { "epoch": 2.57360581918947, "grad_norm": 0.0, "learning_rate": 6.029846894608082e-06, "loss": 0.252, "step": 7430 }, { "epoch": 2.5770696224454452, "grad_norm": 0.0, "learning_rate": 5.934255984690673e-06, "loss": 0.2932, "step": 7440 }, { "epoch": 2.5805334257014203, "grad_norm": 0.0, "learning_rate": 5.839381019445678e-06, "loss": 0.2914, "step": 7450 }, { "epoch": 2.5839972289573954, "grad_norm": 0.0, "learning_rate": 5.7452235403250395e-06, "loss": 0.2566, "step": 7460 }, { "epoch": 2.5874610322133704, "grad_norm": 0.0, "learning_rate": 5.651785077123528e-06, "loss": 0.3103, "step": 7470 }, { "epoch": 2.5909248354693455, "grad_norm": 0.0, "learning_rate": 5.559067147953956e-06, "loss": 0.3098, "step": 7480 }, { "epoch": 2.5943886387253206, "grad_norm": 0.0, "learning_rate": 5.467071259222462e-06, "loss": 0.2428, "step": 7490 }, { "epoch": 2.5978524419812956, "grad_norm": 0.0, "learning_rate": 5.375798905604051e-06, "loss": 0.3563, "step": 7500 }, { "epoch": 2.6013162452372707, "grad_norm": 0.0, "learning_rate": 5.285251570018334e-06, "loss": 0.2741, "step": 7510 }, { "epoch": 2.6047800484932457, "grad_norm": 0.0, "learning_rate": 5.195430723605388e-06, "loss": 0.2481, "step": 7520 }, { "epoch": 2.608243851749221, "grad_norm": 0.0, "learning_rate": 5.106337825701879e-06, "loss": 0.2314, "step": 7530 }, { "epoch": 2.611707655005196, "grad_norm": 0.0, "learning_rate": 5.017974323817382e-06, "loss": 0.3023, "step": 7540 }, { "epoch": 2.615171458261171, "grad_norm": 0.0, "learning_rate": 4.930341653610804e-06, "loss": 0.4322, "step": 7550 }, { "epoch": 2.618635261517146, "grad_norm": 0.0, "learning_rate": 4.8434412388671135e-06, "loss": 0.246, "step": 7560 }, { "epoch": 2.622099064773121, "grad_norm": 0.0, "learning_rate": 4.7572744914741615e-06, "loss": 0.2804, "step": 7570 }, { "epoch": 2.625562868029096, "grad_norm": 0.0, "learning_rate": 4.6718428113997685e-06, "loss": 0.3123, "step": 7580 }, { "epoch": 2.629026671285071, "grad_norm": 0.0, "learning_rate": 4.5871475866689825e-06, "loss": 0.2304, "step": 7590 }, { "epoch": 2.6324904745410462, "grad_norm": 0.0, "learning_rate": 4.5031901933415e-06, "loss": 0.291, "step": 7600 }, { "epoch": 2.6359542777970213, "grad_norm": 0.0, "learning_rate": 4.419971995489352e-06, "loss": 0.2078, "step": 7610 }, { "epoch": 2.6394180810529964, "grad_norm": 0.0, "learning_rate": 4.33749434517472e-06, "loss": 0.278, "step": 7620 }, { "epoch": 2.642881884308971, "grad_norm": 0.0, "learning_rate": 4.255758582427932e-06, "loss": 0.2278, "step": 7630 }, { "epoch": 2.6463456875649465, "grad_norm": 0.0, "learning_rate": 4.174766035225758e-06, "loss": 0.2659, "step": 7640 }, { "epoch": 2.649809490820921, "grad_norm": 0.0, "learning_rate": 4.094518019469784e-06, "loss": 0.2692, "step": 7650 }, { "epoch": 2.6532732940768966, "grad_norm": 0.0, "learning_rate": 4.015015838965052e-06, "loss": 0.2434, "step": 7660 }, { "epoch": 2.6567370973328712, "grad_norm": 0.0, "learning_rate": 3.9362607853988935e-06, "loss": 0.2486, "step": 7670 }, { "epoch": 2.6602009005888467, "grad_norm": 0.0, "learning_rate": 3.858254138319878e-06, "loss": 0.2877, "step": 7680 }, { "epoch": 2.6636647038448213, "grad_norm": 0.0, "learning_rate": 3.7809971651171337e-06, "loss": 0.2563, "step": 7690 }, { "epoch": 2.667128507100797, "grad_norm": 0.0, "learning_rate": 3.7044911209996368e-06, "loss": 0.26, "step": 7700 }, { "epoch": 2.6705923103567715, "grad_norm": 0.0, "learning_rate": 3.6287372489758943e-06, "loss": 0.3145, "step": 7710 }, { "epoch": 2.674056113612747, "grad_norm": 0.0, "learning_rate": 3.5537367798337274e-06, "loss": 0.2902, "step": 7720 }, { "epoch": 2.6775199168687216, "grad_norm": 0.0, "learning_rate": 3.479490932120294e-06, "loss": 0.2476, "step": 7730 }, { "epoch": 2.680983720124697, "grad_norm": 0.0, "learning_rate": 3.4060009121222315e-06, "loss": 0.2856, "step": 7740 }, { "epoch": 2.6844475233806717, "grad_norm": 0.0, "learning_rate": 3.3332679138461388e-06, "loss": 0.2685, "step": 7750 }, { "epoch": 2.6879113266366472, "grad_norm": 0.0, "learning_rate": 3.261293118999098e-06, "loss": 0.2556, "step": 7760 }, { "epoch": 2.691375129892622, "grad_norm": 0.0, "learning_rate": 3.1900776969695454e-06, "loss": 0.2348, "step": 7770 }, { "epoch": 2.6948389331485973, "grad_norm": 0.0, "learning_rate": 3.1196228048082256e-06, "loss": 0.2288, "step": 7780 }, { "epoch": 2.698302736404572, "grad_norm": 0.0, "learning_rate": 3.0499295872093993e-06, "loss": 0.3252, "step": 7790 }, { "epoch": 2.7017665396605475, "grad_norm": 0.0, "learning_rate": 2.9809991764922772e-06, "loss": 0.2792, "step": 7800 }, { "epoch": 2.705230342916522, "grad_norm": 0.0, "learning_rate": 2.9128326925825675e-06, "loss": 0.2443, "step": 7810 }, { "epoch": 2.7086941461724976, "grad_norm": 0.0, "learning_rate": 2.845431242994351e-06, "loss": 0.3146, "step": 7820 }, { "epoch": 2.712157949428472, "grad_norm": 0.0, "learning_rate": 2.778795922812e-06, "loss": 0.252, "step": 7830 }, { "epoch": 2.7156217526844477, "grad_norm": 0.0, "learning_rate": 2.7129278146724834e-06, "loss": 0.2494, "step": 7840 }, { "epoch": 2.7190855559404223, "grad_norm": 0.0, "learning_rate": 2.647827988747681e-06, "loss": 0.2701, "step": 7850 }, { "epoch": 2.722549359196398, "grad_norm": 0.0, "learning_rate": 2.5834975027270837e-06, "loss": 0.3237, "step": 7860 }, { "epoch": 2.7260131624523725, "grad_norm": 0.0, "learning_rate": 2.5199374018005374e-06, "loss": 0.3471, "step": 7870 }, { "epoch": 2.729476965708348, "grad_norm": 0.0, "learning_rate": 2.4571487186413167e-06, "loss": 0.2665, "step": 7880 }, { "epoch": 2.7329407689643226, "grad_norm": 0.0, "learning_rate": 2.3951324733893e-06, "loss": 0.2607, "step": 7890 }, { "epoch": 2.736404572220298, "grad_norm": 0.0, "learning_rate": 2.333889673634432e-06, "loss": 0.2242, "step": 7900 }, { "epoch": 2.7398683754762727, "grad_norm": 0.0, "learning_rate": 2.2734213144003367e-06, "loss": 0.2484, "step": 7910 }, { "epoch": 2.743332178732248, "grad_norm": 0.0, "learning_rate": 2.2137283781281437e-06, "loss": 0.318, "step": 7920 }, { "epoch": 2.746795981988223, "grad_norm": 0.0, "learning_rate": 2.1548118346605528e-06, "loss": 0.2792, "step": 7930 }, { "epoch": 2.7502597852441983, "grad_norm": 0.0, "learning_rate": 2.0966726412260383e-06, "loss": 0.2829, "step": 7940 }, { "epoch": 2.753723588500173, "grad_norm": 0.0, "learning_rate": 2.0393117424233387e-06, "loss": 0.3143, "step": 7950 }, { "epoch": 2.7571873917561485, "grad_norm": 0.0, "learning_rate": 1.9827300702060735e-06, "loss": 0.2696, "step": 7960 }, { "epoch": 2.760651195012123, "grad_norm": 0.0, "learning_rate": 1.926928543867623e-06, "loss": 0.2244, "step": 7970 }, { "epoch": 2.7641149982680986, "grad_norm": 0.0, "learning_rate": 1.871908070026196e-06, "loss": 0.2952, "step": 7980 }, { "epoch": 2.767578801524073, "grad_norm": 0.0, "learning_rate": 1.8176695426100732e-06, "loss": 0.3776, "step": 7990 }, { "epoch": 2.7710426047800487, "grad_norm": 0.0, "learning_rate": 1.7642138428431044e-06, "loss": 0.2693, "step": 8000 }, { "epoch": 2.7745064080360233, "grad_norm": 0.0, "learning_rate": 1.7115418392304017e-06, "loss": 0.2936, "step": 8010 }, { "epoch": 2.777970211291999, "grad_norm": 0.0, "learning_rate": 1.659654387544196e-06, "loss": 0.2496, "step": 8020 }, { "epoch": 2.7814340145479735, "grad_norm": 0.0, "learning_rate": 1.6085523308099592e-06, "loss": 0.2628, "step": 8030 }, { "epoch": 2.784897817803949, "grad_norm": 0.0, "learning_rate": 1.5582364992926979e-06, "loss": 0.2327, "step": 8040 }, { "epoch": 2.7883616210599236, "grad_norm": 0.0, "learning_rate": 1.508707710483459e-06, "loss": 0.257, "step": 8050 }, { "epoch": 2.791825424315899, "grad_norm": 0.0, "learning_rate": 1.4599667690860796e-06, "loss": 0.3409, "step": 8060 }, { "epoch": 2.7952892275718737, "grad_norm": 0.0, "learning_rate": 1.412014467004047e-06, "loss": 0.225, "step": 8070 }, { "epoch": 2.798753030827849, "grad_norm": 0.0, "learning_rate": 1.3648515833277142e-06, "loss": 0.2617, "step": 8080 }, { "epoch": 2.802216834083824, "grad_norm": 0.0, "learning_rate": 1.3184788843215668e-06, "loss": 0.2781, "step": 8090 }, { "epoch": 2.8056806373397993, "grad_norm": 0.0, "learning_rate": 1.272897123411826e-06, "loss": 0.2863, "step": 8100 }, { "epoch": 2.809144440595774, "grad_norm": 0.0, "learning_rate": 1.2281070411741925e-06, "loss": 0.2508, "step": 8110 }, { "epoch": 2.8126082438517495, "grad_norm": 0.0, "learning_rate": 1.184109365321795e-06, "loss": 0.3322, "step": 8120 }, { "epoch": 2.816072047107724, "grad_norm": 0.0, "learning_rate": 1.140904810693383e-06, "loss": 0.2669, "step": 8130 }, { "epoch": 2.8195358503636996, "grad_norm": 0.0, "learning_rate": 1.098494079241724e-06, "loss": 0.2666, "step": 8140 }, { "epoch": 2.822999653619674, "grad_norm": 0.0, "learning_rate": 1.0568778600221818e-06, "loss": 0.2522, "step": 8150 }, { "epoch": 2.8264634568756497, "grad_norm": 0.0, "learning_rate": 1.016056829181533e-06, "loss": 0.2389, "step": 8160 }, { "epoch": 2.8299272601316243, "grad_norm": 0.0, "learning_rate": 9.760316499469568e-07, "loss": 0.2811, "step": 8170 }, { "epoch": 2.8333910633876, "grad_norm": 0.0, "learning_rate": 9.368029726152805e-07, "loss": 0.3168, "step": 8180 }, { "epoch": 2.8368548666435744, "grad_norm": 0.0, "learning_rate": 8.983714345424332e-07, "loss": 0.294, "step": 8190 }, { "epoch": 2.84031866989955, "grad_norm": 0.0, "learning_rate": 8.607376601330485e-07, "loss": 0.2502, "step": 8200 }, { "epoch": 2.8437824731555246, "grad_norm": 0.0, "learning_rate": 8.239022608303337e-07, "loss": 0.249, "step": 8210 }, { "epoch": 2.8472462764115, "grad_norm": 0.0, "learning_rate": 7.878658351061663e-07, "loss": 0.2778, "step": 8220 }, { "epoch": 2.8507100796674747, "grad_norm": 0.0, "learning_rate": 7.526289684513244e-07, "loss": 0.2887, "step": 8230 }, { "epoch": 2.85417388292345, "grad_norm": 0.0, "learning_rate": 7.181922333659996e-07, "loss": 0.2604, "step": 8240 }, { "epoch": 2.857637686179425, "grad_norm": 0.0, "learning_rate": 6.845561893504881e-07, "loss": 0.2574, "step": 8250 }, { "epoch": 2.8611014894354003, "grad_norm": 0.0, "learning_rate": 6.517213828961144e-07, "loss": 0.2609, "step": 8260 }, { "epoch": 2.864565292691375, "grad_norm": 0.0, "learning_rate": 6.196883474763382e-07, "loss": 0.3016, "step": 8270 }, { "epoch": 2.8680290959473504, "grad_norm": 0.0, "learning_rate": 5.884576035380729e-07, "loss": 0.2709, "step": 8280 }, { "epoch": 2.871492899203325, "grad_norm": 0.0, "learning_rate": 5.580296584932699e-07, "loss": 0.2204, "step": 8290 }, { "epoch": 2.8749567024593006, "grad_norm": 0.0, "learning_rate": 5.284050067106416e-07, "loss": 0.3067, "step": 8300 }, { "epoch": 2.878420505715275, "grad_norm": 0.0, "learning_rate": 4.995841295076353e-07, "loss": 0.2448, "step": 8310 }, { "epoch": 2.8818843089712507, "grad_norm": 0.0, "learning_rate": 4.7156749514262187e-07, "loss": 0.3258, "step": 8320 }, { "epoch": 2.8853481122272253, "grad_norm": 0.0, "learning_rate": 4.443555588072912e-07, "loss": 0.2463, "step": 8330 }, { "epoch": 2.888811915483201, "grad_norm": 0.0, "learning_rate": 4.1794876261923023e-07, "loss": 0.2953, "step": 8340 }, { "epoch": 2.8922757187391754, "grad_norm": 0.0, "learning_rate": 3.923475356147788e-07, "loss": 0.3699, "step": 8350 }, { "epoch": 2.895739521995151, "grad_norm": 0.0, "learning_rate": 3.675522937420128e-07, "loss": 0.253, "step": 8360 }, { "epoch": 2.8992033252511256, "grad_norm": 0.0, "learning_rate": 3.4356343985403883e-07, "loss": 0.3307, "step": 8370 }, { "epoch": 2.902667128507101, "grad_norm": 0.0, "learning_rate": 3.2038136370239893e-07, "loss": 0.3008, "step": 8380 }, { "epoch": 2.9061309317630757, "grad_norm": 0.0, "learning_rate": 2.9800644193078154e-07, "loss": 0.2559, "step": 8390 }, { "epoch": 2.909594735019051, "grad_norm": 0.0, "learning_rate": 2.7643903806886527e-07, "loss": 0.2336, "step": 8400 }, { "epoch": 2.913058538275026, "grad_norm": 0.0, "learning_rate": 2.5567950252644026e-07, "loss": 0.2601, "step": 8410 }, { "epoch": 2.9165223415310013, "grad_norm": 0.0, "learning_rate": 2.357281725876903e-07, "loss": 0.2399, "step": 8420 }, { "epoch": 2.919986144786976, "grad_norm": 0.0, "learning_rate": 2.165853724057476e-07, "loss": 0.2663, "step": 8430 }, { "epoch": 2.923449948042951, "grad_norm": 0.0, "learning_rate": 1.982514129973745e-07, "loss": 0.3308, "step": 8440 }, { "epoch": 2.926913751298926, "grad_norm": 0.0, "learning_rate": 1.8072659223797306e-07, "loss": 0.2399, "step": 8450 }, { "epoch": 2.930377554554901, "grad_norm": 0.0, "learning_rate": 1.640111948566947e-07, "loss": 0.2523, "step": 8460 }, { "epoch": 2.933841357810876, "grad_norm": 0.0, "learning_rate": 1.481054924318326e-07, "loss": 0.2828, "step": 8470 }, { "epoch": 2.9373051610668512, "grad_norm": 0.0, "learning_rate": 1.3300974338641415e-07, "loss": 0.2055, "step": 8480 }, { "epoch": 2.9407689643228263, "grad_norm": 0.0, "learning_rate": 1.1872419298399884e-07, "loss": 0.2327, "step": 8490 }, { "epoch": 2.9442327675788014, "grad_norm": 0.0, "learning_rate": 1.0524907332468692e-07, "loss": 0.3232, "step": 8500 }, { "epoch": 2.9476965708347764, "grad_norm": 0.0, "learning_rate": 9.258460334135577e-08, "loss": 0.247, "step": 8510 }, { "epoch": 2.9511603740907515, "grad_norm": 0.0, "learning_rate": 8.073098879610163e-08, "loss": 0.3356, "step": 8520 }, { "epoch": 2.9546241773467266, "grad_norm": 0.0, "learning_rate": 6.968842227689232e-08, "loss": 0.3203, "step": 8530 }, { "epoch": 2.9580879806027016, "grad_norm": 0.0, "learning_rate": 5.9457083194441877e-08, "loss": 0.2883, "step": 8540 }, { "epoch": 2.9615517838586767, "grad_norm": 0.0, "learning_rate": 5.003713777930741e-08, "loss": 0.2537, "step": 8550 }, { "epoch": 2.9650155871146517, "grad_norm": 0.0, "learning_rate": 4.142873907915234e-08, "loss": 0.2455, "step": 8560 }, { "epoch": 2.968479390370627, "grad_norm": 0.0, "learning_rate": 3.363202695629841e-08, "loss": 0.3041, "step": 8570 }, { "epoch": 2.971943193626602, "grad_norm": 0.0, "learning_rate": 2.6647128085444116e-08, "loss": 0.3571, "step": 8580 }, { "epoch": 2.975406996882577, "grad_norm": 0.0, "learning_rate": 2.0474155951588635e-08, "loss": 0.3192, "step": 8590 }, { "epoch": 2.978870800138552, "grad_norm": 0.0, "learning_rate": 1.5113210848211046e-08, "loss": 0.2552, "step": 8600 }, { "epoch": 2.982334603394527, "grad_norm": 0.0, "learning_rate": 1.0564379875610541e-08, "loss": 0.2826, "step": 8610 }, { "epoch": 2.985798406650502, "grad_norm": 0.0, "learning_rate": 6.827736939540863e-09, "loss": 0.2664, "step": 8620 }, { "epoch": 2.989262209906477, "grad_norm": 0.0, "learning_rate": 3.903342749950189e-09, "loss": 0.2982, "step": 8630 }, { "epoch": 2.9927260131624522, "grad_norm": 0.0, "learning_rate": 1.7912448200430031e-09, "loss": 0.3279, "step": 8640 }, { "epoch": 2.9961898164184273, "grad_norm": 0.0, "learning_rate": 4.914774654807275e-10, "loss": 0.3187, "step": 8650 }, { "epoch": 2.9996536196744024, "grad_norm": 0.0, "learning_rate": 4.061803832167144e-12, "loss": 0.2699, "step": 8660 }, { "epoch": 3.0, "step": 8661, "total_flos": 3.120007324041216e+18, "train_loss": 0.4153889326402145, "train_runtime": 17073.3476, "train_samples_per_second": 0.507, "train_steps_per_second": 0.507 } ], "logging_steps": 10, "max_steps": 8661, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.120007324041216e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }