{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.329838575470128, "eval_steps": 500, "global_step": 3500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0006656681644200366, "grad_norm": NaN, "learning_rate": 0.0, "loss": 1.9001, "step": 1 }, { "epoch": 0.0013313363288400732, "grad_norm": NaN, "learning_rate": 0.0, "loss": 1.1836, "step": 2 }, { "epoch": 0.00199700449326011, "grad_norm": Infinity, "learning_rate": 0.0, "loss": 2.0432, "step": 3 }, { "epoch": 0.0026626726576801465, "grad_norm": Infinity, "learning_rate": 0.0, "loss": 2.7364, "step": 4 }, { "epoch": 0.003328340822100183, "grad_norm": 305.05010986328125, "learning_rate": 0.0001, "loss": 2.7278, "step": 5 }, { "epoch": 0.00399400898652022, "grad_norm": 195.2991180419922, "learning_rate": 0.0002, "loss": 1.5935, "step": 6 }, { "epoch": 0.004659677150940256, "grad_norm": 135.4779815673828, "learning_rate": 0.00019995559502664298, "loss": 1.558, "step": 7 }, { "epoch": 0.005325345315360293, "grad_norm": 175.23486328125, "learning_rate": 0.00019991119005328598, "loss": 1.8278, "step": 8 }, { "epoch": 0.005991013479780329, "grad_norm": 70.18768310546875, "learning_rate": 0.00019986678507992895, "loss": 1.2821, "step": 9 }, { "epoch": 0.006656681644200366, "grad_norm": 91.9547348022461, "learning_rate": 0.00019982238010657195, "loss": 1.3009, "step": 10 }, { "epoch": 0.007322349808620403, "grad_norm": 140.80667114257812, "learning_rate": 0.00019977797513321492, "loss": 1.4739, "step": 11 }, { "epoch": 0.00798801797304044, "grad_norm": 117.13510131835938, "learning_rate": 0.00019973357015985792, "loss": 1.0922, "step": 12 }, { "epoch": 0.008653686137460476, "grad_norm": 87.54082489013672, "learning_rate": 0.0001996891651865009, "loss": 1.6633, "step": 13 }, { "epoch": 0.009319354301880512, "grad_norm": 178.7975616455078, "learning_rate": 0.0001996447602131439, "loss": 1.7614, "step": 14 }, { "epoch": 0.00998502246630055, "grad_norm": 199.108154296875, "learning_rate": 0.00019960035523978686, "loss": 1.4873, "step": 15 }, { "epoch": 0.010650690630720586, "grad_norm": 97.36341857910156, "learning_rate": 0.00019955595026642986, "loss": 1.2921, "step": 16 }, { "epoch": 0.011316358795140622, "grad_norm": 145.5021209716797, "learning_rate": 0.00019951154529307283, "loss": 1.4667, "step": 17 }, { "epoch": 0.011982026959560658, "grad_norm": 163.09628295898438, "learning_rate": 0.00019946714031971583, "loss": 1.4959, "step": 18 }, { "epoch": 0.012647695123980696, "grad_norm": 181.53123474121094, "learning_rate": 0.0001994227353463588, "loss": 1.4931, "step": 19 }, { "epoch": 0.013313363288400732, "grad_norm": 134.01087951660156, "learning_rate": 0.00019937833037300177, "loss": 1.1797, "step": 20 }, { "epoch": 0.013979031452820768, "grad_norm": 138.4213409423828, "learning_rate": 0.00019933392539964477, "loss": 1.3021, "step": 21 }, { "epoch": 0.014644699617240806, "grad_norm": 94.12645721435547, "learning_rate": 0.00019928952042628774, "loss": 1.2099, "step": 22 }, { "epoch": 0.015310367781660842, "grad_norm": 68.36018371582031, "learning_rate": 0.00019924511545293074, "loss": 1.0593, "step": 23 }, { "epoch": 0.01597603594608088, "grad_norm": 83.23944854736328, "learning_rate": 0.0001992007104795737, "loss": 1.1152, "step": 24 }, { "epoch": 0.016641704110500914, "grad_norm": 145.29953002929688, "learning_rate": 0.0001991563055062167, "loss": 1.194, "step": 25 }, { "epoch": 0.017307372274920952, "grad_norm": 71.01487731933594, "learning_rate": 0.00019911190053285968, "loss": 1.0201, "step": 26 }, { "epoch": 0.01797304043934099, "grad_norm": 82.99989318847656, "learning_rate": 0.00019906749555950268, "loss": 1.3319, "step": 27 }, { "epoch": 0.018638708603761024, "grad_norm": 51.65462112426758, "learning_rate": 0.00019902309058614565, "loss": 0.8622, "step": 28 }, { "epoch": 0.019304376768181062, "grad_norm": 106.69747161865234, "learning_rate": 0.00019897868561278865, "loss": 1.0051, "step": 29 }, { "epoch": 0.0199700449326011, "grad_norm": 85.53784942626953, "learning_rate": 0.00019893428063943162, "loss": 0.7687, "step": 30 }, { "epoch": 0.020635713097021134, "grad_norm": 141.33482360839844, "learning_rate": 0.00019888987566607462, "loss": 0.9697, "step": 31 }, { "epoch": 0.021301381261441172, "grad_norm": 110.15571594238281, "learning_rate": 0.0001988454706927176, "loss": 0.6955, "step": 32 }, { "epoch": 0.02196704942586121, "grad_norm": 123.28990173339844, "learning_rate": 0.00019880106571936056, "loss": 0.9405, "step": 33 }, { "epoch": 0.022632717590281244, "grad_norm": 104.15376281738281, "learning_rate": 0.00019875666074600356, "loss": 1.1353, "step": 34 }, { "epoch": 0.023298385754701282, "grad_norm": 122.83297729492188, "learning_rate": 0.00019871225577264653, "loss": 0.9912, "step": 35 }, { "epoch": 0.023964053919121316, "grad_norm": 137.1550750732422, "learning_rate": 0.00019866785079928953, "loss": 1.0156, "step": 36 }, { "epoch": 0.024629722083541354, "grad_norm": 95.7591781616211, "learning_rate": 0.0001986234458259325, "loss": 0.9295, "step": 37 }, { "epoch": 0.025295390247961392, "grad_norm": 163.35743713378906, "learning_rate": 0.0001985790408525755, "loss": 0.9697, "step": 38 }, { "epoch": 0.025961058412381426, "grad_norm": 42.57448959350586, "learning_rate": 0.00019853463587921847, "loss": 0.8329, "step": 39 }, { "epoch": 0.026626726576801464, "grad_norm": 161.05844116210938, "learning_rate": 0.00019849023090586147, "loss": 0.8588, "step": 40 }, { "epoch": 0.027292394741221502, "grad_norm": 123.78165435791016, "learning_rate": 0.00019844582593250444, "loss": 0.8082, "step": 41 }, { "epoch": 0.027958062905641536, "grad_norm": 117.62382507324219, "learning_rate": 0.00019840142095914744, "loss": 1.0692, "step": 42 }, { "epoch": 0.028623731070061574, "grad_norm": NaN, "learning_rate": 0.00019840142095914744, "loss": 1.316, "step": 43 }, { "epoch": 0.029289399234481612, "grad_norm": 63.35919952392578, "learning_rate": 0.0001983570159857904, "loss": 0.6028, "step": 44 }, { "epoch": 0.029955067398901646, "grad_norm": 277.1460266113281, "learning_rate": 0.00019831261101243338, "loss": 1.4349, "step": 45 }, { "epoch": 0.030620735563321684, "grad_norm": NaN, "learning_rate": 0.00019831261101243338, "loss": 1.816, "step": 46 }, { "epoch": 0.03128640372774172, "grad_norm": 206.0849151611328, "learning_rate": 0.00019826820603907638, "loss": 1.7035, "step": 47 }, { "epoch": 0.03195207189216176, "grad_norm": 154.10183715820312, "learning_rate": 0.00019822380106571935, "loss": 0.7897, "step": 48 }, { "epoch": 0.032617740056581794, "grad_norm": 182.32650756835938, "learning_rate": 0.00019817939609236235, "loss": 1.2611, "step": 49 }, { "epoch": 0.03328340822100183, "grad_norm": 100.30379486083984, "learning_rate": 0.00019813499111900532, "loss": 0.8746, "step": 50 }, { "epoch": 0.03394907638542187, "grad_norm": 99.05520629882812, "learning_rate": 0.00019809058614564832, "loss": 0.6544, "step": 51 }, { "epoch": 0.034614744549841904, "grad_norm": 75.6047592163086, "learning_rate": 0.0001980461811722913, "loss": 0.5308, "step": 52 }, { "epoch": 0.03528041271426194, "grad_norm": 81.81549835205078, "learning_rate": 0.0001980017761989343, "loss": 1.0103, "step": 53 }, { "epoch": 0.03594608087868198, "grad_norm": 305.5655517578125, "learning_rate": 0.00019795737122557726, "loss": 2.0811, "step": 54 }, { "epoch": 0.036611749043102014, "grad_norm": 254.49461364746094, "learning_rate": 0.00019791296625222026, "loss": 1.5316, "step": 55 }, { "epoch": 0.03727741720752205, "grad_norm": 249.2161407470703, "learning_rate": 0.00019786856127886323, "loss": 1.4851, "step": 56 }, { "epoch": 0.03794308537194209, "grad_norm": 187.69805908203125, "learning_rate": 0.00019782415630550623, "loss": 1.0635, "step": 57 }, { "epoch": 0.038608753536362124, "grad_norm": 135.42141723632812, "learning_rate": 0.0001977797513321492, "loss": 0.8954, "step": 58 }, { "epoch": 0.03927442170078216, "grad_norm": 198.7463836669922, "learning_rate": 0.0001977353463587922, "loss": 1.2765, "step": 59 }, { "epoch": 0.0399400898652022, "grad_norm": 256.67279052734375, "learning_rate": 0.00019769094138543517, "loss": 1.4927, "step": 60 }, { "epoch": 0.040605758029622234, "grad_norm": 64.07623291015625, "learning_rate": 0.00019764653641207817, "loss": 0.7874, "step": 61 }, { "epoch": 0.04127142619404227, "grad_norm": 147.7261962890625, "learning_rate": 0.00019760213143872114, "loss": 0.7924, "step": 62 }, { "epoch": 0.04193709435846231, "grad_norm": 122.39730072021484, "learning_rate": 0.00019755772646536413, "loss": 0.7656, "step": 63 }, { "epoch": 0.042602762522882344, "grad_norm": 94.13729095458984, "learning_rate": 0.0001975133214920071, "loss": 0.9051, "step": 64 }, { "epoch": 0.04326843068730238, "grad_norm": 48.220184326171875, "learning_rate": 0.0001974689165186501, "loss": 0.6262, "step": 65 }, { "epoch": 0.04393409885172242, "grad_norm": 32.33949661254883, "learning_rate": 0.00019742451154529308, "loss": 0.544, "step": 66 }, { "epoch": 0.044599767016142454, "grad_norm": 133.78097534179688, "learning_rate": 0.00019738010657193607, "loss": 0.6472, "step": 67 }, { "epoch": 0.04526543518056249, "grad_norm": 105.52882385253906, "learning_rate": 0.00019733570159857907, "loss": 0.7272, "step": 68 }, { "epoch": 0.04593110334498253, "grad_norm": 40.91939926147461, "learning_rate": 0.00019729129662522204, "loss": 0.6232, "step": 69 }, { "epoch": 0.046596771509402564, "grad_norm": 48.693763732910156, "learning_rate": 0.00019724689165186504, "loss": 0.3698, "step": 70 }, { "epoch": 0.0472624396738226, "grad_norm": 98.95121765136719, "learning_rate": 0.00019720248667850801, "loss": 0.4914, "step": 71 }, { "epoch": 0.04792810783824263, "grad_norm": 217.00401306152344, "learning_rate": 0.00019715808170515098, "loss": 1.3837, "step": 72 }, { "epoch": 0.048593776002662674, "grad_norm": 97.31861877441406, "learning_rate": 0.00019711367673179398, "loss": 0.7037, "step": 73 }, { "epoch": 0.04925944416708271, "grad_norm": 101.28430938720703, "learning_rate": 0.00019706927175843695, "loss": 1.3999, "step": 74 }, { "epoch": 0.04992511233150274, "grad_norm": 86.35800170898438, "learning_rate": 0.00019702486678507995, "loss": 0.8045, "step": 75 }, { "epoch": 0.050590780495922784, "grad_norm": 102.23155212402344, "learning_rate": 0.00019698046181172292, "loss": 0.7521, "step": 76 }, { "epoch": 0.05125644866034282, "grad_norm": 125.80261993408203, "learning_rate": 0.00019693605683836592, "loss": 0.7933, "step": 77 }, { "epoch": 0.05192211682476285, "grad_norm": 160.28123474121094, "learning_rate": 0.0001968916518650089, "loss": 0.6355, "step": 78 }, { "epoch": 0.052587784989182894, "grad_norm": 42.919464111328125, "learning_rate": 0.0001968472468916519, "loss": 0.2693, "step": 79 }, { "epoch": 0.05325345315360293, "grad_norm": 56.61810302734375, "learning_rate": 0.00019680284191829486, "loss": 0.3141, "step": 80 }, { "epoch": 0.05391912131802296, "grad_norm": 299.3166198730469, "learning_rate": 0.00019675843694493786, "loss": 1.7119, "step": 81 }, { "epoch": 0.054584789482443004, "grad_norm": 133.17459106445312, "learning_rate": 0.00019671403197158083, "loss": 1.2885, "step": 82 }, { "epoch": 0.05525045764686304, "grad_norm": 131.6356964111328, "learning_rate": 0.00019666962699822383, "loss": 0.7976, "step": 83 }, { "epoch": 0.05591612581128307, "grad_norm": 122.43846130371094, "learning_rate": 0.0001966252220248668, "loss": 0.7031, "step": 84 }, { "epoch": 0.056581793975703114, "grad_norm": 51.365901947021484, "learning_rate": 0.00019658081705150977, "loss": 0.6684, "step": 85 }, { "epoch": 0.05724746214012315, "grad_norm": 251.70664978027344, "learning_rate": 0.00019653641207815277, "loss": 1.2329, "step": 86 }, { "epoch": 0.05791313030454318, "grad_norm": 94.94196319580078, "learning_rate": 0.00019649200710479574, "loss": 0.746, "step": 87 }, { "epoch": 0.058578798468963224, "grad_norm": 106.89917755126953, "learning_rate": 0.00019644760213143874, "loss": 0.8735, "step": 88 }, { "epoch": 0.05924446663338326, "grad_norm": 130.65577697753906, "learning_rate": 0.00019640319715808171, "loss": 1.0905, "step": 89 }, { "epoch": 0.05991013479780329, "grad_norm": 37.6524543762207, "learning_rate": 0.0001963587921847247, "loss": 0.447, "step": 90 }, { "epoch": 0.060575802962223334, "grad_norm": 55.81675720214844, "learning_rate": 0.00019631438721136768, "loss": 0.8305, "step": 91 }, { "epoch": 0.06124147112664337, "grad_norm": 56.32381057739258, "learning_rate": 0.00019626998223801068, "loss": 0.6437, "step": 92 }, { "epoch": 0.0619071392910634, "grad_norm": 107.72248840332031, "learning_rate": 0.00019622557726465365, "loss": 0.7331, "step": 93 }, { "epoch": 0.06257280745548344, "grad_norm": 60.965084075927734, "learning_rate": 0.00019618117229129665, "loss": 0.1975, "step": 94 }, { "epoch": 0.06323847561990348, "grad_norm": 236.83071899414062, "learning_rate": 0.00019613676731793962, "loss": 1.3465, "step": 95 }, { "epoch": 0.06390414378432352, "grad_norm": 32.12284851074219, "learning_rate": 0.0001960923623445826, "loss": 0.4818, "step": 96 }, { "epoch": 0.06456981194874355, "grad_norm": 139.45631408691406, "learning_rate": 0.0001960479573712256, "loss": 0.7907, "step": 97 }, { "epoch": 0.06523548011316359, "grad_norm": 85.41361999511719, "learning_rate": 0.00019600355239786856, "loss": 0.5648, "step": 98 }, { "epoch": 0.06590114827758363, "grad_norm": 79.85189056396484, "learning_rate": 0.00019595914742451156, "loss": 0.7735, "step": 99 }, { "epoch": 0.06656681644200366, "grad_norm": 64.68962860107422, "learning_rate": 0.00019591474245115453, "loss": 0.4128, "step": 100 }, { "epoch": 0.0672324846064237, "grad_norm": 58.707054138183594, "learning_rate": 0.00019587033747779753, "loss": 0.8143, "step": 101 }, { "epoch": 0.06789815277084374, "grad_norm": 107.83698272705078, "learning_rate": 0.0001958259325044405, "loss": 0.8662, "step": 102 }, { "epoch": 0.06856382093526377, "grad_norm": 221.64857482910156, "learning_rate": 0.0001957815275310835, "loss": 1.2667, "step": 103 }, { "epoch": 0.06922948909968381, "grad_norm": 75.43016815185547, "learning_rate": 0.00019573712255772647, "loss": 0.6056, "step": 104 }, { "epoch": 0.06989515726410385, "grad_norm": 81.48358917236328, "learning_rate": 0.00019569271758436947, "loss": 0.4954, "step": 105 }, { "epoch": 0.07056082542852388, "grad_norm": 100.88373565673828, "learning_rate": 0.00019564831261101244, "loss": 0.8129, "step": 106 }, { "epoch": 0.07122649359294392, "grad_norm": 56.06926727294922, "learning_rate": 0.00019560390763765544, "loss": 0.5694, "step": 107 }, { "epoch": 0.07189216175736396, "grad_norm": 119.89936828613281, "learning_rate": 0.0001955595026642984, "loss": 0.8492, "step": 108 }, { "epoch": 0.07255782992178399, "grad_norm": 226.14344787597656, "learning_rate": 0.00019551509769094138, "loss": 1.5915, "step": 109 }, { "epoch": 0.07322349808620403, "grad_norm": 167.6849365234375, "learning_rate": 0.00019547069271758438, "loss": 0.8734, "step": 110 }, { "epoch": 0.07388916625062407, "grad_norm": 191.29388427734375, "learning_rate": 0.00019542628774422735, "loss": 1.1915, "step": 111 }, { "epoch": 0.0745548344150441, "grad_norm": 74.29916381835938, "learning_rate": 0.00019538188277087035, "loss": 0.399, "step": 112 }, { "epoch": 0.07522050257946414, "grad_norm": 193.20899963378906, "learning_rate": 0.00019533747779751332, "loss": 1.0812, "step": 113 }, { "epoch": 0.07588617074388418, "grad_norm": 104.37834930419922, "learning_rate": 0.00019529307282415632, "loss": 0.7821, "step": 114 }, { "epoch": 0.0765518389083042, "grad_norm": 63.434146881103516, "learning_rate": 0.0001952486678507993, "loss": 0.5636, "step": 115 }, { "epoch": 0.07721750707272425, "grad_norm": 163.28578186035156, "learning_rate": 0.0001952042628774423, "loss": 0.8423, "step": 116 }, { "epoch": 0.07788317523714429, "grad_norm": 94.21780395507812, "learning_rate": 0.00019515985790408526, "loss": 0.6704, "step": 117 }, { "epoch": 0.07854884340156432, "grad_norm": 97.0186767578125, "learning_rate": 0.00019511545293072826, "loss": 0.4906, "step": 118 }, { "epoch": 0.07921451156598436, "grad_norm": 67.47974395751953, "learning_rate": 0.00019507104795737123, "loss": 0.5328, "step": 119 }, { "epoch": 0.0798801797304044, "grad_norm": 113.68756103515625, "learning_rate": 0.00019502664298401423, "loss": 0.7459, "step": 120 }, { "epoch": 0.08054584789482443, "grad_norm": 71.1500244140625, "learning_rate": 0.0001949822380106572, "loss": 0.6177, "step": 121 }, { "epoch": 0.08121151605924447, "grad_norm": 131.37168884277344, "learning_rate": 0.00019493783303730017, "loss": 0.7955, "step": 122 }, { "epoch": 0.08187718422366451, "grad_norm": 72.7484359741211, "learning_rate": 0.00019489342806394317, "loss": 0.7026, "step": 123 }, { "epoch": 0.08254285238808454, "grad_norm": 34.95873260498047, "learning_rate": 0.00019484902309058614, "loss": 0.3607, "step": 124 }, { "epoch": 0.08320852055250458, "grad_norm": 23.322694778442383, "learning_rate": 0.00019480461811722914, "loss": 0.4757, "step": 125 }, { "epoch": 0.08387418871692462, "grad_norm": 102.11334228515625, "learning_rate": 0.0001947602131438721, "loss": 0.9689, "step": 126 }, { "epoch": 0.08453985688134465, "grad_norm": 70.20091247558594, "learning_rate": 0.0001947158081705151, "loss": 0.5523, "step": 127 }, { "epoch": 0.08520552504576469, "grad_norm": 108.84575653076172, "learning_rate": 0.00019467140319715808, "loss": 0.7221, "step": 128 }, { "epoch": 0.08587119321018473, "grad_norm": 74.28465270996094, "learning_rate": 0.00019462699822380108, "loss": 0.3408, "step": 129 }, { "epoch": 0.08653686137460476, "grad_norm": 45.54611587524414, "learning_rate": 0.00019458259325044405, "loss": 0.3064, "step": 130 }, { "epoch": 0.0872025295390248, "grad_norm": 107.5720443725586, "learning_rate": 0.00019453818827708705, "loss": 0.4285, "step": 131 }, { "epoch": 0.08786819770344484, "grad_norm": 113.14404296875, "learning_rate": 0.00019449378330373002, "loss": 0.7893, "step": 132 }, { "epoch": 0.08853386586786487, "grad_norm": 41.36758041381836, "learning_rate": 0.00019444937833037302, "loss": 0.2135, "step": 133 }, { "epoch": 0.08919953403228491, "grad_norm": 111.0009765625, "learning_rate": 0.000194404973357016, "loss": 0.4909, "step": 134 }, { "epoch": 0.08986520219670495, "grad_norm": 119.50690460205078, "learning_rate": 0.00019436056838365896, "loss": 0.8744, "step": 135 }, { "epoch": 0.09053087036112498, "grad_norm": 83.16583251953125, "learning_rate": 0.00019431616341030196, "loss": 0.5229, "step": 136 }, { "epoch": 0.09119653852554502, "grad_norm": 96.94984436035156, "learning_rate": 0.00019427175843694493, "loss": 0.6746, "step": 137 }, { "epoch": 0.09186220668996506, "grad_norm": 65.90393829345703, "learning_rate": 0.00019422735346358793, "loss": 0.3112, "step": 138 }, { "epoch": 0.09252787485438509, "grad_norm": 92.0575180053711, "learning_rate": 0.0001941829484902309, "loss": 0.5264, "step": 139 }, { "epoch": 0.09319354301880513, "grad_norm": 63.846744537353516, "learning_rate": 0.0001941385435168739, "loss": 0.8869, "step": 140 }, { "epoch": 0.09385921118322517, "grad_norm": 129.04086303710938, "learning_rate": 0.00019409413854351687, "loss": 0.7077, "step": 141 }, { "epoch": 0.0945248793476452, "grad_norm": 38.49999237060547, "learning_rate": 0.00019404973357015987, "loss": 0.1854, "step": 142 }, { "epoch": 0.09519054751206524, "grad_norm": 142.83619689941406, "learning_rate": 0.00019400532859680284, "loss": 1.3549, "step": 143 }, { "epoch": 0.09585621567648527, "grad_norm": 58.420223236083984, "learning_rate": 0.00019396092362344584, "loss": 0.3978, "step": 144 }, { "epoch": 0.0965218838409053, "grad_norm": 82.57951354980469, "learning_rate": 0.0001939165186500888, "loss": 0.9771, "step": 145 }, { "epoch": 0.09718755200532535, "grad_norm": 52.69244384765625, "learning_rate": 0.00019387211367673178, "loss": 0.389, "step": 146 }, { "epoch": 0.09785322016974538, "grad_norm": 99.57772827148438, "learning_rate": 0.00019382770870337478, "loss": 0.938, "step": 147 }, { "epoch": 0.09851888833416542, "grad_norm": 66.55644226074219, "learning_rate": 0.00019378330373001775, "loss": 0.4984, "step": 148 }, { "epoch": 0.09918455649858546, "grad_norm": 64.8500747680664, "learning_rate": 0.00019373889875666075, "loss": 0.4189, "step": 149 }, { "epoch": 0.09985022466300549, "grad_norm": 113.6292495727539, "learning_rate": 0.00019369449378330372, "loss": 0.486, "step": 150 }, { "epoch": 0.10051589282742553, "grad_norm": 83.64993286132812, "learning_rate": 0.00019365008880994672, "loss": 0.3288, "step": 151 }, { "epoch": 0.10118156099184557, "grad_norm": 70.0993423461914, "learning_rate": 0.0001936056838365897, "loss": 0.7637, "step": 152 }, { "epoch": 0.1018472291562656, "grad_norm": 68.05122375488281, "learning_rate": 0.0001935612788632327, "loss": 0.3255, "step": 153 }, { "epoch": 0.10251289732068564, "grad_norm": 56.31193923950195, "learning_rate": 0.00019351687388987566, "loss": 0.1978, "step": 154 }, { "epoch": 0.10317856548510568, "grad_norm": 131.5350799560547, "learning_rate": 0.00019347246891651866, "loss": 0.7034, "step": 155 }, { "epoch": 0.1038442336495257, "grad_norm": 143.9272003173828, "learning_rate": 0.00019342806394316163, "loss": 0.6122, "step": 156 }, { "epoch": 0.10450990181394575, "grad_norm": 56.10287094116211, "learning_rate": 0.00019338365896980463, "loss": 0.5982, "step": 157 }, { "epoch": 0.10517556997836579, "grad_norm": 113.03327941894531, "learning_rate": 0.0001933392539964476, "loss": 0.5578, "step": 158 }, { "epoch": 0.10584123814278582, "grad_norm": 156.54730224609375, "learning_rate": 0.0001932948490230906, "loss": 0.9944, "step": 159 }, { "epoch": 0.10650690630720586, "grad_norm": 117.35420989990234, "learning_rate": 0.00019325044404973357, "loss": 0.9612, "step": 160 }, { "epoch": 0.1071725744716259, "grad_norm": 137.27517700195312, "learning_rate": 0.00019320603907637657, "loss": 0.8892, "step": 161 }, { "epoch": 0.10783824263604593, "grad_norm": 121.9662094116211, "learning_rate": 0.00019316163410301954, "loss": 0.8541, "step": 162 }, { "epoch": 0.10850391080046597, "grad_norm": 50.96897506713867, "learning_rate": 0.00019311722912966254, "loss": 0.4626, "step": 163 }, { "epoch": 0.10916957896488601, "grad_norm": 86.26678466796875, "learning_rate": 0.0001930728241563055, "loss": 0.4941, "step": 164 }, { "epoch": 0.10983524712930604, "grad_norm": 92.44398498535156, "learning_rate": 0.0001930284191829485, "loss": 0.6016, "step": 165 }, { "epoch": 0.11050091529372608, "grad_norm": 76.84557342529297, "learning_rate": 0.00019298401420959148, "loss": 0.465, "step": 166 }, { "epoch": 0.11116658345814612, "grad_norm": 120.3193588256836, "learning_rate": 0.00019293960923623448, "loss": 0.7597, "step": 167 }, { "epoch": 0.11183225162256615, "grad_norm": 98.53234100341797, "learning_rate": 0.00019289520426287745, "loss": 0.4161, "step": 168 }, { "epoch": 0.11249791978698619, "grad_norm": 136.8874053955078, "learning_rate": 0.00019285079928952045, "loss": 1.0431, "step": 169 }, { "epoch": 0.11316358795140623, "grad_norm": 145.96075439453125, "learning_rate": 0.00019280639431616342, "loss": 0.8223, "step": 170 }, { "epoch": 0.11382925611582626, "grad_norm": 95.09919738769531, "learning_rate": 0.00019276198934280642, "loss": 0.4007, "step": 171 }, { "epoch": 0.1144949242802463, "grad_norm": 95.6208267211914, "learning_rate": 0.0001927175843694494, "loss": 0.4496, "step": 172 }, { "epoch": 0.11516059244466634, "grad_norm": 131.4598846435547, "learning_rate": 0.00019267317939609239, "loss": 0.7203, "step": 173 }, { "epoch": 0.11582626060908637, "grad_norm": 183.2200469970703, "learning_rate": 0.00019262877442273536, "loss": 0.8879, "step": 174 }, { "epoch": 0.1164919287735064, "grad_norm": 65.30191802978516, "learning_rate": 0.00019258436944937836, "loss": 0.2161, "step": 175 }, { "epoch": 0.11715759693792645, "grad_norm": 100.97784423828125, "learning_rate": 0.00019253996447602133, "loss": 0.8095, "step": 176 }, { "epoch": 0.11782326510234647, "grad_norm": 57.705963134765625, "learning_rate": 0.00019249555950266432, "loss": 0.3389, "step": 177 }, { "epoch": 0.11848893326676652, "grad_norm": 119.22857666015625, "learning_rate": 0.0001924511545293073, "loss": 0.5287, "step": 178 }, { "epoch": 0.11915460143118656, "grad_norm": 97.59465026855469, "learning_rate": 0.0001924067495559503, "loss": 0.4567, "step": 179 }, { "epoch": 0.11982026959560658, "grad_norm": 47.80794906616211, "learning_rate": 0.00019236234458259327, "loss": 0.3795, "step": 180 }, { "epoch": 0.12048593776002663, "grad_norm": 93.19955444335938, "learning_rate": 0.00019231793960923626, "loss": 1.0376, "step": 181 }, { "epoch": 0.12115160592444667, "grad_norm": 57.24619674682617, "learning_rate": 0.00019227353463587924, "loss": 0.3167, "step": 182 }, { "epoch": 0.1218172740888667, "grad_norm": 71.7004165649414, "learning_rate": 0.00019222912966252223, "loss": 0.3515, "step": 183 }, { "epoch": 0.12248294225328674, "grad_norm": 42.93682098388672, "learning_rate": 0.0001921847246891652, "loss": 0.3272, "step": 184 }, { "epoch": 0.12314861041770678, "grad_norm": 88.22802734375, "learning_rate": 0.00019214031971580818, "loss": 0.5751, "step": 185 }, { "epoch": 0.1238142785821268, "grad_norm": 88.80204772949219, "learning_rate": 0.00019209591474245117, "loss": 0.5306, "step": 186 }, { "epoch": 0.12447994674654685, "grad_norm": 92.46751403808594, "learning_rate": 0.00019205150976909415, "loss": 0.4262, "step": 187 }, { "epoch": 0.12514561491096687, "grad_norm": 55.72347640991211, "learning_rate": 0.00019200710479573714, "loss": 0.3939, "step": 188 }, { "epoch": 0.12581128307538691, "grad_norm": 72.36625671386719, "learning_rate": 0.00019196269982238012, "loss": 0.3739, "step": 189 }, { "epoch": 0.12647695123980696, "grad_norm": 96.56300354003906, "learning_rate": 0.00019191829484902311, "loss": 0.4574, "step": 190 }, { "epoch": 0.127142619404227, "grad_norm": 122.78588104248047, "learning_rate": 0.00019187388987566609, "loss": 0.7646, "step": 191 }, { "epoch": 0.12780828756864704, "grad_norm": 74.9114990234375, "learning_rate": 0.00019182948490230908, "loss": 0.2156, "step": 192 }, { "epoch": 0.12847395573306705, "grad_norm": 61.17184829711914, "learning_rate": 0.00019178507992895206, "loss": 0.7067, "step": 193 }, { "epoch": 0.1291396238974871, "grad_norm": 120.3477783203125, "learning_rate": 0.00019174067495559505, "loss": 0.7562, "step": 194 }, { "epoch": 0.12980529206190713, "grad_norm": 204.0528564453125, "learning_rate": 0.00019169626998223802, "loss": 1.2893, "step": 195 }, { "epoch": 0.13047096022632718, "grad_norm": 91.55975341796875, "learning_rate": 0.000191651865008881, "loss": 1.025, "step": 196 }, { "epoch": 0.13113662839074722, "grad_norm": 111.15852355957031, "learning_rate": 0.000191607460035524, "loss": 0.4063, "step": 197 }, { "epoch": 0.13180229655516726, "grad_norm": 90.34394073486328, "learning_rate": 0.00019156305506216697, "loss": 0.6954, "step": 198 }, { "epoch": 0.13246796471958727, "grad_norm": 95.49404907226562, "learning_rate": 0.00019151865008880996, "loss": 0.6267, "step": 199 }, { "epoch": 0.1331336328840073, "grad_norm": 41.02389144897461, "learning_rate": 0.00019147424511545294, "loss": 0.2385, "step": 200 }, { "epoch": 0.13379930104842735, "grad_norm": 91.21604919433594, "learning_rate": 0.00019142984014209593, "loss": 0.4531, "step": 201 }, { "epoch": 0.1344649692128474, "grad_norm": 49.900184631347656, "learning_rate": 0.0001913854351687389, "loss": 0.1657, "step": 202 }, { "epoch": 0.13513063737726744, "grad_norm": 144.78623962402344, "learning_rate": 0.0001913410301953819, "loss": 0.7326, "step": 203 }, { "epoch": 0.13579630554168748, "grad_norm": 143.92132568359375, "learning_rate": 0.00019129662522202487, "loss": 1.0342, "step": 204 }, { "epoch": 0.1364619737061075, "grad_norm": 107.93486022949219, "learning_rate": 0.00019125222024866787, "loss": 0.3643, "step": 205 }, { "epoch": 0.13712764187052753, "grad_norm": 69.4767074584961, "learning_rate": 0.00019120781527531084, "loss": 0.4302, "step": 206 }, { "epoch": 0.13779331003494757, "grad_norm": 102.22624206542969, "learning_rate": 0.00019116341030195384, "loss": 0.7252, "step": 207 }, { "epoch": 0.13845897819936762, "grad_norm": 39.15114974975586, "learning_rate": 0.00019111900532859681, "loss": 0.376, "step": 208 }, { "epoch": 0.13912464636378766, "grad_norm": 65.12492370605469, "learning_rate": 0.00019107460035523979, "loss": 0.6005, "step": 209 }, { "epoch": 0.1397903145282077, "grad_norm": 95.04318237304688, "learning_rate": 0.00019103019538188278, "loss": 0.8371, "step": 210 }, { "epoch": 0.1404559826926277, "grad_norm": 119.12702941894531, "learning_rate": 0.00019098579040852576, "loss": 0.5701, "step": 211 }, { "epoch": 0.14112165085704775, "grad_norm": 56.38259506225586, "learning_rate": 0.00019094138543516875, "loss": 0.6285, "step": 212 }, { "epoch": 0.1417873190214678, "grad_norm": 29.10639762878418, "learning_rate": 0.00019089698046181172, "loss": 0.4639, "step": 213 }, { "epoch": 0.14245298718588784, "grad_norm": 60.445865631103516, "learning_rate": 0.00019085257548845472, "loss": 0.6099, "step": 214 }, { "epoch": 0.14311865535030788, "grad_norm": 83.33357238769531, "learning_rate": 0.0001908081705150977, "loss": 0.6992, "step": 215 }, { "epoch": 0.14378432351472792, "grad_norm": 50.539451599121094, "learning_rate": 0.0001907637655417407, "loss": 0.673, "step": 216 }, { "epoch": 0.14444999167914793, "grad_norm": 108.4279556274414, "learning_rate": 0.00019071936056838366, "loss": 0.7867, "step": 217 }, { "epoch": 0.14511565984356797, "grad_norm": 71.25292205810547, "learning_rate": 0.00019067495559502666, "loss": 0.2414, "step": 218 }, { "epoch": 0.14578132800798801, "grad_norm": 96.37142944335938, "learning_rate": 0.00019063055062166963, "loss": 0.4556, "step": 219 }, { "epoch": 0.14644699617240806, "grad_norm": 105.84346008300781, "learning_rate": 0.00019058614564831263, "loss": 0.5288, "step": 220 }, { "epoch": 0.1471126643368281, "grad_norm": 53.66656494140625, "learning_rate": 0.0001905417406749556, "loss": 0.2665, "step": 221 }, { "epoch": 0.14777833250124814, "grad_norm": 73.43135070800781, "learning_rate": 0.00019049733570159857, "loss": 0.465, "step": 222 }, { "epoch": 0.14844400066566815, "grad_norm": 108.94127655029297, "learning_rate": 0.00019045293072824157, "loss": 0.6516, "step": 223 }, { "epoch": 0.1491096688300882, "grad_norm": 23.10163688659668, "learning_rate": 0.00019040852575488454, "loss": 0.1401, "step": 224 }, { "epoch": 0.14977533699450823, "grad_norm": 182.62322998046875, "learning_rate": 0.00019036412078152754, "loss": 0.7953, "step": 225 }, { "epoch": 0.15044100515892828, "grad_norm": 104.31300354003906, "learning_rate": 0.00019031971580817051, "loss": 0.5117, "step": 226 }, { "epoch": 0.15110667332334832, "grad_norm": 122.88026428222656, "learning_rate": 0.0001902753108348135, "loss": 0.8029, "step": 227 }, { "epoch": 0.15177234148776836, "grad_norm": 72.17605590820312, "learning_rate": 0.00019023090586145648, "loss": 0.3403, "step": 228 }, { "epoch": 0.15243800965218837, "grad_norm": 55.70381164550781, "learning_rate": 0.00019018650088809948, "loss": 0.2811, "step": 229 }, { "epoch": 0.1531036778166084, "grad_norm": 80.8816146850586, "learning_rate": 0.00019014209591474245, "loss": 1.0904, "step": 230 }, { "epoch": 0.15376934598102845, "grad_norm": 77.0992431640625, "learning_rate": 0.00019009769094138545, "loss": 0.8617, "step": 231 }, { "epoch": 0.1544350141454485, "grad_norm": 47.77476119995117, "learning_rate": 0.00019005328596802842, "loss": 0.3185, "step": 232 }, { "epoch": 0.15510068230986854, "grad_norm": 77.98711395263672, "learning_rate": 0.00019000888099467142, "loss": 0.2655, "step": 233 }, { "epoch": 0.15576635047428858, "grad_norm": 63.54255294799805, "learning_rate": 0.0001899644760213144, "loss": 0.3955, "step": 234 }, { "epoch": 0.1564320186387086, "grad_norm": 190.26271057128906, "learning_rate": 0.00018992007104795736, "loss": 1.2886, "step": 235 }, { "epoch": 0.15709768680312863, "grad_norm": 117.44766998291016, "learning_rate": 0.00018987566607460036, "loss": 0.7142, "step": 236 }, { "epoch": 0.15776335496754867, "grad_norm": 133.50717163085938, "learning_rate": 0.00018983126110124333, "loss": 0.8898, "step": 237 }, { "epoch": 0.15842902313196872, "grad_norm": 62.771507263183594, "learning_rate": 0.00018978685612788633, "loss": 0.3062, "step": 238 }, { "epoch": 0.15909469129638876, "grad_norm": 74.36737060546875, "learning_rate": 0.0001897424511545293, "loss": 0.5336, "step": 239 }, { "epoch": 0.1597603594608088, "grad_norm": 137.45458984375, "learning_rate": 0.0001896980461811723, "loss": 0.7025, "step": 240 }, { "epoch": 0.1604260276252288, "grad_norm": 54.84712600708008, "learning_rate": 0.00018965364120781527, "loss": 0.338, "step": 241 }, { "epoch": 0.16109169578964885, "grad_norm": 127.16575622558594, "learning_rate": 0.00018960923623445827, "loss": 1.0382, "step": 242 }, { "epoch": 0.1617573639540689, "grad_norm": 93.72176361083984, "learning_rate": 0.00018956483126110124, "loss": 1.0084, "step": 243 }, { "epoch": 0.16242303211848894, "grad_norm": 56.54580307006836, "learning_rate": 0.00018952042628774424, "loss": 0.5353, "step": 244 }, { "epoch": 0.16308870028290898, "grad_norm": 76.96385955810547, "learning_rate": 0.0001894760213143872, "loss": 0.397, "step": 245 }, { "epoch": 0.16375436844732902, "grad_norm": 91.15630340576172, "learning_rate": 0.00018943161634103018, "loss": 0.4533, "step": 246 }, { "epoch": 0.16442003661174903, "grad_norm": 103.4432373046875, "learning_rate": 0.00018938721136767318, "loss": 0.6514, "step": 247 }, { "epoch": 0.16508570477616907, "grad_norm": 60.81359100341797, "learning_rate": 0.00018934280639431615, "loss": 0.4884, "step": 248 }, { "epoch": 0.16575137294058911, "grad_norm": 46.649139404296875, "learning_rate": 0.00018929840142095915, "loss": 0.271, "step": 249 }, { "epoch": 0.16641704110500916, "grad_norm": 58.7072868347168, "learning_rate": 0.00018925399644760212, "loss": 0.27, "step": 250 }, { "epoch": 0.1670827092694292, "grad_norm": 79.08338928222656, "learning_rate": 0.00018920959147424512, "loss": 0.4671, "step": 251 }, { "epoch": 0.16774837743384924, "grad_norm": 98.57723236083984, "learning_rate": 0.0001891651865008881, "loss": 0.3235, "step": 252 }, { "epoch": 0.16841404559826925, "grad_norm": 131.836181640625, "learning_rate": 0.0001891207815275311, "loss": 0.4688, "step": 253 }, { "epoch": 0.1690797137626893, "grad_norm": 67.51493835449219, "learning_rate": 0.00018907637655417406, "loss": 0.6911, "step": 254 }, { "epoch": 0.16974538192710933, "grad_norm": 51.76738357543945, "learning_rate": 0.00018903197158081706, "loss": 0.1837, "step": 255 }, { "epoch": 0.17041105009152938, "grad_norm": 52.23995590209961, "learning_rate": 0.00018898756660746003, "loss": 0.71, "step": 256 }, { "epoch": 0.17107671825594942, "grad_norm": 49.984336853027344, "learning_rate": 0.00018894316163410303, "loss": 0.3221, "step": 257 }, { "epoch": 0.17174238642036946, "grad_norm": 48.08108901977539, "learning_rate": 0.000188898756660746, "loss": 0.3334, "step": 258 }, { "epoch": 0.17240805458478947, "grad_norm": 24.143003463745117, "learning_rate": 0.00018885435168738897, "loss": 0.2096, "step": 259 }, { "epoch": 0.1730737227492095, "grad_norm": 116.61510467529297, "learning_rate": 0.00018880994671403197, "loss": 0.8852, "step": 260 }, { "epoch": 0.17373939091362955, "grad_norm": 41.168052673339844, "learning_rate": 0.00018876554174067494, "loss": 0.2541, "step": 261 }, { "epoch": 0.1744050590780496, "grad_norm": 19.256343841552734, "learning_rate": 0.00018872113676731794, "loss": 0.2278, "step": 262 }, { "epoch": 0.17507072724246964, "grad_norm": 51.543418884277344, "learning_rate": 0.0001886767317939609, "loss": 0.2494, "step": 263 }, { "epoch": 0.17573639540688968, "grad_norm": 44.67826461791992, "learning_rate": 0.0001886323268206039, "loss": 0.3112, "step": 264 }, { "epoch": 0.1764020635713097, "grad_norm": 41.30339431762695, "learning_rate": 0.00018858792184724688, "loss": 0.5114, "step": 265 }, { "epoch": 0.17706773173572973, "grad_norm": 12.575431823730469, "learning_rate": 0.00018854351687388988, "loss": 0.0375, "step": 266 }, { "epoch": 0.17773339990014977, "grad_norm": 71.91178894042969, "learning_rate": 0.00018849911190053285, "loss": 0.5989, "step": 267 }, { "epoch": 0.17839906806456982, "grad_norm": 83.31620788574219, "learning_rate": 0.00018845470692717585, "loss": 0.51, "step": 268 }, { "epoch": 0.17906473622898986, "grad_norm": 83.13764190673828, "learning_rate": 0.00018841030195381885, "loss": 1.1146, "step": 269 }, { "epoch": 0.1797304043934099, "grad_norm": 151.9849853515625, "learning_rate": 0.00018836589698046182, "loss": 1.1219, "step": 270 }, { "epoch": 0.1803960725578299, "grad_norm": 121.70146179199219, "learning_rate": 0.00018832149200710482, "loss": 0.9397, "step": 271 }, { "epoch": 0.18106174072224995, "grad_norm": 51.62641906738281, "learning_rate": 0.0001882770870337478, "loss": 0.2968, "step": 272 }, { "epoch": 0.18172740888667, "grad_norm": 66.82881164550781, "learning_rate": 0.0001882326820603908, "loss": 0.5365, "step": 273 }, { "epoch": 0.18239307705109004, "grad_norm": 78.71537017822266, "learning_rate": 0.00018818827708703376, "loss": 0.8149, "step": 274 }, { "epoch": 0.18305874521551008, "grad_norm": 137.4581756591797, "learning_rate": 0.00018814387211367676, "loss": 0.6777, "step": 275 }, { "epoch": 0.18372441337993012, "grad_norm": 64.81407165527344, "learning_rate": 0.00018809946714031973, "loss": 0.7677, "step": 276 }, { "epoch": 0.18439008154435013, "grad_norm": 96.37416076660156, "learning_rate": 0.00018805506216696273, "loss": 1.332, "step": 277 }, { "epoch": 0.18505574970877017, "grad_norm": 47.11737823486328, "learning_rate": 0.0001880106571936057, "loss": 0.29, "step": 278 }, { "epoch": 0.18572141787319021, "grad_norm": 111.5100326538086, "learning_rate": 0.0001879662522202487, "loss": 0.5466, "step": 279 }, { "epoch": 0.18638708603761026, "grad_norm": 123.91747283935547, "learning_rate": 0.00018792184724689167, "loss": 0.6733, "step": 280 }, { "epoch": 0.1870527542020303, "grad_norm": 107.9887924194336, "learning_rate": 0.00018787744227353467, "loss": 0.4885, "step": 281 }, { "epoch": 0.18771842236645034, "grad_norm": 76.66865539550781, "learning_rate": 0.00018783303730017764, "loss": 0.6393, "step": 282 }, { "epoch": 0.18838409053087035, "grad_norm": 67.36553192138672, "learning_rate": 0.00018778863232682064, "loss": 0.4087, "step": 283 }, { "epoch": 0.1890497586952904, "grad_norm": 50.0120849609375, "learning_rate": 0.0001877442273534636, "loss": 0.3911, "step": 284 }, { "epoch": 0.18971542685971043, "grad_norm": 109.88864135742188, "learning_rate": 0.00018769982238010658, "loss": 0.6727, "step": 285 }, { "epoch": 0.19038109502413048, "grad_norm": 114.64707946777344, "learning_rate": 0.00018765541740674958, "loss": 0.8044, "step": 286 }, { "epoch": 0.19104676318855052, "grad_norm": 130.69219970703125, "learning_rate": 0.00018761101243339255, "loss": 0.6119, "step": 287 }, { "epoch": 0.19171243135297053, "grad_norm": 50.73808670043945, "learning_rate": 0.00018756660746003555, "loss": 0.4887, "step": 288 }, { "epoch": 0.19237809951739057, "grad_norm": 143.2826385498047, "learning_rate": 0.00018752220248667852, "loss": 0.8228, "step": 289 }, { "epoch": 0.1930437676818106, "grad_norm": 132.3501739501953, "learning_rate": 0.00018747779751332152, "loss": 0.808, "step": 290 }, { "epoch": 0.19370943584623065, "grad_norm": 74.29531860351562, "learning_rate": 0.0001874333925399645, "loss": 0.7815, "step": 291 }, { "epoch": 0.1943751040106507, "grad_norm": 81.61593627929688, "learning_rate": 0.00018738898756660749, "loss": 0.5732, "step": 292 }, { "epoch": 0.19504077217507074, "grad_norm": 86.1830825805664, "learning_rate": 0.00018734458259325046, "loss": 0.5226, "step": 293 }, { "epoch": 0.19570644033949075, "grad_norm": 23.495864868164062, "learning_rate": 0.00018730017761989346, "loss": 0.1476, "step": 294 }, { "epoch": 0.1963721085039108, "grad_norm": 45.22040939331055, "learning_rate": 0.00018725577264653643, "loss": 0.4335, "step": 295 }, { "epoch": 0.19703777666833083, "grad_norm": 102.33810424804688, "learning_rate": 0.0001872113676731794, "loss": 0.6778, "step": 296 }, { "epoch": 0.19770344483275087, "grad_norm": 90.49777221679688, "learning_rate": 0.0001871669626998224, "loss": 0.7652, "step": 297 }, { "epoch": 0.19836911299717092, "grad_norm": 14.38054084777832, "learning_rate": 0.00018712255772646537, "loss": 0.0598, "step": 298 }, { "epoch": 0.19903478116159096, "grad_norm": 24.415199279785156, "learning_rate": 0.00018707815275310837, "loss": 0.3666, "step": 299 }, { "epoch": 0.19970044932601097, "grad_norm": 37.869869232177734, "learning_rate": 0.00018703374777975134, "loss": 0.1511, "step": 300 }, { "epoch": 0.200366117490431, "grad_norm": 115.4464111328125, "learning_rate": 0.00018698934280639434, "loss": 0.7441, "step": 301 }, { "epoch": 0.20103178565485105, "grad_norm": 124.49742889404297, "learning_rate": 0.0001869449378330373, "loss": 0.6898, "step": 302 }, { "epoch": 0.2016974538192711, "grad_norm": 45.529197692871094, "learning_rate": 0.0001869005328596803, "loss": 0.2891, "step": 303 }, { "epoch": 0.20236312198369114, "grad_norm": 107.60458374023438, "learning_rate": 0.00018685612788632328, "loss": 0.6037, "step": 304 }, { "epoch": 0.20302879014811118, "grad_norm": 72.92047119140625, "learning_rate": 0.00018681172291296628, "loss": 0.8364, "step": 305 }, { "epoch": 0.2036944583125312, "grad_norm": 80.35264587402344, "learning_rate": 0.00018676731793960925, "loss": 0.6218, "step": 306 }, { "epoch": 0.20436012647695123, "grad_norm": 181.0091552734375, "learning_rate": 0.00018672291296625225, "loss": 0.8943, "step": 307 }, { "epoch": 0.20502579464137127, "grad_norm": 54.508758544921875, "learning_rate": 0.00018667850799289522, "loss": 0.3581, "step": 308 }, { "epoch": 0.20569146280579131, "grad_norm": 22.76518440246582, "learning_rate": 0.0001866341030195382, "loss": 0.0707, "step": 309 }, { "epoch": 0.20635713097021136, "grad_norm": 144.45138549804688, "learning_rate": 0.00018658969804618119, "loss": 1.199, "step": 310 }, { "epoch": 0.2070227991346314, "grad_norm": 51.15176010131836, "learning_rate": 0.00018654529307282416, "loss": 0.2966, "step": 311 }, { "epoch": 0.2076884672990514, "grad_norm": 84.0188980102539, "learning_rate": 0.00018650088809946716, "loss": 0.3478, "step": 312 }, { "epoch": 0.20835413546347145, "grad_norm": 74.65904235839844, "learning_rate": 0.00018645648312611013, "loss": 0.5906, "step": 313 }, { "epoch": 0.2090198036278915, "grad_norm": 76.40025329589844, "learning_rate": 0.00018641207815275313, "loss": 0.5313, "step": 314 }, { "epoch": 0.20968547179231153, "grad_norm": 67.46442413330078, "learning_rate": 0.0001863676731793961, "loss": 0.5431, "step": 315 }, { "epoch": 0.21035113995673158, "grad_norm": 153.51947021484375, "learning_rate": 0.0001863232682060391, "loss": 1.4613, "step": 316 }, { "epoch": 0.21101680812115162, "grad_norm": 141.46417236328125, "learning_rate": 0.00018627886323268207, "loss": 0.6173, "step": 317 }, { "epoch": 0.21168247628557163, "grad_norm": 117.94355773925781, "learning_rate": 0.00018623445825932506, "loss": 0.703, "step": 318 }, { "epoch": 0.21234814444999167, "grad_norm": 72.00027465820312, "learning_rate": 0.00018619005328596804, "loss": 0.3786, "step": 319 }, { "epoch": 0.2130138126144117, "grad_norm": 54.18503189086914, "learning_rate": 0.00018614564831261103, "loss": 0.3633, "step": 320 }, { "epoch": 0.21367948077883175, "grad_norm": 121.49830627441406, "learning_rate": 0.000186101243339254, "loss": 0.7459, "step": 321 }, { "epoch": 0.2143451489432518, "grad_norm": 176.50872802734375, "learning_rate": 0.00018605683836589698, "loss": 0.8166, "step": 322 }, { "epoch": 0.21501081710767184, "grad_norm": 151.5113067626953, "learning_rate": 0.00018601243339253998, "loss": 0.7919, "step": 323 }, { "epoch": 0.21567648527209185, "grad_norm": 84.63894653320312, "learning_rate": 0.00018596802841918295, "loss": 0.7106, "step": 324 }, { "epoch": 0.2163421534365119, "grad_norm": 102.90968322753906, "learning_rate": 0.00018592362344582595, "loss": 0.689, "step": 325 }, { "epoch": 0.21700782160093193, "grad_norm": 61.89885330200195, "learning_rate": 0.00018587921847246892, "loss": 0.3393, "step": 326 }, { "epoch": 0.21767348976535197, "grad_norm": 106.49290466308594, "learning_rate": 0.00018583481349911191, "loss": 0.3928, "step": 327 }, { "epoch": 0.21833915792977202, "grad_norm": 43.59565353393555, "learning_rate": 0.00018579040852575489, "loss": 0.3696, "step": 328 }, { "epoch": 0.21900482609419206, "grad_norm": 67.05834197998047, "learning_rate": 0.00018574600355239788, "loss": 0.8861, "step": 329 }, { "epoch": 0.21967049425861207, "grad_norm": 108.7596435546875, "learning_rate": 0.00018570159857904086, "loss": 1.1147, "step": 330 }, { "epoch": 0.2203361624230321, "grad_norm": 45.871212005615234, "learning_rate": 0.00018565719360568385, "loss": 0.2634, "step": 331 }, { "epoch": 0.22100183058745215, "grad_norm": 97.05673217773438, "learning_rate": 0.00018561278863232683, "loss": 0.7176, "step": 332 }, { "epoch": 0.2216674987518722, "grad_norm": 37.64607238769531, "learning_rate": 0.00018556838365896982, "loss": 0.405, "step": 333 }, { "epoch": 0.22233316691629224, "grad_norm": 82.15957641601562, "learning_rate": 0.0001855239786856128, "loss": 0.3589, "step": 334 }, { "epoch": 0.22299883508071228, "grad_norm": 48.32238006591797, "learning_rate": 0.00018547957371225577, "loss": 0.2946, "step": 335 }, { "epoch": 0.2236645032451323, "grad_norm": 33.3898811340332, "learning_rate": 0.00018543516873889876, "loss": 0.388, "step": 336 }, { "epoch": 0.22433017140955233, "grad_norm": 42.87260818481445, "learning_rate": 0.00018539076376554174, "loss": 0.2302, "step": 337 }, { "epoch": 0.22499583957397237, "grad_norm": 58.45392608642578, "learning_rate": 0.00018534635879218473, "loss": 0.6616, "step": 338 }, { "epoch": 0.22566150773839241, "grad_norm": 28.114885330200195, "learning_rate": 0.0001853019538188277, "loss": 0.1042, "step": 339 }, { "epoch": 0.22632717590281246, "grad_norm": 51.35744857788086, "learning_rate": 0.0001852575488454707, "loss": 0.4782, "step": 340 }, { "epoch": 0.2269928440672325, "grad_norm": 82.15337371826172, "learning_rate": 0.00018521314387211368, "loss": 0.3746, "step": 341 }, { "epoch": 0.2276585122316525, "grad_norm": 109.32012176513672, "learning_rate": 0.00018516873889875667, "loss": 0.6589, "step": 342 }, { "epoch": 0.22832418039607255, "grad_norm": 155.4207000732422, "learning_rate": 0.00018512433392539965, "loss": 1.1968, "step": 343 }, { "epoch": 0.2289898485604926, "grad_norm": 33.03417205810547, "learning_rate": 0.00018507992895204264, "loss": 0.3845, "step": 344 }, { "epoch": 0.22965551672491263, "grad_norm": 37.060585021972656, "learning_rate": 0.00018503552397868561, "loss": 0.1341, "step": 345 }, { "epoch": 0.23032118488933268, "grad_norm": 106.52214050292969, "learning_rate": 0.00018499111900532859, "loss": 0.7823, "step": 346 }, { "epoch": 0.23098685305375272, "grad_norm": 48.00297164916992, "learning_rate": 0.00018494671403197158, "loss": 0.6375, "step": 347 }, { "epoch": 0.23165252121817273, "grad_norm": 51.31806564331055, "learning_rate": 0.00018490230905861456, "loss": 0.4176, "step": 348 }, { "epoch": 0.23231818938259277, "grad_norm": 57.65751647949219, "learning_rate": 0.00018485790408525755, "loss": 0.2368, "step": 349 }, { "epoch": 0.2329838575470128, "grad_norm": 46.041908264160156, "learning_rate": 0.00018481349911190053, "loss": 0.4219, "step": 350 }, { "epoch": 0.23364952571143285, "grad_norm": 97.23726654052734, "learning_rate": 0.00018476909413854352, "loss": 0.8618, "step": 351 }, { "epoch": 0.2343151938758529, "grad_norm": 39.60933303833008, "learning_rate": 0.0001847246891651865, "loss": 0.2446, "step": 352 }, { "epoch": 0.23498086204027294, "grad_norm": 83.39789581298828, "learning_rate": 0.0001846802841918295, "loss": 0.5283, "step": 353 }, { "epoch": 0.23564653020469295, "grad_norm": 23.28626823425293, "learning_rate": 0.00018463587921847246, "loss": 0.2269, "step": 354 }, { "epoch": 0.236312198369113, "grad_norm": 75.43800354003906, "learning_rate": 0.00018459147424511546, "loss": 0.5959, "step": 355 }, { "epoch": 0.23697786653353303, "grad_norm": 33.72787094116211, "learning_rate": 0.00018454706927175843, "loss": 0.2795, "step": 356 }, { "epoch": 0.23764353469795307, "grad_norm": 56.4429817199707, "learning_rate": 0.00018450266429840143, "loss": 0.6191, "step": 357 }, { "epoch": 0.23830920286237312, "grad_norm": 97.20005798339844, "learning_rate": 0.0001844582593250444, "loss": 0.6252, "step": 358 }, { "epoch": 0.23897487102679316, "grad_norm": NaN, "learning_rate": 0.0001844582593250444, "loss": 1.1109, "step": 359 }, { "epoch": 0.23964053919121317, "grad_norm": 103.0977783203125, "learning_rate": 0.00018441385435168738, "loss": 0.5326, "step": 360 }, { "epoch": 0.2403062073556332, "grad_norm": 33.12345504760742, "learning_rate": 0.00018436944937833037, "loss": 0.4465, "step": 361 }, { "epoch": 0.24097187552005325, "grad_norm": 29.757606506347656, "learning_rate": 0.00018432504440497335, "loss": 0.3853, "step": 362 }, { "epoch": 0.2416375436844733, "grad_norm": 52.76167678833008, "learning_rate": 0.00018428063943161634, "loss": 0.3521, "step": 363 }, { "epoch": 0.24230321184889334, "grad_norm": 80.51199340820312, "learning_rate": 0.00018423623445825931, "loss": 0.4551, "step": 364 }, { "epoch": 0.24296888001331338, "grad_norm": 47.3682861328125, "learning_rate": 0.0001841918294849023, "loss": 0.2393, "step": 365 }, { "epoch": 0.2436345481777334, "grad_norm": 80.18375396728516, "learning_rate": 0.00018414742451154528, "loss": 0.9122, "step": 366 }, { "epoch": 0.24430021634215343, "grad_norm": 65.6572265625, "learning_rate": 0.00018410301953818828, "loss": 0.3345, "step": 367 }, { "epoch": 0.24496588450657347, "grad_norm": 108.21941375732422, "learning_rate": 0.00018405861456483125, "loss": 0.5405, "step": 368 }, { "epoch": 0.2456315526709935, "grad_norm": 57.63541030883789, "learning_rate": 0.00018401420959147425, "loss": 0.2502, "step": 369 }, { "epoch": 0.24629722083541356, "grad_norm": 35.54652786254883, "learning_rate": 0.00018396980461811722, "loss": 0.5846, "step": 370 }, { "epoch": 0.2469628889998336, "grad_norm": 55.58504867553711, "learning_rate": 0.00018392539964476022, "loss": 0.2154, "step": 371 }, { "epoch": 0.2476285571642536, "grad_norm": 55.62077331542969, "learning_rate": 0.0001838809946714032, "loss": 0.4984, "step": 372 }, { "epoch": 0.24829422532867365, "grad_norm": 87.00872039794922, "learning_rate": 0.0001838365896980462, "loss": 0.4538, "step": 373 }, { "epoch": 0.2489598934930937, "grad_norm": 46.597068786621094, "learning_rate": 0.00018379218472468916, "loss": 0.648, "step": 374 }, { "epoch": 0.24962556165751373, "grad_norm": 99.84630584716797, "learning_rate": 0.00018374777975133216, "loss": 0.7931, "step": 375 }, { "epoch": 0.25029122982193375, "grad_norm": 54.02254104614258, "learning_rate": 0.00018370337477797513, "loss": 0.2499, "step": 376 }, { "epoch": 0.2509568979863538, "grad_norm": 29.319000244140625, "learning_rate": 0.00018365896980461813, "loss": 0.4007, "step": 377 }, { "epoch": 0.25162256615077383, "grad_norm": 113.27352905273438, "learning_rate": 0.0001836145648312611, "loss": 0.7379, "step": 378 }, { "epoch": 0.25228823431519387, "grad_norm": 112.12903594970703, "learning_rate": 0.0001835701598579041, "loss": 0.6326, "step": 379 }, { "epoch": 0.2529539024796139, "grad_norm": 123.96865844726562, "learning_rate": 0.00018352575488454707, "loss": 0.5521, "step": 380 }, { "epoch": 0.25361957064403395, "grad_norm": 42.07058334350586, "learning_rate": 0.00018348134991119007, "loss": 0.4259, "step": 381 }, { "epoch": 0.254285238808454, "grad_norm": 55.002960205078125, "learning_rate": 0.00018343694493783304, "loss": 0.4797, "step": 382 }, { "epoch": 0.25495090697287404, "grad_norm": 33.028038024902344, "learning_rate": 0.00018339253996447604, "loss": 0.2893, "step": 383 }, { "epoch": 0.2556165751372941, "grad_norm": 54.003570556640625, "learning_rate": 0.000183348134991119, "loss": 0.4897, "step": 384 }, { "epoch": 0.2562822433017141, "grad_norm": 80.12110137939453, "learning_rate": 0.000183303730017762, "loss": 0.2902, "step": 385 }, { "epoch": 0.2569479114661341, "grad_norm": 46.586753845214844, "learning_rate": 0.00018325932504440498, "loss": 0.3721, "step": 386 }, { "epoch": 0.25761357963055415, "grad_norm": 45.596553802490234, "learning_rate": 0.00018321492007104798, "loss": 0.2719, "step": 387 }, { "epoch": 0.2582792477949742, "grad_norm": 90.06937408447266, "learning_rate": 0.00018317051509769095, "loss": 0.6424, "step": 388 }, { "epoch": 0.25894491595939423, "grad_norm": 44.34123611450195, "learning_rate": 0.00018312611012433395, "loss": 0.2926, "step": 389 }, { "epoch": 0.25961058412381427, "grad_norm": 123.05774688720703, "learning_rate": 0.00018308170515097692, "loss": 0.8703, "step": 390 }, { "epoch": 0.2602762522882343, "grad_norm": 132.67784118652344, "learning_rate": 0.00018303730017761992, "loss": 0.492, "step": 391 }, { "epoch": 0.26094192045265435, "grad_norm": 65.75527954101562, "learning_rate": 0.0001829928952042629, "loss": 0.2371, "step": 392 }, { "epoch": 0.2616075886170744, "grad_norm": 70.54750061035156, "learning_rate": 0.0001829484902309059, "loss": 0.5192, "step": 393 }, { "epoch": 0.26227325678149443, "grad_norm": 85.00897216796875, "learning_rate": 0.00018290408525754886, "loss": 0.7114, "step": 394 }, { "epoch": 0.2629389249459145, "grad_norm": 96.78849792480469, "learning_rate": 0.00018285968028419186, "loss": 0.4515, "step": 395 }, { "epoch": 0.2636045931103345, "grad_norm": 21.13523292541504, "learning_rate": 0.00018281527531083483, "loss": 0.3526, "step": 396 }, { "epoch": 0.26427026127475456, "grad_norm": 29.27351188659668, "learning_rate": 0.0001827708703374778, "loss": 0.5151, "step": 397 }, { "epoch": 0.26493592943917454, "grad_norm": 35.63692855834961, "learning_rate": 0.0001827264653641208, "loss": 0.2374, "step": 398 }, { "epoch": 0.2656015976035946, "grad_norm": 66.08139038085938, "learning_rate": 0.00018268206039076377, "loss": 0.4133, "step": 399 }, { "epoch": 0.2662672657680146, "grad_norm": 34.3465461730957, "learning_rate": 0.00018263765541740677, "loss": 0.3082, "step": 400 }, { "epoch": 0.26693293393243467, "grad_norm": 100.62875366210938, "learning_rate": 0.00018259325044404974, "loss": 0.6884, "step": 401 }, { "epoch": 0.2675986020968547, "grad_norm": 139.31932067871094, "learning_rate": 0.00018254884547069274, "loss": 1.0101, "step": 402 }, { "epoch": 0.26826427026127475, "grad_norm": 118.92940521240234, "learning_rate": 0.0001825044404973357, "loss": 0.9706, "step": 403 }, { "epoch": 0.2689299384256948, "grad_norm": 151.63421630859375, "learning_rate": 0.0001824600355239787, "loss": 0.961, "step": 404 }, { "epoch": 0.26959560659011483, "grad_norm": 27.14082145690918, "learning_rate": 0.00018241563055062168, "loss": 0.227, "step": 405 }, { "epoch": 0.2702612747545349, "grad_norm": 80.59782409667969, "learning_rate": 0.00018237122557726468, "loss": 0.6828, "step": 406 }, { "epoch": 0.2709269429189549, "grad_norm": 49.46958541870117, "learning_rate": 0.00018232682060390765, "loss": 0.3184, "step": 407 }, { "epoch": 0.27159261108337496, "grad_norm": 73.98738098144531, "learning_rate": 0.00018228241563055065, "loss": 0.6858, "step": 408 }, { "epoch": 0.272258279247795, "grad_norm": 86.27637481689453, "learning_rate": 0.00018223801065719362, "loss": 0.5358, "step": 409 }, { "epoch": 0.272923947412215, "grad_norm": 145.42340087890625, "learning_rate": 0.0001821936056838366, "loss": 1.0992, "step": 410 }, { "epoch": 0.273589615576635, "grad_norm": 41.17599105834961, "learning_rate": 0.0001821492007104796, "loss": 0.369, "step": 411 }, { "epoch": 0.27425528374105507, "grad_norm": 55.49694061279297, "learning_rate": 0.00018210479573712256, "loss": 0.2842, "step": 412 }, { "epoch": 0.2749209519054751, "grad_norm": 47.01858139038086, "learning_rate": 0.00018206039076376556, "loss": 0.3368, "step": 413 }, { "epoch": 0.27558662006989515, "grad_norm": 40.550880432128906, "learning_rate": 0.00018201598579040853, "loss": 0.2099, "step": 414 }, { "epoch": 0.2762522882343152, "grad_norm": 38.81869125366211, "learning_rate": 0.00018197158081705153, "loss": 0.28, "step": 415 }, { "epoch": 0.27691795639873523, "grad_norm": 65.78675842285156, "learning_rate": 0.0001819271758436945, "loss": 0.521, "step": 416 }, { "epoch": 0.2775836245631553, "grad_norm": 73.13219451904297, "learning_rate": 0.0001818827708703375, "loss": 0.7937, "step": 417 }, { "epoch": 0.2782492927275753, "grad_norm": 65.54695129394531, "learning_rate": 0.00018183836589698047, "loss": 0.4079, "step": 418 }, { "epoch": 0.27891496089199536, "grad_norm": 34.30622863769531, "learning_rate": 0.00018179396092362347, "loss": 0.0964, "step": 419 }, { "epoch": 0.2795806290564154, "grad_norm": 67.12702941894531, "learning_rate": 0.00018174955595026644, "loss": 1.4374, "step": 420 }, { "epoch": 0.28024629722083544, "grad_norm": 33.8336181640625, "learning_rate": 0.00018170515097690944, "loss": 0.1509, "step": 421 }, { "epoch": 0.2809119653852554, "grad_norm": 66.7548599243164, "learning_rate": 0.0001816607460035524, "loss": 0.2623, "step": 422 }, { "epoch": 0.28157763354967547, "grad_norm": 57.463253021240234, "learning_rate": 0.00018161634103019538, "loss": 0.3713, "step": 423 }, { "epoch": 0.2822433017140955, "grad_norm": 32.50680923461914, "learning_rate": 0.00018157193605683838, "loss": 0.1945, "step": 424 }, { "epoch": 0.28290896987851555, "grad_norm": 72.0036849975586, "learning_rate": 0.00018152753108348135, "loss": 0.3324, "step": 425 }, { "epoch": 0.2835746380429356, "grad_norm": 38.52781295776367, "learning_rate": 0.00018148312611012435, "loss": 0.185, "step": 426 }, { "epoch": 0.28424030620735563, "grad_norm": 74.13616943359375, "learning_rate": 0.00018143872113676732, "loss": 0.2684, "step": 427 }, { "epoch": 0.28490597437177567, "grad_norm": 76.92218017578125, "learning_rate": 0.00018139431616341032, "loss": 0.1456, "step": 428 }, { "epoch": 0.2855716425361957, "grad_norm": 131.1004638671875, "learning_rate": 0.0001813499111900533, "loss": 0.6252, "step": 429 }, { "epoch": 0.28623731070061575, "grad_norm": 112.32926940917969, "learning_rate": 0.0001813055062166963, "loss": 0.8337, "step": 430 }, { "epoch": 0.2869029788650358, "grad_norm": 125.7912368774414, "learning_rate": 0.00018126110124333926, "loss": 0.3741, "step": 431 }, { "epoch": 0.28756864702945584, "grad_norm": 164.3785858154297, "learning_rate": 0.00018121669626998226, "loss": 0.9616, "step": 432 }, { "epoch": 0.2882343151938759, "grad_norm": 191.42367553710938, "learning_rate": 0.00018117229129662523, "loss": 0.8257, "step": 433 }, { "epoch": 0.28889998335829586, "grad_norm": 79.08208465576172, "learning_rate": 0.00018112788632326823, "loss": 0.5202, "step": 434 }, { "epoch": 0.2895656515227159, "grad_norm": 109.46979522705078, "learning_rate": 0.0001810834813499112, "loss": 0.941, "step": 435 }, { "epoch": 0.29023131968713595, "grad_norm": 74.53866577148438, "learning_rate": 0.00018103907637655417, "loss": 0.7279, "step": 436 }, { "epoch": 0.290896987851556, "grad_norm": 137.80848693847656, "learning_rate": 0.00018099467140319717, "loss": 1.4923, "step": 437 }, { "epoch": 0.29156265601597603, "grad_norm": 126.38627624511719, "learning_rate": 0.00018095026642984014, "loss": 0.721, "step": 438 }, { "epoch": 0.29222832418039607, "grad_norm": 98.57434844970703, "learning_rate": 0.00018090586145648314, "loss": 0.5846, "step": 439 }, { "epoch": 0.2928939923448161, "grad_norm": 33.60892868041992, "learning_rate": 0.0001808614564831261, "loss": 0.4007, "step": 440 }, { "epoch": 0.29355966050923615, "grad_norm": 84.4782943725586, "learning_rate": 0.0001808170515097691, "loss": 0.6742, "step": 441 }, { "epoch": 0.2942253286736562, "grad_norm": 78.77970886230469, "learning_rate": 0.00018077264653641208, "loss": 0.5823, "step": 442 }, { "epoch": 0.29489099683807624, "grad_norm": 157.4933319091797, "learning_rate": 0.00018072824156305508, "loss": 0.7795, "step": 443 }, { "epoch": 0.2955566650024963, "grad_norm": 18.716060638427734, "learning_rate": 0.00018068383658969805, "loss": 0.3715, "step": 444 }, { "epoch": 0.2962223331669163, "grad_norm": 85.15080261230469, "learning_rate": 0.00018063943161634105, "loss": 0.606, "step": 445 }, { "epoch": 0.2968880013313363, "grad_norm": 83.07491302490234, "learning_rate": 0.00018059502664298402, "loss": 0.4614, "step": 446 }, { "epoch": 0.29755366949575635, "grad_norm": 52.511085510253906, "learning_rate": 0.000180550621669627, "loss": 0.4219, "step": 447 }, { "epoch": 0.2982193376601764, "grad_norm": 126.2627944946289, "learning_rate": 0.00018050621669627, "loss": 0.6889, "step": 448 }, { "epoch": 0.29888500582459643, "grad_norm": 73.53312683105469, "learning_rate": 0.00018046181172291296, "loss": 0.577, "step": 449 }, { "epoch": 0.29955067398901647, "grad_norm": 103.77356719970703, "learning_rate": 0.00018041740674955596, "loss": 0.7924, "step": 450 }, { "epoch": 0.3002163421534365, "grad_norm": 89.16803741455078, "learning_rate": 0.00018037300177619893, "loss": 0.5117, "step": 451 }, { "epoch": 0.30088201031785655, "grad_norm": 116.20187377929688, "learning_rate": 0.00018032859680284193, "loss": 0.6893, "step": 452 }, { "epoch": 0.3015476784822766, "grad_norm": 130.5649871826172, "learning_rate": 0.0001802841918294849, "loss": 0.6782, "step": 453 }, { "epoch": 0.30221334664669663, "grad_norm": 71.50609588623047, "learning_rate": 0.0001802397868561279, "loss": 0.4608, "step": 454 }, { "epoch": 0.3028790148111167, "grad_norm": 101.877197265625, "learning_rate": 0.00018019538188277087, "loss": 0.5512, "step": 455 }, { "epoch": 0.3035446829755367, "grad_norm": 135.80545043945312, "learning_rate": 0.00018015097690941387, "loss": 0.6684, "step": 456 }, { "epoch": 0.30421035113995676, "grad_norm": 83.38865661621094, "learning_rate": 0.00018010657193605684, "loss": 0.4582, "step": 457 }, { "epoch": 0.30487601930437674, "grad_norm": 93.57701110839844, "learning_rate": 0.00018006216696269984, "loss": 0.7277, "step": 458 }, { "epoch": 0.3055416874687968, "grad_norm": 72.99111938476562, "learning_rate": 0.0001800177619893428, "loss": 0.5737, "step": 459 }, { "epoch": 0.3062073556332168, "grad_norm": 64.9205551147461, "learning_rate": 0.00017997335701598578, "loss": 0.3831, "step": 460 }, { "epoch": 0.30687302379763687, "grad_norm": 36.749935150146484, "learning_rate": 0.00017992895204262878, "loss": 0.4547, "step": 461 }, { "epoch": 0.3075386919620569, "grad_norm": 77.0653076171875, "learning_rate": 0.00017988454706927175, "loss": 0.8629, "step": 462 }, { "epoch": 0.30820436012647695, "grad_norm": 48.22278594970703, "learning_rate": 0.00017984014209591475, "loss": 0.3634, "step": 463 }, { "epoch": 0.308870028290897, "grad_norm": 39.0225715637207, "learning_rate": 0.00017979573712255772, "loss": 0.317, "step": 464 }, { "epoch": 0.30953569645531703, "grad_norm": 68.747314453125, "learning_rate": 0.00017975133214920072, "loss": 0.3624, "step": 465 }, { "epoch": 0.3102013646197371, "grad_norm": 71.55371856689453, "learning_rate": 0.0001797069271758437, "loss": 0.3992, "step": 466 }, { "epoch": 0.3108670327841571, "grad_norm": 36.87495422363281, "learning_rate": 0.00017966252220248669, "loss": 0.2098, "step": 467 }, { "epoch": 0.31153270094857716, "grad_norm": 10.664894104003906, "learning_rate": 0.00017961811722912966, "loss": 0.0542, "step": 468 }, { "epoch": 0.3121983691129972, "grad_norm": 34.32514953613281, "learning_rate": 0.00017957371225577265, "loss": 0.3147, "step": 469 }, { "epoch": 0.3128640372774172, "grad_norm": 24.026748657226562, "learning_rate": 0.00017952930728241563, "loss": 0.0797, "step": 470 }, { "epoch": 0.3135297054418372, "grad_norm": 23.764530181884766, "learning_rate": 0.00017948490230905862, "loss": 0.0965, "step": 471 }, { "epoch": 0.31419537360625727, "grad_norm": 5.237178325653076, "learning_rate": 0.0001794404973357016, "loss": 0.0106, "step": 472 }, { "epoch": 0.3148610417706773, "grad_norm": 20.264196395874023, "learning_rate": 0.0001793960923623446, "loss": 0.0546, "step": 473 }, { "epoch": 0.31552670993509735, "grad_norm": 79.56272888183594, "learning_rate": 0.00017935168738898757, "loss": 0.5564, "step": 474 }, { "epoch": 0.3161923780995174, "grad_norm": 37.59955596923828, "learning_rate": 0.00017930728241563056, "loss": 0.3443, "step": 475 }, { "epoch": 0.31685804626393743, "grad_norm": 49.22467803955078, "learning_rate": 0.00017926287744227354, "loss": 0.3749, "step": 476 }, { "epoch": 0.3175237144283575, "grad_norm": 175.10850524902344, "learning_rate": 0.00017921847246891653, "loss": 1.7856, "step": 477 }, { "epoch": 0.3181893825927775, "grad_norm": 47.71009826660156, "learning_rate": 0.0001791740674955595, "loss": 0.7082, "step": 478 }, { "epoch": 0.31885505075719756, "grad_norm": 82.27182006835938, "learning_rate": 0.0001791296625222025, "loss": 0.9351, "step": 479 }, { "epoch": 0.3195207189216176, "grad_norm": 79.33952331542969, "learning_rate": 0.00017908525754884547, "loss": 0.3707, "step": 480 }, { "epoch": 0.3201863870860376, "grad_norm": 88.69058227539062, "learning_rate": 0.00017904085257548847, "loss": 0.85, "step": 481 }, { "epoch": 0.3208520552504576, "grad_norm": 110.95427703857422, "learning_rate": 0.00017899644760213144, "loss": 0.3507, "step": 482 }, { "epoch": 0.32151772341487767, "grad_norm": 101.72648620605469, "learning_rate": 0.00017895204262877444, "loss": 0.8123, "step": 483 }, { "epoch": 0.3221833915792977, "grad_norm": 172.9845733642578, "learning_rate": 0.00017890763765541741, "loss": 1.1193, "step": 484 }, { "epoch": 0.32284905974371775, "grad_norm": 40.887081146240234, "learning_rate": 0.0001788632326820604, "loss": 0.1791, "step": 485 }, { "epoch": 0.3235147279081378, "grad_norm": 89.45999145507812, "learning_rate": 0.00017881882770870338, "loss": 0.6104, "step": 486 }, { "epoch": 0.32418039607255783, "grad_norm": 101.5306167602539, "learning_rate": 0.00017877442273534638, "loss": 0.5197, "step": 487 }, { "epoch": 0.32484606423697787, "grad_norm": 25.401044845581055, "learning_rate": 0.00017873001776198935, "loss": 0.2328, "step": 488 }, { "epoch": 0.3255117324013979, "grad_norm": 122.61588287353516, "learning_rate": 0.00017868561278863235, "loss": 0.9357, "step": 489 }, { "epoch": 0.32617740056581795, "grad_norm": 51.50986099243164, "learning_rate": 0.00017864120781527532, "loss": 0.506, "step": 490 }, { "epoch": 0.326843068730238, "grad_norm": 64.19971466064453, "learning_rate": 0.00017859680284191832, "loss": 0.2713, "step": 491 }, { "epoch": 0.32750873689465804, "grad_norm": 19.61275863647461, "learning_rate": 0.0001785523978685613, "loss": 0.1013, "step": 492 }, { "epoch": 0.328174405059078, "grad_norm": 66.36785125732422, "learning_rate": 0.0001785079928952043, "loss": 0.3726, "step": 493 }, { "epoch": 0.32884007322349806, "grad_norm": 87.43013763427734, "learning_rate": 0.00017846358792184726, "loss": 0.4851, "step": 494 }, { "epoch": 0.3295057413879181, "grad_norm": 63.96004867553711, "learning_rate": 0.00017841918294849026, "loss": 0.2111, "step": 495 }, { "epoch": 0.33017140955233815, "grad_norm": 81.01036834716797, "learning_rate": 0.00017837477797513323, "loss": 0.6494, "step": 496 }, { "epoch": 0.3308370777167582, "grad_norm": 63.75148010253906, "learning_rate": 0.0001783303730017762, "loss": 0.3041, "step": 497 }, { "epoch": 0.33150274588117823, "grad_norm": 91.487060546875, "learning_rate": 0.0001782859680284192, "loss": 0.689, "step": 498 }, { "epoch": 0.33216841404559827, "grad_norm": 58.02032470703125, "learning_rate": 0.00017824156305506217, "loss": 0.5346, "step": 499 }, { "epoch": 0.3328340822100183, "grad_norm": 73.02722930908203, "learning_rate": 0.00017819715808170517, "loss": 0.541, "step": 500 }, { "epoch": 0.33349975037443835, "grad_norm": 118.74954986572266, "learning_rate": 0.00017815275310834814, "loss": 0.5354, "step": 501 }, { "epoch": 0.3341654185388584, "grad_norm": 40.1008415222168, "learning_rate": 0.00017810834813499114, "loss": 0.2462, "step": 502 }, { "epoch": 0.33483108670327844, "grad_norm": 63.85152053833008, "learning_rate": 0.0001780639431616341, "loss": 0.5912, "step": 503 }, { "epoch": 0.3354967548676985, "grad_norm": 48.006710052490234, "learning_rate": 0.0001780195381882771, "loss": 0.5213, "step": 504 }, { "epoch": 0.33616242303211846, "grad_norm": 72.3712387084961, "learning_rate": 0.00017797513321492008, "loss": 0.3234, "step": 505 }, { "epoch": 0.3368280911965385, "grad_norm": 31.964515686035156, "learning_rate": 0.00017793072824156308, "loss": 0.1817, "step": 506 }, { "epoch": 0.33749375936095855, "grad_norm": 113.52684783935547, "learning_rate": 0.00017788632326820605, "loss": 0.6327, "step": 507 }, { "epoch": 0.3381594275253786, "grad_norm": 50.74462127685547, "learning_rate": 0.00017784191829484905, "loss": 0.3449, "step": 508 }, { "epoch": 0.3388250956897986, "grad_norm": 46.287391662597656, "learning_rate": 0.00017779751332149202, "loss": 0.2474, "step": 509 }, { "epoch": 0.33949076385421867, "grad_norm": 62.21271514892578, "learning_rate": 0.000177753108348135, "loss": 0.2743, "step": 510 }, { "epoch": 0.3401564320186387, "grad_norm": 52.463260650634766, "learning_rate": 0.000177708703374778, "loss": 0.3754, "step": 511 }, { "epoch": 0.34082210018305875, "grad_norm": 70.85062408447266, "learning_rate": 0.00017766429840142096, "loss": 0.3937, "step": 512 }, { "epoch": 0.3414877683474788, "grad_norm": 86.9253158569336, "learning_rate": 0.00017761989342806396, "loss": 0.3501, "step": 513 }, { "epoch": 0.34215343651189883, "grad_norm": 51.851112365722656, "learning_rate": 0.00017757548845470693, "loss": 0.2259, "step": 514 }, { "epoch": 0.3428191046763189, "grad_norm": 68.38616943359375, "learning_rate": 0.00017753108348134993, "loss": 0.4034, "step": 515 }, { "epoch": 0.3434847728407389, "grad_norm": 189.30650329589844, "learning_rate": 0.0001774866785079929, "loss": 0.8131, "step": 516 }, { "epoch": 0.3441504410051589, "grad_norm": 30.575939178466797, "learning_rate": 0.0001774422735346359, "loss": 0.1302, "step": 517 }, { "epoch": 0.34481610916957894, "grad_norm": 14.13813304901123, "learning_rate": 0.00017739786856127887, "loss": 0.1055, "step": 518 }, { "epoch": 0.345481777333999, "grad_norm": 124.42664337158203, "learning_rate": 0.00017735346358792187, "loss": 0.4918, "step": 519 }, { "epoch": 0.346147445498419, "grad_norm": 62.3731575012207, "learning_rate": 0.00017730905861456484, "loss": 0.2553, "step": 520 }, { "epoch": 0.34681311366283907, "grad_norm": 91.36651611328125, "learning_rate": 0.00017726465364120784, "loss": 0.3425, "step": 521 }, { "epoch": 0.3474787818272591, "grad_norm": 102.5283432006836, "learning_rate": 0.0001772202486678508, "loss": 0.4142, "step": 522 }, { "epoch": 0.34814444999167915, "grad_norm": 48.83149337768555, "learning_rate": 0.00017717584369449378, "loss": 0.327, "step": 523 }, { "epoch": 0.3488101181560992, "grad_norm": 68.21099090576172, "learning_rate": 0.00017713143872113678, "loss": 0.5076, "step": 524 }, { "epoch": 0.34947578632051923, "grad_norm": 50.855072021484375, "learning_rate": 0.00017708703374777975, "loss": 0.7123, "step": 525 }, { "epoch": 0.3501414544849393, "grad_norm": 95.03412628173828, "learning_rate": 0.00017704262877442275, "loss": 0.4952, "step": 526 }, { "epoch": 0.3508071226493593, "grad_norm": 45.51231384277344, "learning_rate": 0.00017699822380106572, "loss": 0.2897, "step": 527 }, { "epoch": 0.35147279081377936, "grad_norm": 83.111083984375, "learning_rate": 0.00017695381882770872, "loss": 0.3769, "step": 528 }, { "epoch": 0.35213845897819934, "grad_norm": 222.33758544921875, "learning_rate": 0.0001769094138543517, "loss": 1.4976, "step": 529 }, { "epoch": 0.3528041271426194, "grad_norm": 117.77812957763672, "learning_rate": 0.0001768650088809947, "loss": 1.0537, "step": 530 }, { "epoch": 0.3534697953070394, "grad_norm": 143.2985382080078, "learning_rate": 0.00017682060390763766, "loss": 1.206, "step": 531 }, { "epoch": 0.35413546347145947, "grad_norm": 78.69023132324219, "learning_rate": 0.00017677619893428066, "loss": 0.4841, "step": 532 }, { "epoch": 0.3548011316358795, "grad_norm": 30.563365936279297, "learning_rate": 0.00017673179396092363, "loss": 0.213, "step": 533 }, { "epoch": 0.35546679980029955, "grad_norm": 128.74452209472656, "learning_rate": 0.00017668738898756663, "loss": 0.9367, "step": 534 }, { "epoch": 0.3561324679647196, "grad_norm": 68.94506072998047, "learning_rate": 0.0001766429840142096, "loss": 0.671, "step": 535 }, { "epoch": 0.35679813612913963, "grad_norm": 87.38471984863281, "learning_rate": 0.00017659857904085257, "loss": 0.4509, "step": 536 }, { "epoch": 0.3574638042935597, "grad_norm": 127.65184783935547, "learning_rate": 0.00017655417406749557, "loss": 0.6365, "step": 537 }, { "epoch": 0.3581294724579797, "grad_norm": 23.673858642578125, "learning_rate": 0.00017650976909413854, "loss": 0.1011, "step": 538 }, { "epoch": 0.35879514062239976, "grad_norm": 41.28850173950195, "learning_rate": 0.00017646536412078154, "loss": 1.0469, "step": 539 }, { "epoch": 0.3594608087868198, "grad_norm": 66.7071304321289, "learning_rate": 0.0001764209591474245, "loss": 0.382, "step": 540 }, { "epoch": 0.3601264769512398, "grad_norm": 41.335899353027344, "learning_rate": 0.0001763765541740675, "loss": 0.3328, "step": 541 }, { "epoch": 0.3607921451156598, "grad_norm": 65.9492416381836, "learning_rate": 0.00017633214920071048, "loss": 0.2534, "step": 542 }, { "epoch": 0.36145781328007986, "grad_norm": 51.26821517944336, "learning_rate": 0.00017628774422735348, "loss": 0.2136, "step": 543 }, { "epoch": 0.3621234814444999, "grad_norm": 80.04798126220703, "learning_rate": 0.00017624333925399645, "loss": 0.3201, "step": 544 }, { "epoch": 0.36278914960891995, "grad_norm": 29.61306381225586, "learning_rate": 0.00017619893428063945, "loss": 0.361, "step": 545 }, { "epoch": 0.36345481777334, "grad_norm": 57.36445617675781, "learning_rate": 0.00017615452930728242, "loss": 0.3203, "step": 546 }, { "epoch": 0.36412048593776003, "grad_norm": 22.750394821166992, "learning_rate": 0.0001761101243339254, "loss": 0.0615, "step": 547 }, { "epoch": 0.36478615410218007, "grad_norm": 92.89879608154297, "learning_rate": 0.0001760657193605684, "loss": 0.7158, "step": 548 }, { "epoch": 0.3654518222666001, "grad_norm": 75.37178039550781, "learning_rate": 0.00017602131438721136, "loss": 0.2596, "step": 549 }, { "epoch": 0.36611749043102015, "grad_norm": 77.78299713134766, "learning_rate": 0.00017597690941385436, "loss": 0.5387, "step": 550 }, { "epoch": 0.3667831585954402, "grad_norm": 106.28474426269531, "learning_rate": 0.00017593250444049733, "loss": 0.5178, "step": 551 }, { "epoch": 0.36744882675986024, "grad_norm": 67.27828979492188, "learning_rate": 0.00017588809946714033, "loss": 0.4135, "step": 552 }, { "epoch": 0.3681144949242802, "grad_norm": 128.6137237548828, "learning_rate": 0.0001758436944937833, "loss": 0.591, "step": 553 }, { "epoch": 0.36878016308870026, "grad_norm": 69.33638763427734, "learning_rate": 0.0001757992895204263, "loss": 0.653, "step": 554 }, { "epoch": 0.3694458312531203, "grad_norm": 50.19843292236328, "learning_rate": 0.00017575488454706927, "loss": 0.2516, "step": 555 }, { "epoch": 0.37011149941754035, "grad_norm": 16.58024787902832, "learning_rate": 0.00017571047957371227, "loss": 0.26, "step": 556 }, { "epoch": 0.3707771675819604, "grad_norm": 25.605676651000977, "learning_rate": 0.00017566607460035524, "loss": 0.249, "step": 557 }, { "epoch": 0.37144283574638043, "grad_norm": 56.728515625, "learning_rate": 0.00017562166962699824, "loss": 0.3381, "step": 558 }, { "epoch": 0.37210850391080047, "grad_norm": 27.297210693359375, "learning_rate": 0.0001755772646536412, "loss": 0.111, "step": 559 }, { "epoch": 0.3727741720752205, "grad_norm": 85.54072570800781, "learning_rate": 0.00017553285968028418, "loss": 0.4148, "step": 560 }, { "epoch": 0.37343984023964055, "grad_norm": 80.2583236694336, "learning_rate": 0.00017548845470692718, "loss": 0.4845, "step": 561 }, { "epoch": 0.3741055084040606, "grad_norm": 33.99844741821289, "learning_rate": 0.00017544404973357015, "loss": 0.2521, "step": 562 }, { "epoch": 0.37477117656848064, "grad_norm": 55.42804718017578, "learning_rate": 0.00017539964476021315, "loss": 0.1915, "step": 563 }, { "epoch": 0.3754368447329007, "grad_norm": 34.7382926940918, "learning_rate": 0.00017535523978685612, "loss": 0.1789, "step": 564 }, { "epoch": 0.37610251289732066, "grad_norm": 67.93021392822266, "learning_rate": 0.00017531083481349912, "loss": 0.4769, "step": 565 }, { "epoch": 0.3767681810617407, "grad_norm": 83.29434967041016, "learning_rate": 0.0001752664298401421, "loss": 0.9316, "step": 566 }, { "epoch": 0.37743384922616074, "grad_norm": 29.34658432006836, "learning_rate": 0.0001752220248667851, "loss": 0.1181, "step": 567 }, { "epoch": 0.3780995173905808, "grad_norm": 73.09173583984375, "learning_rate": 0.00017517761989342806, "loss": 0.2341, "step": 568 }, { "epoch": 0.3787651855550008, "grad_norm": 100.28826141357422, "learning_rate": 0.00017513321492007106, "loss": 0.7953, "step": 569 }, { "epoch": 0.37943085371942087, "grad_norm": 34.65471267700195, "learning_rate": 0.00017508880994671403, "loss": 0.1516, "step": 570 }, { "epoch": 0.3800965218838409, "grad_norm": 40.57853317260742, "learning_rate": 0.00017504440497335703, "loss": 0.4384, "step": 571 }, { "epoch": 0.38076219004826095, "grad_norm": 42.92374038696289, "learning_rate": 0.000175, "loss": 0.7244, "step": 572 }, { "epoch": 0.381427858212681, "grad_norm": 43.757568359375, "learning_rate": 0.00017495559502664297, "loss": 0.2562, "step": 573 }, { "epoch": 0.38209352637710103, "grad_norm": 53.40153503417969, "learning_rate": 0.00017491119005328597, "loss": 0.272, "step": 574 }, { "epoch": 0.3827591945415211, "grad_norm": 58.91749572753906, "learning_rate": 0.00017486678507992894, "loss": 0.2955, "step": 575 }, { "epoch": 0.38342486270594106, "grad_norm": 45.21612548828125, "learning_rate": 0.00017482238010657194, "loss": 0.1861, "step": 576 }, { "epoch": 0.3840905308703611, "grad_norm": 179.73159790039062, "learning_rate": 0.0001747779751332149, "loss": 1.1993, "step": 577 }, { "epoch": 0.38475619903478114, "grad_norm": 129.5082550048828, "learning_rate": 0.0001747335701598579, "loss": 0.8181, "step": 578 }, { "epoch": 0.3854218671992012, "grad_norm": 39.42813491821289, "learning_rate": 0.00017468916518650088, "loss": 0.5695, "step": 579 }, { "epoch": 0.3860875353636212, "grad_norm": 57.18281173706055, "learning_rate": 0.00017464476021314388, "loss": 0.2304, "step": 580 }, { "epoch": 0.38675320352804127, "grad_norm": 43.29937744140625, "learning_rate": 0.00017460035523978685, "loss": 0.235, "step": 581 }, { "epoch": 0.3874188716924613, "grad_norm": 60.49991989135742, "learning_rate": 0.00017455595026642985, "loss": 0.1998, "step": 582 }, { "epoch": 0.38808453985688135, "grad_norm": 82.13057708740234, "learning_rate": 0.00017451154529307282, "loss": 0.4337, "step": 583 }, { "epoch": 0.3887502080213014, "grad_norm": 29.797197341918945, "learning_rate": 0.00017446714031971582, "loss": 0.1105, "step": 584 }, { "epoch": 0.38941587618572143, "grad_norm": 55.97276306152344, "learning_rate": 0.0001744227353463588, "loss": 0.3703, "step": 585 }, { "epoch": 0.3900815443501415, "grad_norm": 59.474639892578125, "learning_rate": 0.00017437833037300179, "loss": 0.3482, "step": 586 }, { "epoch": 0.3907472125145615, "grad_norm": 49.60660934448242, "learning_rate": 0.00017433392539964478, "loss": 0.5983, "step": 587 }, { "epoch": 0.3914128806789815, "grad_norm": 108.58460998535156, "learning_rate": 0.00017428952042628776, "loss": 0.9933, "step": 588 }, { "epoch": 0.39207854884340154, "grad_norm": 49.046234130859375, "learning_rate": 0.00017424511545293075, "loss": 0.1581, "step": 589 }, { "epoch": 0.3927442170078216, "grad_norm": 62.635292053222656, "learning_rate": 0.00017420071047957373, "loss": 0.6443, "step": 590 }, { "epoch": 0.3934098851722416, "grad_norm": 65.41645812988281, "learning_rate": 0.00017415630550621672, "loss": 0.5636, "step": 591 }, { "epoch": 0.39407555333666167, "grad_norm": 84.65705871582031, "learning_rate": 0.0001741119005328597, "loss": 0.5554, "step": 592 }, { "epoch": 0.3947412215010817, "grad_norm": 58.44446563720703, "learning_rate": 0.0001740674955595027, "loss": 0.2876, "step": 593 }, { "epoch": 0.39540688966550175, "grad_norm": 91.17931365966797, "learning_rate": 0.00017402309058614566, "loss": 0.4579, "step": 594 }, { "epoch": 0.3960725578299218, "grad_norm": 49.55839920043945, "learning_rate": 0.00017397868561278866, "loss": 0.5159, "step": 595 }, { "epoch": 0.39673822599434183, "grad_norm": 38.55497360229492, "learning_rate": 0.00017393428063943163, "loss": 0.2211, "step": 596 }, { "epoch": 0.3974038941587619, "grad_norm": 72.52485656738281, "learning_rate": 0.00017388987566607463, "loss": 0.8013, "step": 597 }, { "epoch": 0.3980695623231819, "grad_norm": 42.55357360839844, "learning_rate": 0.0001738454706927176, "loss": 0.1123, "step": 598 }, { "epoch": 0.39873523048760195, "grad_norm": 75.76644897460938, "learning_rate": 0.00017380106571936058, "loss": 0.6063, "step": 599 }, { "epoch": 0.39940089865202194, "grad_norm": 49.05486297607422, "learning_rate": 0.00017375666074600357, "loss": 0.3929, "step": 600 }, { "epoch": 0.400066566816442, "grad_norm": 54.81421661376953, "learning_rate": 0.00017371225577264654, "loss": 0.3706, "step": 601 }, { "epoch": 0.400732234980862, "grad_norm": 45.12424087524414, "learning_rate": 0.00017366785079928954, "loss": 0.1521, "step": 602 }, { "epoch": 0.40139790314528206, "grad_norm": 159.4431915283203, "learning_rate": 0.00017362344582593251, "loss": 0.9427, "step": 603 }, { "epoch": 0.4020635713097021, "grad_norm": 37.86516189575195, "learning_rate": 0.0001735790408525755, "loss": 0.3125, "step": 604 }, { "epoch": 0.40272923947412215, "grad_norm": 122.04438781738281, "learning_rate": 0.00017353463587921848, "loss": 0.6369, "step": 605 }, { "epoch": 0.4033949076385422, "grad_norm": 29.7852725982666, "learning_rate": 0.00017349023090586148, "loss": 0.0806, "step": 606 }, { "epoch": 0.40406057580296223, "grad_norm": 87.71656799316406, "learning_rate": 0.00017344582593250445, "loss": 0.624, "step": 607 }, { "epoch": 0.40472624396738227, "grad_norm": 91.41803741455078, "learning_rate": 0.00017340142095914745, "loss": 0.9653, "step": 608 }, { "epoch": 0.4053919121318023, "grad_norm": 116.02742004394531, "learning_rate": 0.00017335701598579042, "loss": 0.6518, "step": 609 }, { "epoch": 0.40605758029622235, "grad_norm": 39.57069778442383, "learning_rate": 0.0001733126110124334, "loss": 0.1943, "step": 610 }, { "epoch": 0.4067232484606424, "grad_norm": 65.77499389648438, "learning_rate": 0.0001732682060390764, "loss": 0.379, "step": 611 }, { "epoch": 0.4073889166250624, "grad_norm": 110.7475814819336, "learning_rate": 0.00017322380106571936, "loss": 0.618, "step": 612 }, { "epoch": 0.4080545847894824, "grad_norm": 43.67226791381836, "learning_rate": 0.00017317939609236236, "loss": 0.2807, "step": 613 }, { "epoch": 0.40872025295390246, "grad_norm": 64.67245483398438, "learning_rate": 0.00017313499111900533, "loss": 0.4116, "step": 614 }, { "epoch": 0.4093859211183225, "grad_norm": 37.3428840637207, "learning_rate": 0.00017309058614564833, "loss": 0.7732, "step": 615 }, { "epoch": 0.41005158928274255, "grad_norm": 106.12779998779297, "learning_rate": 0.0001730461811722913, "loss": 0.7123, "step": 616 }, { "epoch": 0.4107172574471626, "grad_norm": 89.30298614501953, "learning_rate": 0.0001730017761989343, "loss": 0.4929, "step": 617 }, { "epoch": 0.41138292561158263, "grad_norm": 119.3560791015625, "learning_rate": 0.00017295737122557727, "loss": 1.2403, "step": 618 }, { "epoch": 0.41204859377600267, "grad_norm": 40.7421760559082, "learning_rate": 0.00017291296625222027, "loss": 0.2883, "step": 619 }, { "epoch": 0.4127142619404227, "grad_norm": 108.93859100341797, "learning_rate": 0.00017286856127886324, "loss": 0.829, "step": 620 }, { "epoch": 0.41337993010484275, "grad_norm": 67.42476654052734, "learning_rate": 0.00017282415630550624, "loss": 0.5903, "step": 621 }, { "epoch": 0.4140455982692628, "grad_norm": 46.429893493652344, "learning_rate": 0.0001727797513321492, "loss": 0.3084, "step": 622 }, { "epoch": 0.41471126643368283, "grad_norm": 82.67963409423828, "learning_rate": 0.00017273534635879218, "loss": 0.4691, "step": 623 }, { "epoch": 0.4153769345981028, "grad_norm": 93.61371612548828, "learning_rate": 0.00017269094138543518, "loss": 0.4056, "step": 624 }, { "epoch": 0.41604260276252286, "grad_norm": 133.24710083007812, "learning_rate": 0.00017264653641207815, "loss": 0.5845, "step": 625 }, { "epoch": 0.4167082709269429, "grad_norm": 65.41972351074219, "learning_rate": 0.00017260213143872115, "loss": 0.3826, "step": 626 }, { "epoch": 0.41737393909136294, "grad_norm": 103.12113189697266, "learning_rate": 0.00017255772646536412, "loss": 0.3715, "step": 627 }, { "epoch": 0.418039607255783, "grad_norm": 57.001365661621094, "learning_rate": 0.00017251332149200712, "loss": 0.5263, "step": 628 }, { "epoch": 0.418705275420203, "grad_norm": 67.60276794433594, "learning_rate": 0.0001724689165186501, "loss": 0.9148, "step": 629 }, { "epoch": 0.41937094358462307, "grad_norm": 42.20960235595703, "learning_rate": 0.0001724245115452931, "loss": 0.2265, "step": 630 }, { "epoch": 0.4200366117490431, "grad_norm": 61.50966262817383, "learning_rate": 0.00017238010657193606, "loss": 0.2724, "step": 631 }, { "epoch": 0.42070227991346315, "grad_norm": 154.45419311523438, "learning_rate": 0.00017233570159857906, "loss": 0.7941, "step": 632 }, { "epoch": 0.4213679480778832, "grad_norm": 50.674625396728516, "learning_rate": 0.00017229129662522203, "loss": 0.5776, "step": 633 }, { "epoch": 0.42203361624230323, "grad_norm": 93.19532012939453, "learning_rate": 0.00017224689165186503, "loss": 0.442, "step": 634 }, { "epoch": 0.4226992844067233, "grad_norm": 47.145652770996094, "learning_rate": 0.000172202486678508, "loss": 0.3244, "step": 635 }, { "epoch": 0.42336495257114326, "grad_norm": 54.49449157714844, "learning_rate": 0.00017215808170515097, "loss": 0.3336, "step": 636 }, { "epoch": 0.4240306207355633, "grad_norm": 57.44892120361328, "learning_rate": 0.00017211367673179397, "loss": 0.3192, "step": 637 }, { "epoch": 0.42469628889998334, "grad_norm": 32.6402587890625, "learning_rate": 0.00017206927175843694, "loss": 0.3978, "step": 638 }, { "epoch": 0.4253619570644034, "grad_norm": 62.53096389770508, "learning_rate": 0.00017202486678507994, "loss": 0.2868, "step": 639 }, { "epoch": 0.4260276252288234, "grad_norm": 29.77239227294922, "learning_rate": 0.0001719804618117229, "loss": 0.1912, "step": 640 }, { "epoch": 0.42669329339324347, "grad_norm": 23.30158042907715, "learning_rate": 0.0001719360568383659, "loss": 0.0966, "step": 641 }, { "epoch": 0.4273589615576635, "grad_norm": 74.98001861572266, "learning_rate": 0.00017189165186500888, "loss": 0.342, "step": 642 }, { "epoch": 0.42802462972208355, "grad_norm": 57.433895111083984, "learning_rate": 0.00017184724689165188, "loss": 0.1998, "step": 643 }, { "epoch": 0.4286902978865036, "grad_norm": 26.67135238647461, "learning_rate": 0.00017180284191829485, "loss": 0.2448, "step": 644 }, { "epoch": 0.42935596605092363, "grad_norm": 103.6978988647461, "learning_rate": 0.00017175843694493785, "loss": 0.5228, "step": 645 }, { "epoch": 0.4300216342153437, "grad_norm": 95.85309600830078, "learning_rate": 0.00017171403197158082, "loss": 0.4366, "step": 646 }, { "epoch": 0.4306873023797637, "grad_norm": 41.5027961730957, "learning_rate": 0.00017166962699822382, "loss": 0.2968, "step": 647 }, { "epoch": 0.4313529705441837, "grad_norm": 20.21183967590332, "learning_rate": 0.0001716252220248668, "loss": 0.055, "step": 648 }, { "epoch": 0.43201863870860374, "grad_norm": 74.72583770751953, "learning_rate": 0.00017158081705150976, "loss": 0.9381, "step": 649 }, { "epoch": 0.4326843068730238, "grad_norm": 77.46644592285156, "learning_rate": 0.00017153641207815276, "loss": 0.3423, "step": 650 }, { "epoch": 0.4333499750374438, "grad_norm": 37.057373046875, "learning_rate": 0.00017149200710479573, "loss": 0.6322, "step": 651 }, { "epoch": 0.43401564320186387, "grad_norm": 70.27009582519531, "learning_rate": 0.00017144760213143873, "loss": 0.3602, "step": 652 }, { "epoch": 0.4346813113662839, "grad_norm": 49.69029998779297, "learning_rate": 0.0001714031971580817, "loss": 0.2872, "step": 653 }, { "epoch": 0.43534697953070395, "grad_norm": 37.20336151123047, "learning_rate": 0.0001713587921847247, "loss": 0.2243, "step": 654 }, { "epoch": 0.436012647695124, "grad_norm": 46.541385650634766, "learning_rate": 0.00017131438721136767, "loss": 0.5596, "step": 655 }, { "epoch": 0.43667831585954403, "grad_norm": 96.3812026977539, "learning_rate": 0.00017126998223801067, "loss": 0.7193, "step": 656 }, { "epoch": 0.43734398402396407, "grad_norm": 39.73197555541992, "learning_rate": 0.00017122557726465364, "loss": 0.3243, "step": 657 }, { "epoch": 0.4380096521883841, "grad_norm": 64.15336608886719, "learning_rate": 0.00017118117229129664, "loss": 0.4194, "step": 658 }, { "epoch": 0.43867532035280415, "grad_norm": 38.59345245361328, "learning_rate": 0.0001711367673179396, "loss": 0.5887, "step": 659 }, { "epoch": 0.43934098851722414, "grad_norm": 20.948026657104492, "learning_rate": 0.00017109236234458258, "loss": 0.2583, "step": 660 }, { "epoch": 0.4400066566816442, "grad_norm": 50.797611236572266, "learning_rate": 0.00017104795737122558, "loss": 0.4461, "step": 661 }, { "epoch": 0.4406723248460642, "grad_norm": 90.01778411865234, "learning_rate": 0.00017100355239786855, "loss": 0.264, "step": 662 }, { "epoch": 0.44133799301048426, "grad_norm": 66.13937377929688, "learning_rate": 0.00017095914742451155, "loss": 0.9227, "step": 663 }, { "epoch": 0.4420036611749043, "grad_norm": 17.219995498657227, "learning_rate": 0.00017091474245115452, "loss": 0.1018, "step": 664 }, { "epoch": 0.44266932933932435, "grad_norm": 51.22856903076172, "learning_rate": 0.00017087033747779752, "loss": 0.2328, "step": 665 }, { "epoch": 0.4433349975037444, "grad_norm": 109.7330551147461, "learning_rate": 0.0001708259325044405, "loss": 0.6561, "step": 666 }, { "epoch": 0.44400066566816443, "grad_norm": 46.88761901855469, "learning_rate": 0.0001707815275310835, "loss": 0.1331, "step": 667 }, { "epoch": 0.44466633383258447, "grad_norm": 61.69406509399414, "learning_rate": 0.00017073712255772646, "loss": 0.4561, "step": 668 }, { "epoch": 0.4453320019970045, "grad_norm": 75.9193115234375, "learning_rate": 0.00017069271758436946, "loss": 0.3184, "step": 669 }, { "epoch": 0.44599767016142455, "grad_norm": 44.89881896972656, "learning_rate": 0.00017064831261101243, "loss": 0.3697, "step": 670 }, { "epoch": 0.44666333832584454, "grad_norm": 78.53019714355469, "learning_rate": 0.00017060390763765543, "loss": 0.7524, "step": 671 }, { "epoch": 0.4473290064902646, "grad_norm": 61.14173889160156, "learning_rate": 0.0001705595026642984, "loss": 0.465, "step": 672 }, { "epoch": 0.4479946746546846, "grad_norm": 30.314422607421875, "learning_rate": 0.00017051509769094137, "loss": 0.3398, "step": 673 }, { "epoch": 0.44866034281910466, "grad_norm": 68.44725799560547, "learning_rate": 0.00017047069271758437, "loss": 0.3817, "step": 674 }, { "epoch": 0.4493260109835247, "grad_norm": 66.30414581298828, "learning_rate": 0.00017042628774422734, "loss": 0.3968, "step": 675 }, { "epoch": 0.44999167914794475, "grad_norm": 192.29794311523438, "learning_rate": 0.00017038188277087034, "loss": 1.2593, "step": 676 }, { "epoch": 0.4506573473123648, "grad_norm": 90.28983306884766, "learning_rate": 0.0001703374777975133, "loss": 0.3282, "step": 677 }, { "epoch": 0.45132301547678483, "grad_norm": 24.365171432495117, "learning_rate": 0.0001702930728241563, "loss": 0.0878, "step": 678 }, { "epoch": 0.45198868364120487, "grad_norm": 57.6705322265625, "learning_rate": 0.00017024866785079928, "loss": 0.2243, "step": 679 }, { "epoch": 0.4526543518056249, "grad_norm": 83.08112335205078, "learning_rate": 0.00017020426287744228, "loss": 0.4434, "step": 680 }, { "epoch": 0.45332001997004495, "grad_norm": 66.54988098144531, "learning_rate": 0.00017015985790408525, "loss": 0.6353, "step": 681 }, { "epoch": 0.453985688134465, "grad_norm": 62.16245651245117, "learning_rate": 0.00017011545293072825, "loss": 0.5275, "step": 682 }, { "epoch": 0.454651356298885, "grad_norm": 119.9114761352539, "learning_rate": 0.00017007104795737122, "loss": 0.7318, "step": 683 }, { "epoch": 0.455317024463305, "grad_norm": 22.712766647338867, "learning_rate": 0.00017002664298401422, "loss": 0.1221, "step": 684 }, { "epoch": 0.45598269262772506, "grad_norm": 30.54381561279297, "learning_rate": 0.0001699822380106572, "loss": 0.1475, "step": 685 }, { "epoch": 0.4566483607921451, "grad_norm": 50.76087951660156, "learning_rate": 0.0001699378330373002, "loss": 0.2274, "step": 686 }, { "epoch": 0.45731402895656514, "grad_norm": 48.10842514038086, "learning_rate": 0.00016989342806394316, "loss": 0.2283, "step": 687 }, { "epoch": 0.4579796971209852, "grad_norm": 123.0356216430664, "learning_rate": 0.00016984902309058616, "loss": 0.3123, "step": 688 }, { "epoch": 0.4586453652854052, "grad_norm": 84.65762329101562, "learning_rate": 0.00016980461811722913, "loss": 0.2491, "step": 689 }, { "epoch": 0.45931103344982527, "grad_norm": 30.20061683654785, "learning_rate": 0.00016976021314387213, "loss": 0.0821, "step": 690 }, { "epoch": 0.4599767016142453, "grad_norm": 14.239197731018066, "learning_rate": 0.0001697158081705151, "loss": 0.0399, "step": 691 }, { "epoch": 0.46064236977866535, "grad_norm": 17.447071075439453, "learning_rate": 0.0001696714031971581, "loss": 0.1809, "step": 692 }, { "epoch": 0.4613080379430854, "grad_norm": 120.60710144042969, "learning_rate": 0.00016962699822380107, "loss": 0.6193, "step": 693 }, { "epoch": 0.46197370610750543, "grad_norm": 89.69656372070312, "learning_rate": 0.00016958259325044407, "loss": 0.7936, "step": 694 }, { "epoch": 0.4626393742719254, "grad_norm": 60.408546447753906, "learning_rate": 0.00016953818827708704, "loss": 0.2803, "step": 695 }, { "epoch": 0.46330504243634546, "grad_norm": 72.99823760986328, "learning_rate": 0.00016949378330373004, "loss": 0.436, "step": 696 }, { "epoch": 0.4639707106007655, "grad_norm": 117.71556091308594, "learning_rate": 0.000169449378330373, "loss": 1.5063, "step": 697 }, { "epoch": 0.46463637876518554, "grad_norm": 45.7595100402832, "learning_rate": 0.000169404973357016, "loss": 0.361, "step": 698 }, { "epoch": 0.4653020469296056, "grad_norm": 73.0155029296875, "learning_rate": 0.00016936056838365898, "loss": 0.81, "step": 699 }, { "epoch": 0.4659677150940256, "grad_norm": 99.09642791748047, "learning_rate": 0.00016931616341030198, "loss": 0.925, "step": 700 }, { "epoch": 0.46663338325844567, "grad_norm": 77.26228332519531, "learning_rate": 0.00016927175843694495, "loss": 0.5596, "step": 701 }, { "epoch": 0.4672990514228657, "grad_norm": 34.49489212036133, "learning_rate": 0.00016922735346358795, "loss": 0.2152, "step": 702 }, { "epoch": 0.46796471958728575, "grad_norm": 65.58250427246094, "learning_rate": 0.00016918294849023092, "loss": 0.3207, "step": 703 }, { "epoch": 0.4686303877517058, "grad_norm": 69.55067443847656, "learning_rate": 0.00016913854351687392, "loss": 0.4024, "step": 704 }, { "epoch": 0.46929605591612583, "grad_norm": 36.643272399902344, "learning_rate": 0.0001690941385435169, "loss": 0.3132, "step": 705 }, { "epoch": 0.4699617240805459, "grad_norm": 69.05078887939453, "learning_rate": 0.00016904973357015989, "loss": 0.4488, "step": 706 }, { "epoch": 0.47062739224496586, "grad_norm": 30.151975631713867, "learning_rate": 0.00016900532859680286, "loss": 0.3472, "step": 707 }, { "epoch": 0.4712930604093859, "grad_norm": 79.70284271240234, "learning_rate": 0.00016896092362344585, "loss": 0.6793, "step": 708 }, { "epoch": 0.47195872857380594, "grad_norm": 54.25303649902344, "learning_rate": 0.00016891651865008883, "loss": 0.2562, "step": 709 }, { "epoch": 0.472624396738226, "grad_norm": 62.36642837524414, "learning_rate": 0.0001688721136767318, "loss": 0.3446, "step": 710 }, { "epoch": 0.473290064902646, "grad_norm": 43.8344612121582, "learning_rate": 0.0001688277087033748, "loss": 0.3185, "step": 711 }, { "epoch": 0.47395573306706607, "grad_norm": 27.90556526184082, "learning_rate": 0.00016878330373001777, "loss": 0.2685, "step": 712 }, { "epoch": 0.4746214012314861, "grad_norm": 27.03131866455078, "learning_rate": 0.00016873889875666077, "loss": 0.0934, "step": 713 }, { "epoch": 0.47528706939590615, "grad_norm": 123.10067749023438, "learning_rate": 0.00016869449378330374, "loss": 0.7965, "step": 714 }, { "epoch": 0.4759527375603262, "grad_norm": 79.507568359375, "learning_rate": 0.00016865008880994673, "loss": 0.6379, "step": 715 }, { "epoch": 0.47661840572474623, "grad_norm": 62.209964752197266, "learning_rate": 0.0001686056838365897, "loss": 0.187, "step": 716 }, { "epoch": 0.47728407388916627, "grad_norm": 92.40242767333984, "learning_rate": 0.0001685612788632327, "loss": 0.4563, "step": 717 }, { "epoch": 0.4779497420535863, "grad_norm": 54.89033126831055, "learning_rate": 0.00016851687388987568, "loss": 0.3967, "step": 718 }, { "epoch": 0.4786154102180063, "grad_norm": 18.528785705566406, "learning_rate": 0.00016847246891651867, "loss": 0.0683, "step": 719 }, { "epoch": 0.47928107838242634, "grad_norm": 87.40145874023438, "learning_rate": 0.00016842806394316165, "loss": 0.7017, "step": 720 }, { "epoch": 0.4799467465468464, "grad_norm": 61.1949462890625, "learning_rate": 0.00016838365896980464, "loss": 0.7981, "step": 721 }, { "epoch": 0.4806124147112664, "grad_norm": 90.75920104980469, "learning_rate": 0.00016833925399644762, "loss": 0.8007, "step": 722 }, { "epoch": 0.48127808287568646, "grad_norm": 31.92529296875, "learning_rate": 0.0001682948490230906, "loss": 0.1569, "step": 723 }, { "epoch": 0.4819437510401065, "grad_norm": 81.36243438720703, "learning_rate": 0.00016825044404973358, "loss": 0.2877, "step": 724 }, { "epoch": 0.48260941920452655, "grad_norm": 45.11757278442383, "learning_rate": 0.00016820603907637656, "loss": 0.4982, "step": 725 }, { "epoch": 0.4832750873689466, "grad_norm": 47.99403762817383, "learning_rate": 0.00016816163410301955, "loss": 0.1855, "step": 726 }, { "epoch": 0.48394075553336663, "grad_norm": 103.36270141601562, "learning_rate": 0.00016811722912966253, "loss": 0.3909, "step": 727 }, { "epoch": 0.48460642369778667, "grad_norm": 11.007644653320312, "learning_rate": 0.00016807282415630552, "loss": 0.0956, "step": 728 }, { "epoch": 0.4852720918622067, "grad_norm": 51.31671905517578, "learning_rate": 0.0001680284191829485, "loss": 0.2221, "step": 729 }, { "epoch": 0.48593776002662675, "grad_norm": 36.494293212890625, "learning_rate": 0.0001679840142095915, "loss": 0.1623, "step": 730 }, { "epoch": 0.48660342819104674, "grad_norm": 100.02330017089844, "learning_rate": 0.00016793960923623447, "loss": 0.4674, "step": 731 }, { "epoch": 0.4872690963554668, "grad_norm": 41.67621612548828, "learning_rate": 0.00016789520426287746, "loss": 0.1248, "step": 732 }, { "epoch": 0.4879347645198868, "grad_norm": 114.75006103515625, "learning_rate": 0.00016785079928952043, "loss": 0.4371, "step": 733 }, { "epoch": 0.48860043268430686, "grad_norm": 41.169837951660156, "learning_rate": 0.00016780639431616343, "loss": 0.1483, "step": 734 }, { "epoch": 0.4892661008487269, "grad_norm": 77.92648315429688, "learning_rate": 0.0001677619893428064, "loss": 0.4915, "step": 735 }, { "epoch": 0.48993176901314694, "grad_norm": 97.74514770507812, "learning_rate": 0.00016771758436944938, "loss": 0.3256, "step": 736 }, { "epoch": 0.490597437177567, "grad_norm": 63.13382339477539, "learning_rate": 0.00016767317939609237, "loss": 0.4301, "step": 737 }, { "epoch": 0.491263105341987, "grad_norm": 104.03643798828125, "learning_rate": 0.00016762877442273535, "loss": 0.6005, "step": 738 }, { "epoch": 0.49192877350640707, "grad_norm": 102.05120086669922, "learning_rate": 0.00016758436944937834, "loss": 1.173, "step": 739 }, { "epoch": 0.4925944416708271, "grad_norm": 75.59339141845703, "learning_rate": 0.00016753996447602132, "loss": 0.7547, "step": 740 }, { "epoch": 0.49326010983524715, "grad_norm": 59.76469802856445, "learning_rate": 0.00016749555950266431, "loss": 0.302, "step": 741 }, { "epoch": 0.4939257779996672, "grad_norm": 69.68283081054688, "learning_rate": 0.00016745115452930728, "loss": 0.4141, "step": 742 }, { "epoch": 0.4945914461640872, "grad_norm": 8.145694732666016, "learning_rate": 0.00016740674955595028, "loss": 0.0285, "step": 743 }, { "epoch": 0.4952571143285072, "grad_norm": 32.968955993652344, "learning_rate": 0.00016736234458259325, "loss": 0.1136, "step": 744 }, { "epoch": 0.49592278249292726, "grad_norm": 73.82479095458984, "learning_rate": 0.00016731793960923625, "loss": 0.261, "step": 745 }, { "epoch": 0.4965884506573473, "grad_norm": 108.1319808959961, "learning_rate": 0.00016727353463587922, "loss": 0.6509, "step": 746 }, { "epoch": 0.49725411882176734, "grad_norm": 129.9436798095703, "learning_rate": 0.00016722912966252222, "loss": 1.5656, "step": 747 }, { "epoch": 0.4979197869861874, "grad_norm": 42.64841842651367, "learning_rate": 0.0001671847246891652, "loss": 0.0873, "step": 748 }, { "epoch": 0.4985854551506074, "grad_norm": 74.73552703857422, "learning_rate": 0.00016714031971580817, "loss": 0.4388, "step": 749 }, { "epoch": 0.49925112331502747, "grad_norm": 97.19401550292969, "learning_rate": 0.00016709591474245116, "loss": 0.6456, "step": 750 }, { "epoch": 0.4999167914794475, "grad_norm": 31.533082962036133, "learning_rate": 0.00016705150976909413, "loss": 0.123, "step": 751 }, { "epoch": 0.5005824596438675, "grad_norm": 62.29495620727539, "learning_rate": 0.00016700710479573713, "loss": 0.4031, "step": 752 }, { "epoch": 0.5012481278082875, "grad_norm": 49.27903366088867, "learning_rate": 0.0001669626998223801, "loss": 0.3012, "step": 753 }, { "epoch": 0.5019137959727076, "grad_norm": 24.742130279541016, "learning_rate": 0.0001669182948490231, "loss": 0.0864, "step": 754 }, { "epoch": 0.5025794641371276, "grad_norm": 71.86795806884766, "learning_rate": 0.00016687388987566607, "loss": 1.0555, "step": 755 }, { "epoch": 0.5032451323015477, "grad_norm": 26.747037887573242, "learning_rate": 0.00016682948490230907, "loss": 0.1599, "step": 756 }, { "epoch": 0.5039108004659677, "grad_norm": 71.43767547607422, "learning_rate": 0.00016678507992895204, "loss": 0.255, "step": 757 }, { "epoch": 0.5045764686303877, "grad_norm": 45.117950439453125, "learning_rate": 0.00016674067495559504, "loss": 0.4431, "step": 758 }, { "epoch": 0.5052421367948078, "grad_norm": 44.90398025512695, "learning_rate": 0.00016669626998223801, "loss": 0.291, "step": 759 }, { "epoch": 0.5059078049592278, "grad_norm": 24.768451690673828, "learning_rate": 0.00016665186500888098, "loss": 0.109, "step": 760 }, { "epoch": 0.5065734731236479, "grad_norm": 46.24464797973633, "learning_rate": 0.00016660746003552398, "loss": 0.3432, "step": 761 }, { "epoch": 0.5072391412880679, "grad_norm": 19.104442596435547, "learning_rate": 0.00016656305506216695, "loss": 0.091, "step": 762 }, { "epoch": 0.507904809452488, "grad_norm": 95.51792907714844, "learning_rate": 0.00016651865008880995, "loss": 0.5077, "step": 763 }, { "epoch": 0.508570477616908, "grad_norm": 33.28221130371094, "learning_rate": 0.00016647424511545292, "loss": 0.1207, "step": 764 }, { "epoch": 0.509236145781328, "grad_norm": 48.90406036376953, "learning_rate": 0.00016642984014209592, "loss": 0.4793, "step": 765 }, { "epoch": 0.5099018139457481, "grad_norm": 29.814743041992188, "learning_rate": 0.0001663854351687389, "loss": 0.0951, "step": 766 }, { "epoch": 0.5105674821101681, "grad_norm": 96.55838012695312, "learning_rate": 0.0001663410301953819, "loss": 1.0379, "step": 767 }, { "epoch": 0.5112331502745882, "grad_norm": 100.2326431274414, "learning_rate": 0.00016629662522202486, "loss": 0.5686, "step": 768 }, { "epoch": 0.5118988184390082, "grad_norm": 81.53526306152344, "learning_rate": 0.00016625222024866786, "loss": 0.4334, "step": 769 }, { "epoch": 0.5125644866034282, "grad_norm": 50.92794418334961, "learning_rate": 0.00016620781527531083, "loss": 0.4187, "step": 770 }, { "epoch": 0.5132301547678483, "grad_norm": 37.605533599853516, "learning_rate": 0.00016616341030195383, "loss": 0.1725, "step": 771 }, { "epoch": 0.5138958229322682, "grad_norm": 48.62330627441406, "learning_rate": 0.0001661190053285968, "loss": 0.226, "step": 772 }, { "epoch": 0.5145614910966883, "grad_norm": 131.95993041992188, "learning_rate": 0.00016607460035523977, "loss": 0.9214, "step": 773 }, { "epoch": 0.5152271592611083, "grad_norm": 59.95537567138672, "learning_rate": 0.00016603019538188277, "loss": 0.4827, "step": 774 }, { "epoch": 0.5158928274255283, "grad_norm": 126.53961181640625, "learning_rate": 0.00016598579040852574, "loss": 0.6684, "step": 775 }, { "epoch": 0.5165584955899484, "grad_norm": 88.98424530029297, "learning_rate": 0.00016594138543516874, "loss": 0.5835, "step": 776 }, { "epoch": 0.5172241637543684, "grad_norm": 63.73120880126953, "learning_rate": 0.0001658969804618117, "loss": 0.3389, "step": 777 }, { "epoch": 0.5178898319187885, "grad_norm": 73.84934997558594, "learning_rate": 0.0001658525754884547, "loss": 0.6063, "step": 778 }, { "epoch": 0.5185555000832085, "grad_norm": 47.694114685058594, "learning_rate": 0.00016580817051509768, "loss": 0.6936, "step": 779 }, { "epoch": 0.5192211682476285, "grad_norm": 50.0419921875, "learning_rate": 0.00016576376554174068, "loss": 0.3097, "step": 780 }, { "epoch": 0.5198868364120486, "grad_norm": 36.88435745239258, "learning_rate": 0.00016571936056838365, "loss": 0.2563, "step": 781 }, { "epoch": 0.5205525045764686, "grad_norm": 82.27161407470703, "learning_rate": 0.00016567495559502665, "loss": 0.4158, "step": 782 }, { "epoch": 0.5212181727408887, "grad_norm": 46.550270080566406, "learning_rate": 0.00016563055062166962, "loss": 0.2749, "step": 783 }, { "epoch": 0.5218838409053087, "grad_norm": 69.61832427978516, "learning_rate": 0.00016558614564831262, "loss": 0.3109, "step": 784 }, { "epoch": 0.5225495090697287, "grad_norm": 60.874752044677734, "learning_rate": 0.0001655417406749556, "loss": 0.7005, "step": 785 }, { "epoch": 0.5232151772341488, "grad_norm": 33.250511169433594, "learning_rate": 0.00016549733570159856, "loss": 0.6432, "step": 786 }, { "epoch": 0.5238808453985688, "grad_norm": 116.75189971923828, "learning_rate": 0.00016545293072824156, "loss": 0.5193, "step": 787 }, { "epoch": 0.5245465135629889, "grad_norm": 88.6585922241211, "learning_rate": 0.00016540852575488456, "loss": 0.4463, "step": 788 }, { "epoch": 0.5252121817274089, "grad_norm": 29.097097396850586, "learning_rate": 0.00016536412078152753, "loss": 0.1297, "step": 789 }, { "epoch": 0.525877849891829, "grad_norm": 28.999595642089844, "learning_rate": 0.00016531971580817053, "loss": 0.1885, "step": 790 }, { "epoch": 0.526543518056249, "grad_norm": 66.06973266601562, "learning_rate": 0.0001652753108348135, "loss": 0.7004, "step": 791 }, { "epoch": 0.527209186220669, "grad_norm": 31.66838836669922, "learning_rate": 0.0001652309058614565, "loss": 0.1691, "step": 792 }, { "epoch": 0.5278748543850891, "grad_norm": 23.52103042602539, "learning_rate": 0.00016518650088809947, "loss": 0.3266, "step": 793 }, { "epoch": 0.5285405225495091, "grad_norm": 49.926326751708984, "learning_rate": 0.00016514209591474247, "loss": 0.4661, "step": 794 }, { "epoch": 0.5292061907139292, "grad_norm": 38.95738220214844, "learning_rate": 0.00016509769094138544, "loss": 0.5901, "step": 795 }, { "epoch": 0.5298718588783491, "grad_norm": 79.33973693847656, "learning_rate": 0.00016505328596802844, "loss": 0.4434, "step": 796 }, { "epoch": 0.5305375270427691, "grad_norm": 59.31830596923828, "learning_rate": 0.0001650088809946714, "loss": 0.4059, "step": 797 }, { "epoch": 0.5312031952071892, "grad_norm": 70.5051498413086, "learning_rate": 0.0001649644760213144, "loss": 0.4051, "step": 798 }, { "epoch": 0.5318688633716092, "grad_norm": 38.054168701171875, "learning_rate": 0.00016492007104795738, "loss": 0.3508, "step": 799 }, { "epoch": 0.5325345315360293, "grad_norm": 42.29998016357422, "learning_rate": 0.00016487566607460038, "loss": 0.1866, "step": 800 }, { "epoch": 0.5332001997004493, "grad_norm": 36.5222282409668, "learning_rate": 0.00016483126110124335, "loss": 0.436, "step": 801 }, { "epoch": 0.5338658678648693, "grad_norm": 23.83983612060547, "learning_rate": 0.00016478685612788635, "loss": 0.0762, "step": 802 }, { "epoch": 0.5345315360292894, "grad_norm": 36.12353515625, "learning_rate": 0.00016474245115452932, "loss": 0.1777, "step": 803 }, { "epoch": 0.5351972041937094, "grad_norm": 47.710758209228516, "learning_rate": 0.00016469804618117232, "loss": 0.264, "step": 804 }, { "epoch": 0.5358628723581295, "grad_norm": 59.07710266113281, "learning_rate": 0.0001646536412078153, "loss": 0.48, "step": 805 }, { "epoch": 0.5365285405225495, "grad_norm": 31.360429763793945, "learning_rate": 0.0001646092362344583, "loss": 0.1614, "step": 806 }, { "epoch": 0.5371942086869695, "grad_norm": 82.94551086425781, "learning_rate": 0.00016456483126110126, "loss": 0.6375, "step": 807 }, { "epoch": 0.5378598768513896, "grad_norm": 97.9576644897461, "learning_rate": 0.00016452042628774426, "loss": 0.6941, "step": 808 }, { "epoch": 0.5385255450158096, "grad_norm": 55.905860900878906, "learning_rate": 0.00016447602131438723, "loss": 0.4596, "step": 809 }, { "epoch": 0.5391912131802297, "grad_norm": 157.75843811035156, "learning_rate": 0.0001644316163410302, "loss": 0.6038, "step": 810 }, { "epoch": 0.5398568813446497, "grad_norm": 115.88114929199219, "learning_rate": 0.0001643872113676732, "loss": 0.7865, "step": 811 }, { "epoch": 0.5405225495090697, "grad_norm": 50.640113830566406, "learning_rate": 0.00016434280639431617, "loss": 0.5734, "step": 812 }, { "epoch": 0.5411882176734898, "grad_norm": 51.00415802001953, "learning_rate": 0.00016429840142095917, "loss": 0.1352, "step": 813 }, { "epoch": 0.5418538858379098, "grad_norm": 73.97732543945312, "learning_rate": 0.00016425399644760214, "loss": 0.3152, "step": 814 }, { "epoch": 0.5425195540023299, "grad_norm": 28.134428024291992, "learning_rate": 0.00016420959147424514, "loss": 0.2167, "step": 815 }, { "epoch": 0.5431852221667499, "grad_norm": 45.735347747802734, "learning_rate": 0.0001641651865008881, "loss": 0.239, "step": 816 }, { "epoch": 0.54385089033117, "grad_norm": 113.84656524658203, "learning_rate": 0.0001641207815275311, "loss": 0.8154, "step": 817 }, { "epoch": 0.54451655849559, "grad_norm": 33.40947723388672, "learning_rate": 0.00016407637655417408, "loss": 0.264, "step": 818 }, { "epoch": 0.54518222666001, "grad_norm": 63.0031623840332, "learning_rate": 0.00016403197158081708, "loss": 0.4459, "step": 819 }, { "epoch": 0.54584789482443, "grad_norm": 49.237754821777344, "learning_rate": 0.00016398756660746005, "loss": 0.2641, "step": 820 }, { "epoch": 0.54651356298885, "grad_norm": 36.26042175292969, "learning_rate": 0.00016394316163410305, "loss": 0.2145, "step": 821 }, { "epoch": 0.54717923115327, "grad_norm": 66.88060760498047, "learning_rate": 0.00016389875666074602, "loss": 0.5929, "step": 822 }, { "epoch": 0.5478448993176901, "grad_norm": 80.92005157470703, "learning_rate": 0.000163854351687389, "loss": 0.4821, "step": 823 }, { "epoch": 0.5485105674821101, "grad_norm": 46.16059494018555, "learning_rate": 0.000163809946714032, "loss": 0.1213, "step": 824 }, { "epoch": 0.5491762356465302, "grad_norm": 86.4364013671875, "learning_rate": 0.00016376554174067496, "loss": 0.5759, "step": 825 }, { "epoch": 0.5498419038109502, "grad_norm": 70.4107437133789, "learning_rate": 0.00016372113676731796, "loss": 0.4036, "step": 826 }, { "epoch": 0.5505075719753703, "grad_norm": 129.24647521972656, "learning_rate": 0.00016367673179396093, "loss": 0.6871, "step": 827 }, { "epoch": 0.5511732401397903, "grad_norm": 34.301849365234375, "learning_rate": 0.00016363232682060393, "loss": 0.2726, "step": 828 }, { "epoch": 0.5518389083042103, "grad_norm": 57.288421630859375, "learning_rate": 0.0001635879218472469, "loss": 0.1683, "step": 829 }, { "epoch": 0.5525045764686304, "grad_norm": 71.24075317382812, "learning_rate": 0.0001635435168738899, "loss": 0.4904, "step": 830 }, { "epoch": 0.5531702446330504, "grad_norm": 37.57721710205078, "learning_rate": 0.00016349911190053287, "loss": 0.1642, "step": 831 }, { "epoch": 0.5538359127974705, "grad_norm": 34.01001739501953, "learning_rate": 0.00016345470692717587, "loss": 0.3038, "step": 832 }, { "epoch": 0.5545015809618905, "grad_norm": 41.36484909057617, "learning_rate": 0.00016341030195381884, "loss": 0.5156, "step": 833 }, { "epoch": 0.5551672491263105, "grad_norm": 24.977052688598633, "learning_rate": 0.00016336589698046184, "loss": 0.107, "step": 834 }, { "epoch": 0.5558329172907306, "grad_norm": 28.269882202148438, "learning_rate": 0.0001633214920071048, "loss": 0.1388, "step": 835 }, { "epoch": 0.5564985854551506, "grad_norm": 91.3974380493164, "learning_rate": 0.00016327708703374778, "loss": 0.2595, "step": 836 }, { "epoch": 0.5571642536195707, "grad_norm": 60.29289245605469, "learning_rate": 0.00016323268206039078, "loss": 0.4556, "step": 837 }, { "epoch": 0.5578299217839907, "grad_norm": 73.14957427978516, "learning_rate": 0.00016318827708703375, "loss": 0.5624, "step": 838 }, { "epoch": 0.5584955899484108, "grad_norm": 56.330039978027344, "learning_rate": 0.00016314387211367675, "loss": 0.3286, "step": 839 }, { "epoch": 0.5591612581128308, "grad_norm": 41.747867584228516, "learning_rate": 0.00016309946714031972, "loss": 0.2241, "step": 840 }, { "epoch": 0.5598269262772508, "grad_norm": 111.82936096191406, "learning_rate": 0.00016305506216696272, "loss": 0.6143, "step": 841 }, { "epoch": 0.5604925944416709, "grad_norm": 63.755306243896484, "learning_rate": 0.0001630106571936057, "loss": 0.4782, "step": 842 }, { "epoch": 0.5611582626060909, "grad_norm": 58.69625473022461, "learning_rate": 0.00016296625222024869, "loss": 0.1724, "step": 843 }, { "epoch": 0.5618239307705108, "grad_norm": 31.554664611816406, "learning_rate": 0.00016292184724689166, "loss": 0.2879, "step": 844 }, { "epoch": 0.5624895989349309, "grad_norm": 49.19903564453125, "learning_rate": 0.00016287744227353466, "loss": 0.3343, "step": 845 }, { "epoch": 0.5631552670993509, "grad_norm": 48.23353576660156, "learning_rate": 0.00016283303730017763, "loss": 0.3251, "step": 846 }, { "epoch": 0.563820935263771, "grad_norm": 73.30461883544922, "learning_rate": 0.00016278863232682062, "loss": 0.5211, "step": 847 }, { "epoch": 0.564486603428191, "grad_norm": 74.00879669189453, "learning_rate": 0.0001627442273534636, "loss": 0.2641, "step": 848 }, { "epoch": 0.565152271592611, "grad_norm": 18.06185531616211, "learning_rate": 0.00016269982238010657, "loss": 0.0765, "step": 849 }, { "epoch": 0.5658179397570311, "grad_norm": 38.10951614379883, "learning_rate": 0.00016265541740674957, "loss": 0.203, "step": 850 }, { "epoch": 0.5664836079214511, "grad_norm": 53.78894805908203, "learning_rate": 0.00016261101243339254, "loss": 0.2008, "step": 851 }, { "epoch": 0.5671492760858712, "grad_norm": 65.65164947509766, "learning_rate": 0.00016256660746003554, "loss": 0.4519, "step": 852 }, { "epoch": 0.5678149442502912, "grad_norm": 68.0468521118164, "learning_rate": 0.0001625222024866785, "loss": 0.282, "step": 853 }, { "epoch": 0.5684806124147113, "grad_norm": 63.54074478149414, "learning_rate": 0.0001624777975133215, "loss": 0.5825, "step": 854 }, { "epoch": 0.5691462805791313, "grad_norm": 92.11983489990234, "learning_rate": 0.00016243339253996448, "loss": 0.685, "step": 855 }, { "epoch": 0.5698119487435513, "grad_norm": 16.614620208740234, "learning_rate": 0.00016238898756660747, "loss": 0.2696, "step": 856 }, { "epoch": 0.5704776169079714, "grad_norm": 26.756486892700195, "learning_rate": 0.00016234458259325045, "loss": 0.2283, "step": 857 }, { "epoch": 0.5711432850723914, "grad_norm": 48.378456115722656, "learning_rate": 0.00016230017761989344, "loss": 1.2203, "step": 858 }, { "epoch": 0.5718089532368115, "grad_norm": 70.19944763183594, "learning_rate": 0.00016225577264653642, "loss": 0.6875, "step": 859 }, { "epoch": 0.5724746214012315, "grad_norm": 41.7786750793457, "learning_rate": 0.0001622113676731794, "loss": 0.2601, "step": 860 }, { "epoch": 0.5731402895656516, "grad_norm": 32.285614013671875, "learning_rate": 0.00016216696269982239, "loss": 0.2627, "step": 861 }, { "epoch": 0.5738059577300716, "grad_norm": 43.81740951538086, "learning_rate": 0.00016212255772646536, "loss": 0.3652, "step": 862 }, { "epoch": 0.5744716258944916, "grad_norm": 61.80424118041992, "learning_rate": 0.00016207815275310836, "loss": 0.2883, "step": 863 }, { "epoch": 0.5751372940589117, "grad_norm": 59.186954498291016, "learning_rate": 0.00016203374777975133, "loss": 0.3879, "step": 864 }, { "epoch": 0.5758029622233317, "grad_norm": 36.24767303466797, "learning_rate": 0.00016198934280639432, "loss": 0.338, "step": 865 }, { "epoch": 0.5764686303877518, "grad_norm": 53.11578369140625, "learning_rate": 0.0001619449378330373, "loss": 0.5221, "step": 866 }, { "epoch": 0.5771342985521717, "grad_norm": 85.61437225341797, "learning_rate": 0.0001619005328596803, "loss": 0.5565, "step": 867 }, { "epoch": 0.5777999667165917, "grad_norm": 53.73318862915039, "learning_rate": 0.00016185612788632327, "loss": 0.3819, "step": 868 }, { "epoch": 0.5784656348810118, "grad_norm": 107.41029357910156, "learning_rate": 0.00016181172291296626, "loss": 0.6618, "step": 869 }, { "epoch": 0.5791313030454318, "grad_norm": 61.5377197265625, "learning_rate": 0.00016176731793960924, "loss": 0.3592, "step": 870 }, { "epoch": 0.5797969712098519, "grad_norm": 39.31349182128906, "learning_rate": 0.00016172291296625223, "loss": 0.1912, "step": 871 }, { "epoch": 0.5804626393742719, "grad_norm": 67.16478729248047, "learning_rate": 0.0001616785079928952, "loss": 0.5373, "step": 872 }, { "epoch": 0.5811283075386919, "grad_norm": 65.04055786132812, "learning_rate": 0.00016163410301953818, "loss": 0.5011, "step": 873 }, { "epoch": 0.581793975703112, "grad_norm": 30.299055099487305, "learning_rate": 0.00016158969804618117, "loss": 0.4038, "step": 874 }, { "epoch": 0.582459643867532, "grad_norm": 37.89064025878906, "learning_rate": 0.00016154529307282415, "loss": 0.3733, "step": 875 }, { "epoch": 0.5831253120319521, "grad_norm": 35.79871368408203, "learning_rate": 0.00016150088809946714, "loss": 0.4639, "step": 876 }, { "epoch": 0.5837909801963721, "grad_norm": 41.761165618896484, "learning_rate": 0.00016145648312611012, "loss": 0.4087, "step": 877 }, { "epoch": 0.5844566483607921, "grad_norm": 62.74789047241211, "learning_rate": 0.00016141207815275311, "loss": 0.4322, "step": 878 }, { "epoch": 0.5851223165252122, "grad_norm": 66.2603759765625, "learning_rate": 0.00016136767317939609, "loss": 0.2986, "step": 879 }, { "epoch": 0.5857879846896322, "grad_norm": 38.21004867553711, "learning_rate": 0.00016132326820603908, "loss": 0.2088, "step": 880 }, { "epoch": 0.5864536528540523, "grad_norm": 24.51487922668457, "learning_rate": 0.00016127886323268206, "loss": 0.097, "step": 881 }, { "epoch": 0.5871193210184723, "grad_norm": 27.898771286010742, "learning_rate": 0.00016123445825932505, "loss": 0.1775, "step": 882 }, { "epoch": 0.5877849891828923, "grad_norm": 46.40385818481445, "learning_rate": 0.00016119005328596802, "loss": 0.2212, "step": 883 }, { "epoch": 0.5884506573473124, "grad_norm": 56.09572982788086, "learning_rate": 0.00016114564831261102, "loss": 0.4678, "step": 884 }, { "epoch": 0.5891163255117324, "grad_norm": 129.60107421875, "learning_rate": 0.000161101243339254, "loss": 0.809, "step": 885 }, { "epoch": 0.5897819936761525, "grad_norm": 93.75535583496094, "learning_rate": 0.00016105683836589697, "loss": 1.2058, "step": 886 }, { "epoch": 0.5904476618405725, "grad_norm": 109.20541381835938, "learning_rate": 0.00016101243339253996, "loss": 0.6219, "step": 887 }, { "epoch": 0.5911133300049926, "grad_norm": 58.076332092285156, "learning_rate": 0.00016096802841918294, "loss": 0.4415, "step": 888 }, { "epoch": 0.5917789981694126, "grad_norm": 91.718017578125, "learning_rate": 0.00016092362344582593, "loss": 0.4294, "step": 889 }, { "epoch": 0.5924446663338326, "grad_norm": 110.23320770263672, "learning_rate": 0.0001608792184724689, "loss": 0.7701, "step": 890 }, { "epoch": 0.5931103344982526, "grad_norm": 84.98243713378906, "learning_rate": 0.0001608348134991119, "loss": 0.5707, "step": 891 }, { "epoch": 0.5937760026626726, "grad_norm": 82.67597961425781, "learning_rate": 0.00016079040852575487, "loss": 0.7084, "step": 892 }, { "epoch": 0.5944416708270926, "grad_norm": 90.17684936523438, "learning_rate": 0.00016074600355239787, "loss": 0.4872, "step": 893 }, { "epoch": 0.5951073389915127, "grad_norm": 112.2318344116211, "learning_rate": 0.00016070159857904084, "loss": 0.8127, "step": 894 }, { "epoch": 0.5957730071559327, "grad_norm": 16.132047653198242, "learning_rate": 0.00016065719360568384, "loss": 0.0944, "step": 895 }, { "epoch": 0.5964386753203528, "grad_norm": 69.02125549316406, "learning_rate": 0.00016061278863232681, "loss": 0.2236, "step": 896 }, { "epoch": 0.5971043434847728, "grad_norm": 29.546409606933594, "learning_rate": 0.0001605683836589698, "loss": 0.1212, "step": 897 }, { "epoch": 0.5977700116491929, "grad_norm": 54.70003890991211, "learning_rate": 0.00016052397868561278, "loss": 0.3831, "step": 898 }, { "epoch": 0.5984356798136129, "grad_norm": 65.44963073730469, "learning_rate": 0.00016047957371225578, "loss": 0.4181, "step": 899 }, { "epoch": 0.5991013479780329, "grad_norm": 109.88179779052734, "learning_rate": 0.00016043516873889875, "loss": 0.6356, "step": 900 }, { "epoch": 0.599767016142453, "grad_norm": 64.11637115478516, "learning_rate": 0.00016039076376554175, "loss": 0.708, "step": 901 }, { "epoch": 0.600432684306873, "grad_norm": 23.20315170288086, "learning_rate": 0.00016034635879218472, "loss": 0.1969, "step": 902 }, { "epoch": 0.6010983524712931, "grad_norm": 45.66897201538086, "learning_rate": 0.00016030195381882772, "loss": 0.5047, "step": 903 }, { "epoch": 0.6017640206357131, "grad_norm": 27.65083122253418, "learning_rate": 0.0001602575488454707, "loss": 0.1068, "step": 904 }, { "epoch": 0.6024296888001331, "grad_norm": 49.640907287597656, "learning_rate": 0.0001602131438721137, "loss": 0.533, "step": 905 }, { "epoch": 0.6030953569645532, "grad_norm": 79.02021026611328, "learning_rate": 0.0001601687388987567, "loss": 0.9014, "step": 906 }, { "epoch": 0.6037610251289732, "grad_norm": 49.97978591918945, "learning_rate": 0.00016012433392539966, "loss": 0.5546, "step": 907 }, { "epoch": 0.6044266932933933, "grad_norm": 70.31575775146484, "learning_rate": 0.00016007992895204266, "loss": 0.3117, "step": 908 }, { "epoch": 0.6050923614578133, "grad_norm": 53.54672622680664, "learning_rate": 0.00016003552397868563, "loss": 0.8231, "step": 909 }, { "epoch": 0.6057580296222334, "grad_norm": 45.44542694091797, "learning_rate": 0.0001599911190053286, "loss": 0.222, "step": 910 }, { "epoch": 0.6064236977866534, "grad_norm": 27.014482498168945, "learning_rate": 0.0001599467140319716, "loss": 0.2923, "step": 911 }, { "epoch": 0.6070893659510734, "grad_norm": 33.17491149902344, "learning_rate": 0.00015990230905861457, "loss": 0.3998, "step": 912 }, { "epoch": 0.6077550341154935, "grad_norm": 30.22976303100586, "learning_rate": 0.00015985790408525757, "loss": 0.22, "step": 913 }, { "epoch": 0.6084207022799135, "grad_norm": 92.3844223022461, "learning_rate": 0.00015981349911190054, "loss": 0.5543, "step": 914 }, { "epoch": 0.6090863704443334, "grad_norm": 55.11675262451172, "learning_rate": 0.00015976909413854354, "loss": 0.4034, "step": 915 }, { "epoch": 0.6097520386087535, "grad_norm": 51.71201705932617, "learning_rate": 0.0001597246891651865, "loss": 0.4482, "step": 916 }, { "epoch": 0.6104177067731735, "grad_norm": 41.75859069824219, "learning_rate": 0.0001596802841918295, "loss": 0.2557, "step": 917 }, { "epoch": 0.6110833749375936, "grad_norm": 47.229122161865234, "learning_rate": 0.00015963587921847248, "loss": 0.1506, "step": 918 }, { "epoch": 0.6117490431020136, "grad_norm": 29.504947662353516, "learning_rate": 0.00015959147424511548, "loss": 0.1791, "step": 919 }, { "epoch": 0.6124147112664337, "grad_norm": 48.473594665527344, "learning_rate": 0.00015954706927175845, "loss": 0.1924, "step": 920 }, { "epoch": 0.6130803794308537, "grad_norm": 24.240800857543945, "learning_rate": 0.00015950266429840145, "loss": 0.1287, "step": 921 }, { "epoch": 0.6137460475952737, "grad_norm": 55.36631393432617, "learning_rate": 0.00015945825932504442, "loss": 0.3015, "step": 922 }, { "epoch": 0.6144117157596938, "grad_norm": 24.28390884399414, "learning_rate": 0.0001594138543516874, "loss": 0.1358, "step": 923 }, { "epoch": 0.6150773839241138, "grad_norm": 15.040701866149902, "learning_rate": 0.0001593694493783304, "loss": 0.078, "step": 924 }, { "epoch": 0.6157430520885339, "grad_norm": 80.5035171508789, "learning_rate": 0.00015932504440497336, "loss": 0.3659, "step": 925 }, { "epoch": 0.6164087202529539, "grad_norm": 14.009264945983887, "learning_rate": 0.00015928063943161636, "loss": 0.059, "step": 926 }, { "epoch": 0.6170743884173739, "grad_norm": 41.53221130371094, "learning_rate": 0.00015923623445825933, "loss": 0.2141, "step": 927 }, { "epoch": 0.617740056581794, "grad_norm": 56.1119499206543, "learning_rate": 0.00015919182948490233, "loss": 0.352, "step": 928 }, { "epoch": 0.618405724746214, "grad_norm": 150.60353088378906, "learning_rate": 0.0001591474245115453, "loss": 0.8479, "step": 929 }, { "epoch": 0.6190713929106341, "grad_norm": 37.60380554199219, "learning_rate": 0.0001591030195381883, "loss": 0.2058, "step": 930 }, { "epoch": 0.6197370610750541, "grad_norm": 86.55335998535156, "learning_rate": 0.00015905861456483127, "loss": 1.5609, "step": 931 }, { "epoch": 0.6204027292394741, "grad_norm": 33.71519088745117, "learning_rate": 0.00015901420959147427, "loss": 0.2137, "step": 932 }, { "epoch": 0.6210683974038942, "grad_norm": 120.12718200683594, "learning_rate": 0.00015896980461811724, "loss": 0.9256, "step": 933 }, { "epoch": 0.6217340655683142, "grad_norm": 77.55552673339844, "learning_rate": 0.00015892539964476024, "loss": 0.3218, "step": 934 }, { "epoch": 0.6223997337327343, "grad_norm": 72.75386047363281, "learning_rate": 0.0001588809946714032, "loss": 0.6255, "step": 935 }, { "epoch": 0.6230654018971543, "grad_norm": 28.577308654785156, "learning_rate": 0.00015883658969804618, "loss": 0.5123, "step": 936 }, { "epoch": 0.6237310700615744, "grad_norm": 64.8055419921875, "learning_rate": 0.00015879218472468918, "loss": 0.2593, "step": 937 }, { "epoch": 0.6243967382259944, "grad_norm": 32.127960205078125, "learning_rate": 0.00015874777975133215, "loss": 0.0822, "step": 938 }, { "epoch": 0.6250624063904143, "grad_norm": 41.353248596191406, "learning_rate": 0.00015870337477797515, "loss": 0.1516, "step": 939 }, { "epoch": 0.6257280745548344, "grad_norm": 22.283430099487305, "learning_rate": 0.00015865896980461812, "loss": 0.2439, "step": 940 }, { "epoch": 0.6263937427192544, "grad_norm": 87.97146606445312, "learning_rate": 0.00015861456483126112, "loss": 1.0962, "step": 941 }, { "epoch": 0.6270594108836745, "grad_norm": 173.7371063232422, "learning_rate": 0.0001585701598579041, "loss": 1.3254, "step": 942 }, { "epoch": 0.6277250790480945, "grad_norm": 176.40904235839844, "learning_rate": 0.0001585257548845471, "loss": 0.9746, "step": 943 }, { "epoch": 0.6283907472125145, "grad_norm": 18.358781814575195, "learning_rate": 0.00015848134991119006, "loss": 0.0599, "step": 944 }, { "epoch": 0.6290564153769346, "grad_norm": 37.37943649291992, "learning_rate": 0.00015843694493783306, "loss": 0.2559, "step": 945 }, { "epoch": 0.6297220835413546, "grad_norm": 23.679391860961914, "learning_rate": 0.00015839253996447603, "loss": 0.1181, "step": 946 }, { "epoch": 0.6303877517057747, "grad_norm": 114.88152313232422, "learning_rate": 0.00015834813499111903, "loss": 0.7996, "step": 947 }, { "epoch": 0.6310534198701947, "grad_norm": 138.5634002685547, "learning_rate": 0.000158303730017762, "loss": 0.9386, "step": 948 }, { "epoch": 0.6317190880346147, "grad_norm": 65.20646667480469, "learning_rate": 0.00015825932504440497, "loss": 0.626, "step": 949 }, { "epoch": 0.6323847561990348, "grad_norm": 42.87594985961914, "learning_rate": 0.00015821492007104797, "loss": 0.4692, "step": 950 }, { "epoch": 0.6330504243634548, "grad_norm": 18.94576644897461, "learning_rate": 0.00015817051509769094, "loss": 0.1272, "step": 951 }, { "epoch": 0.6337160925278749, "grad_norm": 30.526195526123047, "learning_rate": 0.00015812611012433394, "loss": 0.3469, "step": 952 }, { "epoch": 0.6343817606922949, "grad_norm": 11.924509048461914, "learning_rate": 0.0001580817051509769, "loss": 0.2021, "step": 953 }, { "epoch": 0.635047428856715, "grad_norm": 64.6435775756836, "learning_rate": 0.0001580373001776199, "loss": 0.2674, "step": 954 }, { "epoch": 0.635713097021135, "grad_norm": 45.49868392944336, "learning_rate": 0.00015799289520426288, "loss": 0.2064, "step": 955 }, { "epoch": 0.636378765185555, "grad_norm": 65.23988342285156, "learning_rate": 0.00015794849023090588, "loss": 0.7653, "step": 956 }, { "epoch": 0.6370444333499751, "grad_norm": 27.883058547973633, "learning_rate": 0.00015790408525754885, "loss": 0.1265, "step": 957 }, { "epoch": 0.6377101015143951, "grad_norm": 35.20575714111328, "learning_rate": 0.00015785968028419185, "loss": 0.1687, "step": 958 }, { "epoch": 0.6383757696788152, "grad_norm": 51.54057693481445, "learning_rate": 0.00015781527531083482, "loss": 0.9334, "step": 959 }, { "epoch": 0.6390414378432352, "grad_norm": 46.25690841674805, "learning_rate": 0.0001577708703374778, "loss": 0.3797, "step": 960 }, { "epoch": 0.6397071060076552, "grad_norm": 67.028076171875, "learning_rate": 0.0001577264653641208, "loss": 0.3632, "step": 961 }, { "epoch": 0.6403727741720752, "grad_norm": 73.9373779296875, "learning_rate": 0.00015768206039076376, "loss": 0.631, "step": 962 }, { "epoch": 0.6410384423364952, "grad_norm": 83.83721160888672, "learning_rate": 0.00015763765541740676, "loss": 0.4768, "step": 963 }, { "epoch": 0.6417041105009152, "grad_norm": 29.81329345703125, "learning_rate": 0.00015759325044404973, "loss": 0.1028, "step": 964 }, { "epoch": 0.6423697786653353, "grad_norm": 82.8477554321289, "learning_rate": 0.00015754884547069273, "loss": 1.1949, "step": 965 }, { "epoch": 0.6430354468297553, "grad_norm": 55.87648010253906, "learning_rate": 0.0001575044404973357, "loss": 1.1176, "step": 966 }, { "epoch": 0.6437011149941754, "grad_norm": 55.962013244628906, "learning_rate": 0.0001574600355239787, "loss": 0.3424, "step": 967 }, { "epoch": 0.6443667831585954, "grad_norm": 34.9271125793457, "learning_rate": 0.00015741563055062167, "loss": 0.3968, "step": 968 }, { "epoch": 0.6450324513230155, "grad_norm": 63.50864028930664, "learning_rate": 0.00015737122557726467, "loss": 0.7757, "step": 969 }, { "epoch": 0.6456981194874355, "grad_norm": 13.376656532287598, "learning_rate": 0.00015732682060390764, "loss": 0.0667, "step": 970 }, { "epoch": 0.6463637876518555, "grad_norm": 50.788692474365234, "learning_rate": 0.00015728241563055064, "loss": 0.4594, "step": 971 }, { "epoch": 0.6470294558162756, "grad_norm": 83.55867004394531, "learning_rate": 0.0001572380106571936, "loss": 0.59, "step": 972 }, { "epoch": 0.6476951239806956, "grad_norm": 44.18373107910156, "learning_rate": 0.00015719360568383658, "loss": 0.2554, "step": 973 }, { "epoch": 0.6483607921451157, "grad_norm": 49.03962326049805, "learning_rate": 0.00015714920071047958, "loss": 0.2313, "step": 974 }, { "epoch": 0.6490264603095357, "grad_norm": 130.43885803222656, "learning_rate": 0.00015710479573712255, "loss": 0.5978, "step": 975 }, { "epoch": 0.6496921284739557, "grad_norm": 38.39082336425781, "learning_rate": 0.00015706039076376555, "loss": 0.1913, "step": 976 }, { "epoch": 0.6503577966383758, "grad_norm": 64.96234893798828, "learning_rate": 0.00015701598579040852, "loss": 0.2994, "step": 977 }, { "epoch": 0.6510234648027958, "grad_norm": 33.34348678588867, "learning_rate": 0.00015697158081705152, "loss": 0.2654, "step": 978 }, { "epoch": 0.6516891329672159, "grad_norm": 74.68543243408203, "learning_rate": 0.0001569271758436945, "loss": 0.481, "step": 979 }, { "epoch": 0.6523548011316359, "grad_norm": 70.37760162353516, "learning_rate": 0.00015688277087033749, "loss": 0.6861, "step": 980 }, { "epoch": 0.653020469296056, "grad_norm": 50.27323913574219, "learning_rate": 0.00015683836589698046, "loss": 0.5713, "step": 981 }, { "epoch": 0.653686137460476, "grad_norm": 28.98605728149414, "learning_rate": 0.00015679396092362346, "loss": 0.4162, "step": 982 }, { "epoch": 0.654351805624896, "grad_norm": 93.48116302490234, "learning_rate": 0.00015674955595026643, "loss": 0.6315, "step": 983 }, { "epoch": 0.6550174737893161, "grad_norm": 79.4161148071289, "learning_rate": 0.00015670515097690943, "loss": 0.3387, "step": 984 }, { "epoch": 0.6556831419537361, "grad_norm": 39.957183837890625, "learning_rate": 0.0001566607460035524, "loss": 0.2098, "step": 985 }, { "epoch": 0.656348810118156, "grad_norm": 52.71916961669922, "learning_rate": 0.00015661634103019537, "loss": 0.4032, "step": 986 }, { "epoch": 0.6570144782825761, "grad_norm": 74.70369720458984, "learning_rate": 0.00015657193605683837, "loss": 0.9639, "step": 987 }, { "epoch": 0.6576801464469961, "grad_norm": 78.83412170410156, "learning_rate": 0.00015652753108348134, "loss": 0.4124, "step": 988 }, { "epoch": 0.6583458146114162, "grad_norm": 56.441986083984375, "learning_rate": 0.00015648312611012434, "loss": 0.2709, "step": 989 }, { "epoch": 0.6590114827758362, "grad_norm": 50.2053337097168, "learning_rate": 0.0001564387211367673, "loss": 0.5839, "step": 990 }, { "epoch": 0.6596771509402563, "grad_norm": 46.05778884887695, "learning_rate": 0.0001563943161634103, "loss": 0.1879, "step": 991 }, { "epoch": 0.6603428191046763, "grad_norm": 89.58129119873047, "learning_rate": 0.00015634991119005328, "loss": 0.4644, "step": 992 }, { "epoch": 0.6610084872690963, "grad_norm": 72.7092514038086, "learning_rate": 0.00015630550621669628, "loss": 0.747, "step": 993 }, { "epoch": 0.6616741554335164, "grad_norm": 26.638051986694336, "learning_rate": 0.00015626110124333925, "loss": 0.3813, "step": 994 }, { "epoch": 0.6623398235979364, "grad_norm": 52.18305587768555, "learning_rate": 0.00015621669626998225, "loss": 0.2625, "step": 995 }, { "epoch": 0.6630054917623565, "grad_norm": 79.29065704345703, "learning_rate": 0.00015617229129662522, "loss": 0.4931, "step": 996 }, { "epoch": 0.6636711599267765, "grad_norm": 51.57868576049805, "learning_rate": 0.00015612788632326821, "loss": 0.3116, "step": 997 }, { "epoch": 0.6643368280911965, "grad_norm": 36.20134353637695, "learning_rate": 0.00015608348134991119, "loss": 0.4365, "step": 998 }, { "epoch": 0.6650024962556166, "grad_norm": 48.62169647216797, "learning_rate": 0.00015603907637655418, "loss": 0.3637, "step": 999 }, { "epoch": 0.6656681644200366, "grad_norm": 91.92649841308594, "learning_rate": 0.00015599467140319716, "loss": 0.4506, "step": 1000 }, { "epoch": 0.6663338325844567, "grad_norm": 72.67039489746094, "learning_rate": 0.00015595026642984015, "loss": 0.434, "step": 1001 }, { "epoch": 0.6669995007488767, "grad_norm": 43.33938980102539, "learning_rate": 0.00015590586145648313, "loss": 0.4614, "step": 1002 }, { "epoch": 0.6676651689132967, "grad_norm": 28.011381149291992, "learning_rate": 0.00015586145648312612, "loss": 0.1414, "step": 1003 }, { "epoch": 0.6683308370777168, "grad_norm": 71.00584411621094, "learning_rate": 0.0001558170515097691, "loss": 0.4117, "step": 1004 }, { "epoch": 0.6689965052421368, "grad_norm": 45.19962692260742, "learning_rate": 0.0001557726465364121, "loss": 0.3588, "step": 1005 }, { "epoch": 0.6696621734065569, "grad_norm": 30.098079681396484, "learning_rate": 0.00015572824156305506, "loss": 0.2342, "step": 1006 }, { "epoch": 0.6703278415709769, "grad_norm": 40.0351676940918, "learning_rate": 0.00015568383658969806, "loss": 0.1837, "step": 1007 }, { "epoch": 0.670993509735397, "grad_norm": 58.18196105957031, "learning_rate": 0.00015563943161634103, "loss": 0.2554, "step": 1008 }, { "epoch": 0.671659177899817, "grad_norm": 132.00172424316406, "learning_rate": 0.00015559502664298403, "loss": 0.6316, "step": 1009 }, { "epoch": 0.6723248460642369, "grad_norm": 22.881837844848633, "learning_rate": 0.000155550621669627, "loss": 0.1079, "step": 1010 }, { "epoch": 0.672990514228657, "grad_norm": 39.57730484008789, "learning_rate": 0.00015550621669627, "loss": 0.2325, "step": 1011 }, { "epoch": 0.673656182393077, "grad_norm": 28.402847290039062, "learning_rate": 0.00015546181172291297, "loss": 0.1991, "step": 1012 }, { "epoch": 0.674321850557497, "grad_norm": 43.875694274902344, "learning_rate": 0.00015541740674955597, "loss": 0.2521, "step": 1013 }, { "epoch": 0.6749875187219171, "grad_norm": 39.5247688293457, "learning_rate": 0.00015537300177619894, "loss": 0.0853, "step": 1014 }, { "epoch": 0.6756531868863371, "grad_norm": 67.50533294677734, "learning_rate": 0.00015532859680284194, "loss": 0.3061, "step": 1015 }, { "epoch": 0.6763188550507572, "grad_norm": 87.47017669677734, "learning_rate": 0.0001552841918294849, "loss": 0.8137, "step": 1016 }, { "epoch": 0.6769845232151772, "grad_norm": 63.81171417236328, "learning_rate": 0.0001552397868561279, "loss": 0.2849, "step": 1017 }, { "epoch": 0.6776501913795973, "grad_norm": 60.34880447387695, "learning_rate": 0.00015519538188277088, "loss": 0.2647, "step": 1018 }, { "epoch": 0.6783158595440173, "grad_norm": 69.48861694335938, "learning_rate": 0.00015515097690941388, "loss": 0.4646, "step": 1019 }, { "epoch": 0.6789815277084373, "grad_norm": 32.498104095458984, "learning_rate": 0.00015510657193605685, "loss": 0.214, "step": 1020 }, { "epoch": 0.6796471958728574, "grad_norm": 41.618438720703125, "learning_rate": 0.00015506216696269985, "loss": 0.1724, "step": 1021 }, { "epoch": 0.6803128640372774, "grad_norm": 79.42880249023438, "learning_rate": 0.00015501776198934282, "loss": 0.5835, "step": 1022 }, { "epoch": 0.6809785322016975, "grad_norm": 77.26080322265625, "learning_rate": 0.0001549733570159858, "loss": 0.2289, "step": 1023 }, { "epoch": 0.6816442003661175, "grad_norm": 66.00711822509766, "learning_rate": 0.0001549289520426288, "loss": 0.2978, "step": 1024 }, { "epoch": 0.6823098685305375, "grad_norm": 78.91641235351562, "learning_rate": 0.00015488454706927176, "loss": 1.254, "step": 1025 }, { "epoch": 0.6829755366949576, "grad_norm": 87.50247955322266, "learning_rate": 0.00015484014209591476, "loss": 0.3481, "step": 1026 }, { "epoch": 0.6836412048593776, "grad_norm": 57.35453414916992, "learning_rate": 0.00015479573712255773, "loss": 0.7368, "step": 1027 }, { "epoch": 0.6843068730237977, "grad_norm": 25.701223373413086, "learning_rate": 0.00015475133214920073, "loss": 0.4667, "step": 1028 }, { "epoch": 0.6849725411882177, "grad_norm": 32.501243591308594, "learning_rate": 0.0001547069271758437, "loss": 0.4981, "step": 1029 }, { "epoch": 0.6856382093526378, "grad_norm": 20.496658325195312, "learning_rate": 0.0001546625222024867, "loss": 0.2085, "step": 1030 }, { "epoch": 0.6863038775170578, "grad_norm": 35.525142669677734, "learning_rate": 0.00015461811722912967, "loss": 0.468, "step": 1031 }, { "epoch": 0.6869695456814778, "grad_norm": 48.355430603027344, "learning_rate": 0.00015457371225577267, "loss": 0.544, "step": 1032 }, { "epoch": 0.6876352138458979, "grad_norm": 37.98526382446289, "learning_rate": 0.00015452930728241564, "loss": 0.1542, "step": 1033 }, { "epoch": 0.6883008820103178, "grad_norm": 23.485910415649414, "learning_rate": 0.00015448490230905864, "loss": 0.4195, "step": 1034 }, { "epoch": 0.6889665501747378, "grad_norm": 30.047311782836914, "learning_rate": 0.0001544404973357016, "loss": 0.556, "step": 1035 }, { "epoch": 0.6896322183391579, "grad_norm": 35.63859176635742, "learning_rate": 0.00015439609236234458, "loss": 0.1312, "step": 1036 }, { "epoch": 0.6902978865035779, "grad_norm": 60.22919464111328, "learning_rate": 0.00015435168738898758, "loss": 0.3867, "step": 1037 }, { "epoch": 0.690963554667998, "grad_norm": 45.02543258666992, "learning_rate": 0.00015430728241563055, "loss": 0.3657, "step": 1038 }, { "epoch": 0.691629222832418, "grad_norm": 66.2843017578125, "learning_rate": 0.00015426287744227355, "loss": 0.485, "step": 1039 }, { "epoch": 0.692294890996838, "grad_norm": 44.457950592041016, "learning_rate": 0.00015421847246891652, "loss": 0.3118, "step": 1040 }, { "epoch": 0.6929605591612581, "grad_norm": 43.35429763793945, "learning_rate": 0.00015417406749555952, "loss": 0.4477, "step": 1041 }, { "epoch": 0.6936262273256781, "grad_norm": 88.85895538330078, "learning_rate": 0.0001541296625222025, "loss": 0.4381, "step": 1042 }, { "epoch": 0.6942918954900982, "grad_norm": 81.06390380859375, "learning_rate": 0.0001540852575488455, "loss": 0.2828, "step": 1043 }, { "epoch": 0.6949575636545182, "grad_norm": 73.58795928955078, "learning_rate": 0.00015404085257548846, "loss": 0.4024, "step": 1044 }, { "epoch": 0.6956232318189383, "grad_norm": 57.68181228637695, "learning_rate": 0.00015399644760213146, "loss": 0.2911, "step": 1045 }, { "epoch": 0.6962888999833583, "grad_norm": 73.5315170288086, "learning_rate": 0.00015395204262877443, "loss": 0.291, "step": 1046 }, { "epoch": 0.6969545681477783, "grad_norm": 15.196721076965332, "learning_rate": 0.00015390763765541743, "loss": 0.212, "step": 1047 }, { "epoch": 0.6976202363121984, "grad_norm": 106.10466766357422, "learning_rate": 0.0001538632326820604, "loss": 0.6399, "step": 1048 }, { "epoch": 0.6982859044766184, "grad_norm": 32.85442352294922, "learning_rate": 0.00015381882770870337, "loss": 0.3886, "step": 1049 }, { "epoch": 0.6989515726410385, "grad_norm": 81.06078338623047, "learning_rate": 0.00015377442273534637, "loss": 0.395, "step": 1050 }, { "epoch": 0.6996172408054585, "grad_norm": 45.303585052490234, "learning_rate": 0.00015373001776198934, "loss": 0.6317, "step": 1051 }, { "epoch": 0.7002829089698785, "grad_norm": 56.93617248535156, "learning_rate": 0.00015368561278863234, "loss": 0.262, "step": 1052 }, { "epoch": 0.7009485771342986, "grad_norm": 41.121883392333984, "learning_rate": 0.0001536412078152753, "loss": 0.5429, "step": 1053 }, { "epoch": 0.7016142452987186, "grad_norm": 36.327362060546875, "learning_rate": 0.0001535968028419183, "loss": 0.1532, "step": 1054 }, { "epoch": 0.7022799134631387, "grad_norm": 118.11614227294922, "learning_rate": 0.00015355239786856128, "loss": 1.233, "step": 1055 }, { "epoch": 0.7029455816275587, "grad_norm": 38.730621337890625, "learning_rate": 0.00015350799289520428, "loss": 0.1899, "step": 1056 }, { "epoch": 0.7036112497919786, "grad_norm": 45.68764114379883, "learning_rate": 0.00015346358792184725, "loss": 0.1665, "step": 1057 }, { "epoch": 0.7042769179563987, "grad_norm": 62.38422775268555, "learning_rate": 0.00015341918294849025, "loss": 0.3899, "step": 1058 }, { "epoch": 0.7049425861208187, "grad_norm": 54.563358306884766, "learning_rate": 0.00015337477797513322, "loss": 0.4982, "step": 1059 }, { "epoch": 0.7056082542852388, "grad_norm": 71.28933715820312, "learning_rate": 0.0001533303730017762, "loss": 0.7313, "step": 1060 }, { "epoch": 0.7062739224496588, "grad_norm": 23.5653133392334, "learning_rate": 0.0001532859680284192, "loss": 0.4373, "step": 1061 }, { "epoch": 0.7069395906140788, "grad_norm": 71.70646667480469, "learning_rate": 0.00015324156305506216, "loss": 0.414, "step": 1062 }, { "epoch": 0.7076052587784989, "grad_norm": 16.511093139648438, "learning_rate": 0.00015319715808170516, "loss": 0.1164, "step": 1063 }, { "epoch": 0.7082709269429189, "grad_norm": 33.91377639770508, "learning_rate": 0.00015315275310834813, "loss": 0.2195, "step": 1064 }, { "epoch": 0.708936595107339, "grad_norm": 48.85737991333008, "learning_rate": 0.00015310834813499113, "loss": 0.5585, "step": 1065 }, { "epoch": 0.709602263271759, "grad_norm": 47.1104850769043, "learning_rate": 0.0001530639431616341, "loss": 0.4511, "step": 1066 }, { "epoch": 0.7102679314361791, "grad_norm": 19.834882736206055, "learning_rate": 0.0001530195381882771, "loss": 0.2713, "step": 1067 }, { "epoch": 0.7109335996005991, "grad_norm": 48.64509963989258, "learning_rate": 0.00015297513321492007, "loss": 0.2581, "step": 1068 }, { "epoch": 0.7115992677650191, "grad_norm": 55.4994010925293, "learning_rate": 0.00015293072824156307, "loss": 0.287, "step": 1069 }, { "epoch": 0.7122649359294392, "grad_norm": 49.485355377197266, "learning_rate": 0.00015288632326820604, "loss": 0.3441, "step": 1070 }, { "epoch": 0.7129306040938592, "grad_norm": 47.73970413208008, "learning_rate": 0.00015284191829484904, "loss": 0.2326, "step": 1071 }, { "epoch": 0.7135962722582793, "grad_norm": 165.4192657470703, "learning_rate": 0.000152797513321492, "loss": 1.2375, "step": 1072 }, { "epoch": 0.7142619404226993, "grad_norm": 30.818099975585938, "learning_rate": 0.00015275310834813498, "loss": 0.4808, "step": 1073 }, { "epoch": 0.7149276085871193, "grad_norm": 25.218486785888672, "learning_rate": 0.00015270870337477798, "loss": 0.2016, "step": 1074 }, { "epoch": 0.7155932767515394, "grad_norm": 35.62108612060547, "learning_rate": 0.00015266429840142095, "loss": 0.4267, "step": 1075 }, { "epoch": 0.7162589449159594, "grad_norm": 39.81587600708008, "learning_rate": 0.00015261989342806395, "loss": 0.364, "step": 1076 }, { "epoch": 0.7169246130803795, "grad_norm": 44.70426940917969, "learning_rate": 0.00015257548845470692, "loss": 0.5001, "step": 1077 }, { "epoch": 0.7175902812447995, "grad_norm": 29.57493019104004, "learning_rate": 0.00015253108348134992, "loss": 0.0809, "step": 1078 }, { "epoch": 0.7182559494092196, "grad_norm": 21.22566795349121, "learning_rate": 0.0001524866785079929, "loss": 0.2915, "step": 1079 }, { "epoch": 0.7189216175736396, "grad_norm": 30.201745986938477, "learning_rate": 0.0001524422735346359, "loss": 0.2421, "step": 1080 }, { "epoch": 0.7195872857380595, "grad_norm": 31.989322662353516, "learning_rate": 0.00015239786856127886, "loss": 0.3325, "step": 1081 }, { "epoch": 0.7202529539024796, "grad_norm": 35.1367073059082, "learning_rate": 0.00015235346358792186, "loss": 0.2029, "step": 1082 }, { "epoch": 0.7209186220668996, "grad_norm": 55.973289489746094, "learning_rate": 0.00015230905861456483, "loss": 0.4624, "step": 1083 }, { "epoch": 0.7215842902313196, "grad_norm": 29.908843994140625, "learning_rate": 0.00015226465364120783, "loss": 0.2073, "step": 1084 }, { "epoch": 0.7222499583957397, "grad_norm": 115.7583999633789, "learning_rate": 0.0001522202486678508, "loss": 0.5003, "step": 1085 }, { "epoch": 0.7229156265601597, "grad_norm": 52.41007614135742, "learning_rate": 0.00015217584369449377, "loss": 0.2991, "step": 1086 }, { "epoch": 0.7235812947245798, "grad_norm": 33.20024871826172, "learning_rate": 0.00015213143872113677, "loss": 0.246, "step": 1087 }, { "epoch": 0.7242469628889998, "grad_norm": 66.91655731201172, "learning_rate": 0.00015208703374777974, "loss": 0.4534, "step": 1088 }, { "epoch": 0.7249126310534199, "grad_norm": 23.518030166625977, "learning_rate": 0.00015204262877442274, "loss": 0.0696, "step": 1089 }, { "epoch": 0.7255782992178399, "grad_norm": 92.09255981445312, "learning_rate": 0.0001519982238010657, "loss": 0.4496, "step": 1090 }, { "epoch": 0.7262439673822599, "grad_norm": 43.675048828125, "learning_rate": 0.0001519538188277087, "loss": 0.172, "step": 1091 }, { "epoch": 0.72690963554668, "grad_norm": 34.142303466796875, "learning_rate": 0.00015190941385435168, "loss": 0.5257, "step": 1092 }, { "epoch": 0.7275753037111, "grad_norm": 8.39620590209961, "learning_rate": 0.00015186500888099468, "loss": 0.0263, "step": 1093 }, { "epoch": 0.7282409718755201, "grad_norm": 91.039794921875, "learning_rate": 0.00015182060390763765, "loss": 1.1117, "step": 1094 }, { "epoch": 0.7289066400399401, "grad_norm": 68.07660675048828, "learning_rate": 0.00015177619893428065, "loss": 1.112, "step": 1095 }, { "epoch": 0.7295723082043601, "grad_norm": 22.02355194091797, "learning_rate": 0.00015173179396092362, "loss": 0.1046, "step": 1096 }, { "epoch": 0.7302379763687802, "grad_norm": 79.7904052734375, "learning_rate": 0.00015168738898756662, "loss": 0.5833, "step": 1097 }, { "epoch": 0.7309036445332002, "grad_norm": 28.20608901977539, "learning_rate": 0.0001516429840142096, "loss": 0.1487, "step": 1098 }, { "epoch": 0.7315693126976203, "grad_norm": 44.9555549621582, "learning_rate": 0.00015159857904085256, "loss": 0.1707, "step": 1099 }, { "epoch": 0.7322349808620403, "grad_norm": 27.90171241760254, "learning_rate": 0.00015155417406749556, "loss": 0.1437, "step": 1100 }, { "epoch": 0.7329006490264603, "grad_norm": 28.228458404541016, "learning_rate": 0.00015150976909413853, "loss": 0.3047, "step": 1101 }, { "epoch": 0.7335663171908804, "grad_norm": 39.96080017089844, "learning_rate": 0.00015146536412078153, "loss": 0.5157, "step": 1102 }, { "epoch": 0.7342319853553004, "grad_norm": 46.6057014465332, "learning_rate": 0.0001514209591474245, "loss": 0.5086, "step": 1103 }, { "epoch": 0.7348976535197205, "grad_norm": 101.82032012939453, "learning_rate": 0.0001513765541740675, "loss": 1.0609, "step": 1104 }, { "epoch": 0.7355633216841404, "grad_norm": 45.16305923461914, "learning_rate": 0.00015133214920071047, "loss": 0.4876, "step": 1105 }, { "epoch": 0.7362289898485604, "grad_norm": 28.174299240112305, "learning_rate": 0.00015128774422735347, "loss": 0.1836, "step": 1106 }, { "epoch": 0.7368946580129805, "grad_norm": 24.220523834228516, "learning_rate": 0.00015124333925399647, "loss": 0.1297, "step": 1107 }, { "epoch": 0.7375603261774005, "grad_norm": 39.064697265625, "learning_rate": 0.00015119893428063944, "loss": 0.1811, "step": 1108 }, { "epoch": 0.7382259943418206, "grad_norm": 43.84132385253906, "learning_rate": 0.00015115452930728244, "loss": 0.5146, "step": 1109 }, { "epoch": 0.7388916625062406, "grad_norm": 72.97101593017578, "learning_rate": 0.0001511101243339254, "loss": 0.6201, "step": 1110 }, { "epoch": 0.7395573306706607, "grad_norm": 36.33200454711914, "learning_rate": 0.0001510657193605684, "loss": 0.1652, "step": 1111 }, { "epoch": 0.7402229988350807, "grad_norm": 45.93895721435547, "learning_rate": 0.00015102131438721138, "loss": 0.2961, "step": 1112 }, { "epoch": 0.7408886669995007, "grad_norm": 47.38772201538086, "learning_rate": 0.00015097690941385437, "loss": 0.1987, "step": 1113 }, { "epoch": 0.7415543351639208, "grad_norm": 43.999027252197266, "learning_rate": 0.00015093250444049735, "loss": 0.189, "step": 1114 }, { "epoch": 0.7422200033283408, "grad_norm": 36.21022415161133, "learning_rate": 0.00015088809946714034, "loss": 0.3939, "step": 1115 }, { "epoch": 0.7428856714927609, "grad_norm": 18.37775421142578, "learning_rate": 0.00015084369449378332, "loss": 0.1409, "step": 1116 }, { "epoch": 0.7435513396571809, "grad_norm": 26.144956588745117, "learning_rate": 0.00015079928952042631, "loss": 0.1063, "step": 1117 }, { "epoch": 0.7442170078216009, "grad_norm": 56.28199005126953, "learning_rate": 0.00015075488454706929, "loss": 0.2831, "step": 1118 }, { "epoch": 0.744882675986021, "grad_norm": 6.287634372711182, "learning_rate": 0.00015071047957371228, "loss": 0.0248, "step": 1119 }, { "epoch": 0.745548344150441, "grad_norm": 70.22602844238281, "learning_rate": 0.00015066607460035525, "loss": 0.5079, "step": 1120 }, { "epoch": 0.7462140123148611, "grad_norm": 44.20976257324219, "learning_rate": 0.00015062166962699825, "loss": 0.2915, "step": 1121 }, { "epoch": 0.7468796804792811, "grad_norm": 127.5069808959961, "learning_rate": 0.00015057726465364122, "loss": 0.688, "step": 1122 }, { "epoch": 0.7475453486437011, "grad_norm": 52.48374938964844, "learning_rate": 0.0001505328596802842, "loss": 0.3903, "step": 1123 }, { "epoch": 0.7482110168081212, "grad_norm": 23.693878173828125, "learning_rate": 0.0001504884547069272, "loss": 0.1137, "step": 1124 }, { "epoch": 0.7488766849725412, "grad_norm": 28.868141174316406, "learning_rate": 0.00015044404973357017, "loss": 0.3542, "step": 1125 }, { "epoch": 0.7495423531369613, "grad_norm": 31.06593132019043, "learning_rate": 0.00015039964476021316, "loss": 0.1631, "step": 1126 }, { "epoch": 0.7502080213013813, "grad_norm": 50.74720764160156, "learning_rate": 0.00015035523978685614, "loss": 0.4445, "step": 1127 }, { "epoch": 0.7508736894658014, "grad_norm": 89.12187194824219, "learning_rate": 0.00015031083481349913, "loss": 0.4337, "step": 1128 }, { "epoch": 0.7515393576302213, "grad_norm": 160.0100555419922, "learning_rate": 0.0001502664298401421, "loss": 0.9883, "step": 1129 }, { "epoch": 0.7522050257946413, "grad_norm": 90.32857513427734, "learning_rate": 0.0001502220248667851, "loss": 0.5721, "step": 1130 }, { "epoch": 0.7528706939590614, "grad_norm": 48.3609504699707, "learning_rate": 0.00015017761989342807, "loss": 0.2299, "step": 1131 }, { "epoch": 0.7535363621234814, "grad_norm": 159.57183837890625, "learning_rate": 0.00015013321492007107, "loss": 0.933, "step": 1132 }, { "epoch": 0.7542020302879014, "grad_norm": 76.97663879394531, "learning_rate": 0.00015008880994671404, "loss": 0.3011, "step": 1133 }, { "epoch": 0.7548676984523215, "grad_norm": 65.98152923583984, "learning_rate": 0.00015004440497335704, "loss": 0.5159, "step": 1134 }, { "epoch": 0.7555333666167415, "grad_norm": 55.93942642211914, "learning_rate": 0.00015000000000000001, "loss": 0.5924, "step": 1135 }, { "epoch": 0.7561990347811616, "grad_norm": 68.26671600341797, "learning_rate": 0.00014995559502664299, "loss": 0.1483, "step": 1136 }, { "epoch": 0.7568647029455816, "grad_norm": 43.174015045166016, "learning_rate": 0.00014991119005328598, "loss": 0.2261, "step": 1137 }, { "epoch": 0.7575303711100017, "grad_norm": 7.699881076812744, "learning_rate": 0.00014986678507992895, "loss": 0.0219, "step": 1138 }, { "epoch": 0.7581960392744217, "grad_norm": 31.757158279418945, "learning_rate": 0.00014982238010657195, "loss": 0.1486, "step": 1139 }, { "epoch": 0.7588617074388417, "grad_norm": 139.2758026123047, "learning_rate": 0.00014977797513321492, "loss": 0.9059, "step": 1140 }, { "epoch": 0.7595273756032618, "grad_norm": 114.37268829345703, "learning_rate": 0.00014973357015985792, "loss": 0.4527, "step": 1141 }, { "epoch": 0.7601930437676818, "grad_norm": 87.0638427734375, "learning_rate": 0.0001496891651865009, "loss": 0.528, "step": 1142 }, { "epoch": 0.7608587119321019, "grad_norm": 47.24208450317383, "learning_rate": 0.0001496447602131439, "loss": 0.4367, "step": 1143 }, { "epoch": 0.7615243800965219, "grad_norm": 20.45673370361328, "learning_rate": 0.00014960035523978686, "loss": 0.1657, "step": 1144 }, { "epoch": 0.7621900482609419, "grad_norm": 68.62003326416016, "learning_rate": 0.00014955595026642986, "loss": 0.3808, "step": 1145 }, { "epoch": 0.762855716425362, "grad_norm": 91.08775329589844, "learning_rate": 0.00014951154529307283, "loss": 0.4248, "step": 1146 }, { "epoch": 0.763521384589782, "grad_norm": 46.624488830566406, "learning_rate": 0.00014946714031971583, "loss": 0.1758, "step": 1147 }, { "epoch": 0.7641870527542021, "grad_norm": 65.56462097167969, "learning_rate": 0.0001494227353463588, "loss": 1.0118, "step": 1148 }, { "epoch": 0.7648527209186221, "grad_norm": 45.4866828918457, "learning_rate": 0.00014937833037300177, "loss": 0.4172, "step": 1149 }, { "epoch": 0.7655183890830422, "grad_norm": 34.517486572265625, "learning_rate": 0.00014933392539964477, "loss": 0.1651, "step": 1150 }, { "epoch": 0.7661840572474622, "grad_norm": 17.68457794189453, "learning_rate": 0.00014928952042628774, "loss": 0.1094, "step": 1151 }, { "epoch": 0.7668497254118821, "grad_norm": 75.56631469726562, "learning_rate": 0.00014924511545293074, "loss": 0.3299, "step": 1152 }, { "epoch": 0.7675153935763022, "grad_norm": 104.16996002197266, "learning_rate": 0.00014920071047957371, "loss": 0.8619, "step": 1153 }, { "epoch": 0.7681810617407222, "grad_norm": 32.42017364501953, "learning_rate": 0.0001491563055062167, "loss": 0.1048, "step": 1154 }, { "epoch": 0.7688467299051422, "grad_norm": 63.57094955444336, "learning_rate": 0.00014911190053285968, "loss": 0.2196, "step": 1155 }, { "epoch": 0.7695123980695623, "grad_norm": 35.73279571533203, "learning_rate": 0.00014906749555950268, "loss": 0.3458, "step": 1156 }, { "epoch": 0.7701780662339823, "grad_norm": 40.91196060180664, "learning_rate": 0.00014902309058614565, "loss": 0.2994, "step": 1157 }, { "epoch": 0.7708437343984024, "grad_norm": 16.660181045532227, "learning_rate": 0.00014897868561278865, "loss": 0.114, "step": 1158 }, { "epoch": 0.7715094025628224, "grad_norm": 74.0390853881836, "learning_rate": 0.00014893428063943162, "loss": 0.9675, "step": 1159 }, { "epoch": 0.7721750707272425, "grad_norm": 45.767269134521484, "learning_rate": 0.00014888987566607462, "loss": 0.1265, "step": 1160 }, { "epoch": 0.7728407388916625, "grad_norm": 56.93156051635742, "learning_rate": 0.0001488454706927176, "loss": 0.6565, "step": 1161 }, { "epoch": 0.7735064070560825, "grad_norm": 73.47280883789062, "learning_rate": 0.00014880106571936056, "loss": 0.3325, "step": 1162 }, { "epoch": 0.7741720752205026, "grad_norm": 20.19727897644043, "learning_rate": 0.00014875666074600356, "loss": 0.1534, "step": 1163 }, { "epoch": 0.7748377433849226, "grad_norm": 84.14955139160156, "learning_rate": 0.00014871225577264653, "loss": 0.4423, "step": 1164 }, { "epoch": 0.7755034115493427, "grad_norm": 72.55103302001953, "learning_rate": 0.00014866785079928953, "loss": 0.2602, "step": 1165 }, { "epoch": 0.7761690797137627, "grad_norm": 48.70878982543945, "learning_rate": 0.0001486234458259325, "loss": 0.5364, "step": 1166 }, { "epoch": 0.7768347478781827, "grad_norm": 42.47782897949219, "learning_rate": 0.0001485790408525755, "loss": 0.2075, "step": 1167 }, { "epoch": 0.7775004160426028, "grad_norm": 53.00294876098633, "learning_rate": 0.00014853463587921847, "loss": 0.5774, "step": 1168 }, { "epoch": 0.7781660842070228, "grad_norm": 24.320568084716797, "learning_rate": 0.00014849023090586147, "loss": 0.2326, "step": 1169 }, { "epoch": 0.7788317523714429, "grad_norm": 67.81494140625, "learning_rate": 0.00014844582593250444, "loss": 0.685, "step": 1170 }, { "epoch": 0.7794974205358629, "grad_norm": 33.585205078125, "learning_rate": 0.00014840142095914744, "loss": 0.0862, "step": 1171 }, { "epoch": 0.780163088700283, "grad_norm": 79.48095703125, "learning_rate": 0.0001483570159857904, "loss": 0.3263, "step": 1172 }, { "epoch": 0.780828756864703, "grad_norm": 75.42149353027344, "learning_rate": 0.00014831261101243338, "loss": 0.8082, "step": 1173 }, { "epoch": 0.781494425029123, "grad_norm": 140.92236328125, "learning_rate": 0.00014826820603907638, "loss": 1.1663, "step": 1174 }, { "epoch": 0.7821600931935431, "grad_norm": 32.08091735839844, "learning_rate": 0.00014822380106571935, "loss": 0.1213, "step": 1175 }, { "epoch": 0.782825761357963, "grad_norm": 80.28524017333984, "learning_rate": 0.00014817939609236235, "loss": 0.3669, "step": 1176 }, { "epoch": 0.783491429522383, "grad_norm": 52.19442367553711, "learning_rate": 0.00014813499111900532, "loss": 0.3446, "step": 1177 }, { "epoch": 0.7841570976868031, "grad_norm": 49.308589935302734, "learning_rate": 0.00014809058614564832, "loss": 0.6078, "step": 1178 }, { "epoch": 0.7848227658512231, "grad_norm": 25.709796905517578, "learning_rate": 0.0001480461811722913, "loss": 0.1212, "step": 1179 }, { "epoch": 0.7854884340156432, "grad_norm": 48.620845794677734, "learning_rate": 0.0001480017761989343, "loss": 0.1723, "step": 1180 }, { "epoch": 0.7861541021800632, "grad_norm": 83.81596374511719, "learning_rate": 0.00014795737122557726, "loss": 0.3797, "step": 1181 }, { "epoch": 0.7868197703444832, "grad_norm": 64.71217346191406, "learning_rate": 0.00014791296625222026, "loss": 0.2607, "step": 1182 }, { "epoch": 0.7874854385089033, "grad_norm": 190.05763244628906, "learning_rate": 0.00014786856127886323, "loss": 0.9319, "step": 1183 }, { "epoch": 0.7881511066733233, "grad_norm": 65.00997924804688, "learning_rate": 0.00014782415630550623, "loss": 0.3281, "step": 1184 }, { "epoch": 0.7888167748377434, "grad_norm": 70.21501922607422, "learning_rate": 0.0001477797513321492, "loss": 0.2666, "step": 1185 }, { "epoch": 0.7894824430021634, "grad_norm": 38.21614456176758, "learning_rate": 0.00014773534635879217, "loss": 0.2436, "step": 1186 }, { "epoch": 0.7901481111665835, "grad_norm": 26.931507110595703, "learning_rate": 0.00014769094138543517, "loss": 0.1595, "step": 1187 }, { "epoch": 0.7908137793310035, "grad_norm": 8.515877723693848, "learning_rate": 0.00014764653641207814, "loss": 0.0456, "step": 1188 }, { "epoch": 0.7914794474954235, "grad_norm": 26.995105743408203, "learning_rate": 0.00014760213143872114, "loss": 0.1411, "step": 1189 }, { "epoch": 0.7921451156598436, "grad_norm": 56.647010803222656, "learning_rate": 0.0001475577264653641, "loss": 0.2041, "step": 1190 }, { "epoch": 0.7928107838242636, "grad_norm": 160.20263671875, "learning_rate": 0.0001475133214920071, "loss": 1.7234, "step": 1191 }, { "epoch": 0.7934764519886837, "grad_norm": 17.144086837768555, "learning_rate": 0.00014746891651865008, "loss": 0.0622, "step": 1192 }, { "epoch": 0.7941421201531037, "grad_norm": 89.60062408447266, "learning_rate": 0.00014742451154529308, "loss": 0.7542, "step": 1193 }, { "epoch": 0.7948077883175237, "grad_norm": 36.0671501159668, "learning_rate": 0.00014738010657193605, "loss": 0.3883, "step": 1194 }, { "epoch": 0.7954734564819438, "grad_norm": 110.19222259521484, "learning_rate": 0.00014733570159857905, "loss": 0.5973, "step": 1195 }, { "epoch": 0.7961391246463638, "grad_norm": 47.82674026489258, "learning_rate": 0.00014729129662522202, "loss": 0.4346, "step": 1196 }, { "epoch": 0.7968047928107839, "grad_norm": 49.006752014160156, "learning_rate": 0.00014724689165186502, "loss": 0.3034, "step": 1197 }, { "epoch": 0.7974704609752039, "grad_norm": 19.300506591796875, "learning_rate": 0.000147202486678508, "loss": 0.0486, "step": 1198 }, { "epoch": 0.798136129139624, "grad_norm": 59.16493225097656, "learning_rate": 0.00014715808170515096, "loss": 0.5721, "step": 1199 }, { "epoch": 0.7988017973040439, "grad_norm": 104.55072784423828, "learning_rate": 0.00014711367673179396, "loss": 0.6708, "step": 1200 }, { "epoch": 0.7994674654684639, "grad_norm": 65.50591278076172, "learning_rate": 0.00014706927175843693, "loss": 0.3916, "step": 1201 }, { "epoch": 0.800133133632884, "grad_norm": 45.64435577392578, "learning_rate": 0.00014702486678507993, "loss": 0.2219, "step": 1202 }, { "epoch": 0.800798801797304, "grad_norm": 26.092966079711914, "learning_rate": 0.0001469804618117229, "loss": 0.102, "step": 1203 }, { "epoch": 0.801464469961724, "grad_norm": 6.989476203918457, "learning_rate": 0.0001469360568383659, "loss": 0.0194, "step": 1204 }, { "epoch": 0.8021301381261441, "grad_norm": 54.65571594238281, "learning_rate": 0.00014689165186500887, "loss": 0.5015, "step": 1205 }, { "epoch": 0.8027958062905641, "grad_norm": 63.82543182373047, "learning_rate": 0.00014684724689165187, "loss": 0.2404, "step": 1206 }, { "epoch": 0.8034614744549842, "grad_norm": 43.29622268676758, "learning_rate": 0.00014680284191829484, "loss": 0.2142, "step": 1207 }, { "epoch": 0.8041271426194042, "grad_norm": 145.57347106933594, "learning_rate": 0.00014675843694493784, "loss": 0.8315, "step": 1208 }, { "epoch": 0.8047928107838243, "grad_norm": 43.235164642333984, "learning_rate": 0.0001467140319715808, "loss": 0.2158, "step": 1209 }, { "epoch": 0.8054584789482443, "grad_norm": 71.96846008300781, "learning_rate": 0.0001466696269982238, "loss": 0.5527, "step": 1210 }, { "epoch": 0.8061241471126643, "grad_norm": 80.44132995605469, "learning_rate": 0.00014662522202486678, "loss": 0.5039, "step": 1211 }, { "epoch": 0.8067898152770844, "grad_norm": 37.07688903808594, "learning_rate": 0.00014658081705150978, "loss": 0.2079, "step": 1212 }, { "epoch": 0.8074554834415044, "grad_norm": 50.39672088623047, "learning_rate": 0.00014653641207815275, "loss": 0.4003, "step": 1213 }, { "epoch": 0.8081211516059245, "grad_norm": 14.553754806518555, "learning_rate": 0.00014649200710479575, "loss": 0.0929, "step": 1214 }, { "epoch": 0.8087868197703445, "grad_norm": 29.12799644470215, "learning_rate": 0.00014644760213143872, "loss": 0.217, "step": 1215 }, { "epoch": 0.8094524879347645, "grad_norm": 105.95761108398438, "learning_rate": 0.00014640319715808172, "loss": 0.4326, "step": 1216 }, { "epoch": 0.8101181560991846, "grad_norm": 43.320587158203125, "learning_rate": 0.0001463587921847247, "loss": 0.1519, "step": 1217 }, { "epoch": 0.8107838242636046, "grad_norm": 75.10538482666016, "learning_rate": 0.0001463143872113677, "loss": 0.6495, "step": 1218 }, { "epoch": 0.8114494924280247, "grad_norm": 71.7596435546875, "learning_rate": 0.00014626998223801066, "loss": 0.9403, "step": 1219 }, { "epoch": 0.8121151605924447, "grad_norm": 142.55914306640625, "learning_rate": 0.00014622557726465366, "loss": 1.0025, "step": 1220 }, { "epoch": 0.8127808287568647, "grad_norm": 78.93761444091797, "learning_rate": 0.00014618117229129663, "loss": 0.3843, "step": 1221 }, { "epoch": 0.8134464969212848, "grad_norm": 50.76917266845703, "learning_rate": 0.00014613676731793963, "loss": 0.4467, "step": 1222 }, { "epoch": 0.8141121650857048, "grad_norm": 29.166250228881836, "learning_rate": 0.0001460923623445826, "loss": 0.2112, "step": 1223 }, { "epoch": 0.8147778332501248, "grad_norm": 75.68633270263672, "learning_rate": 0.0001460479573712256, "loss": 0.3437, "step": 1224 }, { "epoch": 0.8154435014145448, "grad_norm": 74.50650787353516, "learning_rate": 0.00014600355239786857, "loss": 0.5997, "step": 1225 }, { "epoch": 0.8161091695789648, "grad_norm": 30.94719123840332, "learning_rate": 0.00014595914742451157, "loss": 0.3563, "step": 1226 }, { "epoch": 0.8167748377433849, "grad_norm": 102.63308715820312, "learning_rate": 0.00014591474245115454, "loss": 0.85, "step": 1227 }, { "epoch": 0.8174405059078049, "grad_norm": 22.703868865966797, "learning_rate": 0.00014587033747779754, "loss": 0.088, "step": 1228 }, { "epoch": 0.818106174072225, "grad_norm": 25.85906410217285, "learning_rate": 0.0001458259325044405, "loss": 0.2035, "step": 1229 }, { "epoch": 0.818771842236645, "grad_norm": 68.95484924316406, "learning_rate": 0.0001457815275310835, "loss": 0.5144, "step": 1230 }, { "epoch": 0.819437510401065, "grad_norm": 29.924816131591797, "learning_rate": 0.00014573712255772648, "loss": 0.1269, "step": 1231 }, { "epoch": 0.8201031785654851, "grad_norm": 73.62071228027344, "learning_rate": 0.00014569271758436948, "loss": 0.5401, "step": 1232 }, { "epoch": 0.8207688467299051, "grad_norm": 55.8017692565918, "learning_rate": 0.00014564831261101245, "loss": 0.9776, "step": 1233 }, { "epoch": 0.8214345148943252, "grad_norm": 25.978734970092773, "learning_rate": 0.00014560390763765545, "loss": 0.145, "step": 1234 }, { "epoch": 0.8221001830587452, "grad_norm": 28.74590492248535, "learning_rate": 0.00014555950266429842, "loss": 0.3171, "step": 1235 }, { "epoch": 0.8227658512231653, "grad_norm": 39.05659484863281, "learning_rate": 0.0001455150976909414, "loss": 0.238, "step": 1236 }, { "epoch": 0.8234315193875853, "grad_norm": 52.03479766845703, "learning_rate": 0.00014547069271758439, "loss": 0.2445, "step": 1237 }, { "epoch": 0.8240971875520053, "grad_norm": 28.90163803100586, "learning_rate": 0.00014542628774422736, "loss": 0.0914, "step": 1238 }, { "epoch": 0.8247628557164254, "grad_norm": 64.31964874267578, "learning_rate": 0.00014538188277087036, "loss": 0.3439, "step": 1239 }, { "epoch": 0.8254285238808454, "grad_norm": 33.80768585205078, "learning_rate": 0.00014533747779751333, "loss": 0.4763, "step": 1240 }, { "epoch": 0.8260941920452655, "grad_norm": 57.27531433105469, "learning_rate": 0.00014529307282415633, "loss": 0.3239, "step": 1241 }, { "epoch": 0.8267598602096855, "grad_norm": 72.25888061523438, "learning_rate": 0.0001452486678507993, "loss": 0.5758, "step": 1242 }, { "epoch": 0.8274255283741055, "grad_norm": 25.7634334564209, "learning_rate": 0.0001452042628774423, "loss": 0.1174, "step": 1243 }, { "epoch": 0.8280911965385256, "grad_norm": 118.91036987304688, "learning_rate": 0.00014515985790408527, "loss": 0.9157, "step": 1244 }, { "epoch": 0.8287568647029456, "grad_norm": 62.879493713378906, "learning_rate": 0.00014511545293072826, "loss": 0.4637, "step": 1245 }, { "epoch": 0.8294225328673657, "grad_norm": 10.97956371307373, "learning_rate": 0.00014507104795737124, "loss": 0.1148, "step": 1246 }, { "epoch": 0.8300882010317856, "grad_norm": 34.82111358642578, "learning_rate": 0.00014502664298401423, "loss": 0.188, "step": 1247 }, { "epoch": 0.8307538691962056, "grad_norm": 63.016754150390625, "learning_rate": 0.0001449822380106572, "loss": 0.5228, "step": 1248 }, { "epoch": 0.8314195373606257, "grad_norm": 61.121826171875, "learning_rate": 0.00014493783303730018, "loss": 0.2117, "step": 1249 }, { "epoch": 0.8320852055250457, "grad_norm": 76.04669952392578, "learning_rate": 0.00014489342806394318, "loss": 0.7473, "step": 1250 }, { "epoch": 0.8327508736894658, "grad_norm": 43.226646423339844, "learning_rate": 0.00014484902309058615, "loss": 0.3819, "step": 1251 }, { "epoch": 0.8334165418538858, "grad_norm": 35.6154899597168, "learning_rate": 0.00014480461811722914, "loss": 0.1301, "step": 1252 }, { "epoch": 0.8340822100183058, "grad_norm": 128.75949096679688, "learning_rate": 0.00014476021314387212, "loss": 0.9622, "step": 1253 }, { "epoch": 0.8347478781827259, "grad_norm": 43.62495803833008, "learning_rate": 0.00014471580817051511, "loss": 0.2774, "step": 1254 }, { "epoch": 0.8354135463471459, "grad_norm": 61.1744499206543, "learning_rate": 0.00014467140319715809, "loss": 0.3987, "step": 1255 }, { "epoch": 0.836079214511566, "grad_norm": 67.01728057861328, "learning_rate": 0.00014462699822380108, "loss": 0.6077, "step": 1256 }, { "epoch": 0.836744882675986, "grad_norm": 76.08590698242188, "learning_rate": 0.00014458259325044406, "loss": 0.3622, "step": 1257 }, { "epoch": 0.837410550840406, "grad_norm": 65.32396697998047, "learning_rate": 0.00014453818827708705, "loss": 0.3597, "step": 1258 }, { "epoch": 0.8380762190048261, "grad_norm": 9.421642303466797, "learning_rate": 0.00014449378330373003, "loss": 0.0372, "step": 1259 }, { "epoch": 0.8387418871692461, "grad_norm": 45.54380798339844, "learning_rate": 0.00014444937833037302, "loss": 0.3783, "step": 1260 }, { "epoch": 0.8394075553336662, "grad_norm": 17.207380294799805, "learning_rate": 0.000144404973357016, "loss": 0.1013, "step": 1261 }, { "epoch": 0.8400732234980862, "grad_norm": 23.20643424987793, "learning_rate": 0.00014436056838365897, "loss": 0.3503, "step": 1262 }, { "epoch": 0.8407388916625063, "grad_norm": 70.00853729248047, "learning_rate": 0.00014431616341030196, "loss": 0.5384, "step": 1263 }, { "epoch": 0.8414045598269263, "grad_norm": 19.45144271850586, "learning_rate": 0.00014427175843694494, "loss": 0.0845, "step": 1264 }, { "epoch": 0.8420702279913463, "grad_norm": 111.85758209228516, "learning_rate": 0.00014422735346358793, "loss": 0.968, "step": 1265 }, { "epoch": 0.8427358961557664, "grad_norm": 31.912893295288086, "learning_rate": 0.0001441829484902309, "loss": 0.094, "step": 1266 }, { "epoch": 0.8434015643201864, "grad_norm": 110.80167388916016, "learning_rate": 0.0001441385435168739, "loss": 0.4886, "step": 1267 }, { "epoch": 0.8440672324846065, "grad_norm": 91.96247100830078, "learning_rate": 0.00014409413854351688, "loss": 0.954, "step": 1268 }, { "epoch": 0.8447329006490265, "grad_norm": 58.025081634521484, "learning_rate": 0.00014404973357015987, "loss": 0.4556, "step": 1269 }, { "epoch": 0.8453985688134465, "grad_norm": 49.4656867980957, "learning_rate": 0.00014400532859680284, "loss": 0.1042, "step": 1270 }, { "epoch": 0.8460642369778665, "grad_norm": 44.73607635498047, "learning_rate": 0.00014396092362344584, "loss": 0.1664, "step": 1271 }, { "epoch": 0.8467299051422865, "grad_norm": 38.9708251953125, "learning_rate": 0.00014391651865008881, "loss": 0.4827, "step": 1272 }, { "epoch": 0.8473955733067066, "grad_norm": 8.022209167480469, "learning_rate": 0.00014387211367673179, "loss": 0.0511, "step": 1273 }, { "epoch": 0.8480612414711266, "grad_norm": 65.75480651855469, "learning_rate": 0.00014382770870337478, "loss": 0.4004, "step": 1274 }, { "epoch": 0.8487269096355466, "grad_norm": 171.89952087402344, "learning_rate": 0.00014378330373001776, "loss": 0.8089, "step": 1275 }, { "epoch": 0.8493925777999667, "grad_norm": 49.746299743652344, "learning_rate": 0.00014373889875666075, "loss": 0.3264, "step": 1276 }, { "epoch": 0.8500582459643867, "grad_norm": 152.19830322265625, "learning_rate": 0.00014369449378330373, "loss": 0.906, "step": 1277 }, { "epoch": 0.8507239141288068, "grad_norm": 84.58346557617188, "learning_rate": 0.00014365008880994672, "loss": 0.3384, "step": 1278 }, { "epoch": 0.8513895822932268, "grad_norm": 65.4444808959961, "learning_rate": 0.0001436056838365897, "loss": 0.4452, "step": 1279 }, { "epoch": 0.8520552504576469, "grad_norm": 59.972782135009766, "learning_rate": 0.0001435612788632327, "loss": 0.5103, "step": 1280 }, { "epoch": 0.8527209186220669, "grad_norm": 27.884855270385742, "learning_rate": 0.00014351687388987566, "loss": 0.1045, "step": 1281 }, { "epoch": 0.8533865867864869, "grad_norm": 59.312259674072266, "learning_rate": 0.00014347246891651866, "loss": 0.4007, "step": 1282 }, { "epoch": 0.854052254950907, "grad_norm": 38.883304595947266, "learning_rate": 0.00014342806394316163, "loss": 0.4014, "step": 1283 }, { "epoch": 0.854717923115327, "grad_norm": 6.253726959228516, "learning_rate": 0.00014338365896980463, "loss": 0.0208, "step": 1284 }, { "epoch": 0.8553835912797471, "grad_norm": 60.496219635009766, "learning_rate": 0.0001433392539964476, "loss": 0.6616, "step": 1285 }, { "epoch": 0.8560492594441671, "grad_norm": 142.0136260986328, "learning_rate": 0.00014329484902309058, "loss": 1.0412, "step": 1286 }, { "epoch": 0.8567149276085871, "grad_norm": 73.22315216064453, "learning_rate": 0.00014325044404973357, "loss": 0.4169, "step": 1287 }, { "epoch": 0.8573805957730072, "grad_norm": 48.91903305053711, "learning_rate": 0.00014320603907637654, "loss": 0.3147, "step": 1288 }, { "epoch": 0.8580462639374272, "grad_norm": 82.37753295898438, "learning_rate": 0.00014316163410301954, "loss": 0.6757, "step": 1289 }, { "epoch": 0.8587119321018473, "grad_norm": 68.59038543701172, "learning_rate": 0.00014311722912966251, "loss": 0.7777, "step": 1290 }, { "epoch": 0.8593776002662673, "grad_norm": 2.0153257846832275, "learning_rate": 0.0001430728241563055, "loss": 0.0086, "step": 1291 }, { "epoch": 0.8600432684306873, "grad_norm": 65.46209716796875, "learning_rate": 0.00014302841918294848, "loss": 0.2463, "step": 1292 }, { "epoch": 0.8607089365951074, "grad_norm": 42.20965576171875, "learning_rate": 0.00014298401420959148, "loss": 0.2652, "step": 1293 }, { "epoch": 0.8613746047595274, "grad_norm": 33.425987243652344, "learning_rate": 0.00014293960923623445, "loss": 0.2774, "step": 1294 }, { "epoch": 0.8620402729239474, "grad_norm": 154.2758331298828, "learning_rate": 0.00014289520426287745, "loss": 1.8189, "step": 1295 }, { "epoch": 0.8627059410883674, "grad_norm": 53.27513885498047, "learning_rate": 0.00014285079928952042, "loss": 0.2672, "step": 1296 }, { "epoch": 0.8633716092527874, "grad_norm": 66.36660766601562, "learning_rate": 0.00014280639431616342, "loss": 0.9157, "step": 1297 }, { "epoch": 0.8640372774172075, "grad_norm": 35.55696487426758, "learning_rate": 0.0001427619893428064, "loss": 0.1728, "step": 1298 }, { "epoch": 0.8647029455816275, "grad_norm": 37.19813919067383, "learning_rate": 0.00014271758436944936, "loss": 0.163, "step": 1299 }, { "epoch": 0.8653686137460476, "grad_norm": 43.158138275146484, "learning_rate": 0.00014267317939609236, "loss": 0.1693, "step": 1300 }, { "epoch": 0.8660342819104676, "grad_norm": 36.93934631347656, "learning_rate": 0.00014262877442273533, "loss": 0.2359, "step": 1301 }, { "epoch": 0.8666999500748876, "grad_norm": 31.97963523864746, "learning_rate": 0.00014258436944937833, "loss": 0.0936, "step": 1302 }, { "epoch": 0.8673656182393077, "grad_norm": 54.78131103515625, "learning_rate": 0.0001425399644760213, "loss": 0.436, "step": 1303 }, { "epoch": 0.8680312864037277, "grad_norm": 28.82598114013672, "learning_rate": 0.0001424955595026643, "loss": 0.1752, "step": 1304 }, { "epoch": 0.8686969545681478, "grad_norm": 70.60225677490234, "learning_rate": 0.00014245115452930727, "loss": 0.4448, "step": 1305 }, { "epoch": 0.8693626227325678, "grad_norm": 52.563079833984375, "learning_rate": 0.00014240674955595027, "loss": 0.3112, "step": 1306 }, { "epoch": 0.8700282908969879, "grad_norm": 42.23539352416992, "learning_rate": 0.00014236234458259324, "loss": 0.1804, "step": 1307 }, { "epoch": 0.8706939590614079, "grad_norm": 70.87139892578125, "learning_rate": 0.00014231793960923624, "loss": 1.0799, "step": 1308 }, { "epoch": 0.8713596272258279, "grad_norm": 36.24211883544922, "learning_rate": 0.0001422735346358792, "loss": 0.2556, "step": 1309 }, { "epoch": 0.872025295390248, "grad_norm": 125.8697509765625, "learning_rate": 0.0001422291296625222, "loss": 0.8421, "step": 1310 }, { "epoch": 0.872690963554668, "grad_norm": 36.61944580078125, "learning_rate": 0.00014218472468916518, "loss": 0.3533, "step": 1311 }, { "epoch": 0.8733566317190881, "grad_norm": 66.49879455566406, "learning_rate": 0.00014214031971580818, "loss": 0.7084, "step": 1312 }, { "epoch": 0.8740222998835081, "grad_norm": 79.41429901123047, "learning_rate": 0.00014209591474245115, "loss": 0.7719, "step": 1313 }, { "epoch": 0.8746879680479281, "grad_norm": 42.24322509765625, "learning_rate": 0.00014205150976909415, "loss": 0.1846, "step": 1314 }, { "epoch": 0.8753536362123482, "grad_norm": 47.408390045166016, "learning_rate": 0.00014200710479573712, "loss": 0.408, "step": 1315 }, { "epoch": 0.8760193043767682, "grad_norm": 45.516883850097656, "learning_rate": 0.00014196269982238012, "loss": 0.3997, "step": 1316 }, { "epoch": 0.8766849725411883, "grad_norm": 34.819068908691406, "learning_rate": 0.0001419182948490231, "loss": 0.2617, "step": 1317 }, { "epoch": 0.8773506407056083, "grad_norm": 52.184749603271484, "learning_rate": 0.0001418738898756661, "loss": 0.2546, "step": 1318 }, { "epoch": 0.8780163088700282, "grad_norm": 58.92740249633789, "learning_rate": 0.00014182948490230906, "loss": 0.5765, "step": 1319 }, { "epoch": 0.8786819770344483, "grad_norm": 67.0827865600586, "learning_rate": 0.00014178507992895206, "loss": 0.3192, "step": 1320 }, { "epoch": 0.8793476451988683, "grad_norm": 45.35605239868164, "learning_rate": 0.00014174067495559503, "loss": 0.1883, "step": 1321 }, { "epoch": 0.8800133133632884, "grad_norm": 84.37458038330078, "learning_rate": 0.00014169626998223803, "loss": 0.5937, "step": 1322 }, { "epoch": 0.8806789815277084, "grad_norm": 38.89516830444336, "learning_rate": 0.000141651865008881, "loss": 0.2136, "step": 1323 }, { "epoch": 0.8813446496921284, "grad_norm": 76.28330993652344, "learning_rate": 0.000141607460035524, "loss": 0.697, "step": 1324 }, { "epoch": 0.8820103178565485, "grad_norm": 48.03358459472656, "learning_rate": 0.00014156305506216697, "loss": 0.2988, "step": 1325 }, { "epoch": 0.8826759860209685, "grad_norm": 36.21906661987305, "learning_rate": 0.00014151865008880997, "loss": 0.7056, "step": 1326 }, { "epoch": 0.8833416541853886, "grad_norm": 78.60859680175781, "learning_rate": 0.00014147424511545294, "loss": 0.3909, "step": 1327 }, { "epoch": 0.8840073223498086, "grad_norm": 20.138051986694336, "learning_rate": 0.00014142984014209594, "loss": 0.2614, "step": 1328 }, { "epoch": 0.8846729905142287, "grad_norm": 45.97512435913086, "learning_rate": 0.0001413854351687389, "loss": 0.3121, "step": 1329 }, { "epoch": 0.8853386586786487, "grad_norm": 53.834503173828125, "learning_rate": 0.0001413410301953819, "loss": 0.7437, "step": 1330 }, { "epoch": 0.8860043268430687, "grad_norm": 62.20562744140625, "learning_rate": 0.00014129662522202488, "loss": 0.5638, "step": 1331 }, { "epoch": 0.8866699950074888, "grad_norm": 29.805953979492188, "learning_rate": 0.00014125222024866788, "loss": 0.3874, "step": 1332 }, { "epoch": 0.8873356631719088, "grad_norm": 34.111080169677734, "learning_rate": 0.00014120781527531085, "loss": 0.253, "step": 1333 }, { "epoch": 0.8880013313363289, "grad_norm": 63.19011688232422, "learning_rate": 0.00014116341030195385, "loss": 0.5686, "step": 1334 }, { "epoch": 0.8886669995007489, "grad_norm": 48.80171203613281, "learning_rate": 0.00014111900532859682, "loss": 0.3301, "step": 1335 }, { "epoch": 0.8893326676651689, "grad_norm": 29.755435943603516, "learning_rate": 0.0001410746003552398, "loss": 0.1524, "step": 1336 }, { "epoch": 0.889998335829589, "grad_norm": 25.09624481201172, "learning_rate": 0.0001410301953818828, "loss": 0.3311, "step": 1337 }, { "epoch": 0.890664003994009, "grad_norm": 28.74748992919922, "learning_rate": 0.00014098579040852576, "loss": 0.2005, "step": 1338 }, { "epoch": 0.8913296721584291, "grad_norm": 57.81682205200195, "learning_rate": 0.00014094138543516876, "loss": 0.6848, "step": 1339 }, { "epoch": 0.8919953403228491, "grad_norm": 34.122703552246094, "learning_rate": 0.00014089698046181173, "loss": 0.6192, "step": 1340 }, { "epoch": 0.8926610084872691, "grad_norm": 87.72877502441406, "learning_rate": 0.00014085257548845473, "loss": 0.4187, "step": 1341 }, { "epoch": 0.8933266766516891, "grad_norm": 27.392183303833008, "learning_rate": 0.0001408081705150977, "loss": 0.1463, "step": 1342 }, { "epoch": 0.8939923448161091, "grad_norm": 14.832976341247559, "learning_rate": 0.0001407637655417407, "loss": 0.0909, "step": 1343 }, { "epoch": 0.8946580129805292, "grad_norm": 32.26533508300781, "learning_rate": 0.00014071936056838367, "loss": 0.4076, "step": 1344 }, { "epoch": 0.8953236811449492, "grad_norm": 37.978240966796875, "learning_rate": 0.00014067495559502667, "loss": 0.3945, "step": 1345 }, { "epoch": 0.8959893493093692, "grad_norm": 10.255926132202148, "learning_rate": 0.00014063055062166964, "loss": 0.0768, "step": 1346 }, { "epoch": 0.8966550174737893, "grad_norm": 36.25493621826172, "learning_rate": 0.00014058614564831264, "loss": 0.1747, "step": 1347 }, { "epoch": 0.8973206856382093, "grad_norm": 74.25749206542969, "learning_rate": 0.0001405417406749556, "loss": 0.8047, "step": 1348 }, { "epoch": 0.8979863538026294, "grad_norm": 76.57219696044922, "learning_rate": 0.00014049733570159858, "loss": 0.5247, "step": 1349 }, { "epoch": 0.8986520219670494, "grad_norm": 82.6986312866211, "learning_rate": 0.00014045293072824158, "loss": 0.6751, "step": 1350 }, { "epoch": 0.8993176901314694, "grad_norm": 61.353492736816406, "learning_rate": 0.00014040852575488455, "loss": 0.2196, "step": 1351 }, { "epoch": 0.8999833582958895, "grad_norm": 78.33421325683594, "learning_rate": 0.00014036412078152755, "loss": 0.7381, "step": 1352 }, { "epoch": 0.9006490264603095, "grad_norm": 26.842025756835938, "learning_rate": 0.00014031971580817052, "loss": 0.1704, "step": 1353 }, { "epoch": 0.9013146946247296, "grad_norm": 31.503564834594727, "learning_rate": 0.00014027531083481352, "loss": 0.5544, "step": 1354 }, { "epoch": 0.9019803627891496, "grad_norm": 47.245601654052734, "learning_rate": 0.0001402309058614565, "loss": 0.2578, "step": 1355 }, { "epoch": 0.9026460309535697, "grad_norm": 63.07925796508789, "learning_rate": 0.0001401865008880995, "loss": 0.3026, "step": 1356 }, { "epoch": 0.9033116991179897, "grad_norm": 42.88800048828125, "learning_rate": 0.00014014209591474246, "loss": 0.3575, "step": 1357 }, { "epoch": 0.9039773672824097, "grad_norm": 30.306550979614258, "learning_rate": 0.00014009769094138546, "loss": 0.2791, "step": 1358 }, { "epoch": 0.9046430354468298, "grad_norm": 85.1515884399414, "learning_rate": 0.00014005328596802843, "loss": 0.4204, "step": 1359 }, { "epoch": 0.9053087036112498, "grad_norm": 23.311250686645508, "learning_rate": 0.00014000888099467143, "loss": 0.1302, "step": 1360 }, { "epoch": 0.9059743717756699, "grad_norm": 25.332386016845703, "learning_rate": 0.0001399644760213144, "loss": 0.1528, "step": 1361 }, { "epoch": 0.9066400399400899, "grad_norm": 59.98972702026367, "learning_rate": 0.00013992007104795737, "loss": 0.6661, "step": 1362 }, { "epoch": 0.90730570810451, "grad_norm": 95.51297760009766, "learning_rate": 0.00013987566607460037, "loss": 0.7696, "step": 1363 }, { "epoch": 0.90797137626893, "grad_norm": 50.402889251708984, "learning_rate": 0.00013983126110124334, "loss": 0.4229, "step": 1364 }, { "epoch": 0.90863704443335, "grad_norm": 34.545589447021484, "learning_rate": 0.00013978685612788634, "loss": 0.1866, "step": 1365 }, { "epoch": 0.90930271259777, "grad_norm": 41.82796859741211, "learning_rate": 0.0001397424511545293, "loss": 0.2048, "step": 1366 }, { "epoch": 0.90996838076219, "grad_norm": 29.8832950592041, "learning_rate": 0.0001396980461811723, "loss": 0.1406, "step": 1367 }, { "epoch": 0.91063404892661, "grad_norm": 48.74016189575195, "learning_rate": 0.00013965364120781528, "loss": 0.2185, "step": 1368 }, { "epoch": 0.9112997170910301, "grad_norm": 36.24103546142578, "learning_rate": 0.00013960923623445828, "loss": 0.1463, "step": 1369 }, { "epoch": 0.9119653852554501, "grad_norm": 45.757957458496094, "learning_rate": 0.00013956483126110125, "loss": 0.5364, "step": 1370 }, { "epoch": 0.9126310534198702, "grad_norm": 42.56910705566406, "learning_rate": 0.00013952042628774425, "loss": 0.1913, "step": 1371 }, { "epoch": 0.9132967215842902, "grad_norm": 30.841646194458008, "learning_rate": 0.00013947602131438722, "loss": 0.1802, "step": 1372 }, { "epoch": 0.9139623897487102, "grad_norm": 41.267974853515625, "learning_rate": 0.0001394316163410302, "loss": 0.1613, "step": 1373 }, { "epoch": 0.9146280579131303, "grad_norm": 43.241153717041016, "learning_rate": 0.0001393872113676732, "loss": 0.261, "step": 1374 }, { "epoch": 0.9152937260775503, "grad_norm": 61.06812286376953, "learning_rate": 0.00013934280639431616, "loss": 0.4302, "step": 1375 }, { "epoch": 0.9159593942419704, "grad_norm": 49.418704986572266, "learning_rate": 0.00013929840142095916, "loss": 0.3749, "step": 1376 }, { "epoch": 0.9166250624063904, "grad_norm": 42.140464782714844, "learning_rate": 0.00013925399644760213, "loss": 0.1752, "step": 1377 }, { "epoch": 0.9172907305708105, "grad_norm": 109.88323974609375, "learning_rate": 0.00013920959147424513, "loss": 0.4314, "step": 1378 }, { "epoch": 0.9179563987352305, "grad_norm": 48.18379211425781, "learning_rate": 0.0001391651865008881, "loss": 0.1955, "step": 1379 }, { "epoch": 0.9186220668996505, "grad_norm": 77.02790832519531, "learning_rate": 0.0001391207815275311, "loss": 0.4732, "step": 1380 }, { "epoch": 0.9192877350640706, "grad_norm": 36.05132293701172, "learning_rate": 0.00013907637655417407, "loss": 0.3354, "step": 1381 }, { "epoch": 0.9199534032284906, "grad_norm": 81.78813934326172, "learning_rate": 0.00013903197158081707, "loss": 0.6251, "step": 1382 }, { "epoch": 0.9206190713929107, "grad_norm": 51.858238220214844, "learning_rate": 0.00013898756660746004, "loss": 0.703, "step": 1383 }, { "epoch": 0.9212847395573307, "grad_norm": 5.450369834899902, "learning_rate": 0.00013894316163410303, "loss": 0.0387, "step": 1384 }, { "epoch": 0.9219504077217507, "grad_norm": 29.26894760131836, "learning_rate": 0.000138898756660746, "loss": 0.1885, "step": 1385 }, { "epoch": 0.9226160758861708, "grad_norm": 43.31132125854492, "learning_rate": 0.00013885435168738898, "loss": 0.3355, "step": 1386 }, { "epoch": 0.9232817440505908, "grad_norm": 45.395751953125, "learning_rate": 0.00013880994671403198, "loss": 0.3363, "step": 1387 }, { "epoch": 0.9239474122150109, "grad_norm": 61.053855895996094, "learning_rate": 0.00013876554174067495, "loss": 0.2261, "step": 1388 }, { "epoch": 0.9246130803794309, "grad_norm": 31.03114128112793, "learning_rate": 0.00013872113676731795, "loss": 0.2035, "step": 1389 }, { "epoch": 0.9252787485438508, "grad_norm": 85.08035278320312, "learning_rate": 0.00013867673179396092, "loss": 0.7085, "step": 1390 }, { "epoch": 0.9259444167082709, "grad_norm": 48.80365753173828, "learning_rate": 0.00013863232682060392, "loss": 0.1879, "step": 1391 }, { "epoch": 0.9266100848726909, "grad_norm": 69.83169555664062, "learning_rate": 0.0001385879218472469, "loss": 0.526, "step": 1392 }, { "epoch": 0.927275753037111, "grad_norm": 46.45308303833008, "learning_rate": 0.00013854351687388988, "loss": 0.3331, "step": 1393 }, { "epoch": 0.927941421201531, "grad_norm": 32.49675750732422, "learning_rate": 0.00013849911190053286, "loss": 0.1529, "step": 1394 }, { "epoch": 0.928607089365951, "grad_norm": 72.4941635131836, "learning_rate": 0.00013845470692717585, "loss": 0.5155, "step": 1395 }, { "epoch": 0.9292727575303711, "grad_norm": 36.44300079345703, "learning_rate": 0.00013841030195381883, "loss": 0.4156, "step": 1396 }, { "epoch": 0.9299384256947911, "grad_norm": 53.10055923461914, "learning_rate": 0.00013836589698046182, "loss": 0.5358, "step": 1397 }, { "epoch": 0.9306040938592112, "grad_norm": 43.32754135131836, "learning_rate": 0.0001383214920071048, "loss": 0.1951, "step": 1398 }, { "epoch": 0.9312697620236312, "grad_norm": 69.81607055664062, "learning_rate": 0.00013827708703374777, "loss": 0.2728, "step": 1399 }, { "epoch": 0.9319354301880513, "grad_norm": 62.127567291259766, "learning_rate": 0.00013823268206039077, "loss": 0.3095, "step": 1400 }, { "epoch": 0.9326010983524713, "grad_norm": 62.99802780151367, "learning_rate": 0.00013818827708703374, "loss": 0.3631, "step": 1401 }, { "epoch": 0.9332667665168913, "grad_norm": 99.65278625488281, "learning_rate": 0.00013814387211367673, "loss": 0.8688, "step": 1402 }, { "epoch": 0.9339324346813114, "grad_norm": 87.81271362304688, "learning_rate": 0.0001380994671403197, "loss": 0.5893, "step": 1403 }, { "epoch": 0.9345981028457314, "grad_norm": 43.909976959228516, "learning_rate": 0.0001380550621669627, "loss": 0.6989, "step": 1404 }, { "epoch": 0.9352637710101515, "grad_norm": 65.83758544921875, "learning_rate": 0.00013801065719360568, "loss": 0.4555, "step": 1405 }, { "epoch": 0.9359294391745715, "grad_norm": 22.164623260498047, "learning_rate": 0.00013796625222024867, "loss": 0.2172, "step": 1406 }, { "epoch": 0.9365951073389915, "grad_norm": 41.753326416015625, "learning_rate": 0.00013792184724689165, "loss": 0.4236, "step": 1407 }, { "epoch": 0.9372607755034116, "grad_norm": 86.22735595703125, "learning_rate": 0.00013787744227353464, "loss": 0.6515, "step": 1408 }, { "epoch": 0.9379264436678316, "grad_norm": 24.779319763183594, "learning_rate": 0.00013783303730017762, "loss": 0.2877, "step": 1409 }, { "epoch": 0.9385921118322517, "grad_norm": 40.450740814208984, "learning_rate": 0.00013778863232682061, "loss": 0.3601, "step": 1410 }, { "epoch": 0.9392577799966717, "grad_norm": 39.6088752746582, "learning_rate": 0.00013774422735346358, "loss": 0.2147, "step": 1411 }, { "epoch": 0.9399234481610917, "grad_norm": 23.805906295776367, "learning_rate": 0.00013769982238010656, "loss": 0.3028, "step": 1412 }, { "epoch": 0.9405891163255118, "grad_norm": 23.02085304260254, "learning_rate": 0.00013765541740674955, "loss": 0.0836, "step": 1413 }, { "epoch": 0.9412547844899317, "grad_norm": 23.69212532043457, "learning_rate": 0.00013761101243339253, "loss": 0.2493, "step": 1414 }, { "epoch": 0.9419204526543518, "grad_norm": 28.35810089111328, "learning_rate": 0.00013756660746003552, "loss": 0.1171, "step": 1415 }, { "epoch": 0.9425861208187718, "grad_norm": 23.166410446166992, "learning_rate": 0.0001375222024866785, "loss": 0.1607, "step": 1416 }, { "epoch": 0.9432517889831918, "grad_norm": 33.61221694946289, "learning_rate": 0.0001374777975133215, "loss": 0.4199, "step": 1417 }, { "epoch": 0.9439174571476119, "grad_norm": 37.31405258178711, "learning_rate": 0.00013743339253996447, "loss": 0.2213, "step": 1418 }, { "epoch": 0.9445831253120319, "grad_norm": 56.77967834472656, "learning_rate": 0.00013738898756660746, "loss": 0.2945, "step": 1419 }, { "epoch": 0.945248793476452, "grad_norm": 23.5822696685791, "learning_rate": 0.00013734458259325043, "loss": 0.253, "step": 1420 }, { "epoch": 0.945914461640872, "grad_norm": 35.9830436706543, "learning_rate": 0.00013730017761989343, "loss": 0.1106, "step": 1421 }, { "epoch": 0.946580129805292, "grad_norm": 14.625288963317871, "learning_rate": 0.0001372557726465364, "loss": 0.1071, "step": 1422 }, { "epoch": 0.9472457979697121, "grad_norm": 121.55781555175781, "learning_rate": 0.0001372113676731794, "loss": 0.6414, "step": 1423 }, { "epoch": 0.9479114661341321, "grad_norm": 47.65458297729492, "learning_rate": 0.0001371669626998224, "loss": 0.2697, "step": 1424 }, { "epoch": 0.9485771342985522, "grad_norm": 93.63408660888672, "learning_rate": 0.00013712255772646537, "loss": 1.0909, "step": 1425 }, { "epoch": 0.9492428024629722, "grad_norm": 49.3912239074707, "learning_rate": 0.00013707815275310837, "loss": 0.3802, "step": 1426 }, { "epoch": 0.9499084706273923, "grad_norm": 40.63557434082031, "learning_rate": 0.00013703374777975134, "loss": 0.2589, "step": 1427 }, { "epoch": 0.9505741387918123, "grad_norm": 6.672457695007324, "learning_rate": 0.00013698934280639434, "loss": 0.0308, "step": 1428 }, { "epoch": 0.9512398069562323, "grad_norm": 44.79387283325195, "learning_rate": 0.0001369449378330373, "loss": 0.4385, "step": 1429 }, { "epoch": 0.9519054751206524, "grad_norm": 59.466426849365234, "learning_rate": 0.0001369005328596803, "loss": 0.4786, "step": 1430 }, { "epoch": 0.9525711432850724, "grad_norm": 30.637113571166992, "learning_rate": 0.00013685612788632328, "loss": 0.2503, "step": 1431 }, { "epoch": 0.9532368114494925, "grad_norm": 42.67710494995117, "learning_rate": 0.00013681172291296628, "loss": 0.2137, "step": 1432 }, { "epoch": 0.9539024796139125, "grad_norm": 63.8383674621582, "learning_rate": 0.00013676731793960925, "loss": 0.4558, "step": 1433 }, { "epoch": 0.9545681477783325, "grad_norm": 77.58506774902344, "learning_rate": 0.00013672291296625225, "loss": 0.4746, "step": 1434 }, { "epoch": 0.9552338159427526, "grad_norm": 52.78440856933594, "learning_rate": 0.00013667850799289522, "loss": 0.3198, "step": 1435 }, { "epoch": 0.9558994841071726, "grad_norm": 26.549100875854492, "learning_rate": 0.0001366341030195382, "loss": 0.0974, "step": 1436 }, { "epoch": 0.9565651522715926, "grad_norm": 36.16482925415039, "learning_rate": 0.0001365896980461812, "loss": 0.1307, "step": 1437 }, { "epoch": 0.9572308204360126, "grad_norm": 80.76986694335938, "learning_rate": 0.00013654529307282416, "loss": 0.3619, "step": 1438 }, { "epoch": 0.9578964886004326, "grad_norm": 38.59693145751953, "learning_rate": 0.00013650088809946716, "loss": 0.1823, "step": 1439 }, { "epoch": 0.9585621567648527, "grad_norm": 31.808364868164062, "learning_rate": 0.00013645648312611013, "loss": 0.1226, "step": 1440 }, { "epoch": 0.9592278249292727, "grad_norm": 53.049137115478516, "learning_rate": 0.00013641207815275313, "loss": 0.4268, "step": 1441 }, { "epoch": 0.9598934930936928, "grad_norm": 27.240991592407227, "learning_rate": 0.0001363676731793961, "loss": 0.0856, "step": 1442 }, { "epoch": 0.9605591612581128, "grad_norm": 35.355796813964844, "learning_rate": 0.0001363232682060391, "loss": 0.4293, "step": 1443 }, { "epoch": 0.9612248294225328, "grad_norm": 42.38942337036133, "learning_rate": 0.00013627886323268207, "loss": 0.2424, "step": 1444 }, { "epoch": 0.9618904975869529, "grad_norm": 100.4751205444336, "learning_rate": 0.00013623445825932507, "loss": 0.6466, "step": 1445 }, { "epoch": 0.9625561657513729, "grad_norm": 117.26541137695312, "learning_rate": 0.00013619005328596804, "loss": 0.5938, "step": 1446 }, { "epoch": 0.963221833915793, "grad_norm": 78.23147583007812, "learning_rate": 0.00013614564831261104, "loss": 0.7117, "step": 1447 }, { "epoch": 0.963887502080213, "grad_norm": 28.035341262817383, "learning_rate": 0.000136101243339254, "loss": 0.1225, "step": 1448 }, { "epoch": 0.964553170244633, "grad_norm": 40.73735427856445, "learning_rate": 0.00013605683836589698, "loss": 0.185, "step": 1449 }, { "epoch": 0.9652188384090531, "grad_norm": 43.69590377807617, "learning_rate": 0.00013601243339253998, "loss": 0.2746, "step": 1450 }, { "epoch": 0.9658845065734731, "grad_norm": 39.62010192871094, "learning_rate": 0.00013596802841918295, "loss": 0.3011, "step": 1451 }, { "epoch": 0.9665501747378932, "grad_norm": 73.28042602539062, "learning_rate": 0.00013592362344582595, "loss": 0.2665, "step": 1452 }, { "epoch": 0.9672158429023132, "grad_norm": 78.28668975830078, "learning_rate": 0.00013587921847246892, "loss": 0.3442, "step": 1453 }, { "epoch": 0.9678815110667333, "grad_norm": 31.521352767944336, "learning_rate": 0.00013583481349911192, "loss": 0.137, "step": 1454 }, { "epoch": 0.9685471792311533, "grad_norm": 78.63130950927734, "learning_rate": 0.0001357904085257549, "loss": 0.4162, "step": 1455 }, { "epoch": 0.9692128473955733, "grad_norm": 44.271484375, "learning_rate": 0.0001357460035523979, "loss": 0.1856, "step": 1456 }, { "epoch": 0.9698785155599934, "grad_norm": 24.507455825805664, "learning_rate": 0.00013570159857904086, "loss": 0.1343, "step": 1457 }, { "epoch": 0.9705441837244134, "grad_norm": 32.432655334472656, "learning_rate": 0.00013565719360568386, "loss": 0.3468, "step": 1458 }, { "epoch": 0.9712098518888335, "grad_norm": 78.22180938720703, "learning_rate": 0.00013561278863232683, "loss": 0.2939, "step": 1459 }, { "epoch": 0.9718755200532535, "grad_norm": 32.28510284423828, "learning_rate": 0.00013556838365896983, "loss": 0.1685, "step": 1460 }, { "epoch": 0.9725411882176734, "grad_norm": 33.94179916381836, "learning_rate": 0.0001355239786856128, "loss": 0.1021, "step": 1461 }, { "epoch": 0.9732068563820935, "grad_norm": 46.704872131347656, "learning_rate": 0.00013547957371225577, "loss": 0.5844, "step": 1462 }, { "epoch": 0.9738725245465135, "grad_norm": 47.116844177246094, "learning_rate": 0.00013543516873889877, "loss": 0.2231, "step": 1463 }, { "epoch": 0.9745381927109336, "grad_norm": 14.879481315612793, "learning_rate": 0.00013539076376554174, "loss": 0.0524, "step": 1464 }, { "epoch": 0.9752038608753536, "grad_norm": 42.56416702270508, "learning_rate": 0.00013534635879218474, "loss": 0.2752, "step": 1465 }, { "epoch": 0.9758695290397736, "grad_norm": 25.45599937438965, "learning_rate": 0.0001353019538188277, "loss": 0.1888, "step": 1466 }, { "epoch": 0.9765351972041937, "grad_norm": 55.972957611083984, "learning_rate": 0.0001352575488454707, "loss": 0.334, "step": 1467 }, { "epoch": 0.9772008653686137, "grad_norm": 39.72566604614258, "learning_rate": 0.00013521314387211368, "loss": 0.3426, "step": 1468 }, { "epoch": 0.9778665335330338, "grad_norm": 77.64808654785156, "learning_rate": 0.00013516873889875668, "loss": 0.7148, "step": 1469 }, { "epoch": 0.9785322016974538, "grad_norm": 116.59876251220703, "learning_rate": 0.00013512433392539965, "loss": 0.6387, "step": 1470 }, { "epoch": 0.9791978698618738, "grad_norm": 29.280254364013672, "learning_rate": 0.00013507992895204265, "loss": 0.1463, "step": 1471 }, { "epoch": 0.9798635380262939, "grad_norm": 17.300331115722656, "learning_rate": 0.00013503552397868562, "loss": 0.136, "step": 1472 }, { "epoch": 0.9805292061907139, "grad_norm": 6.958846569061279, "learning_rate": 0.0001349911190053286, "loss": 0.0218, "step": 1473 }, { "epoch": 0.981194874355134, "grad_norm": 24.05440902709961, "learning_rate": 0.0001349467140319716, "loss": 0.2125, "step": 1474 }, { "epoch": 0.981860542519554, "grad_norm": 74.2096176147461, "learning_rate": 0.00013490230905861456, "loss": 0.4106, "step": 1475 }, { "epoch": 0.982526210683974, "grad_norm": 88.6530990600586, "learning_rate": 0.00013485790408525756, "loss": 0.4514, "step": 1476 }, { "epoch": 0.9831918788483941, "grad_norm": 33.72336196899414, "learning_rate": 0.00013481349911190053, "loss": 0.1088, "step": 1477 }, { "epoch": 0.9838575470128141, "grad_norm": 50.04899978637695, "learning_rate": 0.00013476909413854353, "loss": 0.3598, "step": 1478 }, { "epoch": 0.9845232151772342, "grad_norm": 42.27912521362305, "learning_rate": 0.0001347246891651865, "loss": 0.5035, "step": 1479 }, { "epoch": 0.9851888833416542, "grad_norm": 20.637144088745117, "learning_rate": 0.0001346802841918295, "loss": 0.0514, "step": 1480 }, { "epoch": 0.9858545515060743, "grad_norm": 33.74753189086914, "learning_rate": 0.00013463587921847247, "loss": 0.506, "step": 1481 }, { "epoch": 0.9865202196704943, "grad_norm": 108.18144989013672, "learning_rate": 0.00013459147424511547, "loss": 0.4964, "step": 1482 }, { "epoch": 0.9871858878349143, "grad_norm": 54.47599411010742, "learning_rate": 0.00013454706927175844, "loss": 0.2487, "step": 1483 }, { "epoch": 0.9878515559993344, "grad_norm": 30.460220336914062, "learning_rate": 0.00013450266429840144, "loss": 0.108, "step": 1484 }, { "epoch": 0.9885172241637543, "grad_norm": 63.42583465576172, "learning_rate": 0.0001344582593250444, "loss": 0.5582, "step": 1485 }, { "epoch": 0.9891828923281744, "grad_norm": 57.32478332519531, "learning_rate": 0.00013441385435168738, "loss": 0.4496, "step": 1486 }, { "epoch": 0.9898485604925944, "grad_norm": 36.21717834472656, "learning_rate": 0.00013436944937833038, "loss": 0.3115, "step": 1487 }, { "epoch": 0.9905142286570144, "grad_norm": 99.90687561035156, "learning_rate": 0.00013432504440497335, "loss": 0.75, "step": 1488 }, { "epoch": 0.9911798968214345, "grad_norm": 38.466976165771484, "learning_rate": 0.00013428063943161635, "loss": 0.1495, "step": 1489 }, { "epoch": 0.9918455649858545, "grad_norm": 60.86931610107422, "learning_rate": 0.00013423623445825932, "loss": 0.2404, "step": 1490 }, { "epoch": 0.9925112331502746, "grad_norm": 40.01704025268555, "learning_rate": 0.00013419182948490232, "loss": 0.5159, "step": 1491 }, { "epoch": 0.9931769013146946, "grad_norm": 41.067283630371094, "learning_rate": 0.0001341474245115453, "loss": 0.626, "step": 1492 }, { "epoch": 0.9938425694791146, "grad_norm": 60.9128303527832, "learning_rate": 0.0001341030195381883, "loss": 0.4464, "step": 1493 }, { "epoch": 0.9945082376435347, "grad_norm": 28.736631393432617, "learning_rate": 0.00013405861456483126, "loss": 0.2869, "step": 1494 }, { "epoch": 0.9951739058079547, "grad_norm": 27.92344856262207, "learning_rate": 0.00013401420959147426, "loss": 0.1525, "step": 1495 }, { "epoch": 0.9958395739723748, "grad_norm": 35.2568473815918, "learning_rate": 0.00013396980461811723, "loss": 0.2093, "step": 1496 }, { "epoch": 0.9965052421367948, "grad_norm": 13.281268119812012, "learning_rate": 0.00013392539964476023, "loss": 0.1006, "step": 1497 }, { "epoch": 0.9971709103012149, "grad_norm": 39.88976287841797, "learning_rate": 0.0001338809946714032, "loss": 0.3204, "step": 1498 }, { "epoch": 0.9978365784656349, "grad_norm": 28.20810890197754, "learning_rate": 0.00013383658969804617, "loss": 0.1764, "step": 1499 }, { "epoch": 0.9985022466300549, "grad_norm": 16.875349044799805, "learning_rate": 0.00013379218472468917, "loss": 0.0514, "step": 1500 }, { "epoch": 0.999167914794475, "grad_norm": 34.00355911254883, "learning_rate": 0.00013374777975133214, "loss": 0.2051, "step": 1501 }, { "epoch": 0.999833582958895, "grad_norm": 42.16084289550781, "learning_rate": 0.00013370337477797514, "loss": 0.1819, "step": 1502 }, { "epoch": 1.000499251123315, "grad_norm": 16.130739212036133, "learning_rate": 0.0001336589698046181, "loss": 0.0762, "step": 1503 }, { "epoch": 1.001164919287735, "grad_norm": 55.9047737121582, "learning_rate": 0.0001336145648312611, "loss": 0.2071, "step": 1504 }, { "epoch": 1.0018305874521551, "grad_norm": 31.728910446166992, "learning_rate": 0.00013357015985790408, "loss": 0.2364, "step": 1505 }, { "epoch": 1.002496255616575, "grad_norm": 44.72419738769531, "learning_rate": 0.00013352575488454708, "loss": 0.2769, "step": 1506 }, { "epoch": 1.0031619237809952, "grad_norm": 61.57970428466797, "learning_rate": 0.00013348134991119005, "loss": 0.4729, "step": 1507 }, { "epoch": 1.0038275919454152, "grad_norm": 35.890933990478516, "learning_rate": 0.00013343694493783305, "loss": 0.3703, "step": 1508 }, { "epoch": 1.0044932601098353, "grad_norm": 69.92425537109375, "learning_rate": 0.00013339253996447602, "loss": 0.6699, "step": 1509 }, { "epoch": 1.0051589282742552, "grad_norm": 77.50531768798828, "learning_rate": 0.00013334813499111902, "loss": 0.2947, "step": 1510 }, { "epoch": 1.0058245964386754, "grad_norm": 67.69709777832031, "learning_rate": 0.000133303730017762, "loss": 0.5026, "step": 1511 }, { "epoch": 1.0064902646030953, "grad_norm": 54.660335540771484, "learning_rate": 0.00013325932504440496, "loss": 0.6443, "step": 1512 }, { "epoch": 1.0071559327675155, "grad_norm": 90.5311508178711, "learning_rate": 0.00013321492007104796, "loss": 0.6667, "step": 1513 }, { "epoch": 1.0078216009319354, "grad_norm": 37.90154266357422, "learning_rate": 0.00013317051509769093, "loss": 0.2237, "step": 1514 }, { "epoch": 1.0084872690963556, "grad_norm": 24.480512619018555, "learning_rate": 0.00013312611012433393, "loss": 0.0836, "step": 1515 }, { "epoch": 1.0091529372607755, "grad_norm": 61.944149017333984, "learning_rate": 0.0001330817051509769, "loss": 0.3446, "step": 1516 }, { "epoch": 1.0098186054251956, "grad_norm": 17.98834800720215, "learning_rate": 0.0001330373001776199, "loss": 0.1657, "step": 1517 }, { "epoch": 1.0104842735896156, "grad_norm": 62.744529724121094, "learning_rate": 0.00013299289520426287, "loss": 0.3193, "step": 1518 }, { "epoch": 1.0111499417540357, "grad_norm": 43.202842712402344, "learning_rate": 0.00013294849023090587, "loss": 0.3193, "step": 1519 }, { "epoch": 1.0118156099184556, "grad_norm": 6.932873249053955, "learning_rate": 0.00013290408525754884, "loss": 0.0116, "step": 1520 }, { "epoch": 1.0124812780828756, "grad_norm": 93.96986389160156, "learning_rate": 0.00013285968028419184, "loss": 0.627, "step": 1521 }, { "epoch": 1.0131469462472957, "grad_norm": 59.124019622802734, "learning_rate": 0.0001328152753108348, "loss": 0.2239, "step": 1522 }, { "epoch": 1.0138126144117157, "grad_norm": 70.20691680908203, "learning_rate": 0.0001327708703374778, "loss": 0.2583, "step": 1523 }, { "epoch": 1.0144782825761358, "grad_norm": 50.7972412109375, "learning_rate": 0.00013272646536412078, "loss": 0.2726, "step": 1524 }, { "epoch": 1.0151439507405557, "grad_norm": 35.38340759277344, "learning_rate": 0.00013268206039076377, "loss": 0.2039, "step": 1525 }, { "epoch": 1.015809618904976, "grad_norm": 33.25050354003906, "learning_rate": 0.00013263765541740675, "loss": 0.2169, "step": 1526 }, { "epoch": 1.0164752870693958, "grad_norm": 59.16511154174805, "learning_rate": 0.00013259325044404974, "loss": 0.2251, "step": 1527 }, { "epoch": 1.017140955233816, "grad_norm": 30.06669807434082, "learning_rate": 0.00013254884547069272, "loss": 0.1363, "step": 1528 }, { "epoch": 1.017806623398236, "grad_norm": 29.909889221191406, "learning_rate": 0.00013250444049733571, "loss": 0.1037, "step": 1529 }, { "epoch": 1.018472291562656, "grad_norm": 38.46884536743164, "learning_rate": 0.00013246003552397869, "loss": 0.1356, "step": 1530 }, { "epoch": 1.019137959727076, "grad_norm": 39.98774719238281, "learning_rate": 0.00013241563055062168, "loss": 0.1168, "step": 1531 }, { "epoch": 1.0198036278914961, "grad_norm": 43.331912994384766, "learning_rate": 0.00013237122557726466, "loss": 0.2762, "step": 1532 }, { "epoch": 1.020469296055916, "grad_norm": 35.60062026977539, "learning_rate": 0.00013232682060390765, "loss": 0.1878, "step": 1533 }, { "epoch": 1.0211349642203362, "grad_norm": 38.046566009521484, "learning_rate": 0.00013228241563055062, "loss": 0.3602, "step": 1534 }, { "epoch": 1.0218006323847562, "grad_norm": 25.638227462768555, "learning_rate": 0.00013223801065719362, "loss": 0.2112, "step": 1535 }, { "epoch": 1.0224663005491763, "grad_norm": 74.6210708618164, "learning_rate": 0.0001321936056838366, "loss": 0.6001, "step": 1536 }, { "epoch": 1.0231319687135962, "grad_norm": 35.4752082824707, "learning_rate": 0.0001321492007104796, "loss": 0.1107, "step": 1537 }, { "epoch": 1.0237976368780164, "grad_norm": 30.31052017211914, "learning_rate": 0.00013210479573712256, "loss": 0.1362, "step": 1538 }, { "epoch": 1.0244633050424363, "grad_norm": 31.850797653198242, "learning_rate": 0.00013206039076376556, "loss": 0.0855, "step": 1539 }, { "epoch": 1.0251289732068565, "grad_norm": 36.738197326660156, "learning_rate": 0.00013201598579040853, "loss": 0.2649, "step": 1540 }, { "epoch": 1.0257946413712764, "grad_norm": 28.015552520751953, "learning_rate": 0.00013197158081705153, "loss": 0.1784, "step": 1541 }, { "epoch": 1.0264603095356966, "grad_norm": 104.23088073730469, "learning_rate": 0.0001319271758436945, "loss": 0.5053, "step": 1542 }, { "epoch": 1.0271259777001165, "grad_norm": 128.6961669921875, "learning_rate": 0.0001318827708703375, "loss": 0.8857, "step": 1543 }, { "epoch": 1.0277916458645366, "grad_norm": 68.6438217163086, "learning_rate": 0.00013183836589698047, "loss": 0.6653, "step": 1544 }, { "epoch": 1.0284573140289566, "grad_norm": 0.619352400302887, "learning_rate": 0.00013179396092362347, "loss": 0.0012, "step": 1545 }, { "epoch": 1.0291229821933765, "grad_norm": 130.28871154785156, "learning_rate": 0.00013174955595026644, "loss": 0.7131, "step": 1546 }, { "epoch": 1.0297886503577967, "grad_norm": 76.49749755859375, "learning_rate": 0.00013170515097690944, "loss": 0.4285, "step": 1547 }, { "epoch": 1.0304543185222166, "grad_norm": 22.493146896362305, "learning_rate": 0.0001316607460035524, "loss": 0.0754, "step": 1548 }, { "epoch": 1.0311199866866367, "grad_norm": 3.85562801361084, "learning_rate": 0.00013161634103019538, "loss": 0.0106, "step": 1549 }, { "epoch": 1.0317856548510567, "grad_norm": 22.619091033935547, "learning_rate": 0.00013157193605683838, "loss": 0.0925, "step": 1550 }, { "epoch": 1.0324513230154768, "grad_norm": 51.793067932128906, "learning_rate": 0.00013152753108348135, "loss": 0.1817, "step": 1551 }, { "epoch": 1.0331169911798967, "grad_norm": 8.747608184814453, "learning_rate": 0.00013148312611012435, "loss": 0.0302, "step": 1552 }, { "epoch": 1.033782659344317, "grad_norm": 46.68729019165039, "learning_rate": 0.00013143872113676732, "loss": 0.2148, "step": 1553 }, { "epoch": 1.0344483275087368, "grad_norm": 59.01924133300781, "learning_rate": 0.00013139431616341032, "loss": 0.4396, "step": 1554 }, { "epoch": 1.035113995673157, "grad_norm": 137.48919677734375, "learning_rate": 0.0001313499111900533, "loss": 1.1802, "step": 1555 }, { "epoch": 1.035779663837577, "grad_norm": 86.36365509033203, "learning_rate": 0.0001313055062166963, "loss": 0.148, "step": 1556 }, { "epoch": 1.036445332001997, "grad_norm": 117.85287475585938, "learning_rate": 0.00013126110124333926, "loss": 0.2761, "step": 1557 }, { "epoch": 1.037111000166417, "grad_norm": 84.41642761230469, "learning_rate": 0.00013121669626998226, "loss": 0.5909, "step": 1558 }, { "epoch": 1.0377766683308371, "grad_norm": 109.68749237060547, "learning_rate": 0.00013117229129662523, "loss": 0.4269, "step": 1559 }, { "epoch": 1.038442336495257, "grad_norm": 61.24327850341797, "learning_rate": 0.00013112788632326823, "loss": 0.1997, "step": 1560 }, { "epoch": 1.0391080046596772, "grad_norm": 111.8650131225586, "learning_rate": 0.0001310834813499112, "loss": 1.1649, "step": 1561 }, { "epoch": 1.0397736728240972, "grad_norm": 132.17225646972656, "learning_rate": 0.00013103907637655417, "loss": 0.6326, "step": 1562 }, { "epoch": 1.0404393409885173, "grad_norm": 49.30635452270508, "learning_rate": 0.00013099467140319717, "loss": 0.4752, "step": 1563 }, { "epoch": 1.0411050091529372, "grad_norm": 66.80237579345703, "learning_rate": 0.00013095026642984014, "loss": 0.1477, "step": 1564 }, { "epoch": 1.0417706773173574, "grad_norm": 88.0537109375, "learning_rate": 0.00013090586145648314, "loss": 0.364, "step": 1565 }, { "epoch": 1.0424363454817773, "grad_norm": 210.458740234375, "learning_rate": 0.0001308614564831261, "loss": 1.8179, "step": 1566 }, { "epoch": 1.0431020136461973, "grad_norm": 50.41940689086914, "learning_rate": 0.0001308170515097691, "loss": 0.3693, "step": 1567 }, { "epoch": 1.0437676818106174, "grad_norm": 63.888675689697266, "learning_rate": 0.00013077264653641208, "loss": 0.2518, "step": 1568 }, { "epoch": 1.0444333499750373, "grad_norm": 12.438095092773438, "learning_rate": 0.00013072824156305508, "loss": 0.0361, "step": 1569 }, { "epoch": 1.0450990181394575, "grad_norm": 2.3615851402282715, "learning_rate": 0.00013068383658969805, "loss": 0.0051, "step": 1570 }, { "epoch": 1.0457646863038774, "grad_norm": 56.62102127075195, "learning_rate": 0.00013063943161634105, "loss": 0.423, "step": 1571 }, { "epoch": 1.0464303544682976, "grad_norm": 65.62693786621094, "learning_rate": 0.00013059502664298402, "loss": 0.6279, "step": 1572 }, { "epoch": 1.0470960226327175, "grad_norm": 14.44961929321289, "learning_rate": 0.000130550621669627, "loss": 0.0383, "step": 1573 }, { "epoch": 1.0477616907971377, "grad_norm": 29.319581985473633, "learning_rate": 0.00013050621669627, "loss": 0.0845, "step": 1574 }, { "epoch": 1.0484273589615576, "grad_norm": 47.517276763916016, "learning_rate": 0.00013046181172291296, "loss": 0.1676, "step": 1575 }, { "epoch": 1.0490930271259777, "grad_norm": 16.043867111206055, "learning_rate": 0.00013041740674955596, "loss": 0.0871, "step": 1576 }, { "epoch": 1.0497586952903977, "grad_norm": 87.3962173461914, "learning_rate": 0.00013037300177619893, "loss": 0.6998, "step": 1577 }, { "epoch": 1.0504243634548178, "grad_norm": 46.2629508972168, "learning_rate": 0.00013032859680284193, "loss": 0.1643, "step": 1578 }, { "epoch": 1.0510900316192378, "grad_norm": 7.9480977058410645, "learning_rate": 0.0001302841918294849, "loss": 0.0268, "step": 1579 }, { "epoch": 1.051755699783658, "grad_norm": 74.5625228881836, "learning_rate": 0.0001302397868561279, "loss": 0.5006, "step": 1580 }, { "epoch": 1.0524213679480778, "grad_norm": 51.07015609741211, "learning_rate": 0.00013019538188277087, "loss": 0.1402, "step": 1581 }, { "epoch": 1.053087036112498, "grad_norm": 43.359134674072266, "learning_rate": 0.00013015097690941387, "loss": 0.238, "step": 1582 }, { "epoch": 1.053752704276918, "grad_norm": 69.16613006591797, "learning_rate": 0.00013010657193605684, "loss": 0.6622, "step": 1583 }, { "epoch": 1.054418372441338, "grad_norm": 49.51469039916992, "learning_rate": 0.00013006216696269984, "loss": 0.4137, "step": 1584 }, { "epoch": 1.055084040605758, "grad_norm": 34.495582580566406, "learning_rate": 0.0001300177619893428, "loss": 0.1008, "step": 1585 }, { "epoch": 1.0557497087701782, "grad_norm": 21.827096939086914, "learning_rate": 0.00012997335701598578, "loss": 0.0653, "step": 1586 }, { "epoch": 1.056415376934598, "grad_norm": 29.265003204345703, "learning_rate": 0.00012992895204262878, "loss": 0.1136, "step": 1587 }, { "epoch": 1.0570810450990182, "grad_norm": 62.32855987548828, "learning_rate": 0.00012988454706927175, "loss": 0.5083, "step": 1588 }, { "epoch": 1.0577467132634382, "grad_norm": 24.315889358520508, "learning_rate": 0.00012984014209591475, "loss": 0.0713, "step": 1589 }, { "epoch": 1.0584123814278583, "grad_norm": 49.413063049316406, "learning_rate": 0.00012979573712255772, "loss": 0.4701, "step": 1590 }, { "epoch": 1.0590780495922782, "grad_norm": 26.282974243164062, "learning_rate": 0.00012975133214920072, "loss": 0.0711, "step": 1591 }, { "epoch": 1.0597437177566982, "grad_norm": 54.755165100097656, "learning_rate": 0.0001297069271758437, "loss": 0.3837, "step": 1592 }, { "epoch": 1.0604093859211183, "grad_norm": 44.34020233154297, "learning_rate": 0.0001296625222024867, "loss": 0.3189, "step": 1593 }, { "epoch": 1.0610750540855383, "grad_norm": 49.961700439453125, "learning_rate": 0.00012961811722912966, "loss": 0.7676, "step": 1594 }, { "epoch": 1.0617407222499584, "grad_norm": 13.223048210144043, "learning_rate": 0.00012957371225577266, "loss": 0.0626, "step": 1595 }, { "epoch": 1.0624063904143783, "grad_norm": 50.694759368896484, "learning_rate": 0.00012952930728241563, "loss": 0.1442, "step": 1596 }, { "epoch": 1.0630720585787985, "grad_norm": 54.78448486328125, "learning_rate": 0.00012948490230905863, "loss": 0.3928, "step": 1597 }, { "epoch": 1.0637377267432184, "grad_norm": 52.82820510864258, "learning_rate": 0.0001294404973357016, "loss": 0.2117, "step": 1598 }, { "epoch": 1.0644033949076386, "grad_norm": 72.75708770751953, "learning_rate": 0.00012939609236234457, "loss": 0.411, "step": 1599 }, { "epoch": 1.0650690630720585, "grad_norm": 72.80860900878906, "learning_rate": 0.00012935168738898757, "loss": 0.5258, "step": 1600 }, { "epoch": 1.0657347312364787, "grad_norm": 29.469266891479492, "learning_rate": 0.00012930728241563054, "loss": 0.1163, "step": 1601 }, { "epoch": 1.0664003994008986, "grad_norm": 32.04841995239258, "learning_rate": 0.00012926287744227354, "loss": 0.1921, "step": 1602 }, { "epoch": 1.0670660675653187, "grad_norm": 13.9080171585083, "learning_rate": 0.0001292184724689165, "loss": 0.0954, "step": 1603 }, { "epoch": 1.0677317357297387, "grad_norm": 7.307251453399658, "learning_rate": 0.0001291740674955595, "loss": 0.023, "step": 1604 }, { "epoch": 1.0683974038941588, "grad_norm": 20.18907928466797, "learning_rate": 0.00012912966252220248, "loss": 0.0822, "step": 1605 }, { "epoch": 1.0690630720585788, "grad_norm": 47.4912223815918, "learning_rate": 0.00012908525754884548, "loss": 0.3766, "step": 1606 }, { "epoch": 1.069728740222999, "grad_norm": 119.94644927978516, "learning_rate": 0.00012904085257548845, "loss": 0.1534, "step": 1607 }, { "epoch": 1.0703944083874188, "grad_norm": 52.42042541503906, "learning_rate": 0.00012899644760213145, "loss": 0.3412, "step": 1608 }, { "epoch": 1.071060076551839, "grad_norm": 107.58332824707031, "learning_rate": 0.00012895204262877442, "loss": 0.4823, "step": 1609 }, { "epoch": 1.071725744716259, "grad_norm": 44.404090881347656, "learning_rate": 0.00012890763765541742, "loss": 0.1148, "step": 1610 }, { "epoch": 1.072391412880679, "grad_norm": 38.6839599609375, "learning_rate": 0.0001288632326820604, "loss": 0.2979, "step": 1611 }, { "epoch": 1.073057081045099, "grad_norm": 9.74061393737793, "learning_rate": 0.00012881882770870336, "loss": 0.0265, "step": 1612 }, { "epoch": 1.0737227492095192, "grad_norm": 39.030433654785156, "learning_rate": 0.00012877442273534636, "loss": 0.1866, "step": 1613 }, { "epoch": 1.074388417373939, "grad_norm": 46.30801010131836, "learning_rate": 0.00012873001776198933, "loss": 0.1841, "step": 1614 }, { "epoch": 1.0750540855383592, "grad_norm": 38.60115432739258, "learning_rate": 0.00012868561278863233, "loss": 0.1381, "step": 1615 }, { "epoch": 1.0757197537027792, "grad_norm": 27.713333129882812, "learning_rate": 0.0001286412078152753, "loss": 0.0823, "step": 1616 }, { "epoch": 1.076385421867199, "grad_norm": 99.10941314697266, "learning_rate": 0.0001285968028419183, "loss": 0.5727, "step": 1617 }, { "epoch": 1.0770510900316193, "grad_norm": 5.887019634246826, "learning_rate": 0.00012855239786856127, "loss": 0.0135, "step": 1618 }, { "epoch": 1.0777167581960392, "grad_norm": 69.98397064208984, "learning_rate": 0.00012850799289520427, "loss": 0.2469, "step": 1619 }, { "epoch": 1.0783824263604593, "grad_norm": 15.54038143157959, "learning_rate": 0.00012846358792184724, "loss": 0.0274, "step": 1620 }, { "epoch": 1.0790480945248793, "grad_norm": 27.956186294555664, "learning_rate": 0.00012841918294849024, "loss": 0.0812, "step": 1621 }, { "epoch": 1.0797137626892994, "grad_norm": 16.48492431640625, "learning_rate": 0.0001283747779751332, "loss": 0.074, "step": 1622 }, { "epoch": 1.0803794308537193, "grad_norm": 16.525527954101562, "learning_rate": 0.00012833037300177618, "loss": 0.0337, "step": 1623 }, { "epoch": 1.0810450990181395, "grad_norm": 99.67036437988281, "learning_rate": 0.00012828596802841918, "loss": 0.8465, "step": 1624 }, { "epoch": 1.0817107671825594, "grad_norm": 7.101929187774658, "learning_rate": 0.00012824156305506218, "loss": 0.0128, "step": 1625 }, { "epoch": 1.0823764353469796, "grad_norm": 76.10822296142578, "learning_rate": 0.00012819715808170515, "loss": 0.9137, "step": 1626 }, { "epoch": 1.0830421035113995, "grad_norm": 118.75426483154297, "learning_rate": 0.00012815275310834815, "loss": 0.904, "step": 1627 }, { "epoch": 1.0837077716758197, "grad_norm": 131.6405792236328, "learning_rate": 0.00012810834813499112, "loss": 0.8974, "step": 1628 }, { "epoch": 1.0843734398402396, "grad_norm": 113.39791870117188, "learning_rate": 0.00012806394316163412, "loss": 0.6925, "step": 1629 }, { "epoch": 1.0850391080046597, "grad_norm": 119.22786712646484, "learning_rate": 0.0001280195381882771, "loss": 0.8384, "step": 1630 }, { "epoch": 1.0857047761690797, "grad_norm": 90.8923568725586, "learning_rate": 0.00012797513321492009, "loss": 0.4588, "step": 1631 }, { "epoch": 1.0863704443334998, "grad_norm": 50.50541305541992, "learning_rate": 0.00012793072824156306, "loss": 0.6814, "step": 1632 }, { "epoch": 1.0870361124979198, "grad_norm": 32.594512939453125, "learning_rate": 0.00012788632326820606, "loss": 0.3751, "step": 1633 }, { "epoch": 1.08770178066234, "grad_norm": 64.97533416748047, "learning_rate": 0.00012784191829484903, "loss": 0.1222, "step": 1634 }, { "epoch": 1.0883674488267598, "grad_norm": 54.40082550048828, "learning_rate": 0.00012779751332149203, "loss": 0.2244, "step": 1635 }, { "epoch": 1.08903311699118, "grad_norm": 60.19000244140625, "learning_rate": 0.000127753108348135, "loss": 0.4115, "step": 1636 }, { "epoch": 1.0896987851556, "grad_norm": 126.69731903076172, "learning_rate": 0.000127708703374778, "loss": 1.2657, "step": 1637 }, { "epoch": 1.0903644533200199, "grad_norm": 56.9218864440918, "learning_rate": 0.00012766429840142097, "loss": 0.1505, "step": 1638 }, { "epoch": 1.09103012148444, "grad_norm": 32.740814208984375, "learning_rate": 0.00012761989342806397, "loss": 0.1448, "step": 1639 }, { "epoch": 1.0916957896488602, "grad_norm": 14.416125297546387, "learning_rate": 0.00012757548845470694, "loss": 0.0804, "step": 1640 }, { "epoch": 1.09236145781328, "grad_norm": 33.17654800415039, "learning_rate": 0.00012753108348134993, "loss": 0.6931, "step": 1641 }, { "epoch": 1.0930271259777, "grad_norm": 39.734989166259766, "learning_rate": 0.0001274866785079929, "loss": 0.3041, "step": 1642 }, { "epoch": 1.0936927941421202, "grad_norm": 47.25416564941406, "learning_rate": 0.0001274422735346359, "loss": 0.0974, "step": 1643 }, { "epoch": 1.09435846230654, "grad_norm": 78.47022247314453, "learning_rate": 0.00012739786856127888, "loss": 0.4056, "step": 1644 }, { "epoch": 1.0950241304709603, "grad_norm": 20.88690757751465, "learning_rate": 0.00012735346358792187, "loss": 0.1163, "step": 1645 }, { "epoch": 1.0956897986353802, "grad_norm": 18.71324348449707, "learning_rate": 0.00012730905861456485, "loss": 0.1235, "step": 1646 }, { "epoch": 1.0963554667998003, "grad_norm": 77.86106872558594, "learning_rate": 0.00012726465364120784, "loss": 0.5394, "step": 1647 }, { "epoch": 1.0970211349642203, "grad_norm": 21.836231231689453, "learning_rate": 0.00012722024866785081, "loss": 0.1334, "step": 1648 }, { "epoch": 1.0976868031286404, "grad_norm": 48.32339859008789, "learning_rate": 0.00012717584369449379, "loss": 0.4802, "step": 1649 }, { "epoch": 1.0983524712930604, "grad_norm": 59.782936096191406, "learning_rate": 0.00012713143872113678, "loss": 0.3029, "step": 1650 }, { "epoch": 1.0990181394574805, "grad_norm": 24.361541748046875, "learning_rate": 0.00012708703374777976, "loss": 0.1564, "step": 1651 }, { "epoch": 1.0996838076219004, "grad_norm": 47.993675231933594, "learning_rate": 0.00012704262877442275, "loss": 0.2522, "step": 1652 }, { "epoch": 1.1003494757863206, "grad_norm": 43.45526885986328, "learning_rate": 0.00012699822380106573, "loss": 0.3656, "step": 1653 }, { "epoch": 1.1010151439507405, "grad_norm": 27.111066818237305, "learning_rate": 0.00012695381882770872, "loss": 0.1286, "step": 1654 }, { "epoch": 1.1016808121151607, "grad_norm": 65.31238555908203, "learning_rate": 0.0001269094138543517, "loss": 0.5258, "step": 1655 }, { "epoch": 1.1023464802795806, "grad_norm": 15.900341033935547, "learning_rate": 0.0001268650088809947, "loss": 0.0505, "step": 1656 }, { "epoch": 1.1030121484440008, "grad_norm": 65.35611724853516, "learning_rate": 0.00012682060390763766, "loss": 0.299, "step": 1657 }, { "epoch": 1.1036778166084207, "grad_norm": 46.87686538696289, "learning_rate": 0.00012677619893428066, "loss": 0.2593, "step": 1658 }, { "epoch": 1.1043434847728408, "grad_norm": 101.28860473632812, "learning_rate": 0.00012673179396092363, "loss": 0.7526, "step": 1659 }, { "epoch": 1.1050091529372608, "grad_norm": 34.1505241394043, "learning_rate": 0.00012668738898756663, "loss": 0.0911, "step": 1660 }, { "epoch": 1.105674821101681, "grad_norm": 61.11198043823242, "learning_rate": 0.0001266429840142096, "loss": 0.1532, "step": 1661 }, { "epoch": 1.1063404892661008, "grad_norm": 96.3040542602539, "learning_rate": 0.00012659857904085258, "loss": 1.0265, "step": 1662 }, { "epoch": 1.1070061574305208, "grad_norm": 40.270896911621094, "learning_rate": 0.00012655417406749557, "loss": 0.1925, "step": 1663 }, { "epoch": 1.107671825594941, "grad_norm": 9.517655372619629, "learning_rate": 0.00012650976909413855, "loss": 0.029, "step": 1664 }, { "epoch": 1.1083374937593609, "grad_norm": 58.8939323425293, "learning_rate": 0.00012646536412078154, "loss": 0.2334, "step": 1665 }, { "epoch": 1.109003161923781, "grad_norm": 11.043906211853027, "learning_rate": 0.00012642095914742451, "loss": 0.0339, "step": 1666 }, { "epoch": 1.109668830088201, "grad_norm": 40.188777923583984, "learning_rate": 0.0001263765541740675, "loss": 0.1958, "step": 1667 }, { "epoch": 1.110334498252621, "grad_norm": 13.981783866882324, "learning_rate": 0.00012633214920071048, "loss": 0.045, "step": 1668 }, { "epoch": 1.111000166417041, "grad_norm": 16.876413345336914, "learning_rate": 0.00012628774422735348, "loss": 0.0443, "step": 1669 }, { "epoch": 1.1116658345814612, "grad_norm": 46.84407424926758, "learning_rate": 0.00012624333925399645, "loss": 0.132, "step": 1670 }, { "epoch": 1.112331502745881, "grad_norm": 44.368038177490234, "learning_rate": 0.00012619893428063945, "loss": 0.1648, "step": 1671 }, { "epoch": 1.1129971709103013, "grad_norm": 67.10077667236328, "learning_rate": 0.00012615452930728242, "loss": 0.5709, "step": 1672 }, { "epoch": 1.1136628390747212, "grad_norm": 67.34343719482422, "learning_rate": 0.0001261101243339254, "loss": 0.531, "step": 1673 }, { "epoch": 1.1143285072391413, "grad_norm": 28.69812774658203, "learning_rate": 0.0001260657193605684, "loss": 0.105, "step": 1674 }, { "epoch": 1.1149941754035613, "grad_norm": 86.9577865600586, "learning_rate": 0.00012602131438721136, "loss": 0.4906, "step": 1675 }, { "epoch": 1.1156598435679814, "grad_norm": 37.13364028930664, "learning_rate": 0.00012597690941385436, "loss": 0.3584, "step": 1676 }, { "epoch": 1.1163255117324014, "grad_norm": 69.33271026611328, "learning_rate": 0.00012593250444049733, "loss": 0.1862, "step": 1677 }, { "epoch": 1.1169911798968215, "grad_norm": 36.56421661376953, "learning_rate": 0.00012588809946714033, "loss": 0.1307, "step": 1678 }, { "epoch": 1.1176568480612414, "grad_norm": 40.67729187011719, "learning_rate": 0.0001258436944937833, "loss": 0.1856, "step": 1679 }, { "epoch": 1.1183225162256616, "grad_norm": 42.51642990112305, "learning_rate": 0.0001257992895204263, "loss": 0.0863, "step": 1680 }, { "epoch": 1.1189881843900815, "grad_norm": 52.85578155517578, "learning_rate": 0.00012575488454706927, "loss": 0.27, "step": 1681 }, { "epoch": 1.1196538525545017, "grad_norm": 35.844017028808594, "learning_rate": 0.00012571047957371227, "loss": 0.0727, "step": 1682 }, { "epoch": 1.1203195207189216, "grad_norm": 39.846595764160156, "learning_rate": 0.00012566607460035524, "loss": 0.0679, "step": 1683 }, { "epoch": 1.1209851888833418, "grad_norm": 32.856929779052734, "learning_rate": 0.00012562166962699824, "loss": 0.1318, "step": 1684 }, { "epoch": 1.1216508570477617, "grad_norm": 65.73534393310547, "learning_rate": 0.0001255772646536412, "loss": 0.592, "step": 1685 }, { "epoch": 1.1223165252121818, "grad_norm": 28.59135627746582, "learning_rate": 0.00012553285968028418, "loss": 0.085, "step": 1686 }, { "epoch": 1.1229821933766018, "grad_norm": 24.35097885131836, "learning_rate": 0.00012548845470692718, "loss": 0.0619, "step": 1687 }, { "epoch": 1.1236478615410217, "grad_norm": 9.174190521240234, "learning_rate": 0.00012544404973357015, "loss": 0.0299, "step": 1688 }, { "epoch": 1.1243135297054418, "grad_norm": 22.660593032836914, "learning_rate": 0.00012539964476021315, "loss": 0.0447, "step": 1689 }, { "epoch": 1.1249791978698618, "grad_norm": 136.49009704589844, "learning_rate": 0.00012535523978685612, "loss": 0.9401, "step": 1690 }, { "epoch": 1.125644866034282, "grad_norm": 82.46659088134766, "learning_rate": 0.00012531083481349912, "loss": 0.5236, "step": 1691 }, { "epoch": 1.1263105341987019, "grad_norm": 62.201454162597656, "learning_rate": 0.0001252664298401421, "loss": 0.5939, "step": 1692 }, { "epoch": 1.126976202363122, "grad_norm": 59.48065948486328, "learning_rate": 0.0001252220248667851, "loss": 0.0448, "step": 1693 }, { "epoch": 1.127641870527542, "grad_norm": 36.699031829833984, "learning_rate": 0.00012517761989342806, "loss": 0.4065, "step": 1694 }, { "epoch": 1.128307538691962, "grad_norm": 5.689311504364014, "learning_rate": 0.00012513321492007106, "loss": 0.0186, "step": 1695 }, { "epoch": 1.128973206856382, "grad_norm": 97.11627197265625, "learning_rate": 0.00012508880994671403, "loss": 0.121, "step": 1696 }, { "epoch": 1.1296388750208022, "grad_norm": 46.57074737548828, "learning_rate": 0.00012504440497335703, "loss": 0.2558, "step": 1697 }, { "epoch": 1.130304543185222, "grad_norm": 76.75553894042969, "learning_rate": 0.000125, "loss": 0.3536, "step": 1698 }, { "epoch": 1.1309702113496423, "grad_norm": 52.81585693359375, "learning_rate": 0.00012495559502664297, "loss": 0.5746, "step": 1699 }, { "epoch": 1.1316358795140622, "grad_norm": 47.3369140625, "learning_rate": 0.00012491119005328597, "loss": 0.2703, "step": 1700 }, { "epoch": 1.1323015476784823, "grad_norm": 55.852325439453125, "learning_rate": 0.00012486678507992894, "loss": 0.4363, "step": 1701 }, { "epoch": 1.1329672158429023, "grad_norm": 40.72039794921875, "learning_rate": 0.00012482238010657194, "loss": 0.1875, "step": 1702 }, { "epoch": 1.1336328840073224, "grad_norm": 53.9277458190918, "learning_rate": 0.0001247779751332149, "loss": 0.4369, "step": 1703 }, { "epoch": 1.1342985521717424, "grad_norm": 18.575359344482422, "learning_rate": 0.0001247335701598579, "loss": 0.0685, "step": 1704 }, { "epoch": 1.1349642203361625, "grad_norm": 40.57480239868164, "learning_rate": 0.00012468916518650088, "loss": 0.178, "step": 1705 }, { "epoch": 1.1356298885005824, "grad_norm": 3.1595706939697266, "learning_rate": 0.00012464476021314388, "loss": 0.0096, "step": 1706 }, { "epoch": 1.1362955566650026, "grad_norm": 28.67967414855957, "learning_rate": 0.00012460035523978685, "loss": 0.1, "step": 1707 }, { "epoch": 1.1369612248294225, "grad_norm": 11.445876121520996, "learning_rate": 0.00012455595026642985, "loss": 0.0316, "step": 1708 }, { "epoch": 1.1376268929938425, "grad_norm": 9.125166893005371, "learning_rate": 0.00012451154529307282, "loss": 0.0305, "step": 1709 }, { "epoch": 1.1382925611582626, "grad_norm": 70.58088684082031, "learning_rate": 0.00012446714031971582, "loss": 0.4383, "step": 1710 }, { "epoch": 1.1389582293226828, "grad_norm": 112.8126449584961, "learning_rate": 0.0001244227353463588, "loss": 0.6135, "step": 1711 }, { "epoch": 1.1396238974871027, "grad_norm": 76.1169662475586, "learning_rate": 0.00012437833037300176, "loss": 0.4179, "step": 1712 }, { "epoch": 1.1402895656515226, "grad_norm": 55.31990051269531, "learning_rate": 0.00012433392539964476, "loss": 0.2204, "step": 1713 }, { "epoch": 1.1409552338159428, "grad_norm": 35.56497573852539, "learning_rate": 0.00012428952042628773, "loss": 0.2304, "step": 1714 }, { "epoch": 1.1416209019803627, "grad_norm": 41.59407043457031, "learning_rate": 0.00012424511545293073, "loss": 0.2448, "step": 1715 }, { "epoch": 1.1422865701447829, "grad_norm": 82.35811614990234, "learning_rate": 0.0001242007104795737, "loss": 0.4154, "step": 1716 }, { "epoch": 1.1429522383092028, "grad_norm": 0.9948437213897705, "learning_rate": 0.0001241563055062167, "loss": 0.0034, "step": 1717 }, { "epoch": 1.143617906473623, "grad_norm": 67.5859146118164, "learning_rate": 0.00012411190053285967, "loss": 0.467, "step": 1718 }, { "epoch": 1.1442835746380429, "grad_norm": 33.16363525390625, "learning_rate": 0.00012406749555950267, "loss": 0.0623, "step": 1719 }, { "epoch": 1.144949242802463, "grad_norm": 103.69190979003906, "learning_rate": 0.00012402309058614564, "loss": 0.2009, "step": 1720 }, { "epoch": 1.145614910966883, "grad_norm": 40.75819778442383, "learning_rate": 0.00012397868561278864, "loss": 0.0813, "step": 1721 }, { "epoch": 1.146280579131303, "grad_norm": 101.44440460205078, "learning_rate": 0.0001239342806394316, "loss": 1.1291, "step": 1722 }, { "epoch": 1.146946247295723, "grad_norm": 18.247499465942383, "learning_rate": 0.0001238898756660746, "loss": 0.0228, "step": 1723 }, { "epoch": 1.1476119154601432, "grad_norm": 105.16305541992188, "learning_rate": 0.00012384547069271758, "loss": 0.3756, "step": 1724 }, { "epoch": 1.1482775836245631, "grad_norm": 63.42177200317383, "learning_rate": 0.00012380106571936055, "loss": 0.6663, "step": 1725 }, { "epoch": 1.1489432517889833, "grad_norm": 23.25391387939453, "learning_rate": 0.00012375666074600355, "loss": 0.0291, "step": 1726 }, { "epoch": 1.1496089199534032, "grad_norm": 112.8501205444336, "learning_rate": 0.00012371225577264652, "loss": 0.729, "step": 1727 }, { "epoch": 1.1502745881178233, "grad_norm": 119.95511627197266, "learning_rate": 0.00012366785079928952, "loss": 0.3577, "step": 1728 }, { "epoch": 1.1509402562822433, "grad_norm": 55.90987777709961, "learning_rate": 0.0001236234458259325, "loss": 0.4891, "step": 1729 }, { "epoch": 1.1516059244466634, "grad_norm": 67.94099426269531, "learning_rate": 0.0001235790408525755, "loss": 0.5482, "step": 1730 }, { "epoch": 1.1522715926110834, "grad_norm": 90.74688720703125, "learning_rate": 0.00012353463587921846, "loss": 0.4853, "step": 1731 }, { "epoch": 1.1529372607755035, "grad_norm": 102.6233901977539, "learning_rate": 0.00012349023090586146, "loss": 1.0199, "step": 1732 }, { "epoch": 1.1536029289399234, "grad_norm": 39.753047943115234, "learning_rate": 0.00012344582593250443, "loss": 0.4519, "step": 1733 }, { "epoch": 1.1542685971043434, "grad_norm": 54.624515533447266, "learning_rate": 0.00012340142095914743, "loss": 0.1897, "step": 1734 }, { "epoch": 1.1549342652687635, "grad_norm": 49.87479782104492, "learning_rate": 0.0001233570159857904, "loss": 0.2632, "step": 1735 }, { "epoch": 1.1555999334331837, "grad_norm": 58.85496520996094, "learning_rate": 0.0001233126110124334, "loss": 0.8345, "step": 1736 }, { "epoch": 1.1562656015976036, "grad_norm": 69.97720336914062, "learning_rate": 0.00012326820603907637, "loss": 0.4291, "step": 1737 }, { "epoch": 1.1569312697620235, "grad_norm": 25.539186477661133, "learning_rate": 0.00012322380106571937, "loss": 0.066, "step": 1738 }, { "epoch": 1.1575969379264437, "grad_norm": 18.36098289489746, "learning_rate": 0.00012317939609236234, "loss": 0.0764, "step": 1739 }, { "epoch": 1.1582626060908636, "grad_norm": 28.83622169494629, "learning_rate": 0.00012313499111900534, "loss": 0.2368, "step": 1740 }, { "epoch": 1.1589282742552838, "grad_norm": 25.079330444335938, "learning_rate": 0.0001230905861456483, "loss": 0.125, "step": 1741 }, { "epoch": 1.1595939424197037, "grad_norm": 32.652862548828125, "learning_rate": 0.0001230461811722913, "loss": 0.1689, "step": 1742 }, { "epoch": 1.1602596105841239, "grad_norm": 15.260133743286133, "learning_rate": 0.0001230017761989343, "loss": 0.1008, "step": 1743 }, { "epoch": 1.1609252787485438, "grad_norm": 73.29993438720703, "learning_rate": 0.00012295737122557728, "loss": 0.3519, "step": 1744 }, { "epoch": 1.161590946912964, "grad_norm": 45.03847122192383, "learning_rate": 0.00012291296625222028, "loss": 0.2374, "step": 1745 }, { "epoch": 1.1622566150773839, "grad_norm": 69.38485717773438, "learning_rate": 0.00012286856127886325, "loss": 0.2813, "step": 1746 }, { "epoch": 1.162922283241804, "grad_norm": 26.108379364013672, "learning_rate": 0.00012282415630550625, "loss": 0.0964, "step": 1747 }, { "epoch": 1.163587951406224, "grad_norm": 41.81491470336914, "learning_rate": 0.00012277975133214922, "loss": 0.1642, "step": 1748 }, { "epoch": 1.164253619570644, "grad_norm": 30.141956329345703, "learning_rate": 0.0001227353463587922, "loss": 0.0888, "step": 1749 }, { "epoch": 1.164919287735064, "grad_norm": 20.94239044189453, "learning_rate": 0.0001226909413854352, "loss": 0.0645, "step": 1750 }, { "epoch": 1.1655849558994842, "grad_norm": 28.80684471130371, "learning_rate": 0.00012264653641207816, "loss": 0.0945, "step": 1751 }, { "epoch": 1.1662506240639041, "grad_norm": 44.381229400634766, "learning_rate": 0.00012260213143872116, "loss": 0.1462, "step": 1752 }, { "epoch": 1.1669162922283243, "grad_norm": 14.057089805603027, "learning_rate": 0.00012255772646536413, "loss": 0.0502, "step": 1753 }, { "epoch": 1.1675819603927442, "grad_norm": 28.116914749145508, "learning_rate": 0.00012251332149200713, "loss": 0.0961, "step": 1754 }, { "epoch": 1.1682476285571641, "grad_norm": 48.43739700317383, "learning_rate": 0.0001224689165186501, "loss": 0.2717, "step": 1755 }, { "epoch": 1.1689132967215843, "grad_norm": 62.79151916503906, "learning_rate": 0.0001224245115452931, "loss": 0.5599, "step": 1756 }, { "epoch": 1.1695789648860044, "grad_norm": 24.24590492248535, "learning_rate": 0.00012238010657193607, "loss": 0.1486, "step": 1757 }, { "epoch": 1.1702446330504244, "grad_norm": 52.96484375, "learning_rate": 0.00012233570159857907, "loss": 0.2642, "step": 1758 }, { "epoch": 1.1709103012148443, "grad_norm": 63.703285217285156, "learning_rate": 0.00012229129662522204, "loss": 0.3584, "step": 1759 }, { "epoch": 1.1715759693792644, "grad_norm": 54.90060806274414, "learning_rate": 0.00012224689165186504, "loss": 0.2484, "step": 1760 }, { "epoch": 1.1722416375436844, "grad_norm": 53.68577575683594, "learning_rate": 0.000122202486678508, "loss": 0.2202, "step": 1761 }, { "epoch": 1.1729073057081045, "grad_norm": 101.63054656982422, "learning_rate": 0.00012215808170515098, "loss": 0.1619, "step": 1762 }, { "epoch": 1.1735729738725245, "grad_norm": 106.55372619628906, "learning_rate": 0.00012211367673179398, "loss": 0.6401, "step": 1763 }, { "epoch": 1.1742386420369446, "grad_norm": 9.360029220581055, "learning_rate": 0.00012206927175843695, "loss": 0.0233, "step": 1764 }, { "epoch": 1.1749043102013645, "grad_norm": 24.123807907104492, "learning_rate": 0.00012202486678507993, "loss": 0.0852, "step": 1765 }, { "epoch": 1.1755699783657847, "grad_norm": 32.831092834472656, "learning_rate": 0.00012198046181172292, "loss": 0.0888, "step": 1766 }, { "epoch": 1.1762356465302046, "grad_norm": 3.4706904888153076, "learning_rate": 0.0001219360568383659, "loss": 0.0069, "step": 1767 }, { "epoch": 1.1769013146946248, "grad_norm": 58.15880584716797, "learning_rate": 0.00012189165186500889, "loss": 0.2271, "step": 1768 }, { "epoch": 1.1775669828590447, "grad_norm": 64.16773986816406, "learning_rate": 0.00012184724689165187, "loss": 0.1847, "step": 1769 }, { "epoch": 1.1782326510234649, "grad_norm": 39.20933532714844, "learning_rate": 0.00012180284191829486, "loss": 0.1499, "step": 1770 }, { "epoch": 1.1788983191878848, "grad_norm": 61.72149658203125, "learning_rate": 0.00012175843694493784, "loss": 0.1783, "step": 1771 }, { "epoch": 1.179563987352305, "grad_norm": 7.459522247314453, "learning_rate": 0.00012171403197158083, "loss": 0.01, "step": 1772 }, { "epoch": 1.1802296555167249, "grad_norm": 58.44734573364258, "learning_rate": 0.00012166962699822382, "loss": 0.1375, "step": 1773 }, { "epoch": 1.180895323681145, "grad_norm": 73.8143081665039, "learning_rate": 0.0001216252220248668, "loss": 0.5342, "step": 1774 }, { "epoch": 1.181560991845565, "grad_norm": 124.09207916259766, "learning_rate": 0.00012158081705150977, "loss": 1.5688, "step": 1775 }, { "epoch": 1.182226660009985, "grad_norm": 68.54161071777344, "learning_rate": 0.00012153641207815277, "loss": 0.7073, "step": 1776 }, { "epoch": 1.182892328174405, "grad_norm": 96.21676635742188, "learning_rate": 0.00012149200710479574, "loss": 0.6329, "step": 1777 }, { "epoch": 1.1835579963388252, "grad_norm": 7.976950645446777, "learning_rate": 0.00012144760213143874, "loss": 0.0155, "step": 1778 }, { "epoch": 1.1842236645032451, "grad_norm": 40.75532913208008, "learning_rate": 0.0001214031971580817, "loss": 0.0841, "step": 1779 }, { "epoch": 1.184889332667665, "grad_norm": 81.84866333007812, "learning_rate": 0.0001213587921847247, "loss": 0.2816, "step": 1780 }, { "epoch": 1.1855550008320852, "grad_norm": 58.48090362548828, "learning_rate": 0.00012131438721136768, "loss": 0.2526, "step": 1781 }, { "epoch": 1.1862206689965054, "grad_norm": 50.94480514526367, "learning_rate": 0.00012126998223801067, "loss": 0.476, "step": 1782 }, { "epoch": 1.1868863371609253, "grad_norm": 83.68280792236328, "learning_rate": 0.00012122557726465365, "loss": 0.7236, "step": 1783 }, { "epoch": 1.1875520053253452, "grad_norm": 34.66283416748047, "learning_rate": 0.00012118117229129664, "loss": 0.2787, "step": 1784 }, { "epoch": 1.1882176734897654, "grad_norm": 97.59159088134766, "learning_rate": 0.00012113676731793962, "loss": 0.5378, "step": 1785 }, { "epoch": 1.1888833416541853, "grad_norm": 125.527587890625, "learning_rate": 0.00012109236234458259, "loss": 0.672, "step": 1786 }, { "epoch": 1.1895490098186055, "grad_norm": 88.32150268554688, "learning_rate": 0.00012104795737122559, "loss": 0.7129, "step": 1787 }, { "epoch": 1.1902146779830254, "grad_norm": 9.79939079284668, "learning_rate": 0.00012100355239786856, "loss": 0.0262, "step": 1788 }, { "epoch": 1.1908803461474455, "grad_norm": 27.288618087768555, "learning_rate": 0.00012095914742451155, "loss": 0.0483, "step": 1789 }, { "epoch": 1.1915460143118655, "grad_norm": 40.166934967041016, "learning_rate": 0.00012091474245115453, "loss": 0.1743, "step": 1790 }, { "epoch": 1.1922116824762856, "grad_norm": 19.970264434814453, "learning_rate": 0.00012087033747779752, "loss": 0.0499, "step": 1791 }, { "epoch": 1.1928773506407055, "grad_norm": 46.07417297363281, "learning_rate": 0.0001208259325044405, "loss": 0.2944, "step": 1792 }, { "epoch": 1.1935430188051257, "grad_norm": 33.00217056274414, "learning_rate": 0.0001207815275310835, "loss": 0.1559, "step": 1793 }, { "epoch": 1.1942086869695456, "grad_norm": 40.64289855957031, "learning_rate": 0.00012073712255772647, "loss": 0.1742, "step": 1794 }, { "epoch": 1.1948743551339658, "grad_norm": 79.71502685546875, "learning_rate": 0.00012069271758436946, "loss": 0.2544, "step": 1795 }, { "epoch": 1.1955400232983857, "grad_norm": 31.091110229492188, "learning_rate": 0.00012064831261101244, "loss": 0.069, "step": 1796 }, { "epoch": 1.1962056914628059, "grad_norm": 48.19164276123047, "learning_rate": 0.00012060390763765543, "loss": 0.2599, "step": 1797 }, { "epoch": 1.1968713596272258, "grad_norm": 70.30054473876953, "learning_rate": 0.0001205595026642984, "loss": 0.254, "step": 1798 }, { "epoch": 1.197537027791646, "grad_norm": 61.600242614746094, "learning_rate": 0.00012051509769094138, "loss": 0.4289, "step": 1799 }, { "epoch": 1.1982026959560659, "grad_norm": 84.9737777709961, "learning_rate": 0.00012047069271758437, "loss": 0.5043, "step": 1800 }, { "epoch": 1.198868364120486, "grad_norm": 27.392847061157227, "learning_rate": 0.00012042628774422735, "loss": 0.1486, "step": 1801 }, { "epoch": 1.199534032284906, "grad_norm": 24.098041534423828, "learning_rate": 0.00012038188277087034, "loss": 0.1865, "step": 1802 }, { "epoch": 1.2001997004493261, "grad_norm": 51.285091400146484, "learning_rate": 0.00012033747779751332, "loss": 0.178, "step": 1803 }, { "epoch": 1.200865368613746, "grad_norm": 35.74811553955078, "learning_rate": 0.00012029307282415631, "loss": 0.1275, "step": 1804 }, { "epoch": 1.201531036778166, "grad_norm": 16.749380111694336, "learning_rate": 0.00012024866785079929, "loss": 0.0497, "step": 1805 }, { "epoch": 1.2021967049425861, "grad_norm": 48.10573959350586, "learning_rate": 0.00012020426287744228, "loss": 0.1785, "step": 1806 }, { "epoch": 1.2028623731070063, "grad_norm": 38.705814361572266, "learning_rate": 0.00012015985790408525, "loss": 0.4094, "step": 1807 }, { "epoch": 1.2035280412714262, "grad_norm": 66.93748474121094, "learning_rate": 0.00012011545293072825, "loss": 0.6624, "step": 1808 }, { "epoch": 1.2041937094358461, "grad_norm": 61.66252136230469, "learning_rate": 0.00012007104795737122, "loss": 0.1986, "step": 1809 }, { "epoch": 1.2048593776002663, "grad_norm": 50.47108840942383, "learning_rate": 0.00012002664298401422, "loss": 0.1143, "step": 1810 }, { "epoch": 1.2055250457646862, "grad_norm": 11.109007835388184, "learning_rate": 0.0001199822380106572, "loss": 0.0278, "step": 1811 }, { "epoch": 1.2061907139291064, "grad_norm": 10.17584228515625, "learning_rate": 0.00011993783303730018, "loss": 0.0498, "step": 1812 }, { "epoch": 1.2068563820935263, "grad_norm": 47.093528747558594, "learning_rate": 0.00011989342806394316, "loss": 0.2645, "step": 1813 }, { "epoch": 1.2075220502579465, "grad_norm": 29.003870010375977, "learning_rate": 0.00011984902309058615, "loss": 0.1468, "step": 1814 }, { "epoch": 1.2081877184223664, "grad_norm": 40.20622634887695, "learning_rate": 0.00011980461811722913, "loss": 0.1275, "step": 1815 }, { "epoch": 1.2088533865867865, "grad_norm": 39.93096923828125, "learning_rate": 0.00011976021314387212, "loss": 0.1079, "step": 1816 }, { "epoch": 1.2095190547512065, "grad_norm": 30.0715389251709, "learning_rate": 0.0001197158081705151, "loss": 0.0579, "step": 1817 }, { "epoch": 1.2101847229156266, "grad_norm": 28.30344009399414, "learning_rate": 0.00011967140319715809, "loss": 0.1572, "step": 1818 }, { "epoch": 1.2108503910800466, "grad_norm": 53.62153244018555, "learning_rate": 0.00011962699822380107, "loss": 0.4916, "step": 1819 }, { "epoch": 1.2115160592444667, "grad_norm": 46.005706787109375, "learning_rate": 0.00011958259325044406, "loss": 0.1317, "step": 1820 }, { "epoch": 1.2121817274088866, "grad_norm": 35.67490768432617, "learning_rate": 0.00011953818827708704, "loss": 0.1973, "step": 1821 }, { "epoch": 1.2128473955733068, "grad_norm": 110.17412567138672, "learning_rate": 0.00011949378330373003, "loss": 0.4995, "step": 1822 }, { "epoch": 1.2135130637377267, "grad_norm": 28.53251838684082, "learning_rate": 0.00011944937833037301, "loss": 0.1427, "step": 1823 }, { "epoch": 1.2141787319021469, "grad_norm": 25.906639099121094, "learning_rate": 0.000119404973357016, "loss": 0.0721, "step": 1824 }, { "epoch": 1.2148444000665668, "grad_norm": 26.571443557739258, "learning_rate": 0.00011936056838365897, "loss": 0.0764, "step": 1825 }, { "epoch": 1.2155100682309867, "grad_norm": 48.38930892944336, "learning_rate": 0.00011931616341030197, "loss": 0.4866, "step": 1826 }, { "epoch": 1.2161757363954069, "grad_norm": 43.499629974365234, "learning_rate": 0.00011927175843694494, "loss": 0.1919, "step": 1827 }, { "epoch": 1.216841404559827, "grad_norm": 22.993633270263672, "learning_rate": 0.00011922735346358794, "loss": 0.0617, "step": 1828 }, { "epoch": 1.217507072724247, "grad_norm": 57.592830657958984, "learning_rate": 0.00011918294849023091, "loss": 0.2119, "step": 1829 }, { "epoch": 1.218172740888667, "grad_norm": 28.978059768676758, "learning_rate": 0.0001191385435168739, "loss": 0.0949, "step": 1830 }, { "epoch": 1.218838409053087, "grad_norm": 54.02754592895508, "learning_rate": 0.00011909413854351688, "loss": 0.5451, "step": 1831 }, { "epoch": 1.219504077217507, "grad_norm": 35.086551666259766, "learning_rate": 0.00011904973357015988, "loss": 0.2771, "step": 1832 }, { "epoch": 1.2201697453819271, "grad_norm": 63.2122802734375, "learning_rate": 0.00011900532859680285, "loss": 0.1698, "step": 1833 }, { "epoch": 1.220835413546347, "grad_norm": 54.811553955078125, "learning_rate": 0.00011896092362344585, "loss": 0.3713, "step": 1834 }, { "epoch": 1.2215010817107672, "grad_norm": 78.5242691040039, "learning_rate": 0.00011891651865008882, "loss": 0.292, "step": 1835 }, { "epoch": 1.2221667498751871, "grad_norm": 21.47328758239746, "learning_rate": 0.00011887211367673179, "loss": 0.0674, "step": 1836 }, { "epoch": 1.2228324180396073, "grad_norm": 48.47159194946289, "learning_rate": 0.00011882770870337479, "loss": 0.2254, "step": 1837 }, { "epoch": 1.2234980862040272, "grad_norm": 80.96033477783203, "learning_rate": 0.00011878330373001776, "loss": 0.3805, "step": 1838 }, { "epoch": 1.2241637543684474, "grad_norm": 40.73381042480469, "learning_rate": 0.00011873889875666076, "loss": 0.315, "step": 1839 }, { "epoch": 1.2248294225328673, "grad_norm": 18.109880447387695, "learning_rate": 0.00011869449378330373, "loss": 0.0829, "step": 1840 }, { "epoch": 1.2254950906972875, "grad_norm": 13.151747703552246, "learning_rate": 0.00011865008880994673, "loss": 0.0734, "step": 1841 }, { "epoch": 1.2261607588617074, "grad_norm": 52.29880905151367, "learning_rate": 0.0001186056838365897, "loss": 0.3132, "step": 1842 }, { "epoch": 1.2268264270261275, "grad_norm": 28.357728958129883, "learning_rate": 0.0001185612788632327, "loss": 0.1971, "step": 1843 }, { "epoch": 1.2274920951905475, "grad_norm": 53.71747970581055, "learning_rate": 0.00011851687388987567, "loss": 0.2028, "step": 1844 }, { "epoch": 1.2281577633549676, "grad_norm": 73.06195068359375, "learning_rate": 0.00011847246891651867, "loss": 0.2843, "step": 1845 }, { "epoch": 1.2288234315193876, "grad_norm": 39.00293731689453, "learning_rate": 0.00011842806394316164, "loss": 0.1411, "step": 1846 }, { "epoch": 1.2294890996838077, "grad_norm": 22.640640258789062, "learning_rate": 0.00011838365896980463, "loss": 0.1211, "step": 1847 }, { "epoch": 1.2301547678482276, "grad_norm": 66.06171417236328, "learning_rate": 0.0001183392539964476, "loss": 0.4353, "step": 1848 }, { "epoch": 1.2308204360126478, "grad_norm": 29.97242546081543, "learning_rate": 0.00011829484902309058, "loss": 0.1306, "step": 1849 }, { "epoch": 1.2314861041770677, "grad_norm": 54.64789581298828, "learning_rate": 0.00011825044404973358, "loss": 0.4204, "step": 1850 }, { "epoch": 1.2321517723414876, "grad_norm": 12.896210670471191, "learning_rate": 0.00011820603907637655, "loss": 0.0496, "step": 1851 }, { "epoch": 1.2328174405059078, "grad_norm": 36.92301559448242, "learning_rate": 0.00011816163410301955, "loss": 0.0754, "step": 1852 }, { "epoch": 1.233483108670328, "grad_norm": 42.50762176513672, "learning_rate": 0.00011811722912966252, "loss": 0.2271, "step": 1853 }, { "epoch": 1.2341487768347479, "grad_norm": 61.36509323120117, "learning_rate": 0.00011807282415630552, "loss": 0.4272, "step": 1854 }, { "epoch": 1.2348144449991678, "grad_norm": 128.75918579101562, "learning_rate": 0.00011802841918294849, "loss": 0.2962, "step": 1855 }, { "epoch": 1.235480113163588, "grad_norm": 23.29256820678711, "learning_rate": 0.00011798401420959148, "loss": 0.1236, "step": 1856 }, { "epoch": 1.236145781328008, "grad_norm": 66.76676177978516, "learning_rate": 0.00011793960923623446, "loss": 0.2225, "step": 1857 }, { "epoch": 1.236811449492428, "grad_norm": 20.17377281188965, "learning_rate": 0.00011789520426287745, "loss": 0.0514, "step": 1858 }, { "epoch": 1.237477117656848, "grad_norm": 66.10411834716797, "learning_rate": 0.00011785079928952043, "loss": 0.3403, "step": 1859 }, { "epoch": 1.2381427858212681, "grad_norm": 9.343244552612305, "learning_rate": 0.00011780639431616342, "loss": 0.0187, "step": 1860 }, { "epoch": 1.238808453985688, "grad_norm": 6.085109233856201, "learning_rate": 0.0001177619893428064, "loss": 0.0134, "step": 1861 }, { "epoch": 1.2394741221501082, "grad_norm": 70.27294921875, "learning_rate": 0.00011771758436944938, "loss": 0.3458, "step": 1862 }, { "epoch": 1.2401397903145281, "grad_norm": 28.280113220214844, "learning_rate": 0.00011767317939609237, "loss": 0.0948, "step": 1863 }, { "epoch": 1.2408054584789483, "grad_norm": 31.884614944458008, "learning_rate": 0.00011762877442273535, "loss": 0.5291, "step": 1864 }, { "epoch": 1.2414711266433682, "grad_norm": 42.773033142089844, "learning_rate": 0.00011758436944937833, "loss": 0.5482, "step": 1865 }, { "epoch": 1.2421367948077884, "grad_norm": 38.158565521240234, "learning_rate": 0.00011753996447602132, "loss": 0.1283, "step": 1866 }, { "epoch": 1.2428024629722083, "grad_norm": 28.548717498779297, "learning_rate": 0.0001174955595026643, "loss": 0.0577, "step": 1867 }, { "epoch": 1.2434681311366285, "grad_norm": 78.70352935791016, "learning_rate": 0.00011745115452930729, "loss": 0.0974, "step": 1868 }, { "epoch": 1.2441337993010484, "grad_norm": 53.035743713378906, "learning_rate": 0.00011740674955595027, "loss": 0.3927, "step": 1869 }, { "epoch": 1.2447994674654685, "grad_norm": 45.45488739013672, "learning_rate": 0.00011736234458259326, "loss": 0.1683, "step": 1870 }, { "epoch": 1.2454651356298885, "grad_norm": 83.3730697631836, "learning_rate": 0.00011731793960923624, "loss": 0.2929, "step": 1871 }, { "epoch": 1.2461308037943086, "grad_norm": 72.66669464111328, "learning_rate": 0.00011727353463587923, "loss": 0.6639, "step": 1872 }, { "epoch": 1.2467964719587286, "grad_norm": 35.330299377441406, "learning_rate": 0.00011722912966252221, "loss": 0.163, "step": 1873 }, { "epoch": 1.2474621401231487, "grad_norm": 46.40792465209961, "learning_rate": 0.0001171847246891652, "loss": 0.2716, "step": 1874 }, { "epoch": 1.2481278082875686, "grad_norm": 73.30651092529297, "learning_rate": 0.00011714031971580817, "loss": 0.3321, "step": 1875 }, { "epoch": 1.2487934764519886, "grad_norm": 51.8281135559082, "learning_rate": 0.00011709591474245117, "loss": 0.0683, "step": 1876 }, { "epoch": 1.2494591446164087, "grad_norm": 21.537567138671875, "learning_rate": 0.00011705150976909414, "loss": 0.0516, "step": 1877 }, { "epoch": 1.2501248127808289, "grad_norm": 31.505767822265625, "learning_rate": 0.00011700710479573714, "loss": 0.1234, "step": 1878 }, { "epoch": 1.2507904809452488, "grad_norm": 3.491621732711792, "learning_rate": 0.00011696269982238011, "loss": 0.0076, "step": 1879 }, { "epoch": 1.2514561491096687, "grad_norm": 22.357328414916992, "learning_rate": 0.00011691829484902311, "loss": 0.1106, "step": 1880 }, { "epoch": 1.2521218172740889, "grad_norm": 87.8445816040039, "learning_rate": 0.00011687388987566608, "loss": 0.8877, "step": 1881 }, { "epoch": 1.252787485438509, "grad_norm": 37.13893127441406, "learning_rate": 0.00011682948490230908, "loss": 0.3848, "step": 1882 }, { "epoch": 1.253453153602929, "grad_norm": 71.55647277832031, "learning_rate": 0.00011678507992895205, "loss": 0.5046, "step": 1883 }, { "epoch": 1.254118821767349, "grad_norm": 9.508859634399414, "learning_rate": 0.00011674067495559505, "loss": 0.0205, "step": 1884 }, { "epoch": 1.254784489931769, "grad_norm": 63.116641998291016, "learning_rate": 0.00011669626998223802, "loss": 0.2742, "step": 1885 }, { "epoch": 1.255450158096189, "grad_norm": 67.19985961914062, "learning_rate": 0.00011665186500888099, "loss": 0.2853, "step": 1886 }, { "epoch": 1.2561158262606091, "grad_norm": 38.56999969482422, "learning_rate": 0.00011660746003552399, "loss": 0.1773, "step": 1887 }, { "epoch": 1.256781494425029, "grad_norm": 59.92221450805664, "learning_rate": 0.00011656305506216696, "loss": 0.2415, "step": 1888 }, { "epoch": 1.2574471625894492, "grad_norm": 65.25638580322266, "learning_rate": 0.00011651865008880996, "loss": 0.2284, "step": 1889 }, { "epoch": 1.2581128307538691, "grad_norm": 87.92919921875, "learning_rate": 0.00011647424511545293, "loss": 0.4274, "step": 1890 }, { "epoch": 1.2587784989182893, "grad_norm": 26.441356658935547, "learning_rate": 0.00011642984014209593, "loss": 0.2741, "step": 1891 }, { "epoch": 1.2594441670827092, "grad_norm": 42.815269470214844, "learning_rate": 0.0001163854351687389, "loss": 0.3171, "step": 1892 }, { "epoch": 1.2601098352471294, "grad_norm": 38.57990264892578, "learning_rate": 0.0001163410301953819, "loss": 0.207, "step": 1893 }, { "epoch": 1.2607755034115493, "grad_norm": 43.61090087890625, "learning_rate": 0.00011629662522202487, "loss": 0.2182, "step": 1894 }, { "epoch": 1.2614411715759695, "grad_norm": 68.76473999023438, "learning_rate": 0.00011625222024866787, "loss": 0.2461, "step": 1895 }, { "epoch": 1.2621068397403894, "grad_norm": 82.43309020996094, "learning_rate": 0.00011620781527531084, "loss": 0.7414, "step": 1896 }, { "epoch": 1.2627725079048093, "grad_norm": 9.451969146728516, "learning_rate": 0.00011616341030195384, "loss": 0.0235, "step": 1897 }, { "epoch": 1.2634381760692295, "grad_norm": 26.462921142578125, "learning_rate": 0.00011611900532859681, "loss": 0.0701, "step": 1898 }, { "epoch": 1.2641038442336496, "grad_norm": 35.759002685546875, "learning_rate": 0.00011607460035523978, "loss": 0.0909, "step": 1899 }, { "epoch": 1.2647695123980696, "grad_norm": 73.17784118652344, "learning_rate": 0.00011603019538188278, "loss": 0.7541, "step": 1900 }, { "epoch": 1.2654351805624895, "grad_norm": 10.289645195007324, "learning_rate": 0.00011598579040852575, "loss": 0.0264, "step": 1901 }, { "epoch": 1.2661008487269096, "grad_norm": 19.49089813232422, "learning_rate": 0.00011594138543516875, "loss": 0.0447, "step": 1902 }, { "epoch": 1.2667665168913298, "grad_norm": 111.29071044921875, "learning_rate": 0.00011589698046181172, "loss": 0.8915, "step": 1903 }, { "epoch": 1.2674321850557497, "grad_norm": 5.8084025382995605, "learning_rate": 0.00011585257548845472, "loss": 0.0253, "step": 1904 }, { "epoch": 1.2680978532201697, "grad_norm": 35.514617919921875, "learning_rate": 0.00011580817051509769, "loss": 0.1306, "step": 1905 }, { "epoch": 1.2687635213845898, "grad_norm": 65.45555877685547, "learning_rate": 0.00011576376554174069, "loss": 0.2006, "step": 1906 }, { "epoch": 1.2694291895490097, "grad_norm": 208.85850524902344, "learning_rate": 0.00011571936056838366, "loss": 0.4897, "step": 1907 }, { "epoch": 1.27009485771343, "grad_norm": 47.611881256103516, "learning_rate": 0.00011567495559502666, "loss": 0.2972, "step": 1908 }, { "epoch": 1.2707605258778498, "grad_norm": 41.14829635620117, "learning_rate": 0.00011563055062166963, "loss": 0.3926, "step": 1909 }, { "epoch": 1.27142619404227, "grad_norm": 47.72732162475586, "learning_rate": 0.00011558614564831263, "loss": 0.2769, "step": 1910 }, { "epoch": 1.27209186220669, "grad_norm": 78.26897430419922, "learning_rate": 0.0001155417406749556, "loss": 0.7558, "step": 1911 }, { "epoch": 1.27275753037111, "grad_norm": 87.93765258789062, "learning_rate": 0.00011549733570159857, "loss": 0.2999, "step": 1912 }, { "epoch": 1.27342319853553, "grad_norm": 60.47151184082031, "learning_rate": 0.00011545293072824157, "loss": 0.387, "step": 1913 }, { "epoch": 1.2740888666999501, "grad_norm": 29.86738395690918, "learning_rate": 0.00011540852575488455, "loss": 0.0805, "step": 1914 }, { "epoch": 1.27475453486437, "grad_norm": 73.64708709716797, "learning_rate": 0.00011536412078152754, "loss": 0.5643, "step": 1915 }, { "epoch": 1.2754202030287902, "grad_norm": 36.35663986206055, "learning_rate": 0.00011531971580817052, "loss": 0.0774, "step": 1916 }, { "epoch": 1.2760858711932102, "grad_norm": 50.397621154785156, "learning_rate": 0.0001152753108348135, "loss": 0.3921, "step": 1917 }, { "epoch": 1.27675153935763, "grad_norm": 52.23714065551758, "learning_rate": 0.00011523090586145649, "loss": 0.3007, "step": 1918 }, { "epoch": 1.2774172075220502, "grad_norm": 45.20994186401367, "learning_rate": 0.00011518650088809948, "loss": 0.1925, "step": 1919 }, { "epoch": 1.2780828756864704, "grad_norm": 60.4399528503418, "learning_rate": 0.00011514209591474246, "loss": 0.2833, "step": 1920 }, { "epoch": 1.2787485438508903, "grad_norm": 40.45793151855469, "learning_rate": 0.00011509769094138544, "loss": 0.2307, "step": 1921 }, { "epoch": 1.2794142120153102, "grad_norm": 129.30470275878906, "learning_rate": 0.00011505328596802843, "loss": 0.2982, "step": 1922 }, { "epoch": 1.2800798801797304, "grad_norm": 58.35688781738281, "learning_rate": 0.00011500888099467141, "loss": 0.2976, "step": 1923 }, { "epoch": 1.2807455483441506, "grad_norm": 48.074005126953125, "learning_rate": 0.0001149644760213144, "loss": 0.1858, "step": 1924 }, { "epoch": 1.2814112165085705, "grad_norm": 85.98326110839844, "learning_rate": 0.00011492007104795737, "loss": 0.6106, "step": 1925 }, { "epoch": 1.2820768846729904, "grad_norm": 40.11345291137695, "learning_rate": 0.00011487566607460037, "loss": 0.1455, "step": 1926 }, { "epoch": 1.2827425528374106, "grad_norm": 30.28862762451172, "learning_rate": 0.00011483126110124334, "loss": 0.0769, "step": 1927 }, { "epoch": 1.2834082210018307, "grad_norm": 22.745271682739258, "learning_rate": 0.00011478685612788634, "loss": 0.0745, "step": 1928 }, { "epoch": 1.2840738891662506, "grad_norm": 34.1413688659668, "learning_rate": 0.00011474245115452931, "loss": 0.1079, "step": 1929 }, { "epoch": 1.2847395573306706, "grad_norm": 37.211082458496094, "learning_rate": 0.00011469804618117231, "loss": 0.1064, "step": 1930 }, { "epoch": 1.2854052254950907, "grad_norm": 37.28524398803711, "learning_rate": 0.00011465364120781528, "loss": 0.1949, "step": 1931 }, { "epoch": 1.2860708936595107, "grad_norm": 33.26810836791992, "learning_rate": 0.00011460923623445828, "loss": 0.4688, "step": 1932 }, { "epoch": 1.2867365618239308, "grad_norm": 30.993741989135742, "learning_rate": 0.00011456483126110125, "loss": 0.233, "step": 1933 }, { "epoch": 1.2874022299883507, "grad_norm": 77.94334411621094, "learning_rate": 0.00011452042628774425, "loss": 0.438, "step": 1934 }, { "epoch": 1.288067898152771, "grad_norm": 21.873188018798828, "learning_rate": 0.00011447602131438722, "loss": 0.0736, "step": 1935 }, { "epoch": 1.2887335663171908, "grad_norm": 31.631364822387695, "learning_rate": 0.00011443161634103019, "loss": 0.2632, "step": 1936 }, { "epoch": 1.289399234481611, "grad_norm": 2.678969621658325, "learning_rate": 0.00011438721136767319, "loss": 0.0126, "step": 1937 }, { "epoch": 1.290064902646031, "grad_norm": 40.6650505065918, "learning_rate": 0.00011434280639431616, "loss": 0.15, "step": 1938 }, { "epoch": 1.290730570810451, "grad_norm": 15.994421005249023, "learning_rate": 0.00011429840142095916, "loss": 0.043, "step": 1939 }, { "epoch": 1.291396238974871, "grad_norm": 34.70260238647461, "learning_rate": 0.00011425399644760213, "loss": 0.1108, "step": 1940 }, { "epoch": 1.2920619071392911, "grad_norm": 55.11068344116211, "learning_rate": 0.00011420959147424513, "loss": 0.5279, "step": 1941 }, { "epoch": 1.292727575303711, "grad_norm": 32.15353775024414, "learning_rate": 0.0001141651865008881, "loss": 0.1442, "step": 1942 }, { "epoch": 1.293393243468131, "grad_norm": 53.1467170715332, "learning_rate": 0.0001141207815275311, "loss": 0.1584, "step": 1943 }, { "epoch": 1.2940589116325512, "grad_norm": 54.72723388671875, "learning_rate": 0.00011407637655417407, "loss": 0.5245, "step": 1944 }, { "epoch": 1.2947245797969713, "grad_norm": 41.248130798339844, "learning_rate": 0.00011403197158081707, "loss": 0.1879, "step": 1945 }, { "epoch": 1.2953902479613912, "grad_norm": 48.67994689941406, "learning_rate": 0.00011398756660746004, "loss": 0.212, "step": 1946 }, { "epoch": 1.2960559161258112, "grad_norm": 63.934913635253906, "learning_rate": 0.00011394316163410304, "loss": 0.4503, "step": 1947 }, { "epoch": 1.2967215842902313, "grad_norm": 118.86078643798828, "learning_rate": 0.00011389875666074601, "loss": 1.2752, "step": 1948 }, { "epoch": 1.2973872524546515, "grad_norm": 75.7842788696289, "learning_rate": 0.00011385435168738898, "loss": 0.3613, "step": 1949 }, { "epoch": 1.2980529206190714, "grad_norm": 35.29917526245117, "learning_rate": 0.00011380994671403198, "loss": 0.1588, "step": 1950 }, { "epoch": 1.2987185887834913, "grad_norm": 61.763641357421875, "learning_rate": 0.00011376554174067495, "loss": 0.1867, "step": 1951 }, { "epoch": 1.2993842569479115, "grad_norm": 29.896873474121094, "learning_rate": 0.00011372113676731795, "loss": 0.1215, "step": 1952 }, { "epoch": 1.3000499251123316, "grad_norm": 30.55638885498047, "learning_rate": 0.00011367673179396092, "loss": 0.0979, "step": 1953 }, { "epoch": 1.3007155932767516, "grad_norm": 69.40913391113281, "learning_rate": 0.00011363232682060392, "loss": 0.4368, "step": 1954 }, { "epoch": 1.3013812614411715, "grad_norm": 50.46235656738281, "learning_rate": 0.00011358792184724689, "loss": 0.1402, "step": 1955 }, { "epoch": 1.3020469296055917, "grad_norm": 66.19755554199219, "learning_rate": 0.00011354351687388989, "loss": 0.2939, "step": 1956 }, { "epoch": 1.3027125977700116, "grad_norm": 39.39725875854492, "learning_rate": 0.00011349911190053286, "loss": 0.3057, "step": 1957 }, { "epoch": 1.3033782659344317, "grad_norm": 55.25946807861328, "learning_rate": 0.00011345470692717586, "loss": 1.1272, "step": 1958 }, { "epoch": 1.3040439340988517, "grad_norm": 20.182119369506836, "learning_rate": 0.00011341030195381883, "loss": 0.0773, "step": 1959 }, { "epoch": 1.3047096022632718, "grad_norm": 47.701995849609375, "learning_rate": 0.00011336589698046183, "loss": 0.3354, "step": 1960 }, { "epoch": 1.3053752704276917, "grad_norm": 76.01998138427734, "learning_rate": 0.0001133214920071048, "loss": 0.4734, "step": 1961 }, { "epoch": 1.306040938592112, "grad_norm": 45.11589813232422, "learning_rate": 0.00011327708703374777, "loss": 0.1838, "step": 1962 }, { "epoch": 1.3067066067565318, "grad_norm": 29.065757751464844, "learning_rate": 0.00011323268206039077, "loss": 0.3254, "step": 1963 }, { "epoch": 1.307372274920952, "grad_norm": 48.5753288269043, "learning_rate": 0.00011318827708703374, "loss": 0.2727, "step": 1964 }, { "epoch": 1.308037943085372, "grad_norm": 56.62685775756836, "learning_rate": 0.00011314387211367674, "loss": 0.5166, "step": 1965 }, { "epoch": 1.308703611249792, "grad_norm": 47.98997116088867, "learning_rate": 0.00011309946714031971, "loss": 0.1646, "step": 1966 }, { "epoch": 1.309369279414212, "grad_norm": 40.40205383300781, "learning_rate": 0.00011305506216696271, "loss": 0.2773, "step": 1967 }, { "epoch": 1.310034947578632, "grad_norm": 60.773414611816406, "learning_rate": 0.00011301065719360568, "loss": 0.3931, "step": 1968 }, { "epoch": 1.310700615743052, "grad_norm": 59.695716857910156, "learning_rate": 0.00011296625222024868, "loss": 0.2971, "step": 1969 }, { "epoch": 1.3113662839074722, "grad_norm": 22.084003448486328, "learning_rate": 0.00011292184724689165, "loss": 0.1107, "step": 1970 }, { "epoch": 1.3120319520718922, "grad_norm": 51.63294219970703, "learning_rate": 0.00011287744227353465, "loss": 0.3998, "step": 1971 }, { "epoch": 1.312697620236312, "grad_norm": 61.981201171875, "learning_rate": 0.00011283303730017762, "loss": 0.3074, "step": 1972 }, { "epoch": 1.3133632884007322, "grad_norm": 75.01221466064453, "learning_rate": 0.00011278863232682062, "loss": 0.4185, "step": 1973 }, { "epoch": 1.3140289565651524, "grad_norm": 3.917577028274536, "learning_rate": 0.0001127442273534636, "loss": 0.0141, "step": 1974 }, { "epoch": 1.3146946247295723, "grad_norm": 34.57941818237305, "learning_rate": 0.00011269982238010657, "loss": 0.3449, "step": 1975 }, { "epoch": 1.3153602928939923, "grad_norm": 45.81370162963867, "learning_rate": 0.00011265541740674957, "loss": 0.163, "step": 1976 }, { "epoch": 1.3160259610584124, "grad_norm": 99.58523559570312, "learning_rate": 0.00011261101243339254, "loss": 0.5657, "step": 1977 }, { "epoch": 1.3166916292228323, "grad_norm": 41.988433837890625, "learning_rate": 0.00011256660746003554, "loss": 0.3542, "step": 1978 }, { "epoch": 1.3173572973872525, "grad_norm": 64.77676391601562, "learning_rate": 0.00011252220248667851, "loss": 0.5442, "step": 1979 }, { "epoch": 1.3180229655516724, "grad_norm": 51.408626556396484, "learning_rate": 0.00011247779751332151, "loss": 0.2853, "step": 1980 }, { "epoch": 1.3186886337160926, "grad_norm": 65.16521453857422, "learning_rate": 0.00011243339253996448, "loss": 0.4515, "step": 1981 }, { "epoch": 1.3193543018805125, "grad_norm": 25.483165740966797, "learning_rate": 0.00011238898756660748, "loss": 0.066, "step": 1982 }, { "epoch": 1.3200199700449327, "grad_norm": 36.29672622680664, "learning_rate": 0.00011234458259325045, "loss": 0.1438, "step": 1983 }, { "epoch": 1.3206856382093526, "grad_norm": 46.580299377441406, "learning_rate": 0.00011230017761989345, "loss": 0.3144, "step": 1984 }, { "epoch": 1.3213513063737727, "grad_norm": 82.25044250488281, "learning_rate": 0.00011225577264653642, "loss": 0.1195, "step": 1985 }, { "epoch": 1.3220169745381927, "grad_norm": 82.43309020996094, "learning_rate": 0.00011221136767317939, "loss": 0.4054, "step": 1986 }, { "epoch": 1.3226826427026128, "grad_norm": 2.072540521621704, "learning_rate": 0.00011216696269982239, "loss": 0.0078, "step": 1987 }, { "epoch": 1.3233483108670328, "grad_norm": 29.629587173461914, "learning_rate": 0.00011212255772646536, "loss": 0.3983, "step": 1988 }, { "epoch": 1.324013979031453, "grad_norm": 42.29221725463867, "learning_rate": 0.00011207815275310836, "loss": 0.404, "step": 1989 }, { "epoch": 1.3246796471958728, "grad_norm": 31.84395980834961, "learning_rate": 0.00011203374777975133, "loss": 0.2454, "step": 1990 }, { "epoch": 1.325345315360293, "grad_norm": 12.726351737976074, "learning_rate": 0.00011198934280639433, "loss": 0.0653, "step": 1991 }, { "epoch": 1.326010983524713, "grad_norm": 25.238473892211914, "learning_rate": 0.0001119449378330373, "loss": 0.2034, "step": 1992 }, { "epoch": 1.3266766516891328, "grad_norm": 54.18577575683594, "learning_rate": 0.0001119005328596803, "loss": 0.2633, "step": 1993 }, { "epoch": 1.327342319853553, "grad_norm": 23.165210723876953, "learning_rate": 0.00011185612788632327, "loss": 0.1123, "step": 1994 }, { "epoch": 1.3280079880179732, "grad_norm": 33.137298583984375, "learning_rate": 0.00011181172291296627, "loss": 0.5306, "step": 1995 }, { "epoch": 1.328673656182393, "grad_norm": 27.999927520751953, "learning_rate": 0.00011176731793960924, "loss": 0.1222, "step": 1996 }, { "epoch": 1.329339324346813, "grad_norm": 38.28351974487305, "learning_rate": 0.00011172291296625224, "loss": 0.1211, "step": 1997 }, { "epoch": 1.3300049925112332, "grad_norm": 89.51787567138672, "learning_rate": 0.00011167850799289521, "loss": 0.3123, "step": 1998 }, { "epoch": 1.3306706606756533, "grad_norm": 29.251649856567383, "learning_rate": 0.00011163410301953818, "loss": 0.1055, "step": 1999 }, { "epoch": 1.3313363288400732, "grad_norm": 73.37461853027344, "learning_rate": 0.00011158969804618118, "loss": 0.2544, "step": 2000 }, { "epoch": 1.3320019970044932, "grad_norm": 46.40040969848633, "learning_rate": 0.00011154529307282415, "loss": 0.468, "step": 2001 }, { "epoch": 1.3326676651689133, "grad_norm": 5.108377456665039, "learning_rate": 0.00011150088809946715, "loss": 0.0088, "step": 2002 }, { "epoch": 1.3333333333333333, "grad_norm": 35.744300842285156, "learning_rate": 0.00011145648312611012, "loss": 0.4784, "step": 2003 }, { "epoch": 1.3339990014977534, "grad_norm": 29.662899017333984, "learning_rate": 0.00011141207815275312, "loss": 0.2839, "step": 2004 }, { "epoch": 1.3346646696621733, "grad_norm": 41.16428756713867, "learning_rate": 0.00011136767317939609, "loss": 0.1405, "step": 2005 }, { "epoch": 1.3353303378265935, "grad_norm": 9.077014923095703, "learning_rate": 0.00011132326820603909, "loss": 0.0286, "step": 2006 }, { "epoch": 1.3359960059910134, "grad_norm": 63.33696365356445, "learning_rate": 0.00011127886323268206, "loss": 0.3384, "step": 2007 }, { "epoch": 1.3366616741554336, "grad_norm": 70.9099349975586, "learning_rate": 0.00011123445825932506, "loss": 0.3072, "step": 2008 }, { "epoch": 1.3373273423198535, "grad_norm": 41.488643646240234, "learning_rate": 0.00011119005328596803, "loss": 0.1433, "step": 2009 }, { "epoch": 1.3379930104842737, "grad_norm": 37.45521545410156, "learning_rate": 0.00011114564831261103, "loss": 0.1305, "step": 2010 }, { "epoch": 1.3386586786486936, "grad_norm": 55.42148208618164, "learning_rate": 0.000111101243339254, "loss": 0.2472, "step": 2011 }, { "epoch": 1.3393243468131137, "grad_norm": 14.142194747924805, "learning_rate": 0.00011105683836589697, "loss": 0.0214, "step": 2012 }, { "epoch": 1.3399900149775337, "grad_norm": 13.767867088317871, "learning_rate": 0.00011101243339253997, "loss": 0.0378, "step": 2013 }, { "epoch": 1.3406556831419536, "grad_norm": 13.954246520996094, "learning_rate": 0.00011096802841918294, "loss": 0.0411, "step": 2014 }, { "epoch": 1.3413213513063738, "grad_norm": 29.523969650268555, "learning_rate": 0.00011092362344582594, "loss": 0.1297, "step": 2015 }, { "epoch": 1.341987019470794, "grad_norm": 78.81105041503906, "learning_rate": 0.00011087921847246891, "loss": 0.0598, "step": 2016 }, { "epoch": 1.3426526876352138, "grad_norm": 39.23978805541992, "learning_rate": 0.00011083481349911191, "loss": 0.3338, "step": 2017 }, { "epoch": 1.3433183557996338, "grad_norm": 13.666553497314453, "learning_rate": 0.00011079040852575488, "loss": 0.0254, "step": 2018 }, { "epoch": 1.343984023964054, "grad_norm": 1.8999654054641724, "learning_rate": 0.00011074600355239788, "loss": 0.004, "step": 2019 }, { "epoch": 1.344649692128474, "grad_norm": 50.11959457397461, "learning_rate": 0.00011070159857904085, "loss": 0.2075, "step": 2020 }, { "epoch": 1.345315360292894, "grad_norm": 46.97832107543945, "learning_rate": 0.00011065719360568385, "loss": 0.3377, "step": 2021 }, { "epoch": 1.345981028457314, "grad_norm": 44.09254837036133, "learning_rate": 0.00011061278863232682, "loss": 0.3328, "step": 2022 }, { "epoch": 1.346646696621734, "grad_norm": 17.140277862548828, "learning_rate": 0.00011056838365896982, "loss": 0.0248, "step": 2023 }, { "epoch": 1.3473123647861542, "grad_norm": 9.154749870300293, "learning_rate": 0.00011052397868561279, "loss": 0.0263, "step": 2024 }, { "epoch": 1.3479780329505742, "grad_norm": 97.01370239257812, "learning_rate": 0.00011047957371225577, "loss": 0.4412, "step": 2025 }, { "epoch": 1.348643701114994, "grad_norm": 24.45655632019043, "learning_rate": 0.00011043516873889876, "loss": 0.0942, "step": 2026 }, { "epoch": 1.3493093692794143, "grad_norm": 65.18374633789062, "learning_rate": 0.00011039076376554174, "loss": 0.1591, "step": 2027 }, { "epoch": 1.3499750374438342, "grad_norm": 91.31522369384766, "learning_rate": 0.00011034635879218473, "loss": 0.2206, "step": 2028 }, { "epoch": 1.3506407056082543, "grad_norm": 73.86246490478516, "learning_rate": 0.00011030195381882771, "loss": 0.6312, "step": 2029 }, { "epoch": 1.3513063737726743, "grad_norm": 0.25490066409111023, "learning_rate": 0.0001102575488454707, "loss": 0.0006, "step": 2030 }, { "epoch": 1.3519720419370944, "grad_norm": 17.025897979736328, "learning_rate": 0.00011021314387211368, "loss": 0.0411, "step": 2031 }, { "epoch": 1.3526377101015143, "grad_norm": 16.928932189941406, "learning_rate": 0.00011016873889875668, "loss": 0.0422, "step": 2032 }, { "epoch": 1.3533033782659345, "grad_norm": 55.55219650268555, "learning_rate": 0.00011012433392539965, "loss": 0.1983, "step": 2033 }, { "epoch": 1.3539690464303544, "grad_norm": 50.13077926635742, "learning_rate": 0.00011007992895204265, "loss": 0.4616, "step": 2034 }, { "epoch": 1.3546347145947746, "grad_norm": 56.72121047973633, "learning_rate": 0.00011003552397868562, "loss": 0.1784, "step": 2035 }, { "epoch": 1.3553003827591945, "grad_norm": 52.24368667602539, "learning_rate": 0.00010999111900532859, "loss": 0.1451, "step": 2036 }, { "epoch": 1.3559660509236147, "grad_norm": 38.37303924560547, "learning_rate": 0.00010994671403197159, "loss": 0.2802, "step": 2037 }, { "epoch": 1.3566317190880346, "grad_norm": 50.917232513427734, "learning_rate": 0.00010990230905861456, "loss": 0.2735, "step": 2038 }, { "epoch": 1.3572973872524545, "grad_norm": 11.295807838439941, "learning_rate": 0.00010985790408525756, "loss": 0.0168, "step": 2039 }, { "epoch": 1.3579630554168747, "grad_norm": 82.45463562011719, "learning_rate": 0.00010981349911190053, "loss": 0.1225, "step": 2040 }, { "epoch": 1.3586287235812948, "grad_norm": 37.29206848144531, "learning_rate": 0.00010976909413854353, "loss": 0.0563, "step": 2041 }, { "epoch": 1.3592943917457148, "grad_norm": 80.5683364868164, "learning_rate": 0.0001097246891651865, "loss": 0.3028, "step": 2042 }, { "epoch": 1.3599600599101347, "grad_norm": 50.32986068725586, "learning_rate": 0.0001096802841918295, "loss": 0.2918, "step": 2043 }, { "epoch": 1.3606257280745548, "grad_norm": 7.844069957733154, "learning_rate": 0.00010963587921847247, "loss": 0.023, "step": 2044 }, { "epoch": 1.361291396238975, "grad_norm": 81.17059326171875, "learning_rate": 0.00010959147424511547, "loss": 1.47, "step": 2045 }, { "epoch": 1.361957064403395, "grad_norm": 50.851966857910156, "learning_rate": 0.00010954706927175844, "loss": 0.4534, "step": 2046 }, { "epoch": 1.3626227325678149, "grad_norm": 37.565834045410156, "learning_rate": 0.00010950266429840144, "loss": 0.2016, "step": 2047 }, { "epoch": 1.363288400732235, "grad_norm": 76.7689437866211, "learning_rate": 0.00010945825932504441, "loss": 0.6848, "step": 2048 }, { "epoch": 1.363954068896655, "grad_norm": 50.36044692993164, "learning_rate": 0.00010941385435168738, "loss": 0.1705, "step": 2049 }, { "epoch": 1.364619737061075, "grad_norm": 3.56250262260437, "learning_rate": 0.00010936944937833038, "loss": 0.0092, "step": 2050 }, { "epoch": 1.365285405225495, "grad_norm": 8.063261985778809, "learning_rate": 0.00010932504440497335, "loss": 0.0184, "step": 2051 }, { "epoch": 1.3659510733899152, "grad_norm": 62.019710540771484, "learning_rate": 0.00010928063943161635, "loss": 0.3417, "step": 2052 }, { "epoch": 1.366616741554335, "grad_norm": 44.8753547668457, "learning_rate": 0.00010923623445825932, "loss": 0.1469, "step": 2053 }, { "epoch": 1.3672824097187553, "grad_norm": 2.8787999153137207, "learning_rate": 0.00010919182948490232, "loss": 0.007, "step": 2054 }, { "epoch": 1.3679480778831752, "grad_norm": 36.50013732910156, "learning_rate": 0.00010914742451154529, "loss": 0.1281, "step": 2055 }, { "epoch": 1.3686137460475953, "grad_norm": 47.467002868652344, "learning_rate": 0.00010910301953818829, "loss": 0.1231, "step": 2056 }, { "epoch": 1.3692794142120153, "grad_norm": 27.761484146118164, "learning_rate": 0.00010905861456483126, "loss": 0.1702, "step": 2057 }, { "epoch": 1.3699450823764354, "grad_norm": 54.4456901550293, "learning_rate": 0.00010901420959147426, "loss": 0.1923, "step": 2058 }, { "epoch": 1.3706107505408553, "grad_norm": 63.133426666259766, "learning_rate": 0.00010896980461811723, "loss": 0.3797, "step": 2059 }, { "epoch": 1.3712764187052755, "grad_norm": 75.04270935058594, "learning_rate": 0.00010892539964476023, "loss": 0.4572, "step": 2060 }, { "epoch": 1.3719420868696954, "grad_norm": 6.204617500305176, "learning_rate": 0.0001088809946714032, "loss": 0.012, "step": 2061 }, { "epoch": 1.3726077550341156, "grad_norm": 14.372852325439453, "learning_rate": 0.00010883658969804617, "loss": 0.0421, "step": 2062 }, { "epoch": 1.3732734231985355, "grad_norm": 124.19054412841797, "learning_rate": 0.00010879218472468917, "loss": 0.9077, "step": 2063 }, { "epoch": 1.3739390913629554, "grad_norm": 31.081514358520508, "learning_rate": 0.00010874777975133214, "loss": 0.0727, "step": 2064 }, { "epoch": 1.3746047595273756, "grad_norm": 42.078060150146484, "learning_rate": 0.00010870337477797514, "loss": 0.1112, "step": 2065 }, { "epoch": 1.3752704276917957, "grad_norm": 7.857242107391357, "learning_rate": 0.00010865896980461811, "loss": 0.0133, "step": 2066 }, { "epoch": 1.3759360958562157, "grad_norm": 48.92726516723633, "learning_rate": 0.00010861456483126111, "loss": 0.585, "step": 2067 }, { "epoch": 1.3766017640206356, "grad_norm": 102.62325286865234, "learning_rate": 0.00010857015985790408, "loss": 0.9239, "step": 2068 }, { "epoch": 1.3772674321850558, "grad_norm": 74.60015869140625, "learning_rate": 0.00010852575488454708, "loss": 0.4072, "step": 2069 }, { "epoch": 1.377933100349476, "grad_norm": 89.37828063964844, "learning_rate": 0.00010848134991119005, "loss": 1.1217, "step": 2070 }, { "epoch": 1.3785987685138958, "grad_norm": 11.525237083435059, "learning_rate": 0.00010843694493783305, "loss": 0.0299, "step": 2071 }, { "epoch": 1.3792644366783158, "grad_norm": 70.63633728027344, "learning_rate": 0.00010839253996447602, "loss": 0.5166, "step": 2072 }, { "epoch": 1.379930104842736, "grad_norm": 26.574739456176758, "learning_rate": 0.00010834813499111902, "loss": 0.0863, "step": 2073 }, { "epoch": 1.3805957730071559, "grad_norm": 92.20925903320312, "learning_rate": 0.00010830373001776199, "loss": 0.5285, "step": 2074 }, { "epoch": 1.381261441171576, "grad_norm": 77.48152160644531, "learning_rate": 0.00010825932504440497, "loss": 0.7548, "step": 2075 }, { "epoch": 1.381927109335996, "grad_norm": 34.449485778808594, "learning_rate": 0.00010821492007104796, "loss": 0.2443, "step": 2076 }, { "epoch": 1.382592777500416, "grad_norm": 31.266016006469727, "learning_rate": 0.00010817051509769094, "loss": 0.1419, "step": 2077 }, { "epoch": 1.383258445664836, "grad_norm": 22.793745040893555, "learning_rate": 0.00010812611012433393, "loss": 0.076, "step": 2078 }, { "epoch": 1.3839241138292562, "grad_norm": 64.29205322265625, "learning_rate": 0.00010808170515097691, "loss": 0.2525, "step": 2079 }, { "epoch": 1.384589781993676, "grad_norm": 31.512174606323242, "learning_rate": 0.0001080373001776199, "loss": 0.0492, "step": 2080 }, { "epoch": 1.3852554501580963, "grad_norm": 64.45855712890625, "learning_rate": 0.00010799289520426288, "loss": 0.5128, "step": 2081 }, { "epoch": 1.3859211183225162, "grad_norm": 14.41579532623291, "learning_rate": 0.00010794849023090587, "loss": 0.0526, "step": 2082 }, { "epoch": 1.3865867864869363, "grad_norm": 31.10066795349121, "learning_rate": 0.00010790408525754885, "loss": 0.1704, "step": 2083 }, { "epoch": 1.3872524546513563, "grad_norm": 14.604890823364258, "learning_rate": 0.00010785968028419184, "loss": 0.0704, "step": 2084 }, { "epoch": 1.3879181228157762, "grad_norm": 27.38639259338379, "learning_rate": 0.00010781527531083482, "loss": 0.1087, "step": 2085 }, { "epoch": 1.3885837909801964, "grad_norm": 27.486242294311523, "learning_rate": 0.0001077708703374778, "loss": 0.1148, "step": 2086 }, { "epoch": 1.3892494591446165, "grad_norm": 58.08623504638672, "learning_rate": 0.00010772646536412079, "loss": 0.356, "step": 2087 }, { "epoch": 1.3899151273090364, "grad_norm": 35.439205169677734, "learning_rate": 0.00010768206039076376, "loss": 0.2203, "step": 2088 }, { "epoch": 1.3905807954734564, "grad_norm": 34.60232925415039, "learning_rate": 0.00010763765541740676, "loss": 0.2061, "step": 2089 }, { "epoch": 1.3912464636378765, "grad_norm": 48.03777313232422, "learning_rate": 0.00010759325044404973, "loss": 0.4746, "step": 2090 }, { "epoch": 1.3919121318022967, "grad_norm": 50.73992156982422, "learning_rate": 0.00010754884547069273, "loss": 0.611, "step": 2091 }, { "epoch": 1.3925777999667166, "grad_norm": 40.05353927612305, "learning_rate": 0.0001075044404973357, "loss": 0.2483, "step": 2092 }, { "epoch": 1.3932434681311365, "grad_norm": 47.57456588745117, "learning_rate": 0.0001074600355239787, "loss": 0.4974, "step": 2093 }, { "epoch": 1.3939091362955567, "grad_norm": 26.729158401489258, "learning_rate": 0.00010741563055062167, "loss": 0.0908, "step": 2094 }, { "epoch": 1.3945748044599768, "grad_norm": 65.19374084472656, "learning_rate": 0.00010737122557726467, "loss": 0.2838, "step": 2095 }, { "epoch": 1.3952404726243968, "grad_norm": 40.21290969848633, "learning_rate": 0.00010732682060390764, "loss": 0.1488, "step": 2096 }, { "epoch": 1.3959061407888167, "grad_norm": 32.270484924316406, "learning_rate": 0.00010728241563055064, "loss": 0.1746, "step": 2097 }, { "epoch": 1.3965718089532368, "grad_norm": 40.07378005981445, "learning_rate": 0.00010723801065719361, "loss": 0.4627, "step": 2098 }, { "epoch": 1.3972374771176568, "grad_norm": 32.53923416137695, "learning_rate": 0.00010719360568383658, "loss": 0.5244, "step": 2099 }, { "epoch": 1.397903145282077, "grad_norm": 8.928348541259766, "learning_rate": 0.00010714920071047958, "loss": 0.0544, "step": 2100 }, { "epoch": 1.3985688134464969, "grad_norm": 52.469608306884766, "learning_rate": 0.00010710479573712255, "loss": 0.2908, "step": 2101 }, { "epoch": 1.399234481610917, "grad_norm": 29.16887855529785, "learning_rate": 0.00010706039076376555, "loss": 0.1055, "step": 2102 }, { "epoch": 1.399900149775337, "grad_norm": 43.890296936035156, "learning_rate": 0.00010701598579040852, "loss": 0.457, "step": 2103 }, { "epoch": 1.400565817939757, "grad_norm": 41.00320816040039, "learning_rate": 0.00010697158081705152, "loss": 0.2166, "step": 2104 }, { "epoch": 1.401231486104177, "grad_norm": 27.757577896118164, "learning_rate": 0.00010692717584369449, "loss": 0.1459, "step": 2105 }, { "epoch": 1.4018971542685972, "grad_norm": 44.91120147705078, "learning_rate": 0.00010688277087033749, "loss": 0.1918, "step": 2106 }, { "epoch": 1.402562822433017, "grad_norm": 82.23652648925781, "learning_rate": 0.00010683836589698046, "loss": 0.7, "step": 2107 }, { "epoch": 1.4032284905974373, "grad_norm": 75.12691497802734, "learning_rate": 0.00010679396092362346, "loss": 0.4777, "step": 2108 }, { "epoch": 1.4038941587618572, "grad_norm": 50.852237701416016, "learning_rate": 0.00010674955595026643, "loss": 0.3406, "step": 2109 }, { "epoch": 1.4045598269262771, "grad_norm": 63.66010665893555, "learning_rate": 0.00010670515097690943, "loss": 0.9751, "step": 2110 }, { "epoch": 1.4052254950906973, "grad_norm": 110.55316925048828, "learning_rate": 0.0001066607460035524, "loss": 0.4959, "step": 2111 }, { "epoch": 1.4058911632551174, "grad_norm": 78.97017669677734, "learning_rate": 0.00010661634103019537, "loss": 1.1101, "step": 2112 }, { "epoch": 1.4065568314195374, "grad_norm": 47.116146087646484, "learning_rate": 0.00010657193605683837, "loss": 0.356, "step": 2113 }, { "epoch": 1.4072224995839573, "grad_norm": 69.57429504394531, "learning_rate": 0.00010652753108348134, "loss": 0.3826, "step": 2114 }, { "epoch": 1.4078881677483774, "grad_norm": 2.5150790214538574, "learning_rate": 0.00010648312611012434, "loss": 0.0085, "step": 2115 }, { "epoch": 1.4085538359127976, "grad_norm": 19.398975372314453, "learning_rate": 0.00010643872113676731, "loss": 0.0565, "step": 2116 }, { "epoch": 1.4092195040772175, "grad_norm": 58.04856491088867, "learning_rate": 0.00010639431616341031, "loss": 0.6235, "step": 2117 }, { "epoch": 1.4098851722416375, "grad_norm": 47.56690979003906, "learning_rate": 0.00010634991119005328, "loss": 0.3233, "step": 2118 }, { "epoch": 1.4105508404060576, "grad_norm": 19.039306640625, "learning_rate": 0.00010630550621669628, "loss": 0.1031, "step": 2119 }, { "epoch": 1.4112165085704775, "grad_norm": 63.333980560302734, "learning_rate": 0.00010626110124333925, "loss": 0.2384, "step": 2120 }, { "epoch": 1.4118821767348977, "grad_norm": 40.69241714477539, "learning_rate": 0.00010621669626998225, "loss": 0.2932, "step": 2121 }, { "epoch": 1.4125478448993176, "grad_norm": 42.182498931884766, "learning_rate": 0.00010617229129662522, "loss": 0.2207, "step": 2122 }, { "epoch": 1.4132135130637378, "grad_norm": 6.699224948883057, "learning_rate": 0.00010612788632326822, "loss": 0.0251, "step": 2123 }, { "epoch": 1.4138791812281577, "grad_norm": 25.854949951171875, "learning_rate": 0.00010608348134991119, "loss": 0.1879, "step": 2124 }, { "epoch": 1.4145448493925779, "grad_norm": 17.82037925720215, "learning_rate": 0.00010603907637655418, "loss": 0.0885, "step": 2125 }, { "epoch": 1.4152105175569978, "grad_norm": 73.820068359375, "learning_rate": 0.00010599467140319716, "loss": 0.6528, "step": 2126 }, { "epoch": 1.415876185721418, "grad_norm": 52.913089752197266, "learning_rate": 0.00010595026642984015, "loss": 0.4037, "step": 2127 }, { "epoch": 1.4165418538858379, "grad_norm": 60.46794891357422, "learning_rate": 0.00010590586145648313, "loss": 0.239, "step": 2128 }, { "epoch": 1.417207522050258, "grad_norm": 41.661197662353516, "learning_rate": 0.00010586145648312611, "loss": 0.2722, "step": 2129 }, { "epoch": 1.417873190214678, "grad_norm": 8.981060028076172, "learning_rate": 0.0001058170515097691, "loss": 0.0202, "step": 2130 }, { "epoch": 1.418538858379098, "grad_norm": 23.493610382080078, "learning_rate": 0.00010577264653641208, "loss": 0.1096, "step": 2131 }, { "epoch": 1.419204526543518, "grad_norm": 51.386390686035156, "learning_rate": 0.00010572824156305507, "loss": 0.3341, "step": 2132 }, { "epoch": 1.4198701947079382, "grad_norm": 20.230579376220703, "learning_rate": 0.00010568383658969805, "loss": 0.0919, "step": 2133 }, { "epoch": 1.4205358628723581, "grad_norm": 37.420188903808594, "learning_rate": 0.00010563943161634104, "loss": 0.1113, "step": 2134 }, { "epoch": 1.421201531036778, "grad_norm": 34.74143600463867, "learning_rate": 0.00010559502664298402, "loss": 0.2637, "step": 2135 }, { "epoch": 1.4218671992011982, "grad_norm": 42.98000717163086, "learning_rate": 0.000105550621669627, "loss": 0.5282, "step": 2136 }, { "epoch": 1.4225328673656183, "grad_norm": 11.664846420288086, "learning_rate": 0.00010550621669627, "loss": 0.0394, "step": 2137 }, { "epoch": 1.4231985355300383, "grad_norm": 76.63288879394531, "learning_rate": 0.00010546181172291296, "loss": 0.5377, "step": 2138 }, { "epoch": 1.4238642036944582, "grad_norm": 50.07666015625, "learning_rate": 0.00010541740674955596, "loss": 0.2778, "step": 2139 }, { "epoch": 1.4245298718588784, "grad_norm": 19.871234893798828, "learning_rate": 0.00010537300177619893, "loss": 0.0415, "step": 2140 }, { "epoch": 1.4251955400232985, "grad_norm": 50.02949905395508, "learning_rate": 0.00010532859680284193, "loss": 0.1708, "step": 2141 }, { "epoch": 1.4258612081877184, "grad_norm": 25.459726333618164, "learning_rate": 0.0001052841918294849, "loss": 0.0541, "step": 2142 }, { "epoch": 1.4265268763521384, "grad_norm": 104.60662078857422, "learning_rate": 0.0001052397868561279, "loss": 1.1574, "step": 2143 }, { "epoch": 1.4271925445165585, "grad_norm": 29.339962005615234, "learning_rate": 0.00010519538188277087, "loss": 0.1013, "step": 2144 }, { "epoch": 1.4278582126809785, "grad_norm": 65.1827621459961, "learning_rate": 0.00010515097690941387, "loss": 0.5037, "step": 2145 }, { "epoch": 1.4285238808453986, "grad_norm": 38.673099517822266, "learning_rate": 0.00010510657193605684, "loss": 0.2915, "step": 2146 }, { "epoch": 1.4291895490098185, "grad_norm": 70.0672836303711, "learning_rate": 0.00010506216696269984, "loss": 0.7252, "step": 2147 }, { "epoch": 1.4298552171742387, "grad_norm": 7.679555416107178, "learning_rate": 0.00010501776198934281, "loss": 0.021, "step": 2148 }, { "epoch": 1.4305208853386586, "grad_norm": 139.7518768310547, "learning_rate": 0.00010497335701598578, "loss": 1.6982, "step": 2149 }, { "epoch": 1.4311865535030788, "grad_norm": 49.534420013427734, "learning_rate": 0.00010492895204262878, "loss": 0.2796, "step": 2150 }, { "epoch": 1.4318522216674987, "grad_norm": 35.231990814208984, "learning_rate": 0.00010488454706927175, "loss": 0.2309, "step": 2151 }, { "epoch": 1.4325178898319189, "grad_norm": 74.30767822265625, "learning_rate": 0.00010484014209591475, "loss": 0.7403, "step": 2152 }, { "epoch": 1.4331835579963388, "grad_norm": 61.4307861328125, "learning_rate": 0.00010479573712255772, "loss": 0.6435, "step": 2153 }, { "epoch": 1.433849226160759, "grad_norm": 2.556778907775879, "learning_rate": 0.00010475133214920072, "loss": 0.0078, "step": 2154 }, { "epoch": 1.4345148943251789, "grad_norm": 24.79681396484375, "learning_rate": 0.0001047069271758437, "loss": 0.0745, "step": 2155 }, { "epoch": 1.4351805624895988, "grad_norm": 37.959228515625, "learning_rate": 0.00010466252220248669, "loss": 0.2984, "step": 2156 }, { "epoch": 1.435846230654019, "grad_norm": 35.0047492980957, "learning_rate": 0.00010461811722912966, "loss": 0.291, "step": 2157 }, { "epoch": 1.436511898818439, "grad_norm": 23.24560546875, "learning_rate": 0.00010457371225577266, "loss": 0.1324, "step": 2158 }, { "epoch": 1.437177566982859, "grad_norm": 33.166080474853516, "learning_rate": 0.00010452930728241563, "loss": 0.3158, "step": 2159 }, { "epoch": 1.437843235147279, "grad_norm": 48.673370361328125, "learning_rate": 0.00010448490230905863, "loss": 0.2514, "step": 2160 }, { "epoch": 1.4385089033116991, "grad_norm": 62.832191467285156, "learning_rate": 0.0001044404973357016, "loss": 0.3518, "step": 2161 }, { "epoch": 1.4391745714761193, "grad_norm": 25.648235321044922, "learning_rate": 0.00010439609236234457, "loss": 0.0971, "step": 2162 }, { "epoch": 1.4398402396405392, "grad_norm": 28.441898345947266, "learning_rate": 0.00010435168738898757, "loss": 0.0934, "step": 2163 }, { "epoch": 1.4405059078049591, "grad_norm": 14.695634841918945, "learning_rate": 0.00010430728241563054, "loss": 0.0466, "step": 2164 }, { "epoch": 1.4411715759693793, "grad_norm": 35.00516891479492, "learning_rate": 0.00010426287744227354, "loss": 0.2279, "step": 2165 }, { "epoch": 1.4418372441337994, "grad_norm": 14.605650901794434, "learning_rate": 0.00010421847246891651, "loss": 0.0521, "step": 2166 }, { "epoch": 1.4425029122982194, "grad_norm": 92.80223083496094, "learning_rate": 0.00010417406749555951, "loss": 0.5935, "step": 2167 }, { "epoch": 1.4431685804626393, "grad_norm": 62.45814514160156, "learning_rate": 0.00010412966252220248, "loss": 0.3422, "step": 2168 }, { "epoch": 1.4438342486270594, "grad_norm": 9.82910442352295, "learning_rate": 0.00010408525754884548, "loss": 0.0351, "step": 2169 }, { "epoch": 1.4444999167914794, "grad_norm": 32.248870849609375, "learning_rate": 0.00010404085257548845, "loss": 0.2117, "step": 2170 }, { "epoch": 1.4451655849558995, "grad_norm": 36.799705505371094, "learning_rate": 0.00010399644760213145, "loss": 0.1821, "step": 2171 }, { "epoch": 1.4458312531203195, "grad_norm": 35.92013168334961, "learning_rate": 0.00010395204262877442, "loss": 0.1772, "step": 2172 }, { "epoch": 1.4464969212847396, "grad_norm": 64.58824157714844, "learning_rate": 0.00010390763765541742, "loss": 0.3886, "step": 2173 }, { "epoch": 1.4471625894491595, "grad_norm": 9.116527557373047, "learning_rate": 0.00010386323268206039, "loss": 0.0173, "step": 2174 }, { "epoch": 1.4478282576135797, "grad_norm": 33.91664505004883, "learning_rate": 0.00010381882770870338, "loss": 0.2462, "step": 2175 }, { "epoch": 1.4484939257779996, "grad_norm": 79.28636169433594, "learning_rate": 0.00010377442273534636, "loss": 0.5193, "step": 2176 }, { "epoch": 1.4491595939424198, "grad_norm": 64.25091552734375, "learning_rate": 0.00010373001776198935, "loss": 0.4757, "step": 2177 }, { "epoch": 1.4498252621068397, "grad_norm": 36.867591857910156, "learning_rate": 0.00010368561278863233, "loss": 0.2, "step": 2178 }, { "epoch": 1.4504909302712599, "grad_norm": 25.86852264404297, "learning_rate": 0.00010364120781527532, "loss": 0.0309, "step": 2179 }, { "epoch": 1.4511565984356798, "grad_norm": 25.260671615600586, "learning_rate": 0.0001035968028419183, "loss": 0.0824, "step": 2180 }, { "epoch": 1.4518222666000997, "grad_norm": 47.031131744384766, "learning_rate": 0.00010355239786856129, "loss": 0.5144, "step": 2181 }, { "epoch": 1.4524879347645199, "grad_norm": 94.73207092285156, "learning_rate": 0.00010350799289520427, "loss": 0.6515, "step": 2182 }, { "epoch": 1.45315360292894, "grad_norm": 64.87279510498047, "learning_rate": 0.00010346358792184726, "loss": 0.2847, "step": 2183 }, { "epoch": 1.45381927109336, "grad_norm": 2.1851558685302734, "learning_rate": 0.00010341918294849024, "loss": 0.0064, "step": 2184 }, { "epoch": 1.4544849392577799, "grad_norm": 12.116601943969727, "learning_rate": 0.00010337477797513322, "loss": 0.0412, "step": 2185 }, { "epoch": 1.4551506074222, "grad_norm": 43.704750061035156, "learning_rate": 0.0001033303730017762, "loss": 0.218, "step": 2186 }, { "epoch": 1.4558162755866202, "grad_norm": 76.6130599975586, "learning_rate": 0.0001032859680284192, "loss": 0.4313, "step": 2187 }, { "epoch": 1.4564819437510401, "grad_norm": 5.211718559265137, "learning_rate": 0.00010324156305506217, "loss": 0.0155, "step": 2188 }, { "epoch": 1.45714761191546, "grad_norm": 37.67426300048828, "learning_rate": 0.00010319715808170516, "loss": 0.1681, "step": 2189 }, { "epoch": 1.4578132800798802, "grad_norm": 37.42005920410156, "learning_rate": 0.00010315275310834814, "loss": 0.0969, "step": 2190 }, { "epoch": 1.4584789482443004, "grad_norm": 41.34249496459961, "learning_rate": 0.00010310834813499113, "loss": 0.4342, "step": 2191 }, { "epoch": 1.4591446164087203, "grad_norm": 42.75221633911133, "learning_rate": 0.0001030639431616341, "loss": 0.1815, "step": 2192 }, { "epoch": 1.4598102845731402, "grad_norm": 134.22018432617188, "learning_rate": 0.0001030195381882771, "loss": 0.6194, "step": 2193 }, { "epoch": 1.4604759527375604, "grad_norm": 42.21216583251953, "learning_rate": 0.00010297513321492007, "loss": 0.2403, "step": 2194 }, { "epoch": 1.4611416209019803, "grad_norm": 53.971588134765625, "learning_rate": 0.00010293072824156307, "loss": 0.1596, "step": 2195 }, { "epoch": 1.4618072890664005, "grad_norm": 10.36235237121582, "learning_rate": 0.00010288632326820604, "loss": 0.034, "step": 2196 }, { "epoch": 1.4624729572308204, "grad_norm": 22.25579261779785, "learning_rate": 0.00010284191829484904, "loss": 0.0626, "step": 2197 }, { "epoch": 1.4631386253952405, "grad_norm": 10.987948417663574, "learning_rate": 0.00010279751332149201, "loss": 0.027, "step": 2198 }, { "epoch": 1.4638042935596605, "grad_norm": 63.69123077392578, "learning_rate": 0.00010275310834813499, "loss": 0.285, "step": 2199 }, { "epoch": 1.4644699617240806, "grad_norm": 52.57264709472656, "learning_rate": 0.00010270870337477798, "loss": 0.2292, "step": 2200 }, { "epoch": 1.4651356298885005, "grad_norm": 53.009735107421875, "learning_rate": 0.00010266429840142096, "loss": 0.2156, "step": 2201 }, { "epoch": 1.4658012980529207, "grad_norm": 25.717741012573242, "learning_rate": 0.00010261989342806395, "loss": 0.1236, "step": 2202 }, { "epoch": 1.4664669662173406, "grad_norm": 24.538572311401367, "learning_rate": 0.00010257548845470692, "loss": 0.1124, "step": 2203 }, { "epoch": 1.4671326343817608, "grad_norm": 23.42035484313965, "learning_rate": 0.00010253108348134992, "loss": 0.0567, "step": 2204 }, { "epoch": 1.4677983025461807, "grad_norm": 29.259788513183594, "learning_rate": 0.0001024866785079929, "loss": 0.1598, "step": 2205 }, { "epoch": 1.4684639707106006, "grad_norm": 51.69668960571289, "learning_rate": 0.00010244227353463589, "loss": 0.2295, "step": 2206 }, { "epoch": 1.4691296388750208, "grad_norm": 46.6073112487793, "learning_rate": 0.00010239786856127886, "loss": 0.2923, "step": 2207 }, { "epoch": 1.469795307039441, "grad_norm": 62.206031799316406, "learning_rate": 0.00010235346358792186, "loss": 0.5158, "step": 2208 }, { "epoch": 1.4704609752038609, "grad_norm": 52.34242630004883, "learning_rate": 0.00010230905861456483, "loss": 0.7113, "step": 2209 }, { "epoch": 1.4711266433682808, "grad_norm": 41.1063117980957, "learning_rate": 0.00010226465364120783, "loss": 0.3134, "step": 2210 }, { "epoch": 1.471792311532701, "grad_norm": 80.76094055175781, "learning_rate": 0.0001022202486678508, "loss": 0.5217, "step": 2211 }, { "epoch": 1.472457979697121, "grad_norm": 39.76702117919922, "learning_rate": 0.00010217584369449377, "loss": 0.216, "step": 2212 }, { "epoch": 1.473123647861541, "grad_norm": 55.79610824584961, "learning_rate": 0.00010213143872113677, "loss": 0.3369, "step": 2213 }, { "epoch": 1.473789316025961, "grad_norm": 52.35502243041992, "learning_rate": 0.00010208703374777974, "loss": 0.236, "step": 2214 }, { "epoch": 1.4744549841903811, "grad_norm": 22.270151138305664, "learning_rate": 0.00010204262877442274, "loss": 0.1002, "step": 2215 }, { "epoch": 1.475120652354801, "grad_norm": 10.123306274414062, "learning_rate": 0.00010199822380106571, "loss": 0.0291, "step": 2216 }, { "epoch": 1.4757863205192212, "grad_norm": 62.4449348449707, "learning_rate": 0.00010195381882770871, "loss": 0.3618, "step": 2217 }, { "epoch": 1.4764519886836411, "grad_norm": 68.68087005615234, "learning_rate": 0.00010190941385435168, "loss": 0.2271, "step": 2218 }, { "epoch": 1.4771176568480613, "grad_norm": 3.106027364730835, "learning_rate": 0.00010186500888099468, "loss": 0.0097, "step": 2219 }, { "epoch": 1.4777833250124812, "grad_norm": 1.2246886491775513, "learning_rate": 0.00010182060390763765, "loss": 0.0041, "step": 2220 }, { "epoch": 1.4784489931769014, "grad_norm": 6.962759494781494, "learning_rate": 0.00010177619893428065, "loss": 0.0268, "step": 2221 }, { "epoch": 1.4791146613413213, "grad_norm": 8.942953109741211, "learning_rate": 0.00010173179396092362, "loss": 0.0235, "step": 2222 }, { "epoch": 1.4797803295057415, "grad_norm": 112.1907958984375, "learning_rate": 0.00010168738898756662, "loss": 0.4712, "step": 2223 }, { "epoch": 1.4804459976701614, "grad_norm": 57.0410270690918, "learning_rate": 0.00010164298401420959, "loss": 0.5294, "step": 2224 }, { "epoch": 1.4811116658345815, "grad_norm": 16.46082878112793, "learning_rate": 0.00010159857904085256, "loss": 0.0419, "step": 2225 }, { "epoch": 1.4817773339990015, "grad_norm": 15.291955947875977, "learning_rate": 0.00010155417406749556, "loss": 0.0277, "step": 2226 }, { "epoch": 1.4824430021634214, "grad_norm": 30.236244201660156, "learning_rate": 0.00010150976909413853, "loss": 0.1584, "step": 2227 }, { "epoch": 1.4831086703278415, "grad_norm": 39.912269592285156, "learning_rate": 0.00010146536412078153, "loss": 0.0919, "step": 2228 }, { "epoch": 1.4837743384922617, "grad_norm": 29.772533416748047, "learning_rate": 0.0001014209591474245, "loss": 0.0596, "step": 2229 }, { "epoch": 1.4844400066566816, "grad_norm": 48.643428802490234, "learning_rate": 0.0001013765541740675, "loss": 0.4017, "step": 2230 }, { "epoch": 1.4851056748211016, "grad_norm": 150.6394805908203, "learning_rate": 0.00010133214920071047, "loss": 1.7441, "step": 2231 }, { "epoch": 1.4857713429855217, "grad_norm": 41.88225555419922, "learning_rate": 0.00010128774422735347, "loss": 0.3285, "step": 2232 }, { "epoch": 1.4864370111499419, "grad_norm": 137.64869689941406, "learning_rate": 0.00010124333925399646, "loss": 0.3519, "step": 2233 }, { "epoch": 1.4871026793143618, "grad_norm": 44.37678909301758, "learning_rate": 0.00010119893428063944, "loss": 0.2344, "step": 2234 }, { "epoch": 1.4877683474787817, "grad_norm": 35.5928840637207, "learning_rate": 0.00010115452930728243, "loss": 0.1774, "step": 2235 }, { "epoch": 1.4884340156432019, "grad_norm": 8.928791046142578, "learning_rate": 0.0001011101243339254, "loss": 0.026, "step": 2236 }, { "epoch": 1.489099683807622, "grad_norm": 89.05236053466797, "learning_rate": 0.0001010657193605684, "loss": 0.5801, "step": 2237 }, { "epoch": 1.489765351972042, "grad_norm": 113.47551727294922, "learning_rate": 0.00010102131438721137, "loss": 0.2874, "step": 2238 }, { "epoch": 1.490431020136462, "grad_norm": 63.45227813720703, "learning_rate": 0.00010097690941385437, "loss": 0.7436, "step": 2239 }, { "epoch": 1.491096688300882, "grad_norm": 18.672109603881836, "learning_rate": 0.00010093250444049734, "loss": 0.0378, "step": 2240 }, { "epoch": 1.491762356465302, "grad_norm": 16.842517852783203, "learning_rate": 0.00010088809946714034, "loss": 0.0404, "step": 2241 }, { "epoch": 1.4924280246297221, "grad_norm": 23.478473663330078, "learning_rate": 0.0001008436944937833, "loss": 0.0695, "step": 2242 }, { "epoch": 1.493093692794142, "grad_norm": 20.125471115112305, "learning_rate": 0.0001007992895204263, "loss": 0.0733, "step": 2243 }, { "epoch": 1.4937593609585622, "grad_norm": 28.721750259399414, "learning_rate": 0.00010075488454706928, "loss": 0.1831, "step": 2244 }, { "epoch": 1.4944250291229821, "grad_norm": 50.20158004760742, "learning_rate": 0.00010071047957371227, "loss": 0.6617, "step": 2245 }, { "epoch": 1.4950906972874023, "grad_norm": 37.243309020996094, "learning_rate": 0.00010066607460035525, "loss": 0.1704, "step": 2246 }, { "epoch": 1.4957563654518222, "grad_norm": 3.990790367126465, "learning_rate": 0.00010062166962699824, "loss": 0.0052, "step": 2247 }, { "epoch": 1.4964220336162424, "grad_norm": 15.846875190734863, "learning_rate": 0.00010057726465364122, "loss": 0.029, "step": 2248 }, { "epoch": 1.4970877017806623, "grad_norm": 13.285712242126465, "learning_rate": 0.00010053285968028419, "loss": 0.026, "step": 2249 }, { "epoch": 1.4977533699450825, "grad_norm": 59.312156677246094, "learning_rate": 0.00010048845470692719, "loss": 1.1105, "step": 2250 }, { "epoch": 1.4984190381095024, "grad_norm": 93.45195007324219, "learning_rate": 0.00010044404973357016, "loss": 0.441, "step": 2251 }, { "epoch": 1.4990847062739223, "grad_norm": 39.31588363647461, "learning_rate": 0.00010039964476021315, "loss": 0.657, "step": 2252 }, { "epoch": 1.4997503744383425, "grad_norm": 54.291255950927734, "learning_rate": 0.00010035523978685613, "loss": 0.3593, "step": 2253 }, { "epoch": 1.5004160426027626, "grad_norm": 23.339929580688477, "learning_rate": 0.00010031083481349912, "loss": 0.0568, "step": 2254 }, { "epoch": 1.5010817107671826, "grad_norm": 108.71127319335938, "learning_rate": 0.0001002664298401421, "loss": 0.5498, "step": 2255 }, { "epoch": 1.5017473789316025, "grad_norm": 6.677554607391357, "learning_rate": 0.0001002220248667851, "loss": 0.0166, "step": 2256 }, { "epoch": 1.5024130470960226, "grad_norm": 42.01664352416992, "learning_rate": 0.00010017761989342807, "loss": 0.2364, "step": 2257 }, { "epoch": 1.5030787152604428, "grad_norm": 53.1296272277832, "learning_rate": 0.00010013321492007106, "loss": 0.4037, "step": 2258 }, { "epoch": 1.5037443834248627, "grad_norm": 29.98288917541504, "learning_rate": 0.00010008880994671404, "loss": 0.148, "step": 2259 }, { "epoch": 1.5044100515892826, "grad_norm": 5.959773540496826, "learning_rate": 0.00010004440497335703, "loss": 0.0238, "step": 2260 }, { "epoch": 1.5050757197537028, "grad_norm": 54.67912292480469, "learning_rate": 0.0001, "loss": 0.2246, "step": 2261 }, { "epoch": 1.505741387918123, "grad_norm": 78.04622650146484, "learning_rate": 9.995559502664299e-05, "loss": 0.5267, "step": 2262 }, { "epoch": 1.5064070560825429, "grad_norm": 81.58208465576172, "learning_rate": 9.991119005328597e-05, "loss": 1.0155, "step": 2263 }, { "epoch": 1.5070727242469628, "grad_norm": 71.14002990722656, "learning_rate": 9.986678507992896e-05, "loss": 0.2742, "step": 2264 }, { "epoch": 1.507738392411383, "grad_norm": 36.318626403808594, "learning_rate": 9.982238010657194e-05, "loss": 0.2166, "step": 2265 }, { "epoch": 1.5084040605758031, "grad_norm": 25.122879028320312, "learning_rate": 9.977797513321493e-05, "loss": 0.0831, "step": 2266 }, { "epoch": 1.509069728740223, "grad_norm": 58.27655029296875, "learning_rate": 9.973357015985791e-05, "loss": 0.1612, "step": 2267 }, { "epoch": 1.509735396904643, "grad_norm": 65.42062377929688, "learning_rate": 9.968916518650089e-05, "loss": 0.2846, "step": 2268 }, { "epoch": 1.510401065069063, "grad_norm": 78.4611587524414, "learning_rate": 9.964476021314387e-05, "loss": 0.2751, "step": 2269 }, { "epoch": 1.511066733233483, "grad_norm": 38.92177963256836, "learning_rate": 9.960035523978685e-05, "loss": 0.5032, "step": 2270 }, { "epoch": 1.5117324013979032, "grad_norm": 52.774871826171875, "learning_rate": 9.955595026642984e-05, "loss": 0.4692, "step": 2271 }, { "epoch": 1.5123980695623231, "grad_norm": 36.72901153564453, "learning_rate": 9.951154529307282e-05, "loss": 0.109, "step": 2272 }, { "epoch": 1.513063737726743, "grad_norm": 8.902722358703613, "learning_rate": 9.946714031971581e-05, "loss": 0.03, "step": 2273 }, { "epoch": 1.5137294058911632, "grad_norm": 61.80864334106445, "learning_rate": 9.94227353463588e-05, "loss": 0.553, "step": 2274 }, { "epoch": 1.5143950740555834, "grad_norm": 2.9053337574005127, "learning_rate": 9.937833037300178e-05, "loss": 0.0144, "step": 2275 }, { "epoch": 1.5150607422200033, "grad_norm": 38.89218521118164, "learning_rate": 9.933392539964476e-05, "loss": 0.1599, "step": 2276 }, { "epoch": 1.5157264103844232, "grad_norm": 68.37144470214844, "learning_rate": 9.928952042628775e-05, "loss": 0.5475, "step": 2277 }, { "epoch": 1.5163920785488434, "grad_norm": 12.99852180480957, "learning_rate": 9.924511545293073e-05, "loss": 0.0645, "step": 2278 }, { "epoch": 1.5170577467132635, "grad_norm": 53.400325775146484, "learning_rate": 9.920071047957372e-05, "loss": 0.3663, "step": 2279 }, { "epoch": 1.5177234148776835, "grad_norm": 23.889270782470703, "learning_rate": 9.915630550621669e-05, "loss": 0.0657, "step": 2280 }, { "epoch": 1.5183890830421034, "grad_norm": 6.495269298553467, "learning_rate": 9.911190053285967e-05, "loss": 0.0197, "step": 2281 }, { "epoch": 1.5190547512065236, "grad_norm": 24.418495178222656, "learning_rate": 9.906749555950266e-05, "loss": 0.118, "step": 2282 }, { "epoch": 1.5197204193709437, "grad_norm": 24.35348892211914, "learning_rate": 9.902309058614564e-05, "loss": 0.329, "step": 2283 }, { "epoch": 1.5203860875353636, "grad_norm": 66.30371856689453, "learning_rate": 9.897868561278863e-05, "loss": 0.3274, "step": 2284 }, { "epoch": 1.5210517556997836, "grad_norm": 17.211938858032227, "learning_rate": 9.893428063943161e-05, "loss": 0.0542, "step": 2285 }, { "epoch": 1.5217174238642037, "grad_norm": 6.886520862579346, "learning_rate": 9.88898756660746e-05, "loss": 0.0097, "step": 2286 }, { "epoch": 1.5223830920286239, "grad_norm": 5.7531023025512695, "learning_rate": 9.884547069271758e-05, "loss": 0.0221, "step": 2287 }, { "epoch": 1.5230487601930438, "grad_norm": 26.847780227661133, "learning_rate": 9.880106571936057e-05, "loss": 0.1768, "step": 2288 }, { "epoch": 1.5237144283574637, "grad_norm": 51.29607009887695, "learning_rate": 9.875666074600355e-05, "loss": 0.2611, "step": 2289 }, { "epoch": 1.5243800965218839, "grad_norm": 105.51429748535156, "learning_rate": 9.871225577264654e-05, "loss": 0.9774, "step": 2290 }, { "epoch": 1.525045764686304, "grad_norm": 6.155182838439941, "learning_rate": 9.866785079928954e-05, "loss": 0.0398, "step": 2291 }, { "epoch": 1.525711432850724, "grad_norm": 30.219083786010742, "learning_rate": 9.862344582593252e-05, "loss": 0.0978, "step": 2292 }, { "epoch": 1.526377101015144, "grad_norm": 30.354900360107422, "learning_rate": 9.857904085257549e-05, "loss": 0.1506, "step": 2293 }, { "epoch": 1.5270427691795638, "grad_norm": 45.86323165893555, "learning_rate": 9.853463587921848e-05, "loss": 0.3273, "step": 2294 }, { "epoch": 1.527708437343984, "grad_norm": 3.1487386226654053, "learning_rate": 9.849023090586146e-05, "loss": 0.0094, "step": 2295 }, { "epoch": 1.5283741055084041, "grad_norm": 100.65200805664062, "learning_rate": 9.844582593250445e-05, "loss": 0.7518, "step": 2296 }, { "epoch": 1.529039773672824, "grad_norm": 73.65653228759766, "learning_rate": 9.840142095914743e-05, "loss": 0.4797, "step": 2297 }, { "epoch": 1.529705441837244, "grad_norm": 79.96868896484375, "learning_rate": 9.835701598579042e-05, "loss": 0.3993, "step": 2298 }, { "epoch": 1.5303711100016641, "grad_norm": 53.2676887512207, "learning_rate": 9.83126110124334e-05, "loss": 0.5051, "step": 2299 }, { "epoch": 1.5310367781660843, "grad_norm": 16.816186904907227, "learning_rate": 9.826820603907639e-05, "loss": 0.033, "step": 2300 }, { "epoch": 1.5317024463305042, "grad_norm": 102.59551239013672, "learning_rate": 9.822380106571937e-05, "loss": 0.5525, "step": 2301 }, { "epoch": 1.5323681144949242, "grad_norm": 16.096580505371094, "learning_rate": 9.817939609236236e-05, "loss": 0.0351, "step": 2302 }, { "epoch": 1.5330337826593443, "grad_norm": 59.30455780029297, "learning_rate": 9.813499111900534e-05, "loss": 0.1185, "step": 2303 }, { "epoch": 1.5336994508237645, "grad_norm": 75.46652221679688, "learning_rate": 9.809058614564833e-05, "loss": 0.5784, "step": 2304 }, { "epoch": 1.5343651189881844, "grad_norm": 22.132413864135742, "learning_rate": 9.80461811722913e-05, "loss": 0.1176, "step": 2305 }, { "epoch": 1.5350307871526043, "grad_norm": 27.462966918945312, "learning_rate": 9.800177619893428e-05, "loss": 0.0316, "step": 2306 }, { "epoch": 1.5356964553170245, "grad_norm": 107.68171691894531, "learning_rate": 9.795737122557727e-05, "loss": 0.2219, "step": 2307 }, { "epoch": 1.5363621234814446, "grad_norm": 14.660025596618652, "learning_rate": 9.791296625222025e-05, "loss": 0.0305, "step": 2308 }, { "epoch": 1.5370277916458646, "grad_norm": 69.86400604248047, "learning_rate": 9.786856127886324e-05, "loss": 0.723, "step": 2309 }, { "epoch": 1.5376934598102845, "grad_norm": 3.898486852645874, "learning_rate": 9.782415630550622e-05, "loss": 0.0115, "step": 2310 }, { "epoch": 1.5383591279747046, "grad_norm": 52.487953186035156, "learning_rate": 9.77797513321492e-05, "loss": 0.4256, "step": 2311 }, { "epoch": 1.5390247961391248, "grad_norm": 85.11210632324219, "learning_rate": 9.773534635879219e-05, "loss": 0.5651, "step": 2312 }, { "epoch": 1.5396904643035447, "grad_norm": 4.776047229766846, "learning_rate": 9.769094138543518e-05, "loss": 0.0105, "step": 2313 }, { "epoch": 1.5403561324679647, "grad_norm": 9.567429542541504, "learning_rate": 9.764653641207816e-05, "loss": 0.0198, "step": 2314 }, { "epoch": 1.5410218006323848, "grad_norm": 35.34830856323242, "learning_rate": 9.760213143872115e-05, "loss": 0.1913, "step": 2315 }, { "epoch": 1.5416874687968047, "grad_norm": 43.557884216308594, "learning_rate": 9.755772646536413e-05, "loss": 0.178, "step": 2316 }, { "epoch": 1.542353136961225, "grad_norm": 31.360998153686523, "learning_rate": 9.751332149200711e-05, "loss": 0.0553, "step": 2317 }, { "epoch": 1.5430188051256448, "grad_norm": 53.95760726928711, "learning_rate": 9.746891651865009e-05, "loss": 0.2899, "step": 2318 }, { "epoch": 1.5436844732900648, "grad_norm": 52.672725677490234, "learning_rate": 9.742451154529307e-05, "loss": 0.5019, "step": 2319 }, { "epoch": 1.544350141454485, "grad_norm": 3.912768840789795, "learning_rate": 9.738010657193606e-05, "loss": 0.0127, "step": 2320 }, { "epoch": 1.545015809618905, "grad_norm": 29.462934494018555, "learning_rate": 9.733570159857904e-05, "loss": 0.075, "step": 2321 }, { "epoch": 1.545681477783325, "grad_norm": 39.1092529296875, "learning_rate": 9.729129662522203e-05, "loss": 0.1642, "step": 2322 }, { "epoch": 1.546347145947745, "grad_norm": 29.544994354248047, "learning_rate": 9.724689165186501e-05, "loss": 0.476, "step": 2323 }, { "epoch": 1.547012814112165, "grad_norm": 111.7665786743164, "learning_rate": 9.7202486678508e-05, "loss": 0.4582, "step": 2324 }, { "epoch": 1.5476784822765852, "grad_norm": 12.43513298034668, "learning_rate": 9.715808170515098e-05, "loss": 0.0439, "step": 2325 }, { "epoch": 1.5483441504410052, "grad_norm": 35.20425796508789, "learning_rate": 9.711367673179396e-05, "loss": 0.151, "step": 2326 }, { "epoch": 1.549009818605425, "grad_norm": 91.79122924804688, "learning_rate": 9.706927175843695e-05, "loss": 0.5051, "step": 2327 }, { "epoch": 1.5496754867698452, "grad_norm": 25.28584098815918, "learning_rate": 9.702486678507993e-05, "loss": 0.0884, "step": 2328 }, { "epoch": 1.5503411549342654, "grad_norm": 94.55110931396484, "learning_rate": 9.698046181172292e-05, "loss": 0.5131, "step": 2329 }, { "epoch": 1.5510068230986853, "grad_norm": 2.6768691539764404, "learning_rate": 9.693605683836589e-05, "loss": 0.0055, "step": 2330 }, { "epoch": 1.5516724912631052, "grad_norm": 50.229618072509766, "learning_rate": 9.689165186500888e-05, "loss": 0.4457, "step": 2331 }, { "epoch": 1.5523381594275254, "grad_norm": 62.407711029052734, "learning_rate": 9.684724689165186e-05, "loss": 0.2177, "step": 2332 }, { "epoch": 1.5530038275919456, "grad_norm": 31.880765914916992, "learning_rate": 9.680284191829485e-05, "loss": 0.1747, "step": 2333 }, { "epoch": 1.5536694957563655, "grad_norm": 70.33451843261719, "learning_rate": 9.675843694493783e-05, "loss": 0.5242, "step": 2334 }, { "epoch": 1.5543351639207854, "grad_norm": 54.715091705322266, "learning_rate": 9.671403197158081e-05, "loss": 0.47, "step": 2335 }, { "epoch": 1.5550008320852056, "grad_norm": 32.43948745727539, "learning_rate": 9.66696269982238e-05, "loss": 0.0756, "step": 2336 }, { "epoch": 1.5556665002496257, "grad_norm": 13.764049530029297, "learning_rate": 9.662522202486678e-05, "loss": 0.0512, "step": 2337 }, { "epoch": 1.5563321684140456, "grad_norm": 14.379326820373535, "learning_rate": 9.658081705150977e-05, "loss": 0.04, "step": 2338 }, { "epoch": 1.5569978365784656, "grad_norm": 23.684633255004883, "learning_rate": 9.653641207815275e-05, "loss": 0.1049, "step": 2339 }, { "epoch": 1.5576635047428855, "grad_norm": 58.02132797241211, "learning_rate": 9.649200710479574e-05, "loss": 0.2001, "step": 2340 }, { "epoch": 1.5583291729073057, "grad_norm": 69.31839752197266, "learning_rate": 9.644760213143872e-05, "loss": 0.7277, "step": 2341 }, { "epoch": 1.5589948410717258, "grad_norm": 120.31485748291016, "learning_rate": 9.640319715808171e-05, "loss": 0.1035, "step": 2342 }, { "epoch": 1.5596605092361457, "grad_norm": 58.467567443847656, "learning_rate": 9.63587921847247e-05, "loss": 0.2483, "step": 2343 }, { "epoch": 1.5603261774005657, "grad_norm": 16.387710571289062, "learning_rate": 9.631438721136768e-05, "loss": 0.0536, "step": 2344 }, { "epoch": 1.5609918455649858, "grad_norm": 7.574170112609863, "learning_rate": 9.626998223801066e-05, "loss": 0.0356, "step": 2345 }, { "epoch": 1.561657513729406, "grad_norm": 36.53459930419922, "learning_rate": 9.622557726465365e-05, "loss": 0.1704, "step": 2346 }, { "epoch": 1.562323181893826, "grad_norm": 76.53927612304688, "learning_rate": 9.618117229129663e-05, "loss": 0.4029, "step": 2347 }, { "epoch": 1.5629888500582458, "grad_norm": 46.571990966796875, "learning_rate": 9.613676731793962e-05, "loss": 0.0559, "step": 2348 }, { "epoch": 1.563654518222666, "grad_norm": 5.6426777839660645, "learning_rate": 9.60923623445826e-05, "loss": 0.0169, "step": 2349 }, { "epoch": 1.5643201863870861, "grad_norm": 31.623960494995117, "learning_rate": 9.604795737122559e-05, "loss": 0.0444, "step": 2350 }, { "epoch": 1.564985854551506, "grad_norm": 70.62718963623047, "learning_rate": 9.600355239786857e-05, "loss": 0.3884, "step": 2351 }, { "epoch": 1.565651522715926, "grad_norm": 54.923091888427734, "learning_rate": 9.595914742451156e-05, "loss": 0.1408, "step": 2352 }, { "epoch": 1.5663171908803462, "grad_norm": 34.60984420776367, "learning_rate": 9.591474245115454e-05, "loss": 0.4563, "step": 2353 }, { "epoch": 1.5669828590447663, "grad_norm": 27.241857528686523, "learning_rate": 9.587033747779753e-05, "loss": 0.1023, "step": 2354 }, { "epoch": 1.5676485272091862, "grad_norm": 133.38938903808594, "learning_rate": 9.58259325044405e-05, "loss": 0.5456, "step": 2355 }, { "epoch": 1.5683141953736062, "grad_norm": 35.35673141479492, "learning_rate": 9.578152753108348e-05, "loss": 0.3013, "step": 2356 }, { "epoch": 1.5689798635380263, "grad_norm": 101.0013427734375, "learning_rate": 9.573712255772647e-05, "loss": 0.6429, "step": 2357 }, { "epoch": 1.5696455317024465, "grad_norm": 37.864131927490234, "learning_rate": 9.569271758436945e-05, "loss": 0.1306, "step": 2358 }, { "epoch": 1.5703111998668664, "grad_norm": 17.22524070739746, "learning_rate": 9.564831261101244e-05, "loss": 0.0315, "step": 2359 }, { "epoch": 1.5709768680312863, "grad_norm": 84.59060668945312, "learning_rate": 9.560390763765542e-05, "loss": 0.367, "step": 2360 }, { "epoch": 1.5716425361957065, "grad_norm": 41.82799530029297, "learning_rate": 9.555950266429841e-05, "loss": 0.106, "step": 2361 }, { "epoch": 1.5723082043601266, "grad_norm": 1.6580352783203125, "learning_rate": 9.551509769094139e-05, "loss": 0.0046, "step": 2362 }, { "epoch": 1.5729738725245466, "grad_norm": 14.910322189331055, "learning_rate": 9.547069271758438e-05, "loss": 0.0452, "step": 2363 }, { "epoch": 1.5736395406889665, "grad_norm": 40.421817779541016, "learning_rate": 9.542628774422736e-05, "loss": 0.1751, "step": 2364 }, { "epoch": 1.5743052088533864, "grad_norm": 29.022422790527344, "learning_rate": 9.538188277087035e-05, "loss": 0.1515, "step": 2365 }, { "epoch": 1.5749708770178066, "grad_norm": 55.24034118652344, "learning_rate": 9.533747779751333e-05, "loss": 0.7863, "step": 2366 }, { "epoch": 1.5756365451822267, "grad_norm": 51.957496643066406, "learning_rate": 9.529307282415632e-05, "loss": 0.2734, "step": 2367 }, { "epoch": 1.5763022133466467, "grad_norm": 6.176498889923096, "learning_rate": 9.524866785079929e-05, "loss": 0.0133, "step": 2368 }, { "epoch": 1.5769678815110666, "grad_norm": 74.96781158447266, "learning_rate": 9.520426287744227e-05, "loss": 0.397, "step": 2369 }, { "epoch": 1.5776335496754867, "grad_norm": 23.87809181213379, "learning_rate": 9.515985790408526e-05, "loss": 0.1424, "step": 2370 }, { "epoch": 1.578299217839907, "grad_norm": 10.879682540893555, "learning_rate": 9.511545293072824e-05, "loss": 0.0246, "step": 2371 }, { "epoch": 1.5789648860043268, "grad_norm": 40.713321685791016, "learning_rate": 9.507104795737123e-05, "loss": 0.1568, "step": 2372 }, { "epoch": 1.5796305541687468, "grad_norm": 5.046321392059326, "learning_rate": 9.502664298401421e-05, "loss": 0.0091, "step": 2373 }, { "epoch": 1.580296222333167, "grad_norm": 60.345706939697266, "learning_rate": 9.49822380106572e-05, "loss": 0.6399, "step": 2374 }, { "epoch": 1.580961890497587, "grad_norm": 25.132938385009766, "learning_rate": 9.493783303730018e-05, "loss": 0.0535, "step": 2375 }, { "epoch": 1.581627558662007, "grad_norm": 76.7105712890625, "learning_rate": 9.489342806394317e-05, "loss": 0.4513, "step": 2376 }, { "epoch": 1.582293226826427, "grad_norm": 53.98052978515625, "learning_rate": 9.484902309058615e-05, "loss": 0.4573, "step": 2377 }, { "epoch": 1.582958894990847, "grad_norm": 88.8341293334961, "learning_rate": 9.480461811722914e-05, "loss": 0.8974, "step": 2378 }, { "epoch": 1.5836245631552672, "grad_norm": 34.29758071899414, "learning_rate": 9.476021314387212e-05, "loss": 0.1487, "step": 2379 }, { "epoch": 1.5842902313196872, "grad_norm": 24.019878387451172, "learning_rate": 9.471580817051509e-05, "loss": 0.0445, "step": 2380 }, { "epoch": 1.584955899484107, "grad_norm": 94.3261489868164, "learning_rate": 9.467140319715808e-05, "loss": 0.8572, "step": 2381 }, { "epoch": 1.5856215676485272, "grad_norm": 25.140283584594727, "learning_rate": 9.462699822380106e-05, "loss": 0.1028, "step": 2382 }, { "epoch": 1.5862872358129474, "grad_norm": 7.92868185043335, "learning_rate": 9.458259325044405e-05, "loss": 0.0218, "step": 2383 }, { "epoch": 1.5869529039773673, "grad_norm": 29.709585189819336, "learning_rate": 9.453818827708703e-05, "loss": 0.0792, "step": 2384 }, { "epoch": 1.5876185721417873, "grad_norm": 29.581445693969727, "learning_rate": 9.449378330373002e-05, "loss": 0.1741, "step": 2385 }, { "epoch": 1.5882842403062074, "grad_norm": 73.32685852050781, "learning_rate": 9.4449378330373e-05, "loss": 0.6924, "step": 2386 }, { "epoch": 1.5889499084706273, "grad_norm": 43.63706970214844, "learning_rate": 9.440497335701599e-05, "loss": 0.3174, "step": 2387 }, { "epoch": 1.5896155766350475, "grad_norm": 31.46158218383789, "learning_rate": 9.436056838365897e-05, "loss": 0.0643, "step": 2388 }, { "epoch": 1.5902812447994674, "grad_norm": 33.30908203125, "learning_rate": 9.431616341030196e-05, "loss": 0.1377, "step": 2389 }, { "epoch": 1.5909469129638873, "grad_norm": 17.07303810119629, "learning_rate": 9.427175843694494e-05, "loss": 0.0487, "step": 2390 }, { "epoch": 1.5916125811283075, "grad_norm": 19.82699203491211, "learning_rate": 9.422735346358793e-05, "loss": 0.0577, "step": 2391 }, { "epoch": 1.5922782492927277, "grad_norm": 56.2745361328125, "learning_rate": 9.418294849023091e-05, "loss": 0.3057, "step": 2392 }, { "epoch": 1.5929439174571476, "grad_norm": 45.65102005004883, "learning_rate": 9.41385435168739e-05, "loss": 0.2853, "step": 2393 }, { "epoch": 1.5936095856215675, "grad_norm": 4.959023952484131, "learning_rate": 9.409413854351688e-05, "loss": 0.0075, "step": 2394 }, { "epoch": 1.5942752537859877, "grad_norm": 42.66277313232422, "learning_rate": 9.404973357015986e-05, "loss": 0.5544, "step": 2395 }, { "epoch": 1.5949409219504078, "grad_norm": 5.568058490753174, "learning_rate": 9.400532859680285e-05, "loss": 0.0084, "step": 2396 }, { "epoch": 1.5956065901148277, "grad_norm": 43.342002868652344, "learning_rate": 9.396092362344583e-05, "loss": 0.4109, "step": 2397 }, { "epoch": 1.5962722582792477, "grad_norm": 29.80427360534668, "learning_rate": 9.391651865008882e-05, "loss": 0.1355, "step": 2398 }, { "epoch": 1.5969379264436678, "grad_norm": 11.413454055786133, "learning_rate": 9.38721136767318e-05, "loss": 0.0205, "step": 2399 }, { "epoch": 1.597603594608088, "grad_norm": 50.72407531738281, "learning_rate": 9.382770870337479e-05, "loss": 0.5267, "step": 2400 }, { "epoch": 1.598269262772508, "grad_norm": 27.069740295410156, "learning_rate": 9.378330373001777e-05, "loss": 0.0298, "step": 2401 }, { "epoch": 1.5989349309369278, "grad_norm": 52.162109375, "learning_rate": 9.373889875666076e-05, "loss": 0.4354, "step": 2402 }, { "epoch": 1.599600599101348, "grad_norm": 107.22039031982422, "learning_rate": 9.369449378330374e-05, "loss": 0.8869, "step": 2403 }, { "epoch": 1.6002662672657681, "grad_norm": 50.54841232299805, "learning_rate": 9.365008880994673e-05, "loss": 0.3579, "step": 2404 }, { "epoch": 1.600931935430188, "grad_norm": 81.49309539794922, "learning_rate": 9.36056838365897e-05, "loss": 0.7586, "step": 2405 }, { "epoch": 1.601597603594608, "grad_norm": 33.493980407714844, "learning_rate": 9.356127886323268e-05, "loss": 0.1003, "step": 2406 }, { "epoch": 1.6022632717590282, "grad_norm": 42.038543701171875, "learning_rate": 9.351687388987567e-05, "loss": 0.4266, "step": 2407 }, { "epoch": 1.6029289399234483, "grad_norm": 37.18623733520508, "learning_rate": 9.347246891651865e-05, "loss": 0.0783, "step": 2408 }, { "epoch": 1.6035946080878682, "grad_norm": 9.788639068603516, "learning_rate": 9.342806394316164e-05, "loss": 0.0294, "step": 2409 }, { "epoch": 1.6042602762522882, "grad_norm": 13.086716651916504, "learning_rate": 9.338365896980462e-05, "loss": 0.0446, "step": 2410 }, { "epoch": 1.604925944416708, "grad_norm": 15.407952308654785, "learning_rate": 9.333925399644761e-05, "loss": 0.0533, "step": 2411 }, { "epoch": 1.6055916125811283, "grad_norm": 45.89344024658203, "learning_rate": 9.329484902309059e-05, "loss": 0.1718, "step": 2412 }, { "epoch": 1.6062572807455484, "grad_norm": 36.40590286254883, "learning_rate": 9.325044404973358e-05, "loss": 0.4461, "step": 2413 }, { "epoch": 1.6069229489099683, "grad_norm": 71.3982162475586, "learning_rate": 9.320603907637656e-05, "loss": 0.5587, "step": 2414 }, { "epoch": 1.6075886170743883, "grad_norm": 57.528446197509766, "learning_rate": 9.316163410301955e-05, "loss": 0.7088, "step": 2415 }, { "epoch": 1.6082542852388084, "grad_norm": 61.17235565185547, "learning_rate": 9.311722912966253e-05, "loss": 0.3918, "step": 2416 }, { "epoch": 1.6089199534032286, "grad_norm": 20.595144271850586, "learning_rate": 9.307282415630552e-05, "loss": 0.0845, "step": 2417 }, { "epoch": 1.6095856215676485, "grad_norm": 34.503814697265625, "learning_rate": 9.302841918294849e-05, "loss": 0.2511, "step": 2418 }, { "epoch": 1.6102512897320684, "grad_norm": 16.020893096923828, "learning_rate": 9.298401420959147e-05, "loss": 0.0511, "step": 2419 }, { "epoch": 1.6109169578964886, "grad_norm": 20.62986183166504, "learning_rate": 9.293960923623446e-05, "loss": 0.0779, "step": 2420 }, { "epoch": 1.6115826260609087, "grad_norm": 70.81784057617188, "learning_rate": 9.289520426287744e-05, "loss": 0.677, "step": 2421 }, { "epoch": 1.6122482942253287, "grad_norm": 19.858976364135742, "learning_rate": 9.285079928952043e-05, "loss": 0.1557, "step": 2422 }, { "epoch": 1.6129139623897486, "grad_norm": 32.1212272644043, "learning_rate": 9.280639431616341e-05, "loss": 0.2223, "step": 2423 }, { "epoch": 1.6135796305541688, "grad_norm": 53.579811096191406, "learning_rate": 9.27619893428064e-05, "loss": 0.3631, "step": 2424 }, { "epoch": 1.614245298718589, "grad_norm": 105.37787628173828, "learning_rate": 9.271758436944938e-05, "loss": 0.2168, "step": 2425 }, { "epoch": 1.6149109668830088, "grad_norm": 33.3464469909668, "learning_rate": 9.267317939609237e-05, "loss": 0.2062, "step": 2426 }, { "epoch": 1.6155766350474288, "grad_norm": 51.30805587768555, "learning_rate": 9.262877442273535e-05, "loss": 0.2353, "step": 2427 }, { "epoch": 1.616242303211849, "grad_norm": 28.053707122802734, "learning_rate": 9.258436944937834e-05, "loss": 0.1217, "step": 2428 }, { "epoch": 1.616907971376269, "grad_norm": 25.779495239257812, "learning_rate": 9.253996447602132e-05, "loss": 0.0497, "step": 2429 }, { "epoch": 1.617573639540689, "grad_norm": 50.07307815551758, "learning_rate": 9.249555950266429e-05, "loss": 0.2718, "step": 2430 }, { "epoch": 1.618239307705109, "grad_norm": 70.97674560546875, "learning_rate": 9.245115452930728e-05, "loss": 1.2455, "step": 2431 }, { "epoch": 1.618904975869529, "grad_norm": 44.86273193359375, "learning_rate": 9.240674955595026e-05, "loss": 0.2678, "step": 2432 }, { "epoch": 1.6195706440339492, "grad_norm": 25.8091983795166, "learning_rate": 9.236234458259325e-05, "loss": 0.0959, "step": 2433 }, { "epoch": 1.6202363121983692, "grad_norm": 51.438785552978516, "learning_rate": 9.231793960923623e-05, "loss": 0.2444, "step": 2434 }, { "epoch": 1.620901980362789, "grad_norm": 40.81222915649414, "learning_rate": 9.227353463587922e-05, "loss": 0.2238, "step": 2435 }, { "epoch": 1.621567648527209, "grad_norm": 45.644371032714844, "learning_rate": 9.22291296625222e-05, "loss": 0.3491, "step": 2436 }, { "epoch": 1.6222333166916292, "grad_norm": 25.918373107910156, "learning_rate": 9.218472468916519e-05, "loss": 0.1893, "step": 2437 }, { "epoch": 1.6228989848560493, "grad_norm": 29.952579498291016, "learning_rate": 9.214031971580817e-05, "loss": 0.1154, "step": 2438 }, { "epoch": 1.6235646530204693, "grad_norm": 37.095062255859375, "learning_rate": 9.209591474245116e-05, "loss": 0.1783, "step": 2439 }, { "epoch": 1.6242303211848892, "grad_norm": 6.809981822967529, "learning_rate": 9.205150976909414e-05, "loss": 0.0218, "step": 2440 }, { "epoch": 1.6248959893493093, "grad_norm": 31.979650497436523, "learning_rate": 9.200710479573713e-05, "loss": 0.1733, "step": 2441 }, { "epoch": 1.6255616575137295, "grad_norm": 56.26057052612305, "learning_rate": 9.196269982238011e-05, "loss": 0.1622, "step": 2442 }, { "epoch": 1.6262273256781494, "grad_norm": 39.30649185180664, "learning_rate": 9.19182948490231e-05, "loss": 0.3259, "step": 2443 }, { "epoch": 1.6268929938425694, "grad_norm": 5.297876834869385, "learning_rate": 9.187388987566608e-05, "loss": 0.0343, "step": 2444 }, { "epoch": 1.6275586620069895, "grad_norm": 44.93223571777344, "learning_rate": 9.182948490230907e-05, "loss": 0.3079, "step": 2445 }, { "epoch": 1.6282243301714097, "grad_norm": 17.083660125732422, "learning_rate": 9.178507992895205e-05, "loss": 0.0348, "step": 2446 }, { "epoch": 1.6288899983358296, "grad_norm": 2.987624168395996, "learning_rate": 9.174067495559504e-05, "loss": 0.0118, "step": 2447 }, { "epoch": 1.6295556665002495, "grad_norm": 41.67808532714844, "learning_rate": 9.169626998223802e-05, "loss": 0.431, "step": 2448 }, { "epoch": 1.6302213346646697, "grad_norm": 17.668418884277344, "learning_rate": 9.1651865008881e-05, "loss": 0.0483, "step": 2449 }, { "epoch": 1.6308870028290898, "grad_norm": 21.337465286254883, "learning_rate": 9.160746003552399e-05, "loss": 0.0569, "step": 2450 }, { "epoch": 1.6315526709935098, "grad_norm": 27.76921272277832, "learning_rate": 9.156305506216697e-05, "loss": 0.0879, "step": 2451 }, { "epoch": 1.6322183391579297, "grad_norm": 43.2540283203125, "learning_rate": 9.151865008880996e-05, "loss": 0.1402, "step": 2452 }, { "epoch": 1.6328840073223498, "grad_norm": 25.035865783691406, "learning_rate": 9.147424511545294e-05, "loss": 0.1343, "step": 2453 }, { "epoch": 1.63354967548677, "grad_norm": 61.47492599487305, "learning_rate": 9.142984014209593e-05, "loss": 0.2878, "step": 2454 }, { "epoch": 1.63421534365119, "grad_norm": 8.469744682312012, "learning_rate": 9.13854351687389e-05, "loss": 0.0244, "step": 2455 }, { "epoch": 1.6348810118156099, "grad_norm": 100.8773422241211, "learning_rate": 9.134103019538189e-05, "loss": 0.9253, "step": 2456 }, { "epoch": 1.63554667998003, "grad_norm": 27.50389289855957, "learning_rate": 9.129662522202487e-05, "loss": 0.0563, "step": 2457 }, { "epoch": 1.6362123481444502, "grad_norm": 81.96160125732422, "learning_rate": 9.125222024866785e-05, "loss": 0.2955, "step": 2458 }, { "epoch": 1.63687801630887, "grad_norm": 0.5598182082176208, "learning_rate": 9.120781527531084e-05, "loss": 0.0017, "step": 2459 }, { "epoch": 1.63754368447329, "grad_norm": 70.70709991455078, "learning_rate": 9.116341030195382e-05, "loss": 0.0678, "step": 2460 }, { "epoch": 1.63820935263771, "grad_norm": 33.05805206298828, "learning_rate": 9.111900532859681e-05, "loss": 0.0825, "step": 2461 }, { "epoch": 1.63887502080213, "grad_norm": 96.8099594116211, "learning_rate": 9.10746003552398e-05, "loss": 0.5339, "step": 2462 }, { "epoch": 1.6395406889665503, "grad_norm": 43.461795806884766, "learning_rate": 9.103019538188278e-05, "loss": 0.2964, "step": 2463 }, { "epoch": 1.6402063571309702, "grad_norm": 63.74673843383789, "learning_rate": 9.098579040852576e-05, "loss": 0.1413, "step": 2464 }, { "epoch": 1.6408720252953901, "grad_norm": 0.5713216066360474, "learning_rate": 9.094138543516875e-05, "loss": 0.0017, "step": 2465 }, { "epoch": 1.6415376934598103, "grad_norm": 1.243265986442566, "learning_rate": 9.089698046181173e-05, "loss": 0.0034, "step": 2466 }, { "epoch": 1.6422033616242304, "grad_norm": 84.5959701538086, "learning_rate": 9.085257548845472e-05, "loss": 0.9577, "step": 2467 }, { "epoch": 1.6428690297886503, "grad_norm": 40.408966064453125, "learning_rate": 9.080817051509769e-05, "loss": 0.3578, "step": 2468 }, { "epoch": 1.6435346979530703, "grad_norm": 21.698341369628906, "learning_rate": 9.076376554174067e-05, "loss": 0.0545, "step": 2469 }, { "epoch": 1.6442003661174904, "grad_norm": 43.9974479675293, "learning_rate": 9.071936056838366e-05, "loss": 0.164, "step": 2470 }, { "epoch": 1.6448660342819106, "grad_norm": 31.4620361328125, "learning_rate": 9.067495559502664e-05, "loss": 0.1718, "step": 2471 }, { "epoch": 1.6455317024463305, "grad_norm": 36.63017654418945, "learning_rate": 9.063055062166963e-05, "loss": 0.3096, "step": 2472 }, { "epoch": 1.6461973706107504, "grad_norm": 49.87548828125, "learning_rate": 9.058614564831261e-05, "loss": 0.2474, "step": 2473 }, { "epoch": 1.6468630387751706, "grad_norm": 31.68560791015625, "learning_rate": 9.05417406749556e-05, "loss": 0.1009, "step": 2474 }, { "epoch": 1.6475287069395907, "grad_norm": 43.38981628417969, "learning_rate": 9.049733570159858e-05, "loss": 0.237, "step": 2475 }, { "epoch": 1.6481943751040107, "grad_norm": 8.770151138305664, "learning_rate": 9.045293072824157e-05, "loss": 0.0173, "step": 2476 }, { "epoch": 1.6488600432684306, "grad_norm": 63.738582611083984, "learning_rate": 9.040852575488455e-05, "loss": 0.3641, "step": 2477 }, { "epoch": 1.6495257114328508, "grad_norm": 34.64439010620117, "learning_rate": 9.036412078152754e-05, "loss": 0.1135, "step": 2478 }, { "epoch": 1.650191379597271, "grad_norm": 130.81285095214844, "learning_rate": 9.031971580817052e-05, "loss": 0.4412, "step": 2479 }, { "epoch": 1.6508570477616908, "grad_norm": 49.69594955444336, "learning_rate": 9.02753108348135e-05, "loss": 0.1297, "step": 2480 }, { "epoch": 1.6515227159261108, "grad_norm": 50.79115676879883, "learning_rate": 9.023090586145648e-05, "loss": 0.1415, "step": 2481 }, { "epoch": 1.6521883840905307, "grad_norm": 39.11735534667969, "learning_rate": 9.018650088809946e-05, "loss": 0.1487, "step": 2482 }, { "epoch": 1.6528540522549509, "grad_norm": 6.8349833488464355, "learning_rate": 9.014209591474245e-05, "loss": 0.0163, "step": 2483 }, { "epoch": 1.653519720419371, "grad_norm": 57.06932067871094, "learning_rate": 9.009769094138543e-05, "loss": 0.1804, "step": 2484 }, { "epoch": 1.654185388583791, "grad_norm": 82.49613189697266, "learning_rate": 9.005328596802842e-05, "loss": 0.1611, "step": 2485 }, { "epoch": 1.6548510567482109, "grad_norm": 17.83732795715332, "learning_rate": 9.00088809946714e-05, "loss": 0.0441, "step": 2486 }, { "epoch": 1.655516724912631, "grad_norm": 46.61641311645508, "learning_rate": 8.996447602131439e-05, "loss": 0.1636, "step": 2487 }, { "epoch": 1.6561823930770512, "grad_norm": 9.754440307617188, "learning_rate": 8.992007104795737e-05, "loss": 0.0268, "step": 2488 }, { "epoch": 1.656848061241471, "grad_norm": 80.84696960449219, "learning_rate": 8.987566607460036e-05, "loss": 0.267, "step": 2489 }, { "epoch": 1.657513729405891, "grad_norm": 63.48737335205078, "learning_rate": 8.983126110124334e-05, "loss": 0.4447, "step": 2490 }, { "epoch": 1.6581793975703112, "grad_norm": 60.9344482421875, "learning_rate": 8.978685612788633e-05, "loss": 0.5781, "step": 2491 }, { "epoch": 1.6588450657347313, "grad_norm": 3.709758996963501, "learning_rate": 8.974245115452931e-05, "loss": 0.0067, "step": 2492 }, { "epoch": 1.6595107338991513, "grad_norm": 61.038204193115234, "learning_rate": 8.96980461811723e-05, "loss": 0.3445, "step": 2493 }, { "epoch": 1.6601764020635712, "grad_norm": 42.08723831176758, "learning_rate": 8.965364120781528e-05, "loss": 0.2199, "step": 2494 }, { "epoch": 1.6608420702279914, "grad_norm": 41.67546463012695, "learning_rate": 8.960923623445827e-05, "loss": 0.1738, "step": 2495 }, { "epoch": 1.6615077383924115, "grad_norm": 103.30386352539062, "learning_rate": 8.956483126110125e-05, "loss": 1.2187, "step": 2496 }, { "epoch": 1.6621734065568314, "grad_norm": 62.8175048828125, "learning_rate": 8.952042628774424e-05, "loss": 0.6841, "step": 2497 }, { "epoch": 1.6628390747212514, "grad_norm": 110.06562042236328, "learning_rate": 8.947602131438722e-05, "loss": 0.5955, "step": 2498 }, { "epoch": 1.6635047428856715, "grad_norm": 16.987764358520508, "learning_rate": 8.94316163410302e-05, "loss": 0.0285, "step": 2499 }, { "epoch": 1.6641704110500917, "grad_norm": 41.673072814941406, "learning_rate": 8.938721136767319e-05, "loss": 0.1793, "step": 2500 }, { "epoch": 1.6648360792145116, "grad_norm": 35.432098388671875, "learning_rate": 8.934280639431618e-05, "loss": 0.125, "step": 2501 }, { "epoch": 1.6655017473789315, "grad_norm": 9.721442222595215, "learning_rate": 8.929840142095916e-05, "loss": 0.0164, "step": 2502 }, { "epoch": 1.6661674155433517, "grad_norm": 55.195499420166016, "learning_rate": 8.925399644760215e-05, "loss": 0.199, "step": 2503 }, { "epoch": 1.6668330837077718, "grad_norm": 14.33546257019043, "learning_rate": 8.920959147424513e-05, "loss": 0.0301, "step": 2504 }, { "epoch": 1.6674987518721918, "grad_norm": 45.84988021850586, "learning_rate": 8.91651865008881e-05, "loss": 0.1483, "step": 2505 }, { "epoch": 1.6681644200366117, "grad_norm": 22.817310333251953, "learning_rate": 8.912078152753109e-05, "loss": 0.0552, "step": 2506 }, { "epoch": 1.6688300882010316, "grad_norm": 39.06748580932617, "learning_rate": 8.907637655417407e-05, "loss": 0.4904, "step": 2507 }, { "epoch": 1.6694957563654518, "grad_norm": 30.82869529724121, "learning_rate": 8.903197158081706e-05, "loss": 0.0256, "step": 2508 }, { "epoch": 1.670161424529872, "grad_norm": 33.30055236816406, "learning_rate": 8.898756660746004e-05, "loss": 0.0752, "step": 2509 }, { "epoch": 1.6708270926942919, "grad_norm": 6.687162399291992, "learning_rate": 8.894316163410303e-05, "loss": 0.008, "step": 2510 }, { "epoch": 1.6714927608587118, "grad_norm": 5.699782848358154, "learning_rate": 8.889875666074601e-05, "loss": 0.0073, "step": 2511 }, { "epoch": 1.672158429023132, "grad_norm": 41.28861999511719, "learning_rate": 8.8854351687389e-05, "loss": 0.1698, "step": 2512 }, { "epoch": 1.672824097187552, "grad_norm": 45.939552307128906, "learning_rate": 8.880994671403198e-05, "loss": 0.2492, "step": 2513 }, { "epoch": 1.673489765351972, "grad_norm": 20.018022537231445, "learning_rate": 8.876554174067497e-05, "loss": 0.0513, "step": 2514 }, { "epoch": 1.674155433516392, "grad_norm": 118.777587890625, "learning_rate": 8.872113676731795e-05, "loss": 0.437, "step": 2515 }, { "epoch": 1.674821101680812, "grad_norm": 111.73308563232422, "learning_rate": 8.867673179396093e-05, "loss": 0.7436, "step": 2516 }, { "epoch": 1.6754867698452323, "grad_norm": 28.79439353942871, "learning_rate": 8.863232682060392e-05, "loss": 0.1903, "step": 2517 }, { "epoch": 1.6761524380096522, "grad_norm": 0.7686096429824829, "learning_rate": 8.858792184724689e-05, "loss": 0.0016, "step": 2518 }, { "epoch": 1.6768181061740721, "grad_norm": 38.94993209838867, "learning_rate": 8.854351687388988e-05, "loss": 0.0731, "step": 2519 }, { "epoch": 1.6774837743384923, "grad_norm": 69.95482635498047, "learning_rate": 8.849911190053286e-05, "loss": 1.3336, "step": 2520 }, { "epoch": 1.6781494425029124, "grad_norm": 22.42783546447754, "learning_rate": 8.845470692717585e-05, "loss": 0.0608, "step": 2521 }, { "epoch": 1.6788151106673324, "grad_norm": 22.973590850830078, "learning_rate": 8.841030195381883e-05, "loss": 0.0763, "step": 2522 }, { "epoch": 1.6794807788317523, "grad_norm": 20.757122039794922, "learning_rate": 8.836589698046182e-05, "loss": 0.0698, "step": 2523 }, { "epoch": 1.6801464469961724, "grad_norm": 21.058683395385742, "learning_rate": 8.83214920071048e-05, "loss": 0.049, "step": 2524 }, { "epoch": 1.6808121151605926, "grad_norm": 78.88239288330078, "learning_rate": 8.827708703374778e-05, "loss": 0.4674, "step": 2525 }, { "epoch": 1.6814777833250125, "grad_norm": 53.14613723754883, "learning_rate": 8.823268206039077e-05, "loss": 0.1323, "step": 2526 }, { "epoch": 1.6821434514894325, "grad_norm": 25.038597106933594, "learning_rate": 8.818827708703375e-05, "loss": 0.0857, "step": 2527 }, { "epoch": 1.6828091196538526, "grad_norm": 68.05943298339844, "learning_rate": 8.814387211367674e-05, "loss": 0.2925, "step": 2528 }, { "epoch": 1.6834747878182728, "grad_norm": 23.949832916259766, "learning_rate": 8.809946714031972e-05, "loss": 0.0899, "step": 2529 }, { "epoch": 1.6841404559826927, "grad_norm": 13.184732437133789, "learning_rate": 8.80550621669627e-05, "loss": 0.0357, "step": 2530 }, { "epoch": 1.6848061241471126, "grad_norm": 43.132545471191406, "learning_rate": 8.801065719360568e-05, "loss": 0.1944, "step": 2531 }, { "epoch": 1.6854717923115325, "grad_norm": 59.569637298583984, "learning_rate": 8.796625222024867e-05, "loss": 0.2107, "step": 2532 }, { "epoch": 1.6861374604759527, "grad_norm": 31.548635482788086, "learning_rate": 8.792184724689165e-05, "loss": 0.2095, "step": 2533 }, { "epoch": 1.6868031286403729, "grad_norm": 44.94803237915039, "learning_rate": 8.787744227353463e-05, "loss": 0.1491, "step": 2534 }, { "epoch": 1.6874687968047928, "grad_norm": 76.41862487792969, "learning_rate": 8.783303730017762e-05, "loss": 0.3286, "step": 2535 }, { "epoch": 1.6881344649692127, "grad_norm": 34.66581344604492, "learning_rate": 8.77886323268206e-05, "loss": 0.2722, "step": 2536 }, { "epoch": 1.6888001331336329, "grad_norm": 60.362754821777344, "learning_rate": 8.774422735346359e-05, "loss": 0.8262, "step": 2537 }, { "epoch": 1.689465801298053, "grad_norm": 89.76812744140625, "learning_rate": 8.769982238010657e-05, "loss": 0.6926, "step": 2538 }, { "epoch": 1.690131469462473, "grad_norm": 81.87251281738281, "learning_rate": 8.765541740674956e-05, "loss": 0.2244, "step": 2539 }, { "epoch": 1.6907971376268929, "grad_norm": 3.8928475379943848, "learning_rate": 8.761101243339254e-05, "loss": 0.0108, "step": 2540 }, { "epoch": 1.691462805791313, "grad_norm": 43.844364166259766, "learning_rate": 8.756660746003553e-05, "loss": 0.1803, "step": 2541 }, { "epoch": 1.6921284739557332, "grad_norm": 102.1588363647461, "learning_rate": 8.752220248667851e-05, "loss": 0.754, "step": 2542 }, { "epoch": 1.6927941421201531, "grad_norm": 10.578431129455566, "learning_rate": 8.747779751332148e-05, "loss": 0.0416, "step": 2543 }, { "epoch": 1.693459810284573, "grad_norm": 67.71282958984375, "learning_rate": 8.743339253996447e-05, "loss": 0.4502, "step": 2544 }, { "epoch": 1.6941254784489932, "grad_norm": 37.609004974365234, "learning_rate": 8.738898756660745e-05, "loss": 0.2512, "step": 2545 }, { "epoch": 1.6947911466134133, "grad_norm": 76.71928405761719, "learning_rate": 8.734458259325044e-05, "loss": 0.496, "step": 2546 }, { "epoch": 1.6954568147778333, "grad_norm": 61.35820770263672, "learning_rate": 8.730017761989342e-05, "loss": 0.2788, "step": 2547 }, { "epoch": 1.6961224829422532, "grad_norm": 3.2343103885650635, "learning_rate": 8.725577264653641e-05, "loss": 0.0063, "step": 2548 }, { "epoch": 1.6967881511066734, "grad_norm": 70.34629821777344, "learning_rate": 8.72113676731794e-05, "loss": 0.3065, "step": 2549 }, { "epoch": 1.6974538192710935, "grad_norm": 38.027793884277344, "learning_rate": 8.716696269982239e-05, "loss": 0.1134, "step": 2550 }, { "epoch": 1.6981194874355134, "grad_norm": 36.3894157409668, "learning_rate": 8.712255772646538e-05, "loss": 0.1924, "step": 2551 }, { "epoch": 1.6987851555999334, "grad_norm": 6.59135103225708, "learning_rate": 8.707815275310836e-05, "loss": 0.0192, "step": 2552 }, { "epoch": 1.6994508237643535, "grad_norm": 3.3366539478302, "learning_rate": 8.703374777975135e-05, "loss": 0.0068, "step": 2553 }, { "epoch": 1.7001164919287735, "grad_norm": 16.128673553466797, "learning_rate": 8.698934280639433e-05, "loss": 0.0621, "step": 2554 }, { "epoch": 1.7007821600931936, "grad_norm": 30.56314468383789, "learning_rate": 8.694493783303732e-05, "loss": 0.1178, "step": 2555 }, { "epoch": 1.7014478282576135, "grad_norm": 33.746158599853516, "learning_rate": 8.690053285968029e-05, "loss": 0.1578, "step": 2556 }, { "epoch": 1.7021134964220335, "grad_norm": 6.771616458892822, "learning_rate": 8.685612788632327e-05, "loss": 0.0086, "step": 2557 }, { "epoch": 1.7027791645864536, "grad_norm": 30.768205642700195, "learning_rate": 8.681172291296626e-05, "loss": 0.0447, "step": 2558 }, { "epoch": 1.7034448327508738, "grad_norm": 33.8901481628418, "learning_rate": 8.676731793960924e-05, "loss": 0.2778, "step": 2559 }, { "epoch": 1.7041105009152937, "grad_norm": 58.39024353027344, "learning_rate": 8.672291296625223e-05, "loss": 0.2569, "step": 2560 }, { "epoch": 1.7047761690797136, "grad_norm": 38.764251708984375, "learning_rate": 8.667850799289521e-05, "loss": 0.2971, "step": 2561 }, { "epoch": 1.7054418372441338, "grad_norm": 91.56017303466797, "learning_rate": 8.66341030195382e-05, "loss": 0.1299, "step": 2562 }, { "epoch": 1.706107505408554, "grad_norm": 38.783287048339844, "learning_rate": 8.658969804618118e-05, "loss": 0.24, "step": 2563 }, { "epoch": 1.7067731735729739, "grad_norm": 37.98073196411133, "learning_rate": 8.654529307282417e-05, "loss": 0.1325, "step": 2564 }, { "epoch": 1.7074388417373938, "grad_norm": 35.608394622802734, "learning_rate": 8.650088809946715e-05, "loss": 0.3814, "step": 2565 }, { "epoch": 1.708104509901814, "grad_norm": 32.38398742675781, "learning_rate": 8.645648312611014e-05, "loss": 0.0902, "step": 2566 }, { "epoch": 1.708770178066234, "grad_norm": 0.15740056335926056, "learning_rate": 8.641207815275312e-05, "loss": 0.0005, "step": 2567 }, { "epoch": 1.709435846230654, "grad_norm": 32.83389663696289, "learning_rate": 8.636767317939609e-05, "loss": 0.2976, "step": 2568 }, { "epoch": 1.710101514395074, "grad_norm": 45.68979263305664, "learning_rate": 8.632326820603908e-05, "loss": 0.5236, "step": 2569 }, { "epoch": 1.7107671825594941, "grad_norm": 36.86159896850586, "learning_rate": 8.627886323268206e-05, "loss": 0.1986, "step": 2570 }, { "epoch": 1.7114328507239143, "grad_norm": 3.291645050048828, "learning_rate": 8.623445825932505e-05, "loss": 0.0132, "step": 2571 }, { "epoch": 1.7120985188883342, "grad_norm": 18.64942169189453, "learning_rate": 8.619005328596803e-05, "loss": 0.0419, "step": 2572 }, { "epoch": 1.7127641870527541, "grad_norm": 66.5919189453125, "learning_rate": 8.614564831261102e-05, "loss": 0.1338, "step": 2573 }, { "epoch": 1.7134298552171743, "grad_norm": 40.664485931396484, "learning_rate": 8.6101243339254e-05, "loss": 0.2854, "step": 2574 }, { "epoch": 1.7140955233815944, "grad_norm": 41.47353744506836, "learning_rate": 8.605683836589699e-05, "loss": 0.2047, "step": 2575 }, { "epoch": 1.7147611915460144, "grad_norm": 59.07371520996094, "learning_rate": 8.601243339253997e-05, "loss": 0.2697, "step": 2576 }, { "epoch": 1.7154268597104343, "grad_norm": 78.93032836914062, "learning_rate": 8.596802841918296e-05, "loss": 0.7739, "step": 2577 }, { "epoch": 1.7160925278748542, "grad_norm": 28.15635871887207, "learning_rate": 8.592362344582594e-05, "loss": 0.1331, "step": 2578 }, { "epoch": 1.7167581960392744, "grad_norm": 5.510909080505371, "learning_rate": 8.587921847246893e-05, "loss": 0.0141, "step": 2579 }, { "epoch": 1.7174238642036945, "grad_norm": 13.602741241455078, "learning_rate": 8.583481349911191e-05, "loss": 0.0488, "step": 2580 }, { "epoch": 1.7180895323681145, "grad_norm": 6.326352596282959, "learning_rate": 8.579040852575488e-05, "loss": 0.0151, "step": 2581 }, { "epoch": 1.7187552005325344, "grad_norm": 33.634883880615234, "learning_rate": 8.574600355239787e-05, "loss": 0.2486, "step": 2582 }, { "epoch": 1.7194208686969545, "grad_norm": 27.754230499267578, "learning_rate": 8.570159857904085e-05, "loss": 0.0407, "step": 2583 }, { "epoch": 1.7200865368613747, "grad_norm": 19.805042266845703, "learning_rate": 8.565719360568384e-05, "loss": 0.06, "step": 2584 }, { "epoch": 1.7207522050257946, "grad_norm": 37.06386947631836, "learning_rate": 8.561278863232682e-05, "loss": 0.1041, "step": 2585 }, { "epoch": 1.7214178731902146, "grad_norm": 29.32569694519043, "learning_rate": 8.55683836589698e-05, "loss": 0.4845, "step": 2586 }, { "epoch": 1.7220835413546347, "grad_norm": 43.97129440307617, "learning_rate": 8.552397868561279e-05, "loss": 0.1993, "step": 2587 }, { "epoch": 1.7227492095190549, "grad_norm": 85.27297973632812, "learning_rate": 8.547957371225578e-05, "loss": 0.553, "step": 2588 }, { "epoch": 1.7234148776834748, "grad_norm": 66.77306365966797, "learning_rate": 8.543516873889876e-05, "loss": 0.3985, "step": 2589 }, { "epoch": 1.7240805458478947, "grad_norm": 2.720907211303711, "learning_rate": 8.539076376554174e-05, "loss": 0.0056, "step": 2590 }, { "epoch": 1.7247462140123149, "grad_norm": 29.983379364013672, "learning_rate": 8.534635879218473e-05, "loss": 0.0735, "step": 2591 }, { "epoch": 1.725411882176735, "grad_norm": 13.737260818481445, "learning_rate": 8.530195381882771e-05, "loss": 0.0342, "step": 2592 }, { "epoch": 1.726077550341155, "grad_norm": 51.28158187866211, "learning_rate": 8.525754884547069e-05, "loss": 0.4229, "step": 2593 }, { "epoch": 1.7267432185055749, "grad_norm": 5.936929225921631, "learning_rate": 8.521314387211367e-05, "loss": 0.0113, "step": 2594 }, { "epoch": 1.727408886669995, "grad_norm": 64.77667999267578, "learning_rate": 8.516873889875666e-05, "loss": 0.1697, "step": 2595 }, { "epoch": 1.7280745548344152, "grad_norm": 53.67452621459961, "learning_rate": 8.512433392539964e-05, "loss": 0.286, "step": 2596 }, { "epoch": 1.7287402229988351, "grad_norm": 16.58892822265625, "learning_rate": 8.507992895204263e-05, "loss": 0.0546, "step": 2597 }, { "epoch": 1.729405891163255, "grad_norm": 44.14348602294922, "learning_rate": 8.503552397868561e-05, "loss": 0.1119, "step": 2598 }, { "epoch": 1.7300715593276752, "grad_norm": 67.6447525024414, "learning_rate": 8.49911190053286e-05, "loss": 0.6635, "step": 2599 }, { "epoch": 1.7307372274920954, "grad_norm": 26.156444549560547, "learning_rate": 8.494671403197158e-05, "loss": 0.0854, "step": 2600 }, { "epoch": 1.7314028956565153, "grad_norm": 0.3501136302947998, "learning_rate": 8.490230905861456e-05, "loss": 0.001, "step": 2601 }, { "epoch": 1.7320685638209352, "grad_norm": 16.02205467224121, "learning_rate": 8.485790408525755e-05, "loss": 0.0436, "step": 2602 }, { "epoch": 1.7327342319853551, "grad_norm": 66.35966491699219, "learning_rate": 8.481349911190053e-05, "loss": 0.294, "step": 2603 }, { "epoch": 1.7333999001497753, "grad_norm": 68.58609771728516, "learning_rate": 8.476909413854352e-05, "loss": 0.5437, "step": 2604 }, { "epoch": 1.7340655683141954, "grad_norm": 3.855602502822876, "learning_rate": 8.47246891651865e-05, "loss": 0.013, "step": 2605 }, { "epoch": 1.7347312364786154, "grad_norm": 100.3924331665039, "learning_rate": 8.468028419182949e-05, "loss": 0.5743, "step": 2606 }, { "epoch": 1.7353969046430353, "grad_norm": 36.822364807128906, "learning_rate": 8.463587921847247e-05, "loss": 0.0771, "step": 2607 }, { "epoch": 1.7360625728074555, "grad_norm": 49.90605163574219, "learning_rate": 8.459147424511546e-05, "loss": 0.198, "step": 2608 }, { "epoch": 1.7367282409718756, "grad_norm": 31.521530151367188, "learning_rate": 8.454706927175844e-05, "loss": 0.1149, "step": 2609 }, { "epoch": 1.7373939091362955, "grad_norm": 10.793007850646973, "learning_rate": 8.450266429840143e-05, "loss": 0.0205, "step": 2610 }, { "epoch": 1.7380595773007155, "grad_norm": 40.38314437866211, "learning_rate": 8.445825932504441e-05, "loss": 0.096, "step": 2611 }, { "epoch": 1.7387252454651356, "grad_norm": 5.628844738006592, "learning_rate": 8.44138543516874e-05, "loss": 0.0126, "step": 2612 }, { "epoch": 1.7393909136295558, "grad_norm": 38.91952896118164, "learning_rate": 8.436944937833038e-05, "loss": 0.0853, "step": 2613 }, { "epoch": 1.7400565817939757, "grad_norm": 51.839725494384766, "learning_rate": 8.432504440497337e-05, "loss": 0.2771, "step": 2614 }, { "epoch": 1.7407222499583956, "grad_norm": 24.52115821838379, "learning_rate": 8.428063943161635e-05, "loss": 0.0356, "step": 2615 }, { "epoch": 1.7413879181228158, "grad_norm": 17.77857780456543, "learning_rate": 8.423623445825934e-05, "loss": 0.0405, "step": 2616 }, { "epoch": 1.742053586287236, "grad_norm": 14.166539192199707, "learning_rate": 8.419182948490232e-05, "loss": 0.0287, "step": 2617 }, { "epoch": 1.7427192544516559, "grad_norm": 48.33513641357422, "learning_rate": 8.41474245115453e-05, "loss": 0.4281, "step": 2618 }, { "epoch": 1.7433849226160758, "grad_norm": 58.158668518066406, "learning_rate": 8.410301953818828e-05, "loss": 0.4389, "step": 2619 }, { "epoch": 1.744050590780496, "grad_norm": 84.82035827636719, "learning_rate": 8.405861456483126e-05, "loss": 0.8382, "step": 2620 }, { "epoch": 1.744716258944916, "grad_norm": 55.7158317565918, "learning_rate": 8.401420959147425e-05, "loss": 0.1925, "step": 2621 }, { "epoch": 1.745381927109336, "grad_norm": 27.67528533935547, "learning_rate": 8.396980461811723e-05, "loss": 0.2255, "step": 2622 }, { "epoch": 1.746047595273756, "grad_norm": 19.12830352783203, "learning_rate": 8.392539964476022e-05, "loss": 0.0365, "step": 2623 }, { "epoch": 1.7467132634381761, "grad_norm": 43.7209587097168, "learning_rate": 8.38809946714032e-05, "loss": 0.2897, "step": 2624 }, { "epoch": 1.747378931602596, "grad_norm": 12.265998840332031, "learning_rate": 8.383658969804619e-05, "loss": 0.0176, "step": 2625 }, { "epoch": 1.7480445997670162, "grad_norm": 35.182594299316406, "learning_rate": 8.379218472468917e-05, "loss": 0.259, "step": 2626 }, { "epoch": 1.7487102679314361, "grad_norm": 44.20758819580078, "learning_rate": 8.374777975133216e-05, "loss": 0.1749, "step": 2627 }, { "epoch": 1.749375936095856, "grad_norm": 3.4633185863494873, "learning_rate": 8.370337477797514e-05, "loss": 0.0057, "step": 2628 }, { "epoch": 1.7500416042602762, "grad_norm": 57.24038314819336, "learning_rate": 8.365896980461813e-05, "loss": 0.0928, "step": 2629 }, { "epoch": 1.7507072724246964, "grad_norm": 7.958834648132324, "learning_rate": 8.361456483126111e-05, "loss": 0.0194, "step": 2630 }, { "epoch": 1.7513729405891163, "grad_norm": 39.624088287353516, "learning_rate": 8.357015985790408e-05, "loss": 0.2089, "step": 2631 }, { "epoch": 1.7520386087535362, "grad_norm": 150.1725311279297, "learning_rate": 8.352575488454707e-05, "loss": 0.1185, "step": 2632 }, { "epoch": 1.7527042769179564, "grad_norm": 46.39543914794922, "learning_rate": 8.348134991119005e-05, "loss": 0.1412, "step": 2633 }, { "epoch": 1.7533699450823765, "grad_norm": 2.9092674255371094, "learning_rate": 8.343694493783304e-05, "loss": 0.0058, "step": 2634 }, { "epoch": 1.7540356132467965, "grad_norm": 37.30491256713867, "learning_rate": 8.339253996447602e-05, "loss": 0.2701, "step": 2635 }, { "epoch": 1.7547012814112164, "grad_norm": 59.646766662597656, "learning_rate": 8.334813499111901e-05, "loss": 0.6346, "step": 2636 }, { "epoch": 1.7553669495756365, "grad_norm": 2.656698226928711, "learning_rate": 8.330373001776199e-05, "loss": 0.0042, "step": 2637 }, { "epoch": 1.7560326177400567, "grad_norm": 49.6700553894043, "learning_rate": 8.325932504440498e-05, "loss": 0.3215, "step": 2638 }, { "epoch": 1.7566982859044766, "grad_norm": 54.03047180175781, "learning_rate": 8.321492007104796e-05, "loss": 0.4739, "step": 2639 }, { "epoch": 1.7573639540688966, "grad_norm": 44.02018737792969, "learning_rate": 8.317051509769095e-05, "loss": 0.1842, "step": 2640 }, { "epoch": 1.7580296222333167, "grad_norm": 53.248268127441406, "learning_rate": 8.312611012433393e-05, "loss": 0.3283, "step": 2641 }, { "epoch": 1.7586952903977369, "grad_norm": 68.10962677001953, "learning_rate": 8.308170515097692e-05, "loss": 0.7713, "step": 2642 }, { "epoch": 1.7593609585621568, "grad_norm": 33.34562301635742, "learning_rate": 8.303730017761989e-05, "loss": 0.0923, "step": 2643 }, { "epoch": 1.7600266267265767, "grad_norm": 48.30095291137695, "learning_rate": 8.299289520426287e-05, "loss": 0.5184, "step": 2644 }, { "epoch": 1.7606922948909969, "grad_norm": 16.319461822509766, "learning_rate": 8.294849023090586e-05, "loss": 0.0502, "step": 2645 }, { "epoch": 1.761357963055417, "grad_norm": 54.672061920166016, "learning_rate": 8.290408525754884e-05, "loss": 0.8855, "step": 2646 }, { "epoch": 1.762023631219837, "grad_norm": 35.859275817871094, "learning_rate": 8.285968028419183e-05, "loss": 0.5339, "step": 2647 }, { "epoch": 1.762689299384257, "grad_norm": 41.832763671875, "learning_rate": 8.281527531083481e-05, "loss": 0.183, "step": 2648 }, { "epoch": 1.7633549675486768, "grad_norm": 73.81871032714844, "learning_rate": 8.27708703374778e-05, "loss": 0.2824, "step": 2649 }, { "epoch": 1.764020635713097, "grad_norm": 35.23242950439453, "learning_rate": 8.272646536412078e-05, "loss": 0.2389, "step": 2650 }, { "epoch": 1.7646863038775171, "grad_norm": 10.827171325683594, "learning_rate": 8.268206039076377e-05, "loss": 0.0328, "step": 2651 }, { "epoch": 1.765351972041937, "grad_norm": 8.776627540588379, "learning_rate": 8.263765541740675e-05, "loss": 0.0222, "step": 2652 }, { "epoch": 1.766017640206357, "grad_norm": 12.959379196166992, "learning_rate": 8.259325044404974e-05, "loss": 0.0321, "step": 2653 }, { "epoch": 1.7666833083707771, "grad_norm": 86.70416259765625, "learning_rate": 8.254884547069272e-05, "loss": 1.2803, "step": 2654 }, { "epoch": 1.7673489765351973, "grad_norm": 47.598140716552734, "learning_rate": 8.25044404973357e-05, "loss": 0.2609, "step": 2655 }, { "epoch": 1.7680146446996172, "grad_norm": 35.30906677246094, "learning_rate": 8.246003552397869e-05, "loss": 0.1901, "step": 2656 }, { "epoch": 1.7686803128640372, "grad_norm": 17.873184204101562, "learning_rate": 8.241563055062167e-05, "loss": 0.0594, "step": 2657 }, { "epoch": 1.7693459810284573, "grad_norm": 49.77486038208008, "learning_rate": 8.237122557726466e-05, "loss": 0.182, "step": 2658 }, { "epoch": 1.7700116491928775, "grad_norm": 57.61217498779297, "learning_rate": 8.232682060390764e-05, "loss": 0.441, "step": 2659 }, { "epoch": 1.7706773173572974, "grad_norm": 20.968868255615234, "learning_rate": 8.228241563055063e-05, "loss": 0.2143, "step": 2660 }, { "epoch": 1.7713429855217173, "grad_norm": 78.10819244384766, "learning_rate": 8.223801065719361e-05, "loss": 0.4033, "step": 2661 }, { "epoch": 1.7720086536861375, "grad_norm": 41.15236282348633, "learning_rate": 8.21936056838366e-05, "loss": 0.5091, "step": 2662 }, { "epoch": 1.7726743218505576, "grad_norm": 75.0620346069336, "learning_rate": 8.214920071047958e-05, "loss": 0.1911, "step": 2663 }, { "epoch": 1.7733399900149776, "grad_norm": 80.40105438232422, "learning_rate": 8.210479573712257e-05, "loss": 0.4448, "step": 2664 }, { "epoch": 1.7740056581793975, "grad_norm": 36.618507385253906, "learning_rate": 8.206039076376555e-05, "loss": 0.1713, "step": 2665 }, { "epoch": 1.7746713263438176, "grad_norm": 32.07270050048828, "learning_rate": 8.201598579040854e-05, "loss": 0.6278, "step": 2666 }, { "epoch": 1.7753369945082378, "grad_norm": 32.78857421875, "learning_rate": 8.197158081705152e-05, "loss": 0.3341, "step": 2667 }, { "epoch": 1.7760026626726577, "grad_norm": 20.83468246459961, "learning_rate": 8.19271758436945e-05, "loss": 0.0655, "step": 2668 }, { "epoch": 1.7766683308370776, "grad_norm": 57.2627067565918, "learning_rate": 8.188277087033748e-05, "loss": 0.6385, "step": 2669 }, { "epoch": 1.7773339990014978, "grad_norm": 20.072032928466797, "learning_rate": 8.183836589698046e-05, "loss": 0.1296, "step": 2670 }, { "epoch": 1.777999667165918, "grad_norm": 42.45745086669922, "learning_rate": 8.179396092362345e-05, "loss": 0.1224, "step": 2671 }, { "epoch": 1.7786653353303379, "grad_norm": 43.692317962646484, "learning_rate": 8.174955595026643e-05, "loss": 0.38, "step": 2672 }, { "epoch": 1.7793310034947578, "grad_norm": 22.545961380004883, "learning_rate": 8.170515097690942e-05, "loss": 0.1188, "step": 2673 }, { "epoch": 1.7799966716591777, "grad_norm": 37.031497955322266, "learning_rate": 8.16607460035524e-05, "loss": 0.3607, "step": 2674 }, { "epoch": 1.780662339823598, "grad_norm": 37.777015686035156, "learning_rate": 8.161634103019539e-05, "loss": 0.1665, "step": 2675 }, { "epoch": 1.781328007988018, "grad_norm": 24.817811965942383, "learning_rate": 8.157193605683837e-05, "loss": 0.3103, "step": 2676 }, { "epoch": 1.781993676152438, "grad_norm": 44.2794075012207, "learning_rate": 8.152753108348136e-05, "loss": 0.4044, "step": 2677 }, { "epoch": 1.782659344316858, "grad_norm": 50.487510681152344, "learning_rate": 8.148312611012434e-05, "loss": 0.1662, "step": 2678 }, { "epoch": 1.783325012481278, "grad_norm": 54.592124938964844, "learning_rate": 8.143872113676733e-05, "loss": 0.4239, "step": 2679 }, { "epoch": 1.7839906806456982, "grad_norm": 56.2703742980957, "learning_rate": 8.139431616341031e-05, "loss": 0.5623, "step": 2680 }, { "epoch": 1.7846563488101181, "grad_norm": 25.388029098510742, "learning_rate": 8.134991119005328e-05, "loss": 0.1671, "step": 2681 }, { "epoch": 1.785322016974538, "grad_norm": 24.090496063232422, "learning_rate": 8.130550621669627e-05, "loss": 0.0912, "step": 2682 }, { "epoch": 1.7859876851389582, "grad_norm": 29.685937881469727, "learning_rate": 8.126110124333925e-05, "loss": 0.1262, "step": 2683 }, { "epoch": 1.7866533533033784, "grad_norm": 33.438838958740234, "learning_rate": 8.121669626998224e-05, "loss": 0.2714, "step": 2684 }, { "epoch": 1.7873190214677983, "grad_norm": 38.1854133605957, "learning_rate": 8.117229129662522e-05, "loss": 0.5443, "step": 2685 }, { "epoch": 1.7879846896322182, "grad_norm": 14.902162551879883, "learning_rate": 8.112788632326821e-05, "loss": 0.0826, "step": 2686 }, { "epoch": 1.7886503577966384, "grad_norm": 31.024290084838867, "learning_rate": 8.108348134991119e-05, "loss": 0.3484, "step": 2687 }, { "epoch": 1.7893160259610585, "grad_norm": 2.1075057983398438, "learning_rate": 8.103907637655418e-05, "loss": 0.0106, "step": 2688 }, { "epoch": 1.7899816941254785, "grad_norm": 8.55073070526123, "learning_rate": 8.099467140319716e-05, "loss": 0.0277, "step": 2689 }, { "epoch": 1.7906473622898984, "grad_norm": 24.395856857299805, "learning_rate": 8.095026642984015e-05, "loss": 0.1709, "step": 2690 }, { "epoch": 1.7913130304543186, "grad_norm": 55.90815353393555, "learning_rate": 8.090586145648313e-05, "loss": 0.3854, "step": 2691 }, { "epoch": 1.7919786986187387, "grad_norm": 66.92333221435547, "learning_rate": 8.086145648312612e-05, "loss": 0.4055, "step": 2692 }, { "epoch": 1.7926443667831586, "grad_norm": 31.756406784057617, "learning_rate": 8.081705150976909e-05, "loss": 0.1913, "step": 2693 }, { "epoch": 1.7933100349475786, "grad_norm": 37.485069274902344, "learning_rate": 8.077264653641207e-05, "loss": 0.4055, "step": 2694 }, { "epoch": 1.7939757031119987, "grad_norm": 18.087495803833008, "learning_rate": 8.072824156305506e-05, "loss": 0.0699, "step": 2695 }, { "epoch": 1.7946413712764187, "grad_norm": 13.640735626220703, "learning_rate": 8.068383658969804e-05, "loss": 0.055, "step": 2696 }, { "epoch": 1.7953070394408388, "grad_norm": 45.913238525390625, "learning_rate": 8.063943161634103e-05, "loss": 0.4131, "step": 2697 }, { "epoch": 1.7959727076052587, "grad_norm": 7.278524875640869, "learning_rate": 8.059502664298401e-05, "loss": 0.0444, "step": 2698 }, { "epoch": 1.7966383757696787, "grad_norm": 39.094703674316406, "learning_rate": 8.0550621669627e-05, "loss": 0.3786, "step": 2699 }, { "epoch": 1.7973040439340988, "grad_norm": 59.006591796875, "learning_rate": 8.050621669626998e-05, "loss": 0.3703, "step": 2700 }, { "epoch": 1.797969712098519, "grad_norm": 35.40041732788086, "learning_rate": 8.046181172291297e-05, "loss": 0.1255, "step": 2701 }, { "epoch": 1.798635380262939, "grad_norm": 3.513657331466675, "learning_rate": 8.041740674955595e-05, "loss": 0.0141, "step": 2702 }, { "epoch": 1.7993010484273588, "grad_norm": 16.90599822998047, "learning_rate": 8.037300177619894e-05, "loss": 0.0538, "step": 2703 }, { "epoch": 1.799966716591779, "grad_norm": 9.73635482788086, "learning_rate": 8.032859680284192e-05, "loss": 0.0367, "step": 2704 }, { "epoch": 1.8006323847561991, "grad_norm": 31.437097549438477, "learning_rate": 8.02841918294849e-05, "loss": 0.1218, "step": 2705 }, { "epoch": 1.801298052920619, "grad_norm": 20.632604598999023, "learning_rate": 8.023978685612789e-05, "loss": 0.0696, "step": 2706 }, { "epoch": 1.801963721085039, "grad_norm": 27.330516815185547, "learning_rate": 8.019538188277088e-05, "loss": 0.1132, "step": 2707 }, { "epoch": 1.8026293892494591, "grad_norm": 32.40435028076172, "learning_rate": 8.015097690941386e-05, "loss": 0.1119, "step": 2708 }, { "epoch": 1.8032950574138793, "grad_norm": 18.288724899291992, "learning_rate": 8.010657193605685e-05, "loss": 0.0565, "step": 2709 }, { "epoch": 1.8039607255782992, "grad_norm": 62.09398651123047, "learning_rate": 8.006216696269983e-05, "loss": 0.4318, "step": 2710 }, { "epoch": 1.8046263937427192, "grad_norm": 43.98017120361328, "learning_rate": 8.001776198934282e-05, "loss": 0.2005, "step": 2711 }, { "epoch": 1.8052920619071393, "grad_norm": 36.573822021484375, "learning_rate": 7.99733570159858e-05, "loss": 0.1393, "step": 2712 }, { "epoch": 1.8059577300715595, "grad_norm": 19.97627830505371, "learning_rate": 7.992895204262878e-05, "loss": 0.1151, "step": 2713 }, { "epoch": 1.8066233982359794, "grad_norm": 61.16523361206055, "learning_rate": 7.988454706927177e-05, "loss": 0.5298, "step": 2714 }, { "epoch": 1.8072890664003993, "grad_norm": 27.421443939208984, "learning_rate": 7.984014209591475e-05, "loss": 0.1654, "step": 2715 }, { "epoch": 1.8079547345648195, "grad_norm": 21.24353790283203, "learning_rate": 7.979573712255774e-05, "loss": 0.0634, "step": 2716 }, { "epoch": 1.8086204027292396, "grad_norm": 22.30165672302246, "learning_rate": 7.975133214920072e-05, "loss": 0.1725, "step": 2717 }, { "epoch": 1.8092860708936596, "grad_norm": 13.919240951538086, "learning_rate": 7.97069271758437e-05, "loss": 0.0394, "step": 2718 }, { "epoch": 1.8099517390580795, "grad_norm": 58.61162567138672, "learning_rate": 7.966252220248668e-05, "loss": 0.4469, "step": 2719 }, { "epoch": 1.8106174072224994, "grad_norm": 21.072168350219727, "learning_rate": 7.961811722912967e-05, "loss": 0.0625, "step": 2720 }, { "epoch": 1.8112830753869196, "grad_norm": 74.75184631347656, "learning_rate": 7.957371225577265e-05, "loss": 0.1985, "step": 2721 }, { "epoch": 1.8119487435513397, "grad_norm": 16.468860626220703, "learning_rate": 7.952930728241563e-05, "loss": 0.0424, "step": 2722 }, { "epoch": 1.8126144117157597, "grad_norm": 66.07185363769531, "learning_rate": 7.948490230905862e-05, "loss": 0.0609, "step": 2723 }, { "epoch": 1.8132800798801796, "grad_norm": 24.108409881591797, "learning_rate": 7.94404973357016e-05, "loss": 0.0625, "step": 2724 }, { "epoch": 1.8139457480445997, "grad_norm": 34.122528076171875, "learning_rate": 7.939609236234459e-05, "loss": 0.0927, "step": 2725 }, { "epoch": 1.81461141620902, "grad_norm": 11.17801284790039, "learning_rate": 7.935168738898757e-05, "loss": 0.0543, "step": 2726 }, { "epoch": 1.8152770843734398, "grad_norm": 7.584578037261963, "learning_rate": 7.930728241563056e-05, "loss": 0.015, "step": 2727 }, { "epoch": 1.8159427525378598, "grad_norm": 33.989986419677734, "learning_rate": 7.926287744227354e-05, "loss": 0.1769, "step": 2728 }, { "epoch": 1.81660842070228, "grad_norm": 78.71027374267578, "learning_rate": 7.921847246891653e-05, "loss": 0.3429, "step": 2729 }, { "epoch": 1.8172740888667, "grad_norm": 4.057672500610352, "learning_rate": 7.917406749555951e-05, "loss": 0.0118, "step": 2730 }, { "epoch": 1.81793975703112, "grad_norm": 41.898109436035156, "learning_rate": 7.912966252220248e-05, "loss": 0.4597, "step": 2731 }, { "epoch": 1.81860542519554, "grad_norm": 41.811344146728516, "learning_rate": 7.908525754884547e-05, "loss": 0.114, "step": 2732 }, { "epoch": 1.81927109335996, "grad_norm": 25.377656936645508, "learning_rate": 7.904085257548845e-05, "loss": 0.0903, "step": 2733 }, { "epoch": 1.8199367615243802, "grad_norm": 42.144107818603516, "learning_rate": 7.899644760213144e-05, "loss": 0.1592, "step": 2734 }, { "epoch": 1.8206024296888002, "grad_norm": 42.22956466674805, "learning_rate": 7.895204262877442e-05, "loss": 0.233, "step": 2735 }, { "epoch": 1.82126809785322, "grad_norm": 55.821258544921875, "learning_rate": 7.890763765541741e-05, "loss": 0.2383, "step": 2736 }, { "epoch": 1.8219337660176402, "grad_norm": 50.268856048583984, "learning_rate": 7.88632326820604e-05, "loss": 0.3173, "step": 2737 }, { "epoch": 1.8225994341820604, "grad_norm": 27.03841209411621, "learning_rate": 7.881882770870338e-05, "loss": 0.0833, "step": 2738 }, { "epoch": 1.8232651023464803, "grad_norm": 61.994712829589844, "learning_rate": 7.877442273534636e-05, "loss": 0.3148, "step": 2739 }, { "epoch": 1.8239307705109002, "grad_norm": 114.39158630371094, "learning_rate": 7.873001776198935e-05, "loss": 0.3413, "step": 2740 }, { "epoch": 1.8245964386753204, "grad_norm": 64.72805786132812, "learning_rate": 7.868561278863233e-05, "loss": 0.4494, "step": 2741 }, { "epoch": 1.8252621068397406, "grad_norm": 18.24867057800293, "learning_rate": 7.864120781527532e-05, "loss": 0.0247, "step": 2742 }, { "epoch": 1.8259277750041605, "grad_norm": 10.917926788330078, "learning_rate": 7.859680284191829e-05, "loss": 0.0139, "step": 2743 }, { "epoch": 1.8265934431685804, "grad_norm": 46.26217269897461, "learning_rate": 7.855239786856127e-05, "loss": 0.1141, "step": 2744 }, { "epoch": 1.8272591113330003, "grad_norm": 6.841650009155273, "learning_rate": 7.850799289520426e-05, "loss": 0.0114, "step": 2745 }, { "epoch": 1.8279247794974205, "grad_norm": 52.37775421142578, "learning_rate": 7.846358792184724e-05, "loss": 0.3959, "step": 2746 }, { "epoch": 1.8285904476618406, "grad_norm": 20.178836822509766, "learning_rate": 7.841918294849023e-05, "loss": 0.0452, "step": 2747 }, { "epoch": 1.8292561158262606, "grad_norm": 27.97673988342285, "learning_rate": 7.837477797513321e-05, "loss": 0.0393, "step": 2748 }, { "epoch": 1.8299217839906805, "grad_norm": 35.92624282836914, "learning_rate": 7.83303730017762e-05, "loss": 0.2318, "step": 2749 }, { "epoch": 1.8305874521551007, "grad_norm": 48.933677673339844, "learning_rate": 7.828596802841918e-05, "loss": 0.3686, "step": 2750 }, { "epoch": 1.8312531203195208, "grad_norm": 4.591276168823242, "learning_rate": 7.824156305506217e-05, "loss": 0.0089, "step": 2751 }, { "epoch": 1.8319187884839407, "grad_norm": 37.67312240600586, "learning_rate": 7.819715808170515e-05, "loss": 0.2051, "step": 2752 }, { "epoch": 1.8325844566483607, "grad_norm": 44.68748474121094, "learning_rate": 7.815275310834814e-05, "loss": 0.8232, "step": 2753 }, { "epoch": 1.8332501248127808, "grad_norm": 2.5557971000671387, "learning_rate": 7.810834813499112e-05, "loss": 0.0102, "step": 2754 }, { "epoch": 1.833915792977201, "grad_norm": 57.844093322753906, "learning_rate": 7.806394316163411e-05, "loss": 0.2435, "step": 2755 }, { "epoch": 1.834581461141621, "grad_norm": 1.9619042873382568, "learning_rate": 7.801953818827709e-05, "loss": 0.0036, "step": 2756 }, { "epoch": 1.8352471293060408, "grad_norm": 52.30373764038086, "learning_rate": 7.797513321492008e-05, "loss": 0.23, "step": 2757 }, { "epoch": 1.835912797470461, "grad_norm": 58.690914154052734, "learning_rate": 7.793072824156306e-05, "loss": 0.1715, "step": 2758 }, { "epoch": 1.8365784656348811, "grad_norm": 23.2955322265625, "learning_rate": 7.788632326820605e-05, "loss": 0.0521, "step": 2759 }, { "epoch": 1.837244133799301, "grad_norm": 15.068081855773926, "learning_rate": 7.784191829484903e-05, "loss": 0.0677, "step": 2760 }, { "epoch": 1.837909801963721, "grad_norm": 62.1632080078125, "learning_rate": 7.779751332149202e-05, "loss": 0.2528, "step": 2761 }, { "epoch": 1.8385754701281412, "grad_norm": 30.124385833740234, "learning_rate": 7.7753108348135e-05, "loss": 0.0716, "step": 2762 }, { "epoch": 1.8392411382925613, "grad_norm": 23.337011337280273, "learning_rate": 7.770870337477799e-05, "loss": 0.0548, "step": 2763 }, { "epoch": 1.8399068064569812, "grad_norm": 56.498416900634766, "learning_rate": 7.766429840142097e-05, "loss": 0.3154, "step": 2764 }, { "epoch": 1.8405724746214012, "grad_norm": 15.785268783569336, "learning_rate": 7.761989342806396e-05, "loss": 0.067, "step": 2765 }, { "epoch": 1.8412381427858213, "grad_norm": 21.690275192260742, "learning_rate": 7.757548845470694e-05, "loss": 0.0582, "step": 2766 }, { "epoch": 1.8419038109502412, "grad_norm": 12.153983116149902, "learning_rate": 7.753108348134993e-05, "loss": 0.0279, "step": 2767 }, { "epoch": 1.8425694791146614, "grad_norm": 43.059295654296875, "learning_rate": 7.74866785079929e-05, "loss": 0.1933, "step": 2768 }, { "epoch": 1.8432351472790813, "grad_norm": 81.39176177978516, "learning_rate": 7.744227353463588e-05, "loss": 0.1981, "step": 2769 }, { "epoch": 1.8439008154435013, "grad_norm": 94.04952239990234, "learning_rate": 7.739786856127887e-05, "loss": 0.185, "step": 2770 }, { "epoch": 1.8445664836079214, "grad_norm": 11.107410430908203, "learning_rate": 7.735346358792185e-05, "loss": 0.0211, "step": 2771 }, { "epoch": 1.8452321517723416, "grad_norm": 58.752750396728516, "learning_rate": 7.730905861456484e-05, "loss": 0.7693, "step": 2772 }, { "epoch": 1.8458978199367615, "grad_norm": 21.116273880004883, "learning_rate": 7.726465364120782e-05, "loss": 0.0985, "step": 2773 }, { "epoch": 1.8465634881011814, "grad_norm": 7.067499160766602, "learning_rate": 7.72202486678508e-05, "loss": 0.0155, "step": 2774 }, { "epoch": 1.8472291562656016, "grad_norm": 76.12965393066406, "learning_rate": 7.717584369449379e-05, "loss": 0.751, "step": 2775 }, { "epoch": 1.8478948244300217, "grad_norm": 31.86406135559082, "learning_rate": 7.713143872113678e-05, "loss": 0.0963, "step": 2776 }, { "epoch": 1.8485604925944417, "grad_norm": 12.557707786560059, "learning_rate": 7.708703374777976e-05, "loss": 0.0318, "step": 2777 }, { "epoch": 1.8492261607588616, "grad_norm": 69.96639251708984, "learning_rate": 7.704262877442275e-05, "loss": 0.4184, "step": 2778 }, { "epoch": 1.8498918289232817, "grad_norm": 75.41741943359375, "learning_rate": 7.699822380106573e-05, "loss": 0.4956, "step": 2779 }, { "epoch": 1.850557497087702, "grad_norm": 94.87342834472656, "learning_rate": 7.695381882770871e-05, "loss": 0.2208, "step": 2780 }, { "epoch": 1.8512231652521218, "grad_norm": 12.924460411071777, "learning_rate": 7.690941385435169e-05, "loss": 0.0182, "step": 2781 }, { "epoch": 1.8518888334165418, "grad_norm": 95.09745025634766, "learning_rate": 7.686500888099467e-05, "loss": 0.2722, "step": 2782 }, { "epoch": 1.852554501580962, "grad_norm": 96.17192077636719, "learning_rate": 7.682060390763766e-05, "loss": 1.1858, "step": 2783 }, { "epoch": 1.853220169745382, "grad_norm": 35.83465576171875, "learning_rate": 7.677619893428064e-05, "loss": 0.2793, "step": 2784 }, { "epoch": 1.853885837909802, "grad_norm": 29.156787872314453, "learning_rate": 7.673179396092363e-05, "loss": 0.0599, "step": 2785 }, { "epoch": 1.854551506074222, "grad_norm": 7.492499828338623, "learning_rate": 7.668738898756661e-05, "loss": 0.0119, "step": 2786 }, { "epoch": 1.855217174238642, "grad_norm": 125.66876983642578, "learning_rate": 7.66429840142096e-05, "loss": 1.3937, "step": 2787 }, { "epoch": 1.8558828424030622, "grad_norm": 42.43928146362305, "learning_rate": 7.659857904085258e-05, "loss": 0.3593, "step": 2788 }, { "epoch": 1.8565485105674822, "grad_norm": 49.2476692199707, "learning_rate": 7.655417406749556e-05, "loss": 0.421, "step": 2789 }, { "epoch": 1.857214178731902, "grad_norm": 38.29008483886719, "learning_rate": 7.650976909413855e-05, "loss": 0.2685, "step": 2790 }, { "epoch": 1.857879846896322, "grad_norm": 68.42425537109375, "learning_rate": 7.646536412078153e-05, "loss": 0.2808, "step": 2791 }, { "epoch": 1.8585455150607422, "grad_norm": 39.59495544433594, "learning_rate": 7.642095914742452e-05, "loss": 0.2852, "step": 2792 }, { "epoch": 1.8592111832251623, "grad_norm": 58.05021286010742, "learning_rate": 7.637655417406749e-05, "loss": 0.3073, "step": 2793 }, { "epoch": 1.8598768513895823, "grad_norm": 35.57040786743164, "learning_rate": 7.633214920071048e-05, "loss": 0.216, "step": 2794 }, { "epoch": 1.8605425195540022, "grad_norm": 28.909866333007812, "learning_rate": 7.628774422735346e-05, "loss": 0.0593, "step": 2795 }, { "epoch": 1.8612081877184223, "grad_norm": 32.26399612426758, "learning_rate": 7.624333925399645e-05, "loss": 0.0446, "step": 2796 }, { "epoch": 1.8618738558828425, "grad_norm": 1.3602535724639893, "learning_rate": 7.619893428063943e-05, "loss": 0.0042, "step": 2797 }, { "epoch": 1.8625395240472624, "grad_norm": 31.194215774536133, "learning_rate": 7.615452930728241e-05, "loss": 0.125, "step": 2798 }, { "epoch": 1.8632051922116823, "grad_norm": 7.358043670654297, "learning_rate": 7.61101243339254e-05, "loss": 0.0183, "step": 2799 }, { "epoch": 1.8638708603761025, "grad_norm": 56.010597229003906, "learning_rate": 7.606571936056838e-05, "loss": 0.3711, "step": 2800 }, { "epoch": 1.8645365285405227, "grad_norm": 62.644649505615234, "learning_rate": 7.602131438721137e-05, "loss": 0.322, "step": 2801 }, { "epoch": 1.8652021967049426, "grad_norm": 64.53557586669922, "learning_rate": 7.597690941385435e-05, "loss": 0.3271, "step": 2802 }, { "epoch": 1.8658678648693625, "grad_norm": 30.464277267456055, "learning_rate": 7.593250444049734e-05, "loss": 0.1234, "step": 2803 }, { "epoch": 1.8665335330337827, "grad_norm": 3.7914674282073975, "learning_rate": 7.588809946714032e-05, "loss": 0.0099, "step": 2804 }, { "epoch": 1.8671992011982028, "grad_norm": 22.578453063964844, "learning_rate": 7.584369449378331e-05, "loss": 0.0522, "step": 2805 }, { "epoch": 1.8678648693626227, "grad_norm": 40.23014831542969, "learning_rate": 7.579928952042628e-05, "loss": 0.4768, "step": 2806 }, { "epoch": 1.8685305375270427, "grad_norm": 72.4029769897461, "learning_rate": 7.575488454706926e-05, "loss": 0.3344, "step": 2807 }, { "epoch": 1.8691962056914628, "grad_norm": 4.151562213897705, "learning_rate": 7.571047957371225e-05, "loss": 0.0124, "step": 2808 }, { "epoch": 1.869861873855883, "grad_norm": 59.1591796875, "learning_rate": 7.566607460035523e-05, "loss": 0.2384, "step": 2809 }, { "epoch": 1.870527542020303, "grad_norm": 10.145440101623535, "learning_rate": 7.562166962699823e-05, "loss": 0.0364, "step": 2810 }, { "epoch": 1.8711932101847228, "grad_norm": 46.685367584228516, "learning_rate": 7.557726465364122e-05, "loss": 0.5098, "step": 2811 }, { "epoch": 1.871858878349143, "grad_norm": 17.860754013061523, "learning_rate": 7.55328596802842e-05, "loss": 0.0222, "step": 2812 }, { "epoch": 1.8725245465135631, "grad_norm": 5.977094650268555, "learning_rate": 7.548845470692719e-05, "loss": 0.0133, "step": 2813 }, { "epoch": 1.873190214677983, "grad_norm": 24.474895477294922, "learning_rate": 7.544404973357017e-05, "loss": 0.0757, "step": 2814 }, { "epoch": 1.873855882842403, "grad_norm": 79.36913299560547, "learning_rate": 7.539964476021316e-05, "loss": 0.9514, "step": 2815 }, { "epoch": 1.874521551006823, "grad_norm": 92.79966735839844, "learning_rate": 7.535523978685614e-05, "loss": 0.4391, "step": 2816 }, { "epoch": 1.875187219171243, "grad_norm": 26.47312355041504, "learning_rate": 7.531083481349913e-05, "loss": 0.0718, "step": 2817 }, { "epoch": 1.8758528873356632, "grad_norm": 23.758758544921875, "learning_rate": 7.52664298401421e-05, "loss": 0.0361, "step": 2818 }, { "epoch": 1.8765185555000832, "grad_norm": 62.85113525390625, "learning_rate": 7.522202486678508e-05, "loss": 0.2937, "step": 2819 }, { "epoch": 1.877184223664503, "grad_norm": 53.8686637878418, "learning_rate": 7.517761989342807e-05, "loss": 0.3988, "step": 2820 }, { "epoch": 1.8778498918289233, "grad_norm": 42.02883529663086, "learning_rate": 7.513321492007105e-05, "loss": 0.325, "step": 2821 }, { "epoch": 1.8785155599933434, "grad_norm": 89.2615966796875, "learning_rate": 7.508880994671404e-05, "loss": 0.2937, "step": 2822 }, { "epoch": 1.8791812281577633, "grad_norm": 78.1486587524414, "learning_rate": 7.504440497335702e-05, "loss": 0.452, "step": 2823 }, { "epoch": 1.8798468963221833, "grad_norm": 3.7266769409179688, "learning_rate": 7.500000000000001e-05, "loss": 0.0048, "step": 2824 }, { "epoch": 1.8805125644866034, "grad_norm": 41.9632568359375, "learning_rate": 7.495559502664299e-05, "loss": 0.1052, "step": 2825 }, { "epoch": 1.8811782326510236, "grad_norm": 31.525318145751953, "learning_rate": 7.491119005328598e-05, "loss": 0.1047, "step": 2826 }, { "epoch": 1.8818439008154435, "grad_norm": 69.26793670654297, "learning_rate": 7.486678507992896e-05, "loss": 0.2557, "step": 2827 }, { "epoch": 1.8825095689798634, "grad_norm": 18.465503692626953, "learning_rate": 7.482238010657195e-05, "loss": 0.0549, "step": 2828 }, { "epoch": 1.8831752371442836, "grad_norm": 26.04500389099121, "learning_rate": 7.477797513321493e-05, "loss": 0.0608, "step": 2829 }, { "epoch": 1.8838409053087037, "grad_norm": 4.349294662475586, "learning_rate": 7.473357015985792e-05, "loss": 0.0114, "step": 2830 }, { "epoch": 1.8845065734731237, "grad_norm": 28.608726501464844, "learning_rate": 7.468916518650089e-05, "loss": 0.142, "step": 2831 }, { "epoch": 1.8851722416375436, "grad_norm": 34.37458038330078, "learning_rate": 7.464476021314387e-05, "loss": 0.1136, "step": 2832 }, { "epoch": 1.8858379098019638, "grad_norm": 12.163702011108398, "learning_rate": 7.460035523978686e-05, "loss": 0.0387, "step": 2833 }, { "epoch": 1.886503577966384, "grad_norm": 5.014279842376709, "learning_rate": 7.455595026642984e-05, "loss": 0.0124, "step": 2834 }, { "epoch": 1.8871692461308038, "grad_norm": 36.29486083984375, "learning_rate": 7.451154529307283e-05, "loss": 0.1098, "step": 2835 }, { "epoch": 1.8878349142952238, "grad_norm": 64.902099609375, "learning_rate": 7.446714031971581e-05, "loss": 0.5301, "step": 2836 }, { "epoch": 1.888500582459644, "grad_norm": 56.5096549987793, "learning_rate": 7.44227353463588e-05, "loss": 0.4914, "step": 2837 }, { "epoch": 1.889166250624064, "grad_norm": 15.485305786132812, "learning_rate": 7.437833037300178e-05, "loss": 0.0541, "step": 2838 }, { "epoch": 1.889831918788484, "grad_norm": 67.94281005859375, "learning_rate": 7.433392539964477e-05, "loss": 0.7816, "step": 2839 }, { "epoch": 1.890497586952904, "grad_norm": 5.544710159301758, "learning_rate": 7.428952042628775e-05, "loss": 0.0097, "step": 2840 }, { "epoch": 1.8911632551173239, "grad_norm": 38.6276741027832, "learning_rate": 7.424511545293074e-05, "loss": 0.5267, "step": 2841 }, { "epoch": 1.891828923281744, "grad_norm": 0.9757862687110901, "learning_rate": 7.420071047957372e-05, "loss": 0.0029, "step": 2842 }, { "epoch": 1.8924945914461642, "grad_norm": 31.113445281982422, "learning_rate": 7.415630550621669e-05, "loss": 0.1017, "step": 2843 }, { "epoch": 1.893160259610584, "grad_norm": 28.256975173950195, "learning_rate": 7.411190053285968e-05, "loss": 0.2323, "step": 2844 }, { "epoch": 1.893825927775004, "grad_norm": 87.38008880615234, "learning_rate": 7.406749555950266e-05, "loss": 0.4401, "step": 2845 }, { "epoch": 1.8944915959394242, "grad_norm": 31.434106826782227, "learning_rate": 7.402309058614565e-05, "loss": 0.5407, "step": 2846 }, { "epoch": 1.8951572641038443, "grad_norm": 20.329036712646484, "learning_rate": 7.397868561278863e-05, "loss": 0.0381, "step": 2847 }, { "epoch": 1.8958229322682643, "grad_norm": 16.281925201416016, "learning_rate": 7.393428063943162e-05, "loss": 0.0495, "step": 2848 }, { "epoch": 1.8964886004326842, "grad_norm": 27.099016189575195, "learning_rate": 7.38898756660746e-05, "loss": 0.0789, "step": 2849 }, { "epoch": 1.8971542685971043, "grad_norm": 15.258371353149414, "learning_rate": 7.384547069271759e-05, "loss": 0.0368, "step": 2850 }, { "epoch": 1.8978199367615245, "grad_norm": 41.89741516113281, "learning_rate": 7.380106571936057e-05, "loss": 0.1323, "step": 2851 }, { "epoch": 1.8984856049259444, "grad_norm": 50.287513732910156, "learning_rate": 7.375666074600356e-05, "loss": 0.2183, "step": 2852 }, { "epoch": 1.8991512730903644, "grad_norm": 23.819271087646484, "learning_rate": 7.371225577264654e-05, "loss": 0.059, "step": 2853 }, { "epoch": 1.8998169412547845, "grad_norm": 59.39651107788086, "learning_rate": 7.366785079928952e-05, "loss": 0.3732, "step": 2854 }, { "epoch": 1.9004826094192047, "grad_norm": 44.32624816894531, "learning_rate": 7.362344582593251e-05, "loss": 0.2625, "step": 2855 }, { "epoch": 1.9011482775836246, "grad_norm": 28.70370101928711, "learning_rate": 7.357904085257548e-05, "loss": 0.1195, "step": 2856 }, { "epoch": 1.9018139457480445, "grad_norm": 55.44340515136719, "learning_rate": 7.353463587921847e-05, "loss": 0.2002, "step": 2857 }, { "epoch": 1.9024796139124647, "grad_norm": 80.92920684814453, "learning_rate": 7.349023090586145e-05, "loss": 0.1312, "step": 2858 }, { "epoch": 1.9031452820768848, "grad_norm": 61.36829376220703, "learning_rate": 7.344582593250444e-05, "loss": 0.2713, "step": 2859 }, { "epoch": 1.9038109502413048, "grad_norm": 42.589019775390625, "learning_rate": 7.340142095914742e-05, "loss": 0.0775, "step": 2860 }, { "epoch": 1.9044766184057247, "grad_norm": 35.52931213378906, "learning_rate": 7.33570159857904e-05, "loss": 0.1198, "step": 2861 }, { "epoch": 1.9051422865701446, "grad_norm": 56.813621520996094, "learning_rate": 7.331261101243339e-05, "loss": 0.2514, "step": 2862 }, { "epoch": 1.9058079547345648, "grad_norm": 27.69788932800293, "learning_rate": 7.326820603907637e-05, "loss": 0.0747, "step": 2863 }, { "epoch": 1.906473622898985, "grad_norm": 4.723409652709961, "learning_rate": 7.322380106571936e-05, "loss": 0.0096, "step": 2864 }, { "epoch": 1.9071392910634049, "grad_norm": 0.38779914379119873, "learning_rate": 7.317939609236234e-05, "loss": 0.0011, "step": 2865 }, { "epoch": 1.9078049592278248, "grad_norm": 11.887478828430176, "learning_rate": 7.313499111900533e-05, "loss": 0.0338, "step": 2866 }, { "epoch": 1.908470627392245, "grad_norm": 44.516258239746094, "learning_rate": 7.309058614564831e-05, "loss": 0.1925, "step": 2867 }, { "epoch": 1.909136295556665, "grad_norm": 3.035400867462158, "learning_rate": 7.30461811722913e-05, "loss": 0.0068, "step": 2868 }, { "epoch": 1.909801963721085, "grad_norm": 2.2110989093780518, "learning_rate": 7.300177619893428e-05, "loss": 0.0047, "step": 2869 }, { "epoch": 1.910467631885505, "grad_norm": 42.917640686035156, "learning_rate": 7.295737122557727e-05, "loss": 0.0951, "step": 2870 }, { "epoch": 1.911133300049925, "grad_norm": 10.228337287902832, "learning_rate": 7.291296625222025e-05, "loss": 0.0139, "step": 2871 }, { "epoch": 1.9117989682143453, "grad_norm": 73.12277221679688, "learning_rate": 7.286856127886324e-05, "loss": 0.5454, "step": 2872 }, { "epoch": 1.9124646363787652, "grad_norm": 5.225230693817139, "learning_rate": 7.282415630550622e-05, "loss": 0.0088, "step": 2873 }, { "epoch": 1.9131303045431851, "grad_norm": 9.360527992248535, "learning_rate": 7.277975133214921e-05, "loss": 0.017, "step": 2874 }, { "epoch": 1.9137959727076053, "grad_norm": 104.11754608154297, "learning_rate": 7.273534635879219e-05, "loss": 0.8431, "step": 2875 }, { "epoch": 1.9144616408720254, "grad_norm": 60.74580001831055, "learning_rate": 7.269094138543518e-05, "loss": 0.2242, "step": 2876 }, { "epoch": 1.9151273090364453, "grad_norm": 32.47584533691406, "learning_rate": 7.264653641207816e-05, "loss": 0.0721, "step": 2877 }, { "epoch": 1.9157929772008653, "grad_norm": 9.751727104187012, "learning_rate": 7.260213143872115e-05, "loss": 0.0165, "step": 2878 }, { "epoch": 1.9164586453652854, "grad_norm": 107.07324981689453, "learning_rate": 7.255772646536413e-05, "loss": 0.545, "step": 2879 }, { "epoch": 1.9171243135297056, "grad_norm": 123.25487518310547, "learning_rate": 7.251332149200712e-05, "loss": 0.3001, "step": 2880 }, { "epoch": 1.9177899816941255, "grad_norm": 6.253665924072266, "learning_rate": 7.246891651865009e-05, "loss": 0.005, "step": 2881 }, { "epoch": 1.9184556498585454, "grad_norm": 48.15544509887695, "learning_rate": 7.242451154529307e-05, "loss": 0.1635, "step": 2882 }, { "epoch": 1.9191213180229656, "grad_norm": 30.17596435546875, "learning_rate": 7.238010657193606e-05, "loss": 0.1061, "step": 2883 }, { "epoch": 1.9197869861873857, "grad_norm": 27.58555793762207, "learning_rate": 7.233570159857904e-05, "loss": 0.0615, "step": 2884 }, { "epoch": 1.9204526543518057, "grad_norm": 52.63606643676758, "learning_rate": 7.229129662522203e-05, "loss": 0.2158, "step": 2885 }, { "epoch": 1.9211183225162256, "grad_norm": 54.130271911621094, "learning_rate": 7.224689165186501e-05, "loss": 0.0657, "step": 2886 }, { "epoch": 1.9217839906806455, "grad_norm": 44.94889831542969, "learning_rate": 7.2202486678508e-05, "loss": 0.3186, "step": 2887 }, { "epoch": 1.9224496588450657, "grad_norm": 34.67464828491211, "learning_rate": 7.215808170515098e-05, "loss": 0.1207, "step": 2888 }, { "epoch": 1.9231153270094858, "grad_norm": 64.97206115722656, "learning_rate": 7.211367673179397e-05, "loss": 0.5927, "step": 2889 }, { "epoch": 1.9237809951739058, "grad_norm": 48.11124801635742, "learning_rate": 7.206927175843695e-05, "loss": 0.1286, "step": 2890 }, { "epoch": 1.9244466633383257, "grad_norm": 23.16326904296875, "learning_rate": 7.202486678507994e-05, "loss": 0.0424, "step": 2891 }, { "epoch": 1.9251123315027459, "grad_norm": 14.012351036071777, "learning_rate": 7.198046181172292e-05, "loss": 0.0232, "step": 2892 }, { "epoch": 1.925777999667166, "grad_norm": 30.978113174438477, "learning_rate": 7.193605683836589e-05, "loss": 0.1046, "step": 2893 }, { "epoch": 1.926443667831586, "grad_norm": 6.459052562713623, "learning_rate": 7.189165186500888e-05, "loss": 0.0066, "step": 2894 }, { "epoch": 1.9271093359960059, "grad_norm": 33.82785415649414, "learning_rate": 7.184724689165186e-05, "loss": 0.0672, "step": 2895 }, { "epoch": 1.927775004160426, "grad_norm": 20.29143714904785, "learning_rate": 7.180284191829485e-05, "loss": 0.0475, "step": 2896 }, { "epoch": 1.9284406723248462, "grad_norm": 18.79274559020996, "learning_rate": 7.175843694493783e-05, "loss": 0.0535, "step": 2897 }, { "epoch": 1.929106340489266, "grad_norm": 34.87660217285156, "learning_rate": 7.171403197158082e-05, "loss": 0.1316, "step": 2898 }, { "epoch": 1.929772008653686, "grad_norm": 59.329288482666016, "learning_rate": 7.16696269982238e-05, "loss": 0.23, "step": 2899 }, { "epoch": 1.9304376768181062, "grad_norm": 0.738498866558075, "learning_rate": 7.162522202486679e-05, "loss": 0.0014, "step": 2900 }, { "epoch": 1.9311033449825263, "grad_norm": 6.381306171417236, "learning_rate": 7.158081705150977e-05, "loss": 0.0096, "step": 2901 }, { "epoch": 1.9317690131469463, "grad_norm": 41.281211853027344, "learning_rate": 7.153641207815276e-05, "loss": 0.1178, "step": 2902 }, { "epoch": 1.9324346813113662, "grad_norm": 78.23087310791016, "learning_rate": 7.149200710479574e-05, "loss": 0.4841, "step": 2903 }, { "epoch": 1.9331003494757864, "grad_norm": 9.567252159118652, "learning_rate": 7.144760213143873e-05, "loss": 0.0164, "step": 2904 }, { "epoch": 1.9337660176402065, "grad_norm": 9.694705963134766, "learning_rate": 7.140319715808171e-05, "loss": 0.0135, "step": 2905 }, { "epoch": 1.9344316858046264, "grad_norm": 60.62501907348633, "learning_rate": 7.135879218472468e-05, "loss": 0.4098, "step": 2906 }, { "epoch": 1.9350973539690464, "grad_norm": 51.086116790771484, "learning_rate": 7.131438721136767e-05, "loss": 0.233, "step": 2907 }, { "epoch": 1.9357630221334665, "grad_norm": 30.132762908935547, "learning_rate": 7.126998223801065e-05, "loss": 0.0706, "step": 2908 }, { "epoch": 1.9364286902978867, "grad_norm": 28.5262451171875, "learning_rate": 7.122557726465364e-05, "loss": 0.0808, "step": 2909 }, { "epoch": 1.9370943584623066, "grad_norm": 21.183320999145508, "learning_rate": 7.118117229129662e-05, "loss": 0.064, "step": 2910 }, { "epoch": 1.9377600266267265, "grad_norm": 73.25838470458984, "learning_rate": 7.11367673179396e-05, "loss": 0.4005, "step": 2911 }, { "epoch": 1.9384256947911465, "grad_norm": 80.67166900634766, "learning_rate": 7.109236234458259e-05, "loss": 0.3255, "step": 2912 }, { "epoch": 1.9390913629555666, "grad_norm": 41.56186294555664, "learning_rate": 7.104795737122558e-05, "loss": 0.2098, "step": 2913 }, { "epoch": 1.9397570311199868, "grad_norm": 71.6270751953125, "learning_rate": 7.100355239786856e-05, "loss": 0.8507, "step": 2914 }, { "epoch": 1.9404226992844067, "grad_norm": 56.03753662109375, "learning_rate": 7.095914742451155e-05, "loss": 0.1024, "step": 2915 }, { "epoch": 1.9410883674488266, "grad_norm": 65.22859191894531, "learning_rate": 7.091474245115453e-05, "loss": 0.4686, "step": 2916 }, { "epoch": 1.9417540356132468, "grad_norm": 71.27999114990234, "learning_rate": 7.087033747779752e-05, "loss": 0.4975, "step": 2917 }, { "epoch": 1.942419703777667, "grad_norm": 21.529245376586914, "learning_rate": 7.08259325044405e-05, "loss": 0.0665, "step": 2918 }, { "epoch": 1.9430853719420869, "grad_norm": 76.30394744873047, "learning_rate": 7.078152753108349e-05, "loss": 0.2908, "step": 2919 }, { "epoch": 1.9437510401065068, "grad_norm": 31.704683303833008, "learning_rate": 7.073712255772647e-05, "loss": 0.3829, "step": 2920 }, { "epoch": 1.944416708270927, "grad_norm": 19.12434196472168, "learning_rate": 7.069271758436945e-05, "loss": 0.0356, "step": 2921 }, { "epoch": 1.945082376435347, "grad_norm": 32.388755798339844, "learning_rate": 7.064831261101244e-05, "loss": 0.1451, "step": 2922 }, { "epoch": 1.945748044599767, "grad_norm": 52.293373107910156, "learning_rate": 7.060390763765542e-05, "loss": 0.1838, "step": 2923 }, { "epoch": 1.946413712764187, "grad_norm": 1.9771173000335693, "learning_rate": 7.055950266429841e-05, "loss": 0.0033, "step": 2924 }, { "epoch": 1.947079380928607, "grad_norm": 22.27392578125, "learning_rate": 7.05150976909414e-05, "loss": 0.0345, "step": 2925 }, { "epoch": 1.9477450490930273, "grad_norm": 65.29061889648438, "learning_rate": 7.047069271758438e-05, "loss": 0.1664, "step": 2926 }, { "epoch": 1.9484107172574472, "grad_norm": 5.264341354370117, "learning_rate": 7.042628774422736e-05, "loss": 0.0091, "step": 2927 }, { "epoch": 1.9490763854218671, "grad_norm": 28.4490909576416, "learning_rate": 7.038188277087035e-05, "loss": 0.1536, "step": 2928 }, { "epoch": 1.9497420535862873, "grad_norm": 58.59207534790039, "learning_rate": 7.033747779751333e-05, "loss": 0.3871, "step": 2929 }, { "epoch": 1.9504077217507074, "grad_norm": 1.7915669679641724, "learning_rate": 7.029307282415632e-05, "loss": 0.0047, "step": 2930 }, { "epoch": 1.9510733899151274, "grad_norm": 50.42550277709961, "learning_rate": 7.024866785079929e-05, "loss": 0.8998, "step": 2931 }, { "epoch": 1.9517390580795473, "grad_norm": 83.81277465820312, "learning_rate": 7.020426287744227e-05, "loss": 0.5937, "step": 2932 }, { "epoch": 1.9524047262439674, "grad_norm": 45.88479995727539, "learning_rate": 7.015985790408526e-05, "loss": 0.3059, "step": 2933 }, { "epoch": 1.9530703944083874, "grad_norm": 73.91339111328125, "learning_rate": 7.011545293072824e-05, "loss": 0.1553, "step": 2934 }, { "epoch": 1.9537360625728075, "grad_norm": 8.387800216674805, "learning_rate": 7.007104795737123e-05, "loss": 0.0086, "step": 2935 }, { "epoch": 1.9544017307372274, "grad_norm": 33.290931701660156, "learning_rate": 7.002664298401421e-05, "loss": 0.3854, "step": 2936 }, { "epoch": 1.9550673989016474, "grad_norm": 10.637995719909668, "learning_rate": 6.99822380106572e-05, "loss": 0.0199, "step": 2937 }, { "epoch": 1.9557330670660675, "grad_norm": 96.36773681640625, "learning_rate": 6.993783303730018e-05, "loss": 0.2824, "step": 2938 }, { "epoch": 1.9563987352304877, "grad_norm": 133.07757568359375, "learning_rate": 6.989342806394317e-05, "loss": 0.4727, "step": 2939 }, { "epoch": 1.9570644033949076, "grad_norm": 45.42192459106445, "learning_rate": 6.984902309058615e-05, "loss": 0.0935, "step": 2940 }, { "epoch": 1.9577300715593275, "grad_norm": 46.302982330322266, "learning_rate": 6.980461811722914e-05, "loss": 0.3847, "step": 2941 }, { "epoch": 1.9583957397237477, "grad_norm": 46.72023010253906, "learning_rate": 6.976021314387212e-05, "loss": 0.3753, "step": 2942 }, { "epoch": 1.9590614078881678, "grad_norm": 27.948217391967773, "learning_rate": 6.97158081705151e-05, "loss": 0.0832, "step": 2943 }, { "epoch": 1.9597270760525878, "grad_norm": 16.38991928100586, "learning_rate": 6.967140319715808e-05, "loss": 0.0591, "step": 2944 }, { "epoch": 1.9603927442170077, "grad_norm": 98.27371215820312, "learning_rate": 6.962699822380106e-05, "loss": 0.6357, "step": 2945 }, { "epoch": 1.9610584123814279, "grad_norm": 25.49346160888672, "learning_rate": 6.958259325044405e-05, "loss": 0.1191, "step": 2946 }, { "epoch": 1.961724080545848, "grad_norm": 62.689537048339844, "learning_rate": 6.953818827708703e-05, "loss": 0.2816, "step": 2947 }, { "epoch": 1.962389748710268, "grad_norm": 26.70528221130371, "learning_rate": 6.949378330373002e-05, "loss": 0.213, "step": 2948 }, { "epoch": 1.9630554168746879, "grad_norm": 62.94496536254883, "learning_rate": 6.9449378330373e-05, "loss": 0.7538, "step": 2949 }, { "epoch": 1.963721085039108, "grad_norm": 16.40899085998535, "learning_rate": 6.940497335701599e-05, "loss": 0.0798, "step": 2950 }, { "epoch": 1.9643867532035282, "grad_norm": 10.490519523620605, "learning_rate": 6.936056838365897e-05, "loss": 0.0358, "step": 2951 }, { "epoch": 1.965052421367948, "grad_norm": 74.38216400146484, "learning_rate": 6.931616341030196e-05, "loss": 0.3374, "step": 2952 }, { "epoch": 1.965718089532368, "grad_norm": 22.063823699951172, "learning_rate": 6.927175843694494e-05, "loss": 0.0818, "step": 2953 }, { "epoch": 1.9663837576967882, "grad_norm": 74.95606994628906, "learning_rate": 6.922735346358793e-05, "loss": 0.3157, "step": 2954 }, { "epoch": 1.9670494258612083, "grad_norm": 50.645877838134766, "learning_rate": 6.918294849023091e-05, "loss": 0.7031, "step": 2955 }, { "epoch": 1.9677150940256283, "grad_norm": 17.89077377319336, "learning_rate": 6.913854351687388e-05, "loss": 0.0678, "step": 2956 }, { "epoch": 1.9683807621900482, "grad_norm": 21.998212814331055, "learning_rate": 6.909413854351687e-05, "loss": 0.1826, "step": 2957 }, { "epoch": 1.9690464303544681, "grad_norm": 3.3572635650634766, "learning_rate": 6.904973357015985e-05, "loss": 0.0129, "step": 2958 }, { "epoch": 1.9697120985188883, "grad_norm": 51.26792907714844, "learning_rate": 6.900532859680284e-05, "loss": 0.2118, "step": 2959 }, { "epoch": 1.9703777666833084, "grad_norm": 36.81552505493164, "learning_rate": 6.896092362344582e-05, "loss": 0.4629, "step": 2960 }, { "epoch": 1.9710434348477284, "grad_norm": 58.6724853515625, "learning_rate": 6.891651865008881e-05, "loss": 0.2603, "step": 2961 }, { "epoch": 1.9717091030121483, "grad_norm": 18.310096740722656, "learning_rate": 6.887211367673179e-05, "loss": 0.0441, "step": 2962 }, { "epoch": 1.9723747711765685, "grad_norm": 10.618471145629883, "learning_rate": 6.882770870337478e-05, "loss": 0.0337, "step": 2963 }, { "epoch": 1.9730404393409886, "grad_norm": 22.563739776611328, "learning_rate": 6.878330373001776e-05, "loss": 0.0674, "step": 2964 }, { "epoch": 1.9737061075054085, "grad_norm": 23.736507415771484, "learning_rate": 6.873889875666075e-05, "loss": 0.0672, "step": 2965 }, { "epoch": 1.9743717756698285, "grad_norm": 24.22038459777832, "learning_rate": 6.869449378330373e-05, "loss": 0.0732, "step": 2966 }, { "epoch": 1.9750374438342486, "grad_norm": 75.350341796875, "learning_rate": 6.865008880994672e-05, "loss": 0.645, "step": 2967 }, { "epoch": 1.9757031119986688, "grad_norm": 27.374385833740234, "learning_rate": 6.86056838365897e-05, "loss": 0.0237, "step": 2968 }, { "epoch": 1.9763687801630887, "grad_norm": 3.711238384246826, "learning_rate": 6.856127886323269e-05, "loss": 0.01, "step": 2969 }, { "epoch": 1.9770344483275086, "grad_norm": 28.171049118041992, "learning_rate": 6.851687388987567e-05, "loss": 0.542, "step": 2970 }, { "epoch": 1.9777001164919288, "grad_norm": 37.890995025634766, "learning_rate": 6.847246891651866e-05, "loss": 0.335, "step": 2971 }, { "epoch": 1.978365784656349, "grad_norm": 52.22409439086914, "learning_rate": 6.842806394316164e-05, "loss": 0.382, "step": 2972 }, { "epoch": 1.9790314528207689, "grad_norm": 92.54714965820312, "learning_rate": 6.838365896980463e-05, "loss": 0.8296, "step": 2973 }, { "epoch": 1.9796971209851888, "grad_norm": 55.86698532104492, "learning_rate": 6.833925399644761e-05, "loss": 0.5872, "step": 2974 }, { "epoch": 1.980362789149609, "grad_norm": 8.27837085723877, "learning_rate": 6.82948490230906e-05, "loss": 0.0359, "step": 2975 }, { "epoch": 1.981028457314029, "grad_norm": 51.28205871582031, "learning_rate": 6.825044404973358e-05, "loss": 0.1037, "step": 2976 }, { "epoch": 1.981694125478449, "grad_norm": 29.107093811035156, "learning_rate": 6.820603907637657e-05, "loss": 0.1935, "step": 2977 }, { "epoch": 1.982359793642869, "grad_norm": 25.65674591064453, "learning_rate": 6.816163410301955e-05, "loss": 0.0833, "step": 2978 }, { "epoch": 1.9830254618072891, "grad_norm": 76.73957824707031, "learning_rate": 6.811722912966253e-05, "loss": 0.4848, "step": 2979 }, { "epoch": 1.9836911299717093, "grad_norm": 7.989038467407227, "learning_rate": 6.807282415630552e-05, "loss": 0.0271, "step": 2980 }, { "epoch": 1.9843567981361292, "grad_norm": 67.47351837158203, "learning_rate": 6.802841918294849e-05, "loss": 0.2909, "step": 2981 }, { "epoch": 1.9850224663005491, "grad_norm": 26.227602005004883, "learning_rate": 6.798401420959148e-05, "loss": 0.1116, "step": 2982 }, { "epoch": 1.985688134464969, "grad_norm": 51.05128479003906, "learning_rate": 6.793960923623446e-05, "loss": 0.2999, "step": 2983 }, { "epoch": 1.9863538026293892, "grad_norm": 45.39921569824219, "learning_rate": 6.789520426287745e-05, "loss": 0.2467, "step": 2984 }, { "epoch": 1.9870194707938094, "grad_norm": 35.069889068603516, "learning_rate": 6.785079928952043e-05, "loss": 0.445, "step": 2985 }, { "epoch": 1.9876851389582293, "grad_norm": 48.22510528564453, "learning_rate": 6.780639431616341e-05, "loss": 0.1482, "step": 2986 }, { "epoch": 1.9883508071226492, "grad_norm": 18.450986862182617, "learning_rate": 6.77619893428064e-05, "loss": 0.0561, "step": 2987 }, { "epoch": 1.9890164752870694, "grad_norm": 43.702972412109375, "learning_rate": 6.771758436944938e-05, "loss": 0.3331, "step": 2988 }, { "epoch": 1.9896821434514895, "grad_norm": 43.212623596191406, "learning_rate": 6.767317939609237e-05, "loss": 0.081, "step": 2989 }, { "epoch": 1.9903478116159095, "grad_norm": 12.057387351989746, "learning_rate": 6.762877442273535e-05, "loss": 0.04, "step": 2990 }, { "epoch": 1.9910134797803294, "grad_norm": 55.46007537841797, "learning_rate": 6.758436944937834e-05, "loss": 0.0905, "step": 2991 }, { "epoch": 1.9916791479447495, "grad_norm": 71.17450714111328, "learning_rate": 6.753996447602132e-05, "loss": 0.3097, "step": 2992 }, { "epoch": 1.9923448161091697, "grad_norm": 24.00665283203125, "learning_rate": 6.74955595026643e-05, "loss": 0.1118, "step": 2993 }, { "epoch": 1.9930104842735896, "grad_norm": 44.63752746582031, "learning_rate": 6.745115452930728e-05, "loss": 0.1946, "step": 2994 }, { "epoch": 1.9936761524380096, "grad_norm": 31.687950134277344, "learning_rate": 6.740674955595026e-05, "loss": 0.1593, "step": 2995 }, { "epoch": 1.9943418206024297, "grad_norm": 40.35961151123047, "learning_rate": 6.736234458259325e-05, "loss": 0.1581, "step": 2996 }, { "epoch": 1.9950074887668499, "grad_norm": 63.913299560546875, "learning_rate": 6.731793960923623e-05, "loss": 0.3358, "step": 2997 }, { "epoch": 1.9956731569312698, "grad_norm": 50.905418395996094, "learning_rate": 6.727353463587922e-05, "loss": 0.293, "step": 2998 }, { "epoch": 1.9963388250956897, "grad_norm": 2.7470343112945557, "learning_rate": 6.72291296625222e-05, "loss": 0.0099, "step": 2999 }, { "epoch": 1.9970044932601099, "grad_norm": 91.21598052978516, "learning_rate": 6.718472468916519e-05, "loss": 0.2027, "step": 3000 }, { "epoch": 1.99767016142453, "grad_norm": 33.03544616699219, "learning_rate": 6.714031971580817e-05, "loss": 0.1131, "step": 3001 }, { "epoch": 1.99833582958895, "grad_norm": 56.328895568847656, "learning_rate": 6.709591474245116e-05, "loss": 0.2924, "step": 3002 }, { "epoch": 1.9990014977533699, "grad_norm": 42.984256744384766, "learning_rate": 6.705150976909414e-05, "loss": 0.4401, "step": 3003 }, { "epoch": 1.99966716591779, "grad_norm": 36.00129699707031, "learning_rate": 6.700710479573713e-05, "loss": 0.0906, "step": 3004 }, { "epoch": 2.00033283408221, "grad_norm": 24.757564544677734, "learning_rate": 6.696269982238011e-05, "loss": 0.2198, "step": 3005 }, { "epoch": 2.00099850224663, "grad_norm": 6.742340087890625, "learning_rate": 6.691829484902308e-05, "loss": 0.0275, "step": 3006 }, { "epoch": 2.00166417041105, "grad_norm": 6.727244853973389, "learning_rate": 6.687388987566607e-05, "loss": 0.0135, "step": 3007 }, { "epoch": 2.00232983857547, "grad_norm": 15.755208015441895, "learning_rate": 6.682948490230905e-05, "loss": 0.0483, "step": 3008 }, { "epoch": 2.0029955067398904, "grad_norm": 111.51408386230469, "learning_rate": 6.678507992895204e-05, "loss": 0.3166, "step": 3009 }, { "epoch": 2.0036611749043103, "grad_norm": 79.9453125, "learning_rate": 6.674067495559502e-05, "loss": 0.3846, "step": 3010 }, { "epoch": 2.00432684306873, "grad_norm": 3.69150710105896, "learning_rate": 6.669626998223801e-05, "loss": 0.023, "step": 3011 }, { "epoch": 2.00499251123315, "grad_norm": 5.576863765716553, "learning_rate": 6.6651865008881e-05, "loss": 0.0167, "step": 3012 }, { "epoch": 2.0056581793975705, "grad_norm": 10.010639190673828, "learning_rate": 6.660746003552398e-05, "loss": 0.0347, "step": 3013 }, { "epoch": 2.0063238475619904, "grad_norm": 22.074275970458984, "learning_rate": 6.656305506216696e-05, "loss": 0.0783, "step": 3014 }, { "epoch": 2.0069895157264104, "grad_norm": 33.93080520629883, "learning_rate": 6.651865008880995e-05, "loss": 0.1192, "step": 3015 }, { "epoch": 2.0076551838908303, "grad_norm": 49.922889709472656, "learning_rate": 6.647424511545293e-05, "loss": 0.5718, "step": 3016 }, { "epoch": 2.0083208520552502, "grad_norm": 103.0090103149414, "learning_rate": 6.642984014209592e-05, "loss": 0.2626, "step": 3017 }, { "epoch": 2.0089865202196706, "grad_norm": 32.448341369628906, "learning_rate": 6.63854351687389e-05, "loss": 0.3826, "step": 3018 }, { "epoch": 2.0096521883840905, "grad_norm": 0.4474305808544159, "learning_rate": 6.634103019538189e-05, "loss": 0.0011, "step": 3019 }, { "epoch": 2.0103178565485105, "grad_norm": 23.972492218017578, "learning_rate": 6.629662522202487e-05, "loss": 0.0501, "step": 3020 }, { "epoch": 2.0109835247129304, "grad_norm": 29.796436309814453, "learning_rate": 6.625222024866786e-05, "loss": 0.0857, "step": 3021 }, { "epoch": 2.011649192877351, "grad_norm": 42.373023986816406, "learning_rate": 6.620781527531084e-05, "loss": 0.1659, "step": 3022 }, { "epoch": 2.0123148610417707, "grad_norm": 44.73994445800781, "learning_rate": 6.616341030195383e-05, "loss": 0.2233, "step": 3023 }, { "epoch": 2.0129805292061906, "grad_norm": 2.534695625305176, "learning_rate": 6.611900532859681e-05, "loss": 0.0073, "step": 3024 }, { "epoch": 2.0136461973706106, "grad_norm": 2.861107110977173, "learning_rate": 6.60746003552398e-05, "loss": 0.0086, "step": 3025 }, { "epoch": 2.014311865535031, "grad_norm": 3.224713087081909, "learning_rate": 6.603019538188278e-05, "loss": 0.0079, "step": 3026 }, { "epoch": 2.014977533699451, "grad_norm": 39.02479553222656, "learning_rate": 6.598579040852577e-05, "loss": 0.2043, "step": 3027 }, { "epoch": 2.015643201863871, "grad_norm": 45.73806381225586, "learning_rate": 6.594138543516875e-05, "loss": 0.1223, "step": 3028 }, { "epoch": 2.0163088700282907, "grad_norm": 40.077823638916016, "learning_rate": 6.589698046181174e-05, "loss": 0.1773, "step": 3029 }, { "epoch": 2.016974538192711, "grad_norm": 5.455604553222656, "learning_rate": 6.585257548845472e-05, "loss": 0.0115, "step": 3030 }, { "epoch": 2.017640206357131, "grad_norm": 7.709271430969238, "learning_rate": 6.580817051509769e-05, "loss": 0.0152, "step": 3031 }, { "epoch": 2.018305874521551, "grad_norm": 1.6229524612426758, "learning_rate": 6.576376554174068e-05, "loss": 0.0024, "step": 3032 }, { "epoch": 2.018971542685971, "grad_norm": 0.054127246141433716, "learning_rate": 6.571936056838366e-05, "loss": 0.0002, "step": 3033 }, { "epoch": 2.0196372108503913, "grad_norm": 11.79362678527832, "learning_rate": 6.567495559502665e-05, "loss": 0.0335, "step": 3034 }, { "epoch": 2.020302879014811, "grad_norm": 5.991158485412598, "learning_rate": 6.563055062166963e-05, "loss": 0.0119, "step": 3035 }, { "epoch": 2.020968547179231, "grad_norm": 20.796621322631836, "learning_rate": 6.558614564831262e-05, "loss": 0.0576, "step": 3036 }, { "epoch": 2.021634215343651, "grad_norm": 5.646274089813232, "learning_rate": 6.55417406749556e-05, "loss": 0.0134, "step": 3037 }, { "epoch": 2.0222998835080714, "grad_norm": 44.90586471557617, "learning_rate": 6.549733570159859e-05, "loss": 0.3229, "step": 3038 }, { "epoch": 2.0229655516724914, "grad_norm": 74.36479187011719, "learning_rate": 6.545293072824157e-05, "loss": 0.3941, "step": 3039 }, { "epoch": 2.0236312198369113, "grad_norm": 33.04402542114258, "learning_rate": 6.540852575488456e-05, "loss": 0.4977, "step": 3040 }, { "epoch": 2.0242968880013312, "grad_norm": 44.09501647949219, "learning_rate": 6.536412078152754e-05, "loss": 0.167, "step": 3041 }, { "epoch": 2.024962556165751, "grad_norm": 3.174835205078125, "learning_rate": 6.531971580817053e-05, "loss": 0.0063, "step": 3042 }, { "epoch": 2.0256282243301715, "grad_norm": 1.58350670337677, "learning_rate": 6.52753108348135e-05, "loss": 0.0028, "step": 3043 }, { "epoch": 2.0262938924945915, "grad_norm": 40.45137023925781, "learning_rate": 6.523090586145648e-05, "loss": 0.0941, "step": 3044 }, { "epoch": 2.0269595606590114, "grad_norm": 61.793148040771484, "learning_rate": 6.518650088809947e-05, "loss": 0.0877, "step": 3045 }, { "epoch": 2.0276252288234313, "grad_norm": 48.17112350463867, "learning_rate": 6.514209591474245e-05, "loss": 0.3291, "step": 3046 }, { "epoch": 2.0282908969878517, "grad_norm": 3.12809681892395, "learning_rate": 6.509769094138544e-05, "loss": 0.0072, "step": 3047 }, { "epoch": 2.0289565651522716, "grad_norm": 1.3230597972869873, "learning_rate": 6.505328596802842e-05, "loss": 0.0019, "step": 3048 }, { "epoch": 2.0296222333166916, "grad_norm": 6.544448375701904, "learning_rate": 6.50088809946714e-05, "loss": 0.0119, "step": 3049 }, { "epoch": 2.0302879014811115, "grad_norm": 2.088547706604004, "learning_rate": 6.496447602131439e-05, "loss": 0.0042, "step": 3050 }, { "epoch": 2.030953569645532, "grad_norm": 18.167999267578125, "learning_rate": 6.492007104795738e-05, "loss": 0.036, "step": 3051 }, { "epoch": 2.031619237809952, "grad_norm": 34.093360900878906, "learning_rate": 6.487566607460036e-05, "loss": 0.2126, "step": 3052 }, { "epoch": 2.0322849059743717, "grad_norm": 7.602128028869629, "learning_rate": 6.483126110124334e-05, "loss": 0.0093, "step": 3053 }, { "epoch": 2.0329505741387917, "grad_norm": 32.118194580078125, "learning_rate": 6.478685612788633e-05, "loss": 0.0629, "step": 3054 }, { "epoch": 2.033616242303212, "grad_norm": 2.1596267223358154, "learning_rate": 6.474245115452931e-05, "loss": 0.0035, "step": 3055 }, { "epoch": 2.034281910467632, "grad_norm": 1.1355314254760742, "learning_rate": 6.469804618117229e-05, "loss": 0.0024, "step": 3056 }, { "epoch": 2.034947578632052, "grad_norm": 9.863322257995605, "learning_rate": 6.465364120781527e-05, "loss": 0.0151, "step": 3057 }, { "epoch": 2.035613246796472, "grad_norm": 3.6854867935180664, "learning_rate": 6.460923623445826e-05, "loss": 0.0051, "step": 3058 }, { "epoch": 2.036278914960892, "grad_norm": 0.1391722410917282, "learning_rate": 6.456483126110124e-05, "loss": 0.0004, "step": 3059 }, { "epoch": 2.036944583125312, "grad_norm": 20.485607147216797, "learning_rate": 6.452042628774423e-05, "loss": 0.0295, "step": 3060 }, { "epoch": 2.037610251289732, "grad_norm": 0.16033516824245453, "learning_rate": 6.447602131438721e-05, "loss": 0.0002, "step": 3061 }, { "epoch": 2.038275919454152, "grad_norm": 49.2470588684082, "learning_rate": 6.44316163410302e-05, "loss": 0.1846, "step": 3062 }, { "epoch": 2.0389415876185724, "grad_norm": 0.09907753765583038, "learning_rate": 6.438721136767318e-05, "loss": 0.0002, "step": 3063 }, { "epoch": 2.0396072557829923, "grad_norm": 0.0022163838148117065, "learning_rate": 6.434280639431616e-05, "loss": 0.0, "step": 3064 }, { "epoch": 2.040272923947412, "grad_norm": 45.06425857543945, "learning_rate": 6.429840142095915e-05, "loss": 0.1071, "step": 3065 }, { "epoch": 2.040938592111832, "grad_norm": 3.3526573181152344, "learning_rate": 6.425399644760213e-05, "loss": 0.0041, "step": 3066 }, { "epoch": 2.041604260276252, "grad_norm": 148.12774658203125, "learning_rate": 6.420959147424512e-05, "loss": 0.132, "step": 3067 }, { "epoch": 2.0422699284406725, "grad_norm": 0.3530925512313843, "learning_rate": 6.416518650088809e-05, "loss": 0.0008, "step": 3068 }, { "epoch": 2.0429355966050924, "grad_norm": 1.4099760055541992, "learning_rate": 6.412078152753109e-05, "loss": 0.0021, "step": 3069 }, { "epoch": 2.0436012647695123, "grad_norm": 81.7883071899414, "learning_rate": 6.407637655417407e-05, "loss": 1.2438, "step": 3070 }, { "epoch": 2.0442669329339322, "grad_norm": 0.5354982018470764, "learning_rate": 6.403197158081706e-05, "loss": 0.0013, "step": 3071 }, { "epoch": 2.0449326010983526, "grad_norm": 3.446418523788452, "learning_rate": 6.398756660746004e-05, "loss": 0.0032, "step": 3072 }, { "epoch": 2.0455982692627726, "grad_norm": 0.19478213787078857, "learning_rate": 6.394316163410303e-05, "loss": 0.0002, "step": 3073 }, { "epoch": 2.0462639374271925, "grad_norm": 0.10158519446849823, "learning_rate": 6.389875666074601e-05, "loss": 0.0002, "step": 3074 }, { "epoch": 2.0469296055916124, "grad_norm": 92.99313354492188, "learning_rate": 6.3854351687389e-05, "loss": 0.174, "step": 3075 }, { "epoch": 2.047595273756033, "grad_norm": 0.7470491528511047, "learning_rate": 6.380994671403198e-05, "loss": 0.0006, "step": 3076 }, { "epoch": 2.0482609419204527, "grad_norm": 0.3984420895576477, "learning_rate": 6.376554174067497e-05, "loss": 0.0003, "step": 3077 }, { "epoch": 2.0489266100848726, "grad_norm": 0.9849663376808167, "learning_rate": 6.372113676731795e-05, "loss": 0.0015, "step": 3078 }, { "epoch": 2.0495922782492926, "grad_norm": 21.70440101623535, "learning_rate": 6.367673179396094e-05, "loss": 0.034, "step": 3079 }, { "epoch": 2.050257946413713, "grad_norm": 0.056247320026159286, "learning_rate": 6.363232682060392e-05, "loss": 0.0001, "step": 3080 }, { "epoch": 2.050923614578133, "grad_norm": 0.0022599915973842144, "learning_rate": 6.358792184724689e-05, "loss": 0.0, "step": 3081 }, { "epoch": 2.051589282742553, "grad_norm": 22.711776733398438, "learning_rate": 6.354351687388988e-05, "loss": 0.022, "step": 3082 }, { "epoch": 2.0522549509069727, "grad_norm": 155.9021453857422, "learning_rate": 6.349911190053286e-05, "loss": 1.1014, "step": 3083 }, { "epoch": 2.052920619071393, "grad_norm": 30.160184860229492, "learning_rate": 6.345470692717585e-05, "loss": 0.0467, "step": 3084 }, { "epoch": 2.053586287235813, "grad_norm": 0.9574832320213318, "learning_rate": 6.341030195381883e-05, "loss": 0.0011, "step": 3085 }, { "epoch": 2.054251955400233, "grad_norm": 0.0166773684322834, "learning_rate": 6.336589698046182e-05, "loss": 0.0, "step": 3086 }, { "epoch": 2.054917623564653, "grad_norm": 0.013485881499946117, "learning_rate": 6.33214920071048e-05, "loss": 0.0, "step": 3087 }, { "epoch": 2.0555832917290733, "grad_norm": 137.52781677246094, "learning_rate": 6.327708703374779e-05, "loss": 0.2715, "step": 3088 }, { "epoch": 2.056248959893493, "grad_norm": 0.0010899835033342242, "learning_rate": 6.323268206039077e-05, "loss": 0.0, "step": 3089 }, { "epoch": 2.056914628057913, "grad_norm": 1.2061229944229126, "learning_rate": 6.318827708703376e-05, "loss": 0.0012, "step": 3090 }, { "epoch": 2.057580296222333, "grad_norm": 0.019467955455183983, "learning_rate": 6.314387211367674e-05, "loss": 0.0, "step": 3091 }, { "epoch": 2.058245964386753, "grad_norm": 127.95604705810547, "learning_rate": 6.309946714031973e-05, "loss": 1.6085, "step": 3092 }, { "epoch": 2.0589116325511734, "grad_norm": 0.054994650185108185, "learning_rate": 6.30550621669627e-05, "loss": 0.0001, "step": 3093 }, { "epoch": 2.0595773007155933, "grad_norm": 0.03873397037386894, "learning_rate": 6.301065719360568e-05, "loss": 0.0001, "step": 3094 }, { "epoch": 2.0602429688800132, "grad_norm": 0.006200865376740694, "learning_rate": 6.296625222024867e-05, "loss": 0.0, "step": 3095 }, { "epoch": 2.060908637044433, "grad_norm": 0.2527855336666107, "learning_rate": 6.292184724689165e-05, "loss": 0.0005, "step": 3096 }, { "epoch": 2.0615743052088535, "grad_norm": 133.4344024658203, "learning_rate": 6.287744227353464e-05, "loss": 0.1056, "step": 3097 }, { "epoch": 2.0622399733732735, "grad_norm": 13.203033447265625, "learning_rate": 6.283303730017762e-05, "loss": 0.0044, "step": 3098 }, { "epoch": 2.0629056415376934, "grad_norm": 83.00621795654297, "learning_rate": 6.27886323268206e-05, "loss": 0.5364, "step": 3099 }, { "epoch": 2.0635713097021133, "grad_norm": 0.02336590364575386, "learning_rate": 6.274422735346359e-05, "loss": 0.0, "step": 3100 }, { "epoch": 2.0642369778665337, "grad_norm": 0.04199038818478584, "learning_rate": 6.269982238010658e-05, "loss": 0.0001, "step": 3101 }, { "epoch": 2.0649026460309536, "grad_norm": 0.015436958521604538, "learning_rate": 6.265541740674956e-05, "loss": 0.0, "step": 3102 }, { "epoch": 2.0655683141953736, "grad_norm": 63.20383834838867, "learning_rate": 6.261101243339255e-05, "loss": 0.1689, "step": 3103 }, { "epoch": 2.0662339823597935, "grad_norm": 160.21482849121094, "learning_rate": 6.256660746003553e-05, "loss": 0.0931, "step": 3104 }, { "epoch": 2.066899650524214, "grad_norm": 0.14238587021827698, "learning_rate": 6.252220248667852e-05, "loss": 0.0003, "step": 3105 }, { "epoch": 2.067565318688634, "grad_norm": 32.09461975097656, "learning_rate": 6.247779751332149e-05, "loss": 0.0375, "step": 3106 }, { "epoch": 2.0682309868530537, "grad_norm": 51.706050872802734, "learning_rate": 6.243339253996447e-05, "loss": 0.594, "step": 3107 }, { "epoch": 2.0688966550174737, "grad_norm": 0.014732087031006813, "learning_rate": 6.238898756660746e-05, "loss": 0.0, "step": 3108 }, { "epoch": 2.069562323181894, "grad_norm": 273.5494384765625, "learning_rate": 6.234458259325044e-05, "loss": 0.6999, "step": 3109 }, { "epoch": 2.070227991346314, "grad_norm": 0.09334751218557358, "learning_rate": 6.230017761989343e-05, "loss": 0.0001, "step": 3110 }, { "epoch": 2.070893659510734, "grad_norm": 0.33053353428840637, "learning_rate": 6.225577264653641e-05, "loss": 0.0002, "step": 3111 }, { "epoch": 2.071559327675154, "grad_norm": 62.7347412109375, "learning_rate": 6.22113676731794e-05, "loss": 0.1977, "step": 3112 }, { "epoch": 2.0722249958395738, "grad_norm": 14.21413803100586, "learning_rate": 6.216696269982238e-05, "loss": 0.0065, "step": 3113 }, { "epoch": 2.072890664003994, "grad_norm": 0.1861925572156906, "learning_rate": 6.212255772646537e-05, "loss": 0.0002, "step": 3114 }, { "epoch": 2.073556332168414, "grad_norm": 75.09960174560547, "learning_rate": 6.207815275310835e-05, "loss": 0.2267, "step": 3115 }, { "epoch": 2.074222000332834, "grad_norm": 12.200486183166504, "learning_rate": 6.203374777975134e-05, "loss": 0.0095, "step": 3116 }, { "epoch": 2.074887668497254, "grad_norm": 0.9586191177368164, "learning_rate": 6.198934280639432e-05, "loss": 0.001, "step": 3117 }, { "epoch": 2.0755533366616743, "grad_norm": 0.2495388239622116, "learning_rate": 6.19449378330373e-05, "loss": 0.0004, "step": 3118 }, { "epoch": 2.0762190048260942, "grad_norm": 47.819435119628906, "learning_rate": 6.190053285968028e-05, "loss": 0.1081, "step": 3119 }, { "epoch": 2.076884672990514, "grad_norm": 60.292667388916016, "learning_rate": 6.185612788632326e-05, "loss": 0.053, "step": 3120 }, { "epoch": 2.077550341154934, "grad_norm": 147.84869384765625, "learning_rate": 6.181172291296625e-05, "loss": 0.2529, "step": 3121 }, { "epoch": 2.0782160093193545, "grad_norm": 18.114715576171875, "learning_rate": 6.176731793960923e-05, "loss": 0.0125, "step": 3122 }, { "epoch": 2.0788816774837744, "grad_norm": 12.642441749572754, "learning_rate": 6.172291296625222e-05, "loss": 0.0117, "step": 3123 }, { "epoch": 2.0795473456481943, "grad_norm": 0.2666056156158447, "learning_rate": 6.16785079928952e-05, "loss": 0.0003, "step": 3124 }, { "epoch": 2.0802130138126143, "grad_norm": 0.027772117406129837, "learning_rate": 6.163410301953819e-05, "loss": 0.0001, "step": 3125 }, { "epoch": 2.0808786819770346, "grad_norm": 3.1628427505493164, "learning_rate": 6.158969804618117e-05, "loss": 0.0044, "step": 3126 }, { "epoch": 2.0815443501414546, "grad_norm": 111.79324340820312, "learning_rate": 6.154529307282415e-05, "loss": 0.1923, "step": 3127 }, { "epoch": 2.0822100183058745, "grad_norm": 33.60985565185547, "learning_rate": 6.150088809946715e-05, "loss": 0.0288, "step": 3128 }, { "epoch": 2.0828756864702944, "grad_norm": 0.042768023908138275, "learning_rate": 6.145648312611014e-05, "loss": 0.0001, "step": 3129 }, { "epoch": 2.083541354634715, "grad_norm": 68.28672790527344, "learning_rate": 6.141207815275312e-05, "loss": 0.127, "step": 3130 }, { "epoch": 2.0842070227991347, "grad_norm": 21.603004455566406, "learning_rate": 6.13676731793961e-05, "loss": 0.0246, "step": 3131 }, { "epoch": 2.0848726909635547, "grad_norm": 79.04849243164062, "learning_rate": 6.132326820603908e-05, "loss": 0.9378, "step": 3132 }, { "epoch": 2.0855383591279746, "grad_norm": 0.11011223495006561, "learning_rate": 6.127886323268206e-05, "loss": 0.0001, "step": 3133 }, { "epoch": 2.0862040272923945, "grad_norm": 0.014358668588101864, "learning_rate": 6.123445825932505e-05, "loss": 0.0, "step": 3134 }, { "epoch": 2.086869695456815, "grad_norm": 1.2764637470245361, "learning_rate": 6.119005328596803e-05, "loss": 0.0007, "step": 3135 }, { "epoch": 2.087535363621235, "grad_norm": 42.57699966430664, "learning_rate": 6.114564831261102e-05, "loss": 0.0777, "step": 3136 }, { "epoch": 2.0882010317856547, "grad_norm": 188.3223419189453, "learning_rate": 6.1101243339254e-05, "loss": 0.5125, "step": 3137 }, { "epoch": 2.0888666999500747, "grad_norm": 2.5027549266815186, "learning_rate": 6.105683836589699e-05, "loss": 0.0028, "step": 3138 }, { "epoch": 2.089532368114495, "grad_norm": 0.0883927121758461, "learning_rate": 6.1012433392539966e-05, "loss": 0.0002, "step": 3139 }, { "epoch": 2.090198036278915, "grad_norm": 57.915321350097656, "learning_rate": 6.096802841918295e-05, "loss": 0.1946, "step": 3140 }, { "epoch": 2.090863704443335, "grad_norm": 121.11367797851562, "learning_rate": 6.0923623445825936e-05, "loss": 0.6383, "step": 3141 }, { "epoch": 2.091529372607755, "grad_norm": 0.20855890214443207, "learning_rate": 6.087921847246892e-05, "loss": 0.0002, "step": 3142 }, { "epoch": 2.092195040772175, "grad_norm": 2.5594370365142822, "learning_rate": 6.083481349911191e-05, "loss": 0.0019, "step": 3143 }, { "epoch": 2.092860708936595, "grad_norm": 55.6262321472168, "learning_rate": 6.0790408525754884e-05, "loss": 0.0214, "step": 3144 }, { "epoch": 2.093526377101015, "grad_norm": 14.289521217346191, "learning_rate": 6.074600355239787e-05, "loss": 0.0185, "step": 3145 }, { "epoch": 2.094192045265435, "grad_norm": 10.727511405944824, "learning_rate": 6.070159857904085e-05, "loss": 0.0203, "step": 3146 }, { "epoch": 2.0948577134298554, "grad_norm": 123.15850067138672, "learning_rate": 6.065719360568384e-05, "loss": 0.8844, "step": 3147 }, { "epoch": 2.0955233815942753, "grad_norm": 93.24414825439453, "learning_rate": 6.061278863232682e-05, "loss": 0.1446, "step": 3148 }, { "epoch": 2.0961890497586952, "grad_norm": 4.274571895599365, "learning_rate": 6.056838365896981e-05, "loss": 0.0036, "step": 3149 }, { "epoch": 2.096854717923115, "grad_norm": 73.74907684326172, "learning_rate": 6.052397868561279e-05, "loss": 0.1908, "step": 3150 }, { "epoch": 2.0975203860875355, "grad_norm": 71.38188171386719, "learning_rate": 6.047957371225578e-05, "loss": 0.2344, "step": 3151 }, { "epoch": 2.0981860542519555, "grad_norm": 98.9805908203125, "learning_rate": 6.043516873889876e-05, "loss": 0.2971, "step": 3152 }, { "epoch": 2.0988517224163754, "grad_norm": 0.232648566365242, "learning_rate": 6.039076376554175e-05, "loss": 0.0002, "step": 3153 }, { "epoch": 2.0995173905807953, "grad_norm": 55.20022201538086, "learning_rate": 6.034635879218473e-05, "loss": 0.0815, "step": 3154 }, { "epoch": 2.1001830587452157, "grad_norm": 4.079741477966309, "learning_rate": 6.030195381882772e-05, "loss": 0.004, "step": 3155 }, { "epoch": 2.1008487269096356, "grad_norm": 74.40824127197266, "learning_rate": 6.025754884547069e-05, "loss": 0.1306, "step": 3156 }, { "epoch": 2.1015143950740556, "grad_norm": 89.30941772460938, "learning_rate": 6.021314387211367e-05, "loss": 0.129, "step": 3157 }, { "epoch": 2.1021800632384755, "grad_norm": 0.020798690617084503, "learning_rate": 6.016873889875666e-05, "loss": 0.0, "step": 3158 }, { "epoch": 2.102845731402896, "grad_norm": 0.01616029627621174, "learning_rate": 6.012433392539964e-05, "loss": 0.0, "step": 3159 }, { "epoch": 2.103511399567316, "grad_norm": 63.5081901550293, "learning_rate": 6.007992895204263e-05, "loss": 0.3077, "step": 3160 }, { "epoch": 2.1041770677317357, "grad_norm": 9.187117576599121, "learning_rate": 6.003552397868561e-05, "loss": 0.0123, "step": 3161 }, { "epoch": 2.1048427358961557, "grad_norm": 76.72493743896484, "learning_rate": 5.99911190053286e-05, "loss": 0.1789, "step": 3162 }, { "epoch": 2.1055084040605756, "grad_norm": 0.6776677966117859, "learning_rate": 5.994671403197158e-05, "loss": 0.0007, "step": 3163 }, { "epoch": 2.106174072224996, "grad_norm": 42.19890213012695, "learning_rate": 5.990230905861457e-05, "loss": 0.06, "step": 3164 }, { "epoch": 2.106839740389416, "grad_norm": 44.69881057739258, "learning_rate": 5.985790408525755e-05, "loss": 0.0556, "step": 3165 }, { "epoch": 2.107505408553836, "grad_norm": 51.004451751708984, "learning_rate": 5.9813499111900536e-05, "loss": 0.0948, "step": 3166 }, { "epoch": 2.1081710767182558, "grad_norm": 49.30232620239258, "learning_rate": 5.976909413854352e-05, "loss": 0.4559, "step": 3167 }, { "epoch": 2.108836744882676, "grad_norm": 0.1439058929681778, "learning_rate": 5.9724689165186506e-05, "loss": 0.0002, "step": 3168 }, { "epoch": 2.109502413047096, "grad_norm": 1.3022292852401733, "learning_rate": 5.9680284191829484e-05, "loss": 0.0014, "step": 3169 }, { "epoch": 2.110168081211516, "grad_norm": 52.462345123291016, "learning_rate": 5.963587921847247e-05, "loss": 0.1046, "step": 3170 }, { "epoch": 2.110833749375936, "grad_norm": 60.84833908081055, "learning_rate": 5.9591474245115454e-05, "loss": 0.3525, "step": 3171 }, { "epoch": 2.1114994175403563, "grad_norm": 15.848183631896973, "learning_rate": 5.954706927175844e-05, "loss": 0.0226, "step": 3172 }, { "epoch": 2.1121650857047762, "grad_norm": 19.52475357055664, "learning_rate": 5.9502664298401424e-05, "loss": 0.0215, "step": 3173 }, { "epoch": 2.112830753869196, "grad_norm": 74.64434051513672, "learning_rate": 5.945825932504441e-05, "loss": 0.1695, "step": 3174 }, { "epoch": 2.113496422033616, "grad_norm": 129.00830078125, "learning_rate": 5.941385435168739e-05, "loss": 0.3409, "step": 3175 }, { "epoch": 2.1141620901980365, "grad_norm": 123.70621490478516, "learning_rate": 5.936944937833038e-05, "loss": 0.3179, "step": 3176 }, { "epoch": 2.1148277583624564, "grad_norm": 43.59246063232422, "learning_rate": 5.932504440497336e-05, "loss": 0.1596, "step": 3177 }, { "epoch": 2.1154934265268763, "grad_norm": 59.7046012878418, "learning_rate": 5.928063943161635e-05, "loss": 0.241, "step": 3178 }, { "epoch": 2.1161590946912963, "grad_norm": 1.0017504692077637, "learning_rate": 5.923623445825933e-05, "loss": 0.0019, "step": 3179 }, { "epoch": 2.1168247628557166, "grad_norm": 25.24599266052246, "learning_rate": 5.919182948490232e-05, "loss": 0.0535, "step": 3180 }, { "epoch": 2.1174904310201366, "grad_norm": 3.5230231285095215, "learning_rate": 5.914742451154529e-05, "loss": 0.0049, "step": 3181 }, { "epoch": 2.1181560991845565, "grad_norm": 69.81744384765625, "learning_rate": 5.9103019538188274e-05, "loss": 0.2114, "step": 3182 }, { "epoch": 2.1188217673489764, "grad_norm": 80.32762145996094, "learning_rate": 5.905861456483126e-05, "loss": 0.1762, "step": 3183 }, { "epoch": 2.1194874355133964, "grad_norm": 43.15791702270508, "learning_rate": 5.901420959147424e-05, "loss": 0.0393, "step": 3184 }, { "epoch": 2.1201531036778167, "grad_norm": 2.8857414722442627, "learning_rate": 5.896980461811723e-05, "loss": 0.0044, "step": 3185 }, { "epoch": 2.1208187718422367, "grad_norm": 47.25928497314453, "learning_rate": 5.892539964476021e-05, "loss": 0.2685, "step": 3186 }, { "epoch": 2.1214844400066566, "grad_norm": 12.276843070983887, "learning_rate": 5.88809946714032e-05, "loss": 0.0188, "step": 3187 }, { "epoch": 2.1221501081710765, "grad_norm": 14.323443412780762, "learning_rate": 5.883658969804618e-05, "loss": 0.019, "step": 3188 }, { "epoch": 2.122815776335497, "grad_norm": 2.072903633117676, "learning_rate": 5.879218472468917e-05, "loss": 0.0031, "step": 3189 }, { "epoch": 2.123481444499917, "grad_norm": 0.02759326808154583, "learning_rate": 5.874777975133215e-05, "loss": 0.0001, "step": 3190 }, { "epoch": 2.1241471126643368, "grad_norm": 2.5400266647338867, "learning_rate": 5.870337477797514e-05, "loss": 0.0025, "step": 3191 }, { "epoch": 2.1248127808287567, "grad_norm": 89.05103302001953, "learning_rate": 5.865896980461812e-05, "loss": 0.4793, "step": 3192 }, { "epoch": 2.125478448993177, "grad_norm": 0.060356978327035904, "learning_rate": 5.861456483126111e-05, "loss": 0.0001, "step": 3193 }, { "epoch": 2.126144117157597, "grad_norm": 29.305280685424805, "learning_rate": 5.8570159857904085e-05, "loss": 0.0338, "step": 3194 }, { "epoch": 2.126809785322017, "grad_norm": 50.31920623779297, "learning_rate": 5.852575488454707e-05, "loss": 0.322, "step": 3195 }, { "epoch": 2.127475453486437, "grad_norm": 81.76736450195312, "learning_rate": 5.8481349911190054e-05, "loss": 0.1505, "step": 3196 }, { "epoch": 2.1281411216508572, "grad_norm": 48.77127456665039, "learning_rate": 5.843694493783304e-05, "loss": 0.1587, "step": 3197 }, { "epoch": 2.128806789815277, "grad_norm": 0.34772059321403503, "learning_rate": 5.8392539964476024e-05, "loss": 0.0005, "step": 3198 }, { "epoch": 2.129472457979697, "grad_norm": 55.01288986206055, "learning_rate": 5.834813499111901e-05, "loss": 0.2623, "step": 3199 }, { "epoch": 2.130138126144117, "grad_norm": 1.2655458450317383, "learning_rate": 5.8303730017761994e-05, "loss": 0.0015, "step": 3200 }, { "epoch": 2.1308037943085374, "grad_norm": 3.0290145874023438, "learning_rate": 5.825932504440498e-05, "loss": 0.0032, "step": 3201 }, { "epoch": 2.1314694624729573, "grad_norm": 38.32987976074219, "learning_rate": 5.8214920071047963e-05, "loss": 0.5191, "step": 3202 }, { "epoch": 2.1321351306373773, "grad_norm": 87.5805435180664, "learning_rate": 5.817051509769095e-05, "loss": 0.4812, "step": 3203 }, { "epoch": 2.132800798801797, "grad_norm": 0.4613059163093567, "learning_rate": 5.812611012433393e-05, "loss": 0.0008, "step": 3204 }, { "epoch": 2.133466466966217, "grad_norm": 0.05957324430346489, "learning_rate": 5.808170515097692e-05, "loss": 0.0001, "step": 3205 }, { "epoch": 2.1341321351306375, "grad_norm": 0.4088250994682312, "learning_rate": 5.803730017761989e-05, "loss": 0.0007, "step": 3206 }, { "epoch": 2.1347978032950574, "grad_norm": 97.16470336914062, "learning_rate": 5.7992895204262874e-05, "loss": 0.306, "step": 3207 }, { "epoch": 2.1354634714594773, "grad_norm": 7.639748573303223, "learning_rate": 5.794849023090586e-05, "loss": 0.004, "step": 3208 }, { "epoch": 2.1361291396238977, "grad_norm": 52.792903900146484, "learning_rate": 5.7904085257548844e-05, "loss": 0.1258, "step": 3209 }, { "epoch": 2.1367948077883177, "grad_norm": 0.06396713852882385, "learning_rate": 5.785968028419183e-05, "loss": 0.0001, "step": 3210 }, { "epoch": 2.1374604759527376, "grad_norm": 1.7327839136123657, "learning_rate": 5.7815275310834813e-05, "loss": 0.0019, "step": 3211 }, { "epoch": 2.1381261441171575, "grad_norm": 0.07871033251285553, "learning_rate": 5.77708703374778e-05, "loss": 0.0001, "step": 3212 }, { "epoch": 2.1387918122815774, "grad_norm": 51.73154067993164, "learning_rate": 5.772646536412078e-05, "loss": 0.4961, "step": 3213 }, { "epoch": 2.139457480445998, "grad_norm": 21.93402099609375, "learning_rate": 5.768206039076377e-05, "loss": 0.0176, "step": 3214 }, { "epoch": 2.1401231486104177, "grad_norm": 0.3303959369659424, "learning_rate": 5.763765541740675e-05, "loss": 0.0006, "step": 3215 }, { "epoch": 2.1407888167748377, "grad_norm": 7.288761615753174, "learning_rate": 5.759325044404974e-05, "loss": 0.0118, "step": 3216 }, { "epoch": 2.1414544849392576, "grad_norm": 17.588436126708984, "learning_rate": 5.754884547069272e-05, "loss": 0.0209, "step": 3217 }, { "epoch": 2.142120153103678, "grad_norm": 61.35725402832031, "learning_rate": 5.750444049733571e-05, "loss": 0.4426, "step": 3218 }, { "epoch": 2.142785821268098, "grad_norm": 0.0260478425770998, "learning_rate": 5.7460035523978685e-05, "loss": 0.0001, "step": 3219 }, { "epoch": 2.143451489432518, "grad_norm": 0.0860922709107399, "learning_rate": 5.741563055062167e-05, "loss": 0.0001, "step": 3220 }, { "epoch": 2.1441171575969378, "grad_norm": 42.49921417236328, "learning_rate": 5.7371225577264655e-05, "loss": 0.0084, "step": 3221 }, { "epoch": 2.144782825761358, "grad_norm": 0.07198088616132736, "learning_rate": 5.732682060390764e-05, "loss": 0.0001, "step": 3222 }, { "epoch": 2.145448493925778, "grad_norm": 5.857786655426025, "learning_rate": 5.7282415630550625e-05, "loss": 0.0101, "step": 3223 }, { "epoch": 2.146114162090198, "grad_norm": 8.330015182495117, "learning_rate": 5.723801065719361e-05, "loss": 0.0077, "step": 3224 }, { "epoch": 2.146779830254618, "grad_norm": 8.961620330810547, "learning_rate": 5.7193605683836594e-05, "loss": 0.0095, "step": 3225 }, { "epoch": 2.1474454984190383, "grad_norm": 0.2516133785247803, "learning_rate": 5.714920071047958e-05, "loss": 0.0004, "step": 3226 }, { "epoch": 2.1481111665834582, "grad_norm": 14.264823913574219, "learning_rate": 5.7104795737122564e-05, "loss": 0.0069, "step": 3227 }, { "epoch": 2.148776834747878, "grad_norm": 11.518255233764648, "learning_rate": 5.706039076376555e-05, "loss": 0.0148, "step": 3228 }, { "epoch": 2.149442502912298, "grad_norm": 0.15960028767585754, "learning_rate": 5.7015985790408534e-05, "loss": 0.0004, "step": 3229 }, { "epoch": 2.1501081710767185, "grad_norm": 26.888429641723633, "learning_rate": 5.697158081705152e-05, "loss": 0.0491, "step": 3230 }, { "epoch": 2.1507738392411384, "grad_norm": 4.646658897399902, "learning_rate": 5.692717584369449e-05, "loss": 0.0018, "step": 3231 }, { "epoch": 2.1514395074055583, "grad_norm": 32.405982971191406, "learning_rate": 5.6882770870337475e-05, "loss": 0.0823, "step": 3232 }, { "epoch": 2.1521051755699783, "grad_norm": 0.01114699523895979, "learning_rate": 5.683836589698046e-05, "loss": 0.0, "step": 3233 }, { "epoch": 2.152770843734398, "grad_norm": 0.0644788071513176, "learning_rate": 5.6793960923623444e-05, "loss": 0.0001, "step": 3234 }, { "epoch": 2.1534365118988186, "grad_norm": 18.5268611907959, "learning_rate": 5.674955595026643e-05, "loss": 0.0103, "step": 3235 }, { "epoch": 2.1541021800632385, "grad_norm": 54.99586868286133, "learning_rate": 5.6705150976909414e-05, "loss": 0.1014, "step": 3236 }, { "epoch": 2.1547678482276584, "grad_norm": 109.45361328125, "learning_rate": 5.66607460035524e-05, "loss": 0.5407, "step": 3237 }, { "epoch": 2.1554335163920784, "grad_norm": 39.8967170715332, "learning_rate": 5.6616341030195384e-05, "loss": 0.0492, "step": 3238 }, { "epoch": 2.1560991845564987, "grad_norm": 0.11225156486034393, "learning_rate": 5.657193605683837e-05, "loss": 0.0001, "step": 3239 }, { "epoch": 2.1567648527209187, "grad_norm": 2.299262285232544, "learning_rate": 5.6527531083481353e-05, "loss": 0.0024, "step": 3240 }, { "epoch": 2.1574305208853386, "grad_norm": 2.3159537315368652, "learning_rate": 5.648312611012434e-05, "loss": 0.0018, "step": 3241 }, { "epoch": 2.1580961890497585, "grad_norm": 1.5603997707366943, "learning_rate": 5.643872113676732e-05, "loss": 0.0021, "step": 3242 }, { "epoch": 2.158761857214179, "grad_norm": 2.629176616668701, "learning_rate": 5.639431616341031e-05, "loss": 0.0019, "step": 3243 }, { "epoch": 2.159427525378599, "grad_norm": 19.61549949645996, "learning_rate": 5.6349911190053286e-05, "loss": 0.0196, "step": 3244 }, { "epoch": 2.1600931935430188, "grad_norm": 0.006163885351270437, "learning_rate": 5.630550621669627e-05, "loss": 0.0, "step": 3245 }, { "epoch": 2.1607588617074387, "grad_norm": 2.567134141921997, "learning_rate": 5.6261101243339256e-05, "loss": 0.0037, "step": 3246 }, { "epoch": 2.161424529871859, "grad_norm": 0.020258542150259018, "learning_rate": 5.621669626998224e-05, "loss": 0.0, "step": 3247 }, { "epoch": 2.162090198036279, "grad_norm": 69.08560943603516, "learning_rate": 5.6172291296625225e-05, "loss": 0.2744, "step": 3248 }, { "epoch": 2.162755866200699, "grad_norm": 0.06742933392524719, "learning_rate": 5.612788632326821e-05, "loss": 0.0001, "step": 3249 }, { "epoch": 2.163421534365119, "grad_norm": 2.4400768280029297, "learning_rate": 5.6083481349911195e-05, "loss": 0.0008, "step": 3250 }, { "epoch": 2.1640872025295392, "grad_norm": 36.621482849121094, "learning_rate": 5.603907637655418e-05, "loss": 0.0438, "step": 3251 }, { "epoch": 2.164752870693959, "grad_norm": 0.5460940599441528, "learning_rate": 5.5994671403197165e-05, "loss": 0.0007, "step": 3252 }, { "epoch": 2.165418538858379, "grad_norm": 52.36701583862305, "learning_rate": 5.595026642984015e-05, "loss": 0.1568, "step": 3253 }, { "epoch": 2.166084207022799, "grad_norm": 1.3537335395812988, "learning_rate": 5.5905861456483134e-05, "loss": 0.0005, "step": 3254 }, { "epoch": 2.166749875187219, "grad_norm": 5.750248908996582, "learning_rate": 5.586145648312612e-05, "loss": 0.0059, "step": 3255 }, { "epoch": 2.1674155433516393, "grad_norm": 72.69279479980469, "learning_rate": 5.581705150976909e-05, "loss": 0.4633, "step": 3256 }, { "epoch": 2.1680812115160593, "grad_norm": 14.228470802307129, "learning_rate": 5.5772646536412075e-05, "loss": 0.0116, "step": 3257 }, { "epoch": 2.168746879680479, "grad_norm": 74.15756225585938, "learning_rate": 5.572824156305506e-05, "loss": 0.3953, "step": 3258 }, { "epoch": 2.169412547844899, "grad_norm": 48.27680587768555, "learning_rate": 5.5683836589698045e-05, "loss": 0.0322, "step": 3259 }, { "epoch": 2.1700782160093195, "grad_norm": 78.20919036865234, "learning_rate": 5.563943161634103e-05, "loss": 0.0429, "step": 3260 }, { "epoch": 2.1707438841737394, "grad_norm": 0.03551693260669708, "learning_rate": 5.5595026642984015e-05, "loss": 0.0001, "step": 3261 }, { "epoch": 2.1714095523381594, "grad_norm": 123.9832534790039, "learning_rate": 5.5550621669627e-05, "loss": 0.7517, "step": 3262 }, { "epoch": 2.1720752205025793, "grad_norm": 1.4918420314788818, "learning_rate": 5.5506216696269984e-05, "loss": 0.0013, "step": 3263 }, { "epoch": 2.1727408886669997, "grad_norm": 99.40475463867188, "learning_rate": 5.546181172291297e-05, "loss": 0.5923, "step": 3264 }, { "epoch": 2.1734065568314196, "grad_norm": 47.171356201171875, "learning_rate": 5.5417406749555954e-05, "loss": 0.0698, "step": 3265 }, { "epoch": 2.1740722249958395, "grad_norm": 96.94278717041016, "learning_rate": 5.537300177619894e-05, "loss": 0.0729, "step": 3266 }, { "epoch": 2.1747378931602594, "grad_norm": 0.4139353036880493, "learning_rate": 5.5328596802841924e-05, "loss": 0.0006, "step": 3267 }, { "epoch": 2.17540356132468, "grad_norm": 0.040047936141490936, "learning_rate": 5.528419182948491e-05, "loss": 0.0, "step": 3268 }, { "epoch": 2.1760692294890998, "grad_norm": 0.007366158999502659, "learning_rate": 5.5239786856127887e-05, "loss": 0.0, "step": 3269 }, { "epoch": 2.1767348976535197, "grad_norm": 8.173839569091797, "learning_rate": 5.519538188277087e-05, "loss": 0.0053, "step": 3270 }, { "epoch": 2.1774005658179396, "grad_norm": 5.053240776062012, "learning_rate": 5.5150976909413856e-05, "loss": 0.0057, "step": 3271 }, { "epoch": 2.17806623398236, "grad_norm": 126.83847045898438, "learning_rate": 5.510657193605684e-05, "loss": 0.3408, "step": 3272 }, { "epoch": 2.17873190214678, "grad_norm": 1.3679685592651367, "learning_rate": 5.5062166962699826e-05, "loss": 0.0017, "step": 3273 }, { "epoch": 2.1793975703112, "grad_norm": 1.182540774345398, "learning_rate": 5.501776198934281e-05, "loss": 0.0005, "step": 3274 }, { "epoch": 2.18006323847562, "grad_norm": 0.028353044763207436, "learning_rate": 5.4973357015985796e-05, "loss": 0.0, "step": 3275 }, { "epoch": 2.1807289066400397, "grad_norm": 13.948457717895508, "learning_rate": 5.492895204262878e-05, "loss": 0.011, "step": 3276 }, { "epoch": 2.18139457480446, "grad_norm": 3.0304791927337646, "learning_rate": 5.4884547069271765e-05, "loss": 0.002, "step": 3277 }, { "epoch": 2.18206024296888, "grad_norm": 0.01200586836785078, "learning_rate": 5.484014209591475e-05, "loss": 0.0, "step": 3278 }, { "epoch": 2.1827259111333, "grad_norm": 95.4730224609375, "learning_rate": 5.4795737122557735e-05, "loss": 1.6883, "step": 3279 }, { "epoch": 2.1833915792977203, "grad_norm": 50.965999603271484, "learning_rate": 5.475133214920072e-05, "loss": 1.0391, "step": 3280 }, { "epoch": 2.1840572474621402, "grad_norm": 69.93728637695312, "learning_rate": 5.470692717584369e-05, "loss": 0.4645, "step": 3281 }, { "epoch": 2.18472291562656, "grad_norm": 110.47856903076172, "learning_rate": 5.4662522202486676e-05, "loss": 0.2031, "step": 3282 }, { "epoch": 2.18538858379098, "grad_norm": 49.68052291870117, "learning_rate": 5.461811722912966e-05, "loss": 0.4278, "step": 3283 }, { "epoch": 2.1860542519554, "grad_norm": 117.51255798339844, "learning_rate": 5.4573712255772646e-05, "loss": 0.8065, "step": 3284 }, { "epoch": 2.1867199201198204, "grad_norm": 6.003003120422363, "learning_rate": 5.452930728241563e-05, "loss": 0.0046, "step": 3285 }, { "epoch": 2.1873855882842403, "grad_norm": 158.5895233154297, "learning_rate": 5.4484902309058615e-05, "loss": 0.6537, "step": 3286 }, { "epoch": 2.1880512564486603, "grad_norm": 0.06924758851528168, "learning_rate": 5.44404973357016e-05, "loss": 0.0001, "step": 3287 }, { "epoch": 2.18871692461308, "grad_norm": 54.01683044433594, "learning_rate": 5.4396092362344585e-05, "loss": 0.0814, "step": 3288 }, { "epoch": 2.1893825927775006, "grad_norm": 54.037879943847656, "learning_rate": 5.435168738898757e-05, "loss": 0.5568, "step": 3289 }, { "epoch": 2.1900482609419205, "grad_norm": 2.41611385345459, "learning_rate": 5.4307282415630555e-05, "loss": 0.0024, "step": 3290 }, { "epoch": 2.1907139291063404, "grad_norm": 6.018718242645264, "learning_rate": 5.426287744227354e-05, "loss": 0.0073, "step": 3291 }, { "epoch": 2.1913795972707604, "grad_norm": 49.30658721923828, "learning_rate": 5.4218472468916524e-05, "loss": 0.1661, "step": 3292 }, { "epoch": 2.1920452654351807, "grad_norm": 13.349894523620605, "learning_rate": 5.417406749555951e-05, "loss": 0.0101, "step": 3293 }, { "epoch": 2.1927109335996007, "grad_norm": 0.07320424169301987, "learning_rate": 5.412966252220249e-05, "loss": 0.0002, "step": 3294 }, { "epoch": 2.1933766017640206, "grad_norm": 96.66607666015625, "learning_rate": 5.408525754884547e-05, "loss": 0.1338, "step": 3295 }, { "epoch": 2.1940422699284405, "grad_norm": 57.96388626098633, "learning_rate": 5.404085257548846e-05, "loss": 0.4856, "step": 3296 }, { "epoch": 2.194707938092861, "grad_norm": 17.951644897460938, "learning_rate": 5.399644760213144e-05, "loss": 0.0196, "step": 3297 }, { "epoch": 2.195373606257281, "grad_norm": 27.341413497924805, "learning_rate": 5.3952042628774426e-05, "loss": 0.0416, "step": 3298 }, { "epoch": 2.1960392744217008, "grad_norm": 161.66429138183594, "learning_rate": 5.390763765541741e-05, "loss": 0.5509, "step": 3299 }, { "epoch": 2.1967049425861207, "grad_norm": 3.290213108062744, "learning_rate": 5.3863232682060396e-05, "loss": 0.0018, "step": 3300 }, { "epoch": 2.197370610750541, "grad_norm": 69.64868927001953, "learning_rate": 5.381882770870338e-05, "loss": 0.4097, "step": 3301 }, { "epoch": 2.198036278914961, "grad_norm": 166.9256134033203, "learning_rate": 5.3774422735346366e-05, "loss": 0.4621, "step": 3302 }, { "epoch": 2.198701947079381, "grad_norm": 27.942548751831055, "learning_rate": 5.373001776198935e-05, "loss": 0.0867, "step": 3303 }, { "epoch": 2.199367615243801, "grad_norm": 32.51577377319336, "learning_rate": 5.3685612788632336e-05, "loss": 0.0282, "step": 3304 }, { "epoch": 2.200033283408221, "grad_norm": 0.6533284187316895, "learning_rate": 5.364120781527532e-05, "loss": 0.0011, "step": 3305 }, { "epoch": 2.200698951572641, "grad_norm": 40.40217208862305, "learning_rate": 5.359680284191829e-05, "loss": 0.0615, "step": 3306 }, { "epoch": 2.201364619737061, "grad_norm": 47.67695617675781, "learning_rate": 5.3552397868561276e-05, "loss": 0.0799, "step": 3307 }, { "epoch": 2.202030287901481, "grad_norm": 28.532602310180664, "learning_rate": 5.350799289520426e-05, "loss": 0.0417, "step": 3308 }, { "epoch": 2.202695956065901, "grad_norm": 42.773006439208984, "learning_rate": 5.3463587921847246e-05, "loss": 0.0531, "step": 3309 }, { "epoch": 2.2033616242303213, "grad_norm": 40.260189056396484, "learning_rate": 5.341918294849023e-05, "loss": 0.4232, "step": 3310 }, { "epoch": 2.2040272923947413, "grad_norm": 48.73624038696289, "learning_rate": 5.3374777975133216e-05, "loss": 0.2709, "step": 3311 }, { "epoch": 2.204692960559161, "grad_norm": 44.66926193237305, "learning_rate": 5.33303730017762e-05, "loss": 0.4947, "step": 3312 }, { "epoch": 2.205358628723581, "grad_norm": 71.39413452148438, "learning_rate": 5.3285968028419185e-05, "loss": 0.1511, "step": 3313 }, { "epoch": 2.2060242968880015, "grad_norm": 3.761054039001465, "learning_rate": 5.324156305506217e-05, "loss": 0.0061, "step": 3314 }, { "epoch": 2.2066899650524214, "grad_norm": 0.18515798449516296, "learning_rate": 5.3197158081705155e-05, "loss": 0.0003, "step": 3315 }, { "epoch": 2.2073556332168414, "grad_norm": 78.81092071533203, "learning_rate": 5.315275310834814e-05, "loss": 0.1383, "step": 3316 }, { "epoch": 2.2080213013812613, "grad_norm": 5.082155704498291, "learning_rate": 5.3108348134991125e-05, "loss": 0.0079, "step": 3317 }, { "epoch": 2.2086869695456817, "grad_norm": 8.822120666503906, "learning_rate": 5.306394316163411e-05, "loss": 0.0093, "step": 3318 }, { "epoch": 2.2093526377101016, "grad_norm": 0.3320612609386444, "learning_rate": 5.301953818827709e-05, "loss": 0.001, "step": 3319 }, { "epoch": 2.2100183058745215, "grad_norm": 49.13965606689453, "learning_rate": 5.297513321492007e-05, "loss": 0.0234, "step": 3320 }, { "epoch": 2.2106839740389415, "grad_norm": 0.5889614224433899, "learning_rate": 5.293072824156306e-05, "loss": 0.0008, "step": 3321 }, { "epoch": 2.211349642203362, "grad_norm": 14.017581939697266, "learning_rate": 5.288632326820604e-05, "loss": 0.0234, "step": 3322 }, { "epoch": 2.2120153103677818, "grad_norm": 57.50160217285156, "learning_rate": 5.284191829484903e-05, "loss": 0.1411, "step": 3323 }, { "epoch": 2.2126809785322017, "grad_norm": 52.22507858276367, "learning_rate": 5.279751332149201e-05, "loss": 0.1033, "step": 3324 }, { "epoch": 2.2133466466966216, "grad_norm": 0.24788106977939606, "learning_rate": 5.2753108348135e-05, "loss": 0.0005, "step": 3325 }, { "epoch": 2.2140123148610416, "grad_norm": 0.20796774327754974, "learning_rate": 5.270870337477798e-05, "loss": 0.0004, "step": 3326 }, { "epoch": 2.214677983025462, "grad_norm": 0.6395787000656128, "learning_rate": 5.2664298401420966e-05, "loss": 0.001, "step": 3327 }, { "epoch": 2.215343651189882, "grad_norm": 5.848868370056152, "learning_rate": 5.261989342806395e-05, "loss": 0.0077, "step": 3328 }, { "epoch": 2.216009319354302, "grad_norm": 0.11117543280124664, "learning_rate": 5.2575488454706936e-05, "loss": 0.0002, "step": 3329 }, { "epoch": 2.2166749875187217, "grad_norm": 118.66429901123047, "learning_rate": 5.253108348134992e-05, "loss": 0.2067, "step": 3330 }, { "epoch": 2.217340655683142, "grad_norm": 0.16296948492527008, "learning_rate": 5.248667850799289e-05, "loss": 0.0004, "step": 3331 }, { "epoch": 2.218006323847562, "grad_norm": 3.9955084323883057, "learning_rate": 5.244227353463588e-05, "loss": 0.0043, "step": 3332 }, { "epoch": 2.218671992011982, "grad_norm": 0.8316051959991455, "learning_rate": 5.239786856127886e-05, "loss": 0.0008, "step": 3333 }, { "epoch": 2.219337660176402, "grad_norm": 48.82874298095703, "learning_rate": 5.235346358792185e-05, "loss": 0.0866, "step": 3334 }, { "epoch": 2.2200033283408223, "grad_norm": 40.85750961303711, "learning_rate": 5.230905861456483e-05, "loss": 0.0819, "step": 3335 }, { "epoch": 2.220668996505242, "grad_norm": 44.90069580078125, "learning_rate": 5.2264653641207816e-05, "loss": 0.3354, "step": 3336 }, { "epoch": 2.221334664669662, "grad_norm": 28.032623291015625, "learning_rate": 5.22202486678508e-05, "loss": 0.0318, "step": 3337 }, { "epoch": 2.222000332834082, "grad_norm": 0.22410161793231964, "learning_rate": 5.2175843694493786e-05, "loss": 0.0004, "step": 3338 }, { "epoch": 2.2226660009985024, "grad_norm": 0.015235529281198978, "learning_rate": 5.213143872113677e-05, "loss": 0.0, "step": 3339 }, { "epoch": 2.2233316691629224, "grad_norm": 12.736632347106934, "learning_rate": 5.2087033747779756e-05, "loss": 0.0133, "step": 3340 }, { "epoch": 2.2239973373273423, "grad_norm": 90.50444793701172, "learning_rate": 5.204262877442274e-05, "loss": 0.3768, "step": 3341 }, { "epoch": 2.224663005491762, "grad_norm": 113.64257049560547, "learning_rate": 5.1998223801065725e-05, "loss": 0.4622, "step": 3342 }, { "epoch": 2.2253286736561826, "grad_norm": 46.758460998535156, "learning_rate": 5.195381882770871e-05, "loss": 0.1093, "step": 3343 }, { "epoch": 2.2259943418206025, "grad_norm": 2.153589963912964, "learning_rate": 5.190941385435169e-05, "loss": 0.0032, "step": 3344 }, { "epoch": 2.2266600099850224, "grad_norm": 51.988094329833984, "learning_rate": 5.186500888099467e-05, "loss": 0.2247, "step": 3345 }, { "epoch": 2.2273256781494424, "grad_norm": 1.4930009841918945, "learning_rate": 5.182060390763766e-05, "loss": 0.002, "step": 3346 }, { "epoch": 2.2279913463138623, "grad_norm": 63.254600524902344, "learning_rate": 5.177619893428064e-05, "loss": 0.2912, "step": 3347 }, { "epoch": 2.2286570144782827, "grad_norm": 37.303428649902344, "learning_rate": 5.173179396092363e-05, "loss": 0.1196, "step": 3348 }, { "epoch": 2.2293226826427026, "grad_norm": 0.021773332729935646, "learning_rate": 5.168738898756661e-05, "loss": 0.0001, "step": 3349 }, { "epoch": 2.2299883508071225, "grad_norm": 59.53232192993164, "learning_rate": 5.16429840142096e-05, "loss": 0.3143, "step": 3350 }, { "epoch": 2.230654018971543, "grad_norm": 0.571306049823761, "learning_rate": 5.159857904085258e-05, "loss": 0.0012, "step": 3351 }, { "epoch": 2.231319687135963, "grad_norm": 2.1043150424957275, "learning_rate": 5.155417406749557e-05, "loss": 0.002, "step": 3352 }, { "epoch": 2.231985355300383, "grad_norm": 153.66357421875, "learning_rate": 5.150976909413855e-05, "loss": 0.3388, "step": 3353 }, { "epoch": 2.2326510234648027, "grad_norm": 42.46649932861328, "learning_rate": 5.146536412078154e-05, "loss": 0.1723, "step": 3354 }, { "epoch": 2.2333166916292226, "grad_norm": 91.15131378173828, "learning_rate": 5.142095914742452e-05, "loss": 0.3983, "step": 3355 }, { "epoch": 2.233982359793643, "grad_norm": 109.5458755493164, "learning_rate": 5.137655417406749e-05, "loss": 1.2477, "step": 3356 }, { "epoch": 2.234648027958063, "grad_norm": 19.02757453918457, "learning_rate": 5.133214920071048e-05, "loss": 0.0122, "step": 3357 }, { "epoch": 2.235313696122483, "grad_norm": 28.34870719909668, "learning_rate": 5.128774422735346e-05, "loss": 0.0441, "step": 3358 }, { "epoch": 2.235979364286903, "grad_norm": 28.247955322265625, "learning_rate": 5.124333925399645e-05, "loss": 0.0413, "step": 3359 }, { "epoch": 2.236645032451323, "grad_norm": 7.440523147583008, "learning_rate": 5.119893428063943e-05, "loss": 0.0111, "step": 3360 }, { "epoch": 2.237310700615743, "grad_norm": 0.11113868653774261, "learning_rate": 5.115452930728242e-05, "loss": 0.0003, "step": 3361 }, { "epoch": 2.237976368780163, "grad_norm": 64.81865692138672, "learning_rate": 5.11101243339254e-05, "loss": 0.3363, "step": 3362 }, { "epoch": 2.238642036944583, "grad_norm": 0.9227455854415894, "learning_rate": 5.106571936056839e-05, "loss": 0.0009, "step": 3363 }, { "epoch": 2.2393077051090033, "grad_norm": 28.121381759643555, "learning_rate": 5.102131438721137e-05, "loss": 0.1093, "step": 3364 }, { "epoch": 2.2399733732734233, "grad_norm": 0.6866164207458496, "learning_rate": 5.0976909413854356e-05, "loss": 0.0017, "step": 3365 }, { "epoch": 2.240639041437843, "grad_norm": 22.32615089416504, "learning_rate": 5.093250444049734e-05, "loss": 0.0249, "step": 3366 }, { "epoch": 2.241304709602263, "grad_norm": 19.953475952148438, "learning_rate": 5.0888099467140326e-05, "loss": 0.0324, "step": 3367 }, { "epoch": 2.2419703777666835, "grad_norm": 0.8420886993408203, "learning_rate": 5.084369449378331e-05, "loss": 0.0014, "step": 3368 }, { "epoch": 2.2426360459311034, "grad_norm": 43.63079833984375, "learning_rate": 5.079928952042628e-05, "loss": 0.0659, "step": 3369 }, { "epoch": 2.2433017140955234, "grad_norm": 17.528825759887695, "learning_rate": 5.075488454706927e-05, "loss": 0.0304, "step": 3370 }, { "epoch": 2.2439673822599433, "grad_norm": 44.65475082397461, "learning_rate": 5.071047957371225e-05, "loss": 0.1334, "step": 3371 }, { "epoch": 2.2446330504243637, "grad_norm": 0.42451250553131104, "learning_rate": 5.066607460035524e-05, "loss": 0.0005, "step": 3372 }, { "epoch": 2.2452987185887836, "grad_norm": 0.14530281722545624, "learning_rate": 5.062166962699823e-05, "loss": 0.0004, "step": 3373 }, { "epoch": 2.2459643867532035, "grad_norm": 29.051969528198242, "learning_rate": 5.057726465364121e-05, "loss": 0.0813, "step": 3374 }, { "epoch": 2.2466300549176235, "grad_norm": 94.92222595214844, "learning_rate": 5.05328596802842e-05, "loss": 0.2935, "step": 3375 }, { "epoch": 2.2472957230820434, "grad_norm": 5.433004856109619, "learning_rate": 5.048845470692718e-05, "loss": 0.004, "step": 3376 }, { "epoch": 2.2479613912464638, "grad_norm": 0.020511437207460403, "learning_rate": 5.044404973357017e-05, "loss": 0.0001, "step": 3377 }, { "epoch": 2.2486270594108837, "grad_norm": 73.14897155761719, "learning_rate": 5.039964476021315e-05, "loss": 0.1298, "step": 3378 }, { "epoch": 2.2492927275753036, "grad_norm": 1.4833312034606934, "learning_rate": 5.035523978685614e-05, "loss": 0.0016, "step": 3379 }, { "epoch": 2.2499583957397236, "grad_norm": 49.478431701660156, "learning_rate": 5.031083481349912e-05, "loss": 0.1957, "step": 3380 }, { "epoch": 2.250624063904144, "grad_norm": 118.13292694091797, "learning_rate": 5.026642984014209e-05, "loss": 0.1374, "step": 3381 }, { "epoch": 2.251289732068564, "grad_norm": 0.10966784507036209, "learning_rate": 5.022202486678508e-05, "loss": 0.0003, "step": 3382 }, { "epoch": 2.251955400232984, "grad_norm": 6.063979148864746, "learning_rate": 5.017761989342806e-05, "loss": 0.0086, "step": 3383 }, { "epoch": 2.2526210683974037, "grad_norm": 3.221780776977539, "learning_rate": 5.013321492007105e-05, "loss": 0.0053, "step": 3384 }, { "epoch": 2.253286736561824, "grad_norm": 72.34793090820312, "learning_rate": 5.008880994671403e-05, "loss": 0.149, "step": 3385 }, { "epoch": 2.253952404726244, "grad_norm": 52.66592025756836, "learning_rate": 5.004440497335702e-05, "loss": 0.0411, "step": 3386 }, { "epoch": 2.254618072890664, "grad_norm": 86.9973373413086, "learning_rate": 5e-05, "loss": 0.4768, "step": 3387 }, { "epoch": 2.255283741055084, "grad_norm": 69.72212219238281, "learning_rate": 4.995559502664299e-05, "loss": 0.536, "step": 3388 }, { "epoch": 2.2559494092195043, "grad_norm": 68.34197998046875, "learning_rate": 4.991119005328597e-05, "loss": 0.0825, "step": 3389 }, { "epoch": 2.256615077383924, "grad_norm": 0.2626631259918213, "learning_rate": 4.986678507992896e-05, "loss": 0.0003, "step": 3390 }, { "epoch": 2.257280745548344, "grad_norm": 0.6001341342926025, "learning_rate": 4.9822380106571935e-05, "loss": 0.0008, "step": 3391 }, { "epoch": 2.257946413712764, "grad_norm": 49.093360900878906, "learning_rate": 4.977797513321492e-05, "loss": 0.1277, "step": 3392 }, { "epoch": 2.2586120818771844, "grad_norm": 61.54854965209961, "learning_rate": 4.9733570159857905e-05, "loss": 0.0979, "step": 3393 }, { "epoch": 2.2592777500416044, "grad_norm": 2.8693997859954834, "learning_rate": 4.968916518650089e-05, "loss": 0.0028, "step": 3394 }, { "epoch": 2.2599434182060243, "grad_norm": 11.510348320007324, "learning_rate": 4.9644760213143874e-05, "loss": 0.0165, "step": 3395 }, { "epoch": 2.260609086370444, "grad_norm": 15.866061210632324, "learning_rate": 4.960035523978686e-05, "loss": 0.0193, "step": 3396 }, { "epoch": 2.261274754534864, "grad_norm": 20.701824188232422, "learning_rate": 4.955595026642984e-05, "loss": 0.0271, "step": 3397 }, { "epoch": 2.2619404226992845, "grad_norm": 16.4841365814209, "learning_rate": 4.951154529307282e-05, "loss": 0.0337, "step": 3398 }, { "epoch": 2.2626060908637045, "grad_norm": 139.8430633544922, "learning_rate": 4.946714031971581e-05, "loss": 0.6183, "step": 3399 }, { "epoch": 2.2632717590281244, "grad_norm": 9.820241928100586, "learning_rate": 4.942273534635879e-05, "loss": 0.01, "step": 3400 }, { "epoch": 2.2639374271925448, "grad_norm": 14.376213073730469, "learning_rate": 4.9378330373001777e-05, "loss": 0.0123, "step": 3401 }, { "epoch": 2.2646030953569647, "grad_norm": 1.1032092571258545, "learning_rate": 4.933392539964477e-05, "loss": 0.0013, "step": 3402 }, { "epoch": 2.2652687635213846, "grad_norm": 0.3712501525878906, "learning_rate": 4.9289520426287746e-05, "loss": 0.0005, "step": 3403 }, { "epoch": 2.2659344316858046, "grad_norm": 54.187599182128906, "learning_rate": 4.924511545293073e-05, "loss": 0.0783, "step": 3404 }, { "epoch": 2.2666000998502245, "grad_norm": 45.09128189086914, "learning_rate": 4.9200710479573716e-05, "loss": 0.0699, "step": 3405 }, { "epoch": 2.267265768014645, "grad_norm": 56.12318420410156, "learning_rate": 4.91563055062167e-05, "loss": 0.1388, "step": 3406 }, { "epoch": 2.267931436179065, "grad_norm": 0.012287760153412819, "learning_rate": 4.9111900532859686e-05, "loss": 0.0, "step": 3407 }, { "epoch": 2.2685971043434847, "grad_norm": 98.5062255859375, "learning_rate": 4.906749555950267e-05, "loss": 0.2818, "step": 3408 }, { "epoch": 2.2692627725079046, "grad_norm": 123.96216583251953, "learning_rate": 4.902309058614565e-05, "loss": 0.2142, "step": 3409 }, { "epoch": 2.269928440672325, "grad_norm": 81.85225677490234, "learning_rate": 4.897868561278863e-05, "loss": 0.6109, "step": 3410 }, { "epoch": 2.270594108836745, "grad_norm": 66.12645721435547, "learning_rate": 4.893428063943162e-05, "loss": 0.2973, "step": 3411 }, { "epoch": 2.271259777001165, "grad_norm": 0.3254053294658661, "learning_rate": 4.88898756660746e-05, "loss": 0.0008, "step": 3412 }, { "epoch": 2.271925445165585, "grad_norm": 90.0333251953125, "learning_rate": 4.884547069271759e-05, "loss": 0.5206, "step": 3413 }, { "epoch": 2.272591113330005, "grad_norm": 0.18955622613430023, "learning_rate": 4.880106571936057e-05, "loss": 0.0003, "step": 3414 }, { "epoch": 2.273256781494425, "grad_norm": 0.14609962701797485, "learning_rate": 4.875666074600356e-05, "loss": 0.0002, "step": 3415 }, { "epoch": 2.273922449658845, "grad_norm": 0.04052688926458359, "learning_rate": 4.8712255772646536e-05, "loss": 0.0001, "step": 3416 }, { "epoch": 2.274588117823265, "grad_norm": 0.9392072558403015, "learning_rate": 4.866785079928952e-05, "loss": 0.0006, "step": 3417 }, { "epoch": 2.275253785987685, "grad_norm": 60.568477630615234, "learning_rate": 4.8623445825932505e-05, "loss": 0.0683, "step": 3418 }, { "epoch": 2.2759194541521053, "grad_norm": 0.047519538551568985, "learning_rate": 4.857904085257549e-05, "loss": 0.0001, "step": 3419 }, { "epoch": 2.276585122316525, "grad_norm": 108.40901947021484, "learning_rate": 4.8534635879218475e-05, "loss": 0.0852, "step": 3420 }, { "epoch": 2.277250790480945, "grad_norm": 0.3501035273075104, "learning_rate": 4.849023090586146e-05, "loss": 0.0005, "step": 3421 }, { "epoch": 2.2779164586453655, "grad_norm": 40.92536926269531, "learning_rate": 4.844582593250444e-05, "loss": 0.0398, "step": 3422 }, { "epoch": 2.2785821268097854, "grad_norm": 0.15832233428955078, "learning_rate": 4.840142095914742e-05, "loss": 0.0002, "step": 3423 }, { "epoch": 2.2792477949742054, "grad_norm": 11.366667747497559, "learning_rate": 4.835701598579041e-05, "loss": 0.0115, "step": 3424 }, { "epoch": 2.2799134631386253, "grad_norm": 8.716774940490723, "learning_rate": 4.831261101243339e-05, "loss": 0.0111, "step": 3425 }, { "epoch": 2.2805791313030452, "grad_norm": 72.91868591308594, "learning_rate": 4.826820603907638e-05, "loss": 0.2472, "step": 3426 }, { "epoch": 2.2812447994674656, "grad_norm": 59.80415725708008, "learning_rate": 4.822380106571936e-05, "loss": 0.3362, "step": 3427 }, { "epoch": 2.2819104676318855, "grad_norm": 57.69461441040039, "learning_rate": 4.817939609236235e-05, "loss": 0.4992, "step": 3428 }, { "epoch": 2.2825761357963055, "grad_norm": 0.5054060220718384, "learning_rate": 4.813499111900533e-05, "loss": 0.0005, "step": 3429 }, { "epoch": 2.2832418039607254, "grad_norm": 3.2217907905578613, "learning_rate": 4.8090586145648316e-05, "loss": 0.0019, "step": 3430 }, { "epoch": 2.2839074721251458, "grad_norm": 104.14290618896484, "learning_rate": 4.80461811722913e-05, "loss": 0.1759, "step": 3431 }, { "epoch": 2.2845731402895657, "grad_norm": 0.0831749364733696, "learning_rate": 4.8001776198934286e-05, "loss": 0.0001, "step": 3432 }, { "epoch": 2.2852388084539856, "grad_norm": 68.72409057617188, "learning_rate": 4.795737122557727e-05, "loss": 0.1749, "step": 3433 }, { "epoch": 2.2859044766184056, "grad_norm": 60.877098083496094, "learning_rate": 4.791296625222025e-05, "loss": 0.2019, "step": 3434 }, { "epoch": 2.286570144782826, "grad_norm": 3.352598190307617, "learning_rate": 4.7868561278863234e-05, "loss": 0.0021, "step": 3435 }, { "epoch": 2.287235812947246, "grad_norm": 7.078639030456543, "learning_rate": 4.782415630550622e-05, "loss": 0.0108, "step": 3436 }, { "epoch": 2.287901481111666, "grad_norm": 0.015447895973920822, "learning_rate": 4.7779751332149204e-05, "loss": 0.0, "step": 3437 }, { "epoch": 2.2885671492760857, "grad_norm": 119.02298736572266, "learning_rate": 4.773534635879219e-05, "loss": 0.1888, "step": 3438 }, { "epoch": 2.2892328174405057, "grad_norm": 0.01859729178249836, "learning_rate": 4.769094138543517e-05, "loss": 0.0, "step": 3439 }, { "epoch": 2.289898485604926, "grad_norm": 143.07273864746094, "learning_rate": 4.764653641207816e-05, "loss": 0.242, "step": 3440 }, { "epoch": 2.290564153769346, "grad_norm": 52.5589714050293, "learning_rate": 4.7602131438721136e-05, "loss": 0.0894, "step": 3441 }, { "epoch": 2.291229821933766, "grad_norm": 0.1302657127380371, "learning_rate": 4.755772646536412e-05, "loss": 0.0002, "step": 3442 }, { "epoch": 2.2918954900981863, "grad_norm": 4.09016752243042, "learning_rate": 4.7513321492007106e-05, "loss": 0.0058, "step": 3443 }, { "epoch": 2.292561158262606, "grad_norm": 10.12913703918457, "learning_rate": 4.746891651865009e-05, "loss": 0.0117, "step": 3444 }, { "epoch": 2.293226826427026, "grad_norm": 1.4829157590866089, "learning_rate": 4.7424511545293075e-05, "loss": 0.0022, "step": 3445 }, { "epoch": 2.293892494591446, "grad_norm": 6.108059406280518, "learning_rate": 4.738010657193606e-05, "loss": 0.009, "step": 3446 }, { "epoch": 2.294558162755866, "grad_norm": 36.034156799316406, "learning_rate": 4.733570159857904e-05, "loss": 0.5336, "step": 3447 }, { "epoch": 2.2952238309202864, "grad_norm": 5.494277477264404, "learning_rate": 4.729129662522202e-05, "loss": 0.0086, "step": 3448 }, { "epoch": 2.2958894990847063, "grad_norm": 198.86412048339844, "learning_rate": 4.724689165186501e-05, "loss": 0.6415, "step": 3449 }, { "epoch": 2.2965551672491262, "grad_norm": 0.6709858179092407, "learning_rate": 4.720248667850799e-05, "loss": 0.0007, "step": 3450 }, { "epoch": 2.297220835413546, "grad_norm": 0.020148148760199547, "learning_rate": 4.715808170515098e-05, "loss": 0.0, "step": 3451 }, { "epoch": 2.2978865035779665, "grad_norm": 0.3769901394844055, "learning_rate": 4.711367673179396e-05, "loss": 0.0003, "step": 3452 }, { "epoch": 2.2985521717423865, "grad_norm": 0.012269992381334305, "learning_rate": 4.706927175843695e-05, "loss": 0.0, "step": 3453 }, { "epoch": 2.2992178399068064, "grad_norm": 116.6130142211914, "learning_rate": 4.702486678507993e-05, "loss": 0.3848, "step": 3454 }, { "epoch": 2.2998835080712263, "grad_norm": 98.89693450927734, "learning_rate": 4.698046181172292e-05, "loss": 0.3897, "step": 3455 }, { "epoch": 2.3005491762356467, "grad_norm": 0.39358896017074585, "learning_rate": 4.69360568383659e-05, "loss": 0.0003, "step": 3456 }, { "epoch": 2.3012148444000666, "grad_norm": 1.22605562210083, "learning_rate": 4.689165186500889e-05, "loss": 0.0019, "step": 3457 }, { "epoch": 2.3018805125644866, "grad_norm": 24.561145782470703, "learning_rate": 4.684724689165187e-05, "loss": 0.0265, "step": 3458 }, { "epoch": 2.3025461807289065, "grad_norm": 0.2188664972782135, "learning_rate": 4.680284191829485e-05, "loss": 0.0003, "step": 3459 }, { "epoch": 2.303211848893327, "grad_norm": 91.39602661132812, "learning_rate": 4.6758436944937834e-05, "loss": 0.5047, "step": 3460 }, { "epoch": 2.303877517057747, "grad_norm": 23.383649826049805, "learning_rate": 4.671403197158082e-05, "loss": 0.0302, "step": 3461 }, { "epoch": 2.3045431852221667, "grad_norm": 0.7979440689086914, "learning_rate": 4.6669626998223804e-05, "loss": 0.0012, "step": 3462 }, { "epoch": 2.3052088533865867, "grad_norm": 20.89393424987793, "learning_rate": 4.662522202486679e-05, "loss": 0.0307, "step": 3463 }, { "epoch": 2.305874521551007, "grad_norm": 0.026250893250107765, "learning_rate": 4.6580817051509774e-05, "loss": 0.0, "step": 3464 }, { "epoch": 2.306540189715427, "grad_norm": 0.8734409213066101, "learning_rate": 4.653641207815276e-05, "loss": 0.0015, "step": 3465 }, { "epoch": 2.307205857879847, "grad_norm": 4.414907932281494, "learning_rate": 4.649200710479574e-05, "loss": 0.0027, "step": 3466 }, { "epoch": 2.307871526044267, "grad_norm": 61.49554443359375, "learning_rate": 4.644760213143872e-05, "loss": 0.0844, "step": 3467 }, { "epoch": 2.3085371942086867, "grad_norm": 9.615006446838379, "learning_rate": 4.6403197158081706e-05, "loss": 0.0077, "step": 3468 }, { "epoch": 2.309202862373107, "grad_norm": 45.584136962890625, "learning_rate": 4.635879218472469e-05, "loss": 0.0306, "step": 3469 }, { "epoch": 2.309868530537527, "grad_norm": 16.220415115356445, "learning_rate": 4.6314387211367676e-05, "loss": 0.0161, "step": 3470 }, { "epoch": 2.310534198701947, "grad_norm": 108.12321472167969, "learning_rate": 4.626998223801066e-05, "loss": 0.3449, "step": 3471 }, { "epoch": 2.3111998668663674, "grad_norm": 20.277606964111328, "learning_rate": 4.622557726465364e-05, "loss": 0.0118, "step": 3472 }, { "epoch": 2.3118655350307873, "grad_norm": 2.8607828617095947, "learning_rate": 4.6181172291296624e-05, "loss": 0.0019, "step": 3473 }, { "epoch": 2.312531203195207, "grad_norm": 47.28598403930664, "learning_rate": 4.613676731793961e-05, "loss": 0.0709, "step": 3474 }, { "epoch": 2.313196871359627, "grad_norm": 74.57675170898438, "learning_rate": 4.6092362344582593e-05, "loss": 0.078, "step": 3475 }, { "epoch": 2.313862539524047, "grad_norm": 118.73934936523438, "learning_rate": 4.604795737122558e-05, "loss": 0.3028, "step": 3476 }, { "epoch": 2.3145282076884675, "grad_norm": 104.28363800048828, "learning_rate": 4.600355239786856e-05, "loss": 0.203, "step": 3477 }, { "epoch": 2.3151938758528874, "grad_norm": 3.6722915172576904, "learning_rate": 4.595914742451155e-05, "loss": 0.0048, "step": 3478 }, { "epoch": 2.3158595440173073, "grad_norm": 136.7152099609375, "learning_rate": 4.591474245115453e-05, "loss": 0.4608, "step": 3479 }, { "epoch": 2.3165252121817272, "grad_norm": 76.76419067382812, "learning_rate": 4.587033747779752e-05, "loss": 0.0487, "step": 3480 }, { "epoch": 2.3171908803461476, "grad_norm": 2.0233991146087646, "learning_rate": 4.58259325044405e-05, "loss": 0.0016, "step": 3481 }, { "epoch": 2.3178565485105675, "grad_norm": 143.36451721191406, "learning_rate": 4.578152753108349e-05, "loss": 0.4678, "step": 3482 }, { "epoch": 2.3185222166749875, "grad_norm": 0.030263926833868027, "learning_rate": 4.573712255772647e-05, "loss": 0.0, "step": 3483 }, { "epoch": 2.3191878848394074, "grad_norm": 139.0300750732422, "learning_rate": 4.569271758436945e-05, "loss": 0.3792, "step": 3484 }, { "epoch": 2.319853553003828, "grad_norm": 87.4975357055664, "learning_rate": 4.5648312611012435e-05, "loss": 0.3647, "step": 3485 }, { "epoch": 2.3205192211682477, "grad_norm": 25.736635208129883, "learning_rate": 4.560390763765542e-05, "loss": 0.0251, "step": 3486 }, { "epoch": 2.3211848893326676, "grad_norm": 71.91048431396484, "learning_rate": 4.5559502664298405e-05, "loss": 0.2358, "step": 3487 }, { "epoch": 2.3218505574970876, "grad_norm": 4.577054977416992, "learning_rate": 4.551509769094139e-05, "loss": 0.0036, "step": 3488 }, { "epoch": 2.3225162256615075, "grad_norm": 43.11410903930664, "learning_rate": 4.5470692717584374e-05, "loss": 0.0485, "step": 3489 }, { "epoch": 2.323181893825928, "grad_norm": 70.19703674316406, "learning_rate": 4.542628774422736e-05, "loss": 0.1218, "step": 3490 }, { "epoch": 2.323847561990348, "grad_norm": 0.04665432497859001, "learning_rate": 4.538188277087034e-05, "loss": 0.0, "step": 3491 }, { "epoch": 2.3245132301547677, "grad_norm": 26.18996238708496, "learning_rate": 4.533747779751332e-05, "loss": 0.0445, "step": 3492 }, { "epoch": 2.325178898319188, "grad_norm": 53.239349365234375, "learning_rate": 4.529307282415631e-05, "loss": 0.0117, "step": 3493 }, { "epoch": 2.325844566483608, "grad_norm": 5.625436305999756, "learning_rate": 4.524866785079929e-05, "loss": 0.0074, "step": 3494 }, { "epoch": 2.326510234648028, "grad_norm": 0.1888342797756195, "learning_rate": 4.520426287744228e-05, "loss": 0.0003, "step": 3495 }, { "epoch": 2.327175902812448, "grad_norm": 56.80867385864258, "learning_rate": 4.515985790408526e-05, "loss": 0.1805, "step": 3496 }, { "epoch": 2.327841570976868, "grad_norm": 0.6770138144493103, "learning_rate": 4.511545293072824e-05, "loss": 0.001, "step": 3497 }, { "epoch": 2.328507239141288, "grad_norm": 0.042139191180467606, "learning_rate": 4.5071047957371224e-05, "loss": 0.0001, "step": 3498 }, { "epoch": 2.329172907305708, "grad_norm": 0.12172967940568924, "learning_rate": 4.502664298401421e-05, "loss": 0.0002, "step": 3499 }, { "epoch": 2.329838575470128, "grad_norm": 40.61111068725586, "learning_rate": 4.4982238010657194e-05, "loss": 0.0491, "step": 3500 } ], "logging_steps": 1, "max_steps": 4506, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 6.5653891149312e+17, "train_batch_size": 4, "trial_name": null, "trial_params": null }