{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.2853845929128567, "eval_steps": 500, "global_step": 5500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0002337062896205194, "grad_norm": 9.888903617858887, "learning_rate": 1.1686793922867161e-08, "loss": 3.8209, "step": 1 }, { "epoch": 0.0004674125792410388, "grad_norm": 9.704883575439453, "learning_rate": 2.3373587845734322e-08, "loss": 3.8581, "step": 2 }, { "epoch": 0.0007011188688615583, "grad_norm": 9.647233963012695, "learning_rate": 3.5060381768601485e-08, "loss": 4.0384, "step": 3 }, { "epoch": 0.0009348251584820776, "grad_norm": 9.588553428649902, "learning_rate": 4.6747175691468644e-08, "loss": 3.605, "step": 4 }, { "epoch": 0.0011685314481025971, "grad_norm": 9.429463386535645, "learning_rate": 5.8433969614335804e-08, "loss": 3.71, "step": 5 }, { "epoch": 0.0014022377377231165, "grad_norm": 8.851358413696289, "learning_rate": 7.012076353720297e-08, "loss": 3.9192, "step": 6 }, { "epoch": 0.0016359440273436359, "grad_norm": 8.99773120880127, "learning_rate": 8.180755746007012e-08, "loss": 3.7418, "step": 7 }, { "epoch": 0.0018696503169641553, "grad_norm": 8.851916313171387, "learning_rate": 9.349435138293729e-08, "loss": 3.7371, "step": 8 }, { "epoch": 0.0021033566065846747, "grad_norm": 9.048555374145508, "learning_rate": 1.0518114530580444e-07, "loss": 3.8932, "step": 9 }, { "epoch": 0.0023370628962051943, "grad_norm": 8.662620544433594, "learning_rate": 1.1686793922867161e-07, "loss": 4.1042, "step": 10 }, { "epoch": 0.0025707691858257134, "grad_norm": 9.462686538696289, "learning_rate": 1.2855473315153879e-07, "loss": 3.907, "step": 11 }, { "epoch": 0.002804475475446233, "grad_norm": 9.551268577575684, "learning_rate": 1.4024152707440594e-07, "loss": 3.8902, "step": 12 }, { "epoch": 0.003038181765066752, "grad_norm": 8.862306594848633, "learning_rate": 1.519283209972731e-07, "loss": 3.9089, "step": 13 }, { "epoch": 0.0032718880546872718, "grad_norm": 10.091557502746582, "learning_rate": 1.6361511492014024e-07, "loss": 3.978, "step": 14 }, { "epoch": 0.0035055943443077914, "grad_norm": 9.977960586547852, "learning_rate": 1.753019088430074e-07, "loss": 3.6407, "step": 15 }, { "epoch": 0.0037393006339283105, "grad_norm": 9.279823303222656, "learning_rate": 1.8698870276587458e-07, "loss": 3.9837, "step": 16 }, { "epoch": 0.00397300692354883, "grad_norm": 8.616594314575195, "learning_rate": 1.9867549668874173e-07, "loss": 3.5866, "step": 17 }, { "epoch": 0.004206713213169349, "grad_norm": 9.658316612243652, "learning_rate": 2.1036229061160888e-07, "loss": 3.8372, "step": 18 }, { "epoch": 0.004440419502789869, "grad_norm": 9.043797492980957, "learning_rate": 2.2204908453447604e-07, "loss": 3.7553, "step": 19 }, { "epoch": 0.0046741257924103885, "grad_norm": 10.080089569091797, "learning_rate": 2.3373587845734321e-07, "loss": 3.8823, "step": 20 }, { "epoch": 0.004907832082030907, "grad_norm": 9.960041046142578, "learning_rate": 2.4542267238021037e-07, "loss": 3.7926, "step": 21 }, { "epoch": 0.005141538371651427, "grad_norm": 9.626150131225586, "learning_rate": 2.5710946630307757e-07, "loss": 3.7342, "step": 22 }, { "epoch": 0.005375244661271946, "grad_norm": 8.797398567199707, "learning_rate": 2.687962602259447e-07, "loss": 3.9407, "step": 23 }, { "epoch": 0.005608950950892466, "grad_norm": 10.051013946533203, "learning_rate": 2.804830541488119e-07, "loss": 3.6341, "step": 24 }, { "epoch": 0.005842657240512986, "grad_norm": 9.262906074523926, "learning_rate": 2.9216984807167903e-07, "loss": 3.8188, "step": 25 }, { "epoch": 0.006076363530133504, "grad_norm": 9.197418212890625, "learning_rate": 3.038566419945462e-07, "loss": 3.9247, "step": 26 }, { "epoch": 0.006310069819754024, "grad_norm": 9.185877799987793, "learning_rate": 3.1554343591741334e-07, "loss": 3.8934, "step": 27 }, { "epoch": 0.0065437761093745436, "grad_norm": 9.619216918945312, "learning_rate": 3.272302298402805e-07, "loss": 4.0139, "step": 28 }, { "epoch": 0.006777482398995063, "grad_norm": 9.204328536987305, "learning_rate": 3.3891702376314764e-07, "loss": 3.9126, "step": 29 }, { "epoch": 0.007011188688615583, "grad_norm": 9.461112022399902, "learning_rate": 3.506038176860148e-07, "loss": 3.8085, "step": 30 }, { "epoch": 0.0072448949782361015, "grad_norm": 9.394824981689453, "learning_rate": 3.62290611608882e-07, "loss": 3.834, "step": 31 }, { "epoch": 0.007478601267856621, "grad_norm": 9.439630508422852, "learning_rate": 3.7397740553174915e-07, "loss": 4.0441, "step": 32 }, { "epoch": 0.007712307557477141, "grad_norm": 9.321464538574219, "learning_rate": 3.856641994546163e-07, "loss": 3.8253, "step": 33 }, { "epoch": 0.00794601384709766, "grad_norm": 10.065900802612305, "learning_rate": 3.9735099337748346e-07, "loss": 3.7268, "step": 34 }, { "epoch": 0.00817972013671818, "grad_norm": 9.228426933288574, "learning_rate": 4.090377873003506e-07, "loss": 3.755, "step": 35 }, { "epoch": 0.008413426426338699, "grad_norm": 8.065155029296875, "learning_rate": 4.2072458122321776e-07, "loss": 3.6885, "step": 36 }, { "epoch": 0.008647132715959219, "grad_norm": 10.277901649475098, "learning_rate": 4.324113751460849e-07, "loss": 3.7631, "step": 37 }, { "epoch": 0.008880839005579738, "grad_norm": 9.086091041564941, "learning_rate": 4.4409816906895207e-07, "loss": 3.7613, "step": 38 }, { "epoch": 0.009114545295200257, "grad_norm": 8.642022132873535, "learning_rate": 4.557849629918192e-07, "loss": 3.9923, "step": 39 }, { "epoch": 0.009348251584820777, "grad_norm": 9.411796569824219, "learning_rate": 4.6747175691468643e-07, "loss": 3.7716, "step": 40 }, { "epoch": 0.009581957874441296, "grad_norm": 8.217592239379883, "learning_rate": 4.791585508375536e-07, "loss": 3.825, "step": 41 }, { "epoch": 0.009815664164061814, "grad_norm": 9.274012565612793, "learning_rate": 4.908453447604207e-07, "loss": 3.7281, "step": 42 }, { "epoch": 0.010049370453682335, "grad_norm": 8.585111618041992, "learning_rate": 5.025321386832879e-07, "loss": 3.7746, "step": 43 }, { "epoch": 0.010283076743302854, "grad_norm": 8.701138496398926, "learning_rate": 5.142189326061551e-07, "loss": 3.891, "step": 44 }, { "epoch": 0.010516783032923374, "grad_norm": 8.630434036254883, "learning_rate": 5.259057265290223e-07, "loss": 3.9761, "step": 45 }, { "epoch": 0.010750489322543893, "grad_norm": 7.800134181976318, "learning_rate": 5.375925204518895e-07, "loss": 3.8631, "step": 46 }, { "epoch": 0.010984195612164412, "grad_norm": 9.085697174072266, "learning_rate": 5.492793143747566e-07, "loss": 3.7085, "step": 47 }, { "epoch": 0.011217901901784932, "grad_norm": 8.194122314453125, "learning_rate": 5.609661082976238e-07, "loss": 3.5913, "step": 48 }, { "epoch": 0.01145160819140545, "grad_norm": 9.315595626831055, "learning_rate": 5.726529022204909e-07, "loss": 3.8467, "step": 49 }, { "epoch": 0.011685314481025971, "grad_norm": 8.377971649169922, "learning_rate": 5.843396961433581e-07, "loss": 3.7797, "step": 50 }, { "epoch": 0.01191902077064649, "grad_norm": 8.511436462402344, "learning_rate": 5.960264900662252e-07, "loss": 3.8592, "step": 51 }, { "epoch": 0.012152727060267009, "grad_norm": 9.041543006896973, "learning_rate": 6.077132839890924e-07, "loss": 3.8704, "step": 52 }, { "epoch": 0.01238643334988753, "grad_norm": 8.198445320129395, "learning_rate": 6.194000779119595e-07, "loss": 3.9688, "step": 53 }, { "epoch": 0.012620139639508048, "grad_norm": 7.833465576171875, "learning_rate": 6.310868718348267e-07, "loss": 3.6194, "step": 54 }, { "epoch": 0.012853845929128568, "grad_norm": 7.8408203125, "learning_rate": 6.427736657576938e-07, "loss": 3.627, "step": 55 }, { "epoch": 0.013087552218749087, "grad_norm": 7.109745025634766, "learning_rate": 6.54460459680561e-07, "loss": 3.5262, "step": 56 }, { "epoch": 0.013321258508369606, "grad_norm": 6.644226551055908, "learning_rate": 6.661472536034281e-07, "loss": 3.488, "step": 57 }, { "epoch": 0.013554964797990126, "grad_norm": 7.900940418243408, "learning_rate": 6.778340475262953e-07, "loss": 3.8051, "step": 58 }, { "epoch": 0.013788671087610645, "grad_norm": 7.494518756866455, "learning_rate": 6.895208414491624e-07, "loss": 3.7817, "step": 59 }, { "epoch": 0.014022377377231166, "grad_norm": 7.268681526184082, "learning_rate": 7.012076353720296e-07, "loss": 3.6725, "step": 60 }, { "epoch": 0.014256083666851684, "grad_norm": 7.383267879486084, "learning_rate": 7.128944292948967e-07, "loss": 3.6144, "step": 61 }, { "epoch": 0.014489789956472203, "grad_norm": 7.540992259979248, "learning_rate": 7.24581223217764e-07, "loss": 3.9404, "step": 62 }, { "epoch": 0.014723496246092723, "grad_norm": 7.1084065437316895, "learning_rate": 7.362680171406312e-07, "loss": 3.5333, "step": 63 }, { "epoch": 0.014957202535713242, "grad_norm": 7.060537338256836, "learning_rate": 7.479548110634983e-07, "loss": 3.5918, "step": 64 }, { "epoch": 0.015190908825333763, "grad_norm": 7.590440273284912, "learning_rate": 7.596416049863655e-07, "loss": 3.6108, "step": 65 }, { "epoch": 0.015424615114954281, "grad_norm": 7.42264461517334, "learning_rate": 7.713283989092326e-07, "loss": 3.8631, "step": 66 }, { "epoch": 0.0156583214045748, "grad_norm": 7.035755157470703, "learning_rate": 7.830151928320998e-07, "loss": 3.6776, "step": 67 }, { "epoch": 0.01589202769419532, "grad_norm": 6.689797878265381, "learning_rate": 7.947019867549669e-07, "loss": 3.5969, "step": 68 }, { "epoch": 0.01612573398381584, "grad_norm": 7.453489780426025, "learning_rate": 8.063887806778341e-07, "loss": 3.5063, "step": 69 }, { "epoch": 0.01635944027343636, "grad_norm": 6.9404377937316895, "learning_rate": 8.180755746007012e-07, "loss": 3.6111, "step": 70 }, { "epoch": 0.01659314656305688, "grad_norm": 6.821652412414551, "learning_rate": 8.297623685235684e-07, "loss": 3.7068, "step": 71 }, { "epoch": 0.016826852852677397, "grad_norm": 7.397894859313965, "learning_rate": 8.414491624464355e-07, "loss": 3.528, "step": 72 }, { "epoch": 0.017060559142297916, "grad_norm": 7.220261096954346, "learning_rate": 8.531359563693027e-07, "loss": 3.5468, "step": 73 }, { "epoch": 0.017294265431918438, "grad_norm": 6.151828289031982, "learning_rate": 8.648227502921698e-07, "loss": 3.489, "step": 74 }, { "epoch": 0.017527971721538957, "grad_norm": 6.144110202789307, "learning_rate": 8.76509544215037e-07, "loss": 3.6343, "step": 75 }, { "epoch": 0.017761678011159476, "grad_norm": 6.029625415802002, "learning_rate": 8.881963381379041e-07, "loss": 3.7117, "step": 76 }, { "epoch": 0.017995384300779994, "grad_norm": 5.362909317016602, "learning_rate": 8.998831320607713e-07, "loss": 3.3961, "step": 77 }, { "epoch": 0.018229090590400513, "grad_norm": 5.354233264923096, "learning_rate": 9.115699259836384e-07, "loss": 3.3543, "step": 78 }, { "epoch": 0.018462796880021035, "grad_norm": 5.224942207336426, "learning_rate": 9.232567199065057e-07, "loss": 3.5825, "step": 79 }, { "epoch": 0.018696503169641554, "grad_norm": 5.015356540679932, "learning_rate": 9.349435138293729e-07, "loss": 3.2152, "step": 80 }, { "epoch": 0.018930209459262073, "grad_norm": 4.972865104675293, "learning_rate": 9.466303077522401e-07, "loss": 3.2159, "step": 81 }, { "epoch": 0.01916391574888259, "grad_norm": 5.673475742340088, "learning_rate": 9.583171016751072e-07, "loss": 3.5525, "step": 82 }, { "epoch": 0.01939762203850311, "grad_norm": 5.36541748046875, "learning_rate": 9.700038955979744e-07, "loss": 3.6232, "step": 83 }, { "epoch": 0.01963132832812363, "grad_norm": 5.303091049194336, "learning_rate": 9.816906895208415e-07, "loss": 3.5856, "step": 84 }, { "epoch": 0.01986503461774415, "grad_norm": 4.800100326538086, "learning_rate": 9.933774834437087e-07, "loss": 3.3645, "step": 85 }, { "epoch": 0.02009874090736467, "grad_norm": 5.021236896514893, "learning_rate": 1.0050642773665758e-06, "loss": 3.445, "step": 86 }, { "epoch": 0.02033244719698519, "grad_norm": 5.323349952697754, "learning_rate": 1.016751071289443e-06, "loss": 3.4177, "step": 87 }, { "epoch": 0.020566153486605707, "grad_norm": 4.456082344055176, "learning_rate": 1.0284378652123103e-06, "loss": 3.2071, "step": 88 }, { "epoch": 0.020799859776226226, "grad_norm": 4.565123558044434, "learning_rate": 1.0401246591351773e-06, "loss": 3.2878, "step": 89 }, { "epoch": 0.021033566065846748, "grad_norm": 4.588274002075195, "learning_rate": 1.0518114530580446e-06, "loss": 3.2741, "step": 90 }, { "epoch": 0.021267272355467267, "grad_norm": 4.530331611633301, "learning_rate": 1.0634982469809116e-06, "loss": 3.2817, "step": 91 }, { "epoch": 0.021500978645087786, "grad_norm": 4.626265525817871, "learning_rate": 1.075185040903779e-06, "loss": 3.2938, "step": 92 }, { "epoch": 0.021734684934708304, "grad_norm": 4.871453285217285, "learning_rate": 1.086871834826646e-06, "loss": 3.2824, "step": 93 }, { "epoch": 0.021968391224328823, "grad_norm": 4.361344337463379, "learning_rate": 1.0985586287495132e-06, "loss": 3.1612, "step": 94 }, { "epoch": 0.022202097513949345, "grad_norm": 4.149984359741211, "learning_rate": 1.1102454226723803e-06, "loss": 3.2448, "step": 95 }, { "epoch": 0.022435803803569864, "grad_norm": 4.247848987579346, "learning_rate": 1.1219322165952475e-06, "loss": 3.2124, "step": 96 }, { "epoch": 0.022669510093190383, "grad_norm": 3.9699320793151855, "learning_rate": 1.1336190105181146e-06, "loss": 3.1141, "step": 97 }, { "epoch": 0.0229032163828109, "grad_norm": 4.227936744689941, "learning_rate": 1.1453058044409818e-06, "loss": 3.0503, "step": 98 }, { "epoch": 0.02313692267243142, "grad_norm": 4.027324676513672, "learning_rate": 1.1569925983638489e-06, "loss": 3.0107, "step": 99 }, { "epoch": 0.023370628962051943, "grad_norm": 4.1815996170043945, "learning_rate": 1.1686793922867161e-06, "loss": 3.219, "step": 100 }, { "epoch": 0.02360433525167246, "grad_norm": 3.920253038406372, "learning_rate": 1.1803661862095832e-06, "loss": 3.1854, "step": 101 }, { "epoch": 0.02383804154129298, "grad_norm": 3.9520134925842285, "learning_rate": 1.1920529801324504e-06, "loss": 3.3102, "step": 102 }, { "epoch": 0.0240717478309135, "grad_norm": 3.861772060394287, "learning_rate": 1.2037397740553175e-06, "loss": 3.2015, "step": 103 }, { "epoch": 0.024305454120534017, "grad_norm": 3.2346348762512207, "learning_rate": 1.2154265679781847e-06, "loss": 3.039, "step": 104 }, { "epoch": 0.02453916041015454, "grad_norm": 3.3379340171813965, "learning_rate": 1.227113361901052e-06, "loss": 3.1769, "step": 105 }, { "epoch": 0.02477286669977506, "grad_norm": 3.2951889038085938, "learning_rate": 1.238800155823919e-06, "loss": 3.1018, "step": 106 }, { "epoch": 0.025006572989395577, "grad_norm": 3.3087852001190186, "learning_rate": 1.2504869497467863e-06, "loss": 3.0558, "step": 107 }, { "epoch": 0.025240279279016096, "grad_norm": 2.937974691390991, "learning_rate": 1.2621737436696533e-06, "loss": 3.1105, "step": 108 }, { "epoch": 0.025473985568636615, "grad_norm": 3.0109856128692627, "learning_rate": 1.2738605375925206e-06, "loss": 3.0868, "step": 109 }, { "epoch": 0.025707691858257137, "grad_norm": 2.8333067893981934, "learning_rate": 1.2855473315153877e-06, "loss": 2.9174, "step": 110 }, { "epoch": 0.025941398147877655, "grad_norm": 2.873534679412842, "learning_rate": 1.297234125438255e-06, "loss": 3.1226, "step": 111 }, { "epoch": 0.026175104437498174, "grad_norm": 2.8458433151245117, "learning_rate": 1.308920919361122e-06, "loss": 3.1765, "step": 112 }, { "epoch": 0.026408810727118693, "grad_norm": 3.2382309436798096, "learning_rate": 1.3206077132839892e-06, "loss": 2.9674, "step": 113 }, { "epoch": 0.02664251701673921, "grad_norm": 2.9394209384918213, "learning_rate": 1.3322945072068563e-06, "loss": 3.2018, "step": 114 }, { "epoch": 0.026876223306359734, "grad_norm": 2.71695876121521, "learning_rate": 1.3439813011297235e-06, "loss": 3.0896, "step": 115 }, { "epoch": 0.027109929595980253, "grad_norm": 2.7877562046051025, "learning_rate": 1.3556680950525906e-06, "loss": 3.2961, "step": 116 }, { "epoch": 0.02734363588560077, "grad_norm": 2.8273675441741943, "learning_rate": 1.3673548889754578e-06, "loss": 2.9601, "step": 117 }, { "epoch": 0.02757734217522129, "grad_norm": 2.5691614151000977, "learning_rate": 1.3790416828983249e-06, "loss": 2.9337, "step": 118 }, { "epoch": 0.02781104846484181, "grad_norm": 3.033754825592041, "learning_rate": 1.3907284768211921e-06, "loss": 2.9333, "step": 119 }, { "epoch": 0.02804475475446233, "grad_norm": 2.607788324356079, "learning_rate": 1.4024152707440592e-06, "loss": 3.1652, "step": 120 }, { "epoch": 0.02827846104408285, "grad_norm": 2.6202616691589355, "learning_rate": 1.4141020646669264e-06, "loss": 2.997, "step": 121 }, { "epoch": 0.02851216733370337, "grad_norm": 2.760383129119873, "learning_rate": 1.4257888585897935e-06, "loss": 3.0468, "step": 122 }, { "epoch": 0.028745873623323887, "grad_norm": 2.610387086868286, "learning_rate": 1.4374756525126607e-06, "loss": 2.9691, "step": 123 }, { "epoch": 0.028979579912944406, "grad_norm": 2.5390279293060303, "learning_rate": 1.449162446435528e-06, "loss": 3.0517, "step": 124 }, { "epoch": 0.029213286202564928, "grad_norm": 2.549622058868408, "learning_rate": 1.460849240358395e-06, "loss": 2.7878, "step": 125 }, { "epoch": 0.029446992492185447, "grad_norm": 2.8993265628814697, "learning_rate": 1.4725360342812623e-06, "loss": 2.8751, "step": 126 }, { "epoch": 0.029680698781805966, "grad_norm": 2.4403021335601807, "learning_rate": 1.4842228282041294e-06, "loss": 2.9263, "step": 127 }, { "epoch": 0.029914405071426484, "grad_norm": 3.782134532928467, "learning_rate": 1.4959096221269966e-06, "loss": 2.8193, "step": 128 }, { "epoch": 0.030148111361047003, "grad_norm": 2.441246271133423, "learning_rate": 1.5075964160498637e-06, "loss": 3.0235, "step": 129 }, { "epoch": 0.030381817650667525, "grad_norm": 2.567688226699829, "learning_rate": 1.519283209972731e-06, "loss": 2.9598, "step": 130 }, { "epoch": 0.030615523940288044, "grad_norm": 2.663773775100708, "learning_rate": 1.530970003895598e-06, "loss": 2.9288, "step": 131 }, { "epoch": 0.030849230229908563, "grad_norm": 2.5610382556915283, "learning_rate": 1.5426567978184652e-06, "loss": 3.0479, "step": 132 }, { "epoch": 0.03108293651952908, "grad_norm": 2.8623111248016357, "learning_rate": 1.5543435917413323e-06, "loss": 3.1943, "step": 133 }, { "epoch": 0.0313166428091496, "grad_norm": 2.4769723415374756, "learning_rate": 1.5660303856641995e-06, "loss": 2.9342, "step": 134 }, { "epoch": 0.03155034909877012, "grad_norm": 2.560309886932373, "learning_rate": 1.5777171795870666e-06, "loss": 3.0303, "step": 135 }, { "epoch": 0.03178405538839064, "grad_norm": 2.3928561210632324, "learning_rate": 1.5894039735099338e-06, "loss": 2.9619, "step": 136 }, { "epoch": 0.03201776167801116, "grad_norm": 2.3227617740631104, "learning_rate": 1.6010907674328009e-06, "loss": 2.874, "step": 137 }, { "epoch": 0.03225146796763168, "grad_norm": 2.5808913707733154, "learning_rate": 1.6127775613556681e-06, "loss": 3.0417, "step": 138 }, { "epoch": 0.0324851742572522, "grad_norm": 2.609210729598999, "learning_rate": 1.6244643552785352e-06, "loss": 2.8131, "step": 139 }, { "epoch": 0.03271888054687272, "grad_norm": 2.283381223678589, "learning_rate": 1.6361511492014024e-06, "loss": 2.9647, "step": 140 }, { "epoch": 0.032952586836493235, "grad_norm": 2.7190849781036377, "learning_rate": 1.6478379431242697e-06, "loss": 2.6972, "step": 141 }, { "epoch": 0.03318629312611376, "grad_norm": 2.694617509841919, "learning_rate": 1.6595247370471368e-06, "loss": 3.0141, "step": 142 }, { "epoch": 0.03341999941573428, "grad_norm": 2.715696334838867, "learning_rate": 1.671211530970004e-06, "loss": 3.1011, "step": 143 }, { "epoch": 0.033653705705354794, "grad_norm": 2.800549268722534, "learning_rate": 1.682898324892871e-06, "loss": 2.7334, "step": 144 }, { "epoch": 0.03388741199497532, "grad_norm": 2.1698930263519287, "learning_rate": 1.6945851188157383e-06, "loss": 2.8689, "step": 145 }, { "epoch": 0.03412111828459583, "grad_norm": 2.7338943481445312, "learning_rate": 1.7062719127386054e-06, "loss": 2.8376, "step": 146 }, { "epoch": 0.034354824574216354, "grad_norm": 2.2841286659240723, "learning_rate": 1.7179587066614726e-06, "loss": 2.8879, "step": 147 }, { "epoch": 0.034588530863836876, "grad_norm": 2.4011733531951904, "learning_rate": 1.7296455005843397e-06, "loss": 2.8469, "step": 148 }, { "epoch": 0.03482223715345739, "grad_norm": 2.5494420528411865, "learning_rate": 1.741332294507207e-06, "loss": 2.855, "step": 149 }, { "epoch": 0.035055943443077914, "grad_norm": 2.3133769035339355, "learning_rate": 1.753019088430074e-06, "loss": 3.0183, "step": 150 }, { "epoch": 0.03528964973269843, "grad_norm": 2.3024470806121826, "learning_rate": 1.7647058823529412e-06, "loss": 2.8981, "step": 151 }, { "epoch": 0.03552335602231895, "grad_norm": 2.3038525581359863, "learning_rate": 1.7763926762758083e-06, "loss": 2.8518, "step": 152 }, { "epoch": 0.03575706231193947, "grad_norm": 2.6594038009643555, "learning_rate": 1.7880794701986755e-06, "loss": 3.0442, "step": 153 }, { "epoch": 0.03599076860155999, "grad_norm": 2.2779242992401123, "learning_rate": 1.7997662641215426e-06, "loss": 2.835, "step": 154 }, { "epoch": 0.03622447489118051, "grad_norm": 2.679377555847168, "learning_rate": 1.8114530580444098e-06, "loss": 2.7257, "step": 155 }, { "epoch": 0.036458181180801026, "grad_norm": 2.266465425491333, "learning_rate": 1.8231398519672769e-06, "loss": 2.691, "step": 156 }, { "epoch": 0.03669188747042155, "grad_norm": 2.3050122261047363, "learning_rate": 1.8348266458901442e-06, "loss": 2.7306, "step": 157 }, { "epoch": 0.03692559376004207, "grad_norm": 2.1789355278015137, "learning_rate": 1.8465134398130114e-06, "loss": 2.7266, "step": 158 }, { "epoch": 0.037159300049662586, "grad_norm": 2.1759183406829834, "learning_rate": 1.8582002337358785e-06, "loss": 2.8328, "step": 159 }, { "epoch": 0.03739300633928311, "grad_norm": 2.4635884761810303, "learning_rate": 1.8698870276587457e-06, "loss": 3.0965, "step": 160 }, { "epoch": 0.03762671262890362, "grad_norm": 2.5612127780914307, "learning_rate": 1.8815738215816128e-06, "loss": 2.8069, "step": 161 }, { "epoch": 0.037860418918524145, "grad_norm": 2.4434351921081543, "learning_rate": 1.8932606155044802e-06, "loss": 2.742, "step": 162 }, { "epoch": 0.03809412520814466, "grad_norm": 2.304774045944214, "learning_rate": 1.904947409427347e-06, "loss": 2.7615, "step": 163 }, { "epoch": 0.03832783149776518, "grad_norm": 2.5403335094451904, "learning_rate": 1.9166342033502143e-06, "loss": 2.9959, "step": 164 }, { "epoch": 0.038561537787385705, "grad_norm": 2.305016040802002, "learning_rate": 1.9283209972730814e-06, "loss": 2.7931, "step": 165 }, { "epoch": 0.03879524407700622, "grad_norm": 2.391474723815918, "learning_rate": 1.940007791195949e-06, "loss": 2.8835, "step": 166 }, { "epoch": 0.03902895036662674, "grad_norm": 2.6192288398742676, "learning_rate": 1.951694585118816e-06, "loss": 2.8652, "step": 167 }, { "epoch": 0.03926265665624726, "grad_norm": 2.4850120544433594, "learning_rate": 1.963381379041683e-06, "loss": 2.9293, "step": 168 }, { "epoch": 0.03949636294586778, "grad_norm": 2.246975898742676, "learning_rate": 1.97506817296455e-06, "loss": 2.8021, "step": 169 }, { "epoch": 0.0397300692354883, "grad_norm": 2.2591028213500977, "learning_rate": 1.9867549668874175e-06, "loss": 2.9568, "step": 170 }, { "epoch": 0.03996377552510882, "grad_norm": 2.732700824737549, "learning_rate": 1.9984417608102845e-06, "loss": 2.7486, "step": 171 }, { "epoch": 0.04019748181472934, "grad_norm": 2.632443904876709, "learning_rate": 2.0101285547331515e-06, "loss": 2.7865, "step": 172 }, { "epoch": 0.040431188104349855, "grad_norm": 2.26342511177063, "learning_rate": 2.0218153486560186e-06, "loss": 2.854, "step": 173 }, { "epoch": 0.04066489439397038, "grad_norm": 2.492175579071045, "learning_rate": 2.033502142578886e-06, "loss": 2.9526, "step": 174 }, { "epoch": 0.0408986006835909, "grad_norm": 2.4482855796813965, "learning_rate": 2.045188936501753e-06, "loss": 2.6309, "step": 175 }, { "epoch": 0.041132306973211415, "grad_norm": 2.2271041870117188, "learning_rate": 2.0568757304246206e-06, "loss": 2.8897, "step": 176 }, { "epoch": 0.04136601326283194, "grad_norm": 2.640291452407837, "learning_rate": 2.068562524347487e-06, "loss": 3.1546, "step": 177 }, { "epoch": 0.04159971955245245, "grad_norm": 2.908668279647827, "learning_rate": 2.0802493182703547e-06, "loss": 2.9494, "step": 178 }, { "epoch": 0.041833425842072974, "grad_norm": 2.1896111965179443, "learning_rate": 2.0919361121932217e-06, "loss": 2.9076, "step": 179 }, { "epoch": 0.042067132131693497, "grad_norm": 2.357929229736328, "learning_rate": 2.103622906116089e-06, "loss": 3.0307, "step": 180 }, { "epoch": 0.04230083842131401, "grad_norm": 2.4422006607055664, "learning_rate": 2.115309700038956e-06, "loss": 2.672, "step": 181 }, { "epoch": 0.042534544710934534, "grad_norm": 11.580658912658691, "learning_rate": 2.1269964939618233e-06, "loss": 3.1172, "step": 182 }, { "epoch": 0.04276825100055505, "grad_norm": 2.4096500873565674, "learning_rate": 2.1386832878846903e-06, "loss": 2.7246, "step": 183 }, { "epoch": 0.04300195729017557, "grad_norm": 2.3617992401123047, "learning_rate": 2.150370081807558e-06, "loss": 2.9741, "step": 184 }, { "epoch": 0.043235663579796094, "grad_norm": 2.5660741329193115, "learning_rate": 2.1620568757304244e-06, "loss": 2.7234, "step": 185 }, { "epoch": 0.04346936986941661, "grad_norm": 2.78482985496521, "learning_rate": 2.173743669653292e-06, "loss": 2.9165, "step": 186 }, { "epoch": 0.04370307615903713, "grad_norm": 2.7698397636413574, "learning_rate": 2.185430463576159e-06, "loss": 2.726, "step": 187 }, { "epoch": 0.043936782448657646, "grad_norm": 2.6998486518859863, "learning_rate": 2.1971172574990264e-06, "loss": 2.9335, "step": 188 }, { "epoch": 0.04417048873827817, "grad_norm": 2.146197557449341, "learning_rate": 2.208804051421893e-06, "loss": 2.7091, "step": 189 }, { "epoch": 0.04440419502789869, "grad_norm": 2.2743113040924072, "learning_rate": 2.2204908453447605e-06, "loss": 2.7342, "step": 190 }, { "epoch": 0.044637901317519206, "grad_norm": 2.295213460922241, "learning_rate": 2.2321776392676276e-06, "loss": 2.7434, "step": 191 }, { "epoch": 0.04487160760713973, "grad_norm": 2.3094136714935303, "learning_rate": 2.243864433190495e-06, "loss": 2.7382, "step": 192 }, { "epoch": 0.045105313896760243, "grad_norm": 2.961498498916626, "learning_rate": 2.255551227113362e-06, "loss": 2.7759, "step": 193 }, { "epoch": 0.045339020186380766, "grad_norm": 2.132300615310669, "learning_rate": 2.267238021036229e-06, "loss": 2.8427, "step": 194 }, { "epoch": 0.04557272647600129, "grad_norm": 2.269036054611206, "learning_rate": 2.278924814959096e-06, "loss": 2.7303, "step": 195 }, { "epoch": 0.0458064327656218, "grad_norm": 2.1634879112243652, "learning_rate": 2.2906116088819636e-06, "loss": 2.8111, "step": 196 }, { "epoch": 0.046040139055242325, "grad_norm": 2.3109676837921143, "learning_rate": 2.3022984028048307e-06, "loss": 2.994, "step": 197 }, { "epoch": 0.04627384534486284, "grad_norm": 2.4377570152282715, "learning_rate": 2.3139851967276977e-06, "loss": 2.7262, "step": 198 }, { "epoch": 0.04650755163448336, "grad_norm": 2.56258487701416, "learning_rate": 2.3256719906505648e-06, "loss": 2.8131, "step": 199 }, { "epoch": 0.046741257924103885, "grad_norm": 2.344308376312256, "learning_rate": 2.3373587845734322e-06, "loss": 2.7691, "step": 200 }, { "epoch": 0.0469749642137244, "grad_norm": 2.506953239440918, "learning_rate": 2.3490455784962993e-06, "loss": 2.8234, "step": 201 }, { "epoch": 0.04720867050334492, "grad_norm": 2.524859666824341, "learning_rate": 2.3607323724191663e-06, "loss": 2.8683, "step": 202 }, { "epoch": 0.04744237679296544, "grad_norm": 3.339614152908325, "learning_rate": 2.3724191663420334e-06, "loss": 2.7104, "step": 203 }, { "epoch": 0.04767608308258596, "grad_norm": 2.5800867080688477, "learning_rate": 2.384105960264901e-06, "loss": 2.7546, "step": 204 }, { "epoch": 0.04790978937220648, "grad_norm": 2.53129243850708, "learning_rate": 2.395792754187768e-06, "loss": 2.7902, "step": 205 }, { "epoch": 0.048143495661827, "grad_norm": 2.526665449142456, "learning_rate": 2.407479548110635e-06, "loss": 2.9032, "step": 206 }, { "epoch": 0.04837720195144752, "grad_norm": 2.08148193359375, "learning_rate": 2.419166342033502e-06, "loss": 2.8952, "step": 207 }, { "epoch": 0.048610908241068035, "grad_norm": 2.429283618927002, "learning_rate": 2.4308531359563695e-06, "loss": 2.7022, "step": 208 }, { "epoch": 0.04884461453068856, "grad_norm": 2.7161171436309814, "learning_rate": 2.4425399298792365e-06, "loss": 2.7631, "step": 209 }, { "epoch": 0.04907832082030908, "grad_norm": 2.3051183223724365, "learning_rate": 2.454226723802104e-06, "loss": 2.6537, "step": 210 }, { "epoch": 0.049312027109929595, "grad_norm": 2.471071243286133, "learning_rate": 2.4659135177249706e-06, "loss": 2.8327, "step": 211 }, { "epoch": 0.04954573339955012, "grad_norm": 2.2667319774627686, "learning_rate": 2.477600311647838e-06, "loss": 2.659, "step": 212 }, { "epoch": 0.04977943968917063, "grad_norm": 2.1188113689422607, "learning_rate": 2.489287105570705e-06, "loss": 2.756, "step": 213 }, { "epoch": 0.050013145978791154, "grad_norm": 2.534419298171997, "learning_rate": 2.5009738994935726e-06, "loss": 2.8218, "step": 214 }, { "epoch": 0.050246852268411676, "grad_norm": 2.5264668464660645, "learning_rate": 2.5126606934164392e-06, "loss": 2.7725, "step": 215 }, { "epoch": 0.05048055855803219, "grad_norm": 2.4650156497955322, "learning_rate": 2.5243474873393067e-06, "loss": 2.5758, "step": 216 }, { "epoch": 0.050714264847652714, "grad_norm": 2.1425838470458984, "learning_rate": 2.5360342812621737e-06, "loss": 2.6443, "step": 217 }, { "epoch": 0.05094797113727323, "grad_norm": 2.308124542236328, "learning_rate": 2.547721075185041e-06, "loss": 2.7496, "step": 218 }, { "epoch": 0.05118167742689375, "grad_norm": 2.4274392127990723, "learning_rate": 2.559407869107908e-06, "loss": 2.9126, "step": 219 }, { "epoch": 0.051415383716514274, "grad_norm": 2.1558122634887695, "learning_rate": 2.5710946630307753e-06, "loss": 2.8134, "step": 220 }, { "epoch": 0.05164909000613479, "grad_norm": 2.1862165927886963, "learning_rate": 2.5827814569536424e-06, "loss": 2.6968, "step": 221 }, { "epoch": 0.05188279629575531, "grad_norm": 2.376927614212036, "learning_rate": 2.59446825087651e-06, "loss": 2.7068, "step": 222 }, { "epoch": 0.052116502585375826, "grad_norm": 2.2245872020721436, "learning_rate": 2.6061550447993764e-06, "loss": 2.8441, "step": 223 }, { "epoch": 0.05235020887499635, "grad_norm": 2.313992500305176, "learning_rate": 2.617841838722244e-06, "loss": 2.7204, "step": 224 }, { "epoch": 0.05258391516461687, "grad_norm": 2.0907175540924072, "learning_rate": 2.629528632645111e-06, "loss": 2.8199, "step": 225 }, { "epoch": 0.052817621454237386, "grad_norm": 2.1973090171813965, "learning_rate": 2.6412154265679784e-06, "loss": 2.7549, "step": 226 }, { "epoch": 0.05305132774385791, "grad_norm": 2.17120623588562, "learning_rate": 2.6529022204908455e-06, "loss": 2.7116, "step": 227 }, { "epoch": 0.05328503403347842, "grad_norm": 2.2098426818847656, "learning_rate": 2.6645890144137125e-06, "loss": 2.7174, "step": 228 }, { "epoch": 0.053518740323098946, "grad_norm": 2.0925512313842773, "learning_rate": 2.6762758083365796e-06, "loss": 2.6344, "step": 229 }, { "epoch": 0.05375244661271947, "grad_norm": 2.5590429306030273, "learning_rate": 2.687962602259447e-06, "loss": 2.5982, "step": 230 }, { "epoch": 0.05398615290233998, "grad_norm": 2.4244232177734375, "learning_rate": 2.699649396182314e-06, "loss": 2.8816, "step": 231 }, { "epoch": 0.054219859191960505, "grad_norm": 2.6248135566711426, "learning_rate": 2.711336190105181e-06, "loss": 2.6762, "step": 232 }, { "epoch": 0.05445356548158102, "grad_norm": 2.1550397872924805, "learning_rate": 2.723022984028048e-06, "loss": 2.8013, "step": 233 }, { "epoch": 0.05468727177120154, "grad_norm": 2.182803153991699, "learning_rate": 2.7347097779509157e-06, "loss": 2.8273, "step": 234 }, { "epoch": 0.054920978060822065, "grad_norm": 2.155156373977661, "learning_rate": 2.7463965718737827e-06, "loss": 2.7199, "step": 235 }, { "epoch": 0.05515468435044258, "grad_norm": 2.253246307373047, "learning_rate": 2.7580833657966497e-06, "loss": 2.735, "step": 236 }, { "epoch": 0.0553883906400631, "grad_norm": 3.0656044483184814, "learning_rate": 2.769770159719517e-06, "loss": 2.6681, "step": 237 }, { "epoch": 0.05562209692968362, "grad_norm": 2.460017204284668, "learning_rate": 2.7814569536423843e-06, "loss": 2.9508, "step": 238 }, { "epoch": 0.05585580321930414, "grad_norm": 2.1165571212768555, "learning_rate": 2.7931437475652513e-06, "loss": 2.7702, "step": 239 }, { "epoch": 0.05608950950892466, "grad_norm": 2.5097787380218506, "learning_rate": 2.8048305414881184e-06, "loss": 2.6869, "step": 240 }, { "epoch": 0.05632321579854518, "grad_norm": 2.1822850704193115, "learning_rate": 2.8165173354109854e-06, "loss": 2.7605, "step": 241 }, { "epoch": 0.0565569220881657, "grad_norm": 2.154486894607544, "learning_rate": 2.828204129333853e-06, "loss": 3.0025, "step": 242 }, { "epoch": 0.056790628377786215, "grad_norm": 2.0745983123779297, "learning_rate": 2.83989092325672e-06, "loss": 2.666, "step": 243 }, { "epoch": 0.05702433466740674, "grad_norm": 2.3220934867858887, "learning_rate": 2.851577717179587e-06, "loss": 2.6722, "step": 244 }, { "epoch": 0.05725804095702726, "grad_norm": 2.1299848556518555, "learning_rate": 2.863264511102454e-06, "loss": 2.855, "step": 245 }, { "epoch": 0.057491747246647774, "grad_norm": 2.274822235107422, "learning_rate": 2.8749513050253215e-06, "loss": 2.6673, "step": 246 }, { "epoch": 0.0577254535362683, "grad_norm": 2.1445794105529785, "learning_rate": 2.8866380989481885e-06, "loss": 2.7066, "step": 247 }, { "epoch": 0.05795915982588881, "grad_norm": 2.238879680633545, "learning_rate": 2.898324892871056e-06, "loss": 2.7411, "step": 248 }, { "epoch": 0.058192866115509334, "grad_norm": 2.1916892528533936, "learning_rate": 2.9100116867939226e-06, "loss": 2.6492, "step": 249 }, { "epoch": 0.058426572405129856, "grad_norm": 2.1890151500701904, "learning_rate": 2.92169848071679e-06, "loss": 2.4968, "step": 250 }, { "epoch": 0.05866027869475037, "grad_norm": 2.1759305000305176, "learning_rate": 2.933385274639657e-06, "loss": 2.8366, "step": 251 }, { "epoch": 0.058893984984370894, "grad_norm": 2.3982341289520264, "learning_rate": 2.9450720685625246e-06, "loss": 2.6759, "step": 252 }, { "epoch": 0.05912769127399141, "grad_norm": 2.0752174854278564, "learning_rate": 2.9567588624853912e-06, "loss": 2.7684, "step": 253 }, { "epoch": 0.05936139756361193, "grad_norm": 2.345867872238159, "learning_rate": 2.9684456564082587e-06, "loss": 2.9303, "step": 254 }, { "epoch": 0.05959510385323245, "grad_norm": 2.214048385620117, "learning_rate": 2.9801324503311258e-06, "loss": 2.6016, "step": 255 }, { "epoch": 0.05982881014285297, "grad_norm": 2.4768736362457275, "learning_rate": 2.9918192442539932e-06, "loss": 2.7654, "step": 256 }, { "epoch": 0.06006251643247349, "grad_norm": 2.3219337463378906, "learning_rate": 3.00350603817686e-06, "loss": 2.6846, "step": 257 }, { "epoch": 0.060296222722094006, "grad_norm": 2.2863361835479736, "learning_rate": 3.0151928320997273e-06, "loss": 2.7367, "step": 258 }, { "epoch": 0.06052992901171453, "grad_norm": 2.438624620437622, "learning_rate": 3.0268796260225944e-06, "loss": 2.6296, "step": 259 }, { "epoch": 0.06076363530133505, "grad_norm": 2.3783466815948486, "learning_rate": 3.038566419945462e-06, "loss": 2.8014, "step": 260 }, { "epoch": 0.060997341590955566, "grad_norm": 2.5006086826324463, "learning_rate": 3.0502532138683285e-06, "loss": 2.7174, "step": 261 }, { "epoch": 0.06123104788057609, "grad_norm": 2.5780222415924072, "learning_rate": 3.061940007791196e-06, "loss": 2.7802, "step": 262 }, { "epoch": 0.0614647541701966, "grad_norm": 3.3885936737060547, "learning_rate": 3.073626801714063e-06, "loss": 2.9376, "step": 263 }, { "epoch": 0.061698460459817125, "grad_norm": 2.156909704208374, "learning_rate": 3.0853135956369305e-06, "loss": 2.6183, "step": 264 }, { "epoch": 0.06193216674943765, "grad_norm": 2.1270177364349365, "learning_rate": 3.0970003895597975e-06, "loss": 2.7203, "step": 265 }, { "epoch": 0.06216587303905816, "grad_norm": 2.2606630325317383, "learning_rate": 3.1086871834826645e-06, "loss": 2.6323, "step": 266 }, { "epoch": 0.062399579328678685, "grad_norm": 2.4267592430114746, "learning_rate": 3.1203739774055316e-06, "loss": 2.7319, "step": 267 }, { "epoch": 0.0626332856182992, "grad_norm": 3.287079334259033, "learning_rate": 3.132060771328399e-06, "loss": 2.6197, "step": 268 }, { "epoch": 0.06286699190791972, "grad_norm": 3.0488038063049316, "learning_rate": 3.143747565251266e-06, "loss": 2.7351, "step": 269 }, { "epoch": 0.06310069819754024, "grad_norm": 2.647602081298828, "learning_rate": 3.155434359174133e-06, "loss": 2.7414, "step": 270 }, { "epoch": 0.06333440448716077, "grad_norm": 2.5496394634246826, "learning_rate": 3.1671211530970006e-06, "loss": 2.7351, "step": 271 }, { "epoch": 0.06356811077678128, "grad_norm": 2.298487901687622, "learning_rate": 3.1788079470198677e-06, "loss": 2.6244, "step": 272 }, { "epoch": 0.0638018170664018, "grad_norm": 6.362486362457275, "learning_rate": 3.190494740942735e-06, "loss": 2.8872, "step": 273 }, { "epoch": 0.06403552335602232, "grad_norm": 2.0095975399017334, "learning_rate": 3.2021815348656018e-06, "loss": 2.7018, "step": 274 }, { "epoch": 0.06426922964564284, "grad_norm": 3.3410232067108154, "learning_rate": 3.2138683287884692e-06, "loss": 2.9191, "step": 275 }, { "epoch": 0.06450293593526336, "grad_norm": 3.8017640113830566, "learning_rate": 3.2255551227113363e-06, "loss": 2.8337, "step": 276 }, { "epoch": 0.06473664222488387, "grad_norm": 2.8365821838378906, "learning_rate": 3.2372419166342038e-06, "loss": 2.7917, "step": 277 }, { "epoch": 0.0649703485145044, "grad_norm": 2.819336175918579, "learning_rate": 3.2489287105570704e-06, "loss": 2.6843, "step": 278 }, { "epoch": 0.06520405480412492, "grad_norm": 2.7649731636047363, "learning_rate": 3.260615504479938e-06, "loss": 2.7515, "step": 279 }, { "epoch": 0.06543776109374544, "grad_norm": 2.098273992538452, "learning_rate": 3.272302298402805e-06, "loss": 2.6732, "step": 280 }, { "epoch": 0.06567146738336596, "grad_norm": 2.109799385070801, "learning_rate": 3.2839890923256724e-06, "loss": 2.7584, "step": 281 }, { "epoch": 0.06590517367298647, "grad_norm": 2.106476306915283, "learning_rate": 3.2956758862485394e-06, "loss": 2.6381, "step": 282 }, { "epoch": 0.06613887996260699, "grad_norm": 2.29251766204834, "learning_rate": 3.3073626801714065e-06, "loss": 2.6462, "step": 283 }, { "epoch": 0.06637258625222751, "grad_norm": 2.0889599323272705, "learning_rate": 3.3190494740942735e-06, "loss": 2.5239, "step": 284 }, { "epoch": 0.06660629254184804, "grad_norm": 2.9763450622558594, "learning_rate": 3.330736268017141e-06, "loss": 2.5884, "step": 285 }, { "epoch": 0.06683999883146856, "grad_norm": 2.480184555053711, "learning_rate": 3.342423061940008e-06, "loss": 2.5001, "step": 286 }, { "epoch": 0.06707370512108907, "grad_norm": 2.0718984603881836, "learning_rate": 3.354109855862875e-06, "loss": 2.7369, "step": 287 }, { "epoch": 0.06730741141070959, "grad_norm": 2.099640369415283, "learning_rate": 3.365796649785742e-06, "loss": 2.6099, "step": 288 }, { "epoch": 0.06754111770033011, "grad_norm": 2.1269848346710205, "learning_rate": 3.3774834437086096e-06, "loss": 2.7522, "step": 289 }, { "epoch": 0.06777482398995063, "grad_norm": 2.109318733215332, "learning_rate": 3.3891702376314766e-06, "loss": 2.6463, "step": 290 }, { "epoch": 0.06800853027957116, "grad_norm": 2.231839656829834, "learning_rate": 3.4008570315543437e-06, "loss": 2.7358, "step": 291 }, { "epoch": 0.06824223656919166, "grad_norm": 2.264965057373047, "learning_rate": 3.4125438254772107e-06, "loss": 2.8019, "step": 292 }, { "epoch": 0.06847594285881219, "grad_norm": 2.3735077381134033, "learning_rate": 3.424230619400078e-06, "loss": 2.6858, "step": 293 }, { "epoch": 0.06870964914843271, "grad_norm": 2.389153242111206, "learning_rate": 3.4359174133229452e-06, "loss": 2.5495, "step": 294 }, { "epoch": 0.06894335543805323, "grad_norm": 2.186394214630127, "learning_rate": 3.4476042072458123e-06, "loss": 2.6058, "step": 295 }, { "epoch": 0.06917706172767375, "grad_norm": 2.112577199935913, "learning_rate": 3.4592910011686793e-06, "loss": 2.527, "step": 296 }, { "epoch": 0.06941076801729426, "grad_norm": 2.155726432800293, "learning_rate": 3.470977795091547e-06, "loss": 2.5599, "step": 297 }, { "epoch": 0.06964447430691478, "grad_norm": 2.1444180011749268, "learning_rate": 3.482664589014414e-06, "loss": 2.6343, "step": 298 }, { "epoch": 0.0698781805965353, "grad_norm": 2.2002556324005127, "learning_rate": 3.4943513829372813e-06, "loss": 2.7519, "step": 299 }, { "epoch": 0.07011188688615583, "grad_norm": 2.0305771827697754, "learning_rate": 3.506038176860148e-06, "loss": 2.6046, "step": 300 }, { "epoch": 0.07034559317577635, "grad_norm": 2.263918399810791, "learning_rate": 3.5177249707830154e-06, "loss": 2.6929, "step": 301 }, { "epoch": 0.07057929946539686, "grad_norm": 2.1300063133239746, "learning_rate": 3.5294117647058825e-06, "loss": 2.6831, "step": 302 }, { "epoch": 0.07081300575501738, "grad_norm": 7.931042194366455, "learning_rate": 3.54109855862875e-06, "loss": 2.6735, "step": 303 }, { "epoch": 0.0710467120446379, "grad_norm": 2.0475916862487793, "learning_rate": 3.5527853525516166e-06, "loss": 2.7611, "step": 304 }, { "epoch": 0.07128041833425842, "grad_norm": 2.0590054988861084, "learning_rate": 3.564472146474484e-06, "loss": 2.6392, "step": 305 }, { "epoch": 0.07151412462387895, "grad_norm": 2.1635148525238037, "learning_rate": 3.576158940397351e-06, "loss": 2.5558, "step": 306 }, { "epoch": 0.07174783091349946, "grad_norm": 2.037390947341919, "learning_rate": 3.5878457343202185e-06, "loss": 2.681, "step": 307 }, { "epoch": 0.07198153720311998, "grad_norm": 2.087592601776123, "learning_rate": 3.599532528243085e-06, "loss": 2.7762, "step": 308 }, { "epoch": 0.0722152434927405, "grad_norm": 2.100921154022217, "learning_rate": 3.6112193221659526e-06, "loss": 2.6492, "step": 309 }, { "epoch": 0.07244894978236102, "grad_norm": 1.9774651527404785, "learning_rate": 3.6229061160888197e-06, "loss": 2.6669, "step": 310 }, { "epoch": 0.07268265607198154, "grad_norm": 2.0826027393341064, "learning_rate": 3.634592910011687e-06, "loss": 2.6939, "step": 311 }, { "epoch": 0.07291636236160205, "grad_norm": 2.2995963096618652, "learning_rate": 3.6462797039345538e-06, "loss": 2.4904, "step": 312 }, { "epoch": 0.07315006865122257, "grad_norm": 2.458216667175293, "learning_rate": 3.6579664978574213e-06, "loss": 2.4616, "step": 313 }, { "epoch": 0.0733837749408431, "grad_norm": 2.2194573879241943, "learning_rate": 3.6696532917802883e-06, "loss": 2.4891, "step": 314 }, { "epoch": 0.07361748123046362, "grad_norm": 2.125258207321167, "learning_rate": 3.6813400857031558e-06, "loss": 2.5775, "step": 315 }, { "epoch": 0.07385118752008414, "grad_norm": 2.239152193069458, "learning_rate": 3.693026879626023e-06, "loss": 2.5071, "step": 316 }, { "epoch": 0.07408489380970465, "grad_norm": 2.2092859745025635, "learning_rate": 3.70471367354889e-06, "loss": 2.579, "step": 317 }, { "epoch": 0.07431860009932517, "grad_norm": 2.7326979637145996, "learning_rate": 3.716400467471757e-06, "loss": 2.7663, "step": 318 }, { "epoch": 0.0745523063889457, "grad_norm": 2.053737163543701, "learning_rate": 3.7280872613946244e-06, "loss": 2.6279, "step": 319 }, { "epoch": 0.07478601267856622, "grad_norm": 2.3220927715301514, "learning_rate": 3.7397740553174914e-06, "loss": 2.7977, "step": 320 }, { "epoch": 0.07501971896818674, "grad_norm": 2.083249092102051, "learning_rate": 3.751460849240358e-06, "loss": 2.5886, "step": 321 }, { "epoch": 0.07525342525780725, "grad_norm": 2.1724183559417725, "learning_rate": 3.7631476431632255e-06, "loss": 2.5036, "step": 322 }, { "epoch": 0.07548713154742777, "grad_norm": 2.78222918510437, "learning_rate": 3.774834437086093e-06, "loss": 2.7001, "step": 323 }, { "epoch": 0.07572083783704829, "grad_norm": 2.348860740661621, "learning_rate": 3.7865212310089605e-06, "loss": 2.6683, "step": 324 }, { "epoch": 0.07595454412666881, "grad_norm": 2.2033846378326416, "learning_rate": 3.7982080249318267e-06, "loss": 2.7254, "step": 325 }, { "epoch": 0.07618825041628932, "grad_norm": 2.226972818374634, "learning_rate": 3.809894818854694e-06, "loss": 2.6556, "step": 326 }, { "epoch": 0.07642195670590984, "grad_norm": 2.109581708908081, "learning_rate": 3.821581612777562e-06, "loss": 2.5866, "step": 327 }, { "epoch": 0.07665566299553037, "grad_norm": 2.494746685028076, "learning_rate": 3.833268406700429e-06, "loss": 2.4002, "step": 328 }, { "epoch": 0.07688936928515089, "grad_norm": 2.1985373497009277, "learning_rate": 3.844955200623296e-06, "loss": 2.6772, "step": 329 }, { "epoch": 0.07712307557477141, "grad_norm": 2.370687961578369, "learning_rate": 3.856641994546163e-06, "loss": 2.6394, "step": 330 }, { "epoch": 0.07735678186439192, "grad_norm": 2.2602338790893555, "learning_rate": 3.86832878846903e-06, "loss": 2.4715, "step": 331 }, { "epoch": 0.07759048815401244, "grad_norm": 2.079008102416992, "learning_rate": 3.880015582391898e-06, "loss": 2.6823, "step": 332 }, { "epoch": 0.07782419444363296, "grad_norm": 2.3297197818756104, "learning_rate": 3.891702376314764e-06, "loss": 2.6503, "step": 333 }, { "epoch": 0.07805790073325349, "grad_norm": 2.0664665699005127, "learning_rate": 3.903389170237632e-06, "loss": 2.7426, "step": 334 }, { "epoch": 0.07829160702287401, "grad_norm": 2.311551332473755, "learning_rate": 3.915075964160499e-06, "loss": 2.7054, "step": 335 }, { "epoch": 0.07852531331249452, "grad_norm": 2.2487633228302, "learning_rate": 3.926762758083366e-06, "loss": 2.6822, "step": 336 }, { "epoch": 0.07875901960211504, "grad_norm": 2.1039087772369385, "learning_rate": 3.938449552006233e-06, "loss": 2.635, "step": 337 }, { "epoch": 0.07899272589173556, "grad_norm": 2.275362968444824, "learning_rate": 3.9501363459291e-06, "loss": 2.7479, "step": 338 }, { "epoch": 0.07922643218135608, "grad_norm": 2.1846401691436768, "learning_rate": 3.961823139851968e-06, "loss": 2.8003, "step": 339 }, { "epoch": 0.0794601384709766, "grad_norm": 2.0934717655181885, "learning_rate": 3.973509933774835e-06, "loss": 2.7202, "step": 340 }, { "epoch": 0.07969384476059711, "grad_norm": 2.080775737762451, "learning_rate": 3.985196727697701e-06, "loss": 2.4773, "step": 341 }, { "epoch": 0.07992755105021764, "grad_norm": 2.207667827606201, "learning_rate": 3.996883521620569e-06, "loss": 2.6866, "step": 342 }, { "epoch": 0.08016125733983816, "grad_norm": 2.5906031131744385, "learning_rate": 4.008570315543436e-06, "loss": 2.6275, "step": 343 }, { "epoch": 0.08039496362945868, "grad_norm": 2.134122133255005, "learning_rate": 4.020257109466303e-06, "loss": 2.5579, "step": 344 }, { "epoch": 0.0806286699190792, "grad_norm": 2.175856113433838, "learning_rate": 4.03194390338917e-06, "loss": 2.7872, "step": 345 }, { "epoch": 0.08086237620869971, "grad_norm": 2.2518515586853027, "learning_rate": 4.043630697312037e-06, "loss": 2.8397, "step": 346 }, { "epoch": 0.08109608249832023, "grad_norm": 2.241237163543701, "learning_rate": 4.055317491234905e-06, "loss": 2.5008, "step": 347 }, { "epoch": 0.08132978878794075, "grad_norm": 2.69916033744812, "learning_rate": 4.067004285157772e-06, "loss": 2.6428, "step": 348 }, { "epoch": 0.08156349507756128, "grad_norm": 2.0155446529388428, "learning_rate": 4.078691079080638e-06, "loss": 2.6686, "step": 349 }, { "epoch": 0.0817972013671818, "grad_norm": 2.609933376312256, "learning_rate": 4.090377873003506e-06, "loss": 2.7904, "step": 350 }, { "epoch": 0.08203090765680231, "grad_norm": 2.088484048843384, "learning_rate": 4.102064666926373e-06, "loss": 2.7231, "step": 351 }, { "epoch": 0.08226461394642283, "grad_norm": 2.1402227878570557, "learning_rate": 4.113751460849241e-06, "loss": 2.5142, "step": 352 }, { "epoch": 0.08249832023604335, "grad_norm": 2.147503614425659, "learning_rate": 4.125438254772107e-06, "loss": 2.6515, "step": 353 }, { "epoch": 0.08273202652566387, "grad_norm": 2.1589605808258057, "learning_rate": 4.137125048694974e-06, "loss": 2.5359, "step": 354 }, { "epoch": 0.0829657328152844, "grad_norm": 2.2304258346557617, "learning_rate": 4.148811842617842e-06, "loss": 2.4994, "step": 355 }, { "epoch": 0.0831994391049049, "grad_norm": 2.1476387977600098, "learning_rate": 4.160498636540709e-06, "loss": 2.6149, "step": 356 }, { "epoch": 0.08343314539452543, "grad_norm": 2.3848934173583984, "learning_rate": 4.1721854304635756e-06, "loss": 2.7107, "step": 357 }, { "epoch": 0.08366685168414595, "grad_norm": 1.9041041135787964, "learning_rate": 4.1838722243864434e-06, "loss": 2.7927, "step": 358 }, { "epoch": 0.08390055797376647, "grad_norm": 2.2112481594085693, "learning_rate": 4.1955590183093105e-06, "loss": 2.6362, "step": 359 }, { "epoch": 0.08413426426338699, "grad_norm": 2.0748584270477295, "learning_rate": 4.207245812232178e-06, "loss": 2.7497, "step": 360 }, { "epoch": 0.0843679705530075, "grad_norm": 2.283808708190918, "learning_rate": 4.218932606155045e-06, "loss": 2.6386, "step": 361 }, { "epoch": 0.08460167684262802, "grad_norm": 2.1479828357696533, "learning_rate": 4.230619400077912e-06, "loss": 2.6938, "step": 362 }, { "epoch": 0.08483538313224855, "grad_norm": 2.3284554481506348, "learning_rate": 4.2423061940007795e-06, "loss": 2.4971, "step": 363 }, { "epoch": 0.08506908942186907, "grad_norm": 2.2969963550567627, "learning_rate": 4.2539929879236466e-06, "loss": 2.5215, "step": 364 }, { "epoch": 0.08530279571148959, "grad_norm": 2.339613437652588, "learning_rate": 4.265679781846513e-06, "loss": 2.639, "step": 365 }, { "epoch": 0.0855365020011101, "grad_norm": 2.0073204040527344, "learning_rate": 4.277366575769381e-06, "loss": 2.5229, "step": 366 }, { "epoch": 0.08577020829073062, "grad_norm": 2.616271734237671, "learning_rate": 4.289053369692248e-06, "loss": 2.6308, "step": 367 }, { "epoch": 0.08600391458035114, "grad_norm": 2.310546875, "learning_rate": 4.300740163615116e-06, "loss": 2.5104, "step": 368 }, { "epoch": 0.08623762086997167, "grad_norm": 2.228577136993408, "learning_rate": 4.312426957537982e-06, "loss": 2.666, "step": 369 }, { "epoch": 0.08647132715959219, "grad_norm": 2.45011305809021, "learning_rate": 4.324113751460849e-06, "loss": 2.5716, "step": 370 }, { "epoch": 0.0867050334492127, "grad_norm": 2.2168917655944824, "learning_rate": 4.335800545383717e-06, "loss": 2.6507, "step": 371 }, { "epoch": 0.08693873973883322, "grad_norm": 2.310333728790283, "learning_rate": 4.347487339306584e-06, "loss": 2.5759, "step": 372 }, { "epoch": 0.08717244602845374, "grad_norm": 2.475567102432251, "learning_rate": 4.359174133229451e-06, "loss": 2.5332, "step": 373 }, { "epoch": 0.08740615231807426, "grad_norm": 1.9193233251571655, "learning_rate": 4.370860927152318e-06, "loss": 2.4985, "step": 374 }, { "epoch": 0.08763985860769478, "grad_norm": 2.25281023979187, "learning_rate": 4.382547721075185e-06, "loss": 2.6109, "step": 375 }, { "epoch": 0.08787356489731529, "grad_norm": 2.0688045024871826, "learning_rate": 4.394234514998053e-06, "loss": 2.5108, "step": 376 }, { "epoch": 0.08810727118693581, "grad_norm": 2.2170839309692383, "learning_rate": 4.405921308920919e-06, "loss": 2.7088, "step": 377 }, { "epoch": 0.08834097747655634, "grad_norm": 2.0618555545806885, "learning_rate": 4.417608102843786e-06, "loss": 2.5372, "step": 378 }, { "epoch": 0.08857468376617686, "grad_norm": 2.1324987411499023, "learning_rate": 4.429294896766654e-06, "loss": 2.4911, "step": 379 }, { "epoch": 0.08880839005579738, "grad_norm": 2.0915255546569824, "learning_rate": 4.440981690689521e-06, "loss": 2.7046, "step": 380 }, { "epoch": 0.08904209634541789, "grad_norm": 1.9431495666503906, "learning_rate": 4.452668484612388e-06, "loss": 2.7372, "step": 381 }, { "epoch": 0.08927580263503841, "grad_norm": 2.0027260780334473, "learning_rate": 4.464355278535255e-06, "loss": 2.6583, "step": 382 }, { "epoch": 0.08950950892465893, "grad_norm": 2.9133169651031494, "learning_rate": 4.476042072458122e-06, "loss": 2.5206, "step": 383 }, { "epoch": 0.08974321521427946, "grad_norm": 2.9111132621765137, "learning_rate": 4.48772886638099e-06, "loss": 2.6875, "step": 384 }, { "epoch": 0.08997692150389998, "grad_norm": 1.9994900226593018, "learning_rate": 4.499415660303856e-06, "loss": 2.6258, "step": 385 }, { "epoch": 0.09021062779352049, "grad_norm": 2.10575532913208, "learning_rate": 4.511102454226724e-06, "loss": 2.4375, "step": 386 }, { "epoch": 0.09044433408314101, "grad_norm": 2.2720561027526855, "learning_rate": 4.522789248149591e-06, "loss": 2.7064, "step": 387 }, { "epoch": 0.09067804037276153, "grad_norm": 2.4070560932159424, "learning_rate": 4.534476042072458e-06, "loss": 2.6632, "step": 388 }, { "epoch": 0.09091174666238205, "grad_norm": 2.4298877716064453, "learning_rate": 4.546162835995325e-06, "loss": 2.7034, "step": 389 }, { "epoch": 0.09114545295200258, "grad_norm": 2.139240026473999, "learning_rate": 4.557849629918192e-06, "loss": 2.6523, "step": 390 }, { "epoch": 0.09137915924162308, "grad_norm": 2.772392988204956, "learning_rate": 4.569536423841059e-06, "loss": 2.5517, "step": 391 }, { "epoch": 0.0916128655312436, "grad_norm": 3.047044515609741, "learning_rate": 4.581223217763927e-06, "loss": 2.5625, "step": 392 }, { "epoch": 0.09184657182086413, "grad_norm": 1.9968663454055786, "learning_rate": 4.5929100116867935e-06, "loss": 2.6394, "step": 393 }, { "epoch": 0.09208027811048465, "grad_norm": 2.5252954959869385, "learning_rate": 4.604596805609661e-06, "loss": 2.8302, "step": 394 }, { "epoch": 0.09231398440010517, "grad_norm": 2.080099105834961, "learning_rate": 4.616283599532528e-06, "loss": 2.549, "step": 395 }, { "epoch": 0.09254769068972568, "grad_norm": 2.6646409034729004, "learning_rate": 4.6279703934553955e-06, "loss": 2.6163, "step": 396 }, { "epoch": 0.0927813969793462, "grad_norm": 2.3770973682403564, "learning_rate": 4.6396571873782625e-06, "loss": 2.5446, "step": 397 }, { "epoch": 0.09301510326896673, "grad_norm": 2.0549135208129883, "learning_rate": 4.6513439813011296e-06, "loss": 2.5582, "step": 398 }, { "epoch": 0.09324880955858725, "grad_norm": 2.148503065109253, "learning_rate": 4.663030775223997e-06, "loss": 2.6647, "step": 399 }, { "epoch": 0.09348251584820777, "grad_norm": 2.462367534637451, "learning_rate": 4.6747175691468645e-06, "loss": 2.7017, "step": 400 }, { "epoch": 0.09371622213782828, "grad_norm": 2.2049973011016846, "learning_rate": 4.686404363069731e-06, "loss": 2.5927, "step": 401 }, { "epoch": 0.0939499284274488, "grad_norm": 2.2035958766937256, "learning_rate": 4.698091156992599e-06, "loss": 2.6936, "step": 402 }, { "epoch": 0.09418363471706932, "grad_norm": 1.962731122970581, "learning_rate": 4.709777950915466e-06, "loss": 2.6445, "step": 403 }, { "epoch": 0.09441734100668984, "grad_norm": 2.0610318183898926, "learning_rate": 4.721464744838333e-06, "loss": 2.6677, "step": 404 }, { "epoch": 0.09465104729631037, "grad_norm": 2.2989442348480225, "learning_rate": 4.7331515387612e-06, "loss": 2.6858, "step": 405 }, { "epoch": 0.09488475358593088, "grad_norm": 2.27300763130188, "learning_rate": 4.744838332684067e-06, "loss": 2.6579, "step": 406 }, { "epoch": 0.0951184598755514, "grad_norm": 3.0131170749664307, "learning_rate": 4.756525126606935e-06, "loss": 2.6174, "step": 407 }, { "epoch": 0.09535216616517192, "grad_norm": 2.1515448093414307, "learning_rate": 4.768211920529802e-06, "loss": 2.6546, "step": 408 }, { "epoch": 0.09558587245479244, "grad_norm": 2.154306173324585, "learning_rate": 4.779898714452669e-06, "loss": 2.6893, "step": 409 }, { "epoch": 0.09581957874441296, "grad_norm": 2.457332134246826, "learning_rate": 4.791585508375536e-06, "loss": 2.4654, "step": 410 }, { "epoch": 0.09605328503403347, "grad_norm": 2.106736660003662, "learning_rate": 4.803272302298403e-06, "loss": 2.634, "step": 411 }, { "epoch": 0.096286991323654, "grad_norm": 2.372244119644165, "learning_rate": 4.81495909622127e-06, "loss": 2.4129, "step": 412 }, { "epoch": 0.09652069761327452, "grad_norm": 1.9246770143508911, "learning_rate": 4.826645890144138e-06, "loss": 2.5642, "step": 413 }, { "epoch": 0.09675440390289504, "grad_norm": 2.153226137161255, "learning_rate": 4.838332684067004e-06, "loss": 2.597, "step": 414 }, { "epoch": 0.09698811019251556, "grad_norm": 2.2111752033233643, "learning_rate": 4.850019477989872e-06, "loss": 2.5915, "step": 415 }, { "epoch": 0.09722181648213607, "grad_norm": 2.354423761367798, "learning_rate": 4.861706271912739e-06, "loss": 2.6795, "step": 416 }, { "epoch": 0.09745552277175659, "grad_norm": 2.35514760017395, "learning_rate": 4.873393065835606e-06, "loss": 2.4935, "step": 417 }, { "epoch": 0.09768922906137711, "grad_norm": 2.1565427780151367, "learning_rate": 4.885079859758473e-06, "loss": 2.4821, "step": 418 }, { "epoch": 0.09792293535099764, "grad_norm": 2.3325295448303223, "learning_rate": 4.89676665368134e-06, "loss": 2.7066, "step": 419 }, { "epoch": 0.09815664164061816, "grad_norm": 2.249176263809204, "learning_rate": 4.908453447604208e-06, "loss": 2.6601, "step": 420 }, { "epoch": 0.09839034793023867, "grad_norm": 2.1960561275482178, "learning_rate": 4.920140241527075e-06, "loss": 2.6232, "step": 421 }, { "epoch": 0.09862405421985919, "grad_norm": 1.9900469779968262, "learning_rate": 4.931827035449941e-06, "loss": 2.74, "step": 422 }, { "epoch": 0.09885776050947971, "grad_norm": 2.293816328048706, "learning_rate": 4.943513829372809e-06, "loss": 2.636, "step": 423 }, { "epoch": 0.09909146679910023, "grad_norm": 2.3579294681549072, "learning_rate": 4.955200623295676e-06, "loss": 2.5985, "step": 424 }, { "epoch": 0.09932517308872076, "grad_norm": 2.0622315406799316, "learning_rate": 4.966887417218543e-06, "loss": 2.2942, "step": 425 }, { "epoch": 0.09955887937834126, "grad_norm": 2.297405481338501, "learning_rate": 4.97857421114141e-06, "loss": 2.5316, "step": 426 }, { "epoch": 0.09979258566796179, "grad_norm": 2.4507482051849365, "learning_rate": 4.990261005064277e-06, "loss": 2.6357, "step": 427 }, { "epoch": 0.10002629195758231, "grad_norm": 2.940269708633423, "learning_rate": 5.001947798987145e-06, "loss": 2.5161, "step": 428 }, { "epoch": 0.10025999824720283, "grad_norm": 2.7219841480255127, "learning_rate": 5.013634592910012e-06, "loss": 2.6264, "step": 429 }, { "epoch": 0.10049370453682335, "grad_norm": 2.263099193572998, "learning_rate": 5.0253213868328784e-06, "loss": 2.6336, "step": 430 }, { "epoch": 0.10072741082644386, "grad_norm": 2.704266309738159, "learning_rate": 5.037008180755746e-06, "loss": 2.4343, "step": 431 }, { "epoch": 0.10096111711606438, "grad_norm": 2.8439459800720215, "learning_rate": 5.048694974678613e-06, "loss": 2.6516, "step": 432 }, { "epoch": 0.1011948234056849, "grad_norm": 2.2304348945617676, "learning_rate": 5.0603817686014804e-06, "loss": 2.5907, "step": 433 }, { "epoch": 0.10142852969530543, "grad_norm": 2.42801570892334, "learning_rate": 5.0720685625243475e-06, "loss": 2.5262, "step": 434 }, { "epoch": 0.10166223598492595, "grad_norm": 2.213078498840332, "learning_rate": 5.0837553564472145e-06, "loss": 2.4425, "step": 435 }, { "epoch": 0.10189594227454646, "grad_norm": 2.01448392868042, "learning_rate": 5.095442150370082e-06, "loss": 2.6472, "step": 436 }, { "epoch": 0.10212964856416698, "grad_norm": 3.0677201747894287, "learning_rate": 5.1071289442929495e-06, "loss": 2.667, "step": 437 }, { "epoch": 0.1023633548537875, "grad_norm": 2.716438055038452, "learning_rate": 5.118815738215816e-06, "loss": 2.413, "step": 438 }, { "epoch": 0.10259706114340802, "grad_norm": 2.1353087425231934, "learning_rate": 5.1305025321386836e-06, "loss": 2.5994, "step": 439 }, { "epoch": 0.10283076743302855, "grad_norm": 2.0557751655578613, "learning_rate": 5.142189326061551e-06, "loss": 2.4477, "step": 440 }, { "epoch": 0.10306447372264906, "grad_norm": 2.257211923599243, "learning_rate": 5.1538761199844185e-06, "loss": 2.5285, "step": 441 }, { "epoch": 0.10329818001226958, "grad_norm": 2.044069528579712, "learning_rate": 5.165562913907285e-06, "loss": 2.5438, "step": 442 }, { "epoch": 0.1035318863018901, "grad_norm": 2.490328311920166, "learning_rate": 5.177249707830152e-06, "loss": 2.5647, "step": 443 }, { "epoch": 0.10376559259151062, "grad_norm": 2.2903547286987305, "learning_rate": 5.18893650175302e-06, "loss": 2.5563, "step": 444 }, { "epoch": 0.10399929888113114, "grad_norm": 2.167609453201294, "learning_rate": 5.200623295675887e-06, "loss": 2.6075, "step": 445 }, { "epoch": 0.10423300517075165, "grad_norm": 2.180150032043457, "learning_rate": 5.212310089598753e-06, "loss": 2.6279, "step": 446 }, { "epoch": 0.10446671146037217, "grad_norm": 2.233415365219116, "learning_rate": 5.223996883521621e-06, "loss": 2.5074, "step": 447 }, { "epoch": 0.1047004177499927, "grad_norm": 2.2796525955200195, "learning_rate": 5.235683677444488e-06, "loss": 2.5145, "step": 448 }, { "epoch": 0.10493412403961322, "grad_norm": 2.1205320358276367, "learning_rate": 5.247370471367356e-06, "loss": 2.7673, "step": 449 }, { "epoch": 0.10516783032923374, "grad_norm": 2.0501203536987305, "learning_rate": 5.259057265290222e-06, "loss": 2.4543, "step": 450 }, { "epoch": 0.10540153661885425, "grad_norm": 2.217111110687256, "learning_rate": 5.270744059213089e-06, "loss": 2.5826, "step": 451 }, { "epoch": 0.10563524290847477, "grad_norm": 2.7722010612487793, "learning_rate": 5.282430853135957e-06, "loss": 2.6388, "step": 452 }, { "epoch": 0.1058689491980953, "grad_norm": 2.3739333152770996, "learning_rate": 5.294117647058824e-06, "loss": 2.5491, "step": 453 }, { "epoch": 0.10610265548771582, "grad_norm": 2.1784884929656982, "learning_rate": 5.305804440981691e-06, "loss": 2.5071, "step": 454 }, { "epoch": 0.10633636177733634, "grad_norm": 2.1648623943328857, "learning_rate": 5.317491234904558e-06, "loss": 2.6085, "step": 455 }, { "epoch": 0.10657006806695685, "grad_norm": 2.2620444297790527, "learning_rate": 5.329178028827425e-06, "loss": 2.4862, "step": 456 }, { "epoch": 0.10680377435657737, "grad_norm": 2.4915668964385986, "learning_rate": 5.340864822750293e-06, "loss": 2.4714, "step": 457 }, { "epoch": 0.10703748064619789, "grad_norm": 2.0374605655670166, "learning_rate": 5.352551616673159e-06, "loss": 2.6927, "step": 458 }, { "epoch": 0.10727118693581841, "grad_norm": 2.3814444541931152, "learning_rate": 5.364238410596026e-06, "loss": 2.5146, "step": 459 }, { "epoch": 0.10750489322543894, "grad_norm": 2.2664597034454346, "learning_rate": 5.375925204518894e-06, "loss": 2.5842, "step": 460 }, { "epoch": 0.10773859951505944, "grad_norm": 2.04825496673584, "learning_rate": 5.387611998441761e-06, "loss": 2.5215, "step": 461 }, { "epoch": 0.10797230580467997, "grad_norm": 2.3676891326904297, "learning_rate": 5.399298792364628e-06, "loss": 2.6655, "step": 462 }, { "epoch": 0.10820601209430049, "grad_norm": 1.966156005859375, "learning_rate": 5.410985586287495e-06, "loss": 2.6739, "step": 463 }, { "epoch": 0.10843971838392101, "grad_norm": 2.0377137660980225, "learning_rate": 5.422672380210362e-06, "loss": 2.5394, "step": 464 }, { "epoch": 0.10867342467354153, "grad_norm": 2.4358952045440674, "learning_rate": 5.43435917413323e-06, "loss": 2.6609, "step": 465 }, { "epoch": 0.10890713096316204, "grad_norm": 1.92918062210083, "learning_rate": 5.446045968056096e-06, "loss": 2.6294, "step": 466 }, { "epoch": 0.10914083725278256, "grad_norm": 3.0477094650268555, "learning_rate": 5.457732761978963e-06, "loss": 2.4522, "step": 467 }, { "epoch": 0.10937454354240309, "grad_norm": 2.0663001537323, "learning_rate": 5.469419555901831e-06, "loss": 2.5857, "step": 468 }, { "epoch": 0.10960824983202361, "grad_norm": 2.2881484031677246, "learning_rate": 5.481106349824698e-06, "loss": 2.6603, "step": 469 }, { "epoch": 0.10984195612164413, "grad_norm": 2.274214506149292, "learning_rate": 5.492793143747565e-06, "loss": 2.4473, "step": 470 }, { "epoch": 0.11007566241126464, "grad_norm": 2.2721545696258545, "learning_rate": 5.5044799376704325e-06, "loss": 2.6653, "step": 471 }, { "epoch": 0.11030936870088516, "grad_norm": 2.418056011199951, "learning_rate": 5.5161667315932995e-06, "loss": 2.537, "step": 472 }, { "epoch": 0.11054307499050568, "grad_norm": 2.2122302055358887, "learning_rate": 5.527853525516167e-06, "loss": 2.5567, "step": 473 }, { "epoch": 0.1107767812801262, "grad_norm": 2.1849141120910645, "learning_rate": 5.539540319439034e-06, "loss": 2.4506, "step": 474 }, { "epoch": 0.11101048756974673, "grad_norm": 2.5804591178894043, "learning_rate": 5.5512271133619015e-06, "loss": 2.5629, "step": 475 }, { "epoch": 0.11124419385936724, "grad_norm": 2.026536703109741, "learning_rate": 5.5629139072847685e-06, "loss": 2.7054, "step": 476 }, { "epoch": 0.11147790014898776, "grad_norm": 2.222411870956421, "learning_rate": 5.574600701207636e-06, "loss": 2.387, "step": 477 }, { "epoch": 0.11171160643860828, "grad_norm": 2.1787235736846924, "learning_rate": 5.586287495130503e-06, "loss": 2.3689, "step": 478 }, { "epoch": 0.1119453127282288, "grad_norm": 2.216923236846924, "learning_rate": 5.59797428905337e-06, "loss": 2.4833, "step": 479 }, { "epoch": 0.11217901901784932, "grad_norm": 2.112323045730591, "learning_rate": 5.609661082976237e-06, "loss": 2.5742, "step": 480 }, { "epoch": 0.11241272530746983, "grad_norm": 2.565870761871338, "learning_rate": 5.621347876899105e-06, "loss": 2.563, "step": 481 }, { "epoch": 0.11264643159709035, "grad_norm": 2.1655633449554443, "learning_rate": 5.633034670821971e-06, "loss": 2.5873, "step": 482 }, { "epoch": 0.11288013788671088, "grad_norm": 2.3120620250701904, "learning_rate": 5.644721464744839e-06, "loss": 2.6257, "step": 483 }, { "epoch": 0.1131138441763314, "grad_norm": 2.232757329940796, "learning_rate": 5.656408258667706e-06, "loss": 2.5066, "step": 484 }, { "epoch": 0.11334755046595192, "grad_norm": 2.2909576892852783, "learning_rate": 5.668095052590573e-06, "loss": 2.4479, "step": 485 }, { "epoch": 0.11358125675557243, "grad_norm": 2.1446268558502197, "learning_rate": 5.67978184651344e-06, "loss": 2.4146, "step": 486 }, { "epoch": 0.11381496304519295, "grad_norm": 2.074239492416382, "learning_rate": 5.691468640436307e-06, "loss": 2.4573, "step": 487 }, { "epoch": 0.11404866933481347, "grad_norm": 2.349771738052368, "learning_rate": 5.703155434359174e-06, "loss": 2.516, "step": 488 }, { "epoch": 0.114282375624434, "grad_norm": 2.2607550621032715, "learning_rate": 5.714842228282042e-06, "loss": 2.5077, "step": 489 }, { "epoch": 0.11451608191405452, "grad_norm": 2.0770063400268555, "learning_rate": 5.726529022204908e-06, "loss": 2.2667, "step": 490 }, { "epoch": 0.11474978820367503, "grad_norm": 3.3017396926879883, "learning_rate": 5.738215816127776e-06, "loss": 2.6348, "step": 491 }, { "epoch": 0.11498349449329555, "grad_norm": 2.5660693645477295, "learning_rate": 5.749902610050643e-06, "loss": 2.5033, "step": 492 }, { "epoch": 0.11521720078291607, "grad_norm": 2.231001377105713, "learning_rate": 5.76158940397351e-06, "loss": 2.5837, "step": 493 }, { "epoch": 0.1154509070725366, "grad_norm": 2.0775020122528076, "learning_rate": 5.773276197896377e-06, "loss": 2.4341, "step": 494 }, { "epoch": 0.11568461336215712, "grad_norm": 2.6242716312408447, "learning_rate": 5.784962991819244e-06, "loss": 2.568, "step": 495 }, { "epoch": 0.11591831965177762, "grad_norm": 2.4131526947021484, "learning_rate": 5.796649785742112e-06, "loss": 2.3655, "step": 496 }, { "epoch": 0.11615202594139815, "grad_norm": 2.5470712184906006, "learning_rate": 5.808336579664979e-06, "loss": 2.6426, "step": 497 }, { "epoch": 0.11638573223101867, "grad_norm": 2.779430866241455, "learning_rate": 5.820023373587845e-06, "loss": 2.8223, "step": 498 }, { "epoch": 0.11661943852063919, "grad_norm": 2.8101413249969482, "learning_rate": 5.831710167510713e-06, "loss": 2.565, "step": 499 }, { "epoch": 0.11685314481025971, "grad_norm": 2.0815932750701904, "learning_rate": 5.84339696143358e-06, "loss": 2.7993, "step": 500 }, { "epoch": 0.11708685109988022, "grad_norm": 3.0597944259643555, "learning_rate": 5.855083755356447e-06, "loss": 2.6645, "step": 501 }, { "epoch": 0.11732055738950074, "grad_norm": 2.0409562587738037, "learning_rate": 5.866770549279314e-06, "loss": 2.4496, "step": 502 }, { "epoch": 0.11755426367912127, "grad_norm": 2.4022059440612793, "learning_rate": 5.878457343202181e-06, "loss": 2.4684, "step": 503 }, { "epoch": 0.11778796996874179, "grad_norm": 2.2522599697113037, "learning_rate": 5.890144137125049e-06, "loss": 2.53, "step": 504 }, { "epoch": 0.11802167625836231, "grad_norm": 2.2955751419067383, "learning_rate": 5.901830931047916e-06, "loss": 2.5946, "step": 505 }, { "epoch": 0.11825538254798282, "grad_norm": 2.5334856510162354, "learning_rate": 5.9135177249707825e-06, "loss": 2.3867, "step": 506 }, { "epoch": 0.11848908883760334, "grad_norm": 2.1115760803222656, "learning_rate": 5.92520451889365e-06, "loss": 2.5873, "step": 507 }, { "epoch": 0.11872279512722386, "grad_norm": 2.052091121673584, "learning_rate": 5.936891312816517e-06, "loss": 2.3843, "step": 508 }, { "epoch": 0.11895650141684438, "grad_norm": 2.022075891494751, "learning_rate": 5.948578106739385e-06, "loss": 2.5276, "step": 509 }, { "epoch": 0.1191902077064649, "grad_norm": 2.339672327041626, "learning_rate": 5.9602649006622515e-06, "loss": 2.514, "step": 510 }, { "epoch": 0.11942391399608542, "grad_norm": 2.6430041790008545, "learning_rate": 5.9719516945851186e-06, "loss": 2.707, "step": 511 }, { "epoch": 0.11965762028570594, "grad_norm": 2.9963467121124268, "learning_rate": 5.9836384885079865e-06, "loss": 2.3791, "step": 512 }, { "epoch": 0.11989132657532646, "grad_norm": 2.271695852279663, "learning_rate": 5.9953252824308535e-06, "loss": 2.5766, "step": 513 }, { "epoch": 0.12012503286494698, "grad_norm": 2.209756374359131, "learning_rate": 6.00701207635372e-06, "loss": 2.5944, "step": 514 }, { "epoch": 0.1203587391545675, "grad_norm": 2.378046989440918, "learning_rate": 6.018698870276588e-06, "loss": 2.5161, "step": 515 }, { "epoch": 0.12059244544418801, "grad_norm": 2.1483070850372314, "learning_rate": 6.030385664199455e-06, "loss": 2.5194, "step": 516 }, { "epoch": 0.12082615173380853, "grad_norm": 1.9927738904953003, "learning_rate": 6.0420724581223225e-06, "loss": 2.5088, "step": 517 }, { "epoch": 0.12105985802342906, "grad_norm": 2.4795806407928467, "learning_rate": 6.053759252045189e-06, "loss": 2.4359, "step": 518 }, { "epoch": 0.12129356431304958, "grad_norm": 4.308606147766113, "learning_rate": 6.065446045968056e-06, "loss": 2.431, "step": 519 }, { "epoch": 0.1215272706026701, "grad_norm": 3.570502281188965, "learning_rate": 6.077132839890924e-06, "loss": 2.5219, "step": 520 }, { "epoch": 0.12176097689229061, "grad_norm": 2.404146194458008, "learning_rate": 6.088819633813791e-06, "loss": 2.635, "step": 521 }, { "epoch": 0.12199468318191113, "grad_norm": 2.066732168197632, "learning_rate": 6.100506427736657e-06, "loss": 2.4544, "step": 522 }, { "epoch": 0.12222838947153165, "grad_norm": 3.6416988372802734, "learning_rate": 6.112193221659525e-06, "loss": 2.6701, "step": 523 }, { "epoch": 0.12246209576115218, "grad_norm": 3.531463861465454, "learning_rate": 6.123880015582392e-06, "loss": 2.6253, "step": 524 }, { "epoch": 0.1226958020507727, "grad_norm": 2.105196475982666, "learning_rate": 6.13556680950526e-06, "loss": 2.4686, "step": 525 }, { "epoch": 0.1229295083403932, "grad_norm": 2.7551231384277344, "learning_rate": 6.147253603428126e-06, "loss": 2.6266, "step": 526 }, { "epoch": 0.12316321463001373, "grad_norm": 2.288761615753174, "learning_rate": 6.158940397350993e-06, "loss": 2.53, "step": 527 }, { "epoch": 0.12339692091963425, "grad_norm": 3.797744035720825, "learning_rate": 6.170627191273861e-06, "loss": 2.7028, "step": 528 }, { "epoch": 0.12363062720925477, "grad_norm": 2.3684256076812744, "learning_rate": 6.182313985196728e-06, "loss": 2.5644, "step": 529 }, { "epoch": 0.1238643334988753, "grad_norm": 2.164806842803955, "learning_rate": 6.194000779119595e-06, "loss": 2.5843, "step": 530 }, { "epoch": 0.1240980397884958, "grad_norm": 2.0515739917755127, "learning_rate": 6.205687573042462e-06, "loss": 2.4152, "step": 531 }, { "epoch": 0.12433174607811633, "grad_norm": 2.831336736679077, "learning_rate": 6.217374366965329e-06, "loss": 2.3857, "step": 532 }, { "epoch": 0.12456545236773685, "grad_norm": 2.255380153656006, "learning_rate": 6.229061160888197e-06, "loss": 2.6106, "step": 533 }, { "epoch": 0.12479915865735737, "grad_norm": 2.2005505561828613, "learning_rate": 6.240747954811063e-06, "loss": 2.5453, "step": 534 }, { "epoch": 0.12503286494697788, "grad_norm": 2.7216901779174805, "learning_rate": 6.25243474873393e-06, "loss": 2.3956, "step": 535 }, { "epoch": 0.1252665712365984, "grad_norm": 2.1246674060821533, "learning_rate": 6.264121542656798e-06, "loss": 2.5171, "step": 536 }, { "epoch": 0.12550027752621892, "grad_norm": 2.068380117416382, "learning_rate": 6.275808336579665e-06, "loss": 2.3431, "step": 537 }, { "epoch": 0.12573398381583945, "grad_norm": 2.2472198009490967, "learning_rate": 6.287495130502532e-06, "loss": 2.4875, "step": 538 }, { "epoch": 0.12596769010545997, "grad_norm": 2.6920275688171387, "learning_rate": 6.299181924425399e-06, "loss": 2.4193, "step": 539 }, { "epoch": 0.1262013963950805, "grad_norm": 2.6327545642852783, "learning_rate": 6.310868718348266e-06, "loss": 2.5289, "step": 540 }, { "epoch": 0.126435102684701, "grad_norm": 2.126450300216675, "learning_rate": 6.322555512271134e-06, "loss": 2.6484, "step": 541 }, { "epoch": 0.12666880897432153, "grad_norm": 3.122652292251587, "learning_rate": 6.334242306194001e-06, "loss": 2.4683, "step": 542 }, { "epoch": 0.12690251526394203, "grad_norm": 3.317017078399658, "learning_rate": 6.345929100116868e-06, "loss": 2.5308, "step": 543 }, { "epoch": 0.12713622155356255, "grad_norm": 2.4377167224884033, "learning_rate": 6.357615894039735e-06, "loss": 2.5371, "step": 544 }, { "epoch": 0.12736992784318307, "grad_norm": 2.109522581100464, "learning_rate": 6.369302687962602e-06, "loss": 2.5402, "step": 545 }, { "epoch": 0.1276036341328036, "grad_norm": 2.5240960121154785, "learning_rate": 6.38098948188547e-06, "loss": 2.404, "step": 546 }, { "epoch": 0.12783734042242412, "grad_norm": 2.095463991165161, "learning_rate": 6.3926762758083365e-06, "loss": 2.4161, "step": 547 }, { "epoch": 0.12807104671204464, "grad_norm": 2.9461112022399902, "learning_rate": 6.4043630697312035e-06, "loss": 2.463, "step": 548 }, { "epoch": 0.12830475300166516, "grad_norm": 3.0624125003814697, "learning_rate": 6.4160498636540714e-06, "loss": 2.3634, "step": 549 }, { "epoch": 0.12853845929128568, "grad_norm": 3.53266978263855, "learning_rate": 6.4277366575769385e-06, "loss": 2.5017, "step": 550 }, { "epoch": 0.1287721655809062, "grad_norm": 3.251788377761841, "learning_rate": 6.4394234514998055e-06, "loss": 2.3518, "step": 551 }, { "epoch": 0.12900587187052673, "grad_norm": 2.37613844871521, "learning_rate": 6.4511102454226726e-06, "loss": 2.5037, "step": 552 }, { "epoch": 0.12923957816014722, "grad_norm": 2.925072431564331, "learning_rate": 6.46279703934554e-06, "loss": 2.3537, "step": 553 }, { "epoch": 0.12947328444976774, "grad_norm": 3.726407051086426, "learning_rate": 6.4744838332684075e-06, "loss": 2.6189, "step": 554 }, { "epoch": 0.12970699073938827, "grad_norm": 2.75364089012146, "learning_rate": 6.486170627191274e-06, "loss": 2.5468, "step": 555 }, { "epoch": 0.1299406970290088, "grad_norm": 2.3366622924804688, "learning_rate": 6.497857421114141e-06, "loss": 2.5776, "step": 556 }, { "epoch": 0.1301744033186293, "grad_norm": 2.4038147926330566, "learning_rate": 6.509544215037009e-06, "loss": 2.5428, "step": 557 }, { "epoch": 0.13040810960824983, "grad_norm": 2.1106557846069336, "learning_rate": 6.521231008959876e-06, "loss": 2.5082, "step": 558 }, { "epoch": 0.13064181589787036, "grad_norm": 2.599641799926758, "learning_rate": 6.532917802882743e-06, "loss": 2.449, "step": 559 }, { "epoch": 0.13087552218749088, "grad_norm": 2.080599784851074, "learning_rate": 6.54460459680561e-06, "loss": 2.4692, "step": 560 }, { "epoch": 0.1311092284771114, "grad_norm": 2.8009626865386963, "learning_rate": 6.556291390728477e-06, "loss": 2.4158, "step": 561 }, { "epoch": 0.13134293476673192, "grad_norm": 2.7373383045196533, "learning_rate": 6.567978184651345e-06, "loss": 2.566, "step": 562 }, { "epoch": 0.13157664105635242, "grad_norm": 2.1681101322174072, "learning_rate": 6.579664978574211e-06, "loss": 2.4557, "step": 563 }, { "epoch": 0.13181034734597294, "grad_norm": 2.4806811809539795, "learning_rate": 6.591351772497079e-06, "loss": 2.5767, "step": 564 }, { "epoch": 0.13204405363559346, "grad_norm": 2.505441427230835, "learning_rate": 6.603038566419946e-06, "loss": 2.4933, "step": 565 }, { "epoch": 0.13227775992521398, "grad_norm": 2.2596936225891113, "learning_rate": 6.614725360342813e-06, "loss": 2.5386, "step": 566 }, { "epoch": 0.1325114662148345, "grad_norm": 2.494717597961426, "learning_rate": 6.62641215426568e-06, "loss": 2.4636, "step": 567 }, { "epoch": 0.13274517250445503, "grad_norm": 2.2982261180877686, "learning_rate": 6.638098948188547e-06, "loss": 2.4605, "step": 568 }, { "epoch": 0.13297887879407555, "grad_norm": 2.2732725143432617, "learning_rate": 6.649785742111414e-06, "loss": 2.6024, "step": 569 }, { "epoch": 0.13321258508369607, "grad_norm": 2.0725531578063965, "learning_rate": 6.661472536034282e-06, "loss": 2.6675, "step": 570 }, { "epoch": 0.1334462913733166, "grad_norm": 2.491908311843872, "learning_rate": 6.673159329957148e-06, "loss": 2.5461, "step": 571 }, { "epoch": 0.13367999766293712, "grad_norm": 2.6406736373901367, "learning_rate": 6.684846123880016e-06, "loss": 2.6565, "step": 572 }, { "epoch": 0.1339137039525576, "grad_norm": 2.2929129600524902, "learning_rate": 6.696532917802883e-06, "loss": 2.6085, "step": 573 }, { "epoch": 0.13414741024217813, "grad_norm": 2.0278561115264893, "learning_rate": 6.70821971172575e-06, "loss": 2.5818, "step": 574 }, { "epoch": 0.13438111653179866, "grad_norm": 2.579801082611084, "learning_rate": 6.719906505648617e-06, "loss": 2.336, "step": 575 }, { "epoch": 0.13461482282141918, "grad_norm": 2.216449737548828, "learning_rate": 6.731593299571484e-06, "loss": 2.5361, "step": 576 }, { "epoch": 0.1348485291110397, "grad_norm": 2.1561474800109863, "learning_rate": 6.743280093494352e-06, "loss": 2.4788, "step": 577 }, { "epoch": 0.13508223540066022, "grad_norm": 3.2043609619140625, "learning_rate": 6.754966887417219e-06, "loss": 2.4064, "step": 578 }, { "epoch": 0.13531594169028074, "grad_norm": 2.3214914798736572, "learning_rate": 6.766653681340085e-06, "loss": 2.6497, "step": 579 }, { "epoch": 0.13554964797990127, "grad_norm": 2.2135202884674072, "learning_rate": 6.778340475262953e-06, "loss": 2.471, "step": 580 }, { "epoch": 0.1357833542695218, "grad_norm": 4.016070365905762, "learning_rate": 6.79002726918582e-06, "loss": 2.3472, "step": 581 }, { "epoch": 0.1360170605591423, "grad_norm": 2.2975687980651855, "learning_rate": 6.801714063108687e-06, "loss": 2.3975, "step": 582 }, { "epoch": 0.1362507668487628, "grad_norm": 2.109772205352783, "learning_rate": 6.813400857031554e-06, "loss": 2.4402, "step": 583 }, { "epoch": 0.13648447313838333, "grad_norm": 2.4656577110290527, "learning_rate": 6.8250876509544215e-06, "loss": 2.5441, "step": 584 }, { "epoch": 0.13671817942800385, "grad_norm": 2.3982999324798584, "learning_rate": 6.836774444877289e-06, "loss": 2.4335, "step": 585 }, { "epoch": 0.13695188571762437, "grad_norm": 2.2763545513153076, "learning_rate": 6.848461238800156e-06, "loss": 2.4326, "step": 586 }, { "epoch": 0.1371855920072449, "grad_norm": 3.818749189376831, "learning_rate": 6.860148032723023e-06, "loss": 2.5332, "step": 587 }, { "epoch": 0.13741929829686542, "grad_norm": 3.3831863403320312, "learning_rate": 6.8718348266458905e-06, "loss": 2.3961, "step": 588 }, { "epoch": 0.13765300458648594, "grad_norm": 2.72816801071167, "learning_rate": 6.8835216205687575e-06, "loss": 2.3487, "step": 589 }, { "epoch": 0.13788671087610646, "grad_norm": 2.2882328033447266, "learning_rate": 6.895208414491625e-06, "loss": 2.54, "step": 590 }, { "epoch": 0.13812041716572698, "grad_norm": 2.5532126426696777, "learning_rate": 6.906895208414492e-06, "loss": 2.5228, "step": 591 }, { "epoch": 0.1383541234553475, "grad_norm": 2.5823137760162354, "learning_rate": 6.918582002337359e-06, "loss": 2.6085, "step": 592 }, { "epoch": 0.138587829744968, "grad_norm": 3.915198564529419, "learning_rate": 6.9302687962602266e-06, "loss": 2.5732, "step": 593 }, { "epoch": 0.13882153603458852, "grad_norm": 3.377697706222534, "learning_rate": 6.941955590183094e-06, "loss": 2.3571, "step": 594 }, { "epoch": 0.13905524232420904, "grad_norm": 2.114248514175415, "learning_rate": 6.95364238410596e-06, "loss": 2.3505, "step": 595 }, { "epoch": 0.13928894861382957, "grad_norm": 2.1866402626037598, "learning_rate": 6.965329178028828e-06, "loss": 2.6384, "step": 596 }, { "epoch": 0.1395226549034501, "grad_norm": 3.437279462814331, "learning_rate": 6.977015971951695e-06, "loss": 2.496, "step": 597 }, { "epoch": 0.1397563611930706, "grad_norm": 2.6444029808044434, "learning_rate": 6.988702765874563e-06, "loss": 2.4716, "step": 598 }, { "epoch": 0.13999006748269113, "grad_norm": 2.237706184387207, "learning_rate": 7.000389559797429e-06, "loss": 2.4132, "step": 599 }, { "epoch": 0.14022377377231166, "grad_norm": 2.277073383331299, "learning_rate": 7.012076353720296e-06, "loss": 2.503, "step": 600 }, { "epoch": 0.14045748006193218, "grad_norm": 2.186126470565796, "learning_rate": 7.023763147643164e-06, "loss": 2.6011, "step": 601 }, { "epoch": 0.1406911863515527, "grad_norm": 1.9935530424118042, "learning_rate": 7.035449941566031e-06, "loss": 2.5401, "step": 602 }, { "epoch": 0.1409248926411732, "grad_norm": 2.0473475456237793, "learning_rate": 7.047136735488897e-06, "loss": 2.5594, "step": 603 }, { "epoch": 0.14115859893079372, "grad_norm": 2.1723902225494385, "learning_rate": 7.058823529411765e-06, "loss": 2.4929, "step": 604 }, { "epoch": 0.14139230522041424, "grad_norm": 2.5996267795562744, "learning_rate": 7.070510323334632e-06, "loss": 2.498, "step": 605 }, { "epoch": 0.14162601151003476, "grad_norm": 2.151992082595825, "learning_rate": 7.0821971172575e-06, "loss": 2.6587, "step": 606 }, { "epoch": 0.14185971779965528, "grad_norm": 2.1938679218292236, "learning_rate": 7.093883911180366e-06, "loss": 2.652, "step": 607 }, { "epoch": 0.1420934240892758, "grad_norm": 2.6135871410369873, "learning_rate": 7.105570705103233e-06, "loss": 2.5112, "step": 608 }, { "epoch": 0.14232713037889633, "grad_norm": 2.5082945823669434, "learning_rate": 7.117257499026101e-06, "loss": 2.4989, "step": 609 }, { "epoch": 0.14256083666851685, "grad_norm": 2.335653066635132, "learning_rate": 7.128944292948968e-06, "loss": 2.4919, "step": 610 }, { "epoch": 0.14279454295813737, "grad_norm": 2.2056307792663574, "learning_rate": 7.140631086871835e-06, "loss": 2.5161, "step": 611 }, { "epoch": 0.1430282492477579, "grad_norm": 2.220576763153076, "learning_rate": 7.152317880794702e-06, "loss": 2.5665, "step": 612 }, { "epoch": 0.1432619555373784, "grad_norm": 2.3713126182556152, "learning_rate": 7.164004674717569e-06, "loss": 2.4181, "step": 613 }, { "epoch": 0.1434956618269989, "grad_norm": 2.2263615131378174, "learning_rate": 7.175691468640437e-06, "loss": 2.4899, "step": 614 }, { "epoch": 0.14372936811661943, "grad_norm": 2.4039759635925293, "learning_rate": 7.187378262563303e-06, "loss": 2.6517, "step": 615 }, { "epoch": 0.14396307440623995, "grad_norm": 2.271540403366089, "learning_rate": 7.19906505648617e-06, "loss": 2.7561, "step": 616 }, { "epoch": 0.14419678069586048, "grad_norm": 2.3008718490600586, "learning_rate": 7.210751850409038e-06, "loss": 2.4514, "step": 617 }, { "epoch": 0.144430486985481, "grad_norm": 2.523449420928955, "learning_rate": 7.222438644331905e-06, "loss": 2.5074, "step": 618 }, { "epoch": 0.14466419327510152, "grad_norm": 2.2768819332122803, "learning_rate": 7.234125438254772e-06, "loss": 2.545, "step": 619 }, { "epoch": 0.14489789956472204, "grad_norm": 3.3530800342559814, "learning_rate": 7.245812232177639e-06, "loss": 2.5677, "step": 620 }, { "epoch": 0.14513160585434257, "grad_norm": 2.362889289855957, "learning_rate": 7.2574990261005064e-06, "loss": 2.5873, "step": 621 }, { "epoch": 0.1453653121439631, "grad_norm": 2.1265439987182617, "learning_rate": 7.269185820023374e-06, "loss": 2.4192, "step": 622 }, { "epoch": 0.14559901843358358, "grad_norm": 2.3700616359710693, "learning_rate": 7.2808726139462405e-06, "loss": 2.3793, "step": 623 }, { "epoch": 0.1458327247232041, "grad_norm": 2.2295432090759277, "learning_rate": 7.2925594078691076e-06, "loss": 2.4832, "step": 624 }, { "epoch": 0.14606643101282463, "grad_norm": 2.8807389736175537, "learning_rate": 7.3042462017919755e-06, "loss": 2.3728, "step": 625 }, { "epoch": 0.14630013730244515, "grad_norm": 2.815558671951294, "learning_rate": 7.3159329957148425e-06, "loss": 2.3403, "step": 626 }, { "epoch": 0.14653384359206567, "grad_norm": 2.501708507537842, "learning_rate": 7.3276197896377096e-06, "loss": 2.5517, "step": 627 }, { "epoch": 0.1467675498816862, "grad_norm": 2.42330265045166, "learning_rate": 7.339306583560577e-06, "loss": 2.4816, "step": 628 }, { "epoch": 0.14700125617130672, "grad_norm": 3.3611350059509277, "learning_rate": 7.350993377483444e-06, "loss": 2.6485, "step": 629 }, { "epoch": 0.14723496246092724, "grad_norm": 2.8350884914398193, "learning_rate": 7.3626801714063115e-06, "loss": 2.3502, "step": 630 }, { "epoch": 0.14746866875054776, "grad_norm": 2.8480384349823, "learning_rate": 7.374366965329178e-06, "loss": 2.5161, "step": 631 }, { "epoch": 0.14770237504016828, "grad_norm": 2.1526386737823486, "learning_rate": 7.386053759252046e-06, "loss": 2.5334, "step": 632 }, { "epoch": 0.14793608132978878, "grad_norm": 2.4057161808013916, "learning_rate": 7.397740553174913e-06, "loss": 2.4862, "step": 633 }, { "epoch": 0.1481697876194093, "grad_norm": 2.8833045959472656, "learning_rate": 7.40942734709778e-06, "loss": 2.6932, "step": 634 }, { "epoch": 0.14840349390902982, "grad_norm": 2.2119948863983154, "learning_rate": 7.421114141020647e-06, "loss": 2.5086, "step": 635 }, { "epoch": 0.14863720019865034, "grad_norm": 1.9735182523727417, "learning_rate": 7.432800934943514e-06, "loss": 2.5011, "step": 636 }, { "epoch": 0.14887090648827087, "grad_norm": 2.083779811859131, "learning_rate": 7.444487728866381e-06, "loss": 2.4343, "step": 637 }, { "epoch": 0.1491046127778914, "grad_norm": 2.5325167179107666, "learning_rate": 7.456174522789249e-06, "loss": 2.4725, "step": 638 }, { "epoch": 0.1493383190675119, "grad_norm": 2.451310873031616, "learning_rate": 7.467861316712115e-06, "loss": 2.587, "step": 639 }, { "epoch": 0.14957202535713243, "grad_norm": 2.2491745948791504, "learning_rate": 7.479548110634983e-06, "loss": 2.4016, "step": 640 }, { "epoch": 0.14980573164675295, "grad_norm": 2.086554765701294, "learning_rate": 7.49123490455785e-06, "loss": 2.4736, "step": 641 }, { "epoch": 0.15003943793637348, "grad_norm": 2.331108331680298, "learning_rate": 7.502921698480716e-06, "loss": 2.4363, "step": 642 }, { "epoch": 0.15027314422599397, "grad_norm": 2.2278800010681152, "learning_rate": 7.514608492403585e-06, "loss": 2.5792, "step": 643 }, { "epoch": 0.1505068505156145, "grad_norm": 2.3517189025878906, "learning_rate": 7.526295286326451e-06, "loss": 2.4088, "step": 644 }, { "epoch": 0.15074055680523502, "grad_norm": 2.6084752082824707, "learning_rate": 7.537982080249318e-06, "loss": 2.4671, "step": 645 }, { "epoch": 0.15097426309485554, "grad_norm": 2.137871265411377, "learning_rate": 7.549668874172186e-06, "loss": 2.4864, "step": 646 }, { "epoch": 0.15120796938447606, "grad_norm": 2.358049154281616, "learning_rate": 7.561355668095052e-06, "loss": 2.4262, "step": 647 }, { "epoch": 0.15144167567409658, "grad_norm": 2.619272232055664, "learning_rate": 7.573042462017921e-06, "loss": 2.5075, "step": 648 }, { "epoch": 0.1516753819637171, "grad_norm": 2.3063642978668213, "learning_rate": 7.584729255940787e-06, "loss": 2.3149, "step": 649 }, { "epoch": 0.15190908825333763, "grad_norm": 2.1334166526794434, "learning_rate": 7.596416049863653e-06, "loss": 2.7551, "step": 650 }, { "epoch": 0.15214279454295815, "grad_norm": 2.1670477390289307, "learning_rate": 7.608102843786522e-06, "loss": 2.5241, "step": 651 }, { "epoch": 0.15237650083257864, "grad_norm": 2.263967275619507, "learning_rate": 7.619789637709388e-06, "loss": 2.4313, "step": 652 }, { "epoch": 0.15261020712219917, "grad_norm": 2.4283833503723145, "learning_rate": 7.631476431632254e-06, "loss": 2.5331, "step": 653 }, { "epoch": 0.1528439134118197, "grad_norm": 2.081860303878784, "learning_rate": 7.643163225555123e-06, "loss": 2.551, "step": 654 }, { "epoch": 0.1530776197014402, "grad_norm": 2.1040759086608887, "learning_rate": 7.65485001947799e-06, "loss": 2.5441, "step": 655 }, { "epoch": 0.15331132599106073, "grad_norm": 3.0086660385131836, "learning_rate": 7.666536813400857e-06, "loss": 2.3479, "step": 656 }, { "epoch": 0.15354503228068125, "grad_norm": 2.0043485164642334, "learning_rate": 7.678223607323724e-06, "loss": 2.5306, "step": 657 }, { "epoch": 0.15377873857030178, "grad_norm": 2.2797183990478516, "learning_rate": 7.689910401246591e-06, "loss": 2.3876, "step": 658 }, { "epoch": 0.1540124448599223, "grad_norm": 2.238813877105713, "learning_rate": 7.701597195169458e-06, "loss": 2.5173, "step": 659 }, { "epoch": 0.15424615114954282, "grad_norm": 2.0981605052948, "learning_rate": 7.713283989092325e-06, "loss": 2.5192, "step": 660 }, { "epoch": 0.15447985743916334, "grad_norm": 2.0497324466705322, "learning_rate": 7.724970783015193e-06, "loss": 2.4352, "step": 661 }, { "epoch": 0.15471356372878384, "grad_norm": 2.1874442100524902, "learning_rate": 7.73665757693806e-06, "loss": 2.347, "step": 662 }, { "epoch": 0.15494727001840436, "grad_norm": 2.114384174346924, "learning_rate": 7.748344370860927e-06, "loss": 2.3433, "step": 663 }, { "epoch": 0.15518097630802488, "grad_norm": 2.3342278003692627, "learning_rate": 7.760031164783795e-06, "loss": 2.6232, "step": 664 }, { "epoch": 0.1554146825976454, "grad_norm": 2.8513519763946533, "learning_rate": 7.771717958706662e-06, "loss": 2.5893, "step": 665 }, { "epoch": 0.15564838888726593, "grad_norm": 2.451432228088379, "learning_rate": 7.783404752629528e-06, "loss": 2.4464, "step": 666 }, { "epoch": 0.15588209517688645, "grad_norm": 2.248469829559326, "learning_rate": 7.795091546552397e-06, "loss": 2.6412, "step": 667 }, { "epoch": 0.15611580146650697, "grad_norm": 2.276488780975342, "learning_rate": 7.806778340475264e-06, "loss": 2.5335, "step": 668 }, { "epoch": 0.1563495077561275, "grad_norm": 2.0763018131256104, "learning_rate": 7.818465134398129e-06, "loss": 2.3355, "step": 669 }, { "epoch": 0.15658321404574801, "grad_norm": 2.15838360786438, "learning_rate": 7.830151928320998e-06, "loss": 2.4997, "step": 670 }, { "epoch": 0.15681692033536854, "grad_norm": 2.3858187198638916, "learning_rate": 7.841838722243865e-06, "loss": 2.3064, "step": 671 }, { "epoch": 0.15705062662498903, "grad_norm": 3.5563297271728516, "learning_rate": 7.853525516166732e-06, "loss": 2.2532, "step": 672 }, { "epoch": 0.15728433291460955, "grad_norm": 2.4391679763793945, "learning_rate": 7.865212310089599e-06, "loss": 2.5706, "step": 673 }, { "epoch": 0.15751803920423008, "grad_norm": 2.2624807357788086, "learning_rate": 7.876899104012466e-06, "loss": 2.4757, "step": 674 }, { "epoch": 0.1577517454938506, "grad_norm": 2.3098552227020264, "learning_rate": 7.888585897935333e-06, "loss": 2.4344, "step": 675 }, { "epoch": 0.15798545178347112, "grad_norm": 2.769314765930176, "learning_rate": 7.9002726918582e-06, "loss": 2.4679, "step": 676 }, { "epoch": 0.15821915807309164, "grad_norm": 4.038566589355469, "learning_rate": 7.911959485781069e-06, "loss": 2.5252, "step": 677 }, { "epoch": 0.15845286436271216, "grad_norm": 3.2646915912628174, "learning_rate": 7.923646279703936e-06, "loss": 2.5095, "step": 678 }, { "epoch": 0.1586865706523327, "grad_norm": 2.555297613143921, "learning_rate": 7.935333073626801e-06, "loss": 2.5269, "step": 679 }, { "epoch": 0.1589202769419532, "grad_norm": 2.104368209838867, "learning_rate": 7.94701986754967e-06, "loss": 2.5673, "step": 680 }, { "epoch": 0.15915398323157373, "grad_norm": 3.197493314743042, "learning_rate": 7.958706661472537e-06, "loss": 2.4647, "step": 681 }, { "epoch": 0.15938768952119423, "grad_norm": 3.1871395111083984, "learning_rate": 7.970393455395402e-06, "loss": 2.5052, "step": 682 }, { "epoch": 0.15962139581081475, "grad_norm": 2.282222270965576, "learning_rate": 7.982080249318271e-06, "loss": 2.4029, "step": 683 }, { "epoch": 0.15985510210043527, "grad_norm": 2.541109085083008, "learning_rate": 7.993767043241138e-06, "loss": 2.4551, "step": 684 }, { "epoch": 0.1600888083900558, "grad_norm": 2.5483603477478027, "learning_rate": 8.005453837164005e-06, "loss": 2.6976, "step": 685 }, { "epoch": 0.16032251467967631, "grad_norm": 2.2278764247894287, "learning_rate": 8.017140631086872e-06, "loss": 2.4018, "step": 686 }, { "epoch": 0.16055622096929684, "grad_norm": 2.303342819213867, "learning_rate": 8.028827425009739e-06, "loss": 2.4726, "step": 687 }, { "epoch": 0.16078992725891736, "grad_norm": 2.764023780822754, "learning_rate": 8.040514218932606e-06, "loss": 2.5756, "step": 688 }, { "epoch": 0.16102363354853788, "grad_norm": 3.1167659759521484, "learning_rate": 8.052201012855473e-06, "loss": 2.4013, "step": 689 }, { "epoch": 0.1612573398381584, "grad_norm": 2.5984694957733154, "learning_rate": 8.06388780677834e-06, "loss": 2.5335, "step": 690 }, { "epoch": 0.16149104612777893, "grad_norm": 3.6351609230041504, "learning_rate": 8.075574600701209e-06, "loss": 2.5477, "step": 691 }, { "epoch": 0.16172475241739942, "grad_norm": 2.213742256164551, "learning_rate": 8.087261394624074e-06, "loss": 2.427, "step": 692 }, { "epoch": 0.16195845870701994, "grad_norm": 2.6370818614959717, "learning_rate": 8.098948188546943e-06, "loss": 2.6143, "step": 693 }, { "epoch": 0.16219216499664046, "grad_norm": 2.220344066619873, "learning_rate": 8.11063498246981e-06, "loss": 2.3649, "step": 694 }, { "epoch": 0.162425871286261, "grad_norm": 2.064901828765869, "learning_rate": 8.122321776392676e-06, "loss": 2.5648, "step": 695 }, { "epoch": 0.1626595775758815, "grad_norm": 3.114729404449463, "learning_rate": 8.134008570315544e-06, "loss": 2.5447, "step": 696 }, { "epoch": 0.16289328386550203, "grad_norm": 3.0480310916900635, "learning_rate": 8.145695364238411e-06, "loss": 2.3883, "step": 697 }, { "epoch": 0.16312699015512255, "grad_norm": 2.386691093444824, "learning_rate": 8.157382158161277e-06, "loss": 2.5621, "step": 698 }, { "epoch": 0.16336069644474308, "grad_norm": 2.0870609283447266, "learning_rate": 8.169068952084145e-06, "loss": 2.3842, "step": 699 }, { "epoch": 0.1635944027343636, "grad_norm": 2.559891939163208, "learning_rate": 8.180755746007012e-06, "loss": 2.5684, "step": 700 }, { "epoch": 0.16382810902398412, "grad_norm": 2.262017250061035, "learning_rate": 8.19244253992988e-06, "loss": 2.5824, "step": 701 }, { "epoch": 0.16406181531360461, "grad_norm": 2.3484129905700684, "learning_rate": 8.204129333852747e-06, "loss": 2.3638, "step": 702 }, { "epoch": 0.16429552160322514, "grad_norm": 2.1168582439422607, "learning_rate": 8.215816127775614e-06, "loss": 2.4782, "step": 703 }, { "epoch": 0.16452922789284566, "grad_norm": 2.1998116970062256, "learning_rate": 8.227502921698482e-06, "loss": 2.5748, "step": 704 }, { "epoch": 0.16476293418246618, "grad_norm": 2.2595369815826416, "learning_rate": 8.239189715621348e-06, "loss": 2.46, "step": 705 }, { "epoch": 0.1649966404720867, "grad_norm": 2.4693796634674072, "learning_rate": 8.250876509544215e-06, "loss": 2.5311, "step": 706 }, { "epoch": 0.16523034676170723, "grad_norm": 2.3193697929382324, "learning_rate": 8.262563303467083e-06, "loss": 2.5054, "step": 707 }, { "epoch": 0.16546405305132775, "grad_norm": 2.396204710006714, "learning_rate": 8.274250097389949e-06, "loss": 2.5171, "step": 708 }, { "epoch": 0.16569775934094827, "grad_norm": 2.2344939708709717, "learning_rate": 8.285936891312818e-06, "loss": 2.3789, "step": 709 }, { "epoch": 0.1659314656305688, "grad_norm": 2.2725284099578857, "learning_rate": 8.297623685235685e-06, "loss": 2.4338, "step": 710 }, { "epoch": 0.16616517192018931, "grad_norm": 2.161031484603882, "learning_rate": 8.30931047915855e-06, "loss": 2.3234, "step": 711 }, { "epoch": 0.1663988782098098, "grad_norm": 2.2503321170806885, "learning_rate": 8.320997273081419e-06, "loss": 2.4868, "step": 712 }, { "epoch": 0.16663258449943033, "grad_norm": 2.534825325012207, "learning_rate": 8.332684067004286e-06, "loss": 2.5519, "step": 713 }, { "epoch": 0.16686629078905085, "grad_norm": 3.1620187759399414, "learning_rate": 8.344370860927151e-06, "loss": 2.475, "step": 714 }, { "epoch": 0.16709999707867138, "grad_norm": 2.8278424739837646, "learning_rate": 8.35605765485002e-06, "loss": 2.3621, "step": 715 }, { "epoch": 0.1673337033682919, "grad_norm": 2.6000454425811768, "learning_rate": 8.367744448772887e-06, "loss": 2.535, "step": 716 }, { "epoch": 0.16756740965791242, "grad_norm": 2.1834044456481934, "learning_rate": 8.379431242695754e-06, "loss": 2.4851, "step": 717 }, { "epoch": 0.16780111594753294, "grad_norm": 2.3175716400146484, "learning_rate": 8.391118036618621e-06, "loss": 2.4198, "step": 718 }, { "epoch": 0.16803482223715346, "grad_norm": 2.8295342922210693, "learning_rate": 8.402804830541488e-06, "loss": 2.4588, "step": 719 }, { "epoch": 0.16826852852677399, "grad_norm": 2.2020294666290283, "learning_rate": 8.414491624464357e-06, "loss": 2.3948, "step": 720 }, { "epoch": 0.1685022348163945, "grad_norm": 2.3889083862304688, "learning_rate": 8.426178418387222e-06, "loss": 2.4905, "step": 721 }, { "epoch": 0.168735941106015, "grad_norm": 2.168710708618164, "learning_rate": 8.43786521231009e-06, "loss": 2.4856, "step": 722 }, { "epoch": 0.16896964739563552, "grad_norm": 2.5213334560394287, "learning_rate": 8.449552006232958e-06, "loss": 2.6522, "step": 723 }, { "epoch": 0.16920335368525605, "grad_norm": 2.4963572025299072, "learning_rate": 8.461238800155823e-06, "loss": 2.4293, "step": 724 }, { "epoch": 0.16943705997487657, "grad_norm": 3.518925905227661, "learning_rate": 8.472925594078692e-06, "loss": 2.3173, "step": 725 }, { "epoch": 0.1696707662644971, "grad_norm": 2.3839471340179443, "learning_rate": 8.484612388001559e-06, "loss": 2.5837, "step": 726 }, { "epoch": 0.1699044725541176, "grad_norm": 2.129446268081665, "learning_rate": 8.496299181924424e-06, "loss": 2.5568, "step": 727 }, { "epoch": 0.17013817884373814, "grad_norm": 2.5956435203552246, "learning_rate": 8.507985975847293e-06, "loss": 2.5597, "step": 728 }, { "epoch": 0.17037188513335866, "grad_norm": 3.111035108566284, "learning_rate": 8.51967276977016e-06, "loss": 2.3526, "step": 729 }, { "epoch": 0.17060559142297918, "grad_norm": 3.055586576461792, "learning_rate": 8.531359563693026e-06, "loss": 2.3369, "step": 730 }, { "epoch": 0.1708392977125997, "grad_norm": 2.8054683208465576, "learning_rate": 8.543046357615894e-06, "loss": 2.3907, "step": 731 }, { "epoch": 0.1710730040022202, "grad_norm": 2.291343927383423, "learning_rate": 8.554733151538761e-06, "loss": 2.4289, "step": 732 }, { "epoch": 0.17130671029184072, "grad_norm": 2.2767724990844727, "learning_rate": 8.56641994546163e-06, "loss": 2.2902, "step": 733 }, { "epoch": 0.17154041658146124, "grad_norm": 2.229933500289917, "learning_rate": 8.578106739384495e-06, "loss": 2.4241, "step": 734 }, { "epoch": 0.17177412287108176, "grad_norm": 2.140052080154419, "learning_rate": 8.589793533307362e-06, "loss": 2.3747, "step": 735 }, { "epoch": 0.17200782916070229, "grad_norm": 2.345613956451416, "learning_rate": 8.601480327230231e-06, "loss": 2.3279, "step": 736 }, { "epoch": 0.1722415354503228, "grad_norm": 3.337568998336792, "learning_rate": 8.613167121153097e-06, "loss": 2.5377, "step": 737 }, { "epoch": 0.17247524173994333, "grad_norm": 2.3622679710388184, "learning_rate": 8.624853915075964e-06, "loss": 2.4293, "step": 738 }, { "epoch": 0.17270894802956385, "grad_norm": 2.1647017002105713, "learning_rate": 8.636540708998832e-06, "loss": 2.5273, "step": 739 }, { "epoch": 0.17294265431918437, "grad_norm": 2.3732168674468994, "learning_rate": 8.648227502921698e-06, "loss": 2.3143, "step": 740 }, { "epoch": 0.1731763606088049, "grad_norm": 3.105177879333496, "learning_rate": 8.659914296844566e-06, "loss": 2.332, "step": 741 }, { "epoch": 0.1734100668984254, "grad_norm": 2.6960668563842773, "learning_rate": 8.671601090767433e-06, "loss": 2.3804, "step": 742 }, { "epoch": 0.1736437731880459, "grad_norm": 2.2382943630218506, "learning_rate": 8.683287884690299e-06, "loss": 2.5132, "step": 743 }, { "epoch": 0.17387747947766644, "grad_norm": 2.4285175800323486, "learning_rate": 8.694974678613168e-06, "loss": 2.4546, "step": 744 }, { "epoch": 0.17411118576728696, "grad_norm": 2.7899773120880127, "learning_rate": 8.706661472536035e-06, "loss": 2.3506, "step": 745 }, { "epoch": 0.17434489205690748, "grad_norm": 2.1851911544799805, "learning_rate": 8.718348266458902e-06, "loss": 2.3153, "step": 746 }, { "epoch": 0.174578598346528, "grad_norm": 2.4295687675476074, "learning_rate": 8.730035060381769e-06, "loss": 2.3204, "step": 747 }, { "epoch": 0.17481230463614852, "grad_norm": 2.7670507431030273, "learning_rate": 8.741721854304636e-06, "loss": 2.5051, "step": 748 }, { "epoch": 0.17504601092576905, "grad_norm": 2.3826115131378174, "learning_rate": 8.753408648227505e-06, "loss": 2.3936, "step": 749 }, { "epoch": 0.17527971721538957, "grad_norm": 2.986328601837158, "learning_rate": 8.76509544215037e-06, "loss": 2.4599, "step": 750 }, { "epoch": 0.1755134235050101, "grad_norm": 2.7266275882720947, "learning_rate": 8.776782236073237e-06, "loss": 2.4185, "step": 751 }, { "epoch": 0.17574712979463059, "grad_norm": 2.2913856506347656, "learning_rate": 8.788469029996106e-06, "loss": 2.3193, "step": 752 }, { "epoch": 0.1759808360842511, "grad_norm": 2.2450358867645264, "learning_rate": 8.800155823918971e-06, "loss": 2.4831, "step": 753 }, { "epoch": 0.17621454237387163, "grad_norm": 3.0544638633728027, "learning_rate": 8.811842617841838e-06, "loss": 2.3734, "step": 754 }, { "epoch": 0.17644824866349215, "grad_norm": 2.2578299045562744, "learning_rate": 8.823529411764707e-06, "loss": 2.4677, "step": 755 }, { "epoch": 0.17668195495311267, "grad_norm": 2.33188533782959, "learning_rate": 8.835216205687572e-06, "loss": 2.5774, "step": 756 }, { "epoch": 0.1769156612427332, "grad_norm": 2.233891248703003, "learning_rate": 8.846902999610441e-06, "loss": 2.4207, "step": 757 }, { "epoch": 0.17714936753235372, "grad_norm": 2.5160248279571533, "learning_rate": 8.858589793533308e-06, "loss": 2.5021, "step": 758 }, { "epoch": 0.17738307382197424, "grad_norm": 2.3079090118408203, "learning_rate": 8.870276587456175e-06, "loss": 2.3257, "step": 759 }, { "epoch": 0.17761678011159476, "grad_norm": 2.4692370891571045, "learning_rate": 8.881963381379042e-06, "loss": 2.3522, "step": 760 }, { "epoch": 0.17785048640121529, "grad_norm": 2.1909894943237305, "learning_rate": 8.893650175301909e-06, "loss": 2.327, "step": 761 }, { "epoch": 0.17808419269083578, "grad_norm": 2.068303108215332, "learning_rate": 8.905336969224776e-06, "loss": 2.3896, "step": 762 }, { "epoch": 0.1783178989804563, "grad_norm": 2.321049690246582, "learning_rate": 8.917023763147643e-06, "loss": 2.2532, "step": 763 }, { "epoch": 0.17855160527007682, "grad_norm": 2.260704517364502, "learning_rate": 8.92871055707051e-06, "loss": 2.334, "step": 764 }, { "epoch": 0.17878531155969735, "grad_norm": 2.7114710807800293, "learning_rate": 8.940397350993379e-06, "loss": 2.5278, "step": 765 }, { "epoch": 0.17901901784931787, "grad_norm": 2.3078997135162354, "learning_rate": 8.952084144916244e-06, "loss": 2.2192, "step": 766 }, { "epoch": 0.1792527241389384, "grad_norm": 2.4261951446533203, "learning_rate": 8.963770938839111e-06, "loss": 2.5571, "step": 767 }, { "epoch": 0.1794864304285589, "grad_norm": 2.7658700942993164, "learning_rate": 8.97545773276198e-06, "loss": 2.4441, "step": 768 }, { "epoch": 0.17972013671817944, "grad_norm": 2.3395028114318848, "learning_rate": 8.987144526684845e-06, "loss": 2.3533, "step": 769 }, { "epoch": 0.17995384300779996, "grad_norm": 2.361670970916748, "learning_rate": 8.998831320607713e-06, "loss": 2.2535, "step": 770 }, { "epoch": 0.18018754929742048, "grad_norm": 2.1553263664245605, "learning_rate": 9.010518114530581e-06, "loss": 2.4492, "step": 771 }, { "epoch": 0.18042125558704097, "grad_norm": 2.301483154296875, "learning_rate": 9.022204908453448e-06, "loss": 2.5009, "step": 772 }, { "epoch": 0.1806549618766615, "grad_norm": 2.3756024837493896, "learning_rate": 9.033891702376315e-06, "loss": 2.3814, "step": 773 }, { "epoch": 0.18088866816628202, "grad_norm": 2.8396832942962646, "learning_rate": 9.045578496299182e-06, "loss": 2.5379, "step": 774 }, { "epoch": 0.18112237445590254, "grad_norm": 2.1430435180664062, "learning_rate": 9.05726529022205e-06, "loss": 2.3638, "step": 775 }, { "epoch": 0.18135608074552306, "grad_norm": 2.3005425930023193, "learning_rate": 9.068952084144916e-06, "loss": 2.5398, "step": 776 }, { "epoch": 0.18158978703514358, "grad_norm": 2.6705265045166016, "learning_rate": 9.080638878067784e-06, "loss": 2.4433, "step": 777 }, { "epoch": 0.1818234933247641, "grad_norm": 2.4219281673431396, "learning_rate": 9.09232567199065e-06, "loss": 2.3798, "step": 778 }, { "epoch": 0.18205719961438463, "grad_norm": 2.473064422607422, "learning_rate": 9.104012465913518e-06, "loss": 2.3202, "step": 779 }, { "epoch": 0.18229090590400515, "grad_norm": 2.0566771030426025, "learning_rate": 9.115699259836385e-06, "loss": 2.6096, "step": 780 }, { "epoch": 0.18252461219362567, "grad_norm": 2.0159518718719482, "learning_rate": 9.127386053759253e-06, "loss": 2.5534, "step": 781 }, { "epoch": 0.18275831848324617, "grad_norm": 2.257366418838501, "learning_rate": 9.139072847682119e-06, "loss": 2.5708, "step": 782 }, { "epoch": 0.1829920247728667, "grad_norm": 2.4249541759490967, "learning_rate": 9.150759641604986e-06, "loss": 2.2341, "step": 783 }, { "epoch": 0.1832257310624872, "grad_norm": 2.3003406524658203, "learning_rate": 9.162446435527855e-06, "loss": 2.5848, "step": 784 }, { "epoch": 0.18345943735210773, "grad_norm": 2.503593683242798, "learning_rate": 9.17413322945072e-06, "loss": 2.4749, "step": 785 }, { "epoch": 0.18369314364172826, "grad_norm": 2.39269757270813, "learning_rate": 9.185820023373587e-06, "loss": 2.4464, "step": 786 }, { "epoch": 0.18392684993134878, "grad_norm": 2.193746328353882, "learning_rate": 9.197506817296456e-06, "loss": 2.4893, "step": 787 }, { "epoch": 0.1841605562209693, "grad_norm": 2.3978044986724854, "learning_rate": 9.209193611219323e-06, "loss": 2.3096, "step": 788 }, { "epoch": 0.18439426251058982, "grad_norm": 2.1503632068634033, "learning_rate": 9.22088040514219e-06, "loss": 2.3515, "step": 789 }, { "epoch": 0.18462796880021035, "grad_norm": 2.5877907276153564, "learning_rate": 9.232567199065057e-06, "loss": 2.3331, "step": 790 }, { "epoch": 0.18486167508983087, "grad_norm": 3.331425189971924, "learning_rate": 9.244253992987924e-06, "loss": 2.2708, "step": 791 }, { "epoch": 0.18509538137945136, "grad_norm": 2.1755967140197754, "learning_rate": 9.255940786910791e-06, "loss": 2.324, "step": 792 }, { "epoch": 0.18532908766907188, "grad_norm": 2.0453715324401855, "learning_rate": 9.267627580833658e-06, "loss": 2.5007, "step": 793 }, { "epoch": 0.1855627939586924, "grad_norm": 2.6125898361206055, "learning_rate": 9.279314374756525e-06, "loss": 2.4145, "step": 794 }, { "epoch": 0.18579650024831293, "grad_norm": 2.457577705383301, "learning_rate": 9.291001168679392e-06, "loss": 2.4781, "step": 795 }, { "epoch": 0.18603020653793345, "grad_norm": 2.0882809162139893, "learning_rate": 9.302687962602259e-06, "loss": 2.4856, "step": 796 }, { "epoch": 0.18626391282755397, "grad_norm": 2.073916435241699, "learning_rate": 9.314374756525128e-06, "loss": 2.4601, "step": 797 }, { "epoch": 0.1864976191171745, "grad_norm": 2.3498222827911377, "learning_rate": 9.326061550447993e-06, "loss": 2.4671, "step": 798 }, { "epoch": 0.18673132540679502, "grad_norm": 2.36476469039917, "learning_rate": 9.33774834437086e-06, "loss": 2.4888, "step": 799 }, { "epoch": 0.18696503169641554, "grad_norm": 2.1305270195007324, "learning_rate": 9.349435138293729e-06, "loss": 2.2456, "step": 800 }, { "epoch": 0.18719873798603606, "grad_norm": 2.138932943344116, "learning_rate": 9.361121932216596e-06, "loss": 2.2769, "step": 801 }, { "epoch": 0.18743244427565656, "grad_norm": 2.030771255493164, "learning_rate": 9.372808726139461e-06, "loss": 2.4625, "step": 802 }, { "epoch": 0.18766615056527708, "grad_norm": 2.130605697631836, "learning_rate": 9.38449552006233e-06, "loss": 2.4104, "step": 803 }, { "epoch": 0.1878998568548976, "grad_norm": 2.29727840423584, "learning_rate": 9.396182313985197e-06, "loss": 2.4129, "step": 804 }, { "epoch": 0.18813356314451812, "grad_norm": 2.2407619953155518, "learning_rate": 9.407869107908064e-06, "loss": 2.4315, "step": 805 }, { "epoch": 0.18836726943413865, "grad_norm": 2.231020450592041, "learning_rate": 9.419555901830931e-06, "loss": 2.5515, "step": 806 }, { "epoch": 0.18860097572375917, "grad_norm": 3.7723989486694336, "learning_rate": 9.431242695753798e-06, "loss": 2.284, "step": 807 }, { "epoch": 0.1888346820133797, "grad_norm": 3.0208940505981445, "learning_rate": 9.442929489676665e-06, "loss": 2.3838, "step": 808 }, { "epoch": 0.1890683883030002, "grad_norm": 2.352492094039917, "learning_rate": 9.454616283599532e-06, "loss": 2.4821, "step": 809 }, { "epoch": 0.18930209459262073, "grad_norm": 3.1655259132385254, "learning_rate": 9.4663030775224e-06, "loss": 2.5127, "step": 810 }, { "epoch": 0.18953580088224126, "grad_norm": 3.3927557468414307, "learning_rate": 9.477989871445267e-06, "loss": 2.3888, "step": 811 }, { "epoch": 0.18976950717186175, "grad_norm": 2.3291330337524414, "learning_rate": 9.489676665368134e-06, "loss": 2.3522, "step": 812 }, { "epoch": 0.19000321346148227, "grad_norm": 2.0993239879608154, "learning_rate": 9.501363459291002e-06, "loss": 2.2976, "step": 813 }, { "epoch": 0.1902369197511028, "grad_norm": 2.333217144012451, "learning_rate": 9.51305025321387e-06, "loss": 2.293, "step": 814 }, { "epoch": 0.19047062604072332, "grad_norm": 3.1906702518463135, "learning_rate": 9.524737047136735e-06, "loss": 2.4119, "step": 815 }, { "epoch": 0.19070433233034384, "grad_norm": 2.3688042163848877, "learning_rate": 9.536423841059603e-06, "loss": 2.4597, "step": 816 }, { "epoch": 0.19093803861996436, "grad_norm": 2.3319361209869385, "learning_rate": 9.54811063498247e-06, "loss": 2.4399, "step": 817 }, { "epoch": 0.19117174490958488, "grad_norm": 3.3750627040863037, "learning_rate": 9.559797428905338e-06, "loss": 2.353, "step": 818 }, { "epoch": 0.1914054511992054, "grad_norm": 2.4419777393341064, "learning_rate": 9.571484222828205e-06, "loss": 2.4156, "step": 819 }, { "epoch": 0.19163915748882593, "grad_norm": 2.307007312774658, "learning_rate": 9.583171016751072e-06, "loss": 2.424, "step": 820 }, { "epoch": 0.19187286377844645, "grad_norm": 2.6114206314086914, "learning_rate": 9.594857810673939e-06, "loss": 2.3313, "step": 821 }, { "epoch": 0.19210657006806695, "grad_norm": 2.5788016319274902, "learning_rate": 9.606544604596806e-06, "loss": 2.3328, "step": 822 }, { "epoch": 0.19234027635768747, "grad_norm": 2.248469114303589, "learning_rate": 9.618231398519673e-06, "loss": 2.4832, "step": 823 }, { "epoch": 0.192573982647308, "grad_norm": 2.4719812870025635, "learning_rate": 9.62991819244254e-06, "loss": 2.2878, "step": 824 }, { "epoch": 0.1928076889369285, "grad_norm": 2.2512786388397217, "learning_rate": 9.641604986365407e-06, "loss": 2.3001, "step": 825 }, { "epoch": 0.19304139522654903, "grad_norm": 2.348910331726074, "learning_rate": 9.653291780288276e-06, "loss": 2.5752, "step": 826 }, { "epoch": 0.19327510151616956, "grad_norm": 2.2469632625579834, "learning_rate": 9.664978574211143e-06, "loss": 2.3134, "step": 827 }, { "epoch": 0.19350880780579008, "grad_norm": 2.285038471221924, "learning_rate": 9.676665368134008e-06, "loss": 2.3635, "step": 828 }, { "epoch": 0.1937425140954106, "grad_norm": 2.5802700519561768, "learning_rate": 9.688352162056877e-06, "loss": 2.4627, "step": 829 }, { "epoch": 0.19397622038503112, "grad_norm": 2.7603089809417725, "learning_rate": 9.700038955979744e-06, "loss": 2.452, "step": 830 }, { "epoch": 0.19420992667465165, "grad_norm": 2.113271951675415, "learning_rate": 9.711725749902609e-06, "loss": 2.3225, "step": 831 }, { "epoch": 0.19444363296427214, "grad_norm": 2.9635891914367676, "learning_rate": 9.723412543825478e-06, "loss": 2.3573, "step": 832 }, { "epoch": 0.19467733925389266, "grad_norm": 2.329213857650757, "learning_rate": 9.735099337748345e-06, "loss": 2.3994, "step": 833 }, { "epoch": 0.19491104554351318, "grad_norm": 2.376818895339966, "learning_rate": 9.746786131671212e-06, "loss": 2.3869, "step": 834 }, { "epoch": 0.1951447518331337, "grad_norm": 2.265974283218384, "learning_rate": 9.758472925594079e-06, "loss": 2.3917, "step": 835 }, { "epoch": 0.19537845812275423, "grad_norm": 2.3179209232330322, "learning_rate": 9.770159719516946e-06, "loss": 2.3642, "step": 836 }, { "epoch": 0.19561216441237475, "grad_norm": 2.291776657104492, "learning_rate": 9.781846513439813e-06, "loss": 2.5127, "step": 837 }, { "epoch": 0.19584587070199527, "grad_norm": 2.1870198249816895, "learning_rate": 9.79353330736268e-06, "loss": 2.5431, "step": 838 }, { "epoch": 0.1960795769916158, "grad_norm": 2.3500962257385254, "learning_rate": 9.805220101285547e-06, "loss": 2.3356, "step": 839 }, { "epoch": 0.19631328328123632, "grad_norm": 3.491117000579834, "learning_rate": 9.816906895208416e-06, "loss": 2.3934, "step": 840 }, { "epoch": 0.19654698957085684, "grad_norm": 2.6922240257263184, "learning_rate": 9.828593689131281e-06, "loss": 2.3726, "step": 841 }, { "epoch": 0.19678069586047733, "grad_norm": 2.842912435531616, "learning_rate": 9.84028048305415e-06, "loss": 2.4973, "step": 842 }, { "epoch": 0.19701440215009786, "grad_norm": 3.6348798274993896, "learning_rate": 9.851967276977017e-06, "loss": 2.6222, "step": 843 }, { "epoch": 0.19724810843971838, "grad_norm": 2.95827054977417, "learning_rate": 9.863654070899882e-06, "loss": 2.4829, "step": 844 }, { "epoch": 0.1974818147293389, "grad_norm": 2.2883331775665283, "learning_rate": 9.875340864822751e-06, "loss": 2.4822, "step": 845 }, { "epoch": 0.19771552101895942, "grad_norm": 2.362381935119629, "learning_rate": 9.887027658745618e-06, "loss": 2.5169, "step": 846 }, { "epoch": 0.19794922730857994, "grad_norm": 3.4531972408294678, "learning_rate": 9.898714452668484e-06, "loss": 2.2862, "step": 847 }, { "epoch": 0.19818293359820047, "grad_norm": 3.4169297218322754, "learning_rate": 9.910401246591352e-06, "loss": 2.3949, "step": 848 }, { "epoch": 0.198416639887821, "grad_norm": 2.227917432785034, "learning_rate": 9.92208804051422e-06, "loss": 2.5509, "step": 849 }, { "epoch": 0.1986503461774415, "grad_norm": 2.1612467765808105, "learning_rate": 9.933774834437086e-06, "loss": 2.5217, "step": 850 }, { "epoch": 0.19888405246706203, "grad_norm": 2.6451425552368164, "learning_rate": 9.945461628359953e-06, "loss": 2.4326, "step": 851 }, { "epoch": 0.19911775875668253, "grad_norm": 2.6196353435516357, "learning_rate": 9.95714842228282e-06, "loss": 2.2249, "step": 852 }, { "epoch": 0.19935146504630305, "grad_norm": 2.159616470336914, "learning_rate": 9.968835216205688e-06, "loss": 2.5051, "step": 853 }, { "epoch": 0.19958517133592357, "grad_norm": 2.162959098815918, "learning_rate": 9.980522010128555e-06, "loss": 2.4987, "step": 854 }, { "epoch": 0.1998188776255441, "grad_norm": 2.1190178394317627, "learning_rate": 9.992208804051422e-06, "loss": 2.4533, "step": 855 }, { "epoch": 0.20005258391516462, "grad_norm": 2.3338894844055176, "learning_rate": 1.000389559797429e-05, "loss": 2.5086, "step": 856 }, { "epoch": 0.20028629020478514, "grad_norm": 2.476609230041504, "learning_rate": 1.0015582391897156e-05, "loss": 2.3749, "step": 857 }, { "epoch": 0.20051999649440566, "grad_norm": 2.4573588371276855, "learning_rate": 1.0027269185820024e-05, "loss": 2.4382, "step": 858 }, { "epoch": 0.20075370278402618, "grad_norm": 2.335963249206543, "learning_rate": 1.0038955979742892e-05, "loss": 2.2865, "step": 859 }, { "epoch": 0.2009874090736467, "grad_norm": 2.7505037784576416, "learning_rate": 1.0050642773665757e-05, "loss": 2.6318, "step": 860 }, { "epoch": 0.20122111536326723, "grad_norm": 2.17611026763916, "learning_rate": 1.0062329567588626e-05, "loss": 2.3246, "step": 861 }, { "epoch": 0.20145482165288772, "grad_norm": 2.140289306640625, "learning_rate": 1.0074016361511493e-05, "loss": 2.3993, "step": 862 }, { "epoch": 0.20168852794250824, "grad_norm": 2.581900119781494, "learning_rate": 1.0085703155434358e-05, "loss": 2.3587, "step": 863 }, { "epoch": 0.20192223423212877, "grad_norm": 2.235598564147949, "learning_rate": 1.0097389949357227e-05, "loss": 2.3013, "step": 864 }, { "epoch": 0.2021559405217493, "grad_norm": 2.3454856872558594, "learning_rate": 1.0109076743280094e-05, "loss": 2.3018, "step": 865 }, { "epoch": 0.2023896468113698, "grad_norm": 4.109177589416504, "learning_rate": 1.0120763537202961e-05, "loss": 2.4647, "step": 866 }, { "epoch": 0.20262335310099033, "grad_norm": 3.2276337146759033, "learning_rate": 1.0132450331125828e-05, "loss": 2.5001, "step": 867 }, { "epoch": 0.20285705939061086, "grad_norm": 2.6417288780212402, "learning_rate": 1.0144137125048695e-05, "loss": 2.3791, "step": 868 }, { "epoch": 0.20309076568023138, "grad_norm": 2.2861664295196533, "learning_rate": 1.0155823918971564e-05, "loss": 2.3356, "step": 869 }, { "epoch": 0.2033244719698519, "grad_norm": 3.0578081607818604, "learning_rate": 1.0167510712894429e-05, "loss": 2.3925, "step": 870 }, { "epoch": 0.2035581782594724, "grad_norm": 2.9976625442504883, "learning_rate": 1.0179197506817296e-05, "loss": 2.3758, "step": 871 }, { "epoch": 0.20379188454909292, "grad_norm": 3.4947171211242676, "learning_rate": 1.0190884300740165e-05, "loss": 2.5898, "step": 872 }, { "epoch": 0.20402559083871344, "grad_norm": 2.2827515602111816, "learning_rate": 1.020257109466303e-05, "loss": 2.3922, "step": 873 }, { "epoch": 0.20425929712833396, "grad_norm": 2.244795560836792, "learning_rate": 1.0214257888585899e-05, "loss": 2.5376, "step": 874 }, { "epoch": 0.20449300341795448, "grad_norm": 2.465303421020508, "learning_rate": 1.0225944682508766e-05, "loss": 2.3044, "step": 875 }, { "epoch": 0.204726709707575, "grad_norm": 3.3525490760803223, "learning_rate": 1.0237631476431631e-05, "loss": 2.358, "step": 876 }, { "epoch": 0.20496041599719553, "grad_norm": 2.8462934494018555, "learning_rate": 1.02493182703545e-05, "loss": 2.3821, "step": 877 }, { "epoch": 0.20519412228681605, "grad_norm": 2.904953956604004, "learning_rate": 1.0261005064277367e-05, "loss": 2.3855, "step": 878 }, { "epoch": 0.20542782857643657, "grad_norm": 3.624619722366333, "learning_rate": 1.0272691858200232e-05, "loss": 2.4641, "step": 879 }, { "epoch": 0.2056615348660571, "grad_norm": 3.370133399963379, "learning_rate": 1.0284378652123101e-05, "loss": 2.5051, "step": 880 }, { "epoch": 0.2058952411556776, "grad_norm": 2.254397392272949, "learning_rate": 1.0296065446045968e-05, "loss": 2.2304, "step": 881 }, { "epoch": 0.2061289474452981, "grad_norm": 2.4976119995117188, "learning_rate": 1.0307752239968837e-05, "loss": 2.4392, "step": 882 }, { "epoch": 0.20636265373491863, "grad_norm": 3.1412363052368164, "learning_rate": 1.0319439033891702e-05, "loss": 2.2762, "step": 883 }, { "epoch": 0.20659636002453916, "grad_norm": 3.272939443588257, "learning_rate": 1.033112582781457e-05, "loss": 2.4003, "step": 884 }, { "epoch": 0.20683006631415968, "grad_norm": 3.3397934436798096, "learning_rate": 1.0342812621737438e-05, "loss": 2.2319, "step": 885 }, { "epoch": 0.2070637726037802, "grad_norm": 2.0607926845550537, "learning_rate": 1.0354499415660303e-05, "loss": 2.2523, "step": 886 }, { "epoch": 0.20729747889340072, "grad_norm": 2.685067653656006, "learning_rate": 1.036618620958317e-05, "loss": 2.4847, "step": 887 }, { "epoch": 0.20753118518302124, "grad_norm": 3.134225845336914, "learning_rate": 1.037787300350604e-05, "loss": 2.5313, "step": 888 }, { "epoch": 0.20776489147264177, "grad_norm": 3.546175479888916, "learning_rate": 1.0389559797428905e-05, "loss": 2.3867, "step": 889 }, { "epoch": 0.2079985977622623, "grad_norm": 2.1829631328582764, "learning_rate": 1.0401246591351773e-05, "loss": 2.4028, "step": 890 }, { "epoch": 0.20823230405188278, "grad_norm": 2.4811604022979736, "learning_rate": 1.041293338527464e-05, "loss": 2.609, "step": 891 }, { "epoch": 0.2084660103415033, "grad_norm": 2.314955234527588, "learning_rate": 1.0424620179197506e-05, "loss": 2.6005, "step": 892 }, { "epoch": 0.20869971663112383, "grad_norm": 2.5671112537384033, "learning_rate": 1.0436306973120375e-05, "loss": 2.4345, "step": 893 }, { "epoch": 0.20893342292074435, "grad_norm": 2.212461471557617, "learning_rate": 1.0447993767043242e-05, "loss": 2.4294, "step": 894 }, { "epoch": 0.20916712921036487, "grad_norm": 2.4538114070892334, "learning_rate": 1.0459680560966109e-05, "loss": 2.4016, "step": 895 }, { "epoch": 0.2094008354999854, "grad_norm": 3.226811408996582, "learning_rate": 1.0471367354888976e-05, "loss": 2.5656, "step": 896 }, { "epoch": 0.20963454178960592, "grad_norm": 2.1684622764587402, "learning_rate": 1.0483054148811843e-05, "loss": 2.6239, "step": 897 }, { "epoch": 0.20986824807922644, "grad_norm": 2.484133720397949, "learning_rate": 1.0494740942734711e-05, "loss": 2.4247, "step": 898 }, { "epoch": 0.21010195436884696, "grad_norm": 3.1199333667755127, "learning_rate": 1.0506427736657577e-05, "loss": 2.4486, "step": 899 }, { "epoch": 0.21033566065846748, "grad_norm": 2.184452772140503, "learning_rate": 1.0518114530580444e-05, "loss": 2.4757, "step": 900 }, { "epoch": 0.21056936694808798, "grad_norm": 2.4003286361694336, "learning_rate": 1.0529801324503313e-05, "loss": 2.367, "step": 901 }, { "epoch": 0.2108030732377085, "grad_norm": 1.9814890623092651, "learning_rate": 1.0541488118426178e-05, "loss": 2.4323, "step": 902 }, { "epoch": 0.21103677952732902, "grad_norm": 2.8638477325439453, "learning_rate": 1.0553174912349045e-05, "loss": 2.4104, "step": 903 }, { "epoch": 0.21127048581694954, "grad_norm": 3.626507520675659, "learning_rate": 1.0564861706271914e-05, "loss": 2.6227, "step": 904 }, { "epoch": 0.21150419210657007, "grad_norm": 2.270120620727539, "learning_rate": 1.0576548500194779e-05, "loss": 2.443, "step": 905 }, { "epoch": 0.2117378983961906, "grad_norm": 2.7630674839019775, "learning_rate": 1.0588235294117648e-05, "loss": 2.3765, "step": 906 }, { "epoch": 0.2119716046858111, "grad_norm": 2.191476821899414, "learning_rate": 1.0599922088040515e-05, "loss": 2.5254, "step": 907 }, { "epoch": 0.21220531097543163, "grad_norm": 2.3385491371154785, "learning_rate": 1.0611608881963382e-05, "loss": 2.5484, "step": 908 }, { "epoch": 0.21243901726505215, "grad_norm": 2.37386417388916, "learning_rate": 1.0623295675886249e-05, "loss": 2.3642, "step": 909 }, { "epoch": 0.21267272355467268, "grad_norm": 2.454802989959717, "learning_rate": 1.0634982469809116e-05, "loss": 2.305, "step": 910 }, { "epoch": 0.21290642984429317, "grad_norm": 1.99324369430542, "learning_rate": 1.0646669263731983e-05, "loss": 2.2961, "step": 911 }, { "epoch": 0.2131401361339137, "grad_norm": 2.135446786880493, "learning_rate": 1.065835605765485e-05, "loss": 2.3165, "step": 912 }, { "epoch": 0.21337384242353422, "grad_norm": 2.310868978500366, "learning_rate": 1.0670042851577717e-05, "loss": 2.3825, "step": 913 }, { "epoch": 0.21360754871315474, "grad_norm": 2.3979251384735107, "learning_rate": 1.0681729645500586e-05, "loss": 2.375, "step": 914 }, { "epoch": 0.21384125500277526, "grad_norm": 2.4759466648101807, "learning_rate": 1.0693416439423451e-05, "loss": 2.3222, "step": 915 }, { "epoch": 0.21407496129239578, "grad_norm": 2.231201648712158, "learning_rate": 1.0705103233346318e-05, "loss": 2.376, "step": 916 }, { "epoch": 0.2143086675820163, "grad_norm": 2.189512014389038, "learning_rate": 1.0716790027269187e-05, "loss": 2.4843, "step": 917 }, { "epoch": 0.21454237387163683, "grad_norm": 2.389437437057495, "learning_rate": 1.0728476821192052e-05, "loss": 2.2009, "step": 918 }, { "epoch": 0.21477608016125735, "grad_norm": 2.5855278968811035, "learning_rate": 1.074016361511492e-05, "loss": 2.5364, "step": 919 }, { "epoch": 0.21500978645087787, "grad_norm": 2.633242607116699, "learning_rate": 1.0751850409037788e-05, "loss": 2.4364, "step": 920 }, { "epoch": 0.21524349274049837, "grad_norm": 2.0650265216827393, "learning_rate": 1.0763537202960654e-05, "loss": 2.5198, "step": 921 }, { "epoch": 0.2154771990301189, "grad_norm": 2.4865546226501465, "learning_rate": 1.0775223996883522e-05, "loss": 2.3889, "step": 922 }, { "epoch": 0.2157109053197394, "grad_norm": 2.286130666732788, "learning_rate": 1.078691079080639e-05, "loss": 2.3327, "step": 923 }, { "epoch": 0.21594461160935993, "grad_norm": 2.2409474849700928, "learning_rate": 1.0798597584729256e-05, "loss": 2.4587, "step": 924 }, { "epoch": 0.21617831789898045, "grad_norm": 2.4017820358276367, "learning_rate": 1.0810284378652123e-05, "loss": 2.6565, "step": 925 }, { "epoch": 0.21641202418860098, "grad_norm": 3.219679117202759, "learning_rate": 1.082197117257499e-05, "loss": 2.4095, "step": 926 }, { "epoch": 0.2166457304782215, "grad_norm": 3.520413637161255, "learning_rate": 1.0833657966497858e-05, "loss": 2.4049, "step": 927 }, { "epoch": 0.21687943676784202, "grad_norm": 2.9558701515197754, "learning_rate": 1.0845344760420725e-05, "loss": 2.3605, "step": 928 }, { "epoch": 0.21711314305746254, "grad_norm": 2.8544580936431885, "learning_rate": 1.0857031554343592e-05, "loss": 2.2623, "step": 929 }, { "epoch": 0.21734684934708307, "grad_norm": 3.366443157196045, "learning_rate": 1.086871834826646e-05, "loss": 2.4642, "step": 930 }, { "epoch": 0.21758055563670356, "grad_norm": 3.5377705097198486, "learning_rate": 1.0880405142189326e-05, "loss": 2.5277, "step": 931 }, { "epoch": 0.21781426192632408, "grad_norm": 2.258030891418457, "learning_rate": 1.0892091936112193e-05, "loss": 2.3208, "step": 932 }, { "epoch": 0.2180479682159446, "grad_norm": 2.416288137435913, "learning_rate": 1.0903778730035061e-05, "loss": 2.2709, "step": 933 }, { "epoch": 0.21828167450556513, "grad_norm": 4.150886058807373, "learning_rate": 1.0915465523957927e-05, "loss": 2.2195, "step": 934 }, { "epoch": 0.21851538079518565, "grad_norm": 2.290069580078125, "learning_rate": 1.0927152317880794e-05, "loss": 2.3936, "step": 935 }, { "epoch": 0.21874908708480617, "grad_norm": 2.2298738956451416, "learning_rate": 1.0938839111803663e-05, "loss": 2.4466, "step": 936 }, { "epoch": 0.2189827933744267, "grad_norm": 2.2353670597076416, "learning_rate": 1.095052590572653e-05, "loss": 2.3882, "step": 937 }, { "epoch": 0.21921649966404722, "grad_norm": 2.9713330268859863, "learning_rate": 1.0962212699649397e-05, "loss": 2.3667, "step": 938 }, { "epoch": 0.21945020595366774, "grad_norm": 2.5492913722991943, "learning_rate": 1.0973899493572264e-05, "loss": 2.2812, "step": 939 }, { "epoch": 0.21968391224328826, "grad_norm": 2.2301905155181885, "learning_rate": 1.098558628749513e-05, "loss": 2.4223, "step": 940 }, { "epoch": 0.21991761853290875, "grad_norm": 2.4420523643493652, "learning_rate": 1.0997273081417998e-05, "loss": 2.5318, "step": 941 }, { "epoch": 0.22015132482252928, "grad_norm": 2.631399393081665, "learning_rate": 1.1008959875340865e-05, "loss": 2.4549, "step": 942 }, { "epoch": 0.2203850311121498, "grad_norm": 4.583476543426514, "learning_rate": 1.1020646669263732e-05, "loss": 2.2549, "step": 943 }, { "epoch": 0.22061873740177032, "grad_norm": 2.456970691680908, "learning_rate": 1.1032333463186599e-05, "loss": 2.3968, "step": 944 }, { "epoch": 0.22085244369139084, "grad_norm": 2.3664422035217285, "learning_rate": 1.1044020257109466e-05, "loss": 2.4736, "step": 945 }, { "epoch": 0.22108614998101137, "grad_norm": 2.315542697906494, "learning_rate": 1.1055707051032335e-05, "loss": 2.3155, "step": 946 }, { "epoch": 0.2213198562706319, "grad_norm": 2.318297863006592, "learning_rate": 1.10673938449552e-05, "loss": 2.4181, "step": 947 }, { "epoch": 0.2215535625602524, "grad_norm": 2.773958921432495, "learning_rate": 1.1079080638878067e-05, "loss": 2.3726, "step": 948 }, { "epoch": 0.22178726884987293, "grad_norm": 2.775818109512329, "learning_rate": 1.1090767432800936e-05, "loss": 2.5694, "step": 949 }, { "epoch": 0.22202097513949345, "grad_norm": 2.499495267868042, "learning_rate": 1.1102454226723803e-05, "loss": 2.4949, "step": 950 }, { "epoch": 0.22225468142911395, "grad_norm": 2.8190176486968994, "learning_rate": 1.111414102064667e-05, "loss": 2.3211, "step": 951 }, { "epoch": 0.22248838771873447, "grad_norm": 3.747666120529175, "learning_rate": 1.1125827814569537e-05, "loss": 2.3812, "step": 952 }, { "epoch": 0.222722094008355, "grad_norm": 2.895554780960083, "learning_rate": 1.1137514608492404e-05, "loss": 2.4033, "step": 953 }, { "epoch": 0.22295580029797551, "grad_norm": 2.799398899078369, "learning_rate": 1.1149201402415271e-05, "loss": 2.4304, "step": 954 }, { "epoch": 0.22318950658759604, "grad_norm": 2.321385622024536, "learning_rate": 1.1160888196338138e-05, "loss": 2.3661, "step": 955 }, { "epoch": 0.22342321287721656, "grad_norm": 2.9209110736846924, "learning_rate": 1.1172574990261005e-05, "loss": 2.4386, "step": 956 }, { "epoch": 0.22365691916683708, "grad_norm": 3.2398159503936768, "learning_rate": 1.1184261784183872e-05, "loss": 2.3389, "step": 957 }, { "epoch": 0.2238906254564576, "grad_norm": 2.4435837268829346, "learning_rate": 1.119594857810674e-05, "loss": 2.4502, "step": 958 }, { "epoch": 0.22412433174607813, "grad_norm": 2.2122721672058105, "learning_rate": 1.1207635372029608e-05, "loss": 2.4179, "step": 959 }, { "epoch": 0.22435803803569865, "grad_norm": 2.7928268909454346, "learning_rate": 1.1219322165952473e-05, "loss": 2.4066, "step": 960 }, { "epoch": 0.22459174432531914, "grad_norm": 2.4428555965423584, "learning_rate": 1.123100895987534e-05, "loss": 2.3688, "step": 961 }, { "epoch": 0.22482545061493966, "grad_norm": 2.209172248840332, "learning_rate": 1.124269575379821e-05, "loss": 2.5035, "step": 962 }, { "epoch": 0.2250591569045602, "grad_norm": 2.239774465560913, "learning_rate": 1.1254382547721076e-05, "loss": 2.5302, "step": 963 }, { "epoch": 0.2252928631941807, "grad_norm": 2.6220455169677734, "learning_rate": 1.1266069341643942e-05, "loss": 2.3823, "step": 964 }, { "epoch": 0.22552656948380123, "grad_norm": 2.5216429233551025, "learning_rate": 1.127775613556681e-05, "loss": 2.3861, "step": 965 }, { "epoch": 0.22576027577342175, "grad_norm": 3.005992889404297, "learning_rate": 1.1289442929489677e-05, "loss": 2.2751, "step": 966 }, { "epoch": 0.22599398206304228, "grad_norm": 2.2376365661621094, "learning_rate": 1.1301129723412544e-05, "loss": 2.3812, "step": 967 }, { "epoch": 0.2262276883526628, "grad_norm": 2.765627145767212, "learning_rate": 1.1312816517335412e-05, "loss": 2.4624, "step": 968 }, { "epoch": 0.22646139464228332, "grad_norm": 2.4542171955108643, "learning_rate": 1.1324503311258279e-05, "loss": 2.6393, "step": 969 }, { "epoch": 0.22669510093190384, "grad_norm": 2.293494701385498, "learning_rate": 1.1336190105181146e-05, "loss": 2.5037, "step": 970 }, { "epoch": 0.22692880722152434, "grad_norm": 3.1543734073638916, "learning_rate": 1.1347876899104013e-05, "loss": 2.38, "step": 971 }, { "epoch": 0.22716251351114486, "grad_norm": 2.2049899101257324, "learning_rate": 1.135956369302688e-05, "loss": 2.5463, "step": 972 }, { "epoch": 0.22739621980076538, "grad_norm": 2.5199482440948486, "learning_rate": 1.1371250486949747e-05, "loss": 2.4041, "step": 973 }, { "epoch": 0.2276299260903859, "grad_norm": 2.87957763671875, "learning_rate": 1.1382937280872614e-05, "loss": 2.2994, "step": 974 }, { "epoch": 0.22786363238000643, "grad_norm": 2.4906580448150635, "learning_rate": 1.1394624074795483e-05, "loss": 2.362, "step": 975 }, { "epoch": 0.22809733866962695, "grad_norm": 2.347156524658203, "learning_rate": 1.1406310868718348e-05, "loss": 2.1636, "step": 976 }, { "epoch": 0.22833104495924747, "grad_norm": 2.5170657634735107, "learning_rate": 1.1417997662641215e-05, "loss": 2.4225, "step": 977 }, { "epoch": 0.228564751248868, "grad_norm": 2.2298800945281982, "learning_rate": 1.1429684456564084e-05, "loss": 2.4054, "step": 978 }, { "epoch": 0.22879845753848851, "grad_norm": 2.3139283657073975, "learning_rate": 1.144137125048695e-05, "loss": 2.4659, "step": 979 }, { "epoch": 0.22903216382810904, "grad_norm": 2.6245310306549072, "learning_rate": 1.1453058044409816e-05, "loss": 2.3075, "step": 980 }, { "epoch": 0.22926587011772953, "grad_norm": 1.9937117099761963, "learning_rate": 1.1464744838332685e-05, "loss": 2.3512, "step": 981 }, { "epoch": 0.22949957640735005, "grad_norm": 2.4676902294158936, "learning_rate": 1.1476431632255552e-05, "loss": 2.3836, "step": 982 }, { "epoch": 0.22973328269697058, "grad_norm": 2.8158462047576904, "learning_rate": 1.1488118426178419e-05, "loss": 2.2058, "step": 983 }, { "epoch": 0.2299669889865911, "grad_norm": 2.490894079208374, "learning_rate": 1.1499805220101286e-05, "loss": 2.3269, "step": 984 }, { "epoch": 0.23020069527621162, "grad_norm": 2.9560024738311768, "learning_rate": 1.1511492014024153e-05, "loss": 2.302, "step": 985 }, { "epoch": 0.23043440156583214, "grad_norm": 3.318225622177124, "learning_rate": 1.152317880794702e-05, "loss": 2.343, "step": 986 }, { "epoch": 0.23066810785545266, "grad_norm": 2.54537034034729, "learning_rate": 1.1534865601869887e-05, "loss": 2.3075, "step": 987 }, { "epoch": 0.2309018141450732, "grad_norm": 2.6906590461730957, "learning_rate": 1.1546552395792754e-05, "loss": 2.4144, "step": 988 }, { "epoch": 0.2311355204346937, "grad_norm": 3.331969738006592, "learning_rate": 1.1558239189715621e-05, "loss": 2.3461, "step": 989 }, { "epoch": 0.23136922672431423, "grad_norm": 2.7223522663116455, "learning_rate": 1.1569925983638488e-05, "loss": 2.4982, "step": 990 }, { "epoch": 0.23160293301393473, "grad_norm": 2.7267305850982666, "learning_rate": 1.1581612777561357e-05, "loss": 2.3971, "step": 991 }, { "epoch": 0.23183663930355525, "grad_norm": 2.9850661754608154, "learning_rate": 1.1593299571484224e-05, "loss": 2.3908, "step": 992 }, { "epoch": 0.23207034559317577, "grad_norm": 3.9342994689941406, "learning_rate": 1.160498636540709e-05, "loss": 2.4754, "step": 993 }, { "epoch": 0.2323040518827963, "grad_norm": 2.6131153106689453, "learning_rate": 1.1616673159329958e-05, "loss": 2.4209, "step": 994 }, { "epoch": 0.23253775817241681, "grad_norm": 2.4457080364227295, "learning_rate": 1.1628359953252825e-05, "loss": 2.4984, "step": 995 }, { "epoch": 0.23277146446203734, "grad_norm": 2.4020872116088867, "learning_rate": 1.164004674717569e-05, "loss": 2.3036, "step": 996 }, { "epoch": 0.23300517075165786, "grad_norm": 2.3602139949798584, "learning_rate": 1.165173354109856e-05, "loss": 2.5158, "step": 997 }, { "epoch": 0.23323887704127838, "grad_norm": 2.4484939575195312, "learning_rate": 1.1663420335021426e-05, "loss": 2.2741, "step": 998 }, { "epoch": 0.2334725833308989, "grad_norm": 2.299462080001831, "learning_rate": 1.1675107128944293e-05, "loss": 2.3417, "step": 999 }, { "epoch": 0.23370628962051943, "grad_norm": 3.1917147636413574, "learning_rate": 1.168679392286716e-05, "loss": 2.3088, "step": 1000 }, { "epoch": 0.23393999591013992, "grad_norm": 2.2557573318481445, "learning_rate": 1.1698480716790027e-05, "loss": 2.3533, "step": 1001 }, { "epoch": 0.23417370219976044, "grad_norm": 2.2707135677337646, "learning_rate": 1.1710167510712894e-05, "loss": 2.325, "step": 1002 }, { "epoch": 0.23440740848938096, "grad_norm": 3.222472667694092, "learning_rate": 1.1721854304635762e-05, "loss": 2.6099, "step": 1003 }, { "epoch": 0.2346411147790015, "grad_norm": 2.772439956665039, "learning_rate": 1.1733541098558629e-05, "loss": 2.3262, "step": 1004 }, { "epoch": 0.234874821068622, "grad_norm": 3.4773387908935547, "learning_rate": 1.1745227892481497e-05, "loss": 2.5083, "step": 1005 }, { "epoch": 0.23510852735824253, "grad_norm": 2.7826836109161377, "learning_rate": 1.1756914686404363e-05, "loss": 2.2797, "step": 1006 }, { "epoch": 0.23534223364786305, "grad_norm": 2.511599063873291, "learning_rate": 1.1768601480327231e-05, "loss": 2.4744, "step": 1007 }, { "epoch": 0.23557593993748357, "grad_norm": 2.274261951446533, "learning_rate": 1.1780288274250098e-05, "loss": 2.2942, "step": 1008 }, { "epoch": 0.2358096462271041, "grad_norm": 2.1253907680511475, "learning_rate": 1.1791975068172964e-05, "loss": 2.2262, "step": 1009 }, { "epoch": 0.23604335251672462, "grad_norm": 2.415381908416748, "learning_rate": 1.1803661862095833e-05, "loss": 2.5024, "step": 1010 }, { "epoch": 0.2362770588063451, "grad_norm": 2.4027888774871826, "learning_rate": 1.18153486560187e-05, "loss": 2.4593, "step": 1011 }, { "epoch": 0.23651076509596564, "grad_norm": 2.4717912673950195, "learning_rate": 1.1827035449941565e-05, "loss": 2.3604, "step": 1012 }, { "epoch": 0.23674447138558616, "grad_norm": 2.3351078033447266, "learning_rate": 1.1838722243864434e-05, "loss": 2.4916, "step": 1013 }, { "epoch": 0.23697817767520668, "grad_norm": 3.536163330078125, "learning_rate": 1.18504090377873e-05, "loss": 2.3346, "step": 1014 }, { "epoch": 0.2372118839648272, "grad_norm": 2.15455961227417, "learning_rate": 1.1862095831710168e-05, "loss": 2.5009, "step": 1015 }, { "epoch": 0.23744559025444772, "grad_norm": 2.3651504516601562, "learning_rate": 1.1873782625633035e-05, "loss": 2.4076, "step": 1016 }, { "epoch": 0.23767929654406825, "grad_norm": 3.1010220050811768, "learning_rate": 1.1885469419555902e-05, "loss": 2.4188, "step": 1017 }, { "epoch": 0.23791300283368877, "grad_norm": 2.2707180976867676, "learning_rate": 1.189715621347877e-05, "loss": 2.2737, "step": 1018 }, { "epoch": 0.2381467091233093, "grad_norm": 2.144305944442749, "learning_rate": 1.1908843007401636e-05, "loss": 2.5573, "step": 1019 }, { "epoch": 0.2383804154129298, "grad_norm": 2.2264304161071777, "learning_rate": 1.1920529801324503e-05, "loss": 2.3107, "step": 1020 }, { "epoch": 0.2386141217025503, "grad_norm": 2.2882297039031982, "learning_rate": 1.1932216595247372e-05, "loss": 2.3602, "step": 1021 }, { "epoch": 0.23884782799217083, "grad_norm": 2.460169792175293, "learning_rate": 1.1943903389170237e-05, "loss": 2.2974, "step": 1022 }, { "epoch": 0.23908153428179135, "grad_norm": 2.442979574203491, "learning_rate": 1.1955590183093106e-05, "loss": 2.5174, "step": 1023 }, { "epoch": 0.23931524057141187, "grad_norm": 2.7722232341766357, "learning_rate": 1.1967276977015973e-05, "loss": 2.2983, "step": 1024 }, { "epoch": 0.2395489468610324, "grad_norm": 2.558912754058838, "learning_rate": 1.1978963770938838e-05, "loss": 2.24, "step": 1025 }, { "epoch": 0.23978265315065292, "grad_norm": 2.657616376876831, "learning_rate": 1.1990650564861707e-05, "loss": 2.2731, "step": 1026 }, { "epoch": 0.24001635944027344, "grad_norm": 4.2679057121276855, "learning_rate": 1.2002337358784574e-05, "loss": 2.3553, "step": 1027 }, { "epoch": 0.24025006572989396, "grad_norm": 3.063540458679199, "learning_rate": 1.201402415270744e-05, "loss": 2.3509, "step": 1028 }, { "epoch": 0.24048377201951449, "grad_norm": 2.549604892730713, "learning_rate": 1.2025710946630308e-05, "loss": 2.3471, "step": 1029 }, { "epoch": 0.240717478309135, "grad_norm": 3.3646812438964844, "learning_rate": 1.2037397740553175e-05, "loss": 2.3144, "step": 1030 }, { "epoch": 0.2409511845987555, "grad_norm": 2.770362138748169, "learning_rate": 1.2049084534476042e-05, "loss": 2.4312, "step": 1031 }, { "epoch": 0.24118489088837602, "grad_norm": 2.9667136669158936, "learning_rate": 1.206077132839891e-05, "loss": 2.4685, "step": 1032 }, { "epoch": 0.24141859717799655, "grad_norm": 2.576103448867798, "learning_rate": 1.2072458122321776e-05, "loss": 2.2614, "step": 1033 }, { "epoch": 0.24165230346761707, "grad_norm": 2.792973756790161, "learning_rate": 1.2084144916244645e-05, "loss": 2.5034, "step": 1034 }, { "epoch": 0.2418860097572376, "grad_norm": 2.623237133026123, "learning_rate": 1.209583171016751e-05, "loss": 2.3977, "step": 1035 }, { "epoch": 0.2421197160468581, "grad_norm": 2.7335329055786133, "learning_rate": 1.2107518504090377e-05, "loss": 2.3364, "step": 1036 }, { "epoch": 0.24235342233647864, "grad_norm": 2.8699657917022705, "learning_rate": 1.2119205298013246e-05, "loss": 2.2117, "step": 1037 }, { "epoch": 0.24258712862609916, "grad_norm": 3.6129045486450195, "learning_rate": 1.2130892091936112e-05, "loss": 2.197, "step": 1038 }, { "epoch": 0.24282083491571968, "grad_norm": 3.3150930404663086, "learning_rate": 1.214257888585898e-05, "loss": 2.3557, "step": 1039 }, { "epoch": 0.2430545412053402, "grad_norm": 2.1878955364227295, "learning_rate": 1.2154265679781847e-05, "loss": 2.3048, "step": 1040 }, { "epoch": 0.2432882474949607, "grad_norm": 3.2142045497894287, "learning_rate": 1.2165952473704713e-05, "loss": 2.343, "step": 1041 }, { "epoch": 0.24352195378458122, "grad_norm": 2.5891613960266113, "learning_rate": 1.2177639267627581e-05, "loss": 2.265, "step": 1042 }, { "epoch": 0.24375566007420174, "grad_norm": 2.5772645473480225, "learning_rate": 1.2189326061550448e-05, "loss": 2.4736, "step": 1043 }, { "epoch": 0.24398936636382226, "grad_norm": 3.135657787322998, "learning_rate": 1.2201012855473314e-05, "loss": 2.3502, "step": 1044 }, { "epoch": 0.24422307265344279, "grad_norm": 2.4191153049468994, "learning_rate": 1.2212699649396183e-05, "loss": 2.3932, "step": 1045 }, { "epoch": 0.2444567789430633, "grad_norm": 2.7587969303131104, "learning_rate": 1.222438644331905e-05, "loss": 2.487, "step": 1046 }, { "epoch": 0.24469048523268383, "grad_norm": 3.568873643875122, "learning_rate": 1.2236073237241918e-05, "loss": 2.2445, "step": 1047 }, { "epoch": 0.24492419152230435, "grad_norm": 2.9701695442199707, "learning_rate": 1.2247760031164784e-05, "loss": 2.2732, "step": 1048 }, { "epoch": 0.24515789781192487, "grad_norm": 2.3214805126190186, "learning_rate": 1.225944682508765e-05, "loss": 2.4979, "step": 1049 }, { "epoch": 0.2453916041015454, "grad_norm": 2.379728317260742, "learning_rate": 1.227113361901052e-05, "loss": 2.6615, "step": 1050 }, { "epoch": 0.2456253103911659, "grad_norm": 2.6113178730010986, "learning_rate": 1.2282820412933385e-05, "loss": 2.4831, "step": 1051 }, { "epoch": 0.2458590166807864, "grad_norm": 3.3151767253875732, "learning_rate": 1.2294507206856252e-05, "loss": 2.4627, "step": 1052 }, { "epoch": 0.24609272297040694, "grad_norm": 3.0145456790924072, "learning_rate": 1.230619400077912e-05, "loss": 2.3182, "step": 1053 }, { "epoch": 0.24632642926002746, "grad_norm": 2.3364696502685547, "learning_rate": 1.2317880794701986e-05, "loss": 2.2805, "step": 1054 }, { "epoch": 0.24656013554964798, "grad_norm": 3.4495511054992676, "learning_rate": 1.2329567588624855e-05, "loss": 2.6353, "step": 1055 }, { "epoch": 0.2467938418392685, "grad_norm": 3.2345783710479736, "learning_rate": 1.2341254382547722e-05, "loss": 2.2803, "step": 1056 }, { "epoch": 0.24702754812888902, "grad_norm": 2.699896812438965, "learning_rate": 1.2352941176470587e-05, "loss": 2.4024, "step": 1057 }, { "epoch": 0.24726125441850955, "grad_norm": 3.2410385608673096, "learning_rate": 1.2364627970393456e-05, "loss": 2.3307, "step": 1058 }, { "epoch": 0.24749496070813007, "grad_norm": 2.9061837196350098, "learning_rate": 1.2376314764316323e-05, "loss": 2.2392, "step": 1059 }, { "epoch": 0.2477286669977506, "grad_norm": 2.701958179473877, "learning_rate": 1.238800155823919e-05, "loss": 2.414, "step": 1060 }, { "epoch": 0.24796237328737109, "grad_norm": 2.254629135131836, "learning_rate": 1.2399688352162057e-05, "loss": 2.4151, "step": 1061 }, { "epoch": 0.2481960795769916, "grad_norm": 3.5998878479003906, "learning_rate": 1.2411375146084924e-05, "loss": 2.602, "step": 1062 }, { "epoch": 0.24842978586661213, "grad_norm": 3.3126816749572754, "learning_rate": 1.2423061940007793e-05, "loss": 2.5092, "step": 1063 }, { "epoch": 0.24866349215623265, "grad_norm": 2.969599485397339, "learning_rate": 1.2434748733930658e-05, "loss": 2.4018, "step": 1064 }, { "epoch": 0.24889719844585317, "grad_norm": 2.5923163890838623, "learning_rate": 1.2446435527853525e-05, "loss": 2.3253, "step": 1065 }, { "epoch": 0.2491309047354737, "grad_norm": 2.758068799972534, "learning_rate": 1.2458122321776394e-05, "loss": 2.3168, "step": 1066 }, { "epoch": 0.24936461102509422, "grad_norm": 2.6241321563720703, "learning_rate": 1.246980911569926e-05, "loss": 2.3481, "step": 1067 }, { "epoch": 0.24959831731471474, "grad_norm": 2.6636340618133545, "learning_rate": 1.2481495909622126e-05, "loss": 2.2997, "step": 1068 }, { "epoch": 0.24983202360433526, "grad_norm": 2.9115679264068604, "learning_rate": 1.2493182703544995e-05, "loss": 2.1998, "step": 1069 }, { "epoch": 0.25006572989395576, "grad_norm": 2.734442949295044, "learning_rate": 1.250486949746786e-05, "loss": 2.5224, "step": 1070 }, { "epoch": 0.2502994361835763, "grad_norm": 2.5577914714813232, "learning_rate": 1.251655629139073e-05, "loss": 2.3037, "step": 1071 }, { "epoch": 0.2505331424731968, "grad_norm": 2.502912759780884, "learning_rate": 1.2528243085313596e-05, "loss": 2.3604, "step": 1072 }, { "epoch": 0.25076684876281735, "grad_norm": 2.3205275535583496, "learning_rate": 1.2539929879236463e-05, "loss": 2.411, "step": 1073 }, { "epoch": 0.25100055505243785, "grad_norm": 2.400325059890747, "learning_rate": 1.255161667315933e-05, "loss": 2.3823, "step": 1074 }, { "epoch": 0.2512342613420584, "grad_norm": 2.3555526733398438, "learning_rate": 1.2563303467082197e-05, "loss": 2.4347, "step": 1075 }, { "epoch": 0.2514679676316789, "grad_norm": 2.1325018405914307, "learning_rate": 1.2574990261005064e-05, "loss": 2.3548, "step": 1076 }, { "epoch": 0.2517016739212994, "grad_norm": 2.587979555130005, "learning_rate": 1.2586677054927931e-05, "loss": 2.3448, "step": 1077 }, { "epoch": 0.25193538021091993, "grad_norm": 2.66182279586792, "learning_rate": 1.2598363848850799e-05, "loss": 2.3267, "step": 1078 }, { "epoch": 0.25216908650054043, "grad_norm": 3.7039248943328857, "learning_rate": 1.2610050642773667e-05, "loss": 2.2378, "step": 1079 }, { "epoch": 0.252402792790161, "grad_norm": 2.283888578414917, "learning_rate": 1.2621737436696533e-05, "loss": 2.3307, "step": 1080 }, { "epoch": 0.2526364990797815, "grad_norm": 2.2917752265930176, "learning_rate": 1.26334242306194e-05, "loss": 2.3354, "step": 1081 }, { "epoch": 0.252870205369402, "grad_norm": 2.6990723609924316, "learning_rate": 1.2645111024542268e-05, "loss": 2.4271, "step": 1082 }, { "epoch": 0.2531039116590225, "grad_norm": 2.6894547939300537, "learning_rate": 1.2656797818465134e-05, "loss": 2.256, "step": 1083 }, { "epoch": 0.25333761794864307, "grad_norm": 2.4899730682373047, "learning_rate": 1.2668484612388003e-05, "loss": 2.4301, "step": 1084 }, { "epoch": 0.25357132423826356, "grad_norm": 2.322681188583374, "learning_rate": 1.268017140631087e-05, "loss": 2.2958, "step": 1085 }, { "epoch": 0.25380503052788406, "grad_norm": 2.9679627418518066, "learning_rate": 1.2691858200233737e-05, "loss": 2.4163, "step": 1086 }, { "epoch": 0.2540387368175046, "grad_norm": 2.5355918407440186, "learning_rate": 1.2703544994156604e-05, "loss": 2.4457, "step": 1087 }, { "epoch": 0.2542724431071251, "grad_norm": 3.2842347621917725, "learning_rate": 1.271523178807947e-05, "loss": 2.46, "step": 1088 }, { "epoch": 0.25450614939674565, "grad_norm": 2.5713789463043213, "learning_rate": 1.2726918582002338e-05, "loss": 2.3702, "step": 1089 }, { "epoch": 0.25473985568636615, "grad_norm": 2.645819664001465, "learning_rate": 1.2738605375925205e-05, "loss": 2.2254, "step": 1090 }, { "epoch": 0.2549735619759867, "grad_norm": 2.921215295791626, "learning_rate": 1.2750292169848072e-05, "loss": 2.2034, "step": 1091 }, { "epoch": 0.2552072682656072, "grad_norm": 2.32975435256958, "learning_rate": 1.276197896377094e-05, "loss": 2.4964, "step": 1092 }, { "epoch": 0.25544097455522774, "grad_norm": 2.897944688796997, "learning_rate": 1.2773665757693806e-05, "loss": 2.5045, "step": 1093 }, { "epoch": 0.25567468084484823, "grad_norm": 2.9335973262786865, "learning_rate": 1.2785352551616673e-05, "loss": 2.3335, "step": 1094 }, { "epoch": 0.25590838713446873, "grad_norm": 2.617985486984253, "learning_rate": 1.2797039345539542e-05, "loss": 2.2704, "step": 1095 }, { "epoch": 0.2561420934240893, "grad_norm": 3.7054882049560547, "learning_rate": 1.2808726139462407e-05, "loss": 2.3107, "step": 1096 }, { "epoch": 0.2563757997137098, "grad_norm": 2.959228754043579, "learning_rate": 1.2820412933385274e-05, "loss": 2.2167, "step": 1097 }, { "epoch": 0.2566095060033303, "grad_norm": 2.8711671829223633, "learning_rate": 1.2832099727308143e-05, "loss": 2.459, "step": 1098 }, { "epoch": 0.2568432122929508, "grad_norm": 5.645910263061523, "learning_rate": 1.2843786521231008e-05, "loss": 2.2504, "step": 1099 }, { "epoch": 0.25707691858257137, "grad_norm": 3.6625497341156006, "learning_rate": 1.2855473315153877e-05, "loss": 2.2684, "step": 1100 }, { "epoch": 0.25731062487219186, "grad_norm": 3.60335111618042, "learning_rate": 1.2867160109076744e-05, "loss": 2.4307, "step": 1101 }, { "epoch": 0.2575443311618124, "grad_norm": 2.161379814147949, "learning_rate": 1.2878846902999611e-05, "loss": 2.5125, "step": 1102 }, { "epoch": 0.2577780374514329, "grad_norm": 4.7790045738220215, "learning_rate": 1.2890533696922478e-05, "loss": 2.4792, "step": 1103 }, { "epoch": 0.25801174374105346, "grad_norm": 3.47955584526062, "learning_rate": 1.2902220490845345e-05, "loss": 2.3344, "step": 1104 }, { "epoch": 0.25824545003067395, "grad_norm": 2.201047897338867, "learning_rate": 1.2913907284768212e-05, "loss": 2.323, "step": 1105 }, { "epoch": 0.25847915632029445, "grad_norm": 2.3132381439208984, "learning_rate": 1.292559407869108e-05, "loss": 2.4286, "step": 1106 }, { "epoch": 0.258712862609915, "grad_norm": 2.9580397605895996, "learning_rate": 1.2937280872613946e-05, "loss": 2.2879, "step": 1107 }, { "epoch": 0.2589465688995355, "grad_norm": 2.4791839122772217, "learning_rate": 1.2948967666536815e-05, "loss": 2.4036, "step": 1108 }, { "epoch": 0.25918027518915604, "grad_norm": 2.444129467010498, "learning_rate": 1.296065446045968e-05, "loss": 2.3369, "step": 1109 }, { "epoch": 0.25941398147877653, "grad_norm": 2.3884902000427246, "learning_rate": 1.2972341254382547e-05, "loss": 2.3679, "step": 1110 }, { "epoch": 0.2596476877683971, "grad_norm": 2.830084800720215, "learning_rate": 1.2984028048305416e-05, "loss": 2.2383, "step": 1111 }, { "epoch": 0.2598813940580176, "grad_norm": 2.264315366744995, "learning_rate": 1.2995714842228282e-05, "loss": 2.2925, "step": 1112 }, { "epoch": 0.26011510034763813, "grad_norm": 2.325756311416626, "learning_rate": 1.3007401636151149e-05, "loss": 2.2533, "step": 1113 }, { "epoch": 0.2603488066372586, "grad_norm": 2.760469675064087, "learning_rate": 1.3019088430074017e-05, "loss": 2.2411, "step": 1114 }, { "epoch": 0.2605825129268791, "grad_norm": 2.2485179901123047, "learning_rate": 1.3030775223996884e-05, "loss": 2.3387, "step": 1115 }, { "epoch": 0.26081621921649967, "grad_norm": 3.4566638469696045, "learning_rate": 1.3042462017919751e-05, "loss": 2.2993, "step": 1116 }, { "epoch": 0.26104992550612016, "grad_norm": 2.7335879802703857, "learning_rate": 1.3054148811842618e-05, "loss": 2.1695, "step": 1117 }, { "epoch": 0.2612836317957407, "grad_norm": 2.281440019607544, "learning_rate": 1.3065835605765485e-05, "loss": 2.3155, "step": 1118 }, { "epoch": 0.2615173380853612, "grad_norm": 2.4550509452819824, "learning_rate": 1.3077522399688353e-05, "loss": 2.3911, "step": 1119 }, { "epoch": 0.26175104437498176, "grad_norm": 3.959573745727539, "learning_rate": 1.308920919361122e-05, "loss": 2.2531, "step": 1120 }, { "epoch": 0.26198475066460225, "grad_norm": 3.052020788192749, "learning_rate": 1.3100895987534087e-05, "loss": 2.5305, "step": 1121 }, { "epoch": 0.2622184569542228, "grad_norm": 2.25022029876709, "learning_rate": 1.3112582781456954e-05, "loss": 2.5581, "step": 1122 }, { "epoch": 0.2624521632438433, "grad_norm": 3.0224735736846924, "learning_rate": 1.312426957537982e-05, "loss": 2.3518, "step": 1123 }, { "epoch": 0.26268586953346384, "grad_norm": 3.83501935005188, "learning_rate": 1.313595636930269e-05, "loss": 2.3595, "step": 1124 }, { "epoch": 0.26291957582308434, "grad_norm": 3.702861785888672, "learning_rate": 1.3147643163225555e-05, "loss": 2.5093, "step": 1125 }, { "epoch": 0.26315328211270483, "grad_norm": 2.5671448707580566, "learning_rate": 1.3159329957148422e-05, "loss": 2.2588, "step": 1126 }, { "epoch": 0.2633869884023254, "grad_norm": 2.2282581329345703, "learning_rate": 1.317101675107129e-05, "loss": 2.5036, "step": 1127 }, { "epoch": 0.2636206946919459, "grad_norm": 2.8610637187957764, "learning_rate": 1.3182703544994158e-05, "loss": 2.4192, "step": 1128 }, { "epoch": 0.26385440098156643, "grad_norm": 2.4043500423431396, "learning_rate": 1.3194390338917023e-05, "loss": 2.4451, "step": 1129 }, { "epoch": 0.2640881072711869, "grad_norm": 2.265580892562866, "learning_rate": 1.3206077132839892e-05, "loss": 2.3583, "step": 1130 }, { "epoch": 0.2643218135608075, "grad_norm": 2.1251566410064697, "learning_rate": 1.3217763926762759e-05, "loss": 2.2604, "step": 1131 }, { "epoch": 0.26455551985042797, "grad_norm": 2.706468343734741, "learning_rate": 1.3229450720685626e-05, "loss": 2.2302, "step": 1132 }, { "epoch": 0.2647892261400485, "grad_norm": 2.606767177581787, "learning_rate": 1.3241137514608493e-05, "loss": 2.2761, "step": 1133 }, { "epoch": 0.265022932429669, "grad_norm": 3.1144211292266846, "learning_rate": 1.325282430853136e-05, "loss": 2.5352, "step": 1134 }, { "epoch": 0.2652566387192895, "grad_norm": 3.6481096744537354, "learning_rate": 1.3264511102454227e-05, "loss": 2.1359, "step": 1135 }, { "epoch": 0.26549034500891006, "grad_norm": 3.9983291625976562, "learning_rate": 1.3276197896377094e-05, "loss": 2.4021, "step": 1136 }, { "epoch": 0.26572405129853055, "grad_norm": 2.423210620880127, "learning_rate": 1.3287884690299961e-05, "loss": 2.391, "step": 1137 }, { "epoch": 0.2659577575881511, "grad_norm": 2.4460370540618896, "learning_rate": 1.3299571484222828e-05, "loss": 2.37, "step": 1138 }, { "epoch": 0.2661914638777716, "grad_norm": 3.606884241104126, "learning_rate": 1.3311258278145695e-05, "loss": 2.2452, "step": 1139 }, { "epoch": 0.26642517016739214, "grad_norm": 2.924855947494507, "learning_rate": 1.3322945072068564e-05, "loss": 2.2759, "step": 1140 }, { "epoch": 0.26665887645701264, "grad_norm": 2.6858725547790527, "learning_rate": 1.3334631865991431e-05, "loss": 2.4158, "step": 1141 }, { "epoch": 0.2668925827466332, "grad_norm": 2.5846073627471924, "learning_rate": 1.3346318659914296e-05, "loss": 2.2861, "step": 1142 }, { "epoch": 0.2671262890362537, "grad_norm": 2.5380899906158447, "learning_rate": 1.3358005453837165e-05, "loss": 2.1815, "step": 1143 }, { "epoch": 0.26735999532587423, "grad_norm": 2.528204917907715, "learning_rate": 1.3369692247760032e-05, "loss": 2.3903, "step": 1144 }, { "epoch": 0.26759370161549473, "grad_norm": 3.0689473152160645, "learning_rate": 1.3381379041682897e-05, "loss": 2.2032, "step": 1145 }, { "epoch": 0.2678274079051152, "grad_norm": 2.3379719257354736, "learning_rate": 1.3393065835605766e-05, "loss": 2.4251, "step": 1146 }, { "epoch": 0.2680611141947358, "grad_norm": 2.298018455505371, "learning_rate": 1.3404752629528633e-05, "loss": 2.3677, "step": 1147 }, { "epoch": 0.26829482048435627, "grad_norm": 2.59718656539917, "learning_rate": 1.34164394234515e-05, "loss": 2.2019, "step": 1148 }, { "epoch": 0.2685285267739768, "grad_norm": 2.4605915546417236, "learning_rate": 1.3428126217374367e-05, "loss": 2.444, "step": 1149 }, { "epoch": 0.2687622330635973, "grad_norm": 2.315429210662842, "learning_rate": 1.3439813011297234e-05, "loss": 2.4013, "step": 1150 }, { "epoch": 0.26899593935321786, "grad_norm": 2.5005857944488525, "learning_rate": 1.3451499805220101e-05, "loss": 2.2824, "step": 1151 }, { "epoch": 0.26922964564283836, "grad_norm": 3.183335304260254, "learning_rate": 1.3463186599142968e-05, "loss": 2.3027, "step": 1152 }, { "epoch": 0.2694633519324589, "grad_norm": 3.6563284397125244, "learning_rate": 1.3474873393065836e-05, "loss": 2.1671, "step": 1153 }, { "epoch": 0.2696970582220794, "grad_norm": 4.051114559173584, "learning_rate": 1.3486560186988704e-05, "loss": 2.3629, "step": 1154 }, { "epoch": 0.2699307645116999, "grad_norm": 3.3076088428497314, "learning_rate": 1.349824698091157e-05, "loss": 2.3611, "step": 1155 }, { "epoch": 0.27016447080132044, "grad_norm": 3.7667908668518066, "learning_rate": 1.3509933774834438e-05, "loss": 2.3356, "step": 1156 }, { "epoch": 0.27039817709094094, "grad_norm": 4.8162007331848145, "learning_rate": 1.3521620568757305e-05, "loss": 2.3793, "step": 1157 }, { "epoch": 0.2706318833805615, "grad_norm": 2.8484385013580322, "learning_rate": 1.353330736268017e-05, "loss": 2.3778, "step": 1158 }, { "epoch": 0.270865589670182, "grad_norm": 2.6301393508911133, "learning_rate": 1.354499415660304e-05, "loss": 2.3971, "step": 1159 }, { "epoch": 0.27109929595980253, "grad_norm": 3.547933578491211, "learning_rate": 1.3556680950525907e-05, "loss": 2.4482, "step": 1160 }, { "epoch": 0.271333002249423, "grad_norm": 2.8652589321136475, "learning_rate": 1.3568367744448772e-05, "loss": 2.3093, "step": 1161 }, { "epoch": 0.2715667085390436, "grad_norm": 2.288252353668213, "learning_rate": 1.358005453837164e-05, "loss": 2.4304, "step": 1162 }, { "epoch": 0.27180041482866407, "grad_norm": 2.411295175552368, "learning_rate": 1.3591741332294508e-05, "loss": 2.2973, "step": 1163 }, { "epoch": 0.2720341211182846, "grad_norm": 2.296420097351074, "learning_rate": 1.3603428126217375e-05, "loss": 2.4653, "step": 1164 }, { "epoch": 0.2722678274079051, "grad_norm": 2.283541679382324, "learning_rate": 1.3615114920140242e-05, "loss": 2.4295, "step": 1165 }, { "epoch": 0.2725015336975256, "grad_norm": 2.1518547534942627, "learning_rate": 1.3626801714063109e-05, "loss": 2.2395, "step": 1166 }, { "epoch": 0.27273523998714616, "grad_norm": 2.3232059478759766, "learning_rate": 1.3638488507985976e-05, "loss": 2.459, "step": 1167 }, { "epoch": 0.27296894627676666, "grad_norm": 2.65576434135437, "learning_rate": 1.3650175301908843e-05, "loss": 2.2844, "step": 1168 }, { "epoch": 0.2732026525663872, "grad_norm": 2.2270166873931885, "learning_rate": 1.366186209583171e-05, "loss": 2.314, "step": 1169 }, { "epoch": 0.2734363588560077, "grad_norm": 2.5952019691467285, "learning_rate": 1.3673548889754579e-05, "loss": 2.3054, "step": 1170 }, { "epoch": 0.27367006514562825, "grad_norm": 2.406843662261963, "learning_rate": 1.3685235683677444e-05, "loss": 2.3307, "step": 1171 }, { "epoch": 0.27390377143524874, "grad_norm": 3.5244479179382324, "learning_rate": 1.3696922477600313e-05, "loss": 2.2891, "step": 1172 }, { "epoch": 0.2741374777248693, "grad_norm": 2.6896603107452393, "learning_rate": 1.370860927152318e-05, "loss": 2.3675, "step": 1173 }, { "epoch": 0.2743711840144898, "grad_norm": 2.4778850078582764, "learning_rate": 1.3720296065446045e-05, "loss": 2.3757, "step": 1174 }, { "epoch": 0.2746048903041103, "grad_norm": 14.582777976989746, "learning_rate": 1.3731982859368914e-05, "loss": 2.5804, "step": 1175 }, { "epoch": 0.27483859659373083, "grad_norm": 3.6810755729675293, "learning_rate": 1.3743669653291781e-05, "loss": 2.2452, "step": 1176 }, { "epoch": 0.2750723028833513, "grad_norm": 2.9957010746002197, "learning_rate": 1.3755356447214646e-05, "loss": 2.2564, "step": 1177 }, { "epoch": 0.2753060091729719, "grad_norm": 2.5358471870422363, "learning_rate": 1.3767043241137515e-05, "loss": 2.486, "step": 1178 }, { "epoch": 0.27553971546259237, "grad_norm": 3.0048375129699707, "learning_rate": 1.3778730035060382e-05, "loss": 2.3154, "step": 1179 }, { "epoch": 0.2757734217522129, "grad_norm": 4.167477130889893, "learning_rate": 1.379041682898325e-05, "loss": 2.1783, "step": 1180 }, { "epoch": 0.2760071280418334, "grad_norm": 2.764702320098877, "learning_rate": 1.3802103622906116e-05, "loss": 2.2932, "step": 1181 }, { "epoch": 0.27624083433145397, "grad_norm": 2.735867738723755, "learning_rate": 1.3813790416828983e-05, "loss": 2.5798, "step": 1182 }, { "epoch": 0.27647454062107446, "grad_norm": 3.386700391769409, "learning_rate": 1.3825477210751852e-05, "loss": 2.2791, "step": 1183 }, { "epoch": 0.276708246910695, "grad_norm": 2.310915946960449, "learning_rate": 1.3837164004674717e-05, "loss": 2.2781, "step": 1184 }, { "epoch": 0.2769419532003155, "grad_norm": 2.212953805923462, "learning_rate": 1.3848850798597584e-05, "loss": 2.4242, "step": 1185 }, { "epoch": 0.277175659489936, "grad_norm": 2.3630452156066895, "learning_rate": 1.3860537592520453e-05, "loss": 2.3181, "step": 1186 }, { "epoch": 0.27740936577955655, "grad_norm": 2.3748819828033447, "learning_rate": 1.3872224386443318e-05, "loss": 2.4215, "step": 1187 }, { "epoch": 0.27764307206917704, "grad_norm": 2.538005828857422, "learning_rate": 1.3883911180366187e-05, "loss": 2.3858, "step": 1188 }, { "epoch": 0.2778767783587976, "grad_norm": 2.6854090690612793, "learning_rate": 1.3895597974289054e-05, "loss": 2.199, "step": 1189 }, { "epoch": 0.2781104846484181, "grad_norm": 2.8582849502563477, "learning_rate": 1.390728476821192e-05, "loss": 2.2804, "step": 1190 }, { "epoch": 0.27834419093803864, "grad_norm": 4.258210182189941, "learning_rate": 1.3918971562134788e-05, "loss": 2.3128, "step": 1191 }, { "epoch": 0.27857789722765913, "grad_norm": 3.393566131591797, "learning_rate": 1.3930658356057655e-05, "loss": 2.2371, "step": 1192 }, { "epoch": 0.2788116035172797, "grad_norm": 3.1220881938934326, "learning_rate": 1.394234514998052e-05, "loss": 2.3789, "step": 1193 }, { "epoch": 0.2790453098069002, "grad_norm": 2.502462148666382, "learning_rate": 1.395403194390339e-05, "loss": 2.3521, "step": 1194 }, { "epoch": 0.27927901609652067, "grad_norm": 3.968447685241699, "learning_rate": 1.3965718737826257e-05, "loss": 2.4604, "step": 1195 }, { "epoch": 0.2795127223861412, "grad_norm": 4.565341949462891, "learning_rate": 1.3977405531749125e-05, "loss": 2.3844, "step": 1196 }, { "epoch": 0.2797464286757617, "grad_norm": 4.680972576141357, "learning_rate": 1.398909232567199e-05, "loss": 2.287, "step": 1197 }, { "epoch": 0.27998013496538227, "grad_norm": 2.353515625, "learning_rate": 1.4000779119594858e-05, "loss": 2.2567, "step": 1198 }, { "epoch": 0.28021384125500276, "grad_norm": 4.213557720184326, "learning_rate": 1.4012465913517726e-05, "loss": 2.2509, "step": 1199 }, { "epoch": 0.2804475475446233, "grad_norm": 4.503024101257324, "learning_rate": 1.4024152707440592e-05, "loss": 2.304, "step": 1200 }, { "epoch": 0.2806812538342438, "grad_norm": 4.896185874938965, "learning_rate": 1.4035839501363459e-05, "loss": 2.3222, "step": 1201 }, { "epoch": 0.28091496012386435, "grad_norm": 2.501950263977051, "learning_rate": 1.4047526295286328e-05, "loss": 2.3102, "step": 1202 }, { "epoch": 0.28114866641348485, "grad_norm": 2.6822357177734375, "learning_rate": 1.4059213089209193e-05, "loss": 2.4221, "step": 1203 }, { "epoch": 0.2813823727031054, "grad_norm": 3.0251264572143555, "learning_rate": 1.4070899883132062e-05, "loss": 2.2625, "step": 1204 }, { "epoch": 0.2816160789927259, "grad_norm": 3.8693366050720215, "learning_rate": 1.4082586677054929e-05, "loss": 2.3369, "step": 1205 }, { "epoch": 0.2818497852823464, "grad_norm": 2.536729574203491, "learning_rate": 1.4094273470977794e-05, "loss": 2.065, "step": 1206 }, { "epoch": 0.28208349157196694, "grad_norm": 5.450313568115234, "learning_rate": 1.4105960264900663e-05, "loss": 2.3392, "step": 1207 }, { "epoch": 0.28231719786158743, "grad_norm": 4.566758155822754, "learning_rate": 1.411764705882353e-05, "loss": 2.1727, "step": 1208 }, { "epoch": 0.282550904151208, "grad_norm": 2.4333479404449463, "learning_rate": 1.4129333852746397e-05, "loss": 2.3949, "step": 1209 }, { "epoch": 0.2827846104408285, "grad_norm": 2.6057021617889404, "learning_rate": 1.4141020646669264e-05, "loss": 2.2975, "step": 1210 }, { "epoch": 0.283018316730449, "grad_norm": 2.5320310592651367, "learning_rate": 1.4152707440592131e-05, "loss": 2.4748, "step": 1211 }, { "epoch": 0.2832520230200695, "grad_norm": 2.256586790084839, "learning_rate": 1.4164394234515e-05, "loss": 2.3273, "step": 1212 }, { "epoch": 0.28348572930969007, "grad_norm": 3.026047706604004, "learning_rate": 1.4176081028437865e-05, "loss": 2.3844, "step": 1213 }, { "epoch": 0.28371943559931057, "grad_norm": 2.271399974822998, "learning_rate": 1.4187767822360732e-05, "loss": 2.093, "step": 1214 }, { "epoch": 0.28395314188893106, "grad_norm": 3.237295389175415, "learning_rate": 1.4199454616283601e-05, "loss": 2.2627, "step": 1215 }, { "epoch": 0.2841868481785516, "grad_norm": 3.046438694000244, "learning_rate": 1.4211141410206466e-05, "loss": 2.3699, "step": 1216 }, { "epoch": 0.2844205544681721, "grad_norm": 2.5931637287139893, "learning_rate": 1.4222828204129335e-05, "loss": 2.1176, "step": 1217 }, { "epoch": 0.28465426075779265, "grad_norm": 3.275080680847168, "learning_rate": 1.4234514998052202e-05, "loss": 2.1823, "step": 1218 }, { "epoch": 0.28488796704741315, "grad_norm": 2.4449892044067383, "learning_rate": 1.4246201791975067e-05, "loss": 2.3326, "step": 1219 }, { "epoch": 0.2851216733370337, "grad_norm": 2.3482933044433594, "learning_rate": 1.4257888585897936e-05, "loss": 2.3607, "step": 1220 }, { "epoch": 0.2853553796266542, "grad_norm": 2.357088088989258, "learning_rate": 1.4269575379820803e-05, "loss": 2.3636, "step": 1221 }, { "epoch": 0.28558908591627474, "grad_norm": 3.3943634033203125, "learning_rate": 1.428126217374367e-05, "loss": 2.193, "step": 1222 }, { "epoch": 0.28582279220589524, "grad_norm": 2.5331332683563232, "learning_rate": 1.4292948967666537e-05, "loss": 2.3817, "step": 1223 }, { "epoch": 0.2860564984955158, "grad_norm": 3.4319956302642822, "learning_rate": 1.4304635761589404e-05, "loss": 2.2721, "step": 1224 }, { "epoch": 0.2862902047851363, "grad_norm": 2.813153028488159, "learning_rate": 1.4316322555512273e-05, "loss": 2.4171, "step": 1225 }, { "epoch": 0.2865239110747568, "grad_norm": 2.597738742828369, "learning_rate": 1.4328009349435138e-05, "loss": 2.243, "step": 1226 }, { "epoch": 0.2867576173643773, "grad_norm": 3.1202232837677, "learning_rate": 1.4339696143358005e-05, "loss": 2.3312, "step": 1227 }, { "epoch": 0.2869913236539978, "grad_norm": 2.9898011684417725, "learning_rate": 1.4351382937280874e-05, "loss": 2.2769, "step": 1228 }, { "epoch": 0.28722502994361837, "grad_norm": 2.6181771755218506, "learning_rate": 1.436306973120374e-05, "loss": 2.3091, "step": 1229 }, { "epoch": 0.28745873623323887, "grad_norm": 3.322254180908203, "learning_rate": 1.4374756525126607e-05, "loss": 2.4095, "step": 1230 }, { "epoch": 0.2876924425228594, "grad_norm": 3.223782539367676, "learning_rate": 1.4386443319049475e-05, "loss": 2.292, "step": 1231 }, { "epoch": 0.2879261488124799, "grad_norm": 3.030848979949951, "learning_rate": 1.439813011297234e-05, "loss": 2.2871, "step": 1232 }, { "epoch": 0.28815985510210046, "grad_norm": 2.4271106719970703, "learning_rate": 1.440981690689521e-05, "loss": 2.3004, "step": 1233 }, { "epoch": 0.28839356139172095, "grad_norm": 4.9435272216796875, "learning_rate": 1.4421503700818076e-05, "loss": 2.3416, "step": 1234 }, { "epoch": 0.28862726768134145, "grad_norm": 4.04534912109375, "learning_rate": 1.4433190494740942e-05, "loss": 2.379, "step": 1235 }, { "epoch": 0.288860973970962, "grad_norm": 2.4628379344940186, "learning_rate": 1.444487728866381e-05, "loss": 2.3473, "step": 1236 }, { "epoch": 0.2890946802605825, "grad_norm": 4.0370402336120605, "learning_rate": 1.4456564082586678e-05, "loss": 2.2728, "step": 1237 }, { "epoch": 0.28932838655020304, "grad_norm": 3.2664802074432373, "learning_rate": 1.4468250876509545e-05, "loss": 2.2373, "step": 1238 }, { "epoch": 0.28956209283982354, "grad_norm": 3.2527244091033936, "learning_rate": 1.4479937670432412e-05, "loss": 2.3408, "step": 1239 }, { "epoch": 0.2897957991294441, "grad_norm": 2.9664454460144043, "learning_rate": 1.4491624464355279e-05, "loss": 2.4564, "step": 1240 }, { "epoch": 0.2900295054190646, "grad_norm": 3.2065589427948, "learning_rate": 1.4503311258278147e-05, "loss": 2.3852, "step": 1241 }, { "epoch": 0.29026321170868513, "grad_norm": 3.4139459133148193, "learning_rate": 1.4514998052201013e-05, "loss": 2.2149, "step": 1242 }, { "epoch": 0.2904969179983056, "grad_norm": 2.2355122566223145, "learning_rate": 1.452668484612388e-05, "loss": 2.4131, "step": 1243 }, { "epoch": 0.2907306242879262, "grad_norm": 3.9821693897247314, "learning_rate": 1.4538371640046749e-05, "loss": 2.3362, "step": 1244 }, { "epoch": 0.29096433057754667, "grad_norm": 2.433687210083008, "learning_rate": 1.4550058433969614e-05, "loss": 2.3008, "step": 1245 }, { "epoch": 0.29119803686716716, "grad_norm": 2.4938225746154785, "learning_rate": 1.4561745227892481e-05, "loss": 2.4573, "step": 1246 }, { "epoch": 0.2914317431567877, "grad_norm": 2.6029930114746094, "learning_rate": 1.457343202181535e-05, "loss": 2.3517, "step": 1247 }, { "epoch": 0.2916654494464082, "grad_norm": 2.5971479415893555, "learning_rate": 1.4585118815738215e-05, "loss": 2.223, "step": 1248 }, { "epoch": 0.29189915573602876, "grad_norm": 3.364023208618164, "learning_rate": 1.4596805609661084e-05, "loss": 2.3825, "step": 1249 }, { "epoch": 0.29213286202564925, "grad_norm": 2.919590950012207, "learning_rate": 1.4608492403583951e-05, "loss": 2.323, "step": 1250 }, { "epoch": 0.2923665683152698, "grad_norm": 2.4508447647094727, "learning_rate": 1.4620179197506818e-05, "loss": 2.2485, "step": 1251 }, { "epoch": 0.2926002746048903, "grad_norm": 3.506882429122925, "learning_rate": 1.4631865991429685e-05, "loss": 2.4583, "step": 1252 }, { "epoch": 0.29283398089451085, "grad_norm": 3.6957476139068604, "learning_rate": 1.4643552785352552e-05, "loss": 2.2577, "step": 1253 }, { "epoch": 0.29306768718413134, "grad_norm": 2.526686429977417, "learning_rate": 1.4655239579275419e-05, "loss": 2.2042, "step": 1254 }, { "epoch": 0.29330139347375184, "grad_norm": 3.6228315830230713, "learning_rate": 1.4666926373198286e-05, "loss": 2.2482, "step": 1255 }, { "epoch": 0.2935350997633724, "grad_norm": 3.9001052379608154, "learning_rate": 1.4678613167121153e-05, "loss": 2.3691, "step": 1256 }, { "epoch": 0.2937688060529929, "grad_norm": 3.0730478763580322, "learning_rate": 1.4690299961044022e-05, "loss": 2.2415, "step": 1257 }, { "epoch": 0.29400251234261343, "grad_norm": 2.976280927658081, "learning_rate": 1.4701986754966887e-05, "loss": 2.5838, "step": 1258 }, { "epoch": 0.2942362186322339, "grad_norm": 2.9305214881896973, "learning_rate": 1.4713673548889754e-05, "loss": 2.2738, "step": 1259 }, { "epoch": 0.2944699249218545, "grad_norm": 3.508312225341797, "learning_rate": 1.4725360342812623e-05, "loss": 2.3008, "step": 1260 }, { "epoch": 0.29470363121147497, "grad_norm": 2.517322063446045, "learning_rate": 1.4737047136735488e-05, "loss": 2.3191, "step": 1261 }, { "epoch": 0.2949373375010955, "grad_norm": 2.5755300521850586, "learning_rate": 1.4748733930658355e-05, "loss": 2.3955, "step": 1262 }, { "epoch": 0.295171043790716, "grad_norm": 2.372783899307251, "learning_rate": 1.4760420724581224e-05, "loss": 2.2736, "step": 1263 }, { "epoch": 0.29540475008033656, "grad_norm": 2.5983386039733887, "learning_rate": 1.4772107518504091e-05, "loss": 2.3413, "step": 1264 }, { "epoch": 0.29563845636995706, "grad_norm": 2.520341157913208, "learning_rate": 1.4783794312426958e-05, "loss": 2.3342, "step": 1265 }, { "epoch": 0.29587216265957755, "grad_norm": 2.390874147415161, "learning_rate": 1.4795481106349825e-05, "loss": 2.3754, "step": 1266 }, { "epoch": 0.2961058689491981, "grad_norm": 3.217174530029297, "learning_rate": 1.4807167900272692e-05, "loss": 2.3025, "step": 1267 }, { "epoch": 0.2963395752388186, "grad_norm": 3.6078310012817383, "learning_rate": 1.481885469419556e-05, "loss": 2.2951, "step": 1268 }, { "epoch": 0.29657328152843915, "grad_norm": 2.568143606185913, "learning_rate": 1.4830541488118427e-05, "loss": 2.2874, "step": 1269 }, { "epoch": 0.29680698781805964, "grad_norm": 2.4396026134490967, "learning_rate": 1.4842228282041294e-05, "loss": 2.2308, "step": 1270 }, { "epoch": 0.2970406941076802, "grad_norm": 2.5172600746154785, "learning_rate": 1.485391507596416e-05, "loss": 2.2906, "step": 1271 }, { "epoch": 0.2972744003973007, "grad_norm": 2.5192227363586426, "learning_rate": 1.4865601869887028e-05, "loss": 2.2402, "step": 1272 }, { "epoch": 0.29750810668692124, "grad_norm": 3.3074047565460205, "learning_rate": 1.4877288663809896e-05, "loss": 2.4675, "step": 1273 }, { "epoch": 0.29774181297654173, "grad_norm": 2.49957537651062, "learning_rate": 1.4888975457732762e-05, "loss": 2.3702, "step": 1274 }, { "epoch": 0.2979755192661622, "grad_norm": 2.350102663040161, "learning_rate": 1.4900662251655629e-05, "loss": 2.4729, "step": 1275 }, { "epoch": 0.2982092255557828, "grad_norm": 3.4843363761901855, "learning_rate": 1.4912349045578498e-05, "loss": 2.2939, "step": 1276 }, { "epoch": 0.29844293184540327, "grad_norm": 2.710296392440796, "learning_rate": 1.4924035839501365e-05, "loss": 2.1669, "step": 1277 }, { "epoch": 0.2986766381350238, "grad_norm": 2.862847328186035, "learning_rate": 1.493572263342423e-05, "loss": 2.1824, "step": 1278 }, { "epoch": 0.2989103444246443, "grad_norm": 3.3605027198791504, "learning_rate": 1.4947409427347099e-05, "loss": 2.2619, "step": 1279 }, { "epoch": 0.29914405071426486, "grad_norm": 3.5847361087799072, "learning_rate": 1.4959096221269966e-05, "loss": 2.393, "step": 1280 }, { "epoch": 0.29937775700388536, "grad_norm": 2.3646273612976074, "learning_rate": 1.4970783015192833e-05, "loss": 2.3145, "step": 1281 }, { "epoch": 0.2996114632935059, "grad_norm": 2.217289447784424, "learning_rate": 1.49824698091157e-05, "loss": 2.3222, "step": 1282 }, { "epoch": 0.2998451695831264, "grad_norm": 3.216028928756714, "learning_rate": 1.4994156603038567e-05, "loss": 2.2566, "step": 1283 }, { "epoch": 0.30007887587274695, "grad_norm": 3.456295967102051, "learning_rate": 1.5005843396961432e-05, "loss": 2.2874, "step": 1284 }, { "epoch": 0.30031258216236745, "grad_norm": 3.543412208557129, "learning_rate": 1.5017530190884301e-05, "loss": 2.2424, "step": 1285 }, { "epoch": 0.30054628845198794, "grad_norm": 2.429489850997925, "learning_rate": 1.502921698480717e-05, "loss": 2.1938, "step": 1286 }, { "epoch": 0.3007799947416085, "grad_norm": 3.498807668685913, "learning_rate": 1.5040903778730033e-05, "loss": 2.2362, "step": 1287 }, { "epoch": 0.301013701031229, "grad_norm": 2.7182910442352295, "learning_rate": 1.5052590572652902e-05, "loss": 2.2316, "step": 1288 }, { "epoch": 0.30124740732084954, "grad_norm": 4.120992660522461, "learning_rate": 1.506427736657577e-05, "loss": 2.159, "step": 1289 }, { "epoch": 0.30148111361047003, "grad_norm": 2.3876755237579346, "learning_rate": 1.5075964160498636e-05, "loss": 2.2282, "step": 1290 }, { "epoch": 0.3017148199000906, "grad_norm": 3.018602132797241, "learning_rate": 1.5087650954421503e-05, "loss": 2.3843, "step": 1291 }, { "epoch": 0.3019485261897111, "grad_norm": 2.4224741458892822, "learning_rate": 1.5099337748344372e-05, "loss": 2.2679, "step": 1292 }, { "epoch": 0.3021822324793316, "grad_norm": 2.487619638442993, "learning_rate": 1.511102454226724e-05, "loss": 2.2567, "step": 1293 }, { "epoch": 0.3024159387689521, "grad_norm": 2.271442174911499, "learning_rate": 1.5122711336190104e-05, "loss": 2.3394, "step": 1294 }, { "epoch": 0.3026496450585726, "grad_norm": 2.4295237064361572, "learning_rate": 1.5134398130112973e-05, "loss": 2.3002, "step": 1295 }, { "epoch": 0.30288335134819316, "grad_norm": 2.3637826442718506, "learning_rate": 1.5146084924035842e-05, "loss": 2.423, "step": 1296 }, { "epoch": 0.30311705763781366, "grad_norm": 2.525533676147461, "learning_rate": 1.5157771717958706e-05, "loss": 2.4466, "step": 1297 }, { "epoch": 0.3033507639274342, "grad_norm": 3.241565465927124, "learning_rate": 1.5169458511881574e-05, "loss": 2.3636, "step": 1298 }, { "epoch": 0.3035844702170547, "grad_norm": 3.422006607055664, "learning_rate": 1.5181145305804443e-05, "loss": 2.4563, "step": 1299 }, { "epoch": 0.30381817650667525, "grad_norm": 2.9657394886016846, "learning_rate": 1.5192832099727307e-05, "loss": 2.3342, "step": 1300 }, { "epoch": 0.30405188279629575, "grad_norm": 2.372166156768799, "learning_rate": 1.5204518893650175e-05, "loss": 2.3289, "step": 1301 }, { "epoch": 0.3042855890859163, "grad_norm": 2.4290006160736084, "learning_rate": 1.5216205687573044e-05, "loss": 2.3646, "step": 1302 }, { "epoch": 0.3045192953755368, "grad_norm": 3.037914991378784, "learning_rate": 1.5227892481495908e-05, "loss": 2.3485, "step": 1303 }, { "epoch": 0.3047530016651573, "grad_norm": 2.332611560821533, "learning_rate": 1.5239579275418777e-05, "loss": 2.2408, "step": 1304 }, { "epoch": 0.30498670795477784, "grad_norm": 2.3164358139038086, "learning_rate": 1.5251266069341645e-05, "loss": 2.1614, "step": 1305 }, { "epoch": 0.30522041424439833, "grad_norm": 2.2150087356567383, "learning_rate": 1.526295286326451e-05, "loss": 2.2575, "step": 1306 }, { "epoch": 0.3054541205340189, "grad_norm": 3.3033461570739746, "learning_rate": 1.527463965718738e-05, "loss": 2.1711, "step": 1307 }, { "epoch": 0.3056878268236394, "grad_norm": 2.3355164527893066, "learning_rate": 1.5286326451110246e-05, "loss": 2.2614, "step": 1308 }, { "epoch": 0.3059215331132599, "grad_norm": 4.845826625823975, "learning_rate": 1.5298013245033113e-05, "loss": 2.1824, "step": 1309 }, { "epoch": 0.3061552394028804, "grad_norm": 5.519720554351807, "learning_rate": 1.530970003895598e-05, "loss": 2.4583, "step": 1310 }, { "epoch": 0.30638894569250097, "grad_norm": 2.804513454437256, "learning_rate": 1.5321386832878848e-05, "loss": 2.4619, "step": 1311 }, { "epoch": 0.30662265198212146, "grad_norm": 3.3403704166412354, "learning_rate": 1.5333073626801715e-05, "loss": 2.2943, "step": 1312 }, { "epoch": 0.306856358271742, "grad_norm": 3.7405669689178467, "learning_rate": 1.534476042072458e-05, "loss": 2.2645, "step": 1313 }, { "epoch": 0.3070900645613625, "grad_norm": 4.271100997924805, "learning_rate": 1.535644721464745e-05, "loss": 2.2396, "step": 1314 }, { "epoch": 0.307323770850983, "grad_norm": 2.4641261100769043, "learning_rate": 1.5368134008570316e-05, "loss": 2.192, "step": 1315 }, { "epoch": 0.30755747714060355, "grad_norm": 3.56257700920105, "learning_rate": 1.5379820802493183e-05, "loss": 2.2306, "step": 1316 }, { "epoch": 0.30779118343022405, "grad_norm": 3.2128102779388428, "learning_rate": 1.539150759641605e-05, "loss": 2.3411, "step": 1317 }, { "epoch": 0.3080248897198446, "grad_norm": 3.182624101638794, "learning_rate": 1.5403194390338917e-05, "loss": 2.3898, "step": 1318 }, { "epoch": 0.3082585960094651, "grad_norm": 2.5144176483154297, "learning_rate": 1.5414881184261784e-05, "loss": 2.6013, "step": 1319 }, { "epoch": 0.30849230229908564, "grad_norm": 5.035153865814209, "learning_rate": 1.542656797818465e-05, "loss": 2.2149, "step": 1320 }, { "epoch": 0.30872600858870614, "grad_norm": 3.1745288372039795, "learning_rate": 1.5438254772107518e-05, "loss": 2.3897, "step": 1321 }, { "epoch": 0.3089597148783267, "grad_norm": 2.5395023822784424, "learning_rate": 1.5449941566030385e-05, "loss": 2.3321, "step": 1322 }, { "epoch": 0.3091934211679472, "grad_norm": 3.2046735286712646, "learning_rate": 1.5461628359953252e-05, "loss": 2.31, "step": 1323 }, { "epoch": 0.3094271274575677, "grad_norm": 3.0614919662475586, "learning_rate": 1.547331515387612e-05, "loss": 2.3026, "step": 1324 }, { "epoch": 0.3096608337471882, "grad_norm": 2.716216802597046, "learning_rate": 1.548500194779899e-05, "loss": 2.3571, "step": 1325 }, { "epoch": 0.3098945400368087, "grad_norm": 2.4935269355773926, "learning_rate": 1.5496688741721853e-05, "loss": 2.2407, "step": 1326 }, { "epoch": 0.31012824632642927, "grad_norm": 4.724549293518066, "learning_rate": 1.5508375535644724e-05, "loss": 2.2603, "step": 1327 }, { "epoch": 0.31036195261604976, "grad_norm": 2.9275360107421875, "learning_rate": 1.552006232956759e-05, "loss": 2.4184, "step": 1328 }, { "epoch": 0.3105956589056703, "grad_norm": 2.809265613555908, "learning_rate": 1.5531749123490454e-05, "loss": 2.3128, "step": 1329 }, { "epoch": 0.3108293651952908, "grad_norm": 4.215442180633545, "learning_rate": 1.5543435917413325e-05, "loss": 2.2711, "step": 1330 }, { "epoch": 0.31106307148491136, "grad_norm": 2.798297166824341, "learning_rate": 1.5555122711336192e-05, "loss": 2.3407, "step": 1331 }, { "epoch": 0.31129677777453185, "grad_norm": 2.3231165409088135, "learning_rate": 1.5566809505259056e-05, "loss": 2.2652, "step": 1332 }, { "epoch": 0.3115304840641524, "grad_norm": 3.9674465656280518, "learning_rate": 1.5578496299181926e-05, "loss": 2.3269, "step": 1333 }, { "epoch": 0.3117641903537729, "grad_norm": 2.8501808643341064, "learning_rate": 1.5590183093104793e-05, "loss": 2.2381, "step": 1334 }, { "epoch": 0.3119978966433934, "grad_norm": 2.336064338684082, "learning_rate": 1.5601869887027657e-05, "loss": 2.4299, "step": 1335 }, { "epoch": 0.31223160293301394, "grad_norm": 2.3492794036865234, "learning_rate": 1.5613556680950527e-05, "loss": 2.2469, "step": 1336 }, { "epoch": 0.31246530922263444, "grad_norm": 2.881326198577881, "learning_rate": 1.5625243474873394e-05, "loss": 2.2019, "step": 1337 }, { "epoch": 0.312699015512255, "grad_norm": 2.237765312194824, "learning_rate": 1.5636930268796258e-05, "loss": 2.1775, "step": 1338 }, { "epoch": 0.3129327218018755, "grad_norm": 2.5871572494506836, "learning_rate": 1.5648617062719128e-05, "loss": 2.4299, "step": 1339 }, { "epoch": 0.31316642809149603, "grad_norm": 2.500730276107788, "learning_rate": 1.5660303856641995e-05, "loss": 2.2001, "step": 1340 }, { "epoch": 0.3134001343811165, "grad_norm": 3.340531826019287, "learning_rate": 1.5671990650564862e-05, "loss": 2.2108, "step": 1341 }, { "epoch": 0.3136338406707371, "grad_norm": 3.2704031467437744, "learning_rate": 1.568367744448773e-05, "loss": 2.3094, "step": 1342 }, { "epoch": 0.31386754696035757, "grad_norm": 2.8035941123962402, "learning_rate": 1.5695364238410596e-05, "loss": 2.3682, "step": 1343 }, { "epoch": 0.31410125324997806, "grad_norm": 2.708469867706299, "learning_rate": 1.5707051032333463e-05, "loss": 2.2867, "step": 1344 }, { "epoch": 0.3143349595395986, "grad_norm": 2.3315978050231934, "learning_rate": 1.571873782625633e-05, "loss": 2.1376, "step": 1345 }, { "epoch": 0.3145686658292191, "grad_norm": 3.1524722576141357, "learning_rate": 1.5730424620179198e-05, "loss": 2.4386, "step": 1346 }, { "epoch": 0.31480237211883966, "grad_norm": 3.1788883209228516, "learning_rate": 1.5742111414102065e-05, "loss": 2.3847, "step": 1347 }, { "epoch": 0.31503607840846015, "grad_norm": 3.504059314727783, "learning_rate": 1.575379820802493e-05, "loss": 2.316, "step": 1348 }, { "epoch": 0.3152697846980807, "grad_norm": 2.4764530658721924, "learning_rate": 1.57654850019478e-05, "loss": 2.5161, "step": 1349 }, { "epoch": 0.3155034909877012, "grad_norm": 2.2316174507141113, "learning_rate": 1.5777171795870666e-05, "loss": 2.3138, "step": 1350 }, { "epoch": 0.31573719727732175, "grad_norm": 2.3574604988098145, "learning_rate": 1.5788858589793533e-05, "loss": 2.2637, "step": 1351 }, { "epoch": 0.31597090356694224, "grad_norm": 2.9759974479675293, "learning_rate": 1.58005453837164e-05, "loss": 2.5872, "step": 1352 }, { "epoch": 0.3162046098565628, "grad_norm": 2.8900463581085205, "learning_rate": 1.581223217763927e-05, "loss": 2.2154, "step": 1353 }, { "epoch": 0.3164383161461833, "grad_norm": 3.582139253616333, "learning_rate": 1.5823918971562137e-05, "loss": 2.3189, "step": 1354 }, { "epoch": 0.3166720224358038, "grad_norm": 3.6511197090148926, "learning_rate": 1.5835605765485e-05, "loss": 2.467, "step": 1355 }, { "epoch": 0.31690572872542433, "grad_norm": 2.440196990966797, "learning_rate": 1.584729255940787e-05, "loss": 2.2169, "step": 1356 }, { "epoch": 0.3171394350150448, "grad_norm": 3.0044236183166504, "learning_rate": 1.585897935333074e-05, "loss": 2.1972, "step": 1357 }, { "epoch": 0.3173731413046654, "grad_norm": 2.7334978580474854, "learning_rate": 1.5870666147253602e-05, "loss": 2.2521, "step": 1358 }, { "epoch": 0.31760684759428587, "grad_norm": 2.245666265487671, "learning_rate": 1.5882352941176473e-05, "loss": 2.207, "step": 1359 }, { "epoch": 0.3178405538839064, "grad_norm": 2.414252758026123, "learning_rate": 1.589403973509934e-05, "loss": 2.3268, "step": 1360 }, { "epoch": 0.3180742601735269, "grad_norm": 2.7271742820739746, "learning_rate": 1.5905726529022203e-05, "loss": 2.2012, "step": 1361 }, { "epoch": 0.31830796646314746, "grad_norm": 2.5593559741973877, "learning_rate": 1.5917413322945074e-05, "loss": 2.424, "step": 1362 }, { "epoch": 0.31854167275276796, "grad_norm": 2.3680126667022705, "learning_rate": 1.592910011686794e-05, "loss": 2.2861, "step": 1363 }, { "epoch": 0.31877537904238845, "grad_norm": 3.0441720485687256, "learning_rate": 1.5940786910790804e-05, "loss": 2.4291, "step": 1364 }, { "epoch": 0.319009085332009, "grad_norm": 2.9274628162384033, "learning_rate": 1.5952473704713675e-05, "loss": 2.3231, "step": 1365 }, { "epoch": 0.3192427916216295, "grad_norm": 2.9582931995391846, "learning_rate": 1.5964160498636542e-05, "loss": 2.3729, "step": 1366 }, { "epoch": 0.31947649791125005, "grad_norm": 2.4255290031433105, "learning_rate": 1.5975847292559406e-05, "loss": 2.229, "step": 1367 }, { "epoch": 0.31971020420087054, "grad_norm": 2.8531265258789062, "learning_rate": 1.5987534086482276e-05, "loss": 2.4521, "step": 1368 }, { "epoch": 0.3199439104904911, "grad_norm": 2.935349225997925, "learning_rate": 1.5999220880405143e-05, "loss": 2.3102, "step": 1369 }, { "epoch": 0.3201776167801116, "grad_norm": 2.8539156913757324, "learning_rate": 1.601090767432801e-05, "loss": 2.3142, "step": 1370 }, { "epoch": 0.32041132306973213, "grad_norm": 4.203031063079834, "learning_rate": 1.6022594468250877e-05, "loss": 2.3978, "step": 1371 }, { "epoch": 0.32064502935935263, "grad_norm": 2.6010851860046387, "learning_rate": 1.6034281262173744e-05, "loss": 2.3277, "step": 1372 }, { "epoch": 0.3208787356489732, "grad_norm": 2.6946003437042236, "learning_rate": 1.604596805609661e-05, "loss": 2.2476, "step": 1373 }, { "epoch": 0.3211124419385937, "grad_norm": 2.4215428829193115, "learning_rate": 1.6057654850019478e-05, "loss": 2.4546, "step": 1374 }, { "epoch": 0.32134614822821417, "grad_norm": 2.9222288131713867, "learning_rate": 1.6069341643942345e-05, "loss": 2.3131, "step": 1375 }, { "epoch": 0.3215798545178347, "grad_norm": 2.802823066711426, "learning_rate": 1.6081028437865212e-05, "loss": 2.3966, "step": 1376 }, { "epoch": 0.3218135608074552, "grad_norm": 2.8585894107818604, "learning_rate": 1.609271523178808e-05, "loss": 2.3249, "step": 1377 }, { "epoch": 0.32204726709707576, "grad_norm": 2.606731653213501, "learning_rate": 1.6104402025710946e-05, "loss": 2.3311, "step": 1378 }, { "epoch": 0.32228097338669626, "grad_norm": 2.212160587310791, "learning_rate": 1.6116088819633814e-05, "loss": 2.266, "step": 1379 }, { "epoch": 0.3225146796763168, "grad_norm": 3.7574806213378906, "learning_rate": 1.612777561355668e-05, "loss": 2.3997, "step": 1380 }, { "epoch": 0.3227483859659373, "grad_norm": 3.613532066345215, "learning_rate": 1.6139462407479548e-05, "loss": 2.2966, "step": 1381 }, { "epoch": 0.32298209225555785, "grad_norm": 3.0834741592407227, "learning_rate": 1.6151149201402418e-05, "loss": 2.1968, "step": 1382 }, { "epoch": 0.32321579854517835, "grad_norm": 2.6541733741760254, "learning_rate": 1.6162835995325282e-05, "loss": 2.3382, "step": 1383 }, { "epoch": 0.32344950483479884, "grad_norm": 4.043711185455322, "learning_rate": 1.617452278924815e-05, "loss": 2.3283, "step": 1384 }, { "epoch": 0.3236832111244194, "grad_norm": 2.536165475845337, "learning_rate": 1.618620958317102e-05, "loss": 2.1883, "step": 1385 }, { "epoch": 0.3239169174140399, "grad_norm": 2.830601453781128, "learning_rate": 1.6197896377093886e-05, "loss": 2.3139, "step": 1386 }, { "epoch": 0.32415062370366043, "grad_norm": 3.99747633934021, "learning_rate": 1.620958317101675e-05, "loss": 2.3029, "step": 1387 }, { "epoch": 0.32438432999328093, "grad_norm": 2.9461965560913086, "learning_rate": 1.622126996493962e-05, "loss": 2.3638, "step": 1388 }, { "epoch": 0.3246180362829015, "grad_norm": 2.7273035049438477, "learning_rate": 1.6232956758862487e-05, "loss": 2.2179, "step": 1389 }, { "epoch": 0.324851742572522, "grad_norm": 2.588639736175537, "learning_rate": 1.624464355278535e-05, "loss": 2.2044, "step": 1390 }, { "epoch": 0.3250854488621425, "grad_norm": 2.3338706493377686, "learning_rate": 1.625633034670822e-05, "loss": 2.3086, "step": 1391 }, { "epoch": 0.325319155151763, "grad_norm": 2.5103867053985596, "learning_rate": 1.626801714063109e-05, "loss": 2.2715, "step": 1392 }, { "epoch": 0.32555286144138357, "grad_norm": 2.749269962310791, "learning_rate": 1.6279703934553952e-05, "loss": 2.1844, "step": 1393 }, { "epoch": 0.32578656773100406, "grad_norm": 3.650484561920166, "learning_rate": 1.6291390728476823e-05, "loss": 2.2879, "step": 1394 }, { "epoch": 0.32602027402062456, "grad_norm": 2.5238258838653564, "learning_rate": 1.630307752239969e-05, "loss": 2.1226, "step": 1395 }, { "epoch": 0.3262539803102451, "grad_norm": 2.703458786010742, "learning_rate": 1.6314764316322553e-05, "loss": 2.3183, "step": 1396 }, { "epoch": 0.3264876865998656, "grad_norm": 2.433992385864258, "learning_rate": 1.6326451110245424e-05, "loss": 2.2791, "step": 1397 }, { "epoch": 0.32672139288948615, "grad_norm": 2.5490517616271973, "learning_rate": 1.633813790416829e-05, "loss": 2.1982, "step": 1398 }, { "epoch": 0.32695509917910665, "grad_norm": 2.2691917419433594, "learning_rate": 1.6349824698091154e-05, "loss": 2.2861, "step": 1399 }, { "epoch": 0.3271888054687272, "grad_norm": 2.6952431201934814, "learning_rate": 1.6361511492014025e-05, "loss": 2.1421, "step": 1400 }, { "epoch": 0.3274225117583477, "grad_norm": 3.3676090240478516, "learning_rate": 1.6373198285936892e-05, "loss": 2.2887, "step": 1401 }, { "epoch": 0.32765621804796824, "grad_norm": 2.8645269870758057, "learning_rate": 1.638488507985976e-05, "loss": 2.3963, "step": 1402 }, { "epoch": 0.32788992433758873, "grad_norm": 2.4079253673553467, "learning_rate": 1.6396571873782626e-05, "loss": 2.1867, "step": 1403 }, { "epoch": 0.32812363062720923, "grad_norm": 4.076973915100098, "learning_rate": 1.6408258667705493e-05, "loss": 2.4671, "step": 1404 }, { "epoch": 0.3283573369168298, "grad_norm": 2.8844854831695557, "learning_rate": 1.641994546162836e-05, "loss": 2.2762, "step": 1405 }, { "epoch": 0.3285910432064503, "grad_norm": 2.5844385623931885, "learning_rate": 1.6431632255551227e-05, "loss": 2.2626, "step": 1406 }, { "epoch": 0.3288247494960708, "grad_norm": 3.1600680351257324, "learning_rate": 1.6443319049474094e-05, "loss": 2.2755, "step": 1407 }, { "epoch": 0.3290584557856913, "grad_norm": 4.079757213592529, "learning_rate": 1.6455005843396965e-05, "loss": 2.1065, "step": 1408 }, { "epoch": 0.32929216207531187, "grad_norm": 2.6553499698638916, "learning_rate": 1.646669263731983e-05, "loss": 2.3457, "step": 1409 }, { "epoch": 0.32952586836493236, "grad_norm": 10.21612548828125, "learning_rate": 1.6478379431242695e-05, "loss": 2.5254, "step": 1410 }, { "epoch": 0.3297595746545529, "grad_norm": 2.8903937339782715, "learning_rate": 1.6490066225165566e-05, "loss": 2.2104, "step": 1411 }, { "epoch": 0.3299932809441734, "grad_norm": 3.421772003173828, "learning_rate": 1.650175301908843e-05, "loss": 2.3819, "step": 1412 }, { "epoch": 0.33022698723379396, "grad_norm": 3.801422119140625, "learning_rate": 1.6513439813011297e-05, "loss": 2.2424, "step": 1413 }, { "epoch": 0.33046069352341445, "grad_norm": 2.3630783557891846, "learning_rate": 1.6525126606934167e-05, "loss": 2.2151, "step": 1414 }, { "epoch": 0.33069439981303494, "grad_norm": 3.483093500137329, "learning_rate": 1.653681340085703e-05, "loss": 2.2045, "step": 1415 }, { "epoch": 0.3309281061026555, "grad_norm": 3.003427028656006, "learning_rate": 1.6548500194779898e-05, "loss": 2.1943, "step": 1416 }, { "epoch": 0.331161812392276, "grad_norm": 2.777116060256958, "learning_rate": 1.6560186988702768e-05, "loss": 2.3049, "step": 1417 }, { "epoch": 0.33139551868189654, "grad_norm": 3.4670329093933105, "learning_rate": 1.6571873782625635e-05, "loss": 2.2605, "step": 1418 }, { "epoch": 0.33162922497151703, "grad_norm": 2.692005157470703, "learning_rate": 1.65835605765485e-05, "loss": 2.3438, "step": 1419 }, { "epoch": 0.3318629312611376, "grad_norm": 2.631028175354004, "learning_rate": 1.659524737047137e-05, "loss": 2.2771, "step": 1420 }, { "epoch": 0.3320966375507581, "grad_norm": 2.3305721282958984, "learning_rate": 1.6606934164394236e-05, "loss": 2.3521, "step": 1421 }, { "epoch": 0.33233034384037863, "grad_norm": 2.2624549865722656, "learning_rate": 1.66186209583171e-05, "loss": 2.3775, "step": 1422 }, { "epoch": 0.3325640501299991, "grad_norm": 2.3397061824798584, "learning_rate": 1.663030775223997e-05, "loss": 2.2195, "step": 1423 }, { "epoch": 0.3327977564196196, "grad_norm": 2.5057177543640137, "learning_rate": 1.6641994546162837e-05, "loss": 2.1705, "step": 1424 }, { "epoch": 0.33303146270924017, "grad_norm": 2.5751054286956787, "learning_rate": 1.66536813400857e-05, "loss": 2.2064, "step": 1425 }, { "epoch": 0.33326516899886066, "grad_norm": 2.382983684539795, "learning_rate": 1.666536813400857e-05, "loss": 2.2689, "step": 1426 }, { "epoch": 0.3334988752884812, "grad_norm": 2.3909337520599365, "learning_rate": 1.667705492793144e-05, "loss": 2.3162, "step": 1427 }, { "epoch": 0.3337325815781017, "grad_norm": 3.2040798664093018, "learning_rate": 1.6688741721854302e-05, "loss": 2.2166, "step": 1428 }, { "epoch": 0.33396628786772226, "grad_norm": 3.0267107486724854, "learning_rate": 1.6700428515777173e-05, "loss": 2.2136, "step": 1429 }, { "epoch": 0.33419999415734275, "grad_norm": 2.6430466175079346, "learning_rate": 1.671211530970004e-05, "loss": 2.4536, "step": 1430 }, { "epoch": 0.3344337004469633, "grad_norm": 2.3465027809143066, "learning_rate": 1.6723802103622903e-05, "loss": 2.2935, "step": 1431 }, { "epoch": 0.3346674067365838, "grad_norm": 4.257509231567383, "learning_rate": 1.6735488897545774e-05, "loss": 2.1751, "step": 1432 }, { "epoch": 0.33490111302620434, "grad_norm": 3.463031530380249, "learning_rate": 1.674717569146864e-05, "loss": 2.3691, "step": 1433 }, { "epoch": 0.33513481931582484, "grad_norm": 2.5014231204986572, "learning_rate": 1.6758862485391508e-05, "loss": 2.3033, "step": 1434 }, { "epoch": 0.33536852560544533, "grad_norm": 4.217365264892578, "learning_rate": 1.6770549279314375e-05, "loss": 2.3761, "step": 1435 }, { "epoch": 0.3356022318950659, "grad_norm": 4.266536712646484, "learning_rate": 1.6782236073237242e-05, "loss": 2.1561, "step": 1436 }, { "epoch": 0.3358359381846864, "grad_norm": 3.1609902381896973, "learning_rate": 1.6793922867160112e-05, "loss": 2.219, "step": 1437 }, { "epoch": 0.33606964447430693, "grad_norm": 2.6708686351776123, "learning_rate": 1.6805609661082976e-05, "loss": 2.2729, "step": 1438 }, { "epoch": 0.3363033507639274, "grad_norm": 4.132392406463623, "learning_rate": 1.6817296455005843e-05, "loss": 2.2848, "step": 1439 }, { "epoch": 0.33653705705354797, "grad_norm": 2.8049404621124268, "learning_rate": 1.6828983248928714e-05, "loss": 2.3728, "step": 1440 }, { "epoch": 0.33677076334316847, "grad_norm": 2.5647432804107666, "learning_rate": 1.6840670042851577e-05, "loss": 2.2518, "step": 1441 }, { "epoch": 0.337004469632789, "grad_norm": 2.742530345916748, "learning_rate": 1.6852356836774444e-05, "loss": 2.3921, "step": 1442 }, { "epoch": 0.3372381759224095, "grad_norm": 2.196401357650757, "learning_rate": 1.6864043630697315e-05, "loss": 2.335, "step": 1443 }, { "epoch": 0.33747188221203, "grad_norm": 2.628990888595581, "learning_rate": 1.687573042462018e-05, "loss": 2.1479, "step": 1444 }, { "epoch": 0.33770558850165056, "grad_norm": 2.4397456645965576, "learning_rate": 1.6887417218543045e-05, "loss": 2.3178, "step": 1445 }, { "epoch": 0.33793929479127105, "grad_norm": 3.034738063812256, "learning_rate": 1.6899104012465916e-05, "loss": 2.2009, "step": 1446 }, { "epoch": 0.3381730010808916, "grad_norm": 2.7283244132995605, "learning_rate": 1.691079080638878e-05, "loss": 2.1676, "step": 1447 }, { "epoch": 0.3384067073705121, "grad_norm": 2.755211114883423, "learning_rate": 1.6922477600311647e-05, "loss": 2.1679, "step": 1448 }, { "epoch": 0.33864041366013264, "grad_norm": 3.290980815887451, "learning_rate": 1.6934164394234517e-05, "loss": 2.3948, "step": 1449 }, { "epoch": 0.33887411994975314, "grad_norm": 2.4375593662261963, "learning_rate": 1.6945851188157384e-05, "loss": 2.286, "step": 1450 }, { "epoch": 0.3391078262393737, "grad_norm": 2.8794679641723633, "learning_rate": 1.6957537982080248e-05, "loss": 2.5159, "step": 1451 }, { "epoch": 0.3393415325289942, "grad_norm": 4.234555721282959, "learning_rate": 1.6969224776003118e-05, "loss": 2.1361, "step": 1452 }, { "epoch": 0.33957523881861473, "grad_norm": 3.8656797409057617, "learning_rate": 1.6980911569925985e-05, "loss": 2.2412, "step": 1453 }, { "epoch": 0.3398089451082352, "grad_norm": 2.8643958568573, "learning_rate": 1.699259836384885e-05, "loss": 2.2602, "step": 1454 }, { "epoch": 0.3400426513978557, "grad_norm": 4.204207420349121, "learning_rate": 1.700428515777172e-05, "loss": 2.389, "step": 1455 }, { "epoch": 0.34027635768747627, "grad_norm": 5.575317859649658, "learning_rate": 1.7015971951694586e-05, "loss": 2.4836, "step": 1456 }, { "epoch": 0.34051006397709677, "grad_norm": 3.6205739974975586, "learning_rate": 1.702765874561745e-05, "loss": 2.2827, "step": 1457 }, { "epoch": 0.3407437702667173, "grad_norm": 2.465664863586426, "learning_rate": 1.703934553954032e-05, "loss": 2.2404, "step": 1458 }, { "epoch": 0.3409774765563378, "grad_norm": 2.8389151096343994, "learning_rate": 1.7051032333463187e-05, "loss": 2.2301, "step": 1459 }, { "epoch": 0.34121118284595836, "grad_norm": 3.744534969329834, "learning_rate": 1.706271912738605e-05, "loss": 2.2493, "step": 1460 }, { "epoch": 0.34144488913557886, "grad_norm": 2.752701759338379, "learning_rate": 1.707440592130892e-05, "loss": 2.1915, "step": 1461 }, { "epoch": 0.3416785954251994, "grad_norm": 3.4373786449432373, "learning_rate": 1.708609271523179e-05, "loss": 2.3632, "step": 1462 }, { "epoch": 0.3419123017148199, "grad_norm": 3.4748947620391846, "learning_rate": 1.7097779509154656e-05, "loss": 2.3594, "step": 1463 }, { "epoch": 0.3421460080044404, "grad_norm": 2.3317224979400635, "learning_rate": 1.7109466303077523e-05, "loss": 2.3551, "step": 1464 }, { "epoch": 0.34237971429406094, "grad_norm": 2.900691032409668, "learning_rate": 1.712115309700039e-05, "loss": 2.5321, "step": 1465 }, { "epoch": 0.34261342058368144, "grad_norm": 2.8849165439605713, "learning_rate": 1.713283989092326e-05, "loss": 2.3373, "step": 1466 }, { "epoch": 0.342847126873302, "grad_norm": 2.978261947631836, "learning_rate": 1.7144526684846124e-05, "loss": 2.1887, "step": 1467 }, { "epoch": 0.3430808331629225, "grad_norm": 2.344787359237671, "learning_rate": 1.715621347876899e-05, "loss": 2.1446, "step": 1468 }, { "epoch": 0.34331453945254303, "grad_norm": 2.4905948638916016, "learning_rate": 1.716790027269186e-05, "loss": 2.2119, "step": 1469 }, { "epoch": 0.3435482457421635, "grad_norm": 2.6582190990448, "learning_rate": 1.7179587066614725e-05, "loss": 2.1811, "step": 1470 }, { "epoch": 0.3437819520317841, "grad_norm": 2.8836324214935303, "learning_rate": 1.7191273860537592e-05, "loss": 2.2061, "step": 1471 }, { "epoch": 0.34401565832140457, "grad_norm": 2.4177825450897217, "learning_rate": 1.7202960654460462e-05, "loss": 2.1347, "step": 1472 }, { "epoch": 0.3442493646110251, "grad_norm": 3.6461477279663086, "learning_rate": 1.7214647448383326e-05, "loss": 2.2975, "step": 1473 }, { "epoch": 0.3444830709006456, "grad_norm": 3.7861216068267822, "learning_rate": 1.7226334242306193e-05, "loss": 2.2455, "step": 1474 }, { "epoch": 0.3447167771902661, "grad_norm": 2.665552854537964, "learning_rate": 1.7238021036229064e-05, "loss": 2.3334, "step": 1475 }, { "epoch": 0.34495048347988666, "grad_norm": 2.402681827545166, "learning_rate": 1.7249707830151927e-05, "loss": 2.2179, "step": 1476 }, { "epoch": 0.34518418976950715, "grad_norm": 2.567901372909546, "learning_rate": 1.7261394624074794e-05, "loss": 2.2106, "step": 1477 }, { "epoch": 0.3454178960591277, "grad_norm": 2.4071826934814453, "learning_rate": 1.7273081417997665e-05, "loss": 2.3474, "step": 1478 }, { "epoch": 0.3456516023487482, "grad_norm": 3.6077182292938232, "learning_rate": 1.728476821192053e-05, "loss": 2.1682, "step": 1479 }, { "epoch": 0.34588530863836875, "grad_norm": 2.5693228244781494, "learning_rate": 1.7296455005843395e-05, "loss": 2.1306, "step": 1480 }, { "epoch": 0.34611901492798924, "grad_norm": 2.7436304092407227, "learning_rate": 1.7308141799766266e-05, "loss": 2.344, "step": 1481 }, { "epoch": 0.3463527212176098, "grad_norm": 4.2990031242370605, "learning_rate": 1.7319828593689133e-05, "loss": 2.1709, "step": 1482 }, { "epoch": 0.3465864275072303, "grad_norm": 3.4644789695739746, "learning_rate": 1.7331515387611997e-05, "loss": 2.1865, "step": 1483 }, { "epoch": 0.3468201337968508, "grad_norm": 2.420722246170044, "learning_rate": 1.7343202181534867e-05, "loss": 2.3721, "step": 1484 }, { "epoch": 0.34705384008647133, "grad_norm": 3.7064285278320312, "learning_rate": 1.7354888975457734e-05, "loss": 2.3722, "step": 1485 }, { "epoch": 0.3472875463760918, "grad_norm": 3.141497850418091, "learning_rate": 1.7366575769380598e-05, "loss": 2.3187, "step": 1486 }, { "epoch": 0.3475212526657124, "grad_norm": 2.379324436187744, "learning_rate": 1.7378262563303468e-05, "loss": 2.3911, "step": 1487 }, { "epoch": 0.34775495895533287, "grad_norm": 2.9103786945343018, "learning_rate": 1.7389949357226335e-05, "loss": 2.3822, "step": 1488 }, { "epoch": 0.3479886652449534, "grad_norm": 4.254236221313477, "learning_rate": 1.7401636151149202e-05, "loss": 2.1694, "step": 1489 }, { "epoch": 0.3482223715345739, "grad_norm": 2.399022340774536, "learning_rate": 1.741332294507207e-05, "loss": 2.3985, "step": 1490 }, { "epoch": 0.34845607782419447, "grad_norm": 3.1901538372039795, "learning_rate": 1.7425009738994936e-05, "loss": 2.1896, "step": 1491 }, { "epoch": 0.34868978411381496, "grad_norm": 3.1737446784973145, "learning_rate": 1.7436696532917803e-05, "loss": 2.3872, "step": 1492 }, { "epoch": 0.3489234904034355, "grad_norm": 2.6282968521118164, "learning_rate": 1.744838332684067e-05, "loss": 2.2138, "step": 1493 }, { "epoch": 0.349157196693056, "grad_norm": 2.9080917835235596, "learning_rate": 1.7460070120763537e-05, "loss": 2.4371, "step": 1494 }, { "epoch": 0.3493909029826765, "grad_norm": 2.377211809158325, "learning_rate": 1.7471756914686408e-05, "loss": 2.3837, "step": 1495 }, { "epoch": 0.34962460927229705, "grad_norm": 2.5678935050964355, "learning_rate": 1.748344370860927e-05, "loss": 2.1585, "step": 1496 }, { "epoch": 0.34985831556191754, "grad_norm": 2.612926483154297, "learning_rate": 1.749513050253214e-05, "loss": 2.2512, "step": 1497 }, { "epoch": 0.3500920218515381, "grad_norm": 2.6118743419647217, "learning_rate": 1.750681729645501e-05, "loss": 2.2089, "step": 1498 }, { "epoch": 0.3503257281411586, "grad_norm": 2.54852032661438, "learning_rate": 1.7518504090377873e-05, "loss": 2.2308, "step": 1499 }, { "epoch": 0.35055943443077914, "grad_norm": 2.307025194168091, "learning_rate": 1.753019088430074e-05, "loss": 2.2712, "step": 1500 }, { "epoch": 0.35079314072039963, "grad_norm": 2.4851632118225098, "learning_rate": 1.754187767822361e-05, "loss": 2.1929, "step": 1501 }, { "epoch": 0.3510268470100202, "grad_norm": 2.502549886703491, "learning_rate": 1.7553564472146474e-05, "loss": 2.3648, "step": 1502 }, { "epoch": 0.3512605532996407, "grad_norm": 2.696831226348877, "learning_rate": 1.756525126606934e-05, "loss": 2.3501, "step": 1503 }, { "epoch": 0.35149425958926117, "grad_norm": 2.4532153606414795, "learning_rate": 1.757693805999221e-05, "loss": 2.1705, "step": 1504 }, { "epoch": 0.3517279658788817, "grad_norm": 2.589041233062744, "learning_rate": 1.7588624853915075e-05, "loss": 2.1581, "step": 1505 }, { "epoch": 0.3519616721685022, "grad_norm": 2.384111166000366, "learning_rate": 1.7600311647837942e-05, "loss": 2.2631, "step": 1506 }, { "epoch": 0.35219537845812277, "grad_norm": 4.04292631149292, "learning_rate": 1.7611998441760812e-05, "loss": 2.3517, "step": 1507 }, { "epoch": 0.35242908474774326, "grad_norm": 3.586549758911133, "learning_rate": 1.7623685235683676e-05, "loss": 2.137, "step": 1508 }, { "epoch": 0.3526627910373638, "grad_norm": 2.759960412979126, "learning_rate": 1.7635372029606543e-05, "loss": 2.2453, "step": 1509 }, { "epoch": 0.3528964973269843, "grad_norm": 2.687126874923706, "learning_rate": 1.7647058823529414e-05, "loss": 2.1894, "step": 1510 }, { "epoch": 0.35313020361660485, "grad_norm": 5.244961261749268, "learning_rate": 1.765874561745228e-05, "loss": 2.2488, "step": 1511 }, { "epoch": 0.35336390990622535, "grad_norm": 3.7604329586029053, "learning_rate": 1.7670432411375144e-05, "loss": 2.3907, "step": 1512 }, { "epoch": 0.35359761619584584, "grad_norm": 2.6936159133911133, "learning_rate": 1.7682119205298015e-05, "loss": 2.2061, "step": 1513 }, { "epoch": 0.3538313224854664, "grad_norm": 3.7002673149108887, "learning_rate": 1.7693805999220882e-05, "loss": 2.2907, "step": 1514 }, { "epoch": 0.3540650287750869, "grad_norm": 2.73586106300354, "learning_rate": 1.7705492793143745e-05, "loss": 2.1745, "step": 1515 }, { "epoch": 0.35429873506470744, "grad_norm": 2.840665578842163, "learning_rate": 1.7717179587066616e-05, "loss": 2.2653, "step": 1516 }, { "epoch": 0.35453244135432793, "grad_norm": 2.7260360717773438, "learning_rate": 1.7728866380989483e-05, "loss": 2.3837, "step": 1517 }, { "epoch": 0.3547661476439485, "grad_norm": 2.290071725845337, "learning_rate": 1.774055317491235e-05, "loss": 2.2152, "step": 1518 }, { "epoch": 0.354999853933569, "grad_norm": 3.1425247192382812, "learning_rate": 1.7752239968835217e-05, "loss": 2.3267, "step": 1519 }, { "epoch": 0.3552335602231895, "grad_norm": 3.2097771167755127, "learning_rate": 1.7763926762758084e-05, "loss": 2.0763, "step": 1520 }, { "epoch": 0.35546726651281, "grad_norm": 2.720566749572754, "learning_rate": 1.777561355668095e-05, "loss": 2.2833, "step": 1521 }, { "epoch": 0.35570097280243057, "grad_norm": 2.3080949783325195, "learning_rate": 1.7787300350603818e-05, "loss": 2.1674, "step": 1522 }, { "epoch": 0.35593467909205107, "grad_norm": 2.290910482406616, "learning_rate": 1.7798987144526685e-05, "loss": 2.1344, "step": 1523 }, { "epoch": 0.35616838538167156, "grad_norm": 2.6790013313293457, "learning_rate": 1.7810673938449552e-05, "loss": 2.3783, "step": 1524 }, { "epoch": 0.3564020916712921, "grad_norm": 2.7726643085479736, "learning_rate": 1.782236073237242e-05, "loss": 2.3852, "step": 1525 }, { "epoch": 0.3566357979609126, "grad_norm": 3.805816650390625, "learning_rate": 1.7834047526295286e-05, "loss": 2.3911, "step": 1526 }, { "epoch": 0.35686950425053315, "grad_norm": 2.658165693283081, "learning_rate": 1.7845734320218157e-05, "loss": 2.2461, "step": 1527 }, { "epoch": 0.35710321054015365, "grad_norm": 5.174410343170166, "learning_rate": 1.785742111414102e-05, "loss": 2.4686, "step": 1528 }, { "epoch": 0.3573369168297742, "grad_norm": 3.084965944290161, "learning_rate": 1.7869107908063887e-05, "loss": 2.1221, "step": 1529 }, { "epoch": 0.3575706231193947, "grad_norm": 3.1827642917633057, "learning_rate": 1.7880794701986758e-05, "loss": 2.3624, "step": 1530 }, { "epoch": 0.35780432940901524, "grad_norm": 2.779664993286133, "learning_rate": 1.789248149590962e-05, "loss": 2.2266, "step": 1531 }, { "epoch": 0.35803803569863574, "grad_norm": 3.214322805404663, "learning_rate": 1.790416828983249e-05, "loss": 2.1413, "step": 1532 }, { "epoch": 0.35827174198825623, "grad_norm": 2.5505142211914062, "learning_rate": 1.791585508375536e-05, "loss": 2.3137, "step": 1533 }, { "epoch": 0.3585054482778768, "grad_norm": 2.874760389328003, "learning_rate": 1.7927541877678223e-05, "loss": 2.1792, "step": 1534 }, { "epoch": 0.3587391545674973, "grad_norm": 2.5523197650909424, "learning_rate": 1.793922867160109e-05, "loss": 2.138, "step": 1535 }, { "epoch": 0.3589728608571178, "grad_norm": 2.5927765369415283, "learning_rate": 1.795091546552396e-05, "loss": 2.341, "step": 1536 }, { "epoch": 0.3592065671467383, "grad_norm": 3.7224764823913574, "learning_rate": 1.7962602259446824e-05, "loss": 2.1765, "step": 1537 }, { "epoch": 0.35944027343635887, "grad_norm": 2.3791534900665283, "learning_rate": 1.797428905336969e-05, "loss": 2.2439, "step": 1538 }, { "epoch": 0.35967397972597936, "grad_norm": 2.6876301765441895, "learning_rate": 1.798597584729256e-05, "loss": 2.2122, "step": 1539 }, { "epoch": 0.3599076860155999, "grad_norm": 2.6992197036743164, "learning_rate": 1.7997662641215425e-05, "loss": 2.3232, "step": 1540 }, { "epoch": 0.3601413923052204, "grad_norm": 3.9674527645111084, "learning_rate": 1.8009349435138292e-05, "loss": 2.4887, "step": 1541 }, { "epoch": 0.36037509859484096, "grad_norm": 2.5024924278259277, "learning_rate": 1.8021036229061162e-05, "loss": 2.2029, "step": 1542 }, { "epoch": 0.36060880488446145, "grad_norm": 2.6637680530548096, "learning_rate": 1.803272302298403e-05, "loss": 2.2832, "step": 1543 }, { "epoch": 0.36084251117408195, "grad_norm": 2.5790493488311768, "learning_rate": 1.8044409816906897e-05, "loss": 2.3826, "step": 1544 }, { "epoch": 0.3610762174637025, "grad_norm": 2.344949245452881, "learning_rate": 1.8056096610829764e-05, "loss": 2.3239, "step": 1545 }, { "epoch": 0.361309923753323, "grad_norm": 2.495091438293457, "learning_rate": 1.806778340475263e-05, "loss": 2.186, "step": 1546 }, { "epoch": 0.36154363004294354, "grad_norm": 2.734325885772705, "learning_rate": 1.8079470198675498e-05, "loss": 2.1871, "step": 1547 }, { "epoch": 0.36177733633256404, "grad_norm": 2.549595832824707, "learning_rate": 1.8091156992598365e-05, "loss": 2.2078, "step": 1548 }, { "epoch": 0.3620110426221846, "grad_norm": 2.1654863357543945, "learning_rate": 1.8102843786521232e-05, "loss": 2.2943, "step": 1549 }, { "epoch": 0.3622447489118051, "grad_norm": 2.566162586212158, "learning_rate": 1.81145305804441e-05, "loss": 2.1412, "step": 1550 }, { "epoch": 0.36247845520142563, "grad_norm": 2.399627208709717, "learning_rate": 1.8126217374366966e-05, "loss": 2.3063, "step": 1551 }, { "epoch": 0.3627121614910461, "grad_norm": 2.4734272956848145, "learning_rate": 1.8137904168289833e-05, "loss": 2.3352, "step": 1552 }, { "epoch": 0.3629458677806666, "grad_norm": 2.2894725799560547, "learning_rate": 1.81495909622127e-05, "loss": 2.3087, "step": 1553 }, { "epoch": 0.36317957407028717, "grad_norm": 2.49192476272583, "learning_rate": 1.8161277756135567e-05, "loss": 2.3076, "step": 1554 }, { "epoch": 0.36341328035990766, "grad_norm": 2.816631317138672, "learning_rate": 1.8172964550058434e-05, "loss": 2.2435, "step": 1555 }, { "epoch": 0.3636469866495282, "grad_norm": 2.8430070877075195, "learning_rate": 1.81846513439813e-05, "loss": 2.2738, "step": 1556 }, { "epoch": 0.3638806929391487, "grad_norm": 2.5051329135894775, "learning_rate": 1.8196338137904168e-05, "loss": 2.2274, "step": 1557 }, { "epoch": 0.36411439922876926, "grad_norm": 2.703772783279419, "learning_rate": 1.8208024931827035e-05, "loss": 2.084, "step": 1558 }, { "epoch": 0.36434810551838975, "grad_norm": 2.432318925857544, "learning_rate": 1.8219711725749906e-05, "loss": 2.1935, "step": 1559 }, { "epoch": 0.3645818118080103, "grad_norm": 3.804211139678955, "learning_rate": 1.823139851967277e-05, "loss": 2.1931, "step": 1560 }, { "epoch": 0.3648155180976308, "grad_norm": 2.6605684757232666, "learning_rate": 1.8243085313595636e-05, "loss": 2.3495, "step": 1561 }, { "epoch": 0.36504922438725135, "grad_norm": 2.810617685317993, "learning_rate": 1.8254772107518507e-05, "loss": 2.2324, "step": 1562 }, { "epoch": 0.36528293067687184, "grad_norm": 4.046947002410889, "learning_rate": 1.826645890144137e-05, "loss": 2.3727, "step": 1563 }, { "epoch": 0.36551663696649234, "grad_norm": 2.3859875202178955, "learning_rate": 1.8278145695364238e-05, "loss": 2.1031, "step": 1564 }, { "epoch": 0.3657503432561129, "grad_norm": 4.069711685180664, "learning_rate": 1.8289832489287108e-05, "loss": 2.4649, "step": 1565 }, { "epoch": 0.3659840495457334, "grad_norm": 2.887402296066284, "learning_rate": 1.830151928320997e-05, "loss": 2.2608, "step": 1566 }, { "epoch": 0.36621775583535393, "grad_norm": 2.922914981842041, "learning_rate": 1.831320607713284e-05, "loss": 2.3911, "step": 1567 }, { "epoch": 0.3664514621249744, "grad_norm": 2.457986354827881, "learning_rate": 1.832489287105571e-05, "loss": 2.354, "step": 1568 }, { "epoch": 0.366685168414595, "grad_norm": 3.4330968856811523, "learning_rate": 1.8336579664978573e-05, "loss": 2.3667, "step": 1569 }, { "epoch": 0.36691887470421547, "grad_norm": 2.5094902515411377, "learning_rate": 1.834826645890144e-05, "loss": 2.3008, "step": 1570 }, { "epoch": 0.367152580993836, "grad_norm": 3.0011024475097656, "learning_rate": 1.835995325282431e-05, "loss": 2.0759, "step": 1571 }, { "epoch": 0.3673862872834565, "grad_norm": 3.078911304473877, "learning_rate": 1.8371640046747174e-05, "loss": 2.2035, "step": 1572 }, { "epoch": 0.367619993573077, "grad_norm": 2.950028896331787, "learning_rate": 1.8383326840670044e-05, "loss": 2.3028, "step": 1573 }, { "epoch": 0.36785369986269756, "grad_norm": 3.331878900527954, "learning_rate": 1.839501363459291e-05, "loss": 2.2995, "step": 1574 }, { "epoch": 0.36808740615231805, "grad_norm": 3.0512709617614746, "learning_rate": 1.840670042851578e-05, "loss": 2.1782, "step": 1575 }, { "epoch": 0.3683211124419386, "grad_norm": 2.5685715675354004, "learning_rate": 1.8418387222438645e-05, "loss": 2.3739, "step": 1576 }, { "epoch": 0.3685548187315591, "grad_norm": 2.85890793800354, "learning_rate": 1.8430074016361513e-05, "loss": 2.2704, "step": 1577 }, { "epoch": 0.36878852502117965, "grad_norm": 3.763599157333374, "learning_rate": 1.844176081028438e-05, "loss": 2.1166, "step": 1578 }, { "epoch": 0.36902223131080014, "grad_norm": 2.9239554405212402, "learning_rate": 1.8453447604207247e-05, "loss": 2.1514, "step": 1579 }, { "epoch": 0.3692559376004207, "grad_norm": 3.677985191345215, "learning_rate": 1.8465134398130114e-05, "loss": 2.2598, "step": 1580 }, { "epoch": 0.3694896438900412, "grad_norm": 2.780100107192993, "learning_rate": 1.847682119205298e-05, "loss": 2.1405, "step": 1581 }, { "epoch": 0.36972335017966174, "grad_norm": 2.546570062637329, "learning_rate": 1.8488507985975848e-05, "loss": 2.4232, "step": 1582 }, { "epoch": 0.36995705646928223, "grad_norm": 2.5448477268218994, "learning_rate": 1.8500194779898715e-05, "loss": 2.2143, "step": 1583 }, { "epoch": 0.3701907627589027, "grad_norm": 3.2392921447753906, "learning_rate": 1.8511881573821582e-05, "loss": 2.2663, "step": 1584 }, { "epoch": 0.3704244690485233, "grad_norm": 2.8229191303253174, "learning_rate": 1.852356836774445e-05, "loss": 2.1454, "step": 1585 }, { "epoch": 0.37065817533814377, "grad_norm": 6.921701431274414, "learning_rate": 1.8535255161667316e-05, "loss": 2.2784, "step": 1586 }, { "epoch": 0.3708918816277643, "grad_norm": 3.8775174617767334, "learning_rate": 1.8546941955590183e-05, "loss": 2.3381, "step": 1587 }, { "epoch": 0.3711255879173848, "grad_norm": 2.5520191192626953, "learning_rate": 1.855862874951305e-05, "loss": 2.4026, "step": 1588 }, { "epoch": 0.37135929420700536, "grad_norm": 4.276261329650879, "learning_rate": 1.8570315543435917e-05, "loss": 2.1469, "step": 1589 }, { "epoch": 0.37159300049662586, "grad_norm": 3.336637258529663, "learning_rate": 1.8582002337358784e-05, "loss": 2.3285, "step": 1590 }, { "epoch": 0.3718267067862464, "grad_norm": 2.9612581729888916, "learning_rate": 1.8593689131281655e-05, "loss": 2.387, "step": 1591 }, { "epoch": 0.3720604130758669, "grad_norm": 2.555177688598633, "learning_rate": 1.8605375925204518e-05, "loss": 2.1931, "step": 1592 }, { "epoch": 0.3722941193654874, "grad_norm": 3.7833240032196045, "learning_rate": 1.8617062719127385e-05, "loss": 2.1723, "step": 1593 }, { "epoch": 0.37252782565510795, "grad_norm": 3.475759506225586, "learning_rate": 1.8628749513050256e-05, "loss": 2.2645, "step": 1594 }, { "epoch": 0.37276153194472844, "grad_norm": 2.7209858894348145, "learning_rate": 1.864043630697312e-05, "loss": 2.4037, "step": 1595 }, { "epoch": 0.372995238234349, "grad_norm": 4.347095489501953, "learning_rate": 1.8652123100895986e-05, "loss": 2.2423, "step": 1596 }, { "epoch": 0.3732289445239695, "grad_norm": 2.6195452213287354, "learning_rate": 1.8663809894818857e-05, "loss": 2.4819, "step": 1597 }, { "epoch": 0.37346265081359004, "grad_norm": 2.9487316608428955, "learning_rate": 1.867549668874172e-05, "loss": 2.3066, "step": 1598 }, { "epoch": 0.37369635710321053, "grad_norm": 2.54582142829895, "learning_rate": 1.868718348266459e-05, "loss": 2.22, "step": 1599 }, { "epoch": 0.3739300633928311, "grad_norm": 3.1158885955810547, "learning_rate": 1.8698870276587458e-05, "loss": 2.2895, "step": 1600 }, { "epoch": 0.3741637696824516, "grad_norm": 2.577423334121704, "learning_rate": 1.871055707051032e-05, "loss": 2.2224, "step": 1601 }, { "epoch": 0.3743974759720721, "grad_norm": 3.1754541397094727, "learning_rate": 1.8722243864433192e-05, "loss": 2.1581, "step": 1602 }, { "epoch": 0.3746311822616926, "grad_norm": 3.7751686573028564, "learning_rate": 1.873393065835606e-05, "loss": 2.2786, "step": 1603 }, { "epoch": 0.3748648885513131, "grad_norm": 2.738820791244507, "learning_rate": 1.8745617452278923e-05, "loss": 2.3621, "step": 1604 }, { "epoch": 0.37509859484093366, "grad_norm": 3.815706729888916, "learning_rate": 1.8757304246201793e-05, "loss": 2.2009, "step": 1605 }, { "epoch": 0.37533230113055416, "grad_norm": 2.498344659805298, "learning_rate": 1.876899104012466e-05, "loss": 2.1575, "step": 1606 }, { "epoch": 0.3755660074201747, "grad_norm": 2.9437344074249268, "learning_rate": 1.8780677834047527e-05, "loss": 2.3581, "step": 1607 }, { "epoch": 0.3757997137097952, "grad_norm": 2.9395675659179688, "learning_rate": 1.8792364627970394e-05, "loss": 2.2805, "step": 1608 }, { "epoch": 0.37603341999941575, "grad_norm": 3.153411865234375, "learning_rate": 1.880405142189326e-05, "loss": 2.2842, "step": 1609 }, { "epoch": 0.37626712628903625, "grad_norm": 2.4842371940612793, "learning_rate": 1.881573821581613e-05, "loss": 2.3166, "step": 1610 }, { "epoch": 0.3765008325786568, "grad_norm": 2.952996253967285, "learning_rate": 1.8827425009738996e-05, "loss": 2.1317, "step": 1611 }, { "epoch": 0.3767345388682773, "grad_norm": 3.6180169582366943, "learning_rate": 1.8839111803661863e-05, "loss": 2.3152, "step": 1612 }, { "epoch": 0.3769682451578978, "grad_norm": 2.682856798171997, "learning_rate": 1.885079859758473e-05, "loss": 2.1162, "step": 1613 }, { "epoch": 0.37720195144751834, "grad_norm": 2.3334410190582275, "learning_rate": 1.8862485391507597e-05, "loss": 2.245, "step": 1614 }, { "epoch": 0.37743565773713883, "grad_norm": 3.300863742828369, "learning_rate": 1.8874172185430464e-05, "loss": 2.326, "step": 1615 }, { "epoch": 0.3776693640267594, "grad_norm": 3.6240217685699463, "learning_rate": 1.888585897935333e-05, "loss": 2.2308, "step": 1616 }, { "epoch": 0.3779030703163799, "grad_norm": 2.532698154449463, "learning_rate": 1.8897545773276198e-05, "loss": 2.0909, "step": 1617 }, { "epoch": 0.3781367766060004, "grad_norm": 2.738300085067749, "learning_rate": 1.8909232567199065e-05, "loss": 2.1746, "step": 1618 }, { "epoch": 0.3783704828956209, "grad_norm": 3.024104356765747, "learning_rate": 1.8920919361121932e-05, "loss": 2.3173, "step": 1619 }, { "epoch": 0.37860418918524147, "grad_norm": 2.480562925338745, "learning_rate": 1.89326061550448e-05, "loss": 2.2827, "step": 1620 }, { "epoch": 0.37883789547486196, "grad_norm": 2.728797435760498, "learning_rate": 1.8944292948967666e-05, "loss": 2.264, "step": 1621 }, { "epoch": 0.3790716017644825, "grad_norm": 2.722193479537964, "learning_rate": 1.8955979742890533e-05, "loss": 2.3787, "step": 1622 }, { "epoch": 0.379305308054103, "grad_norm": 2.90635085105896, "learning_rate": 1.8967666536813403e-05, "loss": 2.462, "step": 1623 }, { "epoch": 0.3795390143437235, "grad_norm": 2.7500782012939453, "learning_rate": 1.8979353330736267e-05, "loss": 2.3447, "step": 1624 }, { "epoch": 0.37977272063334405, "grad_norm": 4.458848476409912, "learning_rate": 1.8991040124659138e-05, "loss": 2.4567, "step": 1625 }, { "epoch": 0.38000642692296455, "grad_norm": 4.164292812347412, "learning_rate": 1.9002726918582005e-05, "loss": 2.3224, "step": 1626 }, { "epoch": 0.3802401332125851, "grad_norm": 2.5836124420166016, "learning_rate": 1.9014413712504868e-05, "loss": 2.2155, "step": 1627 }, { "epoch": 0.3804738395022056, "grad_norm": 2.9103660583496094, "learning_rate": 1.902610050642774e-05, "loss": 2.3221, "step": 1628 }, { "epoch": 0.38070754579182614, "grad_norm": 3.5016121864318848, "learning_rate": 1.9037787300350606e-05, "loss": 2.1352, "step": 1629 }, { "epoch": 0.38094125208144664, "grad_norm": 2.3652334213256836, "learning_rate": 1.904947409427347e-05, "loss": 2.172, "step": 1630 }, { "epoch": 0.3811749583710672, "grad_norm": 2.934231996536255, "learning_rate": 1.906116088819634e-05, "loss": 2.2861, "step": 1631 }, { "epoch": 0.3814086646606877, "grad_norm": 2.813208818435669, "learning_rate": 1.9072847682119207e-05, "loss": 2.2614, "step": 1632 }, { "epoch": 0.3816423709503082, "grad_norm": 2.45536732673645, "learning_rate": 1.908453447604207e-05, "loss": 2.1229, "step": 1633 }, { "epoch": 0.3818760772399287, "grad_norm": 2.5868945121765137, "learning_rate": 1.909622126996494e-05, "loss": 2.191, "step": 1634 }, { "epoch": 0.3821097835295492, "grad_norm": 3.0425076484680176, "learning_rate": 1.9107908063887808e-05, "loss": 2.3147, "step": 1635 }, { "epoch": 0.38234348981916977, "grad_norm": 2.746837854385376, "learning_rate": 1.9119594857810675e-05, "loss": 2.022, "step": 1636 }, { "epoch": 0.38257719610879026, "grad_norm": 2.9763400554656982, "learning_rate": 1.9131281651733542e-05, "loss": 2.3522, "step": 1637 }, { "epoch": 0.3828109023984108, "grad_norm": 2.609694242477417, "learning_rate": 1.914296844565641e-05, "loss": 2.4095, "step": 1638 }, { "epoch": 0.3830446086880313, "grad_norm": 4.8327226638793945, "learning_rate": 1.9154655239579276e-05, "loss": 2.1391, "step": 1639 }, { "epoch": 0.38327831497765186, "grad_norm": 3.373361349105835, "learning_rate": 1.9166342033502143e-05, "loss": 2.1482, "step": 1640 }, { "epoch": 0.38351202126727235, "grad_norm": 2.7832565307617188, "learning_rate": 1.917802882742501e-05, "loss": 2.318, "step": 1641 }, { "epoch": 0.3837457275568929, "grad_norm": 3.369396448135376, "learning_rate": 1.9189715621347877e-05, "loss": 2.1941, "step": 1642 }, { "epoch": 0.3839794338465134, "grad_norm": 5.624741077423096, "learning_rate": 1.9201402415270744e-05, "loss": 2.2569, "step": 1643 }, { "epoch": 0.3842131401361339, "grad_norm": 3.8083744049072266, "learning_rate": 1.921308920919361e-05, "loss": 2.2967, "step": 1644 }, { "epoch": 0.38444684642575444, "grad_norm": 2.711231231689453, "learning_rate": 1.922477600311648e-05, "loss": 2.2581, "step": 1645 }, { "epoch": 0.38468055271537493, "grad_norm": 2.4093472957611084, "learning_rate": 1.9236462797039346e-05, "loss": 2.3886, "step": 1646 }, { "epoch": 0.3849142590049955, "grad_norm": 3.8755903244018555, "learning_rate": 1.9248149590962213e-05, "loss": 2.2148, "step": 1647 }, { "epoch": 0.385147965294616, "grad_norm": 2.400176763534546, "learning_rate": 1.925983638488508e-05, "loss": 2.2752, "step": 1648 }, { "epoch": 0.38538167158423653, "grad_norm": 3.4974827766418457, "learning_rate": 1.9271523178807947e-05, "loss": 2.3095, "step": 1649 }, { "epoch": 0.385615377873857, "grad_norm": 2.5755484104156494, "learning_rate": 1.9283209972730814e-05, "loss": 2.3254, "step": 1650 }, { "epoch": 0.3858490841634776, "grad_norm": 3.608700752258301, "learning_rate": 1.929489676665368e-05, "loss": 2.2354, "step": 1651 }, { "epoch": 0.38608279045309807, "grad_norm": 2.8424830436706543, "learning_rate": 1.930658356057655e-05, "loss": 2.3052, "step": 1652 }, { "epoch": 0.38631649674271856, "grad_norm": 2.599135160446167, "learning_rate": 1.9318270354499415e-05, "loss": 2.2118, "step": 1653 }, { "epoch": 0.3865502030323391, "grad_norm": 2.3573176860809326, "learning_rate": 1.9329957148422285e-05, "loss": 2.1702, "step": 1654 }, { "epoch": 0.3867839093219596, "grad_norm": 2.641078233718872, "learning_rate": 1.9341643942345152e-05, "loss": 2.2197, "step": 1655 }, { "epoch": 0.38701761561158016, "grad_norm": 3.4408504962921143, "learning_rate": 1.9353330736268016e-05, "loss": 2.2721, "step": 1656 }, { "epoch": 0.38725132190120065, "grad_norm": 3.290590524673462, "learning_rate": 1.9365017530190886e-05, "loss": 2.2846, "step": 1657 }, { "epoch": 0.3874850281908212, "grad_norm": 2.862835168838501, "learning_rate": 1.9376704324113753e-05, "loss": 2.1783, "step": 1658 }, { "epoch": 0.3877187344804417, "grad_norm": 2.7554502487182617, "learning_rate": 1.9388391118036617e-05, "loss": 2.1506, "step": 1659 }, { "epoch": 0.38795244077006225, "grad_norm": 2.93245005607605, "learning_rate": 1.9400077911959488e-05, "loss": 2.3379, "step": 1660 }, { "epoch": 0.38818614705968274, "grad_norm": 2.467510461807251, "learning_rate": 1.9411764705882355e-05, "loss": 2.2036, "step": 1661 }, { "epoch": 0.3884198533493033, "grad_norm": 3.0846328735351562, "learning_rate": 1.9423451499805218e-05, "loss": 2.2945, "step": 1662 }, { "epoch": 0.3886535596389238, "grad_norm": 2.3872251510620117, "learning_rate": 1.943513829372809e-05, "loss": 2.2398, "step": 1663 }, { "epoch": 0.3888872659285443, "grad_norm": 2.489973545074463, "learning_rate": 1.9446825087650956e-05, "loss": 2.363, "step": 1664 }, { "epoch": 0.38912097221816483, "grad_norm": 2.774144172668457, "learning_rate": 1.945851188157382e-05, "loss": 2.3444, "step": 1665 }, { "epoch": 0.3893546785077853, "grad_norm": 3.5305938720703125, "learning_rate": 1.947019867549669e-05, "loss": 2.051, "step": 1666 }, { "epoch": 0.3895883847974059, "grad_norm": 2.385781764984131, "learning_rate": 1.9481885469419557e-05, "loss": 2.3612, "step": 1667 }, { "epoch": 0.38982209108702637, "grad_norm": 2.835238218307495, "learning_rate": 1.9493572263342424e-05, "loss": 2.2917, "step": 1668 }, { "epoch": 0.3900557973766469, "grad_norm": 2.483445405960083, "learning_rate": 1.950525905726529e-05, "loss": 2.2023, "step": 1669 }, { "epoch": 0.3902895036662674, "grad_norm": 2.5532610416412354, "learning_rate": 1.9516945851188158e-05, "loss": 2.277, "step": 1670 }, { "epoch": 0.39052320995588796, "grad_norm": 2.6013386249542236, "learning_rate": 1.9528632645111025e-05, "loss": 2.3742, "step": 1671 }, { "epoch": 0.39075691624550846, "grad_norm": 3.11236572265625, "learning_rate": 1.9540319439033892e-05, "loss": 2.2201, "step": 1672 }, { "epoch": 0.39099062253512895, "grad_norm": 3.958367109298706, "learning_rate": 1.955200623295676e-05, "loss": 2.3956, "step": 1673 }, { "epoch": 0.3912243288247495, "grad_norm": 2.924663543701172, "learning_rate": 1.9563693026879626e-05, "loss": 2.2662, "step": 1674 }, { "epoch": 0.39145803511437, "grad_norm": 2.4594147205352783, "learning_rate": 1.9575379820802493e-05, "loss": 2.2464, "step": 1675 }, { "epoch": 0.39169174140399055, "grad_norm": 2.267164945602417, "learning_rate": 1.958706661472536e-05, "loss": 2.2388, "step": 1676 }, { "epoch": 0.39192544769361104, "grad_norm": 3.335792303085327, "learning_rate": 1.9598753408648227e-05, "loss": 2.2227, "step": 1677 }, { "epoch": 0.3921591539832316, "grad_norm": 2.3891568183898926, "learning_rate": 1.9610440202571094e-05, "loss": 2.2357, "step": 1678 }, { "epoch": 0.3923928602728521, "grad_norm": 3.24432373046875, "learning_rate": 1.962212699649396e-05, "loss": 2.133, "step": 1679 }, { "epoch": 0.39262656656247263, "grad_norm": 2.5881803035736084, "learning_rate": 1.9633813790416832e-05, "loss": 2.0874, "step": 1680 }, { "epoch": 0.39286027285209313, "grad_norm": 2.8288302421569824, "learning_rate": 1.9645500584339696e-05, "loss": 2.0618, "step": 1681 }, { "epoch": 0.3930939791417137, "grad_norm": 3.1557810306549072, "learning_rate": 1.9657187378262563e-05, "loss": 2.2373, "step": 1682 }, { "epoch": 0.3933276854313342, "grad_norm": 2.541653871536255, "learning_rate": 1.9668874172185433e-05, "loss": 2.2639, "step": 1683 }, { "epoch": 0.39356139172095467, "grad_norm": 4.1180806159973145, "learning_rate": 1.96805609661083e-05, "loss": 2.1961, "step": 1684 }, { "epoch": 0.3937950980105752, "grad_norm": 3.60068941116333, "learning_rate": 1.9692247760031164e-05, "loss": 2.2794, "step": 1685 }, { "epoch": 0.3940288043001957, "grad_norm": 2.6843395233154297, "learning_rate": 1.9703934553954034e-05, "loss": 2.2645, "step": 1686 }, { "epoch": 0.39426251058981626, "grad_norm": 2.4972751140594482, "learning_rate": 1.97156213478769e-05, "loss": 2.1679, "step": 1687 }, { "epoch": 0.39449621687943676, "grad_norm": 3.3874101638793945, "learning_rate": 1.9727308141799765e-05, "loss": 2.2092, "step": 1688 }, { "epoch": 0.3947299231690573, "grad_norm": 2.72406005859375, "learning_rate": 1.9738994935722635e-05, "loss": 2.4237, "step": 1689 }, { "epoch": 0.3949636294586778, "grad_norm": 2.8997254371643066, "learning_rate": 1.9750681729645502e-05, "loss": 2.3388, "step": 1690 }, { "epoch": 0.39519733574829835, "grad_norm": 3.8124876022338867, "learning_rate": 1.9762368523568366e-05, "loss": 2.1952, "step": 1691 }, { "epoch": 0.39543104203791885, "grad_norm": 3.5125608444213867, "learning_rate": 1.9774055317491236e-05, "loss": 2.2861, "step": 1692 }, { "epoch": 0.39566474832753934, "grad_norm": 3.324810743331909, "learning_rate": 1.9785742111414104e-05, "loss": 2.1304, "step": 1693 }, { "epoch": 0.3958984546171599, "grad_norm": 5.894842624664307, "learning_rate": 1.9797428905336967e-05, "loss": 2.3239, "step": 1694 }, { "epoch": 0.3961321609067804, "grad_norm": 4.007997512817383, "learning_rate": 1.9809115699259838e-05, "loss": 2.2566, "step": 1695 }, { "epoch": 0.39636586719640093, "grad_norm": 3.0880353450775146, "learning_rate": 1.9820802493182705e-05, "loss": 2.2509, "step": 1696 }, { "epoch": 0.39659957348602143, "grad_norm": 3.8782668113708496, "learning_rate": 1.983248928710557e-05, "loss": 2.2674, "step": 1697 }, { "epoch": 0.396833279775642, "grad_norm": 2.836682081222534, "learning_rate": 1.984417608102844e-05, "loss": 2.3031, "step": 1698 }, { "epoch": 0.3970669860652625, "grad_norm": 2.2341432571411133, "learning_rate": 1.9855862874951306e-05, "loss": 2.2116, "step": 1699 }, { "epoch": 0.397300692354883, "grad_norm": 2.2509653568267822, "learning_rate": 1.9867549668874173e-05, "loss": 2.0819, "step": 1700 }, { "epoch": 0.3975343986445035, "grad_norm": 3.2661352157592773, "learning_rate": 1.987923646279704e-05, "loss": 2.2616, "step": 1701 }, { "epoch": 0.39776810493412407, "grad_norm": 2.9777262210845947, "learning_rate": 1.9890923256719907e-05, "loss": 2.1468, "step": 1702 }, { "epoch": 0.39800181122374456, "grad_norm": 2.862814426422119, "learning_rate": 1.9902610050642774e-05, "loss": 2.1824, "step": 1703 }, { "epoch": 0.39823551751336506, "grad_norm": 3.5084779262542725, "learning_rate": 1.991429684456564e-05, "loss": 2.2041, "step": 1704 }, { "epoch": 0.3984692238029856, "grad_norm": 3.4468185901641846, "learning_rate": 1.9925983638488508e-05, "loss": 2.2061, "step": 1705 }, { "epoch": 0.3987029300926061, "grad_norm": 3.1630094051361084, "learning_rate": 1.9937670432411375e-05, "loss": 2.4288, "step": 1706 }, { "epoch": 0.39893663638222665, "grad_norm": 4.187211990356445, "learning_rate": 1.9949357226334242e-05, "loss": 2.242, "step": 1707 }, { "epoch": 0.39917034267184714, "grad_norm": 2.688222646713257, "learning_rate": 1.996104402025711e-05, "loss": 2.151, "step": 1708 }, { "epoch": 0.3994040489614677, "grad_norm": 3.2538034915924072, "learning_rate": 1.997273081417998e-05, "loss": 2.2065, "step": 1709 }, { "epoch": 0.3996377552510882, "grad_norm": 2.903949499130249, "learning_rate": 1.9984417608102843e-05, "loss": 2.2441, "step": 1710 }, { "epoch": 0.39987146154070874, "grad_norm": 2.9958813190460205, "learning_rate": 1.999610440202571e-05, "loss": 2.2297, "step": 1711 }, { "epoch": 0.40010516783032923, "grad_norm": 2.912386655807495, "learning_rate": 2.000779119594858e-05, "loss": 2.1494, "step": 1712 }, { "epoch": 0.40033887411994973, "grad_norm": 2.922437906265259, "learning_rate": 2.0019477989871444e-05, "loss": 2.2928, "step": 1713 }, { "epoch": 0.4005725804095703, "grad_norm": 3.0066215991973877, "learning_rate": 2.003116478379431e-05, "loss": 2.4795, "step": 1714 }, { "epoch": 0.4008062866991908, "grad_norm": 2.7229812145233154, "learning_rate": 2.0042851577717182e-05, "loss": 2.3062, "step": 1715 }, { "epoch": 0.4010399929888113, "grad_norm": 4.3844733238220215, "learning_rate": 2.005453837164005e-05, "loss": 2.1817, "step": 1716 }, { "epoch": 0.4012736992784318, "grad_norm": 2.9256279468536377, "learning_rate": 2.0066225165562913e-05, "loss": 2.2878, "step": 1717 }, { "epoch": 0.40150740556805237, "grad_norm": 2.5990025997161865, "learning_rate": 2.0077911959485783e-05, "loss": 2.4015, "step": 1718 }, { "epoch": 0.40174111185767286, "grad_norm": 3.0485355854034424, "learning_rate": 2.008959875340865e-05, "loss": 2.2334, "step": 1719 }, { "epoch": 0.4019748181472934, "grad_norm": 3.748810291290283, "learning_rate": 2.0101285547331514e-05, "loss": 2.1625, "step": 1720 }, { "epoch": 0.4022085244369139, "grad_norm": 2.9118120670318604, "learning_rate": 2.0112972341254384e-05, "loss": 2.2398, "step": 1721 }, { "epoch": 0.40244223072653446, "grad_norm": 3.047092914581299, "learning_rate": 2.012465913517725e-05, "loss": 2.2153, "step": 1722 }, { "epoch": 0.40267593701615495, "grad_norm": 3.1205809116363525, "learning_rate": 2.0136345929100115e-05, "loss": 2.1217, "step": 1723 }, { "epoch": 0.40290964330577544, "grad_norm": 2.897765636444092, "learning_rate": 2.0148032723022985e-05, "loss": 2.4046, "step": 1724 }, { "epoch": 0.403143349595396, "grad_norm": 4.260271072387695, "learning_rate": 2.0159719516945852e-05, "loss": 2.2624, "step": 1725 }, { "epoch": 0.4033770558850165, "grad_norm": 2.3374242782592773, "learning_rate": 2.0171406310868716e-05, "loss": 2.2185, "step": 1726 }, { "epoch": 0.40361076217463704, "grad_norm": 2.8095736503601074, "learning_rate": 2.0183093104791587e-05, "loss": 2.2081, "step": 1727 }, { "epoch": 0.40384446846425753, "grad_norm": 3.399904727935791, "learning_rate": 2.0194779898714454e-05, "loss": 2.2619, "step": 1728 }, { "epoch": 0.4040781747538781, "grad_norm": 3.952247381210327, "learning_rate": 2.0206466692637317e-05, "loss": 2.4182, "step": 1729 }, { "epoch": 0.4043118810434986, "grad_norm": 3.0013701915740967, "learning_rate": 2.0218153486560188e-05, "loss": 2.397, "step": 1730 }, { "epoch": 0.4045455873331191, "grad_norm": 2.540700674057007, "learning_rate": 2.0229840280483055e-05, "loss": 2.1654, "step": 1731 }, { "epoch": 0.4047792936227396, "grad_norm": 2.5160763263702393, "learning_rate": 2.0241527074405922e-05, "loss": 2.4739, "step": 1732 }, { "epoch": 0.4050129999123601, "grad_norm": 2.5366134643554688, "learning_rate": 2.025321386832879e-05, "loss": 2.1861, "step": 1733 }, { "epoch": 0.40524670620198067, "grad_norm": 3.8280606269836426, "learning_rate": 2.0264900662251656e-05, "loss": 2.1927, "step": 1734 }, { "epoch": 0.40548041249160116, "grad_norm": 2.4586246013641357, "learning_rate": 2.0276587456174526e-05, "loss": 2.3042, "step": 1735 }, { "epoch": 0.4057141187812217, "grad_norm": 3.155426263809204, "learning_rate": 2.028827425009739e-05, "loss": 2.2703, "step": 1736 }, { "epoch": 0.4059478250708422, "grad_norm": 3.325408458709717, "learning_rate": 2.0299961044020257e-05, "loss": 2.1758, "step": 1737 }, { "epoch": 0.40618153136046276, "grad_norm": 2.474064588546753, "learning_rate": 2.0311647837943127e-05, "loss": 2.3982, "step": 1738 }, { "epoch": 0.40641523765008325, "grad_norm": 2.6662850379943848, "learning_rate": 2.032333463186599e-05, "loss": 2.2742, "step": 1739 }, { "epoch": 0.4066489439397038, "grad_norm": 3.2467055320739746, "learning_rate": 2.0335021425788858e-05, "loss": 2.162, "step": 1740 }, { "epoch": 0.4068826502293243, "grad_norm": 2.631587266921997, "learning_rate": 2.034670821971173e-05, "loss": 2.2979, "step": 1741 }, { "epoch": 0.4071163565189448, "grad_norm": 2.494954824447632, "learning_rate": 2.0358395013634592e-05, "loss": 2.249, "step": 1742 }, { "epoch": 0.40735006280856534, "grad_norm": 3.3449084758758545, "learning_rate": 2.037008180755746e-05, "loss": 2.259, "step": 1743 }, { "epoch": 0.40758376909818583, "grad_norm": 2.478496551513672, "learning_rate": 2.038176860148033e-05, "loss": 2.195, "step": 1744 }, { "epoch": 0.4078174753878064, "grad_norm": 2.945042610168457, "learning_rate": 2.0393455395403193e-05, "loss": 2.4121, "step": 1745 }, { "epoch": 0.4080511816774269, "grad_norm": 2.608442783355713, "learning_rate": 2.040514218932606e-05, "loss": 2.2172, "step": 1746 }, { "epoch": 0.4082848879670474, "grad_norm": 3.509392499923706, "learning_rate": 2.041682898324893e-05, "loss": 2.0934, "step": 1747 }, { "epoch": 0.4085185942566679, "grad_norm": 2.6957411766052246, "learning_rate": 2.0428515777171798e-05, "loss": 2.4313, "step": 1748 }, { "epoch": 0.40875230054628847, "grad_norm": 2.795567035675049, "learning_rate": 2.044020257109466e-05, "loss": 2.0283, "step": 1749 }, { "epoch": 0.40898600683590897, "grad_norm": 2.7247445583343506, "learning_rate": 2.0451889365017532e-05, "loss": 2.2972, "step": 1750 }, { "epoch": 0.4092197131255295, "grad_norm": 3.4571239948272705, "learning_rate": 2.04635761589404e-05, "loss": 2.2806, "step": 1751 }, { "epoch": 0.40945341941515, "grad_norm": 2.4788930416107178, "learning_rate": 2.0475262952863263e-05, "loss": 2.2699, "step": 1752 }, { "epoch": 0.4096871257047705, "grad_norm": 2.535564661026001, "learning_rate": 2.0486949746786133e-05, "loss": 2.2333, "step": 1753 }, { "epoch": 0.40992083199439106, "grad_norm": 2.760268211364746, "learning_rate": 2.0498636540709e-05, "loss": 2.1735, "step": 1754 }, { "epoch": 0.41015453828401155, "grad_norm": 2.6688430309295654, "learning_rate": 2.0510323334631864e-05, "loss": 2.3855, "step": 1755 }, { "epoch": 0.4103882445736321, "grad_norm": 2.578131675720215, "learning_rate": 2.0522010128554734e-05, "loss": 2.1822, "step": 1756 }, { "epoch": 0.4106219508632526, "grad_norm": 3.337296962738037, "learning_rate": 2.05336969224776e-05, "loss": 2.0692, "step": 1757 }, { "epoch": 0.41085565715287314, "grad_norm": 2.628603935241699, "learning_rate": 2.0545383716400465e-05, "loss": 2.2422, "step": 1758 }, { "epoch": 0.41108936344249364, "grad_norm": 3.5080888271331787, "learning_rate": 2.0557070510323335e-05, "loss": 2.225, "step": 1759 }, { "epoch": 0.4113230697321142, "grad_norm": 3.3690171241760254, "learning_rate": 2.0568757304246202e-05, "loss": 2.2826, "step": 1760 }, { "epoch": 0.4115567760217347, "grad_norm": 2.7393958568573, "learning_rate": 2.058044409816907e-05, "loss": 2.1676, "step": 1761 }, { "epoch": 0.4117904823113552, "grad_norm": 2.6133720874786377, "learning_rate": 2.0592130892091937e-05, "loss": 2.1774, "step": 1762 }, { "epoch": 0.4120241886009757, "grad_norm": 3.1564385890960693, "learning_rate": 2.0603817686014804e-05, "loss": 2.1966, "step": 1763 }, { "epoch": 0.4122578948905962, "grad_norm": 2.341870069503784, "learning_rate": 2.0615504479937674e-05, "loss": 2.14, "step": 1764 }, { "epoch": 0.41249160118021677, "grad_norm": 3.9195399284362793, "learning_rate": 2.0627191273860538e-05, "loss": 2.3085, "step": 1765 }, { "epoch": 0.41272530746983727, "grad_norm": 4.6981940269470215, "learning_rate": 2.0638878067783405e-05, "loss": 2.1377, "step": 1766 }, { "epoch": 0.4129590137594578, "grad_norm": 3.0217175483703613, "learning_rate": 2.0650564861706275e-05, "loss": 2.3457, "step": 1767 }, { "epoch": 0.4131927200490783, "grad_norm": 3.6863443851470947, "learning_rate": 2.066225165562914e-05, "loss": 2.3419, "step": 1768 }, { "epoch": 0.41342642633869886, "grad_norm": 6.072162628173828, "learning_rate": 2.0673938449552006e-05, "loss": 2.3119, "step": 1769 }, { "epoch": 0.41366013262831935, "grad_norm": 5.182234287261963, "learning_rate": 2.0685625243474876e-05, "loss": 2.4054, "step": 1770 }, { "epoch": 0.4138938389179399, "grad_norm": 3.1476171016693115, "learning_rate": 2.069731203739774e-05, "loss": 2.1476, "step": 1771 }, { "epoch": 0.4141275452075604, "grad_norm": 4.844021320343018, "learning_rate": 2.0708998831320607e-05, "loss": 2.2203, "step": 1772 }, { "epoch": 0.4143612514971809, "grad_norm": 6.345508575439453, "learning_rate": 2.0720685625243477e-05, "loss": 2.1876, "step": 1773 }, { "epoch": 0.41459495778680144, "grad_norm": 5.475948333740234, "learning_rate": 2.073237241916634e-05, "loss": 2.2321, "step": 1774 }, { "epoch": 0.41482866407642194, "grad_norm": 2.217520236968994, "learning_rate": 2.0744059213089208e-05, "loss": 2.1696, "step": 1775 }, { "epoch": 0.4150623703660425, "grad_norm": 2.437932252883911, "learning_rate": 2.075574600701208e-05, "loss": 2.2329, "step": 1776 }, { "epoch": 0.415296076655663, "grad_norm": 3.363109588623047, "learning_rate": 2.0767432800934946e-05, "loss": 2.0879, "step": 1777 }, { "epoch": 0.41552978294528353, "grad_norm": 2.589115858078003, "learning_rate": 2.077911959485781e-05, "loss": 2.3231, "step": 1778 }, { "epoch": 0.415763489234904, "grad_norm": 2.349133253097534, "learning_rate": 2.079080638878068e-05, "loss": 2.1839, "step": 1779 }, { "epoch": 0.4159971955245246, "grad_norm": 2.2252578735351562, "learning_rate": 2.0802493182703547e-05, "loss": 2.2749, "step": 1780 }, { "epoch": 0.41623090181414507, "grad_norm": 2.4152281284332275, "learning_rate": 2.081417997662641e-05, "loss": 2.2227, "step": 1781 }, { "epoch": 0.41646460810376557, "grad_norm": 2.3015530109405518, "learning_rate": 2.082586677054928e-05, "loss": 2.2437, "step": 1782 }, { "epoch": 0.4166983143933861, "grad_norm": 2.469292402267456, "learning_rate": 2.0837553564472148e-05, "loss": 2.0407, "step": 1783 }, { "epoch": 0.4169320206830066, "grad_norm": 2.5074687004089355, "learning_rate": 2.084924035839501e-05, "loss": 2.0935, "step": 1784 }, { "epoch": 0.41716572697262716, "grad_norm": 4.1460676193237305, "learning_rate": 2.0860927152317882e-05, "loss": 2.311, "step": 1785 }, { "epoch": 0.41739943326224765, "grad_norm": 2.951481580734253, "learning_rate": 2.087261394624075e-05, "loss": 2.2395, "step": 1786 }, { "epoch": 0.4176331395518682, "grad_norm": 3.1482882499694824, "learning_rate": 2.0884300740163613e-05, "loss": 2.1736, "step": 1787 }, { "epoch": 0.4178668458414887, "grad_norm": 3.038201332092285, "learning_rate": 2.0895987534086483e-05, "loss": 2.2392, "step": 1788 }, { "epoch": 0.41810055213110925, "grad_norm": 3.3910598754882812, "learning_rate": 2.090767432800935e-05, "loss": 2.0943, "step": 1789 }, { "epoch": 0.41833425842072974, "grad_norm": 4.134871006011963, "learning_rate": 2.0919361121932217e-05, "loss": 2.2714, "step": 1790 }, { "epoch": 0.4185679647103503, "grad_norm": 2.5107617378234863, "learning_rate": 2.0931047915855084e-05, "loss": 2.2705, "step": 1791 }, { "epoch": 0.4188016709999708, "grad_norm": 4.4322004318237305, "learning_rate": 2.094273470977795e-05, "loss": 2.0605, "step": 1792 }, { "epoch": 0.4190353772895913, "grad_norm": 5.005577087402344, "learning_rate": 2.0954421503700822e-05, "loss": 2.1641, "step": 1793 }, { "epoch": 0.41926908357921183, "grad_norm": 2.642115592956543, "learning_rate": 2.0966108297623685e-05, "loss": 1.9537, "step": 1794 }, { "epoch": 0.4195027898688323, "grad_norm": 2.6922860145568848, "learning_rate": 2.0977795091546552e-05, "loss": 2.2521, "step": 1795 }, { "epoch": 0.4197364961584529, "grad_norm": 6.68074893951416, "learning_rate": 2.0989481885469423e-05, "loss": 2.1593, "step": 1796 }, { "epoch": 0.41997020244807337, "grad_norm": 4.836878299713135, "learning_rate": 2.1001168679392287e-05, "loss": 2.0914, "step": 1797 }, { "epoch": 0.4202039087376939, "grad_norm": 3.068497657775879, "learning_rate": 2.1012855473315154e-05, "loss": 2.2329, "step": 1798 }, { "epoch": 0.4204376150273144, "grad_norm": 3.789597988128662, "learning_rate": 2.1024542267238024e-05, "loss": 2.1904, "step": 1799 }, { "epoch": 0.42067132131693497, "grad_norm": 3.5978941917419434, "learning_rate": 2.1036229061160888e-05, "loss": 2.2594, "step": 1800 }, { "epoch": 0.42090502760655546, "grad_norm": 4.086334228515625, "learning_rate": 2.1047915855083755e-05, "loss": 2.309, "step": 1801 }, { "epoch": 0.42113873389617595, "grad_norm": 3.4311609268188477, "learning_rate": 2.1059602649006625e-05, "loss": 2.1472, "step": 1802 }, { "epoch": 0.4213724401857965, "grad_norm": 5.322911739349365, "learning_rate": 2.107128944292949e-05, "loss": 2.1661, "step": 1803 }, { "epoch": 0.421606146475417, "grad_norm": 5.446767807006836, "learning_rate": 2.1082976236852356e-05, "loss": 2.2906, "step": 1804 }, { "epoch": 0.42183985276503755, "grad_norm": 3.786336898803711, "learning_rate": 2.1094663030775226e-05, "loss": 2.3178, "step": 1805 }, { "epoch": 0.42207355905465804, "grad_norm": 2.339181423187256, "learning_rate": 2.110634982469809e-05, "loss": 2.2809, "step": 1806 }, { "epoch": 0.4223072653442786, "grad_norm": 4.112423419952393, "learning_rate": 2.1118036618620957e-05, "loss": 2.2058, "step": 1807 }, { "epoch": 0.4225409716338991, "grad_norm": 5.111412525177002, "learning_rate": 2.1129723412543827e-05, "loss": 2.2611, "step": 1808 }, { "epoch": 0.42277467792351964, "grad_norm": 3.4328200817108154, "learning_rate": 2.1141410206466695e-05, "loss": 2.1628, "step": 1809 }, { "epoch": 0.42300838421314013, "grad_norm": 4.463654518127441, "learning_rate": 2.1153097000389558e-05, "loss": 2.1539, "step": 1810 }, { "epoch": 0.4232420905027607, "grad_norm": 5.1102824211120605, "learning_rate": 2.116478379431243e-05, "loss": 2.272, "step": 1811 }, { "epoch": 0.4234757967923812, "grad_norm": 4.275919437408447, "learning_rate": 2.1176470588235296e-05, "loss": 2.058, "step": 1812 }, { "epoch": 0.42370950308200167, "grad_norm": 2.647378444671631, "learning_rate": 2.118815738215816e-05, "loss": 2.256, "step": 1813 }, { "epoch": 0.4239432093716222, "grad_norm": 3.3381876945495605, "learning_rate": 2.119984417608103e-05, "loss": 2.3396, "step": 1814 }, { "epoch": 0.4241769156612427, "grad_norm": 2.234646797180176, "learning_rate": 2.1211530970003897e-05, "loss": 2.1584, "step": 1815 }, { "epoch": 0.42441062195086326, "grad_norm": 2.762970447540283, "learning_rate": 2.1223217763926764e-05, "loss": 2.333, "step": 1816 }, { "epoch": 0.42464432824048376, "grad_norm": 2.5469446182250977, "learning_rate": 2.123490455784963e-05, "loss": 2.124, "step": 1817 }, { "epoch": 0.4248780345301043, "grad_norm": 4.005640983581543, "learning_rate": 2.1246591351772498e-05, "loss": 2.1934, "step": 1818 }, { "epoch": 0.4251117408197248, "grad_norm": 3.615823984146118, "learning_rate": 2.1258278145695365e-05, "loss": 2.2931, "step": 1819 }, { "epoch": 0.42534544710934535, "grad_norm": 3.9270546436309814, "learning_rate": 2.1269964939618232e-05, "loss": 2.1482, "step": 1820 }, { "epoch": 0.42557915339896585, "grad_norm": 6.004262447357178, "learning_rate": 2.12816517335411e-05, "loss": 2.2735, "step": 1821 }, { "epoch": 0.42581285968858634, "grad_norm": 2.8427653312683105, "learning_rate": 2.1293338527463966e-05, "loss": 2.2924, "step": 1822 }, { "epoch": 0.4260465659782069, "grad_norm": 3.5893962383270264, "learning_rate": 2.1305025321386833e-05, "loss": 2.3892, "step": 1823 }, { "epoch": 0.4262802722678274, "grad_norm": 2.582458257675171, "learning_rate": 2.13167121153097e-05, "loss": 2.1743, "step": 1824 }, { "epoch": 0.42651397855744794, "grad_norm": 2.663409471511841, "learning_rate": 2.132839890923257e-05, "loss": 2.2066, "step": 1825 }, { "epoch": 0.42674768484706843, "grad_norm": 2.646984338760376, "learning_rate": 2.1340085703155434e-05, "loss": 2.3637, "step": 1826 }, { "epoch": 0.426981391136689, "grad_norm": 2.3513104915618896, "learning_rate": 2.13517724970783e-05, "loss": 2.3091, "step": 1827 }, { "epoch": 0.4272150974263095, "grad_norm": 2.5237252712249756, "learning_rate": 2.1363459291001172e-05, "loss": 2.148, "step": 1828 }, { "epoch": 0.42744880371593, "grad_norm": 2.6149680614471436, "learning_rate": 2.1375146084924035e-05, "loss": 2.0079, "step": 1829 }, { "epoch": 0.4276825100055505, "grad_norm": 2.2871391773223877, "learning_rate": 2.1386832878846902e-05, "loss": 2.3028, "step": 1830 }, { "epoch": 0.42791621629517107, "grad_norm": 2.489077568054199, "learning_rate": 2.1398519672769773e-05, "loss": 2.2074, "step": 1831 }, { "epoch": 0.42814992258479156, "grad_norm": 2.595181465148926, "learning_rate": 2.1410206466692637e-05, "loss": 2.1339, "step": 1832 }, { "epoch": 0.42838362887441206, "grad_norm": 2.6180458068847656, "learning_rate": 2.1421893260615504e-05, "loss": 2.2226, "step": 1833 }, { "epoch": 0.4286173351640326, "grad_norm": 2.358647108078003, "learning_rate": 2.1433580054538374e-05, "loss": 2.135, "step": 1834 }, { "epoch": 0.4288510414536531, "grad_norm": 2.5909101963043213, "learning_rate": 2.1445266848461238e-05, "loss": 2.3494, "step": 1835 }, { "epoch": 0.42908474774327365, "grad_norm": 2.591754674911499, "learning_rate": 2.1456953642384105e-05, "loss": 2.1801, "step": 1836 }, { "epoch": 0.42931845403289415, "grad_norm": 2.9683399200439453, "learning_rate": 2.1468640436306975e-05, "loss": 2.2897, "step": 1837 }, { "epoch": 0.4295521603225147, "grad_norm": 3.6518337726593018, "learning_rate": 2.148032723022984e-05, "loss": 2.1763, "step": 1838 }, { "epoch": 0.4297858666121352, "grad_norm": 3.1660211086273193, "learning_rate": 2.1492014024152706e-05, "loss": 2.3077, "step": 1839 }, { "epoch": 0.43001957290175574, "grad_norm": 3.5366477966308594, "learning_rate": 2.1503700818075576e-05, "loss": 2.3012, "step": 1840 }, { "epoch": 0.43025327919137624, "grad_norm": 2.6486666202545166, "learning_rate": 2.1515387611998443e-05, "loss": 2.2002, "step": 1841 }, { "epoch": 0.43048698548099673, "grad_norm": 3.07226824760437, "learning_rate": 2.1527074405921307e-05, "loss": 2.1422, "step": 1842 }, { "epoch": 0.4307206917706173, "grad_norm": 2.8641581535339355, "learning_rate": 2.1538761199844177e-05, "loss": 2.0971, "step": 1843 }, { "epoch": 0.4309543980602378, "grad_norm": 2.68418288230896, "learning_rate": 2.1550447993767045e-05, "loss": 2.1905, "step": 1844 }, { "epoch": 0.4311881043498583, "grad_norm": 3.9012677669525146, "learning_rate": 2.156213478768991e-05, "loss": 2.2842, "step": 1845 }, { "epoch": 0.4314218106394788, "grad_norm": 2.538569688796997, "learning_rate": 2.157382158161278e-05, "loss": 2.1996, "step": 1846 }, { "epoch": 0.43165551692909937, "grad_norm": 2.4971024990081787, "learning_rate": 2.1585508375535646e-05, "loss": 2.1257, "step": 1847 }, { "epoch": 0.43188922321871986, "grad_norm": 3.522017002105713, "learning_rate": 2.1597195169458513e-05, "loss": 2.4112, "step": 1848 }, { "epoch": 0.4321229295083404, "grad_norm": 3.3988215923309326, "learning_rate": 2.160888196338138e-05, "loss": 2.1419, "step": 1849 }, { "epoch": 0.4323566357979609, "grad_norm": 2.5511953830718994, "learning_rate": 2.1620568757304247e-05, "loss": 2.1964, "step": 1850 }, { "epoch": 0.43259034208758146, "grad_norm": 3.124946355819702, "learning_rate": 2.1632255551227114e-05, "loss": 2.165, "step": 1851 }, { "epoch": 0.43282404837720195, "grad_norm": 2.684237241744995, "learning_rate": 2.164394234514998e-05, "loss": 2.1792, "step": 1852 }, { "epoch": 0.43305775466682245, "grad_norm": 2.659742593765259, "learning_rate": 2.1655629139072848e-05, "loss": 2.253, "step": 1853 }, { "epoch": 0.433291460956443, "grad_norm": 2.734706163406372, "learning_rate": 2.1667315932995715e-05, "loss": 2.1423, "step": 1854 }, { "epoch": 0.4335251672460635, "grad_norm": 2.2656590938568115, "learning_rate": 2.1679002726918582e-05, "loss": 2.2352, "step": 1855 }, { "epoch": 0.43375887353568404, "grad_norm": 3.908499002456665, "learning_rate": 2.169068952084145e-05, "loss": 2.1415, "step": 1856 }, { "epoch": 0.43399257982530454, "grad_norm": 2.437976598739624, "learning_rate": 2.170237631476432e-05, "loss": 2.1827, "step": 1857 }, { "epoch": 0.4342262861149251, "grad_norm": 2.480855941772461, "learning_rate": 2.1714063108687183e-05, "loss": 2.1494, "step": 1858 }, { "epoch": 0.4344599924045456, "grad_norm": 2.518155813217163, "learning_rate": 2.172574990261005e-05, "loss": 2.3519, "step": 1859 }, { "epoch": 0.43469369869416613, "grad_norm": 2.5638678073883057, "learning_rate": 2.173743669653292e-05, "loss": 2.154, "step": 1860 }, { "epoch": 0.4349274049837866, "grad_norm": 2.7160134315490723, "learning_rate": 2.1749123490455784e-05, "loss": 2.1488, "step": 1861 }, { "epoch": 0.4351611112734071, "grad_norm": 3.6549265384674072, "learning_rate": 2.176081028437865e-05, "loss": 2.2551, "step": 1862 }, { "epoch": 0.43539481756302767, "grad_norm": 2.4785799980163574, "learning_rate": 2.1772497078301522e-05, "loss": 2.3285, "step": 1863 }, { "epoch": 0.43562852385264816, "grad_norm": 2.9209864139556885, "learning_rate": 2.1784183872224385e-05, "loss": 2.2629, "step": 1864 }, { "epoch": 0.4358622301422687, "grad_norm": 3.881571054458618, "learning_rate": 2.1795870666147253e-05, "loss": 2.1501, "step": 1865 }, { "epoch": 0.4360959364318892, "grad_norm": 2.625666856765747, "learning_rate": 2.1807557460070123e-05, "loss": 2.1928, "step": 1866 }, { "epoch": 0.43632964272150976, "grad_norm": 4.40388298034668, "learning_rate": 2.1819244253992987e-05, "loss": 2.1834, "step": 1867 }, { "epoch": 0.43656334901113025, "grad_norm": 4.573095798492432, "learning_rate": 2.1830931047915854e-05, "loss": 2.2237, "step": 1868 }, { "epoch": 0.4367970553007508, "grad_norm": 3.420083999633789, "learning_rate": 2.1842617841838724e-05, "loss": 2.3288, "step": 1869 }, { "epoch": 0.4370307615903713, "grad_norm": 3.57715106010437, "learning_rate": 2.1854304635761588e-05, "loss": 2.2094, "step": 1870 }, { "epoch": 0.43726446787999185, "grad_norm": 2.758606433868408, "learning_rate": 2.1865991429684458e-05, "loss": 2.3192, "step": 1871 }, { "epoch": 0.43749817416961234, "grad_norm": 3.8590521812438965, "learning_rate": 2.1877678223607325e-05, "loss": 2.1927, "step": 1872 }, { "epoch": 0.43773188045923284, "grad_norm": 3.2787325382232666, "learning_rate": 2.1889365017530192e-05, "loss": 2.3504, "step": 1873 }, { "epoch": 0.4379655867488534, "grad_norm": 2.8428993225097656, "learning_rate": 2.190105181145306e-05, "loss": 2.2139, "step": 1874 }, { "epoch": 0.4381992930384739, "grad_norm": 2.576422691345215, "learning_rate": 2.1912738605375926e-05, "loss": 2.159, "step": 1875 }, { "epoch": 0.43843299932809443, "grad_norm": 3.856890916824341, "learning_rate": 2.1924425399298793e-05, "loss": 2.2421, "step": 1876 }, { "epoch": 0.4386667056177149, "grad_norm": 2.9476161003112793, "learning_rate": 2.193611219322166e-05, "loss": 2.278, "step": 1877 }, { "epoch": 0.4389004119073355, "grad_norm": 2.443061351776123, "learning_rate": 2.1947798987144528e-05, "loss": 2.0499, "step": 1878 }, { "epoch": 0.43913411819695597, "grad_norm": 3.3275203704833984, "learning_rate": 2.1959485781067395e-05, "loss": 2.1959, "step": 1879 }, { "epoch": 0.4393678244865765, "grad_norm": 2.72053861618042, "learning_rate": 2.197117257499026e-05, "loss": 2.2572, "step": 1880 }, { "epoch": 0.439601530776197, "grad_norm": 2.404989004135132, "learning_rate": 2.198285936891313e-05, "loss": 2.3074, "step": 1881 }, { "epoch": 0.4398352370658175, "grad_norm": 3.9661264419555664, "learning_rate": 2.1994546162835996e-05, "loss": 2.2, "step": 1882 }, { "epoch": 0.44006894335543806, "grad_norm": 3.4152066707611084, "learning_rate": 2.2006232956758863e-05, "loss": 2.3272, "step": 1883 }, { "epoch": 0.44030264964505855, "grad_norm": 3.0517780780792236, "learning_rate": 2.201791975068173e-05, "loss": 2.2363, "step": 1884 }, { "epoch": 0.4405363559346791, "grad_norm": 3.6660842895507812, "learning_rate": 2.2029606544604597e-05, "loss": 2.184, "step": 1885 }, { "epoch": 0.4407700622242996, "grad_norm": 3.8653509616851807, "learning_rate": 2.2041293338527464e-05, "loss": 2.1057, "step": 1886 }, { "epoch": 0.44100376851392015, "grad_norm": 2.7694954872131348, "learning_rate": 2.205298013245033e-05, "loss": 2.2304, "step": 1887 }, { "epoch": 0.44123747480354064, "grad_norm": 4.463608741760254, "learning_rate": 2.2064666926373198e-05, "loss": 2.1777, "step": 1888 }, { "epoch": 0.4414711810931612, "grad_norm": 4.638602256774902, "learning_rate": 2.207635372029607e-05, "loss": 2.2497, "step": 1889 }, { "epoch": 0.4417048873827817, "grad_norm": 5.182187080383301, "learning_rate": 2.2088040514218932e-05, "loss": 2.1138, "step": 1890 }, { "epoch": 0.44193859367240224, "grad_norm": 2.553220272064209, "learning_rate": 2.20997273081418e-05, "loss": 2.0993, "step": 1891 }, { "epoch": 0.44217229996202273, "grad_norm": 4.143064498901367, "learning_rate": 2.211141410206467e-05, "loss": 2.1931, "step": 1892 }, { "epoch": 0.4424060062516432, "grad_norm": 2.606637477874756, "learning_rate": 2.2123100895987533e-05, "loss": 2.2717, "step": 1893 }, { "epoch": 0.4426397125412638, "grad_norm": 3.0655479431152344, "learning_rate": 2.21347876899104e-05, "loss": 2.1673, "step": 1894 }, { "epoch": 0.44287341883088427, "grad_norm": 3.112384080886841, "learning_rate": 2.214647448383327e-05, "loss": 2.1179, "step": 1895 }, { "epoch": 0.4431071251205048, "grad_norm": 2.6592495441436768, "learning_rate": 2.2158161277756134e-05, "loss": 2.301, "step": 1896 }, { "epoch": 0.4433408314101253, "grad_norm": 2.4565036296844482, "learning_rate": 2.2169848071679e-05, "loss": 2.0589, "step": 1897 }, { "epoch": 0.44357453769974586, "grad_norm": 2.79766845703125, "learning_rate": 2.2181534865601872e-05, "loss": 2.2496, "step": 1898 }, { "epoch": 0.44380824398936636, "grad_norm": 2.6436970233917236, "learning_rate": 2.2193221659524736e-05, "loss": 2.2677, "step": 1899 }, { "epoch": 0.4440419502789869, "grad_norm": 3.4952657222747803, "learning_rate": 2.2204908453447606e-05, "loss": 2.2845, "step": 1900 }, { "epoch": 0.4442756565686074, "grad_norm": 2.7374351024627686, "learning_rate": 2.2216595247370473e-05, "loss": 2.2148, "step": 1901 }, { "epoch": 0.4445093628582279, "grad_norm": 2.7588796615600586, "learning_rate": 2.222828204129334e-05, "loss": 2.1595, "step": 1902 }, { "epoch": 0.44474306914784845, "grad_norm": 3.5166049003601074, "learning_rate": 2.2239968835216207e-05, "loss": 2.1675, "step": 1903 }, { "epoch": 0.44497677543746894, "grad_norm": 2.6390292644500732, "learning_rate": 2.2251655629139074e-05, "loss": 2.0649, "step": 1904 }, { "epoch": 0.4452104817270895, "grad_norm": 2.6998863220214844, "learning_rate": 2.226334242306194e-05, "loss": 2.1821, "step": 1905 }, { "epoch": 0.44544418801671, "grad_norm": 2.552177667617798, "learning_rate": 2.2275029216984808e-05, "loss": 2.165, "step": 1906 }, { "epoch": 0.44567789430633054, "grad_norm": 2.6083121299743652, "learning_rate": 2.2286716010907675e-05, "loss": 2.1762, "step": 1907 }, { "epoch": 0.44591160059595103, "grad_norm": 2.6626012325286865, "learning_rate": 2.2298402804830542e-05, "loss": 2.1291, "step": 1908 }, { "epoch": 0.4461453068855716, "grad_norm": 3.158504009246826, "learning_rate": 2.231008959875341e-05, "loss": 2.2343, "step": 1909 }, { "epoch": 0.4463790131751921, "grad_norm": 2.941570520401001, "learning_rate": 2.2321776392676276e-05, "loss": 2.2602, "step": 1910 }, { "epoch": 0.4466127194648126, "grad_norm": 2.8536713123321533, "learning_rate": 2.2333463186599143e-05, "loss": 2.2622, "step": 1911 }, { "epoch": 0.4468464257544331, "grad_norm": 4.014939785003662, "learning_rate": 2.234514998052201e-05, "loss": 2.2892, "step": 1912 }, { "epoch": 0.4470801320440536, "grad_norm": 2.887481689453125, "learning_rate": 2.2356836774444878e-05, "loss": 2.1178, "step": 1913 }, { "epoch": 0.44731383833367416, "grad_norm": 4.079108715057373, "learning_rate": 2.2368523568367745e-05, "loss": 2.4012, "step": 1914 }, { "epoch": 0.44754754462329466, "grad_norm": 3.6631476879119873, "learning_rate": 2.238021036229061e-05, "loss": 2.1684, "step": 1915 }, { "epoch": 0.4477812509129152, "grad_norm": 2.7395315170288086, "learning_rate": 2.239189715621348e-05, "loss": 2.0844, "step": 1916 }, { "epoch": 0.4480149572025357, "grad_norm": 4.039689064025879, "learning_rate": 2.2403583950136346e-05, "loss": 2.2617, "step": 1917 }, { "epoch": 0.44824866349215625, "grad_norm": 3.42154598236084, "learning_rate": 2.2415270744059216e-05, "loss": 2.2515, "step": 1918 }, { "epoch": 0.44848236978177675, "grad_norm": 3.3102076053619385, "learning_rate": 2.242695753798208e-05, "loss": 2.3461, "step": 1919 }, { "epoch": 0.4487160760713973, "grad_norm": 2.414551019668579, "learning_rate": 2.2438644331904947e-05, "loss": 2.0728, "step": 1920 }, { "epoch": 0.4489497823610178, "grad_norm": 2.686201810836792, "learning_rate": 2.2450331125827817e-05, "loss": 2.1373, "step": 1921 }, { "epoch": 0.4491834886506383, "grad_norm": 3.258359670639038, "learning_rate": 2.246201791975068e-05, "loss": 2.3743, "step": 1922 }, { "epoch": 0.44941719494025884, "grad_norm": 2.6427414417266846, "learning_rate": 2.2473704713673548e-05, "loss": 2.2059, "step": 1923 }, { "epoch": 0.44965090122987933, "grad_norm": 2.7689976692199707, "learning_rate": 2.248539150759642e-05, "loss": 2.2681, "step": 1924 }, { "epoch": 0.4498846075194999, "grad_norm": 2.962830066680908, "learning_rate": 2.2497078301519282e-05, "loss": 2.2035, "step": 1925 }, { "epoch": 0.4501183138091204, "grad_norm": 2.551330089569092, "learning_rate": 2.2508765095442153e-05, "loss": 2.1672, "step": 1926 }, { "epoch": 0.4503520200987409, "grad_norm": 2.5886523723602295, "learning_rate": 2.252045188936502e-05, "loss": 2.3876, "step": 1927 }, { "epoch": 0.4505857263883614, "grad_norm": 2.6324360370635986, "learning_rate": 2.2532138683287883e-05, "loss": 2.1597, "step": 1928 }, { "epoch": 0.45081943267798197, "grad_norm": 3.6057631969451904, "learning_rate": 2.2543825477210754e-05, "loss": 2.0366, "step": 1929 }, { "epoch": 0.45105313896760246, "grad_norm": 2.675565004348755, "learning_rate": 2.255551227113362e-05, "loss": 2.1592, "step": 1930 }, { "epoch": 0.451286845257223, "grad_norm": 3.5541207790374756, "learning_rate": 2.2567199065056484e-05, "loss": 2.1298, "step": 1931 }, { "epoch": 0.4515205515468435, "grad_norm": 3.838644027709961, "learning_rate": 2.2578885858979355e-05, "loss": 2.2105, "step": 1932 }, { "epoch": 0.451754257836464, "grad_norm": 2.563831090927124, "learning_rate": 2.2590572652902222e-05, "loss": 2.3543, "step": 1933 }, { "epoch": 0.45198796412608455, "grad_norm": 3.5777831077575684, "learning_rate": 2.260225944682509e-05, "loss": 2.1599, "step": 1934 }, { "epoch": 0.45222167041570505, "grad_norm": 3.4820361137390137, "learning_rate": 2.2613946240747956e-05, "loss": 2.255, "step": 1935 }, { "epoch": 0.4524553767053256, "grad_norm": 2.8237740993499756, "learning_rate": 2.2625633034670823e-05, "loss": 2.1398, "step": 1936 }, { "epoch": 0.4526890829949461, "grad_norm": 5.319874286651611, "learning_rate": 2.263731982859369e-05, "loss": 2.0326, "step": 1937 }, { "epoch": 0.45292278928456664, "grad_norm": 4.188897132873535, "learning_rate": 2.2649006622516557e-05, "loss": 2.2741, "step": 1938 }, { "epoch": 0.45315649557418713, "grad_norm": 2.5689687728881836, "learning_rate": 2.2660693416439424e-05, "loss": 2.1303, "step": 1939 }, { "epoch": 0.4533902018638077, "grad_norm": 4.273180961608887, "learning_rate": 2.267238021036229e-05, "loss": 2.1385, "step": 1940 }, { "epoch": 0.4536239081534282, "grad_norm": 4.693446159362793, "learning_rate": 2.2684067004285158e-05, "loss": 2.3906, "step": 1941 }, { "epoch": 0.4538576144430487, "grad_norm": 3.2082204818725586, "learning_rate": 2.2695753798208025e-05, "loss": 2.3361, "step": 1942 }, { "epoch": 0.4540913207326692, "grad_norm": 3.040693759918213, "learning_rate": 2.2707440592130892e-05, "loss": 2.2164, "step": 1943 }, { "epoch": 0.4543250270222897, "grad_norm": 3.6446752548217773, "learning_rate": 2.271912738605376e-05, "loss": 2.2581, "step": 1944 }, { "epoch": 0.45455873331191027, "grad_norm": 3.172974109649658, "learning_rate": 2.2730814179976626e-05, "loss": 2.1611, "step": 1945 }, { "epoch": 0.45479243960153076, "grad_norm": 2.9480648040771484, "learning_rate": 2.2742500973899493e-05, "loss": 2.2537, "step": 1946 }, { "epoch": 0.4550261458911513, "grad_norm": 4.165976524353027, "learning_rate": 2.275418776782236e-05, "loss": 2.0951, "step": 1947 }, { "epoch": 0.4552598521807718, "grad_norm": 2.7630527019500732, "learning_rate": 2.2765874561745228e-05, "loss": 2.2027, "step": 1948 }, { "epoch": 0.45549355847039236, "grad_norm": 2.8796286582946777, "learning_rate": 2.2777561355668095e-05, "loss": 2.2526, "step": 1949 }, { "epoch": 0.45572726476001285, "grad_norm": 2.7398648262023926, "learning_rate": 2.2789248149590965e-05, "loss": 2.0792, "step": 1950 }, { "epoch": 0.45596097104963335, "grad_norm": 2.6058292388916016, "learning_rate": 2.280093494351383e-05, "loss": 2.2718, "step": 1951 }, { "epoch": 0.4561946773392539, "grad_norm": 3.3854219913482666, "learning_rate": 2.2812621737436696e-05, "loss": 2.12, "step": 1952 }, { "epoch": 0.4564283836288744, "grad_norm": 2.6787493228912354, "learning_rate": 2.2824308531359566e-05, "loss": 2.1537, "step": 1953 }, { "epoch": 0.45666208991849494, "grad_norm": 2.5289745330810547, "learning_rate": 2.283599532528243e-05, "loss": 2.0478, "step": 1954 }, { "epoch": 0.45689579620811543, "grad_norm": 2.3863182067871094, "learning_rate": 2.28476821192053e-05, "loss": 2.1552, "step": 1955 }, { "epoch": 0.457129502497736, "grad_norm": 3.6234970092773438, "learning_rate": 2.2859368913128167e-05, "loss": 2.1816, "step": 1956 }, { "epoch": 0.4573632087873565, "grad_norm": 5.35477876663208, "learning_rate": 2.287105570705103e-05, "loss": 2.1324, "step": 1957 }, { "epoch": 0.45759691507697703, "grad_norm": 2.764617919921875, "learning_rate": 2.28827425009739e-05, "loss": 2.0598, "step": 1958 }, { "epoch": 0.4578306213665975, "grad_norm": 4.258030891418457, "learning_rate": 2.289442929489677e-05, "loss": 2.2463, "step": 1959 }, { "epoch": 0.4580643276562181, "grad_norm": 3.475505828857422, "learning_rate": 2.2906116088819632e-05, "loss": 2.3579, "step": 1960 }, { "epoch": 0.45829803394583857, "grad_norm": 2.633333444595337, "learning_rate": 2.2917802882742503e-05, "loss": 2.1109, "step": 1961 }, { "epoch": 0.45853174023545906, "grad_norm": 3.467999219894409, "learning_rate": 2.292948967666537e-05, "loss": 2.22, "step": 1962 }, { "epoch": 0.4587654465250796, "grad_norm": 3.0065510272979736, "learning_rate": 2.2941176470588233e-05, "loss": 2.4781, "step": 1963 }, { "epoch": 0.4589991528147001, "grad_norm": 3.27762508392334, "learning_rate": 2.2952863264511104e-05, "loss": 2.1207, "step": 1964 }, { "epoch": 0.45923285910432066, "grad_norm": 2.7734293937683105, "learning_rate": 2.296455005843397e-05, "loss": 2.1434, "step": 1965 }, { "epoch": 0.45946656539394115, "grad_norm": 2.433542013168335, "learning_rate": 2.2976236852356838e-05, "loss": 2.218, "step": 1966 }, { "epoch": 0.4597002716835617, "grad_norm": 2.3464291095733643, "learning_rate": 2.2987923646279705e-05, "loss": 2.2366, "step": 1967 }, { "epoch": 0.4599339779731822, "grad_norm": 3.266556978225708, "learning_rate": 2.2999610440202572e-05, "loss": 2.3157, "step": 1968 }, { "epoch": 0.46016768426280275, "grad_norm": 2.773423671722412, "learning_rate": 2.301129723412544e-05, "loss": 2.0629, "step": 1969 }, { "epoch": 0.46040139055242324, "grad_norm": 2.4997825622558594, "learning_rate": 2.3022984028048306e-05, "loss": 2.1051, "step": 1970 }, { "epoch": 0.46063509684204373, "grad_norm": 2.380723714828491, "learning_rate": 2.3034670821971173e-05, "loss": 2.1867, "step": 1971 }, { "epoch": 0.4608688031316643, "grad_norm": 2.6484904289245605, "learning_rate": 2.304635761589404e-05, "loss": 2.1194, "step": 1972 }, { "epoch": 0.4611025094212848, "grad_norm": 2.4232683181762695, "learning_rate": 2.3058044409816907e-05, "loss": 2.1653, "step": 1973 }, { "epoch": 0.46133621571090533, "grad_norm": 2.5574703216552734, "learning_rate": 2.3069731203739774e-05, "loss": 2.1748, "step": 1974 }, { "epoch": 0.4615699220005258, "grad_norm": 2.567248582839966, "learning_rate": 2.308141799766264e-05, "loss": 2.1267, "step": 1975 }, { "epoch": 0.4618036282901464, "grad_norm": 2.8763649463653564, "learning_rate": 2.3093104791585508e-05, "loss": 2.2167, "step": 1976 }, { "epoch": 0.46203733457976687, "grad_norm": 2.668464422225952, "learning_rate": 2.3104791585508375e-05, "loss": 2.3392, "step": 1977 }, { "epoch": 0.4622710408693874, "grad_norm": 4.446616172790527, "learning_rate": 2.3116478379431242e-05, "loss": 2.1595, "step": 1978 }, { "epoch": 0.4625047471590079, "grad_norm": 2.67913556098938, "learning_rate": 2.312816517335411e-05, "loss": 2.2763, "step": 1979 }, { "epoch": 0.46273845344862846, "grad_norm": 3.1912686824798584, "learning_rate": 2.3139851967276976e-05, "loss": 2.1949, "step": 1980 }, { "epoch": 0.46297215973824896, "grad_norm": 2.9722824096679688, "learning_rate": 2.3151538761199847e-05, "loss": 2.12, "step": 1981 }, { "epoch": 0.46320586602786945, "grad_norm": 3.1884450912475586, "learning_rate": 2.3163225555122714e-05, "loss": 2.0937, "step": 1982 }, { "epoch": 0.46343957231749, "grad_norm": 2.526158094406128, "learning_rate": 2.3174912349045578e-05, "loss": 2.2814, "step": 1983 }, { "epoch": 0.4636732786071105, "grad_norm": 4.107946872711182, "learning_rate": 2.3186599142968448e-05, "loss": 2.2901, "step": 1984 }, { "epoch": 0.46390698489673104, "grad_norm": 3.688542366027832, "learning_rate": 2.3198285936891315e-05, "loss": 2.1236, "step": 1985 }, { "epoch": 0.46414069118635154, "grad_norm": 2.962937593460083, "learning_rate": 2.320997273081418e-05, "loss": 2.209, "step": 1986 }, { "epoch": 0.4643743974759721, "grad_norm": 4.002509593963623, "learning_rate": 2.322165952473705e-05, "loss": 2.1887, "step": 1987 }, { "epoch": 0.4646081037655926, "grad_norm": 3.2456717491149902, "learning_rate": 2.3233346318659916e-05, "loss": 2.1871, "step": 1988 }, { "epoch": 0.46484181005521313, "grad_norm": 2.8218657970428467, "learning_rate": 2.324503311258278e-05, "loss": 2.119, "step": 1989 }, { "epoch": 0.46507551634483363, "grad_norm": 2.351463556289673, "learning_rate": 2.325671990650565e-05, "loss": 2.196, "step": 1990 }, { "epoch": 0.4653092226344541, "grad_norm": 3.2824254035949707, "learning_rate": 2.3268406700428517e-05, "loss": 2.0599, "step": 1991 }, { "epoch": 0.4655429289240747, "grad_norm": 3.039299726486206, "learning_rate": 2.328009349435138e-05, "loss": 2.014, "step": 1992 }, { "epoch": 0.46577663521369517, "grad_norm": 2.8978114128112793, "learning_rate": 2.329178028827425e-05, "loss": 2.004, "step": 1993 }, { "epoch": 0.4660103415033157, "grad_norm": 2.994581460952759, "learning_rate": 2.330346708219712e-05, "loss": 2.0842, "step": 1994 }, { "epoch": 0.4662440477929362, "grad_norm": 3.170658588409424, "learning_rate": 2.3315153876119982e-05, "loss": 2.3338, "step": 1995 }, { "epoch": 0.46647775408255676, "grad_norm": 2.771014451980591, "learning_rate": 2.3326840670042853e-05, "loss": 2.1639, "step": 1996 }, { "epoch": 0.46671146037217726, "grad_norm": 4.649882793426514, "learning_rate": 2.333852746396572e-05, "loss": 2.109, "step": 1997 }, { "epoch": 0.4669451666617978, "grad_norm": 5.079276084899902, "learning_rate": 2.3350214257888587e-05, "loss": 2.152, "step": 1998 }, { "epoch": 0.4671788729514183, "grad_norm": 2.7853446006774902, "learning_rate": 2.3361901051811454e-05, "loss": 2.2542, "step": 1999 }, { "epoch": 0.46741257924103885, "grad_norm": 3.383317708969116, "learning_rate": 2.337358784573432e-05, "loss": 2.4034, "step": 2000 }, { "epoch": 0.46764628553065934, "grad_norm": 3.684896230697632, "learning_rate": 2.3385274639657188e-05, "loss": 2.2791, "step": 2001 }, { "epoch": 0.46787999182027984, "grad_norm": 2.8884284496307373, "learning_rate": 2.3396961433580055e-05, "loss": 2.1674, "step": 2002 }, { "epoch": 0.4681136981099004, "grad_norm": 3.166696310043335, "learning_rate": 2.3408648227502922e-05, "loss": 2.1946, "step": 2003 }, { "epoch": 0.4683474043995209, "grad_norm": 3.3002355098724365, "learning_rate": 2.342033502142579e-05, "loss": 2.0832, "step": 2004 }, { "epoch": 0.46858111068914143, "grad_norm": 2.430129051208496, "learning_rate": 2.3432021815348656e-05, "loss": 2.2163, "step": 2005 }, { "epoch": 0.46881481697876193, "grad_norm": 2.3263721466064453, "learning_rate": 2.3443708609271523e-05, "loss": 2.1388, "step": 2006 }, { "epoch": 0.4690485232683825, "grad_norm": 2.6231958866119385, "learning_rate": 2.345539540319439e-05, "loss": 2.2939, "step": 2007 }, { "epoch": 0.469282229558003, "grad_norm": 2.5458898544311523, "learning_rate": 2.3467082197117257e-05, "loss": 2.2537, "step": 2008 }, { "epoch": 0.4695159358476235, "grad_norm": 2.677114486694336, "learning_rate": 2.3478768991040124e-05, "loss": 2.0056, "step": 2009 }, { "epoch": 0.469749642137244, "grad_norm": 2.756920337677002, "learning_rate": 2.3490455784962995e-05, "loss": 2.2205, "step": 2010 }, { "epoch": 0.4699833484268645, "grad_norm": 2.708707571029663, "learning_rate": 2.350214257888586e-05, "loss": 2.1337, "step": 2011 }, { "epoch": 0.47021705471648506, "grad_norm": 4.486354827880859, "learning_rate": 2.3513829372808725e-05, "loss": 2.3048, "step": 2012 }, { "epoch": 0.47045076100610556, "grad_norm": 2.7190353870391846, "learning_rate": 2.3525516166731596e-05, "loss": 2.1882, "step": 2013 }, { "epoch": 0.4706844672957261, "grad_norm": 2.800949811935425, "learning_rate": 2.3537202960654463e-05, "loss": 2.0719, "step": 2014 }, { "epoch": 0.4709181735853466, "grad_norm": 4.396462440490723, "learning_rate": 2.3548889754577326e-05, "loss": 2.1952, "step": 2015 }, { "epoch": 0.47115187987496715, "grad_norm": 6.085193634033203, "learning_rate": 2.3560576548500197e-05, "loss": 2.0783, "step": 2016 }, { "epoch": 0.47138558616458764, "grad_norm": 2.8275158405303955, "learning_rate": 2.3572263342423064e-05, "loss": 2.3408, "step": 2017 }, { "epoch": 0.4716192924542082, "grad_norm": 4.530135631561279, "learning_rate": 2.3583950136345928e-05, "loss": 2.2727, "step": 2018 }, { "epoch": 0.4718529987438287, "grad_norm": 4.2954325675964355, "learning_rate": 2.3595636930268798e-05, "loss": 2.2664, "step": 2019 }, { "epoch": 0.47208670503344924, "grad_norm": 2.955198049545288, "learning_rate": 2.3607323724191665e-05, "loss": 2.3035, "step": 2020 }, { "epoch": 0.47232041132306973, "grad_norm": 3.7517993450164795, "learning_rate": 2.361901051811453e-05, "loss": 2.1864, "step": 2021 }, { "epoch": 0.4725541176126902, "grad_norm": 2.9317328929901123, "learning_rate": 2.36306973120374e-05, "loss": 2.2872, "step": 2022 }, { "epoch": 0.4727878239023108, "grad_norm": 2.4745349884033203, "learning_rate": 2.3642384105960266e-05, "loss": 2.3005, "step": 2023 }, { "epoch": 0.47302153019193127, "grad_norm": 2.38458251953125, "learning_rate": 2.365407089988313e-05, "loss": 2.2994, "step": 2024 }, { "epoch": 0.4732552364815518, "grad_norm": 2.7426817417144775, "learning_rate": 2.3665757693806e-05, "loss": 2.1154, "step": 2025 }, { "epoch": 0.4734889427711723, "grad_norm": 2.9106028079986572, "learning_rate": 2.3677444487728867e-05, "loss": 2.287, "step": 2026 }, { "epoch": 0.47372264906079287, "grad_norm": 2.53344464302063, "learning_rate": 2.3689131281651734e-05, "loss": 2.2268, "step": 2027 }, { "epoch": 0.47395635535041336, "grad_norm": 2.3219408988952637, "learning_rate": 2.37008180755746e-05, "loss": 2.1284, "step": 2028 }, { "epoch": 0.4741900616400339, "grad_norm": 2.2951860427856445, "learning_rate": 2.371250486949747e-05, "loss": 2.058, "step": 2029 }, { "epoch": 0.4744237679296544, "grad_norm": 3.1306824684143066, "learning_rate": 2.3724191663420336e-05, "loss": 2.1431, "step": 2030 }, { "epoch": 0.4746574742192749, "grad_norm": 3.402435064315796, "learning_rate": 2.3735878457343203e-05, "loss": 2.264, "step": 2031 }, { "epoch": 0.47489118050889545, "grad_norm": 3.982570171356201, "learning_rate": 2.374756525126607e-05, "loss": 2.2689, "step": 2032 }, { "epoch": 0.47512488679851594, "grad_norm": 3.8462817668914795, "learning_rate": 2.3759252045188937e-05, "loss": 2.1771, "step": 2033 }, { "epoch": 0.4753585930881365, "grad_norm": 3.2540156841278076, "learning_rate": 2.3770938839111804e-05, "loss": 2.1548, "step": 2034 }, { "epoch": 0.475592299377757, "grad_norm": 3.125035285949707, "learning_rate": 2.378262563303467e-05, "loss": 2.1444, "step": 2035 }, { "epoch": 0.47582600566737754, "grad_norm": 5.10337495803833, "learning_rate": 2.379431242695754e-05, "loss": 2.1234, "step": 2036 }, { "epoch": 0.47605971195699803, "grad_norm": 2.645731210708618, "learning_rate": 2.3805999220880405e-05, "loss": 2.1892, "step": 2037 }, { "epoch": 0.4762934182466186, "grad_norm": 3.5539369583129883, "learning_rate": 2.3817686014803272e-05, "loss": 2.259, "step": 2038 }, { "epoch": 0.4765271245362391, "grad_norm": 2.7073750495910645, "learning_rate": 2.3829372808726142e-05, "loss": 2.1711, "step": 2039 }, { "epoch": 0.4767608308258596, "grad_norm": 2.398108959197998, "learning_rate": 2.3841059602649006e-05, "loss": 2.3504, "step": 2040 }, { "epoch": 0.4769945371154801, "grad_norm": 2.397775411605835, "learning_rate": 2.3852746396571873e-05, "loss": 2.1602, "step": 2041 }, { "epoch": 0.4772282434051006, "grad_norm": 2.4967358112335205, "learning_rate": 2.3864433190494744e-05, "loss": 2.3158, "step": 2042 }, { "epoch": 0.47746194969472117, "grad_norm": 2.6652872562408447, "learning_rate": 2.387611998441761e-05, "loss": 2.0266, "step": 2043 }, { "epoch": 0.47769565598434166, "grad_norm": 2.6790459156036377, "learning_rate": 2.3887806778340474e-05, "loss": 2.192, "step": 2044 }, { "epoch": 0.4779293622739622, "grad_norm": 2.496969223022461, "learning_rate": 2.3899493572263345e-05, "loss": 2.1882, "step": 2045 }, { "epoch": 0.4781630685635827, "grad_norm": 2.697309970855713, "learning_rate": 2.3911180366186212e-05, "loss": 2.0754, "step": 2046 }, { "epoch": 0.47839677485320325, "grad_norm": 2.549802541732788, "learning_rate": 2.3922867160109075e-05, "loss": 2.2833, "step": 2047 }, { "epoch": 0.47863048114282375, "grad_norm": 2.657839059829712, "learning_rate": 2.3934553954031946e-05, "loss": 2.1973, "step": 2048 }, { "epoch": 0.4788641874324443, "grad_norm": 2.6649465560913086, "learning_rate": 2.3946240747954813e-05, "loss": 2.2662, "step": 2049 }, { "epoch": 0.4790978937220648, "grad_norm": 2.9459502696990967, "learning_rate": 2.3957927541877677e-05, "loss": 2.1763, "step": 2050 }, { "epoch": 0.4793316000116853, "grad_norm": 2.5791468620300293, "learning_rate": 2.3969614335800547e-05, "loss": 2.1379, "step": 2051 }, { "epoch": 0.47956530630130584, "grad_norm": 2.540917158126831, "learning_rate": 2.3981301129723414e-05, "loss": 2.2305, "step": 2052 }, { "epoch": 0.47979901259092633, "grad_norm": 3.522838592529297, "learning_rate": 2.3992987923646278e-05, "loss": 2.2691, "step": 2053 }, { "epoch": 0.4800327188805469, "grad_norm": 2.6310484409332275, "learning_rate": 2.4004674717569148e-05, "loss": 2.2241, "step": 2054 }, { "epoch": 0.4802664251701674, "grad_norm": 2.569125175476074, "learning_rate": 2.4016361511492015e-05, "loss": 2.3445, "step": 2055 }, { "epoch": 0.4805001314597879, "grad_norm": 2.6243481636047363, "learning_rate": 2.402804830541488e-05, "loss": 2.2791, "step": 2056 }, { "epoch": 0.4807338377494084, "grad_norm": 2.4727535247802734, "learning_rate": 2.403973509933775e-05, "loss": 2.2584, "step": 2057 }, { "epoch": 0.48096754403902897, "grad_norm": 2.674720525741577, "learning_rate": 2.4051421893260616e-05, "loss": 2.1166, "step": 2058 }, { "epoch": 0.48120125032864947, "grad_norm": 2.882211446762085, "learning_rate": 2.4063108687183483e-05, "loss": 2.2931, "step": 2059 }, { "epoch": 0.48143495661827, "grad_norm": 3.022773504257202, "learning_rate": 2.407479548110635e-05, "loss": 2.3536, "step": 2060 }, { "epoch": 0.4816686629078905, "grad_norm": 2.544262170791626, "learning_rate": 2.4086482275029217e-05, "loss": 2.223, "step": 2061 }, { "epoch": 0.481902369197511, "grad_norm": 2.7344131469726562, "learning_rate": 2.4098169068952084e-05, "loss": 2.1843, "step": 2062 }, { "epoch": 0.48213607548713155, "grad_norm": 2.4483582973480225, "learning_rate": 2.410985586287495e-05, "loss": 2.2767, "step": 2063 }, { "epoch": 0.48236978177675205, "grad_norm": 2.725341796875, "learning_rate": 2.412154265679782e-05, "loss": 2.3068, "step": 2064 }, { "epoch": 0.4826034880663726, "grad_norm": 2.771304130554199, "learning_rate": 2.413322945072069e-05, "loss": 2.1982, "step": 2065 }, { "epoch": 0.4828371943559931, "grad_norm": 2.7458605766296387, "learning_rate": 2.4144916244643553e-05, "loss": 2.382, "step": 2066 }, { "epoch": 0.48307090064561364, "grad_norm": 2.558445930480957, "learning_rate": 2.415660303856642e-05, "loss": 2.0191, "step": 2067 }, { "epoch": 0.48330460693523414, "grad_norm": 3.1313650608062744, "learning_rate": 2.416828983248929e-05, "loss": 2.1953, "step": 2068 }, { "epoch": 0.4835383132248547, "grad_norm": 2.812892436981201, "learning_rate": 2.4179976626412154e-05, "loss": 2.1438, "step": 2069 }, { "epoch": 0.4837720195144752, "grad_norm": 3.358776330947876, "learning_rate": 2.419166342033502e-05, "loss": 2.156, "step": 2070 }, { "epoch": 0.4840057258040957, "grad_norm": 2.682061195373535, "learning_rate": 2.420335021425789e-05, "loss": 2.2242, "step": 2071 }, { "epoch": 0.4842394320937162, "grad_norm": 6.9602460861206055, "learning_rate": 2.4215037008180755e-05, "loss": 2.2769, "step": 2072 }, { "epoch": 0.4844731383833367, "grad_norm": 3.075552463531494, "learning_rate": 2.4226723802103622e-05, "loss": 2.1808, "step": 2073 }, { "epoch": 0.48470684467295727, "grad_norm": 3.1153039932250977, "learning_rate": 2.4238410596026492e-05, "loss": 2.1796, "step": 2074 }, { "epoch": 0.48494055096257777, "grad_norm": 5.425137519836426, "learning_rate": 2.425009738994936e-05, "loss": 2.2521, "step": 2075 }, { "epoch": 0.4851742572521983, "grad_norm": 3.748281240463257, "learning_rate": 2.4261784183872223e-05, "loss": 2.1969, "step": 2076 }, { "epoch": 0.4854079635418188, "grad_norm": 5.47175407409668, "learning_rate": 2.4273470977795094e-05, "loss": 2.0811, "step": 2077 }, { "epoch": 0.48564166983143936, "grad_norm": 4.707736492156982, "learning_rate": 2.428515777171796e-05, "loss": 2.2264, "step": 2078 }, { "epoch": 0.48587537612105985, "grad_norm": 2.6535332202911377, "learning_rate": 2.4296844565640824e-05, "loss": 2.2459, "step": 2079 }, { "epoch": 0.4861090824106804, "grad_norm": 2.5406551361083984, "learning_rate": 2.4308531359563695e-05, "loss": 2.1033, "step": 2080 }, { "epoch": 0.4863427887003009, "grad_norm": 2.8931167125701904, "learning_rate": 2.4320218153486562e-05, "loss": 2.212, "step": 2081 }, { "epoch": 0.4865764949899214, "grad_norm": 3.2210910320281982, "learning_rate": 2.4331904947409425e-05, "loss": 2.2021, "step": 2082 }, { "epoch": 0.48681020127954194, "grad_norm": 3.018939733505249, "learning_rate": 2.4343591741332296e-05, "loss": 2.1437, "step": 2083 }, { "epoch": 0.48704390756916244, "grad_norm": 2.8773252964019775, "learning_rate": 2.4355278535255163e-05, "loss": 2.1797, "step": 2084 }, { "epoch": 0.487277613858783, "grad_norm": 3.15427303314209, "learning_rate": 2.4366965329178027e-05, "loss": 2.1568, "step": 2085 }, { "epoch": 0.4875113201484035, "grad_norm": 2.564553737640381, "learning_rate": 2.4378652123100897e-05, "loss": 2.19, "step": 2086 }, { "epoch": 0.48774502643802403, "grad_norm": 2.658665418624878, "learning_rate": 2.4390338917023764e-05, "loss": 2.2977, "step": 2087 }, { "epoch": 0.4879787327276445, "grad_norm": 4.085999488830566, "learning_rate": 2.4402025710946628e-05, "loss": 2.1638, "step": 2088 }, { "epoch": 0.4882124390172651, "grad_norm": 3.405120611190796, "learning_rate": 2.4413712504869498e-05, "loss": 2.5164, "step": 2089 }, { "epoch": 0.48844614530688557, "grad_norm": 2.977567434310913, "learning_rate": 2.4425399298792365e-05, "loss": 2.1867, "step": 2090 }, { "epoch": 0.48867985159650607, "grad_norm": 3.5622479915618896, "learning_rate": 2.4437086092715236e-05, "loss": 2.1015, "step": 2091 }, { "epoch": 0.4889135578861266, "grad_norm": 2.503800392150879, "learning_rate": 2.44487728866381e-05, "loss": 2.1756, "step": 2092 }, { "epoch": 0.4891472641757471, "grad_norm": 3.1085376739501953, "learning_rate": 2.4460459680560966e-05, "loss": 2.071, "step": 2093 }, { "epoch": 0.48938097046536766, "grad_norm": 2.716881513595581, "learning_rate": 2.4472146474483837e-05, "loss": 2.1304, "step": 2094 }, { "epoch": 0.48961467675498815, "grad_norm": 2.4962210655212402, "learning_rate": 2.44838332684067e-05, "loss": 2.2602, "step": 2095 }, { "epoch": 0.4898483830446087, "grad_norm": 3.1439356803894043, "learning_rate": 2.4495520062329567e-05, "loss": 2.1626, "step": 2096 }, { "epoch": 0.4900820893342292, "grad_norm": 2.3198962211608887, "learning_rate": 2.4507206856252438e-05, "loss": 2.1743, "step": 2097 }, { "epoch": 0.49031579562384975, "grad_norm": 2.695725202560425, "learning_rate": 2.45188936501753e-05, "loss": 1.9994, "step": 2098 }, { "epoch": 0.49054950191347024, "grad_norm": 2.9755871295928955, "learning_rate": 2.453058044409817e-05, "loss": 2.406, "step": 2099 }, { "epoch": 0.4907832082030908, "grad_norm": 2.6927602291107178, "learning_rate": 2.454226723802104e-05, "loss": 2.0576, "step": 2100 }, { "epoch": 0.4910169144927113, "grad_norm": 2.9859201908111572, "learning_rate": 2.4553954031943903e-05, "loss": 2.143, "step": 2101 }, { "epoch": 0.4912506207823318, "grad_norm": 2.8760581016540527, "learning_rate": 2.456564082586677e-05, "loss": 2.2044, "step": 2102 }, { "epoch": 0.49148432707195233, "grad_norm": 3.0201075077056885, "learning_rate": 2.457732761978964e-05, "loss": 2.0182, "step": 2103 }, { "epoch": 0.4917180333615728, "grad_norm": 2.6895105838775635, "learning_rate": 2.4589014413712504e-05, "loss": 2.1408, "step": 2104 }, { "epoch": 0.4919517396511934, "grad_norm": 3.8730130195617676, "learning_rate": 2.460070120763537e-05, "loss": 2.1809, "step": 2105 }, { "epoch": 0.49218544594081387, "grad_norm": 3.517266273498535, "learning_rate": 2.461238800155824e-05, "loss": 2.1138, "step": 2106 }, { "epoch": 0.4924191522304344, "grad_norm": 2.7919392585754395, "learning_rate": 2.462407479548111e-05, "loss": 2.3518, "step": 2107 }, { "epoch": 0.4926528585200549, "grad_norm": 4.085084438323975, "learning_rate": 2.4635761589403972e-05, "loss": 2.1895, "step": 2108 }, { "epoch": 0.49288656480967546, "grad_norm": 4.627925872802734, "learning_rate": 2.4647448383326842e-05, "loss": 2.2471, "step": 2109 }, { "epoch": 0.49312027109929596, "grad_norm": 2.8159713745117188, "learning_rate": 2.465913517724971e-05, "loss": 2.2051, "step": 2110 }, { "epoch": 0.49335397738891645, "grad_norm": 4.154223918914795, "learning_rate": 2.4670821971172573e-05, "loss": 2.0993, "step": 2111 }, { "epoch": 0.493587683678537, "grad_norm": 3.7139980792999268, "learning_rate": 2.4682508765095444e-05, "loss": 2.1929, "step": 2112 }, { "epoch": 0.4938213899681575, "grad_norm": 2.9377665519714355, "learning_rate": 2.469419555901831e-05, "loss": 2.1095, "step": 2113 }, { "epoch": 0.49405509625777805, "grad_norm": 4.533810615539551, "learning_rate": 2.4705882352941174e-05, "loss": 2.1914, "step": 2114 }, { "epoch": 0.49428880254739854, "grad_norm": 4.395666122436523, "learning_rate": 2.4717569146864045e-05, "loss": 2.1606, "step": 2115 }, { "epoch": 0.4945225088370191, "grad_norm": 2.984097480773926, "learning_rate": 2.4729255940786912e-05, "loss": 2.1742, "step": 2116 }, { "epoch": 0.4947562151266396, "grad_norm": 4.6165900230407715, "learning_rate": 2.474094273470978e-05, "loss": 2.2237, "step": 2117 }, { "epoch": 0.49498992141626014, "grad_norm": 3.9990220069885254, "learning_rate": 2.4752629528632646e-05, "loss": 2.1968, "step": 2118 }, { "epoch": 0.49522362770588063, "grad_norm": 3.3016717433929443, "learning_rate": 2.4764316322555513e-05, "loss": 2.2419, "step": 2119 }, { "epoch": 0.4954573339955012, "grad_norm": 4.425976276397705, "learning_rate": 2.477600311647838e-05, "loss": 2.1067, "step": 2120 }, { "epoch": 0.4956910402851217, "grad_norm": 5.896582126617432, "learning_rate": 2.4787689910401247e-05, "loss": 2.1917, "step": 2121 }, { "epoch": 0.49592474657474217, "grad_norm": 4.773500919342041, "learning_rate": 2.4799376704324114e-05, "loss": 2.2587, "step": 2122 }, { "epoch": 0.4961584528643627, "grad_norm": 3.699228048324585, "learning_rate": 2.4811063498246985e-05, "loss": 2.1069, "step": 2123 }, { "epoch": 0.4963921591539832, "grad_norm": 7.4672698974609375, "learning_rate": 2.4822750292169848e-05, "loss": 2.157, "step": 2124 }, { "epoch": 0.49662586544360376, "grad_norm": 5.398828506469727, "learning_rate": 2.4834437086092715e-05, "loss": 2.3181, "step": 2125 }, { "epoch": 0.49685957173322426, "grad_norm": 3.3385019302368164, "learning_rate": 2.4846123880015586e-05, "loss": 2.1555, "step": 2126 }, { "epoch": 0.4970932780228448, "grad_norm": 4.809932708740234, "learning_rate": 2.485781067393845e-05, "loss": 2.1742, "step": 2127 }, { "epoch": 0.4973269843124653, "grad_norm": 8.498217582702637, "learning_rate": 2.4869497467861316e-05, "loss": 2.2021, "step": 2128 }, { "epoch": 0.49756069060208585, "grad_norm": 6.318469524383545, "learning_rate": 2.4881184261784187e-05, "loss": 2.2481, "step": 2129 }, { "epoch": 0.49779439689170635, "grad_norm": 3.5035417079925537, "learning_rate": 2.489287105570705e-05, "loss": 2.292, "step": 2130 }, { "epoch": 0.49802810318132684, "grad_norm": 5.547337055206299, "learning_rate": 2.4904557849629917e-05, "loss": 2.1612, "step": 2131 }, { "epoch": 0.4982618094709474, "grad_norm": 6.736724376678467, "learning_rate": 2.4916244643552788e-05, "loss": 2.2372, "step": 2132 }, { "epoch": 0.4984955157605679, "grad_norm": 7.314239978790283, "learning_rate": 2.492793143747565e-05, "loss": 2.2527, "step": 2133 }, { "epoch": 0.49872922205018844, "grad_norm": 3.1197683811187744, "learning_rate": 2.493961823139852e-05, "loss": 2.0569, "step": 2134 }, { "epoch": 0.49896292833980893, "grad_norm": 4.1725945472717285, "learning_rate": 2.495130502532139e-05, "loss": 2.1061, "step": 2135 }, { "epoch": 0.4991966346294295, "grad_norm": 5.87784481048584, "learning_rate": 2.4962991819244253e-05, "loss": 2.2004, "step": 2136 }, { "epoch": 0.49943034091905, "grad_norm": 5.577396869659424, "learning_rate": 2.497467861316712e-05, "loss": 2.1775, "step": 2137 }, { "epoch": 0.4996640472086705, "grad_norm": 3.2867727279663086, "learning_rate": 2.498636540708999e-05, "loss": 2.1, "step": 2138 }, { "epoch": 0.499897753498291, "grad_norm": 3.7097952365875244, "learning_rate": 2.4998052201012857e-05, "loss": 2.3094, "step": 2139 }, { "epoch": 0.5001314597879115, "grad_norm": 7.429637908935547, "learning_rate": 2.500973899493572e-05, "loss": 2.1712, "step": 2140 }, { "epoch": 0.500365166077532, "grad_norm": 5.724938869476318, "learning_rate": 2.502142578885859e-05, "loss": 2.2791, "step": 2141 }, { "epoch": 0.5005988723671526, "grad_norm": 3.20817494392395, "learning_rate": 2.503311258278146e-05, "loss": 2.2413, "step": 2142 }, { "epoch": 0.5008325786567731, "grad_norm": 2.7340738773345947, "learning_rate": 2.5044799376704322e-05, "loss": 2.0911, "step": 2143 }, { "epoch": 0.5010662849463936, "grad_norm": 3.0645527839660645, "learning_rate": 2.5056486170627192e-05, "loss": 2.0894, "step": 2144 }, { "epoch": 0.5012999912360141, "grad_norm": 2.9315378665924072, "learning_rate": 2.506817296455006e-05, "loss": 2.1132, "step": 2145 }, { "epoch": 0.5015336975256347, "grad_norm": 2.4468164443969727, "learning_rate": 2.5079859758472927e-05, "loss": 2.2452, "step": 2146 }, { "epoch": 0.5017674038152552, "grad_norm": 3.1717991828918457, "learning_rate": 2.5091546552395794e-05, "loss": 2.1794, "step": 2147 }, { "epoch": 0.5020011101048757, "grad_norm": 3.2002482414245605, "learning_rate": 2.510323334631866e-05, "loss": 2.1068, "step": 2148 }, { "epoch": 0.5022348163944962, "grad_norm": 5.008562088012695, "learning_rate": 2.5114920140241528e-05, "loss": 2.085, "step": 2149 }, { "epoch": 0.5024685226841168, "grad_norm": 2.640658140182495, "learning_rate": 2.5126606934164395e-05, "loss": 2.2903, "step": 2150 }, { "epoch": 0.5027022289737373, "grad_norm": 2.72247314453125, "learning_rate": 2.5138293728087262e-05, "loss": 2.194, "step": 2151 }, { "epoch": 0.5029359352633578, "grad_norm": 4.120837688446045, "learning_rate": 2.514998052201013e-05, "loss": 2.1122, "step": 2152 }, { "epoch": 0.5031696415529783, "grad_norm": 5.896668910980225, "learning_rate": 2.5161667315932996e-05, "loss": 2.1359, "step": 2153 }, { "epoch": 0.5034033478425988, "grad_norm": 2.6827526092529297, "learning_rate": 2.5173354109855863e-05, "loss": 2.1754, "step": 2154 }, { "epoch": 0.5036370541322194, "grad_norm": 2.797638416290283, "learning_rate": 2.5185040903778733e-05, "loss": 2.234, "step": 2155 }, { "epoch": 0.5038707604218399, "grad_norm": 4.017022132873535, "learning_rate": 2.5196727697701597e-05, "loss": 2.1774, "step": 2156 }, { "epoch": 0.5041044667114604, "grad_norm": 4.023656845092773, "learning_rate": 2.5208414491624464e-05, "loss": 2.1912, "step": 2157 }, { "epoch": 0.5043381730010809, "grad_norm": 3.2395853996276855, "learning_rate": 2.5220101285547335e-05, "loss": 2.1383, "step": 2158 }, { "epoch": 0.5045718792907015, "grad_norm": 5.154934406280518, "learning_rate": 2.5231788079470198e-05, "loss": 2.1628, "step": 2159 }, { "epoch": 0.504805585580322, "grad_norm": 3.8504951000213623, "learning_rate": 2.5243474873393065e-05, "loss": 2.103, "step": 2160 }, { "epoch": 0.5050392918699425, "grad_norm": 2.8218538761138916, "learning_rate": 2.5255161667315936e-05, "loss": 2.1733, "step": 2161 }, { "epoch": 0.505272998159563, "grad_norm": 3.580767869949341, "learning_rate": 2.52668484612388e-05, "loss": 2.2973, "step": 2162 }, { "epoch": 0.5055067044491834, "grad_norm": 2.679971218109131, "learning_rate": 2.5278535255161666e-05, "loss": 2.2686, "step": 2163 }, { "epoch": 0.505740410738804, "grad_norm": 2.309824228286743, "learning_rate": 2.5290222049084537e-05, "loss": 2.283, "step": 2164 }, { "epoch": 0.5059741170284245, "grad_norm": 2.719669818878174, "learning_rate": 2.53019088430074e-05, "loss": 2.0202, "step": 2165 }, { "epoch": 0.506207823318045, "grad_norm": 3.2217671871185303, "learning_rate": 2.5313595636930268e-05, "loss": 2.1303, "step": 2166 }, { "epoch": 0.5064415296076655, "grad_norm": 5.185467720031738, "learning_rate": 2.5325282430853138e-05, "loss": 2.1315, "step": 2167 }, { "epoch": 0.5066752358972861, "grad_norm": 3.309536933898926, "learning_rate": 2.5336969224776005e-05, "loss": 2.2217, "step": 2168 }, { "epoch": 0.5069089421869066, "grad_norm": 4.039063453674316, "learning_rate": 2.534865601869887e-05, "loss": 2.2301, "step": 2169 }, { "epoch": 0.5071426484765271, "grad_norm": 5.3974151611328125, "learning_rate": 2.536034281262174e-05, "loss": 2.2154, "step": 2170 }, { "epoch": 0.5073763547661476, "grad_norm": 2.960235357284546, "learning_rate": 2.5372029606544606e-05, "loss": 2.1173, "step": 2171 }, { "epoch": 0.5076100610557681, "grad_norm": 3.2524166107177734, "learning_rate": 2.5383716400467473e-05, "loss": 2.2383, "step": 2172 }, { "epoch": 0.5078437673453887, "grad_norm": 5.065480709075928, "learning_rate": 2.539540319439034e-05, "loss": 2.0693, "step": 2173 }, { "epoch": 0.5080774736350092, "grad_norm": 2.588942766189575, "learning_rate": 2.5407089988313207e-05, "loss": 2.0686, "step": 2174 }, { "epoch": 0.5083111799246297, "grad_norm": 3.175208568572998, "learning_rate": 2.5418776782236074e-05, "loss": 2.2472, "step": 2175 }, { "epoch": 0.5085448862142502, "grad_norm": 3.6915953159332275, "learning_rate": 2.543046357615894e-05, "loss": 2.1842, "step": 2176 }, { "epoch": 0.5087785925038708, "grad_norm": 2.526501178741455, "learning_rate": 2.544215037008181e-05, "loss": 2.2454, "step": 2177 }, { "epoch": 0.5090122987934913, "grad_norm": 3.4151501655578613, "learning_rate": 2.5453837164004675e-05, "loss": 2.1454, "step": 2178 }, { "epoch": 0.5092460050831118, "grad_norm": 2.5340874195098877, "learning_rate": 2.5465523957927543e-05, "loss": 2.2051, "step": 2179 }, { "epoch": 0.5094797113727323, "grad_norm": 3.4087893962860107, "learning_rate": 2.547721075185041e-05, "loss": 2.1478, "step": 2180 }, { "epoch": 0.5097134176623528, "grad_norm": 2.4013209342956543, "learning_rate": 2.5488897545773277e-05, "loss": 2.1739, "step": 2181 }, { "epoch": 0.5099471239519734, "grad_norm": 3.9971301555633545, "learning_rate": 2.5500584339696144e-05, "loss": 2.247, "step": 2182 }, { "epoch": 0.5101808302415939, "grad_norm": 3.507291316986084, "learning_rate": 2.551227113361901e-05, "loss": 2.2931, "step": 2183 }, { "epoch": 0.5104145365312144, "grad_norm": 2.381869316101074, "learning_rate": 2.552395792754188e-05, "loss": 2.1918, "step": 2184 }, { "epoch": 0.5106482428208349, "grad_norm": 3.4910006523132324, "learning_rate": 2.5535644721464745e-05, "loss": 2.1467, "step": 2185 }, { "epoch": 0.5108819491104555, "grad_norm": 3.3110921382904053, "learning_rate": 2.5547331515387612e-05, "loss": 2.1506, "step": 2186 }, { "epoch": 0.511115655400076, "grad_norm": 3.5510475635528564, "learning_rate": 2.5559018309310482e-05, "loss": 2.1961, "step": 2187 }, { "epoch": 0.5113493616896965, "grad_norm": 2.8515594005584717, "learning_rate": 2.5570705103233346e-05, "loss": 2.0989, "step": 2188 }, { "epoch": 0.511583067979317, "grad_norm": 2.5655899047851562, "learning_rate": 2.5582391897156213e-05, "loss": 2.1765, "step": 2189 }, { "epoch": 0.5118167742689375, "grad_norm": 4.535583972930908, "learning_rate": 2.5594078691079083e-05, "loss": 2.2549, "step": 2190 }, { "epoch": 0.5120504805585581, "grad_norm": 2.6671698093414307, "learning_rate": 2.5605765485001947e-05, "loss": 2.1948, "step": 2191 }, { "epoch": 0.5122841868481786, "grad_norm": 2.6006970405578613, "learning_rate": 2.5617452278924814e-05, "loss": 2.1317, "step": 2192 }, { "epoch": 0.512517893137799, "grad_norm": 2.439560651779175, "learning_rate": 2.5629139072847685e-05, "loss": 2.3159, "step": 2193 }, { "epoch": 0.5127515994274195, "grad_norm": 3.452054500579834, "learning_rate": 2.5640825866770548e-05, "loss": 2.2398, "step": 2194 }, { "epoch": 0.5129853057170402, "grad_norm": 4.046244144439697, "learning_rate": 2.5652512660693415e-05, "loss": 2.2148, "step": 2195 }, { "epoch": 0.5132190120066606, "grad_norm": 5.220508098602295, "learning_rate": 2.5664199454616286e-05, "loss": 2.1427, "step": 2196 }, { "epoch": 0.5134527182962811, "grad_norm": 2.9452414512634277, "learning_rate": 2.567588624853915e-05, "loss": 2.1828, "step": 2197 }, { "epoch": 0.5136864245859016, "grad_norm": 3.2816991806030273, "learning_rate": 2.5687573042462016e-05, "loss": 2.1672, "step": 2198 }, { "epoch": 0.5139201308755222, "grad_norm": 4.043905258178711, "learning_rate": 2.5699259836384887e-05, "loss": 2.2318, "step": 2199 }, { "epoch": 0.5141538371651427, "grad_norm": 2.7599196434020996, "learning_rate": 2.5710946630307754e-05, "loss": 2.1329, "step": 2200 }, { "epoch": 0.5143875434547632, "grad_norm": 3.294286012649536, "learning_rate": 2.572263342423062e-05, "loss": 2.2459, "step": 2201 }, { "epoch": 0.5146212497443837, "grad_norm": 2.35380220413208, "learning_rate": 2.5734320218153488e-05, "loss": 2.1708, "step": 2202 }, { "epoch": 0.5148549560340042, "grad_norm": 2.6427769660949707, "learning_rate": 2.5746007012076355e-05, "loss": 2.1209, "step": 2203 }, { "epoch": 0.5150886623236248, "grad_norm": 2.8065221309661865, "learning_rate": 2.5757693805999222e-05, "loss": 2.1572, "step": 2204 }, { "epoch": 0.5153223686132453, "grad_norm": 5.078792095184326, "learning_rate": 2.576938059992209e-05, "loss": 2.0858, "step": 2205 }, { "epoch": 0.5155560749028658, "grad_norm": 3.1518356800079346, "learning_rate": 2.5781067393844956e-05, "loss": 2.1424, "step": 2206 }, { "epoch": 0.5157897811924863, "grad_norm": 2.927814483642578, "learning_rate": 2.5792754187767823e-05, "loss": 2.1176, "step": 2207 }, { "epoch": 0.5160234874821069, "grad_norm": 3.1504032611846924, "learning_rate": 2.580444098169069e-05, "loss": 2.1692, "step": 2208 }, { "epoch": 0.5162571937717274, "grad_norm": 3.2908363342285156, "learning_rate": 2.5816127775613557e-05, "loss": 2.1338, "step": 2209 }, { "epoch": 0.5164909000613479, "grad_norm": 3.2791049480438232, "learning_rate": 2.5827814569536424e-05, "loss": 2.1347, "step": 2210 }, { "epoch": 0.5167246063509684, "grad_norm": 2.729269504547119, "learning_rate": 2.583950136345929e-05, "loss": 2.1165, "step": 2211 }, { "epoch": 0.5169583126405889, "grad_norm": 4.493456840515137, "learning_rate": 2.585118815738216e-05, "loss": 2.1492, "step": 2212 }, { "epoch": 0.5171920189302095, "grad_norm": 2.917104721069336, "learning_rate": 2.5862874951305026e-05, "loss": 2.168, "step": 2213 }, { "epoch": 0.51742572521983, "grad_norm": 2.752470016479492, "learning_rate": 2.5874561745227893e-05, "loss": 2.176, "step": 2214 }, { "epoch": 0.5176594315094505, "grad_norm": 4.404149055480957, "learning_rate": 2.588624853915076e-05, "loss": 2.1367, "step": 2215 }, { "epoch": 0.517893137799071, "grad_norm": 6.702399730682373, "learning_rate": 2.589793533307363e-05, "loss": 2.0941, "step": 2216 }, { "epoch": 0.5181268440886916, "grad_norm": 4.2174787521362305, "learning_rate": 2.5909622126996494e-05, "loss": 2.1217, "step": 2217 }, { "epoch": 0.5183605503783121, "grad_norm": 3.211970329284668, "learning_rate": 2.592130892091936e-05, "loss": 2.1925, "step": 2218 }, { "epoch": 0.5185942566679326, "grad_norm": 3.9498913288116455, "learning_rate": 2.593299571484223e-05, "loss": 2.1631, "step": 2219 }, { "epoch": 0.5188279629575531, "grad_norm": 2.640514373779297, "learning_rate": 2.5944682508765095e-05, "loss": 2.1922, "step": 2220 }, { "epoch": 0.5190616692471736, "grad_norm": 2.6358397006988525, "learning_rate": 2.5956369302687962e-05, "loss": 2.2848, "step": 2221 }, { "epoch": 0.5192953755367942, "grad_norm": 2.8177220821380615, "learning_rate": 2.5968056096610832e-05, "loss": 2.2095, "step": 2222 }, { "epoch": 0.5195290818264147, "grad_norm": 2.719440221786499, "learning_rate": 2.5979742890533696e-05, "loss": 2.2617, "step": 2223 }, { "epoch": 0.5197627881160352, "grad_norm": 2.8127541542053223, "learning_rate": 2.5991429684456563e-05, "loss": 2.1884, "step": 2224 }, { "epoch": 0.5199964944056557, "grad_norm": 3.0078983306884766, "learning_rate": 2.6003116478379433e-05, "loss": 2.309, "step": 2225 }, { "epoch": 0.5202302006952763, "grad_norm": 5.873018264770508, "learning_rate": 2.6014803272302297e-05, "loss": 2.1536, "step": 2226 }, { "epoch": 0.5204639069848968, "grad_norm": 5.439422607421875, "learning_rate": 2.6026490066225168e-05, "loss": 2.2097, "step": 2227 }, { "epoch": 0.5206976132745172, "grad_norm": 3.2012734413146973, "learning_rate": 2.6038176860148035e-05, "loss": 2.248, "step": 2228 }, { "epoch": 0.5209313195641377, "grad_norm": 3.086343765258789, "learning_rate": 2.6049863654070898e-05, "loss": 2.2304, "step": 2229 }, { "epoch": 0.5211650258537582, "grad_norm": 2.88130784034729, "learning_rate": 2.606155044799377e-05, "loss": 2.2124, "step": 2230 }, { "epoch": 0.5213987321433788, "grad_norm": 3.125000238418579, "learning_rate": 2.6073237241916636e-05, "loss": 2.1141, "step": 2231 }, { "epoch": 0.5216324384329993, "grad_norm": 2.35744047164917, "learning_rate": 2.6084924035839503e-05, "loss": 2.1371, "step": 2232 }, { "epoch": 0.5218661447226198, "grad_norm": 2.9730064868927, "learning_rate": 2.609661082976237e-05, "loss": 2.1298, "step": 2233 }, { "epoch": 0.5220998510122403, "grad_norm": 3.0633959770202637, "learning_rate": 2.6108297623685237e-05, "loss": 2.3307, "step": 2234 }, { "epoch": 0.5223335573018609, "grad_norm": 2.4943277835845947, "learning_rate": 2.6119984417608104e-05, "loss": 2.1947, "step": 2235 }, { "epoch": 0.5225672635914814, "grad_norm": 2.52315354347229, "learning_rate": 2.613167121153097e-05, "loss": 2.2604, "step": 2236 }, { "epoch": 0.5228009698811019, "grad_norm": 2.940913677215576, "learning_rate": 2.6143358005453838e-05, "loss": 2.0715, "step": 2237 }, { "epoch": 0.5230346761707224, "grad_norm": 2.6245384216308594, "learning_rate": 2.6155044799376705e-05, "loss": 2.1728, "step": 2238 }, { "epoch": 0.523268382460343, "grad_norm": 3.019918203353882, "learning_rate": 2.6166731593299572e-05, "loss": 2.1027, "step": 2239 }, { "epoch": 0.5235020887499635, "grad_norm": 2.44087290763855, "learning_rate": 2.617841838722244e-05, "loss": 2.0302, "step": 2240 }, { "epoch": 0.523735795039584, "grad_norm": 2.5202062129974365, "learning_rate": 2.6190105181145306e-05, "loss": 2.3122, "step": 2241 }, { "epoch": 0.5239695013292045, "grad_norm": 3.031083345413208, "learning_rate": 2.6201791975068173e-05, "loss": 2.1247, "step": 2242 }, { "epoch": 0.524203207618825, "grad_norm": 2.711836099624634, "learning_rate": 2.621347876899104e-05, "loss": 2.1778, "step": 2243 }, { "epoch": 0.5244369139084456, "grad_norm": 2.9822580814361572, "learning_rate": 2.6225165562913907e-05, "loss": 2.0413, "step": 2244 }, { "epoch": 0.5246706201980661, "grad_norm": 2.6571247577667236, "learning_rate": 2.6236852356836774e-05, "loss": 2.1542, "step": 2245 }, { "epoch": 0.5249043264876866, "grad_norm": 2.69323992729187, "learning_rate": 2.624853915075964e-05, "loss": 2.2399, "step": 2246 }, { "epoch": 0.5251380327773071, "grad_norm": 2.7768781185150146, "learning_rate": 2.626022594468251e-05, "loss": 2.1371, "step": 2247 }, { "epoch": 0.5253717390669277, "grad_norm": 3.4779319763183594, "learning_rate": 2.627191273860538e-05, "loss": 2.1103, "step": 2248 }, { "epoch": 0.5256054453565482, "grad_norm": 2.6631529331207275, "learning_rate": 2.6283599532528243e-05, "loss": 2.1682, "step": 2249 }, { "epoch": 0.5258391516461687, "grad_norm": 4.572608470916748, "learning_rate": 2.629528632645111e-05, "loss": 2.3522, "step": 2250 }, { "epoch": 0.5260728579357892, "grad_norm": 3.2518768310546875, "learning_rate": 2.630697312037398e-05, "loss": 2.2384, "step": 2251 }, { "epoch": 0.5263065642254097, "grad_norm": 2.470762014389038, "learning_rate": 2.6318659914296844e-05, "loss": 2.2167, "step": 2252 }, { "epoch": 0.5265402705150303, "grad_norm": 4.183166027069092, "learning_rate": 2.6330346708219714e-05, "loss": 2.1935, "step": 2253 }, { "epoch": 0.5267739768046508, "grad_norm": 3.456010580062866, "learning_rate": 2.634203350214258e-05, "loss": 2.0733, "step": 2254 }, { "epoch": 0.5270076830942713, "grad_norm": 2.9213569164276123, "learning_rate": 2.6353720296065445e-05, "loss": 2.283, "step": 2255 }, { "epoch": 0.5272413893838918, "grad_norm": 3.80102276802063, "learning_rate": 2.6365407089988315e-05, "loss": 2.3788, "step": 2256 }, { "epoch": 0.5274750956735124, "grad_norm": 2.5271711349487305, "learning_rate": 2.6377093883911182e-05, "loss": 2.1881, "step": 2257 }, { "epoch": 0.5277088019631329, "grad_norm": 2.616734027862549, "learning_rate": 2.6388780677834046e-05, "loss": 2.199, "step": 2258 }, { "epoch": 0.5279425082527534, "grad_norm": 2.5961644649505615, "learning_rate": 2.6400467471756916e-05, "loss": 2.1397, "step": 2259 }, { "epoch": 0.5281762145423738, "grad_norm": 2.5145599842071533, "learning_rate": 2.6412154265679783e-05, "loss": 2.1065, "step": 2260 }, { "epoch": 0.5284099208319943, "grad_norm": 2.628901243209839, "learning_rate": 2.6423841059602647e-05, "loss": 2.0925, "step": 2261 }, { "epoch": 0.528643627121615, "grad_norm": 2.647941827774048, "learning_rate": 2.6435527853525518e-05, "loss": 2.1223, "step": 2262 }, { "epoch": 0.5288773334112354, "grad_norm": 4.278921127319336, "learning_rate": 2.6447214647448385e-05, "loss": 2.0242, "step": 2263 }, { "epoch": 0.5291110397008559, "grad_norm": 2.904160499572754, "learning_rate": 2.645890144137125e-05, "loss": 2.1463, "step": 2264 }, { "epoch": 0.5293447459904764, "grad_norm": 2.724797010421753, "learning_rate": 2.647058823529412e-05, "loss": 2.0944, "step": 2265 }, { "epoch": 0.529578452280097, "grad_norm": 3.432840347290039, "learning_rate": 2.6482275029216986e-05, "loss": 2.2764, "step": 2266 }, { "epoch": 0.5298121585697175, "grad_norm": 2.827646017074585, "learning_rate": 2.6493961823139853e-05, "loss": 2.092, "step": 2267 }, { "epoch": 0.530045864859338, "grad_norm": 3.542837619781494, "learning_rate": 2.650564861706272e-05, "loss": 2.3561, "step": 2268 }, { "epoch": 0.5302795711489585, "grad_norm": 3.4905312061309814, "learning_rate": 2.6517335410985587e-05, "loss": 2.1795, "step": 2269 }, { "epoch": 0.530513277438579, "grad_norm": 2.630706787109375, "learning_rate": 2.6529022204908454e-05, "loss": 2.1385, "step": 2270 }, { "epoch": 0.5307469837281996, "grad_norm": 3.2232234477996826, "learning_rate": 2.654070899883132e-05, "loss": 2.1388, "step": 2271 }, { "epoch": 0.5309806900178201, "grad_norm": 3.9522624015808105, "learning_rate": 2.6552395792754188e-05, "loss": 2.1411, "step": 2272 }, { "epoch": 0.5312143963074406, "grad_norm": 2.7879254817962646, "learning_rate": 2.6564082586677055e-05, "loss": 2.1128, "step": 2273 }, { "epoch": 0.5314481025970611, "grad_norm": 3.5809855461120605, "learning_rate": 2.6575769380599922e-05, "loss": 2.2401, "step": 2274 }, { "epoch": 0.5316818088866817, "grad_norm": 3.738654136657715, "learning_rate": 2.658745617452279e-05, "loss": 2.2057, "step": 2275 }, { "epoch": 0.5319155151763022, "grad_norm": 3.7559080123901367, "learning_rate": 2.6599142968445656e-05, "loss": 2.3364, "step": 2276 }, { "epoch": 0.5321492214659227, "grad_norm": 5.806914329528809, "learning_rate": 2.6610829762368523e-05, "loss": 2.3645, "step": 2277 }, { "epoch": 0.5323829277555432, "grad_norm": 3.213313102722168, "learning_rate": 2.662251655629139e-05, "loss": 2.1666, "step": 2278 }, { "epoch": 0.5326166340451638, "grad_norm": 5.232606887817383, "learning_rate": 2.6634203350214257e-05, "loss": 2.2504, "step": 2279 }, { "epoch": 0.5328503403347843, "grad_norm": 3.821277141571045, "learning_rate": 2.6645890144137128e-05, "loss": 2.1847, "step": 2280 }, { "epoch": 0.5330840466244048, "grad_norm": 2.923469305038452, "learning_rate": 2.665757693805999e-05, "loss": 2.0786, "step": 2281 }, { "epoch": 0.5333177529140253, "grad_norm": 2.956148624420166, "learning_rate": 2.6669263731982862e-05, "loss": 2.1162, "step": 2282 }, { "epoch": 0.5335514592036458, "grad_norm": 3.1233131885528564, "learning_rate": 2.668095052590573e-05, "loss": 2.1581, "step": 2283 }, { "epoch": 0.5337851654932664, "grad_norm": 3.2639780044555664, "learning_rate": 2.6692637319828593e-05, "loss": 2.0231, "step": 2284 }, { "epoch": 0.5340188717828869, "grad_norm": 3.5521180629730225, "learning_rate": 2.6704324113751463e-05, "loss": 2.3259, "step": 2285 }, { "epoch": 0.5342525780725074, "grad_norm": 4.078639030456543, "learning_rate": 2.671601090767433e-05, "loss": 2.2076, "step": 2286 }, { "epoch": 0.5344862843621279, "grad_norm": 3.3295607566833496, "learning_rate": 2.6727697701597194e-05, "loss": 2.0984, "step": 2287 }, { "epoch": 0.5347199906517485, "grad_norm": 4.48197078704834, "learning_rate": 2.6739384495520064e-05, "loss": 2.2753, "step": 2288 }, { "epoch": 0.534953696941369, "grad_norm": 5.509686470031738, "learning_rate": 2.675107128944293e-05, "loss": 2.1645, "step": 2289 }, { "epoch": 0.5351874032309895, "grad_norm": 3.0880885124206543, "learning_rate": 2.6762758083365795e-05, "loss": 2.2864, "step": 2290 }, { "epoch": 0.53542110952061, "grad_norm": 2.848376750946045, "learning_rate": 2.6774444877288665e-05, "loss": 2.1903, "step": 2291 }, { "epoch": 0.5356548158102304, "grad_norm": 4.597163200378418, "learning_rate": 2.6786131671211532e-05, "loss": 2.1314, "step": 2292 }, { "epoch": 0.535888522099851, "grad_norm": 4.1699419021606445, "learning_rate": 2.6797818465134396e-05, "loss": 2.2478, "step": 2293 }, { "epoch": 0.5361222283894715, "grad_norm": 2.827033758163452, "learning_rate": 2.6809505259057266e-05, "loss": 2.1374, "step": 2294 }, { "epoch": 0.536355934679092, "grad_norm": 6.774947166442871, "learning_rate": 2.6821192052980134e-05, "loss": 2.2007, "step": 2295 }, { "epoch": 0.5365896409687125, "grad_norm": 7.040114402770996, "learning_rate": 2.6832878846903e-05, "loss": 2.0978, "step": 2296 }, { "epoch": 0.5368233472583331, "grad_norm": 4.7586164474487305, "learning_rate": 2.6844565640825868e-05, "loss": 2.1, "step": 2297 }, { "epoch": 0.5370570535479536, "grad_norm": 6.121994495391846, "learning_rate": 2.6856252434748735e-05, "loss": 2.1329, "step": 2298 }, { "epoch": 0.5372907598375741, "grad_norm": 6.781256198883057, "learning_rate": 2.6867939228671602e-05, "loss": 2.2109, "step": 2299 }, { "epoch": 0.5375244661271946, "grad_norm": 5.69451379776001, "learning_rate": 2.687962602259447e-05, "loss": 2.301, "step": 2300 }, { "epoch": 0.5377581724168151, "grad_norm": 2.941096782684326, "learning_rate": 2.6891312816517336e-05, "loss": 2.2496, "step": 2301 }, { "epoch": 0.5379918787064357, "grad_norm": 5.286913871765137, "learning_rate": 2.6902999610440203e-05, "loss": 2.1542, "step": 2302 }, { "epoch": 0.5382255849960562, "grad_norm": 8.79991626739502, "learning_rate": 2.691468640436307e-05, "loss": 2.2507, "step": 2303 }, { "epoch": 0.5384592912856767, "grad_norm": 7.684869289398193, "learning_rate": 2.6926373198285937e-05, "loss": 2.1462, "step": 2304 }, { "epoch": 0.5386929975752972, "grad_norm": 3.4352214336395264, "learning_rate": 2.6938059992208804e-05, "loss": 2.2627, "step": 2305 }, { "epoch": 0.5389267038649178, "grad_norm": 3.913557291030884, "learning_rate": 2.694974678613167e-05, "loss": 2.2733, "step": 2306 }, { "epoch": 0.5391604101545383, "grad_norm": 8.024800300598145, "learning_rate": 2.6961433580054538e-05, "loss": 2.0632, "step": 2307 }, { "epoch": 0.5393941164441588, "grad_norm": 7.429574966430664, "learning_rate": 2.697312037397741e-05, "loss": 2.1866, "step": 2308 }, { "epoch": 0.5396278227337793, "grad_norm": 5.802492618560791, "learning_rate": 2.6984807167900276e-05, "loss": 2.2643, "step": 2309 }, { "epoch": 0.5398615290233998, "grad_norm": 4.2379841804504395, "learning_rate": 2.699649396182314e-05, "loss": 2.2252, "step": 2310 }, { "epoch": 0.5400952353130204, "grad_norm": 6.661357402801514, "learning_rate": 2.700818075574601e-05, "loss": 2.0971, "step": 2311 }, { "epoch": 0.5403289416026409, "grad_norm": 5.492388725280762, "learning_rate": 2.7019867549668877e-05, "loss": 2.298, "step": 2312 }, { "epoch": 0.5405626478922614, "grad_norm": 3.260664701461792, "learning_rate": 2.703155434359174e-05, "loss": 2.1069, "step": 2313 }, { "epoch": 0.5407963541818819, "grad_norm": 4.623117923736572, "learning_rate": 2.704324113751461e-05, "loss": 2.1895, "step": 2314 }, { "epoch": 0.5410300604715025, "grad_norm": 4.166358470916748, "learning_rate": 2.7054927931437478e-05, "loss": 2.166, "step": 2315 }, { "epoch": 0.541263766761123, "grad_norm": 3.790564775466919, "learning_rate": 2.706661472536034e-05, "loss": 2.0927, "step": 2316 }, { "epoch": 0.5414974730507435, "grad_norm": 3.7851271629333496, "learning_rate": 2.7078301519283212e-05, "loss": 2.0753, "step": 2317 }, { "epoch": 0.541731179340364, "grad_norm": 4.774055004119873, "learning_rate": 2.708998831320608e-05, "loss": 2.1131, "step": 2318 }, { "epoch": 0.5419648856299846, "grad_norm": 2.7086424827575684, "learning_rate": 2.7101675107128943e-05, "loss": 2.3077, "step": 2319 }, { "epoch": 0.5421985919196051, "grad_norm": 2.877131223678589, "learning_rate": 2.7113361901051813e-05, "loss": 2.1638, "step": 2320 }, { "epoch": 0.5424322982092256, "grad_norm": 3.0546956062316895, "learning_rate": 2.712504869497468e-05, "loss": 2.1262, "step": 2321 }, { "epoch": 0.542666004498846, "grad_norm": 2.4742352962493896, "learning_rate": 2.7136735488897544e-05, "loss": 2.1292, "step": 2322 }, { "epoch": 0.5428997107884665, "grad_norm": 3.5513319969177246, "learning_rate": 2.7148422282820414e-05, "loss": 2.2102, "step": 2323 }, { "epoch": 0.5431334170780872, "grad_norm": 2.6665804386138916, "learning_rate": 2.716010907674328e-05, "loss": 2.0767, "step": 2324 }, { "epoch": 0.5433671233677076, "grad_norm": 2.474043130874634, "learning_rate": 2.717179587066615e-05, "loss": 2.1586, "step": 2325 }, { "epoch": 0.5436008296573281, "grad_norm": 2.4898228645324707, "learning_rate": 2.7183482664589015e-05, "loss": 1.9741, "step": 2326 }, { "epoch": 0.5438345359469486, "grad_norm": 2.588074207305908, "learning_rate": 2.7195169458511882e-05, "loss": 2.1276, "step": 2327 }, { "epoch": 0.5440682422365692, "grad_norm": 2.8259589672088623, "learning_rate": 2.720685625243475e-05, "loss": 2.1455, "step": 2328 }, { "epoch": 0.5443019485261897, "grad_norm": 2.6817939281463623, "learning_rate": 2.7218543046357616e-05, "loss": 2.0193, "step": 2329 }, { "epoch": 0.5445356548158102, "grad_norm": 3.058067798614502, "learning_rate": 2.7230229840280484e-05, "loss": 2.167, "step": 2330 }, { "epoch": 0.5447693611054307, "grad_norm": 3.891463279724121, "learning_rate": 2.724191663420335e-05, "loss": 2.0765, "step": 2331 }, { "epoch": 0.5450030673950512, "grad_norm": 4.128868579864502, "learning_rate": 2.7253603428126218e-05, "loss": 2.0954, "step": 2332 }, { "epoch": 0.5452367736846718, "grad_norm": 2.975797653198242, "learning_rate": 2.7265290222049085e-05, "loss": 2.1991, "step": 2333 }, { "epoch": 0.5454704799742923, "grad_norm": 3.466961622238159, "learning_rate": 2.7276977015971952e-05, "loss": 2.0778, "step": 2334 }, { "epoch": 0.5457041862639128, "grad_norm": 3.010869264602661, "learning_rate": 2.728866380989482e-05, "loss": 2.1147, "step": 2335 }, { "epoch": 0.5459378925535333, "grad_norm": 3.4614980220794678, "learning_rate": 2.7300350603817686e-05, "loss": 2.2703, "step": 2336 }, { "epoch": 0.5461715988431539, "grad_norm": 3.2410390377044678, "learning_rate": 2.7312037397740556e-05, "loss": 2.1652, "step": 2337 }, { "epoch": 0.5464053051327744, "grad_norm": 3.071322202682495, "learning_rate": 2.732372419166342e-05, "loss": 2.3884, "step": 2338 }, { "epoch": 0.5466390114223949, "grad_norm": 2.587831735610962, "learning_rate": 2.7335410985586287e-05, "loss": 2.2391, "step": 2339 }, { "epoch": 0.5468727177120154, "grad_norm": 2.471184015274048, "learning_rate": 2.7347097779509157e-05, "loss": 2.1724, "step": 2340 }, { "epoch": 0.5471064240016359, "grad_norm": 3.5572288036346436, "learning_rate": 2.7358784573432024e-05, "loss": 2.2564, "step": 2341 }, { "epoch": 0.5473401302912565, "grad_norm": 2.550633430480957, "learning_rate": 2.7370471367354888e-05, "loss": 2.187, "step": 2342 }, { "epoch": 0.547573836580877, "grad_norm": 3.0593698024749756, "learning_rate": 2.738215816127776e-05, "loss": 2.207, "step": 2343 }, { "epoch": 0.5478075428704975, "grad_norm": 2.9830429553985596, "learning_rate": 2.7393844955200626e-05, "loss": 1.9944, "step": 2344 }, { "epoch": 0.548041249160118, "grad_norm": 2.8285000324249268, "learning_rate": 2.740553174912349e-05, "loss": 2.0792, "step": 2345 }, { "epoch": 0.5482749554497386, "grad_norm": 3.3514227867126465, "learning_rate": 2.741721854304636e-05, "loss": 2.1198, "step": 2346 }, { "epoch": 0.5485086617393591, "grad_norm": 3.801743745803833, "learning_rate": 2.7428905336969227e-05, "loss": 2.1641, "step": 2347 }, { "epoch": 0.5487423680289796, "grad_norm": 2.6850740909576416, "learning_rate": 2.744059213089209e-05, "loss": 2.2937, "step": 2348 }, { "epoch": 0.5489760743186001, "grad_norm": 3.981412410736084, "learning_rate": 2.745227892481496e-05, "loss": 2.0148, "step": 2349 }, { "epoch": 0.5492097806082206, "grad_norm": 4.1551408767700195, "learning_rate": 2.7463965718737828e-05, "loss": 2.1266, "step": 2350 }, { "epoch": 0.5494434868978412, "grad_norm": 3.4520533084869385, "learning_rate": 2.747565251266069e-05, "loss": 2.194, "step": 2351 }, { "epoch": 0.5496771931874617, "grad_norm": 4.90723991394043, "learning_rate": 2.7487339306583562e-05, "loss": 2.4858, "step": 2352 }, { "epoch": 0.5499108994770822, "grad_norm": 5.305671691894531, "learning_rate": 2.749902610050643e-05, "loss": 2.1439, "step": 2353 }, { "epoch": 0.5501446057667027, "grad_norm": 3.7922425270080566, "learning_rate": 2.7510712894429293e-05, "loss": 2.1548, "step": 2354 }, { "epoch": 0.5503783120563233, "grad_norm": 4.421918869018555, "learning_rate": 2.7522399688352163e-05, "loss": 2.1734, "step": 2355 }, { "epoch": 0.5506120183459438, "grad_norm": 3.0494143962860107, "learning_rate": 2.753408648227503e-05, "loss": 2.0142, "step": 2356 }, { "epoch": 0.5508457246355642, "grad_norm": 4.033827304840088, "learning_rate": 2.7545773276197897e-05, "loss": 2.1378, "step": 2357 }, { "epoch": 0.5510794309251847, "grad_norm": 3.214677095413208, "learning_rate": 2.7557460070120764e-05, "loss": 2.0556, "step": 2358 }, { "epoch": 0.5513131372148053, "grad_norm": 4.440398693084717, "learning_rate": 2.756914686404363e-05, "loss": 2.1889, "step": 2359 }, { "epoch": 0.5515468435044258, "grad_norm": 2.9814534187316895, "learning_rate": 2.75808336579665e-05, "loss": 2.0326, "step": 2360 }, { "epoch": 0.5517805497940463, "grad_norm": 2.8428385257720947, "learning_rate": 2.7592520451889365e-05, "loss": 2.2053, "step": 2361 }, { "epoch": 0.5520142560836668, "grad_norm": 2.653456449508667, "learning_rate": 2.7604207245812232e-05, "loss": 2.2456, "step": 2362 }, { "epoch": 0.5522479623732873, "grad_norm": 3.0831074714660645, "learning_rate": 2.7615894039735103e-05, "loss": 2.2814, "step": 2363 }, { "epoch": 0.5524816686629079, "grad_norm": 2.9387760162353516, "learning_rate": 2.7627580833657967e-05, "loss": 2.0905, "step": 2364 }, { "epoch": 0.5527153749525284, "grad_norm": 3.4398131370544434, "learning_rate": 2.7639267627580834e-05, "loss": 1.9913, "step": 2365 }, { "epoch": 0.5529490812421489, "grad_norm": 3.740269660949707, "learning_rate": 2.7650954421503704e-05, "loss": 2.2809, "step": 2366 }, { "epoch": 0.5531827875317694, "grad_norm": 3.424461603164673, "learning_rate": 2.7662641215426568e-05, "loss": 2.3017, "step": 2367 }, { "epoch": 0.55341649382139, "grad_norm": 3.983975648880005, "learning_rate": 2.7674328009349435e-05, "loss": 2.0173, "step": 2368 }, { "epoch": 0.5536502001110105, "grad_norm": 2.7730178833007812, "learning_rate": 2.7686014803272305e-05, "loss": 2.151, "step": 2369 }, { "epoch": 0.553883906400631, "grad_norm": 2.488633155822754, "learning_rate": 2.769770159719517e-05, "loss": 1.9484, "step": 2370 }, { "epoch": 0.5541176126902515, "grad_norm": 2.51981520652771, "learning_rate": 2.7709388391118036e-05, "loss": 2.177, "step": 2371 }, { "epoch": 0.554351318979872, "grad_norm": 2.691351890563965, "learning_rate": 2.7721075185040906e-05, "loss": 2.2748, "step": 2372 }, { "epoch": 0.5545850252694926, "grad_norm": 2.454040288925171, "learning_rate": 2.7732761978963773e-05, "loss": 2.2262, "step": 2373 }, { "epoch": 0.5548187315591131, "grad_norm": 4.283604145050049, "learning_rate": 2.7744448772886637e-05, "loss": 2.332, "step": 2374 }, { "epoch": 0.5550524378487336, "grad_norm": 2.8588852882385254, "learning_rate": 2.7756135566809507e-05, "loss": 2.2459, "step": 2375 }, { "epoch": 0.5552861441383541, "grad_norm": 2.4599437713623047, "learning_rate": 2.7767822360732374e-05, "loss": 2.131, "step": 2376 }, { "epoch": 0.5555198504279747, "grad_norm": 3.4006288051605225, "learning_rate": 2.7779509154655238e-05, "loss": 1.9738, "step": 2377 }, { "epoch": 0.5557535567175952, "grad_norm": 3.0084915161132812, "learning_rate": 2.779119594857811e-05, "loss": 2.0772, "step": 2378 }, { "epoch": 0.5559872630072157, "grad_norm": 2.9962828159332275, "learning_rate": 2.7802882742500976e-05, "loss": 2.1244, "step": 2379 }, { "epoch": 0.5562209692968362, "grad_norm": 3.39836049079895, "learning_rate": 2.781456953642384e-05, "loss": 2.1936, "step": 2380 }, { "epoch": 0.5564546755864567, "grad_norm": 2.7771658897399902, "learning_rate": 2.782625633034671e-05, "loss": 2.1751, "step": 2381 }, { "epoch": 0.5566883818760773, "grad_norm": 3.655604600906372, "learning_rate": 2.7837943124269577e-05, "loss": 2.0701, "step": 2382 }, { "epoch": 0.5569220881656978, "grad_norm": 3.3601090908050537, "learning_rate": 2.784962991819244e-05, "loss": 2.2097, "step": 2383 }, { "epoch": 0.5571557944553183, "grad_norm": 3.014648675918579, "learning_rate": 2.786131671211531e-05, "loss": 1.9624, "step": 2384 }, { "epoch": 0.5573895007449388, "grad_norm": 4.595667839050293, "learning_rate": 2.7873003506038178e-05, "loss": 2.2942, "step": 2385 }, { "epoch": 0.5576232070345594, "grad_norm": 3.616365909576416, "learning_rate": 2.788469029996104e-05, "loss": 2.1126, "step": 2386 }, { "epoch": 0.5578569133241799, "grad_norm": 2.5228993892669678, "learning_rate": 2.7896377093883912e-05, "loss": 2.1358, "step": 2387 }, { "epoch": 0.5580906196138004, "grad_norm": 2.7745771408081055, "learning_rate": 2.790806388780678e-05, "loss": 2.1873, "step": 2388 }, { "epoch": 0.5583243259034208, "grad_norm": 2.4881720542907715, "learning_rate": 2.7919750681729646e-05, "loss": 2.0751, "step": 2389 }, { "epoch": 0.5585580321930413, "grad_norm": 2.8162777423858643, "learning_rate": 2.7931437475652513e-05, "loss": 2.0936, "step": 2390 }, { "epoch": 0.558791738482662, "grad_norm": 4.017678260803223, "learning_rate": 2.794312426957538e-05, "loss": 2.3341, "step": 2391 }, { "epoch": 0.5590254447722824, "grad_norm": 3.034943103790283, "learning_rate": 2.795481106349825e-05, "loss": 2.2152, "step": 2392 }, { "epoch": 0.5592591510619029, "grad_norm": 2.763063430786133, "learning_rate": 2.7966497857421114e-05, "loss": 2.2009, "step": 2393 }, { "epoch": 0.5594928573515234, "grad_norm": 2.5668673515319824, "learning_rate": 2.797818465134398e-05, "loss": 2.046, "step": 2394 }, { "epoch": 0.559726563641144, "grad_norm": 3.477714776992798, "learning_rate": 2.7989871445266852e-05, "loss": 2.2393, "step": 2395 }, { "epoch": 0.5599602699307645, "grad_norm": 3.201491594314575, "learning_rate": 2.8001558239189715e-05, "loss": 2.0082, "step": 2396 }, { "epoch": 0.560193976220385, "grad_norm": 4.208128929138184, "learning_rate": 2.8013245033112582e-05, "loss": 2.1549, "step": 2397 }, { "epoch": 0.5604276825100055, "grad_norm": 5.8077569007873535, "learning_rate": 2.8024931827035453e-05, "loss": 2.2014, "step": 2398 }, { "epoch": 0.560661388799626, "grad_norm": 3.014476776123047, "learning_rate": 2.8036618620958317e-05, "loss": 2.2659, "step": 2399 }, { "epoch": 0.5608950950892466, "grad_norm": 4.083223342895508, "learning_rate": 2.8048305414881184e-05, "loss": 2.093, "step": 2400 }, { "epoch": 0.5611288013788671, "grad_norm": 4.3579020500183105, "learning_rate": 2.8059992208804054e-05, "loss": 2.3788, "step": 2401 }, { "epoch": 0.5613625076684876, "grad_norm": 2.9448251724243164, "learning_rate": 2.8071679002726918e-05, "loss": 2.0947, "step": 2402 }, { "epoch": 0.5615962139581081, "grad_norm": 4.830577850341797, "learning_rate": 2.8083365796649785e-05, "loss": 2.2161, "step": 2403 }, { "epoch": 0.5618299202477287, "grad_norm": 6.260195732116699, "learning_rate": 2.8095052590572655e-05, "loss": 2.315, "step": 2404 }, { "epoch": 0.5620636265373492, "grad_norm": 3.2091450691223145, "learning_rate": 2.8106739384495522e-05, "loss": 2.1352, "step": 2405 }, { "epoch": 0.5622973328269697, "grad_norm": 4.523221492767334, "learning_rate": 2.8118426178418386e-05, "loss": 2.2307, "step": 2406 }, { "epoch": 0.5625310391165902, "grad_norm": 6.6581878662109375, "learning_rate": 2.8130112972341256e-05, "loss": 2.2226, "step": 2407 }, { "epoch": 0.5627647454062108, "grad_norm": 3.722830057144165, "learning_rate": 2.8141799766264123e-05, "loss": 1.9927, "step": 2408 }, { "epoch": 0.5629984516958313, "grad_norm": 4.873662948608398, "learning_rate": 2.8153486560186987e-05, "loss": 2.3211, "step": 2409 }, { "epoch": 0.5632321579854518, "grad_norm": 4.86455774307251, "learning_rate": 2.8165173354109857e-05, "loss": 2.1848, "step": 2410 }, { "epoch": 0.5634658642750723, "grad_norm": 4.253104209899902, "learning_rate": 2.8176860148032725e-05, "loss": 2.105, "step": 2411 }, { "epoch": 0.5636995705646928, "grad_norm": 4.090255260467529, "learning_rate": 2.8188546941955588e-05, "loss": 2.1764, "step": 2412 }, { "epoch": 0.5639332768543134, "grad_norm": 3.759225368499756, "learning_rate": 2.820023373587846e-05, "loss": 2.094, "step": 2413 }, { "epoch": 0.5641669831439339, "grad_norm": 2.5006308555603027, "learning_rate": 2.8211920529801326e-05, "loss": 2.3371, "step": 2414 }, { "epoch": 0.5644006894335544, "grad_norm": 4.22935676574707, "learning_rate": 2.822360732372419e-05, "loss": 2.0755, "step": 2415 }, { "epoch": 0.5646343957231749, "grad_norm": 3.4602813720703125, "learning_rate": 2.823529411764706e-05, "loss": 2.2383, "step": 2416 }, { "epoch": 0.5648681020127955, "grad_norm": 2.468663215637207, "learning_rate": 2.8246980911569927e-05, "loss": 2.1559, "step": 2417 }, { "epoch": 0.565101808302416, "grad_norm": 3.834351062774658, "learning_rate": 2.8258667705492794e-05, "loss": 1.976, "step": 2418 }, { "epoch": 0.5653355145920365, "grad_norm": 4.817743301391602, "learning_rate": 2.827035449941566e-05, "loss": 2.0809, "step": 2419 }, { "epoch": 0.565569220881657, "grad_norm": 2.7215569019317627, "learning_rate": 2.8282041293338528e-05, "loss": 2.0705, "step": 2420 }, { "epoch": 0.5658029271712774, "grad_norm": 4.8717122077941895, "learning_rate": 2.82937280872614e-05, "loss": 2.1825, "step": 2421 }, { "epoch": 0.566036633460898, "grad_norm": 2.7486326694488525, "learning_rate": 2.8305414881184262e-05, "loss": 2.076, "step": 2422 }, { "epoch": 0.5662703397505185, "grad_norm": 2.617115020751953, "learning_rate": 2.831710167510713e-05, "loss": 2.0584, "step": 2423 }, { "epoch": 0.566504046040139, "grad_norm": 3.318089008331299, "learning_rate": 2.832878846903e-05, "loss": 2.0232, "step": 2424 }, { "epoch": 0.5667377523297595, "grad_norm": 4.134700775146484, "learning_rate": 2.8340475262952863e-05, "loss": 2.043, "step": 2425 }, { "epoch": 0.5669714586193801, "grad_norm": 3.008373260498047, "learning_rate": 2.835216205687573e-05, "loss": 2.157, "step": 2426 }, { "epoch": 0.5672051649090006, "grad_norm": 2.3605434894561768, "learning_rate": 2.83638488507986e-05, "loss": 2.1691, "step": 2427 }, { "epoch": 0.5674388711986211, "grad_norm": 3.027925491333008, "learning_rate": 2.8375535644721464e-05, "loss": 1.9663, "step": 2428 }, { "epoch": 0.5676725774882416, "grad_norm": 2.6626899242401123, "learning_rate": 2.838722243864433e-05, "loss": 2.1661, "step": 2429 }, { "epoch": 0.5679062837778621, "grad_norm": 3.1282644271850586, "learning_rate": 2.8398909232567202e-05, "loss": 2.0372, "step": 2430 }, { "epoch": 0.5681399900674827, "grad_norm": 2.8540027141571045, "learning_rate": 2.8410596026490065e-05, "loss": 2.0635, "step": 2431 }, { "epoch": 0.5683736963571032, "grad_norm": 2.568732261657715, "learning_rate": 2.8422282820412932e-05, "loss": 2.1845, "step": 2432 }, { "epoch": 0.5686074026467237, "grad_norm": 3.4821276664733887, "learning_rate": 2.8433969614335803e-05, "loss": 2.0609, "step": 2433 }, { "epoch": 0.5688411089363442, "grad_norm": 3.5887227058410645, "learning_rate": 2.844565640825867e-05, "loss": 2.1785, "step": 2434 }, { "epoch": 0.5690748152259648, "grad_norm": 2.5579843521118164, "learning_rate": 2.8457343202181534e-05, "loss": 2.1505, "step": 2435 }, { "epoch": 0.5693085215155853, "grad_norm": 4.12794303894043, "learning_rate": 2.8469029996104404e-05, "loss": 2.14, "step": 2436 }, { "epoch": 0.5695422278052058, "grad_norm": 2.921217441558838, "learning_rate": 2.848071679002727e-05, "loss": 2.2447, "step": 2437 }, { "epoch": 0.5697759340948263, "grad_norm": 2.5334649085998535, "learning_rate": 2.8492403583950135e-05, "loss": 2.1637, "step": 2438 }, { "epoch": 0.5700096403844468, "grad_norm": 3.5134117603302, "learning_rate": 2.8504090377873005e-05, "loss": 2.2353, "step": 2439 }, { "epoch": 0.5702433466740674, "grad_norm": 2.6205689907073975, "learning_rate": 2.8515777171795872e-05, "loss": 2.1574, "step": 2440 }, { "epoch": 0.5704770529636879, "grad_norm": 4.750428676605225, "learning_rate": 2.8527463965718736e-05, "loss": 2.3176, "step": 2441 }, { "epoch": 0.5707107592533084, "grad_norm": 2.7638378143310547, "learning_rate": 2.8539150759641606e-05, "loss": 2.1794, "step": 2442 }, { "epoch": 0.5709444655429289, "grad_norm": 3.365548610687256, "learning_rate": 2.8550837553564473e-05, "loss": 2.23, "step": 2443 }, { "epoch": 0.5711781718325495, "grad_norm": 3.0451323986053467, "learning_rate": 2.856252434748734e-05, "loss": 2.1371, "step": 2444 }, { "epoch": 0.57141187812217, "grad_norm": 3.017514705657959, "learning_rate": 2.8574211141410207e-05, "loss": 2.2064, "step": 2445 }, { "epoch": 0.5716455844117905, "grad_norm": 3.7177891731262207, "learning_rate": 2.8585897935333075e-05, "loss": 2.2321, "step": 2446 }, { "epoch": 0.571879290701411, "grad_norm": 2.773831844329834, "learning_rate": 2.859758472925594e-05, "loss": 2.169, "step": 2447 }, { "epoch": 0.5721129969910316, "grad_norm": 2.521056652069092, "learning_rate": 2.860927152317881e-05, "loss": 2.2433, "step": 2448 }, { "epoch": 0.5723467032806521, "grad_norm": 3.4069299697875977, "learning_rate": 2.8620958317101676e-05, "loss": 2.3515, "step": 2449 }, { "epoch": 0.5725804095702726, "grad_norm": 4.9752678871154785, "learning_rate": 2.8632645111024546e-05, "loss": 2.2627, "step": 2450 }, { "epoch": 0.5728141158598931, "grad_norm": 2.911116361618042, "learning_rate": 2.864433190494741e-05, "loss": 2.1835, "step": 2451 }, { "epoch": 0.5730478221495136, "grad_norm": 5.46993350982666, "learning_rate": 2.8656018698870277e-05, "loss": 2.1497, "step": 2452 }, { "epoch": 0.5732815284391342, "grad_norm": 3.9915506839752197, "learning_rate": 2.8667705492793147e-05, "loss": 2.2156, "step": 2453 }, { "epoch": 0.5735152347287547, "grad_norm": 3.6647703647613525, "learning_rate": 2.867939228671601e-05, "loss": 2.0863, "step": 2454 }, { "epoch": 0.5737489410183751, "grad_norm": 4.3379669189453125, "learning_rate": 2.8691079080638878e-05, "loss": 2.1459, "step": 2455 }, { "epoch": 0.5739826473079956, "grad_norm": 3.0604031085968018, "learning_rate": 2.870276587456175e-05, "loss": 2.1232, "step": 2456 }, { "epoch": 0.5742163535976162, "grad_norm": 3.7250287532806396, "learning_rate": 2.8714452668484612e-05, "loss": 2.127, "step": 2457 }, { "epoch": 0.5744500598872367, "grad_norm": 2.549146890640259, "learning_rate": 2.872613946240748e-05, "loss": 2.1154, "step": 2458 }, { "epoch": 0.5746837661768572, "grad_norm": 3.209848642349243, "learning_rate": 2.873782625633035e-05, "loss": 2.2868, "step": 2459 }, { "epoch": 0.5749174724664777, "grad_norm": 2.855752468109131, "learning_rate": 2.8749513050253213e-05, "loss": 2.1296, "step": 2460 }, { "epoch": 0.5751511787560982, "grad_norm": 3.96943998336792, "learning_rate": 2.876119984417608e-05, "loss": 2.2636, "step": 2461 }, { "epoch": 0.5753848850457188, "grad_norm": 3.301464080810547, "learning_rate": 2.877288663809895e-05, "loss": 2.2812, "step": 2462 }, { "epoch": 0.5756185913353393, "grad_norm": 2.8840487003326416, "learning_rate": 2.8784573432021814e-05, "loss": 2.257, "step": 2463 }, { "epoch": 0.5758522976249598, "grad_norm": 2.9724700450897217, "learning_rate": 2.879626022594468e-05, "loss": 2.2758, "step": 2464 }, { "epoch": 0.5760860039145803, "grad_norm": 3.142043352127075, "learning_rate": 2.8807947019867552e-05, "loss": 2.3049, "step": 2465 }, { "epoch": 0.5763197102042009, "grad_norm": 3.4536795616149902, "learning_rate": 2.881963381379042e-05, "loss": 2.2366, "step": 2466 }, { "epoch": 0.5765534164938214, "grad_norm": 2.6716623306274414, "learning_rate": 2.8831320607713283e-05, "loss": 2.1018, "step": 2467 }, { "epoch": 0.5767871227834419, "grad_norm": 3.1909024715423584, "learning_rate": 2.8843007401636153e-05, "loss": 2.3261, "step": 2468 }, { "epoch": 0.5770208290730624, "grad_norm": 3.398157835006714, "learning_rate": 2.885469419555902e-05, "loss": 2.1716, "step": 2469 }, { "epoch": 0.5772545353626829, "grad_norm": 3.773055076599121, "learning_rate": 2.8866380989481884e-05, "loss": 2.2607, "step": 2470 }, { "epoch": 0.5774882416523035, "grad_norm": 3.3960659503936768, "learning_rate": 2.8878067783404754e-05, "loss": 2.1821, "step": 2471 }, { "epoch": 0.577721947941924, "grad_norm": 3.6259591579437256, "learning_rate": 2.888975457732762e-05, "loss": 2.1484, "step": 2472 }, { "epoch": 0.5779556542315445, "grad_norm": 5.429389953613281, "learning_rate": 2.8901441371250488e-05, "loss": 2.1686, "step": 2473 }, { "epoch": 0.578189360521165, "grad_norm": 3.6038265228271484, "learning_rate": 2.8913128165173355e-05, "loss": 2.2837, "step": 2474 }, { "epoch": 0.5784230668107856, "grad_norm": 3.3097236156463623, "learning_rate": 2.8924814959096222e-05, "loss": 2.1307, "step": 2475 }, { "epoch": 0.5786567731004061, "grad_norm": 3.9699056148529053, "learning_rate": 2.893650175301909e-05, "loss": 2.0957, "step": 2476 }, { "epoch": 0.5788904793900266, "grad_norm": 3.8720507621765137, "learning_rate": 2.8948188546941956e-05, "loss": 2.2102, "step": 2477 }, { "epoch": 0.5791241856796471, "grad_norm": 2.8082895278930664, "learning_rate": 2.8959875340864823e-05, "loss": 2.1644, "step": 2478 }, { "epoch": 0.5793578919692676, "grad_norm": 4.04162073135376, "learning_rate": 2.897156213478769e-05, "loss": 2.0541, "step": 2479 }, { "epoch": 0.5795915982588882, "grad_norm": 3.3876869678497314, "learning_rate": 2.8983248928710558e-05, "loss": 2.1004, "step": 2480 }, { "epoch": 0.5798253045485087, "grad_norm": 3.510697603225708, "learning_rate": 2.8994935722633425e-05, "loss": 2.1276, "step": 2481 }, { "epoch": 0.5800590108381292, "grad_norm": 3.3848257064819336, "learning_rate": 2.9006622516556295e-05, "loss": 2.1566, "step": 2482 }, { "epoch": 0.5802927171277497, "grad_norm": 3.094491958618164, "learning_rate": 2.901830931047916e-05, "loss": 2.1327, "step": 2483 }, { "epoch": 0.5805264234173703, "grad_norm": 2.9426684379577637, "learning_rate": 2.9029996104402026e-05, "loss": 2.28, "step": 2484 }, { "epoch": 0.5807601297069908, "grad_norm": 2.9364757537841797, "learning_rate": 2.9041682898324896e-05, "loss": 2.1582, "step": 2485 }, { "epoch": 0.5809938359966113, "grad_norm": 2.4094226360321045, "learning_rate": 2.905336969224776e-05, "loss": 2.0981, "step": 2486 }, { "epoch": 0.5812275422862317, "grad_norm": 2.791027545928955, "learning_rate": 2.9065056486170627e-05, "loss": 2.0381, "step": 2487 }, { "epoch": 0.5814612485758524, "grad_norm": 2.6490793228149414, "learning_rate": 2.9076743280093497e-05, "loss": 2.1193, "step": 2488 }, { "epoch": 0.5816949548654728, "grad_norm": 3.120192050933838, "learning_rate": 2.908843007401636e-05, "loss": 2.1473, "step": 2489 }, { "epoch": 0.5819286611550933, "grad_norm": 2.8741061687469482, "learning_rate": 2.9100116867939228e-05, "loss": 2.1569, "step": 2490 }, { "epoch": 0.5821623674447138, "grad_norm": 3.1385505199432373, "learning_rate": 2.91118036618621e-05, "loss": 2.1477, "step": 2491 }, { "epoch": 0.5823960737343343, "grad_norm": 3.143723964691162, "learning_rate": 2.9123490455784962e-05, "loss": 2.2036, "step": 2492 }, { "epoch": 0.5826297800239549, "grad_norm": 2.6612253189086914, "learning_rate": 2.913517724970783e-05, "loss": 2.1452, "step": 2493 }, { "epoch": 0.5828634863135754, "grad_norm": 4.432351112365723, "learning_rate": 2.91468640436307e-05, "loss": 1.9991, "step": 2494 }, { "epoch": 0.5830971926031959, "grad_norm": 3.1005115509033203, "learning_rate": 2.9158550837553563e-05, "loss": 2.1268, "step": 2495 }, { "epoch": 0.5833308988928164, "grad_norm": 4.415120601654053, "learning_rate": 2.917023763147643e-05, "loss": 2.1111, "step": 2496 }, { "epoch": 0.583564605182437, "grad_norm": 3.849726438522339, "learning_rate": 2.91819244253993e-05, "loss": 2.157, "step": 2497 }, { "epoch": 0.5837983114720575, "grad_norm": 3.102363348007202, "learning_rate": 2.9193611219322168e-05, "loss": 2.2001, "step": 2498 }, { "epoch": 0.584032017761678, "grad_norm": 2.5337793827056885, "learning_rate": 2.9205298013245035e-05, "loss": 2.1279, "step": 2499 }, { "epoch": 0.5842657240512985, "grad_norm": 3.1100988388061523, "learning_rate": 2.9216984807167902e-05, "loss": 2.1321, "step": 2500 }, { "epoch": 0.584499430340919, "grad_norm": 2.9294424057006836, "learning_rate": 2.922867160109077e-05, "loss": 2.1797, "step": 2501 }, { "epoch": 0.5847331366305396, "grad_norm": 2.904578924179077, "learning_rate": 2.9240358395013636e-05, "loss": 2.1877, "step": 2502 }, { "epoch": 0.5849668429201601, "grad_norm": 3.13269305229187, "learning_rate": 2.9252045188936503e-05, "loss": 1.9934, "step": 2503 }, { "epoch": 0.5852005492097806, "grad_norm": 3.125417709350586, "learning_rate": 2.926373198285937e-05, "loss": 2.0801, "step": 2504 }, { "epoch": 0.5854342554994011, "grad_norm": 5.294067859649658, "learning_rate": 2.9275418776782237e-05, "loss": 2.2655, "step": 2505 }, { "epoch": 0.5856679617890217, "grad_norm": 3.7901675701141357, "learning_rate": 2.9287105570705104e-05, "loss": 2.1283, "step": 2506 }, { "epoch": 0.5859016680786422, "grad_norm": 5.6818528175354, "learning_rate": 2.929879236462797e-05, "loss": 2.0371, "step": 2507 }, { "epoch": 0.5861353743682627, "grad_norm": 3.791768789291382, "learning_rate": 2.9310479158550838e-05, "loss": 2.2359, "step": 2508 }, { "epoch": 0.5863690806578832, "grad_norm": 2.964843273162842, "learning_rate": 2.9322165952473705e-05, "loss": 2.3852, "step": 2509 }, { "epoch": 0.5866027869475037, "grad_norm": 2.397624969482422, "learning_rate": 2.9333852746396572e-05, "loss": 2.2263, "step": 2510 }, { "epoch": 0.5868364932371243, "grad_norm": 4.622359275817871, "learning_rate": 2.934553954031944e-05, "loss": 2.4177, "step": 2511 }, { "epoch": 0.5870701995267448, "grad_norm": 3.048053026199341, "learning_rate": 2.9357226334242306e-05, "loss": 2.1981, "step": 2512 }, { "epoch": 0.5873039058163653, "grad_norm": 3.6790499687194824, "learning_rate": 2.9368913128165173e-05, "loss": 2.1337, "step": 2513 }, { "epoch": 0.5875376121059858, "grad_norm": 3.522935152053833, "learning_rate": 2.9380599922088044e-05, "loss": 2.2567, "step": 2514 }, { "epoch": 0.5877713183956064, "grad_norm": 3.234828233718872, "learning_rate": 2.9392286716010908e-05, "loss": 2.1613, "step": 2515 }, { "epoch": 0.5880050246852269, "grad_norm": 2.51900315284729, "learning_rate": 2.9403973509933775e-05, "loss": 1.9679, "step": 2516 }, { "epoch": 0.5882387309748474, "grad_norm": 4.0977702140808105, "learning_rate": 2.9415660303856645e-05, "loss": 2.1652, "step": 2517 }, { "epoch": 0.5884724372644679, "grad_norm": 3.3580315113067627, "learning_rate": 2.942734709777951e-05, "loss": 2.5233, "step": 2518 }, { "epoch": 0.5887061435540883, "grad_norm": 2.927455186843872, "learning_rate": 2.9439033891702376e-05, "loss": 2.2781, "step": 2519 }, { "epoch": 0.588939849843709, "grad_norm": 2.424816370010376, "learning_rate": 2.9450720685625246e-05, "loss": 2.125, "step": 2520 }, { "epoch": 0.5891735561333294, "grad_norm": 3.4802849292755127, "learning_rate": 2.946240747954811e-05, "loss": 2.0665, "step": 2521 }, { "epoch": 0.5894072624229499, "grad_norm": 2.7616896629333496, "learning_rate": 2.9474094273470977e-05, "loss": 2.2143, "step": 2522 }, { "epoch": 0.5896409687125704, "grad_norm": 3.4480326175689697, "learning_rate": 2.9485781067393847e-05, "loss": 2.1032, "step": 2523 }, { "epoch": 0.589874675002191, "grad_norm": 2.888767957687378, "learning_rate": 2.949746786131671e-05, "loss": 2.0319, "step": 2524 }, { "epoch": 0.5901083812918115, "grad_norm": 2.880016326904297, "learning_rate": 2.9509154655239578e-05, "loss": 2.1006, "step": 2525 }, { "epoch": 0.590342087581432, "grad_norm": 3.239976644515991, "learning_rate": 2.952084144916245e-05, "loss": 2.1569, "step": 2526 }, { "epoch": 0.5905757938710525, "grad_norm": 2.719024181365967, "learning_rate": 2.9532528243085312e-05, "loss": 2.1975, "step": 2527 }, { "epoch": 0.5908095001606731, "grad_norm": 2.8089523315429688, "learning_rate": 2.9544215037008183e-05, "loss": 2.0164, "step": 2528 }, { "epoch": 0.5910432064502936, "grad_norm": 3.2550249099731445, "learning_rate": 2.955590183093105e-05, "loss": 2.029, "step": 2529 }, { "epoch": 0.5912769127399141, "grad_norm": 3.1312761306762695, "learning_rate": 2.9567588624853917e-05, "loss": 2.1693, "step": 2530 }, { "epoch": 0.5915106190295346, "grad_norm": 4.111201763153076, "learning_rate": 2.9579275418776784e-05, "loss": 2.0676, "step": 2531 }, { "epoch": 0.5917443253191551, "grad_norm": 2.7065796852111816, "learning_rate": 2.959096221269965e-05, "loss": 2.0691, "step": 2532 }, { "epoch": 0.5919780316087757, "grad_norm": 2.463132858276367, "learning_rate": 2.9602649006622518e-05, "loss": 2.1846, "step": 2533 }, { "epoch": 0.5922117378983962, "grad_norm": 2.4689924716949463, "learning_rate": 2.9614335800545385e-05, "loss": 2.0568, "step": 2534 }, { "epoch": 0.5924454441880167, "grad_norm": 2.6377158164978027, "learning_rate": 2.9626022594468252e-05, "loss": 2.1884, "step": 2535 }, { "epoch": 0.5926791504776372, "grad_norm": 3.0178098678588867, "learning_rate": 2.963770938839112e-05, "loss": 2.0615, "step": 2536 }, { "epoch": 0.5929128567672578, "grad_norm": 3.7813854217529297, "learning_rate": 2.9649396182313986e-05, "loss": 2.1359, "step": 2537 }, { "epoch": 0.5931465630568783, "grad_norm": 2.92142653465271, "learning_rate": 2.9661082976236853e-05, "loss": 2.2462, "step": 2538 }, { "epoch": 0.5933802693464988, "grad_norm": 3.2125000953674316, "learning_rate": 2.967276977015972e-05, "loss": 2.14, "step": 2539 }, { "epoch": 0.5936139756361193, "grad_norm": 4.530763626098633, "learning_rate": 2.9684456564082587e-05, "loss": 2.1641, "step": 2540 }, { "epoch": 0.5938476819257398, "grad_norm": 3.7777483463287354, "learning_rate": 2.9696143358005454e-05, "loss": 2.0734, "step": 2541 }, { "epoch": 0.5940813882153604, "grad_norm": 2.736964702606201, "learning_rate": 2.970783015192832e-05, "loss": 2.2387, "step": 2542 }, { "epoch": 0.5943150945049809, "grad_norm": 3.8529319763183594, "learning_rate": 2.9719516945851188e-05, "loss": 2.0763, "step": 2543 }, { "epoch": 0.5945488007946014, "grad_norm": 3.2259414196014404, "learning_rate": 2.9731203739774055e-05, "loss": 2.166, "step": 2544 }, { "epoch": 0.5947825070842219, "grad_norm": 2.8292245864868164, "learning_rate": 2.9742890533696922e-05, "loss": 2.1753, "step": 2545 }, { "epoch": 0.5950162133738425, "grad_norm": 2.8199098110198975, "learning_rate": 2.9754577327619793e-05, "loss": 2.1657, "step": 2546 }, { "epoch": 0.595249919663463, "grad_norm": 2.5240378379821777, "learning_rate": 2.9766264121542656e-05, "loss": 2.0562, "step": 2547 }, { "epoch": 0.5954836259530835, "grad_norm": 3.258937358856201, "learning_rate": 2.9777950915465523e-05, "loss": 2.2562, "step": 2548 }, { "epoch": 0.595717332242704, "grad_norm": 3.8967549800872803, "learning_rate": 2.9789637709388394e-05, "loss": 2.047, "step": 2549 }, { "epoch": 0.5959510385323245, "grad_norm": 6.161501884460449, "learning_rate": 2.9801324503311258e-05, "loss": 2.0373, "step": 2550 }, { "epoch": 0.5961847448219451, "grad_norm": 3.104236125946045, "learning_rate": 2.9813011297234125e-05, "loss": 2.0503, "step": 2551 }, { "epoch": 0.5964184511115656, "grad_norm": 3.1579174995422363, "learning_rate": 2.9824698091156995e-05, "loss": 2.0291, "step": 2552 }, { "epoch": 0.596652157401186, "grad_norm": 5.586901664733887, "learning_rate": 2.983638488507986e-05, "loss": 2.0595, "step": 2553 }, { "epoch": 0.5968858636908065, "grad_norm": 4.6942644119262695, "learning_rate": 2.984807167900273e-05, "loss": 2.2235, "step": 2554 }, { "epoch": 0.5971195699804271, "grad_norm": 3.359220027923584, "learning_rate": 2.9859758472925596e-05, "loss": 2.1947, "step": 2555 }, { "epoch": 0.5973532762700476, "grad_norm": 4.341405391693115, "learning_rate": 2.987144526684846e-05, "loss": 2.0753, "step": 2556 }, { "epoch": 0.5975869825596681, "grad_norm": 6.22213888168335, "learning_rate": 2.988313206077133e-05, "loss": 2.0541, "step": 2557 }, { "epoch": 0.5978206888492886, "grad_norm": 3.183227062225342, "learning_rate": 2.9894818854694197e-05, "loss": 2.2203, "step": 2558 }, { "epoch": 0.5980543951389091, "grad_norm": 4.432919979095459, "learning_rate": 2.990650564861706e-05, "loss": 2.0474, "step": 2559 }, { "epoch": 0.5982881014285297, "grad_norm": 2.4522604942321777, "learning_rate": 2.991819244253993e-05, "loss": 2.1711, "step": 2560 }, { "epoch": 0.5985218077181502, "grad_norm": 2.505354404449463, "learning_rate": 2.99298792364628e-05, "loss": 2.1418, "step": 2561 }, { "epoch": 0.5987555140077707, "grad_norm": 2.6819798946380615, "learning_rate": 2.9941566030385666e-05, "loss": 2.1497, "step": 2562 }, { "epoch": 0.5989892202973912, "grad_norm": 3.0800893306732178, "learning_rate": 2.9953252824308533e-05, "loss": 2.0697, "step": 2563 }, { "epoch": 0.5992229265870118, "grad_norm": 4.087061882019043, "learning_rate": 2.99649396182314e-05, "loss": 2.1767, "step": 2564 }, { "epoch": 0.5994566328766323, "grad_norm": 3.0218846797943115, "learning_rate": 2.9976626412154267e-05, "loss": 2.1893, "step": 2565 }, { "epoch": 0.5996903391662528, "grad_norm": 2.4910664558410645, "learning_rate": 2.9988313206077134e-05, "loss": 2.1559, "step": 2566 }, { "epoch": 0.5999240454558733, "grad_norm": 2.595533847808838, "learning_rate": 3e-05, "loss": 2.0712, "step": 2567 }, { "epoch": 0.6001577517454939, "grad_norm": 4.685715675354004, "learning_rate": 2.9999999989253226e-05, "loss": 2.0364, "step": 2568 }, { "epoch": 0.6003914580351144, "grad_norm": 3.008789300918579, "learning_rate": 2.9999999957012906e-05, "loss": 2.1225, "step": 2569 }, { "epoch": 0.6006251643247349, "grad_norm": 3.9363131523132324, "learning_rate": 2.9999999903279043e-05, "loss": 2.1458, "step": 2570 }, { "epoch": 0.6008588706143554, "grad_norm": 3.7594757080078125, "learning_rate": 2.999999982805163e-05, "loss": 2.1709, "step": 2571 }, { "epoch": 0.6010925769039759, "grad_norm": 3.269941568374634, "learning_rate": 2.9999999731330674e-05, "loss": 2.1287, "step": 2572 }, { "epoch": 0.6013262831935965, "grad_norm": 2.9613802433013916, "learning_rate": 2.999999961311617e-05, "loss": 2.2053, "step": 2573 }, { "epoch": 0.601559989483217, "grad_norm": 2.762586832046509, "learning_rate": 2.9999999473408118e-05, "loss": 2.4794, "step": 2574 }, { "epoch": 0.6017936957728375, "grad_norm": 2.885895252227783, "learning_rate": 2.9999999312206527e-05, "loss": 2.129, "step": 2575 }, { "epoch": 0.602027402062458, "grad_norm": 2.8239784240722656, "learning_rate": 2.9999999129511386e-05, "loss": 2.3126, "step": 2576 }, { "epoch": 0.6022611083520786, "grad_norm": 4.248277187347412, "learning_rate": 2.99999989253227e-05, "loss": 2.2123, "step": 2577 }, { "epoch": 0.6024948146416991, "grad_norm": 3.0241947174072266, "learning_rate": 2.9999998699640468e-05, "loss": 1.9929, "step": 2578 }, { "epoch": 0.6027285209313196, "grad_norm": 2.71213698387146, "learning_rate": 2.9999998452464694e-05, "loss": 2.1593, "step": 2579 }, { "epoch": 0.6029622272209401, "grad_norm": 3.1236660480499268, "learning_rate": 2.9999998183795374e-05, "loss": 2.107, "step": 2580 }, { "epoch": 0.6031959335105606, "grad_norm": 3.084170341491699, "learning_rate": 2.9999997893632515e-05, "loss": 2.1578, "step": 2581 }, { "epoch": 0.6034296398001812, "grad_norm": 3.2737526893615723, "learning_rate": 2.9999997581976107e-05, "loss": 2.1131, "step": 2582 }, { "epoch": 0.6036633460898017, "grad_norm": 2.550945520401001, "learning_rate": 2.999999724882616e-05, "loss": 2.2228, "step": 2583 }, { "epoch": 0.6038970523794222, "grad_norm": 5.200470924377441, "learning_rate": 2.9999996894182667e-05, "loss": 2.1193, "step": 2584 }, { "epoch": 0.6041307586690426, "grad_norm": 2.91137433052063, "learning_rate": 2.9999996518045635e-05, "loss": 2.1286, "step": 2585 }, { "epoch": 0.6043644649586633, "grad_norm": 2.6061344146728516, "learning_rate": 2.999999612041506e-05, "loss": 2.1019, "step": 2586 }, { "epoch": 0.6045981712482837, "grad_norm": 5.360860824584961, "learning_rate": 2.9999995701290944e-05, "loss": 1.9602, "step": 2587 }, { "epoch": 0.6048318775379042, "grad_norm": 5.315706729888916, "learning_rate": 2.999999526067329e-05, "loss": 2.208, "step": 2588 }, { "epoch": 0.6050655838275247, "grad_norm": 3.1824769973754883, "learning_rate": 2.9999994798562098e-05, "loss": 2.2043, "step": 2589 }, { "epoch": 0.6052992901171452, "grad_norm": 4.176323413848877, "learning_rate": 2.9999994314957364e-05, "loss": 2.1905, "step": 2590 }, { "epoch": 0.6055329964067658, "grad_norm": 3.3403685092926025, "learning_rate": 2.9999993809859095e-05, "loss": 2.0963, "step": 2591 }, { "epoch": 0.6057667026963863, "grad_norm": 2.4068517684936523, "learning_rate": 2.9999993283267284e-05, "loss": 2.1581, "step": 2592 }, { "epoch": 0.6060004089860068, "grad_norm": 3.5643503665924072, "learning_rate": 2.9999992735181937e-05, "loss": 2.2434, "step": 2593 }, { "epoch": 0.6062341152756273, "grad_norm": 3.053550958633423, "learning_rate": 2.9999992165603055e-05, "loss": 2.2872, "step": 2594 }, { "epoch": 0.6064678215652479, "grad_norm": 2.759333848953247, "learning_rate": 2.999999157453064e-05, "loss": 2.1105, "step": 2595 }, { "epoch": 0.6067015278548684, "grad_norm": 3.761321544647217, "learning_rate": 2.9999990961964686e-05, "loss": 2.2225, "step": 2596 }, { "epoch": 0.6069352341444889, "grad_norm": 3.386653423309326, "learning_rate": 2.9999990327905207e-05, "loss": 2.2115, "step": 2597 }, { "epoch": 0.6071689404341094, "grad_norm": 4.085807800292969, "learning_rate": 2.999998967235219e-05, "loss": 2.1587, "step": 2598 }, { "epoch": 0.6074026467237299, "grad_norm": 4.0244879722595215, "learning_rate": 2.999998899530564e-05, "loss": 2.2147, "step": 2599 }, { "epoch": 0.6076363530133505, "grad_norm": 3.0277178287506104, "learning_rate": 2.9999988296765563e-05, "loss": 2.1604, "step": 2600 }, { "epoch": 0.607870059302971, "grad_norm": 3.526428699493408, "learning_rate": 2.9999987576731952e-05, "loss": 2.2537, "step": 2601 }, { "epoch": 0.6081037655925915, "grad_norm": 4.145070552825928, "learning_rate": 2.9999986835204816e-05, "loss": 2.2353, "step": 2602 }, { "epoch": 0.608337471882212, "grad_norm": 2.4170050621032715, "learning_rate": 2.9999986072184154e-05, "loss": 2.0331, "step": 2603 }, { "epoch": 0.6085711781718326, "grad_norm": 2.6266112327575684, "learning_rate": 2.9999985287669963e-05, "loss": 2.0315, "step": 2604 }, { "epoch": 0.6088048844614531, "grad_norm": 3.03143572807312, "learning_rate": 2.9999984481662247e-05, "loss": 2.0534, "step": 2605 }, { "epoch": 0.6090385907510736, "grad_norm": 3.3712687492370605, "learning_rate": 2.999998365416101e-05, "loss": 2.1265, "step": 2606 }, { "epoch": 0.6092722970406941, "grad_norm": 5.443599224090576, "learning_rate": 2.9999982805166245e-05, "loss": 2.1021, "step": 2607 }, { "epoch": 0.6095060033303146, "grad_norm": 6.2539143562316895, "learning_rate": 2.9999981934677962e-05, "loss": 2.0587, "step": 2608 }, { "epoch": 0.6097397096199352, "grad_norm": 3.988706111907959, "learning_rate": 2.9999981042696158e-05, "loss": 2.1443, "step": 2609 }, { "epoch": 0.6099734159095557, "grad_norm": 5.274906635284424, "learning_rate": 2.9999980129220832e-05, "loss": 2.0631, "step": 2610 }, { "epoch": 0.6102071221991762, "grad_norm": 7.324644565582275, "learning_rate": 2.999997919425199e-05, "loss": 2.2357, "step": 2611 }, { "epoch": 0.6104408284887967, "grad_norm": 4.458303928375244, "learning_rate": 2.9999978237789633e-05, "loss": 2.0598, "step": 2612 }, { "epoch": 0.6106745347784173, "grad_norm": 3.0338971614837646, "learning_rate": 2.999997725983376e-05, "loss": 2.0535, "step": 2613 }, { "epoch": 0.6109082410680378, "grad_norm": 4.448062896728516, "learning_rate": 2.9999976260384374e-05, "loss": 2.1924, "step": 2614 }, { "epoch": 0.6111419473576583, "grad_norm": 3.0000782012939453, "learning_rate": 2.9999975239441475e-05, "loss": 2.1309, "step": 2615 }, { "epoch": 0.6113756536472787, "grad_norm": 4.2394537925720215, "learning_rate": 2.9999974197005067e-05, "loss": 2.1139, "step": 2616 }, { "epoch": 0.6116093599368994, "grad_norm": 2.4474847316741943, "learning_rate": 2.9999973133075144e-05, "loss": 2.2217, "step": 2617 }, { "epoch": 0.6118430662265198, "grad_norm": 2.4926631450653076, "learning_rate": 2.999997204765172e-05, "loss": 2.0336, "step": 2618 }, { "epoch": 0.6120767725161403, "grad_norm": 2.637083053588867, "learning_rate": 2.9999970940734787e-05, "loss": 2.108, "step": 2619 }, { "epoch": 0.6123104788057608, "grad_norm": 2.929912805557251, "learning_rate": 2.999996981232435e-05, "loss": 2.0976, "step": 2620 }, { "epoch": 0.6125441850953813, "grad_norm": 3.455371141433716, "learning_rate": 2.9999968662420405e-05, "loss": 2.0392, "step": 2621 }, { "epoch": 0.6127778913850019, "grad_norm": 4.508673667907715, "learning_rate": 2.9999967491022964e-05, "loss": 2.1288, "step": 2622 }, { "epoch": 0.6130115976746224, "grad_norm": 2.8019919395446777, "learning_rate": 2.999996629813202e-05, "loss": 2.1493, "step": 2623 }, { "epoch": 0.6132453039642429, "grad_norm": 4.768520355224609, "learning_rate": 2.9999965083747582e-05, "loss": 2.072, "step": 2624 }, { "epoch": 0.6134790102538634, "grad_norm": 2.5282702445983887, "learning_rate": 2.9999963847869646e-05, "loss": 2.0902, "step": 2625 }, { "epoch": 0.613712716543484, "grad_norm": 2.7508108615875244, "learning_rate": 2.9999962590498215e-05, "loss": 2.1356, "step": 2626 }, { "epoch": 0.6139464228331045, "grad_norm": 3.701738119125366, "learning_rate": 2.999996131163329e-05, "loss": 2.1896, "step": 2627 }, { "epoch": 0.614180129122725, "grad_norm": 2.95188570022583, "learning_rate": 2.9999960011274877e-05, "loss": 2.1965, "step": 2628 }, { "epoch": 0.6144138354123455, "grad_norm": 2.850818634033203, "learning_rate": 2.9999958689422975e-05, "loss": 2.0692, "step": 2629 }, { "epoch": 0.614647541701966, "grad_norm": 2.4347269535064697, "learning_rate": 2.9999957346077584e-05, "loss": 2.1933, "step": 2630 }, { "epoch": 0.6148812479915866, "grad_norm": 2.840049982070923, "learning_rate": 2.9999955981238705e-05, "loss": 2.0456, "step": 2631 }, { "epoch": 0.6151149542812071, "grad_norm": 2.672170877456665, "learning_rate": 2.9999954594906348e-05, "loss": 2.0693, "step": 2632 }, { "epoch": 0.6153486605708276, "grad_norm": 2.676084280014038, "learning_rate": 2.9999953187080507e-05, "loss": 2.0174, "step": 2633 }, { "epoch": 0.6155823668604481, "grad_norm": 3.214237928390503, "learning_rate": 2.999995175776119e-05, "loss": 2.0745, "step": 2634 }, { "epoch": 0.6158160731500687, "grad_norm": 2.5140833854675293, "learning_rate": 2.9999950306948393e-05, "loss": 2.0671, "step": 2635 }, { "epoch": 0.6160497794396892, "grad_norm": 3.019166946411133, "learning_rate": 2.999994883464212e-05, "loss": 2.1263, "step": 2636 }, { "epoch": 0.6162834857293097, "grad_norm": 2.42059063911438, "learning_rate": 2.9999947340842378e-05, "loss": 2.217, "step": 2637 }, { "epoch": 0.6165171920189302, "grad_norm": 2.4747068881988525, "learning_rate": 2.9999945825549163e-05, "loss": 2.033, "step": 2638 }, { "epoch": 0.6167508983085507, "grad_norm": 3.3977277278900146, "learning_rate": 2.9999944288762477e-05, "loss": 2.2182, "step": 2639 }, { "epoch": 0.6169846045981713, "grad_norm": 4.414750576019287, "learning_rate": 2.9999942730482327e-05, "loss": 2.1266, "step": 2640 }, { "epoch": 0.6172183108877918, "grad_norm": 2.905215263366699, "learning_rate": 2.999994115070871e-05, "loss": 2.219, "step": 2641 }, { "epoch": 0.6174520171774123, "grad_norm": 2.7055718898773193, "learning_rate": 2.9999939549441634e-05, "loss": 2.3335, "step": 2642 }, { "epoch": 0.6176857234670328, "grad_norm": 2.622222900390625, "learning_rate": 2.9999937926681097e-05, "loss": 2.2312, "step": 2643 }, { "epoch": 0.6179194297566534, "grad_norm": 4.182762145996094, "learning_rate": 2.9999936282427103e-05, "loss": 2.1755, "step": 2644 }, { "epoch": 0.6181531360462739, "grad_norm": 2.6352741718292236, "learning_rate": 2.9999934616679657e-05, "loss": 2.0284, "step": 2645 }, { "epoch": 0.6183868423358944, "grad_norm": 3.5920791625976562, "learning_rate": 2.999993292943875e-05, "loss": 2.1406, "step": 2646 }, { "epoch": 0.6186205486255149, "grad_norm": 3.255211591720581, "learning_rate": 2.99999312207044e-05, "loss": 2.2181, "step": 2647 }, { "epoch": 0.6188542549151353, "grad_norm": 3.2128357887268066, "learning_rate": 2.99999294904766e-05, "loss": 2.1151, "step": 2648 }, { "epoch": 0.619087961204756, "grad_norm": 3.165515422821045, "learning_rate": 2.9999927738755358e-05, "loss": 2.0405, "step": 2649 }, { "epoch": 0.6193216674943764, "grad_norm": 2.4857869148254395, "learning_rate": 2.9999925965540672e-05, "loss": 2.1412, "step": 2650 }, { "epoch": 0.6195553737839969, "grad_norm": 4.291042804718018, "learning_rate": 2.9999924170832542e-05, "loss": 2.2098, "step": 2651 }, { "epoch": 0.6197890800736174, "grad_norm": 2.588829755783081, "learning_rate": 2.9999922354630978e-05, "loss": 2.0719, "step": 2652 }, { "epoch": 0.620022786363238, "grad_norm": 3.379207134246826, "learning_rate": 2.9999920516935976e-05, "loss": 2.2043, "step": 2653 }, { "epoch": 0.6202564926528585, "grad_norm": 3.7099010944366455, "learning_rate": 2.9999918657747545e-05, "loss": 2.1026, "step": 2654 }, { "epoch": 0.620490198942479, "grad_norm": 2.919857978820801, "learning_rate": 2.9999916777065682e-05, "loss": 2.0708, "step": 2655 }, { "epoch": 0.6207239052320995, "grad_norm": 2.93373703956604, "learning_rate": 2.9999914874890393e-05, "loss": 2.0164, "step": 2656 }, { "epoch": 0.6209576115217201, "grad_norm": 2.584329843521118, "learning_rate": 2.999991295122168e-05, "loss": 2.227, "step": 2657 }, { "epoch": 0.6211913178113406, "grad_norm": 2.7373573780059814, "learning_rate": 2.9999911006059542e-05, "loss": 2.1248, "step": 2658 }, { "epoch": 0.6214250241009611, "grad_norm": 2.932347059249878, "learning_rate": 2.9999909039403988e-05, "loss": 2.014, "step": 2659 }, { "epoch": 0.6216587303905816, "grad_norm": 3.0579302310943604, "learning_rate": 2.999990705125502e-05, "loss": 2.1118, "step": 2660 }, { "epoch": 0.6218924366802021, "grad_norm": 2.686251640319824, "learning_rate": 2.9999905041612636e-05, "loss": 2.0675, "step": 2661 }, { "epoch": 0.6221261429698227, "grad_norm": 4.208503723144531, "learning_rate": 2.999990301047684e-05, "loss": 2.1902, "step": 2662 }, { "epoch": 0.6223598492594432, "grad_norm": 2.669306755065918, "learning_rate": 2.9999900957847642e-05, "loss": 2.1972, "step": 2663 }, { "epoch": 0.6225935555490637, "grad_norm": 3.342355966567993, "learning_rate": 2.9999898883725034e-05, "loss": 2.1328, "step": 2664 }, { "epoch": 0.6228272618386842, "grad_norm": 4.7574663162231445, "learning_rate": 2.9999896788109025e-05, "loss": 2.0784, "step": 2665 }, { "epoch": 0.6230609681283048, "grad_norm": 2.9626080989837646, "learning_rate": 2.9999894670999627e-05, "loss": 1.9989, "step": 2666 }, { "epoch": 0.6232946744179253, "grad_norm": 3.7141306400299072, "learning_rate": 2.9999892532396825e-05, "loss": 2.0806, "step": 2667 }, { "epoch": 0.6235283807075458, "grad_norm": 3.350193977355957, "learning_rate": 2.999989037230063e-05, "loss": 2.1644, "step": 2668 }, { "epoch": 0.6237620869971663, "grad_norm": 2.6916379928588867, "learning_rate": 2.999988819071105e-05, "loss": 2.1931, "step": 2669 }, { "epoch": 0.6239957932867868, "grad_norm": 2.8656485080718994, "learning_rate": 2.9999885987628083e-05, "loss": 1.9949, "step": 2670 }, { "epoch": 0.6242294995764074, "grad_norm": 3.4843719005584717, "learning_rate": 2.9999883763051732e-05, "loss": 2.0978, "step": 2671 }, { "epoch": 0.6244632058660279, "grad_norm": 2.869832992553711, "learning_rate": 2.9999881516982e-05, "loss": 2.038, "step": 2672 }, { "epoch": 0.6246969121556484, "grad_norm": 4.77819299697876, "learning_rate": 2.9999879249418893e-05, "loss": 2.1607, "step": 2673 }, { "epoch": 0.6249306184452689, "grad_norm": 4.403614044189453, "learning_rate": 2.9999876960362413e-05, "loss": 2.0415, "step": 2674 }, { "epoch": 0.6251643247348895, "grad_norm": 2.916949987411499, "learning_rate": 2.9999874649812563e-05, "loss": 2.2575, "step": 2675 }, { "epoch": 0.62539803102451, "grad_norm": 4.130530834197998, "learning_rate": 2.9999872317769347e-05, "loss": 2.2186, "step": 2676 }, { "epoch": 0.6256317373141305, "grad_norm": 3.182192802429199, "learning_rate": 2.9999869964232765e-05, "loss": 2.0524, "step": 2677 }, { "epoch": 0.625865443603751, "grad_norm": 2.8393378257751465, "learning_rate": 2.9999867589202823e-05, "loss": 2.0501, "step": 2678 }, { "epoch": 0.6260991498933715, "grad_norm": 2.771392822265625, "learning_rate": 2.9999865192679526e-05, "loss": 2.0559, "step": 2679 }, { "epoch": 0.6263328561829921, "grad_norm": 3.1582775115966797, "learning_rate": 2.9999862774662876e-05, "loss": 2.0646, "step": 2680 }, { "epoch": 0.6265665624726126, "grad_norm": 2.7898800373077393, "learning_rate": 2.9999860335152876e-05, "loss": 2.1373, "step": 2681 }, { "epoch": 0.626800268762233, "grad_norm": 2.744086265563965, "learning_rate": 2.999985787414953e-05, "loss": 2.0577, "step": 2682 }, { "epoch": 0.6270339750518535, "grad_norm": 2.971482038497925, "learning_rate": 2.9999855391652843e-05, "loss": 2.0666, "step": 2683 }, { "epoch": 0.6272676813414741, "grad_norm": 3.4273033142089844, "learning_rate": 2.9999852887662816e-05, "loss": 2.1828, "step": 2684 }, { "epoch": 0.6275013876310946, "grad_norm": 2.765570640563965, "learning_rate": 2.9999850362179454e-05, "loss": 2.1075, "step": 2685 }, { "epoch": 0.6277350939207151, "grad_norm": 4.369833946228027, "learning_rate": 2.9999847815202756e-05, "loss": 2.0716, "step": 2686 }, { "epoch": 0.6279688002103356, "grad_norm": 4.738552570343018, "learning_rate": 2.9999845246732732e-05, "loss": 2.1409, "step": 2687 }, { "epoch": 0.6282025064999561, "grad_norm": 3.6895999908447266, "learning_rate": 2.9999842656769386e-05, "loss": 2.2659, "step": 2688 }, { "epoch": 0.6284362127895767, "grad_norm": 6.05429220199585, "learning_rate": 2.999984004531272e-05, "loss": 2.0924, "step": 2689 }, { "epoch": 0.6286699190791972, "grad_norm": 7.985546112060547, "learning_rate": 2.9999837412362732e-05, "loss": 2.1579, "step": 2690 }, { "epoch": 0.6289036253688177, "grad_norm": 3.9234232902526855, "learning_rate": 2.9999834757919433e-05, "loss": 2.1266, "step": 2691 }, { "epoch": 0.6291373316584382, "grad_norm": 3.139462471008301, "learning_rate": 2.9999832081982823e-05, "loss": 2.3579, "step": 2692 }, { "epoch": 0.6293710379480588, "grad_norm": 6.156429290771484, "learning_rate": 2.999982938455291e-05, "loss": 2.1535, "step": 2693 }, { "epoch": 0.6296047442376793, "grad_norm": 4.352836608886719, "learning_rate": 2.999982666562969e-05, "loss": 2.1099, "step": 2694 }, { "epoch": 0.6298384505272998, "grad_norm": 3.107897996902466, "learning_rate": 2.9999823925213177e-05, "loss": 1.9814, "step": 2695 }, { "epoch": 0.6300721568169203, "grad_norm": 4.870451927185059, "learning_rate": 2.999982116330337e-05, "loss": 2.1073, "step": 2696 }, { "epoch": 0.6303058631065409, "grad_norm": 3.253415107727051, "learning_rate": 2.9999818379900273e-05, "loss": 2.05, "step": 2697 }, { "epoch": 0.6305395693961614, "grad_norm": 3.402700662612915, "learning_rate": 2.9999815575003886e-05, "loss": 2.1368, "step": 2698 }, { "epoch": 0.6307732756857819, "grad_norm": 4.175210475921631, "learning_rate": 2.9999812748614224e-05, "loss": 2.2525, "step": 2699 }, { "epoch": 0.6310069819754024, "grad_norm": 3.09706974029541, "learning_rate": 2.9999809900731277e-05, "loss": 2.1306, "step": 2700 }, { "epoch": 0.6312406882650229, "grad_norm": 3.4874353408813477, "learning_rate": 2.999980703135506e-05, "loss": 2.0253, "step": 2701 }, { "epoch": 0.6314743945546435, "grad_norm": 4.170726299285889, "learning_rate": 2.9999804140485572e-05, "loss": 2.0545, "step": 2702 }, { "epoch": 0.631708100844264, "grad_norm": 2.8059186935424805, "learning_rate": 2.9999801228122815e-05, "loss": 1.9911, "step": 2703 }, { "epoch": 0.6319418071338845, "grad_norm": 4.360985279083252, "learning_rate": 2.9999798294266806e-05, "loss": 2.1, "step": 2704 }, { "epoch": 0.632175513423505, "grad_norm": 2.7380282878875732, "learning_rate": 2.9999795338917533e-05, "loss": 2.06, "step": 2705 }, { "epoch": 0.6324092197131256, "grad_norm": 2.48465895652771, "learning_rate": 2.9999792362075006e-05, "loss": 2.1445, "step": 2706 }, { "epoch": 0.6326429260027461, "grad_norm": 2.4069533348083496, "learning_rate": 2.9999789363739234e-05, "loss": 2.244, "step": 2707 }, { "epoch": 0.6328766322923666, "grad_norm": 2.6356983184814453, "learning_rate": 2.9999786343910215e-05, "loss": 2.0691, "step": 2708 }, { "epoch": 0.6331103385819871, "grad_norm": 2.729182243347168, "learning_rate": 2.9999783302587952e-05, "loss": 2.0982, "step": 2709 }, { "epoch": 0.6333440448716076, "grad_norm": 2.287994146347046, "learning_rate": 2.9999780239772458e-05, "loss": 1.9695, "step": 2710 }, { "epoch": 0.6335777511612282, "grad_norm": 2.838627576828003, "learning_rate": 2.9999777155463733e-05, "loss": 2.22, "step": 2711 }, { "epoch": 0.6338114574508487, "grad_norm": 2.6748294830322266, "learning_rate": 2.999977404966178e-05, "loss": 2.1862, "step": 2712 }, { "epoch": 0.6340451637404692, "grad_norm": 2.8758163452148438, "learning_rate": 2.9999770922366606e-05, "loss": 2.0995, "step": 2713 }, { "epoch": 0.6342788700300896, "grad_norm": 2.815192699432373, "learning_rate": 2.9999767773578213e-05, "loss": 2.0989, "step": 2714 }, { "epoch": 0.6345125763197103, "grad_norm": 2.7390928268432617, "learning_rate": 2.9999764603296603e-05, "loss": 2.0615, "step": 2715 }, { "epoch": 0.6347462826093307, "grad_norm": 3.6655654907226562, "learning_rate": 2.9999761411521786e-05, "loss": 2.1722, "step": 2716 }, { "epoch": 0.6349799888989512, "grad_norm": 2.6300182342529297, "learning_rate": 2.999975819825377e-05, "loss": 2.0769, "step": 2717 }, { "epoch": 0.6352136951885717, "grad_norm": 2.8899412155151367, "learning_rate": 2.999975496349255e-05, "loss": 2.1992, "step": 2718 }, { "epoch": 0.6354474014781922, "grad_norm": 2.4816513061523438, "learning_rate": 2.9999751707238135e-05, "loss": 2.0922, "step": 2719 }, { "epoch": 0.6356811077678128, "grad_norm": 2.4735114574432373, "learning_rate": 2.9999748429490527e-05, "loss": 2.1522, "step": 2720 }, { "epoch": 0.6359148140574333, "grad_norm": 2.7698895931243896, "learning_rate": 2.9999745130249737e-05, "loss": 2.1579, "step": 2721 }, { "epoch": 0.6361485203470538, "grad_norm": 2.7373266220092773, "learning_rate": 2.9999741809515763e-05, "loss": 1.9738, "step": 2722 }, { "epoch": 0.6363822266366743, "grad_norm": 3.2279748916625977, "learning_rate": 2.9999738467288613e-05, "loss": 2.0636, "step": 2723 }, { "epoch": 0.6366159329262949, "grad_norm": 2.622744560241699, "learning_rate": 2.999973510356829e-05, "loss": 2.1387, "step": 2724 }, { "epoch": 0.6368496392159154, "grad_norm": 2.5458085536956787, "learning_rate": 2.9999731718354804e-05, "loss": 2.0794, "step": 2725 }, { "epoch": 0.6370833455055359, "grad_norm": 2.45892333984375, "learning_rate": 2.9999728311648155e-05, "loss": 2.186, "step": 2726 }, { "epoch": 0.6373170517951564, "grad_norm": 2.5312259197235107, "learning_rate": 2.9999724883448346e-05, "loss": 2.1387, "step": 2727 }, { "epoch": 0.6375507580847769, "grad_norm": 2.573387622833252, "learning_rate": 2.999972143375539e-05, "loss": 2.1566, "step": 2728 }, { "epoch": 0.6377844643743975, "grad_norm": 2.4328453540802, "learning_rate": 2.9999717962569285e-05, "loss": 2.1528, "step": 2729 }, { "epoch": 0.638018170664018, "grad_norm": 3.9412620067596436, "learning_rate": 2.999971446989004e-05, "loss": 2.014, "step": 2730 }, { "epoch": 0.6382518769536385, "grad_norm": 4.400397777557373, "learning_rate": 2.9999710955717653e-05, "loss": 2.0629, "step": 2731 }, { "epoch": 0.638485583243259, "grad_norm": 3.376992702484131, "learning_rate": 2.9999707420052135e-05, "loss": 2.0574, "step": 2732 }, { "epoch": 0.6387192895328796, "grad_norm": 5.23610258102417, "learning_rate": 2.9999703862893492e-05, "loss": 2.2517, "step": 2733 }, { "epoch": 0.6389529958225001, "grad_norm": 3.1684165000915527, "learning_rate": 2.9999700284241726e-05, "loss": 2.0036, "step": 2734 }, { "epoch": 0.6391867021121206, "grad_norm": 3.0307207107543945, "learning_rate": 2.999969668409684e-05, "loss": 2.1979, "step": 2735 }, { "epoch": 0.6394204084017411, "grad_norm": 2.6151881217956543, "learning_rate": 2.999969306245885e-05, "loss": 2.0497, "step": 2736 }, { "epoch": 0.6396541146913617, "grad_norm": 3.659677505493164, "learning_rate": 2.9999689419327747e-05, "loss": 2.1978, "step": 2737 }, { "epoch": 0.6398878209809822, "grad_norm": 3.6011176109313965, "learning_rate": 2.9999685754703543e-05, "loss": 2.0417, "step": 2738 }, { "epoch": 0.6401215272706027, "grad_norm": 3.6481544971466064, "learning_rate": 2.9999682068586244e-05, "loss": 2.1552, "step": 2739 }, { "epoch": 0.6403552335602232, "grad_norm": 2.8419687747955322, "learning_rate": 2.9999678360975857e-05, "loss": 1.9847, "step": 2740 }, { "epoch": 0.6405889398498437, "grad_norm": 2.871457099914551, "learning_rate": 2.999967463187238e-05, "loss": 2.0837, "step": 2741 }, { "epoch": 0.6408226461394643, "grad_norm": 2.494598865509033, "learning_rate": 2.9999670881275828e-05, "loss": 2.1816, "step": 2742 }, { "epoch": 0.6410563524290848, "grad_norm": 3.1708016395568848, "learning_rate": 2.99996671091862e-05, "loss": 2.134, "step": 2743 }, { "epoch": 0.6412900587187053, "grad_norm": 2.7395026683807373, "learning_rate": 2.99996633156035e-05, "loss": 2.2234, "step": 2744 }, { "epoch": 0.6415237650083258, "grad_norm": 2.602565050125122, "learning_rate": 2.9999659500527735e-05, "loss": 2.2016, "step": 2745 }, { "epoch": 0.6417574712979464, "grad_norm": 3.309537649154663, "learning_rate": 2.9999655663958916e-05, "loss": 2.1464, "step": 2746 }, { "epoch": 0.6419911775875669, "grad_norm": 2.8590340614318848, "learning_rate": 2.9999651805897042e-05, "loss": 2.1983, "step": 2747 }, { "epoch": 0.6422248838771873, "grad_norm": 3.719193458557129, "learning_rate": 2.999964792634212e-05, "loss": 2.0216, "step": 2748 }, { "epoch": 0.6424585901668078, "grad_norm": 4.418211460113525, "learning_rate": 2.999964402529416e-05, "loss": 2.1913, "step": 2749 }, { "epoch": 0.6426922964564283, "grad_norm": 4.066451072692871, "learning_rate": 2.999964010275316e-05, "loss": 2.075, "step": 2750 }, { "epoch": 0.6429260027460489, "grad_norm": 4.477896690368652, "learning_rate": 2.999963615871913e-05, "loss": 2.0561, "step": 2751 }, { "epoch": 0.6431597090356694, "grad_norm": 4.082113742828369, "learning_rate": 2.999963219319208e-05, "loss": 2.0114, "step": 2752 }, { "epoch": 0.6433934153252899, "grad_norm": 3.8595221042633057, "learning_rate": 2.9999628206172004e-05, "loss": 2.0973, "step": 2753 }, { "epoch": 0.6436271216149104, "grad_norm": 4.282974720001221, "learning_rate": 2.9999624197658915e-05, "loss": 2.1102, "step": 2754 }, { "epoch": 0.643860827904531, "grad_norm": 2.9734723567962646, "learning_rate": 2.999962016765282e-05, "loss": 2.127, "step": 2755 }, { "epoch": 0.6440945341941515, "grad_norm": 3.488223075866699, "learning_rate": 2.9999616116153723e-05, "loss": 2.0107, "step": 2756 }, { "epoch": 0.644328240483772, "grad_norm": 2.7114622592926025, "learning_rate": 2.9999612043161627e-05, "loss": 2.1381, "step": 2757 }, { "epoch": 0.6445619467733925, "grad_norm": 2.720616579055786, "learning_rate": 2.999960794867654e-05, "loss": 2.0791, "step": 2758 }, { "epoch": 0.644795653063013, "grad_norm": 3.3265161514282227, "learning_rate": 2.9999603832698478e-05, "loss": 2.0025, "step": 2759 }, { "epoch": 0.6450293593526336, "grad_norm": 2.9880425930023193, "learning_rate": 2.999959969522743e-05, "loss": 2.2008, "step": 2760 }, { "epoch": 0.6452630656422541, "grad_norm": 3.2620391845703125, "learning_rate": 2.999959553626341e-05, "loss": 2.0024, "step": 2761 }, { "epoch": 0.6454967719318746, "grad_norm": 3.3956120014190674, "learning_rate": 2.9999591355806424e-05, "loss": 1.9424, "step": 2762 }, { "epoch": 0.6457304782214951, "grad_norm": 2.409242630004883, "learning_rate": 2.9999587153856475e-05, "loss": 2.0225, "step": 2763 }, { "epoch": 0.6459641845111157, "grad_norm": 5.314191818237305, "learning_rate": 2.9999582930413574e-05, "loss": 2.1169, "step": 2764 }, { "epoch": 0.6461978908007362, "grad_norm": 3.6014811992645264, "learning_rate": 2.999957868547772e-05, "loss": 1.9342, "step": 2765 }, { "epoch": 0.6464315970903567, "grad_norm": 3.09942889213562, "learning_rate": 2.9999574419048926e-05, "loss": 2.0303, "step": 2766 }, { "epoch": 0.6466653033799772, "grad_norm": 7.320970058441162, "learning_rate": 2.9999570131127196e-05, "loss": 2.0504, "step": 2767 }, { "epoch": 0.6468990096695977, "grad_norm": 6.503530979156494, "learning_rate": 2.9999565821712538e-05, "loss": 2.1967, "step": 2768 }, { "epoch": 0.6471327159592183, "grad_norm": 3.755702495574951, "learning_rate": 2.999956149080495e-05, "loss": 2.1326, "step": 2769 }, { "epoch": 0.6473664222488388, "grad_norm": 3.18879771232605, "learning_rate": 2.999955713840445e-05, "loss": 2.0751, "step": 2770 }, { "epoch": 0.6476001285384593, "grad_norm": 4.953183650970459, "learning_rate": 2.999955276451104e-05, "loss": 2.1449, "step": 2771 }, { "epoch": 0.6478338348280798, "grad_norm": 5.07566499710083, "learning_rate": 2.9999548369124715e-05, "loss": 1.966, "step": 2772 }, { "epoch": 0.6480675411177004, "grad_norm": 4.180307388305664, "learning_rate": 2.9999543952245502e-05, "loss": 2.0841, "step": 2773 }, { "epoch": 0.6483012474073209, "grad_norm": 3.4557547569274902, "learning_rate": 2.999953951387339e-05, "loss": 2.0471, "step": 2774 }, { "epoch": 0.6485349536969414, "grad_norm": 2.461749315261841, "learning_rate": 2.999953505400839e-05, "loss": 2.0607, "step": 2775 }, { "epoch": 0.6487686599865619, "grad_norm": 3.9475345611572266, "learning_rate": 2.9999530572650514e-05, "loss": 2.0103, "step": 2776 }, { "epoch": 0.6490023662761825, "grad_norm": 3.097691535949707, "learning_rate": 2.9999526069799764e-05, "loss": 2.1484, "step": 2777 }, { "epoch": 0.649236072565803, "grad_norm": 4.2841901779174805, "learning_rate": 2.9999521545456144e-05, "loss": 2.2063, "step": 2778 }, { "epoch": 0.6494697788554235, "grad_norm": 3.403226375579834, "learning_rate": 2.999951699961966e-05, "loss": 2.3115, "step": 2779 }, { "epoch": 0.649703485145044, "grad_norm": 4.000845909118652, "learning_rate": 2.9999512432290328e-05, "loss": 2.1108, "step": 2780 }, { "epoch": 0.6499371914346644, "grad_norm": 3.1933062076568604, "learning_rate": 2.9999507843468147e-05, "loss": 2.1102, "step": 2781 }, { "epoch": 0.650170897724285, "grad_norm": 2.9120750427246094, "learning_rate": 2.9999503233153125e-05, "loss": 2.1156, "step": 2782 }, { "epoch": 0.6504046040139055, "grad_norm": 2.811765432357788, "learning_rate": 2.999949860134527e-05, "loss": 1.9136, "step": 2783 }, { "epoch": 0.650638310303526, "grad_norm": 3.4414560794830322, "learning_rate": 2.9999493948044585e-05, "loss": 2.1371, "step": 2784 }, { "epoch": 0.6508720165931465, "grad_norm": 2.81860089302063, "learning_rate": 2.9999489273251077e-05, "loss": 1.9998, "step": 2785 }, { "epoch": 0.6511057228827671, "grad_norm": 3.236013889312744, "learning_rate": 2.9999484576964755e-05, "loss": 1.9795, "step": 2786 }, { "epoch": 0.6513394291723876, "grad_norm": 3.953721284866333, "learning_rate": 2.999947985918563e-05, "loss": 2.1823, "step": 2787 }, { "epoch": 0.6515731354620081, "grad_norm": 3.101938486099243, "learning_rate": 2.99994751199137e-05, "loss": 2.2435, "step": 2788 }, { "epoch": 0.6518068417516286, "grad_norm": 3.7461047172546387, "learning_rate": 2.9999470359148977e-05, "loss": 2.2238, "step": 2789 }, { "epoch": 0.6520405480412491, "grad_norm": 3.2453508377075195, "learning_rate": 2.9999465576891468e-05, "loss": 2.0315, "step": 2790 }, { "epoch": 0.6522742543308697, "grad_norm": 4.7244486808776855, "learning_rate": 2.9999460773141178e-05, "loss": 2.1458, "step": 2791 }, { "epoch": 0.6525079606204902, "grad_norm": 3.9589269161224365, "learning_rate": 2.999945594789811e-05, "loss": 2.2088, "step": 2792 }, { "epoch": 0.6527416669101107, "grad_norm": 2.6379621028900146, "learning_rate": 2.999945110116228e-05, "loss": 1.9873, "step": 2793 }, { "epoch": 0.6529753731997312, "grad_norm": 3.0476229190826416, "learning_rate": 2.9999446232933686e-05, "loss": 2.0402, "step": 2794 }, { "epoch": 0.6532090794893518, "grad_norm": 2.6884682178497314, "learning_rate": 2.9999441343212344e-05, "loss": 2.0485, "step": 2795 }, { "epoch": 0.6534427857789723, "grad_norm": 3.275876045227051, "learning_rate": 2.9999436431998253e-05, "loss": 2.1093, "step": 2796 }, { "epoch": 0.6536764920685928, "grad_norm": 2.5873169898986816, "learning_rate": 2.9999431499291425e-05, "loss": 2.039, "step": 2797 }, { "epoch": 0.6539101983582133, "grad_norm": 2.439955711364746, "learning_rate": 2.9999426545091862e-05, "loss": 2.2697, "step": 2798 }, { "epoch": 0.6541439046478338, "grad_norm": 2.556126117706299, "learning_rate": 2.9999421569399576e-05, "loss": 2.1372, "step": 2799 }, { "epoch": 0.6543776109374544, "grad_norm": 2.727961778640747, "learning_rate": 2.9999416572214578e-05, "loss": 2.1006, "step": 2800 }, { "epoch": 0.6546113172270749, "grad_norm": 2.572901487350464, "learning_rate": 2.9999411553536864e-05, "loss": 2.0121, "step": 2801 }, { "epoch": 0.6548450235166954, "grad_norm": 3.2354576587677, "learning_rate": 2.9999406513366445e-05, "loss": 2.2725, "step": 2802 }, { "epoch": 0.6550787298063159, "grad_norm": 3.241799831390381, "learning_rate": 2.9999401451703334e-05, "loss": 1.9783, "step": 2803 }, { "epoch": 0.6553124360959365, "grad_norm": 2.978919506072998, "learning_rate": 2.9999396368547534e-05, "loss": 2.2176, "step": 2804 }, { "epoch": 0.655546142385557, "grad_norm": 5.342133045196533, "learning_rate": 2.9999391263899056e-05, "loss": 2.1983, "step": 2805 }, { "epoch": 0.6557798486751775, "grad_norm": 3.583592414855957, "learning_rate": 2.9999386137757896e-05, "loss": 2.1662, "step": 2806 }, { "epoch": 0.656013554964798, "grad_norm": 4.396713733673096, "learning_rate": 2.9999380990124078e-05, "loss": 2.0154, "step": 2807 }, { "epoch": 0.6562472612544185, "grad_norm": 3.858100175857544, "learning_rate": 2.999937582099759e-05, "loss": 2.049, "step": 2808 }, { "epoch": 0.6564809675440391, "grad_norm": 2.712195634841919, "learning_rate": 2.9999370630378458e-05, "loss": 2.2236, "step": 2809 }, { "epoch": 0.6567146738336596, "grad_norm": 3.2071115970611572, "learning_rate": 2.999936541826668e-05, "loss": 2.0452, "step": 2810 }, { "epoch": 0.65694838012328, "grad_norm": 2.4464738368988037, "learning_rate": 2.9999360184662263e-05, "loss": 2.0319, "step": 2811 }, { "epoch": 0.6571820864129005, "grad_norm": 2.390638589859009, "learning_rate": 2.999935492956522e-05, "loss": 2.1921, "step": 2812 }, { "epoch": 0.6574157927025212, "grad_norm": 4.667612552642822, "learning_rate": 2.9999349652975554e-05, "loss": 1.9113, "step": 2813 }, { "epoch": 0.6576494989921416, "grad_norm": 3.586907386779785, "learning_rate": 2.9999344354893276e-05, "loss": 2.0449, "step": 2814 }, { "epoch": 0.6578832052817621, "grad_norm": 3.010568380355835, "learning_rate": 2.9999339035318383e-05, "loss": 2.0908, "step": 2815 }, { "epoch": 0.6581169115713826, "grad_norm": 5.835478782653809, "learning_rate": 2.99993336942509e-05, "loss": 2.0816, "step": 2816 }, { "epoch": 0.6583506178610031, "grad_norm": 4.072377681732178, "learning_rate": 2.999932833169082e-05, "loss": 2.2017, "step": 2817 }, { "epoch": 0.6585843241506237, "grad_norm": 3.06284761428833, "learning_rate": 2.9999322947638157e-05, "loss": 2.0907, "step": 2818 }, { "epoch": 0.6588180304402442, "grad_norm": 3.658717393875122, "learning_rate": 2.9999317542092917e-05, "loss": 2.1928, "step": 2819 }, { "epoch": 0.6590517367298647, "grad_norm": 2.541919469833374, "learning_rate": 2.999931211505511e-05, "loss": 2.1453, "step": 2820 }, { "epoch": 0.6592854430194852, "grad_norm": 2.583425521850586, "learning_rate": 2.9999306666524743e-05, "loss": 2.0862, "step": 2821 }, { "epoch": 0.6595191493091058, "grad_norm": 2.5969879627227783, "learning_rate": 2.999930119650182e-05, "loss": 2.0866, "step": 2822 }, { "epoch": 0.6597528555987263, "grad_norm": 2.4719419479370117, "learning_rate": 2.9999295704986358e-05, "loss": 2.2309, "step": 2823 }, { "epoch": 0.6599865618883468, "grad_norm": 2.631267547607422, "learning_rate": 2.9999290191978356e-05, "loss": 2.2085, "step": 2824 }, { "epoch": 0.6602202681779673, "grad_norm": 3.134850025177002, "learning_rate": 2.9999284657477825e-05, "loss": 2.014, "step": 2825 }, { "epoch": 0.6604539744675879, "grad_norm": 2.729823350906372, "learning_rate": 2.999927910148477e-05, "loss": 2.087, "step": 2826 }, { "epoch": 0.6606876807572084, "grad_norm": 3.204702854156494, "learning_rate": 2.9999273523999205e-05, "loss": 2.2062, "step": 2827 }, { "epoch": 0.6609213870468289, "grad_norm": 2.630276679992676, "learning_rate": 2.999926792502113e-05, "loss": 2.2233, "step": 2828 }, { "epoch": 0.6611550933364494, "grad_norm": 2.8759477138519287, "learning_rate": 2.9999262304550563e-05, "loss": 2.0768, "step": 2829 }, { "epoch": 0.6613887996260699, "grad_norm": 3.6771204471588135, "learning_rate": 2.9999256662587504e-05, "loss": 2.1793, "step": 2830 }, { "epoch": 0.6616225059156905, "grad_norm": 3.1622836589813232, "learning_rate": 2.9999250999131963e-05, "loss": 2.2337, "step": 2831 }, { "epoch": 0.661856212205311, "grad_norm": 3.7809534072875977, "learning_rate": 2.999924531418395e-05, "loss": 2.1038, "step": 2832 }, { "epoch": 0.6620899184949315, "grad_norm": 4.510531902313232, "learning_rate": 2.9999239607743472e-05, "loss": 2.163, "step": 2833 }, { "epoch": 0.662323624784552, "grad_norm": 3.060248851776123, "learning_rate": 2.999923387981054e-05, "loss": 2.0641, "step": 2834 }, { "epoch": 0.6625573310741726, "grad_norm": 4.23712158203125, "learning_rate": 2.9999228130385156e-05, "loss": 2.0676, "step": 2835 }, { "epoch": 0.6627910373637931, "grad_norm": 4.544587135314941, "learning_rate": 2.9999222359467335e-05, "loss": 2.0873, "step": 2836 }, { "epoch": 0.6630247436534136, "grad_norm": 2.7874855995178223, "learning_rate": 2.999921656705708e-05, "loss": 2.0926, "step": 2837 }, { "epoch": 0.6632584499430341, "grad_norm": 5.2000041007995605, "learning_rate": 2.99992107531544e-05, "loss": 2.2466, "step": 2838 }, { "epoch": 0.6634921562326546, "grad_norm": 3.771899938583374, "learning_rate": 2.9999204917759304e-05, "loss": 2.0066, "step": 2839 }, { "epoch": 0.6637258625222752, "grad_norm": 2.7471160888671875, "learning_rate": 2.9999199060871806e-05, "loss": 2.2777, "step": 2840 }, { "epoch": 0.6639595688118957, "grad_norm": 5.920521259307861, "learning_rate": 2.9999193182491903e-05, "loss": 2.0004, "step": 2841 }, { "epoch": 0.6641932751015162, "grad_norm": 6.1903462409973145, "learning_rate": 2.9999187282619614e-05, "loss": 2.1606, "step": 2842 }, { "epoch": 0.6644269813911367, "grad_norm": 2.939709186553955, "learning_rate": 2.999918136125494e-05, "loss": 2.0059, "step": 2843 }, { "epoch": 0.6646606876807573, "grad_norm": 4.0779547691345215, "learning_rate": 2.9999175418397898e-05, "loss": 2.0352, "step": 2844 }, { "epoch": 0.6648943939703778, "grad_norm": 5.83911657333374, "learning_rate": 2.999916945404849e-05, "loss": 2.0737, "step": 2845 }, { "epoch": 0.6651281002599982, "grad_norm": 3.437966823577881, "learning_rate": 2.9999163468206726e-05, "loss": 2.0733, "step": 2846 }, { "epoch": 0.6653618065496187, "grad_norm": 2.9170584678649902, "learning_rate": 2.9999157460872607e-05, "loss": 2.0504, "step": 2847 }, { "epoch": 0.6655955128392392, "grad_norm": 5.882424354553223, "learning_rate": 2.9999151432046158e-05, "loss": 2.1333, "step": 2848 }, { "epoch": 0.6658292191288598, "grad_norm": 6.40337610244751, "learning_rate": 2.9999145381727373e-05, "loss": 2.0777, "step": 2849 }, { "epoch": 0.6660629254184803, "grad_norm": 4.169745445251465, "learning_rate": 2.999913930991627e-05, "loss": 2.3942, "step": 2850 }, { "epoch": 0.6662966317081008, "grad_norm": 4.347919940948486, "learning_rate": 2.9999133216612853e-05, "loss": 2.1501, "step": 2851 }, { "epoch": 0.6665303379977213, "grad_norm": 3.9642107486724854, "learning_rate": 2.9999127101817127e-05, "loss": 2.1069, "step": 2852 }, { "epoch": 0.6667640442873419, "grad_norm": 3.8516175746917725, "learning_rate": 2.999912096552911e-05, "loss": 2.1273, "step": 2853 }, { "epoch": 0.6669977505769624, "grad_norm": 4.187371253967285, "learning_rate": 2.999911480774881e-05, "loss": 2.2059, "step": 2854 }, { "epoch": 0.6672314568665829, "grad_norm": 3.4064104557037354, "learning_rate": 2.9999108628476225e-05, "loss": 2.0957, "step": 2855 }, { "epoch": 0.6674651631562034, "grad_norm": 3.056704521179199, "learning_rate": 2.9999102427711375e-05, "loss": 2.0996, "step": 2856 }, { "epoch": 0.6676988694458239, "grad_norm": 3.065167188644409, "learning_rate": 2.9999096205454263e-05, "loss": 1.9862, "step": 2857 }, { "epoch": 0.6679325757354445, "grad_norm": 2.3082287311553955, "learning_rate": 2.99990899617049e-05, "loss": 2.2473, "step": 2858 }, { "epoch": 0.668166282025065, "grad_norm": 5.562440395355225, "learning_rate": 2.9999083696463295e-05, "loss": 2.0695, "step": 2859 }, { "epoch": 0.6683999883146855, "grad_norm": 3.7036445140838623, "learning_rate": 2.9999077409729456e-05, "loss": 2.0038, "step": 2860 }, { "epoch": 0.668633694604306, "grad_norm": 2.841172695159912, "learning_rate": 2.9999071101503393e-05, "loss": 1.932, "step": 2861 }, { "epoch": 0.6688674008939266, "grad_norm": 4.2582621574401855, "learning_rate": 2.9999064771785116e-05, "loss": 2.2025, "step": 2862 }, { "epoch": 0.6691011071835471, "grad_norm": 5.523399353027344, "learning_rate": 2.999905842057463e-05, "loss": 2.0785, "step": 2863 }, { "epoch": 0.6693348134731676, "grad_norm": 2.9947681427001953, "learning_rate": 2.999905204787195e-05, "loss": 2.1683, "step": 2864 }, { "epoch": 0.6695685197627881, "grad_norm": 4.3989105224609375, "learning_rate": 2.9999045653677082e-05, "loss": 2.1203, "step": 2865 }, { "epoch": 0.6698022260524087, "grad_norm": 5.113658905029297, "learning_rate": 2.9999039237990036e-05, "loss": 2.0607, "step": 2866 }, { "epoch": 0.6700359323420292, "grad_norm": 4.185638427734375, "learning_rate": 2.9999032800810816e-05, "loss": 2.0752, "step": 2867 }, { "epoch": 0.6702696386316497, "grad_norm": 4.895598411560059, "learning_rate": 2.9999026342139436e-05, "loss": 1.8983, "step": 2868 }, { "epoch": 0.6705033449212702, "grad_norm": 7.8318963050842285, "learning_rate": 2.999901986197591e-05, "loss": 2.078, "step": 2869 }, { "epoch": 0.6707370512108907, "grad_norm": 3.4859516620635986, "learning_rate": 2.9999013360320237e-05, "loss": 2.0487, "step": 2870 }, { "epoch": 0.6709707575005113, "grad_norm": 4.014725208282471, "learning_rate": 2.9999006837172435e-05, "loss": 2.1645, "step": 2871 }, { "epoch": 0.6712044637901318, "grad_norm": 5.483249187469482, "learning_rate": 2.999900029253251e-05, "loss": 2.1176, "step": 2872 }, { "epoch": 0.6714381700797523, "grad_norm": 4.262908935546875, "learning_rate": 2.999899372640047e-05, "loss": 2.0834, "step": 2873 }, { "epoch": 0.6716718763693728, "grad_norm": 2.9162089824676514, "learning_rate": 2.9998987138776323e-05, "loss": 2.1594, "step": 2874 }, { "epoch": 0.6719055826589934, "grad_norm": 4.639408111572266, "learning_rate": 2.9998980529660084e-05, "loss": 2.2306, "step": 2875 }, { "epoch": 0.6721392889486139, "grad_norm": 3.5227694511413574, "learning_rate": 2.999897389905176e-05, "loss": 2.0593, "step": 2876 }, { "epoch": 0.6723729952382343, "grad_norm": 2.5642237663269043, "learning_rate": 2.9998967246951356e-05, "loss": 1.9686, "step": 2877 }, { "epoch": 0.6726067015278548, "grad_norm": 4.479180812835693, "learning_rate": 2.9998960573358888e-05, "loss": 2.1793, "step": 2878 }, { "epoch": 0.6728404078174753, "grad_norm": 4.063263893127441, "learning_rate": 2.999895387827436e-05, "loss": 2.0655, "step": 2879 }, { "epoch": 0.6730741141070959, "grad_norm": 3.100548505783081, "learning_rate": 2.9998947161697794e-05, "loss": 2.0348, "step": 2880 }, { "epoch": 0.6733078203967164, "grad_norm": 6.123075008392334, "learning_rate": 2.9998940423629184e-05, "loss": 2.1718, "step": 2881 }, { "epoch": 0.6735415266863369, "grad_norm": 6.60188627243042, "learning_rate": 2.9998933664068547e-05, "loss": 2.1876, "step": 2882 }, { "epoch": 0.6737752329759574, "grad_norm": 2.800860643386841, "learning_rate": 2.9998926883015886e-05, "loss": 2.1288, "step": 2883 }, { "epoch": 0.674008939265578, "grad_norm": 2.766401767730713, "learning_rate": 2.9998920080471226e-05, "loss": 2.0662, "step": 2884 }, { "epoch": 0.6742426455551985, "grad_norm": 4.491683483123779, "learning_rate": 2.999891325643456e-05, "loss": 2.1557, "step": 2885 }, { "epoch": 0.674476351844819, "grad_norm": 3.5621252059936523, "learning_rate": 2.9998906410905906e-05, "loss": 2.0494, "step": 2886 }, { "epoch": 0.6747100581344395, "grad_norm": 4.105093479156494, "learning_rate": 2.9998899543885275e-05, "loss": 1.9812, "step": 2887 }, { "epoch": 0.67494376442406, "grad_norm": 5.215368747711182, "learning_rate": 2.9998892655372674e-05, "loss": 2.0553, "step": 2888 }, { "epoch": 0.6751774707136806, "grad_norm": 3.3819684982299805, "learning_rate": 2.9998885745368113e-05, "loss": 2.1315, "step": 2889 }, { "epoch": 0.6754111770033011, "grad_norm": 4.857898235321045, "learning_rate": 2.9998878813871602e-05, "loss": 2.1409, "step": 2890 }, { "epoch": 0.6756448832929216, "grad_norm": 5.640170574188232, "learning_rate": 2.999887186088315e-05, "loss": 2.1625, "step": 2891 }, { "epoch": 0.6758785895825421, "grad_norm": 3.2499351501464844, "learning_rate": 2.999886488640277e-05, "loss": 2.0272, "step": 2892 }, { "epoch": 0.6761122958721627, "grad_norm": 2.785362720489502, "learning_rate": 2.999885789043047e-05, "loss": 2.038, "step": 2893 }, { "epoch": 0.6763460021617832, "grad_norm": 5.024521827697754, "learning_rate": 2.9998850872966262e-05, "loss": 2.2178, "step": 2894 }, { "epoch": 0.6765797084514037, "grad_norm": 3.316071033477783, "learning_rate": 2.9998843834010154e-05, "loss": 2.092, "step": 2895 }, { "epoch": 0.6768134147410242, "grad_norm": 2.6291565895080566, "learning_rate": 2.9998836773562156e-05, "loss": 2.1565, "step": 2896 }, { "epoch": 0.6770471210306447, "grad_norm": 4.448175430297852, "learning_rate": 2.9998829691622274e-05, "loss": 2.0966, "step": 2897 }, { "epoch": 0.6772808273202653, "grad_norm": 3.4707348346710205, "learning_rate": 2.999882258819053e-05, "loss": 1.9846, "step": 2898 }, { "epoch": 0.6775145336098858, "grad_norm": 5.837862491607666, "learning_rate": 2.999881546326692e-05, "loss": 2.1463, "step": 2899 }, { "epoch": 0.6777482398995063, "grad_norm": 6.3077569007873535, "learning_rate": 2.9998808316851465e-05, "loss": 1.9731, "step": 2900 }, { "epoch": 0.6779819461891268, "grad_norm": 3.7029573917388916, "learning_rate": 2.999880114894417e-05, "loss": 2.0392, "step": 2901 }, { "epoch": 0.6782156524787474, "grad_norm": 4.303221225738525, "learning_rate": 2.9998793959545048e-05, "loss": 2.0708, "step": 2902 }, { "epoch": 0.6784493587683679, "grad_norm": 5.693370342254639, "learning_rate": 2.9998786748654106e-05, "loss": 2.097, "step": 2903 }, { "epoch": 0.6786830650579884, "grad_norm": 4.223989486694336, "learning_rate": 2.9998779516271356e-05, "loss": 2.1189, "step": 2904 }, { "epoch": 0.6789167713476089, "grad_norm": 2.75368332862854, "learning_rate": 2.999877226239681e-05, "loss": 2.2672, "step": 2905 }, { "epoch": 0.6791504776372295, "grad_norm": 3.80670166015625, "learning_rate": 2.9998764987030475e-05, "loss": 2.0463, "step": 2906 }, { "epoch": 0.67938418392685, "grad_norm": 5.706357479095459, "learning_rate": 2.9998757690172366e-05, "loss": 1.9968, "step": 2907 }, { "epoch": 0.6796178902164705, "grad_norm": 3.1795835494995117, "learning_rate": 2.9998750371822485e-05, "loss": 2.0107, "step": 2908 }, { "epoch": 0.679851596506091, "grad_norm": 3.339547634124756, "learning_rate": 2.9998743031980856e-05, "loss": 2.0264, "step": 2909 }, { "epoch": 0.6800853027957114, "grad_norm": 2.718374013900757, "learning_rate": 2.9998735670647478e-05, "loss": 2.073, "step": 2910 }, { "epoch": 0.680319009085332, "grad_norm": 3.037734270095825, "learning_rate": 2.9998728287822363e-05, "loss": 2.14, "step": 2911 }, { "epoch": 0.6805527153749525, "grad_norm": 2.493868827819824, "learning_rate": 2.999872088350553e-05, "loss": 1.9202, "step": 2912 }, { "epoch": 0.680786421664573, "grad_norm": 3.7167305946350098, "learning_rate": 2.9998713457696984e-05, "loss": 2.0447, "step": 2913 }, { "epoch": 0.6810201279541935, "grad_norm": 2.8679304122924805, "learning_rate": 2.9998706010396727e-05, "loss": 2.0994, "step": 2914 }, { "epoch": 0.6812538342438141, "grad_norm": 2.7433297634124756, "learning_rate": 2.999869854160478e-05, "loss": 2.1599, "step": 2915 }, { "epoch": 0.6814875405334346, "grad_norm": 3.7137959003448486, "learning_rate": 2.9998691051321154e-05, "loss": 1.9613, "step": 2916 }, { "epoch": 0.6817212468230551, "grad_norm": 3.6095938682556152, "learning_rate": 2.999868353954586e-05, "loss": 2.1467, "step": 2917 }, { "epoch": 0.6819549531126756, "grad_norm": 3.1476094722747803, "learning_rate": 2.9998676006278902e-05, "loss": 2.0083, "step": 2918 }, { "epoch": 0.6821886594022961, "grad_norm": 5.047992706298828, "learning_rate": 2.9998668451520298e-05, "loss": 1.9773, "step": 2919 }, { "epoch": 0.6824223656919167, "grad_norm": 6.415522575378418, "learning_rate": 2.9998660875270053e-05, "loss": 1.9993, "step": 2920 }, { "epoch": 0.6826560719815372, "grad_norm": 4.506679534912109, "learning_rate": 2.999865327752818e-05, "loss": 2.1864, "step": 2921 }, { "epoch": 0.6828897782711577, "grad_norm": 4.226666450500488, "learning_rate": 2.9998645658294692e-05, "loss": 2.0894, "step": 2922 }, { "epoch": 0.6831234845607782, "grad_norm": 5.414323329925537, "learning_rate": 2.9998638017569596e-05, "loss": 2.2651, "step": 2923 }, { "epoch": 0.6833571908503988, "grad_norm": 4.446050643920898, "learning_rate": 2.999863035535291e-05, "loss": 2.1186, "step": 2924 }, { "epoch": 0.6835908971400193, "grad_norm": 3.017512798309326, "learning_rate": 2.999862267164464e-05, "loss": 2.0233, "step": 2925 }, { "epoch": 0.6838246034296398, "grad_norm": 3.7686383724212646, "learning_rate": 2.9998614966444795e-05, "loss": 2.0581, "step": 2926 }, { "epoch": 0.6840583097192603, "grad_norm": 2.5159645080566406, "learning_rate": 2.9998607239753387e-05, "loss": 2.1348, "step": 2927 }, { "epoch": 0.6842920160088808, "grad_norm": 2.549560070037842, "learning_rate": 2.999859949157043e-05, "loss": 2.1475, "step": 2928 }, { "epoch": 0.6845257222985014, "grad_norm": 2.8199851512908936, "learning_rate": 2.999859172189593e-05, "loss": 2.1001, "step": 2929 }, { "epoch": 0.6847594285881219, "grad_norm": 2.697706460952759, "learning_rate": 2.999858393072991e-05, "loss": 2.1662, "step": 2930 }, { "epoch": 0.6849931348777424, "grad_norm": 2.7947874069213867, "learning_rate": 2.9998576118072366e-05, "loss": 2.1037, "step": 2931 }, { "epoch": 0.6852268411673629, "grad_norm": 3.162850856781006, "learning_rate": 2.9998568283923316e-05, "loss": 2.1017, "step": 2932 }, { "epoch": 0.6854605474569835, "grad_norm": 2.6347250938415527, "learning_rate": 2.9998560428282774e-05, "loss": 2.1653, "step": 2933 }, { "epoch": 0.685694253746604, "grad_norm": 3.318610191345215, "learning_rate": 2.999855255115075e-05, "loss": 2.2094, "step": 2934 }, { "epoch": 0.6859279600362245, "grad_norm": 2.7956974506378174, "learning_rate": 2.9998544652527248e-05, "loss": 2.0922, "step": 2935 }, { "epoch": 0.686161666325845, "grad_norm": 3.7436749935150146, "learning_rate": 2.999853673241229e-05, "loss": 2.1744, "step": 2936 }, { "epoch": 0.6863953726154655, "grad_norm": 2.77530574798584, "learning_rate": 2.9998528790805878e-05, "loss": 1.9841, "step": 2937 }, { "epoch": 0.6866290789050861, "grad_norm": 2.6034092903137207, "learning_rate": 2.999852082770803e-05, "loss": 1.9168, "step": 2938 }, { "epoch": 0.6868627851947066, "grad_norm": 2.760157346725464, "learning_rate": 2.999851284311876e-05, "loss": 2.0073, "step": 2939 }, { "epoch": 0.687096491484327, "grad_norm": 2.7882213592529297, "learning_rate": 2.999850483703807e-05, "loss": 2.0988, "step": 2940 }, { "epoch": 0.6873301977739475, "grad_norm": 2.494877338409424, "learning_rate": 2.9998496809465975e-05, "loss": 2.1756, "step": 2941 }, { "epoch": 0.6875639040635682, "grad_norm": 4.468499183654785, "learning_rate": 2.9998488760402488e-05, "loss": 2.0338, "step": 2942 }, { "epoch": 0.6877976103531886, "grad_norm": 3.071725845336914, "learning_rate": 2.999848068984762e-05, "loss": 1.9921, "step": 2943 }, { "epoch": 0.6880313166428091, "grad_norm": 2.7028684616088867, "learning_rate": 2.9998472597801384e-05, "loss": 2.1043, "step": 2944 }, { "epoch": 0.6882650229324296, "grad_norm": 3.133124589920044, "learning_rate": 2.9998464484263792e-05, "loss": 2.0999, "step": 2945 }, { "epoch": 0.6884987292220502, "grad_norm": 3.12408185005188, "learning_rate": 2.999845634923485e-05, "loss": 2.0836, "step": 2946 }, { "epoch": 0.6887324355116707, "grad_norm": 3.290550708770752, "learning_rate": 2.9998448192714577e-05, "loss": 2.2078, "step": 2947 }, { "epoch": 0.6889661418012912, "grad_norm": 2.799563407897949, "learning_rate": 2.999844001470298e-05, "loss": 1.923, "step": 2948 }, { "epoch": 0.6891998480909117, "grad_norm": 2.6998703479766846, "learning_rate": 2.9998431815200072e-05, "loss": 2.163, "step": 2949 }, { "epoch": 0.6894335543805322, "grad_norm": 2.6494853496551514, "learning_rate": 2.9998423594205868e-05, "loss": 2.1015, "step": 2950 }, { "epoch": 0.6896672606701528, "grad_norm": 2.6283469200134277, "learning_rate": 2.999841535172037e-05, "loss": 2.1138, "step": 2951 }, { "epoch": 0.6899009669597733, "grad_norm": 2.570309638977051, "learning_rate": 2.9998407087743597e-05, "loss": 2.1168, "step": 2952 }, { "epoch": 0.6901346732493938, "grad_norm": 3.0607221126556396, "learning_rate": 2.9998398802275566e-05, "loss": 2.0529, "step": 2953 }, { "epoch": 0.6903683795390143, "grad_norm": 2.552840232849121, "learning_rate": 2.999839049531628e-05, "loss": 2.0664, "step": 2954 }, { "epoch": 0.6906020858286349, "grad_norm": 2.739959716796875, "learning_rate": 2.9998382166865753e-05, "loss": 2.1665, "step": 2955 }, { "epoch": 0.6908357921182554, "grad_norm": 2.792731761932373, "learning_rate": 2.9998373816924e-05, "loss": 1.8581, "step": 2956 }, { "epoch": 0.6910694984078759, "grad_norm": 2.991180419921875, "learning_rate": 2.999836544549103e-05, "loss": 2.1327, "step": 2957 }, { "epoch": 0.6913032046974964, "grad_norm": 2.7806508541107178, "learning_rate": 2.9998357052566854e-05, "loss": 1.8278, "step": 2958 }, { "epoch": 0.6915369109871169, "grad_norm": 3.0575101375579834, "learning_rate": 2.999834863815149e-05, "loss": 2.0466, "step": 2959 }, { "epoch": 0.6917706172767375, "grad_norm": 3.5834126472473145, "learning_rate": 2.9998340202244942e-05, "loss": 2.0581, "step": 2960 }, { "epoch": 0.692004323566358, "grad_norm": 3.1015872955322266, "learning_rate": 2.9998331744847226e-05, "loss": 1.9146, "step": 2961 }, { "epoch": 0.6922380298559785, "grad_norm": 3.146606206893921, "learning_rate": 2.9998323265958357e-05, "loss": 2.0303, "step": 2962 }, { "epoch": 0.692471736145599, "grad_norm": 2.6989383697509766, "learning_rate": 2.9998314765578344e-05, "loss": 2.2361, "step": 2963 }, { "epoch": 0.6927054424352196, "grad_norm": 2.7099153995513916, "learning_rate": 2.9998306243707195e-05, "loss": 2.0278, "step": 2964 }, { "epoch": 0.6929391487248401, "grad_norm": 3.062739849090576, "learning_rate": 2.999829770034493e-05, "loss": 2.0486, "step": 2965 }, { "epoch": 0.6931728550144606, "grad_norm": 2.664715528488159, "learning_rate": 2.9998289135491557e-05, "loss": 1.916, "step": 2966 }, { "epoch": 0.6934065613040811, "grad_norm": 3.1332929134368896, "learning_rate": 2.999828054914709e-05, "loss": 1.9973, "step": 2967 }, { "epoch": 0.6936402675937016, "grad_norm": 2.8519127368927, "learning_rate": 2.9998271941311544e-05, "loss": 2.1619, "step": 2968 }, { "epoch": 0.6938739738833222, "grad_norm": 2.9288394451141357, "learning_rate": 2.9998263311984924e-05, "loss": 2.1901, "step": 2969 }, { "epoch": 0.6941076801729427, "grad_norm": 4.220013618469238, "learning_rate": 2.999825466116725e-05, "loss": 2.0491, "step": 2970 }, { "epoch": 0.6943413864625632, "grad_norm": 3.268754482269287, "learning_rate": 2.9998245988858524e-05, "loss": 2.1638, "step": 2971 }, { "epoch": 0.6945750927521837, "grad_norm": 2.5511863231658936, "learning_rate": 2.999823729505877e-05, "loss": 2.1903, "step": 2972 }, { "epoch": 0.6948087990418043, "grad_norm": 2.3619983196258545, "learning_rate": 2.999822857976799e-05, "loss": 2.1148, "step": 2973 }, { "epoch": 0.6950425053314248, "grad_norm": 2.5998611450195312, "learning_rate": 2.9998219842986206e-05, "loss": 2.0653, "step": 2974 }, { "epoch": 0.6952762116210452, "grad_norm": 2.7766432762145996, "learning_rate": 2.999821108471343e-05, "loss": 2.1283, "step": 2975 }, { "epoch": 0.6955099179106657, "grad_norm": 3.751664400100708, "learning_rate": 2.9998202304949668e-05, "loss": 2.199, "step": 2976 }, { "epoch": 0.6957436242002862, "grad_norm": 2.785196304321289, "learning_rate": 2.9998193503694933e-05, "loss": 2.0859, "step": 2977 }, { "epoch": 0.6959773304899068, "grad_norm": 4.310913562774658, "learning_rate": 2.999818468094924e-05, "loss": 2.053, "step": 2978 }, { "epoch": 0.6962110367795273, "grad_norm": 2.803645372390747, "learning_rate": 2.9998175836712605e-05, "loss": 2.0455, "step": 2979 }, { "epoch": 0.6964447430691478, "grad_norm": 5.460148811340332, "learning_rate": 2.9998166970985035e-05, "loss": 2.1584, "step": 2980 }, { "epoch": 0.6966784493587683, "grad_norm": 4.645954132080078, "learning_rate": 2.999815808376655e-05, "loss": 2.0119, "step": 2981 }, { "epoch": 0.6969121556483889, "grad_norm": 3.124809980392456, "learning_rate": 2.9998149175057154e-05, "loss": 2.2729, "step": 2982 }, { "epoch": 0.6971458619380094, "grad_norm": 4.802783012390137, "learning_rate": 2.9998140244856862e-05, "loss": 2.1454, "step": 2983 }, { "epoch": 0.6973795682276299, "grad_norm": 4.088217258453369, "learning_rate": 2.9998131293165694e-05, "loss": 2.2641, "step": 2984 }, { "epoch": 0.6976132745172504, "grad_norm": 2.5498149394989014, "learning_rate": 2.9998122319983654e-05, "loss": 2.1417, "step": 2985 }, { "epoch": 0.697846980806871, "grad_norm": 6.9096574783325195, "learning_rate": 2.9998113325310758e-05, "loss": 2.168, "step": 2986 }, { "epoch": 0.6980806870964915, "grad_norm": 6.380491733551025, "learning_rate": 2.999810430914702e-05, "loss": 2.2098, "step": 2987 }, { "epoch": 0.698314393386112, "grad_norm": 3.911074638366699, "learning_rate": 2.9998095271492453e-05, "loss": 2.206, "step": 2988 }, { "epoch": 0.6985480996757325, "grad_norm": 3.8760986328125, "learning_rate": 2.999808621234707e-05, "loss": 2.0734, "step": 2989 }, { "epoch": 0.698781805965353, "grad_norm": 5.42279052734375, "learning_rate": 2.999807713171088e-05, "loss": 2.0012, "step": 2990 }, { "epoch": 0.6990155122549736, "grad_norm": 4.948220729827881, "learning_rate": 2.9998068029583902e-05, "loss": 2.0455, "step": 2991 }, { "epoch": 0.6992492185445941, "grad_norm": 3.359808921813965, "learning_rate": 2.999805890596614e-05, "loss": 2.1517, "step": 2992 }, { "epoch": 0.6994829248342146, "grad_norm": 4.725801467895508, "learning_rate": 2.9998049760857615e-05, "loss": 2.0709, "step": 2993 }, { "epoch": 0.6997166311238351, "grad_norm": 3.207820177078247, "learning_rate": 2.9998040594258342e-05, "loss": 2.2058, "step": 2994 }, { "epoch": 0.6999503374134557, "grad_norm": 2.9111545085906982, "learning_rate": 2.9998031406168325e-05, "loss": 2.0565, "step": 2995 }, { "epoch": 0.7001840437030762, "grad_norm": 2.9157304763793945, "learning_rate": 2.999802219658759e-05, "loss": 2.1433, "step": 2996 }, { "epoch": 0.7004177499926967, "grad_norm": 3.1129767894744873, "learning_rate": 2.999801296551614e-05, "loss": 2.1655, "step": 2997 }, { "epoch": 0.7006514562823172, "grad_norm": 2.6938772201538086, "learning_rate": 2.999800371295399e-05, "loss": 2.1131, "step": 2998 }, { "epoch": 0.7008851625719377, "grad_norm": 4.2311930656433105, "learning_rate": 2.9997994438901153e-05, "loss": 1.9478, "step": 2999 }, { "epoch": 0.7011188688615583, "grad_norm": 2.6969032287597656, "learning_rate": 2.9997985143357645e-05, "loss": 1.932, "step": 3000 }, { "epoch": 0.7013525751511788, "grad_norm": 2.5001230239868164, "learning_rate": 2.9997975826323474e-05, "loss": 2.0885, "step": 3001 }, { "epoch": 0.7015862814407993, "grad_norm": 2.643470287322998, "learning_rate": 2.999796648779866e-05, "loss": 2.0607, "step": 3002 }, { "epoch": 0.7018199877304198, "grad_norm": 2.6343870162963867, "learning_rate": 2.9997957127783216e-05, "loss": 2.0952, "step": 3003 }, { "epoch": 0.7020536940200404, "grad_norm": 2.5970072746276855, "learning_rate": 2.9997947746277153e-05, "loss": 2.0721, "step": 3004 }, { "epoch": 0.7022874003096609, "grad_norm": 2.5831985473632812, "learning_rate": 2.999793834328048e-05, "loss": 2.0914, "step": 3005 }, { "epoch": 0.7025211065992814, "grad_norm": 2.975984573364258, "learning_rate": 2.9997928918793215e-05, "loss": 2.0707, "step": 3006 }, { "epoch": 0.7027548128889018, "grad_norm": 2.667062520980835, "learning_rate": 2.9997919472815375e-05, "loss": 2.0158, "step": 3007 }, { "epoch": 0.7029885191785223, "grad_norm": 2.6537370681762695, "learning_rate": 2.999791000534697e-05, "loss": 2.0475, "step": 3008 }, { "epoch": 0.703222225468143, "grad_norm": 4.16664457321167, "learning_rate": 2.999790051638801e-05, "loss": 2.0331, "step": 3009 }, { "epoch": 0.7034559317577634, "grad_norm": 2.6289424896240234, "learning_rate": 2.9997891005938516e-05, "loss": 2.0699, "step": 3010 }, { "epoch": 0.7036896380473839, "grad_norm": 3.0824356079101562, "learning_rate": 2.9997881473998493e-05, "loss": 2.0915, "step": 3011 }, { "epoch": 0.7039233443370044, "grad_norm": 2.520820140838623, "learning_rate": 2.999787192056796e-05, "loss": 1.9359, "step": 3012 }, { "epoch": 0.704157050626625, "grad_norm": 2.5559916496276855, "learning_rate": 2.9997862345646935e-05, "loss": 2.1054, "step": 3013 }, { "epoch": 0.7043907569162455, "grad_norm": 2.5860233306884766, "learning_rate": 2.9997852749235426e-05, "loss": 2.1608, "step": 3014 }, { "epoch": 0.704624463205866, "grad_norm": 2.437238931655884, "learning_rate": 2.9997843131333445e-05, "loss": 2.0505, "step": 3015 }, { "epoch": 0.7048581694954865, "grad_norm": 2.646415948867798, "learning_rate": 2.9997833491941008e-05, "loss": 2.0259, "step": 3016 }, { "epoch": 0.705091875785107, "grad_norm": 2.575070858001709, "learning_rate": 2.9997823831058125e-05, "loss": 2.0896, "step": 3017 }, { "epoch": 0.7053255820747276, "grad_norm": 4.418297290802002, "learning_rate": 2.9997814148684824e-05, "loss": 2.2097, "step": 3018 }, { "epoch": 0.7055592883643481, "grad_norm": 3.3733408451080322, "learning_rate": 2.99978044448211e-05, "loss": 2.1186, "step": 3019 }, { "epoch": 0.7057929946539686, "grad_norm": 3.0227622985839844, "learning_rate": 2.999779471946698e-05, "loss": 2.1581, "step": 3020 }, { "epoch": 0.7060267009435891, "grad_norm": 5.9544243812561035, "learning_rate": 2.999778497262247e-05, "loss": 2.1943, "step": 3021 }, { "epoch": 0.7062604072332097, "grad_norm": 3.7844583988189697, "learning_rate": 2.9997775204287594e-05, "loss": 1.9916, "step": 3022 }, { "epoch": 0.7064941135228302, "grad_norm": 3.0435009002685547, "learning_rate": 2.9997765414462352e-05, "loss": 2.0962, "step": 3023 }, { "epoch": 0.7067278198124507, "grad_norm": 4.225766181945801, "learning_rate": 2.9997755603146774e-05, "loss": 2.1257, "step": 3024 }, { "epoch": 0.7069615261020712, "grad_norm": 2.838860273361206, "learning_rate": 2.999774577034086e-05, "loss": 2.2125, "step": 3025 }, { "epoch": 0.7071952323916917, "grad_norm": 4.047363758087158, "learning_rate": 2.999773591604463e-05, "loss": 2.081, "step": 3026 }, { "epoch": 0.7074289386813123, "grad_norm": 3.2504448890686035, "learning_rate": 2.9997726040258097e-05, "loss": 2.0556, "step": 3027 }, { "epoch": 0.7076626449709328, "grad_norm": 2.9121456146240234, "learning_rate": 2.9997716142981282e-05, "loss": 2.1397, "step": 3028 }, { "epoch": 0.7078963512605533, "grad_norm": 2.979457139968872, "learning_rate": 2.999770622421419e-05, "loss": 2.0487, "step": 3029 }, { "epoch": 0.7081300575501738, "grad_norm": 2.552992582321167, "learning_rate": 2.9997696283956836e-05, "loss": 2.0956, "step": 3030 }, { "epoch": 0.7083637638397944, "grad_norm": 3.8653175830841064, "learning_rate": 2.999768632220924e-05, "loss": 2.0505, "step": 3031 }, { "epoch": 0.7085974701294149, "grad_norm": 4.135956764221191, "learning_rate": 2.9997676338971408e-05, "loss": 2.1551, "step": 3032 }, { "epoch": 0.7088311764190354, "grad_norm": 3.621908664703369, "learning_rate": 2.9997666334243364e-05, "loss": 1.9869, "step": 3033 }, { "epoch": 0.7090648827086559, "grad_norm": 4.63511848449707, "learning_rate": 2.9997656308025116e-05, "loss": 2.0174, "step": 3034 }, { "epoch": 0.7092985889982765, "grad_norm": 4.428648948669434, "learning_rate": 2.9997646260316682e-05, "loss": 2.2216, "step": 3035 }, { "epoch": 0.709532295287897, "grad_norm": 4.18833589553833, "learning_rate": 2.999763619111807e-05, "loss": 2.083, "step": 3036 }, { "epoch": 0.7097660015775175, "grad_norm": 4.297928810119629, "learning_rate": 2.99976261004293e-05, "loss": 2.1113, "step": 3037 }, { "epoch": 0.709999707867138, "grad_norm": 2.540461301803589, "learning_rate": 2.999761598825039e-05, "loss": 2.0188, "step": 3038 }, { "epoch": 0.7102334141567584, "grad_norm": 3.1472368240356445, "learning_rate": 2.9997605854581345e-05, "loss": 2.1262, "step": 3039 }, { "epoch": 0.710467120446379, "grad_norm": 3.8245396614074707, "learning_rate": 2.9997595699422184e-05, "loss": 2.0337, "step": 3040 }, { "epoch": 0.7107008267359995, "grad_norm": 4.054488658905029, "learning_rate": 2.9997585522772927e-05, "loss": 2.1748, "step": 3041 }, { "epoch": 0.71093453302562, "grad_norm": 3.0311622619628906, "learning_rate": 2.999757532463358e-05, "loss": 2.356, "step": 3042 }, { "epoch": 0.7111682393152405, "grad_norm": 3.8170218467712402, "learning_rate": 2.999756510500416e-05, "loss": 2.2023, "step": 3043 }, { "epoch": 0.7114019456048611, "grad_norm": 4.435495376586914, "learning_rate": 2.9997554863884688e-05, "loss": 2.2864, "step": 3044 }, { "epoch": 0.7116356518944816, "grad_norm": 2.8177621364593506, "learning_rate": 2.9997544601275163e-05, "loss": 2.1626, "step": 3045 }, { "epoch": 0.7118693581841021, "grad_norm": 3.2265923023223877, "learning_rate": 2.999753431717562e-05, "loss": 2.2126, "step": 3046 }, { "epoch": 0.7121030644737226, "grad_norm": 4.3068013191223145, "learning_rate": 2.9997524011586056e-05, "loss": 1.9846, "step": 3047 }, { "epoch": 0.7123367707633431, "grad_norm": 3.0867927074432373, "learning_rate": 2.99975136845065e-05, "loss": 2.2816, "step": 3048 }, { "epoch": 0.7125704770529637, "grad_norm": 2.6415317058563232, "learning_rate": 2.9997503335936957e-05, "loss": 2.0876, "step": 3049 }, { "epoch": 0.7128041833425842, "grad_norm": 2.7689626216888428, "learning_rate": 2.9997492965877447e-05, "loss": 2.1537, "step": 3050 }, { "epoch": 0.7130378896322047, "grad_norm": 4.058720111846924, "learning_rate": 2.9997482574327982e-05, "loss": 1.9087, "step": 3051 }, { "epoch": 0.7132715959218252, "grad_norm": 3.286090850830078, "learning_rate": 2.9997472161288577e-05, "loss": 2.0902, "step": 3052 }, { "epoch": 0.7135053022114458, "grad_norm": 2.3934550285339355, "learning_rate": 2.9997461726759248e-05, "loss": 2.05, "step": 3053 }, { "epoch": 0.7137390085010663, "grad_norm": 2.873445749282837, "learning_rate": 2.9997451270740012e-05, "loss": 2.142, "step": 3054 }, { "epoch": 0.7139727147906868, "grad_norm": 4.188594818115234, "learning_rate": 2.999744079323088e-05, "loss": 2.127, "step": 3055 }, { "epoch": 0.7142064210803073, "grad_norm": 4.901712417602539, "learning_rate": 2.999743029423187e-05, "loss": 1.8862, "step": 3056 }, { "epoch": 0.7144401273699278, "grad_norm": 2.911168098449707, "learning_rate": 2.9997419773742995e-05, "loss": 2.0581, "step": 3057 }, { "epoch": 0.7146738336595484, "grad_norm": 2.4151530265808105, "learning_rate": 2.999740923176427e-05, "loss": 2.1416, "step": 3058 }, { "epoch": 0.7149075399491689, "grad_norm": 2.4791276454925537, "learning_rate": 2.9997398668295714e-05, "loss": 2.1514, "step": 3059 }, { "epoch": 0.7151412462387894, "grad_norm": 2.7415223121643066, "learning_rate": 2.9997388083337334e-05, "loss": 2.0701, "step": 3060 }, { "epoch": 0.7153749525284099, "grad_norm": 3.199341058731079, "learning_rate": 2.9997377476889154e-05, "loss": 1.9751, "step": 3061 }, { "epoch": 0.7156086588180305, "grad_norm": 2.456761360168457, "learning_rate": 2.9997366848951184e-05, "loss": 1.9537, "step": 3062 }, { "epoch": 0.715842365107651, "grad_norm": 2.4423329830169678, "learning_rate": 2.9997356199523444e-05, "loss": 2.0815, "step": 3063 }, { "epoch": 0.7160760713972715, "grad_norm": 2.513082504272461, "learning_rate": 2.9997345528605942e-05, "loss": 2.1682, "step": 3064 }, { "epoch": 0.716309777686892, "grad_norm": 3.525508403778076, "learning_rate": 2.9997334836198703e-05, "loss": 2.2535, "step": 3065 }, { "epoch": 0.7165434839765125, "grad_norm": 2.479215383529663, "learning_rate": 2.9997324122301733e-05, "loss": 2.1027, "step": 3066 }, { "epoch": 0.7167771902661331, "grad_norm": 3.8647849559783936, "learning_rate": 2.9997313386915045e-05, "loss": 2.1127, "step": 3067 }, { "epoch": 0.7170108965557536, "grad_norm": 3.047086715698242, "learning_rate": 2.9997302630038668e-05, "loss": 1.9765, "step": 3068 }, { "epoch": 0.7172446028453741, "grad_norm": 2.5798349380493164, "learning_rate": 2.999729185167261e-05, "loss": 2.2146, "step": 3069 }, { "epoch": 0.7174783091349946, "grad_norm": 4.248068809509277, "learning_rate": 2.9997281051816884e-05, "loss": 2.0022, "step": 3070 }, { "epoch": 0.7177120154246152, "grad_norm": 3.353222370147705, "learning_rate": 2.9997270230471508e-05, "loss": 2.0442, "step": 3071 }, { "epoch": 0.7179457217142357, "grad_norm": 3.149198055267334, "learning_rate": 2.9997259387636496e-05, "loss": 2.0847, "step": 3072 }, { "epoch": 0.7181794280038561, "grad_norm": 4.5494561195373535, "learning_rate": 2.9997248523311865e-05, "loss": 2.1281, "step": 3073 }, { "epoch": 0.7184131342934766, "grad_norm": 2.641202926635742, "learning_rate": 2.9997237637497632e-05, "loss": 1.9564, "step": 3074 }, { "epoch": 0.7186468405830972, "grad_norm": 2.558511972427368, "learning_rate": 2.9997226730193808e-05, "loss": 2.1478, "step": 3075 }, { "epoch": 0.7188805468727177, "grad_norm": 2.844970226287842, "learning_rate": 2.999721580140042e-05, "loss": 2.1974, "step": 3076 }, { "epoch": 0.7191142531623382, "grad_norm": 3.1146657466888428, "learning_rate": 2.9997204851117464e-05, "loss": 2.1873, "step": 3077 }, { "epoch": 0.7193479594519587, "grad_norm": 3.1705353260040283, "learning_rate": 2.9997193879344973e-05, "loss": 2.191, "step": 3078 }, { "epoch": 0.7195816657415792, "grad_norm": 3.9618518352508545, "learning_rate": 2.9997182886082954e-05, "loss": 2.0945, "step": 3079 }, { "epoch": 0.7198153720311998, "grad_norm": 3.7074782848358154, "learning_rate": 2.9997171871331424e-05, "loss": 1.8796, "step": 3080 }, { "epoch": 0.7200490783208203, "grad_norm": 3.609999656677246, "learning_rate": 2.9997160835090405e-05, "loss": 2.051, "step": 3081 }, { "epoch": 0.7202827846104408, "grad_norm": 2.854372501373291, "learning_rate": 2.9997149777359902e-05, "loss": 2.0058, "step": 3082 }, { "epoch": 0.7205164909000613, "grad_norm": 2.5392799377441406, "learning_rate": 2.9997138698139942e-05, "loss": 2.1481, "step": 3083 }, { "epoch": 0.7207501971896819, "grad_norm": 2.3814072608947754, "learning_rate": 2.9997127597430537e-05, "loss": 2.0616, "step": 3084 }, { "epoch": 0.7209839034793024, "grad_norm": 2.8804893493652344, "learning_rate": 2.9997116475231692e-05, "loss": 2.0086, "step": 3085 }, { "epoch": 0.7212176097689229, "grad_norm": 2.6600146293640137, "learning_rate": 2.9997105331543442e-05, "loss": 2.1964, "step": 3086 }, { "epoch": 0.7214513160585434, "grad_norm": 2.785337448120117, "learning_rate": 2.9997094166365793e-05, "loss": 2.1451, "step": 3087 }, { "epoch": 0.7216850223481639, "grad_norm": 2.738849639892578, "learning_rate": 2.999708297969876e-05, "loss": 2.0911, "step": 3088 }, { "epoch": 0.7219187286377845, "grad_norm": 2.6911020278930664, "learning_rate": 2.9997071771542357e-05, "loss": 2.2355, "step": 3089 }, { "epoch": 0.722152434927405, "grad_norm": 3.311570882797241, "learning_rate": 2.9997060541896607e-05, "loss": 2.2214, "step": 3090 }, { "epoch": 0.7223861412170255, "grad_norm": 2.44940447807312, "learning_rate": 2.999704929076152e-05, "loss": 2.1459, "step": 3091 }, { "epoch": 0.722619847506646, "grad_norm": 3.4848430156707764, "learning_rate": 2.9997038018137118e-05, "loss": 2.1154, "step": 3092 }, { "epoch": 0.7228535537962666, "grad_norm": 3.176057815551758, "learning_rate": 2.9997026724023413e-05, "loss": 2.0552, "step": 3093 }, { "epoch": 0.7230872600858871, "grad_norm": 2.7971203327178955, "learning_rate": 2.999701540842042e-05, "loss": 2.0199, "step": 3094 }, { "epoch": 0.7233209663755076, "grad_norm": 2.84104323387146, "learning_rate": 2.999700407132816e-05, "loss": 2.0563, "step": 3095 }, { "epoch": 0.7235546726651281, "grad_norm": 2.489424467086792, "learning_rate": 2.9996992712746647e-05, "loss": 2.0363, "step": 3096 }, { "epoch": 0.7237883789547486, "grad_norm": 2.6997742652893066, "learning_rate": 2.9996981332675896e-05, "loss": 2.0539, "step": 3097 }, { "epoch": 0.7240220852443692, "grad_norm": 3.133901357650757, "learning_rate": 2.999696993111592e-05, "loss": 1.9531, "step": 3098 }, { "epoch": 0.7242557915339897, "grad_norm": 4.255972385406494, "learning_rate": 2.9996958508066746e-05, "loss": 2.0255, "step": 3099 }, { "epoch": 0.7244894978236102, "grad_norm": 3.1530070304870605, "learning_rate": 2.999694706352838e-05, "loss": 1.9595, "step": 3100 }, { "epoch": 0.7247232041132307, "grad_norm": 3.6418025493621826, "learning_rate": 2.9996935597500842e-05, "loss": 2.0847, "step": 3101 }, { "epoch": 0.7249569104028513, "grad_norm": 3.2136762142181396, "learning_rate": 2.9996924109984148e-05, "loss": 2.0473, "step": 3102 }, { "epoch": 0.7251906166924718, "grad_norm": 3.359095335006714, "learning_rate": 2.999691260097832e-05, "loss": 1.8941, "step": 3103 }, { "epoch": 0.7254243229820923, "grad_norm": 2.603652238845825, "learning_rate": 2.9996901070483367e-05, "loss": 2.2193, "step": 3104 }, { "epoch": 0.7256580292717127, "grad_norm": 2.3877604007720947, "learning_rate": 2.9996889518499307e-05, "loss": 2.1368, "step": 3105 }, { "epoch": 0.7258917355613332, "grad_norm": 4.229381084442139, "learning_rate": 2.999687794502616e-05, "loss": 2.0238, "step": 3106 }, { "epoch": 0.7261254418509538, "grad_norm": 4.197469234466553, "learning_rate": 2.9996866350063936e-05, "loss": 2.0963, "step": 3107 }, { "epoch": 0.7263591481405743, "grad_norm": 2.8419275283813477, "learning_rate": 2.999685473361266e-05, "loss": 2.1259, "step": 3108 }, { "epoch": 0.7265928544301948, "grad_norm": 3.0350775718688965, "learning_rate": 2.9996843095672342e-05, "loss": 2.1726, "step": 3109 }, { "epoch": 0.7268265607198153, "grad_norm": 2.4867966175079346, "learning_rate": 2.9996831436243003e-05, "loss": 2.0062, "step": 3110 }, { "epoch": 0.7270602670094359, "grad_norm": 2.690397024154663, "learning_rate": 2.9996819755324655e-05, "loss": 2.1225, "step": 3111 }, { "epoch": 0.7272939732990564, "grad_norm": 2.6100170612335205, "learning_rate": 2.9996808052917322e-05, "loss": 1.9848, "step": 3112 }, { "epoch": 0.7275276795886769, "grad_norm": 2.699329137802124, "learning_rate": 2.9996796329021017e-05, "loss": 2.2524, "step": 3113 }, { "epoch": 0.7277613858782974, "grad_norm": 3.1745665073394775, "learning_rate": 2.999678458363575e-05, "loss": 2.1174, "step": 3114 }, { "epoch": 0.727995092167918, "grad_norm": 3.431861639022827, "learning_rate": 2.999677281676155e-05, "loss": 2.2749, "step": 3115 }, { "epoch": 0.7282287984575385, "grad_norm": 4.470460891723633, "learning_rate": 2.9996761028398427e-05, "loss": 2.0625, "step": 3116 }, { "epoch": 0.728462504747159, "grad_norm": 4.392298221588135, "learning_rate": 2.9996749218546397e-05, "loss": 2.0645, "step": 3117 }, { "epoch": 0.7286962110367795, "grad_norm": 2.9748430252075195, "learning_rate": 2.9996737387205477e-05, "loss": 2.1171, "step": 3118 }, { "epoch": 0.7289299173264, "grad_norm": 3.6987576484680176, "learning_rate": 2.999672553437569e-05, "loss": 2.1746, "step": 3119 }, { "epoch": 0.7291636236160206, "grad_norm": 3.3007848262786865, "learning_rate": 2.9996713660057047e-05, "loss": 2.2286, "step": 3120 }, { "epoch": 0.7293973299056411, "grad_norm": 2.8390488624572754, "learning_rate": 2.9996701764249564e-05, "loss": 2.1433, "step": 3121 }, { "epoch": 0.7296310361952616, "grad_norm": 2.8289971351623535, "learning_rate": 2.9996689846953265e-05, "loss": 2.0243, "step": 3122 }, { "epoch": 0.7298647424848821, "grad_norm": 2.888181447982788, "learning_rate": 2.999667790816816e-05, "loss": 2.0709, "step": 3123 }, { "epoch": 0.7300984487745027, "grad_norm": 2.5714306831359863, "learning_rate": 2.999666594789427e-05, "loss": 1.95, "step": 3124 }, { "epoch": 0.7303321550641232, "grad_norm": 2.687187910079956, "learning_rate": 2.9996653966131616e-05, "loss": 1.9806, "step": 3125 }, { "epoch": 0.7305658613537437, "grad_norm": 2.8114893436431885, "learning_rate": 2.9996641962880202e-05, "loss": 2.0496, "step": 3126 }, { "epoch": 0.7307995676433642, "grad_norm": 2.6623175144195557, "learning_rate": 2.9996629938140058e-05, "loss": 2.2084, "step": 3127 }, { "epoch": 0.7310332739329847, "grad_norm": 3.0764567852020264, "learning_rate": 2.99966178919112e-05, "loss": 2.058, "step": 3128 }, { "epoch": 0.7312669802226053, "grad_norm": 2.7083888053894043, "learning_rate": 2.9996605824193633e-05, "loss": 1.9986, "step": 3129 }, { "epoch": 0.7315006865122258, "grad_norm": 2.404050827026367, "learning_rate": 2.999659373498739e-05, "loss": 2.1884, "step": 3130 }, { "epoch": 0.7317343928018463, "grad_norm": 2.7730143070220947, "learning_rate": 2.9996581624292476e-05, "loss": 2.0951, "step": 3131 }, { "epoch": 0.7319680990914668, "grad_norm": 2.6640708446502686, "learning_rate": 2.999656949210892e-05, "loss": 2.0619, "step": 3132 }, { "epoch": 0.7322018053810874, "grad_norm": 2.5459797382354736, "learning_rate": 2.999655733843673e-05, "loss": 2.0021, "step": 3133 }, { "epoch": 0.7324355116707079, "grad_norm": 2.8478288650512695, "learning_rate": 2.9996545163275924e-05, "loss": 1.8781, "step": 3134 }, { "epoch": 0.7326692179603284, "grad_norm": 2.6415092945098877, "learning_rate": 2.9996532966626526e-05, "loss": 2.0114, "step": 3135 }, { "epoch": 0.7329029242499489, "grad_norm": 4.539912223815918, "learning_rate": 2.9996520748488553e-05, "loss": 2.1211, "step": 3136 }, { "epoch": 0.7331366305395693, "grad_norm": 4.147520542144775, "learning_rate": 2.9996508508862014e-05, "loss": 2.107, "step": 3137 }, { "epoch": 0.73337033682919, "grad_norm": 2.7846550941467285, "learning_rate": 2.999649624774693e-05, "loss": 1.918, "step": 3138 }, { "epoch": 0.7336040431188104, "grad_norm": 3.925058603286743, "learning_rate": 2.9996483965143323e-05, "loss": 2.1151, "step": 3139 }, { "epoch": 0.7338377494084309, "grad_norm": 4.44582986831665, "learning_rate": 2.999647166105121e-05, "loss": 2.0863, "step": 3140 }, { "epoch": 0.7340714556980514, "grad_norm": 2.632319450378418, "learning_rate": 2.99964593354706e-05, "loss": 2.1306, "step": 3141 }, { "epoch": 0.734305161987672, "grad_norm": 5.093031883239746, "learning_rate": 2.9996446988401525e-05, "loss": 2.0509, "step": 3142 }, { "epoch": 0.7345388682772925, "grad_norm": 5.4459686279296875, "learning_rate": 2.999643461984399e-05, "loss": 2.1304, "step": 3143 }, { "epoch": 0.734772574566913, "grad_norm": 2.730285882949829, "learning_rate": 2.999642222979802e-05, "loss": 2.0965, "step": 3144 }, { "epoch": 0.7350062808565335, "grad_norm": 5.014401912689209, "learning_rate": 2.999640981826363e-05, "loss": 1.9829, "step": 3145 }, { "epoch": 0.735239987146154, "grad_norm": 6.9723005294799805, "learning_rate": 2.9996397385240835e-05, "loss": 2.1501, "step": 3146 }, { "epoch": 0.7354736934357746, "grad_norm": 4.840221405029297, "learning_rate": 2.9996384930729658e-05, "loss": 2.2203, "step": 3147 }, { "epoch": 0.7357073997253951, "grad_norm": 2.6442534923553467, "learning_rate": 2.9996372454730115e-05, "loss": 1.9846, "step": 3148 }, { "epoch": 0.7359411060150156, "grad_norm": 7.970198631286621, "learning_rate": 2.9996359957242224e-05, "loss": 2.1253, "step": 3149 }, { "epoch": 0.7361748123046361, "grad_norm": 7.449532508850098, "learning_rate": 2.9996347438266002e-05, "loss": 2.0488, "step": 3150 }, { "epoch": 0.7364085185942567, "grad_norm": 7.14539098739624, "learning_rate": 2.999633489780147e-05, "loss": 2.1467, "step": 3151 }, { "epoch": 0.7366422248838772, "grad_norm": 3.4003117084503174, "learning_rate": 2.999632233584864e-05, "loss": 1.8623, "step": 3152 }, { "epoch": 0.7368759311734977, "grad_norm": 4.628963947296143, "learning_rate": 2.9996309752407533e-05, "loss": 2.0063, "step": 3153 }, { "epoch": 0.7371096374631182, "grad_norm": 5.436675071716309, "learning_rate": 2.999629714747817e-05, "loss": 2.0252, "step": 3154 }, { "epoch": 0.7373433437527388, "grad_norm": 2.8963563442230225, "learning_rate": 2.9996284521060567e-05, "loss": 1.9695, "step": 3155 }, { "epoch": 0.7375770500423593, "grad_norm": 4.3256964683532715, "learning_rate": 2.9996271873154735e-05, "loss": 1.9599, "step": 3156 }, { "epoch": 0.7378107563319798, "grad_norm": 4.0457282066345215, "learning_rate": 2.9996259203760705e-05, "loss": 1.9589, "step": 3157 }, { "epoch": 0.7380444626216003, "grad_norm": 2.591634511947632, "learning_rate": 2.999624651287849e-05, "loss": 2.166, "step": 3158 }, { "epoch": 0.7382781689112208, "grad_norm": 2.792654275894165, "learning_rate": 2.99962338005081e-05, "loss": 2.0704, "step": 3159 }, { "epoch": 0.7385118752008414, "grad_norm": 2.6428890228271484, "learning_rate": 2.9996221066649567e-05, "loss": 2.1974, "step": 3160 }, { "epoch": 0.7387455814904619, "grad_norm": 2.324049234390259, "learning_rate": 2.9996208311302902e-05, "loss": 2.1844, "step": 3161 }, { "epoch": 0.7389792877800824, "grad_norm": 2.9441144466400146, "learning_rate": 2.9996195534468122e-05, "loss": 1.8537, "step": 3162 }, { "epoch": 0.7392129940697029, "grad_norm": 3.4957869052886963, "learning_rate": 2.9996182736145245e-05, "loss": 1.9332, "step": 3163 }, { "epoch": 0.7394467003593235, "grad_norm": 3.685129165649414, "learning_rate": 2.9996169916334297e-05, "loss": 2.1358, "step": 3164 }, { "epoch": 0.739680406648944, "grad_norm": 2.8463168144226074, "learning_rate": 2.9996157075035286e-05, "loss": 2.1171, "step": 3165 }, { "epoch": 0.7399141129385645, "grad_norm": 3.2770471572875977, "learning_rate": 2.9996144212248238e-05, "loss": 1.943, "step": 3166 }, { "epoch": 0.740147819228185, "grad_norm": 2.6923458576202393, "learning_rate": 2.9996131327973167e-05, "loss": 2.0213, "step": 3167 }, { "epoch": 0.7403815255178054, "grad_norm": 2.6955368518829346, "learning_rate": 2.9996118422210093e-05, "loss": 2.0276, "step": 3168 }, { "epoch": 0.740615231807426, "grad_norm": 2.6911966800689697, "learning_rate": 2.9996105494959037e-05, "loss": 1.9737, "step": 3169 }, { "epoch": 0.7408489380970465, "grad_norm": 5.405184268951416, "learning_rate": 2.9996092546220015e-05, "loss": 2.0408, "step": 3170 }, { "epoch": 0.741082644386667, "grad_norm": 3.1705405712127686, "learning_rate": 2.9996079575993044e-05, "loss": 1.9895, "step": 3171 }, { "epoch": 0.7413163506762875, "grad_norm": 3.9101197719573975, "learning_rate": 2.9996066584278146e-05, "loss": 2.1342, "step": 3172 }, { "epoch": 0.7415500569659081, "grad_norm": 4.6401047706604, "learning_rate": 2.9996053571075333e-05, "loss": 1.8982, "step": 3173 }, { "epoch": 0.7417837632555286, "grad_norm": 3.4791009426116943, "learning_rate": 2.9996040536384632e-05, "loss": 2.0452, "step": 3174 }, { "epoch": 0.7420174695451491, "grad_norm": 4.62009859085083, "learning_rate": 2.9996027480206057e-05, "loss": 2.1614, "step": 3175 }, { "epoch": 0.7422511758347696, "grad_norm": 4.742685794830322, "learning_rate": 2.9996014402539632e-05, "loss": 2.1324, "step": 3176 }, { "epoch": 0.7424848821243901, "grad_norm": 3.050679922103882, "learning_rate": 2.9996001303385367e-05, "loss": 2.0291, "step": 3177 }, { "epoch": 0.7427185884140107, "grad_norm": 4.341999530792236, "learning_rate": 2.999598818274329e-05, "loss": 2.0346, "step": 3178 }, { "epoch": 0.7429522947036312, "grad_norm": 4.691953659057617, "learning_rate": 2.999597504061341e-05, "loss": 1.9951, "step": 3179 }, { "epoch": 0.7431860009932517, "grad_norm": 3.006284236907959, "learning_rate": 2.9995961876995754e-05, "loss": 1.9483, "step": 3180 }, { "epoch": 0.7434197072828722, "grad_norm": 4.420060634613037, "learning_rate": 2.999594869189034e-05, "loss": 1.9728, "step": 3181 }, { "epoch": 0.7436534135724928, "grad_norm": 5.056088924407959, "learning_rate": 2.999593548529718e-05, "loss": 2.1532, "step": 3182 }, { "epoch": 0.7438871198621133, "grad_norm": 3.5533764362335205, "learning_rate": 2.9995922257216305e-05, "loss": 2.0972, "step": 3183 }, { "epoch": 0.7441208261517338, "grad_norm": 3.8034026622772217, "learning_rate": 2.9995909007647718e-05, "loss": 2.0362, "step": 3184 }, { "epoch": 0.7443545324413543, "grad_norm": 4.354186058044434, "learning_rate": 2.999589573659145e-05, "loss": 2.1795, "step": 3185 }, { "epoch": 0.7445882387309748, "grad_norm": 3.500171184539795, "learning_rate": 2.9995882444047518e-05, "loss": 2.0999, "step": 3186 }, { "epoch": 0.7448219450205954, "grad_norm": 4.195639610290527, "learning_rate": 2.999586913001594e-05, "loss": 2.0657, "step": 3187 }, { "epoch": 0.7450556513102159, "grad_norm": 2.9961304664611816, "learning_rate": 2.9995855794496737e-05, "loss": 2.0143, "step": 3188 }, { "epoch": 0.7452893575998364, "grad_norm": 2.600064277648926, "learning_rate": 2.999584243748992e-05, "loss": 1.9927, "step": 3189 }, { "epoch": 0.7455230638894569, "grad_norm": 3.1248531341552734, "learning_rate": 2.9995829058995517e-05, "loss": 1.9938, "step": 3190 }, { "epoch": 0.7457567701790775, "grad_norm": 2.5247702598571777, "learning_rate": 2.9995815659013547e-05, "loss": 2.0901, "step": 3191 }, { "epoch": 0.745990476468698, "grad_norm": 3.8182897567749023, "learning_rate": 2.999580223754402e-05, "loss": 2.0403, "step": 3192 }, { "epoch": 0.7462241827583185, "grad_norm": 3.212054967880249, "learning_rate": 2.9995788794586967e-05, "loss": 2.0817, "step": 3193 }, { "epoch": 0.746457889047939, "grad_norm": 3.4597864151000977, "learning_rate": 2.99957753301424e-05, "loss": 2.1611, "step": 3194 }, { "epoch": 0.7466915953375596, "grad_norm": 3.925178289413452, "learning_rate": 2.999576184421034e-05, "loss": 2.1159, "step": 3195 }, { "epoch": 0.7469253016271801, "grad_norm": 2.544508934020996, "learning_rate": 2.999574833679081e-05, "loss": 2.0032, "step": 3196 }, { "epoch": 0.7471590079168006, "grad_norm": 2.330156087875366, "learning_rate": 2.999573480788382e-05, "loss": 2.092, "step": 3197 }, { "epoch": 0.7473927142064211, "grad_norm": 3.9056479930877686, "learning_rate": 2.99957212574894e-05, "loss": 1.9765, "step": 3198 }, { "epoch": 0.7476264204960416, "grad_norm": 2.5170865058898926, "learning_rate": 2.9995707685607564e-05, "loss": 2.1507, "step": 3199 }, { "epoch": 0.7478601267856622, "grad_norm": 5.3600053787231445, "learning_rate": 2.999569409223833e-05, "loss": 2.0344, "step": 3200 }, { "epoch": 0.7480938330752827, "grad_norm": 4.232366561889648, "learning_rate": 2.9995680477381718e-05, "loss": 2.103, "step": 3201 }, { "epoch": 0.7483275393649031, "grad_norm": 2.6918344497680664, "learning_rate": 2.9995666841037755e-05, "loss": 2.0429, "step": 3202 }, { "epoch": 0.7485612456545236, "grad_norm": 4.4109110832214355, "learning_rate": 2.999565318320645e-05, "loss": 1.9764, "step": 3203 }, { "epoch": 0.7487949519441442, "grad_norm": 3.9743971824645996, "learning_rate": 2.999563950388783e-05, "loss": 2.0024, "step": 3204 }, { "epoch": 0.7490286582337647, "grad_norm": 2.6440064907073975, "learning_rate": 2.9995625803081908e-05, "loss": 2.0466, "step": 3205 }, { "epoch": 0.7492623645233852, "grad_norm": 6.128126621246338, "learning_rate": 2.9995612080788716e-05, "loss": 2.1236, "step": 3206 }, { "epoch": 0.7494960708130057, "grad_norm": 6.1943135261535645, "learning_rate": 2.9995598337008256e-05, "loss": 1.9593, "step": 3207 }, { "epoch": 0.7497297771026262, "grad_norm": 4.3618059158325195, "learning_rate": 2.9995584571740562e-05, "loss": 1.9418, "step": 3208 }, { "epoch": 0.7499634833922468, "grad_norm": 3.974355459213257, "learning_rate": 2.9995570784985645e-05, "loss": 2.0562, "step": 3209 }, { "epoch": 0.7501971896818673, "grad_norm": 6.3982930183410645, "learning_rate": 2.999555697674353e-05, "loss": 2.1777, "step": 3210 }, { "epoch": 0.7504308959714878, "grad_norm": 5.720801830291748, "learning_rate": 2.999554314701424e-05, "loss": 1.9619, "step": 3211 }, { "epoch": 0.7506646022611083, "grad_norm": 2.7282605171203613, "learning_rate": 2.9995529295797783e-05, "loss": 2.2809, "step": 3212 }, { "epoch": 0.7508983085507289, "grad_norm": 4.603346824645996, "learning_rate": 2.9995515423094186e-05, "loss": 2.1646, "step": 3213 }, { "epoch": 0.7511320148403494, "grad_norm": 5.508745193481445, "learning_rate": 2.9995501528903473e-05, "loss": 2.0765, "step": 3214 }, { "epoch": 0.7513657211299699, "grad_norm": 5.3963236808776855, "learning_rate": 2.999548761322566e-05, "loss": 2.1618, "step": 3215 }, { "epoch": 0.7515994274195904, "grad_norm": 3.3703689575195312, "learning_rate": 2.9995473676060764e-05, "loss": 2.003, "step": 3216 }, { "epoch": 0.7518331337092109, "grad_norm": 4.2792840003967285, "learning_rate": 2.99954597174088e-05, "loss": 2.1167, "step": 3217 }, { "epoch": 0.7520668399988315, "grad_norm": 3.0171306133270264, "learning_rate": 2.9995445737269807e-05, "loss": 2.1004, "step": 3218 }, { "epoch": 0.752300546288452, "grad_norm": 2.514120578765869, "learning_rate": 2.999543173564379e-05, "loss": 1.9552, "step": 3219 }, { "epoch": 0.7525342525780725, "grad_norm": 5.465395927429199, "learning_rate": 2.9995417712530774e-05, "loss": 2.0772, "step": 3220 }, { "epoch": 0.752767958867693, "grad_norm": 3.8286798000335693, "learning_rate": 2.9995403667930774e-05, "loss": 1.8876, "step": 3221 }, { "epoch": 0.7530016651573136, "grad_norm": 2.563162088394165, "learning_rate": 2.9995389601843815e-05, "loss": 2.0792, "step": 3222 }, { "epoch": 0.7532353714469341, "grad_norm": 4.455422878265381, "learning_rate": 2.9995375514269917e-05, "loss": 2.129, "step": 3223 }, { "epoch": 0.7534690777365546, "grad_norm": 4.246941089630127, "learning_rate": 2.9995361405209097e-05, "loss": 1.8957, "step": 3224 }, { "epoch": 0.7537027840261751, "grad_norm": 2.7022781372070312, "learning_rate": 2.9995347274661383e-05, "loss": 2.1389, "step": 3225 }, { "epoch": 0.7539364903157956, "grad_norm": 3.509553909301758, "learning_rate": 2.9995333122626783e-05, "loss": 2.0838, "step": 3226 }, { "epoch": 0.7541701966054162, "grad_norm": 2.721708297729492, "learning_rate": 2.9995318949105327e-05, "loss": 2.2053, "step": 3227 }, { "epoch": 0.7544039028950367, "grad_norm": 3.4734559059143066, "learning_rate": 2.9995304754097033e-05, "loss": 2.0529, "step": 3228 }, { "epoch": 0.7546376091846572, "grad_norm": 4.005742073059082, "learning_rate": 2.9995290537601918e-05, "loss": 2.0714, "step": 3229 }, { "epoch": 0.7548713154742777, "grad_norm": 5.260129451751709, "learning_rate": 2.9995276299620004e-05, "loss": 2.1515, "step": 3230 }, { "epoch": 0.7551050217638983, "grad_norm": 2.6963565349578857, "learning_rate": 2.9995262040151317e-05, "loss": 2.0822, "step": 3231 }, { "epoch": 0.7553387280535188, "grad_norm": 3.8045506477355957, "learning_rate": 2.9995247759195867e-05, "loss": 2.0782, "step": 3232 }, { "epoch": 0.7555724343431393, "grad_norm": 2.5693533420562744, "learning_rate": 2.9995233456753685e-05, "loss": 2.0265, "step": 3233 }, { "epoch": 0.7558061406327597, "grad_norm": 3.139824151992798, "learning_rate": 2.9995219132824787e-05, "loss": 1.9326, "step": 3234 }, { "epoch": 0.7560398469223802, "grad_norm": 5.229424476623535, "learning_rate": 2.999520478740919e-05, "loss": 1.9306, "step": 3235 }, { "epoch": 0.7562735532120008, "grad_norm": 3.79227876663208, "learning_rate": 2.9995190420506918e-05, "loss": 1.9957, "step": 3236 }, { "epoch": 0.7565072595016213, "grad_norm": 3.3169784545898438, "learning_rate": 2.9995176032117993e-05, "loss": 1.985, "step": 3237 }, { "epoch": 0.7567409657912418, "grad_norm": 3.9203665256500244, "learning_rate": 2.9995161622242435e-05, "loss": 1.9703, "step": 3238 }, { "epoch": 0.7569746720808623, "grad_norm": 7.144460201263428, "learning_rate": 2.999514719088026e-05, "loss": 2.0906, "step": 3239 }, { "epoch": 0.7572083783704829, "grad_norm": 3.8159146308898926, "learning_rate": 2.9995132738031494e-05, "loss": 1.9936, "step": 3240 }, { "epoch": 0.7574420846601034, "grad_norm": 4.870062828063965, "learning_rate": 2.999511826369616e-05, "loss": 2.057, "step": 3241 }, { "epoch": 0.7576757909497239, "grad_norm": 5.47717809677124, "learning_rate": 2.9995103767874268e-05, "loss": 1.9447, "step": 3242 }, { "epoch": 0.7579094972393444, "grad_norm": 3.607241630554199, "learning_rate": 2.9995089250565848e-05, "loss": 2.0544, "step": 3243 }, { "epoch": 0.758143203528965, "grad_norm": 3.9772391319274902, "learning_rate": 2.999507471177092e-05, "loss": 2.0958, "step": 3244 }, { "epoch": 0.7583769098185855, "grad_norm": 5.466676235198975, "learning_rate": 2.9995060151489506e-05, "loss": 2.1398, "step": 3245 }, { "epoch": 0.758610616108206, "grad_norm": 2.9913723468780518, "learning_rate": 2.9995045569721617e-05, "loss": 1.9653, "step": 3246 }, { "epoch": 0.7588443223978265, "grad_norm": 2.6291756629943848, "learning_rate": 2.9995030966467285e-05, "loss": 2.037, "step": 3247 }, { "epoch": 0.759078028687447, "grad_norm": 4.826300144195557, "learning_rate": 2.9995016341726525e-05, "loss": 2.2167, "step": 3248 }, { "epoch": 0.7593117349770676, "grad_norm": 3.0058202743530273, "learning_rate": 2.9995001695499362e-05, "loss": 1.9666, "step": 3249 }, { "epoch": 0.7595454412666881, "grad_norm": 3.313514471054077, "learning_rate": 2.9994987027785812e-05, "loss": 2.0259, "step": 3250 }, { "epoch": 0.7597791475563086, "grad_norm": 4.791034698486328, "learning_rate": 2.9994972338585903e-05, "loss": 2.1127, "step": 3251 }, { "epoch": 0.7600128538459291, "grad_norm": 3.325100898742676, "learning_rate": 2.9994957627899648e-05, "loss": 2.0891, "step": 3252 }, { "epoch": 0.7602465601355497, "grad_norm": 3.2335305213928223, "learning_rate": 2.999494289572707e-05, "loss": 1.966, "step": 3253 }, { "epoch": 0.7604802664251702, "grad_norm": 6.0774993896484375, "learning_rate": 2.9994928142068198e-05, "loss": 2.3105, "step": 3254 }, { "epoch": 0.7607139727147907, "grad_norm": 3.212923765182495, "learning_rate": 2.9994913366923043e-05, "loss": 2.1402, "step": 3255 }, { "epoch": 0.7609476790044112, "grad_norm": 2.4429147243499756, "learning_rate": 2.9994898570291628e-05, "loss": 1.9344, "step": 3256 }, { "epoch": 0.7611813852940317, "grad_norm": 3.5728158950805664, "learning_rate": 2.999488375217398e-05, "loss": 1.9421, "step": 3257 }, { "epoch": 0.7614150915836523, "grad_norm": 3.3937206268310547, "learning_rate": 2.999486891257012e-05, "loss": 2.1714, "step": 3258 }, { "epoch": 0.7616487978732728, "grad_norm": 3.346144437789917, "learning_rate": 2.9994854051480064e-05, "loss": 1.9779, "step": 3259 }, { "epoch": 0.7618825041628933, "grad_norm": 3.458651065826416, "learning_rate": 2.9994839168903834e-05, "loss": 2.0039, "step": 3260 }, { "epoch": 0.7621162104525138, "grad_norm": 2.484585762023926, "learning_rate": 2.999482426484145e-05, "loss": 1.9248, "step": 3261 }, { "epoch": 0.7623499167421344, "grad_norm": 3.042677879333496, "learning_rate": 2.9994809339292943e-05, "loss": 2.0426, "step": 3262 }, { "epoch": 0.7625836230317549, "grad_norm": 4.596987247467041, "learning_rate": 2.999479439225832e-05, "loss": 2.0217, "step": 3263 }, { "epoch": 0.7628173293213754, "grad_norm": 2.837996244430542, "learning_rate": 2.9994779423737616e-05, "loss": 2.1909, "step": 3264 }, { "epoch": 0.7630510356109959, "grad_norm": 3.1045846939086914, "learning_rate": 2.9994764433730845e-05, "loss": 2.0901, "step": 3265 }, { "epoch": 0.7632847419006163, "grad_norm": 3.7354230880737305, "learning_rate": 2.9994749422238026e-05, "loss": 2.0058, "step": 3266 }, { "epoch": 0.763518448190237, "grad_norm": 2.9585647583007812, "learning_rate": 2.9994734389259187e-05, "loss": 2.0751, "step": 3267 }, { "epoch": 0.7637521544798574, "grad_norm": 2.640845775604248, "learning_rate": 2.9994719334794348e-05, "loss": 2.1184, "step": 3268 }, { "epoch": 0.7639858607694779, "grad_norm": 2.3456485271453857, "learning_rate": 2.9994704258843528e-05, "loss": 2.1304, "step": 3269 }, { "epoch": 0.7642195670590984, "grad_norm": 3.364197015762329, "learning_rate": 2.999468916140675e-05, "loss": 2.126, "step": 3270 }, { "epoch": 0.764453273348719, "grad_norm": 3.921332836151123, "learning_rate": 2.9994674042484037e-05, "loss": 1.8875, "step": 3271 }, { "epoch": 0.7646869796383395, "grad_norm": 2.7153806686401367, "learning_rate": 2.9994658902075406e-05, "loss": 2.105, "step": 3272 }, { "epoch": 0.76492068592796, "grad_norm": 2.5769734382629395, "learning_rate": 2.9994643740180883e-05, "loss": 2.0391, "step": 3273 }, { "epoch": 0.7651543922175805, "grad_norm": 2.536255359649658, "learning_rate": 2.9994628556800488e-05, "loss": 1.8978, "step": 3274 }, { "epoch": 0.765388098507201, "grad_norm": 3.84814453125, "learning_rate": 2.9994613351934246e-05, "loss": 1.8925, "step": 3275 }, { "epoch": 0.7656218047968216, "grad_norm": 3.8747828006744385, "learning_rate": 2.999459812558218e-05, "loss": 2.1706, "step": 3276 }, { "epoch": 0.7658555110864421, "grad_norm": 3.0243515968322754, "learning_rate": 2.99945828777443e-05, "loss": 2.0837, "step": 3277 }, { "epoch": 0.7660892173760626, "grad_norm": 5.114602088928223, "learning_rate": 2.999456760842064e-05, "loss": 2.1048, "step": 3278 }, { "epoch": 0.7663229236656831, "grad_norm": 5.837532043457031, "learning_rate": 2.9994552317611214e-05, "loss": 2.008, "step": 3279 }, { "epoch": 0.7665566299553037, "grad_norm": 2.8692336082458496, "learning_rate": 2.999453700531605e-05, "loss": 2.1931, "step": 3280 }, { "epoch": 0.7667903362449242, "grad_norm": 5.473630905151367, "learning_rate": 2.999452167153517e-05, "loss": 1.9529, "step": 3281 }, { "epoch": 0.7670240425345447, "grad_norm": 6.794943809509277, "learning_rate": 2.9994506316268593e-05, "loss": 1.9614, "step": 3282 }, { "epoch": 0.7672577488241652, "grad_norm": 2.6867995262145996, "learning_rate": 2.9994490939516343e-05, "loss": 2.0442, "step": 3283 }, { "epoch": 0.7674914551137858, "grad_norm": 3.162748098373413, "learning_rate": 2.9994475541278438e-05, "loss": 2.0354, "step": 3284 }, { "epoch": 0.7677251614034063, "grad_norm": 4.634817600250244, "learning_rate": 2.99944601215549e-05, "loss": 2.1845, "step": 3285 }, { "epoch": 0.7679588676930268, "grad_norm": 4.043803691864014, "learning_rate": 2.9994444680345753e-05, "loss": 2.148, "step": 3286 }, { "epoch": 0.7681925739826473, "grad_norm": 3.4864842891693115, "learning_rate": 2.999442921765103e-05, "loss": 1.9821, "step": 3287 }, { "epoch": 0.7684262802722678, "grad_norm": 3.204197406768799, "learning_rate": 2.9994413733470733e-05, "loss": 2.1539, "step": 3288 }, { "epoch": 0.7686599865618884, "grad_norm": 3.425311803817749, "learning_rate": 2.99943982278049e-05, "loss": 2.0621, "step": 3289 }, { "epoch": 0.7688936928515089, "grad_norm": 2.904505491256714, "learning_rate": 2.999438270065354e-05, "loss": 2.0484, "step": 3290 }, { "epoch": 0.7691273991411294, "grad_norm": 3.0420095920562744, "learning_rate": 2.999436715201669e-05, "loss": 1.9693, "step": 3291 }, { "epoch": 0.7693611054307499, "grad_norm": 2.5990233421325684, "learning_rate": 2.9994351581894367e-05, "loss": 2.186, "step": 3292 }, { "epoch": 0.7695948117203705, "grad_norm": 3.0857656002044678, "learning_rate": 2.999433599028659e-05, "loss": 2.0854, "step": 3293 }, { "epoch": 0.769828518009991, "grad_norm": 3.1427009105682373, "learning_rate": 2.9994320377193377e-05, "loss": 1.9894, "step": 3294 }, { "epoch": 0.7700622242996115, "grad_norm": 2.6041200160980225, "learning_rate": 2.9994304742614757e-05, "loss": 2.0059, "step": 3295 }, { "epoch": 0.770295930589232, "grad_norm": 2.3574702739715576, "learning_rate": 2.9994289086550757e-05, "loss": 2.0599, "step": 3296 }, { "epoch": 0.7705296368788525, "grad_norm": 2.5468430519104004, "learning_rate": 2.999427340900139e-05, "loss": 2.0113, "step": 3297 }, { "epoch": 0.7707633431684731, "grad_norm": 2.782317638397217, "learning_rate": 2.9994257709966683e-05, "loss": 2.1255, "step": 3298 }, { "epoch": 0.7709970494580936, "grad_norm": 3.5581319332122803, "learning_rate": 2.9994241989446657e-05, "loss": 2.152, "step": 3299 }, { "epoch": 0.771230755747714, "grad_norm": 2.8792552947998047, "learning_rate": 2.9994226247441335e-05, "loss": 2.2219, "step": 3300 }, { "epoch": 0.7714644620373345, "grad_norm": 3.2028825283050537, "learning_rate": 2.9994210483950745e-05, "loss": 1.8962, "step": 3301 }, { "epoch": 0.7716981683269551, "grad_norm": 2.488284111022949, "learning_rate": 2.9994194698974897e-05, "loss": 2.0749, "step": 3302 }, { "epoch": 0.7719318746165756, "grad_norm": 5.121581554412842, "learning_rate": 2.9994178892513827e-05, "loss": 2.0847, "step": 3303 }, { "epoch": 0.7721655809061961, "grad_norm": 3.027299642562866, "learning_rate": 2.999416306456755e-05, "loss": 2.0052, "step": 3304 }, { "epoch": 0.7723992871958166, "grad_norm": 3.247889757156372, "learning_rate": 2.999414721513609e-05, "loss": 2.0212, "step": 3305 }, { "epoch": 0.7726329934854371, "grad_norm": 3.328782796859741, "learning_rate": 2.9994131344219468e-05, "loss": 2.0041, "step": 3306 }, { "epoch": 0.7728666997750577, "grad_norm": 2.6650025844573975, "learning_rate": 2.9994115451817714e-05, "loss": 2.0831, "step": 3307 }, { "epoch": 0.7731004060646782, "grad_norm": 3.9119949340820312, "learning_rate": 2.9994099537930843e-05, "loss": 1.989, "step": 3308 }, { "epoch": 0.7733341123542987, "grad_norm": 2.71500825881958, "learning_rate": 2.9994083602558878e-05, "loss": 2.1998, "step": 3309 }, { "epoch": 0.7735678186439192, "grad_norm": 2.7153444290161133, "learning_rate": 2.999406764570185e-05, "loss": 2.1766, "step": 3310 }, { "epoch": 0.7738015249335398, "grad_norm": 3.5285301208496094, "learning_rate": 2.9994051667359774e-05, "loss": 1.9579, "step": 3311 }, { "epoch": 0.7740352312231603, "grad_norm": 2.4805595874786377, "learning_rate": 2.999403566753267e-05, "loss": 2.0035, "step": 3312 }, { "epoch": 0.7742689375127808, "grad_norm": 2.3176400661468506, "learning_rate": 2.999401964622057e-05, "loss": 1.9511, "step": 3313 }, { "epoch": 0.7745026438024013, "grad_norm": 2.7844672203063965, "learning_rate": 2.9994003603423493e-05, "loss": 2.0815, "step": 3314 }, { "epoch": 0.7747363500920218, "grad_norm": 3.3939566612243652, "learning_rate": 2.9993987539141464e-05, "loss": 1.999, "step": 3315 }, { "epoch": 0.7749700563816424, "grad_norm": 2.9033491611480713, "learning_rate": 2.99939714533745e-05, "loss": 1.977, "step": 3316 }, { "epoch": 0.7752037626712629, "grad_norm": 3.4498767852783203, "learning_rate": 2.999395534612263e-05, "loss": 2.1368, "step": 3317 }, { "epoch": 0.7754374689608834, "grad_norm": 2.87088942527771, "learning_rate": 2.9993939217385874e-05, "loss": 2.1176, "step": 3318 }, { "epoch": 0.7756711752505039, "grad_norm": 4.0419135093688965, "learning_rate": 2.999392306716426e-05, "loss": 1.9828, "step": 3319 }, { "epoch": 0.7759048815401245, "grad_norm": 4.761674404144287, "learning_rate": 2.9993906895457805e-05, "loss": 1.9453, "step": 3320 }, { "epoch": 0.776138587829745, "grad_norm": 3.928194761276245, "learning_rate": 2.999389070226653e-05, "loss": 2.0814, "step": 3321 }, { "epoch": 0.7763722941193655, "grad_norm": 3.9717366695404053, "learning_rate": 2.999387448759047e-05, "loss": 2.0959, "step": 3322 }, { "epoch": 0.776606000408986, "grad_norm": 3.184352159500122, "learning_rate": 2.9993858251429638e-05, "loss": 1.942, "step": 3323 }, { "epoch": 0.7768397066986066, "grad_norm": 3.935795307159424, "learning_rate": 2.9993841993784058e-05, "loss": 2.0751, "step": 3324 }, { "epoch": 0.7770734129882271, "grad_norm": 2.329821825027466, "learning_rate": 2.999382571465376e-05, "loss": 2.1485, "step": 3325 }, { "epoch": 0.7773071192778476, "grad_norm": 4.205030918121338, "learning_rate": 2.999380941403876e-05, "loss": 2.1908, "step": 3326 }, { "epoch": 0.7775408255674681, "grad_norm": 3.856536626815796, "learning_rate": 2.9993793091939082e-05, "loss": 1.9942, "step": 3327 }, { "epoch": 0.7777745318570886, "grad_norm": 3.1505279541015625, "learning_rate": 2.999377674835476e-05, "loss": 1.9404, "step": 3328 }, { "epoch": 0.7780082381467092, "grad_norm": 2.7641866207122803, "learning_rate": 2.9993760383285798e-05, "loss": 2.0403, "step": 3329 }, { "epoch": 0.7782419444363297, "grad_norm": 2.534954786300659, "learning_rate": 2.9993743996732236e-05, "loss": 1.9957, "step": 3330 }, { "epoch": 0.7784756507259502, "grad_norm": 2.460357189178467, "learning_rate": 2.9993727588694093e-05, "loss": 2.0365, "step": 3331 }, { "epoch": 0.7787093570155706, "grad_norm": 2.776667833328247, "learning_rate": 2.999371115917139e-05, "loss": 2.0416, "step": 3332 }, { "epoch": 0.7789430633051913, "grad_norm": 3.030038595199585, "learning_rate": 2.999369470816415e-05, "loss": 2.0931, "step": 3333 }, { "epoch": 0.7791767695948117, "grad_norm": 3.4263837337493896, "learning_rate": 2.9993678235672403e-05, "loss": 2.1502, "step": 3334 }, { "epoch": 0.7794104758844322, "grad_norm": 5.08666467666626, "learning_rate": 2.9993661741696168e-05, "loss": 2.1561, "step": 3335 }, { "epoch": 0.7796441821740527, "grad_norm": 2.627048969268799, "learning_rate": 2.9993645226235463e-05, "loss": 1.8963, "step": 3336 }, { "epoch": 0.7798778884636732, "grad_norm": 3.3110835552215576, "learning_rate": 2.9993628689290324e-05, "loss": 2.033, "step": 3337 }, { "epoch": 0.7801115947532938, "grad_norm": 5.84624719619751, "learning_rate": 2.9993612130860766e-05, "loss": 2.0657, "step": 3338 }, { "epoch": 0.7803453010429143, "grad_norm": 2.674447774887085, "learning_rate": 2.9993595550946814e-05, "loss": 1.9109, "step": 3339 }, { "epoch": 0.7805790073325348, "grad_norm": 3.931779384613037, "learning_rate": 2.999357894954849e-05, "loss": 1.9877, "step": 3340 }, { "epoch": 0.7808127136221553, "grad_norm": 3.3496756553649902, "learning_rate": 2.9993562326665825e-05, "loss": 2.0873, "step": 3341 }, { "epoch": 0.7810464199117759, "grad_norm": 2.65419340133667, "learning_rate": 2.9993545682298835e-05, "loss": 2.1449, "step": 3342 }, { "epoch": 0.7812801262013964, "grad_norm": 5.254113674163818, "learning_rate": 2.999352901644755e-05, "loss": 2.0141, "step": 3343 }, { "epoch": 0.7815138324910169, "grad_norm": 3.0666489601135254, "learning_rate": 2.9993512329111988e-05, "loss": 2.0087, "step": 3344 }, { "epoch": 0.7817475387806374, "grad_norm": 2.581458330154419, "learning_rate": 2.999349562029218e-05, "loss": 2.0783, "step": 3345 }, { "epoch": 0.7819812450702579, "grad_norm": 2.8119752407073975, "learning_rate": 2.9993478889988143e-05, "loss": 2.0686, "step": 3346 }, { "epoch": 0.7822149513598785, "grad_norm": 5.085569858551025, "learning_rate": 2.9993462138199903e-05, "loss": 2.0875, "step": 3347 }, { "epoch": 0.782448657649499, "grad_norm": 2.730224132537842, "learning_rate": 2.9993445364927485e-05, "loss": 1.9578, "step": 3348 }, { "epoch": 0.7826823639391195, "grad_norm": 3.531280755996704, "learning_rate": 2.9993428570170918e-05, "loss": 2.0076, "step": 3349 }, { "epoch": 0.78291607022874, "grad_norm": 4.263237476348877, "learning_rate": 2.9993411753930213e-05, "loss": 1.9994, "step": 3350 }, { "epoch": 0.7831497765183606, "grad_norm": 3.0235917568206787, "learning_rate": 2.999339491620541e-05, "loss": 2.0799, "step": 3351 }, { "epoch": 0.7833834828079811, "grad_norm": 4.162120819091797, "learning_rate": 2.999337805699652e-05, "loss": 1.9639, "step": 3352 }, { "epoch": 0.7836171890976016, "grad_norm": 6.144708633422852, "learning_rate": 2.9993361176303572e-05, "loss": 2.1491, "step": 3353 }, { "epoch": 0.7838508953872221, "grad_norm": 2.9943652153015137, "learning_rate": 2.999334427412659e-05, "loss": 2.0694, "step": 3354 }, { "epoch": 0.7840846016768426, "grad_norm": 3.759629726409912, "learning_rate": 2.99933273504656e-05, "loss": 1.8818, "step": 3355 }, { "epoch": 0.7843183079664632, "grad_norm": 4.282783508300781, "learning_rate": 2.9993310405320627e-05, "loss": 2.1359, "step": 3356 }, { "epoch": 0.7845520142560837, "grad_norm": 3.3744633197784424, "learning_rate": 2.9993293438691693e-05, "loss": 2.203, "step": 3357 }, { "epoch": 0.7847857205457042, "grad_norm": 4.19279146194458, "learning_rate": 2.9993276450578825e-05, "loss": 2.0244, "step": 3358 }, { "epoch": 0.7850194268353247, "grad_norm": 6.4810004234313965, "learning_rate": 2.999325944098204e-05, "loss": 2.107, "step": 3359 }, { "epoch": 0.7852531331249453, "grad_norm": 4.581513404846191, "learning_rate": 2.9993242409901365e-05, "loss": 2.1381, "step": 3360 }, { "epoch": 0.7854868394145658, "grad_norm": 2.7063307762145996, "learning_rate": 2.9993225357336832e-05, "loss": 2.1083, "step": 3361 }, { "epoch": 0.7857205457041863, "grad_norm": 4.576045989990234, "learning_rate": 2.999320828328846e-05, "loss": 2.015, "step": 3362 }, { "epoch": 0.7859542519938068, "grad_norm": 4.651428699493408, "learning_rate": 2.999319118775627e-05, "loss": 2.0617, "step": 3363 }, { "epoch": 0.7861879582834274, "grad_norm": 2.8072829246520996, "learning_rate": 2.999317407074029e-05, "loss": 2.0574, "step": 3364 }, { "epoch": 0.7864216645730479, "grad_norm": 2.6198089122772217, "learning_rate": 2.999315693224055e-05, "loss": 2.0403, "step": 3365 }, { "epoch": 0.7866553708626683, "grad_norm": 4.449140548706055, "learning_rate": 2.9993139772257064e-05, "loss": 1.9408, "step": 3366 }, { "epoch": 0.7868890771522888, "grad_norm": 3.140977144241333, "learning_rate": 2.9993122590789864e-05, "loss": 2.0347, "step": 3367 }, { "epoch": 0.7871227834419093, "grad_norm": 2.6577744483947754, "learning_rate": 2.999310538783897e-05, "loss": 2.2408, "step": 3368 }, { "epoch": 0.7873564897315299, "grad_norm": 2.4138333797454834, "learning_rate": 2.9993088163404412e-05, "loss": 1.8993, "step": 3369 }, { "epoch": 0.7875901960211504, "grad_norm": 2.4356517791748047, "learning_rate": 2.999307091748621e-05, "loss": 2.0811, "step": 3370 }, { "epoch": 0.7878239023107709, "grad_norm": 2.6219253540039062, "learning_rate": 2.999305365008439e-05, "loss": 2.0454, "step": 3371 }, { "epoch": 0.7880576086003914, "grad_norm": 2.9795753955841064, "learning_rate": 2.999303636119898e-05, "loss": 2.0471, "step": 3372 }, { "epoch": 0.788291314890012, "grad_norm": 2.505953550338745, "learning_rate": 2.999301905083e-05, "loss": 2.0007, "step": 3373 }, { "epoch": 0.7885250211796325, "grad_norm": 3.2809221744537354, "learning_rate": 2.9993001718977475e-05, "loss": 2.0064, "step": 3374 }, { "epoch": 0.788758727469253, "grad_norm": 2.4582533836364746, "learning_rate": 2.9992984365641435e-05, "loss": 1.933, "step": 3375 }, { "epoch": 0.7889924337588735, "grad_norm": 2.3818771839141846, "learning_rate": 2.99929669908219e-05, "loss": 2.0638, "step": 3376 }, { "epoch": 0.789226140048494, "grad_norm": 2.693329095840454, "learning_rate": 2.9992949594518896e-05, "loss": 2.1452, "step": 3377 }, { "epoch": 0.7894598463381146, "grad_norm": 2.716068744659424, "learning_rate": 2.9992932176732445e-05, "loss": 1.9219, "step": 3378 }, { "epoch": 0.7896935526277351, "grad_norm": 2.44929838180542, "learning_rate": 2.999291473746258e-05, "loss": 1.9771, "step": 3379 }, { "epoch": 0.7899272589173556, "grad_norm": 2.270587682723999, "learning_rate": 2.999289727670932e-05, "loss": 1.9422, "step": 3380 }, { "epoch": 0.7901609652069761, "grad_norm": 3.033668279647827, "learning_rate": 2.999287979447269e-05, "loss": 2.0317, "step": 3381 }, { "epoch": 0.7903946714965967, "grad_norm": 2.936354637145996, "learning_rate": 2.999286229075272e-05, "loss": 2.1603, "step": 3382 }, { "epoch": 0.7906283777862172, "grad_norm": 3.532137632369995, "learning_rate": 2.9992844765549427e-05, "loss": 1.9515, "step": 3383 }, { "epoch": 0.7908620840758377, "grad_norm": 3.8498880863189697, "learning_rate": 2.9992827218862843e-05, "loss": 2.0304, "step": 3384 }, { "epoch": 0.7910957903654582, "grad_norm": 2.641995668411255, "learning_rate": 2.999280965069299e-05, "loss": 1.9776, "step": 3385 }, { "epoch": 0.7913294966550787, "grad_norm": 2.763641834259033, "learning_rate": 2.999279206103989e-05, "loss": 2.0068, "step": 3386 }, { "epoch": 0.7915632029446993, "grad_norm": 2.471973180770874, "learning_rate": 2.9992774449903577e-05, "loss": 2.0604, "step": 3387 }, { "epoch": 0.7917969092343198, "grad_norm": 2.744142770767212, "learning_rate": 2.999275681728407e-05, "loss": 2.0935, "step": 3388 }, { "epoch": 0.7920306155239403, "grad_norm": 2.589233636856079, "learning_rate": 2.9992739163181397e-05, "loss": 2.037, "step": 3389 }, { "epoch": 0.7922643218135608, "grad_norm": 2.8565170764923096, "learning_rate": 2.9992721487595576e-05, "loss": 2.0459, "step": 3390 }, { "epoch": 0.7924980281031814, "grad_norm": 3.123812198638916, "learning_rate": 2.9992703790526644e-05, "loss": 2.1465, "step": 3391 }, { "epoch": 0.7927317343928019, "grad_norm": 4.054316997528076, "learning_rate": 2.999268607197462e-05, "loss": 2.0468, "step": 3392 }, { "epoch": 0.7929654406824224, "grad_norm": 4.816154479980469, "learning_rate": 2.9992668331939524e-05, "loss": 2.0494, "step": 3393 }, { "epoch": 0.7931991469720429, "grad_norm": 3.214200019836426, "learning_rate": 2.9992650570421397e-05, "loss": 2.0723, "step": 3394 }, { "epoch": 0.7934328532616634, "grad_norm": 3.4159276485443115, "learning_rate": 2.9992632787420245e-05, "loss": 1.9822, "step": 3395 }, { "epoch": 0.793666559551284, "grad_norm": 5.344353199005127, "learning_rate": 2.999261498293611e-05, "loss": 2.0533, "step": 3396 }, { "epoch": 0.7939002658409045, "grad_norm": 3.983415365219116, "learning_rate": 2.9992597156969008e-05, "loss": 2.074, "step": 3397 }, { "epoch": 0.794133972130525, "grad_norm": 3.727174997329712, "learning_rate": 2.999257930951897e-05, "loss": 2.0442, "step": 3398 }, { "epoch": 0.7943676784201454, "grad_norm": 4.234236717224121, "learning_rate": 2.9992561440586012e-05, "loss": 1.9805, "step": 3399 }, { "epoch": 0.794601384709766, "grad_norm": 3.4327991008758545, "learning_rate": 2.999254355017017e-05, "loss": 2.0154, "step": 3400 }, { "epoch": 0.7948350909993865, "grad_norm": 3.0727078914642334, "learning_rate": 2.999252563827147e-05, "loss": 2.0812, "step": 3401 }, { "epoch": 0.795068797289007, "grad_norm": 3.3549387454986572, "learning_rate": 2.999250770488993e-05, "loss": 2.2418, "step": 3402 }, { "epoch": 0.7953025035786275, "grad_norm": 3.392911911010742, "learning_rate": 2.999248975002558e-05, "loss": 2.0195, "step": 3403 }, { "epoch": 0.7955362098682481, "grad_norm": 3.349036693572998, "learning_rate": 2.9992471773678446e-05, "loss": 2.0954, "step": 3404 }, { "epoch": 0.7957699161578686, "grad_norm": 3.3527674674987793, "learning_rate": 2.9992453775848552e-05, "loss": 2.0209, "step": 3405 }, { "epoch": 0.7960036224474891, "grad_norm": 2.6308507919311523, "learning_rate": 2.9992435756535924e-05, "loss": 2.0391, "step": 3406 }, { "epoch": 0.7962373287371096, "grad_norm": 3.2587532997131348, "learning_rate": 2.9992417715740585e-05, "loss": 1.9687, "step": 3407 }, { "epoch": 0.7964710350267301, "grad_norm": 2.938230514526367, "learning_rate": 2.9992399653462573e-05, "loss": 1.9863, "step": 3408 }, { "epoch": 0.7967047413163507, "grad_norm": 2.9066758155822754, "learning_rate": 2.9992381569701898e-05, "loss": 2.0512, "step": 3409 }, { "epoch": 0.7969384476059712, "grad_norm": 3.1025094985961914, "learning_rate": 2.9992363464458597e-05, "loss": 2.0709, "step": 3410 }, { "epoch": 0.7971721538955917, "grad_norm": 5.804746150970459, "learning_rate": 2.999234533773269e-05, "loss": 1.9261, "step": 3411 }, { "epoch": 0.7974058601852122, "grad_norm": 2.9595882892608643, "learning_rate": 2.999232718952421e-05, "loss": 2.1027, "step": 3412 }, { "epoch": 0.7976395664748328, "grad_norm": 3.712160110473633, "learning_rate": 2.999230901983317e-05, "loss": 2.0026, "step": 3413 }, { "epoch": 0.7978732727644533, "grad_norm": 2.8556811809539795, "learning_rate": 2.999229082865961e-05, "loss": 2.0326, "step": 3414 }, { "epoch": 0.7981069790540738, "grad_norm": 3.7536890506744385, "learning_rate": 2.9992272616003546e-05, "loss": 2.3056, "step": 3415 }, { "epoch": 0.7983406853436943, "grad_norm": 2.714385747909546, "learning_rate": 2.9992254381865012e-05, "loss": 2.2041, "step": 3416 }, { "epoch": 0.7985743916333148, "grad_norm": 5.373486042022705, "learning_rate": 2.9992236126244032e-05, "loss": 1.9545, "step": 3417 }, { "epoch": 0.7988080979229354, "grad_norm": 3.5477285385131836, "learning_rate": 2.999221784914063e-05, "loss": 1.9691, "step": 3418 }, { "epoch": 0.7990418042125559, "grad_norm": 2.610372304916382, "learning_rate": 2.999219955055483e-05, "loss": 1.9599, "step": 3419 }, { "epoch": 0.7992755105021764, "grad_norm": 4.054826259613037, "learning_rate": 2.999218123048666e-05, "loss": 2.1816, "step": 3420 }, { "epoch": 0.7995092167917969, "grad_norm": 2.844397783279419, "learning_rate": 2.9992162888936148e-05, "loss": 2.1238, "step": 3421 }, { "epoch": 0.7997429230814175, "grad_norm": 2.373462677001953, "learning_rate": 2.999214452590332e-05, "loss": 2.1154, "step": 3422 }, { "epoch": 0.799976629371038, "grad_norm": 3.268120527267456, "learning_rate": 2.9992126141388205e-05, "loss": 2.0202, "step": 3423 }, { "epoch": 0.8002103356606585, "grad_norm": 3.8934383392333984, "learning_rate": 2.9992107735390823e-05, "loss": 2.048, "step": 3424 }, { "epoch": 0.800444041950279, "grad_norm": 2.960533618927002, "learning_rate": 2.99920893079112e-05, "loss": 2.182, "step": 3425 }, { "epoch": 0.8006777482398995, "grad_norm": 5.22256326675415, "learning_rate": 2.9992070858949374e-05, "loss": 2.1216, "step": 3426 }, { "epoch": 0.8009114545295201, "grad_norm": 2.846740484237671, "learning_rate": 2.999205238850536e-05, "loss": 2.002, "step": 3427 }, { "epoch": 0.8011451608191406, "grad_norm": 2.9824411869049072, "learning_rate": 2.9992033896579184e-05, "loss": 2.1014, "step": 3428 }, { "epoch": 0.801378867108761, "grad_norm": 4.696046352386475, "learning_rate": 2.9992015383170883e-05, "loss": 2.1434, "step": 3429 }, { "epoch": 0.8016125733983815, "grad_norm": 2.794316530227661, "learning_rate": 2.9991996848280474e-05, "loss": 2.2677, "step": 3430 }, { "epoch": 0.8018462796880022, "grad_norm": 2.5763099193573, "learning_rate": 2.9991978291907984e-05, "loss": 2.0034, "step": 3431 }, { "epoch": 0.8020799859776226, "grad_norm": 2.4520421028137207, "learning_rate": 2.999195971405344e-05, "loss": 2.0229, "step": 3432 }, { "epoch": 0.8023136922672431, "grad_norm": 3.0337860584259033, "learning_rate": 2.999194111471688e-05, "loss": 1.8384, "step": 3433 }, { "epoch": 0.8025473985568636, "grad_norm": 2.745734930038452, "learning_rate": 2.9991922493898315e-05, "loss": 2.0806, "step": 3434 }, { "epoch": 0.8027811048464841, "grad_norm": 2.8267874717712402, "learning_rate": 2.999190385159778e-05, "loss": 2.0586, "step": 3435 }, { "epoch": 0.8030148111361047, "grad_norm": 4.977199077606201, "learning_rate": 2.9991885187815294e-05, "loss": 1.9482, "step": 3436 }, { "epoch": 0.8032485174257252, "grad_norm": 3.3046979904174805, "learning_rate": 2.99918665025509e-05, "loss": 2.0153, "step": 3437 }, { "epoch": 0.8034822237153457, "grad_norm": 3.168452024459839, "learning_rate": 2.9991847795804606e-05, "loss": 2.0172, "step": 3438 }, { "epoch": 0.8037159300049662, "grad_norm": 2.510862112045288, "learning_rate": 2.999182906757645e-05, "loss": 2.0251, "step": 3439 }, { "epoch": 0.8039496362945868, "grad_norm": 2.9580330848693848, "learning_rate": 2.9991810317866454e-05, "loss": 2.1551, "step": 3440 }, { "epoch": 0.8041833425842073, "grad_norm": 2.775876045227051, "learning_rate": 2.9991791546674646e-05, "loss": 2.1022, "step": 3441 }, { "epoch": 0.8044170488738278, "grad_norm": 2.5822908878326416, "learning_rate": 2.9991772754001055e-05, "loss": 1.8734, "step": 3442 }, { "epoch": 0.8046507551634483, "grad_norm": 2.582120895385742, "learning_rate": 2.999175393984571e-05, "loss": 2.128, "step": 3443 }, { "epoch": 0.8048844614530689, "grad_norm": 2.5434536933898926, "learning_rate": 2.9991735104208632e-05, "loss": 2.0662, "step": 3444 }, { "epoch": 0.8051181677426894, "grad_norm": 3.9599897861480713, "learning_rate": 2.999171624708985e-05, "loss": 1.9254, "step": 3445 }, { "epoch": 0.8053518740323099, "grad_norm": 3.5740795135498047, "learning_rate": 2.999169736848939e-05, "loss": 2.1237, "step": 3446 }, { "epoch": 0.8055855803219304, "grad_norm": 2.747279405593872, "learning_rate": 2.9991678468407285e-05, "loss": 2.073, "step": 3447 }, { "epoch": 0.8058192866115509, "grad_norm": 4.091653823852539, "learning_rate": 2.9991659546843557e-05, "loss": 1.9716, "step": 3448 }, { "epoch": 0.8060529929011715, "grad_norm": 2.588785171508789, "learning_rate": 2.999164060379823e-05, "loss": 1.9894, "step": 3449 }, { "epoch": 0.806286699190792, "grad_norm": 2.593823194503784, "learning_rate": 2.9991621639271334e-05, "loss": 2.0047, "step": 3450 }, { "epoch": 0.8065204054804125, "grad_norm": 2.4059696197509766, "learning_rate": 2.9991602653262903e-05, "loss": 1.9559, "step": 3451 }, { "epoch": 0.806754111770033, "grad_norm": 2.6553876399993896, "learning_rate": 2.999158364577295e-05, "loss": 2.0629, "step": 3452 }, { "epoch": 0.8069878180596536, "grad_norm": 2.864245891571045, "learning_rate": 2.9991564616801516e-05, "loss": 2.0693, "step": 3453 }, { "epoch": 0.8072215243492741, "grad_norm": 2.9695346355438232, "learning_rate": 2.9991545566348623e-05, "loss": 2.0226, "step": 3454 }, { "epoch": 0.8074552306388946, "grad_norm": 2.6266770362854004, "learning_rate": 2.9991526494414297e-05, "loss": 2.0397, "step": 3455 }, { "epoch": 0.8076889369285151, "grad_norm": 2.531005859375, "learning_rate": 2.9991507400998565e-05, "loss": 1.9267, "step": 3456 }, { "epoch": 0.8079226432181356, "grad_norm": 2.5505683422088623, "learning_rate": 2.999148828610146e-05, "loss": 2.0286, "step": 3457 }, { "epoch": 0.8081563495077562, "grad_norm": 3.087002754211426, "learning_rate": 2.9991469149723e-05, "loss": 2.0213, "step": 3458 }, { "epoch": 0.8083900557973767, "grad_norm": 2.8799946308135986, "learning_rate": 2.999144999186322e-05, "loss": 2.0745, "step": 3459 }, { "epoch": 0.8086237620869972, "grad_norm": 2.548454523086548, "learning_rate": 2.9991430812522146e-05, "loss": 2.0142, "step": 3460 }, { "epoch": 0.8088574683766176, "grad_norm": 2.6719255447387695, "learning_rate": 2.9991411611699807e-05, "loss": 1.9638, "step": 3461 }, { "epoch": 0.8090911746662383, "grad_norm": 2.7751119136810303, "learning_rate": 2.9991392389396225e-05, "loss": 2.1125, "step": 3462 }, { "epoch": 0.8093248809558587, "grad_norm": 2.7600796222686768, "learning_rate": 2.9991373145611426e-05, "loss": 2.0896, "step": 3463 }, { "epoch": 0.8095585872454792, "grad_norm": 3.2622721195220947, "learning_rate": 2.999135388034545e-05, "loss": 2.0375, "step": 3464 }, { "epoch": 0.8097922935350997, "grad_norm": 2.7820377349853516, "learning_rate": 2.999133459359831e-05, "loss": 2.0199, "step": 3465 }, { "epoch": 0.8100259998247202, "grad_norm": 3.5239570140838623, "learning_rate": 2.9991315285370048e-05, "loss": 1.9831, "step": 3466 }, { "epoch": 0.8102597061143408, "grad_norm": 2.507491111755371, "learning_rate": 2.9991295955660676e-05, "loss": 1.9593, "step": 3467 }, { "epoch": 0.8104934124039613, "grad_norm": 3.5046544075012207, "learning_rate": 2.9991276604470236e-05, "loss": 1.9511, "step": 3468 }, { "epoch": 0.8107271186935818, "grad_norm": 2.575998544692993, "learning_rate": 2.9991257231798743e-05, "loss": 2.0642, "step": 3469 }, { "epoch": 0.8109608249832023, "grad_norm": 3.1376147270202637, "learning_rate": 2.999123783764624e-05, "loss": 1.9686, "step": 3470 }, { "epoch": 0.8111945312728229, "grad_norm": 3.2778799533843994, "learning_rate": 2.9991218422012737e-05, "loss": 1.8959, "step": 3471 }, { "epoch": 0.8114282375624434, "grad_norm": 2.861820936203003, "learning_rate": 2.9991198984898276e-05, "loss": 2.0345, "step": 3472 }, { "epoch": 0.8116619438520639, "grad_norm": 4.724711894989014, "learning_rate": 2.999117952630288e-05, "loss": 2.178, "step": 3473 }, { "epoch": 0.8118956501416844, "grad_norm": 2.5825321674346924, "learning_rate": 2.9991160046226575e-05, "loss": 1.9546, "step": 3474 }, { "epoch": 0.8121293564313049, "grad_norm": 2.420713424682617, "learning_rate": 2.999114054466939e-05, "loss": 1.9918, "step": 3475 }, { "epoch": 0.8123630627209255, "grad_norm": 2.7124178409576416, "learning_rate": 2.9991121021631354e-05, "loss": 2.1605, "step": 3476 }, { "epoch": 0.812596769010546, "grad_norm": 2.763007640838623, "learning_rate": 2.9991101477112494e-05, "loss": 1.949, "step": 3477 }, { "epoch": 0.8128304753001665, "grad_norm": 2.72509503364563, "learning_rate": 2.9991081911112843e-05, "loss": 2.0867, "step": 3478 }, { "epoch": 0.813064181589787, "grad_norm": 3.267976999282837, "learning_rate": 2.9991062323632418e-05, "loss": 2.085, "step": 3479 }, { "epoch": 0.8132978878794076, "grad_norm": 2.7142817974090576, "learning_rate": 2.9991042714671255e-05, "loss": 1.9417, "step": 3480 }, { "epoch": 0.8135315941690281, "grad_norm": 2.57413911819458, "learning_rate": 2.9991023084229384e-05, "loss": 2.0449, "step": 3481 }, { "epoch": 0.8137653004586486, "grad_norm": 2.4721693992614746, "learning_rate": 2.999100343230683e-05, "loss": 2.0727, "step": 3482 }, { "epoch": 0.8139990067482691, "grad_norm": 2.6719295978546143, "learning_rate": 2.9990983758903612e-05, "loss": 1.8711, "step": 3483 }, { "epoch": 0.8142327130378896, "grad_norm": 3.8135287761688232, "learning_rate": 2.9990964064019776e-05, "loss": 2.1416, "step": 3484 }, { "epoch": 0.8144664193275102, "grad_norm": 3.092350959777832, "learning_rate": 2.999094434765534e-05, "loss": 2.1879, "step": 3485 }, { "epoch": 0.8147001256171307, "grad_norm": 2.7638907432556152, "learning_rate": 2.9990924609810332e-05, "loss": 2.1091, "step": 3486 }, { "epoch": 0.8149338319067512, "grad_norm": 2.9776172637939453, "learning_rate": 2.999090485048478e-05, "loss": 2.1343, "step": 3487 }, { "epoch": 0.8151675381963717, "grad_norm": 2.988506317138672, "learning_rate": 2.999088506967872e-05, "loss": 1.9771, "step": 3488 }, { "epoch": 0.8154012444859923, "grad_norm": 3.653111696243286, "learning_rate": 2.9990865267392173e-05, "loss": 2.1176, "step": 3489 }, { "epoch": 0.8156349507756128, "grad_norm": 2.551478862762451, "learning_rate": 2.9990845443625162e-05, "loss": 2.0296, "step": 3490 }, { "epoch": 0.8158686570652333, "grad_norm": 2.7947676181793213, "learning_rate": 2.9990825598377728e-05, "loss": 2.2141, "step": 3491 }, { "epoch": 0.8161023633548538, "grad_norm": 3.0279455184936523, "learning_rate": 2.9990805731649897e-05, "loss": 2.1376, "step": 3492 }, { "epoch": 0.8163360696444744, "grad_norm": 2.5188584327697754, "learning_rate": 2.9990785843441687e-05, "loss": 1.9773, "step": 3493 }, { "epoch": 0.8165697759340949, "grad_norm": 2.847593307495117, "learning_rate": 2.9990765933753138e-05, "loss": 2.1582, "step": 3494 }, { "epoch": 0.8168034822237153, "grad_norm": 2.5099103450775146, "learning_rate": 2.9990746002584267e-05, "loss": 2.0099, "step": 3495 }, { "epoch": 0.8170371885133358, "grad_norm": 2.8451027870178223, "learning_rate": 2.9990726049935118e-05, "loss": 2.0123, "step": 3496 }, { "epoch": 0.8172708948029563, "grad_norm": 2.5861897468566895, "learning_rate": 2.9990706075805712e-05, "loss": 2.0908, "step": 3497 }, { "epoch": 0.8175046010925769, "grad_norm": 2.644139051437378, "learning_rate": 2.999068608019607e-05, "loss": 2.0097, "step": 3498 }, { "epoch": 0.8177383073821974, "grad_norm": 2.5331780910491943, "learning_rate": 2.9990666063106236e-05, "loss": 1.9506, "step": 3499 }, { "epoch": 0.8179720136718179, "grad_norm": 2.9316999912261963, "learning_rate": 2.9990646024536223e-05, "loss": 1.8951, "step": 3500 }, { "epoch": 0.8182057199614384, "grad_norm": 3.1178317070007324, "learning_rate": 2.9990625964486075e-05, "loss": 1.8935, "step": 3501 }, { "epoch": 0.818439426251059, "grad_norm": 2.5937182903289795, "learning_rate": 2.9990605882955805e-05, "loss": 1.9295, "step": 3502 }, { "epoch": 0.8186731325406795, "grad_norm": 3.2518062591552734, "learning_rate": 2.999058577994545e-05, "loss": 2.0313, "step": 3503 }, { "epoch": 0.8189068388303, "grad_norm": 2.9900081157684326, "learning_rate": 2.9990565655455045e-05, "loss": 2.0984, "step": 3504 }, { "epoch": 0.8191405451199205, "grad_norm": 2.828932046890259, "learning_rate": 2.9990545509484605e-05, "loss": 2.0748, "step": 3505 }, { "epoch": 0.819374251409541, "grad_norm": 2.828972101211548, "learning_rate": 2.999052534203417e-05, "loss": 2.1164, "step": 3506 }, { "epoch": 0.8196079576991616, "grad_norm": 2.7671494483947754, "learning_rate": 2.9990505153103764e-05, "loss": 2.0729, "step": 3507 }, { "epoch": 0.8198416639887821, "grad_norm": 3.3570032119750977, "learning_rate": 2.9990484942693416e-05, "loss": 2.0397, "step": 3508 }, { "epoch": 0.8200753702784026, "grad_norm": 2.507689952850342, "learning_rate": 2.999046471080316e-05, "loss": 2.0523, "step": 3509 }, { "epoch": 0.8203090765680231, "grad_norm": 3.4658706188201904, "learning_rate": 2.9990444457433015e-05, "loss": 1.9645, "step": 3510 }, { "epoch": 0.8205427828576437, "grad_norm": 2.9490320682525635, "learning_rate": 2.999042418258302e-05, "loss": 2.1034, "step": 3511 }, { "epoch": 0.8207764891472642, "grad_norm": 3.091395139694214, "learning_rate": 2.9990403886253197e-05, "loss": 1.9599, "step": 3512 }, { "epoch": 0.8210101954368847, "grad_norm": 2.5536882877349854, "learning_rate": 2.999038356844358e-05, "loss": 2.0342, "step": 3513 }, { "epoch": 0.8212439017265052, "grad_norm": 4.170446872711182, "learning_rate": 2.9990363229154195e-05, "loss": 2.0919, "step": 3514 }, { "epoch": 0.8214776080161257, "grad_norm": 2.659024953842163, "learning_rate": 2.9990342868385074e-05, "loss": 2.0915, "step": 3515 }, { "epoch": 0.8217113143057463, "grad_norm": 2.9967901706695557, "learning_rate": 2.9990322486136244e-05, "loss": 2.0521, "step": 3516 }, { "epoch": 0.8219450205953668, "grad_norm": 3.6216201782226562, "learning_rate": 2.9990302082407732e-05, "loss": 2.007, "step": 3517 }, { "epoch": 0.8221787268849873, "grad_norm": 2.4318759441375732, "learning_rate": 2.9990281657199573e-05, "loss": 2.0764, "step": 3518 }, { "epoch": 0.8224124331746078, "grad_norm": 4.0883026123046875, "learning_rate": 2.9990261210511794e-05, "loss": 2.017, "step": 3519 }, { "epoch": 0.8226461394642284, "grad_norm": 2.7787766456604004, "learning_rate": 2.999024074234442e-05, "loss": 2.0712, "step": 3520 }, { "epoch": 0.8228798457538489, "grad_norm": 3.79752779006958, "learning_rate": 2.999022025269748e-05, "loss": 2.1991, "step": 3521 }, { "epoch": 0.8231135520434694, "grad_norm": 2.748656749725342, "learning_rate": 2.999019974157102e-05, "loss": 1.9769, "step": 3522 }, { "epoch": 0.8233472583330899, "grad_norm": 2.468137502670288, "learning_rate": 2.9990179208965046e-05, "loss": 2.0754, "step": 3523 }, { "epoch": 0.8235809646227104, "grad_norm": 3.3170204162597656, "learning_rate": 2.99901586548796e-05, "loss": 2.0289, "step": 3524 }, { "epoch": 0.823814670912331, "grad_norm": 2.670727014541626, "learning_rate": 2.999013807931471e-05, "loss": 2.11, "step": 3525 }, { "epoch": 0.8240483772019515, "grad_norm": 2.8838531970977783, "learning_rate": 2.999011748227041e-05, "loss": 2.1387, "step": 3526 }, { "epoch": 0.824282083491572, "grad_norm": 3.6064491271972656, "learning_rate": 2.9990096863746715e-05, "loss": 2.0979, "step": 3527 }, { "epoch": 0.8245157897811924, "grad_norm": 2.497849941253662, "learning_rate": 2.999007622374367e-05, "loss": 1.9786, "step": 3528 }, { "epoch": 0.824749496070813, "grad_norm": 2.7282798290252686, "learning_rate": 2.9990055562261293e-05, "loss": 1.9232, "step": 3529 }, { "epoch": 0.8249832023604335, "grad_norm": 2.4462177753448486, "learning_rate": 2.9990034879299623e-05, "loss": 2.0791, "step": 3530 }, { "epoch": 0.825216908650054, "grad_norm": 2.401097297668457, "learning_rate": 2.9990014174858687e-05, "loss": 1.9723, "step": 3531 }, { "epoch": 0.8254506149396745, "grad_norm": 2.5625011920928955, "learning_rate": 2.9989993448938514e-05, "loss": 2.106, "step": 3532 }, { "epoch": 0.8256843212292951, "grad_norm": 3.6226325035095215, "learning_rate": 2.998997270153913e-05, "loss": 1.9991, "step": 3533 }, { "epoch": 0.8259180275189156, "grad_norm": 4.230058670043945, "learning_rate": 2.998995193266057e-05, "loss": 2.0357, "step": 3534 }, { "epoch": 0.8261517338085361, "grad_norm": 3.4583029747009277, "learning_rate": 2.998993114230286e-05, "loss": 2.1001, "step": 3535 }, { "epoch": 0.8263854400981566, "grad_norm": 4.095943450927734, "learning_rate": 2.998991033046603e-05, "loss": 2.1189, "step": 3536 }, { "epoch": 0.8266191463877771, "grad_norm": 3.0992558002471924, "learning_rate": 2.9989889497150117e-05, "loss": 1.9793, "step": 3537 }, { "epoch": 0.8268528526773977, "grad_norm": 2.662564754486084, "learning_rate": 2.998986864235514e-05, "loss": 2.0459, "step": 3538 }, { "epoch": 0.8270865589670182, "grad_norm": 3.5491302013397217, "learning_rate": 2.998984776608114e-05, "loss": 2.1671, "step": 3539 }, { "epoch": 0.8273202652566387, "grad_norm": 3.2298495769500732, "learning_rate": 2.9989826868328128e-05, "loss": 2.0114, "step": 3540 }, { "epoch": 0.8275539715462592, "grad_norm": 3.559555768966675, "learning_rate": 2.998980594909616e-05, "loss": 2.1294, "step": 3541 }, { "epoch": 0.8277876778358798, "grad_norm": 2.7220396995544434, "learning_rate": 2.9989785008385248e-05, "loss": 1.9819, "step": 3542 }, { "epoch": 0.8280213841255003, "grad_norm": 3.983184814453125, "learning_rate": 2.9989764046195426e-05, "loss": 1.8959, "step": 3543 }, { "epoch": 0.8282550904151208, "grad_norm": 2.8353328704833984, "learning_rate": 2.9989743062526725e-05, "loss": 1.9757, "step": 3544 }, { "epoch": 0.8284887967047413, "grad_norm": 2.8889989852905273, "learning_rate": 2.9989722057379176e-05, "loss": 2.0383, "step": 3545 }, { "epoch": 0.8287225029943618, "grad_norm": 2.385155200958252, "learning_rate": 2.9989701030752806e-05, "loss": 2.1037, "step": 3546 }, { "epoch": 0.8289562092839824, "grad_norm": 2.6974072456359863, "learning_rate": 2.998967998264765e-05, "loss": 2.0344, "step": 3547 }, { "epoch": 0.8291899155736029, "grad_norm": 3.169640064239502, "learning_rate": 2.9989658913063733e-05, "loss": 2.0827, "step": 3548 }, { "epoch": 0.8294236218632234, "grad_norm": 3.8360342979431152, "learning_rate": 2.998963782200109e-05, "loss": 2.1023, "step": 3549 }, { "epoch": 0.8296573281528439, "grad_norm": 3.9685492515563965, "learning_rate": 2.9989616709459745e-05, "loss": 1.9974, "step": 3550 }, { "epoch": 0.8298910344424645, "grad_norm": 2.902735948562622, "learning_rate": 2.9989595575439735e-05, "loss": 1.8809, "step": 3551 }, { "epoch": 0.830124740732085, "grad_norm": 3.8103222846984863, "learning_rate": 2.9989574419941084e-05, "loss": 2.0777, "step": 3552 }, { "epoch": 0.8303584470217055, "grad_norm": 4.423947334289551, "learning_rate": 2.9989553242963828e-05, "loss": 2.0343, "step": 3553 }, { "epoch": 0.830592153311326, "grad_norm": 2.762244462966919, "learning_rate": 2.9989532044507996e-05, "loss": 2.0041, "step": 3554 }, { "epoch": 0.8308258596009465, "grad_norm": 2.9382145404815674, "learning_rate": 2.9989510824573617e-05, "loss": 2.1104, "step": 3555 }, { "epoch": 0.8310595658905671, "grad_norm": 3.634178400039673, "learning_rate": 2.9989489583160722e-05, "loss": 2.0739, "step": 3556 }, { "epoch": 0.8312932721801876, "grad_norm": 2.5795223712921143, "learning_rate": 2.9989468320269338e-05, "loss": 2.1628, "step": 3557 }, { "epoch": 0.831526978469808, "grad_norm": 2.4644670486450195, "learning_rate": 2.99894470358995e-05, "loss": 1.9963, "step": 3558 }, { "epoch": 0.8317606847594285, "grad_norm": 3.914637327194214, "learning_rate": 2.9989425730051236e-05, "loss": 2.0932, "step": 3559 }, { "epoch": 0.8319943910490492, "grad_norm": 2.8253817558288574, "learning_rate": 2.9989404402724578e-05, "loss": 2.1035, "step": 3560 }, { "epoch": 0.8322280973386696, "grad_norm": 2.7441489696502686, "learning_rate": 2.9989383053919557e-05, "loss": 1.9539, "step": 3561 }, { "epoch": 0.8324618036282901, "grad_norm": 3.282763719558716, "learning_rate": 2.9989361683636203e-05, "loss": 1.9168, "step": 3562 }, { "epoch": 0.8326955099179106, "grad_norm": 2.8129334449768066, "learning_rate": 2.9989340291874547e-05, "loss": 2.1915, "step": 3563 }, { "epoch": 0.8329292162075311, "grad_norm": 3.0995280742645264, "learning_rate": 2.998931887863462e-05, "loss": 1.8346, "step": 3564 }, { "epoch": 0.8331629224971517, "grad_norm": 4.118931770324707, "learning_rate": 2.998929744391645e-05, "loss": 1.8998, "step": 3565 }, { "epoch": 0.8333966287867722, "grad_norm": 2.7246503829956055, "learning_rate": 2.9989275987720067e-05, "loss": 1.9988, "step": 3566 }, { "epoch": 0.8336303350763927, "grad_norm": 2.919879674911499, "learning_rate": 2.9989254510045506e-05, "loss": 2.1304, "step": 3567 }, { "epoch": 0.8338640413660132, "grad_norm": 2.3953709602355957, "learning_rate": 2.99892330108928e-05, "loss": 1.9526, "step": 3568 }, { "epoch": 0.8340977476556338, "grad_norm": 2.5750339031219482, "learning_rate": 2.9989211490261973e-05, "loss": 2.0722, "step": 3569 }, { "epoch": 0.8343314539452543, "grad_norm": 2.655714750289917, "learning_rate": 2.9989189948153056e-05, "loss": 2.1008, "step": 3570 }, { "epoch": 0.8345651602348748, "grad_norm": 2.932671546936035, "learning_rate": 2.9989168384566087e-05, "loss": 1.9661, "step": 3571 }, { "epoch": 0.8347988665244953, "grad_norm": 3.943315029144287, "learning_rate": 2.9989146799501087e-05, "loss": 2.0555, "step": 3572 }, { "epoch": 0.8350325728141159, "grad_norm": 3.555257558822632, "learning_rate": 2.9989125192958096e-05, "loss": 1.8215, "step": 3573 }, { "epoch": 0.8352662791037364, "grad_norm": 4.653425693511963, "learning_rate": 2.9989103564937137e-05, "loss": 2.1072, "step": 3574 }, { "epoch": 0.8354999853933569, "grad_norm": 2.6674656867980957, "learning_rate": 2.9989081915438253e-05, "loss": 2.1055, "step": 3575 }, { "epoch": 0.8357336916829774, "grad_norm": 2.77213978767395, "learning_rate": 2.9989060244461462e-05, "loss": 2.0206, "step": 3576 }, { "epoch": 0.8359673979725979, "grad_norm": 2.315324306488037, "learning_rate": 2.99890385520068e-05, "loss": 1.9662, "step": 3577 }, { "epoch": 0.8362011042622185, "grad_norm": 3.771214723587036, "learning_rate": 2.9989016838074297e-05, "loss": 1.9423, "step": 3578 }, { "epoch": 0.836434810551839, "grad_norm": 3.6398251056671143, "learning_rate": 2.998899510266399e-05, "loss": 2.0349, "step": 3579 }, { "epoch": 0.8366685168414595, "grad_norm": 2.8037631511688232, "learning_rate": 2.99889733457759e-05, "loss": 2.0067, "step": 3580 }, { "epoch": 0.83690222313108, "grad_norm": 2.6713297367095947, "learning_rate": 2.998895156741007e-05, "loss": 2.1242, "step": 3581 }, { "epoch": 0.8371359294207006, "grad_norm": 2.6254935264587402, "learning_rate": 2.9988929767566525e-05, "loss": 2.0392, "step": 3582 }, { "epoch": 0.8373696357103211, "grad_norm": 3.094463348388672, "learning_rate": 2.998890794624529e-05, "loss": 1.9062, "step": 3583 }, { "epoch": 0.8376033419999416, "grad_norm": 2.990570306777954, "learning_rate": 2.9988886103446408e-05, "loss": 1.9354, "step": 3584 }, { "epoch": 0.8378370482895621, "grad_norm": 3.3486242294311523, "learning_rate": 2.9988864239169905e-05, "loss": 2.0462, "step": 3585 }, { "epoch": 0.8380707545791826, "grad_norm": 3.0036139488220215, "learning_rate": 2.9988842353415808e-05, "loss": 1.9665, "step": 3586 }, { "epoch": 0.8383044608688032, "grad_norm": 2.627617597579956, "learning_rate": 2.9988820446184154e-05, "loss": 2.1525, "step": 3587 }, { "epoch": 0.8385381671584237, "grad_norm": 2.2995548248291016, "learning_rate": 2.9988798517474973e-05, "loss": 2.037, "step": 3588 }, { "epoch": 0.8387718734480442, "grad_norm": 2.827878713607788, "learning_rate": 2.99887765672883e-05, "loss": 2.0025, "step": 3589 }, { "epoch": 0.8390055797376647, "grad_norm": 2.781935691833496, "learning_rate": 2.9988754595624157e-05, "loss": 1.9927, "step": 3590 }, { "epoch": 0.8392392860272853, "grad_norm": 2.669185161590576, "learning_rate": 2.9988732602482586e-05, "loss": 2.0448, "step": 3591 }, { "epoch": 0.8394729923169058, "grad_norm": 2.6677374839782715, "learning_rate": 2.9988710587863614e-05, "loss": 2.0584, "step": 3592 }, { "epoch": 0.8397066986065262, "grad_norm": 2.82561993598938, "learning_rate": 2.9988688551767267e-05, "loss": 1.9624, "step": 3593 }, { "epoch": 0.8399404048961467, "grad_norm": 2.7274439334869385, "learning_rate": 2.9988666494193584e-05, "loss": 2.1974, "step": 3594 }, { "epoch": 0.8401741111857672, "grad_norm": 3.126363754272461, "learning_rate": 2.9988644415142597e-05, "loss": 2.1804, "step": 3595 }, { "epoch": 0.8404078174753878, "grad_norm": 2.605231285095215, "learning_rate": 2.9988622314614335e-05, "loss": 2.1157, "step": 3596 }, { "epoch": 0.8406415237650083, "grad_norm": 3.5053677558898926, "learning_rate": 2.9988600192608827e-05, "loss": 1.9497, "step": 3597 }, { "epoch": 0.8408752300546288, "grad_norm": 2.7964587211608887, "learning_rate": 2.998857804912611e-05, "loss": 2.1209, "step": 3598 }, { "epoch": 0.8411089363442493, "grad_norm": 2.658301591873169, "learning_rate": 2.9988555884166214e-05, "loss": 1.944, "step": 3599 }, { "epoch": 0.8413426426338699, "grad_norm": 2.4397854804992676, "learning_rate": 2.9988533697729166e-05, "loss": 2.077, "step": 3600 }, { "epoch": 0.8415763489234904, "grad_norm": 2.6812353134155273, "learning_rate": 2.9988511489815006e-05, "loss": 1.8656, "step": 3601 }, { "epoch": 0.8418100552131109, "grad_norm": 2.9350764751434326, "learning_rate": 2.9988489260423762e-05, "loss": 1.7852, "step": 3602 }, { "epoch": 0.8420437615027314, "grad_norm": 4.85158109664917, "learning_rate": 2.9988467009555462e-05, "loss": 2.0748, "step": 3603 }, { "epoch": 0.8422774677923519, "grad_norm": 2.8048861026763916, "learning_rate": 2.998844473721014e-05, "loss": 2.16, "step": 3604 }, { "epoch": 0.8425111740819725, "grad_norm": 3.1161959171295166, "learning_rate": 2.9988422443387832e-05, "loss": 2.0173, "step": 3605 }, { "epoch": 0.842744880371593, "grad_norm": 2.6960439682006836, "learning_rate": 2.9988400128088572e-05, "loss": 2.062, "step": 3606 }, { "epoch": 0.8429785866612135, "grad_norm": 2.7037782669067383, "learning_rate": 2.998837779131238e-05, "loss": 2.0803, "step": 3607 }, { "epoch": 0.843212292950834, "grad_norm": 2.9768052101135254, "learning_rate": 2.9988355433059295e-05, "loss": 2.052, "step": 3608 }, { "epoch": 0.8434459992404546, "grad_norm": 2.6794044971466064, "learning_rate": 2.9988333053329353e-05, "loss": 1.9888, "step": 3609 }, { "epoch": 0.8436797055300751, "grad_norm": 2.51910400390625, "learning_rate": 2.998831065212258e-05, "loss": 2.0453, "step": 3610 }, { "epoch": 0.8439134118196956, "grad_norm": 2.8307342529296875, "learning_rate": 2.9988288229439012e-05, "loss": 1.9674, "step": 3611 }, { "epoch": 0.8441471181093161, "grad_norm": 2.916210889816284, "learning_rate": 2.9988265785278682e-05, "loss": 2.0453, "step": 3612 }, { "epoch": 0.8443808243989367, "grad_norm": 2.496386766433716, "learning_rate": 2.9988243319641618e-05, "loss": 2.072, "step": 3613 }, { "epoch": 0.8446145306885572, "grad_norm": 2.6899492740631104, "learning_rate": 2.998822083252785e-05, "loss": 1.9902, "step": 3614 }, { "epoch": 0.8448482369781777, "grad_norm": 2.574500560760498, "learning_rate": 2.9988198323937413e-05, "loss": 2.1237, "step": 3615 }, { "epoch": 0.8450819432677982, "grad_norm": 3.892793893814087, "learning_rate": 2.998817579387035e-05, "loss": 2.1182, "step": 3616 }, { "epoch": 0.8453156495574187, "grad_norm": 2.755615234375, "learning_rate": 2.9988153242326677e-05, "loss": 2.0824, "step": 3617 }, { "epoch": 0.8455493558470393, "grad_norm": 2.5642740726470947, "learning_rate": 2.998813066930643e-05, "loss": 2.1232, "step": 3618 }, { "epoch": 0.8457830621366598, "grad_norm": 2.6337647438049316, "learning_rate": 2.998810807480965e-05, "loss": 2.0768, "step": 3619 }, { "epoch": 0.8460167684262803, "grad_norm": 3.0571892261505127, "learning_rate": 2.9988085458836363e-05, "loss": 2.0056, "step": 3620 }, { "epoch": 0.8462504747159008, "grad_norm": 2.6333484649658203, "learning_rate": 2.99880628213866e-05, "loss": 2.0249, "step": 3621 }, { "epoch": 0.8464841810055214, "grad_norm": 2.654442310333252, "learning_rate": 2.9988040162460395e-05, "loss": 2.1393, "step": 3622 }, { "epoch": 0.8467178872951419, "grad_norm": 3.2004473209381104, "learning_rate": 2.9988017482057785e-05, "loss": 2.1281, "step": 3623 }, { "epoch": 0.8469515935847624, "grad_norm": 3.075993537902832, "learning_rate": 2.9987994780178793e-05, "loss": 2.0635, "step": 3624 }, { "epoch": 0.8471852998743828, "grad_norm": 3.1156952381134033, "learning_rate": 2.9987972056823464e-05, "loss": 1.9967, "step": 3625 }, { "epoch": 0.8474190061640033, "grad_norm": 2.9521186351776123, "learning_rate": 2.9987949311991818e-05, "loss": 2.1513, "step": 3626 }, { "epoch": 0.847652712453624, "grad_norm": 2.353376865386963, "learning_rate": 2.9987926545683892e-05, "loss": 1.9653, "step": 3627 }, { "epoch": 0.8478864187432444, "grad_norm": 2.7011878490448, "learning_rate": 2.9987903757899724e-05, "loss": 2.0644, "step": 3628 }, { "epoch": 0.8481201250328649, "grad_norm": 3.0545005798339844, "learning_rate": 2.998788094863934e-05, "loss": 2.2081, "step": 3629 }, { "epoch": 0.8483538313224854, "grad_norm": 3.4772024154663086, "learning_rate": 2.9987858117902778e-05, "loss": 2.1086, "step": 3630 }, { "epoch": 0.848587537612106, "grad_norm": 4.968886852264404, "learning_rate": 2.998783526569006e-05, "loss": 2.1631, "step": 3631 }, { "epoch": 0.8488212439017265, "grad_norm": 4.929360866546631, "learning_rate": 2.9987812392001234e-05, "loss": 2.1667, "step": 3632 }, { "epoch": 0.849054950191347, "grad_norm": 4.842472553253174, "learning_rate": 2.9987789496836324e-05, "loss": 2.0563, "step": 3633 }, { "epoch": 0.8492886564809675, "grad_norm": 3.736696720123291, "learning_rate": 2.9987766580195364e-05, "loss": 2.085, "step": 3634 }, { "epoch": 0.849522362770588, "grad_norm": 4.811767101287842, "learning_rate": 2.998774364207839e-05, "loss": 2.0269, "step": 3635 }, { "epoch": 0.8497560690602086, "grad_norm": 4.516146659851074, "learning_rate": 2.998772068248543e-05, "loss": 2.0352, "step": 3636 }, { "epoch": 0.8499897753498291, "grad_norm": 3.255499839782715, "learning_rate": 2.9987697701416518e-05, "loss": 2.0404, "step": 3637 }, { "epoch": 0.8502234816394496, "grad_norm": 3.565493106842041, "learning_rate": 2.9987674698871685e-05, "loss": 2.0273, "step": 3638 }, { "epoch": 0.8504571879290701, "grad_norm": 2.6357834339141846, "learning_rate": 2.9987651674850966e-05, "loss": 2.1231, "step": 3639 }, { "epoch": 0.8506908942186907, "grad_norm": 2.3442022800445557, "learning_rate": 2.99876286293544e-05, "loss": 2.0016, "step": 3640 }, { "epoch": 0.8509246005083112, "grad_norm": 3.3475804328918457, "learning_rate": 2.9987605562382014e-05, "loss": 1.9061, "step": 3641 }, { "epoch": 0.8511583067979317, "grad_norm": 2.5351755619049072, "learning_rate": 2.9987582473933838e-05, "loss": 1.9532, "step": 3642 }, { "epoch": 0.8513920130875522, "grad_norm": 2.8852760791778564, "learning_rate": 2.998755936400991e-05, "loss": 1.9388, "step": 3643 }, { "epoch": 0.8516257193771727, "grad_norm": 2.878812551498413, "learning_rate": 2.9987536232610266e-05, "loss": 2.0331, "step": 3644 }, { "epoch": 0.8518594256667933, "grad_norm": 2.7095651626586914, "learning_rate": 2.9987513079734927e-05, "loss": 2.0563, "step": 3645 }, { "epoch": 0.8520931319564138, "grad_norm": 2.8648693561553955, "learning_rate": 2.9987489905383943e-05, "loss": 2.0055, "step": 3646 }, { "epoch": 0.8523268382460343, "grad_norm": 2.9346277713775635, "learning_rate": 2.9987466709557332e-05, "loss": 2.0554, "step": 3647 }, { "epoch": 0.8525605445356548, "grad_norm": 2.6011710166931152, "learning_rate": 2.9987443492255137e-05, "loss": 1.9598, "step": 3648 }, { "epoch": 0.8527942508252754, "grad_norm": 3.4797656536102295, "learning_rate": 2.9987420253477386e-05, "loss": 1.9202, "step": 3649 }, { "epoch": 0.8530279571148959, "grad_norm": 2.6962263584136963, "learning_rate": 2.9987396993224118e-05, "loss": 1.9595, "step": 3650 }, { "epoch": 0.8532616634045164, "grad_norm": 2.681260347366333, "learning_rate": 2.998737371149536e-05, "loss": 2.0286, "step": 3651 }, { "epoch": 0.8534953696941369, "grad_norm": 2.721203088760376, "learning_rate": 2.9987350408291145e-05, "loss": 2.0118, "step": 3652 }, { "epoch": 0.8537290759837575, "grad_norm": 3.392770767211914, "learning_rate": 2.9987327083611514e-05, "loss": 2.0261, "step": 3653 }, { "epoch": 0.853962782273378, "grad_norm": 2.6269803047180176, "learning_rate": 2.9987303737456494e-05, "loss": 1.9869, "step": 3654 }, { "epoch": 0.8541964885629985, "grad_norm": 2.561781406402588, "learning_rate": 2.9987280369826117e-05, "loss": 1.9059, "step": 3655 }, { "epoch": 0.854430194852619, "grad_norm": 3.0518813133239746, "learning_rate": 2.9987256980720425e-05, "loss": 2.0564, "step": 3656 }, { "epoch": 0.8546639011422394, "grad_norm": 3.074598550796509, "learning_rate": 2.9987233570139446e-05, "loss": 2.2436, "step": 3657 }, { "epoch": 0.85489760743186, "grad_norm": 2.4059553146362305, "learning_rate": 2.998721013808321e-05, "loss": 1.8852, "step": 3658 }, { "epoch": 0.8551313137214805, "grad_norm": 2.4781947135925293, "learning_rate": 2.998718668455176e-05, "loss": 2.069, "step": 3659 }, { "epoch": 0.855365020011101, "grad_norm": 2.9990830421447754, "learning_rate": 2.9987163209545118e-05, "loss": 2.0801, "step": 3660 }, { "epoch": 0.8555987263007215, "grad_norm": 2.9777328968048096, "learning_rate": 2.9987139713063326e-05, "loss": 1.9859, "step": 3661 }, { "epoch": 0.8558324325903421, "grad_norm": 3.4832923412323, "learning_rate": 2.9987116195106413e-05, "loss": 1.9658, "step": 3662 }, { "epoch": 0.8560661388799626, "grad_norm": 3.2714240550994873, "learning_rate": 2.998709265567442e-05, "loss": 2.1052, "step": 3663 }, { "epoch": 0.8562998451695831, "grad_norm": 2.650588274002075, "learning_rate": 2.998706909476737e-05, "loss": 2.1421, "step": 3664 }, { "epoch": 0.8565335514592036, "grad_norm": 5.550221920013428, "learning_rate": 2.9987045512385306e-05, "loss": 1.9445, "step": 3665 }, { "epoch": 0.8567672577488241, "grad_norm": 3.583881378173828, "learning_rate": 2.9987021908528258e-05, "loss": 2.0226, "step": 3666 }, { "epoch": 0.8570009640384447, "grad_norm": 3.169607639312744, "learning_rate": 2.998699828319626e-05, "loss": 2.0669, "step": 3667 }, { "epoch": 0.8572346703280652, "grad_norm": 3.4073495864868164, "learning_rate": 2.9986974636389343e-05, "loss": 1.9732, "step": 3668 }, { "epoch": 0.8574683766176857, "grad_norm": 2.8394296169281006, "learning_rate": 2.9986950968107548e-05, "loss": 2.192, "step": 3669 }, { "epoch": 0.8577020829073062, "grad_norm": 2.76822566986084, "learning_rate": 2.99869272783509e-05, "loss": 2.0189, "step": 3670 }, { "epoch": 0.8579357891969268, "grad_norm": 3.708615303039551, "learning_rate": 2.9986903567119442e-05, "loss": 1.8904, "step": 3671 }, { "epoch": 0.8581694954865473, "grad_norm": 3.553253173828125, "learning_rate": 2.99868798344132e-05, "loss": 1.8895, "step": 3672 }, { "epoch": 0.8584032017761678, "grad_norm": 2.764413356781006, "learning_rate": 2.9986856080232217e-05, "loss": 1.9422, "step": 3673 }, { "epoch": 0.8586369080657883, "grad_norm": 3.1624865531921387, "learning_rate": 2.9986832304576517e-05, "loss": 1.968, "step": 3674 }, { "epoch": 0.8588706143554088, "grad_norm": 4.59089994430542, "learning_rate": 2.998680850744614e-05, "loss": 2.0568, "step": 3675 }, { "epoch": 0.8591043206450294, "grad_norm": 3.9822843074798584, "learning_rate": 2.998678468884112e-05, "loss": 2.0052, "step": 3676 }, { "epoch": 0.8593380269346499, "grad_norm": 2.6335155963897705, "learning_rate": 2.998676084876149e-05, "loss": 2.0291, "step": 3677 }, { "epoch": 0.8595717332242704, "grad_norm": 4.82760763168335, "learning_rate": 2.998673698720728e-05, "loss": 2.1693, "step": 3678 }, { "epoch": 0.8598054395138909, "grad_norm": 5.376944541931152, "learning_rate": 2.998671310417853e-05, "loss": 2.0643, "step": 3679 }, { "epoch": 0.8600391458035115, "grad_norm": 3.580988645553589, "learning_rate": 2.9986689199675277e-05, "loss": 2.0386, "step": 3680 }, { "epoch": 0.860272852093132, "grad_norm": 2.635108709335327, "learning_rate": 2.9986665273697548e-05, "loss": 1.9857, "step": 3681 }, { "epoch": 0.8605065583827525, "grad_norm": 2.852592945098877, "learning_rate": 2.998664132624538e-05, "loss": 1.9326, "step": 3682 }, { "epoch": 0.860740264672373, "grad_norm": 3.3275933265686035, "learning_rate": 2.9986617357318804e-05, "loss": 2.0972, "step": 3683 }, { "epoch": 0.8609739709619935, "grad_norm": 2.5379960536956787, "learning_rate": 2.9986593366917864e-05, "loss": 1.9038, "step": 3684 }, { "epoch": 0.8612076772516141, "grad_norm": 2.484492540359497, "learning_rate": 2.998656935504258e-05, "loss": 1.8413, "step": 3685 }, { "epoch": 0.8614413835412346, "grad_norm": 3.4459049701690674, "learning_rate": 2.9986545321693005e-05, "loss": 2.003, "step": 3686 }, { "epoch": 0.8616750898308551, "grad_norm": 3.1465063095092773, "learning_rate": 2.9986521266869154e-05, "loss": 2.0926, "step": 3687 }, { "epoch": 0.8619087961204756, "grad_norm": 3.459064245223999, "learning_rate": 2.9986497190571077e-05, "loss": 2.0896, "step": 3688 }, { "epoch": 0.8621425024100962, "grad_norm": 2.946610450744629, "learning_rate": 2.99864730927988e-05, "loss": 2.0087, "step": 3689 }, { "epoch": 0.8623762086997167, "grad_norm": 2.441649913787842, "learning_rate": 2.9986448973552354e-05, "loss": 2.0621, "step": 3690 }, { "epoch": 0.8626099149893371, "grad_norm": 2.8407628536224365, "learning_rate": 2.9986424832831783e-05, "loss": 1.9096, "step": 3691 }, { "epoch": 0.8628436212789576, "grad_norm": 2.8425283432006836, "learning_rate": 2.9986400670637118e-05, "loss": 1.9653, "step": 3692 }, { "epoch": 0.8630773275685781, "grad_norm": 3.4648232460021973, "learning_rate": 2.9986376486968392e-05, "loss": 2.1678, "step": 3693 }, { "epoch": 0.8633110338581987, "grad_norm": 3.3163938522338867, "learning_rate": 2.9986352281825643e-05, "loss": 1.9646, "step": 3694 }, { "epoch": 0.8635447401478192, "grad_norm": 3.007972240447998, "learning_rate": 2.9986328055208904e-05, "loss": 2.0267, "step": 3695 }, { "epoch": 0.8637784464374397, "grad_norm": 2.6560964584350586, "learning_rate": 2.9986303807118206e-05, "loss": 1.9066, "step": 3696 }, { "epoch": 0.8640121527270602, "grad_norm": 5.332592487335205, "learning_rate": 2.998627953755359e-05, "loss": 2.0898, "step": 3697 }, { "epoch": 0.8642458590166808, "grad_norm": 4.517512321472168, "learning_rate": 2.9986255246515087e-05, "loss": 1.8854, "step": 3698 }, { "epoch": 0.8644795653063013, "grad_norm": 2.750135898590088, "learning_rate": 2.998623093400273e-05, "loss": 2.0774, "step": 3699 }, { "epoch": 0.8647132715959218, "grad_norm": 3.3750243186950684, "learning_rate": 2.998620660001656e-05, "loss": 1.9685, "step": 3700 }, { "epoch": 0.8649469778855423, "grad_norm": 3.0901849269866943, "learning_rate": 2.9986182244556608e-05, "loss": 2.0832, "step": 3701 }, { "epoch": 0.8651806841751629, "grad_norm": 3.2934210300445557, "learning_rate": 2.9986157867622905e-05, "loss": 1.8242, "step": 3702 }, { "epoch": 0.8654143904647834, "grad_norm": 4.056787490844727, "learning_rate": 2.9986133469215495e-05, "loss": 1.9454, "step": 3703 }, { "epoch": 0.8656480967544039, "grad_norm": 3.8067333698272705, "learning_rate": 2.9986109049334406e-05, "loss": 1.9687, "step": 3704 }, { "epoch": 0.8658818030440244, "grad_norm": 2.6783359050750732, "learning_rate": 2.9986084607979677e-05, "loss": 1.9128, "step": 3705 }, { "epoch": 0.8661155093336449, "grad_norm": 4.7297186851501465, "learning_rate": 2.9986060145151338e-05, "loss": 1.8424, "step": 3706 }, { "epoch": 0.8663492156232655, "grad_norm": 3.6706454753875732, "learning_rate": 2.998603566084943e-05, "loss": 1.966, "step": 3707 }, { "epoch": 0.866582921912886, "grad_norm": 2.5593786239624023, "learning_rate": 2.9986011155073983e-05, "loss": 2.0584, "step": 3708 }, { "epoch": 0.8668166282025065, "grad_norm": 2.9069406986236572, "learning_rate": 2.9985986627825034e-05, "loss": 1.9922, "step": 3709 }, { "epoch": 0.867050334492127, "grad_norm": 4.109007835388184, "learning_rate": 2.998596207910262e-05, "loss": 2.1179, "step": 3710 }, { "epoch": 0.8672840407817476, "grad_norm": 2.6978800296783447, "learning_rate": 2.9985937508906775e-05, "loss": 1.9874, "step": 3711 }, { "epoch": 0.8675177470713681, "grad_norm": 2.900658130645752, "learning_rate": 2.9985912917237535e-05, "loss": 1.963, "step": 3712 }, { "epoch": 0.8677514533609886, "grad_norm": 2.96570086479187, "learning_rate": 2.998588830409493e-05, "loss": 2.1052, "step": 3713 }, { "epoch": 0.8679851596506091, "grad_norm": 2.615204334259033, "learning_rate": 2.9985863669479004e-05, "loss": 1.9446, "step": 3714 }, { "epoch": 0.8682188659402296, "grad_norm": 5.402909278869629, "learning_rate": 2.9985839013389785e-05, "loss": 2.1043, "step": 3715 }, { "epoch": 0.8684525722298502, "grad_norm": 5.530276298522949, "learning_rate": 2.998581433582731e-05, "loss": 2.0604, "step": 3716 }, { "epoch": 0.8686862785194707, "grad_norm": 2.781984329223633, "learning_rate": 2.9985789636791618e-05, "loss": 2.0419, "step": 3717 }, { "epoch": 0.8689199848090912, "grad_norm": 3.640930414199829, "learning_rate": 2.9985764916282744e-05, "loss": 2.0281, "step": 3718 }, { "epoch": 0.8691536910987117, "grad_norm": 2.960834503173828, "learning_rate": 2.9985740174300715e-05, "loss": 1.9112, "step": 3719 }, { "epoch": 0.8693873973883323, "grad_norm": 2.6281909942626953, "learning_rate": 2.9985715410845573e-05, "loss": 2.042, "step": 3720 }, { "epoch": 0.8696211036779528, "grad_norm": 2.7327661514282227, "learning_rate": 2.9985690625917357e-05, "loss": 2.0357, "step": 3721 }, { "epoch": 0.8698548099675733, "grad_norm": 2.5361690521240234, "learning_rate": 2.9985665819516097e-05, "loss": 2.0209, "step": 3722 }, { "epoch": 0.8700885162571937, "grad_norm": 2.4407918453216553, "learning_rate": 2.9985640991641833e-05, "loss": 1.9561, "step": 3723 }, { "epoch": 0.8703222225468142, "grad_norm": 3.0160632133483887, "learning_rate": 2.9985616142294593e-05, "loss": 2.0487, "step": 3724 }, { "epoch": 0.8705559288364348, "grad_norm": 2.5621604919433594, "learning_rate": 2.998559127147442e-05, "loss": 2.1172, "step": 3725 }, { "epoch": 0.8707896351260553, "grad_norm": 2.7467403411865234, "learning_rate": 2.998556637918135e-05, "loss": 2.138, "step": 3726 }, { "epoch": 0.8710233414156758, "grad_norm": 2.5390067100524902, "learning_rate": 2.998554146541541e-05, "loss": 1.9245, "step": 3727 }, { "epoch": 0.8712570477052963, "grad_norm": 2.498828887939453, "learning_rate": 2.9985516530176648e-05, "loss": 1.9548, "step": 3728 }, { "epoch": 0.8714907539949169, "grad_norm": 2.982391595840454, "learning_rate": 2.9985491573465085e-05, "loss": 1.9511, "step": 3729 }, { "epoch": 0.8717244602845374, "grad_norm": 2.590423107147217, "learning_rate": 2.998546659528077e-05, "loss": 1.9568, "step": 3730 }, { "epoch": 0.8719581665741579, "grad_norm": 2.5173888206481934, "learning_rate": 2.998544159562373e-05, "loss": 2.1082, "step": 3731 }, { "epoch": 0.8721918728637784, "grad_norm": 2.556691884994507, "learning_rate": 2.9985416574494008e-05, "loss": 1.964, "step": 3732 }, { "epoch": 0.8724255791533989, "grad_norm": 2.376100540161133, "learning_rate": 2.9985391531891637e-05, "loss": 2.1091, "step": 3733 }, { "epoch": 0.8726592854430195, "grad_norm": 2.793853759765625, "learning_rate": 2.998536646781665e-05, "loss": 2.0319, "step": 3734 }, { "epoch": 0.87289299173264, "grad_norm": 2.7417478561401367, "learning_rate": 2.9985341382269084e-05, "loss": 2.0114, "step": 3735 }, { "epoch": 0.8731266980222605, "grad_norm": 2.7300941944122314, "learning_rate": 2.998531627524898e-05, "loss": 1.9418, "step": 3736 }, { "epoch": 0.873360404311881, "grad_norm": 2.708343029022217, "learning_rate": 2.9985291146756367e-05, "loss": 2.0661, "step": 3737 }, { "epoch": 0.8735941106015016, "grad_norm": 2.7827095985412598, "learning_rate": 2.9985265996791283e-05, "loss": 1.9522, "step": 3738 }, { "epoch": 0.8738278168911221, "grad_norm": 2.4923434257507324, "learning_rate": 2.9985240825353767e-05, "loss": 2.0129, "step": 3739 }, { "epoch": 0.8740615231807426, "grad_norm": 2.5919766426086426, "learning_rate": 2.9985215632443855e-05, "loss": 2.0696, "step": 3740 }, { "epoch": 0.8742952294703631, "grad_norm": 2.9827773571014404, "learning_rate": 2.998519041806158e-05, "loss": 1.9657, "step": 3741 }, { "epoch": 0.8745289357599837, "grad_norm": 3.388490676879883, "learning_rate": 2.9985165182206975e-05, "loss": 2.1846, "step": 3742 }, { "epoch": 0.8747626420496042, "grad_norm": 3.0380911827087402, "learning_rate": 2.9985139924880085e-05, "loss": 1.8589, "step": 3743 }, { "epoch": 0.8749963483392247, "grad_norm": 2.785538911819458, "learning_rate": 2.9985114646080944e-05, "loss": 2.0306, "step": 3744 }, { "epoch": 0.8752300546288452, "grad_norm": 3.6978936195373535, "learning_rate": 2.998508934580958e-05, "loss": 2.1383, "step": 3745 }, { "epoch": 0.8754637609184657, "grad_norm": 2.7753567695617676, "learning_rate": 2.998506402406604e-05, "loss": 1.9415, "step": 3746 }, { "epoch": 0.8756974672080863, "grad_norm": 2.6149380207061768, "learning_rate": 2.998503868085035e-05, "loss": 2.1503, "step": 3747 }, { "epoch": 0.8759311734977068, "grad_norm": 3.0545597076416016, "learning_rate": 2.9985013316162563e-05, "loss": 1.9543, "step": 3748 }, { "epoch": 0.8761648797873273, "grad_norm": 2.6237778663635254, "learning_rate": 2.9984987930002694e-05, "loss": 2.0042, "step": 3749 }, { "epoch": 0.8763985860769478, "grad_norm": 2.687669515609741, "learning_rate": 2.9984962522370793e-05, "loss": 1.9853, "step": 3750 }, { "epoch": 0.8766322923665684, "grad_norm": 3.0753557682037354, "learning_rate": 2.9984937093266892e-05, "loss": 2.0804, "step": 3751 }, { "epoch": 0.8768659986561889, "grad_norm": 2.519646406173706, "learning_rate": 2.9984911642691028e-05, "loss": 2.0576, "step": 3752 }, { "epoch": 0.8770997049458094, "grad_norm": 2.789714813232422, "learning_rate": 2.9984886170643237e-05, "loss": 2.0166, "step": 3753 }, { "epoch": 0.8773334112354298, "grad_norm": 2.9217634201049805, "learning_rate": 2.998486067712356e-05, "loss": 1.9034, "step": 3754 }, { "epoch": 0.8775671175250503, "grad_norm": 3.360980272293091, "learning_rate": 2.9984835162132027e-05, "loss": 1.9578, "step": 3755 }, { "epoch": 0.877800823814671, "grad_norm": 2.623255491256714, "learning_rate": 2.9984809625668675e-05, "loss": 2.0251, "step": 3756 }, { "epoch": 0.8780345301042914, "grad_norm": 2.6379053592681885, "learning_rate": 2.998478406773355e-05, "loss": 2.0789, "step": 3757 }, { "epoch": 0.8782682363939119, "grad_norm": 3.3912525177001953, "learning_rate": 2.9984758488326674e-05, "loss": 1.8786, "step": 3758 }, { "epoch": 0.8785019426835324, "grad_norm": 2.630387544631958, "learning_rate": 2.99847328874481e-05, "loss": 1.9203, "step": 3759 }, { "epoch": 0.878735648973153, "grad_norm": 3.6282222270965576, "learning_rate": 2.998470726509785e-05, "loss": 1.933, "step": 3760 }, { "epoch": 0.8789693552627735, "grad_norm": 4.466384410858154, "learning_rate": 2.998468162127597e-05, "loss": 1.9205, "step": 3761 }, { "epoch": 0.879203061552394, "grad_norm": 2.8582704067230225, "learning_rate": 2.9984655955982487e-05, "loss": 1.9088, "step": 3762 }, { "epoch": 0.8794367678420145, "grad_norm": 4.402927875518799, "learning_rate": 2.998463026921745e-05, "loss": 2.0578, "step": 3763 }, { "epoch": 0.879670474131635, "grad_norm": 4.546339988708496, "learning_rate": 2.998460456098089e-05, "loss": 1.9612, "step": 3764 }, { "epoch": 0.8799041804212556, "grad_norm": 2.7503647804260254, "learning_rate": 2.9984578831272843e-05, "loss": 1.9498, "step": 3765 }, { "epoch": 0.8801378867108761, "grad_norm": 3.024109125137329, "learning_rate": 2.9984553080093347e-05, "loss": 2.0137, "step": 3766 }, { "epoch": 0.8803715930004966, "grad_norm": 3.450015068054199, "learning_rate": 2.9984527307442434e-05, "loss": 2.0563, "step": 3767 }, { "epoch": 0.8806052992901171, "grad_norm": 2.929361581802368, "learning_rate": 2.998450151332015e-05, "loss": 2.0684, "step": 3768 }, { "epoch": 0.8808390055797377, "grad_norm": 3.4177963733673096, "learning_rate": 2.9984475697726526e-05, "loss": 1.9375, "step": 3769 }, { "epoch": 0.8810727118693582, "grad_norm": 3.1939029693603516, "learning_rate": 2.9984449860661603e-05, "loss": 1.9594, "step": 3770 }, { "epoch": 0.8813064181589787, "grad_norm": 2.287409782409668, "learning_rate": 2.9984424002125412e-05, "loss": 1.8983, "step": 3771 }, { "epoch": 0.8815401244485992, "grad_norm": 2.3156046867370605, "learning_rate": 2.9984398122118e-05, "loss": 1.9365, "step": 3772 }, { "epoch": 0.8817738307382197, "grad_norm": 2.373547077178955, "learning_rate": 2.9984372220639395e-05, "loss": 2.2268, "step": 3773 }, { "epoch": 0.8820075370278403, "grad_norm": 3.621675968170166, "learning_rate": 2.9984346297689634e-05, "loss": 1.9935, "step": 3774 }, { "epoch": 0.8822412433174608, "grad_norm": 2.7700579166412354, "learning_rate": 2.9984320353268753e-05, "loss": 2.0689, "step": 3775 }, { "epoch": 0.8824749496070813, "grad_norm": 2.586986541748047, "learning_rate": 2.9984294387376797e-05, "loss": 1.9082, "step": 3776 }, { "epoch": 0.8827086558967018, "grad_norm": 2.873481512069702, "learning_rate": 2.9984268400013802e-05, "loss": 2.075, "step": 3777 }, { "epoch": 0.8829423621863224, "grad_norm": 2.4888551235198975, "learning_rate": 2.9984242391179803e-05, "loss": 2.0713, "step": 3778 }, { "epoch": 0.8831760684759429, "grad_norm": 2.4298346042633057, "learning_rate": 2.9984216360874837e-05, "loss": 2.1433, "step": 3779 }, { "epoch": 0.8834097747655634, "grad_norm": 2.3473777770996094, "learning_rate": 2.9984190309098937e-05, "loss": 1.8744, "step": 3780 }, { "epoch": 0.8836434810551839, "grad_norm": 2.4035391807556152, "learning_rate": 2.9984164235852148e-05, "loss": 2.1153, "step": 3781 }, { "epoch": 0.8838771873448045, "grad_norm": 3.847459316253662, "learning_rate": 2.99841381411345e-05, "loss": 2.1065, "step": 3782 }, { "epoch": 0.884110893634425, "grad_norm": 3.851841449737549, "learning_rate": 2.9984112024946034e-05, "loss": 1.949, "step": 3783 }, { "epoch": 0.8843445999240455, "grad_norm": 2.670600175857544, "learning_rate": 2.998408588728679e-05, "loss": 2.0124, "step": 3784 }, { "epoch": 0.884578306213666, "grad_norm": 3.413591146469116, "learning_rate": 2.9984059728156807e-05, "loss": 1.9563, "step": 3785 }, { "epoch": 0.8848120125032864, "grad_norm": 2.801719903945923, "learning_rate": 2.9984033547556117e-05, "loss": 1.9626, "step": 3786 }, { "epoch": 0.885045718792907, "grad_norm": 2.480353832244873, "learning_rate": 2.9984007345484755e-05, "loss": 1.8781, "step": 3787 }, { "epoch": 0.8852794250825275, "grad_norm": 3.057408332824707, "learning_rate": 2.9983981121942764e-05, "loss": 2.0466, "step": 3788 }, { "epoch": 0.885513131372148, "grad_norm": 3.5773725509643555, "learning_rate": 2.9983954876930182e-05, "loss": 2.1126, "step": 3789 }, { "epoch": 0.8857468376617685, "grad_norm": 3.2768685817718506, "learning_rate": 2.998392861044704e-05, "loss": 2.1342, "step": 3790 }, { "epoch": 0.8859805439513891, "grad_norm": 3.0999162197113037, "learning_rate": 2.9983902322493386e-05, "loss": 2.1102, "step": 3791 }, { "epoch": 0.8862142502410096, "grad_norm": 3.9843461513519287, "learning_rate": 2.998387601306925e-05, "loss": 2.1062, "step": 3792 }, { "epoch": 0.8864479565306301, "grad_norm": 2.392793655395508, "learning_rate": 2.9983849682174674e-05, "loss": 2.3126, "step": 3793 }, { "epoch": 0.8866816628202506, "grad_norm": 2.4408864974975586, "learning_rate": 2.998382332980969e-05, "loss": 1.9276, "step": 3794 }, { "epoch": 0.8869153691098711, "grad_norm": 2.598998546600342, "learning_rate": 2.998379695597434e-05, "loss": 1.9561, "step": 3795 }, { "epoch": 0.8871490753994917, "grad_norm": 2.3942625522613525, "learning_rate": 2.9983770560668662e-05, "loss": 1.9607, "step": 3796 }, { "epoch": 0.8873827816891122, "grad_norm": 2.8528151512145996, "learning_rate": 2.9983744143892696e-05, "loss": 1.9693, "step": 3797 }, { "epoch": 0.8876164879787327, "grad_norm": 2.6715714931488037, "learning_rate": 2.998371770564647e-05, "loss": 2.0298, "step": 3798 }, { "epoch": 0.8878501942683532, "grad_norm": 3.0273449420928955, "learning_rate": 2.998369124593003e-05, "loss": 1.964, "step": 3799 }, { "epoch": 0.8880839005579738, "grad_norm": 4.268172740936279, "learning_rate": 2.9983664764743416e-05, "loss": 1.9561, "step": 3800 }, { "epoch": 0.8883176068475943, "grad_norm": 2.6332876682281494, "learning_rate": 2.9983638262086664e-05, "loss": 2.1063, "step": 3801 }, { "epoch": 0.8885513131372148, "grad_norm": 2.8699429035186768, "learning_rate": 2.9983611737959805e-05, "loss": 1.9253, "step": 3802 }, { "epoch": 0.8887850194268353, "grad_norm": 4.498633861541748, "learning_rate": 2.9983585192362888e-05, "loss": 2.1152, "step": 3803 }, { "epoch": 0.8890187257164558, "grad_norm": 2.708646059036255, "learning_rate": 2.998355862529594e-05, "loss": 2.0107, "step": 3804 }, { "epoch": 0.8892524320060764, "grad_norm": 3.3724513053894043, "learning_rate": 2.9983532036759007e-05, "loss": 2.1588, "step": 3805 }, { "epoch": 0.8894861382956969, "grad_norm": 3.026920795440674, "learning_rate": 2.9983505426752128e-05, "loss": 2.0272, "step": 3806 }, { "epoch": 0.8897198445853174, "grad_norm": 3.0057520866394043, "learning_rate": 2.9983478795275333e-05, "loss": 1.9584, "step": 3807 }, { "epoch": 0.8899535508749379, "grad_norm": 2.833111524581909, "learning_rate": 2.9983452142328668e-05, "loss": 1.8445, "step": 3808 }, { "epoch": 0.8901872571645585, "grad_norm": 2.610947847366333, "learning_rate": 2.9983425467912164e-05, "loss": 2.0967, "step": 3809 }, { "epoch": 0.890420963454179, "grad_norm": 2.7222635746002197, "learning_rate": 2.998339877202587e-05, "loss": 1.9151, "step": 3810 }, { "epoch": 0.8906546697437995, "grad_norm": 4.27071475982666, "learning_rate": 2.9983372054669812e-05, "loss": 1.885, "step": 3811 }, { "epoch": 0.89088837603342, "grad_norm": 2.7722415924072266, "learning_rate": 2.9983345315844035e-05, "loss": 1.9563, "step": 3812 }, { "epoch": 0.8911220823230405, "grad_norm": 2.6755151748657227, "learning_rate": 2.998331855554858e-05, "loss": 1.9726, "step": 3813 }, { "epoch": 0.8913557886126611, "grad_norm": 4.254831790924072, "learning_rate": 2.9983291773783476e-05, "loss": 2.2068, "step": 3814 }, { "epoch": 0.8915894949022816, "grad_norm": 2.739992380142212, "learning_rate": 2.9983264970548772e-05, "loss": 1.9931, "step": 3815 }, { "epoch": 0.8918232011919021, "grad_norm": 3.103700637817383, "learning_rate": 2.99832381458445e-05, "loss": 1.9275, "step": 3816 }, { "epoch": 0.8920569074815226, "grad_norm": 2.754610300064087, "learning_rate": 2.9983211299670697e-05, "loss": 1.9397, "step": 3817 }, { "epoch": 0.8922906137711432, "grad_norm": 2.4768409729003906, "learning_rate": 2.9983184432027407e-05, "loss": 2.0551, "step": 3818 }, { "epoch": 0.8925243200607637, "grad_norm": 2.550839424133301, "learning_rate": 2.998315754291466e-05, "loss": 1.9622, "step": 3819 }, { "epoch": 0.8927580263503841, "grad_norm": 2.9336912631988525, "learning_rate": 2.998313063233251e-05, "loss": 2.0321, "step": 3820 }, { "epoch": 0.8929917326400046, "grad_norm": 2.6492502689361572, "learning_rate": 2.998310370028098e-05, "loss": 2.2182, "step": 3821 }, { "epoch": 0.8932254389296252, "grad_norm": 2.6279711723327637, "learning_rate": 2.9983076746760115e-05, "loss": 1.9926, "step": 3822 }, { "epoch": 0.8934591452192457, "grad_norm": 2.36104416847229, "learning_rate": 2.9983049771769955e-05, "loss": 2.0158, "step": 3823 }, { "epoch": 0.8936928515088662, "grad_norm": 2.8456571102142334, "learning_rate": 2.998302277531054e-05, "loss": 2.0954, "step": 3824 }, { "epoch": 0.8939265577984867, "grad_norm": 2.8550825119018555, "learning_rate": 2.99829957573819e-05, "loss": 1.9647, "step": 3825 }, { "epoch": 0.8941602640881072, "grad_norm": 3.928532838821411, "learning_rate": 2.998296871798408e-05, "loss": 1.9858, "step": 3826 }, { "epoch": 0.8943939703777278, "grad_norm": 2.762599229812622, "learning_rate": 2.9982941657117118e-05, "loss": 1.8859, "step": 3827 }, { "epoch": 0.8946276766673483, "grad_norm": 3.3166935443878174, "learning_rate": 2.998291457478105e-05, "loss": 1.9464, "step": 3828 }, { "epoch": 0.8948613829569688, "grad_norm": 3.3591408729553223, "learning_rate": 2.998288747097592e-05, "loss": 2.0041, "step": 3829 }, { "epoch": 0.8950950892465893, "grad_norm": 2.7237088680267334, "learning_rate": 2.9982860345701764e-05, "loss": 1.8695, "step": 3830 }, { "epoch": 0.8953287955362099, "grad_norm": 2.499424695968628, "learning_rate": 2.9982833198958624e-05, "loss": 1.8912, "step": 3831 }, { "epoch": 0.8955625018258304, "grad_norm": 3.3871138095855713, "learning_rate": 2.998280603074653e-05, "loss": 2.1808, "step": 3832 }, { "epoch": 0.8957962081154509, "grad_norm": 2.3571577072143555, "learning_rate": 2.9982778841065533e-05, "loss": 1.9048, "step": 3833 }, { "epoch": 0.8960299144050714, "grad_norm": 2.990544557571411, "learning_rate": 2.9982751629915665e-05, "loss": 2.0305, "step": 3834 }, { "epoch": 0.8962636206946919, "grad_norm": 2.7507107257843018, "learning_rate": 2.9982724397296964e-05, "loss": 2.0796, "step": 3835 }, { "epoch": 0.8964973269843125, "grad_norm": 3.6376144886016846, "learning_rate": 2.9982697143209468e-05, "loss": 2.0523, "step": 3836 }, { "epoch": 0.896731033273933, "grad_norm": 3.8768093585968018, "learning_rate": 2.9982669867653224e-05, "loss": 1.9498, "step": 3837 }, { "epoch": 0.8969647395635535, "grad_norm": 3.2789089679718018, "learning_rate": 2.9982642570628266e-05, "loss": 1.8512, "step": 3838 }, { "epoch": 0.897198445853174, "grad_norm": 3.8770899772644043, "learning_rate": 2.998261525213463e-05, "loss": 1.9561, "step": 3839 }, { "epoch": 0.8974321521427946, "grad_norm": 4.491934776306152, "learning_rate": 2.998258791217236e-05, "loss": 1.961, "step": 3840 }, { "epoch": 0.8976658584324151, "grad_norm": 5.407653331756592, "learning_rate": 2.9982560550741494e-05, "loss": 1.928, "step": 3841 }, { "epoch": 0.8978995647220356, "grad_norm": 2.7559974193573, "learning_rate": 2.9982533167842072e-05, "loss": 1.9718, "step": 3842 }, { "epoch": 0.8981332710116561, "grad_norm": 4.593653678894043, "learning_rate": 2.998250576347413e-05, "loss": 2.0769, "step": 3843 }, { "epoch": 0.8983669773012766, "grad_norm": 4.146289348602295, "learning_rate": 2.998247833763771e-05, "loss": 2.0553, "step": 3844 }, { "epoch": 0.8986006835908972, "grad_norm": 2.3645989894866943, "learning_rate": 2.998245089033285e-05, "loss": 1.9502, "step": 3845 }, { "epoch": 0.8988343898805177, "grad_norm": 2.998253107070923, "learning_rate": 2.998242342155959e-05, "loss": 1.9233, "step": 3846 }, { "epoch": 0.8990680961701382, "grad_norm": 3.7990550994873047, "learning_rate": 2.9982395931317966e-05, "loss": 2.0349, "step": 3847 }, { "epoch": 0.8993018024597587, "grad_norm": 2.952136993408203, "learning_rate": 2.998236841960803e-05, "loss": 1.948, "step": 3848 }, { "epoch": 0.8995355087493793, "grad_norm": 2.7440083026885986, "learning_rate": 2.9982340886429804e-05, "loss": 2.0285, "step": 3849 }, { "epoch": 0.8997692150389998, "grad_norm": 4.198298931121826, "learning_rate": 2.9982313331783335e-05, "loss": 2.0249, "step": 3850 }, { "epoch": 0.9000029213286203, "grad_norm": 2.528331756591797, "learning_rate": 2.9982285755668668e-05, "loss": 2.0683, "step": 3851 }, { "epoch": 0.9002366276182407, "grad_norm": 3.080138921737671, "learning_rate": 2.9982258158085836e-05, "loss": 2.084, "step": 3852 }, { "epoch": 0.9004703339078612, "grad_norm": 3.0648040771484375, "learning_rate": 2.998223053903488e-05, "loss": 2.2285, "step": 3853 }, { "epoch": 0.9007040401974818, "grad_norm": 4.669230937957764, "learning_rate": 2.9982202898515838e-05, "loss": 1.9491, "step": 3854 }, { "epoch": 0.9009377464871023, "grad_norm": 3.2519829273223877, "learning_rate": 2.998217523652875e-05, "loss": 2.071, "step": 3855 }, { "epoch": 0.9011714527767228, "grad_norm": 2.7252776622772217, "learning_rate": 2.998214755307366e-05, "loss": 2.0945, "step": 3856 }, { "epoch": 0.9014051590663433, "grad_norm": 2.3827877044677734, "learning_rate": 2.99821198481506e-05, "loss": 2.0012, "step": 3857 }, { "epoch": 0.9016388653559639, "grad_norm": 3.5458076000213623, "learning_rate": 2.998209212175962e-05, "loss": 2.0213, "step": 3858 }, { "epoch": 0.9018725716455844, "grad_norm": 2.562760829925537, "learning_rate": 2.9982064373900753e-05, "loss": 1.9964, "step": 3859 }, { "epoch": 0.9021062779352049, "grad_norm": 4.149973392486572, "learning_rate": 2.9982036604574036e-05, "loss": 1.953, "step": 3860 }, { "epoch": 0.9023399842248254, "grad_norm": 2.4166455268859863, "learning_rate": 2.9982008813779515e-05, "loss": 1.9459, "step": 3861 }, { "epoch": 0.902573690514446, "grad_norm": 3.050300359725952, "learning_rate": 2.998198100151723e-05, "loss": 2.0537, "step": 3862 }, { "epoch": 0.9028073968040665, "grad_norm": 3.1289072036743164, "learning_rate": 2.998195316778721e-05, "loss": 2.0663, "step": 3863 }, { "epoch": 0.903041103093687, "grad_norm": 2.881776809692383, "learning_rate": 2.998192531258951e-05, "loss": 1.8794, "step": 3864 }, { "epoch": 0.9032748093833075, "grad_norm": 2.4973108768463135, "learning_rate": 2.9981897435924165e-05, "loss": 1.9305, "step": 3865 }, { "epoch": 0.903508515672928, "grad_norm": 3.629563570022583, "learning_rate": 2.9981869537791207e-05, "loss": 2.099, "step": 3866 }, { "epoch": 0.9037422219625486, "grad_norm": 2.864359140396118, "learning_rate": 2.9981841618190683e-05, "loss": 2.0569, "step": 3867 }, { "epoch": 0.9039759282521691, "grad_norm": 2.8346633911132812, "learning_rate": 2.998181367712263e-05, "loss": 1.9539, "step": 3868 }, { "epoch": 0.9042096345417896, "grad_norm": 2.450988292694092, "learning_rate": 2.9981785714587095e-05, "loss": 1.9655, "step": 3869 }, { "epoch": 0.9044433408314101, "grad_norm": 3.7050468921661377, "learning_rate": 2.9981757730584115e-05, "loss": 2.2224, "step": 3870 }, { "epoch": 0.9046770471210307, "grad_norm": 2.795464038848877, "learning_rate": 2.9981729725113722e-05, "loss": 2.0172, "step": 3871 }, { "epoch": 0.9049107534106512, "grad_norm": 2.456627607345581, "learning_rate": 2.998170169817596e-05, "loss": 2.239, "step": 3872 }, { "epoch": 0.9051444597002717, "grad_norm": 3.707822561264038, "learning_rate": 2.9981673649770877e-05, "loss": 1.8785, "step": 3873 }, { "epoch": 0.9053781659898922, "grad_norm": 2.4926719665527344, "learning_rate": 2.998164557989851e-05, "loss": 1.9185, "step": 3874 }, { "epoch": 0.9056118722795127, "grad_norm": 2.5382001399993896, "learning_rate": 2.998161748855889e-05, "loss": 1.9503, "step": 3875 }, { "epoch": 0.9058455785691333, "grad_norm": 2.4987030029296875, "learning_rate": 2.998158937575207e-05, "loss": 1.8428, "step": 3876 }, { "epoch": 0.9060792848587538, "grad_norm": 2.54748272895813, "learning_rate": 2.9981561241478083e-05, "loss": 2.1355, "step": 3877 }, { "epoch": 0.9063129911483743, "grad_norm": 2.3874380588531494, "learning_rate": 2.9981533085736964e-05, "loss": 2.0854, "step": 3878 }, { "epoch": 0.9065466974379948, "grad_norm": 2.42724609375, "learning_rate": 2.9981504908528768e-05, "loss": 1.8968, "step": 3879 }, { "epoch": 0.9067804037276154, "grad_norm": 3.1080591678619385, "learning_rate": 2.998147670985352e-05, "loss": 1.8482, "step": 3880 }, { "epoch": 0.9070141100172359, "grad_norm": 2.901418685913086, "learning_rate": 2.9981448489711274e-05, "loss": 1.9605, "step": 3881 }, { "epoch": 0.9072478163068564, "grad_norm": 2.651219367980957, "learning_rate": 2.9981420248102064e-05, "loss": 2.0832, "step": 3882 }, { "epoch": 0.9074815225964769, "grad_norm": 4.005091190338135, "learning_rate": 2.9981391985025927e-05, "loss": 1.8797, "step": 3883 }, { "epoch": 0.9077152288860973, "grad_norm": 2.8755581378936768, "learning_rate": 2.9981363700482914e-05, "loss": 1.9479, "step": 3884 }, { "epoch": 0.907948935175718, "grad_norm": 3.154618740081787, "learning_rate": 2.998133539447305e-05, "loss": 2.0895, "step": 3885 }, { "epoch": 0.9081826414653384, "grad_norm": 2.752101421356201, "learning_rate": 2.998130706699639e-05, "loss": 2.1769, "step": 3886 }, { "epoch": 0.9084163477549589, "grad_norm": 3.1197731494903564, "learning_rate": 2.9981278718052967e-05, "loss": 2.0766, "step": 3887 }, { "epoch": 0.9086500540445794, "grad_norm": 3.4723222255706787, "learning_rate": 2.9981250347642827e-05, "loss": 1.895, "step": 3888 }, { "epoch": 0.9088837603342, "grad_norm": 2.4811184406280518, "learning_rate": 2.9981221955766e-05, "loss": 1.8247, "step": 3889 }, { "epoch": 0.9091174666238205, "grad_norm": 4.830767631530762, "learning_rate": 2.998119354242254e-05, "loss": 1.9556, "step": 3890 }, { "epoch": 0.909351172913441, "grad_norm": 3.619582176208496, "learning_rate": 2.9981165107612478e-05, "loss": 2.0501, "step": 3891 }, { "epoch": 0.9095848792030615, "grad_norm": 2.725105047225952, "learning_rate": 2.9981136651335856e-05, "loss": 2.1013, "step": 3892 }, { "epoch": 0.909818585492682, "grad_norm": 3.0309033393859863, "learning_rate": 2.9981108173592723e-05, "loss": 1.952, "step": 3893 }, { "epoch": 0.9100522917823026, "grad_norm": 2.7024130821228027, "learning_rate": 2.9981079674383112e-05, "loss": 1.9683, "step": 3894 }, { "epoch": 0.9102859980719231, "grad_norm": 3.463804244995117, "learning_rate": 2.9981051153707067e-05, "loss": 2.0223, "step": 3895 }, { "epoch": 0.9105197043615436, "grad_norm": 2.3949062824249268, "learning_rate": 2.9981022611564625e-05, "loss": 1.8722, "step": 3896 }, { "epoch": 0.9107534106511641, "grad_norm": 5.853272438049316, "learning_rate": 2.9980994047955827e-05, "loss": 2.0064, "step": 3897 }, { "epoch": 0.9109871169407847, "grad_norm": 3.4265849590301514, "learning_rate": 2.9980965462880718e-05, "loss": 2.0192, "step": 3898 }, { "epoch": 0.9112208232304052, "grad_norm": 4.4958038330078125, "learning_rate": 2.9980936856339344e-05, "loss": 2.0392, "step": 3899 }, { "epoch": 0.9114545295200257, "grad_norm": 3.5189504623413086, "learning_rate": 2.9980908228331732e-05, "loss": 1.9657, "step": 3900 }, { "epoch": 0.9116882358096462, "grad_norm": 2.9667954444885254, "learning_rate": 2.998087957885793e-05, "loss": 1.9421, "step": 3901 }, { "epoch": 0.9119219420992667, "grad_norm": 3.9411351680755615, "learning_rate": 2.9980850907917983e-05, "loss": 2.056, "step": 3902 }, { "epoch": 0.9121556483888873, "grad_norm": 2.456244468688965, "learning_rate": 2.9980822215511927e-05, "loss": 2.0123, "step": 3903 }, { "epoch": 0.9123893546785078, "grad_norm": 2.6821961402893066, "learning_rate": 2.9980793501639807e-05, "loss": 2.0639, "step": 3904 }, { "epoch": 0.9126230609681283, "grad_norm": 2.3959121704101562, "learning_rate": 2.998076476630166e-05, "loss": 1.9434, "step": 3905 }, { "epoch": 0.9128567672577488, "grad_norm": 2.860842704772949, "learning_rate": 2.9980736009497527e-05, "loss": 1.9015, "step": 3906 }, { "epoch": 0.9130904735473694, "grad_norm": 2.954437732696533, "learning_rate": 2.9980707231227453e-05, "loss": 1.8903, "step": 3907 }, { "epoch": 0.9133241798369899, "grad_norm": 2.7986106872558594, "learning_rate": 2.9980678431491477e-05, "loss": 1.8345, "step": 3908 }, { "epoch": 0.9135578861266104, "grad_norm": 3.544926404953003, "learning_rate": 2.9980649610289644e-05, "loss": 2.0579, "step": 3909 }, { "epoch": 0.9137915924162309, "grad_norm": 3.5264816284179688, "learning_rate": 2.9980620767621986e-05, "loss": 2.0295, "step": 3910 }, { "epoch": 0.9140252987058515, "grad_norm": 3.0400259494781494, "learning_rate": 2.998059190348855e-05, "loss": 2.0282, "step": 3911 }, { "epoch": 0.914259004995472, "grad_norm": 3.616438388824463, "learning_rate": 2.9980563017889388e-05, "loss": 1.905, "step": 3912 }, { "epoch": 0.9144927112850925, "grad_norm": 3.1530003547668457, "learning_rate": 2.998053411082452e-05, "loss": 2.1666, "step": 3913 }, { "epoch": 0.914726417574713, "grad_norm": 3.655817985534668, "learning_rate": 2.9980505182294e-05, "loss": 1.8634, "step": 3914 }, { "epoch": 0.9149601238643335, "grad_norm": 2.760308027267456, "learning_rate": 2.9980476232297874e-05, "loss": 2.0022, "step": 3915 }, { "epoch": 0.9151938301539541, "grad_norm": 2.3874094486236572, "learning_rate": 2.998044726083617e-05, "loss": 2.0591, "step": 3916 }, { "epoch": 0.9154275364435746, "grad_norm": 2.4253406524658203, "learning_rate": 2.998041826790894e-05, "loss": 1.9645, "step": 3917 }, { "epoch": 0.915661242733195, "grad_norm": 2.495539665222168, "learning_rate": 2.9980389253516224e-05, "loss": 2.073, "step": 3918 }, { "epoch": 0.9158949490228155, "grad_norm": 2.4705185890197754, "learning_rate": 2.998036021765806e-05, "loss": 1.906, "step": 3919 }, { "epoch": 0.9161286553124361, "grad_norm": 2.7264657020568848, "learning_rate": 2.998033116033449e-05, "loss": 1.9483, "step": 3920 }, { "epoch": 0.9163623616020566, "grad_norm": 3.5300464630126953, "learning_rate": 2.998030208154556e-05, "loss": 2.0749, "step": 3921 }, { "epoch": 0.9165960678916771, "grad_norm": 2.7107064723968506, "learning_rate": 2.998027298129131e-05, "loss": 1.9902, "step": 3922 }, { "epoch": 0.9168297741812976, "grad_norm": 3.30118727684021, "learning_rate": 2.9980243859571778e-05, "loss": 1.9628, "step": 3923 }, { "epoch": 0.9170634804709181, "grad_norm": 2.9525654315948486, "learning_rate": 2.9980214716387012e-05, "loss": 2.0496, "step": 3924 }, { "epoch": 0.9172971867605387, "grad_norm": 2.9104790687561035, "learning_rate": 2.9980185551737045e-05, "loss": 2.1194, "step": 3925 }, { "epoch": 0.9175308930501592, "grad_norm": 3.687438488006592, "learning_rate": 2.9980156365621925e-05, "loss": 1.7617, "step": 3926 }, { "epoch": 0.9177645993397797, "grad_norm": 2.784647226333618, "learning_rate": 2.99801271580417e-05, "loss": 1.9464, "step": 3927 }, { "epoch": 0.9179983056294002, "grad_norm": 3.806386709213257, "learning_rate": 2.9980097928996395e-05, "loss": 1.9953, "step": 3928 }, { "epoch": 0.9182320119190208, "grad_norm": 2.8398354053497314, "learning_rate": 2.998006867848607e-05, "loss": 2.0302, "step": 3929 }, { "epoch": 0.9184657182086413, "grad_norm": 3.019193649291992, "learning_rate": 2.998003940651075e-05, "loss": 1.9254, "step": 3930 }, { "epoch": 0.9186994244982618, "grad_norm": 2.721101760864258, "learning_rate": 2.998001011307049e-05, "loss": 2.0622, "step": 3931 }, { "epoch": 0.9189331307878823, "grad_norm": 4.152740478515625, "learning_rate": 2.9979980798165322e-05, "loss": 1.8494, "step": 3932 }, { "epoch": 0.9191668370775028, "grad_norm": 3.265141248703003, "learning_rate": 2.9979951461795302e-05, "loss": 2.1038, "step": 3933 }, { "epoch": 0.9194005433671234, "grad_norm": 2.616511106491089, "learning_rate": 2.9979922103960458e-05, "loss": 1.9627, "step": 3934 }, { "epoch": 0.9196342496567439, "grad_norm": 3.165388822555542, "learning_rate": 2.997989272466084e-05, "loss": 1.9703, "step": 3935 }, { "epoch": 0.9198679559463644, "grad_norm": 3.487091541290283, "learning_rate": 2.997986332389648e-05, "loss": 2.0042, "step": 3936 }, { "epoch": 0.9201016622359849, "grad_norm": 2.771810293197632, "learning_rate": 2.9979833901667434e-05, "loss": 1.9399, "step": 3937 }, { "epoch": 0.9203353685256055, "grad_norm": 4.175945281982422, "learning_rate": 2.9979804457973738e-05, "loss": 2.0637, "step": 3938 }, { "epoch": 0.920569074815226, "grad_norm": 3.947758674621582, "learning_rate": 2.997977499281543e-05, "loss": 2.0264, "step": 3939 }, { "epoch": 0.9208027811048465, "grad_norm": 2.898775339126587, "learning_rate": 2.997974550619256e-05, "loss": 2.0949, "step": 3940 }, { "epoch": 0.921036487394467, "grad_norm": 4.721043109893799, "learning_rate": 2.9979715998105168e-05, "loss": 1.9289, "step": 3941 }, { "epoch": 0.9212701936840875, "grad_norm": 3.4696648120880127, "learning_rate": 2.9979686468553292e-05, "loss": 2.0195, "step": 3942 }, { "epoch": 0.9215038999737081, "grad_norm": 2.4588358402252197, "learning_rate": 2.9979656917536978e-05, "loss": 2.0598, "step": 3943 }, { "epoch": 0.9217376062633286, "grad_norm": 3.492486000061035, "learning_rate": 2.9979627345056266e-05, "loss": 1.9717, "step": 3944 }, { "epoch": 0.9219713125529491, "grad_norm": 4.298289775848389, "learning_rate": 2.9979597751111198e-05, "loss": 1.972, "step": 3945 }, { "epoch": 0.9222050188425696, "grad_norm": 2.92063045501709, "learning_rate": 2.997956813570182e-05, "loss": 2.0828, "step": 3946 }, { "epoch": 0.9224387251321902, "grad_norm": 3.6879844665527344, "learning_rate": 2.9979538498828172e-05, "loss": 2.0503, "step": 3947 }, { "epoch": 0.9226724314218107, "grad_norm": 2.6671504974365234, "learning_rate": 2.99795088404903e-05, "loss": 2.0667, "step": 3948 }, { "epoch": 0.9229061377114312, "grad_norm": 2.353581666946411, "learning_rate": 2.9979479160688244e-05, "loss": 1.9066, "step": 3949 }, { "epoch": 0.9231398440010516, "grad_norm": 2.6144487857818604, "learning_rate": 2.9979449459422045e-05, "loss": 2.1224, "step": 3950 }, { "epoch": 0.9233735502906723, "grad_norm": 3.392246723175049, "learning_rate": 2.997941973669174e-05, "loss": 2.1033, "step": 3951 }, { "epoch": 0.9236072565802927, "grad_norm": 2.6912975311279297, "learning_rate": 2.9979389992497384e-05, "loss": 1.8745, "step": 3952 }, { "epoch": 0.9238409628699132, "grad_norm": 3.4590251445770264, "learning_rate": 2.9979360226839015e-05, "loss": 1.9041, "step": 3953 }, { "epoch": 0.9240746691595337, "grad_norm": 2.4621715545654297, "learning_rate": 2.997933043971667e-05, "loss": 2.1362, "step": 3954 }, { "epoch": 0.9243083754491542, "grad_norm": 3.054464101791382, "learning_rate": 2.99793006311304e-05, "loss": 1.8486, "step": 3955 }, { "epoch": 0.9245420817387748, "grad_norm": 2.8599398136138916, "learning_rate": 2.997927080108024e-05, "loss": 1.9488, "step": 3956 }, { "epoch": 0.9247757880283953, "grad_norm": 2.5337064266204834, "learning_rate": 2.9979240949566243e-05, "loss": 1.9316, "step": 3957 }, { "epoch": 0.9250094943180158, "grad_norm": 3.9438984394073486, "learning_rate": 2.9979211076588444e-05, "loss": 1.9833, "step": 3958 }, { "epoch": 0.9252432006076363, "grad_norm": 2.782378911972046, "learning_rate": 2.997918118214688e-05, "loss": 1.9954, "step": 3959 }, { "epoch": 0.9254769068972569, "grad_norm": 2.9628376960754395, "learning_rate": 2.997915126624161e-05, "loss": 1.9755, "step": 3960 }, { "epoch": 0.9257106131868774, "grad_norm": 2.707446575164795, "learning_rate": 2.9979121328872664e-05, "loss": 1.9983, "step": 3961 }, { "epoch": 0.9259443194764979, "grad_norm": 2.5058956146240234, "learning_rate": 2.997909137004009e-05, "loss": 2.2229, "step": 3962 }, { "epoch": 0.9261780257661184, "grad_norm": 5.065980434417725, "learning_rate": 2.9979061389743927e-05, "loss": 1.9501, "step": 3963 }, { "epoch": 0.9264117320557389, "grad_norm": 3.1028947830200195, "learning_rate": 2.997903138798422e-05, "loss": 1.9662, "step": 3964 }, { "epoch": 0.9266454383453595, "grad_norm": 3.3062331676483154, "learning_rate": 2.9979001364761022e-05, "loss": 1.9498, "step": 3965 }, { "epoch": 0.92687914463498, "grad_norm": 6.078021049499512, "learning_rate": 2.9978971320074357e-05, "loss": 2.0531, "step": 3966 }, { "epoch": 0.9271128509246005, "grad_norm": 3.6231958866119385, "learning_rate": 2.997894125392428e-05, "loss": 1.831, "step": 3967 }, { "epoch": 0.927346557214221, "grad_norm": 2.8997678756713867, "learning_rate": 2.9978911166310832e-05, "loss": 2.1439, "step": 3968 }, { "epoch": 0.9275802635038416, "grad_norm": 4.288653373718262, "learning_rate": 2.9978881057234057e-05, "loss": 2.0287, "step": 3969 }, { "epoch": 0.9278139697934621, "grad_norm": 3.8452298641204834, "learning_rate": 2.9978850926693996e-05, "loss": 1.9191, "step": 3970 }, { "epoch": 0.9280476760830826, "grad_norm": 3.0903778076171875, "learning_rate": 2.9978820774690695e-05, "loss": 2.0763, "step": 3971 }, { "epoch": 0.9282813823727031, "grad_norm": 5.125584602355957, "learning_rate": 2.997879060122419e-05, "loss": 2.0104, "step": 3972 }, { "epoch": 0.9285150886623236, "grad_norm": 4.724825859069824, "learning_rate": 2.9978760406294538e-05, "loss": 2.0851, "step": 3973 }, { "epoch": 0.9287487949519442, "grad_norm": 4.310937404632568, "learning_rate": 2.9978730189901765e-05, "loss": 2.1288, "step": 3974 }, { "epoch": 0.9289825012415647, "grad_norm": 3.039923667907715, "learning_rate": 2.9978699952045933e-05, "loss": 1.94, "step": 3975 }, { "epoch": 0.9292162075311852, "grad_norm": 4.276183605194092, "learning_rate": 2.997866969272707e-05, "loss": 2.0334, "step": 3976 }, { "epoch": 0.9294499138208057, "grad_norm": 3.76261568069458, "learning_rate": 2.997863941194523e-05, "loss": 1.9382, "step": 3977 }, { "epoch": 0.9296836201104263, "grad_norm": 2.427396774291992, "learning_rate": 2.9978609109700442e-05, "loss": 2.0674, "step": 3978 }, { "epoch": 0.9299173264000468, "grad_norm": 2.8285460472106934, "learning_rate": 2.9978578785992762e-05, "loss": 2.0501, "step": 3979 }, { "epoch": 0.9301510326896673, "grad_norm": 2.344780445098877, "learning_rate": 2.9978548440822233e-05, "loss": 2.012, "step": 3980 }, { "epoch": 0.9303847389792878, "grad_norm": 2.5097601413726807, "learning_rate": 2.99785180741889e-05, "loss": 2.0958, "step": 3981 }, { "epoch": 0.9306184452689082, "grad_norm": 2.2550203800201416, "learning_rate": 2.9978487686092793e-05, "loss": 2.0534, "step": 3982 }, { "epoch": 0.9308521515585289, "grad_norm": 3.018582582473755, "learning_rate": 2.9978457276533967e-05, "loss": 1.9585, "step": 3983 }, { "epoch": 0.9310858578481493, "grad_norm": 2.471696615219116, "learning_rate": 2.9978426845512464e-05, "loss": 2.0968, "step": 3984 }, { "epoch": 0.9313195641377698, "grad_norm": 3.155282497406006, "learning_rate": 2.9978396393028332e-05, "loss": 1.965, "step": 3985 }, { "epoch": 0.9315532704273903, "grad_norm": 2.817662239074707, "learning_rate": 2.9978365919081605e-05, "loss": 2.0859, "step": 3986 }, { "epoch": 0.9317869767170109, "grad_norm": 2.3745288848876953, "learning_rate": 2.9978335423672334e-05, "loss": 1.9887, "step": 3987 }, { "epoch": 0.9320206830066314, "grad_norm": 2.739480972290039, "learning_rate": 2.997830490680056e-05, "loss": 1.903, "step": 3988 }, { "epoch": 0.9322543892962519, "grad_norm": 3.2871363162994385, "learning_rate": 2.9978274368466325e-05, "loss": 2.0063, "step": 3989 }, { "epoch": 0.9324880955858724, "grad_norm": 2.6946418285369873, "learning_rate": 2.997824380866967e-05, "loss": 2.0307, "step": 3990 }, { "epoch": 0.932721801875493, "grad_norm": 3.6705708503723145, "learning_rate": 2.9978213227410646e-05, "loss": 2.0089, "step": 3991 }, { "epoch": 0.9329555081651135, "grad_norm": 2.532118082046509, "learning_rate": 2.9978182624689293e-05, "loss": 2.0923, "step": 3992 }, { "epoch": 0.933189214454734, "grad_norm": 2.4994006156921387, "learning_rate": 2.9978152000505657e-05, "loss": 2.1173, "step": 3993 }, { "epoch": 0.9334229207443545, "grad_norm": 2.40881609916687, "learning_rate": 2.9978121354859785e-05, "loss": 1.9359, "step": 3994 }, { "epoch": 0.933656627033975, "grad_norm": 2.3887617588043213, "learning_rate": 2.9978090687751715e-05, "loss": 2.0496, "step": 3995 }, { "epoch": 0.9338903333235956, "grad_norm": 2.794273614883423, "learning_rate": 2.997805999918149e-05, "loss": 2.1302, "step": 3996 }, { "epoch": 0.9341240396132161, "grad_norm": 2.678359270095825, "learning_rate": 2.9978029289149156e-05, "loss": 1.9733, "step": 3997 }, { "epoch": 0.9343577459028366, "grad_norm": 2.919119119644165, "learning_rate": 2.997799855765476e-05, "loss": 1.9789, "step": 3998 }, { "epoch": 0.9345914521924571, "grad_norm": 2.5730392932891846, "learning_rate": 2.9977967804698345e-05, "loss": 1.9746, "step": 3999 }, { "epoch": 0.9348251584820777, "grad_norm": 4.1210126876831055, "learning_rate": 2.997793703027995e-05, "loss": 1.9619, "step": 4000 }, { "epoch": 0.9350588647716982, "grad_norm": 2.6475138664245605, "learning_rate": 2.9977906234399627e-05, "loss": 2.0747, "step": 4001 }, { "epoch": 0.9352925710613187, "grad_norm": 3.6821181774139404, "learning_rate": 2.9977875417057415e-05, "loss": 2.1143, "step": 4002 }, { "epoch": 0.9355262773509392, "grad_norm": 3.9601874351501465, "learning_rate": 2.9977844578253354e-05, "loss": 2.0296, "step": 4003 }, { "epoch": 0.9357599836405597, "grad_norm": 2.372251033782959, "learning_rate": 2.9977813717987496e-05, "loss": 2.048, "step": 4004 }, { "epoch": 0.9359936899301803, "grad_norm": 4.736337184906006, "learning_rate": 2.9977782836259884e-05, "loss": 1.8979, "step": 4005 }, { "epoch": 0.9362273962198008, "grad_norm": 2.891854763031006, "learning_rate": 2.9977751933070558e-05, "loss": 1.9741, "step": 4006 }, { "epoch": 0.9364611025094213, "grad_norm": 2.442065715789795, "learning_rate": 2.9977721008419572e-05, "loss": 2.0286, "step": 4007 }, { "epoch": 0.9366948087990418, "grad_norm": 4.43684196472168, "learning_rate": 2.9977690062306956e-05, "loss": 1.8937, "step": 4008 }, { "epoch": 0.9369285150886624, "grad_norm": 4.801257610321045, "learning_rate": 2.9977659094732765e-05, "loss": 2.0352, "step": 4009 }, { "epoch": 0.9371622213782829, "grad_norm": 3.381510019302368, "learning_rate": 2.9977628105697036e-05, "loss": 1.9749, "step": 4010 }, { "epoch": 0.9373959276679034, "grad_norm": 3.5784459114074707, "learning_rate": 2.9977597095199823e-05, "loss": 2.0432, "step": 4011 }, { "epoch": 0.9376296339575239, "grad_norm": 2.615068197250366, "learning_rate": 2.9977566063241163e-05, "loss": 2.1415, "step": 4012 }, { "epoch": 0.9378633402471444, "grad_norm": 4.730086326599121, "learning_rate": 2.99775350098211e-05, "loss": 1.9832, "step": 4013 }, { "epoch": 0.938097046536765, "grad_norm": 4.890897274017334, "learning_rate": 2.9977503934939684e-05, "loss": 1.9536, "step": 4014 }, { "epoch": 0.9383307528263855, "grad_norm": 3.615889072418213, "learning_rate": 2.9977472838596952e-05, "loss": 1.8639, "step": 4015 }, { "epoch": 0.938564459116006, "grad_norm": 4.251618385314941, "learning_rate": 2.997744172079296e-05, "loss": 1.8948, "step": 4016 }, { "epoch": 0.9387981654056264, "grad_norm": 6.364307880401611, "learning_rate": 2.997741058152774e-05, "loss": 1.9746, "step": 4017 }, { "epoch": 0.939031871695247, "grad_norm": 3.7009806632995605, "learning_rate": 2.9977379420801345e-05, "loss": 1.9911, "step": 4018 }, { "epoch": 0.9392655779848675, "grad_norm": 3.081549882888794, "learning_rate": 2.9977348238613816e-05, "loss": 2.0452, "step": 4019 }, { "epoch": 0.939499284274488, "grad_norm": 2.5002360343933105, "learning_rate": 2.9977317034965196e-05, "loss": 1.9045, "step": 4020 }, { "epoch": 0.9397329905641085, "grad_norm": 2.6571836471557617, "learning_rate": 2.9977285809855534e-05, "loss": 1.9822, "step": 4021 }, { "epoch": 0.939966696853729, "grad_norm": 2.5877790451049805, "learning_rate": 2.9977254563284875e-05, "loss": 2.0033, "step": 4022 }, { "epoch": 0.9402004031433496, "grad_norm": 2.3249332904815674, "learning_rate": 2.9977223295253257e-05, "loss": 1.8502, "step": 4023 }, { "epoch": 0.9404341094329701, "grad_norm": 2.9796173572540283, "learning_rate": 2.9977192005760735e-05, "loss": 2.0224, "step": 4024 }, { "epoch": 0.9406678157225906, "grad_norm": 2.5006978511810303, "learning_rate": 2.9977160694807343e-05, "loss": 2.1413, "step": 4025 }, { "epoch": 0.9409015220122111, "grad_norm": 2.603334903717041, "learning_rate": 2.9977129362393133e-05, "loss": 1.977, "step": 4026 }, { "epoch": 0.9411352283018317, "grad_norm": 2.391407012939453, "learning_rate": 2.9977098008518152e-05, "loss": 2.006, "step": 4027 }, { "epoch": 0.9413689345914522, "grad_norm": 2.529733180999756, "learning_rate": 2.997706663318244e-05, "loss": 1.9487, "step": 4028 }, { "epoch": 0.9416026408810727, "grad_norm": 4.32262659072876, "learning_rate": 2.997703523638604e-05, "loss": 2.1819, "step": 4029 }, { "epoch": 0.9418363471706932, "grad_norm": 3.3366599082946777, "learning_rate": 2.9977003818129e-05, "loss": 1.9062, "step": 4030 }, { "epoch": 0.9420700534603138, "grad_norm": 3.7939300537109375, "learning_rate": 2.9976972378411367e-05, "loss": 1.9267, "step": 4031 }, { "epoch": 0.9423037597499343, "grad_norm": 2.2596435546875, "learning_rate": 2.997694091723318e-05, "loss": 2.0439, "step": 4032 }, { "epoch": 0.9425374660395548, "grad_norm": 2.9911437034606934, "learning_rate": 2.9976909434594495e-05, "loss": 1.9888, "step": 4033 }, { "epoch": 0.9427711723291753, "grad_norm": 2.8331754207611084, "learning_rate": 2.9976877930495347e-05, "loss": 1.8875, "step": 4034 }, { "epoch": 0.9430048786187958, "grad_norm": 2.6040756702423096, "learning_rate": 2.9976846404935783e-05, "loss": 2.0629, "step": 4035 }, { "epoch": 0.9432385849084164, "grad_norm": 4.321868896484375, "learning_rate": 2.9976814857915846e-05, "loss": 1.9004, "step": 4036 }, { "epoch": 0.9434722911980369, "grad_norm": 3.5257115364074707, "learning_rate": 2.997678328943559e-05, "loss": 2.0923, "step": 4037 }, { "epoch": 0.9437059974876574, "grad_norm": 2.8698229789733887, "learning_rate": 2.9976751699495058e-05, "loss": 2.1291, "step": 4038 }, { "epoch": 0.9439397037772779, "grad_norm": 2.3575754165649414, "learning_rate": 2.9976720088094285e-05, "loss": 1.9328, "step": 4039 }, { "epoch": 0.9441734100668985, "grad_norm": 2.52990460395813, "learning_rate": 2.997668845523333e-05, "loss": 1.8568, "step": 4040 }, { "epoch": 0.944407116356519, "grad_norm": 4.288946151733398, "learning_rate": 2.9976656800912227e-05, "loss": 2.1804, "step": 4041 }, { "epoch": 0.9446408226461395, "grad_norm": 2.6112546920776367, "learning_rate": 2.997662512513103e-05, "loss": 2.0129, "step": 4042 }, { "epoch": 0.94487452893576, "grad_norm": 3.230710744857788, "learning_rate": 2.9976593427889777e-05, "loss": 1.997, "step": 4043 }, { "epoch": 0.9451082352253805, "grad_norm": 2.577296733856201, "learning_rate": 2.997656170918852e-05, "loss": 1.972, "step": 4044 }, { "epoch": 0.9453419415150011, "grad_norm": 2.6618049144744873, "learning_rate": 2.99765299690273e-05, "loss": 1.9822, "step": 4045 }, { "epoch": 0.9455756478046216, "grad_norm": 2.686800241470337, "learning_rate": 2.9976498207406162e-05, "loss": 2.0575, "step": 4046 }, { "epoch": 0.945809354094242, "grad_norm": 2.889042377471924, "learning_rate": 2.9976466424325157e-05, "loss": 2.0129, "step": 4047 }, { "epoch": 0.9460430603838625, "grad_norm": 2.540853261947632, "learning_rate": 2.9976434619784324e-05, "loss": 1.8474, "step": 4048 }, { "epoch": 0.9462767666734831, "grad_norm": 2.480734348297119, "learning_rate": 2.9976402793783718e-05, "loss": 2.056, "step": 4049 }, { "epoch": 0.9465104729631036, "grad_norm": 2.9036478996276855, "learning_rate": 2.997637094632337e-05, "loss": 2.1351, "step": 4050 }, { "epoch": 0.9467441792527241, "grad_norm": 3.4695987701416016, "learning_rate": 2.9976339077403337e-05, "loss": 1.9266, "step": 4051 }, { "epoch": 0.9469778855423446, "grad_norm": 3.176738977432251, "learning_rate": 2.997630718702366e-05, "loss": 2.0303, "step": 4052 }, { "epoch": 0.9472115918319651, "grad_norm": 2.6312477588653564, "learning_rate": 2.997627527518439e-05, "loss": 1.9767, "step": 4053 }, { "epoch": 0.9474452981215857, "grad_norm": 4.667984962463379, "learning_rate": 2.9976243341885568e-05, "loss": 1.8986, "step": 4054 }, { "epoch": 0.9476790044112062, "grad_norm": 3.3262012004852295, "learning_rate": 2.997621138712724e-05, "loss": 1.9293, "step": 4055 }, { "epoch": 0.9479127107008267, "grad_norm": 4.086404323577881, "learning_rate": 2.997617941090945e-05, "loss": 1.9476, "step": 4056 }, { "epoch": 0.9481464169904472, "grad_norm": 3.728101968765259, "learning_rate": 2.997614741323225e-05, "loss": 1.9278, "step": 4057 }, { "epoch": 0.9483801232800678, "grad_norm": 2.3951029777526855, "learning_rate": 2.9976115394095682e-05, "loss": 1.924, "step": 4058 }, { "epoch": 0.9486138295696883, "grad_norm": 2.5701639652252197, "learning_rate": 2.997608335349979e-05, "loss": 1.8805, "step": 4059 }, { "epoch": 0.9488475358593088, "grad_norm": 6.440473556518555, "learning_rate": 2.997605129144462e-05, "loss": 2.1412, "step": 4060 }, { "epoch": 0.9490812421489293, "grad_norm": 4.397747993469238, "learning_rate": 2.9976019207930225e-05, "loss": 2.0206, "step": 4061 }, { "epoch": 0.9493149484385498, "grad_norm": 3.9037275314331055, "learning_rate": 2.9975987102956645e-05, "loss": 1.9358, "step": 4062 }, { "epoch": 0.9495486547281704, "grad_norm": 2.7302989959716797, "learning_rate": 2.9975954976523922e-05, "loss": 1.9018, "step": 4063 }, { "epoch": 0.9497823610177909, "grad_norm": 2.6597790718078613, "learning_rate": 2.9975922828632114e-05, "loss": 2.2484, "step": 4064 }, { "epoch": 0.9500160673074114, "grad_norm": 2.9977498054504395, "learning_rate": 2.9975890659281255e-05, "loss": 2.1061, "step": 4065 }, { "epoch": 0.9502497735970319, "grad_norm": 3.058886766433716, "learning_rate": 2.9975858468471396e-05, "loss": 1.9367, "step": 4066 }, { "epoch": 0.9504834798866525, "grad_norm": 2.6499054431915283, "learning_rate": 2.9975826256202587e-05, "loss": 1.8552, "step": 4067 }, { "epoch": 0.950717186176273, "grad_norm": 3.155825138092041, "learning_rate": 2.9975794022474866e-05, "loss": 2.0621, "step": 4068 }, { "epoch": 0.9509508924658935, "grad_norm": 2.7151174545288086, "learning_rate": 2.9975761767288287e-05, "loss": 2.0527, "step": 4069 }, { "epoch": 0.951184598755514, "grad_norm": 2.5109035968780518, "learning_rate": 2.997572949064289e-05, "loss": 1.8989, "step": 4070 }, { "epoch": 0.9514183050451346, "grad_norm": 3.6283700466156006, "learning_rate": 2.9975697192538724e-05, "loss": 2.024, "step": 4071 }, { "epoch": 0.9516520113347551, "grad_norm": 2.7026939392089844, "learning_rate": 2.997566487297584e-05, "loss": 1.9635, "step": 4072 }, { "epoch": 0.9518857176243756, "grad_norm": 3.4863736629486084, "learning_rate": 2.997563253195427e-05, "loss": 2.0012, "step": 4073 }, { "epoch": 0.9521194239139961, "grad_norm": 3.38071608543396, "learning_rate": 2.9975600169474076e-05, "loss": 1.8825, "step": 4074 }, { "epoch": 0.9523531302036166, "grad_norm": 3.72908878326416, "learning_rate": 2.9975567785535297e-05, "loss": 1.853, "step": 4075 }, { "epoch": 0.9525868364932372, "grad_norm": 2.640024185180664, "learning_rate": 2.9975535380137985e-05, "loss": 1.9708, "step": 4076 }, { "epoch": 0.9528205427828577, "grad_norm": 2.2921183109283447, "learning_rate": 2.997550295328218e-05, "loss": 1.919, "step": 4077 }, { "epoch": 0.9530542490724782, "grad_norm": 2.7906625270843506, "learning_rate": 2.997547050496793e-05, "loss": 2.0094, "step": 4078 }, { "epoch": 0.9532879553620986, "grad_norm": 2.4772541522979736, "learning_rate": 2.997543803519528e-05, "loss": 2.0596, "step": 4079 }, { "epoch": 0.9535216616517193, "grad_norm": 3.822584867477417, "learning_rate": 2.997540554396428e-05, "loss": 1.8791, "step": 4080 }, { "epoch": 0.9537553679413397, "grad_norm": 2.6158154010772705, "learning_rate": 2.9975373031274974e-05, "loss": 1.8573, "step": 4081 }, { "epoch": 0.9539890742309602, "grad_norm": 2.64974308013916, "learning_rate": 2.997534049712741e-05, "loss": 2.0553, "step": 4082 }, { "epoch": 0.9542227805205807, "grad_norm": 2.8808562755584717, "learning_rate": 2.9975307941521637e-05, "loss": 1.8917, "step": 4083 }, { "epoch": 0.9544564868102012, "grad_norm": 3.6852500438690186, "learning_rate": 2.99752753644577e-05, "loss": 1.9586, "step": 4084 }, { "epoch": 0.9546901930998218, "grad_norm": 2.4878058433532715, "learning_rate": 2.9975242765935637e-05, "loss": 1.9133, "step": 4085 }, { "epoch": 0.9549238993894423, "grad_norm": 3.90189790725708, "learning_rate": 2.997521014595551e-05, "loss": 1.9809, "step": 4086 }, { "epoch": 0.9551576056790628, "grad_norm": 2.812544345855713, "learning_rate": 2.9975177504517354e-05, "loss": 1.9208, "step": 4087 }, { "epoch": 0.9553913119686833, "grad_norm": 2.7316577434539795, "learning_rate": 2.9975144841621224e-05, "loss": 1.856, "step": 4088 }, { "epoch": 0.9556250182583039, "grad_norm": 5.278995513916016, "learning_rate": 2.997511215726716e-05, "loss": 2.0539, "step": 4089 }, { "epoch": 0.9558587245479244, "grad_norm": 5.875057220458984, "learning_rate": 2.997507945145521e-05, "loss": 1.9385, "step": 4090 }, { "epoch": 0.9560924308375449, "grad_norm": 3.0188026428222656, "learning_rate": 2.9975046724185427e-05, "loss": 2.1329, "step": 4091 }, { "epoch": 0.9563261371271654, "grad_norm": 4.914309978485107, "learning_rate": 2.9975013975457852e-05, "loss": 2.0485, "step": 4092 }, { "epoch": 0.9565598434167859, "grad_norm": 5.406667709350586, "learning_rate": 2.997498120527253e-05, "loss": 2.1878, "step": 4093 }, { "epoch": 0.9567935497064065, "grad_norm": 4.267186641693115, "learning_rate": 2.9974948413629513e-05, "loss": 1.979, "step": 4094 }, { "epoch": 0.957027255996027, "grad_norm": 3.3527908325195312, "learning_rate": 2.9974915600528844e-05, "loss": 1.9457, "step": 4095 }, { "epoch": 0.9572609622856475, "grad_norm": 4.5584001541137695, "learning_rate": 2.997488276597058e-05, "loss": 1.9764, "step": 4096 }, { "epoch": 0.957494668575268, "grad_norm": 3.152674913406372, "learning_rate": 2.9974849909954753e-05, "loss": 1.9885, "step": 4097 }, { "epoch": 0.9577283748648886, "grad_norm": 2.4512367248535156, "learning_rate": 2.997481703248142e-05, "loss": 2.091, "step": 4098 }, { "epoch": 0.9579620811545091, "grad_norm": 3.818575143814087, "learning_rate": 2.9974784133550624e-05, "loss": 2.2648, "step": 4099 }, { "epoch": 0.9581957874441296, "grad_norm": 5.7874674797058105, "learning_rate": 2.9974751213162413e-05, "loss": 2.0748, "step": 4100 }, { "epoch": 0.9584294937337501, "grad_norm": 3.378045082092285, "learning_rate": 2.997471827131684e-05, "loss": 2.0296, "step": 4101 }, { "epoch": 0.9586632000233706, "grad_norm": 2.828965187072754, "learning_rate": 2.997468530801394e-05, "loss": 1.9057, "step": 4102 }, { "epoch": 0.9588969063129912, "grad_norm": 4.485560417175293, "learning_rate": 2.997465232325377e-05, "loss": 2.0098, "step": 4103 }, { "epoch": 0.9591306126026117, "grad_norm": 4.0572686195373535, "learning_rate": 2.9974619317036373e-05, "loss": 2.0106, "step": 4104 }, { "epoch": 0.9593643188922322, "grad_norm": 2.6409761905670166, "learning_rate": 2.9974586289361804e-05, "loss": 2.1029, "step": 4105 }, { "epoch": 0.9595980251818527, "grad_norm": 4.169966697692871, "learning_rate": 2.9974553240230098e-05, "loss": 1.9687, "step": 4106 }, { "epoch": 0.9598317314714733, "grad_norm": 3.7101523876190186, "learning_rate": 2.997452016964131e-05, "loss": 1.9872, "step": 4107 }, { "epoch": 0.9600654377610938, "grad_norm": 2.4433252811431885, "learning_rate": 2.9974487077595486e-05, "loss": 1.9348, "step": 4108 }, { "epoch": 0.9602991440507143, "grad_norm": 5.31918478012085, "learning_rate": 2.9974453964092674e-05, "loss": 1.9657, "step": 4109 }, { "epoch": 0.9605328503403348, "grad_norm": 8.148451805114746, "learning_rate": 2.997442082913292e-05, "loss": 2.0433, "step": 4110 }, { "epoch": 0.9607665566299552, "grad_norm": 2.9391565322875977, "learning_rate": 2.9974387672716274e-05, "loss": 2.0094, "step": 4111 }, { "epoch": 0.9610002629195759, "grad_norm": 4.145353317260742, "learning_rate": 2.9974354494842777e-05, "loss": 2.0473, "step": 4112 }, { "epoch": 0.9612339692091963, "grad_norm": 4.145709991455078, "learning_rate": 2.997432129551248e-05, "loss": 2.0784, "step": 4113 }, { "epoch": 0.9614676754988168, "grad_norm": 2.2941510677337646, "learning_rate": 2.997428807472544e-05, "loss": 2.0204, "step": 4114 }, { "epoch": 0.9617013817884373, "grad_norm": 2.2833526134490967, "learning_rate": 2.997425483248169e-05, "loss": 1.9833, "step": 4115 }, { "epoch": 0.9619350880780579, "grad_norm": 2.6562602519989014, "learning_rate": 2.9974221568781285e-05, "loss": 2.008, "step": 4116 }, { "epoch": 0.9621687943676784, "grad_norm": 2.662088632583618, "learning_rate": 2.9974188283624275e-05, "loss": 1.9005, "step": 4117 }, { "epoch": 0.9624025006572989, "grad_norm": 2.696044921875, "learning_rate": 2.99741549770107e-05, "loss": 1.9107, "step": 4118 }, { "epoch": 0.9626362069469194, "grad_norm": 2.8368477821350098, "learning_rate": 2.9974121648940617e-05, "loss": 2.0902, "step": 4119 }, { "epoch": 0.96286991323654, "grad_norm": 3.143216133117676, "learning_rate": 2.9974088299414063e-05, "loss": 1.9222, "step": 4120 }, { "epoch": 0.9631036195261605, "grad_norm": 2.6011104583740234, "learning_rate": 2.9974054928431098e-05, "loss": 2.0077, "step": 4121 }, { "epoch": 0.963337325815781, "grad_norm": 2.996828079223633, "learning_rate": 2.997402153599176e-05, "loss": 1.8329, "step": 4122 }, { "epoch": 0.9635710321054015, "grad_norm": 4.342305660247803, "learning_rate": 2.99739881220961e-05, "loss": 2.0305, "step": 4123 }, { "epoch": 0.963804738395022, "grad_norm": 3.208226442337036, "learning_rate": 2.9973954686744165e-05, "loss": 1.9932, "step": 4124 }, { "epoch": 0.9640384446846426, "grad_norm": 4.130649089813232, "learning_rate": 2.9973921229936006e-05, "loss": 1.8994, "step": 4125 }, { "epoch": 0.9642721509742631, "grad_norm": 4.535831451416016, "learning_rate": 2.9973887751671673e-05, "loss": 1.9603, "step": 4126 }, { "epoch": 0.9645058572638836, "grad_norm": 2.8010666370391846, "learning_rate": 2.9973854251951202e-05, "loss": 1.9303, "step": 4127 }, { "epoch": 0.9647395635535041, "grad_norm": 4.035982608795166, "learning_rate": 2.9973820730774652e-05, "loss": 1.9723, "step": 4128 }, { "epoch": 0.9649732698431247, "grad_norm": 5.068716526031494, "learning_rate": 2.9973787188142073e-05, "loss": 1.9797, "step": 4129 }, { "epoch": 0.9652069761327452, "grad_norm": 4.392560958862305, "learning_rate": 2.99737536240535e-05, "loss": 1.956, "step": 4130 }, { "epoch": 0.9654406824223657, "grad_norm": 3.5831334590911865, "learning_rate": 2.9973720038508993e-05, "loss": 2.0715, "step": 4131 }, { "epoch": 0.9656743887119862, "grad_norm": 4.443781852722168, "learning_rate": 2.9973686431508598e-05, "loss": 1.9311, "step": 4132 }, { "epoch": 0.9659080950016067, "grad_norm": 2.906735420227051, "learning_rate": 2.997365280305236e-05, "loss": 1.9667, "step": 4133 }, { "epoch": 0.9661418012912273, "grad_norm": 2.8488786220550537, "learning_rate": 2.997361915314033e-05, "loss": 2.0408, "step": 4134 }, { "epoch": 0.9663755075808478, "grad_norm": 4.402078151702881, "learning_rate": 2.9973585481772553e-05, "loss": 2.0017, "step": 4135 }, { "epoch": 0.9666092138704683, "grad_norm": 2.7283051013946533, "learning_rate": 2.9973551788949082e-05, "loss": 2.1669, "step": 4136 }, { "epoch": 0.9668429201600888, "grad_norm": 5.920908451080322, "learning_rate": 2.9973518074669962e-05, "loss": 1.9343, "step": 4137 }, { "epoch": 0.9670766264497094, "grad_norm": 3.739349842071533, "learning_rate": 2.997348433893524e-05, "loss": 2.0087, "step": 4138 }, { "epoch": 0.9673103327393299, "grad_norm": 5.584187984466553, "learning_rate": 2.9973450581744966e-05, "loss": 2.0521, "step": 4139 }, { "epoch": 0.9675440390289504, "grad_norm": 2.7397050857543945, "learning_rate": 2.9973416803099185e-05, "loss": 2.1764, "step": 4140 }, { "epoch": 0.9677777453185709, "grad_norm": 2.7138941287994385, "learning_rate": 2.9973383002997955e-05, "loss": 2.004, "step": 4141 }, { "epoch": 0.9680114516081914, "grad_norm": 2.59317946434021, "learning_rate": 2.9973349181441315e-05, "loss": 1.9967, "step": 4142 }, { "epoch": 0.968245157897812, "grad_norm": 2.3398616313934326, "learning_rate": 2.9973315338429318e-05, "loss": 2.0162, "step": 4143 }, { "epoch": 0.9684788641874325, "grad_norm": 2.333987236022949, "learning_rate": 2.997328147396201e-05, "loss": 2.0479, "step": 4144 }, { "epoch": 0.968712570477053, "grad_norm": 2.2605655193328857, "learning_rate": 2.9973247588039442e-05, "loss": 2.1061, "step": 4145 }, { "epoch": 0.9689462767666734, "grad_norm": 2.5932776927948, "learning_rate": 2.997321368066166e-05, "loss": 2.033, "step": 4146 }, { "epoch": 0.969179983056294, "grad_norm": 3.330770254135132, "learning_rate": 2.9973179751828716e-05, "loss": 1.7875, "step": 4147 }, { "epoch": 0.9694136893459145, "grad_norm": 2.7219293117523193, "learning_rate": 2.9973145801540653e-05, "loss": 2.1122, "step": 4148 }, { "epoch": 0.969647395635535, "grad_norm": 5.2655134201049805, "learning_rate": 2.9973111829797527e-05, "loss": 1.9414, "step": 4149 }, { "epoch": 0.9698811019251555, "grad_norm": 3.42136812210083, "learning_rate": 2.997307783659938e-05, "loss": 2.0577, "step": 4150 }, { "epoch": 0.970114808214776, "grad_norm": 2.4765870571136475, "learning_rate": 2.9973043821946268e-05, "loss": 2.0089, "step": 4151 }, { "epoch": 0.9703485145043966, "grad_norm": 3.450800657272339, "learning_rate": 2.997300978583823e-05, "loss": 1.8527, "step": 4152 }, { "epoch": 0.9705822207940171, "grad_norm": 3.6654088497161865, "learning_rate": 2.9972975728275324e-05, "loss": 1.9995, "step": 4153 }, { "epoch": 0.9708159270836376, "grad_norm": 2.712965965270996, "learning_rate": 2.9972941649257592e-05, "loss": 2.0699, "step": 4154 }, { "epoch": 0.9710496333732581, "grad_norm": 3.8475914001464844, "learning_rate": 2.9972907548785086e-05, "loss": 1.995, "step": 4155 }, { "epoch": 0.9712833396628787, "grad_norm": 4.305470943450928, "learning_rate": 2.9972873426857858e-05, "loss": 2.0477, "step": 4156 }, { "epoch": 0.9715170459524992, "grad_norm": 2.698617696762085, "learning_rate": 2.997283928347595e-05, "loss": 2.0008, "step": 4157 }, { "epoch": 0.9717507522421197, "grad_norm": 2.524899959564209, "learning_rate": 2.9972805118639415e-05, "loss": 1.9464, "step": 4158 }, { "epoch": 0.9719844585317402, "grad_norm": 3.318150758743286, "learning_rate": 2.99727709323483e-05, "loss": 1.9546, "step": 4159 }, { "epoch": 0.9722181648213608, "grad_norm": 3.734119176864624, "learning_rate": 2.9972736724602656e-05, "loss": 1.9357, "step": 4160 }, { "epoch": 0.9724518711109813, "grad_norm": 4.502737045288086, "learning_rate": 2.9972702495402534e-05, "loss": 1.9293, "step": 4161 }, { "epoch": 0.9726855774006018, "grad_norm": 3.531345844268799, "learning_rate": 2.9972668244747976e-05, "loss": 1.9968, "step": 4162 }, { "epoch": 0.9729192836902223, "grad_norm": 2.4950175285339355, "learning_rate": 2.9972633972639038e-05, "loss": 1.8763, "step": 4163 }, { "epoch": 0.9731529899798428, "grad_norm": 3.530059576034546, "learning_rate": 2.9972599679075766e-05, "loss": 1.9795, "step": 4164 }, { "epoch": 0.9733866962694634, "grad_norm": 4.079337120056152, "learning_rate": 2.9972565364058212e-05, "loss": 1.9692, "step": 4165 }, { "epoch": 0.9736204025590839, "grad_norm": 2.762408494949341, "learning_rate": 2.997253102758642e-05, "loss": 2.0446, "step": 4166 }, { "epoch": 0.9738541088487044, "grad_norm": 4.0837554931640625, "learning_rate": 2.9972496669660444e-05, "loss": 2.0894, "step": 4167 }, { "epoch": 0.9740878151383249, "grad_norm": 4.463430881500244, "learning_rate": 2.9972462290280333e-05, "loss": 1.9597, "step": 4168 }, { "epoch": 0.9743215214279455, "grad_norm": 2.8307459354400635, "learning_rate": 2.997242788944613e-05, "loss": 2.1571, "step": 4169 }, { "epoch": 0.974555227717566, "grad_norm": 2.51045560836792, "learning_rate": 2.9972393467157893e-05, "loss": 2.0341, "step": 4170 }, { "epoch": 0.9747889340071865, "grad_norm": 3.2568066120147705, "learning_rate": 2.9972359023415666e-05, "loss": 1.9504, "step": 4171 }, { "epoch": 0.975022640296807, "grad_norm": 2.429187297821045, "learning_rate": 2.9972324558219496e-05, "loss": 2.0597, "step": 4172 }, { "epoch": 0.9752563465864275, "grad_norm": 3.1446642875671387, "learning_rate": 2.997229007156944e-05, "loss": 1.8647, "step": 4173 }, { "epoch": 0.9754900528760481, "grad_norm": 2.6339051723480225, "learning_rate": 2.9972255563465542e-05, "loss": 1.8976, "step": 4174 }, { "epoch": 0.9757237591656686, "grad_norm": 2.9041225910186768, "learning_rate": 2.9972221033907852e-05, "loss": 2.033, "step": 4175 }, { "epoch": 0.975957465455289, "grad_norm": 2.9690351486206055, "learning_rate": 2.9972186482896423e-05, "loss": 1.8494, "step": 4176 }, { "epoch": 0.9761911717449095, "grad_norm": 2.5803136825561523, "learning_rate": 2.9972151910431297e-05, "loss": 2.0776, "step": 4177 }, { "epoch": 0.9764248780345302, "grad_norm": 2.6247496604919434, "learning_rate": 2.9972117316512533e-05, "loss": 1.9164, "step": 4178 }, { "epoch": 0.9766585843241506, "grad_norm": 3.781344175338745, "learning_rate": 2.9972082701140173e-05, "loss": 1.9198, "step": 4179 }, { "epoch": 0.9768922906137711, "grad_norm": 4.248897552490234, "learning_rate": 2.9972048064314268e-05, "loss": 1.7886, "step": 4180 }, { "epoch": 0.9771259969033916, "grad_norm": 2.6956026554107666, "learning_rate": 2.997201340603487e-05, "loss": 2.0235, "step": 4181 }, { "epoch": 0.9773597031930121, "grad_norm": 2.5485713481903076, "learning_rate": 2.9971978726302035e-05, "loss": 1.8482, "step": 4182 }, { "epoch": 0.9775934094826327, "grad_norm": 2.2776012420654297, "learning_rate": 2.9971944025115795e-05, "loss": 1.9724, "step": 4183 }, { "epoch": 0.9778271157722532, "grad_norm": 2.4609615802764893, "learning_rate": 2.997190930247622e-05, "loss": 1.9345, "step": 4184 }, { "epoch": 0.9780608220618737, "grad_norm": 3.0937082767486572, "learning_rate": 2.9971874558383342e-05, "loss": 1.929, "step": 4185 }, { "epoch": 0.9782945283514942, "grad_norm": 2.591714382171631, "learning_rate": 2.997183979283722e-05, "loss": 1.9649, "step": 4186 }, { "epoch": 0.9785282346411148, "grad_norm": 3.0134150981903076, "learning_rate": 2.9971805005837907e-05, "loss": 2.0439, "step": 4187 }, { "epoch": 0.9787619409307353, "grad_norm": 2.7262871265411377, "learning_rate": 2.9971770197385448e-05, "loss": 2.0269, "step": 4188 }, { "epoch": 0.9789956472203558, "grad_norm": 2.748950958251953, "learning_rate": 2.9971735367479886e-05, "loss": 1.968, "step": 4189 }, { "epoch": 0.9792293535099763, "grad_norm": 2.6083984375, "learning_rate": 2.9971700516121284e-05, "loss": 1.9307, "step": 4190 }, { "epoch": 0.9794630597995968, "grad_norm": 2.580885410308838, "learning_rate": 2.9971665643309686e-05, "loss": 1.8629, "step": 4191 }, { "epoch": 0.9796967660892174, "grad_norm": 3.201083183288574, "learning_rate": 2.9971630749045142e-05, "loss": 2.1017, "step": 4192 }, { "epoch": 0.9799304723788379, "grad_norm": 2.259486675262451, "learning_rate": 2.99715958333277e-05, "loss": 1.9156, "step": 4193 }, { "epoch": 0.9801641786684584, "grad_norm": 2.8032631874084473, "learning_rate": 2.9971560896157413e-05, "loss": 1.8449, "step": 4194 }, { "epoch": 0.9803978849580789, "grad_norm": 2.5799944400787354, "learning_rate": 2.9971525937534333e-05, "loss": 2.0382, "step": 4195 }, { "epoch": 0.9806315912476995, "grad_norm": 2.323934316635132, "learning_rate": 2.9971490957458503e-05, "loss": 1.9439, "step": 4196 }, { "epoch": 0.98086529753732, "grad_norm": 2.5449018478393555, "learning_rate": 2.9971455955929983e-05, "loss": 1.9851, "step": 4197 }, { "epoch": 0.9810990038269405, "grad_norm": 2.757823944091797, "learning_rate": 2.997142093294881e-05, "loss": 1.9461, "step": 4198 }, { "epoch": 0.981332710116561, "grad_norm": 2.590665578842163, "learning_rate": 2.9971385888515046e-05, "loss": 2.0135, "step": 4199 }, { "epoch": 0.9815664164061816, "grad_norm": 2.929641008377075, "learning_rate": 2.9971350822628738e-05, "loss": 1.9306, "step": 4200 }, { "epoch": 0.9818001226958021, "grad_norm": 3.320493459701538, "learning_rate": 2.9971315735289935e-05, "loss": 1.9613, "step": 4201 }, { "epoch": 0.9820338289854226, "grad_norm": 2.766523838043213, "learning_rate": 2.9971280626498684e-05, "loss": 1.871, "step": 4202 }, { "epoch": 0.9822675352750431, "grad_norm": 4.051684379577637, "learning_rate": 2.9971245496255043e-05, "loss": 1.9077, "step": 4203 }, { "epoch": 0.9825012415646636, "grad_norm": 2.947244167327881, "learning_rate": 2.9971210344559053e-05, "loss": 1.9144, "step": 4204 }, { "epoch": 0.9827349478542842, "grad_norm": 2.707946538925171, "learning_rate": 2.9971175171410775e-05, "loss": 2.0501, "step": 4205 }, { "epoch": 0.9829686541439047, "grad_norm": 2.6913113594055176, "learning_rate": 2.9971139976810255e-05, "loss": 2.0172, "step": 4206 }, { "epoch": 0.9832023604335252, "grad_norm": 2.676081895828247, "learning_rate": 2.997110476075753e-05, "loss": 2.2038, "step": 4207 }, { "epoch": 0.9834360667231457, "grad_norm": 2.3923404216766357, "learning_rate": 2.997106952325268e-05, "loss": 1.9951, "step": 4208 }, { "epoch": 0.9836697730127663, "grad_norm": 2.563236951828003, "learning_rate": 2.9971034264295724e-05, "loss": 1.9245, "step": 4209 }, { "epoch": 0.9839034793023868, "grad_norm": 3.735856294631958, "learning_rate": 2.997099898388674e-05, "loss": 2.026, "step": 4210 }, { "epoch": 0.9841371855920072, "grad_norm": 2.480625629425049, "learning_rate": 2.9970963682025756e-05, "loss": 1.9743, "step": 4211 }, { "epoch": 0.9843708918816277, "grad_norm": 3.6213154792785645, "learning_rate": 2.997092835871283e-05, "loss": 1.9698, "step": 4212 }, { "epoch": 0.9846045981712482, "grad_norm": 2.5149736404418945, "learning_rate": 2.997089301394802e-05, "loss": 1.9678, "step": 4213 }, { "epoch": 0.9848383044608688, "grad_norm": 2.925884962081909, "learning_rate": 2.997085764773137e-05, "loss": 1.9086, "step": 4214 }, { "epoch": 0.9850720107504893, "grad_norm": 2.7639260292053223, "learning_rate": 2.997082226006293e-05, "loss": 1.8932, "step": 4215 }, { "epoch": 0.9853057170401098, "grad_norm": 2.66447377204895, "learning_rate": 2.9970786850942756e-05, "loss": 2.0382, "step": 4216 }, { "epoch": 0.9855394233297303, "grad_norm": 2.672908306121826, "learning_rate": 2.9970751420370892e-05, "loss": 1.9261, "step": 4217 }, { "epoch": 0.9857731296193509, "grad_norm": 2.5821025371551514, "learning_rate": 2.9970715968347393e-05, "loss": 1.9602, "step": 4218 }, { "epoch": 0.9860068359089714, "grad_norm": 2.457000970840454, "learning_rate": 2.9970680494872307e-05, "loss": 1.8926, "step": 4219 }, { "epoch": 0.9862405421985919, "grad_norm": 2.531712055206299, "learning_rate": 2.9970644999945692e-05, "loss": 1.9562, "step": 4220 }, { "epoch": 0.9864742484882124, "grad_norm": 2.9034905433654785, "learning_rate": 2.997060948356759e-05, "loss": 2.0682, "step": 4221 }, { "epoch": 0.9867079547778329, "grad_norm": 2.3093175888061523, "learning_rate": 2.997057394573806e-05, "loss": 2.0388, "step": 4222 }, { "epoch": 0.9869416610674535, "grad_norm": 2.7819888591766357, "learning_rate": 2.9970538386457143e-05, "loss": 2.041, "step": 4223 }, { "epoch": 0.987175367357074, "grad_norm": 2.6262710094451904, "learning_rate": 2.9970502805724898e-05, "loss": 1.9651, "step": 4224 }, { "epoch": 0.9874090736466945, "grad_norm": 2.5735599994659424, "learning_rate": 2.997046720354137e-05, "loss": 2.0055, "step": 4225 }, { "epoch": 0.987642779936315, "grad_norm": 3.0846312046051025, "learning_rate": 2.9970431579906618e-05, "loss": 1.8786, "step": 4226 }, { "epoch": 0.9878764862259356, "grad_norm": 3.0060994625091553, "learning_rate": 2.9970395934820685e-05, "loss": 1.8254, "step": 4227 }, { "epoch": 0.9881101925155561, "grad_norm": 2.6071596145629883, "learning_rate": 2.9970360268283626e-05, "loss": 2.0773, "step": 4228 }, { "epoch": 0.9883438988051766, "grad_norm": 3.257136583328247, "learning_rate": 2.997032458029549e-05, "loss": 1.8953, "step": 4229 }, { "epoch": 0.9885776050947971, "grad_norm": 2.9076087474823, "learning_rate": 2.9970288870856332e-05, "loss": 2.0351, "step": 4230 }, { "epoch": 0.9888113113844176, "grad_norm": 2.9801056385040283, "learning_rate": 2.9970253139966202e-05, "loss": 1.9566, "step": 4231 }, { "epoch": 0.9890450176740382, "grad_norm": 3.190408229827881, "learning_rate": 2.997021738762515e-05, "loss": 1.9003, "step": 4232 }, { "epoch": 0.9892787239636587, "grad_norm": 2.4900259971618652, "learning_rate": 2.9970181613833225e-05, "loss": 1.9817, "step": 4233 }, { "epoch": 0.9895124302532792, "grad_norm": 3.1888110637664795, "learning_rate": 2.9970145818590478e-05, "loss": 1.9028, "step": 4234 }, { "epoch": 0.9897461365428997, "grad_norm": 2.7422871589660645, "learning_rate": 2.9970110001896968e-05, "loss": 2.0143, "step": 4235 }, { "epoch": 0.9899798428325203, "grad_norm": 3.2592599391937256, "learning_rate": 2.9970074163752745e-05, "loss": 1.9148, "step": 4236 }, { "epoch": 0.9902135491221408, "grad_norm": 2.721808671951294, "learning_rate": 2.997003830415785e-05, "loss": 1.9642, "step": 4237 }, { "epoch": 0.9904472554117613, "grad_norm": 3.486818790435791, "learning_rate": 2.9970002423112342e-05, "loss": 1.8645, "step": 4238 }, { "epoch": 0.9906809617013818, "grad_norm": 2.700803279876709, "learning_rate": 2.9969966520616268e-05, "loss": 1.8589, "step": 4239 }, { "epoch": 0.9909146679910024, "grad_norm": 3.072617530822754, "learning_rate": 2.9969930596669688e-05, "loss": 1.8657, "step": 4240 }, { "epoch": 0.9911483742806229, "grad_norm": 2.740347385406494, "learning_rate": 2.9969894651272645e-05, "loss": 1.9422, "step": 4241 }, { "epoch": 0.9913820805702434, "grad_norm": 2.9905033111572266, "learning_rate": 2.9969858684425194e-05, "loss": 2.0362, "step": 4242 }, { "epoch": 0.9916157868598638, "grad_norm": 3.9980623722076416, "learning_rate": 2.9969822696127386e-05, "loss": 2.0379, "step": 4243 }, { "epoch": 0.9918494931494843, "grad_norm": 2.9621403217315674, "learning_rate": 2.996978668637928e-05, "loss": 1.8271, "step": 4244 }, { "epoch": 0.992083199439105, "grad_norm": 3.099782705307007, "learning_rate": 2.9969750655180912e-05, "loss": 1.9014, "step": 4245 }, { "epoch": 0.9923169057287254, "grad_norm": 2.375370979309082, "learning_rate": 2.9969714602532344e-05, "loss": 1.8441, "step": 4246 }, { "epoch": 0.9925506120183459, "grad_norm": 4.170222759246826, "learning_rate": 2.996967852843363e-05, "loss": 2.0006, "step": 4247 }, { "epoch": 0.9927843183079664, "grad_norm": 2.4499337673187256, "learning_rate": 2.996964243288481e-05, "loss": 1.9573, "step": 4248 }, { "epoch": 0.993018024597587, "grad_norm": 2.625404119491577, "learning_rate": 2.996960631588595e-05, "loss": 2.0855, "step": 4249 }, { "epoch": 0.9932517308872075, "grad_norm": 2.33793044090271, "learning_rate": 2.996957017743709e-05, "loss": 1.965, "step": 4250 }, { "epoch": 0.993485437176828, "grad_norm": 2.305114507675171, "learning_rate": 2.9969534017538286e-05, "loss": 1.9414, "step": 4251 }, { "epoch": 0.9937191434664485, "grad_norm": 2.3155603408813477, "learning_rate": 2.9969497836189597e-05, "loss": 1.8839, "step": 4252 }, { "epoch": 0.993952849756069, "grad_norm": 3.659715175628662, "learning_rate": 2.9969461633391063e-05, "loss": 1.703, "step": 4253 }, { "epoch": 0.9941865560456896, "grad_norm": 2.4802122116088867, "learning_rate": 2.9969425409142744e-05, "loss": 1.7985, "step": 4254 }, { "epoch": 0.9944202623353101, "grad_norm": 3.826115608215332, "learning_rate": 2.996938916344469e-05, "loss": 1.9156, "step": 4255 }, { "epoch": 0.9946539686249306, "grad_norm": 2.599834680557251, "learning_rate": 2.9969352896296946e-05, "loss": 1.933, "step": 4256 }, { "epoch": 0.9948876749145511, "grad_norm": 2.597306251525879, "learning_rate": 2.9969316607699575e-05, "loss": 2.1331, "step": 4257 }, { "epoch": 0.9951213812041717, "grad_norm": 2.3994576930999756, "learning_rate": 2.9969280297652622e-05, "loss": 1.9672, "step": 4258 }, { "epoch": 0.9953550874937922, "grad_norm": 2.3446879386901855, "learning_rate": 2.9969243966156143e-05, "loss": 1.9635, "step": 4259 }, { "epoch": 0.9955887937834127, "grad_norm": 2.5164363384246826, "learning_rate": 2.9969207613210187e-05, "loss": 1.872, "step": 4260 }, { "epoch": 0.9958225000730332, "grad_norm": 2.9235949516296387, "learning_rate": 2.9969171238814805e-05, "loss": 1.9529, "step": 4261 }, { "epoch": 0.9960562063626537, "grad_norm": 3.011176347732544, "learning_rate": 2.9969134842970053e-05, "loss": 1.9767, "step": 4262 }, { "epoch": 0.9962899126522743, "grad_norm": 2.6119580268859863, "learning_rate": 2.9969098425675983e-05, "loss": 1.9803, "step": 4263 }, { "epoch": 0.9965236189418948, "grad_norm": 4.154172897338867, "learning_rate": 2.9969061986932643e-05, "loss": 1.9023, "step": 4264 }, { "epoch": 0.9967573252315153, "grad_norm": 2.717188596725464, "learning_rate": 2.9969025526740092e-05, "loss": 1.9367, "step": 4265 }, { "epoch": 0.9969910315211358, "grad_norm": 2.6337783336639404, "learning_rate": 2.996898904509838e-05, "loss": 2.0358, "step": 4266 }, { "epoch": 0.9972247378107564, "grad_norm": 3.1223509311676025, "learning_rate": 2.996895254200755e-05, "loss": 1.7783, "step": 4267 }, { "epoch": 0.9974584441003769, "grad_norm": 2.503702163696289, "learning_rate": 2.9968916017467663e-05, "loss": 2.0938, "step": 4268 }, { "epoch": 0.9976921503899974, "grad_norm": 2.789860963821411, "learning_rate": 2.9968879471478775e-05, "loss": 1.9677, "step": 4269 }, { "epoch": 0.9979258566796179, "grad_norm": 4.018794059753418, "learning_rate": 2.996884290404093e-05, "loss": 2.0419, "step": 4270 }, { "epoch": 0.9981595629692384, "grad_norm": 2.4916012287139893, "learning_rate": 2.9968806315154183e-05, "loss": 2.0531, "step": 4271 }, { "epoch": 0.998393269258859, "grad_norm": 2.7116336822509766, "learning_rate": 2.996876970481859e-05, "loss": 1.9413, "step": 4272 }, { "epoch": 0.9986269755484795, "grad_norm": 2.665799379348755, "learning_rate": 2.9968733073034196e-05, "loss": 2.1474, "step": 4273 }, { "epoch": 0.9988606818381, "grad_norm": 2.584745168685913, "learning_rate": 2.9968696419801064e-05, "loss": 1.7959, "step": 4274 }, { "epoch": 0.9990943881277204, "grad_norm": 2.975522756576538, "learning_rate": 2.996865974511924e-05, "loss": 1.8584, "step": 4275 }, { "epoch": 0.999328094417341, "grad_norm": 2.6016507148742676, "learning_rate": 2.9968623048988773e-05, "loss": 2.0582, "step": 4276 }, { "epoch": 0.9995618007069615, "grad_norm": 3.4948620796203613, "learning_rate": 2.9968586331409723e-05, "loss": 2.1913, "step": 4277 }, { "epoch": 0.999795506996582, "grad_norm": 3.0642826557159424, "learning_rate": 2.9968549592382134e-05, "loss": 2.0201, "step": 4278 }, { "epoch": 1.0000292132862025, "grad_norm": 2.6817879676818848, "learning_rate": 2.9968512831906073e-05, "loss": 1.9131, "step": 4279 }, { "epoch": 1.000262919575823, "grad_norm": 3.507995843887329, "learning_rate": 2.996847604998158e-05, "loss": 1.806, "step": 4280 }, { "epoch": 1.0004966258654435, "grad_norm": 3.5531961917877197, "learning_rate": 2.996843924660871e-05, "loss": 1.9028, "step": 4281 }, { "epoch": 1.000730332155064, "grad_norm": 3.315309762954712, "learning_rate": 2.9968402421787518e-05, "loss": 1.9638, "step": 4282 }, { "epoch": 1.0009640384446847, "grad_norm": 2.5635900497436523, "learning_rate": 2.9968365575518054e-05, "loss": 1.9027, "step": 4283 }, { "epoch": 1.0011977447343052, "grad_norm": 2.4452192783355713, "learning_rate": 2.9968328707800377e-05, "loss": 1.875, "step": 4284 }, { "epoch": 1.0014314510239257, "grad_norm": 2.538828134536743, "learning_rate": 2.9968291818634532e-05, "loss": 1.9126, "step": 4285 }, { "epoch": 1.0016651573135462, "grad_norm": 2.578315019607544, "learning_rate": 2.9968254908020576e-05, "loss": 1.8965, "step": 4286 }, { "epoch": 1.0018988636031667, "grad_norm": 2.7060329914093018, "learning_rate": 2.996821797595856e-05, "loss": 1.8834, "step": 4287 }, { "epoch": 1.0021325698927872, "grad_norm": 2.4530344009399414, "learning_rate": 2.996818102244854e-05, "loss": 1.9, "step": 4288 }, { "epoch": 1.0023662761824077, "grad_norm": 3.630093574523926, "learning_rate": 2.996814404749057e-05, "loss": 1.8826, "step": 4289 }, { "epoch": 1.0025999824720282, "grad_norm": 2.4834794998168945, "learning_rate": 2.9968107051084697e-05, "loss": 1.8896, "step": 4290 }, { "epoch": 1.0028336887616487, "grad_norm": 3.3435208797454834, "learning_rate": 2.996807003323098e-05, "loss": 1.8854, "step": 4291 }, { "epoch": 1.0030673950512694, "grad_norm": 3.0034284591674805, "learning_rate": 2.9968032993929466e-05, "loss": 1.7882, "step": 4292 }, { "epoch": 1.00330110134089, "grad_norm": 2.3247275352478027, "learning_rate": 2.9967995933180214e-05, "loss": 2.0322, "step": 4293 }, { "epoch": 1.0035348076305104, "grad_norm": 3.2269883155822754, "learning_rate": 2.9967958850983272e-05, "loss": 2.064, "step": 4294 }, { "epoch": 1.003768513920131, "grad_norm": 3.4670493602752686, "learning_rate": 2.9967921747338696e-05, "loss": 1.8575, "step": 4295 }, { "epoch": 1.0040022202097514, "grad_norm": 3.4267992973327637, "learning_rate": 2.996788462224654e-05, "loss": 1.9363, "step": 4296 }, { "epoch": 1.0042359264993719, "grad_norm": 3.2384209632873535, "learning_rate": 2.9967847475706856e-05, "loss": 1.9607, "step": 4297 }, { "epoch": 1.0044696327889924, "grad_norm": 2.5272178649902344, "learning_rate": 2.9967810307719696e-05, "loss": 1.8868, "step": 4298 }, { "epoch": 1.0047033390786129, "grad_norm": 2.6574625968933105, "learning_rate": 2.9967773118285116e-05, "loss": 2.0511, "step": 4299 }, { "epoch": 1.0049370453682336, "grad_norm": 2.398665189743042, "learning_rate": 2.9967735907403168e-05, "loss": 1.9463, "step": 4300 }, { "epoch": 1.005170751657854, "grad_norm": 2.5124871730804443, "learning_rate": 2.99676986750739e-05, "loss": 1.9674, "step": 4301 }, { "epoch": 1.0054044579474746, "grad_norm": 2.3859825134277344, "learning_rate": 2.9967661421297374e-05, "loss": 1.9418, "step": 4302 }, { "epoch": 1.005638164237095, "grad_norm": 2.5386087894439697, "learning_rate": 2.996762414607364e-05, "loss": 1.8828, "step": 4303 }, { "epoch": 1.0058718705267156, "grad_norm": 3.5010826587677, "learning_rate": 2.996758684940275e-05, "loss": 1.852, "step": 4304 }, { "epoch": 1.006105576816336, "grad_norm": 2.559016466140747, "learning_rate": 2.996754953128476e-05, "loss": 1.8513, "step": 4305 }, { "epoch": 1.0063392831059566, "grad_norm": 2.473668336868286, "learning_rate": 2.996751219171972e-05, "loss": 1.8991, "step": 4306 }, { "epoch": 1.006572989395577, "grad_norm": 2.734959125518799, "learning_rate": 2.996747483070769e-05, "loss": 2.0251, "step": 4307 }, { "epoch": 1.0068066956851975, "grad_norm": 2.7396864891052246, "learning_rate": 2.9967437448248716e-05, "loss": 1.8426, "step": 4308 }, { "epoch": 1.0070404019748183, "grad_norm": 3.061062812805176, "learning_rate": 2.9967400044342853e-05, "loss": 1.9184, "step": 4309 }, { "epoch": 1.0072741082644388, "grad_norm": 2.3727591037750244, "learning_rate": 2.996736261899016e-05, "loss": 1.8873, "step": 4310 }, { "epoch": 1.0075078145540592, "grad_norm": 2.8231024742126465, "learning_rate": 2.9967325172190684e-05, "loss": 1.8676, "step": 4311 }, { "epoch": 1.0077415208436797, "grad_norm": 2.842597484588623, "learning_rate": 2.9967287703944483e-05, "loss": 1.8964, "step": 4312 }, { "epoch": 1.0079752271333002, "grad_norm": 2.7946250438690186, "learning_rate": 2.996725021425161e-05, "loss": 1.9857, "step": 4313 }, { "epoch": 1.0082089334229207, "grad_norm": 2.994807481765747, "learning_rate": 2.9967212703112115e-05, "loss": 1.9121, "step": 4314 }, { "epoch": 1.0084426397125412, "grad_norm": 2.478501081466675, "learning_rate": 2.9967175170526057e-05, "loss": 1.924, "step": 4315 }, { "epoch": 1.0086763460021617, "grad_norm": 3.167609214782715, "learning_rate": 2.9967137616493484e-05, "loss": 1.8248, "step": 4316 }, { "epoch": 1.0089100522917822, "grad_norm": 2.920461654663086, "learning_rate": 2.9967100041014457e-05, "loss": 1.8169, "step": 4317 }, { "epoch": 1.009143758581403, "grad_norm": 2.600160598754883, "learning_rate": 2.9967062444089026e-05, "loss": 2.0233, "step": 4318 }, { "epoch": 1.0093774648710234, "grad_norm": 2.61716365814209, "learning_rate": 2.9967024825717246e-05, "loss": 1.825, "step": 4319 }, { "epoch": 1.009611171160644, "grad_norm": 2.5726656913757324, "learning_rate": 2.9966987185899172e-05, "loss": 2.1178, "step": 4320 }, { "epoch": 1.0098448774502644, "grad_norm": 2.8814470767974854, "learning_rate": 2.996694952463485e-05, "loss": 1.8202, "step": 4321 }, { "epoch": 1.010078583739885, "grad_norm": 2.5026679039001465, "learning_rate": 2.9966911841924342e-05, "loss": 1.9715, "step": 4322 }, { "epoch": 1.0103122900295054, "grad_norm": 2.5481526851654053, "learning_rate": 2.9966874137767698e-05, "loss": 1.9138, "step": 4323 }, { "epoch": 1.010545996319126, "grad_norm": 2.5394463539123535, "learning_rate": 2.9966836412164976e-05, "loss": 1.9355, "step": 4324 }, { "epoch": 1.0107797026087464, "grad_norm": 2.543947458267212, "learning_rate": 2.9966798665116227e-05, "loss": 1.9176, "step": 4325 }, { "epoch": 1.0110134088983669, "grad_norm": 2.7116432189941406, "learning_rate": 2.9966760896621508e-05, "loss": 2.1125, "step": 4326 }, { "epoch": 1.0112471151879876, "grad_norm": 2.7122809886932373, "learning_rate": 2.996672310668087e-05, "loss": 1.9131, "step": 4327 }, { "epoch": 1.011480821477608, "grad_norm": 2.4446911811828613, "learning_rate": 2.996668529529437e-05, "loss": 1.8488, "step": 4328 }, { "epoch": 1.0117145277672286, "grad_norm": 2.5627005100250244, "learning_rate": 2.9966647462462056e-05, "loss": 1.9047, "step": 4329 }, { "epoch": 1.011948234056849, "grad_norm": 3.5130746364593506, "learning_rate": 2.996660960818399e-05, "loss": 1.8698, "step": 4330 }, { "epoch": 1.0121819403464696, "grad_norm": 2.6239216327667236, "learning_rate": 2.9966571732460224e-05, "loss": 1.9806, "step": 4331 }, { "epoch": 1.01241564663609, "grad_norm": 3.6350173950195312, "learning_rate": 2.9966533835290806e-05, "loss": 2.0024, "step": 4332 }, { "epoch": 1.0126493529257106, "grad_norm": 2.8312013149261475, "learning_rate": 2.99664959166758e-05, "loss": 1.9909, "step": 4333 }, { "epoch": 1.012883059215331, "grad_norm": 2.275416851043701, "learning_rate": 2.9966457976615257e-05, "loss": 1.9848, "step": 4334 }, { "epoch": 1.0131167655049516, "grad_norm": 3.8841192722320557, "learning_rate": 2.9966420015109224e-05, "loss": 1.9262, "step": 4335 }, { "epoch": 1.0133504717945723, "grad_norm": 2.5677216053009033, "learning_rate": 2.9966382032157767e-05, "loss": 1.8493, "step": 4336 }, { "epoch": 1.0135841780841928, "grad_norm": 2.8408772945404053, "learning_rate": 2.9966344027760934e-05, "loss": 1.8058, "step": 4337 }, { "epoch": 1.0138178843738133, "grad_norm": 3.9724626541137695, "learning_rate": 2.9966306001918778e-05, "loss": 1.8309, "step": 4338 }, { "epoch": 1.0140515906634338, "grad_norm": 2.726712942123413, "learning_rate": 2.996626795463136e-05, "loss": 1.9002, "step": 4339 }, { "epoch": 1.0142852969530542, "grad_norm": 3.8235299587249756, "learning_rate": 2.996622988589873e-05, "loss": 1.8454, "step": 4340 }, { "epoch": 1.0145190032426747, "grad_norm": 3.614804744720459, "learning_rate": 2.9966191795720938e-05, "loss": 2.0102, "step": 4341 }, { "epoch": 1.0147527095322952, "grad_norm": 2.612334966659546, "learning_rate": 2.9966153684098045e-05, "loss": 1.9962, "step": 4342 }, { "epoch": 1.0149864158219157, "grad_norm": 4.927137851715088, "learning_rate": 2.996611555103011e-05, "loss": 1.9201, "step": 4343 }, { "epoch": 1.0152201221115362, "grad_norm": 4.868696689605713, "learning_rate": 2.9966077396517174e-05, "loss": 1.8732, "step": 4344 }, { "epoch": 1.015453828401157, "grad_norm": 2.4816195964813232, "learning_rate": 2.9966039220559306e-05, "loss": 1.965, "step": 4345 }, { "epoch": 1.0156875346907774, "grad_norm": 2.660860061645508, "learning_rate": 2.996600102315655e-05, "loss": 1.9482, "step": 4346 }, { "epoch": 1.015921240980398, "grad_norm": 4.884754657745361, "learning_rate": 2.9965962804308968e-05, "loss": 2.014, "step": 4347 }, { "epoch": 1.0161549472700184, "grad_norm": 3.4369421005249023, "learning_rate": 2.996592456401661e-05, "loss": 1.9318, "step": 4348 }, { "epoch": 1.016388653559639, "grad_norm": 2.63602614402771, "learning_rate": 2.9965886302279532e-05, "loss": 1.9406, "step": 4349 }, { "epoch": 1.0166223598492594, "grad_norm": 5.598435401916504, "learning_rate": 2.996584801909779e-05, "loss": 1.9439, "step": 4350 }, { "epoch": 1.01685606613888, "grad_norm": 3.719611167907715, "learning_rate": 2.996580971447144e-05, "loss": 1.9371, "step": 4351 }, { "epoch": 1.0170897724285004, "grad_norm": 2.7052416801452637, "learning_rate": 2.996577138840053e-05, "loss": 2.0654, "step": 4352 }, { "epoch": 1.017323478718121, "grad_norm": 4.4865312576293945, "learning_rate": 2.9965733040885125e-05, "loss": 2.0131, "step": 4353 }, { "epoch": 1.0175571850077416, "grad_norm": 3.620802879333496, "learning_rate": 2.9965694671925274e-05, "loss": 1.8894, "step": 4354 }, { "epoch": 1.017790891297362, "grad_norm": 3.034599781036377, "learning_rate": 2.9965656281521032e-05, "loss": 1.8445, "step": 4355 }, { "epoch": 1.0180245975869826, "grad_norm": 2.412224531173706, "learning_rate": 2.9965617869672454e-05, "loss": 1.8334, "step": 4356 }, { "epoch": 1.018258303876603, "grad_norm": 3.959059953689575, "learning_rate": 2.9965579436379596e-05, "loss": 2.1236, "step": 4357 }, { "epoch": 1.0184920101662236, "grad_norm": 3.8619253635406494, "learning_rate": 2.9965540981642514e-05, "loss": 1.8927, "step": 4358 }, { "epoch": 1.018725716455844, "grad_norm": 2.5891919136047363, "learning_rate": 2.9965502505461258e-05, "loss": 1.9311, "step": 4359 }, { "epoch": 1.0189594227454646, "grad_norm": 5.176835060119629, "learning_rate": 2.9965464007835896e-05, "loss": 1.9782, "step": 4360 }, { "epoch": 1.019193129035085, "grad_norm": 3.699603319168091, "learning_rate": 2.9965425488766472e-05, "loss": 1.9724, "step": 4361 }, { "epoch": 1.0194268353247056, "grad_norm": 3.191511392593384, "learning_rate": 2.9965386948253037e-05, "loss": 1.8451, "step": 4362 }, { "epoch": 1.0196605416143263, "grad_norm": 4.530267238616943, "learning_rate": 2.9965348386295658e-05, "loss": 1.7844, "step": 4363 }, { "epoch": 1.0198942479039468, "grad_norm": 6.4062957763671875, "learning_rate": 2.9965309802894383e-05, "loss": 1.8486, "step": 4364 }, { "epoch": 1.0201279541935673, "grad_norm": 4.7291975021362305, "learning_rate": 2.9965271198049273e-05, "loss": 1.7179, "step": 4365 }, { "epoch": 1.0203616604831878, "grad_norm": 2.9599907398223877, "learning_rate": 2.996523257176038e-05, "loss": 1.8084, "step": 4366 }, { "epoch": 1.0205953667728083, "grad_norm": 5.2306084632873535, "learning_rate": 2.9965193924027754e-05, "loss": 1.9779, "step": 4367 }, { "epoch": 1.0208290730624288, "grad_norm": 4.354454040527344, "learning_rate": 2.996515525485146e-05, "loss": 1.9879, "step": 4368 }, { "epoch": 1.0210627793520493, "grad_norm": 4.381011486053467, "learning_rate": 2.9965116564231546e-05, "loss": 2.0431, "step": 4369 }, { "epoch": 1.0212964856416697, "grad_norm": 2.386573553085327, "learning_rate": 2.996507785216807e-05, "loss": 1.8115, "step": 4370 }, { "epoch": 1.0215301919312902, "grad_norm": 4.798203468322754, "learning_rate": 2.9965039118661088e-05, "loss": 1.8242, "step": 4371 }, { "epoch": 1.021763898220911, "grad_norm": 5.627052307128906, "learning_rate": 2.9965000363710657e-05, "loss": 1.895, "step": 4372 }, { "epoch": 1.0219976045105315, "grad_norm": 3.078312397003174, "learning_rate": 2.9964961587316832e-05, "loss": 2.0973, "step": 4373 }, { "epoch": 1.022231310800152, "grad_norm": 3.632584810256958, "learning_rate": 2.996492278947966e-05, "loss": 1.8272, "step": 4374 }, { "epoch": 1.0224650170897724, "grad_norm": 2.566725730895996, "learning_rate": 2.996488397019921e-05, "loss": 2.0117, "step": 4375 }, { "epoch": 1.022698723379393, "grad_norm": 2.7417919635772705, "learning_rate": 2.9964845129475533e-05, "loss": 1.8926, "step": 4376 }, { "epoch": 1.0229324296690134, "grad_norm": 2.261195182800293, "learning_rate": 2.996480626730868e-05, "loss": 2.0603, "step": 4377 }, { "epoch": 1.023166135958634, "grad_norm": 3.263007164001465, "learning_rate": 2.996476738369871e-05, "loss": 2.0248, "step": 4378 }, { "epoch": 1.0233998422482544, "grad_norm": 2.464589834213257, "learning_rate": 2.9964728478645682e-05, "loss": 1.8139, "step": 4379 }, { "epoch": 1.023633548537875, "grad_norm": 2.327956199645996, "learning_rate": 2.9964689552149646e-05, "loss": 1.9316, "step": 4380 }, { "epoch": 1.0238672548274956, "grad_norm": 2.676375389099121, "learning_rate": 2.9964650604210656e-05, "loss": 1.8334, "step": 4381 }, { "epoch": 1.0241009611171161, "grad_norm": 2.2607250213623047, "learning_rate": 2.996461163482878e-05, "loss": 1.864, "step": 4382 }, { "epoch": 1.0243346674067366, "grad_norm": 2.537022590637207, "learning_rate": 2.996457264400406e-05, "loss": 1.8039, "step": 4383 }, { "epoch": 1.0245683736963571, "grad_norm": 2.7561235427856445, "learning_rate": 2.996453363173656e-05, "loss": 1.9029, "step": 4384 }, { "epoch": 1.0248020799859776, "grad_norm": 2.833418130874634, "learning_rate": 2.9964494598026333e-05, "loss": 1.8698, "step": 4385 }, { "epoch": 1.025035786275598, "grad_norm": 2.8359627723693848, "learning_rate": 2.9964455542873436e-05, "loss": 1.8423, "step": 4386 }, { "epoch": 1.0252694925652186, "grad_norm": 3.5730233192443848, "learning_rate": 2.9964416466277923e-05, "loss": 1.8877, "step": 4387 }, { "epoch": 1.025503198854839, "grad_norm": 3.3591468334198, "learning_rate": 2.9964377368239853e-05, "loss": 1.8974, "step": 4388 }, { "epoch": 1.0257369051444596, "grad_norm": 3.027245283126831, "learning_rate": 2.996433824875928e-05, "loss": 1.8372, "step": 4389 }, { "epoch": 1.0259706114340803, "grad_norm": 2.1382265090942383, "learning_rate": 2.996429910783626e-05, "loss": 1.7802, "step": 4390 }, { "epoch": 1.0262043177237008, "grad_norm": 2.5195977687835693, "learning_rate": 2.9964259945470852e-05, "loss": 1.7926, "step": 4391 }, { "epoch": 1.0264380240133213, "grad_norm": 2.781874656677246, "learning_rate": 2.9964220761663107e-05, "loss": 1.8705, "step": 4392 }, { "epoch": 1.0266717303029418, "grad_norm": 2.498100996017456, "learning_rate": 2.9964181556413085e-05, "loss": 1.9917, "step": 4393 }, { "epoch": 1.0269054365925623, "grad_norm": 2.6613688468933105, "learning_rate": 2.996414232972084e-05, "loss": 1.9858, "step": 4394 }, { "epoch": 1.0271391428821828, "grad_norm": 2.296081066131592, "learning_rate": 2.9964103081586432e-05, "loss": 1.9526, "step": 4395 }, { "epoch": 1.0273728491718033, "grad_norm": 3.1573212146759033, "learning_rate": 2.9964063812009912e-05, "loss": 1.8048, "step": 4396 }, { "epoch": 1.0276065554614238, "grad_norm": 2.6704745292663574, "learning_rate": 2.996402452099134e-05, "loss": 1.95, "step": 4397 }, { "epoch": 1.0278402617510445, "grad_norm": 3.364154577255249, "learning_rate": 2.996398520853077e-05, "loss": 1.8487, "step": 4398 }, { "epoch": 1.028073968040665, "grad_norm": 3.9360547065734863, "learning_rate": 2.996394587462826e-05, "loss": 1.9413, "step": 4399 }, { "epoch": 1.0283076743302855, "grad_norm": 3.35306715965271, "learning_rate": 2.9963906519283864e-05, "loss": 1.9182, "step": 4400 }, { "epoch": 1.028541380619906, "grad_norm": 3.02185320854187, "learning_rate": 2.9963867142497643e-05, "loss": 2.0364, "step": 4401 }, { "epoch": 1.0287750869095265, "grad_norm": 2.596798896789551, "learning_rate": 2.996382774426965e-05, "loss": 2.0727, "step": 4402 }, { "epoch": 1.029008793199147, "grad_norm": 3.3950917720794678, "learning_rate": 2.996378832459994e-05, "loss": 1.9778, "step": 4403 }, { "epoch": 1.0292424994887674, "grad_norm": 3.401742696762085, "learning_rate": 2.996374888348857e-05, "loss": 1.8336, "step": 4404 }, { "epoch": 1.029476205778388, "grad_norm": 3.381256580352783, "learning_rate": 2.9963709420935605e-05, "loss": 1.7896, "step": 4405 }, { "epoch": 1.0297099120680084, "grad_norm": 3.2461421489715576, "learning_rate": 2.9963669936941088e-05, "loss": 1.9008, "step": 4406 }, { "epoch": 1.0299436183576292, "grad_norm": 2.4780383110046387, "learning_rate": 2.9963630431505086e-05, "loss": 1.9609, "step": 4407 }, { "epoch": 1.0301773246472496, "grad_norm": 2.407440423965454, "learning_rate": 2.996359090462765e-05, "loss": 1.9469, "step": 4408 }, { "epoch": 1.0304110309368701, "grad_norm": 2.644472360610962, "learning_rate": 2.996355135630884e-05, "loss": 1.8426, "step": 4409 }, { "epoch": 1.0306447372264906, "grad_norm": 2.771419048309326, "learning_rate": 2.9963511786548712e-05, "loss": 1.8807, "step": 4410 }, { "epoch": 1.0308784435161111, "grad_norm": 3.762500047683716, "learning_rate": 2.9963472195347318e-05, "loss": 1.9111, "step": 4411 }, { "epoch": 1.0311121498057316, "grad_norm": 3.143662452697754, "learning_rate": 2.996343258270472e-05, "loss": 2.0023, "step": 4412 }, { "epoch": 1.0313458560953521, "grad_norm": 3.0872182846069336, "learning_rate": 2.9963392948620973e-05, "loss": 1.7903, "step": 4413 }, { "epoch": 1.0315795623849726, "grad_norm": 3.14103627204895, "learning_rate": 2.9963353293096137e-05, "loss": 1.9815, "step": 4414 }, { "epoch": 1.031813268674593, "grad_norm": 4.021389484405518, "learning_rate": 2.996331361613026e-05, "loss": 1.8019, "step": 4415 }, { "epoch": 1.0320469749642138, "grad_norm": 5.203418731689453, "learning_rate": 2.9963273917723408e-05, "loss": 1.8335, "step": 4416 }, { "epoch": 1.0322806812538343, "grad_norm": 2.6383724212646484, "learning_rate": 2.996323419787564e-05, "loss": 1.8803, "step": 4417 }, { "epoch": 1.0325143875434548, "grad_norm": 2.657977342605591, "learning_rate": 2.9963194456586996e-05, "loss": 2.0853, "step": 4418 }, { "epoch": 1.0327480938330753, "grad_norm": 3.154982089996338, "learning_rate": 2.9963154693857554e-05, "loss": 2.0423, "step": 4419 }, { "epoch": 1.0329818001226958, "grad_norm": 2.889148235321045, "learning_rate": 2.9963114909687357e-05, "loss": 1.9582, "step": 4420 }, { "epoch": 1.0332155064123163, "grad_norm": 3.227574348449707, "learning_rate": 2.996307510407647e-05, "loss": 1.8519, "step": 4421 }, { "epoch": 1.0334492127019368, "grad_norm": 2.529141426086426, "learning_rate": 2.9963035277024944e-05, "loss": 1.9466, "step": 4422 }, { "epoch": 1.0336829189915573, "grad_norm": 3.2485625743865967, "learning_rate": 2.9962995428532837e-05, "loss": 1.8291, "step": 4423 }, { "epoch": 1.0339166252811778, "grad_norm": 3.3081631660461426, "learning_rate": 2.9962955558600212e-05, "loss": 1.9149, "step": 4424 }, { "epoch": 1.0341503315707985, "grad_norm": 3.4621784687042236, "learning_rate": 2.9962915667227116e-05, "loss": 1.8481, "step": 4425 }, { "epoch": 1.034384037860419, "grad_norm": 4.099974155426025, "learning_rate": 2.9962875754413617e-05, "loss": 1.8542, "step": 4426 }, { "epoch": 1.0346177441500395, "grad_norm": 2.581453800201416, "learning_rate": 2.9962835820159763e-05, "loss": 2.0051, "step": 4427 }, { "epoch": 1.03485145043966, "grad_norm": 3.630143880844116, "learning_rate": 2.9962795864465617e-05, "loss": 1.9134, "step": 4428 }, { "epoch": 1.0350851567292805, "grad_norm": 3.3334953784942627, "learning_rate": 2.996275588733124e-05, "loss": 1.9275, "step": 4429 }, { "epoch": 1.035318863018901, "grad_norm": 2.478864908218384, "learning_rate": 2.9962715888756677e-05, "loss": 1.8589, "step": 4430 }, { "epoch": 1.0355525693085215, "grad_norm": 5.6558966636657715, "learning_rate": 2.9962675868741994e-05, "loss": 1.8518, "step": 4431 }, { "epoch": 1.035786275598142, "grad_norm": 5.447789192199707, "learning_rate": 2.9962635827287246e-05, "loss": 1.8677, "step": 4432 }, { "epoch": 1.0360199818877625, "grad_norm": 3.05165696144104, "learning_rate": 2.996259576439249e-05, "loss": 1.9228, "step": 4433 }, { "epoch": 1.0362536881773832, "grad_norm": 4.450710296630859, "learning_rate": 2.996255568005779e-05, "loss": 1.7619, "step": 4434 }, { "epoch": 1.0364873944670037, "grad_norm": 3.7927234172821045, "learning_rate": 2.9962515574283193e-05, "loss": 2.0387, "step": 4435 }, { "epoch": 1.0367211007566242, "grad_norm": 3.470546007156372, "learning_rate": 2.9962475447068765e-05, "loss": 1.8698, "step": 4436 }, { "epoch": 1.0369548070462447, "grad_norm": 2.819930076599121, "learning_rate": 2.9962435298414554e-05, "loss": 2.0122, "step": 4437 }, { "epoch": 1.0371885133358651, "grad_norm": 4.722240447998047, "learning_rate": 2.996239512832063e-05, "loss": 1.9291, "step": 4438 }, { "epoch": 1.0374222196254856, "grad_norm": 3.609114646911621, "learning_rate": 2.996235493678704e-05, "loss": 1.7659, "step": 4439 }, { "epoch": 1.0376559259151061, "grad_norm": 2.9713761806488037, "learning_rate": 2.9962314723813842e-05, "loss": 1.9534, "step": 4440 }, { "epoch": 1.0378896322047266, "grad_norm": 3.394681215286255, "learning_rate": 2.99622744894011e-05, "loss": 2.0002, "step": 4441 }, { "epoch": 1.0381233384943471, "grad_norm": 4.3981733322143555, "learning_rate": 2.9962234233548873e-05, "loss": 1.7936, "step": 4442 }, { "epoch": 1.0383570447839678, "grad_norm": 2.736701011657715, "learning_rate": 2.9962193956257206e-05, "loss": 1.8141, "step": 4443 }, { "epoch": 1.0385907510735883, "grad_norm": 3.838289260864258, "learning_rate": 2.9962153657526175e-05, "loss": 1.8593, "step": 4444 }, { "epoch": 1.0388244573632088, "grad_norm": 4.222703456878662, "learning_rate": 2.996211333735582e-05, "loss": 1.992, "step": 4445 }, { "epoch": 1.0390581636528293, "grad_norm": 2.845642566680908, "learning_rate": 2.9962072995746213e-05, "loss": 1.9915, "step": 4446 }, { "epoch": 1.0392918699424498, "grad_norm": 2.927908182144165, "learning_rate": 2.99620326326974e-05, "loss": 1.7639, "step": 4447 }, { "epoch": 1.0395255762320703, "grad_norm": 5.225662708282471, "learning_rate": 2.9961992248209446e-05, "loss": 1.9105, "step": 4448 }, { "epoch": 1.0397592825216908, "grad_norm": 3.1211490631103516, "learning_rate": 2.9961951842282405e-05, "loss": 2.1194, "step": 4449 }, { "epoch": 1.0399929888113113, "grad_norm": 2.9994776248931885, "learning_rate": 2.996191141491634e-05, "loss": 1.8257, "step": 4450 }, { "epoch": 1.0402266951009318, "grad_norm": 3.103757381439209, "learning_rate": 2.9961870966111305e-05, "loss": 2.0413, "step": 4451 }, { "epoch": 1.0404604013905525, "grad_norm": 3.0422732830047607, "learning_rate": 2.9961830495867362e-05, "loss": 1.9155, "step": 4452 }, { "epoch": 1.040694107680173, "grad_norm": 2.5649514198303223, "learning_rate": 2.996179000418456e-05, "loss": 2.0466, "step": 4453 }, { "epoch": 1.0409278139697935, "grad_norm": 6.063696384429932, "learning_rate": 2.9961749491062966e-05, "loss": 1.7967, "step": 4454 }, { "epoch": 1.041161520259414, "grad_norm": 5.453652381896973, "learning_rate": 2.9961708956502635e-05, "loss": 1.9066, "step": 4455 }, { "epoch": 1.0413952265490345, "grad_norm": 2.4096693992614746, "learning_rate": 2.9961668400503622e-05, "loss": 1.9192, "step": 4456 }, { "epoch": 1.041628932838655, "grad_norm": 3.5683999061584473, "learning_rate": 2.9961627823065993e-05, "loss": 1.7666, "step": 4457 }, { "epoch": 1.0418626391282755, "grad_norm": 36.624244689941406, "learning_rate": 2.99615872241898e-05, "loss": 2.721, "step": 4458 }, { "epoch": 1.042096345417896, "grad_norm": 5.053011894226074, "learning_rate": 2.99615466038751e-05, "loss": 1.9247, "step": 4459 }, { "epoch": 1.0423300517075167, "grad_norm": 5.275794506072998, "learning_rate": 2.9961505962121958e-05, "loss": 2.022, "step": 4460 }, { "epoch": 1.0425637579971372, "grad_norm": 2.9678308963775635, "learning_rate": 2.9961465298930424e-05, "loss": 1.9023, "step": 4461 }, { "epoch": 1.0427974642867577, "grad_norm": 2.7753279209136963, "learning_rate": 2.9961424614300565e-05, "loss": 1.8386, "step": 4462 }, { "epoch": 1.0430311705763782, "grad_norm": 2.317884922027588, "learning_rate": 2.996138390823243e-05, "loss": 1.7559, "step": 4463 }, { "epoch": 1.0432648768659987, "grad_norm": 3.3295586109161377, "learning_rate": 2.9961343180726082e-05, "loss": 1.9193, "step": 4464 }, { "epoch": 1.0434985831556192, "grad_norm": 2.681025743484497, "learning_rate": 2.9961302431781583e-05, "loss": 1.9948, "step": 4465 }, { "epoch": 1.0437322894452397, "grad_norm": 3.064382791519165, "learning_rate": 2.9961261661398983e-05, "loss": 1.7808, "step": 4466 }, { "epoch": 1.0439659957348602, "grad_norm": 4.727543354034424, "learning_rate": 2.9961220869578344e-05, "loss": 1.8582, "step": 4467 }, { "epoch": 1.0441997020244806, "grad_norm": 3.097378730773926, "learning_rate": 2.996118005631973e-05, "loss": 1.9045, "step": 4468 }, { "epoch": 1.0444334083141014, "grad_norm": 3.020700693130493, "learning_rate": 2.9961139221623196e-05, "loss": 1.6974, "step": 4469 }, { "epoch": 1.0446671146037219, "grad_norm": 5.246773719787598, "learning_rate": 2.9961098365488792e-05, "loss": 1.9024, "step": 4470 }, { "epoch": 1.0449008208933424, "grad_norm": 3.0008254051208496, "learning_rate": 2.996105748791659e-05, "loss": 2.0027, "step": 4471 }, { "epoch": 1.0451345271829628, "grad_norm": 2.3152554035186768, "learning_rate": 2.996101658890664e-05, "loss": 1.9665, "step": 4472 }, { "epoch": 1.0453682334725833, "grad_norm": 2.7834036350250244, "learning_rate": 2.9960975668459007e-05, "loss": 1.9747, "step": 4473 }, { "epoch": 1.0456019397622038, "grad_norm": 2.831861972808838, "learning_rate": 2.9960934726573742e-05, "loss": 2.0055, "step": 4474 }, { "epoch": 1.0458356460518243, "grad_norm": 2.2264328002929688, "learning_rate": 2.9960893763250906e-05, "loss": 1.9449, "step": 4475 }, { "epoch": 1.0460693523414448, "grad_norm": 2.7473554611206055, "learning_rate": 2.9960852778490562e-05, "loss": 1.9422, "step": 4476 }, { "epoch": 1.0463030586310653, "grad_norm": 2.337923765182495, "learning_rate": 2.996081177229277e-05, "loss": 1.9054, "step": 4477 }, { "epoch": 1.046536764920686, "grad_norm": 3.116574287414551, "learning_rate": 2.9960770744657575e-05, "loss": 1.936, "step": 4478 }, { "epoch": 1.0467704712103065, "grad_norm": 2.910395622253418, "learning_rate": 2.9960729695585054e-05, "loss": 1.8066, "step": 4479 }, { "epoch": 1.047004177499927, "grad_norm": 2.3942487239837646, "learning_rate": 2.9960688625075254e-05, "loss": 1.8487, "step": 4480 }, { "epoch": 1.0472378837895475, "grad_norm": 2.40763783454895, "learning_rate": 2.9960647533128237e-05, "loss": 1.9051, "step": 4481 }, { "epoch": 1.047471590079168, "grad_norm": 3.1842336654663086, "learning_rate": 2.996060641974406e-05, "loss": 1.8993, "step": 4482 }, { "epoch": 1.0477052963687885, "grad_norm": 2.8916139602661133, "learning_rate": 2.9960565284922785e-05, "loss": 1.9511, "step": 4483 }, { "epoch": 1.047939002658409, "grad_norm": 3.1731529235839844, "learning_rate": 2.9960524128664472e-05, "loss": 1.9346, "step": 4484 }, { "epoch": 1.0481727089480295, "grad_norm": 3.5428454875946045, "learning_rate": 2.9960482950969175e-05, "loss": 1.9174, "step": 4485 }, { "epoch": 1.04840641523765, "grad_norm": 2.385262966156006, "learning_rate": 2.9960441751836958e-05, "loss": 1.8719, "step": 4486 }, { "epoch": 1.0486401215272707, "grad_norm": 2.7694976329803467, "learning_rate": 2.996040053126788e-05, "loss": 1.8572, "step": 4487 }, { "epoch": 1.0488738278168912, "grad_norm": 4.135123252868652, "learning_rate": 2.9960359289261995e-05, "loss": 1.9237, "step": 4488 }, { "epoch": 1.0491075341065117, "grad_norm": 2.6089892387390137, "learning_rate": 2.9960318025819364e-05, "loss": 1.8746, "step": 4489 }, { "epoch": 1.0493412403961322, "grad_norm": 3.375791549682617, "learning_rate": 2.9960276740940047e-05, "loss": 1.9213, "step": 4490 }, { "epoch": 1.0495749466857527, "grad_norm": 5.632355213165283, "learning_rate": 2.9960235434624102e-05, "loss": 1.8828, "step": 4491 }, { "epoch": 1.0498086529753732, "grad_norm": 3.0932888984680176, "learning_rate": 2.9960194106871596e-05, "loss": 2.0117, "step": 4492 }, { "epoch": 1.0500423592649937, "grad_norm": 2.8003971576690674, "learning_rate": 2.9960152757682576e-05, "loss": 2.0233, "step": 4493 }, { "epoch": 1.0502760655546142, "grad_norm": 3.3662431240081787, "learning_rate": 2.9960111387057107e-05, "loss": 1.9218, "step": 4494 }, { "epoch": 1.0505097718442347, "grad_norm": 3.5001189708709717, "learning_rate": 2.996006999499525e-05, "loss": 1.8344, "step": 4495 }, { "epoch": 1.0507434781338554, "grad_norm": 3.335792303085327, "learning_rate": 2.996002858149706e-05, "loss": 2.0101, "step": 4496 }, { "epoch": 1.0509771844234759, "grad_norm": 3.5623459815979004, "learning_rate": 2.9959987146562604e-05, "loss": 2.0482, "step": 4497 }, { "epoch": 1.0512108907130964, "grad_norm": 2.5307116508483887, "learning_rate": 2.995994569019193e-05, "loss": 1.7988, "step": 4498 }, { "epoch": 1.0514445970027169, "grad_norm": 3.126789093017578, "learning_rate": 2.9959904212385107e-05, "loss": 1.947, "step": 4499 }, { "epoch": 1.0516783032923374, "grad_norm": 3.737931966781616, "learning_rate": 2.9959862713142192e-05, "loss": 1.9063, "step": 4500 }, { "epoch": 1.0519120095819579, "grad_norm": 2.6783738136291504, "learning_rate": 2.995982119246324e-05, "loss": 1.8457, "step": 4501 }, { "epoch": 1.0521457158715783, "grad_norm": 3.267517328262329, "learning_rate": 2.995977965034832e-05, "loss": 1.9048, "step": 4502 }, { "epoch": 1.0523794221611988, "grad_norm": 3.6487228870391846, "learning_rate": 2.9959738086797483e-05, "loss": 1.8085, "step": 4503 }, { "epoch": 1.0526131284508193, "grad_norm": 2.5676817893981934, "learning_rate": 2.995969650181079e-05, "loss": 1.9814, "step": 4504 }, { "epoch": 1.05284683474044, "grad_norm": 3.8639187812805176, "learning_rate": 2.99596548953883e-05, "loss": 1.9392, "step": 4505 }, { "epoch": 1.0530805410300605, "grad_norm": 3.5747549533843994, "learning_rate": 2.9959613267530077e-05, "loss": 1.8558, "step": 4506 }, { "epoch": 1.053314247319681, "grad_norm": 2.4773201942443848, "learning_rate": 2.9959571618236176e-05, "loss": 2.0244, "step": 4507 }, { "epoch": 1.0535479536093015, "grad_norm": 2.844351291656494, "learning_rate": 2.9959529947506663e-05, "loss": 1.9234, "step": 4508 }, { "epoch": 1.053781659898922, "grad_norm": 5.53937292098999, "learning_rate": 2.995948825534159e-05, "loss": 1.9221, "step": 4509 }, { "epoch": 1.0540153661885425, "grad_norm": 4.24689245223999, "learning_rate": 2.9959446541741018e-05, "loss": 1.9851, "step": 4510 }, { "epoch": 1.054249072478163, "grad_norm": 2.61197566986084, "learning_rate": 2.995940480670501e-05, "loss": 1.9497, "step": 4511 }, { "epoch": 1.0544827787677835, "grad_norm": 4.750768184661865, "learning_rate": 2.9959363050233626e-05, "loss": 1.9801, "step": 4512 }, { "epoch": 1.054716485057404, "grad_norm": 2.879762649536133, "learning_rate": 2.9959321272326927e-05, "loss": 2.0037, "step": 4513 }, { "epoch": 1.0549501913470247, "grad_norm": 2.421328067779541, "learning_rate": 2.9959279472984965e-05, "loss": 1.8422, "step": 4514 }, { "epoch": 1.0551838976366452, "grad_norm": 2.141350507736206, "learning_rate": 2.995923765220781e-05, "loss": 1.8653, "step": 4515 }, { "epoch": 1.0554176039262657, "grad_norm": 3.466594696044922, "learning_rate": 2.9959195809995516e-05, "loss": 1.8971, "step": 4516 }, { "epoch": 1.0556513102158862, "grad_norm": 2.480499505996704, "learning_rate": 2.9959153946348144e-05, "loss": 1.9555, "step": 4517 }, { "epoch": 1.0558850165055067, "grad_norm": 2.8998985290527344, "learning_rate": 2.9959112061265753e-05, "loss": 1.9801, "step": 4518 }, { "epoch": 1.0561187227951272, "grad_norm": 2.8063039779663086, "learning_rate": 2.9959070154748407e-05, "loss": 1.8285, "step": 4519 }, { "epoch": 1.0563524290847477, "grad_norm": 2.9730048179626465, "learning_rate": 2.995902822679616e-05, "loss": 1.7697, "step": 4520 }, { "epoch": 1.0565861353743682, "grad_norm": 3.021963596343994, "learning_rate": 2.9958986277409074e-05, "loss": 1.887, "step": 4521 }, { "epoch": 1.0568198416639887, "grad_norm": 2.8356621265411377, "learning_rate": 2.9958944306587213e-05, "loss": 1.8087, "step": 4522 }, { "epoch": 1.0570535479536094, "grad_norm": 3.527040719985962, "learning_rate": 2.9958902314330634e-05, "loss": 1.8185, "step": 4523 }, { "epoch": 1.05728725424323, "grad_norm": 2.717069625854492, "learning_rate": 2.99588603006394e-05, "loss": 2.1134, "step": 4524 }, { "epoch": 1.0575209605328504, "grad_norm": 2.5559256076812744, "learning_rate": 2.9958818265513564e-05, "loss": 1.8568, "step": 4525 }, { "epoch": 1.0577546668224709, "grad_norm": 2.380368232727051, "learning_rate": 2.99587762089532e-05, "loss": 1.8464, "step": 4526 }, { "epoch": 1.0579883731120914, "grad_norm": 3.940307140350342, "learning_rate": 2.995873413095835e-05, "loss": 1.9295, "step": 4527 }, { "epoch": 1.0582220794017119, "grad_norm": 3.2397661209106445, "learning_rate": 2.9958692031529087e-05, "loss": 1.8455, "step": 4528 }, { "epoch": 1.0584557856913324, "grad_norm": 2.4018301963806152, "learning_rate": 2.9958649910665466e-05, "loss": 1.8565, "step": 4529 }, { "epoch": 1.0586894919809529, "grad_norm": 3.17878794670105, "learning_rate": 2.995860776836755e-05, "loss": 2.0241, "step": 4530 }, { "epoch": 1.0589231982705734, "grad_norm": 3.0975823402404785, "learning_rate": 2.9958565604635403e-05, "loss": 1.7968, "step": 4531 }, { "epoch": 1.059156904560194, "grad_norm": 2.5414719581604004, "learning_rate": 2.995852341946908e-05, "loss": 1.7596, "step": 4532 }, { "epoch": 1.0593906108498146, "grad_norm": 3.454547882080078, "learning_rate": 2.995848121286864e-05, "loss": 1.9083, "step": 4533 }, { "epoch": 1.059624317139435, "grad_norm": 2.6060633659362793, "learning_rate": 2.9958438984834147e-05, "loss": 1.7849, "step": 4534 }, { "epoch": 1.0598580234290556, "grad_norm": 2.7905819416046143, "learning_rate": 2.995839673536566e-05, "loss": 2.0754, "step": 4535 }, { "epoch": 1.060091729718676, "grad_norm": 2.742567300796509, "learning_rate": 2.995835446446324e-05, "loss": 1.9793, "step": 4536 }, { "epoch": 1.0603254360082965, "grad_norm": 2.663233518600464, "learning_rate": 2.9958312172126948e-05, "loss": 1.7654, "step": 4537 }, { "epoch": 1.060559142297917, "grad_norm": 2.4854443073272705, "learning_rate": 2.9958269858356845e-05, "loss": 1.8802, "step": 4538 }, { "epoch": 1.0607928485875375, "grad_norm": 3.89007830619812, "learning_rate": 2.995822752315299e-05, "loss": 1.8608, "step": 4539 }, { "epoch": 1.061026554877158, "grad_norm": 3.796112060546875, "learning_rate": 2.9958185166515445e-05, "loss": 1.8356, "step": 4540 }, { "epoch": 1.0612602611667787, "grad_norm": 2.919677972793579, "learning_rate": 2.9958142788444273e-05, "loss": 1.8491, "step": 4541 }, { "epoch": 1.0614939674563992, "grad_norm": 3.1313343048095703, "learning_rate": 2.995810038893953e-05, "loss": 2.0122, "step": 4542 }, { "epoch": 1.0617276737460197, "grad_norm": 3.404546022415161, "learning_rate": 2.995805796800128e-05, "loss": 1.9861, "step": 4543 }, { "epoch": 1.0619613800356402, "grad_norm": 2.3897037506103516, "learning_rate": 2.995801552562958e-05, "loss": 1.8637, "step": 4544 }, { "epoch": 1.0621950863252607, "grad_norm": 3.461196184158325, "learning_rate": 2.9957973061824494e-05, "loss": 2.0044, "step": 4545 }, { "epoch": 1.0624287926148812, "grad_norm": 2.7528698444366455, "learning_rate": 2.995793057658608e-05, "loss": 1.9428, "step": 4546 }, { "epoch": 1.0626624989045017, "grad_norm": 2.561742067337036, "learning_rate": 2.9957888069914405e-05, "loss": 1.9341, "step": 4547 }, { "epoch": 1.0628962051941222, "grad_norm": 3.2076332569122314, "learning_rate": 2.9957845541809527e-05, "loss": 2.0455, "step": 4548 }, { "epoch": 1.0631299114837427, "grad_norm": 2.9265949726104736, "learning_rate": 2.99578029922715e-05, "loss": 2.0027, "step": 4549 }, { "epoch": 1.0633636177733634, "grad_norm": 2.5865817070007324, "learning_rate": 2.99577604213004e-05, "loss": 1.8638, "step": 4550 }, { "epoch": 1.063597324062984, "grad_norm": 2.358288049697876, "learning_rate": 2.995771782889627e-05, "loss": 1.8575, "step": 4551 }, { "epoch": 1.0638310303526044, "grad_norm": 2.8411056995391846, "learning_rate": 2.9957675215059183e-05, "loss": 1.8722, "step": 4552 }, { "epoch": 1.064064736642225, "grad_norm": 4.146001815795898, "learning_rate": 2.99576325797892e-05, "loss": 1.8641, "step": 4553 }, { "epoch": 1.0642984429318454, "grad_norm": 3.098193645477295, "learning_rate": 2.9957589923086375e-05, "loss": 1.7802, "step": 4554 }, { "epoch": 1.0645321492214659, "grad_norm": 2.792424440383911, "learning_rate": 2.9957547244950772e-05, "loss": 1.9557, "step": 4555 }, { "epoch": 1.0647658555110864, "grad_norm": 2.7452125549316406, "learning_rate": 2.9957504545382458e-05, "loss": 1.8872, "step": 4556 }, { "epoch": 1.0649995618007069, "grad_norm": 2.3016605377197266, "learning_rate": 2.995746182438149e-05, "loss": 2.0366, "step": 4557 }, { "epoch": 1.0652332680903274, "grad_norm": 2.8408803939819336, "learning_rate": 2.9957419081947928e-05, "loss": 1.9439, "step": 4558 }, { "epoch": 1.065466974379948, "grad_norm": 2.4430935382843018, "learning_rate": 2.995737631808183e-05, "loss": 2.0326, "step": 4559 }, { "epoch": 1.0657006806695686, "grad_norm": 2.9466447830200195, "learning_rate": 2.9957333532783265e-05, "loss": 1.9151, "step": 4560 }, { "epoch": 1.065934386959189, "grad_norm": 2.9197893142700195, "learning_rate": 2.995729072605229e-05, "loss": 2.0256, "step": 4561 }, { "epoch": 1.0661680932488096, "grad_norm": 4.063087463378906, "learning_rate": 2.995724789788897e-05, "loss": 2.0379, "step": 4562 }, { "epoch": 1.06640179953843, "grad_norm": 3.944281816482544, "learning_rate": 2.9957205048293355e-05, "loss": 1.8577, "step": 4563 }, { "epoch": 1.0666355058280506, "grad_norm": 3.3508384227752686, "learning_rate": 2.995716217726552e-05, "loss": 1.9892, "step": 4564 }, { "epoch": 1.066869212117671, "grad_norm": 3.5753870010375977, "learning_rate": 2.995711928480552e-05, "loss": 1.8261, "step": 4565 }, { "epoch": 1.0671029184072915, "grad_norm": 4.820699691772461, "learning_rate": 2.995707637091342e-05, "loss": 1.9926, "step": 4566 }, { "epoch": 1.067336624696912, "grad_norm": 2.9436779022216797, "learning_rate": 2.9957033435589276e-05, "loss": 1.7961, "step": 4567 }, { "epoch": 1.0675703309865328, "grad_norm": 3.8054256439208984, "learning_rate": 2.9956990478833157e-05, "loss": 1.7996, "step": 4568 }, { "epoch": 1.0678040372761533, "grad_norm": 2.92482852935791, "learning_rate": 2.9956947500645115e-05, "loss": 1.9507, "step": 4569 }, { "epoch": 1.0680377435657737, "grad_norm": 2.862746477127075, "learning_rate": 2.995690450102522e-05, "loss": 2.0098, "step": 4570 }, { "epoch": 1.0682714498553942, "grad_norm": 3.392899513244629, "learning_rate": 2.9956861479973528e-05, "loss": 1.8185, "step": 4571 }, { "epoch": 1.0685051561450147, "grad_norm": 2.3941702842712402, "learning_rate": 2.9956818437490105e-05, "loss": 1.8104, "step": 4572 }, { "epoch": 1.0687388624346352, "grad_norm": 2.191145181655884, "learning_rate": 2.995677537357501e-05, "loss": 1.8527, "step": 4573 }, { "epoch": 1.0689725687242557, "grad_norm": 2.3247289657592773, "learning_rate": 2.995673228822831e-05, "loss": 1.789, "step": 4574 }, { "epoch": 1.0692062750138762, "grad_norm": 2.3848659992218018, "learning_rate": 2.9956689181450053e-05, "loss": 1.6934, "step": 4575 }, { "epoch": 1.069439981303497, "grad_norm": 2.487088680267334, "learning_rate": 2.9956646053240317e-05, "loss": 1.9736, "step": 4576 }, { "epoch": 1.0696736875931174, "grad_norm": 2.7963240146636963, "learning_rate": 2.9956602903599155e-05, "loss": 2.0283, "step": 4577 }, { "epoch": 1.069907393882738, "grad_norm": 3.1603808403015137, "learning_rate": 2.9956559732526627e-05, "loss": 2.1406, "step": 4578 }, { "epoch": 1.0701411001723584, "grad_norm": 2.5174403190612793, "learning_rate": 2.9956516540022804e-05, "loss": 1.9609, "step": 4579 }, { "epoch": 1.070374806461979, "grad_norm": 3.351663112640381, "learning_rate": 2.9956473326087735e-05, "loss": 2.0008, "step": 4580 }, { "epoch": 1.0706085127515994, "grad_norm": 4.363175392150879, "learning_rate": 2.9956430090721496e-05, "loss": 1.9315, "step": 4581 }, { "epoch": 1.07084221904122, "grad_norm": 3.0196597576141357, "learning_rate": 2.9956386833924142e-05, "loss": 1.8318, "step": 4582 }, { "epoch": 1.0710759253308404, "grad_norm": 4.107675075531006, "learning_rate": 2.9956343555695732e-05, "loss": 1.9273, "step": 4583 }, { "epoch": 1.071309631620461, "grad_norm": 4.857968807220459, "learning_rate": 2.995630025603633e-05, "loss": 1.9295, "step": 4584 }, { "epoch": 1.0715433379100816, "grad_norm": 3.8214271068573, "learning_rate": 2.9956256934946005e-05, "loss": 1.8201, "step": 4585 }, { "epoch": 1.071777044199702, "grad_norm": 2.9964864253997803, "learning_rate": 2.995621359242481e-05, "loss": 1.9637, "step": 4586 }, { "epoch": 1.0720107504893226, "grad_norm": 3.825049638748169, "learning_rate": 2.995617022847281e-05, "loss": 1.9837, "step": 4587 }, { "epoch": 1.072244456778943, "grad_norm": 2.1906113624572754, "learning_rate": 2.9956126843090067e-05, "loss": 1.9209, "step": 4588 }, { "epoch": 1.0724781630685636, "grad_norm": 2.648927927017212, "learning_rate": 2.9956083436276645e-05, "loss": 1.9171, "step": 4589 }, { "epoch": 1.072711869358184, "grad_norm": 2.735002279281616, "learning_rate": 2.9956040008032602e-05, "loss": 1.8258, "step": 4590 }, { "epoch": 1.0729455756478046, "grad_norm": 2.2884180545806885, "learning_rate": 2.9955996558358008e-05, "loss": 1.7956, "step": 4591 }, { "epoch": 1.073179281937425, "grad_norm": 3.0246829986572266, "learning_rate": 2.9955953087252915e-05, "loss": 1.9216, "step": 4592 }, { "epoch": 1.0734129882270456, "grad_norm": 2.409738779067993, "learning_rate": 2.995590959471739e-05, "loss": 1.9923, "step": 4593 }, { "epoch": 1.0736466945166663, "grad_norm": 2.8427703380584717, "learning_rate": 2.9955866080751502e-05, "loss": 1.8992, "step": 4594 }, { "epoch": 1.0738804008062868, "grad_norm": 2.788825273513794, "learning_rate": 2.9955822545355308e-05, "loss": 1.9795, "step": 4595 }, { "epoch": 1.0741141070959073, "grad_norm": 2.7329795360565186, "learning_rate": 2.9955778988528864e-05, "loss": 1.922, "step": 4596 }, { "epoch": 1.0743478133855278, "grad_norm": 2.5364017486572266, "learning_rate": 2.9955735410272237e-05, "loss": 1.9309, "step": 4597 }, { "epoch": 1.0745815196751483, "grad_norm": 3.3056139945983887, "learning_rate": 2.9955691810585496e-05, "loss": 1.7961, "step": 4598 }, { "epoch": 1.0748152259647688, "grad_norm": 2.144900321960449, "learning_rate": 2.99556481894687e-05, "loss": 1.9189, "step": 4599 }, { "epoch": 1.0750489322543892, "grad_norm": 2.580411911010742, "learning_rate": 2.9955604546921902e-05, "loss": 2.0291, "step": 4600 }, { "epoch": 1.0752826385440097, "grad_norm": 4.3442840576171875, "learning_rate": 2.9955560882945175e-05, "loss": 1.8489, "step": 4601 }, { "epoch": 1.0755163448336302, "grad_norm": 2.7355875968933105, "learning_rate": 2.995551719753858e-05, "loss": 1.9528, "step": 4602 }, { "epoch": 1.075750051123251, "grad_norm": 2.3350768089294434, "learning_rate": 2.995547349070218e-05, "loss": 1.8023, "step": 4603 }, { "epoch": 1.0759837574128714, "grad_norm": 4.360716342926025, "learning_rate": 2.995542976243603e-05, "loss": 1.8244, "step": 4604 }, { "epoch": 1.076217463702492, "grad_norm": 2.711874485015869, "learning_rate": 2.9955386012740205e-05, "loss": 1.7882, "step": 4605 }, { "epoch": 1.0764511699921124, "grad_norm": 2.7198173999786377, "learning_rate": 2.9955342241614758e-05, "loss": 1.927, "step": 4606 }, { "epoch": 1.076684876281733, "grad_norm": 3.7226874828338623, "learning_rate": 2.9955298449059754e-05, "loss": 1.9142, "step": 4607 }, { "epoch": 1.0769185825713534, "grad_norm": 3.942476749420166, "learning_rate": 2.9955254635075256e-05, "loss": 1.973, "step": 4608 }, { "epoch": 1.077152288860974, "grad_norm": 2.463541269302368, "learning_rate": 2.9955210799661334e-05, "loss": 2.1254, "step": 4609 }, { "epoch": 1.0773859951505944, "grad_norm": 2.996103525161743, "learning_rate": 2.9955166942818036e-05, "loss": 1.9887, "step": 4610 }, { "epoch": 1.0776197014402151, "grad_norm": 3.0181984901428223, "learning_rate": 2.995512306454544e-05, "loss": 1.7984, "step": 4611 }, { "epoch": 1.0778534077298356, "grad_norm": 2.892425060272217, "learning_rate": 2.9955079164843596e-05, "loss": 1.8507, "step": 4612 }, { "epoch": 1.0780871140194561, "grad_norm": 3.6065874099731445, "learning_rate": 2.9955035243712574e-05, "loss": 1.7435, "step": 4613 }, { "epoch": 1.0783208203090766, "grad_norm": 2.6435041427612305, "learning_rate": 2.995499130115244e-05, "loss": 2.0578, "step": 4614 }, { "epoch": 1.078554526598697, "grad_norm": 2.570807456970215, "learning_rate": 2.995494733716325e-05, "loss": 1.7701, "step": 4615 }, { "epoch": 1.0787882328883176, "grad_norm": 3.2761316299438477, "learning_rate": 2.995490335174507e-05, "loss": 1.8251, "step": 4616 }, { "epoch": 1.079021939177938, "grad_norm": 2.6309802532196045, "learning_rate": 2.9954859344897963e-05, "loss": 1.7121, "step": 4617 }, { "epoch": 1.0792556454675586, "grad_norm": 2.521202325820923, "learning_rate": 2.995481531662199e-05, "loss": 1.7476, "step": 4618 }, { "epoch": 1.079489351757179, "grad_norm": 2.4507012367248535, "learning_rate": 2.995477126691722e-05, "loss": 2.0086, "step": 4619 }, { "epoch": 1.0797230580467998, "grad_norm": 2.685401439666748, "learning_rate": 2.9954727195783707e-05, "loss": 1.9083, "step": 4620 }, { "epoch": 1.0799567643364203, "grad_norm": 2.608172655105591, "learning_rate": 2.9954683103221525e-05, "loss": 1.9043, "step": 4621 }, { "epoch": 1.0801904706260408, "grad_norm": 2.719574213027954, "learning_rate": 2.9954638989230728e-05, "loss": 1.8646, "step": 4622 }, { "epoch": 1.0804241769156613, "grad_norm": 3.560607671737671, "learning_rate": 2.9954594853811382e-05, "loss": 1.8544, "step": 4623 }, { "epoch": 1.0806578832052818, "grad_norm": 2.3234574794769287, "learning_rate": 2.9954550696963554e-05, "loss": 2.0591, "step": 4624 }, { "epoch": 1.0808915894949023, "grad_norm": 3.12205171585083, "learning_rate": 2.99545065186873e-05, "loss": 1.9901, "step": 4625 }, { "epoch": 1.0811252957845228, "grad_norm": 3.4650542736053467, "learning_rate": 2.995446231898269e-05, "loss": 2.1017, "step": 4626 }, { "epoch": 1.0813590020741433, "grad_norm": 2.4943344593048096, "learning_rate": 2.9954418097849786e-05, "loss": 1.9725, "step": 4627 }, { "epoch": 1.0815927083637638, "grad_norm": 2.737377643585205, "learning_rate": 2.9954373855288647e-05, "loss": 1.8548, "step": 4628 }, { "epoch": 1.0818264146533845, "grad_norm": 4.587695121765137, "learning_rate": 2.9954329591299343e-05, "loss": 2.0341, "step": 4629 }, { "epoch": 1.082060120943005, "grad_norm": 3.8532135486602783, "learning_rate": 2.995428530588193e-05, "loss": 1.905, "step": 4630 }, { "epoch": 1.0822938272326255, "grad_norm": 2.437424421310425, "learning_rate": 2.995424099903648e-05, "loss": 1.8534, "step": 4631 }, { "epoch": 1.082527533522246, "grad_norm": 5.602038860321045, "learning_rate": 2.9954196670763047e-05, "loss": 1.8559, "step": 4632 }, { "epoch": 1.0827612398118664, "grad_norm": 6.427145481109619, "learning_rate": 2.9954152321061703e-05, "loss": 1.8348, "step": 4633 }, { "epoch": 1.082994946101487, "grad_norm": 3.421205520629883, "learning_rate": 2.9954107949932506e-05, "loss": 1.8586, "step": 4634 }, { "epoch": 1.0832286523911074, "grad_norm": 3.291947841644287, "learning_rate": 2.995406355737552e-05, "loss": 1.9597, "step": 4635 }, { "epoch": 1.083462358680728, "grad_norm": 4.382709980010986, "learning_rate": 2.995401914339082e-05, "loss": 1.8473, "step": 4636 }, { "epoch": 1.0836960649703484, "grad_norm": 3.735769271850586, "learning_rate": 2.995397470797845e-05, "loss": 1.8686, "step": 4637 }, { "epoch": 1.0839297712599691, "grad_norm": 2.5964274406433105, "learning_rate": 2.9953930251138486e-05, "loss": 1.8974, "step": 4638 }, { "epoch": 1.0841634775495896, "grad_norm": 5.170013904571533, "learning_rate": 2.9953885772870988e-05, "loss": 2.0581, "step": 4639 }, { "epoch": 1.0843971838392101, "grad_norm": 5.322552680969238, "learning_rate": 2.995384127317602e-05, "loss": 1.8628, "step": 4640 }, { "epoch": 1.0846308901288306, "grad_norm": 2.9637932777404785, "learning_rate": 2.995379675205365e-05, "loss": 1.8747, "step": 4641 }, { "epoch": 1.0848645964184511, "grad_norm": 5.284841537475586, "learning_rate": 2.9953752209503936e-05, "loss": 1.8441, "step": 4642 }, { "epoch": 1.0850983027080716, "grad_norm": 4.74448299407959, "learning_rate": 2.995370764552695e-05, "loss": 1.9895, "step": 4643 }, { "epoch": 1.085332008997692, "grad_norm": 2.8021903038024902, "learning_rate": 2.995366306012274e-05, "loss": 1.9684, "step": 4644 }, { "epoch": 1.0855657152873126, "grad_norm": 2.711268186569214, "learning_rate": 2.995361845329139e-05, "loss": 1.9079, "step": 4645 }, { "epoch": 1.085799421576933, "grad_norm": 3.2748067378997803, "learning_rate": 2.995357382503295e-05, "loss": 1.8703, "step": 4646 }, { "epoch": 1.0860331278665538, "grad_norm": 4.048442840576172, "learning_rate": 2.995352917534749e-05, "loss": 1.8534, "step": 4647 }, { "epoch": 1.0862668341561743, "grad_norm": 2.8826262950897217, "learning_rate": 2.9953484504235064e-05, "loss": 1.8306, "step": 4648 }, { "epoch": 1.0865005404457948, "grad_norm": 3.7025976181030273, "learning_rate": 2.995343981169575e-05, "loss": 1.8868, "step": 4649 }, { "epoch": 1.0867342467354153, "grad_norm": 4.114192485809326, "learning_rate": 2.9953395097729604e-05, "loss": 1.9328, "step": 4650 }, { "epoch": 1.0869679530250358, "grad_norm": 2.700291156768799, "learning_rate": 2.9953350362336697e-05, "loss": 1.8284, "step": 4651 }, { "epoch": 1.0872016593146563, "grad_norm": 4.790151119232178, "learning_rate": 2.995330560551708e-05, "loss": 1.8729, "step": 4652 }, { "epoch": 1.0874353656042768, "grad_norm": 4.13860559463501, "learning_rate": 2.9953260827270832e-05, "loss": 1.8416, "step": 4653 }, { "epoch": 1.0876690718938973, "grad_norm": 2.530042886734009, "learning_rate": 2.995321602759801e-05, "loss": 1.8355, "step": 4654 }, { "epoch": 1.0879027781835178, "grad_norm": 2.7081027030944824, "learning_rate": 2.9953171206498674e-05, "loss": 1.8331, "step": 4655 }, { "epoch": 1.0881364844731385, "grad_norm": 2.537808656692505, "learning_rate": 2.9953126363972897e-05, "loss": 2.0258, "step": 4656 }, { "epoch": 1.088370190762759, "grad_norm": 2.4794821739196777, "learning_rate": 2.9953081500020738e-05, "loss": 2.0554, "step": 4657 }, { "epoch": 1.0886038970523795, "grad_norm": 2.642946481704712, "learning_rate": 2.995303661464226e-05, "loss": 1.8808, "step": 4658 }, { "epoch": 1.088837603342, "grad_norm": 4.005585670471191, "learning_rate": 2.9952991707837534e-05, "loss": 1.8453, "step": 4659 }, { "epoch": 1.0890713096316205, "grad_norm": 3.0712389945983887, "learning_rate": 2.9952946779606612e-05, "loss": 1.8569, "step": 4660 }, { "epoch": 1.089305015921241, "grad_norm": 3.018270969390869, "learning_rate": 2.995290182994958e-05, "loss": 1.9322, "step": 4661 }, { "epoch": 1.0895387222108615, "grad_norm": 3.5379834175109863, "learning_rate": 2.9952856858866477e-05, "loss": 1.8841, "step": 4662 }, { "epoch": 1.089772428500482, "grad_norm": 2.9522039890289307, "learning_rate": 2.9952811866357383e-05, "loss": 1.8685, "step": 4663 }, { "epoch": 1.0900061347901024, "grad_norm": 3.736117362976074, "learning_rate": 2.995276685242236e-05, "loss": 1.8689, "step": 4664 }, { "epoch": 1.0902398410797232, "grad_norm": 3.4053947925567627, "learning_rate": 2.995272181706147e-05, "loss": 1.9449, "step": 4665 }, { "epoch": 1.0904735473693437, "grad_norm": 2.6173853874206543, "learning_rate": 2.9952676760274777e-05, "loss": 1.8985, "step": 4666 }, { "epoch": 1.0907072536589641, "grad_norm": 3.3016600608825684, "learning_rate": 2.995263168206235e-05, "loss": 1.8027, "step": 4667 }, { "epoch": 1.0909409599485846, "grad_norm": 2.72590708732605, "learning_rate": 2.995258658242425e-05, "loss": 2.0156, "step": 4668 }, { "epoch": 1.0911746662382051, "grad_norm": 2.547665596008301, "learning_rate": 2.995254146136054e-05, "loss": 1.8129, "step": 4669 }, { "epoch": 1.0914083725278256, "grad_norm": 3.3159563541412354, "learning_rate": 2.995249631887129e-05, "loss": 1.8133, "step": 4670 }, { "epoch": 1.0916420788174461, "grad_norm": 3.740271806716919, "learning_rate": 2.9952451154956564e-05, "loss": 1.881, "step": 4671 }, { "epoch": 1.0918757851070666, "grad_norm": 2.451007604598999, "learning_rate": 2.9952405969616425e-05, "loss": 1.7957, "step": 4672 }, { "epoch": 1.0921094913966871, "grad_norm": 5.747876167297363, "learning_rate": 2.9952360762850932e-05, "loss": 1.8509, "step": 4673 }, { "epoch": 1.0923431976863078, "grad_norm": 4.954195976257324, "learning_rate": 2.9952315534660155e-05, "loss": 1.8455, "step": 4674 }, { "epoch": 1.0925769039759283, "grad_norm": 2.9669933319091797, "learning_rate": 2.9952270285044167e-05, "loss": 1.7996, "step": 4675 }, { "epoch": 1.0928106102655488, "grad_norm": 3.084340810775757, "learning_rate": 2.9952225014003016e-05, "loss": 1.9101, "step": 4676 }, { "epoch": 1.0930443165551693, "grad_norm": 5.994510173797607, "learning_rate": 2.995217972153678e-05, "loss": 1.8703, "step": 4677 }, { "epoch": 1.0932780228447898, "grad_norm": 4.197116374969482, "learning_rate": 2.995213440764552e-05, "loss": 1.8894, "step": 4678 }, { "epoch": 1.0935117291344103, "grad_norm": 2.3213069438934326, "learning_rate": 2.99520890723293e-05, "loss": 1.8846, "step": 4679 }, { "epoch": 1.0937454354240308, "grad_norm": 4.953786373138428, "learning_rate": 2.995204371558818e-05, "loss": 2.0058, "step": 4680 }, { "epoch": 1.0939791417136513, "grad_norm": 5.727564811706543, "learning_rate": 2.9951998337422237e-05, "loss": 1.9959, "step": 4681 }, { "epoch": 1.0942128480032718, "grad_norm": 4.216311931610107, "learning_rate": 2.9951952937831524e-05, "loss": 1.9581, "step": 4682 }, { "epoch": 1.0944465542928925, "grad_norm": 2.320638418197632, "learning_rate": 2.9951907516816115e-05, "loss": 1.8508, "step": 4683 }, { "epoch": 1.094680260582513, "grad_norm": 3.092881441116333, "learning_rate": 2.9951862074376072e-05, "loss": 2.0911, "step": 4684 }, { "epoch": 1.0949139668721335, "grad_norm": 3.5668418407440186, "learning_rate": 2.995181661051146e-05, "loss": 1.9891, "step": 4685 }, { "epoch": 1.095147673161754, "grad_norm": 3.389676094055176, "learning_rate": 2.9951771125222343e-05, "loss": 1.9232, "step": 4686 }, { "epoch": 1.0953813794513745, "grad_norm": 3.759333848953247, "learning_rate": 2.9951725618508786e-05, "loss": 1.8272, "step": 4687 }, { "epoch": 1.095615085740995, "grad_norm": 3.50325608253479, "learning_rate": 2.9951680090370855e-05, "loss": 1.9817, "step": 4688 }, { "epoch": 1.0958487920306155, "grad_norm": 2.576349973678589, "learning_rate": 2.995163454080862e-05, "loss": 1.9667, "step": 4689 }, { "epoch": 1.096082498320236, "grad_norm": 2.972372531890869, "learning_rate": 2.9951588969822137e-05, "loss": 1.8824, "step": 4690 }, { "epoch": 1.0963162046098565, "grad_norm": 2.642824411392212, "learning_rate": 2.9951543377411475e-05, "loss": 1.9684, "step": 4691 }, { "epoch": 1.0965499108994772, "grad_norm": 3.05505633354187, "learning_rate": 2.9951497763576703e-05, "loss": 1.8487, "step": 4692 }, { "epoch": 1.0967836171890977, "grad_norm": 2.712388038635254, "learning_rate": 2.9951452128317883e-05, "loss": 1.928, "step": 4693 }, { "epoch": 1.0970173234787182, "grad_norm": 4.969226837158203, "learning_rate": 2.9951406471635083e-05, "loss": 1.8243, "step": 4694 }, { "epoch": 1.0972510297683387, "grad_norm": 2.446974515914917, "learning_rate": 2.9951360793528366e-05, "loss": 1.9818, "step": 4695 }, { "epoch": 1.0974847360579592, "grad_norm": 2.6695263385772705, "learning_rate": 2.9951315093997794e-05, "loss": 1.8947, "step": 4696 }, { "epoch": 1.0977184423475796, "grad_norm": 3.199275493621826, "learning_rate": 2.995126937304344e-05, "loss": 1.7841, "step": 4697 }, { "epoch": 1.0979521486372001, "grad_norm": 3.9524292945861816, "learning_rate": 2.9951223630665366e-05, "loss": 1.9196, "step": 4698 }, { "epoch": 1.0981858549268206, "grad_norm": 3.6479134559631348, "learning_rate": 2.9951177866863638e-05, "loss": 1.8394, "step": 4699 }, { "epoch": 1.0984195612164411, "grad_norm": 4.67551851272583, "learning_rate": 2.995113208163832e-05, "loss": 1.7766, "step": 4700 }, { "epoch": 1.0986532675060618, "grad_norm": 3.430398941040039, "learning_rate": 2.9951086274989477e-05, "loss": 1.9298, "step": 4701 }, { "epoch": 1.0988869737956823, "grad_norm": 3.8556911945343018, "learning_rate": 2.9951040446917176e-05, "loss": 1.9673, "step": 4702 }, { "epoch": 1.0991206800853028, "grad_norm": 4.188712120056152, "learning_rate": 2.9950994597421488e-05, "loss": 1.9732, "step": 4703 }, { "epoch": 1.0993543863749233, "grad_norm": 3.206634283065796, "learning_rate": 2.995094872650247e-05, "loss": 1.954, "step": 4704 }, { "epoch": 1.0995880926645438, "grad_norm": 2.6717662811279297, "learning_rate": 2.995090283416019e-05, "loss": 1.8226, "step": 4705 }, { "epoch": 1.0998217989541643, "grad_norm": 3.18412184715271, "learning_rate": 2.9950856920394716e-05, "loss": 1.8859, "step": 4706 }, { "epoch": 1.1000555052437848, "grad_norm": 2.5489890575408936, "learning_rate": 2.9950810985206117e-05, "loss": 1.9041, "step": 4707 }, { "epoch": 1.1002892115334053, "grad_norm": 2.9468114376068115, "learning_rate": 2.9950765028594452e-05, "loss": 1.7998, "step": 4708 }, { "epoch": 1.1005229178230258, "grad_norm": 2.465294122695923, "learning_rate": 2.9950719050559786e-05, "loss": 1.8393, "step": 4709 }, { "epoch": 1.1007566241126465, "grad_norm": 2.912867784500122, "learning_rate": 2.9950673051102193e-05, "loss": 1.9567, "step": 4710 }, { "epoch": 1.100990330402267, "grad_norm": 3.236738681793213, "learning_rate": 2.9950627030221734e-05, "loss": 1.8396, "step": 4711 }, { "epoch": 1.1012240366918875, "grad_norm": 2.4547500610351562, "learning_rate": 2.9950580987918473e-05, "loss": 2.0008, "step": 4712 }, { "epoch": 1.101457742981508, "grad_norm": 2.6608734130859375, "learning_rate": 2.9950534924192486e-05, "loss": 1.958, "step": 4713 }, { "epoch": 1.1016914492711285, "grad_norm": 2.352980375289917, "learning_rate": 2.9950488839043825e-05, "loss": 2.0038, "step": 4714 }, { "epoch": 1.101925155560749, "grad_norm": 3.0667247772216797, "learning_rate": 2.9950442732472563e-05, "loss": 1.9714, "step": 4715 }, { "epoch": 1.1021588618503695, "grad_norm": 2.6929268836975098, "learning_rate": 2.9950396604478764e-05, "loss": 1.8157, "step": 4716 }, { "epoch": 1.10239256813999, "grad_norm": 2.468709945678711, "learning_rate": 2.9950350455062495e-05, "loss": 1.9087, "step": 4717 }, { "epoch": 1.1026262744296105, "grad_norm": 3.6659858226776123, "learning_rate": 2.9950304284223828e-05, "loss": 1.7902, "step": 4718 }, { "epoch": 1.1028599807192312, "grad_norm": 3.098226308822632, "learning_rate": 2.995025809196282e-05, "loss": 1.7868, "step": 4719 }, { "epoch": 1.1030936870088517, "grad_norm": 2.8654963970184326, "learning_rate": 2.9950211878279542e-05, "loss": 2.1675, "step": 4720 }, { "epoch": 1.1033273932984722, "grad_norm": 3.139399528503418, "learning_rate": 2.9950165643174056e-05, "loss": 1.9034, "step": 4721 }, { "epoch": 1.1035610995880927, "grad_norm": 2.5964791774749756, "learning_rate": 2.9950119386646436e-05, "loss": 1.802, "step": 4722 }, { "epoch": 1.1037948058777132, "grad_norm": 3.706488847732544, "learning_rate": 2.9950073108696742e-05, "loss": 1.7859, "step": 4723 }, { "epoch": 1.1040285121673337, "grad_norm": 3.545280933380127, "learning_rate": 2.9950026809325044e-05, "loss": 1.8663, "step": 4724 }, { "epoch": 1.1042622184569542, "grad_norm": 2.679776906967163, "learning_rate": 2.9949980488531404e-05, "loss": 1.8342, "step": 4725 }, { "epoch": 1.1044959247465747, "grad_norm": 3.0657310485839844, "learning_rate": 2.994993414631589e-05, "loss": 1.9434, "step": 4726 }, { "epoch": 1.1047296310361951, "grad_norm": 3.281479597091675, "learning_rate": 2.9949887782678573e-05, "loss": 1.904, "step": 4727 }, { "epoch": 1.1049633373258159, "grad_norm": 2.4006381034851074, "learning_rate": 2.994984139761951e-05, "loss": 1.7552, "step": 4728 }, { "epoch": 1.1051970436154364, "grad_norm": 3.4246580600738525, "learning_rate": 2.994979499113878e-05, "loss": 1.7515, "step": 4729 }, { "epoch": 1.1054307499050569, "grad_norm": 2.4805142879486084, "learning_rate": 2.9949748563236436e-05, "loss": 1.9782, "step": 4730 }, { "epoch": 1.1056644561946773, "grad_norm": 2.9366438388824463, "learning_rate": 2.9949702113912556e-05, "loss": 1.9502, "step": 4731 }, { "epoch": 1.1058981624842978, "grad_norm": 5.1688642501831055, "learning_rate": 2.99496556431672e-05, "loss": 1.9579, "step": 4732 }, { "epoch": 1.1061318687739183, "grad_norm": 2.96500563621521, "learning_rate": 2.9949609151000435e-05, "loss": 1.9044, "step": 4733 }, { "epoch": 1.1063655750635388, "grad_norm": 3.4893572330474854, "learning_rate": 2.994956263741233e-05, "loss": 1.7956, "step": 4734 }, { "epoch": 1.1065992813531593, "grad_norm": 4.5520405769348145, "learning_rate": 2.9949516102402955e-05, "loss": 1.9356, "step": 4735 }, { "epoch": 1.1068329876427798, "grad_norm": 4.009535312652588, "learning_rate": 2.9949469545972363e-05, "loss": 1.8959, "step": 4736 }, { "epoch": 1.1070666939324005, "grad_norm": 2.703685998916626, "learning_rate": 2.994942296812064e-05, "loss": 1.8992, "step": 4737 }, { "epoch": 1.107300400222021, "grad_norm": 3.0166726112365723, "learning_rate": 2.9949376368847835e-05, "loss": 1.9599, "step": 4738 }, { "epoch": 1.1075341065116415, "grad_norm": 2.7328691482543945, "learning_rate": 2.9949329748154025e-05, "loss": 1.884, "step": 4739 }, { "epoch": 1.107767812801262, "grad_norm": 2.6855218410491943, "learning_rate": 2.9949283106039274e-05, "loss": 1.9406, "step": 4740 }, { "epoch": 1.1080015190908825, "grad_norm": 2.327160358428955, "learning_rate": 2.994923644250365e-05, "loss": 1.8717, "step": 4741 }, { "epoch": 1.108235225380503, "grad_norm": 4.2993083000183105, "learning_rate": 2.9949189757547218e-05, "loss": 1.9279, "step": 4742 }, { "epoch": 1.1084689316701235, "grad_norm": 2.4609274864196777, "learning_rate": 2.9949143051170045e-05, "loss": 1.8975, "step": 4743 }, { "epoch": 1.108702637959744, "grad_norm": 2.4554383754730225, "learning_rate": 2.9949096323372202e-05, "loss": 1.9865, "step": 4744 }, { "epoch": 1.1089363442493647, "grad_norm": 4.217386245727539, "learning_rate": 2.994904957415375e-05, "loss": 1.8473, "step": 4745 }, { "epoch": 1.1091700505389852, "grad_norm": 3.305041790008545, "learning_rate": 2.994900280351476e-05, "loss": 1.7735, "step": 4746 }, { "epoch": 1.1094037568286057, "grad_norm": 3.5100655555725098, "learning_rate": 2.9948956011455298e-05, "loss": 1.6716, "step": 4747 }, { "epoch": 1.1096374631182262, "grad_norm": 3.8004045486450195, "learning_rate": 2.9948909197975427e-05, "loss": 1.7884, "step": 4748 }, { "epoch": 1.1098711694078467, "grad_norm": 2.6852331161499023, "learning_rate": 2.994886236307522e-05, "loss": 1.9698, "step": 4749 }, { "epoch": 1.1101048756974672, "grad_norm": 2.586697816848755, "learning_rate": 2.9948815506754745e-05, "loss": 1.8906, "step": 4750 }, { "epoch": 1.1103385819870877, "grad_norm": 2.7019100189208984, "learning_rate": 2.9948768629014062e-05, "loss": 1.9702, "step": 4751 }, { "epoch": 1.1105722882767082, "grad_norm": 2.3367090225219727, "learning_rate": 2.9948721729853248e-05, "loss": 1.7698, "step": 4752 }, { "epoch": 1.1108059945663287, "grad_norm": 2.2510759830474854, "learning_rate": 2.994867480927236e-05, "loss": 1.8858, "step": 4753 }, { "epoch": 1.1110397008559494, "grad_norm": 2.2682793140411377, "learning_rate": 2.994862786727147e-05, "loss": 1.9161, "step": 4754 }, { "epoch": 1.1112734071455699, "grad_norm": 2.529003620147705, "learning_rate": 2.9948580903850645e-05, "loss": 1.8098, "step": 4755 }, { "epoch": 1.1115071134351904, "grad_norm": 2.2421534061431885, "learning_rate": 2.9948533919009955e-05, "loss": 1.9202, "step": 4756 }, { "epoch": 1.1117408197248109, "grad_norm": 2.7284927368164062, "learning_rate": 2.994848691274946e-05, "loss": 1.9595, "step": 4757 }, { "epoch": 1.1119745260144314, "grad_norm": 2.255598783493042, "learning_rate": 2.9948439885069236e-05, "loss": 1.9579, "step": 4758 }, { "epoch": 1.1122082323040519, "grad_norm": 2.490274667739868, "learning_rate": 2.9948392835969343e-05, "loss": 1.8508, "step": 4759 }, { "epoch": 1.1124419385936724, "grad_norm": 3.2041330337524414, "learning_rate": 2.9948345765449857e-05, "loss": 2.0338, "step": 4760 }, { "epoch": 1.1126756448832928, "grad_norm": 2.5262932777404785, "learning_rate": 2.9948298673510832e-05, "loss": 1.9423, "step": 4761 }, { "epoch": 1.1129093511729133, "grad_norm": 3.3868322372436523, "learning_rate": 2.994825156015235e-05, "loss": 1.8908, "step": 4762 }, { "epoch": 1.113143057462534, "grad_norm": 2.4140522480010986, "learning_rate": 2.994820442537447e-05, "loss": 2.0999, "step": 4763 }, { "epoch": 1.1133767637521546, "grad_norm": 2.1386351585388184, "learning_rate": 2.994815726917726e-05, "loss": 1.745, "step": 4764 }, { "epoch": 1.113610470041775, "grad_norm": 3.227491855621338, "learning_rate": 2.9948110091560795e-05, "loss": 1.9806, "step": 4765 }, { "epoch": 1.1138441763313955, "grad_norm": 2.696993827819824, "learning_rate": 2.9948062892525133e-05, "loss": 1.9724, "step": 4766 }, { "epoch": 1.114077882621016, "grad_norm": 2.397810220718384, "learning_rate": 2.9948015672070346e-05, "loss": 1.9523, "step": 4767 }, { "epoch": 1.1143115889106365, "grad_norm": 4.308778762817383, "learning_rate": 2.9947968430196502e-05, "loss": 1.9321, "step": 4768 }, { "epoch": 1.114545295200257, "grad_norm": 2.737929105758667, "learning_rate": 2.994792116690367e-05, "loss": 2.089, "step": 4769 }, { "epoch": 1.1147790014898775, "grad_norm": 2.366579294204712, "learning_rate": 2.994787388219191e-05, "loss": 1.9057, "step": 4770 }, { "epoch": 1.115012707779498, "grad_norm": 4.383698463439941, "learning_rate": 2.9947826576061296e-05, "loss": 1.8917, "step": 4771 }, { "epoch": 1.1152464140691187, "grad_norm": 3.049950361251831, "learning_rate": 2.99477792485119e-05, "loss": 1.9961, "step": 4772 }, { "epoch": 1.1154801203587392, "grad_norm": 3.1260340213775635, "learning_rate": 2.9947731899543782e-05, "loss": 1.9014, "step": 4773 }, { "epoch": 1.1157138266483597, "grad_norm": 2.8506412506103516, "learning_rate": 2.994768452915701e-05, "loss": 1.9524, "step": 4774 }, { "epoch": 1.1159475329379802, "grad_norm": 2.3828718662261963, "learning_rate": 2.994763713735166e-05, "loss": 1.7955, "step": 4775 }, { "epoch": 1.1161812392276007, "grad_norm": 2.444481372833252, "learning_rate": 2.994758972412779e-05, "loss": 1.8827, "step": 4776 }, { "epoch": 1.1164149455172212, "grad_norm": 3.5413012504577637, "learning_rate": 2.9947542289485476e-05, "loss": 1.773, "step": 4777 }, { "epoch": 1.1166486518068417, "grad_norm": 2.3785929679870605, "learning_rate": 2.994749483342478e-05, "loss": 1.8737, "step": 4778 }, { "epoch": 1.1168823580964622, "grad_norm": 3.915722370147705, "learning_rate": 2.9947447355945778e-05, "loss": 1.8893, "step": 4779 }, { "epoch": 1.117116064386083, "grad_norm": 3.7312774658203125, "learning_rate": 2.9947399857048527e-05, "loss": 1.9032, "step": 4780 }, { "epoch": 1.1173497706757034, "grad_norm": 2.475222110748291, "learning_rate": 2.9947352336733102e-05, "loss": 1.8941, "step": 4781 }, { "epoch": 1.117583476965324, "grad_norm": 2.8735456466674805, "learning_rate": 2.9947304794999572e-05, "loss": 1.9376, "step": 4782 }, { "epoch": 1.1178171832549444, "grad_norm": 4.693267822265625, "learning_rate": 2.9947257231848e-05, "loss": 1.6934, "step": 4783 }, { "epoch": 1.1180508895445649, "grad_norm": 2.37031626701355, "learning_rate": 2.9947209647278456e-05, "loss": 1.9997, "step": 4784 }, { "epoch": 1.1182845958341854, "grad_norm": 3.1202807426452637, "learning_rate": 2.9947162041291013e-05, "loss": 1.8594, "step": 4785 }, { "epoch": 1.1185183021238059, "grad_norm": 5.841130256652832, "learning_rate": 2.994711441388573e-05, "loss": 1.926, "step": 4786 }, { "epoch": 1.1187520084134264, "grad_norm": 3.3635926246643066, "learning_rate": 2.9947066765062686e-05, "loss": 1.8737, "step": 4787 }, { "epoch": 1.1189857147030469, "grad_norm": 2.9261362552642822, "learning_rate": 2.9947019094821942e-05, "loss": 1.8255, "step": 4788 }, { "epoch": 1.1192194209926676, "grad_norm": 7.018653392791748, "learning_rate": 2.9946971403163567e-05, "loss": 1.9322, "step": 4789 }, { "epoch": 1.119453127282288, "grad_norm": 4.253231048583984, "learning_rate": 2.994692369008763e-05, "loss": 1.8792, "step": 4790 }, { "epoch": 1.1196868335719086, "grad_norm": 3.795048952102661, "learning_rate": 2.9946875955594207e-05, "loss": 1.8896, "step": 4791 }, { "epoch": 1.119920539861529, "grad_norm": 4.127931118011475, "learning_rate": 2.994682819968335e-05, "loss": 2.0049, "step": 4792 }, { "epoch": 1.1201542461511496, "grad_norm": 3.365974187850952, "learning_rate": 2.994678042235514e-05, "loss": 1.9147, "step": 4793 }, { "epoch": 1.12038795244077, "grad_norm": 3.617187976837158, "learning_rate": 2.9946732623609644e-05, "loss": 1.8128, "step": 4794 }, { "epoch": 1.1206216587303905, "grad_norm": 3.6723835468292236, "learning_rate": 2.9946684803446928e-05, "loss": 1.8606, "step": 4795 }, { "epoch": 1.120855365020011, "grad_norm": 2.4938690662384033, "learning_rate": 2.9946636961867058e-05, "loss": 1.8116, "step": 4796 }, { "epoch": 1.1210890713096315, "grad_norm": 2.484403133392334, "learning_rate": 2.9946589098870113e-05, "loss": 2.1646, "step": 4797 }, { "epoch": 1.1213227775992523, "grad_norm": 3.1368632316589355, "learning_rate": 2.9946541214456146e-05, "loss": 1.9326, "step": 4798 }, { "epoch": 1.1215564838888727, "grad_norm": 3.513066530227661, "learning_rate": 2.994649330862524e-05, "loss": 1.9429, "step": 4799 }, { "epoch": 1.1217901901784932, "grad_norm": 2.9655470848083496, "learning_rate": 2.9946445381377453e-05, "loss": 1.7131, "step": 4800 }, { "epoch": 1.1220238964681137, "grad_norm": 3.425867795944214, "learning_rate": 2.994639743271286e-05, "loss": 1.9591, "step": 4801 }, { "epoch": 1.1222576027577342, "grad_norm": 5.817063808441162, "learning_rate": 2.994634946263153e-05, "loss": 2.0339, "step": 4802 }, { "epoch": 1.1224913090473547, "grad_norm": 3.6517889499664307, "learning_rate": 2.994630147113353e-05, "loss": 1.9921, "step": 4803 }, { "epoch": 1.1227250153369752, "grad_norm": 5.739150524139404, "learning_rate": 2.9946253458218925e-05, "loss": 1.8606, "step": 4804 }, { "epoch": 1.1229587216265957, "grad_norm": 4.332708835601807, "learning_rate": 2.9946205423887787e-05, "loss": 1.7848, "step": 4805 }, { "epoch": 1.1231924279162162, "grad_norm": 3.1311087608337402, "learning_rate": 2.9946157368140187e-05, "loss": 1.8761, "step": 4806 }, { "epoch": 1.123426134205837, "grad_norm": 2.60465145111084, "learning_rate": 2.9946109290976196e-05, "loss": 1.8528, "step": 4807 }, { "epoch": 1.1236598404954574, "grad_norm": 3.979818105697632, "learning_rate": 2.994606119239587e-05, "loss": 1.8969, "step": 4808 }, { "epoch": 1.123893546785078, "grad_norm": 2.960944414138794, "learning_rate": 2.9946013072399297e-05, "loss": 1.9482, "step": 4809 }, { "epoch": 1.1241272530746984, "grad_norm": 2.928323268890381, "learning_rate": 2.9945964930986525e-05, "loss": 1.7705, "step": 4810 }, { "epoch": 1.124360959364319, "grad_norm": 2.8501992225646973, "learning_rate": 2.994591676815764e-05, "loss": 1.8991, "step": 4811 }, { "epoch": 1.1245946656539394, "grad_norm": 2.9355103969573975, "learning_rate": 2.994586858391271e-05, "loss": 1.891, "step": 4812 }, { "epoch": 1.12482837194356, "grad_norm": 2.9252400398254395, "learning_rate": 2.994582037825179e-05, "loss": 1.8885, "step": 4813 }, { "epoch": 1.1250620782331804, "grad_norm": 2.694383144378662, "learning_rate": 2.9945772151174958e-05, "loss": 1.8589, "step": 4814 }, { "epoch": 1.1252957845228009, "grad_norm": 4.060123920440674, "learning_rate": 2.9945723902682283e-05, "loss": 1.8761, "step": 4815 }, { "epoch": 1.1255294908124216, "grad_norm": 2.48164439201355, "learning_rate": 2.994567563277384e-05, "loss": 1.9469, "step": 4816 }, { "epoch": 1.125763197102042, "grad_norm": 2.478538990020752, "learning_rate": 2.9945627341449688e-05, "loss": 1.8791, "step": 4817 }, { "epoch": 1.1259969033916626, "grad_norm": 2.956639289855957, "learning_rate": 2.9945579028709903e-05, "loss": 1.949, "step": 4818 }, { "epoch": 1.126230609681283, "grad_norm": 2.448975086212158, "learning_rate": 2.9945530694554547e-05, "loss": 2.0174, "step": 4819 }, { "epoch": 1.1264643159709036, "grad_norm": 2.9069101810455322, "learning_rate": 2.99454823389837e-05, "loss": 1.8153, "step": 4820 }, { "epoch": 1.126698022260524, "grad_norm": 2.4675800800323486, "learning_rate": 2.9945433961997422e-05, "loss": 1.8465, "step": 4821 }, { "epoch": 1.1269317285501446, "grad_norm": 2.569063901901245, "learning_rate": 2.9945385563595783e-05, "loss": 1.9429, "step": 4822 }, { "epoch": 1.127165434839765, "grad_norm": 2.812976121902466, "learning_rate": 2.9945337143778863e-05, "loss": 1.8027, "step": 4823 }, { "epoch": 1.1273991411293856, "grad_norm": 3.3030781745910645, "learning_rate": 2.9945288702546718e-05, "loss": 1.9191, "step": 4824 }, { "epoch": 1.1276328474190063, "grad_norm": 2.397214889526367, "learning_rate": 2.994524023989942e-05, "loss": 1.9199, "step": 4825 }, { "epoch": 1.1278665537086268, "grad_norm": 3.2789597511291504, "learning_rate": 2.994519175583704e-05, "loss": 1.8969, "step": 4826 }, { "epoch": 1.1281002599982473, "grad_norm": 3.0918643474578857, "learning_rate": 2.9945143250359656e-05, "loss": 1.8087, "step": 4827 }, { "epoch": 1.1283339662878678, "grad_norm": 2.3441293239593506, "learning_rate": 2.9945094723467326e-05, "loss": 1.8481, "step": 4828 }, { "epoch": 1.1285676725774882, "grad_norm": 4.563718318939209, "learning_rate": 2.9945046175160123e-05, "loss": 1.8748, "step": 4829 }, { "epoch": 1.1288013788671087, "grad_norm": 4.715076446533203, "learning_rate": 2.9944997605438118e-05, "loss": 1.8276, "step": 4830 }, { "epoch": 1.1290350851567292, "grad_norm": 2.6419589519500732, "learning_rate": 2.9944949014301385e-05, "loss": 1.9581, "step": 4831 }, { "epoch": 1.1292687914463497, "grad_norm": 5.889037132263184, "learning_rate": 2.994490040174998e-05, "loss": 2.0236, "step": 4832 }, { "epoch": 1.1295024977359702, "grad_norm": 6.8945465087890625, "learning_rate": 2.9944851767783983e-05, "loss": 1.9282, "step": 4833 }, { "epoch": 1.129736204025591, "grad_norm": 3.6678521633148193, "learning_rate": 2.9944803112403467e-05, "loss": 1.9359, "step": 4834 }, { "epoch": 1.1299699103152114, "grad_norm": 2.9171931743621826, "learning_rate": 2.994475443560849e-05, "loss": 1.885, "step": 4835 }, { "epoch": 1.130203616604832, "grad_norm": 3.0913233757019043, "learning_rate": 2.9944705737399134e-05, "loss": 1.791, "step": 4836 }, { "epoch": 1.1304373228944524, "grad_norm": 3.422466516494751, "learning_rate": 2.9944657017775458e-05, "loss": 1.925, "step": 4837 }, { "epoch": 1.130671029184073, "grad_norm": 2.376732349395752, "learning_rate": 2.994460827673754e-05, "loss": 1.9365, "step": 4838 }, { "epoch": 1.1309047354736934, "grad_norm": 5.986307621002197, "learning_rate": 2.9944559514285445e-05, "loss": 2.1395, "step": 4839 }, { "epoch": 1.131138441763314, "grad_norm": 4.822865962982178, "learning_rate": 2.9944510730419243e-05, "loss": 1.9409, "step": 4840 }, { "epoch": 1.1313721480529344, "grad_norm": 3.771383762359619, "learning_rate": 2.9944461925139008e-05, "loss": 1.8663, "step": 4841 }, { "epoch": 1.131605854342555, "grad_norm": 4.636715888977051, "learning_rate": 2.9944413098444808e-05, "loss": 1.8545, "step": 4842 }, { "epoch": 1.1318395606321756, "grad_norm": 6.150402069091797, "learning_rate": 2.9944364250336714e-05, "loss": 1.8768, "step": 4843 }, { "epoch": 1.132073266921796, "grad_norm": 4.767149448394775, "learning_rate": 2.994431538081479e-05, "loss": 1.9236, "step": 4844 }, { "epoch": 1.1323069732114166, "grad_norm": 2.4104862213134766, "learning_rate": 2.994426648987911e-05, "loss": 1.9022, "step": 4845 }, { "epoch": 1.132540679501037, "grad_norm": 5.133652687072754, "learning_rate": 2.9944217577529747e-05, "loss": 2.0427, "step": 4846 }, { "epoch": 1.1327743857906576, "grad_norm": 6.778326511383057, "learning_rate": 2.9944168643766767e-05, "loss": 2.0009, "step": 4847 }, { "epoch": 1.133008092080278, "grad_norm": 4.382068157196045, "learning_rate": 2.9944119688590246e-05, "loss": 1.7348, "step": 4848 }, { "epoch": 1.1332417983698986, "grad_norm": 2.5078675746917725, "learning_rate": 2.9944070712000247e-05, "loss": 1.9233, "step": 4849 }, { "epoch": 1.133475504659519, "grad_norm": 5.402267932891846, "learning_rate": 2.9944021713996842e-05, "loss": 1.7411, "step": 4850 }, { "epoch": 1.1337092109491396, "grad_norm": 6.272580146789551, "learning_rate": 2.9943972694580102e-05, "loss": 1.9475, "step": 4851 }, { "epoch": 1.1339429172387603, "grad_norm": 5.874711036682129, "learning_rate": 2.9943923653750098e-05, "loss": 1.8539, "step": 4852 }, { "epoch": 1.1341766235283808, "grad_norm": 2.613043785095215, "learning_rate": 2.9943874591506897e-05, "loss": 1.9601, "step": 4853 }, { "epoch": 1.1344103298180013, "grad_norm": 5.653476715087891, "learning_rate": 2.994382550785058e-05, "loss": 1.864, "step": 4854 }, { "epoch": 1.1346440361076218, "grad_norm": 6.698519229888916, "learning_rate": 2.99437764027812e-05, "loss": 1.9208, "step": 4855 }, { "epoch": 1.1348777423972423, "grad_norm": 4.274765968322754, "learning_rate": 2.9943727276298837e-05, "loss": 1.8962, "step": 4856 }, { "epoch": 1.1351114486868628, "grad_norm": 3.321429967880249, "learning_rate": 2.9943678128403567e-05, "loss": 1.977, "step": 4857 }, { "epoch": 1.1353451549764833, "grad_norm": 6.760220527648926, "learning_rate": 2.9943628959095453e-05, "loss": 1.8326, "step": 4858 }, { "epoch": 1.1355788612661037, "grad_norm": 5.127604007720947, "learning_rate": 2.9943579768374566e-05, "loss": 1.9322, "step": 4859 }, { "epoch": 1.1358125675557242, "grad_norm": 2.603623390197754, "learning_rate": 2.9943530556240974e-05, "loss": 1.8293, "step": 4860 }, { "epoch": 1.136046273845345, "grad_norm": 4.794036865234375, "learning_rate": 2.994348132269476e-05, "loss": 1.8854, "step": 4861 }, { "epoch": 1.1362799801349655, "grad_norm": 6.102962017059326, "learning_rate": 2.9943432067735977e-05, "loss": 1.9791, "step": 4862 }, { "epoch": 1.136513686424586, "grad_norm": 5.309488296508789, "learning_rate": 2.9943382791364705e-05, "loss": 1.8787, "step": 4863 }, { "epoch": 1.1367473927142064, "grad_norm": 2.57517147064209, "learning_rate": 2.994333349358102e-05, "loss": 1.782, "step": 4864 }, { "epoch": 1.136981099003827, "grad_norm": 6.486574172973633, "learning_rate": 2.9943284174384976e-05, "loss": 1.9536, "step": 4865 }, { "epoch": 1.1372148052934474, "grad_norm": 7.970876693725586, "learning_rate": 2.9943234833776666e-05, "loss": 1.854, "step": 4866 }, { "epoch": 1.137448511583068, "grad_norm": 5.54341459274292, "learning_rate": 2.994318547175614e-05, "loss": 2.0479, "step": 4867 }, { "epoch": 1.1376822178726884, "grad_norm": 2.462040662765503, "learning_rate": 2.994313608832348e-05, "loss": 1.8605, "step": 4868 }, { "epoch": 1.137915924162309, "grad_norm": 3.469777822494507, "learning_rate": 2.9943086683478755e-05, "loss": 1.8834, "step": 4869 }, { "epoch": 1.1381496304519296, "grad_norm": 4.16858434677124, "learning_rate": 2.9943037257222035e-05, "loss": 1.984, "step": 4870 }, { "epoch": 1.1383833367415501, "grad_norm": 2.579202651977539, "learning_rate": 2.9942987809553388e-05, "loss": 1.8027, "step": 4871 }, { "epoch": 1.1386170430311706, "grad_norm": 2.4150619506835938, "learning_rate": 2.994293834047289e-05, "loss": 1.7094, "step": 4872 }, { "epoch": 1.1388507493207911, "grad_norm": 2.8850014209747314, "learning_rate": 2.994288884998061e-05, "loss": 1.8612, "step": 4873 }, { "epoch": 1.1390844556104116, "grad_norm": 2.9954769611358643, "learning_rate": 2.9942839338076617e-05, "loss": 1.7537, "step": 4874 }, { "epoch": 1.139318161900032, "grad_norm": 2.644989490509033, "learning_rate": 2.9942789804760983e-05, "loss": 1.825, "step": 4875 }, { "epoch": 1.1395518681896526, "grad_norm": 2.300931453704834, "learning_rate": 2.9942740250033782e-05, "loss": 1.8637, "step": 4876 }, { "epoch": 1.139785574479273, "grad_norm": 2.390087366104126, "learning_rate": 2.9942690673895076e-05, "loss": 1.787, "step": 4877 }, { "epoch": 1.1400192807688936, "grad_norm": 2.2483394145965576, "learning_rate": 2.994264107634495e-05, "loss": 1.86, "step": 4878 }, { "epoch": 1.1402529870585143, "grad_norm": 3.3074402809143066, "learning_rate": 2.9942591457383462e-05, "loss": 1.8488, "step": 4879 }, { "epoch": 1.1404866933481348, "grad_norm": 2.575815439224243, "learning_rate": 2.9942541817010693e-05, "loss": 1.8455, "step": 4880 }, { "epoch": 1.1407203996377553, "grad_norm": 3.175995349884033, "learning_rate": 2.994249215522671e-05, "loss": 1.7955, "step": 4881 }, { "epoch": 1.1409541059273758, "grad_norm": 2.528167724609375, "learning_rate": 2.994244247203158e-05, "loss": 1.9857, "step": 4882 }, { "epoch": 1.1411878122169963, "grad_norm": 3.084804058074951, "learning_rate": 2.994239276742538e-05, "loss": 1.7807, "step": 4883 }, { "epoch": 1.1414215185066168, "grad_norm": 4.662459373474121, "learning_rate": 2.994234304140818e-05, "loss": 1.755, "step": 4884 }, { "epoch": 1.1416552247962373, "grad_norm": 4.8751654624938965, "learning_rate": 2.994229329398005e-05, "loss": 1.7512, "step": 4885 }, { "epoch": 1.1418889310858578, "grad_norm": 3.305230140686035, "learning_rate": 2.994224352514106e-05, "loss": 1.7401, "step": 4886 }, { "epoch": 1.1421226373754783, "grad_norm": 3.6142256259918213, "learning_rate": 2.9942193734891284e-05, "loss": 1.8318, "step": 4887 }, { "epoch": 1.142356343665099, "grad_norm": 4.133058071136475, "learning_rate": 2.9942143923230795e-05, "loss": 2.022, "step": 4888 }, { "epoch": 1.1425900499547195, "grad_norm": 2.4651594161987305, "learning_rate": 2.9942094090159657e-05, "loss": 1.8311, "step": 4889 }, { "epoch": 1.14282375624434, "grad_norm": 2.5920825004577637, "learning_rate": 2.9942044235677952e-05, "loss": 2.0941, "step": 4890 }, { "epoch": 1.1430574625339605, "grad_norm": 2.815997362136841, "learning_rate": 2.994199435978574e-05, "loss": 1.8658, "step": 4891 }, { "epoch": 1.143291168823581, "grad_norm": 2.297421932220459, "learning_rate": 2.9941944462483102e-05, "loss": 1.8724, "step": 4892 }, { "epoch": 1.1435248751132014, "grad_norm": 2.6830735206604004, "learning_rate": 2.9941894543770106e-05, "loss": 1.9122, "step": 4893 }, { "epoch": 1.143758581402822, "grad_norm": 5.029969215393066, "learning_rate": 2.994184460364682e-05, "loss": 1.9833, "step": 4894 }, { "epoch": 1.1439922876924424, "grad_norm": 2.403576612472534, "learning_rate": 2.994179464211332e-05, "loss": 1.7889, "step": 4895 }, { "epoch": 1.144225993982063, "grad_norm": 2.728407144546509, "learning_rate": 2.994174465916968e-05, "loss": 1.9067, "step": 4896 }, { "epoch": 1.1444597002716836, "grad_norm": 3.757385492324829, "learning_rate": 2.9941694654815962e-05, "loss": 1.8559, "step": 4897 }, { "epoch": 1.1446934065613041, "grad_norm": 2.5582475662231445, "learning_rate": 2.994164462905225e-05, "loss": 1.9873, "step": 4898 }, { "epoch": 1.1449271128509246, "grad_norm": 2.30798602104187, "learning_rate": 2.9941594581878606e-05, "loss": 1.8685, "step": 4899 }, { "epoch": 1.1451608191405451, "grad_norm": 2.757939577102661, "learning_rate": 2.9941544513295103e-05, "loss": 1.9871, "step": 4900 }, { "epoch": 1.1453945254301656, "grad_norm": 3.7043962478637695, "learning_rate": 2.9941494423301817e-05, "loss": 1.7866, "step": 4901 }, { "epoch": 1.1456282317197861, "grad_norm": 2.637655735015869, "learning_rate": 2.9941444311898817e-05, "loss": 1.9376, "step": 4902 }, { "epoch": 1.1458619380094066, "grad_norm": 2.5638887882232666, "learning_rate": 2.9941394179086175e-05, "loss": 1.8631, "step": 4903 }, { "epoch": 1.146095644299027, "grad_norm": 2.4567394256591797, "learning_rate": 2.9941344024863962e-05, "loss": 1.6935, "step": 4904 }, { "epoch": 1.1463293505886476, "grad_norm": 2.946922540664673, "learning_rate": 2.9941293849232253e-05, "loss": 1.8101, "step": 4905 }, { "epoch": 1.1465630568782683, "grad_norm": 2.919825315475464, "learning_rate": 2.994124365219112e-05, "loss": 1.9144, "step": 4906 }, { "epoch": 1.1467967631678888, "grad_norm": 3.064405918121338, "learning_rate": 2.994119343374063e-05, "loss": 1.7553, "step": 4907 }, { "epoch": 1.1470304694575093, "grad_norm": 3.5507736206054688, "learning_rate": 2.994114319388086e-05, "loss": 1.847, "step": 4908 }, { "epoch": 1.1472641757471298, "grad_norm": 2.4889075756073, "learning_rate": 2.9941092932611873e-05, "loss": 1.8314, "step": 4909 }, { "epoch": 1.1474978820367503, "grad_norm": 2.538064479827881, "learning_rate": 2.9941042649933756e-05, "loss": 1.8762, "step": 4910 }, { "epoch": 1.1477315883263708, "grad_norm": 2.605147361755371, "learning_rate": 2.994099234584657e-05, "loss": 1.9652, "step": 4911 }, { "epoch": 1.1479652946159913, "grad_norm": 2.862868309020996, "learning_rate": 2.9940942020350388e-05, "loss": 1.9447, "step": 4912 }, { "epoch": 1.1481990009056118, "grad_norm": 2.638637065887451, "learning_rate": 2.9940891673445287e-05, "loss": 1.8845, "step": 4913 }, { "epoch": 1.1484327071952323, "grad_norm": 3.5457653999328613, "learning_rate": 2.9940841305131336e-05, "loss": 1.8621, "step": 4914 }, { "epoch": 1.148666413484853, "grad_norm": 2.671140193939209, "learning_rate": 2.9940790915408607e-05, "loss": 2.0525, "step": 4915 }, { "epoch": 1.1489001197744735, "grad_norm": 3.281653642654419, "learning_rate": 2.994074050427717e-05, "loss": 1.7934, "step": 4916 }, { "epoch": 1.149133826064094, "grad_norm": 2.9837818145751953, "learning_rate": 2.99406900717371e-05, "loss": 1.8593, "step": 4917 }, { "epoch": 1.1493675323537145, "grad_norm": 2.3966150283813477, "learning_rate": 2.994063961778847e-05, "loss": 1.9749, "step": 4918 }, { "epoch": 1.149601238643335, "grad_norm": 3.6396124362945557, "learning_rate": 2.9940589142431354e-05, "loss": 1.8042, "step": 4919 }, { "epoch": 1.1498349449329555, "grad_norm": 2.9052135944366455, "learning_rate": 2.994053864566582e-05, "loss": 1.975, "step": 4920 }, { "epoch": 1.150068651222576, "grad_norm": 2.733325719833374, "learning_rate": 2.994048812749194e-05, "loss": 1.849, "step": 4921 }, { "epoch": 1.1503023575121967, "grad_norm": 2.998831272125244, "learning_rate": 2.9940437587909794e-05, "loss": 1.8104, "step": 4922 }, { "epoch": 1.150536063801817, "grad_norm": 2.4916188716888428, "learning_rate": 2.9940387026919442e-05, "loss": 1.8987, "step": 4923 }, { "epoch": 1.1507697700914377, "grad_norm": 2.8902182579040527, "learning_rate": 2.994033644452097e-05, "loss": 1.8592, "step": 4924 }, { "epoch": 1.1510034763810582, "grad_norm": 2.520920515060425, "learning_rate": 2.9940285840714438e-05, "loss": 2.0056, "step": 4925 }, { "epoch": 1.1512371826706786, "grad_norm": 2.748351812362671, "learning_rate": 2.9940235215499924e-05, "loss": 1.8302, "step": 4926 }, { "epoch": 1.1514708889602991, "grad_norm": 3.045684814453125, "learning_rate": 2.994018456887751e-05, "loss": 1.847, "step": 4927 }, { "epoch": 1.1517045952499196, "grad_norm": 2.992365598678589, "learning_rate": 2.994013390084725e-05, "loss": 1.9673, "step": 4928 }, { "epoch": 1.1519383015395401, "grad_norm": 4.621305465698242, "learning_rate": 2.994008321140923e-05, "loss": 1.9238, "step": 4929 }, { "epoch": 1.1521720078291606, "grad_norm": 6.386037349700928, "learning_rate": 2.9940032500563514e-05, "loss": 1.871, "step": 4930 }, { "epoch": 1.1524057141187813, "grad_norm": 3.2887730598449707, "learning_rate": 2.9939981768310182e-05, "loss": 2.0211, "step": 4931 }, { "epoch": 1.1526394204084018, "grad_norm": 4.496822834014893, "learning_rate": 2.993993101464931e-05, "loss": 1.9587, "step": 4932 }, { "epoch": 1.1528731266980223, "grad_norm": 4.261423587799072, "learning_rate": 2.9939880239580958e-05, "loss": 2.0945, "step": 4933 }, { "epoch": 1.1531068329876428, "grad_norm": 2.9836606979370117, "learning_rate": 2.9939829443105205e-05, "loss": 1.9244, "step": 4934 }, { "epoch": 1.1533405392772633, "grad_norm": 2.492703437805176, "learning_rate": 2.9939778625222128e-05, "loss": 1.9361, "step": 4935 }, { "epoch": 1.1535742455668838, "grad_norm": 3.613537549972534, "learning_rate": 2.9939727785931793e-05, "loss": 1.9652, "step": 4936 }, { "epoch": 1.1538079518565043, "grad_norm": 2.9863994121551514, "learning_rate": 2.993967692523428e-05, "loss": 1.8324, "step": 4937 }, { "epoch": 1.1540416581461248, "grad_norm": 2.7126622200012207, "learning_rate": 2.9939626043129655e-05, "loss": 1.9148, "step": 4938 }, { "epoch": 1.1542753644357453, "grad_norm": 5.127562046051025, "learning_rate": 2.993957513961799e-05, "loss": 1.686, "step": 4939 }, { "epoch": 1.154509070725366, "grad_norm": 3.8987412452697754, "learning_rate": 2.9939524214699368e-05, "loss": 1.7317, "step": 4940 }, { "epoch": 1.1547427770149865, "grad_norm": 2.485335111618042, "learning_rate": 2.9939473268373854e-05, "loss": 1.8823, "step": 4941 }, { "epoch": 1.154976483304607, "grad_norm": 3.0372915267944336, "learning_rate": 2.9939422300641524e-05, "loss": 1.9169, "step": 4942 }, { "epoch": 1.1552101895942275, "grad_norm": 4.01839017868042, "learning_rate": 2.9939371311502445e-05, "loss": 1.8626, "step": 4943 }, { "epoch": 1.155443895883848, "grad_norm": 2.2769410610198975, "learning_rate": 2.99393203009567e-05, "loss": 1.8933, "step": 4944 }, { "epoch": 1.1556776021734685, "grad_norm": 3.7040555477142334, "learning_rate": 2.9939269269004354e-05, "loss": 1.8945, "step": 4945 }, { "epoch": 1.155911308463089, "grad_norm": 3.3034908771514893, "learning_rate": 2.993921821564548e-05, "loss": 1.8264, "step": 4946 }, { "epoch": 1.1561450147527095, "grad_norm": 2.6885814666748047, "learning_rate": 2.9939167140880156e-05, "loss": 1.9804, "step": 4947 }, { "epoch": 1.15637872104233, "grad_norm": 3.1934118270874023, "learning_rate": 2.9939116044708455e-05, "loss": 1.8523, "step": 4948 }, { "epoch": 1.1566124273319507, "grad_norm": 3.590927839279175, "learning_rate": 2.9939064927130454e-05, "loss": 1.832, "step": 4949 }, { "epoch": 1.1568461336215712, "grad_norm": 2.9860010147094727, "learning_rate": 2.9939013788146212e-05, "loss": 1.7897, "step": 4950 }, { "epoch": 1.1570798399111917, "grad_norm": 2.191967487335205, "learning_rate": 2.9938962627755814e-05, "loss": 1.8494, "step": 4951 }, { "epoch": 1.1573135462008122, "grad_norm": 4.2651824951171875, "learning_rate": 2.9938911445959325e-05, "loss": 1.9342, "step": 4952 }, { "epoch": 1.1575472524904327, "grad_norm": 4.862129211425781, "learning_rate": 2.9938860242756837e-05, "loss": 1.7499, "step": 4953 }, { "epoch": 1.1577809587800532, "grad_norm": 2.7454206943511963, "learning_rate": 2.99388090181484e-05, "loss": 1.8789, "step": 4954 }, { "epoch": 1.1580146650696737, "grad_norm": 4.043206214904785, "learning_rate": 2.9938757772134094e-05, "loss": 1.7875, "step": 4955 }, { "epoch": 1.1582483713592941, "grad_norm": 3.0007686614990234, "learning_rate": 2.9938706504714006e-05, "loss": 1.8336, "step": 4956 }, { "epoch": 1.1584820776489146, "grad_norm": 2.8830795288085938, "learning_rate": 2.9938655215888192e-05, "loss": 1.7786, "step": 4957 }, { "epoch": 1.1587157839385354, "grad_norm": 2.618178606033325, "learning_rate": 2.9938603905656738e-05, "loss": 1.9851, "step": 4958 }, { "epoch": 1.1589494902281559, "grad_norm": 4.514334678649902, "learning_rate": 2.9938552574019707e-05, "loss": 1.9462, "step": 4959 }, { "epoch": 1.1591831965177763, "grad_norm": 2.6963818073272705, "learning_rate": 2.9938501220977185e-05, "loss": 1.7867, "step": 4960 }, { "epoch": 1.1594169028073968, "grad_norm": 2.376190185546875, "learning_rate": 2.9938449846529232e-05, "loss": 1.8627, "step": 4961 }, { "epoch": 1.1596506090970173, "grad_norm": 3.370925188064575, "learning_rate": 2.9938398450675927e-05, "loss": 1.9322, "step": 4962 }, { "epoch": 1.1598843153866378, "grad_norm": 2.465212821960449, "learning_rate": 2.993834703341735e-05, "loss": 1.9125, "step": 4963 }, { "epoch": 1.1601180216762583, "grad_norm": 2.881185531616211, "learning_rate": 2.993829559475357e-05, "loss": 1.8588, "step": 4964 }, { "epoch": 1.1603517279658788, "grad_norm": 2.569396734237671, "learning_rate": 2.9938244134684657e-05, "loss": 2.0081, "step": 4965 }, { "epoch": 1.1605854342554993, "grad_norm": 3.093289375305176, "learning_rate": 2.9938192653210686e-05, "loss": 1.8082, "step": 4966 }, { "epoch": 1.16081914054512, "grad_norm": 2.689709424972534, "learning_rate": 2.9938141150331735e-05, "loss": 1.8234, "step": 4967 }, { "epoch": 1.1610528468347405, "grad_norm": 3.1806013584136963, "learning_rate": 2.9938089626047876e-05, "loss": 1.9591, "step": 4968 }, { "epoch": 1.161286553124361, "grad_norm": 3.5953288078308105, "learning_rate": 2.9938038080359182e-05, "loss": 1.7958, "step": 4969 }, { "epoch": 1.1615202594139815, "grad_norm": 2.663175106048584, "learning_rate": 2.993798651326573e-05, "loss": 1.931, "step": 4970 }, { "epoch": 1.161753965703602, "grad_norm": 3.1595726013183594, "learning_rate": 2.9937934924767587e-05, "loss": 1.8798, "step": 4971 }, { "epoch": 1.1619876719932225, "grad_norm": 3.4886441230773926, "learning_rate": 2.9937883314864834e-05, "loss": 1.8487, "step": 4972 }, { "epoch": 1.162221378282843, "grad_norm": 2.605485200881958, "learning_rate": 2.9937831683557538e-05, "loss": 1.7599, "step": 4973 }, { "epoch": 1.1624550845724635, "grad_norm": 2.9974069595336914, "learning_rate": 2.9937780030845777e-05, "loss": 1.8602, "step": 4974 }, { "epoch": 1.162688790862084, "grad_norm": 2.7362592220306396, "learning_rate": 2.9937728356729626e-05, "loss": 1.9177, "step": 4975 }, { "epoch": 1.1629224971517047, "grad_norm": 2.507800340652466, "learning_rate": 2.993767666120916e-05, "loss": 1.8212, "step": 4976 }, { "epoch": 1.1631562034413252, "grad_norm": 2.679985761642456, "learning_rate": 2.993762494428445e-05, "loss": 1.9689, "step": 4977 }, { "epoch": 1.1633899097309457, "grad_norm": 2.291863203048706, "learning_rate": 2.993757320595557e-05, "loss": 1.8311, "step": 4978 }, { "epoch": 1.1636236160205662, "grad_norm": 2.934187412261963, "learning_rate": 2.9937521446222594e-05, "loss": 1.8788, "step": 4979 }, { "epoch": 1.1638573223101867, "grad_norm": 3.1400890350341797, "learning_rate": 2.99374696650856e-05, "loss": 1.7941, "step": 4980 }, { "epoch": 1.1640910285998072, "grad_norm": 3.4715723991394043, "learning_rate": 2.993741786254466e-05, "loss": 2.0255, "step": 4981 }, { "epoch": 1.1643247348894277, "grad_norm": 2.3359878063201904, "learning_rate": 2.9937366038599846e-05, "loss": 1.7746, "step": 4982 }, { "epoch": 1.1645584411790482, "grad_norm": 3.0084640979766846, "learning_rate": 2.9937314193251238e-05, "loss": 1.8131, "step": 4983 }, { "epoch": 1.1647921474686687, "grad_norm": 2.4924862384796143, "learning_rate": 2.99372623264989e-05, "loss": 1.8396, "step": 4984 }, { "epoch": 1.1650258537582894, "grad_norm": 2.780672073364258, "learning_rate": 2.993721043834292e-05, "loss": 1.7922, "step": 4985 }, { "epoch": 1.1652595600479099, "grad_norm": 2.472055435180664, "learning_rate": 2.9937158528783357e-05, "loss": 1.7726, "step": 4986 }, { "epoch": 1.1654932663375304, "grad_norm": 2.593956470489502, "learning_rate": 2.9937106597820297e-05, "loss": 1.9707, "step": 4987 }, { "epoch": 1.1657269726271509, "grad_norm": 2.3817107677459717, "learning_rate": 2.9937054645453813e-05, "loss": 1.8373, "step": 4988 }, { "epoch": 1.1659606789167714, "grad_norm": 2.2850756645202637, "learning_rate": 2.9937002671683975e-05, "loss": 1.8367, "step": 4989 }, { "epoch": 1.1661943852063918, "grad_norm": 2.584538459777832, "learning_rate": 2.993695067651086e-05, "loss": 1.9197, "step": 4990 }, { "epoch": 1.1664280914960123, "grad_norm": 2.5685670375823975, "learning_rate": 2.9936898659934542e-05, "loss": 1.8024, "step": 4991 }, { "epoch": 1.1666617977856328, "grad_norm": 2.5498344898223877, "learning_rate": 2.9936846621955097e-05, "loss": 1.9822, "step": 4992 }, { "epoch": 1.1668955040752533, "grad_norm": 2.8127667903900146, "learning_rate": 2.9936794562572597e-05, "loss": 1.9292, "step": 4993 }, { "epoch": 1.167129210364874, "grad_norm": 3.206029176712036, "learning_rate": 2.9936742481787122e-05, "loss": 1.7607, "step": 4994 }, { "epoch": 1.1673629166544945, "grad_norm": 4.568969249725342, "learning_rate": 2.9936690379598737e-05, "loss": 1.9268, "step": 4995 }, { "epoch": 1.167596622944115, "grad_norm": 3.3478946685791016, "learning_rate": 2.9936638256007527e-05, "loss": 1.7804, "step": 4996 }, { "epoch": 1.1678303292337355, "grad_norm": 2.42445707321167, "learning_rate": 2.9936586111013557e-05, "loss": 1.8761, "step": 4997 }, { "epoch": 1.168064035523356, "grad_norm": 2.383197784423828, "learning_rate": 2.993653394461691e-05, "loss": 1.7504, "step": 4998 }, { "epoch": 1.1682977418129765, "grad_norm": 3.509746789932251, "learning_rate": 2.9936481756817655e-05, "loss": 1.8322, "step": 4999 }, { "epoch": 1.168531448102597, "grad_norm": 3.12420392036438, "learning_rate": 2.993642954761587e-05, "loss": 1.9458, "step": 5000 }, { "epoch": 1.1687651543922175, "grad_norm": 2.6161656379699707, "learning_rate": 2.993637731701163e-05, "loss": 1.9343, "step": 5001 }, { "epoch": 1.168998860681838, "grad_norm": 2.4736294746398926, "learning_rate": 2.993632506500501e-05, "loss": 2.0003, "step": 5002 }, { "epoch": 1.1692325669714587, "grad_norm": 2.642141342163086, "learning_rate": 2.9936272791596084e-05, "loss": 1.8111, "step": 5003 }, { "epoch": 1.1694662732610792, "grad_norm": 2.433363914489746, "learning_rate": 2.9936220496784926e-05, "loss": 2.0401, "step": 5004 }, { "epoch": 1.1696999795506997, "grad_norm": 2.5433170795440674, "learning_rate": 2.993616818057161e-05, "loss": 1.812, "step": 5005 }, { "epoch": 1.1699336858403202, "grad_norm": 4.522892475128174, "learning_rate": 2.9936115842956212e-05, "loss": 1.8331, "step": 5006 }, { "epoch": 1.1701673921299407, "grad_norm": 2.4926631450653076, "learning_rate": 2.9936063483938808e-05, "loss": 1.8141, "step": 5007 }, { "epoch": 1.1704010984195612, "grad_norm": 3.0272083282470703, "learning_rate": 2.993601110351947e-05, "loss": 1.9285, "step": 5008 }, { "epoch": 1.1706348047091817, "grad_norm": 2.5249812602996826, "learning_rate": 2.9935958701698283e-05, "loss": 1.7819, "step": 5009 }, { "epoch": 1.1708685109988022, "grad_norm": 2.739302396774292, "learning_rate": 2.9935906278475308e-05, "loss": 1.9878, "step": 5010 }, { "epoch": 1.1711022172884227, "grad_norm": 2.312244176864624, "learning_rate": 2.993585383385063e-05, "loss": 1.7907, "step": 5011 }, { "epoch": 1.1713359235780434, "grad_norm": 2.2818775177001953, "learning_rate": 2.993580136782432e-05, "loss": 1.9103, "step": 5012 }, { "epoch": 1.1715696298676639, "grad_norm": 2.9249069690704346, "learning_rate": 2.9935748880396457e-05, "loss": 1.8655, "step": 5013 }, { "epoch": 1.1718033361572844, "grad_norm": 2.7421224117279053, "learning_rate": 2.9935696371567107e-05, "loss": 1.7971, "step": 5014 }, { "epoch": 1.1720370424469049, "grad_norm": 2.9082465171813965, "learning_rate": 2.9935643841336355e-05, "loss": 1.8865, "step": 5015 }, { "epoch": 1.1722707487365254, "grad_norm": 2.6128997802734375, "learning_rate": 2.9935591289704273e-05, "loss": 1.8306, "step": 5016 }, { "epoch": 1.1725044550261459, "grad_norm": 2.6009743213653564, "learning_rate": 2.9935538716670937e-05, "loss": 1.9068, "step": 5017 }, { "epoch": 1.1727381613157664, "grad_norm": 2.7381293773651123, "learning_rate": 2.993548612223642e-05, "loss": 1.9332, "step": 5018 }, { "epoch": 1.1729718676053869, "grad_norm": 3.306993007659912, "learning_rate": 2.99354335064008e-05, "loss": 1.7315, "step": 5019 }, { "epoch": 1.1732055738950073, "grad_norm": 2.8062009811401367, "learning_rate": 2.9935380869164153e-05, "loss": 1.8527, "step": 5020 }, { "epoch": 1.173439280184628, "grad_norm": 3.2301578521728516, "learning_rate": 2.9935328210526547e-05, "loss": 1.9734, "step": 5021 }, { "epoch": 1.1736729864742486, "grad_norm": 2.5307817459106445, "learning_rate": 2.9935275530488067e-05, "loss": 1.788, "step": 5022 }, { "epoch": 1.173906692763869, "grad_norm": 2.8355813026428223, "learning_rate": 2.9935222829048786e-05, "loss": 1.8374, "step": 5023 }, { "epoch": 1.1741403990534895, "grad_norm": 2.281914234161377, "learning_rate": 2.9935170106208772e-05, "loss": 1.9541, "step": 5024 }, { "epoch": 1.17437410534311, "grad_norm": 2.9005184173583984, "learning_rate": 2.9935117361968113e-05, "loss": 1.9069, "step": 5025 }, { "epoch": 1.1746078116327305, "grad_norm": 2.8345439434051514, "learning_rate": 2.9935064596326874e-05, "loss": 1.8043, "step": 5026 }, { "epoch": 1.174841517922351, "grad_norm": 2.858928680419922, "learning_rate": 2.9935011809285138e-05, "loss": 1.8804, "step": 5027 }, { "epoch": 1.1750752242119715, "grad_norm": 3.033958673477173, "learning_rate": 2.993495900084298e-05, "loss": 1.8436, "step": 5028 }, { "epoch": 1.175308930501592, "grad_norm": 3.6013739109039307, "learning_rate": 2.993490617100047e-05, "loss": 1.855, "step": 5029 }, { "epoch": 1.1755426367912127, "grad_norm": 3.7911911010742188, "learning_rate": 2.9934853319757684e-05, "loss": 1.9096, "step": 5030 }, { "epoch": 1.1757763430808332, "grad_norm": 2.964615821838379, "learning_rate": 2.9934800447114703e-05, "loss": 1.8261, "step": 5031 }, { "epoch": 1.1760100493704537, "grad_norm": 3.5566904544830322, "learning_rate": 2.99347475530716e-05, "loss": 2.0355, "step": 5032 }, { "epoch": 1.1762437556600742, "grad_norm": 5.6918745040893555, "learning_rate": 2.9934694637628453e-05, "loss": 1.9573, "step": 5033 }, { "epoch": 1.1764774619496947, "grad_norm": 5.048044681549072, "learning_rate": 2.993464170078533e-05, "loss": 1.996, "step": 5034 }, { "epoch": 1.1767111682393152, "grad_norm": 3.4721639156341553, "learning_rate": 2.993458874254232e-05, "loss": 1.9634, "step": 5035 }, { "epoch": 1.1769448745289357, "grad_norm": 3.116865396499634, "learning_rate": 2.9934535762899487e-05, "loss": 1.7836, "step": 5036 }, { "epoch": 1.1771785808185562, "grad_norm": 3.5251970291137695, "learning_rate": 2.9934482761856913e-05, "loss": 1.9674, "step": 5037 }, { "epoch": 1.1774122871081767, "grad_norm": 2.8563473224639893, "learning_rate": 2.9934429739414676e-05, "loss": 1.8467, "step": 5038 }, { "epoch": 1.1776459933977974, "grad_norm": 4.125290393829346, "learning_rate": 2.993437669557284e-05, "loss": 1.9872, "step": 5039 }, { "epoch": 1.177879699687418, "grad_norm": 4.774967193603516, "learning_rate": 2.9934323630331503e-05, "loss": 1.736, "step": 5040 }, { "epoch": 1.1781134059770384, "grad_norm": 2.5418426990509033, "learning_rate": 2.9934270543690717e-05, "loss": 1.8586, "step": 5041 }, { "epoch": 1.178347112266659, "grad_norm": 3.9297423362731934, "learning_rate": 2.993421743565057e-05, "loss": 1.7969, "step": 5042 }, { "epoch": 1.1785808185562794, "grad_norm": 4.731308937072754, "learning_rate": 2.993416430621114e-05, "loss": 1.8607, "step": 5043 }, { "epoch": 1.1788145248458999, "grad_norm": 3.740420341491699, "learning_rate": 2.9934111155372496e-05, "loss": 2.0478, "step": 5044 }, { "epoch": 1.1790482311355204, "grad_norm": 3.8444416522979736, "learning_rate": 2.993405798313472e-05, "loss": 1.9788, "step": 5045 }, { "epoch": 1.1792819374251409, "grad_norm": 3.750337600708008, "learning_rate": 2.9934004789497885e-05, "loss": 1.9181, "step": 5046 }, { "epoch": 1.1795156437147614, "grad_norm": 3.049084424972534, "learning_rate": 2.993395157446207e-05, "loss": 1.8284, "step": 5047 }, { "epoch": 1.179749350004382, "grad_norm": 2.7965893745422363, "learning_rate": 2.993389833802735e-05, "loss": 1.8417, "step": 5048 }, { "epoch": 1.1799830562940026, "grad_norm": 4.596904277801514, "learning_rate": 2.99338450801938e-05, "loss": 1.8861, "step": 5049 }, { "epoch": 1.180216762583623, "grad_norm": 2.9704952239990234, "learning_rate": 2.9933791800961497e-05, "loss": 1.9348, "step": 5050 }, { "epoch": 1.1804504688732436, "grad_norm": 2.921517848968506, "learning_rate": 2.993373850033052e-05, "loss": 1.8552, "step": 5051 }, { "epoch": 1.180684175162864, "grad_norm": 3.199302911758423, "learning_rate": 2.9933685178300937e-05, "loss": 1.8386, "step": 5052 }, { "epoch": 1.1809178814524846, "grad_norm": 3.0136892795562744, "learning_rate": 2.9933631834872836e-05, "loss": 2.0654, "step": 5053 }, { "epoch": 1.181151587742105, "grad_norm": 3.487574815750122, "learning_rate": 2.9933578470046288e-05, "loss": 1.8868, "step": 5054 }, { "epoch": 1.1813852940317255, "grad_norm": 3.669865131378174, "learning_rate": 2.993352508382137e-05, "loss": 1.8053, "step": 5055 }, { "epoch": 1.181619000321346, "grad_norm": 2.4492335319519043, "learning_rate": 2.9933471676198153e-05, "loss": 1.9267, "step": 5056 }, { "epoch": 1.1818527066109668, "grad_norm": 2.775965929031372, "learning_rate": 2.993341824717672e-05, "loss": 1.7494, "step": 5057 }, { "epoch": 1.1820864129005872, "grad_norm": 3.871509313583374, "learning_rate": 2.993336479675715e-05, "loss": 1.8207, "step": 5058 }, { "epoch": 1.1823201191902077, "grad_norm": 2.942333698272705, "learning_rate": 2.9933311324939508e-05, "loss": 1.9277, "step": 5059 }, { "epoch": 1.1825538254798282, "grad_norm": 3.144721746444702, "learning_rate": 2.9933257831723886e-05, "loss": 1.8192, "step": 5060 }, { "epoch": 1.1827875317694487, "grad_norm": 3.770504951477051, "learning_rate": 2.993320431711035e-05, "loss": 2.0572, "step": 5061 }, { "epoch": 1.1830212380590692, "grad_norm": 2.60502552986145, "learning_rate": 2.9933150781098978e-05, "loss": 1.8604, "step": 5062 }, { "epoch": 1.1832549443486897, "grad_norm": 4.20154333114624, "learning_rate": 2.9933097223689846e-05, "loss": 1.8828, "step": 5063 }, { "epoch": 1.1834886506383102, "grad_norm": 3.4495956897735596, "learning_rate": 2.9933043644883036e-05, "loss": 1.8893, "step": 5064 }, { "epoch": 1.1837223569279307, "grad_norm": 2.813206672668457, "learning_rate": 2.9932990044678622e-05, "loss": 1.9404, "step": 5065 }, { "epoch": 1.1839560632175514, "grad_norm": 16.298553466796875, "learning_rate": 2.9932936423076682e-05, "loss": 2.1776, "step": 5066 }, { "epoch": 1.184189769507172, "grad_norm": 2.630251407623291, "learning_rate": 2.9932882780077287e-05, "loss": 1.88, "step": 5067 }, { "epoch": 1.1844234757967924, "grad_norm": 2.884572982788086, "learning_rate": 2.993282911568052e-05, "loss": 1.7622, "step": 5068 }, { "epoch": 1.184657182086413, "grad_norm": 2.349046468734741, "learning_rate": 2.9932775429886457e-05, "loss": 1.9085, "step": 5069 }, { "epoch": 1.1848908883760334, "grad_norm": 2.7503597736358643, "learning_rate": 2.9932721722695175e-05, "loss": 2.0494, "step": 5070 }, { "epoch": 1.185124594665654, "grad_norm": 2.7544760704040527, "learning_rate": 2.9932667994106746e-05, "loss": 1.733, "step": 5071 }, { "epoch": 1.1853583009552744, "grad_norm": 2.302659749984741, "learning_rate": 2.9932614244121255e-05, "loss": 1.8274, "step": 5072 }, { "epoch": 1.1855920072448949, "grad_norm": 4.168538570404053, "learning_rate": 2.9932560472738774e-05, "loss": 1.8804, "step": 5073 }, { "epoch": 1.1858257135345154, "grad_norm": 2.907733917236328, "learning_rate": 2.9932506679959376e-05, "loss": 1.9039, "step": 5074 }, { "epoch": 1.186059419824136, "grad_norm": 2.886643409729004, "learning_rate": 2.993245286578315e-05, "loss": 1.902, "step": 5075 }, { "epoch": 1.1862931261137566, "grad_norm": 3.097490072250366, "learning_rate": 2.9932399030210164e-05, "loss": 1.9088, "step": 5076 }, { "epoch": 1.186526832403377, "grad_norm": 2.4092891216278076, "learning_rate": 2.99323451732405e-05, "loss": 1.8356, "step": 5077 }, { "epoch": 1.1867605386929976, "grad_norm": 2.9882287979125977, "learning_rate": 2.9932291294874226e-05, "loss": 1.9594, "step": 5078 }, { "epoch": 1.186994244982618, "grad_norm": 2.2680184841156006, "learning_rate": 2.993223739511143e-05, "loss": 1.7989, "step": 5079 }, { "epoch": 1.1872279512722386, "grad_norm": 3.1557557582855225, "learning_rate": 2.9932183473952184e-05, "loss": 1.8215, "step": 5080 }, { "epoch": 1.187461657561859, "grad_norm": 2.6368489265441895, "learning_rate": 2.9932129531396565e-05, "loss": 1.8496, "step": 5081 }, { "epoch": 1.1876953638514796, "grad_norm": 3.331880569458008, "learning_rate": 2.9932075567444653e-05, "loss": 1.8864, "step": 5082 }, { "epoch": 1.1879290701411, "grad_norm": 3.0686378479003906, "learning_rate": 2.9932021582096523e-05, "loss": 1.7681, "step": 5083 }, { "epoch": 1.1881627764307208, "grad_norm": 2.4160006046295166, "learning_rate": 2.9931967575352256e-05, "loss": 1.8473, "step": 5084 }, { "epoch": 1.1883964827203413, "grad_norm": 2.873589515686035, "learning_rate": 2.9931913547211923e-05, "loss": 2.031, "step": 5085 }, { "epoch": 1.1886301890099618, "grad_norm": 3.360154151916504, "learning_rate": 2.9931859497675606e-05, "loss": 1.9612, "step": 5086 }, { "epoch": 1.1888638952995823, "grad_norm": 3.8029332160949707, "learning_rate": 2.9931805426743382e-05, "loss": 1.9232, "step": 5087 }, { "epoch": 1.1890976015892027, "grad_norm": 2.5671679973602295, "learning_rate": 2.993175133441533e-05, "loss": 1.9066, "step": 5088 }, { "epoch": 1.1893313078788232, "grad_norm": 2.8512144088745117, "learning_rate": 2.9931697220691523e-05, "loss": 1.9903, "step": 5089 }, { "epoch": 1.1895650141684437, "grad_norm": 2.4718406200408936, "learning_rate": 2.993164308557204e-05, "loss": 2.0, "step": 5090 }, { "epoch": 1.1897987204580645, "grad_norm": 2.6868979930877686, "learning_rate": 2.993158892905696e-05, "loss": 1.8585, "step": 5091 }, { "epoch": 1.1900324267476847, "grad_norm": 4.064939022064209, "learning_rate": 2.993153475114636e-05, "loss": 1.7759, "step": 5092 }, { "epoch": 1.1902661330373054, "grad_norm": 3.06034779548645, "learning_rate": 2.993148055184032e-05, "loss": 1.8446, "step": 5093 }, { "epoch": 1.190499839326926, "grad_norm": 2.5708463191986084, "learning_rate": 2.9931426331138915e-05, "loss": 2.0019, "step": 5094 }, { "epoch": 1.1907335456165464, "grad_norm": 4.31316614151001, "learning_rate": 2.993137208904222e-05, "loss": 1.9225, "step": 5095 }, { "epoch": 1.190967251906167, "grad_norm": 4.441995143890381, "learning_rate": 2.993131782555032e-05, "loss": 1.8505, "step": 5096 }, { "epoch": 1.1912009581957874, "grad_norm": 2.5422017574310303, "learning_rate": 2.9931263540663286e-05, "loss": 1.8961, "step": 5097 }, { "epoch": 1.191434664485408, "grad_norm": 4.219817638397217, "learning_rate": 2.9931209234381196e-05, "loss": 1.8212, "step": 5098 }, { "epoch": 1.1916683707750284, "grad_norm": 4.5107269287109375, "learning_rate": 2.9931154906704135e-05, "loss": 1.8525, "step": 5099 }, { "epoch": 1.1919020770646491, "grad_norm": 2.856609344482422, "learning_rate": 2.9931100557632176e-05, "loss": 1.9518, "step": 5100 }, { "epoch": 1.1921357833542696, "grad_norm": 3.7280211448669434, "learning_rate": 2.9931046187165395e-05, "loss": 1.7709, "step": 5101 }, { "epoch": 1.1923694896438901, "grad_norm": 3.8663346767425537, "learning_rate": 2.9930991795303873e-05, "loss": 1.8477, "step": 5102 }, { "epoch": 1.1926031959335106, "grad_norm": 2.4834084510803223, "learning_rate": 2.9930937382047685e-05, "loss": 1.9031, "step": 5103 }, { "epoch": 1.192836902223131, "grad_norm": 3.672680139541626, "learning_rate": 2.993088294739691e-05, "loss": 1.9106, "step": 5104 }, { "epoch": 1.1930706085127516, "grad_norm": 2.409191370010376, "learning_rate": 2.9930828491351627e-05, "loss": 1.8389, "step": 5105 }, { "epoch": 1.193304314802372, "grad_norm": 2.297537326812744, "learning_rate": 2.9930774013911917e-05, "loss": 1.9654, "step": 5106 }, { "epoch": 1.1935380210919926, "grad_norm": 3.7799019813537598, "learning_rate": 2.9930719515077852e-05, "loss": 1.8902, "step": 5107 }, { "epoch": 1.193771727381613, "grad_norm": 3.25345778465271, "learning_rate": 2.9930664994849513e-05, "loss": 1.8224, "step": 5108 }, { "epoch": 1.1940054336712338, "grad_norm": 4.5086259841918945, "learning_rate": 2.993061045322698e-05, "loss": 1.8373, "step": 5109 }, { "epoch": 1.1942391399608543, "grad_norm": 3.1507809162139893, "learning_rate": 2.9930555890210324e-05, "loss": 1.8875, "step": 5110 }, { "epoch": 1.1944728462504748, "grad_norm": 2.5572969913482666, "learning_rate": 2.9930501305799633e-05, "loss": 1.8709, "step": 5111 }, { "epoch": 1.1947065525400953, "grad_norm": 2.918262243270874, "learning_rate": 2.9930446699994983e-05, "loss": 1.9147, "step": 5112 }, { "epoch": 1.1949402588297158, "grad_norm": 3.531287431716919, "learning_rate": 2.9930392072796443e-05, "loss": 2.0042, "step": 5113 }, { "epoch": 1.1951739651193363, "grad_norm": 3.043297052383423, "learning_rate": 2.9930337424204107e-05, "loss": 1.8344, "step": 5114 }, { "epoch": 1.1954076714089568, "grad_norm": 2.4809062480926514, "learning_rate": 2.993028275421804e-05, "loss": 1.8084, "step": 5115 }, { "epoch": 1.1956413776985773, "grad_norm": 5.736629486083984, "learning_rate": 2.9930228062838322e-05, "loss": 1.8821, "step": 5116 }, { "epoch": 1.1958750839881978, "grad_norm": 3.7259178161621094, "learning_rate": 2.993017335006504e-05, "loss": 1.9355, "step": 5117 }, { "epoch": 1.1961087902778185, "grad_norm": 2.761483907699585, "learning_rate": 2.993011861589826e-05, "loss": 1.9745, "step": 5118 }, { "epoch": 1.196342496567439, "grad_norm": 3.7936313152313232, "learning_rate": 2.9930063860338065e-05, "loss": 1.8445, "step": 5119 }, { "epoch": 1.1965762028570595, "grad_norm": 4.506153583526611, "learning_rate": 2.9930009083384543e-05, "loss": 1.8231, "step": 5120 }, { "epoch": 1.19680990914668, "grad_norm": 3.6453564167022705, "learning_rate": 2.992995428503776e-05, "loss": 1.8494, "step": 5121 }, { "epoch": 1.1970436154363004, "grad_norm": 4.464837074279785, "learning_rate": 2.9929899465297804e-05, "loss": 1.9527, "step": 5122 }, { "epoch": 1.197277321725921, "grad_norm": 5.722427845001221, "learning_rate": 2.9929844624164745e-05, "loss": 1.8302, "step": 5123 }, { "epoch": 1.1975110280155414, "grad_norm": 4.517265319824219, "learning_rate": 2.992978976163867e-05, "loss": 1.8419, "step": 5124 }, { "epoch": 1.197744734305162, "grad_norm": 2.5028843879699707, "learning_rate": 2.992973487771965e-05, "loss": 1.8638, "step": 5125 }, { "epoch": 1.1979784405947824, "grad_norm": 4.422055721282959, "learning_rate": 2.9929679972407764e-05, "loss": 1.8477, "step": 5126 }, { "epoch": 1.1982121468844031, "grad_norm": 4.950786590576172, "learning_rate": 2.9929625045703096e-05, "loss": 1.8459, "step": 5127 }, { "epoch": 1.1984458531740236, "grad_norm": 4.446730136871338, "learning_rate": 2.9929570097605724e-05, "loss": 1.8684, "step": 5128 }, { "epoch": 1.1986795594636441, "grad_norm": 2.5321006774902344, "learning_rate": 2.9929515128115725e-05, "loss": 1.8459, "step": 5129 }, { "epoch": 1.1989132657532646, "grad_norm": 3.9405083656311035, "learning_rate": 2.9929460137233174e-05, "loss": 1.7535, "step": 5130 }, { "epoch": 1.1991469720428851, "grad_norm": 4.370143890380859, "learning_rate": 2.9929405124958155e-05, "loss": 1.9717, "step": 5131 }, { "epoch": 1.1993806783325056, "grad_norm": 3.121629238128662, "learning_rate": 2.9929350091290746e-05, "loss": 1.9504, "step": 5132 }, { "epoch": 1.199614384622126, "grad_norm": 4.321176528930664, "learning_rate": 2.992929503623103e-05, "loss": 1.9157, "step": 5133 }, { "epoch": 1.1998480909117466, "grad_norm": 5.108989238739014, "learning_rate": 2.9929239959779075e-05, "loss": 1.9967, "step": 5134 }, { "epoch": 1.200081797201367, "grad_norm": 2.996492624282837, "learning_rate": 2.9929184861934965e-05, "loss": 1.7702, "step": 5135 }, { "epoch": 1.2003155034909878, "grad_norm": 2.479118585586548, "learning_rate": 2.9929129742698784e-05, "loss": 1.9224, "step": 5136 }, { "epoch": 1.2005492097806083, "grad_norm": 4.919881343841553, "learning_rate": 2.9929074602070602e-05, "loss": 1.6625, "step": 5137 }, { "epoch": 1.2007829160702288, "grad_norm": 4.036565780639648, "learning_rate": 2.992901944005051e-05, "loss": 1.6716, "step": 5138 }, { "epoch": 1.2010166223598493, "grad_norm": 2.722683906555176, "learning_rate": 2.992896425663857e-05, "loss": 1.8277, "step": 5139 }, { "epoch": 1.2012503286494698, "grad_norm": 3.450547218322754, "learning_rate": 2.9928909051834885e-05, "loss": 1.9062, "step": 5140 }, { "epoch": 1.2014840349390903, "grad_norm": 3.1411120891571045, "learning_rate": 2.9928853825639506e-05, "loss": 2.0594, "step": 5141 }, { "epoch": 1.2017177412287108, "grad_norm": 2.430792808532715, "learning_rate": 2.9928798578052534e-05, "loss": 1.8602, "step": 5142 }, { "epoch": 1.2019514475183313, "grad_norm": 2.1518778800964355, "learning_rate": 2.992874330907404e-05, "loss": 1.8282, "step": 5143 }, { "epoch": 1.2021851538079518, "grad_norm": 2.5724129676818848, "learning_rate": 2.99286880187041e-05, "loss": 1.6983, "step": 5144 }, { "epoch": 1.2024188600975725, "grad_norm": 2.3612494468688965, "learning_rate": 2.99286327069428e-05, "loss": 1.8768, "step": 5145 }, { "epoch": 1.202652566387193, "grad_norm": 2.6650447845458984, "learning_rate": 2.9928577373790216e-05, "loss": 1.764, "step": 5146 }, { "epoch": 1.2028862726768135, "grad_norm": 3.525217056274414, "learning_rate": 2.992852201924643e-05, "loss": 1.8626, "step": 5147 }, { "epoch": 1.203119978966434, "grad_norm": 3.3083808422088623, "learning_rate": 2.9928466643311515e-05, "loss": 2.002, "step": 5148 }, { "epoch": 1.2033536852560545, "grad_norm": 2.6805012226104736, "learning_rate": 2.9928411245985554e-05, "loss": 2.0464, "step": 5149 }, { "epoch": 1.203587391545675, "grad_norm": 2.309476375579834, "learning_rate": 2.9928355827268626e-05, "loss": 1.8852, "step": 5150 }, { "epoch": 1.2038210978352955, "grad_norm": 2.2254302501678467, "learning_rate": 2.9928300387160813e-05, "loss": 1.7683, "step": 5151 }, { "epoch": 1.204054804124916, "grad_norm": 3.4201345443725586, "learning_rate": 2.9928244925662192e-05, "loss": 1.7578, "step": 5152 }, { "epoch": 1.2042885104145364, "grad_norm": 2.6162667274475098, "learning_rate": 2.9928189442772842e-05, "loss": 1.8452, "step": 5153 }, { "epoch": 1.2045222167041572, "grad_norm": 2.4983608722686768, "learning_rate": 2.9928133938492843e-05, "loss": 1.8317, "step": 5154 }, { "epoch": 1.2047559229937777, "grad_norm": 2.7319726943969727, "learning_rate": 2.9928078412822275e-05, "loss": 1.7625, "step": 5155 }, { "epoch": 1.2049896292833981, "grad_norm": 3.0554163455963135, "learning_rate": 2.9928022865761217e-05, "loss": 1.8444, "step": 5156 }, { "epoch": 1.2052233355730186, "grad_norm": 2.599454879760742, "learning_rate": 2.992796729730975e-05, "loss": 1.9902, "step": 5157 }, { "epoch": 1.2054570418626391, "grad_norm": 2.8434395790100098, "learning_rate": 2.992791170746795e-05, "loss": 2.0797, "step": 5158 }, { "epoch": 1.2056907481522596, "grad_norm": 2.6320061683654785, "learning_rate": 2.9927856096235904e-05, "loss": 1.9299, "step": 5159 }, { "epoch": 1.2059244544418801, "grad_norm": 3.0945241451263428, "learning_rate": 2.992780046361368e-05, "loss": 1.9433, "step": 5160 }, { "epoch": 1.2061581607315006, "grad_norm": 3.729675531387329, "learning_rate": 2.992774480960137e-05, "loss": 2.0197, "step": 5161 }, { "epoch": 1.2063918670211211, "grad_norm": 2.5299336910247803, "learning_rate": 2.9927689134199046e-05, "loss": 1.8685, "step": 5162 }, { "epoch": 1.2066255733107418, "grad_norm": 2.5496737957000732, "learning_rate": 2.9927633437406796e-05, "loss": 1.9659, "step": 5163 }, { "epoch": 1.2068592796003623, "grad_norm": 3.7229998111724854, "learning_rate": 2.992757771922469e-05, "loss": 1.9052, "step": 5164 }, { "epoch": 1.2070929858899828, "grad_norm": 2.885373592376709, "learning_rate": 2.992752197965281e-05, "loss": 1.8709, "step": 5165 }, { "epoch": 1.2073266921796033, "grad_norm": 2.6525728702545166, "learning_rate": 2.9927466218691237e-05, "loss": 1.7996, "step": 5166 }, { "epoch": 1.2075603984692238, "grad_norm": 2.329023838043213, "learning_rate": 2.9927410436340053e-05, "loss": 1.7828, "step": 5167 }, { "epoch": 1.2077941047588443, "grad_norm": 2.5958495140075684, "learning_rate": 2.9927354632599338e-05, "loss": 1.8491, "step": 5168 }, { "epoch": 1.2080278110484648, "grad_norm": 2.544625997543335, "learning_rate": 2.992729880746917e-05, "loss": 1.7121, "step": 5169 }, { "epoch": 1.2082615173380853, "grad_norm": 3.391549587249756, "learning_rate": 2.9927242960949628e-05, "loss": 1.8207, "step": 5170 }, { "epoch": 1.2084952236277058, "grad_norm": 2.5298826694488525, "learning_rate": 2.9927187093040794e-05, "loss": 1.831, "step": 5171 }, { "epoch": 1.2087289299173265, "grad_norm": 3.8447437286376953, "learning_rate": 2.992713120374275e-05, "loss": 1.8345, "step": 5172 }, { "epoch": 1.208962636206947, "grad_norm": 4.772557735443115, "learning_rate": 2.9927075293055573e-05, "loss": 1.9672, "step": 5173 }, { "epoch": 1.2091963424965675, "grad_norm": 3.630009651184082, "learning_rate": 2.992701936097934e-05, "loss": 1.9987, "step": 5174 }, { "epoch": 1.209430048786188, "grad_norm": 2.509167432785034, "learning_rate": 2.9926963407514133e-05, "loss": 1.8661, "step": 5175 }, { "epoch": 1.2096637550758085, "grad_norm": 7.738816261291504, "learning_rate": 2.992690743266004e-05, "loss": 1.8255, "step": 5176 }, { "epoch": 1.209897461365429, "grad_norm": 6.706874370574951, "learning_rate": 2.9926851436417133e-05, "loss": 1.8011, "step": 5177 }, { "epoch": 1.2101311676550495, "grad_norm": 3.185638189315796, "learning_rate": 2.9926795418785496e-05, "loss": 1.7685, "step": 5178 }, { "epoch": 1.21036487394467, "grad_norm": 3.258502960205078, "learning_rate": 2.992673937976521e-05, "loss": 1.8361, "step": 5179 }, { "epoch": 1.2105985802342905, "grad_norm": 3.707811117172241, "learning_rate": 2.9926683319356347e-05, "loss": 1.7548, "step": 5180 }, { "epoch": 1.2108322865239112, "grad_norm": 4.7889404296875, "learning_rate": 2.9926627237559e-05, "loss": 1.8326, "step": 5181 }, { "epoch": 1.2110659928135317, "grad_norm": 2.86456036567688, "learning_rate": 2.992657113437324e-05, "loss": 1.8144, "step": 5182 }, { "epoch": 1.2112996991031522, "grad_norm": 2.455597162246704, "learning_rate": 2.9926515009799144e-05, "loss": 1.7598, "step": 5183 }, { "epoch": 1.2115334053927727, "grad_norm": 5.1320085525512695, "learning_rate": 2.9926458863836804e-05, "loss": 1.8129, "step": 5184 }, { "epoch": 1.2117671116823931, "grad_norm": 3.4813344478607178, "learning_rate": 2.9926402696486296e-05, "loss": 1.7249, "step": 5185 }, { "epoch": 1.2120008179720136, "grad_norm": 2.3913767337799072, "learning_rate": 2.9926346507747698e-05, "loss": 1.6854, "step": 5186 }, { "epoch": 1.2122345242616341, "grad_norm": 3.698119640350342, "learning_rate": 2.992629029762109e-05, "loss": 1.7952, "step": 5187 }, { "epoch": 1.2124682305512546, "grad_norm": 2.813288450241089, "learning_rate": 2.992623406610656e-05, "loss": 1.7947, "step": 5188 }, { "epoch": 1.2127019368408751, "grad_norm": 3.196633815765381, "learning_rate": 2.992617781320418e-05, "loss": 1.9298, "step": 5189 }, { "epoch": 1.2129356431304958, "grad_norm": 2.989546537399292, "learning_rate": 2.9926121538914035e-05, "loss": 1.7655, "step": 5190 }, { "epoch": 1.2131693494201163, "grad_norm": 2.737414598464966, "learning_rate": 2.9926065243236203e-05, "loss": 1.8694, "step": 5191 }, { "epoch": 1.2134030557097368, "grad_norm": 2.8985650539398193, "learning_rate": 2.9926008926170764e-05, "loss": 1.9936, "step": 5192 }, { "epoch": 1.2136367619993573, "grad_norm": 3.6602396965026855, "learning_rate": 2.99259525877178e-05, "loss": 1.8413, "step": 5193 }, { "epoch": 1.2138704682889778, "grad_norm": 2.56015944480896, "learning_rate": 2.99258962278774e-05, "loss": 1.9087, "step": 5194 }, { "epoch": 1.2141041745785983, "grad_norm": 2.711444616317749, "learning_rate": 2.992583984664963e-05, "loss": 1.7654, "step": 5195 }, { "epoch": 1.2143378808682188, "grad_norm": 2.707228899002075, "learning_rate": 2.9925783444034584e-05, "loss": 1.7535, "step": 5196 }, { "epoch": 1.2145715871578393, "grad_norm": 2.951141119003296, "learning_rate": 2.992572702003233e-05, "loss": 1.8701, "step": 5197 }, { "epoch": 1.2148052934474598, "grad_norm": 2.4505791664123535, "learning_rate": 2.9925670574642962e-05, "loss": 1.7522, "step": 5198 }, { "epoch": 1.2150389997370805, "grad_norm": 2.181155204772949, "learning_rate": 2.9925614107866548e-05, "loss": 1.9557, "step": 5199 }, { "epoch": 1.215272706026701, "grad_norm": 2.8266966342926025, "learning_rate": 2.992555761970318e-05, "loss": 1.7405, "step": 5200 }, { "epoch": 1.2155064123163215, "grad_norm": 2.6279444694519043, "learning_rate": 2.9925501110152934e-05, "loss": 1.8113, "step": 5201 }, { "epoch": 1.215740118605942, "grad_norm": 2.398808479309082, "learning_rate": 2.992544457921589e-05, "loss": 2.0103, "step": 5202 }, { "epoch": 1.2159738248955625, "grad_norm": 2.698316812515259, "learning_rate": 2.9925388026892134e-05, "loss": 1.7719, "step": 5203 }, { "epoch": 1.216207531185183, "grad_norm": 2.318401336669922, "learning_rate": 2.992533145318174e-05, "loss": 1.7841, "step": 5204 }, { "epoch": 1.2164412374748035, "grad_norm": 2.4731063842773438, "learning_rate": 2.9925274858084794e-05, "loss": 1.9386, "step": 5205 }, { "epoch": 1.216674943764424, "grad_norm": 2.560741424560547, "learning_rate": 2.9925218241601375e-05, "loss": 1.8325, "step": 5206 }, { "epoch": 1.2169086500540445, "grad_norm": 2.908576726913452, "learning_rate": 2.9925161603731565e-05, "loss": 1.9535, "step": 5207 }, { "epoch": 1.2171423563436652, "grad_norm": 2.853936195373535, "learning_rate": 2.9925104944475445e-05, "loss": 1.8753, "step": 5208 }, { "epoch": 1.2173760626332857, "grad_norm": 2.2979207038879395, "learning_rate": 2.992504826383309e-05, "loss": 1.8995, "step": 5209 }, { "epoch": 1.2176097689229062, "grad_norm": 2.7582764625549316, "learning_rate": 2.9924991561804598e-05, "loss": 1.9297, "step": 5210 }, { "epoch": 1.2178434752125267, "grad_norm": 2.522467851638794, "learning_rate": 2.9924934838390035e-05, "loss": 1.8197, "step": 5211 }, { "epoch": 1.2180771815021472, "grad_norm": 2.557239055633545, "learning_rate": 2.9924878093589482e-05, "loss": 1.9583, "step": 5212 }, { "epoch": 1.2183108877917677, "grad_norm": 2.8196375370025635, "learning_rate": 2.992482132740303e-05, "loss": 1.8587, "step": 5213 }, { "epoch": 1.2185445940813882, "grad_norm": 2.577538013458252, "learning_rate": 2.9924764539830753e-05, "loss": 1.8399, "step": 5214 }, { "epoch": 1.2187783003710086, "grad_norm": 2.354130268096924, "learning_rate": 2.992470773087274e-05, "loss": 1.8786, "step": 5215 }, { "epoch": 1.2190120066606291, "grad_norm": 2.5944435596466064, "learning_rate": 2.9924650900529066e-05, "loss": 2.0095, "step": 5216 }, { "epoch": 1.2192457129502499, "grad_norm": 3.513112783432007, "learning_rate": 2.9924594048799806e-05, "loss": 1.8084, "step": 5217 }, { "epoch": 1.2194794192398704, "grad_norm": 2.504774808883667, "learning_rate": 2.9924537175685058e-05, "loss": 1.7444, "step": 5218 }, { "epoch": 1.2197131255294908, "grad_norm": 2.8055777549743652, "learning_rate": 2.9924480281184892e-05, "loss": 1.7647, "step": 5219 }, { "epoch": 1.2199468318191113, "grad_norm": 2.729574203491211, "learning_rate": 2.9924423365299388e-05, "loss": 1.8157, "step": 5220 }, { "epoch": 1.2201805381087318, "grad_norm": 2.686854124069214, "learning_rate": 2.9924366428028636e-05, "loss": 1.8044, "step": 5221 }, { "epoch": 1.2204142443983523, "grad_norm": 3.239799737930298, "learning_rate": 2.9924309469372712e-05, "loss": 1.8976, "step": 5222 }, { "epoch": 1.2206479506879728, "grad_norm": 4.216409683227539, "learning_rate": 2.99242524893317e-05, "loss": 1.7699, "step": 5223 }, { "epoch": 1.2208816569775933, "grad_norm": 2.436958074569702, "learning_rate": 2.992419548790568e-05, "loss": 1.8429, "step": 5224 }, { "epoch": 1.2211153632672138, "grad_norm": 3.2657406330108643, "learning_rate": 2.9924138465094733e-05, "loss": 1.7955, "step": 5225 }, { "epoch": 1.2213490695568345, "grad_norm": 4.175999164581299, "learning_rate": 2.9924081420898947e-05, "loss": 1.8337, "step": 5226 }, { "epoch": 1.221582775846455, "grad_norm": 2.4765493869781494, "learning_rate": 2.992402435531839e-05, "loss": 1.7837, "step": 5227 }, { "epoch": 1.2218164821360755, "grad_norm": 3.818920612335205, "learning_rate": 2.9923967268353156e-05, "loss": 1.7107, "step": 5228 }, { "epoch": 1.222050188425696, "grad_norm": 2.6085681915283203, "learning_rate": 2.992391016000333e-05, "loss": 1.8532, "step": 5229 }, { "epoch": 1.2222838947153165, "grad_norm": 2.510899543762207, "learning_rate": 2.9923853030268978e-05, "loss": 1.853, "step": 5230 }, { "epoch": 1.222517601004937, "grad_norm": 2.403486490249634, "learning_rate": 2.9923795879150193e-05, "loss": 1.9355, "step": 5231 }, { "epoch": 1.2227513072945575, "grad_norm": 3.854386329650879, "learning_rate": 2.9923738706647055e-05, "loss": 1.7226, "step": 5232 }, { "epoch": 1.222985013584178, "grad_norm": 3.49428653717041, "learning_rate": 2.992368151275965e-05, "loss": 1.8995, "step": 5233 }, { "epoch": 1.2232187198737985, "grad_norm": 2.964296579360962, "learning_rate": 2.992362429748805e-05, "loss": 1.8729, "step": 5234 }, { "epoch": 1.2234524261634192, "grad_norm": 3.136737823486328, "learning_rate": 2.9923567060832347e-05, "loss": 1.9771, "step": 5235 }, { "epoch": 1.2236861324530397, "grad_norm": 2.3988189697265625, "learning_rate": 2.9923509802792617e-05, "loss": 1.7679, "step": 5236 }, { "epoch": 1.2239198387426602, "grad_norm": 2.5053584575653076, "learning_rate": 2.9923452523368944e-05, "loss": 1.8871, "step": 5237 }, { "epoch": 1.2241535450322807, "grad_norm": 2.7048497200012207, "learning_rate": 2.992339522256141e-05, "loss": 1.8474, "step": 5238 }, { "epoch": 1.2243872513219012, "grad_norm": 3.880764961242676, "learning_rate": 2.9923337900370093e-05, "loss": 1.8154, "step": 5239 }, { "epoch": 1.2246209576115217, "grad_norm": 3.573888063430786, "learning_rate": 2.9923280556795086e-05, "loss": 1.7871, "step": 5240 }, { "epoch": 1.2248546639011422, "grad_norm": 3.114422559738159, "learning_rate": 2.9923223191836454e-05, "loss": 1.9384, "step": 5241 }, { "epoch": 1.2250883701907627, "grad_norm": 3.9370243549346924, "learning_rate": 2.9923165805494302e-05, "loss": 1.8594, "step": 5242 }, { "epoch": 1.2253220764803832, "grad_norm": 2.4819023609161377, "learning_rate": 2.992310839776869e-05, "loss": 1.8446, "step": 5243 }, { "epoch": 1.2255557827700039, "grad_norm": 2.280585289001465, "learning_rate": 2.9923050968659714e-05, "loss": 1.8693, "step": 5244 }, { "epoch": 1.2257894890596244, "grad_norm": 2.4351742267608643, "learning_rate": 2.992299351816745e-05, "loss": 1.8766, "step": 5245 }, { "epoch": 1.2260231953492449, "grad_norm": 2.791652202606201, "learning_rate": 2.9922936046291987e-05, "loss": 1.7689, "step": 5246 }, { "epoch": 1.2262569016388654, "grad_norm": 2.3425543308258057, "learning_rate": 2.99228785530334e-05, "loss": 1.8452, "step": 5247 }, { "epoch": 1.2264906079284859, "grad_norm": 2.609469175338745, "learning_rate": 2.9922821038391777e-05, "loss": 1.8313, "step": 5248 }, { "epoch": 1.2267243142181063, "grad_norm": 2.2943410873413086, "learning_rate": 2.992276350236719e-05, "loss": 1.7707, "step": 5249 }, { "epoch": 1.2269580205077268, "grad_norm": 4.353989601135254, "learning_rate": 2.992270594495974e-05, "loss": 1.8782, "step": 5250 }, { "epoch": 1.2271917267973473, "grad_norm": 2.9891133308410645, "learning_rate": 2.992264836616949e-05, "loss": 1.9075, "step": 5251 }, { "epoch": 1.2274254330869678, "grad_norm": 2.75846791267395, "learning_rate": 2.9922590765996536e-05, "loss": 1.8882, "step": 5252 }, { "epoch": 1.2276591393765885, "grad_norm": 3.6984522342681885, "learning_rate": 2.992253314444095e-05, "loss": 1.8884, "step": 5253 }, { "epoch": 1.227892845666209, "grad_norm": 2.9008467197418213, "learning_rate": 2.9922475501502826e-05, "loss": 1.7892, "step": 5254 }, { "epoch": 1.2281265519558295, "grad_norm": 2.419351100921631, "learning_rate": 2.9922417837182235e-05, "loss": 1.7931, "step": 5255 }, { "epoch": 1.22836025824545, "grad_norm": 2.32649827003479, "learning_rate": 2.9922360151479273e-05, "loss": 1.9963, "step": 5256 }, { "epoch": 1.2285939645350705, "grad_norm": 2.2520763874053955, "learning_rate": 2.992230244439401e-05, "loss": 1.8518, "step": 5257 }, { "epoch": 1.228827670824691, "grad_norm": 2.605008363723755, "learning_rate": 2.9922244715926535e-05, "loss": 1.829, "step": 5258 }, { "epoch": 1.2290613771143115, "grad_norm": 2.678783893585205, "learning_rate": 2.992218696607693e-05, "loss": 1.8752, "step": 5259 }, { "epoch": 1.2292950834039322, "grad_norm": 2.544220209121704, "learning_rate": 2.9922129194845278e-05, "loss": 1.9359, "step": 5260 }, { "epoch": 1.2295287896935525, "grad_norm": 2.2041733264923096, "learning_rate": 2.9922071402231657e-05, "loss": 1.8517, "step": 5261 }, { "epoch": 1.2297624959831732, "grad_norm": 2.8031091690063477, "learning_rate": 2.992201358823616e-05, "loss": 1.859, "step": 5262 }, { "epoch": 1.2299962022727937, "grad_norm": 3.6947181224823, "learning_rate": 2.9921955752858858e-05, "loss": 1.7443, "step": 5263 }, { "epoch": 1.2302299085624142, "grad_norm": 2.990607261657715, "learning_rate": 2.9921897896099842e-05, "loss": 1.8997, "step": 5264 }, { "epoch": 1.2304636148520347, "grad_norm": 3.7056570053100586, "learning_rate": 2.992184001795919e-05, "loss": 1.7365, "step": 5265 }, { "epoch": 1.2306973211416552, "grad_norm": 2.435682535171509, "learning_rate": 2.992178211843699e-05, "loss": 1.9441, "step": 5266 }, { "epoch": 1.2309310274312757, "grad_norm": 2.7780473232269287, "learning_rate": 2.9921724197533326e-05, "loss": 1.7822, "step": 5267 }, { "epoch": 1.2311647337208962, "grad_norm": 2.3155415058135986, "learning_rate": 2.9921666255248268e-05, "loss": 1.7927, "step": 5268 }, { "epoch": 1.231398440010517, "grad_norm": 2.2029483318328857, "learning_rate": 2.9921608291581914e-05, "loss": 1.9198, "step": 5269 }, { "epoch": 1.2316321463001374, "grad_norm": 2.489924669265747, "learning_rate": 2.9921550306534344e-05, "loss": 2.0051, "step": 5270 }, { "epoch": 1.231865852589758, "grad_norm": 2.650909900665283, "learning_rate": 2.9921492300105633e-05, "loss": 1.7724, "step": 5271 }, { "epoch": 1.2320995588793784, "grad_norm": 2.465228319168091, "learning_rate": 2.9921434272295872e-05, "loss": 1.8907, "step": 5272 }, { "epoch": 1.2323332651689989, "grad_norm": 4.6985344886779785, "learning_rate": 2.9921376223105143e-05, "loss": 1.6837, "step": 5273 }, { "epoch": 1.2325669714586194, "grad_norm": 2.7131404876708984, "learning_rate": 2.9921318152533527e-05, "loss": 2.035, "step": 5274 }, { "epoch": 1.2328006777482399, "grad_norm": 2.485325813293457, "learning_rate": 2.9921260060581106e-05, "loss": 1.7835, "step": 5275 }, { "epoch": 1.2330343840378604, "grad_norm": 2.533496379852295, "learning_rate": 2.9921201947247967e-05, "loss": 2.0288, "step": 5276 }, { "epoch": 1.2332680903274809, "grad_norm": 2.4344615936279297, "learning_rate": 2.9921143812534195e-05, "loss": 1.8768, "step": 5277 }, { "epoch": 1.2335017966171016, "grad_norm": 3.254096508026123, "learning_rate": 2.9921085656439864e-05, "loss": 1.9806, "step": 5278 }, { "epoch": 1.233735502906722, "grad_norm": 2.3625056743621826, "learning_rate": 2.9921027478965067e-05, "loss": 1.8669, "step": 5279 }, { "epoch": 1.2339692091963426, "grad_norm": 2.295971155166626, "learning_rate": 2.9920969280109885e-05, "loss": 1.8641, "step": 5280 }, { "epoch": 1.234202915485963, "grad_norm": 2.4850733280181885, "learning_rate": 2.9920911059874398e-05, "loss": 1.8746, "step": 5281 }, { "epoch": 1.2344366217755836, "grad_norm": 2.4018962383270264, "learning_rate": 2.9920852818258688e-05, "loss": 1.8908, "step": 5282 }, { "epoch": 1.234670328065204, "grad_norm": 2.809826612472534, "learning_rate": 2.9920794555262847e-05, "loss": 1.9453, "step": 5283 }, { "epoch": 1.2349040343548245, "grad_norm": 2.440370559692383, "learning_rate": 2.9920736270886955e-05, "loss": 1.7229, "step": 5284 }, { "epoch": 1.235137740644445, "grad_norm": 2.3229405879974365, "learning_rate": 2.9920677965131092e-05, "loss": 2.0347, "step": 5285 }, { "epoch": 1.2353714469340655, "grad_norm": 2.502596616744995, "learning_rate": 2.992061963799534e-05, "loss": 1.913, "step": 5286 }, { "epoch": 1.2356051532236862, "grad_norm": 2.5280399322509766, "learning_rate": 2.9920561289479787e-05, "loss": 1.7604, "step": 5287 }, { "epoch": 1.2358388595133067, "grad_norm": 2.574737548828125, "learning_rate": 2.992050291958452e-05, "loss": 2.0648, "step": 5288 }, { "epoch": 1.2360725658029272, "grad_norm": 2.571057081222534, "learning_rate": 2.992044452830961e-05, "loss": 1.9427, "step": 5289 }, { "epoch": 1.2363062720925477, "grad_norm": 3.050816774368286, "learning_rate": 2.9920386115655157e-05, "loss": 1.9058, "step": 5290 }, { "epoch": 1.2365399783821682, "grad_norm": 2.622025966644287, "learning_rate": 2.9920327681621232e-05, "loss": 1.8113, "step": 5291 }, { "epoch": 1.2367736846717887, "grad_norm": 3.0775015354156494, "learning_rate": 2.9920269226207932e-05, "loss": 1.8527, "step": 5292 }, { "epoch": 1.2370073909614092, "grad_norm": 3.086092472076416, "learning_rate": 2.9920210749415323e-05, "loss": 1.9146, "step": 5293 }, { "epoch": 1.2372410972510297, "grad_norm": 4.0921783447265625, "learning_rate": 2.99201522512435e-05, "loss": 1.7826, "step": 5294 }, { "epoch": 1.2374748035406502, "grad_norm": 3.3955976963043213, "learning_rate": 2.9920093731692545e-05, "loss": 1.7786, "step": 5295 }, { "epoch": 1.237708509830271, "grad_norm": 2.7933688163757324, "learning_rate": 2.992003519076254e-05, "loss": 2.0763, "step": 5296 }, { "epoch": 1.2379422161198914, "grad_norm": 5.355695724487305, "learning_rate": 2.9919976628453574e-05, "loss": 1.7743, "step": 5297 }, { "epoch": 1.238175922409512, "grad_norm": 4.045856475830078, "learning_rate": 2.9919918044765726e-05, "loss": 1.7936, "step": 5298 }, { "epoch": 1.2384096286991324, "grad_norm": 3.1107091903686523, "learning_rate": 2.9919859439699085e-05, "loss": 1.9009, "step": 5299 }, { "epoch": 1.238643334988753, "grad_norm": 2.72139573097229, "learning_rate": 2.9919800813253725e-05, "loss": 1.8565, "step": 5300 }, { "epoch": 1.2388770412783734, "grad_norm": 4.397364616394043, "learning_rate": 2.9919742165429742e-05, "loss": 1.8664, "step": 5301 }, { "epoch": 1.2391107475679939, "grad_norm": 3.427030563354492, "learning_rate": 2.991968349622721e-05, "loss": 1.8064, "step": 5302 }, { "epoch": 1.2393444538576144, "grad_norm": 2.497904062271118, "learning_rate": 2.9919624805646216e-05, "loss": 1.783, "step": 5303 }, { "epoch": 1.2395781601472349, "grad_norm": 2.7125470638275146, "learning_rate": 2.991956609368685e-05, "loss": 1.7683, "step": 5304 }, { "epoch": 1.2398118664368556, "grad_norm": 2.7908966541290283, "learning_rate": 2.991950736034919e-05, "loss": 1.8652, "step": 5305 }, { "epoch": 1.240045572726476, "grad_norm": 3.1803834438323975, "learning_rate": 2.9919448605633323e-05, "loss": 1.9272, "step": 5306 }, { "epoch": 1.2402792790160966, "grad_norm": 3.0511200428009033, "learning_rate": 2.9919389829539332e-05, "loss": 1.9423, "step": 5307 }, { "epoch": 1.240512985305717, "grad_norm": 2.7044150829315186, "learning_rate": 2.99193310320673e-05, "loss": 1.8654, "step": 5308 }, { "epoch": 1.2407466915953376, "grad_norm": 3.5812528133392334, "learning_rate": 2.9919272213217312e-05, "loss": 1.8296, "step": 5309 }, { "epoch": 1.240980397884958, "grad_norm": 2.3918681144714355, "learning_rate": 2.9919213372989453e-05, "loss": 1.8505, "step": 5310 }, { "epoch": 1.2412141041745786, "grad_norm": 2.356167793273926, "learning_rate": 2.991915451138381e-05, "loss": 1.9111, "step": 5311 }, { "epoch": 1.241447810464199, "grad_norm": 3.4222052097320557, "learning_rate": 2.991909562840046e-05, "loss": 1.8016, "step": 5312 }, { "epoch": 1.2416815167538195, "grad_norm": 3.938783645629883, "learning_rate": 2.9919036724039495e-05, "loss": 1.8443, "step": 5313 }, { "epoch": 1.2419152230434403, "grad_norm": 2.445117473602295, "learning_rate": 2.9918977798300997e-05, "loss": 1.7262, "step": 5314 }, { "epoch": 1.2421489293330608, "grad_norm": 3.032679319381714, "learning_rate": 2.991891885118505e-05, "loss": 2.0211, "step": 5315 }, { "epoch": 1.2423826356226813, "grad_norm": 4.482065677642822, "learning_rate": 2.9918859882691734e-05, "loss": 1.9049, "step": 5316 }, { "epoch": 1.2426163419123017, "grad_norm": 2.471666097640991, "learning_rate": 2.991880089282114e-05, "loss": 1.8812, "step": 5317 }, { "epoch": 1.2428500482019222, "grad_norm": 3.180670738220215, "learning_rate": 2.991874188157335e-05, "loss": 2.0576, "step": 5318 }, { "epoch": 1.2430837544915427, "grad_norm": 3.8092269897460938, "learning_rate": 2.9918682848948447e-05, "loss": 1.9539, "step": 5319 }, { "epoch": 1.2433174607811632, "grad_norm": 3.246593713760376, "learning_rate": 2.991862379494652e-05, "loss": 1.8508, "step": 5320 }, { "epoch": 1.2435511670707837, "grad_norm": 3.6141340732574463, "learning_rate": 2.991856471956765e-05, "loss": 1.9491, "step": 5321 }, { "epoch": 1.2437848733604042, "grad_norm": 3.868481397628784, "learning_rate": 2.9918505622811924e-05, "loss": 1.8698, "step": 5322 }, { "epoch": 1.244018579650025, "grad_norm": 3.7932024002075195, "learning_rate": 2.991844650467942e-05, "loss": 1.8288, "step": 5323 }, { "epoch": 1.2442522859396454, "grad_norm": 2.3799102306365967, "learning_rate": 2.9918387365170234e-05, "loss": 1.9408, "step": 5324 }, { "epoch": 1.244485992229266, "grad_norm": 5.963438987731934, "learning_rate": 2.991832820428444e-05, "loss": 1.9406, "step": 5325 }, { "epoch": 1.2447196985188864, "grad_norm": 5.391637802124023, "learning_rate": 2.991826902202213e-05, "loss": 1.8708, "step": 5326 }, { "epoch": 1.244953404808507, "grad_norm": 3.8306448459625244, "learning_rate": 2.9918209818383388e-05, "loss": 1.7807, "step": 5327 }, { "epoch": 1.2451871110981274, "grad_norm": 3.124098539352417, "learning_rate": 2.9918150593368295e-05, "loss": 1.9266, "step": 5328 }, { "epoch": 1.245420817387748, "grad_norm": 3.338428258895874, "learning_rate": 2.991809134697694e-05, "loss": 1.9245, "step": 5329 }, { "epoch": 1.2456545236773684, "grad_norm": 3.685152292251587, "learning_rate": 2.9918032079209403e-05, "loss": 1.8537, "step": 5330 }, { "epoch": 1.245888229966989, "grad_norm": 2.3703317642211914, "learning_rate": 2.991797279006577e-05, "loss": 1.8321, "step": 5331 }, { "epoch": 1.2461219362566096, "grad_norm": 3.4644789695739746, "learning_rate": 2.9917913479546134e-05, "loss": 1.7923, "step": 5332 }, { "epoch": 1.24635564254623, "grad_norm": 4.153054237365723, "learning_rate": 2.991785414765057e-05, "loss": 1.882, "step": 5333 }, { "epoch": 1.2465893488358506, "grad_norm": 3.2747271060943604, "learning_rate": 2.9917794794379167e-05, "loss": 1.8137, "step": 5334 }, { "epoch": 1.246823055125471, "grad_norm": 2.6982548236846924, "learning_rate": 2.9917735419732007e-05, "loss": 1.9003, "step": 5335 }, { "epoch": 1.2470567614150916, "grad_norm": 3.0020639896392822, "learning_rate": 2.991767602370918e-05, "loss": 1.8057, "step": 5336 }, { "epoch": 1.247290467704712, "grad_norm": 2.3880412578582764, "learning_rate": 2.9917616606310775e-05, "loss": 1.8657, "step": 5337 }, { "epoch": 1.2475241739943326, "grad_norm": 3.1763787269592285, "learning_rate": 2.9917557167536865e-05, "loss": 1.9978, "step": 5338 }, { "epoch": 1.247757880283953, "grad_norm": 3.226304531097412, "learning_rate": 2.991749770738754e-05, "loss": 1.841, "step": 5339 }, { "epoch": 1.2479915865735736, "grad_norm": 2.3483734130859375, "learning_rate": 2.9917438225862887e-05, "loss": 1.8553, "step": 5340 }, { "epoch": 1.2482252928631943, "grad_norm": 2.2725706100463867, "learning_rate": 2.9917378722962992e-05, "loss": 1.8956, "step": 5341 }, { "epoch": 1.2484589991528148, "grad_norm": 2.201361656188965, "learning_rate": 2.991731919868794e-05, "loss": 1.8234, "step": 5342 }, { "epoch": 1.2486927054424353, "grad_norm": 2.4822866916656494, "learning_rate": 2.9917259653037815e-05, "loss": 1.8896, "step": 5343 }, { "epoch": 1.2489264117320558, "grad_norm": 2.2560997009277344, "learning_rate": 2.99172000860127e-05, "loss": 1.9485, "step": 5344 }, { "epoch": 1.2491601180216763, "grad_norm": 2.643930435180664, "learning_rate": 2.9917140497612688e-05, "loss": 2.0239, "step": 5345 }, { "epoch": 1.2493938243112968, "grad_norm": 2.410156011581421, "learning_rate": 2.991708088783786e-05, "loss": 1.7723, "step": 5346 }, { "epoch": 1.2496275306009172, "grad_norm": 2.6129417419433594, "learning_rate": 2.9917021256688292e-05, "loss": 1.9296, "step": 5347 }, { "epoch": 1.2498612368905377, "grad_norm": 3.0727009773254395, "learning_rate": 2.9916961604164086e-05, "loss": 1.7239, "step": 5348 }, { "epoch": 1.2500949431801582, "grad_norm": 2.7676570415496826, "learning_rate": 2.9916901930265316e-05, "loss": 1.8537, "step": 5349 }, { "epoch": 1.250328649469779, "grad_norm": 3.6114723682403564, "learning_rate": 2.9916842234992072e-05, "loss": 1.7439, "step": 5350 }, { "epoch": 1.2505623557593994, "grad_norm": 2.6283953189849854, "learning_rate": 2.991678251834444e-05, "loss": 1.9222, "step": 5351 }, { "epoch": 1.25079606204902, "grad_norm": 2.725404977798462, "learning_rate": 2.9916722780322504e-05, "loss": 1.9698, "step": 5352 }, { "epoch": 1.2510297683386404, "grad_norm": 2.3647360801696777, "learning_rate": 2.991666302092635e-05, "loss": 1.911, "step": 5353 }, { "epoch": 1.251263474628261, "grad_norm": 2.462169647216797, "learning_rate": 2.9916603240156062e-05, "loss": 1.8244, "step": 5354 }, { "epoch": 1.2514971809178814, "grad_norm": 3.535574436187744, "learning_rate": 2.9916543438011726e-05, "loss": 1.8662, "step": 5355 }, { "epoch": 1.251730887207502, "grad_norm": 2.4723246097564697, "learning_rate": 2.9916483614493432e-05, "loss": 1.9354, "step": 5356 }, { "epoch": 1.2519645934971224, "grad_norm": 3.03352952003479, "learning_rate": 2.991642376960126e-05, "loss": 1.8845, "step": 5357 }, { "epoch": 1.252198299786743, "grad_norm": 3.328505754470825, "learning_rate": 2.99163639033353e-05, "loss": 1.9024, "step": 5358 }, { "epoch": 1.2524320060763636, "grad_norm": 2.9011940956115723, "learning_rate": 2.9916304015695633e-05, "loss": 1.7837, "step": 5359 }, { "epoch": 1.2526657123659841, "grad_norm": 3.618051052093506, "learning_rate": 2.9916244106682353e-05, "loss": 1.865, "step": 5360 }, { "epoch": 1.2528994186556046, "grad_norm": 3.7109363079071045, "learning_rate": 2.991618417629554e-05, "loss": 1.8593, "step": 5361 }, { "epoch": 1.253133124945225, "grad_norm": 2.9940104484558105, "learning_rate": 2.9916124224535277e-05, "loss": 2.0169, "step": 5362 }, { "epoch": 1.2533668312348456, "grad_norm": 2.442319393157959, "learning_rate": 2.9916064251401657e-05, "loss": 2.1622, "step": 5363 }, { "epoch": 1.253600537524466, "grad_norm": 3.1016926765441895, "learning_rate": 2.991600425689476e-05, "loss": 1.8621, "step": 5364 }, { "epoch": 1.2538342438140866, "grad_norm": 3.3214354515075684, "learning_rate": 2.9915944241014677e-05, "loss": 1.827, "step": 5365 }, { "epoch": 1.254067950103707, "grad_norm": 2.2854793071746826, "learning_rate": 2.9915884203761492e-05, "loss": 1.9179, "step": 5366 }, { "epoch": 1.2543016563933276, "grad_norm": 3.8238654136657715, "learning_rate": 2.9915824145135285e-05, "loss": 1.9832, "step": 5367 }, { "epoch": 1.2545353626829483, "grad_norm": 5.884975910186768, "learning_rate": 2.9915764065136153e-05, "loss": 1.899, "step": 5368 }, { "epoch": 1.2547690689725688, "grad_norm": 2.943368911743164, "learning_rate": 2.9915703963764174e-05, "loss": 1.8704, "step": 5369 }, { "epoch": 1.2550027752621893, "grad_norm": 3.953937530517578, "learning_rate": 2.991564384101944e-05, "loss": 1.9231, "step": 5370 }, { "epoch": 1.2552364815518098, "grad_norm": 7.087413787841797, "learning_rate": 2.9915583696902028e-05, "loss": 1.8977, "step": 5371 }, { "epoch": 1.2554701878414303, "grad_norm": 4.521431922912598, "learning_rate": 2.9915523531412037e-05, "loss": 1.8035, "step": 5372 }, { "epoch": 1.2557038941310508, "grad_norm": 2.5243327617645264, "learning_rate": 2.991546334454954e-05, "loss": 1.8652, "step": 5373 }, { "epoch": 1.2559376004206713, "grad_norm": 3.401113748550415, "learning_rate": 2.9915403136314635e-05, "loss": 1.9039, "step": 5374 }, { "epoch": 1.256171306710292, "grad_norm": 5.214142799377441, "learning_rate": 2.99153429067074e-05, "loss": 1.9361, "step": 5375 }, { "epoch": 1.2564050129999123, "grad_norm": 4.005946159362793, "learning_rate": 2.9915282655727922e-05, "loss": 1.8105, "step": 5376 }, { "epoch": 1.256638719289533, "grad_norm": 2.3630003929138184, "learning_rate": 2.9915222383376295e-05, "loss": 1.791, "step": 5377 }, { "epoch": 1.2568724255791535, "grad_norm": 3.845897912979126, "learning_rate": 2.9915162089652597e-05, "loss": 1.9888, "step": 5378 }, { "epoch": 1.257106131868774, "grad_norm": 5.671634674072266, "learning_rate": 2.9915101774556916e-05, "loss": 1.9733, "step": 5379 }, { "epoch": 1.2573398381583945, "grad_norm": 4.589301586151123, "learning_rate": 2.991504143808934e-05, "loss": 1.8986, "step": 5380 }, { "epoch": 1.257573544448015, "grad_norm": 2.3469574451446533, "learning_rate": 2.9914981080249957e-05, "loss": 1.8597, "step": 5381 }, { "epoch": 1.2578072507376354, "grad_norm": 2.7045629024505615, "learning_rate": 2.9914920701038845e-05, "loss": 1.7593, "step": 5382 }, { "epoch": 1.258040957027256, "grad_norm": 5.034801959991455, "learning_rate": 2.9914860300456103e-05, "loss": 1.8978, "step": 5383 }, { "epoch": 1.2582746633168767, "grad_norm": 4.7710394859313965, "learning_rate": 2.991479987850181e-05, "loss": 1.8331, "step": 5384 }, { "epoch": 1.258508369606497, "grad_norm": 2.703012466430664, "learning_rate": 2.991473943517606e-05, "loss": 1.9264, "step": 5385 }, { "epoch": 1.2587420758961176, "grad_norm": 4.439531326293945, "learning_rate": 2.9914678970478926e-05, "loss": 1.9028, "step": 5386 }, { "epoch": 1.2589757821857381, "grad_norm": 7.076721668243408, "learning_rate": 2.9914618484410505e-05, "loss": 1.8002, "step": 5387 }, { "epoch": 1.2592094884753586, "grad_norm": 6.079005241394043, "learning_rate": 2.9914557976970883e-05, "loss": 1.8251, "step": 5388 }, { "epoch": 1.2594431947649791, "grad_norm": 2.6909615993499756, "learning_rate": 2.991449744816014e-05, "loss": 1.7533, "step": 5389 }, { "epoch": 1.2596769010545996, "grad_norm": 5.306320667266846, "learning_rate": 2.9914436897978372e-05, "loss": 1.7447, "step": 5390 }, { "epoch": 1.2599106073442201, "grad_norm": 6.10982084274292, "learning_rate": 2.991437632642566e-05, "loss": 1.9454, "step": 5391 }, { "epoch": 1.2601443136338406, "grad_norm": 4.460124969482422, "learning_rate": 2.9914315733502093e-05, "loss": 1.951, "step": 5392 }, { "epoch": 1.2603780199234613, "grad_norm": 2.643627166748047, "learning_rate": 2.9914255119207757e-05, "loss": 1.8001, "step": 5393 }, { "epoch": 1.2606117262130816, "grad_norm": 2.8697633743286133, "learning_rate": 2.991419448354274e-05, "loss": 1.9543, "step": 5394 }, { "epoch": 1.2608454325027023, "grad_norm": 2.808346748352051, "learning_rate": 2.9914133826507126e-05, "loss": 1.8498, "step": 5395 }, { "epoch": 1.2610791387923228, "grad_norm": 2.356900930404663, "learning_rate": 2.9914073148101e-05, "loss": 1.7454, "step": 5396 }, { "epoch": 1.2613128450819433, "grad_norm": 3.365917444229126, "learning_rate": 2.9914012448324457e-05, "loss": 1.8146, "step": 5397 }, { "epoch": 1.2615465513715638, "grad_norm": 3.794388771057129, "learning_rate": 2.9913951727177576e-05, "loss": 1.8483, "step": 5398 }, { "epoch": 1.2617802576611843, "grad_norm": 2.531236171722412, "learning_rate": 2.9913890984660454e-05, "loss": 1.7433, "step": 5399 }, { "epoch": 1.2620139639508048, "grad_norm": 3.158489227294922, "learning_rate": 2.9913830220773165e-05, "loss": 1.8328, "step": 5400 }, { "epoch": 1.2622476702404253, "grad_norm": 2.3754096031188965, "learning_rate": 2.9913769435515807e-05, "loss": 1.7876, "step": 5401 }, { "epoch": 1.262481376530046, "grad_norm": 2.6399784088134766, "learning_rate": 2.9913708628888458e-05, "loss": 1.8631, "step": 5402 }, { "epoch": 1.2627150828196663, "grad_norm": 2.4771647453308105, "learning_rate": 2.9913647800891217e-05, "loss": 1.7122, "step": 5403 }, { "epoch": 1.262948789109287, "grad_norm": 5.343264579772949, "learning_rate": 2.9913586951524158e-05, "loss": 1.8654, "step": 5404 }, { "epoch": 1.2631824953989075, "grad_norm": 3.7674853801727295, "learning_rate": 2.9913526080787375e-05, "loss": 1.9579, "step": 5405 }, { "epoch": 1.263416201688528, "grad_norm": 2.5356321334838867, "learning_rate": 2.9913465188680958e-05, "loss": 1.8174, "step": 5406 }, { "epoch": 1.2636499079781485, "grad_norm": 3.545452117919922, "learning_rate": 2.991340427520499e-05, "loss": 1.9094, "step": 5407 }, { "epoch": 1.263883614267769, "grad_norm": 3.6794469356536865, "learning_rate": 2.9913343340359554e-05, "loss": 1.8126, "step": 5408 }, { "epoch": 1.2641173205573895, "grad_norm": 2.6962931156158447, "learning_rate": 2.9913282384144744e-05, "loss": 1.9574, "step": 5409 }, { "epoch": 1.26435102684701, "grad_norm": 3.1809074878692627, "learning_rate": 2.991322140656065e-05, "loss": 1.7661, "step": 5410 }, { "epoch": 1.2645847331366307, "grad_norm": 2.635692596435547, "learning_rate": 2.9913160407607352e-05, "loss": 1.8617, "step": 5411 }, { "epoch": 1.264818439426251, "grad_norm": 2.3733224868774414, "learning_rate": 2.9913099387284938e-05, "loss": 1.7659, "step": 5412 }, { "epoch": 1.2650521457158717, "grad_norm": 3.9666240215301514, "learning_rate": 2.9913038345593505e-05, "loss": 1.9467, "step": 5413 }, { "epoch": 1.2652858520054922, "grad_norm": 4.492288112640381, "learning_rate": 2.9912977282533125e-05, "loss": 1.8677, "step": 5414 }, { "epoch": 1.2655195582951126, "grad_norm": 3.374823570251465, "learning_rate": 2.99129161981039e-05, "loss": 1.8199, "step": 5415 }, { "epoch": 1.2657532645847331, "grad_norm": 3.578963041305542, "learning_rate": 2.9912855092305906e-05, "loss": 1.9211, "step": 5416 }, { "epoch": 1.2659869708743536, "grad_norm": 4.532707214355469, "learning_rate": 2.991279396513924e-05, "loss": 2.0483, "step": 5417 }, { "epoch": 1.2662206771639741, "grad_norm": 2.7283480167388916, "learning_rate": 2.9912732816603987e-05, "loss": 1.8997, "step": 5418 }, { "epoch": 1.2664543834535946, "grad_norm": 2.935046672821045, "learning_rate": 2.9912671646700226e-05, "loss": 1.9206, "step": 5419 }, { "epoch": 1.2666880897432153, "grad_norm": 2.558006525039673, "learning_rate": 2.9912610455428057e-05, "loss": 1.7741, "step": 5420 }, { "epoch": 1.2669217960328356, "grad_norm": 3.807387351989746, "learning_rate": 2.991254924278756e-05, "loss": 1.8263, "step": 5421 }, { "epoch": 1.2671555023224563, "grad_norm": 2.8151111602783203, "learning_rate": 2.991248800877883e-05, "loss": 1.9515, "step": 5422 }, { "epoch": 1.2673892086120768, "grad_norm": 2.9159786701202393, "learning_rate": 2.991242675340195e-05, "loss": 1.9559, "step": 5423 }, { "epoch": 1.2676229149016973, "grad_norm": 3.703958511352539, "learning_rate": 2.9912365476657003e-05, "loss": 1.8102, "step": 5424 }, { "epoch": 1.2678566211913178, "grad_norm": 2.8941335678100586, "learning_rate": 2.9912304178544082e-05, "loss": 1.9704, "step": 5425 }, { "epoch": 1.2680903274809383, "grad_norm": 2.611711263656616, "learning_rate": 2.9912242859063278e-05, "loss": 1.8922, "step": 5426 }, { "epoch": 1.2683240337705588, "grad_norm": 2.3323469161987305, "learning_rate": 2.991218151821467e-05, "loss": 1.959, "step": 5427 }, { "epoch": 1.2685577400601793, "grad_norm": 3.630850315093994, "learning_rate": 2.9912120155998355e-05, "loss": 1.8022, "step": 5428 }, { "epoch": 1.2687914463498, "grad_norm": 2.7204208374023438, "learning_rate": 2.9912058772414413e-05, "loss": 1.8625, "step": 5429 }, { "epoch": 1.2690251526394203, "grad_norm": 2.2438488006591797, "learning_rate": 2.991199736746294e-05, "loss": 1.8416, "step": 5430 }, { "epoch": 1.269258858929041, "grad_norm": 2.23524808883667, "learning_rate": 2.9911935941144016e-05, "loss": 1.7025, "step": 5431 }, { "epoch": 1.2694925652186615, "grad_norm": 2.2937734127044678, "learning_rate": 2.991187449345774e-05, "loss": 1.9336, "step": 5432 }, { "epoch": 1.269726271508282, "grad_norm": 2.4135758876800537, "learning_rate": 2.9911813024404184e-05, "loss": 1.9443, "step": 5433 }, { "epoch": 1.2699599777979025, "grad_norm": 2.2429468631744385, "learning_rate": 2.9911751533983448e-05, "loss": 1.862, "step": 5434 }, { "epoch": 1.270193684087523, "grad_norm": 2.305826425552368, "learning_rate": 2.991169002219562e-05, "loss": 1.7953, "step": 5435 }, { "epoch": 1.2704273903771435, "grad_norm": 2.367145299911499, "learning_rate": 2.9911628489040788e-05, "loss": 1.8462, "step": 5436 }, { "epoch": 1.270661096666764, "grad_norm": 2.504910469055176, "learning_rate": 2.991156693451903e-05, "loss": 1.6887, "step": 5437 }, { "epoch": 1.2708948029563847, "grad_norm": 2.3415465354919434, "learning_rate": 2.9911505358630447e-05, "loss": 1.7876, "step": 5438 }, { "epoch": 1.271128509246005, "grad_norm": 2.2876899242401123, "learning_rate": 2.9911443761375118e-05, "loss": 1.8732, "step": 5439 }, { "epoch": 1.2713622155356257, "grad_norm": 2.4596407413482666, "learning_rate": 2.991138214275314e-05, "loss": 1.978, "step": 5440 }, { "epoch": 1.2715959218252462, "grad_norm": 3.688632011413574, "learning_rate": 2.9911320502764592e-05, "loss": 1.8382, "step": 5441 }, { "epoch": 1.2718296281148667, "grad_norm": 2.6388444900512695, "learning_rate": 2.991125884140957e-05, "loss": 1.8295, "step": 5442 }, { "epoch": 1.2720633344044872, "grad_norm": 3.428713798522949, "learning_rate": 2.991119715868816e-05, "loss": 1.935, "step": 5443 }, { "epoch": 1.2722970406941077, "grad_norm": 2.6972458362579346, "learning_rate": 2.9911135454600446e-05, "loss": 1.8054, "step": 5444 }, { "epoch": 1.2725307469837281, "grad_norm": 2.8951025009155273, "learning_rate": 2.991107372914652e-05, "loss": 1.7527, "step": 5445 }, { "epoch": 1.2727644532733486, "grad_norm": 2.505312919616699, "learning_rate": 2.9911011982326473e-05, "loss": 1.7353, "step": 5446 }, { "epoch": 1.2729981595629694, "grad_norm": 2.429412364959717, "learning_rate": 2.991095021414039e-05, "loss": 1.8538, "step": 5447 }, { "epoch": 1.2732318658525896, "grad_norm": 2.339062213897705, "learning_rate": 2.9910888424588362e-05, "loss": 1.8899, "step": 5448 }, { "epoch": 1.2734655721422103, "grad_norm": 3.0291614532470703, "learning_rate": 2.991082661367047e-05, "loss": 1.8708, "step": 5449 }, { "epoch": 1.2736992784318308, "grad_norm": 2.7568554878234863, "learning_rate": 2.9910764781386813e-05, "loss": 1.9029, "step": 5450 }, { "epoch": 1.2739329847214513, "grad_norm": 2.4763357639312744, "learning_rate": 2.9910702927737474e-05, "loss": 1.8714, "step": 5451 }, { "epoch": 1.2741666910110718, "grad_norm": 2.456021308898926, "learning_rate": 2.9910641052722546e-05, "loss": 2.0418, "step": 5452 }, { "epoch": 1.2744003973006923, "grad_norm": 2.456859827041626, "learning_rate": 2.9910579156342113e-05, "loss": 1.7806, "step": 5453 }, { "epoch": 1.2746341035903128, "grad_norm": 2.5200202465057373, "learning_rate": 2.9910517238596263e-05, "loss": 1.962, "step": 5454 }, { "epoch": 1.2748678098799333, "grad_norm": 2.787384033203125, "learning_rate": 2.9910455299485088e-05, "loss": 1.8854, "step": 5455 }, { "epoch": 1.275101516169554, "grad_norm": 2.577845573425293, "learning_rate": 2.9910393339008676e-05, "loss": 1.9192, "step": 5456 }, { "epoch": 1.2753352224591743, "grad_norm": 2.6101362705230713, "learning_rate": 2.9910331357167114e-05, "loss": 1.879, "step": 5457 }, { "epoch": 1.275568928748795, "grad_norm": 3.3608081340789795, "learning_rate": 2.9910269353960492e-05, "loss": 1.9231, "step": 5458 }, { "epoch": 1.2758026350384155, "grad_norm": 4.082651615142822, "learning_rate": 2.99102073293889e-05, "loss": 1.8154, "step": 5459 }, { "epoch": 1.276036341328036, "grad_norm": 2.5483365058898926, "learning_rate": 2.9910145283452427e-05, "loss": 1.9969, "step": 5460 }, { "epoch": 1.2762700476176565, "grad_norm": 4.208503723144531, "learning_rate": 2.991008321615116e-05, "loss": 1.7836, "step": 5461 }, { "epoch": 1.276503753907277, "grad_norm": 2.8424222469329834, "learning_rate": 2.9910021127485188e-05, "loss": 1.8443, "step": 5462 }, { "epoch": 1.2767374601968975, "grad_norm": 2.531346559524536, "learning_rate": 2.99099590174546e-05, "loss": 1.8069, "step": 5463 }, { "epoch": 1.276971166486518, "grad_norm": 2.646132707595825, "learning_rate": 2.9909896886059488e-05, "loss": 1.6915, "step": 5464 }, { "epoch": 1.2772048727761387, "grad_norm": 2.875366449356079, "learning_rate": 2.990983473329994e-05, "loss": 1.8508, "step": 5465 }, { "epoch": 1.277438579065759, "grad_norm": 2.3381872177124023, "learning_rate": 2.990977255917604e-05, "loss": 1.8984, "step": 5466 }, { "epoch": 1.2776722853553797, "grad_norm": 2.4475231170654297, "learning_rate": 2.9909710363687884e-05, "loss": 1.8128, "step": 5467 }, { "epoch": 1.2779059916450002, "grad_norm": 2.7009055614471436, "learning_rate": 2.9909648146835554e-05, "loss": 1.8243, "step": 5468 }, { "epoch": 1.2781396979346207, "grad_norm": 3.03523588180542, "learning_rate": 2.9909585908619146e-05, "loss": 1.9103, "step": 5469 }, { "epoch": 1.2783734042242412, "grad_norm": 2.291141986846924, "learning_rate": 2.9909523649038744e-05, "loss": 1.8565, "step": 5470 }, { "epoch": 1.2786071105138617, "grad_norm": 2.4249725341796875, "learning_rate": 2.9909461368094443e-05, "loss": 1.8047, "step": 5471 }, { "epoch": 1.2788408168034822, "grad_norm": 2.2737369537353516, "learning_rate": 2.9909399065786328e-05, "loss": 1.9375, "step": 5472 }, { "epoch": 1.2790745230931027, "grad_norm": 3.0009162425994873, "learning_rate": 2.9909336742114486e-05, "loss": 1.7722, "step": 5473 }, { "epoch": 1.2793082293827234, "grad_norm": 2.619511604309082, "learning_rate": 2.9909274397079014e-05, "loss": 1.8152, "step": 5474 }, { "epoch": 1.2795419356723436, "grad_norm": 2.4228320121765137, "learning_rate": 2.9909212030679992e-05, "loss": 2.0655, "step": 5475 }, { "epoch": 1.2797756419619644, "grad_norm": 3.6145565509796143, "learning_rate": 2.9909149642917515e-05, "loss": 1.8262, "step": 5476 }, { "epoch": 1.2800093482515849, "grad_norm": 3.9942409992218018, "learning_rate": 2.9909087233791674e-05, "loss": 1.7176, "step": 5477 }, { "epoch": 1.2802430545412053, "grad_norm": 2.573423147201538, "learning_rate": 2.9909024803302557e-05, "loss": 1.7831, "step": 5478 }, { "epoch": 1.2804767608308258, "grad_norm": 2.7351443767547607, "learning_rate": 2.9908962351450247e-05, "loss": 1.777, "step": 5479 }, { "epoch": 1.2807104671204463, "grad_norm": 6.049731254577637, "learning_rate": 2.9908899878234844e-05, "loss": 1.9504, "step": 5480 }, { "epoch": 1.2809441734100668, "grad_norm": 4.665690898895264, "learning_rate": 2.990883738365643e-05, "loss": 1.8632, "step": 5481 }, { "epoch": 1.2811778796996873, "grad_norm": 2.511673927307129, "learning_rate": 2.9908774867715096e-05, "loss": 1.8624, "step": 5482 }, { "epoch": 1.281411585989308, "grad_norm": 2.934884786605835, "learning_rate": 2.9908712330410934e-05, "loss": 1.8962, "step": 5483 }, { "epoch": 1.2816452922789285, "grad_norm": 2.133509397506714, "learning_rate": 2.9908649771744035e-05, "loss": 1.8588, "step": 5484 }, { "epoch": 1.281878998568549, "grad_norm": 2.5540430545806885, "learning_rate": 2.990858719171448e-05, "loss": 1.7467, "step": 5485 }, { "epoch": 1.2821127048581695, "grad_norm": 2.741940975189209, "learning_rate": 2.9908524590322366e-05, "loss": 1.9353, "step": 5486 }, { "epoch": 1.28234641114779, "grad_norm": 2.418943405151367, "learning_rate": 2.9908461967567785e-05, "loss": 1.8957, "step": 5487 }, { "epoch": 1.2825801174374105, "grad_norm": 2.432703733444214, "learning_rate": 2.9908399323450817e-05, "loss": 1.9212, "step": 5488 }, { "epoch": 1.282813823727031, "grad_norm": 2.433086633682251, "learning_rate": 2.990833665797156e-05, "loss": 1.6774, "step": 5489 }, { "epoch": 1.2830475300166515, "grad_norm": 2.920679807662964, "learning_rate": 2.99082739711301e-05, "loss": 1.7749, "step": 5490 }, { "epoch": 1.283281236306272, "grad_norm": 3.059351921081543, "learning_rate": 2.990821126292653e-05, "loss": 1.9804, "step": 5491 }, { "epoch": 1.2835149425958927, "grad_norm": 3.5613198280334473, "learning_rate": 2.9908148533360943e-05, "loss": 1.9546, "step": 5492 }, { "epoch": 1.2837486488855132, "grad_norm": 3.8165934085845947, "learning_rate": 2.9908085782433415e-05, "loss": 1.927, "step": 5493 }, { "epoch": 1.2839823551751337, "grad_norm": 2.5177645683288574, "learning_rate": 2.9908023010144048e-05, "loss": 1.8876, "step": 5494 }, { "epoch": 1.2842160614647542, "grad_norm": 2.4242360591888428, "learning_rate": 2.990796021649293e-05, "loss": 1.8386, "step": 5495 }, { "epoch": 1.2844497677543747, "grad_norm": 2.9905169010162354, "learning_rate": 2.990789740148015e-05, "loss": 1.8675, "step": 5496 }, { "epoch": 1.2846834740439952, "grad_norm": 2.271868944168091, "learning_rate": 2.9907834565105798e-05, "loss": 1.863, "step": 5497 }, { "epoch": 1.2849171803336157, "grad_norm": 2.3556365966796875, "learning_rate": 2.9907771707369963e-05, "loss": 1.8884, "step": 5498 }, { "epoch": 1.2851508866232362, "grad_norm": 2.6082067489624023, "learning_rate": 2.9907708828272736e-05, "loss": 1.7414, "step": 5499 }, { "epoch": 1.2853845929128567, "grad_norm": 2.489637613296509, "learning_rate": 2.9907645927814206e-05, "loss": 1.7778, "step": 5500 } ], "logging_steps": 1.0, "max_steps": 85560, "num_input_tokens_seen": 0, "num_train_epochs": 20, "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": 2.410404119400415e+18, "train_batch_size": 2, "trial_name": null, "trial_params": null }