{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.9990442816183498, "eval_steps": 500, "global_step": 3138, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0006371455877668047, "grad_norm": 45.95752780609825, "learning_rate": 2.105263157894737e-07, "loss": 3.0986, "step": 1 }, { "epoch": 0.0012742911755336094, "grad_norm": 36.24569437603479, "learning_rate": 4.210526315789474e-07, "loss": 2.7002, "step": 2 }, { "epoch": 0.001911436763300414, "grad_norm": 23.4237243623578, "learning_rate": 6.315789473684211e-07, "loss": 2.5681, "step": 3 }, { "epoch": 0.002548582351067219, "grad_norm": 214.67771689359765, "learning_rate": 8.421052631578948e-07, "loss": 3.7021, "step": 4 }, { "epoch": 0.0031857279388340237, "grad_norm": 17.64079860772153, "learning_rate": 1.0526315789473685e-06, "loss": 2.4258, "step": 5 }, { "epoch": 0.003822873526600828, "grad_norm": 18.677405718623927, "learning_rate": 1.2631578947368422e-06, "loss": 2.124, "step": 6 }, { "epoch": 0.004460019114367633, "grad_norm": 26.133181681163308, "learning_rate": 1.4736842105263159e-06, "loss": 2.8677, "step": 7 }, { "epoch": 0.005097164702134438, "grad_norm": 11.719861634496823, "learning_rate": 1.6842105263157895e-06, "loss": 1.2725, "step": 8 }, { "epoch": 0.005734310289901242, "grad_norm": 12.47019263116219, "learning_rate": 1.8947368421052634e-06, "loss": 1.4929, "step": 9 }, { "epoch": 0.0063714558776680474, "grad_norm": 27.245765244096937, "learning_rate": 2.105263157894737e-06, "loss": 3.124, "step": 10 }, { "epoch": 0.007008601465434852, "grad_norm": 15.376442446025528, "learning_rate": 2.3157894736842105e-06, "loss": 1.8757, "step": 11 }, { "epoch": 0.007645747053201656, "grad_norm": 13.164809083621904, "learning_rate": 2.5263157894736844e-06, "loss": 1.9058, "step": 12 }, { "epoch": 0.00828289264096846, "grad_norm": 17.556797173153324, "learning_rate": 2.7368421052631583e-06, "loss": 2.3081, "step": 13 }, { "epoch": 0.008920038228735267, "grad_norm": 17.416402016182435, "learning_rate": 2.9473684210526317e-06, "loss": 2.3413, "step": 14 }, { "epoch": 0.009557183816502071, "grad_norm": 16.491404812982495, "learning_rate": 3.157894736842105e-06, "loss": 2.9268, "step": 15 }, { "epoch": 0.010194329404268876, "grad_norm": 9.361970104417082, "learning_rate": 3.368421052631579e-06, "loss": 1.2715, "step": 16 }, { "epoch": 0.01083147499203568, "grad_norm": 9.608515406439366, "learning_rate": 3.578947368421053e-06, "loss": 1.3601, "step": 17 }, { "epoch": 0.011468620579802484, "grad_norm": 11.0406392603677, "learning_rate": 3.789473684210527e-06, "loss": 1.8503, "step": 18 }, { "epoch": 0.012105766167569289, "grad_norm": 7.709409861182795, "learning_rate": 4.000000000000001e-06, "loss": 1.3721, "step": 19 }, { "epoch": 0.012742911755336095, "grad_norm": 13.590957858691095, "learning_rate": 4.210526315789474e-06, "loss": 1.9077, "step": 20 }, { "epoch": 0.0133800573431029, "grad_norm": 12.764950075484615, "learning_rate": 4.4210526315789476e-06, "loss": 1.804, "step": 21 }, { "epoch": 0.014017202930869704, "grad_norm": 11.572857970744545, "learning_rate": 4.631578947368421e-06, "loss": 2.0754, "step": 22 }, { "epoch": 0.014654348518636508, "grad_norm": 9.531248481558716, "learning_rate": 4.842105263157895e-06, "loss": 1.9243, "step": 23 }, { "epoch": 0.015291494106403312, "grad_norm": 10.178081080235318, "learning_rate": 5.052631578947369e-06, "loss": 2.1104, "step": 24 }, { "epoch": 0.01592863969417012, "grad_norm": 6.943687912934176, "learning_rate": 5.263157894736842e-06, "loss": 1.4335, "step": 25 }, { "epoch": 0.01656578528193692, "grad_norm": 5.685076881965541, "learning_rate": 5.4736842105263165e-06, "loss": 1.0725, "step": 26 }, { "epoch": 0.017202930869703727, "grad_norm": 9.995826418219114, "learning_rate": 5.68421052631579e-06, "loss": 1.7068, "step": 27 }, { "epoch": 0.017840076457470534, "grad_norm": 11.19128221437317, "learning_rate": 5.8947368421052634e-06, "loss": 1.7932, "step": 28 }, { "epoch": 0.018477222045237336, "grad_norm": 11.126786260168483, "learning_rate": 6.105263157894738e-06, "loss": 1.8376, "step": 29 }, { "epoch": 0.019114367633004142, "grad_norm": 11.073610139805014, "learning_rate": 6.31578947368421e-06, "loss": 2.2104, "step": 30 }, { "epoch": 0.019751513220770945, "grad_norm": 9.88727472954615, "learning_rate": 6.526315789473685e-06, "loss": 1.9231, "step": 31 }, { "epoch": 0.02038865880853775, "grad_norm": 4.335911350133859, "learning_rate": 6.736842105263158e-06, "loss": 1.0359, "step": 32 }, { "epoch": 0.021025804396304557, "grad_norm": 9.478199887668643, "learning_rate": 6.947368421052632e-06, "loss": 1.9712, "step": 33 }, { "epoch": 0.02166294998407136, "grad_norm": 6.970126813945282, "learning_rate": 7.157894736842106e-06, "loss": 1.3826, "step": 34 }, { "epoch": 0.022300095571838166, "grad_norm": 8.233870274952354, "learning_rate": 7.368421052631579e-06, "loss": 1.7405, "step": 35 }, { "epoch": 0.02293724115960497, "grad_norm": 10.511931630484213, "learning_rate": 7.578947368421054e-06, "loss": 2.0059, "step": 36 }, { "epoch": 0.023574386747371775, "grad_norm": 9.396532562585158, "learning_rate": 7.789473684210526e-06, "loss": 2.1755, "step": 37 }, { "epoch": 0.024211532335138577, "grad_norm": 6.241276407655751, "learning_rate": 8.000000000000001e-06, "loss": 1.5938, "step": 38 }, { "epoch": 0.024848677922905384, "grad_norm": 6.105998224968607, "learning_rate": 8.210526315789475e-06, "loss": 1.3384, "step": 39 }, { "epoch": 0.02548582351067219, "grad_norm": 7.5224413432247905, "learning_rate": 8.421052631578948e-06, "loss": 2.1572, "step": 40 }, { "epoch": 0.026122969098438992, "grad_norm": 7.486029990735986, "learning_rate": 8.631578947368422e-06, "loss": 2.0088, "step": 41 }, { "epoch": 0.0267601146862058, "grad_norm": 5.270461711847247, "learning_rate": 8.842105263157895e-06, "loss": 1.2095, "step": 42 }, { "epoch": 0.0273972602739726, "grad_norm": 7.594559848183423, "learning_rate": 9.05263157894737e-06, "loss": 1.402, "step": 43 }, { "epoch": 0.028034405861739407, "grad_norm": 7.139162454501502, "learning_rate": 9.263157894736842e-06, "loss": 1.9497, "step": 44 }, { "epoch": 0.028671551449506213, "grad_norm": 7.4383460465025095, "learning_rate": 9.473684210526315e-06, "loss": 1.5212, "step": 45 }, { "epoch": 0.029308697037273016, "grad_norm": 7.15393903675183, "learning_rate": 9.68421052631579e-06, "loss": 1.4102, "step": 46 }, { "epoch": 0.029945842625039822, "grad_norm": 6.469850364787783, "learning_rate": 9.894736842105264e-06, "loss": 1.3447, "step": 47 }, { "epoch": 0.030582988212806625, "grad_norm": 9.434050181040856, "learning_rate": 1.0105263157894738e-05, "loss": 1.6428, "step": 48 }, { "epoch": 0.03122013380057343, "grad_norm": 8.07271204485426, "learning_rate": 1.0315789473684213e-05, "loss": 1.873, "step": 49 }, { "epoch": 0.03185727938834024, "grad_norm": 7.501955375493345, "learning_rate": 1.0526315789473684e-05, "loss": 1.7109, "step": 50 }, { "epoch": 0.03249442497610704, "grad_norm": 7.595561276670744, "learning_rate": 1.073684210526316e-05, "loss": 1.6691, "step": 51 }, { "epoch": 0.03313157056387384, "grad_norm": 7.693575864534909, "learning_rate": 1.0947368421052633e-05, "loss": 2.0122, "step": 52 }, { "epoch": 0.03376871615164065, "grad_norm": 5.26898958766443, "learning_rate": 1.1157894736842105e-05, "loss": 1.2305, "step": 53 }, { "epoch": 0.034405861739407455, "grad_norm": 9.87311821647473, "learning_rate": 1.136842105263158e-05, "loss": 2.4785, "step": 54 }, { "epoch": 0.03504300732717426, "grad_norm": 13.855480324215948, "learning_rate": 1.1578947368421053e-05, "loss": 2.5938, "step": 55 }, { "epoch": 0.03568015291494107, "grad_norm": 30.49559348372227, "learning_rate": 1.1789473684210527e-05, "loss": 1.7581, "step": 56 }, { "epoch": 0.036317298502707866, "grad_norm": 6.458047218008244, "learning_rate": 1.2e-05, "loss": 1.4075, "step": 57 }, { "epoch": 0.03695444409047467, "grad_norm": 6.605711107003584, "learning_rate": 1.2210526315789475e-05, "loss": 1.1396, "step": 58 }, { "epoch": 0.03759158967824148, "grad_norm": 7.1433782541905915, "learning_rate": 1.2421052631578949e-05, "loss": 1.4478, "step": 59 }, { "epoch": 0.038228735266008285, "grad_norm": 6.253969832201868, "learning_rate": 1.263157894736842e-05, "loss": 1.5005, "step": 60 }, { "epoch": 0.03886588085377509, "grad_norm": 11.999160929552698, "learning_rate": 1.2842105263157896e-05, "loss": 1.8115, "step": 61 }, { "epoch": 0.03950302644154189, "grad_norm": 8.540533042671003, "learning_rate": 1.305263157894737e-05, "loss": 1.8071, "step": 62 }, { "epoch": 0.040140172029308696, "grad_norm": 6.781280537745764, "learning_rate": 1.3263157894736843e-05, "loss": 1.6135, "step": 63 }, { "epoch": 0.0407773176170755, "grad_norm": 3.18916866137568, "learning_rate": 1.3473684210526316e-05, "loss": 0.6443, "step": 64 }, { "epoch": 0.04141446320484231, "grad_norm": 5.177250581967893, "learning_rate": 1.3684210526315791e-05, "loss": 1.1587, "step": 65 }, { "epoch": 0.042051608792609114, "grad_norm": 9.709316792246275, "learning_rate": 1.3894736842105265e-05, "loss": 2.085, "step": 66 }, { "epoch": 0.042688754380375914, "grad_norm": 7.51445518087513, "learning_rate": 1.4105263157894738e-05, "loss": 2.1582, "step": 67 }, { "epoch": 0.04332589996814272, "grad_norm": 7.773754891171214, "learning_rate": 1.4315789473684212e-05, "loss": 1.5769, "step": 68 }, { "epoch": 0.043963045555909526, "grad_norm": 8.088168980512949, "learning_rate": 1.4526315789473687e-05, "loss": 1.4109, "step": 69 }, { "epoch": 0.04460019114367633, "grad_norm": 22.89571852687989, "learning_rate": 1.4736842105263159e-05, "loss": 2.0304, "step": 70 }, { "epoch": 0.04523733673144314, "grad_norm": 6.593900434973564, "learning_rate": 1.4947368421052632e-05, "loss": 1.4399, "step": 71 }, { "epoch": 0.04587448231920994, "grad_norm": 8.744401934830632, "learning_rate": 1.5157894736842107e-05, "loss": 1.8789, "step": 72 }, { "epoch": 0.046511627906976744, "grad_norm": 9.690468237590709, "learning_rate": 1.536842105263158e-05, "loss": 1.7295, "step": 73 }, { "epoch": 0.04714877349474355, "grad_norm": 5.738081510659548, "learning_rate": 1.5578947368421052e-05, "loss": 1.7048, "step": 74 }, { "epoch": 0.047785919082510356, "grad_norm": 7.244241606326725, "learning_rate": 1.578947368421053e-05, "loss": 1.3792, "step": 75 }, { "epoch": 0.048423064670277155, "grad_norm": 8.880056854492686, "learning_rate": 1.6000000000000003e-05, "loss": 1.8694, "step": 76 }, { "epoch": 0.04906021025804396, "grad_norm": 6.7057419960253135, "learning_rate": 1.6210526315789473e-05, "loss": 1.7863, "step": 77 }, { "epoch": 0.04969735584581077, "grad_norm": 4.643440404281383, "learning_rate": 1.642105263157895e-05, "loss": 0.9589, "step": 78 }, { "epoch": 0.05033450143357757, "grad_norm": 16.187416673414575, "learning_rate": 1.6631578947368423e-05, "loss": 1.6807, "step": 79 }, { "epoch": 0.05097164702134438, "grad_norm": 5.290422043656508, "learning_rate": 1.6842105263157896e-05, "loss": 1.1682, "step": 80 }, { "epoch": 0.05160879260911118, "grad_norm": 6.581805973558028, "learning_rate": 1.705263157894737e-05, "loss": 1.4731, "step": 81 }, { "epoch": 0.052245938196877985, "grad_norm": 13.09448506643181, "learning_rate": 1.7263157894736843e-05, "loss": 1.697, "step": 82 }, { "epoch": 0.05288308378464479, "grad_norm": 5.041635155803943, "learning_rate": 1.7473684210526317e-05, "loss": 1.012, "step": 83 }, { "epoch": 0.0535202293724116, "grad_norm": 4.861003066774403, "learning_rate": 1.768421052631579e-05, "loss": 1.1045, "step": 84 }, { "epoch": 0.0541573749601784, "grad_norm": 7.2551689420615615, "learning_rate": 1.7894736842105264e-05, "loss": 1.98, "step": 85 }, { "epoch": 0.0547945205479452, "grad_norm": 9.845160551134803, "learning_rate": 1.810526315789474e-05, "loss": 1.7896, "step": 86 }, { "epoch": 0.05543166613571201, "grad_norm": 12.240087822923565, "learning_rate": 1.831578947368421e-05, "loss": 1.5967, "step": 87 }, { "epoch": 0.056068811723478815, "grad_norm": 7.255024423389392, "learning_rate": 1.8526315789473684e-05, "loss": 1.9863, "step": 88 }, { "epoch": 0.05670595731124562, "grad_norm": 7.657518038467798, "learning_rate": 1.873684210526316e-05, "loss": 1.738, "step": 89 }, { "epoch": 0.05734310289901243, "grad_norm": 6.042075278696005, "learning_rate": 1.894736842105263e-05, "loss": 1.4049, "step": 90 }, { "epoch": 0.057980248486779226, "grad_norm": 4.121579639953916, "learning_rate": 1.9157894736842108e-05, "loss": 0.981, "step": 91 }, { "epoch": 0.05861739407454603, "grad_norm": 19.393174660054704, "learning_rate": 1.936842105263158e-05, "loss": 1.6194, "step": 92 }, { "epoch": 0.05925453966231284, "grad_norm": 6.093414836079516, "learning_rate": 1.9578947368421055e-05, "loss": 1.4207, "step": 93 }, { "epoch": 0.059891685250079645, "grad_norm": 6.60794020422448, "learning_rate": 1.9789473684210528e-05, "loss": 1.4172, "step": 94 }, { "epoch": 0.06052883083784645, "grad_norm": 6.36605460707131, "learning_rate": 2e-05, "loss": 1.3411, "step": 95 }, { "epoch": 0.06116597642561325, "grad_norm": 7.015393707406119, "learning_rate": 1.9999995203678174e-05, "loss": 1.5781, "step": 96 }, { "epoch": 0.061803122013380056, "grad_norm": 5.580221025308569, "learning_rate": 1.999998081471781e-05, "loss": 1.5864, "step": 97 }, { "epoch": 0.06244026760114686, "grad_norm": 6.581838441202273, "learning_rate": 1.9999956833134246e-05, "loss": 1.5352, "step": 98 }, { "epoch": 0.06307741318891366, "grad_norm": 4.330865580440961, "learning_rate": 1.9999923258953037e-05, "loss": 1.1902, "step": 99 }, { "epoch": 0.06371455877668047, "grad_norm": 4.131897272287731, "learning_rate": 1.9999880092209972e-05, "loss": 1.0498, "step": 100 }, { "epoch": 0.06435170436444727, "grad_norm": 8.84568953831988, "learning_rate": 1.9999827332951055e-05, "loss": 1.7427, "step": 101 }, { "epoch": 0.06498884995221409, "grad_norm": 5.2802214128342815, "learning_rate": 1.999976498123253e-05, "loss": 1.3831, "step": 102 }, { "epoch": 0.06562599553998089, "grad_norm": 8.5822422601699, "learning_rate": 1.9999693037120848e-05, "loss": 2.0273, "step": 103 }, { "epoch": 0.06626314112774769, "grad_norm": 9.042949179597462, "learning_rate": 1.999961150069269e-05, "loss": 1.6223, "step": 104 }, { "epoch": 0.0669002867155145, "grad_norm": 6.2398647716593185, "learning_rate": 1.999952037203496e-05, "loss": 1.697, "step": 105 }, { "epoch": 0.0675374323032813, "grad_norm": 5.251600576357235, "learning_rate": 1.9999419651244794e-05, "loss": 1.2766, "step": 106 }, { "epoch": 0.06817457789104811, "grad_norm": 7.232082083824022, "learning_rate": 1.999930933842954e-05, "loss": 1.2905, "step": 107 }, { "epoch": 0.06881172347881491, "grad_norm": 5.697789891362125, "learning_rate": 1.999918943370677e-05, "loss": 1.5046, "step": 108 }, { "epoch": 0.06944886906658171, "grad_norm": 5.751911081850657, "learning_rate": 1.9999059937204298e-05, "loss": 1.0449, "step": 109 }, { "epoch": 0.07008601465434852, "grad_norm": 6.584244845525387, "learning_rate": 1.9998920849060138e-05, "loss": 1.344, "step": 110 }, { "epoch": 0.07072316024211532, "grad_norm": 7.085399129976097, "learning_rate": 1.999877216942254e-05, "loss": 1.6775, "step": 111 }, { "epoch": 0.07136030582988213, "grad_norm": 5.976214935143067, "learning_rate": 1.9998613898449972e-05, "loss": 1.8533, "step": 112 }, { "epoch": 0.07199745141764893, "grad_norm": 7.442628712911392, "learning_rate": 1.9998446036311127e-05, "loss": 1.7399, "step": 113 }, { "epoch": 0.07263459700541573, "grad_norm": 7.668576032376383, "learning_rate": 1.9998268583184926e-05, "loss": 1.8801, "step": 114 }, { "epoch": 0.07327174259318255, "grad_norm": 6.534954436697609, "learning_rate": 1.9998081539260503e-05, "loss": 1.5098, "step": 115 }, { "epoch": 0.07390888818094934, "grad_norm": 8.303163163057453, "learning_rate": 1.9997884904737212e-05, "loss": 1.3811, "step": 116 }, { "epoch": 0.07454603376871616, "grad_norm": 5.152422259450472, "learning_rate": 1.999767867982465e-05, "loss": 1.3572, "step": 117 }, { "epoch": 0.07518317935648296, "grad_norm": 2.967323247384616, "learning_rate": 1.999746286474261e-05, "loss": 1.0052, "step": 118 }, { "epoch": 0.07582032494424976, "grad_norm": 5.429468337945769, "learning_rate": 1.9997237459721125e-05, "loss": 1.3198, "step": 119 }, { "epoch": 0.07645747053201657, "grad_norm": 4.065233780555616, "learning_rate": 1.9997002465000438e-05, "loss": 1.23, "step": 120 }, { "epoch": 0.07709461611978337, "grad_norm": 8.89595388249666, "learning_rate": 1.9996757880831023e-05, "loss": 1.3774, "step": 121 }, { "epoch": 0.07773176170755018, "grad_norm": 9.790998090238158, "learning_rate": 1.999650370747357e-05, "loss": 1.6335, "step": 122 }, { "epoch": 0.07836890729531698, "grad_norm": 5.582464667359614, "learning_rate": 1.999623994519898e-05, "loss": 1.2546, "step": 123 }, { "epoch": 0.07900605288308378, "grad_norm": 9.518362969384498, "learning_rate": 1.9995966594288395e-05, "loss": 1.7686, "step": 124 }, { "epoch": 0.07964319847085059, "grad_norm": 5.956986895842951, "learning_rate": 1.999568365503316e-05, "loss": 1.5254, "step": 125 }, { "epoch": 0.08028034405861739, "grad_norm": 15.069071033207734, "learning_rate": 1.999539112773485e-05, "loss": 1.4268, "step": 126 }, { "epoch": 0.0809174896463842, "grad_norm": 7.440836341987204, "learning_rate": 1.9995089012705248e-05, "loss": 1.6788, "step": 127 }, { "epoch": 0.081554635234151, "grad_norm": 7.691017963191412, "learning_rate": 1.9994777310266367e-05, "loss": 1.8555, "step": 128 }, { "epoch": 0.0821917808219178, "grad_norm": 4.063538039293008, "learning_rate": 1.999445602075044e-05, "loss": 1.1339, "step": 129 }, { "epoch": 0.08282892640968462, "grad_norm": 5.3898301408753895, "learning_rate": 1.9994125144499905e-05, "loss": 1.2683, "step": 130 }, { "epoch": 0.08346607199745142, "grad_norm": 18.692831432248262, "learning_rate": 1.9993784681867426e-05, "loss": 1.4407, "step": 131 }, { "epoch": 0.08410321758521823, "grad_norm": 8.269603282900995, "learning_rate": 1.9993434633215892e-05, "loss": 1.5667, "step": 132 }, { "epoch": 0.08474036317298503, "grad_norm": 8.179947862429042, "learning_rate": 1.99930749989184e-05, "loss": 1.406, "step": 133 }, { "epoch": 0.08537750876075183, "grad_norm": 7.31731250508769, "learning_rate": 1.9992705779358264e-05, "loss": 1.9463, "step": 134 }, { "epoch": 0.08601465434851864, "grad_norm": 6.39907874702958, "learning_rate": 1.9992326974929016e-05, "loss": 1.3799, "step": 135 }, { "epoch": 0.08665179993628544, "grad_norm": 5.634893472521617, "learning_rate": 1.9991938586034406e-05, "loss": 1.4388, "step": 136 }, { "epoch": 0.08728894552405225, "grad_norm": 5.74147244187628, "learning_rate": 1.9991540613088393e-05, "loss": 1.1838, "step": 137 }, { "epoch": 0.08792609111181905, "grad_norm": 9.917361408871427, "learning_rate": 1.9991133056515166e-05, "loss": 1.7678, "step": 138 }, { "epoch": 0.08856323669958585, "grad_norm": 3.8338518641340587, "learning_rate": 1.999071591674911e-05, "loss": 0.9486, "step": 139 }, { "epoch": 0.08920038228735266, "grad_norm": 6.500832639033168, "learning_rate": 1.999028919423484e-05, "loss": 1.594, "step": 140 }, { "epoch": 0.08983752787511946, "grad_norm": 8.76647617864947, "learning_rate": 1.9989852889427172e-05, "loss": 1.7522, "step": 141 }, { "epoch": 0.09047467346288628, "grad_norm": 6.601130144352926, "learning_rate": 1.9989407002791144e-05, "loss": 1.678, "step": 142 }, { "epoch": 0.09111181905065308, "grad_norm": 12.750206404572252, "learning_rate": 1.9988951534802e-05, "loss": 1.55, "step": 143 }, { "epoch": 0.09174896463841987, "grad_norm": 6.84411071741464, "learning_rate": 1.9988486485945206e-05, "loss": 1.5515, "step": 144 }, { "epoch": 0.09238611022618669, "grad_norm": 10.64829009923721, "learning_rate": 1.9988011856716434e-05, "loss": 1.9155, "step": 145 }, { "epoch": 0.09302325581395349, "grad_norm": 11.12978135105461, "learning_rate": 1.9987527647621563e-05, "loss": 1.864, "step": 146 }, { "epoch": 0.09366040140172029, "grad_norm": 7.229592348350161, "learning_rate": 1.9987033859176686e-05, "loss": 1.6748, "step": 147 }, { "epoch": 0.0942975469894871, "grad_norm": 10.139341703115708, "learning_rate": 1.9986530491908117e-05, "loss": 1.637, "step": 148 }, { "epoch": 0.0949346925772539, "grad_norm": 9.62343371030914, "learning_rate": 1.9986017546352356e-05, "loss": 1.616, "step": 149 }, { "epoch": 0.09557183816502071, "grad_norm": 6.9222067988324865, "learning_rate": 1.998549502305614e-05, "loss": 1.8215, "step": 150 }, { "epoch": 0.09620898375278751, "grad_norm": 3.6561394625203123, "learning_rate": 1.9984962922576382e-05, "loss": 0.9147, "step": 151 }, { "epoch": 0.09684612934055431, "grad_norm": 84.93979515484092, "learning_rate": 1.998442124548024e-05, "loss": 1.4602, "step": 152 }, { "epoch": 0.09748327492832112, "grad_norm": 8.768925680606557, "learning_rate": 1.9983869992345046e-05, "loss": 1.9592, "step": 153 }, { "epoch": 0.09812042051608792, "grad_norm": 10.651364188695549, "learning_rate": 1.998330916375836e-05, "loss": 1.7549, "step": 154 }, { "epoch": 0.09875756610385474, "grad_norm": 8.196700496385358, "learning_rate": 1.998273876031794e-05, "loss": 2.2612, "step": 155 }, { "epoch": 0.09939471169162153, "grad_norm": 4.635176067234948, "learning_rate": 1.9982158782631746e-05, "loss": 1.3376, "step": 156 }, { "epoch": 0.10003185727938833, "grad_norm": 6.574406140303575, "learning_rate": 1.9981569231317953e-05, "loss": 1.5762, "step": 157 }, { "epoch": 0.10066900286715515, "grad_norm": 4.833003686477079, "learning_rate": 1.998097010700493e-05, "loss": 1.1989, "step": 158 }, { "epoch": 0.10130614845492195, "grad_norm": 3.7594249070906023, "learning_rate": 1.9980361410331253e-05, "loss": 0.8507, "step": 159 }, { "epoch": 0.10194329404268876, "grad_norm": 5.158189641275894, "learning_rate": 1.9979743141945702e-05, "loss": 0.9296, "step": 160 }, { "epoch": 0.10258043963045556, "grad_norm": 17.967097551030427, "learning_rate": 1.9979115302507257e-05, "loss": 1.2236, "step": 161 }, { "epoch": 0.10321758521822236, "grad_norm": 7.317205263024187, "learning_rate": 1.99784778926851e-05, "loss": 1.2971, "step": 162 }, { "epoch": 0.10385473080598917, "grad_norm": 6.141461743958987, "learning_rate": 1.997783091315862e-05, "loss": 1.2284, "step": 163 }, { "epoch": 0.10449187639375597, "grad_norm": 9.210315439980231, "learning_rate": 1.9977174364617392e-05, "loss": 1.2604, "step": 164 }, { "epoch": 0.10512902198152278, "grad_norm": 7.464597912894261, "learning_rate": 1.9976508247761196e-05, "loss": 1.7004, "step": 165 }, { "epoch": 0.10576616756928958, "grad_norm": 4.632733147595637, "learning_rate": 1.9975832563300024e-05, "loss": 0.9186, "step": 166 }, { "epoch": 0.10640331315705638, "grad_norm": 7.415054412281283, "learning_rate": 1.9975147311954042e-05, "loss": 1.3074, "step": 167 }, { "epoch": 0.1070404587448232, "grad_norm": 4.34197792439413, "learning_rate": 1.9974452494453628e-05, "loss": 0.9679, "step": 168 }, { "epoch": 0.10767760433259, "grad_norm": 4.594783820916168, "learning_rate": 1.997374811153935e-05, "loss": 1.3367, "step": 169 }, { "epoch": 0.1083147499203568, "grad_norm": 5.002601282999489, "learning_rate": 1.9973034163961987e-05, "loss": 1.2422, "step": 170 }, { "epoch": 0.1089518955081236, "grad_norm": 6.111958570303732, "learning_rate": 1.9972310652482486e-05, "loss": 1.3787, "step": 171 }, { "epoch": 0.1095890410958904, "grad_norm": 6.028459897630762, "learning_rate": 1.9971577577872003e-05, "loss": 1.4543, "step": 172 }, { "epoch": 0.11022618668365722, "grad_norm": 2.692924576439833, "learning_rate": 1.997083494091189e-05, "loss": 0.7395, "step": 173 }, { "epoch": 0.11086333227142402, "grad_norm": 8.404327197202436, "learning_rate": 1.9970082742393683e-05, "loss": 1.9019, "step": 174 }, { "epoch": 0.11150047785919083, "grad_norm": 8.362583295949792, "learning_rate": 1.996932098311911e-05, "loss": 1.506, "step": 175 }, { "epoch": 0.11213762344695763, "grad_norm": 5.496804401513692, "learning_rate": 1.99685496639001e-05, "loss": 1.1753, "step": 176 }, { "epoch": 0.11277476903472443, "grad_norm": 7.097233360597305, "learning_rate": 1.9967768785558753e-05, "loss": 1.2671, "step": 177 }, { "epoch": 0.11341191462249124, "grad_norm": 9.561672919079236, "learning_rate": 1.9966978348927376e-05, "loss": 1.3762, "step": 178 }, { "epoch": 0.11404906021025804, "grad_norm": 8.365142912616124, "learning_rate": 1.996617835484845e-05, "loss": 1.3091, "step": 179 }, { "epoch": 0.11468620579802485, "grad_norm": 7.904992296888948, "learning_rate": 1.9965368804174645e-05, "loss": 1.9351, "step": 180 }, { "epoch": 0.11532335138579165, "grad_norm": 3.8498315716561082, "learning_rate": 1.9964549697768827e-05, "loss": 0.9954, "step": 181 }, { "epoch": 0.11596049697355845, "grad_norm": 7.163198916135486, "learning_rate": 1.996372103650404e-05, "loss": 1.6108, "step": 182 }, { "epoch": 0.11659764256132527, "grad_norm": 7.065639064697123, "learning_rate": 1.996288282126351e-05, "loss": 1.6089, "step": 183 }, { "epoch": 0.11723478814909206, "grad_norm": 7.732066896087199, "learning_rate": 1.9962035052940642e-05, "loss": 1.4573, "step": 184 }, { "epoch": 0.11787193373685888, "grad_norm": 5.073464443196689, "learning_rate": 1.996117773243904e-05, "loss": 1.0519, "step": 185 }, { "epoch": 0.11850907932462568, "grad_norm": 6.384650153565918, "learning_rate": 1.996031086067247e-05, "loss": 1.5272, "step": 186 }, { "epoch": 0.11914622491239248, "grad_norm": 12.60129277241348, "learning_rate": 1.9959434438564895e-05, "loss": 2.2119, "step": 187 }, { "epoch": 0.11978337050015929, "grad_norm": 7.464178454053418, "learning_rate": 1.995854846705044e-05, "loss": 1.7761, "step": 188 }, { "epoch": 0.12042051608792609, "grad_norm": 7.460585497774486, "learning_rate": 1.995765294707342e-05, "loss": 1.4878, "step": 189 }, { "epoch": 0.1210576616756929, "grad_norm": 5.41911247285965, "learning_rate": 1.995674787958833e-05, "loss": 1.5476, "step": 190 }, { "epoch": 0.1216948072634597, "grad_norm": 5.589469583114178, "learning_rate": 1.9955833265559828e-05, "loss": 1.2112, "step": 191 }, { "epoch": 0.1223319528512265, "grad_norm": 2.396446670843477, "learning_rate": 1.995490910596276e-05, "loss": 0.6198, "step": 192 }, { "epoch": 0.12296909843899331, "grad_norm": 9.157213527068446, "learning_rate": 1.9953975401782134e-05, "loss": 0.9635, "step": 193 }, { "epoch": 0.12360624402676011, "grad_norm": 10.175701352479368, "learning_rate": 1.9953032154013146e-05, "loss": 2.0396, "step": 194 }, { "epoch": 0.12424338961452693, "grad_norm": 11.679676854092229, "learning_rate": 1.995207936366115e-05, "loss": 1.812, "step": 195 }, { "epoch": 0.12488053520229372, "grad_norm": 6.426852284040646, "learning_rate": 1.995111703174168e-05, "loss": 1.2733, "step": 196 }, { "epoch": 0.12551768079006054, "grad_norm": 12.271456672215066, "learning_rate": 1.995014515928044e-05, "loss": 1.217, "step": 197 }, { "epoch": 0.12615482637782732, "grad_norm": 15.376403723347785, "learning_rate": 1.9949163747313287e-05, "loss": 1.853, "step": 198 }, { "epoch": 0.12679197196559414, "grad_norm": 75.56259163398282, "learning_rate": 1.9948172796886272e-05, "loss": 1.3674, "step": 199 }, { "epoch": 0.12742911755336095, "grad_norm": 5.389234703295916, "learning_rate": 1.9947172309055595e-05, "loss": 1.4507, "step": 200 }, { "epoch": 0.12806626314112773, "grad_norm": 6.5437194834375365, "learning_rate": 1.9946162284887624e-05, "loss": 1.5703, "step": 201 }, { "epoch": 0.12870340872889455, "grad_norm": 36.28335160564543, "learning_rate": 1.9945142725458893e-05, "loss": 1.625, "step": 202 }, { "epoch": 0.12934055431666136, "grad_norm": 11.317957742778804, "learning_rate": 1.9944113631856096e-05, "loss": 1.3004, "step": 203 }, { "epoch": 0.12997769990442817, "grad_norm": 12.36736138668665, "learning_rate": 1.9943075005176093e-05, "loss": 1.5249, "step": 204 }, { "epoch": 0.13061484549219496, "grad_norm": 8.55919379814451, "learning_rate": 1.9942026846525908e-05, "loss": 1.6096, "step": 205 }, { "epoch": 0.13125199107996177, "grad_norm": 5.678995984834073, "learning_rate": 1.9940969157022713e-05, "loss": 0.9714, "step": 206 }, { "epoch": 0.13188913666772858, "grad_norm": 5.251064213161032, "learning_rate": 1.9939901937793854e-05, "loss": 1.2939, "step": 207 }, { "epoch": 0.13252628225549537, "grad_norm": 6.415934351959902, "learning_rate": 1.9938825189976816e-05, "loss": 1.316, "step": 208 }, { "epoch": 0.13316342784326218, "grad_norm": 15.006459583801933, "learning_rate": 1.9937738914719256e-05, "loss": 1.729, "step": 209 }, { "epoch": 0.133800573431029, "grad_norm": 7.09140166904399, "learning_rate": 1.9936643113178983e-05, "loss": 1.7256, "step": 210 }, { "epoch": 0.13443771901879578, "grad_norm": 8.226813197760722, "learning_rate": 1.993553778652395e-05, "loss": 1.6333, "step": 211 }, { "epoch": 0.1350748646065626, "grad_norm": 8.910437230654765, "learning_rate": 1.9934422935932267e-05, "loss": 1.5723, "step": 212 }, { "epoch": 0.1357120101943294, "grad_norm": 8.300846910147735, "learning_rate": 1.99332985625922e-05, "loss": 1.6309, "step": 213 }, { "epoch": 0.13634915578209622, "grad_norm": 14.425196455716126, "learning_rate": 1.9932164667702168e-05, "loss": 1.6108, "step": 214 }, { "epoch": 0.136986301369863, "grad_norm": 8.102314736494703, "learning_rate": 1.993102125247072e-05, "loss": 1.3403, "step": 215 }, { "epoch": 0.13762344695762982, "grad_norm": 3.125675130790247, "learning_rate": 1.9929868318116575e-05, "loss": 0.827, "step": 216 }, { "epoch": 0.13826059254539663, "grad_norm": 5.227192975144215, "learning_rate": 1.992870586586858e-05, "loss": 1.2073, "step": 217 }, { "epoch": 0.13889773813316342, "grad_norm": 14.557717491603174, "learning_rate": 1.9927533896965735e-05, "loss": 1.3281, "step": 218 }, { "epoch": 0.13953488372093023, "grad_norm": 5.216798826344003, "learning_rate": 1.9926352412657188e-05, "loss": 1.2173, "step": 219 }, { "epoch": 0.14017202930869704, "grad_norm": 6.717830649058362, "learning_rate": 1.9925161414202215e-05, "loss": 1.3738, "step": 220 }, { "epoch": 0.14080917489646383, "grad_norm": 5.612247608643661, "learning_rate": 1.992396090287025e-05, "loss": 1.5601, "step": 221 }, { "epoch": 0.14144632048423064, "grad_norm": 6.415900445065662, "learning_rate": 1.9922750879940846e-05, "loss": 1.2378, "step": 222 }, { "epoch": 0.14208346607199746, "grad_norm": 6.623002991415273, "learning_rate": 1.992153134670371e-05, "loss": 0.8511, "step": 223 }, { "epoch": 0.14272061165976427, "grad_norm": 10.13903526732582, "learning_rate": 1.9920302304458687e-05, "loss": 1.2318, "step": 224 }, { "epoch": 0.14335775724753105, "grad_norm": 6.5456409820035395, "learning_rate": 1.9919063754515746e-05, "loss": 1.4858, "step": 225 }, { "epoch": 0.14399490283529787, "grad_norm": 4.463411903426627, "learning_rate": 1.991781569819499e-05, "loss": 1.0828, "step": 226 }, { "epoch": 0.14463204842306468, "grad_norm": 6.702318213529139, "learning_rate": 1.9916558136826663e-05, "loss": 1.6008, "step": 227 }, { "epoch": 0.14526919401083146, "grad_norm": 6.039540948056287, "learning_rate": 1.991529107175114e-05, "loss": 1.4917, "step": 228 }, { "epoch": 0.14590633959859828, "grad_norm": 5.988921378681106, "learning_rate": 1.991401450431891e-05, "loss": 1.578, "step": 229 }, { "epoch": 0.1465434851863651, "grad_norm": 4.621996233522874, "learning_rate": 1.9912728435890615e-05, "loss": 1.3596, "step": 230 }, { "epoch": 0.14718063077413188, "grad_norm": 8.011238299435371, "learning_rate": 1.9911432867837003e-05, "loss": 1.6416, "step": 231 }, { "epoch": 0.1478177763618987, "grad_norm": 6.523306243649803, "learning_rate": 1.9910127801538955e-05, "loss": 1.3394, "step": 232 }, { "epoch": 0.1484549219496655, "grad_norm": 6.947763881276892, "learning_rate": 1.9908813238387476e-05, "loss": 1.5945, "step": 233 }, { "epoch": 0.14909206753743232, "grad_norm": 5.076389732303654, "learning_rate": 1.9907489179783687e-05, "loss": 1.1843, "step": 234 }, { "epoch": 0.1497292131251991, "grad_norm": 5.809367419320748, "learning_rate": 1.9906155627138843e-05, "loss": 1.3683, "step": 235 }, { "epoch": 0.15036635871296591, "grad_norm": 11.13165097847537, "learning_rate": 1.990481258187431e-05, "loss": 1.6697, "step": 236 }, { "epoch": 0.15100350430073273, "grad_norm": 6.770470898504377, "learning_rate": 1.990346004542156e-05, "loss": 1.6196, "step": 237 }, { "epoch": 0.1516406498884995, "grad_norm": 9.119043347822311, "learning_rate": 1.9902098019222208e-05, "loss": 1.8105, "step": 238 }, { "epoch": 0.15227779547626633, "grad_norm": 6.997503560087028, "learning_rate": 1.9900726504727958e-05, "loss": 1.2678, "step": 239 }, { "epoch": 0.15291494106403314, "grad_norm": 6.0211956624002045, "learning_rate": 1.9899345503400643e-05, "loss": 1.5166, "step": 240 }, { "epoch": 0.15355208665179992, "grad_norm": 5.356469761369459, "learning_rate": 1.9897955016712202e-05, "loss": 1.4453, "step": 241 }, { "epoch": 0.15418923223956674, "grad_norm": 5.08516803681534, "learning_rate": 1.989655504614468e-05, "loss": 1.1664, "step": 242 }, { "epoch": 0.15482637782733355, "grad_norm": 5.237615904163628, "learning_rate": 1.9895145593190246e-05, "loss": 1.1743, "step": 243 }, { "epoch": 0.15546352341510036, "grad_norm": 6.12283551771803, "learning_rate": 1.9893726659351148e-05, "loss": 1.6526, "step": 244 }, { "epoch": 0.15610066900286715, "grad_norm": 4.558503385234118, "learning_rate": 1.989229824613977e-05, "loss": 1.5659, "step": 245 }, { "epoch": 0.15673781459063396, "grad_norm": 10.03558396486273, "learning_rate": 1.9890860355078573e-05, "loss": 1.5151, "step": 246 }, { "epoch": 0.15737496017840077, "grad_norm": 6.75814250536217, "learning_rate": 1.988941298770014e-05, "loss": 1.3384, "step": 247 }, { "epoch": 0.15801210576616756, "grad_norm": 7.96751945616568, "learning_rate": 1.9887956145547148e-05, "loss": 1.5898, "step": 248 }, { "epoch": 0.15864925135393437, "grad_norm": 6.945812130367764, "learning_rate": 1.988648983017237e-05, "loss": 1.3641, "step": 249 }, { "epoch": 0.15928639694170119, "grad_norm": 5.006557934432128, "learning_rate": 1.988501404313867e-05, "loss": 1.1799, "step": 250 }, { "epoch": 0.15992354252946797, "grad_norm": 5.87318463133434, "learning_rate": 1.9883528786019023e-05, "loss": 1.3333, "step": 251 }, { "epoch": 0.16056068811723478, "grad_norm": 5.948961124628064, "learning_rate": 1.988203406039649e-05, "loss": 1.0991, "step": 252 }, { "epoch": 0.1611978337050016, "grad_norm": 5.6587786851186745, "learning_rate": 1.988052986786421e-05, "loss": 1.3186, "step": 253 }, { "epoch": 0.1618349792927684, "grad_norm": 9.72941997334175, "learning_rate": 1.987901621002544e-05, "loss": 1.1982, "step": 254 }, { "epoch": 0.1624721248805352, "grad_norm": 7.515642836032979, "learning_rate": 1.9877493088493508e-05, "loss": 1.6185, "step": 255 }, { "epoch": 0.163109270468302, "grad_norm": 10.678821236419989, "learning_rate": 1.9875960504891828e-05, "loss": 1.3796, "step": 256 }, { "epoch": 0.16374641605606882, "grad_norm": 5.0931248001797575, "learning_rate": 1.9874418460853903e-05, "loss": 1.2561, "step": 257 }, { "epoch": 0.1643835616438356, "grad_norm": 8.314142063113119, "learning_rate": 1.9872866958023324e-05, "loss": 1.8608, "step": 258 }, { "epoch": 0.16502070723160242, "grad_norm": 6.667558058652838, "learning_rate": 1.987130599805376e-05, "loss": 1.5686, "step": 259 }, { "epoch": 0.16565785281936923, "grad_norm": 4.7544317706831745, "learning_rate": 1.9869735582608953e-05, "loss": 1.3201, "step": 260 }, { "epoch": 0.16629499840713602, "grad_norm": 5.707620027367956, "learning_rate": 1.986815571336273e-05, "loss": 1.08, "step": 261 }, { "epoch": 0.16693214399490283, "grad_norm": 3.129390772344477, "learning_rate": 1.9866566391998995e-05, "loss": 0.6615, "step": 262 }, { "epoch": 0.16756928958266964, "grad_norm": 6.8360137780257855, "learning_rate": 1.986496762021173e-05, "loss": 1.6982, "step": 263 }, { "epoch": 0.16820643517043646, "grad_norm": 5.424541242215366, "learning_rate": 1.986335939970498e-05, "loss": 1.6484, "step": 264 }, { "epoch": 0.16884358075820324, "grad_norm": 4.913342758851015, "learning_rate": 1.9861741732192866e-05, "loss": 1.2371, "step": 265 }, { "epoch": 0.16948072634597006, "grad_norm": 4.967322218053952, "learning_rate": 1.986011461939958e-05, "loss": 1.3108, "step": 266 }, { "epoch": 0.17011787193373687, "grad_norm": 6.458050292483352, "learning_rate": 1.9858478063059373e-05, "loss": 1.9893, "step": 267 }, { "epoch": 0.17075501752150365, "grad_norm": 3.9611899785431386, "learning_rate": 1.9856832064916575e-05, "loss": 1.0947, "step": 268 }, { "epoch": 0.17139216310927047, "grad_norm": 7.136112451961382, "learning_rate": 1.9855176626725567e-05, "loss": 1.4167, "step": 269 }, { "epoch": 0.17202930869703728, "grad_norm": 5.454198024872833, "learning_rate": 1.98535117502508e-05, "loss": 1.3955, "step": 270 }, { "epoch": 0.17266645428480407, "grad_norm": 5.2674083932157645, "learning_rate": 1.9851837437266776e-05, "loss": 1.1174, "step": 271 }, { "epoch": 0.17330359987257088, "grad_norm": 7.024772782937848, "learning_rate": 1.9850153689558064e-05, "loss": 1.7683, "step": 272 }, { "epoch": 0.1739407454603377, "grad_norm": 6.939523968791263, "learning_rate": 1.9848460508919287e-05, "loss": 1.6138, "step": 273 }, { "epoch": 0.1745778910481045, "grad_norm": 14.52808451832565, "learning_rate": 1.984675789715512e-05, "loss": 1.8848, "step": 274 }, { "epoch": 0.1752150366358713, "grad_norm": 5.8637357200164475, "learning_rate": 1.9845045856080286e-05, "loss": 1.056, "step": 275 }, { "epoch": 0.1758521822236381, "grad_norm": 6.265450083532673, "learning_rate": 1.9843324387519568e-05, "loss": 1.4075, "step": 276 }, { "epoch": 0.17648932781140492, "grad_norm": 4.35051131171262, "learning_rate": 1.9841593493307788e-05, "loss": 1.332, "step": 277 }, { "epoch": 0.1771264733991717, "grad_norm": 6.955320111966914, "learning_rate": 1.983985317528982e-05, "loss": 1.6685, "step": 278 }, { "epoch": 0.17776361898693852, "grad_norm": 14.48459583526567, "learning_rate": 1.9838103435320583e-05, "loss": 1.3062, "step": 279 }, { "epoch": 0.17840076457470533, "grad_norm": 7.859407707845617, "learning_rate": 1.9836344275265032e-05, "loss": 1.7349, "step": 280 }, { "epoch": 0.1790379101624721, "grad_norm": 7.139767500417876, "learning_rate": 1.9834575696998174e-05, "loss": 1.6907, "step": 281 }, { "epoch": 0.17967505575023893, "grad_norm": 8.40004605229434, "learning_rate": 1.9832797702405035e-05, "loss": 1.0089, "step": 282 }, { "epoch": 0.18031220133800574, "grad_norm": 11.802213573471763, "learning_rate": 1.9831010293380698e-05, "loss": 1.4015, "step": 283 }, { "epoch": 0.18094934692577255, "grad_norm": 8.215895404620946, "learning_rate": 1.982921347183027e-05, "loss": 1.5378, "step": 284 }, { "epoch": 0.18158649251353934, "grad_norm": 10.287942304294447, "learning_rate": 1.9827407239668892e-05, "loss": 1.5464, "step": 285 }, { "epoch": 0.18222363810130615, "grad_norm": 7.016773002569816, "learning_rate": 1.9825591598821734e-05, "loss": 1.3909, "step": 286 }, { "epoch": 0.18286078368907296, "grad_norm": 8.914862233300935, "learning_rate": 1.9823766551224e-05, "loss": 1.7092, "step": 287 }, { "epoch": 0.18349792927683975, "grad_norm": 7.943482006351951, "learning_rate": 1.9821932098820903e-05, "loss": 1.531, "step": 288 }, { "epoch": 0.18413507486460656, "grad_norm": 7.553012029873069, "learning_rate": 1.982008824356771e-05, "loss": 1.7407, "step": 289 }, { "epoch": 0.18477222045237338, "grad_norm": 4.002884729933982, "learning_rate": 1.981823498742968e-05, "loss": 1.0311, "step": 290 }, { "epoch": 0.18540936604014016, "grad_norm": 8.440779104740699, "learning_rate": 1.981637233238211e-05, "loss": 1.6492, "step": 291 }, { "epoch": 0.18604651162790697, "grad_norm": 9.703704609182564, "learning_rate": 1.9814500280410313e-05, "loss": 1.8496, "step": 292 }, { "epoch": 0.1866836572156738, "grad_norm": 7.524776719107354, "learning_rate": 1.9812618833509605e-05, "loss": 1.4939, "step": 293 }, { "epoch": 0.18732080280344057, "grad_norm": 5.854513252613075, "learning_rate": 1.9810727993685335e-05, "loss": 1.1879, "step": 294 }, { "epoch": 0.18795794839120739, "grad_norm": 10.10343562723306, "learning_rate": 1.9808827762952845e-05, "loss": 1.4651, "step": 295 }, { "epoch": 0.1885950939789742, "grad_norm": 6.142973075149954, "learning_rate": 1.9806918143337498e-05, "loss": 1.8418, "step": 296 }, { "epoch": 0.189232239566741, "grad_norm": 10.593259478683024, "learning_rate": 1.980499913687466e-05, "loss": 1.5104, "step": 297 }, { "epoch": 0.1898693851545078, "grad_norm": 3.6317577674094523, "learning_rate": 1.9803070745609705e-05, "loss": 0.9141, "step": 298 }, { "epoch": 0.1905065307422746, "grad_norm": 5.4644475334067995, "learning_rate": 1.9801132971598002e-05, "loss": 1.4099, "step": 299 }, { "epoch": 0.19114367633004142, "grad_norm": 5.870339987602588, "learning_rate": 1.9799185816904935e-05, "loss": 1.4714, "step": 300 }, { "epoch": 0.1917808219178082, "grad_norm": 7.955549091208996, "learning_rate": 1.979722928360587e-05, "loss": 1.6382, "step": 301 }, { "epoch": 0.19241796750557502, "grad_norm": 5.635374790795237, "learning_rate": 1.979526337378618e-05, "loss": 1.7222, "step": 302 }, { "epoch": 0.19305511309334183, "grad_norm": 4.656377085760168, "learning_rate": 1.979328808954122e-05, "loss": 1.3047, "step": 303 }, { "epoch": 0.19369225868110862, "grad_norm": 8.689677375085521, "learning_rate": 1.9791303432976364e-05, "loss": 1.5824, "step": 304 }, { "epoch": 0.19432940426887543, "grad_norm": 9.309206112178432, "learning_rate": 1.9789309406206944e-05, "loss": 1.5188, "step": 305 }, { "epoch": 0.19496654985664225, "grad_norm": 6.417790139698107, "learning_rate": 1.9787306011358287e-05, "loss": 1.532, "step": 306 }, { "epoch": 0.19560369544440906, "grad_norm": 5.120453989566787, "learning_rate": 1.9785293250565718e-05, "loss": 1.2029, "step": 307 }, { "epoch": 0.19624084103217584, "grad_norm": 5.30167072211929, "learning_rate": 1.978327112597454e-05, "loss": 1.3949, "step": 308 }, { "epoch": 0.19687798661994266, "grad_norm": 6.013440184590243, "learning_rate": 1.9781239639740027e-05, "loss": 1.478, "step": 309 }, { "epoch": 0.19751513220770947, "grad_norm": 7.9887829156728225, "learning_rate": 1.9779198794027438e-05, "loss": 1.7664, "step": 310 }, { "epoch": 0.19815227779547626, "grad_norm": 6.180445910028873, "learning_rate": 1.9777148591012003e-05, "loss": 1.8013, "step": 311 }, { "epoch": 0.19878942338324307, "grad_norm": 8.05115332261221, "learning_rate": 1.977508903287894e-05, "loss": 1.1071, "step": 312 }, { "epoch": 0.19942656897100988, "grad_norm": 6.110482373175694, "learning_rate": 1.9773020121823416e-05, "loss": 1.7769, "step": 313 }, { "epoch": 0.20006371455877667, "grad_norm": 5.113718605826186, "learning_rate": 1.977094186005059e-05, "loss": 1.179, "step": 314 }, { "epoch": 0.20070086014654348, "grad_norm": 5.930714961288337, "learning_rate": 1.9768854249775566e-05, "loss": 1.5605, "step": 315 }, { "epoch": 0.2013380057343103, "grad_norm": 25.13765422652374, "learning_rate": 1.9766757293223427e-05, "loss": 1.6272, "step": 316 }, { "epoch": 0.2019751513220771, "grad_norm": 9.653540463478336, "learning_rate": 1.9764650992629207e-05, "loss": 1.8308, "step": 317 }, { "epoch": 0.2026122969098439, "grad_norm": 6.917930259871953, "learning_rate": 1.9762535350237917e-05, "loss": 1.6565, "step": 318 }, { "epoch": 0.2032494424976107, "grad_norm": 6.416558453760042, "learning_rate": 1.97604103683045e-05, "loss": 1.6729, "step": 319 }, { "epoch": 0.20388658808537752, "grad_norm": 9.078858674703769, "learning_rate": 1.9758276049093875e-05, "loss": 1.8381, "step": 320 }, { "epoch": 0.2045237336731443, "grad_norm": 6.756623247773793, "learning_rate": 1.97561323948809e-05, "loss": 1.5848, "step": 321 }, { "epoch": 0.20516087926091112, "grad_norm": 7.237379323630484, "learning_rate": 1.9753979407950387e-05, "loss": 1.3774, "step": 322 }, { "epoch": 0.20579802484867793, "grad_norm": 6.267392029396905, "learning_rate": 1.9751817090597102e-05, "loss": 1.4907, "step": 323 }, { "epoch": 0.20643517043644471, "grad_norm": 12.301828947604308, "learning_rate": 1.974964544512574e-05, "loss": 1.4913, "step": 324 }, { "epoch": 0.20707231602421153, "grad_norm": 146.88277842337945, "learning_rate": 1.9747464473850954e-05, "loss": 1.7483, "step": 325 }, { "epoch": 0.20770946161197834, "grad_norm": 9.544220648256683, "learning_rate": 1.9745274179097332e-05, "loss": 1.3585, "step": 326 }, { "epoch": 0.20834660719974515, "grad_norm": 9.074312688159273, "learning_rate": 1.9743074563199392e-05, "loss": 1.1932, "step": 327 }, { "epoch": 0.20898375278751194, "grad_norm": 10.19243475068208, "learning_rate": 1.9740865628501598e-05, "loss": 1.4702, "step": 328 }, { "epoch": 0.20962089837527875, "grad_norm": 6.02571871907873, "learning_rate": 1.973864737735834e-05, "loss": 1.4519, "step": 329 }, { "epoch": 0.21025804396304557, "grad_norm": 9.285678840298111, "learning_rate": 1.9736419812133938e-05, "loss": 1.8496, "step": 330 }, { "epoch": 0.21089518955081235, "grad_norm": 5.120151127534322, "learning_rate": 1.9734182935202645e-05, "loss": 1.4297, "step": 331 }, { "epoch": 0.21153233513857916, "grad_norm": 6.414485775307414, "learning_rate": 1.9731936748948628e-05, "loss": 1.1678, "step": 332 }, { "epoch": 0.21216948072634598, "grad_norm": 6.798568610213098, "learning_rate": 1.972968125576599e-05, "loss": 1.687, "step": 333 }, { "epoch": 0.21280662631411276, "grad_norm": 4.663862502352313, "learning_rate": 1.9727416458058742e-05, "loss": 1.1145, "step": 334 }, { "epoch": 0.21344377190187958, "grad_norm": 7.2506780125568016, "learning_rate": 1.9725142358240818e-05, "loss": 1.4985, "step": 335 }, { "epoch": 0.2140809174896464, "grad_norm": 5.483874313989314, "learning_rate": 1.972285895873607e-05, "loss": 1.0322, "step": 336 }, { "epoch": 0.2147180630774132, "grad_norm": 8.955556845810928, "learning_rate": 1.972056626197825e-05, "loss": 1.3965, "step": 337 }, { "epoch": 0.21535520866518, "grad_norm": 7.038822341078918, "learning_rate": 1.9718264270411035e-05, "loss": 1.4778, "step": 338 }, { "epoch": 0.2159923542529468, "grad_norm": 6.382928031371734, "learning_rate": 1.9715952986487995e-05, "loss": 1.2041, "step": 339 }, { "epoch": 0.2166294998407136, "grad_norm": 7.8678972691570905, "learning_rate": 1.9713632412672613e-05, "loss": 1.1692, "step": 340 }, { "epoch": 0.2172666454284804, "grad_norm": 6.70656368751133, "learning_rate": 1.971130255143827e-05, "loss": 1.4924, "step": 341 }, { "epoch": 0.2179037910162472, "grad_norm": 7.83328136432225, "learning_rate": 1.9708963405268253e-05, "loss": 1.4849, "step": 342 }, { "epoch": 0.21854093660401402, "grad_norm": 7.149141396279329, "learning_rate": 1.9706614976655728e-05, "loss": 1.7542, "step": 343 }, { "epoch": 0.2191780821917808, "grad_norm": 11.778189121552845, "learning_rate": 1.9704257268103776e-05, "loss": 1.4756, "step": 344 }, { "epoch": 0.21981522777954762, "grad_norm": 4.031627257987123, "learning_rate": 1.9701890282125354e-05, "loss": 1.054, "step": 345 }, { "epoch": 0.22045237336731444, "grad_norm": 5.987707949047307, "learning_rate": 1.969951402124331e-05, "loss": 1.261, "step": 346 }, { "epoch": 0.22108951895508125, "grad_norm": 7.436144778030723, "learning_rate": 1.969712848799039e-05, "loss": 1.6799, "step": 347 }, { "epoch": 0.22172666454284803, "grad_norm": 3.3859207174202104, "learning_rate": 1.96947336849092e-05, "loss": 1.0166, "step": 348 }, { "epoch": 0.22236381013061485, "grad_norm": 6.67351747465036, "learning_rate": 1.9692329614552246e-05, "loss": 1.6997, "step": 349 }, { "epoch": 0.22300095571838166, "grad_norm": 5.57351028965415, "learning_rate": 1.968991627948191e-05, "loss": 1.2667, "step": 350 }, { "epoch": 0.22363810130614845, "grad_norm": 7.293770579539804, "learning_rate": 1.9687493682270434e-05, "loss": 1.6855, "step": 351 }, { "epoch": 0.22427524689391526, "grad_norm": 4.100095144455806, "learning_rate": 1.9685061825499946e-05, "loss": 1.1191, "step": 352 }, { "epoch": 0.22491239248168207, "grad_norm": 9.741767983161065, "learning_rate": 1.9682620711762434e-05, "loss": 1.6831, "step": 353 }, { "epoch": 0.22554953806944886, "grad_norm": 8.80844641298855, "learning_rate": 1.9680170343659766e-05, "loss": 1.1444, "step": 354 }, { "epoch": 0.22618668365721567, "grad_norm": 10.067962685002888, "learning_rate": 1.9677710723803657e-05, "loss": 1.9448, "step": 355 }, { "epoch": 0.22682382924498248, "grad_norm": 7.755320421184453, "learning_rate": 1.9675241854815696e-05, "loss": 1.3875, "step": 356 }, { "epoch": 0.2274609748327493, "grad_norm": 7.74553034059425, "learning_rate": 1.9672763739327326e-05, "loss": 1.3889, "step": 357 }, { "epoch": 0.22809812042051608, "grad_norm": 4.363203130079606, "learning_rate": 1.967027637997984e-05, "loss": 1.2577, "step": 358 }, { "epoch": 0.2287352660082829, "grad_norm": 9.179278895983199, "learning_rate": 1.9667779779424394e-05, "loss": 1.3091, "step": 359 }, { "epoch": 0.2293724115960497, "grad_norm": 4.490632339628285, "learning_rate": 1.966527394032198e-05, "loss": 1.009, "step": 360 }, { "epoch": 0.2300095571838165, "grad_norm": 7.569148493719714, "learning_rate": 1.9662758865343454e-05, "loss": 1.8237, "step": 361 }, { "epoch": 0.2306467027715833, "grad_norm": 15.64079910807429, "learning_rate": 1.96602345571695e-05, "loss": 1.5281, "step": 362 }, { "epoch": 0.23128384835935012, "grad_norm": 8.003506891564358, "learning_rate": 1.965770101849065e-05, "loss": 1.616, "step": 363 }, { "epoch": 0.2319209939471169, "grad_norm": 11.279892491193054, "learning_rate": 1.965515825200728e-05, "loss": 1.4951, "step": 364 }, { "epoch": 0.23255813953488372, "grad_norm": 7.004722899290254, "learning_rate": 1.965260626042959e-05, "loss": 1.2694, "step": 365 }, { "epoch": 0.23319528512265053, "grad_norm": 4.77331070946512, "learning_rate": 1.9650045046477622e-05, "loss": 1.1638, "step": 366 }, { "epoch": 0.23383243071041734, "grad_norm": 5.679788766609498, "learning_rate": 1.9647474612881246e-05, "loss": 1.5017, "step": 367 }, { "epoch": 0.23446957629818413, "grad_norm": 6.204328506076152, "learning_rate": 1.964489496238015e-05, "loss": 1.2729, "step": 368 }, { "epoch": 0.23510672188595094, "grad_norm": 5.70663667558071, "learning_rate": 1.964230609772386e-05, "loss": 1.6365, "step": 369 }, { "epoch": 0.23574386747371776, "grad_norm": 4.604459998537063, "learning_rate": 1.9639708021671714e-05, "loss": 1.1558, "step": 370 }, { "epoch": 0.23638101306148454, "grad_norm": 10.387439342783747, "learning_rate": 1.9637100736992863e-05, "loss": 1.6147, "step": 371 }, { "epoch": 0.23701815864925135, "grad_norm": 6.678202553349704, "learning_rate": 1.963448424646629e-05, "loss": 1.0551, "step": 372 }, { "epoch": 0.23765530423701817, "grad_norm": 9.011228338305935, "learning_rate": 1.963185855288077e-05, "loss": 1.5483, "step": 373 }, { "epoch": 0.23829244982478495, "grad_norm": 6.169986624861777, "learning_rate": 1.96292236590349e-05, "loss": 1.5879, "step": 374 }, { "epoch": 0.23892959541255177, "grad_norm": 6.993326437849051, "learning_rate": 1.962657956773708e-05, "loss": 1.678, "step": 375 }, { "epoch": 0.23956674100031858, "grad_norm": 8.118426208137658, "learning_rate": 1.9623926281805515e-05, "loss": 1.6943, "step": 376 }, { "epoch": 0.2402038865880854, "grad_norm": 3.057908981166614, "learning_rate": 1.9621263804068207e-05, "loss": 0.9873, "step": 377 }, { "epoch": 0.24084103217585218, "grad_norm": 5.086386300565994, "learning_rate": 1.9618592137362954e-05, "loss": 1.3638, "step": 378 }, { "epoch": 0.241478177763619, "grad_norm": 7.039774644257911, "learning_rate": 1.9615911284537354e-05, "loss": 1.5714, "step": 379 }, { "epoch": 0.2421153233513858, "grad_norm": 6.757182183030164, "learning_rate": 1.961322124844878e-05, "loss": 1.5571, "step": 380 }, { "epoch": 0.2427524689391526, "grad_norm": 9.151257557910302, "learning_rate": 1.9610522031964426e-05, "loss": 1.2935, "step": 381 }, { "epoch": 0.2433896145269194, "grad_norm": 5.617954400108297, "learning_rate": 1.960781363796123e-05, "loss": 1.52, "step": 382 }, { "epoch": 0.24402676011468621, "grad_norm": 4.684455932632781, "learning_rate": 1.9605096069325946e-05, "loss": 1.2598, "step": 383 }, { "epoch": 0.244663905702453, "grad_norm": 5.613974067312931, "learning_rate": 1.9602369328955083e-05, "loss": 1.6602, "step": 384 }, { "epoch": 0.2453010512902198, "grad_norm": 8.73978670796969, "learning_rate": 1.959963341975494e-05, "loss": 1.6077, "step": 385 }, { "epoch": 0.24593819687798663, "grad_norm": 4.396907213986035, "learning_rate": 1.9596888344641585e-05, "loss": 1.3541, "step": 386 }, { "epoch": 0.2465753424657534, "grad_norm": 6.433791238021473, "learning_rate": 1.9594134106540847e-05, "loss": 1.5024, "step": 387 }, { "epoch": 0.24721248805352022, "grad_norm": 3.8813612127656847, "learning_rate": 1.9591370708388336e-05, "loss": 0.959, "step": 388 }, { "epoch": 0.24784963364128704, "grad_norm": 5.425296070654179, "learning_rate": 1.958859815312942e-05, "loss": 1.2339, "step": 389 }, { "epoch": 0.24848677922905385, "grad_norm": 6.042129949541454, "learning_rate": 1.958581644371922e-05, "loss": 1.3446, "step": 390 }, { "epoch": 0.24912392481682064, "grad_norm": 5.696506619646398, "learning_rate": 1.9583025583122618e-05, "loss": 1.6599, "step": 391 }, { "epoch": 0.24976107040458745, "grad_norm": 8.418509757767909, "learning_rate": 1.9580225574314255e-05, "loss": 1.7075, "step": 392 }, { "epoch": 0.25039821599235423, "grad_norm": 13.184358725814295, "learning_rate": 1.9577416420278518e-05, "loss": 1.5049, "step": 393 }, { "epoch": 0.2510353615801211, "grad_norm": 10.923867532611279, "learning_rate": 1.957459812400954e-05, "loss": 1.606, "step": 394 }, { "epoch": 0.25167250716788786, "grad_norm": 5.026832030736399, "learning_rate": 1.9571770688511194e-05, "loss": 1.0654, "step": 395 }, { "epoch": 0.25230965275565465, "grad_norm": 8.098245432804974, "learning_rate": 1.956893411679711e-05, "loss": 1.9092, "step": 396 }, { "epoch": 0.2529467983434215, "grad_norm": 6.936832509879017, "learning_rate": 1.9566088411890644e-05, "loss": 1.657, "step": 397 }, { "epoch": 0.25358394393118827, "grad_norm": 6.2016592781178845, "learning_rate": 1.9563233576824886e-05, "loss": 1.2726, "step": 398 }, { "epoch": 0.25422108951895506, "grad_norm": 7.452698469950273, "learning_rate": 1.956036961464266e-05, "loss": 1.2905, "step": 399 }, { "epoch": 0.2548582351067219, "grad_norm": 5.694939778378254, "learning_rate": 1.955749652839652e-05, "loss": 1.5803, "step": 400 }, { "epoch": 0.2554953806944887, "grad_norm": 8.07337565737307, "learning_rate": 1.9554614321148743e-05, "loss": 1.6437, "step": 401 }, { "epoch": 0.25613252628225547, "grad_norm": 10.551905292575496, "learning_rate": 1.9551722995971322e-05, "loss": 1.6514, "step": 402 }, { "epoch": 0.2567696718700223, "grad_norm": 7.775323495063576, "learning_rate": 1.9548822555945984e-05, "loss": 1.4036, "step": 403 }, { "epoch": 0.2574068174577891, "grad_norm": 8.496589838799357, "learning_rate": 1.9545913004164154e-05, "loss": 1.2445, "step": 404 }, { "epoch": 0.25804396304555594, "grad_norm": 8.15817455296583, "learning_rate": 1.9542994343726976e-05, "loss": 1.8914, "step": 405 }, { "epoch": 0.2586811086333227, "grad_norm": 5.255644401403822, "learning_rate": 1.9540066577745303e-05, "loss": 1.2024, "step": 406 }, { "epoch": 0.2593182542210895, "grad_norm": 22.673352939420443, "learning_rate": 1.953712970933969e-05, "loss": 1.364, "step": 407 }, { "epoch": 0.25995539980885635, "grad_norm": 5.38994208982444, "learning_rate": 1.9534183741640402e-05, "loss": 1.4082, "step": 408 }, { "epoch": 0.26059254539662313, "grad_norm": 4.3643973706100825, "learning_rate": 1.9531228677787393e-05, "loss": 1.4453, "step": 409 }, { "epoch": 0.2612296909843899, "grad_norm": 5.664716590816333, "learning_rate": 1.9528264520930317e-05, "loss": 1.5352, "step": 410 }, { "epoch": 0.26186683657215676, "grad_norm": 6.184146305865928, "learning_rate": 1.9525291274228515e-05, "loss": 1.853, "step": 411 }, { "epoch": 0.26250398215992354, "grad_norm": 6.190633126097247, "learning_rate": 1.9522308940851024e-05, "loss": 1.6458, "step": 412 }, { "epoch": 0.26314112774769033, "grad_norm": 2.979564995798331, "learning_rate": 1.951931752397656e-05, "loss": 0.6304, "step": 413 }, { "epoch": 0.26377827333545717, "grad_norm": 3.060835640384856, "learning_rate": 1.951631702679352e-05, "loss": 0.7302, "step": 414 }, { "epoch": 0.26441541892322395, "grad_norm": 6.4391495811695725, "learning_rate": 1.951330745249999e-05, "loss": 1.8398, "step": 415 }, { "epoch": 0.26505256451099074, "grad_norm": 8.422184634371158, "learning_rate": 1.9510288804303715e-05, "loss": 1.5994, "step": 416 }, { "epoch": 0.2656897100987576, "grad_norm": 6.409808654451682, "learning_rate": 1.950726108542212e-05, "loss": 1.5758, "step": 417 }, { "epoch": 0.26632685568652437, "grad_norm": 7.0911894663830335, "learning_rate": 1.95042242990823e-05, "loss": 1.1545, "step": 418 }, { "epoch": 0.26696400127429115, "grad_norm": 14.300017380678293, "learning_rate": 1.950117844852101e-05, "loss": 1.458, "step": 419 }, { "epoch": 0.267601146862058, "grad_norm": 4.651692967797939, "learning_rate": 1.9498123536984667e-05, "loss": 1.2205, "step": 420 }, { "epoch": 0.2682382924498248, "grad_norm": 7.181575027498083, "learning_rate": 1.9495059567729346e-05, "loss": 1.8384, "step": 421 }, { "epoch": 0.26887543803759156, "grad_norm": 8.031266319843127, "learning_rate": 1.949198654402078e-05, "loss": 1.4388, "step": 422 }, { "epoch": 0.2695125836253584, "grad_norm": 24.17901978539513, "learning_rate": 1.948890446913434e-05, "loss": 2.3589, "step": 423 }, { "epoch": 0.2701497292131252, "grad_norm": 7.130972720780271, "learning_rate": 1.9485813346355067e-05, "loss": 1.7173, "step": 424 }, { "epoch": 0.27078687480089203, "grad_norm": 6.744496627151225, "learning_rate": 1.9482713178977623e-05, "loss": 1.5208, "step": 425 }, { "epoch": 0.2714240203886588, "grad_norm": 5.548139251882077, "learning_rate": 1.947960397030632e-05, "loss": 1.2986, "step": 426 }, { "epoch": 0.2720611659764256, "grad_norm": 24.432401945775172, "learning_rate": 1.9476485723655112e-05, "loss": 1.8115, "step": 427 }, { "epoch": 0.27269831156419244, "grad_norm": 4.3802809816459005, "learning_rate": 1.947335844234757e-05, "loss": 1.4414, "step": 428 }, { "epoch": 0.2733354571519592, "grad_norm": 6.955653766081546, "learning_rate": 1.947022212971691e-05, "loss": 1.801, "step": 429 }, { "epoch": 0.273972602739726, "grad_norm": 6.022222109299027, "learning_rate": 1.946707678910597e-05, "loss": 1.5146, "step": 430 }, { "epoch": 0.27460974832749285, "grad_norm": 6.508539886525709, "learning_rate": 1.9463922423867205e-05, "loss": 1.3033, "step": 431 }, { "epoch": 0.27524689391525964, "grad_norm": 6.8412099442609025, "learning_rate": 1.9460759037362694e-05, "loss": 1.4102, "step": 432 }, { "epoch": 0.2758840395030264, "grad_norm": 4.053103686568388, "learning_rate": 1.9457586632964137e-05, "loss": 1.1273, "step": 433 }, { "epoch": 0.27652118509079326, "grad_norm": 11.232569869224521, "learning_rate": 1.945440521405283e-05, "loss": 1.3582, "step": 434 }, { "epoch": 0.27715833067856005, "grad_norm": 3.947263848536659, "learning_rate": 1.945121478401969e-05, "loss": 1.1462, "step": 435 }, { "epoch": 0.27779547626632684, "grad_norm": 11.331551295103289, "learning_rate": 1.944801534626523e-05, "loss": 1.4016, "step": 436 }, { "epoch": 0.2784326218540937, "grad_norm": 3.050009695502577, "learning_rate": 1.944480690419958e-05, "loss": 0.6826, "step": 437 }, { "epoch": 0.27906976744186046, "grad_norm": 8.73119426613881, "learning_rate": 1.9441589461242444e-05, "loss": 1.3074, "step": 438 }, { "epoch": 0.27970691302962725, "grad_norm": 9.11039850040637, "learning_rate": 1.943836302082314e-05, "loss": 1.7163, "step": 439 }, { "epoch": 0.2803440586173941, "grad_norm": 8.354915298480245, "learning_rate": 1.9435127586380557e-05, "loss": 1.4579, "step": 440 }, { "epoch": 0.2809812042051609, "grad_norm": 6.405419580950263, "learning_rate": 1.943188316136319e-05, "loss": 1.2852, "step": 441 }, { "epoch": 0.28161834979292766, "grad_norm": 5.586742346966785, "learning_rate": 1.9428629749229104e-05, "loss": 1.4656, "step": 442 }, { "epoch": 0.2822554953806945, "grad_norm": 5.218246372898348, "learning_rate": 1.9425367353445943e-05, "loss": 0.9009, "step": 443 }, { "epoch": 0.2828926409684613, "grad_norm": 14.90270793635325, "learning_rate": 1.9422095977490934e-05, "loss": 1.6021, "step": 444 }, { "epoch": 0.2835297865562281, "grad_norm": 10.941186805973627, "learning_rate": 1.9418815624850868e-05, "loss": 1.7517, "step": 445 }, { "epoch": 0.2841669321439949, "grad_norm": 8.956320274086089, "learning_rate": 1.9415526299022104e-05, "loss": 1.7039, "step": 446 }, { "epoch": 0.2848040777317617, "grad_norm": 7.116298614260081, "learning_rate": 1.9412228003510576e-05, "loss": 2.0039, "step": 447 }, { "epoch": 0.28544122331952854, "grad_norm": 5.737107609113058, "learning_rate": 1.940892074183176e-05, "loss": 1.4355, "step": 448 }, { "epoch": 0.2860783689072953, "grad_norm": 6.053595019494591, "learning_rate": 1.9405604517510707e-05, "loss": 1.1902, "step": 449 }, { "epoch": 0.2867155144950621, "grad_norm": 3.967608246633206, "learning_rate": 1.940227933408201e-05, "loss": 0.9749, "step": 450 }, { "epoch": 0.28735266008282895, "grad_norm": 5.959061955739262, "learning_rate": 1.9398945195089805e-05, "loss": 1.5303, "step": 451 }, { "epoch": 0.28798980567059573, "grad_norm": 7.335999865933237, "learning_rate": 1.9395602104087793e-05, "loss": 1.7532, "step": 452 }, { "epoch": 0.2886269512583625, "grad_norm": 8.228021168532802, "learning_rate": 1.9392250064639202e-05, "loss": 1.0482, "step": 453 }, { "epoch": 0.28926409684612936, "grad_norm": 5.547830970012715, "learning_rate": 1.9388889080316802e-05, "loss": 1.3717, "step": 454 }, { "epoch": 0.28990124243389614, "grad_norm": 6.664074608081776, "learning_rate": 1.9385519154702888e-05, "loss": 1.7642, "step": 455 }, { "epoch": 0.29053838802166293, "grad_norm": 9.108556126316644, "learning_rate": 1.9382140291389303e-05, "loss": 1.7695, "step": 456 }, { "epoch": 0.29117553360942977, "grad_norm": 6.492356867748741, "learning_rate": 1.9378752493977397e-05, "loss": 1.9717, "step": 457 }, { "epoch": 0.29181267919719656, "grad_norm": 5.208605401377055, "learning_rate": 1.9375355766078056e-05, "loss": 1.2622, "step": 458 }, { "epoch": 0.29244982478496334, "grad_norm": 3.8331589784615865, "learning_rate": 1.9371950111311683e-05, "loss": 1.1775, "step": 459 }, { "epoch": 0.2930869703727302, "grad_norm": 6.239477506145884, "learning_rate": 1.9368535533308188e-05, "loss": 1.573, "step": 460 }, { "epoch": 0.29372411596049697, "grad_norm": 4.673228847308328, "learning_rate": 1.9365112035707e-05, "loss": 1.4166, "step": 461 }, { "epoch": 0.29436126154826375, "grad_norm": 6.667766287810246, "learning_rate": 1.936167962215705e-05, "loss": 1.056, "step": 462 }, { "epoch": 0.2949984071360306, "grad_norm": 4.573840744891376, "learning_rate": 1.935823829631677e-05, "loss": 1.328, "step": 463 }, { "epoch": 0.2956355527237974, "grad_norm": 4.657821498088626, "learning_rate": 1.93547880618541e-05, "loss": 0.87, "step": 464 }, { "epoch": 0.2962726983115642, "grad_norm": 6.90026756803054, "learning_rate": 1.935132892244647e-05, "loss": 1.6267, "step": 465 }, { "epoch": 0.296909843899331, "grad_norm": 5.673269549854018, "learning_rate": 1.9347860881780796e-05, "loss": 1.0804, "step": 466 }, { "epoch": 0.2975469894870978, "grad_norm": 7.636887370363844, "learning_rate": 1.934438394355349e-05, "loss": 1.6233, "step": 467 }, { "epoch": 0.29818413507486463, "grad_norm": 4.624257841688103, "learning_rate": 1.9340898111470448e-05, "loss": 0.9662, "step": 468 }, { "epoch": 0.2988212806626314, "grad_norm": 7.1223163032052925, "learning_rate": 1.9337403389247032e-05, "loss": 1.6499, "step": 469 }, { "epoch": 0.2994584262503982, "grad_norm": 7.312405139107185, "learning_rate": 1.9333899780608097e-05, "loss": 1.7617, "step": 470 }, { "epoch": 0.30009557183816504, "grad_norm": 4.943541557872431, "learning_rate": 1.9330387289287964e-05, "loss": 1.2427, "step": 471 }, { "epoch": 0.30073271742593183, "grad_norm": 7.4934135572854, "learning_rate": 1.932686591903041e-05, "loss": 1.5895, "step": 472 }, { "epoch": 0.3013698630136986, "grad_norm": 4.401658514829763, "learning_rate": 1.93233356735887e-05, "loss": 1.0177, "step": 473 }, { "epoch": 0.30200700860146545, "grad_norm": 6.857371529934064, "learning_rate": 1.931979655672552e-05, "loss": 1.2346, "step": 474 }, { "epoch": 0.30264415418923224, "grad_norm": 3.3794524498954814, "learning_rate": 1.9316248572213058e-05, "loss": 0.8444, "step": 475 }, { "epoch": 0.303281299776999, "grad_norm": 7.17620538700122, "learning_rate": 1.9312691723832917e-05, "loss": 1.9585, "step": 476 }, { "epoch": 0.30391844536476587, "grad_norm": 7.7528762665108895, "learning_rate": 1.930912601537617e-05, "loss": 1.405, "step": 477 }, { "epoch": 0.30455559095253265, "grad_norm": 8.488772629045242, "learning_rate": 1.930555145064332e-05, "loss": 1.7275, "step": 478 }, { "epoch": 0.30519273654029944, "grad_norm": 7.6192949420749345, "learning_rate": 1.9301968033444315e-05, "loss": 1.3848, "step": 479 }, { "epoch": 0.3058298821280663, "grad_norm": 7.112130904429123, "learning_rate": 1.9298375767598536e-05, "loss": 1.5381, "step": 480 }, { "epoch": 0.30646702771583306, "grad_norm": 4.451440871023836, "learning_rate": 1.92947746569348e-05, "loss": 1.0142, "step": 481 }, { "epoch": 0.30710417330359985, "grad_norm": 9.863891607331137, "learning_rate": 1.929116470529135e-05, "loss": 1.5298, "step": 482 }, { "epoch": 0.3077413188913667, "grad_norm": 4.7059183359376835, "learning_rate": 1.9287545916515847e-05, "loss": 1.2681, "step": 483 }, { "epoch": 0.3083784644791335, "grad_norm": 42.099401194936036, "learning_rate": 1.9283918294465372e-05, "loss": 1.2211, "step": 484 }, { "epoch": 0.30901561006690026, "grad_norm": 6.105635437690791, "learning_rate": 1.928028184300643e-05, "loss": 1.1956, "step": 485 }, { "epoch": 0.3096527556546671, "grad_norm": 4.313144917106101, "learning_rate": 1.9276636566014923e-05, "loss": 1.0017, "step": 486 }, { "epoch": 0.3102899012424339, "grad_norm": 6.522854079433015, "learning_rate": 1.9272982467376172e-05, "loss": 1.6064, "step": 487 }, { "epoch": 0.3109270468302007, "grad_norm": 8.774360756236613, "learning_rate": 1.9269319550984897e-05, "loss": 1.0757, "step": 488 }, { "epoch": 0.3115641924179675, "grad_norm": 5.827989312708804, "learning_rate": 1.926564782074521e-05, "loss": 1.2583, "step": 489 }, { "epoch": 0.3122013380057343, "grad_norm": 10.307699201243036, "learning_rate": 1.9261967280570627e-05, "loss": 1.2644, "step": 490 }, { "epoch": 0.31283848359350114, "grad_norm": 6.6022328231279666, "learning_rate": 1.9258277934384045e-05, "loss": 1.6509, "step": 491 }, { "epoch": 0.3134756291812679, "grad_norm": 7.349728262644769, "learning_rate": 1.9254579786117752e-05, "loss": 1.415, "step": 492 }, { "epoch": 0.3141127747690347, "grad_norm": 4.924969146901724, "learning_rate": 1.9250872839713416e-05, "loss": 1.3442, "step": 493 }, { "epoch": 0.31474992035680155, "grad_norm": 7.9513642010977845, "learning_rate": 1.924715709912208e-05, "loss": 1.4739, "step": 494 }, { "epoch": 0.31538706594456833, "grad_norm": 7.17566230652925, "learning_rate": 1.9243432568304176e-05, "loss": 1.5403, "step": 495 }, { "epoch": 0.3160242115323351, "grad_norm": 5.756897023441995, "learning_rate": 1.9239699251229476e-05, "loss": 1.4321, "step": 496 }, { "epoch": 0.31666135712010196, "grad_norm": 8.398277083539178, "learning_rate": 1.9235957151877147e-05, "loss": 1.6772, "step": 497 }, { "epoch": 0.31729850270786875, "grad_norm": 6.264215893366967, "learning_rate": 1.9232206274235697e-05, "loss": 1.5132, "step": 498 }, { "epoch": 0.31793564829563553, "grad_norm": 6.292092537065723, "learning_rate": 1.9228446622302997e-05, "loss": 1.8403, "step": 499 }, { "epoch": 0.31857279388340237, "grad_norm": 5.1121350763407225, "learning_rate": 1.9224678200086267e-05, "loss": 1.1786, "step": 500 }, { "epoch": 0.31920993947116916, "grad_norm": 9.178032499931707, "learning_rate": 1.922090101160208e-05, "loss": 1.5228, "step": 501 }, { "epoch": 0.31984708505893594, "grad_norm": 7.573150455362142, "learning_rate": 1.921711506087635e-05, "loss": 1.2692, "step": 502 }, { "epoch": 0.3204842306467028, "grad_norm": 12.757575326804039, "learning_rate": 1.921332035194433e-05, "loss": 0.9896, "step": 503 }, { "epoch": 0.32112137623446957, "grad_norm": 7.724211146587327, "learning_rate": 1.9209516888850605e-05, "loss": 1.6084, "step": 504 }, { "epoch": 0.32175852182223635, "grad_norm": 4.926779004642543, "learning_rate": 1.9205704675649094e-05, "loss": 0.8942, "step": 505 }, { "epoch": 0.3223956674100032, "grad_norm": 6.925597320833541, "learning_rate": 1.920188371640305e-05, "loss": 1.6863, "step": 506 }, { "epoch": 0.32303281299777, "grad_norm": 4.940307367100383, "learning_rate": 1.9198054015185023e-05, "loss": 1.4202, "step": 507 }, { "epoch": 0.3236699585855368, "grad_norm": 4.841886987063608, "learning_rate": 1.9194215576076917e-05, "loss": 1.1821, "step": 508 }, { "epoch": 0.3243071041733036, "grad_norm": 24.313527891856488, "learning_rate": 1.919036840316991e-05, "loss": 1.5969, "step": 509 }, { "epoch": 0.3249442497610704, "grad_norm": 3.6198251900878833, "learning_rate": 1.918651250056452e-05, "loss": 0.6881, "step": 510 }, { "epoch": 0.32558139534883723, "grad_norm": 8.828394376169639, "learning_rate": 1.9182647872370562e-05, "loss": 1.9697, "step": 511 }, { "epoch": 0.326218540936604, "grad_norm": 8.21302607672626, "learning_rate": 1.9178774522707138e-05, "loss": 1.9307, "step": 512 }, { "epoch": 0.3268556865243708, "grad_norm": 7.1138762414505665, "learning_rate": 1.917489245570266e-05, "loss": 1.4691, "step": 513 }, { "epoch": 0.32749283211213764, "grad_norm": 7.563207289898115, "learning_rate": 1.9171001675494826e-05, "loss": 1.8335, "step": 514 }, { "epoch": 0.32812997769990443, "grad_norm": 4.746425087419655, "learning_rate": 1.9167102186230616e-05, "loss": 1.2227, "step": 515 }, { "epoch": 0.3287671232876712, "grad_norm": 39.86213369274787, "learning_rate": 1.9163193992066303e-05, "loss": 1.4958, "step": 516 }, { "epoch": 0.32940426887543806, "grad_norm": 9.319373291746757, "learning_rate": 1.9159277097167434e-05, "loss": 1.7163, "step": 517 }, { "epoch": 0.33004141446320484, "grad_norm": 6.320311991839263, "learning_rate": 1.915535150570883e-05, "loss": 1.5364, "step": 518 }, { "epoch": 0.3306785600509716, "grad_norm": 5.800961485424098, "learning_rate": 1.9151417221874572e-05, "loss": 0.8535, "step": 519 }, { "epoch": 0.33131570563873847, "grad_norm": 4.946906375782722, "learning_rate": 1.9147474249858017e-05, "loss": 1.0492, "step": 520 }, { "epoch": 0.33195285122650525, "grad_norm": 10.365988433048367, "learning_rate": 1.9143522593861783e-05, "loss": 1.3884, "step": 521 }, { "epoch": 0.33258999681427204, "grad_norm": 6.441315414443533, "learning_rate": 1.9139562258097737e-05, "loss": 1.5549, "step": 522 }, { "epoch": 0.3332271424020389, "grad_norm": 7.429719204077702, "learning_rate": 1.9135593246786997e-05, "loss": 1.2518, "step": 523 }, { "epoch": 0.33386428798980566, "grad_norm": 7.900032038849596, "learning_rate": 1.9131615564159938e-05, "loss": 1.5083, "step": 524 }, { "epoch": 0.33450143357757245, "grad_norm": 3.836877037324927, "learning_rate": 1.9127629214456164e-05, "loss": 0.9795, "step": 525 }, { "epoch": 0.3351385791653393, "grad_norm": 4.942923376802814, "learning_rate": 1.9123634201924527e-05, "loss": 1.2783, "step": 526 }, { "epoch": 0.3357757247531061, "grad_norm": 5.683739149378437, "learning_rate": 1.9119630530823108e-05, "loss": 1.1853, "step": 527 }, { "epoch": 0.3364128703408729, "grad_norm": 5.812140358026368, "learning_rate": 1.911561820541921e-05, "loss": 1.333, "step": 528 }, { "epoch": 0.3370500159286397, "grad_norm": 5.071543088579532, "learning_rate": 1.9111597229989378e-05, "loss": 1.3726, "step": 529 }, { "epoch": 0.3376871615164065, "grad_norm": 4.561309020637667, "learning_rate": 1.910756760881936e-05, "loss": 1.2571, "step": 530 }, { "epoch": 0.3383243071041733, "grad_norm": 8.759423331637308, "learning_rate": 1.910352934620412e-05, "loss": 1.8774, "step": 531 }, { "epoch": 0.3389614526919401, "grad_norm": 6.118997282885562, "learning_rate": 1.909948244644785e-05, "loss": 1.6487, "step": 532 }, { "epoch": 0.3395985982797069, "grad_norm": 7.142413614956808, "learning_rate": 1.9095426913863923e-05, "loss": 1.8328, "step": 533 }, { "epoch": 0.34023574386747374, "grad_norm": 7.737775983731822, "learning_rate": 1.9091362752774935e-05, "loss": 1.4905, "step": 534 }, { "epoch": 0.3408728894552405, "grad_norm": 5.8149056101928185, "learning_rate": 1.9087289967512663e-05, "loss": 1.5991, "step": 535 }, { "epoch": 0.3415100350430073, "grad_norm": 8.717947489665539, "learning_rate": 1.9083208562418082e-05, "loss": 1.4661, "step": 536 }, { "epoch": 0.34214718063077415, "grad_norm": 4.881896359511361, "learning_rate": 1.9079118541841356e-05, "loss": 1.0228, "step": 537 }, { "epoch": 0.34278432621854094, "grad_norm": 3.863322291121575, "learning_rate": 1.907501991014184e-05, "loss": 0.697, "step": 538 }, { "epoch": 0.3434214718063077, "grad_norm": 5.990762881999785, "learning_rate": 1.907091267168804e-05, "loss": 0.8732, "step": 539 }, { "epoch": 0.34405861739407456, "grad_norm": 6.257796336846997, "learning_rate": 1.9066796830857665e-05, "loss": 1.5068, "step": 540 }, { "epoch": 0.34469576298184135, "grad_norm": 8.321550440371396, "learning_rate": 1.906267239203758e-05, "loss": 1.4341, "step": 541 }, { "epoch": 0.34533290856960813, "grad_norm": 7.576130812038169, "learning_rate": 1.9058539359623813e-05, "loss": 1.2504, "step": 542 }, { "epoch": 0.345970054157375, "grad_norm": 8.594940897222155, "learning_rate": 1.9054397738021552e-05, "loss": 0.6022, "step": 543 }, { "epoch": 0.34660719974514176, "grad_norm": 4.384434126060305, "learning_rate": 1.9050247531645147e-05, "loss": 0.9529, "step": 544 }, { "epoch": 0.34724434533290854, "grad_norm": 7.283373826700764, "learning_rate": 1.904608874491809e-05, "loss": 1.6819, "step": 545 }, { "epoch": 0.3478814909206754, "grad_norm": 7.73011366917788, "learning_rate": 1.9041921382273015e-05, "loss": 1.8271, "step": 546 }, { "epoch": 0.34851863650844217, "grad_norm": 16.287984406699394, "learning_rate": 1.903774544815171e-05, "loss": 1.4885, "step": 547 }, { "epoch": 0.349155782096209, "grad_norm": 4.107702215968391, "learning_rate": 1.9033560947005092e-05, "loss": 1.1228, "step": 548 }, { "epoch": 0.3497929276839758, "grad_norm": 4.114494175074306, "learning_rate": 1.9029367883293205e-05, "loss": 0.912, "step": 549 }, { "epoch": 0.3504300732717426, "grad_norm": 7.0577330605524216, "learning_rate": 1.902516626148522e-05, "loss": 1.7869, "step": 550 }, { "epoch": 0.3510672188595094, "grad_norm": 7.954762516882862, "learning_rate": 1.9020956086059438e-05, "loss": 1.6406, "step": 551 }, { "epoch": 0.3517043644472762, "grad_norm": 5.784268695952804, "learning_rate": 1.9016737361503272e-05, "loss": 1.5233, "step": 552 }, { "epoch": 0.352341510035043, "grad_norm": 10.095472159104535, "learning_rate": 1.9012510092313244e-05, "loss": 0.9026, "step": 553 }, { "epoch": 0.35297865562280983, "grad_norm": 8.130188353977426, "learning_rate": 1.9008274282994986e-05, "loss": 1.3577, "step": 554 }, { "epoch": 0.3536158012105766, "grad_norm": 6.968693859932392, "learning_rate": 1.900402993806323e-05, "loss": 1.8757, "step": 555 }, { "epoch": 0.3542529467983434, "grad_norm": 9.11550035639889, "learning_rate": 1.8999777062041817e-05, "loss": 1.707, "step": 556 }, { "epoch": 0.35489009238611025, "grad_norm": 5.132962246661145, "learning_rate": 1.899551565946366e-05, "loss": 0.8754, "step": 557 }, { "epoch": 0.35552723797387703, "grad_norm": 7.564109530939992, "learning_rate": 1.8991245734870777e-05, "loss": 1.7949, "step": 558 }, { "epoch": 0.3561643835616438, "grad_norm": 7.091779027429183, "learning_rate": 1.8986967292814268e-05, "loss": 1.9185, "step": 559 }, { "epoch": 0.35680152914941066, "grad_norm": 8.312728991493048, "learning_rate": 1.8982680337854306e-05, "loss": 1.3672, "step": 560 }, { "epoch": 0.35743867473717744, "grad_norm": 6.759142727621221, "learning_rate": 1.8978384874560133e-05, "loss": 1.1958, "step": 561 }, { "epoch": 0.3580758203249442, "grad_norm": 11.505266644993956, "learning_rate": 1.8974080907510074e-05, "loss": 1.7075, "step": 562 }, { "epoch": 0.35871296591271107, "grad_norm": 5.861312706412572, "learning_rate": 1.89697684412915e-05, "loss": 1.6252, "step": 563 }, { "epoch": 0.35935011150047785, "grad_norm": 6.946365057861703, "learning_rate": 1.8965447480500857e-05, "loss": 1.562, "step": 564 }, { "epoch": 0.35998725708824464, "grad_norm": 9.998592693316477, "learning_rate": 1.8961118029743636e-05, "loss": 1.2676, "step": 565 }, { "epoch": 0.3606244026760115, "grad_norm": 10.435876911182717, "learning_rate": 1.895678009363438e-05, "loss": 1.3425, "step": 566 }, { "epoch": 0.36126154826377826, "grad_norm": 4.867394797298838, "learning_rate": 1.8952433676796674e-05, "loss": 1.2014, "step": 567 }, { "epoch": 0.3618986938515451, "grad_norm": 5.758118281691437, "learning_rate": 1.8948078783863143e-05, "loss": 1.0498, "step": 568 }, { "epoch": 0.3625358394393119, "grad_norm": 7.830691597130755, "learning_rate": 1.8943715419475443e-05, "loss": 1.4474, "step": 569 }, { "epoch": 0.3631729850270787, "grad_norm": 9.452675892254563, "learning_rate": 1.8939343588284275e-05, "loss": 1.6587, "step": 570 }, { "epoch": 0.3638101306148455, "grad_norm": 8.125676108696972, "learning_rate": 1.893496329494934e-05, "loss": 1.0813, "step": 571 }, { "epoch": 0.3644472762026123, "grad_norm": 8.505967746061815, "learning_rate": 1.8930574544139373e-05, "loss": 1.5625, "step": 572 }, { "epoch": 0.3650844217903791, "grad_norm": 7.501863731103875, "learning_rate": 1.8926177340532123e-05, "loss": 0.9609, "step": 573 }, { "epoch": 0.36572156737814593, "grad_norm": 8.241117929620719, "learning_rate": 1.8921771688814348e-05, "loss": 1.3475, "step": 574 }, { "epoch": 0.3663587129659127, "grad_norm": 6.594360744337388, "learning_rate": 1.8917357593681807e-05, "loss": 1.3994, "step": 575 }, { "epoch": 0.3669958585536795, "grad_norm": 7.658301699138836, "learning_rate": 1.8912935059839258e-05, "loss": 1.5317, "step": 576 }, { "epoch": 0.36763300414144634, "grad_norm": 7.363872255284409, "learning_rate": 1.8908504092000458e-05, "loss": 1.1602, "step": 577 }, { "epoch": 0.3682701497292131, "grad_norm": 12.623832248191235, "learning_rate": 1.890406469488815e-05, "loss": 1.3644, "step": 578 }, { "epoch": 0.3689072953169799, "grad_norm": 7.473696035419751, "learning_rate": 1.8899616873234067e-05, "loss": 1.4988, "step": 579 }, { "epoch": 0.36954444090474675, "grad_norm": 9.683897279120899, "learning_rate": 1.8895160631778906e-05, "loss": 1.5051, "step": 580 }, { "epoch": 0.37018158649251354, "grad_norm": 5.519418305591046, "learning_rate": 1.8890695975272358e-05, "loss": 0.9052, "step": 581 }, { "epoch": 0.3708187320802803, "grad_norm": 6.537903397853653, "learning_rate": 1.8886222908473067e-05, "loss": 1.4604, "step": 582 }, { "epoch": 0.37145587766804716, "grad_norm": 8.393118525905619, "learning_rate": 1.888174143614866e-05, "loss": 0.9928, "step": 583 }, { "epoch": 0.37209302325581395, "grad_norm": 10.196077108475311, "learning_rate": 1.8877251563075698e-05, "loss": 1.915, "step": 584 }, { "epoch": 0.37273016884358073, "grad_norm": 5.565324025442431, "learning_rate": 1.8872753294039717e-05, "loss": 1.056, "step": 585 }, { "epoch": 0.3733673144313476, "grad_norm": 18.87368274849809, "learning_rate": 1.886824663383519e-05, "loss": 1.094, "step": 586 }, { "epoch": 0.37400446001911436, "grad_norm": 6.648396752771901, "learning_rate": 1.886373158726554e-05, "loss": 1.188, "step": 587 }, { "epoch": 0.37464160560688115, "grad_norm": 6.01053297384486, "learning_rate": 1.8859208159143124e-05, "loss": 1.3564, "step": 588 }, { "epoch": 0.375278751194648, "grad_norm": 6.899139308471648, "learning_rate": 1.8854676354289235e-05, "loss": 1.3191, "step": 589 }, { "epoch": 0.37591589678241477, "grad_norm": 7.555411585438631, "learning_rate": 1.8850136177534094e-05, "loss": 1.1593, "step": 590 }, { "epoch": 0.3765530423701816, "grad_norm": 8.284703091997324, "learning_rate": 1.8845587633716846e-05, "loss": 1.0007, "step": 591 }, { "epoch": 0.3771901879579484, "grad_norm": 14.91903194118992, "learning_rate": 1.8841030727685554e-05, "loss": 1.7371, "step": 592 }, { "epoch": 0.3778273335457152, "grad_norm": 6.391418640239891, "learning_rate": 1.8836465464297196e-05, "loss": 1.363, "step": 593 }, { "epoch": 0.378464479133482, "grad_norm": 6.739849309240288, "learning_rate": 1.8831891848417643e-05, "loss": 1.269, "step": 594 }, { "epoch": 0.3791016247212488, "grad_norm": 8.688137082620784, "learning_rate": 1.8827309884921696e-05, "loss": 1.6472, "step": 595 }, { "epoch": 0.3797387703090156, "grad_norm": 7.425601800213068, "learning_rate": 1.8822719578693025e-05, "loss": 1.4353, "step": 596 }, { "epoch": 0.38037591589678243, "grad_norm": 14.389533505413876, "learning_rate": 1.8818120934624214e-05, "loss": 1.5251, "step": 597 }, { "epoch": 0.3810130614845492, "grad_norm": 13.095574813373196, "learning_rate": 1.8813513957616715e-05, "loss": 1.7744, "step": 598 }, { "epoch": 0.381650207072316, "grad_norm": 8.715611911617826, "learning_rate": 1.880889865258088e-05, "loss": 1.7146, "step": 599 }, { "epoch": 0.38228735266008285, "grad_norm": 8.805754962484528, "learning_rate": 1.8804275024435922e-05, "loss": 1.1886, "step": 600 }, { "epoch": 0.38292449824784963, "grad_norm": 6.978777270024981, "learning_rate": 1.8799643078109936e-05, "loss": 1.3433, "step": 601 }, { "epoch": 0.3835616438356164, "grad_norm": 4.655275305328869, "learning_rate": 1.8795002818539875e-05, "loss": 0.9392, "step": 602 }, { "epoch": 0.38419878942338326, "grad_norm": 11.363343662300705, "learning_rate": 1.879035425067156e-05, "loss": 1.8555, "step": 603 }, { "epoch": 0.38483593501115004, "grad_norm": 11.65577702350492, "learning_rate": 1.8785697379459665e-05, "loss": 1.5278, "step": 604 }, { "epoch": 0.38547308059891683, "grad_norm": 5.824443545108649, "learning_rate": 1.8781032209867706e-05, "loss": 1.0482, "step": 605 }, { "epoch": 0.38611022618668367, "grad_norm": 7.049895868690771, "learning_rate": 1.8776358746868056e-05, "loss": 1.4268, "step": 606 }, { "epoch": 0.38674737177445045, "grad_norm": 8.94153033857395, "learning_rate": 1.8771676995441925e-05, "loss": 1.4316, "step": 607 }, { "epoch": 0.38738451736221724, "grad_norm": 10.449608036420356, "learning_rate": 1.876698696057935e-05, "loss": 1.77, "step": 608 }, { "epoch": 0.3880216629499841, "grad_norm": 5.117739313116863, "learning_rate": 1.87622886472792e-05, "loss": 0.8916, "step": 609 }, { "epoch": 0.38865880853775087, "grad_norm": 13.902041560937397, "learning_rate": 1.875758206054917e-05, "loss": 1.7883, "step": 610 }, { "epoch": 0.3892959541255177, "grad_norm": 8.719318091935039, "learning_rate": 1.8752867205405775e-05, "loss": 1.384, "step": 611 }, { "epoch": 0.3899330997132845, "grad_norm": 12.080962286510273, "learning_rate": 1.874814408687434e-05, "loss": 1.1958, "step": 612 }, { "epoch": 0.3905702453010513, "grad_norm": 6.755080938884501, "learning_rate": 1.8743412709988993e-05, "loss": 1.4204, "step": 613 }, { "epoch": 0.3912073908888181, "grad_norm": 16.82191077270928, "learning_rate": 1.8738673079792672e-05, "loss": 1.3416, "step": 614 }, { "epoch": 0.3918445364765849, "grad_norm": 8.9820657194822, "learning_rate": 1.8733925201337108e-05, "loss": 1.2946, "step": 615 }, { "epoch": 0.3924816820643517, "grad_norm": 5.881794506054397, "learning_rate": 1.8729169079682827e-05, "loss": 0.8982, "step": 616 }, { "epoch": 0.39311882765211853, "grad_norm": 4.793888089499336, "learning_rate": 1.872440471989913e-05, "loss": 0.7941, "step": 617 }, { "epoch": 0.3937559732398853, "grad_norm": 12.243703753042137, "learning_rate": 1.8719632127064114e-05, "loss": 1.1608, "step": 618 }, { "epoch": 0.3943931188276521, "grad_norm": 9.215665374243448, "learning_rate": 1.8714851306264642e-05, "loss": 0.869, "step": 619 }, { "epoch": 0.39503026441541894, "grad_norm": 6.260441743031529, "learning_rate": 1.8710062262596346e-05, "loss": 0.8655, "step": 620 }, { "epoch": 0.3956674100031857, "grad_norm": 7.69929072831766, "learning_rate": 1.870526500116363e-05, "loss": 1.7656, "step": 621 }, { "epoch": 0.3963045555909525, "grad_norm": 4.165460907447116, "learning_rate": 1.870045952707964e-05, "loss": 0.8022, "step": 622 }, { "epoch": 0.39694170117871935, "grad_norm": 8.78236466652637, "learning_rate": 1.8695645845466303e-05, "loss": 1.3505, "step": 623 }, { "epoch": 0.39757884676648614, "grad_norm": 11.957463998053585, "learning_rate": 1.8690823961454264e-05, "loss": 1.269, "step": 624 }, { "epoch": 0.3982159923542529, "grad_norm": 6.837488950876851, "learning_rate": 1.8685993880182935e-05, "loss": 1.3892, "step": 625 }, { "epoch": 0.39885313794201976, "grad_norm": 11.396287135602318, "learning_rate": 1.8681155606800448e-05, "loss": 1.7893, "step": 626 }, { "epoch": 0.39949028352978655, "grad_norm": 9.179151097301714, "learning_rate": 1.867630914646368e-05, "loss": 1.3608, "step": 627 }, { "epoch": 0.40012742911755333, "grad_norm": 7.756426805680213, "learning_rate": 1.8671454504338223e-05, "loss": 1.1926, "step": 628 }, { "epoch": 0.4007645747053202, "grad_norm": 8.881951811403653, "learning_rate": 1.866659168559839e-05, "loss": 1.0845, "step": 629 }, { "epoch": 0.40140172029308696, "grad_norm": 8.211751979880958, "learning_rate": 1.866172069542722e-05, "loss": 1.213, "step": 630 }, { "epoch": 0.4020388658808538, "grad_norm": 3.5044210404924763, "learning_rate": 1.865684153901645e-05, "loss": 0.7021, "step": 631 }, { "epoch": 0.4026760114686206, "grad_norm": 7.260486515771138, "learning_rate": 1.8651954221566527e-05, "loss": 1.2253, "step": 632 }, { "epoch": 0.4033131570563874, "grad_norm": 6.4770926386491094, "learning_rate": 1.8647058748286597e-05, "loss": 1.2839, "step": 633 }, { "epoch": 0.4039503026441542, "grad_norm": 6.94790003126207, "learning_rate": 1.8642155124394496e-05, "loss": 0.9984, "step": 634 }, { "epoch": 0.404587448231921, "grad_norm": 7.547608456223474, "learning_rate": 1.8637243355116743e-05, "loss": 1.3373, "step": 635 }, { "epoch": 0.4052245938196878, "grad_norm": 5.679484506547999, "learning_rate": 1.8632323445688548e-05, "loss": 1.0269, "step": 636 }, { "epoch": 0.4058617394074546, "grad_norm": 10.069281106747521, "learning_rate": 1.8627395401353792e-05, "loss": 1.0973, "step": 637 }, { "epoch": 0.4064988849952214, "grad_norm": 8.68097294808428, "learning_rate": 1.862245922736503e-05, "loss": 1.8403, "step": 638 }, { "epoch": 0.4071360305829882, "grad_norm": 4.392894885520022, "learning_rate": 1.8617514928983477e-05, "loss": 0.6149, "step": 639 }, { "epoch": 0.40777317617075504, "grad_norm": 8.669955960721822, "learning_rate": 1.8612562511479014e-05, "loss": 1.6311, "step": 640 }, { "epoch": 0.4084103217585218, "grad_norm": 8.20870312518082, "learning_rate": 1.8607601980130162e-05, "loss": 1.4972, "step": 641 }, { "epoch": 0.4090474673462886, "grad_norm": 11.767704956660124, "learning_rate": 1.860263334022411e-05, "loss": 1.5321, "step": 642 }, { "epoch": 0.40968461293405545, "grad_norm": 11.186432169962384, "learning_rate": 1.8597656597056678e-05, "loss": 1.021, "step": 643 }, { "epoch": 0.41032175852182223, "grad_norm": 7.823692600625712, "learning_rate": 1.8592671755932314e-05, "loss": 1.115, "step": 644 }, { "epoch": 0.410958904109589, "grad_norm": 15.368556400056772, "learning_rate": 1.8587678822164122e-05, "loss": 1.4385, "step": 645 }, { "epoch": 0.41159604969735586, "grad_norm": 7.432362872134659, "learning_rate": 1.85826778010738e-05, "loss": 1.0833, "step": 646 }, { "epoch": 0.41223319528512264, "grad_norm": 10.148062013308977, "learning_rate": 1.8577668697991706e-05, "loss": 1.7778, "step": 647 }, { "epoch": 0.41287034087288943, "grad_norm": 6.279421487889011, "learning_rate": 1.857265151825677e-05, "loss": 1.0151, "step": 648 }, { "epoch": 0.41350748646065627, "grad_norm": 21.242641548681785, "learning_rate": 1.8567626267216553e-05, "loss": 1.3311, "step": 649 }, { "epoch": 0.41414463204842306, "grad_norm": 12.260469986428514, "learning_rate": 1.856259295022722e-05, "loss": 1.5498, "step": 650 }, { "epoch": 0.4147817776361899, "grad_norm": 8.333458095298315, "learning_rate": 1.8557551572653523e-05, "loss": 1.38, "step": 651 }, { "epoch": 0.4154189232239567, "grad_norm": 12.412994038428938, "learning_rate": 1.855250213986882e-05, "loss": 1.5632, "step": 652 }, { "epoch": 0.41605606881172347, "grad_norm": 10.319866931253134, "learning_rate": 1.8547444657255034e-05, "loss": 1.4127, "step": 653 }, { "epoch": 0.4166932143994903, "grad_norm": 10.838793809132284, "learning_rate": 1.8542379130202682e-05, "loss": 1.5046, "step": 654 }, { "epoch": 0.4173303599872571, "grad_norm": 7.17564148728505, "learning_rate": 1.853730556411086e-05, "loss": 1.2603, "step": 655 }, { "epoch": 0.4179675055750239, "grad_norm": 13.466319358965638, "learning_rate": 1.853222396438722e-05, "loss": 1.243, "step": 656 }, { "epoch": 0.4186046511627907, "grad_norm": 10.347673754292645, "learning_rate": 1.852713433644798e-05, "loss": 1.4386, "step": 657 }, { "epoch": 0.4192417967505575, "grad_norm": 10.35936945895772, "learning_rate": 1.8522036685717916e-05, "loss": 1.4709, "step": 658 }, { "epoch": 0.4198789423383243, "grad_norm": 8.717979825270646, "learning_rate": 1.8516931017630354e-05, "loss": 1.0054, "step": 659 }, { "epoch": 0.42051608792609113, "grad_norm": 10.924937342315928, "learning_rate": 1.851181733762717e-05, "loss": 1.4076, "step": 660 }, { "epoch": 0.4211532335138579, "grad_norm": 16.00308218979736, "learning_rate": 1.850669565115878e-05, "loss": 1.532, "step": 661 }, { "epoch": 0.4217903791016247, "grad_norm": 6.32529771817146, "learning_rate": 1.8501565963684123e-05, "loss": 1.0786, "step": 662 }, { "epoch": 0.42242752468939154, "grad_norm": 13.024406410677235, "learning_rate": 1.8496428280670677e-05, "loss": 1.4282, "step": 663 }, { "epoch": 0.42306467027715833, "grad_norm": 8.594846941715888, "learning_rate": 1.8491282607594434e-05, "loss": 1.351, "step": 664 }, { "epoch": 0.4237018158649251, "grad_norm": 8.493779686151722, "learning_rate": 1.848612894993991e-05, "loss": 1.2251, "step": 665 }, { "epoch": 0.42433896145269195, "grad_norm": 19.225542389868956, "learning_rate": 1.8480967313200127e-05, "loss": 0.8062, "step": 666 }, { "epoch": 0.42497610704045874, "grad_norm": 6.9059318448466325, "learning_rate": 1.847579770287661e-05, "loss": 1.1665, "step": 667 }, { "epoch": 0.4256132526282255, "grad_norm": 8.53531439584581, "learning_rate": 1.8470620124479383e-05, "loss": 1.2251, "step": 668 }, { "epoch": 0.42625039821599237, "grad_norm": 8.246355426627197, "learning_rate": 1.8465434583526968e-05, "loss": 1.3787, "step": 669 }, { "epoch": 0.42688754380375915, "grad_norm": 10.560571856526144, "learning_rate": 1.8460241085546366e-05, "loss": 1.6487, "step": 670 }, { "epoch": 0.42752468939152594, "grad_norm": 7.428926348002717, "learning_rate": 1.8455039636073068e-05, "loss": 1.1389, "step": 671 }, { "epoch": 0.4281618349792928, "grad_norm": 6.821621163901021, "learning_rate": 1.844983024065103e-05, "loss": 1.1521, "step": 672 }, { "epoch": 0.42879898056705956, "grad_norm": 15.168875844183727, "learning_rate": 1.844461290483269e-05, "loss": 1.8525, "step": 673 }, { "epoch": 0.4294361261548264, "grad_norm": 9.074959942632928, "learning_rate": 1.8439387634178934e-05, "loss": 1.0719, "step": 674 }, { "epoch": 0.4300732717425932, "grad_norm": 8.326921285548593, "learning_rate": 1.8434154434259116e-05, "loss": 1.4326, "step": 675 }, { "epoch": 0.43071041733036, "grad_norm": 8.004427293573071, "learning_rate": 1.842891331065104e-05, "loss": 1.3552, "step": 676 }, { "epoch": 0.4313475629181268, "grad_norm": 9.39556217388559, "learning_rate": 1.842366426894095e-05, "loss": 1.2856, "step": 677 }, { "epoch": 0.4319847085058936, "grad_norm": 19.434041979442522, "learning_rate": 1.841840731472354e-05, "loss": 1.6765, "step": 678 }, { "epoch": 0.4326218540936604, "grad_norm": 8.404327889325856, "learning_rate": 1.841314245360192e-05, "loss": 1.0294, "step": 679 }, { "epoch": 0.4332589996814272, "grad_norm": 9.989095156760403, "learning_rate": 1.8407869691187656e-05, "loss": 1.3129, "step": 680 }, { "epoch": 0.433896145269194, "grad_norm": 5.809788359027117, "learning_rate": 1.8402589033100698e-05, "loss": 0.7548, "step": 681 }, { "epoch": 0.4345332908569608, "grad_norm": 9.328175592508245, "learning_rate": 1.8397300484969443e-05, "loss": 1.4766, "step": 682 }, { "epoch": 0.43517043644472764, "grad_norm": 11.871740504778703, "learning_rate": 1.8392004052430686e-05, "loss": 1.3417, "step": 683 }, { "epoch": 0.4358075820324944, "grad_norm": 5.68497837394419, "learning_rate": 1.8386699741129626e-05, "loss": 0.928, "step": 684 }, { "epoch": 0.4364447276202612, "grad_norm": 8.437057524856968, "learning_rate": 1.8381387556719853e-05, "loss": 1.2335, "step": 685 }, { "epoch": 0.43708187320802805, "grad_norm": 24.949777622444394, "learning_rate": 1.8376067504863365e-05, "loss": 1.6614, "step": 686 }, { "epoch": 0.43771901879579483, "grad_norm": 25.251284443283048, "learning_rate": 1.8370739591230525e-05, "loss": 1.261, "step": 687 }, { "epoch": 0.4383561643835616, "grad_norm": 4.190382840962181, "learning_rate": 1.8365403821500095e-05, "loss": 0.6351, "step": 688 }, { "epoch": 0.43899330997132846, "grad_norm": 10.636603546426814, "learning_rate": 1.8360060201359194e-05, "loss": 1.25, "step": 689 }, { "epoch": 0.43963045555909525, "grad_norm": 9.5394395770038, "learning_rate": 1.8354708736503324e-05, "loss": 1.0797, "step": 690 }, { "epoch": 0.44026760114686203, "grad_norm": 11.73253283239547, "learning_rate": 1.8349349432636333e-05, "loss": 1.7026, "step": 691 }, { "epoch": 0.44090474673462887, "grad_norm": 12.97425505458184, "learning_rate": 1.8343982295470432e-05, "loss": 1.6843, "step": 692 }, { "epoch": 0.44154189232239566, "grad_norm": 11.986930184042022, "learning_rate": 1.8338607330726186e-05, "loss": 1.2327, "step": 693 }, { "epoch": 0.4421790379101625, "grad_norm": 22.623591583264883, "learning_rate": 1.833322454413249e-05, "loss": 1.5806, "step": 694 }, { "epoch": 0.4428161834979293, "grad_norm": 11.771079150061446, "learning_rate": 1.832783394142659e-05, "loss": 1.344, "step": 695 }, { "epoch": 0.44345332908569607, "grad_norm": 9.372790917034518, "learning_rate": 1.832243552835405e-05, "loss": 1.259, "step": 696 }, { "epoch": 0.4440904746734629, "grad_norm": 11.727780287981775, "learning_rate": 1.8317029310668768e-05, "loss": 1.2522, "step": 697 }, { "epoch": 0.4447276202612297, "grad_norm": 9.013625530172517, "learning_rate": 1.831161529413296e-05, "loss": 1.361, "step": 698 }, { "epoch": 0.4453647658489965, "grad_norm": 10.949030321921308, "learning_rate": 1.8306193484517142e-05, "loss": 1.0176, "step": 699 }, { "epoch": 0.4460019114367633, "grad_norm": 12.11965171296925, "learning_rate": 1.830076388760016e-05, "loss": 1.4393, "step": 700 }, { "epoch": 0.4466390570245301, "grad_norm": 16.329249735147922, "learning_rate": 1.8295326509169135e-05, "loss": 1.202, "step": 701 }, { "epoch": 0.4472762026122969, "grad_norm": 8.202843997162052, "learning_rate": 1.8289881355019498e-05, "loss": 1.0369, "step": 702 }, { "epoch": 0.44791334820006373, "grad_norm": 9.251882358173928, "learning_rate": 1.8284428430954958e-05, "loss": 1.153, "step": 703 }, { "epoch": 0.4485504937878305, "grad_norm": 14.57464771959806, "learning_rate": 1.8278967742787514e-05, "loss": 1.7939, "step": 704 }, { "epoch": 0.4491876393755973, "grad_norm": 13.052174794146705, "learning_rate": 1.827349929633744e-05, "loss": 1.707, "step": 705 }, { "epoch": 0.44982478496336414, "grad_norm": 9.367847827524315, "learning_rate": 1.8268023097433266e-05, "loss": 1.6746, "step": 706 }, { "epoch": 0.45046193055113093, "grad_norm": 6.935294687429904, "learning_rate": 1.8262539151911805e-05, "loss": 1.0927, "step": 707 }, { "epoch": 0.4510990761388977, "grad_norm": 11.303423304722939, "learning_rate": 1.8257047465618114e-05, "loss": 1.7964, "step": 708 }, { "epoch": 0.45173622172666456, "grad_norm": 7.833227109883839, "learning_rate": 1.8251548044405497e-05, "loss": 0.9355, "step": 709 }, { "epoch": 0.45237336731443134, "grad_norm": 10.242104186540413, "learning_rate": 1.8246040894135515e-05, "loss": 1.4675, "step": 710 }, { "epoch": 0.4530105129021981, "grad_norm": 11.482968103090863, "learning_rate": 1.8240526020677955e-05, "loss": 1.5066, "step": 711 }, { "epoch": 0.45364765848996497, "grad_norm": 10.901293315997302, "learning_rate": 1.8235003429910846e-05, "loss": 1.5203, "step": 712 }, { "epoch": 0.45428480407773175, "grad_norm": 12.602431106194745, "learning_rate": 1.8229473127720434e-05, "loss": 1.3892, "step": 713 }, { "epoch": 0.4549219496654986, "grad_norm": 8.765578004235874, "learning_rate": 1.822393512000119e-05, "loss": 1.322, "step": 714 }, { "epoch": 0.4555590952532654, "grad_norm": 9.881302385573655, "learning_rate": 1.821838941265579e-05, "loss": 1.3086, "step": 715 }, { "epoch": 0.45619624084103216, "grad_norm": 9.846181442097519, "learning_rate": 1.821283601159513e-05, "loss": 1.626, "step": 716 }, { "epoch": 0.456833386428799, "grad_norm": 7.799077557671758, "learning_rate": 1.8207274922738296e-05, "loss": 1.0677, "step": 717 }, { "epoch": 0.4574705320165658, "grad_norm": 8.306159070157836, "learning_rate": 1.8201706152012567e-05, "loss": 1.1455, "step": 718 }, { "epoch": 0.4581076776043326, "grad_norm": 8.092988522713656, "learning_rate": 1.8196129705353422e-05, "loss": 1.0431, "step": 719 }, { "epoch": 0.4587448231920994, "grad_norm": 24.086116676345306, "learning_rate": 1.81905455887045e-05, "loss": 1.9832, "step": 720 }, { "epoch": 0.4593819687798662, "grad_norm": 10.517156689211204, "learning_rate": 1.8184953808017638e-05, "loss": 1.2823, "step": 721 }, { "epoch": 0.460019114367633, "grad_norm": 9.992582893394118, "learning_rate": 1.8179354369252827e-05, "loss": 1.705, "step": 722 }, { "epoch": 0.4606562599553998, "grad_norm": 7.727351988036586, "learning_rate": 1.817374727837823e-05, "loss": 1.3733, "step": 723 }, { "epoch": 0.4612934055431666, "grad_norm": 7.298529007391542, "learning_rate": 1.8168132541370148e-05, "loss": 1.4963, "step": 724 }, { "epoch": 0.4619305511309334, "grad_norm": 5.507019804372582, "learning_rate": 1.816251016421306e-05, "loss": 1.134, "step": 725 }, { "epoch": 0.46256769671870024, "grad_norm": 8.611966147424427, "learning_rate": 1.8156880152899565e-05, "loss": 1.146, "step": 726 }, { "epoch": 0.463204842306467, "grad_norm": 10.051440871979823, "learning_rate": 1.8151242513430408e-05, "loss": 1.5713, "step": 727 }, { "epoch": 0.4638419878942338, "grad_norm": 11.944139439757734, "learning_rate": 1.8145597251814465e-05, "loss": 1.666, "step": 728 }, { "epoch": 0.46447913348200065, "grad_norm": 8.592077937313016, "learning_rate": 1.813994437406873e-05, "loss": 1.3586, "step": 729 }, { "epoch": 0.46511627906976744, "grad_norm": 11.390190765073054, "learning_rate": 1.813428388621832e-05, "loss": 1.677, "step": 730 }, { "epoch": 0.4657534246575342, "grad_norm": 8.72184962902982, "learning_rate": 1.812861579429646e-05, "loss": 1.4181, "step": 731 }, { "epoch": 0.46639057024530106, "grad_norm": 8.35050279406086, "learning_rate": 1.812294010434449e-05, "loss": 1.4966, "step": 732 }, { "epoch": 0.46702771583306785, "grad_norm": 5.919176978715163, "learning_rate": 1.8117256822411834e-05, "loss": 1.1121, "step": 733 }, { "epoch": 0.4676648614208347, "grad_norm": 9.589169380605863, "learning_rate": 1.8111565954556006e-05, "loss": 1.1682, "step": 734 }, { "epoch": 0.4683020070086015, "grad_norm": 8.865122874539031, "learning_rate": 1.8105867506842634e-05, "loss": 1.8735, "step": 735 }, { "epoch": 0.46893915259636826, "grad_norm": 8.24776579052446, "learning_rate": 1.8100161485345386e-05, "loss": 1.236, "step": 736 }, { "epoch": 0.4695762981841351, "grad_norm": 12.20987857192294, "learning_rate": 1.8094447896146028e-05, "loss": 1.3652, "step": 737 }, { "epoch": 0.4702134437719019, "grad_norm": 7.166959328002267, "learning_rate": 1.8088726745334387e-05, "loss": 1.3313, "step": 738 }, { "epoch": 0.47085058935966867, "grad_norm": 15.976558403748067, "learning_rate": 1.8082998039008338e-05, "loss": 1.8872, "step": 739 }, { "epoch": 0.4714877349474355, "grad_norm": 8.932928484195468, "learning_rate": 1.8077261783273834e-05, "loss": 1.2764, "step": 740 }, { "epoch": 0.4721248805352023, "grad_norm": 10.394382232076312, "learning_rate": 1.8071517984244846e-05, "loss": 1.5923, "step": 741 }, { "epoch": 0.4727620261229691, "grad_norm": 14.25506779392472, "learning_rate": 1.8065766648043406e-05, "loss": 1.4734, "step": 742 }, { "epoch": 0.4733991717107359, "grad_norm": 9.776594487594899, "learning_rate": 1.8060007780799567e-05, "loss": 1.1721, "step": 743 }, { "epoch": 0.4740363172985027, "grad_norm": 9.637390836637067, "learning_rate": 1.8054241388651415e-05, "loss": 1.4009, "step": 744 }, { "epoch": 0.4746734628862695, "grad_norm": 7.3287836165057705, "learning_rate": 1.8048467477745056e-05, "loss": 0.9749, "step": 745 }, { "epoch": 0.47531060847403633, "grad_norm": 10.276291165075165, "learning_rate": 1.804268605423461e-05, "loss": 1.6113, "step": 746 }, { "epoch": 0.4759477540618031, "grad_norm": 11.255766954948182, "learning_rate": 1.8036897124282195e-05, "loss": 1.2866, "step": 747 }, { "epoch": 0.4765848996495699, "grad_norm": 10.996398224008434, "learning_rate": 1.803110069405795e-05, "loss": 1.0771, "step": 748 }, { "epoch": 0.47722204523733674, "grad_norm": 7.044272471535901, "learning_rate": 1.8025296769739985e-05, "loss": 0.929, "step": 749 }, { "epoch": 0.47785919082510353, "grad_norm": 14.747682180051937, "learning_rate": 1.801948535751442e-05, "loss": 1.7593, "step": 750 }, { "epoch": 0.4784963364128703, "grad_norm": 8.634712855697606, "learning_rate": 1.8013666463575335e-05, "loss": 1.2983, "step": 751 }, { "epoch": 0.47913348200063716, "grad_norm": 6.613325036799877, "learning_rate": 1.8007840094124798e-05, "loss": 1.2593, "step": 752 }, { "epoch": 0.47977062758840394, "grad_norm": 5.916099566787368, "learning_rate": 1.8002006255372842e-05, "loss": 0.9178, "step": 753 }, { "epoch": 0.4804077731761708, "grad_norm": 7.606095983391201, "learning_rate": 1.7996164953537458e-05, "loss": 1.6115, "step": 754 }, { "epoch": 0.48104491876393757, "grad_norm": 19.770761207548244, "learning_rate": 1.7990316194844588e-05, "loss": 1.2864, "step": 755 }, { "epoch": 0.48168206435170435, "grad_norm": 9.836392324686432, "learning_rate": 1.7984459985528138e-05, "loss": 1.6636, "step": 756 }, { "epoch": 0.4823192099394712, "grad_norm": 11.523177207246166, "learning_rate": 1.7978596331829938e-05, "loss": 1.8081, "step": 757 }, { "epoch": 0.482956355527238, "grad_norm": 8.669927161667117, "learning_rate": 1.797272523999976e-05, "loss": 1.4709, "step": 758 }, { "epoch": 0.48359350111500476, "grad_norm": 17.17007138759077, "learning_rate": 1.7966846716295304e-05, "loss": 1.2307, "step": 759 }, { "epoch": 0.4842306467027716, "grad_norm": 6.567564147655352, "learning_rate": 1.796096076698219e-05, "loss": 1.2932, "step": 760 }, { "epoch": 0.4848677922905384, "grad_norm": 20.24102704728638, "learning_rate": 1.7955067398333957e-05, "loss": 1.2484, "step": 761 }, { "epoch": 0.4855049378783052, "grad_norm": 6.9179546487308015, "learning_rate": 1.7949166616632038e-05, "loss": 0.7979, "step": 762 }, { "epoch": 0.486142083466072, "grad_norm": 6.2330603160339075, "learning_rate": 1.7943258428165787e-05, "loss": 1.1598, "step": 763 }, { "epoch": 0.4867792290538388, "grad_norm": 11.78772291028572, "learning_rate": 1.7937342839232435e-05, "loss": 1.4521, "step": 764 }, { "epoch": 0.4874163746416056, "grad_norm": 5.895160989594719, "learning_rate": 1.7931419856137113e-05, "loss": 0.9808, "step": 765 }, { "epoch": 0.48805352022937243, "grad_norm": 7.83807354207524, "learning_rate": 1.7925489485192825e-05, "loss": 1.2937, "step": 766 }, { "epoch": 0.4886906658171392, "grad_norm": 10.449612684331433, "learning_rate": 1.791955173272046e-05, "loss": 1.6387, "step": 767 }, { "epoch": 0.489327811404906, "grad_norm": 9.43518173902069, "learning_rate": 1.791360660504876e-05, "loss": 1.4939, "step": 768 }, { "epoch": 0.48996495699267284, "grad_norm": 4.4424888786448795, "learning_rate": 1.7907654108514335e-05, "loss": 0.9164, "step": 769 }, { "epoch": 0.4906021025804396, "grad_norm": 9.84408589315613, "learning_rate": 1.790169424946165e-05, "loss": 1.4226, "step": 770 }, { "epoch": 0.4912392481682064, "grad_norm": 14.370775702456456, "learning_rate": 1.789572703424302e-05, "loss": 1.4392, "step": 771 }, { "epoch": 0.49187639375597325, "grad_norm": 36.22826821348787, "learning_rate": 1.78897524692186e-05, "loss": 1.3636, "step": 772 }, { "epoch": 0.49251353934374004, "grad_norm": 5.7339518393899604, "learning_rate": 1.7883770560756365e-05, "loss": 0.7671, "step": 773 }, { "epoch": 0.4931506849315068, "grad_norm": 27.663081759421377, "learning_rate": 1.7877781315232136e-05, "loss": 1.7053, "step": 774 }, { "epoch": 0.49378783051927366, "grad_norm": 7.156634202287303, "learning_rate": 1.787178473902954e-05, "loss": 1.0111, "step": 775 }, { "epoch": 0.49442497610704045, "grad_norm": 19.54706584910433, "learning_rate": 1.786578083854003e-05, "loss": 1.6091, "step": 776 }, { "epoch": 0.4950621216948073, "grad_norm": 10.056439320581092, "learning_rate": 1.7859769620162856e-05, "loss": 0.9786, "step": 777 }, { "epoch": 0.4956992672825741, "grad_norm": 9.93137095941016, "learning_rate": 1.7853751090305077e-05, "loss": 1.4406, "step": 778 }, { "epoch": 0.49633641287034086, "grad_norm": 6.144734282968319, "learning_rate": 1.7847725255381525e-05, "loss": 0.8041, "step": 779 }, { "epoch": 0.4969735584581077, "grad_norm": 7.14013497017461, "learning_rate": 1.7841692121814848e-05, "loss": 0.8255, "step": 780 }, { "epoch": 0.4976107040458745, "grad_norm": 8.542890722349338, "learning_rate": 1.7835651696035442e-05, "loss": 1.4165, "step": 781 }, { "epoch": 0.49824784963364127, "grad_norm": 10.81820955294959, "learning_rate": 1.78296039844815e-05, "loss": 1.6035, "step": 782 }, { "epoch": 0.4988849952214081, "grad_norm": 8.255472567960666, "learning_rate": 1.7823548993598972e-05, "loss": 1.3441, "step": 783 }, { "epoch": 0.4995221408091749, "grad_norm": 7.315853106369783, "learning_rate": 1.781748672984156e-05, "loss": 1.2731, "step": 784 }, { "epoch": 0.5001592863969417, "grad_norm": 8.053445756096135, "learning_rate": 1.7811417199670728e-05, "loss": 0.9917, "step": 785 }, { "epoch": 0.5007964319847085, "grad_norm": 8.280560761283978, "learning_rate": 1.7805340409555676e-05, "loss": 1.261, "step": 786 }, { "epoch": 0.5014335775724753, "grad_norm": 7.028597462564036, "learning_rate": 1.779925636597335e-05, "loss": 1.2163, "step": 787 }, { "epoch": 0.5020707231602421, "grad_norm": 14.938934663311874, "learning_rate": 1.779316507540842e-05, "loss": 1.4268, "step": 788 }, { "epoch": 0.5027078687480089, "grad_norm": 6.954662897649293, "learning_rate": 1.7787066544353294e-05, "loss": 1.4154, "step": 789 }, { "epoch": 0.5033450143357757, "grad_norm": 11.645221415517485, "learning_rate": 1.7780960779308074e-05, "loss": 1.0251, "step": 790 }, { "epoch": 0.5039821599235426, "grad_norm": 7.041483459425362, "learning_rate": 1.7774847786780593e-05, "loss": 1.1538, "step": 791 }, { "epoch": 0.5046193055113093, "grad_norm": 6.0303225443946324, "learning_rate": 1.7768727573286378e-05, "loss": 0.9609, "step": 792 }, { "epoch": 0.5052564510990761, "grad_norm": 7.693332954814591, "learning_rate": 1.7762600145348653e-05, "loss": 1.4873, "step": 793 }, { "epoch": 0.505893596686843, "grad_norm": 7.456361301623696, "learning_rate": 1.7756465509498334e-05, "loss": 1.4337, "step": 794 }, { "epoch": 0.5065307422746097, "grad_norm": 10.122401362183414, "learning_rate": 1.7750323672274016e-05, "loss": 1.0754, "step": 795 }, { "epoch": 0.5071678878623765, "grad_norm": 8.270308445837701, "learning_rate": 1.774417464022198e-05, "loss": 1.6157, "step": 796 }, { "epoch": 0.5078050334501434, "grad_norm": 4.07122502918916, "learning_rate": 1.773801841989615e-05, "loss": 0.6831, "step": 797 }, { "epoch": 0.5084421790379101, "grad_norm": 6.889932911508637, "learning_rate": 1.773185501785815e-05, "loss": 0.99, "step": 798 }, { "epoch": 0.509079324625677, "grad_norm": 8.386622106751442, "learning_rate": 1.7725684440677224e-05, "loss": 1.3381, "step": 799 }, { "epoch": 0.5097164702134438, "grad_norm": 5.299346550620101, "learning_rate": 1.7719506694930284e-05, "loss": 0.9896, "step": 800 }, { "epoch": 0.5103536158012105, "grad_norm": 7.644019521885516, "learning_rate": 1.771332178720187e-05, "loss": 1.2479, "step": 801 }, { "epoch": 0.5109907613889774, "grad_norm": 10.304024242198023, "learning_rate": 1.7707129724084168e-05, "loss": 1.9851, "step": 802 }, { "epoch": 0.5116279069767442, "grad_norm": 26.78125331031643, "learning_rate": 1.770093051217698e-05, "loss": 1.3008, "step": 803 }, { "epoch": 0.5122650525645109, "grad_norm": 10.461012095812832, "learning_rate": 1.769472415808773e-05, "loss": 1.5879, "step": 804 }, { "epoch": 0.5129021981522778, "grad_norm": 8.856781869621758, "learning_rate": 1.7688510668431462e-05, "loss": 1.2219, "step": 805 }, { "epoch": 0.5135393437400446, "grad_norm": 5.08338664845283, "learning_rate": 1.7682290049830817e-05, "loss": 0.8947, "step": 806 }, { "epoch": 0.5141764893278115, "grad_norm": 10.22312343506265, "learning_rate": 1.7676062308916034e-05, "loss": 1.363, "step": 807 }, { "epoch": 0.5148136349155782, "grad_norm": 9.990744987103136, "learning_rate": 1.7669827452324947e-05, "loss": 2.0928, "step": 808 }, { "epoch": 0.515450780503345, "grad_norm": 11.064242927400896, "learning_rate": 1.766358548670298e-05, "loss": 1.3915, "step": 809 }, { "epoch": 0.5160879260911119, "grad_norm": 7.430627942939803, "learning_rate": 1.765733641870312e-05, "loss": 1.3218, "step": 810 }, { "epoch": 0.5167250716788786, "grad_norm": 6.3439831880454065, "learning_rate": 1.7651080254985933e-05, "loss": 1.1973, "step": 811 }, { "epoch": 0.5173622172666454, "grad_norm": 7.189641218914793, "learning_rate": 1.764481700221955e-05, "loss": 0.9835, "step": 812 }, { "epoch": 0.5179993628544123, "grad_norm": 10.213279779830895, "learning_rate": 1.7638546667079652e-05, "loss": 1.4592, "step": 813 }, { "epoch": 0.518636508442179, "grad_norm": 8.785162523113573, "learning_rate": 1.763226925624947e-05, "loss": 1.2134, "step": 814 }, { "epoch": 0.5192736540299459, "grad_norm": 12.35704434591847, "learning_rate": 1.7625984776419785e-05, "loss": 1.5583, "step": 815 }, { "epoch": 0.5199107996177127, "grad_norm": 25.979361977426304, "learning_rate": 1.7619693234288894e-05, "loss": 1.1443, "step": 816 }, { "epoch": 0.5205479452054794, "grad_norm": 6.373613342355757, "learning_rate": 1.7613394636562644e-05, "loss": 1.0818, "step": 817 }, { "epoch": 0.5211850907932463, "grad_norm": 7.722358295310241, "learning_rate": 1.7607088989954385e-05, "loss": 1.3517, "step": 818 }, { "epoch": 0.5218222363810131, "grad_norm": 14.127084547905302, "learning_rate": 1.7600776301184984e-05, "loss": 1.5952, "step": 819 }, { "epoch": 0.5224593819687798, "grad_norm": 4.696304339904449, "learning_rate": 1.759445657698282e-05, "loss": 0.6958, "step": 820 }, { "epoch": 0.5230965275565467, "grad_norm": 11.828784286788895, "learning_rate": 1.758812982408377e-05, "loss": 1.6118, "step": 821 }, { "epoch": 0.5237336731443135, "grad_norm": 9.927798290725898, "learning_rate": 1.758179604923119e-05, "loss": 1.2435, "step": 822 }, { "epoch": 0.5243708187320802, "grad_norm": 13.379914959798551, "learning_rate": 1.7575455259175932e-05, "loss": 1.531, "step": 823 }, { "epoch": 0.5250079643198471, "grad_norm": 7.711645739635623, "learning_rate": 1.7569107460676327e-05, "loss": 1.0643, "step": 824 }, { "epoch": 0.5256451099076139, "grad_norm": 5.116454108928259, "learning_rate": 1.7562752660498172e-05, "loss": 0.8143, "step": 825 }, { "epoch": 0.5262822554953807, "grad_norm": 9.081905026659857, "learning_rate": 1.755639086541472e-05, "loss": 1.2587, "step": 826 }, { "epoch": 0.5269194010831475, "grad_norm": 7.268224316830114, "learning_rate": 1.755002208220669e-05, "loss": 1.5273, "step": 827 }, { "epoch": 0.5275565466709143, "grad_norm": 14.322914597224631, "learning_rate": 1.7543646317662243e-05, "loss": 0.8706, "step": 828 }, { "epoch": 0.5281936922586811, "grad_norm": 8.81798547747795, "learning_rate": 1.7537263578576986e-05, "loss": 1.3669, "step": 829 }, { "epoch": 0.5288308378464479, "grad_norm": 5.525185844152531, "learning_rate": 1.7530873871753954e-05, "loss": 0.6248, "step": 830 }, { "epoch": 0.5294679834342148, "grad_norm": 5.273782761537048, "learning_rate": 1.752447720400361e-05, "loss": 0.9062, "step": 831 }, { "epoch": 0.5301051290219815, "grad_norm": 7.111852446872557, "learning_rate": 1.7518073582143848e-05, "loss": 1.2993, "step": 832 }, { "epoch": 0.5307422746097483, "grad_norm": 9.819073767139827, "learning_rate": 1.7511663012999953e-05, "loss": 1.6646, "step": 833 }, { "epoch": 0.5313794201975152, "grad_norm": 9.30048216926761, "learning_rate": 1.750524550340463e-05, "loss": 1.4673, "step": 834 }, { "epoch": 0.5320165657852819, "grad_norm": 7.203507331802977, "learning_rate": 1.749882106019797e-05, "loss": 1.3455, "step": 835 }, { "epoch": 0.5326537113730487, "grad_norm": 8.293168217210722, "learning_rate": 1.7492389690227476e-05, "loss": 1.168, "step": 836 }, { "epoch": 0.5332908569608156, "grad_norm": 23.181141746529814, "learning_rate": 1.7485951400348006e-05, "loss": 1.5051, "step": 837 }, { "epoch": 0.5339280025485823, "grad_norm": 7.856703135374817, "learning_rate": 1.7479506197421815e-05, "loss": 1.5376, "step": 838 }, { "epoch": 0.5345651481363491, "grad_norm": 7.717991354971533, "learning_rate": 1.7473054088318507e-05, "loss": 1.3257, "step": 839 }, { "epoch": 0.535202293724116, "grad_norm": 7.661164562986751, "learning_rate": 1.746659507991507e-05, "loss": 1.2419, "step": 840 }, { "epoch": 0.5358394393118827, "grad_norm": 4.5862824908409205, "learning_rate": 1.746012917909583e-05, "loss": 0.8512, "step": 841 }, { "epoch": 0.5364765848996496, "grad_norm": 3.7185538492441315, "learning_rate": 1.7453656392752462e-05, "loss": 0.6517, "step": 842 }, { "epoch": 0.5371137304874164, "grad_norm": 7.434626259362521, "learning_rate": 1.744717672778398e-05, "loss": 1.2754, "step": 843 }, { "epoch": 0.5377508760751831, "grad_norm": 7.06782321362731, "learning_rate": 1.7440690191096728e-05, "loss": 1.1304, "step": 844 }, { "epoch": 0.53838802166295, "grad_norm": 11.364588356786593, "learning_rate": 1.743419678960438e-05, "loss": 1.2544, "step": 845 }, { "epoch": 0.5390251672507168, "grad_norm": 7.029141283138012, "learning_rate": 1.7427696530227925e-05, "loss": 1.3532, "step": 846 }, { "epoch": 0.5396623128384836, "grad_norm": 6.959707701140308, "learning_rate": 1.7421189419895657e-05, "loss": 1.0623, "step": 847 }, { "epoch": 0.5402994584262504, "grad_norm": 7.492745842110722, "learning_rate": 1.7414675465543175e-05, "loss": 1.1799, "step": 848 }, { "epoch": 0.5409366040140172, "grad_norm": 9.614394987722715, "learning_rate": 1.7408154674113375e-05, "loss": 1.0842, "step": 849 }, { "epoch": 0.5415737496017841, "grad_norm": 7.0088915347472005, "learning_rate": 1.7401627052556433e-05, "loss": 1.0242, "step": 850 }, { "epoch": 0.5422108951895508, "grad_norm": 30.445472813535716, "learning_rate": 1.7395092607829813e-05, "loss": 1.1207, "step": 851 }, { "epoch": 0.5428480407773176, "grad_norm": 9.023021896818044, "learning_rate": 1.7388551346898248e-05, "loss": 1.0468, "step": 852 }, { "epoch": 0.5434851863650845, "grad_norm": 5.085372119810797, "learning_rate": 1.738200327673374e-05, "loss": 0.78, "step": 853 }, { "epoch": 0.5441223319528512, "grad_norm": 20.608447269754624, "learning_rate": 1.737544840431554e-05, "loss": 1.3967, "step": 854 }, { "epoch": 0.544759477540618, "grad_norm": 8.153800722610725, "learning_rate": 1.7368886736630152e-05, "loss": 1.4526, "step": 855 }, { "epoch": 0.5453966231283849, "grad_norm": 10.658428981073683, "learning_rate": 1.7362318280671334e-05, "loss": 1.0271, "step": 856 }, { "epoch": 0.5460337687161516, "grad_norm": 6.0452735939970745, "learning_rate": 1.7355743043440063e-05, "loss": 0.8462, "step": 857 }, { "epoch": 0.5466709143039185, "grad_norm": 9.759657422380466, "learning_rate": 1.734916103194455e-05, "loss": 1.1383, "step": 858 }, { "epoch": 0.5473080598916853, "grad_norm": 11.721631770249298, "learning_rate": 1.734257225320023e-05, "loss": 1.3347, "step": 859 }, { "epoch": 0.547945205479452, "grad_norm": 9.09395118921857, "learning_rate": 1.733597671422975e-05, "loss": 1.5189, "step": 860 }, { "epoch": 0.5485823510672189, "grad_norm": 9.85405346643734, "learning_rate": 1.7329374422062958e-05, "loss": 0.8989, "step": 861 }, { "epoch": 0.5492194966549857, "grad_norm": 11.224730233644813, "learning_rate": 1.7322765383736907e-05, "loss": 1.5735, "step": 862 }, { "epoch": 0.5498566422427524, "grad_norm": 10.84104815006806, "learning_rate": 1.7316149606295825e-05, "loss": 1.7036, "step": 863 }, { "epoch": 0.5504937878305193, "grad_norm": 11.482410531779294, "learning_rate": 1.7309527096791147e-05, "loss": 1.5637, "step": 864 }, { "epoch": 0.5511309334182861, "grad_norm": 9.769378215045641, "learning_rate": 1.7302897862281463e-05, "loss": 1.4548, "step": 865 }, { "epoch": 0.5517680790060528, "grad_norm": 11.49296920810241, "learning_rate": 1.7296261909832538e-05, "loss": 1.3562, "step": 866 }, { "epoch": 0.5524052245938197, "grad_norm": 15.051699550681233, "learning_rate": 1.7289619246517298e-05, "loss": 1.6494, "step": 867 }, { "epoch": 0.5530423701815865, "grad_norm": 8.472559111029048, "learning_rate": 1.7282969879415822e-05, "loss": 1.459, "step": 868 }, { "epoch": 0.5536795157693533, "grad_norm": 8.626855095487814, "learning_rate": 1.7276313815615328e-05, "loss": 1.2993, "step": 869 }, { "epoch": 0.5543166613571201, "grad_norm": 11.507946891379277, "learning_rate": 1.7269651062210183e-05, "loss": 1.2371, "step": 870 }, { "epoch": 0.5549538069448869, "grad_norm": 10.598943587933919, "learning_rate": 1.726298162630188e-05, "loss": 1.6453, "step": 871 }, { "epoch": 0.5555909525326537, "grad_norm": 7.187903466093254, "learning_rate": 1.7256305514999023e-05, "loss": 0.8583, "step": 872 }, { "epoch": 0.5562280981204205, "grad_norm": 10.839885003562001, "learning_rate": 1.7249622735417348e-05, "loss": 1.7891, "step": 873 }, { "epoch": 0.5568652437081874, "grad_norm": 12.92191993324444, "learning_rate": 1.724293329467969e-05, "loss": 1.4304, "step": 874 }, { "epoch": 0.5575023892959541, "grad_norm": 9.925873574116064, "learning_rate": 1.7236237199915985e-05, "loss": 1.4565, "step": 875 }, { "epoch": 0.5581395348837209, "grad_norm": 7.7941936490344785, "learning_rate": 1.722953445826326e-05, "loss": 1.396, "step": 876 }, { "epoch": 0.5587766804714878, "grad_norm": 9.569496217309016, "learning_rate": 1.7222825076865627e-05, "loss": 1.3545, "step": 877 }, { "epoch": 0.5594138260592545, "grad_norm": 16.025465235059237, "learning_rate": 1.721610906287427e-05, "loss": 1.2527, "step": 878 }, { "epoch": 0.5600509716470213, "grad_norm": 10.382945623901547, "learning_rate": 1.7209386423447457e-05, "loss": 1.5848, "step": 879 }, { "epoch": 0.5606881172347882, "grad_norm": 7.233412834735024, "learning_rate": 1.7202657165750507e-05, "loss": 0.885, "step": 880 }, { "epoch": 0.5613252628225549, "grad_norm": 10.563657744973815, "learning_rate": 1.7195921296955792e-05, "loss": 1.3169, "step": 881 }, { "epoch": 0.5619624084103217, "grad_norm": 8.840361013995276, "learning_rate": 1.7189178824242734e-05, "loss": 1.4343, "step": 882 }, { "epoch": 0.5625995539980886, "grad_norm": 9.468888956496798, "learning_rate": 1.718242975479779e-05, "loss": 1.4607, "step": 883 }, { "epoch": 0.5632366995858553, "grad_norm": 43.20237365844101, "learning_rate": 1.717567409581445e-05, "loss": 1.6893, "step": 884 }, { "epoch": 0.5638738451736222, "grad_norm": 8.717588056059462, "learning_rate": 1.716891185449324e-05, "loss": 1.2969, "step": 885 }, { "epoch": 0.564510990761389, "grad_norm": 13.002589148937641, "learning_rate": 1.7162143038041673e-05, "loss": 1.2106, "step": 886 }, { "epoch": 0.5651481363491557, "grad_norm": 11.557022603476886, "learning_rate": 1.71553676536743e-05, "loss": 1.5674, "step": 887 }, { "epoch": 0.5657852819369226, "grad_norm": 14.354158286437698, "learning_rate": 1.714858570861266e-05, "loss": 1.3762, "step": 888 }, { "epoch": 0.5664224275246894, "grad_norm": 5.52355935879636, "learning_rate": 1.7141797210085273e-05, "loss": 0.6676, "step": 889 }, { "epoch": 0.5670595731124562, "grad_norm": 10.185815917091059, "learning_rate": 1.7135002165327665e-05, "loss": 1.7466, "step": 890 }, { "epoch": 0.567696718700223, "grad_norm": 8.045237530207986, "learning_rate": 1.7128200581582325e-05, "loss": 1.4009, "step": 891 }, { "epoch": 0.5683338642879898, "grad_norm": 8.058839607957335, "learning_rate": 1.7121392466098717e-05, "loss": 1.488, "step": 892 }, { "epoch": 0.5689710098757567, "grad_norm": 14.20875232398316, "learning_rate": 1.711457782613327e-05, "loss": 1.5381, "step": 893 }, { "epoch": 0.5696081554635234, "grad_norm": 10.353000468050801, "learning_rate": 1.710775666894936e-05, "loss": 1.2668, "step": 894 }, { "epoch": 0.5702453010512902, "grad_norm": 17.211569493235185, "learning_rate": 1.7100929001817313e-05, "loss": 0.9755, "step": 895 }, { "epoch": 0.5708824466390571, "grad_norm": 10.998245940373158, "learning_rate": 1.7094094832014396e-05, "loss": 1.2517, "step": 896 }, { "epoch": 0.5715195922268238, "grad_norm": 9.697539848416248, "learning_rate": 1.7087254166824798e-05, "loss": 1.5378, "step": 897 }, { "epoch": 0.5721567378145906, "grad_norm": 7.6523815955703975, "learning_rate": 1.7080407013539638e-05, "loss": 1.4531, "step": 898 }, { "epoch": 0.5727938834023575, "grad_norm": 8.417670935338915, "learning_rate": 1.7073553379456957e-05, "loss": 1.3857, "step": 899 }, { "epoch": 0.5734310289901242, "grad_norm": 8.55641978822052, "learning_rate": 1.706669327188169e-05, "loss": 1.4131, "step": 900 }, { "epoch": 0.574068174577891, "grad_norm": 7.75464903439232, "learning_rate": 1.705982669812568e-05, "loss": 1.314, "step": 901 }, { "epoch": 0.5747053201656579, "grad_norm": 10.214841906539466, "learning_rate": 1.7052953665507653e-05, "loss": 1.7358, "step": 902 }, { "epoch": 0.5753424657534246, "grad_norm": 7.022761253480784, "learning_rate": 1.7046074181353235e-05, "loss": 1.231, "step": 903 }, { "epoch": 0.5759796113411915, "grad_norm": 8.435119503479307, "learning_rate": 1.7039188252994916e-05, "loss": 1.178, "step": 904 }, { "epoch": 0.5766167569289583, "grad_norm": 8.272125453444938, "learning_rate": 1.7032295887772057e-05, "loss": 1.2527, "step": 905 }, { "epoch": 0.577253902516725, "grad_norm": 13.149651972859328, "learning_rate": 1.7025397093030882e-05, "loss": 1.0558, "step": 906 }, { "epoch": 0.5778910481044919, "grad_norm": 8.156657447680958, "learning_rate": 1.7018491876124465e-05, "loss": 1.1062, "step": 907 }, { "epoch": 0.5785281936922587, "grad_norm": 6.628818888653732, "learning_rate": 1.701158024441273e-05, "loss": 1.1189, "step": 908 }, { "epoch": 0.5791653392800254, "grad_norm": 9.873679743456293, "learning_rate": 1.7004662205262432e-05, "loss": 1.5143, "step": 909 }, { "epoch": 0.5798024848677923, "grad_norm": 14.264391488382179, "learning_rate": 1.6997737766047154e-05, "loss": 1.8679, "step": 910 }, { "epoch": 0.5804396304555591, "grad_norm": 8.848575458386435, "learning_rate": 1.699080693414731e-05, "loss": 1.3826, "step": 911 }, { "epoch": 0.5810767760433259, "grad_norm": 9.007096468159133, "learning_rate": 1.6983869716950127e-05, "loss": 1.3496, "step": 912 }, { "epoch": 0.5817139216310927, "grad_norm": 3.9624711866093665, "learning_rate": 1.697692612184962e-05, "loss": 0.7385, "step": 913 }, { "epoch": 0.5823510672188595, "grad_norm": 11.470856574607662, "learning_rate": 1.6969976156246628e-05, "loss": 1.8501, "step": 914 }, { "epoch": 0.5829882128066263, "grad_norm": 8.11694962254874, "learning_rate": 1.6963019827548762e-05, "loss": 1.3372, "step": 915 }, { "epoch": 0.5836253583943931, "grad_norm": 5.742517644681737, "learning_rate": 1.6956057143170412e-05, "loss": 1.0586, "step": 916 }, { "epoch": 0.58426250398216, "grad_norm": 9.734885774547992, "learning_rate": 1.6949088110532762e-05, "loss": 1.7229, "step": 917 }, { "epoch": 0.5848996495699267, "grad_norm": 7.829844165147248, "learning_rate": 1.694211273706375e-05, "loss": 0.9406, "step": 918 }, { "epoch": 0.5855367951576935, "grad_norm": 10.032374738760893, "learning_rate": 1.6935131030198068e-05, "loss": 1.296, "step": 919 }, { "epoch": 0.5861739407454604, "grad_norm": 8.559844954227037, "learning_rate": 1.6928142997377155e-05, "loss": 1.405, "step": 920 }, { "epoch": 0.5868110863332271, "grad_norm": 14.469897908711127, "learning_rate": 1.6921148646049218e-05, "loss": 1.9341, "step": 921 }, { "epoch": 0.5874482319209939, "grad_norm": 17.93845412806622, "learning_rate": 1.6914147983669173e-05, "loss": 1.6841, "step": 922 }, { "epoch": 0.5880853775087608, "grad_norm": 7.981484838677479, "learning_rate": 1.690714101769867e-05, "loss": 1.2393, "step": 923 }, { "epoch": 0.5887225230965275, "grad_norm": 17.21943922194113, "learning_rate": 1.6900127755606083e-05, "loss": 1.605, "step": 924 }, { "epoch": 0.5893596686842943, "grad_norm": 16.690343816340295, "learning_rate": 1.6893108204866487e-05, "loss": 1.6636, "step": 925 }, { "epoch": 0.5899968142720612, "grad_norm": 14.763419367420262, "learning_rate": 1.6886082372961666e-05, "loss": 1.1694, "step": 926 }, { "epoch": 0.5906339598598279, "grad_norm": 11.200108799489563, "learning_rate": 1.6879050267380104e-05, "loss": 1.26, "step": 927 }, { "epoch": 0.5912711054475948, "grad_norm": 9.82649961657811, "learning_rate": 1.6872011895616965e-05, "loss": 1.3411, "step": 928 }, { "epoch": 0.5919082510353616, "grad_norm": 12.144040386524846, "learning_rate": 1.686496726517408e-05, "loss": 2.0146, "step": 929 }, { "epoch": 0.5925453966231284, "grad_norm": 7.585077222308106, "learning_rate": 1.6857916383559982e-05, "loss": 1.2465, "step": 930 }, { "epoch": 0.5931825422108952, "grad_norm": 9.899855786369608, "learning_rate": 1.6850859258289832e-05, "loss": 1.262, "step": 931 }, { "epoch": 0.593819687798662, "grad_norm": 10.716171937652529, "learning_rate": 1.6843795896885474e-05, "loss": 1.4916, "step": 932 }, { "epoch": 0.5944568333864289, "grad_norm": 7.2065588087423755, "learning_rate": 1.683672630687538e-05, "loss": 1.0215, "step": 933 }, { "epoch": 0.5950939789741956, "grad_norm": 6.113754205325509, "learning_rate": 1.6829650495794668e-05, "loss": 1.1216, "step": 934 }, { "epoch": 0.5957311245619624, "grad_norm": 7.038442028323394, "learning_rate": 1.682256847118509e-05, "loss": 1.1219, "step": 935 }, { "epoch": 0.5963682701497293, "grad_norm": 9.478265152165221, "learning_rate": 1.681548024059501e-05, "loss": 1.3901, "step": 936 }, { "epoch": 0.597005415737496, "grad_norm": 8.5174207404453, "learning_rate": 1.6808385811579423e-05, "loss": 1.1765, "step": 937 }, { "epoch": 0.5976425613252628, "grad_norm": 7.78782351691358, "learning_rate": 1.680128519169991e-05, "loss": 1.2964, "step": 938 }, { "epoch": 0.5982797069130297, "grad_norm": 6.104818948738051, "learning_rate": 1.6794178388524674e-05, "loss": 1.0082, "step": 939 }, { "epoch": 0.5989168525007964, "grad_norm": 9.14426655877977, "learning_rate": 1.6787065409628486e-05, "loss": 1.6621, "step": 940 }, { "epoch": 0.5995539980885632, "grad_norm": 11.141288886489859, "learning_rate": 1.6779946262592713e-05, "loss": 1.3661, "step": 941 }, { "epoch": 0.6001911436763301, "grad_norm": 5.736086688153554, "learning_rate": 1.67728209550053e-05, "loss": 1.1797, "step": 942 }, { "epoch": 0.6008282892640968, "grad_norm": 7.681860966622187, "learning_rate": 1.676568949446074e-05, "loss": 1.1167, "step": 943 }, { "epoch": 0.6014654348518637, "grad_norm": 6.774160962755834, "learning_rate": 1.6758551888560098e-05, "loss": 1.0789, "step": 944 }, { "epoch": 0.6021025804396305, "grad_norm": 8.680864267074364, "learning_rate": 1.675140814491099e-05, "loss": 1.4636, "step": 945 }, { "epoch": 0.6027397260273972, "grad_norm": 16.652757090779485, "learning_rate": 1.6744258271127566e-05, "loss": 1.7935, "step": 946 }, { "epoch": 0.6033768716151641, "grad_norm": 6.553173952253008, "learning_rate": 1.6737102274830517e-05, "loss": 0.8745, "step": 947 }, { "epoch": 0.6040140172029309, "grad_norm": 30.260344578846432, "learning_rate": 1.6729940163647057e-05, "loss": 1.1472, "step": 948 }, { "epoch": 0.6046511627906976, "grad_norm": 11.558852846172265, "learning_rate": 1.6722771945210912e-05, "loss": 1.6702, "step": 949 }, { "epoch": 0.6052883083784645, "grad_norm": 6.655551969790025, "learning_rate": 1.671559762716233e-05, "loss": 0.7786, "step": 950 }, { "epoch": 0.6059254539662313, "grad_norm": 8.03749518927872, "learning_rate": 1.6708417217148047e-05, "loss": 0.9847, "step": 951 }, { "epoch": 0.606562599553998, "grad_norm": 8.034162664448456, "learning_rate": 1.6701230722821302e-05, "loss": 1.6409, "step": 952 }, { "epoch": 0.6071997451417649, "grad_norm": 7.217517305330075, "learning_rate": 1.669403815184181e-05, "loss": 1.1355, "step": 953 }, { "epoch": 0.6078368907295317, "grad_norm": 17.140571630808267, "learning_rate": 1.6686839511875774e-05, "loss": 1.45, "step": 954 }, { "epoch": 0.6084740363172985, "grad_norm": 10.365483662858008, "learning_rate": 1.667963481059585e-05, "loss": 1.0114, "step": 955 }, { "epoch": 0.6091111819050653, "grad_norm": 10.647124638358882, "learning_rate": 1.6672424055681177e-05, "loss": 1.3331, "step": 956 }, { "epoch": 0.6097483274928321, "grad_norm": 12.776125258705214, "learning_rate": 1.6665207254817323e-05, "loss": 1.1321, "step": 957 }, { "epoch": 0.6103854730805989, "grad_norm": 9.928939338038145, "learning_rate": 1.665798441569631e-05, "loss": 1.4209, "step": 958 }, { "epoch": 0.6110226186683657, "grad_norm": 7.733852596468951, "learning_rate": 1.66507555460166e-05, "loss": 1.5049, "step": 959 }, { "epoch": 0.6116597642561326, "grad_norm": 10.20037547441489, "learning_rate": 1.6643520653483082e-05, "loss": 1.4513, "step": 960 }, { "epoch": 0.6122969098438993, "grad_norm": 12.837209218121389, "learning_rate": 1.6636279745807053e-05, "loss": 1.2841, "step": 961 }, { "epoch": 0.6129340554316661, "grad_norm": 12.422881821888696, "learning_rate": 1.6629032830706232e-05, "loss": 1.6414, "step": 962 }, { "epoch": 0.613571201019433, "grad_norm": 9.325450266306708, "learning_rate": 1.6621779915904742e-05, "loss": 1.1669, "step": 963 }, { "epoch": 0.6142083466071997, "grad_norm": 13.260997705202644, "learning_rate": 1.6614521009133094e-05, "loss": 1.2371, "step": 964 }, { "epoch": 0.6148454921949665, "grad_norm": 8.547791402770114, "learning_rate": 1.6607256118128186e-05, "loss": 1.1542, "step": 965 }, { "epoch": 0.6154826377827334, "grad_norm": 53.13471660187148, "learning_rate": 1.65999852506333e-05, "loss": 1.8438, "step": 966 }, { "epoch": 0.6161197833705001, "grad_norm": 9.46333758509313, "learning_rate": 1.659270841439809e-05, "loss": 1.5522, "step": 967 }, { "epoch": 0.616756928958267, "grad_norm": 14.486584000702882, "learning_rate": 1.658542561717856e-05, "loss": 1.6924, "step": 968 }, { "epoch": 0.6173940745460338, "grad_norm": 17.083426519401677, "learning_rate": 1.6578136866737073e-05, "loss": 1.8242, "step": 969 }, { "epoch": 0.6180312201338005, "grad_norm": 8.528624029659582, "learning_rate": 1.6570842170842346e-05, "loss": 1.0693, "step": 970 }, { "epoch": 0.6186683657215674, "grad_norm": 8.59419704848926, "learning_rate": 1.6563541537269418e-05, "loss": 1.5994, "step": 971 }, { "epoch": 0.6193055113093342, "grad_norm": 11.209378687027383, "learning_rate": 1.6556234973799667e-05, "loss": 1.2139, "step": 972 }, { "epoch": 0.619942656897101, "grad_norm": 16.60891370814216, "learning_rate": 1.654892248822079e-05, "loss": 1.5676, "step": 973 }, { "epoch": 0.6205798024848678, "grad_norm": 12.091540221371504, "learning_rate": 1.6541604088326792e-05, "loss": 1.4452, "step": 974 }, { "epoch": 0.6212169480726346, "grad_norm": 5.920346606767888, "learning_rate": 1.6534279781917985e-05, "loss": 1.0886, "step": 975 }, { "epoch": 0.6218540936604015, "grad_norm": 15.176353782488098, "learning_rate": 1.652694957680098e-05, "loss": 1.7241, "step": 976 }, { "epoch": 0.6224912392481682, "grad_norm": 9.636783537210228, "learning_rate": 1.6519613480788666e-05, "loss": 1.396, "step": 977 }, { "epoch": 0.623128384835935, "grad_norm": 25.907304552691585, "learning_rate": 1.6512271501700215e-05, "loss": 1.3992, "step": 978 }, { "epoch": 0.6237655304237019, "grad_norm": 10.917415825767288, "learning_rate": 1.6504923647361076e-05, "loss": 1.8647, "step": 979 }, { "epoch": 0.6244026760114686, "grad_norm": 12.960087337691663, "learning_rate": 1.649756992560295e-05, "loss": 1.0295, "step": 980 }, { "epoch": 0.6250398215992354, "grad_norm": 13.112567563537153, "learning_rate": 1.6490210344263796e-05, "loss": 1.4905, "step": 981 }, { "epoch": 0.6256769671870023, "grad_norm": 13.74438631346764, "learning_rate": 1.648284491118782e-05, "loss": 1.6272, "step": 982 }, { "epoch": 0.626314112774769, "grad_norm": 5.162172162205254, "learning_rate": 1.6475473634225458e-05, "loss": 0.6803, "step": 983 }, { "epoch": 0.6269512583625358, "grad_norm": 18.22646450692492, "learning_rate": 1.646809652123339e-05, "loss": 1.4697, "step": 984 }, { "epoch": 0.6275884039503027, "grad_norm": 9.585279363645153, "learning_rate": 1.6460713580074507e-05, "loss": 1.1858, "step": 985 }, { "epoch": 0.6282255495380694, "grad_norm": 10.642845659939338, "learning_rate": 1.64533248186179e-05, "loss": 1.4419, "step": 986 }, { "epoch": 0.6288626951258363, "grad_norm": 13.377242893957739, "learning_rate": 1.6445930244738884e-05, "loss": 1.0808, "step": 987 }, { "epoch": 0.6294998407136031, "grad_norm": 13.690582076355225, "learning_rate": 1.6438529866318957e-05, "loss": 1.4651, "step": 988 }, { "epoch": 0.6301369863013698, "grad_norm": 10.559738890315074, "learning_rate": 1.6431123691245813e-05, "loss": 1.3953, "step": 989 }, { "epoch": 0.6307741318891367, "grad_norm": 11.494390162081741, "learning_rate": 1.6423711727413308e-05, "loss": 1.4553, "step": 990 }, { "epoch": 0.6314112774769035, "grad_norm": 17.681408897039095, "learning_rate": 1.641629398272149e-05, "loss": 1.3831, "step": 991 }, { "epoch": 0.6320484230646702, "grad_norm": 11.380266849987054, "learning_rate": 1.6408870465076548e-05, "loss": 1.3464, "step": 992 }, { "epoch": 0.6326855686524371, "grad_norm": 9.666042741782721, "learning_rate": 1.6401441182390836e-05, "loss": 1.1018, "step": 993 }, { "epoch": 0.6333227142402039, "grad_norm": 11.772949686151486, "learning_rate": 1.639400614258285e-05, "loss": 1.3113, "step": 994 }, { "epoch": 0.6339598598279707, "grad_norm": 9.188462703469792, "learning_rate": 1.6386565353577225e-05, "loss": 1.1888, "step": 995 }, { "epoch": 0.6345970054157375, "grad_norm": 10.836114487450542, "learning_rate": 1.6379118823304715e-05, "loss": 1.2688, "step": 996 }, { "epoch": 0.6352341510035043, "grad_norm": 12.44194872079717, "learning_rate": 1.6371666559702205e-05, "loss": 1.2678, "step": 997 }, { "epoch": 0.6358712965912711, "grad_norm": 12.210112907104703, "learning_rate": 1.6364208570712677e-05, "loss": 1.5342, "step": 998 }, { "epoch": 0.6365084421790379, "grad_norm": 14.560126578300654, "learning_rate": 1.635674486428523e-05, "loss": 1.2804, "step": 999 }, { "epoch": 0.6371455877668047, "grad_norm": 13.916930362446493, "learning_rate": 1.634927544837505e-05, "loss": 1.824, "step": 1000 }, { "epoch": 0.6377827333545715, "grad_norm": 14.45888767520857, "learning_rate": 1.6341800330943406e-05, "loss": 1.4158, "step": 1001 }, { "epoch": 0.6384198789423383, "grad_norm": 9.577863412675365, "learning_rate": 1.6334319519957645e-05, "loss": 1.2697, "step": 1002 }, { "epoch": 0.6390570245301052, "grad_norm": 10.272328050346236, "learning_rate": 1.6326833023391192e-05, "loss": 1.2456, "step": 1003 }, { "epoch": 0.6396941701178719, "grad_norm": 15.97252690157074, "learning_rate": 1.6319340849223514e-05, "loss": 2.1597, "step": 1004 }, { "epoch": 0.6403313157056387, "grad_norm": 12.680493399345838, "learning_rate": 1.6311843005440147e-05, "loss": 1.7126, "step": 1005 }, { "epoch": 0.6409684612934056, "grad_norm": 10.07762398948109, "learning_rate": 1.630433950003266e-05, "loss": 1.3815, "step": 1006 }, { "epoch": 0.6416056068811723, "grad_norm": 9.220128463349058, "learning_rate": 1.6296830340998654e-05, "loss": 1.2756, "step": 1007 }, { "epoch": 0.6422427524689391, "grad_norm": 13.010654412223523, "learning_rate": 1.628931553634177e-05, "loss": 1.4597, "step": 1008 }, { "epoch": 0.642879898056706, "grad_norm": 11.305839906080106, "learning_rate": 1.6281795094071652e-05, "loss": 1.3386, "step": 1009 }, { "epoch": 0.6435170436444727, "grad_norm": 9.888783912165938, "learning_rate": 1.6274269022203955e-05, "loss": 1.3425, "step": 1010 }, { "epoch": 0.6441541892322395, "grad_norm": 10.748016292138635, "learning_rate": 1.6266737328760347e-05, "loss": 1.6348, "step": 1011 }, { "epoch": 0.6447913348200064, "grad_norm": 7.970739679782896, "learning_rate": 1.625920002176847e-05, "loss": 1.3314, "step": 1012 }, { "epoch": 0.6454284804077732, "grad_norm": 9.180406250250147, "learning_rate": 1.625165710926196e-05, "loss": 1.437, "step": 1013 }, { "epoch": 0.64606562599554, "grad_norm": 12.611443327246624, "learning_rate": 1.624410859928043e-05, "loss": 1.3176, "step": 1014 }, { "epoch": 0.6467027715833068, "grad_norm": 15.929161861933668, "learning_rate": 1.623655449986944e-05, "loss": 1.7842, "step": 1015 }, { "epoch": 0.6473399171710736, "grad_norm": 7.688361814148123, "learning_rate": 1.622899481908053e-05, "loss": 0.937, "step": 1016 }, { "epoch": 0.6479770627588404, "grad_norm": 10.893510079084914, "learning_rate": 1.622142956497119e-05, "loss": 0.8359, "step": 1017 }, { "epoch": 0.6486142083466072, "grad_norm": 8.845097385457475, "learning_rate": 1.6213858745604827e-05, "loss": 1.4001, "step": 1018 }, { "epoch": 0.649251353934374, "grad_norm": 21.017507403239705, "learning_rate": 1.6206282369050797e-05, "loss": 1.6538, "step": 1019 }, { "epoch": 0.6498884995221408, "grad_norm": 8.36669815364719, "learning_rate": 1.6198700443384378e-05, "loss": 1.2673, "step": 1020 }, { "epoch": 0.6505256451099076, "grad_norm": 10.137129071397846, "learning_rate": 1.6191112976686757e-05, "loss": 1.5272, "step": 1021 }, { "epoch": 0.6511627906976745, "grad_norm": 6.128662521896482, "learning_rate": 1.618351997704503e-05, "loss": 0.8789, "step": 1022 }, { "epoch": 0.6517999362854412, "grad_norm": 9.871631829869305, "learning_rate": 1.6175921452552193e-05, "loss": 0.8782, "step": 1023 }, { "epoch": 0.652437081873208, "grad_norm": 10.034181224456184, "learning_rate": 1.616831741130713e-05, "loss": 1.4182, "step": 1024 }, { "epoch": 0.6530742274609749, "grad_norm": 11.290038295604898, "learning_rate": 1.61607078614146e-05, "loss": 1.4082, "step": 1025 }, { "epoch": 0.6537113730487416, "grad_norm": 9.819260391193287, "learning_rate": 1.6153092810985235e-05, "loss": 1.3833, "step": 1026 }, { "epoch": 0.6543485186365084, "grad_norm": 29.935257934052817, "learning_rate": 1.6145472268135533e-05, "loss": 1.5125, "step": 1027 }, { "epoch": 0.6549856642242753, "grad_norm": 8.502038153220827, "learning_rate": 1.6137846240987845e-05, "loss": 1.2581, "step": 1028 }, { "epoch": 0.655622809812042, "grad_norm": 7.859831401623185, "learning_rate": 1.6130214737670364e-05, "loss": 1.0219, "step": 1029 }, { "epoch": 0.6562599553998089, "grad_norm": 7.5011391661216855, "learning_rate": 1.6122577766317125e-05, "loss": 1.2732, "step": 1030 }, { "epoch": 0.6568971009875757, "grad_norm": 29.18089356248298, "learning_rate": 1.6114935335067985e-05, "loss": 0.8611, "step": 1031 }, { "epoch": 0.6575342465753424, "grad_norm": 11.757293222889121, "learning_rate": 1.6107287452068628e-05, "loss": 1.1597, "step": 1032 }, { "epoch": 0.6581713921631093, "grad_norm": 17.993704031601908, "learning_rate": 1.6099634125470544e-05, "loss": 1.1656, "step": 1033 }, { "epoch": 0.6588085377508761, "grad_norm": 14.697046809782963, "learning_rate": 1.6091975363431022e-05, "loss": 1.0737, "step": 1034 }, { "epoch": 0.6594456833386428, "grad_norm": 12.071422872644964, "learning_rate": 1.6084311174113153e-05, "loss": 1.5576, "step": 1035 }, { "epoch": 0.6600828289264097, "grad_norm": 14.316777671315592, "learning_rate": 1.60766415656858e-05, "loss": 0.9448, "step": 1036 }, { "epoch": 0.6607199745141765, "grad_norm": 17.7952228931381, "learning_rate": 1.6068966546323625e-05, "loss": 1.4592, "step": 1037 }, { "epoch": 0.6613571201019433, "grad_norm": 8.926830878481743, "learning_rate": 1.6061286124207027e-05, "loss": 1.1454, "step": 1038 }, { "epoch": 0.6619942656897101, "grad_norm": 15.885913348840091, "learning_rate": 1.6053600307522184e-05, "loss": 1.5259, "step": 1039 }, { "epoch": 0.6626314112774769, "grad_norm": 10.49822827017974, "learning_rate": 1.6045909104461022e-05, "loss": 1.0508, "step": 1040 }, { "epoch": 0.6632685568652437, "grad_norm": 14.405649941214334, "learning_rate": 1.6038212523221194e-05, "loss": 1.3494, "step": 1041 }, { "epoch": 0.6639057024530105, "grad_norm": 15.187281597490426, "learning_rate": 1.6030510572006112e-05, "loss": 1.231, "step": 1042 }, { "epoch": 0.6645428480407773, "grad_norm": 10.349600639248406, "learning_rate": 1.6022803259024874e-05, "loss": 1.3286, "step": 1043 }, { "epoch": 0.6651799936285441, "grad_norm": 13.667696081885234, "learning_rate": 1.6015090592492334e-05, "loss": 1.5286, "step": 1044 }, { "epoch": 0.6658171392163109, "grad_norm": 14.299249915104168, "learning_rate": 1.600737258062902e-05, "loss": 1.0453, "step": 1045 }, { "epoch": 0.6664542848040778, "grad_norm": 8.589112113930211, "learning_rate": 1.5999649231661172e-05, "loss": 1.0219, "step": 1046 }, { "epoch": 0.6670914303918445, "grad_norm": 9.990081232483119, "learning_rate": 1.599192055382072e-05, "loss": 1.3064, "step": 1047 }, { "epoch": 0.6677285759796113, "grad_norm": 16.193166711506446, "learning_rate": 1.5984186555345267e-05, "loss": 1.5791, "step": 1048 }, { "epoch": 0.6683657215673782, "grad_norm": 12.002052467306697, "learning_rate": 1.597644724447809e-05, "loss": 1.3679, "step": 1049 }, { "epoch": 0.6690028671551449, "grad_norm": 10.26151285700156, "learning_rate": 1.5968702629468127e-05, "loss": 1.1672, "step": 1050 }, { "epoch": 0.6696400127429117, "grad_norm": 10.12820644301558, "learning_rate": 1.5960952718569973e-05, "loss": 0.8223, "step": 1051 }, { "epoch": 0.6702771583306786, "grad_norm": 10.02943947085702, "learning_rate": 1.5953197520043864e-05, "loss": 1.2484, "step": 1052 }, { "epoch": 0.6709143039184453, "grad_norm": 21.157917040129927, "learning_rate": 1.594543704215567e-05, "loss": 1.7375, "step": 1053 }, { "epoch": 0.6715514495062122, "grad_norm": 10.984601688200772, "learning_rate": 1.5937671293176896e-05, "loss": 1.1047, "step": 1054 }, { "epoch": 0.672188595093979, "grad_norm": 13.215159808431494, "learning_rate": 1.5929900281384655e-05, "loss": 1.0599, "step": 1055 }, { "epoch": 0.6728257406817458, "grad_norm": 23.987500757322156, "learning_rate": 1.5922124015061676e-05, "loss": 1.3707, "step": 1056 }, { "epoch": 0.6734628862695126, "grad_norm": 26.0468175761627, "learning_rate": 1.5914342502496288e-05, "loss": 1.1429, "step": 1057 }, { "epoch": 0.6741000318572794, "grad_norm": 12.59966186040277, "learning_rate": 1.5906555751982415e-05, "loss": 1.6523, "step": 1058 }, { "epoch": 0.6747371774450462, "grad_norm": 14.396941749096008, "learning_rate": 1.589876377181955e-05, "loss": 1.5229, "step": 1059 }, { "epoch": 0.675374323032813, "grad_norm": 15.248946623440167, "learning_rate": 1.5890966570312773e-05, "loss": 1.4739, "step": 1060 }, { "epoch": 0.6760114686205798, "grad_norm": 13.3361218673708, "learning_rate": 1.5883164155772732e-05, "loss": 0.9435, "step": 1061 }, { "epoch": 0.6766486142083467, "grad_norm": 15.896514250982145, "learning_rate": 1.587535653651562e-05, "loss": 1.5955, "step": 1062 }, { "epoch": 0.6772857597961134, "grad_norm": 8.55728832901544, "learning_rate": 1.586754372086318e-05, "loss": 0.7501, "step": 1063 }, { "epoch": 0.6779229053838802, "grad_norm": 8.677953332208903, "learning_rate": 1.58597257171427e-05, "loss": 1.1783, "step": 1064 }, { "epoch": 0.6785600509716471, "grad_norm": 21.089025286024388, "learning_rate": 1.5851902533686995e-05, "loss": 1.583, "step": 1065 }, { "epoch": 0.6791971965594138, "grad_norm": 11.932431084246586, "learning_rate": 1.5844074178834393e-05, "loss": 1.2305, "step": 1066 }, { "epoch": 0.6798343421471806, "grad_norm": 25.254111582083098, "learning_rate": 1.583624066092875e-05, "loss": 1.6851, "step": 1067 }, { "epoch": 0.6804714877349475, "grad_norm": 15.071327666923995, "learning_rate": 1.5828401988319407e-05, "loss": 1.0439, "step": 1068 }, { "epoch": 0.6811086333227142, "grad_norm": 13.177846489943649, "learning_rate": 1.582055816936121e-05, "loss": 1.2037, "step": 1069 }, { "epoch": 0.681745778910481, "grad_norm": 7.463836513735721, "learning_rate": 1.5812709212414487e-05, "loss": 0.9667, "step": 1070 }, { "epoch": 0.6823829244982479, "grad_norm": 10.987865399196092, "learning_rate": 1.580485512584505e-05, "loss": 1.4253, "step": 1071 }, { "epoch": 0.6830200700860146, "grad_norm": 18.945720584839197, "learning_rate": 1.5796995918024162e-05, "loss": 1.3973, "step": 1072 }, { "epoch": 0.6836572156737815, "grad_norm": 14.468282045516068, "learning_rate": 1.5789131597328562e-05, "loss": 1.5134, "step": 1073 }, { "epoch": 0.6842943612615483, "grad_norm": 10.639446272240697, "learning_rate": 1.5781262172140426e-05, "loss": 1.0349, "step": 1074 }, { "epoch": 0.684931506849315, "grad_norm": 12.911692453433307, "learning_rate": 1.577338765084738e-05, "loss": 1.306, "step": 1075 }, { "epoch": 0.6855686524370819, "grad_norm": 28.323037125362625, "learning_rate": 1.5765508041842464e-05, "loss": 1.2432, "step": 1076 }, { "epoch": 0.6862057980248487, "grad_norm": 17.580566978411014, "learning_rate": 1.5757623353524172e-05, "loss": 1.2932, "step": 1077 }, { "epoch": 0.6868429436126154, "grad_norm": 11.149569242164223, "learning_rate": 1.5749733594296384e-05, "loss": 1.1902, "step": 1078 }, { "epoch": 0.6874800892003823, "grad_norm": 11.89897024502198, "learning_rate": 1.5741838772568393e-05, "loss": 1.4255, "step": 1079 }, { "epoch": 0.6881172347881491, "grad_norm": 14.336278432991755, "learning_rate": 1.5733938896754892e-05, "loss": 1.6282, "step": 1080 }, { "epoch": 0.6887543803759159, "grad_norm": 8.406623948486148, "learning_rate": 1.572603397527596e-05, "loss": 1.1219, "step": 1081 }, { "epoch": 0.6893915259636827, "grad_norm": 14.661903277073062, "learning_rate": 1.5718124016557045e-05, "loss": 1.2357, "step": 1082 }, { "epoch": 0.6900286715514495, "grad_norm": 24.269864914922692, "learning_rate": 1.5710209029028977e-05, "loss": 1.439, "step": 1083 }, { "epoch": 0.6906658171392163, "grad_norm": 12.639418646557527, "learning_rate": 1.570228902112794e-05, "loss": 1.4238, "step": 1084 }, { "epoch": 0.6913029627269831, "grad_norm": 12.321535773272766, "learning_rate": 1.5694364001295464e-05, "loss": 1.2787, "step": 1085 }, { "epoch": 0.69194010831475, "grad_norm": 9.201010907078023, "learning_rate": 1.568643397797843e-05, "loss": 1.4237, "step": 1086 }, { "epoch": 0.6925772539025167, "grad_norm": 9.089832636753332, "learning_rate": 1.5678498959629047e-05, "loss": 1.1556, "step": 1087 }, { "epoch": 0.6932143994902835, "grad_norm": 19.595988905513714, "learning_rate": 1.567055895470485e-05, "loss": 1.6287, "step": 1088 }, { "epoch": 0.6938515450780504, "grad_norm": 17.658157150556953, "learning_rate": 1.5662613971668682e-05, "loss": 1.6489, "step": 1089 }, { "epoch": 0.6944886906658171, "grad_norm": 9.129873901769164, "learning_rate": 1.5654664018988703e-05, "loss": 1.0597, "step": 1090 }, { "epoch": 0.6951258362535839, "grad_norm": 47.26316890931682, "learning_rate": 1.5646709105138363e-05, "loss": 1.3323, "step": 1091 }, { "epoch": 0.6957629818413508, "grad_norm": 25.35595198427831, "learning_rate": 1.5638749238596402e-05, "loss": 1.5898, "step": 1092 }, { "epoch": 0.6964001274291175, "grad_norm": 13.20574331543404, "learning_rate": 1.5630784427846834e-05, "loss": 1.2483, "step": 1093 }, { "epoch": 0.6970372730168843, "grad_norm": 17.524329990937076, "learning_rate": 1.5622814681378955e-05, "loss": 1.489, "step": 1094 }, { "epoch": 0.6976744186046512, "grad_norm": 12.503467012305759, "learning_rate": 1.5614840007687303e-05, "loss": 1.5391, "step": 1095 }, { "epoch": 0.698311564192418, "grad_norm": 21.602145393086314, "learning_rate": 1.560686041527169e-05, "loss": 1.9133, "step": 1096 }, { "epoch": 0.6989487097801848, "grad_norm": 10.740990342627617, "learning_rate": 1.5598875912637152e-05, "loss": 1.4014, "step": 1097 }, { "epoch": 0.6995858553679516, "grad_norm": 14.957487835949134, "learning_rate": 1.5590886508293967e-05, "loss": 1.4927, "step": 1098 }, { "epoch": 0.7002230009557184, "grad_norm": 26.806131468208203, "learning_rate": 1.5582892210757634e-05, "loss": 1.1614, "step": 1099 }, { "epoch": 0.7008601465434852, "grad_norm": 11.337094806919875, "learning_rate": 1.5574893028548873e-05, "loss": 1.5188, "step": 1100 }, { "epoch": 0.701497292131252, "grad_norm": 25.649842639889208, "learning_rate": 1.55668889701936e-05, "loss": 1.0939, "step": 1101 }, { "epoch": 0.7021344377190188, "grad_norm": 12.150043143878026, "learning_rate": 1.5558880044222943e-05, "loss": 1.1367, "step": 1102 }, { "epoch": 0.7027715833067856, "grad_norm": 15.542378689843074, "learning_rate": 1.555086625917321e-05, "loss": 1.0977, "step": 1103 }, { "epoch": 0.7034087288945524, "grad_norm": 11.24443130345626, "learning_rate": 1.554284762358588e-05, "loss": 1.5906, "step": 1104 }, { "epoch": 0.7040458744823193, "grad_norm": 13.729532395008798, "learning_rate": 1.5534824146007617e-05, "loss": 1.3696, "step": 1105 }, { "epoch": 0.704683020070086, "grad_norm": 17.225969462790932, "learning_rate": 1.552679583499024e-05, "loss": 1.7827, "step": 1106 }, { "epoch": 0.7053201656578528, "grad_norm": 11.440161555747224, "learning_rate": 1.551876269909071e-05, "loss": 1.0999, "step": 1107 }, { "epoch": 0.7059573112456197, "grad_norm": 9.380289083198852, "learning_rate": 1.5510724746871148e-05, "loss": 1.5562, "step": 1108 }, { "epoch": 0.7065944568333864, "grad_norm": 9.541425178770663, "learning_rate": 1.5502681986898796e-05, "loss": 0.9559, "step": 1109 }, { "epoch": 0.7072316024211532, "grad_norm": 17.736567323562966, "learning_rate": 1.5494634427746026e-05, "loss": 1.3462, "step": 1110 }, { "epoch": 0.7078687480089201, "grad_norm": 18.790668709703926, "learning_rate": 1.548658207799032e-05, "loss": 1.5667, "step": 1111 }, { "epoch": 0.7085058935966868, "grad_norm": 13.392205361864303, "learning_rate": 1.5478524946214275e-05, "loss": 1.532, "step": 1112 }, { "epoch": 0.7091430391844536, "grad_norm": 9.176415716993294, "learning_rate": 1.547046304100557e-05, "loss": 1.2648, "step": 1113 }, { "epoch": 0.7097801847722205, "grad_norm": 14.19825730817287, "learning_rate": 1.546239637095699e-05, "loss": 1.4121, "step": 1114 }, { "epoch": 0.7104173303599872, "grad_norm": 10.256332312940112, "learning_rate": 1.545432494466638e-05, "loss": 1.1256, "step": 1115 }, { "epoch": 0.7110544759477541, "grad_norm": 38.60917150598471, "learning_rate": 1.5446248770736677e-05, "loss": 1.5042, "step": 1116 }, { "epoch": 0.7116916215355209, "grad_norm": 7.543048309039195, "learning_rate": 1.5438167857775856e-05, "loss": 0.9374, "step": 1117 }, { "epoch": 0.7123287671232876, "grad_norm": 9.481407155527167, "learning_rate": 1.5430082214396955e-05, "loss": 1.4705, "step": 1118 }, { "epoch": 0.7129659127110545, "grad_norm": 13.933885892144925, "learning_rate": 1.5421991849218046e-05, "loss": 1.7815, "step": 1119 }, { "epoch": 0.7136030582988213, "grad_norm": 10.66772749152227, "learning_rate": 1.541389677086225e-05, "loss": 1.2462, "step": 1120 }, { "epoch": 0.714240203886588, "grad_norm": 11.647584969448232, "learning_rate": 1.5405796987957695e-05, "loss": 1.1442, "step": 1121 }, { "epoch": 0.7148773494743549, "grad_norm": 10.258870678606664, "learning_rate": 1.5397692509137527e-05, "loss": 1.3013, "step": 1122 }, { "epoch": 0.7155144950621217, "grad_norm": 11.038009395359209, "learning_rate": 1.5389583343039896e-05, "loss": 1.4558, "step": 1123 }, { "epoch": 0.7161516406498885, "grad_norm": 19.799291159119417, "learning_rate": 1.538146949830796e-05, "loss": 1.5825, "step": 1124 }, { "epoch": 0.7167887862376553, "grad_norm": 6.833122514076152, "learning_rate": 1.5373350983589848e-05, "loss": 0.9163, "step": 1125 }, { "epoch": 0.7174259318254221, "grad_norm": 10.810070321589048, "learning_rate": 1.5365227807538678e-05, "loss": 1.115, "step": 1126 }, { "epoch": 0.7180630774131889, "grad_norm": 10.071402962400088, "learning_rate": 1.5357099978812524e-05, "loss": 0.9993, "step": 1127 }, { "epoch": 0.7187002230009557, "grad_norm": 14.41732264765947, "learning_rate": 1.5348967506074435e-05, "loss": 1.5667, "step": 1128 }, { "epoch": 0.7193373685887225, "grad_norm": 8.779743688391157, "learning_rate": 1.5340830397992395e-05, "loss": 1.146, "step": 1129 }, { "epoch": 0.7199745141764893, "grad_norm": 12.747336885310087, "learning_rate": 1.5332688663239337e-05, "loss": 1.0703, "step": 1130 }, { "epoch": 0.7206116597642561, "grad_norm": 16.434315457631374, "learning_rate": 1.532454231049313e-05, "loss": 1.8403, "step": 1131 }, { "epoch": 0.721248805352023, "grad_norm": 12.666207986316548, "learning_rate": 1.5316391348436547e-05, "loss": 1.5375, "step": 1132 }, { "epoch": 0.7218859509397897, "grad_norm": 6.694618640824554, "learning_rate": 1.530823578575729e-05, "loss": 0.9602, "step": 1133 }, { "epoch": 0.7225230965275565, "grad_norm": 11.857264099924333, "learning_rate": 1.530007563114796e-05, "loss": 1.6272, "step": 1134 }, { "epoch": 0.7231602421153234, "grad_norm": 12.960333537222605, "learning_rate": 1.5291910893306055e-05, "loss": 1.3228, "step": 1135 }, { "epoch": 0.7237973877030902, "grad_norm": 7.8420753360927975, "learning_rate": 1.5283741580933945e-05, "loss": 1.0858, "step": 1136 }, { "epoch": 0.7244345332908569, "grad_norm": 16.246271212142545, "learning_rate": 1.52755677027389e-05, "loss": 1.3477, "step": 1137 }, { "epoch": 0.7250716788786238, "grad_norm": 44.4337194962715, "learning_rate": 1.5267389267433034e-05, "loss": 1.2312, "step": 1138 }, { "epoch": 0.7257088244663906, "grad_norm": 12.649093137774878, "learning_rate": 1.5259206283733324e-05, "loss": 1.0977, "step": 1139 }, { "epoch": 0.7263459700541574, "grad_norm": 9.064049223182355, "learning_rate": 1.5251018760361605e-05, "loss": 1.297, "step": 1140 }, { "epoch": 0.7269831156419242, "grad_norm": 10.1931352489353, "learning_rate": 1.5242826706044535e-05, "loss": 1.1292, "step": 1141 }, { "epoch": 0.727620261229691, "grad_norm": 16.371459070038043, "learning_rate": 1.5234630129513616e-05, "loss": 1.4644, "step": 1142 }, { "epoch": 0.7282574068174578, "grad_norm": 26.255572002499044, "learning_rate": 1.5226429039505162e-05, "loss": 1.4563, "step": 1143 }, { "epoch": 0.7288945524052246, "grad_norm": 10.763511299737013, "learning_rate": 1.5218223444760296e-05, "loss": 1.1399, "step": 1144 }, { "epoch": 0.7295316979929914, "grad_norm": 30.864592234912415, "learning_rate": 1.5210013354024951e-05, "loss": 1.6042, "step": 1145 }, { "epoch": 0.7301688435807582, "grad_norm": 20.826117908525468, "learning_rate": 1.5201798776049837e-05, "loss": 1.7393, "step": 1146 }, { "epoch": 0.730805989168525, "grad_norm": 8.010297968760176, "learning_rate": 1.5193579719590463e-05, "loss": 1.2698, "step": 1147 }, { "epoch": 0.7314431347562919, "grad_norm": 10.029761640429411, "learning_rate": 1.5185356193407104e-05, "loss": 1.2745, "step": 1148 }, { "epoch": 0.7320802803440586, "grad_norm": 16.057788912128103, "learning_rate": 1.5177128206264795e-05, "loss": 1.7812, "step": 1149 }, { "epoch": 0.7327174259318254, "grad_norm": 12.306563013762327, "learning_rate": 1.5168895766933339e-05, "loss": 1.418, "step": 1150 }, { "epoch": 0.7333545715195923, "grad_norm": 18.48875291715215, "learning_rate": 1.5160658884187269e-05, "loss": 1.4187, "step": 1151 }, { "epoch": 0.733991717107359, "grad_norm": 19.60337373337071, "learning_rate": 1.5152417566805863e-05, "loss": 1.2738, "step": 1152 }, { "epoch": 0.7346288626951258, "grad_norm": 13.817639596691935, "learning_rate": 1.5144171823573122e-05, "loss": 1.4578, "step": 1153 }, { "epoch": 0.7352660082828927, "grad_norm": 9.473522856242877, "learning_rate": 1.5135921663277768e-05, "loss": 1.1077, "step": 1154 }, { "epoch": 0.7359031538706594, "grad_norm": 10.92222780663691, "learning_rate": 1.5127667094713222e-05, "loss": 1.2961, "step": 1155 }, { "epoch": 0.7365402994584263, "grad_norm": 16.853991508399762, "learning_rate": 1.511940812667762e-05, "loss": 0.929, "step": 1156 }, { "epoch": 0.7371774450461931, "grad_norm": 13.995932606580375, "learning_rate": 1.5111144767973769e-05, "loss": 1.6667, "step": 1157 }, { "epoch": 0.7378145906339598, "grad_norm": 12.380391933498034, "learning_rate": 1.510287702740917e-05, "loss": 1.5942, "step": 1158 }, { "epoch": 0.7384517362217267, "grad_norm": 30.488885400105723, "learning_rate": 1.5094604913795981e-05, "loss": 1.5557, "step": 1159 }, { "epoch": 0.7390888818094935, "grad_norm": 13.050000748677409, "learning_rate": 1.5086328435951039e-05, "loss": 1.1689, "step": 1160 }, { "epoch": 0.7397260273972602, "grad_norm": 14.404512016456195, "learning_rate": 1.5078047602695813e-05, "loss": 1.5023, "step": 1161 }, { "epoch": 0.7403631729850271, "grad_norm": 20.95785532819619, "learning_rate": 1.5069762422856427e-05, "loss": 1.2493, "step": 1162 }, { "epoch": 0.7410003185727939, "grad_norm": 46.90674527151889, "learning_rate": 1.5061472905263631e-05, "loss": 1.332, "step": 1163 }, { "epoch": 0.7416374641605606, "grad_norm": 29.243881701736775, "learning_rate": 1.5053179058752807e-05, "loss": 2.0811, "step": 1164 }, { "epoch": 0.7422746097483275, "grad_norm": 7.148930556779064, "learning_rate": 1.5044880892163944e-05, "loss": 0.8643, "step": 1165 }, { "epoch": 0.7429117553360943, "grad_norm": 21.97729272073257, "learning_rate": 1.5036578414341629e-05, "loss": 1.2655, "step": 1166 }, { "epoch": 0.7435489009238611, "grad_norm": 11.433254867316817, "learning_rate": 1.5028271634135065e-05, "loss": 1.5005, "step": 1167 }, { "epoch": 0.7441860465116279, "grad_norm": 15.476177944448278, "learning_rate": 1.5019960560398019e-05, "loss": 1.6433, "step": 1168 }, { "epoch": 0.7448231920993947, "grad_norm": 7.229098805992748, "learning_rate": 1.5011645201988847e-05, "loss": 0.9703, "step": 1169 }, { "epoch": 0.7454603376871615, "grad_norm": 13.525564496459749, "learning_rate": 1.5003325567770463e-05, "loss": 1.2874, "step": 1170 }, { "epoch": 0.7460974832749283, "grad_norm": 14.059529234309581, "learning_rate": 1.499500166661035e-05, "loss": 1.3461, "step": 1171 }, { "epoch": 0.7467346288626951, "grad_norm": 9.295195143066874, "learning_rate": 1.4986673507380528e-05, "loss": 1.0439, "step": 1172 }, { "epoch": 0.7473717744504619, "grad_norm": 9.510206322760963, "learning_rate": 1.497834109895756e-05, "loss": 1.1101, "step": 1173 }, { "epoch": 0.7480089200382287, "grad_norm": 10.488871771249778, "learning_rate": 1.497000445022254e-05, "loss": 1.2278, "step": 1174 }, { "epoch": 0.7486460656259956, "grad_norm": 6.4102520754650065, "learning_rate": 1.4961663570061073e-05, "loss": 1.084, "step": 1175 }, { "epoch": 0.7492832112137623, "grad_norm": 16.47578449780838, "learning_rate": 1.4953318467363287e-05, "loss": 1.5796, "step": 1176 }, { "epoch": 0.7499203568015291, "grad_norm": 11.334332369876373, "learning_rate": 1.49449691510238e-05, "loss": 1.6587, "step": 1177 }, { "epoch": 0.750557502389296, "grad_norm": 10.110962991937397, "learning_rate": 1.4936615629941731e-05, "loss": 1.6792, "step": 1178 }, { "epoch": 0.7511946479770628, "grad_norm": 15.828089719424433, "learning_rate": 1.4928257913020662e-05, "loss": 1.5245, "step": 1179 }, { "epoch": 0.7518317935648295, "grad_norm": 9.558438274145399, "learning_rate": 1.4919896009168676e-05, "loss": 1.034, "step": 1180 }, { "epoch": 0.7524689391525964, "grad_norm": 10.837642970335004, "learning_rate": 1.491152992729829e-05, "loss": 1.2908, "step": 1181 }, { "epoch": 0.7531060847403632, "grad_norm": 11.846285172903626, "learning_rate": 1.490315967632649e-05, "loss": 1.6289, "step": 1182 }, { "epoch": 0.75374323032813, "grad_norm": 9.430897951509195, "learning_rate": 1.48947852651747e-05, "loss": 1.2871, "step": 1183 }, { "epoch": 0.7543803759158968, "grad_norm": 11.626721456307095, "learning_rate": 1.4886406702768786e-05, "loss": 1.374, "step": 1184 }, { "epoch": 0.7550175215036636, "grad_norm": 12.316448648410974, "learning_rate": 1.4878023998039028e-05, "loss": 1.1536, "step": 1185 }, { "epoch": 0.7556546670914304, "grad_norm": 18.811445860266126, "learning_rate": 1.4869637159920129e-05, "loss": 1.7615, "step": 1186 }, { "epoch": 0.7562918126791972, "grad_norm": 12.74424253526663, "learning_rate": 1.4861246197351194e-05, "loss": 1.5186, "step": 1187 }, { "epoch": 0.756928958266964, "grad_norm": 8.434386257877023, "learning_rate": 1.485285111927572e-05, "loss": 1.0654, "step": 1188 }, { "epoch": 0.7575661038547308, "grad_norm": 11.0600504311186, "learning_rate": 1.4844451934641596e-05, "loss": 1.5396, "step": 1189 }, { "epoch": 0.7582032494424976, "grad_norm": 9.616840263893675, "learning_rate": 1.4836048652401096e-05, "loss": 1.3547, "step": 1190 }, { "epoch": 0.7588403950302645, "grad_norm": 12.40122379851735, "learning_rate": 1.4827641281510843e-05, "loss": 1.7104, "step": 1191 }, { "epoch": 0.7594775406180312, "grad_norm": 16.65778150387942, "learning_rate": 1.481922983093183e-05, "loss": 1.7944, "step": 1192 }, { "epoch": 0.760114686205798, "grad_norm": 18.019205547303518, "learning_rate": 1.4810814309629396e-05, "loss": 1.1772, "step": 1193 }, { "epoch": 0.7607518317935649, "grad_norm": 13.410780602859479, "learning_rate": 1.480239472657322e-05, "loss": 1.4917, "step": 1194 }, { "epoch": 0.7613889773813316, "grad_norm": 11.212385632603354, "learning_rate": 1.4793971090737307e-05, "loss": 0.7836, "step": 1195 }, { "epoch": 0.7620261229690984, "grad_norm": 9.007608281941408, "learning_rate": 1.4785543411099979e-05, "loss": 1.0559, "step": 1196 }, { "epoch": 0.7626632685568653, "grad_norm": 12.537416668025221, "learning_rate": 1.477711169664388e-05, "loss": 1.3801, "step": 1197 }, { "epoch": 0.763300414144632, "grad_norm": 22.609733174025873, "learning_rate": 1.4768675956355942e-05, "loss": 1.0881, "step": 1198 }, { "epoch": 0.7639375597323989, "grad_norm": 14.082461317366366, "learning_rate": 1.4760236199227396e-05, "loss": 1.0557, "step": 1199 }, { "epoch": 0.7645747053201657, "grad_norm": 34.10999276746803, "learning_rate": 1.4751792434253751e-05, "loss": 1.5974, "step": 1200 }, { "epoch": 0.7652118509079324, "grad_norm": 13.833568299812123, "learning_rate": 1.4743344670434784e-05, "loss": 1.509, "step": 1201 }, { "epoch": 0.7658489964956993, "grad_norm": 12.398785632083033, "learning_rate": 1.4734892916774543e-05, "loss": 1.4352, "step": 1202 }, { "epoch": 0.7664861420834661, "grad_norm": 9.29093141990189, "learning_rate": 1.4726437182281319e-05, "loss": 1.1816, "step": 1203 }, { "epoch": 0.7671232876712328, "grad_norm": 8.391324534934563, "learning_rate": 1.4717977475967655e-05, "loss": 1.2782, "step": 1204 }, { "epoch": 0.7677604332589997, "grad_norm": 11.575814943693764, "learning_rate": 1.470951380685032e-05, "loss": 1.615, "step": 1205 }, { "epoch": 0.7683975788467665, "grad_norm": 19.90440140448828, "learning_rate": 1.4701046183950312e-05, "loss": 1.4551, "step": 1206 }, { "epoch": 0.7690347244345332, "grad_norm": 9.801289027439504, "learning_rate": 1.4692574616292837e-05, "loss": 1.4067, "step": 1207 }, { "epoch": 0.7696718700223001, "grad_norm": 14.174056939677607, "learning_rate": 1.4684099112907314e-05, "loss": 1.0779, "step": 1208 }, { "epoch": 0.7703090156100669, "grad_norm": 17.644744638789092, "learning_rate": 1.467561968282735e-05, "loss": 1.6675, "step": 1209 }, { "epoch": 0.7709461611978337, "grad_norm": 9.913553120279998, "learning_rate": 1.4667136335090741e-05, "loss": 0.8268, "step": 1210 }, { "epoch": 0.7715833067856005, "grad_norm": 13.145552509098787, "learning_rate": 1.4658649078739454e-05, "loss": 1.267, "step": 1211 }, { "epoch": 0.7722204523733673, "grad_norm": 10.869155202710232, "learning_rate": 1.4650157922819626e-05, "loss": 1.498, "step": 1212 }, { "epoch": 0.7728575979611341, "grad_norm": 13.645843300461095, "learning_rate": 1.4641662876381555e-05, "loss": 1.1536, "step": 1213 }, { "epoch": 0.7734947435489009, "grad_norm": 18.712561722593914, "learning_rate": 1.4633163948479674e-05, "loss": 1.5361, "step": 1214 }, { "epoch": 0.7741318891366678, "grad_norm": 14.034957793692085, "learning_rate": 1.4624661148172564e-05, "loss": 1.3351, "step": 1215 }, { "epoch": 0.7747690347244345, "grad_norm": 8.579551292252768, "learning_rate": 1.4616154484522924e-05, "loss": 0.917, "step": 1216 }, { "epoch": 0.7754061803122013, "grad_norm": 12.81730810967236, "learning_rate": 1.4607643966597579e-05, "loss": 1.3445, "step": 1217 }, { "epoch": 0.7760433258999682, "grad_norm": 13.04849440721216, "learning_rate": 1.4599129603467453e-05, "loss": 1.491, "step": 1218 }, { "epoch": 0.776680471487735, "grad_norm": 8.0896337678192, "learning_rate": 1.4590611404207579e-05, "loss": 0.7924, "step": 1219 }, { "epoch": 0.7773176170755017, "grad_norm": 17.437949924316626, "learning_rate": 1.4582089377897075e-05, "loss": 1.6641, "step": 1220 }, { "epoch": 0.7779547626632686, "grad_norm": 21.1316171049501, "learning_rate": 1.4573563533619136e-05, "loss": 1.5947, "step": 1221 }, { "epoch": 0.7785919082510354, "grad_norm": 14.460804470217038, "learning_rate": 1.4565033880461023e-05, "loss": 1.7568, "step": 1222 }, { "epoch": 0.7792290538388021, "grad_norm": 7.5080737767056185, "learning_rate": 1.4556500427514065e-05, "loss": 0.9889, "step": 1223 }, { "epoch": 0.779866199426569, "grad_norm": 9.506806002207751, "learning_rate": 1.4547963183873637e-05, "loss": 1.2169, "step": 1224 }, { "epoch": 0.7805033450143358, "grad_norm": 33.524102331250035, "learning_rate": 1.4539422158639148e-05, "loss": 1.395, "step": 1225 }, { "epoch": 0.7811404906021026, "grad_norm": 16.878216168450294, "learning_rate": 1.4530877360914046e-05, "loss": 1.4709, "step": 1226 }, { "epoch": 0.7817776361898694, "grad_norm": 14.434939718256642, "learning_rate": 1.4522328799805805e-05, "loss": 1.5815, "step": 1227 }, { "epoch": 0.7824147817776362, "grad_norm": 18.66828498128432, "learning_rate": 1.4513776484425897e-05, "loss": 1.0913, "step": 1228 }, { "epoch": 0.783051927365403, "grad_norm": 10.21767965909488, "learning_rate": 1.4505220423889801e-05, "loss": 0.959, "step": 1229 }, { "epoch": 0.7836890729531698, "grad_norm": 16.261515706430234, "learning_rate": 1.449666062731699e-05, "loss": 1.3503, "step": 1230 }, { "epoch": 0.7843262185409366, "grad_norm": 27.180422546810888, "learning_rate": 1.4488097103830914e-05, "loss": 1.6174, "step": 1231 }, { "epoch": 0.7849633641287034, "grad_norm": 16.629409911901092, "learning_rate": 1.4479529862559e-05, "loss": 1.3947, "step": 1232 }, { "epoch": 0.7856005097164702, "grad_norm": 37.54209374140125, "learning_rate": 1.4470958912632636e-05, "loss": 1.5063, "step": 1233 }, { "epoch": 0.7862376553042371, "grad_norm": 12.183040207371587, "learning_rate": 1.4462384263187165e-05, "loss": 1.4033, "step": 1234 }, { "epoch": 0.7868748008920038, "grad_norm": 8.390112334635674, "learning_rate": 1.445380592336187e-05, "loss": 1.2621, "step": 1235 }, { "epoch": 0.7875119464797706, "grad_norm": 14.386326914171853, "learning_rate": 1.444522390229997e-05, "loss": 1.6018, "step": 1236 }, { "epoch": 0.7881490920675375, "grad_norm": 13.234250089953363, "learning_rate": 1.4436638209148598e-05, "loss": 1.229, "step": 1237 }, { "epoch": 0.7887862376553042, "grad_norm": 16.270087016405963, "learning_rate": 1.4428048853058822e-05, "loss": 1.3113, "step": 1238 }, { "epoch": 0.789423383243071, "grad_norm": 11.557965366914308, "learning_rate": 1.441945584318559e-05, "loss": 0.8928, "step": 1239 }, { "epoch": 0.7900605288308379, "grad_norm": 14.551152526000774, "learning_rate": 1.441085918868776e-05, "loss": 1.5977, "step": 1240 }, { "epoch": 0.7906976744186046, "grad_norm": 14.946315679885245, "learning_rate": 1.4402258898728072e-05, "loss": 1.3154, "step": 1241 }, { "epoch": 0.7913348200063715, "grad_norm": 30.525422437388094, "learning_rate": 1.4393654982473133e-05, "loss": 1.9502, "step": 1242 }, { "epoch": 0.7919719655941383, "grad_norm": 21.986803336300746, "learning_rate": 1.438504744909343e-05, "loss": 1.1816, "step": 1243 }, { "epoch": 0.792609111181905, "grad_norm": 9.593385386584627, "learning_rate": 1.437643630776329e-05, "loss": 1.2869, "step": 1244 }, { "epoch": 0.7932462567696719, "grad_norm": 18.710931428778732, "learning_rate": 1.436782156766089e-05, "loss": 1.6814, "step": 1245 }, { "epoch": 0.7938834023574387, "grad_norm": 18.13991374508064, "learning_rate": 1.4359203237968251e-05, "loss": 1.8857, "step": 1246 }, { "epoch": 0.7945205479452054, "grad_norm": 14.025622364506567, "learning_rate": 1.4350581327871206e-05, "loss": 1.7358, "step": 1247 }, { "epoch": 0.7951576935329723, "grad_norm": 11.710919664304352, "learning_rate": 1.4341955846559423e-05, "loss": 1.6707, "step": 1248 }, { "epoch": 0.7957948391207391, "grad_norm": 12.01161734813932, "learning_rate": 1.4333326803226355e-05, "loss": 1.3793, "step": 1249 }, { "epoch": 0.7964319847085058, "grad_norm": 17.886641236532228, "learning_rate": 1.4324694207069265e-05, "loss": 1.4312, "step": 1250 }, { "epoch": 0.7970691302962727, "grad_norm": 11.572857753396455, "learning_rate": 1.43160580672892e-05, "loss": 1.0122, "step": 1251 }, { "epoch": 0.7977062758840395, "grad_norm": 12.891323005285107, "learning_rate": 1.4307418393090981e-05, "loss": 1.2678, "step": 1252 }, { "epoch": 0.7983434214718063, "grad_norm": 17.73612598854489, "learning_rate": 1.4298775193683205e-05, "loss": 1.0668, "step": 1253 }, { "epoch": 0.7989805670595731, "grad_norm": 11.557332196991679, "learning_rate": 1.4290128478278213e-05, "loss": 1.6448, "step": 1254 }, { "epoch": 0.7996177126473399, "grad_norm": 13.081145860809627, "learning_rate": 1.4281478256092106e-05, "loss": 1.0801, "step": 1255 }, { "epoch": 0.8002548582351067, "grad_norm": 10.486039759717409, "learning_rate": 1.4272824536344712e-05, "loss": 1.3812, "step": 1256 }, { "epoch": 0.8008920038228735, "grad_norm": 6.836603413019294, "learning_rate": 1.4264167328259597e-05, "loss": 1.1096, "step": 1257 }, { "epoch": 0.8015291494106404, "grad_norm": 13.07880982310632, "learning_rate": 1.4255506641064036e-05, "loss": 1.5383, "step": 1258 }, { "epoch": 0.8021662949984071, "grad_norm": 5.1494761529504665, "learning_rate": 1.4246842483989019e-05, "loss": 0.7633, "step": 1259 }, { "epoch": 0.8028034405861739, "grad_norm": 12.697192118380633, "learning_rate": 1.423817486626923e-05, "loss": 1.7405, "step": 1260 }, { "epoch": 0.8034405861739408, "grad_norm": 7.8513270308056695, "learning_rate": 1.4229503797143045e-05, "loss": 1.3167, "step": 1261 }, { "epoch": 0.8040777317617076, "grad_norm": 7.886969752889762, "learning_rate": 1.4220829285852514e-05, "loss": 1.0425, "step": 1262 }, { "epoch": 0.8047148773494743, "grad_norm": 10.178404441400074, "learning_rate": 1.4212151341643364e-05, "loss": 1.3226, "step": 1263 }, { "epoch": 0.8053520229372412, "grad_norm": 4.76620082989673, "learning_rate": 1.4203469973764969e-05, "loss": 0.7178, "step": 1264 }, { "epoch": 0.805989168525008, "grad_norm": 12.407216801266351, "learning_rate": 1.4194785191470365e-05, "loss": 1.0225, "step": 1265 }, { "epoch": 0.8066263141127747, "grad_norm": 13.639889310652224, "learning_rate": 1.4186097004016217e-05, "loss": 1.7183, "step": 1266 }, { "epoch": 0.8072634597005416, "grad_norm": 13.124825420521425, "learning_rate": 1.4177405420662826e-05, "loss": 1.3776, "step": 1267 }, { "epoch": 0.8079006052883084, "grad_norm": 11.916791928770326, "learning_rate": 1.4168710450674108e-05, "loss": 1.4401, "step": 1268 }, { "epoch": 0.8085377508760752, "grad_norm": 9.468238446035803, "learning_rate": 1.416001210331759e-05, "loss": 1.387, "step": 1269 }, { "epoch": 0.809174896463842, "grad_norm": 21.06467145893861, "learning_rate": 1.4151310387864405e-05, "loss": 1.3385, "step": 1270 }, { "epoch": 0.8098120420516088, "grad_norm": 6.363271011504864, "learning_rate": 1.4142605313589264e-05, "loss": 0.7174, "step": 1271 }, { "epoch": 0.8104491876393756, "grad_norm": 10.868418867185815, "learning_rate": 1.4133896889770467e-05, "loss": 1.0919, "step": 1272 }, { "epoch": 0.8110863332271424, "grad_norm": 7.76993199964701, "learning_rate": 1.412518512568988e-05, "loss": 0.8414, "step": 1273 }, { "epoch": 0.8117234788149092, "grad_norm": 10.590349772853074, "learning_rate": 1.4116470030632929e-05, "loss": 1.0212, "step": 1274 }, { "epoch": 0.812360624402676, "grad_norm": 10.298812869898471, "learning_rate": 1.4107751613888599e-05, "loss": 1.5021, "step": 1275 }, { "epoch": 0.8129977699904428, "grad_norm": 9.135585313413454, "learning_rate": 1.4099029884749398e-05, "loss": 0.8893, "step": 1276 }, { "epoch": 0.8136349155782097, "grad_norm": 10.172503599818826, "learning_rate": 1.4090304852511386e-05, "loss": 1.1677, "step": 1277 }, { "epoch": 0.8142720611659764, "grad_norm": 10.35441258642809, "learning_rate": 1.4081576526474123e-05, "loss": 1.163, "step": 1278 }, { "epoch": 0.8149092067537432, "grad_norm": 13.540810148117341, "learning_rate": 1.4072844915940696e-05, "loss": 1.3663, "step": 1279 }, { "epoch": 0.8155463523415101, "grad_norm": 10.913993390687551, "learning_rate": 1.4064110030217684e-05, "loss": 1.079, "step": 1280 }, { "epoch": 0.8161834979292768, "grad_norm": 25.561975044242395, "learning_rate": 1.4055371878615153e-05, "loss": 1.6718, "step": 1281 }, { "epoch": 0.8168206435170436, "grad_norm": 10.533948541523529, "learning_rate": 1.4046630470446663e-05, "loss": 1.1237, "step": 1282 }, { "epoch": 0.8174577891048105, "grad_norm": 4.886574733985856, "learning_rate": 1.4037885815029236e-05, "loss": 0.8188, "step": 1283 }, { "epoch": 0.8180949346925772, "grad_norm": 12.822838354337366, "learning_rate": 1.4029137921683356e-05, "loss": 1.4412, "step": 1284 }, { "epoch": 0.818732080280344, "grad_norm": 7.29891730693425, "learning_rate": 1.402038679973296e-05, "loss": 0.8319, "step": 1285 }, { "epoch": 0.8193692258681109, "grad_norm": 8.073355204774396, "learning_rate": 1.401163245850542e-05, "loss": 1.1199, "step": 1286 }, { "epoch": 0.8200063714558776, "grad_norm": 10.916070814791361, "learning_rate": 1.4002874907331552e-05, "loss": 1.4792, "step": 1287 }, { "epoch": 0.8206435170436445, "grad_norm": 8.498791309398747, "learning_rate": 1.399411415554558e-05, "loss": 1.3813, "step": 1288 }, { "epoch": 0.8212806626314113, "grad_norm": 10.541746136233488, "learning_rate": 1.3985350212485147e-05, "loss": 1.0624, "step": 1289 }, { "epoch": 0.821917808219178, "grad_norm": 7.605906142276422, "learning_rate": 1.39765830874913e-05, "loss": 1.2975, "step": 1290 }, { "epoch": 0.8225549538069449, "grad_norm": 10.741594551306948, "learning_rate": 1.3967812789908467e-05, "loss": 0.9452, "step": 1291 }, { "epoch": 0.8231920993947117, "grad_norm": 11.609802576899803, "learning_rate": 1.3959039329084465e-05, "loss": 1.2593, "step": 1292 }, { "epoch": 0.8238292449824784, "grad_norm": 7.160432794781688, "learning_rate": 1.3950262714370481e-05, "loss": 1.0988, "step": 1293 }, { "epoch": 0.8244663905702453, "grad_norm": 15.770955712973194, "learning_rate": 1.3941482955121066e-05, "loss": 1.8071, "step": 1294 }, { "epoch": 0.8251035361580121, "grad_norm": 8.504518228251529, "learning_rate": 1.393270006069412e-05, "loss": 1.2515, "step": 1295 }, { "epoch": 0.8257406817457789, "grad_norm": 10.729038053497188, "learning_rate": 1.3923914040450877e-05, "loss": 1.0088, "step": 1296 }, { "epoch": 0.8263778273335457, "grad_norm": 9.613745671563315, "learning_rate": 1.391512490375592e-05, "loss": 0.9758, "step": 1297 }, { "epoch": 0.8270149729213125, "grad_norm": 16.35617223529865, "learning_rate": 1.3906332659977142e-05, "loss": 1.3552, "step": 1298 }, { "epoch": 0.8276521185090793, "grad_norm": 11.287441390772464, "learning_rate": 1.3897537318485748e-05, "loss": 1.4524, "step": 1299 }, { "epoch": 0.8282892640968461, "grad_norm": 9.76132504149058, "learning_rate": 1.3888738888656247e-05, "loss": 1.3978, "step": 1300 }, { "epoch": 0.828926409684613, "grad_norm": 71.43082043168982, "learning_rate": 1.387993737986644e-05, "loss": 1.5786, "step": 1301 }, { "epoch": 0.8295635552723798, "grad_norm": 10.366835428243403, "learning_rate": 1.3871132801497404e-05, "loss": 1.3369, "step": 1302 }, { "epoch": 0.8302007008601465, "grad_norm": 13.915670029329675, "learning_rate": 1.38623251629335e-05, "loss": 1.5024, "step": 1303 }, { "epoch": 0.8308378464479134, "grad_norm": 19.56479061774396, "learning_rate": 1.3853514473562342e-05, "loss": 2.0005, "step": 1304 }, { "epoch": 0.8314749920356802, "grad_norm": 8.086160924991217, "learning_rate": 1.3844700742774796e-05, "loss": 1.0304, "step": 1305 }, { "epoch": 0.8321121376234469, "grad_norm": 11.445007595056655, "learning_rate": 1.3835883979964964e-05, "loss": 1.4434, "step": 1306 }, { "epoch": 0.8327492832112138, "grad_norm": 17.4708278780384, "learning_rate": 1.3827064194530198e-05, "loss": 1.364, "step": 1307 }, { "epoch": 0.8333864287989806, "grad_norm": 8.821495558014208, "learning_rate": 1.3818241395871053e-05, "loss": 1.3301, "step": 1308 }, { "epoch": 0.8340235743867473, "grad_norm": 5.833283126525502, "learning_rate": 1.3809415593391303e-05, "loss": 0.8774, "step": 1309 }, { "epoch": 0.8346607199745142, "grad_norm": 7.9548580687303225, "learning_rate": 1.3800586796497923e-05, "loss": 1.2496, "step": 1310 }, { "epoch": 0.835297865562281, "grad_norm": 10.561912861117033, "learning_rate": 1.3791755014601085e-05, "loss": 1.3899, "step": 1311 }, { "epoch": 0.8359350111500478, "grad_norm": 7.50760517070452, "learning_rate": 1.3782920257114133e-05, "loss": 1.3931, "step": 1312 }, { "epoch": 0.8365721567378146, "grad_norm": 15.647942216354185, "learning_rate": 1.377408253345359e-05, "loss": 1.2522, "step": 1313 }, { "epoch": 0.8372093023255814, "grad_norm": 11.576924015202007, "learning_rate": 1.3765241853039133e-05, "loss": 1.2858, "step": 1314 }, { "epoch": 0.8378464479133482, "grad_norm": 14.32438777980592, "learning_rate": 1.3756398225293602e-05, "loss": 1.6624, "step": 1315 }, { "epoch": 0.838483593501115, "grad_norm": 11.35723562497473, "learning_rate": 1.3747551659642962e-05, "loss": 1.6897, "step": 1316 }, { "epoch": 0.8391207390888819, "grad_norm": 14.707978136341396, "learning_rate": 1.3738702165516325e-05, "loss": 1.5789, "step": 1317 }, { "epoch": 0.8397578846766486, "grad_norm": 44.415384110565654, "learning_rate": 1.3729849752345918e-05, "loss": 1.6677, "step": 1318 }, { "epoch": 0.8403950302644154, "grad_norm": 12.455126507070874, "learning_rate": 1.3720994429567074e-05, "loss": 1.6157, "step": 1319 }, { "epoch": 0.8410321758521823, "grad_norm": 12.699251380065283, "learning_rate": 1.3712136206618243e-05, "loss": 1.25, "step": 1320 }, { "epoch": 0.841669321439949, "grad_norm": 10.93334169163756, "learning_rate": 1.3703275092940946e-05, "loss": 1.7998, "step": 1321 }, { "epoch": 0.8423064670277158, "grad_norm": 9.848436731365732, "learning_rate": 1.3694411097979798e-05, "loss": 1.3508, "step": 1322 }, { "epoch": 0.8429436126154827, "grad_norm": 7.71098817638587, "learning_rate": 1.3685544231182481e-05, "loss": 1.2351, "step": 1323 }, { "epoch": 0.8435807582032494, "grad_norm": 11.281869887170771, "learning_rate": 1.3676674501999742e-05, "loss": 1.4417, "step": 1324 }, { "epoch": 0.8442179037910162, "grad_norm": 15.76388283138257, "learning_rate": 1.3667801919885374e-05, "loss": 1.4624, "step": 1325 }, { "epoch": 0.8448550493787831, "grad_norm": 11.152364854885205, "learning_rate": 1.365892649429621e-05, "loss": 1.3503, "step": 1326 }, { "epoch": 0.8454921949665498, "grad_norm": 12.875609325730832, "learning_rate": 1.3650048234692117e-05, "loss": 1.3081, "step": 1327 }, { "epoch": 0.8461293405543167, "grad_norm": 14.16521110376584, "learning_rate": 1.3641167150535985e-05, "loss": 1.2218, "step": 1328 }, { "epoch": 0.8467664861420835, "grad_norm": 12.675532756679434, "learning_rate": 1.3632283251293705e-05, "loss": 1.5918, "step": 1329 }, { "epoch": 0.8474036317298502, "grad_norm": 23.228638897703657, "learning_rate": 1.3623396546434183e-05, "loss": 1.3491, "step": 1330 }, { "epoch": 0.8480407773176171, "grad_norm": 10.846894151201456, "learning_rate": 1.36145070454293e-05, "loss": 1.3575, "step": 1331 }, { "epoch": 0.8486779229053839, "grad_norm": 11.766091923047364, "learning_rate": 1.3605614757753924e-05, "loss": 1.1938, "step": 1332 }, { "epoch": 0.8493150684931506, "grad_norm": 12.193172088906145, "learning_rate": 1.3596719692885902e-05, "loss": 1.3918, "step": 1333 }, { "epoch": 0.8499522140809175, "grad_norm": 16.129722433809622, "learning_rate": 1.3587821860306026e-05, "loss": 1.4443, "step": 1334 }, { "epoch": 0.8505893596686843, "grad_norm": 8.990881839026862, "learning_rate": 1.3578921269498046e-05, "loss": 1.3635, "step": 1335 }, { "epoch": 0.851226505256451, "grad_norm": 14.436716568195566, "learning_rate": 1.3570017929948649e-05, "loss": 1.2189, "step": 1336 }, { "epoch": 0.8518636508442179, "grad_norm": 13.528595552569586, "learning_rate": 1.3561111851147452e-05, "loss": 1.6355, "step": 1337 }, { "epoch": 0.8525007964319847, "grad_norm": 10.004867796584286, "learning_rate": 1.3552203042586997e-05, "loss": 1.2148, "step": 1338 }, { "epoch": 0.8531379420197515, "grad_norm": 4.188933164961296, "learning_rate": 1.3543291513762728e-05, "loss": 0.5929, "step": 1339 }, { "epoch": 0.8537750876075183, "grad_norm": 9.706513835529034, "learning_rate": 1.3534377274172999e-05, "loss": 1.1354, "step": 1340 }, { "epoch": 0.8544122331952851, "grad_norm": 10.488181362147115, "learning_rate": 1.3525460333319035e-05, "loss": 1.6494, "step": 1341 }, { "epoch": 0.8550493787830519, "grad_norm": 12.733177103722616, "learning_rate": 1.3516540700704956e-05, "loss": 1.5125, "step": 1342 }, { "epoch": 0.8556865243708187, "grad_norm": 6.956417695442206, "learning_rate": 1.3507618385837751e-05, "loss": 1.1884, "step": 1343 }, { "epoch": 0.8563236699585856, "grad_norm": 9.511321261350686, "learning_rate": 1.3498693398227254e-05, "loss": 1.134, "step": 1344 }, { "epoch": 0.8569608155463524, "grad_norm": 9.519041327794527, "learning_rate": 1.348976574738616e-05, "loss": 1.3634, "step": 1345 }, { "epoch": 0.8575979611341191, "grad_norm": 14.071108451695897, "learning_rate": 1.3480835442830003e-05, "loss": 1.3918, "step": 1346 }, { "epoch": 0.858235106721886, "grad_norm": 10.729654605932277, "learning_rate": 1.347190249407714e-05, "loss": 1.2976, "step": 1347 }, { "epoch": 0.8588722523096528, "grad_norm": 12.519608540995973, "learning_rate": 1.3462966910648749e-05, "loss": 1.5555, "step": 1348 }, { "epoch": 0.8595093978974195, "grad_norm": 12.277726276564561, "learning_rate": 1.3454028702068808e-05, "loss": 1.3171, "step": 1349 }, { "epoch": 0.8601465434851864, "grad_norm": 5.057744274231072, "learning_rate": 1.3445087877864112e-05, "loss": 0.7473, "step": 1350 }, { "epoch": 0.8607836890729532, "grad_norm": 31.799900476057992, "learning_rate": 1.3436144447564224e-05, "loss": 1.3081, "step": 1351 }, { "epoch": 0.86142083466072, "grad_norm": 12.57757966659432, "learning_rate": 1.342719842070149e-05, "loss": 1.3163, "step": 1352 }, { "epoch": 0.8620579802484868, "grad_norm": 13.78812302803317, "learning_rate": 1.341824980681104e-05, "loss": 1.4009, "step": 1353 }, { "epoch": 0.8626951258362536, "grad_norm": 17.43983457121057, "learning_rate": 1.3409298615430734e-05, "loss": 1.1957, "step": 1354 }, { "epoch": 0.8633322714240204, "grad_norm": 11.122287712488106, "learning_rate": 1.3400344856101202e-05, "loss": 1.3112, "step": 1355 }, { "epoch": 0.8639694170117872, "grad_norm": 10.422955004497961, "learning_rate": 1.3391388538365795e-05, "loss": 1.0634, "step": 1356 }, { "epoch": 0.864606562599554, "grad_norm": 10.220790825067331, "learning_rate": 1.3382429671770602e-05, "loss": 1.3156, "step": 1357 }, { "epoch": 0.8652437081873208, "grad_norm": 15.747475614874409, "learning_rate": 1.3373468265864424e-05, "loss": 1.7014, "step": 1358 }, { "epoch": 0.8658808537750876, "grad_norm": 13.170791315847026, "learning_rate": 1.336450433019877e-05, "loss": 1.1744, "step": 1359 }, { "epoch": 0.8665179993628545, "grad_norm": 7.644214315508008, "learning_rate": 1.3355537874327844e-05, "loss": 0.9462, "step": 1360 }, { "epoch": 0.8671551449506212, "grad_norm": 15.02398327703111, "learning_rate": 1.3346568907808536e-05, "loss": 1.4031, "step": 1361 }, { "epoch": 0.867792290538388, "grad_norm": 11.253396719166734, "learning_rate": 1.3337597440200413e-05, "loss": 1.2889, "step": 1362 }, { "epoch": 0.8684294361261549, "grad_norm": 28.74063354675122, "learning_rate": 1.3328623481065707e-05, "loss": 1.7856, "step": 1363 }, { "epoch": 0.8690665817139216, "grad_norm": 10.202650901040611, "learning_rate": 1.331964703996931e-05, "loss": 1.217, "step": 1364 }, { "epoch": 0.8697037273016884, "grad_norm": 28.805750067234126, "learning_rate": 1.3310668126478745e-05, "loss": 0.8665, "step": 1365 }, { "epoch": 0.8703408728894553, "grad_norm": 10.941321497434368, "learning_rate": 1.330168675016419e-05, "loss": 1.2437, "step": 1366 }, { "epoch": 0.870978018477222, "grad_norm": 12.558511701255375, "learning_rate": 1.3292702920598437e-05, "loss": 1.5645, "step": 1367 }, { "epoch": 0.8716151640649888, "grad_norm": 12.684006547639699, "learning_rate": 1.3283716647356895e-05, "loss": 1.5322, "step": 1368 }, { "epoch": 0.8722523096527557, "grad_norm": 10.437798127591016, "learning_rate": 1.3274727940017574e-05, "loss": 1.0267, "step": 1369 }, { "epoch": 0.8728894552405224, "grad_norm": 11.22207808275918, "learning_rate": 1.326573680816108e-05, "loss": 1.0308, "step": 1370 }, { "epoch": 0.8735266008282893, "grad_norm": 10.79168322962705, "learning_rate": 1.325674326137061e-05, "loss": 1.033, "step": 1371 }, { "epoch": 0.8741637464160561, "grad_norm": 12.93389392946053, "learning_rate": 1.324774730923193e-05, "loss": 1.2356, "step": 1372 }, { "epoch": 0.8748008920038228, "grad_norm": 8.561885493811154, "learning_rate": 1.323874896133336e-05, "loss": 1.1736, "step": 1373 }, { "epoch": 0.8754380375915897, "grad_norm": 9.145694911419938, "learning_rate": 1.322974822726579e-05, "loss": 1.2688, "step": 1374 }, { "epoch": 0.8760751831793565, "grad_norm": 15.996950419174647, "learning_rate": 1.3220745116622649e-05, "loss": 1.0916, "step": 1375 }, { "epoch": 0.8767123287671232, "grad_norm": 12.299747080149645, "learning_rate": 1.3211739638999888e-05, "loss": 1.1041, "step": 1376 }, { "epoch": 0.8773494743548901, "grad_norm": 23.941921402487434, "learning_rate": 1.3202731803995996e-05, "loss": 1.2468, "step": 1377 }, { "epoch": 0.8779866199426569, "grad_norm": 28.628997828020953, "learning_rate": 1.3193721621211964e-05, "loss": 1.5007, "step": 1378 }, { "epoch": 0.8786237655304237, "grad_norm": 26.20825363744769, "learning_rate": 1.3184709100251285e-05, "loss": 1.4692, "step": 1379 }, { "epoch": 0.8792609111181905, "grad_norm": 12.231054606095364, "learning_rate": 1.3175694250719955e-05, "loss": 1.2311, "step": 1380 }, { "epoch": 0.8798980567059573, "grad_norm": 21.228730723307745, "learning_rate": 1.3166677082226446e-05, "loss": 0.9648, "step": 1381 }, { "epoch": 0.8805352022937241, "grad_norm": 8.454785146239349, "learning_rate": 1.3157657604381693e-05, "loss": 0.8867, "step": 1382 }, { "epoch": 0.8811723478814909, "grad_norm": 8.622824331596787, "learning_rate": 1.3148635826799108e-05, "loss": 0.8558, "step": 1383 }, { "epoch": 0.8818094934692577, "grad_norm": 14.264541142585644, "learning_rate": 1.3139611759094544e-05, "loss": 1.411, "step": 1384 }, { "epoch": 0.8824466390570246, "grad_norm": 10.687193491304887, "learning_rate": 1.3130585410886292e-05, "loss": 1.2878, "step": 1385 }, { "epoch": 0.8830837846447913, "grad_norm": 16.39473200711394, "learning_rate": 1.3121556791795082e-05, "loss": 1.6509, "step": 1386 }, { "epoch": 0.8837209302325582, "grad_norm": 15.28435340565428, "learning_rate": 1.3112525911444065e-05, "loss": 1.4504, "step": 1387 }, { "epoch": 0.884358075820325, "grad_norm": 8.486687897835353, "learning_rate": 1.3103492779458796e-05, "loss": 0.7864, "step": 1388 }, { "epoch": 0.8849952214080917, "grad_norm": 11.788359345927613, "learning_rate": 1.3094457405467229e-05, "loss": 1.5391, "step": 1389 }, { "epoch": 0.8856323669958586, "grad_norm": 8.030165962464604, "learning_rate": 1.3085419799099717e-05, "loss": 0.8912, "step": 1390 }, { "epoch": 0.8862695125836254, "grad_norm": 11.812909622192814, "learning_rate": 1.3076379969988982e-05, "loss": 0.9908, "step": 1391 }, { "epoch": 0.8869066581713921, "grad_norm": 15.833970424309253, "learning_rate": 1.3067337927770121e-05, "loss": 1.2273, "step": 1392 }, { "epoch": 0.887543803759159, "grad_norm": 10.524985529759833, "learning_rate": 1.3058293682080585e-05, "loss": 0.9202, "step": 1393 }, { "epoch": 0.8881809493469258, "grad_norm": 7.557626782619985, "learning_rate": 1.304924724256018e-05, "loss": 0.8778, "step": 1394 }, { "epoch": 0.8888180949346925, "grad_norm": 7.626485245120705, "learning_rate": 1.3040198618851054e-05, "loss": 0.986, "step": 1395 }, { "epoch": 0.8894552405224594, "grad_norm": 15.294857509285064, "learning_rate": 1.3031147820597662e-05, "loss": 1.7427, "step": 1396 }, { "epoch": 0.8900923861102262, "grad_norm": 19.19366247484842, "learning_rate": 1.3022094857446803e-05, "loss": 1.1166, "step": 1397 }, { "epoch": 0.890729531697993, "grad_norm": 10.966759162803388, "learning_rate": 1.3013039739047567e-05, "loss": 1.2286, "step": 1398 }, { "epoch": 0.8913666772857598, "grad_norm": 13.154780241970368, "learning_rate": 1.300398247505134e-05, "loss": 1.3348, "step": 1399 }, { "epoch": 0.8920038228735266, "grad_norm": 15.967742732732864, "learning_rate": 1.2994923075111806e-05, "loss": 1.1523, "step": 1400 }, { "epoch": 0.8926409684612934, "grad_norm": 11.908033239236584, "learning_rate": 1.2985861548884926e-05, "loss": 1.1804, "step": 1401 }, { "epoch": 0.8932781140490602, "grad_norm": 6.096160031222386, "learning_rate": 1.2976797906028908e-05, "loss": 0.928, "step": 1402 }, { "epoch": 0.893915259636827, "grad_norm": 12.119654580223008, "learning_rate": 1.2967732156204243e-05, "loss": 1.2834, "step": 1403 }, { "epoch": 0.8945524052245938, "grad_norm": 10.954117159921479, "learning_rate": 1.2958664309073645e-05, "loss": 1.1763, "step": 1404 }, { "epoch": 0.8951895508123606, "grad_norm": 9.660452261355715, "learning_rate": 1.2949594374302075e-05, "loss": 1.072, "step": 1405 }, { "epoch": 0.8958266964001275, "grad_norm": 13.950098107360317, "learning_rate": 1.2940522361556713e-05, "loss": 1.1016, "step": 1406 }, { "epoch": 0.8964638419878942, "grad_norm": 11.59863569715782, "learning_rate": 1.2931448280506962e-05, "loss": 1.4583, "step": 1407 }, { "epoch": 0.897100987575661, "grad_norm": 8.222426178998989, "learning_rate": 1.2922372140824427e-05, "loss": 1.1221, "step": 1408 }, { "epoch": 0.8977381331634279, "grad_norm": 12.773213962563187, "learning_rate": 1.2913293952182899e-05, "loss": 1.1333, "step": 1409 }, { "epoch": 0.8983752787511946, "grad_norm": 19.68479728791302, "learning_rate": 1.290421372425836e-05, "loss": 1.3369, "step": 1410 }, { "epoch": 0.8990124243389614, "grad_norm": 26.882014084591933, "learning_rate": 1.2895131466728972e-05, "loss": 1.0121, "step": 1411 }, { "epoch": 0.8996495699267283, "grad_norm": 7.7306500774968185, "learning_rate": 1.2886047189275042e-05, "loss": 0.988, "step": 1412 }, { "epoch": 0.900286715514495, "grad_norm": 8.598391058819416, "learning_rate": 1.2876960901579048e-05, "loss": 1.2914, "step": 1413 }, { "epoch": 0.9009238611022619, "grad_norm": 17.078234706263906, "learning_rate": 1.2867872613325601e-05, "loss": 1.5918, "step": 1414 }, { "epoch": 0.9015610066900287, "grad_norm": 16.392722183921748, "learning_rate": 1.2858782334201447e-05, "loss": 1.6753, "step": 1415 }, { "epoch": 0.9021981522777954, "grad_norm": 15.628862479443772, "learning_rate": 1.2849690073895452e-05, "loss": 1.3389, "step": 1416 }, { "epoch": 0.9028352978655623, "grad_norm": 13.47431621259894, "learning_rate": 1.2840595842098597e-05, "loss": 1.6484, "step": 1417 }, { "epoch": 0.9034724434533291, "grad_norm": 12.708618872541829, "learning_rate": 1.2831499648503964e-05, "loss": 1.2798, "step": 1418 }, { "epoch": 0.9041095890410958, "grad_norm": 22.083697098043977, "learning_rate": 1.282240150280672e-05, "loss": 1.6206, "step": 1419 }, { "epoch": 0.9047467346288627, "grad_norm": 11.264645275187421, "learning_rate": 1.2813301414704125e-05, "loss": 1.5415, "step": 1420 }, { "epoch": 0.9053838802166295, "grad_norm": 10.474938719549858, "learning_rate": 1.2804199393895488e-05, "loss": 1.6963, "step": 1421 }, { "epoch": 0.9060210258043963, "grad_norm": 13.293058299394763, "learning_rate": 1.2795095450082204e-05, "loss": 1.439, "step": 1422 }, { "epoch": 0.9066581713921631, "grad_norm": 19.74715280759693, "learning_rate": 1.27859895929677e-05, "loss": 1.7554, "step": 1423 }, { "epoch": 0.9072953169799299, "grad_norm": 11.04644675238145, "learning_rate": 1.2776881832257451e-05, "loss": 1.0232, "step": 1424 }, { "epoch": 0.9079324625676968, "grad_norm": 11.388766535802617, "learning_rate": 1.2767772177658955e-05, "loss": 1.3186, "step": 1425 }, { "epoch": 0.9085696081554635, "grad_norm": 13.548724780976228, "learning_rate": 1.2758660638881728e-05, "loss": 1.4695, "step": 1426 }, { "epoch": 0.9092067537432303, "grad_norm": 10.47227753198024, "learning_rate": 1.2749547225637298e-05, "loss": 1.361, "step": 1427 }, { "epoch": 0.9098438993309972, "grad_norm": 6.963814689394079, "learning_rate": 1.2740431947639193e-05, "loss": 0.8407, "step": 1428 }, { "epoch": 0.9104810449187639, "grad_norm": 10.578064649052052, "learning_rate": 1.2731314814602925e-05, "loss": 1.6313, "step": 1429 }, { "epoch": 0.9111181905065308, "grad_norm": 6.976521805762002, "learning_rate": 1.2722195836245982e-05, "loss": 0.8491, "step": 1430 }, { "epoch": 0.9117553360942976, "grad_norm": 54.60887717892646, "learning_rate": 1.2713075022287824e-05, "loss": 1.9531, "step": 1431 }, { "epoch": 0.9123924816820643, "grad_norm": 10.32821260407392, "learning_rate": 1.270395238244986e-05, "loss": 1.6877, "step": 1432 }, { "epoch": 0.9130296272698312, "grad_norm": 10.618483199697323, "learning_rate": 1.2694827926455454e-05, "loss": 0.9467, "step": 1433 }, { "epoch": 0.913666772857598, "grad_norm": 11.721738487033385, "learning_rate": 1.2685701664029897e-05, "loss": 1.8291, "step": 1434 }, { "epoch": 0.9143039184453647, "grad_norm": 9.700947010519045, "learning_rate": 1.2676573604900408e-05, "loss": 1.2358, "step": 1435 }, { "epoch": 0.9149410640331316, "grad_norm": 12.432934964150565, "learning_rate": 1.2667443758796125e-05, "loss": 1.6895, "step": 1436 }, { "epoch": 0.9155782096208984, "grad_norm": 9.863540024959198, "learning_rate": 1.2658312135448092e-05, "loss": 1.324, "step": 1437 }, { "epoch": 0.9162153552086651, "grad_norm": 13.323653939658938, "learning_rate": 1.2649178744589236e-05, "loss": 1.2869, "step": 1438 }, { "epoch": 0.916852500796432, "grad_norm": 12.398947388754399, "learning_rate": 1.264004359595438e-05, "loss": 1.0604, "step": 1439 }, { "epoch": 0.9174896463841988, "grad_norm": 13.965995486502115, "learning_rate": 1.2630906699280214e-05, "loss": 1.1385, "step": 1440 }, { "epoch": 0.9181267919719656, "grad_norm": 9.370382856202882, "learning_rate": 1.2621768064305294e-05, "loss": 1.3398, "step": 1441 }, { "epoch": 0.9187639375597324, "grad_norm": 11.479466227213496, "learning_rate": 1.2612627700770024e-05, "loss": 1.537, "step": 1442 }, { "epoch": 0.9194010831474992, "grad_norm": 11.601078550441706, "learning_rate": 1.2603485618416655e-05, "loss": 1.5107, "step": 1443 }, { "epoch": 0.920038228735266, "grad_norm": 12.354021231753649, "learning_rate": 1.2594341826989276e-05, "loss": 1.4041, "step": 1444 }, { "epoch": 0.9206753743230328, "grad_norm": 14.938976486947958, "learning_rate": 1.2585196336233784e-05, "loss": 0.9036, "step": 1445 }, { "epoch": 0.9213125199107997, "grad_norm": 13.734455596961082, "learning_rate": 1.2576049155897893e-05, "loss": 1.219, "step": 1446 }, { "epoch": 0.9219496654985664, "grad_norm": 8.249481038550043, "learning_rate": 1.2566900295731123e-05, "loss": 1.1738, "step": 1447 }, { "epoch": 0.9225868110863332, "grad_norm": 11.866867658119205, "learning_rate": 1.255774976548478e-05, "loss": 1.1418, "step": 1448 }, { "epoch": 0.9232239566741001, "grad_norm": 13.023966731067846, "learning_rate": 1.2548597574911945e-05, "loss": 1.4894, "step": 1449 }, { "epoch": 0.9238611022618668, "grad_norm": 67.44558277813292, "learning_rate": 1.2539443733767478e-05, "loss": 1.2388, "step": 1450 }, { "epoch": 0.9244982478496336, "grad_norm": 16.087567452490877, "learning_rate": 1.2530288251807994e-05, "loss": 1.5281, "step": 1451 }, { "epoch": 0.9251353934374005, "grad_norm": 10.866527349036328, "learning_rate": 1.2521131138791853e-05, "loss": 1.1013, "step": 1452 }, { "epoch": 0.9257725390251672, "grad_norm": 14.303275386646229, "learning_rate": 1.2511972404479167e-05, "loss": 1.2249, "step": 1453 }, { "epoch": 0.926409684612934, "grad_norm": 11.887491064409303, "learning_rate": 1.2502812058631756e-05, "loss": 1.7246, "step": 1454 }, { "epoch": 0.9270468302007009, "grad_norm": 11.814507244051319, "learning_rate": 1.2493650111013173e-05, "loss": 1.499, "step": 1455 }, { "epoch": 0.9276839757884676, "grad_norm": 9.783719650196437, "learning_rate": 1.2484486571388672e-05, "loss": 1.2489, "step": 1456 }, { "epoch": 0.9283211213762345, "grad_norm": 34.761190760574536, "learning_rate": 1.2475321449525208e-05, "loss": 1.2639, "step": 1457 }, { "epoch": 0.9289582669640013, "grad_norm": 14.978925174542175, "learning_rate": 1.2466154755191418e-05, "loss": 1.7339, "step": 1458 }, { "epoch": 0.929595412551768, "grad_norm": 6.4300580051086795, "learning_rate": 1.2456986498157613e-05, "loss": 1.0251, "step": 1459 }, { "epoch": 0.9302325581395349, "grad_norm": 12.135977870895367, "learning_rate": 1.2447816688195779e-05, "loss": 1.422, "step": 1460 }, { "epoch": 0.9308697037273017, "grad_norm": 6.8407895232957525, "learning_rate": 1.243864533507955e-05, "loss": 0.7548, "step": 1461 }, { "epoch": 0.9315068493150684, "grad_norm": 19.9244103882623, "learning_rate": 1.2429472448584206e-05, "loss": 1.4153, "step": 1462 }, { "epoch": 0.9321439949028353, "grad_norm": 10.301717588263017, "learning_rate": 1.242029803848666e-05, "loss": 0.8855, "step": 1463 }, { "epoch": 0.9327811404906021, "grad_norm": 9.442786481996073, "learning_rate": 1.2411122114565452e-05, "loss": 1.165, "step": 1464 }, { "epoch": 0.9334182860783689, "grad_norm": 9.23935515173557, "learning_rate": 1.2401944686600734e-05, "loss": 0.8652, "step": 1465 }, { "epoch": 0.9340554316661357, "grad_norm": 11.811072868330275, "learning_rate": 1.239276576437426e-05, "loss": 1.311, "step": 1466 }, { "epoch": 0.9346925772539025, "grad_norm": 14.438048748637152, "learning_rate": 1.2383585357669376e-05, "loss": 1.3872, "step": 1467 }, { "epoch": 0.9353297228416694, "grad_norm": 12.067964030812794, "learning_rate": 1.237440347627102e-05, "loss": 1.4127, "step": 1468 }, { "epoch": 0.9359668684294361, "grad_norm": 27.115887580943415, "learning_rate": 1.2365220129965686e-05, "loss": 1.6797, "step": 1469 }, { "epoch": 0.936604014017203, "grad_norm": 13.69385171245648, "learning_rate": 1.2356035328541436e-05, "loss": 1.2371, "step": 1470 }, { "epoch": 0.9372411596049698, "grad_norm": 12.64812046720311, "learning_rate": 1.2346849081787891e-05, "loss": 1.4519, "step": 1471 }, { "epoch": 0.9378783051927365, "grad_norm": 6.555956632481931, "learning_rate": 1.2337661399496199e-05, "loss": 1.0244, "step": 1472 }, { "epoch": 0.9385154507805034, "grad_norm": 15.252112894218307, "learning_rate": 1.2328472291459052e-05, "loss": 1.3368, "step": 1473 }, { "epoch": 0.9391525963682702, "grad_norm": 8.792402745657093, "learning_rate": 1.2319281767470652e-05, "loss": 1.2034, "step": 1474 }, { "epoch": 0.9397897419560369, "grad_norm": 9.759239307980879, "learning_rate": 1.2310089837326708e-05, "loss": 1.3413, "step": 1475 }, { "epoch": 0.9404268875438038, "grad_norm": 10.214073409437734, "learning_rate": 1.2300896510824434e-05, "loss": 1.5927, "step": 1476 }, { "epoch": 0.9410640331315706, "grad_norm": 11.900421435051076, "learning_rate": 1.2291701797762532e-05, "loss": 1.0081, "step": 1477 }, { "epoch": 0.9417011787193373, "grad_norm": 16.55924186153805, "learning_rate": 1.2282505707941181e-05, "loss": 1.2579, "step": 1478 }, { "epoch": 0.9423383243071042, "grad_norm": 14.394824750093054, "learning_rate": 1.2273308251162023e-05, "loss": 1.0913, "step": 1479 }, { "epoch": 0.942975469894871, "grad_norm": 18.54619355657725, "learning_rate": 1.2264109437228168e-05, "loss": 1.4017, "step": 1480 }, { "epoch": 0.9436126154826378, "grad_norm": 19.66719557270391, "learning_rate": 1.2254909275944162e-05, "loss": 1.5178, "step": 1481 }, { "epoch": 0.9442497610704046, "grad_norm": 7.232342959533586, "learning_rate": 1.2245707777115984e-05, "loss": 1.1707, "step": 1482 }, { "epoch": 0.9448869066581714, "grad_norm": 11.627562826422025, "learning_rate": 1.2236504950551058e-05, "loss": 1.2227, "step": 1483 }, { "epoch": 0.9455240522459382, "grad_norm": 11.649291956984726, "learning_rate": 1.2227300806058195e-05, "loss": 1.6331, "step": 1484 }, { "epoch": 0.946161197833705, "grad_norm": 9.17777082185266, "learning_rate": 1.2218095353447637e-05, "loss": 0.8933, "step": 1485 }, { "epoch": 0.9467983434214718, "grad_norm": 11.480947082359439, "learning_rate": 1.2208888602531007e-05, "loss": 1.1055, "step": 1486 }, { "epoch": 0.9474354890092386, "grad_norm": 9.087628983908903, "learning_rate": 1.219968056312131e-05, "loss": 1.4143, "step": 1487 }, { "epoch": 0.9480726345970054, "grad_norm": 10.659283494558215, "learning_rate": 1.2190471245032935e-05, "loss": 1.614, "step": 1488 }, { "epoch": 0.9487097801847723, "grad_norm": 17.362779015724684, "learning_rate": 1.218126065808162e-05, "loss": 1.7419, "step": 1489 }, { "epoch": 0.949346925772539, "grad_norm": 8.446396136873288, "learning_rate": 1.2172048812084467e-05, "loss": 0.8059, "step": 1490 }, { "epoch": 0.9499840713603058, "grad_norm": 11.358555393004881, "learning_rate": 1.2162835716859913e-05, "loss": 1.26, "step": 1491 }, { "epoch": 0.9506212169480727, "grad_norm": 15.961313977718008, "learning_rate": 1.2153621382227731e-05, "loss": 1.4424, "step": 1492 }, { "epoch": 0.9512583625358394, "grad_norm": 14.355459359047048, "learning_rate": 1.2144405818009013e-05, "loss": 1.4241, "step": 1493 }, { "epoch": 0.9518955081236062, "grad_norm": 7.8429832088575315, "learning_rate": 1.2135189034026162e-05, "loss": 0.8213, "step": 1494 }, { "epoch": 0.9525326537113731, "grad_norm": 11.393876629431633, "learning_rate": 1.212597104010288e-05, "loss": 1.4944, "step": 1495 }, { "epoch": 0.9531697992991398, "grad_norm": 12.748469470347388, "learning_rate": 1.2116751846064158e-05, "loss": 1.4348, "step": 1496 }, { "epoch": 0.9538069448869066, "grad_norm": 8.935224117122875, "learning_rate": 1.2107531461736271e-05, "loss": 1.2827, "step": 1497 }, { "epoch": 0.9544440904746735, "grad_norm": 14.186398168756819, "learning_rate": 1.2098309896946751e-05, "loss": 1.4802, "step": 1498 }, { "epoch": 0.9550812360624402, "grad_norm": 12.063986859033719, "learning_rate": 1.2089087161524406e-05, "loss": 1.5134, "step": 1499 }, { "epoch": 0.9557183816502071, "grad_norm": 15.256992117885424, "learning_rate": 1.207986326529928e-05, "loss": 1.4187, "step": 1500 }, { "epoch": 0.9563555272379739, "grad_norm": 21.902510621984813, "learning_rate": 1.2070638218102652e-05, "loss": 0.9198, "step": 1501 }, { "epoch": 0.9569926728257406, "grad_norm": 4.701507373898286, "learning_rate": 1.2061412029767034e-05, "loss": 0.5461, "step": 1502 }, { "epoch": 0.9576298184135075, "grad_norm": 7.950665182479189, "learning_rate": 1.2052184710126156e-05, "loss": 0.9663, "step": 1503 }, { "epoch": 0.9582669640012743, "grad_norm": 11.995257547725727, "learning_rate": 1.2042956269014948e-05, "loss": 1.1248, "step": 1504 }, { "epoch": 0.958904109589041, "grad_norm": 11.038724139905352, "learning_rate": 1.203372671626953e-05, "loss": 1.0635, "step": 1505 }, { "epoch": 0.9595412551768079, "grad_norm": 12.163068070900959, "learning_rate": 1.2024496061727222e-05, "loss": 1.3087, "step": 1506 }, { "epoch": 0.9601784007645747, "grad_norm": 9.896622352561009, "learning_rate": 1.2015264315226513e-05, "loss": 0.9034, "step": 1507 }, { "epoch": 0.9608155463523416, "grad_norm": 14.405361932939856, "learning_rate": 1.2006031486607048e-05, "loss": 1.0532, "step": 1508 }, { "epoch": 0.9614526919401083, "grad_norm": 13.919819022055185, "learning_rate": 1.1996797585709634e-05, "loss": 0.9979, "step": 1509 }, { "epoch": 0.9620898375278751, "grad_norm": 11.160039566177344, "learning_rate": 1.1987562622376214e-05, "loss": 0.9971, "step": 1510 }, { "epoch": 0.962726983115642, "grad_norm": 10.103923237082256, "learning_rate": 1.1978326606449872e-05, "loss": 1.1095, "step": 1511 }, { "epoch": 0.9633641287034087, "grad_norm": 49.40559603735989, "learning_rate": 1.1969089547774806e-05, "loss": 1.0231, "step": 1512 }, { "epoch": 0.9640012742911755, "grad_norm": 42.32538436071784, "learning_rate": 1.1959851456196326e-05, "loss": 1.4611, "step": 1513 }, { "epoch": 0.9646384198789424, "grad_norm": 8.348262945639423, "learning_rate": 1.1950612341560852e-05, "loss": 1.0416, "step": 1514 }, { "epoch": 0.9652755654667091, "grad_norm": 15.137289885792466, "learning_rate": 1.1941372213715885e-05, "loss": 1.5908, "step": 1515 }, { "epoch": 0.965912711054476, "grad_norm": 16.98166743513247, "learning_rate": 1.1932131082510005e-05, "loss": 1.6016, "step": 1516 }, { "epoch": 0.9665498566422428, "grad_norm": 19.192538618419512, "learning_rate": 1.192288895779287e-05, "loss": 1.9539, "step": 1517 }, { "epoch": 0.9671870022300095, "grad_norm": 18.31255881096285, "learning_rate": 1.1913645849415189e-05, "loss": 1.6282, "step": 1518 }, { "epoch": 0.9678241478177764, "grad_norm": 11.547493170791247, "learning_rate": 1.1904401767228725e-05, "loss": 1.5437, "step": 1519 }, { "epoch": 0.9684612934055432, "grad_norm": 6.114231359315228, "learning_rate": 1.1895156721086275e-05, "loss": 0.7581, "step": 1520 }, { "epoch": 0.9690984389933099, "grad_norm": 12.857293986302807, "learning_rate": 1.1885910720841669e-05, "loss": 1.2845, "step": 1521 }, { "epoch": 0.9697355845810768, "grad_norm": 11.139332752075736, "learning_rate": 1.1876663776349742e-05, "loss": 1.0518, "step": 1522 }, { "epoch": 0.9703727301688436, "grad_norm": 13.10756876506012, "learning_rate": 1.1867415897466348e-05, "loss": 1.6594, "step": 1523 }, { "epoch": 0.9710098757566104, "grad_norm": 10.77096000233611, "learning_rate": 1.1858167094048332e-05, "loss": 1.4307, "step": 1524 }, { "epoch": 0.9716470213443772, "grad_norm": 12.511550920467217, "learning_rate": 1.1848917375953526e-05, "loss": 1.2803, "step": 1525 }, { "epoch": 0.972284166932144, "grad_norm": 8.026589950327216, "learning_rate": 1.1839666753040728e-05, "loss": 1.0891, "step": 1526 }, { "epoch": 0.9729213125199108, "grad_norm": 21.475721098948767, "learning_rate": 1.1830415235169718e-05, "loss": 1.4158, "step": 1527 }, { "epoch": 0.9735584581076776, "grad_norm": 12.881979092536074, "learning_rate": 1.1821162832201216e-05, "loss": 1.3279, "step": 1528 }, { "epoch": 0.9741956036954444, "grad_norm": 9.941061570827427, "learning_rate": 1.1811909553996887e-05, "loss": 1.1196, "step": 1529 }, { "epoch": 0.9748327492832112, "grad_norm": 12.099217534246355, "learning_rate": 1.1802655410419335e-05, "loss": 1.5857, "step": 1530 }, { "epoch": 0.975469894870978, "grad_norm": 11.923830674848267, "learning_rate": 1.1793400411332077e-05, "loss": 1.614, "step": 1531 }, { "epoch": 0.9761070404587449, "grad_norm": 20.719894063281373, "learning_rate": 1.178414456659955e-05, "loss": 1.0996, "step": 1532 }, { "epoch": 0.9767441860465116, "grad_norm": 10.788679219333883, "learning_rate": 1.1774887886087091e-05, "loss": 1.3098, "step": 1533 }, { "epoch": 0.9773813316342784, "grad_norm": 18.358445332805143, "learning_rate": 1.1765630379660928e-05, "loss": 1.4583, "step": 1534 }, { "epoch": 0.9780184772220453, "grad_norm": 17.570880706808033, "learning_rate": 1.1756372057188161e-05, "loss": 1.0471, "step": 1535 }, { "epoch": 0.978655622809812, "grad_norm": 12.121399177208758, "learning_rate": 1.174711292853677e-05, "loss": 1.2021, "step": 1536 }, { "epoch": 0.9792927683975788, "grad_norm": 14.443556311154124, "learning_rate": 1.1737853003575595e-05, "loss": 1.4282, "step": 1537 }, { "epoch": 0.9799299139853457, "grad_norm": 24.790162622732723, "learning_rate": 1.1728592292174316e-05, "loss": 1.626, "step": 1538 }, { "epoch": 0.9805670595731124, "grad_norm": 12.050283673394414, "learning_rate": 1.1719330804203453e-05, "loss": 0.9722, "step": 1539 }, { "epoch": 0.9812042051608793, "grad_norm": 18.797576535609643, "learning_rate": 1.1710068549534362e-05, "loss": 1.4873, "step": 1540 }, { "epoch": 0.9818413507486461, "grad_norm": 11.195620423381103, "learning_rate": 1.1700805538039208e-05, "loss": 1.05, "step": 1541 }, { "epoch": 0.9824784963364128, "grad_norm": 18.627998444335034, "learning_rate": 1.1691541779590965e-05, "loss": 1.3871, "step": 1542 }, { "epoch": 0.9831156419241797, "grad_norm": 8.69750834769125, "learning_rate": 1.1682277284063406e-05, "loss": 0.9272, "step": 1543 }, { "epoch": 0.9837527875119465, "grad_norm": 16.51135713841813, "learning_rate": 1.1673012061331083e-05, "loss": 1.2019, "step": 1544 }, { "epoch": 0.9843899330997132, "grad_norm": 10.867432409269478, "learning_rate": 1.166374612126933e-05, "loss": 1.2197, "step": 1545 }, { "epoch": 0.9850270786874801, "grad_norm": 9.866264895613797, "learning_rate": 1.1654479473754243e-05, "loss": 1.0112, "step": 1546 }, { "epoch": 0.9856642242752469, "grad_norm": 30.09854744996536, "learning_rate": 1.164521212866267e-05, "loss": 1.2742, "step": 1547 }, { "epoch": 0.9863013698630136, "grad_norm": 14.853517203378729, "learning_rate": 1.1635944095872205e-05, "loss": 1.4187, "step": 1548 }, { "epoch": 0.9869385154507805, "grad_norm": 12.678505196334067, "learning_rate": 1.162667538526117e-05, "loss": 1.2759, "step": 1549 }, { "epoch": 0.9875756610385473, "grad_norm": 12.954172452830125, "learning_rate": 1.1617406006708624e-05, "loss": 1.4697, "step": 1550 }, { "epoch": 0.9882128066263142, "grad_norm": 5.729890223798183, "learning_rate": 1.1608135970094316e-05, "loss": 0.665, "step": 1551 }, { "epoch": 0.9888499522140809, "grad_norm": 17.5398753984428, "learning_rate": 1.1598865285298717e-05, "loss": 1.3994, "step": 1552 }, { "epoch": 0.9894870978018477, "grad_norm": 33.14046875355449, "learning_rate": 1.1589593962202974e-05, "loss": 1.6694, "step": 1553 }, { "epoch": 0.9901242433896146, "grad_norm": 10.283042378301278, "learning_rate": 1.1580322010688922e-05, "loss": 1.2501, "step": 1554 }, { "epoch": 0.9907613889773813, "grad_norm": 14.497335211894034, "learning_rate": 1.1571049440639066e-05, "loss": 0.884, "step": 1555 }, { "epoch": 0.9913985345651481, "grad_norm": 17.792043733053287, "learning_rate": 1.1561776261936566e-05, "loss": 0.895, "step": 1556 }, { "epoch": 0.992035680152915, "grad_norm": 12.037664793168721, "learning_rate": 1.1552502484465236e-05, "loss": 0.8796, "step": 1557 }, { "epoch": 0.9926728257406817, "grad_norm": 16.44357628690183, "learning_rate": 1.154322811810952e-05, "loss": 1.561, "step": 1558 }, { "epoch": 0.9933099713284486, "grad_norm": 16.208673531838553, "learning_rate": 1.1533953172754495e-05, "loss": 1.3335, "step": 1559 }, { "epoch": 0.9939471169162154, "grad_norm": 13.05661535836861, "learning_rate": 1.1524677658285863e-05, "loss": 1.6147, "step": 1560 }, { "epoch": 0.9945842625039821, "grad_norm": 15.34036517058158, "learning_rate": 1.1515401584589913e-05, "loss": 1.1497, "step": 1561 }, { "epoch": 0.995221408091749, "grad_norm": 12.545359771409451, "learning_rate": 1.1506124961553549e-05, "loss": 0.8784, "step": 1562 }, { "epoch": 0.9958585536795158, "grad_norm": 18.75741919546992, "learning_rate": 1.1496847799064255e-05, "loss": 1.6145, "step": 1563 }, { "epoch": 0.9964956992672825, "grad_norm": 33.74214905037755, "learning_rate": 1.1487570107010085e-05, "loss": 1.3106, "step": 1564 }, { "epoch": 0.9971328448550494, "grad_norm": 28.543900700941883, "learning_rate": 1.1478291895279658e-05, "loss": 1.5497, "step": 1565 }, { "epoch": 0.9977699904428162, "grad_norm": 18.779602281688835, "learning_rate": 1.1469013173762154e-05, "loss": 1.3762, "step": 1566 }, { "epoch": 0.998407136030583, "grad_norm": 11.137360169882484, "learning_rate": 1.145973395234729e-05, "loss": 1.0115, "step": 1567 }, { "epoch": 0.9990442816183498, "grad_norm": 8.347610014950819, "learning_rate": 1.145045424092531e-05, "loss": 0.8384, "step": 1568 }, { "epoch": 0.9996814272061166, "grad_norm": 21.690079590678515, "learning_rate": 1.1441174049387e-05, "loss": 1.8062, "step": 1569 }, { "epoch": 1.0, "grad_norm": 21.690079590678515, "learning_rate": 1.1431893387623645e-05, "loss": 0.6318, "step": 1570 }, { "epoch": 1.0006371455877667, "grad_norm": 30.124978333458508, "learning_rate": 1.1422612265527025e-05, "loss": 0.8132, "step": 1571 }, { "epoch": 1.0012742911755337, "grad_norm": 6.569489168172049, "learning_rate": 1.1413330692989415e-05, "loss": 0.762, "step": 1572 }, { "epoch": 1.0019114367633004, "grad_norm": 14.763677143597455, "learning_rate": 1.1404048679903583e-05, "loss": 1.0181, "step": 1573 }, { "epoch": 1.0025485823510671, "grad_norm": 17.65389085788204, "learning_rate": 1.1394766236162749e-05, "loss": 1.3479, "step": 1574 }, { "epoch": 1.003185727938834, "grad_norm": 14.075548693162235, "learning_rate": 1.13854833716606e-05, "loss": 1.2529, "step": 1575 }, { "epoch": 1.0038228735266008, "grad_norm": 11.038813434708398, "learning_rate": 1.1376200096291267e-05, "loss": 0.734, "step": 1576 }, { "epoch": 1.0044600191143676, "grad_norm": 6.145634777952769, "learning_rate": 1.1366916419949335e-05, "loss": 1.293, "step": 1577 }, { "epoch": 1.0050971647021345, "grad_norm": 21.746043395689185, "learning_rate": 1.1357632352529792e-05, "loss": 1.3811, "step": 1578 }, { "epoch": 1.0057343102899012, "grad_norm": 178.39400021638994, "learning_rate": 1.1348347903928057e-05, "loss": 1.2625, "step": 1579 }, { "epoch": 1.006371455877668, "grad_norm": 13.803415048110066, "learning_rate": 1.1339063084039956e-05, "loss": 1.5039, "step": 1580 }, { "epoch": 1.007008601465435, "grad_norm": 11.04960443141559, "learning_rate": 1.1329777902761708e-05, "loss": 1.2012, "step": 1581 }, { "epoch": 1.0076457470532016, "grad_norm": 10.745936997843494, "learning_rate": 1.1320492369989911e-05, "loss": 1.1868, "step": 1582 }, { "epoch": 1.0082828926409684, "grad_norm": 11.434023963737681, "learning_rate": 1.1311206495621547e-05, "loss": 1.2117, "step": 1583 }, { "epoch": 1.0089200382287353, "grad_norm": 15.486062864960031, "learning_rate": 1.1301920289553963e-05, "loss": 1.6421, "step": 1584 }, { "epoch": 1.009557183816502, "grad_norm": 16.908401483016583, "learning_rate": 1.1292633761684848e-05, "loss": 1.3379, "step": 1585 }, { "epoch": 1.0101943294042688, "grad_norm": 13.249007892746402, "learning_rate": 1.1283346921912245e-05, "loss": 0.8193, "step": 1586 }, { "epoch": 1.0108314749920357, "grad_norm": 12.993700025668577, "learning_rate": 1.127405978013453e-05, "loss": 1.5688, "step": 1587 }, { "epoch": 1.0114686205798025, "grad_norm": 14.145826934346058, "learning_rate": 1.1264772346250388e-05, "loss": 1.5039, "step": 1588 }, { "epoch": 1.0121057661675692, "grad_norm": 17.429128761814265, "learning_rate": 1.1255484630158823e-05, "loss": 1.6711, "step": 1589 }, { "epoch": 1.0127429117553362, "grad_norm": 13.883677305325024, "learning_rate": 1.1246196641759152e-05, "loss": 1.1655, "step": 1590 }, { "epoch": 1.0133800573431029, "grad_norm": 14.80614969133142, "learning_rate": 1.1236908390950963e-05, "loss": 0.9309, "step": 1591 }, { "epoch": 1.0140172029308696, "grad_norm": 6.936320206919998, "learning_rate": 1.1227619887634131e-05, "loss": 0.7709, "step": 1592 }, { "epoch": 1.0146543485186366, "grad_norm": 18.568527724198198, "learning_rate": 1.1218331141708802e-05, "loss": 1.5447, "step": 1593 }, { "epoch": 1.0152914941064033, "grad_norm": 16.698542042662865, "learning_rate": 1.1209042163075383e-05, "loss": 1.1583, "step": 1594 }, { "epoch": 1.01592863969417, "grad_norm": 11.610098206702881, "learning_rate": 1.1199752961634516e-05, "loss": 0.816, "step": 1595 }, { "epoch": 1.016565785281937, "grad_norm": 16.743785231489728, "learning_rate": 1.11904635472871e-05, "loss": 1.3925, "step": 1596 }, { "epoch": 1.0172029308697037, "grad_norm": 19.391877818302685, "learning_rate": 1.1181173929934246e-05, "loss": 1.5552, "step": 1597 }, { "epoch": 1.0178400764574704, "grad_norm": 9.851027849964408, "learning_rate": 1.1171884119477287e-05, "loss": 1.1034, "step": 1598 }, { "epoch": 1.0184772220452374, "grad_norm": 15.910876466809325, "learning_rate": 1.1162594125817758e-05, "loss": 0.9546, "step": 1599 }, { "epoch": 1.0191143676330041, "grad_norm": 18.85260485024362, "learning_rate": 1.1153303958857395e-05, "loss": 1.5122, "step": 1600 }, { "epoch": 1.0197515132207708, "grad_norm": 18.01249307913508, "learning_rate": 1.114401362849812e-05, "loss": 1.4352, "step": 1601 }, { "epoch": 1.0203886588085378, "grad_norm": 32.948508605226706, "learning_rate": 1.1134723144642017e-05, "loss": 1.7119, "step": 1602 }, { "epoch": 1.0210258043963045, "grad_norm": 28.752496498571357, "learning_rate": 1.1125432517191346e-05, "loss": 1.406, "step": 1603 }, { "epoch": 1.0216629499840713, "grad_norm": 19.330363001328198, "learning_rate": 1.1116141756048513e-05, "loss": 1.4861, "step": 1604 }, { "epoch": 1.0223000955718382, "grad_norm": 13.265830483971056, "learning_rate": 1.1106850871116073e-05, "loss": 0.9542, "step": 1605 }, { "epoch": 1.022937241159605, "grad_norm": 28.211992218899436, "learning_rate": 1.1097559872296708e-05, "loss": 1.0613, "step": 1606 }, { "epoch": 1.0235743867473717, "grad_norm": 12.520221553482374, "learning_rate": 1.1088268769493225e-05, "loss": 1.2673, "step": 1607 }, { "epoch": 1.0242115323351386, "grad_norm": 8.78345571265968, "learning_rate": 1.1078977572608532e-05, "loss": 1.142, "step": 1608 }, { "epoch": 1.0248486779229053, "grad_norm": 18.91550654550647, "learning_rate": 1.1069686291545649e-05, "loss": 0.7063, "step": 1609 }, { "epoch": 1.0254858235106723, "grad_norm": 15.51584292260518, "learning_rate": 1.1060394936207678e-05, "loss": 1.1089, "step": 1610 }, { "epoch": 1.026122969098439, "grad_norm": 13.740053503730797, "learning_rate": 1.1051103516497814e-05, "loss": 1.132, "step": 1611 }, { "epoch": 1.0267601146862058, "grad_norm": 13.628414213406684, "learning_rate": 1.10418120423193e-05, "loss": 1.3971, "step": 1612 }, { "epoch": 1.0273972602739727, "grad_norm": 16.029681026593284, "learning_rate": 1.103252052357545e-05, "loss": 1.3444, "step": 1613 }, { "epoch": 1.0280344058617394, "grad_norm": 10.773273800442515, "learning_rate": 1.1023228970169623e-05, "loss": 1.3944, "step": 1614 }, { "epoch": 1.0286715514495062, "grad_norm": 18.564544604522283, "learning_rate": 1.1013937392005214e-05, "loss": 1.2255, "step": 1615 }, { "epoch": 1.0293086970372731, "grad_norm": 8.942903358603836, "learning_rate": 1.1004645798985646e-05, "loss": 1.1735, "step": 1616 }, { "epoch": 1.0299458426250399, "grad_norm": 8.354738697563002, "learning_rate": 1.0995354201014356e-05, "loss": 1.1851, "step": 1617 }, { "epoch": 1.0305829882128066, "grad_norm": 17.12549009733859, "learning_rate": 1.0986062607994787e-05, "loss": 0.8748, "step": 1618 }, { "epoch": 1.0312201338005735, "grad_norm": 20.22650205967701, "learning_rate": 1.097677102983038e-05, "loss": 1.1445, "step": 1619 }, { "epoch": 1.0318572793883403, "grad_norm": 18.45565556172361, "learning_rate": 1.0967479476424554e-05, "loss": 1.3702, "step": 1620 }, { "epoch": 1.032494424976107, "grad_norm": 19.060211860001864, "learning_rate": 1.0958187957680704e-05, "loss": 1.8853, "step": 1621 }, { "epoch": 1.033131570563874, "grad_norm": 16.003642710260717, "learning_rate": 1.0948896483502192e-05, "loss": 1.488, "step": 1622 }, { "epoch": 1.0337687161516407, "grad_norm": 15.163652101551179, "learning_rate": 1.0939605063792324e-05, "loss": 1.2632, "step": 1623 }, { "epoch": 1.0344058617394074, "grad_norm": 15.064683383011644, "learning_rate": 1.0930313708454354e-05, "loss": 1.3583, "step": 1624 }, { "epoch": 1.0350430073271744, "grad_norm": 12.940307182328105, "learning_rate": 1.0921022427391473e-05, "loss": 1.2375, "step": 1625 }, { "epoch": 1.035680152914941, "grad_norm": 13.9407628104505, "learning_rate": 1.0911731230506782e-05, "loss": 1.177, "step": 1626 }, { "epoch": 1.0363172985027078, "grad_norm": 47.80940901664529, "learning_rate": 1.0902440127703295e-05, "loss": 1.5068, "step": 1627 }, { "epoch": 1.0369544440904748, "grad_norm": 17.3070935991362, "learning_rate": 1.0893149128883932e-05, "loss": 1.2974, "step": 1628 }, { "epoch": 1.0375915896782415, "grad_norm": 17.71173111732036, "learning_rate": 1.088385824395149e-05, "loss": 1.2012, "step": 1629 }, { "epoch": 1.0382287352660082, "grad_norm": 9.988061053843888, "learning_rate": 1.0874567482808659e-05, "loss": 0.526, "step": 1630 }, { "epoch": 1.0388658808537752, "grad_norm": 25.564550713702445, "learning_rate": 1.0865276855357987e-05, "loss": 1.0334, "step": 1631 }, { "epoch": 1.039503026441542, "grad_norm": 24.319500176907614, "learning_rate": 1.0855986371501884e-05, "loss": 1.8091, "step": 1632 }, { "epoch": 1.0401401720293086, "grad_norm": 14.944405863694342, "learning_rate": 1.0846696041142606e-05, "loss": 1.2732, "step": 1633 }, { "epoch": 1.0407773176170756, "grad_norm": 48.83815194682277, "learning_rate": 1.0837405874182245e-05, "loss": 1.28, "step": 1634 }, { "epoch": 1.0414144632048423, "grad_norm": 8.171820247893427, "learning_rate": 1.0828115880522719e-05, "loss": 0.9598, "step": 1635 }, { "epoch": 1.042051608792609, "grad_norm": 154.22138720621618, "learning_rate": 1.081882607006576e-05, "loss": 1.4292, "step": 1636 }, { "epoch": 1.042688754380376, "grad_norm": 22.027170228981323, "learning_rate": 1.0809536452712904e-05, "loss": 1.4436, "step": 1637 }, { "epoch": 1.0433258999681427, "grad_norm": 16.547164000339965, "learning_rate": 1.0800247038365486e-05, "loss": 1.0706, "step": 1638 }, { "epoch": 1.0439630455559095, "grad_norm": 21.08747759260983, "learning_rate": 1.0790957836924623e-05, "loss": 1.0388, "step": 1639 }, { "epoch": 1.0446001911436764, "grad_norm": 27.279543828129164, "learning_rate": 1.07816688582912e-05, "loss": 1.6752, "step": 1640 }, { "epoch": 1.0452373367314431, "grad_norm": 19.985329324269188, "learning_rate": 1.0772380112365874e-05, "loss": 1.4856, "step": 1641 }, { "epoch": 1.0458744823192099, "grad_norm": 24.948820685883454, "learning_rate": 1.0763091609049043e-05, "loss": 1.4614, "step": 1642 }, { "epoch": 1.0465116279069768, "grad_norm": 18.085792394151852, "learning_rate": 1.0753803358240854e-05, "loss": 1.4163, "step": 1643 }, { "epoch": 1.0471487734947436, "grad_norm": 13.6168292483925, "learning_rate": 1.0744515369841178e-05, "loss": 0.9064, "step": 1644 }, { "epoch": 1.0477859190825103, "grad_norm": 14.479937288098725, "learning_rate": 1.0735227653749618e-05, "loss": 1.444, "step": 1645 }, { "epoch": 1.0484230646702772, "grad_norm": 13.941990378948267, "learning_rate": 1.0725940219865473e-05, "loss": 1.1219, "step": 1646 }, { "epoch": 1.049060210258044, "grad_norm": 20.304051439244088, "learning_rate": 1.0716653078087756e-05, "loss": 0.8368, "step": 1647 }, { "epoch": 1.0496973558458107, "grad_norm": 11.315939748381282, "learning_rate": 1.0707366238315156e-05, "loss": 1.488, "step": 1648 }, { "epoch": 1.0503345014335776, "grad_norm": 16.110377692080448, "learning_rate": 1.0698079710446043e-05, "loss": 1.343, "step": 1649 }, { "epoch": 1.0509716470213444, "grad_norm": 8.670093991363531, "learning_rate": 1.0688793504378455e-05, "loss": 1.21, "step": 1650 }, { "epoch": 1.051608792609111, "grad_norm": 16.511489831086333, "learning_rate": 1.0679507630010093e-05, "loss": 1.4858, "step": 1651 }, { "epoch": 1.052245938196878, "grad_norm": 15.041686715407392, "learning_rate": 1.0670222097238297e-05, "loss": 0.8894, "step": 1652 }, { "epoch": 1.0528830837846448, "grad_norm": 11.873037262144214, "learning_rate": 1.0660936915960047e-05, "loss": 1.0427, "step": 1653 }, { "epoch": 1.0535202293724115, "grad_norm": 11.219628997090902, "learning_rate": 1.0651652096071948e-05, "loss": 1.0977, "step": 1654 }, { "epoch": 1.0541573749601785, "grad_norm": 10.842420220499292, "learning_rate": 1.0642367647470213e-05, "loss": 1.1082, "step": 1655 }, { "epoch": 1.0547945205479452, "grad_norm": 19.106242615105337, "learning_rate": 1.0633083580050671e-05, "loss": 1.5144, "step": 1656 }, { "epoch": 1.055431666135712, "grad_norm": 31.338804933260757, "learning_rate": 1.0623799903708732e-05, "loss": 0.856, "step": 1657 }, { "epoch": 1.0560688117234789, "grad_norm": 17.09206690043829, "learning_rate": 1.0614516628339405e-05, "loss": 1.2853, "step": 1658 }, { "epoch": 1.0567059573112456, "grad_norm": 28.836624875164052, "learning_rate": 1.0605233763837254e-05, "loss": 1.4797, "step": 1659 }, { "epoch": 1.0573431028990123, "grad_norm": 21.293446102783008, "learning_rate": 1.059595132009642e-05, "loss": 0.6671, "step": 1660 }, { "epoch": 1.0579802484867793, "grad_norm": 7.095439499049638, "learning_rate": 1.0586669307010587e-05, "loss": 0.5393, "step": 1661 }, { "epoch": 1.058617394074546, "grad_norm": 11.576946683064143, "learning_rate": 1.0577387734472981e-05, "loss": 1.4507, "step": 1662 }, { "epoch": 1.0592545396623128, "grad_norm": 18.773966409833516, "learning_rate": 1.0568106612376362e-05, "loss": 1.3032, "step": 1663 }, { "epoch": 1.0598916852500797, "grad_norm": 14.647949749191744, "learning_rate": 1.0558825950613001e-05, "loss": 1.1909, "step": 1664 }, { "epoch": 1.0605288308378464, "grad_norm": 15.115713206461786, "learning_rate": 1.054954575907469e-05, "loss": 1.1406, "step": 1665 }, { "epoch": 1.0611659764256132, "grad_norm": 14.227801355418714, "learning_rate": 1.0540266047652717e-05, "loss": 1.0256, "step": 1666 }, { "epoch": 1.0618031220133801, "grad_norm": 35.14183261432549, "learning_rate": 1.053098682623785e-05, "loss": 1.3306, "step": 1667 }, { "epoch": 1.0624402676011468, "grad_norm": 21.3809840757498, "learning_rate": 1.0521708104720345e-05, "loss": 1.5099, "step": 1668 }, { "epoch": 1.0630774131889136, "grad_norm": 9.566520013824935, "learning_rate": 1.0512429892989921e-05, "loss": 0.8292, "step": 1669 }, { "epoch": 1.0637145587766805, "grad_norm": 16.906991384056692, "learning_rate": 1.0503152200935746e-05, "loss": 1.498, "step": 1670 }, { "epoch": 1.0643517043644473, "grad_norm": 15.908989350554966, "learning_rate": 1.049387503844645e-05, "loss": 0.8605, "step": 1671 }, { "epoch": 1.064988849952214, "grad_norm": 14.956493336785188, "learning_rate": 1.048459841541009e-05, "loss": 1.582, "step": 1672 }, { "epoch": 1.065625995539981, "grad_norm": 26.04978453963156, "learning_rate": 1.0475322341714143e-05, "loss": 1.8105, "step": 1673 }, { "epoch": 1.0662631411277477, "grad_norm": 21.66649795398796, "learning_rate": 1.0466046827245509e-05, "loss": 1.1307, "step": 1674 }, { "epoch": 1.0669002867155144, "grad_norm": 11.962606574547882, "learning_rate": 1.0456771881890486e-05, "loss": 0.9666, "step": 1675 }, { "epoch": 1.0675374323032814, "grad_norm": 11.508123508441276, "learning_rate": 1.044749751553477e-05, "loss": 0.9415, "step": 1676 }, { "epoch": 1.068174577891048, "grad_norm": 17.406693436735228, "learning_rate": 1.0438223738063435e-05, "loss": 1.2388, "step": 1677 }, { "epoch": 1.0688117234788148, "grad_norm": 12.767509562067996, "learning_rate": 1.0428950559360937e-05, "loss": 1.4, "step": 1678 }, { "epoch": 1.0694488690665818, "grad_norm": 11.228999383545032, "learning_rate": 1.041967798931108e-05, "loss": 1.0918, "step": 1679 }, { "epoch": 1.0700860146543485, "grad_norm": 11.833092202273058, "learning_rate": 1.0410406037797027e-05, "loss": 1.0144, "step": 1680 }, { "epoch": 1.0707231602421152, "grad_norm": 9.04690303093443, "learning_rate": 1.0401134714701287e-05, "loss": 0.9051, "step": 1681 }, { "epoch": 1.0713603058298822, "grad_norm": 16.044240313651308, "learning_rate": 1.0391864029905687e-05, "loss": 1.407, "step": 1682 }, { "epoch": 1.071997451417649, "grad_norm": 20.766416237385087, "learning_rate": 1.0382593993291384e-05, "loss": 1.7046, "step": 1683 }, { "epoch": 1.0726345970054156, "grad_norm": 18.463836162413575, "learning_rate": 1.037332461473883e-05, "loss": 1.1772, "step": 1684 }, { "epoch": 1.0732717425931826, "grad_norm": 18.158459591773433, "learning_rate": 1.0364055904127801e-05, "loss": 1.5779, "step": 1685 }, { "epoch": 1.0739088881809493, "grad_norm": 13.665062501741675, "learning_rate": 1.0354787871337335e-05, "loss": 1.1346, "step": 1686 }, { "epoch": 1.074546033768716, "grad_norm": 11.907505052628279, "learning_rate": 1.034552052624576e-05, "loss": 1.4902, "step": 1687 }, { "epoch": 1.075183179356483, "grad_norm": 14.725389564228436, "learning_rate": 1.0336253878730673e-05, "loss": 1.2977, "step": 1688 }, { "epoch": 1.0758203249442497, "grad_norm": 25.324617534742387, "learning_rate": 1.032698793866892e-05, "loss": 1.6328, "step": 1689 }, { "epoch": 1.0764574705320165, "grad_norm": 16.353482239581396, "learning_rate": 1.0317722715936595e-05, "loss": 1.3738, "step": 1690 }, { "epoch": 1.0770946161197834, "grad_norm": 14.736827210800953, "learning_rate": 1.0308458220409037e-05, "loss": 1.1143, "step": 1691 }, { "epoch": 1.0777317617075501, "grad_norm": 26.702619245602957, "learning_rate": 1.0299194461960795e-05, "loss": 1.2295, "step": 1692 }, { "epoch": 1.0783689072953169, "grad_norm": 19.6037464149453, "learning_rate": 1.0289931450465643e-05, "loss": 1.2324, "step": 1693 }, { "epoch": 1.0790060528830838, "grad_norm": 10.05972859464142, "learning_rate": 1.0280669195796551e-05, "loss": 1.1005, "step": 1694 }, { "epoch": 1.0796431984708506, "grad_norm": 13.53577138770359, "learning_rate": 1.0271407707825688e-05, "loss": 1.7759, "step": 1695 }, { "epoch": 1.0802803440586173, "grad_norm": 16.501754904107887, "learning_rate": 1.026214699642441e-05, "loss": 1.5752, "step": 1696 }, { "epoch": 1.0809174896463842, "grad_norm": 24.570166474936382, "learning_rate": 1.0252887071463229e-05, "loss": 1.5785, "step": 1697 }, { "epoch": 1.081554635234151, "grad_norm": 15.093830586051672, "learning_rate": 1.024362794281184e-05, "loss": 1.7141, "step": 1698 }, { "epoch": 1.0821917808219177, "grad_norm": 21.391544267334368, "learning_rate": 1.0234369620339077e-05, "loss": 1.0477, "step": 1699 }, { "epoch": 1.0828289264096846, "grad_norm": 42.29073360598398, "learning_rate": 1.0225112113912912e-05, "loss": 1.7039, "step": 1700 }, { "epoch": 1.0834660719974514, "grad_norm": 16.158951811408198, "learning_rate": 1.0215855433400452e-05, "loss": 0.8566, "step": 1701 }, { "epoch": 1.0841032175852183, "grad_norm": 14.816508808512525, "learning_rate": 1.0206599588667927e-05, "loss": 1.7471, "step": 1702 }, { "epoch": 1.084740363172985, "grad_norm": 32.73822354325728, "learning_rate": 1.0197344589580671e-05, "loss": 0.99, "step": 1703 }, { "epoch": 1.0853775087607518, "grad_norm": 23.403125890186537, "learning_rate": 1.0188090446003114e-05, "loss": 1.0554, "step": 1704 }, { "epoch": 1.0860146543485187, "grad_norm": 22.69161472732119, "learning_rate": 1.0178837167798788e-05, "loss": 1.3525, "step": 1705 }, { "epoch": 1.0866517999362855, "grad_norm": 16.719882340751127, "learning_rate": 1.0169584764830285e-05, "loss": 1.209, "step": 1706 }, { "epoch": 1.0872889455240522, "grad_norm": 18.456279799711815, "learning_rate": 1.0160333246959275e-05, "loss": 1.1295, "step": 1707 }, { "epoch": 1.0879260911118191, "grad_norm": 12.075247312714357, "learning_rate": 1.015108262404648e-05, "loss": 1.3191, "step": 1708 }, { "epoch": 1.0885632366995859, "grad_norm": 18.541743037028272, "learning_rate": 1.0141832905951672e-05, "loss": 1.2493, "step": 1709 }, { "epoch": 1.0892003822873526, "grad_norm": 14.206669111687967, "learning_rate": 1.0132584102533653e-05, "loss": 1.2059, "step": 1710 }, { "epoch": 1.0898375278751196, "grad_norm": 17.536393459716336, "learning_rate": 1.012333622365026e-05, "loss": 1.1626, "step": 1711 }, { "epoch": 1.0904746734628863, "grad_norm": 35.81787311947373, "learning_rate": 1.0114089279158334e-05, "loss": 1.0325, "step": 1712 }, { "epoch": 1.091111819050653, "grad_norm": 14.162468383400421, "learning_rate": 1.0104843278913727e-05, "loss": 1.1178, "step": 1713 }, { "epoch": 1.09174896463842, "grad_norm": 48.12857722431001, "learning_rate": 1.0095598232771278e-05, "loss": 1.0574, "step": 1714 }, { "epoch": 1.0923861102261867, "grad_norm": 12.073877596490222, "learning_rate": 1.0086354150584812e-05, "loss": 0.7966, "step": 1715 }, { "epoch": 1.0930232558139534, "grad_norm": 15.925895015144155, "learning_rate": 1.0077111042207136e-05, "loss": 1.5013, "step": 1716 }, { "epoch": 1.0936604014017204, "grad_norm": 14.386722983187175, "learning_rate": 1.0067868917489996e-05, "loss": 1.0288, "step": 1717 }, { "epoch": 1.0942975469894871, "grad_norm": 11.505370081480857, "learning_rate": 1.0058627786284118e-05, "loss": 0.9436, "step": 1718 }, { "epoch": 1.0949346925772538, "grad_norm": 19.61014589986847, "learning_rate": 1.004938765843915e-05, "loss": 1.3971, "step": 1719 }, { "epoch": 1.0955718381650208, "grad_norm": 14.571158609121222, "learning_rate": 1.0040148543803675e-05, "loss": 1.4124, "step": 1720 }, { "epoch": 1.0962089837527875, "grad_norm": 15.192039637983564, "learning_rate": 1.0030910452225198e-05, "loss": 1.1833, "step": 1721 }, { "epoch": 1.0968461293405543, "grad_norm": 22.521210145495793, "learning_rate": 1.0021673393550132e-05, "loss": 1.4297, "step": 1722 }, { "epoch": 1.0974832749283212, "grad_norm": 23.766346288139953, "learning_rate": 1.0012437377623789e-05, "loss": 2.0801, "step": 1723 }, { "epoch": 1.098120420516088, "grad_norm": 25.360923282211864, "learning_rate": 1.0003202414290367e-05, "loss": 0.9269, "step": 1724 }, { "epoch": 1.0987575661038547, "grad_norm": 12.220870288623162, "learning_rate": 9.993968513392955e-06, "loss": 1.1556, "step": 1725 }, { "epoch": 1.0993947116916216, "grad_norm": 53.28973391334579, "learning_rate": 9.98473568477349e-06, "loss": 1.0505, "step": 1726 }, { "epoch": 1.1000318572793883, "grad_norm": 14.664071298391471, "learning_rate": 9.975503938272781e-06, "loss": 1.4746, "step": 1727 }, { "epoch": 1.100669002867155, "grad_norm": 17.58059236491245, "learning_rate": 9.966273283730471e-06, "loss": 0.6558, "step": 1728 }, { "epoch": 1.101306148454922, "grad_norm": 13.369739641077425, "learning_rate": 9.957043730985062e-06, "loss": 1.5627, "step": 1729 }, { "epoch": 1.1019432940426888, "grad_norm": 20.917547300602273, "learning_rate": 9.947815289873846e-06, "loss": 1.5464, "step": 1730 }, { "epoch": 1.1025804396304555, "grad_norm": 22.67669356212688, "learning_rate": 9.938587970232967e-06, "loss": 1.3367, "step": 1731 }, { "epoch": 1.1032175852182224, "grad_norm": 28.700854009473616, "learning_rate": 9.92936178189735e-06, "loss": 1.1124, "step": 1732 }, { "epoch": 1.1038547308059892, "grad_norm": 22.13742079028797, "learning_rate": 9.920136734700726e-06, "loss": 0.7887, "step": 1733 }, { "epoch": 1.104491876393756, "grad_norm": 29.525848842103617, "learning_rate": 9.910912838475597e-06, "loss": 1.1987, "step": 1734 }, { "epoch": 1.1051290219815229, "grad_norm": 18.46785231994123, "learning_rate": 9.901690103053252e-06, "loss": 1.2234, "step": 1735 }, { "epoch": 1.1057661675692896, "grad_norm": 59.649367224890895, "learning_rate": 9.892468538263735e-06, "loss": 1.3723, "step": 1736 }, { "epoch": 1.1064033131570563, "grad_norm": 29.0765507294825, "learning_rate": 9.883248153935843e-06, "loss": 1.1067, "step": 1737 }, { "epoch": 1.1070404587448233, "grad_norm": 13.824564611496903, "learning_rate": 9.874028959897123e-06, "loss": 1.3796, "step": 1738 }, { "epoch": 1.10767760433259, "grad_norm": 24.79904758765088, "learning_rate": 9.86481096597384e-06, "loss": 1.3364, "step": 1739 }, { "epoch": 1.1083147499203567, "grad_norm": 19.865627236202066, "learning_rate": 9.855594181990988e-06, "loss": 1.4946, "step": 1740 }, { "epoch": 1.1089518955081237, "grad_norm": 34.75378215370497, "learning_rate": 9.846378617772275e-06, "loss": 1.3691, "step": 1741 }, { "epoch": 1.1095890410958904, "grad_norm": 29.501528512894662, "learning_rate": 9.83716428314009e-06, "loss": 1.2012, "step": 1742 }, { "epoch": 1.1102261866836571, "grad_norm": 13.200650355079594, "learning_rate": 9.827951187915539e-06, "loss": 1.1915, "step": 1743 }, { "epoch": 1.110863332271424, "grad_norm": 11.704543159229784, "learning_rate": 9.818739341918381e-06, "loss": 1.2064, "step": 1744 }, { "epoch": 1.1115004778591908, "grad_norm": 21.45559505134991, "learning_rate": 9.809528754967067e-06, "loss": 0.9391, "step": 1745 }, { "epoch": 1.1121376234469575, "grad_norm": 17.466665265684483, "learning_rate": 9.800319436878691e-06, "loss": 1.3522, "step": 1746 }, { "epoch": 1.1127747690347245, "grad_norm": 17.110590451094417, "learning_rate": 9.791111397468998e-06, "loss": 1.2175, "step": 1747 }, { "epoch": 1.1134119146224912, "grad_norm": 14.48728693044128, "learning_rate": 9.781904646552368e-06, "loss": 1.0627, "step": 1748 }, { "epoch": 1.114049060210258, "grad_norm": 16.681909011679643, "learning_rate": 9.772699193941808e-06, "loss": 1.5488, "step": 1749 }, { "epoch": 1.114686205798025, "grad_norm": 16.76082583517391, "learning_rate": 9.763495049448949e-06, "loss": 1.5815, "step": 1750 }, { "epoch": 1.1153233513857916, "grad_norm": 22.616563695165475, "learning_rate": 9.754292222884017e-06, "loss": 0.9833, "step": 1751 }, { "epoch": 1.1159604969735584, "grad_norm": 17.167974167379345, "learning_rate": 9.74509072405584e-06, "loss": 2.043, "step": 1752 }, { "epoch": 1.1165976425613253, "grad_norm": 28.598460994044572, "learning_rate": 9.735890562771833e-06, "loss": 1.4421, "step": 1753 }, { "epoch": 1.117234788149092, "grad_norm": 59.85443663626299, "learning_rate": 9.726691748837978e-06, "loss": 1.0735, "step": 1754 }, { "epoch": 1.1178719337368588, "grad_norm": 12.544827706437914, "learning_rate": 9.717494292058822e-06, "loss": 0.7865, "step": 1755 }, { "epoch": 1.1185090793246257, "grad_norm": 15.74870802589939, "learning_rate": 9.708298202237472e-06, "loss": 1.3621, "step": 1756 }, { "epoch": 1.1191462249123925, "grad_norm": 31.518005324180688, "learning_rate": 9.699103489175567e-06, "loss": 1.842, "step": 1757 }, { "epoch": 1.1197833705001592, "grad_norm": 16.446623981836527, "learning_rate": 9.689910162673295e-06, "loss": 1.4812, "step": 1758 }, { "epoch": 1.1204205160879261, "grad_norm": 21.62602969571571, "learning_rate": 9.680718232529351e-06, "loss": 1.9092, "step": 1759 }, { "epoch": 1.1210576616756929, "grad_norm": 35.264888821722586, "learning_rate": 9.671527708540949e-06, "loss": 1.0614, "step": 1760 }, { "epoch": 1.1216948072634596, "grad_norm": 17.678575185978556, "learning_rate": 9.662338600503803e-06, "loss": 1.5933, "step": 1761 }, { "epoch": 1.1223319528512266, "grad_norm": 22.537578426221746, "learning_rate": 9.653150918212113e-06, "loss": 1.0829, "step": 1762 }, { "epoch": 1.1229690984389933, "grad_norm": 16.46414097996275, "learning_rate": 9.643964671458565e-06, "loss": 1.1632, "step": 1763 }, { "epoch": 1.12360624402676, "grad_norm": 33.545953394523735, "learning_rate": 9.63477987003432e-06, "loss": 0.7699, "step": 1764 }, { "epoch": 1.124243389614527, "grad_norm": 20.200718285032163, "learning_rate": 9.625596523728984e-06, "loss": 1.0031, "step": 1765 }, { "epoch": 1.1248805352022937, "grad_norm": 10.79182524819518, "learning_rate": 9.616414642330625e-06, "loss": 0.9805, "step": 1766 }, { "epoch": 1.1255176807900606, "grad_norm": 18.897405751798285, "learning_rate": 9.607234235625746e-06, "loss": 1.3552, "step": 1767 }, { "epoch": 1.1261548263778274, "grad_norm": 68.5181331714571, "learning_rate": 9.598055313399272e-06, "loss": 1.3997, "step": 1768 }, { "epoch": 1.126791971965594, "grad_norm": 42.48974320106021, "learning_rate": 9.588877885434556e-06, "loss": 1.2161, "step": 1769 }, { "epoch": 1.127429117553361, "grad_norm": 14.181502781077718, "learning_rate": 9.579701961513343e-06, "loss": 0.8688, "step": 1770 }, { "epoch": 1.1280662631411278, "grad_norm": 17.594236685311373, "learning_rate": 9.570527551415797e-06, "loss": 1.7737, "step": 1771 }, { "epoch": 1.1287034087288945, "grad_norm": 32.171722741035985, "learning_rate": 9.56135466492045e-06, "loss": 0.8787, "step": 1772 }, { "epoch": 1.1293405543166615, "grad_norm": 26.05368792834925, "learning_rate": 9.552183311804222e-06, "loss": 0.8617, "step": 1773 }, { "epoch": 1.1299776999044282, "grad_norm": 15.42357176725072, "learning_rate": 9.543013501842388e-06, "loss": 1.3096, "step": 1774 }, { "epoch": 1.130614845492195, "grad_norm": 22.526390982190744, "learning_rate": 9.533845244808588e-06, "loss": 1.0553, "step": 1775 }, { "epoch": 1.1312519910799619, "grad_norm": 30.087367834143556, "learning_rate": 9.524678550474797e-06, "loss": 1.6128, "step": 1776 }, { "epoch": 1.1318891366677286, "grad_norm": 15.89093789144686, "learning_rate": 9.515513428611329e-06, "loss": 1.3577, "step": 1777 }, { "epoch": 1.1325262822554953, "grad_norm": 20.722298623116732, "learning_rate": 9.506349888986833e-06, "loss": 1.5288, "step": 1778 }, { "epoch": 1.1331634278432623, "grad_norm": 19.51128774219621, "learning_rate": 9.497187941368247e-06, "loss": 1.1565, "step": 1779 }, { "epoch": 1.133800573431029, "grad_norm": 33.32970144381471, "learning_rate": 9.488027595520839e-06, "loss": 1.6006, "step": 1780 }, { "epoch": 1.1344377190187958, "grad_norm": 78.0607318854288, "learning_rate": 9.478868861208148e-06, "loss": 1.75, "step": 1781 }, { "epoch": 1.1350748646065627, "grad_norm": 16.32218075915462, "learning_rate": 9.469711748192012e-06, "loss": 1.3887, "step": 1782 }, { "epoch": 1.1357120101943294, "grad_norm": 15.480978123198211, "learning_rate": 9.460556266232525e-06, "loss": 0.6793, "step": 1783 }, { "epoch": 1.1363491557820962, "grad_norm": 8.077839520446902, "learning_rate": 9.451402425088058e-06, "loss": 0.9092, "step": 1784 }, { "epoch": 1.1369863013698631, "grad_norm": 18.3517534053024, "learning_rate": 9.442250234515223e-06, "loss": 1.7385, "step": 1785 }, { "epoch": 1.1376234469576298, "grad_norm": 29.56850969278638, "learning_rate": 9.43309970426888e-06, "loss": 1.5176, "step": 1786 }, { "epoch": 1.1382605925453966, "grad_norm": 25.538245029507852, "learning_rate": 9.42395084410211e-06, "loss": 1.1653, "step": 1787 }, { "epoch": 1.1388977381331635, "grad_norm": 17.704684217855846, "learning_rate": 9.41480366376622e-06, "loss": 0.8282, "step": 1788 }, { "epoch": 1.1395348837209303, "grad_norm": 31.693787586372267, "learning_rate": 9.40565817301073e-06, "loss": 1.5708, "step": 1789 }, { "epoch": 1.140172029308697, "grad_norm": 22.478523689714905, "learning_rate": 9.396514381583344e-06, "loss": 1.2927, "step": 1790 }, { "epoch": 1.140809174896464, "grad_norm": 16.670841086908123, "learning_rate": 9.38737229922998e-06, "loss": 1.1897, "step": 1791 }, { "epoch": 1.1414463204842307, "grad_norm": 17.57960904767866, "learning_rate": 9.37823193569471e-06, "loss": 1.0474, "step": 1792 }, { "epoch": 1.1420834660719974, "grad_norm": 20.177147749518205, "learning_rate": 9.369093300719789e-06, "loss": 1.4447, "step": 1793 }, { "epoch": 1.1427206116597644, "grad_norm": 28.179121735944562, "learning_rate": 9.359956404045625e-06, "loss": 1.0372, "step": 1794 }, { "epoch": 1.143357757247531, "grad_norm": 13.693237236242515, "learning_rate": 9.350821255410767e-06, "loss": 1.1414, "step": 1795 }, { "epoch": 1.1439949028352978, "grad_norm": 14.96497440641425, "learning_rate": 9.341687864551913e-06, "loss": 1.3015, "step": 1796 }, { "epoch": 1.1446320484230648, "grad_norm": 15.920671894708043, "learning_rate": 9.332556241203876e-06, "loss": 1.0864, "step": 1797 }, { "epoch": 1.1452691940108315, "grad_norm": 17.345059477854786, "learning_rate": 9.323426395099595e-06, "loss": 1.2493, "step": 1798 }, { "epoch": 1.1459063395985982, "grad_norm": 23.431018756628898, "learning_rate": 9.314298335970107e-06, "loss": 1.2687, "step": 1799 }, { "epoch": 1.1465434851863652, "grad_norm": 73.7826446208117, "learning_rate": 9.305172073544549e-06, "loss": 1.6433, "step": 1800 }, { "epoch": 1.147180630774132, "grad_norm": 23.193513272485575, "learning_rate": 9.296047617550144e-06, "loss": 1.1384, "step": 1801 }, { "epoch": 1.1478177763618986, "grad_norm": 23.426391373792917, "learning_rate": 9.286924977712179e-06, "loss": 1.2429, "step": 1802 }, { "epoch": 1.1484549219496656, "grad_norm": 17.67157920494736, "learning_rate": 9.277804163754019e-06, "loss": 1.3701, "step": 1803 }, { "epoch": 1.1490920675374323, "grad_norm": 18.118326865722302, "learning_rate": 9.268685185397078e-06, "loss": 0.9291, "step": 1804 }, { "epoch": 1.149729213125199, "grad_norm": 13.172087998612893, "learning_rate": 9.25956805236081e-06, "loss": 1.0358, "step": 1805 }, { "epoch": 1.150366358712966, "grad_norm": 16.96767264570214, "learning_rate": 9.250452774362704e-06, "loss": 1.0253, "step": 1806 }, { "epoch": 1.1510035043007327, "grad_norm": 17.18498217689128, "learning_rate": 9.241339361118278e-06, "loss": 1.0968, "step": 1807 }, { "epoch": 1.1516406498884995, "grad_norm": 16.805758101474197, "learning_rate": 9.23222782234105e-06, "loss": 1.4558, "step": 1808 }, { "epoch": 1.1522777954762664, "grad_norm": 20.22982684670958, "learning_rate": 9.223118167742553e-06, "loss": 1.6023, "step": 1809 }, { "epoch": 1.1529149410640331, "grad_norm": 25.1143482494465, "learning_rate": 9.214010407032299e-06, "loss": 1.295, "step": 1810 }, { "epoch": 1.1535520866517999, "grad_norm": 15.232016166222662, "learning_rate": 9.204904549917798e-06, "loss": 1.7695, "step": 1811 }, { "epoch": 1.1541892322395668, "grad_norm": 28.731533315454126, "learning_rate": 9.195800606104515e-06, "loss": 1.3748, "step": 1812 }, { "epoch": 1.1548263778273336, "grad_norm": 16.71641074000269, "learning_rate": 9.186698585295881e-06, "loss": 1.1995, "step": 1813 }, { "epoch": 1.1554635234151003, "grad_norm": 32.46532158106598, "learning_rate": 9.177598497193283e-06, "loss": 1.4961, "step": 1814 }, { "epoch": 1.1561006690028672, "grad_norm": 18.74546523130735, "learning_rate": 9.168500351496039e-06, "loss": 1.2856, "step": 1815 }, { "epoch": 1.156737814590634, "grad_norm": 16.46965883082128, "learning_rate": 9.159404157901405e-06, "loss": 1.1354, "step": 1816 }, { "epoch": 1.1573749601784007, "grad_norm": 14.7845499281058, "learning_rate": 9.150309926104548e-06, "loss": 1.2738, "step": 1817 }, { "epoch": 1.1580121057661676, "grad_norm": 16.795198046047823, "learning_rate": 9.141217665798557e-06, "loss": 1.1415, "step": 1818 }, { "epoch": 1.1586492513539344, "grad_norm": 13.583570159691417, "learning_rate": 9.132127386674404e-06, "loss": 1.1062, "step": 1819 }, { "epoch": 1.159286396941701, "grad_norm": 10.421141128677549, "learning_rate": 9.123039098420953e-06, "loss": 1.2437, "step": 1820 }, { "epoch": 1.159923542529468, "grad_norm": 17.26997583931632, "learning_rate": 9.11395281072496e-06, "loss": 1.4426, "step": 1821 }, { "epoch": 1.1605606881172348, "grad_norm": 23.328605696883002, "learning_rate": 9.104868533271034e-06, "loss": 0.9863, "step": 1822 }, { "epoch": 1.1611978337050015, "grad_norm": 14.163005428445578, "learning_rate": 9.095786275741639e-06, "loss": 1.3627, "step": 1823 }, { "epoch": 1.1618349792927685, "grad_norm": 18.100226202659897, "learning_rate": 9.086706047817104e-06, "loss": 1.4159, "step": 1824 }, { "epoch": 1.1624721248805352, "grad_norm": 11.860836669936324, "learning_rate": 9.077627859175578e-06, "loss": 1.113, "step": 1825 }, { "epoch": 1.163109270468302, "grad_norm": 17.06806338374739, "learning_rate": 9.068551719493039e-06, "loss": 0.9193, "step": 1826 }, { "epoch": 1.1637464160560689, "grad_norm": 21.728744765432403, "learning_rate": 9.05947763844329e-06, "loss": 0.9226, "step": 1827 }, { "epoch": 1.1643835616438356, "grad_norm": 17.454439970902467, "learning_rate": 9.050405625697932e-06, "loss": 1.8135, "step": 1828 }, { "epoch": 1.1650207072316023, "grad_norm": 29.151422254209667, "learning_rate": 9.04133569092636e-06, "loss": 1.6289, "step": 1829 }, { "epoch": 1.1656578528193693, "grad_norm": 21.44036870538139, "learning_rate": 9.03226784379576e-06, "loss": 1.5195, "step": 1830 }, { "epoch": 1.166294998407136, "grad_norm": 39.87991774671935, "learning_rate": 9.023202093971091e-06, "loss": 1.3181, "step": 1831 }, { "epoch": 1.1669321439949027, "grad_norm": 36.835376617690066, "learning_rate": 9.014138451115078e-06, "loss": 1.1226, "step": 1832 }, { "epoch": 1.1675692895826697, "grad_norm": 19.867978368343152, "learning_rate": 9.005076924888195e-06, "loss": 0.7682, "step": 1833 }, { "epoch": 1.1682064351704364, "grad_norm": 13.334440784341309, "learning_rate": 8.996017524948662e-06, "loss": 1.1086, "step": 1834 }, { "epoch": 1.1688435807582032, "grad_norm": 15.969594890041792, "learning_rate": 8.98696026095244e-06, "loss": 1.5143, "step": 1835 }, { "epoch": 1.1694807263459701, "grad_norm": 53.8270100051073, "learning_rate": 8.9779051425532e-06, "loss": 1.8289, "step": 1836 }, { "epoch": 1.1701178719337368, "grad_norm": 32.982974672420525, "learning_rate": 8.968852179402336e-06, "loss": 1.0618, "step": 1837 }, { "epoch": 1.1707550175215036, "grad_norm": 28.792269231311945, "learning_rate": 8.959801381148949e-06, "loss": 1.4312, "step": 1838 }, { "epoch": 1.1713921631092705, "grad_norm": 31.699072315700143, "learning_rate": 8.95075275743982e-06, "loss": 1.5618, "step": 1839 }, { "epoch": 1.1720293086970373, "grad_norm": 16.657881900841907, "learning_rate": 8.941706317919418e-06, "loss": 1.6902, "step": 1840 }, { "epoch": 1.172666454284804, "grad_norm": 35.62713575229665, "learning_rate": 8.932662072229884e-06, "loss": 1.1279, "step": 1841 }, { "epoch": 1.173303599872571, "grad_norm": 17.656752964944314, "learning_rate": 8.923620030011023e-06, "loss": 0.9291, "step": 1842 }, { "epoch": 1.1739407454603377, "grad_norm": 64.18029503044514, "learning_rate": 8.914580200900285e-06, "loss": 0.9545, "step": 1843 }, { "epoch": 1.1745778910481044, "grad_norm": 12.710703300352149, "learning_rate": 8.905542594532773e-06, "loss": 1.61, "step": 1844 }, { "epoch": 1.1752150366358713, "grad_norm": 41.41564968955933, "learning_rate": 8.896507220541209e-06, "loss": 1.4303, "step": 1845 }, { "epoch": 1.175852182223638, "grad_norm": 19.089589773492836, "learning_rate": 8.887474088555939e-06, "loss": 0.8831, "step": 1846 }, { "epoch": 1.1764893278114048, "grad_norm": 26.28844731489874, "learning_rate": 8.878443208204919e-06, "loss": 1.0542, "step": 1847 }, { "epoch": 1.1771264733991718, "grad_norm": 14.877487709054938, "learning_rate": 8.869414589113712e-06, "loss": 1.3125, "step": 1848 }, { "epoch": 1.1777636189869385, "grad_norm": 27.659236508907867, "learning_rate": 8.860388240905462e-06, "loss": 1.3365, "step": 1849 }, { "epoch": 1.1784007645747052, "grad_norm": 18.956981837976052, "learning_rate": 8.851364173200892e-06, "loss": 1.5415, "step": 1850 }, { "epoch": 1.1790379101624722, "grad_norm": 25.36925796944392, "learning_rate": 8.842342395618307e-06, "loss": 1.123, "step": 1851 }, { "epoch": 1.179675055750239, "grad_norm": 23.559306314696325, "learning_rate": 8.833322917773557e-06, "loss": 0.7867, "step": 1852 }, { "epoch": 1.1803122013380056, "grad_norm": 12.328484638280631, "learning_rate": 8.824305749280046e-06, "loss": 1.6157, "step": 1853 }, { "epoch": 1.1809493469257726, "grad_norm": 15.503022266722018, "learning_rate": 8.815290899748716e-06, "loss": 0.9601, "step": 1854 }, { "epoch": 1.1815864925135393, "grad_norm": 19.63721335213346, "learning_rate": 8.806278378788042e-06, "loss": 1.9878, "step": 1855 }, { "epoch": 1.182223638101306, "grad_norm": 27.514682326327584, "learning_rate": 8.797268196004009e-06, "loss": 0.9189, "step": 1856 }, { "epoch": 1.182860783689073, "grad_norm": 12.366384397027506, "learning_rate": 8.788260361000113e-06, "loss": 1.0806, "step": 1857 }, { "epoch": 1.1834979292768397, "grad_norm": 10.931295849365103, "learning_rate": 8.779254883377356e-06, "loss": 1.1006, "step": 1858 }, { "epoch": 1.1841350748646065, "grad_norm": 21.409133594750536, "learning_rate": 8.770251772734213e-06, "loss": 1.2146, "step": 1859 }, { "epoch": 1.1847722204523734, "grad_norm": 16.879476709355753, "learning_rate": 8.761251038666645e-06, "loss": 1.2991, "step": 1860 }, { "epoch": 1.1854093660401401, "grad_norm": 14.675022533395854, "learning_rate": 8.752252690768077e-06, "loss": 0.7166, "step": 1861 }, { "epoch": 1.1860465116279069, "grad_norm": 28.589744323483597, "learning_rate": 8.743256738629394e-06, "loss": 1.6865, "step": 1862 }, { "epoch": 1.1866836572156738, "grad_norm": 18.638426634281984, "learning_rate": 8.734263191838921e-06, "loss": 0.7351, "step": 1863 }, { "epoch": 1.1873208028034405, "grad_norm": 9.850890026523555, "learning_rate": 8.72527205998243e-06, "loss": 1.0576, "step": 1864 }, { "epoch": 1.1879579483912073, "grad_norm": 11.060944543303671, "learning_rate": 8.71628335264311e-06, "loss": 1.0044, "step": 1865 }, { "epoch": 1.1885950939789742, "grad_norm": 19.909989378477608, "learning_rate": 8.707297079401567e-06, "loss": 1.7319, "step": 1866 }, { "epoch": 1.189232239566741, "grad_norm": 15.92757657036251, "learning_rate": 8.698313249835812e-06, "loss": 0.9789, "step": 1867 }, { "epoch": 1.1898693851545077, "grad_norm": 19.53899416792729, "learning_rate": 8.689331873521258e-06, "loss": 1.2434, "step": 1868 }, { "epoch": 1.1905065307422746, "grad_norm": 14.620554877622936, "learning_rate": 8.680352960030698e-06, "loss": 1.2197, "step": 1869 }, { "epoch": 1.1911436763300414, "grad_norm": 25.533463553369675, "learning_rate": 8.671376518934294e-06, "loss": 1.3599, "step": 1870 }, { "epoch": 1.191780821917808, "grad_norm": 21.02987424464482, "learning_rate": 8.66240255979959e-06, "loss": 1.3295, "step": 1871 }, { "epoch": 1.192417967505575, "grad_norm": 14.618171496090643, "learning_rate": 8.653431092191466e-06, "loss": 1.0653, "step": 1872 }, { "epoch": 1.1930551130933418, "grad_norm": 142.35534693717815, "learning_rate": 8.644462125672157e-06, "loss": 1.3651, "step": 1873 }, { "epoch": 1.1936922586811085, "grad_norm": 30.041431884468594, "learning_rate": 8.635495669801232e-06, "loss": 1.0618, "step": 1874 }, { "epoch": 1.1943294042688755, "grad_norm": 11.673620704568046, "learning_rate": 8.626531734135577e-06, "loss": 1.5164, "step": 1875 }, { "epoch": 1.1949665498566422, "grad_norm": 14.170074759361416, "learning_rate": 8.6175703282294e-06, "loss": 1.6387, "step": 1876 }, { "epoch": 1.1956036954444091, "grad_norm": 25.37231314940131, "learning_rate": 8.608611461634206e-06, "loss": 1.7007, "step": 1877 }, { "epoch": 1.1962408410321759, "grad_norm": 15.078198454218079, "learning_rate": 8.599655143898799e-06, "loss": 1.2832, "step": 1878 }, { "epoch": 1.1968779866199426, "grad_norm": 20.263022312660368, "learning_rate": 8.590701384569269e-06, "loss": 1.1005, "step": 1879 }, { "epoch": 1.1975151322077096, "grad_norm": 18.555366439950053, "learning_rate": 8.581750193188965e-06, "loss": 1.6786, "step": 1880 }, { "epoch": 1.1981522777954763, "grad_norm": 16.361437303851794, "learning_rate": 8.57280157929851e-06, "loss": 1.2843, "step": 1881 }, { "epoch": 1.198789423383243, "grad_norm": 17.909727469568402, "learning_rate": 8.56385555243578e-06, "loss": 1.1627, "step": 1882 }, { "epoch": 1.19942656897101, "grad_norm": 15.227490470787286, "learning_rate": 8.55491212213589e-06, "loss": 1.126, "step": 1883 }, { "epoch": 1.2000637145587767, "grad_norm": 13.612352907980895, "learning_rate": 8.545971297931193e-06, "loss": 1.446, "step": 1884 }, { "epoch": 1.2007008601465434, "grad_norm": 52.6334539552954, "learning_rate": 8.537033089351256e-06, "loss": 1.5122, "step": 1885 }, { "epoch": 1.2013380057343104, "grad_norm": 16.401211864744905, "learning_rate": 8.528097505922863e-06, "loss": 1.1984, "step": 1886 }, { "epoch": 1.201975151322077, "grad_norm": 14.8901266164447, "learning_rate": 8.519164557169998e-06, "loss": 1.2122, "step": 1887 }, { "epoch": 1.2026122969098438, "grad_norm": 16.042356717448722, "learning_rate": 8.510234252613843e-06, "loss": 0.7169, "step": 1888 }, { "epoch": 1.2032494424976108, "grad_norm": 20.754513067624742, "learning_rate": 8.501306601772752e-06, "loss": 1.7644, "step": 1889 }, { "epoch": 1.2038865880853775, "grad_norm": 29.46006881293003, "learning_rate": 8.492381614162253e-06, "loss": 1.0916, "step": 1890 }, { "epoch": 1.2045237336731442, "grad_norm": 13.339160421301292, "learning_rate": 8.483459299295045e-06, "loss": 0.9351, "step": 1891 }, { "epoch": 1.2051608792609112, "grad_norm": 13.351351900599052, "learning_rate": 8.474539666680968e-06, "loss": 1.3145, "step": 1892 }, { "epoch": 1.205798024848678, "grad_norm": 18.312061999592753, "learning_rate": 8.465622725827006e-06, "loss": 1.3279, "step": 1893 }, { "epoch": 1.2064351704364447, "grad_norm": 12.777063523110078, "learning_rate": 8.456708486237273e-06, "loss": 1.2407, "step": 1894 }, { "epoch": 1.2070723160242116, "grad_norm": 19.5002684246185, "learning_rate": 8.447796957413009e-06, "loss": 1.6658, "step": 1895 }, { "epoch": 1.2077094616119783, "grad_norm": 28.823678977967393, "learning_rate": 8.438888148852551e-06, "loss": 1.6333, "step": 1896 }, { "epoch": 1.208346607199745, "grad_norm": 23.26480485294834, "learning_rate": 8.429982070051354e-06, "loss": 0.9536, "step": 1897 }, { "epoch": 1.208983752787512, "grad_norm": 28.10942175601999, "learning_rate": 8.421078730501957e-06, "loss": 1.6287, "step": 1898 }, { "epoch": 1.2096208983752788, "grad_norm": 24.721310410034736, "learning_rate": 8.412178139693977e-06, "loss": 1.5272, "step": 1899 }, { "epoch": 1.2102580439630455, "grad_norm": 29.907643055648165, "learning_rate": 8.403280307114099e-06, "loss": 1.0469, "step": 1900 }, { "epoch": 1.2108951895508124, "grad_norm": 19.41364255308731, "learning_rate": 8.394385242246078e-06, "loss": 1.6875, "step": 1901 }, { "epoch": 1.2115323351385792, "grad_norm": 23.9410420469222, "learning_rate": 8.385492954570705e-06, "loss": 1.3441, "step": 1902 }, { "epoch": 1.212169480726346, "grad_norm": 17.241060974825174, "learning_rate": 8.37660345356582e-06, "loss": 1.5016, "step": 1903 }, { "epoch": 1.2128066263141128, "grad_norm": 23.253968651682598, "learning_rate": 8.367716748706296e-06, "loss": 1.8364, "step": 1904 }, { "epoch": 1.2134437719018796, "grad_norm": 17.611845717627332, "learning_rate": 8.358832849464018e-06, "loss": 1.3052, "step": 1905 }, { "epoch": 1.2140809174896463, "grad_norm": 29.14946179057472, "learning_rate": 8.349951765307884e-06, "loss": 1.3254, "step": 1906 }, { "epoch": 1.2147180630774133, "grad_norm": 17.102446941223604, "learning_rate": 8.341073505703794e-06, "loss": 1.5059, "step": 1907 }, { "epoch": 1.21535520866518, "grad_norm": 22.8230841843699, "learning_rate": 8.33219808011463e-06, "loss": 1.0753, "step": 1908 }, { "epoch": 1.2159923542529467, "grad_norm": 15.236755378611305, "learning_rate": 8.323325498000264e-06, "loss": 1.0929, "step": 1909 }, { "epoch": 1.2166294998407137, "grad_norm": 14.87266719559423, "learning_rate": 8.31445576881752e-06, "loss": 0.9543, "step": 1910 }, { "epoch": 1.2172666454284804, "grad_norm": 15.907043083689393, "learning_rate": 8.305588902020205e-06, "loss": 1.4458, "step": 1911 }, { "epoch": 1.2179037910162471, "grad_norm": 19.46226117145178, "learning_rate": 8.296724907059057e-06, "loss": 1.4502, "step": 1912 }, { "epoch": 1.218540936604014, "grad_norm": 35.52111767427054, "learning_rate": 8.287863793381761e-06, "loss": 1.1919, "step": 1913 }, { "epoch": 1.2191780821917808, "grad_norm": 19.663813189011478, "learning_rate": 8.279005570432927e-06, "loss": 1.5453, "step": 1914 }, { "epoch": 1.2198152277795475, "grad_norm": 32.83268629975652, "learning_rate": 8.270150247654088e-06, "loss": 1.7407, "step": 1915 }, { "epoch": 1.2204523733673145, "grad_norm": 23.386069858880752, "learning_rate": 8.26129783448368e-06, "loss": 1.3496, "step": 1916 }, { "epoch": 1.2210895189550812, "grad_norm": 14.310651085840012, "learning_rate": 8.252448340357043e-06, "loss": 1.4441, "step": 1917 }, { "epoch": 1.221726664542848, "grad_norm": 16.223196385235262, "learning_rate": 8.243601774706403e-06, "loss": 0.998, "step": 1918 }, { "epoch": 1.222363810130615, "grad_norm": 13.092550425173053, "learning_rate": 8.234758146960868e-06, "loss": 1.5078, "step": 1919 }, { "epoch": 1.2230009557183816, "grad_norm": 13.297209875165223, "learning_rate": 8.225917466546413e-06, "loss": 1.1316, "step": 1920 }, { "epoch": 1.2236381013061484, "grad_norm": 16.0189129277461, "learning_rate": 8.21707974288587e-06, "loss": 0.9211, "step": 1921 }, { "epoch": 1.2242752468939153, "grad_norm": 11.039722696513687, "learning_rate": 8.208244985398919e-06, "loss": 0.7841, "step": 1922 }, { "epoch": 1.224912392481682, "grad_norm": 28.186495392104604, "learning_rate": 8.199413203502075e-06, "loss": 1.5403, "step": 1923 }, { "epoch": 1.2255495380694488, "grad_norm": 20.53521194439237, "learning_rate": 8.190584406608702e-06, "loss": 1.4084, "step": 1924 }, { "epoch": 1.2261866836572157, "grad_norm": 19.84732359169261, "learning_rate": 8.181758604128952e-06, "loss": 1.1964, "step": 1925 }, { "epoch": 1.2268238292449825, "grad_norm": 30.804995643003497, "learning_rate": 8.172935805469806e-06, "loss": 1.1323, "step": 1926 }, { "epoch": 1.2274609748327494, "grad_norm": 24.387382284158708, "learning_rate": 8.16411602003504e-06, "loss": 1.6274, "step": 1927 }, { "epoch": 1.2280981204205161, "grad_norm": 20.966587532756726, "learning_rate": 8.15529925722521e-06, "loss": 1.5801, "step": 1928 }, { "epoch": 1.2287352660082829, "grad_norm": 17.683109088112026, "learning_rate": 8.146485526437665e-06, "loss": 1.6628, "step": 1929 }, { "epoch": 1.2293724115960498, "grad_norm": 13.734861187550994, "learning_rate": 8.137674837066502e-06, "loss": 1.1101, "step": 1930 }, { "epoch": 1.2300095571838165, "grad_norm": 18.063747784742347, "learning_rate": 8.128867198502599e-06, "loss": 1.1423, "step": 1931 }, { "epoch": 1.2306467027715833, "grad_norm": 17.407742567992084, "learning_rate": 8.120062620133564e-06, "loss": 1.6144, "step": 1932 }, { "epoch": 1.2312838483593502, "grad_norm": 22.09396748047628, "learning_rate": 8.111261111343756e-06, "loss": 0.8585, "step": 1933 }, { "epoch": 1.231920993947117, "grad_norm": 17.93411129863866, "learning_rate": 8.102462681514256e-06, "loss": 1.2588, "step": 1934 }, { "epoch": 1.2325581395348837, "grad_norm": 18.544596078097985, "learning_rate": 8.093667340022862e-06, "loss": 1.1736, "step": 1935 }, { "epoch": 1.2331952851226506, "grad_norm": 20.145358615608306, "learning_rate": 8.084875096244084e-06, "loss": 1.4397, "step": 1936 }, { "epoch": 1.2338324307104174, "grad_norm": 76.85851866632409, "learning_rate": 8.076085959549122e-06, "loss": 1.7278, "step": 1937 }, { "epoch": 1.234469576298184, "grad_norm": 20.063122689534683, "learning_rate": 8.067299939305886e-06, "loss": 1.626, "step": 1938 }, { "epoch": 1.235106721885951, "grad_norm": 21.898026614667106, "learning_rate": 8.058517044878935e-06, "loss": 1.8389, "step": 1939 }, { "epoch": 1.2357438674737178, "grad_norm": 19.315256563590616, "learning_rate": 8.04973728562952e-06, "loss": 1.6729, "step": 1940 }, { "epoch": 1.2363810130614845, "grad_norm": 17.90420817371799, "learning_rate": 8.040960670915536e-06, "loss": 1.0569, "step": 1941 }, { "epoch": 1.2370181586492515, "grad_norm": 37.03097473252867, "learning_rate": 8.032187210091537e-06, "loss": 1.5339, "step": 1942 }, { "epoch": 1.2376553042370182, "grad_norm": 21.31307753988263, "learning_rate": 8.023416912508701e-06, "loss": 1.2289, "step": 1943 }, { "epoch": 1.238292449824785, "grad_norm": 22.083619955848864, "learning_rate": 8.014649787514853e-06, "loss": 1.2101, "step": 1944 }, { "epoch": 1.2389295954125519, "grad_norm": 11.739443441516308, "learning_rate": 8.005885844454421e-06, "loss": 1.2838, "step": 1945 }, { "epoch": 1.2395667410003186, "grad_norm": 24.926662370641797, "learning_rate": 7.997125092668452e-06, "loss": 1.1387, "step": 1946 }, { "epoch": 1.2402038865880853, "grad_norm": 16.700710523181375, "learning_rate": 7.988367541494585e-06, "loss": 1.5237, "step": 1947 }, { "epoch": 1.2408410321758523, "grad_norm": 18.200880376447884, "learning_rate": 7.979613200267047e-06, "loss": 1.7837, "step": 1948 }, { "epoch": 1.241478177763619, "grad_norm": 19.776093661373, "learning_rate": 7.97086207831665e-06, "loss": 0.9766, "step": 1949 }, { "epoch": 1.2421153233513857, "grad_norm": 14.810241373305804, "learning_rate": 7.962114184970767e-06, "loss": 1.7422, "step": 1950 }, { "epoch": 1.2427524689391527, "grad_norm": 37.37618330820152, "learning_rate": 7.953369529553341e-06, "loss": 1.4788, "step": 1951 }, { "epoch": 1.2433896145269194, "grad_norm": 15.303344031901347, "learning_rate": 7.944628121384853e-06, "loss": 1.1608, "step": 1952 }, { "epoch": 1.2440267601146862, "grad_norm": 29.927097600762803, "learning_rate": 7.93588996978232e-06, "loss": 1.3043, "step": 1953 }, { "epoch": 1.244663905702453, "grad_norm": 20.36701563401396, "learning_rate": 7.927155084059309e-06, "loss": 1.4146, "step": 1954 }, { "epoch": 1.2453010512902198, "grad_norm": 17.183218629831007, "learning_rate": 7.91842347352588e-06, "loss": 1.5344, "step": 1955 }, { "epoch": 1.2459381968779866, "grad_norm": 25.06416694617986, "learning_rate": 7.90969514748862e-06, "loss": 1.3441, "step": 1956 }, { "epoch": 1.2465753424657535, "grad_norm": 11.884015124591476, "learning_rate": 7.900970115250602e-06, "loss": 1.0415, "step": 1957 }, { "epoch": 1.2472124880535203, "grad_norm": 36.22443892294525, "learning_rate": 7.892248386111406e-06, "loss": 1.4272, "step": 1958 }, { "epoch": 1.247849633641287, "grad_norm": 13.827604877994448, "learning_rate": 7.883529969367074e-06, "loss": 1.0334, "step": 1959 }, { "epoch": 1.248486779229054, "grad_norm": 13.565450649419871, "learning_rate": 7.874814874310123e-06, "loss": 1.5444, "step": 1960 }, { "epoch": 1.2491239248168207, "grad_norm": 12.795027393489745, "learning_rate": 7.866103110229538e-06, "loss": 1.1781, "step": 1961 }, { "epoch": 1.2497610704045874, "grad_norm": 13.222013667940791, "learning_rate": 7.85739468641074e-06, "loss": 1.2161, "step": 1962 }, { "epoch": 1.2503982159923543, "grad_norm": 21.61377135944457, "learning_rate": 7.848689612135596e-06, "loss": 1.5603, "step": 1963 }, { "epoch": 1.251035361580121, "grad_norm": 37.72814279236647, "learning_rate": 7.83998789668241e-06, "loss": 1.1202, "step": 1964 }, { "epoch": 1.2516725071678878, "grad_norm": 46.27287893930173, "learning_rate": 7.831289549325893e-06, "loss": 1.6641, "step": 1965 }, { "epoch": 1.2523096527556548, "grad_norm": 19.57500084352955, "learning_rate": 7.822594579337177e-06, "loss": 0.8746, "step": 1966 }, { "epoch": 1.2529467983434215, "grad_norm": 7.039887963521517, "learning_rate": 7.813902995983785e-06, "loss": 0.5032, "step": 1967 }, { "epoch": 1.2535839439311882, "grad_norm": 17.55728016065635, "learning_rate": 7.805214808529638e-06, "loss": 1.4148, "step": 1968 }, { "epoch": 1.2542210895189552, "grad_norm": 43.53658084196563, "learning_rate": 7.796530026235034e-06, "loss": 1.6116, "step": 1969 }, { "epoch": 1.254858235106722, "grad_norm": 17.17347039791626, "learning_rate": 7.787848658356638e-06, "loss": 1.176, "step": 1970 }, { "epoch": 1.2554953806944886, "grad_norm": 16.090812760997643, "learning_rate": 7.779170714147486e-06, "loss": 1.0468, "step": 1971 }, { "epoch": 1.2561325262822556, "grad_norm": 22.088230389635914, "learning_rate": 7.77049620285696e-06, "loss": 1.0281, "step": 1972 }, { "epoch": 1.2567696718700223, "grad_norm": 21.90447474847686, "learning_rate": 7.761825133730774e-06, "loss": 1.3809, "step": 1973 }, { "epoch": 1.257406817457789, "grad_norm": 16.618117066158504, "learning_rate": 7.753157516010985e-06, "loss": 1.099, "step": 1974 }, { "epoch": 1.258043963045556, "grad_norm": 20.49979209635337, "learning_rate": 7.744493358935967e-06, "loss": 1.8086, "step": 1975 }, { "epoch": 1.2586811086333227, "grad_norm": 18.65237519772387, "learning_rate": 7.73583267174041e-06, "loss": 1.0114, "step": 1976 }, { "epoch": 1.2593182542210895, "grad_norm": 17.848059206493225, "learning_rate": 7.72717546365529e-06, "loss": 1.552, "step": 1977 }, { "epoch": 1.2599553998088564, "grad_norm": 17.333658798779915, "learning_rate": 7.718521743907897e-06, "loss": 1.1104, "step": 1978 }, { "epoch": 1.2605925453966231, "grad_norm": 13.363059534130969, "learning_rate": 7.70987152172179e-06, "loss": 1.1913, "step": 1979 }, { "epoch": 1.2612296909843899, "grad_norm": 26.33508200029296, "learning_rate": 7.701224806316798e-06, "loss": 1.2153, "step": 1980 }, { "epoch": 1.2618668365721568, "grad_norm": 11.597282044324498, "learning_rate": 7.69258160690902e-06, "loss": 0.9539, "step": 1981 }, { "epoch": 1.2625039821599235, "grad_norm": 21.49355194586029, "learning_rate": 7.683941932710806e-06, "loss": 1.027, "step": 1982 }, { "epoch": 1.2631411277476903, "grad_norm": 16.984241705352684, "learning_rate": 7.675305792930737e-06, "loss": 0.9958, "step": 1983 }, { "epoch": 1.2637782733354572, "grad_norm": 15.146136571701966, "learning_rate": 7.666673196773648e-06, "loss": 1.3601, "step": 1984 }, { "epoch": 1.264415418923224, "grad_norm": 25.647307934935323, "learning_rate": 7.65804415344058e-06, "loss": 1.605, "step": 1985 }, { "epoch": 1.2650525645109907, "grad_norm": 15.38398587636435, "learning_rate": 7.649418672128795e-06, "loss": 0.7827, "step": 1986 }, { "epoch": 1.2656897100987576, "grad_norm": 12.089561547184791, "learning_rate": 7.640796762031753e-06, "loss": 1.3809, "step": 1987 }, { "epoch": 1.2663268556865244, "grad_norm": 28.619577577466572, "learning_rate": 7.632178432339112e-06, "loss": 1.2677, "step": 1988 }, { "epoch": 1.266964001274291, "grad_norm": 20.836807764878184, "learning_rate": 7.623563692236715e-06, "loss": 1.1888, "step": 1989 }, { "epoch": 1.267601146862058, "grad_norm": 28.36680460435664, "learning_rate": 7.614952550906572e-06, "loss": 1.5166, "step": 1990 }, { "epoch": 1.2682382924498248, "grad_norm": 14.543000444834128, "learning_rate": 7.606345017526866e-06, "loss": 1.0933, "step": 1991 }, { "epoch": 1.2688754380375915, "grad_norm": 23.354217401284203, "learning_rate": 7.5977411012719315e-06, "loss": 1.4958, "step": 1992 }, { "epoch": 1.2695125836253585, "grad_norm": 27.55606371857269, "learning_rate": 7.589140811312243e-06, "loss": 1.1882, "step": 1993 }, { "epoch": 1.2701497292131252, "grad_norm": 18.680408214513943, "learning_rate": 7.580544156814411e-06, "loss": 1.0774, "step": 1994 }, { "epoch": 1.270786874800892, "grad_norm": 20.068935832967853, "learning_rate": 7.571951146941183e-06, "loss": 1.3276, "step": 1995 }, { "epoch": 1.2714240203886589, "grad_norm": 18.265811224812786, "learning_rate": 7.5633617908514045e-06, "loss": 1.5073, "step": 1996 }, { "epoch": 1.2720611659764256, "grad_norm": 36.290817707395256, "learning_rate": 7.554776097700034e-06, "loss": 1.175, "step": 1997 }, { "epoch": 1.2726983115641923, "grad_norm": 17.73215811957424, "learning_rate": 7.546194076638132e-06, "loss": 1.4448, "step": 1998 }, { "epoch": 1.2733354571519593, "grad_norm": 14.993184339442939, "learning_rate": 7.537615736812837e-06, "loss": 1.1548, "step": 1999 }, { "epoch": 1.273972602739726, "grad_norm": 25.49261674750794, "learning_rate": 7.529041087367366e-06, "loss": 1.4041, "step": 2000 }, { "epoch": 1.2746097483274927, "grad_norm": 22.68154491060999, "learning_rate": 7.520470137441005e-06, "loss": 1.0321, "step": 2001 }, { "epoch": 1.2752468939152597, "grad_norm": 14.040807552979459, "learning_rate": 7.511902896169092e-06, "loss": 1.0883, "step": 2002 }, { "epoch": 1.2758840395030264, "grad_norm": 22.125550518572513, "learning_rate": 7.503339372683014e-06, "loss": 1.499, "step": 2003 }, { "epoch": 1.2765211850907932, "grad_norm": 23.450585819735053, "learning_rate": 7.494779576110203e-06, "loss": 1.3037, "step": 2004 }, { "epoch": 1.27715833067856, "grad_norm": 50.62092729953359, "learning_rate": 7.486223515574107e-06, "loss": 1.2085, "step": 2005 }, { "epoch": 1.2777954762663268, "grad_norm": 14.614473787905393, "learning_rate": 7.4776712001941985e-06, "loss": 1.6553, "step": 2006 }, { "epoch": 1.2784326218540936, "grad_norm": 18.4352835566885, "learning_rate": 7.469122639085958e-06, "loss": 1.0117, "step": 2007 }, { "epoch": 1.2790697674418605, "grad_norm": 16.1296419492854, "learning_rate": 7.460577841360856e-06, "loss": 1.0179, "step": 2008 }, { "epoch": 1.2797069130296272, "grad_norm": 35.87518589325864, "learning_rate": 7.452036816126371e-06, "loss": 1.5435, "step": 2009 }, { "epoch": 1.280344058617394, "grad_norm": 15.928613095502602, "learning_rate": 7.443499572485937e-06, "loss": 1.5088, "step": 2010 }, { "epoch": 1.280981204205161, "grad_norm": 20.72907302451433, "learning_rate": 7.434966119538978e-06, "loss": 1.5918, "step": 2011 }, { "epoch": 1.2816183497929277, "grad_norm": 20.270259327764276, "learning_rate": 7.426436466380869e-06, "loss": 1.1416, "step": 2012 }, { "epoch": 1.2822554953806944, "grad_norm": 23.936505343061146, "learning_rate": 7.417910622102926e-06, "loss": 1.594, "step": 2013 }, { "epoch": 1.2828926409684613, "grad_norm": 19.75310050333401, "learning_rate": 7.409388595792424e-06, "loss": 1.3755, "step": 2014 }, { "epoch": 1.283529786556228, "grad_norm": 15.76382275451788, "learning_rate": 7.400870396532551e-06, "loss": 1.2512, "step": 2015 }, { "epoch": 1.2841669321439948, "grad_norm": 11.710934435806239, "learning_rate": 7.392356033402428e-06, "loss": 1.1123, "step": 2016 }, { "epoch": 1.2848040777317618, "grad_norm": 13.307682419784653, "learning_rate": 7.383845515477079e-06, "loss": 1.3431, "step": 2017 }, { "epoch": 1.2854412233195285, "grad_norm": 25.454058983281328, "learning_rate": 7.37533885182744e-06, "loss": 1.2612, "step": 2018 }, { "epoch": 1.2860783689072952, "grad_norm": 21.79126547751339, "learning_rate": 7.36683605152033e-06, "loss": 1.4292, "step": 2019 }, { "epoch": 1.2867155144950622, "grad_norm": 9.087677551286749, "learning_rate": 7.358337123618449e-06, "loss": 1.0538, "step": 2020 }, { "epoch": 1.287352660082829, "grad_norm": 26.614844798861995, "learning_rate": 7.349842077180376e-06, "loss": 1.2449, "step": 2021 }, { "epoch": 1.2879898056705956, "grad_norm": 25.69074210778717, "learning_rate": 7.341350921260551e-06, "loss": 1.4612, "step": 2022 }, { "epoch": 1.2886269512583626, "grad_norm": 36.39633656567229, "learning_rate": 7.3328636649092645e-06, "loss": 0.9644, "step": 2023 }, { "epoch": 1.2892640968461293, "grad_norm": 13.95008706495136, "learning_rate": 7.324380317172653e-06, "loss": 0.7449, "step": 2024 }, { "epoch": 1.289901242433896, "grad_norm": 14.722669534567013, "learning_rate": 7.315900887092691e-06, "loss": 1.3456, "step": 2025 }, { "epoch": 1.290538388021663, "grad_norm": 19.78389636104912, "learning_rate": 7.3074253837071675e-06, "loss": 1.6157, "step": 2026 }, { "epoch": 1.2911755336094297, "grad_norm": 18.760688424478346, "learning_rate": 7.298953816049694e-06, "loss": 1.7754, "step": 2027 }, { "epoch": 1.2918126791971964, "grad_norm": 25.470771216660356, "learning_rate": 7.290486193149684e-06, "loss": 1.5784, "step": 2028 }, { "epoch": 1.2924498247849634, "grad_norm": 22.3378022068526, "learning_rate": 7.282022524032348e-06, "loss": 1.4641, "step": 2029 }, { "epoch": 1.2930869703727301, "grad_norm": 15.991989154735492, "learning_rate": 7.273562817718682e-06, "loss": 1.5469, "step": 2030 }, { "epoch": 1.2937241159604969, "grad_norm": 19.160069224234334, "learning_rate": 7.26510708322546e-06, "loss": 0.8649, "step": 2031 }, { "epoch": 1.2943612615482638, "grad_norm": 15.584712895098502, "learning_rate": 7.256655329565216e-06, "loss": 0.8362, "step": 2032 }, { "epoch": 1.2949984071360305, "grad_norm": 25.93649558262536, "learning_rate": 7.24820756574625e-06, "loss": 1.1956, "step": 2033 }, { "epoch": 1.2956355527237973, "grad_norm": 13.345399112895453, "learning_rate": 7.2397638007726055e-06, "loss": 1.1442, "step": 2034 }, { "epoch": 1.2962726983115642, "grad_norm": 15.217086293245297, "learning_rate": 7.231324043644059e-06, "loss": 1.2827, "step": 2035 }, { "epoch": 1.296909843899331, "grad_norm": 13.47178854825813, "learning_rate": 7.222888303356122e-06, "loss": 1.1409, "step": 2036 }, { "epoch": 1.2975469894870977, "grad_norm": 12.485438153272963, "learning_rate": 7.2144565889000206e-06, "loss": 1.2236, "step": 2037 }, { "epoch": 1.2981841350748646, "grad_norm": 14.511731512666522, "learning_rate": 7.206028909262699e-06, "loss": 1.6392, "step": 2038 }, { "epoch": 1.2988212806626314, "grad_norm": 25.01338120519881, "learning_rate": 7.197605273426782e-06, "loss": 1.0345, "step": 2039 }, { "epoch": 1.299458426250398, "grad_norm": 11.569054147484211, "learning_rate": 7.189185690370603e-06, "loss": 1.4331, "step": 2040 }, { "epoch": 1.300095571838165, "grad_norm": 20.843031532458088, "learning_rate": 7.180770169068174e-06, "loss": 1.2496, "step": 2041 }, { "epoch": 1.3007327174259318, "grad_norm": 11.70335207371257, "learning_rate": 7.17235871848916e-06, "loss": 1.22, "step": 2042 }, { "epoch": 1.3013698630136985, "grad_norm": 15.007983214679186, "learning_rate": 7.1639513475989075e-06, "loss": 0.9237, "step": 2043 }, { "epoch": 1.3020070086014655, "grad_norm": 17.336711331235076, "learning_rate": 7.155548065358402e-06, "loss": 0.9055, "step": 2044 }, { "epoch": 1.3026441541892322, "grad_norm": 9.400213675125604, "learning_rate": 7.147148880724284e-06, "loss": 1.1803, "step": 2045 }, { "epoch": 1.303281299776999, "grad_norm": 22.3053132283713, "learning_rate": 7.13875380264881e-06, "loss": 1.5259, "step": 2046 }, { "epoch": 1.3039184453647659, "grad_norm": 19.27988274703124, "learning_rate": 7.130362840079874e-06, "loss": 1.2046, "step": 2047 }, { "epoch": 1.3045555909525326, "grad_norm": 12.456510850739807, "learning_rate": 7.121976001960975e-06, "loss": 1.0947, "step": 2048 }, { "epoch": 1.3051927365402993, "grad_norm": 11.923151417923675, "learning_rate": 7.113593297231216e-06, "loss": 1.0684, "step": 2049 }, { "epoch": 1.3058298821280663, "grad_norm": 10.94926334268741, "learning_rate": 7.105214734825301e-06, "loss": 1.4258, "step": 2050 }, { "epoch": 1.306467027715833, "grad_norm": 17.523873943716026, "learning_rate": 7.0968403236735126e-06, "loss": 1.3064, "step": 2051 }, { "epoch": 1.3071041733035997, "grad_norm": 41.702107662007315, "learning_rate": 7.088470072701717e-06, "loss": 1.5552, "step": 2052 }, { "epoch": 1.3077413188913667, "grad_norm": 23.825619860951555, "learning_rate": 7.080103990831329e-06, "loss": 1.6936, "step": 2053 }, { "epoch": 1.3083784644791334, "grad_norm": 17.876888291409742, "learning_rate": 7.071742086979336e-06, "loss": 1.2642, "step": 2054 }, { "epoch": 1.3090156100669001, "grad_norm": 19.51100829435562, "learning_rate": 7.063384370058275e-06, "loss": 1.5215, "step": 2055 }, { "epoch": 1.309652755654667, "grad_norm": 59.20142591220812, "learning_rate": 7.055030848976199e-06, "loss": 1.7729, "step": 2056 }, { "epoch": 1.3102899012424338, "grad_norm": 17.32870627402887, "learning_rate": 7.046681532636715e-06, "loss": 1.3234, "step": 2057 }, { "epoch": 1.3109270468302008, "grad_norm": 21.528892755331675, "learning_rate": 7.0383364299389275e-06, "loss": 1.1614, "step": 2058 }, { "epoch": 1.3115641924179675, "grad_norm": 11.976131742424036, "learning_rate": 7.0299955497774664e-06, "loss": 1.1981, "step": 2059 }, { "epoch": 1.3122013380057342, "grad_norm": 31.48391400407821, "learning_rate": 7.021658901042446e-06, "loss": 0.8711, "step": 2060 }, { "epoch": 1.3128384835935012, "grad_norm": 9.581145758694436, "learning_rate": 7.013326492619474e-06, "loss": 1.2373, "step": 2061 }, { "epoch": 1.313475629181268, "grad_norm": 15.121551532190344, "learning_rate": 7.004998333389654e-06, "loss": 0.9204, "step": 2062 }, { "epoch": 1.3141127747690347, "grad_norm": 17.676714831824402, "learning_rate": 6.9966744322295365e-06, "loss": 1.6919, "step": 2063 }, { "epoch": 1.3147499203568016, "grad_norm": 31.456359214001655, "learning_rate": 6.988354798011157e-06, "loss": 1.5811, "step": 2064 }, { "epoch": 1.3153870659445683, "grad_norm": 15.174536631794895, "learning_rate": 6.980039439601982e-06, "loss": 1.3486, "step": 2065 }, { "epoch": 1.316024211532335, "grad_norm": 12.117151865605187, "learning_rate": 6.971728365864939e-06, "loss": 1.1575, "step": 2066 }, { "epoch": 1.316661357120102, "grad_norm": 13.543361574990467, "learning_rate": 6.963421585658372e-06, "loss": 1.6196, "step": 2067 }, { "epoch": 1.3172985027078687, "grad_norm": 15.14722968978732, "learning_rate": 6.95511910783606e-06, "loss": 1.166, "step": 2068 }, { "epoch": 1.3179356482956355, "grad_norm": 13.146656070215565, "learning_rate": 6.946820941247198e-06, "loss": 1.1014, "step": 2069 }, { "epoch": 1.3185727938834024, "grad_norm": 74.66541237174894, "learning_rate": 6.938527094736368e-06, "loss": 1.3069, "step": 2070 }, { "epoch": 1.3192099394711692, "grad_norm": 15.149502925257037, "learning_rate": 6.930237577143577e-06, "loss": 1.0195, "step": 2071 }, { "epoch": 1.3198470850589359, "grad_norm": 18.329845787000806, "learning_rate": 6.92195239730419e-06, "loss": 1.5408, "step": 2072 }, { "epoch": 1.3204842306467028, "grad_norm": 10.95146590174113, "learning_rate": 6.913671564048967e-06, "loss": 1.4747, "step": 2073 }, { "epoch": 1.3211213762344696, "grad_norm": 13.67821312250801, "learning_rate": 6.905395086204021e-06, "loss": 1.3029, "step": 2074 }, { "epoch": 1.3217585218222363, "grad_norm": 35.25853934387668, "learning_rate": 6.897122972590834e-06, "loss": 1.1389, "step": 2075 }, { "epoch": 1.3223956674100033, "grad_norm": 13.336136694206239, "learning_rate": 6.888855232026232e-06, "loss": 1.4348, "step": 2076 }, { "epoch": 1.32303281299777, "grad_norm": 12.92389700154025, "learning_rate": 6.880591873322382e-06, "loss": 1.5308, "step": 2077 }, { "epoch": 1.323669958585537, "grad_norm": 19.86718506652111, "learning_rate": 6.87233290528678e-06, "loss": 1.1656, "step": 2078 }, { "epoch": 1.3243071041733037, "grad_norm": 23.383994283407414, "learning_rate": 6.8640783367222364e-06, "loss": 1.0642, "step": 2079 }, { "epoch": 1.3249442497610704, "grad_norm": 20.77807489992674, "learning_rate": 6.855828176426883e-06, "loss": 1.0348, "step": 2080 }, { "epoch": 1.3255813953488373, "grad_norm": 14.611572264828686, "learning_rate": 6.8475824331941395e-06, "loss": 1.9702, "step": 2081 }, { "epoch": 1.326218540936604, "grad_norm": 15.931762639421502, "learning_rate": 6.839341115812732e-06, "loss": 1.4031, "step": 2082 }, { "epoch": 1.3268556865243708, "grad_norm": 15.24856322643738, "learning_rate": 6.831104233066661e-06, "loss": 1.6589, "step": 2083 }, { "epoch": 1.3274928321121378, "grad_norm": 20.971120127166092, "learning_rate": 6.822871793735204e-06, "loss": 1.2058, "step": 2084 }, { "epoch": 1.3281299776999045, "grad_norm": 12.777600743797253, "learning_rate": 6.8146438065929e-06, "loss": 1.2592, "step": 2085 }, { "epoch": 1.3287671232876712, "grad_norm": 17.65740829332342, "learning_rate": 6.806420280409539e-06, "loss": 1.7856, "step": 2086 }, { "epoch": 1.3294042688754382, "grad_norm": 17.204439085693352, "learning_rate": 6.79820122395017e-06, "loss": 1.6399, "step": 2087 }, { "epoch": 1.330041414463205, "grad_norm": 25.995376385104247, "learning_rate": 6.789986645975056e-06, "loss": 1.1875, "step": 2088 }, { "epoch": 1.3306785600509716, "grad_norm": 11.562495208262241, "learning_rate": 6.781776555239705e-06, "loss": 1.202, "step": 2089 }, { "epoch": 1.3313157056387386, "grad_norm": 18.598809484316792, "learning_rate": 6.7735709604948396e-06, "loss": 1.3395, "step": 2090 }, { "epoch": 1.3319528512265053, "grad_norm": 20.343882759278166, "learning_rate": 6.765369870486382e-06, "loss": 0.9003, "step": 2091 }, { "epoch": 1.332589996814272, "grad_norm": 15.00826520901423, "learning_rate": 6.757173293955466e-06, "loss": 1.0854, "step": 2092 }, { "epoch": 1.333227142402039, "grad_norm": 9.686066214174714, "learning_rate": 6.748981239638397e-06, "loss": 0.9435, "step": 2093 }, { "epoch": 1.3338642879898057, "grad_norm": 13.359600783512544, "learning_rate": 6.74079371626668e-06, "loss": 1.7375, "step": 2094 }, { "epoch": 1.3345014335775724, "grad_norm": 11.871751380486703, "learning_rate": 6.732610732566971e-06, "loss": 1.2498, "step": 2095 }, { "epoch": 1.3351385791653394, "grad_norm": 35.02327360961624, "learning_rate": 6.724432297261103e-06, "loss": 1.522, "step": 2096 }, { "epoch": 1.3357757247531061, "grad_norm": 11.384388425467007, "learning_rate": 6.7162584190660555e-06, "loss": 1.0751, "step": 2097 }, { "epoch": 1.3364128703408729, "grad_norm": 21.163982029783945, "learning_rate": 6.708089106693948e-06, "loss": 1.2587, "step": 2098 }, { "epoch": 1.3370500159286398, "grad_norm": 16.888683146905635, "learning_rate": 6.699924368852044e-06, "loss": 1.5959, "step": 2099 }, { "epoch": 1.3376871615164065, "grad_norm": 26.172939569590316, "learning_rate": 6.691764214242713e-06, "loss": 1.5229, "step": 2100 }, { "epoch": 1.3383243071041733, "grad_norm": 15.279579061944373, "learning_rate": 6.683608651563459e-06, "loss": 1.6836, "step": 2101 }, { "epoch": 1.3389614526919402, "grad_norm": 14.231200656902189, "learning_rate": 6.675457689506876e-06, "loss": 1.406, "step": 2102 }, { "epoch": 1.339598598279707, "grad_norm": 18.529351623736073, "learning_rate": 6.6673113367606645e-06, "loss": 1.5527, "step": 2103 }, { "epoch": 1.3402357438674737, "grad_norm": 18.41386817821341, "learning_rate": 6.6591696020076065e-06, "loss": 1.2859, "step": 2104 }, { "epoch": 1.3408728894552406, "grad_norm": 15.846599784165038, "learning_rate": 6.651032493925567e-06, "loss": 0.9359, "step": 2105 }, { "epoch": 1.3415100350430074, "grad_norm": 14.171783952664823, "learning_rate": 6.642900021187479e-06, "loss": 0.9674, "step": 2106 }, { "epoch": 1.342147180630774, "grad_norm": 13.040222780115075, "learning_rate": 6.6347721924613275e-06, "loss": 0.9884, "step": 2107 }, { "epoch": 1.342784326218541, "grad_norm": 17.529992897657092, "learning_rate": 6.626649016410158e-06, "loss": 1.1672, "step": 2108 }, { "epoch": 1.3434214718063078, "grad_norm": 39.95597961070566, "learning_rate": 6.618530501692044e-06, "loss": 1.1561, "step": 2109 }, { "epoch": 1.3440586173940745, "grad_norm": 12.65557695643918, "learning_rate": 6.610416656960107e-06, "loss": 0.8899, "step": 2110 }, { "epoch": 1.3446957629818415, "grad_norm": 24.066823678271174, "learning_rate": 6.602307490862479e-06, "loss": 2.103, "step": 2111 }, { "epoch": 1.3453329085696082, "grad_norm": 82.12029557588562, "learning_rate": 6.594203012042308e-06, "loss": 1.1877, "step": 2112 }, { "epoch": 1.345970054157375, "grad_norm": 14.111011877101175, "learning_rate": 6.586103229137752e-06, "loss": 0.839, "step": 2113 }, { "epoch": 1.3466071997451419, "grad_norm": 11.227140678942293, "learning_rate": 6.578008150781953e-06, "loss": 1.0317, "step": 2114 }, { "epoch": 1.3472443453329086, "grad_norm": 47.60701690000064, "learning_rate": 6.569917785603052e-06, "loss": 1.73, "step": 2115 }, { "epoch": 1.3478814909206753, "grad_norm": 20.24204661566761, "learning_rate": 6.561832142224145e-06, "loss": 1.4509, "step": 2116 }, { "epoch": 1.3485186365084423, "grad_norm": 15.891367738037328, "learning_rate": 6.553751229263326e-06, "loss": 1.2825, "step": 2117 }, { "epoch": 1.349155782096209, "grad_norm": 17.09383235177548, "learning_rate": 6.5456750553336206e-06, "loss": 1.1437, "step": 2118 }, { "epoch": 1.3497929276839757, "grad_norm": 18.693631490974546, "learning_rate": 6.537603629043013e-06, "loss": 1.1533, "step": 2119 }, { "epoch": 1.3504300732717427, "grad_norm": 21.065356913239942, "learning_rate": 6.529536958994436e-06, "loss": 1.1863, "step": 2120 }, { "epoch": 1.3510672188595094, "grad_norm": 8.860386746159124, "learning_rate": 6.521475053785731e-06, "loss": 1.079, "step": 2121 }, { "epoch": 1.3517043644472762, "grad_norm": 16.77816126353163, "learning_rate": 6.513417922009685e-06, "loss": 1.536, "step": 2122 }, { "epoch": 1.352341510035043, "grad_norm": 12.513309876167739, "learning_rate": 6.505365572253974e-06, "loss": 1.2961, "step": 2123 }, { "epoch": 1.3529786556228098, "grad_norm": 13.329550126394864, "learning_rate": 6.497318013101206e-06, "loss": 1.0979, "step": 2124 }, { "epoch": 1.3536158012105766, "grad_norm": 16.817656782478615, "learning_rate": 6.489275253128853e-06, "loss": 0.61, "step": 2125 }, { "epoch": 1.3542529467983435, "grad_norm": 10.948595437306784, "learning_rate": 6.48123730090929e-06, "loss": 1.5485, "step": 2126 }, { "epoch": 1.3548900923861102, "grad_norm": 13.543655785666045, "learning_rate": 6.473204165009765e-06, "loss": 1.3666, "step": 2127 }, { "epoch": 1.355527237973877, "grad_norm": 14.494036713124936, "learning_rate": 6.4651758539923846e-06, "loss": 1.3704, "step": 2128 }, { "epoch": 1.356164383561644, "grad_norm": 18.290447744746988, "learning_rate": 6.457152376414124e-06, "loss": 1.1825, "step": 2129 }, { "epoch": 1.3568015291494107, "grad_norm": 34.453343230140156, "learning_rate": 6.449133740826792e-06, "loss": 1.4783, "step": 2130 }, { "epoch": 1.3574386747371774, "grad_norm": 15.522203711348432, "learning_rate": 6.4411199557770576e-06, "loss": 1.1099, "step": 2131 }, { "epoch": 1.3580758203249443, "grad_norm": 27.912863437800404, "learning_rate": 6.433111029806401e-06, "loss": 1.4631, "step": 2132 }, { "epoch": 1.358712965912711, "grad_norm": 13.381377035042984, "learning_rate": 6.425106971451129e-06, "loss": 1.8186, "step": 2133 }, { "epoch": 1.3593501115004778, "grad_norm": 22.093598073449808, "learning_rate": 6.41710778924237e-06, "loss": 1.2944, "step": 2134 }, { "epoch": 1.3599872570882447, "grad_norm": 45.48053026190749, "learning_rate": 6.409113491706037e-06, "loss": 1.2983, "step": 2135 }, { "epoch": 1.3606244026760115, "grad_norm": 8.747130997920156, "learning_rate": 6.401124087362851e-06, "loss": 0.6539, "step": 2136 }, { "epoch": 1.3612615482637782, "grad_norm": 16.815635409535506, "learning_rate": 6.393139584728311e-06, "loss": 1.1519, "step": 2137 }, { "epoch": 1.3618986938515452, "grad_norm": 18.17724298235253, "learning_rate": 6.3851599923126975e-06, "loss": 1.4292, "step": 2138 }, { "epoch": 1.362535839439312, "grad_norm": 13.370787410083807, "learning_rate": 6.377185318621049e-06, "loss": 0.8986, "step": 2139 }, { "epoch": 1.3631729850270786, "grad_norm": 17.071905777771413, "learning_rate": 6.3692155721531665e-06, "loss": 1.2372, "step": 2140 }, { "epoch": 1.3638101306148456, "grad_norm": 20.97172996181436, "learning_rate": 6.3612507614036036e-06, "loss": 1.4592, "step": 2141 }, { "epoch": 1.3644472762026123, "grad_norm": 14.413755651287442, "learning_rate": 6.353290894861641e-06, "loss": 1.1289, "step": 2142 }, { "epoch": 1.365084421790379, "grad_norm": 29.700975379413944, "learning_rate": 6.3453359810113e-06, "loss": 1.473, "step": 2143 }, { "epoch": 1.365721567378146, "grad_norm": 18.239364027605724, "learning_rate": 6.3373860283313206e-06, "loss": 1.1337, "step": 2144 }, { "epoch": 1.3663587129659127, "grad_norm": 14.924436520804674, "learning_rate": 6.329441045295157e-06, "loss": 1.2397, "step": 2145 }, { "epoch": 1.3669958585536794, "grad_norm": 23.51184150274292, "learning_rate": 6.321501040370956e-06, "loss": 0.9343, "step": 2146 }, { "epoch": 1.3676330041414464, "grad_norm": 36.85808420858987, "learning_rate": 6.313566022021571e-06, "loss": 0.9943, "step": 2147 }, { "epoch": 1.3682701497292131, "grad_norm": 26.39652674236415, "learning_rate": 6.30563599870454e-06, "loss": 0.7241, "step": 2148 }, { "epoch": 1.3689072953169799, "grad_norm": 10.363840666978618, "learning_rate": 6.297710978872065e-06, "loss": 1.3765, "step": 2149 }, { "epoch": 1.3695444409047468, "grad_norm": 25.498597152979407, "learning_rate": 6.289790970971026e-06, "loss": 1.5415, "step": 2150 }, { "epoch": 1.3701815864925135, "grad_norm": 16.614393102032246, "learning_rate": 6.2818759834429565e-06, "loss": 1.738, "step": 2151 }, { "epoch": 1.3708187320802803, "grad_norm": 14.540907565953502, "learning_rate": 6.273966024724043e-06, "loss": 1.1936, "step": 2152 }, { "epoch": 1.3714558776680472, "grad_norm": 18.885950967807993, "learning_rate": 6.26606110324511e-06, "loss": 1.4481, "step": 2153 }, { "epoch": 1.372093023255814, "grad_norm": 14.551921818593703, "learning_rate": 6.258161227431609e-06, "loss": 0.6362, "step": 2154 }, { "epoch": 1.3727301688435807, "grad_norm": 16.706950890804706, "learning_rate": 6.250266405703621e-06, "loss": 1.4883, "step": 2155 }, { "epoch": 1.3733673144313476, "grad_norm": 16.161122465404162, "learning_rate": 6.2423766464758275e-06, "loss": 1.7595, "step": 2156 }, { "epoch": 1.3740044600191144, "grad_norm": 74.2518246628955, "learning_rate": 6.2344919581575355e-06, "loss": 1.3855, "step": 2157 }, { "epoch": 1.374641605606881, "grad_norm": 114.83239800602526, "learning_rate": 6.226612349152626e-06, "loss": 0.782, "step": 2158 }, { "epoch": 1.375278751194648, "grad_norm": 13.693596550891739, "learning_rate": 6.218737827859575e-06, "loss": 1.3174, "step": 2159 }, { "epoch": 1.3759158967824148, "grad_norm": 13.39024536135737, "learning_rate": 6.210868402671442e-06, "loss": 0.9283, "step": 2160 }, { "epoch": 1.3765530423701815, "grad_norm": 26.55741760191655, "learning_rate": 6.20300408197584e-06, "loss": 0.9049, "step": 2161 }, { "epoch": 1.3771901879579485, "grad_norm": 16.257359413491724, "learning_rate": 6.195144874154957e-06, "loss": 1.4736, "step": 2162 }, { "epoch": 1.3778273335457152, "grad_norm": 17.550753710660093, "learning_rate": 6.187290787585511e-06, "loss": 1.3448, "step": 2163 }, { "epoch": 1.378464479133482, "grad_norm": 12.039266839354966, "learning_rate": 6.179441830638795e-06, "loss": 0.9067, "step": 2164 }, { "epoch": 1.3791016247212489, "grad_norm": 10.908626327704468, "learning_rate": 6.171598011680598e-06, "loss": 1.1584, "step": 2165 }, { "epoch": 1.3797387703090156, "grad_norm": 25.41127954995898, "learning_rate": 6.1637593390712534e-06, "loss": 1.2977, "step": 2166 }, { "epoch": 1.3803759158967823, "grad_norm": 12.445617597305683, "learning_rate": 6.1559258211656095e-06, "loss": 0.6239, "step": 2167 }, { "epoch": 1.3810130614845493, "grad_norm": 6.213784725267618, "learning_rate": 6.148097466313008e-06, "loss": 1.1941, "step": 2168 }, { "epoch": 1.381650207072316, "grad_norm": 17.544272628654426, "learning_rate": 6.140274282857304e-06, "loss": 1.364, "step": 2169 }, { "epoch": 1.3822873526600827, "grad_norm": 19.80707910854222, "learning_rate": 6.13245627913682e-06, "loss": 1.5605, "step": 2170 }, { "epoch": 1.3829244982478497, "grad_norm": 14.717360932179904, "learning_rate": 6.124643463484383e-06, "loss": 1.2244, "step": 2171 }, { "epoch": 1.3835616438356164, "grad_norm": 15.976170386637104, "learning_rate": 6.116835844227269e-06, "loss": 1.3875, "step": 2172 }, { "epoch": 1.3841987894233831, "grad_norm": 40.54120153119264, "learning_rate": 6.109033429687227e-06, "loss": 1.3374, "step": 2173 }, { "epoch": 1.38483593501115, "grad_norm": 15.537312734637425, "learning_rate": 6.1012362281804534e-06, "loss": 1.0872, "step": 2174 }, { "epoch": 1.3854730805989168, "grad_norm": 14.37881505870341, "learning_rate": 6.093444248017591e-06, "loss": 1.2559, "step": 2175 }, { "epoch": 1.3861102261866836, "grad_norm": 15.754963200810028, "learning_rate": 6.085657497503713e-06, "loss": 1.3101, "step": 2176 }, { "epoch": 1.3867473717744505, "grad_norm": 14.982993897261158, "learning_rate": 6.077875984938325e-06, "loss": 1.168, "step": 2177 }, { "epoch": 1.3873845173622172, "grad_norm": 17.445655041643906, "learning_rate": 6.07009971861535e-06, "loss": 0.9681, "step": 2178 }, { "epoch": 1.388021662949984, "grad_norm": 16.56199015178178, "learning_rate": 6.062328706823109e-06, "loss": 1.1285, "step": 2179 }, { "epoch": 1.388658808537751, "grad_norm": 15.461256789975195, "learning_rate": 6.054562957844332e-06, "loss": 1.6245, "step": 2180 }, { "epoch": 1.3892959541255177, "grad_norm": 16.263015953508635, "learning_rate": 6.046802479956142e-06, "loss": 1.079, "step": 2181 }, { "epoch": 1.3899330997132844, "grad_norm": 11.561414298626877, "learning_rate": 6.03904728143003e-06, "loss": 1.5571, "step": 2182 }, { "epoch": 1.3905702453010513, "grad_norm": 20.288538416277145, "learning_rate": 6.031297370531875e-06, "loss": 1.0078, "step": 2183 }, { "epoch": 1.391207390888818, "grad_norm": 35.482266170351274, "learning_rate": 6.023552755521911e-06, "loss": 1.1755, "step": 2184 }, { "epoch": 1.3918445364765848, "grad_norm": 17.81737841177004, "learning_rate": 6.0158134446547354e-06, "loss": 1.2655, "step": 2185 }, { "epoch": 1.3924816820643517, "grad_norm": 14.532269608270305, "learning_rate": 6.0080794461792825e-06, "loss": 0.9561, "step": 2186 }, { "epoch": 1.3931188276521185, "grad_norm": 14.399221671353576, "learning_rate": 6.000350768338829e-06, "loss": 1.49, "step": 2187 }, { "epoch": 1.3937559732398852, "grad_norm": 18.75454849225662, "learning_rate": 5.992627419370984e-06, "loss": 1.833, "step": 2188 }, { "epoch": 1.3943931188276522, "grad_norm": 36.56796071512866, "learning_rate": 5.984909407507671e-06, "loss": 1.833, "step": 2189 }, { "epoch": 1.3950302644154189, "grad_norm": 16.927289198457938, "learning_rate": 5.977196740975126e-06, "loss": 1.0486, "step": 2190 }, { "epoch": 1.3956674100031856, "grad_norm": 11.557324240166162, "learning_rate": 5.969489427993893e-06, "loss": 0.9514, "step": 2191 }, { "epoch": 1.3963045555909526, "grad_norm": 61.19125924541549, "learning_rate": 5.961787476778806e-06, "loss": 1.2535, "step": 2192 }, { "epoch": 1.3969417011787193, "grad_norm": 17.034066166298633, "learning_rate": 5.954090895538982e-06, "loss": 1.366, "step": 2193 }, { "epoch": 1.397578846766486, "grad_norm": 12.167517767329953, "learning_rate": 5.946399692477815e-06, "loss": 0.8298, "step": 2194 }, { "epoch": 1.398215992354253, "grad_norm": 22.56154111506585, "learning_rate": 5.938713875792976e-06, "loss": 1.146, "step": 2195 }, { "epoch": 1.3988531379420197, "grad_norm": 26.08199865147237, "learning_rate": 5.931033453676376e-06, "loss": 1.3909, "step": 2196 }, { "epoch": 1.3994902835297864, "grad_norm": 11.549178114954165, "learning_rate": 5.923358434314198e-06, "loss": 1.1534, "step": 2197 }, { "epoch": 1.4001274291175534, "grad_norm": 20.47487852911026, "learning_rate": 5.91568882588685e-06, "loss": 1.3647, "step": 2198 }, { "epoch": 1.4007645747053201, "grad_norm": 19.686447870659578, "learning_rate": 5.908024636568982e-06, "loss": 1.1129, "step": 2199 }, { "epoch": 1.4014017202930868, "grad_norm": 16.90656450093098, "learning_rate": 5.900365874529461e-06, "loss": 1.6426, "step": 2200 }, { "epoch": 1.4020388658808538, "grad_norm": 16.45469378618153, "learning_rate": 5.892712547931375e-06, "loss": 1.2283, "step": 2201 }, { "epoch": 1.4026760114686205, "grad_norm": 19.90275176644662, "learning_rate": 5.8850646649320195e-06, "loss": 1.5659, "step": 2202 }, { "epoch": 1.4033131570563873, "grad_norm": 13.828449270758819, "learning_rate": 5.877422233682877e-06, "loss": 0.8319, "step": 2203 }, { "epoch": 1.4039503026441542, "grad_norm": 39.79486857957291, "learning_rate": 5.869785262329639e-06, "loss": 1.097, "step": 2204 }, { "epoch": 1.404587448231921, "grad_norm": 75.92883364712401, "learning_rate": 5.862153759012157e-06, "loss": 1.55, "step": 2205 }, { "epoch": 1.4052245938196877, "grad_norm": 21.82711497067787, "learning_rate": 5.85452773186447e-06, "loss": 1.328, "step": 2206 }, { "epoch": 1.4058617394074546, "grad_norm": 19.250260577286713, "learning_rate": 5.846907189014768e-06, "loss": 1.1589, "step": 2207 }, { "epoch": 1.4064988849952214, "grad_norm": 14.974969933696343, "learning_rate": 5.839292138585402e-06, "loss": 1.1383, "step": 2208 }, { "epoch": 1.407136030582988, "grad_norm": 51.92036762315189, "learning_rate": 5.831682588692873e-06, "loss": 1.9131, "step": 2209 }, { "epoch": 1.407773176170755, "grad_norm": 65.33071816462262, "learning_rate": 5.824078547447804e-06, "loss": 1.5532, "step": 2210 }, { "epoch": 1.4084103217585218, "grad_norm": 19.467564624926734, "learning_rate": 5.816480022954971e-06, "loss": 1.2476, "step": 2211 }, { "epoch": 1.4090474673462885, "grad_norm": 22.60113184941007, "learning_rate": 5.808887023313247e-06, "loss": 1.3146, "step": 2212 }, { "epoch": 1.4096846129340554, "grad_norm": 38.5673610964027, "learning_rate": 5.801299556615628e-06, "loss": 1.3157, "step": 2213 }, { "epoch": 1.4103217585218222, "grad_norm": 13.376811729795957, "learning_rate": 5.793717630949207e-06, "loss": 1.2549, "step": 2214 }, { "epoch": 1.410958904109589, "grad_norm": 18.786063099055088, "learning_rate": 5.786141254395177e-06, "loss": 1.4395, "step": 2215 }, { "epoch": 1.4115960496973559, "grad_norm": 15.588305369048003, "learning_rate": 5.778570435028812e-06, "loss": 1.499, "step": 2216 }, { "epoch": 1.4122331952851226, "grad_norm": 36.83783906341447, "learning_rate": 5.771005180919467e-06, "loss": 1.3333, "step": 2217 }, { "epoch": 1.4128703408728893, "grad_norm": 19.540180407651736, "learning_rate": 5.7634455001305625e-06, "loss": 1.5615, "step": 2218 }, { "epoch": 1.4135074864606563, "grad_norm": 29.943234093525103, "learning_rate": 5.755891400719577e-06, "loss": 1.3079, "step": 2219 }, { "epoch": 1.414144632048423, "grad_norm": 20.593440982923944, "learning_rate": 5.7483428907380436e-06, "loss": 1.4851, "step": 2220 }, { "epoch": 1.41478177763619, "grad_norm": 20.439357796694395, "learning_rate": 5.740799978231533e-06, "loss": 1.4473, "step": 2221 }, { "epoch": 1.4154189232239567, "grad_norm": 15.219647646305047, "learning_rate": 5.733262671239654e-06, "loss": 1.4338, "step": 2222 }, { "epoch": 1.4160560688117234, "grad_norm": 29.680547692568386, "learning_rate": 5.725730977796044e-06, "loss": 1.5618, "step": 2223 }, { "epoch": 1.4166932143994904, "grad_norm": 19.72249773683745, "learning_rate": 5.71820490592835e-06, "loss": 0.9097, "step": 2224 }, { "epoch": 1.417330359987257, "grad_norm": 13.884345446090999, "learning_rate": 5.710684463658234e-06, "loss": 0.9551, "step": 2225 }, { "epoch": 1.4179675055750238, "grad_norm": 11.93017263740438, "learning_rate": 5.7031696590013485e-06, "loss": 1.3999, "step": 2226 }, { "epoch": 1.4186046511627908, "grad_norm": 25.847342106283197, "learning_rate": 5.695660499967348e-06, "loss": 1.4049, "step": 2227 }, { "epoch": 1.4192417967505575, "grad_norm": 17.922117891593334, "learning_rate": 5.688156994559858e-06, "loss": 0.7632, "step": 2228 }, { "epoch": 1.4198789423383242, "grad_norm": 17.983561180186015, "learning_rate": 5.680659150776489e-06, "loss": 0.9435, "step": 2229 }, { "epoch": 1.4205160879260912, "grad_norm": 7.335770545768071, "learning_rate": 5.673166976608811e-06, "loss": 1.0131, "step": 2230 }, { "epoch": 1.421153233513858, "grad_norm": 20.637889402966742, "learning_rate": 5.665680480042355e-06, "loss": 1.0459, "step": 2231 }, { "epoch": 1.4217903791016246, "grad_norm": 13.353615764637215, "learning_rate": 5.658199669056598e-06, "loss": 1.1221, "step": 2232 }, { "epoch": 1.4224275246893916, "grad_norm": 18.18489121695622, "learning_rate": 5.650724551624953e-06, "loss": 1.7156, "step": 2233 }, { "epoch": 1.4230646702771583, "grad_norm": 15.036711339678364, "learning_rate": 5.643255135714773e-06, "loss": 0.9052, "step": 2234 }, { "epoch": 1.423701815864925, "grad_norm": 16.074471253531666, "learning_rate": 5.635791429287326e-06, "loss": 1.543, "step": 2235 }, { "epoch": 1.424338961452692, "grad_norm": 14.172543562828839, "learning_rate": 5.6283334402978e-06, "loss": 0.9109, "step": 2236 }, { "epoch": 1.4249761070404587, "grad_norm": 17.362608427989755, "learning_rate": 5.6208811766952855e-06, "loss": 1.0851, "step": 2237 }, { "epoch": 1.4256132526282255, "grad_norm": 26.23737256764091, "learning_rate": 5.613434646422776e-06, "loss": 1.087, "step": 2238 }, { "epoch": 1.4262503982159924, "grad_norm": 14.785238728633193, "learning_rate": 5.605993857417151e-06, "loss": 1.512, "step": 2239 }, { "epoch": 1.4268875438037592, "grad_norm": 14.427164959661004, "learning_rate": 5.598558817609166e-06, "loss": 1.0306, "step": 2240 }, { "epoch": 1.4275246893915259, "grad_norm": 16.023702041467594, "learning_rate": 5.591129534923458e-06, "loss": 1.1001, "step": 2241 }, { "epoch": 1.4281618349792928, "grad_norm": 19.50094005517075, "learning_rate": 5.583706017278516e-06, "loss": 0.9641, "step": 2242 }, { "epoch": 1.4287989805670596, "grad_norm": 13.470154853393314, "learning_rate": 5.576288272586694e-06, "loss": 1.0638, "step": 2243 }, { "epoch": 1.4294361261548265, "grad_norm": 11.973729380809573, "learning_rate": 5.568876308754191e-06, "loss": 0.6914, "step": 2244 }, { "epoch": 1.4300732717425932, "grad_norm": 22.776895426867586, "learning_rate": 5.561470133681044e-06, "loss": 1.3779, "step": 2245 }, { "epoch": 1.43071041733036, "grad_norm": 14.668249876332316, "learning_rate": 5.55406975526112e-06, "loss": 1.085, "step": 2246 }, { "epoch": 1.431347562918127, "grad_norm": 16.612590350531622, "learning_rate": 5.5466751813821044e-06, "loss": 1.3611, "step": 2247 }, { "epoch": 1.4319847085058937, "grad_norm": 12.79712214301427, "learning_rate": 5.5392864199255e-06, "loss": 0.9849, "step": 2248 }, { "epoch": 1.4326218540936604, "grad_norm": 20.8027649155241, "learning_rate": 5.531903478766612e-06, "loss": 1.4823, "step": 2249 }, { "epoch": 1.4332589996814273, "grad_norm": 18.008062750353343, "learning_rate": 5.524526365774543e-06, "loss": 1.2722, "step": 2250 }, { "epoch": 1.433896145269194, "grad_norm": 22.927327756226095, "learning_rate": 5.517155088812185e-06, "loss": 0.9875, "step": 2251 }, { "epoch": 1.4345332908569608, "grad_norm": 12.731450934523714, "learning_rate": 5.5097896557362084e-06, "loss": 1.4636, "step": 2252 }, { "epoch": 1.4351704364447277, "grad_norm": 22.06432225373696, "learning_rate": 5.502430074397056e-06, "loss": 1.2209, "step": 2253 }, { "epoch": 1.4358075820324945, "grad_norm": 15.050837633931863, "learning_rate": 5.495076352638929e-06, "loss": 1.73, "step": 2254 }, { "epoch": 1.4364447276202612, "grad_norm": 18.7197960253423, "learning_rate": 5.487728498299789e-06, "loss": 1.7021, "step": 2255 }, { "epoch": 1.4370818732080282, "grad_norm": 17.64756357924925, "learning_rate": 5.4803865192113356e-06, "loss": 0.9971, "step": 2256 }, { "epoch": 1.437719018795795, "grad_norm": 18.852818783326473, "learning_rate": 5.4730504231990235e-06, "loss": 1.0411, "step": 2257 }, { "epoch": 1.4383561643835616, "grad_norm": 25.508955549738197, "learning_rate": 5.465720218082016e-06, "loss": 1.512, "step": 2258 }, { "epoch": 1.4389933099713286, "grad_norm": 15.086326622675161, "learning_rate": 5.4583959116732084e-06, "loss": 1.4385, "step": 2259 }, { "epoch": 1.4396304555590953, "grad_norm": 15.614558259462388, "learning_rate": 5.451077511779215e-06, "loss": 0.9702, "step": 2260 }, { "epoch": 1.440267601146862, "grad_norm": 8.535917312462917, "learning_rate": 5.443765026200336e-06, "loss": 0.9309, "step": 2261 }, { "epoch": 1.440904746734629, "grad_norm": 14.0623644154073, "learning_rate": 5.436458462730587e-06, "loss": 1.2324, "step": 2262 }, { "epoch": 1.4415418923223957, "grad_norm": 11.817938958775168, "learning_rate": 5.429157829157656e-06, "loss": 1.4004, "step": 2263 }, { "epoch": 1.4421790379101624, "grad_norm": 14.184269830430447, "learning_rate": 5.42186313326293e-06, "loss": 1.3728, "step": 2264 }, { "epoch": 1.4428161834979294, "grad_norm": 15.665586209799546, "learning_rate": 5.4145743828214435e-06, "loss": 1.6189, "step": 2265 }, { "epoch": 1.4434533290856961, "grad_norm": 25.836482707937787, "learning_rate": 5.407291585601913e-06, "loss": 1.1804, "step": 2266 }, { "epoch": 1.4440904746734629, "grad_norm": 14.282596062566306, "learning_rate": 5.400014749366702e-06, "loss": 1.4569, "step": 2267 }, { "epoch": 1.4447276202612298, "grad_norm": 17.8714375088293, "learning_rate": 5.392743881871817e-06, "loss": 1.489, "step": 2268 }, { "epoch": 1.4453647658489965, "grad_norm": 27.543222690152657, "learning_rate": 5.3854789908669126e-06, "loss": 0.9519, "step": 2269 }, { "epoch": 1.4460019114367633, "grad_norm": 13.263641549751807, "learning_rate": 5.37822008409526e-06, "loss": 1.342, "step": 2270 }, { "epoch": 1.4466390570245302, "grad_norm": 27.497148349539525, "learning_rate": 5.37096716929377e-06, "loss": 1.5476, "step": 2271 }, { "epoch": 1.447276202612297, "grad_norm": 25.829345000432287, "learning_rate": 5.36372025419295e-06, "loss": 1.2861, "step": 2272 }, { "epoch": 1.4479133482000637, "grad_norm": 18.184748407546326, "learning_rate": 5.356479346516921e-06, "loss": 1.0249, "step": 2273 }, { "epoch": 1.4485504937878306, "grad_norm": 22.708898387960804, "learning_rate": 5.3492444539834025e-06, "loss": 1.2241, "step": 2274 }, { "epoch": 1.4491876393755974, "grad_norm": 36.48054781610626, "learning_rate": 5.342015584303692e-06, "loss": 1.48, "step": 2275 }, { "epoch": 1.449824784963364, "grad_norm": 17.117758458121738, "learning_rate": 5.33479274518268e-06, "loss": 1.1544, "step": 2276 }, { "epoch": 1.450461930551131, "grad_norm": 18.073784206544314, "learning_rate": 5.3275759443188245e-06, "loss": 1.9956, "step": 2277 }, { "epoch": 1.4510990761388978, "grad_norm": 17.617997875953414, "learning_rate": 5.3203651894041505e-06, "loss": 1.2234, "step": 2278 }, { "epoch": 1.4517362217266645, "grad_norm": 16.63067628234081, "learning_rate": 5.31316048812423e-06, "loss": 1.3264, "step": 2279 }, { "epoch": 1.4523733673144315, "grad_norm": 71.53440997963608, "learning_rate": 5.305961848158192e-06, "loss": 1.4341, "step": 2280 }, { "epoch": 1.4530105129021982, "grad_norm": 14.512511254244092, "learning_rate": 5.298769277178704e-06, "loss": 1.3828, "step": 2281 }, { "epoch": 1.453647658489965, "grad_norm": 14.689411040337413, "learning_rate": 5.291582782851956e-06, "loss": 1.1958, "step": 2282 }, { "epoch": 1.4542848040777319, "grad_norm": 12.35978862293902, "learning_rate": 5.2844023728376735e-06, "loss": 1.043, "step": 2283 }, { "epoch": 1.4549219496654986, "grad_norm": 15.641834868267452, "learning_rate": 5.277228054789088e-06, "loss": 0.8524, "step": 2284 }, { "epoch": 1.4555590952532653, "grad_norm": 9.206767161249354, "learning_rate": 5.270059836352945e-06, "loss": 1.2708, "step": 2285 }, { "epoch": 1.4561962408410323, "grad_norm": 14.216017094069038, "learning_rate": 5.2628977251694855e-06, "loss": 1.1107, "step": 2286 }, { "epoch": 1.456833386428799, "grad_norm": 19.41252069789052, "learning_rate": 5.255741728872436e-06, "loss": 1.1594, "step": 2287 }, { "epoch": 1.4574705320165657, "grad_norm": 11.797863843206104, "learning_rate": 5.248591855089016e-06, "loss": 1.1284, "step": 2288 }, { "epoch": 1.4581076776043327, "grad_norm": 14.041880926823971, "learning_rate": 5.241448111439907e-06, "loss": 1.3953, "step": 2289 }, { "epoch": 1.4587448231920994, "grad_norm": 22.675679155402545, "learning_rate": 5.234310505539264e-06, "loss": 1.3633, "step": 2290 }, { "epoch": 1.4593819687798661, "grad_norm": 45.11174675089733, "learning_rate": 5.227179044994703e-06, "loss": 1.6716, "step": 2291 }, { "epoch": 1.460019114367633, "grad_norm": 41.77331142447014, "learning_rate": 5.220053737407284e-06, "loss": 1.7715, "step": 2292 }, { "epoch": 1.4606562599553998, "grad_norm": 19.622015804940343, "learning_rate": 5.212934590371517e-06, "loss": 1.5105, "step": 2293 }, { "epoch": 1.4612934055431666, "grad_norm": 25.952517702407253, "learning_rate": 5.20582161147533e-06, "loss": 1.3345, "step": 2294 }, { "epoch": 1.4619305511309335, "grad_norm": 16.512792837603442, "learning_rate": 5.198714808300093e-06, "loss": 1.5483, "step": 2295 }, { "epoch": 1.4625676967187002, "grad_norm": 26.384698989464887, "learning_rate": 5.19161418842058e-06, "loss": 1.0668, "step": 2296 }, { "epoch": 1.463204842306467, "grad_norm": 24.299370989932477, "learning_rate": 5.184519759404993e-06, "loss": 1.1804, "step": 2297 }, { "epoch": 1.463841987894234, "grad_norm": 13.774669594741793, "learning_rate": 5.177431528814916e-06, "loss": 1.5288, "step": 2298 }, { "epoch": 1.4644791334820007, "grad_norm": 23.391600275899496, "learning_rate": 5.170349504205335e-06, "loss": 1.2572, "step": 2299 }, { "epoch": 1.4651162790697674, "grad_norm": 16.481382546290757, "learning_rate": 5.163273693124624e-06, "loss": 1.2356, "step": 2300 }, { "epoch": 1.4657534246575343, "grad_norm": 21.999575370856125, "learning_rate": 5.156204103114529e-06, "loss": 1.1094, "step": 2301 }, { "epoch": 1.466390570245301, "grad_norm": 30.646892836697237, "learning_rate": 5.149140741710171e-06, "loss": 0.9553, "step": 2302 }, { "epoch": 1.4670277158330678, "grad_norm": 29.28758147570212, "learning_rate": 5.14208361644002e-06, "loss": 1.1758, "step": 2303 }, { "epoch": 1.4676648614208347, "grad_norm": 16.200920588156396, "learning_rate": 5.13503273482592e-06, "loss": 1.2134, "step": 2304 }, { "epoch": 1.4683020070086015, "grad_norm": 17.635400701936266, "learning_rate": 5.127988104383039e-06, "loss": 1.5894, "step": 2305 }, { "epoch": 1.4689391525963682, "grad_norm": 34.03334000281187, "learning_rate": 5.120949732619895e-06, "loss": 1.8755, "step": 2306 }, { "epoch": 1.4695762981841352, "grad_norm": 25.8279715160543, "learning_rate": 5.113917627038334e-06, "loss": 1.006, "step": 2307 }, { "epoch": 1.4702134437719019, "grad_norm": 11.809951271498061, "learning_rate": 5.1068917951335154e-06, "loss": 1.2556, "step": 2308 }, { "epoch": 1.4708505893596686, "grad_norm": 69.64577052095996, "learning_rate": 5.09987224439392e-06, "loss": 1.7671, "step": 2309 }, { "epoch": 1.4714877349474356, "grad_norm": 14.044646446966276, "learning_rate": 5.09285898230133e-06, "loss": 1.3179, "step": 2310 }, { "epoch": 1.4721248805352023, "grad_norm": 18.98948262527537, "learning_rate": 5.08585201633083e-06, "loss": 1.231, "step": 2311 }, { "epoch": 1.472762026122969, "grad_norm": 22.84996896066259, "learning_rate": 5.078851353950783e-06, "loss": 1.312, "step": 2312 }, { "epoch": 1.473399171710736, "grad_norm": 20.017878736949857, "learning_rate": 5.071857002622844e-06, "loss": 1.197, "step": 2313 }, { "epoch": 1.4740363172985027, "grad_norm": 20.298408018701046, "learning_rate": 5.06486896980194e-06, "loss": 1.1417, "step": 2314 }, { "epoch": 1.4746734628862694, "grad_norm": 12.216176050131688, "learning_rate": 5.057887262936254e-06, "loss": 0.9041, "step": 2315 }, { "epoch": 1.4753106084740364, "grad_norm": 20.913835286483582, "learning_rate": 5.050911889467238e-06, "loss": 1.2616, "step": 2316 }, { "epoch": 1.4759477540618031, "grad_norm": 14.192166468016246, "learning_rate": 5.043942856829586e-06, "loss": 1.0581, "step": 2317 }, { "epoch": 1.4765848996495698, "grad_norm": 26.93539303315801, "learning_rate": 5.036980172451243e-06, "loss": 1.2988, "step": 2318 }, { "epoch": 1.4772220452373368, "grad_norm": 13.317698820857563, "learning_rate": 5.030023843753374e-06, "loss": 1.2135, "step": 2319 }, { "epoch": 1.4778591908251035, "grad_norm": 20.44797165639501, "learning_rate": 5.023073878150379e-06, "loss": 1.2954, "step": 2320 }, { "epoch": 1.4784963364128703, "grad_norm": 15.111036595095294, "learning_rate": 5.016130283049878e-06, "loss": 1.1401, "step": 2321 }, { "epoch": 1.4791334820006372, "grad_norm": 13.948381143782154, "learning_rate": 5.009193065852691e-06, "loss": 1.354, "step": 2322 }, { "epoch": 1.479770627588404, "grad_norm": 25.9152870501955, "learning_rate": 5.002262233952848e-06, "loss": 1.2454, "step": 2323 }, { "epoch": 1.4804077731761707, "grad_norm": 15.39536848686781, "learning_rate": 4.995337794737572e-06, "loss": 1.4856, "step": 2324 }, { "epoch": 1.4810449187639376, "grad_norm": 20.567352763607555, "learning_rate": 4.988419755587275e-06, "loss": 1.0679, "step": 2325 }, { "epoch": 1.4816820643517044, "grad_norm": 11.785757367080532, "learning_rate": 4.981508123875537e-06, "loss": 0.6875, "step": 2326 }, { "epoch": 1.482319209939471, "grad_norm": 14.823702184205013, "learning_rate": 4.97460290696912e-06, "loss": 1.4893, "step": 2327 }, { "epoch": 1.482956355527238, "grad_norm": 37.29225935910662, "learning_rate": 4.967704112227947e-06, "loss": 1.7637, "step": 2328 }, { "epoch": 1.4835935011150048, "grad_norm": 14.1071027501518, "learning_rate": 4.960811747005085e-06, "loss": 1.1075, "step": 2329 }, { "epoch": 1.4842306467027715, "grad_norm": 22.747688924816504, "learning_rate": 4.9539258186467685e-06, "loss": 0.8715, "step": 2330 }, { "epoch": 1.4848677922905384, "grad_norm": 12.239173144234922, "learning_rate": 4.947046334492349e-06, "loss": 1.2401, "step": 2331 }, { "epoch": 1.4855049378783052, "grad_norm": 38.69923090352668, "learning_rate": 4.940173301874328e-06, "loss": 1.8269, "step": 2332 }, { "epoch": 1.486142083466072, "grad_norm": 24.99535836943953, "learning_rate": 4.933306728118314e-06, "loss": 0.7803, "step": 2333 }, { "epoch": 1.4867792290538389, "grad_norm": 13.649435906996054, "learning_rate": 4.926446620543046e-06, "loss": 1.2756, "step": 2334 }, { "epoch": 1.4874163746416056, "grad_norm": 32.068120467386706, "learning_rate": 4.919592986460363e-06, "loss": 1.6143, "step": 2335 }, { "epoch": 1.4880535202293723, "grad_norm": 18.304477178736615, "learning_rate": 4.912745833175204e-06, "loss": 0.8374, "step": 2336 }, { "epoch": 1.4886906658171393, "grad_norm": 14.148995480039181, "learning_rate": 4.905905167985608e-06, "loss": 1.5623, "step": 2337 }, { "epoch": 1.489327811404906, "grad_norm": 15.892343908022442, "learning_rate": 4.8990709981826875e-06, "loss": 1.4177, "step": 2338 }, { "epoch": 1.4899649569926727, "grad_norm": 13.546041440432068, "learning_rate": 4.8922433310506425e-06, "loss": 1.1489, "step": 2339 }, { "epoch": 1.4906021025804397, "grad_norm": 17.006821988327374, "learning_rate": 4.885422173866732e-06, "loss": 1.4297, "step": 2340 }, { "epoch": 1.4912392481682064, "grad_norm": 23.834865395838097, "learning_rate": 4.878607533901284e-06, "loss": 1.3566, "step": 2341 }, { "epoch": 1.4918763937559731, "grad_norm": 15.785492721079159, "learning_rate": 4.87179941841768e-06, "loss": 1.6614, "step": 2342 }, { "epoch": 1.49251353934374, "grad_norm": 12.472597961426137, "learning_rate": 4.864997834672338e-06, "loss": 0.9875, "step": 2343 }, { "epoch": 1.4931506849315068, "grad_norm": 18.21449880820592, "learning_rate": 4.858202789914731e-06, "loss": 1.2874, "step": 2344 }, { "epoch": 1.4937878305192736, "grad_norm": 18.51415889181591, "learning_rate": 4.851414291387345e-06, "loss": 1.1318, "step": 2345 }, { "epoch": 1.4944249761070405, "grad_norm": 11.764716804509415, "learning_rate": 4.844632346325703e-06, "loss": 1.1304, "step": 2346 }, { "epoch": 1.4950621216948072, "grad_norm": 15.68933837198415, "learning_rate": 4.837856961958329e-06, "loss": 0.8661, "step": 2347 }, { "epoch": 1.495699267282574, "grad_norm": 19.242444478679744, "learning_rate": 4.8310881455067644e-06, "loss": 1.3831, "step": 2348 }, { "epoch": 1.496336412870341, "grad_norm": 20.746828326752006, "learning_rate": 4.82432590418555e-06, "loss": 1.5752, "step": 2349 }, { "epoch": 1.4969735584581076, "grad_norm": 20.073182947402003, "learning_rate": 4.817570245202212e-06, "loss": 1.4844, "step": 2350 }, { "epoch": 1.4976107040458744, "grad_norm": 18.825415416973573, "learning_rate": 4.810821175757271e-06, "loss": 0.9164, "step": 2351 }, { "epoch": 1.4982478496336413, "grad_norm": 17.6652270049946, "learning_rate": 4.804078703044211e-06, "loss": 1.4532, "step": 2352 }, { "epoch": 1.498884995221408, "grad_norm": 22.767328795777175, "learning_rate": 4.797342834249497e-06, "loss": 1.5347, "step": 2353 }, { "epoch": 1.4995221408091748, "grad_norm": 15.603588693909998, "learning_rate": 4.790613576552544e-06, "loss": 1.5708, "step": 2354 }, { "epoch": 1.5001592863969417, "grad_norm": 25.699346159299292, "learning_rate": 4.783890937125731e-06, "loss": 1.4473, "step": 2355 }, { "epoch": 1.5007964319847085, "grad_norm": 25.02175162587817, "learning_rate": 4.777174923134377e-06, "loss": 1.1001, "step": 2356 }, { "epoch": 1.5014335775724752, "grad_norm": 18.48425264138435, "learning_rate": 4.770465541736742e-06, "loss": 1.2495, "step": 2357 }, { "epoch": 1.5020707231602421, "grad_norm": 23.25190303937209, "learning_rate": 4.763762800084018e-06, "loss": 1.2118, "step": 2358 }, { "epoch": 1.5027078687480089, "grad_norm": 47.97121958879385, "learning_rate": 4.757066705320311e-06, "loss": 1.0724, "step": 2359 }, { "epoch": 1.5033450143357756, "grad_norm": 21.125317734038212, "learning_rate": 4.7503772645826545e-06, "loss": 0.9171, "step": 2360 }, { "epoch": 1.5039821599235426, "grad_norm": 15.442306989806296, "learning_rate": 4.74369448500098e-06, "loss": 0.9091, "step": 2361 }, { "epoch": 1.5046193055113093, "grad_norm": 20.342733479391235, "learning_rate": 4.737018373698125e-06, "loss": 1.2177, "step": 2362 }, { "epoch": 1.505256451099076, "grad_norm": 13.773703230027868, "learning_rate": 4.730348937789818e-06, "loss": 0.7959, "step": 2363 }, { "epoch": 1.505893596686843, "grad_norm": 19.843989860446925, "learning_rate": 4.7236861843846724e-06, "loss": 1.6372, "step": 2364 }, { "epoch": 1.5065307422746097, "grad_norm": 25.310256131530014, "learning_rate": 4.717030120584183e-06, "loss": 1.5115, "step": 2365 }, { "epoch": 1.5071678878623764, "grad_norm": 18.970752435090674, "learning_rate": 4.710380753482705e-06, "loss": 1.4955, "step": 2366 }, { "epoch": 1.5078050334501434, "grad_norm": 16.79894462846904, "learning_rate": 4.703738090167466e-06, "loss": 0.9054, "step": 2367 }, { "epoch": 1.5084421790379101, "grad_norm": 19.20213823267543, "learning_rate": 4.69710213771854e-06, "loss": 1.2585, "step": 2368 }, { "epoch": 1.5090793246256768, "grad_norm": 23.901764527684573, "learning_rate": 4.690472903208854e-06, "loss": 0.687, "step": 2369 }, { "epoch": 1.5097164702134438, "grad_norm": 18.054034665011336, "learning_rate": 4.683850393704173e-06, "loss": 1.9478, "step": 2370 }, { "epoch": 1.5103536158012105, "grad_norm": 40.83204726210891, "learning_rate": 4.6772346162630944e-06, "loss": 1.3396, "step": 2371 }, { "epoch": 1.5109907613889773, "grad_norm": 33.72490243739008, "learning_rate": 4.670625577937043e-06, "loss": 1.4329, "step": 2372 }, { "epoch": 1.5116279069767442, "grad_norm": 21.63581743260693, "learning_rate": 4.664023285770251e-06, "loss": 1.1155, "step": 2373 }, { "epoch": 1.512265052564511, "grad_norm": 16.92409480552317, "learning_rate": 4.657427746799773e-06, "loss": 1.618, "step": 2374 }, { "epoch": 1.5129021981522777, "grad_norm": 21.565335500059575, "learning_rate": 4.650838968055454e-06, "loss": 0.9458, "step": 2375 }, { "epoch": 1.5135393437400446, "grad_norm": 16.805123540561652, "learning_rate": 4.644256956559943e-06, "loss": 1.0764, "step": 2376 }, { "epoch": 1.5141764893278116, "grad_norm": 22.022586587749437, "learning_rate": 4.63768171932867e-06, "loss": 1.0363, "step": 2377 }, { "epoch": 1.514813634915578, "grad_norm": 14.07728124900937, "learning_rate": 4.631113263369848e-06, "loss": 1.6782, "step": 2378 }, { "epoch": 1.515450780503345, "grad_norm": 16.254994886991604, "learning_rate": 4.624551595684466e-06, "loss": 1.6316, "step": 2379 }, { "epoch": 1.516087926091112, "grad_norm": 21.625506695699375, "learning_rate": 4.6179967232662645e-06, "loss": 1.2717, "step": 2380 }, { "epoch": 1.5167250716788785, "grad_norm": 43.68840656641778, "learning_rate": 4.611448653101756e-06, "loss": 1.2788, "step": 2381 }, { "epoch": 1.5173622172666454, "grad_norm": 11.210988137758402, "learning_rate": 4.604907392170191e-06, "loss": 0.8887, "step": 2382 }, { "epoch": 1.5179993628544124, "grad_norm": 15.14782972046012, "learning_rate": 4.598372947443572e-06, "loss": 1.4578, "step": 2383 }, { "epoch": 1.518636508442179, "grad_norm": 23.630026381988518, "learning_rate": 4.59184532588663e-06, "loss": 1.6826, "step": 2384 }, { "epoch": 1.5192736540299459, "grad_norm": 27.766495829039517, "learning_rate": 4.585324534456827e-06, "loss": 1.2034, "step": 2385 }, { "epoch": 1.5199107996177128, "grad_norm": 25.357640055442143, "learning_rate": 4.578810580104347e-06, "loss": 0.9424, "step": 2386 }, { "epoch": 1.5205479452054793, "grad_norm": 18.77473651096339, "learning_rate": 4.572303469772077e-06, "loss": 1.8306, "step": 2387 }, { "epoch": 1.5211850907932463, "grad_norm": 22.214842563080914, "learning_rate": 4.565803210395622e-06, "loss": 1.173, "step": 2388 }, { "epoch": 1.5218222363810132, "grad_norm": 22.79999258581565, "learning_rate": 4.559309808903272e-06, "loss": 1.2166, "step": 2389 }, { "epoch": 1.5224593819687797, "grad_norm": 35.74374989094812, "learning_rate": 4.5528232722160235e-06, "loss": 1.4829, "step": 2390 }, { "epoch": 1.5230965275565467, "grad_norm": 20.735728041678755, "learning_rate": 4.546343607247541e-06, "loss": 1.8684, "step": 2391 }, { "epoch": 1.5237336731443136, "grad_norm": 34.789929494593366, "learning_rate": 4.5398708209041695e-06, "loss": 1.0129, "step": 2392 }, { "epoch": 1.5243708187320801, "grad_norm": 25.624715058389096, "learning_rate": 4.53340492008493e-06, "loss": 1.4963, "step": 2393 }, { "epoch": 1.525007964319847, "grad_norm": 19.24668706671426, "learning_rate": 4.526945911681493e-06, "loss": 1.1466, "step": 2394 }, { "epoch": 1.525645109907614, "grad_norm": 23.32429418784675, "learning_rate": 4.520493802578191e-06, "loss": 1.6782, "step": 2395 }, { "epoch": 1.5262822554953805, "grad_norm": 24.396876099190333, "learning_rate": 4.514048599651994e-06, "loss": 1.0566, "step": 2396 }, { "epoch": 1.5269194010831475, "grad_norm": 18.026747713312005, "learning_rate": 4.507610309772529e-06, "loss": 1.0197, "step": 2397 }, { "epoch": 1.5275565466709145, "grad_norm": 28.79706277135271, "learning_rate": 4.50117893980203e-06, "loss": 1.2812, "step": 2398 }, { "epoch": 1.528193692258681, "grad_norm": 18.255186055649236, "learning_rate": 4.494754496595374e-06, "loss": 1.0851, "step": 2399 }, { "epoch": 1.528830837846448, "grad_norm": 16.902681194870816, "learning_rate": 4.488336987000051e-06, "loss": 0.8806, "step": 2400 }, { "epoch": 1.5294679834342149, "grad_norm": 26.12458830907757, "learning_rate": 4.481926417856155e-06, "loss": 1.2649, "step": 2401 }, { "epoch": 1.5301051290219814, "grad_norm": 19.732560599449222, "learning_rate": 4.4755227959963915e-06, "loss": 1.4238, "step": 2402 }, { "epoch": 1.5307422746097483, "grad_norm": 28.011160520818173, "learning_rate": 4.469126128246047e-06, "loss": 1.5204, "step": 2403 }, { "epoch": 1.5313794201975153, "grad_norm": 108.97594959437154, "learning_rate": 4.462736421423019e-06, "loss": 1.0696, "step": 2404 }, { "epoch": 1.5320165657852818, "grad_norm": 24.170992060091645, "learning_rate": 4.45635368233776e-06, "loss": 1.2417, "step": 2405 }, { "epoch": 1.5326537113730487, "grad_norm": 44.81672966838737, "learning_rate": 4.449977917793313e-06, "loss": 0.8695, "step": 2406 }, { "epoch": 1.5332908569608157, "grad_norm": 16.31323655479564, "learning_rate": 4.443609134585285e-06, "loss": 1.2461, "step": 2407 }, { "epoch": 1.5339280025485822, "grad_norm": 24.492153806301236, "learning_rate": 4.437247339501833e-06, "loss": 1.1803, "step": 2408 }, { "epoch": 1.5345651481363491, "grad_norm": 14.141080752616377, "learning_rate": 4.430892539323674e-06, "loss": 1.2952, "step": 2409 }, { "epoch": 1.535202293724116, "grad_norm": 82.35202689881153, "learning_rate": 4.424544740824068e-06, "loss": 1.1538, "step": 2410 }, { "epoch": 1.5358394393118826, "grad_norm": 11.378016704390145, "learning_rate": 4.418203950768815e-06, "loss": 1.3057, "step": 2411 }, { "epoch": 1.5364765848996496, "grad_norm": 24.039073225901692, "learning_rate": 4.411870175916234e-06, "loss": 1.2324, "step": 2412 }, { "epoch": 1.5371137304874165, "grad_norm": 16.142825187459668, "learning_rate": 4.4055434230171804e-06, "loss": 0.8583, "step": 2413 }, { "epoch": 1.537750876075183, "grad_norm": 37.066226677744176, "learning_rate": 4.399223698815018e-06, "loss": 1.4272, "step": 2414 }, { "epoch": 1.53838802166295, "grad_norm": 15.954495071327193, "learning_rate": 4.3929110100456185e-06, "loss": 1.2987, "step": 2415 }, { "epoch": 1.539025167250717, "grad_norm": 31.491722867520984, "learning_rate": 4.386605363437358e-06, "loss": 1.2627, "step": 2416 }, { "epoch": 1.5396623128384836, "grad_norm": 43.645543383808764, "learning_rate": 4.3803067657111054e-06, "loss": 1.2227, "step": 2417 }, { "epoch": 1.5402994584262504, "grad_norm": 47.77954340206793, "learning_rate": 4.3740152235802165e-06, "loss": 0.7384, "step": 2418 }, { "epoch": 1.5409366040140173, "grad_norm": 18.490632672830234, "learning_rate": 4.3677307437505304e-06, "loss": 1.5093, "step": 2419 }, { "epoch": 1.541573749601784, "grad_norm": 33.079040654259636, "learning_rate": 4.36145333292035e-06, "loss": 1.2898, "step": 2420 }, { "epoch": 1.5422108951895508, "grad_norm": 15.902760348244469, "learning_rate": 4.355182997780454e-06, "loss": 1.1045, "step": 2421 }, { "epoch": 1.5428480407773177, "grad_norm": 20.751367678289604, "learning_rate": 4.34891974501407e-06, "loss": 1.6372, "step": 2422 }, { "epoch": 1.5434851863650845, "grad_norm": 25.860648067757907, "learning_rate": 4.342663581296883e-06, "loss": 1.4221, "step": 2423 }, { "epoch": 1.5441223319528512, "grad_norm": 29.743745928432613, "learning_rate": 4.336414513297023e-06, "loss": 1.4607, "step": 2424 }, { "epoch": 1.5447594775406182, "grad_norm": 32.070279221467246, "learning_rate": 4.330172547675052e-06, "loss": 1.6213, "step": 2425 }, { "epoch": 1.5453966231283849, "grad_norm": 17.13179406467998, "learning_rate": 4.32393769108397e-06, "loss": 0.9426, "step": 2426 }, { "epoch": 1.5460337687161516, "grad_norm": 18.321417110128355, "learning_rate": 4.3177099501691864e-06, "loss": 1.1259, "step": 2427 }, { "epoch": 1.5466709143039186, "grad_norm": 31.041115737697037, "learning_rate": 4.311489331568542e-06, "loss": 1.3953, "step": 2428 }, { "epoch": 1.5473080598916853, "grad_norm": 22.77693471687558, "learning_rate": 4.30527584191227e-06, "loss": 1.6499, "step": 2429 }, { "epoch": 1.547945205479452, "grad_norm": 31.69717282247815, "learning_rate": 4.299069487823023e-06, "loss": 0.9901, "step": 2430 }, { "epoch": 1.548582351067219, "grad_norm": 12.956503524852188, "learning_rate": 4.292870275915835e-06, "loss": 1.2106, "step": 2431 }, { "epoch": 1.5492194966549857, "grad_norm": 30.26130134436587, "learning_rate": 4.2866782127981305e-06, "loss": 1.0925, "step": 2432 }, { "epoch": 1.5498566422427524, "grad_norm": 19.434399701210566, "learning_rate": 4.280493305069721e-06, "loss": 1.8862, "step": 2433 }, { "epoch": 1.5504937878305194, "grad_norm": 19.6654612373016, "learning_rate": 4.274315559322778e-06, "loss": 1.2983, "step": 2434 }, { "epoch": 1.5511309334182861, "grad_norm": 22.46514866703086, "learning_rate": 4.268144982141854e-06, "loss": 0.9792, "step": 2435 }, { "epoch": 1.5517680790060528, "grad_norm": 21.018635910083727, "learning_rate": 4.261981580103848e-06, "loss": 1.3401, "step": 2436 }, { "epoch": 1.5524052245938198, "grad_norm": 28.300689513767423, "learning_rate": 4.255825359778028e-06, "loss": 1.1918, "step": 2437 }, { "epoch": 1.5530423701815865, "grad_norm": 15.389356407099028, "learning_rate": 4.2496763277259855e-06, "loss": 0.6129, "step": 2438 }, { "epoch": 1.5536795157693533, "grad_norm": 22.937389839230743, "learning_rate": 4.243534490501668e-06, "loss": 1.3494, "step": 2439 }, { "epoch": 1.5543166613571202, "grad_norm": 34.169243946539524, "learning_rate": 4.2373998546513505e-06, "loss": 1.5488, "step": 2440 }, { "epoch": 1.554953806944887, "grad_norm": 28.826096847163875, "learning_rate": 4.2312724267136254e-06, "loss": 1.3591, "step": 2441 }, { "epoch": 1.5555909525326537, "grad_norm": 19.98654939637575, "learning_rate": 4.22515221321941e-06, "loss": 0.9591, "step": 2442 }, { "epoch": 1.5562280981204206, "grad_norm": 12.872215910767448, "learning_rate": 4.219039220691925e-06, "loss": 1.1672, "step": 2443 }, { "epoch": 1.5568652437081874, "grad_norm": 12.921751289173478, "learning_rate": 4.2129334556467095e-06, "loss": 1.1022, "step": 2444 }, { "epoch": 1.557502389295954, "grad_norm": 30.03725127127231, "learning_rate": 4.206834924591578e-06, "loss": 1.4998, "step": 2445 }, { "epoch": 1.558139534883721, "grad_norm": 28.119358161700948, "learning_rate": 4.200743634026651e-06, "loss": 1.3743, "step": 2446 }, { "epoch": 1.5587766804714878, "grad_norm": 17.45954639800886, "learning_rate": 4.194659590444328e-06, "loss": 0.9833, "step": 2447 }, { "epoch": 1.5594138260592545, "grad_norm": 17.42213185683615, "learning_rate": 4.188582800329276e-06, "loss": 1.2217, "step": 2448 }, { "epoch": 1.5600509716470214, "grad_norm": 24.833658430606125, "learning_rate": 4.182513270158443e-06, "loss": 1.1119, "step": 2449 }, { "epoch": 1.5606881172347882, "grad_norm": 12.661924963495524, "learning_rate": 4.176451006401031e-06, "loss": 0.8668, "step": 2450 }, { "epoch": 1.561325262822555, "grad_norm": 27.404422504746748, "learning_rate": 4.170396015518503e-06, "loss": 1.3174, "step": 2451 }, { "epoch": 1.5619624084103219, "grad_norm": 12.969789724373415, "learning_rate": 4.16434830396456e-06, "loss": 1.231, "step": 2452 }, { "epoch": 1.5625995539980886, "grad_norm": 25.365798157560953, "learning_rate": 4.158307878185157e-06, "loss": 1.1808, "step": 2453 }, { "epoch": 1.5632366995858553, "grad_norm": 42.05604032851017, "learning_rate": 4.152274744618477e-06, "loss": 1.1189, "step": 2454 }, { "epoch": 1.5638738451736223, "grad_norm": 18.573605045767874, "learning_rate": 4.146248909694928e-06, "loss": 0.8489, "step": 2455 }, { "epoch": 1.564510990761389, "grad_norm": 108.15028977717142, "learning_rate": 4.140230379837145e-06, "loss": 1.0532, "step": 2456 }, { "epoch": 1.5651481363491557, "grad_norm": 21.945416114743495, "learning_rate": 4.134219161459972e-06, "loss": 1.6587, "step": 2457 }, { "epoch": 1.5657852819369227, "grad_norm": 31.276233347982185, "learning_rate": 4.128215260970462e-06, "loss": 0.9181, "step": 2458 }, { "epoch": 1.5664224275246894, "grad_norm": 75.20439907379281, "learning_rate": 4.122218684767869e-06, "loss": 0.879, "step": 2459 }, { "epoch": 1.5670595731124561, "grad_norm": 17.447111610989662, "learning_rate": 4.116229439243638e-06, "loss": 1.2112, "step": 2460 }, { "epoch": 1.567696718700223, "grad_norm": 18.07528639797717, "learning_rate": 4.110247530781406e-06, "loss": 1.0286, "step": 2461 }, { "epoch": 1.5683338642879898, "grad_norm": 24.078294787406527, "learning_rate": 4.1042729657569805e-06, "loss": 1.2563, "step": 2462 }, { "epoch": 1.5689710098757566, "grad_norm": 38.041000491626015, "learning_rate": 4.0983057505383505e-06, "loss": 1.0382, "step": 2463 }, { "epoch": 1.5696081554635235, "grad_norm": 16.31062992076628, "learning_rate": 4.092345891485666e-06, "loss": 1.3826, "step": 2464 }, { "epoch": 1.5702453010512902, "grad_norm": 28.85454176385331, "learning_rate": 4.086393394951243e-06, "loss": 1.4448, "step": 2465 }, { "epoch": 1.570882446639057, "grad_norm": 44.88727329422949, "learning_rate": 4.0804482672795435e-06, "loss": 0.9264, "step": 2466 }, { "epoch": 1.571519592226824, "grad_norm": 12.045428031913012, "learning_rate": 4.074510514807174e-06, "loss": 0.931, "step": 2467 }, { "epoch": 1.5721567378145906, "grad_norm": 45.56997052242361, "learning_rate": 4.068580143862889e-06, "loss": 1.3872, "step": 2468 }, { "epoch": 1.5727938834023574, "grad_norm": 27.62641855278439, "learning_rate": 4.062657160767565e-06, "loss": 1.1759, "step": 2469 }, { "epoch": 1.5734310289901243, "grad_norm": 39.95490954216083, "learning_rate": 4.056741571834216e-06, "loss": 1.0126, "step": 2470 }, { "epoch": 1.574068174577891, "grad_norm": 19.721638922902756, "learning_rate": 4.050833383367963e-06, "loss": 0.9912, "step": 2471 }, { "epoch": 1.5747053201656578, "grad_norm": 23.03168196211596, "learning_rate": 4.044932601666048e-06, "loss": 1.2772, "step": 2472 }, { "epoch": 1.5753424657534247, "grad_norm": 90.31693200461007, "learning_rate": 4.0390392330178115e-06, "loss": 1.2401, "step": 2473 }, { "epoch": 1.5759796113411915, "grad_norm": 25.043203195789143, "learning_rate": 4.033153283704697e-06, "loss": 0.9099, "step": 2474 }, { "epoch": 1.5766167569289582, "grad_norm": 19.074878031591265, "learning_rate": 4.027274760000243e-06, "loss": 1.6294, "step": 2475 }, { "epoch": 1.5772539025167251, "grad_norm": 22.716352953091796, "learning_rate": 4.021403668170063e-06, "loss": 0.9216, "step": 2476 }, { "epoch": 1.5778910481044919, "grad_norm": 16.428048366538146, "learning_rate": 4.015540014471865e-06, "loss": 1.104, "step": 2477 }, { "epoch": 1.5785281936922586, "grad_norm": 41.94604576005231, "learning_rate": 4.009683805155413e-06, "loss": 1.168, "step": 2478 }, { "epoch": 1.5791653392800256, "grad_norm": 197.64123417057968, "learning_rate": 4.003835046462549e-06, "loss": 0.8607, "step": 2479 }, { "epoch": 1.5798024848677923, "grad_norm": 15.19522855162574, "learning_rate": 3.9979937446271625e-06, "loss": 1.1974, "step": 2480 }, { "epoch": 1.580439630455559, "grad_norm": 27.413738545035255, "learning_rate": 3.992159905875204e-06, "loss": 1.7935, "step": 2481 }, { "epoch": 1.581076776043326, "grad_norm": 26.876948828559474, "learning_rate": 3.986333536424669e-06, "loss": 1.4629, "step": 2482 }, { "epoch": 1.5817139216310927, "grad_norm": 20.536531069178796, "learning_rate": 3.980514642485581e-06, "loss": 1.6431, "step": 2483 }, { "epoch": 1.5823510672188594, "grad_norm": 32.3529673337124, "learning_rate": 3.974703230260015e-06, "loss": 1.5017, "step": 2484 }, { "epoch": 1.5829882128066264, "grad_norm": 14.720819009335479, "learning_rate": 3.968899305942052e-06, "loss": 1.4004, "step": 2485 }, { "epoch": 1.5836253583943931, "grad_norm": 32.82916287568615, "learning_rate": 3.963102875717807e-06, "loss": 1.1096, "step": 2486 }, { "epoch": 1.5842625039821598, "grad_norm": 27.26842123349532, "learning_rate": 3.957313945765395e-06, "loss": 1.5128, "step": 2487 }, { "epoch": 1.5848996495699268, "grad_norm": 32.66643095503693, "learning_rate": 3.951532522254946e-06, "loss": 1.3264, "step": 2488 }, { "epoch": 1.5855367951576935, "grad_norm": 25.386340720916323, "learning_rate": 3.945758611348586e-06, "loss": 1.3127, "step": 2489 }, { "epoch": 1.5861739407454603, "grad_norm": 18.836307335205948, "learning_rate": 3.939992219200434e-06, "loss": 1.5051, "step": 2490 }, { "epoch": 1.5868110863332272, "grad_norm": 69.33988408099633, "learning_rate": 3.934233351956597e-06, "loss": 1.752, "step": 2491 }, { "epoch": 1.587448231920994, "grad_norm": 31.77213585083978, "learning_rate": 3.928482015755154e-06, "loss": 1.3733, "step": 2492 }, { "epoch": 1.5880853775087607, "grad_norm": 28.92984841891655, "learning_rate": 3.92273821672617e-06, "loss": 1.1453, "step": 2493 }, { "epoch": 1.5887225230965276, "grad_norm": 20.70521152639441, "learning_rate": 3.917001960991662e-06, "loss": 0.9346, "step": 2494 }, { "epoch": 1.5893596686842943, "grad_norm": 13.654734418010891, "learning_rate": 3.911273254665617e-06, "loss": 1.6067, "step": 2495 }, { "epoch": 1.589996814272061, "grad_norm": 29.737483194371517, "learning_rate": 3.905552103853974e-06, "loss": 1.1658, "step": 2496 }, { "epoch": 1.590633959859828, "grad_norm": 17.158237507469863, "learning_rate": 3.899838514654616e-06, "loss": 0.7788, "step": 2497 }, { "epoch": 1.5912711054475948, "grad_norm": 14.56379956594654, "learning_rate": 3.894132493157371e-06, "loss": 1.3228, "step": 2498 }, { "epoch": 1.5919082510353615, "grad_norm": 20.306244478522927, "learning_rate": 3.888434045443993e-06, "loss": 1.2834, "step": 2499 }, { "epoch": 1.5925453966231284, "grad_norm": 20.835945760296124, "learning_rate": 3.8827431775881716e-06, "loss": 0.9692, "step": 2500 }, { "epoch": 1.5931825422108952, "grad_norm": 13.876972120565778, "learning_rate": 3.877059895655513e-06, "loss": 1.0305, "step": 2501 }, { "epoch": 1.593819687798662, "grad_norm": 22.315818756438535, "learning_rate": 3.871384205703541e-06, "loss": 1.244, "step": 2502 }, { "epoch": 1.5944568333864289, "grad_norm": 35.689513006241974, "learning_rate": 3.865716113781683e-06, "loss": 1.3907, "step": 2503 }, { "epoch": 1.5950939789741956, "grad_norm": 19.961070242244055, "learning_rate": 3.860055625931271e-06, "loss": 1.0912, "step": 2504 }, { "epoch": 1.5957311245619623, "grad_norm": 15.180702798788296, "learning_rate": 3.854402748185538e-06, "loss": 0.8669, "step": 2505 }, { "epoch": 1.5963682701497293, "grad_norm": 11.323310088987663, "learning_rate": 3.848757486569592e-06, "loss": 0.5908, "step": 2506 }, { "epoch": 1.597005415737496, "grad_norm": 22.726587794464127, "learning_rate": 3.843119847100437e-06, "loss": 1.6111, "step": 2507 }, { "epoch": 1.5976425613252627, "grad_norm": 19.736250780272144, "learning_rate": 3.837489835786942e-06, "loss": 1.4656, "step": 2508 }, { "epoch": 1.5982797069130297, "grad_norm": 20.96022255428504, "learning_rate": 3.831867458629852e-06, "loss": 1.446, "step": 2509 }, { "epoch": 1.5989168525007964, "grad_norm": 31.863987201563333, "learning_rate": 3.826252721621775e-06, "loss": 1.5977, "step": 2510 }, { "epoch": 1.5995539980885631, "grad_norm": 49.90862521243167, "learning_rate": 3.820645630747174e-06, "loss": 1.134, "step": 2511 }, { "epoch": 1.60019114367633, "grad_norm": 46.26339638861929, "learning_rate": 3.815046191982365e-06, "loss": 1.485, "step": 2512 }, { "epoch": 1.6008282892640968, "grad_norm": 19.542753619399292, "learning_rate": 3.8094544112955027e-06, "loss": 1.0344, "step": 2513 }, { "epoch": 1.6014654348518635, "grad_norm": 13.812019115239586, "learning_rate": 3.8038702946465857e-06, "loss": 1.3246, "step": 2514 }, { "epoch": 1.6021025804396305, "grad_norm": 23.3188769616663, "learning_rate": 3.7982938479874354e-06, "loss": 1.3772, "step": 2515 }, { "epoch": 1.6027397260273972, "grad_norm": 30.733688637171262, "learning_rate": 3.7927250772617073e-06, "loss": 1.2496, "step": 2516 }, { "epoch": 1.603376871615164, "grad_norm": 22.16956247067201, "learning_rate": 3.7871639884048717e-06, "loss": 1.1572, "step": 2517 }, { "epoch": 1.604014017202931, "grad_norm": 28.78770714808474, "learning_rate": 3.7816105873442112e-06, "loss": 1.6938, "step": 2518 }, { "epoch": 1.6046511627906976, "grad_norm": 16.548824656371785, "learning_rate": 3.7760648799988152e-06, "loss": 0.7279, "step": 2519 }, { "epoch": 1.6052883083784644, "grad_norm": 15.588737294197234, "learning_rate": 3.77052687227957e-06, "loss": 1.5042, "step": 2520 }, { "epoch": 1.6059254539662313, "grad_norm": 28.642032607054524, "learning_rate": 3.7649965700891587e-06, "loss": 1.6428, "step": 2521 }, { "epoch": 1.606562599553998, "grad_norm": 34.93714915024378, "learning_rate": 3.759473979322048e-06, "loss": 1.0205, "step": 2522 }, { "epoch": 1.6071997451417648, "grad_norm": 19.350934517650742, "learning_rate": 3.7539591058644888e-06, "loss": 1.3142, "step": 2523 }, { "epoch": 1.6078368907295317, "grad_norm": 25.45114342065329, "learning_rate": 3.7484519555945054e-06, "loss": 1.7583, "step": 2524 }, { "epoch": 1.6084740363172985, "grad_norm": 27.70242835034207, "learning_rate": 3.7429525343818894e-06, "loss": 1.5396, "step": 2525 }, { "epoch": 1.6091111819050652, "grad_norm": 31.300847645248624, "learning_rate": 3.737460848088197e-06, "loss": 1.6047, "step": 2526 }, { "epoch": 1.6097483274928321, "grad_norm": 40.93232520901023, "learning_rate": 3.7319769025667352e-06, "loss": 1.0322, "step": 2527 }, { "epoch": 1.6103854730805989, "grad_norm": 17.179483694884627, "learning_rate": 3.7265007036625665e-06, "loss": 0.8728, "step": 2528 }, { "epoch": 1.6110226186683656, "grad_norm": 19.17963881902007, "learning_rate": 3.7210322572124863e-06, "loss": 1.1775, "step": 2529 }, { "epoch": 1.6116597642561326, "grad_norm": 14.439442252193235, "learning_rate": 3.7155715690450454e-06, "loss": 1.4321, "step": 2530 }, { "epoch": 1.6122969098438993, "grad_norm": 32.710368765182324, "learning_rate": 3.7101186449805073e-06, "loss": 1.3748, "step": 2531 }, { "epoch": 1.612934055431666, "grad_norm": 30.232402159079573, "learning_rate": 3.7046734908308674e-06, "loss": 1.1715, "step": 2532 }, { "epoch": 1.613571201019433, "grad_norm": 19.28766024735767, "learning_rate": 3.6992361123998442e-06, "loss": 1.2209, "step": 2533 }, { "epoch": 1.6142083466071997, "grad_norm": 20.3628657857816, "learning_rate": 3.693806515482859e-06, "loss": 1.6919, "step": 2534 }, { "epoch": 1.6148454921949664, "grad_norm": 23.27982537265124, "learning_rate": 3.6883847058670462e-06, "loss": 0.9086, "step": 2535 }, { "epoch": 1.6154826377827334, "grad_norm": 13.820865416422492, "learning_rate": 3.6829706893312327e-06, "loss": 0.6619, "step": 2536 }, { "epoch": 1.6161197833705, "grad_norm": 71.07462515908088, "learning_rate": 3.6775644716459525e-06, "loss": 1.3218, "step": 2537 }, { "epoch": 1.6167569289582668, "grad_norm": 24.89952787532141, "learning_rate": 3.672166058573413e-06, "loss": 1.343, "step": 2538 }, { "epoch": 1.6173940745460338, "grad_norm": 23.952943556646854, "learning_rate": 3.6667754558675106e-06, "loss": 1.1813, "step": 2539 }, { "epoch": 1.6180312201338005, "grad_norm": 19.617290903620088, "learning_rate": 3.661392669273817e-06, "loss": 1.1421, "step": 2540 }, { "epoch": 1.6186683657215672, "grad_norm": 16.33473437481388, "learning_rate": 3.6560177045295685e-06, "loss": 1.4362, "step": 2541 }, { "epoch": 1.6193055113093342, "grad_norm": 25.894299098313187, "learning_rate": 3.650650567363669e-06, "loss": 1.328, "step": 2542 }, { "epoch": 1.6199426568971012, "grad_norm": 25.058660930009474, "learning_rate": 3.645291263496677e-06, "loss": 1.2745, "step": 2543 }, { "epoch": 1.6205798024848677, "grad_norm": 23.95841167195612, "learning_rate": 3.639939798640807e-06, "loss": 1.2069, "step": 2544 }, { "epoch": 1.6212169480726346, "grad_norm": 28.43337779623744, "learning_rate": 3.634596178499907e-06, "loss": 1.4556, "step": 2545 }, { "epoch": 1.6218540936604016, "grad_norm": 21.5997870701206, "learning_rate": 3.629260408769476e-06, "loss": 1.4055, "step": 2546 }, { "epoch": 1.622491239248168, "grad_norm": 19.77675735195598, "learning_rate": 3.62393249513664e-06, "loss": 1.6211, "step": 2547 }, { "epoch": 1.623128384835935, "grad_norm": 20.176413116042657, "learning_rate": 3.6186124432801484e-06, "loss": 1.1428, "step": 2548 }, { "epoch": 1.623765530423702, "grad_norm": 17.982958657599696, "learning_rate": 3.613300258870377e-06, "loss": 1.0527, "step": 2549 }, { "epoch": 1.6244026760114685, "grad_norm": 20.371454455948665, "learning_rate": 3.607995947569314e-06, "loss": 1.5315, "step": 2550 }, { "epoch": 1.6250398215992354, "grad_norm": 26.449224337234742, "learning_rate": 3.6026995150305553e-06, "loss": 1.2109, "step": 2551 }, { "epoch": 1.6256769671870024, "grad_norm": 37.284340853754244, "learning_rate": 3.597410966899304e-06, "loss": 1.3536, "step": 2552 }, { "epoch": 1.626314112774769, "grad_norm": 45.84463196272492, "learning_rate": 3.5921303088123488e-06, "loss": 1.0317, "step": 2553 }, { "epoch": 1.6269512583625358, "grad_norm": 30.72320554661904, "learning_rate": 3.5868575463980804e-06, "loss": 1.5354, "step": 2554 }, { "epoch": 1.6275884039503028, "grad_norm": 27.472637636264906, "learning_rate": 3.5815926852764633e-06, "loss": 1.1445, "step": 2555 }, { "epoch": 1.6282255495380693, "grad_norm": 14.68755222003923, "learning_rate": 3.5763357310590506e-06, "loss": 1.0342, "step": 2556 }, { "epoch": 1.6288626951258363, "grad_norm": 25.489658000553323, "learning_rate": 3.5710866893489614e-06, "loss": 1.4236, "step": 2557 }, { "epoch": 1.6294998407136032, "grad_norm": 27.468249941119414, "learning_rate": 3.5658455657408843e-06, "loss": 1.5137, "step": 2558 }, { "epoch": 1.6301369863013697, "grad_norm": 35.13442287763208, "learning_rate": 3.560612365821069e-06, "loss": 1.3644, "step": 2559 }, { "epoch": 1.6307741318891367, "grad_norm": 19.32719054445389, "learning_rate": 3.555387095167313e-06, "loss": 1.0068, "step": 2560 }, { "epoch": 1.6314112774769036, "grad_norm": 19.003205379837404, "learning_rate": 3.550169759348972e-06, "loss": 1.3821, "step": 2561 }, { "epoch": 1.6320484230646701, "grad_norm": 51.95901341650048, "learning_rate": 3.544960363926936e-06, "loss": 1.1523, "step": 2562 }, { "epoch": 1.632685568652437, "grad_norm": 29.529636723974235, "learning_rate": 3.539758914453637e-06, "loss": 1.1925, "step": 2563 }, { "epoch": 1.633322714240204, "grad_norm": 34.95698954507153, "learning_rate": 3.534565416473036e-06, "loss": 1.7375, "step": 2564 }, { "epoch": 1.6339598598279705, "grad_norm": 20.64451813191161, "learning_rate": 3.5293798755206186e-06, "loss": 1.4475, "step": 2565 }, { "epoch": 1.6345970054157375, "grad_norm": 25.318487034873947, "learning_rate": 3.5242022971233947e-06, "loss": 1.0564, "step": 2566 }, { "epoch": 1.6352341510035044, "grad_norm": 42.747863394468325, "learning_rate": 3.5190326867998764e-06, "loss": 0.7622, "step": 2567 }, { "epoch": 1.635871296591271, "grad_norm": 17.367606177110957, "learning_rate": 3.5138710500600925e-06, "loss": 1.4082, "step": 2568 }, { "epoch": 1.636508442179038, "grad_norm": 24.231027151798024, "learning_rate": 3.508717392405566e-06, "loss": 1.2555, "step": 2569 }, { "epoch": 1.6371455877668049, "grad_norm": 23.4386487237221, "learning_rate": 3.503571719329325e-06, "loss": 1.7446, "step": 2570 }, { "epoch": 1.6377827333545714, "grad_norm": 34.63415535484522, "learning_rate": 3.4984340363158785e-06, "loss": 1.7329, "step": 2571 }, { "epoch": 1.6384198789423383, "grad_norm": 21.415629407876843, "learning_rate": 3.4933043488412212e-06, "loss": 1.4996, "step": 2572 }, { "epoch": 1.6390570245301053, "grad_norm": 20.27573529726606, "learning_rate": 3.4881826623728308e-06, "loss": 1.123, "step": 2573 }, { "epoch": 1.6396941701178718, "grad_norm": 24.0358857306477, "learning_rate": 3.4830689823696482e-06, "loss": 1.7588, "step": 2574 }, { "epoch": 1.6403313157056387, "grad_norm": 25.354589578337258, "learning_rate": 3.4779633142820913e-06, "loss": 1.2484, "step": 2575 }, { "epoch": 1.6409684612934057, "grad_norm": 24.655311816755983, "learning_rate": 3.472865663552024e-06, "loss": 1.043, "step": 2576 }, { "epoch": 1.6416056068811722, "grad_norm": 26.82233564399391, "learning_rate": 3.4677760356127844e-06, "loss": 1.6267, "step": 2577 }, { "epoch": 1.6422427524689391, "grad_norm": 65.82279570467759, "learning_rate": 3.4626944358891412e-06, "loss": 1.1218, "step": 2578 }, { "epoch": 1.642879898056706, "grad_norm": 27.985377613637176, "learning_rate": 3.4576208697973167e-06, "loss": 1.3008, "step": 2579 }, { "epoch": 1.6435170436444726, "grad_norm": 26.60020682858107, "learning_rate": 3.45255534274497e-06, "loss": 1.1639, "step": 2580 }, { "epoch": 1.6441541892322395, "grad_norm": 22.170351905676906, "learning_rate": 3.4474978601311837e-06, "loss": 1.4421, "step": 2581 }, { "epoch": 1.6447913348200065, "grad_norm": 14.820903150548123, "learning_rate": 3.4424484273464787e-06, "loss": 1.3982, "step": 2582 }, { "epoch": 1.6454284804077732, "grad_norm": 28.871504408779415, "learning_rate": 3.437407049772781e-06, "loss": 0.7944, "step": 2583 }, { "epoch": 1.64606562599554, "grad_norm": 13.48984597298236, "learning_rate": 3.4323737327834495e-06, "loss": 0.7647, "step": 2584 }, { "epoch": 1.646702771583307, "grad_norm": 11.944596866049482, "learning_rate": 3.4273484817432335e-06, "loss": 1.3257, "step": 2585 }, { "epoch": 1.6473399171710736, "grad_norm": 27.12771978900244, "learning_rate": 3.422331302008296e-06, "loss": 1.4817, "step": 2586 }, { "epoch": 1.6479770627588404, "grad_norm": 22.692975444220878, "learning_rate": 3.4173221989261985e-06, "loss": 1.0148, "step": 2587 }, { "epoch": 1.6486142083466073, "grad_norm": 36.375963459084765, "learning_rate": 3.412321177835882e-06, "loss": 0.8401, "step": 2588 }, { "epoch": 1.649251353934374, "grad_norm": 19.15966984627985, "learning_rate": 3.4073282440676868e-06, "loss": 1.2345, "step": 2589 }, { "epoch": 1.6498884995221408, "grad_norm": 36.203719393041126, "learning_rate": 3.4023434029433267e-06, "loss": 0.9531, "step": 2590 }, { "epoch": 1.6505256451099077, "grad_norm": 37.556996559397554, "learning_rate": 3.3973666597758925e-06, "loss": 1.6621, "step": 2591 }, { "epoch": 1.6511627906976745, "grad_norm": 18.939853119766973, "learning_rate": 3.3923980198698393e-06, "loss": 1.4983, "step": 2592 }, { "epoch": 1.6517999362854412, "grad_norm": 56.55665286096557, "learning_rate": 3.387437488520989e-06, "loss": 1.2974, "step": 2593 }, { "epoch": 1.6524370818732081, "grad_norm": 28.11703537416843, "learning_rate": 3.3824850710165247e-06, "loss": 0.7644, "step": 2594 }, { "epoch": 1.6530742274609749, "grad_norm": 22.412377454674683, "learning_rate": 3.3775407726349714e-06, "loss": 1.3365, "step": 2595 }, { "epoch": 1.6537113730487416, "grad_norm": 20.93908826224356, "learning_rate": 3.3726045986462074e-06, "loss": 1.5977, "step": 2596 }, { "epoch": 1.6543485186365086, "grad_norm": 22.69765875330249, "learning_rate": 3.367676554311453e-06, "loss": 1.1047, "step": 2597 }, { "epoch": 1.6549856642242753, "grad_norm": 18.618733237880644, "learning_rate": 3.3627566448832598e-06, "loss": 0.9849, "step": 2598 }, { "epoch": 1.655622809812042, "grad_norm": 17.138199455155068, "learning_rate": 3.357844875605507e-06, "loss": 1.3079, "step": 2599 }, { "epoch": 1.656259955399809, "grad_norm": 21.058538498129984, "learning_rate": 3.352941251713404e-06, "loss": 1.119, "step": 2600 }, { "epoch": 1.6568971009875757, "grad_norm": 21.241297416436513, "learning_rate": 3.348045778433474e-06, "loss": 1.3757, "step": 2601 }, { "epoch": 1.6575342465753424, "grad_norm": 25.480266478911272, "learning_rate": 3.3431584609835506e-06, "loss": 1.2188, "step": 2602 }, { "epoch": 1.6581713921631094, "grad_norm": 23.623799149326906, "learning_rate": 3.338279304572783e-06, "loss": 1.175, "step": 2603 }, { "epoch": 1.658808537750876, "grad_norm": 28.530620482797282, "learning_rate": 3.333408314401612e-06, "loss": 1.5342, "step": 2604 }, { "epoch": 1.6594456833386428, "grad_norm": 18.105527595257932, "learning_rate": 3.3285454956617826e-06, "loss": 1.0505, "step": 2605 }, { "epoch": 1.6600828289264098, "grad_norm": 20.842233289059614, "learning_rate": 3.3236908535363237e-06, "loss": 1.2683, "step": 2606 }, { "epoch": 1.6607199745141765, "grad_norm": 25.277865919491134, "learning_rate": 3.318844393199553e-06, "loss": 1.3894, "step": 2607 }, { "epoch": 1.6613571201019433, "grad_norm": 13.53872704955822, "learning_rate": 3.3140061198170685e-06, "loss": 0.6028, "step": 2608 }, { "epoch": 1.6619942656897102, "grad_norm": 20.00871550512412, "learning_rate": 3.3091760385457363e-06, "loss": 1.3608, "step": 2609 }, { "epoch": 1.662631411277477, "grad_norm": 19.03466546263796, "learning_rate": 3.304354154533702e-06, "loss": 0.6169, "step": 2610 }, { "epoch": 1.6632685568652437, "grad_norm": 16.44823313941527, "learning_rate": 3.2995404729203608e-06, "loss": 0.9932, "step": 2611 }, { "epoch": 1.6639057024530106, "grad_norm": 15.819136916587501, "learning_rate": 3.2947349988363774e-06, "loss": 1.1918, "step": 2612 }, { "epoch": 1.6645428480407773, "grad_norm": 17.496439310316514, "learning_rate": 3.2899377374036565e-06, "loss": 1.0682, "step": 2613 }, { "epoch": 1.665179993628544, "grad_norm": 32.58085493661352, "learning_rate": 3.2851486937353597e-06, "loss": 1.1688, "step": 2614 }, { "epoch": 1.665817139216311, "grad_norm": 29.20301160329673, "learning_rate": 3.280367872935888e-06, "loss": 1.7114, "step": 2615 }, { "epoch": 1.6664542848040778, "grad_norm": 31.648345717060277, "learning_rate": 3.2755952801008696e-06, "loss": 1.174, "step": 2616 }, { "epoch": 1.6670914303918445, "grad_norm": 18.270067770434345, "learning_rate": 3.270830920317176e-06, "loss": 1.1838, "step": 2617 }, { "epoch": 1.6677285759796114, "grad_norm": 24.761815867509032, "learning_rate": 3.266074798662893e-06, "loss": 1.77, "step": 2618 }, { "epoch": 1.6683657215673782, "grad_norm": 24.150419940682927, "learning_rate": 3.2613269202073306e-06, "loss": 1.4033, "step": 2619 }, { "epoch": 1.669002867155145, "grad_norm": 20.35670322929247, "learning_rate": 3.256587290011009e-06, "loss": 0.8032, "step": 2620 }, { "epoch": 1.6696400127429118, "grad_norm": 22.070935900585408, "learning_rate": 3.251855913125662e-06, "loss": 1.5266, "step": 2621 }, { "epoch": 1.6702771583306786, "grad_norm": 89.47936118266324, "learning_rate": 3.2471327945942253e-06, "loss": 0.9204, "step": 2622 }, { "epoch": 1.6709143039184453, "grad_norm": 16.877564737111253, "learning_rate": 3.242417939450829e-06, "loss": 1.3264, "step": 2623 }, { "epoch": 1.6715514495062123, "grad_norm": 28.935036822592597, "learning_rate": 3.2377113527208026e-06, "loss": 1.5476, "step": 2624 }, { "epoch": 1.672188595093979, "grad_norm": 22.645226868168987, "learning_rate": 3.233013039420653e-06, "loss": 1.5737, "step": 2625 }, { "epoch": 1.6728257406817457, "grad_norm": 23.242946637549714, "learning_rate": 3.2283230045580783e-06, "loss": 1.6628, "step": 2626 }, { "epoch": 1.6734628862695127, "grad_norm": 44.906103560145425, "learning_rate": 3.2236412531319444e-06, "loss": 1.1763, "step": 2627 }, { "epoch": 1.6741000318572794, "grad_norm": 14.985049151892898, "learning_rate": 3.2189677901322945e-06, "loss": 1.2466, "step": 2628 }, { "epoch": 1.6747371774450461, "grad_norm": 22.15324298020636, "learning_rate": 3.2143026205403376e-06, "loss": 1.1487, "step": 2629 }, { "epoch": 1.675374323032813, "grad_norm": 19.304247651530336, "learning_rate": 3.2096457493284395e-06, "loss": 1.2708, "step": 2630 }, { "epoch": 1.6760114686205798, "grad_norm": 18.278747901311572, "learning_rate": 3.204997181460128e-06, "loss": 1.499, "step": 2631 }, { "epoch": 1.6766486142083465, "grad_norm": 82.3243373450291, "learning_rate": 3.2003569218900675e-06, "loss": 1.1285, "step": 2632 }, { "epoch": 1.6772857597961135, "grad_norm": 14.298717195542167, "learning_rate": 3.195724975564082e-06, "loss": 1.1556, "step": 2633 }, { "epoch": 1.6779229053838802, "grad_norm": 30.280514515221757, "learning_rate": 3.1911013474191243e-06, "loss": 1.4292, "step": 2634 }, { "epoch": 1.678560050971647, "grad_norm": 27.245350797991858, "learning_rate": 3.1864860423832874e-06, "loss": 0.7321, "step": 2635 }, { "epoch": 1.679197196559414, "grad_norm": 12.219442685516514, "learning_rate": 3.1818790653757893e-06, "loss": 0.64, "step": 2636 }, { "epoch": 1.6798343421471806, "grad_norm": 16.568819648487576, "learning_rate": 3.1772804213069746e-06, "loss": 1.1316, "step": 2637 }, { "epoch": 1.6804714877349474, "grad_norm": 39.77132677799872, "learning_rate": 3.172690115078307e-06, "loss": 1.1243, "step": 2638 }, { "epoch": 1.6811086333227143, "grad_norm": 16.539427087900474, "learning_rate": 3.1681081515823566e-06, "loss": 1.2678, "step": 2639 }, { "epoch": 1.681745778910481, "grad_norm": 47.36249469289484, "learning_rate": 3.163534535702809e-06, "loss": 1.2549, "step": 2640 }, { "epoch": 1.6823829244982478, "grad_norm": 37.223069449349495, "learning_rate": 3.158969272314447e-06, "loss": 1.4846, "step": 2641 }, { "epoch": 1.6830200700860147, "grad_norm": 33.10997074085698, "learning_rate": 3.154412366283155e-06, "loss": 1.811, "step": 2642 }, { "epoch": 1.6836572156737815, "grad_norm": 28.22298107631648, "learning_rate": 3.1498638224659075e-06, "loss": 1.1179, "step": 2643 }, { "epoch": 1.6842943612615482, "grad_norm": 17.80316522693762, "learning_rate": 3.1453236457107673e-06, "loss": 1.8574, "step": 2644 }, { "epoch": 1.6849315068493151, "grad_norm": 133.2438147011908, "learning_rate": 3.14079184085688e-06, "loss": 1.0862, "step": 2645 }, { "epoch": 1.6855686524370819, "grad_norm": 33.55076848409059, "learning_rate": 3.1362684127344632e-06, "loss": 1.068, "step": 2646 }, { "epoch": 1.6862057980248486, "grad_norm": 36.22381178689432, "learning_rate": 3.131753366164814e-06, "loss": 1.3003, "step": 2647 }, { "epoch": 1.6868429436126156, "grad_norm": 16.27081052504699, "learning_rate": 3.127246705960286e-06, "loss": 1.668, "step": 2648 }, { "epoch": 1.6874800892003823, "grad_norm": 70.38245663620789, "learning_rate": 3.1227484369243037e-06, "loss": 1.3699, "step": 2649 }, { "epoch": 1.688117234788149, "grad_norm": 22.416029176214803, "learning_rate": 3.1182585638513422e-06, "loss": 1.2888, "step": 2650 }, { "epoch": 1.688754380375916, "grad_norm": 14.861346837607785, "learning_rate": 3.1137770915269305e-06, "loss": 1.1138, "step": 2651 }, { "epoch": 1.6893915259636827, "grad_norm": 25.916767579647367, "learning_rate": 3.1093040247276456e-06, "loss": 1.5049, "step": 2652 }, { "epoch": 1.6900286715514494, "grad_norm": 23.725329116406513, "learning_rate": 3.104839368221097e-06, "loss": 1.4885, "step": 2653 }, { "epoch": 1.6906658171392164, "grad_norm": 30.670567024530474, "learning_rate": 3.10038312676594e-06, "loss": 0.9633, "step": 2654 }, { "epoch": 1.691302962726983, "grad_norm": 15.434703126693902, "learning_rate": 3.0959353051118523e-06, "loss": 0.8976, "step": 2655 }, { "epoch": 1.6919401083147498, "grad_norm": 15.787955607517949, "learning_rate": 3.0914959079995442e-06, "loss": 1.1277, "step": 2656 }, { "epoch": 1.6925772539025168, "grad_norm": 22.520752078863772, "learning_rate": 3.0870649401607443e-06, "loss": 1.4875, "step": 2657 }, { "epoch": 1.6932143994902835, "grad_norm": 32.306271816426765, "learning_rate": 3.0826424063181954e-06, "loss": 0.8936, "step": 2658 }, { "epoch": 1.6938515450780502, "grad_norm": 16.652628919418298, "learning_rate": 3.0782283111856543e-06, "loss": 1.5553, "step": 2659 }, { "epoch": 1.6944886906658172, "grad_norm": 112.21596196353329, "learning_rate": 3.073822659467879e-06, "loss": 1.2905, "step": 2660 }, { "epoch": 1.695125836253584, "grad_norm": 20.70384387696636, "learning_rate": 3.069425455860631e-06, "loss": 1.7263, "step": 2661 }, { "epoch": 1.6957629818413507, "grad_norm": 18.968255025585794, "learning_rate": 3.065036705050663e-06, "loss": 1.0365, "step": 2662 }, { "epoch": 1.6964001274291176, "grad_norm": 16.59452548861851, "learning_rate": 3.060656411715729e-06, "loss": 0.8696, "step": 2663 }, { "epoch": 1.6970372730168843, "grad_norm": 13.260721128980524, "learning_rate": 3.0562845805245567e-06, "loss": 0.8036, "step": 2664 }, { "epoch": 1.697674418604651, "grad_norm": 12.001664249194825, "learning_rate": 3.0519212161368596e-06, "loss": 0.5402, "step": 2665 }, { "epoch": 1.698311564192418, "grad_norm": 11.590706323586149, "learning_rate": 3.0475663232033296e-06, "loss": 1.2556, "step": 2666 }, { "epoch": 1.6989487097801848, "grad_norm": 18.32136552705927, "learning_rate": 3.043219906365623e-06, "loss": 0.9893, "step": 2667 }, { "epoch": 1.6995858553679515, "grad_norm": 11.158253970933094, "learning_rate": 3.0388819702563667e-06, "loss": 1.1731, "step": 2668 }, { "epoch": 1.7002230009557184, "grad_norm": 35.82800805637861, "learning_rate": 3.0345525194991438e-06, "loss": 1.0154, "step": 2669 }, { "epoch": 1.7008601465434852, "grad_norm": 18.668631678726136, "learning_rate": 3.0302315587085025e-06, "loss": 1.1304, "step": 2670 }, { "epoch": 1.701497292131252, "grad_norm": 18.241501373163644, "learning_rate": 3.0259190924899305e-06, "loss": 1.161, "step": 2671 }, { "epoch": 1.7021344377190188, "grad_norm": 21.391728009337037, "learning_rate": 3.0216151254398677e-06, "loss": 0.8666, "step": 2672 }, { "epoch": 1.7027715833067856, "grad_norm": 24.150748589913384, "learning_rate": 3.017319662145698e-06, "loss": 1.7666, "step": 2673 }, { "epoch": 1.7034087288945523, "grad_norm": 29.278525366293724, "learning_rate": 3.013032707185733e-06, "loss": 1.3354, "step": 2674 }, { "epoch": 1.7040458744823193, "grad_norm": 24.982904301188928, "learning_rate": 3.008754265129223e-06, "loss": 0.8937, "step": 2675 }, { "epoch": 1.704683020070086, "grad_norm": 14.492586975811633, "learning_rate": 3.004484340536342e-06, "loss": 1.5708, "step": 2676 }, { "epoch": 1.7053201656578527, "grad_norm": 26.72810636400064, "learning_rate": 3.000222937958188e-06, "loss": 1.7729, "step": 2677 }, { "epoch": 1.7059573112456197, "grad_norm": 27.858618322932998, "learning_rate": 2.995970061936771e-06, "loss": 1.2897, "step": 2678 }, { "epoch": 1.7065944568333864, "grad_norm": 31.930498902999776, "learning_rate": 2.9917257170050166e-06, "loss": 0.9428, "step": 2679 }, { "epoch": 1.7072316024211531, "grad_norm": 34.961624299916856, "learning_rate": 2.9874899076867596e-06, "loss": 1.311, "step": 2680 }, { "epoch": 1.70786874800892, "grad_norm": 11.679554101255244, "learning_rate": 2.98326263849673e-06, "loss": 1.4424, "step": 2681 }, { "epoch": 1.7085058935966868, "grad_norm": 20.085853253697493, "learning_rate": 2.979043913940563e-06, "loss": 1.244, "step": 2682 }, { "epoch": 1.7091430391844535, "grad_norm": 16.211430559471438, "learning_rate": 2.9748337385147805e-06, "loss": 1.4521, "step": 2683 }, { "epoch": 1.7097801847722205, "grad_norm": 21.169679048365083, "learning_rate": 2.970632116706798e-06, "loss": 1.2681, "step": 2684 }, { "epoch": 1.7104173303599872, "grad_norm": 42.10648959703698, "learning_rate": 2.9664390529949116e-06, "loss": 1.4247, "step": 2685 }, { "epoch": 1.711054475947754, "grad_norm": 23.005023212318335, "learning_rate": 2.9622545518482904e-06, "loss": 1.2942, "step": 2686 }, { "epoch": 1.711691621535521, "grad_norm": 19.3095040264909, "learning_rate": 2.9580786177269884e-06, "loss": 1.1228, "step": 2687 }, { "epoch": 1.7123287671232876, "grad_norm": 32.09285722475379, "learning_rate": 2.953911255081915e-06, "loss": 1.6078, "step": 2688 }, { "epoch": 1.7129659127110544, "grad_norm": 16.892299837574328, "learning_rate": 2.949752468354855e-06, "loss": 1.303, "step": 2689 }, { "epoch": 1.7136030582988213, "grad_norm": 24.043097988216477, "learning_rate": 2.9456022619784487e-06, "loss": 0.9609, "step": 2690 }, { "epoch": 1.714240203886588, "grad_norm": 20.706633529127984, "learning_rate": 2.941460640376189e-06, "loss": 1.1484, "step": 2691 }, { "epoch": 1.7148773494743548, "grad_norm": 31.50248418163455, "learning_rate": 2.937327607962423e-06, "loss": 1.4045, "step": 2692 }, { "epoch": 1.7155144950621217, "grad_norm": 22.80048975904876, "learning_rate": 2.9332031691423364e-06, "loss": 1.5227, "step": 2693 }, { "epoch": 1.7161516406498885, "grad_norm": 23.362510784020404, "learning_rate": 2.929087328311963e-06, "loss": 1.0842, "step": 2694 }, { "epoch": 1.7167887862376552, "grad_norm": 36.409970349486244, "learning_rate": 2.9249800898581647e-06, "loss": 1.1503, "step": 2695 }, { "epoch": 1.7174259318254221, "grad_norm": 21.770521616565976, "learning_rate": 2.920881458158643e-06, "loss": 1.2349, "step": 2696 }, { "epoch": 1.7180630774131889, "grad_norm": 18.830849658965693, "learning_rate": 2.9167914375819188e-06, "loss": 1.6003, "step": 2697 }, { "epoch": 1.7187002230009556, "grad_norm": 22.34007334241361, "learning_rate": 2.912710032487339e-06, "loss": 1.2141, "step": 2698 }, { "epoch": 1.7193373685887225, "grad_norm": 16.21752461238933, "learning_rate": 2.9086372472250678e-06, "loss": 1.2476, "step": 2699 }, { "epoch": 1.7199745141764893, "grad_norm": 26.257257729795974, "learning_rate": 2.904573086136077e-06, "loss": 1.261, "step": 2700 }, { "epoch": 1.720611659764256, "grad_norm": 17.430393046038972, "learning_rate": 2.9005175535521533e-06, "loss": 0.7399, "step": 2701 }, { "epoch": 1.721248805352023, "grad_norm": 13.68518979048532, "learning_rate": 2.896470653795879e-06, "loss": 1.1777, "step": 2702 }, { "epoch": 1.7218859509397897, "grad_norm": 21.273750391213277, "learning_rate": 2.892432391180643e-06, "loss": 1.6223, "step": 2703 }, { "epoch": 1.7225230965275564, "grad_norm": 17.80038930996724, "learning_rate": 2.888402770010624e-06, "loss": 0.9247, "step": 2704 }, { "epoch": 1.7231602421153234, "grad_norm": 50.63803579277661, "learning_rate": 2.884381794580789e-06, "loss": 0.8588, "step": 2705 }, { "epoch": 1.7237973877030903, "grad_norm": 21.055035280975083, "learning_rate": 2.8803694691768957e-06, "loss": 1.3782, "step": 2706 }, { "epoch": 1.7244345332908568, "grad_norm": 13.427782368233801, "learning_rate": 2.876365798075474e-06, "loss": 0.9528, "step": 2707 }, { "epoch": 1.7250716788786238, "grad_norm": 16.2257611243128, "learning_rate": 2.872370785543837e-06, "loss": 1.1799, "step": 2708 }, { "epoch": 1.7257088244663907, "grad_norm": 19.13292296758699, "learning_rate": 2.868384435840062e-06, "loss": 1.2036, "step": 2709 }, { "epoch": 1.7263459700541572, "grad_norm": 14.168815571268409, "learning_rate": 2.8644067532130037e-06, "loss": 1.2408, "step": 2710 }, { "epoch": 1.7269831156419242, "grad_norm": 15.963407110857881, "learning_rate": 2.860437741902265e-06, "loss": 1.0891, "step": 2711 }, { "epoch": 1.7276202612296911, "grad_norm": 14.253267044628988, "learning_rate": 2.8564774061382185e-06, "loss": 1.4771, "step": 2712 }, { "epoch": 1.7282574068174577, "grad_norm": 17.544359445303908, "learning_rate": 2.852525750141986e-06, "loss": 1.1907, "step": 2713 }, { "epoch": 1.7288945524052246, "grad_norm": 16.978486840519366, "learning_rate": 2.8485827781254317e-06, "loss": 1.0391, "step": 2714 }, { "epoch": 1.7295316979929916, "grad_norm": 16.260947764875525, "learning_rate": 2.8446484942911767e-06, "loss": 1.3323, "step": 2715 }, { "epoch": 1.730168843580758, "grad_norm": 25.344161167725883, "learning_rate": 2.8407229028325657e-06, "loss": 1.3796, "step": 2716 }, { "epoch": 1.730805989168525, "grad_norm": 24.741950450572613, "learning_rate": 2.8368060079336984e-06, "loss": 1.0181, "step": 2717 }, { "epoch": 1.731443134756292, "grad_norm": 23.662546879987225, "learning_rate": 2.832897813769386e-06, "loss": 1.2994, "step": 2718 }, { "epoch": 1.7320802803440585, "grad_norm": 22.049376444678508, "learning_rate": 2.828998324505178e-06, "loss": 1.2427, "step": 2719 }, { "epoch": 1.7327174259318254, "grad_norm": 34.59576153677471, "learning_rate": 2.825107544297343e-06, "loss": 1.1941, "step": 2720 }, { "epoch": 1.7333545715195924, "grad_norm": 18.82795929534228, "learning_rate": 2.8212254772928627e-06, "loss": 1.4333, "step": 2721 }, { "epoch": 1.7339917171073589, "grad_norm": 19.346413313505412, "learning_rate": 2.817352127629439e-06, "loss": 1.2898, "step": 2722 }, { "epoch": 1.7346288626951258, "grad_norm": 12.979603939642336, "learning_rate": 2.8134874994354778e-06, "loss": 1.4672, "step": 2723 }, { "epoch": 1.7352660082828928, "grad_norm": 36.147910058382315, "learning_rate": 2.809631596830092e-06, "loss": 1.3301, "step": 2724 }, { "epoch": 1.7359031538706593, "grad_norm": 20.566121198687654, "learning_rate": 2.805784423923088e-06, "loss": 0.9576, "step": 2725 }, { "epoch": 1.7365402994584263, "grad_norm": 31.51971485783815, "learning_rate": 2.8019459848149767e-06, "loss": 1.0505, "step": 2726 }, { "epoch": 1.7371774450461932, "grad_norm": 14.666099439182265, "learning_rate": 2.798116283596956e-06, "loss": 1.3491, "step": 2727 }, { "epoch": 1.7378145906339597, "grad_norm": 13.50124317701223, "learning_rate": 2.7942953243509063e-06, "loss": 0.5328, "step": 2728 }, { "epoch": 1.7384517362217267, "grad_norm": 11.723123810838048, "learning_rate": 2.7904831111493976e-06, "loss": 1.3956, "step": 2729 }, { "epoch": 1.7390888818094936, "grad_norm": 21.99900765676368, "learning_rate": 2.7866796480556723e-06, "loss": 1.4219, "step": 2730 }, { "epoch": 1.7397260273972601, "grad_norm": 16.06752845724641, "learning_rate": 2.782884939123652e-06, "loss": 0.9077, "step": 2731 }, { "epoch": 1.740363172985027, "grad_norm": 12.61269093507539, "learning_rate": 2.779098988397921e-06, "loss": 1.1316, "step": 2732 }, { "epoch": 1.741000318572794, "grad_norm": 36.11675043392593, "learning_rate": 2.775321799913735e-06, "loss": 1.1378, "step": 2733 }, { "epoch": 1.7416374641605605, "grad_norm": 30.06339492647996, "learning_rate": 2.7715533776970074e-06, "loss": 1.4227, "step": 2734 }, { "epoch": 1.7422746097483275, "grad_norm": 20.634713149823096, "learning_rate": 2.767793725764306e-06, "loss": 1.5239, "step": 2735 }, { "epoch": 1.7429117553360944, "grad_norm": 20.242126471397786, "learning_rate": 2.7640428481228555e-06, "loss": 0.9338, "step": 2736 }, { "epoch": 1.743548900923861, "grad_norm": 14.229292135094695, "learning_rate": 2.7603007487705245e-06, "loss": 0.9438, "step": 2737 }, { "epoch": 1.744186046511628, "grad_norm": 23.400778159537985, "learning_rate": 2.7565674316958296e-06, "loss": 1.5962, "step": 2738 }, { "epoch": 1.7448231920993948, "grad_norm": 24.571413129042387, "learning_rate": 2.752842900877921e-06, "loss": 0.6995, "step": 2739 }, { "epoch": 1.7454603376871614, "grad_norm": 16.193603411883544, "learning_rate": 2.7491271602865897e-06, "loss": 1.2788, "step": 2740 }, { "epoch": 1.7460974832749283, "grad_norm": 16.364492721450087, "learning_rate": 2.7454202138822542e-06, "loss": 1.3153, "step": 2741 }, { "epoch": 1.7467346288626953, "grad_norm": 19.230248799748686, "learning_rate": 2.741722065615959e-06, "loss": 1.3433, "step": 2742 }, { "epoch": 1.7473717744504618, "grad_norm": 25.509421762808735, "learning_rate": 2.738032719429376e-06, "loss": 1.2705, "step": 2743 }, { "epoch": 1.7480089200382287, "grad_norm": 17.56807574921326, "learning_rate": 2.73435217925479e-06, "loss": 1.3127, "step": 2744 }, { "epoch": 1.7486460656259957, "grad_norm": 18.45690315884839, "learning_rate": 2.730680449015104e-06, "loss": 1.1246, "step": 2745 }, { "epoch": 1.7492832112137622, "grad_norm": 19.92706506916137, "learning_rate": 2.727017532623828e-06, "loss": 1.3756, "step": 2746 }, { "epoch": 1.7499203568015291, "grad_norm": 30.02506083914947, "learning_rate": 2.723363433985077e-06, "loss": 1.2134, "step": 2747 }, { "epoch": 1.750557502389296, "grad_norm": 13.69017115924802, "learning_rate": 2.7197181569935737e-06, "loss": 1.4673, "step": 2748 }, { "epoch": 1.7511946479770628, "grad_norm": 22.43080793431401, "learning_rate": 2.7160817055346287e-06, "loss": 1.3512, "step": 2749 }, { "epoch": 1.7518317935648295, "grad_norm": 20.569938128012794, "learning_rate": 2.7124540834841563e-06, "loss": 1.1699, "step": 2750 }, { "epoch": 1.7524689391525965, "grad_norm": 22.343926146994992, "learning_rate": 2.708835294708652e-06, "loss": 1.0979, "step": 2751 }, { "epoch": 1.7531060847403632, "grad_norm": 24.418669153182353, "learning_rate": 2.7052253430652013e-06, "loss": 1.3107, "step": 2752 }, { "epoch": 1.75374323032813, "grad_norm": 19.989466259128815, "learning_rate": 2.7016242324014656e-06, "loss": 1.2645, "step": 2753 }, { "epoch": 1.754380375915897, "grad_norm": 24.84121869320209, "learning_rate": 2.6980319665556883e-06, "loss": 1.2939, "step": 2754 }, { "epoch": 1.7550175215036636, "grad_norm": 13.781637654464562, "learning_rate": 2.694448549356684e-06, "loss": 1.01, "step": 2755 }, { "epoch": 1.7556546670914304, "grad_norm": 13.187412059529235, "learning_rate": 2.6908739846238317e-06, "loss": 1.3579, "step": 2756 }, { "epoch": 1.7562918126791973, "grad_norm": 18.835825224165045, "learning_rate": 2.6873082761670845e-06, "loss": 1.2261, "step": 2757 }, { "epoch": 1.756928958266964, "grad_norm": 38.772655474967806, "learning_rate": 2.6837514277869446e-06, "loss": 1.3542, "step": 2758 }, { "epoch": 1.7575661038547308, "grad_norm": 27.646417567520903, "learning_rate": 2.6802034432744815e-06, "loss": 1.5459, "step": 2759 }, { "epoch": 1.7582032494424977, "grad_norm": 30.787605936725768, "learning_rate": 2.676664326411308e-06, "loss": 1.3735, "step": 2760 }, { "epoch": 1.7588403950302645, "grad_norm": 23.014948093565557, "learning_rate": 2.67313408096959e-06, "loss": 1.7, "step": 2761 }, { "epoch": 1.7594775406180312, "grad_norm": 25.030439638128687, "learning_rate": 2.6696127107120375e-06, "loss": 1.1577, "step": 2762 }, { "epoch": 1.7601146862057981, "grad_norm": 31.553541643032222, "learning_rate": 2.6661002193919012e-06, "loss": 1.5435, "step": 2763 }, { "epoch": 1.7607518317935649, "grad_norm": 19.17383396771835, "learning_rate": 2.662596610752969e-06, "loss": 1.4011, "step": 2764 }, { "epoch": 1.7613889773813316, "grad_norm": 17.77713537101136, "learning_rate": 2.6591018885295554e-06, "loss": 0.8496, "step": 2765 }, { "epoch": 1.7620261229690986, "grad_norm": 17.25776735722316, "learning_rate": 2.6556160564465124e-06, "loss": 1.1061, "step": 2766 }, { "epoch": 1.7626632685568653, "grad_norm": 40.21332118565099, "learning_rate": 2.652139118219207e-06, "loss": 1.5303, "step": 2767 }, { "epoch": 1.763300414144632, "grad_norm": 14.558848647721664, "learning_rate": 2.648671077553534e-06, "loss": 1.0928, "step": 2768 }, { "epoch": 1.763937559732399, "grad_norm": 27.05490396670792, "learning_rate": 2.6452119381459014e-06, "loss": 1.7036, "step": 2769 }, { "epoch": 1.7645747053201657, "grad_norm": 13.125222935004011, "learning_rate": 2.6417617036832303e-06, "loss": 0.8651, "step": 2770 }, { "epoch": 1.7652118509079324, "grad_norm": 15.501715657938304, "learning_rate": 2.6383203778429536e-06, "loss": 1.3094, "step": 2771 }, { "epoch": 1.7658489964956994, "grad_norm": 50.72522064669373, "learning_rate": 2.634887964293002e-06, "loss": 1.3888, "step": 2772 }, { "epoch": 1.766486142083466, "grad_norm": 32.051463285094734, "learning_rate": 2.631464466691814e-06, "loss": 1.4146, "step": 2773 }, { "epoch": 1.7671232876712328, "grad_norm": 27.115748258627626, "learning_rate": 2.6280498886883182e-06, "loss": 1.4441, "step": 2774 }, { "epoch": 1.7677604332589998, "grad_norm": 19.720186121336802, "learning_rate": 2.6246442339219434e-06, "loss": 1.5928, "step": 2775 }, { "epoch": 1.7683975788467665, "grad_norm": 28.47197491975781, "learning_rate": 2.621247506022605e-06, "loss": 1.3872, "step": 2776 }, { "epoch": 1.7690347244345332, "grad_norm": 27.88118055796459, "learning_rate": 2.6178597086107005e-06, "loss": 0.4589, "step": 2777 }, { "epoch": 1.7696718700223002, "grad_norm": 11.88859164475557, "learning_rate": 2.6144808452971133e-06, "loss": 0.6812, "step": 2778 }, { "epoch": 1.770309015610067, "grad_norm": 14.709729808484052, "learning_rate": 2.6111109196832017e-06, "loss": 1.0745, "step": 2779 }, { "epoch": 1.7709461611978337, "grad_norm": 19.31275193167072, "learning_rate": 2.6077499353607988e-06, "loss": 1.2444, "step": 2780 }, { "epoch": 1.7715833067856006, "grad_norm": 64.05105186851704, "learning_rate": 2.6043978959122067e-06, "loss": 1.1963, "step": 2781 }, { "epoch": 1.7722204523733673, "grad_norm": 22.617322806760892, "learning_rate": 2.601054804910195e-06, "loss": 1.3396, "step": 2782 }, { "epoch": 1.772857597961134, "grad_norm": 18.554977783561604, "learning_rate": 2.597720665917993e-06, "loss": 1.3499, "step": 2783 }, { "epoch": 1.773494743548901, "grad_norm": 22.983184459483766, "learning_rate": 2.5943954824892933e-06, "loss": 0.6693, "step": 2784 }, { "epoch": 1.7741318891366678, "grad_norm": 7.108915415201197, "learning_rate": 2.59107925816824e-06, "loss": 0.8119, "step": 2785 }, { "epoch": 1.7747690347244345, "grad_norm": 16.717929753564842, "learning_rate": 2.587771996489426e-06, "loss": 1.1318, "step": 2786 }, { "epoch": 1.7754061803122014, "grad_norm": 36.68036613743465, "learning_rate": 2.5844737009778977e-06, "loss": 1.5017, "step": 2787 }, { "epoch": 1.7760433258999682, "grad_norm": 21.74438953282182, "learning_rate": 2.5811843751491355e-06, "loss": 1.0809, "step": 2788 }, { "epoch": 1.776680471487735, "grad_norm": 13.63311518880439, "learning_rate": 2.577904022509069e-06, "loss": 1.2693, "step": 2789 }, { "epoch": 1.7773176170755018, "grad_norm": 22.827650148004203, "learning_rate": 2.574632646554059e-06, "loss": 1.0183, "step": 2790 }, { "epoch": 1.7779547626632686, "grad_norm": 17.092690747278585, "learning_rate": 2.5713702507708994e-06, "loss": 0.9946, "step": 2791 }, { "epoch": 1.7785919082510353, "grad_norm": 14.13688442145234, "learning_rate": 2.568116838636813e-06, "loss": 0.9216, "step": 2792 }, { "epoch": 1.7792290538388023, "grad_norm": 30.42805488783326, "learning_rate": 2.5648724136194453e-06, "loss": 1.1562, "step": 2793 }, { "epoch": 1.779866199426569, "grad_norm": 15.748809729696442, "learning_rate": 2.5616369791768657e-06, "loss": 1.3345, "step": 2794 }, { "epoch": 1.7805033450143357, "grad_norm": 15.419999484155717, "learning_rate": 2.558410538757558e-06, "loss": 0.6656, "step": 2795 }, { "epoch": 1.7811404906021027, "grad_norm": 14.382347747833911, "learning_rate": 2.5551930958004234e-06, "loss": 1.1951, "step": 2796 }, { "epoch": 1.7817776361898694, "grad_norm": 21.264492416845407, "learning_rate": 2.5519846537347685e-06, "loss": 1.5999, "step": 2797 }, { "epoch": 1.7824147817776361, "grad_norm": 16.83883549358605, "learning_rate": 2.5487852159803126e-06, "loss": 0.8663, "step": 2798 }, { "epoch": 1.783051927365403, "grad_norm": 13.996992336113946, "learning_rate": 2.545594785947174e-06, "loss": 0.7162, "step": 2799 }, { "epoch": 1.7836890729531698, "grad_norm": 10.93931405929045, "learning_rate": 2.542413367035867e-06, "loss": 1.2992, "step": 2800 }, { "epoch": 1.7843262185409365, "grad_norm": 21.219206443554764, "learning_rate": 2.5392409626373075e-06, "loss": 1.5444, "step": 2801 }, { "epoch": 1.7849633641287035, "grad_norm": 20.19291056630913, "learning_rate": 2.536077576132797e-06, "loss": 1.3647, "step": 2802 }, { "epoch": 1.7856005097164702, "grad_norm": 15.611130814298464, "learning_rate": 2.532923210894034e-06, "loss": 1.4248, "step": 2803 }, { "epoch": 1.786237655304237, "grad_norm": 26.97991352863327, "learning_rate": 2.5297778702830927e-06, "loss": 1.1499, "step": 2804 }, { "epoch": 1.786874800892004, "grad_norm": 17.504292769012398, "learning_rate": 2.5266415576524334e-06, "loss": 0.9742, "step": 2805 }, { "epoch": 1.7875119464797706, "grad_norm": 13.995863392769557, "learning_rate": 2.523514276344893e-06, "loss": 1.1333, "step": 2806 }, { "epoch": 1.7881490920675374, "grad_norm": 20.842425080975374, "learning_rate": 2.520396029693681e-06, "loss": 1.811, "step": 2807 }, { "epoch": 1.7887862376553043, "grad_norm": 17.04540632635262, "learning_rate": 2.5172868210223777e-06, "loss": 1.1172, "step": 2808 }, { "epoch": 1.789423383243071, "grad_norm": 43.703568553152245, "learning_rate": 2.514186653644933e-06, "loss": 1.384, "step": 2809 }, { "epoch": 1.7900605288308378, "grad_norm": 12.088161899173954, "learning_rate": 2.5110955308656576e-06, "loss": 1.1215, "step": 2810 }, { "epoch": 1.7906976744186047, "grad_norm": 14.872811664489031, "learning_rate": 2.508013455979224e-06, "loss": 1.0675, "step": 2811 }, { "epoch": 1.7913348200063715, "grad_norm": 19.04845258335751, "learning_rate": 2.504940432270656e-06, "loss": 1.2961, "step": 2812 }, { "epoch": 1.7919719655941382, "grad_norm": 13.525258968591617, "learning_rate": 2.501876463015337e-06, "loss": 1.75, "step": 2813 }, { "epoch": 1.7926091111819051, "grad_norm": 16.99171407651165, "learning_rate": 2.4988215514789934e-06, "loss": 0.8751, "step": 2814 }, { "epoch": 1.7932462567696719, "grad_norm": 17.102462617807426, "learning_rate": 2.495775700917703e-06, "loss": 1.1846, "step": 2815 }, { "epoch": 1.7938834023574386, "grad_norm": 17.118891860031454, "learning_rate": 2.4927389145778817e-06, "loss": 1.366, "step": 2816 }, { "epoch": 1.7945205479452055, "grad_norm": 24.71448499955079, "learning_rate": 2.489711195696287e-06, "loss": 1.7217, "step": 2817 }, { "epoch": 1.7951576935329723, "grad_norm": 33.229602994636444, "learning_rate": 2.4866925475000125e-06, "loss": 1.1924, "step": 2818 }, { "epoch": 1.795794839120739, "grad_norm": 27.149277879646082, "learning_rate": 2.4836829732064795e-06, "loss": 1.4517, "step": 2819 }, { "epoch": 1.796431984708506, "grad_norm": 24.846919161481917, "learning_rate": 2.480682476023443e-06, "loss": 1.1793, "step": 2820 }, { "epoch": 1.7970691302962727, "grad_norm": 19.554047184853772, "learning_rate": 2.477691059148979e-06, "loss": 1.4385, "step": 2821 }, { "epoch": 1.7977062758840394, "grad_norm": 17.836430004592657, "learning_rate": 2.474708725771487e-06, "loss": 1.1232, "step": 2822 }, { "epoch": 1.7983434214718064, "grad_norm": 17.490368509233864, "learning_rate": 2.471735479069685e-06, "loss": 1.6313, "step": 2823 }, { "epoch": 1.798980567059573, "grad_norm": 19.305233097377428, "learning_rate": 2.468771322212608e-06, "loss": 1.3511, "step": 2824 }, { "epoch": 1.7996177126473398, "grad_norm": 18.19896831084442, "learning_rate": 2.465816258359599e-06, "loss": 1.1627, "step": 2825 }, { "epoch": 1.8002548582351068, "grad_norm": 41.74751063216226, "learning_rate": 2.4628702906603106e-06, "loss": 1.1729, "step": 2826 }, { "epoch": 1.8008920038228735, "grad_norm": 15.364010091875604, "learning_rate": 2.4599334222547014e-06, "loss": 1.5601, "step": 2827 }, { "epoch": 1.8015291494106402, "grad_norm": 16.58432296114247, "learning_rate": 2.4570056562730285e-06, "loss": 1.7786, "step": 2828 }, { "epoch": 1.8021662949984072, "grad_norm": 18.99819023834427, "learning_rate": 2.4540869958358503e-06, "loss": 0.8408, "step": 2829 }, { "epoch": 1.802803440586174, "grad_norm": 8.35726809955348, "learning_rate": 2.4511774440540185e-06, "loss": 1.0656, "step": 2830 }, { "epoch": 1.8034405861739407, "grad_norm": 24.900961083058178, "learning_rate": 2.4482770040286775e-06, "loss": 0.9102, "step": 2831 }, { "epoch": 1.8040777317617076, "grad_norm": 13.347211249250623, "learning_rate": 2.4453856788512608e-06, "loss": 0.5181, "step": 2832 }, { "epoch": 1.8047148773494743, "grad_norm": 12.625186319206826, "learning_rate": 2.4425034716034825e-06, "loss": 1.2711, "step": 2833 }, { "epoch": 1.805352022937241, "grad_norm": 28.685875084918486, "learning_rate": 2.4396303853573428e-06, "loss": 1.2577, "step": 2834 }, { "epoch": 1.805989168525008, "grad_norm": 15.070789530958855, "learning_rate": 2.436766423175117e-06, "loss": 0.9458, "step": 2835 }, { "epoch": 1.8066263141127747, "grad_norm": 16.788793089313565, "learning_rate": 2.4339115881093585e-06, "loss": 1.3113, "step": 2836 }, { "epoch": 1.8072634597005415, "grad_norm": 16.647851783554096, "learning_rate": 2.4310658832028914e-06, "loss": 1.344, "step": 2837 }, { "epoch": 1.8079006052883084, "grad_norm": 36.814330930266635, "learning_rate": 2.428229311488807e-06, "loss": 0.927, "step": 2838 }, { "epoch": 1.8085377508760752, "grad_norm": 14.612488173926957, "learning_rate": 2.4254018759904664e-06, "loss": 1.4659, "step": 2839 }, { "epoch": 1.8091748964638419, "grad_norm": 35.94780341001998, "learning_rate": 2.4225835797214855e-06, "loss": 1.3804, "step": 2840 }, { "epoch": 1.8098120420516088, "grad_norm": 17.951921887195056, "learning_rate": 2.4197744256857477e-06, "loss": 0.9131, "step": 2841 }, { "epoch": 1.8104491876393756, "grad_norm": 6.818660045343089, "learning_rate": 2.416974416877382e-06, "loss": 1.0958, "step": 2842 }, { "epoch": 1.8110863332271423, "grad_norm": 15.677243052694571, "learning_rate": 2.414183556280782e-06, "loss": 1.1927, "step": 2843 }, { "epoch": 1.8117234788149092, "grad_norm": 13.079321873851015, "learning_rate": 2.41140184687058e-06, "loss": 0.967, "step": 2844 }, { "epoch": 1.812360624402676, "grad_norm": 27.728227691355663, "learning_rate": 2.4086292916116617e-06, "loss": 1.262, "step": 2845 }, { "epoch": 1.8129977699904427, "grad_norm": 23.984032413808052, "learning_rate": 2.4058658934591532e-06, "loss": 1.7568, "step": 2846 }, { "epoch": 1.8136349155782097, "grad_norm": 19.780829307725973, "learning_rate": 2.4031116553584185e-06, "loss": 1.2987, "step": 2847 }, { "epoch": 1.8142720611659764, "grad_norm": 82.40233242677202, "learning_rate": 2.4003665802450634e-06, "loss": 1.3607, "step": 2848 }, { "epoch": 1.8149092067537431, "grad_norm": 12.719349471239834, "learning_rate": 2.397630671044919e-06, "loss": 1.344, "step": 2849 }, { "epoch": 1.81554635234151, "grad_norm": 17.426698655380154, "learning_rate": 2.3949039306740576e-06, "loss": 1.2571, "step": 2850 }, { "epoch": 1.8161834979292768, "grad_norm": 14.359624919547871, "learning_rate": 2.3921863620387704e-06, "loss": 1.4401, "step": 2851 }, { "epoch": 1.8168206435170435, "grad_norm": 53.994463414370486, "learning_rate": 2.3894779680355774e-06, "loss": 1.0942, "step": 2852 }, { "epoch": 1.8174577891048105, "grad_norm": 14.628515211089537, "learning_rate": 2.386778751551219e-06, "loss": 1.2783, "step": 2853 }, { "epoch": 1.8180949346925772, "grad_norm": 21.264652149177355, "learning_rate": 2.384088715462651e-06, "loss": 1.2104, "step": 2854 }, { "epoch": 1.818732080280344, "grad_norm": 13.776936431512237, "learning_rate": 2.3814078626370494e-06, "loss": 1.1345, "step": 2855 }, { "epoch": 1.819369225868111, "grad_norm": 16.168528167400705, "learning_rate": 2.378736195931794e-06, "loss": 1.313, "step": 2856 }, { "epoch": 1.8200063714558776, "grad_norm": 22.258651634603332, "learning_rate": 2.376073718194486e-06, "loss": 0.6803, "step": 2857 }, { "epoch": 1.8206435170436444, "grad_norm": 15.04857726817744, "learning_rate": 2.37342043226292e-06, "loss": 1.3604, "step": 2858 }, { "epoch": 1.8212806626314113, "grad_norm": 76.16459816560227, "learning_rate": 2.3707763409651013e-06, "loss": 1.5967, "step": 2859 }, { "epoch": 1.821917808219178, "grad_norm": 22.991561448183585, "learning_rate": 2.3681414471192334e-06, "loss": 1.3767, "step": 2860 }, { "epoch": 1.8225549538069448, "grad_norm": 20.66515212283535, "learning_rate": 2.3655157535337133e-06, "loss": 1.4256, "step": 2861 }, { "epoch": 1.8231920993947117, "grad_norm": 27.095242716313958, "learning_rate": 2.362899263007137e-06, "loss": 0.9598, "step": 2862 }, { "epoch": 1.8238292449824784, "grad_norm": 31.32016532612421, "learning_rate": 2.3602919783282875e-06, "loss": 0.9896, "step": 2863 }, { "epoch": 1.8244663905702452, "grad_norm": 21.56637654833653, "learning_rate": 2.3576939022761395e-06, "loss": 1.582, "step": 2864 }, { "epoch": 1.8251035361580121, "grad_norm": 19.747592835612423, "learning_rate": 2.3551050376198486e-06, "loss": 0.9697, "step": 2865 }, { "epoch": 1.8257406817457789, "grad_norm": 38.17675147446833, "learning_rate": 2.3525253871187547e-06, "loss": 1.4536, "step": 2866 }, { "epoch": 1.8263778273335456, "grad_norm": 20.632097814292695, "learning_rate": 2.3499549535223783e-06, "loss": 1.5991, "step": 2867 }, { "epoch": 1.8270149729213125, "grad_norm": 61.208708379993695, "learning_rate": 2.3473937395704102e-06, "loss": 1.1665, "step": 2868 }, { "epoch": 1.8276521185090793, "grad_norm": 21.33775707136015, "learning_rate": 2.3448417479927217e-06, "loss": 1.2324, "step": 2869 }, { "epoch": 1.828289264096846, "grad_norm": 13.805884164197424, "learning_rate": 2.342298981509351e-06, "loss": 1.2463, "step": 2870 }, { "epoch": 1.828926409684613, "grad_norm": 34.407545600577194, "learning_rate": 2.339765442830504e-06, "loss": 1.5364, "step": 2871 }, { "epoch": 1.82956355527238, "grad_norm": 25.447040329106684, "learning_rate": 2.3372411346565502e-06, "loss": 1.6426, "step": 2872 }, { "epoch": 1.8302007008601464, "grad_norm": 20.978668837548835, "learning_rate": 2.3347260596780215e-06, "loss": 1.1494, "step": 2873 }, { "epoch": 1.8308378464479134, "grad_norm": 22.63496446587322, "learning_rate": 2.3322202205756105e-06, "loss": 1.4468, "step": 2874 }, { "epoch": 1.8314749920356803, "grad_norm": 32.208719709862514, "learning_rate": 2.32972362002016e-06, "loss": 1.0891, "step": 2875 }, { "epoch": 1.8321121376234468, "grad_norm": 10.713736082745578, "learning_rate": 2.327236260672676e-06, "loss": 1.0756, "step": 2876 }, { "epoch": 1.8327492832112138, "grad_norm": 18.207510323569448, "learning_rate": 2.3247581451843036e-06, "loss": 1.6069, "step": 2877 }, { "epoch": 1.8333864287989807, "grad_norm": 22.95782044381334, "learning_rate": 2.3222892761963446e-06, "loss": 1.6106, "step": 2878 }, { "epoch": 1.8340235743867472, "grad_norm": 18.76389936072524, "learning_rate": 2.319829656340237e-06, "loss": 1.7627, "step": 2879 }, { "epoch": 1.8346607199745142, "grad_norm": 23.171469400585313, "learning_rate": 2.317379288237567e-06, "loss": 1.4458, "step": 2880 }, { "epoch": 1.8352978655622811, "grad_norm": 30.77212775200375, "learning_rate": 2.314938174500059e-06, "loss": 1.5098, "step": 2881 }, { "epoch": 1.8359350111500476, "grad_norm": 21.292953602255373, "learning_rate": 2.3125063177295688e-06, "loss": 1.1558, "step": 2882 }, { "epoch": 1.8365721567378146, "grad_norm": 12.30485921278015, "learning_rate": 2.3100837205180927e-06, "loss": 0.7242, "step": 2883 }, { "epoch": 1.8372093023255816, "grad_norm": 32.95797290314697, "learning_rate": 2.3076703854477535e-06, "loss": 2.0366, "step": 2884 }, { "epoch": 1.837846447913348, "grad_norm": 21.233975923196194, "learning_rate": 2.305266315090802e-06, "loss": 1.2628, "step": 2885 }, { "epoch": 1.838483593501115, "grad_norm": 23.003442711041455, "learning_rate": 2.3028715120096147e-06, "loss": 1.4111, "step": 2886 }, { "epoch": 1.839120739088882, "grad_norm": 18.609498413369643, "learning_rate": 2.30048597875669e-06, "loss": 1.2795, "step": 2887 }, { "epoch": 1.8397578846766485, "grad_norm": 11.6251874918685, "learning_rate": 2.2981097178746497e-06, "loss": 1.1306, "step": 2888 }, { "epoch": 1.8403950302644154, "grad_norm": 31.6608986157923, "learning_rate": 2.2957427318962265e-06, "loss": 1.1227, "step": 2889 }, { "epoch": 1.8410321758521824, "grad_norm": 20.47743497583027, "learning_rate": 2.2933850233442735e-06, "loss": 1.4819, "step": 2890 }, { "epoch": 1.8416693214399489, "grad_norm": 22.190592411161663, "learning_rate": 2.291036594731751e-06, "loss": 1.2373, "step": 2891 }, { "epoch": 1.8423064670277158, "grad_norm": 16.447842184188787, "learning_rate": 2.2886974485617305e-06, "loss": 1.3245, "step": 2892 }, { "epoch": 1.8429436126154828, "grad_norm": 26.49707452365479, "learning_rate": 2.286367587327389e-06, "loss": 1.709, "step": 2893 }, { "epoch": 1.8435807582032493, "grad_norm": 22.906639644047083, "learning_rate": 2.2840470135120075e-06, "loss": 1.2957, "step": 2894 }, { "epoch": 1.8442179037910162, "grad_norm": 13.887135686776528, "learning_rate": 2.2817357295889687e-06, "loss": 0.9918, "step": 2895 }, { "epoch": 1.8448550493787832, "grad_norm": 13.499698994641737, "learning_rate": 2.2794337380217517e-06, "loss": 1.2815, "step": 2896 }, { "epoch": 1.8454921949665497, "grad_norm": 22.186328093635083, "learning_rate": 2.2771410412639345e-06, "loss": 1.4153, "step": 2897 }, { "epoch": 1.8461293405543167, "grad_norm": 34.25969802282154, "learning_rate": 2.2748576417591838e-06, "loss": 1.4915, "step": 2898 }, { "epoch": 1.8467664861420836, "grad_norm": 17.466096507824638, "learning_rate": 2.272583541941262e-06, "loss": 1.4155, "step": 2899 }, { "epoch": 1.8474036317298501, "grad_norm": 15.765196274154503, "learning_rate": 2.2703187442340144e-06, "loss": 1.0779, "step": 2900 }, { "epoch": 1.848040777317617, "grad_norm": 14.170486365810039, "learning_rate": 2.2680632510513752e-06, "loss": 1.416, "step": 2901 }, { "epoch": 1.848677922905384, "grad_norm": 21.306270170229872, "learning_rate": 2.265817064797359e-06, "loss": 1.3264, "step": 2902 }, { "epoch": 1.8493150684931505, "grad_norm": 24.885732846417195, "learning_rate": 2.263580187866064e-06, "loss": 1.7173, "step": 2903 }, { "epoch": 1.8499522140809175, "grad_norm": 37.75108599949561, "learning_rate": 2.261352622641664e-06, "loss": 1.3074, "step": 2904 }, { "epoch": 1.8505893596686844, "grad_norm": 16.19396601744116, "learning_rate": 2.2591343714984047e-06, "loss": 1.3538, "step": 2905 }, { "epoch": 1.851226505256451, "grad_norm": 32.18961164240791, "learning_rate": 2.2569254368006103e-06, "loss": 1.1008, "step": 2906 }, { "epoch": 1.851863650844218, "grad_norm": 23.596356594910745, "learning_rate": 2.2547258209026723e-06, "loss": 1.2714, "step": 2907 }, { "epoch": 1.8525007964319848, "grad_norm": 15.65254115361861, "learning_rate": 2.2525355261490472e-06, "loss": 1.1348, "step": 2908 }, { "epoch": 1.8531379420197513, "grad_norm": 20.00554485061519, "learning_rate": 2.2503545548742614e-06, "loss": 1.3269, "step": 2909 }, { "epoch": 1.8537750876075183, "grad_norm": 49.90672220420858, "learning_rate": 2.248182909402901e-06, "loss": 1.3738, "step": 2910 }, { "epoch": 1.8544122331952853, "grad_norm": 14.888391952337775, "learning_rate": 2.2460205920496136e-06, "loss": 0.9785, "step": 2911 }, { "epoch": 1.8550493787830518, "grad_norm": 17.492702698758986, "learning_rate": 2.243867605119103e-06, "loss": 1.297, "step": 2912 }, { "epoch": 1.8556865243708187, "grad_norm": 25.510176577536782, "learning_rate": 2.2417239509061274e-06, "loss": 1.1841, "step": 2913 }, { "epoch": 1.8563236699585857, "grad_norm": 86.71790315190687, "learning_rate": 2.239589631695501e-06, "loss": 1.6177, "step": 2914 }, { "epoch": 1.8569608155463524, "grad_norm": 26.430177560415366, "learning_rate": 2.2374646497620854e-06, "loss": 1.4847, "step": 2915 }, { "epoch": 1.8575979611341191, "grad_norm": 18.091622676670717, "learning_rate": 2.2353490073707917e-06, "loss": 1.0355, "step": 2916 }, { "epoch": 1.858235106721886, "grad_norm": 23.98645564001689, "learning_rate": 2.2332427067765757e-06, "loss": 1.394, "step": 2917 }, { "epoch": 1.8588722523096528, "grad_norm": 19.000060557273077, "learning_rate": 2.2311457502244365e-06, "loss": 1.5796, "step": 2918 }, { "epoch": 1.8595093978974195, "grad_norm": 17.301844503031102, "learning_rate": 2.2290581399494137e-06, "loss": 1.1219, "step": 2919 }, { "epoch": 1.8601465434851865, "grad_norm": 20.98850965923807, "learning_rate": 2.226979878176585e-06, "loss": 1.3335, "step": 2920 }, { "epoch": 1.8607836890729532, "grad_norm": 29.492432043085895, "learning_rate": 2.224910967121064e-06, "loss": 1.0413, "step": 2921 }, { "epoch": 1.86142083466072, "grad_norm": 65.31047622900914, "learning_rate": 2.222851408987999e-06, "loss": 1.0199, "step": 2922 }, { "epoch": 1.862057980248487, "grad_norm": 21.765739877671667, "learning_rate": 2.2208012059725666e-06, "loss": 1.3098, "step": 2923 }, { "epoch": 1.8626951258362536, "grad_norm": 21.87594880119286, "learning_rate": 2.2187603602599773e-06, "loss": 1.4381, "step": 2924 }, { "epoch": 1.8633322714240204, "grad_norm": 60.778467879747204, "learning_rate": 2.2167288740254637e-06, "loss": 1.4346, "step": 2925 }, { "epoch": 1.8639694170117873, "grad_norm": 12.719913666777572, "learning_rate": 2.2147067494342827e-06, "loss": 1.2479, "step": 2926 }, { "epoch": 1.864606562599554, "grad_norm": 16.416698999833013, "learning_rate": 2.2126939886417158e-06, "loss": 1.2422, "step": 2927 }, { "epoch": 1.8652437081873208, "grad_norm": 21.366250142930078, "learning_rate": 2.2106905937930625e-06, "loss": 0.6155, "step": 2928 }, { "epoch": 1.8658808537750877, "grad_norm": 19.623720942222352, "learning_rate": 2.2086965670236386e-06, "loss": 1.3699, "step": 2929 }, { "epoch": 1.8665179993628545, "grad_norm": 16.97875397444886, "learning_rate": 2.206711910458778e-06, "loss": 1.5786, "step": 2930 }, { "epoch": 1.8671551449506212, "grad_norm": 14.050614469375486, "learning_rate": 2.204736626213824e-06, "loss": 0.9672, "step": 2931 }, { "epoch": 1.8677922905383881, "grad_norm": 22.33766775429125, "learning_rate": 2.2027707163941337e-06, "loss": 1.1709, "step": 2932 }, { "epoch": 1.8684294361261549, "grad_norm": 14.978307826517208, "learning_rate": 2.200814183095068e-06, "loss": 0.8639, "step": 2933 }, { "epoch": 1.8690665817139216, "grad_norm": 33.77567264111755, "learning_rate": 2.1988670284019994e-06, "loss": 1.4121, "step": 2934 }, { "epoch": 1.8697037273016885, "grad_norm": 16.0447034771722, "learning_rate": 2.1969292543902985e-06, "loss": 0.5309, "step": 2935 }, { "epoch": 1.8703408728894553, "grad_norm": 13.54933534444099, "learning_rate": 2.195000863125343e-06, "loss": 1.5315, "step": 2936 }, { "epoch": 1.870978018477222, "grad_norm": 17.13070454370637, "learning_rate": 2.193081856662505e-06, "loss": 1.4541, "step": 2937 }, { "epoch": 1.871615164064989, "grad_norm": 15.129654614511827, "learning_rate": 2.1911722370471584e-06, "loss": 1.0861, "step": 2938 }, { "epoch": 1.8722523096527557, "grad_norm": 18.37714602939438, "learning_rate": 2.1892720063146688e-06, "loss": 0.7765, "step": 2939 }, { "epoch": 1.8728894552405224, "grad_norm": 17.817759445421768, "learning_rate": 2.187381166490396e-06, "loss": 1.265, "step": 2940 }, { "epoch": 1.8735266008282894, "grad_norm": 20.49312681108832, "learning_rate": 2.1854997195896895e-06, "loss": 1.3224, "step": 2941 }, { "epoch": 1.874163746416056, "grad_norm": 17.667819770798385, "learning_rate": 2.1836276676178896e-06, "loss": 0.9016, "step": 2942 }, { "epoch": 1.8748008920038228, "grad_norm": 8.523202838243936, "learning_rate": 2.1817650125703197e-06, "loss": 0.8738, "step": 2943 }, { "epoch": 1.8754380375915898, "grad_norm": 20.008387745601, "learning_rate": 2.1799117564322925e-06, "loss": 1.0737, "step": 2944 }, { "epoch": 1.8760751831793565, "grad_norm": 16.453047937436764, "learning_rate": 2.1780679011790965e-06, "loss": 0.7657, "step": 2945 }, { "epoch": 1.8767123287671232, "grad_norm": 16.491361807625843, "learning_rate": 2.1762334487760056e-06, "loss": 1.0598, "step": 2946 }, { "epoch": 1.8773494743548902, "grad_norm": 16.480048907662514, "learning_rate": 2.174408401178268e-06, "loss": 1.1177, "step": 2947 }, { "epoch": 1.877986619942657, "grad_norm": 49.51598257463096, "learning_rate": 2.1725927603311096e-06, "loss": 1.6006, "step": 2948 }, { "epoch": 1.8786237655304237, "grad_norm": 15.800995588312027, "learning_rate": 2.170786528169731e-06, "loss": 1.3049, "step": 2949 }, { "epoch": 1.8792609111181906, "grad_norm": 19.782960153616088, "learning_rate": 2.1689897066193027e-06, "loss": 1.5945, "step": 2950 }, { "epoch": 1.8798980567059573, "grad_norm": 29.97527795477022, "learning_rate": 2.167202297594967e-06, "loss": 1.6714, "step": 2951 }, { "epoch": 1.880535202293724, "grad_norm": 31.85447634466334, "learning_rate": 2.1654243030018294e-06, "loss": 1.6213, "step": 2952 }, { "epoch": 1.881172347881491, "grad_norm": 20.542169211686467, "learning_rate": 2.1636557247349693e-06, "loss": 1.1787, "step": 2953 }, { "epoch": 1.8818094934692577, "grad_norm": 13.107492328073972, "learning_rate": 2.161896564679419e-06, "loss": 1.5546, "step": 2954 }, { "epoch": 1.8824466390570245, "grad_norm": 16.345108748832537, "learning_rate": 2.160146824710181e-06, "loss": 0.9586, "step": 2955 }, { "epoch": 1.8830837846447914, "grad_norm": 26.610100528522416, "learning_rate": 2.158406506692215e-06, "loss": 1.5781, "step": 2956 }, { "epoch": 1.8837209302325582, "grad_norm": 32.20466324523368, "learning_rate": 2.156675612480436e-06, "loss": 1.2524, "step": 2957 }, { "epoch": 1.8843580758203249, "grad_norm": 18.518944885542894, "learning_rate": 2.1549541439197176e-06, "loss": 0.8224, "step": 2958 }, { "epoch": 1.8849952214080918, "grad_norm": 17.30712816513351, "learning_rate": 2.153242102844884e-06, "loss": 1.1172, "step": 2959 }, { "epoch": 1.8856323669958586, "grad_norm": 13.795788021246993, "learning_rate": 2.151539491080716e-06, "loss": 0.889, "step": 2960 }, { "epoch": 1.8862695125836253, "grad_norm": 21.623478232488782, "learning_rate": 2.149846310441939e-06, "loss": 1.3015, "step": 2961 }, { "epoch": 1.8869066581713922, "grad_norm": 17.896461769809346, "learning_rate": 2.1481625627332285e-06, "loss": 1.4834, "step": 2962 }, { "epoch": 1.887543803759159, "grad_norm": 19.37512916638427, "learning_rate": 2.1464882497492055e-06, "loss": 1.3594, "step": 2963 }, { "epoch": 1.8881809493469257, "grad_norm": 29.761327427836708, "learning_rate": 2.1448233732744367e-06, "loss": 1.2129, "step": 2964 }, { "epoch": 1.8888180949346927, "grad_norm": 25.072530616997515, "learning_rate": 2.1431679350834284e-06, "loss": 0.6123, "step": 2965 }, { "epoch": 1.8894552405224594, "grad_norm": 10.137373059009471, "learning_rate": 2.141521936940629e-06, "loss": 1.1373, "step": 2966 }, { "epoch": 1.8900923861102261, "grad_norm": 21.876867679672802, "learning_rate": 2.1398853806004248e-06, "loss": 1.4401, "step": 2967 }, { "epoch": 1.890729531697993, "grad_norm": 24.24226921850383, "learning_rate": 2.138258267807136e-06, "loss": 1.5967, "step": 2968 }, { "epoch": 1.8913666772857598, "grad_norm": 22.816109600745982, "learning_rate": 2.136640600295022e-06, "loss": 0.9197, "step": 2969 }, { "epoch": 1.8920038228735265, "grad_norm": 14.88510204471734, "learning_rate": 2.135032379788271e-06, "loss": 1.2788, "step": 2970 }, { "epoch": 1.8926409684612935, "grad_norm": 16.237104467319433, "learning_rate": 2.133433608001004e-06, "loss": 1.1409, "step": 2971 }, { "epoch": 1.8932781140490602, "grad_norm": 13.878747949992183, "learning_rate": 2.131844286637273e-06, "loss": 0.7848, "step": 2972 }, { "epoch": 1.893915259636827, "grad_norm": 17.902995445221904, "learning_rate": 2.1302644173910524e-06, "loss": 1.1948, "step": 2973 }, { "epoch": 1.894552405224594, "grad_norm": 19.236201943046936, "learning_rate": 2.1286940019462453e-06, "loss": 0.9675, "step": 2974 }, { "epoch": 1.8951895508123606, "grad_norm": 14.177951599295632, "learning_rate": 2.127133041976677e-06, "loss": 1.5862, "step": 2975 }, { "epoch": 1.8958266964001274, "grad_norm": 18.48876121457051, "learning_rate": 2.1255815391460993e-06, "loss": 1.9502, "step": 2976 }, { "epoch": 1.8964638419878943, "grad_norm": 22.921031609923084, "learning_rate": 2.124039495108175e-06, "loss": 0.8668, "step": 2977 }, { "epoch": 1.897100987575661, "grad_norm": 11.680991830592514, "learning_rate": 2.1225069115064954e-06, "loss": 1.2058, "step": 2978 }, { "epoch": 1.8977381331634278, "grad_norm": 19.888365040616605, "learning_rate": 2.120983789974562e-06, "loss": 1.4336, "step": 2979 }, { "epoch": 1.8983752787511947, "grad_norm": 20.692201675486373, "learning_rate": 2.1194701321357907e-06, "loss": 1.2373, "step": 2980 }, { "epoch": 1.8990124243389614, "grad_norm": 20.01421788717973, "learning_rate": 2.117965939603517e-06, "loss": 1.4927, "step": 2981 }, { "epoch": 1.8996495699267282, "grad_norm": 18.372470847400685, "learning_rate": 2.1164712139809777e-06, "loss": 1.3127, "step": 2982 }, { "epoch": 1.9002867155144951, "grad_norm": 24.719707290989454, "learning_rate": 2.1149859568613304e-06, "loss": 1.481, "step": 2983 }, { "epoch": 1.9009238611022619, "grad_norm": 23.664474303748555, "learning_rate": 2.1135101698276326e-06, "loss": 1.3804, "step": 2984 }, { "epoch": 1.9015610066900286, "grad_norm": 27.601531902772372, "learning_rate": 2.1120438544528517e-06, "loss": 1.0546, "step": 2985 }, { "epoch": 1.9021981522777955, "grad_norm": 15.97308353279539, "learning_rate": 2.110587012299859e-06, "loss": 1.0688, "step": 2986 }, { "epoch": 1.9028352978655623, "grad_norm": 27.974135427323393, "learning_rate": 2.1091396449214275e-06, "loss": 1.0045, "step": 2987 }, { "epoch": 1.903472443453329, "grad_norm": 12.748837786173613, "learning_rate": 2.1077017538602344e-06, "loss": 1.1455, "step": 2988 }, { "epoch": 1.904109589041096, "grad_norm": 23.976724103062786, "learning_rate": 2.1062733406488526e-06, "loss": 1.2693, "step": 2989 }, { "epoch": 1.9047467346288627, "grad_norm": 14.595177311547532, "learning_rate": 2.1048544068097588e-06, "loss": 1.3276, "step": 2990 }, { "epoch": 1.9053838802166294, "grad_norm": 17.70342307118104, "learning_rate": 2.1034449538553194e-06, "loss": 0.8513, "step": 2991 }, { "epoch": 1.9060210258043964, "grad_norm": 27.400102872847263, "learning_rate": 2.1020449832878e-06, "loss": 1.4995, "step": 2992 }, { "epoch": 1.906658171392163, "grad_norm": 22.614029530331564, "learning_rate": 2.100654496599359e-06, "loss": 1.0491, "step": 2993 }, { "epoch": 1.9072953169799298, "grad_norm": 24.041682200678682, "learning_rate": 2.0992734952720437e-06, "loss": 1.6672, "step": 2994 }, { "epoch": 1.9079324625676968, "grad_norm": 22.093319907478328, "learning_rate": 2.097901980777796e-06, "loss": 1.2046, "step": 2995 }, { "epoch": 1.9085696081554635, "grad_norm": 32.191386747188815, "learning_rate": 2.0965399545784405e-06, "loss": 1.4949, "step": 2996 }, { "epoch": 1.9092067537432302, "grad_norm": 28.242155514555044, "learning_rate": 2.095187418125696e-06, "loss": 1.7727, "step": 2997 }, { "epoch": 1.9098438993309972, "grad_norm": 26.017758148105088, "learning_rate": 2.0938443728611587e-06, "loss": 1.855, "step": 2998 }, { "epoch": 1.910481044918764, "grad_norm": 19.0962817314754, "learning_rate": 2.0925108202163137e-06, "loss": 1.3289, "step": 2999 }, { "epoch": 1.9111181905065306, "grad_norm": 20.72209611834307, "learning_rate": 2.091186761612528e-06, "loss": 1.2362, "step": 3000 }, { "epoch": 1.9117553360942976, "grad_norm": 15.40998775431217, "learning_rate": 2.0898721984610475e-06, "loss": 1.2368, "step": 3001 }, { "epoch": 1.9123924816820643, "grad_norm": 25.12888627111021, "learning_rate": 2.088567132162999e-06, "loss": 1.2864, "step": 3002 }, { "epoch": 1.913029627269831, "grad_norm": 15.091182101902499, "learning_rate": 2.0872715641093855e-06, "loss": 1.094, "step": 3003 }, { "epoch": 1.913666772857598, "grad_norm": 16.07466893724092, "learning_rate": 2.0859854956810895e-06, "loss": 0.9751, "step": 3004 }, { "epoch": 1.9143039184453647, "grad_norm": 18.11514160910553, "learning_rate": 2.0847089282488637e-06, "loss": 1.2625, "step": 3005 }, { "epoch": 1.9149410640331315, "grad_norm": 17.274832963144664, "learning_rate": 2.083441863173338e-06, "loss": 0.8479, "step": 3006 }, { "epoch": 1.9155782096208984, "grad_norm": 17.783272415368277, "learning_rate": 2.0821843018050123e-06, "loss": 1.1965, "step": 3007 }, { "epoch": 1.9162153552086651, "grad_norm": 16.34384728646077, "learning_rate": 2.0809362454842576e-06, "loss": 1.274, "step": 3008 }, { "epoch": 1.9168525007964319, "grad_norm": 16.3127342398779, "learning_rate": 2.079697695541314e-06, "loss": 1.5098, "step": 3009 }, { "epoch": 1.9174896463841988, "grad_norm": 34.16117007601095, "learning_rate": 2.0784686532962885e-06, "loss": 0.9651, "step": 3010 }, { "epoch": 1.9181267919719656, "grad_norm": 28.753058069568123, "learning_rate": 2.0772491200591564e-06, "loss": 1.3384, "step": 3011 }, { "epoch": 1.9187639375597323, "grad_norm": 11.465150361151418, "learning_rate": 2.0760390971297544e-06, "loss": 1.0862, "step": 3012 }, { "epoch": 1.9194010831474992, "grad_norm": 13.5573391310706, "learning_rate": 2.0748385857977857e-06, "loss": 1.05, "step": 3013 }, { "epoch": 1.920038228735266, "grad_norm": 81.67720737221065, "learning_rate": 2.073647587342814e-06, "loss": 1.0592, "step": 3014 }, { "epoch": 1.9206753743230327, "grad_norm": 27.29672498271581, "learning_rate": 2.072466103034264e-06, "loss": 1.36, "step": 3015 }, { "epoch": 1.9213125199107997, "grad_norm": 91.84469223474558, "learning_rate": 2.071294134131422e-06, "loss": 1.3458, "step": 3016 }, { "epoch": 1.9219496654985664, "grad_norm": 22.433993918837356, "learning_rate": 2.0701316818834265e-06, "loss": 0.9097, "step": 3017 }, { "epoch": 1.9225868110863331, "grad_norm": 14.473457725283922, "learning_rate": 2.0689787475292805e-06, "loss": 1.2249, "step": 3018 }, { "epoch": 1.9232239566741, "grad_norm": 12.24520568162093, "learning_rate": 2.0678353322978343e-06, "loss": 1.2542, "step": 3019 }, { "epoch": 1.9238611022618668, "grad_norm": 27.024012249445363, "learning_rate": 2.0667014374077994e-06, "loss": 1.0117, "step": 3020 }, { "epoch": 1.9244982478496335, "grad_norm": 21.008317859859353, "learning_rate": 2.065577064067736e-06, "loss": 1.2332, "step": 3021 }, { "epoch": 1.9251353934374005, "grad_norm": 29.039769232996907, "learning_rate": 2.0644622134760545e-06, "loss": 1.4326, "step": 3022 }, { "epoch": 1.9257725390251672, "grad_norm": 22.91224269269131, "learning_rate": 2.0633568868210202e-06, "loss": 1.1843, "step": 3023 }, { "epoch": 1.926409684612934, "grad_norm": 24.13397926002867, "learning_rate": 2.062261085280744e-06, "loss": 1.0452, "step": 3024 }, { "epoch": 1.9270468302007009, "grad_norm": 16.17525642794105, "learning_rate": 2.061174810023186e-06, "loss": 1.1689, "step": 3025 }, { "epoch": 1.9276839757884676, "grad_norm": 25.979693890660723, "learning_rate": 2.06009806220615e-06, "loss": 0.7527, "step": 3026 }, { "epoch": 1.9283211213762343, "grad_norm": 11.728383355048802, "learning_rate": 2.0590308429772875e-06, "loss": 1.0939, "step": 3027 }, { "epoch": 1.9289582669640013, "grad_norm": 22.525113405615414, "learning_rate": 2.0579731534740945e-06, "loss": 1.1831, "step": 3028 }, { "epoch": 1.929595412551768, "grad_norm": 20.136516756635032, "learning_rate": 2.056924994823907e-06, "loss": 1.6528, "step": 3029 }, { "epoch": 1.9302325581395348, "grad_norm": 34.87413219566701, "learning_rate": 2.055886368143906e-06, "loss": 0.7018, "step": 3030 }, { "epoch": 1.9308697037273017, "grad_norm": 15.22448491776063, "learning_rate": 2.0548572745411103e-06, "loss": 1.7544, "step": 3031 }, { "epoch": 1.9315068493150684, "grad_norm": 21.651538878962917, "learning_rate": 2.0538377151123782e-06, "loss": 1.1667, "step": 3032 }, { "epoch": 1.9321439949028352, "grad_norm": 19.153757347712386, "learning_rate": 2.0528276909444057e-06, "loss": 1.5332, "step": 3033 }, { "epoch": 1.9327811404906021, "grad_norm": 19.487648237812326, "learning_rate": 2.051827203113728e-06, "loss": 1.2615, "step": 3034 }, { "epoch": 1.9334182860783689, "grad_norm": 18.985030886156853, "learning_rate": 2.050836252686713e-06, "loss": 1.553, "step": 3035 }, { "epoch": 1.9340554316661356, "grad_norm": 22.39154053388185, "learning_rate": 2.0498548407195652e-06, "loss": 1.5967, "step": 3036 }, { "epoch": 1.9346925772539025, "grad_norm": 31.010417849197303, "learning_rate": 2.048882968258322e-06, "loss": 1.6733, "step": 3037 }, { "epoch": 1.9353297228416695, "grad_norm": 20.92694406227875, "learning_rate": 2.047920636338852e-06, "loss": 1.2777, "step": 3038 }, { "epoch": 1.935966868429436, "grad_norm": 29.44839533868108, "learning_rate": 2.0469678459868575e-06, "loss": 1.4087, "step": 3039 }, { "epoch": 1.936604014017203, "grad_norm": 18.647673920480987, "learning_rate": 2.0460245982178684e-06, "loss": 1.1344, "step": 3040 }, { "epoch": 1.93724115960497, "grad_norm": 21.195945485304897, "learning_rate": 2.045090894037244e-06, "loss": 1.1971, "step": 3041 }, { "epoch": 1.9378783051927364, "grad_norm": 16.95365894834779, "learning_rate": 2.0441667344401743e-06, "loss": 1.2688, "step": 3042 }, { "epoch": 1.9385154507805034, "grad_norm": 33.309308630470966, "learning_rate": 2.043252120411672e-06, "loss": 1.4954, "step": 3043 }, { "epoch": 1.9391525963682703, "grad_norm": 31.495642176041255, "learning_rate": 2.0423470529265794e-06, "loss": 1.6074, "step": 3044 }, { "epoch": 1.9397897419560368, "grad_norm": 19.118761640119384, "learning_rate": 2.0414515329495614e-06, "loss": 1.3536, "step": 3045 }, { "epoch": 1.9404268875438038, "grad_norm": 14.36843776578736, "learning_rate": 2.0405655614351076e-06, "loss": 1.1335, "step": 3046 }, { "epoch": 1.9410640331315707, "grad_norm": 14.292250408907643, "learning_rate": 2.0396891393275292e-06, "loss": 1.1344, "step": 3047 }, { "epoch": 1.9417011787193372, "grad_norm": 12.054337085770811, "learning_rate": 2.038822267560962e-06, "loss": 1.3145, "step": 3048 }, { "epoch": 1.9423383243071042, "grad_norm": 16.43411874451034, "learning_rate": 2.037964947059358e-06, "loss": 1.1221, "step": 3049 }, { "epoch": 1.9429754698948711, "grad_norm": 18.17865298288559, "learning_rate": 2.0371171787364937e-06, "loss": 1.5278, "step": 3050 }, { "epoch": 1.9436126154826376, "grad_norm": 37.34255330422712, "learning_rate": 2.0362789634959614e-06, "loss": 0.8707, "step": 3051 }, { "epoch": 1.9442497610704046, "grad_norm": 17.608188199352977, "learning_rate": 2.035450302231173e-06, "loss": 1.3562, "step": 3052 }, { "epoch": 1.9448869066581715, "grad_norm": 13.741547283935049, "learning_rate": 2.0346311958253563e-06, "loss": 1.4729, "step": 3053 }, { "epoch": 1.945524052245938, "grad_norm": 24.973473504772496, "learning_rate": 2.033821645151555e-06, "loss": 1.2299, "step": 3054 }, { "epoch": 1.946161197833705, "grad_norm": 16.01598029968431, "learning_rate": 2.0330216510726277e-06, "loss": 1.1539, "step": 3055 }, { "epoch": 1.946798343421472, "grad_norm": 22.08296415256315, "learning_rate": 2.032231214441248e-06, "loss": 1.396, "step": 3056 }, { "epoch": 1.9474354890092385, "grad_norm": 18.3591308503942, "learning_rate": 2.0314503360999022e-06, "loss": 1.562, "step": 3057 }, { "epoch": 1.9480726345970054, "grad_norm": 21.46056966897051, "learning_rate": 2.0306790168808893e-06, "loss": 1.2759, "step": 3058 }, { "epoch": 1.9487097801847724, "grad_norm": 36.832549264435585, "learning_rate": 2.0299172576063194e-06, "loss": 1.8745, "step": 3059 }, { "epoch": 1.9493469257725389, "grad_norm": 42.58608094317416, "learning_rate": 2.029165059088113e-06, "loss": 1.0488, "step": 3060 }, { "epoch": 1.9499840713603058, "grad_norm": 16.454333447134125, "learning_rate": 2.0284224221279987e-06, "loss": 1.0972, "step": 3061 }, { "epoch": 1.9506212169480728, "grad_norm": 28.216258578701407, "learning_rate": 2.0276893475175177e-06, "loss": 1.4851, "step": 3062 }, { "epoch": 1.9512583625358393, "grad_norm": 20.92171727751267, "learning_rate": 2.026965836038016e-06, "loss": 0.7823, "step": 3063 }, { "epoch": 1.9518955081236062, "grad_norm": 19.11281906526087, "learning_rate": 2.0262518884606482e-06, "loss": 0.9215, "step": 3064 }, { "epoch": 1.9525326537113732, "grad_norm": 13.158021886739286, "learning_rate": 2.0255475055463754e-06, "loss": 1.316, "step": 3065 }, { "epoch": 1.9531697992991397, "grad_norm": 31.212790884396387, "learning_rate": 2.024852688045962e-06, "loss": 1.2249, "step": 3066 }, { "epoch": 1.9538069448869066, "grad_norm": 15.694013761505602, "learning_rate": 2.0241674366999805e-06, "loss": 1.1406, "step": 3067 }, { "epoch": 1.9544440904746736, "grad_norm": 15.829258051144404, "learning_rate": 2.0234917522388035e-06, "loss": 1.6812, "step": 3068 }, { "epoch": 1.95508123606244, "grad_norm": 45.81581380708545, "learning_rate": 2.0228256353826114e-06, "loss": 1.4038, "step": 3069 }, { "epoch": 1.955718381650207, "grad_norm": 42.59728151987077, "learning_rate": 2.0221690868413817e-06, "loss": 1.7017, "step": 3070 }, { "epoch": 1.956355527237974, "grad_norm": 24.976065648595995, "learning_rate": 2.0215221073148988e-06, "loss": 1.1882, "step": 3071 }, { "epoch": 1.9569926728257405, "grad_norm": 19.664642246432837, "learning_rate": 2.020884697492744e-06, "loss": 1.3706, "step": 3072 }, { "epoch": 1.9576298184135075, "grad_norm": 22.724190384250907, "learning_rate": 2.0202568580542994e-06, "loss": 1.3923, "step": 3073 }, { "epoch": 1.9582669640012744, "grad_norm": 20.62527861490947, "learning_rate": 2.0196385896687493e-06, "loss": 1.5156, "step": 3074 }, { "epoch": 1.958904109589041, "grad_norm": 25.146336942784693, "learning_rate": 2.0190298929950724e-06, "loss": 1.4581, "step": 3075 }, { "epoch": 1.9595412551768079, "grad_norm": 27.179494418196423, "learning_rate": 2.0184307686820494e-06, "loss": 1.1376, "step": 3076 }, { "epoch": 1.9601784007645748, "grad_norm": 14.285497379383015, "learning_rate": 2.0178412173682542e-06, "loss": 1.5002, "step": 3077 }, { "epoch": 1.9608155463523416, "grad_norm": 22.44408830543417, "learning_rate": 2.0172612396820625e-06, "loss": 1.4968, "step": 3078 }, { "epoch": 1.9614526919401083, "grad_norm": 25.139092426008908, "learning_rate": 2.0166908362416415e-06, "loss": 1.2983, "step": 3079 }, { "epoch": 1.9620898375278752, "grad_norm": 22.275454621301105, "learning_rate": 2.0161300076549557e-06, "loss": 1.6807, "step": 3080 }, { "epoch": 1.962726983115642, "grad_norm": 30.081485225980767, "learning_rate": 2.0155787545197636e-06, "loss": 1.2888, "step": 3081 }, { "epoch": 1.9633641287034087, "grad_norm": 22.45327071916814, "learning_rate": 2.015037077423618e-06, "loss": 1.5392, "step": 3082 }, { "epoch": 1.9640012742911757, "grad_norm": 24.528036111956705, "learning_rate": 2.0145049769438655e-06, "loss": 1.0516, "step": 3083 }, { "epoch": 1.9646384198789424, "grad_norm": 20.428866648719715, "learning_rate": 2.0139824536476442e-06, "loss": 1.5129, "step": 3084 }, { "epoch": 1.9652755654667091, "grad_norm": 92.73736875797489, "learning_rate": 2.0134695080918855e-06, "loss": 1.5173, "step": 3085 }, { "epoch": 1.965912711054476, "grad_norm": 26.906912390728408, "learning_rate": 2.012966140823313e-06, "loss": 1.6353, "step": 3086 }, { "epoch": 1.9665498566422428, "grad_norm": 21.608682990105677, "learning_rate": 2.012472352378439e-06, "loss": 1.4733, "step": 3087 }, { "epoch": 1.9671870022300095, "grad_norm": 31.544203306235623, "learning_rate": 2.0119881432835683e-06, "loss": 1.6245, "step": 3088 }, { "epoch": 1.9678241478177765, "grad_norm": 26.936484375161424, "learning_rate": 2.011513514054794e-06, "loss": 1.0867, "step": 3089 }, { "epoch": 1.9684612934055432, "grad_norm": 36.30011108767737, "learning_rate": 2.011048465198001e-06, "loss": 1.5681, "step": 3090 }, { "epoch": 1.96909843899331, "grad_norm": 17.964415672957685, "learning_rate": 2.01059299720886e-06, "loss": 1.1572, "step": 3091 }, { "epoch": 1.969735584581077, "grad_norm": 15.498592431291849, "learning_rate": 2.010147110572831e-06, "loss": 1.0759, "step": 3092 }, { "epoch": 1.9703727301688436, "grad_norm": 18.796542936526517, "learning_rate": 2.0097108057651635e-06, "loss": 1.4402, "step": 3093 }, { "epoch": 1.9710098757566104, "grad_norm": 13.205957734712634, "learning_rate": 2.0092840832508907e-06, "loss": 1.4089, "step": 3094 }, { "epoch": 1.9716470213443773, "grad_norm": 24.62727522260786, "learning_rate": 2.0088669434848353e-06, "loss": 1.2104, "step": 3095 }, { "epoch": 1.972284166932144, "grad_norm": 29.911818848591132, "learning_rate": 2.0084593869116067e-06, "loss": 1.6738, "step": 3096 }, { "epoch": 1.9729213125199108, "grad_norm": 25.16197982108377, "learning_rate": 2.008061413965597e-06, "loss": 1.3374, "step": 3097 }, { "epoch": 1.9735584581076777, "grad_norm": 14.248735878848187, "learning_rate": 2.007673025070986e-06, "loss": 0.6725, "step": 3098 }, { "epoch": 1.9741956036954444, "grad_norm": 18.547074495627268, "learning_rate": 2.007294220641738e-06, "loss": 1.4629, "step": 3099 }, { "epoch": 1.9748327492832112, "grad_norm": 62.937360794438575, "learning_rate": 2.0069250010816004e-06, "loss": 1.49, "step": 3100 }, { "epoch": 1.9754698948709781, "grad_norm": 23.30971305634586, "learning_rate": 2.0065653667841072e-06, "loss": 1.6875, "step": 3101 }, { "epoch": 1.9761070404587449, "grad_norm": 22.388064292936388, "learning_rate": 2.006215318132573e-06, "loss": 1.373, "step": 3102 }, { "epoch": 1.9767441860465116, "grad_norm": 27.74379646891686, "learning_rate": 2.0058748555000977e-06, "loss": 0.9711, "step": 3103 }, { "epoch": 1.9773813316342785, "grad_norm": 11.174521833924118, "learning_rate": 2.0055439792495635e-06, "loss": 0.9667, "step": 3104 }, { "epoch": 1.9780184772220453, "grad_norm": 14.118554388268693, "learning_rate": 2.0052226897336334e-06, "loss": 1.0258, "step": 3105 }, { "epoch": 1.978655622809812, "grad_norm": 15.206697090901663, "learning_rate": 2.0049109872947546e-06, "loss": 1.118, "step": 3106 }, { "epoch": 1.979292768397579, "grad_norm": 15.65813434062292, "learning_rate": 2.004608872265155e-06, "loss": 1.2227, "step": 3107 }, { "epoch": 1.9799299139853457, "grad_norm": 61.16943094739625, "learning_rate": 2.0043163449668424e-06, "loss": 1.2988, "step": 3108 }, { "epoch": 1.9805670595731124, "grad_norm": 38.13112424031628, "learning_rate": 2.0040334057116074e-06, "loss": 1.2322, "step": 3109 }, { "epoch": 1.9812042051608794, "grad_norm": 19.018173758945828, "learning_rate": 2.0037600548010216e-06, "loss": 0.6494, "step": 3110 }, { "epoch": 1.981841350748646, "grad_norm": 12.363994297764082, "learning_rate": 2.0034962925264353e-06, "loss": 1.5039, "step": 3111 }, { "epoch": 1.9824784963364128, "grad_norm": 26.77601246613456, "learning_rate": 2.003242119168979e-06, "loss": 1.4288, "step": 3112 }, { "epoch": 1.9831156419241798, "grad_norm": 16.875667951307456, "learning_rate": 2.002997534999563e-06, "loss": 1.1261, "step": 3113 }, { "epoch": 1.9837527875119465, "grad_norm": 38.34477872469023, "learning_rate": 2.0027625402788773e-06, "loss": 1.5413, "step": 3114 }, { "epoch": 1.9843899330997132, "grad_norm": 14.971804544509421, "learning_rate": 2.0025371352573917e-06, "loss": 1.4648, "step": 3115 }, { "epoch": 1.9850270786874802, "grad_norm": 33.55541665212372, "learning_rate": 2.0023213201753537e-06, "loss": 1.0627, "step": 3116 }, { "epoch": 1.985664224275247, "grad_norm": 20.07733965883875, "learning_rate": 2.0021150952627887e-06, "loss": 1.8398, "step": 3117 }, { "epoch": 1.9863013698630136, "grad_norm": 35.12469764746989, "learning_rate": 2.001918460739503e-06, "loss": 1.2485, "step": 3118 }, { "epoch": 1.9869385154507806, "grad_norm": 15.907746172181305, "learning_rate": 2.0017314168150776e-06, "loss": 1.0824, "step": 3119 }, { "epoch": 1.9875756610385473, "grad_norm": 15.12590755720015, "learning_rate": 2.0015539636888737e-06, "loss": 1.2354, "step": 3120 }, { "epoch": 1.988212806626314, "grad_norm": 20.751715007544004, "learning_rate": 2.0013861015500305e-06, "loss": 1.2771, "step": 3121 }, { "epoch": 1.988849952214081, "grad_norm": 16.14288512014973, "learning_rate": 2.0012278305774625e-06, "loss": 0.8149, "step": 3122 }, { "epoch": 1.9894870978018477, "grad_norm": 16.155454912602842, "learning_rate": 2.001079150939863e-06, "loss": 1.5295, "step": 3123 }, { "epoch": 1.9901242433896145, "grad_norm": 44.20192518211058, "learning_rate": 2.000940062795703e-06, "loss": 1.6943, "step": 3124 }, { "epoch": 1.9907613889773814, "grad_norm": 13.938144089940748, "learning_rate": 2.000810566293229e-06, "loss": 0.9874, "step": 3125 }, { "epoch": 1.9913985345651481, "grad_norm": 27.321188504464192, "learning_rate": 2.000690661570464e-06, "loss": 1.6238, "step": 3126 }, { "epoch": 1.9920356801529149, "grad_norm": 19.428559341459952, "learning_rate": 2.0005803487552097e-06, "loss": 1.5474, "step": 3127 }, { "epoch": 1.9926728257406818, "grad_norm": 16.846965983130133, "learning_rate": 2.0004796279650416e-06, "loss": 0.7284, "step": 3128 }, { "epoch": 1.9933099713284486, "grad_norm": 11.794674832020137, "learning_rate": 2.0003884993073135e-06, "loss": 1.0189, "step": 3129 }, { "epoch": 1.9939471169162153, "grad_norm": 9.844677897004939, "learning_rate": 2.0003069628791556e-06, "loss": 1.1426, "step": 3130 }, { "epoch": 1.9945842625039822, "grad_norm": 18.370307174122345, "learning_rate": 2.0002350187674713e-06, "loss": 0.9001, "step": 3131 }, { "epoch": 1.995221408091749, "grad_norm": 20.11099559020925, "learning_rate": 2.0001726670489454e-06, "loss": 1.584, "step": 3132 }, { "epoch": 1.9958585536795157, "grad_norm": 29.340952521174767, "learning_rate": 2.0001199077900315e-06, "loss": 1.2808, "step": 3133 }, { "epoch": 1.9964956992672827, "grad_norm": 16.01716865395361, "learning_rate": 2.0000767410469662e-06, "loss": 1.208, "step": 3134 }, { "epoch": 1.9971328448550494, "grad_norm": 35.184263387859865, "learning_rate": 2.000043166865757e-06, "loss": 1.8032, "step": 3135 }, { "epoch": 1.9977699904428161, "grad_norm": 23.35925438786841, "learning_rate": 2.0000191852821902e-06, "loss": 1.3252, "step": 3136 }, { "epoch": 1.998407136030583, "grad_norm": 14.718121304976952, "learning_rate": 2.0000047963218263e-06, "loss": 1.2488, "step": 3137 }, { "epoch": 1.9990442816183498, "grad_norm": 20.617343538423512, "learning_rate": 2.0000000000000003e-06, "loss": 1.1246, "step": 3138 }, { "epoch": 1.9990442816183498, "step": 3138, "total_flos": 519506685820928.0, "train_loss": 1.3205695361981655, "train_runtime": 190872.0691, "train_samples_per_second": 2.105, "train_steps_per_second": 0.016 } ], "logging_steps": 1.0, "max_steps": 3138, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 250, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 519506685820928.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }