diff --git "a/llmga-Phi-3-mini-128k-pretrain/trainer_state.json" "b/llmga-Phi-3-mini-128k-pretrain/trainer_state.json" new file mode 100644--- /dev/null +++ "b/llmga-Phi-3-mini-128k-pretrain/trainer_state.json" @@ -0,0 +1,15302 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9998853342506593, + "eval_steps": 500, + "global_step": 2180, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00045866299736268775, + "grad_norm": 81.61013321561762, + "learning_rate": 1.5151515151515153e-05, + "loss": 6.922, + "step": 1 + }, + { + "epoch": 0.0009173259947253755, + "grad_norm": 102.77615476982184, + "learning_rate": 3.0303030303030306e-05, + "loss": 6.8432, + "step": 2 + }, + { + "epoch": 0.0013759889920880633, + "grad_norm": 68.32648005874286, + "learning_rate": 4.545454545454546e-05, + "loss": 6.6275, + "step": 3 + }, + { + "epoch": 0.001834651989450751, + "grad_norm": 64.99913148980559, + "learning_rate": 6.060606060606061e-05, + "loss": 5.8454, + "step": 4 + }, + { + "epoch": 0.002293314986813439, + "grad_norm": 324.81776351076377, + "learning_rate": 7.575757575757576e-05, + "loss": 6.2271, + "step": 5 + }, + { + "epoch": 0.0027519779841761265, + "grad_norm": 101.21177823061869, + "learning_rate": 9.090909090909092e-05, + "loss": 6.1464, + "step": 6 + }, + { + "epoch": 0.0032106409815388145, + "grad_norm": 60.26372696775186, + "learning_rate": 0.00010606060606060606, + "loss": 5.6737, + "step": 7 + }, + { + "epoch": 0.003669303978901502, + "grad_norm": 52.96186226133536, + "learning_rate": 0.00012121212121212122, + "loss": 6.0687, + "step": 8 + }, + { + "epoch": 0.0041279669762641896, + "grad_norm": 44.53220788772306, + "learning_rate": 0.00013636363636363637, + "loss": 5.6797, + "step": 9 + }, + { + "epoch": 0.004586629973626878, + "grad_norm": 22.21479984412214, + "learning_rate": 0.00015151515151515152, + "loss": 5.0171, + "step": 10 + }, + { + "epoch": 0.0050452929709895655, + "grad_norm": 12.23491248289535, + "learning_rate": 0.00016666666666666666, + "loss": 4.7953, + "step": 11 + }, + { + "epoch": 0.005503955968352253, + "grad_norm": 24.127187736428418, + "learning_rate": 0.00018181818181818183, + "loss": 4.9654, + "step": 12 + }, + { + "epoch": 0.0059626189657149406, + "grad_norm": 16.88739614356292, + "learning_rate": 0.00019696969696969695, + "loss": 4.8523, + "step": 13 + }, + { + "epoch": 0.006421281963077629, + "grad_norm": 8.27887286529328, + "learning_rate": 0.00021212121212121213, + "loss": 4.7525, + "step": 14 + }, + { + "epoch": 0.0068799449604403165, + "grad_norm": 6.739074597032506, + "learning_rate": 0.00022727272727272727, + "loss": 4.6121, + "step": 15 + }, + { + "epoch": 0.007338607957803004, + "grad_norm": 7.939800781471982, + "learning_rate": 0.00024242424242424245, + "loss": 4.5331, + "step": 16 + }, + { + "epoch": 0.007797270955165692, + "grad_norm": 8.808308009749643, + "learning_rate": 0.00025757575757575756, + "loss": 4.5282, + "step": 17 + }, + { + "epoch": 0.008255933952528379, + "grad_norm": 5.461296042377687, + "learning_rate": 0.00027272727272727274, + "loss": 4.2904, + "step": 18 + }, + { + "epoch": 0.008714596949891068, + "grad_norm": 4.021987858750789, + "learning_rate": 0.0002878787878787879, + "loss": 4.2635, + "step": 19 + }, + { + "epoch": 0.009173259947253756, + "grad_norm": 6.707915163562537, + "learning_rate": 0.00030303030303030303, + "loss": 4.3416, + "step": 20 + }, + { + "epoch": 0.009631922944616443, + "grad_norm": 5.6255475302350755, + "learning_rate": 0.0003181818181818182, + "loss": 4.3245, + "step": 21 + }, + { + "epoch": 0.010090585941979131, + "grad_norm": 6.069957614307274, + "learning_rate": 0.0003333333333333333, + "loss": 4.168, + "step": 22 + }, + { + "epoch": 0.010549248939341819, + "grad_norm": 2.596168617947962, + "learning_rate": 0.0003484848484848485, + "loss": 4.1157, + "step": 23 + }, + { + "epoch": 0.011007911936704506, + "grad_norm": 2.849451199963196, + "learning_rate": 0.00036363636363636367, + "loss": 4.1274, + "step": 24 + }, + { + "epoch": 0.011466574934067194, + "grad_norm": 3.481531479661074, + "learning_rate": 0.0003787878787878788, + "loss": 4.0587, + "step": 25 + }, + { + "epoch": 0.011925237931429881, + "grad_norm": 2.8905791701369195, + "learning_rate": 0.0003939393939393939, + "loss": 3.9762, + "step": 26 + }, + { + "epoch": 0.01238390092879257, + "grad_norm": 2.603466838278235, + "learning_rate": 0.00040909090909090913, + "loss": 4.0254, + "step": 27 + }, + { + "epoch": 0.012842563926155258, + "grad_norm": 2.065412467188646, + "learning_rate": 0.00042424242424242425, + "loss": 3.8674, + "step": 28 + }, + { + "epoch": 0.013301226923517945, + "grad_norm": 2.363102590681577, + "learning_rate": 0.0004393939393939394, + "loss": 4.0022, + "step": 29 + }, + { + "epoch": 0.013759889920880633, + "grad_norm": 1.90987991198935, + "learning_rate": 0.00045454545454545455, + "loss": 3.862, + "step": 30 + }, + { + "epoch": 0.01421855291824332, + "grad_norm": 2.2855304819839604, + "learning_rate": 0.0004696969696969697, + "loss": 3.8894, + "step": 31 + }, + { + "epoch": 0.014677215915606008, + "grad_norm": 1.8528627257771246, + "learning_rate": 0.0004848484848484849, + "loss": 3.9092, + "step": 32 + }, + { + "epoch": 0.015135878912968696, + "grad_norm": 1.3864459304000019, + "learning_rate": 0.0005, + "loss": 3.8944, + "step": 33 + }, + { + "epoch": 0.015594541910331383, + "grad_norm": 1.1792218712619478, + "learning_rate": 0.0005151515151515151, + "loss": 3.873, + "step": 34 + }, + { + "epoch": 0.01605320490769407, + "grad_norm": 1.0972044167455648, + "learning_rate": 0.0005303030303030302, + "loss": 3.7967, + "step": 35 + }, + { + "epoch": 0.016511867905056758, + "grad_norm": 1.2434439058742266, + "learning_rate": 0.0005454545454545455, + "loss": 3.735, + "step": 36 + }, + { + "epoch": 0.016970530902419446, + "grad_norm": 1.1544960714128294, + "learning_rate": 0.0005606060606060606, + "loss": 3.7615, + "step": 37 + }, + { + "epoch": 0.017429193899782137, + "grad_norm": 1.2017394949096787, + "learning_rate": 0.0005757575757575758, + "loss": 3.7415, + "step": 38 + }, + { + "epoch": 0.017887856897144824, + "grad_norm": 1.2049697573912759, + "learning_rate": 0.0005909090909090909, + "loss": 3.7988, + "step": 39 + }, + { + "epoch": 0.018346519894507512, + "grad_norm": 1.2644382378458148, + "learning_rate": 0.0006060606060606061, + "loss": 3.7909, + "step": 40 + }, + { + "epoch": 0.0188051828918702, + "grad_norm": 0.9387058625841772, + "learning_rate": 0.0006212121212121212, + "loss": 3.571, + "step": 41 + }, + { + "epoch": 0.019263845889232887, + "grad_norm": 0.9449168750378817, + "learning_rate": 0.0006363636363636364, + "loss": 3.6863, + "step": 42 + }, + { + "epoch": 0.019722508886595574, + "grad_norm": 0.9725496211609492, + "learning_rate": 0.0006515151515151515, + "loss": 3.7863, + "step": 43 + }, + { + "epoch": 0.020181171883958262, + "grad_norm": 1.052216827303055, + "learning_rate": 0.0006666666666666666, + "loss": 3.6725, + "step": 44 + }, + { + "epoch": 0.02063983488132095, + "grad_norm": 0.9640972630073015, + "learning_rate": 0.0006818181818181818, + "loss": 3.6352, + "step": 45 + }, + { + "epoch": 0.021098497878683637, + "grad_norm": 1.1722011860151713, + "learning_rate": 0.000696969696969697, + "loss": 3.7247, + "step": 46 + }, + { + "epoch": 0.021557160876046325, + "grad_norm": 0.8253249488872249, + "learning_rate": 0.0007121212121212122, + "loss": 3.6106, + "step": 47 + }, + { + "epoch": 0.022015823873409012, + "grad_norm": 0.8180161691112922, + "learning_rate": 0.0007272727272727273, + "loss": 3.5795, + "step": 48 + }, + { + "epoch": 0.0224744868707717, + "grad_norm": 0.9487313385741145, + "learning_rate": 0.0007424242424242425, + "loss": 3.8047, + "step": 49 + }, + { + "epoch": 0.022933149868134387, + "grad_norm": 0.8457288932684505, + "learning_rate": 0.0007575757575757576, + "loss": 3.6836, + "step": 50 + }, + { + "epoch": 0.023391812865497075, + "grad_norm": 0.8655586113209142, + "learning_rate": 0.0007727272727272727, + "loss": 3.6827, + "step": 51 + }, + { + "epoch": 0.023850475862859762, + "grad_norm": 0.8748687859867682, + "learning_rate": 0.0007878787878787878, + "loss": 3.7566, + "step": 52 + }, + { + "epoch": 0.024309138860222453, + "grad_norm": 0.8983732434054743, + "learning_rate": 0.000803030303030303, + "loss": 3.7326, + "step": 53 + }, + { + "epoch": 0.02476780185758514, + "grad_norm": 0.887250356572744, + "learning_rate": 0.0008181818181818183, + "loss": 3.6119, + "step": 54 + }, + { + "epoch": 0.02522646485494783, + "grad_norm": 0.75096646139792, + "learning_rate": 0.0008333333333333334, + "loss": 3.6629, + "step": 55 + }, + { + "epoch": 0.025685127852310516, + "grad_norm": 0.8606633544255665, + "learning_rate": 0.0008484848484848485, + "loss": 3.7052, + "step": 56 + }, + { + "epoch": 0.026143790849673203, + "grad_norm": 0.7458855113938381, + "learning_rate": 0.0008636363636363636, + "loss": 3.6434, + "step": 57 + }, + { + "epoch": 0.02660245384703589, + "grad_norm": 0.70705232312968, + "learning_rate": 0.0008787878787878789, + "loss": 3.6799, + "step": 58 + }, + { + "epoch": 0.02706111684439858, + "grad_norm": 0.7698534281963185, + "learning_rate": 0.000893939393939394, + "loss": 3.5982, + "step": 59 + }, + { + "epoch": 0.027519779841761266, + "grad_norm": 0.823236947689985, + "learning_rate": 0.0009090909090909091, + "loss": 3.5524, + "step": 60 + }, + { + "epoch": 0.027978442839123954, + "grad_norm": 0.7485182080923718, + "learning_rate": 0.0009242424242424242, + "loss": 3.5144, + "step": 61 + }, + { + "epoch": 0.02843710583648664, + "grad_norm": 0.8376453553093982, + "learning_rate": 0.0009393939393939394, + "loss": 3.5321, + "step": 62 + }, + { + "epoch": 0.02889576883384933, + "grad_norm": 0.9597356165048272, + "learning_rate": 0.0009545454545454546, + "loss": 3.5853, + "step": 63 + }, + { + "epoch": 0.029354431831212016, + "grad_norm": 0.8856165493749967, + "learning_rate": 0.0009696969696969698, + "loss": 3.7146, + "step": 64 + }, + { + "epoch": 0.029813094828574704, + "grad_norm": 0.817865707173511, + "learning_rate": 0.000984848484848485, + "loss": 3.5218, + "step": 65 + }, + { + "epoch": 0.03027175782593739, + "grad_norm": 0.8416056081655476, + "learning_rate": 0.001, + "loss": 3.5656, + "step": 66 + }, + { + "epoch": 0.03073042082330008, + "grad_norm": 0.7633644132921669, + "learning_rate": 0.0009999994478847943, + "loss": 3.5547, + "step": 67 + }, + { + "epoch": 0.031189083820662766, + "grad_norm": 0.8705010289708955, + "learning_rate": 0.0009999977915403962, + "loss": 3.4937, + "step": 68 + }, + { + "epoch": 0.03164774681802546, + "grad_norm": 0.9511426103646287, + "learning_rate": 0.0009999950309704639, + "loss": 3.5824, + "step": 69 + }, + { + "epoch": 0.03210640981538814, + "grad_norm": 0.7848564806093093, + "learning_rate": 0.000999991166181094, + "loss": 3.4495, + "step": 70 + }, + { + "epoch": 0.03256507281275083, + "grad_norm": 0.764009584016194, + "learning_rate": 0.0009999861971808216, + "loss": 3.4964, + "step": 71 + }, + { + "epoch": 0.033023735810113516, + "grad_norm": 0.9323236501535059, + "learning_rate": 0.0009999801239806208, + "loss": 3.58, + "step": 72 + }, + { + "epoch": 0.03348239880747621, + "grad_norm": 0.887301348144735, + "learning_rate": 0.0009999729465939035, + "loss": 3.4626, + "step": 73 + }, + { + "epoch": 0.03394106180483889, + "grad_norm": 0.716690469560034, + "learning_rate": 0.0009999646650365212, + "loss": 3.5119, + "step": 74 + }, + { + "epoch": 0.03439972480220158, + "grad_norm": 0.8429566579587359, + "learning_rate": 0.0009999552793267634, + "loss": 3.5216, + "step": 75 + }, + { + "epoch": 0.034858387799564274, + "grad_norm": 0.8773059607574444, + "learning_rate": 0.0009999447894853577, + "loss": 3.4817, + "step": 76 + }, + { + "epoch": 0.03531705079692696, + "grad_norm": 0.9024057495012673, + "learning_rate": 0.0009999331955354708, + "loss": 3.5387, + "step": 77 + }, + { + "epoch": 0.03577571379428965, + "grad_norm": 0.8942405788945225, + "learning_rate": 0.0009999204975027073, + "loss": 3.4907, + "step": 78 + }, + { + "epoch": 0.03623437679165233, + "grad_norm": 0.994278882175091, + "learning_rate": 0.0009999066954151103, + "loss": 3.4892, + "step": 79 + }, + { + "epoch": 0.036693039789015024, + "grad_norm": 0.9643351586199568, + "learning_rate": 0.0009998917893031614, + "loss": 3.5265, + "step": 80 + }, + { + "epoch": 0.03715170278637771, + "grad_norm": 0.9377025959534837, + "learning_rate": 0.0009998757791997801, + "loss": 3.3998, + "step": 81 + }, + { + "epoch": 0.0376103657837404, + "grad_norm": 0.9065636490898342, + "learning_rate": 0.0009998586651403238, + "loss": 3.4957, + "step": 82 + }, + { + "epoch": 0.03806902878110308, + "grad_norm": 1.2786641451327276, + "learning_rate": 0.0009998404471625885, + "loss": 3.472, + "step": 83 + }, + { + "epoch": 0.038527691778465774, + "grad_norm": 1.0559713761547163, + "learning_rate": 0.0009998211253068078, + "loss": 3.4515, + "step": 84 + }, + { + "epoch": 0.03898635477582846, + "grad_norm": 1.2812544273839257, + "learning_rate": 0.0009998006996156535, + "loss": 3.5111, + "step": 85 + }, + { + "epoch": 0.03944501777319115, + "grad_norm": 1.072264190193131, + "learning_rate": 0.0009997791701342347, + "loss": 3.5403, + "step": 86 + }, + { + "epoch": 0.03990368077055383, + "grad_norm": 0.9386293034857816, + "learning_rate": 0.0009997565369100983, + "loss": 3.3371, + "step": 87 + }, + { + "epoch": 0.040362343767916524, + "grad_norm": 1.1609488016800573, + "learning_rate": 0.0009997327999932291, + "loss": 3.416, + "step": 88 + }, + { + "epoch": 0.04082100676527921, + "grad_norm": 1.151211600806735, + "learning_rate": 0.000999707959436049, + "loss": 3.4117, + "step": 89 + }, + { + "epoch": 0.0412796697626419, + "grad_norm": 1.020571135405344, + "learning_rate": 0.0009996820152934176, + "loss": 3.4397, + "step": 90 + }, + { + "epoch": 0.04173833276000459, + "grad_norm": 1.0971324081264977, + "learning_rate": 0.000999654967622631, + "loss": 3.3461, + "step": 91 + }, + { + "epoch": 0.042196995757367274, + "grad_norm": 1.1153787154579105, + "learning_rate": 0.0009996268164834238, + "loss": 3.3231, + "step": 92 + }, + { + "epoch": 0.042655658754729965, + "grad_norm": 1.3281450650125863, + "learning_rate": 0.000999597561937966, + "loss": 3.3664, + "step": 93 + }, + { + "epoch": 0.04311432175209265, + "grad_norm": 0.9945739168704084, + "learning_rate": 0.0009995672040508656, + "loss": 3.3906, + "step": 94 + }, + { + "epoch": 0.04357298474945534, + "grad_norm": 1.1804048546119599, + "learning_rate": 0.0009995357428891662, + "loss": 3.3899, + "step": 95 + }, + { + "epoch": 0.044031647746818024, + "grad_norm": 1.1716190453019866, + "learning_rate": 0.0009995031785223491, + "loss": 3.4427, + "step": 96 + }, + { + "epoch": 0.044490310744180715, + "grad_norm": 1.2623658606547155, + "learning_rate": 0.000999469511022331, + "loss": 3.4134, + "step": 97 + }, + { + "epoch": 0.0449489737415434, + "grad_norm": 0.9463546773684918, + "learning_rate": 0.0009994347404634657, + "loss": 3.3663, + "step": 98 + }, + { + "epoch": 0.04540763673890609, + "grad_norm": 1.0862877079302296, + "learning_rate": 0.0009993988669225423, + "loss": 3.3401, + "step": 99 + }, + { + "epoch": 0.045866299736268774, + "grad_norm": 1.4863267518724976, + "learning_rate": 0.000999361890478786, + "loss": 3.3735, + "step": 100 + }, + { + "epoch": 0.046324962733631465, + "grad_norm": 1.0704995719709178, + "learning_rate": 0.0009993238112138583, + "loss": 3.4312, + "step": 101 + }, + { + "epoch": 0.04678362573099415, + "grad_norm": 0.9788360713612421, + "learning_rate": 0.0009992846292118554, + "loss": 3.2405, + "step": 102 + }, + { + "epoch": 0.04724228872835684, + "grad_norm": 1.0760145634595113, + "learning_rate": 0.000999244344559309, + "loss": 3.3221, + "step": 103 + }, + { + "epoch": 0.047700951725719525, + "grad_norm": 1.014820492619447, + "learning_rate": 0.0009992029573451869, + "loss": 3.4179, + "step": 104 + }, + { + "epoch": 0.048159614723082216, + "grad_norm": 1.2647743328061363, + "learning_rate": 0.0009991604676608905, + "loss": 3.3121, + "step": 105 + }, + { + "epoch": 0.04861827772044491, + "grad_norm": 1.2261523560539749, + "learning_rate": 0.0009991168756002568, + "loss": 3.3619, + "step": 106 + }, + { + "epoch": 0.04907694071780759, + "grad_norm": 1.0341228002908578, + "learning_rate": 0.0009990721812595574, + "loss": 3.2522, + "step": 107 + }, + { + "epoch": 0.04953560371517028, + "grad_norm": 1.1901418767529102, + "learning_rate": 0.0009990263847374976, + "loss": 3.2563, + "step": 108 + }, + { + "epoch": 0.049994266712532966, + "grad_norm": 1.2815243335525268, + "learning_rate": 0.0009989794861352173, + "loss": 3.2976, + "step": 109 + }, + { + "epoch": 0.05045292970989566, + "grad_norm": 1.2101863533226296, + "learning_rate": 0.0009989314855562905, + "loss": 3.3684, + "step": 110 + }, + { + "epoch": 0.05091159270725834, + "grad_norm": 1.0834609907477937, + "learning_rate": 0.0009988823831067245, + "loss": 3.3341, + "step": 111 + }, + { + "epoch": 0.05137025570462103, + "grad_norm": 1.0691553423065325, + "learning_rate": 0.0009988321788949597, + "loss": 3.2692, + "step": 112 + }, + { + "epoch": 0.051828918701983716, + "grad_norm": 1.1866836745354197, + "learning_rate": 0.0009987808730318709, + "loss": 3.2566, + "step": 113 + }, + { + "epoch": 0.05228758169934641, + "grad_norm": 1.2340506456041191, + "learning_rate": 0.0009987284656307644, + "loss": 3.3568, + "step": 114 + }, + { + "epoch": 0.05274624469670909, + "grad_norm": 1.1590151381725209, + "learning_rate": 0.0009986749568073802, + "loss": 3.3033, + "step": 115 + }, + { + "epoch": 0.05320490769407178, + "grad_norm": 2.3041052650979723, + "learning_rate": 0.0009986203466798905, + "loss": 3.3595, + "step": 116 + }, + { + "epoch": 0.053663570691434466, + "grad_norm": 1.0440915938139463, + "learning_rate": 0.0009985646353688996, + "loss": 3.3448, + "step": 117 + }, + { + "epoch": 0.05412223368879716, + "grad_norm": 1.1335359402273164, + "learning_rate": 0.0009985078229974437, + "loss": 3.2632, + "step": 118 + }, + { + "epoch": 0.05458089668615984, + "grad_norm": 1.1968035263604115, + "learning_rate": 0.0009984499096909905, + "loss": 3.2989, + "step": 119 + }, + { + "epoch": 0.05503955968352253, + "grad_norm": 1.294936924983272, + "learning_rate": 0.0009983908955774397, + "loss": 3.2727, + "step": 120 + }, + { + "epoch": 0.05549822268088522, + "grad_norm": 1.1063504023991162, + "learning_rate": 0.0009983307807871211, + "loss": 3.2366, + "step": 121 + }, + { + "epoch": 0.05595688567824791, + "grad_norm": 1.1332076683700085, + "learning_rate": 0.0009982695654527965, + "loss": 3.2315, + "step": 122 + }, + { + "epoch": 0.0564155486756106, + "grad_norm": 1.132851759335386, + "learning_rate": 0.0009982072497096571, + "loss": 3.2301, + "step": 123 + }, + { + "epoch": 0.05687421167297328, + "grad_norm": 1.2268835882648246, + "learning_rate": 0.000998143833695325, + "loss": 3.2791, + "step": 124 + }, + { + "epoch": 0.05733287467033597, + "grad_norm": 1.269526164150147, + "learning_rate": 0.0009980793175498517, + "loss": 3.2742, + "step": 125 + }, + { + "epoch": 0.05779153766769866, + "grad_norm": 1.1407408890707014, + "learning_rate": 0.000998013701415719, + "loss": 3.2533, + "step": 126 + }, + { + "epoch": 0.05825020066506135, + "grad_norm": 1.128708337225489, + "learning_rate": 0.0009979469854378372, + "loss": 3.2007, + "step": 127 + }, + { + "epoch": 0.05870886366242403, + "grad_norm": 1.2304844649923454, + "learning_rate": 0.000997879169763546, + "loss": 3.1835, + "step": 128 + }, + { + "epoch": 0.05916752665978672, + "grad_norm": 1.1423866983933786, + "learning_rate": 0.000997810254542614, + "loss": 3.1712, + "step": 129 + }, + { + "epoch": 0.05962618965714941, + "grad_norm": 1.1483489382186338, + "learning_rate": 0.0009977402399272374, + "loss": 3.2221, + "step": 130 + }, + { + "epoch": 0.0600848526545121, + "grad_norm": 1.2378549659054756, + "learning_rate": 0.0009976691260720407, + "loss": 3.056, + "step": 131 + }, + { + "epoch": 0.06054351565187478, + "grad_norm": 1.376510325418578, + "learning_rate": 0.0009975969131340763, + "loss": 3.229, + "step": 132 + }, + { + "epoch": 0.06100217864923747, + "grad_norm": 1.4719169093488742, + "learning_rate": 0.0009975236012728236, + "loss": 3.2012, + "step": 133 + }, + { + "epoch": 0.06146084164660016, + "grad_norm": 1.2634729316273439, + "learning_rate": 0.0009974491906501886, + "loss": 3.1974, + "step": 134 + }, + { + "epoch": 0.06191950464396285, + "grad_norm": 1.3738887803535356, + "learning_rate": 0.0009973736814305049, + "loss": 3.2443, + "step": 135 + }, + { + "epoch": 0.06237816764132553, + "grad_norm": 1.2753325542885747, + "learning_rate": 0.0009972970737805312, + "loss": 3.2148, + "step": 136 + }, + { + "epoch": 0.06283683063868822, + "grad_norm": 1.1989231663495221, + "learning_rate": 0.0009972193678694525, + "loss": 3.1979, + "step": 137 + }, + { + "epoch": 0.06329549363605091, + "grad_norm": 1.4324150705123475, + "learning_rate": 0.0009971405638688794, + "loss": 3.1972, + "step": 138 + }, + { + "epoch": 0.0637541566334136, + "grad_norm": 1.1326866114427132, + "learning_rate": 0.0009970606619528475, + "loss": 3.0834, + "step": 139 + }, + { + "epoch": 0.06421281963077628, + "grad_norm": 1.2043592144794524, + "learning_rate": 0.000996979662297817, + "loss": 3.1695, + "step": 140 + }, + { + "epoch": 0.06467148262813897, + "grad_norm": 1.315289030124334, + "learning_rate": 0.0009968975650826721, + "loss": 3.1377, + "step": 141 + }, + { + "epoch": 0.06513014562550166, + "grad_norm": 1.2835565598032082, + "learning_rate": 0.000996814370488722, + "loss": 3.2131, + "step": 142 + }, + { + "epoch": 0.06558880862286436, + "grad_norm": 1.5832615677734347, + "learning_rate": 0.000996730078699698, + "loss": 3.1657, + "step": 143 + }, + { + "epoch": 0.06604747162022703, + "grad_norm": 1.3892307100385157, + "learning_rate": 0.0009966446899017558, + "loss": 3.1778, + "step": 144 + }, + { + "epoch": 0.06650613461758972, + "grad_norm": 1.2375533882739782, + "learning_rate": 0.0009965582042834728, + "loss": 3.1852, + "step": 145 + }, + { + "epoch": 0.06696479761495241, + "grad_norm": 1.1080988516354482, + "learning_rate": 0.0009964706220358492, + "loss": 3.1581, + "step": 146 + }, + { + "epoch": 0.0674234606123151, + "grad_norm": 1.2210369877301381, + "learning_rate": 0.000996381943352307, + "loss": 3.1295, + "step": 147 + }, + { + "epoch": 0.06788212360967778, + "grad_norm": 1.489638887554701, + "learning_rate": 0.0009962921684286896, + "loss": 3.1719, + "step": 148 + }, + { + "epoch": 0.06834078660704047, + "grad_norm": 1.4232506686234756, + "learning_rate": 0.0009962012974632614, + "loss": 3.1939, + "step": 149 + }, + { + "epoch": 0.06879944960440317, + "grad_norm": 1.4129539135000884, + "learning_rate": 0.0009961093306567075, + "loss": 3.1143, + "step": 150 + }, + { + "epoch": 0.06925811260176586, + "grad_norm": 1.4568286113831337, + "learning_rate": 0.0009960162682121328, + "loss": 3.1343, + "step": 151 + }, + { + "epoch": 0.06971677559912855, + "grad_norm": 1.4537384670430762, + "learning_rate": 0.0009959221103350623, + "loss": 3.1598, + "step": 152 + }, + { + "epoch": 0.07017543859649122, + "grad_norm": 1.3325812186016823, + "learning_rate": 0.0009958268572334394, + "loss": 3.001, + "step": 153 + }, + { + "epoch": 0.07063410159385392, + "grad_norm": 1.4044529617588124, + "learning_rate": 0.0009957305091176274, + "loss": 3.1231, + "step": 154 + }, + { + "epoch": 0.0710927645912166, + "grad_norm": 1.2881732357208018, + "learning_rate": 0.0009956330662004075, + "loss": 3.1165, + "step": 155 + }, + { + "epoch": 0.0715514275885793, + "grad_norm": 1.4156300135914575, + "learning_rate": 0.0009955345286969779, + "loss": 3.1105, + "step": 156 + }, + { + "epoch": 0.07201009058594197, + "grad_norm": 1.3029023432575337, + "learning_rate": 0.0009954348968249551, + "loss": 3.1178, + "step": 157 + }, + { + "epoch": 0.07246875358330467, + "grad_norm": 1.2023265483616365, + "learning_rate": 0.0009953341708043724, + "loss": 3.1191, + "step": 158 + }, + { + "epoch": 0.07292741658066736, + "grad_norm": 1.1253379349772592, + "learning_rate": 0.0009952323508576793, + "loss": 3.1327, + "step": 159 + }, + { + "epoch": 0.07338607957803005, + "grad_norm": 1.4366502252287159, + "learning_rate": 0.0009951294372097406, + "loss": 3.1246, + "step": 160 + }, + { + "epoch": 0.07384474257539272, + "grad_norm": 1.293772926564736, + "learning_rate": 0.0009950254300878378, + "loss": 3.0712, + "step": 161 + }, + { + "epoch": 0.07430340557275542, + "grad_norm": 1.1954499039157431, + "learning_rate": 0.000994920329721666, + "loss": 2.9986, + "step": 162 + }, + { + "epoch": 0.0747620685701181, + "grad_norm": 1.3070852604417027, + "learning_rate": 0.0009948141363433356, + "loss": 3.0911, + "step": 163 + }, + { + "epoch": 0.0752207315674808, + "grad_norm": 1.5231120908024403, + "learning_rate": 0.00099470685018737, + "loss": 3.0152, + "step": 164 + }, + { + "epoch": 0.07567939456484347, + "grad_norm": 1.4442950545304176, + "learning_rate": 0.0009945984714907073, + "loss": 3.0304, + "step": 165 + }, + { + "epoch": 0.07613805756220617, + "grad_norm": 1.2531437099080245, + "learning_rate": 0.000994489000492697, + "loss": 3.0709, + "step": 166 + }, + { + "epoch": 0.07659672055956886, + "grad_norm": 1.5305968985219325, + "learning_rate": 0.0009943784374351016, + "loss": 3.1352, + "step": 167 + }, + { + "epoch": 0.07705538355693155, + "grad_norm": 1.6898005079048353, + "learning_rate": 0.0009942667825620951, + "loss": 2.9832, + "step": 168 + }, + { + "epoch": 0.07751404655429424, + "grad_norm": 1.27040609672997, + "learning_rate": 0.0009941540361202634, + "loss": 2.8956, + "step": 169 + }, + { + "epoch": 0.07797270955165692, + "grad_norm": 1.2225302631090635, + "learning_rate": 0.0009940401983586022, + "loss": 3.0038, + "step": 170 + }, + { + "epoch": 0.0784313725490196, + "grad_norm": 1.579414089996605, + "learning_rate": 0.000993925269528518, + "loss": 3.1269, + "step": 171 + }, + { + "epoch": 0.0788900355463823, + "grad_norm": 1.2513157283744154, + "learning_rate": 0.0009938092498838265, + "loss": 3.0469, + "step": 172 + }, + { + "epoch": 0.07934869854374499, + "grad_norm": 1.2215269101530832, + "learning_rate": 0.0009936921396807524, + "loss": 3.0877, + "step": 173 + }, + { + "epoch": 0.07980736154110767, + "grad_norm": 1.3873786842220641, + "learning_rate": 0.0009935739391779292, + "loss": 2.9722, + "step": 174 + }, + { + "epoch": 0.08026602453847036, + "grad_norm": 1.305273444947428, + "learning_rate": 0.000993454648636398, + "loss": 2.9874, + "step": 175 + }, + { + "epoch": 0.08072468753583305, + "grad_norm": 1.1387260640139856, + "learning_rate": 0.0009933342683196074, + "loss": 2.985, + "step": 176 + }, + { + "epoch": 0.08118335053319574, + "grad_norm": 1.4237096739739086, + "learning_rate": 0.0009932127984934125, + "loss": 2.9921, + "step": 177 + }, + { + "epoch": 0.08164201353055842, + "grad_norm": 1.3297587217748534, + "learning_rate": 0.0009930902394260745, + "loss": 3.076, + "step": 178 + }, + { + "epoch": 0.08210067652792111, + "grad_norm": 1.205797663222874, + "learning_rate": 0.0009929665913882607, + "loss": 2.9667, + "step": 179 + }, + { + "epoch": 0.0825593395252838, + "grad_norm": 1.318246904114062, + "learning_rate": 0.0009928418546530425, + "loss": 2.9589, + "step": 180 + }, + { + "epoch": 0.08301800252264649, + "grad_norm": 1.3205544576956543, + "learning_rate": 0.0009927160294958964, + "loss": 3.0542, + "step": 181 + }, + { + "epoch": 0.08347666552000918, + "grad_norm": 1.274807968645388, + "learning_rate": 0.000992589116194702, + "loss": 3.015, + "step": 182 + }, + { + "epoch": 0.08393532851737186, + "grad_norm": 1.2913644430757565, + "learning_rate": 0.000992461115029743, + "loss": 2.9101, + "step": 183 + }, + { + "epoch": 0.08439399151473455, + "grad_norm": 1.2534829839486443, + "learning_rate": 0.000992332026283704, + "loss": 2.9789, + "step": 184 + }, + { + "epoch": 0.08485265451209724, + "grad_norm": 1.3133399551422689, + "learning_rate": 0.0009922018502416736, + "loss": 2.8793, + "step": 185 + }, + { + "epoch": 0.08531131750945993, + "grad_norm": 1.2144611029313201, + "learning_rate": 0.0009920705871911395, + "loss": 2.9244, + "step": 186 + }, + { + "epoch": 0.08576998050682261, + "grad_norm": 1.1894040045743424, + "learning_rate": 0.0009919382374219915, + "loss": 2.9329, + "step": 187 + }, + { + "epoch": 0.0862286435041853, + "grad_norm": 1.4620349051582728, + "learning_rate": 0.0009918048012265187, + "loss": 2.9189, + "step": 188 + }, + { + "epoch": 0.08668730650154799, + "grad_norm": 3.1732915237414825, + "learning_rate": 0.0009916702788994097, + "loss": 2.9729, + "step": 189 + }, + { + "epoch": 0.08714596949891068, + "grad_norm": 1.4125612022961562, + "learning_rate": 0.0009915346707377519, + "loss": 2.9626, + "step": 190 + }, + { + "epoch": 0.08760463249627336, + "grad_norm": 1.3704643270247066, + "learning_rate": 0.0009913979770410305, + "loss": 2.9888, + "step": 191 + }, + { + "epoch": 0.08806329549363605, + "grad_norm": 1.3462370775983368, + "learning_rate": 0.0009912601981111285, + "loss": 2.9918, + "step": 192 + }, + { + "epoch": 0.08852195849099874, + "grad_norm": 1.2029737160305511, + "learning_rate": 0.0009911213342523248, + "loss": 2.9966, + "step": 193 + }, + { + "epoch": 0.08898062148836143, + "grad_norm": 1.3892565127666059, + "learning_rate": 0.000990981385771295, + "loss": 2.9839, + "step": 194 + }, + { + "epoch": 0.08943928448572411, + "grad_norm": 1.126268042827079, + "learning_rate": 0.00099084035297711, + "loss": 2.9102, + "step": 195 + }, + { + "epoch": 0.0898979474830868, + "grad_norm": 1.238596510806253, + "learning_rate": 0.000990698236181235, + "loss": 2.9046, + "step": 196 + }, + { + "epoch": 0.09035661048044949, + "grad_norm": 1.14893720331777, + "learning_rate": 0.0009905550356975293, + "loss": 2.9227, + "step": 197 + }, + { + "epoch": 0.09081527347781218, + "grad_norm": 1.218593803888788, + "learning_rate": 0.0009904107518422457, + "loss": 2.9434, + "step": 198 + }, + { + "epoch": 0.09127393647517487, + "grad_norm": 1.1142594878888605, + "learning_rate": 0.0009902653849340295, + "loss": 2.9191, + "step": 199 + }, + { + "epoch": 0.09173259947253755, + "grad_norm": 1.1963000339379373, + "learning_rate": 0.0009901189352939177, + "loss": 2.9423, + "step": 200 + }, + { + "epoch": 0.09219126246990024, + "grad_norm": 1.1832940766994748, + "learning_rate": 0.0009899714032453387, + "loss": 2.8947, + "step": 201 + }, + { + "epoch": 0.09264992546726293, + "grad_norm": 1.3498887097127683, + "learning_rate": 0.000989822789114111, + "loss": 2.8502, + "step": 202 + }, + { + "epoch": 0.09310858846462562, + "grad_norm": 1.2149262036559643, + "learning_rate": 0.0009896730932284434, + "loss": 2.8951, + "step": 203 + }, + { + "epoch": 0.0935672514619883, + "grad_norm": 1.3529043077976657, + "learning_rate": 0.0009895223159189332, + "loss": 2.88, + "step": 204 + }, + { + "epoch": 0.09402591445935099, + "grad_norm": 1.39482293830198, + "learning_rate": 0.0009893704575185663, + "loss": 2.9637, + "step": 205 + }, + { + "epoch": 0.09448457745671368, + "grad_norm": 1.3251310251807606, + "learning_rate": 0.000989217518362716, + "loss": 2.9086, + "step": 206 + }, + { + "epoch": 0.09494324045407637, + "grad_norm": 1.3068977355053064, + "learning_rate": 0.0009890634987891425, + "loss": 2.9056, + "step": 207 + }, + { + "epoch": 0.09540190345143905, + "grad_norm": 1.552549719247358, + "learning_rate": 0.0009889083991379917, + "loss": 2.9351, + "step": 208 + }, + { + "epoch": 0.09586056644880174, + "grad_norm": 1.277604443248944, + "learning_rate": 0.0009887522197517954, + "loss": 2.9416, + "step": 209 + }, + { + "epoch": 0.09631922944616443, + "grad_norm": 1.8817768369327545, + "learning_rate": 0.0009885949609754693, + "loss": 3.0097, + "step": 210 + }, + { + "epoch": 0.09677789244352712, + "grad_norm": 1.3349838760684052, + "learning_rate": 0.000988436623156314, + "loss": 2.92, + "step": 211 + }, + { + "epoch": 0.09723655544088981, + "grad_norm": 1.4620729384720046, + "learning_rate": 0.0009882772066440114, + "loss": 2.7885, + "step": 212 + }, + { + "epoch": 0.09769521843825249, + "grad_norm": 1.3736505110856025, + "learning_rate": 0.0009881167117906276, + "loss": 2.6955, + "step": 213 + }, + { + "epoch": 0.09815388143561518, + "grad_norm": 1.1725731133731772, + "learning_rate": 0.0009879551389506084, + "loss": 2.9121, + "step": 214 + }, + { + "epoch": 0.09861254443297787, + "grad_norm": 1.2640817931990103, + "learning_rate": 0.0009877924884807814, + "loss": 2.9223, + "step": 215 + }, + { + "epoch": 0.09907120743034056, + "grad_norm": 1.6090567304725392, + "learning_rate": 0.000987628760740354, + "loss": 2.9248, + "step": 216 + }, + { + "epoch": 0.09952987042770324, + "grad_norm": 1.3589473461992507, + "learning_rate": 0.0009874639560909118, + "loss": 2.8959, + "step": 217 + }, + { + "epoch": 0.09998853342506593, + "grad_norm": 1.2626810321047346, + "learning_rate": 0.0009872980748964202, + "loss": 2.8273, + "step": 218 + }, + { + "epoch": 0.10044719642242862, + "grad_norm": 1.382641184247785, + "learning_rate": 0.000987131117523221, + "loss": 2.81, + "step": 219 + }, + { + "epoch": 0.10090585941979131, + "grad_norm": 1.5253422895587763, + "learning_rate": 0.000986963084340033, + "loss": 2.8615, + "step": 220 + }, + { + "epoch": 0.10136452241715399, + "grad_norm": 1.4149728239017028, + "learning_rate": 0.0009867939757179508, + "loss": 2.827, + "step": 221 + }, + { + "epoch": 0.10182318541451668, + "grad_norm": 1.5398471497181703, + "learning_rate": 0.0009866237920304443, + "loss": 2.8721, + "step": 222 + }, + { + "epoch": 0.10228184841187937, + "grad_norm": 1.1935283414477655, + "learning_rate": 0.0009864525336533577, + "loss": 2.8206, + "step": 223 + }, + { + "epoch": 0.10274051140924206, + "grad_norm": 1.3478197492483663, + "learning_rate": 0.000986280200964908, + "loss": 2.8302, + "step": 224 + }, + { + "epoch": 0.10319917440660474, + "grad_norm": 1.2155271434395551, + "learning_rate": 0.0009861067943456856, + "loss": 2.884, + "step": 225 + }, + { + "epoch": 0.10365783740396743, + "grad_norm": 1.1390471756669205, + "learning_rate": 0.000985932314178652, + "loss": 2.8122, + "step": 226 + }, + { + "epoch": 0.10411650040133012, + "grad_norm": 1.1166582880982483, + "learning_rate": 0.00098575676084914, + "loss": 2.746, + "step": 227 + }, + { + "epoch": 0.10457516339869281, + "grad_norm": 1.2907615562472738, + "learning_rate": 0.0009855801347448518, + "loss": 2.874, + "step": 228 + }, + { + "epoch": 0.1050338263960555, + "grad_norm": 1.3559916948459891, + "learning_rate": 0.0009854024362558596, + "loss": 2.7916, + "step": 229 + }, + { + "epoch": 0.10549248939341818, + "grad_norm": 1.3625473756492517, + "learning_rate": 0.0009852236657746035, + "loss": 2.8394, + "step": 230 + }, + { + "epoch": 0.10595115239078087, + "grad_norm": 1.5068759327868735, + "learning_rate": 0.0009850438236958911, + "loss": 2.7505, + "step": 231 + }, + { + "epoch": 0.10640981538814356, + "grad_norm": 1.4116567527364652, + "learning_rate": 0.0009848629104168966, + "loss": 2.8485, + "step": 232 + }, + { + "epoch": 0.10686847838550625, + "grad_norm": 1.507077759831815, + "learning_rate": 0.00098468092633716, + "loss": 2.8764, + "step": 233 + }, + { + "epoch": 0.10732714138286893, + "grad_norm": 1.534325509765213, + "learning_rate": 0.0009844978718585855, + "loss": 2.8178, + "step": 234 + }, + { + "epoch": 0.10778580438023162, + "grad_norm": 1.3721669485631427, + "learning_rate": 0.0009843137473854423, + "loss": 2.8072, + "step": 235 + }, + { + "epoch": 0.10824446737759431, + "grad_norm": 1.289360987382948, + "learning_rate": 0.000984128553324362, + "loss": 2.855, + "step": 236 + }, + { + "epoch": 0.108703130374957, + "grad_norm": 1.1205547238680778, + "learning_rate": 0.0009839422900843383, + "loss": 2.7709, + "step": 237 + }, + { + "epoch": 0.10916179337231968, + "grad_norm": 1.2907949066811188, + "learning_rate": 0.0009837549580767261, + "loss": 2.8838, + "step": 238 + }, + { + "epoch": 0.10962045636968237, + "grad_norm": 1.1659951901452819, + "learning_rate": 0.0009835665577152411, + "loss": 2.7029, + "step": 239 + }, + { + "epoch": 0.11007911936704506, + "grad_norm": 1.3740502759612314, + "learning_rate": 0.000983377089415958, + "loss": 2.8705, + "step": 240 + }, + { + "epoch": 0.11053778236440776, + "grad_norm": 1.3825884394555616, + "learning_rate": 0.0009831865535973102, + "loss": 2.8583, + "step": 241 + }, + { + "epoch": 0.11099644536177045, + "grad_norm": 1.3144901389413353, + "learning_rate": 0.0009829949506800885, + "loss": 2.7708, + "step": 242 + }, + { + "epoch": 0.11145510835913312, + "grad_norm": 1.249416224136499, + "learning_rate": 0.0009828022810874405, + "loss": 2.7613, + "step": 243 + }, + { + "epoch": 0.11191377135649581, + "grad_norm": 4.588102239358526, + "learning_rate": 0.0009826085452448693, + "loss": 2.6686, + "step": 244 + }, + { + "epoch": 0.1123724343538585, + "grad_norm": 1.237283720244033, + "learning_rate": 0.000982413743580233, + "loss": 2.7931, + "step": 245 + }, + { + "epoch": 0.1128310973512212, + "grad_norm": 1.7193016865583248, + "learning_rate": 0.0009822178765237436, + "loss": 2.7464, + "step": 246 + }, + { + "epoch": 0.11328976034858387, + "grad_norm": 1.4758152973506948, + "learning_rate": 0.0009820209445079654, + "loss": 2.8202, + "step": 247 + }, + { + "epoch": 0.11374842334594656, + "grad_norm": 1.5668356605547336, + "learning_rate": 0.0009818229479678158, + "loss": 2.7312, + "step": 248 + }, + { + "epoch": 0.11420708634330926, + "grad_norm": 2.221352936583685, + "learning_rate": 0.0009816238873405615, + "loss": 2.7042, + "step": 249 + }, + { + "epoch": 0.11466574934067195, + "grad_norm": 1.278424153701712, + "learning_rate": 0.0009814237630658207, + "loss": 2.782, + "step": 250 + }, + { + "epoch": 0.11512441233803462, + "grad_norm": 1.3319647123731495, + "learning_rate": 0.00098122257558556, + "loss": 2.8338, + "step": 251 + }, + { + "epoch": 0.11558307533539731, + "grad_norm": 1.1759158169897217, + "learning_rate": 0.0009810203253440937, + "loss": 2.7809, + "step": 252 + }, + { + "epoch": 0.11604173833276, + "grad_norm": 1.343363223389528, + "learning_rate": 0.0009808170127880837, + "loss": 2.8346, + "step": 253 + }, + { + "epoch": 0.1165004013301227, + "grad_norm": 1.3833401516494517, + "learning_rate": 0.000980612638366538, + "loss": 2.8139, + "step": 254 + }, + { + "epoch": 0.11695906432748537, + "grad_norm": 1.2771729869732147, + "learning_rate": 0.0009804072025308096, + "loss": 2.6335, + "step": 255 + }, + { + "epoch": 0.11741772732484806, + "grad_norm": 2.4242683037339594, + "learning_rate": 0.000980200705734595, + "loss": 2.6936, + "step": 256 + }, + { + "epoch": 0.11787639032221076, + "grad_norm": 1.1308115803495011, + "learning_rate": 0.0009799931484339344, + "loss": 2.7117, + "step": 257 + }, + { + "epoch": 0.11833505331957345, + "grad_norm": 1.1623859793207256, + "learning_rate": 0.0009797845310872103, + "loss": 2.7103, + "step": 258 + }, + { + "epoch": 0.11879371631693614, + "grad_norm": 1.3875105702865704, + "learning_rate": 0.0009795748541551457, + "loss": 2.6966, + "step": 259 + }, + { + "epoch": 0.11925237931429881, + "grad_norm": 1.2176879708570025, + "learning_rate": 0.000979364118100804, + "loss": 2.7589, + "step": 260 + }, + { + "epoch": 0.1197110423116615, + "grad_norm": 1.4220433892728062, + "learning_rate": 0.0009791523233895875, + "loss": 2.8413, + "step": 261 + }, + { + "epoch": 0.1201697053090242, + "grad_norm": 1.222563169457215, + "learning_rate": 0.0009789394704892364, + "loss": 2.7007, + "step": 262 + }, + { + "epoch": 0.12062836830638689, + "grad_norm": 1.393227069420099, + "learning_rate": 0.0009787255598698282, + "loss": 2.7053, + "step": 263 + }, + { + "epoch": 0.12108703130374956, + "grad_norm": 1.5665152508376026, + "learning_rate": 0.0009785105920037758, + "loss": 2.7599, + "step": 264 + }, + { + "epoch": 0.12154569430111226, + "grad_norm": 1.3504398163112887, + "learning_rate": 0.0009782945673658275, + "loss": 2.697, + "step": 265 + }, + { + "epoch": 0.12200435729847495, + "grad_norm": 1.273852510148791, + "learning_rate": 0.0009780774864330654, + "loss": 2.6829, + "step": 266 + }, + { + "epoch": 0.12246302029583764, + "grad_norm": 1.4680719838324223, + "learning_rate": 0.000977859349684904, + "loss": 2.6624, + "step": 267 + }, + { + "epoch": 0.12292168329320032, + "grad_norm": 1.193573645925073, + "learning_rate": 0.00097764015760309, + "loss": 2.6864, + "step": 268 + }, + { + "epoch": 0.123380346290563, + "grad_norm": 1.2401477677882684, + "learning_rate": 0.0009774199106717004, + "loss": 2.6902, + "step": 269 + }, + { + "epoch": 0.1238390092879257, + "grad_norm": 1.256844333183089, + "learning_rate": 0.0009771986093771417, + "loss": 2.6357, + "step": 270 + }, + { + "epoch": 0.12429767228528839, + "grad_norm": 1.2210721654256746, + "learning_rate": 0.0009769762542081496, + "loss": 2.5844, + "step": 271 + }, + { + "epoch": 0.12475633528265107, + "grad_norm": 1.270762193996546, + "learning_rate": 0.000976752845655786, + "loss": 2.6786, + "step": 272 + }, + { + "epoch": 0.12521499828001376, + "grad_norm": 1.429411394245012, + "learning_rate": 0.0009765283842134411, + "loss": 2.6257, + "step": 273 + }, + { + "epoch": 0.12567366127737645, + "grad_norm": 1.1244969053505134, + "learning_rate": 0.0009763028703768282, + "loss": 2.681, + "step": 274 + }, + { + "epoch": 0.12613232427473914, + "grad_norm": 1.1562445746935874, + "learning_rate": 0.0009760763046439862, + "loss": 2.6466, + "step": 275 + }, + { + "epoch": 0.12659098727210183, + "grad_norm": 1.1556887144179153, + "learning_rate": 0.0009758486875152766, + "loss": 2.7035, + "step": 276 + }, + { + "epoch": 0.12704965026946452, + "grad_norm": 1.0416389198122553, + "learning_rate": 0.0009756200194933829, + "loss": 2.6587, + "step": 277 + }, + { + "epoch": 0.1275083132668272, + "grad_norm": 1.2049003378564036, + "learning_rate": 0.0009753903010833094, + "loss": 2.6367, + "step": 278 + }, + { + "epoch": 0.12796697626418987, + "grad_norm": 1.2382822834639315, + "learning_rate": 0.0009751595327923803, + "loss": 2.6684, + "step": 279 + }, + { + "epoch": 0.12842563926155257, + "grad_norm": 1.553934452274605, + "learning_rate": 0.0009749277151302382, + "loss": 2.7023, + "step": 280 + }, + { + "epoch": 0.12888430225891526, + "grad_norm": 1.282074574919836, + "learning_rate": 0.0009746948486088435, + "loss": 2.6766, + "step": 281 + }, + { + "epoch": 0.12934296525627795, + "grad_norm": 1.1895058361813624, + "learning_rate": 0.0009744609337424727, + "loss": 2.6605, + "step": 282 + }, + { + "epoch": 0.12980162825364064, + "grad_norm": 1.121821151199899, + "learning_rate": 0.0009742259710477177, + "loss": 2.6372, + "step": 283 + }, + { + "epoch": 0.13026029125100333, + "grad_norm": 1.3587658700804188, + "learning_rate": 0.0009739899610434841, + "loss": 2.7462, + "step": 284 + }, + { + "epoch": 0.13071895424836602, + "grad_norm": 1.1732742461104708, + "learning_rate": 0.0009737529042509913, + "loss": 2.5889, + "step": 285 + }, + { + "epoch": 0.1311776172457287, + "grad_norm": 1.358768868932136, + "learning_rate": 0.0009735148011937693, + "loss": 2.5982, + "step": 286 + }, + { + "epoch": 0.13163628024309137, + "grad_norm": 1.7421617035920605, + "learning_rate": 0.00097327565239766, + "loss": 2.7005, + "step": 287 + }, + { + "epoch": 0.13209494324045407, + "grad_norm": 1.2147603792992787, + "learning_rate": 0.0009730354583908136, + "loss": 2.672, + "step": 288 + }, + { + "epoch": 0.13255360623781676, + "grad_norm": 1.1599289977145477, + "learning_rate": 0.0009727942197036895, + "loss": 2.6634, + "step": 289 + }, + { + "epoch": 0.13301226923517945, + "grad_norm": 1.360030324973643, + "learning_rate": 0.0009725519368690539, + "loss": 2.6829, + "step": 290 + }, + { + "epoch": 0.13347093223254214, + "grad_norm": 1.1314056169019695, + "learning_rate": 0.0009723086104219787, + "loss": 2.5916, + "step": 291 + }, + { + "epoch": 0.13392959522990483, + "grad_norm": 1.1044064653291132, + "learning_rate": 0.0009720642408998409, + "loss": 2.6166, + "step": 292 + }, + { + "epoch": 0.13438825822726752, + "grad_norm": 1.2853812061271286, + "learning_rate": 0.0009718188288423211, + "loss": 2.5614, + "step": 293 + }, + { + "epoch": 0.1348469212246302, + "grad_norm": 1.2484568184144644, + "learning_rate": 0.0009715723747914022, + "loss": 2.6151, + "step": 294 + }, + { + "epoch": 0.1353055842219929, + "grad_norm": 1.2736810542440145, + "learning_rate": 0.0009713248792913685, + "loss": 2.6385, + "step": 295 + }, + { + "epoch": 0.13576424721935557, + "grad_norm": 1.0951403577380991, + "learning_rate": 0.0009710763428888037, + "loss": 2.6411, + "step": 296 + }, + { + "epoch": 0.13622291021671826, + "grad_norm": 1.098820621194863, + "learning_rate": 0.0009708267661325909, + "loss": 2.7025, + "step": 297 + }, + { + "epoch": 0.13668157321408095, + "grad_norm": 1.589027513426579, + "learning_rate": 0.0009705761495739107, + "loss": 2.6808, + "step": 298 + }, + { + "epoch": 0.13714023621144364, + "grad_norm": 1.1469028252897138, + "learning_rate": 0.0009703244937662399, + "loss": 2.5489, + "step": 299 + }, + { + "epoch": 0.13759889920880633, + "grad_norm": 1.229141427683149, + "learning_rate": 0.0009700717992653505, + "loss": 2.6271, + "step": 300 + }, + { + "epoch": 0.13805756220616902, + "grad_norm": 1.1903757059303692, + "learning_rate": 0.0009698180666293083, + "loss": 2.6505, + "step": 301 + }, + { + "epoch": 0.1385162252035317, + "grad_norm": 1.1594671530530236, + "learning_rate": 0.000969563296418472, + "loss": 2.6294, + "step": 302 + }, + { + "epoch": 0.1389748882008944, + "grad_norm": 1.0630694448989244, + "learning_rate": 0.0009693074891954914, + "loss": 2.6271, + "step": 303 + }, + { + "epoch": 0.1394335511982571, + "grad_norm": 1.0383651220381203, + "learning_rate": 0.0009690506455253072, + "loss": 2.6129, + "step": 304 + }, + { + "epoch": 0.13989221419561976, + "grad_norm": 1.322593003877876, + "learning_rate": 0.0009687927659751481, + "loss": 2.5421, + "step": 305 + }, + { + "epoch": 0.14035087719298245, + "grad_norm": 1.1895415013202513, + "learning_rate": 0.0009685338511145312, + "loss": 2.57, + "step": 306 + }, + { + "epoch": 0.14080954019034514, + "grad_norm": 1.2378310828556829, + "learning_rate": 0.0009682739015152598, + "loss": 2.6192, + "step": 307 + }, + { + "epoch": 0.14126820318770783, + "grad_norm": 1.259977080179307, + "learning_rate": 0.0009680129177514226, + "loss": 2.6032, + "step": 308 + }, + { + "epoch": 0.14172686618507052, + "grad_norm": 1.20561826879324, + "learning_rate": 0.0009677509003993915, + "loss": 2.6077, + "step": 309 + }, + { + "epoch": 0.1421855291824332, + "grad_norm": 1.1955572583600662, + "learning_rate": 0.0009674878500378221, + "loss": 2.6491, + "step": 310 + }, + { + "epoch": 0.1426441921797959, + "grad_norm": 1.1965361860727692, + "learning_rate": 0.0009672237672476505, + "loss": 2.5292, + "step": 311 + }, + { + "epoch": 0.1431028551771586, + "grad_norm": 1.3073407653426001, + "learning_rate": 0.0009669586526120935, + "loss": 2.5455, + "step": 312 + }, + { + "epoch": 0.14356151817452126, + "grad_norm": 1.206974848102603, + "learning_rate": 0.0009666925067166459, + "loss": 2.6697, + "step": 313 + }, + { + "epoch": 0.14402018117188395, + "grad_norm": 1.2435093905381118, + "learning_rate": 0.000966425330149081, + "loss": 2.6148, + "step": 314 + }, + { + "epoch": 0.14447884416924664, + "grad_norm": 1.0611228356585451, + "learning_rate": 0.0009661571234994475, + "loss": 2.537, + "step": 315 + }, + { + "epoch": 0.14493750716660933, + "grad_norm": 1.1808305842149853, + "learning_rate": 0.0009658878873600691, + "loss": 2.5628, + "step": 316 + }, + { + "epoch": 0.14539617016397202, + "grad_norm": 1.3385419730795647, + "learning_rate": 0.0009656176223255438, + "loss": 2.6248, + "step": 317 + }, + { + "epoch": 0.1458548331613347, + "grad_norm": 1.2795718978898984, + "learning_rate": 0.000965346328992741, + "loss": 2.5798, + "step": 318 + }, + { + "epoch": 0.1463134961586974, + "grad_norm": 1.4998067462282485, + "learning_rate": 0.0009650740079608014, + "loss": 2.6209, + "step": 319 + }, + { + "epoch": 0.1467721591560601, + "grad_norm": 1.1031907174734614, + "learning_rate": 0.0009648006598311353, + "loss": 2.5969, + "step": 320 + }, + { + "epoch": 0.14723082215342279, + "grad_norm": 1.3422822066934004, + "learning_rate": 0.0009645262852074214, + "loss": 2.5787, + "step": 321 + }, + { + "epoch": 0.14768948515078545, + "grad_norm": 1.0849384030401727, + "learning_rate": 0.0009642508846956053, + "loss": 2.5724, + "step": 322 + }, + { + "epoch": 0.14814814814814814, + "grad_norm": 1.1576680384228046, + "learning_rate": 0.0009639744589038983, + "loss": 2.5663, + "step": 323 + }, + { + "epoch": 0.14860681114551083, + "grad_norm": 1.1567584680295586, + "learning_rate": 0.0009636970084427759, + "loss": 2.6386, + "step": 324 + }, + { + "epoch": 0.14906547414287352, + "grad_norm": 1.0885604528084332, + "learning_rate": 0.0009634185339249766, + "loss": 2.6171, + "step": 325 + }, + { + "epoch": 0.1495241371402362, + "grad_norm": 1.1261465131695667, + "learning_rate": 0.0009631390359655003, + "loss": 2.6327, + "step": 326 + }, + { + "epoch": 0.1499828001375989, + "grad_norm": 1.0325086894143858, + "learning_rate": 0.0009628585151816074, + "loss": 2.4973, + "step": 327 + }, + { + "epoch": 0.1504414631349616, + "grad_norm": 1.0085219816782793, + "learning_rate": 0.0009625769721928172, + "loss": 2.6054, + "step": 328 + }, + { + "epoch": 0.15090012613232429, + "grad_norm": 1.0769369313334238, + "learning_rate": 0.0009622944076209061, + "loss": 2.5287, + "step": 329 + }, + { + "epoch": 0.15135878912968695, + "grad_norm": 1.0797179676244955, + "learning_rate": 0.0009620108220899071, + "loss": 2.5519, + "step": 330 + }, + { + "epoch": 0.15181745212704964, + "grad_norm": 1.0906265304752472, + "learning_rate": 0.0009617262162261075, + "loss": 2.6619, + "step": 331 + }, + { + "epoch": 0.15227611512441233, + "grad_norm": 1.0925963823833142, + "learning_rate": 0.0009614405906580486, + "loss": 2.5735, + "step": 332 + }, + { + "epoch": 0.15273477812177502, + "grad_norm": 1.2129383689687185, + "learning_rate": 0.000961153946016523, + "loss": 2.5919, + "step": 333 + }, + { + "epoch": 0.1531934411191377, + "grad_norm": 1.08238677052952, + "learning_rate": 0.000960866282934574, + "loss": 2.5083, + "step": 334 + }, + { + "epoch": 0.1536521041165004, + "grad_norm": 2.302980225133164, + "learning_rate": 0.0009605776020474945, + "loss": 2.5524, + "step": 335 + }, + { + "epoch": 0.1541107671138631, + "grad_norm": 1.0781315479121656, + "learning_rate": 0.0009602879039928249, + "loss": 2.5751, + "step": 336 + }, + { + "epoch": 0.1545694301112258, + "grad_norm": 1.1869790680930872, + "learning_rate": 0.0009599971894103521, + "loss": 2.6163, + "step": 337 + }, + { + "epoch": 0.15502809310858848, + "grad_norm": 1.1151191844134312, + "learning_rate": 0.0009597054589421077, + "loss": 2.5567, + "step": 338 + }, + { + "epoch": 0.15548675610595114, + "grad_norm": 1.0115114690836386, + "learning_rate": 0.0009594127132323669, + "loss": 2.5388, + "step": 339 + }, + { + "epoch": 0.15594541910331383, + "grad_norm": 1.0380949591757096, + "learning_rate": 0.0009591189529276474, + "loss": 2.6238, + "step": 340 + }, + { + "epoch": 0.15640408210067652, + "grad_norm": 1.0837456628657418, + "learning_rate": 0.0009588241786767072, + "loss": 2.4882, + "step": 341 + }, + { + "epoch": 0.1568627450980392, + "grad_norm": 1.0008573722719534, + "learning_rate": 0.0009585283911305436, + "loss": 2.4055, + "step": 342 + }, + { + "epoch": 0.1573214080954019, + "grad_norm": 1.0376388785713508, + "learning_rate": 0.000958231590942392, + "loss": 2.5337, + "step": 343 + }, + { + "epoch": 0.1577800710927646, + "grad_norm": 0.9777129424313156, + "learning_rate": 0.0009579337787677238, + "loss": 2.584, + "step": 344 + }, + { + "epoch": 0.1582387340901273, + "grad_norm": 1.0536488310306682, + "learning_rate": 0.0009576349552642456, + "loss": 2.6786, + "step": 345 + }, + { + "epoch": 0.15869739708748998, + "grad_norm": 1.025202677325547, + "learning_rate": 0.0009573351210918975, + "loss": 2.4762, + "step": 346 + }, + { + "epoch": 0.15915606008485264, + "grad_norm": 0.9371174318821094, + "learning_rate": 0.0009570342769128514, + "loss": 2.6013, + "step": 347 + }, + { + "epoch": 0.15961472308221533, + "grad_norm": 1.2445708544980074, + "learning_rate": 0.0009567324233915099, + "loss": 2.5369, + "step": 348 + }, + { + "epoch": 0.16007338607957802, + "grad_norm": 1.1627976695286435, + "learning_rate": 0.0009564295611945047, + "loss": 2.5946, + "step": 349 + }, + { + "epoch": 0.1605320490769407, + "grad_norm": 1.7727617593181249, + "learning_rate": 0.000956125690990695, + "loss": 2.5377, + "step": 350 + }, + { + "epoch": 0.1609907120743034, + "grad_norm": 0.9997440942827818, + "learning_rate": 0.0009558208134511665, + "loss": 2.6782, + "step": 351 + }, + { + "epoch": 0.1614493750716661, + "grad_norm": 1.2158839092141658, + "learning_rate": 0.0009555149292492289, + "loss": 2.5184, + "step": 352 + }, + { + "epoch": 0.1619080380690288, + "grad_norm": 1.0114879369642145, + "learning_rate": 0.0009552080390604159, + "loss": 2.4894, + "step": 353 + }, + { + "epoch": 0.16236670106639148, + "grad_norm": 1.6062929938794395, + "learning_rate": 0.0009549001435624823, + "loss": 2.5933, + "step": 354 + }, + { + "epoch": 0.16282536406375417, + "grad_norm": 1.0074903395395398, + "learning_rate": 0.0009545912434354029, + "loss": 2.5704, + "step": 355 + }, + { + "epoch": 0.16328402706111683, + "grad_norm": 1.3638484126823418, + "learning_rate": 0.0009542813393613721, + "loss": 2.624, + "step": 356 + }, + { + "epoch": 0.16374269005847952, + "grad_norm": 1.0178760484563565, + "learning_rate": 0.0009539704320248006, + "loss": 2.5371, + "step": 357 + }, + { + "epoch": 0.16420135305584221, + "grad_norm": 1.0773786204207696, + "learning_rate": 0.0009536585221123151, + "loss": 2.3874, + "step": 358 + }, + { + "epoch": 0.1646600160532049, + "grad_norm": 1.0838743280407517, + "learning_rate": 0.0009533456103127565, + "loss": 2.669, + "step": 359 + }, + { + "epoch": 0.1651186790505676, + "grad_norm": 0.9713065533437762, + "learning_rate": 0.000953031697317178, + "loss": 2.5264, + "step": 360 + }, + { + "epoch": 0.1655773420479303, + "grad_norm": 0.9158491806445213, + "learning_rate": 0.0009527167838188445, + "loss": 2.5253, + "step": 361 + }, + { + "epoch": 0.16603600504529298, + "grad_norm": 0.8376861896446467, + "learning_rate": 0.0009524008705132299, + "loss": 2.5394, + "step": 362 + }, + { + "epoch": 0.16649466804265567, + "grad_norm": 1.0113226520940304, + "learning_rate": 0.0009520839580980166, + "loss": 2.5212, + "step": 363 + }, + { + "epoch": 0.16695333104001836, + "grad_norm": 1.8215877549224508, + "learning_rate": 0.0009517660472730929, + "loss": 2.4477, + "step": 364 + }, + { + "epoch": 0.16741199403738102, + "grad_norm": 1.0014538494411909, + "learning_rate": 0.0009514471387405526, + "loss": 2.5308, + "step": 365 + }, + { + "epoch": 0.16787065703474371, + "grad_norm": 1.1383961813992591, + "learning_rate": 0.0009511272332046926, + "loss": 2.5469, + "step": 366 + }, + { + "epoch": 0.1683293200321064, + "grad_norm": 0.9788853720784348, + "learning_rate": 0.0009508063313720119, + "loss": 2.5462, + "step": 367 + }, + { + "epoch": 0.1687879830294691, + "grad_norm": 1.0695667109387954, + "learning_rate": 0.0009504844339512095, + "loss": 2.5239, + "step": 368 + }, + { + "epoch": 0.1692466460268318, + "grad_norm": 1.1053514508538156, + "learning_rate": 0.0009501615416531835, + "loss": 2.6246, + "step": 369 + }, + { + "epoch": 0.16970530902419448, + "grad_norm": 0.9768110293942509, + "learning_rate": 0.0009498376551910285, + "loss": 2.5629, + "step": 370 + }, + { + "epoch": 0.17016397202155717, + "grad_norm": 0.9516048168591165, + "learning_rate": 0.0009495127752800352, + "loss": 2.5317, + "step": 371 + }, + { + "epoch": 0.17062263501891986, + "grad_norm": 0.9491650336632849, + "learning_rate": 0.0009491869026376882, + "loss": 2.5822, + "step": 372 + }, + { + "epoch": 0.17108129801628252, + "grad_norm": 0.9683389141034281, + "learning_rate": 0.0009488600379836648, + "loss": 2.5099, + "step": 373 + }, + { + "epoch": 0.17153996101364521, + "grad_norm": 1.1314976433340618, + "learning_rate": 0.0009485321820398321, + "loss": 2.4701, + "step": 374 + }, + { + "epoch": 0.1719986240110079, + "grad_norm": 1.1029435992544767, + "learning_rate": 0.0009482033355302475, + "loss": 2.5979, + "step": 375 + }, + { + "epoch": 0.1724572870083706, + "grad_norm": 1.0906919791864598, + "learning_rate": 0.0009478734991811556, + "loss": 2.5316, + "step": 376 + }, + { + "epoch": 0.1729159500057333, + "grad_norm": 1.0261780042769484, + "learning_rate": 0.0009475426737209871, + "loss": 2.6562, + "step": 377 + }, + { + "epoch": 0.17337461300309598, + "grad_norm": 2.203631712237197, + "learning_rate": 0.000947210859880357, + "loss": 2.5532, + "step": 378 + }, + { + "epoch": 0.17383327600045867, + "grad_norm": 1.2873435442610301, + "learning_rate": 0.0009468780583920631, + "loss": 2.5098, + "step": 379 + }, + { + "epoch": 0.17429193899782136, + "grad_norm": 1.0575479113304005, + "learning_rate": 0.0009465442699910846, + "loss": 2.4606, + "step": 380 + }, + { + "epoch": 0.17475060199518405, + "grad_norm": 1.16124564808449, + "learning_rate": 0.0009462094954145801, + "loss": 2.662, + "step": 381 + }, + { + "epoch": 0.17520926499254671, + "grad_norm": 0.9392704520423946, + "learning_rate": 0.0009458737354018859, + "loss": 2.5174, + "step": 382 + }, + { + "epoch": 0.1756679279899094, + "grad_norm": 2.0414108719573085, + "learning_rate": 0.000945536990694515, + "loss": 2.5639, + "step": 383 + }, + { + "epoch": 0.1761265909872721, + "grad_norm": 0.9755552570693197, + "learning_rate": 0.0009451992620361551, + "loss": 2.6598, + "step": 384 + }, + { + "epoch": 0.1765852539846348, + "grad_norm": 1.0032580519402974, + "learning_rate": 0.0009448605501726664, + "loss": 2.4604, + "step": 385 + }, + { + "epoch": 0.17704391698199748, + "grad_norm": 0.9889207219742117, + "learning_rate": 0.000944520855852081, + "loss": 2.5818, + "step": 386 + }, + { + "epoch": 0.17750257997936017, + "grad_norm": 1.256839575049198, + "learning_rate": 0.0009441801798246002, + "loss": 2.4989, + "step": 387 + }, + { + "epoch": 0.17796124297672286, + "grad_norm": 0.9511173212957855, + "learning_rate": 0.0009438385228425939, + "loss": 2.6057, + "step": 388 + }, + { + "epoch": 0.17841990597408555, + "grad_norm": 1.292792853249691, + "learning_rate": 0.0009434958856605982, + "loss": 2.5934, + "step": 389 + }, + { + "epoch": 0.17887856897144822, + "grad_norm": 0.9813791559879659, + "learning_rate": 0.0009431522690353137, + "loss": 2.5679, + "step": 390 + }, + { + "epoch": 0.1793372319688109, + "grad_norm": 1.0201188244910138, + "learning_rate": 0.0009428076737256044, + "loss": 2.5297, + "step": 391 + }, + { + "epoch": 0.1797958949661736, + "grad_norm": 1.028407794007259, + "learning_rate": 0.0009424621004924954, + "loss": 2.548, + "step": 392 + }, + { + "epoch": 0.1802545579635363, + "grad_norm": 1.272875828087249, + "learning_rate": 0.0009421155500991719, + "loss": 2.4684, + "step": 393 + }, + { + "epoch": 0.18071322096089898, + "grad_norm": 1.169946487016137, + "learning_rate": 0.0009417680233109767, + "loss": 2.6603, + "step": 394 + }, + { + "epoch": 0.18117188395826167, + "grad_norm": 1.0334175872954354, + "learning_rate": 0.000941419520895409, + "loss": 2.4092, + "step": 395 + }, + { + "epoch": 0.18163054695562436, + "grad_norm": 0.9606729614024848, + "learning_rate": 0.0009410700436221229, + "loss": 2.5348, + "step": 396 + }, + { + "epoch": 0.18208920995298705, + "grad_norm": 1.0901764082598115, + "learning_rate": 0.0009407195922629252, + "loss": 2.4603, + "step": 397 + }, + { + "epoch": 0.18254787295034974, + "grad_norm": 0.8437685760924681, + "learning_rate": 0.000940368167591774, + "loss": 2.3915, + "step": 398 + }, + { + "epoch": 0.1830065359477124, + "grad_norm": 0.9186364242258162, + "learning_rate": 0.0009400157703847769, + "loss": 2.5076, + "step": 399 + }, + { + "epoch": 0.1834651989450751, + "grad_norm": 1.002904768691321, + "learning_rate": 0.0009396624014201895, + "loss": 2.4806, + "step": 400 + }, + { + "epoch": 0.1839238619424378, + "grad_norm": 1.0678062663695496, + "learning_rate": 0.000939308061478413, + "loss": 2.5523, + "step": 401 + }, + { + "epoch": 0.18438252493980048, + "grad_norm": 0.9406510463042821, + "learning_rate": 0.0009389527513419935, + "loss": 2.5146, + "step": 402 + }, + { + "epoch": 0.18484118793716317, + "grad_norm": 0.9707168249796773, + "learning_rate": 0.0009385964717956195, + "loss": 2.492, + "step": 403 + }, + { + "epoch": 0.18529985093452586, + "grad_norm": 0.9661018444205177, + "learning_rate": 0.0009382392236261201, + "loss": 2.4687, + "step": 404 + }, + { + "epoch": 0.18575851393188855, + "grad_norm": 0.9176862948144718, + "learning_rate": 0.0009378810076224644, + "loss": 2.5263, + "step": 405 + }, + { + "epoch": 0.18621717692925124, + "grad_norm": 0.8970092076803013, + "learning_rate": 0.0009375218245757582, + "loss": 2.5145, + "step": 406 + }, + { + "epoch": 0.1866758399266139, + "grad_norm": 1.024208819872216, + "learning_rate": 0.0009371616752792432, + "loss": 2.5269, + "step": 407 + }, + { + "epoch": 0.1871345029239766, + "grad_norm": 0.9989004777404514, + "learning_rate": 0.0009368005605282949, + "loss": 2.4553, + "step": 408 + }, + { + "epoch": 0.1875931659213393, + "grad_norm": 0.9877464206673017, + "learning_rate": 0.0009364384811204212, + "loss": 2.5439, + "step": 409 + }, + { + "epoch": 0.18805182891870198, + "grad_norm": 0.9337779896221502, + "learning_rate": 0.00093607543785526, + "loss": 2.5029, + "step": 410 + }, + { + "epoch": 0.18851049191606467, + "grad_norm": 1.0891732504066467, + "learning_rate": 0.0009357114315345787, + "loss": 2.5158, + "step": 411 + }, + { + "epoch": 0.18896915491342736, + "grad_norm": 1.5198073228947546, + "learning_rate": 0.0009353464629622705, + "loss": 2.5018, + "step": 412 + }, + { + "epoch": 0.18942781791079005, + "grad_norm": 0.9148557194571809, + "learning_rate": 0.0009349805329443544, + "loss": 2.4836, + "step": 413 + }, + { + "epoch": 0.18988648090815274, + "grad_norm": 0.9477982432142353, + "learning_rate": 0.0009346136422889724, + "loss": 2.4882, + "step": 414 + }, + { + "epoch": 0.19034514390551543, + "grad_norm": 1.0493710522847037, + "learning_rate": 0.0009342457918063882, + "loss": 2.5235, + "step": 415 + }, + { + "epoch": 0.1908038069028781, + "grad_norm": 0.88642690646891, + "learning_rate": 0.0009338769823089853, + "loss": 2.4657, + "step": 416 + }, + { + "epoch": 0.1912624699002408, + "grad_norm": 0.932382340814617, + "learning_rate": 0.0009335072146112648, + "loss": 2.485, + "step": 417 + }, + { + "epoch": 0.19172113289760348, + "grad_norm": 0.9771570328959786, + "learning_rate": 0.0009331364895298444, + "loss": 2.5672, + "step": 418 + }, + { + "epoch": 0.19217979589496617, + "grad_norm": 0.936131582892936, + "learning_rate": 0.0009327648078834559, + "loss": 2.4032, + "step": 419 + }, + { + "epoch": 0.19263845889232886, + "grad_norm": 0.8768545933267284, + "learning_rate": 0.0009323921704929434, + "loss": 2.575, + "step": 420 + }, + { + "epoch": 0.19309712188969155, + "grad_norm": 0.9277259078305828, + "learning_rate": 0.0009320185781812623, + "loss": 2.4949, + "step": 421 + }, + { + "epoch": 0.19355578488705424, + "grad_norm": 0.9329058621473814, + "learning_rate": 0.0009316440317734762, + "loss": 2.42, + "step": 422 + }, + { + "epoch": 0.19401444788441694, + "grad_norm": 0.9634063434955307, + "learning_rate": 0.0009312685320967565, + "loss": 2.4583, + "step": 423 + }, + { + "epoch": 0.19447311088177963, + "grad_norm": 0.9941024469958015, + "learning_rate": 0.0009308920799803793, + "loss": 2.5763, + "step": 424 + }, + { + "epoch": 0.1949317738791423, + "grad_norm": 1.1077135006862129, + "learning_rate": 0.0009305146762557246, + "loss": 2.4588, + "step": 425 + }, + { + "epoch": 0.19539043687650498, + "grad_norm": 0.944764738876573, + "learning_rate": 0.0009301363217562736, + "loss": 2.4982, + "step": 426 + }, + { + "epoch": 0.19584909987386767, + "grad_norm": 0.8924687385030895, + "learning_rate": 0.0009297570173176074, + "loss": 2.4944, + "step": 427 + }, + { + "epoch": 0.19630776287123036, + "grad_norm": 0.9472139459273335, + "learning_rate": 0.000929376763777405, + "loss": 2.5134, + "step": 428 + }, + { + "epoch": 0.19676642586859305, + "grad_norm": 0.9614151121701311, + "learning_rate": 0.0009289955619754413, + "loss": 2.4818, + "step": 429 + }, + { + "epoch": 0.19722508886595574, + "grad_norm": 0.9898484519081638, + "learning_rate": 0.0009286134127535859, + "loss": 2.5415, + "step": 430 + }, + { + "epoch": 0.19768375186331844, + "grad_norm": 0.8375413967357035, + "learning_rate": 0.0009282303169558, + "loss": 2.4765, + "step": 431 + }, + { + "epoch": 0.19814241486068113, + "grad_norm": 0.8719423103357349, + "learning_rate": 0.0009278462754281359, + "loss": 2.4737, + "step": 432 + }, + { + "epoch": 0.1986010778580438, + "grad_norm": 0.8884780278284764, + "learning_rate": 0.0009274612890187342, + "loss": 2.4954, + "step": 433 + }, + { + "epoch": 0.19905974085540648, + "grad_norm": 0.8902784923582241, + "learning_rate": 0.0009270753585778222, + "loss": 2.5501, + "step": 434 + }, + { + "epoch": 0.19951840385276917, + "grad_norm": 0.9915820216889567, + "learning_rate": 0.0009266884849577124, + "loss": 2.3976, + "step": 435 + }, + { + "epoch": 0.19997706685013186, + "grad_norm": 1.0756670482811415, + "learning_rate": 0.0009263006690127998, + "loss": 2.4378, + "step": 436 + }, + { + "epoch": 0.20043572984749455, + "grad_norm": 0.9920576325987812, + "learning_rate": 0.0009259119115995609, + "loss": 2.5172, + "step": 437 + }, + { + "epoch": 0.20089439284485724, + "grad_norm": 0.9608966221708257, + "learning_rate": 0.0009255222135765511, + "loss": 2.4501, + "step": 438 + }, + { + "epoch": 0.20135305584221994, + "grad_norm": 1.0451292029331767, + "learning_rate": 0.0009251315758044032, + "loss": 2.5653, + "step": 439 + }, + { + "epoch": 0.20181171883958263, + "grad_norm": 0.964250736696325, + "learning_rate": 0.0009247399991458255, + "loss": 2.4586, + "step": 440 + }, + { + "epoch": 0.20227038183694532, + "grad_norm": 0.9650448834451658, + "learning_rate": 0.0009243474844655996, + "loss": 2.5166, + "step": 441 + }, + { + "epoch": 0.20272904483430798, + "grad_norm": 0.9219411327725426, + "learning_rate": 0.0009239540326305791, + "loss": 2.3861, + "step": 442 + }, + { + "epoch": 0.20318770783167067, + "grad_norm": 0.9024695273911946, + "learning_rate": 0.0009235596445096864, + "loss": 2.6271, + "step": 443 + }, + { + "epoch": 0.20364637082903336, + "grad_norm": 0.9168862602334338, + "learning_rate": 0.0009231643209739127, + "loss": 2.553, + "step": 444 + }, + { + "epoch": 0.20410503382639605, + "grad_norm": 0.92854044288624, + "learning_rate": 0.0009227680628963145, + "loss": 2.501, + "step": 445 + }, + { + "epoch": 0.20456369682375874, + "grad_norm": 0.8946575926623643, + "learning_rate": 0.000922370871152012, + "loss": 2.537, + "step": 446 + }, + { + "epoch": 0.20502235982112144, + "grad_norm": 0.9096102676506979, + "learning_rate": 0.0009219727466181877, + "loss": 2.4909, + "step": 447 + }, + { + "epoch": 0.20548102281848413, + "grad_norm": 0.9453648777414424, + "learning_rate": 0.0009215736901740841, + "loss": 2.498, + "step": 448 + }, + { + "epoch": 0.20593968581584682, + "grad_norm": 0.9577740911382892, + "learning_rate": 0.0009211737027010016, + "loss": 2.5707, + "step": 449 + }, + { + "epoch": 0.20639834881320948, + "grad_norm": 0.8903929585522322, + "learning_rate": 0.0009207727850822971, + "loss": 2.4874, + "step": 450 + }, + { + "epoch": 0.20685701181057217, + "grad_norm": 0.8855260339066233, + "learning_rate": 0.0009203709382033814, + "loss": 2.5534, + "step": 451 + }, + { + "epoch": 0.20731567480793486, + "grad_norm": 0.9519698765739639, + "learning_rate": 0.0009199681629517173, + "loss": 2.47, + "step": 452 + }, + { + "epoch": 0.20777433780529755, + "grad_norm": 0.8848815823522478, + "learning_rate": 0.0009195644602168184, + "loss": 2.5019, + "step": 453 + }, + { + "epoch": 0.20823300080266025, + "grad_norm": 0.9702586648781519, + "learning_rate": 0.0009191598308902464, + "loss": 2.4819, + "step": 454 + }, + { + "epoch": 0.20869166380002294, + "grad_norm": 0.9878715531833094, + "learning_rate": 0.0009187542758656091, + "loss": 2.4987, + "step": 455 + }, + { + "epoch": 0.20915032679738563, + "grad_norm": 1.0013211735145715, + "learning_rate": 0.0009183477960385591, + "loss": 2.3832, + "step": 456 + }, + { + "epoch": 0.20960898979474832, + "grad_norm": 0.8838496673852833, + "learning_rate": 0.0009179403923067912, + "loss": 2.5325, + "step": 457 + }, + { + "epoch": 0.210067652792111, + "grad_norm": 0.9144758379059125, + "learning_rate": 0.0009175320655700406, + "loss": 2.4353, + "step": 458 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 0.9184846750814274, + "learning_rate": 0.0009171228167300805, + "loss": 2.4888, + "step": 459 + }, + { + "epoch": 0.21098497878683636, + "grad_norm": 0.8848707484710656, + "learning_rate": 0.0009167126466907215, + "loss": 2.4645, + "step": 460 + }, + { + "epoch": 0.21144364178419905, + "grad_norm": 0.8538235338352425, + "learning_rate": 0.0009163015563578074, + "loss": 2.4666, + "step": 461 + }, + { + "epoch": 0.21190230478156175, + "grad_norm": 0.8598476127463165, + "learning_rate": 0.0009158895466392158, + "loss": 2.5715, + "step": 462 + }, + { + "epoch": 0.21236096777892444, + "grad_norm": 1.0157389557138423, + "learning_rate": 0.0009154766184448535, + "loss": 2.449, + "step": 463 + }, + { + "epoch": 0.21281963077628713, + "grad_norm": 1.0130266040008897, + "learning_rate": 0.0009150627726866568, + "loss": 2.3851, + "step": 464 + }, + { + "epoch": 0.21327829377364982, + "grad_norm": 0.9171790641864079, + "learning_rate": 0.000914648010278587, + "loss": 2.4031, + "step": 465 + }, + { + "epoch": 0.2137369567710125, + "grad_norm": 0.9169923812728696, + "learning_rate": 0.0009142323321366315, + "loss": 2.5277, + "step": 466 + }, + { + "epoch": 0.21419561976837517, + "grad_norm": 1.0434205042053541, + "learning_rate": 0.0009138157391787986, + "loss": 2.5763, + "step": 467 + }, + { + "epoch": 0.21465428276573786, + "grad_norm": 0.8510356930064098, + "learning_rate": 0.0009133982323251177, + "loss": 2.4723, + "step": 468 + }, + { + "epoch": 0.21511294576310055, + "grad_norm": 1.3553998922434278, + "learning_rate": 0.0009129798124976365, + "loss": 2.5746, + "step": 469 + }, + { + "epoch": 0.21557160876046325, + "grad_norm": 0.9157227754940461, + "learning_rate": 0.0009125604806204187, + "loss": 2.4883, + "step": 470 + }, + { + "epoch": 0.21603027175782594, + "grad_norm": 0.9773883240294716, + "learning_rate": 0.0009121402376195421, + "loss": 2.4182, + "step": 471 + }, + { + "epoch": 0.21648893475518863, + "grad_norm": 4.4292397017670915, + "learning_rate": 0.0009117190844230972, + "loss": 2.5227, + "step": 472 + }, + { + "epoch": 0.21694759775255132, + "grad_norm": 0.9436267500408707, + "learning_rate": 0.0009112970219611841, + "loss": 2.4457, + "step": 473 + }, + { + "epoch": 0.217406260749914, + "grad_norm": 1.0116790584887982, + "learning_rate": 0.0009108740511659115, + "loss": 2.5686, + "step": 474 + }, + { + "epoch": 0.2178649237472767, + "grad_norm": 0.9199579328222662, + "learning_rate": 0.0009104501729713935, + "loss": 2.4734, + "step": 475 + }, + { + "epoch": 0.21832358674463936, + "grad_norm": 0.8143028630561481, + "learning_rate": 0.0009100253883137488, + "loss": 2.4805, + "step": 476 + }, + { + "epoch": 0.21878224974200206, + "grad_norm": 0.999373977902029, + "learning_rate": 0.0009095996981310974, + "loss": 2.5182, + "step": 477 + }, + { + "epoch": 0.21924091273936475, + "grad_norm": 0.9956383787942932, + "learning_rate": 0.0009091731033635596, + "loss": 2.5123, + "step": 478 + }, + { + "epoch": 0.21969957573672744, + "grad_norm": 0.8691088930995258, + "learning_rate": 0.0009087456049532529, + "loss": 2.4718, + "step": 479 + }, + { + "epoch": 0.22015823873409013, + "grad_norm": 0.8141229030983909, + "learning_rate": 0.0009083172038442914, + "loss": 2.4762, + "step": 480 + }, + { + "epoch": 0.22061690173145282, + "grad_norm": 0.8734580159583423, + "learning_rate": 0.0009078879009827817, + "loss": 2.402, + "step": 481 + }, + { + "epoch": 0.2210755647288155, + "grad_norm": 0.9626327140459192, + "learning_rate": 0.0009074576973168223, + "loss": 2.504, + "step": 482 + }, + { + "epoch": 0.2215342277261782, + "grad_norm": 0.8846998630465084, + "learning_rate": 0.0009070265937965015, + "loss": 2.4179, + "step": 483 + }, + { + "epoch": 0.2219928907235409, + "grad_norm": 0.8778066114943069, + "learning_rate": 0.0009065945913738942, + "loss": 2.4859, + "step": 484 + }, + { + "epoch": 0.22245155372090356, + "grad_norm": 1.0228157879845547, + "learning_rate": 0.0009061616910030609, + "loss": 2.4642, + "step": 485 + }, + { + "epoch": 0.22291021671826625, + "grad_norm": 0.8471206901342737, + "learning_rate": 0.0009057278936400453, + "loss": 2.4414, + "step": 486 + }, + { + "epoch": 0.22336887971562894, + "grad_norm": 0.8461000298874403, + "learning_rate": 0.0009052932002428715, + "loss": 2.4912, + "step": 487 + }, + { + "epoch": 0.22382754271299163, + "grad_norm": 1.1327323221332781, + "learning_rate": 0.0009048576117715435, + "loss": 2.4959, + "step": 488 + }, + { + "epoch": 0.22428620571035432, + "grad_norm": 0.9531509495303944, + "learning_rate": 0.0009044211291880407, + "loss": 2.4675, + "step": 489 + }, + { + "epoch": 0.224744868707717, + "grad_norm": 0.8773264067116902, + "learning_rate": 0.000903983753456318, + "loss": 2.4176, + "step": 490 + }, + { + "epoch": 0.2252035317050797, + "grad_norm": 0.8892079204383692, + "learning_rate": 0.0009035454855423026, + "loss": 2.5109, + "step": 491 + }, + { + "epoch": 0.2256621947024424, + "grad_norm": 1.0270176417825871, + "learning_rate": 0.0009031063264138922, + "loss": 2.4564, + "step": 492 + }, + { + "epoch": 0.22612085769980506, + "grad_norm": 0.9041131041747281, + "learning_rate": 0.0009026662770409522, + "loss": 2.4995, + "step": 493 + }, + { + "epoch": 0.22657952069716775, + "grad_norm": 0.8552496880025156, + "learning_rate": 0.0009022253383953147, + "loss": 2.5115, + "step": 494 + }, + { + "epoch": 0.22703818369453044, + "grad_norm": 0.8469990774489283, + "learning_rate": 0.0009017835114507753, + "loss": 2.4326, + "step": 495 + }, + { + "epoch": 0.22749684669189313, + "grad_norm": 0.8589968145415151, + "learning_rate": 0.0009013407971830914, + "loss": 2.4684, + "step": 496 + }, + { + "epoch": 0.22795550968925582, + "grad_norm": 0.9212261300703677, + "learning_rate": 0.0009008971965699801, + "loss": 2.5679, + "step": 497 + }, + { + "epoch": 0.2284141726866185, + "grad_norm": 0.9318625205283657, + "learning_rate": 0.0009004527105911163, + "loss": 2.5031, + "step": 498 + }, + { + "epoch": 0.2288728356839812, + "grad_norm": 0.8921431175030273, + "learning_rate": 0.0009000073402281295, + "loss": 2.5107, + "step": 499 + }, + { + "epoch": 0.2293314986813439, + "grad_norm": 0.8492920202746853, + "learning_rate": 0.0008995610864646028, + "loss": 2.4078, + "step": 500 + }, + { + "epoch": 0.22979016167870658, + "grad_norm": 0.9766381636459567, + "learning_rate": 0.0008991139502860703, + "loss": 2.4173, + "step": 501 + }, + { + "epoch": 0.23024882467606925, + "grad_norm": 0.8460176112904854, + "learning_rate": 0.0008986659326800146, + "loss": 2.3551, + "step": 502 + }, + { + "epoch": 0.23070748767343194, + "grad_norm": 1.0938465840022817, + "learning_rate": 0.0008982170346358651, + "loss": 2.4802, + "step": 503 + }, + { + "epoch": 0.23116615067079463, + "grad_norm": 1.0334213812703288, + "learning_rate": 0.0008977672571449956, + "loss": 2.4036, + "step": 504 + }, + { + "epoch": 0.23162481366815732, + "grad_norm": 0.93303137409271, + "learning_rate": 0.0008973166012007217, + "loss": 2.5824, + "step": 505 + }, + { + "epoch": 0.23208347666552, + "grad_norm": 0.9112958011688143, + "learning_rate": 0.0008968650677982998, + "loss": 2.4727, + "step": 506 + }, + { + "epoch": 0.2325421396628827, + "grad_norm": 0.9080614540530094, + "learning_rate": 0.0008964126579349236, + "loss": 2.4187, + "step": 507 + }, + { + "epoch": 0.2330008026602454, + "grad_norm": 0.9279619326747133, + "learning_rate": 0.0008959593726097226, + "loss": 2.4305, + "step": 508 + }, + { + "epoch": 0.23345946565760808, + "grad_norm": 0.9807680939776316, + "learning_rate": 0.0008955052128237596, + "loss": 2.4095, + "step": 509 + }, + { + "epoch": 0.23391812865497075, + "grad_norm": 0.8994068566226616, + "learning_rate": 0.0008950501795800288, + "loss": 2.5365, + "step": 510 + }, + { + "epoch": 0.23437679165233344, + "grad_norm": 0.9178706992685652, + "learning_rate": 0.0008945942738834532, + "loss": 2.3954, + "step": 511 + }, + { + "epoch": 0.23483545464969613, + "grad_norm": 1.0265243769501937, + "learning_rate": 0.0008941374967408826, + "loss": 2.4983, + "step": 512 + }, + { + "epoch": 0.23529411764705882, + "grad_norm": 1.290346575601095, + "learning_rate": 0.0008936798491610916, + "loss": 2.4397, + "step": 513 + }, + { + "epoch": 0.2357527806444215, + "grad_norm": 0.9377351093962103, + "learning_rate": 0.0008932213321547768, + "loss": 2.4972, + "step": 514 + }, + { + "epoch": 0.2362114436417842, + "grad_norm": 0.9293371782393633, + "learning_rate": 0.0008927619467345554, + "loss": 2.4404, + "step": 515 + }, + { + "epoch": 0.2366701066391469, + "grad_norm": 0.9250705187045419, + "learning_rate": 0.0008923016939149615, + "loss": 2.4883, + "step": 516 + }, + { + "epoch": 0.23712876963650958, + "grad_norm": 0.8544360470029596, + "learning_rate": 0.0008918405747124458, + "loss": 2.3747, + "step": 517 + }, + { + "epoch": 0.23758743263387228, + "grad_norm": 0.9347899102583163, + "learning_rate": 0.0008913785901453721, + "loss": 2.4895, + "step": 518 + }, + { + "epoch": 0.23804609563123494, + "grad_norm": 0.8844495950864275, + "learning_rate": 0.000890915741234015, + "loss": 2.4499, + "step": 519 + }, + { + "epoch": 0.23850475862859763, + "grad_norm": 0.8772999705972551, + "learning_rate": 0.0008904520290005582, + "loss": 2.5083, + "step": 520 + }, + { + "epoch": 0.23896342162596032, + "grad_norm": 1.2307602016255719, + "learning_rate": 0.000889987454469092, + "loss": 2.5676, + "step": 521 + }, + { + "epoch": 0.239422084623323, + "grad_norm": 0.8237805633253273, + "learning_rate": 0.0008895220186656111, + "loss": 2.4054, + "step": 522 + }, + { + "epoch": 0.2398807476206857, + "grad_norm": 0.8600742426386667, + "learning_rate": 0.0008890557226180122, + "loss": 2.423, + "step": 523 + }, + { + "epoch": 0.2403394106180484, + "grad_norm": 0.8768319720666145, + "learning_rate": 0.0008885885673560921, + "loss": 2.3781, + "step": 524 + }, + { + "epoch": 0.24079807361541108, + "grad_norm": 1.0433034624583348, + "learning_rate": 0.0008881205539115444, + "loss": 2.5172, + "step": 525 + }, + { + "epoch": 0.24125673661277378, + "grad_norm": 0.8413039071877446, + "learning_rate": 0.0008876516833179589, + "loss": 2.4865, + "step": 526 + }, + { + "epoch": 0.24171539961013644, + "grad_norm": 0.7784609486433713, + "learning_rate": 0.0008871819566108177, + "loss": 2.3554, + "step": 527 + }, + { + "epoch": 0.24217406260749913, + "grad_norm": 0.9225845057933544, + "learning_rate": 0.000886711374827494, + "loss": 2.4116, + "step": 528 + }, + { + "epoch": 0.24263272560486182, + "grad_norm": 0.8856517565393851, + "learning_rate": 0.0008862399390072491, + "loss": 2.4622, + "step": 529 + }, + { + "epoch": 0.2430913886022245, + "grad_norm": 0.7982008531732973, + "learning_rate": 0.0008857676501912305, + "loss": 2.4177, + "step": 530 + }, + { + "epoch": 0.2435500515995872, + "grad_norm": 0.8168803824267935, + "learning_rate": 0.0008852945094224697, + "loss": 2.4243, + "step": 531 + }, + { + "epoch": 0.2440087145969499, + "grad_norm": 0.8397985848341906, + "learning_rate": 0.0008848205177458795, + "loss": 2.4475, + "step": 532 + }, + { + "epoch": 0.24446737759431258, + "grad_norm": 0.8478714962601426, + "learning_rate": 0.0008843456762082518, + "loss": 2.352, + "step": 533 + }, + { + "epoch": 0.24492604059167528, + "grad_norm": 0.946614861992968, + "learning_rate": 0.0008838699858582557, + "loss": 2.4891, + "step": 534 + }, + { + "epoch": 0.24538470358903797, + "grad_norm": 0.8508565866448745, + "learning_rate": 0.0008833934477464347, + "loss": 2.503, + "step": 535 + }, + { + "epoch": 0.24584336658640063, + "grad_norm": 0.8953856239334935, + "learning_rate": 0.0008829160629252045, + "loss": 2.48, + "step": 536 + }, + { + "epoch": 0.24630202958376332, + "grad_norm": 0.9300961965830582, + "learning_rate": 0.0008824378324488509, + "loss": 2.5084, + "step": 537 + }, + { + "epoch": 0.246760692581126, + "grad_norm": 0.8814609901283131, + "learning_rate": 0.0008819587573735268, + "loss": 2.5174, + "step": 538 + }, + { + "epoch": 0.2472193555784887, + "grad_norm": 0.8550444049938698, + "learning_rate": 0.0008814788387572513, + "loss": 2.3388, + "step": 539 + }, + { + "epoch": 0.2476780185758514, + "grad_norm": 0.8183429730299434, + "learning_rate": 0.0008809980776599053, + "loss": 2.4325, + "step": 540 + }, + { + "epoch": 0.24813668157321409, + "grad_norm": 0.9788282287891815, + "learning_rate": 0.0008805164751432312, + "loss": 2.3554, + "step": 541 + }, + { + "epoch": 0.24859534457057678, + "grad_norm": 0.8317378522493669, + "learning_rate": 0.0008800340322708292, + "loss": 2.5041, + "step": 542 + }, + { + "epoch": 0.24905400756793947, + "grad_norm": 0.822051324493021, + "learning_rate": 0.0008795507501081555, + "loss": 2.3911, + "step": 543 + }, + { + "epoch": 0.24951267056530213, + "grad_norm": 0.8768513279009821, + "learning_rate": 0.0008790666297225196, + "loss": 2.3524, + "step": 544 + }, + { + "epoch": 0.24997133356266482, + "grad_norm": 0.9464411757505273, + "learning_rate": 0.0008785816721830829, + "loss": 2.5047, + "step": 545 + }, + { + "epoch": 0.2504299965600275, + "grad_norm": 0.965325868548125, + "learning_rate": 0.0008780958785608546, + "loss": 2.5239, + "step": 546 + }, + { + "epoch": 0.2508886595573902, + "grad_norm": 0.84323037132264, + "learning_rate": 0.0008776092499286912, + "loss": 2.4115, + "step": 547 + }, + { + "epoch": 0.2513473225547529, + "grad_norm": 0.8108609055013464, + "learning_rate": 0.0008771217873612929, + "loss": 2.3956, + "step": 548 + }, + { + "epoch": 0.2518059855521156, + "grad_norm": 0.8551644280785637, + "learning_rate": 0.0008766334919352017, + "loss": 2.5226, + "step": 549 + }, + { + "epoch": 0.2522646485494783, + "grad_norm": 0.8282503718033429, + "learning_rate": 0.0008761443647287987, + "loss": 2.3886, + "step": 550 + }, + { + "epoch": 0.25272331154684097, + "grad_norm": 0.8637993594192993, + "learning_rate": 0.0008756544068223026, + "loss": 2.4426, + "step": 551 + }, + { + "epoch": 0.25318197454420366, + "grad_norm": 0.8414971773479316, + "learning_rate": 0.0008751636192977659, + "loss": 2.434, + "step": 552 + }, + { + "epoch": 0.25364063754156635, + "grad_norm": 0.8380499523329216, + "learning_rate": 0.0008746720032390737, + "loss": 2.5292, + "step": 553 + }, + { + "epoch": 0.25409930053892904, + "grad_norm": 0.8873430397275823, + "learning_rate": 0.0008741795597319408, + "loss": 2.462, + "step": 554 + }, + { + "epoch": 0.25455796353629173, + "grad_norm": 0.9209104928384224, + "learning_rate": 0.0008736862898639095, + "loss": 2.4546, + "step": 555 + }, + { + "epoch": 0.2550166265336544, + "grad_norm": 0.8099674858586006, + "learning_rate": 0.0008731921947243468, + "loss": 2.4332, + "step": 556 + }, + { + "epoch": 0.25547528953101706, + "grad_norm": 0.823413045659857, + "learning_rate": 0.0008726972754044427, + "loss": 2.4887, + "step": 557 + }, + { + "epoch": 0.25593395252837975, + "grad_norm": 0.8594896620813187, + "learning_rate": 0.0008722015329972069, + "loss": 2.4928, + "step": 558 + }, + { + "epoch": 0.25639261552574244, + "grad_norm": 0.8578042447666914, + "learning_rate": 0.0008717049685974672, + "loss": 2.46, + "step": 559 + }, + { + "epoch": 0.25685127852310513, + "grad_norm": 0.8647974655545693, + "learning_rate": 0.0008712075833018665, + "loss": 2.4276, + "step": 560 + }, + { + "epoch": 0.2573099415204678, + "grad_norm": 0.8709795922620308, + "learning_rate": 0.0008707093782088608, + "loss": 2.4127, + "step": 561 + }, + { + "epoch": 0.2577686045178305, + "grad_norm": 0.9069356332437817, + "learning_rate": 0.0008702103544187167, + "loss": 2.394, + "step": 562 + }, + { + "epoch": 0.2582272675151932, + "grad_norm": 0.8419295420999467, + "learning_rate": 0.0008697105130335085, + "loss": 2.46, + "step": 563 + }, + { + "epoch": 0.2586859305125559, + "grad_norm": 0.8973995215847101, + "learning_rate": 0.0008692098551571164, + "loss": 2.5097, + "step": 564 + }, + { + "epoch": 0.2591445935099186, + "grad_norm": 0.8798937039496577, + "learning_rate": 0.0008687083818952235, + "loss": 2.4628, + "step": 565 + }, + { + "epoch": 0.2596032565072813, + "grad_norm": 0.8631335658250031, + "learning_rate": 0.0008682060943553143, + "loss": 2.4409, + "step": 566 + }, + { + "epoch": 0.26006191950464397, + "grad_norm": 0.9236061206149685, + "learning_rate": 0.0008677029936466707, + "loss": 2.4485, + "step": 567 + }, + { + "epoch": 0.26052058250200666, + "grad_norm": 0.8202733625813916, + "learning_rate": 0.0008671990808803711, + "loss": 2.5685, + "step": 568 + }, + { + "epoch": 0.26097924549936935, + "grad_norm": 0.8614138983332668, + "learning_rate": 0.0008666943571692871, + "loss": 2.4417, + "step": 569 + }, + { + "epoch": 0.26143790849673204, + "grad_norm": 0.8720572990087369, + "learning_rate": 0.0008661888236280813, + "loss": 2.4817, + "step": 570 + }, + { + "epoch": 0.26189657149409473, + "grad_norm": 0.7836327863930004, + "learning_rate": 0.0008656824813732045, + "loss": 2.3356, + "step": 571 + }, + { + "epoch": 0.2623552344914574, + "grad_norm": 0.935384270325341, + "learning_rate": 0.000865175331522894, + "loss": 2.4154, + "step": 572 + }, + { + "epoch": 0.2628138974888201, + "grad_norm": 0.8013227083537534, + "learning_rate": 0.0008646673751971703, + "loss": 2.3505, + "step": 573 + }, + { + "epoch": 0.26327256048618275, + "grad_norm": 0.8141278445412143, + "learning_rate": 0.000864158613517835, + "loss": 2.3483, + "step": 574 + }, + { + "epoch": 0.26373122348354544, + "grad_norm": 0.8496708715009139, + "learning_rate": 0.0008636490476084681, + "loss": 2.4629, + "step": 575 + }, + { + "epoch": 0.26418988648090813, + "grad_norm": 0.8292260559645915, + "learning_rate": 0.0008631386785944264, + "loss": 2.4333, + "step": 576 + }, + { + "epoch": 0.2646485494782708, + "grad_norm": 0.9005695879984328, + "learning_rate": 0.0008626275076028397, + "loss": 2.463, + "step": 577 + }, + { + "epoch": 0.2651072124756335, + "grad_norm": 0.8366209467506159, + "learning_rate": 0.0008621155357626091, + "loss": 2.4822, + "step": 578 + }, + { + "epoch": 0.2655658754729962, + "grad_norm": 0.946289608511057, + "learning_rate": 0.0008616027642044042, + "loss": 2.425, + "step": 579 + }, + { + "epoch": 0.2660245384703589, + "grad_norm": 0.8271880490329367, + "learning_rate": 0.000861089194060661, + "loss": 2.5182, + "step": 580 + }, + { + "epoch": 0.2664832014677216, + "grad_norm": 0.8013761821651919, + "learning_rate": 0.000860574826465579, + "loss": 2.3928, + "step": 581 + }, + { + "epoch": 0.2669418644650843, + "grad_norm": 0.8657282022453442, + "learning_rate": 0.0008600596625551191, + "loss": 2.5122, + "step": 582 + }, + { + "epoch": 0.26740052746244697, + "grad_norm": 0.8589717122852175, + "learning_rate": 0.0008595437034670006, + "loss": 2.4324, + "step": 583 + }, + { + "epoch": 0.26785919045980966, + "grad_norm": 0.8446956859188163, + "learning_rate": 0.0008590269503406985, + "loss": 2.416, + "step": 584 + }, + { + "epoch": 0.26831785345717235, + "grad_norm": 0.8586872967643314, + "learning_rate": 0.0008585094043174423, + "loss": 2.493, + "step": 585 + }, + { + "epoch": 0.26877651645453504, + "grad_norm": 0.8420559350069495, + "learning_rate": 0.0008579910665402118, + "loss": 2.4797, + "step": 586 + }, + { + "epoch": 0.26923517945189773, + "grad_norm": 0.8960544633441714, + "learning_rate": 0.000857471938153736, + "loss": 2.5327, + "step": 587 + }, + { + "epoch": 0.2696938424492604, + "grad_norm": 0.7958749831446542, + "learning_rate": 0.0008569520203044892, + "loss": 2.4679, + "step": 588 + }, + { + "epoch": 0.2701525054466231, + "grad_norm": 0.8410268280820914, + "learning_rate": 0.0008564313141406901, + "loss": 2.3974, + "step": 589 + }, + { + "epoch": 0.2706111684439858, + "grad_norm": 1.1689235799052498, + "learning_rate": 0.0008559098208122973, + "loss": 2.4335, + "step": 590 + }, + { + "epoch": 0.2710698314413485, + "grad_norm": 0.8595751100171332, + "learning_rate": 0.0008553875414710089, + "loss": 2.447, + "step": 591 + }, + { + "epoch": 0.27152849443871113, + "grad_norm": 0.8181739774922024, + "learning_rate": 0.0008548644772702579, + "loss": 2.3308, + "step": 592 + }, + { + "epoch": 0.2719871574360738, + "grad_norm": 0.7884442878048739, + "learning_rate": 0.0008543406293652116, + "loss": 2.4547, + "step": 593 + }, + { + "epoch": 0.2724458204334365, + "grad_norm": 0.8419829089788782, + "learning_rate": 0.0008538159989127671, + "loss": 2.4564, + "step": 594 + }, + { + "epoch": 0.2729044834307992, + "grad_norm": 0.781705633155259, + "learning_rate": 0.0008532905870715505, + "loss": 2.4153, + "step": 595 + }, + { + "epoch": 0.2733631464281619, + "grad_norm": 0.8546660795111982, + "learning_rate": 0.0008527643950019131, + "loss": 2.5143, + "step": 596 + }, + { + "epoch": 0.2738218094255246, + "grad_norm": 0.8336780864000869, + "learning_rate": 0.0008522374238659296, + "loss": 2.4183, + "step": 597 + }, + { + "epoch": 0.2742804724228873, + "grad_norm": 0.8796147420187108, + "learning_rate": 0.0008517096748273951, + "loss": 2.3958, + "step": 598 + }, + { + "epoch": 0.27473913542024997, + "grad_norm": 0.8204209094387331, + "learning_rate": 0.0008511811490518227, + "loss": 2.3958, + "step": 599 + }, + { + "epoch": 0.27519779841761266, + "grad_norm": 0.8368277932224492, + "learning_rate": 0.0008506518477064405, + "loss": 2.5112, + "step": 600 + }, + { + "epoch": 0.27565646141497535, + "grad_norm": 0.8179727022179314, + "learning_rate": 0.0008501217719601903, + "loss": 2.3653, + "step": 601 + }, + { + "epoch": 0.27611512441233804, + "grad_norm": 0.8643229163654725, + "learning_rate": 0.0008495909229837233, + "loss": 2.5007, + "step": 602 + }, + { + "epoch": 0.27657378740970073, + "grad_norm": 1.6682856371639803, + "learning_rate": 0.000849059301949399, + "loss": 2.3807, + "step": 603 + }, + { + "epoch": 0.2770324504070634, + "grad_norm": 0.7653071905138391, + "learning_rate": 0.0008485269100312812, + "loss": 2.3852, + "step": 604 + }, + { + "epoch": 0.2774911134044261, + "grad_norm": 0.8301963093791922, + "learning_rate": 0.0008479937484051368, + "loss": 2.4297, + "step": 605 + }, + { + "epoch": 0.2779497764017888, + "grad_norm": 0.8858930529552053, + "learning_rate": 0.0008474598182484323, + "loss": 2.4438, + "step": 606 + }, + { + "epoch": 0.2784084393991515, + "grad_norm": 0.8416959390090784, + "learning_rate": 0.0008469251207403317, + "loss": 2.431, + "step": 607 + }, + { + "epoch": 0.2788671023965142, + "grad_norm": 0.8666732950442635, + "learning_rate": 0.0008463896570616934, + "loss": 2.4236, + "step": 608 + }, + { + "epoch": 0.2793257653938768, + "grad_norm": 0.8012320381027087, + "learning_rate": 0.0008458534283950678, + "loss": 2.464, + "step": 609 + }, + { + "epoch": 0.2797844283912395, + "grad_norm": 0.8516739667226084, + "learning_rate": 0.0008453164359246952, + "loss": 2.406, + "step": 610 + }, + { + "epoch": 0.2802430913886022, + "grad_norm": 0.9346353111923094, + "learning_rate": 0.0008447786808365022, + "loss": 2.4047, + "step": 611 + }, + { + "epoch": 0.2807017543859649, + "grad_norm": 0.796653470922592, + "learning_rate": 0.0008442401643181, + "loss": 2.4984, + "step": 612 + }, + { + "epoch": 0.2811604173833276, + "grad_norm": 1.1602626808447492, + "learning_rate": 0.0008437008875587811, + "loss": 2.4448, + "step": 613 + }, + { + "epoch": 0.2816190803806903, + "grad_norm": 0.8101049912950282, + "learning_rate": 0.0008431608517495171, + "loss": 2.3904, + "step": 614 + }, + { + "epoch": 0.28207774337805297, + "grad_norm": 0.8649919354246952, + "learning_rate": 0.0008426200580829561, + "loss": 2.3427, + "step": 615 + }, + { + "epoch": 0.28253640637541566, + "grad_norm": 0.7968219575358425, + "learning_rate": 0.0008420785077534195, + "loss": 2.4201, + "step": 616 + }, + { + "epoch": 0.28299506937277835, + "grad_norm": 0.7982902990252788, + "learning_rate": 0.0008415362019569001, + "loss": 2.4238, + "step": 617 + }, + { + "epoch": 0.28345373237014104, + "grad_norm": 0.7875969162665155, + "learning_rate": 0.0008409931418910591, + "loss": 2.4146, + "step": 618 + }, + { + "epoch": 0.28391239536750373, + "grad_norm": 0.7896846851164362, + "learning_rate": 0.0008404493287552232, + "loss": 2.4268, + "step": 619 + }, + { + "epoch": 0.2843710583648664, + "grad_norm": 0.7477799360904257, + "learning_rate": 0.0008399047637503825, + "loss": 2.4239, + "step": 620 + }, + { + "epoch": 0.2848297213622291, + "grad_norm": 0.7586912061239717, + "learning_rate": 0.0008393594480791875, + "loss": 2.371, + "step": 621 + }, + { + "epoch": 0.2852883843595918, + "grad_norm": 0.8112073961803687, + "learning_rate": 0.0008388133829459463, + "loss": 2.464, + "step": 622 + }, + { + "epoch": 0.2857470473569545, + "grad_norm": 0.9975375826880689, + "learning_rate": 0.0008382665695566227, + "loss": 2.4791, + "step": 623 + }, + { + "epoch": 0.2862057103543172, + "grad_norm": 2.075453863268213, + "learning_rate": 0.0008377190091188324, + "loss": 2.4051, + "step": 624 + }, + { + "epoch": 0.2866643733516799, + "grad_norm": 0.8786555436499865, + "learning_rate": 0.0008371707028418413, + "loss": 2.4234, + "step": 625 + }, + { + "epoch": 0.2871230363490425, + "grad_norm": 0.8443861962325123, + "learning_rate": 0.0008366216519365621, + "loss": 2.4771, + "step": 626 + }, + { + "epoch": 0.2875816993464052, + "grad_norm": 0.7968112417578864, + "learning_rate": 0.0008360718576155525, + "loss": 2.378, + "step": 627 + }, + { + "epoch": 0.2880403623437679, + "grad_norm": 0.8323157830088884, + "learning_rate": 0.0008355213210930118, + "loss": 2.3584, + "step": 628 + }, + { + "epoch": 0.2884990253411306, + "grad_norm": 0.8181279681714728, + "learning_rate": 0.0008349700435847778, + "loss": 2.5623, + "step": 629 + }, + { + "epoch": 0.2889576883384933, + "grad_norm": 0.8083394292232203, + "learning_rate": 0.0008344180263083256, + "loss": 2.4575, + "step": 630 + }, + { + "epoch": 0.28941635133585597, + "grad_norm": 1.5217188302082743, + "learning_rate": 0.000833865270482764, + "loss": 2.3304, + "step": 631 + }, + { + "epoch": 0.28987501433321866, + "grad_norm": 0.7678124699512735, + "learning_rate": 0.0008333117773288324, + "loss": 2.3848, + "step": 632 + }, + { + "epoch": 0.29033367733058135, + "grad_norm": 0.7677040300483599, + "learning_rate": 0.0008327575480688985, + "loss": 2.3675, + "step": 633 + }, + { + "epoch": 0.29079234032794404, + "grad_norm": 0.8181379965571308, + "learning_rate": 0.000832202583926956, + "loss": 2.3805, + "step": 634 + }, + { + "epoch": 0.29125100332530673, + "grad_norm": 0.7619986575161547, + "learning_rate": 0.0008316468861286217, + "loss": 2.4507, + "step": 635 + }, + { + "epoch": 0.2917096663226694, + "grad_norm": 0.7994438181137277, + "learning_rate": 0.0008310904559011323, + "loss": 2.4764, + "step": 636 + }, + { + "epoch": 0.2921683293200321, + "grad_norm": 0.7698611669513289, + "learning_rate": 0.0008305332944733419, + "loss": 2.431, + "step": 637 + }, + { + "epoch": 0.2926269923173948, + "grad_norm": 0.8302241413724637, + "learning_rate": 0.0008299754030757202, + "loss": 2.4246, + "step": 638 + }, + { + "epoch": 0.2930856553147575, + "grad_norm": 0.8110592861974777, + "learning_rate": 0.0008294167829403481, + "loss": 2.4901, + "step": 639 + }, + { + "epoch": 0.2935443183121202, + "grad_norm": 0.817739408197779, + "learning_rate": 0.0008288574353009164, + "loss": 2.4505, + "step": 640 + }, + { + "epoch": 0.2940029813094829, + "grad_norm": 0.8246350150287589, + "learning_rate": 0.0008282973613927225, + "loss": 2.3968, + "step": 641 + }, + { + "epoch": 0.29446164430684557, + "grad_norm": 0.7780271735081714, + "learning_rate": 0.0008277365624526675, + "loss": 2.4561, + "step": 642 + }, + { + "epoch": 0.2949203073042082, + "grad_norm": 0.846441023273551, + "learning_rate": 0.0008271750397192541, + "loss": 2.3555, + "step": 643 + }, + { + "epoch": 0.2953789703015709, + "grad_norm": 0.7875789460164012, + "learning_rate": 0.0008266127944325832, + "loss": 2.3992, + "step": 644 + }, + { + "epoch": 0.2958376332989336, + "grad_norm": 0.8145853780270296, + "learning_rate": 0.0008260498278343513, + "loss": 2.4502, + "step": 645 + }, + { + "epoch": 0.2962962962962963, + "grad_norm": 0.750768852434945, + "learning_rate": 0.0008254861411678485, + "loss": 2.4036, + "step": 646 + }, + { + "epoch": 0.29675495929365897, + "grad_norm": 0.7935706903827764, + "learning_rate": 0.0008249217356779544, + "loss": 2.3599, + "step": 647 + }, + { + "epoch": 0.29721362229102166, + "grad_norm": 0.7459007009223835, + "learning_rate": 0.0008243566126111363, + "loss": 2.3712, + "step": 648 + }, + { + "epoch": 0.29767228528838435, + "grad_norm": 0.7938023317911982, + "learning_rate": 0.0008237907732154466, + "loss": 2.5129, + "step": 649 + }, + { + "epoch": 0.29813094828574704, + "grad_norm": 0.7355354498152386, + "learning_rate": 0.0008232242187405194, + "loss": 2.4267, + "step": 650 + }, + { + "epoch": 0.29858961128310973, + "grad_norm": 0.7965141221020842, + "learning_rate": 0.000822656950437568, + "loss": 2.4049, + "step": 651 + }, + { + "epoch": 0.2990482742804724, + "grad_norm": 0.8398475670843923, + "learning_rate": 0.0008220889695593823, + "loss": 2.4217, + "step": 652 + }, + { + "epoch": 0.2995069372778351, + "grad_norm": 0.8272614748932868, + "learning_rate": 0.0008215202773603259, + "loss": 2.452, + "step": 653 + }, + { + "epoch": 0.2999656002751978, + "grad_norm": 0.8105582062488457, + "learning_rate": 0.0008209508750963328, + "loss": 2.3839, + "step": 654 + }, + { + "epoch": 0.3004242632725605, + "grad_norm": 0.8447206108526784, + "learning_rate": 0.0008203807640249062, + "loss": 2.4399, + "step": 655 + }, + { + "epoch": 0.3008829262699232, + "grad_norm": 0.7421520640781, + "learning_rate": 0.0008198099454051136, + "loss": 2.5215, + "step": 656 + }, + { + "epoch": 0.3013415892672859, + "grad_norm": 1.0330236862718911, + "learning_rate": 0.0008192384204975857, + "loss": 2.3618, + "step": 657 + }, + { + "epoch": 0.30180025226464857, + "grad_norm": 0.7996490622222933, + "learning_rate": 0.000818666190564513, + "loss": 2.3802, + "step": 658 + }, + { + "epoch": 0.30225891526201126, + "grad_norm": 0.7208641345758762, + "learning_rate": 0.0008180932568696426, + "loss": 2.3673, + "step": 659 + }, + { + "epoch": 0.3027175782593739, + "grad_norm": 0.8447548160102745, + "learning_rate": 0.0008175196206782764, + "loss": 2.3796, + "step": 660 + }, + { + "epoch": 0.3031762412567366, + "grad_norm": 0.8113693168781959, + "learning_rate": 0.0008169452832572675, + "loss": 2.3417, + "step": 661 + }, + { + "epoch": 0.3036349042540993, + "grad_norm": 0.7819680497191763, + "learning_rate": 0.0008163702458750173, + "loss": 2.481, + "step": 662 + }, + { + "epoch": 0.30409356725146197, + "grad_norm": 0.810164866205099, + "learning_rate": 0.0008157945098014734, + "loss": 2.4319, + "step": 663 + }, + { + "epoch": 0.30455223024882466, + "grad_norm": 1.5410929439288308, + "learning_rate": 0.0008152180763081267, + "loss": 2.4645, + "step": 664 + }, + { + "epoch": 0.30501089324618735, + "grad_norm": 0.7608474825043635, + "learning_rate": 0.0008146409466680076, + "loss": 2.437, + "step": 665 + }, + { + "epoch": 0.30546955624355004, + "grad_norm": 0.9353774749433554, + "learning_rate": 0.0008140631221556845, + "loss": 2.4215, + "step": 666 + }, + { + "epoch": 0.30592821924091274, + "grad_norm": 0.8999868146624684, + "learning_rate": 0.0008134846040472599, + "loss": 2.4185, + "step": 667 + }, + { + "epoch": 0.3063868822382754, + "grad_norm": 0.8629585430917012, + "learning_rate": 0.0008129053936203688, + "loss": 2.3927, + "step": 668 + }, + { + "epoch": 0.3068455452356381, + "grad_norm": 0.7428630278250291, + "learning_rate": 0.0008123254921541745, + "loss": 2.3954, + "step": 669 + }, + { + "epoch": 0.3073042082330008, + "grad_norm": 0.7996302758801848, + "learning_rate": 0.0008117449009293668, + "loss": 2.3975, + "step": 670 + }, + { + "epoch": 0.3077628712303635, + "grad_norm": 0.7946898912449113, + "learning_rate": 0.0008111636212281586, + "loss": 2.4619, + "step": 671 + }, + { + "epoch": 0.3082215342277262, + "grad_norm": 0.837690334113698, + "learning_rate": 0.0008105816543342833, + "loss": 2.4186, + "step": 672 + }, + { + "epoch": 0.3086801972250889, + "grad_norm": 0.7299243428309126, + "learning_rate": 0.0008099990015329919, + "loss": 2.431, + "step": 673 + }, + { + "epoch": 0.3091388602224516, + "grad_norm": 0.8096814837743994, + "learning_rate": 0.0008094156641110504, + "loss": 2.3954, + "step": 674 + }, + { + "epoch": 0.30959752321981426, + "grad_norm": 0.7649261476085886, + "learning_rate": 0.0008088316433567369, + "loss": 2.5299, + "step": 675 + }, + { + "epoch": 0.31005618621717695, + "grad_norm": 0.7391444747179154, + "learning_rate": 0.0008082469405598378, + "loss": 2.4769, + "step": 676 + }, + { + "epoch": 0.3105148492145396, + "grad_norm": 0.7613055621876477, + "learning_rate": 0.0008076615570116468, + "loss": 2.4258, + "step": 677 + }, + { + "epoch": 0.3109735122119023, + "grad_norm": 0.7888063028498611, + "learning_rate": 0.0008070754940049603, + "loss": 2.4967, + "step": 678 + }, + { + "epoch": 0.31143217520926497, + "grad_norm": 0.7414411038860752, + "learning_rate": 0.0008064887528340756, + "loss": 2.3863, + "step": 679 + }, + { + "epoch": 0.31189083820662766, + "grad_norm": 0.7448954355524465, + "learning_rate": 0.0008059013347947874, + "loss": 2.4467, + "step": 680 + }, + { + "epoch": 0.31234950120399035, + "grad_norm": 0.8078290417819287, + "learning_rate": 0.0008053132411843857, + "loss": 2.3595, + "step": 681 + }, + { + "epoch": 0.31280816420135305, + "grad_norm": 0.8147968288797134, + "learning_rate": 0.0008047244733016521, + "loss": 2.4675, + "step": 682 + }, + { + "epoch": 0.31326682719871574, + "grad_norm": 0.8941397939562651, + "learning_rate": 0.0008041350324468573, + "loss": 2.3597, + "step": 683 + }, + { + "epoch": 0.3137254901960784, + "grad_norm": 0.7634742084478972, + "learning_rate": 0.0008035449199217583, + "loss": 2.4717, + "step": 684 + }, + { + "epoch": 0.3141841531934411, + "grad_norm": 0.8215795339826173, + "learning_rate": 0.0008029541370295957, + "loss": 2.4073, + "step": 685 + }, + { + "epoch": 0.3146428161908038, + "grad_norm": 0.8564847638678296, + "learning_rate": 0.0008023626850750903, + "loss": 2.4321, + "step": 686 + }, + { + "epoch": 0.3151014791881665, + "grad_norm": 0.7491125061617957, + "learning_rate": 0.0008017705653644406, + "loss": 2.4546, + "step": 687 + }, + { + "epoch": 0.3155601421855292, + "grad_norm": 0.7694046209486628, + "learning_rate": 0.0008011777792053195, + "loss": 2.2961, + "step": 688 + }, + { + "epoch": 0.3160188051828919, + "grad_norm": 0.9676154557448575, + "learning_rate": 0.0008005843279068725, + "loss": 2.4458, + "step": 689 + }, + { + "epoch": 0.3164774681802546, + "grad_norm": 0.7885731783977579, + "learning_rate": 0.000799990212779713, + "loss": 2.4766, + "step": 690 + }, + { + "epoch": 0.31693613117761726, + "grad_norm": 1.0609569155257992, + "learning_rate": 0.0007993954351359214, + "loss": 2.4738, + "step": 691 + }, + { + "epoch": 0.31739479417497996, + "grad_norm": 0.8571827856040641, + "learning_rate": 0.0007987999962890406, + "loss": 2.5547, + "step": 692 + }, + { + "epoch": 0.31785345717234265, + "grad_norm": 0.932831413713626, + "learning_rate": 0.0007982038975540742, + "loss": 2.3701, + "step": 693 + }, + { + "epoch": 0.3183121201697053, + "grad_norm": 0.8155060550529012, + "learning_rate": 0.0007976071402474826, + "loss": 2.3589, + "step": 694 + }, + { + "epoch": 0.318770783167068, + "grad_norm": 0.9083305491959367, + "learning_rate": 0.0007970097256871811, + "loss": 2.4627, + "step": 695 + }, + { + "epoch": 0.31922944616443066, + "grad_norm": 0.7311420421116644, + "learning_rate": 0.0007964116551925364, + "loss": 2.4368, + "step": 696 + }, + { + "epoch": 0.31968810916179335, + "grad_norm": 0.9165656147817087, + "learning_rate": 0.0007958129300843637, + "loss": 2.3688, + "step": 697 + }, + { + "epoch": 0.32014677215915605, + "grad_norm": 0.783480290376345, + "learning_rate": 0.0007952135516849239, + "loss": 2.336, + "step": 698 + }, + { + "epoch": 0.32060543515651874, + "grad_norm": 0.7531292748057887, + "learning_rate": 0.0007946135213179207, + "loss": 2.465, + "step": 699 + }, + { + "epoch": 0.3210640981538814, + "grad_norm": 0.8315002456164453, + "learning_rate": 0.0007940128403084977, + "loss": 2.5049, + "step": 700 + }, + { + "epoch": 0.3215227611512441, + "grad_norm": 0.7672063541297209, + "learning_rate": 0.0007934115099832355, + "loss": 2.4293, + "step": 701 + }, + { + "epoch": 0.3219814241486068, + "grad_norm": 0.8322079367528697, + "learning_rate": 0.0007928095316701483, + "loss": 2.416, + "step": 702 + }, + { + "epoch": 0.3224400871459695, + "grad_norm": 0.7838036977158092, + "learning_rate": 0.0007922069066986819, + "loss": 2.4494, + "step": 703 + }, + { + "epoch": 0.3228987501433322, + "grad_norm": 0.7848023087914686, + "learning_rate": 0.0007916036363997097, + "loss": 2.4098, + "step": 704 + }, + { + "epoch": 0.3233574131406949, + "grad_norm": 0.8013589481891592, + "learning_rate": 0.0007909997221055308, + "loss": 2.4782, + "step": 705 + }, + { + "epoch": 0.3238160761380576, + "grad_norm": 0.7302966058096704, + "learning_rate": 0.0007903951651498658, + "loss": 2.3916, + "step": 706 + }, + { + "epoch": 0.32427473913542026, + "grad_norm": 0.7858848438780899, + "learning_rate": 0.0007897899668678557, + "loss": 2.37, + "step": 707 + }, + { + "epoch": 0.32473340213278296, + "grad_norm": 0.804026379596347, + "learning_rate": 0.0007891841285960566, + "loss": 2.4367, + "step": 708 + }, + { + "epoch": 0.32519206513014565, + "grad_norm": 0.8334837914578676, + "learning_rate": 0.0007885776516724388, + "loss": 2.3938, + "step": 709 + }, + { + "epoch": 0.32565072812750834, + "grad_norm": 0.8038528843289264, + "learning_rate": 0.0007879705374363831, + "loss": 2.3759, + "step": 710 + }, + { + "epoch": 0.32610939112487103, + "grad_norm": 0.7598483470501433, + "learning_rate": 0.000787362787228677, + "loss": 2.418, + "step": 711 + }, + { + "epoch": 0.32656805412223366, + "grad_norm": 0.8065032246836125, + "learning_rate": 0.0007867544023915134, + "loss": 2.4807, + "step": 712 + }, + { + "epoch": 0.32702671711959636, + "grad_norm": 0.8237509549771336, + "learning_rate": 0.0007861453842684861, + "loss": 2.3219, + "step": 713 + }, + { + "epoch": 0.32748538011695905, + "grad_norm": 0.7481779164577824, + "learning_rate": 0.0007855357342045882, + "loss": 2.4094, + "step": 714 + }, + { + "epoch": 0.32794404311432174, + "grad_norm": 0.7961972794055875, + "learning_rate": 0.0007849254535462074, + "loss": 2.3419, + "step": 715 + }, + { + "epoch": 0.32840270611168443, + "grad_norm": 0.8090024288451333, + "learning_rate": 0.0007843145436411252, + "loss": 2.3792, + "step": 716 + }, + { + "epoch": 0.3288613691090471, + "grad_norm": 0.8043223331736598, + "learning_rate": 0.0007837030058385117, + "loss": 2.3668, + "step": 717 + }, + { + "epoch": 0.3293200321064098, + "grad_norm": 0.7980886677020386, + "learning_rate": 0.0007830908414889246, + "loss": 2.5353, + "step": 718 + }, + { + "epoch": 0.3297786951037725, + "grad_norm": 0.7888641309603991, + "learning_rate": 0.0007824780519443046, + "loss": 2.442, + "step": 719 + }, + { + "epoch": 0.3302373581011352, + "grad_norm": 0.7572509995003309, + "learning_rate": 0.0007818646385579735, + "loss": 2.3433, + "step": 720 + }, + { + "epoch": 0.3306960210984979, + "grad_norm": 0.770528181362742, + "learning_rate": 0.0007812506026846307, + "loss": 2.417, + "step": 721 + }, + { + "epoch": 0.3311546840958606, + "grad_norm": 0.9566673867895013, + "learning_rate": 0.0007806359456803504, + "loss": 2.4715, + "step": 722 + }, + { + "epoch": 0.33161334709322327, + "grad_norm": 0.7875119042651693, + "learning_rate": 0.0007800206689025785, + "loss": 2.4102, + "step": 723 + }, + { + "epoch": 0.33207201009058596, + "grad_norm": 0.8050098379639888, + "learning_rate": 0.0007794047737101297, + "loss": 2.476, + "step": 724 + }, + { + "epoch": 0.33253067308794865, + "grad_norm": 0.8077092425893975, + "learning_rate": 0.0007787882614631843, + "loss": 2.4146, + "step": 725 + }, + { + "epoch": 0.33298933608531134, + "grad_norm": 0.7369356532302926, + "learning_rate": 0.0007781711335232856, + "loss": 2.3496, + "step": 726 + }, + { + "epoch": 0.33344799908267403, + "grad_norm": 0.7679964700685465, + "learning_rate": 0.0007775533912533363, + "loss": 2.4313, + "step": 727 + }, + { + "epoch": 0.3339066620800367, + "grad_norm": 0.8429767638460902, + "learning_rate": 0.0007769350360175962, + "loss": 2.535, + "step": 728 + }, + { + "epoch": 0.33436532507739936, + "grad_norm": 0.7779204900249019, + "learning_rate": 0.0007763160691816784, + "loss": 2.4559, + "step": 729 + }, + { + "epoch": 0.33482398807476205, + "grad_norm": 0.7461506650991777, + "learning_rate": 0.000775696492112547, + "loss": 2.3066, + "step": 730 + }, + { + "epoch": 0.33528265107212474, + "grad_norm": 0.884287519650087, + "learning_rate": 0.0007750763061785137, + "loss": 2.3861, + "step": 731 + }, + { + "epoch": 0.33574131406948743, + "grad_norm": 0.7658880921783572, + "learning_rate": 0.000774455512749235, + "loss": 2.432, + "step": 732 + }, + { + "epoch": 0.3361999770668501, + "grad_norm": 0.774746354007081, + "learning_rate": 0.0007738341131957085, + "loss": 2.3763, + "step": 733 + }, + { + "epoch": 0.3366586400642128, + "grad_norm": 0.7886037136471208, + "learning_rate": 0.000773212108890271, + "loss": 2.4106, + "step": 734 + }, + { + "epoch": 0.3371173030615755, + "grad_norm": 0.9011616018114815, + "learning_rate": 0.0007725895012065947, + "loss": 2.2835, + "step": 735 + }, + { + "epoch": 0.3375759660589382, + "grad_norm": 0.7740863129314842, + "learning_rate": 0.0007719662915196844, + "loss": 2.4016, + "step": 736 + }, + { + "epoch": 0.3380346290563009, + "grad_norm": 0.7617759374824994, + "learning_rate": 0.0007713424812058736, + "loss": 2.3697, + "step": 737 + }, + { + "epoch": 0.3384932920536636, + "grad_norm": 0.9004278201291823, + "learning_rate": 0.0007707180716428237, + "loss": 2.4549, + "step": 738 + }, + { + "epoch": 0.33895195505102627, + "grad_norm": 0.8078207421680591, + "learning_rate": 0.0007700930642095184, + "loss": 2.4264, + "step": 739 + }, + { + "epoch": 0.33941061804838896, + "grad_norm": 0.8567122764010793, + "learning_rate": 0.0007694674602862621, + "loss": 2.3859, + "step": 740 + }, + { + "epoch": 0.33986928104575165, + "grad_norm": 0.8776383221736019, + "learning_rate": 0.0007688412612546769, + "loss": 2.438, + "step": 741 + }, + { + "epoch": 0.34032794404311434, + "grad_norm": 0.8320259778146196, + "learning_rate": 0.0007682144684976983, + "loss": 2.3281, + "step": 742 + }, + { + "epoch": 0.34078660704047703, + "grad_norm": 0.8304678048445581, + "learning_rate": 0.0007675870833995739, + "loss": 2.3282, + "step": 743 + }, + { + "epoch": 0.3412452700378397, + "grad_norm": 0.7622550270000481, + "learning_rate": 0.0007669591073458592, + "loss": 2.4813, + "step": 744 + }, + { + "epoch": 0.3417039330352024, + "grad_norm": 0.7690085106589749, + "learning_rate": 0.0007663305417234146, + "loss": 2.3414, + "step": 745 + }, + { + "epoch": 0.34216259603256505, + "grad_norm": 0.8528661247855669, + "learning_rate": 0.0007657013879204022, + "loss": 2.3683, + "step": 746 + }, + { + "epoch": 0.34262125902992774, + "grad_norm": 0.7830386940259885, + "learning_rate": 0.0007650716473262842, + "loss": 2.2991, + "step": 747 + }, + { + "epoch": 0.34307992202729043, + "grad_norm": 0.7896118378394325, + "learning_rate": 0.0007644413213318177, + "loss": 2.3847, + "step": 748 + }, + { + "epoch": 0.3435385850246531, + "grad_norm": 0.8201027859997383, + "learning_rate": 0.0007638104113290531, + "loss": 2.3792, + "step": 749 + }, + { + "epoch": 0.3439972480220158, + "grad_norm": 0.8702948772367166, + "learning_rate": 0.0007631789187113303, + "loss": 2.384, + "step": 750 + }, + { + "epoch": 0.3444559110193785, + "grad_norm": 0.7637855675841474, + "learning_rate": 0.000762546844873276, + "loss": 2.3662, + "step": 751 + }, + { + "epoch": 0.3449145740167412, + "grad_norm": 0.7683984908541728, + "learning_rate": 0.0007619141912108007, + "loss": 2.5102, + "step": 752 + }, + { + "epoch": 0.3453732370141039, + "grad_norm": 0.9765488722828396, + "learning_rate": 0.000761280959121095, + "loss": 2.4065, + "step": 753 + }, + { + "epoch": 0.3458319000114666, + "grad_norm": 0.7145811460129042, + "learning_rate": 0.0007606471500026273, + "loss": 2.4427, + "step": 754 + }, + { + "epoch": 0.34629056300882927, + "grad_norm": 0.8631253657234366, + "learning_rate": 0.0007600127652551401, + "loss": 2.3445, + "step": 755 + }, + { + "epoch": 0.34674922600619196, + "grad_norm": 0.800288555678986, + "learning_rate": 0.0007593778062796472, + "loss": 2.4157, + "step": 756 + }, + { + "epoch": 0.34720788900355465, + "grad_norm": 0.7482927500559754, + "learning_rate": 0.000758742274478431, + "loss": 2.5175, + "step": 757 + }, + { + "epoch": 0.34766655200091734, + "grad_norm": 0.8656581267283838, + "learning_rate": 0.0007581061712550381, + "loss": 2.3857, + "step": 758 + }, + { + "epoch": 0.34812521499828003, + "grad_norm": 0.7990967725425315, + "learning_rate": 0.0007574694980142779, + "loss": 2.3874, + "step": 759 + }, + { + "epoch": 0.3485838779956427, + "grad_norm": 0.7389554738524283, + "learning_rate": 0.0007568322561622183, + "loss": 2.4099, + "step": 760 + }, + { + "epoch": 0.3490425409930054, + "grad_norm": 0.7426801376433322, + "learning_rate": 0.0007561944471061826, + "loss": 2.3755, + "step": 761 + }, + { + "epoch": 0.3495012039903681, + "grad_norm": 0.7333298783457516, + "learning_rate": 0.0007555560722547475, + "loss": 2.2828, + "step": 762 + }, + { + "epoch": 0.34995986698773074, + "grad_norm": 0.7244599784981033, + "learning_rate": 0.0007549171330177387, + "loss": 2.5393, + "step": 763 + }, + { + "epoch": 0.35041852998509343, + "grad_norm": 0.7411579657894272, + "learning_rate": 0.0007542776308062285, + "loss": 2.3902, + "step": 764 + }, + { + "epoch": 0.3508771929824561, + "grad_norm": 0.805499806960988, + "learning_rate": 0.0007536375670325325, + "loss": 2.3271, + "step": 765 + }, + { + "epoch": 0.3513358559798188, + "grad_norm": 0.7246501299733149, + "learning_rate": 0.0007529969431102063, + "loss": 2.438, + "step": 766 + }, + { + "epoch": 0.3517945189771815, + "grad_norm": 0.7292301594953408, + "learning_rate": 0.000752355760454043, + "loss": 2.4482, + "step": 767 + }, + { + "epoch": 0.3522531819745442, + "grad_norm": 0.8221263449081223, + "learning_rate": 0.0007517140204800693, + "loss": 2.445, + "step": 768 + }, + { + "epoch": 0.3527118449719069, + "grad_norm": 0.8204383612779542, + "learning_rate": 0.0007510717246055425, + "loss": 2.4381, + "step": 769 + }, + { + "epoch": 0.3531705079692696, + "grad_norm": 0.8165349639606174, + "learning_rate": 0.0007504288742489482, + "loss": 2.3873, + "step": 770 + }, + { + "epoch": 0.35362917096663227, + "grad_norm": 0.7915571278257483, + "learning_rate": 0.0007497854708299963, + "loss": 2.3548, + "step": 771 + }, + { + "epoch": 0.35408783396399496, + "grad_norm": 0.7337302679607395, + "learning_rate": 0.0007491415157696178, + "loss": 2.3466, + "step": 772 + }, + { + "epoch": 0.35454649696135765, + "grad_norm": 0.7935846289597412, + "learning_rate": 0.0007484970104899623, + "loss": 2.4, + "step": 773 + }, + { + "epoch": 0.35500515995872034, + "grad_norm": 0.7325951799142862, + "learning_rate": 0.0007478519564143945, + "loss": 2.4013, + "step": 774 + }, + { + "epoch": 0.35546382295608303, + "grad_norm": 0.7658761978977395, + "learning_rate": 0.000747206354967491, + "loss": 2.3804, + "step": 775 + }, + { + "epoch": 0.3559224859534457, + "grad_norm": 0.8450662781403074, + "learning_rate": 0.0007465602075750373, + "loss": 2.3484, + "step": 776 + }, + { + "epoch": 0.3563811489508084, + "grad_norm": 0.7453112308929764, + "learning_rate": 0.0007459135156640247, + "loss": 2.4155, + "step": 777 + }, + { + "epoch": 0.3568398119481711, + "grad_norm": 0.783869498188983, + "learning_rate": 0.0007452662806626468, + "loss": 2.3879, + "step": 778 + }, + { + "epoch": 0.3572984749455338, + "grad_norm": 0.7815216898892576, + "learning_rate": 0.0007446185040002967, + "loss": 2.269, + "step": 779 + }, + { + "epoch": 0.35775713794289643, + "grad_norm": 1.051357023838775, + "learning_rate": 0.0007439701871075642, + "loss": 2.404, + "step": 780 + }, + { + "epoch": 0.3582158009402591, + "grad_norm": 0.7570378809191287, + "learning_rate": 0.0007433213314162313, + "loss": 2.3792, + "step": 781 + }, + { + "epoch": 0.3586744639376218, + "grad_norm": 0.8186237498416177, + "learning_rate": 0.0007426719383592705, + "loss": 2.3559, + "step": 782 + }, + { + "epoch": 0.3591331269349845, + "grad_norm": 0.8042349213950369, + "learning_rate": 0.000742022009370841, + "loss": 2.2803, + "step": 783 + }, + { + "epoch": 0.3595917899323472, + "grad_norm": 0.7719654890482284, + "learning_rate": 0.0007413715458862855, + "loss": 2.4273, + "step": 784 + }, + { + "epoch": 0.3600504529297099, + "grad_norm": 0.7446471615288832, + "learning_rate": 0.0007407205493421272, + "loss": 2.3736, + "step": 785 + }, + { + "epoch": 0.3605091159270726, + "grad_norm": 0.7571028028809516, + "learning_rate": 0.0007400690211760661, + "loss": 2.3098, + "step": 786 + }, + { + "epoch": 0.36096777892443527, + "grad_norm": 0.8781030280958132, + "learning_rate": 0.0007394169628269771, + "loss": 2.2917, + "step": 787 + }, + { + "epoch": 0.36142644192179796, + "grad_norm": 0.7832410086502902, + "learning_rate": 0.0007387643757349051, + "loss": 2.3156, + "step": 788 + }, + { + "epoch": 0.36188510491916065, + "grad_norm": 0.8823324210102942, + "learning_rate": 0.0007381112613410635, + "loss": 2.3267, + "step": 789 + }, + { + "epoch": 0.36234376791652334, + "grad_norm": 0.7934092377111689, + "learning_rate": 0.0007374576210878298, + "loss": 2.5148, + "step": 790 + }, + { + "epoch": 0.36280243091388603, + "grad_norm": 0.7704628752741737, + "learning_rate": 0.0007368034564187425, + "loss": 2.3876, + "step": 791 + }, + { + "epoch": 0.3632610939112487, + "grad_norm": 0.7827195098607439, + "learning_rate": 0.0007361487687784989, + "loss": 2.4504, + "step": 792 + }, + { + "epoch": 0.3637197569086114, + "grad_norm": 0.7804901212227076, + "learning_rate": 0.0007354935596129513, + "loss": 2.4721, + "step": 793 + }, + { + "epoch": 0.3641784199059741, + "grad_norm": 0.7604838984013046, + "learning_rate": 0.000734837830369103, + "loss": 2.3431, + "step": 794 + }, + { + "epoch": 0.3646370829033368, + "grad_norm": 0.7450809758475156, + "learning_rate": 0.0007341815824951066, + "loss": 2.3465, + "step": 795 + }, + { + "epoch": 0.3650957459006995, + "grad_norm": 0.7035014138521629, + "learning_rate": 0.0007335248174402597, + "loss": 2.3389, + "step": 796 + }, + { + "epoch": 0.3655544088980621, + "grad_norm": 0.6956048895596296, + "learning_rate": 0.0007328675366550023, + "loss": 2.3139, + "step": 797 + }, + { + "epoch": 0.3660130718954248, + "grad_norm": 0.8107958701690775, + "learning_rate": 0.0007322097415909134, + "loss": 2.3742, + "step": 798 + }, + { + "epoch": 0.3664717348927875, + "grad_norm": 0.7740462701810895, + "learning_rate": 0.0007315514337007071, + "loss": 2.3226, + "step": 799 + }, + { + "epoch": 0.3669303978901502, + "grad_norm": 0.8336348955170941, + "learning_rate": 0.0007308926144382312, + "loss": 2.3962, + "step": 800 + }, + { + "epoch": 0.3673890608875129, + "grad_norm": 0.7947791204846696, + "learning_rate": 0.0007302332852584619, + "loss": 2.3718, + "step": 801 + }, + { + "epoch": 0.3678477238848756, + "grad_norm": 0.851509771397248, + "learning_rate": 0.0007295734476175018, + "loss": 2.4083, + "step": 802 + }, + { + "epoch": 0.36830638688223827, + "grad_norm": 0.8132888944277684, + "learning_rate": 0.0007289131029725768, + "loss": 2.3512, + "step": 803 + }, + { + "epoch": 0.36876504987960096, + "grad_norm": 1.2608511500202617, + "learning_rate": 0.0007282522527820319, + "loss": 2.3537, + "step": 804 + }, + { + "epoch": 0.36922371287696365, + "grad_norm": 0.8795263284613015, + "learning_rate": 0.000727590898505329, + "loss": 2.3688, + "step": 805 + }, + { + "epoch": 0.36968237587432634, + "grad_norm": 0.7829316551545304, + "learning_rate": 0.0007269290416030429, + "loss": 2.3259, + "step": 806 + }, + { + "epoch": 0.37014103887168903, + "grad_norm": 0.7520754507307954, + "learning_rate": 0.000726266683536859, + "loss": 2.3196, + "step": 807 + }, + { + "epoch": 0.3705997018690517, + "grad_norm": 0.7758014189625201, + "learning_rate": 0.0007256038257695687, + "loss": 2.3565, + "step": 808 + }, + { + "epoch": 0.3710583648664144, + "grad_norm": 0.7957527265870592, + "learning_rate": 0.0007249404697650678, + "loss": 2.3425, + "step": 809 + }, + { + "epoch": 0.3715170278637771, + "grad_norm": 0.7338150149010338, + "learning_rate": 0.0007242766169883518, + "loss": 2.3828, + "step": 810 + }, + { + "epoch": 0.3719756908611398, + "grad_norm": 0.8013839686657119, + "learning_rate": 0.0007236122689055138, + "loss": 2.378, + "step": 811 + }, + { + "epoch": 0.3724343538585025, + "grad_norm": 0.7986991145145933, + "learning_rate": 0.0007229474269837401, + "loss": 2.4582, + "step": 812 + }, + { + "epoch": 0.3728930168558652, + "grad_norm": 0.8861858116969749, + "learning_rate": 0.0007222820926913085, + "loss": 2.3363, + "step": 813 + }, + { + "epoch": 0.3733516798532278, + "grad_norm": 0.7625939513279882, + "learning_rate": 0.0007216162674975833, + "loss": 2.4212, + "step": 814 + }, + { + "epoch": 0.3738103428505905, + "grad_norm": 0.7294789253720322, + "learning_rate": 0.0007209499528730138, + "loss": 2.3737, + "step": 815 + }, + { + "epoch": 0.3742690058479532, + "grad_norm": 0.9124957742858358, + "learning_rate": 0.0007202831502891294, + "loss": 2.3381, + "step": 816 + }, + { + "epoch": 0.3747276688453159, + "grad_norm": 0.7767273521327148, + "learning_rate": 0.0007196158612185375, + "loss": 2.4828, + "step": 817 + }, + { + "epoch": 0.3751863318426786, + "grad_norm": 0.7455208941784632, + "learning_rate": 0.0007189480871349201, + "loss": 2.2826, + "step": 818 + }, + { + "epoch": 0.37564499484004127, + "grad_norm": 0.8428303008821749, + "learning_rate": 0.0007182798295130299, + "loss": 2.3169, + "step": 819 + }, + { + "epoch": 0.37610365783740396, + "grad_norm": 0.8252355060094554, + "learning_rate": 0.0007176110898286878, + "loss": 2.3863, + "step": 820 + }, + { + "epoch": 0.37656232083476665, + "grad_norm": 0.7535335218503296, + "learning_rate": 0.0007169418695587791, + "loss": 2.35, + "step": 821 + }, + { + "epoch": 0.37702098383212934, + "grad_norm": 0.77694062440927, + "learning_rate": 0.0007162721701812506, + "loss": 2.2691, + "step": 822 + }, + { + "epoch": 0.37747964682949203, + "grad_norm": 0.7785240582806469, + "learning_rate": 0.0007156019931751072, + "loss": 2.2746, + "step": 823 + }, + { + "epoch": 0.3779383098268547, + "grad_norm": 0.7955520757274679, + "learning_rate": 0.0007149313400204082, + "loss": 2.368, + "step": 824 + }, + { + "epoch": 0.3783969728242174, + "grad_norm": 0.8254631297799135, + "learning_rate": 0.0007142602121982653, + "loss": 2.4442, + "step": 825 + }, + { + "epoch": 0.3788556358215801, + "grad_norm": 0.7843112521847936, + "learning_rate": 0.0007135886111908379, + "loss": 2.4232, + "step": 826 + }, + { + "epoch": 0.3793142988189428, + "grad_norm": 0.958699757654021, + "learning_rate": 0.0007129165384813303, + "loss": 2.3372, + "step": 827 + }, + { + "epoch": 0.3797729618163055, + "grad_norm": 0.7558246273485991, + "learning_rate": 0.0007122439955539888, + "loss": 2.3977, + "step": 828 + }, + { + "epoch": 0.3802316248136682, + "grad_norm": 0.7050608445680318, + "learning_rate": 0.0007115709838940983, + "loss": 2.466, + "step": 829 + }, + { + "epoch": 0.38069028781103087, + "grad_norm": 0.7502501382842127, + "learning_rate": 0.0007108975049879785, + "loss": 2.3347, + "step": 830 + }, + { + "epoch": 0.3811489508083935, + "grad_norm": 0.8964929323716384, + "learning_rate": 0.0007102235603229814, + "loss": 2.3871, + "step": 831 + }, + { + "epoch": 0.3816076138057562, + "grad_norm": 0.7195172375175003, + "learning_rate": 0.000709549151387487, + "loss": 2.468, + "step": 832 + }, + { + "epoch": 0.3820662768031189, + "grad_norm": 0.7555192063726833, + "learning_rate": 0.0007088742796709013, + "loss": 2.4027, + "step": 833 + }, + { + "epoch": 0.3825249398004816, + "grad_norm": 0.7028346402025711, + "learning_rate": 0.000708198946663652, + "loss": 2.4059, + "step": 834 + }, + { + "epoch": 0.38298360279784427, + "grad_norm": 0.7082355088073351, + "learning_rate": 0.0007075231538571856, + "loss": 2.3995, + "step": 835 + }, + { + "epoch": 0.38344226579520696, + "grad_norm": 0.7000145433159809, + "learning_rate": 0.0007068469027439641, + "loss": 2.4305, + "step": 836 + }, + { + "epoch": 0.38390092879256965, + "grad_norm": 0.7631362872195674, + "learning_rate": 0.0007061701948174613, + "loss": 2.3685, + "step": 837 + }, + { + "epoch": 0.38435959178993234, + "grad_norm": 0.685455221547008, + "learning_rate": 0.0007054930315721606, + "loss": 2.2443, + "step": 838 + }, + { + "epoch": 0.38481825478729503, + "grad_norm": 0.7688414565081579, + "learning_rate": 0.0007048154145035501, + "loss": 2.3078, + "step": 839 + }, + { + "epoch": 0.3852769177846577, + "grad_norm": 0.8248617341979392, + "learning_rate": 0.0007041373451081207, + "loss": 2.3403, + "step": 840 + }, + { + "epoch": 0.3857355807820204, + "grad_norm": 0.7798598304740172, + "learning_rate": 0.0007034588248833621, + "loss": 2.3229, + "step": 841 + }, + { + "epoch": 0.3861942437793831, + "grad_norm": 0.7934116818508113, + "learning_rate": 0.0007027798553277595, + "loss": 2.4643, + "step": 842 + }, + { + "epoch": 0.3866529067767458, + "grad_norm": 1.1708474527468427, + "learning_rate": 0.0007021004379407909, + "loss": 2.3837, + "step": 843 + }, + { + "epoch": 0.3871115697741085, + "grad_norm": 0.7767017232391452, + "learning_rate": 0.0007014205742229227, + "loss": 2.3375, + "step": 844 + }, + { + "epoch": 0.3875702327714712, + "grad_norm": 0.8149641870992606, + "learning_rate": 0.0007007402656756072, + "loss": 2.3692, + "step": 845 + }, + { + "epoch": 0.38802889576883387, + "grad_norm": 3.014915144740135, + "learning_rate": 0.0007000595138012797, + "loss": 2.4354, + "step": 846 + }, + { + "epoch": 0.38848755876619656, + "grad_norm": 0.7966434390924643, + "learning_rate": 0.0006993783201033535, + "loss": 2.4273, + "step": 847 + }, + { + "epoch": 0.38894622176355925, + "grad_norm": 0.7966016817245629, + "learning_rate": 0.0006986966860862182, + "loss": 2.3729, + "step": 848 + }, + { + "epoch": 0.3894048847609219, + "grad_norm": 0.8009283725712834, + "learning_rate": 0.000698014613255236, + "loss": 2.3345, + "step": 849 + }, + { + "epoch": 0.3898635477582846, + "grad_norm": 0.7244175106464799, + "learning_rate": 0.0006973321031167382, + "loss": 2.3316, + "step": 850 + }, + { + "epoch": 0.39032221075564727, + "grad_norm": 0.8337185012109274, + "learning_rate": 0.0006966491571780216, + "loss": 2.4512, + "step": 851 + }, + { + "epoch": 0.39078087375300996, + "grad_norm": 0.8698374980115159, + "learning_rate": 0.0006959657769473453, + "loss": 2.297, + "step": 852 + }, + { + "epoch": 0.39123953675037265, + "grad_norm": 0.7225323132904605, + "learning_rate": 0.000695281963933928, + "loss": 2.4174, + "step": 853 + }, + { + "epoch": 0.39169819974773534, + "grad_norm": 0.7455395798303737, + "learning_rate": 0.0006945977196479438, + "loss": 2.3993, + "step": 854 + }, + { + "epoch": 0.39215686274509803, + "grad_norm": 0.7496422281814061, + "learning_rate": 0.0006939130456005196, + "loss": 2.4792, + "step": 855 + }, + { + "epoch": 0.3926155257424607, + "grad_norm": 0.7666101814243061, + "learning_rate": 0.0006932279433037311, + "loss": 2.3688, + "step": 856 + }, + { + "epoch": 0.3930741887398234, + "grad_norm": 0.6974811036624291, + "learning_rate": 0.0006925424142705997, + "loss": 2.3556, + "step": 857 + }, + { + "epoch": 0.3935328517371861, + "grad_norm": 0.7299158157116274, + "learning_rate": 0.0006918564600150896, + "loss": 2.3455, + "step": 858 + }, + { + "epoch": 0.3939915147345488, + "grad_norm": 0.7585918793242484, + "learning_rate": 0.0006911700820521042, + "loss": 2.3468, + "step": 859 + }, + { + "epoch": 0.3944501777319115, + "grad_norm": 0.7472145462820317, + "learning_rate": 0.0006904832818974818, + "loss": 2.4254, + "step": 860 + }, + { + "epoch": 0.3949088407292742, + "grad_norm": 0.7238042863292762, + "learning_rate": 0.0006897960610679939, + "loss": 2.5144, + "step": 861 + }, + { + "epoch": 0.39536750372663687, + "grad_norm": 0.7001289743792715, + "learning_rate": 0.0006891084210813407, + "loss": 2.4111, + "step": 862 + }, + { + "epoch": 0.39582616672399956, + "grad_norm": 0.8138117482168981, + "learning_rate": 0.0006884203634561483, + "loss": 2.3906, + "step": 863 + }, + { + "epoch": 0.39628482972136225, + "grad_norm": 0.6772987552687569, + "learning_rate": 0.0006877318897119651, + "loss": 2.3458, + "step": 864 + }, + { + "epoch": 0.39674349271872494, + "grad_norm": 0.758487931251777, + "learning_rate": 0.0006870430013692579, + "loss": 2.3379, + "step": 865 + }, + { + "epoch": 0.3972021557160876, + "grad_norm": 0.7146880570766357, + "learning_rate": 0.0006863536999494101, + "loss": 2.368, + "step": 866 + }, + { + "epoch": 0.39766081871345027, + "grad_norm": 0.7674510285845524, + "learning_rate": 0.0006856639869747167, + "loss": 2.4376, + "step": 867 + }, + { + "epoch": 0.39811948171081296, + "grad_norm": 0.8110957934503258, + "learning_rate": 0.0006849738639683818, + "loss": 2.3329, + "step": 868 + }, + { + "epoch": 0.39857814470817565, + "grad_norm": 0.7510639745424744, + "learning_rate": 0.000684283332454515, + "loss": 2.3619, + "step": 869 + }, + { + "epoch": 0.39903680770553834, + "grad_norm": 0.9483303348244958, + "learning_rate": 0.0006835923939581281, + "loss": 2.3379, + "step": 870 + }, + { + "epoch": 0.39949547070290103, + "grad_norm": 0.8260536243737734, + "learning_rate": 0.0006829010500051318, + "loss": 2.4276, + "step": 871 + }, + { + "epoch": 0.3999541337002637, + "grad_norm": 0.805180008102778, + "learning_rate": 0.0006822093021223321, + "loss": 2.3681, + "step": 872 + }, + { + "epoch": 0.4004127966976264, + "grad_norm": 0.768908628884231, + "learning_rate": 0.0006815171518374268, + "loss": 2.3862, + "step": 873 + }, + { + "epoch": 0.4008714596949891, + "grad_norm": 0.7494532422195165, + "learning_rate": 0.0006808246006790031, + "loss": 2.3465, + "step": 874 + }, + { + "epoch": 0.4013301226923518, + "grad_norm": 0.7899372052134073, + "learning_rate": 0.0006801316501765329, + "loss": 2.3768, + "step": 875 + }, + { + "epoch": 0.4017887856897145, + "grad_norm": 0.7234789454037845, + "learning_rate": 0.0006794383018603704, + "loss": 2.4408, + "step": 876 + }, + { + "epoch": 0.4022474486870772, + "grad_norm": 0.7747384440233951, + "learning_rate": 0.0006787445572617481, + "loss": 2.4307, + "step": 877 + }, + { + "epoch": 0.40270611168443987, + "grad_norm": 0.6886158963204584, + "learning_rate": 0.0006780504179127734, + "loss": 2.3871, + "step": 878 + }, + { + "epoch": 0.40316477468180256, + "grad_norm": 0.7912887052156912, + "learning_rate": 0.0006773558853464265, + "loss": 2.38, + "step": 879 + }, + { + "epoch": 0.40362343767916525, + "grad_norm": 0.6769011887454174, + "learning_rate": 0.000676660961096555, + "loss": 2.3349, + "step": 880 + }, + { + "epoch": 0.40408210067652794, + "grad_norm": 0.7494139630143282, + "learning_rate": 0.000675965646697872, + "loss": 2.3684, + "step": 881 + }, + { + "epoch": 0.40454076367389064, + "grad_norm": 0.7233910427646464, + "learning_rate": 0.0006752699436859519, + "loss": 2.3381, + "step": 882 + }, + { + "epoch": 0.40499942667125327, + "grad_norm": 0.7135139371800977, + "learning_rate": 0.0006745738535972279, + "loss": 2.2963, + "step": 883 + }, + { + "epoch": 0.40545808966861596, + "grad_norm": 0.8137029421081956, + "learning_rate": 0.0006738773779689874, + "loss": 2.3543, + "step": 884 + }, + { + "epoch": 0.40591675266597865, + "grad_norm": 0.7336540217565746, + "learning_rate": 0.0006731805183393696, + "loss": 2.3361, + "step": 885 + }, + { + "epoch": 0.40637541566334134, + "grad_norm": 0.8135218454863882, + "learning_rate": 0.0006724832762473618, + "loss": 2.358, + "step": 886 + }, + { + "epoch": 0.40683407866070403, + "grad_norm": 0.8315450064279764, + "learning_rate": 0.0006717856532327956, + "loss": 2.3404, + "step": 887 + }, + { + "epoch": 0.4072927416580667, + "grad_norm": 0.7454651021856797, + "learning_rate": 0.0006710876508363444, + "loss": 2.3467, + "step": 888 + }, + { + "epoch": 0.4077514046554294, + "grad_norm": 0.841990766977827, + "learning_rate": 0.0006703892705995189, + "loss": 2.3912, + "step": 889 + }, + { + "epoch": 0.4082100676527921, + "grad_norm": 0.7300181250032899, + "learning_rate": 0.0006696905140646647, + "loss": 2.4073, + "step": 890 + }, + { + "epoch": 0.4086687306501548, + "grad_norm": 0.7074394300302365, + "learning_rate": 0.0006689913827749581, + "loss": 2.3451, + "step": 891 + }, + { + "epoch": 0.4091273936475175, + "grad_norm": 0.749342015209368, + "learning_rate": 0.0006682918782744032, + "loss": 2.3661, + "step": 892 + }, + { + "epoch": 0.4095860566448802, + "grad_norm": 0.8767111112488528, + "learning_rate": 0.0006675920021078282, + "loss": 2.1809, + "step": 893 + }, + { + "epoch": 0.41004471964224287, + "grad_norm": 0.739830704451208, + "learning_rate": 0.0006668917558208823, + "loss": 2.5216, + "step": 894 + }, + { + "epoch": 0.41050338263960556, + "grad_norm": 0.7974038580484438, + "learning_rate": 0.0006661911409600321, + "loss": 2.3456, + "step": 895 + }, + { + "epoch": 0.41096204563696825, + "grad_norm": 0.7616529245833997, + "learning_rate": 0.0006654901590725577, + "loss": 2.3687, + "step": 896 + }, + { + "epoch": 0.41142070863433094, + "grad_norm": 0.8965929271225344, + "learning_rate": 0.0006647888117065507, + "loss": 2.3892, + "step": 897 + }, + { + "epoch": 0.41187937163169364, + "grad_norm": 0.7582198521251975, + "learning_rate": 0.0006640871004109086, + "loss": 2.3282, + "step": 898 + }, + { + "epoch": 0.4123380346290563, + "grad_norm": 0.7727105737266937, + "learning_rate": 0.000663385026735334, + "loss": 2.3246, + "step": 899 + }, + { + "epoch": 0.41279669762641896, + "grad_norm": 0.7433133055637263, + "learning_rate": 0.0006626825922303285, + "loss": 2.4088, + "step": 900 + }, + { + "epoch": 0.41325536062378165, + "grad_norm": 0.7205701501111299, + "learning_rate": 0.0006619797984471915, + "loss": 2.3894, + "step": 901 + }, + { + "epoch": 0.41371402362114434, + "grad_norm": 0.7113472915792617, + "learning_rate": 0.0006612766469380158, + "loss": 2.3683, + "step": 902 + }, + { + "epoch": 0.41417268661850704, + "grad_norm": 0.7692626301855585, + "learning_rate": 0.0006605731392556833, + "loss": 2.5313, + "step": 903 + }, + { + "epoch": 0.4146313496158697, + "grad_norm": 0.7536004485694296, + "learning_rate": 0.0006598692769538637, + "loss": 2.3742, + "step": 904 + }, + { + "epoch": 0.4150900126132324, + "grad_norm": 0.7013662930951287, + "learning_rate": 0.0006591650615870091, + "loss": 2.3867, + "step": 905 + }, + { + "epoch": 0.4155486756105951, + "grad_norm": 0.6922074084620502, + "learning_rate": 0.0006584604947103514, + "loss": 2.3985, + "step": 906 + }, + { + "epoch": 0.4160073386079578, + "grad_norm": 0.7359983385606577, + "learning_rate": 0.0006577555778798993, + "loss": 2.3543, + "step": 907 + }, + { + "epoch": 0.4164660016053205, + "grad_norm": 0.7934607085201227, + "learning_rate": 0.0006570503126524336, + "loss": 2.332, + "step": 908 + }, + { + "epoch": 0.4169246646026832, + "grad_norm": 0.6868356499707018, + "learning_rate": 0.0006563447005855054, + "loss": 2.3526, + "step": 909 + }, + { + "epoch": 0.4173833276000459, + "grad_norm": 0.7827295451592374, + "learning_rate": 0.000655638743237431, + "loss": 2.3866, + "step": 910 + }, + { + "epoch": 0.41784199059740856, + "grad_norm": 0.748051284122138, + "learning_rate": 0.0006549324421672894, + "loss": 2.4146, + "step": 911 + }, + { + "epoch": 0.41830065359477125, + "grad_norm": 0.9268066979893296, + "learning_rate": 0.0006542257989349194, + "loss": 2.3674, + "step": 912 + }, + { + "epoch": 0.41875931659213395, + "grad_norm": 0.7205615096839769, + "learning_rate": 0.0006535188151009142, + "loss": 2.3753, + "step": 913 + }, + { + "epoch": 0.41921797958949664, + "grad_norm": 0.7493710328135269, + "learning_rate": 0.0006528114922266204, + "loss": 2.4321, + "step": 914 + }, + { + "epoch": 0.4196766425868593, + "grad_norm": 1.4550584754681894, + "learning_rate": 0.0006521038318741327, + "loss": 2.3824, + "step": 915 + }, + { + "epoch": 0.420135305584222, + "grad_norm": 0.7726876769187094, + "learning_rate": 0.0006513958356062912, + "loss": 2.414, + "step": 916 + }, + { + "epoch": 0.42059396858158465, + "grad_norm": 0.72069920909975, + "learning_rate": 0.0006506875049866781, + "loss": 2.3103, + "step": 917 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 0.7206240529675546, + "learning_rate": 0.0006499788415796137, + "loss": 2.3864, + "step": 918 + }, + { + "epoch": 0.42151129457631004, + "grad_norm": 0.7606574980198649, + "learning_rate": 0.0006492698469501532, + "loss": 2.3171, + "step": 919 + }, + { + "epoch": 0.4219699575736727, + "grad_norm": 0.7404164376419237, + "learning_rate": 0.0006485605226640837, + "loss": 2.3483, + "step": 920 + }, + { + "epoch": 0.4224286205710354, + "grad_norm": 0.9111550895199443, + "learning_rate": 0.00064785087028792, + "loss": 2.3893, + "step": 921 + }, + { + "epoch": 0.4228872835683981, + "grad_norm": 0.7625789180932001, + "learning_rate": 0.0006471408913889019, + "loss": 2.2904, + "step": 922 + }, + { + "epoch": 0.4233459465657608, + "grad_norm": 0.7534674333768265, + "learning_rate": 0.0006464305875349892, + "loss": 2.3316, + "step": 923 + }, + { + "epoch": 0.4238046095631235, + "grad_norm": 0.7180717477224986, + "learning_rate": 0.000645719960294861, + "loss": 2.3187, + "step": 924 + }, + { + "epoch": 0.4242632725604862, + "grad_norm": 0.7545390927784381, + "learning_rate": 0.0006450090112379092, + "loss": 2.2646, + "step": 925 + }, + { + "epoch": 0.4247219355578489, + "grad_norm": 0.7660971915408298, + "learning_rate": 0.0006442977419342371, + "loss": 2.2953, + "step": 926 + }, + { + "epoch": 0.42518059855521156, + "grad_norm": 0.6785709095341108, + "learning_rate": 0.000643586153954655, + "loss": 2.3514, + "step": 927 + }, + { + "epoch": 0.42563926155257426, + "grad_norm": 0.7792143329623858, + "learning_rate": 0.0006428742488706772, + "loss": 2.3269, + "step": 928 + }, + { + "epoch": 0.42609792454993695, + "grad_norm": 0.9179980120235869, + "learning_rate": 0.0006421620282545182, + "loss": 2.2814, + "step": 929 + }, + { + "epoch": 0.42655658754729964, + "grad_norm": 0.7144446519395787, + "learning_rate": 0.0006414494936790892, + "loss": 2.3607, + "step": 930 + }, + { + "epoch": 0.42701525054466233, + "grad_norm": 0.805339154825918, + "learning_rate": 0.0006407366467179951, + "loss": 2.3334, + "step": 931 + }, + { + "epoch": 0.427473913542025, + "grad_norm": 0.7455276911107238, + "learning_rate": 0.0006400234889455301, + "loss": 2.4167, + "step": 932 + }, + { + "epoch": 0.4279325765393877, + "grad_norm": 0.7261047745662765, + "learning_rate": 0.0006393100219366755, + "loss": 2.4112, + "step": 933 + }, + { + "epoch": 0.42839123953675035, + "grad_norm": 2.4440051583516738, + "learning_rate": 0.0006385962472670953, + "loss": 2.3034, + "step": 934 + }, + { + "epoch": 0.42884990253411304, + "grad_norm": 0.7202645730246002, + "learning_rate": 0.0006378821665131328, + "loss": 2.3469, + "step": 935 + }, + { + "epoch": 0.4293085655314757, + "grad_norm": 0.7353972352976533, + "learning_rate": 0.0006371677812518072, + "loss": 2.4299, + "step": 936 + }, + { + "epoch": 0.4297672285288384, + "grad_norm": 0.733226670375434, + "learning_rate": 0.0006364530930608107, + "loss": 2.3456, + "step": 937 + }, + { + "epoch": 0.4302258915262011, + "grad_norm": 0.7504820907530235, + "learning_rate": 0.0006357381035185038, + "loss": 2.3877, + "step": 938 + }, + { + "epoch": 0.4306845545235638, + "grad_norm": 0.7334596786259189, + "learning_rate": 0.0006350228142039131, + "loss": 2.4229, + "step": 939 + }, + { + "epoch": 0.4311432175209265, + "grad_norm": 0.8419571077700503, + "learning_rate": 0.000634307226696727, + "loss": 2.3876, + "step": 940 + }, + { + "epoch": 0.4316018805182892, + "grad_norm": 0.7670877795447945, + "learning_rate": 0.0006335913425772926, + "loss": 2.4456, + "step": 941 + }, + { + "epoch": 0.4320605435156519, + "grad_norm": 0.719130780167132, + "learning_rate": 0.0006328751634266117, + "loss": 2.2918, + "step": 942 + }, + { + "epoch": 0.43251920651301456, + "grad_norm": 0.8774706279944038, + "learning_rate": 0.0006321586908263382, + "loss": 2.3743, + "step": 943 + }, + { + "epoch": 0.43297786951037726, + "grad_norm": 0.802259181201537, + "learning_rate": 0.0006314419263587732, + "loss": 2.4087, + "step": 944 + }, + { + "epoch": 0.43343653250773995, + "grad_norm": 0.7789759366065465, + "learning_rate": 0.0006307248716068637, + "loss": 2.3601, + "step": 945 + }, + { + "epoch": 0.43389519550510264, + "grad_norm": 0.8523766536417178, + "learning_rate": 0.0006300075281541964, + "loss": 2.2752, + "step": 946 + }, + { + "epoch": 0.43435385850246533, + "grad_norm": 0.756942744189373, + "learning_rate": 0.0006292898975849966, + "loss": 2.5129, + "step": 947 + }, + { + "epoch": 0.434812521499828, + "grad_norm": 0.790475696449528, + "learning_rate": 0.000628571981484123, + "loss": 2.3671, + "step": 948 + }, + { + "epoch": 0.4352711844971907, + "grad_norm": 0.8999198805525266, + "learning_rate": 0.0006278537814370654, + "loss": 2.5329, + "step": 949 + }, + { + "epoch": 0.4357298474945534, + "grad_norm": 0.7263455128063303, + "learning_rate": 0.0006271352990299406, + "loss": 2.4014, + "step": 950 + }, + { + "epoch": 0.43618851049191604, + "grad_norm": 0.7542227483536568, + "learning_rate": 0.0006264165358494885, + "loss": 2.3469, + "step": 951 + }, + { + "epoch": 0.43664717348927873, + "grad_norm": 0.7454286270077304, + "learning_rate": 0.0006256974934830694, + "loss": 2.3995, + "step": 952 + }, + { + "epoch": 0.4371058364866414, + "grad_norm": 0.7361031781480551, + "learning_rate": 0.0006249781735186606, + "loss": 2.3245, + "step": 953 + }, + { + "epoch": 0.4375644994840041, + "grad_norm": 0.7728940836252534, + "learning_rate": 0.0006242585775448518, + "loss": 2.3696, + "step": 954 + }, + { + "epoch": 0.4380231624813668, + "grad_norm": 0.7497380470534151, + "learning_rate": 0.0006235387071508427, + "loss": 2.2936, + "step": 955 + }, + { + "epoch": 0.4384818254787295, + "grad_norm": 0.7353758444849192, + "learning_rate": 0.0006228185639264384, + "loss": 2.3704, + "step": 956 + }, + { + "epoch": 0.4389404884760922, + "grad_norm": 0.7374806693609556, + "learning_rate": 0.0006220981494620475, + "loss": 2.3393, + "step": 957 + }, + { + "epoch": 0.4393991514734549, + "grad_norm": 0.8474490786860961, + "learning_rate": 0.000621377465348677, + "loss": 2.4471, + "step": 958 + }, + { + "epoch": 0.43985781447081757, + "grad_norm": 0.745657649128481, + "learning_rate": 0.0006206565131779293, + "loss": 2.338, + "step": 959 + }, + { + "epoch": 0.44031647746818026, + "grad_norm": 0.7663244852255668, + "learning_rate": 0.0006199352945419994, + "loss": 2.3598, + "step": 960 + }, + { + "epoch": 0.44077514046554295, + "grad_norm": 0.7383058720895178, + "learning_rate": 0.00061921381103367, + "loss": 2.3662, + "step": 961 + }, + { + "epoch": 0.44123380346290564, + "grad_norm": 0.6963816937283623, + "learning_rate": 0.0006184920642463094, + "loss": 2.4132, + "step": 962 + }, + { + "epoch": 0.44169246646026833, + "grad_norm": 0.7416886319867618, + "learning_rate": 0.0006177700557738672, + "loss": 2.3909, + "step": 963 + }, + { + "epoch": 0.442151129457631, + "grad_norm": 0.7571624460048223, + "learning_rate": 0.0006170477872108706, + "loss": 2.3128, + "step": 964 + }, + { + "epoch": 0.4426097924549937, + "grad_norm": 0.6957055609599142, + "learning_rate": 0.0006163252601524216, + "loss": 2.3689, + "step": 965 + }, + { + "epoch": 0.4430684554523564, + "grad_norm": 0.7477397856603907, + "learning_rate": 0.0006156024761941925, + "loss": 2.3807, + "step": 966 + }, + { + "epoch": 0.4435271184497191, + "grad_norm": 0.6891520617774844, + "learning_rate": 0.000614879436932424, + "loss": 2.3877, + "step": 967 + }, + { + "epoch": 0.4439857814470818, + "grad_norm": 0.7285893347409235, + "learning_rate": 0.0006141561439639196, + "loss": 2.3252, + "step": 968 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 0.7299527266695685, + "learning_rate": 0.0006134325988860433, + "loss": 2.302, + "step": 969 + }, + { + "epoch": 0.4449031074418071, + "grad_norm": 0.7265748440328289, + "learning_rate": 0.0006127088032967165, + "loss": 2.3689, + "step": 970 + }, + { + "epoch": 0.4453617704391698, + "grad_norm": 0.7062705783995827, + "learning_rate": 0.0006119847587944131, + "loss": 2.3321, + "step": 971 + }, + { + "epoch": 0.4458204334365325, + "grad_norm": 0.7612018260686806, + "learning_rate": 0.0006112604669781572, + "loss": 2.3743, + "step": 972 + }, + { + "epoch": 0.4462790964338952, + "grad_norm": 0.7582103742743924, + "learning_rate": 0.0006105359294475188, + "loss": 2.4867, + "step": 973 + }, + { + "epoch": 0.4467377594312579, + "grad_norm": 0.6721677205650349, + "learning_rate": 0.0006098111478026107, + "loss": 2.3673, + "step": 974 + }, + { + "epoch": 0.44719642242862057, + "grad_norm": 0.7242677528592562, + "learning_rate": 0.0006090861236440848, + "loss": 2.4162, + "step": 975 + }, + { + "epoch": 0.44765508542598326, + "grad_norm": 0.7925723245628338, + "learning_rate": 0.0006083608585731282, + "loss": 2.3287, + "step": 976 + }, + { + "epoch": 0.44811374842334595, + "grad_norm": 0.7310198054265543, + "learning_rate": 0.0006076353541914609, + "loss": 2.3004, + "step": 977 + }, + { + "epoch": 0.44857241142070864, + "grad_norm": 0.6652082631276356, + "learning_rate": 0.0006069096121013307, + "loss": 2.3973, + "step": 978 + }, + { + "epoch": 0.44903107441807133, + "grad_norm": 0.708339525057551, + "learning_rate": 0.0006061836339055105, + "loss": 2.3248, + "step": 979 + }, + { + "epoch": 0.449489737415434, + "grad_norm": 0.7234437664732958, + "learning_rate": 0.0006054574212072948, + "loss": 2.3427, + "step": 980 + }, + { + "epoch": 0.4499484004127967, + "grad_norm": 0.7214211796558098, + "learning_rate": 0.0006047309756104958, + "loss": 2.3573, + "step": 981 + }, + { + "epoch": 0.4504070634101594, + "grad_norm": 0.7745594968910117, + "learning_rate": 0.00060400429871944, + "loss": 2.3417, + "step": 982 + }, + { + "epoch": 0.4508657264075221, + "grad_norm": 0.7440418925438044, + "learning_rate": 0.0006032773921389654, + "loss": 2.3054, + "step": 983 + }, + { + "epoch": 0.4513243894048848, + "grad_norm": 0.7318685679544856, + "learning_rate": 0.0006025502574744162, + "loss": 2.4159, + "step": 984 + }, + { + "epoch": 0.4517830524022475, + "grad_norm": 0.8846492208805433, + "learning_rate": 0.000601822896331641, + "loss": 2.3276, + "step": 985 + }, + { + "epoch": 0.4522417153996101, + "grad_norm": 0.827407835922663, + "learning_rate": 0.0006010953103169883, + "loss": 2.3557, + "step": 986 + }, + { + "epoch": 0.4527003783969728, + "grad_norm": 0.7132880914587941, + "learning_rate": 0.0006003675010373034, + "loss": 2.4186, + "step": 987 + }, + { + "epoch": 0.4531590413943355, + "grad_norm": 0.8342219103912915, + "learning_rate": 0.0005996394700999246, + "loss": 2.3761, + "step": 988 + }, + { + "epoch": 0.4536177043916982, + "grad_norm": 0.752632194328458, + "learning_rate": 0.0005989112191126794, + "loss": 2.3004, + "step": 989 + }, + { + "epoch": 0.4540763673890609, + "grad_norm": 0.7166605776198001, + "learning_rate": 0.0005981827496838822, + "loss": 2.2864, + "step": 990 + }, + { + "epoch": 0.45453503038642357, + "grad_norm": 0.7665351937782928, + "learning_rate": 0.0005974540634223286, + "loss": 2.3596, + "step": 991 + }, + { + "epoch": 0.45499369338378626, + "grad_norm": 0.7141283402114681, + "learning_rate": 0.0005967251619372939, + "loss": 2.3868, + "step": 992 + }, + { + "epoch": 0.45545235638114895, + "grad_norm": 0.8006502587725668, + "learning_rate": 0.0005959960468385284, + "loss": 2.4254, + "step": 993 + }, + { + "epoch": 0.45591101937851164, + "grad_norm": 0.7555445619080006, + "learning_rate": 0.0005952667197362542, + "loss": 2.2809, + "step": 994 + }, + { + "epoch": 0.45636968237587433, + "grad_norm": 0.719920413280221, + "learning_rate": 0.0005945371822411621, + "loss": 2.3944, + "step": 995 + }, + { + "epoch": 0.456828345373237, + "grad_norm": 0.7557890313701164, + "learning_rate": 0.0005938074359644063, + "loss": 2.281, + "step": 996 + }, + { + "epoch": 0.4572870083705997, + "grad_norm": 0.6908560346279087, + "learning_rate": 0.0005930774825176034, + "loss": 2.3688, + "step": 997 + }, + { + "epoch": 0.4577456713679624, + "grad_norm": 0.7219977787968322, + "learning_rate": 0.0005923473235128268, + "loss": 2.3502, + "step": 998 + }, + { + "epoch": 0.4582043343653251, + "grad_norm": 0.8076192870214803, + "learning_rate": 0.0005916169605626042, + "loss": 2.338, + "step": 999 + }, + { + "epoch": 0.4586629973626878, + "grad_norm": 0.7089674543241029, + "learning_rate": 0.0005908863952799134, + "loss": 2.3789, + "step": 1000 + }, + { + "epoch": 0.4591216603600505, + "grad_norm": 0.806568803588219, + "learning_rate": 0.0005901556292781793, + "loss": 2.3414, + "step": 1001 + }, + { + "epoch": 0.45958032335741317, + "grad_norm": 0.7027385896874405, + "learning_rate": 0.0005894246641712698, + "loss": 2.3253, + "step": 1002 + }, + { + "epoch": 0.4600389863547758, + "grad_norm": 0.7190293270119469, + "learning_rate": 0.0005886935015734931, + "loss": 2.3936, + "step": 1003 + }, + { + "epoch": 0.4604976493521385, + "grad_norm": 0.7088615954881797, + "learning_rate": 0.0005879621430995928, + "loss": 2.3298, + "step": 1004 + }, + { + "epoch": 0.4609563123495012, + "grad_norm": 0.6736402849682561, + "learning_rate": 0.0005872305903647455, + "loss": 2.3746, + "step": 1005 + }, + { + "epoch": 0.4614149753468639, + "grad_norm": 0.6513824750987495, + "learning_rate": 0.0005864988449845569, + "loss": 2.3071, + "step": 1006 + }, + { + "epoch": 0.46187363834422657, + "grad_norm": 0.7203870343616434, + "learning_rate": 0.0005857669085750578, + "loss": 2.3731, + "step": 1007 + }, + { + "epoch": 0.46233230134158926, + "grad_norm": 0.7039839556475195, + "learning_rate": 0.0005850347827527013, + "loss": 2.3678, + "step": 1008 + }, + { + "epoch": 0.46279096433895195, + "grad_norm": 0.721699207040889, + "learning_rate": 0.0005843024691343584, + "loss": 2.3114, + "step": 1009 + }, + { + "epoch": 0.46324962733631464, + "grad_norm": 0.6924742414984866, + "learning_rate": 0.000583569969337315, + "loss": 2.336, + "step": 1010 + }, + { + "epoch": 0.46370829033367733, + "grad_norm": 0.6741596462423349, + "learning_rate": 0.0005828372849792686, + "loss": 2.3324, + "step": 1011 + }, + { + "epoch": 0.46416695333104, + "grad_norm": 0.8018892774870944, + "learning_rate": 0.0005821044176783234, + "loss": 2.3844, + "step": 1012 + }, + { + "epoch": 0.4646256163284027, + "grad_norm": 0.7235614895294196, + "learning_rate": 0.0005813713690529886, + "loss": 2.3825, + "step": 1013 + }, + { + "epoch": 0.4650842793257654, + "grad_norm": 0.6692175291596106, + "learning_rate": 0.0005806381407221729, + "loss": 2.3989, + "step": 1014 + }, + { + "epoch": 0.4655429423231281, + "grad_norm": 0.716403013765753, + "learning_rate": 0.0005799047343051826, + "loss": 2.2973, + "step": 1015 + }, + { + "epoch": 0.4660016053204908, + "grad_norm": 0.6806984800551409, + "learning_rate": 0.0005791711514217171, + "loss": 2.393, + "step": 1016 + }, + { + "epoch": 0.4664602683178535, + "grad_norm": 0.763622053557696, + "learning_rate": 0.0005784373936918654, + "loss": 2.4367, + "step": 1017 + }, + { + "epoch": 0.46691893131521617, + "grad_norm": 0.7118474226132839, + "learning_rate": 0.0005777034627361025, + "loss": 2.3751, + "step": 1018 + }, + { + "epoch": 0.46737759431257886, + "grad_norm": 0.7191081571730317, + "learning_rate": 0.0005769693601752864, + "loss": 2.2759, + "step": 1019 + }, + { + "epoch": 0.4678362573099415, + "grad_norm": 0.806607219056567, + "learning_rate": 0.0005762350876306537, + "loss": 2.4168, + "step": 1020 + }, + { + "epoch": 0.4682949203073042, + "grad_norm": 0.6712961843794507, + "learning_rate": 0.0005755006467238168, + "loss": 2.3682, + "step": 1021 + }, + { + "epoch": 0.4687535833046669, + "grad_norm": 0.703099431659453, + "learning_rate": 0.0005747660390767593, + "loss": 2.375, + "step": 1022 + }, + { + "epoch": 0.46921224630202957, + "grad_norm": 0.6798027127601068, + "learning_rate": 0.0005740312663118338, + "loss": 2.2456, + "step": 1023 + }, + { + "epoch": 0.46967090929939226, + "grad_norm": 0.7308799104370874, + "learning_rate": 0.0005732963300517568, + "loss": 2.2563, + "step": 1024 + }, + { + "epoch": 0.47012957229675495, + "grad_norm": 0.7733164640356353, + "learning_rate": 0.0005725612319196064, + "loss": 2.4211, + "step": 1025 + }, + { + "epoch": 0.47058823529411764, + "grad_norm": 0.7064197718295134, + "learning_rate": 0.000571825973538818, + "loss": 2.331, + "step": 1026 + }, + { + "epoch": 0.47104689829148033, + "grad_norm": 0.7132260483493038, + "learning_rate": 0.0005710905565331811, + "loss": 2.3622, + "step": 1027 + }, + { + "epoch": 0.471505561288843, + "grad_norm": 0.7361070709219814, + "learning_rate": 0.0005703549825268353, + "loss": 2.3915, + "step": 1028 + }, + { + "epoch": 0.4719642242862057, + "grad_norm": 0.7172260139105476, + "learning_rate": 0.0005696192531442667, + "loss": 2.3314, + "step": 1029 + }, + { + "epoch": 0.4724228872835684, + "grad_norm": 0.6653116447166156, + "learning_rate": 0.000568883370010305, + "loss": 2.4461, + "step": 1030 + }, + { + "epoch": 0.4728815502809311, + "grad_norm": 0.7462447327526615, + "learning_rate": 0.0005681473347501192, + "loss": 2.5715, + "step": 1031 + }, + { + "epoch": 0.4733402132782938, + "grad_norm": 0.8282354946029855, + "learning_rate": 0.0005674111489892144, + "loss": 2.297, + "step": 1032 + }, + { + "epoch": 0.4737988762756565, + "grad_norm": 0.6939777943203792, + "learning_rate": 0.0005666748143534282, + "loss": 2.3882, + "step": 1033 + }, + { + "epoch": 0.47425753927301917, + "grad_norm": 0.6975144604518554, + "learning_rate": 0.0005659383324689266, + "loss": 2.3215, + "step": 1034 + }, + { + "epoch": 0.47471620227038186, + "grad_norm": 0.7468482095613771, + "learning_rate": 0.0005652017049622007, + "loss": 2.3393, + "step": 1035 + }, + { + "epoch": 0.47517486526774455, + "grad_norm": 0.759403157552217, + "learning_rate": 0.0005644649334600641, + "loss": 2.3492, + "step": 1036 + }, + { + "epoch": 0.4756335282651072, + "grad_norm": 0.7275292927417828, + "learning_rate": 0.0005637280195896474, + "loss": 2.4238, + "step": 1037 + }, + { + "epoch": 0.4760921912624699, + "grad_norm": 0.7772183424957455, + "learning_rate": 0.0005629909649783961, + "loss": 2.4051, + "step": 1038 + }, + { + "epoch": 0.47655085425983257, + "grad_norm": 0.6775883173753181, + "learning_rate": 0.0005622537712540664, + "loss": 2.5465, + "step": 1039 + }, + { + "epoch": 0.47700951725719526, + "grad_norm": 0.7780627695821611, + "learning_rate": 0.0005615164400447218, + "loss": 2.3337, + "step": 1040 + }, + { + "epoch": 0.47746818025455795, + "grad_norm": 0.7595704751770299, + "learning_rate": 0.0005607789729787294, + "loss": 2.3837, + "step": 1041 + }, + { + "epoch": 0.47792684325192064, + "grad_norm": 0.686218807200755, + "learning_rate": 0.0005600413716847564, + "loss": 2.288, + "step": 1042 + }, + { + "epoch": 0.47838550624928333, + "grad_norm": 0.7626604279774272, + "learning_rate": 0.000559303637791766, + "loss": 2.3139, + "step": 1043 + }, + { + "epoch": 0.478844169246646, + "grad_norm": 0.7139999723828158, + "learning_rate": 0.0005585657729290151, + "loss": 2.3106, + "step": 1044 + }, + { + "epoch": 0.4793028322440087, + "grad_norm": 0.7385225130570024, + "learning_rate": 0.000557827778726049, + "loss": 2.3343, + "step": 1045 + }, + { + "epoch": 0.4797614952413714, + "grad_norm": 0.7957781134927029, + "learning_rate": 0.0005570896568126993, + "loss": 2.2641, + "step": 1046 + }, + { + "epoch": 0.4802201582387341, + "grad_norm": 0.7412711022835641, + "learning_rate": 0.0005563514088190788, + "loss": 2.333, + "step": 1047 + }, + { + "epoch": 0.4806788212360968, + "grad_norm": 0.7816979265907771, + "learning_rate": 0.0005556130363755798, + "loss": 2.3256, + "step": 1048 + }, + { + "epoch": 0.4811374842334595, + "grad_norm": 0.8116795086500815, + "learning_rate": 0.0005548745411128688, + "loss": 2.2085, + "step": 1049 + }, + { + "epoch": 0.48159614723082217, + "grad_norm": 0.8101296106758742, + "learning_rate": 0.0005541359246618835, + "loss": 2.3307, + "step": 1050 + }, + { + "epoch": 0.48205481022818486, + "grad_norm": 0.7461999598226511, + "learning_rate": 0.0005533971886538293, + "loss": 2.3844, + "step": 1051 + }, + { + "epoch": 0.48251347322554755, + "grad_norm": 0.7936408349406735, + "learning_rate": 0.000552658334720176, + "loss": 2.3772, + "step": 1052 + }, + { + "epoch": 0.48297213622291024, + "grad_norm": 0.7960105729448774, + "learning_rate": 0.0005519193644926535, + "loss": 2.2798, + "step": 1053 + }, + { + "epoch": 0.4834307992202729, + "grad_norm": 0.696743146481541, + "learning_rate": 0.0005511802796032485, + "loss": 2.273, + "step": 1054 + }, + { + "epoch": 0.48388946221763557, + "grad_norm": 0.6895932115731436, + "learning_rate": 0.0005504410816842009, + "loss": 2.3394, + "step": 1055 + }, + { + "epoch": 0.48434812521499826, + "grad_norm": 0.7047505303299945, + "learning_rate": 0.0005497017723680009, + "loss": 2.2827, + "step": 1056 + }, + { + "epoch": 0.48480678821236095, + "grad_norm": 0.7480981964053715, + "learning_rate": 0.0005489623532873836, + "loss": 2.3381, + "step": 1057 + }, + { + "epoch": 0.48526545120972364, + "grad_norm": 0.6623872456700465, + "learning_rate": 0.0005482228260753273, + "loss": 2.3497, + "step": 1058 + }, + { + "epoch": 0.48572411420708633, + "grad_norm": 0.6751521645707396, + "learning_rate": 0.0005474831923650488, + "loss": 2.3659, + "step": 1059 + }, + { + "epoch": 0.486182777204449, + "grad_norm": 0.7896498336767525, + "learning_rate": 0.00054674345379, + "loss": 2.2839, + "step": 1060 + }, + { + "epoch": 0.4866414402018117, + "grad_norm": 0.7635097335998425, + "learning_rate": 0.000546003611983865, + "loss": 2.3629, + "step": 1061 + }, + { + "epoch": 0.4871001031991744, + "grad_norm": 0.7033871562523616, + "learning_rate": 0.0005452636685805552, + "loss": 2.3036, + "step": 1062 + }, + { + "epoch": 0.4875587661965371, + "grad_norm": 0.7818943109979322, + "learning_rate": 0.0005445236252142066, + "loss": 2.4725, + "step": 1063 + }, + { + "epoch": 0.4880174291938998, + "grad_norm": 0.7826097792865764, + "learning_rate": 0.000543783483519176, + "loss": 2.3502, + "step": 1064 + }, + { + "epoch": 0.4884760921912625, + "grad_norm": 0.7867797289510179, + "learning_rate": 0.0005430432451300374, + "loss": 2.2777, + "step": 1065 + }, + { + "epoch": 0.48893475518862517, + "grad_norm": 0.6559826229653739, + "learning_rate": 0.0005423029116815781, + "loss": 2.2833, + "step": 1066 + }, + { + "epoch": 0.48939341818598786, + "grad_norm": 0.7321644148250357, + "learning_rate": 0.0005415624848087959, + "loss": 2.3351, + "step": 1067 + }, + { + "epoch": 0.48985208118335055, + "grad_norm": 0.833962898953224, + "learning_rate": 0.000540821966146894, + "loss": 2.4361, + "step": 1068 + }, + { + "epoch": 0.49031074418071324, + "grad_norm": 0.7397483293283564, + "learning_rate": 0.0005400813573312793, + "loss": 2.2673, + "step": 1069 + }, + { + "epoch": 0.49076940717807593, + "grad_norm": 0.7244531889239589, + "learning_rate": 0.0005393406599975572, + "loss": 2.3752, + "step": 1070 + }, + { + "epoch": 0.49122807017543857, + "grad_norm": 0.7254044487348418, + "learning_rate": 0.0005385998757815287, + "loss": 2.3812, + "step": 1071 + }, + { + "epoch": 0.49168673317280126, + "grad_norm": 0.7540194928503902, + "learning_rate": 0.0005378590063191867, + "loss": 2.3062, + "step": 1072 + }, + { + "epoch": 0.49214539617016395, + "grad_norm": 0.7618602606603975, + "learning_rate": 0.0005371180532467124, + "loss": 2.3779, + "step": 1073 + }, + { + "epoch": 0.49260405916752664, + "grad_norm": 0.7167992630400797, + "learning_rate": 0.000536377018200472, + "loss": 2.3256, + "step": 1074 + }, + { + "epoch": 0.49306272216488933, + "grad_norm": 0.7564537404848328, + "learning_rate": 0.0005356359028170118, + "loss": 2.3332, + "step": 1075 + }, + { + "epoch": 0.493521385162252, + "grad_norm": 0.6514487074777742, + "learning_rate": 0.0005348947087330564, + "loss": 2.3804, + "step": 1076 + }, + { + "epoch": 0.4939800481596147, + "grad_norm": 0.7151894900548901, + "learning_rate": 0.0005341534375855037, + "loss": 2.4131, + "step": 1077 + }, + { + "epoch": 0.4944387111569774, + "grad_norm": 0.7658980218027751, + "learning_rate": 0.0005334120910114222, + "loss": 2.3579, + "step": 1078 + }, + { + "epoch": 0.4948973741543401, + "grad_norm": 0.6878095061588041, + "learning_rate": 0.0005326706706480467, + "loss": 2.3395, + "step": 1079 + }, + { + "epoch": 0.4953560371517028, + "grad_norm": 0.6982487901712734, + "learning_rate": 0.0005319291781327749, + "loss": 2.2769, + "step": 1080 + }, + { + "epoch": 0.4958147001490655, + "grad_norm": 0.773112023750417, + "learning_rate": 0.0005311876151031642, + "loss": 2.3305, + "step": 1081 + }, + { + "epoch": 0.49627336314642817, + "grad_norm": 0.7215322763575871, + "learning_rate": 0.0005304459831969274, + "loss": 2.2976, + "step": 1082 + }, + { + "epoch": 0.49673202614379086, + "grad_norm": 0.7458674571653674, + "learning_rate": 0.0005297042840519294, + "loss": 2.2991, + "step": 1083 + }, + { + "epoch": 0.49719068914115355, + "grad_norm": 0.732002430534551, + "learning_rate": 0.0005289625193061838, + "loss": 2.2948, + "step": 1084 + }, + { + "epoch": 0.49764935213851624, + "grad_norm": 0.7379677949370402, + "learning_rate": 0.0005282206905978489, + "loss": 2.2531, + "step": 1085 + }, + { + "epoch": 0.49810801513587893, + "grad_norm": 0.6887993196610556, + "learning_rate": 0.0005274787995652246, + "loss": 2.2947, + "step": 1086 + }, + { + "epoch": 0.4985666781332416, + "grad_norm": 0.7344461714458195, + "learning_rate": 0.000526736847846748, + "loss": 2.2785, + "step": 1087 + }, + { + "epoch": 0.49902534113060426, + "grad_norm": 0.6973660570459533, + "learning_rate": 0.0005259948370809901, + "loss": 2.3743, + "step": 1088 + }, + { + "epoch": 0.49948400412796695, + "grad_norm": 0.6667215831575656, + "learning_rate": 0.0005252527689066533, + "loss": 2.3922, + "step": 1089 + }, + { + "epoch": 0.49994266712532964, + "grad_norm": 0.6875796534643218, + "learning_rate": 0.0005245106449625654, + "loss": 2.3624, + "step": 1090 + }, + { + "epoch": 0.5004013301226924, + "grad_norm": 0.7226696062456723, + "learning_rate": 0.0005237684668876785, + "loss": 2.323, + "step": 1091 + }, + { + "epoch": 0.500859993120055, + "grad_norm": 0.7554049092887116, + "learning_rate": 0.0005230262363210637, + "loss": 2.4197, + "step": 1092 + }, + { + "epoch": 0.5013186561174178, + "grad_norm": 0.7152082078323221, + "learning_rate": 0.0005222839549019079, + "loss": 2.3822, + "step": 1093 + }, + { + "epoch": 0.5017773191147804, + "grad_norm": 0.7394814715669289, + "learning_rate": 0.0005215416242695108, + "loss": 2.2996, + "step": 1094 + }, + { + "epoch": 0.5022359821121432, + "grad_norm": 0.6617882735239301, + "learning_rate": 0.0005207992460632804, + "loss": 2.3581, + "step": 1095 + }, + { + "epoch": 0.5026946451095058, + "grad_norm": 0.7472871438835622, + "learning_rate": 0.0005200568219227299, + "loss": 2.2664, + "step": 1096 + }, + { + "epoch": 0.5031533081068684, + "grad_norm": 0.7522779770332627, + "learning_rate": 0.000519314353487474, + "loss": 2.3505, + "step": 1097 + }, + { + "epoch": 0.5036119711042312, + "grad_norm": 0.7203739927529927, + "learning_rate": 0.0005185718423972251, + "loss": 2.326, + "step": 1098 + }, + { + "epoch": 0.5040706341015938, + "grad_norm": 0.7070088622888988, + "learning_rate": 0.0005178292902917898, + "loss": 2.3771, + "step": 1099 + }, + { + "epoch": 0.5045292970989566, + "grad_norm": 0.6923421329126254, + "learning_rate": 0.0005170866988110656, + "loss": 2.3642, + "step": 1100 + }, + { + "epoch": 0.5049879600963192, + "grad_norm": 0.7343965792157483, + "learning_rate": 0.0005163440695950362, + "loss": 2.2655, + "step": 1101 + }, + { + "epoch": 0.5054466230936819, + "grad_norm": 0.7076232647308535, + "learning_rate": 0.0005156014042837695, + "loss": 2.2986, + "step": 1102 + }, + { + "epoch": 0.5059052860910446, + "grad_norm": 0.7194434839990177, + "learning_rate": 0.0005148587045174128, + "loss": 2.4362, + "step": 1103 + }, + { + "epoch": 0.5063639490884073, + "grad_norm": 0.7265144160432816, + "learning_rate": 0.0005141159719361891, + "loss": 2.3185, + "step": 1104 + }, + { + "epoch": 0.50682261208577, + "grad_norm": 0.6461552736354166, + "learning_rate": 0.0005133732081803945, + "loss": 2.3496, + "step": 1105 + }, + { + "epoch": 0.5072812750831327, + "grad_norm": 0.7268076949194873, + "learning_rate": 0.0005126304148903936, + "loss": 2.3328, + "step": 1106 + }, + { + "epoch": 0.5077399380804953, + "grad_norm": 0.7110003769275122, + "learning_rate": 0.0005118875937066161, + "loss": 2.4131, + "step": 1107 + }, + { + "epoch": 0.5081986010778581, + "grad_norm": 0.6861362393230458, + "learning_rate": 0.0005111447462695537, + "loss": 2.3672, + "step": 1108 + }, + { + "epoch": 0.5086572640752207, + "grad_norm": 0.6714225810162389, + "learning_rate": 0.0005104018742197557, + "loss": 2.3251, + "step": 1109 + }, + { + "epoch": 0.5091159270725835, + "grad_norm": 0.6495236835677141, + "learning_rate": 0.0005096589791978261, + "loss": 2.2889, + "step": 1110 + }, + { + "epoch": 0.5095745900699461, + "grad_norm": 0.7080589093120262, + "learning_rate": 0.0005089160628444192, + "loss": 2.3771, + "step": 1111 + }, + { + "epoch": 0.5100332530673088, + "grad_norm": 0.6603785719616789, + "learning_rate": 0.0005081731268002371, + "loss": 2.3645, + "step": 1112 + }, + { + "epoch": 0.5104919160646715, + "grad_norm": 0.6745823872057706, + "learning_rate": 0.0005074301727060243, + "loss": 2.252, + "step": 1113 + }, + { + "epoch": 0.5109505790620341, + "grad_norm": 0.7321327447622302, + "learning_rate": 0.0005066872022025663, + "loss": 2.2656, + "step": 1114 + }, + { + "epoch": 0.5114092420593969, + "grad_norm": 0.6936322347994426, + "learning_rate": 0.0005059442169306844, + "loss": 2.3696, + "step": 1115 + }, + { + "epoch": 0.5118679050567595, + "grad_norm": 0.7498410366607514, + "learning_rate": 0.0005052012185312321, + "loss": 2.3676, + "step": 1116 + }, + { + "epoch": 0.5123265680541222, + "grad_norm": 0.6955469960164541, + "learning_rate": 0.0005044582086450925, + "loss": 2.2998, + "step": 1117 + }, + { + "epoch": 0.5127852310514849, + "grad_norm": 0.6920389789884202, + "learning_rate": 0.0005037151889131737, + "loss": 2.3765, + "step": 1118 + }, + { + "epoch": 0.5132438940488476, + "grad_norm": 0.7553382814600957, + "learning_rate": 0.0005029721609764059, + "loss": 2.249, + "step": 1119 + }, + { + "epoch": 0.5137025570462103, + "grad_norm": 0.6827968931621355, + "learning_rate": 0.000502229126475737, + "loss": 2.3784, + "step": 1120 + }, + { + "epoch": 0.514161220043573, + "grad_norm": 0.7557759058179622, + "learning_rate": 0.0005014860870521293, + "loss": 2.2134, + "step": 1121 + }, + { + "epoch": 0.5146198830409356, + "grad_norm": 0.6930096547251532, + "learning_rate": 0.0005007430443465569, + "loss": 2.4054, + "step": 1122 + }, + { + "epoch": 0.5150785460382984, + "grad_norm": 0.7537661103249673, + "learning_rate": 0.0005, + "loss": 2.3493, + "step": 1123 + }, + { + "epoch": 0.515537209035661, + "grad_norm": 0.7743398765125217, + "learning_rate": 0.0004992569556534432, + "loss": 2.365, + "step": 1124 + }, + { + "epoch": 0.5159958720330238, + "grad_norm": 0.7290649607127428, + "learning_rate": 0.0004985139129478707, + "loss": 2.3505, + "step": 1125 + }, + { + "epoch": 0.5164545350303864, + "grad_norm": 0.7603541060448682, + "learning_rate": 0.0004977708735242633, + "loss": 2.315, + "step": 1126 + }, + { + "epoch": 0.5169131980277492, + "grad_norm": 0.731287912441591, + "learning_rate": 0.0004970278390235942, + "loss": 2.3538, + "step": 1127 + }, + { + "epoch": 0.5173718610251118, + "grad_norm": 0.7929681053098886, + "learning_rate": 0.0004962848110868262, + "loss": 2.3448, + "step": 1128 + }, + { + "epoch": 0.5178305240224745, + "grad_norm": 0.7134398463715416, + "learning_rate": 0.0004955417913549074, + "loss": 2.2996, + "step": 1129 + }, + { + "epoch": 0.5182891870198372, + "grad_norm": 0.6997358652318024, + "learning_rate": 0.0004947987814687679, + "loss": 2.4527, + "step": 1130 + }, + { + "epoch": 0.5187478500171998, + "grad_norm": 0.7633354374790329, + "learning_rate": 0.0004940557830693157, + "loss": 2.2952, + "step": 1131 + }, + { + "epoch": 0.5192065130145626, + "grad_norm": 0.7806460078672218, + "learning_rate": 0.0004933127977974338, + "loss": 2.3383, + "step": 1132 + }, + { + "epoch": 0.5196651760119252, + "grad_norm": 0.6781693715600023, + "learning_rate": 0.0004925698272939757, + "loss": 2.3301, + "step": 1133 + }, + { + "epoch": 0.5201238390092879, + "grad_norm": 0.6788599958572494, + "learning_rate": 0.0004918268731997632, + "loss": 2.3, + "step": 1134 + }, + { + "epoch": 0.5205825020066506, + "grad_norm": 0.7029384483080904, + "learning_rate": 0.0004910839371555809, + "loss": 2.3057, + "step": 1135 + }, + { + "epoch": 0.5210411650040133, + "grad_norm": 0.7020026986612986, + "learning_rate": 0.0004903410208021739, + "loss": 2.2952, + "step": 1136 + }, + { + "epoch": 0.521499828001376, + "grad_norm": 0.6506845375319315, + "learning_rate": 0.0004895981257802443, + "loss": 2.2391, + "step": 1137 + }, + { + "epoch": 0.5219584909987387, + "grad_norm": 0.6839317538693314, + "learning_rate": 0.0004888552537304463, + "loss": 2.2604, + "step": 1138 + }, + { + "epoch": 0.5224171539961013, + "grad_norm": 0.8397252651392505, + "learning_rate": 0.00048811240629338394, + "loss": 2.2486, + "step": 1139 + }, + { + "epoch": 0.5228758169934641, + "grad_norm": 0.7390284388466395, + "learning_rate": 0.00048736958510960663, + "loss": 2.311, + "step": 1140 + }, + { + "epoch": 0.5233344799908267, + "grad_norm": 0.7058226776504977, + "learning_rate": 0.00048662679181960564, + "loss": 2.3711, + "step": 1141 + }, + { + "epoch": 0.5237931429881895, + "grad_norm": 0.7030752077324364, + "learning_rate": 0.00048588402806381094, + "loss": 2.3221, + "step": 1142 + }, + { + "epoch": 0.5242518059855521, + "grad_norm": 0.7809311606475811, + "learning_rate": 0.0004851412954825874, + "loss": 2.2936, + "step": 1143 + }, + { + "epoch": 0.5247104689829148, + "grad_norm": 0.7016179714392349, + "learning_rate": 0.00048439859571623034, + "loss": 2.3181, + "step": 1144 + }, + { + "epoch": 0.5251691319802775, + "grad_norm": 0.7255445041174645, + "learning_rate": 0.00048365593040496373, + "loss": 2.401, + "step": 1145 + }, + { + "epoch": 0.5256277949776402, + "grad_norm": 0.6926742666997815, + "learning_rate": 0.00048291330118893443, + "loss": 2.3519, + "step": 1146 + }, + { + "epoch": 0.5260864579750029, + "grad_norm": 0.7019166713929316, + "learning_rate": 0.0004821707097082102, + "loss": 2.3405, + "step": 1147 + }, + { + "epoch": 0.5265451209723655, + "grad_norm": 0.7253946866213568, + "learning_rate": 0.0004814281576027749, + "loss": 2.421, + "step": 1148 + }, + { + "epoch": 0.5270037839697282, + "grad_norm": 0.666285359695125, + "learning_rate": 0.000480685646512526, + "loss": 2.3892, + "step": 1149 + }, + { + "epoch": 0.5274624469670909, + "grad_norm": 0.7094184021496854, + "learning_rate": 0.00047994317807727025, + "loss": 2.4254, + "step": 1150 + }, + { + "epoch": 0.5279211099644536, + "grad_norm": 0.7278079955774219, + "learning_rate": 0.00047920075393671974, + "loss": 2.2559, + "step": 1151 + }, + { + "epoch": 0.5283797729618163, + "grad_norm": 0.7850924590832233, + "learning_rate": 0.0004784583757304893, + "loss": 2.3917, + "step": 1152 + }, + { + "epoch": 0.528838435959179, + "grad_norm": 0.6584511513687065, + "learning_rate": 0.00047771604509809214, + "loss": 2.3657, + "step": 1153 + }, + { + "epoch": 0.5292970989565416, + "grad_norm": 0.7038755982648288, + "learning_rate": 0.0004769737636789364, + "loss": 2.3662, + "step": 1154 + }, + { + "epoch": 0.5297557619539044, + "grad_norm": 0.7093642443215273, + "learning_rate": 0.00047623153311232157, + "loss": 2.2226, + "step": 1155 + }, + { + "epoch": 0.530214424951267, + "grad_norm": 0.6720123016459876, + "learning_rate": 0.0004754893550374346, + "loss": 2.2634, + "step": 1156 + }, + { + "epoch": 0.5306730879486298, + "grad_norm": 0.6833684351239346, + "learning_rate": 0.00047474723109334685, + "loss": 2.2661, + "step": 1157 + }, + { + "epoch": 0.5311317509459924, + "grad_norm": 0.7094920467130852, + "learning_rate": 0.00047400516291900993, + "loss": 2.3438, + "step": 1158 + }, + { + "epoch": 0.5315904139433552, + "grad_norm": 0.7354164972399502, + "learning_rate": 0.0004732631521532522, + "loss": 2.3595, + "step": 1159 + }, + { + "epoch": 0.5320490769407178, + "grad_norm": 0.7195330641442192, + "learning_rate": 0.0004725212004347755, + "loss": 2.3125, + "step": 1160 + }, + { + "epoch": 0.5325077399380805, + "grad_norm": 0.686735634371495, + "learning_rate": 0.00047177930940215095, + "loss": 2.3009, + "step": 1161 + }, + { + "epoch": 0.5329664029354432, + "grad_norm": 0.6699205018663491, + "learning_rate": 0.00047103748069381624, + "loss": 2.2808, + "step": 1162 + }, + { + "epoch": 0.5334250659328059, + "grad_norm": 0.6981488499548693, + "learning_rate": 0.0004702957159480707, + "loss": 2.4086, + "step": 1163 + }, + { + "epoch": 0.5338837289301686, + "grad_norm": 0.7098340900928889, + "learning_rate": 0.00046955401680307267, + "loss": 2.4618, + "step": 1164 + }, + { + "epoch": 0.5343423919275313, + "grad_norm": 0.8700861801585815, + "learning_rate": 0.0004688123848968359, + "loss": 2.3284, + "step": 1165 + }, + { + "epoch": 0.5348010549248939, + "grad_norm": 0.7031207479863996, + "learning_rate": 0.00046807082186722516, + "loss": 2.3282, + "step": 1166 + }, + { + "epoch": 0.5352597179222566, + "grad_norm": 0.6586549637350291, + "learning_rate": 0.0004673293293519535, + "loss": 2.3307, + "step": 1167 + }, + { + "epoch": 0.5357183809196193, + "grad_norm": 0.650086124671687, + "learning_rate": 0.00046658790898857806, + "loss": 2.393, + "step": 1168 + }, + { + "epoch": 0.536177043916982, + "grad_norm": 0.7225508505256387, + "learning_rate": 0.0004658465624144963, + "loss": 2.2823, + "step": 1169 + }, + { + "epoch": 0.5366357069143447, + "grad_norm": 0.6903591515011483, + "learning_rate": 0.0004651052912669438, + "loss": 2.4015, + "step": 1170 + }, + { + "epoch": 0.5370943699117073, + "grad_norm": 0.6602654342193528, + "learning_rate": 0.0004643640971829883, + "loss": 2.2741, + "step": 1171 + }, + { + "epoch": 0.5375530329090701, + "grad_norm": 0.7159356106671477, + "learning_rate": 0.0004636229817995281, + "loss": 2.3575, + "step": 1172 + }, + { + "epoch": 0.5380116959064327, + "grad_norm": 0.6696943700349886, + "learning_rate": 0.0004628819467532876, + "loss": 2.2135, + "step": 1173 + }, + { + "epoch": 0.5384703589037955, + "grad_norm": 0.6981420986379476, + "learning_rate": 0.00046214099368081335, + "loss": 2.3749, + "step": 1174 + }, + { + "epoch": 0.5389290219011581, + "grad_norm": 0.7222848628616158, + "learning_rate": 0.0004614001242184714, + "loss": 2.3101, + "step": 1175 + }, + { + "epoch": 0.5393876848985208, + "grad_norm": 0.7250293911766736, + "learning_rate": 0.000460659340002443, + "loss": 2.3423, + "step": 1176 + }, + { + "epoch": 0.5398463478958835, + "grad_norm": 0.7189144828693277, + "learning_rate": 0.00045991864266872073, + "loss": 2.3155, + "step": 1177 + }, + { + "epoch": 0.5403050108932462, + "grad_norm": 0.7642787883147378, + "learning_rate": 0.00045917803385310595, + "loss": 2.3312, + "step": 1178 + }, + { + "epoch": 0.5407636738906089, + "grad_norm": 0.6793480950361692, + "learning_rate": 0.00045843751519120417, + "loss": 2.3387, + "step": 1179 + }, + { + "epoch": 0.5412223368879716, + "grad_norm": 0.7384266623049037, + "learning_rate": 0.00045769708831842193, + "loss": 2.3991, + "step": 1180 + }, + { + "epoch": 0.5416809998853342, + "grad_norm": 0.6943772289992626, + "learning_rate": 0.00045695675486996266, + "loss": 2.3149, + "step": 1181 + }, + { + "epoch": 0.542139662882697, + "grad_norm": 0.7925501006048682, + "learning_rate": 0.00045621651648082405, + "loss": 2.2497, + "step": 1182 + }, + { + "epoch": 0.5425983258800596, + "grad_norm": 0.6909505667512327, + "learning_rate": 0.00045547637478579356, + "loss": 2.2324, + "step": 1183 + }, + { + "epoch": 0.5430569888774223, + "grad_norm": 0.7102845819162674, + "learning_rate": 0.0004547363314194449, + "loss": 2.4028, + "step": 1184 + }, + { + "epoch": 0.543515651874785, + "grad_norm": 0.6279179191731096, + "learning_rate": 0.000453996388016135, + "loss": 2.3833, + "step": 1185 + }, + { + "epoch": 0.5439743148721476, + "grad_norm": 0.6831694855359374, + "learning_rate": 0.0004532565462099999, + "loss": 2.3218, + "step": 1186 + }, + { + "epoch": 0.5444329778695104, + "grad_norm": 0.6852766220181321, + "learning_rate": 0.0004525168076349513, + "loss": 2.3453, + "step": 1187 + }, + { + "epoch": 0.544891640866873, + "grad_norm": 0.7278185551097761, + "learning_rate": 0.0004517771739246729, + "loss": 2.3434, + "step": 1188 + }, + { + "epoch": 0.5453503038642358, + "grad_norm": 0.7237644090513412, + "learning_rate": 0.0004510376467126165, + "loss": 2.3541, + "step": 1189 + }, + { + "epoch": 0.5458089668615984, + "grad_norm": 0.6901236640387476, + "learning_rate": 0.0004502982276319992, + "loss": 2.3034, + "step": 1190 + }, + { + "epoch": 0.5462676298589612, + "grad_norm": 0.6911899362636125, + "learning_rate": 0.0004495589183157991, + "loss": 2.2217, + "step": 1191 + }, + { + "epoch": 0.5467262928563238, + "grad_norm": 0.7026500600487752, + "learning_rate": 0.0004488197203967517, + "loss": 2.3263, + "step": 1192 + }, + { + "epoch": 0.5471849558536865, + "grad_norm": 0.6806002145123267, + "learning_rate": 0.0004480806355073467, + "loss": 2.2739, + "step": 1193 + }, + { + "epoch": 0.5476436188510492, + "grad_norm": 0.7419980144164756, + "learning_rate": 0.000447341665279824, + "loss": 2.2918, + "step": 1194 + }, + { + "epoch": 0.5481022818484119, + "grad_norm": 0.7116690162400316, + "learning_rate": 0.0004466028113461708, + "loss": 2.3056, + "step": 1195 + }, + { + "epoch": 0.5485609448457746, + "grad_norm": 0.659677565751824, + "learning_rate": 0.0004458640753381167, + "loss": 2.2933, + "step": 1196 + }, + { + "epoch": 0.5490196078431373, + "grad_norm": 0.6565278034997855, + "learning_rate": 0.0004451254588871313, + "loss": 2.3527, + "step": 1197 + }, + { + "epoch": 0.5494782708404999, + "grad_norm": 0.6781180036209883, + "learning_rate": 0.0004443869636244203, + "loss": 2.3165, + "step": 1198 + }, + { + "epoch": 0.5499369338378627, + "grad_norm": 0.6546371945675067, + "learning_rate": 0.0004436485911809212, + "loss": 2.2654, + "step": 1199 + }, + { + "epoch": 0.5503955968352253, + "grad_norm": 0.6785564213531915, + "learning_rate": 0.00044291034318730087, + "loss": 2.2997, + "step": 1200 + }, + { + "epoch": 0.550854259832588, + "grad_norm": 0.6991784293725265, + "learning_rate": 0.0004421722212739511, + "loss": 2.2659, + "step": 1201 + }, + { + "epoch": 0.5513129228299507, + "grad_norm": 0.7423550866832677, + "learning_rate": 0.0004414342270709848, + "loss": 2.2919, + "step": 1202 + }, + { + "epoch": 0.5517715858273133, + "grad_norm": 0.6732543209190459, + "learning_rate": 0.00044069636220823397, + "loss": 2.3009, + "step": 1203 + }, + { + "epoch": 0.5522302488246761, + "grad_norm": 0.679075699293886, + "learning_rate": 0.0004399586283152437, + "loss": 2.3808, + "step": 1204 + }, + { + "epoch": 0.5526889118220387, + "grad_norm": 0.7219781125920871, + "learning_rate": 0.0004392210270212706, + "loss": 2.3053, + "step": 1205 + }, + { + "epoch": 0.5531475748194015, + "grad_norm": 0.697258419011195, + "learning_rate": 0.00043848355995527825, + "loss": 2.3721, + "step": 1206 + }, + { + "epoch": 0.5536062378167641, + "grad_norm": 0.7294663461112869, + "learning_rate": 0.00043774622874593374, + "loss": 2.3144, + "step": 1207 + }, + { + "epoch": 0.5540649008141268, + "grad_norm": 0.7171650879263656, + "learning_rate": 0.000437009035021604, + "loss": 2.3308, + "step": 1208 + }, + { + "epoch": 0.5545235638114895, + "grad_norm": 0.717182764769441, + "learning_rate": 0.00043627198041035274, + "loss": 2.3229, + "step": 1209 + }, + { + "epoch": 0.5549822268088522, + "grad_norm": 0.6716682202665368, + "learning_rate": 0.00043553506653993597, + "loss": 2.369, + "step": 1210 + }, + { + "epoch": 0.5554408898062149, + "grad_norm": 0.7241353264835427, + "learning_rate": 0.0004347982950377992, + "loss": 2.2693, + "step": 1211 + }, + { + "epoch": 0.5558995528035776, + "grad_norm": 0.6747811756238499, + "learning_rate": 0.0004340616675310735, + "loss": 2.257, + "step": 1212 + }, + { + "epoch": 0.5563582158009402, + "grad_norm": 0.6773463252497344, + "learning_rate": 0.00043332518564657193, + "loss": 2.2212, + "step": 1213 + }, + { + "epoch": 0.556816878798303, + "grad_norm": 0.7693189181481275, + "learning_rate": 0.0004325888510107856, + "loss": 2.4636, + "step": 1214 + }, + { + "epoch": 0.5572755417956656, + "grad_norm": 0.6834871136410707, + "learning_rate": 0.0004318526652498809, + "loss": 2.2893, + "step": 1215 + }, + { + "epoch": 0.5577342047930284, + "grad_norm": 0.7244197504858353, + "learning_rate": 0.00043111662998969523, + "loss": 2.3773, + "step": 1216 + }, + { + "epoch": 0.558192867790391, + "grad_norm": 0.7270312216667245, + "learning_rate": 0.0004303807468557335, + "loss": 2.429, + "step": 1217 + }, + { + "epoch": 0.5586515307877536, + "grad_norm": 0.7427344592050403, + "learning_rate": 0.0004296450174731648, + "loss": 2.3049, + "step": 1218 + }, + { + "epoch": 0.5591101937851164, + "grad_norm": 0.7060381130384076, + "learning_rate": 0.0004289094434668188, + "loss": 2.2657, + "step": 1219 + }, + { + "epoch": 0.559568856782479, + "grad_norm": 0.6960508620502024, + "learning_rate": 0.00042817402646118185, + "loss": 2.3703, + "step": 1220 + }, + { + "epoch": 0.5600275197798418, + "grad_norm": 0.6781663653367506, + "learning_rate": 0.0004274387680803936, + "loss": 2.3637, + "step": 1221 + }, + { + "epoch": 0.5604861827772044, + "grad_norm": 0.6888322154658024, + "learning_rate": 0.00042670366994824327, + "loss": 2.2906, + "step": 1222 + }, + { + "epoch": 0.5609448457745672, + "grad_norm": 0.6467322617780281, + "learning_rate": 0.0004259687336881663, + "loss": 2.2995, + "step": 1223 + }, + { + "epoch": 0.5614035087719298, + "grad_norm": 0.7206655303507443, + "learning_rate": 0.0004252339609232408, + "loss": 2.3316, + "step": 1224 + }, + { + "epoch": 0.5618621717692925, + "grad_norm": 0.7099170540515396, + "learning_rate": 0.0004244993532761834, + "loss": 2.2803, + "step": 1225 + }, + { + "epoch": 0.5623208347666552, + "grad_norm": 0.7255280395998805, + "learning_rate": 0.00042376491236934634, + "loss": 2.3226, + "step": 1226 + }, + { + "epoch": 0.5627794977640179, + "grad_norm": 0.6884966951415659, + "learning_rate": 0.0004230306398247136, + "loss": 2.2357, + "step": 1227 + }, + { + "epoch": 0.5632381607613806, + "grad_norm": 0.7107796122908977, + "learning_rate": 0.0004222965372638976, + "loss": 2.4063, + "step": 1228 + }, + { + "epoch": 0.5636968237587433, + "grad_norm": 0.6735320707956048, + "learning_rate": 0.0004215626063081348, + "loss": 2.426, + "step": 1229 + }, + { + "epoch": 0.5641554867561059, + "grad_norm": 0.842750958878517, + "learning_rate": 0.000420828848578283, + "loss": 2.3692, + "step": 1230 + }, + { + "epoch": 0.5646141497534687, + "grad_norm": 0.6762388112273582, + "learning_rate": 0.0004200952656948175, + "loss": 2.4301, + "step": 1231 + }, + { + "epoch": 0.5650728127508313, + "grad_norm": 0.6981693966528351, + "learning_rate": 0.0004193618592778272, + "loss": 2.3466, + "step": 1232 + }, + { + "epoch": 0.5655314757481941, + "grad_norm": 0.6785399387129915, + "learning_rate": 0.0004186286309470116, + "loss": 2.3113, + "step": 1233 + }, + { + "epoch": 0.5659901387455567, + "grad_norm": 0.6782486005994033, + "learning_rate": 0.0004178955823216767, + "loss": 2.3919, + "step": 1234 + }, + { + "epoch": 0.5664488017429193, + "grad_norm": 0.7180397314929104, + "learning_rate": 0.00041716271502073137, + "loss": 2.4534, + "step": 1235 + }, + { + "epoch": 0.5669074647402821, + "grad_norm": 0.7033731299649755, + "learning_rate": 0.000416430030662685, + "loss": 2.253, + "step": 1236 + }, + { + "epoch": 0.5673661277376447, + "grad_norm": 0.7080751340291378, + "learning_rate": 0.00041569753086564173, + "loss": 2.2827, + "step": 1237 + }, + { + "epoch": 0.5678247907350075, + "grad_norm": 0.7482022185181566, + "learning_rate": 0.0004149652172472988, + "loss": 2.352, + "step": 1238 + }, + { + "epoch": 0.5682834537323701, + "grad_norm": 0.6799072739415754, + "learning_rate": 0.00041423309142494234, + "loss": 2.3282, + "step": 1239 + }, + { + "epoch": 0.5687421167297328, + "grad_norm": 0.7069563677066022, + "learning_rate": 0.0004135011550154433, + "loss": 2.3111, + "step": 1240 + }, + { + "epoch": 0.5692007797270955, + "grad_norm": 0.685312966019222, + "learning_rate": 0.0004127694096352546, + "loss": 2.3473, + "step": 1241 + }, + { + "epoch": 0.5696594427244582, + "grad_norm": 0.6632591227239287, + "learning_rate": 0.00041203785690040743, + "loss": 2.3424, + "step": 1242 + }, + { + "epoch": 0.5701181057218209, + "grad_norm": 0.7209152846409713, + "learning_rate": 0.00041130649842650694, + "loss": 2.3037, + "step": 1243 + }, + { + "epoch": 0.5705767687191836, + "grad_norm": 0.7657571604085872, + "learning_rate": 0.00041057533582873016, + "loss": 2.3856, + "step": 1244 + }, + { + "epoch": 0.5710354317165462, + "grad_norm": 0.6951132996597695, + "learning_rate": 0.0004098443707218208, + "loss": 2.3664, + "step": 1245 + }, + { + "epoch": 0.571494094713909, + "grad_norm": 0.6978132498470553, + "learning_rate": 0.00040911360472008673, + "loss": 2.3899, + "step": 1246 + }, + { + "epoch": 0.5719527577112716, + "grad_norm": 0.6880579721663324, + "learning_rate": 0.0004083830394373959, + "loss": 2.3129, + "step": 1247 + }, + { + "epoch": 0.5724114207086344, + "grad_norm": 0.718818249822869, + "learning_rate": 0.00040765267648717324, + "loss": 2.3203, + "step": 1248 + }, + { + "epoch": 0.572870083705997, + "grad_norm": 0.6696319894215647, + "learning_rate": 0.00040692251748239677, + "loss": 2.2562, + "step": 1249 + }, + { + "epoch": 0.5733287467033598, + "grad_norm": 0.6942332904668204, + "learning_rate": 0.00040619256403559383, + "loss": 2.2084, + "step": 1250 + }, + { + "epoch": 0.5737874097007224, + "grad_norm": 0.7371432501943153, + "learning_rate": 0.000405462817758838, + "loss": 2.318, + "step": 1251 + }, + { + "epoch": 0.574246072698085, + "grad_norm": 0.7010221645412343, + "learning_rate": 0.0004047332802637457, + "loss": 2.2839, + "step": 1252 + }, + { + "epoch": 0.5747047356954478, + "grad_norm": 0.7352929811045607, + "learning_rate": 0.00040400395316147157, + "loss": 2.3196, + "step": 1253 + }, + { + "epoch": 0.5751633986928104, + "grad_norm": 0.7682321256220577, + "learning_rate": 0.00040327483806270627, + "loss": 2.407, + "step": 1254 + }, + { + "epoch": 0.5756220616901732, + "grad_norm": 0.6896339856940585, + "learning_rate": 0.0004025459365776715, + "loss": 2.265, + "step": 1255 + }, + { + "epoch": 0.5760807246875358, + "grad_norm": 0.6834062996424943, + "learning_rate": 0.00040181725031611794, + "loss": 2.4048, + "step": 1256 + }, + { + "epoch": 0.5765393876848985, + "grad_norm": 0.7406208386232527, + "learning_rate": 0.0004010887808873206, + "loss": 2.2792, + "step": 1257 + }, + { + "epoch": 0.5769980506822612, + "grad_norm": 0.6622154951996428, + "learning_rate": 0.00040036052990007553, + "loss": 2.3091, + "step": 1258 + }, + { + "epoch": 0.5774567136796239, + "grad_norm": 0.7264258145631778, + "learning_rate": 0.0003996324989626967, + "loss": 2.3006, + "step": 1259 + }, + { + "epoch": 0.5779153766769866, + "grad_norm": 0.6766353984897876, + "learning_rate": 0.00039890468968301166, + "loss": 2.2988, + "step": 1260 + }, + { + "epoch": 0.5783740396743493, + "grad_norm": 0.7465376338936989, + "learning_rate": 0.0003981771036683591, + "loss": 2.3588, + "step": 1261 + }, + { + "epoch": 0.5788327026717119, + "grad_norm": 0.7022099459601163, + "learning_rate": 0.00039744974252558385, + "loss": 2.291, + "step": 1262 + }, + { + "epoch": 0.5792913656690747, + "grad_norm": 0.7588946872287147, + "learning_rate": 0.00039672260786103463, + "loss": 2.386, + "step": 1263 + }, + { + "epoch": 0.5797500286664373, + "grad_norm": 0.6994653568622695, + "learning_rate": 0.00039599570128055994, + "loss": 2.3649, + "step": 1264 + }, + { + "epoch": 0.5802086916638001, + "grad_norm": 0.7494814942688619, + "learning_rate": 0.0003952690243895044, + "loss": 2.316, + "step": 1265 + }, + { + "epoch": 0.5806673546611627, + "grad_norm": 0.6896286621819717, + "learning_rate": 0.0003945425787927054, + "loss": 2.4498, + "step": 1266 + }, + { + "epoch": 0.5811260176585255, + "grad_norm": 0.6765097589398119, + "learning_rate": 0.00039381636609448975, + "loss": 2.2767, + "step": 1267 + }, + { + "epoch": 0.5815846806558881, + "grad_norm": 0.6990494059218245, + "learning_rate": 0.0003930903878986693, + "loss": 2.341, + "step": 1268 + }, + { + "epoch": 0.5820433436532507, + "grad_norm": 0.6692288503425996, + "learning_rate": 0.00039236464580853916, + "loss": 2.4137, + "step": 1269 + }, + { + "epoch": 0.5825020066506135, + "grad_norm": 0.6728912036801612, + "learning_rate": 0.0003916391414268718, + "loss": 2.3401, + "step": 1270 + }, + { + "epoch": 0.5829606696479761, + "grad_norm": 0.6775203155211199, + "learning_rate": 0.00039091387635591536, + "loss": 2.4168, + "step": 1271 + }, + { + "epoch": 0.5834193326453389, + "grad_norm": 0.659035538662857, + "learning_rate": 0.0003901888521973894, + "loss": 2.274, + "step": 1272 + }, + { + "epoch": 0.5838779956427015, + "grad_norm": 0.6810323049330841, + "learning_rate": 0.0003894640705524813, + "loss": 2.3478, + "step": 1273 + }, + { + "epoch": 0.5843366586400642, + "grad_norm": 0.7107754870888944, + "learning_rate": 0.00038873953302184284, + "loss": 2.205, + "step": 1274 + }, + { + "epoch": 0.5847953216374269, + "grad_norm": 0.7035080212239799, + "learning_rate": 0.000388015241205587, + "loss": 2.2621, + "step": 1275 + }, + { + "epoch": 0.5852539846347896, + "grad_norm": 0.7313853245203271, + "learning_rate": 0.00038729119670328355, + "loss": 2.3689, + "step": 1276 + }, + { + "epoch": 0.5857126476321523, + "grad_norm": 0.712850864356601, + "learning_rate": 0.00038656740111395665, + "loss": 2.3802, + "step": 1277 + }, + { + "epoch": 0.586171310629515, + "grad_norm": 0.7004334728817954, + "learning_rate": 0.00038584385603608053, + "loss": 2.3783, + "step": 1278 + }, + { + "epoch": 0.5866299736268776, + "grad_norm": 0.6995942780496917, + "learning_rate": 0.00038512056306757615, + "loss": 2.4014, + "step": 1279 + }, + { + "epoch": 0.5870886366242404, + "grad_norm": 0.6967734867444161, + "learning_rate": 0.0003843975238058075, + "loss": 2.2924, + "step": 1280 + }, + { + "epoch": 0.587547299621603, + "grad_norm": 0.6657248653751776, + "learning_rate": 0.00038367473984757863, + "loss": 2.3015, + "step": 1281 + }, + { + "epoch": 0.5880059626189658, + "grad_norm": 0.670934410496444, + "learning_rate": 0.0003829522127891296, + "loss": 2.3823, + "step": 1282 + }, + { + "epoch": 0.5884646256163284, + "grad_norm": 0.6867048305414355, + "learning_rate": 0.0003822299442261329, + "loss": 2.2071, + "step": 1283 + }, + { + "epoch": 0.5889232886136911, + "grad_norm": 0.6598298879681221, + "learning_rate": 0.00038150793575369063, + "loss": 2.349, + "step": 1284 + }, + { + "epoch": 0.5893819516110538, + "grad_norm": 0.6876545363844566, + "learning_rate": 0.0003807861889663299, + "loss": 2.2849, + "step": 1285 + }, + { + "epoch": 0.5898406146084164, + "grad_norm": 0.6994039766946237, + "learning_rate": 0.0003800647054580006, + "loss": 2.2769, + "step": 1286 + }, + { + "epoch": 0.5902992776057792, + "grad_norm": 1.9516831511743429, + "learning_rate": 0.00037934348682207064, + "loss": 2.3296, + "step": 1287 + }, + { + "epoch": 0.5907579406031418, + "grad_norm": 0.6847398830654833, + "learning_rate": 0.00037862253465132306, + "loss": 2.2752, + "step": 1288 + }, + { + "epoch": 0.5912166036005045, + "grad_norm": 0.71902602789842, + "learning_rate": 0.00037790185053795245, + "loss": 2.334, + "step": 1289 + }, + { + "epoch": 0.5916752665978672, + "grad_norm": 0.7342927337809919, + "learning_rate": 0.0003771814360735616, + "loss": 2.3947, + "step": 1290 + }, + { + "epoch": 0.5921339295952299, + "grad_norm": 0.7491386459647532, + "learning_rate": 0.00037646129284915755, + "loss": 2.2908, + "step": 1291 + }, + { + "epoch": 0.5925925925925926, + "grad_norm": 0.6938799472886247, + "learning_rate": 0.00037574142245514825, + "loss": 2.3442, + "step": 1292 + }, + { + "epoch": 0.5930512555899553, + "grad_norm": 0.6683065310907996, + "learning_rate": 0.0003750218264813393, + "loss": 2.3655, + "step": 1293 + }, + { + "epoch": 0.5935099185873179, + "grad_norm": 0.6484120505794201, + "learning_rate": 0.0003743025065169305, + "loss": 2.3628, + "step": 1294 + }, + { + "epoch": 0.5939685815846807, + "grad_norm": 0.6971058123094389, + "learning_rate": 0.0003735834641505116, + "loss": 2.4407, + "step": 1295 + }, + { + "epoch": 0.5944272445820433, + "grad_norm": 0.7382903165355585, + "learning_rate": 0.00037286470097005954, + "loss": 2.2844, + "step": 1296 + }, + { + "epoch": 0.5948859075794061, + "grad_norm": 0.6831112091676969, + "learning_rate": 0.0003721462185629347, + "loss": 2.2627, + "step": 1297 + }, + { + "epoch": 0.5953445705767687, + "grad_norm": 0.6794032431977197, + "learning_rate": 0.00037142801851587707, + "loss": 2.3131, + "step": 1298 + }, + { + "epoch": 0.5958032335741315, + "grad_norm": 0.6631618877856097, + "learning_rate": 0.00037071010241500357, + "loss": 2.3558, + "step": 1299 + }, + { + "epoch": 0.5962618965714941, + "grad_norm": 0.7140433606150681, + "learning_rate": 0.00036999247184580383, + "loss": 2.3254, + "step": 1300 + }, + { + "epoch": 0.5967205595688568, + "grad_norm": 0.6650187864530366, + "learning_rate": 0.00036927512839313636, + "loss": 2.3621, + "step": 1301 + }, + { + "epoch": 0.5971792225662195, + "grad_norm": 0.6736013049265582, + "learning_rate": 0.0003685580736412268, + "loss": 2.2881, + "step": 1302 + }, + { + "epoch": 0.5976378855635821, + "grad_norm": 0.6986964310904881, + "learning_rate": 0.000367841309173662, + "loss": 2.2806, + "step": 1303 + }, + { + "epoch": 0.5980965485609449, + "grad_norm": 0.6953416112072396, + "learning_rate": 0.0003671248365733883, + "loss": 2.2329, + "step": 1304 + }, + { + "epoch": 0.5985552115583075, + "grad_norm": 0.6915674519812777, + "learning_rate": 0.0003664086574227075, + "loss": 2.2349, + "step": 1305 + }, + { + "epoch": 0.5990138745556702, + "grad_norm": 0.6617971753176759, + "learning_rate": 0.000365692773303273, + "loss": 2.3351, + "step": 1306 + }, + { + "epoch": 0.5994725375530329, + "grad_norm": 0.7015799721851648, + "learning_rate": 0.00036497718579608696, + "loss": 2.2759, + "step": 1307 + }, + { + "epoch": 0.5999312005503956, + "grad_norm": 0.6798821267557863, + "learning_rate": 0.0003642618964814964, + "loss": 2.3401, + "step": 1308 + }, + { + "epoch": 0.6003898635477583, + "grad_norm": 0.7588807289573614, + "learning_rate": 0.00036354690693918946, + "loss": 2.4204, + "step": 1309 + }, + { + "epoch": 0.600848526545121, + "grad_norm": 0.6585922525332214, + "learning_rate": 0.00036283221874819284, + "loss": 2.352, + "step": 1310 + }, + { + "epoch": 0.6013071895424836, + "grad_norm": 0.797475144486915, + "learning_rate": 0.0003621178334868672, + "loss": 2.3544, + "step": 1311 + }, + { + "epoch": 0.6017658525398464, + "grad_norm": 0.6998354814420867, + "learning_rate": 0.00036140375273290476, + "loss": 2.3377, + "step": 1312 + }, + { + "epoch": 0.602224515537209, + "grad_norm": 0.7100868565664069, + "learning_rate": 0.0003606899780633245, + "loss": 2.2323, + "step": 1313 + }, + { + "epoch": 0.6026831785345718, + "grad_norm": 0.718252817446738, + "learning_rate": 0.0003599765110544699, + "loss": 2.2253, + "step": 1314 + }, + { + "epoch": 0.6031418415319344, + "grad_norm": 0.7436115929150479, + "learning_rate": 0.0003592633532820052, + "loss": 2.2919, + "step": 1315 + }, + { + "epoch": 0.6036005045292971, + "grad_norm": 0.7096479543803113, + "learning_rate": 0.0003585505063209109, + "loss": 2.3649, + "step": 1316 + }, + { + "epoch": 0.6040591675266598, + "grad_norm": 0.6622857132318672, + "learning_rate": 0.00035783797174548194, + "loss": 2.331, + "step": 1317 + }, + { + "epoch": 0.6045178305240225, + "grad_norm": 0.677580511365933, + "learning_rate": 0.00035712575112932277, + "loss": 2.2558, + "step": 1318 + }, + { + "epoch": 0.6049764935213852, + "grad_norm": 0.7059256649323506, + "learning_rate": 0.000356413846045345, + "loss": 2.4057, + "step": 1319 + }, + { + "epoch": 0.6054351565187478, + "grad_norm": 0.6830675435259801, + "learning_rate": 0.000355702258065763, + "loss": 2.3249, + "step": 1320 + }, + { + "epoch": 0.6058938195161105, + "grad_norm": 0.7250191788094296, + "learning_rate": 0.0003549909887620909, + "loss": 2.2954, + "step": 1321 + }, + { + "epoch": 0.6063524825134732, + "grad_norm": 0.689519138854757, + "learning_rate": 0.00035428003970513914, + "loss": 2.3726, + "step": 1322 + }, + { + "epoch": 0.6068111455108359, + "grad_norm": 0.6678198182750561, + "learning_rate": 0.00035356941246501085, + "loss": 2.3754, + "step": 1323 + }, + { + "epoch": 0.6072698085081986, + "grad_norm": 0.7339960233723217, + "learning_rate": 0.0003528591086110984, + "loss": 2.3204, + "step": 1324 + }, + { + "epoch": 0.6077284715055613, + "grad_norm": 0.6773514807330608, + "learning_rate": 0.00035214912971208, + "loss": 2.2722, + "step": 1325 + }, + { + "epoch": 0.6081871345029239, + "grad_norm": 0.6534596702930112, + "learning_rate": 0.0003514394773359163, + "loss": 2.1326, + "step": 1326 + }, + { + "epoch": 0.6086457975002867, + "grad_norm": 0.7450658807231055, + "learning_rate": 0.0003507301530498469, + "loss": 2.3055, + "step": 1327 + }, + { + "epoch": 0.6091044604976493, + "grad_norm": 0.6420517868724647, + "learning_rate": 0.00035002115842038646, + "loss": 2.3093, + "step": 1328 + }, + { + "epoch": 0.6095631234950121, + "grad_norm": 0.660783369547422, + "learning_rate": 0.00034931249501332195, + "loss": 2.3299, + "step": 1329 + }, + { + "epoch": 0.6100217864923747, + "grad_norm": 0.6933764271639634, + "learning_rate": 0.00034860416439370885, + "loss": 2.3505, + "step": 1330 + }, + { + "epoch": 0.6104804494897375, + "grad_norm": 0.6555327813503782, + "learning_rate": 0.0003478961681258674, + "loss": 2.3502, + "step": 1331 + }, + { + "epoch": 0.6109391124871001, + "grad_norm": 0.6404969644787613, + "learning_rate": 0.0003471885077733796, + "loss": 2.2823, + "step": 1332 + }, + { + "epoch": 0.6113977754844628, + "grad_norm": 0.7081736220029167, + "learning_rate": 0.0003464811848990859, + "loss": 2.3579, + "step": 1333 + }, + { + "epoch": 0.6118564384818255, + "grad_norm": 0.7011413335901163, + "learning_rate": 0.00034577420106508063, + "loss": 2.3678, + "step": 1334 + }, + { + "epoch": 0.6123151014791882, + "grad_norm": 0.6774955065377761, + "learning_rate": 0.0003450675578327105, + "loss": 2.3329, + "step": 1335 + }, + { + "epoch": 0.6127737644765509, + "grad_norm": 0.7599102927475775, + "learning_rate": 0.000344361256762569, + "loss": 2.3473, + "step": 1336 + }, + { + "epoch": 0.6132324274739135, + "grad_norm": 0.7480770152289044, + "learning_rate": 0.00034365529941449456, + "loss": 2.2175, + "step": 1337 + }, + { + "epoch": 0.6136910904712762, + "grad_norm": 0.707226107938764, + "learning_rate": 0.0003429496873475664, + "loss": 2.2702, + "step": 1338 + }, + { + "epoch": 0.6141497534686389, + "grad_norm": 0.6782569295732697, + "learning_rate": 0.0003422444221201009, + "loss": 2.2419, + "step": 1339 + }, + { + "epoch": 0.6146084164660016, + "grad_norm": 0.7823454973778349, + "learning_rate": 0.0003415395052896487, + "loss": 2.3692, + "step": 1340 + }, + { + "epoch": 0.6150670794633643, + "grad_norm": 0.7113106856858624, + "learning_rate": 0.0003408349384129912, + "loss": 2.3468, + "step": 1341 + }, + { + "epoch": 0.615525742460727, + "grad_norm": 0.6983556965071304, + "learning_rate": 0.00034013072304613643, + "loss": 2.3249, + "step": 1342 + }, + { + "epoch": 0.6159844054580896, + "grad_norm": 0.7003663907707655, + "learning_rate": 0.00033942686074431674, + "loss": 2.3993, + "step": 1343 + }, + { + "epoch": 0.6164430684554524, + "grad_norm": 0.6763691715786097, + "learning_rate": 0.0003387233530619843, + "loss": 2.3467, + "step": 1344 + }, + { + "epoch": 0.616901731452815, + "grad_norm": 0.6733244132596565, + "learning_rate": 0.0003380202015528084, + "loss": 2.4259, + "step": 1345 + }, + { + "epoch": 0.6173603944501778, + "grad_norm": 0.6507350143797076, + "learning_rate": 0.0003373174077696715, + "loss": 2.365, + "step": 1346 + }, + { + "epoch": 0.6178190574475404, + "grad_norm": 0.6597280929867216, + "learning_rate": 0.0003366149732646661, + "loss": 2.2618, + "step": 1347 + }, + { + "epoch": 0.6182777204449031, + "grad_norm": 0.6593781442956078, + "learning_rate": 0.00033591289958909143, + "loss": 2.2554, + "step": 1348 + }, + { + "epoch": 0.6187363834422658, + "grad_norm": 0.6786694867337414, + "learning_rate": 0.00033521118829344954, + "loss": 2.2603, + "step": 1349 + }, + { + "epoch": 0.6191950464396285, + "grad_norm": 0.6891214205969167, + "learning_rate": 0.0003345098409274423, + "loss": 2.449, + "step": 1350 + }, + { + "epoch": 0.6196537094369912, + "grad_norm": 0.6680739642915868, + "learning_rate": 0.00033380885903996796, + "loss": 2.3012, + "step": 1351 + }, + { + "epoch": 0.6201123724343539, + "grad_norm": 0.6757996408571035, + "learning_rate": 0.00033310824417911766, + "loss": 2.3606, + "step": 1352 + }, + { + "epoch": 0.6205710354317165, + "grad_norm": 0.7220986890947022, + "learning_rate": 0.00033240799789217184, + "loss": 2.3378, + "step": 1353 + }, + { + "epoch": 0.6210296984290792, + "grad_norm": 0.6645203427945862, + "learning_rate": 0.00033170812172559694, + "loss": 2.2567, + "step": 1354 + }, + { + "epoch": 0.6214883614264419, + "grad_norm": 0.7006455055505116, + "learning_rate": 0.000331008617225042, + "loss": 2.3089, + "step": 1355 + }, + { + "epoch": 0.6219470244238046, + "grad_norm": 0.7170994010989409, + "learning_rate": 0.0003303094859353355, + "loss": 2.2531, + "step": 1356 + }, + { + "epoch": 0.6224056874211673, + "grad_norm": 0.7049755171317864, + "learning_rate": 0.0003296107294004812, + "loss": 2.1684, + "step": 1357 + }, + { + "epoch": 0.6228643504185299, + "grad_norm": 0.7140444416241413, + "learning_rate": 0.0003289123491636559, + "loss": 2.3006, + "step": 1358 + }, + { + "epoch": 0.6233230134158927, + "grad_norm": 0.6955459611123781, + "learning_rate": 0.00032821434676720443, + "loss": 2.4244, + "step": 1359 + }, + { + "epoch": 0.6237816764132553, + "grad_norm": 0.7342507054454862, + "learning_rate": 0.00032751672375263836, + "loss": 2.3393, + "step": 1360 + }, + { + "epoch": 0.6242403394106181, + "grad_norm": 0.6846758015994775, + "learning_rate": 0.0003268194816606305, + "loss": 2.3733, + "step": 1361 + }, + { + "epoch": 0.6246990024079807, + "grad_norm": 0.7130988374710049, + "learning_rate": 0.00032612262203101267, + "loss": 2.3048, + "step": 1362 + }, + { + "epoch": 0.6251576654053435, + "grad_norm": 0.674617941641313, + "learning_rate": 0.00032542614640277225, + "loss": 2.204, + "step": 1363 + }, + { + "epoch": 0.6256163284027061, + "grad_norm": 0.665595626219869, + "learning_rate": 0.0003247300563140481, + "loss": 2.2385, + "step": 1364 + }, + { + "epoch": 0.6260749914000688, + "grad_norm": 0.6757511267111173, + "learning_rate": 0.00032403435330212807, + "loss": 2.3498, + "step": 1365 + }, + { + "epoch": 0.6265336543974315, + "grad_norm": 0.6979058598585912, + "learning_rate": 0.00032333903890344515, + "loss": 2.2587, + "step": 1366 + }, + { + "epoch": 0.6269923173947942, + "grad_norm": 0.6822910929713205, + "learning_rate": 0.00032264411465357333, + "loss": 2.362, + "step": 1367 + }, + { + "epoch": 0.6274509803921569, + "grad_norm": 0.6871826132055223, + "learning_rate": 0.00032194958208722654, + "loss": 2.2413, + "step": 1368 + }, + { + "epoch": 0.6279096433895196, + "grad_norm": 0.6795917289359783, + "learning_rate": 0.00032125544273825204, + "loss": 2.2414, + "step": 1369 + }, + { + "epoch": 0.6283683063868822, + "grad_norm": 0.679367895349414, + "learning_rate": 0.0003205616981396297, + "loss": 2.3959, + "step": 1370 + }, + { + "epoch": 0.6288269693842449, + "grad_norm": 0.6631693633595794, + "learning_rate": 0.00031986834982346713, + "loss": 2.2398, + "step": 1371 + }, + { + "epoch": 0.6292856323816076, + "grad_norm": 0.7514597638626956, + "learning_rate": 0.00031917539932099694, + "loss": 2.2948, + "step": 1372 + }, + { + "epoch": 0.6297442953789703, + "grad_norm": 0.7269843393780507, + "learning_rate": 0.00031848284816257336, + "loss": 2.3627, + "step": 1373 + }, + { + "epoch": 0.630202958376333, + "grad_norm": 0.664041049919249, + "learning_rate": 0.0003177906978776682, + "loss": 2.3081, + "step": 1374 + }, + { + "epoch": 0.6306616213736956, + "grad_norm": 0.6721633849535043, + "learning_rate": 0.0003170989499948683, + "loss": 2.3711, + "step": 1375 + }, + { + "epoch": 0.6311202843710584, + "grad_norm": 0.6977704927696363, + "learning_rate": 0.0003164076060418719, + "loss": 2.3527, + "step": 1376 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 0.6964149392201732, + "learning_rate": 0.000315716667545485, + "loss": 2.201, + "step": 1377 + }, + { + "epoch": 0.6320376103657838, + "grad_norm": 0.7416782215234252, + "learning_rate": 0.00031502613603161836, + "loss": 2.3357, + "step": 1378 + }, + { + "epoch": 0.6324962733631464, + "grad_norm": 0.6835408841884674, + "learning_rate": 0.00031433601302528335, + "loss": 2.2971, + "step": 1379 + }, + { + "epoch": 0.6329549363605091, + "grad_norm": 0.7037701966632788, + "learning_rate": 0.00031364630005058995, + "loss": 2.2255, + "step": 1380 + }, + { + "epoch": 0.6334135993578718, + "grad_norm": 0.7662744408704222, + "learning_rate": 0.0003129569986307422, + "loss": 2.3846, + "step": 1381 + }, + { + "epoch": 0.6338722623552345, + "grad_norm": 0.671926946514943, + "learning_rate": 0.00031226811028803515, + "loss": 2.2567, + "step": 1382 + }, + { + "epoch": 0.6343309253525972, + "grad_norm": 0.714244190651207, + "learning_rate": 0.00031157963654385173, + "loss": 2.3809, + "step": 1383 + }, + { + "epoch": 0.6347895883499599, + "grad_norm": 0.7231186182109107, + "learning_rate": 0.0003108915789186592, + "loss": 2.3606, + "step": 1384 + }, + { + "epoch": 0.6352482513473225, + "grad_norm": 0.6779057465825709, + "learning_rate": 0.00031020393893200604, + "loss": 2.3187, + "step": 1385 + }, + { + "epoch": 0.6357069143446853, + "grad_norm": 0.7143708333110406, + "learning_rate": 0.00030951671810251823, + "loss": 2.3175, + "step": 1386 + }, + { + "epoch": 0.6361655773420479, + "grad_norm": 0.7300914917449114, + "learning_rate": 0.0003088299179478959, + "loss": 2.3206, + "step": 1387 + }, + { + "epoch": 0.6366242403394106, + "grad_norm": 0.7171236372908123, + "learning_rate": 0.0003081435399849104, + "loss": 2.351, + "step": 1388 + }, + { + "epoch": 0.6370829033367733, + "grad_norm": 0.7413763296998042, + "learning_rate": 0.0003074575857294004, + "loss": 2.3002, + "step": 1389 + }, + { + "epoch": 0.637541566334136, + "grad_norm": 0.7343884133569856, + "learning_rate": 0.0003067720566962691, + "loss": 2.3177, + "step": 1390 + }, + { + "epoch": 0.6380002293314987, + "grad_norm": 0.709965969321868, + "learning_rate": 0.0003060869543994806, + "loss": 2.3988, + "step": 1391 + }, + { + "epoch": 0.6384588923288613, + "grad_norm": 0.6849858517100875, + "learning_rate": 0.0003054022803520562, + "loss": 2.2955, + "step": 1392 + }, + { + "epoch": 0.6389175553262241, + "grad_norm": 0.7230456131306704, + "learning_rate": 0.0003047180360660721, + "loss": 2.3263, + "step": 1393 + }, + { + "epoch": 0.6393762183235867, + "grad_norm": 0.7117439509678791, + "learning_rate": 0.00030403422305265475, + "loss": 2.2548, + "step": 1394 + }, + { + "epoch": 0.6398348813209495, + "grad_norm": 0.7195355490531298, + "learning_rate": 0.0003033508428219785, + "loss": 2.3267, + "step": 1395 + }, + { + "epoch": 0.6402935443183121, + "grad_norm": 0.7161719515598243, + "learning_rate": 0.00030266789688326184, + "loss": 2.3962, + "step": 1396 + }, + { + "epoch": 0.6407522073156748, + "grad_norm": 0.7019798798985919, + "learning_rate": 0.00030198538674476393, + "loss": 2.2897, + "step": 1397 + }, + { + "epoch": 0.6412108703130375, + "grad_norm": 0.6724188733350598, + "learning_rate": 0.00030130331391378185, + "loss": 2.2396, + "step": 1398 + }, + { + "epoch": 0.6416695333104002, + "grad_norm": 0.7189127394838241, + "learning_rate": 0.0003006216798966468, + "loss": 2.2738, + "step": 1399 + }, + { + "epoch": 0.6421281963077629, + "grad_norm": 0.6574773140151825, + "learning_rate": 0.00029994048619872034, + "loss": 2.2793, + "step": 1400 + }, + { + "epoch": 0.6425868593051256, + "grad_norm": 0.6772776919193734, + "learning_rate": 0.0002992597343243927, + "loss": 2.2438, + "step": 1401 + }, + { + "epoch": 0.6430455223024882, + "grad_norm": 0.6570584387666335, + "learning_rate": 0.0002985794257770773, + "loss": 2.3869, + "step": 1402 + }, + { + "epoch": 0.643504185299851, + "grad_norm": 0.7546935510622429, + "learning_rate": 0.0002978995620592092, + "loss": 2.4059, + "step": 1403 + }, + { + "epoch": 0.6439628482972136, + "grad_norm": 0.7176667239152166, + "learning_rate": 0.0002972201446722405, + "loss": 2.3374, + "step": 1404 + }, + { + "epoch": 0.6444215112945763, + "grad_norm": 0.6852076763468496, + "learning_rate": 0.00029654117511663803, + "loss": 2.2796, + "step": 1405 + }, + { + "epoch": 0.644880174291939, + "grad_norm": 0.7567422766470125, + "learning_rate": 0.0002958626548918795, + "loss": 2.2153, + "step": 1406 + }, + { + "epoch": 0.6453388372893016, + "grad_norm": 0.6476622620320833, + "learning_rate": 0.00029518458549645014, + "loss": 2.4081, + "step": 1407 + }, + { + "epoch": 0.6457975002866644, + "grad_norm": 0.7549309461854302, + "learning_rate": 0.00029450696842783954, + "loss": 2.2608, + "step": 1408 + }, + { + "epoch": 0.646256163284027, + "grad_norm": 0.708390163473901, + "learning_rate": 0.00029382980518253865, + "loss": 2.3014, + "step": 1409 + }, + { + "epoch": 0.6467148262813898, + "grad_norm": 0.7607780975627342, + "learning_rate": 0.00029315309725603595, + "loss": 2.3411, + "step": 1410 + }, + { + "epoch": 0.6471734892787524, + "grad_norm": 0.6718877840657892, + "learning_rate": 0.00029247684614281446, + "loss": 2.2888, + "step": 1411 + }, + { + "epoch": 0.6476321522761151, + "grad_norm": 0.6928210781566306, + "learning_rate": 0.0002918010533363481, + "loss": 2.331, + "step": 1412 + }, + { + "epoch": 0.6480908152734778, + "grad_norm": 0.6676399372003173, + "learning_rate": 0.0002911257203290987, + "loss": 2.2626, + "step": 1413 + }, + { + "epoch": 0.6485494782708405, + "grad_norm": 0.8024207350450154, + "learning_rate": 0.00029045084861251314, + "loss": 2.3826, + "step": 1414 + }, + { + "epoch": 0.6490081412682032, + "grad_norm": 0.6749840649975564, + "learning_rate": 0.00028977643967701897, + "loss": 2.3767, + "step": 1415 + }, + { + "epoch": 0.6494668042655659, + "grad_norm": 0.7418975250817433, + "learning_rate": 0.00028910249501202156, + "loss": 2.2954, + "step": 1416 + }, + { + "epoch": 0.6499254672629285, + "grad_norm": 0.6907355853788575, + "learning_rate": 0.00028842901610590165, + "loss": 2.3502, + "step": 1417 + }, + { + "epoch": 0.6503841302602913, + "grad_norm": 0.69573005961123, + "learning_rate": 0.00028775600444601123, + "loss": 2.3803, + "step": 1418 + }, + { + "epoch": 0.6508427932576539, + "grad_norm": 0.6837063255413205, + "learning_rate": 0.00028708346151866973, + "loss": 2.2455, + "step": 1419 + }, + { + "epoch": 0.6513014562550167, + "grad_norm": 0.7961410972064091, + "learning_rate": 0.0002864113888091622, + "loss": 2.3553, + "step": 1420 + }, + { + "epoch": 0.6517601192523793, + "grad_norm": 0.6965613671258876, + "learning_rate": 0.0002857397878017348, + "loss": 2.3028, + "step": 1421 + }, + { + "epoch": 0.6522187822497421, + "grad_norm": 0.6865352367060681, + "learning_rate": 0.00028506865997959173, + "loss": 2.292, + "step": 1422 + }, + { + "epoch": 0.6526774452471047, + "grad_norm": 0.7363852489754819, + "learning_rate": 0.000284398006824893, + "loss": 2.387, + "step": 1423 + }, + { + "epoch": 0.6531361082444673, + "grad_norm": 0.6842606444587346, + "learning_rate": 0.00028372782981874963, + "loss": 2.267, + "step": 1424 + }, + { + "epoch": 0.6535947712418301, + "grad_norm": 0.7067310419378116, + "learning_rate": 0.00028305813044122096, + "loss": 2.2995, + "step": 1425 + }, + { + "epoch": 0.6540534342391927, + "grad_norm": 0.6765078514849194, + "learning_rate": 0.0002823889101713122, + "loss": 2.2967, + "step": 1426 + }, + { + "epoch": 0.6545120972365555, + "grad_norm": 0.6352108452069228, + "learning_rate": 0.0002817201704869701, + "loss": 2.3798, + "step": 1427 + }, + { + "epoch": 0.6549707602339181, + "grad_norm": 0.7617784743837548, + "learning_rate": 0.00028105191286508, + "loss": 2.4201, + "step": 1428 + }, + { + "epoch": 0.6554294232312808, + "grad_norm": 0.7123512358986197, + "learning_rate": 0.00028038413878146245, + "loss": 2.2283, + "step": 1429 + }, + { + "epoch": 0.6558880862286435, + "grad_norm": 0.7348188573353313, + "learning_rate": 0.00027971684971087073, + "loss": 2.2221, + "step": 1430 + }, + { + "epoch": 0.6563467492260062, + "grad_norm": 0.6709455508177935, + "learning_rate": 0.00027905004712698643, + "loss": 2.2975, + "step": 1431 + }, + { + "epoch": 0.6568054122233689, + "grad_norm": 0.7017898248043621, + "learning_rate": 0.0002783837325024167, + "loss": 2.3457, + "step": 1432 + }, + { + "epoch": 0.6572640752207316, + "grad_norm": 0.7455945038359215, + "learning_rate": 0.00027771790730869153, + "loss": 2.3635, + "step": 1433 + }, + { + "epoch": 0.6577227382180942, + "grad_norm": 0.7489767127326602, + "learning_rate": 0.0002770525730162599, + "loss": 2.2816, + "step": 1434 + }, + { + "epoch": 0.658181401215457, + "grad_norm": 0.7358315878311293, + "learning_rate": 0.00027638773109448645, + "loss": 2.2487, + "step": 1435 + }, + { + "epoch": 0.6586400642128196, + "grad_norm": 0.744022754287691, + "learning_rate": 0.00027572338301164824, + "loss": 2.3815, + "step": 1436 + }, + { + "epoch": 0.6590987272101824, + "grad_norm": 0.7005147230075939, + "learning_rate": 0.0002750595302349324, + "loss": 2.2729, + "step": 1437 + }, + { + "epoch": 0.659557390207545, + "grad_norm": 0.7174726159220899, + "learning_rate": 0.00027439617423043145, + "loss": 2.279, + "step": 1438 + }, + { + "epoch": 0.6600160532049077, + "grad_norm": 0.7310511836931458, + "learning_rate": 0.00027373331646314114, + "loss": 2.2855, + "step": 1439 + }, + { + "epoch": 0.6604747162022704, + "grad_norm": 0.6720011150244805, + "learning_rate": 0.0002730709583969572, + "loss": 2.2603, + "step": 1440 + }, + { + "epoch": 0.660933379199633, + "grad_norm": 0.6822150492132332, + "learning_rate": 0.0002724091014946711, + "loss": 2.4198, + "step": 1441 + }, + { + "epoch": 0.6613920421969958, + "grad_norm": 0.7086146983051445, + "learning_rate": 0.00027174774721796824, + "loss": 2.3251, + "step": 1442 + }, + { + "epoch": 0.6618507051943584, + "grad_norm": 0.6738489408920738, + "learning_rate": 0.0002710868970274232, + "loss": 2.2453, + "step": 1443 + }, + { + "epoch": 0.6623093681917211, + "grad_norm": 0.6985792686390668, + "learning_rate": 0.0002704265523824982, + "loss": 2.3071, + "step": 1444 + }, + { + "epoch": 0.6627680311890838, + "grad_norm": 0.67106556691538, + "learning_rate": 0.00026976671474153826, + "loss": 2.3597, + "step": 1445 + }, + { + "epoch": 0.6632266941864465, + "grad_norm": 0.7449061256511186, + "learning_rate": 0.00026910738556176886, + "loss": 2.2751, + "step": 1446 + }, + { + "epoch": 0.6636853571838092, + "grad_norm": 0.7538274722108518, + "learning_rate": 0.0002684485662992929, + "loss": 2.3791, + "step": 1447 + }, + { + "epoch": 0.6641440201811719, + "grad_norm": 0.6718804128026443, + "learning_rate": 0.0002677902584090869, + "loss": 2.3392, + "step": 1448 + }, + { + "epoch": 0.6646026831785345, + "grad_norm": 0.701447152753004, + "learning_rate": 0.00026713246334499774, + "loss": 2.2366, + "step": 1449 + }, + { + "epoch": 0.6650613461758973, + "grad_norm": 0.7378058289711878, + "learning_rate": 0.00026647518255974023, + "loss": 2.2611, + "step": 1450 + }, + { + "epoch": 0.6655200091732599, + "grad_norm": 0.8074619717733589, + "learning_rate": 0.0002658184175048934, + "loss": 2.2763, + "step": 1451 + }, + { + "epoch": 0.6659786721706227, + "grad_norm": 0.7423671064929601, + "learning_rate": 0.00026516216963089694, + "loss": 2.3237, + "step": 1452 + }, + { + "epoch": 0.6664373351679853, + "grad_norm": 0.6568136374466864, + "learning_rate": 0.0002645064403870488, + "loss": 2.3512, + "step": 1453 + }, + { + "epoch": 0.6668959981653481, + "grad_norm": 0.6721130753693918, + "learning_rate": 0.0002638512312215011, + "loss": 2.3041, + "step": 1454 + }, + { + "epoch": 0.6673546611627107, + "grad_norm": 0.6775128556961096, + "learning_rate": 0.0002631965435812575, + "loss": 2.2856, + "step": 1455 + }, + { + "epoch": 0.6678133241600734, + "grad_norm": 0.6797906166521944, + "learning_rate": 0.00026254237891217046, + "loss": 2.2911, + "step": 1456 + }, + { + "epoch": 0.6682719871574361, + "grad_norm": 0.6650034013117075, + "learning_rate": 0.0002618887386589367, + "loss": 2.3377, + "step": 1457 + }, + { + "epoch": 0.6687306501547987, + "grad_norm": 0.693136186943613, + "learning_rate": 0.0002612356242650949, + "loss": 2.2786, + "step": 1458 + }, + { + "epoch": 0.6691893131521615, + "grad_norm": 0.7169573638777681, + "learning_rate": 0.0002605830371730229, + "loss": 2.1696, + "step": 1459 + }, + { + "epoch": 0.6696479761495241, + "grad_norm": 0.7633834628142501, + "learning_rate": 0.0002599309788239339, + "loss": 2.3714, + "step": 1460 + }, + { + "epoch": 0.6701066391468868, + "grad_norm": 0.7030539978486471, + "learning_rate": 0.00025927945065787306, + "loss": 2.2599, + "step": 1461 + }, + { + "epoch": 0.6705653021442495, + "grad_norm": 0.7387171104301649, + "learning_rate": 0.0002586284541137145, + "loss": 2.3463, + "step": 1462 + }, + { + "epoch": 0.6710239651416122, + "grad_norm": 0.7073451869003547, + "learning_rate": 0.00025797799062915905, + "loss": 2.3298, + "step": 1463 + }, + { + "epoch": 0.6714826281389749, + "grad_norm": 0.6761913729964026, + "learning_rate": 0.00025732806164072966, + "loss": 2.2944, + "step": 1464 + }, + { + "epoch": 0.6719412911363376, + "grad_norm": 0.6845179181256706, + "learning_rate": 0.00025667866858376874, + "loss": 2.336, + "step": 1465 + }, + { + "epoch": 0.6723999541337002, + "grad_norm": 0.6893630933701841, + "learning_rate": 0.0002560298128924358, + "loss": 2.3568, + "step": 1466 + }, + { + "epoch": 0.672858617131063, + "grad_norm": 0.7536520987211416, + "learning_rate": 0.0002553814959997032, + "loss": 2.3368, + "step": 1467 + }, + { + "epoch": 0.6733172801284256, + "grad_norm": 0.7370501335495139, + "learning_rate": 0.00025473371933735334, + "loss": 2.2117, + "step": 1468 + }, + { + "epoch": 0.6737759431257884, + "grad_norm": 0.7039240314545293, + "learning_rate": 0.00025408648433597534, + "loss": 2.3491, + "step": 1469 + }, + { + "epoch": 0.674234606123151, + "grad_norm": 0.6764495944340776, + "learning_rate": 0.00025343979242496283, + "loss": 2.2561, + "step": 1470 + }, + { + "epoch": 0.6746932691205138, + "grad_norm": 0.7062421450547706, + "learning_rate": 0.00025279364503250925, + "loss": 2.3999, + "step": 1471 + }, + { + "epoch": 0.6751519321178764, + "grad_norm": 0.7145159959372371, + "learning_rate": 0.0002521480435856056, + "loss": 2.2935, + "step": 1472 + }, + { + "epoch": 0.6756105951152391, + "grad_norm": 0.7585334926834822, + "learning_rate": 0.0002515029895100378, + "loss": 2.2065, + "step": 1473 + }, + { + "epoch": 0.6760692581126018, + "grad_norm": 0.7360862874329702, + "learning_rate": 0.0002508584842303822, + "loss": 2.4099, + "step": 1474 + }, + { + "epoch": 0.6765279211099644, + "grad_norm": 0.6440346206211753, + "learning_rate": 0.0002502145291700038, + "loss": 2.25, + "step": 1475 + }, + { + "epoch": 0.6769865841073271, + "grad_norm": 0.7731484324246132, + "learning_rate": 0.0002495711257510517, + "loss": 2.3535, + "step": 1476 + }, + { + "epoch": 0.6774452471046898, + "grad_norm": 0.757998381135436, + "learning_rate": 0.0002489282753944575, + "loss": 2.3402, + "step": 1477 + }, + { + "epoch": 0.6779039101020525, + "grad_norm": 0.7005613956923955, + "learning_rate": 0.00024828597951993093, + "loss": 2.2542, + "step": 1478 + }, + { + "epoch": 0.6783625730994152, + "grad_norm": 0.6686723252100304, + "learning_rate": 0.00024764423954595706, + "loss": 2.2811, + "step": 1479 + }, + { + "epoch": 0.6788212360967779, + "grad_norm": 0.6955744821973517, + "learning_rate": 0.0002470030568897938, + "loss": 2.2695, + "step": 1480 + }, + { + "epoch": 0.6792798990941405, + "grad_norm": 0.6935295730377267, + "learning_rate": 0.00024636243296746773, + "loss": 2.3029, + "step": 1481 + }, + { + "epoch": 0.6797385620915033, + "grad_norm": 0.7082334434053319, + "learning_rate": 0.0002457223691937716, + "loss": 2.2529, + "step": 1482 + }, + { + "epoch": 0.6801972250888659, + "grad_norm": 0.6872232111312099, + "learning_rate": 0.0002450828669822613, + "loss": 2.2906, + "step": 1483 + }, + { + "epoch": 0.6806558880862287, + "grad_norm": 0.7094130263124537, + "learning_rate": 0.00024444392774525253, + "loss": 2.2718, + "step": 1484 + }, + { + "epoch": 0.6811145510835913, + "grad_norm": 0.6919237134692277, + "learning_rate": 0.00024380555289381733, + "loss": 2.3635, + "step": 1485 + }, + { + "epoch": 0.6815732140809541, + "grad_norm": 0.6980621335297547, + "learning_rate": 0.00024316774383778184, + "loss": 2.2872, + "step": 1486 + }, + { + "epoch": 0.6820318770783167, + "grad_norm": 0.691914693322672, + "learning_rate": 0.0002425305019857222, + "loss": 2.2933, + "step": 1487 + }, + { + "epoch": 0.6824905400756794, + "grad_norm": 0.6940946575238147, + "learning_rate": 0.00024189382874496184, + "loss": 2.335, + "step": 1488 + }, + { + "epoch": 0.6829492030730421, + "grad_norm": 0.7045280922298124, + "learning_rate": 0.00024125772552156916, + "loss": 2.3357, + "step": 1489 + }, + { + "epoch": 0.6834078660704048, + "grad_norm": 0.6951363255275695, + "learning_rate": 0.00024062219372035292, + "loss": 2.348, + "step": 1490 + }, + { + "epoch": 0.6838665290677675, + "grad_norm": 0.6686591994306005, + "learning_rate": 0.00023998723474486007, + "loss": 2.3394, + "step": 1491 + }, + { + "epoch": 0.6843251920651301, + "grad_norm": 0.7174084104305413, + "learning_rate": 0.00023935284999737272, + "loss": 2.3498, + "step": 1492 + }, + { + "epoch": 0.6847838550624928, + "grad_norm": 0.7492673361826145, + "learning_rate": 0.00023871904087890505, + "loss": 2.3221, + "step": 1493 + }, + { + "epoch": 0.6852425180598555, + "grad_norm": 0.6843889199060997, + "learning_rate": 0.00023808580878919945, + "loss": 2.3247, + "step": 1494 + }, + { + "epoch": 0.6857011810572182, + "grad_norm": 0.7078104768158258, + "learning_rate": 0.00023745315512672398, + "loss": 2.2778, + "step": 1495 + }, + { + "epoch": 0.6861598440545809, + "grad_norm": 0.6705317839582818, + "learning_rate": 0.0002368210812886698, + "loss": 2.2546, + "step": 1496 + }, + { + "epoch": 0.6866185070519436, + "grad_norm": 0.7051157281342358, + "learning_rate": 0.0002361895886709471, + "loss": 2.3483, + "step": 1497 + }, + { + "epoch": 0.6870771700493062, + "grad_norm": 0.7241784280163879, + "learning_rate": 0.0002355586786681823, + "loss": 2.2995, + "step": 1498 + }, + { + "epoch": 0.687535833046669, + "grad_norm": 0.6815490751435586, + "learning_rate": 0.00023492835267371575, + "loss": 2.2781, + "step": 1499 + }, + { + "epoch": 0.6879944960440316, + "grad_norm": 0.671287973891657, + "learning_rate": 0.0002342986120795978, + "loss": 2.3601, + "step": 1500 + }, + { + "epoch": 0.6884531590413944, + "grad_norm": 0.6726114533967881, + "learning_rate": 0.0002336694582765857, + "loss": 2.3199, + "step": 1501 + }, + { + "epoch": 0.688911822038757, + "grad_norm": 0.7042233388187636, + "learning_rate": 0.00023304089265414085, + "loss": 2.3807, + "step": 1502 + }, + { + "epoch": 0.6893704850361198, + "grad_norm": 0.6446053262707889, + "learning_rate": 0.00023241291660042613, + "loss": 2.3462, + "step": 1503 + }, + { + "epoch": 0.6898291480334824, + "grad_norm": 0.6766890024714356, + "learning_rate": 0.00023178553150230186, + "loss": 2.322, + "step": 1504 + }, + { + "epoch": 0.6902878110308451, + "grad_norm": 0.7145449261494504, + "learning_rate": 0.00023115873874532324, + "loss": 2.3083, + "step": 1505 + }, + { + "epoch": 0.6907464740282078, + "grad_norm": 0.6815298536690803, + "learning_rate": 0.00023053253971373796, + "loss": 2.3742, + "step": 1506 + }, + { + "epoch": 0.6912051370255705, + "grad_norm": 0.6694005278091946, + "learning_rate": 0.00022990693579048166, + "loss": 2.286, + "step": 1507 + }, + { + "epoch": 0.6916638000229332, + "grad_norm": 0.6822244586972716, + "learning_rate": 0.00022928192835717644, + "loss": 2.2942, + "step": 1508 + }, + { + "epoch": 0.6921224630202958, + "grad_norm": 0.719841156212866, + "learning_rate": 0.00022865751879412634, + "loss": 2.315, + "step": 1509 + }, + { + "epoch": 0.6925811260176585, + "grad_norm": 0.6894530126213981, + "learning_rate": 0.00022803370848031585, + "loss": 2.3757, + "step": 1510 + }, + { + "epoch": 0.6930397890150212, + "grad_norm": 0.7053703974447829, + "learning_rate": 0.00022741049879340542, + "loss": 2.3384, + "step": 1511 + }, + { + "epoch": 0.6934984520123839, + "grad_norm": 0.6503510196730354, + "learning_rate": 0.00022678789110972897, + "loss": 2.2393, + "step": 1512 + }, + { + "epoch": 0.6939571150097466, + "grad_norm": 0.7286828196446015, + "learning_rate": 0.00022616588680429155, + "loss": 2.2723, + "step": 1513 + }, + { + "epoch": 0.6944157780071093, + "grad_norm": 0.7127325915879452, + "learning_rate": 0.00022554448725076526, + "loss": 2.3346, + "step": 1514 + }, + { + "epoch": 0.6948744410044719, + "grad_norm": 0.7027300114872529, + "learning_rate": 0.0002249236938214863, + "loss": 2.2955, + "step": 1515 + }, + { + "epoch": 0.6953331040018347, + "grad_norm": 0.7091997293585337, + "learning_rate": 0.00022430350788745296, + "loss": 2.2954, + "step": 1516 + }, + { + "epoch": 0.6957917669991973, + "grad_norm": 0.6966442490088455, + "learning_rate": 0.00022368393081832166, + "loss": 2.3737, + "step": 1517 + }, + { + "epoch": 0.6962504299965601, + "grad_norm": 0.6620347781749569, + "learning_rate": 0.00022306496398240383, + "loss": 2.3392, + "step": 1518 + }, + { + "epoch": 0.6967090929939227, + "grad_norm": 0.654050505081007, + "learning_rate": 0.00022244660874666373, + "loss": 2.2616, + "step": 1519 + }, + { + "epoch": 0.6971677559912854, + "grad_norm": 0.705103265144589, + "learning_rate": 0.00022182886647671452, + "loss": 2.2869, + "step": 1520 + }, + { + "epoch": 0.6976264189886481, + "grad_norm": 0.6919992971130465, + "learning_rate": 0.0002212117385368157, + "loss": 2.2455, + "step": 1521 + }, + { + "epoch": 0.6980850819860108, + "grad_norm": 0.6631908405523237, + "learning_rate": 0.00022059522628987038, + "loss": 2.2603, + "step": 1522 + }, + { + "epoch": 0.6985437449833735, + "grad_norm": 0.7046589189656042, + "learning_rate": 0.00021997933109742162, + "loss": 2.3051, + "step": 1523 + }, + { + "epoch": 0.6990024079807362, + "grad_norm": 0.7025081203530055, + "learning_rate": 0.00021936405431964969, + "loss": 2.2822, + "step": 1524 + }, + { + "epoch": 0.6994610709780988, + "grad_norm": 0.7311092309498679, + "learning_rate": 0.00021874939731536926, + "loss": 2.2194, + "step": 1525 + }, + { + "epoch": 0.6999197339754615, + "grad_norm": 0.6529230706883197, + "learning_rate": 0.00021813536144202656, + "loss": 2.2333, + "step": 1526 + }, + { + "epoch": 0.7003783969728242, + "grad_norm": 0.6632411885482562, + "learning_rate": 0.00021752194805569553, + "loss": 2.2161, + "step": 1527 + }, + { + "epoch": 0.7008370599701869, + "grad_norm": 0.7481420259689362, + "learning_rate": 0.0002169091585110754, + "loss": 2.2504, + "step": 1528 + }, + { + "epoch": 0.7012957229675496, + "grad_norm": 0.6847020235590993, + "learning_rate": 0.00021629699416148828, + "loss": 2.2531, + "step": 1529 + }, + { + "epoch": 0.7017543859649122, + "grad_norm": 0.69487113257775, + "learning_rate": 0.000215685456358875, + "loss": 2.3705, + "step": 1530 + }, + { + "epoch": 0.702213048962275, + "grad_norm": 0.7587850180102813, + "learning_rate": 0.00021507454645379258, + "loss": 2.3663, + "step": 1531 + }, + { + "epoch": 0.7026717119596376, + "grad_norm": 0.7117651711948642, + "learning_rate": 0.00021446426579541184, + "loss": 2.3401, + "step": 1532 + }, + { + "epoch": 0.7031303749570004, + "grad_norm": 0.6882020705840624, + "learning_rate": 0.00021385461573151387, + "loss": 2.2426, + "step": 1533 + }, + { + "epoch": 0.703589037954363, + "grad_norm": 0.6974365368043225, + "learning_rate": 0.00021324559760848677, + "loss": 2.3243, + "step": 1534 + }, + { + "epoch": 0.7040477009517258, + "grad_norm": 0.6813330512407868, + "learning_rate": 0.00021263721277132303, + "loss": 2.2071, + "step": 1535 + }, + { + "epoch": 0.7045063639490884, + "grad_norm": 0.6698985881730712, + "learning_rate": 0.0002120294625636171, + "loss": 2.2804, + "step": 1536 + }, + { + "epoch": 0.7049650269464511, + "grad_norm": 0.7000638226048644, + "learning_rate": 0.0002114223483275613, + "loss": 2.3091, + "step": 1537 + }, + { + "epoch": 0.7054236899438138, + "grad_norm": 0.7212817068151397, + "learning_rate": 0.0002108158714039435, + "loss": 2.3451, + "step": 1538 + }, + { + "epoch": 0.7058823529411765, + "grad_norm": 0.7422250880144734, + "learning_rate": 0.00021021003313214455, + "loss": 2.4055, + "step": 1539 + }, + { + "epoch": 0.7063410159385392, + "grad_norm": 0.6962088264452451, + "learning_rate": 0.00020960483485013432, + "loss": 2.3439, + "step": 1540 + }, + { + "epoch": 0.7067996789359019, + "grad_norm": 0.7205477783048959, + "learning_rate": 0.0002090002778944694, + "loss": 2.3232, + "step": 1541 + }, + { + "epoch": 0.7072583419332645, + "grad_norm": 0.6917758157692065, + "learning_rate": 0.00020839636360029025, + "loss": 2.3062, + "step": 1542 + }, + { + "epoch": 0.7077170049306272, + "grad_norm": 0.7661013450610675, + "learning_rate": 0.00020779309330131818, + "loss": 2.272, + "step": 1543 + }, + { + "epoch": 0.7081756679279899, + "grad_norm": 0.7463041530130167, + "learning_rate": 0.00020719046832985184, + "loss": 2.3199, + "step": 1544 + }, + { + "epoch": 0.7086343309253526, + "grad_norm": 0.6632147784203298, + "learning_rate": 0.0002065884900167646, + "loss": 2.2552, + "step": 1545 + }, + { + "epoch": 0.7090929939227153, + "grad_norm": 0.6905132097876959, + "learning_rate": 0.0002059871596915024, + "loss": 2.3969, + "step": 1546 + }, + { + "epoch": 0.7095516569200779, + "grad_norm": 0.6871928500232692, + "learning_rate": 0.0002053864786820795, + "loss": 2.4169, + "step": 1547 + }, + { + "epoch": 0.7100103199174407, + "grad_norm": 0.7080872349884519, + "learning_rate": 0.00020478644831507627, + "loss": 2.1971, + "step": 1548 + }, + { + "epoch": 0.7104689829148033, + "grad_norm": 0.7045573740688712, + "learning_rate": 0.00020418706991563634, + "loss": 2.3789, + "step": 1549 + }, + { + "epoch": 0.7109276459121661, + "grad_norm": 0.7092605184599129, + "learning_rate": 0.00020358834480746363, + "loss": 2.3154, + "step": 1550 + }, + { + "epoch": 0.7113863089095287, + "grad_norm": 0.6968100785868249, + "learning_rate": 0.0002029902743128188, + "loss": 2.3552, + "step": 1551 + }, + { + "epoch": 0.7118449719068914, + "grad_norm": 0.727707937854455, + "learning_rate": 0.0002023928597525174, + "loss": 2.3351, + "step": 1552 + }, + { + "epoch": 0.7123036349042541, + "grad_norm": 0.6641293040985145, + "learning_rate": 0.00020179610244592595, + "loss": 2.2579, + "step": 1553 + }, + { + "epoch": 0.7127622979016168, + "grad_norm": 0.6844587296123071, + "learning_rate": 0.00020120000371095937, + "loss": 2.274, + "step": 1554 + }, + { + "epoch": 0.7132209608989795, + "grad_norm": 0.7175681713162823, + "learning_rate": 0.0002006045648640787, + "loss": 2.244, + "step": 1555 + }, + { + "epoch": 0.7136796238963422, + "grad_norm": 0.6790687311756209, + "learning_rate": 0.00020000978722028713, + "loss": 2.2675, + "step": 1556 + }, + { + "epoch": 0.7141382868937048, + "grad_norm": 0.7766236128686704, + "learning_rate": 0.00019941567209312767, + "loss": 2.4426, + "step": 1557 + }, + { + "epoch": 0.7145969498910676, + "grad_norm": 0.6900254274491463, + "learning_rate": 0.00019882222079468036, + "loss": 2.2987, + "step": 1558 + }, + { + "epoch": 0.7150556128884302, + "grad_norm": 0.7054734024989732, + "learning_rate": 0.0001982294346355595, + "loss": 2.2321, + "step": 1559 + }, + { + "epoch": 0.7155142758857929, + "grad_norm": 0.6749954642970553, + "learning_rate": 0.00019763731492490976, + "loss": 2.2974, + "step": 1560 + }, + { + "epoch": 0.7159729388831556, + "grad_norm": 0.7340270053468406, + "learning_rate": 0.00019704586297040422, + "loss": 2.2628, + "step": 1561 + }, + { + "epoch": 0.7164316018805182, + "grad_norm": 0.6967360375034656, + "learning_rate": 0.0001964550800782417, + "loss": 2.2936, + "step": 1562 + }, + { + "epoch": 0.716890264877881, + "grad_norm": 0.6802127009261599, + "learning_rate": 0.00019586496755314288, + "loss": 2.2618, + "step": 1563 + }, + { + "epoch": 0.7173489278752436, + "grad_norm": 0.7531447038785155, + "learning_rate": 0.00019527552669834798, + "loss": 2.2449, + "step": 1564 + }, + { + "epoch": 0.7178075908726064, + "grad_norm": 0.7056603678139538, + "learning_rate": 0.0001946867588156142, + "loss": 2.2869, + "step": 1565 + }, + { + "epoch": 0.718266253869969, + "grad_norm": 0.7611610267456042, + "learning_rate": 0.00019409866520521258, + "loss": 2.2541, + "step": 1566 + }, + { + "epoch": 0.7187249168673318, + "grad_norm": 0.7478926929655729, + "learning_rate": 0.00019351124716592455, + "loss": 2.3042, + "step": 1567 + }, + { + "epoch": 0.7191835798646944, + "grad_norm": 0.7238709400842842, + "learning_rate": 0.0001929245059950397, + "loss": 2.1931, + "step": 1568 + }, + { + "epoch": 0.7196422428620571, + "grad_norm": 0.7958852885573587, + "learning_rate": 0.0001923384429883533, + "loss": 2.3835, + "step": 1569 + }, + { + "epoch": 0.7201009058594198, + "grad_norm": 0.737043424069362, + "learning_rate": 0.00019175305944016237, + "loss": 2.2106, + "step": 1570 + }, + { + "epoch": 0.7205595688567825, + "grad_norm": 0.6721629211744985, + "learning_rate": 0.00019116835664326326, + "loss": 2.3482, + "step": 1571 + }, + { + "epoch": 0.7210182318541452, + "grad_norm": 0.7221715927910537, + "learning_rate": 0.0001905843358889497, + "loss": 2.284, + "step": 1572 + }, + { + "epoch": 0.7214768948515079, + "grad_norm": 0.7007041545020293, + "learning_rate": 0.00019000099846700836, + "loss": 2.2876, + "step": 1573 + }, + { + "epoch": 0.7219355578488705, + "grad_norm": 0.65965493192873, + "learning_rate": 0.00018941834566571692, + "loss": 2.292, + "step": 1574 + }, + { + "epoch": 0.7223942208462333, + "grad_norm": 0.6821195384591222, + "learning_rate": 0.00018883637877184145, + "loss": 2.1832, + "step": 1575 + }, + { + "epoch": 0.7228528838435959, + "grad_norm": 0.690151008492755, + "learning_rate": 0.00018825509907063325, + "loss": 2.3009, + "step": 1576 + }, + { + "epoch": 0.7233115468409586, + "grad_norm": 0.6706545463172845, + "learning_rate": 0.00018767450784582557, + "loss": 2.3057, + "step": 1577 + }, + { + "epoch": 0.7237702098383213, + "grad_norm": 0.6672433865997219, + "learning_rate": 0.00018709460637963122, + "loss": 2.2685, + "step": 1578 + }, + { + "epoch": 0.7242288728356839, + "grad_norm": 0.6626615483637744, + "learning_rate": 0.00018651539595274013, + "loss": 2.2565, + "step": 1579 + }, + { + "epoch": 0.7246875358330467, + "grad_norm": 0.7272684958889077, + "learning_rate": 0.00018593687784431578, + "loss": 2.2503, + "step": 1580 + }, + { + "epoch": 0.7251461988304093, + "grad_norm": 0.6929734924753138, + "learning_rate": 0.00018535905333199248, + "loss": 2.201, + "step": 1581 + }, + { + "epoch": 0.7256048618277721, + "grad_norm": 0.7183663959594941, + "learning_rate": 0.0001847819236918733, + "loss": 2.3042, + "step": 1582 + }, + { + "epoch": 0.7260635248251347, + "grad_norm": 0.766085861020442, + "learning_rate": 0.00018420549019852655, + "loss": 2.3988, + "step": 1583 + }, + { + "epoch": 0.7265221878224974, + "grad_norm": 0.7313396268937833, + "learning_rate": 0.00018362975412498266, + "loss": 2.3191, + "step": 1584 + }, + { + "epoch": 0.7269808508198601, + "grad_norm": 0.7039600700576623, + "learning_rate": 0.00018305471674273261, + "loss": 2.2608, + "step": 1585 + }, + { + "epoch": 0.7274395138172228, + "grad_norm": 0.7172289209012052, + "learning_rate": 0.0001824803793217237, + "loss": 2.2876, + "step": 1586 + }, + { + "epoch": 0.7278981768145855, + "grad_norm": 0.743150826301944, + "learning_rate": 0.00018190674313035737, + "loss": 2.2851, + "step": 1587 + }, + { + "epoch": 0.7283568398119482, + "grad_norm": 0.6837837969755917, + "learning_rate": 0.00018133380943548716, + "loss": 2.334, + "step": 1588 + }, + { + "epoch": 0.7288155028093108, + "grad_norm": 0.7046678502776893, + "learning_rate": 0.00018076157950241452, + "loss": 2.2526, + "step": 1589 + }, + { + "epoch": 0.7292741658066736, + "grad_norm": 0.6876340281484491, + "learning_rate": 0.00018019005459488652, + "loss": 2.3156, + "step": 1590 + }, + { + "epoch": 0.7297328288040362, + "grad_norm": 0.7532064273911471, + "learning_rate": 0.00017961923597509388, + "loss": 2.3964, + "step": 1591 + }, + { + "epoch": 0.730191491801399, + "grad_norm": 0.7509055796596787, + "learning_rate": 0.00017904912490366722, + "loss": 2.2562, + "step": 1592 + }, + { + "epoch": 0.7306501547987616, + "grad_norm": 0.6795168828350704, + "learning_rate": 0.00017847972263967433, + "loss": 2.3615, + "step": 1593 + }, + { + "epoch": 0.7311088177961242, + "grad_norm": 0.706119260210384, + "learning_rate": 0.0001779110304406177, + "loss": 2.3012, + "step": 1594 + }, + { + "epoch": 0.731567480793487, + "grad_norm": 0.7034583793945517, + "learning_rate": 0.000177343049562432, + "loss": 2.23, + "step": 1595 + }, + { + "epoch": 0.7320261437908496, + "grad_norm": 0.711319357455059, + "learning_rate": 0.0001767757812594807, + "loss": 2.3108, + "step": 1596 + }, + { + "epoch": 0.7324848067882124, + "grad_norm": 0.7097730386000733, + "learning_rate": 0.0001762092267845534, + "loss": 2.2535, + "step": 1597 + }, + { + "epoch": 0.732943469785575, + "grad_norm": 0.7085615960605287, + "learning_rate": 0.00017564338738886365, + "loss": 2.2851, + "step": 1598 + }, + { + "epoch": 0.7334021327829378, + "grad_norm": 0.7025609234093599, + "learning_rate": 0.0001750782643220457, + "loss": 2.3931, + "step": 1599 + }, + { + "epoch": 0.7338607957803004, + "grad_norm": 0.6854706185487751, + "learning_rate": 0.00017451385883215166, + "loss": 2.3699, + "step": 1600 + }, + { + "epoch": 0.7343194587776631, + "grad_norm": 0.6914555266475665, + "learning_rate": 0.00017395017216564863, + "loss": 2.162, + "step": 1601 + }, + { + "epoch": 0.7347781217750258, + "grad_norm": 0.6800912031617445, + "learning_rate": 0.00017338720556741687, + "loss": 2.2997, + "step": 1602 + }, + { + "epoch": 0.7352367847723885, + "grad_norm": 0.7226725706788094, + "learning_rate": 0.00017282496028074606, + "loss": 2.3449, + "step": 1603 + }, + { + "epoch": 0.7356954477697512, + "grad_norm": 0.6814344593459196, + "learning_rate": 0.00017226343754733254, + "loss": 2.2723, + "step": 1604 + }, + { + "epoch": 0.7361541107671139, + "grad_norm": 0.6509210132755576, + "learning_rate": 0.00017170263860727769, + "loss": 2.2757, + "step": 1605 + }, + { + "epoch": 0.7366127737644765, + "grad_norm": 0.6743696824839019, + "learning_rate": 0.0001711425646990838, + "loss": 2.2913, + "step": 1606 + }, + { + "epoch": 0.7370714367618393, + "grad_norm": 0.7104569621098739, + "learning_rate": 0.00017058321705965202, + "loss": 2.3275, + "step": 1607 + }, + { + "epoch": 0.7375300997592019, + "grad_norm": 0.6973588546663567, + "learning_rate": 0.0001700245969242798, + "loss": 2.2764, + "step": 1608 + }, + { + "epoch": 0.7379887627565647, + "grad_norm": 0.7050025898486147, + "learning_rate": 0.00016946670552665804, + "loss": 2.2479, + "step": 1609 + }, + { + "epoch": 0.7384474257539273, + "grad_norm": 0.7300958542079369, + "learning_rate": 0.00016890954409886795, + "loss": 2.2917, + "step": 1610 + }, + { + "epoch": 0.7389060887512899, + "grad_norm": 0.7095576513520432, + "learning_rate": 0.00016835311387137836, + "loss": 2.3109, + "step": 1611 + }, + { + "epoch": 0.7393647517486527, + "grad_norm": 0.7479474978239146, + "learning_rate": 0.0001677974160730441, + "loss": 2.2904, + "step": 1612 + }, + { + "epoch": 0.7398234147460153, + "grad_norm": 0.744102613373866, + "learning_rate": 0.00016724245193110176, + "loss": 2.3434, + "step": 1613 + }, + { + "epoch": 0.7402820777433781, + "grad_norm": 0.7538836332045407, + "learning_rate": 0.00016668822267116784, + "loss": 2.2913, + "step": 1614 + }, + { + "epoch": 0.7407407407407407, + "grad_norm": 0.6726315929601805, + "learning_rate": 0.00016613472951723597, + "loss": 2.2842, + "step": 1615 + }, + { + "epoch": 0.7411994037381034, + "grad_norm": 0.6953772559916598, + "learning_rate": 0.00016558197369167434, + "loss": 2.2767, + "step": 1616 + }, + { + "epoch": 0.7416580667354661, + "grad_norm": 0.7322457772059554, + "learning_rate": 0.00016502995641522216, + "loss": 2.3182, + "step": 1617 + }, + { + "epoch": 0.7421167297328288, + "grad_norm": 0.7440837520754332, + "learning_rate": 0.00016447867890698843, + "loss": 2.3498, + "step": 1618 + }, + { + "epoch": 0.7425753927301915, + "grad_norm": 0.7212822129303975, + "learning_rate": 0.00016392814238444753, + "loss": 2.3316, + "step": 1619 + }, + { + "epoch": 0.7430340557275542, + "grad_norm": 0.6571108626077831, + "learning_rate": 0.00016337834806343782, + "loss": 2.2265, + "step": 1620 + }, + { + "epoch": 0.7434927187249168, + "grad_norm": 0.7036737101063599, + "learning_rate": 0.0001628292971581588, + "loss": 2.2503, + "step": 1621 + }, + { + "epoch": 0.7439513817222796, + "grad_norm": 0.6992002088705661, + "learning_rate": 0.00016228099088116772, + "loss": 2.3845, + "step": 1622 + }, + { + "epoch": 0.7444100447196422, + "grad_norm": 0.6678892655008846, + "learning_rate": 0.00016173343044337734, + "loss": 2.3327, + "step": 1623 + }, + { + "epoch": 0.744868707717005, + "grad_norm": 0.7106438792997931, + "learning_rate": 0.00016118661705405356, + "loss": 2.185, + "step": 1624 + }, + { + "epoch": 0.7453273707143676, + "grad_norm": 0.6821226180389975, + "learning_rate": 0.00016064055192081255, + "loss": 2.3407, + "step": 1625 + }, + { + "epoch": 0.7457860337117304, + "grad_norm": 0.6691838041457273, + "learning_rate": 0.00016009523624961757, + "loss": 2.2865, + "step": 1626 + }, + { + "epoch": 0.746244696709093, + "grad_norm": 0.6616306490294303, + "learning_rate": 0.00015955067124477678, + "loss": 2.2004, + "step": 1627 + }, + { + "epoch": 0.7467033597064556, + "grad_norm": 0.6832757766174954, + "learning_rate": 0.000159006858108941, + "loss": 2.3296, + "step": 1628 + }, + { + "epoch": 0.7471620227038184, + "grad_norm": 0.6908748000632551, + "learning_rate": 0.00015846379804310002, + "loss": 2.2922, + "step": 1629 + }, + { + "epoch": 0.747620685701181, + "grad_norm": 0.6754698059653936, + "learning_rate": 0.00015792149224658054, + "loss": 2.2447, + "step": 1630 + }, + { + "epoch": 0.7480793486985438, + "grad_norm": 0.6655048726679108, + "learning_rate": 0.00015737994191704385, + "loss": 2.334, + "step": 1631 + }, + { + "epoch": 0.7485380116959064, + "grad_norm": 0.709099676217221, + "learning_rate": 0.0001568391482504829, + "loss": 2.2558, + "step": 1632 + }, + { + "epoch": 0.7489966746932691, + "grad_norm": 0.7890191735284179, + "learning_rate": 0.00015629911244121903, + "loss": 2.2568, + "step": 1633 + }, + { + "epoch": 0.7494553376906318, + "grad_norm": 0.7129857794879079, + "learning_rate": 0.0001557598356819, + "loss": 2.4234, + "step": 1634 + }, + { + "epoch": 0.7499140006879945, + "grad_norm": 0.774262025183474, + "learning_rate": 0.00015522131916349786, + "loss": 2.2741, + "step": 1635 + }, + { + "epoch": 0.7503726636853572, + "grad_norm": 0.7285823375256697, + "learning_rate": 0.00015468356407530493, + "loss": 2.3378, + "step": 1636 + }, + { + "epoch": 0.7508313266827199, + "grad_norm": 0.7433853707759984, + "learning_rate": 0.00015414657160493217, + "loss": 2.3809, + "step": 1637 + }, + { + "epoch": 0.7512899896800825, + "grad_norm": 0.6764802075680784, + "learning_rate": 0.00015361034293830673, + "loss": 2.3213, + "step": 1638 + }, + { + "epoch": 0.7517486526774453, + "grad_norm": 0.6900673624812047, + "learning_rate": 0.00015307487925966844, + "loss": 2.2655, + "step": 1639 + }, + { + "epoch": 0.7522073156748079, + "grad_norm": 0.6761095906304826, + "learning_rate": 0.00015254018175156776, + "loss": 2.3565, + "step": 1640 + }, + { + "epoch": 0.7526659786721707, + "grad_norm": 0.7584939935827673, + "learning_rate": 0.0001520062515948632, + "loss": 2.3321, + "step": 1641 + }, + { + "epoch": 0.7531246416695333, + "grad_norm": 0.7155139711642099, + "learning_rate": 0.0001514730899687189, + "loss": 2.1726, + "step": 1642 + }, + { + "epoch": 0.753583304666896, + "grad_norm": 0.6867072952617154, + "learning_rate": 0.00015094069805060122, + "loss": 2.29, + "step": 1643 + }, + { + "epoch": 0.7540419676642587, + "grad_norm": 0.662292841245428, + "learning_rate": 0.00015040907701627666, + "loss": 2.3044, + "step": 1644 + }, + { + "epoch": 0.7545006306616213, + "grad_norm": 0.7015579718642819, + "learning_rate": 0.00014987822803980976, + "loss": 2.22, + "step": 1645 + }, + { + "epoch": 0.7549592936589841, + "grad_norm": 0.6986324106290669, + "learning_rate": 0.00014934815229355965, + "loss": 2.3406, + "step": 1646 + }, + { + "epoch": 0.7554179566563467, + "grad_norm": 0.7051582060602358, + "learning_rate": 0.00014881885094817748, + "loss": 2.2532, + "step": 1647 + }, + { + "epoch": 0.7558766196537094, + "grad_norm": 0.6959241445918751, + "learning_rate": 0.00014829032517260488, + "loss": 2.3801, + "step": 1648 + }, + { + "epoch": 0.7563352826510721, + "grad_norm": 0.6617699494599807, + "learning_rate": 0.0001477625761340704, + "loss": 2.3363, + "step": 1649 + }, + { + "epoch": 0.7567939456484348, + "grad_norm": 0.6992539290479437, + "learning_rate": 0.0001472356049980868, + "loss": 2.3302, + "step": 1650 + }, + { + "epoch": 0.7572526086457975, + "grad_norm": 0.6759445304779764, + "learning_rate": 0.00014670941292844954, + "loss": 2.2709, + "step": 1651 + }, + { + "epoch": 0.7577112716431602, + "grad_norm": 0.670757503242246, + "learning_rate": 0.00014618400108723295, + "loss": 2.2023, + "step": 1652 + }, + { + "epoch": 0.7581699346405228, + "grad_norm": 0.6713765169536584, + "learning_rate": 0.00014565937063478862, + "loss": 2.1933, + "step": 1653 + }, + { + "epoch": 0.7586285976378856, + "grad_norm": 0.705042266628489, + "learning_rate": 0.00014513552272974207, + "loss": 2.2788, + "step": 1654 + }, + { + "epoch": 0.7590872606352482, + "grad_norm": 0.7250666471342984, + "learning_rate": 0.0001446124585289913, + "loss": 2.2632, + "step": 1655 + }, + { + "epoch": 0.759545923632611, + "grad_norm": 0.7137212005696806, + "learning_rate": 0.00014409017918770266, + "loss": 2.2571, + "step": 1656 + }, + { + "epoch": 0.7600045866299736, + "grad_norm": 0.6879968650505062, + "learning_rate": 0.00014356868585930994, + "loss": 2.3746, + "step": 1657 + }, + { + "epoch": 0.7604632496273364, + "grad_norm": 0.7085467020735041, + "learning_rate": 0.00014304797969551077, + "loss": 2.2152, + "step": 1658 + }, + { + "epoch": 0.760921912624699, + "grad_norm": 0.6672966415670788, + "learning_rate": 0.00014252806184626417, + "loss": 2.4349, + "step": 1659 + }, + { + "epoch": 0.7613805756220617, + "grad_norm": 0.7647769926032842, + "learning_rate": 0.00014200893345978817, + "loss": 2.3529, + "step": 1660 + }, + { + "epoch": 0.7618392386194244, + "grad_norm": 0.6762907247855076, + "learning_rate": 0.00014149059568255778, + "loss": 2.3202, + "step": 1661 + }, + { + "epoch": 0.762297901616787, + "grad_norm": 0.7176823314388493, + "learning_rate": 0.00014097304965930157, + "loss": 2.2881, + "step": 1662 + }, + { + "epoch": 0.7627565646141498, + "grad_norm": 0.6803767940555752, + "learning_rate": 0.00014045629653299953, + "loss": 2.3611, + "step": 1663 + }, + { + "epoch": 0.7632152276115124, + "grad_norm": 0.657323613197789, + "learning_rate": 0.00013994033744488076, + "loss": 2.2576, + "step": 1664 + }, + { + "epoch": 0.7636738906088751, + "grad_norm": 0.7188339565822586, + "learning_rate": 0.00013942517353442092, + "loss": 2.2338, + "step": 1665 + }, + { + "epoch": 0.7641325536062378, + "grad_norm": 0.6985492648041384, + "learning_rate": 0.0001389108059393391, + "loss": 2.2706, + "step": 1666 + }, + { + "epoch": 0.7645912166036005, + "grad_norm": 0.7115455778520479, + "learning_rate": 0.00013839723579559581, + "loss": 2.3301, + "step": 1667 + }, + { + "epoch": 0.7650498796009632, + "grad_norm": 0.69267130202298, + "learning_rate": 0.00013788446423739103, + "loss": 2.2684, + "step": 1668 + }, + { + "epoch": 0.7655085425983259, + "grad_norm": 0.6796045489381276, + "learning_rate": 0.00013737249239716042, + "loss": 2.2964, + "step": 1669 + }, + { + "epoch": 0.7659672055956885, + "grad_norm": 0.7056729071634361, + "learning_rate": 0.00013686132140557355, + "loss": 2.2639, + "step": 1670 + }, + { + "epoch": 0.7664258685930513, + "grad_norm": 0.7137588842893978, + "learning_rate": 0.00013635095239153188, + "loss": 2.3055, + "step": 1671 + }, + { + "epoch": 0.7668845315904139, + "grad_norm": 0.7299394047509932, + "learning_rate": 0.00013584138648216527, + "loss": 2.2688, + "step": 1672 + }, + { + "epoch": 0.7673431945877767, + "grad_norm": 0.7014259961448484, + "learning_rate": 0.0001353326248028298, + "loss": 2.3091, + "step": 1673 + }, + { + "epoch": 0.7678018575851393, + "grad_norm": 0.7070570419407934, + "learning_rate": 0.00013482466847710594, + "loss": 2.3247, + "step": 1674 + }, + { + "epoch": 0.768260520582502, + "grad_norm": 0.9329814387104707, + "learning_rate": 0.00013431751862679554, + "loss": 2.3128, + "step": 1675 + }, + { + "epoch": 0.7687191835798647, + "grad_norm": 0.7043667827143364, + "learning_rate": 0.00013381117637191887, + "loss": 2.211, + "step": 1676 + }, + { + "epoch": 0.7691778465772274, + "grad_norm": 0.7026356706330597, + "learning_rate": 0.00013330564283071293, + "loss": 2.2294, + "step": 1677 + }, + { + "epoch": 0.7696365095745901, + "grad_norm": 0.7050887642533927, + "learning_rate": 0.000132800919119629, + "loss": 2.234, + "step": 1678 + }, + { + "epoch": 0.7700951725719528, + "grad_norm": 0.7079354996581078, + "learning_rate": 0.00013229700635332948, + "loss": 2.2692, + "step": 1679 + }, + { + "epoch": 0.7705538355693154, + "grad_norm": 0.7513397212253825, + "learning_rate": 0.00013179390564468585, + "loss": 2.2773, + "step": 1680 + }, + { + "epoch": 0.7710124985666781, + "grad_norm": 0.7625588020611889, + "learning_rate": 0.00013129161810477641, + "loss": 2.3157, + "step": 1681 + }, + { + "epoch": 0.7714711615640408, + "grad_norm": 0.7177987099704437, + "learning_rate": 0.0001307901448428837, + "loss": 2.3263, + "step": 1682 + }, + { + "epoch": 0.7719298245614035, + "grad_norm": 0.6994336936141918, + "learning_rate": 0.0001302894869664916, + "loss": 2.3085, + "step": 1683 + }, + { + "epoch": 0.7723884875587662, + "grad_norm": 0.7109660159194658, + "learning_rate": 0.00012978964558128336, + "loss": 2.2387, + "step": 1684 + }, + { + "epoch": 0.7728471505561288, + "grad_norm": 0.7740156909369987, + "learning_rate": 0.00012929062179113925, + "loss": 2.2587, + "step": 1685 + }, + { + "epoch": 0.7733058135534916, + "grad_norm": 0.6684615433690411, + "learning_rate": 0.00012879241669813368, + "loss": 2.2383, + "step": 1686 + }, + { + "epoch": 0.7737644765508542, + "grad_norm": 0.682444249800194, + "learning_rate": 0.00012829503140253295, + "loss": 2.3841, + "step": 1687 + }, + { + "epoch": 0.774223139548217, + "grad_norm": 0.6896005732259246, + "learning_rate": 0.0001277984670027933, + "loss": 2.263, + "step": 1688 + }, + { + "epoch": 0.7746818025455796, + "grad_norm": 0.6705098220049528, + "learning_rate": 0.00012730272459555737, + "loss": 2.3105, + "step": 1689 + }, + { + "epoch": 0.7751404655429424, + "grad_norm": 0.6735986098719778, + "learning_rate": 0.00012680780527565312, + "loss": 2.2724, + "step": 1690 + }, + { + "epoch": 0.775599128540305, + "grad_norm": 0.7009155758728849, + "learning_rate": 0.0001263137101360905, + "loss": 2.239, + "step": 1691 + }, + { + "epoch": 0.7760577915376677, + "grad_norm": 0.6941165961689142, + "learning_rate": 0.00012582044026805922, + "loss": 2.262, + "step": 1692 + }, + { + "epoch": 0.7765164545350304, + "grad_norm": 0.7251884955701422, + "learning_rate": 0.00012532799676092627, + "loss": 2.2514, + "step": 1693 + }, + { + "epoch": 0.7769751175323931, + "grad_norm": 0.6947049687056887, + "learning_rate": 0.00012483638070223414, + "loss": 2.3308, + "step": 1694 + }, + { + "epoch": 0.7774337805297558, + "grad_norm": 0.7508303486066332, + "learning_rate": 0.00012434559317769752, + "loss": 2.2693, + "step": 1695 + }, + { + "epoch": 0.7778924435271185, + "grad_norm": 0.7271536923828813, + "learning_rate": 0.0001238556352712012, + "loss": 2.341, + "step": 1696 + }, + { + "epoch": 0.7783511065244811, + "grad_norm": 0.7054536705470276, + "learning_rate": 0.00012336650806479827, + "loss": 2.3134, + "step": 1697 + }, + { + "epoch": 0.7788097695218438, + "grad_norm": 0.7019655409675036, + "learning_rate": 0.00012287821263870708, + "loss": 2.3275, + "step": 1698 + }, + { + "epoch": 0.7792684325192065, + "grad_norm": 0.7100918656549229, + "learning_rate": 0.00012239075007130885, + "loss": 2.3536, + "step": 1699 + }, + { + "epoch": 0.7797270955165692, + "grad_norm": 0.6762116630481748, + "learning_rate": 0.00012190412143914536, + "loss": 2.2005, + "step": 1700 + }, + { + "epoch": 0.7801857585139319, + "grad_norm": 0.680482244866114, + "learning_rate": 0.0001214183278169172, + "loss": 2.1524, + "step": 1701 + }, + { + "epoch": 0.7806444215112945, + "grad_norm": 0.6749542185642956, + "learning_rate": 0.00012093337027748042, + "loss": 2.426, + "step": 1702 + }, + { + "epoch": 0.7811030845086573, + "grad_norm": 0.6843656614996422, + "learning_rate": 0.00012044924989184459, + "loss": 2.2735, + "step": 1703 + }, + { + "epoch": 0.7815617475060199, + "grad_norm": 0.6881082218576136, + "learning_rate": 0.0001199659677291709, + "loss": 2.2902, + "step": 1704 + }, + { + "epoch": 0.7820204105033827, + "grad_norm": 0.6850216339276893, + "learning_rate": 0.00011948352485676895, + "loss": 2.344, + "step": 1705 + }, + { + "epoch": 0.7824790735007453, + "grad_norm": 0.6939091284235243, + "learning_rate": 0.00011900192234009477, + "loss": 2.2874, + "step": 1706 + }, + { + "epoch": 0.782937736498108, + "grad_norm": 0.6562929230684251, + "learning_rate": 0.00011852116124274875, + "loss": 2.2861, + "step": 1707 + }, + { + "epoch": 0.7833963994954707, + "grad_norm": 0.6701751127955953, + "learning_rate": 0.00011804124262647314, + "loss": 2.2881, + "step": 1708 + }, + { + "epoch": 0.7838550624928334, + "grad_norm": 0.7275222218512101, + "learning_rate": 0.00011756216755114929, + "loss": 2.2344, + "step": 1709 + }, + { + "epoch": 0.7843137254901961, + "grad_norm": 0.7037610685639114, + "learning_rate": 0.00011708393707479548, + "loss": 2.38, + "step": 1710 + }, + { + "epoch": 0.7847723884875588, + "grad_norm": 0.6991342216086996, + "learning_rate": 0.00011660655225356531, + "loss": 2.2529, + "step": 1711 + }, + { + "epoch": 0.7852310514849214, + "grad_norm": 0.6559475562492989, + "learning_rate": 0.0001161300141417444, + "loss": 2.2658, + "step": 1712 + }, + { + "epoch": 0.7856897144822842, + "grad_norm": 0.708375251126766, + "learning_rate": 0.00011565432379174823, + "loss": 2.3255, + "step": 1713 + }, + { + "epoch": 0.7861483774796468, + "grad_norm": 0.672704796150377, + "learning_rate": 0.00011517948225412056, + "loss": 2.2903, + "step": 1714 + }, + { + "epoch": 0.7866070404770095, + "grad_norm": 0.71384435368978, + "learning_rate": 0.00011470549057753032, + "loss": 2.3369, + "step": 1715 + }, + { + "epoch": 0.7870657034743722, + "grad_norm": 0.7123765133029795, + "learning_rate": 0.00011423234980876957, + "loss": 2.329, + "step": 1716 + }, + { + "epoch": 0.7875243664717348, + "grad_norm": 0.6737532450475071, + "learning_rate": 0.00011376006099275099, + "loss": 2.2816, + "step": 1717 + }, + { + "epoch": 0.7879830294690976, + "grad_norm": 0.7147542553152161, + "learning_rate": 0.00011328862517250609, + "loss": 2.3147, + "step": 1718 + }, + { + "epoch": 0.7884416924664602, + "grad_norm": 0.7168190610577074, + "learning_rate": 0.00011281804338918239, + "loss": 2.3077, + "step": 1719 + }, + { + "epoch": 0.788900355463823, + "grad_norm": 0.7445557221167395, + "learning_rate": 0.00011234831668204115, + "loss": 2.2535, + "step": 1720 + }, + { + "epoch": 0.7893590184611856, + "grad_norm": 0.6892286872684518, + "learning_rate": 0.00011187944608845569, + "loss": 2.2987, + "step": 1721 + }, + { + "epoch": 0.7898176814585484, + "grad_norm": 0.6764061687188548, + "learning_rate": 0.00011141143264390801, + "loss": 2.321, + "step": 1722 + }, + { + "epoch": 0.790276344455911, + "grad_norm": 0.7002086767589462, + "learning_rate": 0.0001109442773819877, + "loss": 2.2818, + "step": 1723 + }, + { + "epoch": 0.7907350074532737, + "grad_norm": 0.6953661933645283, + "learning_rate": 0.0001104779813343889, + "loss": 2.248, + "step": 1724 + }, + { + "epoch": 0.7911936704506364, + "grad_norm": 0.700799837348972, + "learning_rate": 0.00011001254553090812, + "loss": 2.3129, + "step": 1725 + }, + { + "epoch": 0.7916523334479991, + "grad_norm": 0.6887276203505827, + "learning_rate": 0.00010954797099944186, + "loss": 2.2933, + "step": 1726 + }, + { + "epoch": 0.7921109964453618, + "grad_norm": 0.7247717514928976, + "learning_rate": 0.0001090842587659851, + "loss": 2.2189, + "step": 1727 + }, + { + "epoch": 0.7925696594427245, + "grad_norm": 0.6970770344760642, + "learning_rate": 0.00010862140985462804, + "loss": 2.3265, + "step": 1728 + }, + { + "epoch": 0.7930283224400871, + "grad_norm": 0.6847614672672256, + "learning_rate": 0.00010815942528755418, + "loss": 2.2547, + "step": 1729 + }, + { + "epoch": 0.7934869854374499, + "grad_norm": 0.704366645148874, + "learning_rate": 0.00010769830608503844, + "loss": 2.3291, + "step": 1730 + }, + { + "epoch": 0.7939456484348125, + "grad_norm": 0.7633404145378921, + "learning_rate": 0.00010723805326544473, + "loss": 2.3268, + "step": 1731 + }, + { + "epoch": 0.7944043114321752, + "grad_norm": 0.742110281659889, + "learning_rate": 0.00010677866784522316, + "loss": 2.3434, + "step": 1732 + }, + { + "epoch": 0.7948629744295379, + "grad_norm": 0.7253043759244873, + "learning_rate": 0.00010632015083890839, + "loss": 2.1919, + "step": 1733 + }, + { + "epoch": 0.7953216374269005, + "grad_norm": 0.7257767590005169, + "learning_rate": 0.00010586250325911745, + "loss": 2.3835, + "step": 1734 + }, + { + "epoch": 0.7957803004242633, + "grad_norm": 0.6670408054713816, + "learning_rate": 0.00010540572611654697, + "loss": 2.3348, + "step": 1735 + }, + { + "epoch": 0.7962389634216259, + "grad_norm": 0.7128550965594296, + "learning_rate": 0.00010494982041997126, + "loss": 2.2737, + "step": 1736 + }, + { + "epoch": 0.7966976264189887, + "grad_norm": 0.7617473486735481, + "learning_rate": 0.0001044947871762405, + "loss": 2.3314, + "step": 1737 + }, + { + "epoch": 0.7971562894163513, + "grad_norm": 0.7115426336132749, + "learning_rate": 0.00010404062739027753, + "loss": 2.309, + "step": 1738 + }, + { + "epoch": 0.797614952413714, + "grad_norm": 0.7087298760817305, + "learning_rate": 0.00010358734206507641, + "loss": 2.392, + "step": 1739 + }, + { + "epoch": 0.7980736154110767, + "grad_norm": 0.7026755960039178, + "learning_rate": 0.00010313493220170017, + "loss": 2.2342, + "step": 1740 + }, + { + "epoch": 0.7985322784084394, + "grad_norm": 0.7149433467361734, + "learning_rate": 0.00010268339879927836, + "loss": 2.227, + "step": 1741 + }, + { + "epoch": 0.7989909414058021, + "grad_norm": 0.7151754464293006, + "learning_rate": 0.00010223274285500466, + "loss": 2.2761, + "step": 1742 + }, + { + "epoch": 0.7994496044031648, + "grad_norm": 0.7175585806179591, + "learning_rate": 0.00010178296536413495, + "loss": 2.3286, + "step": 1743 + }, + { + "epoch": 0.7999082674005275, + "grad_norm": 0.6894743116669287, + "learning_rate": 0.00010133406731998546, + "loss": 2.2628, + "step": 1744 + }, + { + "epoch": 0.8003669303978902, + "grad_norm": 0.7146763622404778, + "learning_rate": 0.00010088604971392979, + "loss": 2.289, + "step": 1745 + }, + { + "epoch": 0.8008255933952528, + "grad_norm": 0.7355790764159569, + "learning_rate": 0.0001004389135353972, + "loss": 2.3364, + "step": 1746 + }, + { + "epoch": 0.8012842563926156, + "grad_norm": 0.6881710353838401, + "learning_rate": 9.999265977187049e-05, + "loss": 2.3176, + "step": 1747 + }, + { + "epoch": 0.8017429193899782, + "grad_norm": 0.7076945375373263, + "learning_rate": 9.95472894088838e-05, + "loss": 2.322, + "step": 1748 + }, + { + "epoch": 0.8022015823873409, + "grad_norm": 0.6737475350961428, + "learning_rate": 9.910280343001993e-05, + "loss": 2.2337, + "step": 1749 + }, + { + "epoch": 0.8026602453847036, + "grad_norm": 0.6562635376808982, + "learning_rate": 9.865920281690866e-05, + "loss": 2.309, + "step": 1750 + }, + { + "epoch": 0.8031189083820662, + "grad_norm": 0.6971992338292768, + "learning_rate": 9.821648854922482e-05, + "loss": 2.2307, + "step": 1751 + }, + { + "epoch": 0.803577571379429, + "grad_norm": 0.6934182897146998, + "learning_rate": 9.77746616046854e-05, + "loss": 2.2639, + "step": 1752 + }, + { + "epoch": 0.8040362343767916, + "grad_norm": 0.7028627152535848, + "learning_rate": 9.733372295904774e-05, + "loss": 2.2954, + "step": 1753 + }, + { + "epoch": 0.8044948973741544, + "grad_norm": 0.7368678765934761, + "learning_rate": 9.68936735861079e-05, + "loss": 2.3241, + "step": 1754 + }, + { + "epoch": 0.804953560371517, + "grad_norm": 0.7031576693423969, + "learning_rate": 9.645451445769737e-05, + "loss": 2.4166, + "step": 1755 + }, + { + "epoch": 0.8054122233688797, + "grad_norm": 0.7225004487995771, + "learning_rate": 9.601624654368196e-05, + "loss": 2.2549, + "step": 1756 + }, + { + "epoch": 0.8058708863662424, + "grad_norm": 0.697342359096865, + "learning_rate": 9.557887081195938e-05, + "loss": 2.2643, + "step": 1757 + }, + { + "epoch": 0.8063295493636051, + "grad_norm": 0.7140808454447979, + "learning_rate": 9.514238822845667e-05, + "loss": 2.3953, + "step": 1758 + }, + { + "epoch": 0.8067882123609678, + "grad_norm": 0.7116000121673751, + "learning_rate": 9.470679975712837e-05, + "loss": 2.1873, + "step": 1759 + }, + { + "epoch": 0.8072468753583305, + "grad_norm": 0.7070598934446812, + "learning_rate": 9.427210635995481e-05, + "loss": 2.284, + "step": 1760 + }, + { + "epoch": 0.8077055383556931, + "grad_norm": 0.6950741099829922, + "learning_rate": 9.383830899693923e-05, + "loss": 2.2944, + "step": 1761 + }, + { + "epoch": 0.8081642013530559, + "grad_norm": 0.705274217755901, + "learning_rate": 9.340540862610591e-05, + "loss": 2.2316, + "step": 1762 + }, + { + "epoch": 0.8086228643504185, + "grad_norm": 0.7132636086878784, + "learning_rate": 9.297340620349854e-05, + "loss": 2.3325, + "step": 1763 + }, + { + "epoch": 0.8090815273477813, + "grad_norm": 0.730524813030107, + "learning_rate": 9.25423026831777e-05, + "loss": 2.2638, + "step": 1764 + }, + { + "epoch": 0.8095401903451439, + "grad_norm": 0.7131941505547937, + "learning_rate": 9.211209901721846e-05, + "loss": 2.3067, + "step": 1765 + }, + { + "epoch": 0.8099988533425065, + "grad_norm": 0.778193967145314, + "learning_rate": 9.168279615570863e-05, + "loss": 2.1927, + "step": 1766 + }, + { + "epoch": 0.8104575163398693, + "grad_norm": 0.7303639397067796, + "learning_rate": 9.125439504674699e-05, + "loss": 2.2916, + "step": 1767 + }, + { + "epoch": 0.8109161793372319, + "grad_norm": 0.6775512203359243, + "learning_rate": 9.082689663644057e-05, + "loss": 2.209, + "step": 1768 + }, + { + "epoch": 0.8113748423345947, + "grad_norm": 0.6855056750926561, + "learning_rate": 9.040030186890264e-05, + "loss": 2.2134, + "step": 1769 + }, + { + "epoch": 0.8118335053319573, + "grad_norm": 0.7008679099497767, + "learning_rate": 8.997461168625138e-05, + "loss": 2.3052, + "step": 1770 + }, + { + "epoch": 0.81229216832932, + "grad_norm": 0.6873240151424441, + "learning_rate": 8.954982702860664e-05, + "loss": 2.3262, + "step": 1771 + }, + { + "epoch": 0.8127508313266827, + "grad_norm": 0.6863183807097538, + "learning_rate": 8.912594883408865e-05, + "loss": 2.2852, + "step": 1772 + }, + { + "epoch": 0.8132094943240454, + "grad_norm": 0.7152925092741234, + "learning_rate": 8.870297803881589e-05, + "loss": 2.3116, + "step": 1773 + }, + { + "epoch": 0.8136681573214081, + "grad_norm": 0.6770281789495027, + "learning_rate": 8.828091557690287e-05, + "loss": 2.2817, + "step": 1774 + }, + { + "epoch": 0.8141268203187708, + "grad_norm": 0.7165381522890387, + "learning_rate": 8.785976238045801e-05, + "loss": 2.2649, + "step": 1775 + }, + { + "epoch": 0.8145854833161335, + "grad_norm": 0.6987501614875652, + "learning_rate": 8.743951937958144e-05, + "loss": 2.197, + "step": 1776 + }, + { + "epoch": 0.8150441463134962, + "grad_norm": 0.7403520317451955, + "learning_rate": 8.702018750236357e-05, + "loss": 2.2799, + "step": 1777 + }, + { + "epoch": 0.8155028093108588, + "grad_norm": 0.7180341513273862, + "learning_rate": 8.660176767488237e-05, + "loss": 2.31, + "step": 1778 + }, + { + "epoch": 0.8159614723082216, + "grad_norm": 0.7262987901214387, + "learning_rate": 8.618426082120146e-05, + "loss": 2.3189, + "step": 1779 + }, + { + "epoch": 0.8164201353055842, + "grad_norm": 0.6888018526419398, + "learning_rate": 8.576766786336854e-05, + "loss": 2.2086, + "step": 1780 + }, + { + "epoch": 0.816878798302947, + "grad_norm": 0.7086187035345188, + "learning_rate": 8.535198972141294e-05, + "loss": 2.2852, + "step": 1781 + }, + { + "epoch": 0.8173374613003096, + "grad_norm": 0.7044308593679924, + "learning_rate": 8.493722731334347e-05, + "loss": 2.3289, + "step": 1782 + }, + { + "epoch": 0.8177961242976722, + "grad_norm": 0.699047759147558, + "learning_rate": 8.452338155514644e-05, + "loss": 2.3688, + "step": 1783 + }, + { + "epoch": 0.818254787295035, + "grad_norm": 0.7698376903194974, + "learning_rate": 8.411045336078426e-05, + "loss": 2.3491, + "step": 1784 + }, + { + "epoch": 0.8187134502923976, + "grad_norm": 0.7123606820666789, + "learning_rate": 8.369844364219264e-05, + "loss": 2.3415, + "step": 1785 + }, + { + "epoch": 0.8191721132897604, + "grad_norm": 0.7009087897673546, + "learning_rate": 8.328735330927873e-05, + "loss": 2.2655, + "step": 1786 + }, + { + "epoch": 0.819630776287123, + "grad_norm": 0.6737555085591047, + "learning_rate": 8.287718326991961e-05, + "loss": 2.2582, + "step": 1787 + }, + { + "epoch": 0.8200894392844857, + "grad_norm": 0.7047652502104095, + "learning_rate": 8.246793442995954e-05, + "loss": 2.3836, + "step": 1788 + }, + { + "epoch": 0.8205481022818484, + "grad_norm": 0.6931146218728257, + "learning_rate": 8.205960769320875e-05, + "loss": 2.2992, + "step": 1789 + }, + { + "epoch": 0.8210067652792111, + "grad_norm": 0.7871383796633685, + "learning_rate": 8.165220396144085e-05, + "loss": 2.2649, + "step": 1790 + }, + { + "epoch": 0.8214654282765738, + "grad_norm": 0.6833046451626584, + "learning_rate": 8.12457241343909e-05, + "loss": 2.3465, + "step": 1791 + }, + { + "epoch": 0.8219240912739365, + "grad_norm": 0.721294844816923, + "learning_rate": 8.084016910975367e-05, + "loss": 2.2653, + "step": 1792 + }, + { + "epoch": 0.8223827542712991, + "grad_norm": 0.688391194290499, + "learning_rate": 8.043553978318169e-05, + "loss": 2.2571, + "step": 1793 + }, + { + "epoch": 0.8228414172686619, + "grad_norm": 0.7064658151441972, + "learning_rate": 8.003183704828281e-05, + "loss": 2.2719, + "step": 1794 + }, + { + "epoch": 0.8233000802660245, + "grad_norm": 0.7162672945498755, + "learning_rate": 7.962906179661872e-05, + "loss": 2.3247, + "step": 1795 + }, + { + "epoch": 0.8237587432633873, + "grad_norm": 0.6997357628847514, + "learning_rate": 7.922721491770296e-05, + "loss": 2.2801, + "step": 1796 + }, + { + "epoch": 0.8242174062607499, + "grad_norm": 0.6856134463016502, + "learning_rate": 7.882629729899832e-05, + "loss": 2.3027, + "step": 1797 + }, + { + "epoch": 0.8246760692581127, + "grad_norm": 0.7224234065830683, + "learning_rate": 7.842630982591598e-05, + "loss": 2.1512, + "step": 1798 + }, + { + "epoch": 0.8251347322554753, + "grad_norm": 0.6621779559316382, + "learning_rate": 7.802725338181232e-05, + "loss": 2.2614, + "step": 1799 + }, + { + "epoch": 0.8255933952528379, + "grad_norm": 0.6951601697683927, + "learning_rate": 7.762912884798812e-05, + "loss": 2.2883, + "step": 1800 + }, + { + "epoch": 0.8260520582502007, + "grad_norm": 0.699202675514589, + "learning_rate": 7.723193710368564e-05, + "loss": 2.2919, + "step": 1801 + }, + { + "epoch": 0.8265107212475633, + "grad_norm": 0.7033125168854022, + "learning_rate": 7.683567902608729e-05, + "loss": 2.2222, + "step": 1802 + }, + { + "epoch": 0.826969384244926, + "grad_norm": 0.6730452663073688, + "learning_rate": 7.644035549031364e-05, + "loss": 2.3391, + "step": 1803 + }, + { + "epoch": 0.8274280472422887, + "grad_norm": 0.6956852362765976, + "learning_rate": 7.604596736942115e-05, + "loss": 2.2187, + "step": 1804 + }, + { + "epoch": 0.8278867102396514, + "grad_norm": 0.7181722590324493, + "learning_rate": 7.56525155344004e-05, + "loss": 2.2608, + "step": 1805 + }, + { + "epoch": 0.8283453732370141, + "grad_norm": 0.7295436054328155, + "learning_rate": 7.52600008541745e-05, + "loss": 2.2518, + "step": 1806 + }, + { + "epoch": 0.8288040362343768, + "grad_norm": 0.696853438876534, + "learning_rate": 7.486842419559681e-05, + "loss": 2.3377, + "step": 1807 + }, + { + "epoch": 0.8292626992317395, + "grad_norm": 0.7157388786070757, + "learning_rate": 7.447778642344898e-05, + "loss": 2.2799, + "step": 1808 + }, + { + "epoch": 0.8297213622291022, + "grad_norm": 0.7187783374907528, + "learning_rate": 7.408808840043912e-05, + "loss": 2.2928, + "step": 1809 + }, + { + "epoch": 0.8301800252264648, + "grad_norm": 0.6729090461087182, + "learning_rate": 7.369933098720021e-05, + "loss": 2.2479, + "step": 1810 + }, + { + "epoch": 0.8306386882238276, + "grad_norm": 0.6841930203118186, + "learning_rate": 7.331151504228767e-05, + "loss": 2.2325, + "step": 1811 + }, + { + "epoch": 0.8310973512211902, + "grad_norm": 0.6559352305142369, + "learning_rate": 7.292464142217775e-05, + "loss": 2.2838, + "step": 1812 + }, + { + "epoch": 0.831556014218553, + "grad_norm": 0.6771257495695191, + "learning_rate": 7.25387109812658e-05, + "loss": 2.229, + "step": 1813 + }, + { + "epoch": 0.8320146772159156, + "grad_norm": 0.7225958996366174, + "learning_rate": 7.215372457186415e-05, + "loss": 2.2728, + "step": 1814 + }, + { + "epoch": 0.8324733402132783, + "grad_norm": 0.6942902464698832, + "learning_rate": 7.176968304420007e-05, + "loss": 2.2539, + "step": 1815 + }, + { + "epoch": 0.832932003210641, + "grad_norm": 0.6957520439309287, + "learning_rate": 7.138658724641417e-05, + "loss": 2.2638, + "step": 1816 + }, + { + "epoch": 0.8333906662080036, + "grad_norm": 0.6642655520903189, + "learning_rate": 7.10044380245587e-05, + "loss": 2.2448, + "step": 1817 + }, + { + "epoch": 0.8338493292053664, + "grad_norm": 0.7112930541404876, + "learning_rate": 7.062323622259515e-05, + "loss": 2.1965, + "step": 1818 + }, + { + "epoch": 0.834307992202729, + "grad_norm": 0.6901244941709275, + "learning_rate": 7.024298268239265e-05, + "loss": 2.1811, + "step": 1819 + }, + { + "epoch": 0.8347666552000917, + "grad_norm": 0.6983680924507724, + "learning_rate": 6.986367824372647e-05, + "loss": 2.3191, + "step": 1820 + }, + { + "epoch": 0.8352253181974544, + "grad_norm": 0.6540553166095312, + "learning_rate": 6.948532374427541e-05, + "loss": 2.2657, + "step": 1821 + }, + { + "epoch": 0.8356839811948171, + "grad_norm": 0.6936244156809047, + "learning_rate": 6.910792001962063e-05, + "loss": 2.3348, + "step": 1822 + }, + { + "epoch": 0.8361426441921798, + "grad_norm": 0.7367898564608892, + "learning_rate": 6.873146790324358e-05, + "loss": 2.3008, + "step": 1823 + }, + { + "epoch": 0.8366013071895425, + "grad_norm": 0.7364339518064877, + "learning_rate": 6.83559682265239e-05, + "loss": 2.2968, + "step": 1824 + }, + { + "epoch": 0.8370599701869051, + "grad_norm": 0.7822373843208832, + "learning_rate": 6.798142181873784e-05, + "loss": 2.2397, + "step": 1825 + }, + { + "epoch": 0.8375186331842679, + "grad_norm": 0.7628678194866907, + "learning_rate": 6.760782950705662e-05, + "loss": 2.3679, + "step": 1826 + }, + { + "epoch": 0.8379772961816305, + "grad_norm": 0.6783244366704323, + "learning_rate": 6.723519211654422e-05, + "loss": 2.2715, + "step": 1827 + }, + { + "epoch": 0.8384359591789933, + "grad_norm": 0.6972265594816555, + "learning_rate": 6.686351047015554e-05, + "loss": 2.308, + "step": 1828 + }, + { + "epoch": 0.8388946221763559, + "grad_norm": 0.6925229536429637, + "learning_rate": 6.649278538873515e-05, + "loss": 2.1969, + "step": 1829 + }, + { + "epoch": 0.8393532851737187, + "grad_norm": 0.6681646629778479, + "learning_rate": 6.612301769101465e-05, + "loss": 2.2914, + "step": 1830 + }, + { + "epoch": 0.8398119481710813, + "grad_norm": 0.7147114393033783, + "learning_rate": 6.575420819361177e-05, + "loss": 2.3936, + "step": 1831 + }, + { + "epoch": 0.840270611168444, + "grad_norm": 0.7451893975986733, + "learning_rate": 6.538635771102757e-05, + "loss": 2.2426, + "step": 1832 + }, + { + "epoch": 0.8407292741658067, + "grad_norm": 0.7276222861917824, + "learning_rate": 6.501946705564566e-05, + "loss": 2.3841, + "step": 1833 + }, + { + "epoch": 0.8411879371631693, + "grad_norm": 0.6722882311120644, + "learning_rate": 6.465353703772959e-05, + "loss": 2.1801, + "step": 1834 + }, + { + "epoch": 0.841646600160532, + "grad_norm": 0.6971786248410711, + "learning_rate": 6.428856846542136e-05, + "loss": 2.2693, + "step": 1835 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 0.6848052936366067, + "learning_rate": 6.392456214473996e-05, + "loss": 2.3169, + "step": 1836 + }, + { + "epoch": 0.8425639261552574, + "grad_norm": 0.7010946989824027, + "learning_rate": 6.3561518879579e-05, + "loss": 2.223, + "step": 1837 + }, + { + "epoch": 0.8430225891526201, + "grad_norm": 0.729712554269084, + "learning_rate": 6.31994394717052e-05, + "loss": 2.182, + "step": 1838 + }, + { + "epoch": 0.8434812521499828, + "grad_norm": 0.9156522475586636, + "learning_rate": 6.283832472075685e-05, + "loss": 2.2692, + "step": 1839 + }, + { + "epoch": 0.8439399151473455, + "grad_norm": 0.6847788055584897, + "learning_rate": 6.247817542424178e-05, + "loss": 2.3047, + "step": 1840 + }, + { + "epoch": 0.8443985781447082, + "grad_norm": 0.7375098772376666, + "learning_rate": 6.211899237753559e-05, + "loss": 2.2752, + "step": 1841 + }, + { + "epoch": 0.8448572411420708, + "grad_norm": 0.8247600717014337, + "learning_rate": 6.176077637387984e-05, + "loss": 2.2729, + "step": 1842 + }, + { + "epoch": 0.8453159041394336, + "grad_norm": 0.704663788240708, + "learning_rate": 6.140352820438066e-05, + "loss": 2.2312, + "step": 1843 + }, + { + "epoch": 0.8457745671367962, + "grad_norm": 0.7138592406470984, + "learning_rate": 6.104724865800665e-05, + "loss": 2.2478, + "step": 1844 + }, + { + "epoch": 0.846233230134159, + "grad_norm": 0.687084712497627, + "learning_rate": 6.069193852158711e-05, + "loss": 2.2896, + "step": 1845 + }, + { + "epoch": 0.8466918931315216, + "grad_norm": 0.6976634349109444, + "learning_rate": 6.0337598579810584e-05, + "loss": 2.2437, + "step": 1846 + }, + { + "epoch": 0.8471505561288843, + "grad_norm": 0.7373236445872707, + "learning_rate": 5.9984229615223096e-05, + "loss": 2.2166, + "step": 1847 + }, + { + "epoch": 0.847609219126247, + "grad_norm": 0.6907595851405884, + "learning_rate": 5.963183240822606e-05, + "loss": 2.2883, + "step": 1848 + }, + { + "epoch": 0.8480678821236097, + "grad_norm": 0.729730108126497, + "learning_rate": 5.9280407737074825e-05, + "loss": 2.2445, + "step": 1849 + }, + { + "epoch": 0.8485265451209724, + "grad_norm": 0.7725039169719077, + "learning_rate": 5.8929956377877125e-05, + "loss": 2.3256, + "step": 1850 + }, + { + "epoch": 0.848985208118335, + "grad_norm": 0.7544778999897808, + "learning_rate": 5.8580479104591075e-05, + "loss": 2.2256, + "step": 1851 + }, + { + "epoch": 0.8494438711156977, + "grad_norm": 0.7241688599081765, + "learning_rate": 5.823197668902341e-05, + "loss": 2.29, + "step": 1852 + }, + { + "epoch": 0.8499025341130604, + "grad_norm": 0.7361934819040359, + "learning_rate": 5.78844499008282e-05, + "loss": 2.2096, + "step": 1853 + }, + { + "epoch": 0.8503611971104231, + "grad_norm": 0.6772293060587765, + "learning_rate": 5.753789950750454e-05, + "loss": 2.2694, + "step": 1854 + }, + { + "epoch": 0.8508198601077858, + "grad_norm": 0.7588995665455117, + "learning_rate": 5.719232627439558e-05, + "loss": 2.2806, + "step": 1855 + }, + { + "epoch": 0.8512785231051485, + "grad_norm": 0.8206257588313984, + "learning_rate": 5.6847730964686315e-05, + "loss": 2.1576, + "step": 1856 + }, + { + "epoch": 0.8517371861025111, + "grad_norm": 0.7163415685914516, + "learning_rate": 5.650411433940189e-05, + "loss": 2.3331, + "step": 1857 + }, + { + "epoch": 0.8521958490998739, + "grad_norm": 0.7800340227756511, + "learning_rate": 5.61614771574061e-05, + "loss": 2.2253, + "step": 1858 + }, + { + "epoch": 0.8526545120972365, + "grad_norm": 0.6800787830638754, + "learning_rate": 5.581982017539988e-05, + "loss": 2.1883, + "step": 1859 + }, + { + "epoch": 0.8531131750945993, + "grad_norm": 0.7962197205542233, + "learning_rate": 5.5479144147919216e-05, + "loss": 2.2452, + "step": 1860 + }, + { + "epoch": 0.8535718380919619, + "grad_norm": 0.7032178455826968, + "learning_rate": 5.51394498273336e-05, + "loss": 2.2099, + "step": 1861 + }, + { + "epoch": 0.8540305010893247, + "grad_norm": 0.7072761133155955, + "learning_rate": 5.480073796384494e-05, + "loss": 2.3844, + "step": 1862 + }, + { + "epoch": 0.8544891640866873, + "grad_norm": 0.7042705809600021, + "learning_rate": 5.446300930548492e-05, + "loss": 2.3572, + "step": 1863 + }, + { + "epoch": 0.85494782708405, + "grad_norm": 0.6861450882064589, + "learning_rate": 5.412626459811415e-05, + "loss": 2.2199, + "step": 1864 + }, + { + "epoch": 0.8554064900814127, + "grad_norm": 0.735900338830298, + "learning_rate": 5.3790504585419954e-05, + "loss": 2.2517, + "step": 1865 + }, + { + "epoch": 0.8558651530787754, + "grad_norm": 0.7105935012076691, + "learning_rate": 5.345573000891541e-05, + "loss": 2.1409, + "step": 1866 + }, + { + "epoch": 0.8563238160761381, + "grad_norm": 0.7051983541705241, + "learning_rate": 5.312194160793693e-05, + "loss": 2.2544, + "step": 1867 + }, + { + "epoch": 0.8567824790735007, + "grad_norm": 0.7061372588472463, + "learning_rate": 5.278914011964303e-05, + "loss": 2.3111, + "step": 1868 + }, + { + "epoch": 0.8572411420708634, + "grad_norm": 0.7056293609962258, + "learning_rate": 5.2457326279013006e-05, + "loss": 2.2748, + "step": 1869 + }, + { + "epoch": 0.8576998050682261, + "grad_norm": 0.6860723169115913, + "learning_rate": 5.2126500818844514e-05, + "loss": 2.3712, + "step": 1870 + }, + { + "epoch": 0.8581584680655888, + "grad_norm": 0.6743037357035669, + "learning_rate": 5.1796664469752566e-05, + "loss": 2.3395, + "step": 1871 + }, + { + "epoch": 0.8586171310629515, + "grad_norm": 0.7264903623616964, + "learning_rate": 5.1467817960167975e-05, + "loss": 2.2793, + "step": 1872 + }, + { + "epoch": 0.8590757940603142, + "grad_norm": 0.7105602432264334, + "learning_rate": 5.113996201633536e-05, + "loss": 2.2648, + "step": 1873 + }, + { + "epoch": 0.8595344570576768, + "grad_norm": 0.726099196736472, + "learning_rate": 5.0813097362311765e-05, + "loss": 2.2623, + "step": 1874 + }, + { + "epoch": 0.8599931200550396, + "grad_norm": 0.7158843288247695, + "learning_rate": 5.048722471996475e-05, + "loss": 2.152, + "step": 1875 + }, + { + "epoch": 0.8604517830524022, + "grad_norm": 0.7100627349806634, + "learning_rate": 5.016234480897158e-05, + "loss": 2.3239, + "step": 1876 + }, + { + "epoch": 0.860910446049765, + "grad_norm": 0.6706997788302599, + "learning_rate": 4.9838458346816664e-05, + "loss": 2.1927, + "step": 1877 + }, + { + "epoch": 0.8613691090471276, + "grad_norm": 0.6817723056734342, + "learning_rate": 4.9515566048790485e-05, + "loss": 2.2959, + "step": 1878 + }, + { + "epoch": 0.8618277720444903, + "grad_norm": 0.7888017187700392, + "learning_rate": 4.9193668627988074e-05, + "loss": 2.2735, + "step": 1879 + }, + { + "epoch": 0.862286435041853, + "grad_norm": 0.6724522337554605, + "learning_rate": 4.887276679530744e-05, + "loss": 2.3031, + "step": 1880 + }, + { + "epoch": 0.8627450980392157, + "grad_norm": 0.7392412512896153, + "learning_rate": 4.855286125944752e-05, + "loss": 2.3114, + "step": 1881 + }, + { + "epoch": 0.8632037610365784, + "grad_norm": 0.6780931140607096, + "learning_rate": 4.8233952726907224e-05, + "loss": 2.2948, + "step": 1882 + }, + { + "epoch": 0.8636624240339411, + "grad_norm": 0.7294762391885835, + "learning_rate": 4.7916041901983565e-05, + "loss": 2.3068, + "step": 1883 + }, + { + "epoch": 0.8641210870313037, + "grad_norm": 0.7222303537935454, + "learning_rate": 4.7599129486770145e-05, + "loss": 2.2673, + "step": 1884 + }, + { + "epoch": 0.8645797500286664, + "grad_norm": 0.7368149023014149, + "learning_rate": 4.728321618115555e-05, + "loss": 2.2396, + "step": 1885 + }, + { + "epoch": 0.8650384130260291, + "grad_norm": 0.7213454517673417, + "learning_rate": 4.696830268282204e-05, + "loss": 2.2209, + "step": 1886 + }, + { + "epoch": 0.8654970760233918, + "grad_norm": 0.6693876698634056, + "learning_rate": 4.665438968724361e-05, + "loss": 2.3052, + "step": 1887 + }, + { + "epoch": 0.8659557390207545, + "grad_norm": 0.695766398597989, + "learning_rate": 4.634147788768489e-05, + "loss": 2.2305, + "step": 1888 + }, + { + "epoch": 0.8664144020181171, + "grad_norm": 0.6708466974744939, + "learning_rate": 4.6029567975199414e-05, + "loss": 2.319, + "step": 1889 + }, + { + "epoch": 0.8668730650154799, + "grad_norm": 0.6941333220293577, + "learning_rate": 4.571866063862795e-05, + "loss": 2.271, + "step": 1890 + }, + { + "epoch": 0.8673317280128425, + "grad_norm": 0.7035306982099311, + "learning_rate": 4.540875656459703e-05, + "loss": 2.3943, + "step": 1891 + }, + { + "epoch": 0.8677903910102053, + "grad_norm": 0.7163913195372087, + "learning_rate": 4.509985643751785e-05, + "loss": 2.2928, + "step": 1892 + }, + { + "epoch": 0.8682490540075679, + "grad_norm": 0.7004170020872968, + "learning_rate": 4.479196093958421e-05, + "loss": 2.244, + "step": 1893 + }, + { + "epoch": 0.8687077170049307, + "grad_norm": 0.6807182541836552, + "learning_rate": 4.4485070750771187e-05, + "loss": 2.3488, + "step": 1894 + }, + { + "epoch": 0.8691663800022933, + "grad_norm": 0.694265859805762, + "learning_rate": 4.417918654883363e-05, + "loss": 2.2359, + "step": 1895 + }, + { + "epoch": 0.869625042999656, + "grad_norm": 0.7697401947470625, + "learning_rate": 4.3874309009305e-05, + "loss": 2.2838, + "step": 1896 + }, + { + "epoch": 0.8700837059970187, + "grad_norm": 0.7503324850927082, + "learning_rate": 4.357043880549538e-05, + "loss": 2.3835, + "step": 1897 + }, + { + "epoch": 0.8705423689943814, + "grad_norm": 0.6911983913162103, + "learning_rate": 4.326757660849012e-05, + "loss": 2.3557, + "step": 1898 + }, + { + "epoch": 0.8710010319917441, + "grad_norm": 0.696428601391198, + "learning_rate": 4.2965723087148635e-05, + "loss": 2.2471, + "step": 1899 + }, + { + "epoch": 0.8714596949891068, + "grad_norm": 0.6917064429579151, + "learning_rate": 4.266487890810256e-05, + "loss": 2.2278, + "step": 1900 + }, + { + "epoch": 0.8719183579864694, + "grad_norm": 0.6899767279027903, + "learning_rate": 4.2365044735754365e-05, + "loss": 2.2129, + "step": 1901 + }, + { + "epoch": 0.8723770209838321, + "grad_norm": 0.6909653055323446, + "learning_rate": 4.2066221232276266e-05, + "loss": 2.23, + "step": 1902 + }, + { + "epoch": 0.8728356839811948, + "grad_norm": 0.7371118537244409, + "learning_rate": 4.176840905760815e-05, + "loss": 2.228, + "step": 1903 + }, + { + "epoch": 0.8732943469785575, + "grad_norm": 0.6937328685714104, + "learning_rate": 4.1471608869456443e-05, + "loss": 2.1793, + "step": 1904 + }, + { + "epoch": 0.8737530099759202, + "grad_norm": 0.7218889975368293, + "learning_rate": 4.117582132329284e-05, + "loss": 2.3919, + "step": 1905 + }, + { + "epoch": 0.8742116729732828, + "grad_norm": 0.6923403177418579, + "learning_rate": 4.088104707235263e-05, + "loss": 2.2961, + "step": 1906 + }, + { + "epoch": 0.8746703359706456, + "grad_norm": 0.6869174534854894, + "learning_rate": 4.058728676763313e-05, + "loss": 2.2993, + "step": 1907 + }, + { + "epoch": 0.8751289989680082, + "grad_norm": 0.6817680550478871, + "learning_rate": 4.0294541057892375e-05, + "loss": 2.2343, + "step": 1908 + }, + { + "epoch": 0.875587661965371, + "grad_norm": 0.6970445481343477, + "learning_rate": 4.000281058964794e-05, + "loss": 2.2049, + "step": 1909 + }, + { + "epoch": 0.8760463249627336, + "grad_norm": 0.7084838648071485, + "learning_rate": 3.971209600717507e-05, + "loss": 2.2331, + "step": 1910 + }, + { + "epoch": 0.8765049879600963, + "grad_norm": 0.7419554398176575, + "learning_rate": 3.9422397952505465e-05, + "loss": 2.1835, + "step": 1911 + }, + { + "epoch": 0.876963650957459, + "grad_norm": 0.6807444270666781, + "learning_rate": 3.913371706542596e-05, + "loss": 2.3159, + "step": 1912 + }, + { + "epoch": 0.8774223139548217, + "grad_norm": 0.6937387207867636, + "learning_rate": 3.884605398347707e-05, + "loss": 2.1888, + "step": 1913 + }, + { + "epoch": 0.8778809769521844, + "grad_norm": 0.7225281837858918, + "learning_rate": 3.8559409341951456e-05, + "loss": 2.3792, + "step": 1914 + }, + { + "epoch": 0.8783396399495471, + "grad_norm": 0.7059303675521572, + "learning_rate": 3.8273783773892404e-05, + "loss": 2.3027, + "step": 1915 + }, + { + "epoch": 0.8787983029469097, + "grad_norm": 0.6952126861932374, + "learning_rate": 3.798917791009293e-05, + "loss": 2.2227, + "step": 1916 + }, + { + "epoch": 0.8792569659442725, + "grad_norm": 0.7132087439098247, + "learning_rate": 3.770559237909393e-05, + "loss": 2.2174, + "step": 1917 + }, + { + "epoch": 0.8797156289416351, + "grad_norm": 0.6895822130460606, + "learning_rate": 3.742302780718288e-05, + "loss": 2.2138, + "step": 1918 + }, + { + "epoch": 0.8801742919389978, + "grad_norm": 0.6686860723790699, + "learning_rate": 3.7141484818392635e-05, + "loss": 2.1926, + "step": 1919 + }, + { + "epoch": 0.8806329549363605, + "grad_norm": 0.6622738324169312, + "learning_rate": 3.686096403449973e-05, + "loss": 2.2122, + "step": 1920 + }, + { + "epoch": 0.8810916179337231, + "grad_norm": 0.7150054428418974, + "learning_rate": 3.658146607502344e-05, + "loss": 2.2667, + "step": 1921 + }, + { + "epoch": 0.8815502809310859, + "grad_norm": 0.7029281960822855, + "learning_rate": 3.630299155722411e-05, + "loss": 2.3134, + "step": 1922 + }, + { + "epoch": 0.8820089439284485, + "grad_norm": 0.679974694789932, + "learning_rate": 3.6025541096101676e-05, + "loss": 2.2572, + "step": 1923 + }, + { + "epoch": 0.8824676069258113, + "grad_norm": 0.7453546489442789, + "learning_rate": 3.574911530439473e-05, + "loss": 2.3697, + "step": 1924 + }, + { + "epoch": 0.8829262699231739, + "grad_norm": 0.7234591434800279, + "learning_rate": 3.5473714792578606e-05, + "loss": 2.32, + "step": 1925 + }, + { + "epoch": 0.8833849329205367, + "grad_norm": 0.723165754732247, + "learning_rate": 3.519934016886478e-05, + "loss": 2.3789, + "step": 1926 + }, + { + "epoch": 0.8838435959178993, + "grad_norm": 0.6941586159097628, + "learning_rate": 3.4925992039198776e-05, + "loss": 2.36, + "step": 1927 + }, + { + "epoch": 0.884302258915262, + "grad_norm": 0.6982792815590584, + "learning_rate": 3.465367100725908e-05, + "loss": 2.3007, + "step": 1928 + }, + { + "epoch": 0.8847609219126247, + "grad_norm": 0.6842834909809307, + "learning_rate": 3.438237767445618e-05, + "loss": 2.2222, + "step": 1929 + }, + { + "epoch": 0.8852195849099874, + "grad_norm": 0.6878952558966556, + "learning_rate": 3.4112112639930804e-05, + "loss": 2.3103, + "step": 1930 + }, + { + "epoch": 0.8856782479073501, + "grad_norm": 0.6860485761071389, + "learning_rate": 3.3842876500552564e-05, + "loss": 2.2201, + "step": 1931 + }, + { + "epoch": 0.8861369109047128, + "grad_norm": 0.6848208191587728, + "learning_rate": 3.357466985091906e-05, + "loss": 2.3243, + "step": 1932 + }, + { + "epoch": 0.8865955739020754, + "grad_norm": 0.6688953380085922, + "learning_rate": 3.330749328335414e-05, + "loss": 2.2421, + "step": 1933 + }, + { + "epoch": 0.8870542368994382, + "grad_norm": 0.7275312343169247, + "learning_rate": 3.304134738790659e-05, + "loss": 2.2465, + "step": 1934 + }, + { + "epoch": 0.8875128998968008, + "grad_norm": 0.6954522109837235, + "learning_rate": 3.277623275234953e-05, + "loss": 2.2938, + "step": 1935 + }, + { + "epoch": 0.8879715628941636, + "grad_norm": 0.6764898040287263, + "learning_rate": 3.2512149962177994e-05, + "loss": 2.3305, + "step": 1936 + }, + { + "epoch": 0.8884302258915262, + "grad_norm": 0.7159526723942244, + "learning_rate": 3.224909960060851e-05, + "loss": 2.2024, + "step": 1937 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 0.7117319154949354, + "learning_rate": 3.198708224857755e-05, + "loss": 2.2951, + "step": 1938 + }, + { + "epoch": 0.8893475518862516, + "grad_norm": 0.6851520403959456, + "learning_rate": 3.172609848474023e-05, + "loss": 2.2314, + "step": 1939 + }, + { + "epoch": 0.8898062148836142, + "grad_norm": 0.660654301906724, + "learning_rate": 3.1466148885468895e-05, + "loss": 2.2977, + "step": 1940 + }, + { + "epoch": 0.890264877880977, + "grad_norm": 0.7001557462185042, + "learning_rate": 3.120723402485198e-05, + "loss": 2.2955, + "step": 1941 + }, + { + "epoch": 0.8907235408783396, + "grad_norm": 0.7177547841180721, + "learning_rate": 3.094935447469294e-05, + "loss": 2.2475, + "step": 1942 + }, + { + "epoch": 0.8911822038757024, + "grad_norm": 0.6930207577061464, + "learning_rate": 3.069251080450863e-05, + "loss": 2.3179, + "step": 1943 + }, + { + "epoch": 0.891640866873065, + "grad_norm": 0.6800868595068151, + "learning_rate": 3.0436703581528113e-05, + "loss": 2.2493, + "step": 1944 + }, + { + "epoch": 0.8920995298704277, + "grad_norm": 0.7412637141007179, + "learning_rate": 3.0181933370691694e-05, + "loss": 2.315, + "step": 1945 + }, + { + "epoch": 0.8925581928677904, + "grad_norm": 0.682772285839726, + "learning_rate": 2.9928200734649523e-05, + "loss": 2.1905, + "step": 1946 + }, + { + "epoch": 0.8930168558651531, + "grad_norm": 0.7111100561391451, + "learning_rate": 2.9675506233760142e-05, + "loss": 2.2478, + "step": 1947 + }, + { + "epoch": 0.8934755188625157, + "grad_norm": 0.744088239153943, + "learning_rate": 2.942385042608925e-05, + "loss": 2.277, + "step": 1948 + }, + { + "epoch": 0.8939341818598785, + "grad_norm": 0.6635145949657727, + "learning_rate": 2.9173233867409054e-05, + "loss": 2.2493, + "step": 1949 + }, + { + "epoch": 0.8943928448572411, + "grad_norm": 0.7035318817219004, + "learning_rate": 2.892365711119638e-05, + "loss": 2.2844, + "step": 1950 + }, + { + "epoch": 0.8948515078546039, + "grad_norm": 0.7002614885428391, + "learning_rate": 2.8675120708631596e-05, + "loss": 2.275, + "step": 1951 + }, + { + "epoch": 0.8953101708519665, + "grad_norm": 0.6749502166668139, + "learning_rate": 2.8427625208597764e-05, + "loss": 2.2385, + "step": 1952 + }, + { + "epoch": 0.8957688338493293, + "grad_norm": 0.7681318750528394, + "learning_rate": 2.8181171157678874e-05, + "loss": 2.3011, + "step": 1953 + }, + { + "epoch": 0.8962274968466919, + "grad_norm": 0.7003342016613688, + "learning_rate": 2.7935759100159053e-05, + "loss": 2.2904, + "step": 1954 + }, + { + "epoch": 0.8966861598440545, + "grad_norm": 0.6834217962078739, + "learning_rate": 2.7691389578021365e-05, + "loss": 2.2649, + "step": 1955 + }, + { + "epoch": 0.8971448228414173, + "grad_norm": 0.6757718364135981, + "learning_rate": 2.7448063130946223e-05, + "loss": 2.2486, + "step": 1956 + }, + { + "epoch": 0.8976034858387799, + "grad_norm": 0.6661339678010918, + "learning_rate": 2.7205780296310544e-05, + "loss": 2.2598, + "step": 1957 + }, + { + "epoch": 0.8980621488361427, + "grad_norm": 0.6717039441936317, + "learning_rate": 2.6964541609186378e-05, + "loss": 2.2486, + "step": 1958 + }, + { + "epoch": 0.8985208118335053, + "grad_norm": 0.7065877297256975, + "learning_rate": 2.6724347602340104e-05, + "loss": 2.2529, + "step": 1959 + }, + { + "epoch": 0.898979474830868, + "grad_norm": 0.6904485635553338, + "learning_rate": 2.6485198806230682e-05, + "loss": 2.2648, + "step": 1960 + }, + { + "epoch": 0.8994381378282307, + "grad_norm": 0.7488157672876494, + "learning_rate": 2.6247095749008797e-05, + "loss": 2.3428, + "step": 1961 + }, + { + "epoch": 0.8998968008255934, + "grad_norm": 0.6939747127509311, + "learning_rate": 2.6010038956515826e-05, + "loss": 2.351, + "step": 1962 + }, + { + "epoch": 0.9003554638229561, + "grad_norm": 0.7276433693918084, + "learning_rate": 2.5774028952282423e-05, + "loss": 2.267, + "step": 1963 + }, + { + "epoch": 0.9008141268203188, + "grad_norm": 0.7020216343802209, + "learning_rate": 2.5539066257527277e-05, + "loss": 2.2724, + "step": 1964 + }, + { + "epoch": 0.9012727898176814, + "grad_norm": 0.6770893796176066, + "learning_rate": 2.530515139115652e-05, + "loss": 2.3031, + "step": 1965 + }, + { + "epoch": 0.9017314528150442, + "grad_norm": 0.7041885203371918, + "learning_rate": 2.5072284869761874e-05, + "loss": 2.2776, + "step": 1966 + }, + { + "epoch": 0.9021901158124068, + "grad_norm": 0.7386870263577819, + "learning_rate": 2.4840467207619786e-05, + "loss": 2.2729, + "step": 1967 + }, + { + "epoch": 0.9026487788097696, + "grad_norm": 0.6783465394016758, + "learning_rate": 2.460969891669068e-05, + "loss": 2.2737, + "step": 1968 + }, + { + "epoch": 0.9031074418071322, + "grad_norm": 0.7065696306055351, + "learning_rate": 2.4379980506617272e-05, + "loss": 2.3234, + "step": 1969 + }, + { + "epoch": 0.903566104804495, + "grad_norm": 0.7026549446055456, + "learning_rate": 2.4151312484723464e-05, + "loss": 2.2908, + "step": 1970 + }, + { + "epoch": 0.9040247678018576, + "grad_norm": 0.7053719226881741, + "learning_rate": 2.3923695356013798e-05, + "loss": 2.2309, + "step": 1971 + }, + { + "epoch": 0.9044834307992202, + "grad_norm": 0.661332068461737, + "learning_rate": 2.3697129623171833e-05, + "loss": 2.244, + "step": 1972 + }, + { + "epoch": 0.904942093796583, + "grad_norm": 0.732645160539467, + "learning_rate": 2.3471615786559042e-05, + "loss": 2.3525, + "step": 1973 + }, + { + "epoch": 0.9054007567939456, + "grad_norm": 0.7219189632796258, + "learning_rate": 2.3247154344213818e-05, + "loss": 2.2508, + "step": 1974 + }, + { + "epoch": 0.9058594197913084, + "grad_norm": 0.7337638678741344, + "learning_rate": 2.3023745791850625e-05, + "loss": 2.2725, + "step": 1975 + }, + { + "epoch": 0.906318082788671, + "grad_norm": 0.689539515730944, + "learning_rate": 2.2801390622858354e-05, + "loss": 2.3403, + "step": 1976 + }, + { + "epoch": 0.9067767457860337, + "grad_norm": 0.7420128453867441, + "learning_rate": 2.2580089328299746e-05, + "loss": 2.223, + "step": 1977 + }, + { + "epoch": 0.9072354087833964, + "grad_norm": 0.6765874821594484, + "learning_rate": 2.235984239690997e-05, + "loss": 2.2979, + "step": 1978 + }, + { + "epoch": 0.9076940717807591, + "grad_norm": 0.67830749526862, + "learning_rate": 2.2140650315095934e-05, + "loss": 2.2604, + "step": 1979 + }, + { + "epoch": 0.9081527347781218, + "grad_norm": 0.702036966610248, + "learning_rate": 2.192251356693459e-05, + "loss": 2.3441, + "step": 1980 + }, + { + "epoch": 0.9086113977754845, + "grad_norm": 0.6681560428814803, + "learning_rate": 2.170543263417246e-05, + "loss": 2.3361, + "step": 1981 + }, + { + "epoch": 0.9090700607728471, + "grad_norm": 0.7168997949062959, + "learning_rate": 2.1489407996224286e-05, + "loss": 2.3434, + "step": 1982 + }, + { + "epoch": 0.9095287237702099, + "grad_norm": 0.7349071113188225, + "learning_rate": 2.127444013017199e-05, + "loss": 2.2945, + "step": 1983 + }, + { + "epoch": 0.9099873867675725, + "grad_norm": 0.6612247549284189, + "learning_rate": 2.1060529510763648e-05, + "loss": 2.2497, + "step": 1984 + }, + { + "epoch": 0.9104460497649353, + "grad_norm": 0.7313733412629088, + "learning_rate": 2.084767661041259e-05, + "loss": 2.3149, + "step": 1985 + }, + { + "epoch": 0.9109047127622979, + "grad_norm": 0.7112616075208962, + "learning_rate": 2.063588189919596e-05, + "loss": 2.2864, + "step": 1986 + }, + { + "epoch": 0.9113633757596606, + "grad_norm": 0.7382793630701779, + "learning_rate": 2.0425145844854275e-05, + "loss": 2.3087, + "step": 1987 + }, + { + "epoch": 0.9118220387570233, + "grad_norm": 0.7328317543148697, + "learning_rate": 2.0215468912789693e-05, + "loss": 2.3042, + "step": 1988 + }, + { + "epoch": 0.9122807017543859, + "grad_norm": 0.7170606217025706, + "learning_rate": 2.0006851566065575e-05, + "loss": 2.2563, + "step": 1989 + }, + { + "epoch": 0.9127393647517487, + "grad_norm": 0.7061245386221018, + "learning_rate": 1.9799294265405166e-05, + "loss": 2.2116, + "step": 1990 + }, + { + "epoch": 0.9131980277491113, + "grad_norm": 0.7141394887906306, + "learning_rate": 1.9592797469190572e-05, + "loss": 2.3159, + "step": 1991 + }, + { + "epoch": 0.913656690746474, + "grad_norm": 0.7015058534189565, + "learning_rate": 1.938736163346194e-05, + "loss": 2.2562, + "step": 1992 + }, + { + "epoch": 0.9141153537438367, + "grad_norm": 0.7155021843141384, + "learning_rate": 1.9182987211916246e-05, + "loss": 2.1964, + "step": 1993 + }, + { + "epoch": 0.9145740167411994, + "grad_norm": 0.6902614218806029, + "learning_rate": 1.8979674655906332e-05, + "loss": 2.2969, + "step": 1994 + }, + { + "epoch": 0.9150326797385621, + "grad_norm": 0.7157923628549034, + "learning_rate": 1.8777424414440024e-05, + "loss": 2.1866, + "step": 1995 + }, + { + "epoch": 0.9154913427359248, + "grad_norm": 0.7386998633747174, + "learning_rate": 1.8576236934179202e-05, + "loss": 2.3281, + "step": 1996 + }, + { + "epoch": 0.9159500057332874, + "grad_norm": 0.7280175233827451, + "learning_rate": 1.8376112659438393e-05, + "loss": 2.3567, + "step": 1997 + }, + { + "epoch": 0.9164086687306502, + "grad_norm": 0.7287977466489638, + "learning_rate": 1.8177052032184282e-05, + "loss": 2.2312, + "step": 1998 + }, + { + "epoch": 0.9168673317280128, + "grad_norm": 0.7393193265592333, + "learning_rate": 1.7979055492034435e-05, + "loss": 2.3529, + "step": 1999 + }, + { + "epoch": 0.9173259947253756, + "grad_norm": 0.7194058775463095, + "learning_rate": 1.7782123476256407e-05, + "loss": 2.2796, + "step": 2000 + }, + { + "epoch": 0.9177846577227382, + "grad_norm": 0.6843501291223925, + "learning_rate": 1.7586256419766965e-05, + "loss": 2.2987, + "step": 2001 + }, + { + "epoch": 0.918243320720101, + "grad_norm": 0.6749228346850139, + "learning_rate": 1.7391454755130766e-05, + "loss": 2.2362, + "step": 2002 + }, + { + "epoch": 0.9187019837174636, + "grad_norm": 0.7019626036153696, + "learning_rate": 1.7197718912559557e-05, + "loss": 2.2444, + "step": 2003 + }, + { + "epoch": 0.9191606467148263, + "grad_norm": 0.7007129651711157, + "learning_rate": 1.700504931991148e-05, + "loss": 2.2636, + "step": 2004 + }, + { + "epoch": 0.919619309712189, + "grad_norm": 0.7288139348028221, + "learning_rate": 1.681344640268978e-05, + "loss": 2.2799, + "step": 2005 + }, + { + "epoch": 0.9200779727095516, + "grad_norm": 0.7185020774694729, + "learning_rate": 1.6622910584041974e-05, + "loss": 2.371, + "step": 2006 + }, + { + "epoch": 0.9205366357069144, + "grad_norm": 0.677975927433444, + "learning_rate": 1.6433442284758903e-05, + "loss": 2.2226, + "step": 2007 + }, + { + "epoch": 0.920995298704277, + "grad_norm": 0.6847766208133587, + "learning_rate": 1.624504192327392e-05, + "loss": 2.2374, + "step": 2008 + }, + { + "epoch": 0.9214539617016397, + "grad_norm": 0.7061492630771982, + "learning_rate": 1.6057709915661856e-05, + "loss": 2.2413, + "step": 2009 + }, + { + "epoch": 0.9219126246990024, + "grad_norm": 0.6916690165295729, + "learning_rate": 1.5871446675638057e-05, + "loss": 2.3529, + "step": 2010 + }, + { + "epoch": 0.9223712876963651, + "grad_norm": 0.6824950964252795, + "learning_rate": 1.5686252614557638e-05, + "loss": 2.2235, + "step": 2011 + }, + { + "epoch": 0.9228299506937278, + "grad_norm": 0.7257113452836015, + "learning_rate": 1.5502128141414497e-05, + "loss": 2.3575, + "step": 2012 + }, + { + "epoch": 0.9232886136910905, + "grad_norm": 0.7425965797682782, + "learning_rate": 1.5319073662840188e-05, + "loss": 2.22, + "step": 2013 + }, + { + "epoch": 0.9237472766884531, + "grad_norm": 0.9138355102794862, + "learning_rate": 1.5137089583103391e-05, + "loss": 2.3632, + "step": 2014 + }, + { + "epoch": 0.9242059396858159, + "grad_norm": 0.7626317966249929, + "learning_rate": 1.4956176304108893e-05, + "loss": 2.3661, + "step": 2015 + }, + { + "epoch": 0.9246646026831785, + "grad_norm": 0.7814597514656433, + "learning_rate": 1.4776334225396481e-05, + "loss": 2.3341, + "step": 2016 + }, + { + "epoch": 0.9251232656805413, + "grad_norm": 0.6564225321777013, + "learning_rate": 1.4597563744140397e-05, + "loss": 2.2115, + "step": 2017 + }, + { + "epoch": 0.9255819286779039, + "grad_norm": 0.6963560590905644, + "learning_rate": 1.4419865255148269e-05, + "loss": 2.3318, + "step": 2018 + }, + { + "epoch": 0.9260405916752666, + "grad_norm": 0.7049613567789225, + "learning_rate": 1.4243239150860122e-05, + "loss": 2.1376, + "step": 2019 + }, + { + "epoch": 0.9264992546726293, + "grad_norm": 0.7274430022045003, + "learning_rate": 1.4067685821347932e-05, + "loss": 2.2224, + "step": 2020 + }, + { + "epoch": 0.926957917669992, + "grad_norm": 0.7076560650884703, + "learning_rate": 1.389320565431429e-05, + "loss": 2.3359, + "step": 2021 + }, + { + "epoch": 0.9274165806673547, + "grad_norm": 0.6885921408671922, + "learning_rate": 1.3719799035091851e-05, + "loss": 2.2377, + "step": 2022 + }, + { + "epoch": 0.9278752436647173, + "grad_norm": 0.6668382569283798, + "learning_rate": 1.3547466346642278e-05, + "loss": 2.2582, + "step": 2023 + }, + { + "epoch": 0.92833390666208, + "grad_norm": 0.6933545714348582, + "learning_rate": 1.3376207969555577e-05, + "loss": 2.2238, + "step": 2024 + }, + { + "epoch": 0.9287925696594427, + "grad_norm": 0.7110496508597225, + "learning_rate": 1.32060242820492e-05, + "loss": 2.2343, + "step": 2025 + }, + { + "epoch": 0.9292512326568054, + "grad_norm": 0.674409602613792, + "learning_rate": 1.3036915659967118e-05, + "loss": 2.2289, + "step": 2026 + }, + { + "epoch": 0.9297098956541681, + "grad_norm": 0.7051551422556654, + "learning_rate": 1.2868882476779087e-05, + "loss": 2.2873, + "step": 2027 + }, + { + "epoch": 0.9301685586515308, + "grad_norm": 0.7039612838193023, + "learning_rate": 1.2701925103579815e-05, + "loss": 2.2183, + "step": 2028 + }, + { + "epoch": 0.9306272216488934, + "grad_norm": 0.7019561674566873, + "learning_rate": 1.2536043909088191e-05, + "loss": 2.3011, + "step": 2029 + }, + { + "epoch": 0.9310858846462562, + "grad_norm": 0.7274694382170487, + "learning_rate": 1.2371239259646228e-05, + "loss": 2.2943, + "step": 2030 + }, + { + "epoch": 0.9315445476436188, + "grad_norm": 0.707734657330743, + "learning_rate": 1.2207511519218672e-05, + "loss": 2.3383, + "step": 2031 + }, + { + "epoch": 0.9320032106409816, + "grad_norm": 0.7247953086322461, + "learning_rate": 1.2044861049391676e-05, + "loss": 2.2768, + "step": 2032 + }, + { + "epoch": 0.9324618736383442, + "grad_norm": 0.6787909470997703, + "learning_rate": 1.1883288209372512e-05, + "loss": 2.2118, + "step": 2033 + }, + { + "epoch": 0.932920536635707, + "grad_norm": 0.7100095812702936, + "learning_rate": 1.1722793355988471e-05, + "loss": 2.3263, + "step": 2034 + }, + { + "epoch": 0.9333791996330696, + "grad_norm": 0.7347054819602289, + "learning_rate": 1.1563376843686135e-05, + "loss": 2.2878, + "step": 2035 + }, + { + "epoch": 0.9338378626304323, + "grad_norm": 0.7154185071397458, + "learning_rate": 1.140503902453055e-05, + "loss": 2.1805, + "step": 2036 + }, + { + "epoch": 0.934296525627795, + "grad_norm": 0.7611428664099517, + "learning_rate": 1.1247780248204665e-05, + "loss": 2.2512, + "step": 2037 + }, + { + "epoch": 0.9347551886251577, + "grad_norm": 0.7180969144783093, + "learning_rate": 1.1091600862008333e-05, + "loss": 2.3319, + "step": 2038 + }, + { + "epoch": 0.9352138516225204, + "grad_norm": 0.7008750368145901, + "learning_rate": 1.0936501210857652e-05, + "loss": 2.2827, + "step": 2039 + }, + { + "epoch": 0.935672514619883, + "grad_norm": 0.7448011769811299, + "learning_rate": 1.0782481637284013e-05, + "loss": 2.3246, + "step": 2040 + }, + { + "epoch": 0.9361311776172457, + "grad_norm": 0.7090388080484019, + "learning_rate": 1.0629542481433663e-05, + "loss": 2.2579, + "step": 2041 + }, + { + "epoch": 0.9365898406146084, + "grad_norm": 0.6822610901518209, + "learning_rate": 1.0477684081066751e-05, + "loss": 2.2898, + "step": 2042 + }, + { + "epoch": 0.9370485036119711, + "grad_norm": 0.7742047044700402, + "learning_rate": 1.0326906771556566e-05, + "loss": 2.3251, + "step": 2043 + }, + { + "epoch": 0.9375071666093338, + "grad_norm": 0.6931578474882044, + "learning_rate": 1.017721088588891e-05, + "loss": 2.2693, + "step": 2044 + }, + { + "epoch": 0.9379658296066965, + "grad_norm": 0.6819609707746846, + "learning_rate": 1.0028596754661334e-05, + "loss": 2.2094, + "step": 2045 + }, + { + "epoch": 0.9384244926040591, + "grad_norm": 0.7002117079513207, + "learning_rate": 9.881064706082298e-06, + "loss": 2.2794, + "step": 2046 + }, + { + "epoch": 0.9388831556014219, + "grad_norm": 0.8105192358098282, + "learning_rate": 9.734615065970454e-06, + "loss": 2.251, + "step": 2047 + }, + { + "epoch": 0.9393418185987845, + "grad_norm": 0.7321487166555516, + "learning_rate": 9.58924815775425e-06, + "loss": 2.2955, + "step": 2048 + }, + { + "epoch": 0.9398004815961473, + "grad_norm": 0.7209608286168983, + "learning_rate": 9.444964302470715e-06, + "loss": 2.2577, + "step": 2049 + }, + { + "epoch": 0.9402591445935099, + "grad_norm": 0.6886141911027771, + "learning_rate": 9.301763818765018e-06, + "loss": 2.2511, + "step": 2050 + }, + { + "epoch": 0.9407178075908726, + "grad_norm": 0.7086951581660775, + "learning_rate": 9.15964702288996e-06, + "loss": 2.3148, + "step": 2051 + }, + { + "epoch": 0.9411764705882353, + "grad_norm": 0.7379626157302152, + "learning_rate": 9.018614228704925e-06, + "loss": 2.2206, + "step": 2052 + }, + { + "epoch": 0.941635133585598, + "grad_norm": 0.6814767468645024, + "learning_rate": 8.878665747675152e-06, + "loss": 2.2258, + "step": 2053 + }, + { + "epoch": 0.9420937965829607, + "grad_norm": 0.7008878025680049, + "learning_rate": 8.739801888871469e-06, + "loss": 2.27, + "step": 2054 + }, + { + "epoch": 0.9425524595803234, + "grad_norm": 0.6678428620856237, + "learning_rate": 8.602022958969336e-06, + "loss": 2.3367, + "step": 2055 + }, + { + "epoch": 0.943011122577686, + "grad_norm": 0.7050069933970942, + "learning_rate": 8.465329262248078e-06, + "loss": 2.2343, + "step": 2056 + }, + { + "epoch": 0.9434697855750487, + "grad_norm": 0.669192659051319, + "learning_rate": 8.32972110059027e-06, + "loss": 2.2194, + "step": 2057 + }, + { + "epoch": 0.9439284485724114, + "grad_norm": 0.6849375323586148, + "learning_rate": 8.195198773481406e-06, + "loss": 2.3577, + "step": 2058 + }, + { + "epoch": 0.9443871115697741, + "grad_norm": 0.718882877350788, + "learning_rate": 8.061762578008613e-06, + "loss": 2.2742, + "step": 2059 + }, + { + "epoch": 0.9448457745671368, + "grad_norm": 0.7289768876549809, + "learning_rate": 7.929412808860559e-06, + "loss": 2.256, + "step": 2060 + }, + { + "epoch": 0.9453044375644994, + "grad_norm": 0.7076175452106162, + "learning_rate": 7.79814975832649e-06, + "loss": 2.3237, + "step": 2061 + }, + { + "epoch": 0.9457631005618622, + "grad_norm": 0.7323207310367837, + "learning_rate": 7.667973716295851e-06, + "loss": 2.2898, + "step": 2062 + }, + { + "epoch": 0.9462217635592248, + "grad_norm": 0.6832944453832507, + "learning_rate": 7.5388849702571205e-06, + "loss": 2.2768, + "step": 2063 + }, + { + "epoch": 0.9466804265565876, + "grad_norm": 0.6987167924731372, + "learning_rate": 7.4108838052979185e-06, + "loss": 2.2312, + "step": 2064 + }, + { + "epoch": 0.9471390895539502, + "grad_norm": 0.6907853951752295, + "learning_rate": 7.283970504103732e-06, + "loss": 2.2161, + "step": 2065 + }, + { + "epoch": 0.947597752551313, + "grad_norm": 0.7041429986256197, + "learning_rate": 7.1581453469575785e-06, + "loss": 2.2603, + "step": 2066 + }, + { + "epoch": 0.9480564155486756, + "grad_norm": 0.6856316792449985, + "learning_rate": 7.033408611739456e-06, + "loss": 2.2489, + "step": 2067 + }, + { + "epoch": 0.9485150785460383, + "grad_norm": 0.7423234268221894, + "learning_rate": 6.909760573925561e-06, + "loss": 2.3449, + "step": 2068 + }, + { + "epoch": 0.948973741543401, + "grad_norm": 0.6974697211877561, + "learning_rate": 6.787201506587626e-06, + "loss": 2.2118, + "step": 2069 + }, + { + "epoch": 0.9494324045407637, + "grad_norm": 0.6957574972905289, + "learning_rate": 6.66573168039264e-06, + "loss": 2.3174, + "step": 2070 + }, + { + "epoch": 0.9498910675381264, + "grad_norm": 0.678959485153809, + "learning_rate": 6.545351363601959e-06, + "loss": 2.3341, + "step": 2071 + }, + { + "epoch": 0.9503497305354891, + "grad_norm": 0.7476167387916053, + "learning_rate": 6.426060822070812e-06, + "loss": 2.2405, + "step": 2072 + }, + { + "epoch": 0.9508083935328517, + "grad_norm": 0.6772001637960124, + "learning_rate": 6.3078603192475716e-06, + "loss": 2.3442, + "step": 2073 + }, + { + "epoch": 0.9512670565302144, + "grad_norm": 0.7324120039745138, + "learning_rate": 6.1907501161735934e-06, + "loss": 2.3387, + "step": 2074 + }, + { + "epoch": 0.9517257195275771, + "grad_norm": 0.712075856106819, + "learning_rate": 6.074730471482049e-06, + "loss": 2.2152, + "step": 2075 + }, + { + "epoch": 0.9521843825249398, + "grad_norm": 0.7082985761360565, + "learning_rate": 5.959801641397755e-06, + "loss": 2.2463, + "step": 2076 + }, + { + "epoch": 0.9526430455223025, + "grad_norm": 0.6746405843245094, + "learning_rate": 5.845963879736627e-06, + "loss": 2.1901, + "step": 2077 + }, + { + "epoch": 0.9531017085196651, + "grad_norm": 0.6940912074101669, + "learning_rate": 5.733217437904892e-06, + "loss": 2.2558, + "step": 2078 + }, + { + "epoch": 0.9535603715170279, + "grad_norm": 0.6982729839861115, + "learning_rate": 5.621562564898597e-06, + "loss": 2.2931, + "step": 2079 + }, + { + "epoch": 0.9540190345143905, + "grad_norm": 0.6843802381435035, + "learning_rate": 5.51099950730316e-06, + "loss": 2.2657, + "step": 2080 + }, + { + "epoch": 0.9544776975117533, + "grad_norm": 0.6673287263918511, + "learning_rate": 5.401528509292763e-06, + "loss": 2.3676, + "step": 2081 + }, + { + "epoch": 0.9549363605091159, + "grad_norm": 0.7184815006427208, + "learning_rate": 5.2931498126298495e-06, + "loss": 2.3124, + "step": 2082 + }, + { + "epoch": 0.9553950235064786, + "grad_norm": 0.6782573277418295, + "learning_rate": 5.1858636566645135e-06, + "loss": 2.2903, + "step": 2083 + }, + { + "epoch": 0.9558536865038413, + "grad_norm": 0.7455559833965424, + "learning_rate": 5.0796702783340035e-06, + "loss": 2.2053, + "step": 2084 + }, + { + "epoch": 0.956312349501204, + "grad_norm": 0.7780413580077346, + "learning_rate": 4.97456991216233e-06, + "loss": 2.2273, + "step": 2085 + }, + { + "epoch": 0.9567710124985667, + "grad_norm": 0.7396011353428682, + "learning_rate": 4.870562790259325e-06, + "loss": 2.3201, + "step": 2086 + }, + { + "epoch": 0.9572296754959294, + "grad_norm": 0.7330606308183435, + "learning_rate": 4.7676491423208625e-06, + "loss": 2.198, + "step": 2087 + }, + { + "epoch": 0.957688338493292, + "grad_norm": 0.6754538762097755, + "learning_rate": 4.66582919562758e-06, + "loss": 2.2577, + "step": 2088 + }, + { + "epoch": 0.9581470014906548, + "grad_norm": 0.6941476232736947, + "learning_rate": 4.5651031750448825e-06, + "loss": 2.3032, + "step": 2089 + }, + { + "epoch": 0.9586056644880174, + "grad_norm": 0.7235376781092846, + "learning_rate": 4.465471303022217e-06, + "loss": 2.2238, + "step": 2090 + }, + { + "epoch": 0.9590643274853801, + "grad_norm": 0.7273462779241814, + "learning_rate": 4.366933799592743e-06, + "loss": 2.3346, + "step": 2091 + }, + { + "epoch": 0.9595229904827428, + "grad_norm": 0.7246055704730971, + "learning_rate": 4.269490882372551e-06, + "loss": 2.2269, + "step": 2092 + }, + { + "epoch": 0.9599816534801054, + "grad_norm": 0.6867353221092496, + "learning_rate": 4.1731427665606115e-06, + "loss": 2.289, + "step": 2093 + }, + { + "epoch": 0.9604403164774682, + "grad_norm": 0.729270478205683, + "learning_rate": 4.077889664937884e-06, + "loss": 2.2261, + "step": 2094 + }, + { + "epoch": 0.9608989794748308, + "grad_norm": 0.7066007249466947, + "learning_rate": 3.983731787867207e-06, + "loss": 2.2053, + "step": 2095 + }, + { + "epoch": 0.9613576424721936, + "grad_norm": 0.7265170150729597, + "learning_rate": 3.890669343292464e-06, + "loss": 2.3056, + "step": 2096 + }, + { + "epoch": 0.9618163054695562, + "grad_norm": 0.6877395016744653, + "learning_rate": 3.7987025367384743e-06, + "loss": 2.303, + "step": 2097 + }, + { + "epoch": 0.962274968466919, + "grad_norm": 0.675077745660936, + "learning_rate": 3.707831571310327e-06, + "loss": 2.2784, + "step": 2098 + }, + { + "epoch": 0.9627336314642816, + "grad_norm": 0.706813777606977, + "learning_rate": 3.6180566476929912e-06, + "loss": 2.2662, + "step": 2099 + }, + { + "epoch": 0.9631922944616443, + "grad_norm": 0.6756203427592616, + "learning_rate": 3.529377964150815e-06, + "loss": 2.3007, + "step": 2100 + }, + { + "epoch": 0.963650957459007, + "grad_norm": 0.7118617767882741, + "learning_rate": 3.441795716527307e-06, + "loss": 2.3446, + "step": 2101 + }, + { + "epoch": 0.9641096204563697, + "grad_norm": 0.7368666256329742, + "learning_rate": 3.355310098244302e-06, + "loss": 2.2708, + "step": 2102 + }, + { + "epoch": 0.9645682834537324, + "grad_norm": 0.7460522114380744, + "learning_rate": 3.269921300301959e-06, + "loss": 2.2374, + "step": 2103 + }, + { + "epoch": 0.9650269464510951, + "grad_norm": 0.7088673749686109, + "learning_rate": 3.1856295112780988e-06, + "loss": 2.2914, + "step": 2104 + }, + { + "epoch": 0.9654856094484577, + "grad_norm": 0.7058744329566917, + "learning_rate": 3.102434917327812e-06, + "loss": 2.2626, + "step": 2105 + }, + { + "epoch": 0.9659442724458205, + "grad_norm": 0.693389759984262, + "learning_rate": 3.0203377021831292e-06, + "loss": 2.3671, + "step": 2106 + }, + { + "epoch": 0.9664029354431831, + "grad_norm": 0.67703034530128, + "learning_rate": 2.939338047152573e-06, + "loss": 2.2022, + "step": 2107 + }, + { + "epoch": 0.9668615984405458, + "grad_norm": 0.7270409972222868, + "learning_rate": 2.8594361311206073e-06, + "loss": 2.215, + "step": 2108 + }, + { + "epoch": 0.9673202614379085, + "grad_norm": 0.6812455639651694, + "learning_rate": 2.7806321305475225e-06, + "loss": 2.2813, + "step": 2109 + }, + { + "epoch": 0.9677789244352711, + "grad_norm": 0.6848849286671119, + "learning_rate": 2.7029262194688818e-06, + "loss": 2.2639, + "step": 2110 + }, + { + "epoch": 0.9682375874326339, + "grad_norm": 0.669649488353452, + "learning_rate": 2.626318569495134e-06, + "loss": 2.2846, + "step": 2111 + }, + { + "epoch": 0.9686962504299965, + "grad_norm": 0.7009207664732416, + "learning_rate": 2.550809349811334e-06, + "loss": 2.2958, + "step": 2112 + }, + { + "epoch": 0.9691549134273593, + "grad_norm": 0.8522562198420134, + "learning_rate": 2.476398727176532e-06, + "loss": 2.2802, + "step": 2113 + }, + { + "epoch": 0.9696135764247219, + "grad_norm": 0.7136382263148967, + "learning_rate": 2.4030868659237204e-06, + "loss": 2.3009, + "step": 2114 + }, + { + "epoch": 0.9700722394220846, + "grad_norm": 0.6940289123088472, + "learning_rate": 2.3308739279593317e-06, + "loss": 2.262, + "step": 2115 + }, + { + "epoch": 0.9705309024194473, + "grad_norm": 0.705336940649416, + "learning_rate": 2.2597600727626845e-06, + "loss": 2.2237, + "step": 2116 + }, + { + "epoch": 0.97098956541681, + "grad_norm": 0.7099788959036498, + "learning_rate": 2.1897454573860387e-06, + "loss": 2.2689, + "step": 2117 + }, + { + "epoch": 0.9714482284141727, + "grad_norm": 0.6962298916544127, + "learning_rate": 2.1208302364538746e-06, + "loss": 2.2149, + "step": 2118 + }, + { + "epoch": 0.9719068914115354, + "grad_norm": 0.7048611211663033, + "learning_rate": 2.0530145621627804e-06, + "loss": 2.2763, + "step": 2119 + }, + { + "epoch": 0.972365554408898, + "grad_norm": 0.7107747594295706, + "learning_rate": 1.9862985842810653e-06, + "loss": 2.2598, + "step": 2120 + }, + { + "epoch": 0.9728242174062608, + "grad_norm": 0.7275012597652352, + "learning_rate": 1.920682450148259e-06, + "loss": 2.3096, + "step": 2121 + }, + { + "epoch": 0.9732828804036234, + "grad_norm": 0.696535673596614, + "learning_rate": 1.856166304675111e-06, + "loss": 2.3426, + "step": 2122 + }, + { + "epoch": 0.9737415434009862, + "grad_norm": 0.7160883587468454, + "learning_rate": 1.792750290342926e-06, + "loss": 2.3334, + "step": 2123 + }, + { + "epoch": 0.9742002063983488, + "grad_norm": 0.7817564068330422, + "learning_rate": 1.7304345472035632e-06, + "loss": 2.2799, + "step": 2124 + }, + { + "epoch": 0.9746588693957114, + "grad_norm": 0.6947163173021912, + "learning_rate": 1.6692192128788253e-06, + "loss": 2.3248, + "step": 2125 + }, + { + "epoch": 0.9751175323930742, + "grad_norm": 0.6913336079691447, + "learning_rate": 1.6091044225604035e-06, + "loss": 2.2614, + "step": 2126 + }, + { + "epoch": 0.9755761953904368, + "grad_norm": 0.669218108818466, + "learning_rate": 1.5500903090094888e-06, + "loss": 2.2004, + "step": 2127 + }, + { + "epoch": 0.9760348583877996, + "grad_norm": 0.7180761551226951, + "learning_rate": 1.492177002556383e-06, + "loss": 2.2257, + "step": 2128 + }, + { + "epoch": 0.9764935213851622, + "grad_norm": 0.6988763736366413, + "learning_rate": 1.4353646311004443e-06, + "loss": 2.2022, + "step": 2129 + }, + { + "epoch": 0.976952184382525, + "grad_norm": 0.7186752985709972, + "learning_rate": 1.3796533201094752e-06, + "loss": 2.2252, + "step": 2130 + }, + { + "epoch": 0.9774108473798876, + "grad_norm": 0.7509001640848512, + "learning_rate": 1.3250431926197793e-06, + "loss": 2.3587, + "step": 2131 + }, + { + "epoch": 0.9778695103772503, + "grad_norm": 0.7021147819742736, + "learning_rate": 1.2715343692356607e-06, + "loss": 2.3371, + "step": 2132 + }, + { + "epoch": 0.978328173374613, + "grad_norm": 0.7263923829389014, + "learning_rate": 1.2191269681292582e-06, + "loss": 2.32, + "step": 2133 + }, + { + "epoch": 0.9787868363719757, + "grad_norm": 0.7009895215427007, + "learning_rate": 1.1678211050402676e-06, + "loss": 2.3315, + "step": 2134 + }, + { + "epoch": 0.9792454993693384, + "grad_norm": 0.7304282897441534, + "learning_rate": 1.117616893275719e-06, + "loss": 2.3074, + "step": 2135 + }, + { + "epoch": 0.9797041623667011, + "grad_norm": 0.7020243154688917, + "learning_rate": 1.068514443709534e-06, + "loss": 2.2792, + "step": 2136 + }, + { + "epoch": 0.9801628253640637, + "grad_norm": 0.7143340057557146, + "learning_rate": 1.0205138647826905e-06, + "loss": 2.2841, + "step": 2137 + }, + { + "epoch": 0.9806214883614265, + "grad_norm": 0.7252533173174366, + "learning_rate": 9.73615262502503e-07, + "loss": 2.3577, + "step": 2138 + }, + { + "epoch": 0.9810801513587891, + "grad_norm": 0.8085019293301081, + "learning_rate": 9.278187404426763e-07, + "loss": 2.2985, + "step": 2139 + }, + { + "epoch": 0.9815388143561519, + "grad_norm": 0.7485940100536262, + "learning_rate": 8.831243997431404e-07, + "loss": 2.3652, + "step": 2140 + }, + { + "epoch": 0.9819974773535145, + "grad_norm": 0.6907909693074781, + "learning_rate": 8.395323391094944e-07, + "loss": 2.2984, + "step": 2141 + }, + { + "epoch": 0.9824561403508771, + "grad_norm": 0.7052604355321495, + "learning_rate": 7.970426548131183e-07, + "loss": 2.2199, + "step": 2142 + }, + { + "epoch": 0.9829148033482399, + "grad_norm": 0.688496142495421, + "learning_rate": 7.556554406908389e-07, + "loss": 2.1953, + "step": 2143 + }, + { + "epoch": 0.9833734663456025, + "grad_norm": 0.8168017084282585, + "learning_rate": 7.153707881446536e-07, + "loss": 2.4111, + "step": 2144 + }, + { + "epoch": 0.9838321293429653, + "grad_norm": 0.7226868527773749, + "learning_rate": 6.761887861417293e-07, + "loss": 2.205, + "step": 2145 + }, + { + "epoch": 0.9842907923403279, + "grad_norm": 0.7140175007180481, + "learning_rate": 6.381095212139032e-07, + "loss": 2.2185, + "step": 2146 + }, + { + "epoch": 0.9847494553376906, + "grad_norm": 0.7320008802178339, + "learning_rate": 6.011330774577384e-07, + "loss": 2.3254, + "step": 2147 + }, + { + "epoch": 0.9852081183350533, + "grad_norm": 0.7483980435669229, + "learning_rate": 5.652595365343016e-07, + "loss": 2.2379, + "step": 2148 + }, + { + "epoch": 0.985666781332416, + "grad_norm": 0.7139919366900324, + "learning_rate": 5.304889776688859e-07, + "loss": 2.3681, + "step": 2149 + }, + { + "epoch": 0.9861254443297787, + "grad_norm": 0.7017770535487478, + "learning_rate": 4.968214776508994e-07, + "loss": 2.2866, + "step": 2150 + }, + { + "epoch": 0.9865841073271414, + "grad_norm": 0.6886458351568808, + "learning_rate": 4.6425711083375454e-07, + "loss": 2.1937, + "step": 2151 + }, + { + "epoch": 0.987042770324504, + "grad_norm": 0.7068088413380834, + "learning_rate": 4.3279594913447906e-07, + "loss": 2.2407, + "step": 2152 + }, + { + "epoch": 0.9875014333218668, + "grad_norm": 0.6837871673086611, + "learning_rate": 4.02438062033883e-07, + "loss": 2.1921, + "step": 2153 + }, + { + "epoch": 0.9879600963192294, + "grad_norm": 0.7011713149767097, + "learning_rate": 3.7318351657616987e-07, + "loss": 2.2455, + "step": 2154 + }, + { + "epoch": 0.9884187593165922, + "grad_norm": 0.7129165361407003, + "learning_rate": 3.4503237736882573e-07, + "loss": 2.3229, + "step": 2155 + }, + { + "epoch": 0.9888774223139548, + "grad_norm": 0.6843801418660341, + "learning_rate": 3.179847065825081e-07, + "loss": 2.2591, + "step": 2156 + }, + { + "epoch": 0.9893360853113176, + "grad_norm": 0.7150720658623877, + "learning_rate": 2.9204056395104594e-07, + "loss": 2.2432, + "step": 2157 + }, + { + "epoch": 0.9897947483086802, + "grad_norm": 0.6664999852971562, + "learning_rate": 2.672000067709956e-07, + "loss": 2.166, + "step": 2158 + }, + { + "epoch": 0.9902534113060428, + "grad_norm": 0.7498301910526124, + "learning_rate": 2.4346308990175204e-07, + "loss": 2.3155, + "step": 2159 + }, + { + "epoch": 0.9907120743034056, + "grad_norm": 0.6981694948181465, + "learning_rate": 2.208298657653818e-07, + "loss": 2.2716, + "step": 2160 + }, + { + "epoch": 0.9911707373007682, + "grad_norm": 0.6831694192415236, + "learning_rate": 1.9930038434645692e-07, + "loss": 2.2821, + "step": 2161 + }, + { + "epoch": 0.991629400298131, + "grad_norm": 0.7101010641844268, + "learning_rate": 1.7887469319205484e-07, + "loss": 2.2752, + "step": 2162 + }, + { + "epoch": 0.9920880632954936, + "grad_norm": 0.6817862149045488, + "learning_rate": 1.5955283741142523e-07, + "loss": 2.3083, + "step": 2163 + }, + { + "epoch": 0.9925467262928563, + "grad_norm": 0.6947188112480146, + "learning_rate": 1.4133485967615655e-07, + "loss": 2.2868, + "step": 2164 + }, + { + "epoch": 0.993005389290219, + "grad_norm": 0.7288559074721155, + "learning_rate": 1.2422080021995407e-07, + "loss": 2.355, + "step": 2165 + }, + { + "epoch": 0.9934640522875817, + "grad_norm": 0.6905960762326931, + "learning_rate": 1.082106968385288e-07, + "loss": 2.1719, + "step": 2166 + }, + { + "epoch": 0.9939227152849444, + "grad_norm": 0.6949396461878191, + "learning_rate": 9.330458488959748e-08, + "loss": 2.357, + "step": 2167 + }, + { + "epoch": 0.9943813782823071, + "grad_norm": 0.7051173425678199, + "learning_rate": 7.950249729271608e-08, + "loss": 2.2244, + "step": 2168 + }, + { + "epoch": 0.9948400412796697, + "grad_norm": 0.7148391929436407, + "learning_rate": 6.680446452922429e-08, + "loss": 2.2988, + "step": 2169 + }, + { + "epoch": 0.9952987042770325, + "grad_norm": 0.6788134547463939, + "learning_rate": 5.521051464230098e-08, + "loss": 2.2214, + "step": 2170 + }, + { + "epoch": 0.9957573672743951, + "grad_norm": 0.6885536546296773, + "learning_rate": 4.4720673236631206e-08, + "loss": 2.3834, + "step": 2171 + }, + { + "epoch": 0.9962160302717579, + "grad_norm": 0.6909490358842927, + "learning_rate": 3.53349634786837e-08, + "loss": 2.2585, + "step": 2172 + }, + { + "epoch": 0.9966746932691205, + "grad_norm": 0.6737895475393393, + "learning_rate": 2.7053406096433365e-08, + "loss": 2.3381, + "step": 2173 + }, + { + "epoch": 0.9971333562664833, + "grad_norm": 0.6831075478376473, + "learning_rate": 1.987601937930572e-08, + "loss": 2.3047, + "step": 2174 + }, + { + "epoch": 0.9975920192638459, + "grad_norm": 0.6946109985279607, + "learning_rate": 1.3802819178398984e-08, + "loss": 2.2807, + "step": 2175 + }, + { + "epoch": 0.9980506822612085, + "grad_norm": 0.6927186349965557, + "learning_rate": 8.833818906039959e-09, + "loss": 2.1902, + "step": 2176 + }, + { + "epoch": 0.9985093452585713, + "grad_norm": 0.6651049568798221, + "learning_rate": 4.969029536061598e-09, + "loss": 2.2678, + "step": 2177 + }, + { + "epoch": 0.9989680082559339, + "grad_norm": 0.6962389717658996, + "learning_rate": 2.2084596038030037e-09, + "loss": 2.3043, + "step": 2178 + }, + { + "epoch": 0.9994266712532967, + "grad_norm": 0.7169864763919171, + "learning_rate": 5.521152057763601e-10, + "loss": 2.2911, + "step": 2179 + }, + { + "epoch": 0.9998853342506593, + "grad_norm": 0.7053586545453231, + "learning_rate": 0.0, + "loss": 2.2275, + "step": 2180 + }, + { + "epoch": 0.9998853342506593, + "step": 2180, + "total_flos": 415415144595456.0, + "train_loss": 2.480288564721379, + "train_runtime": 10343.1931, + "train_samples_per_second": 53.961, + "train_steps_per_second": 0.211 + } + ], + "logging_steps": 1.0, + "max_steps": 2180, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 24000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": false, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 415415144595456.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}