{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.000677048070413, "eval_steps": 500, "global_step": 2214, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0013540961408259986, "grad_norm": 2.4126635555318527, "learning_rate": 2.5000000000000004e-07, "loss": 3.1189, "step": 1 }, { "epoch": 0.002708192281651997, "grad_norm": 2.8141112841232654, "learning_rate": 5.000000000000001e-07, "loss": 2.6861, "step": 2 }, { "epoch": 0.004062288422477996, "grad_norm": 2.6738862039513274, "learning_rate": 7.5e-07, "loss": 3.3382, "step": 3 }, { "epoch": 0.005416384563303994, "grad_norm": 2.30287728842049, "learning_rate": 1.0000000000000002e-06, "loss": 2.4565, "step": 4 }, { "epoch": 0.006770480704129994, "grad_norm": 42.907191662846444, "learning_rate": 1.25e-06, "loss": 2.5327, "step": 5 }, { "epoch": 0.008124576844955992, "grad_norm": 2.9404364361574724, "learning_rate": 1.5e-06, "loss": 2.7222, "step": 6 }, { "epoch": 0.009478672985781991, "grad_norm": 1.9739382530413463, "learning_rate": 1.75e-06, "loss": 2.1649, "step": 7 }, { "epoch": 0.010832769126607989, "grad_norm": 3.463912464916458, "learning_rate": 2.0000000000000003e-06, "loss": 2.9378, "step": 8 }, { "epoch": 0.012186865267433988, "grad_norm": 2.207345459970691, "learning_rate": 2.25e-06, "loss": 3.6521, "step": 9 }, { "epoch": 0.013540961408259987, "grad_norm": 2.752714935099265, "learning_rate": 2.5e-06, "loss": 2.5672, "step": 10 }, { "epoch": 0.014895057549085985, "grad_norm": 2.7722408021236187, "learning_rate": 2.7500000000000004e-06, "loss": 3.0411, "step": 11 }, { "epoch": 0.016249153689911984, "grad_norm": 2.9367100387488185, "learning_rate": 3e-06, "loss": 2.8985, "step": 12 }, { "epoch": 0.017603249830737983, "grad_norm": 2.7601030286888797, "learning_rate": 3.2500000000000002e-06, "loss": 2.8232, "step": 13 }, { "epoch": 0.018957345971563982, "grad_norm": 3.184703964873984, "learning_rate": 3.5e-06, "loss": 2.9731, "step": 14 }, { "epoch": 0.020311442112389978, "grad_norm": 2.2952697179626105, "learning_rate": 3.7500000000000005e-06, "loss": 2.823, "step": 15 }, { "epoch": 0.021665538253215978, "grad_norm": 2.4601547818230913, "learning_rate": 4.000000000000001e-06, "loss": 2.7683, "step": 16 }, { "epoch": 0.023019634394041977, "grad_norm": 2.0169369154885217, "learning_rate": 4.25e-06, "loss": 2.566, "step": 17 }, { "epoch": 0.024373730534867976, "grad_norm": 1.8874822039928005, "learning_rate": 4.5e-06, "loss": 2.4032, "step": 18 }, { "epoch": 0.025727826675693975, "grad_norm": 1.4345346790759714, "learning_rate": 4.75e-06, "loss": 2.8099, "step": 19 }, { "epoch": 0.027081922816519974, "grad_norm": 2.5317082125734025, "learning_rate": 5e-06, "loss": 2.4169, "step": 20 }, { "epoch": 0.02843601895734597, "grad_norm": 1.4417852513317821, "learning_rate": 5.2500000000000006e-06, "loss": 3.0868, "step": 21 }, { "epoch": 0.02979011509817197, "grad_norm": 0.9044314800944161, "learning_rate": 5.500000000000001e-06, "loss": 2.7374, "step": 22 }, { "epoch": 0.03114421123899797, "grad_norm": 1.4184446315710213, "learning_rate": 5.75e-06, "loss": 2.5317, "step": 23 }, { "epoch": 0.03249830737982397, "grad_norm": 1.6900469714019433, "learning_rate": 6e-06, "loss": 2.4404, "step": 24 }, { "epoch": 0.033852403520649964, "grad_norm": 1.0886842936641719, "learning_rate": 6.25e-06, "loss": 2.8003, "step": 25 }, { "epoch": 0.035206499661475966, "grad_norm": 1.463135149887311, "learning_rate": 6.5000000000000004e-06, "loss": 1.8819, "step": 26 }, { "epoch": 0.03656059580230196, "grad_norm": 1.1222026842497739, "learning_rate": 6.750000000000001e-06, "loss": 2.2671, "step": 27 }, { "epoch": 0.037914691943127965, "grad_norm": 1.0087919882501093, "learning_rate": 7e-06, "loss": 2.4491, "step": 28 }, { "epoch": 0.03926878808395396, "grad_norm": 1.2761649114763567, "learning_rate": 7.25e-06, "loss": 2.7121, "step": 29 }, { "epoch": 0.040622884224779957, "grad_norm": 1.0185872249128933, "learning_rate": 7.500000000000001e-06, "loss": 2.6497, "step": 30 }, { "epoch": 0.04197698036560596, "grad_norm": 1.2789270965641044, "learning_rate": 7.75e-06, "loss": 2.668, "step": 31 }, { "epoch": 0.043331076506431955, "grad_norm": 1.1115471485822677, "learning_rate": 8.000000000000001e-06, "loss": 2.6767, "step": 32 }, { "epoch": 0.04468517264725796, "grad_norm": 1.0627476896958792, "learning_rate": 8.25e-06, "loss": 2.6262, "step": 33 }, { "epoch": 0.046039268788083954, "grad_norm": 1.1212584420019067, "learning_rate": 8.5e-06, "loss": 2.4854, "step": 34 }, { "epoch": 0.04739336492890995, "grad_norm": 0.714156466912055, "learning_rate": 8.750000000000001e-06, "loss": 2.3662, "step": 35 }, { "epoch": 0.04874746106973595, "grad_norm": 1.4520911693345544, "learning_rate": 9e-06, "loss": 2.7126, "step": 36 }, { "epoch": 0.05010155721056195, "grad_norm": 1.493660601298817, "learning_rate": 9.250000000000001e-06, "loss": 3.0131, "step": 37 }, { "epoch": 0.05145565335138795, "grad_norm": 1.2721754658047073, "learning_rate": 9.5e-06, "loss": 3.0784, "step": 38 }, { "epoch": 0.052809749492213946, "grad_norm": 0.8803541178840371, "learning_rate": 9.75e-06, "loss": 2.1007, "step": 39 }, { "epoch": 0.05416384563303995, "grad_norm": 2.0655028083700504, "learning_rate": 1e-05, "loss": 2.5182, "step": 40 }, { "epoch": 0.055517941773865945, "grad_norm": 1.0487372085128044, "learning_rate": 9.999997090241333e-06, "loss": 2.2771, "step": 41 }, { "epoch": 0.05687203791469194, "grad_norm": 0.8160509238097648, "learning_rate": 9.999988360968714e-06, "loss": 2.2042, "step": 42 }, { "epoch": 0.05822613405551794, "grad_norm": 0.8864211341294704, "learning_rate": 9.999973812192306e-06, "loss": 2.7951, "step": 43 }, { "epoch": 0.05958023019634394, "grad_norm": 0.7681880276765591, "learning_rate": 9.99995344392904e-06, "loss": 2.859, "step": 44 }, { "epoch": 0.06093432633716994, "grad_norm": 1.7239414196388092, "learning_rate": 9.999927256202626e-06, "loss": 2.7866, "step": 45 }, { "epoch": 0.06228842247799594, "grad_norm": 0.7879562468694528, "learning_rate": 9.999895249043542e-06, "loss": 2.1671, "step": 46 }, { "epoch": 0.06364251861882193, "grad_norm": 0.8376315627603867, "learning_rate": 9.99985742248904e-06, "loss": 2.6406, "step": 47 }, { "epoch": 0.06499661475964794, "grad_norm": 0.8839739394970294, "learning_rate": 9.999813776583148e-06, "loss": 2.3163, "step": 48 }, { "epoch": 0.06635071090047394, "grad_norm": 0.7373151185194805, "learning_rate": 9.999764311376664e-06, "loss": 2.5735, "step": 49 }, { "epoch": 0.06770480704129993, "grad_norm": 1.1273015474311354, "learning_rate": 9.999709026927162e-06, "loss": 2.2133, "step": 50 }, { "epoch": 0.06905890318212593, "grad_norm": 0.8478718145728893, "learning_rate": 9.99964792329899e-06, "loss": 2.4075, "step": 51 }, { "epoch": 0.07041299932295193, "grad_norm": 1.0667576833889847, "learning_rate": 9.999581000563265e-06, "loss": 2.6946, "step": 52 }, { "epoch": 0.07176709546377792, "grad_norm": 1.0620013368810508, "learning_rate": 9.999508258797876e-06, "loss": 2.6205, "step": 53 }, { "epoch": 0.07312119160460392, "grad_norm": 0.8072331151341536, "learning_rate": 9.999429698087491e-06, "loss": 2.5132, "step": 54 }, { "epoch": 0.07447528774542993, "grad_norm": 0.6900996192892075, "learning_rate": 9.999345318523544e-06, "loss": 2.5071, "step": 55 }, { "epoch": 0.07582938388625593, "grad_norm": 0.7145430930616551, "learning_rate": 9.999255120204248e-06, "loss": 2.6579, "step": 56 }, { "epoch": 0.07718348002708192, "grad_norm": 0.9813015245466205, "learning_rate": 9.999159103234582e-06, "loss": 2.1667, "step": 57 }, { "epoch": 0.07853757616790792, "grad_norm": 0.6801613421398072, "learning_rate": 9.999057267726304e-06, "loss": 2.2016, "step": 58 }, { "epoch": 0.07989167230873392, "grad_norm": 1.340722366857769, "learning_rate": 9.998949613797937e-06, "loss": 2.5416, "step": 59 }, { "epoch": 0.08124576844955991, "grad_norm": 0.8044848626001605, "learning_rate": 9.998836141574781e-06, "loss": 2.6258, "step": 60 }, { "epoch": 0.08259986459038592, "grad_norm": 0.9118107600823588, "learning_rate": 9.99871685118891e-06, "loss": 2.5182, "step": 61 }, { "epoch": 0.08395396073121192, "grad_norm": 0.8103939602292338, "learning_rate": 9.99859174277916e-06, "loss": 2.4597, "step": 62 }, { "epoch": 0.08530805687203792, "grad_norm": 0.7953248414753887, "learning_rate": 9.99846081649115e-06, "loss": 2.6165, "step": 63 }, { "epoch": 0.08666215301286391, "grad_norm": 0.6885480399283996, "learning_rate": 9.998324072477266e-06, "loss": 2.3687, "step": 64 }, { "epoch": 0.08801624915368991, "grad_norm": 0.968337926868628, "learning_rate": 9.99818151089666e-06, "loss": 2.4591, "step": 65 }, { "epoch": 0.08937034529451592, "grad_norm": 0.7154001784843524, "learning_rate": 9.998033131915266e-06, "loss": 2.2175, "step": 66 }, { "epoch": 0.0907244414353419, "grad_norm": 1.18262426992356, "learning_rate": 9.997878935705778e-06, "loss": 3.1191, "step": 67 }, { "epoch": 0.09207853757616791, "grad_norm": 0.7400521396635642, "learning_rate": 9.997718922447669e-06, "loss": 2.5125, "step": 68 }, { "epoch": 0.09343263371699391, "grad_norm": 0.8148295313823943, "learning_rate": 9.997553092327174e-06, "loss": 2.7714, "step": 69 }, { "epoch": 0.0947867298578199, "grad_norm": 0.7378483977717124, "learning_rate": 9.997381445537309e-06, "loss": 2.3631, "step": 70 }, { "epoch": 0.0961408259986459, "grad_norm": 0.6266190246235469, "learning_rate": 9.997203982277852e-06, "loss": 2.4936, "step": 71 }, { "epoch": 0.0974949221394719, "grad_norm": 1.50407707220557, "learning_rate": 9.997020702755353e-06, "loss": 2.7048, "step": 72 }, { "epoch": 0.0988490182802979, "grad_norm": 0.8300611098600416, "learning_rate": 9.996831607183132e-06, "loss": 2.7146, "step": 73 }, { "epoch": 0.1002031144211239, "grad_norm": 0.7552213758897641, "learning_rate": 9.996636695781276e-06, "loss": 2.5399, "step": 74 }, { "epoch": 0.1015572105619499, "grad_norm": 0.705291949087134, "learning_rate": 9.996435968776646e-06, "loss": 2.6843, "step": 75 }, { "epoch": 0.1029113067027759, "grad_norm": 0.7673837145200612, "learning_rate": 9.996229426402867e-06, "loss": 2.7562, "step": 76 }, { "epoch": 0.10426540284360189, "grad_norm": 0.7439456610505197, "learning_rate": 9.996017068900335e-06, "loss": 2.0488, "step": 77 }, { "epoch": 0.10561949898442789, "grad_norm": 0.8945277132428896, "learning_rate": 9.995798896516215e-06, "loss": 2.48, "step": 78 }, { "epoch": 0.1069735951252539, "grad_norm": 0.8901395084271732, "learning_rate": 9.995574909504434e-06, "loss": 2.6099, "step": 79 }, { "epoch": 0.1083276912660799, "grad_norm": 0.824778949464918, "learning_rate": 9.995345108125698e-06, "loss": 2.69, "step": 80 }, { "epoch": 0.10968178740690589, "grad_norm": 0.8965013093653279, "learning_rate": 9.995109492647467e-06, "loss": 1.8424, "step": 81 }, { "epoch": 0.11103588354773189, "grad_norm": 1.0115160015541755, "learning_rate": 9.99486806334398e-06, "loss": 2.3693, "step": 82 }, { "epoch": 0.11238997968855789, "grad_norm": 0.8624086155070334, "learning_rate": 9.994620820496234e-06, "loss": 2.6173, "step": 83 }, { "epoch": 0.11374407582938388, "grad_norm": 0.7454258690069137, "learning_rate": 9.994367764391998e-06, "loss": 2.6624, "step": 84 }, { "epoch": 0.11509817197020988, "grad_norm": 1.0680963212316428, "learning_rate": 9.994108895325802e-06, "loss": 2.478, "step": 85 }, { "epoch": 0.11645226811103589, "grad_norm": 0.8212102801807549, "learning_rate": 9.993844213598949e-06, "loss": 2.0231, "step": 86 }, { "epoch": 0.11780636425186188, "grad_norm": 1.0684938905029466, "learning_rate": 9.993573719519498e-06, "loss": 2.595, "step": 87 }, { "epoch": 0.11916046039268788, "grad_norm": 0.9867071296775374, "learning_rate": 9.993297413402282e-06, "loss": 2.3636, "step": 88 }, { "epoch": 0.12051455653351388, "grad_norm": 1.404157430750734, "learning_rate": 9.993015295568893e-06, "loss": 2.1992, "step": 89 }, { "epoch": 0.12186865267433988, "grad_norm": 0.8693790763825426, "learning_rate": 9.992727366347688e-06, "loss": 2.7016, "step": 90 }, { "epoch": 0.12322274881516587, "grad_norm": 0.7501248710359847, "learning_rate": 9.99243362607379e-06, "loss": 3.2797, "step": 91 }, { "epoch": 0.12457684495599188, "grad_norm": 1.0643003422752406, "learning_rate": 9.992134075089085e-06, "loss": 2.5437, "step": 92 }, { "epoch": 0.12593094109681788, "grad_norm": 0.9103830863991241, "learning_rate": 9.991828713742218e-06, "loss": 3.0088, "step": 93 }, { "epoch": 0.12728503723764387, "grad_norm": 1.142388030489728, "learning_rate": 9.991517542388605e-06, "loss": 2.4977, "step": 94 }, { "epoch": 0.12863913337846988, "grad_norm": 0.9920213403910143, "learning_rate": 9.991200561390417e-06, "loss": 2.5349, "step": 95 }, { "epoch": 0.12999322951929587, "grad_norm": 0.8981930618864583, "learning_rate": 9.990877771116588e-06, "loss": 2.8531, "step": 96 }, { "epoch": 0.13134732566012186, "grad_norm": 1.154244133374351, "learning_rate": 9.990549171942817e-06, "loss": 2.5311, "step": 97 }, { "epoch": 0.13270142180094788, "grad_norm": 0.7466294529414305, "learning_rate": 9.99021476425156e-06, "loss": 2.4112, "step": 98 }, { "epoch": 0.13405551794177387, "grad_norm": 1.1851681560394296, "learning_rate": 9.989874548432037e-06, "loss": 2.2467, "step": 99 }, { "epoch": 0.13540961408259986, "grad_norm": 1.295896953150764, "learning_rate": 9.989528524880225e-06, "loss": 2.5688, "step": 100 }, { "epoch": 0.13676371022342587, "grad_norm": 0.8892799849200678, "learning_rate": 9.989176693998863e-06, "loss": 2.151, "step": 101 }, { "epoch": 0.13811780636425186, "grad_norm": 1.133636165142649, "learning_rate": 9.988819056197448e-06, "loss": 1.8402, "step": 102 }, { "epoch": 0.13947190250507785, "grad_norm": 0.8016376414959372, "learning_rate": 9.988455611892237e-06, "loss": 2.3469, "step": 103 }, { "epoch": 0.14082599864590387, "grad_norm": 0.964428420697917, "learning_rate": 9.98808636150624e-06, "loss": 2.3062, "step": 104 }, { "epoch": 0.14218009478672985, "grad_norm": 0.7711620898648806, "learning_rate": 9.987711305469232e-06, "loss": 2.7206, "step": 105 }, { "epoch": 0.14353419092755584, "grad_norm": 0.7778439840699306, "learning_rate": 9.987330444217739e-06, "loss": 2.4899, "step": 106 }, { "epoch": 0.14488828706838186, "grad_norm": 0.9642875586985667, "learning_rate": 9.986943778195052e-06, "loss": 2.8345, "step": 107 }, { "epoch": 0.14624238320920785, "grad_norm": 0.8268272296488341, "learning_rate": 9.98655130785121e-06, "loss": 2.5926, "step": 108 }, { "epoch": 0.14759647935003387, "grad_norm": 0.7977529748864721, "learning_rate": 9.986153033643011e-06, "loss": 1.9835, "step": 109 }, { "epoch": 0.14895057549085985, "grad_norm": 1.140904527079228, "learning_rate": 9.985748956034007e-06, "loss": 2.2239, "step": 110 }, { "epoch": 0.15030467163168584, "grad_norm": 1.020890884482194, "learning_rate": 9.985339075494504e-06, "loss": 2.6764, "step": 111 }, { "epoch": 0.15165876777251186, "grad_norm": 0.8869762247213703, "learning_rate": 9.984923392501567e-06, "loss": 2.2497, "step": 112 }, { "epoch": 0.15301286391333785, "grad_norm": 0.9590605445485386, "learning_rate": 9.98450190753901e-06, "loss": 2.3991, "step": 113 }, { "epoch": 0.15436696005416384, "grad_norm": 1.27578790045338, "learning_rate": 9.984074621097397e-06, "loss": 2.3551, "step": 114 }, { "epoch": 0.15572105619498985, "grad_norm": 1.0374791947621171, "learning_rate": 9.983641533674053e-06, "loss": 2.6919, "step": 115 }, { "epoch": 0.15707515233581584, "grad_norm": 0.923336386037252, "learning_rate": 9.983202645773049e-06, "loss": 2.6477, "step": 116 }, { "epoch": 0.15842924847664183, "grad_norm": 0.8526247474365115, "learning_rate": 9.982757957905204e-06, "loss": 2.2264, "step": 117 }, { "epoch": 0.15978334461746785, "grad_norm": 1.1258963168949678, "learning_rate": 9.982307470588097e-06, "loss": 2.5224, "step": 118 }, { "epoch": 0.16113744075829384, "grad_norm": 0.7569639815489783, "learning_rate": 9.98185118434605e-06, "loss": 1.8855, "step": 119 }, { "epoch": 0.16249153689911983, "grad_norm": 1.0355226629532543, "learning_rate": 9.981389099710132e-06, "loss": 3.0085, "step": 120 }, { "epoch": 0.16384563303994584, "grad_norm": 1.1486309630139306, "learning_rate": 9.980921217218173e-06, "loss": 2.7178, "step": 121 }, { "epoch": 0.16519972918077183, "grad_norm": 0.7466296304493713, "learning_rate": 9.980447537414736e-06, "loss": 2.251, "step": 122 }, { "epoch": 0.16655382532159782, "grad_norm": 0.651170263411467, "learning_rate": 9.979968060851144e-06, "loss": 2.1519, "step": 123 }, { "epoch": 0.16790792146242384, "grad_norm": 1.008261184550054, "learning_rate": 9.979482788085455e-06, "loss": 2.2199, "step": 124 }, { "epoch": 0.16926201760324983, "grad_norm": 2.416142037739212, "learning_rate": 9.978991719682486e-06, "loss": 2.3314, "step": 125 }, { "epoch": 0.17061611374407584, "grad_norm": 0.8174937699006882, "learning_rate": 9.97849485621379e-06, "loss": 2.5571, "step": 126 }, { "epoch": 0.17197020988490183, "grad_norm": 1.0052455897777324, "learning_rate": 9.977992198257668e-06, "loss": 2.5827, "step": 127 }, { "epoch": 0.17332430602572782, "grad_norm": 0.8663685595604006, "learning_rate": 9.977483746399168e-06, "loss": 2.5957, "step": 128 }, { "epoch": 0.17467840216655384, "grad_norm": 0.8454778683126596, "learning_rate": 9.976969501230074e-06, "loss": 2.2387, "step": 129 }, { "epoch": 0.17603249830737983, "grad_norm": 1.1589088141787116, "learning_rate": 9.976449463348924e-06, "loss": 2.721, "step": 130 }, { "epoch": 0.17738659444820581, "grad_norm": 0.8357485092208772, "learning_rate": 9.975923633360985e-06, "loss": 2.1065, "step": 131 }, { "epoch": 0.17874069058903183, "grad_norm": 1.0483474542727864, "learning_rate": 9.975392011878278e-06, "loss": 2.5342, "step": 132 }, { "epoch": 0.18009478672985782, "grad_norm": 0.8310501434875607, "learning_rate": 9.974854599519557e-06, "loss": 2.9813, "step": 133 }, { "epoch": 0.1814488828706838, "grad_norm": 1.1998035390299338, "learning_rate": 9.974311396910317e-06, "loss": 2.5218, "step": 134 }, { "epoch": 0.18280297901150983, "grad_norm": 1.1877384820263168, "learning_rate": 9.973762404682795e-06, "loss": 3.0512, "step": 135 }, { "epoch": 0.18415707515233581, "grad_norm": 0.7431690298915412, "learning_rate": 9.973207623475964e-06, "loss": 1.9473, "step": 136 }, { "epoch": 0.1855111712931618, "grad_norm": 0.8085958372588746, "learning_rate": 9.972647053935536e-06, "loss": 2.2599, "step": 137 }, { "epoch": 0.18686526743398782, "grad_norm": 1.0088420757147314, "learning_rate": 9.972080696713962e-06, "loss": 2.3532, "step": 138 }, { "epoch": 0.1882193635748138, "grad_norm": 0.871056602633167, "learning_rate": 9.971508552470424e-06, "loss": 2.1344, "step": 139 }, { "epoch": 0.1895734597156398, "grad_norm": 0.9551730724308998, "learning_rate": 9.970930621870843e-06, "loss": 1.7945, "step": 140 }, { "epoch": 0.1909275558564658, "grad_norm": 0.9099059345338785, "learning_rate": 9.970346905587875e-06, "loss": 2.4697, "step": 141 }, { "epoch": 0.1922816519972918, "grad_norm": 2.3602640235819288, "learning_rate": 9.969757404300911e-06, "loss": 2.3396, "step": 142 }, { "epoch": 0.19363574813811782, "grad_norm": 1.0387196911625636, "learning_rate": 9.969162118696072e-06, "loss": 2.2526, "step": 143 }, { "epoch": 0.1949898442789438, "grad_norm": 1.0660871258756586, "learning_rate": 9.968561049466214e-06, "loss": 2.2518, "step": 144 }, { "epoch": 0.1963439404197698, "grad_norm": 0.994954020536128, "learning_rate": 9.967954197310922e-06, "loss": 2.1365, "step": 145 }, { "epoch": 0.1976980365605958, "grad_norm": 1.1172902304571317, "learning_rate": 9.967341562936515e-06, "loss": 2.4633, "step": 146 }, { "epoch": 0.1990521327014218, "grad_norm": 0.8851246859608983, "learning_rate": 9.966723147056036e-06, "loss": 2.3302, "step": 147 }, { "epoch": 0.2004062288422478, "grad_norm": 0.9673278398671098, "learning_rate": 9.966098950389268e-06, "loss": 2.3481, "step": 148 }, { "epoch": 0.2017603249830738, "grad_norm": 1.0923553526656322, "learning_rate": 9.965468973662712e-06, "loss": 2.7291, "step": 149 }, { "epoch": 0.2031144211238998, "grad_norm": 0.6885203668960329, "learning_rate": 9.9648332176096e-06, "loss": 2.4216, "step": 150 }, { "epoch": 0.20446851726472579, "grad_norm": 0.8835456349684598, "learning_rate": 9.964191682969891e-06, "loss": 2.5524, "step": 151 }, { "epoch": 0.2058226134055518, "grad_norm": 1.110246201094945, "learning_rate": 9.96354437049027e-06, "loss": 2.7307, "step": 152 }, { "epoch": 0.2071767095463778, "grad_norm": 0.8335391435494296, "learning_rate": 9.962891280924148e-06, "loss": 2.2845, "step": 153 }, { "epoch": 0.20853080568720378, "grad_norm": 0.7269532952637685, "learning_rate": 9.962232415031653e-06, "loss": 2.2107, "step": 154 }, { "epoch": 0.2098849018280298, "grad_norm": 0.9516173733706272, "learning_rate": 9.961567773579645e-06, "loss": 2.549, "step": 155 }, { "epoch": 0.21123899796885579, "grad_norm": 0.9192964960712486, "learning_rate": 9.960897357341703e-06, "loss": 2.382, "step": 156 }, { "epoch": 0.21259309410968177, "grad_norm": 0.8697364214950628, "learning_rate": 9.960221167098124e-06, "loss": 2.7404, "step": 157 }, { "epoch": 0.2139471902505078, "grad_norm": 0.854037317845471, "learning_rate": 9.959539203635931e-06, "loss": 2.2796, "step": 158 }, { "epoch": 0.21530128639133378, "grad_norm": 9.059153514312463, "learning_rate": 9.958851467748863e-06, "loss": 2.1798, "step": 159 }, { "epoch": 0.2166553825321598, "grad_norm": 0.9333843627998799, "learning_rate": 9.958157960237376e-06, "loss": 2.3693, "step": 160 }, { "epoch": 0.21800947867298578, "grad_norm": 1.3475036120106114, "learning_rate": 9.957458681908647e-06, "loss": 2.4024, "step": 161 }, { "epoch": 0.21936357481381177, "grad_norm": 1.2121402320200159, "learning_rate": 9.956753633576571e-06, "loss": 2.5439, "step": 162 }, { "epoch": 0.2207176709546378, "grad_norm": 0.7792210046361225, "learning_rate": 9.956042816061752e-06, "loss": 2.0299, "step": 163 }, { "epoch": 0.22207176709546378, "grad_norm": 0.8226985573354776, "learning_rate": 9.955326230191517e-06, "loss": 2.8253, "step": 164 }, { "epoch": 0.22342586323628977, "grad_norm": 0.8424020935830455, "learning_rate": 9.9546038767999e-06, "loss": 2.2754, "step": 165 }, { "epoch": 0.22477995937711578, "grad_norm": 1.1561344660760495, "learning_rate": 9.95387575672765e-06, "loss": 1.9543, "step": 166 }, { "epoch": 0.22613405551794177, "grad_norm": 0.8810904992087591, "learning_rate": 9.953141870822232e-06, "loss": 2.8316, "step": 167 }, { "epoch": 0.22748815165876776, "grad_norm": 1.1327457989077157, "learning_rate": 9.952402219937817e-06, "loss": 2.1659, "step": 168 }, { "epoch": 0.22884224779959378, "grad_norm": 0.847237764337931, "learning_rate": 9.951656804935284e-06, "loss": 2.5104, "step": 169 }, { "epoch": 0.23019634394041977, "grad_norm": 1.3311662339401327, "learning_rate": 9.950905626682229e-06, "loss": 2.7411, "step": 170 }, { "epoch": 0.23155044008124576, "grad_norm": 0.8754683872906716, "learning_rate": 9.950148686052948e-06, "loss": 2.2843, "step": 171 }, { "epoch": 0.23290453622207177, "grad_norm": 1.1984097361442936, "learning_rate": 9.949385983928446e-06, "loss": 2.6191, "step": 172 }, { "epoch": 0.23425863236289776, "grad_norm": 1.042317105372754, "learning_rate": 9.948617521196438e-06, "loss": 2.5479, "step": 173 }, { "epoch": 0.23561272850372375, "grad_norm": 0.7787896671463191, "learning_rate": 9.947843298751337e-06, "loss": 2.3346, "step": 174 }, { "epoch": 0.23696682464454977, "grad_norm": 1.2170235533278835, "learning_rate": 9.947063317494265e-06, "loss": 2.2332, "step": 175 }, { "epoch": 0.23832092078537576, "grad_norm": 1.0390706514693488, "learning_rate": 9.946277578333045e-06, "loss": 2.5912, "step": 176 }, { "epoch": 0.23967501692620177, "grad_norm": 1.4816304376017728, "learning_rate": 9.945486082182201e-06, "loss": 2.4719, "step": 177 }, { "epoch": 0.24102911306702776, "grad_norm": 0.8487826972928669, "learning_rate": 9.944688829962957e-06, "loss": 2.2601, "step": 178 }, { "epoch": 0.24238320920785375, "grad_norm": 1.0267911713585076, "learning_rate": 9.94388582260324e-06, "loss": 2.4376, "step": 179 }, { "epoch": 0.24373730534867977, "grad_norm": 0.8526086813518827, "learning_rate": 9.943077061037672e-06, "loss": 2.6301, "step": 180 }, { "epoch": 0.24509140148950576, "grad_norm": 1.356267698059979, "learning_rate": 9.942262546207572e-06, "loss": 2.723, "step": 181 }, { "epoch": 0.24644549763033174, "grad_norm": 0.7426455232138849, "learning_rate": 9.94144227906096e-06, "loss": 2.4959, "step": 182 }, { "epoch": 0.24779959377115776, "grad_norm": 0.8750438781520365, "learning_rate": 9.940616260552545e-06, "loss": 2.3425, "step": 183 }, { "epoch": 0.24915368991198375, "grad_norm": 0.8785910140656594, "learning_rate": 9.939784491643734e-06, "loss": 2.2364, "step": 184 }, { "epoch": 0.25050778605280977, "grad_norm": 0.8938618124220896, "learning_rate": 9.938946973302624e-06, "loss": 2.19, "step": 185 }, { "epoch": 0.25186188219363576, "grad_norm": 1.021044776141451, "learning_rate": 9.938103706504007e-06, "loss": 2.7688, "step": 186 }, { "epoch": 0.25321597833446174, "grad_norm": 1.2815667789992267, "learning_rate": 9.937254692229363e-06, "loss": 2.1036, "step": 187 }, { "epoch": 0.25457007447528773, "grad_norm": 0.9597963383831872, "learning_rate": 9.936399931466866e-06, "loss": 2.7931, "step": 188 }, { "epoch": 0.2559241706161137, "grad_norm": 1.1496492688566942, "learning_rate": 9.935539425211371e-06, "loss": 2.0287, "step": 189 }, { "epoch": 0.25727826675693977, "grad_norm": 0.9029859331791751, "learning_rate": 9.934673174464426e-06, "loss": 2.555, "step": 190 }, { "epoch": 0.25863236289776576, "grad_norm": 0.8623354344497336, "learning_rate": 9.933801180234263e-06, "loss": 2.4571, "step": 191 }, { "epoch": 0.25998645903859174, "grad_norm": 1.172827727876581, "learning_rate": 9.932923443535798e-06, "loss": 2.5339, "step": 192 }, { "epoch": 0.26134055517941773, "grad_norm": 0.7844413148427782, "learning_rate": 9.932039965390634e-06, "loss": 2.232, "step": 193 }, { "epoch": 0.2626946513202437, "grad_norm": 2.4674022908729993, "learning_rate": 9.931150746827055e-06, "loss": 2.4686, "step": 194 }, { "epoch": 0.2640487474610697, "grad_norm": 0.8938856211560369, "learning_rate": 9.930255788880021e-06, "loss": 2.9519, "step": 195 }, { "epoch": 0.26540284360189575, "grad_norm": 0.9967031966791289, "learning_rate": 9.92935509259118e-06, "loss": 2.3506, "step": 196 }, { "epoch": 0.26675693974272174, "grad_norm": 0.8181579212619565, "learning_rate": 9.928448659008856e-06, "loss": 2.3992, "step": 197 }, { "epoch": 0.26811103588354773, "grad_norm": 0.9019033329287899, "learning_rate": 9.927536489188047e-06, "loss": 1.8896, "step": 198 }, { "epoch": 0.2694651320243737, "grad_norm": 1.196294290686767, "learning_rate": 9.926618584190435e-06, "loss": 2.6578, "step": 199 }, { "epoch": 0.2708192281651997, "grad_norm": 0.8815838170577491, "learning_rate": 9.925694945084369e-06, "loss": 2.8371, "step": 200 }, { "epoch": 0.27217332430602575, "grad_norm": 0.933774886972981, "learning_rate": 9.924765572944879e-06, "loss": 2.0347, "step": 201 }, { "epoch": 0.27352742044685174, "grad_norm": 0.9459867483651119, "learning_rate": 9.923830468853662e-06, "loss": 2.3687, "step": 202 }, { "epoch": 0.27488151658767773, "grad_norm": 0.8281062486347639, "learning_rate": 9.92288963389909e-06, "loss": 2.1651, "step": 203 }, { "epoch": 0.2762356127285037, "grad_norm": 0.7955305725702541, "learning_rate": 9.921943069176203e-06, "loss": 2.5368, "step": 204 }, { "epoch": 0.2775897088693297, "grad_norm": 1.026304391265509, "learning_rate": 9.920990775786712e-06, "loss": 2.9058, "step": 205 }, { "epoch": 0.2789438050101557, "grad_norm": 0.9971825829470802, "learning_rate": 9.920032754838994e-06, "loss": 2.8694, "step": 206 }, { "epoch": 0.28029790115098174, "grad_norm": 1.2484687747540484, "learning_rate": 9.919069007448093e-06, "loss": 2.4603, "step": 207 }, { "epoch": 0.28165199729180773, "grad_norm": 0.864015784119935, "learning_rate": 9.91809953473572e-06, "loss": 2.2217, "step": 208 }, { "epoch": 0.2830060934326337, "grad_norm": 0.9647460693812894, "learning_rate": 9.917124337830242e-06, "loss": 2.4799, "step": 209 }, { "epoch": 0.2843601895734597, "grad_norm": 1.0623146158662309, "learning_rate": 9.916143417866702e-06, "loss": 2.3235, "step": 210 }, { "epoch": 0.2857142857142857, "grad_norm": 1.5128627766738143, "learning_rate": 9.915156775986789e-06, "loss": 2.6927, "step": 211 }, { "epoch": 0.2870683818551117, "grad_norm": 0.9157505604273191, "learning_rate": 9.914164413338863e-06, "loss": 2.056, "step": 212 }, { "epoch": 0.28842247799593773, "grad_norm": 2.1672553239870114, "learning_rate": 9.913166331077937e-06, "loss": 2.3637, "step": 213 }, { "epoch": 0.2897765741367637, "grad_norm": 0.8565503250451203, "learning_rate": 9.912162530365683e-06, "loss": 2.3108, "step": 214 }, { "epoch": 0.2911306702775897, "grad_norm": 0.8071561412028346, "learning_rate": 9.911153012370427e-06, "loss": 2.4094, "step": 215 }, { "epoch": 0.2924847664184157, "grad_norm": 0.960350260627165, "learning_rate": 9.910137778267153e-06, "loss": 2.2326, "step": 216 }, { "epoch": 0.2938388625592417, "grad_norm": 1.1434301374173532, "learning_rate": 9.909116829237492e-06, "loss": 2.1396, "step": 217 }, { "epoch": 0.29519295870006773, "grad_norm": 1.0512962564460284, "learning_rate": 9.908090166469733e-06, "loss": 2.4862, "step": 218 }, { "epoch": 0.2965470548408937, "grad_norm": 0.8803334872680652, "learning_rate": 9.90705779115881e-06, "loss": 2.3063, "step": 219 }, { "epoch": 0.2979011509817197, "grad_norm": 0.7957632137271137, "learning_rate": 9.90601970450631e-06, "loss": 2.1209, "step": 220 }, { "epoch": 0.2992552471225457, "grad_norm": 1.5559922150186727, "learning_rate": 9.904975907720465e-06, "loss": 2.745, "step": 221 }, { "epoch": 0.3006093432633717, "grad_norm": 1.0287819888236789, "learning_rate": 9.903926402016153e-06, "loss": 2.3034, "step": 222 }, { "epoch": 0.3019634394041977, "grad_norm": 0.8755596479469875, "learning_rate": 9.902871188614898e-06, "loss": 2.6008, "step": 223 }, { "epoch": 0.3033175355450237, "grad_norm": 2.222977688536351, "learning_rate": 9.901810268744868e-06, "loss": 2.5897, "step": 224 }, { "epoch": 0.3046716316858497, "grad_norm": 0.8294734280823934, "learning_rate": 9.90074364364087e-06, "loss": 2.373, "step": 225 }, { "epoch": 0.3060257278266757, "grad_norm": 0.9955963385213202, "learning_rate": 9.899671314544352e-06, "loss": 2.8267, "step": 226 }, { "epoch": 0.3073798239675017, "grad_norm": 0.9863487128858249, "learning_rate": 9.898593282703402e-06, "loss": 2.3585, "step": 227 }, { "epoch": 0.3087339201083277, "grad_norm": 1.3476414208683485, "learning_rate": 9.897509549372745e-06, "loss": 2.0764, "step": 228 }, { "epoch": 0.31008801624915366, "grad_norm": 1.1281976478830502, "learning_rate": 9.896420115813741e-06, "loss": 2.1232, "step": 229 }, { "epoch": 0.3114421123899797, "grad_norm": 0.9905003772016358, "learning_rate": 9.89532498329439e-06, "loss": 2.0276, "step": 230 }, { "epoch": 0.3127962085308057, "grad_norm": 0.9642354729606564, "learning_rate": 9.894224153089313e-06, "loss": 1.7903, "step": 231 }, { "epoch": 0.3141503046716317, "grad_norm": 1.8609542881386758, "learning_rate": 9.893117626479778e-06, "loss": 2.6118, "step": 232 }, { "epoch": 0.3155044008124577, "grad_norm": 1.0931602473722466, "learning_rate": 9.892005404753669e-06, "loss": 2.4775, "step": 233 }, { "epoch": 0.31685849695328366, "grad_norm": 0.7800560012460497, "learning_rate": 9.890887489205507e-06, "loss": 2.1569, "step": 234 }, { "epoch": 0.3182125930941097, "grad_norm": 1.150743273577776, "learning_rate": 9.889763881136439e-06, "loss": 2.4256, "step": 235 }, { "epoch": 0.3195666892349357, "grad_norm": 0.9159836558576258, "learning_rate": 9.888634581854235e-06, "loss": 2.3495, "step": 236 }, { "epoch": 0.3209207853757617, "grad_norm": 1.0174141433627475, "learning_rate": 9.88749959267329e-06, "loss": 2.8219, "step": 237 }, { "epoch": 0.3222748815165877, "grad_norm": 0.85962610230906, "learning_rate": 9.886358914914624e-06, "loss": 2.3995, "step": 238 }, { "epoch": 0.32362897765741366, "grad_norm": 0.8714138576280448, "learning_rate": 9.885212549905874e-06, "loss": 2.1309, "step": 239 }, { "epoch": 0.32498307379823965, "grad_norm": 0.985120553099594, "learning_rate": 9.884060498981297e-06, "loss": 2.3078, "step": 240 }, { "epoch": 0.3263371699390657, "grad_norm": 1.1509228640558309, "learning_rate": 9.88290276348177e-06, "loss": 2.762, "step": 241 }, { "epoch": 0.3276912660798917, "grad_norm": 0.8569907806447795, "learning_rate": 9.881739344754789e-06, "loss": 2.4162, "step": 242 }, { "epoch": 0.3290453622207177, "grad_norm": 0.8953655946187061, "learning_rate": 9.880570244154455e-06, "loss": 2.1708, "step": 243 }, { "epoch": 0.33039945836154366, "grad_norm": 2.6172719168994782, "learning_rate": 9.879395463041493e-06, "loss": 2.4244, "step": 244 }, { "epoch": 0.33175355450236965, "grad_norm": 1.1692362620244634, "learning_rate": 9.87821500278323e-06, "loss": 2.3573, "step": 245 }, { "epoch": 0.33310765064319564, "grad_norm": 1.3260737783720347, "learning_rate": 9.877028864753614e-06, "loss": 2.2204, "step": 246 }, { "epoch": 0.3344617467840217, "grad_norm": 1.1284130571617974, "learning_rate": 9.87583705033319e-06, "loss": 2.3806, "step": 247 }, { "epoch": 0.3358158429248477, "grad_norm": 0.8740876722841778, "learning_rate": 9.874639560909118e-06, "loss": 1.8413, "step": 248 }, { "epoch": 0.33716993906567366, "grad_norm": 0.8380346921284223, "learning_rate": 9.87343639787516e-06, "loss": 2.0977, "step": 249 }, { "epoch": 0.33852403520649965, "grad_norm": 0.8382850128591168, "learning_rate": 9.87222756263168e-06, "loss": 2.1709, "step": 250 }, { "epoch": 0.33987813134732564, "grad_norm": 1.1238927174638607, "learning_rate": 9.871013056585646e-06, "loss": 2.4974, "step": 251 }, { "epoch": 0.3412322274881517, "grad_norm": 0.7904179666008488, "learning_rate": 9.869792881150624e-06, "loss": 2.6544, "step": 252 }, { "epoch": 0.3425863236289777, "grad_norm": 1.5604245138186315, "learning_rate": 9.868567037746784e-06, "loss": 2.4196, "step": 253 }, { "epoch": 0.34394041976980366, "grad_norm": 0.9541984067972759, "learning_rate": 9.867335527800887e-06, "loss": 2.437, "step": 254 }, { "epoch": 0.34529451591062965, "grad_norm": 1.1385820836347318, "learning_rate": 9.866098352746295e-06, "loss": 2.6164, "step": 255 }, { "epoch": 0.34664861205145564, "grad_norm": 0.8579010274341586, "learning_rate": 9.864855514022955e-06, "loss": 2.0521, "step": 256 }, { "epoch": 0.34800270819228163, "grad_norm": 1.0456339797858754, "learning_rate": 9.863607013077414e-06, "loss": 1.8313, "step": 257 }, { "epoch": 0.3493568043331077, "grad_norm": 0.9189984940983889, "learning_rate": 9.862352851362808e-06, "loss": 2.2923, "step": 258 }, { "epoch": 0.35071090047393366, "grad_norm": 1.0071081750591437, "learning_rate": 9.861093030338859e-06, "loss": 2.5982, "step": 259 }, { "epoch": 0.35206499661475965, "grad_norm": 0.7791457829988662, "learning_rate": 9.859827551471877e-06, "loss": 2.1629, "step": 260 }, { "epoch": 0.35341909275558564, "grad_norm": 1.3508700131079088, "learning_rate": 9.858556416234755e-06, "loss": 2.3283, "step": 261 }, { "epoch": 0.35477318889641163, "grad_norm": 1.4779336761646127, "learning_rate": 9.857279626106975e-06, "loss": 2.1788, "step": 262 }, { "epoch": 0.3561272850372376, "grad_norm": 0.8287354033035986, "learning_rate": 9.855997182574598e-06, "loss": 2.4326, "step": 263 }, { "epoch": 0.35748138117806366, "grad_norm": 1.2376776693679805, "learning_rate": 9.854709087130261e-06, "loss": 1.9501, "step": 264 }, { "epoch": 0.35883547731888965, "grad_norm": 0.9140324089635932, "learning_rate": 9.853415341273185e-06, "loss": 2.1236, "step": 265 }, { "epoch": 0.36018957345971564, "grad_norm": 0.8399357799885816, "learning_rate": 9.852115946509163e-06, "loss": 2.2648, "step": 266 }, { "epoch": 0.36154366960054163, "grad_norm": 1.0405854607468001, "learning_rate": 9.85081090435057e-06, "loss": 2.2106, "step": 267 }, { "epoch": 0.3628977657413676, "grad_norm": 0.8825320659362774, "learning_rate": 9.849500216316346e-06, "loss": 1.9139, "step": 268 }, { "epoch": 0.36425186188219366, "grad_norm": 0.8884433268338788, "learning_rate": 9.848183883932003e-06, "loss": 2.3441, "step": 269 }, { "epoch": 0.36560595802301965, "grad_norm": 0.8144408969478175, "learning_rate": 9.846861908729628e-06, "loss": 2.7794, "step": 270 }, { "epoch": 0.36696005416384564, "grad_norm": 3.0643775465082093, "learning_rate": 9.845534292247872e-06, "loss": 2.6927, "step": 271 }, { "epoch": 0.36831415030467163, "grad_norm": 1.0918166740808275, "learning_rate": 9.844201036031952e-06, "loss": 2.6845, "step": 272 }, { "epoch": 0.3696682464454976, "grad_norm": 1.0782629441989242, "learning_rate": 9.84286214163365e-06, "loss": 2.1323, "step": 273 }, { "epoch": 0.3710223425863236, "grad_norm": 0.899216159697703, "learning_rate": 9.841517610611309e-06, "loss": 2.5939, "step": 274 }, { "epoch": 0.37237643872714965, "grad_norm": 0.8784697558034725, "learning_rate": 9.840167444529834e-06, "loss": 2.0398, "step": 275 }, { "epoch": 0.37373053486797564, "grad_norm": 0.7893703178184768, "learning_rate": 9.838811644960686e-06, "loss": 2.3489, "step": 276 }, { "epoch": 0.37508463100880163, "grad_norm": 1.5759025892565637, "learning_rate": 9.837450213481888e-06, "loss": 2.1618, "step": 277 }, { "epoch": 0.3764387271496276, "grad_norm": 0.8915508527852146, "learning_rate": 9.836083151678014e-06, "loss": 2.2966, "step": 278 }, { "epoch": 0.3777928232904536, "grad_norm": 0.9656457496381649, "learning_rate": 9.834710461140191e-06, "loss": 2.2487, "step": 279 }, { "epoch": 0.3791469194312796, "grad_norm": 1.0512896665744822, "learning_rate": 9.833332143466099e-06, "loss": 2.4065, "step": 280 }, { "epoch": 0.38050101557210564, "grad_norm": 1.0658387523528818, "learning_rate": 9.831948200259966e-06, "loss": 2.4469, "step": 281 }, { "epoch": 0.3818551117129316, "grad_norm": 1.117757922229749, "learning_rate": 9.830558633132568e-06, "loss": 2.2469, "step": 282 }, { "epoch": 0.3832092078537576, "grad_norm": 0.8846882799705557, "learning_rate": 9.82916344370123e-06, "loss": 1.8672, "step": 283 }, { "epoch": 0.3845633039945836, "grad_norm": 0.8829678982284991, "learning_rate": 9.827762633589813e-06, "loss": 2.3709, "step": 284 }, { "epoch": 0.3859174001354096, "grad_norm": 1.5559611510853741, "learning_rate": 9.826356204428726e-06, "loss": 2.531, "step": 285 }, { "epoch": 0.38727149627623564, "grad_norm": 0.8832988192797496, "learning_rate": 9.82494415785492e-06, "loss": 2.0908, "step": 286 }, { "epoch": 0.3886255924170616, "grad_norm": 0.893313058823839, "learning_rate": 9.82352649551188e-06, "loss": 2.2141, "step": 287 }, { "epoch": 0.3899796885578876, "grad_norm": 1.915840114647375, "learning_rate": 9.822103219049625e-06, "loss": 2.3759, "step": 288 }, { "epoch": 0.3913337846987136, "grad_norm": 1.091811159406276, "learning_rate": 9.820674330124716e-06, "loss": 1.949, "step": 289 }, { "epoch": 0.3926878808395396, "grad_norm": 0.9101585142251976, "learning_rate": 9.819239830400238e-06, "loss": 2.2837, "step": 290 }, { "epoch": 0.3940419769803656, "grad_norm": 0.8669554630795423, "learning_rate": 9.81779972154581e-06, "loss": 2.3568, "step": 291 }, { "epoch": 0.3953960731211916, "grad_norm": 1.0812216653348674, "learning_rate": 9.816354005237583e-06, "loss": 2.5594, "step": 292 }, { "epoch": 0.3967501692620176, "grad_norm": 1.0135729926732555, "learning_rate": 9.814902683158227e-06, "loss": 2.2677, "step": 293 }, { "epoch": 0.3981042654028436, "grad_norm": 1.1354280196245004, "learning_rate": 9.813445756996946e-06, "loss": 2.1235, "step": 294 }, { "epoch": 0.3994583615436696, "grad_norm": 0.8872371772513353, "learning_rate": 9.811983228449457e-06, "loss": 2.6214, "step": 295 }, { "epoch": 0.4008124576844956, "grad_norm": 1.198158875138794, "learning_rate": 9.810515099218004e-06, "loss": 2.5482, "step": 296 }, { "epoch": 0.40216655382532157, "grad_norm": 1.122073024930223, "learning_rate": 9.809041371011347e-06, "loss": 2.3414, "step": 297 }, { "epoch": 0.4035206499661476, "grad_norm": 0.7450105063153143, "learning_rate": 9.807562045544764e-06, "loss": 2.5729, "step": 298 }, { "epoch": 0.4048747461069736, "grad_norm": 1.096232419286508, "learning_rate": 9.806077124540045e-06, "loss": 2.5012, "step": 299 }, { "epoch": 0.4062288422477996, "grad_norm": 1.0045118062203406, "learning_rate": 9.804586609725499e-06, "loss": 2.0684, "step": 300 }, { "epoch": 0.4075829383886256, "grad_norm": 1.0368766866449026, "learning_rate": 9.803090502835938e-06, "loss": 2.1287, "step": 301 }, { "epoch": 0.40893703452945157, "grad_norm": 1.0208588144634514, "learning_rate": 9.801588805612685e-06, "loss": 2.3341, "step": 302 }, { "epoch": 0.4102911306702776, "grad_norm": 1.0454209454651813, "learning_rate": 9.800081519803575e-06, "loss": 2.088, "step": 303 }, { "epoch": 0.4116452268111036, "grad_norm": 1.0546734132886375, "learning_rate": 9.798568647162939e-06, "loss": 2.3576, "step": 304 }, { "epoch": 0.4129993229519296, "grad_norm": 1.1121493752708365, "learning_rate": 9.797050189451615e-06, "loss": 2.1073, "step": 305 }, { "epoch": 0.4143534190927556, "grad_norm": 1.1189378911389287, "learning_rate": 9.795526148436945e-06, "loss": 2.308, "step": 306 }, { "epoch": 0.41570751523358157, "grad_norm": 0.9899429575476786, "learning_rate": 9.793996525892762e-06, "loss": 2.2555, "step": 307 }, { "epoch": 0.41706161137440756, "grad_norm": 1.0479984863056457, "learning_rate": 9.7924613235994e-06, "loss": 2.3882, "step": 308 }, { "epoch": 0.4184157075152336, "grad_norm": 0.8437095271965467, "learning_rate": 9.790920543343686e-06, "loss": 2.8099, "step": 309 }, { "epoch": 0.4197698036560596, "grad_norm": 1.1267645343143333, "learning_rate": 9.78937418691894e-06, "loss": 2.377, "step": 310 }, { "epoch": 0.4211238997968856, "grad_norm": 0.954265542677354, "learning_rate": 9.787822256124972e-06, "loss": 2.2119, "step": 311 }, { "epoch": 0.42247799593771157, "grad_norm": 1.0897928699498936, "learning_rate": 9.78626475276808e-06, "loss": 2.5392, "step": 312 }, { "epoch": 0.42383209207853756, "grad_norm": 0.8557954442168854, "learning_rate": 9.784701678661045e-06, "loss": 2.0243, "step": 313 }, { "epoch": 0.42518618821936355, "grad_norm": 1.007998713387866, "learning_rate": 9.783133035623136e-06, "loss": 2.0593, "step": 314 }, { "epoch": 0.4265402843601896, "grad_norm": 0.8245127331943697, "learning_rate": 9.781558825480104e-06, "loss": 1.9544, "step": 315 }, { "epoch": 0.4278943805010156, "grad_norm": 1.1343475642376186, "learning_rate": 9.779979050064174e-06, "loss": 2.1917, "step": 316 }, { "epoch": 0.42924847664184157, "grad_norm": 1.1336908348453985, "learning_rate": 9.778393711214054e-06, "loss": 2.3615, "step": 317 }, { "epoch": 0.43060257278266756, "grad_norm": 1.0074004622651764, "learning_rate": 9.776802810774924e-06, "loss": 2.4642, "step": 318 }, { "epoch": 0.43195666892349355, "grad_norm": 0.869836917179775, "learning_rate": 9.77520635059844e-06, "loss": 2.2773, "step": 319 }, { "epoch": 0.4333107650643196, "grad_norm": 0.8822131547317252, "learning_rate": 9.77360433254273e-06, "loss": 2.5491, "step": 320 }, { "epoch": 0.4346648612051456, "grad_norm": 0.9871053514468092, "learning_rate": 9.771996758472381e-06, "loss": 2.5086, "step": 321 }, { "epoch": 0.43601895734597157, "grad_norm": 0.8810949134669234, "learning_rate": 9.770383630258463e-06, "loss": 2.4492, "step": 322 }, { "epoch": 0.43737305348679756, "grad_norm": 0.9634447042117443, "learning_rate": 9.768764949778495e-06, "loss": 2.4698, "step": 323 }, { "epoch": 0.43872714962762355, "grad_norm": 0.9578841465826307, "learning_rate": 9.767140718916467e-06, "loss": 2.8468, "step": 324 }, { "epoch": 0.44008124576844954, "grad_norm": 1.045773660121023, "learning_rate": 9.765510939562827e-06, "loss": 2.4331, "step": 325 }, { "epoch": 0.4414353419092756, "grad_norm": 1.5236244191611785, "learning_rate": 9.763875613614482e-06, "loss": 2.4025, "step": 326 }, { "epoch": 0.44278943805010157, "grad_norm": 0.9193628457370401, "learning_rate": 9.762234742974793e-06, "loss": 2.2136, "step": 327 }, { "epoch": 0.44414353419092756, "grad_norm": 0.8949925268143644, "learning_rate": 9.76058832955357e-06, "loss": 2.4088, "step": 328 }, { "epoch": 0.44549763033175355, "grad_norm": 2.0556356286018413, "learning_rate": 9.758936375267087e-06, "loss": 2.1351, "step": 329 }, { "epoch": 0.44685172647257954, "grad_norm": 1.0444826822145854, "learning_rate": 9.757278882038056e-06, "loss": 2.7799, "step": 330 }, { "epoch": 0.4482058226134055, "grad_norm": 1.0824437177931403, "learning_rate": 9.755615851795639e-06, "loss": 1.8668, "step": 331 }, { "epoch": 0.44955991875423157, "grad_norm": 1.1111629995250478, "learning_rate": 9.753947286475442e-06, "loss": 2.754, "step": 332 }, { "epoch": 0.45091401489505756, "grad_norm": 0.9573737568401282, "learning_rate": 9.752273188019514e-06, "loss": 2.5329, "step": 333 }, { "epoch": 0.45226811103588355, "grad_norm": 1.0357238977124108, "learning_rate": 9.750593558376347e-06, "loss": 2.2007, "step": 334 }, { "epoch": 0.45362220717670954, "grad_norm": 0.9095487712099384, "learning_rate": 9.748908399500863e-06, "loss": 1.8825, "step": 335 }, { "epoch": 0.4549763033175355, "grad_norm": 1.7445295348550816, "learning_rate": 9.747217713354428e-06, "loss": 2.5305, "step": 336 }, { "epoch": 0.45633039945836157, "grad_norm": 0.9070558170854458, "learning_rate": 9.745521501904835e-06, "loss": 2.1951, "step": 337 }, { "epoch": 0.45768449559918756, "grad_norm": 0.8995516521874707, "learning_rate": 9.743819767126312e-06, "loss": 2.2137, "step": 338 }, { "epoch": 0.45903859174001355, "grad_norm": 1.9426471289696312, "learning_rate": 9.742112510999516e-06, "loss": 2.3716, "step": 339 }, { "epoch": 0.46039268788083954, "grad_norm": 0.9725576998378328, "learning_rate": 9.740399735511524e-06, "loss": 2.6405, "step": 340 }, { "epoch": 0.4617467840216655, "grad_norm": 0.8305802783581133, "learning_rate": 9.738681442655842e-06, "loss": 2.2231, "step": 341 }, { "epoch": 0.4631008801624915, "grad_norm": 0.9734526439933455, "learning_rate": 9.736957634432398e-06, "loss": 2.3041, "step": 342 }, { "epoch": 0.46445497630331756, "grad_norm": 0.824808147309629, "learning_rate": 9.73522831284754e-06, "loss": 2.3024, "step": 343 }, { "epoch": 0.46580907244414355, "grad_norm": 0.9599600802177057, "learning_rate": 9.733493479914031e-06, "loss": 2.3563, "step": 344 }, { "epoch": 0.46716316858496953, "grad_norm": 0.9340063834780592, "learning_rate": 9.731753137651047e-06, "loss": 2.1237, "step": 345 }, { "epoch": 0.4685172647257955, "grad_norm": 1.562533249030556, "learning_rate": 9.730007288084178e-06, "loss": 2.3575, "step": 346 }, { "epoch": 0.4698713608666215, "grad_norm": 0.9490993069469829, "learning_rate": 9.728255933245428e-06, "loss": 2.8822, "step": 347 }, { "epoch": 0.4712254570074475, "grad_norm": 0.9566724872274933, "learning_rate": 9.726499075173201e-06, "loss": 2.3402, "step": 348 }, { "epoch": 0.47257955314827355, "grad_norm": 1.008513385617003, "learning_rate": 9.724736715912313e-06, "loss": 2.1406, "step": 349 }, { "epoch": 0.47393364928909953, "grad_norm": 1.1766610712644368, "learning_rate": 9.72296885751398e-06, "loss": 2.1934, "step": 350 }, { "epoch": 0.4752877454299255, "grad_norm": 0.9325244223791722, "learning_rate": 9.721195502035817e-06, "loss": 1.9835, "step": 351 }, { "epoch": 0.4766418415707515, "grad_norm": 0.837427966240756, "learning_rate": 9.719416651541839e-06, "loss": 2.481, "step": 352 }, { "epoch": 0.4779959377115775, "grad_norm": 1.02007044942659, "learning_rate": 9.717632308102455e-06, "loss": 2.7148, "step": 353 }, { "epoch": 0.47935003385240355, "grad_norm": 0.959116878107154, "learning_rate": 9.715842473794472e-06, "loss": 2.4738, "step": 354 }, { "epoch": 0.48070412999322953, "grad_norm": 1.0102624905029922, "learning_rate": 9.714047150701082e-06, "loss": 2.0945, "step": 355 }, { "epoch": 0.4820582261340555, "grad_norm": 2.0984222479912438, "learning_rate": 9.712246340911866e-06, "loss": 2.1697, "step": 356 }, { "epoch": 0.4834123222748815, "grad_norm": 0.917396475775467, "learning_rate": 9.710440046522797e-06, "loss": 2.2475, "step": 357 }, { "epoch": 0.4847664184157075, "grad_norm": 1.005361407668732, "learning_rate": 9.708628269636224e-06, "loss": 2.2307, "step": 358 }, { "epoch": 0.4861205145565335, "grad_norm": 1.0309046462532379, "learning_rate": 9.706811012360882e-06, "loss": 2.4654, "step": 359 }, { "epoch": 0.48747461069735953, "grad_norm": 3.0659420771722834, "learning_rate": 9.704988276811883e-06, "loss": 2.7097, "step": 360 }, { "epoch": 0.4888287068381855, "grad_norm": 1.0772251854315056, "learning_rate": 9.703160065110716e-06, "loss": 2.3728, "step": 361 }, { "epoch": 0.4901828029790115, "grad_norm": 1.1656105776873809, "learning_rate": 9.701326379385238e-06, "loss": 2.6172, "step": 362 }, { "epoch": 0.4915368991198375, "grad_norm": 1.0293153712278362, "learning_rate": 9.699487221769687e-06, "loss": 2.427, "step": 363 }, { "epoch": 0.4928909952606635, "grad_norm": 1.1554858971025437, "learning_rate": 9.697642594404666e-06, "loss": 2.3854, "step": 364 }, { "epoch": 0.4942450914014895, "grad_norm": 1.1552799937404215, "learning_rate": 9.69579249943714e-06, "loss": 2.2642, "step": 365 }, { "epoch": 0.4955991875423155, "grad_norm": 1.3024770839893858, "learning_rate": 9.693936939020441e-06, "loss": 2.7555, "step": 366 }, { "epoch": 0.4969532836831415, "grad_norm": 0.9139562471712885, "learning_rate": 9.692075915314265e-06, "loss": 1.9849, "step": 367 }, { "epoch": 0.4983073798239675, "grad_norm": 1.1108979731718802, "learning_rate": 9.69020943048466e-06, "loss": 2.7987, "step": 368 }, { "epoch": 0.4996614759647935, "grad_norm": 1.043554360812148, "learning_rate": 9.688337486704038e-06, "loss": 2.379, "step": 369 }, { "epoch": 0.5010155721056195, "grad_norm": 0.8482819857067955, "learning_rate": 9.686460086151159e-06, "loss": 2.3031, "step": 370 }, { "epoch": 0.5023696682464455, "grad_norm": 0.882893031313843, "learning_rate": 9.684577231011134e-06, "loss": 1.9398, "step": 371 }, { "epoch": 0.5037237643872715, "grad_norm": 0.8900569461037698, "learning_rate": 9.68268892347543e-06, "loss": 2.3579, "step": 372 }, { "epoch": 0.5050778605280974, "grad_norm": 1.027440419577199, "learning_rate": 9.680795165741849e-06, "loss": 1.8953, "step": 373 }, { "epoch": 0.5064319566689235, "grad_norm": 1.2020524441224305, "learning_rate": 9.678895960014545e-06, "loss": 2.6735, "step": 374 }, { "epoch": 0.5077860528097495, "grad_norm": 1.6359343554936903, "learning_rate": 9.676991308504012e-06, "loss": 2.5313, "step": 375 }, { "epoch": 0.5091401489505755, "grad_norm": 0.9318914822552815, "learning_rate": 9.675081213427076e-06, "loss": 2.7319, "step": 376 }, { "epoch": 0.5104942450914015, "grad_norm": 1.0629414040228036, "learning_rate": 9.673165677006906e-06, "loss": 2.5041, "step": 377 }, { "epoch": 0.5118483412322274, "grad_norm": 0.9015543787336042, "learning_rate": 9.671244701472999e-06, "loss": 2.2052, "step": 378 }, { "epoch": 0.5132024373730535, "grad_norm": 0.9603399577043747, "learning_rate": 9.669318289061191e-06, "loss": 2.5069, "step": 379 }, { "epoch": 0.5145565335138795, "grad_norm": 0.8421778244290412, "learning_rate": 9.667386442013634e-06, "loss": 2.6313, "step": 380 }, { "epoch": 0.5159106296547055, "grad_norm": 0.8356138567382542, "learning_rate": 9.665449162578814e-06, "loss": 2.0145, "step": 381 }, { "epoch": 0.5172647257955315, "grad_norm": 0.9669755721462949, "learning_rate": 9.663506453011538e-06, "loss": 2.2227, "step": 382 }, { "epoch": 0.5186188219363574, "grad_norm": 1.163959225455655, "learning_rate": 9.66155831557293e-06, "loss": 2.5775, "step": 383 }, { "epoch": 0.5199729180771835, "grad_norm": 0.9362192144637581, "learning_rate": 9.659604752530434e-06, "loss": 2.2014, "step": 384 }, { "epoch": 0.5213270142180095, "grad_norm": 0.9892445662855336, "learning_rate": 9.657645766157813e-06, "loss": 2.4195, "step": 385 }, { "epoch": 0.5226811103588355, "grad_norm": 0.8722826858897498, "learning_rate": 9.655681358735134e-06, "loss": 2.6966, "step": 386 }, { "epoch": 0.5240352064996615, "grad_norm": 1.0196349369430946, "learning_rate": 9.653711532548778e-06, "loss": 2.4237, "step": 387 }, { "epoch": 0.5253893026404874, "grad_norm": 1.0011773469402157, "learning_rate": 9.651736289891434e-06, "loss": 2.3765, "step": 388 }, { "epoch": 0.5267433987813135, "grad_norm": 0.9843650150083161, "learning_rate": 9.649755633062092e-06, "loss": 2.4698, "step": 389 }, { "epoch": 0.5280974949221394, "grad_norm": 1.2199752234970775, "learning_rate": 9.647769564366048e-06, "loss": 2.2025, "step": 390 }, { "epoch": 0.5294515910629655, "grad_norm": 1.048526738243433, "learning_rate": 9.645778086114892e-06, "loss": 2.5315, "step": 391 }, { "epoch": 0.5308056872037915, "grad_norm": 1.0079058862169128, "learning_rate": 9.643781200626512e-06, "loss": 2.2688, "step": 392 }, { "epoch": 0.5321597833446174, "grad_norm": 0.9175841000628026, "learning_rate": 9.641778910225093e-06, "loss": 2.4924, "step": 393 }, { "epoch": 0.5335138794854435, "grad_norm": 0.9394126148321758, "learning_rate": 9.639771217241104e-06, "loss": 2.6604, "step": 394 }, { "epoch": 0.5348679756262694, "grad_norm": 1.8536643574976448, "learning_rate": 9.637758124011307e-06, "loss": 2.3891, "step": 395 }, { "epoch": 0.5362220717670955, "grad_norm": 1.353118661391877, "learning_rate": 9.63573963287875e-06, "loss": 2.2486, "step": 396 }, { "epoch": 0.5375761679079215, "grad_norm": 0.9114437873284958, "learning_rate": 9.633715746192762e-06, "loss": 2.9009, "step": 397 }, { "epoch": 0.5389302640487474, "grad_norm": 0.9178272536878976, "learning_rate": 9.631686466308947e-06, "loss": 2.6545, "step": 398 }, { "epoch": 0.5402843601895735, "grad_norm": 0.8349824532422451, "learning_rate": 9.629651795589197e-06, "loss": 2.3238, "step": 399 }, { "epoch": 0.5416384563303994, "grad_norm": 0.8002054931156397, "learning_rate": 9.627611736401668e-06, "loss": 2.1503, "step": 400 }, { "epoch": 0.5429925524712255, "grad_norm": 1.014043339956385, "learning_rate": 9.625566291120794e-06, "loss": 2.3366, "step": 401 }, { "epoch": 0.5443466486120515, "grad_norm": 0.9753524140226457, "learning_rate": 9.623515462127276e-06, "loss": 2.4777, "step": 402 }, { "epoch": 0.5457007447528774, "grad_norm": 0.9904758899407581, "learning_rate": 9.621459251808078e-06, "loss": 2.5355, "step": 403 }, { "epoch": 0.5470548408937035, "grad_norm": 0.9176434471710123, "learning_rate": 9.619397662556434e-06, "loss": 2.1775, "step": 404 }, { "epoch": 0.5484089370345294, "grad_norm": 0.9994849139704428, "learning_rate": 9.617330696771834e-06, "loss": 2.3618, "step": 405 }, { "epoch": 0.5497630331753555, "grad_norm": 0.9718579619606619, "learning_rate": 9.615258356860027e-06, "loss": 2.065, "step": 406 }, { "epoch": 0.5511171293161814, "grad_norm": 0.8890137365695405, "learning_rate": 9.613180645233014e-06, "loss": 2.153, "step": 407 }, { "epoch": 0.5524712254570074, "grad_norm": 1.001837019128747, "learning_rate": 9.611097564309054e-06, "loss": 2.4168, "step": 408 }, { "epoch": 0.5538253215978335, "grad_norm": 1.205118639472601, "learning_rate": 9.609009116512648e-06, "loss": 2.5229, "step": 409 }, { "epoch": 0.5551794177386594, "grad_norm": 0.9757578479246173, "learning_rate": 9.60691530427455e-06, "loss": 2.1174, "step": 410 }, { "epoch": 0.5565335138794855, "grad_norm": 0.9985946290929901, "learning_rate": 9.60481613003176e-06, "loss": 1.9395, "step": 411 }, { "epoch": 0.5578876100203114, "grad_norm": 0.9058862928790223, "learning_rate": 9.602711596227507e-06, "loss": 2.1988, "step": 412 }, { "epoch": 0.5592417061611374, "grad_norm": 0.9620888369560685, "learning_rate": 9.600601705311267e-06, "loss": 2.3826, "step": 413 }, { "epoch": 0.5605958023019635, "grad_norm": 1.0478032599035778, "learning_rate": 9.598486459738751e-06, "loss": 2.4528, "step": 414 }, { "epoch": 0.5619498984427894, "grad_norm": 0.8838781940692847, "learning_rate": 9.5963658619719e-06, "loss": 2.2238, "step": 415 }, { "epoch": 0.5633039945836155, "grad_norm": 1.044653451464992, "learning_rate": 9.594239914478886e-06, "loss": 2.8379, "step": 416 }, { "epoch": 0.5646580907244414, "grad_norm": 0.9143098376517602, "learning_rate": 9.592108619734107e-06, "loss": 2.3621, "step": 417 }, { "epoch": 0.5660121868652674, "grad_norm": 1.3056702719267006, "learning_rate": 9.58997198021818e-06, "loss": 2.3495, "step": 418 }, { "epoch": 0.5673662830060935, "grad_norm": 1.2114169289052528, "learning_rate": 9.587829998417953e-06, "loss": 2.6471, "step": 419 }, { "epoch": 0.5687203791469194, "grad_norm": 1.046324944660913, "learning_rate": 9.58568267682648e-06, "loss": 2.402, "step": 420 }, { "epoch": 0.5700744752877455, "grad_norm": 0.9349583810327691, "learning_rate": 9.58353001794304e-06, "loss": 2.2225, "step": 421 }, { "epoch": 0.5714285714285714, "grad_norm": 1.481772765240176, "learning_rate": 9.581372024273121e-06, "loss": 1.915, "step": 422 }, { "epoch": 0.5727826675693974, "grad_norm": 0.928476736332317, "learning_rate": 9.579208698328419e-06, "loss": 1.8008, "step": 423 }, { "epoch": 0.5741367637102234, "grad_norm": 0.9260270264242221, "learning_rate": 9.577040042626832e-06, "loss": 2.417, "step": 424 }, { "epoch": 0.5754908598510494, "grad_norm": 1.1064706208462298, "learning_rate": 9.574866059692471e-06, "loss": 2.5265, "step": 425 }, { "epoch": 0.5768449559918755, "grad_norm": 2.139248004116943, "learning_rate": 9.57268675205564e-06, "loss": 2.1862, "step": 426 }, { "epoch": 0.5781990521327014, "grad_norm": 1.2117707663910726, "learning_rate": 9.570502122252844e-06, "loss": 2.8615, "step": 427 }, { "epoch": 0.5795531482735274, "grad_norm": 0.8678663266783263, "learning_rate": 9.568312172826779e-06, "loss": 2.2196, "step": 428 }, { "epoch": 0.5809072444143534, "grad_norm": 1.183740118318461, "learning_rate": 9.566116906326336e-06, "loss": 2.7205, "step": 429 }, { "epoch": 0.5822613405551794, "grad_norm": 0.9774576381702014, "learning_rate": 9.563916325306595e-06, "loss": 2.2203, "step": 430 }, { "epoch": 0.5836154366960055, "grad_norm": 0.9824933411377383, "learning_rate": 9.561710432328817e-06, "loss": 2.7149, "step": 431 }, { "epoch": 0.5849695328368314, "grad_norm": 1.8285370131236038, "learning_rate": 9.55949922996045e-06, "loss": 2.1549, "step": 432 }, { "epoch": 0.5863236289776574, "grad_norm": 0.92698096294728, "learning_rate": 9.55728272077512e-06, "loss": 2.0743, "step": 433 }, { "epoch": 0.5876777251184834, "grad_norm": 1.0344108156736416, "learning_rate": 9.555060907352632e-06, "loss": 2.3762, "step": 434 }, { "epoch": 0.5890318212593094, "grad_norm": 1.0045554696436734, "learning_rate": 9.552833792278957e-06, "loss": 2.1954, "step": 435 }, { "epoch": 0.5903859174001355, "grad_norm": 0.7967159103822394, "learning_rate": 9.550601378146246e-06, "loss": 2.0138, "step": 436 }, { "epoch": 0.5917400135409614, "grad_norm": 0.8111896941287193, "learning_rate": 9.54836366755281e-06, "loss": 2.3701, "step": 437 }, { "epoch": 0.5930941096817874, "grad_norm": 0.9726524616144652, "learning_rate": 9.546120663103134e-06, "loss": 1.9217, "step": 438 }, { "epoch": 0.5944482058226134, "grad_norm": 1.0256397319731523, "learning_rate": 9.543872367407854e-06, "loss": 2.0282, "step": 439 }, { "epoch": 0.5958023019634394, "grad_norm": 1.0079882867447063, "learning_rate": 9.54161878308377e-06, "loss": 2.3361, "step": 440 }, { "epoch": 0.5971563981042654, "grad_norm": 0.9800767836635148, "learning_rate": 9.539359912753839e-06, "loss": 2.6659, "step": 441 }, { "epoch": 0.5985104942450914, "grad_norm": 1.0649583722690574, "learning_rate": 9.537095759047163e-06, "loss": 2.4143, "step": 442 }, { "epoch": 0.5998645903859174, "grad_norm": 1.174278203502222, "learning_rate": 9.534826324599002e-06, "loss": 2.22, "step": 443 }, { "epoch": 0.6012186865267434, "grad_norm": 1.3010142344912385, "learning_rate": 9.53255161205076e-06, "loss": 2.1634, "step": 444 }, { "epoch": 0.6025727826675694, "grad_norm": 0.8409212163726476, "learning_rate": 9.530271624049979e-06, "loss": 1.8782, "step": 445 }, { "epoch": 0.6039268788083954, "grad_norm": 0.8027451978851314, "learning_rate": 9.527986363250348e-06, "loss": 2.203, "step": 446 }, { "epoch": 0.6052809749492214, "grad_norm": 0.9039555297703499, "learning_rate": 9.525695832311688e-06, "loss": 2.3687, "step": 447 }, { "epoch": 0.6066350710900474, "grad_norm": 0.8912812069256285, "learning_rate": 9.523400033899957e-06, "loss": 2.2258, "step": 448 }, { "epoch": 0.6079891672308734, "grad_norm": 0.8488642026117442, "learning_rate": 9.52109897068724e-06, "loss": 2.4719, "step": 449 }, { "epoch": 0.6093432633716994, "grad_norm": 0.9189612986590668, "learning_rate": 9.518792645351757e-06, "loss": 2.1323, "step": 450 }, { "epoch": 0.6106973595125254, "grad_norm": 1.0241173909729102, "learning_rate": 9.516481060577847e-06, "loss": 2.3125, "step": 451 }, { "epoch": 0.6120514556533514, "grad_norm": 0.9700124960446453, "learning_rate": 9.514164219055969e-06, "loss": 2.1262, "step": 452 }, { "epoch": 0.6134055517941774, "grad_norm": 0.9865209216124129, "learning_rate": 9.511842123482703e-06, "loss": 2.0346, "step": 453 }, { "epoch": 0.6147596479350034, "grad_norm": 0.8339628050854324, "learning_rate": 9.509514776560747e-06, "loss": 2.2793, "step": 454 }, { "epoch": 0.6161137440758294, "grad_norm": 1.2124630572926367, "learning_rate": 9.507182180998906e-06, "loss": 2.1059, "step": 455 }, { "epoch": 0.6174678402166554, "grad_norm": 1.1523483755668846, "learning_rate": 9.504844339512096e-06, "loss": 2.1094, "step": 456 }, { "epoch": 0.6188219363574814, "grad_norm": 0.9250806368586671, "learning_rate": 9.50250125482134e-06, "loss": 2.6985, "step": 457 }, { "epoch": 0.6201760324983073, "grad_norm": 1.2028362769677976, "learning_rate": 9.500152929653764e-06, "loss": 2.3153, "step": 458 }, { "epoch": 0.6215301286391334, "grad_norm": 1.071969908607831, "learning_rate": 9.497799366742586e-06, "loss": 2.0759, "step": 459 }, { "epoch": 0.6228842247799594, "grad_norm": 1.3535021074481768, "learning_rate": 9.49544056882713e-06, "loss": 2.2538, "step": 460 }, { "epoch": 0.6242383209207854, "grad_norm": 1.1103088277032604, "learning_rate": 9.49307653865281e-06, "loss": 2.0217, "step": 461 }, { "epoch": 0.6255924170616114, "grad_norm": 0.9163278193375946, "learning_rate": 9.490707278971127e-06, "loss": 2.2543, "step": 462 }, { "epoch": 0.6269465132024373, "grad_norm": 0.953261634167317, "learning_rate": 9.488332792539672e-06, "loss": 2.6539, "step": 463 }, { "epoch": 0.6283006093432634, "grad_norm": 1.029379230332492, "learning_rate": 9.485953082122116e-06, "loss": 2.3507, "step": 464 }, { "epoch": 0.6296547054840894, "grad_norm": 0.9618296595544104, "learning_rate": 9.483568150488215e-06, "loss": 2.4448, "step": 465 }, { "epoch": 0.6310088016249153, "grad_norm": 0.9594228312002971, "learning_rate": 9.481178000413796e-06, "loss": 2.6108, "step": 466 }, { "epoch": 0.6323628977657414, "grad_norm": 0.9624426030456921, "learning_rate": 9.478782634680765e-06, "loss": 2.3803, "step": 467 }, { "epoch": 0.6337169939065673, "grad_norm": 0.9808684308680445, "learning_rate": 9.476382056077097e-06, "loss": 2.5205, "step": 468 }, { "epoch": 0.6350710900473934, "grad_norm": 1.0917634456153018, "learning_rate": 9.473976267396831e-06, "loss": 2.008, "step": 469 }, { "epoch": 0.6364251861882194, "grad_norm": 1.85759777161883, "learning_rate": 9.471565271440075e-06, "loss": 1.9461, "step": 470 }, { "epoch": 0.6377792823290453, "grad_norm": 0.9263896064370456, "learning_rate": 9.469149071012996e-06, "loss": 2.1459, "step": 471 }, { "epoch": 0.6391333784698714, "grad_norm": 1.2831155066680264, "learning_rate": 9.466727668927817e-06, "loss": 2.5968, "step": 472 }, { "epoch": 0.6404874746106973, "grad_norm": 2.5926011027880405, "learning_rate": 9.464301068002815e-06, "loss": 2.8532, "step": 473 }, { "epoch": 0.6418415707515234, "grad_norm": 1.3517479249404691, "learning_rate": 9.461869271062322e-06, "loss": 2.5085, "step": 474 }, { "epoch": 0.6431956668923493, "grad_norm": 1.0653079693612253, "learning_rate": 9.459432280936714e-06, "loss": 1.926, "step": 475 }, { "epoch": 0.6445497630331753, "grad_norm": 1.0222358313347504, "learning_rate": 9.456990100462411e-06, "loss": 2.0835, "step": 476 }, { "epoch": 0.6459038591740014, "grad_norm": 0.9472201990170999, "learning_rate": 9.454542732481876e-06, "loss": 2.5298, "step": 477 }, { "epoch": 0.6472579553148273, "grad_norm": 1.05236931626042, "learning_rate": 9.452090179843609e-06, "loss": 2.5688, "step": 478 }, { "epoch": 0.6486120514556534, "grad_norm": 0.9205333314410238, "learning_rate": 9.449632445402146e-06, "loss": 2.2376, "step": 479 }, { "epoch": 0.6499661475964793, "grad_norm": 0.936047746739389, "learning_rate": 9.44716953201805e-06, "loss": 2.2657, "step": 480 }, { "epoch": 0.6513202437373053, "grad_norm": 0.984012799628733, "learning_rate": 9.444701442557917e-06, "loss": 2.2747, "step": 481 }, { "epoch": 0.6526743398781314, "grad_norm": 1.1605171743997218, "learning_rate": 9.442228179894362e-06, "loss": 2.5819, "step": 482 }, { "epoch": 0.6540284360189573, "grad_norm": 1.0367078942789454, "learning_rate": 9.439749746906027e-06, "loss": 2.3326, "step": 483 }, { "epoch": 0.6553825321597834, "grad_norm": 0.8511526169640604, "learning_rate": 9.437266146477567e-06, "loss": 2.3916, "step": 484 }, { "epoch": 0.6567366283006093, "grad_norm": 0.9546785433848596, "learning_rate": 9.434777381499654e-06, "loss": 2.2932, "step": 485 }, { "epoch": 0.6580907244414353, "grad_norm": 0.897715373597665, "learning_rate": 9.43228345486897e-06, "loss": 2.2329, "step": 486 }, { "epoch": 0.6594448205822614, "grad_norm": 1.1576802918228548, "learning_rate": 9.429784369488205e-06, "loss": 2.5906, "step": 487 }, { "epoch": 0.6607989167230873, "grad_norm": 1.2669978365090186, "learning_rate": 9.427280128266049e-06, "loss": 2.0969, "step": 488 }, { "epoch": 0.6621530128639134, "grad_norm": 1.116706571792338, "learning_rate": 9.424770734117206e-06, "loss": 2.0978, "step": 489 }, { "epoch": 0.6635071090047393, "grad_norm": 1.0731484382258585, "learning_rate": 9.42225618996236e-06, "loss": 2.2606, "step": 490 }, { "epoch": 0.6648612051455653, "grad_norm": 1.0217610153976007, "learning_rate": 9.419736498728203e-06, "loss": 2.0359, "step": 491 }, { "epoch": 0.6662153012863913, "grad_norm": 0.9470105904564678, "learning_rate": 9.417211663347407e-06, "loss": 2.3133, "step": 492 }, { "epoch": 0.6675693974272173, "grad_norm": 1.1496175064197292, "learning_rate": 9.414681686758645e-06, "loss": 2.6914, "step": 493 }, { "epoch": 0.6689234935680434, "grad_norm": 1.1088711952773571, "learning_rate": 9.412146571906556e-06, "loss": 2.1629, "step": 494 }, { "epoch": 0.6702775897088693, "grad_norm": 2.4048301880979315, "learning_rate": 9.409606321741776e-06, "loss": 2.3511, "step": 495 }, { "epoch": 0.6716316858496953, "grad_norm": 1.0855961625720172, "learning_rate": 9.407060939220907e-06, "loss": 3.0198, "step": 496 }, { "epoch": 0.6729857819905213, "grad_norm": 0.831939027060162, "learning_rate": 9.404510427306533e-06, "loss": 2.0721, "step": 497 }, { "epoch": 0.6743398781313473, "grad_norm": 0.9189030221620978, "learning_rate": 9.401954788967199e-06, "loss": 2.1533, "step": 498 }, { "epoch": 0.6756939742721734, "grad_norm": 1.1215682593702503, "learning_rate": 9.39939402717742e-06, "loss": 2.3017, "step": 499 }, { "epoch": 0.6770480704129993, "grad_norm": 0.9997121388139094, "learning_rate": 9.396828144917682e-06, "loss": 2.1092, "step": 500 }, { "epoch": 0.6784021665538253, "grad_norm": 1.0419306713261767, "learning_rate": 9.39425714517442e-06, "loss": 2.3772, "step": 501 }, { "epoch": 0.6797562626946513, "grad_norm": 0.8546730463318373, "learning_rate": 9.391681030940031e-06, "loss": 2.4692, "step": 502 }, { "epoch": 0.6811103588354773, "grad_norm": 1.036666833583794, "learning_rate": 9.389099805212862e-06, "loss": 2.4967, "step": 503 }, { "epoch": 0.6824644549763034, "grad_norm": 0.9028903653709767, "learning_rate": 9.38651347099721e-06, "loss": 2.2658, "step": 504 }, { "epoch": 0.6838185511171293, "grad_norm": 1.004993430273087, "learning_rate": 9.38392203130332e-06, "loss": 2.2413, "step": 505 }, { "epoch": 0.6851726472579553, "grad_norm": 4.244116799684234, "learning_rate": 9.38132548914738e-06, "loss": 2.2353, "step": 506 }, { "epoch": 0.6865267433987813, "grad_norm": 2.0623005376684516, "learning_rate": 9.37872384755151e-06, "loss": 2.3859, "step": 507 }, { "epoch": 0.6878808395396073, "grad_norm": 1.099884816816885, "learning_rate": 9.376117109543769e-06, "loss": 1.8825, "step": 508 }, { "epoch": 0.6892349356804333, "grad_norm": 0.9736281444679528, "learning_rate": 9.373505278158152e-06, "loss": 2.4243, "step": 509 }, { "epoch": 0.6905890318212593, "grad_norm": 0.9265991736695152, "learning_rate": 9.370888356434577e-06, "loss": 2.5581, "step": 510 }, { "epoch": 0.6919431279620853, "grad_norm": 0.9742740668751364, "learning_rate": 9.368266347418891e-06, "loss": 2.2436, "step": 511 }, { "epoch": 0.6932972241029113, "grad_norm": 0.903250624410521, "learning_rate": 9.365639254162855e-06, "loss": 2.3704, "step": 512 }, { "epoch": 0.6946513202437373, "grad_norm": 1.011884451867896, "learning_rate": 9.363007079724153e-06, "loss": 2.3554, "step": 513 }, { "epoch": 0.6960054163845633, "grad_norm": 2.1268078439425504, "learning_rate": 9.360369827166385e-06, "loss": 2.2531, "step": 514 }, { "epoch": 0.6973595125253893, "grad_norm": 0.9226076560386388, "learning_rate": 9.357727499559055e-06, "loss": 2.1536, "step": 515 }, { "epoch": 0.6987136086662153, "grad_norm": 0.9235717043788523, "learning_rate": 9.355080099977579e-06, "loss": 2.5795, "step": 516 }, { "epoch": 0.7000677048070413, "grad_norm": 0.997624756104421, "learning_rate": 9.352427631503274e-06, "loss": 2.4086, "step": 517 }, { "epoch": 0.7014218009478673, "grad_norm": 1.0325935042936567, "learning_rate": 9.349770097223356e-06, "loss": 2.1905, "step": 518 }, { "epoch": 0.7027758970886933, "grad_norm": 0.9779488073579523, "learning_rate": 9.347107500230941e-06, "loss": 2.1902, "step": 519 }, { "epoch": 0.7041299932295193, "grad_norm": 0.9071035111230149, "learning_rate": 9.344439843625034e-06, "loss": 2.4966, "step": 520 }, { "epoch": 0.7054840893703453, "grad_norm": 0.9294082146423078, "learning_rate": 9.341767130510529e-06, "loss": 2.1556, "step": 521 }, { "epoch": 0.7068381855111713, "grad_norm": 0.8010254757576656, "learning_rate": 9.339089363998206e-06, "loss": 2.2359, "step": 522 }, { "epoch": 0.7081922816519973, "grad_norm": 1.2492381753400523, "learning_rate": 9.336406547204726e-06, "loss": 2.0768, "step": 523 }, { "epoch": 0.7095463777928233, "grad_norm": 0.960396750578361, "learning_rate": 9.333718683252631e-06, "loss": 2.2373, "step": 524 }, { "epoch": 0.7109004739336493, "grad_norm": 0.9089783184513304, "learning_rate": 9.331025775270335e-06, "loss": 2.0008, "step": 525 }, { "epoch": 0.7122545700744752, "grad_norm": 1.0902851866790608, "learning_rate": 9.32832782639212e-06, "loss": 2.452, "step": 526 }, { "epoch": 0.7136086662153013, "grad_norm": 1.0819110431428438, "learning_rate": 9.325624839758142e-06, "loss": 2.1797, "step": 527 }, { "epoch": 0.7149627623561273, "grad_norm": 1.035728859841243, "learning_rate": 9.322916818514414e-06, "loss": 2.0016, "step": 528 }, { "epoch": 0.7163168584969533, "grad_norm": 0.8429686871803489, "learning_rate": 9.32020376581281e-06, "loss": 2.1271, "step": 529 }, { "epoch": 0.7176709546377793, "grad_norm": 0.948124830150651, "learning_rate": 9.317485684811065e-06, "loss": 2.463, "step": 530 }, { "epoch": 0.7190250507786052, "grad_norm": 1.0966777201484916, "learning_rate": 9.31476257867276e-06, "loss": 2.7078, "step": 531 }, { "epoch": 0.7203791469194313, "grad_norm": 0.9601018625399497, "learning_rate": 9.312034450567331e-06, "loss": 2.0091, "step": 532 }, { "epoch": 0.7217332430602573, "grad_norm": 1.0195440951721928, "learning_rate": 9.309301303670053e-06, "loss": 2.2595, "step": 533 }, { "epoch": 0.7230873392010833, "grad_norm": 1.0365088177530009, "learning_rate": 9.306563141162046e-06, "loss": 2.5509, "step": 534 }, { "epoch": 0.7244414353419093, "grad_norm": 0.89557097021405, "learning_rate": 9.303819966230265e-06, "loss": 2.184, "step": 535 }, { "epoch": 0.7257955314827352, "grad_norm": 1.1201396043943863, "learning_rate": 9.301071782067504e-06, "loss": 2.2719, "step": 536 }, { "epoch": 0.7271496276235613, "grad_norm": 1.0416986531898793, "learning_rate": 9.298318591872381e-06, "loss": 2.9307, "step": 537 }, { "epoch": 0.7285037237643873, "grad_norm": 0.8874369002318888, "learning_rate": 9.295560398849348e-06, "loss": 2.0185, "step": 538 }, { "epoch": 0.7298578199052133, "grad_norm": 1.3062496312023244, "learning_rate": 9.29279720620867e-06, "loss": 2.4988, "step": 539 }, { "epoch": 0.7312119160460393, "grad_norm": 0.8723686506256735, "learning_rate": 9.290029017166439e-06, "loss": 2.1358, "step": 540 }, { "epoch": 0.7325660121868652, "grad_norm": 0.9558509794048232, "learning_rate": 9.287255834944563e-06, "loss": 2.5911, "step": 541 }, { "epoch": 0.7339201083276913, "grad_norm": 0.9765924363071334, "learning_rate": 9.284477662770753e-06, "loss": 2.2083, "step": 542 }, { "epoch": 0.7352742044685172, "grad_norm": 0.9439430068504877, "learning_rate": 9.281694503878536e-06, "loss": 2.2064, "step": 543 }, { "epoch": 0.7366283006093433, "grad_norm": 1.1103666583629448, "learning_rate": 9.278906361507238e-06, "loss": 2.2118, "step": 544 }, { "epoch": 0.7379823967501693, "grad_norm": 1.015972947443489, "learning_rate": 9.276113238901992e-06, "loss": 2.3865, "step": 545 }, { "epoch": 0.7393364928909952, "grad_norm": 0.9787500482208973, "learning_rate": 9.273315139313719e-06, "loss": 2.0949, "step": 546 }, { "epoch": 0.7406905890318213, "grad_norm": 1.1247301478380771, "learning_rate": 9.270512065999139e-06, "loss": 1.9341, "step": 547 }, { "epoch": 0.7420446851726472, "grad_norm": 0.8630459962648358, "learning_rate": 9.267704022220758e-06, "loss": 2.1934, "step": 548 }, { "epoch": 0.7433987813134733, "grad_norm": 0.9456846452275267, "learning_rate": 9.264891011246867e-06, "loss": 2.1091, "step": 549 }, { "epoch": 0.7447528774542993, "grad_norm": 0.9901864770446673, "learning_rate": 9.26207303635154e-06, "loss": 2.0778, "step": 550 }, { "epoch": 0.7461069735951252, "grad_norm": 0.9659375715716257, "learning_rate": 9.25925010081463e-06, "loss": 2.2679, "step": 551 }, { "epoch": 0.7474610697359513, "grad_norm": 0.9764027265044928, "learning_rate": 9.256422207921757e-06, "loss": 2.7468, "step": 552 }, { "epoch": 0.7488151658767772, "grad_norm": 0.899162453540954, "learning_rate": 9.25358936096432e-06, "loss": 2.312, "step": 553 }, { "epoch": 0.7501692620176033, "grad_norm": 0.7884101704032132, "learning_rate": 9.250751563239473e-06, "loss": 1.8254, "step": 554 }, { "epoch": 0.7515233581584293, "grad_norm": 1.0492022449068705, "learning_rate": 9.247908818050146e-06, "loss": 2.6885, "step": 555 }, { "epoch": 0.7528774542992552, "grad_norm": 1.0133767522172044, "learning_rate": 9.245061128705017e-06, "loss": 2.2732, "step": 556 }, { "epoch": 0.7542315504400813, "grad_norm": 1.1740684334910383, "learning_rate": 9.242208498518522e-06, "loss": 2.4037, "step": 557 }, { "epoch": 0.7555856465809072, "grad_norm": 1.0034773672404191, "learning_rate": 9.239350930810843e-06, "loss": 2.6555, "step": 558 }, { "epoch": 0.7569397427217333, "grad_norm": 1.0559371991027349, "learning_rate": 9.236488428907919e-06, "loss": 1.8506, "step": 559 }, { "epoch": 0.7582938388625592, "grad_norm": 0.9422190702574703, "learning_rate": 9.233620996141421e-06, "loss": 2.393, "step": 560 }, { "epoch": 0.7596479350033852, "grad_norm": 1.701435588173234, "learning_rate": 9.230748635848768e-06, "loss": 2.4996, "step": 561 }, { "epoch": 0.7610020311442113, "grad_norm": 0.9848241894053293, "learning_rate": 9.227871351373108e-06, "loss": 2.2563, "step": 562 }, { "epoch": 0.7623561272850372, "grad_norm": 0.9551626201609839, "learning_rate": 9.224989146063322e-06, "loss": 2.3359, "step": 563 }, { "epoch": 0.7637102234258633, "grad_norm": 1.0501196818584786, "learning_rate": 9.22210202327402e-06, "loss": 2.3484, "step": 564 }, { "epoch": 0.7650643195666892, "grad_norm": 1.0118693595069355, "learning_rate": 9.219209986365533e-06, "loss": 2.0198, "step": 565 }, { "epoch": 0.7664184157075152, "grad_norm": 1.000596790716306, "learning_rate": 9.216313038703914e-06, "loss": 2.4147, "step": 566 }, { "epoch": 0.7677725118483413, "grad_norm": 1.0925850682201712, "learning_rate": 9.21341118366093e-06, "loss": 2.0607, "step": 567 }, { "epoch": 0.7691266079891672, "grad_norm": 0.8195614498636685, "learning_rate": 9.21050442461406e-06, "loss": 2.3161, "step": 568 }, { "epoch": 0.7704807041299933, "grad_norm": 1.4095142340461897, "learning_rate": 9.207592764946491e-06, "loss": 2.3281, "step": 569 }, { "epoch": 0.7718348002708192, "grad_norm": 1.0305488353359658, "learning_rate": 9.204676208047112e-06, "loss": 2.209, "step": 570 }, { "epoch": 0.7731888964116452, "grad_norm": 0.9457457073106245, "learning_rate": 9.201754757310518e-06, "loss": 1.9644, "step": 571 }, { "epoch": 0.7745429925524713, "grad_norm": 0.8798589503356994, "learning_rate": 9.198828416136991e-06, "loss": 2.5588, "step": 572 }, { "epoch": 0.7758970886932972, "grad_norm": 1.1021679000741063, "learning_rate": 9.195897187932513e-06, "loss": 2.2556, "step": 573 }, { "epoch": 0.7772511848341233, "grad_norm": 1.2881391770879362, "learning_rate": 9.192961076108748e-06, "loss": 2.2786, "step": 574 }, { "epoch": 0.7786052809749492, "grad_norm": 0.8079890380239451, "learning_rate": 9.190020084083048e-06, "loss": 1.9979, "step": 575 }, { "epoch": 0.7799593771157752, "grad_norm": 2.3299606691457706, "learning_rate": 9.187074215278444e-06, "loss": 2.3944, "step": 576 }, { "epoch": 0.7813134732566012, "grad_norm": 0.8875887654620506, "learning_rate": 9.184123473123643e-06, "loss": 3.159, "step": 577 }, { "epoch": 0.7826675693974272, "grad_norm": 1.2732756612102565, "learning_rate": 9.181167861053024e-06, "loss": 2.5586, "step": 578 }, { "epoch": 0.7840216655382533, "grad_norm": 1.1186141048074447, "learning_rate": 9.178207382506634e-06, "loss": 2.3656, "step": 579 }, { "epoch": 0.7853757616790792, "grad_norm": 1.0038802401880105, "learning_rate": 9.175242040930185e-06, "loss": 2.3621, "step": 580 }, { "epoch": 0.7867298578199052, "grad_norm": 1.0084194487112716, "learning_rate": 9.172271839775046e-06, "loss": 2.2325, "step": 581 }, { "epoch": 0.7880839539607312, "grad_norm": 0.8495420286871743, "learning_rate": 9.169296782498249e-06, "loss": 2.4249, "step": 582 }, { "epoch": 0.7894380501015572, "grad_norm": 1.0108960715157491, "learning_rate": 9.166316872562467e-06, "loss": 2.0571, "step": 583 }, { "epoch": 0.7907921462423833, "grad_norm": 0.7599417123654713, "learning_rate": 9.163332113436031e-06, "loss": 2.2401, "step": 584 }, { "epoch": 0.7921462423832092, "grad_norm": 0.9153720931911937, "learning_rate": 9.160342508592916e-06, "loss": 2.0296, "step": 585 }, { "epoch": 0.7935003385240352, "grad_norm": 1.0606631317988464, "learning_rate": 9.157348061512728e-06, "loss": 1.995, "step": 586 }, { "epoch": 0.7948544346648612, "grad_norm": 0.9696771933921041, "learning_rate": 9.154348775680714e-06, "loss": 2.4553, "step": 587 }, { "epoch": 0.7962085308056872, "grad_norm": 1.1359820431460632, "learning_rate": 9.151344654587758e-06, "loss": 2.6147, "step": 588 }, { "epoch": 0.7975626269465133, "grad_norm": 1.1620764175753133, "learning_rate": 9.148335701730363e-06, "loss": 2.1694, "step": 589 }, { "epoch": 0.7989167230873392, "grad_norm": 1.0906791903175577, "learning_rate": 9.145321920610662e-06, "loss": 2.0159, "step": 590 }, { "epoch": 0.8002708192281652, "grad_norm": 0.9262503779509312, "learning_rate": 9.142303314736405e-06, "loss": 1.8704, "step": 591 }, { "epoch": 0.8016249153689912, "grad_norm": 0.8897378574716819, "learning_rate": 9.139279887620955e-06, "loss": 2.2212, "step": 592 }, { "epoch": 0.8029790115098172, "grad_norm": 0.8785520958211444, "learning_rate": 9.136251642783294e-06, "loss": 2.2422, "step": 593 }, { "epoch": 0.8043331076506431, "grad_norm": 1.4708082564367666, "learning_rate": 9.133218583748002e-06, "loss": 2.4264, "step": 594 }, { "epoch": 0.8056872037914692, "grad_norm": 0.9127946818898369, "learning_rate": 9.130180714045271e-06, "loss": 1.985, "step": 595 }, { "epoch": 0.8070412999322952, "grad_norm": 1.0273656011649834, "learning_rate": 9.127138037210884e-06, "loss": 2.1752, "step": 596 }, { "epoch": 0.8083953960731212, "grad_norm": 1.0442611846851202, "learning_rate": 9.124090556786227e-06, "loss": 2.2373, "step": 597 }, { "epoch": 0.8097494922139472, "grad_norm": 1.0020569110815791, "learning_rate": 9.12103827631827e-06, "loss": 2.6107, "step": 598 }, { "epoch": 0.8111035883547731, "grad_norm": 1.0433877333186745, "learning_rate": 9.117981199359575e-06, "loss": 2.2095, "step": 599 }, { "epoch": 0.8124576844955992, "grad_norm": 0.9172575390929391, "learning_rate": 9.114919329468283e-06, "loss": 2.0521, "step": 600 }, { "epoch": 0.8138117806364252, "grad_norm": 1.0426202271128402, "learning_rate": 9.111852670208116e-06, "loss": 1.8817, "step": 601 }, { "epoch": 0.8151658767772512, "grad_norm": 1.1301209307772506, "learning_rate": 9.108781225148369e-06, "loss": 3.0088, "step": 602 }, { "epoch": 0.8165199729180772, "grad_norm": 0.9133026858806589, "learning_rate": 9.105704997863907e-06, "loss": 2.5472, "step": 603 }, { "epoch": 0.8178740690589031, "grad_norm": 1.1108413215075008, "learning_rate": 9.102623991935163e-06, "loss": 2.5752, "step": 604 }, { "epoch": 0.8192281651997292, "grad_norm": 0.9702100994035456, "learning_rate": 9.099538210948128e-06, "loss": 2.4404, "step": 605 }, { "epoch": 0.8205822613405552, "grad_norm": 0.9485690500451401, "learning_rate": 9.096447658494357e-06, "loss": 2.2257, "step": 606 }, { "epoch": 0.8219363574813812, "grad_norm": 26.292288625318744, "learning_rate": 9.09335233817095e-06, "loss": 2.676, "step": 607 }, { "epoch": 0.8232904536222072, "grad_norm": 0.9757213752439725, "learning_rate": 9.090252253580565e-06, "loss": 2.083, "step": 608 }, { "epoch": 0.8246445497630331, "grad_norm": 0.977754402583003, "learning_rate": 9.087147408331399e-06, "loss": 2.396, "step": 609 }, { "epoch": 0.8259986459038592, "grad_norm": 1.1224622683564631, "learning_rate": 9.084037806037194e-06, "loss": 2.5127, "step": 610 }, { "epoch": 0.8273527420446851, "grad_norm": 1.0702372697619176, "learning_rate": 9.080923450317226e-06, "loss": 2.44, "step": 611 }, { "epoch": 0.8287068381855112, "grad_norm": 0.9854940413106023, "learning_rate": 9.077804344796302e-06, "loss": 2.3179, "step": 612 }, { "epoch": 0.8300609343263372, "grad_norm": 1.0317420348048132, "learning_rate": 9.074680493104764e-06, "loss": 2.2255, "step": 613 }, { "epoch": 0.8314150304671631, "grad_norm": 1.2553199368667127, "learning_rate": 9.071551898878471e-06, "loss": 2.196, "step": 614 }, { "epoch": 0.8327691266079892, "grad_norm": 0.9428355967192141, "learning_rate": 9.068418565758805e-06, "loss": 2.4896, "step": 615 }, { "epoch": 0.8341232227488151, "grad_norm": 0.919605019517011, "learning_rate": 9.065280497392663e-06, "loss": 2.724, "step": 616 }, { "epoch": 0.8354773188896412, "grad_norm": 0.9006367820269072, "learning_rate": 9.062137697432457e-06, "loss": 2.3021, "step": 617 }, { "epoch": 0.8368314150304672, "grad_norm": 0.9254413019481625, "learning_rate": 9.058990169536098e-06, "loss": 2.8054, "step": 618 }, { "epoch": 0.8381855111712931, "grad_norm": 0.9723010342216781, "learning_rate": 9.055837917367006e-06, "loss": 2.1242, "step": 619 }, { "epoch": 0.8395396073121192, "grad_norm": 0.9646095993807202, "learning_rate": 9.052680944594101e-06, "loss": 2.7603, "step": 620 }, { "epoch": 0.8408937034529451, "grad_norm": 1.3152197929562175, "learning_rate": 9.049519254891793e-06, "loss": 2.1762, "step": 621 }, { "epoch": 0.8422477995937712, "grad_norm": 0.8097257743614227, "learning_rate": 9.046352851939981e-06, "loss": 2.4746, "step": 622 }, { "epoch": 0.8436018957345972, "grad_norm": 0.9993019751259021, "learning_rate": 9.043181739424054e-06, "loss": 1.8778, "step": 623 }, { "epoch": 0.8449559918754231, "grad_norm": 1.1394511401717762, "learning_rate": 9.040005921034884e-06, "loss": 2.2527, "step": 624 }, { "epoch": 0.8463100880162492, "grad_norm": 1.0167098829317125, "learning_rate": 9.036825400468814e-06, "loss": 2.4564, "step": 625 }, { "epoch": 0.8476641841570751, "grad_norm": 1.0799560174241793, "learning_rate": 9.033640181427661e-06, "loss": 2.1033, "step": 626 }, { "epoch": 0.8490182802979012, "grad_norm": 1.174120607908624, "learning_rate": 9.030450267618717e-06, "loss": 2.4757, "step": 627 }, { "epoch": 0.8503723764387271, "grad_norm": 1.050685050196375, "learning_rate": 9.02725566275473e-06, "loss": 2.238, "step": 628 }, { "epoch": 0.8517264725795531, "grad_norm": 0.9338038071047584, "learning_rate": 9.024056370553916e-06, "loss": 2.1313, "step": 629 }, { "epoch": 0.8530805687203792, "grad_norm": 1.156893096752152, "learning_rate": 9.020852394739936e-06, "loss": 2.2162, "step": 630 }, { "epoch": 0.8544346648612051, "grad_norm": 1.2811232350478499, "learning_rate": 9.017643739041914e-06, "loss": 2.9062, "step": 631 }, { "epoch": 0.8557887610020312, "grad_norm": 1.3380924542089319, "learning_rate": 9.014430407194413e-06, "loss": 2.4905, "step": 632 }, { "epoch": 0.8571428571428571, "grad_norm": 1.068909763409773, "learning_rate": 9.011212402937441e-06, "loss": 1.973, "step": 633 }, { "epoch": 0.8584969532836831, "grad_norm": 1.8932248755263266, "learning_rate": 9.007989730016446e-06, "loss": 2.4409, "step": 634 }, { "epoch": 0.8598510494245092, "grad_norm": 1.2891934792560118, "learning_rate": 9.004762392182307e-06, "loss": 2.3235, "step": 635 }, { "epoch": 0.8612051455653351, "grad_norm": 0.858242619324268, "learning_rate": 9.001530393191334e-06, "loss": 1.9865, "step": 636 }, { "epoch": 0.8625592417061612, "grad_norm": 1.1164311194466396, "learning_rate": 8.998293736805262e-06, "loss": 2.3765, "step": 637 }, { "epoch": 0.8639133378469871, "grad_norm": 1.2157891571025403, "learning_rate": 8.995052426791247e-06, "loss": 2.9348, "step": 638 }, { "epoch": 0.8652674339878131, "grad_norm": 1.0909914707099782, "learning_rate": 8.991806466921858e-06, "loss": 2.8517, "step": 639 }, { "epoch": 0.8666215301286392, "grad_norm": 0.9001967266754044, "learning_rate": 8.988555860975082e-06, "loss": 2.3269, "step": 640 }, { "epoch": 0.8679756262694651, "grad_norm": 1.0105961392957552, "learning_rate": 8.98530061273431e-06, "loss": 2.101, "step": 641 }, { "epoch": 0.8693297224102912, "grad_norm": 0.9677003303068539, "learning_rate": 8.982040725988337e-06, "loss": 2.2758, "step": 642 }, { "epoch": 0.8706838185511171, "grad_norm": 0.9813548892015744, "learning_rate": 8.978776204531354e-06, "loss": 2.33, "step": 643 }, { "epoch": 0.8720379146919431, "grad_norm": 1.0150274086140143, "learning_rate": 8.975507052162954e-06, "loss": 2.3066, "step": 644 }, { "epoch": 0.8733920108327691, "grad_norm": 1.0301765250355548, "learning_rate": 8.972233272688107e-06, "loss": 2.1587, "step": 645 }, { "epoch": 0.8747461069735951, "grad_norm": 1.1970855668426001, "learning_rate": 8.968954869917183e-06, "loss": 2.3845, "step": 646 }, { "epoch": 0.8761002031144212, "grad_norm": 1.1034962442803717, "learning_rate": 8.965671847665925e-06, "loss": 1.7817, "step": 647 }, { "epoch": 0.8774542992552471, "grad_norm": 0.9455749614353804, "learning_rate": 8.962384209755453e-06, "loss": 1.8517, "step": 648 }, { "epoch": 0.8788083953960731, "grad_norm": 0.9934755324282731, "learning_rate": 8.95909196001226e-06, "loss": 2.2541, "step": 649 }, { "epoch": 0.8801624915368991, "grad_norm": 1.0284578656567072, "learning_rate": 8.955795102268206e-06, "loss": 2.3968, "step": 650 }, { "epoch": 0.8815165876777251, "grad_norm": 1.037663714131099, "learning_rate": 8.952493640360518e-06, "loss": 2.6138, "step": 651 }, { "epoch": 0.8828706838185512, "grad_norm": 1.0926666615727456, "learning_rate": 8.949187578131777e-06, "loss": 1.9031, "step": 652 }, { "epoch": 0.8842247799593771, "grad_norm": 1.110544847466399, "learning_rate": 8.945876919429922e-06, "loss": 2.2067, "step": 653 }, { "epoch": 0.8855788761002031, "grad_norm": 1.0868074118477926, "learning_rate": 8.942561668108236e-06, "loss": 2.3532, "step": 654 }, { "epoch": 0.8869329722410291, "grad_norm": 1.1283691827694036, "learning_rate": 8.939241828025356e-06, "loss": 2.2511, "step": 655 }, { "epoch": 0.8882870683818551, "grad_norm": 1.9053437775194237, "learning_rate": 8.935917403045251e-06, "loss": 2.1626, "step": 656 }, { "epoch": 0.8896411645226812, "grad_norm": 1.2141944808281238, "learning_rate": 8.932588397037236e-06, "loss": 2.3409, "step": 657 }, { "epoch": 0.8909952606635071, "grad_norm": 1.0922208185194722, "learning_rate": 8.92925481387595e-06, "loss": 1.7841, "step": 658 }, { "epoch": 0.8923493568043331, "grad_norm": 1.5140227295690092, "learning_rate": 8.92591665744136e-06, "loss": 2.369, "step": 659 }, { "epoch": 0.8937034529451591, "grad_norm": 0.8526913110142397, "learning_rate": 8.922573931618763e-06, "loss": 2.0121, "step": 660 }, { "epoch": 0.8950575490859851, "grad_norm": 1.043736998424973, "learning_rate": 8.919226640298763e-06, "loss": 2.2168, "step": 661 }, { "epoch": 0.896411645226811, "grad_norm": 0.8649236144416748, "learning_rate": 8.915874787377289e-06, "loss": 1.9414, "step": 662 }, { "epoch": 0.8977657413676371, "grad_norm": 1.0840827090139065, "learning_rate": 8.912518376755572e-06, "loss": 3.087, "step": 663 }, { "epoch": 0.8991198375084631, "grad_norm": 1.1561825732631927, "learning_rate": 8.90915741234015e-06, "loss": 2.1006, "step": 664 }, { "epoch": 0.9004739336492891, "grad_norm": 0.9793291395752391, "learning_rate": 8.905791898042861e-06, "loss": 1.85, "step": 665 }, { "epoch": 0.9018280297901151, "grad_norm": 1.2185211503560396, "learning_rate": 8.902421837780839e-06, "loss": 2.7844, "step": 666 }, { "epoch": 0.903182125930941, "grad_norm": 1.0236959874012588, "learning_rate": 8.89904723547651e-06, "loss": 2.1372, "step": 667 }, { "epoch": 0.9045362220717671, "grad_norm": 1.0598101921330767, "learning_rate": 8.895668095057584e-06, "loss": 2.413, "step": 668 }, { "epoch": 0.9058903182125931, "grad_norm": 0.9973400634976877, "learning_rate": 8.892284420457054e-06, "loss": 2.1688, "step": 669 }, { "epoch": 0.9072444143534191, "grad_norm": 1.11854939708825, "learning_rate": 8.888896215613192e-06, "loss": 2.3154, "step": 670 }, { "epoch": 0.9085985104942451, "grad_norm": 1.0425526725686658, "learning_rate": 8.885503484469539e-06, "loss": 2.2115, "step": 671 }, { "epoch": 0.909952606635071, "grad_norm": 0.9579083361963218, "learning_rate": 8.88210623097491e-06, "loss": 2.0552, "step": 672 }, { "epoch": 0.9113067027758971, "grad_norm": 0.8237407553812768, "learning_rate": 8.878704459083376e-06, "loss": 2.8482, "step": 673 }, { "epoch": 0.9126607989167231, "grad_norm": 0.8200385146933777, "learning_rate": 8.875298172754274e-06, "loss": 2.1437, "step": 674 }, { "epoch": 0.9140148950575491, "grad_norm": 1.2074180626236117, "learning_rate": 8.871887375952192e-06, "loss": 2.0688, "step": 675 }, { "epoch": 0.9153689911983751, "grad_norm": 1.2749052909022498, "learning_rate": 8.868472072646968e-06, "loss": 2.4633, "step": 676 }, { "epoch": 0.916723087339201, "grad_norm": 1.219617129030152, "learning_rate": 8.865052266813686e-06, "loss": 2.3246, "step": 677 }, { "epoch": 0.9180771834800271, "grad_norm": 0.981406917677342, "learning_rate": 8.861627962432669e-06, "loss": 2.2168, "step": 678 }, { "epoch": 0.919431279620853, "grad_norm": 1.1941352888187047, "learning_rate": 8.858199163489476e-06, "loss": 2.0688, "step": 679 }, { "epoch": 0.9207853757616791, "grad_norm": 1.0302188684262572, "learning_rate": 8.854765873974898e-06, "loss": 2.6023, "step": 680 }, { "epoch": 0.9221394719025051, "grad_norm": 1.0315351443360778, "learning_rate": 8.851328097884956e-06, "loss": 2.4473, "step": 681 }, { "epoch": 0.923493568043331, "grad_norm": 0.9270825061496434, "learning_rate": 8.847885839220884e-06, "loss": 2.3396, "step": 682 }, { "epoch": 0.9248476641841571, "grad_norm": 1.5668051937578615, "learning_rate": 8.844439101989145e-06, "loss": 2.1618, "step": 683 }, { "epoch": 0.926201760324983, "grad_norm": 1.063281721298216, "learning_rate": 8.840987890201404e-06, "loss": 2.0825, "step": 684 }, { "epoch": 0.9275558564658091, "grad_norm": 5.49994743188365, "learning_rate": 8.837532207874539e-06, "loss": 2.3203, "step": 685 }, { "epoch": 0.9289099526066351, "grad_norm": 0.805384496899143, "learning_rate": 8.834072059030631e-06, "loss": 2.2733, "step": 686 }, { "epoch": 0.930264048747461, "grad_norm": 1.2279223754931932, "learning_rate": 8.83060744769696e-06, "loss": 2.519, "step": 687 }, { "epoch": 0.9316181448882871, "grad_norm": 0.9224586328427765, "learning_rate": 8.827138377905999e-06, "loss": 2.8326, "step": 688 }, { "epoch": 0.932972241029113, "grad_norm": 1.6089966176184487, "learning_rate": 8.823664853695408e-06, "loss": 1.6853, "step": 689 }, { "epoch": 0.9343263371699391, "grad_norm": 0.8870008681490367, "learning_rate": 8.820186879108038e-06, "loss": 2.4482, "step": 690 }, { "epoch": 0.9356804333107651, "grad_norm": 0.91606230626226, "learning_rate": 8.816704458191913e-06, "loss": 2.4619, "step": 691 }, { "epoch": 0.937034529451591, "grad_norm": 1.103378118700351, "learning_rate": 8.813217595000234e-06, "loss": 2.0875, "step": 692 }, { "epoch": 0.9383886255924171, "grad_norm": 1.1782658973751619, "learning_rate": 8.809726293591376e-06, "loss": 2.2522, "step": 693 }, { "epoch": 0.939742721733243, "grad_norm": 3.1082108827531383, "learning_rate": 8.806230558028874e-06, "loss": 2.4234, "step": 694 }, { "epoch": 0.9410968178740691, "grad_norm": 1.2771263022311399, "learning_rate": 8.80273039238143e-06, "loss": 2.2953, "step": 695 }, { "epoch": 0.942450914014895, "grad_norm": 1.0851177832663736, "learning_rate": 8.799225800722895e-06, "loss": 2.1023, "step": 696 }, { "epoch": 0.943805010155721, "grad_norm": 0.8982930726085163, "learning_rate": 8.795716787132278e-06, "loss": 2.261, "step": 697 }, { "epoch": 0.9451591062965471, "grad_norm": 0.9710609235588211, "learning_rate": 8.792203355693731e-06, "loss": 2.0593, "step": 698 }, { "epoch": 0.946513202437373, "grad_norm": 1.1589767335158037, "learning_rate": 8.788685510496549e-06, "loss": 2.2502, "step": 699 }, { "epoch": 0.9478672985781991, "grad_norm": 0.923602460813735, "learning_rate": 8.785163255635166e-06, "loss": 2.1052, "step": 700 }, { "epoch": 0.949221394719025, "grad_norm": 1.1331497754597915, "learning_rate": 8.781636595209145e-06, "loss": 2.2342, "step": 701 }, { "epoch": 0.950575490859851, "grad_norm": 1.0911649916589345, "learning_rate": 8.778105533323177e-06, "loss": 2.5234, "step": 702 }, { "epoch": 0.9519295870006771, "grad_norm": 0.7939604108123635, "learning_rate": 8.77457007408708e-06, "loss": 2.3947, "step": 703 }, { "epoch": 0.953283683141503, "grad_norm": 0.9952541607703701, "learning_rate": 8.771030221615786e-06, "loss": 2.4793, "step": 704 }, { "epoch": 0.9546377792823291, "grad_norm": 1.0107073531709874, "learning_rate": 8.767485980029342e-06, "loss": 2.0951, "step": 705 }, { "epoch": 0.955991875423155, "grad_norm": 1.0100357247296368, "learning_rate": 8.763937353452902e-06, "loss": 2.1011, "step": 706 }, { "epoch": 0.957345971563981, "grad_norm": 0.9804384454423593, "learning_rate": 8.760384346016726e-06, "loss": 1.7807, "step": 707 }, { "epoch": 0.9587000677048071, "grad_norm": 1.948525190001482, "learning_rate": 8.756826961856171e-06, "loss": 2.2514, "step": 708 }, { "epoch": 0.960054163845633, "grad_norm": 2.325650876165679, "learning_rate": 8.75326520511169e-06, "loss": 2.43, "step": 709 }, { "epoch": 0.9614082599864591, "grad_norm": 1.4411906872847189, "learning_rate": 8.74969907992882e-06, "loss": 2.5396, "step": 710 }, { "epoch": 0.962762356127285, "grad_norm": 1.1588565975941216, "learning_rate": 8.746128590458191e-06, "loss": 2.2853, "step": 711 }, { "epoch": 0.964116452268111, "grad_norm": 1.018506381733888, "learning_rate": 8.742553740855507e-06, "loss": 2.3352, "step": 712 }, { "epoch": 0.965470548408937, "grad_norm": 0.9885672968116678, "learning_rate": 8.738974535281545e-06, "loss": 2.3807, "step": 713 }, { "epoch": 0.966824644549763, "grad_norm": 0.9360638976400335, "learning_rate": 8.735390977902159e-06, "loss": 2.522, "step": 714 }, { "epoch": 0.9681787406905891, "grad_norm": 1.0138074790076992, "learning_rate": 8.731803072888262e-06, "loss": 2.0687, "step": 715 }, { "epoch": 0.969532836831415, "grad_norm": 0.8593670335614861, "learning_rate": 8.728210824415829e-06, "loss": 2.3171, "step": 716 }, { "epoch": 0.970886932972241, "grad_norm": 0.8782947340827698, "learning_rate": 8.724614236665889e-06, "loss": 2.1826, "step": 717 }, { "epoch": 0.972241029113067, "grad_norm": 1.041877641264968, "learning_rate": 8.721013313824527e-06, "loss": 2.5999, "step": 718 }, { "epoch": 0.973595125253893, "grad_norm": 1.0184265790630844, "learning_rate": 8.717408060082865e-06, "loss": 2.3649, "step": 719 }, { "epoch": 0.9749492213947191, "grad_norm": 0.8805171555606713, "learning_rate": 8.713798479637073e-06, "loss": 2.5529, "step": 720 }, { "epoch": 0.976303317535545, "grad_norm": 0.8250466080220131, "learning_rate": 8.710184576688353e-06, "loss": 2.1707, "step": 721 }, { "epoch": 0.977657413676371, "grad_norm": 0.9735738962932305, "learning_rate": 8.70656635544294e-06, "loss": 1.9474, "step": 722 }, { "epoch": 0.979011509817197, "grad_norm": 0.9574755177161404, "learning_rate": 8.702943820112094e-06, "loss": 2.0991, "step": 723 }, { "epoch": 0.980365605958023, "grad_norm": 1.1008036232420575, "learning_rate": 8.699316974912097e-06, "loss": 2.6044, "step": 724 }, { "epoch": 0.9817197020988491, "grad_norm": 1.0106221199017615, "learning_rate": 8.695685824064246e-06, "loss": 2.4405, "step": 725 }, { "epoch": 0.983073798239675, "grad_norm": 1.167558559334278, "learning_rate": 8.692050371794849e-06, "loss": 2.6635, "step": 726 }, { "epoch": 0.984427894380501, "grad_norm": 0.8542048524860696, "learning_rate": 8.688410622335222e-06, "loss": 2.5169, "step": 727 }, { "epoch": 0.985781990521327, "grad_norm": 1.0044160146409207, "learning_rate": 8.684766579921684e-06, "loss": 2.477, "step": 728 }, { "epoch": 0.987136086662153, "grad_norm": 0.9438656631341891, "learning_rate": 8.681118248795548e-06, "loss": 1.9811, "step": 729 }, { "epoch": 0.988490182802979, "grad_norm": 0.9095360753452296, "learning_rate": 8.677465633203117e-06, "loss": 2.3501, "step": 730 }, { "epoch": 0.989844278943805, "grad_norm": 1.5751129180562193, "learning_rate": 8.673808737395686e-06, "loss": 1.9711, "step": 731 }, { "epoch": 0.991198375084631, "grad_norm": 1.0244041655385774, "learning_rate": 8.670147565629526e-06, "loss": 2.1411, "step": 732 }, { "epoch": 0.992552471225457, "grad_norm": 1.0999564192877611, "learning_rate": 8.66648212216589e-06, "loss": 2.7566, "step": 733 }, { "epoch": 0.993906567366283, "grad_norm": 1.0783864346910368, "learning_rate": 8.662812411270997e-06, "loss": 2.2272, "step": 734 }, { "epoch": 0.995260663507109, "grad_norm": 1.011628347758097, "learning_rate": 8.65913843721604e-06, "loss": 2.5275, "step": 735 }, { "epoch": 0.996614759647935, "grad_norm": 1.2637326655889927, "learning_rate": 8.655460204277167e-06, "loss": 2.7871, "step": 736 }, { "epoch": 0.997968855788761, "grad_norm": 0.8940421539736993, "learning_rate": 8.651777716735488e-06, "loss": 2.8055, "step": 737 }, { "epoch": 0.999322951929587, "grad_norm": 0.8894453146981818, "learning_rate": 8.648090978877063e-06, "loss": 2.2385, "step": 738 }, { "epoch": 1.000677048070413, "grad_norm": 0.9588436493567126, "learning_rate": 8.644399994992898e-06, "loss": 2.1316, "step": 739 }, { "epoch": 1.001354096140826, "grad_norm": 1.042596861537762, "learning_rate": 8.640704769378943e-06, "loss": 2.1595, "step": 740 }, { "epoch": 1.002708192281652, "grad_norm": 0.9542700464860409, "learning_rate": 8.637005306336083e-06, "loss": 2.2683, "step": 741 }, { "epoch": 1.004062288422478, "grad_norm": 1.0792626697240038, "learning_rate": 8.633301610170136e-06, "loss": 1.9535, "step": 742 }, { "epoch": 1.005416384563304, "grad_norm": 0.9813632090639985, "learning_rate": 8.629593685191846e-06, "loss": 2.3908, "step": 743 }, { "epoch": 1.00677048070413, "grad_norm": 1.0359954057575655, "learning_rate": 8.625881535716883e-06, "loss": 2.6186, "step": 744 }, { "epoch": 1.008124576844956, "grad_norm": 0.9759453555493578, "learning_rate": 8.622165166065828e-06, "loss": 1.926, "step": 745 }, { "epoch": 1.009478672985782, "grad_norm": 1.086383812605352, "learning_rate": 8.618444580564175e-06, "loss": 2.4117, "step": 746 }, { "epoch": 1.010832769126608, "grad_norm": 1.066485208921513, "learning_rate": 8.61471978354233e-06, "loss": 2.2121, "step": 747 }, { "epoch": 1.012186865267434, "grad_norm": 1.0075256921761013, "learning_rate": 8.610990779335594e-06, "loss": 2.2686, "step": 748 }, { "epoch": 1.01354096140826, "grad_norm": 4.05838919324133, "learning_rate": 8.60725757228417e-06, "loss": 2.3371, "step": 749 }, { "epoch": 1.014895057549086, "grad_norm": 1.0649638809266402, "learning_rate": 8.603520166733153e-06, "loss": 2.2376, "step": 750 }, { "epoch": 1.016249153689912, "grad_norm": 0.9312921856108541, "learning_rate": 8.599778567032515e-06, "loss": 2.2602, "step": 751 }, { "epoch": 1.017603249830738, "grad_norm": 1.013555848121877, "learning_rate": 8.596032777537124e-06, "loss": 2.2348, "step": 752 }, { "epoch": 1.018957345971564, "grad_norm": 1.0949907267977028, "learning_rate": 8.592282802606713e-06, "loss": 2.2799, "step": 753 }, { "epoch": 1.02031144211239, "grad_norm": 0.8061510705350392, "learning_rate": 8.588528646605893e-06, "loss": 2.1358, "step": 754 }, { "epoch": 1.021665538253216, "grad_norm": 0.8912985567283312, "learning_rate": 8.584770313904138e-06, "loss": 2.1013, "step": 755 }, { "epoch": 1.0230196343940419, "grad_norm": 0.938063703434926, "learning_rate": 8.581007808875786e-06, "loss": 2.4889, "step": 756 }, { "epoch": 1.024373730534868, "grad_norm": 0.7577265350168257, "learning_rate": 8.577241135900026e-06, "loss": 1.8219, "step": 757 }, { "epoch": 1.025727826675694, "grad_norm": 1.0606203403633938, "learning_rate": 8.573470299360906e-06, "loss": 2.4437, "step": 758 }, { "epoch": 1.02708192281652, "grad_norm": 1.0551285967576851, "learning_rate": 8.569695303647313e-06, "loss": 2.1422, "step": 759 }, { "epoch": 1.028436018957346, "grad_norm": 0.9409375423377818, "learning_rate": 8.565916153152982e-06, "loss": 2.1203, "step": 760 }, { "epoch": 1.0297901150981719, "grad_norm": 1.1003263863457742, "learning_rate": 8.562132852276474e-06, "loss": 2.338, "step": 761 }, { "epoch": 1.031144211238998, "grad_norm": 0.9457106296289286, "learning_rate": 8.558345405421189e-06, "loss": 1.9891, "step": 762 }, { "epoch": 1.032498307379824, "grad_norm": 1.0094972443672454, "learning_rate": 8.554553816995347e-06, "loss": 2.064, "step": 763 }, { "epoch": 1.03385240352065, "grad_norm": 0.9931533657371708, "learning_rate": 8.550758091411993e-06, "loss": 2.5872, "step": 764 }, { "epoch": 1.035206499661476, "grad_norm": 1.00970787542275, "learning_rate": 8.546958233088986e-06, "loss": 1.6308, "step": 765 }, { "epoch": 1.0365605958023019, "grad_norm": 0.9046067132912955, "learning_rate": 8.543154246448992e-06, "loss": 2.4595, "step": 766 }, { "epoch": 1.037914691943128, "grad_norm": 0.8869062626220933, "learning_rate": 8.539346135919485e-06, "loss": 2.2769, "step": 767 }, { "epoch": 1.039268788083954, "grad_norm": 1.1369978829634468, "learning_rate": 8.535533905932739e-06, "loss": 2.2687, "step": 768 }, { "epoch": 1.04062288422478, "grad_norm": 1.000069084621953, "learning_rate": 8.531717560925819e-06, "loss": 1.9722, "step": 769 }, { "epoch": 1.041976980365606, "grad_norm": 1.1369770696046089, "learning_rate": 8.527897105340587e-06, "loss": 2.3523, "step": 770 }, { "epoch": 1.0433310765064319, "grad_norm": 1.2290321379198472, "learning_rate": 8.52407254362368e-06, "loss": 2.2146, "step": 771 }, { "epoch": 1.044685172647258, "grad_norm": 1.0959412542423825, "learning_rate": 8.52024388022652e-06, "loss": 2.6636, "step": 772 }, { "epoch": 1.046039268788084, "grad_norm": 1.0258500561432276, "learning_rate": 8.516411119605303e-06, "loss": 2.8278, "step": 773 }, { "epoch": 1.04739336492891, "grad_norm": 0.9866371420929005, "learning_rate": 8.512574266220991e-06, "loss": 2.5915, "step": 774 }, { "epoch": 1.048747461069736, "grad_norm": 1.0612020165548255, "learning_rate": 8.50873332453931e-06, "loss": 2.4831, "step": 775 }, { "epoch": 1.0501015572105619, "grad_norm": 0.9966103250455697, "learning_rate": 8.504888299030748e-06, "loss": 2.1428, "step": 776 }, { "epoch": 1.051455653351388, "grad_norm": 1.0897590636471566, "learning_rate": 8.501039194170542e-06, "loss": 2.0724, "step": 777 }, { "epoch": 1.052809749492214, "grad_norm": 1.0670474685447335, "learning_rate": 8.497186014438677e-06, "loss": 1.9695, "step": 778 }, { "epoch": 1.05416384563304, "grad_norm": 1.1907785256874888, "learning_rate": 8.493328764319886e-06, "loss": 2.4858, "step": 779 }, { "epoch": 1.055517941773866, "grad_norm": 1.2893351808110196, "learning_rate": 8.489467448303633e-06, "loss": 2.7964, "step": 780 }, { "epoch": 1.0568720379146919, "grad_norm": 1.0302272518222444, "learning_rate": 8.485602070884118e-06, "loss": 2.3342, "step": 781 }, { "epoch": 1.058226134055518, "grad_norm": 1.0206240865882117, "learning_rate": 8.481732636560266e-06, "loss": 2.4793, "step": 782 }, { "epoch": 1.059580230196344, "grad_norm": 1.0121397248564201, "learning_rate": 8.477859149835726e-06, "loss": 1.8307, "step": 783 }, { "epoch": 1.06093432633717, "grad_norm": 1.0578580561059918, "learning_rate": 8.473981615218863e-06, "loss": 2.1949, "step": 784 }, { "epoch": 1.062288422477996, "grad_norm": 0.9969690177718763, "learning_rate": 8.470100037222751e-06, "loss": 2.1533, "step": 785 }, { "epoch": 1.0636425186188219, "grad_norm": 1.1212678356984322, "learning_rate": 8.466214420365176e-06, "loss": 2.1861, "step": 786 }, { "epoch": 1.064996614759648, "grad_norm": 0.9923066032834624, "learning_rate": 8.462324769168618e-06, "loss": 2.4372, "step": 787 }, { "epoch": 1.066350710900474, "grad_norm": 1.2404978604054542, "learning_rate": 8.458431088160255e-06, "loss": 2.0584, "step": 788 }, { "epoch": 1.0677048070413, "grad_norm": 1.060465239616255, "learning_rate": 8.454533381871957e-06, "loss": 2.3261, "step": 789 }, { "epoch": 1.069058903182126, "grad_norm": 1.0780439600876397, "learning_rate": 8.450631654840279e-06, "loss": 2.2541, "step": 790 }, { "epoch": 1.0704129993229519, "grad_norm": 1.0543355817090738, "learning_rate": 8.446725911606451e-06, "loss": 2.3729, "step": 791 }, { "epoch": 1.071767095463778, "grad_norm": 1.5929978362842498, "learning_rate": 8.442816156716386e-06, "loss": 2.2915, "step": 792 }, { "epoch": 1.073121191604604, "grad_norm": 0.8559195523501567, "learning_rate": 8.438902394720659e-06, "loss": 1.9731, "step": 793 }, { "epoch": 1.07447528774543, "grad_norm": 1.0220594117408883, "learning_rate": 8.43498463017451e-06, "loss": 2.0958, "step": 794 }, { "epoch": 1.0758293838862558, "grad_norm": 1.0326806004121811, "learning_rate": 8.431062867637838e-06, "loss": 2.0738, "step": 795 }, { "epoch": 1.0771834800270819, "grad_norm": 1.18276220665135, "learning_rate": 8.4271371116752e-06, "loss": 2.3479, "step": 796 }, { "epoch": 1.078537576167908, "grad_norm": 1.1320006680638006, "learning_rate": 8.423207366855794e-06, "loss": 1.8838, "step": 797 }, { "epoch": 1.079891672308734, "grad_norm": 0.8967800856440162, "learning_rate": 8.419273637753463e-06, "loss": 2.2114, "step": 798 }, { "epoch": 1.08124576844956, "grad_norm": 0.9935120415012008, "learning_rate": 8.415335928946691e-06, "loss": 2.3093, "step": 799 }, { "epoch": 1.0825998645903858, "grad_norm": 0.9813682594289855, "learning_rate": 8.411394245018589e-06, "loss": 2.2393, "step": 800 }, { "epoch": 1.0839539607312119, "grad_norm": 1.0470052038748332, "learning_rate": 8.407448590556895e-06, "loss": 2.0396, "step": 801 }, { "epoch": 1.085308056872038, "grad_norm": 1.0097113780081433, "learning_rate": 8.403498970153973e-06, "loss": 2.1874, "step": 802 }, { "epoch": 1.086662153012864, "grad_norm": 0.9573725041864002, "learning_rate": 8.399545388406798e-06, "loss": 2.2734, "step": 803 }, { "epoch": 1.08801624915369, "grad_norm": 0.9578847774313853, "learning_rate": 8.39558784991696e-06, "loss": 2.186, "step": 804 }, { "epoch": 1.0893703452945158, "grad_norm": 0.9151868078460803, "learning_rate": 8.391626359290648e-06, "loss": 2.0554, "step": 805 }, { "epoch": 1.0907244414353419, "grad_norm": 0.9630038514154003, "learning_rate": 8.387660921138657e-06, "loss": 1.6792, "step": 806 }, { "epoch": 1.092078537576168, "grad_norm": 1.0284639983718127, "learning_rate": 8.383691540076372e-06, "loss": 2.795, "step": 807 }, { "epoch": 1.093432633716994, "grad_norm": 1.0684029705952454, "learning_rate": 8.379718220723772e-06, "loss": 2.2904, "step": 808 }, { "epoch": 1.09478672985782, "grad_norm": 1.0034018245502179, "learning_rate": 8.375740967705418e-06, "loss": 2.0063, "step": 809 }, { "epoch": 1.0961408259986458, "grad_norm": 0.9850203635464737, "learning_rate": 8.371759785650444e-06, "loss": 2.0104, "step": 810 }, { "epoch": 1.0974949221394719, "grad_norm": 1.0272152976204258, "learning_rate": 8.367774679192566e-06, "loss": 2.4944, "step": 811 }, { "epoch": 1.098849018280298, "grad_norm": 1.014673449942826, "learning_rate": 8.363785652970062e-06, "loss": 2.4625, "step": 812 }, { "epoch": 1.100203114421124, "grad_norm": 0.9983178702745071, "learning_rate": 8.359792711625773e-06, "loss": 2.6321, "step": 813 }, { "epoch": 1.10155721056195, "grad_norm": 1.2478151045690562, "learning_rate": 8.355795859807097e-06, "loss": 2.6152, "step": 814 }, { "epoch": 1.1029113067027758, "grad_norm": 1.1326333593128552, "learning_rate": 8.351795102165983e-06, "loss": 2.3933, "step": 815 }, { "epoch": 1.1042654028436019, "grad_norm": 1.2156960926171398, "learning_rate": 8.34779044335893e-06, "loss": 2.5298, "step": 816 }, { "epoch": 1.105619498984428, "grad_norm": 1.0345252180165674, "learning_rate": 8.343781888046971e-06, "loss": 2.0308, "step": 817 }, { "epoch": 1.106973595125254, "grad_norm": 0.8846702899770009, "learning_rate": 8.339769440895678e-06, "loss": 2.0716, "step": 818 }, { "epoch": 1.10832769126608, "grad_norm": 0.9633796785097113, "learning_rate": 8.335753106575153e-06, "loss": 1.9232, "step": 819 }, { "epoch": 1.1096817874069058, "grad_norm": 1.0958725343390576, "learning_rate": 8.331732889760021e-06, "loss": 2.2685, "step": 820 }, { "epoch": 1.1110358835477319, "grad_norm": 0.9434748502922011, "learning_rate": 8.327708795129426e-06, "loss": 2.5492, "step": 821 }, { "epoch": 1.112389979688558, "grad_norm": 1.0218749260607791, "learning_rate": 8.323680827367027e-06, "loss": 2.1566, "step": 822 }, { "epoch": 1.113744075829384, "grad_norm": 0.9991712465528062, "learning_rate": 8.319648991160987e-06, "loss": 1.9646, "step": 823 }, { "epoch": 1.1150981719702098, "grad_norm": 0.7872947693815001, "learning_rate": 8.315613291203977e-06, "loss": 2.1754, "step": 824 }, { "epoch": 1.1164522681110358, "grad_norm": 1.0951406212655457, "learning_rate": 8.311573732193162e-06, "loss": 2.3643, "step": 825 }, { "epoch": 1.1178063642518619, "grad_norm": 1.162631966134576, "learning_rate": 8.307530318830195e-06, "loss": 2.447, "step": 826 }, { "epoch": 1.119160460392688, "grad_norm": 1.1057256649020668, "learning_rate": 8.303483055821222e-06, "loss": 2.037, "step": 827 }, { "epoch": 1.120514556533514, "grad_norm": 1.3428986372890304, "learning_rate": 8.299431947876868e-06, "loss": 2.3625, "step": 828 }, { "epoch": 1.12186865267434, "grad_norm": 1.1140355194533162, "learning_rate": 8.29537699971223e-06, "loss": 2.6255, "step": 829 }, { "epoch": 1.1232227488151658, "grad_norm": 1.3177560768420005, "learning_rate": 8.291318216046874e-06, "loss": 2.193, "step": 830 }, { "epoch": 1.1245768449559919, "grad_norm": 1.0528043127642381, "learning_rate": 8.287255601604834e-06, "loss": 1.8949, "step": 831 }, { "epoch": 1.125930941096818, "grad_norm": 1.1208525533921565, "learning_rate": 8.283189161114602e-06, "loss": 2.8527, "step": 832 }, { "epoch": 1.127285037237644, "grad_norm": 1.0730370677327041, "learning_rate": 8.279118899309121e-06, "loss": 2.4503, "step": 833 }, { "epoch": 1.1286391333784698, "grad_norm": 0.9202040373418652, "learning_rate": 8.275044820925784e-06, "loss": 2.066, "step": 834 }, { "epoch": 1.1299932295192958, "grad_norm": 1.0579688082270045, "learning_rate": 8.270966930706424e-06, "loss": 2.9167, "step": 835 }, { "epoch": 1.1313473256601219, "grad_norm": 1.0848914875005917, "learning_rate": 8.266885233397312e-06, "loss": 2.8931, "step": 836 }, { "epoch": 1.132701421800948, "grad_norm": 0.9167254802739708, "learning_rate": 8.262799733749147e-06, "loss": 2.5057, "step": 837 }, { "epoch": 1.134055517941774, "grad_norm": 1.066012346340753, "learning_rate": 8.25871043651706e-06, "loss": 2.3063, "step": 838 }, { "epoch": 1.1354096140825998, "grad_norm": 1.0335819197341622, "learning_rate": 8.254617346460598e-06, "loss": 2.333, "step": 839 }, { "epoch": 1.1367637102234258, "grad_norm": 1.2740330227860488, "learning_rate": 8.250520468343722e-06, "loss": 2.1914, "step": 840 }, { "epoch": 1.1381178063642519, "grad_norm": 0.9819332138988384, "learning_rate": 8.2464198069348e-06, "loss": 2.2113, "step": 841 }, { "epoch": 1.139471902505078, "grad_norm": 0.876882108695887, "learning_rate": 8.24231536700661e-06, "loss": 2.0397, "step": 842 }, { "epoch": 1.140825998645904, "grad_norm": 1.0295962020138234, "learning_rate": 8.238207153336322e-06, "loss": 2.0076, "step": 843 }, { "epoch": 1.1421800947867298, "grad_norm": 0.934240502035731, "learning_rate": 8.2340951707055e-06, "loss": 2.1013, "step": 844 }, { "epoch": 1.1435341909275558, "grad_norm": 1.0551844052623185, "learning_rate": 8.229979423900095e-06, "loss": 2.0243, "step": 845 }, { "epoch": 1.1448882870683819, "grad_norm": 1.0967095459364826, "learning_rate": 8.22585991771044e-06, "loss": 2.1247, "step": 846 }, { "epoch": 1.146242383209208, "grad_norm": 1.752846560905714, "learning_rate": 8.22173665693124e-06, "loss": 2.0735, "step": 847 }, { "epoch": 1.147596479350034, "grad_norm": 1.3734604707116838, "learning_rate": 8.217609646361574e-06, "loss": 1.8235, "step": 848 }, { "epoch": 1.1489505754908598, "grad_norm": 1.1360548028045079, "learning_rate": 8.213478890804884e-06, "loss": 2.4021, "step": 849 }, { "epoch": 1.1503046716316858, "grad_norm": 1.0221199810893757, "learning_rate": 8.209344395068972e-06, "loss": 1.9002, "step": 850 }, { "epoch": 1.1516587677725119, "grad_norm": 0.8652009856618762, "learning_rate": 8.20520616396599e-06, "loss": 1.6506, "step": 851 }, { "epoch": 1.153012863913338, "grad_norm": 0.8365704134586867, "learning_rate": 8.20106420231244e-06, "loss": 2.4525, "step": 852 }, { "epoch": 1.1543669600541637, "grad_norm": 1.1184140556213078, "learning_rate": 8.196918514929166e-06, "loss": 1.9014, "step": 853 }, { "epoch": 1.1557210561949898, "grad_norm": 1.0082190304094705, "learning_rate": 8.192769106641349e-06, "loss": 2.3114, "step": 854 }, { "epoch": 1.1570751523358158, "grad_norm": 1.2193028904213525, "learning_rate": 8.188615982278496e-06, "loss": 2.6133, "step": 855 }, { "epoch": 1.1584292484766419, "grad_norm": 1.361443418854842, "learning_rate": 8.184459146674447e-06, "loss": 2.4022, "step": 856 }, { "epoch": 1.159783344617468, "grad_norm": 1.1918066000555871, "learning_rate": 8.180298604667354e-06, "loss": 2.4705, "step": 857 }, { "epoch": 1.161137440758294, "grad_norm": 1.061895871052199, "learning_rate": 8.17613436109969e-06, "loss": 2.2618, "step": 858 }, { "epoch": 1.1624915368991198, "grad_norm": 1.107285284892013, "learning_rate": 8.171966420818227e-06, "loss": 2.2433, "step": 859 }, { "epoch": 1.1638456330399458, "grad_norm": 1.1870517898017734, "learning_rate": 8.167794788674051e-06, "loss": 2.2089, "step": 860 }, { "epoch": 1.1651997291807719, "grad_norm": 0.9482677548752911, "learning_rate": 8.163619469522537e-06, "loss": 2.2702, "step": 861 }, { "epoch": 1.166553825321598, "grad_norm": 0.9617964781357382, "learning_rate": 8.159440468223352e-06, "loss": 1.9168, "step": 862 }, { "epoch": 1.1679079214624237, "grad_norm": 0.9780264000997955, "learning_rate": 8.155257789640452e-06, "loss": 2.1226, "step": 863 }, { "epoch": 1.1692620176032498, "grad_norm": 1.0030932897097675, "learning_rate": 8.15107143864207e-06, "loss": 2.2428, "step": 864 }, { "epoch": 1.1706161137440758, "grad_norm": 1.0659315727721794, "learning_rate": 8.146881420100715e-06, "loss": 2.3583, "step": 865 }, { "epoch": 1.1719702098849019, "grad_norm": 1.1995965464770133, "learning_rate": 8.142687738893161e-06, "loss": 2.1148, "step": 866 }, { "epoch": 1.173324306025728, "grad_norm": 1.0100674325958858, "learning_rate": 8.138490399900454e-06, "loss": 1.7792, "step": 867 }, { "epoch": 1.174678402166554, "grad_norm": 1.1528480305040367, "learning_rate": 8.134289408007887e-06, "loss": 2.2195, "step": 868 }, { "epoch": 1.1760324983073798, "grad_norm": 0.9508957577380588, "learning_rate": 8.130084768105011e-06, "loss": 1.5612, "step": 869 }, { "epoch": 1.1773865944482058, "grad_norm": 1.0003595585872833, "learning_rate": 8.12587648508562e-06, "loss": 1.8605, "step": 870 }, { "epoch": 1.1787406905890319, "grad_norm": 1.1420205399905372, "learning_rate": 8.12166456384775e-06, "loss": 2.0476, "step": 871 }, { "epoch": 1.180094786729858, "grad_norm": 1.091147689337014, "learning_rate": 8.117449009293668e-06, "loss": 2.238, "step": 872 }, { "epoch": 1.1814488828706837, "grad_norm": 1.1033755210296698, "learning_rate": 8.113229826329876e-06, "loss": 2.3103, "step": 873 }, { "epoch": 1.1828029790115098, "grad_norm": 0.9843912490480584, "learning_rate": 8.109007019867097e-06, "loss": 2.2559, "step": 874 }, { "epoch": 1.1841570751523358, "grad_norm": 1.0569456356202054, "learning_rate": 8.104780594820265e-06, "loss": 2.0617, "step": 875 }, { "epoch": 1.1855111712931619, "grad_norm": 0.9431626809735268, "learning_rate": 8.100550556108534e-06, "loss": 2.216, "step": 876 }, { "epoch": 1.186865267433988, "grad_norm": 0.9619228085042159, "learning_rate": 8.09631690865526e-06, "loss": 2.6777, "step": 877 }, { "epoch": 1.1882193635748137, "grad_norm": 1.1921389359070207, "learning_rate": 8.092079657388e-06, "loss": 2.1743, "step": 878 }, { "epoch": 1.1895734597156398, "grad_norm": 0.9898902262989339, "learning_rate": 8.087838807238506e-06, "loss": 2.0959, "step": 879 }, { "epoch": 1.1909275558564658, "grad_norm": 1.143265112227052, "learning_rate": 8.083594363142717e-06, "loss": 1.7081, "step": 880 }, { "epoch": 1.1922816519972919, "grad_norm": 1.036462321572088, "learning_rate": 8.079346330040757e-06, "loss": 1.7609, "step": 881 }, { "epoch": 1.193635748138118, "grad_norm": 1.0246337259282456, "learning_rate": 8.075094712876925e-06, "loss": 2.4473, "step": 882 }, { "epoch": 1.1949898442789437, "grad_norm": 1.1209881301108675, "learning_rate": 8.070839516599695e-06, "loss": 1.7249, "step": 883 }, { "epoch": 1.1963439404197698, "grad_norm": 1.128942962423783, "learning_rate": 8.066580746161705e-06, "loss": 2.8118, "step": 884 }, { "epoch": 1.1976980365605958, "grad_norm": 1.1416681649771092, "learning_rate": 8.062318406519751e-06, "loss": 2.3983, "step": 885 }, { "epoch": 1.1990521327014219, "grad_norm": 1.1873758496607623, "learning_rate": 8.058052502634786e-06, "loss": 1.9744, "step": 886 }, { "epoch": 1.200406228842248, "grad_norm": 1.0805673821547641, "learning_rate": 8.053783039471909e-06, "loss": 2.4546, "step": 887 }, { "epoch": 1.2017603249830737, "grad_norm": 1.0642882838498282, "learning_rate": 8.049510022000365e-06, "loss": 2.063, "step": 888 }, { "epoch": 1.2031144211238998, "grad_norm": 1.0624621525792766, "learning_rate": 8.045233455193531e-06, "loss": 2.1409, "step": 889 }, { "epoch": 1.2044685172647258, "grad_norm": 1.2346691816252744, "learning_rate": 8.040953344028921e-06, "loss": 2.0547, "step": 890 }, { "epoch": 1.2058226134055519, "grad_norm": 0.9557989229824211, "learning_rate": 8.03666969348817e-06, "loss": 1.9712, "step": 891 }, { "epoch": 1.2071767095463777, "grad_norm": 0.8441961698009585, "learning_rate": 8.032382508557033e-06, "loss": 2.2598, "step": 892 }, { "epoch": 1.2085308056872037, "grad_norm": 1.009678392621935, "learning_rate": 8.02809179422538e-06, "loss": 2.2036, "step": 893 }, { "epoch": 1.2098849018280298, "grad_norm": 1.4057498442542968, "learning_rate": 8.023797555487188e-06, "loss": 2.648, "step": 894 }, { "epoch": 1.2112389979688558, "grad_norm": 1.1266529635992393, "learning_rate": 8.019499797340537e-06, "loss": 2.0414, "step": 895 }, { "epoch": 1.2125930941096819, "grad_norm": 1.2136121598603382, "learning_rate": 8.015198524787603e-06, "loss": 2.2459, "step": 896 }, { "epoch": 1.213947190250508, "grad_norm": 1.1213074117332114, "learning_rate": 8.01089374283465e-06, "loss": 2.2054, "step": 897 }, { "epoch": 1.2153012863913337, "grad_norm": 1.0818726596445376, "learning_rate": 8.00658545649203e-06, "loss": 2.2397, "step": 898 }, { "epoch": 1.2166553825321598, "grad_norm": 1.0597814871253108, "learning_rate": 8.002273670774172e-06, "loss": 2.4344, "step": 899 }, { "epoch": 1.2180094786729858, "grad_norm": 1.0180557577675222, "learning_rate": 7.99795839069958e-06, "loss": 1.9542, "step": 900 }, { "epoch": 1.2193635748138119, "grad_norm": 1.0732743540144358, "learning_rate": 7.99363962129082e-06, "loss": 1.9053, "step": 901 }, { "epoch": 1.2207176709546377, "grad_norm": 1.1464961982270008, "learning_rate": 7.989317367574528e-06, "loss": 2.2012, "step": 902 }, { "epoch": 1.2220717670954637, "grad_norm": 0.990300248509241, "learning_rate": 7.984991634581384e-06, "loss": 1.9006, "step": 903 }, { "epoch": 1.2234258632362898, "grad_norm": 1.0378186156083944, "learning_rate": 7.980662427346127e-06, "loss": 1.8861, "step": 904 }, { "epoch": 1.2247799593771158, "grad_norm": 4.430236914366399, "learning_rate": 7.976329750907536e-06, "loss": 2.1467, "step": 905 }, { "epoch": 1.2261340555179419, "grad_norm": 1.0098877057349103, "learning_rate": 7.971993610308428e-06, "loss": 2.0393, "step": 906 }, { "epoch": 1.2274881516587677, "grad_norm": 1.0949603281546505, "learning_rate": 7.967654010595653e-06, "loss": 2.2995, "step": 907 }, { "epoch": 1.2288422477995937, "grad_norm": 1.1108897890970795, "learning_rate": 7.963310956820085e-06, "loss": 2.2542, "step": 908 }, { "epoch": 1.2301963439404198, "grad_norm": 1.1882474635276283, "learning_rate": 7.95896445403662e-06, "loss": 1.9627, "step": 909 }, { "epoch": 1.2315504400812458, "grad_norm": 1.0695557890668739, "learning_rate": 7.954614507304166e-06, "loss": 2.8768, "step": 910 }, { "epoch": 1.2329045362220719, "grad_norm": 1.1617487541356466, "learning_rate": 7.950261121685642e-06, "loss": 2.1828, "step": 911 }, { "epoch": 1.2342586323628977, "grad_norm": 1.1384341697301983, "learning_rate": 7.945904302247968e-06, "loss": 2.3497, "step": 912 }, { "epoch": 1.2356127285037237, "grad_norm": 1.0784446459158834, "learning_rate": 7.941544054062065e-06, "loss": 2.2683, "step": 913 }, { "epoch": 1.2369668246445498, "grad_norm": 1.038286976843538, "learning_rate": 7.937180382202836e-06, "loss": 2.2668, "step": 914 }, { "epoch": 1.2383209207853758, "grad_norm": 1.0029564115724932, "learning_rate": 7.932813291749177e-06, "loss": 2.0261, "step": 915 }, { "epoch": 1.2396750169262019, "grad_norm": 1.0333541710726606, "learning_rate": 7.928442787783958e-06, "loss": 2.2212, "step": 916 }, { "epoch": 1.2410291130670277, "grad_norm": 1.03717078747682, "learning_rate": 7.924068875394024e-06, "loss": 2.0631, "step": 917 }, { "epoch": 1.2423832092078537, "grad_norm": 0.9341937079507571, "learning_rate": 7.919691559670188e-06, "loss": 2.0421, "step": 918 }, { "epoch": 1.2437373053486798, "grad_norm": 1.1684039509398019, "learning_rate": 7.915310845707223e-06, "loss": 2.0931, "step": 919 }, { "epoch": 1.2450914014895058, "grad_norm": 0.9994376608594767, "learning_rate": 7.910926738603855e-06, "loss": 2.6169, "step": 920 }, { "epoch": 1.2464454976303316, "grad_norm": 1.3781109917383612, "learning_rate": 7.906539243462763e-06, "loss": 2.8901, "step": 921 }, { "epoch": 1.2477995937711577, "grad_norm": 1.1519867240702086, "learning_rate": 7.902148365390567e-06, "loss": 2.3781, "step": 922 }, { "epoch": 1.2491536899119837, "grad_norm": 1.2921377062772406, "learning_rate": 7.897754109497826e-06, "loss": 2.4671, "step": 923 }, { "epoch": 1.2505077860528098, "grad_norm": 1.0832508824791298, "learning_rate": 7.89335648089903e-06, "loss": 2.3355, "step": 924 }, { "epoch": 1.2518618821936358, "grad_norm": 1.156819119690463, "learning_rate": 7.888955484712592e-06, "loss": 2.6164, "step": 925 }, { "epoch": 1.2532159783344619, "grad_norm": 1.0295051950872909, "learning_rate": 7.88455112606085e-06, "loss": 1.9817, "step": 926 }, { "epoch": 1.2545700744752877, "grad_norm": 0.9867097354242217, "learning_rate": 7.880143410070051e-06, "loss": 1.9863, "step": 927 }, { "epoch": 1.2559241706161137, "grad_norm": 1.1107662231325113, "learning_rate": 7.875732341870349e-06, "loss": 2.0024, "step": 928 }, { "epoch": 1.2572782667569398, "grad_norm": 0.9832002800266002, "learning_rate": 7.871317926595804e-06, "loss": 2.1088, "step": 929 }, { "epoch": 1.2586323628977658, "grad_norm": 0.9422090734118035, "learning_rate": 7.866900169384368e-06, "loss": 2.4123, "step": 930 }, { "epoch": 1.2599864590385916, "grad_norm": 1.0997264185351083, "learning_rate": 7.862479075377884e-06, "loss": 2.2158, "step": 931 }, { "epoch": 1.2613405551794177, "grad_norm": 1.1188615261512294, "learning_rate": 7.85805464972208e-06, "loss": 2.3005, "step": 932 }, { "epoch": 1.2626946513202437, "grad_norm": 1.0299930093784284, "learning_rate": 7.85362689756656e-06, "loss": 2.2651, "step": 933 }, { "epoch": 1.2640487474610698, "grad_norm": 0.9937695521744077, "learning_rate": 7.849195824064797e-06, "loss": 2.2419, "step": 934 }, { "epoch": 1.2654028436018958, "grad_norm": 1.0175704885653492, "learning_rate": 7.844761434374138e-06, "loss": 2.2974, "step": 935 }, { "epoch": 1.2667569397427219, "grad_norm": 1.1723623677027686, "learning_rate": 7.84032373365578e-06, "loss": 2.0875, "step": 936 }, { "epoch": 1.2681110358835477, "grad_norm": 1.2528863590890928, "learning_rate": 7.835882727074779e-06, "loss": 2.1546, "step": 937 }, { "epoch": 1.2694651320243737, "grad_norm": 1.3110354157785997, "learning_rate": 7.831438419800042e-06, "loss": 2.1202, "step": 938 }, { "epoch": 1.2708192281651998, "grad_norm": 1.0304392874864414, "learning_rate": 7.826990817004307e-06, "loss": 2.1668, "step": 939 }, { "epoch": 1.2721733243060258, "grad_norm": 1.0116042176431888, "learning_rate": 7.82253992386416e-06, "loss": 2.2464, "step": 940 }, { "epoch": 1.2735274204468516, "grad_norm": 0.9962700643735357, "learning_rate": 7.818085745560009e-06, "loss": 2.2486, "step": 941 }, { "epoch": 1.2748815165876777, "grad_norm": 1.7303323063471459, "learning_rate": 7.813628287276087e-06, "loss": 2.2405, "step": 942 }, { "epoch": 1.2762356127285037, "grad_norm": 1.0526531855129573, "learning_rate": 7.809167554200446e-06, "loss": 2.4385, "step": 943 }, { "epoch": 1.2775897088693298, "grad_norm": 0.9488843491039597, "learning_rate": 7.804703551524948e-06, "loss": 2.0235, "step": 944 }, { "epoch": 1.2789438050101558, "grad_norm": 1.0271145865465217, "learning_rate": 7.800236284445262e-06, "loss": 1.9179, "step": 945 }, { "epoch": 1.2802979011509819, "grad_norm": 1.0978694506737823, "learning_rate": 7.795765758160855e-06, "loss": 2.6674, "step": 946 }, { "epoch": 1.2816519972918077, "grad_norm": 0.9144971550877312, "learning_rate": 7.791291977874989e-06, "loss": 2.1267, "step": 947 }, { "epoch": 1.2830060934326337, "grad_norm": 1.0316859906377736, "learning_rate": 7.78681494879471e-06, "loss": 2.2941, "step": 948 }, { "epoch": 1.2843601895734598, "grad_norm": 1.2313895009261298, "learning_rate": 7.78233467613085e-06, "loss": 2.1893, "step": 949 }, { "epoch": 1.2857142857142856, "grad_norm": 1.158158610918687, "learning_rate": 7.777851165098012e-06, "loss": 1.8157, "step": 950 }, { "epoch": 1.2870683818551116, "grad_norm": 1.0349810102864025, "learning_rate": 7.773364420914572e-06, "loss": 2.4442, "step": 951 }, { "epoch": 1.2884224779959377, "grad_norm": 1.1215212355798856, "learning_rate": 7.768874448802665e-06, "loss": 2.1174, "step": 952 }, { "epoch": 1.2897765741367637, "grad_norm": 0.9965099934668875, "learning_rate": 7.764381253988189e-06, "loss": 2.5512, "step": 953 }, { "epoch": 1.2911306702775898, "grad_norm": 1.104679475403435, "learning_rate": 7.759884841700785e-06, "loss": 2.0502, "step": 954 }, { "epoch": 1.2924847664184158, "grad_norm": 1.660982051919665, "learning_rate": 7.755385217173845e-06, "loss": 2.4249, "step": 955 }, { "epoch": 1.2938388625592416, "grad_norm": 1.0755785420417305, "learning_rate": 7.750882385644495e-06, "loss": 2.3396, "step": 956 }, { "epoch": 1.2951929587000677, "grad_norm": 1.0925433109915934, "learning_rate": 7.746376352353599e-06, "loss": 2.249, "step": 957 }, { "epoch": 1.2965470548408937, "grad_norm": 0.9542648024212192, "learning_rate": 7.741867122545746e-06, "loss": 2.2574, "step": 958 }, { "epoch": 1.2979011509817198, "grad_norm": 1.1252306171349105, "learning_rate": 7.73735470146924e-06, "loss": 2.1923, "step": 959 }, { "epoch": 1.2992552471225456, "grad_norm": 1.0754233683086076, "learning_rate": 7.732839094376106e-06, "loss": 2.0116, "step": 960 }, { "epoch": 1.3006093432633716, "grad_norm": 1.1306744896780743, "learning_rate": 7.728320306522074e-06, "loss": 2.3124, "step": 961 }, { "epoch": 1.3019634394041977, "grad_norm": 1.072223164700719, "learning_rate": 7.723798343166578e-06, "loss": 2.0477, "step": 962 }, { "epoch": 1.3033175355450237, "grad_norm": 1.107055290773056, "learning_rate": 7.719273209572745e-06, "loss": 2.029, "step": 963 }, { "epoch": 1.3046716316858498, "grad_norm": 1.0396566849531226, "learning_rate": 7.714744911007395e-06, "loss": 2.1424, "step": 964 }, { "epoch": 1.3060257278266758, "grad_norm": 1.1282128940500473, "learning_rate": 7.710213452741028e-06, "loss": 2.3644, "step": 965 }, { "epoch": 1.3073798239675016, "grad_norm": 1.0660837047209504, "learning_rate": 7.70567884004783e-06, "loss": 2.3568, "step": 966 }, { "epoch": 1.3087339201083277, "grad_norm": 1.0098446863045805, "learning_rate": 7.701141078205648e-06, "loss": 2.2343, "step": 967 }, { "epoch": 1.3100880162491537, "grad_norm": 9.794773052164214, "learning_rate": 7.696600172495997e-06, "loss": 2.7452, "step": 968 }, { "epoch": 1.3114421123899798, "grad_norm": 1.3318037186474474, "learning_rate": 7.692056128204056e-06, "loss": 1.8471, "step": 969 }, { "epoch": 1.3127962085308056, "grad_norm": 1.151203349471892, "learning_rate": 7.687508950618654e-06, "loss": 2.6144, "step": 970 }, { "epoch": 1.3141503046716316, "grad_norm": 1.0959293878977527, "learning_rate": 7.682958645032265e-06, "loss": 2.22, "step": 971 }, { "epoch": 1.3155044008124577, "grad_norm": 1.0050631329121955, "learning_rate": 7.678405216741008e-06, "loss": 2.4293, "step": 972 }, { "epoch": 1.3168584969532837, "grad_norm": 1.0613207475015503, "learning_rate": 7.67384867104463e-06, "loss": 1.9871, "step": 973 }, { "epoch": 1.3182125930941098, "grad_norm": 1.1967438926432024, "learning_rate": 7.669289013246512e-06, "loss": 2.085, "step": 974 }, { "epoch": 1.3195666892349358, "grad_norm": 1.2081442687520356, "learning_rate": 7.664726248653658e-06, "loss": 2.1026, "step": 975 }, { "epoch": 1.3209207853757616, "grad_norm": 1.117288478243434, "learning_rate": 7.660160382576683e-06, "loss": 2.274, "step": 976 }, { "epoch": 1.3222748815165877, "grad_norm": 1.146272553819504, "learning_rate": 7.655591420329816e-06, "loss": 2.3935, "step": 977 }, { "epoch": 1.3236289776574137, "grad_norm": 1.658633099303849, "learning_rate": 7.651019367230886e-06, "loss": 2.0573, "step": 978 }, { "epoch": 1.3249830737982395, "grad_norm": 1.4532457972038215, "learning_rate": 7.646444228601321e-06, "loss": 2.0818, "step": 979 }, { "epoch": 1.3263371699390656, "grad_norm": 0.9729335641505342, "learning_rate": 7.641866009766145e-06, "loss": 2.1297, "step": 980 }, { "epoch": 1.3276912660798916, "grad_norm": 1.0010902233959411, "learning_rate": 7.637284716053959e-06, "loss": 1.9135, "step": 981 }, { "epoch": 1.3290453622207177, "grad_norm": 1.1644542386857344, "learning_rate": 7.63270035279695e-06, "loss": 2.0995, "step": 982 }, { "epoch": 1.3303994583615437, "grad_norm": 1.17699731802077, "learning_rate": 7.628112925330867e-06, "loss": 2.3395, "step": 983 }, { "epoch": 1.3317535545023698, "grad_norm": 1.9239341747345802, "learning_rate": 7.62352243899504e-06, "loss": 2.2552, "step": 984 }, { "epoch": 1.3331076506431956, "grad_norm": 1.104296574087454, "learning_rate": 7.6189288991323505e-06, "loss": 2.12, "step": 985 }, { "epoch": 1.3344617467840216, "grad_norm": 1.1563231154836133, "learning_rate": 7.614332311089234e-06, "loss": 2.4421, "step": 986 }, { "epoch": 1.3358158429248477, "grad_norm": 1.1224897324302643, "learning_rate": 7.609732680215676e-06, "loss": 2.5472, "step": 987 }, { "epoch": 1.3371699390656737, "grad_norm": 1.0798698576819643, "learning_rate": 7.605130011865201e-06, "loss": 1.9996, "step": 988 }, { "epoch": 1.3385240352064995, "grad_norm": 1.1427247534168647, "learning_rate": 7.600524311394873e-06, "loss": 2.4247, "step": 989 }, { "epoch": 1.3398781313473256, "grad_norm": 1.1436870392125549, "learning_rate": 7.595915584165283e-06, "loss": 2.2447, "step": 990 }, { "epoch": 1.3412322274881516, "grad_norm": 1.0123472207553952, "learning_rate": 7.591303835540544e-06, "loss": 2.5301, "step": 991 }, { "epoch": 1.3425863236289777, "grad_norm": 1.2308969840782176, "learning_rate": 7.586689070888284e-06, "loss": 2.4497, "step": 992 }, { "epoch": 1.3439404197698037, "grad_norm": 1.2549119339933823, "learning_rate": 7.582071295579647e-06, "loss": 2.4945, "step": 993 }, { "epoch": 1.3452945159106298, "grad_norm": 1.1551978030637418, "learning_rate": 7.577450514989275e-06, "loss": 1.7533, "step": 994 }, { "epoch": 1.3466486120514556, "grad_norm": 0.8995145500568735, "learning_rate": 7.572826734495313e-06, "loss": 2.0904, "step": 995 }, { "epoch": 1.3480027081922816, "grad_norm": 1.0456560935840085, "learning_rate": 7.568199959479393e-06, "loss": 2.3237, "step": 996 }, { "epoch": 1.3493568043331077, "grad_norm": 1.1047521602796329, "learning_rate": 7.563570195326635e-06, "loss": 2.2112, "step": 997 }, { "epoch": 1.3507109004739337, "grad_norm": 1.0312134506506243, "learning_rate": 7.558937447425638e-06, "loss": 1.9692, "step": 998 }, { "epoch": 1.3520649966147595, "grad_norm": 1.114956621453039, "learning_rate": 7.5543017211684745e-06, "loss": 2.5477, "step": 999 }, { "epoch": 1.3534190927555856, "grad_norm": 1.0810952077309708, "learning_rate": 7.5496630219506805e-06, "loss": 2.5975, "step": 1000 }, { "epoch": 1.3547731888964116, "grad_norm": 1.0751552768256476, "learning_rate": 7.545021355171253e-06, "loss": 2.4863, "step": 1001 }, { "epoch": 1.3561272850372377, "grad_norm": 1.1285166971489493, "learning_rate": 7.540376726232648e-06, "loss": 2.1198, "step": 1002 }, { "epoch": 1.3574813811780637, "grad_norm": 1.3021965465577348, "learning_rate": 7.535729140540762e-06, "loss": 2.5985, "step": 1003 }, { "epoch": 1.3588354773188898, "grad_norm": 1.1282925324454272, "learning_rate": 7.531078603504938e-06, "loss": 2.4989, "step": 1004 }, { "epoch": 1.3601895734597156, "grad_norm": 1.1221756439884651, "learning_rate": 7.52642512053795e-06, "loss": 2.2481, "step": 1005 }, { "epoch": 1.3615436696005416, "grad_norm": 1.057396172107965, "learning_rate": 7.521768697056004e-06, "loss": 2.181, "step": 1006 }, { "epoch": 1.3628977657413677, "grad_norm": 1.0186231893170619, "learning_rate": 7.517109338478729e-06, "loss": 2.1148, "step": 1007 }, { "epoch": 1.3642518618821937, "grad_norm": 0.996322134726628, "learning_rate": 7.512447050229166e-06, "loss": 2.0567, "step": 1008 }, { "epoch": 1.3656059580230195, "grad_norm": 1.0628848521017902, "learning_rate": 7.507781837733771e-06, "loss": 2.0321, "step": 1009 }, { "epoch": 1.3669600541638456, "grad_norm": 1.2154305322289178, "learning_rate": 7.503113706422398e-06, "loss": 2.6109, "step": 1010 }, { "epoch": 1.3683141503046716, "grad_norm": 1.2116773535315877, "learning_rate": 7.498442661728305e-06, "loss": 2.2339, "step": 1011 }, { "epoch": 1.3696682464454977, "grad_norm": 1.0098078280775096, "learning_rate": 7.493768709088133e-06, "loss": 2.9048, "step": 1012 }, { "epoch": 1.3710223425863237, "grad_norm": 0.9136677144647644, "learning_rate": 7.489091853941914e-06, "loss": 1.5463, "step": 1013 }, { "epoch": 1.3723764387271498, "grad_norm": 1.075788414941271, "learning_rate": 7.4844121017330575e-06, "loss": 2.0282, "step": 1014 }, { "epoch": 1.3737305348679756, "grad_norm": 0.9532034628399626, "learning_rate": 7.4797294579083405e-06, "loss": 2.0474, "step": 1015 }, { "epoch": 1.3750846310088016, "grad_norm": 1.0957243116565307, "learning_rate": 7.475043927917908e-06, "loss": 2.1565, "step": 1016 }, { "epoch": 1.3764387271496277, "grad_norm": 1.0904772853908138, "learning_rate": 7.470355517215267e-06, "loss": 2.3966, "step": 1017 }, { "epoch": 1.3777928232904535, "grad_norm": 1.3101240056799253, "learning_rate": 7.4656642312572734e-06, "loss": 2.1686, "step": 1018 }, { "epoch": 1.3791469194312795, "grad_norm": 1.7456227009220926, "learning_rate": 7.460970075504132e-06, "loss": 2.2421, "step": 1019 }, { "epoch": 1.3805010155721056, "grad_norm": 1.1602149134117044, "learning_rate": 7.4562730554193875e-06, "loss": 2.5002, "step": 1020 }, { "epoch": 1.3818551117129316, "grad_norm": 1.292010568667381, "learning_rate": 7.451573176469916e-06, "loss": 2.4289, "step": 1021 }, { "epoch": 1.3832092078537577, "grad_norm": 1.135185428169095, "learning_rate": 7.446870444125926e-06, "loss": 2.5667, "step": 1022 }, { "epoch": 1.3845633039945837, "grad_norm": 1.0225658545428182, "learning_rate": 7.442164863860939e-06, "loss": 2.0494, "step": 1023 }, { "epoch": 1.3859174001354095, "grad_norm": 1.057697531883615, "learning_rate": 7.4374564411518e-06, "loss": 1.9321, "step": 1024 }, { "epoch": 1.3872714962762356, "grad_norm": 0.9672332524393876, "learning_rate": 7.432745181478659e-06, "loss": 2.3441, "step": 1025 }, { "epoch": 1.3886255924170616, "grad_norm": 1.1227113972766776, "learning_rate": 7.428031090324966e-06, "loss": 3.3247, "step": 1026 }, { "epoch": 1.3899796885578877, "grad_norm": 1.158636683545892, "learning_rate": 7.423314173177467e-06, "loss": 2.4228, "step": 1027 }, { "epoch": 1.3913337846987135, "grad_norm": 1.050811409742003, "learning_rate": 7.4185944355261996e-06, "loss": 2.2984, "step": 1028 }, { "epoch": 1.3926878808395395, "grad_norm": 1.130180934126097, "learning_rate": 7.413871882864483e-06, "loss": 2.0937, "step": 1029 }, { "epoch": 1.3940419769803656, "grad_norm": 1.1323994964140216, "learning_rate": 7.4091465206889115e-06, "loss": 1.8611, "step": 1030 }, { "epoch": 1.3953960731211916, "grad_norm": 1.1272097248462323, "learning_rate": 7.404418354499352e-06, "loss": 2.3052, "step": 1031 }, { "epoch": 1.3967501692620177, "grad_norm": 1.0357729628233883, "learning_rate": 7.399687389798933e-06, "loss": 1.8565, "step": 1032 }, { "epoch": 1.3981042654028437, "grad_norm": 1.0211169637817514, "learning_rate": 7.39495363209404e-06, "loss": 2.1472, "step": 1033 }, { "epoch": 1.3994583615436695, "grad_norm": 1.0575088136735646, "learning_rate": 7.390217086894309e-06, "loss": 1.925, "step": 1034 }, { "epoch": 1.4008124576844956, "grad_norm": 1.153744768232039, "learning_rate": 7.385477759712625e-06, "loss": 2.6485, "step": 1035 }, { "epoch": 1.4021665538253216, "grad_norm": 1.1427109111587352, "learning_rate": 7.380735656065103e-06, "loss": 2.0497, "step": 1036 }, { "epoch": 1.4035206499661477, "grad_norm": 1.1900700043389405, "learning_rate": 7.375990781471098e-06, "loss": 2.3127, "step": 1037 }, { "epoch": 1.4048747461069735, "grad_norm": 1.1628543548305619, "learning_rate": 7.371243141453185e-06, "loss": 2.0708, "step": 1038 }, { "epoch": 1.4062288422477995, "grad_norm": 1.1557498507007162, "learning_rate": 7.366492741537156e-06, "loss": 2.0788, "step": 1039 }, { "epoch": 1.4075829383886256, "grad_norm": 1.0074214592668214, "learning_rate": 7.361739587252019e-06, "loss": 2.2388, "step": 1040 }, { "epoch": 1.4089370345294516, "grad_norm": 0.9850303377628948, "learning_rate": 7.3569836841299905e-06, "loss": 2.3422, "step": 1041 }, { "epoch": 1.4102911306702777, "grad_norm": 0.9362862055958489, "learning_rate": 7.352225037706476e-06, "loss": 1.9123, "step": 1042 }, { "epoch": 1.4116452268111037, "grad_norm": 1.0225754131079985, "learning_rate": 7.347463653520086e-06, "loss": 2.1162, "step": 1043 }, { "epoch": 1.4129993229519295, "grad_norm": 1.1046568630708087, "learning_rate": 7.34269953711261e-06, "loss": 2.744, "step": 1044 }, { "epoch": 1.4143534190927556, "grad_norm": 1.0769083803347521, "learning_rate": 7.337932694029018e-06, "loss": 2.1198, "step": 1045 }, { "epoch": 1.4157075152335816, "grad_norm": 0.9800626580167843, "learning_rate": 7.3331631298174596e-06, "loss": 2.3196, "step": 1046 }, { "epoch": 1.4170616113744074, "grad_norm": 1.0549618215499794, "learning_rate": 7.3283908500292434e-06, "loss": 2.1395, "step": 1047 }, { "epoch": 1.4184157075152335, "grad_norm": 1.060574227209818, "learning_rate": 7.323615860218844e-06, "loss": 2.8704, "step": 1048 }, { "epoch": 1.4197698036560595, "grad_norm": 1.0781456745775502, "learning_rate": 7.3188381659438865e-06, "loss": 2.2806, "step": 1049 }, { "epoch": 1.4211238997968856, "grad_norm": 1.551162578778134, "learning_rate": 7.314057772765148e-06, "loss": 2.116, "step": 1050 }, { "epoch": 1.4224779959377116, "grad_norm": 1.3756005806034488, "learning_rate": 7.309274686246547e-06, "loss": 3.2393, "step": 1051 }, { "epoch": 1.4238320920785377, "grad_norm": 1.0635108173294194, "learning_rate": 7.304488911955129e-06, "loss": 2.2153, "step": 1052 }, { "epoch": 1.4251861882193635, "grad_norm": 1.1523212588317262, "learning_rate": 7.299700455461076e-06, "loss": 1.8101, "step": 1053 }, { "epoch": 1.4265402843601895, "grad_norm": 1.1351217638799846, "learning_rate": 7.294909322337689e-06, "loss": 2.459, "step": 1054 }, { "epoch": 1.4278943805010156, "grad_norm": 1.113321093812555, "learning_rate": 7.290115518161385e-06, "loss": 1.9065, "step": 1055 }, { "epoch": 1.4292484766418416, "grad_norm": 1.1651609776327243, "learning_rate": 7.28531904851169e-06, "loss": 2.2955, "step": 1056 }, { "epoch": 1.4306025727826674, "grad_norm": 0.9391837640644604, "learning_rate": 7.280519918971228e-06, "loss": 2.5399, "step": 1057 }, { "epoch": 1.4319566689234935, "grad_norm": 1.1481164415158314, "learning_rate": 7.275718135125726e-06, "loss": 2.3987, "step": 1058 }, { "epoch": 1.4333107650643195, "grad_norm": 0.9819604190905024, "learning_rate": 7.270913702563995e-06, "loss": 2.0484, "step": 1059 }, { "epoch": 1.4346648612051456, "grad_norm": 1.241655388978992, "learning_rate": 7.266106626877933e-06, "loss": 2.1864, "step": 1060 }, { "epoch": 1.4360189573459716, "grad_norm": 1.0531564756435357, "learning_rate": 7.26129691366251e-06, "loss": 2.1386, "step": 1061 }, { "epoch": 1.4373730534867977, "grad_norm": 1.1218122300065656, "learning_rate": 7.256484568515769e-06, "loss": 2.4484, "step": 1062 }, { "epoch": 1.4387271496276235, "grad_norm": 1.0001136323518225, "learning_rate": 7.251669597038814e-06, "loss": 2.2824, "step": 1063 }, { "epoch": 1.4400812457684495, "grad_norm": 1.2896307069641786, "learning_rate": 7.246852004835807e-06, "loss": 2.4539, "step": 1064 }, { "epoch": 1.4414353419092756, "grad_norm": 1.0510167874878298, "learning_rate": 7.242031797513962e-06, "loss": 2.6035, "step": 1065 }, { "epoch": 1.4427894380501016, "grad_norm": 1.2155395455915952, "learning_rate": 7.2372089806835335e-06, "loss": 1.9087, "step": 1066 }, { "epoch": 1.4441435341909274, "grad_norm": 1.1199078701755178, "learning_rate": 7.232383559957815e-06, "loss": 2.1759, "step": 1067 }, { "epoch": 1.4454976303317535, "grad_norm": 1.159792012820624, "learning_rate": 7.227555540953131e-06, "loss": 2.4314, "step": 1068 }, { "epoch": 1.4468517264725795, "grad_norm": 1.0730576607550595, "learning_rate": 7.22272492928883e-06, "loss": 2.3359, "step": 1069 }, { "epoch": 1.4482058226134056, "grad_norm": 0.8222048650029028, "learning_rate": 7.217891730587276e-06, "loss": 2.2215, "step": 1070 }, { "epoch": 1.4495599187542316, "grad_norm": 1.0285016809620628, "learning_rate": 7.2130559504738464e-06, "loss": 2.239, "step": 1071 }, { "epoch": 1.4509140148950577, "grad_norm": 1.0693497499216194, "learning_rate": 7.2082175945769226e-06, "loss": 2.2506, "step": 1072 }, { "epoch": 1.4522681110358835, "grad_norm": 1.1167979192807023, "learning_rate": 7.203376668527884e-06, "loss": 2.2712, "step": 1073 }, { "epoch": 1.4536222071767095, "grad_norm": 1.1307443613901718, "learning_rate": 7.198533177961102e-06, "loss": 2.0017, "step": 1074 }, { "epoch": 1.4549763033175356, "grad_norm": 1.0564264494899962, "learning_rate": 7.19368712851393e-06, "loss": 2.1056, "step": 1075 }, { "epoch": 1.4563303994583616, "grad_norm": 1.1101530483846054, "learning_rate": 7.188838525826702e-06, "loss": 2.1903, "step": 1076 }, { "epoch": 1.4576844955991874, "grad_norm": 1.0103673641252706, "learning_rate": 7.183987375542726e-06, "loss": 2.382, "step": 1077 }, { "epoch": 1.4590385917400135, "grad_norm": 1.253353521477901, "learning_rate": 7.17913368330827e-06, "loss": 2.0939, "step": 1078 }, { "epoch": 1.4603926878808395, "grad_norm": 1.2508634701536092, "learning_rate": 7.174277454772566e-06, "loss": 1.9324, "step": 1079 }, { "epoch": 1.4617467840216656, "grad_norm": 8.122448301970948, "learning_rate": 7.169418695587791e-06, "loss": 2.1723, "step": 1080 }, { "epoch": 1.4631008801624916, "grad_norm": 1.3165876873084112, "learning_rate": 7.164557411409075e-06, "loss": 2.0454, "step": 1081 }, { "epoch": 1.4644549763033177, "grad_norm": 1.3433865812537615, "learning_rate": 7.1596936078944825e-06, "loss": 1.5743, "step": 1082 }, { "epoch": 1.4658090724441435, "grad_norm": 1.1708156611040208, "learning_rate": 7.154827290705012e-06, "loss": 2.7672, "step": 1083 }, { "epoch": 1.4671631685849695, "grad_norm": 1.0739066488240625, "learning_rate": 7.1499584655045866e-06, "loss": 2.0953, "step": 1084 }, { "epoch": 1.4685172647257956, "grad_norm": 1.0810850204219578, "learning_rate": 7.145087137960047e-06, "loss": 2.6921, "step": 1085 }, { "epoch": 1.4698713608666214, "grad_norm": 0.9939405297389377, "learning_rate": 7.14021331374115e-06, "loss": 2.0588, "step": 1086 }, { "epoch": 1.4712254570074474, "grad_norm": 0.9226185898131262, "learning_rate": 7.135336998520557e-06, "loss": 1.9489, "step": 1087 }, { "epoch": 1.4725795531482735, "grad_norm": 1.0783455175484244, "learning_rate": 7.130458197973828e-06, "loss": 1.8894, "step": 1088 }, { "epoch": 1.4739336492890995, "grad_norm": 1.0986213365125383, "learning_rate": 7.125576917779414e-06, "loss": 2.3981, "step": 1089 }, { "epoch": 1.4752877454299256, "grad_norm": 0.9907037990576693, "learning_rate": 7.120693163618656e-06, "loss": 1.9927, "step": 1090 }, { "epoch": 1.4766418415707516, "grad_norm": 1.067381598945353, "learning_rate": 7.115806941175771e-06, "loss": 1.9368, "step": 1091 }, { "epoch": 1.4779959377115774, "grad_norm": 1.00952047194142, "learning_rate": 7.1109182561378505e-06, "loss": 2.1691, "step": 1092 }, { "epoch": 1.4793500338524035, "grad_norm": 1.107779870373698, "learning_rate": 7.106027114194856e-06, "loss": 1.8634, "step": 1093 }, { "epoch": 1.4807041299932295, "grad_norm": 1.1243485392179005, "learning_rate": 7.101133521039597e-06, "loss": 1.9522, "step": 1094 }, { "epoch": 1.4820582261340556, "grad_norm": 1.0572503465978025, "learning_rate": 7.09623748236775e-06, "loss": 1.8363, "step": 1095 }, { "epoch": 1.4834123222748814, "grad_norm": 1.0216562069583985, "learning_rate": 7.091339003877826e-06, "loss": 1.8552, "step": 1096 }, { "epoch": 1.4847664184157074, "grad_norm": 1.089956344974216, "learning_rate": 7.086438091271186e-06, "loss": 2.5221, "step": 1097 }, { "epoch": 1.4861205145565335, "grad_norm": 1.11408247722042, "learning_rate": 7.0815347502520185e-06, "loss": 2.2585, "step": 1098 }, { "epoch": 1.4874746106973595, "grad_norm": 0.9710069262269226, "learning_rate": 7.076628986527335e-06, "loss": 2.3226, "step": 1099 }, { "epoch": 1.4888287068381856, "grad_norm": 1.0917833338553558, "learning_rate": 7.0717208058069755e-06, "loss": 1.9066, "step": 1100 }, { "epoch": 1.4901828029790116, "grad_norm": 1.1161624453210208, "learning_rate": 7.066810213803586e-06, "loss": 2.1559, "step": 1101 }, { "epoch": 1.4915368991198374, "grad_norm": 1.16178506916721, "learning_rate": 7.061897216232622e-06, "loss": 2.3097, "step": 1102 }, { "epoch": 1.4928909952606635, "grad_norm": 1.0384328202163011, "learning_rate": 7.05698181881234e-06, "loss": 1.938, "step": 1103 }, { "epoch": 1.4942450914014895, "grad_norm": 1.216771459951468, "learning_rate": 7.052064027263785e-06, "loss": 2.4431, "step": 1104 }, { "epoch": 1.4955991875423156, "grad_norm": 1.3088163468638734, "learning_rate": 7.047143847310794e-06, "loss": 2.0962, "step": 1105 }, { "epoch": 1.4969532836831414, "grad_norm": 1.1011041050321808, "learning_rate": 7.042221284679982e-06, "loss": 1.9349, "step": 1106 }, { "epoch": 1.4983073798239674, "grad_norm": 1.2350163332118804, "learning_rate": 7.037296345100734e-06, "loss": 1.8952, "step": 1107 }, { "epoch": 1.4996614759647935, "grad_norm": 1.0028965498141413, "learning_rate": 7.032369034305209e-06, "loss": 2.4324, "step": 1108 }, { "epoch": 1.5010155721056195, "grad_norm": 1.0240481324400763, "learning_rate": 7.027439358028315e-06, "loss": 2.4041, "step": 1109 }, { "epoch": 1.5023696682464456, "grad_norm": 1.2289451784867027, "learning_rate": 7.022507322007723e-06, "loss": 2.296, "step": 1110 }, { "epoch": 1.5037237643872716, "grad_norm": 1.1846045097053997, "learning_rate": 7.017572931983846e-06, "loss": 1.9807, "step": 1111 }, { "epoch": 1.5050778605280974, "grad_norm": 1.3679366822185766, "learning_rate": 7.012636193699838e-06, "loss": 2.3165, "step": 1112 }, { "epoch": 1.5064319566689235, "grad_norm": 1.0735093643875895, "learning_rate": 7.007697112901586e-06, "loss": 2.0637, "step": 1113 }, { "epoch": 1.5077860528097495, "grad_norm": 0.9805492134387425, "learning_rate": 7.002755695337703e-06, "loss": 2.2236, "step": 1114 }, { "epoch": 1.5091401489505754, "grad_norm": 1.1540910897325873, "learning_rate": 6.997811946759522e-06, "loss": 2.5638, "step": 1115 }, { "epoch": 1.5104942450914014, "grad_norm": 1.045564752122656, "learning_rate": 6.9928658729210885e-06, "loss": 2.5629, "step": 1116 }, { "epoch": 1.5118483412322274, "grad_norm": 1.5327917333436447, "learning_rate": 6.987917479579156e-06, "loss": 2.3958, "step": 1117 }, { "epoch": 1.5132024373730535, "grad_norm": 1.0061412022965626, "learning_rate": 6.982966772493176e-06, "loss": 2.3869, "step": 1118 }, { "epoch": 1.5145565335138795, "grad_norm": 1.143443700093151, "learning_rate": 6.978013757425295e-06, "loss": 2.2255, "step": 1119 }, { "epoch": 1.5159106296547056, "grad_norm": 1.076563769412944, "learning_rate": 6.973058440140341e-06, "loss": 2.0117, "step": 1120 }, { "epoch": 1.5172647257955316, "grad_norm": 0.9932180957280397, "learning_rate": 6.96810082640583e-06, "loss": 2.4315, "step": 1121 }, { "epoch": 1.5186188219363574, "grad_norm": 1.7614630897566435, "learning_rate": 6.963140921991941e-06, "loss": 1.8154, "step": 1122 }, { "epoch": 1.5199729180771835, "grad_norm": 1.0435661710662918, "learning_rate": 6.958178732671527e-06, "loss": 2.1161, "step": 1123 }, { "epoch": 1.5213270142180095, "grad_norm": 1.034141937866288, "learning_rate": 6.953214264220095e-06, "loss": 2.2153, "step": 1124 }, { "epoch": 1.5226811103588354, "grad_norm": 1.1333203921847073, "learning_rate": 6.948247522415811e-06, "loss": 2.1679, "step": 1125 }, { "epoch": 1.5240352064996614, "grad_norm": 1.1523266319281296, "learning_rate": 6.943278513039477e-06, "loss": 2.0316, "step": 1126 }, { "epoch": 1.5253893026404874, "grad_norm": 1.1175733614955354, "learning_rate": 6.938307241874547e-06, "loss": 1.9331, "step": 1127 }, { "epoch": 1.5267433987813135, "grad_norm": 1.0816050768685381, "learning_rate": 6.933333714707094e-06, "loss": 2.1337, "step": 1128 }, { "epoch": 1.5280974949221395, "grad_norm": 0.9555568730660431, "learning_rate": 6.928357937325829e-06, "loss": 2.1897, "step": 1129 }, { "epoch": 1.5294515910629656, "grad_norm": 1.1330718777414148, "learning_rate": 6.923379915522075e-06, "loss": 2.5013, "step": 1130 }, { "epoch": 1.5308056872037916, "grad_norm": 1.4860432492374922, "learning_rate": 6.918399655089768e-06, "loss": 2.2661, "step": 1131 }, { "epoch": 1.5321597833446174, "grad_norm": 1.0608341510421102, "learning_rate": 6.913417161825449e-06, "loss": 2.6459, "step": 1132 }, { "epoch": 1.5335138794854435, "grad_norm": 1.2535263483596926, "learning_rate": 6.908432441528262e-06, "loss": 2.5183, "step": 1133 }, { "epoch": 1.5348679756262693, "grad_norm": 1.1881324730918719, "learning_rate": 6.903445499999939e-06, "loss": 1.9834, "step": 1134 }, { "epoch": 1.5362220717670954, "grad_norm": 0.9936992242467968, "learning_rate": 6.8984563430448e-06, "loss": 2.185, "step": 1135 }, { "epoch": 1.5375761679079214, "grad_norm": 0.9856859764312084, "learning_rate": 6.893464976469739e-06, "loss": 2.413, "step": 1136 }, { "epoch": 1.5389302640487474, "grad_norm": 0.9728057769285352, "learning_rate": 6.888471406084227e-06, "loss": 2.2291, "step": 1137 }, { "epoch": 1.5402843601895735, "grad_norm": 1.9592487450519758, "learning_rate": 6.883475637700298e-06, "loss": 2.2332, "step": 1138 }, { "epoch": 1.5416384563303995, "grad_norm": 2.989310700259001, "learning_rate": 6.8784776771325426e-06, "loss": 2.5173, "step": 1139 }, { "epoch": 1.5429925524712256, "grad_norm": 1.0753052749288403, "learning_rate": 6.873477530198107e-06, "loss": 2.2734, "step": 1140 }, { "epoch": 1.5443466486120516, "grad_norm": 1.1348199790348215, "learning_rate": 6.868475202716677e-06, "loss": 3.0435, "step": 1141 }, { "epoch": 1.5457007447528774, "grad_norm": 0.9253248443470898, "learning_rate": 6.863470700510479e-06, "loss": 2.2282, "step": 1142 }, { "epoch": 1.5470548408937035, "grad_norm": 1.0212577294991076, "learning_rate": 6.858464029404272e-06, "loss": 1.9986, "step": 1143 }, { "epoch": 1.5484089370345293, "grad_norm": 1.0119821954044654, "learning_rate": 6.8534551952253395e-06, "loss": 2.236, "step": 1144 }, { "epoch": 1.5497630331753554, "grad_norm": 0.9376658731281714, "learning_rate": 6.848444203803476e-06, "loss": 1.9798, "step": 1145 }, { "epoch": 1.5511171293161814, "grad_norm": 1.0518223950174321, "learning_rate": 6.843431060970995e-06, "loss": 2.2833, "step": 1146 }, { "epoch": 1.5524712254570074, "grad_norm": 1.6688430719575653, "learning_rate": 6.838415772562711e-06, "loss": 2.2911, "step": 1147 }, { "epoch": 1.5538253215978335, "grad_norm": 2.211386740791148, "learning_rate": 6.833398344415933e-06, "loss": 2.6092, "step": 1148 }, { "epoch": 1.5551794177386595, "grad_norm": 1.302197054959585, "learning_rate": 6.8283787823704685e-06, "loss": 2.1764, "step": 1149 }, { "epoch": 1.5565335138794856, "grad_norm": 1.365025659237606, "learning_rate": 6.823357092268596e-06, "loss": 1.9187, "step": 1150 }, { "epoch": 1.5578876100203114, "grad_norm": 0.9995629551418774, "learning_rate": 6.8183332799550836e-06, "loss": 2.0254, "step": 1151 }, { "epoch": 1.5592417061611374, "grad_norm": 1.091795853304648, "learning_rate": 6.813307351277161e-06, "loss": 2.5035, "step": 1152 }, { "epoch": 1.5605958023019635, "grad_norm": 0.9113959325465628, "learning_rate": 6.808279312084525e-06, "loss": 2.2859, "step": 1153 }, { "epoch": 1.5619498984427893, "grad_norm": 1.105743242638601, "learning_rate": 6.803249168229329e-06, "loss": 1.9762, "step": 1154 }, { "epoch": 1.5633039945836154, "grad_norm": 1.3089749040803054, "learning_rate": 6.798216925566171e-06, "loss": 2.245, "step": 1155 }, { "epoch": 1.5646580907244414, "grad_norm": 1.4155040937471588, "learning_rate": 6.7931825899521e-06, "loss": 2.151, "step": 1156 }, { "epoch": 1.5660121868652674, "grad_norm": 1.2355148875524264, "learning_rate": 6.788146167246594e-06, "loss": 2.2848, "step": 1157 }, { "epoch": 1.5673662830060935, "grad_norm": 1.134385408399196, "learning_rate": 6.783107663311566e-06, "loss": 1.9886, "step": 1158 }, { "epoch": 1.5687203791469195, "grad_norm": 1.1445134576197042, "learning_rate": 6.7780670840113434e-06, "loss": 2.2995, "step": 1159 }, { "epoch": 1.5700744752877456, "grad_norm": 1.1766480093443303, "learning_rate": 6.773024435212678e-06, "loss": 2.2219, "step": 1160 }, { "epoch": 1.5714285714285714, "grad_norm": 1.5615268999323522, "learning_rate": 6.767979722784725e-06, "loss": 2.0472, "step": 1161 }, { "epoch": 1.5727826675693974, "grad_norm": 1.1706173436180667, "learning_rate": 6.762932952599043e-06, "loss": 2.4683, "step": 1162 }, { "epoch": 1.5741367637102233, "grad_norm": 1.1480375058434151, "learning_rate": 6.757884130529584e-06, "loss": 1.9804, "step": 1163 }, { "epoch": 1.5754908598510493, "grad_norm": 1.1079469797078965, "learning_rate": 6.75283326245269e-06, "loss": 2.3322, "step": 1164 }, { "epoch": 1.5768449559918754, "grad_norm": 1.1362279207408492, "learning_rate": 6.7477803542470834e-06, "loss": 1.998, "step": 1165 }, { "epoch": 1.5781990521327014, "grad_norm": 0.982450102859919, "learning_rate": 6.742725411793862e-06, "loss": 2.3529, "step": 1166 }, { "epoch": 1.5795531482735274, "grad_norm": 1.1119736244356198, "learning_rate": 6.737668440976494e-06, "loss": 2.0098, "step": 1167 }, { "epoch": 1.5809072444143535, "grad_norm": 1.1735566109179396, "learning_rate": 6.7326094476808e-06, "loss": 2.4483, "step": 1168 }, { "epoch": 1.5822613405551795, "grad_norm": 0.9600432883498903, "learning_rate": 6.727548437794963e-06, "loss": 2.4039, "step": 1169 }, { "epoch": 1.5836154366960056, "grad_norm": 0.9713832736002586, "learning_rate": 6.722485417209509e-06, "loss": 2.129, "step": 1170 }, { "epoch": 1.5849695328368314, "grad_norm": 1.536541973608301, "learning_rate": 6.717420391817306e-06, "loss": 2.0486, "step": 1171 }, { "epoch": 1.5863236289776574, "grad_norm": 1.1215408259768265, "learning_rate": 6.712353367513555e-06, "loss": 1.8101, "step": 1172 }, { "epoch": 1.5876777251184833, "grad_norm": 0.9400933392391618, "learning_rate": 6.707284350195779e-06, "loss": 2.129, "step": 1173 }, { "epoch": 1.5890318212593093, "grad_norm": 1.0955863112973159, "learning_rate": 6.70221334576383e-06, "loss": 1.8608, "step": 1174 }, { "epoch": 1.5903859174001354, "grad_norm": 1.290340133465167, "learning_rate": 6.697140360119867e-06, "loss": 2.6089, "step": 1175 }, { "epoch": 1.5917400135409614, "grad_norm": 1.0935001366917838, "learning_rate": 6.692065399168352e-06, "loss": 2.3835, "step": 1176 }, { "epoch": 1.5930941096817874, "grad_norm": 1.1310710971297633, "learning_rate": 6.686988468816055e-06, "loss": 2.3186, "step": 1177 }, { "epoch": 1.5944482058226135, "grad_norm": 1.1106261511672113, "learning_rate": 6.681909574972028e-06, "loss": 2.278, "step": 1178 }, { "epoch": 1.5958023019634395, "grad_norm": 1.1099983154470332, "learning_rate": 6.676828723547614e-06, "loss": 1.8412, "step": 1179 }, { "epoch": 1.5971563981042654, "grad_norm": 1.115352426872196, "learning_rate": 6.671745920456434e-06, "loss": 2.6848, "step": 1180 }, { "epoch": 1.5985104942450914, "grad_norm": 1.9617683740846232, "learning_rate": 6.666661171614382e-06, "loss": 2.4141, "step": 1181 }, { "epoch": 1.5998645903859174, "grad_norm": 1.0340361767560753, "learning_rate": 6.661574482939615e-06, "loss": 2.029, "step": 1182 }, { "epoch": 1.6012186865267433, "grad_norm": 1.1104672220605092, "learning_rate": 6.656485860352544e-06, "loss": 2.3655, "step": 1183 }, { "epoch": 1.6025727826675693, "grad_norm": 1.2030765915104757, "learning_rate": 6.651395309775837e-06, "loss": 2.1954, "step": 1184 }, { "epoch": 1.6039268788083954, "grad_norm": 1.2573381048562495, "learning_rate": 6.6463028371344015e-06, "loss": 2.5574, "step": 1185 }, { "epoch": 1.6052809749492214, "grad_norm": 1.1722502953655352, "learning_rate": 6.641208448355388e-06, "loss": 1.9912, "step": 1186 }, { "epoch": 1.6066350710900474, "grad_norm": 0.9237524987059714, "learning_rate": 6.63611214936817e-06, "loss": 1.7042, "step": 1187 }, { "epoch": 1.6079891672308735, "grad_norm": 1.0114862987523483, "learning_rate": 6.631013946104348e-06, "loss": 2.1457, "step": 1188 }, { "epoch": 1.6093432633716995, "grad_norm": 1.2504620286352166, "learning_rate": 6.625913844497739e-06, "loss": 2.9242, "step": 1189 }, { "epoch": 1.6106973595125254, "grad_norm": 1.0716280700456458, "learning_rate": 6.620811850484368e-06, "loss": 2.1747, "step": 1190 }, { "epoch": 1.6120514556533514, "grad_norm": 1.0218159941143157, "learning_rate": 6.6157079700024665e-06, "loss": 2.1052, "step": 1191 }, { "epoch": 1.6134055517941774, "grad_norm": 0.9964735852419978, "learning_rate": 6.6106022089924535e-06, "loss": 2.1085, "step": 1192 }, { "epoch": 1.6147596479350033, "grad_norm": 1.1164408166829398, "learning_rate": 6.605494573396947e-06, "loss": 2.146, "step": 1193 }, { "epoch": 1.6161137440758293, "grad_norm": 1.1424753368875373, "learning_rate": 6.600385069160739e-06, "loss": 2.8387, "step": 1194 }, { "epoch": 1.6174678402166554, "grad_norm": 1.3178474061043657, "learning_rate": 6.5952737022308e-06, "loss": 1.902, "step": 1195 }, { "epoch": 1.6188219363574814, "grad_norm": 1.10966169321405, "learning_rate": 6.590160478556269e-06, "loss": 2.4412, "step": 1196 }, { "epoch": 1.6201760324983074, "grad_norm": 1.1587333892126894, "learning_rate": 6.585045404088442e-06, "loss": 2.2305, "step": 1197 }, { "epoch": 1.6215301286391335, "grad_norm": 1.0092191905987942, "learning_rate": 6.579928484780773e-06, "loss": 2.1454, "step": 1198 }, { "epoch": 1.6228842247799595, "grad_norm": 1.0732773418214976, "learning_rate": 6.5748097265888624e-06, "loss": 2.3336, "step": 1199 }, { "epoch": 1.6242383209207854, "grad_norm": 1.174781988020797, "learning_rate": 6.569689135470451e-06, "loss": 2.6344, "step": 1200 }, { "epoch": 1.6255924170616114, "grad_norm": 1.1510963746609166, "learning_rate": 6.564566717385412e-06, "loss": 2.2199, "step": 1201 }, { "epoch": 1.6269465132024372, "grad_norm": 0.8890344864420157, "learning_rate": 6.559442478295745e-06, "loss": 1.9674, "step": 1202 }, { "epoch": 1.6283006093432633, "grad_norm": 1.0562702319805766, "learning_rate": 6.5543164241655724e-06, "loss": 2.0231, "step": 1203 }, { "epoch": 1.6296547054840893, "grad_norm": 1.1686330096691793, "learning_rate": 6.549188560961124e-06, "loss": 2.401, "step": 1204 }, { "epoch": 1.6310088016249153, "grad_norm": 0.9835702255574785, "learning_rate": 6.544058894650736e-06, "loss": 1.9566, "step": 1205 }, { "epoch": 1.6323628977657414, "grad_norm": 1.1799470524015878, "learning_rate": 6.5389274312048476e-06, "loss": 2.3821, "step": 1206 }, { "epoch": 1.6337169939065674, "grad_norm": 1.0305646424878727, "learning_rate": 6.533794176595987e-06, "loss": 1.7532, "step": 1207 }, { "epoch": 1.6350710900473935, "grad_norm": 0.9411916034993967, "learning_rate": 6.5286591367987655e-06, "loss": 2.1668, "step": 1208 }, { "epoch": 1.6364251861882195, "grad_norm": 1.1619947231820393, "learning_rate": 6.523522317789874e-06, "loss": 2.0694, "step": 1209 }, { "epoch": 1.6377792823290453, "grad_norm": 1.1054945884631686, "learning_rate": 6.518383725548074e-06, "loss": 2.1855, "step": 1210 }, { "epoch": 1.6391333784698714, "grad_norm": 1.0751426497234493, "learning_rate": 6.513243366054191e-06, "loss": 2.2038, "step": 1211 }, { "epoch": 1.6404874746106972, "grad_norm": 1.0718361928319553, "learning_rate": 6.508101245291109e-06, "loss": 2.2788, "step": 1212 }, { "epoch": 1.6418415707515233, "grad_norm": 1.1446897139146628, "learning_rate": 6.502957369243757e-06, "loss": 1.8886, "step": 1213 }, { "epoch": 1.6431956668923493, "grad_norm": 1.0780227955508757, "learning_rate": 6.497811743899112e-06, "loss": 2.9019, "step": 1214 }, { "epoch": 1.6445497630331753, "grad_norm": 1.1093588985005516, "learning_rate": 6.492664375246185e-06, "loss": 2.3378, "step": 1215 }, { "epoch": 1.6459038591740014, "grad_norm": 1.2680934914767075, "learning_rate": 6.487515269276015e-06, "loss": 2.2777, "step": 1216 }, { "epoch": 1.6472579553148274, "grad_norm": 1.0496344613229611, "learning_rate": 6.482364431981667e-06, "loss": 2.1846, "step": 1217 }, { "epoch": 1.6486120514556535, "grad_norm": 0.8830408833414931, "learning_rate": 6.4772118693582155e-06, "loss": 2.4639, "step": 1218 }, { "epoch": 1.6499661475964793, "grad_norm": 1.3617840820836493, "learning_rate": 6.472057587402748e-06, "loss": 2.1004, "step": 1219 }, { "epoch": 1.6513202437373053, "grad_norm": 1.0634770023039701, "learning_rate": 6.46690159211435e-06, "loss": 2.3582, "step": 1220 }, { "epoch": 1.6526743398781314, "grad_norm": 1.0462993743834461, "learning_rate": 6.461743889494103e-06, "loss": 2.1793, "step": 1221 }, { "epoch": 1.6540284360189572, "grad_norm": 1.12270603128975, "learning_rate": 6.456584485545075e-06, "loss": 2.0312, "step": 1222 }, { "epoch": 1.6553825321597833, "grad_norm": 1.0474924934857333, "learning_rate": 6.451423386272312e-06, "loss": 2.1907, "step": 1223 }, { "epoch": 1.6567366283006093, "grad_norm": 1.2139372873126406, "learning_rate": 6.4462605976828395e-06, "loss": 2.1359, "step": 1224 }, { "epoch": 1.6580907244414353, "grad_norm": 1.2375074680264349, "learning_rate": 6.441096125785641e-06, "loss": 2.7886, "step": 1225 }, { "epoch": 1.6594448205822614, "grad_norm": 1.2726063282105258, "learning_rate": 6.435929976591665e-06, "loss": 1.8611, "step": 1226 }, { "epoch": 1.6607989167230874, "grad_norm": 1.0325742896045942, "learning_rate": 6.4307621561138114e-06, "loss": 2.1808, "step": 1227 }, { "epoch": 1.6621530128639135, "grad_norm": 1.1451127258616212, "learning_rate": 6.425592670366923e-06, "loss": 2.3795, "step": 1228 }, { "epoch": 1.6635071090047393, "grad_norm": 1.0588080148815049, "learning_rate": 6.420421525367783e-06, "loss": 2.3691, "step": 1229 }, { "epoch": 1.6648612051455653, "grad_norm": 1.1248420963082115, "learning_rate": 6.415248727135103e-06, "loss": 2.2255, "step": 1230 }, { "epoch": 1.6662153012863912, "grad_norm": 1.12193324651476, "learning_rate": 6.410074281689522e-06, "loss": 2.2692, "step": 1231 }, { "epoch": 1.6675693974272172, "grad_norm": 1.1007368907096031, "learning_rate": 6.4048981950535975e-06, "loss": 2.0518, "step": 1232 }, { "epoch": 1.6689234935680433, "grad_norm": 1.06269170578331, "learning_rate": 6.3997204732517915e-06, "loss": 2.3531, "step": 1233 }, { "epoch": 1.6702775897088693, "grad_norm": 1.1163412698869442, "learning_rate": 6.394541122310473e-06, "loss": 2.2524, "step": 1234 }, { "epoch": 1.6716316858496953, "grad_norm": 1.0554457718352965, "learning_rate": 6.389360148257908e-06, "loss": 2.4751, "step": 1235 }, { "epoch": 1.6729857819905214, "grad_norm": 1.1997008894725945, "learning_rate": 6.384177557124247e-06, "loss": 1.9252, "step": 1236 }, { "epoch": 1.6743398781313474, "grad_norm": 1.079055031664645, "learning_rate": 6.378993354941529e-06, "loss": 2.3641, "step": 1237 }, { "epoch": 1.6756939742721735, "grad_norm": 1.1896072037787495, "learning_rate": 6.3738075477436655e-06, "loss": 1.9261, "step": 1238 }, { "epoch": 1.6770480704129993, "grad_norm": 0.9256492202916144, "learning_rate": 6.368620141566432e-06, "loss": 2.3076, "step": 1239 }, { "epoch": 1.6784021665538253, "grad_norm": 1.1122465104394996, "learning_rate": 6.363431142447469e-06, "loss": 2.0237, "step": 1240 }, { "epoch": 1.6797562626946512, "grad_norm": 0.9678398341378385, "learning_rate": 6.358240556426274e-06, "loss": 2.4974, "step": 1241 }, { "epoch": 1.6811103588354772, "grad_norm": 0.9925297077996342, "learning_rate": 6.353048389544185e-06, "loss": 2.244, "step": 1242 }, { "epoch": 1.6824644549763033, "grad_norm": 1.0616822439065718, "learning_rate": 6.3478546478443844e-06, "loss": 2.3992, "step": 1243 }, { "epoch": 1.6838185511171293, "grad_norm": 1.1733624673460672, "learning_rate": 6.342659337371884e-06, "loss": 1.8391, "step": 1244 }, { "epoch": 1.6851726472579553, "grad_norm": 1.2040137843680947, "learning_rate": 6.3374624641735275e-06, "loss": 1.8877, "step": 1245 }, { "epoch": 1.6865267433987814, "grad_norm": 1.0453714976813977, "learning_rate": 6.33226403429797e-06, "loss": 2.0552, "step": 1246 }, { "epoch": 1.6878808395396074, "grad_norm": 1.1426720858647532, "learning_rate": 6.327064053795684e-06, "loss": 2.2107, "step": 1247 }, { "epoch": 1.6892349356804333, "grad_norm": 1.199062064451871, "learning_rate": 6.321862528718945e-06, "loss": 2.0693, "step": 1248 }, { "epoch": 1.6905890318212593, "grad_norm": 1.091725236727943, "learning_rate": 6.3166594651218235e-06, "loss": 2.1291, "step": 1249 }, { "epoch": 1.6919431279620853, "grad_norm": 1.1275607912830679, "learning_rate": 6.311454869060188e-06, "loss": 2.2288, "step": 1250 }, { "epoch": 1.6932972241029112, "grad_norm": 1.236386694185598, "learning_rate": 6.3062487465916825e-06, "loss": 2.4416, "step": 1251 }, { "epoch": 1.6946513202437372, "grad_norm": 0.9827244912709826, "learning_rate": 6.301041103775731e-06, "loss": 2.3147, "step": 1252 }, { "epoch": 1.6960054163845633, "grad_norm": 1.2144425166077264, "learning_rate": 6.295831946673527e-06, "loss": 2.4805, "step": 1253 }, { "epoch": 1.6973595125253893, "grad_norm": 0.9983619215236832, "learning_rate": 6.290621281348028e-06, "loss": 2.3773, "step": 1254 }, { "epoch": 1.6987136086662153, "grad_norm": 1.465782148598409, "learning_rate": 6.285409113863944e-06, "loss": 2.6457, "step": 1255 }, { "epoch": 1.7000677048070414, "grad_norm": 0.9944582045815571, "learning_rate": 6.280195450287736e-06, "loss": 2.3837, "step": 1256 }, { "epoch": 1.7014218009478674, "grad_norm": 1.1585939861872792, "learning_rate": 6.274980296687605e-06, "loss": 2.2119, "step": 1257 }, { "epoch": 1.7027758970886933, "grad_norm": 1.1437189700305217, "learning_rate": 6.269763659133486e-06, "loss": 1.8079, "step": 1258 }, { "epoch": 1.7041299932295193, "grad_norm": 1.2582747800473197, "learning_rate": 6.264545543697039e-06, "loss": 2.5012, "step": 1259 }, { "epoch": 1.7054840893703453, "grad_norm": 1.2978378068690752, "learning_rate": 6.259325956451651e-06, "loss": 2.9485, "step": 1260 }, { "epoch": 1.7068381855111712, "grad_norm": 1.0424842976922624, "learning_rate": 6.254104903472417e-06, "loss": 2.2624, "step": 1261 }, { "epoch": 1.7081922816519972, "grad_norm": 1.0868588376583306, "learning_rate": 6.248882390836135e-06, "loss": 2.3471, "step": 1262 }, { "epoch": 1.7095463777928233, "grad_norm": 1.0151211458884728, "learning_rate": 6.243658424621308e-06, "loss": 1.9188, "step": 1263 }, { "epoch": 1.7109004739336493, "grad_norm": 1.1267719574629707, "learning_rate": 6.238433010908131e-06, "loss": 2.0653, "step": 1264 }, { "epoch": 1.7122545700744753, "grad_norm": 1.1710384477346152, "learning_rate": 6.233206155778476e-06, "loss": 2.8079, "step": 1265 }, { "epoch": 1.7136086662153014, "grad_norm": 1.2399841383381511, "learning_rate": 6.227977865315902e-06, "loss": 2.3833, "step": 1266 }, { "epoch": 1.7149627623561274, "grad_norm": 1.0116588623292786, "learning_rate": 6.222748145605631e-06, "loss": 2.4164, "step": 1267 }, { "epoch": 1.7163168584969533, "grad_norm": 0.8527451840083117, "learning_rate": 6.217517002734554e-06, "loss": 2.1877, "step": 1268 }, { "epoch": 1.7176709546377793, "grad_norm": 1.079554004708601, "learning_rate": 6.212284442791217e-06, "loss": 1.7948, "step": 1269 }, { "epoch": 1.7190250507786051, "grad_norm": 0.9137877079458239, "learning_rate": 6.207050471865814e-06, "loss": 1.841, "step": 1270 }, { "epoch": 1.7203791469194312, "grad_norm": 1.1473597440593684, "learning_rate": 6.201815096050181e-06, "loss": 1.9544, "step": 1271 }, { "epoch": 1.7217332430602572, "grad_norm": 1.189027252854009, "learning_rate": 6.1965783214377895e-06, "loss": 2.2543, "step": 1272 }, { "epoch": 1.7230873392010833, "grad_norm": 0.8466476995754849, "learning_rate": 6.1913401541237414e-06, "loss": 1.8128, "step": 1273 }, { "epoch": 1.7244414353419093, "grad_norm": 1.320497220055551, "learning_rate": 6.186100600204757e-06, "loss": 1.9606, "step": 1274 }, { "epoch": 1.7257955314827353, "grad_norm": 1.1503966715251264, "learning_rate": 6.180859665779173e-06, "loss": 2.0067, "step": 1275 }, { "epoch": 1.7271496276235614, "grad_norm": 1.197795743633595, "learning_rate": 6.175617356946928e-06, "loss": 2.4943, "step": 1276 }, { "epoch": 1.7285037237643874, "grad_norm": 1.153983741344922, "learning_rate": 6.170373679809565e-06, "loss": 2.6468, "step": 1277 }, { "epoch": 1.7298578199052133, "grad_norm": 1.1521006556182283, "learning_rate": 6.165128640470217e-06, "loss": 2.3347, "step": 1278 }, { "epoch": 1.7312119160460393, "grad_norm": 0.9816413381160185, "learning_rate": 6.159882245033606e-06, "loss": 2.4498, "step": 1279 }, { "epoch": 1.7325660121868651, "grad_norm": 1.07736534022899, "learning_rate": 6.1546344996060294e-06, "loss": 2.1578, "step": 1280 }, { "epoch": 1.7339201083276912, "grad_norm": 1.1324233134707422, "learning_rate": 6.149385410295353e-06, "loss": 2.1432, "step": 1281 }, { "epoch": 1.7352742044685172, "grad_norm": 0.9905290656285861, "learning_rate": 6.144134983211012e-06, "loss": 1.872, "step": 1282 }, { "epoch": 1.7366283006093433, "grad_norm": 1.121201508813207, "learning_rate": 6.138883224463999e-06, "loss": 1.9648, "step": 1283 }, { "epoch": 1.7379823967501693, "grad_norm": 1.1862628290117698, "learning_rate": 6.13363014016685e-06, "loss": 2.1744, "step": 1284 }, { "epoch": 1.7393364928909953, "grad_norm": 0.9830015960619232, "learning_rate": 6.128375736433652e-06, "loss": 2.5532, "step": 1285 }, { "epoch": 1.7406905890318214, "grad_norm": 1.3321125394117272, "learning_rate": 6.123120019380021e-06, "loss": 1.8057, "step": 1286 }, { "epoch": 1.7420446851726472, "grad_norm": 0.9980976929176649, "learning_rate": 6.117862995123105e-06, "loss": 2.337, "step": 1287 }, { "epoch": 1.7433987813134733, "grad_norm": 1.3173652394689053, "learning_rate": 6.112604669781572e-06, "loss": 2.0408, "step": 1288 }, { "epoch": 1.7447528774542993, "grad_norm": 1.126084316928292, "learning_rate": 6.107345049475609e-06, "loss": 2.079, "step": 1289 }, { "epoch": 1.7461069735951251, "grad_norm": 1.0413819596877105, "learning_rate": 6.102084140326901e-06, "loss": 2.0356, "step": 1290 }, { "epoch": 1.7474610697359512, "grad_norm": 1.189218565313232, "learning_rate": 6.0968219484586415e-06, "loss": 2.1497, "step": 1291 }, { "epoch": 1.7488151658767772, "grad_norm": 1.2345681545643294, "learning_rate": 6.091558479995512e-06, "loss": 2.0314, "step": 1292 }, { "epoch": 1.7501692620176033, "grad_norm": 1.2541579573695316, "learning_rate": 6.086293741063685e-06, "loss": 2.2602, "step": 1293 }, { "epoch": 1.7515233581584293, "grad_norm": 1.4163932516550155, "learning_rate": 6.081027737790804e-06, "loss": 2.649, "step": 1294 }, { "epoch": 1.7528774542992553, "grad_norm": 1.0677445213609345, "learning_rate": 6.075760476305992e-06, "loss": 1.6769, "step": 1295 }, { "epoch": 1.7542315504400814, "grad_norm": 1.1305409453258632, "learning_rate": 6.070491962739831e-06, "loss": 2.1413, "step": 1296 }, { "epoch": 1.7555856465809072, "grad_norm": 0.8589085134430466, "learning_rate": 6.065222203224363e-06, "loss": 2.2007, "step": 1297 }, { "epoch": 1.7569397427217333, "grad_norm": 1.198177616657121, "learning_rate": 6.05995120389308e-06, "loss": 2.463, "step": 1298 }, { "epoch": 1.758293838862559, "grad_norm": 1.0517431847975949, "learning_rate": 6.054678970880915e-06, "loss": 1.8492, "step": 1299 }, { "epoch": 1.7596479350033851, "grad_norm": 1.0877464246501793, "learning_rate": 6.049405510324237e-06, "loss": 2.2445, "step": 1300 }, { "epoch": 1.7610020311442112, "grad_norm": 0.9182819543184888, "learning_rate": 6.04413082836085e-06, "loss": 1.923, "step": 1301 }, { "epoch": 1.7623561272850372, "grad_norm": 0.9479159720049781, "learning_rate": 6.0388549311299696e-06, "loss": 1.8528, "step": 1302 }, { "epoch": 1.7637102234258633, "grad_norm": 1.1330012872888824, "learning_rate": 6.033577824772234e-06, "loss": 2.3405, "step": 1303 }, { "epoch": 1.7650643195666893, "grad_norm": 1.156479517032631, "learning_rate": 6.028299515429683e-06, "loss": 2.4133, "step": 1304 }, { "epoch": 1.7664184157075153, "grad_norm": 1.2849276017784446, "learning_rate": 6.023020009245761e-06, "loss": 2.0552, "step": 1305 }, { "epoch": 1.7677725118483414, "grad_norm": 1.1081263537914658, "learning_rate": 6.017739312365304e-06, "loss": 1.9643, "step": 1306 }, { "epoch": 1.7691266079891672, "grad_norm": 1.192468338343465, "learning_rate": 6.012457430934532e-06, "loss": 2.2097, "step": 1307 }, { "epoch": 1.7704807041299933, "grad_norm": 1.3083267149971682, "learning_rate": 6.007174371101045e-06, "loss": 2.0671, "step": 1308 }, { "epoch": 1.771834800270819, "grad_norm": 1.0164272321056016, "learning_rate": 6.001890139013816e-06, "loss": 2.1438, "step": 1309 }, { "epoch": 1.7731888964116451, "grad_norm": 1.0169184208499216, "learning_rate": 5.99660474082318e-06, "loss": 1.6461, "step": 1310 }, { "epoch": 1.7745429925524712, "grad_norm": 1.1208962249464, "learning_rate": 5.99131818268083e-06, "loss": 2.0891, "step": 1311 }, { "epoch": 1.7758970886932972, "grad_norm": 0.9880243308990688, "learning_rate": 5.986030470739811e-06, "loss": 2.2771, "step": 1312 }, { "epoch": 1.7772511848341233, "grad_norm": 1.0164372276741276, "learning_rate": 5.980741611154508e-06, "loss": 1.9089, "step": 1313 }, { "epoch": 1.7786052809749493, "grad_norm": 1.1780962664030357, "learning_rate": 5.975451610080643e-06, "loss": 2.2283, "step": 1314 }, { "epoch": 1.7799593771157753, "grad_norm": 1.2337816535300354, "learning_rate": 5.970160473675266e-06, "loss": 2.0219, "step": 1315 }, { "epoch": 1.7813134732566012, "grad_norm": 1.231311259477204, "learning_rate": 5.96486820809675e-06, "loss": 2.263, "step": 1316 }, { "epoch": 1.7826675693974272, "grad_norm": 1.1820145455048345, "learning_rate": 5.959574819504782e-06, "loss": 2.1722, "step": 1317 }, { "epoch": 1.7840216655382533, "grad_norm": 1.0602621258357094, "learning_rate": 5.954280314060353e-06, "loss": 2.0836, "step": 1318 }, { "epoch": 1.785375761679079, "grad_norm": 1.0769310348498187, "learning_rate": 5.9489846979257596e-06, "loss": 2.1107, "step": 1319 }, { "epoch": 1.7867298578199051, "grad_norm": 1.1685570897175144, "learning_rate": 5.943687977264584e-06, "loss": 2.3701, "step": 1320 }, { "epoch": 1.7880839539607312, "grad_norm": 1.0639504620411393, "learning_rate": 5.938390158241701e-06, "loss": 1.8519, "step": 1321 }, { "epoch": 1.7894380501015572, "grad_norm": 1.0471961016733244, "learning_rate": 5.933091247023258e-06, "loss": 2.3008, "step": 1322 }, { "epoch": 1.7907921462423833, "grad_norm": 1.079334192472969, "learning_rate": 5.927791249776678e-06, "loss": 1.9762, "step": 1323 }, { "epoch": 1.7921462423832093, "grad_norm": 1.0698384472116922, "learning_rate": 5.922490172670645e-06, "loss": 2.3354, "step": 1324 }, { "epoch": 1.7935003385240353, "grad_norm": 0.9830724595088538, "learning_rate": 5.917188021875101e-06, "loss": 2.0081, "step": 1325 }, { "epoch": 1.7948544346648612, "grad_norm": 1.340171291705083, "learning_rate": 5.91188480356124e-06, "loss": 2.1949, "step": 1326 }, { "epoch": 1.7962085308056872, "grad_norm": 1.0913921699387112, "learning_rate": 5.906580523901493e-06, "loss": 2.0597, "step": 1327 }, { "epoch": 1.7975626269465133, "grad_norm": 1.0873779097080578, "learning_rate": 5.90127518906953e-06, "loss": 2.8079, "step": 1328 }, { "epoch": 1.798916723087339, "grad_norm": 1.143892818041929, "learning_rate": 5.89596880524025e-06, "loss": 2.1394, "step": 1329 }, { "epoch": 1.8002708192281651, "grad_norm": 0.9347818074040029, "learning_rate": 5.89066137858977e-06, "loss": 2.0658, "step": 1330 }, { "epoch": 1.8016249153689912, "grad_norm": 1.0104844185810784, "learning_rate": 5.885352915295426e-06, "loss": 2.4573, "step": 1331 }, { "epoch": 1.8029790115098172, "grad_norm": 1.2621189889416795, "learning_rate": 5.880043421535751e-06, "loss": 2.2771, "step": 1332 }, { "epoch": 1.8043331076506433, "grad_norm": 0.9776230416310355, "learning_rate": 5.874732903490489e-06, "loss": 2.2122, "step": 1333 }, { "epoch": 1.8056872037914693, "grad_norm": 1.1935436794161547, "learning_rate": 5.869421367340565e-06, "loss": 2.6218, "step": 1334 }, { "epoch": 1.8070412999322953, "grad_norm": 1.166409774943222, "learning_rate": 5.864108819268098e-06, "loss": 2.1516, "step": 1335 }, { "epoch": 1.8083953960731212, "grad_norm": 0.9563193799285069, "learning_rate": 5.858795265456382e-06, "loss": 2.1492, "step": 1336 }, { "epoch": 1.8097494922139472, "grad_norm": 1.0173612877579, "learning_rate": 5.853480712089875e-06, "loss": 2.5045, "step": 1337 }, { "epoch": 1.811103588354773, "grad_norm": 1.075936558534605, "learning_rate": 5.8481651653542105e-06, "loss": 1.8214, "step": 1338 }, { "epoch": 1.812457684495599, "grad_norm": 1.2662166496893965, "learning_rate": 5.842848631436169e-06, "loss": 2.3341, "step": 1339 }, { "epoch": 1.8138117806364251, "grad_norm": 1.114110647827857, "learning_rate": 5.837531116523683e-06, "loss": 1.9848, "step": 1340 }, { "epoch": 1.8151658767772512, "grad_norm": 1.133467071972051, "learning_rate": 5.832212626805825e-06, "loss": 1.7728, "step": 1341 }, { "epoch": 1.8165199729180772, "grad_norm": 1.1904363405626495, "learning_rate": 5.826893168472807e-06, "loss": 2.2031, "step": 1342 }, { "epoch": 1.8178740690589033, "grad_norm": 1.16982781000476, "learning_rate": 5.821572747715961e-06, "loss": 2.2635, "step": 1343 }, { "epoch": 1.8192281651997293, "grad_norm": 2.3913374538342995, "learning_rate": 5.816251370727748e-06, "loss": 2.7674, "step": 1344 }, { "epoch": 1.8205822613405553, "grad_norm": 1.155071954814068, "learning_rate": 5.810929043701733e-06, "loss": 1.6222, "step": 1345 }, { "epoch": 1.8219363574813812, "grad_norm": 1.1195068129462349, "learning_rate": 5.805605772832592e-06, "loss": 2.2405, "step": 1346 }, { "epoch": 1.8232904536222072, "grad_norm": 1.0607883154488784, "learning_rate": 5.800281564316098e-06, "loss": 2.3167, "step": 1347 }, { "epoch": 1.824644549763033, "grad_norm": 0.9198180067594192, "learning_rate": 5.794956424349117e-06, "loss": 2.2837, "step": 1348 }, { "epoch": 1.825998645903859, "grad_norm": 1.1141503225834095, "learning_rate": 5.789630359129599e-06, "loss": 1.9332, "step": 1349 }, { "epoch": 1.8273527420446851, "grad_norm": 1.1494560282733117, "learning_rate": 5.784303374856566e-06, "loss": 2.0177, "step": 1350 }, { "epoch": 1.8287068381855112, "grad_norm": 0.9838628228119614, "learning_rate": 5.778975477730117e-06, "loss": 1.9256, "step": 1351 }, { "epoch": 1.8300609343263372, "grad_norm": 1.0379092917122728, "learning_rate": 5.773646673951406e-06, "loss": 2.0192, "step": 1352 }, { "epoch": 1.8314150304671633, "grad_norm": 1.0735535500692883, "learning_rate": 5.768316969722651e-06, "loss": 1.7708, "step": 1353 }, { "epoch": 1.8327691266079893, "grad_norm": 1.1525689914627582, "learning_rate": 5.762986371247111e-06, "loss": 2.0803, "step": 1354 }, { "epoch": 1.8341232227488151, "grad_norm": 1.2969233616374964, "learning_rate": 5.757654884729087e-06, "loss": 1.7907, "step": 1355 }, { "epoch": 1.8354773188896412, "grad_norm": 1.2729837968386013, "learning_rate": 5.752322516373916e-06, "loss": 1.9638, "step": 1356 }, { "epoch": 1.8368314150304672, "grad_norm": 1.123620270200479, "learning_rate": 5.746989272387959e-06, "loss": 2.4644, "step": 1357 }, { "epoch": 1.838185511171293, "grad_norm": 1.1478076870568643, "learning_rate": 5.741655158978598e-06, "loss": 2.3496, "step": 1358 }, { "epoch": 1.839539607312119, "grad_norm": 0.9606164841242715, "learning_rate": 5.736320182354227e-06, "loss": 2.3232, "step": 1359 }, { "epoch": 1.8408937034529451, "grad_norm": 0.9434545645826506, "learning_rate": 5.730984348724242e-06, "loss": 2.1405, "step": 1360 }, { "epoch": 1.8422477995937712, "grad_norm": 1.1057923006106751, "learning_rate": 5.725647664299039e-06, "loss": 2.9284, "step": 1361 }, { "epoch": 1.8436018957345972, "grad_norm": 1.0188797974653283, "learning_rate": 5.720310135290002e-06, "loss": 2.1123, "step": 1362 }, { "epoch": 1.8449559918754233, "grad_norm": 1.4800000718890725, "learning_rate": 5.7149717679095026e-06, "loss": 2.5262, "step": 1363 }, { "epoch": 1.8463100880162493, "grad_norm": 1.4201072710479334, "learning_rate": 5.709632568370884e-06, "loss": 1.9675, "step": 1364 }, { "epoch": 1.8476641841570751, "grad_norm": 1.0271104317135864, "learning_rate": 5.704292542888458e-06, "loss": 2.3992, "step": 1365 }, { "epoch": 1.8490182802979012, "grad_norm": 1.1894861709189837, "learning_rate": 5.698951697677498e-06, "loss": 2.3255, "step": 1366 }, { "epoch": 1.850372376438727, "grad_norm": 0.9818606413266471, "learning_rate": 5.6936100389542345e-06, "loss": 2.068, "step": 1367 }, { "epoch": 1.851726472579553, "grad_norm": 1.1613550136148025, "learning_rate": 5.688267572935843e-06, "loss": 2.0782, "step": 1368 }, { "epoch": 1.853080568720379, "grad_norm": 1.2604918956006983, "learning_rate": 5.682924305840435e-06, "loss": 2.3411, "step": 1369 }, { "epoch": 1.8544346648612051, "grad_norm": 1.0429238519002082, "learning_rate": 5.6775802438870596e-06, "loss": 2.3507, "step": 1370 }, { "epoch": 1.8557887610020312, "grad_norm": 1.1476866829926053, "learning_rate": 5.6722353932956895e-06, "loss": 2.1169, "step": 1371 }, { "epoch": 1.8571428571428572, "grad_norm": 1.2071680463300853, "learning_rate": 5.6668897602872145e-06, "loss": 2.3788, "step": 1372 }, { "epoch": 1.8584969532836833, "grad_norm": 1.330454708741957, "learning_rate": 5.661543351083435e-06, "loss": 2.3059, "step": 1373 }, { "epoch": 1.8598510494245093, "grad_norm": 1.0666387576576266, "learning_rate": 5.656196171907055e-06, "loss": 1.988, "step": 1374 }, { "epoch": 1.8612051455653351, "grad_norm": 1.1222282028435293, "learning_rate": 5.650848228981676e-06, "loss": 2.3535, "step": 1375 }, { "epoch": 1.8625592417061612, "grad_norm": 1.0691636063361554, "learning_rate": 5.645499528531785e-06, "loss": 2.3795, "step": 1376 }, { "epoch": 1.863913337846987, "grad_norm": 1.0416420607905428, "learning_rate": 5.640150076782755e-06, "loss": 2.2143, "step": 1377 }, { "epoch": 1.865267433987813, "grad_norm": 1.1126322928122026, "learning_rate": 5.634799879960833e-06, "loss": 1.8687, "step": 1378 }, { "epoch": 1.866621530128639, "grad_norm": 1.0283378507513847, "learning_rate": 5.629448944293128e-06, "loss": 2.3348, "step": 1379 }, { "epoch": 1.8679756262694651, "grad_norm": 1.1086979474304333, "learning_rate": 5.624097276007614e-06, "loss": 2.4503, "step": 1380 }, { "epoch": 1.8693297224102912, "grad_norm": 1.0114717774494804, "learning_rate": 5.618744881333117e-06, "loss": 2.2857, "step": 1381 }, { "epoch": 1.8706838185511172, "grad_norm": 1.271521217949957, "learning_rate": 5.613391766499308e-06, "loss": 2.1678, "step": 1382 }, { "epoch": 1.8720379146919433, "grad_norm": 1.0690493430782453, "learning_rate": 5.6080379377366936e-06, "loss": 2.0153, "step": 1383 }, { "epoch": 1.873392010832769, "grad_norm": 1.139241558729131, "learning_rate": 5.6026834012766155e-06, "loss": 2.6149, "step": 1384 }, { "epoch": 1.8747461069735951, "grad_norm": 1.0202843691038828, "learning_rate": 5.597328163351237e-06, "loss": 2.3195, "step": 1385 }, { "epoch": 1.8761002031144212, "grad_norm": 1.0587205093081162, "learning_rate": 5.59197223019354e-06, "loss": 2.369, "step": 1386 }, { "epoch": 1.877454299255247, "grad_norm": 1.0246205891822728, "learning_rate": 5.586615608037309e-06, "loss": 1.7189, "step": 1387 }, { "epoch": 1.878808395396073, "grad_norm": 1.2556625533663983, "learning_rate": 5.581258303117138e-06, "loss": 2.6877, "step": 1388 }, { "epoch": 1.880162491536899, "grad_norm": 1.0753392702185522, "learning_rate": 5.575900321668414e-06, "loss": 2.2089, "step": 1389 }, { "epoch": 1.8815165876777251, "grad_norm": 1.0364442011380954, "learning_rate": 5.5705416699273085e-06, "loss": 2.2061, "step": 1390 }, { "epoch": 1.8828706838185512, "grad_norm": 0.9721557814407294, "learning_rate": 5.565182354130776e-06, "loss": 2.0579, "step": 1391 }, { "epoch": 1.8842247799593772, "grad_norm": 1.0946729786176126, "learning_rate": 5.559822380516539e-06, "loss": 1.8052, "step": 1392 }, { "epoch": 1.8855788761002032, "grad_norm": 1.2816555939803111, "learning_rate": 5.554461755323095e-06, "loss": 2.5817, "step": 1393 }, { "epoch": 1.886932972241029, "grad_norm": 1.132710357655246, "learning_rate": 5.54910048478969e-06, "loss": 2.0912, "step": 1394 }, { "epoch": 1.8882870683818551, "grad_norm": 1.4036561544628778, "learning_rate": 5.5437385751563265e-06, "loss": 1.9642, "step": 1395 }, { "epoch": 1.8896411645226812, "grad_norm": 0.9338991067427596, "learning_rate": 5.5383760326637525e-06, "loss": 2.107, "step": 1396 }, { "epoch": 1.890995260663507, "grad_norm": 1.0328732328422332, "learning_rate": 5.533012863553445e-06, "loss": 2.0279, "step": 1397 }, { "epoch": 1.892349356804333, "grad_norm": 1.0913343945197256, "learning_rate": 5.527649074067618e-06, "loss": 2.037, "step": 1398 }, { "epoch": 1.893703452945159, "grad_norm": 1.1464521194384492, "learning_rate": 5.522284670449204e-06, "loss": 1.9054, "step": 1399 }, { "epoch": 1.8950575490859851, "grad_norm": 1.7305402471538387, "learning_rate": 5.51691965894185e-06, "loss": 2.1952, "step": 1400 }, { "epoch": 1.8964116452268112, "grad_norm": 0.9382393928786104, "learning_rate": 5.511554045789915e-06, "loss": 1.937, "step": 1401 }, { "epoch": 1.8977657413676372, "grad_norm": 1.042649655218192, "learning_rate": 5.506187837238451e-06, "loss": 1.9652, "step": 1402 }, { "epoch": 1.8991198375084632, "grad_norm": 1.1266190255444637, "learning_rate": 5.5008210395332095e-06, "loss": 2.5897, "step": 1403 }, { "epoch": 1.900473933649289, "grad_norm": 1.0568485706741606, "learning_rate": 5.495453658920622e-06, "loss": 2.1456, "step": 1404 }, { "epoch": 1.9018280297901151, "grad_norm": 2.3042054688617517, "learning_rate": 5.490085701647805e-06, "loss": 2.4831, "step": 1405 }, { "epoch": 1.903182125930941, "grad_norm": 1.0661084879580494, "learning_rate": 5.484717173962538e-06, "loss": 1.979, "step": 1406 }, { "epoch": 1.904536222071767, "grad_norm": 1.8624940514969899, "learning_rate": 5.479348082113273e-06, "loss": 2.2721, "step": 1407 }, { "epoch": 1.905890318212593, "grad_norm": 1.0349458998467256, "learning_rate": 5.4739784323491115e-06, "loss": 2.1855, "step": 1408 }, { "epoch": 1.907244414353419, "grad_norm": 1.0420374247356121, "learning_rate": 5.468608230919811e-06, "loss": 1.9857, "step": 1409 }, { "epoch": 1.9085985104942451, "grad_norm": 1.0571913625970193, "learning_rate": 5.463237484075765e-06, "loss": 1.9137, "step": 1410 }, { "epoch": 1.9099526066350712, "grad_norm": 1.148192788565203, "learning_rate": 5.457866198068006e-06, "loss": 1.9036, "step": 1411 }, { "epoch": 1.9113067027758972, "grad_norm": 0.9825967829135833, "learning_rate": 5.45249437914819e-06, "loss": 2.4171, "step": 1412 }, { "epoch": 1.9126607989167232, "grad_norm": 1.070670848522326, "learning_rate": 5.4471220335685985e-06, "loss": 2.2573, "step": 1413 }, { "epoch": 1.914014895057549, "grad_norm": 1.1613970345507367, "learning_rate": 5.44174916758212e-06, "loss": 2.1134, "step": 1414 }, { "epoch": 1.9153689911983751, "grad_norm": 0.9179865288261003, "learning_rate": 5.436375787442256e-06, "loss": 2.2131, "step": 1415 }, { "epoch": 1.916723087339201, "grad_norm": 1.276005671850693, "learning_rate": 5.4310018994030974e-06, "loss": 2.0346, "step": 1416 }, { "epoch": 1.918077183480027, "grad_norm": 1.1781979348827227, "learning_rate": 5.425627509719336e-06, "loss": 2.0975, "step": 1417 }, { "epoch": 1.919431279620853, "grad_norm": 1.1588736839015075, "learning_rate": 5.420252624646238e-06, "loss": 2.8176, "step": 1418 }, { "epoch": 1.920785375761679, "grad_norm": 1.0396625706559195, "learning_rate": 5.414877250439654e-06, "loss": 2.2559, "step": 1419 }, { "epoch": 1.9221394719025051, "grad_norm": 1.1691760754641374, "learning_rate": 5.409501393356001e-06, "loss": 2.134, "step": 1420 }, { "epoch": 1.9234935680433312, "grad_norm": 1.0188156446108358, "learning_rate": 5.404125059652255e-06, "loss": 3.1594, "step": 1421 }, { "epoch": 1.9248476641841572, "grad_norm": 1.1384064676805719, "learning_rate": 5.398748255585952e-06, "loss": 2.3359, "step": 1422 }, { "epoch": 1.926201760324983, "grad_norm": 1.0703628157280833, "learning_rate": 5.39337098741517e-06, "loss": 2.322, "step": 1423 }, { "epoch": 1.927555856465809, "grad_norm": 3.5972509087878834, "learning_rate": 5.387993261398532e-06, "loss": 2.0459, "step": 1424 }, { "epoch": 1.9289099526066351, "grad_norm": 1.1278074196782333, "learning_rate": 5.3826150837951925e-06, "loss": 2.1835, "step": 1425 }, { "epoch": 1.930264048747461, "grad_norm": 1.1684466435119685, "learning_rate": 5.3772364608648304e-06, "loss": 2.5017, "step": 1426 }, { "epoch": 1.931618144888287, "grad_norm": 1.1988587658843286, "learning_rate": 5.371857398867644e-06, "loss": 2.4385, "step": 1427 }, { "epoch": 1.932972241029113, "grad_norm": 0.943349610365291, "learning_rate": 5.36647790406434e-06, "loss": 2.1582, "step": 1428 }, { "epoch": 1.934326337169939, "grad_norm": 1.0186861984009383, "learning_rate": 5.361097982716133e-06, "loss": 2.0697, "step": 1429 }, { "epoch": 1.9356804333107651, "grad_norm": 1.0863559782982186, "learning_rate": 5.355717641084733e-06, "loss": 2.6047, "step": 1430 }, { "epoch": 1.9370345294515912, "grad_norm": 1.2753139571991683, "learning_rate": 5.350336885432337e-06, "loss": 2.5542, "step": 1431 }, { "epoch": 1.9383886255924172, "grad_norm": 1.2432570870357769, "learning_rate": 5.3449557220216245e-06, "loss": 2.465, "step": 1432 }, { "epoch": 1.939742721733243, "grad_norm": 1.1214426550882808, "learning_rate": 5.339574157115752e-06, "loss": 1.8317, "step": 1433 }, { "epoch": 1.941096817874069, "grad_norm": 1.1278145331998868, "learning_rate": 5.334192196978341e-06, "loss": 1.8582, "step": 1434 }, { "epoch": 1.942450914014895, "grad_norm": 0.9703824311730281, "learning_rate": 5.328809847873472e-06, "loss": 2.0282, "step": 1435 }, { "epoch": 1.943805010155721, "grad_norm": 1.1313991479024939, "learning_rate": 5.32342711606568e-06, "loss": 2.2019, "step": 1436 }, { "epoch": 1.945159106296547, "grad_norm": 1.1285464646294614, "learning_rate": 5.318044007819948e-06, "loss": 2.2181, "step": 1437 }, { "epoch": 1.946513202437373, "grad_norm": 1.2286900340042946, "learning_rate": 5.312660529401693e-06, "loss": 1.8868, "step": 1438 }, { "epoch": 1.947867298578199, "grad_norm": 1.0620287314695505, "learning_rate": 5.307276687076762e-06, "loss": 2.5188, "step": 1439 }, { "epoch": 1.9492213947190251, "grad_norm": 1.0392077445928851, "learning_rate": 5.301892487111431e-06, "loss": 2.0289, "step": 1440 }, { "epoch": 1.9505754908598512, "grad_norm": 1.3207900398189212, "learning_rate": 5.296507935772386e-06, "loss": 2.2526, "step": 1441 }, { "epoch": 1.9519295870006772, "grad_norm": 1.407831485125338, "learning_rate": 5.291123039326728e-06, "loss": 2.3924, "step": 1442 }, { "epoch": 1.953283683141503, "grad_norm": 1.092640851593957, "learning_rate": 5.285737804041955e-06, "loss": 2.2856, "step": 1443 }, { "epoch": 1.954637779282329, "grad_norm": 1.0720656311911092, "learning_rate": 5.2803522361859596e-06, "loss": 2.3745, "step": 1444 }, { "epoch": 1.955991875423155, "grad_norm": 1.0470209379216529, "learning_rate": 5.274966342027025e-06, "loss": 2.4369, "step": 1445 }, { "epoch": 1.957345971563981, "grad_norm": 1.7795808290406938, "learning_rate": 5.269580127833812e-06, "loss": 2.2505, "step": 1446 }, { "epoch": 1.958700067704807, "grad_norm": 0.949115783937015, "learning_rate": 5.264193599875353e-06, "loss": 2.1088, "step": 1447 }, { "epoch": 1.960054163845633, "grad_norm": 1.0307094217484138, "learning_rate": 5.258806764421048e-06, "loss": 2.0897, "step": 1448 }, { "epoch": 1.961408259986459, "grad_norm": 1.0944949048376986, "learning_rate": 5.253419627740651e-06, "loss": 2.2959, "step": 1449 }, { "epoch": 1.9627623561272851, "grad_norm": 1.0529310025241865, "learning_rate": 5.248032196104271e-06, "loss": 2.5453, "step": 1450 }, { "epoch": 1.9641164522681112, "grad_norm": 1.135872662601797, "learning_rate": 5.2426444757823595e-06, "loss": 2.1203, "step": 1451 }, { "epoch": 1.965470548408937, "grad_norm": 1.0238104015543514, "learning_rate": 5.2372564730457e-06, "loss": 2.2604, "step": 1452 }, { "epoch": 1.966824644549763, "grad_norm": 0.9503239116070696, "learning_rate": 5.231868194165409e-06, "loss": 2.4823, "step": 1453 }, { "epoch": 1.968178740690589, "grad_norm": 1.1160937498521495, "learning_rate": 5.226479645412923e-06, "loss": 2.2582, "step": 1454 }, { "epoch": 1.969532836831415, "grad_norm": 1.0681616120180237, "learning_rate": 5.221090833059992e-06, "loss": 2.1064, "step": 1455 }, { "epoch": 1.970886932972241, "grad_norm": 1.2336888814525633, "learning_rate": 5.215701763378673e-06, "loss": 2.2147, "step": 1456 }, { "epoch": 1.972241029113067, "grad_norm": 2.70975758573399, "learning_rate": 5.210312442641327e-06, "loss": 2.0919, "step": 1457 }, { "epoch": 1.973595125253893, "grad_norm": 1.062018871284572, "learning_rate": 5.204922877120597e-06, "loss": 2.1447, "step": 1458 }, { "epoch": 1.974949221394719, "grad_norm": 1.0661389981794334, "learning_rate": 5.19953307308942e-06, "loss": 1.9739, "step": 1459 }, { "epoch": 1.9763033175355451, "grad_norm": 1.0931038252658167, "learning_rate": 5.194143036821008e-06, "loss": 1.9725, "step": 1460 }, { "epoch": 1.9776574136763712, "grad_norm": 1.0840157871848553, "learning_rate": 5.188752774588841e-06, "loss": 2.2702, "step": 1461 }, { "epoch": 1.979011509817197, "grad_norm": 1.0991100379851682, "learning_rate": 5.183362292666667e-06, "loss": 2.2273, "step": 1462 }, { "epoch": 1.980365605958023, "grad_norm": 1.0930537215096972, "learning_rate": 5.177971597328484e-06, "loss": 2.2848, "step": 1463 }, { "epoch": 1.981719702098849, "grad_norm": 1.1727805767142325, "learning_rate": 5.172580694848541e-06, "loss": 2.2295, "step": 1464 }, { "epoch": 1.983073798239675, "grad_norm": 1.0331418445485365, "learning_rate": 5.16718959150133e-06, "loss": 2.1311, "step": 1465 }, { "epoch": 1.984427894380501, "grad_norm": 1.1782815464908813, "learning_rate": 5.161798293561573e-06, "loss": 2.2843, "step": 1466 }, { "epoch": 1.985781990521327, "grad_norm": 1.1287633720503536, "learning_rate": 5.156406807304223e-06, "loss": 2.2715, "step": 1467 }, { "epoch": 1.987136086662153, "grad_norm": 1.190713586693506, "learning_rate": 5.151015139004445e-06, "loss": 1.8253, "step": 1468 }, { "epoch": 1.988490182802979, "grad_norm": 1.070858654181211, "learning_rate": 5.145623294937624e-06, "loss": 2.1808, "step": 1469 }, { "epoch": 1.9898442789438051, "grad_norm": 1.1615785503909837, "learning_rate": 5.140231281379345e-06, "loss": 1.8325, "step": 1470 }, { "epoch": 1.9911983750846312, "grad_norm": 0.983503119817654, "learning_rate": 5.134839104605394e-06, "loss": 2.0685, "step": 1471 }, { "epoch": 1.992552471225457, "grad_norm": 1.1464176386439313, "learning_rate": 5.129446770891738e-06, "loss": 2.325, "step": 1472 }, { "epoch": 1.993906567366283, "grad_norm": 1.62246860837646, "learning_rate": 5.124054286514538e-06, "loss": 2.6138, "step": 1473 }, { "epoch": 1.9952606635071088, "grad_norm": 1.0277305356811535, "learning_rate": 5.118661657750122e-06, "loss": 2.1659, "step": 1474 }, { "epoch": 1.996614759647935, "grad_norm": 1.054556455145067, "learning_rate": 5.113268890874994e-06, "loss": 2.5919, "step": 1475 }, { "epoch": 2.001354096140826, "grad_norm": 1.1099987816340893, "learning_rate": 5.107875992165809e-06, "loss": 1.6184, "step": 1476 }, { "epoch": 2.002708192281652, "grad_norm": 0.9633957304966732, "learning_rate": 5.102482967899383e-06, "loss": 2.3234, "step": 1477 }, { "epoch": 2.004062288422478, "grad_norm": 0.9972865266518288, "learning_rate": 5.097089824352677e-06, "loss": 2.0429, "step": 1478 }, { "epoch": 2.005416384563304, "grad_norm": 1.1014804057689878, "learning_rate": 5.091696567802786e-06, "loss": 2.2343, "step": 1479 }, { "epoch": 2.0067704807041298, "grad_norm": 1.0811318725191603, "learning_rate": 5.0863032045269435e-06, "loss": 1.9575, "step": 1480 }, { "epoch": 2.008124576844956, "grad_norm": 2.1426169689874497, "learning_rate": 5.080909740802501e-06, "loss": 2.5731, "step": 1481 }, { "epoch": 2.009478672985782, "grad_norm": 1.1280877174822774, "learning_rate": 5.07551618290693e-06, "loss": 2.1989, "step": 1482 }, { "epoch": 2.010832769126608, "grad_norm": 1.0387751426576015, "learning_rate": 5.070122537117812e-06, "loss": 2.2141, "step": 1483 }, { "epoch": 2.012186865267434, "grad_norm": 1.1301144485709258, "learning_rate": 5.06472880971283e-06, "loss": 2.2053, "step": 1484 }, { "epoch": 2.01354096140826, "grad_norm": 1.083821235977125, "learning_rate": 5.059335006969763e-06, "loss": 2.0799, "step": 1485 }, { "epoch": 2.014895057549086, "grad_norm": 1.067398368449801, "learning_rate": 5.053941135166476e-06, "loss": 2.0641, "step": 1486 }, { "epoch": 2.016249153689912, "grad_norm": 1.0736623832641965, "learning_rate": 5.048547200580913e-06, "loss": 2.3372, "step": 1487 }, { "epoch": 2.017603249830738, "grad_norm": 1.1436139873449696, "learning_rate": 5.043153209491095e-06, "loss": 2.4194, "step": 1488 }, { "epoch": 2.018957345971564, "grad_norm": 1.151749715320049, "learning_rate": 5.037759168175109e-06, "loss": 1.9552, "step": 1489 }, { "epoch": 2.0203114421123898, "grad_norm": 1.1187695871940815, "learning_rate": 5.0323650829110945e-06, "loss": 1.8136, "step": 1490 }, { "epoch": 2.021665538253216, "grad_norm": 1.261982150598327, "learning_rate": 5.026970959977248e-06, "loss": 2.0799, "step": 1491 }, { "epoch": 2.023019634394042, "grad_norm": 1.0612746571999898, "learning_rate": 5.021576805651806e-06, "loss": 2.0989, "step": 1492 }, { "epoch": 2.024373730534868, "grad_norm": 1.0543489050545822, "learning_rate": 5.016182626213047e-06, "loss": 2.1228, "step": 1493 }, { "epoch": 2.025727826675694, "grad_norm": 0.986423192015442, "learning_rate": 5.010788427939275e-06, "loss": 1.9118, "step": 1494 }, { "epoch": 2.02708192281652, "grad_norm": 1.0503602628442998, "learning_rate": 5.0053942171088125e-06, "loss": 1.8566, "step": 1495 }, { "epoch": 2.028436018957346, "grad_norm": 1.1722167526158584, "learning_rate": 5e-06, "loss": 2.1443, "step": 1496 }, { "epoch": 2.029790115098172, "grad_norm": 1.2427395448191687, "learning_rate": 4.99460578289119e-06, "loss": 2.3088, "step": 1497 }, { "epoch": 2.031144211238998, "grad_norm": 1.1655323261951853, "learning_rate": 4.989211572060727e-06, "loss": 2.7667, "step": 1498 }, { "epoch": 2.032498307379824, "grad_norm": 1.1110856338031794, "learning_rate": 4.983817373786954e-06, "loss": 1.9205, "step": 1499 }, { "epoch": 2.0338524035206498, "grad_norm": 1.046067847813546, "learning_rate": 4.978423194348195e-06, "loss": 2.1045, "step": 1500 }, { "epoch": 2.035206499661476, "grad_norm": 1.2302003508661763, "learning_rate": 4.973029040022754e-06, "loss": 2.3181, "step": 1501 }, { "epoch": 2.036560595802302, "grad_norm": 1.135349688929259, "learning_rate": 4.967634917088907e-06, "loss": 2.3167, "step": 1502 }, { "epoch": 2.037914691943128, "grad_norm": 1.1800517406403312, "learning_rate": 4.9622408318248925e-06, "loss": 2.1869, "step": 1503 }, { "epoch": 2.039268788083954, "grad_norm": 0.9448466409715037, "learning_rate": 4.956846790508906e-06, "loss": 2.4484, "step": 1504 }, { "epoch": 2.04062288422478, "grad_norm": 1.2928756280873905, "learning_rate": 4.9514527994190885e-06, "loss": 1.9511, "step": 1505 }, { "epoch": 2.041976980365606, "grad_norm": 1.0720514579615468, "learning_rate": 4.946058864833526e-06, "loss": 2.2906, "step": 1506 }, { "epoch": 2.043331076506432, "grad_norm": 1.0653649392471873, "learning_rate": 4.940664993030238e-06, "loss": 2.0507, "step": 1507 }, { "epoch": 2.044685172647258, "grad_norm": 1.135639769179713, "learning_rate": 4.935271190287171e-06, "loss": 2.1605, "step": 1508 }, { "epoch": 2.0460392687880837, "grad_norm": 1.1660308195349796, "learning_rate": 4.92987746288219e-06, "loss": 2.258, "step": 1509 }, { "epoch": 2.0473933649289098, "grad_norm": 1.0555106320650247, "learning_rate": 4.924483817093071e-06, "loss": 1.9635, "step": 1510 }, { "epoch": 2.048747461069736, "grad_norm": 1.161574450971683, "learning_rate": 4.9190902591975005e-06, "loss": 1.89, "step": 1511 }, { "epoch": 2.050101557210562, "grad_norm": 0.8844191832858331, "learning_rate": 4.913696795473058e-06, "loss": 2.3977, "step": 1512 }, { "epoch": 2.051455653351388, "grad_norm": 1.1202420669878905, "learning_rate": 4.908303432197215e-06, "loss": 2.0963, "step": 1513 }, { "epoch": 2.052809749492214, "grad_norm": 1.0672046562799546, "learning_rate": 4.902910175647325e-06, "loss": 2.4326, "step": 1514 }, { "epoch": 2.05416384563304, "grad_norm": 1.032894269405954, "learning_rate": 4.897517032100617e-06, "loss": 2.1201, "step": 1515 }, { "epoch": 2.055517941773866, "grad_norm": 1.1101842667323178, "learning_rate": 4.892124007834192e-06, "loss": 2.8059, "step": 1516 }, { "epoch": 2.056872037914692, "grad_norm": 1.1335931625514726, "learning_rate": 4.886731109125007e-06, "loss": 2.0627, "step": 1517 }, { "epoch": 2.058226134055518, "grad_norm": 1.1461874883569612, "learning_rate": 4.8813383422498795e-06, "loss": 2.3601, "step": 1518 }, { "epoch": 2.0595802301963437, "grad_norm": 1.1863031869713507, "learning_rate": 4.8759457134854645e-06, "loss": 2.1693, "step": 1519 }, { "epoch": 2.0609343263371698, "grad_norm": 1.1384913524334759, "learning_rate": 4.8705532291082644e-06, "loss": 2.0691, "step": 1520 }, { "epoch": 2.062288422477996, "grad_norm": 1.4267866229603507, "learning_rate": 4.8651608953946095e-06, "loss": 2.0377, "step": 1521 }, { "epoch": 2.063642518618822, "grad_norm": 1.1092482886444381, "learning_rate": 4.859768718620656e-06, "loss": 2.3096, "step": 1522 }, { "epoch": 2.064996614759648, "grad_norm": 1.1069908008833043, "learning_rate": 4.854376705062378e-06, "loss": 2.4173, "step": 1523 }, { "epoch": 2.066350710900474, "grad_norm": 1.2173162669460578, "learning_rate": 4.848984860995557e-06, "loss": 1.9191, "step": 1524 }, { "epoch": 2.0677048070413, "grad_norm": 1.3162083246670064, "learning_rate": 4.84359319269578e-06, "loss": 2.3459, "step": 1525 }, { "epoch": 2.069058903182126, "grad_norm": 1.1452290922812693, "learning_rate": 4.838201706438428e-06, "loss": 2.0403, "step": 1526 }, { "epoch": 2.070412999322952, "grad_norm": 1.1296537536889968, "learning_rate": 4.832810408498671e-06, "loss": 2.1389, "step": 1527 }, { "epoch": 2.071767095463778, "grad_norm": 1.245874366651633, "learning_rate": 4.827419305151461e-06, "loss": 2.0979, "step": 1528 }, { "epoch": 2.0731211916046037, "grad_norm": 1.1295942946447188, "learning_rate": 4.8220284026715185e-06, "loss": 2.6153, "step": 1529 }, { "epoch": 2.0744752877454298, "grad_norm": 1.2668073230934784, "learning_rate": 4.816637707333336e-06, "loss": 1.9422, "step": 1530 }, { "epoch": 2.075829383886256, "grad_norm": 1.6366313985835925, "learning_rate": 4.81124722541116e-06, "loss": 2.3686, "step": 1531 }, { "epoch": 2.077183480027082, "grad_norm": 1.0378803629392122, "learning_rate": 4.805856963178996e-06, "loss": 2.3727, "step": 1532 }, { "epoch": 2.078537576167908, "grad_norm": 1.1246262827678393, "learning_rate": 4.800466926910582e-06, "loss": 1.8426, "step": 1533 }, { "epoch": 2.079891672308734, "grad_norm": 1.1506088445178124, "learning_rate": 4.7950771228794056e-06, "loss": 1.7417, "step": 1534 }, { "epoch": 2.08124576844956, "grad_norm": 0.9568321392137888, "learning_rate": 4.789687557358676e-06, "loss": 2.0365, "step": 1535 }, { "epoch": 2.082599864590386, "grad_norm": 1.080563568463781, "learning_rate": 4.7842982366213275e-06, "loss": 2.0559, "step": 1536 }, { "epoch": 2.083953960731212, "grad_norm": 1.13301896950942, "learning_rate": 4.778909166940011e-06, "loss": 2.129, "step": 1537 }, { "epoch": 2.085308056872038, "grad_norm": 1.137742320059762, "learning_rate": 4.7735203545870794e-06, "loss": 2.051, "step": 1538 }, { "epoch": 2.0866621530128637, "grad_norm": 1.234501634109048, "learning_rate": 4.768131805834594e-06, "loss": 2.2144, "step": 1539 }, { "epoch": 2.0880162491536898, "grad_norm": 1.1602709681570815, "learning_rate": 4.762743526954302e-06, "loss": 2.1123, "step": 1540 }, { "epoch": 2.089370345294516, "grad_norm": 1.3067404824964808, "learning_rate": 4.757355524217643e-06, "loss": 2.2777, "step": 1541 }, { "epoch": 2.090724441435342, "grad_norm": 0.9312850510780861, "learning_rate": 4.751967803895729e-06, "loss": 2.0286, "step": 1542 }, { "epoch": 2.092078537576168, "grad_norm": 1.2880188746624512, "learning_rate": 4.746580372259349e-06, "loss": 1.9872, "step": 1543 }, { "epoch": 2.093432633716994, "grad_norm": 1.1356683167193133, "learning_rate": 4.741193235578953e-06, "loss": 2.2151, "step": 1544 }, { "epoch": 2.09478672985782, "grad_norm": 1.2296847360245158, "learning_rate": 4.735806400124648e-06, "loss": 2.0566, "step": 1545 }, { "epoch": 2.096140825998646, "grad_norm": 1.0406248078824951, "learning_rate": 4.7304198721661884e-06, "loss": 2.184, "step": 1546 }, { "epoch": 2.097494922139472, "grad_norm": 1.311049482691511, "learning_rate": 4.725033657972975e-06, "loss": 1.8853, "step": 1547 }, { "epoch": 2.0988490182802977, "grad_norm": 1.3283344688094014, "learning_rate": 4.719647763814041e-06, "loss": 1.9299, "step": 1548 }, { "epoch": 2.1002031144211237, "grad_norm": 1.0428035980723034, "learning_rate": 4.714262195958047e-06, "loss": 2.0196, "step": 1549 }, { "epoch": 2.1015572105619498, "grad_norm": 1.1579964017634277, "learning_rate": 4.708876960673273e-06, "loss": 2.1329, "step": 1550 }, { "epoch": 2.102911306702776, "grad_norm": 1.0576288691929832, "learning_rate": 4.703492064227614e-06, "loss": 2.0175, "step": 1551 }, { "epoch": 2.104265402843602, "grad_norm": 1.0866916247911798, "learning_rate": 4.69810751288857e-06, "loss": 2.0619, "step": 1552 }, { "epoch": 2.105619498984428, "grad_norm": 2.304676104467671, "learning_rate": 4.692723312923238e-06, "loss": 2.1748, "step": 1553 }, { "epoch": 2.106973595125254, "grad_norm": 1.185452693779257, "learning_rate": 4.687339470598308e-06, "loss": 2.4933, "step": 1554 }, { "epoch": 2.10832769126608, "grad_norm": 1.0970400294111906, "learning_rate": 4.681955992180052e-06, "loss": 2.1228, "step": 1555 }, { "epoch": 2.109681787406906, "grad_norm": 1.1178728146653045, "learning_rate": 4.676572883934319e-06, "loss": 1.8178, "step": 1556 }, { "epoch": 2.111035883547732, "grad_norm": 1.2952454007071246, "learning_rate": 4.671190152126528e-06, "loss": 2.0092, "step": 1557 }, { "epoch": 2.1123899796885577, "grad_norm": 1.1676976415967362, "learning_rate": 4.665807803021659e-06, "loss": 2.2843, "step": 1558 }, { "epoch": 2.1137440758293837, "grad_norm": 1.0535623466354502, "learning_rate": 4.660425842884249e-06, "loss": 1.9981, "step": 1559 }, { "epoch": 2.1150981719702098, "grad_norm": 0.9313987504909503, "learning_rate": 4.6550442779783755e-06, "loss": 1.9104, "step": 1560 }, { "epoch": 2.116452268111036, "grad_norm": 1.1084369317244605, "learning_rate": 4.649663114567663e-06, "loss": 1.9962, "step": 1561 }, { "epoch": 2.117806364251862, "grad_norm": 1.2273961498810957, "learning_rate": 4.644282358915267e-06, "loss": 2.1789, "step": 1562 }, { "epoch": 2.119160460392688, "grad_norm": 1.0352472555668162, "learning_rate": 4.638902017283867e-06, "loss": 1.8401, "step": 1563 }, { "epoch": 2.120514556533514, "grad_norm": 1.0726491386124741, "learning_rate": 4.6335220959356605e-06, "loss": 2.1011, "step": 1564 }, { "epoch": 2.12186865267434, "grad_norm": 1.135410546277804, "learning_rate": 4.628142601132358e-06, "loss": 2.1076, "step": 1565 }, { "epoch": 2.123222748815166, "grad_norm": 1.0333310129766364, "learning_rate": 4.62276353913517e-06, "loss": 1.7628, "step": 1566 }, { "epoch": 2.124576844955992, "grad_norm": 1.11050363211758, "learning_rate": 4.617384916204808e-06, "loss": 1.8538, "step": 1567 }, { "epoch": 2.1259309410968177, "grad_norm": 1.093612224379343, "learning_rate": 4.612006738601469e-06, "loss": 2.1253, "step": 1568 }, { "epoch": 2.1272850372376437, "grad_norm": 1.1237634537047279, "learning_rate": 4.606629012584832e-06, "loss": 2.1641, "step": 1569 }, { "epoch": 2.1286391333784698, "grad_norm": 1.166335975344414, "learning_rate": 4.601251744414051e-06, "loss": 2.1319, "step": 1570 }, { "epoch": 2.129993229519296, "grad_norm": 1.2806092788823242, "learning_rate": 4.595874940347746e-06, "loss": 2.0234, "step": 1571 }, { "epoch": 2.131347325660122, "grad_norm": 1.2122990216252296, "learning_rate": 4.590498606644001e-06, "loss": 1.9265, "step": 1572 }, { "epoch": 2.132701421800948, "grad_norm": 1.1023577932522146, "learning_rate": 4.585122749560347e-06, "loss": 2.3221, "step": 1573 }, { "epoch": 2.134055517941774, "grad_norm": 1.47306076341711, "learning_rate": 4.579747375353763e-06, "loss": 1.9067, "step": 1574 }, { "epoch": 2.1354096140826, "grad_norm": 1.3931738365078317, "learning_rate": 4.574372490280666e-06, "loss": 2.0922, "step": 1575 }, { "epoch": 2.136763710223426, "grad_norm": 1.372903750413496, "learning_rate": 4.568998100596903e-06, "loss": 2.2256, "step": 1576 }, { "epoch": 2.138117806364252, "grad_norm": 1.2347870944106532, "learning_rate": 4.563624212557746e-06, "loss": 2.1726, "step": 1577 }, { "epoch": 2.1394719025050777, "grad_norm": 1.2352502492223656, "learning_rate": 4.558250832417882e-06, "loss": 1.98, "step": 1578 }, { "epoch": 2.1408259986459037, "grad_norm": 1.1857194178661712, "learning_rate": 4.552877966431404e-06, "loss": 2.2132, "step": 1579 }, { "epoch": 2.1421800947867298, "grad_norm": 1.8837742302996179, "learning_rate": 4.547505620851812e-06, "loss": 2.1429, "step": 1580 }, { "epoch": 2.143534190927556, "grad_norm": 1.1156760041554328, "learning_rate": 4.542133801931996e-06, "loss": 2.6146, "step": 1581 }, { "epoch": 2.144888287068382, "grad_norm": 1.2461115120002666, "learning_rate": 4.536762515924236e-06, "loss": 2.5796, "step": 1582 }, { "epoch": 2.146242383209208, "grad_norm": 1.1930848492526376, "learning_rate": 4.531391769080191e-06, "loss": 2.0747, "step": 1583 }, { "epoch": 2.147596479350034, "grad_norm": 1.1880358431537892, "learning_rate": 4.526021567650889e-06, "loss": 2.2914, "step": 1584 }, { "epoch": 2.14895057549086, "grad_norm": 1.2497945778465804, "learning_rate": 4.520651917886729e-06, "loss": 2.3135, "step": 1585 }, { "epoch": 2.150304671631686, "grad_norm": 1.1153960095391755, "learning_rate": 4.515282826037463e-06, "loss": 2.2851, "step": 1586 }, { "epoch": 2.1516587677725116, "grad_norm": 1.1453793761355706, "learning_rate": 4.509914298352197e-06, "loss": 2.5681, "step": 1587 }, { "epoch": 2.1530128639133377, "grad_norm": 1.1235008579227046, "learning_rate": 4.50454634107938e-06, "loss": 2.2522, "step": 1588 }, { "epoch": 2.1543669600541637, "grad_norm": 1.223535551184846, "learning_rate": 4.499178960466792e-06, "loss": 1.9945, "step": 1589 }, { "epoch": 2.1557210561949898, "grad_norm": 1.1910407720274174, "learning_rate": 4.4938121627615495e-06, "loss": 2.2432, "step": 1590 }, { "epoch": 2.157075152335816, "grad_norm": 1.097499963545351, "learning_rate": 4.488445954210086e-06, "loss": 2.075, "step": 1591 }, { "epoch": 2.158429248476642, "grad_norm": 1.0837423130459314, "learning_rate": 4.4830803410581506e-06, "loss": 2.2556, "step": 1592 }, { "epoch": 2.159783344617468, "grad_norm": 1.2065862379928247, "learning_rate": 4.477715329550797e-06, "loss": 2.3815, "step": 1593 }, { "epoch": 2.161137440758294, "grad_norm": 1.1926926464235388, "learning_rate": 4.472350925932384e-06, "loss": 1.9872, "step": 1594 }, { "epoch": 2.16249153689912, "grad_norm": 1.0737372726387304, "learning_rate": 4.466987136446556e-06, "loss": 2.153, "step": 1595 }, { "epoch": 2.163845633039946, "grad_norm": 1.1530425985515302, "learning_rate": 4.461623967336249e-06, "loss": 1.9165, "step": 1596 }, { "epoch": 2.1651997291807716, "grad_norm": 1.2060757397302357, "learning_rate": 4.456261424843675e-06, "loss": 2.0127, "step": 1597 }, { "epoch": 2.1665538253215977, "grad_norm": 1.0746785079550367, "learning_rate": 4.4508995152103116e-06, "loss": 2.0357, "step": 1598 }, { "epoch": 2.1679079214624237, "grad_norm": 1.1419182114508935, "learning_rate": 4.445538244676907e-06, "loss": 2.3228, "step": 1599 }, { "epoch": 2.1692620176032498, "grad_norm": 1.3579856746137162, "learning_rate": 4.4401776194834615e-06, "loss": 2.1991, "step": 1600 }, { "epoch": 2.170616113744076, "grad_norm": 1.1609210651655175, "learning_rate": 4.434817645869226e-06, "loss": 2.1688, "step": 1601 }, { "epoch": 2.171970209884902, "grad_norm": 1.2507237479225968, "learning_rate": 4.429458330072693e-06, "loss": 2.0951, "step": 1602 }, { "epoch": 2.173324306025728, "grad_norm": 1.1756864814287347, "learning_rate": 4.424099678331587e-06, "loss": 2.2656, "step": 1603 }, { "epoch": 2.174678402166554, "grad_norm": 1.1902481513970695, "learning_rate": 4.418741696882863e-06, "loss": 2.8234, "step": 1604 }, { "epoch": 2.17603249830738, "grad_norm": 1.1301334146636766, "learning_rate": 4.4133843919626925e-06, "loss": 1.982, "step": 1605 }, { "epoch": 2.1773865944482056, "grad_norm": 1.042788785013733, "learning_rate": 4.408027769806462e-06, "loss": 3.0665, "step": 1606 }, { "epoch": 2.1787406905890316, "grad_norm": 1.2402217857663635, "learning_rate": 4.402671836648764e-06, "loss": 1.8366, "step": 1607 }, { "epoch": 2.1800947867298577, "grad_norm": 1.0280298132953454, "learning_rate": 4.397316598723385e-06, "loss": 2.1774, "step": 1608 }, { "epoch": 2.1814488828706837, "grad_norm": 1.2158910131402605, "learning_rate": 4.391962062263308e-06, "loss": 1.8601, "step": 1609 }, { "epoch": 2.1828029790115098, "grad_norm": 1.1406998787275462, "learning_rate": 4.386608233500694e-06, "loss": 1.9644, "step": 1610 }, { "epoch": 2.184157075152336, "grad_norm": 0.9266322032604677, "learning_rate": 4.381255118666885e-06, "loss": 2.0725, "step": 1611 }, { "epoch": 2.185511171293162, "grad_norm": 1.2923640198564135, "learning_rate": 4.375902723992388e-06, "loss": 1.88, "step": 1612 }, { "epoch": 2.186865267433988, "grad_norm": 1.2203033908934358, "learning_rate": 4.3705510557068746e-06, "loss": 2.263, "step": 1613 }, { "epoch": 2.188219363574814, "grad_norm": 1.1574828018776115, "learning_rate": 4.365200120039169e-06, "loss": 2.2875, "step": 1614 }, { "epoch": 2.18957345971564, "grad_norm": 1.1003584808353581, "learning_rate": 4.359849923217246e-06, "loss": 2.4866, "step": 1615 }, { "epoch": 2.190927555856466, "grad_norm": 1.1764714484543735, "learning_rate": 4.354500471468217e-06, "loss": 2.1726, "step": 1616 }, { "epoch": 2.1922816519972916, "grad_norm": 1.109296700586937, "learning_rate": 4.3491517710183275e-06, "loss": 2.44, "step": 1617 }, { "epoch": 2.1936357481381177, "grad_norm": 1.063875424222137, "learning_rate": 4.343803828092947e-06, "loss": 1.8824, "step": 1618 }, { "epoch": 2.1949898442789437, "grad_norm": 1.0448414422619752, "learning_rate": 4.338456648916567e-06, "loss": 2.263, "step": 1619 }, { "epoch": 2.1963439404197698, "grad_norm": 1.25112317931986, "learning_rate": 4.333110239712788e-06, "loss": 2.4117, "step": 1620 }, { "epoch": 2.197698036560596, "grad_norm": 1.2917980883647506, "learning_rate": 4.327764606704313e-06, "loss": 2.3118, "step": 1621 }, { "epoch": 2.199052132701422, "grad_norm": 1.1987958985580311, "learning_rate": 4.322419756112943e-06, "loss": 2.1132, "step": 1622 }, { "epoch": 2.200406228842248, "grad_norm": 1.339749545237993, "learning_rate": 4.317075694159567e-06, "loss": 2.2394, "step": 1623 }, { "epoch": 2.201760324983074, "grad_norm": 1.1998654597010925, "learning_rate": 4.31173242706416e-06, "loss": 1.8347, "step": 1624 }, { "epoch": 2.2031144211239, "grad_norm": 1.095910470513959, "learning_rate": 4.306389961045767e-06, "loss": 2.1596, "step": 1625 }, { "epoch": 2.2044685172647256, "grad_norm": 1.0871385280387196, "learning_rate": 4.3010483023225045e-06, "loss": 1.8082, "step": 1626 }, { "epoch": 2.2058226134055516, "grad_norm": 1.2099864152673303, "learning_rate": 4.295707457111545e-06, "loss": 1.9518, "step": 1627 }, { "epoch": 2.2071767095463777, "grad_norm": 1.1342817848587583, "learning_rate": 4.290367431629119e-06, "loss": 2.8945, "step": 1628 }, { "epoch": 2.2085308056872037, "grad_norm": 1.2261803055050031, "learning_rate": 4.285028232090499e-06, "loss": 2.2579, "step": 1629 }, { "epoch": 2.2098849018280298, "grad_norm": 1.205283434883726, "learning_rate": 4.27968986471e-06, "loss": 2.0904, "step": 1630 }, { "epoch": 2.211238997968856, "grad_norm": 1.0939560971887643, "learning_rate": 4.274352335700964e-06, "loss": 1.816, "step": 1631 }, { "epoch": 2.212593094109682, "grad_norm": 1.1149798418376844, "learning_rate": 4.269015651275761e-06, "loss": 2.0789, "step": 1632 }, { "epoch": 2.213947190250508, "grad_norm": 1.1538450924534014, "learning_rate": 4.263679817645775e-06, "loss": 2.4258, "step": 1633 }, { "epoch": 2.215301286391334, "grad_norm": 1.1179951823339618, "learning_rate": 4.258344841021403e-06, "loss": 1.9159, "step": 1634 }, { "epoch": 2.21665538253216, "grad_norm": 1.1684602515702702, "learning_rate": 4.253010727612042e-06, "loss": 2.4231, "step": 1635 }, { "epoch": 2.2180094786729856, "grad_norm": 1.0157380652460308, "learning_rate": 4.247677483626085e-06, "loss": 1.9744, "step": 1636 }, { "epoch": 2.2193635748138116, "grad_norm": 1.1680351754748413, "learning_rate": 4.242345115270914e-06, "loss": 1.9646, "step": 1637 }, { "epoch": 2.2207176709546377, "grad_norm": 1.1860802698747048, "learning_rate": 4.2370136287528905e-06, "loss": 2.3634, "step": 1638 }, { "epoch": 2.2220717670954637, "grad_norm": 1.1623275484358482, "learning_rate": 4.231683030277349e-06, "loss": 2.324, "step": 1639 }, { "epoch": 2.2234258632362898, "grad_norm": 1.0683124715390833, "learning_rate": 4.226353326048594e-06, "loss": 2.2478, "step": 1640 }, { "epoch": 2.224779959377116, "grad_norm": 1.0941651792915803, "learning_rate": 4.221024522269884e-06, "loss": 2.0946, "step": 1641 }, { "epoch": 2.226134055517942, "grad_norm": 1.3547392965324918, "learning_rate": 4.215696625143434e-06, "loss": 1.8724, "step": 1642 }, { "epoch": 2.227488151658768, "grad_norm": 1.1662802032623258, "learning_rate": 4.210369640870403e-06, "loss": 1.922, "step": 1643 }, { "epoch": 2.228842247799594, "grad_norm": 1.1746707623920845, "learning_rate": 4.205043575650884e-06, "loss": 1.9508, "step": 1644 }, { "epoch": 2.2301963439404195, "grad_norm": 1.1908224223990358, "learning_rate": 4.199718435683902e-06, "loss": 2.4077, "step": 1645 }, { "epoch": 2.2315504400812456, "grad_norm": 1.1439669672796644, "learning_rate": 4.194394227167409e-06, "loss": 2.0898, "step": 1646 }, { "epoch": 2.2329045362220716, "grad_norm": 1.3067429767218774, "learning_rate": 4.189070956298269e-06, "loss": 2.0362, "step": 1647 }, { "epoch": 2.2342586323628977, "grad_norm": 1.073330395456851, "learning_rate": 4.183748629272254e-06, "loss": 1.878, "step": 1648 }, { "epoch": 2.2356127285037237, "grad_norm": 1.1629255939548435, "learning_rate": 4.178427252284039e-06, "loss": 1.7567, "step": 1649 }, { "epoch": 2.2369668246445498, "grad_norm": 1.1371967688615032, "learning_rate": 4.173106831527194e-06, "loss": 2.0284, "step": 1650 }, { "epoch": 2.238320920785376, "grad_norm": 1.4840544099057114, "learning_rate": 4.167787373194175e-06, "loss": 2.1274, "step": 1651 }, { "epoch": 2.239675016926202, "grad_norm": 1.191844972399583, "learning_rate": 4.162468883476319e-06, "loss": 2.1309, "step": 1652 }, { "epoch": 2.241029113067028, "grad_norm": 1.1315079396239087, "learning_rate": 4.157151368563832e-06, "loss": 2.0473, "step": 1653 }, { "epoch": 2.242383209207854, "grad_norm": 1.058024081319093, "learning_rate": 4.1518348346457895e-06, "loss": 2.0161, "step": 1654 }, { "epoch": 2.24373730534868, "grad_norm": 1.2675977862246117, "learning_rate": 4.146519287910124e-06, "loss": 2.1973, "step": 1655 }, { "epoch": 2.2450914014895056, "grad_norm": 1.1412508576223728, "learning_rate": 4.14120473454362e-06, "loss": 2.377, "step": 1656 }, { "epoch": 2.2464454976303316, "grad_norm": 1.0899854425232176, "learning_rate": 4.135891180731903e-06, "loss": 1.8871, "step": 1657 }, { "epoch": 2.2477995937711577, "grad_norm": 1.1627149659156846, "learning_rate": 4.130578632659436e-06, "loss": 2.1091, "step": 1658 }, { "epoch": 2.2491536899119837, "grad_norm": 1.1452779953576764, "learning_rate": 4.125267096509513e-06, "loss": 2.2526, "step": 1659 }, { "epoch": 2.2505077860528098, "grad_norm": 1.3565369903794207, "learning_rate": 4.11995657846425e-06, "loss": 2.477, "step": 1660 }, { "epoch": 2.251861882193636, "grad_norm": 1.3173723291108739, "learning_rate": 4.114647084704575e-06, "loss": 2.3661, "step": 1661 }, { "epoch": 2.253215978334462, "grad_norm": 1.021914924841384, "learning_rate": 4.109338621410231e-06, "loss": 2.1235, "step": 1662 }, { "epoch": 2.254570074475288, "grad_norm": 1.0746890128820057, "learning_rate": 4.104031194759752e-06, "loss": 1.974, "step": 1663 }, { "epoch": 2.2559241706161135, "grad_norm": 1.4923113523539697, "learning_rate": 4.098724810930472e-06, "loss": 2.0366, "step": 1664 }, { "epoch": 2.2572782667569395, "grad_norm": 1.0166719359621823, "learning_rate": 4.0934194760985095e-06, "loss": 2.0029, "step": 1665 }, { "epoch": 2.2586323628977656, "grad_norm": 1.0910263057406069, "learning_rate": 4.088115196438763e-06, "loss": 1.9998, "step": 1666 }, { "epoch": 2.2599864590385916, "grad_norm": 1.4321157239578748, "learning_rate": 4.0828119781249e-06, "loss": 2.1868, "step": 1667 }, { "epoch": 2.2613405551794177, "grad_norm": 1.1054628441127339, "learning_rate": 4.077509827329357e-06, "loss": 1.9817, "step": 1668 }, { "epoch": 2.2626946513202437, "grad_norm": 1.1668394403143847, "learning_rate": 4.072208750223324e-06, "loss": 1.8308, "step": 1669 }, { "epoch": 2.2640487474610698, "grad_norm": 1.2899936947835928, "learning_rate": 4.066908752976743e-06, "loss": 2.2294, "step": 1670 }, { "epoch": 2.265402843601896, "grad_norm": 1.547516563075485, "learning_rate": 4.061609841758302e-06, "loss": 2.2889, "step": 1671 }, { "epoch": 2.266756939742722, "grad_norm": 1.0503622123088423, "learning_rate": 4.056312022735417e-06, "loss": 2.2654, "step": 1672 }, { "epoch": 2.268111035883548, "grad_norm": 1.2481918302674415, "learning_rate": 4.051015302074243e-06, "loss": 2.6836, "step": 1673 }, { "epoch": 2.269465132024374, "grad_norm": 1.2299671525086586, "learning_rate": 4.045719685939648e-06, "loss": 2.9113, "step": 1674 }, { "epoch": 2.2708192281651995, "grad_norm": 1.256944791567761, "learning_rate": 4.040425180495219e-06, "loss": 1.9272, "step": 1675 }, { "epoch": 2.2721733243060256, "grad_norm": 1.2044589275520543, "learning_rate": 4.035131791903252e-06, "loss": 2.1785, "step": 1676 }, { "epoch": 2.2735274204468516, "grad_norm": 1.186230314361689, "learning_rate": 4.029839526324736e-06, "loss": 1.8621, "step": 1677 }, { "epoch": 2.2748815165876777, "grad_norm": 1.1433008988333722, "learning_rate": 4.02454838991936e-06, "loss": 2.4923, "step": 1678 }, { "epoch": 2.2762356127285037, "grad_norm": 1.3971522001704908, "learning_rate": 4.019258388845494e-06, "loss": 2.1802, "step": 1679 }, { "epoch": 2.2775897088693298, "grad_norm": 1.1321068783829218, "learning_rate": 4.013969529260191e-06, "loss": 1.9983, "step": 1680 }, { "epoch": 2.278943805010156, "grad_norm": 1.1040588304473815, "learning_rate": 4.008681817319172e-06, "loss": 1.8928, "step": 1681 }, { "epoch": 2.280297901150982, "grad_norm": 1.0988523181486751, "learning_rate": 4.003395259176822e-06, "loss": 1.7453, "step": 1682 }, { "epoch": 2.281651997291808, "grad_norm": 1.3976771401331631, "learning_rate": 3.998109860986186e-06, "loss": 2.0783, "step": 1683 }, { "epoch": 2.2830060934326335, "grad_norm": 1.127197396739673, "learning_rate": 3.9928256288989556e-06, "loss": 2.3781, "step": 1684 }, { "epoch": 2.2843601895734595, "grad_norm": 1.1436749511471185, "learning_rate": 3.987542569065469e-06, "loss": 2.3184, "step": 1685 }, { "epoch": 2.2857142857142856, "grad_norm": 1.09832131495446, "learning_rate": 3.982260687634698e-06, "loss": 2.03, "step": 1686 }, { "epoch": 2.2870683818551116, "grad_norm": 1.250103007503673, "learning_rate": 3.97697999075424e-06, "loss": 2.0276, "step": 1687 }, { "epoch": 2.2884224779959377, "grad_norm": 1.072931382143416, "learning_rate": 3.9717004845703175e-06, "loss": 2.1576, "step": 1688 }, { "epoch": 2.2897765741367637, "grad_norm": 0.9550559478392131, "learning_rate": 3.966422175227767e-06, "loss": 2.0699, "step": 1689 }, { "epoch": 2.2911306702775898, "grad_norm": 1.0752132263819945, "learning_rate": 3.961145068870032e-06, "loss": 2.3097, "step": 1690 }, { "epoch": 2.292484766418416, "grad_norm": 1.0803808533463632, "learning_rate": 3.955869171639151e-06, "loss": 1.7846, "step": 1691 }, { "epoch": 2.293838862559242, "grad_norm": 1.1197583210895903, "learning_rate": 3.9505944896757635e-06, "loss": 2.3472, "step": 1692 }, { "epoch": 2.295192958700068, "grad_norm": 1.081860958378646, "learning_rate": 3.945321029119087e-06, "loss": 2.363, "step": 1693 }, { "epoch": 2.296547054840894, "grad_norm": 1.1614031763131691, "learning_rate": 3.940048796106921e-06, "loss": 2.2561, "step": 1694 }, { "epoch": 2.2979011509817195, "grad_norm": 1.1621692326486843, "learning_rate": 3.934777796775639e-06, "loss": 2.0617, "step": 1695 }, { "epoch": 2.2992552471225456, "grad_norm": 1.0471833259566803, "learning_rate": 3.92950803726017e-06, "loss": 2.3505, "step": 1696 }, { "epoch": 2.3006093432633716, "grad_norm": 0.9888104881880921, "learning_rate": 3.92423952369401e-06, "loss": 2.3782, "step": 1697 }, { "epoch": 2.3019634394041977, "grad_norm": 1.1710462303933948, "learning_rate": 3.918972262209197e-06, "loss": 2.3728, "step": 1698 }, { "epoch": 2.3033175355450237, "grad_norm": 1.3284903669863712, "learning_rate": 3.913706258936317e-06, "loss": 1.9305, "step": 1699 }, { "epoch": 2.3046716316858498, "grad_norm": 1.0958983716178416, "learning_rate": 3.90844152000449e-06, "loss": 1.8812, "step": 1700 }, { "epoch": 2.306025727826676, "grad_norm": 1.27975156841522, "learning_rate": 3.903178051541361e-06, "loss": 1.2622, "step": 1701 }, { "epoch": 2.307379823967502, "grad_norm": 1.269713622936882, "learning_rate": 3.897915859673101e-06, "loss": 2.4716, "step": 1702 }, { "epoch": 2.3087339201083275, "grad_norm": 1.189476182489876, "learning_rate": 3.892654950524394e-06, "loss": 2.1382, "step": 1703 }, { "epoch": 2.3100880162491535, "grad_norm": 1.018132696944682, "learning_rate": 3.887395330218429e-06, "loss": 1.926, "step": 1704 }, { "epoch": 2.3114421123899795, "grad_norm": 1.1871078224967542, "learning_rate": 3.882137004876898e-06, "loss": 2.1995, "step": 1705 }, { "epoch": 2.3127962085308056, "grad_norm": 1.1429296267894478, "learning_rate": 3.876879980619982e-06, "loss": 1.9517, "step": 1706 }, { "epoch": 2.3141503046716316, "grad_norm": 1.4508823986316959, "learning_rate": 3.871624263566351e-06, "loss": 1.7302, "step": 1707 }, { "epoch": 2.3155044008124577, "grad_norm": 1.28033596964489, "learning_rate": 3.866369859833152e-06, "loss": 2.4238, "step": 1708 }, { "epoch": 2.3168584969532837, "grad_norm": 1.1255781179482813, "learning_rate": 3.861116775536004e-06, "loss": 2.3669, "step": 1709 }, { "epoch": 2.3182125930941098, "grad_norm": 1.1835230854995868, "learning_rate": 3.85586501678899e-06, "loss": 2.0611, "step": 1710 }, { "epoch": 2.319566689234936, "grad_norm": 1.2488495069241405, "learning_rate": 3.85061458970465e-06, "loss": 2.7457, "step": 1711 }, { "epoch": 2.320920785375762, "grad_norm": 1.1153007878225367, "learning_rate": 3.845365500393974e-06, "loss": 1.8768, "step": 1712 }, { "epoch": 2.322274881516588, "grad_norm": 1.1748340216351245, "learning_rate": 3.840117754966396e-06, "loss": 2.5446, "step": 1713 }, { "epoch": 2.3236289776574135, "grad_norm": 1.149687138464526, "learning_rate": 3.834871359529785e-06, "loss": 1.8704, "step": 1714 }, { "epoch": 2.3249830737982395, "grad_norm": 1.1843277957382676, "learning_rate": 3.829626320190438e-06, "loss": 2.3361, "step": 1715 }, { "epoch": 2.3263371699390656, "grad_norm": 1.2294816388156478, "learning_rate": 3.824382643053075e-06, "loss": 2.2094, "step": 1716 }, { "epoch": 2.3276912660798916, "grad_norm": 1.2287368389642799, "learning_rate": 3.81914033422083e-06, "loss": 2.2145, "step": 1717 }, { "epoch": 2.3290453622207177, "grad_norm": 1.0838440885981533, "learning_rate": 3.8138993997952446e-06, "loss": 1.8142, "step": 1718 }, { "epoch": 2.3303994583615437, "grad_norm": 1.1061489940249543, "learning_rate": 3.8086598458762606e-06, "loss": 1.8989, "step": 1719 }, { "epoch": 2.3317535545023698, "grad_norm": 1.1317906776428788, "learning_rate": 3.803421678562213e-06, "loss": 1.9887, "step": 1720 }, { "epoch": 2.333107650643196, "grad_norm": 1.25685995846726, "learning_rate": 3.798184903949822e-06, "loss": 1.9188, "step": 1721 }, { "epoch": 2.334461746784022, "grad_norm": 1.0579968475285397, "learning_rate": 3.7929495281341885e-06, "loss": 2.125, "step": 1722 }, { "epoch": 2.3358158429248475, "grad_norm": 1.2620340974069424, "learning_rate": 3.787715557208784e-06, "loss": 2.3681, "step": 1723 }, { "epoch": 2.3371699390656735, "grad_norm": 1.3509848906395134, "learning_rate": 3.782482997265445e-06, "loss": 2.1517, "step": 1724 }, { "epoch": 2.3385240352064995, "grad_norm": 2.0245960532641805, "learning_rate": 3.777251854394369e-06, "loss": 2.3628, "step": 1725 }, { "epoch": 2.3398781313473256, "grad_norm": 1.2815567329693103, "learning_rate": 3.7720221346840986e-06, "loss": 2.0548, "step": 1726 }, { "epoch": 2.3412322274881516, "grad_norm": 1.0971041556639027, "learning_rate": 3.7667938442215247e-06, "loss": 1.9249, "step": 1727 }, { "epoch": 2.3425863236289777, "grad_norm": 1.221881020739655, "learning_rate": 3.7615669890918706e-06, "loss": 2.5982, "step": 1728 }, { "epoch": 2.3439404197698037, "grad_norm": 1.2955975852266357, "learning_rate": 3.7563415753786916e-06, "loss": 1.984, "step": 1729 }, { "epoch": 2.3452945159106298, "grad_norm": 1.0487840870386045, "learning_rate": 3.751117609163865e-06, "loss": 2.3018, "step": 1730 }, { "epoch": 2.346648612051456, "grad_norm": 1.251971797662659, "learning_rate": 3.7458950965275855e-06, "loss": 2.1572, "step": 1731 }, { "epoch": 2.348002708192282, "grad_norm": 1.3139116481270519, "learning_rate": 3.7406740435483492e-06, "loss": 2.0531, "step": 1732 }, { "epoch": 2.349356804333108, "grad_norm": 1.1374306915548074, "learning_rate": 3.735454456302961e-06, "loss": 2.1181, "step": 1733 }, { "epoch": 2.3507109004739335, "grad_norm": 1.2601705526516274, "learning_rate": 3.7302363408665155e-06, "loss": 2.2112, "step": 1734 }, { "epoch": 2.3520649966147595, "grad_norm": 1.0740226683703784, "learning_rate": 3.725019703312396e-06, "loss": 1.8978, "step": 1735 }, { "epoch": 2.3534190927555856, "grad_norm": 1.3431575575124168, "learning_rate": 3.7198045497122647e-06, "loss": 2.1701, "step": 1736 }, { "epoch": 2.3547731888964116, "grad_norm": 1.2857893072832423, "learning_rate": 3.7145908861360565e-06, "loss": 2.4854, "step": 1737 }, { "epoch": 2.3561272850372377, "grad_norm": 1.3703205667792302, "learning_rate": 3.7093787186519724e-06, "loss": 2.2, "step": 1738 }, { "epoch": 2.3574813811780637, "grad_norm": 1.928102646644817, "learning_rate": 3.7041680533264733e-06, "loss": 2.1497, "step": 1739 }, { "epoch": 2.3588354773188898, "grad_norm": 1.3049363959268678, "learning_rate": 3.6989588962242694e-06, "loss": 2.5344, "step": 1740 }, { "epoch": 2.360189573459716, "grad_norm": 1.0978648777207975, "learning_rate": 3.693751253408319e-06, "loss": 2.2206, "step": 1741 }, { "epoch": 2.3615436696005414, "grad_norm": 1.2300932172689376, "learning_rate": 3.688545130939813e-06, "loss": 2.3625, "step": 1742 }, { "epoch": 2.3628977657413675, "grad_norm": 1.1478931419371352, "learning_rate": 3.683340534878176e-06, "loss": 2.222, "step": 1743 }, { "epoch": 2.3642518618821935, "grad_norm": 1.1696561423092657, "learning_rate": 3.6781374712810558e-06, "loss": 1.9408, "step": 1744 }, { "epoch": 2.3656059580230195, "grad_norm": 1.2096103440081174, "learning_rate": 3.6729359462043155e-06, "loss": 2.1108, "step": 1745 }, { "epoch": 2.3669600541638456, "grad_norm": 1.073111333155137, "learning_rate": 3.667735965702031e-06, "loss": 2.3945, "step": 1746 }, { "epoch": 2.3683141503046716, "grad_norm": 2.0067982241165008, "learning_rate": 3.6625375358264737e-06, "loss": 2.262, "step": 1747 }, { "epoch": 2.3696682464454977, "grad_norm": 1.069111491906859, "learning_rate": 3.657340662628116e-06, "loss": 2.0538, "step": 1748 }, { "epoch": 2.3710223425863237, "grad_norm": 1.3472346897288008, "learning_rate": 3.6521453521556172e-06, "loss": 2.4316, "step": 1749 }, { "epoch": 2.3723764387271498, "grad_norm": 1.4270641314977643, "learning_rate": 3.646951610455817e-06, "loss": 2.3077, "step": 1750 }, { "epoch": 2.373730534867976, "grad_norm": 1.3242480415890008, "learning_rate": 3.641759443573728e-06, "loss": 2.272, "step": 1751 }, { "epoch": 2.375084631008802, "grad_norm": 1.1664460983367577, "learning_rate": 3.6365688575525315e-06, "loss": 2.1208, "step": 1752 }, { "epoch": 2.3764387271496275, "grad_norm": 1.131827952104003, "learning_rate": 3.6313798584335696e-06, "loss": 1.9157, "step": 1753 }, { "epoch": 2.3777928232904535, "grad_norm": 1.2132213858371663, "learning_rate": 3.626192452256336e-06, "loss": 2.3505, "step": 1754 }, { "epoch": 2.3791469194312795, "grad_norm": 1.1522214826726387, "learning_rate": 3.621006645058472e-06, "loss": 2.0083, "step": 1755 }, { "epoch": 2.3805010155721056, "grad_norm": 1.2894077526829606, "learning_rate": 3.6158224428757538e-06, "loss": 2.1545, "step": 1756 }, { "epoch": 2.3818551117129316, "grad_norm": 1.6576516579103815, "learning_rate": 3.6106398517420936e-06, "loss": 2.2247, "step": 1757 }, { "epoch": 2.3832092078537577, "grad_norm": 1.948134290118367, "learning_rate": 3.605458877689528e-06, "loss": 2.0332, "step": 1758 }, { "epoch": 2.3845633039945837, "grad_norm": 1.0459980046424817, "learning_rate": 3.6002795267482093e-06, "loss": 1.9783, "step": 1759 }, { "epoch": 2.3859174001354098, "grad_norm": 1.054545662841371, "learning_rate": 3.595101804946404e-06, "loss": 2.1693, "step": 1760 }, { "epoch": 2.387271496276236, "grad_norm": 1.2411262781746522, "learning_rate": 3.5899257183104787e-06, "loss": 1.8118, "step": 1761 }, { "epoch": 2.3886255924170614, "grad_norm": 1.377181774367648, "learning_rate": 3.584751272864899e-06, "loss": 2.2351, "step": 1762 }, { "epoch": 2.3899796885578874, "grad_norm": 1.1475829572692609, "learning_rate": 3.579578474632219e-06, "loss": 2.5598, "step": 1763 }, { "epoch": 2.3913337846987135, "grad_norm": 1.1323761927513982, "learning_rate": 3.574407329633078e-06, "loss": 1.8837, "step": 1764 }, { "epoch": 2.3926878808395395, "grad_norm": 1.023764992138311, "learning_rate": 3.5692378438861907e-06, "loss": 2.3301, "step": 1765 }, { "epoch": 2.3940419769803656, "grad_norm": 1.1328303335527867, "learning_rate": 3.564070023408336e-06, "loss": 1.9072, "step": 1766 }, { "epoch": 2.3953960731211916, "grad_norm": 1.122120761868644, "learning_rate": 3.55890387421436e-06, "loss": 2.7061, "step": 1767 }, { "epoch": 2.3967501692620177, "grad_norm": 1.2280466143188116, "learning_rate": 3.553739402317162e-06, "loss": 2.1304, "step": 1768 }, { "epoch": 2.3981042654028437, "grad_norm": 1.2192909086555546, "learning_rate": 3.5485766137276894e-06, "loss": 2.0742, "step": 1769 }, { "epoch": 2.3994583615436698, "grad_norm": 1.1883813449947513, "learning_rate": 3.5434155144549274e-06, "loss": 2.093, "step": 1770 }, { "epoch": 2.400812457684496, "grad_norm": 1.1955302282581568, "learning_rate": 3.538256110505899e-06, "loss": 2.6478, "step": 1771 }, { "epoch": 2.4021665538253214, "grad_norm": 1.188535233274131, "learning_rate": 3.5330984078856512e-06, "loss": 1.9688, "step": 1772 }, { "epoch": 2.4035206499661474, "grad_norm": 1.273909260786848, "learning_rate": 3.527942412597253e-06, "loss": 2.3278, "step": 1773 }, { "epoch": 2.4048747461069735, "grad_norm": 1.2117502034405825, "learning_rate": 3.522788130641786e-06, "loss": 2.2397, "step": 1774 }, { "epoch": 2.4062288422477995, "grad_norm": 1.1271706497606906, "learning_rate": 3.517635568018335e-06, "loss": 2.051, "step": 1775 }, { "epoch": 2.4075829383886256, "grad_norm": 1.2654718407786778, "learning_rate": 3.5124847307239863e-06, "loss": 2.4296, "step": 1776 }, { "epoch": 2.4089370345294516, "grad_norm": 1.0588589902191545, "learning_rate": 3.5073356247538165e-06, "loss": 2.2326, "step": 1777 }, { "epoch": 2.4102911306702777, "grad_norm": 1.671546422118894, "learning_rate": 3.502188256100889e-06, "loss": 1.855, "step": 1778 }, { "epoch": 2.4116452268111037, "grad_norm": 1.114393579318603, "learning_rate": 3.497042630756245e-06, "loss": 2.2051, "step": 1779 }, { "epoch": 2.4129993229519298, "grad_norm": 1.0291964553913637, "learning_rate": 3.491898754708893e-06, "loss": 2.1031, "step": 1780 }, { "epoch": 2.4143534190927554, "grad_norm": 1.0236129532903953, "learning_rate": 3.4867566339458096e-06, "loss": 1.8354, "step": 1781 }, { "epoch": 2.4157075152335814, "grad_norm": 1.3818904484532941, "learning_rate": 3.4816162744519266e-06, "loss": 2.4883, "step": 1782 }, { "epoch": 2.4170616113744074, "grad_norm": 1.2568760608340415, "learning_rate": 3.4764776822101275e-06, "loss": 2.0817, "step": 1783 }, { "epoch": 2.4184157075152335, "grad_norm": 1.2344642037591336, "learning_rate": 3.471340863201237e-06, "loss": 2.0976, "step": 1784 }, { "epoch": 2.4197698036560595, "grad_norm": 1.2362548526413457, "learning_rate": 3.4662058234040152e-06, "loss": 2.2987, "step": 1785 }, { "epoch": 2.4211238997968856, "grad_norm": 1.0701064249894996, "learning_rate": 3.4610725687951537e-06, "loss": 1.8269, "step": 1786 }, { "epoch": 2.4224779959377116, "grad_norm": 1.0645902419768913, "learning_rate": 3.4559411053492653e-06, "loss": 1.957, "step": 1787 }, { "epoch": 2.4238320920785377, "grad_norm": 0.9684803955067628, "learning_rate": 3.450811439038878e-06, "loss": 2.3672, "step": 1788 }, { "epoch": 2.4251861882193637, "grad_norm": 1.105400484443266, "learning_rate": 3.4456835758344297e-06, "loss": 2.1318, "step": 1789 }, { "epoch": 2.4265402843601898, "grad_norm": 1.3041407139948142, "learning_rate": 3.440557521704256e-06, "loss": 2.2747, "step": 1790 }, { "epoch": 2.427894380501016, "grad_norm": 1.2274798424093714, "learning_rate": 3.4354332826145897e-06, "loss": 1.8344, "step": 1791 }, { "epoch": 2.4292484766418414, "grad_norm": 1.3408591122266145, "learning_rate": 3.43031086452955e-06, "loss": 2.1231, "step": 1792 }, { "epoch": 2.4306025727826674, "grad_norm": 1.2330065391440848, "learning_rate": 3.4251902734111396e-06, "loss": 2.8748, "step": 1793 }, { "epoch": 2.4319566689234935, "grad_norm": 1.1607739927831968, "learning_rate": 3.4200715152192297e-06, "loss": 2.1515, "step": 1794 }, { "epoch": 2.4333107650643195, "grad_norm": 1.2456709995423394, "learning_rate": 3.4149545959115604e-06, "loss": 2.0915, "step": 1795 }, { "epoch": 2.4346648612051456, "grad_norm": 1.2472913998217265, "learning_rate": 3.409839521443734e-06, "loss": 2.15, "step": 1796 }, { "epoch": 2.4360189573459716, "grad_norm": 1.2932408761061893, "learning_rate": 3.4047262977692014e-06, "loss": 2.1991, "step": 1797 }, { "epoch": 2.4373730534867977, "grad_norm": 1.4934208702654748, "learning_rate": 3.3996149308392633e-06, "loss": 2.238, "step": 1798 }, { "epoch": 2.4387271496276237, "grad_norm": 1.1675924017113282, "learning_rate": 3.394505426603055e-06, "loss": 2.6683, "step": 1799 }, { "epoch": 2.4400812457684493, "grad_norm": 1.0739239953562374, "learning_rate": 3.389397791007548e-06, "loss": 1.787, "step": 1800 }, { "epoch": 2.4414353419092754, "grad_norm": 1.3870663358817112, "learning_rate": 3.384292029997537e-06, "loss": 2.2352, "step": 1801 }, { "epoch": 2.4427894380501014, "grad_norm": 1.2869661007931448, "learning_rate": 3.3791881495156333e-06, "loss": 2.1586, "step": 1802 }, { "epoch": 2.4441435341909274, "grad_norm": 1.2207006258055066, "learning_rate": 3.374086155502264e-06, "loss": 2.0053, "step": 1803 }, { "epoch": 2.4454976303317535, "grad_norm": 1.1625296726355647, "learning_rate": 3.3689860538956547e-06, "loss": 1.9323, "step": 1804 }, { "epoch": 2.4468517264725795, "grad_norm": 1.3206523663923264, "learning_rate": 3.363887850631833e-06, "loss": 2.1342, "step": 1805 }, { "epoch": 2.4482058226134056, "grad_norm": 1.0980080592464498, "learning_rate": 3.3587915516446144e-06, "loss": 2.1337, "step": 1806 }, { "epoch": 2.4495599187542316, "grad_norm": 1.3835132946685904, "learning_rate": 3.3536971628655997e-06, "loss": 1.9523, "step": 1807 }, { "epoch": 2.4509140148950577, "grad_norm": 1.2279153775519225, "learning_rate": 3.3486046902241663e-06, "loss": 1.8501, "step": 1808 }, { "epoch": 2.4522681110358837, "grad_norm": 1.0237715488689143, "learning_rate": 3.3435141396474593e-06, "loss": 2.4008, "step": 1809 }, { "epoch": 2.4536222071767098, "grad_norm": 1.064898768520828, "learning_rate": 3.3384255170603886e-06, "loss": 2.5488, "step": 1810 }, { "epoch": 2.4549763033175354, "grad_norm": 1.1844761405317972, "learning_rate": 3.3333388283856195e-06, "loss": 2.4141, "step": 1811 }, { "epoch": 2.4563303994583614, "grad_norm": 1.1842625365482378, "learning_rate": 3.328254079543568e-06, "loss": 2.2771, "step": 1812 }, { "epoch": 2.4576844955991874, "grad_norm": 1.2134105770216446, "learning_rate": 3.3231712764523895e-06, "loss": 1.8451, "step": 1813 }, { "epoch": 2.4590385917400135, "grad_norm": 1.3240334840150958, "learning_rate": 3.3180904250279753e-06, "loss": 2.3674, "step": 1814 }, { "epoch": 2.4603926878808395, "grad_norm": 1.2138349626864064, "learning_rate": 3.313011531183947e-06, "loss": 2.1554, "step": 1815 }, { "epoch": 2.4617467840216656, "grad_norm": 1.0571299498177822, "learning_rate": 3.307934600831648e-06, "loss": 1.7158, "step": 1816 }, { "epoch": 2.4631008801624916, "grad_norm": 1.2045912258333282, "learning_rate": 3.3028596398801337e-06, "loss": 2.1693, "step": 1817 }, { "epoch": 2.4644549763033177, "grad_norm": 1.1102595091465026, "learning_rate": 3.297786654236169e-06, "loss": 2.4208, "step": 1818 }, { "epoch": 2.4658090724441437, "grad_norm": 1.1861190363523093, "learning_rate": 3.2927156498042203e-06, "loss": 2.3872, "step": 1819 }, { "epoch": 2.4671631685849693, "grad_norm": 1.4350957301927922, "learning_rate": 3.287646632486447e-06, "loss": 2.4011, "step": 1820 }, { "epoch": 2.4685172647257954, "grad_norm": 1.1412223208517722, "learning_rate": 3.2825796081826943e-06, "loss": 2.326, "step": 1821 }, { "epoch": 2.4698713608666214, "grad_norm": 1.3258122511636292, "learning_rate": 3.2775145827904907e-06, "loss": 2.5215, "step": 1822 }, { "epoch": 2.4712254570074474, "grad_norm": 1.1041321084851496, "learning_rate": 3.2724515622050373e-06, "loss": 1.9432, "step": 1823 }, { "epoch": 2.4725795531482735, "grad_norm": 1.1587251997975352, "learning_rate": 3.2673905523192e-06, "loss": 2.2471, "step": 1824 }, { "epoch": 2.4739336492890995, "grad_norm": 1.4836238535722521, "learning_rate": 3.2623315590235076e-06, "loss": 1.9917, "step": 1825 }, { "epoch": 2.4752877454299256, "grad_norm": 1.14182181918578, "learning_rate": 3.2572745882061376e-06, "loss": 2.2237, "step": 1826 }, { "epoch": 2.4766418415707516, "grad_norm": 1.0631712784159546, "learning_rate": 3.2522196457529174e-06, "loss": 1.8656, "step": 1827 }, { "epoch": 2.4779959377115777, "grad_norm": 1.3176068549747726, "learning_rate": 3.2471667375473116e-06, "loss": 2.3724, "step": 1828 }, { "epoch": 2.4793500338524037, "grad_norm": 1.1232164894180032, "learning_rate": 3.242115869470418e-06, "loss": 1.9391, "step": 1829 }, { "epoch": 2.4807041299932298, "grad_norm": 1.6876891815018402, "learning_rate": 3.237067047400959e-06, "loss": 2.7208, "step": 1830 }, { "epoch": 2.4820582261340554, "grad_norm": 1.2098903327193213, "learning_rate": 3.2320202772152754e-06, "loss": 2.4239, "step": 1831 }, { "epoch": 2.4834123222748814, "grad_norm": 1.0916100535576414, "learning_rate": 3.226975564787322e-06, "loss": 2.2478, "step": 1832 }, { "epoch": 2.4847664184157074, "grad_norm": 1.1298440911272671, "learning_rate": 3.2219329159886557e-06, "loss": 1.8048, "step": 1833 }, { "epoch": 2.4861205145565335, "grad_norm": 1.069974876640358, "learning_rate": 3.216892336688435e-06, "loss": 2.2146, "step": 1834 }, { "epoch": 2.4874746106973595, "grad_norm": 1.268266227744998, "learning_rate": 3.211853832753406e-06, "loss": 2.0349, "step": 1835 }, { "epoch": 2.4888287068381856, "grad_norm": 1.0951091814234022, "learning_rate": 3.2068174100479003e-06, "loss": 1.5965, "step": 1836 }, { "epoch": 2.4901828029790116, "grad_norm": 1.0676619961656297, "learning_rate": 3.201783074433829e-06, "loss": 1.6985, "step": 1837 }, { "epoch": 2.4915368991198377, "grad_norm": 1.1927517486852435, "learning_rate": 3.1967508317706726e-06, "loss": 1.9756, "step": 1838 }, { "epoch": 2.4928909952606633, "grad_norm": 1.1506662054446082, "learning_rate": 3.1917206879154762e-06, "loss": 2.1803, "step": 1839 }, { "epoch": 2.4942450914014893, "grad_norm": 1.0911188712770328, "learning_rate": 3.18669264872284e-06, "loss": 2.4311, "step": 1840 }, { "epoch": 2.4955991875423154, "grad_norm": 1.2190515392350634, "learning_rate": 3.181666720044918e-06, "loss": 2.1749, "step": 1841 }, { "epoch": 2.4969532836831414, "grad_norm": 1.1062369100585125, "learning_rate": 3.176642907731404e-06, "loss": 2.1382, "step": 1842 }, { "epoch": 2.4983073798239674, "grad_norm": 1.2694735723130735, "learning_rate": 3.1716212176295336e-06, "loss": 2.5935, "step": 1843 }, { "epoch": 2.4996614759647935, "grad_norm": 1.18899966550945, "learning_rate": 3.1666016555840674e-06, "loss": 1.9713, "step": 1844 }, { "epoch": 2.5010155721056195, "grad_norm": 1.244795978447979, "learning_rate": 3.161584227437291e-06, "loss": 1.9336, "step": 1845 }, { "epoch": 2.5023696682464456, "grad_norm": 1.3105695737694865, "learning_rate": 3.1565689390290067e-06, "loss": 2.4228, "step": 1846 }, { "epoch": 2.5037237643872716, "grad_norm": 1.2083854762838475, "learning_rate": 3.1515557961965254e-06, "loss": 1.9489, "step": 1847 }, { "epoch": 2.5050778605280977, "grad_norm": 1.1732618113199662, "learning_rate": 3.1465448047746626e-06, "loss": 2.374, "step": 1848 }, { "epoch": 2.5064319566689237, "grad_norm": 1.1559715411605875, "learning_rate": 3.1415359705957286e-06, "loss": 1.9008, "step": 1849 }, { "epoch": 2.5077860528097498, "grad_norm": 1.0952765578007255, "learning_rate": 3.136529299489522e-06, "loss": 2.1997, "step": 1850 }, { "epoch": 2.5091401489505754, "grad_norm": 1.1724112854770499, "learning_rate": 3.1315247972833245e-06, "loss": 1.97, "step": 1851 }, { "epoch": 2.5104942450914014, "grad_norm": 0.8630387077179754, "learning_rate": 3.1265224698018947e-06, "loss": 2.0949, "step": 1852 }, { "epoch": 2.5118483412322274, "grad_norm": 1.309954867244814, "learning_rate": 3.1215223228674587e-06, "loss": 1.8376, "step": 1853 }, { "epoch": 2.5132024373730535, "grad_norm": 1.1334346803585604, "learning_rate": 3.1165243622997038e-06, "loss": 2.3345, "step": 1854 }, { "epoch": 2.5145565335138795, "grad_norm": 1.162759568561119, "learning_rate": 3.111528593915774e-06, "loss": 2.3126, "step": 1855 }, { "epoch": 2.5159106296547056, "grad_norm": 0.9929365078394864, "learning_rate": 3.106535023530262e-06, "loss": 1.9735, "step": 1856 }, { "epoch": 2.5172647257955316, "grad_norm": 1.159051939882048, "learning_rate": 3.1015436569552014e-06, "loss": 1.9053, "step": 1857 }, { "epoch": 2.518618821936357, "grad_norm": 1.2404297463043783, "learning_rate": 3.0965545000000617e-06, "loss": 2.006, "step": 1858 }, { "epoch": 2.5199729180771833, "grad_norm": 1.4812426930782647, "learning_rate": 3.0915675584717385e-06, "loss": 2.3902, "step": 1859 }, { "epoch": 2.5213270142180093, "grad_norm": 1.1813991851469383, "learning_rate": 3.0865828381745515e-06, "loss": 2.6203, "step": 1860 }, { "epoch": 2.5226811103588354, "grad_norm": 0.9815495710330431, "learning_rate": 3.0816003449102337e-06, "loss": 2.1236, "step": 1861 }, { "epoch": 2.5240352064996614, "grad_norm": 1.121459487252647, "learning_rate": 3.076620084477926e-06, "loss": 2.3479, "step": 1862 }, { "epoch": 2.5253893026404874, "grad_norm": 1.2269898613029497, "learning_rate": 3.0716420626741716e-06, "loss": 1.9193, "step": 1863 }, { "epoch": 2.5267433987813135, "grad_norm": 1.0125323399813895, "learning_rate": 3.0666662852929063e-06, "loss": 2.5693, "step": 1864 }, { "epoch": 2.5280974949221395, "grad_norm": 1.0792380250318991, "learning_rate": 3.0616927581254554e-06, "loss": 2.0622, "step": 1865 }, { "epoch": 2.5294515910629656, "grad_norm": 1.0503014257482366, "learning_rate": 3.056721486960523e-06, "loss": 2.2628, "step": 1866 }, { "epoch": 2.5308056872037916, "grad_norm": 4.333532829084429, "learning_rate": 3.051752477584191e-06, "loss": 2.1919, "step": 1867 }, { "epoch": 2.5321597833446177, "grad_norm": 1.4015593424528954, "learning_rate": 3.046785735779906e-06, "loss": 2.1371, "step": 1868 }, { "epoch": 2.5335138794854437, "grad_norm": 1.1198530962176534, "learning_rate": 3.041821267328475e-06, "loss": 1.9688, "step": 1869 }, { "epoch": 2.5348679756262693, "grad_norm": 1.1952140985764694, "learning_rate": 3.0368590780080605e-06, "loss": 1.9966, "step": 1870 }, { "epoch": 2.5362220717670954, "grad_norm": 1.1294127406898105, "learning_rate": 3.0318991735941716e-06, "loss": 1.985, "step": 1871 }, { "epoch": 2.5375761679079214, "grad_norm": 1.1718681027855775, "learning_rate": 3.0269415598596604e-06, "loss": 2.123, "step": 1872 }, { "epoch": 2.5389302640487474, "grad_norm": 1.4123573867095767, "learning_rate": 3.021986242574707e-06, "loss": 2.1868, "step": 1873 }, { "epoch": 2.5402843601895735, "grad_norm": 1.037472327238385, "learning_rate": 3.0170332275068247e-06, "loss": 2.0131, "step": 1874 }, { "epoch": 2.5416384563303995, "grad_norm": 1.0794105338283817, "learning_rate": 3.012082520420845e-06, "loss": 2.3318, "step": 1875 }, { "epoch": 2.5429925524712256, "grad_norm": 1.291675825491076, "learning_rate": 3.0071341270789123e-06, "loss": 2.3596, "step": 1876 }, { "epoch": 2.5443466486120516, "grad_norm": 1.4631630014723893, "learning_rate": 3.0021880532404803e-06, "loss": 1.8752, "step": 1877 }, { "epoch": 2.545700744752877, "grad_norm": 1.2734557737543577, "learning_rate": 2.997244304662299e-06, "loss": 1.9219, "step": 1878 }, { "epoch": 2.5470548408937033, "grad_norm": 1.2693507698156585, "learning_rate": 2.992302887098415e-06, "loss": 2.7161, "step": 1879 }, { "epoch": 2.5484089370345293, "grad_norm": 1.0924842417286675, "learning_rate": 2.9873638063001633e-06, "loss": 2.1427, "step": 1880 }, { "epoch": 2.5497630331753554, "grad_norm": 1.0730832298666715, "learning_rate": 2.982427068016155e-06, "loss": 1.8215, "step": 1881 }, { "epoch": 2.5511171293161814, "grad_norm": 1.1960864801585642, "learning_rate": 2.9774926779922795e-06, "loss": 2.0656, "step": 1882 }, { "epoch": 2.5524712254570074, "grad_norm": 1.2631705112654903, "learning_rate": 2.972560641971688e-06, "loss": 2.3724, "step": 1883 }, { "epoch": 2.5538253215978335, "grad_norm": 1.087500374115741, "learning_rate": 2.967630965694794e-06, "loss": 2.371, "step": 1884 }, { "epoch": 2.5551794177386595, "grad_norm": 1.179742189093041, "learning_rate": 2.9627036548992673e-06, "loss": 1.8838, "step": 1885 }, { "epoch": 2.5565335138794856, "grad_norm": 1.2264409137324177, "learning_rate": 2.95777871532002e-06, "loss": 2.4891, "step": 1886 }, { "epoch": 2.5578876100203116, "grad_norm": 1.9335576653791682, "learning_rate": 2.952856152689208e-06, "loss": 2.2255, "step": 1887 }, { "epoch": 2.5592417061611377, "grad_norm": 1.214938764580777, "learning_rate": 2.947935972736217e-06, "loss": 1.7472, "step": 1888 }, { "epoch": 2.5605958023019637, "grad_norm": 1.2883537676407941, "learning_rate": 2.943018181187663e-06, "loss": 2.2533, "step": 1889 }, { "epoch": 2.5619498984427893, "grad_norm": 1.1451384643522544, "learning_rate": 2.93810278376738e-06, "loss": 2.5156, "step": 1890 }, { "epoch": 2.5633039945836154, "grad_norm": 1.1255912053122272, "learning_rate": 2.9331897861964165e-06, "loss": 1.7971, "step": 1891 }, { "epoch": 2.5646580907244414, "grad_norm": 1.0156856376793972, "learning_rate": 2.928279194193028e-06, "loss": 2.7232, "step": 1892 }, { "epoch": 2.5660121868652674, "grad_norm": 1.2661251663190862, "learning_rate": 2.923371013472667e-06, "loss": 2.2971, "step": 1893 }, { "epoch": 2.5673662830060935, "grad_norm": 1.0369774871786672, "learning_rate": 2.918465249747986e-06, "loss": 1.9141, "step": 1894 }, { "epoch": 2.5687203791469195, "grad_norm": 1.1449622386811824, "learning_rate": 2.9135619087288153e-06, "loss": 1.741, "step": 1895 }, { "epoch": 2.5700744752877456, "grad_norm": 1.0876296591590318, "learning_rate": 2.9086609961221758e-06, "loss": 1.7262, "step": 1896 }, { "epoch": 2.571428571428571, "grad_norm": 1.1113713629846054, "learning_rate": 2.9037625176322524e-06, "loss": 1.7936, "step": 1897 }, { "epoch": 2.572782667569397, "grad_norm": 1.027983684542803, "learning_rate": 2.898866478960405e-06, "loss": 2.3066, "step": 1898 }, { "epoch": 2.5741367637102233, "grad_norm": 1.093099539697463, "learning_rate": 2.893972885805148e-06, "loss": 2.3458, "step": 1899 }, { "epoch": 2.5754908598510493, "grad_norm": 1.182253093626154, "learning_rate": 2.88908174386215e-06, "loss": 1.9524, "step": 1900 }, { "epoch": 2.5768449559918754, "grad_norm": 1.0884038048228366, "learning_rate": 2.884193058824231e-06, "loss": 2.4982, "step": 1901 }, { "epoch": 2.5781990521327014, "grad_norm": 0.9579106114604485, "learning_rate": 2.879306836381345e-06, "loss": 2.2826, "step": 1902 }, { "epoch": 2.5795531482735274, "grad_norm": 1.2345913942874547, "learning_rate": 2.874423082220588e-06, "loss": 2.0347, "step": 1903 }, { "epoch": 2.5809072444143535, "grad_norm": 1.216905641002978, "learning_rate": 2.8695418020261755e-06, "loss": 2.1111, "step": 1904 }, { "epoch": 2.5822613405551795, "grad_norm": 1.0608513502716586, "learning_rate": 2.8646630014794446e-06, "loss": 2.2819, "step": 1905 }, { "epoch": 2.5836154366960056, "grad_norm": 1.2559204505085202, "learning_rate": 2.8597866862588498e-06, "loss": 2.3679, "step": 1906 }, { "epoch": 2.5849695328368316, "grad_norm": 1.242129760306491, "learning_rate": 2.854912862039955e-06, "loss": 2.2504, "step": 1907 }, { "epoch": 2.5863236289776577, "grad_norm": 1.1763076755156585, "learning_rate": 2.8500415344954147e-06, "loss": 2.4541, "step": 1908 }, { "epoch": 2.5876777251184833, "grad_norm": 1.152476534671673, "learning_rate": 2.845172709294989e-06, "loss": 1.7831, "step": 1909 }, { "epoch": 2.5890318212593093, "grad_norm": 1.0544805656899612, "learning_rate": 2.8403063921055166e-06, "loss": 1.9058, "step": 1910 }, { "epoch": 2.5903859174001354, "grad_norm": 1.5062625664066123, "learning_rate": 2.8354425885909252e-06, "loss": 2.2472, "step": 1911 }, { "epoch": 2.5917400135409614, "grad_norm": 1.2435609781023294, "learning_rate": 2.83058130441221e-06, "loss": 2.518, "step": 1912 }, { "epoch": 2.5930941096817874, "grad_norm": 1.2985177022376626, "learning_rate": 2.825722545227435e-06, "loss": 2.1948, "step": 1913 }, { "epoch": 2.5944482058226135, "grad_norm": 1.4010721969881315, "learning_rate": 2.82086631669173e-06, "loss": 1.9286, "step": 1914 }, { "epoch": 2.5958023019634395, "grad_norm": 1.0138835708229614, "learning_rate": 2.816012624457273e-06, "loss": 1.854, "step": 1915 }, { "epoch": 2.597156398104265, "grad_norm": 1.5554149350648236, "learning_rate": 2.8111614741732975e-06, "loss": 2.4939, "step": 1916 }, { "epoch": 2.598510494245091, "grad_norm": 1.2132520888909704, "learning_rate": 2.806312871486071e-06, "loss": 2.1247, "step": 1917 }, { "epoch": 2.599864590385917, "grad_norm": 1.1500891152278072, "learning_rate": 2.8014668220388986e-06, "loss": 2.5258, "step": 1918 }, { "epoch": 2.6012186865267433, "grad_norm": 1.2326784570791647, "learning_rate": 2.796623331472116e-06, "loss": 1.9567, "step": 1919 }, { "epoch": 2.6025727826675693, "grad_norm": 1.0507124423257452, "learning_rate": 2.7917824054230787e-06, "loss": 1.834, "step": 1920 }, { "epoch": 2.6039268788083954, "grad_norm": 1.0650265976083109, "learning_rate": 2.7869440495261535e-06, "loss": 2.2804, "step": 1921 }, { "epoch": 2.6052809749492214, "grad_norm": 1.234328733445062, "learning_rate": 2.7821082694127255e-06, "loss": 2.143, "step": 1922 }, { "epoch": 2.6066350710900474, "grad_norm": 1.2422589520152854, "learning_rate": 2.77727507071117e-06, "loss": 1.5843, "step": 1923 }, { "epoch": 2.6079891672308735, "grad_norm": 1.1234569860296861, "learning_rate": 2.7724444590468692e-06, "loss": 1.7655, "step": 1924 }, { "epoch": 2.6093432633716995, "grad_norm": 1.2012955311309237, "learning_rate": 2.7676164400421864e-06, "loss": 2.446, "step": 1925 }, { "epoch": 2.6106973595125256, "grad_norm": 1.015266989628815, "learning_rate": 2.762791019316467e-06, "loss": 2.0344, "step": 1926 }, { "epoch": 2.6120514556533516, "grad_norm": 1.1395783437156008, "learning_rate": 2.7579682024860395e-06, "loss": 2.1059, "step": 1927 }, { "epoch": 2.6134055517941777, "grad_norm": 1.5328388130326827, "learning_rate": 2.7531479951641928e-06, "loss": 2.5828, "step": 1928 }, { "epoch": 2.6147596479350033, "grad_norm": 1.2975386707402667, "learning_rate": 2.748330402961188e-06, "loss": 1.8261, "step": 1929 }, { "epoch": 2.6161137440758293, "grad_norm": 1.0936917651584523, "learning_rate": 2.7435154314842337e-06, "loss": 1.8666, "step": 1930 }, { "epoch": 2.6174678402166554, "grad_norm": 1.034196500585431, "learning_rate": 2.7387030863374906e-06, "loss": 1.9812, "step": 1931 }, { "epoch": 2.6188219363574814, "grad_norm": 1.2351778752000186, "learning_rate": 2.7338933731220687e-06, "loss": 2.1411, "step": 1932 }, { "epoch": 2.6201760324983074, "grad_norm": 1.1997522273311612, "learning_rate": 2.729086297436005e-06, "loss": 2.5226, "step": 1933 }, { "epoch": 2.6215301286391335, "grad_norm": 1.1054297747923265, "learning_rate": 2.7242818648742754e-06, "loss": 2.4148, "step": 1934 }, { "epoch": 2.6228842247799595, "grad_norm": 1.2701961141032059, "learning_rate": 2.719480081028774e-06, "loss": 2.7507, "step": 1935 }, { "epoch": 2.624238320920785, "grad_norm": 1.094893145191458, "learning_rate": 2.714680951488312e-06, "loss": 2.1012, "step": 1936 }, { "epoch": 2.625592417061611, "grad_norm": 1.2433123218269542, "learning_rate": 2.7098844818386164e-06, "loss": 1.8294, "step": 1937 }, { "epoch": 2.626946513202437, "grad_norm": 1.0334713107594276, "learning_rate": 2.705090677662311e-06, "loss": 1.9691, "step": 1938 }, { "epoch": 2.6283006093432633, "grad_norm": 1.0894188781803227, "learning_rate": 2.700299544538926e-06, "loss": 2.0785, "step": 1939 }, { "epoch": 2.6296547054840893, "grad_norm": 1.0428258082041535, "learning_rate": 2.695511088044873e-06, "loss": 2.2801, "step": 1940 }, { "epoch": 2.6310088016249153, "grad_norm": 1.049307730170318, "learning_rate": 2.690725313753455e-06, "loss": 2.018, "step": 1941 }, { "epoch": 2.6323628977657414, "grad_norm": 1.3005063942782003, "learning_rate": 2.685942227234852e-06, "loss": 1.9157, "step": 1942 }, { "epoch": 2.6337169939065674, "grad_norm": 1.1765678048494919, "learning_rate": 2.681161834056113e-06, "loss": 1.9543, "step": 1943 }, { "epoch": 2.6350710900473935, "grad_norm": 1.469697862763876, "learning_rate": 2.6763841397811576e-06, "loss": 2.7294, "step": 1944 }, { "epoch": 2.6364251861882195, "grad_norm": 1.3778679025735157, "learning_rate": 2.6716091499707586e-06, "loss": 2.329, "step": 1945 }, { "epoch": 2.6377792823290456, "grad_norm": 1.2097180303807733, "learning_rate": 2.6668368701825413e-06, "loss": 1.7027, "step": 1946 }, { "epoch": 2.6391333784698716, "grad_norm": 1.414983316196713, "learning_rate": 2.662067305970983e-06, "loss": 1.9115, "step": 1947 }, { "epoch": 2.640487474610697, "grad_norm": 1.2825199854627847, "learning_rate": 2.6573004628873912e-06, "loss": 2.1633, "step": 1948 }, { "epoch": 2.6418415707515233, "grad_norm": 1.252941499373871, "learning_rate": 2.6525363464799157e-06, "loss": 2.3245, "step": 1949 }, { "epoch": 2.6431956668923493, "grad_norm": 1.2157544999566485, "learning_rate": 2.647774962293526e-06, "loss": 2.4426, "step": 1950 }, { "epoch": 2.6445497630331753, "grad_norm": 1.278841031544822, "learning_rate": 2.6430163158700116e-06, "loss": 2.4591, "step": 1951 }, { "epoch": 2.6459038591740014, "grad_norm": 1.3220629342969958, "learning_rate": 2.6382604127479815e-06, "loss": 2.6564, "step": 1952 }, { "epoch": 2.6472579553148274, "grad_norm": 1.195027394931894, "learning_rate": 2.633507258462845e-06, "loss": 2.1951, "step": 1953 }, { "epoch": 2.6486120514556535, "grad_norm": 1.2056547583795174, "learning_rate": 2.628756858546817e-06, "loss": 2.3016, "step": 1954 }, { "epoch": 2.649966147596479, "grad_norm": 1.1821670168326954, "learning_rate": 2.6240092185289034e-06, "loss": 1.8315, "step": 1955 }, { "epoch": 2.651320243737305, "grad_norm": 1.3540933011230412, "learning_rate": 2.6192643439348965e-06, "loss": 2.6121, "step": 1956 }, { "epoch": 2.652674339878131, "grad_norm": 1.32198429501895, "learning_rate": 2.614522240287377e-06, "loss": 1.9956, "step": 1957 }, { "epoch": 2.654028436018957, "grad_norm": 1.119914562125602, "learning_rate": 2.609782913105691e-06, "loss": 1.6065, "step": 1958 }, { "epoch": 2.6553825321597833, "grad_norm": 1.2055737577352175, "learning_rate": 2.6050463679059625e-06, "loss": 2.0703, "step": 1959 }, { "epoch": 2.6567366283006093, "grad_norm": 1.2873447778780456, "learning_rate": 2.6003126102010696e-06, "loss": 2.2986, "step": 1960 }, { "epoch": 2.6580907244414353, "grad_norm": 1.194148834870276, "learning_rate": 2.595581645500649e-06, "loss": 2.0265, "step": 1961 }, { "epoch": 2.6594448205822614, "grad_norm": 1.270449850429165, "learning_rate": 2.5908534793110897e-06, "loss": 1.8929, "step": 1962 }, { "epoch": 2.6607989167230874, "grad_norm": 1.1071026496916718, "learning_rate": 2.5861281171355197e-06, "loss": 2.0383, "step": 1963 }, { "epoch": 2.6621530128639135, "grad_norm": 1.1373829690742603, "learning_rate": 2.5814055644738013e-06, "loss": 2.0847, "step": 1964 }, { "epoch": 2.6635071090047395, "grad_norm": 1.319445119132419, "learning_rate": 2.576685826822535e-06, "loss": 2.113, "step": 1965 }, { "epoch": 2.6648612051455656, "grad_norm": 1.4753849357220579, "learning_rate": 2.5719689096750356e-06, "loss": 2.0612, "step": 1966 }, { "epoch": 2.666215301286391, "grad_norm": 1.1861021373714846, "learning_rate": 2.5672548185213426e-06, "loss": 2.2497, "step": 1967 }, { "epoch": 2.667569397427217, "grad_norm": 1.1862868294572253, "learning_rate": 2.5625435588482017e-06, "loss": 2.0994, "step": 1968 }, { "epoch": 2.6689234935680433, "grad_norm": 1.1810442166146469, "learning_rate": 2.557835136139062e-06, "loss": 2.5427, "step": 1969 }, { "epoch": 2.6702775897088693, "grad_norm": 1.3777743229545107, "learning_rate": 2.553129555874077e-06, "loss": 2.2, "step": 1970 }, { "epoch": 2.6716316858496953, "grad_norm": 1.2001825535567758, "learning_rate": 2.5484268235300847e-06, "loss": 2.5004, "step": 1971 }, { "epoch": 2.6729857819905214, "grad_norm": 1.5286836748977617, "learning_rate": 2.5437269445806146e-06, "loss": 1.7329, "step": 1972 }, { "epoch": 2.6743398781313474, "grad_norm": 1.352355703126981, "learning_rate": 2.53902992449587e-06, "loss": 2.1822, "step": 1973 }, { "epoch": 2.6756939742721735, "grad_norm": 1.2957383447397062, "learning_rate": 2.5343357687427274e-06, "loss": 2.0813, "step": 1974 }, { "epoch": 2.677048070412999, "grad_norm": 1.4030280737220668, "learning_rate": 2.5296444827847355e-06, "loss": 2.3392, "step": 1975 }, { "epoch": 2.678402166553825, "grad_norm": 1.1224666256266516, "learning_rate": 2.524956072082093e-06, "loss": 2.1601, "step": 1976 }, { "epoch": 2.679756262694651, "grad_norm": 1.1849188836131852, "learning_rate": 2.520270542091663e-06, "loss": 2.3316, "step": 1977 }, { "epoch": 2.681110358835477, "grad_norm": 1.181908261397778, "learning_rate": 2.5155878982669463e-06, "loss": 1.8901, "step": 1978 }, { "epoch": 2.6824644549763033, "grad_norm": 1.350032736695083, "learning_rate": 2.5109081460580875e-06, "loss": 1.7695, "step": 1979 }, { "epoch": 2.6838185511171293, "grad_norm": 1.092635129761096, "learning_rate": 2.5062312909118695e-06, "loss": 1.8718, "step": 1980 }, { "epoch": 2.6851726472579553, "grad_norm": 1.1151256999149468, "learning_rate": 2.501557338271697e-06, "loss": 2.2951, "step": 1981 }, { "epoch": 2.6865267433987814, "grad_norm": 1.3235428206914133, "learning_rate": 2.496886293577604e-06, "loss": 2.1622, "step": 1982 }, { "epoch": 2.6878808395396074, "grad_norm": 1.1857430087640528, "learning_rate": 2.4922181622662327e-06, "loss": 2.0714, "step": 1983 }, { "epoch": 2.6892349356804335, "grad_norm": 1.1354138420997812, "learning_rate": 2.4875529497708356e-06, "loss": 2.0024, "step": 1984 }, { "epoch": 2.6905890318212595, "grad_norm": 1.2885985123313086, "learning_rate": 2.4828906615212737e-06, "loss": 2.2063, "step": 1985 }, { "epoch": 2.6919431279620856, "grad_norm": 1.1832060463895393, "learning_rate": 2.478231302943997e-06, "loss": 1.8353, "step": 1986 }, { "epoch": 2.693297224102911, "grad_norm": 1.079393913086519, "learning_rate": 2.473574879462052e-06, "loss": 2.6435, "step": 1987 }, { "epoch": 2.694651320243737, "grad_norm": 1.1811276695001673, "learning_rate": 2.4689213964950652e-06, "loss": 1.8575, "step": 1988 }, { "epoch": 2.6960054163845633, "grad_norm": 1.185223943618491, "learning_rate": 2.46427085945924e-06, "loss": 2.132, "step": 1989 }, { "epoch": 2.6973595125253893, "grad_norm": 1.1761787211591375, "learning_rate": 2.4596232737673544e-06, "loss": 2.3717, "step": 1990 }, { "epoch": 2.6987136086662153, "grad_norm": 1.2656081981249776, "learning_rate": 2.454978644828747e-06, "loss": 2.0824, "step": 1991 }, { "epoch": 2.7000677048070414, "grad_norm": 1.4669653270141612, "learning_rate": 2.450336978049322e-06, "loss": 1.9719, "step": 1992 }, { "epoch": 2.7014218009478674, "grad_norm": 0.9045670377260661, "learning_rate": 2.445698278831528e-06, "loss": 1.7157, "step": 1993 }, { "epoch": 2.702775897088693, "grad_norm": 1.0078666373099046, "learning_rate": 2.4410625525743624e-06, "loss": 2.3803, "step": 1994 }, { "epoch": 2.704129993229519, "grad_norm": 1.2630842527041577, "learning_rate": 2.4364298046733664e-06, "loss": 2.1814, "step": 1995 }, { "epoch": 2.705484089370345, "grad_norm": 1.179923885219862, "learning_rate": 2.4318000405206084e-06, "loss": 2.0703, "step": 1996 }, { "epoch": 2.706838185511171, "grad_norm": 1.0660454800548254, "learning_rate": 2.427173265504687e-06, "loss": 2.215, "step": 1997 }, { "epoch": 2.708192281651997, "grad_norm": 1.1679924054564998, "learning_rate": 2.422549485010725e-06, "loss": 2.1496, "step": 1998 }, { "epoch": 2.7095463777928233, "grad_norm": 1.118233437691708, "learning_rate": 2.4179287044203525e-06, "loss": 2.0787, "step": 1999 }, { "epoch": 2.7109004739336493, "grad_norm": 1.2342255039535306, "learning_rate": 2.4133109291117156e-06, "loss": 2.2956, "step": 2000 }, { "epoch": 2.7122545700744753, "grad_norm": 1.4100237119025654, "learning_rate": 2.4086961644594573e-06, "loss": 2.4108, "step": 2001 }, { "epoch": 2.7136086662153014, "grad_norm": 1.1904324271308306, "learning_rate": 2.4040844158347166e-06, "loss": 1.9077, "step": 2002 }, { "epoch": 2.7149627623561274, "grad_norm": 1.241743591889358, "learning_rate": 2.3994756886051267e-06, "loss": 2.4324, "step": 2003 }, { "epoch": 2.7163168584969535, "grad_norm": 1.1351580442565894, "learning_rate": 2.3948699881348003e-06, "loss": 1.8315, "step": 2004 }, { "epoch": 2.7176709546377795, "grad_norm": 1.1479599380366055, "learning_rate": 2.3902673197843253e-06, "loss": 2.3356, "step": 2005 }, { "epoch": 2.719025050778605, "grad_norm": 1.1688098262322058, "learning_rate": 2.385667688910768e-06, "loss": 1.8856, "step": 2006 }, { "epoch": 2.720379146919431, "grad_norm": 1.301384869557324, "learning_rate": 2.3810711008676495e-06, "loss": 2.0653, "step": 2007 }, { "epoch": 2.721733243060257, "grad_norm": 1.1458551631024618, "learning_rate": 2.37647756100496e-06, "loss": 2.0737, "step": 2008 }, { "epoch": 2.7230873392010833, "grad_norm": 1.1573926533607735, "learning_rate": 2.3718870746691336e-06, "loss": 2.1739, "step": 2009 }, { "epoch": 2.7244414353419093, "grad_norm": 1.3536133768478036, "learning_rate": 2.3672996472030523e-06, "loss": 2.273, "step": 2010 }, { "epoch": 2.7257955314827353, "grad_norm": 1.2039966447477246, "learning_rate": 2.3627152839460417e-06, "loss": 2.0116, "step": 2011 }, { "epoch": 2.7271496276235614, "grad_norm": 1.2188846162819493, "learning_rate": 2.358133990233854e-06, "loss": 1.9059, "step": 2012 }, { "epoch": 2.7285037237643874, "grad_norm": 1.2192208362258712, "learning_rate": 2.353555771398678e-06, "loss": 2.0995, "step": 2013 }, { "epoch": 2.729857819905213, "grad_norm": 1.280320922357782, "learning_rate": 2.3489806327691156e-06, "loss": 1.8701, "step": 2014 }, { "epoch": 2.731211916046039, "grad_norm": 1.1966461518114693, "learning_rate": 2.344408579670185e-06, "loss": 2.0868, "step": 2015 }, { "epoch": 2.732566012186865, "grad_norm": 1.219227531579492, "learning_rate": 2.339839617423318e-06, "loss": 2.0725, "step": 2016 }, { "epoch": 2.733920108327691, "grad_norm": 1.112848624813568, "learning_rate": 2.335273751346342e-06, "loss": 2.2796, "step": 2017 }, { "epoch": 2.735274204468517, "grad_norm": 1.255617736150562, "learning_rate": 2.330710986753488e-06, "loss": 2.1968, "step": 2018 }, { "epoch": 2.7366283006093433, "grad_norm": 1.3126109137289483, "learning_rate": 2.326151328955372e-06, "loss": 2.261, "step": 2019 }, { "epoch": 2.7379823967501693, "grad_norm": 1.2433604905579898, "learning_rate": 2.321594783258994e-06, "loss": 2.0331, "step": 2020 }, { "epoch": 2.7393364928909953, "grad_norm": 1.2415506634843163, "learning_rate": 2.3170413549677367e-06, "loss": 2.1415, "step": 2021 }, { "epoch": 2.7406905890318214, "grad_norm": 1.193109130964379, "learning_rate": 2.312491049381347e-06, "loss": 2.2734, "step": 2022 }, { "epoch": 2.7420446851726474, "grad_norm": 1.0282353629728356, "learning_rate": 2.307943871795945e-06, "loss": 2.1257, "step": 2023 }, { "epoch": 2.7433987813134735, "grad_norm": 1.2209243529896048, "learning_rate": 2.3033998275040047e-06, "loss": 1.9774, "step": 2024 }, { "epoch": 2.7447528774542995, "grad_norm": 1.16381904107931, "learning_rate": 2.298858921794354e-06, "loss": 2.4332, "step": 2025 }, { "epoch": 2.746106973595125, "grad_norm": 1.1219774061092502, "learning_rate": 2.2943211599521713e-06, "loss": 2.0022, "step": 2026 }, { "epoch": 2.747461069735951, "grad_norm": 1.0877209755780604, "learning_rate": 2.2897865472589707e-06, "loss": 2.2494, "step": 2027 }, { "epoch": 2.748815165876777, "grad_norm": 1.1641643532315749, "learning_rate": 2.2852550889926067e-06, "loss": 2.3295, "step": 2028 }, { "epoch": 2.7501692620176033, "grad_norm": 1.0971463061405777, "learning_rate": 2.280726790427258e-06, "loss": 2.0522, "step": 2029 }, { "epoch": 2.7515233581584293, "grad_norm": 1.0954456035691678, "learning_rate": 2.276201656833424e-06, "loss": 1.8994, "step": 2030 }, { "epoch": 2.7528774542992553, "grad_norm": 1.2776543904654059, "learning_rate": 2.2716796934779283e-06, "loss": 2.1554, "step": 2031 }, { "epoch": 2.7542315504400814, "grad_norm": 1.0394272753439568, "learning_rate": 2.2671609056238953e-06, "loss": 1.952, "step": 2032 }, { "epoch": 2.755585646580907, "grad_norm": 1.1253810775034214, "learning_rate": 2.262645298530762e-06, "loss": 2.3973, "step": 2033 }, { "epoch": 2.756939742721733, "grad_norm": 1.3345296559246174, "learning_rate": 2.2581328774542567e-06, "loss": 2.6639, "step": 2034 }, { "epoch": 2.758293838862559, "grad_norm": 0.9796809558518422, "learning_rate": 2.2536236476464007e-06, "loss": 1.9827, "step": 2035 }, { "epoch": 2.759647935003385, "grad_norm": 1.1655529301741654, "learning_rate": 2.249117614355506e-06, "loss": 2.3146, "step": 2036 }, { "epoch": 2.761002031144211, "grad_norm": 1.2099057667985653, "learning_rate": 2.2446147828261562e-06, "loss": 2.3993, "step": 2037 }, { "epoch": 2.762356127285037, "grad_norm": 1.3541399204887774, "learning_rate": 2.2401151582992157e-06, "loss": 2.2597, "step": 2038 }, { "epoch": 2.7637102234258633, "grad_norm": 0.9707466204095863, "learning_rate": 2.235618746011813e-06, "loss": 1.8865, "step": 2039 }, { "epoch": 2.7650643195666893, "grad_norm": 1.2003161344667284, "learning_rate": 2.2311255511973347e-06, "loss": 2.2053, "step": 2040 }, { "epoch": 2.7664184157075153, "grad_norm": 1.2427080146924223, "learning_rate": 2.226635579085429e-06, "loss": 2.5333, "step": 2041 }, { "epoch": 2.7677725118483414, "grad_norm": 1.2482486915110669, "learning_rate": 2.2221488349019903e-06, "loss": 2.081, "step": 2042 }, { "epoch": 2.7691266079891674, "grad_norm": 1.2721482948050402, "learning_rate": 2.217665323869152e-06, "loss": 2.5127, "step": 2043 }, { "epoch": 2.7704807041299935, "grad_norm": 1.1928191386170535, "learning_rate": 2.2131850512052928e-06, "loss": 2.3704, "step": 2044 }, { "epoch": 2.771834800270819, "grad_norm": 1.0851493590766315, "learning_rate": 2.208708022125013e-06, "loss": 2.3767, "step": 2045 }, { "epoch": 2.773188896411645, "grad_norm": 1.2230116523152432, "learning_rate": 2.2042342418391467e-06, "loss": 2.0559, "step": 2046 }, { "epoch": 2.774542992552471, "grad_norm": 1.2982061930202065, "learning_rate": 2.1997637155547404e-06, "loss": 1.9753, "step": 2047 }, { "epoch": 2.775897088693297, "grad_norm": 1.4225699287390874, "learning_rate": 2.1952964484750527e-06, "loss": 2.3966, "step": 2048 }, { "epoch": 2.7772511848341233, "grad_norm": 1.2699513520653058, "learning_rate": 2.1908324457995556e-06, "loss": 2.1857, "step": 2049 }, { "epoch": 2.7786052809749493, "grad_norm": 1.1131391184478643, "learning_rate": 2.186371712723913e-06, "loss": 2.476, "step": 2050 }, { "epoch": 2.7799593771157753, "grad_norm": 1.1015068907662482, "learning_rate": 2.1819142544399917e-06, "loss": 2.4802, "step": 2051 }, { "epoch": 2.781313473256601, "grad_norm": 1.3818031208847354, "learning_rate": 2.177460076135841e-06, "loss": 2.2723, "step": 2052 }, { "epoch": 2.782667569397427, "grad_norm": 1.1395789757657575, "learning_rate": 2.173009182995693e-06, "loss": 2.0286, "step": 2053 }, { "epoch": 2.784021665538253, "grad_norm": 1.1604187136522945, "learning_rate": 2.1685615801999606e-06, "loss": 2.2005, "step": 2054 }, { "epoch": 2.785375761679079, "grad_norm": 1.229026030647276, "learning_rate": 2.1641172729252206e-06, "loss": 2.1318, "step": 2055 }, { "epoch": 2.786729857819905, "grad_norm": 1.0060463367078358, "learning_rate": 2.159676266344222e-06, "loss": 2.5203, "step": 2056 }, { "epoch": 2.788083953960731, "grad_norm": 1.167507696393823, "learning_rate": 2.155238565625865e-06, "loss": 1.7903, "step": 2057 }, { "epoch": 2.789438050101557, "grad_norm": 1.3078513803870446, "learning_rate": 2.1508041759352034e-06, "loss": 2.4311, "step": 2058 }, { "epoch": 2.7907921462423833, "grad_norm": 1.2568424750257208, "learning_rate": 2.1463731024334424e-06, "loss": 2.2506, "step": 2059 }, { "epoch": 2.7921462423832093, "grad_norm": 1.2916256075350547, "learning_rate": 2.14194535027792e-06, "loss": 2.2703, "step": 2060 }, { "epoch": 2.7935003385240353, "grad_norm": 1.195880677547239, "learning_rate": 2.1375209246221164e-06, "loss": 1.7292, "step": 2061 }, { "epoch": 2.7948544346648614, "grad_norm": 1.2343623869418596, "learning_rate": 2.1330998306156337e-06, "loss": 2.0974, "step": 2062 }, { "epoch": 2.7962085308056874, "grad_norm": 1.5038792989657825, "learning_rate": 2.128682073404197e-06, "loss": 2.0363, "step": 2063 }, { "epoch": 2.7975626269465135, "grad_norm": 1.2628180745624695, "learning_rate": 2.1242676581296527e-06, "loss": 2.2605, "step": 2064 }, { "epoch": 2.798916723087339, "grad_norm": 1.251372861723447, "learning_rate": 2.1198565899299505e-06, "loss": 2.5106, "step": 2065 }, { "epoch": 2.800270819228165, "grad_norm": 1.181610869162084, "learning_rate": 2.1154488739391514e-06, "loss": 2.2867, "step": 2066 }, { "epoch": 2.801624915368991, "grad_norm": 1.1507959248642738, "learning_rate": 2.1110445152874097e-06, "loss": 2.319, "step": 2067 }, { "epoch": 2.802979011509817, "grad_norm": 1.1688254643806266, "learning_rate": 2.1066435191009717e-06, "loss": 2.4526, "step": 2068 }, { "epoch": 2.8043331076506433, "grad_norm": 1.4143274765288207, "learning_rate": 2.1022458905021764e-06, "loss": 2.2593, "step": 2069 }, { "epoch": 2.8056872037914693, "grad_norm": 1.0941888822986356, "learning_rate": 2.0978516346094342e-06, "loss": 2.5379, "step": 2070 }, { "epoch": 2.8070412999322953, "grad_norm": 1.0961460587511338, "learning_rate": 2.0934607565372394e-06, "loss": 2.0686, "step": 2071 }, { "epoch": 2.808395396073121, "grad_norm": 1.1313149099706428, "learning_rate": 2.089073261396148e-06, "loss": 1.9343, "step": 2072 }, { "epoch": 2.809749492213947, "grad_norm": 1.2488003968430963, "learning_rate": 2.0846891542927794e-06, "loss": 2.5037, "step": 2073 }, { "epoch": 2.811103588354773, "grad_norm": 1.3526779217410578, "learning_rate": 2.0803084403298142e-06, "loss": 2.3099, "step": 2074 }, { "epoch": 2.812457684495599, "grad_norm": 1.181990597260296, "learning_rate": 2.0759311246059764e-06, "loss": 1.9261, "step": 2075 }, { "epoch": 2.813811780636425, "grad_norm": 1.1909588040536576, "learning_rate": 2.071557212216044e-06, "loss": 2.0809, "step": 2076 }, { "epoch": 2.815165876777251, "grad_norm": 1.19354549824163, "learning_rate": 2.067186708250826e-06, "loss": 2.3626, "step": 2077 }, { "epoch": 2.816519972918077, "grad_norm": 1.5303114069512946, "learning_rate": 2.0628196177971654e-06, "loss": 2.3279, "step": 2078 }, { "epoch": 2.8178740690589033, "grad_norm": 1.3153194325064834, "learning_rate": 2.058455945937938e-06, "loss": 2.5325, "step": 2079 }, { "epoch": 2.8192281651997293, "grad_norm": 1.3578691903015818, "learning_rate": 2.054095697752032e-06, "loss": 2.0425, "step": 2080 }, { "epoch": 2.8205822613405553, "grad_norm": 1.109588432142048, "learning_rate": 2.04973887831436e-06, "loss": 2.2762, "step": 2081 }, { "epoch": 2.8219363574813814, "grad_norm": 1.0009686102290198, "learning_rate": 2.0453854926958374e-06, "loss": 2.666, "step": 2082 }, { "epoch": 2.8232904536222074, "grad_norm": 1.8548674081037086, "learning_rate": 2.0410355459633823e-06, "loss": 2.2673, "step": 2083 }, { "epoch": 2.824644549763033, "grad_norm": 1.113124229891495, "learning_rate": 2.036689043179917e-06, "loss": 2.2299, "step": 2084 }, { "epoch": 2.825998645903859, "grad_norm": 1.1430362720866236, "learning_rate": 2.0323459894043497e-06, "loss": 2.2362, "step": 2085 }, { "epoch": 2.827352742044685, "grad_norm": 1.3231970310329217, "learning_rate": 2.0280063896915724e-06, "loss": 2.1857, "step": 2086 }, { "epoch": 2.828706838185511, "grad_norm": 1.0366312826936661, "learning_rate": 2.0236702490924653e-06, "loss": 1.8564, "step": 2087 }, { "epoch": 2.830060934326337, "grad_norm": 1.1592760192547105, "learning_rate": 2.0193375726538737e-06, "loss": 2.7656, "step": 2088 }, { "epoch": 2.8314150304671633, "grad_norm": 1.5009484111746818, "learning_rate": 2.0150083654186163e-06, "loss": 1.8773, "step": 2089 }, { "epoch": 2.8327691266079893, "grad_norm": 1.1661610309674306, "learning_rate": 2.010682632425474e-06, "loss": 2.1206, "step": 2090 }, { "epoch": 2.834123222748815, "grad_norm": 1.0455004004854604, "learning_rate": 2.0063603787091788e-06, "loss": 2.099, "step": 2091 }, { "epoch": 2.835477318889641, "grad_norm": 1.2573170420483708, "learning_rate": 2.0020416093004215e-06, "loss": 2.0071, "step": 2092 }, { "epoch": 2.836831415030467, "grad_norm": 1.3472390637335112, "learning_rate": 1.997726329225829e-06, "loss": 2.3401, "step": 2093 }, { "epoch": 2.838185511171293, "grad_norm": 1.3454963273508525, "learning_rate": 1.9934145435079705e-06, "loss": 2.125, "step": 2094 }, { "epoch": 2.839539607312119, "grad_norm": 1.3088468975041136, "learning_rate": 1.9891062571653513e-06, "loss": 2.3926, "step": 2095 }, { "epoch": 2.840893703452945, "grad_norm": 1.2362661310514473, "learning_rate": 1.9848014752123977e-06, "loss": 1.8366, "step": 2096 }, { "epoch": 2.842247799593771, "grad_norm": 1.231207007524463, "learning_rate": 1.980500202659464e-06, "loss": 2.1338, "step": 2097 }, { "epoch": 2.843601895734597, "grad_norm": 1.1956079183379362, "learning_rate": 1.976202444512813e-06, "loss": 2.2051, "step": 2098 }, { "epoch": 2.8449559918754233, "grad_norm": 1.396527464715163, "learning_rate": 1.9719082057746202e-06, "loss": 2.0318, "step": 2099 }, { "epoch": 2.8463100880162493, "grad_norm": 1.124212719908014, "learning_rate": 1.9676174914429678e-06, "loss": 2.4668, "step": 2100 }, { "epoch": 2.8476641841570753, "grad_norm": 1.3327609510554048, "learning_rate": 1.96333030651183e-06, "loss": 2.416, "step": 2101 }, { "epoch": 2.8490182802979014, "grad_norm": 1.3533962283742291, "learning_rate": 1.959046655971079e-06, "loss": 2.133, "step": 2102 }, { "epoch": 2.850372376438727, "grad_norm": 1.1254876249976722, "learning_rate": 1.9547665448064695e-06, "loss": 2.0591, "step": 2103 }, { "epoch": 2.851726472579553, "grad_norm": 1.3458204752024827, "learning_rate": 1.9504899779996354e-06, "loss": 1.9248, "step": 2104 }, { "epoch": 2.853080568720379, "grad_norm": 1.2574950070426385, "learning_rate": 1.946216960528092e-06, "loss": 1.9106, "step": 2105 }, { "epoch": 2.854434664861205, "grad_norm": 1.0710252913879086, "learning_rate": 1.941947497365214e-06, "loss": 2.013, "step": 2106 }, { "epoch": 2.855788761002031, "grad_norm": 1.25582685422548, "learning_rate": 1.9376815934802496e-06, "loss": 2.0768, "step": 2107 }, { "epoch": 2.857142857142857, "grad_norm": 1.0267706194019677, "learning_rate": 1.933419253838296e-06, "loss": 2.01, "step": 2108 }, { "epoch": 2.8584969532836833, "grad_norm": 1.2904448969859916, "learning_rate": 1.929160483400305e-06, "loss": 1.9768, "step": 2109 }, { "epoch": 2.8598510494245093, "grad_norm": 1.3373422907621348, "learning_rate": 1.9249052871230762e-06, "loss": 2.0565, "step": 2110 }, { "epoch": 2.861205145565335, "grad_norm": 1.1020250575352282, "learning_rate": 1.920653669959244e-06, "loss": 2.0645, "step": 2111 }, { "epoch": 2.862559241706161, "grad_norm": 1.0440980138533542, "learning_rate": 1.9164056368572847e-06, "loss": 2.1169, "step": 2112 }, { "epoch": 2.863913337846987, "grad_norm": 1.0570192976923376, "learning_rate": 1.9121611927614965e-06, "loss": 2.1198, "step": 2113 }, { "epoch": 2.865267433987813, "grad_norm": 1.2811020058606082, "learning_rate": 1.9079203426120007e-06, "loss": 1.9467, "step": 2114 }, { "epoch": 2.866621530128639, "grad_norm": 1.215277745458542, "learning_rate": 1.9036830913447418e-06, "loss": 2.2175, "step": 2115 }, { "epoch": 2.867975626269465, "grad_norm": 1.1326070134691888, "learning_rate": 1.8994494438914669e-06, "loss": 2.4989, "step": 2116 }, { "epoch": 2.869329722410291, "grad_norm": 1.557900149617448, "learning_rate": 1.8952194051797362e-06, "loss": 2.3981, "step": 2117 }, { "epoch": 2.870683818551117, "grad_norm": 1.4088259039808713, "learning_rate": 1.8909929801329057e-06, "loss": 2.2609, "step": 2118 }, { "epoch": 2.8720379146919433, "grad_norm": 1.1122390052020512, "learning_rate": 1.8867701736701238e-06, "loss": 2.0194, "step": 2119 }, { "epoch": 2.8733920108327693, "grad_norm": 1.6626997638245213, "learning_rate": 1.8825509907063328e-06, "loss": 2.5041, "step": 2120 }, { "epoch": 2.8747461069735953, "grad_norm": 1.2021954308723752, "learning_rate": 1.8783354361522515e-06, "loss": 1.9558, "step": 2121 }, { "epoch": 2.8761002031144214, "grad_norm": 1.189053074107543, "learning_rate": 1.8741235149143815e-06, "loss": 2.1995, "step": 2122 }, { "epoch": 2.877454299255247, "grad_norm": 1.1609608293450915, "learning_rate": 1.8699152318949915e-06, "loss": 2.2626, "step": 2123 }, { "epoch": 2.878808395396073, "grad_norm": 1.177143045452091, "learning_rate": 1.8657105919921142e-06, "loss": 2.1996, "step": 2124 }, { "epoch": 2.880162491536899, "grad_norm": 1.5787704494541075, "learning_rate": 1.8615096000995482e-06, "loss": 2.1233, "step": 2125 }, { "epoch": 2.881516587677725, "grad_norm": 0.9998198530880283, "learning_rate": 1.8573122611068406e-06, "loss": 2.1223, "step": 2126 }, { "epoch": 2.882870683818551, "grad_norm": 1.3298134177797156, "learning_rate": 1.853118579899288e-06, "loss": 2.115, "step": 2127 }, { "epoch": 2.884224779959377, "grad_norm": 1.042057656736215, "learning_rate": 1.8489285613579328e-06, "loss": 2.2095, "step": 2128 }, { "epoch": 2.8855788761002032, "grad_norm": 1.0171565014816397, "learning_rate": 1.844742210359549e-06, "loss": 1.6516, "step": 2129 }, { "epoch": 2.886932972241029, "grad_norm": 1.2889249025896115, "learning_rate": 1.8405595317766494e-06, "loss": 2.1873, "step": 2130 }, { "epoch": 2.888287068381855, "grad_norm": 1.1435912604739802, "learning_rate": 1.836380530477465e-06, "loss": 2.324, "step": 2131 }, { "epoch": 2.889641164522681, "grad_norm": 1.0689294471126107, "learning_rate": 1.8322052113259492e-06, "loss": 2.2091, "step": 2132 }, { "epoch": 2.890995260663507, "grad_norm": 1.0567713258079983, "learning_rate": 1.8280335791817733e-06, "loss": 1.8378, "step": 2133 }, { "epoch": 2.892349356804333, "grad_norm": 1.0627783400929425, "learning_rate": 1.8238656389003112e-06, "loss": 2.0609, "step": 2134 }, { "epoch": 2.893703452945159, "grad_norm": 1.1713820085504252, "learning_rate": 1.8197013953326464e-06, "loss": 2.0228, "step": 2135 }, { "epoch": 2.895057549085985, "grad_norm": 1.358198147161062, "learning_rate": 1.8155408533255553e-06, "loss": 2.3418, "step": 2136 }, { "epoch": 2.896411645226811, "grad_norm": 1.1910114735895758, "learning_rate": 1.8113840177215042e-06, "loss": 1.9185, "step": 2137 }, { "epoch": 2.897765741367637, "grad_norm": 1.0438200130516968, "learning_rate": 1.807230893358653e-06, "loss": 2.331, "step": 2138 }, { "epoch": 2.8991198375084632, "grad_norm": 1.120315824767046, "learning_rate": 1.803081485070834e-06, "loss": 2.1703, "step": 2139 }, { "epoch": 2.9004739336492893, "grad_norm": 1.2622978540853298, "learning_rate": 1.7989357976875603e-06, "loss": 2.0372, "step": 2140 }, { "epoch": 2.9018280297901153, "grad_norm": 1.1471832482429662, "learning_rate": 1.7947938360340118e-06, "loss": 1.9602, "step": 2141 }, { "epoch": 2.903182125930941, "grad_norm": 1.2925593645649263, "learning_rate": 1.7906556049310287e-06, "loss": 1.7702, "step": 2142 }, { "epoch": 2.904536222071767, "grad_norm": 1.1733600543199965, "learning_rate": 1.7865211091951168e-06, "loss": 1.702, "step": 2143 }, { "epoch": 2.905890318212593, "grad_norm": 1.0495647809558728, "learning_rate": 1.7823903536384262e-06, "loss": 1.8785, "step": 2144 }, { "epoch": 2.907244414353419, "grad_norm": 1.3724266737000936, "learning_rate": 1.7782633430687618e-06, "loss": 2.2293, "step": 2145 }, { "epoch": 2.908598510494245, "grad_norm": 1.094169187997442, "learning_rate": 1.7741400822895633e-06, "loss": 2.1643, "step": 2146 }, { "epoch": 2.909952606635071, "grad_norm": 1.2660626465199718, "learning_rate": 1.7700205760999061e-06, "loss": 2.3012, "step": 2147 }, { "epoch": 2.911306702775897, "grad_norm": 1.2263371195739785, "learning_rate": 1.7659048292945019e-06, "loss": 1.8481, "step": 2148 }, { "epoch": 2.9126607989167232, "grad_norm": 1.2062354300027056, "learning_rate": 1.7617928466636786e-06, "loss": 2.1278, "step": 2149 }, { "epoch": 2.914014895057549, "grad_norm": 1.396023621940613, "learning_rate": 1.7576846329933906e-06, "loss": 2.3786, "step": 2150 }, { "epoch": 2.915368991198375, "grad_norm": 1.1972452069648623, "learning_rate": 1.7535801930652009e-06, "loss": 1.6638, "step": 2151 }, { "epoch": 2.916723087339201, "grad_norm": 1.021542263655166, "learning_rate": 1.7494795316562791e-06, "loss": 2.0394, "step": 2152 }, { "epoch": 2.918077183480027, "grad_norm": 1.2033557332753488, "learning_rate": 1.745382653539403e-06, "loss": 2.4409, "step": 2153 }, { "epoch": 2.919431279620853, "grad_norm": 1.0785062785447148, "learning_rate": 1.7412895634829391e-06, "loss": 2.0973, "step": 2154 }, { "epoch": 2.920785375761679, "grad_norm": 1.2270595757010991, "learning_rate": 1.7372002662508537e-06, "loss": 1.9301, "step": 2155 }, { "epoch": 2.922139471902505, "grad_norm": 1.2298792448020188, "learning_rate": 1.733114766602691e-06, "loss": 2.2803, "step": 2156 }, { "epoch": 2.923493568043331, "grad_norm": 1.3120973971257666, "learning_rate": 1.7290330692935775e-06, "loss": 2.2858, "step": 2157 }, { "epoch": 2.924847664184157, "grad_norm": 1.2885618789074844, "learning_rate": 1.724955179074218e-06, "loss": 2.0856, "step": 2158 }, { "epoch": 2.9262017603249832, "grad_norm": 1.08833991594575, "learning_rate": 1.7208811006908798e-06, "loss": 1.9095, "step": 2159 }, { "epoch": 2.9275558564658093, "grad_norm": 1.3512141243511449, "learning_rate": 1.7168108388853999e-06, "loss": 2.3918, "step": 2160 }, { "epoch": 2.9289099526066353, "grad_norm": 1.2335733112034404, "learning_rate": 1.7127443983951687e-06, "loss": 1.8729, "step": 2161 }, { "epoch": 2.930264048747461, "grad_norm": 1.3057931319108373, "learning_rate": 1.7086817839531278e-06, "loss": 2.4506, "step": 2162 }, { "epoch": 2.931618144888287, "grad_norm": 1.0631756137857529, "learning_rate": 1.7046230002877735e-06, "loss": 2.1407, "step": 2163 }, { "epoch": 2.932972241029113, "grad_norm": 1.3461802245139118, "learning_rate": 1.7005680521231328e-06, "loss": 2.1968, "step": 2164 }, { "epoch": 2.934326337169939, "grad_norm": 1.2583827741690212, "learning_rate": 1.6965169441787782e-06, "loss": 1.951, "step": 2165 }, { "epoch": 2.935680433310765, "grad_norm": 1.4314758537559276, "learning_rate": 1.6924696811698072e-06, "loss": 2.0765, "step": 2166 }, { "epoch": 2.937034529451591, "grad_norm": 1.1326804889902262, "learning_rate": 1.688426267806841e-06, "loss": 2.1307, "step": 2167 }, { "epoch": 2.938388625592417, "grad_norm": 1.0483784713217499, "learning_rate": 1.6843867087960252e-06, "loss": 1.9878, "step": 2168 }, { "epoch": 2.939742721733243, "grad_norm": 1.2184954478957704, "learning_rate": 1.6803510088390158e-06, "loss": 1.9964, "step": 2169 }, { "epoch": 2.941096817874069, "grad_norm": 1.2861862698127093, "learning_rate": 1.6763191726329753e-06, "loss": 2.184, "step": 2170 }, { "epoch": 2.942450914014895, "grad_norm": 1.1404659174923684, "learning_rate": 1.6722912048705764e-06, "loss": 2.4739, "step": 2171 }, { "epoch": 2.943805010155721, "grad_norm": 1.2147396795004435, "learning_rate": 1.6682671102399806e-06, "loss": 2.1952, "step": 2172 }, { "epoch": 2.945159106296547, "grad_norm": 1.1895679535389942, "learning_rate": 1.6642468934248495e-06, "loss": 1.8479, "step": 2173 }, { "epoch": 2.946513202437373, "grad_norm": 1.2467456458845403, "learning_rate": 1.660230559104325e-06, "loss": 2.1058, "step": 2174 }, { "epoch": 2.947867298578199, "grad_norm": 1.6924620574535818, "learning_rate": 1.6562181119530314e-06, "loss": 2.3561, "step": 2175 }, { "epoch": 2.949221394719025, "grad_norm": 1.0845170117138176, "learning_rate": 1.6522095566410728e-06, "loss": 2.6286, "step": 2176 }, { "epoch": 2.950575490859851, "grad_norm": 1.288053541033133, "learning_rate": 1.6482048978340177e-06, "loss": 2.0054, "step": 2177 }, { "epoch": 2.951929587000677, "grad_norm": 1.2866263858556009, "learning_rate": 1.6442041401929055e-06, "loss": 1.8053, "step": 2178 }, { "epoch": 2.9532836831415032, "grad_norm": 1.2770087302330293, "learning_rate": 1.6402072883742282e-06, "loss": 2.3504, "step": 2179 }, { "epoch": 2.9546377792823293, "grad_norm": 1.209845038682216, "learning_rate": 1.636214347029938e-06, "loss": 2.2815, "step": 2180 }, { "epoch": 2.955991875423155, "grad_norm": 1.0934831741155357, "learning_rate": 1.632225320807434e-06, "loss": 2.076, "step": 2181 }, { "epoch": 2.957345971563981, "grad_norm": 1.152929805068558, "learning_rate": 1.6282402143495568e-06, "loss": 1.9276, "step": 2182 }, { "epoch": 2.958700067704807, "grad_norm": 1.1936452561523994, "learning_rate": 1.6242590322945833e-06, "loss": 2.0273, "step": 2183 }, { "epoch": 2.960054163845633, "grad_norm": 1.288878935798258, "learning_rate": 1.6202817792762283e-06, "loss": 2.5339, "step": 2184 }, { "epoch": 2.961408259986459, "grad_norm": 1.3502029610964787, "learning_rate": 1.6163084599236278e-06, "loss": 1.9991, "step": 2185 }, { "epoch": 2.962762356127285, "grad_norm": 1.3048211810161428, "learning_rate": 1.6123390788613447e-06, "loss": 2.3083, "step": 2186 }, { "epoch": 2.964116452268111, "grad_norm": 1.3110572114980938, "learning_rate": 1.6083736407093541e-06, "loss": 1.7873, "step": 2187 }, { "epoch": 2.9654705484089368, "grad_norm": 1.305186075725804, "learning_rate": 1.6044121500830413e-06, "loss": 2.5161, "step": 2188 }, { "epoch": 2.966824644549763, "grad_norm": 1.1931640892840685, "learning_rate": 1.6004546115932023e-06, "loss": 1.8781, "step": 2189 }, { "epoch": 2.968178740690589, "grad_norm": 1.1329814905618647, "learning_rate": 1.596501029846027e-06, "loss": 1.8183, "step": 2190 }, { "epoch": 2.969532836831415, "grad_norm": 1.256058797607824, "learning_rate": 1.5925514094431054e-06, "loss": 1.9353, "step": 2191 }, { "epoch": 2.970886932972241, "grad_norm": 1.2078766782518078, "learning_rate": 1.5886057549814133e-06, "loss": 2.2515, "step": 2192 }, { "epoch": 2.972241029113067, "grad_norm": 1.2594876666386745, "learning_rate": 1.5846640710533096e-06, "loss": 1.9451, "step": 2193 }, { "epoch": 2.973595125253893, "grad_norm": 2.0283531784822317, "learning_rate": 1.5807263622465375e-06, "loss": 2.1079, "step": 2194 }, { "epoch": 2.974949221394719, "grad_norm": 1.3933091735063965, "learning_rate": 1.5767926331442069e-06, "loss": 2.3794, "step": 2195 }, { "epoch": 2.976303317535545, "grad_norm": 1.8309556056295924, "learning_rate": 1.572862888324801e-06, "loss": 2.5542, "step": 2196 }, { "epoch": 2.977657413676371, "grad_norm": 1.2230217418918525, "learning_rate": 1.5689371323621626e-06, "loss": 2.2557, "step": 2197 }, { "epoch": 2.979011509817197, "grad_norm": 1.1682613662079826, "learning_rate": 1.5650153698254916e-06, "loss": 1.9746, "step": 2198 }, { "epoch": 2.9803656059580232, "grad_norm": 1.9080960547810655, "learning_rate": 1.561097605279343e-06, "loss": 2.034, "step": 2199 }, { "epoch": 2.9817197020988493, "grad_norm": 1.1116323713868257, "learning_rate": 1.557183843283614e-06, "loss": 2.2085, "step": 2200 }, { "epoch": 2.983073798239675, "grad_norm": 1.2122411032785838, "learning_rate": 1.5532740883935487e-06, "loss": 1.8254, "step": 2201 }, { "epoch": 2.984427894380501, "grad_norm": 1.1862771421276836, "learning_rate": 1.549368345159723e-06, "loss": 1.8274, "step": 2202 }, { "epoch": 2.985781990521327, "grad_norm": 1.275270557678822, "learning_rate": 1.5454666181280437e-06, "loss": 1.9233, "step": 2203 }, { "epoch": 2.987136086662153, "grad_norm": 1.0805942989876476, "learning_rate": 1.5415689118397465e-06, "loss": 1.9249, "step": 2204 }, { "epoch": 2.988490182802979, "grad_norm": 1.0896311360207829, "learning_rate": 1.5376752308313847e-06, "loss": 2.2418, "step": 2205 }, { "epoch": 2.989844278943805, "grad_norm": 1.3875258063079425, "learning_rate": 1.5337855796348255e-06, "loss": 2.1242, "step": 2206 }, { "epoch": 2.991198375084631, "grad_norm": 1.2203604406862303, "learning_rate": 1.5298999627772497e-06, "loss": 2.2783, "step": 2207 }, { "epoch": 2.9925524712254568, "grad_norm": 1.0980233593746727, "learning_rate": 1.5260183847811383e-06, "loss": 1.9593, "step": 2208 }, { "epoch": 2.993906567366283, "grad_norm": 1.2415947966462062, "learning_rate": 1.5221408501642759e-06, "loss": 2.341, "step": 2209 }, { "epoch": 2.995260663507109, "grad_norm": 1.188061537994219, "learning_rate": 1.5182673634397365e-06, "loss": 2.2105, "step": 2210 }, { "epoch": 2.996614759647935, "grad_norm": 1.0814079186989471, "learning_rate": 1.514397929115884e-06, "loss": 2.4295, "step": 2211 }, { "epoch": 2.997968855788761, "grad_norm": 1.0354623291225793, "learning_rate": 1.5105325516963693e-06, "loss": 1.8617, "step": 2212 }, { "epoch": 2.999322951929587, "grad_norm": 1.0023079653760507, "learning_rate": 1.506671235680115e-06, "loss": 2.0759, "step": 2213 }, { "epoch": 3.000677048070413, "grad_norm": 1.262291881250163, "learning_rate": 1.5028139855613233e-06, "loss": 2.3376, "step": 2214 } ], "logging_steps": 1, "max_steps": 2952, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 738, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 5.849584295215104e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }