{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 1022, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0019569471624266144, "grad_norm": 14.974896244654243, "learning_rate": 9.99997637686018e-06, "loss": 0.5134, "step": 1 }, { "epoch": 0.003913894324853229, "grad_norm": 11.268155081482439, "learning_rate": 9.999905507663936e-06, "loss": 0.4507, "step": 2 }, { "epoch": 0.005870841487279843, "grad_norm": 5.675201946627105, "learning_rate": 9.999787393080931e-06, "loss": 0.3333, "step": 3 }, { "epoch": 0.007827788649706457, "grad_norm": 4.76369447674635, "learning_rate": 9.99962203422726e-06, "loss": 0.2501, "step": 4 }, { "epoch": 0.009784735812133072, "grad_norm": 3.86430652077425, "learning_rate": 9.999409432665442e-06, "loss": 0.2352, "step": 5 }, { "epoch": 0.011741682974559686, "grad_norm": 2.9391281463544927, "learning_rate": 9.9991495904044e-06, "loss": 0.1801, "step": 6 }, { "epoch": 0.0136986301369863, "grad_norm": 4.935305290808968, "learning_rate": 9.998842509899456e-06, "loss": 0.2022, "step": 7 }, { "epoch": 0.015655577299412915, "grad_norm": 3.394287544469333, "learning_rate": 9.998488194052287e-06, "loss": 0.1986, "step": 8 }, { "epoch": 0.01761252446183953, "grad_norm": 3.191867498997973, "learning_rate": 9.998086646210916e-06, "loss": 0.2218, "step": 9 }, { "epoch": 0.019569471624266144, "grad_norm": 3.736922584025223, "learning_rate": 9.997637870169673e-06, "loss": 0.2301, "step": 10 }, { "epoch": 0.021526418786692758, "grad_norm": 2.574511962253989, "learning_rate": 9.997141870169154e-06, "loss": 0.1974, "step": 11 }, { "epoch": 0.023483365949119372, "grad_norm": 1.9771550245902525, "learning_rate": 9.996598650896191e-06, "loss": 0.1684, "step": 12 }, { "epoch": 0.025440313111545987, "grad_norm": 2.7153119809106325, "learning_rate": 9.996008217483806e-06, "loss": 0.2245, "step": 13 }, { "epoch": 0.0273972602739726, "grad_norm": 2.8742546745263717, "learning_rate": 9.995370575511151e-06, "loss": 0.2288, "step": 14 }, { "epoch": 0.029354207436399216, "grad_norm": 2.169907498218494, "learning_rate": 9.994685731003469e-06, "loss": 0.168, "step": 15 }, { "epoch": 0.03131115459882583, "grad_norm": 2.694869369895634, "learning_rate": 9.993953690432032e-06, "loss": 0.1979, "step": 16 }, { "epoch": 0.033268101761252444, "grad_norm": 2.3481784811262036, "learning_rate": 9.99317446071408e-06, "loss": 0.1669, "step": 17 }, { "epoch": 0.03522504892367906, "grad_norm": 2.2778587852537915, "learning_rate": 9.99234804921275e-06, "loss": 0.1886, "step": 18 }, { "epoch": 0.03718199608610567, "grad_norm": 3.8605535253750856, "learning_rate": 9.991474463737018e-06, "loss": 0.2186, "step": 19 }, { "epoch": 0.03913894324853229, "grad_norm": 3.6712798376081253, "learning_rate": 9.990553712541617e-06, "loss": 0.2869, "step": 20 }, { "epoch": 0.0410958904109589, "grad_norm": 3.020136583581482, "learning_rate": 9.989585804326963e-06, "loss": 0.2047, "step": 21 }, { "epoch": 0.043052837573385516, "grad_norm": 2.980755180928037, "learning_rate": 9.988570748239062e-06, "loss": 0.2078, "step": 22 }, { "epoch": 0.04500978473581213, "grad_norm": 2.577995405397823, "learning_rate": 9.987508553869444e-06, "loss": 0.2251, "step": 23 }, { "epoch": 0.046966731898238745, "grad_norm": 2.3582990191328683, "learning_rate": 9.986399231255057e-06, "loss": 0.1693, "step": 24 }, { "epoch": 0.04892367906066536, "grad_norm": 2.5170243359845474, "learning_rate": 9.985242790878168e-06, "loss": 0.2251, "step": 25 }, { "epoch": 0.050880626223091974, "grad_norm": 2.4413347887710692, "learning_rate": 9.984039243666284e-06, "loss": 0.1831, "step": 26 }, { "epoch": 0.05283757338551859, "grad_norm": 2.428602560977613, "learning_rate": 9.982788600992027e-06, "loss": 0.2028, "step": 27 }, { "epoch": 0.0547945205479452, "grad_norm": 2.354947407625226, "learning_rate": 9.98149087467304e-06, "loss": 0.182, "step": 28 }, { "epoch": 0.05675146771037182, "grad_norm": 2.3234698701184047, "learning_rate": 9.980146076971873e-06, "loss": 0.2235, "step": 29 }, { "epoch": 0.05870841487279843, "grad_norm": 3.972250754175844, "learning_rate": 9.978754220595861e-06, "loss": 0.2506, "step": 30 }, { "epoch": 0.060665362035225046, "grad_norm": 2.7300875684817942, "learning_rate": 9.977315318697013e-06, "loss": 0.2115, "step": 31 }, { "epoch": 0.06262230919765166, "grad_norm": 2.0538364281382964, "learning_rate": 9.975829384871884e-06, "loss": 0.1811, "step": 32 }, { "epoch": 0.06457925636007827, "grad_norm": 2.1943202715227343, "learning_rate": 9.974296433161437e-06, "loss": 0.1954, "step": 33 }, { "epoch": 0.06653620352250489, "grad_norm": 2.8202422319956577, "learning_rate": 9.97271647805093e-06, "loss": 0.2304, "step": 34 }, { "epoch": 0.0684931506849315, "grad_norm": 2.386195936248472, "learning_rate": 9.97108953446976e-06, "loss": 0.1763, "step": 35 }, { "epoch": 0.07045009784735812, "grad_norm": 2.2756713060895346, "learning_rate": 9.969415617791336e-06, "loss": 0.2069, "step": 36 }, { "epoch": 0.07240704500978473, "grad_norm": 2.1497696128664443, "learning_rate": 9.967694743832923e-06, "loss": 0.1558, "step": 37 }, { "epoch": 0.07436399217221135, "grad_norm": 1.6713386371324828, "learning_rate": 9.965926928855498e-06, "loss": 0.136, "step": 38 }, { "epoch": 0.07632093933463796, "grad_norm": 2.327515952013986, "learning_rate": 9.964112189563601e-06, "loss": 0.1882, "step": 39 }, { "epoch": 0.07827788649706457, "grad_norm": 2.0334916496810114, "learning_rate": 9.962250543105167e-06, "loss": 0.1825, "step": 40 }, { "epoch": 0.08023483365949119, "grad_norm": 2.113562795396453, "learning_rate": 9.960342007071368e-06, "loss": 0.1628, "step": 41 }, { "epoch": 0.0821917808219178, "grad_norm": 3.002209123109109, "learning_rate": 9.95838659949645e-06, "loss": 0.2282, "step": 42 }, { "epoch": 0.08414872798434442, "grad_norm": 2.453208594204405, "learning_rate": 9.956384338857561e-06, "loss": 0.1697, "step": 43 }, { "epoch": 0.08610567514677103, "grad_norm": 2.1966470429777827, "learning_rate": 9.954335244074575e-06, "loss": 0.178, "step": 44 }, { "epoch": 0.08806262230919765, "grad_norm": 1.9008965951915378, "learning_rate": 9.952239334509909e-06, "loss": 0.1522, "step": 45 }, { "epoch": 0.09001956947162426, "grad_norm": 2.5235937804916326, "learning_rate": 9.950096629968353e-06, "loss": 0.1658, "step": 46 }, { "epoch": 0.09197651663405088, "grad_norm": 2.2467443614313236, "learning_rate": 9.947907150696868e-06, "loss": 0.1716, "step": 47 }, { "epoch": 0.09393346379647749, "grad_norm": 1.993130048799841, "learning_rate": 9.945670917384404e-06, "loss": 0.1395, "step": 48 }, { "epoch": 0.0958904109589041, "grad_norm": 2.4830606887455273, "learning_rate": 9.943387951161702e-06, "loss": 0.2096, "step": 49 }, { "epoch": 0.09784735812133072, "grad_norm": 2.242698337068841, "learning_rate": 9.941058273601097e-06, "loss": 0.2017, "step": 50 }, { "epoch": 0.09980430528375733, "grad_norm": 2.346468108540763, "learning_rate": 9.938681906716305e-06, "loss": 0.1845, "step": 51 }, { "epoch": 0.10176125244618395, "grad_norm": 3.1153352037858935, "learning_rate": 9.936258872962229e-06, "loss": 0.2309, "step": 52 }, { "epoch": 0.10371819960861056, "grad_norm": 2.618321092869847, "learning_rate": 9.93378919523473e-06, "loss": 0.1928, "step": 53 }, { "epoch": 0.10567514677103718, "grad_norm": 2.2861717118265625, "learning_rate": 9.931272896870427e-06, "loss": 0.1787, "step": 54 }, { "epoch": 0.10763209393346379, "grad_norm": 1.5383754217965275, "learning_rate": 9.928710001646467e-06, "loss": 0.1394, "step": 55 }, { "epoch": 0.1095890410958904, "grad_norm": 2.1101019772665377, "learning_rate": 9.926100533780304e-06, "loss": 0.2006, "step": 56 }, { "epoch": 0.11154598825831702, "grad_norm": 2.1920301027614864, "learning_rate": 9.923444517929467e-06, "loss": 0.2206, "step": 57 }, { "epoch": 0.11350293542074363, "grad_norm": 2.5255054965904997, "learning_rate": 9.92074197919133e-06, "loss": 0.1894, "step": 58 }, { "epoch": 0.11545988258317025, "grad_norm": 2.6673373257327038, "learning_rate": 9.917992943102872e-06, "loss": 0.1983, "step": 59 }, { "epoch": 0.11741682974559686, "grad_norm": 2.60298171545859, "learning_rate": 9.91519743564044e-06, "loss": 0.2287, "step": 60 }, { "epoch": 0.11937377690802348, "grad_norm": 2.051201114821871, "learning_rate": 9.912355483219498e-06, "loss": 0.1754, "step": 61 }, { "epoch": 0.12133072407045009, "grad_norm": 2.065215274409325, "learning_rate": 9.909467112694385e-06, "loss": 0.1744, "step": 62 }, { "epoch": 0.1232876712328767, "grad_norm": 3.4138720691523523, "learning_rate": 9.906532351358047e-06, "loss": 0.2377, "step": 63 }, { "epoch": 0.12524461839530332, "grad_norm": 2.538608622468828, "learning_rate": 9.903551226941801e-06, "loss": 0.2066, "step": 64 }, { "epoch": 0.12720156555772993, "grad_norm": 2.058456318748694, "learning_rate": 9.900523767615052e-06, "loss": 0.1668, "step": 65 }, { "epoch": 0.12915851272015655, "grad_norm": 2.2985972063087954, "learning_rate": 9.897450001985038e-06, "loss": 0.1933, "step": 66 }, { "epoch": 0.13111545988258316, "grad_norm": 2.554593542768687, "learning_rate": 9.894329959096559e-06, "loss": 0.2213, "step": 67 }, { "epoch": 0.13307240704500978, "grad_norm": 3.352703367792943, "learning_rate": 9.891163668431696e-06, "loss": 0.2129, "step": 68 }, { "epoch": 0.1350293542074364, "grad_norm": 2.528637288061403, "learning_rate": 9.887951159909541e-06, "loss": 0.2001, "step": 69 }, { "epoch": 0.136986301369863, "grad_norm": 2.3701562189646332, "learning_rate": 9.88469246388591e-06, "loss": 0.1893, "step": 70 }, { "epoch": 0.13894324853228962, "grad_norm": 2.325790997999378, "learning_rate": 9.881387611153053e-06, "loss": 0.1624, "step": 71 }, { "epoch": 0.14090019569471623, "grad_norm": 2.9142325402204405, "learning_rate": 9.878036632939374e-06, "loss": 0.2046, "step": 72 }, { "epoch": 0.14285714285714285, "grad_norm": 2.528347383651363, "learning_rate": 9.874639560909118e-06, "loss": 0.2037, "step": 73 }, { "epoch": 0.14481409001956946, "grad_norm": 2.6083162779817766, "learning_rate": 9.871196427162094e-06, "loss": 0.167, "step": 74 }, { "epoch": 0.14677103718199608, "grad_norm": 2.5701899877203545, "learning_rate": 9.867707264233349e-06, "loss": 0.1893, "step": 75 }, { "epoch": 0.1487279843444227, "grad_norm": 1.8868351735321773, "learning_rate": 9.86417210509288e-06, "loss": 0.1631, "step": 76 }, { "epoch": 0.1506849315068493, "grad_norm": 2.2202772662950507, "learning_rate": 9.860590983145307e-06, "loss": 0.2088, "step": 77 }, { "epoch": 0.15264187866927592, "grad_norm": 2.0680728264239003, "learning_rate": 9.85696393222957e-06, "loss": 0.1604, "step": 78 }, { "epoch": 0.15459882583170254, "grad_norm": 2.8869753035077235, "learning_rate": 9.853290986618604e-06, "loss": 0.1898, "step": 79 }, { "epoch": 0.15655577299412915, "grad_norm": 2.011353859010772, "learning_rate": 9.849572181019008e-06, "loss": 0.1908, "step": 80 }, { "epoch": 0.15851272015655576, "grad_norm": 2.576393691448709, "learning_rate": 9.845807550570727e-06, "loss": 0.2372, "step": 81 }, { "epoch": 0.16046966731898238, "grad_norm": 2.163806683679211, "learning_rate": 9.84199713084672e-06, "loss": 0.1743, "step": 82 }, { "epoch": 0.162426614481409, "grad_norm": 2.2970455026827556, "learning_rate": 9.83814095785262e-06, "loss": 0.1845, "step": 83 }, { "epoch": 0.1643835616438356, "grad_norm": 2.506283235893472, "learning_rate": 9.834239068026388e-06, "loss": 0.1894, "step": 84 }, { "epoch": 0.16634050880626222, "grad_norm": 2.2760793040271916, "learning_rate": 9.830291498237983e-06, "loss": 0.1901, "step": 85 }, { "epoch": 0.16829745596868884, "grad_norm": 1.877744445840279, "learning_rate": 9.826298285789002e-06, "loss": 0.1733, "step": 86 }, { "epoch": 0.17025440313111545, "grad_norm": 2.158463407750118, "learning_rate": 9.822259468412329e-06, "loss": 0.1754, "step": 87 }, { "epoch": 0.17221135029354206, "grad_norm": 3.29567747173194, "learning_rate": 9.818175084271786e-06, "loss": 0.2131, "step": 88 }, { "epoch": 0.17416829745596868, "grad_norm": 2.3502661441307455, "learning_rate": 9.814045171961762e-06, "loss": 0.2078, "step": 89 }, { "epoch": 0.1761252446183953, "grad_norm": 2.363217223914891, "learning_rate": 9.809869770506855e-06, "loss": 0.2103, "step": 90 }, { "epoch": 0.1780821917808219, "grad_norm": 2.2660004995906036, "learning_rate": 9.805648919361505e-06, "loss": 0.1681, "step": 91 }, { "epoch": 0.18003913894324852, "grad_norm": 2.1548580753007793, "learning_rate": 9.801382658409611e-06, "loss": 0.2007, "step": 92 }, { "epoch": 0.18199608610567514, "grad_norm": 2.314214803839974, "learning_rate": 9.797071027964166e-06, "loss": 0.1932, "step": 93 }, { "epoch": 0.18395303326810175, "grad_norm": 2.6088213640266926, "learning_rate": 9.792714068766872e-06, "loss": 0.2024, "step": 94 }, { "epoch": 0.18590998043052837, "grad_norm": 2.101865278823249, "learning_rate": 9.788311821987748e-06, "loss": 0.1662, "step": 95 }, { "epoch": 0.18786692759295498, "grad_norm": 1.8869320266092935, "learning_rate": 9.783864329224752e-06, "loss": 0.1729, "step": 96 }, { "epoch": 0.1898238747553816, "grad_norm": 2.2933888235709525, "learning_rate": 9.779371632503382e-06, "loss": 0.2095, "step": 97 }, { "epoch": 0.1917808219178082, "grad_norm": 2.3401544826431486, "learning_rate": 9.774833774276278e-06, "loss": 0.2187, "step": 98 }, { "epoch": 0.19373776908023482, "grad_norm": 2.714768649678859, "learning_rate": 9.770250797422822e-06, "loss": 0.1864, "step": 99 }, { "epoch": 0.19569471624266144, "grad_norm": 2.539642396343944, "learning_rate": 9.765622745248739e-06, "loss": 0.1867, "step": 100 }, { "epoch": 0.19765166340508805, "grad_norm": 2.842419533391284, "learning_rate": 9.760949661485676e-06, "loss": 0.2057, "step": 101 }, { "epoch": 0.19960861056751467, "grad_norm": 2.0165674459964866, "learning_rate": 9.7562315902908e-06, "loss": 0.1701, "step": 102 }, { "epoch": 0.20156555772994128, "grad_norm": 2.6728143260900477, "learning_rate": 9.75146857624637e-06, "loss": 0.1889, "step": 103 }, { "epoch": 0.2035225048923679, "grad_norm": 1.9946590456213484, "learning_rate": 9.746660664359326e-06, "loss": 0.1964, "step": 104 }, { "epoch": 0.2054794520547945, "grad_norm": 2.37486581376911, "learning_rate": 9.741807900060858e-06, "loss": 0.1995, "step": 105 }, { "epoch": 0.20743639921722112, "grad_norm": 2.837224439520139, "learning_rate": 9.73691032920598e-06, "loss": 0.1874, "step": 106 }, { "epoch": 0.20939334637964774, "grad_norm": 2.5989351026818737, "learning_rate": 9.731967998073087e-06, "loss": 0.1748, "step": 107 }, { "epoch": 0.21135029354207435, "grad_norm": 2.0889698689227787, "learning_rate": 9.726980953363536e-06, "loss": 0.1767, "step": 108 }, { "epoch": 0.21330724070450097, "grad_norm": 2.1471364387204064, "learning_rate": 9.721949242201185e-06, "loss": 0.1999, "step": 109 }, { "epoch": 0.21526418786692758, "grad_norm": 2.641138246267923, "learning_rate": 9.716872912131964e-06, "loss": 0.1916, "step": 110 }, { "epoch": 0.2172211350293542, "grad_norm": 2.5380983562545008, "learning_rate": 9.711752011123412e-06, "loss": 0.2278, "step": 111 }, { "epoch": 0.2191780821917808, "grad_norm": 2.194901832364108, "learning_rate": 9.706586587564236e-06, "loss": 0.1653, "step": 112 }, { "epoch": 0.22113502935420742, "grad_norm": 2.7696358575386504, "learning_rate": 9.701376690263845e-06, "loss": 0.2252, "step": 113 }, { "epoch": 0.22309197651663404, "grad_norm": 2.4845270754032267, "learning_rate": 9.696122368451887e-06, "loss": 0.2359, "step": 114 }, { "epoch": 0.22504892367906065, "grad_norm": 2.39879428931212, "learning_rate": 9.690823671777799e-06, "loss": 0.2078, "step": 115 }, { "epoch": 0.22700587084148727, "grad_norm": 1.5365655684112836, "learning_rate": 9.685480650310319e-06, "loss": 0.1179, "step": 116 }, { "epoch": 0.22896281800391388, "grad_norm": 1.9032477520882891, "learning_rate": 9.680093354537027e-06, "loss": 0.1732, "step": 117 }, { "epoch": 0.2309197651663405, "grad_norm": 2.526793495598011, "learning_rate": 9.67466183536386e-06, "loss": 0.211, "step": 118 }, { "epoch": 0.2328767123287671, "grad_norm": 1.9175798607590133, "learning_rate": 9.669186144114627e-06, "loss": 0.1449, "step": 119 }, { "epoch": 0.23483365949119372, "grad_norm": 2.6481069327991937, "learning_rate": 9.663666332530541e-06, "loss": 0.239, "step": 120 }, { "epoch": 0.23679060665362034, "grad_norm": 2.287350406623119, "learning_rate": 9.658102452769712e-06, "loss": 0.2139, "step": 121 }, { "epoch": 0.23874755381604695, "grad_norm": 2.0343684178439236, "learning_rate": 9.652494557406666e-06, "loss": 0.205, "step": 122 }, { "epoch": 0.24070450097847357, "grad_norm": 2.4931874551608515, "learning_rate": 9.646842699431842e-06, "loss": 0.263, "step": 123 }, { "epoch": 0.24266144814090018, "grad_norm": 2.2755457031101405, "learning_rate": 9.641146932251088e-06, "loss": 0.1872, "step": 124 }, { "epoch": 0.2446183953033268, "grad_norm": 2.106933822517455, "learning_rate": 9.635407309685171e-06, "loss": 0.164, "step": 125 }, { "epoch": 0.2465753424657534, "grad_norm": 2.199373482317882, "learning_rate": 9.62962388596925e-06, "loss": 0.1922, "step": 126 }, { "epoch": 0.24853228962818003, "grad_norm": 1.8826906309857112, "learning_rate": 9.623796715752378e-06, "loss": 0.1391, "step": 127 }, { "epoch": 0.25048923679060664, "grad_norm": 2.045212898720031, "learning_rate": 9.617925854096975e-06, "loss": 0.2025, "step": 128 }, { "epoch": 0.25244618395303325, "grad_norm": 2.2672309182448407, "learning_rate": 9.61201135647832e-06, "loss": 0.1956, "step": 129 }, { "epoch": 0.25440313111545987, "grad_norm": 2.3887364310142045, "learning_rate": 9.606053278784009e-06, "loss": 0.2228, "step": 130 }, { "epoch": 0.2563600782778865, "grad_norm": 2.07394809445296, "learning_rate": 9.600051677313446e-06, "loss": 0.1819, "step": 131 }, { "epoch": 0.2583170254403131, "grad_norm": 2.4959261932305528, "learning_rate": 9.5940066087773e-06, "loss": 0.2079, "step": 132 }, { "epoch": 0.2602739726027397, "grad_norm": 1.971351687876825, "learning_rate": 9.587918130296969e-06, "loss": 0.1587, "step": 133 }, { "epoch": 0.2622309197651663, "grad_norm": 2.76991712964986, "learning_rate": 9.581786299404046e-06, "loss": 0.1818, "step": 134 }, { "epoch": 0.26418786692759294, "grad_norm": 2.067772303514755, "learning_rate": 9.57561117403977e-06, "loss": 0.1881, "step": 135 }, { "epoch": 0.26614481409001955, "grad_norm": 2.012569369643021, "learning_rate": 9.56939281255448e-06, "loss": 0.1642, "step": 136 }, { "epoch": 0.26810176125244617, "grad_norm": 2.5657028933311237, "learning_rate": 9.563131273707065e-06, "loss": 0.192, "step": 137 }, { "epoch": 0.2700587084148728, "grad_norm": 2.2331355911318616, "learning_rate": 9.556826616664408e-06, "loss": 0.2156, "step": 138 }, { "epoch": 0.2720156555772994, "grad_norm": 2.3608166591087016, "learning_rate": 9.550478901000827e-06, "loss": 0.1981, "step": 139 }, { "epoch": 0.273972602739726, "grad_norm": 3.5251113985327844, "learning_rate": 9.544088186697515e-06, "loss": 0.2531, "step": 140 }, { "epoch": 0.2759295499021526, "grad_norm": 1.9992605701701254, "learning_rate": 9.537654534141964e-06, "loss": 0.1639, "step": 141 }, { "epoch": 0.27788649706457924, "grad_norm": 2.3010614293608715, "learning_rate": 9.531178004127404e-06, "loss": 0.2007, "step": 142 }, { "epoch": 0.27984344422700586, "grad_norm": 2.104994845948875, "learning_rate": 9.524658657852223e-06, "loss": 0.1803, "step": 143 }, { "epoch": 0.28180039138943247, "grad_norm": 2.0383949197659366, "learning_rate": 9.518096556919396e-06, "loss": 0.2103, "step": 144 }, { "epoch": 0.2837573385518591, "grad_norm": 2.645456861377443, "learning_rate": 9.511491763335893e-06, "loss": 0.1903, "step": 145 }, { "epoch": 0.2857142857142857, "grad_norm": 2.267974700181682, "learning_rate": 9.504844339512096e-06, "loss": 0.1859, "step": 146 }, { "epoch": 0.2876712328767123, "grad_norm": 2.661747586804285, "learning_rate": 9.498154348261217e-06, "loss": 0.209, "step": 147 }, { "epoch": 0.2896281800391389, "grad_norm": 1.7300688863380262, "learning_rate": 9.491421852798695e-06, "loss": 0.1559, "step": 148 }, { "epoch": 0.29158512720156554, "grad_norm": 2.1115166329825437, "learning_rate": 9.484646916741602e-06, "loss": 0.1776, "step": 149 }, { "epoch": 0.29354207436399216, "grad_norm": 2.2896687163421006, "learning_rate": 9.477829604108044e-06, "loss": 0.179, "step": 150 }, { "epoch": 0.29549902152641877, "grad_norm": 2.680411643514485, "learning_rate": 9.47096997931655e-06, "loss": 0.2004, "step": 151 }, { "epoch": 0.2974559686888454, "grad_norm": 1.783504748923097, "learning_rate": 9.464068107185476e-06, "loss": 0.1706, "step": 152 }, { "epoch": 0.299412915851272, "grad_norm": 2.6450496087784034, "learning_rate": 9.45712405293237e-06, "loss": 0.2552, "step": 153 }, { "epoch": 0.3013698630136986, "grad_norm": 2.467768619445763, "learning_rate": 9.450137882173385e-06, "loss": 0.2343, "step": 154 }, { "epoch": 0.30332681017612523, "grad_norm": 2.875954106585473, "learning_rate": 9.443109660922632e-06, "loss": 0.1883, "step": 155 }, { "epoch": 0.30528375733855184, "grad_norm": 2.141532859100036, "learning_rate": 9.436039455591574e-06, "loss": 0.1673, "step": 156 }, { "epoch": 0.30724070450097846, "grad_norm": 2.088990067684385, "learning_rate": 9.42892733298839e-06, "loss": 0.193, "step": 157 }, { "epoch": 0.30919765166340507, "grad_norm": 2.5539004361373063, "learning_rate": 9.421773360317348e-06, "loss": 0.226, "step": 158 }, { "epoch": 0.3111545988258317, "grad_norm": 2.276029839206782, "learning_rate": 9.414577605178166e-06, "loss": 0.2149, "step": 159 }, { "epoch": 0.3131115459882583, "grad_norm": 2.1692685720976663, "learning_rate": 9.407340135565375e-06, "loss": 0.2116, "step": 160 }, { "epoch": 0.3150684931506849, "grad_norm": 2.466804623763645, "learning_rate": 9.40006101986768e-06, "loss": 0.2279, "step": 161 }, { "epoch": 0.31702544031311153, "grad_norm": 2.004348996122027, "learning_rate": 9.392740326867304e-06, "loss": 0.1809, "step": 162 }, { "epoch": 0.31898238747553814, "grad_norm": 1.853458016380264, "learning_rate": 9.385378125739352e-06, "loss": 0.1826, "step": 163 }, { "epoch": 0.32093933463796476, "grad_norm": 2.0541740353658104, "learning_rate": 9.377974486051149e-06, "loss": 0.1874, "step": 164 }, { "epoch": 0.32289628180039137, "grad_norm": 1.9561140958446408, "learning_rate": 9.370529477761574e-06, "loss": 0.159, "step": 165 }, { "epoch": 0.324853228962818, "grad_norm": 2.0907068609361907, "learning_rate": 9.363043171220423e-06, "loss": 0.1768, "step": 166 }, { "epoch": 0.3268101761252446, "grad_norm": 1.9106766699400612, "learning_rate": 9.355515637167719e-06, "loss": 0.1887, "step": 167 }, { "epoch": 0.3287671232876712, "grad_norm": 2.179281323108424, "learning_rate": 9.347946946733055e-06, "loss": 0.1898, "step": 168 }, { "epoch": 0.33072407045009783, "grad_norm": 2.007791174027861, "learning_rate": 9.34033717143493e-06, "loss": 0.1954, "step": 169 }, { "epoch": 0.33268101761252444, "grad_norm": 2.0535821031370607, "learning_rate": 9.332686383180055e-06, "loss": 0.1908, "step": 170 }, { "epoch": 0.33463796477495106, "grad_norm": 2.6591309379434938, "learning_rate": 9.324994654262687e-06, "loss": 0.2504, "step": 171 }, { "epoch": 0.33659491193737767, "grad_norm": 2.41811095667973, "learning_rate": 9.31726205736394e-06, "loss": 0.1985, "step": 172 }, { "epoch": 0.3385518590998043, "grad_norm": 2.4198631490303537, "learning_rate": 9.309488665551104e-06, "loss": 0.1864, "step": 173 }, { "epoch": 0.3405088062622309, "grad_norm": 2.333457588281709, "learning_rate": 9.301674552276942e-06, "loss": 0.1848, "step": 174 }, { "epoch": 0.3424657534246575, "grad_norm": 2.204664249713328, "learning_rate": 9.293819791379016e-06, "loss": 0.1969, "step": 175 }, { "epoch": 0.34442270058708413, "grad_norm": 1.697357119388807, "learning_rate": 9.28592445707897e-06, "loss": 0.1554, "step": 176 }, { "epoch": 0.34637964774951074, "grad_norm": 2.3843282012067353, "learning_rate": 9.277988623981837e-06, "loss": 0.1718, "step": 177 }, { "epoch": 0.34833659491193736, "grad_norm": 2.2703118748704063, "learning_rate": 9.270012367075337e-06, "loss": 0.1906, "step": 178 }, { "epoch": 0.350293542074364, "grad_norm": 2.629412922174708, "learning_rate": 9.26199576172916e-06, "loss": 0.2346, "step": 179 }, { "epoch": 0.3522504892367906, "grad_norm": 2.083801792811563, "learning_rate": 9.253938883694266e-06, "loss": 0.2122, "step": 180 }, { "epoch": 0.3542074363992172, "grad_norm": 1.9273334744872532, "learning_rate": 9.245841809102156e-06, "loss": 0.1828, "step": 181 }, { "epoch": 0.3561643835616438, "grad_norm": 2.6856359733216975, "learning_rate": 9.237704614464157e-06, "loss": 0.1956, "step": 182 }, { "epoch": 0.35812133072407043, "grad_norm": 2.012782311372279, "learning_rate": 9.229527376670706e-06, "loss": 0.1673, "step": 183 }, { "epoch": 0.36007827788649704, "grad_norm": 1.7413241271004036, "learning_rate": 9.221310172990616e-06, "loss": 0.1586, "step": 184 }, { "epoch": 0.36203522504892366, "grad_norm": 1.8202573242650377, "learning_rate": 9.213053081070346e-06, "loss": 0.1577, "step": 185 }, { "epoch": 0.3639921722113503, "grad_norm": 2.495626259177546, "learning_rate": 9.204756178933274e-06, "loss": 0.1813, "step": 186 }, { "epoch": 0.3659491193737769, "grad_norm": 2.103311481647965, "learning_rate": 9.19641954497895e-06, "loss": 0.1679, "step": 187 }, { "epoch": 0.3679060665362035, "grad_norm": 2.2478571413783657, "learning_rate": 9.18804325798236e-06, "loss": 0.2404, "step": 188 }, { "epoch": 0.3698630136986301, "grad_norm": 2.106552610701366, "learning_rate": 9.179627397093184e-06, "loss": 0.1809, "step": 189 }, { "epoch": 0.37181996086105673, "grad_norm": 2.2420107376651432, "learning_rate": 9.171172041835048e-06, "loss": 0.1626, "step": 190 }, { "epoch": 0.37377690802348335, "grad_norm": 1.9267166044248112, "learning_rate": 9.162677272104767e-06, "loss": 0.1863, "step": 191 }, { "epoch": 0.37573385518590996, "grad_norm": 2.0936954713929956, "learning_rate": 9.154143168171594e-06, "loss": 0.1706, "step": 192 }, { "epoch": 0.3776908023483366, "grad_norm": 1.757686640682866, "learning_rate": 9.145569810676458e-06, "loss": 0.1514, "step": 193 }, { "epoch": 0.3796477495107632, "grad_norm": 1.92132065534092, "learning_rate": 9.136957280631212e-06, "loss": 0.1392, "step": 194 }, { "epoch": 0.3816046966731898, "grad_norm": 2.0559547811253833, "learning_rate": 9.128305659417856e-06, "loss": 0.162, "step": 195 }, { "epoch": 0.3835616438356164, "grad_norm": 1.8368572598248154, "learning_rate": 9.119615028787771e-06, "loss": 0.1728, "step": 196 }, { "epoch": 0.38551859099804303, "grad_norm": 2.232236063570786, "learning_rate": 9.110885470860953e-06, "loss": 0.2227, "step": 197 }, { "epoch": 0.38747553816046965, "grad_norm": 2.3551421348686916, "learning_rate": 9.102117068125227e-06, "loss": 0.222, "step": 198 }, { "epoch": 0.38943248532289626, "grad_norm": 1.884160399258136, "learning_rate": 9.093309903435475e-06, "loss": 0.1758, "step": 199 }, { "epoch": 0.3913894324853229, "grad_norm": 2.909006293130323, "learning_rate": 9.084464060012849e-06, "loss": 0.2101, "step": 200 }, { "epoch": 0.3933463796477495, "grad_norm": 1.708397953998908, "learning_rate": 9.075579621443991e-06, "loss": 0.1431, "step": 201 }, { "epoch": 0.3953033268101761, "grad_norm": 2.4253091484291622, "learning_rate": 9.066656671680231e-06, "loss": 0.1752, "step": 202 }, { "epoch": 0.3972602739726027, "grad_norm": 2.083118895068761, "learning_rate": 9.057695295036806e-06, "loss": 0.1847, "step": 203 }, { "epoch": 0.39921722113502933, "grad_norm": 2.3372354931659225, "learning_rate": 9.048695576192058e-06, "loss": 0.1945, "step": 204 }, { "epoch": 0.40117416829745595, "grad_norm": 1.9672985346550766, "learning_rate": 9.039657600186633e-06, "loss": 0.1654, "step": 205 }, { "epoch": 0.40313111545988256, "grad_norm": 2.290049444795431, "learning_rate": 9.03058145242268e-06, "loss": 0.185, "step": 206 }, { "epoch": 0.4050880626223092, "grad_norm": 2.1219716130733284, "learning_rate": 9.021467218663042e-06, "loss": 0.1929, "step": 207 }, { "epoch": 0.4070450097847358, "grad_norm": 2.267021821037719, "learning_rate": 9.012314985030445e-06, "loss": 0.1908, "step": 208 }, { "epoch": 0.4090019569471624, "grad_norm": 2.5291016894303473, "learning_rate": 9.003124838006689e-06, "loss": 0.196, "step": 209 }, { "epoch": 0.410958904109589, "grad_norm": 1.6409829164357497, "learning_rate": 8.993896864431825e-06, "loss": 0.1548, "step": 210 }, { "epoch": 0.41291585127201563, "grad_norm": 1.8232246442960565, "learning_rate": 8.984631151503335e-06, "loss": 0.1401, "step": 211 }, { "epoch": 0.41487279843444225, "grad_norm": 2.0668274214588127, "learning_rate": 8.975327786775316e-06, "loss": 0.1802, "step": 212 }, { "epoch": 0.41682974559686886, "grad_norm": 2.408691875995061, "learning_rate": 8.965986858157636e-06, "loss": 0.1874, "step": 213 }, { "epoch": 0.4187866927592955, "grad_norm": 2.373086256676123, "learning_rate": 8.956608453915126e-06, "loss": 0.2004, "step": 214 }, { "epoch": 0.4207436399217221, "grad_norm": 2.0168186910015455, "learning_rate": 8.947192662666724e-06, "loss": 0.1621, "step": 215 }, { "epoch": 0.4227005870841487, "grad_norm": 2.2396474440895657, "learning_rate": 8.937739573384653e-06, "loss": 0.1505, "step": 216 }, { "epoch": 0.4246575342465753, "grad_norm": 2.428725519073936, "learning_rate": 8.928249275393572e-06, "loss": 0.2176, "step": 217 }, { "epoch": 0.42661448140900193, "grad_norm": 1.6515267386212091, "learning_rate": 8.918721858369738e-06, "loss": 0.1375, "step": 218 }, { "epoch": 0.42857142857142855, "grad_norm": 1.8034020547316982, "learning_rate": 8.90915741234015e-06, "loss": 0.2002, "step": 219 }, { "epoch": 0.43052837573385516, "grad_norm": 2.3982364821207502, "learning_rate": 8.899556027681708e-06, "loss": 0.175, "step": 220 }, { "epoch": 0.4324853228962818, "grad_norm": 1.8534318281975142, "learning_rate": 8.88991779512035e-06, "loss": 0.1595, "step": 221 }, { "epoch": 0.4344422700587084, "grad_norm": 2.1890195026567145, "learning_rate": 8.880242805730208e-06, "loss": 0.192, "step": 222 }, { "epoch": 0.436399217221135, "grad_norm": 1.7950602502675235, "learning_rate": 8.870531150932727e-06, "loss": 0.1961, "step": 223 }, { "epoch": 0.4383561643835616, "grad_norm": 2.344609546525777, "learning_rate": 8.860782922495821e-06, "loss": 0.1542, "step": 224 }, { "epoch": 0.44031311154598823, "grad_norm": 1.9299884012856776, "learning_rate": 8.850998212532998e-06, "loss": 0.1888, "step": 225 }, { "epoch": 0.44227005870841485, "grad_norm": 2.158709464028409, "learning_rate": 8.84117711350248e-06, "loss": 0.1885, "step": 226 }, { "epoch": 0.44422700587084146, "grad_norm": 2.0025777361816406, "learning_rate": 8.831319718206353e-06, "loss": 0.1621, "step": 227 }, { "epoch": 0.4461839530332681, "grad_norm": 2.195834851911987, "learning_rate": 8.821426119789662e-06, "loss": 0.1797, "step": 228 }, { "epoch": 0.4481409001956947, "grad_norm": 1.9966357964729864, "learning_rate": 8.811496411739552e-06, "loss": 0.1779, "step": 229 }, { "epoch": 0.4500978473581213, "grad_norm": 2.4306247694641336, "learning_rate": 8.801530687884378e-06, "loss": 0.2296, "step": 230 }, { "epoch": 0.4520547945205479, "grad_norm": 2.084307447587978, "learning_rate": 8.791529042392813e-06, "loss": 0.1798, "step": 231 }, { "epoch": 0.45401174168297453, "grad_norm": 2.49060258234432, "learning_rate": 8.781491569772966e-06, "loss": 0.235, "step": 232 }, { "epoch": 0.45596868884540115, "grad_norm": 1.8685581777124989, "learning_rate": 8.771418364871483e-06, "loss": 0.1678, "step": 233 }, { "epoch": 0.45792563600782776, "grad_norm": 2.419601316035739, "learning_rate": 8.761309522872657e-06, "loss": 0.1401, "step": 234 }, { "epoch": 0.4598825831702544, "grad_norm": 1.9789310526904225, "learning_rate": 8.751165139297522e-06, "loss": 0.2304, "step": 235 }, { "epoch": 0.461839530332681, "grad_norm": 1.9933502096750053, "learning_rate": 8.740985310002956e-06, "loss": 0.1509, "step": 236 }, { "epoch": 0.4637964774951076, "grad_norm": 1.7777907988307426, "learning_rate": 8.730770131180771e-06, "loss": 0.1314, "step": 237 }, { "epoch": 0.4657534246575342, "grad_norm": 2.134939285855735, "learning_rate": 8.720519699356804e-06, "loss": 0.1974, "step": 238 }, { "epoch": 0.46771037181996084, "grad_norm": 1.771483391447446, "learning_rate": 8.71023411139001e-06, "loss": 0.1219, "step": 239 }, { "epoch": 0.46966731898238745, "grad_norm": 2.297080058977332, "learning_rate": 8.699913464471543e-06, "loss": 0.1761, "step": 240 }, { "epoch": 0.47162426614481406, "grad_norm": 1.986317361762347, "learning_rate": 8.689557856123838e-06, "loss": 0.1806, "step": 241 }, { "epoch": 0.4735812133072407, "grad_norm": 1.858506671195948, "learning_rate": 8.679167384199686e-06, "loss": 0.1571, "step": 242 }, { "epoch": 0.4755381604696673, "grad_norm": 2.2310935891754173, "learning_rate": 8.668742146881316e-06, "loss": 0.1837, "step": 243 }, { "epoch": 0.4774951076320939, "grad_norm": 1.905330981900993, "learning_rate": 8.658282242679461e-06, "loss": 0.1631, "step": 244 }, { "epoch": 0.4794520547945205, "grad_norm": 1.8485653384537961, "learning_rate": 8.647787770432439e-06, "loss": 0.1424, "step": 245 }, { "epoch": 0.48140900195694714, "grad_norm": 2.243680140592553, "learning_rate": 8.6372588293052e-06, "loss": 0.2144, "step": 246 }, { "epoch": 0.48336594911937375, "grad_norm": 1.693058551501593, "learning_rate": 8.626695518788403e-06, "loss": 0.1619, "step": 247 }, { "epoch": 0.48532289628180036, "grad_norm": 2.401070711371156, "learning_rate": 8.616097938697476e-06, "loss": 0.2077, "step": 248 }, { "epoch": 0.487279843444227, "grad_norm": 1.8012387800182308, "learning_rate": 8.60546618917166e-06, "loss": 0.1957, "step": 249 }, { "epoch": 0.4892367906066536, "grad_norm": 2.289655321438479, "learning_rate": 8.594800370673083e-06, "loss": 0.2255, "step": 250 }, { "epoch": 0.4911937377690802, "grad_norm": 1.8079750976106275, "learning_rate": 8.584100583985791e-06, "loss": 0.1941, "step": 251 }, { "epoch": 0.4931506849315068, "grad_norm": 1.7718034129885667, "learning_rate": 8.573366930214807e-06, "loss": 0.1589, "step": 252 }, { "epoch": 0.49510763209393344, "grad_norm": 2.026525384886368, "learning_rate": 8.562599510785171e-06, "loss": 0.2021, "step": 253 }, { "epoch": 0.49706457925636005, "grad_norm": 2.2066756698093513, "learning_rate": 8.551798427440985e-06, "loss": 0.1945, "step": 254 }, { "epoch": 0.49902152641878667, "grad_norm": 1.8560579225167089, "learning_rate": 8.540963782244455e-06, "loss": 0.1586, "step": 255 }, { "epoch": 0.5009784735812133, "grad_norm": 2.4568690226426795, "learning_rate": 8.53009567757491e-06, "loss": 0.1967, "step": 256 }, { "epoch": 0.50293542074364, "grad_norm": 1.8472898016661812, "learning_rate": 8.519194216127854e-06, "loss": 0.1709, "step": 257 }, { "epoch": 0.5048923679060665, "grad_norm": 1.4850311919964274, "learning_rate": 8.50825950091399e-06, "loss": 0.1541, "step": 258 }, { "epoch": 0.5068493150684932, "grad_norm": 2.290885407508907, "learning_rate": 8.497291635258235e-06, "loss": 0.2281, "step": 259 }, { "epoch": 0.5088062622309197, "grad_norm": 2.247567268205798, "learning_rate": 8.486290722798765e-06, "loss": 0.1931, "step": 260 }, { "epoch": 0.5107632093933464, "grad_norm": 1.8653759054374042, "learning_rate": 8.475256867486011e-06, "loss": 0.1585, "step": 261 }, { "epoch": 0.512720156555773, "grad_norm": 1.9186753676037949, "learning_rate": 8.464190173581698e-06, "loss": 0.1536, "step": 262 }, { "epoch": 0.5146771037181996, "grad_norm": 1.5517910525189156, "learning_rate": 8.45309074565785e-06, "loss": 0.1379, "step": 263 }, { "epoch": 0.5166340508806262, "grad_norm": 1.9894797507306439, "learning_rate": 8.441958688595802e-06, "loss": 0.1867, "step": 264 }, { "epoch": 0.5185909980430529, "grad_norm": 1.7381019556095967, "learning_rate": 8.430794107585208e-06, "loss": 0.171, "step": 265 }, { "epoch": 0.5205479452054794, "grad_norm": 2.488154754731573, "learning_rate": 8.419597108123054e-06, "loss": 0.228, "step": 266 }, { "epoch": 0.5225048923679061, "grad_norm": 1.788223978703378, "learning_rate": 8.408367796012652e-06, "loss": 0.1674, "step": 267 }, { "epoch": 0.5244618395303327, "grad_norm": 1.679541627280701, "learning_rate": 8.397106277362647e-06, "loss": 0.1662, "step": 268 }, { "epoch": 0.5264187866927593, "grad_norm": 2.2646330772853958, "learning_rate": 8.38581265858601e-06, "loss": 0.152, "step": 269 }, { "epoch": 0.5283757338551859, "grad_norm": 2.2813559239270425, "learning_rate": 8.374487046399035e-06, "loss": 0.1992, "step": 270 }, { "epoch": 0.5303326810176126, "grad_norm": 1.791152622464371, "learning_rate": 8.363129547820333e-06, "loss": 0.182, "step": 271 }, { "epoch": 0.5322896281800391, "grad_norm": 2.0490917756310267, "learning_rate": 8.35174027016981e-06, "loss": 0.1556, "step": 272 }, { "epoch": 0.5342465753424658, "grad_norm": 2.4565634995885746, "learning_rate": 8.340319321067668e-06, "loss": 0.2128, "step": 273 }, { "epoch": 0.5362035225048923, "grad_norm": 2.244252399290381, "learning_rate": 8.328866808433378e-06, "loss": 0.2062, "step": 274 }, { "epoch": 0.538160469667319, "grad_norm": 1.9952336179248558, "learning_rate": 8.317382840484663e-06, "loss": 0.1954, "step": 275 }, { "epoch": 0.5401174168297456, "grad_norm": 2.278877673356308, "learning_rate": 8.305867525736475e-06, "loss": 0.2083, "step": 276 }, { "epoch": 0.5420743639921722, "grad_norm": 2.166267832336264, "learning_rate": 8.29432097299997e-06, "loss": 0.1776, "step": 277 }, { "epoch": 0.5440313111545988, "grad_norm": 2.106611535194033, "learning_rate": 8.28274329138148e-06, "loss": 0.2122, "step": 278 }, { "epoch": 0.5459882583170255, "grad_norm": 2.605623070324236, "learning_rate": 8.271134590281482e-06, "loss": 0.1872, "step": 279 }, { "epoch": 0.547945205479452, "grad_norm": 1.8874904167790443, "learning_rate": 8.259494979393563e-06, "loss": 0.2001, "step": 280 }, { "epoch": 0.5499021526418787, "grad_norm": 1.8775270285087355, "learning_rate": 8.247824568703386e-06, "loss": 0.1603, "step": 281 }, { "epoch": 0.5518590998043053, "grad_norm": 1.9046669064295658, "learning_rate": 8.236123468487649e-06, "loss": 0.1657, "step": 282 }, { "epoch": 0.5538160469667319, "grad_norm": 2.6022000895419484, "learning_rate": 8.224391789313038e-06, "loss": 0.196, "step": 283 }, { "epoch": 0.5557729941291585, "grad_norm": 2.080635262355894, "learning_rate": 8.2126296420352e-06, "loss": 0.1827, "step": 284 }, { "epoch": 0.5577299412915852, "grad_norm": 2.2866313251275616, "learning_rate": 8.200837137797665e-06, "loss": 0.2451, "step": 285 }, { "epoch": 0.5596868884540117, "grad_norm": 1.9752467321582332, "learning_rate": 8.189014388030834e-06, "loss": 0.1849, "step": 286 }, { "epoch": 0.5616438356164384, "grad_norm": 1.559767067112115, "learning_rate": 8.177161504450887e-06, "loss": 0.1336, "step": 287 }, { "epoch": 0.5636007827788649, "grad_norm": 1.6970997545202657, "learning_rate": 8.16527859905876e-06, "loss": 0.1564, "step": 288 }, { "epoch": 0.5655577299412916, "grad_norm": 1.7605194148405368, "learning_rate": 8.153365784139065e-06, "loss": 0.1696, "step": 289 }, { "epoch": 0.5675146771037182, "grad_norm": 1.7104480036418666, "learning_rate": 8.141423172259038e-06, "loss": 0.1485, "step": 290 }, { "epoch": 0.5694716242661448, "grad_norm": 2.533331915343481, "learning_rate": 8.129450876267475e-06, "loss": 0.1965, "step": 291 }, { "epoch": 0.5714285714285714, "grad_norm": 3.542421114815424, "learning_rate": 8.117449009293668e-06, "loss": 0.193, "step": 292 }, { "epoch": 0.5733855185909981, "grad_norm": 1.9547569242431988, "learning_rate": 8.105417684746327e-06, "loss": 0.1821, "step": 293 }, { "epoch": 0.5753424657534246, "grad_norm": 1.6554608535956121, "learning_rate": 8.093357016312518e-06, "loss": 0.1558, "step": 294 }, { "epoch": 0.5772994129158513, "grad_norm": 2.0212523928237576, "learning_rate": 8.081267117956581e-06, "loss": 0.1469, "step": 295 }, { "epoch": 0.5792563600782779, "grad_norm": 1.969212165281817, "learning_rate": 8.069148103919064e-06, "loss": 0.13, "step": 296 }, { "epoch": 0.5812133072407045, "grad_norm": 2.5571093818746373, "learning_rate": 8.057000088715631e-06, "loss": 0.2494, "step": 297 }, { "epoch": 0.5831702544031311, "grad_norm": 1.8240588278905052, "learning_rate": 8.044823187135984e-06, "loss": 0.1434, "step": 298 }, { "epoch": 0.5851272015655578, "grad_norm": 2.17447227309471, "learning_rate": 8.032617514242784e-06, "loss": 0.2344, "step": 299 }, { "epoch": 0.5870841487279843, "grad_norm": 1.9349219091709542, "learning_rate": 8.020383185370559e-06, "loss": 0.1344, "step": 300 }, { "epoch": 0.589041095890411, "grad_norm": 2.689991943951801, "learning_rate": 8.008120316124612e-06, "loss": 0.2232, "step": 301 }, { "epoch": 0.5909980430528375, "grad_norm": 1.9024901172037625, "learning_rate": 7.995829022379937e-06, "loss": 0.1844, "step": 302 }, { "epoch": 0.5929549902152642, "grad_norm": 1.7623051628915145, "learning_rate": 7.983509420280107e-06, "loss": 0.1506, "step": 303 }, { "epoch": 0.5949119373776908, "grad_norm": 2.2628483905778176, "learning_rate": 7.9711616262362e-06, "loss": 0.1702, "step": 304 }, { "epoch": 0.5968688845401174, "grad_norm": 2.493473373186895, "learning_rate": 7.958785756925684e-06, "loss": 0.2085, "step": 305 }, { "epoch": 0.598825831702544, "grad_norm": 1.6315685957398147, "learning_rate": 7.94638192929131e-06, "loss": 0.1432, "step": 306 }, { "epoch": 0.6007827788649707, "grad_norm": 1.8894687460732125, "learning_rate": 7.933950260540023e-06, "loss": 0.1802, "step": 307 }, { "epoch": 0.6027397260273972, "grad_norm": 1.951380004371473, "learning_rate": 7.921490868141843e-06, "loss": 0.1842, "step": 308 }, { "epoch": 0.6046966731898239, "grad_norm": 1.6529494524031472, "learning_rate": 7.909003869828757e-06, "loss": 0.1601, "step": 309 }, { "epoch": 0.6066536203522505, "grad_norm": 1.9312126402795802, "learning_rate": 7.896489383593606e-06, "loss": 0.1225, "step": 310 }, { "epoch": 0.6086105675146771, "grad_norm": 1.886135721395858, "learning_rate": 7.883947527688976e-06, "loss": 0.1401, "step": 311 }, { "epoch": 0.6105675146771037, "grad_norm": 2.0009086990594276, "learning_rate": 7.871378420626072e-06, "loss": 0.1786, "step": 312 }, { "epoch": 0.6125244618395304, "grad_norm": 2.328336503581875, "learning_rate": 7.858782181173602e-06, "loss": 0.1721, "step": 313 }, { "epoch": 0.6144814090019569, "grad_norm": 1.9418571152494646, "learning_rate": 7.84615892835666e-06, "loss": 0.1584, "step": 314 }, { "epoch": 0.6164383561643836, "grad_norm": 2.0653750559473147, "learning_rate": 7.83350878145559e-06, "loss": 0.1977, "step": 315 }, { "epoch": 0.6183953033268101, "grad_norm": 2.4047078796990924, "learning_rate": 7.820831860004867e-06, "loss": 0.1573, "step": 316 }, { "epoch": 0.6203522504892368, "grad_norm": 2.1238681183064365, "learning_rate": 7.808128283791967e-06, "loss": 0.1869, "step": 317 }, { "epoch": 0.6223091976516634, "grad_norm": 1.5562890330930153, "learning_rate": 7.795398172856234e-06, "loss": 0.1295, "step": 318 }, { "epoch": 0.62426614481409, "grad_norm": 1.7923435077027958, "learning_rate": 7.782641647487741e-06, "loss": 0.1674, "step": 319 }, { "epoch": 0.6262230919765166, "grad_norm": 1.9964812138260328, "learning_rate": 7.769858828226165e-06, "loss": 0.1631, "step": 320 }, { "epoch": 0.6281800391389433, "grad_norm": 1.68851778281113, "learning_rate": 7.757049835859635e-06, "loss": 0.1405, "step": 321 }, { "epoch": 0.6301369863013698, "grad_norm": 2.366839833451117, "learning_rate": 7.744214791423597e-06, "loss": 0.1943, "step": 322 }, { "epoch": 0.6320939334637965, "grad_norm": 2.0470616719137604, "learning_rate": 7.731353816199672e-06, "loss": 0.1486, "step": 323 }, { "epoch": 0.6340508806262231, "grad_norm": 2.53362305052429, "learning_rate": 7.718467031714506e-06, "loss": 0.1877, "step": 324 }, { "epoch": 0.6360078277886497, "grad_norm": 2.12240328286498, "learning_rate": 7.705554559738623e-06, "loss": 0.215, "step": 325 }, { "epoch": 0.6379647749510763, "grad_norm": 1.9609152594756358, "learning_rate": 7.692616522285278e-06, "loss": 0.1528, "step": 326 }, { "epoch": 0.639921722113503, "grad_norm": 1.6972094958050024, "learning_rate": 7.679653041609296e-06, "loss": 0.1437, "step": 327 }, { "epoch": 0.6418786692759295, "grad_norm": 2.0174008819360805, "learning_rate": 7.666664240205922e-06, "loss": 0.2119, "step": 328 }, { "epoch": 0.6438356164383562, "grad_norm": 2.155851142211515, "learning_rate": 7.653650240809667e-06, "loss": 0.1836, "step": 329 }, { "epoch": 0.6457925636007827, "grad_norm": 1.9647796104488986, "learning_rate": 7.640611166393142e-06, "loss": 0.162, "step": 330 }, { "epoch": 0.6477495107632094, "grad_norm": 1.6365446711331375, "learning_rate": 7.627547140165899e-06, "loss": 0.1234, "step": 331 }, { "epoch": 0.649706457925636, "grad_norm": 1.8440837288829663, "learning_rate": 7.614458285573262e-06, "loss": 0.1498, "step": 332 }, { "epoch": 0.6516634050880626, "grad_norm": 2.3936621815284234, "learning_rate": 7.60134472629517e-06, "loss": 0.1618, "step": 333 }, { "epoch": 0.6536203522504892, "grad_norm": 2.432433653648141, "learning_rate": 7.588206586245001e-06, "loss": 0.2118, "step": 334 }, { "epoch": 0.6555772994129159, "grad_norm": 1.9913048808597624, "learning_rate": 7.575043989568401e-06, "loss": 0.1613, "step": 335 }, { "epoch": 0.6575342465753424, "grad_norm": 2.14094099361202, "learning_rate": 7.56185706064212e-06, "loss": 0.1856, "step": 336 }, { "epoch": 0.6594911937377691, "grad_norm": 2.9942838659811586, "learning_rate": 7.548645924072816e-06, "loss": 0.1906, "step": 337 }, { "epoch": 0.6614481409001957, "grad_norm": 2.298959647034085, "learning_rate": 7.535410704695907e-06, "loss": 0.1825, "step": 338 }, { "epoch": 0.6634050880626223, "grad_norm": 2.288771608869839, "learning_rate": 7.522151527574362e-06, "loss": 0.2473, "step": 339 }, { "epoch": 0.6653620352250489, "grad_norm": 2.2989244548307286, "learning_rate": 7.508868517997544e-06, "loss": 0.1709, "step": 340 }, { "epoch": 0.6673189823874756, "grad_norm": 1.9322292514474384, "learning_rate": 7.495561801480009e-06, "loss": 0.181, "step": 341 }, { "epoch": 0.6692759295499021, "grad_norm": 2.2325287027619565, "learning_rate": 7.4822315037603245e-06, "loss": 0.1826, "step": 342 }, { "epoch": 0.6712328767123288, "grad_norm": 1.7235494101472766, "learning_rate": 7.468877750799887e-06, "loss": 0.1334, "step": 343 }, { "epoch": 0.6731898238747553, "grad_norm": 1.9879060555447008, "learning_rate": 7.455500668781725e-06, "loss": 0.1579, "step": 344 }, { "epoch": 0.675146771037182, "grad_norm": 2.006853139856201, "learning_rate": 7.442100384109313e-06, "loss": 0.1793, "step": 345 }, { "epoch": 0.6771037181996086, "grad_norm": 2.100084438391705, "learning_rate": 7.428677023405366e-06, "loss": 0.1736, "step": 346 }, { "epoch": 0.6790606653620352, "grad_norm": 1.6622756006593093, "learning_rate": 7.415230713510657e-06, "loss": 0.1392, "step": 347 }, { "epoch": 0.6810176125244618, "grad_norm": 1.8303639460522547, "learning_rate": 7.40176158148281e-06, "loss": 0.142, "step": 348 }, { "epoch": 0.6829745596868885, "grad_norm": 2.12110043511208, "learning_rate": 7.3882697545951e-06, "loss": 0.1582, "step": 349 }, { "epoch": 0.684931506849315, "grad_norm": 1.8225745109883198, "learning_rate": 7.374755360335253e-06, "loss": 0.134, "step": 350 }, { "epoch": 0.6868884540117417, "grad_norm": 1.9042037455795249, "learning_rate": 7.361218526404237e-06, "loss": 0.16, "step": 351 }, { "epoch": 0.6888454011741683, "grad_norm": 2.13766777776365, "learning_rate": 7.3476593807150606e-06, "loss": 0.1898, "step": 352 }, { "epoch": 0.6908023483365949, "grad_norm": 2.139504143861707, "learning_rate": 7.334078051391563e-06, "loss": 0.1485, "step": 353 }, { "epoch": 0.6927592954990215, "grad_norm": 1.6656865341188443, "learning_rate": 7.320474666767201e-06, "loss": 0.1267, "step": 354 }, { "epoch": 0.6947162426614482, "grad_norm": 2.283337414874692, "learning_rate": 7.3068493553838386e-06, "loss": 0.1599, "step": 355 }, { "epoch": 0.6966731898238747, "grad_norm": 2.1159618262030815, "learning_rate": 7.293202245990526e-06, "loss": 0.2157, "step": 356 }, { "epoch": 0.6986301369863014, "grad_norm": 2.222489024028487, "learning_rate": 7.279533467542295e-06, "loss": 0.1838, "step": 357 }, { "epoch": 0.700587084148728, "grad_norm": 2.162550509863416, "learning_rate": 7.265843149198931e-06, "loss": 0.1721, "step": 358 }, { "epoch": 0.7025440313111546, "grad_norm": 1.8530436181472092, "learning_rate": 7.252131420323757e-06, "loss": 0.1526, "step": 359 }, { "epoch": 0.7045009784735812, "grad_norm": 1.9098908416944502, "learning_rate": 7.238398410482408e-06, "loss": 0.1664, "step": 360 }, { "epoch": 0.7064579256360078, "grad_norm": 1.8116502984612795, "learning_rate": 7.224644249441606e-06, "loss": 0.1539, "step": 361 }, { "epoch": 0.7084148727984344, "grad_norm": 2.432105624659133, "learning_rate": 7.210869067167942e-06, "loss": 0.157, "step": 362 }, { "epoch": 0.7103718199608611, "grad_norm": 2.1806998736052994, "learning_rate": 7.197072993826637e-06, "loss": 0.1602, "step": 363 }, { "epoch": 0.7123287671232876, "grad_norm": 2.4627576998080833, "learning_rate": 7.183256159780321e-06, "loss": 0.2519, "step": 364 }, { "epoch": 0.7142857142857143, "grad_norm": 2.4308940613593606, "learning_rate": 7.169418695587791e-06, "loss": 0.2135, "step": 365 }, { "epoch": 0.7162426614481409, "grad_norm": 2.4941058101075537, "learning_rate": 7.155560732002792e-06, "loss": 0.1997, "step": 366 }, { "epoch": 0.7181996086105675, "grad_norm": 2.3165754573475503, "learning_rate": 7.141682399972765e-06, "loss": 0.1924, "step": 367 }, { "epoch": 0.7201565557729941, "grad_norm": 2.255287968047063, "learning_rate": 7.127783830637625e-06, "loss": 0.2021, "step": 368 }, { "epoch": 0.7221135029354208, "grad_norm": 2.183333205701938, "learning_rate": 7.113865155328506e-06, "loss": 0.1881, "step": 369 }, { "epoch": 0.7240704500978473, "grad_norm": 2.1733355043960323, "learning_rate": 7.099926505566537e-06, "loss": 0.1532, "step": 370 }, { "epoch": 0.726027397260274, "grad_norm": 2.2928508214387127, "learning_rate": 7.085968013061585e-06, "loss": 0.1817, "step": 371 }, { "epoch": 0.7279843444227005, "grad_norm": 1.9639249874086246, "learning_rate": 7.071989809711018e-06, "loss": 0.1597, "step": 372 }, { "epoch": 0.7299412915851272, "grad_norm": 2.54835903318302, "learning_rate": 7.057992027598459e-06, "loss": 0.2492, "step": 373 }, { "epoch": 0.7318982387475538, "grad_norm": 1.360830606365982, "learning_rate": 7.043974798992532e-06, "loss": 0.1118, "step": 374 }, { "epoch": 0.7338551859099804, "grad_norm": 2.2152003292813554, "learning_rate": 7.029938256345617e-06, "loss": 0.156, "step": 375 }, { "epoch": 0.735812133072407, "grad_norm": 1.9889617633501924, "learning_rate": 7.015882532292598e-06, "loss": 0.1796, "step": 376 }, { "epoch": 0.7377690802348337, "grad_norm": 2.382140854226086, "learning_rate": 7.001807759649609e-06, "loss": 0.1582, "step": 377 }, { "epoch": 0.7397260273972602, "grad_norm": 1.6738141310398935, "learning_rate": 6.987714071412781e-06, "loss": 0.1496, "step": 378 }, { "epoch": 0.7416829745596869, "grad_norm": 1.8244368571357694, "learning_rate": 6.973601600756978e-06, "loss": 0.1659, "step": 379 }, { "epoch": 0.7436399217221135, "grad_norm": 1.96186269940632, "learning_rate": 6.959470481034547e-06, "loss": 0.1756, "step": 380 }, { "epoch": 0.7455968688845401, "grad_norm": 1.7325900267696301, "learning_rate": 6.945320845774057e-06, "loss": 0.136, "step": 381 }, { "epoch": 0.7475538160469667, "grad_norm": 1.6190578652073147, "learning_rate": 6.931152828679033e-06, "loss": 0.1187, "step": 382 }, { "epoch": 0.7495107632093934, "grad_norm": 2.0592592705414225, "learning_rate": 6.9169665636266925e-06, "loss": 0.1824, "step": 383 }, { "epoch": 0.7514677103718199, "grad_norm": 2.300352366726294, "learning_rate": 6.902762184666687e-06, "loss": 0.2021, "step": 384 }, { "epoch": 0.7534246575342466, "grad_norm": 1.8611660470391644, "learning_rate": 6.888539826019824e-06, "loss": 0.1532, "step": 385 }, { "epoch": 0.7553816046966731, "grad_norm": 2.114915652767351, "learning_rate": 6.874299622076817e-06, "loss": 0.1871, "step": 386 }, { "epoch": 0.7573385518590998, "grad_norm": 2.0148515403923275, "learning_rate": 6.860041707396993e-06, "loss": 0.1688, "step": 387 }, { "epoch": 0.7592954990215264, "grad_norm": 1.9093322663721404, "learning_rate": 6.845766216707037e-06, "loss": 0.1427, "step": 388 }, { "epoch": 0.761252446183953, "grad_norm": 1.9322420099953308, "learning_rate": 6.831473284899715e-06, "loss": 0.1675, "step": 389 }, { "epoch": 0.7632093933463796, "grad_norm": 1.9251339147307147, "learning_rate": 6.817163047032598e-06, "loss": 0.1809, "step": 390 }, { "epoch": 0.7651663405088063, "grad_norm": 2.042342715597983, "learning_rate": 6.802835638326784e-06, "loss": 0.1736, "step": 391 }, { "epoch": 0.7671232876712328, "grad_norm": 2.230726896070188, "learning_rate": 6.788491194165629e-06, "loss": 0.1936, "step": 392 }, { "epoch": 0.7690802348336595, "grad_norm": 1.8954008980818953, "learning_rate": 6.774129850093451e-06, "loss": 0.1324, "step": 393 }, { "epoch": 0.7710371819960861, "grad_norm": 1.8425797897806795, "learning_rate": 6.759751741814271e-06, "loss": 0.1984, "step": 394 }, { "epoch": 0.7729941291585127, "grad_norm": 1.8294613615814772, "learning_rate": 6.74535700519051e-06, "loss": 0.1368, "step": 395 }, { "epoch": 0.7749510763209393, "grad_norm": 1.5556194999338477, "learning_rate": 6.7309457762417214e-06, "loss": 0.1291, "step": 396 }, { "epoch": 0.776908023483366, "grad_norm": 2.128242918340814, "learning_rate": 6.7165181911432955e-06, "loss": 0.1931, "step": 397 }, { "epoch": 0.7788649706457925, "grad_norm": 2.3977013494114714, "learning_rate": 6.702074386225175e-06, "loss": 0.179, "step": 398 }, { "epoch": 0.7808219178082192, "grad_norm": 2.0392202034505265, "learning_rate": 6.687614497970567e-06, "loss": 0.166, "step": 399 }, { "epoch": 0.7827788649706457, "grad_norm": 1.8061414751466955, "learning_rate": 6.67313866301466e-06, "loss": 0.1463, "step": 400 }, { "epoch": 0.7847358121330724, "grad_norm": 2.317606084731522, "learning_rate": 6.658647018143322e-06, "loss": 0.2322, "step": 401 }, { "epoch": 0.786692759295499, "grad_norm": 1.8223132173852006, "learning_rate": 6.6441397002918165e-06, "loss": 0.1622, "step": 402 }, { "epoch": 0.7886497064579256, "grad_norm": 1.7974723380887363, "learning_rate": 6.629616846543497e-06, "loss": 0.1829, "step": 403 }, { "epoch": 0.7906066536203522, "grad_norm": 1.8446451299456803, "learning_rate": 6.615078594128531e-06, "loss": 0.15, "step": 404 }, { "epoch": 0.7925636007827789, "grad_norm": 1.6119498478158396, "learning_rate": 6.600525080422583e-06, "loss": 0.1307, "step": 405 }, { "epoch": 0.7945205479452054, "grad_norm": 1.8111026513076607, "learning_rate": 6.585956442945531e-06, "loss": 0.176, "step": 406 }, { "epoch": 0.7964774951076321, "grad_norm": 2.3929646833655194, "learning_rate": 6.571372819360158e-06, "loss": 0.2013, "step": 407 }, { "epoch": 0.7984344422700587, "grad_norm": 2.0832257354011037, "learning_rate": 6.5567743474708545e-06, "loss": 0.1678, "step": 408 }, { "epoch": 0.8003913894324853, "grad_norm": 2.1169740601266325, "learning_rate": 6.5421611652223185e-06, "loss": 0.1706, "step": 409 }, { "epoch": 0.8023483365949119, "grad_norm": 1.571896350263203, "learning_rate": 6.52753341069825e-06, "loss": 0.12, "step": 410 }, { "epoch": 0.8043052837573386, "grad_norm": 1.9028475574087562, "learning_rate": 6.512891222120044e-06, "loss": 0.1667, "step": 411 }, { "epoch": 0.8062622309197651, "grad_norm": 1.8560679268204048, "learning_rate": 6.498234737845488e-06, "loss": 0.1569, "step": 412 }, { "epoch": 0.8082191780821918, "grad_norm": 1.807054424057428, "learning_rate": 6.483564096367452e-06, "loss": 0.1584, "step": 413 }, { "epoch": 0.8101761252446184, "grad_norm": 2.0645128884080277, "learning_rate": 6.468879436312584e-06, "loss": 0.1905, "step": 414 }, { "epoch": 0.812133072407045, "grad_norm": 2.052996263903671, "learning_rate": 6.454180896439993e-06, "loss": 0.1531, "step": 415 }, { "epoch": 0.8140900195694716, "grad_norm": 2.236874635757778, "learning_rate": 6.439468615639947e-06, "loss": 0.1633, "step": 416 }, { "epoch": 0.8160469667318982, "grad_norm": 2.553160265792283, "learning_rate": 6.424742732932547e-06, "loss": 0.1891, "step": 417 }, { "epoch": 0.8180039138943248, "grad_norm": 2.0732490472951333, "learning_rate": 6.410003387466433e-06, "loss": 0.1725, "step": 418 }, { "epoch": 0.8199608610567515, "grad_norm": 1.9843913839910987, "learning_rate": 6.3952507185174505e-06, "loss": 0.1835, "step": 419 }, { "epoch": 0.821917808219178, "grad_norm": 1.8409166702741069, "learning_rate": 6.380484865487346e-06, "loss": 0.1363, "step": 420 }, { "epoch": 0.8238747553816047, "grad_norm": 2.554343745058397, "learning_rate": 6.365705967902439e-06, "loss": 0.1641, "step": 421 }, { "epoch": 0.8258317025440313, "grad_norm": 2.259599788173401, "learning_rate": 6.35091416541232e-06, "loss": 0.2063, "step": 422 }, { "epoch": 0.8277886497064579, "grad_norm": 1.7920821685358985, "learning_rate": 6.3361095977885125e-06, "loss": 0.166, "step": 423 }, { "epoch": 0.8297455968688845, "grad_norm": 2.3380499115294207, "learning_rate": 6.3212924049231676e-06, "loss": 0.1926, "step": 424 }, { "epoch": 0.8317025440313112, "grad_norm": 2.0932589958181125, "learning_rate": 6.306462726827733e-06, "loss": 0.1823, "step": 425 }, { "epoch": 0.8336594911937377, "grad_norm": 2.0058886855372875, "learning_rate": 6.29162070363163e-06, "loss": 0.1445, "step": 426 }, { "epoch": 0.8356164383561644, "grad_norm": 2.025959355874046, "learning_rate": 6.276766475580935e-06, "loss": 0.1972, "step": 427 }, { "epoch": 0.837573385518591, "grad_norm": 1.6955016858034482, "learning_rate": 6.261900183037053e-06, "loss": 0.1838, "step": 428 }, { "epoch": 0.8395303326810176, "grad_norm": 1.9921670163234422, "learning_rate": 6.247021966475385e-06, "loss": 0.1815, "step": 429 }, { "epoch": 0.8414872798434442, "grad_norm": 1.7798648486521123, "learning_rate": 6.232131966484007e-06, "loss": 0.1144, "step": 430 }, { "epoch": 0.8434442270058709, "grad_norm": 2.1146637818651777, "learning_rate": 6.217230323762338e-06, "loss": 0.1858, "step": 431 }, { "epoch": 0.8454011741682974, "grad_norm": 1.8394527448850762, "learning_rate": 6.202317179119817e-06, "loss": 0.1429, "step": 432 }, { "epoch": 0.8473581213307241, "grad_norm": 1.9374318336428438, "learning_rate": 6.187392673474562e-06, "loss": 0.1693, "step": 433 }, { "epoch": 0.8493150684931506, "grad_norm": 1.82052906702598, "learning_rate": 6.1724569478520495e-06, "loss": 0.1496, "step": 434 }, { "epoch": 0.8512720156555773, "grad_norm": 1.96532391288847, "learning_rate": 6.15751014338377e-06, "loss": 0.1764, "step": 435 }, { "epoch": 0.8532289628180039, "grad_norm": 2.3874055736380804, "learning_rate": 6.142552401305907e-06, "loss": 0.172, "step": 436 }, { "epoch": 0.8551859099804305, "grad_norm": 2.1759096285986326, "learning_rate": 6.1275838629579906e-06, "loss": 0.2244, "step": 437 }, { "epoch": 0.8571428571428571, "grad_norm": 2.44362009290459, "learning_rate": 6.112604669781572e-06, "loss": 0.1836, "step": 438 }, { "epoch": 0.8590998043052838, "grad_norm": 1.880105292315407, "learning_rate": 6.097614963318883e-06, "loss": 0.1705, "step": 439 }, { "epoch": 0.8610567514677103, "grad_norm": 1.9656146308633662, "learning_rate": 6.0826148852114945e-06, "loss": 0.1719, "step": 440 }, { "epoch": 0.863013698630137, "grad_norm": 2.363795853721546, "learning_rate": 6.067604577198981e-06, "loss": 0.1929, "step": 441 }, { "epoch": 0.8649706457925636, "grad_norm": 1.988332433107494, "learning_rate": 6.052584181117589e-06, "loss": 0.1695, "step": 442 }, { "epoch": 0.8669275929549902, "grad_norm": 1.933775090584915, "learning_rate": 6.037553838898882e-06, "loss": 0.1656, "step": 443 }, { "epoch": 0.8688845401174168, "grad_norm": 1.909721965580417, "learning_rate": 6.022513692568412e-06, "loss": 0.1361, "step": 444 }, { "epoch": 0.8708414872798435, "grad_norm": 1.733936870600194, "learning_rate": 6.00746388424437e-06, "loss": 0.1588, "step": 445 }, { "epoch": 0.87279843444227, "grad_norm": 1.9929832970143146, "learning_rate": 5.992404556136247e-06, "loss": 0.1809, "step": 446 }, { "epoch": 0.8747553816046967, "grad_norm": 2.9735340597945132, "learning_rate": 5.977335850543488e-06, "loss": 0.2255, "step": 447 }, { "epoch": 0.8767123287671232, "grad_norm": 2.0804510634037765, "learning_rate": 5.96225790985415e-06, "loss": 0.1596, "step": 448 }, { "epoch": 0.8786692759295499, "grad_norm": 2.50133422344921, "learning_rate": 5.947170876543553e-06, "loss": 0.1638, "step": 449 }, { "epoch": 0.8806262230919765, "grad_norm": 2.1540486319860483, "learning_rate": 5.932074893172934e-06, "loss": 0.2044, "step": 450 }, { "epoch": 0.8825831702544031, "grad_norm": 1.7027589371308625, "learning_rate": 5.916970102388104e-06, "loss": 0.1562, "step": 451 }, { "epoch": 0.8845401174168297, "grad_norm": 2.052759478757739, "learning_rate": 5.9018566469181e-06, "loss": 0.1521, "step": 452 }, { "epoch": 0.8864970645792564, "grad_norm": 2.1812410825484188, "learning_rate": 5.886734669573827e-06, "loss": 0.1964, "step": 453 }, { "epoch": 0.8884540117416829, "grad_norm": 2.687400352021544, "learning_rate": 5.87160431324672e-06, "loss": 0.1744, "step": 454 }, { "epoch": 0.8904109589041096, "grad_norm": 2.331754019498941, "learning_rate": 5.856465720907388e-06, "loss": 0.1786, "step": 455 }, { "epoch": 0.8923679060665362, "grad_norm": 1.8228584834511061, "learning_rate": 5.841319035604267e-06, "loss": 0.172, "step": 456 }, { "epoch": 0.8943248532289628, "grad_norm": 2.0816990702254086, "learning_rate": 5.826164400462259e-06, "loss": 0.1813, "step": 457 }, { "epoch": 0.8962818003913894, "grad_norm": 2.365191904002113, "learning_rate": 5.8110019586813944e-06, "loss": 0.1741, "step": 458 }, { "epoch": 0.898238747553816, "grad_norm": 1.9778576234600616, "learning_rate": 5.795831853535461e-06, "loss": 0.1724, "step": 459 }, { "epoch": 0.9001956947162426, "grad_norm": 2.6636549752410525, "learning_rate": 5.780654228370669e-06, "loss": 0.2285, "step": 460 }, { "epoch": 0.9021526418786693, "grad_norm": 2.2656141746660046, "learning_rate": 5.76546922660428e-06, "loss": 0.1726, "step": 461 }, { "epoch": 0.9041095890410958, "grad_norm": 2.2658339295689442, "learning_rate": 5.7502769917232635e-06, "loss": 0.178, "step": 462 }, { "epoch": 0.9060665362035225, "grad_norm": 2.3811367605307496, "learning_rate": 5.735077667282935e-06, "loss": 0.1818, "step": 463 }, { "epoch": 0.9080234833659491, "grad_norm": 2.123263727226605, "learning_rate": 5.719871396905603e-06, "loss": 0.1996, "step": 464 }, { "epoch": 0.9099804305283757, "grad_norm": 1.6991092883194507, "learning_rate": 5.704658324279203e-06, "loss": 0.1775, "step": 465 }, { "epoch": 0.9119373776908023, "grad_norm": 2.025838725680335, "learning_rate": 5.6894385931559555e-06, "loss": 0.1522, "step": 466 }, { "epoch": 0.913894324853229, "grad_norm": 1.6034025410862527, "learning_rate": 5.674212347350996e-06, "loss": 0.1456, "step": 467 }, { "epoch": 0.9158512720156555, "grad_norm": 1.7227609498181402, "learning_rate": 5.658979730741014e-06, "loss": 0.1359, "step": 468 }, { "epoch": 0.9178082191780822, "grad_norm": 1.9835709060856443, "learning_rate": 5.643740887262905e-06, "loss": 0.171, "step": 469 }, { "epoch": 0.9197651663405088, "grad_norm": 1.6297846067607547, "learning_rate": 5.6284959609124e-06, "loss": 0.127, "step": 470 }, { "epoch": 0.9217221135029354, "grad_norm": 2.2367255512345716, "learning_rate": 5.613245095742711e-06, "loss": 0.2276, "step": 471 }, { "epoch": 0.923679060665362, "grad_norm": 2.0682546713193894, "learning_rate": 5.5979884358631665e-06, "loss": 0.2244, "step": 472 }, { "epoch": 0.9256360078277887, "grad_norm": 1.9064727254028155, "learning_rate": 5.582726125437846e-06, "loss": 0.1571, "step": 473 }, { "epoch": 0.9275929549902152, "grad_norm": 2.0440422055077563, "learning_rate": 5.567458308684233e-06, "loss": 0.1634, "step": 474 }, { "epoch": 0.9295499021526419, "grad_norm": 2.0289471422356007, "learning_rate": 5.5521851298718295e-06, "loss": 0.1789, "step": 475 }, { "epoch": 0.9315068493150684, "grad_norm": 2.273683967116904, "learning_rate": 5.536906733320816e-06, "loss": 0.1581, "step": 476 }, { "epoch": 0.9334637964774951, "grad_norm": 2.5706206578734823, "learning_rate": 5.5216232634006695e-06, "loss": 0.228, "step": 477 }, { "epoch": 0.9354207436399217, "grad_norm": 2.118636087512207, "learning_rate": 5.506334864528808e-06, "loss": 0.1954, "step": 478 }, { "epoch": 0.9373776908023483, "grad_norm": 1.9582340027542617, "learning_rate": 5.491041681169226e-06, "loss": 0.1523, "step": 479 }, { "epoch": 0.9393346379647749, "grad_norm": 1.945697540082699, "learning_rate": 5.4757438578311275e-06, "loss": 0.1937, "step": 480 }, { "epoch": 0.9412915851272016, "grad_norm": 2.262248835249883, "learning_rate": 5.460441539067559e-06, "loss": 0.2039, "step": 481 }, { "epoch": 0.9432485322896281, "grad_norm": 2.222534309983352, "learning_rate": 5.4451348694740495e-06, "loss": 0.1538, "step": 482 }, { "epoch": 0.9452054794520548, "grad_norm": 1.9999160038314057, "learning_rate": 5.429823993687234e-06, "loss": 0.1495, "step": 483 }, { "epoch": 0.9471624266144814, "grad_norm": 1.6354324035199221, "learning_rate": 5.414509056383498e-06, "loss": 0.1309, "step": 484 }, { "epoch": 0.949119373776908, "grad_norm": 1.6615194088601373, "learning_rate": 5.399190202277602e-06, "loss": 0.1419, "step": 485 }, { "epoch": 0.9510763209393346, "grad_norm": 2.286650061764155, "learning_rate": 5.383867576121324e-06, "loss": 0.1648, "step": 486 }, { "epoch": 0.9530332681017613, "grad_norm": 1.7539968417523317, "learning_rate": 5.368541322702074e-06, "loss": 0.1447, "step": 487 }, { "epoch": 0.9549902152641878, "grad_norm": 2.34120310133991, "learning_rate": 5.353211586841547e-06, "loss": 0.1535, "step": 488 }, { "epoch": 0.9569471624266145, "grad_norm": 2.0881045243146605, "learning_rate": 5.337878513394337e-06, "loss": 0.1196, "step": 489 }, { "epoch": 0.958904109589041, "grad_norm": 2.0265714888890494, "learning_rate": 5.322542247246583e-06, "loss": 0.1812, "step": 490 }, { "epoch": 0.9608610567514677, "grad_norm": 2.3421981871226136, "learning_rate": 5.307202933314586e-06, "loss": 0.1694, "step": 491 }, { "epoch": 0.9628180039138943, "grad_norm": 2.3568288617617568, "learning_rate": 5.29186071654345e-06, "loss": 0.1915, "step": 492 }, { "epoch": 0.9647749510763209, "grad_norm": 1.91597048069046, "learning_rate": 5.276515741905708e-06, "loss": 0.1675, "step": 493 }, { "epoch": 0.9667318982387475, "grad_norm": 2.110318250374213, "learning_rate": 5.261168154399953e-06, "loss": 0.1875, "step": 494 }, { "epoch": 0.9686888454011742, "grad_norm": 2.115820868740146, "learning_rate": 5.245818099049467e-06, "loss": 0.1752, "step": 495 }, { "epoch": 0.9706457925636007, "grad_norm": 1.9254162461019477, "learning_rate": 5.23046572090085e-06, "loss": 0.1722, "step": 496 }, { "epoch": 0.9726027397260274, "grad_norm": 2.045714268881818, "learning_rate": 5.215111165022653e-06, "loss": 0.1993, "step": 497 }, { "epoch": 0.974559686888454, "grad_norm": 1.8112233122675736, "learning_rate": 5.199754576504006e-06, "loss": 0.14, "step": 498 }, { "epoch": 0.9765166340508806, "grad_norm": 2.2897903323660573, "learning_rate": 5.184396100453242e-06, "loss": 0.2058, "step": 499 }, { "epoch": 0.9784735812133072, "grad_norm": 2.008131743208414, "learning_rate": 5.1690358819965334e-06, "loss": 0.1586, "step": 500 }, { "epoch": 0.9784735812133072, "eval_loss": 0.16941680014133453, "eval_runtime": 2.5973, "eval_samples_per_second": 16.171, "eval_steps_per_second": 4.235, "step": 500 }, { "epoch": 0.9804305283757339, "grad_norm": 1.7834382209293849, "learning_rate": 5.153674066276513e-06, "loss": 0.1531, "step": 501 }, { "epoch": 0.9823874755381604, "grad_norm": 2.0556344295729874, "learning_rate": 5.138310798450912e-06, "loss": 0.1802, "step": 502 }, { "epoch": 0.9843444227005871, "grad_norm": 2.1807149018564425, "learning_rate": 5.122946223691177e-06, "loss": 0.1977, "step": 503 }, { "epoch": 0.9863013698630136, "grad_norm": 1.8311516957056617, "learning_rate": 5.107580487181112e-06, "loss": 0.1179, "step": 504 }, { "epoch": 0.9882583170254403, "grad_norm": 2.0243269327329148, "learning_rate": 5.092213734115489e-06, "loss": 0.1485, "step": 505 }, { "epoch": 0.9902152641878669, "grad_norm": 1.8946858729214702, "learning_rate": 5.0768461096986935e-06, "loss": 0.1569, "step": 506 }, { "epoch": 0.9921722113502935, "grad_norm": 2.100319035915609, "learning_rate": 5.0614777591433375e-06, "loss": 0.1585, "step": 507 }, { "epoch": 0.9941291585127201, "grad_norm": 2.190267401787019, "learning_rate": 5.0461088276689026e-06, "loss": 0.1768, "step": 508 }, { "epoch": 0.9960861056751468, "grad_norm": 1.7148427898749872, "learning_rate": 5.0307394605003554e-06, "loss": 0.1586, "step": 509 }, { "epoch": 0.9980430528375733, "grad_norm": 1.599383828166455, "learning_rate": 5.01536980286678e-06, "loss": 0.1038, "step": 510 }, { "epoch": 1.0, "grad_norm": 1.9163254835986636, "learning_rate": 5e-06, "loss": 0.1508, "step": 511 }, { "epoch": 1.0019569471624266, "grad_norm": 1.5882345744225046, "learning_rate": 4.984630197133224e-06, "loss": 0.089, "step": 512 }, { "epoch": 1.0039138943248531, "grad_norm": 1.7217558962772432, "learning_rate": 4.969260539499645e-06, "loss": 0.1032, "step": 513 }, { "epoch": 1.00587084148728, "grad_norm": 1.7198923302987532, "learning_rate": 4.953891172331098e-06, "loss": 0.0863, "step": 514 }, { "epoch": 1.0078277886497065, "grad_norm": 1.4293336061434976, "learning_rate": 4.938522240856664e-06, "loss": 0.0808, "step": 515 }, { "epoch": 1.009784735812133, "grad_norm": 1.3991050387789723, "learning_rate": 4.92315389030131e-06, "loss": 0.0781, "step": 516 }, { "epoch": 1.0117416829745598, "grad_norm": 1.2869008028844684, "learning_rate": 4.907786265884511e-06, "loss": 0.073, "step": 517 }, { "epoch": 1.0136986301369864, "grad_norm": 1.5857851411264983, "learning_rate": 4.89241951281889e-06, "loss": 0.0742, "step": 518 }, { "epoch": 1.015655577299413, "grad_norm": 1.4149723792900126, "learning_rate": 4.8770537763088234e-06, "loss": 0.0619, "step": 519 }, { "epoch": 1.0176125244618395, "grad_norm": 1.0116676906891107, "learning_rate": 4.8616892015490905e-06, "loss": 0.0604, "step": 520 }, { "epoch": 1.019569471624266, "grad_norm": 1.4240287571753596, "learning_rate": 4.846325933723487e-06, "loss": 0.0738, "step": 521 }, { "epoch": 1.0215264187866928, "grad_norm": 1.9143863101173124, "learning_rate": 4.830964118003468e-06, "loss": 0.0812, "step": 522 }, { "epoch": 1.0234833659491194, "grad_norm": 1.6898900478412748, "learning_rate": 4.8156038995467606e-06, "loss": 0.0869, "step": 523 }, { "epoch": 1.025440313111546, "grad_norm": 1.3970003455157622, "learning_rate": 4.800245423495997e-06, "loss": 0.0868, "step": 524 }, { "epoch": 1.0273972602739727, "grad_norm": 1.4715449668862843, "learning_rate": 4.784888834977347e-06, "loss": 0.0745, "step": 525 }, { "epoch": 1.0293542074363993, "grad_norm": 1.664073023766418, "learning_rate": 4.769534279099151e-06, "loss": 0.0731, "step": 526 }, { "epoch": 1.0313111545988258, "grad_norm": 1.8718920775573955, "learning_rate": 4.7541819009505354e-06, "loss": 0.0604, "step": 527 }, { "epoch": 1.0332681017612524, "grad_norm": 1.591478729689564, "learning_rate": 4.73883184560005e-06, "loss": 0.0858, "step": 528 }, { "epoch": 1.035225048923679, "grad_norm": 1.45991548111859, "learning_rate": 4.723484258094293e-06, "loss": 0.0513, "step": 529 }, { "epoch": 1.0371819960861057, "grad_norm": 2.0044101391802855, "learning_rate": 4.708139283456551e-06, "loss": 0.0599, "step": 530 }, { "epoch": 1.0391389432485323, "grad_norm": 1.7922616526792323, "learning_rate": 4.692797066685415e-06, "loss": 0.0785, "step": 531 }, { "epoch": 1.0410958904109588, "grad_norm": 1.7620370923796478, "learning_rate": 4.6774577527534195e-06, "loss": 0.0628, "step": 532 }, { "epoch": 1.0430528375733856, "grad_norm": 1.6974182275432634, "learning_rate": 4.662121486605663e-06, "loss": 0.064, "step": 533 }, { "epoch": 1.0450097847358122, "grad_norm": 1.7578637671544346, "learning_rate": 4.646788413158455e-06, "loss": 0.0591, "step": 534 }, { "epoch": 1.0469667318982387, "grad_norm": 1.7568430452557353, "learning_rate": 4.631458677297927e-06, "loss": 0.0653, "step": 535 }, { "epoch": 1.0489236790606653, "grad_norm": 1.9577065194675067, "learning_rate": 4.616132423878679e-06, "loss": 0.0661, "step": 536 }, { "epoch": 1.0508806262230919, "grad_norm": 1.3896703161353012, "learning_rate": 4.600809797722398e-06, "loss": 0.0528, "step": 537 }, { "epoch": 1.0528375733855186, "grad_norm": 1.7239414674761873, "learning_rate": 4.585490943616504e-06, "loss": 0.0458, "step": 538 }, { "epoch": 1.0547945205479452, "grad_norm": 1.8551419636007658, "learning_rate": 4.570176006312769e-06, "loss": 0.0707, "step": 539 }, { "epoch": 1.0567514677103718, "grad_norm": 2.3192434227420446, "learning_rate": 4.554865130525953e-06, "loss": 0.0713, "step": 540 }, { "epoch": 1.0587084148727985, "grad_norm": 2.3023058824121754, "learning_rate": 4.539558460932442e-06, "loss": 0.0768, "step": 541 }, { "epoch": 1.060665362035225, "grad_norm": 1.8575157597443823, "learning_rate": 4.524256142168874e-06, "loss": 0.0782, "step": 542 }, { "epoch": 1.0626223091976517, "grad_norm": 1.8021993823354285, "learning_rate": 4.508958318830776e-06, "loss": 0.0765, "step": 543 }, { "epoch": 1.0645792563600782, "grad_norm": 2.428542241283769, "learning_rate": 4.493665135471194e-06, "loss": 0.0897, "step": 544 }, { "epoch": 1.0665362035225048, "grad_norm": 1.7505925113138043, "learning_rate": 4.478376736599331e-06, "loss": 0.0551, "step": 545 }, { "epoch": 1.0684931506849316, "grad_norm": 2.0268330656897406, "learning_rate": 4.463093266679185e-06, "loss": 0.0653, "step": 546 }, { "epoch": 1.0704500978473581, "grad_norm": 1.9818870151799146, "learning_rate": 4.447814870128172e-06, "loss": 0.0788, "step": 547 }, { "epoch": 1.0724070450097847, "grad_norm": 1.952229513568739, "learning_rate": 4.4325416913157706e-06, "loss": 0.0619, "step": 548 }, { "epoch": 1.0743639921722115, "grad_norm": 1.9025622142698373, "learning_rate": 4.417273874562154e-06, "loss": 0.0615, "step": 549 }, { "epoch": 1.076320939334638, "grad_norm": 1.7105619420125273, "learning_rate": 4.402011564136835e-06, "loss": 0.0656, "step": 550 }, { "epoch": 1.0782778864970646, "grad_norm": 2.1311693983375584, "learning_rate": 4.386754904257291e-06, "loss": 0.0618, "step": 551 }, { "epoch": 1.0802348336594911, "grad_norm": 2.084110009569339, "learning_rate": 4.371504039087602e-06, "loss": 0.0782, "step": 552 }, { "epoch": 1.0821917808219177, "grad_norm": 1.7349573673193466, "learning_rate": 4.356259112737096e-06, "loss": 0.071, "step": 553 }, { "epoch": 1.0841487279843445, "grad_norm": 1.6155783609911654, "learning_rate": 4.341020269258987e-06, "loss": 0.04, "step": 554 }, { "epoch": 1.086105675146771, "grad_norm": 1.5428414649454543, "learning_rate": 4.325787652649006e-06, "loss": 0.0582, "step": 555 }, { "epoch": 1.0880626223091976, "grad_norm": 2.21722256615053, "learning_rate": 4.310561406844045e-06, "loss": 0.0719, "step": 556 }, { "epoch": 1.0900195694716244, "grad_norm": 2.3109992701434496, "learning_rate": 4.295341675720797e-06, "loss": 0.0685, "step": 557 }, { "epoch": 1.091976516634051, "grad_norm": 1.5764890564698777, "learning_rate": 4.280128603094399e-06, "loss": 0.0576, "step": 558 }, { "epoch": 1.0939334637964775, "grad_norm": 1.967194986657686, "learning_rate": 4.264922332717066e-06, "loss": 0.0845, "step": 559 }, { "epoch": 1.095890410958904, "grad_norm": 1.377888518887198, "learning_rate": 4.249723008276737e-06, "loss": 0.0446, "step": 560 }, { "epoch": 1.0978473581213306, "grad_norm": 1.5878498921811268, "learning_rate": 4.234530773395721e-06, "loss": 0.06, "step": 561 }, { "epoch": 1.0998043052837574, "grad_norm": 1.5360435582896006, "learning_rate": 4.219345771629333e-06, "loss": 0.0616, "step": 562 }, { "epoch": 1.101761252446184, "grad_norm": 1.630344464811579, "learning_rate": 4.20416814646454e-06, "loss": 0.0663, "step": 563 }, { "epoch": 1.1037181996086105, "grad_norm": 2.5141146740126543, "learning_rate": 4.188998041318608e-06, "loss": 0.0922, "step": 564 }, { "epoch": 1.1056751467710373, "grad_norm": 1.5009369526245524, "learning_rate": 4.173835599537741e-06, "loss": 0.0687, "step": 565 }, { "epoch": 1.1076320939334638, "grad_norm": 1.6915265232548404, "learning_rate": 4.158680964395734e-06, "loss": 0.0493, "step": 566 }, { "epoch": 1.1095890410958904, "grad_norm": 2.1161604268890315, "learning_rate": 4.143534279092613e-06, "loss": 0.0907, "step": 567 }, { "epoch": 1.111545988258317, "grad_norm": 1.9087817209408955, "learning_rate": 4.1283956867532825e-06, "loss": 0.0591, "step": 568 }, { "epoch": 1.1135029354207435, "grad_norm": 1.2705335669203723, "learning_rate": 4.113265330426175e-06, "loss": 0.0507, "step": 569 }, { "epoch": 1.1154598825831703, "grad_norm": 1.9033053509020057, "learning_rate": 4.098143353081902e-06, "loss": 0.0646, "step": 570 }, { "epoch": 1.1174168297455969, "grad_norm": 2.522451917975451, "learning_rate": 4.083029897611897e-06, "loss": 0.0815, "step": 571 }, { "epoch": 1.1193737769080234, "grad_norm": 2.0594728654106738, "learning_rate": 4.067925106827068e-06, "loss": 0.0668, "step": 572 }, { "epoch": 1.1213307240704502, "grad_norm": 1.6009192024750505, "learning_rate": 4.052829123456448e-06, "loss": 0.058, "step": 573 }, { "epoch": 1.1232876712328768, "grad_norm": 1.7255815330623046, "learning_rate": 4.037742090145851e-06, "loss": 0.0558, "step": 574 }, { "epoch": 1.1252446183953033, "grad_norm": 2.144299027812674, "learning_rate": 4.0226641494565125e-06, "loss": 0.0917, "step": 575 }, { "epoch": 1.1272015655577299, "grad_norm": 1.848319443838593, "learning_rate": 4.007595443863755e-06, "loss": 0.0734, "step": 576 }, { "epoch": 1.1291585127201564, "grad_norm": 1.9473759226487113, "learning_rate": 3.992536115755631e-06, "loss": 0.0785, "step": 577 }, { "epoch": 1.1311154598825832, "grad_norm": 1.5498981785754624, "learning_rate": 3.977486307431589e-06, "loss": 0.0392, "step": 578 }, { "epoch": 1.1330724070450098, "grad_norm": 1.7650662799519117, "learning_rate": 3.962446161101119e-06, "loss": 0.0766, "step": 579 }, { "epoch": 1.1350293542074363, "grad_norm": 1.986318385092273, "learning_rate": 3.9474158188824145e-06, "loss": 0.0779, "step": 580 }, { "epoch": 1.1369863013698631, "grad_norm": 1.2411504862242564, "learning_rate": 3.93239542280102e-06, "loss": 0.0566, "step": 581 }, { "epoch": 1.1389432485322897, "grad_norm": 1.854248640517382, "learning_rate": 3.917385114788508e-06, "loss": 0.0715, "step": 582 }, { "epoch": 1.1409001956947162, "grad_norm": 1.7558676166454183, "learning_rate": 3.902385036681118e-06, "loss": 0.0469, "step": 583 }, { "epoch": 1.1428571428571428, "grad_norm": 1.7167329160091431, "learning_rate": 3.887395330218429e-06, "loss": 0.0571, "step": 584 }, { "epoch": 1.1448140900195694, "grad_norm": 1.6912857380991848, "learning_rate": 3.87241613704201e-06, "loss": 0.0455, "step": 585 }, { "epoch": 1.1467710371819961, "grad_norm": 1.7187971760722889, "learning_rate": 3.857447598694095e-06, "loss": 0.0655, "step": 586 }, { "epoch": 1.1487279843444227, "grad_norm": 1.6943702752611791, "learning_rate": 3.842489856616231e-06, "loss": 0.0614, "step": 587 }, { "epoch": 1.1506849315068493, "grad_norm": 1.7317372746537878, "learning_rate": 3.827543052147952e-06, "loss": 0.06, "step": 588 }, { "epoch": 1.152641878669276, "grad_norm": 2.3086948355229784, "learning_rate": 3.8126073265254377e-06, "loss": 0.0821, "step": 589 }, { "epoch": 1.1545988258317026, "grad_norm": 1.846473948721207, "learning_rate": 3.797682820880184e-06, "loss": 0.0871, "step": 590 }, { "epoch": 1.1565557729941291, "grad_norm": 1.8261049668348723, "learning_rate": 3.782769676237663e-06, "loss": 0.0579, "step": 591 }, { "epoch": 1.1585127201565557, "grad_norm": 1.5193382322991347, "learning_rate": 3.7678680335159955e-06, "loss": 0.0606, "step": 592 }, { "epoch": 1.1604696673189823, "grad_norm": 1.8743428851581523, "learning_rate": 3.7529780335246164e-06, "loss": 0.0839, "step": 593 }, { "epoch": 1.162426614481409, "grad_norm": 1.5640950376439107, "learning_rate": 3.7380998169629477e-06, "loss": 0.0773, "step": 594 }, { "epoch": 1.1643835616438356, "grad_norm": 1.9630526700340132, "learning_rate": 3.7232335244190656e-06, "loss": 0.06, "step": 595 }, { "epoch": 1.1663405088062622, "grad_norm": 1.7141549376207557, "learning_rate": 3.708379296368372e-06, "loss": 0.0627, "step": 596 }, { "epoch": 1.168297455968689, "grad_norm": 1.9077983611309683, "learning_rate": 3.6935372731722686e-06, "loss": 0.0749, "step": 597 }, { "epoch": 1.1702544031311155, "grad_norm": 2.2086661792895526, "learning_rate": 3.6787075950768337e-06, "loss": 0.0992, "step": 598 }, { "epoch": 1.172211350293542, "grad_norm": 1.9531471437720231, "learning_rate": 3.663890402211489e-06, "loss": 0.084, "step": 599 }, { "epoch": 1.1741682974559686, "grad_norm": 2.1310426647026177, "learning_rate": 3.649085834587683e-06, "loss": 0.1117, "step": 600 }, { "epoch": 1.1761252446183952, "grad_norm": 1.978838452832595, "learning_rate": 3.6342940320975613e-06, "loss": 0.0671, "step": 601 }, { "epoch": 1.178082191780822, "grad_norm": 1.5835559069538645, "learning_rate": 3.6195151345126556e-06, "loss": 0.0552, "step": 602 }, { "epoch": 1.1800391389432485, "grad_norm": 1.8077109597809675, "learning_rate": 3.6047492814825508e-06, "loss": 0.0594, "step": 603 }, { "epoch": 1.181996086105675, "grad_norm": 1.981881660619778, "learning_rate": 3.5899966125335684e-06, "loss": 0.0798, "step": 604 }, { "epoch": 1.1839530332681019, "grad_norm": 1.4971249462062561, "learning_rate": 3.575257267067453e-06, "loss": 0.0507, "step": 605 }, { "epoch": 1.1859099804305284, "grad_norm": 1.5620324161789834, "learning_rate": 3.5605313843600555e-06, "loss": 0.0598, "step": 606 }, { "epoch": 1.187866927592955, "grad_norm": 1.4321758392064026, "learning_rate": 3.5458191035600077e-06, "loss": 0.0452, "step": 607 }, { "epoch": 1.1898238747553815, "grad_norm": 1.9036851568096298, "learning_rate": 3.531120563687419e-06, "loss": 0.0644, "step": 608 }, { "epoch": 1.191780821917808, "grad_norm": 1.804067796662997, "learning_rate": 3.5164359036325483e-06, "loss": 0.079, "step": 609 }, { "epoch": 1.1937377690802349, "grad_norm": 1.4048160175547275, "learning_rate": 3.5017652621545133e-06, "loss": 0.0409, "step": 610 }, { "epoch": 1.1956947162426614, "grad_norm": 2.114479892956619, "learning_rate": 3.487108777879957e-06, "loss": 0.058, "step": 611 }, { "epoch": 1.197651663405088, "grad_norm": 1.285301321987489, "learning_rate": 3.4724665893017517e-06, "loss": 0.0478, "step": 612 }, { "epoch": 1.1996086105675148, "grad_norm": 1.689956654738057, "learning_rate": 3.457838834777682e-06, "loss": 0.0529, "step": 613 }, { "epoch": 1.2015655577299413, "grad_norm": 1.8876175214063495, "learning_rate": 3.4432256525291468e-06, "loss": 0.0672, "step": 614 }, { "epoch": 1.203522504892368, "grad_norm": 3.0195829231012907, "learning_rate": 3.428627180639844e-06, "loss": 0.0509, "step": 615 }, { "epoch": 1.2054794520547945, "grad_norm": 4.575100817985506, "learning_rate": 3.4140435570544708e-06, "loss": 0.1073, "step": 616 }, { "epoch": 1.207436399217221, "grad_norm": 2.588696753094895, "learning_rate": 3.3994749195774173e-06, "loss": 0.076, "step": 617 }, { "epoch": 1.2093933463796478, "grad_norm": 2.0873697764271197, "learning_rate": 3.3849214058714707e-06, "loss": 0.0522, "step": 618 }, { "epoch": 1.2113502935420744, "grad_norm": 1.8975782500088405, "learning_rate": 3.370383153456504e-06, "loss": 0.0586, "step": 619 }, { "epoch": 1.213307240704501, "grad_norm": 1.6837440803801198, "learning_rate": 3.355860299708187e-06, "loss": 0.0603, "step": 620 }, { "epoch": 1.2152641878669277, "grad_norm": 1.563152180016155, "learning_rate": 3.341352981856678e-06, "loss": 0.0546, "step": 621 }, { "epoch": 1.2172211350293543, "grad_norm": 1.6550538676265394, "learning_rate": 3.326861336985341e-06, "loss": 0.0472, "step": 622 }, { "epoch": 1.2191780821917808, "grad_norm": 1.4994991023696322, "learning_rate": 3.3123855020294344e-06, "loss": 0.049, "step": 623 }, { "epoch": 1.2211350293542074, "grad_norm": 1.9631354045255305, "learning_rate": 3.2979256137748283e-06, "loss": 0.0712, "step": 624 }, { "epoch": 1.223091976516634, "grad_norm": 2.064358822180799, "learning_rate": 3.283481808856706e-06, "loss": 0.073, "step": 625 }, { "epoch": 1.2250489236790607, "grad_norm": 2.0612184142213192, "learning_rate": 3.269054223758279e-06, "loss": 0.0838, "step": 626 }, { "epoch": 1.2270058708414873, "grad_norm": 1.7801925788931876, "learning_rate": 3.2546429948094904e-06, "loss": 0.0633, "step": 627 }, { "epoch": 1.2289628180039138, "grad_norm": 1.97752681595341, "learning_rate": 3.240248258185731e-06, "loss": 0.0582, "step": 628 }, { "epoch": 1.2309197651663406, "grad_norm": 1.744781443738379, "learning_rate": 3.225870149906549e-06, "loss": 0.0661, "step": 629 }, { "epoch": 1.2328767123287672, "grad_norm": 2.109438878664079, "learning_rate": 3.2115088058343725e-06, "loss": 0.0746, "step": 630 }, { "epoch": 1.2348336594911937, "grad_norm": 1.8032004643640032, "learning_rate": 3.1971643616732174e-06, "loss": 0.0574, "step": 631 }, { "epoch": 1.2367906066536203, "grad_norm": 1.9666924035620796, "learning_rate": 3.182836952967405e-06, "loss": 0.0669, "step": 632 }, { "epoch": 1.2387475538160468, "grad_norm": 1.7296588590203812, "learning_rate": 3.1685267151002856e-06, "loss": 0.0537, "step": 633 }, { "epoch": 1.2407045009784736, "grad_norm": 1.1948538834994598, "learning_rate": 3.154233783292964e-06, "loss": 0.0633, "step": 634 }, { "epoch": 1.2426614481409002, "grad_norm": 1.6108955009153085, "learning_rate": 3.139958292603009e-06, "loss": 0.0542, "step": 635 }, { "epoch": 1.2446183953033267, "grad_norm": 1.9903797262827012, "learning_rate": 3.125700377923186e-06, "loss": 0.0718, "step": 636 }, { "epoch": 1.2465753424657535, "grad_norm": 1.5910737310395764, "learning_rate": 3.111460173980175e-06, "loss": 0.0641, "step": 637 }, { "epoch": 1.24853228962818, "grad_norm": 1.938855354605317, "learning_rate": 3.097237815333315e-06, "loss": 0.063, "step": 638 }, { "epoch": 1.2504892367906066, "grad_norm": 1.842736937552211, "learning_rate": 3.0830334363733083e-06, "loss": 0.0582, "step": 639 }, { "epoch": 1.2524461839530332, "grad_norm": 2.0129358229868, "learning_rate": 3.068847171320969e-06, "loss": 0.0554, "step": 640 }, { "epoch": 1.2544031311154598, "grad_norm": 1.73082234411363, "learning_rate": 3.054679154225943e-06, "loss": 0.0688, "step": 641 }, { "epoch": 1.2563600782778865, "grad_norm": 2.1695558389696674, "learning_rate": 3.0405295189654537e-06, "loss": 0.0654, "step": 642 }, { "epoch": 1.258317025440313, "grad_norm": 1.7338543355228768, "learning_rate": 3.0263983992430245e-06, "loss": 0.0622, "step": 643 }, { "epoch": 1.2602739726027397, "grad_norm": 2.816928466932874, "learning_rate": 3.0122859285872214e-06, "loss": 0.0496, "step": 644 }, { "epoch": 1.2622309197651664, "grad_norm": 1.6292278215450644, "learning_rate": 2.9981922403503914e-06, "loss": 0.0839, "step": 645 }, { "epoch": 1.264187866927593, "grad_norm": 1.7997187342454206, "learning_rate": 2.9841174677074037e-06, "loss": 0.0615, "step": 646 }, { "epoch": 1.2661448140900196, "grad_norm": 2.109284767710085, "learning_rate": 2.9700617436543854e-06, "loss": 0.073, "step": 647 }, { "epoch": 1.2681017612524461, "grad_norm": 1.8896351238607167, "learning_rate": 2.956025201007471e-06, "loss": 0.058, "step": 648 }, { "epoch": 1.2700587084148727, "grad_norm": 1.6198675171517074, "learning_rate": 2.942007972401542e-06, "loss": 0.0548, "step": 649 }, { "epoch": 1.2720156555772995, "grad_norm": 1.3600624950706857, "learning_rate": 2.9280101902889823e-06, "loss": 0.0572, "step": 650 }, { "epoch": 1.273972602739726, "grad_norm": 1.7709996147410414, "learning_rate": 2.914031986938417e-06, "loss": 0.068, "step": 651 }, { "epoch": 1.2759295499021526, "grad_norm": 1.5798001632636973, "learning_rate": 2.900073494433464e-06, "loss": 0.0412, "step": 652 }, { "epoch": 1.2778864970645794, "grad_norm": 1.7793326759272565, "learning_rate": 2.8861348446714942e-06, "loss": 0.0658, "step": 653 }, { "epoch": 1.279843444227006, "grad_norm": 1.8408206261252553, "learning_rate": 2.8722161693623775e-06, "loss": 0.0532, "step": 654 }, { "epoch": 1.2818003913894325, "grad_norm": 2.093440351255318, "learning_rate": 2.858317600027235e-06, "loss": 0.056, "step": 655 }, { "epoch": 1.283757338551859, "grad_norm": 1.488307313614249, "learning_rate": 2.8444392679972107e-06, "loss": 0.0571, "step": 656 }, { "epoch": 1.2857142857142856, "grad_norm": 1.7289185819259836, "learning_rate": 2.83058130441221e-06, "loss": 0.0656, "step": 657 }, { "epoch": 1.2876712328767124, "grad_norm": 1.765394496457208, "learning_rate": 2.816743840219681e-06, "loss": 0.0693, "step": 658 }, { "epoch": 1.289628180039139, "grad_norm": 2.066485552498083, "learning_rate": 2.802927006173365e-06, "loss": 0.087, "step": 659 }, { "epoch": 1.2915851272015655, "grad_norm": 2.2119628930686295, "learning_rate": 2.789130932832059e-06, "loss": 0.0762, "step": 660 }, { "epoch": 1.2935420743639923, "grad_norm": 1.4562056282072915, "learning_rate": 2.7753557505583945e-06, "loss": 0.0528, "step": 661 }, { "epoch": 1.2954990215264188, "grad_norm": 1.4790988410227712, "learning_rate": 2.7616015895175953e-06, "loss": 0.0505, "step": 662 }, { "epoch": 1.2974559686888454, "grad_norm": 1.8371392121671533, "learning_rate": 2.747868579676244e-06, "loss": 0.0597, "step": 663 }, { "epoch": 1.299412915851272, "grad_norm": 1.7770982180203057, "learning_rate": 2.734156850801071e-06, "loss": 0.0578, "step": 664 }, { "epoch": 1.3013698630136985, "grad_norm": 1.3783167235929052, "learning_rate": 2.720466532457707e-06, "loss": 0.0477, "step": 665 }, { "epoch": 1.3033268101761253, "grad_norm": 1.8959466884440794, "learning_rate": 2.706797754009476e-06, "loss": 0.0683, "step": 666 }, { "epoch": 1.3052837573385518, "grad_norm": 1.883596829064341, "learning_rate": 2.693150644616165e-06, "loss": 0.0832, "step": 667 }, { "epoch": 1.3072407045009784, "grad_norm": 3.4552562741625046, "learning_rate": 2.6795253332327996e-06, "loss": 0.0854, "step": 668 }, { "epoch": 1.3091976516634052, "grad_norm": 2.096143109924001, "learning_rate": 2.6659219486084366e-06, "loss": 0.0662, "step": 669 }, { "epoch": 1.3111545988258317, "grad_norm": 1.4237458116792328, "learning_rate": 2.652340619284939e-06, "loss": 0.0597, "step": 670 }, { "epoch": 1.3131115459882583, "grad_norm": 1.5898944331553393, "learning_rate": 2.638781473595765e-06, "loss": 0.0616, "step": 671 }, { "epoch": 1.3150684931506849, "grad_norm": 1.9277383750835024, "learning_rate": 2.6252446396647503e-06, "loss": 0.0772, "step": 672 }, { "epoch": 1.3170254403131114, "grad_norm": 1.7823324143617338, "learning_rate": 2.6117302454049e-06, "loss": 0.0833, "step": 673 }, { "epoch": 1.3189823874755382, "grad_norm": 1.8623966302284694, "learning_rate": 2.5982384185171905e-06, "loss": 0.0822, "step": 674 }, { "epoch": 1.3209393346379648, "grad_norm": 1.8452305744413227, "learning_rate": 2.5847692864893423e-06, "loss": 0.0459, "step": 675 }, { "epoch": 1.3228962818003913, "grad_norm": 1.9836562752599032, "learning_rate": 2.5713229765946355e-06, "loss": 0.0772, "step": 676 }, { "epoch": 1.324853228962818, "grad_norm": 1.9291819391648588, "learning_rate": 2.5578996158906887e-06, "loss": 0.0923, "step": 677 }, { "epoch": 1.3268101761252447, "grad_norm": 2.215890629502689, "learning_rate": 2.544499331218274e-06, "loss": 0.0759, "step": 678 }, { "epoch": 1.3287671232876712, "grad_norm": 1.5171109993467264, "learning_rate": 2.531122249200114e-06, "loss": 0.0418, "step": 679 }, { "epoch": 1.3307240704500978, "grad_norm": 1.869289105757668, "learning_rate": 2.5177684962396764e-06, "loss": 0.0585, "step": 680 }, { "epoch": 1.3326810176125243, "grad_norm": 1.7526805762024966, "learning_rate": 2.504438198519992e-06, "loss": 0.0485, "step": 681 }, { "epoch": 1.3346379647749511, "grad_norm": 2.01228049206706, "learning_rate": 2.4911314820024566e-06, "loss": 0.0755, "step": 682 }, { "epoch": 1.3365949119373777, "grad_norm": 1.7723201778383686, "learning_rate": 2.477848472425638e-06, "loss": 0.0796, "step": 683 }, { "epoch": 1.3385518590998042, "grad_norm": 1.7035567557492177, "learning_rate": 2.464589295304096e-06, "loss": 0.0564, "step": 684 }, { "epoch": 1.340508806262231, "grad_norm": 1.7093321560131791, "learning_rate": 2.4513540759271855e-06, "loss": 0.045, "step": 685 }, { "epoch": 1.3424657534246576, "grad_norm": 2.0193427117784317, "learning_rate": 2.438142939357882e-06, "loss": 0.0693, "step": 686 }, { "epoch": 1.3444227005870841, "grad_norm": 2.1352590163037797, "learning_rate": 2.4249560104315992e-06, "loss": 0.0897, "step": 687 }, { "epoch": 1.3463796477495107, "grad_norm": 2.206410381464822, "learning_rate": 2.4117934137550003e-06, "loss": 0.0637, "step": 688 }, { "epoch": 1.3483365949119372, "grad_norm": 1.889613868633253, "learning_rate": 2.398655273704831e-06, "loss": 0.0873, "step": 689 }, { "epoch": 1.350293542074364, "grad_norm": 2.162005014568479, "learning_rate": 2.38554171442674e-06, "loss": 0.0779, "step": 690 }, { "epoch": 1.3522504892367906, "grad_norm": 1.9715776375838776, "learning_rate": 2.3724528598341033e-06, "loss": 0.0833, "step": 691 }, { "epoch": 1.3542074363992171, "grad_norm": 1.5350993428937318, "learning_rate": 2.35938883360686e-06, "loss": 0.0551, "step": 692 }, { "epoch": 1.356164383561644, "grad_norm": 1.424304838183845, "learning_rate": 2.346349759190332e-06, "loss": 0.0427, "step": 693 }, { "epoch": 1.3581213307240705, "grad_norm": 1.3745481951222147, "learning_rate": 2.3333357597940794e-06, "loss": 0.0549, "step": 694 }, { "epoch": 1.360078277886497, "grad_norm": 1.6116580503901854, "learning_rate": 2.3203469583907074e-06, "loss": 0.0642, "step": 695 }, { "epoch": 1.3620352250489236, "grad_norm": 1.506247819833498, "learning_rate": 2.3073834777147237e-06, "loss": 0.0515, "step": 696 }, { "epoch": 1.3639921722113502, "grad_norm": 1.4114996686281258, "learning_rate": 2.294445440261377e-06, "loss": 0.0561, "step": 697 }, { "epoch": 1.365949119373777, "grad_norm": 1.4937466596330902, "learning_rate": 2.2815329682854944e-06, "loss": 0.0527, "step": 698 }, { "epoch": 1.3679060665362035, "grad_norm": 1.7804167165753735, "learning_rate": 2.2686461838003294e-06, "loss": 0.0881, "step": 699 }, { "epoch": 1.36986301369863, "grad_norm": 1.9608753634166096, "learning_rate": 2.2557852085764053e-06, "loss": 0.074, "step": 700 }, { "epoch": 1.3718199608610568, "grad_norm": 1.4106657742479354, "learning_rate": 2.2429501641403657e-06, "loss": 0.0537, "step": 701 }, { "epoch": 1.3737769080234834, "grad_norm": 1.2179955558689581, "learning_rate": 2.230141171773836e-06, "loss": 0.0371, "step": 702 }, { "epoch": 1.37573385518591, "grad_norm": 1.86763580125151, "learning_rate": 2.2173583525122595e-06, "loss": 0.0548, "step": 703 }, { "epoch": 1.3776908023483365, "grad_norm": 2.1525042046027907, "learning_rate": 2.204601827143769e-06, "loss": 0.0804, "step": 704 }, { "epoch": 1.379647749510763, "grad_norm": 1.7316488971121782, "learning_rate": 2.1918717162080343e-06, "loss": 0.0578, "step": 705 }, { "epoch": 1.3816046966731899, "grad_norm": 2.304413523887218, "learning_rate": 2.179168139995134e-06, "loss": 0.0754, "step": 706 }, { "epoch": 1.3835616438356164, "grad_norm": 1.3775997388811496, "learning_rate": 2.1664912185444127e-06, "loss": 0.0362, "step": 707 }, { "epoch": 1.385518590998043, "grad_norm": 1.693175777271181, "learning_rate": 2.1538410716433415e-06, "loss": 0.0591, "step": 708 }, { "epoch": 1.3874755381604698, "grad_norm": 1.5045108537657308, "learning_rate": 2.1412178188263975e-06, "loss": 0.0541, "step": 709 }, { "epoch": 1.3894324853228963, "grad_norm": 3.4031353194684755, "learning_rate": 2.12862157937393e-06, "loss": 0.0409, "step": 710 }, { "epoch": 1.3913894324853229, "grad_norm": 2.0116476853993612, "learning_rate": 2.1160524723110253e-06, "loss": 0.0819, "step": 711 }, { "epoch": 1.3933463796477494, "grad_norm": 2.112821385727089, "learning_rate": 2.103510616406396e-06, "loss": 0.0592, "step": 712 }, { "epoch": 1.395303326810176, "grad_norm": 1.2884634591652226, "learning_rate": 2.090996130171245e-06, "loss": 0.0405, "step": 713 }, { "epoch": 1.3972602739726028, "grad_norm": 2.0544596091337723, "learning_rate": 2.0785091318581577e-06, "loss": 0.0605, "step": 714 }, { "epoch": 1.3992172211350293, "grad_norm": 1.9534615514544305, "learning_rate": 2.0660497394599786e-06, "loss": 0.0812, "step": 715 }, { "epoch": 1.401174168297456, "grad_norm": 1.2063397297159189, "learning_rate": 2.053618070708691e-06, "loss": 0.0308, "step": 716 }, { "epoch": 1.4031311154598827, "grad_norm": 2.0939542607820347, "learning_rate": 2.0412142430743175e-06, "loss": 0.0649, "step": 717 }, { "epoch": 1.4050880626223092, "grad_norm": 1.7458740900062284, "learning_rate": 2.0288383737638007e-06, "loss": 0.0528, "step": 718 }, { "epoch": 1.4070450097847358, "grad_norm": 1.6670093199742038, "learning_rate": 2.0164905797198936e-06, "loss": 0.0581, "step": 719 }, { "epoch": 1.4090019569471623, "grad_norm": 1.6628581645620761, "learning_rate": 2.004170977620066e-06, "loss": 0.0521, "step": 720 }, { "epoch": 1.410958904109589, "grad_norm": 1.5485506887689364, "learning_rate": 1.991879683875386e-06, "loss": 0.0537, "step": 721 }, { "epoch": 1.4129158512720157, "grad_norm": 1.758230405359101, "learning_rate": 1.9796168146294415e-06, "loss": 0.0699, "step": 722 }, { "epoch": 1.4148727984344422, "grad_norm": 1.4650265151694846, "learning_rate": 1.9673824857572173e-06, "loss": 0.0458, "step": 723 }, { "epoch": 1.4168297455968688, "grad_norm": 2.4230845439068336, "learning_rate": 1.9551768128640173e-06, "loss": 0.0694, "step": 724 }, { "epoch": 1.4187866927592956, "grad_norm": 1.7959628834962165, "learning_rate": 1.94299991128437e-06, "loss": 0.0624, "step": 725 }, { "epoch": 1.4207436399217221, "grad_norm": 1.5369701980825707, "learning_rate": 1.9308518960809353e-06, "loss": 0.0493, "step": 726 }, { "epoch": 1.4227005870841487, "grad_norm": 1.9127793746209953, "learning_rate": 1.9187328820434185e-06, "loss": 0.0437, "step": 727 }, { "epoch": 1.4246575342465753, "grad_norm": 2.1136343475550157, "learning_rate": 1.9066429836874844e-06, "loss": 0.0664, "step": 728 }, { "epoch": 1.4266144814090018, "grad_norm": 2.4694212629957106, "learning_rate": 1.894582315253673e-06, "loss": 0.0839, "step": 729 }, { "epoch": 1.4285714285714286, "grad_norm": 1.520994435390283, "learning_rate": 1.8825509907063328e-06, "loss": 0.0471, "step": 730 }, { "epoch": 1.4305283757338552, "grad_norm": 1.7137680245081792, "learning_rate": 1.8705491237325246e-06, "loss": 0.0706, "step": 731 }, { "epoch": 1.4324853228962817, "grad_norm": 1.496040187186284, "learning_rate": 1.8585768277409639e-06, "loss": 0.0466, "step": 732 }, { "epoch": 1.4344422700587085, "grad_norm": 2.2014614096377008, "learning_rate": 1.8466342158609367e-06, "loss": 0.0791, "step": 733 }, { "epoch": 1.436399217221135, "grad_norm": 1.9085021278231573, "learning_rate": 1.83472140094124e-06, "loss": 0.0717, "step": 734 }, { "epoch": 1.4383561643835616, "grad_norm": 2.0940068266788074, "learning_rate": 1.8228384955491136e-06, "loss": 0.0614, "step": 735 }, { "epoch": 1.4403131115459882, "grad_norm": 1.6505304232614793, "learning_rate": 1.8109856119691672e-06, "loss": 0.0474, "step": 736 }, { "epoch": 1.4422700587084147, "grad_norm": 2.086870524642832, "learning_rate": 1.7991628622023338e-06, "loss": 0.0757, "step": 737 }, { "epoch": 1.4442270058708415, "grad_norm": 1.3797632708072753, "learning_rate": 1.7873703579648033e-06, "loss": 0.0419, "step": 738 }, { "epoch": 1.446183953033268, "grad_norm": 1.6056885886417327, "learning_rate": 1.775608210686962e-06, "loss": 0.0493, "step": 739 }, { "epoch": 1.4481409001956946, "grad_norm": 1.5147993779274453, "learning_rate": 1.763876531512354e-06, "loss": 0.0637, "step": 740 }, { "epoch": 1.4500978473581214, "grad_norm": 2.6826343334917846, "learning_rate": 1.7521754312966155e-06, "loss": 0.073, "step": 741 }, { "epoch": 1.452054794520548, "grad_norm": 1.8973880871931268, "learning_rate": 1.7405050206064372e-06, "loss": 0.0666, "step": 742 }, { "epoch": 1.4540117416829745, "grad_norm": 2.1262766209789463, "learning_rate": 1.7288654097185193e-06, "loss": 0.0728, "step": 743 }, { "epoch": 1.455968688845401, "grad_norm": 1.7244455698308037, "learning_rate": 1.7172567086185205e-06, "loss": 0.0521, "step": 744 }, { "epoch": 1.4579256360078277, "grad_norm": 2.035320374250215, "learning_rate": 1.7056790270000302e-06, "loss": 0.0614, "step": 745 }, { "epoch": 1.4598825831702544, "grad_norm": 1.523270916799851, "learning_rate": 1.6941324742635262e-06, "loss": 0.0523, "step": 746 }, { "epoch": 1.461839530332681, "grad_norm": 2.081663532097519, "learning_rate": 1.682617159515338e-06, "loss": 0.0675, "step": 747 }, { "epoch": 1.4637964774951076, "grad_norm": 2.4862479107058864, "learning_rate": 1.671133191566624e-06, "loss": 0.1067, "step": 748 }, { "epoch": 1.4657534246575343, "grad_norm": 1.50534454377478, "learning_rate": 1.6596806789323317e-06, "loss": 0.0506, "step": 749 }, { "epoch": 1.467710371819961, "grad_norm": 1.7146737093942759, "learning_rate": 1.6482597298301917e-06, "loss": 0.0548, "step": 750 }, { "epoch": 1.4696673189823874, "grad_norm": 1.837792847586084, "learning_rate": 1.6368704521796702e-06, "loss": 0.0764, "step": 751 }, { "epoch": 1.471624266144814, "grad_norm": 1.8643851197147219, "learning_rate": 1.625512953600966e-06, "loss": 0.0566, "step": 752 }, { "epoch": 1.4735812133072406, "grad_norm": 1.1492672457170598, "learning_rate": 1.6141873414139914e-06, "loss": 0.0282, "step": 753 }, { "epoch": 1.4755381604696673, "grad_norm": 1.8855465884604348, "learning_rate": 1.6028937226373536e-06, "loss": 0.0846, "step": 754 }, { "epoch": 1.477495107632094, "grad_norm": 1.821363319217877, "learning_rate": 1.5916322039873494e-06, "loss": 0.0479, "step": 755 }, { "epoch": 1.4794520547945205, "grad_norm": 1.5691998214359502, "learning_rate": 1.5804028918769488e-06, "loss": 0.0396, "step": 756 }, { "epoch": 1.4814090019569472, "grad_norm": 1.829010657356671, "learning_rate": 1.5692058924147924e-06, "loss": 0.0664, "step": 757 }, { "epoch": 1.4833659491193738, "grad_norm": 1.3821387971875276, "learning_rate": 1.5580413114042003e-06, "loss": 0.0537, "step": 758 }, { "epoch": 1.4853228962818004, "grad_norm": 1.4433782325216176, "learning_rate": 1.546909254342151e-06, "loss": 0.0486, "step": 759 }, { "epoch": 1.487279843444227, "grad_norm": 1.7173420874456216, "learning_rate": 1.535809826418304e-06, "loss": 0.0605, "step": 760 }, { "epoch": 1.4892367906066535, "grad_norm": 1.2355612002283043, "learning_rate": 1.524743132513991e-06, "loss": 0.0365, "step": 761 }, { "epoch": 1.4911937377690803, "grad_norm": 1.5540166712697212, "learning_rate": 1.513709277201237e-06, "loss": 0.077, "step": 762 }, { "epoch": 1.4931506849315068, "grad_norm": 2.110626861953574, "learning_rate": 1.5027083647417657e-06, "loss": 0.0726, "step": 763 }, { "epoch": 1.4951076320939334, "grad_norm": 2.0559100169521702, "learning_rate": 1.4917404990860112e-06, "loss": 0.0783, "step": 764 }, { "epoch": 1.4970645792563602, "grad_norm": 1.7091301124460199, "learning_rate": 1.4808057838721451e-06, "loss": 0.0586, "step": 765 }, { "epoch": 1.4990215264187867, "grad_norm": 2.8154082459383756, "learning_rate": 1.4699043224250919e-06, "loss": 0.0617, "step": 766 }, { "epoch": 1.5009784735812133, "grad_norm": 1.8277095487303259, "learning_rate": 1.4590362177555467e-06, "loss": 0.0643, "step": 767 }, { "epoch": 1.5029354207436398, "grad_norm": 1.5965599224445597, "learning_rate": 1.4482015725590159e-06, "loss": 0.0462, "step": 768 }, { "epoch": 1.5048923679060664, "grad_norm": 1.7941970933170703, "learning_rate": 1.4374004892148312e-06, "loss": 0.0631, "step": 769 }, { "epoch": 1.5068493150684932, "grad_norm": 1.784409742993431, "learning_rate": 1.4266330697851955e-06, "loss": 0.0668, "step": 770 }, { "epoch": 1.5088062622309197, "grad_norm": 1.9652161692139385, "learning_rate": 1.4158994160142114e-06, "loss": 0.0606, "step": 771 }, { "epoch": 1.5107632093933465, "grad_norm": 1.596945499218886, "learning_rate": 1.4051996293269182e-06, "loss": 0.0489, "step": 772 }, { "epoch": 1.512720156555773, "grad_norm": 1.1580975504383753, "learning_rate": 1.3945338108283402e-06, "loss": 0.0358, "step": 773 }, { "epoch": 1.5146771037181996, "grad_norm": 1.5635863712227143, "learning_rate": 1.3839020613025267e-06, "loss": 0.0559, "step": 774 }, { "epoch": 1.5166340508806262, "grad_norm": 2.175044888942092, "learning_rate": 1.373304481211598e-06, "loss": 0.0471, "step": 775 }, { "epoch": 1.5185909980430528, "grad_norm": 1.9036281757817788, "learning_rate": 1.3627411706948029e-06, "loss": 0.0652, "step": 776 }, { "epoch": 1.5205479452054793, "grad_norm": 2.223700082700391, "learning_rate": 1.3522122295675616e-06, "loss": 0.0654, "step": 777 }, { "epoch": 1.522504892367906, "grad_norm": 2.8343283295021546, "learning_rate": 1.3417177573205399e-06, "loss": 0.0634, "step": 778 }, { "epoch": 1.5244618395303327, "grad_norm": 2.2833821919816963, "learning_rate": 1.3312578531186881e-06, "loss": 0.0651, "step": 779 }, { "epoch": 1.5264187866927594, "grad_norm": 1.7025298865547591, "learning_rate": 1.3208326158003171e-06, "loss": 0.0461, "step": 780 }, { "epoch": 1.528375733855186, "grad_norm": 1.729270457031336, "learning_rate": 1.310442143876164e-06, "loss": 0.0625, "step": 781 }, { "epoch": 1.5303326810176126, "grad_norm": 1.708891663470466, "learning_rate": 1.3000865355284565e-06, "loss": 0.0627, "step": 782 }, { "epoch": 1.532289628180039, "grad_norm": 1.9846182433451856, "learning_rate": 1.2897658886099906e-06, "loss": 0.0607, "step": 783 }, { "epoch": 1.5342465753424657, "grad_norm": 1.415407235278623, "learning_rate": 1.2794803006431984e-06, "loss": 0.0437, "step": 784 }, { "epoch": 1.5362035225048922, "grad_norm": 1.9928798739540419, "learning_rate": 1.2692298688192301e-06, "loss": 0.0621, "step": 785 }, { "epoch": 1.538160469667319, "grad_norm": 1.7787709139723962, "learning_rate": 1.2590146899970446e-06, "loss": 0.0655, "step": 786 }, { "epoch": 1.5401174168297456, "grad_norm": 2.5438877649729164, "learning_rate": 1.2488348607024781e-06, "loss": 0.0506, "step": 787 }, { "epoch": 1.5420743639921723, "grad_norm": 1.8481061686802278, "learning_rate": 1.2386904771273444e-06, "loss": 0.0571, "step": 788 }, { "epoch": 1.544031311154599, "grad_norm": 1.7111367582060732, "learning_rate": 1.228581635128518e-06, "loss": 0.0655, "step": 789 }, { "epoch": 1.5459882583170255, "grad_norm": 2.3567101003362647, "learning_rate": 1.218508430227035e-06, "loss": 0.0607, "step": 790 }, { "epoch": 1.547945205479452, "grad_norm": 2.3065629932234595, "learning_rate": 1.2084709576071885e-06, "loss": 0.0723, "step": 791 }, { "epoch": 1.5499021526418786, "grad_norm": 1.9329060558710456, "learning_rate": 1.1984693121156226e-06, "loss": 0.0693, "step": 792 }, { "epoch": 1.5518590998043051, "grad_norm": 1.6936220133096964, "learning_rate": 1.1885035882604478e-06, "loss": 0.0544, "step": 793 }, { "epoch": 1.553816046966732, "grad_norm": 2.319681534507439, "learning_rate": 1.1785738802103396e-06, "loss": 0.0583, "step": 794 }, { "epoch": 1.5557729941291585, "grad_norm": 2.3378849229183456, "learning_rate": 1.1686802817936488e-06, "loss": 0.0728, "step": 795 }, { "epoch": 1.5577299412915853, "grad_norm": 1.7356726578982005, "learning_rate": 1.158822886497521e-06, "loss": 0.0422, "step": 796 }, { "epoch": 1.5596868884540118, "grad_norm": 1.4160976934390033, "learning_rate": 1.1490017874670046e-06, "loss": 0.0472, "step": 797 }, { "epoch": 1.5616438356164384, "grad_norm": 1.927730450827665, "learning_rate": 1.1392170775041788e-06, "loss": 0.0694, "step": 798 }, { "epoch": 1.563600782778865, "grad_norm": 1.965149814395182, "learning_rate": 1.1294688490672744e-06, "loss": 0.0613, "step": 799 }, { "epoch": 1.5655577299412915, "grad_norm": 1.69119556556352, "learning_rate": 1.1197571942697938e-06, "loss": 0.0628, "step": 800 }, { "epoch": 1.567514677103718, "grad_norm": 1.9659861437660722, "learning_rate": 1.1100822048796494e-06, "loss": 0.063, "step": 801 }, { "epoch": 1.5694716242661448, "grad_norm": 1.6843863057591004, "learning_rate": 1.1004439723182942e-06, "loss": 0.0405, "step": 802 }, { "epoch": 1.5714285714285714, "grad_norm": 1.7530941761672543, "learning_rate": 1.0908425876598512e-06, "loss": 0.0533, "step": 803 }, { "epoch": 1.5733855185909982, "grad_norm": 1.8545811681752948, "learning_rate": 1.0812781416302642e-06, "loss": 0.0583, "step": 804 }, { "epoch": 1.5753424657534247, "grad_norm": 2.1237829875382306, "learning_rate": 1.0717507246064273e-06, "loss": 0.0637, "step": 805 }, { "epoch": 1.5772994129158513, "grad_norm": 1.6538638167391746, "learning_rate": 1.0622604266153485e-06, "loss": 0.0508, "step": 806 }, { "epoch": 1.5792563600782779, "grad_norm": 2.3004217402209743, "learning_rate": 1.0528073373332782e-06, "loss": 0.0845, "step": 807 }, { "epoch": 1.5812133072407044, "grad_norm": 1.5580509752401939, "learning_rate": 1.0433915460848764e-06, "loss": 0.0702, "step": 808 }, { "epoch": 1.583170254403131, "grad_norm": 1.6628813489071526, "learning_rate": 1.0340131418423648e-06, "loss": 0.0647, "step": 809 }, { "epoch": 1.5851272015655578, "grad_norm": 1.452954774165001, "learning_rate": 1.0246722132246857e-06, "loss": 0.0428, "step": 810 }, { "epoch": 1.5870841487279843, "grad_norm": 1.9531701214960568, "learning_rate": 1.015368848496665e-06, "loss": 0.0664, "step": 811 }, { "epoch": 1.589041095890411, "grad_norm": 2.106148046913457, "learning_rate": 1.0061031355681766e-06, "loss": 0.0672, "step": 812 }, { "epoch": 1.5909980430528377, "grad_norm": 2.5549663672672, "learning_rate": 9.968751619933104e-07, "loss": 0.0747, "step": 813 }, { "epoch": 1.5929549902152642, "grad_norm": 1.4850974563366262, "learning_rate": 9.876850149695556e-07, "loss": 0.0481, "step": 814 }, { "epoch": 1.5949119373776908, "grad_norm": 2.081322694051549, "learning_rate": 9.785327813369589e-07, "loss": 0.0939, "step": 815 }, { "epoch": 1.5968688845401173, "grad_norm": 1.6448636174928861, "learning_rate": 9.694185475773216e-07, "loss": 0.068, "step": 816 }, { "epoch": 1.5988258317025439, "grad_norm": 1.7008280241027351, "learning_rate": 9.603423998133682e-07, "loss": 0.0773, "step": 817 }, { "epoch": 1.6007827788649707, "grad_norm": 1.8891713381290396, "learning_rate": 9.513044238079428e-07, "loss": 0.0576, "step": 818 }, { "epoch": 1.6027397260273972, "grad_norm": 2.0569282113697063, "learning_rate": 9.423047049631956e-07, "loss": 0.0709, "step": 819 }, { "epoch": 1.604696673189824, "grad_norm": 1.8892847167319666, "learning_rate": 9.333433283197702e-07, "loss": 0.0569, "step": 820 }, { "epoch": 1.6066536203522506, "grad_norm": 1.6380803004264342, "learning_rate": 9.244203785560096e-07, "loss": 0.064, "step": 821 }, { "epoch": 1.6086105675146771, "grad_norm": 1.9865219362675468, "learning_rate": 9.15535939987151e-07, "loss": 0.0551, "step": 822 }, { "epoch": 1.6105675146771037, "grad_norm": 2.0101074149238904, "learning_rate": 9.066900965645259e-07, "loss": 0.0878, "step": 823 }, { "epoch": 1.6125244618395302, "grad_norm": 1.6107294337985545, "learning_rate": 8.978829318747745e-07, "loss": 0.0485, "step": 824 }, { "epoch": 1.6144814090019568, "grad_norm": 1.8449635958082884, "learning_rate": 8.891145291390479e-07, "loss": 0.0602, "step": 825 }, { "epoch": 1.6164383561643836, "grad_norm": 2.4563639991758475, "learning_rate": 8.803849712122292e-07, "loss": 0.0789, "step": 826 }, { "epoch": 1.6183953033268101, "grad_norm": 2.0062876877311235, "learning_rate": 8.71694340582146e-07, "loss": 0.0677, "step": 827 }, { "epoch": 1.620352250489237, "grad_norm": 2.365184984034521, "learning_rate": 8.63042719368789e-07, "loss": 0.0627, "step": 828 }, { "epoch": 1.6223091976516635, "grad_norm": 2.0604141356919112, "learning_rate": 8.544301893235424e-07, "loss": 0.0725, "step": 829 }, { "epoch": 1.62426614481409, "grad_norm": 1.704546727549742, "learning_rate": 8.458568318284088e-07, "loss": 0.0528, "step": 830 }, { "epoch": 1.6262230919765166, "grad_norm": 1.7918636924250761, "learning_rate": 8.37322727895234e-07, "loss": 0.0533, "step": 831 }, { "epoch": 1.6281800391389432, "grad_norm": 1.7405799110354743, "learning_rate": 8.288279581649528e-07, "loss": 0.0439, "step": 832 }, { "epoch": 1.6301369863013697, "grad_norm": 1.5238231231663932, "learning_rate": 8.203726029068149e-07, "loss": 0.0432, "step": 833 }, { "epoch": 1.6320939334637965, "grad_norm": 2.113434945919066, "learning_rate": 8.119567420176411e-07, "loss": 0.0692, "step": 834 }, { "epoch": 1.634050880626223, "grad_norm": 1.905350428533381, "learning_rate": 8.035804550210524e-07, "loss": 0.0717, "step": 835 }, { "epoch": 1.6360078277886498, "grad_norm": 1.9311428397198607, "learning_rate": 7.952438210667268e-07, "loss": 0.0606, "step": 836 }, { "epoch": 1.6379647749510764, "grad_norm": 1.9647833813997708, "learning_rate": 7.869469189296536e-07, "loss": 0.0432, "step": 837 }, { "epoch": 1.639921722113503, "grad_norm": 1.8984862869657522, "learning_rate": 7.786898270093846e-07, "loss": 0.0574, "step": 838 }, { "epoch": 1.6418786692759295, "grad_norm": 1.4087401367068324, "learning_rate": 7.704726233292953e-07, "loss": 0.0465, "step": 839 }, { "epoch": 1.643835616438356, "grad_norm": 3.5930037331252707, "learning_rate": 7.622953855358456e-07, "loss": 0.0474, "step": 840 }, { "epoch": 1.6457925636007826, "grad_norm": 1.8748110615358404, "learning_rate": 7.541581908978457e-07, "loss": 0.0626, "step": 841 }, { "epoch": 1.6477495107632094, "grad_norm": 1.6453169630124926, "learning_rate": 7.460611163057347e-07, "loss": 0.0448, "step": 842 }, { "epoch": 1.649706457925636, "grad_norm": 1.5256125801165736, "learning_rate": 7.380042382708396e-07, "loss": 0.0509, "step": 843 }, { "epoch": 1.6516634050880628, "grad_norm": 1.9515694037049551, "learning_rate": 7.299876329246652e-07, "loss": 0.0597, "step": 844 }, { "epoch": 1.6536203522504893, "grad_norm": 1.7292450688323209, "learning_rate": 7.220113760181641e-07, "loss": 0.0578, "step": 845 }, { "epoch": 1.6555772994129159, "grad_norm": 1.9653488507320427, "learning_rate": 7.140755429210316e-07, "loss": 0.0593, "step": 846 }, { "epoch": 1.6575342465753424, "grad_norm": 1.7814008989460584, "learning_rate": 7.061802086209857e-07, "loss": 0.0631, "step": 847 }, { "epoch": 1.659491193737769, "grad_norm": 1.9082384491777782, "learning_rate": 6.983254477230589e-07, "loss": 0.0558, "step": 848 }, { "epoch": 1.6614481409001955, "grad_norm": 1.6207231990052715, "learning_rate": 6.905113344488984e-07, "loss": 0.0731, "step": 849 }, { "epoch": 1.6634050880626223, "grad_norm": 1.770202346744306, "learning_rate": 6.827379426360614e-07, "loss": 0.0529, "step": 850 }, { "epoch": 1.6653620352250489, "grad_norm": 1.8962344284058512, "learning_rate": 6.750053457373146e-07, "loss": 0.0823, "step": 851 }, { "epoch": 1.6673189823874757, "grad_norm": 2.647364243015607, "learning_rate": 6.673136168199467e-07, "loss": 0.0807, "step": 852 }, { "epoch": 1.6692759295499022, "grad_norm": 1.889079480666935, "learning_rate": 6.596628285650713e-07, "loss": 0.0613, "step": 853 }, { "epoch": 1.6712328767123288, "grad_norm": 1.88206776814547, "learning_rate": 6.52053053266945e-07, "loss": 0.0525, "step": 854 }, { "epoch": 1.6731898238747553, "grad_norm": 1.6224313830611528, "learning_rate": 6.444843628322845e-07, "loss": 0.0583, "step": 855 }, { "epoch": 1.675146771037182, "grad_norm": 1.8154782521145592, "learning_rate": 6.36956828779579e-07, "loss": 0.0574, "step": 856 }, { "epoch": 1.6771037181996085, "grad_norm": 1.6209641386991935, "learning_rate": 6.294705222384262e-07, "loss": 0.0521, "step": 857 }, { "epoch": 1.6790606653620352, "grad_norm": 1.7562746050410885, "learning_rate": 6.22025513948854e-07, "loss": 0.0448, "step": 858 }, { "epoch": 1.6810176125244618, "grad_norm": 1.5745455714137646, "learning_rate": 6.14621874260648e-07, "loss": 0.0405, "step": 859 }, { "epoch": 1.6829745596868886, "grad_norm": 1.6006255957242048, "learning_rate": 6.072596731326974e-07, "loss": 0.0494, "step": 860 }, { "epoch": 1.6849315068493151, "grad_norm": 1.8323580038394844, "learning_rate": 5.999389801323219e-07, "loss": 0.0714, "step": 861 }, { "epoch": 1.6868884540117417, "grad_norm": 1.6103216647143175, "learning_rate": 5.92659864434626e-07, "loss": 0.0419, "step": 862 }, { "epoch": 1.6888454011741683, "grad_norm": 1.9190542795442047, "learning_rate": 5.854223948218362e-07, "loss": 0.0686, "step": 863 }, { "epoch": 1.6908023483365948, "grad_norm": 1.8448636925113142, "learning_rate": 5.782266396826536e-07, "loss": 0.0645, "step": 864 }, { "epoch": 1.6927592954990214, "grad_norm": 1.6555773337181234, "learning_rate": 5.710726670116107e-07, "loss": 0.0596, "step": 865 }, { "epoch": 1.6947162426614482, "grad_norm": 1.4361859312042813, "learning_rate": 5.639605444084273e-07, "loss": 0.0591, "step": 866 }, { "epoch": 1.6966731898238747, "grad_norm": 1.7193186858734555, "learning_rate": 5.5689033907737e-07, "loss": 0.0637, "step": 867 }, { "epoch": 1.6986301369863015, "grad_norm": 1.6990152171374995, "learning_rate": 5.498621178266167e-07, "loss": 0.0541, "step": 868 }, { "epoch": 1.700587084148728, "grad_norm": 1.658919195335709, "learning_rate": 5.428759470676304e-07, "loss": 0.0526, "step": 869 }, { "epoch": 1.7025440313111546, "grad_norm": 1.8029067782524621, "learning_rate": 5.359318928145262e-07, "loss": 0.049, "step": 870 }, { "epoch": 1.7045009784735812, "grad_norm": 1.879948508885519, "learning_rate": 5.290300206834492e-07, "loss": 0.0506, "step": 871 }, { "epoch": 1.7064579256360077, "grad_norm": 1.8041907721287118, "learning_rate": 5.221703958919572e-07, "loss": 0.0634, "step": 872 }, { "epoch": 1.7084148727984343, "grad_norm": 1.8403077743484653, "learning_rate": 5.15353083258398e-07, "loss": 0.0598, "step": 873 }, { "epoch": 1.710371819960861, "grad_norm": 2.129564556122822, "learning_rate": 5.085781472013051e-07, "loss": 0.0635, "step": 874 }, { "epoch": 1.7123287671232876, "grad_norm": 1.2906373560573414, "learning_rate": 5.018456517387837e-07, "loss": 0.0466, "step": 875 }, { "epoch": 1.7142857142857144, "grad_norm": 1.6638722510270263, "learning_rate": 4.951556604879049e-07, "loss": 0.0567, "step": 876 }, { "epoch": 1.716242661448141, "grad_norm": 1.4813170653976526, "learning_rate": 4.885082366641081e-07, "loss": 0.0432, "step": 877 }, { "epoch": 1.7181996086105675, "grad_norm": 1.8982578906425085, "learning_rate": 4.819034430806047e-07, "loss": 0.0336, "step": 878 }, { "epoch": 1.720156555772994, "grad_norm": 1.8415254205571303, "learning_rate": 4.7534134214777704e-07, "loss": 0.0545, "step": 879 }, { "epoch": 1.7221135029354206, "grad_norm": 2.3502941754237354, "learning_rate": 4.688219958725981e-07, "loss": 0.0645, "step": 880 }, { "epoch": 1.7240704500978472, "grad_norm": 2.1128020191495804, "learning_rate": 4.623454658580373e-07, "loss": 0.0584, "step": 881 }, { "epoch": 1.726027397260274, "grad_norm": 1.6626854294106876, "learning_rate": 4.5591181330248534e-07, "loss": 0.0419, "step": 882 }, { "epoch": 1.7279843444227005, "grad_norm": 3.291013686863315, "learning_rate": 4.495210989991733e-07, "loss": 0.086, "step": 883 }, { "epoch": 1.7299412915851273, "grad_norm": 2.391076557291267, "learning_rate": 4.4317338333559335e-07, "loss": 0.0657, "step": 884 }, { "epoch": 1.7318982387475539, "grad_norm": 1.464471507976512, "learning_rate": 4.368687262929372e-07, "loss": 0.0523, "step": 885 }, { "epoch": 1.7338551859099804, "grad_norm": 1.782227229166943, "learning_rate": 4.306071874455225e-07, "loss": 0.0503, "step": 886 }, { "epoch": 1.735812133072407, "grad_norm": 1.7803088871078996, "learning_rate": 4.243888259602319e-07, "loss": 0.0558, "step": 887 }, { "epoch": 1.7377690802348336, "grad_norm": 2.3859279357436147, "learning_rate": 4.1821370059595577e-07, "loss": 0.0638, "step": 888 }, { "epoch": 1.7397260273972601, "grad_norm": 1.5722542968527555, "learning_rate": 4.1208186970303097e-07, "loss": 0.0495, "step": 889 }, { "epoch": 1.741682974559687, "grad_norm": 1.6490633447411003, "learning_rate": 4.0599339122270097e-07, "loss": 0.0553, "step": 890 }, { "epoch": 1.7436399217221135, "grad_norm": 2.809159546787939, "learning_rate": 3.999483226865547e-07, "loss": 0.0694, "step": 891 }, { "epoch": 1.7455968688845402, "grad_norm": 1.7142510113391978, "learning_rate": 3.939467212159925e-07, "loss": 0.0505, "step": 892 }, { "epoch": 1.7475538160469668, "grad_norm": 2.1329944689126528, "learning_rate": 3.879886435216823e-07, "loss": 0.073, "step": 893 }, { "epoch": 1.7495107632093934, "grad_norm": 1.6547387575508183, "learning_rate": 3.8207414590302525e-07, "loss": 0.0521, "step": 894 }, { "epoch": 1.75146771037182, "grad_norm": 1.8804947461221329, "learning_rate": 3.762032842476238e-07, "loss": 0.0645, "step": 895 }, { "epoch": 1.7534246575342465, "grad_norm": 2.2734049484939667, "learning_rate": 3.70376114030751e-07, "loss": 0.0518, "step": 896 }, { "epoch": 1.755381604696673, "grad_norm": 1.5011440164967709, "learning_rate": 3.6459269031483e-07, "loss": 0.0409, "step": 897 }, { "epoch": 1.7573385518590998, "grad_norm": 2.827267548073318, "learning_rate": 3.5885306774891214e-07, "loss": 0.0938, "step": 898 }, { "epoch": 1.7592954990215264, "grad_norm": 1.8914210587355387, "learning_rate": 3.5315730056815933e-07, "loss": 0.056, "step": 899 }, { "epoch": 1.7612524461839532, "grad_norm": 2.1699847999018953, "learning_rate": 3.475054425933344e-07, "loss": 0.0827, "step": 900 }, { "epoch": 1.7632093933463797, "grad_norm": 1.4227624894414017, "learning_rate": 3.4189754723028825e-07, "loss": 0.0474, "step": 901 }, { "epoch": 1.7651663405088063, "grad_norm": 1.594662485814448, "learning_rate": 3.363336674694601e-07, "loss": 0.0512, "step": 902 }, { "epoch": 1.7671232876712328, "grad_norm": 1.6442010211283895, "learning_rate": 3.308138558853746e-07, "loss": 0.0465, "step": 903 }, { "epoch": 1.7690802348336594, "grad_norm": 1.6903281327965027, "learning_rate": 3.2533816463614253e-07, "loss": 0.0629, "step": 904 }, { "epoch": 1.771037181996086, "grad_norm": 1.5159988737538432, "learning_rate": 3.1990664546297287e-07, "loss": 0.0452, "step": 905 }, { "epoch": 1.7729941291585127, "grad_norm": 2.1819201593291684, "learning_rate": 3.145193496896809e-07, "loss": 0.0727, "step": 906 }, { "epoch": 1.7749510763209393, "grad_norm": 1.4860976764856664, "learning_rate": 3.0917632822220214e-07, "loss": 0.0573, "step": 907 }, { "epoch": 1.776908023483366, "grad_norm": 2.1301474976416834, "learning_rate": 3.0387763154811457e-07, "loss": 0.0601, "step": 908 }, { "epoch": 1.7788649706457926, "grad_norm": 1.6230396517124395, "learning_rate": 2.986233097361579e-07, "loss": 0.0413, "step": 909 }, { "epoch": 1.7808219178082192, "grad_norm": 1.657328127431171, "learning_rate": 2.934134124357646e-07, "loss": 0.0455, "step": 910 }, { "epoch": 1.7827788649706457, "grad_norm": 1.7667737758763842, "learning_rate": 2.882479888765888e-07, "loss": 0.0664, "step": 911 }, { "epoch": 1.7847358121330723, "grad_norm": 1.4196100523327353, "learning_rate": 2.83127087868037e-07, "loss": 0.0447, "step": 912 }, { "epoch": 1.7866927592954989, "grad_norm": 2.130491762036191, "learning_rate": 2.780507577988156e-07, "loss": 0.0594, "step": 913 }, { "epoch": 1.7886497064579256, "grad_norm": 2.3763083307919737, "learning_rate": 2.730190466364652e-07, "loss": 0.0879, "step": 914 }, { "epoch": 1.7906066536203522, "grad_norm": 1.6302863029500756, "learning_rate": 2.6803200192691393e-07, "loss": 0.0679, "step": 915 }, { "epoch": 1.792563600782779, "grad_norm": 2.2128451559681532, "learning_rate": 2.630896707940228e-07, "loss": 0.0669, "step": 916 }, { "epoch": 1.7945205479452055, "grad_norm": 1.4757746959509797, "learning_rate": 2.5819209993914185e-07, "loss": 0.0582, "step": 917 }, { "epoch": 1.796477495107632, "grad_norm": 2.0233269436918957, "learning_rate": 2.533393356406749e-07, "loss": 0.061, "step": 918 }, { "epoch": 1.7984344422700587, "grad_norm": 1.8371258649633007, "learning_rate": 2.485314237536307e-07, "loss": 0.0586, "step": 919 }, { "epoch": 1.8003913894324852, "grad_norm": 1.9256687208410885, "learning_rate": 2.4376840970920133e-07, "loss": 0.0613, "step": 920 }, { "epoch": 1.8023483365949118, "grad_norm": 1.679071066470637, "learning_rate": 2.3905033851432393e-07, "loss": 0.0555, "step": 921 }, { "epoch": 1.8043052837573386, "grad_norm": 2.2609593197948965, "learning_rate": 2.3437725475126126e-07, "loss": 0.0948, "step": 922 }, { "epoch": 1.8062622309197651, "grad_norm": 2.2303619826789722, "learning_rate": 2.297492025771786e-07, "loss": 0.0814, "step": 923 }, { "epoch": 1.808219178082192, "grad_norm": 2.069297086152837, "learning_rate": 2.2516622572372416e-07, "loss": 0.0587, "step": 924 }, { "epoch": 1.8101761252446185, "grad_norm": 1.8326030260904544, "learning_rate": 2.206283674966192e-07, "loss": 0.0537, "step": 925 }, { "epoch": 1.812133072407045, "grad_norm": 2.3371568714440887, "learning_rate": 2.1613567077524878e-07, "loss": 0.0635, "step": 926 }, { "epoch": 1.8140900195694716, "grad_norm": 1.574593373477464, "learning_rate": 2.1168817801225262e-07, "loss": 0.0586, "step": 927 }, { "epoch": 1.8160469667318981, "grad_norm": 1.5475539070878428, "learning_rate": 2.0728593123312934e-07, "loss": 0.0363, "step": 928 }, { "epoch": 1.8180039138943247, "grad_norm": 1.857489286650982, "learning_rate": 2.0292897203583405e-07, "loss": 0.0673, "step": 929 }, { "epoch": 1.8199608610567515, "grad_norm": 1.8071714740379123, "learning_rate": 1.986173415903897e-07, "loss": 0.0495, "step": 930 }, { "epoch": 1.821917808219178, "grad_norm": 1.7891544372252968, "learning_rate": 1.9435108063849684e-07, "loss": 0.0737, "step": 931 }, { "epoch": 1.8238747553816048, "grad_norm": 2.505045917883914, "learning_rate": 1.9013022949314563e-07, "loss": 0.0773, "step": 932 }, { "epoch": 1.8258317025440314, "grad_norm": 2.243530652970425, "learning_rate": 1.8595482803823894e-07, "loss": 0.1031, "step": 933 }, { "epoch": 1.827788649706458, "grad_norm": 1.93997149384046, "learning_rate": 1.8182491572821536e-07, "loss": 0.058, "step": 934 }, { "epoch": 1.8297455968688845, "grad_norm": 1.7304135242773278, "learning_rate": 1.777405315876718e-07, "loss": 0.0551, "step": 935 }, { "epoch": 1.831702544031311, "grad_norm": 1.8817114029087807, "learning_rate": 1.737017142109998e-07, "loss": 0.0559, "step": 936 }, { "epoch": 1.8336594911937376, "grad_norm": 1.5074880432728548, "learning_rate": 1.6970850176201704e-07, "loss": 0.0463, "step": 937 }, { "epoch": 1.8356164383561644, "grad_norm": 1.5908606604195117, "learning_rate": 1.6576093197361253e-07, "loss": 0.0601, "step": 938 }, { "epoch": 1.837573385518591, "grad_norm": 1.8002470321730617, "learning_rate": 1.6185904214738203e-07, "loss": 0.0673, "step": 939 }, { "epoch": 1.8395303326810177, "grad_norm": 1.4694762837396316, "learning_rate": 1.5800286915328034e-07, "loss": 0.0418, "step": 940 }, { "epoch": 1.8414872798434443, "grad_norm": 2.079743857219496, "learning_rate": 1.5419244942927402e-07, "loss": 0.0812, "step": 941 }, { "epoch": 1.8434442270058709, "grad_norm": 1.571025453453051, "learning_rate": 1.5042781898099433e-07, "loss": 0.0513, "step": 942 }, { "epoch": 1.8454011741682974, "grad_norm": 1.5972821719838626, "learning_rate": 1.4670901338139754e-07, "loss": 0.0651, "step": 943 }, { "epoch": 1.847358121330724, "grad_norm": 1.860523944298364, "learning_rate": 1.430360677704301e-07, "loss": 0.0582, "step": 944 }, { "epoch": 1.8493150684931505, "grad_norm": 1.2751667754181393, "learning_rate": 1.39409016854693e-07, "loss": 0.0447, "step": 945 }, { "epoch": 1.8512720156555773, "grad_norm": 2.157680516745502, "learning_rate": 1.3582789490712179e-07, "loss": 0.0566, "step": 946 }, { "epoch": 1.8532289628180039, "grad_norm": 1.441620870920015, "learning_rate": 1.3229273576665148e-07, "loss": 0.0493, "step": 947 }, { "epoch": 1.8551859099804306, "grad_norm": 1.7266804971212752, "learning_rate": 1.2880357283790778e-07, "loss": 0.0552, "step": 948 }, { "epoch": 1.8571428571428572, "grad_norm": 1.9859694070838683, "learning_rate": 1.253604390908819e-07, "loss": 0.0519, "step": 949 }, { "epoch": 1.8590998043052838, "grad_norm": 1.6184016227826419, "learning_rate": 1.219633670606274e-07, "loss": 0.0438, "step": 950 }, { "epoch": 1.8610567514677103, "grad_norm": 2.144613609282388, "learning_rate": 1.1861238884694759e-07, "loss": 0.0725, "step": 951 }, { "epoch": 1.8630136986301369, "grad_norm": 1.7466058198356005, "learning_rate": 1.1530753611409151e-07, "loss": 0.053, "step": 952 }, { "epoch": 1.8649706457925634, "grad_norm": 1.8399981542831136, "learning_rate": 1.1204884009046013e-07, "loss": 0.0606, "step": 953 }, { "epoch": 1.8669275929549902, "grad_norm": 1.593256733749682, "learning_rate": 1.0883633156830553e-07, "loss": 0.0493, "step": 954 }, { "epoch": 1.8688845401174168, "grad_norm": 1.5127579937533648, "learning_rate": 1.0567004090344224e-07, "loss": 0.068, "step": 955 }, { "epoch": 1.8708414872798436, "grad_norm": 1.3562950588272449, "learning_rate": 1.0254999801496246e-07, "loss": 0.0462, "step": 956 }, { "epoch": 1.8727984344422701, "grad_norm": 1.9016630446801348, "learning_rate": 9.947623238494853e-08, "loss": 0.0609, "step": 957 }, { "epoch": 1.8747553816046967, "grad_norm": 1.737171149561637, "learning_rate": 9.644877305819977e-08, "loss": 0.0505, "step": 958 }, { "epoch": 1.8767123287671232, "grad_norm": 1.7003092729424667, "learning_rate": 9.346764864195335e-08, "loss": 0.0534, "step": 959 }, { "epoch": 1.8786692759295498, "grad_norm": 2.071325222333672, "learning_rate": 9.053288730561716e-08, "loss": 0.0529, "step": 960 }, { "epoch": 1.8806262230919764, "grad_norm": 1.6927806033528883, "learning_rate": 8.764451678050234e-08, "loss": 0.0602, "step": 961 }, { "epoch": 1.8825831702544031, "grad_norm": 1.7756293239918934, "learning_rate": 8.480256435956124e-08, "loss": 0.074, "step": 962 }, { "epoch": 1.8845401174168297, "grad_norm": 1.3371800307434571, "learning_rate": 8.200705689712873e-08, "loss": 0.037, "step": 963 }, { "epoch": 1.8864970645792565, "grad_norm": 1.7650692446175578, "learning_rate": 7.925802080867129e-08, "loss": 0.048, "step": 964 }, { "epoch": 1.888454011741683, "grad_norm": 1.8260684292383653, "learning_rate": 7.655548207053332e-08, "loss": 0.0596, "step": 965 }, { "epoch": 1.8904109589041096, "grad_norm": 2.0178520709376566, "learning_rate": 7.389946621969679e-08, "loss": 0.0741, "step": 966 }, { "epoch": 1.8923679060665362, "grad_norm": 1.3693878504898722, "learning_rate": 7.128999835353423e-08, "loss": 0.044, "step": 967 }, { "epoch": 1.8943248532289627, "grad_norm": 1.3766652915225508, "learning_rate": 6.872710312957498e-08, "loss": 0.043, "step": 968 }, { "epoch": 1.8962818003913893, "grad_norm": 1.6390845296936902, "learning_rate": 6.621080476527209e-08, "loss": 0.0508, "step": 969 }, { "epoch": 1.898238747553816, "grad_norm": 1.62778831224083, "learning_rate": 6.374112703777302e-08, "loss": 0.0389, "step": 970 }, { "epoch": 1.9001956947162426, "grad_norm": 1.7721278344241929, "learning_rate": 6.131809328369542e-08, "loss": 0.0668, "step": 971 }, { "epoch": 1.9021526418786694, "grad_norm": 2.3672704464455454, "learning_rate": 5.8941726398903944e-08, "loss": 0.0599, "step": 972 }, { "epoch": 1.904109589041096, "grad_norm": 2.026571937092977, "learning_rate": 5.661204883829763e-08, "loss": 0.0594, "step": 973 }, { "epoch": 1.9060665362035225, "grad_norm": 1.6795423986436178, "learning_rate": 5.432908261559733e-08, "loss": 0.0456, "step": 974 }, { "epoch": 1.908023483365949, "grad_norm": 1.825865197095102, "learning_rate": 5.209284930313418e-08, "loss": 0.0759, "step": 975 }, { "epoch": 1.9099804305283756, "grad_norm": 1.4517722868556924, "learning_rate": 4.9903370031649246e-08, "loss": 0.055, "step": 976 }, { "epoch": 1.9119373776908022, "grad_norm": 2.4958998540178663, "learning_rate": 4.776066549009195e-08, "loss": 0.0985, "step": 977 }, { "epoch": 1.913894324853229, "grad_norm": 1.8603568484535031, "learning_rate": 4.5664755925426406e-08, "loss": 0.0557, "step": 978 }, { "epoch": 1.9158512720156555, "grad_norm": 2.219223227390303, "learning_rate": 4.3615661142439316e-08, "loss": 0.0564, "step": 979 }, { "epoch": 1.9178082191780823, "grad_norm": 1.3892775573999814, "learning_rate": 4.1613400503550114e-08, "loss": 0.0404, "step": 980 }, { "epoch": 1.9197651663405089, "grad_norm": 1.708275130011913, "learning_rate": 3.9657992928632813e-08, "loss": 0.0567, "step": 981 }, { "epoch": 1.9217221135029354, "grad_norm": 2.0231330289157983, "learning_rate": 3.774945689483445e-08, "loss": 0.0578, "step": 982 }, { "epoch": 1.923679060665362, "grad_norm": 1.6563898062971611, "learning_rate": 3.5887810436399106e-08, "loss": 0.0517, "step": 983 }, { "epoch": 1.9256360078277885, "grad_norm": 2.2122731166734333, "learning_rate": 3.407307114450253e-08, "loss": 0.0874, "step": 984 }, { "epoch": 1.927592954990215, "grad_norm": 1.7703374316957134, "learning_rate": 3.2305256167078916e-08, "loss": 0.0523, "step": 985 }, { "epoch": 1.9295499021526419, "grad_norm": 1.601749808627953, "learning_rate": 3.058438220866544e-08, "loss": 0.0441, "step": 986 }, { "epoch": 1.9315068493150684, "grad_norm": 2.081620873644945, "learning_rate": 2.8910465530240793e-08, "loss": 0.0585, "step": 987 }, { "epoch": 1.9334637964774952, "grad_norm": 1.7162451329464556, "learning_rate": 2.7283521949070802e-08, "loss": 0.0596, "step": 988 }, { "epoch": 1.9354207436399218, "grad_norm": 1.4578812461368706, "learning_rate": 2.570356683856301e-08, "loss": 0.0534, "step": 989 }, { "epoch": 1.9373776908023483, "grad_norm": 1.555685866191036, "learning_rate": 2.4170615128117358e-08, "loss": 0.0573, "step": 990 }, { "epoch": 1.939334637964775, "grad_norm": 1.7794556481554296, "learning_rate": 2.2684681302986288e-08, "loss": 0.0572, "step": 991 }, { "epoch": 1.9412915851272015, "grad_norm": 1.8036239391267412, "learning_rate": 2.12457794041393e-08, "loss": 0.0485, "step": 992 }, { "epoch": 1.943248532289628, "grad_norm": 1.5739278137903592, "learning_rate": 1.9853923028128607e-08, "loss": 0.0495, "step": 993 }, { "epoch": 1.9452054794520548, "grad_norm": 1.4911001807236324, "learning_rate": 1.850912532696092e-08, "loss": 0.0403, "step": 994 }, { "epoch": 1.9471624266144814, "grad_norm": 1.1897050271331737, "learning_rate": 1.7211399007974748e-08, "loss": 0.0492, "step": 995 }, { "epoch": 1.9491193737769081, "grad_norm": 1.5506980035744726, "learning_rate": 1.596075633371774e-08, "loss": 0.0621, "step": 996 }, { "epoch": 1.9510763209393347, "grad_norm": 1.6593208402645558, "learning_rate": 1.4757209121832317e-08, "loss": 0.041, "step": 997 }, { "epoch": 1.9530332681017613, "grad_norm": 2.262593837545751, "learning_rate": 1.3600768744944648e-08, "loss": 0.0773, "step": 998 }, { "epoch": 1.9549902152641878, "grad_norm": 1.4884929486069522, "learning_rate": 1.2491446130555306e-08, "loss": 0.0494, "step": 999 }, { "epoch": 1.9569471624266144, "grad_norm": 1.7277417075115977, "learning_rate": 1.1429251760938232e-08, "loss": 0.0479, "step": 1000 }, { "epoch": 1.9569471624266144, "eval_loss": 0.18533086776733398, "eval_runtime": 2.608, "eval_samples_per_second": 16.105, "eval_steps_per_second": 4.218, "step": 1000 }, { "epoch": 1.958904109589041, "grad_norm": 1.4706114219719566, "learning_rate": 1.041419567303914e-08, "loss": 0.0484, "step": 1001 }, { "epoch": 1.9608610567514677, "grad_norm": 1.4380435729500263, "learning_rate": 9.446287458383385e-09, "loss": 0.0458, "step": 1002 }, { "epoch": 1.9628180039138943, "grad_norm": 1.5276409016323822, "learning_rate": 8.525536262982692e-09, "loss": 0.0627, "step": 1003 }, { "epoch": 1.964774951076321, "grad_norm": 1.9223259626766216, "learning_rate": 7.651950787251339e-09, "loss": 0.0653, "step": 1004 }, { "epoch": 1.9667318982387476, "grad_norm": 1.67242957482699, "learning_rate": 6.825539285922334e-09, "loss": 0.0647, "step": 1005 }, { "epoch": 1.9686888454011742, "grad_norm": 2.433392518166898, "learning_rate": 6.046309567968589e-09, "loss": 0.0744, "step": 1006 }, { "epoch": 1.9706457925636007, "grad_norm": 2.2341174202276988, "learning_rate": 5.3142689965318685e-09, "loss": 0.0845, "step": 1007 }, { "epoch": 1.9726027397260273, "grad_norm": 1.5097715088847043, "learning_rate": 4.629424488850065e-09, "loss": 0.0556, "step": 1008 }, { "epoch": 1.9745596868884538, "grad_norm": 2.1060602758418674, "learning_rate": 3.9917825161950305e-09, "loss": 0.054, "step": 1009 }, { "epoch": 1.9765166340508806, "grad_norm": 1.2344336769572644, "learning_rate": 3.4013491038087376e-09, "loss": 0.0445, "step": 1010 }, { "epoch": 1.9784735812133072, "grad_norm": 1.8990070974218787, "learning_rate": 2.858129830847767e-09, "loss": 0.0603, "step": 1011 }, { "epoch": 1.980430528375734, "grad_norm": 2.1551421643817057, "learning_rate": 2.3621298303294626e-09, "loss": 0.0646, "step": 1012 }, { "epoch": 1.9823874755381605, "grad_norm": 2.2862296315926987, "learning_rate": 1.9133537890853036e-09, "loss": 0.0604, "step": 1013 }, { "epoch": 1.984344422700587, "grad_norm": 2.098801348498227, "learning_rate": 1.5118059477142732e-09, "loss": 0.0701, "step": 1014 }, { "epoch": 1.9863013698630136, "grad_norm": 1.6201405795659745, "learning_rate": 1.1574901005456662e-09, "loss": 0.057, "step": 1015 }, { "epoch": 1.9882583170254402, "grad_norm": 1.6103634075408357, "learning_rate": 8.504095956002323e-10, "loss": 0.0674, "step": 1016 }, { "epoch": 1.9902152641878668, "grad_norm": 2.0004367067382147, "learning_rate": 5.905673345601992e-10, "loss": 0.0565, "step": 1017 }, { "epoch": 1.9921722113502935, "grad_norm": 1.5658385827324133, "learning_rate": 3.7796577274096244e-10, "loss": 0.045, "step": 1018 }, { "epoch": 1.99412915851272, "grad_norm": 1.9970284668327416, "learning_rate": 2.1260691906999087e-10, "loss": 0.066, "step": 1019 }, { "epoch": 1.9960861056751469, "grad_norm": 1.803319690321688, "learning_rate": 9.449233606573238e-11, "loss": 0.0492, "step": 1020 }, { "epoch": 1.9980430528375734, "grad_norm": 1.5515325420435353, "learning_rate": 2.3623139821515874e-11, "loss": 0.0534, "step": 1021 }, { "epoch": 2.0, "grad_norm": 1.2630751031116068, "learning_rate": 0.0, "loss": 0.0354, "step": 1022 }, { "epoch": 2.0, "step": 1022, "total_flos": 4941754269696.0, "train_loss": 0.12209972871901238, "train_runtime": 1147.8539, "train_samples_per_second": 7.116, "train_steps_per_second": 0.89 } ], "logging_steps": 1, "max_steps": 1022, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 50000000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4941754269696.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }