{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.990374331550802, "eval_steps": 500, "global_step": 699, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0042780748663101605, "grad_norm": 6.290021896362305, "learning_rate": 9.999949500779842e-06, "loss": 0.3094, "step": 1 }, { "epoch": 0.008556149732620321, "grad_norm": 4.1442551612854, "learning_rate": 9.999798004139435e-06, "loss": 0.2263, "step": 2 }, { "epoch": 0.012834224598930482, "grad_norm": 7.077147483825684, "learning_rate": 9.999545513138964e-06, "loss": 0.2549, "step": 3 }, { "epoch": 0.017112299465240642, "grad_norm": 3.0407891273498535, "learning_rate": 9.999192032878667e-06, "loss": 0.216, "step": 4 }, { "epoch": 0.0213903743315508, "grad_norm": 4.231607913970947, "learning_rate": 9.998737570498737e-06, "loss": 0.2501, "step": 5 }, { "epoch": 0.025668449197860963, "grad_norm": 3.214677095413208, "learning_rate": 9.998182135179173e-06, "loss": 0.2261, "step": 6 }, { "epoch": 0.029946524064171122, "grad_norm": 2.9068033695220947, "learning_rate": 9.997525738139595e-06, "loss": 0.2013, "step": 7 }, { "epoch": 0.034224598930481284, "grad_norm": 2.8735599517822266, "learning_rate": 9.996768392639015e-06, "loss": 0.2379, "step": 8 }, { "epoch": 0.038502673796791446, "grad_norm": 2.869037389755249, "learning_rate": 9.99591011397558e-06, "loss": 0.2079, "step": 9 }, { "epoch": 0.0427807486631016, "grad_norm": 2.877206325531006, "learning_rate": 9.994950919486248e-06, "loss": 0.2156, "step": 10 }, { "epoch": 0.047058823529411764, "grad_norm": 2.555624008178711, "learning_rate": 9.99389082854645e-06, "loss": 0.202, "step": 11 }, { "epoch": 0.051336898395721926, "grad_norm": 2.7975668907165527, "learning_rate": 9.992729862569694e-06, "loss": 0.2073, "step": 12 }, { "epoch": 0.05561497326203209, "grad_norm": 2.7021799087524414, "learning_rate": 9.99146804500713e-06, "loss": 0.1987, "step": 13 }, { "epoch": 0.059893048128342244, "grad_norm": 2.764227867126465, "learning_rate": 9.990105401347075e-06, "loss": 0.1994, "step": 14 }, { "epoch": 0.06417112299465241, "grad_norm": 2.3483848571777344, "learning_rate": 9.988641959114512e-06, "loss": 0.185, "step": 15 }, { "epoch": 0.06844919786096257, "grad_norm": 2.655330181121826, "learning_rate": 9.987077747870512e-06, "loss": 0.2183, "step": 16 }, { "epoch": 0.07272727272727272, "grad_norm": 2.580486536026001, "learning_rate": 9.985412799211658e-06, "loss": 0.2131, "step": 17 }, { "epoch": 0.07700534759358289, "grad_norm": 2.819974660873413, "learning_rate": 9.98364714676939e-06, "loss": 0.1954, "step": 18 }, { "epoch": 0.08128342245989305, "grad_norm": 2.8293025493621826, "learning_rate": 9.981780826209342e-06, "loss": 0.2, "step": 19 }, { "epoch": 0.0855614973262032, "grad_norm": 2.651710033416748, "learning_rate": 9.979813875230604e-06, "loss": 0.1873, "step": 20 }, { "epoch": 0.08983957219251337, "grad_norm": 2.6966662406921387, "learning_rate": 9.97774633356497e-06, "loss": 0.2199, "step": 21 }, { "epoch": 0.09411764705882353, "grad_norm": 2.494934558868408, "learning_rate": 9.97557824297614e-06, "loss": 0.1907, "step": 22 }, { "epoch": 0.09839572192513368, "grad_norm": 2.639111042022705, "learning_rate": 9.97330964725887e-06, "loss": 0.1847, "step": 23 }, { "epoch": 0.10267379679144385, "grad_norm": 2.6100220680236816, "learning_rate": 9.970940592238077e-06, "loss": 0.2099, "step": 24 }, { "epoch": 0.10695187165775401, "grad_norm": 2.4530348777770996, "learning_rate": 9.968471125767942e-06, "loss": 0.1983, "step": 25 }, { "epoch": 0.11122994652406418, "grad_norm": 2.5759458541870117, "learning_rate": 9.965901297730914e-06, "loss": 0.1875, "step": 26 }, { "epoch": 0.11550802139037433, "grad_norm": 2.7890784740448, "learning_rate": 9.963231160036716e-06, "loss": 0.216, "step": 27 }, { "epoch": 0.11978609625668449, "grad_norm": 2.4862325191497803, "learning_rate": 9.960460766621299e-06, "loss": 0.1984, "step": 28 }, { "epoch": 0.12406417112299466, "grad_norm": 2.451136827468872, "learning_rate": 9.957590173445746e-06, "loss": 0.2138, "step": 29 }, { "epoch": 0.12834224598930483, "grad_norm": 2.4993553161621094, "learning_rate": 9.954619438495142e-06, "loss": 0.1895, "step": 30 }, { "epoch": 0.13262032085561498, "grad_norm": 2.3289599418640137, "learning_rate": 9.951548621777409e-06, "loss": 0.1859, "step": 31 }, { "epoch": 0.13689839572192514, "grad_norm": 2.421508312225342, "learning_rate": 9.948377785322082e-06, "loss": 0.1786, "step": 32 }, { "epoch": 0.1411764705882353, "grad_norm": 2.337279796600342, "learning_rate": 9.945106993179074e-06, "loss": 0.1918, "step": 33 }, { "epoch": 0.14545454545454545, "grad_norm": 2.7197248935699463, "learning_rate": 9.941736311417362e-06, "loss": 0.2228, "step": 34 }, { "epoch": 0.1497326203208556, "grad_norm": 2.701374053955078, "learning_rate": 9.938265808123667e-06, "loss": 0.1984, "step": 35 }, { "epoch": 0.15401069518716579, "grad_norm": 2.559514045715332, "learning_rate": 9.934695553401076e-06, "loss": 0.2047, "step": 36 }, { "epoch": 0.15828877005347594, "grad_norm": 2.4385311603546143, "learning_rate": 9.931025619367617e-06, "loss": 0.1835, "step": 37 }, { "epoch": 0.1625668449197861, "grad_norm": 2.552826404571533, "learning_rate": 9.927256080154813e-06, "loss": 0.1876, "step": 38 }, { "epoch": 0.16684491978609625, "grad_norm": 2.4144608974456787, "learning_rate": 9.923387011906183e-06, "loss": 0.1934, "step": 39 }, { "epoch": 0.1711229946524064, "grad_norm": 2.489952325820923, "learning_rate": 9.919418492775694e-06, "loss": 0.1786, "step": 40 }, { "epoch": 0.1754010695187166, "grad_norm": 2.494995594024658, "learning_rate": 9.915350602926198e-06, "loss": 0.1809, "step": 41 }, { "epoch": 0.17967914438502675, "grad_norm": 2.69183611869812, "learning_rate": 9.911183424527802e-06, "loss": 0.224, "step": 42 }, { "epoch": 0.1839572192513369, "grad_norm": 2.377275228500366, "learning_rate": 9.906917041756208e-06, "loss": 0.1604, "step": 43 }, { "epoch": 0.18823529411764706, "grad_norm": 2.2849512100219727, "learning_rate": 9.902551540791016e-06, "loss": 0.1692, "step": 44 }, { "epoch": 0.1925133689839572, "grad_norm": 2.5412590503692627, "learning_rate": 9.898087009813985e-06, "loss": 0.1988, "step": 45 }, { "epoch": 0.19679144385026737, "grad_norm": 2.724073648452759, "learning_rate": 9.893523539007248e-06, "loss": 0.1951, "step": 46 }, { "epoch": 0.20106951871657755, "grad_norm": 2.281855821609497, "learning_rate": 9.888861220551494e-06, "loss": 0.1663, "step": 47 }, { "epoch": 0.2053475935828877, "grad_norm": 2.5755255222320557, "learning_rate": 9.884100148624096e-06, "loss": 0.1894, "step": 48 }, { "epoch": 0.20962566844919786, "grad_norm": 2.42708420753479, "learning_rate": 9.879240419397227e-06, "loss": 0.1804, "step": 49 }, { "epoch": 0.21390374331550802, "grad_norm": 2.489633560180664, "learning_rate": 9.874282131035899e-06, "loss": 0.1788, "step": 50 }, { "epoch": 0.21818181818181817, "grad_norm": 2.472947359085083, "learning_rate": 9.86922538369599e-06, "loss": 0.183, "step": 51 }, { "epoch": 0.22245989304812835, "grad_norm": 2.3927743434906006, "learning_rate": 9.864070279522222e-06, "loss": 0.1895, "step": 52 }, { "epoch": 0.2267379679144385, "grad_norm": 2.3870928287506104, "learning_rate": 9.858816922646088e-06, "loss": 0.2009, "step": 53 }, { "epoch": 0.23101604278074866, "grad_norm": 2.503772735595703, "learning_rate": 9.853465419183759e-06, "loss": 0.1758, "step": 54 }, { "epoch": 0.23529411764705882, "grad_norm": 2.552257537841797, "learning_rate": 9.848015877233935e-06, "loss": 0.1843, "step": 55 }, { "epoch": 0.23957219251336898, "grad_norm": 2.4590237140655518, "learning_rate": 9.842468406875665e-06, "loss": 0.1642, "step": 56 }, { "epoch": 0.24385026737967913, "grad_norm": 2.636364698410034, "learning_rate": 9.836823120166116e-06, "loss": 0.1739, "step": 57 }, { "epoch": 0.24812834224598931, "grad_norm": 2.378171682357788, "learning_rate": 9.831080131138325e-06, "loss": 0.1839, "step": 58 }, { "epoch": 0.25240641711229944, "grad_norm": 2.4146363735198975, "learning_rate": 9.825239555798875e-06, "loss": 0.1921, "step": 59 }, { "epoch": 0.25668449197860965, "grad_norm": 2.578003168106079, "learning_rate": 9.819301512125565e-06, "loss": 0.186, "step": 60 }, { "epoch": 0.2609625668449198, "grad_norm": 2.491945743560791, "learning_rate": 9.813266120065028e-06, "loss": 0.1986, "step": 61 }, { "epoch": 0.26524064171122996, "grad_norm": 2.3031694889068604, "learning_rate": 9.807133501530297e-06, "loss": 0.193, "step": 62 }, { "epoch": 0.2695187165775401, "grad_norm": 2.3124427795410156, "learning_rate": 9.800903780398357e-06, "loss": 0.1927, "step": 63 }, { "epoch": 0.2737967914438503, "grad_norm": 2.406841278076172, "learning_rate": 9.794577082507631e-06, "loss": 0.1905, "step": 64 }, { "epoch": 0.27807486631016043, "grad_norm": 2.2214150428771973, "learning_rate": 9.788153535655442e-06, "loss": 0.1801, "step": 65 }, { "epoch": 0.2823529411764706, "grad_norm": 2.38136625289917, "learning_rate": 9.781633269595432e-06, "loss": 0.1809, "step": 66 }, { "epoch": 0.28663101604278074, "grad_norm": 2.236647605895996, "learning_rate": 9.77501641603494e-06, "loss": 0.1596, "step": 67 }, { "epoch": 0.2909090909090909, "grad_norm": 2.192596673965454, "learning_rate": 9.76830310863235e-06, "loss": 0.1829, "step": 68 }, { "epoch": 0.29518716577540105, "grad_norm": 2.4073846340179443, "learning_rate": 9.761493482994374e-06, "loss": 0.1681, "step": 69 }, { "epoch": 0.2994652406417112, "grad_norm": 2.2297794818878174, "learning_rate": 9.754587676673323e-06, "loss": 0.1577, "step": 70 }, { "epoch": 0.3037433155080214, "grad_norm": 2.5169358253479004, "learning_rate": 9.747585829164332e-06, "loss": 0.1903, "step": 71 }, { "epoch": 0.30802139037433157, "grad_norm": 2.425218343734741, "learning_rate": 9.74048808190254e-06, "loss": 0.1641, "step": 72 }, { "epoch": 0.3122994652406417, "grad_norm": 2.553696632385254, "learning_rate": 9.733294578260224e-06, "loss": 0.1816, "step": 73 }, { "epoch": 0.3165775401069519, "grad_norm": 2.4855220317840576, "learning_rate": 9.726005463543913e-06, "loss": 0.1537, "step": 74 }, { "epoch": 0.32085561497326204, "grad_norm": 2.5661401748657227, "learning_rate": 9.718620884991455e-06, "loss": 0.2037, "step": 75 }, { "epoch": 0.3251336898395722, "grad_norm": 2.5457708835601807, "learning_rate": 9.711140991769028e-06, "loss": 0.207, "step": 76 }, { "epoch": 0.32941176470588235, "grad_norm": 2.4419784545898438, "learning_rate": 9.703565934968146e-06, "loss": 0.1714, "step": 77 }, { "epoch": 0.3336898395721925, "grad_norm": 2.441663980484009, "learning_rate": 9.695895867602591e-06, "loss": 0.1669, "step": 78 }, { "epoch": 0.33796791443850266, "grad_norm": 2.087768316268921, "learning_rate": 9.688130944605332e-06, "loss": 0.1716, "step": 79 }, { "epoch": 0.3422459893048128, "grad_norm": 2.0874452590942383, "learning_rate": 9.680271322825392e-06, "loss": 0.172, "step": 80 }, { "epoch": 0.34652406417112297, "grad_norm": 2.3413197994232178, "learning_rate": 9.672317161024679e-06, "loss": 0.1678, "step": 81 }, { "epoch": 0.3508021390374332, "grad_norm": 2.2530531883239746, "learning_rate": 9.664268619874776e-06, "loss": 0.1796, "step": 82 }, { "epoch": 0.35508021390374334, "grad_norm": 2.2014973163604736, "learning_rate": 9.656125861953711e-06, "loss": 0.1797, "step": 83 }, { "epoch": 0.3593582887700535, "grad_norm": 2.2256293296813965, "learning_rate": 9.647889051742649e-06, "loss": 0.1657, "step": 84 }, { "epoch": 0.36363636363636365, "grad_norm": 2.323029041290283, "learning_rate": 9.639558355622589e-06, "loss": 0.1537, "step": 85 }, { "epoch": 0.3679144385026738, "grad_norm": 2.3917860984802246, "learning_rate": 9.631133941870993e-06, "loss": 0.1801, "step": 86 }, { "epoch": 0.37219251336898396, "grad_norm": 2.34854793548584, "learning_rate": 9.622615980658391e-06, "loss": 0.1802, "step": 87 }, { "epoch": 0.3764705882352941, "grad_norm": 2.6876931190490723, "learning_rate": 9.614004644044943e-06, "loss": 0.2088, "step": 88 }, { "epoch": 0.38074866310160427, "grad_norm": 2.2763171195983887, "learning_rate": 9.60530010597696e-06, "loss": 0.1574, "step": 89 }, { "epoch": 0.3850267379679144, "grad_norm": 2.2022109031677246, "learning_rate": 9.596502542283399e-06, "loss": 0.155, "step": 90 }, { "epoch": 0.3893048128342246, "grad_norm": 2.397830009460449, "learning_rate": 9.587612130672302e-06, "loss": 0.161, "step": 91 }, { "epoch": 0.39358288770053473, "grad_norm": 2.157370090484619, "learning_rate": 9.578629050727208e-06, "loss": 0.1821, "step": 92 }, { "epoch": 0.39786096256684494, "grad_norm": 2.1560373306274414, "learning_rate": 9.569553483903531e-06, "loss": 0.166, "step": 93 }, { "epoch": 0.4021390374331551, "grad_norm": 2.4403135776519775, "learning_rate": 9.56038561352489e-06, "loss": 0.1845, "step": 94 }, { "epoch": 0.40641711229946526, "grad_norm": 2.1271605491638184, "learning_rate": 9.551125624779407e-06, "loss": 0.1697, "step": 95 }, { "epoch": 0.4106951871657754, "grad_norm": 2.0434744358062744, "learning_rate": 9.541773704715966e-06, "loss": 0.1525, "step": 96 }, { "epoch": 0.41497326203208557, "grad_norm": 2.054521322250366, "learning_rate": 9.532330042240434e-06, "loss": 0.1583, "step": 97 }, { "epoch": 0.4192513368983957, "grad_norm": 2.3837196826934814, "learning_rate": 9.522794828111849e-06, "loss": 0.1911, "step": 98 }, { "epoch": 0.4235294117647059, "grad_norm": 2.1054368019104004, "learning_rate": 9.51316825493856e-06, "loss": 0.1574, "step": 99 }, { "epoch": 0.42780748663101603, "grad_norm": 2.444197654724121, "learning_rate": 9.503450517174344e-06, "loss": 0.1731, "step": 100 }, { "epoch": 0.4320855614973262, "grad_norm": 2.2197964191436768, "learning_rate": 9.493641811114472e-06, "loss": 0.1658, "step": 101 }, { "epoch": 0.43636363636363634, "grad_norm": 2.124176263809204, "learning_rate": 9.483742334891747e-06, "loss": 0.1578, "step": 102 }, { "epoch": 0.4406417112299465, "grad_norm": 2.160501718521118, "learning_rate": 9.473752288472499e-06, "loss": 0.1772, "step": 103 }, { "epoch": 0.4449197860962567, "grad_norm": 2.333179235458374, "learning_rate": 9.463671873652551e-06, "loss": 0.1989, "step": 104 }, { "epoch": 0.44919786096256686, "grad_norm": 2.2235171794891357, "learning_rate": 9.453501294053139e-06, "loss": 0.1782, "step": 105 }, { "epoch": 0.453475935828877, "grad_norm": 2.164055347442627, "learning_rate": 9.443240755116797e-06, "loss": 0.1716, "step": 106 }, { "epoch": 0.4577540106951872, "grad_norm": 2.4989864826202393, "learning_rate": 9.432890464103208e-06, "loss": 0.1884, "step": 107 }, { "epoch": 0.46203208556149733, "grad_norm": 2.4820518493652344, "learning_rate": 9.422450630085026e-06, "loss": 0.165, "step": 108 }, { "epoch": 0.4663101604278075, "grad_norm": 2.080653429031372, "learning_rate": 9.411921463943641e-06, "loss": 0.1564, "step": 109 }, { "epoch": 0.47058823529411764, "grad_norm": 2.1066722869873047, "learning_rate": 9.401303178364923e-06, "loss": 0.173, "step": 110 }, { "epoch": 0.4748663101604278, "grad_norm": 2.6190359592437744, "learning_rate": 9.39059598783493e-06, "loss": 0.1754, "step": 111 }, { "epoch": 0.47914438502673795, "grad_norm": 2.1717638969421387, "learning_rate": 9.37980010863557e-06, "loss": 0.1565, "step": 112 }, { "epoch": 0.4834224598930481, "grad_norm": 2.4137771129608154, "learning_rate": 9.368915758840235e-06, "loss": 0.1699, "step": 113 }, { "epoch": 0.48770053475935826, "grad_norm": 2.33005952835083, "learning_rate": 9.357943158309396e-06, "loss": 0.181, "step": 114 }, { "epoch": 0.4919786096256685, "grad_norm": 2.27974009513855, "learning_rate": 9.346882528686159e-06, "loss": 0.1788, "step": 115 }, { "epoch": 0.49625668449197863, "grad_norm": 2.276630401611328, "learning_rate": 9.335734093391797e-06, "loss": 0.2059, "step": 116 }, { "epoch": 0.5005347593582887, "grad_norm": 2.310633897781372, "learning_rate": 9.32449807762122e-06, "loss": 0.1794, "step": 117 }, { "epoch": 0.5048128342245989, "grad_norm": 2.3753018379211426, "learning_rate": 9.313174708338446e-06, "loss": 0.1587, "step": 118 }, { "epoch": 0.509090909090909, "grad_norm": 2.2157907485961914, "learning_rate": 9.301764214272e-06, "loss": 0.1655, "step": 119 }, { "epoch": 0.5133689839572193, "grad_norm": 2.1588330268859863, "learning_rate": 9.2902668259103e-06, "loss": 0.1726, "step": 120 }, { "epoch": 0.5176470588235295, "grad_norm": 2.2454464435577393, "learning_rate": 9.278682775497012e-06, "loss": 0.1573, "step": 121 }, { "epoch": 0.5219251336898396, "grad_norm": 2.451761245727539, "learning_rate": 9.267012297026334e-06, "loss": 0.1819, "step": 122 }, { "epoch": 0.5262032085561498, "grad_norm": 2.2201905250549316, "learning_rate": 9.255255626238295e-06, "loss": 0.1866, "step": 123 }, { "epoch": 0.5304812834224599, "grad_norm": 2.273721694946289, "learning_rate": 9.243413000613974e-06, "loss": 0.1756, "step": 124 }, { "epoch": 0.5347593582887701, "grad_norm": 2.0575530529022217, "learning_rate": 9.231484659370717e-06, "loss": 0.1573, "step": 125 }, { "epoch": 0.5390374331550802, "grad_norm": 2.2469186782836914, "learning_rate": 9.219470843457294e-06, "loss": 0.177, "step": 126 }, { "epoch": 0.5433155080213904, "grad_norm": 2.0902717113494873, "learning_rate": 9.207371795549043e-06, "loss": 0.1544, "step": 127 }, { "epoch": 0.5475935828877005, "grad_norm": 2.0983753204345703, "learning_rate": 9.195187760042952e-06, "loss": 0.1579, "step": 128 }, { "epoch": 0.5518716577540107, "grad_norm": 2.0997560024261475, "learning_rate": 9.182918983052743e-06, "loss": 0.163, "step": 129 }, { "epoch": 0.5561497326203209, "grad_norm": 2.2666311264038086, "learning_rate": 9.17056571240388e-06, "loss": 0.1833, "step": 130 }, { "epoch": 0.560427807486631, "grad_norm": 2.4803414344787598, "learning_rate": 9.158128197628578e-06, "loss": 0.1662, "step": 131 }, { "epoch": 0.5647058823529412, "grad_norm": 2.2313413619995117, "learning_rate": 9.145606689960756e-06, "loss": 0.1621, "step": 132 }, { "epoch": 0.5689839572192513, "grad_norm": 2.558098554611206, "learning_rate": 9.133001442330964e-06, "loss": 0.1865, "step": 133 }, { "epoch": 0.5732620320855615, "grad_norm": 2.387596368789673, "learning_rate": 9.120312709361271e-06, "loss": 0.1784, "step": 134 }, { "epoch": 0.5775401069518716, "grad_norm": 2.392971992492676, "learning_rate": 9.107540747360124e-06, "loss": 0.1662, "step": 135 }, { "epoch": 0.5818181818181818, "grad_norm": 2.194112777709961, "learning_rate": 9.094685814317174e-06, "loss": 0.1479, "step": 136 }, { "epoch": 0.586096256684492, "grad_norm": 2.356377124786377, "learning_rate": 9.081748169898054e-06, "loss": 0.144, "step": 137 }, { "epoch": 0.5903743315508021, "grad_norm": 2.1277108192443848, "learning_rate": 9.068728075439153e-06, "loss": 0.1519, "step": 138 }, { "epoch": 0.5946524064171123, "grad_norm": 2.2070817947387695, "learning_rate": 9.055625793942308e-06, "loss": 0.1784, "step": 139 }, { "epoch": 0.5989304812834224, "grad_norm": 2.1418747901916504, "learning_rate": 9.042441590069526e-06, "loss": 0.1575, "step": 140 }, { "epoch": 0.6032085561497326, "grad_norm": 2.3820064067840576, "learning_rate": 9.029175730137611e-06, "loss": 0.2117, "step": 141 }, { "epoch": 0.6074866310160428, "grad_norm": 2.36942195892334, "learning_rate": 9.015828482112793e-06, "loss": 0.1866, "step": 142 }, { "epoch": 0.611764705882353, "grad_norm": 2.4794163703918457, "learning_rate": 9.002400115605319e-06, "loss": 0.1941, "step": 143 }, { "epoch": 0.6160427807486631, "grad_norm": 2.1948654651641846, "learning_rate": 8.988890901864006e-06, "loss": 0.1696, "step": 144 }, { "epoch": 0.6203208556149733, "grad_norm": 1.9899914264678955, "learning_rate": 8.975301113770756e-06, "loss": 0.1659, "step": 145 }, { "epoch": 0.6245989304812835, "grad_norm": 2.251145839691162, "learning_rate": 8.96163102583505e-06, "loss": 0.1663, "step": 146 }, { "epoch": 0.6288770053475936, "grad_norm": 2.156743049621582, "learning_rate": 8.947880914188397e-06, "loss": 0.1829, "step": 147 }, { "epoch": 0.6331550802139038, "grad_norm": 2.2125680446624756, "learning_rate": 8.934051056578768e-06, "loss": 0.1813, "step": 148 }, { "epoch": 0.6374331550802139, "grad_norm": 2.3191659450531006, "learning_rate": 8.920141732364971e-06, "loss": 0.1849, "step": 149 }, { "epoch": 0.6417112299465241, "grad_norm": 2.1926629543304443, "learning_rate": 8.906153222511014e-06, "loss": 0.1597, "step": 150 }, { "epoch": 0.6459893048128342, "grad_norm": 2.1265509128570557, "learning_rate": 8.892085809580435e-06, "loss": 0.1694, "step": 151 }, { "epoch": 0.6502673796791444, "grad_norm": 2.4438436031341553, "learning_rate": 8.877939777730585e-06, "loss": 0.1957, "step": 152 }, { "epoch": 0.6545454545454545, "grad_norm": 2.3390517234802246, "learning_rate": 8.863715412706897e-06, "loss": 0.1826, "step": 153 }, { "epoch": 0.6588235294117647, "grad_norm": 2.164252519607544, "learning_rate": 8.849413001837105e-06, "loss": 0.1613, "step": 154 }, { "epoch": 0.6631016042780749, "grad_norm": 2.3636364936828613, "learning_rate": 8.83503283402545e-06, "loss": 0.1674, "step": 155 }, { "epoch": 0.667379679144385, "grad_norm": 2.314908504486084, "learning_rate": 8.820575199746835e-06, "loss": 0.1666, "step": 156 }, { "epoch": 0.6716577540106952, "grad_norm": 2.3792495727539062, "learning_rate": 8.806040391040962e-06, "loss": 0.1832, "step": 157 }, { "epoch": 0.6759358288770053, "grad_norm": 2.072890043258667, "learning_rate": 8.791428701506433e-06, "loss": 0.1662, "step": 158 }, { "epoch": 0.6802139037433155, "grad_norm": 2.1991868019104004, "learning_rate": 8.776740426294818e-06, "loss": 0.1748, "step": 159 }, { "epoch": 0.6844919786096256, "grad_norm": 2.133894443511963, "learning_rate": 8.761975862104694e-06, "loss": 0.1837, "step": 160 }, { "epoch": 0.6887700534759358, "grad_norm": 2.057680130004883, "learning_rate": 8.747135307175657e-06, "loss": 0.1741, "step": 161 }, { "epoch": 0.6930481283422459, "grad_norm": 2.2167131900787354, "learning_rate": 8.73221906128228e-06, "loss": 0.1484, "step": 162 }, { "epoch": 0.6973262032085561, "grad_norm": 2.142549753189087, "learning_rate": 8.71722742572808e-06, "loss": 0.1733, "step": 163 }, { "epoch": 0.7016042780748664, "grad_norm": 2.182915210723877, "learning_rate": 8.702160703339422e-06, "loss": 0.1573, "step": 164 }, { "epoch": 0.7058823529411765, "grad_norm": 2.269728422164917, "learning_rate": 8.687019198459395e-06, "loss": 0.1509, "step": 165 }, { "epoch": 0.7101604278074867, "grad_norm": 2.3940067291259766, "learning_rate": 8.671803216941674e-06, "loss": 0.1738, "step": 166 }, { "epoch": 0.7144385026737968, "grad_norm": 1.9983761310577393, "learning_rate": 8.656513066144342e-06, "loss": 0.1493, "step": 167 }, { "epoch": 0.718716577540107, "grad_norm": 2.2644615173339844, "learning_rate": 8.641149054923673e-06, "loss": 0.1618, "step": 168 }, { "epoch": 0.7229946524064171, "grad_norm": 2.1839921474456787, "learning_rate": 8.625711493627902e-06, "loss": 0.1595, "step": 169 }, { "epoch": 0.7272727272727273, "grad_norm": 2.339559555053711, "learning_rate": 8.610200694090951e-06, "loss": 0.1788, "step": 170 }, { "epoch": 0.7315508021390374, "grad_norm": 2.1269736289978027, "learning_rate": 8.594616969626134e-06, "loss": 0.1665, "step": 171 }, { "epoch": 0.7358288770053476, "grad_norm": 2.1057722568511963, "learning_rate": 8.578960635019822e-06, "loss": 0.1716, "step": 172 }, { "epoch": 0.7401069518716578, "grad_norm": 2.206456422805786, "learning_rate": 8.563232006525093e-06, "loss": 0.1522, "step": 173 }, { "epoch": 0.7443850267379679, "grad_norm": 2.352834463119507, "learning_rate": 8.547431401855333e-06, "loss": 0.1788, "step": 174 }, { "epoch": 0.7486631016042781, "grad_norm": 2.3688833713531494, "learning_rate": 8.531559140177828e-06, "loss": 0.1541, "step": 175 }, { "epoch": 0.7529411764705882, "grad_norm": 2.4845242500305176, "learning_rate": 8.515615542107317e-06, "loss": 0.1865, "step": 176 }, { "epoch": 0.7572192513368984, "grad_norm": 2.3540215492248535, "learning_rate": 8.499600929699501e-06, "loss": 0.1741, "step": 177 }, { "epoch": 0.7614973262032085, "grad_norm": 2.173281192779541, "learning_rate": 8.48351562644456e-06, "loss": 0.1922, "step": 178 }, { "epoch": 0.7657754010695187, "grad_norm": 1.9506014585494995, "learning_rate": 8.4673599572606e-06, "loss": 0.1482, "step": 179 }, { "epoch": 0.7700534759358288, "grad_norm": 2.305087089538574, "learning_rate": 8.4511342484871e-06, "loss": 0.187, "step": 180 }, { "epoch": 0.774331550802139, "grad_norm": 2.127180814743042, "learning_rate": 8.434838827878315e-06, "loss": 0.1597, "step": 181 }, { "epoch": 0.7786096256684492, "grad_norm": 2.2191245555877686, "learning_rate": 8.418474024596659e-06, "loss": 0.1715, "step": 182 }, { "epoch": 0.7828877005347593, "grad_norm": 2.0329484939575195, "learning_rate": 8.402040169206054e-06, "loss": 0.1722, "step": 183 }, { "epoch": 0.7871657754010695, "grad_norm": 2.0768754482269287, "learning_rate": 8.38553759366525e-06, "loss": 0.189, "step": 184 }, { "epoch": 0.7914438502673797, "grad_norm": 2.217754602432251, "learning_rate": 8.36896663132113e-06, "loss": 0.1815, "step": 185 }, { "epoch": 0.7957219251336899, "grad_norm": 2.0363705158233643, "learning_rate": 8.352327616901956e-06, "loss": 0.1698, "step": 186 }, { "epoch": 0.8, "grad_norm": 2.1970932483673096, "learning_rate": 8.335620886510637e-06, "loss": 0.1953, "step": 187 }, { "epoch": 0.8042780748663102, "grad_norm": 2.1952292919158936, "learning_rate": 8.318846777617913e-06, "loss": 0.1887, "step": 188 }, { "epoch": 0.8085561497326204, "grad_norm": 2.0600810050964355, "learning_rate": 8.302005629055549e-06, "loss": 0.1543, "step": 189 }, { "epoch": 0.8128342245989305, "grad_norm": 2.0978524684906006, "learning_rate": 8.285097781009497e-06, "loss": 0.1676, "step": 190 }, { "epoch": 0.8171122994652407, "grad_norm": 2.0804502964019775, "learning_rate": 8.268123575013008e-06, "loss": 0.1796, "step": 191 }, { "epoch": 0.8213903743315508, "grad_norm": 2.0107009410858154, "learning_rate": 8.251083353939752e-06, "loss": 0.1453, "step": 192 }, { "epoch": 0.825668449197861, "grad_norm": 2.2001097202301025, "learning_rate": 8.233977461996879e-06, "loss": 0.1642, "step": 193 }, { "epoch": 0.8299465240641711, "grad_norm": 2.205216646194458, "learning_rate": 8.216806244718068e-06, "loss": 0.1662, "step": 194 }, { "epoch": 0.8342245989304813, "grad_norm": 1.9660472869873047, "learning_rate": 8.199570048956553e-06, "loss": 0.1349, "step": 195 }, { "epoch": 0.8385026737967914, "grad_norm": 2.0300116539001465, "learning_rate": 8.182269222878112e-06, "loss": 0.1443, "step": 196 }, { "epoch": 0.8427807486631016, "grad_norm": 2.3219339847564697, "learning_rate": 8.164904115954036e-06, "loss": 0.1753, "step": 197 }, { "epoch": 0.8470588235294118, "grad_norm": 2.4101979732513428, "learning_rate": 8.147475078954067e-06, "loss": 0.181, "step": 198 }, { "epoch": 0.8513368983957219, "grad_norm": 2.243069648742676, "learning_rate": 8.129982463939313e-06, "loss": 0.1656, "step": 199 }, { "epoch": 0.8556149732620321, "grad_norm": 2.026871681213379, "learning_rate": 8.112426624255145e-06, "loss": 0.1566, "step": 200 }, { "epoch": 0.8598930481283422, "grad_norm": 2.2261993885040283, "learning_rate": 8.094807914524048e-06, "loss": 0.1793, "step": 201 }, { "epoch": 0.8641711229946524, "grad_norm": 2.064054250717163, "learning_rate": 8.07712669063846e-06, "loss": 0.1553, "step": 202 }, { "epoch": 0.8684491978609625, "grad_norm": 2.2260794639587402, "learning_rate": 8.059383309753587e-06, "loss": 0.1709, "step": 203 }, { "epoch": 0.8727272727272727, "grad_norm": 2.0984110832214355, "learning_rate": 8.041578130280194e-06, "loss": 0.1664, "step": 204 }, { "epoch": 0.8770053475935828, "grad_norm": 2.1063830852508545, "learning_rate": 8.023711511877347e-06, "loss": 0.1567, "step": 205 }, { "epoch": 0.881283422459893, "grad_norm": 2.295722007751465, "learning_rate": 8.005783815445168e-06, "loss": 0.1775, "step": 206 }, { "epoch": 0.8855614973262033, "grad_norm": 2.045088768005371, "learning_rate": 7.987795403117528e-06, "loss": 0.1631, "step": 207 }, { "epoch": 0.8898395721925134, "grad_norm": 2.2451844215393066, "learning_rate": 7.96974663825475e-06, "loss": 0.1713, "step": 208 }, { "epoch": 0.8941176470588236, "grad_norm": 2.0437660217285156, "learning_rate": 7.95163788543625e-06, "loss": 0.1703, "step": 209 }, { "epoch": 0.8983957219251337, "grad_norm": 2.0744357109069824, "learning_rate": 7.933469510453189e-06, "loss": 0.1593, "step": 210 }, { "epoch": 0.9026737967914439, "grad_norm": 2.1859078407287598, "learning_rate": 7.915241880301075e-06, "loss": 0.1666, "step": 211 }, { "epoch": 0.906951871657754, "grad_norm": 1.9547103643417358, "learning_rate": 7.896955363172347e-06, "loss": 0.1512, "step": 212 }, { "epoch": 0.9112299465240642, "grad_norm": 2.2689852714538574, "learning_rate": 7.878610328448948e-06, "loss": 0.1717, "step": 213 }, { "epoch": 0.9155080213903743, "grad_norm": 2.109738826751709, "learning_rate": 7.86020714669486e-06, "loss": 0.1528, "step": 214 }, { "epoch": 0.9197860962566845, "grad_norm": 2.14335298538208, "learning_rate": 7.84174618964861e-06, "loss": 0.1492, "step": 215 }, { "epoch": 0.9240641711229947, "grad_norm": 2.1951773166656494, "learning_rate": 7.823227830215776e-06, "loss": 0.1942, "step": 216 }, { "epoch": 0.9283422459893048, "grad_norm": 2.0184476375579834, "learning_rate": 7.804652442461438e-06, "loss": 0.1578, "step": 217 }, { "epoch": 0.932620320855615, "grad_norm": 2.1921589374542236, "learning_rate": 7.786020401602638e-06, "loss": 0.1708, "step": 218 }, { "epoch": 0.9368983957219251, "grad_norm": 2.3244035243988037, "learning_rate": 7.767332084000784e-06, "loss": 0.1734, "step": 219 }, { "epoch": 0.9411764705882353, "grad_norm": 2.016603469848633, "learning_rate": 7.748587867154068e-06, "loss": 0.1611, "step": 220 }, { "epoch": 0.9454545454545454, "grad_norm": 1.9263108968734741, "learning_rate": 7.72978812968982e-06, "loss": 0.1519, "step": 221 }, { "epoch": 0.9497326203208556, "grad_norm": 1.985710620880127, "learning_rate": 7.71093325135687e-06, "loss": 0.1668, "step": 222 }, { "epoch": 0.9540106951871657, "grad_norm": 2.146576404571533, "learning_rate": 7.692023613017884e-06, "loss": 0.1669, "step": 223 }, { "epoch": 0.9582887700534759, "grad_norm": 2.262817859649658, "learning_rate": 7.673059596641657e-06, "loss": 0.1638, "step": 224 }, { "epoch": 0.9625668449197861, "grad_norm": 2.1441292762756348, "learning_rate": 7.6540415852954e-06, "loss": 0.1377, "step": 225 }, { "epoch": 0.9668449197860962, "grad_norm": 2.04654598236084, "learning_rate": 7.634969963137015e-06, "loss": 0.1495, "step": 226 }, { "epoch": 0.9711229946524064, "grad_norm": 2.2831695079803467, "learning_rate": 7.615845115407316e-06, "loss": 0.1622, "step": 227 }, { "epoch": 0.9754010695187165, "grad_norm": 2.1478681564331055, "learning_rate": 7.596667428422264e-06, "loss": 0.159, "step": 228 }, { "epoch": 0.9796791443850268, "grad_norm": 2.215297222137451, "learning_rate": 7.5774372895651545e-06, "loss": 0.1743, "step": 229 }, { "epoch": 0.983957219251337, "grad_norm": 1.9584442377090454, "learning_rate": 7.558155087278791e-06, "loss": 0.1517, "step": 230 }, { "epoch": 0.9882352941176471, "grad_norm": 2.235250234603882, "learning_rate": 7.538821211057648e-06, "loss": 0.1697, "step": 231 }, { "epoch": 0.9925133689839573, "grad_norm": 2.22978138923645, "learning_rate": 7.519436051439991e-06, "loss": 0.1607, "step": 232 }, { "epoch": 0.9967914438502674, "grad_norm": 2.2365949153900146, "learning_rate": 7.500000000000001e-06, "loss": 0.1732, "step": 233 }, { "epoch": 1.0010695187165775, "grad_norm": 4.869343280792236, "learning_rate": 7.480513449339851e-06, "loss": 0.2068, "step": 234 }, { "epoch": 1.0053475935828877, "grad_norm": 2.029161214828491, "learning_rate": 7.460976793081789e-06, "loss": 0.1195, "step": 235 }, { "epoch": 1.0096256684491978, "grad_norm": 1.9457329511642456, "learning_rate": 7.441390425860172e-06, "loss": 0.1141, "step": 236 }, { "epoch": 1.013903743315508, "grad_norm": 1.9417977333068848, "learning_rate": 7.421754743313514e-06, "loss": 0.104, "step": 237 }, { "epoch": 1.018181818181818, "grad_norm": 1.8331741094589233, "learning_rate": 7.402070142076475e-06, "loss": 0.0984, "step": 238 }, { "epoch": 1.0224598930481283, "grad_norm": 1.773482322692871, "learning_rate": 7.382337019771859e-06, "loss": 0.1083, "step": 239 }, { "epoch": 1.0267379679144386, "grad_norm": 2.074436664581299, "learning_rate": 7.36255577500258e-06, "loss": 0.1081, "step": 240 }, { "epoch": 1.0310160427807487, "grad_norm": 1.9538006782531738, "learning_rate": 7.342726807343615e-06, "loss": 0.1061, "step": 241 }, { "epoch": 1.035294117647059, "grad_norm": 2.145508289337158, "learning_rate": 7.322850517333924e-06, "loss": 0.1016, "step": 242 }, { "epoch": 1.039572192513369, "grad_norm": 2.154336929321289, "learning_rate": 7.302927306468365e-06, "loss": 0.0851, "step": 243 }, { "epoch": 1.0438502673796792, "grad_norm": 2.3053643703460693, "learning_rate": 7.282957577189581e-06, "loss": 0.0922, "step": 244 }, { "epoch": 1.0481283422459893, "grad_norm": 2.3355484008789062, "learning_rate": 7.2629417328798755e-06, "loss": 0.0848, "step": 245 }, { "epoch": 1.0524064171122995, "grad_norm": 2.6695733070373535, "learning_rate": 7.242880177853062e-06, "loss": 0.0954, "step": 246 }, { "epoch": 1.0566844919786096, "grad_norm": 2.4661552906036377, "learning_rate": 7.222773317346291e-06, "loss": 0.1074, "step": 247 }, { "epoch": 1.0609625668449199, "grad_norm": 2.4028871059417725, "learning_rate": 7.202621557511874e-06, "loss": 0.1058, "step": 248 }, { "epoch": 1.06524064171123, "grad_norm": 2.568382501602173, "learning_rate": 7.1824253054090735e-06, "loss": 0.1168, "step": 249 }, { "epoch": 1.0695187165775402, "grad_norm": 2.346371650695801, "learning_rate": 7.162184968995882e-06, "loss": 0.0936, "step": 250 }, { "epoch": 1.0737967914438502, "grad_norm": 2.163280487060547, "learning_rate": 7.141900957120781e-06, "loss": 0.086, "step": 251 }, { "epoch": 1.0780748663101605, "grad_norm": 2.0890259742736816, "learning_rate": 7.121573679514484e-06, "loss": 0.1047, "step": 252 }, { "epoch": 1.0823529411764705, "grad_norm": 1.9685567617416382, "learning_rate": 7.101203546781655e-06, "loss": 0.0933, "step": 253 }, { "epoch": 1.0866310160427808, "grad_norm": 2.208024024963379, "learning_rate": 7.080790970392626e-06, "loss": 0.1026, "step": 254 }, { "epoch": 1.0909090909090908, "grad_norm": 2.1724014282226562, "learning_rate": 7.060336362675069e-06, "loss": 0.1003, "step": 255 }, { "epoch": 1.095187165775401, "grad_norm": 1.8007311820983887, "learning_rate": 7.039840136805679e-06, "loss": 0.0779, "step": 256 }, { "epoch": 1.0994652406417111, "grad_norm": 1.898895025253296, "learning_rate": 7.019302706801826e-06, "loss": 0.0885, "step": 257 }, { "epoch": 1.1037433155080214, "grad_norm": 1.9565967321395874, "learning_rate": 6.998724487513191e-06, "loss": 0.0948, "step": 258 }, { "epoch": 1.1080213903743314, "grad_norm": 1.9110807180404663, "learning_rate": 6.978105894613385e-06, "loss": 0.0833, "step": 259 }, { "epoch": 1.1122994652406417, "grad_norm": 1.9192978143692017, "learning_rate": 6.9574473445915495e-06, "loss": 0.0932, "step": 260 }, { "epoch": 1.116577540106952, "grad_norm": 2.101365566253662, "learning_rate": 6.936749254743951e-06, "loss": 0.1016, "step": 261 }, { "epoch": 1.120855614973262, "grad_norm": 2.0373008251190186, "learning_rate": 6.916012043165552e-06, "loss": 0.0864, "step": 262 }, { "epoch": 1.125133689839572, "grad_norm": 2.0774929523468018, "learning_rate": 6.895236128741554e-06, "loss": 0.0842, "step": 263 }, { "epoch": 1.1294117647058823, "grad_norm": 2.0311055183410645, "learning_rate": 6.87442193113895e-06, "loss": 0.083, "step": 264 }, { "epoch": 1.1336898395721926, "grad_norm": 2.03928542137146, "learning_rate": 6.8535698707980356e-06, "loss": 0.0903, "step": 265 }, { "epoch": 1.1379679144385026, "grad_norm": 1.9919860363006592, "learning_rate": 6.83268036892393e-06, "loss": 0.0893, "step": 266 }, { "epoch": 1.142245989304813, "grad_norm": 2.215510368347168, "learning_rate": 6.811753847478051e-06, "loss": 0.0913, "step": 267 }, { "epoch": 1.146524064171123, "grad_norm": 2.1406352519989014, "learning_rate": 6.790790729169604e-06, "loss": 0.0935, "step": 268 }, { "epoch": 1.1508021390374332, "grad_norm": 2.2298877239227295, "learning_rate": 6.769791437447042e-06, "loss": 0.0955, "step": 269 }, { "epoch": 1.1550802139037433, "grad_norm": 2.185397148132324, "learning_rate": 6.7487563964895066e-06, "loss": 0.1009, "step": 270 }, { "epoch": 1.1593582887700535, "grad_norm": 2.0531294345855713, "learning_rate": 6.7276860311982614e-06, "loss": 0.1009, "step": 271 }, { "epoch": 1.1636363636363636, "grad_norm": 2.1386170387268066, "learning_rate": 6.7065807671881155e-06, "loss": 0.1033, "step": 272 }, { "epoch": 1.1679144385026738, "grad_norm": 1.994783639907837, "learning_rate": 6.6854410307788175e-06, "loss": 0.1, "step": 273 }, { "epoch": 1.172192513368984, "grad_norm": 1.9821001291275024, "learning_rate": 6.664267248986447e-06, "loss": 0.0909, "step": 274 }, { "epoch": 1.1764705882352942, "grad_norm": 2.178961753845215, "learning_rate": 6.643059849514795e-06, "loss": 0.1047, "step": 275 }, { "epoch": 1.1807486631016042, "grad_norm": 2.1259138584136963, "learning_rate": 6.621819260746713e-06, "loss": 0.0945, "step": 276 }, { "epoch": 1.1850267379679145, "grad_norm": 2.221371650695801, "learning_rate": 6.600545911735468e-06, "loss": 0.0932, "step": 277 }, { "epoch": 1.1893048128342245, "grad_norm": 2.518772840499878, "learning_rate": 6.579240232196073e-06, "loss": 0.0984, "step": 278 }, { "epoch": 1.1935828877005348, "grad_norm": 2.1421663761138916, "learning_rate": 6.5579026524966106e-06, "loss": 0.107, "step": 279 }, { "epoch": 1.1978609625668448, "grad_norm": 1.9127075672149658, "learning_rate": 6.536533603649536e-06, "loss": 0.0957, "step": 280 }, { "epoch": 1.202139037433155, "grad_norm": 2.00793719291687, "learning_rate": 6.515133517302969e-06, "loss": 0.0925, "step": 281 }, { "epoch": 1.2064171122994654, "grad_norm": 1.9672478437423706, "learning_rate": 6.493702825731977e-06, "loss": 0.0908, "step": 282 }, { "epoch": 1.2106951871657754, "grad_norm": 2.3069119453430176, "learning_rate": 6.472241961829846e-06, "loss": 0.0972, "step": 283 }, { "epoch": 1.2149732620320854, "grad_norm": 1.9063525199890137, "learning_rate": 6.450751359099332e-06, "loss": 0.0867, "step": 284 }, { "epoch": 1.2192513368983957, "grad_norm": 2.1970877647399902, "learning_rate": 6.429231451643907e-06, "loss": 0.1077, "step": 285 }, { "epoch": 1.223529411764706, "grad_norm": 2.0319154262542725, "learning_rate": 6.407682674158988e-06, "loss": 0.0966, "step": 286 }, { "epoch": 1.227807486631016, "grad_norm": 2.0063929557800293, "learning_rate": 6.386105461923159e-06, "loss": 0.0938, "step": 287 }, { "epoch": 1.2320855614973263, "grad_norm": 1.9851186275482178, "learning_rate": 6.364500250789375e-06, "loss": 0.0872, "step": 288 }, { "epoch": 1.2363636363636363, "grad_norm": 2.383897304534912, "learning_rate": 6.342867477176164e-06, "loss": 0.1185, "step": 289 }, { "epoch": 1.2406417112299466, "grad_norm": 2.2189831733703613, "learning_rate": 6.321207578058803e-06, "loss": 0.1042, "step": 290 }, { "epoch": 1.2449197860962566, "grad_norm": 2.208324432373047, "learning_rate": 6.299520990960497e-06, "loss": 0.1209, "step": 291 }, { "epoch": 1.249197860962567, "grad_norm": 2.279092311859131, "learning_rate": 6.2778081539435436e-06, "loss": 0.0957, "step": 292 }, { "epoch": 1.253475935828877, "grad_norm": 1.989281415939331, "learning_rate": 6.256069505600474e-06, "loss": 0.0909, "step": 293 }, { "epoch": 1.2577540106951872, "grad_norm": 2.0415663719177246, "learning_rate": 6.234305485045205e-06, "loss": 0.1003, "step": 294 }, { "epoch": 1.2620320855614973, "grad_norm": 1.9271050691604614, "learning_rate": 6.212516531904164e-06, "loss": 0.0901, "step": 295 }, { "epoch": 1.2663101604278075, "grad_norm": 1.998005986213684, "learning_rate": 6.1907030863074055e-06, "loss": 0.0931, "step": 296 }, { "epoch": 1.2705882352941176, "grad_norm": 1.9580942392349243, "learning_rate": 6.16886558887973e-06, "loss": 0.0798, "step": 297 }, { "epoch": 1.2748663101604278, "grad_norm": 2.072495222091675, "learning_rate": 6.1470044807317695e-06, "loss": 0.0883, "step": 298 }, { "epoch": 1.279144385026738, "grad_norm": 2.0811755657196045, "learning_rate": 6.1251202034510905e-06, "loss": 0.097, "step": 299 }, { "epoch": 1.2834224598930482, "grad_norm": 1.9838371276855469, "learning_rate": 6.103213199093267e-06, "loss": 0.1117, "step": 300 }, { "epoch": 1.2877005347593582, "grad_norm": 1.8616009950637817, "learning_rate": 6.081283910172956e-06, "loss": 0.0988, "step": 301 }, { "epoch": 1.2919786096256685, "grad_norm": 1.9417067766189575, "learning_rate": 6.059332779654953e-06, "loss": 0.083, "step": 302 }, { "epoch": 1.2962566844919787, "grad_norm": 1.8084677457809448, "learning_rate": 6.037360250945243e-06, "loss": 0.0915, "step": 303 }, { "epoch": 1.3005347593582888, "grad_norm": 2.0521080493927, "learning_rate": 6.015366767882054e-06, "loss": 0.0923, "step": 304 }, { "epoch": 1.3048128342245988, "grad_norm": 2.042677879333496, "learning_rate": 5.993352774726885e-06, "loss": 0.0825, "step": 305 }, { "epoch": 1.309090909090909, "grad_norm": 2.153756856918335, "learning_rate": 5.97131871615553e-06, "loss": 0.0878, "step": 306 }, { "epoch": 1.3133689839572193, "grad_norm": 2.1429085731506348, "learning_rate": 5.949265037249096e-06, "loss": 0.1033, "step": 307 }, { "epoch": 1.3176470588235294, "grad_norm": 2.3176093101501465, "learning_rate": 5.927192183485023e-06, "loss": 0.1035, "step": 308 }, { "epoch": 1.3219251336898394, "grad_norm": 2.045224666595459, "learning_rate": 5.905100600728067e-06, "loss": 0.0919, "step": 309 }, { "epoch": 1.3262032085561497, "grad_norm": 2.2196807861328125, "learning_rate": 5.882990735221312e-06, "loss": 0.1003, "step": 310 }, { "epoch": 1.33048128342246, "grad_norm": 2.4191091060638428, "learning_rate": 5.860863033577141e-06, "loss": 0.1003, "step": 311 }, { "epoch": 1.33475935828877, "grad_norm": 2.096099376678467, "learning_rate": 5.8387179427682265e-06, "loss": 0.0855, "step": 312 }, { "epoch": 1.3390374331550803, "grad_norm": 2.179696798324585, "learning_rate": 5.8165559101184955e-06, "loss": 0.0936, "step": 313 }, { "epoch": 1.3433155080213903, "grad_norm": 2.0698587894439697, "learning_rate": 5.794377383294094e-06, "loss": 0.0873, "step": 314 }, { "epoch": 1.3475935828877006, "grad_norm": 1.9027916193008423, "learning_rate": 5.7721828102943445e-06, "loss": 0.0865, "step": 315 }, { "epoch": 1.3518716577540106, "grad_norm": 1.8766309022903442, "learning_rate": 5.749972639442698e-06, "loss": 0.0807, "step": 316 }, { "epoch": 1.356149732620321, "grad_norm": 2.1122074127197266, "learning_rate": 5.72774731937768e-06, "loss": 0.0864, "step": 317 }, { "epoch": 1.360427807486631, "grad_norm": 2.0253007411956787, "learning_rate": 5.705507299043822e-06, "loss": 0.0895, "step": 318 }, { "epoch": 1.3647058823529412, "grad_norm": 2.0981879234313965, "learning_rate": 5.683253027682597e-06, "loss": 0.0843, "step": 319 }, { "epoch": 1.3689839572192513, "grad_norm": 2.0904572010040283, "learning_rate": 5.660984954823342e-06, "loss": 0.0921, "step": 320 }, { "epoch": 1.3732620320855615, "grad_norm": 1.9523817300796509, "learning_rate": 5.638703530274187e-06, "loss": 0.0834, "step": 321 }, { "epoch": 1.3775401069518716, "grad_norm": 2.0361475944519043, "learning_rate": 5.6164092041129544e-06, "loss": 0.0919, "step": 322 }, { "epoch": 1.3818181818181818, "grad_norm": 2.2082595825195312, "learning_rate": 5.594102426678082e-06, "loss": 0.0946, "step": 323 }, { "epoch": 1.386096256684492, "grad_norm": 2.1961371898651123, "learning_rate": 5.57178364855951e-06, "loss": 0.0892, "step": 324 }, { "epoch": 1.3903743315508021, "grad_norm": 1.9156455993652344, "learning_rate": 5.549453320589598e-06, "loss": 0.0936, "step": 325 }, { "epoch": 1.3946524064171122, "grad_norm": 2.183152914047241, "learning_rate": 5.527111893834004e-06, "loss": 0.0979, "step": 326 }, { "epoch": 1.3989304812834225, "grad_norm": 2.356243371963501, "learning_rate": 5.504759819582581e-06, "loss": 0.0923, "step": 327 }, { "epoch": 1.4032085561497327, "grad_norm": 2.1394667625427246, "learning_rate": 5.482397549340256e-06, "loss": 0.0893, "step": 328 }, { "epoch": 1.4074866310160428, "grad_norm": 1.9834396839141846, "learning_rate": 5.460025534817911e-06, "loss": 0.0927, "step": 329 }, { "epoch": 1.4117647058823528, "grad_norm": 1.8735346794128418, "learning_rate": 5.437644227923261e-06, "loss": 0.0901, "step": 330 }, { "epoch": 1.416042780748663, "grad_norm": 2.1759555339813232, "learning_rate": 5.415254080751725e-06, "loss": 0.1009, "step": 331 }, { "epoch": 1.4203208556149733, "grad_norm": 1.9985833168029785, "learning_rate": 5.39285554557729e-06, "loss": 0.1012, "step": 332 }, { "epoch": 1.4245989304812834, "grad_norm": 2.2998223304748535, "learning_rate": 5.37044907484338e-06, "loss": 0.0982, "step": 333 }, { "epoch": 1.4288770053475937, "grad_norm": 2.192631721496582, "learning_rate": 5.348035121153716e-06, "loss": 0.1057, "step": 334 }, { "epoch": 1.4331550802139037, "grad_norm": 1.9574298858642578, "learning_rate": 5.32561413726317e-06, "loss": 0.0937, "step": 335 }, { "epoch": 1.437433155080214, "grad_norm": 2.097489833831787, "learning_rate": 5.303186576068621e-06, "loss": 0.0936, "step": 336 }, { "epoch": 1.441711229946524, "grad_norm": 2.246332883834839, "learning_rate": 5.28075289059981e-06, "loss": 0.1037, "step": 337 }, { "epoch": 1.4459893048128343, "grad_norm": 2.124225378036499, "learning_rate": 5.258313534010187e-06, "loss": 0.0807, "step": 338 }, { "epoch": 1.4502673796791443, "grad_norm": 1.9736027717590332, "learning_rate": 5.235868959567755e-06, "loss": 0.0945, "step": 339 }, { "epoch": 1.4545454545454546, "grad_norm": 2.101895332336426, "learning_rate": 5.213419620645914e-06, "loss": 0.0943, "step": 340 }, { "epoch": 1.4588235294117646, "grad_norm": 2.1007871627807617, "learning_rate": 5.1909659707143105e-06, "loss": 0.0893, "step": 341 }, { "epoch": 1.463101604278075, "grad_norm": 2.1706559658050537, "learning_rate": 5.1685084633296665e-06, "loss": 0.0964, "step": 342 }, { "epoch": 1.467379679144385, "grad_norm": 2.181288957595825, "learning_rate": 5.14604755212663e-06, "loss": 0.0856, "step": 343 }, { "epoch": 1.4716577540106952, "grad_norm": 2.169097661972046, "learning_rate": 5.123583690808596e-06, "loss": 0.0993, "step": 344 }, { "epoch": 1.4759358288770055, "grad_norm": 2.1425249576568604, "learning_rate": 5.101117333138558e-06, "loss": 0.0975, "step": 345 }, { "epoch": 1.4802139037433155, "grad_norm": 2.147711753845215, "learning_rate": 5.078648932929933e-06, "loss": 0.1071, "step": 346 }, { "epoch": 1.4844919786096256, "grad_norm": 2.0641732215881348, "learning_rate": 5.056178944037396e-06, "loss": 0.0926, "step": 347 }, { "epoch": 1.4887700534759358, "grad_norm": 2.0812582969665527, "learning_rate": 5.033707820347715e-06, "loss": 0.0924, "step": 348 }, { "epoch": 1.493048128342246, "grad_norm": 2.1125471591949463, "learning_rate": 5.011236015770577e-06, "loss": 0.0978, "step": 349 }, { "epoch": 1.4973262032085561, "grad_norm": 2.063429117202759, "learning_rate": 4.988763984229425e-06, "loss": 0.1016, "step": 350 }, { "epoch": 1.5016042780748662, "grad_norm": 2.123988628387451, "learning_rate": 4.9662921796522856e-06, "loss": 0.0926, "step": 351 }, { "epoch": 1.5058823529411764, "grad_norm": 1.9401912689208984, "learning_rate": 4.9438210559626045e-06, "loss": 0.0931, "step": 352 }, { "epoch": 1.5101604278074867, "grad_norm": 2.213088035583496, "learning_rate": 4.921351067070068e-06, "loss": 0.1124, "step": 353 }, { "epoch": 1.5144385026737968, "grad_norm": 2.0530011653900146, "learning_rate": 4.898882666861444e-06, "loss": 0.0867, "step": 354 }, { "epoch": 1.5187165775401068, "grad_norm": 2.306276321411133, "learning_rate": 4.876416309191406e-06, "loss": 0.0968, "step": 355 }, { "epoch": 1.522994652406417, "grad_norm": 2.0044631958007812, "learning_rate": 4.853952447873371e-06, "loss": 0.082, "step": 356 }, { "epoch": 1.5272727272727273, "grad_norm": 2.2179064750671387, "learning_rate": 4.831491536670334e-06, "loss": 0.0991, "step": 357 }, { "epoch": 1.5315508021390374, "grad_norm": 2.140604257583618, "learning_rate": 4.809034029285691e-06, "loss": 0.0882, "step": 358 }, { "epoch": 1.5358288770053476, "grad_norm": 1.9317160844802856, "learning_rate": 4.786580379354087e-06, "loss": 0.0812, "step": 359 }, { "epoch": 1.5401069518716577, "grad_norm": 2.0457096099853516, "learning_rate": 4.7641310404322475e-06, "loss": 0.0936, "step": 360 }, { "epoch": 1.544385026737968, "grad_norm": 2.2014577388763428, "learning_rate": 4.741686465989814e-06, "loss": 0.0813, "step": 361 }, { "epoch": 1.5486631016042782, "grad_norm": 1.9533830881118774, "learning_rate": 4.719247109400192e-06, "loss": 0.0949, "step": 362 }, { "epoch": 1.5529411764705883, "grad_norm": 1.9284648895263672, "learning_rate": 4.696813423931381e-06, "loss": 0.0858, "step": 363 }, { "epoch": 1.5572192513368983, "grad_norm": 2.015251398086548, "learning_rate": 4.674385862736832e-06, "loss": 0.0993, "step": 364 }, { "epoch": 1.5614973262032086, "grad_norm": 2.260143518447876, "learning_rate": 4.651964878846285e-06, "loss": 0.1086, "step": 365 }, { "epoch": 1.5657754010695188, "grad_norm": 2.37188982963562, "learning_rate": 4.62955092515662e-06, "loss": 0.1233, "step": 366 }, { "epoch": 1.570053475935829, "grad_norm": 1.9625977277755737, "learning_rate": 4.607144454422711e-06, "loss": 0.0893, "step": 367 }, { "epoch": 1.574331550802139, "grad_norm": 2.1813342571258545, "learning_rate": 4.584745919248275e-06, "loss": 0.0923, "step": 368 }, { "epoch": 1.5786096256684492, "grad_norm": 1.9883830547332764, "learning_rate": 4.56235577207674e-06, "loss": 0.103, "step": 369 }, { "epoch": 1.5828877005347595, "grad_norm": 2.1959195137023926, "learning_rate": 4.5399744651820915e-06, "loss": 0.0959, "step": 370 }, { "epoch": 1.5871657754010695, "grad_norm": 1.9669009447097778, "learning_rate": 4.517602450659746e-06, "loss": 0.0885, "step": 371 }, { "epoch": 1.5914438502673796, "grad_norm": 1.9892116785049438, "learning_rate": 4.49524018041742e-06, "loss": 0.0909, "step": 372 }, { "epoch": 1.5957219251336898, "grad_norm": 2.0287320613861084, "learning_rate": 4.472888106165995e-06, "loss": 0.0847, "step": 373 }, { "epoch": 1.6, "grad_norm": 2.1762945652008057, "learning_rate": 4.450546679410403e-06, "loss": 0.0886, "step": 374 }, { "epoch": 1.6042780748663101, "grad_norm": 1.9880300760269165, "learning_rate": 4.428216351440492e-06, "loss": 0.1015, "step": 375 }, { "epoch": 1.6085561497326202, "grad_norm": 2.06316876411438, "learning_rate": 4.40589757332192e-06, "loss": 0.0997, "step": 376 }, { "epoch": 1.6128342245989304, "grad_norm": 2.0702295303344727, "learning_rate": 4.383590795887046e-06, "loss": 0.0929, "step": 377 }, { "epoch": 1.6171122994652407, "grad_norm": 2.1991732120513916, "learning_rate": 4.361296469725813e-06, "loss": 0.1022, "step": 378 }, { "epoch": 1.6213903743315508, "grad_norm": 2.10152006149292, "learning_rate": 4.339015045176659e-06, "loss": 0.0855, "step": 379 }, { "epoch": 1.6256684491978608, "grad_norm": 2.2753965854644775, "learning_rate": 4.316746972317406e-06, "loss": 0.1037, "step": 380 }, { "epoch": 1.629946524064171, "grad_norm": 1.9243597984313965, "learning_rate": 4.2944927009561786e-06, "loss": 0.0841, "step": 381 }, { "epoch": 1.6342245989304813, "grad_norm": 1.9146336317062378, "learning_rate": 4.272252680622321e-06, "loss": 0.0833, "step": 382 }, { "epoch": 1.6385026737967916, "grad_norm": 1.7657594680786133, "learning_rate": 4.250027360557302e-06, "loss": 0.0818, "step": 383 }, { "epoch": 1.6427807486631016, "grad_norm": 2.1305899620056152, "learning_rate": 4.227817189705657e-06, "loss": 0.0962, "step": 384 }, { "epoch": 1.6470588235294117, "grad_norm": 2.021326780319214, "learning_rate": 4.205622616705909e-06, "loss": 0.0859, "step": 385 }, { "epoch": 1.651336898395722, "grad_norm": 2.08622407913208, "learning_rate": 4.183444089881506e-06, "loss": 0.0836, "step": 386 }, { "epoch": 1.6556149732620322, "grad_norm": 2.1002938747406006, "learning_rate": 4.161282057231776e-06, "loss": 0.086, "step": 387 }, { "epoch": 1.6598930481283423, "grad_norm": 2.04650616645813, "learning_rate": 4.13913696642286e-06, "loss": 0.0878, "step": 388 }, { "epoch": 1.6641711229946523, "grad_norm": 1.8128232955932617, "learning_rate": 4.1170092647786895e-06, "loss": 0.0836, "step": 389 }, { "epoch": 1.6684491978609626, "grad_norm": 2.1174373626708984, "learning_rate": 4.094899399271935e-06, "loss": 0.0964, "step": 390 }, { "epoch": 1.6727272727272728, "grad_norm": 2.191866397857666, "learning_rate": 4.072807816514978e-06, "loss": 0.107, "step": 391 }, { "epoch": 1.6770053475935829, "grad_norm": 2.061434507369995, "learning_rate": 4.0507349627509045e-06, "loss": 0.0876, "step": 392 }, { "epoch": 1.681283422459893, "grad_norm": 2.02978253364563, "learning_rate": 4.028681283844471e-06, "loss": 0.0844, "step": 393 }, { "epoch": 1.6855614973262032, "grad_norm": 2.3137338161468506, "learning_rate": 4.006647225273116e-06, "loss": 0.0956, "step": 394 }, { "epoch": 1.6898395721925135, "grad_norm": 2.265906810760498, "learning_rate": 3.984633232117948e-06, "loss": 0.085, "step": 395 }, { "epoch": 1.6941176470588235, "grad_norm": 2.138777017593384, "learning_rate": 3.96263974905476e-06, "loss": 0.0881, "step": 396 }, { "epoch": 1.6983957219251336, "grad_norm": 2.1694891452789307, "learning_rate": 3.94066722034505e-06, "loss": 0.0968, "step": 397 }, { "epoch": 1.7026737967914438, "grad_norm": 1.892051339149475, "learning_rate": 3.9187160898270435e-06, "loss": 0.0941, "step": 398 }, { "epoch": 1.706951871657754, "grad_norm": 2.209155321121216, "learning_rate": 3.896786800906734e-06, "loss": 0.0891, "step": 399 }, { "epoch": 1.7112299465240641, "grad_norm": 2.2992265224456787, "learning_rate": 3.87487979654891e-06, "loss": 0.0928, "step": 400 }, { "epoch": 1.7155080213903742, "grad_norm": 2.0164313316345215, "learning_rate": 3.852995519268231e-06, "loss": 0.1001, "step": 401 }, { "epoch": 1.7197860962566844, "grad_norm": 2.269282102584839, "learning_rate": 3.831134411120273e-06, "loss": 0.0952, "step": 402 }, { "epoch": 1.7240641711229947, "grad_norm": 2.073052406311035, "learning_rate": 3.809296913692594e-06, "loss": 0.0811, "step": 403 }, { "epoch": 1.728342245989305, "grad_norm": 1.98734712600708, "learning_rate": 3.787483468095838e-06, "loss": 0.0987, "step": 404 }, { "epoch": 1.732620320855615, "grad_norm": 1.9203572273254395, "learning_rate": 3.765694514954796e-06, "loss": 0.084, "step": 405 }, { "epoch": 1.736898395721925, "grad_norm": 2.080029249191284, "learning_rate": 3.7439304943995274e-06, "loss": 0.0896, "step": 406 }, { "epoch": 1.7411764705882353, "grad_norm": 2.0088043212890625, "learning_rate": 3.72219184605646e-06, "loss": 0.0872, "step": 407 }, { "epoch": 1.7454545454545456, "grad_norm": 2.2481374740600586, "learning_rate": 3.7004790090395043e-06, "loss": 0.1012, "step": 408 }, { "epoch": 1.7497326203208556, "grad_norm": 2.1036605834960938, "learning_rate": 3.678792421941199e-06, "loss": 0.0926, "step": 409 }, { "epoch": 1.7540106951871657, "grad_norm": 2.260814666748047, "learning_rate": 3.657132522823837e-06, "loss": 0.1097, "step": 410 }, { "epoch": 1.758288770053476, "grad_norm": 2.0131592750549316, "learning_rate": 3.6354997492106258e-06, "loss": 0.086, "step": 411 }, { "epoch": 1.7625668449197862, "grad_norm": 1.936078667640686, "learning_rate": 3.6138945380768442e-06, "loss": 0.0808, "step": 412 }, { "epoch": 1.7668449197860963, "grad_norm": 2.0518951416015625, "learning_rate": 3.592317325841014e-06, "loss": 0.0929, "step": 413 }, { "epoch": 1.7711229946524063, "grad_norm": 2.1253445148468018, "learning_rate": 3.5707685483560948e-06, "loss": 0.1027, "step": 414 }, { "epoch": 1.7754010695187166, "grad_norm": 2.0566675662994385, "learning_rate": 3.5492486409006684e-06, "loss": 0.0875, "step": 415 }, { "epoch": 1.7796791443850268, "grad_norm": 1.9728094339370728, "learning_rate": 3.5277580381701553e-06, "loss": 0.08, "step": 416 }, { "epoch": 1.7839572192513369, "grad_norm": 2.156283140182495, "learning_rate": 3.5062971742680244e-06, "loss": 0.0887, "step": 417 }, { "epoch": 1.788235294117647, "grad_norm": 2.1552810668945312, "learning_rate": 3.484866482697032e-06, "loss": 0.0863, "step": 418 }, { "epoch": 1.7925133689839572, "grad_norm": 2.292252779006958, "learning_rate": 3.4634663963504654e-06, "loss": 0.084, "step": 419 }, { "epoch": 1.7967914438502675, "grad_norm": 2.240377187728882, "learning_rate": 3.4420973475033894e-06, "loss": 0.1, "step": 420 }, { "epoch": 1.8010695187165775, "grad_norm": 2.1546454429626465, "learning_rate": 3.4207597678039293e-06, "loss": 0.0797, "step": 421 }, { "epoch": 1.8053475935828875, "grad_norm": 2.0921828746795654, "learning_rate": 3.3994540882645353e-06, "loss": 0.0901, "step": 422 }, { "epoch": 1.8096256684491978, "grad_norm": 1.9114705324172974, "learning_rate": 3.3781807392532893e-06, "loss": 0.0775, "step": 423 }, { "epoch": 1.813903743315508, "grad_norm": 2.1772425174713135, "learning_rate": 3.3569401504852073e-06, "loss": 0.0891, "step": 424 }, { "epoch": 1.8181818181818183, "grad_norm": 1.884969711303711, "learning_rate": 3.335732751013553e-06, "loss": 0.0922, "step": 425 }, { "epoch": 1.8224598930481284, "grad_norm": 1.8992670774459839, "learning_rate": 3.3145589692211837e-06, "loss": 0.0968, "step": 426 }, { "epoch": 1.8267379679144384, "grad_norm": 1.957726240158081, "learning_rate": 3.2934192328118866e-06, "loss": 0.0896, "step": 427 }, { "epoch": 1.8310160427807487, "grad_norm": 2.0114240646362305, "learning_rate": 3.27231396880174e-06, "loss": 0.1016, "step": 428 }, { "epoch": 1.835294117647059, "grad_norm": 2.00063157081604, "learning_rate": 3.2512436035104968e-06, "loss": 0.083, "step": 429 }, { "epoch": 1.839572192513369, "grad_norm": 2.367206335067749, "learning_rate": 3.2302085625529596e-06, "loss": 0.0948, "step": 430 }, { "epoch": 1.843850267379679, "grad_norm": 2.1237783432006836, "learning_rate": 3.2092092708303973e-06, "loss": 0.0821, "step": 431 }, { "epoch": 1.8481283422459893, "grad_norm": 2.1839144229888916, "learning_rate": 3.18824615252195e-06, "loss": 0.0977, "step": 432 }, { "epoch": 1.8524064171122996, "grad_norm": 2.173175811767578, "learning_rate": 3.1673196310760723e-06, "loss": 0.0994, "step": 433 }, { "epoch": 1.8566844919786096, "grad_norm": 2.2018539905548096, "learning_rate": 3.146430129201965e-06, "loss": 0.0938, "step": 434 }, { "epoch": 1.8609625668449197, "grad_norm": 1.9908114671707153, "learning_rate": 3.125578068861051e-06, "loss": 0.0846, "step": 435 }, { "epoch": 1.86524064171123, "grad_norm": 2.0504186153411865, "learning_rate": 3.104763871258447e-06, "loss": 0.0838, "step": 436 }, { "epoch": 1.8695187165775402, "grad_norm": 1.9001001119613647, "learning_rate": 3.083987956834449e-06, "loss": 0.0801, "step": 437 }, { "epoch": 1.8737967914438503, "grad_norm": 2.0107505321502686, "learning_rate": 3.06325074525605e-06, "loss": 0.0908, "step": 438 }, { "epoch": 1.8780748663101603, "grad_norm": 1.8863651752471924, "learning_rate": 3.0425526554084526e-06, "loss": 0.0771, "step": 439 }, { "epoch": 1.8823529411764706, "grad_norm": 2.1086320877075195, "learning_rate": 3.0218941053866167e-06, "loss": 0.0819, "step": 440 }, { "epoch": 1.8866310160427808, "grad_norm": 1.9322139024734497, "learning_rate": 3.00127551248681e-06, "loss": 0.0905, "step": 441 }, { "epoch": 1.8909090909090909, "grad_norm": 2.0057458877563477, "learning_rate": 2.980697293198174e-06, "loss": 0.0849, "step": 442 }, { "epoch": 1.895187165775401, "grad_norm": 2.083625316619873, "learning_rate": 2.960159863194322e-06, "loss": 0.0957, "step": 443 }, { "epoch": 1.8994652406417112, "grad_norm": 1.8596922159194946, "learning_rate": 2.939663637324934e-06, "loss": 0.0954, "step": 444 }, { "epoch": 1.9037433155080214, "grad_norm": 1.8501924276351929, "learning_rate": 2.9192090296073755e-06, "loss": 0.0752, "step": 445 }, { "epoch": 1.9080213903743317, "grad_norm": 2.0312633514404297, "learning_rate": 2.8987964532183454e-06, "loss": 0.0855, "step": 446 }, { "epoch": 1.9122994652406418, "grad_norm": 2.077679395675659, "learning_rate": 2.878426320485518e-06, "loss": 0.099, "step": 447 }, { "epoch": 1.9165775401069518, "grad_norm": 1.9612921476364136, "learning_rate": 2.8580990428792205e-06, "loss": 0.0758, "step": 448 }, { "epoch": 1.920855614973262, "grad_norm": 1.9715369939804077, "learning_rate": 2.8378150310041197e-06, "loss": 0.0765, "step": 449 }, { "epoch": 1.9251336898395723, "grad_norm": 2.0577006340026855, "learning_rate": 2.8175746945909277e-06, "loss": 0.0968, "step": 450 }, { "epoch": 1.9294117647058824, "grad_norm": 1.8270310163497925, "learning_rate": 2.7973784424881273e-06, "loss": 0.0829, "step": 451 }, { "epoch": 1.9336898395721924, "grad_norm": 2.1296803951263428, "learning_rate": 2.7772266826537103e-06, "loss": 0.0996, "step": 452 }, { "epoch": 1.9379679144385027, "grad_norm": 1.9016512632369995, "learning_rate": 2.75711982214694e-06, "loss": 0.0858, "step": 453 }, { "epoch": 1.942245989304813, "grad_norm": 2.4038286209106445, "learning_rate": 2.7370582671201253e-06, "loss": 0.0963, "step": 454 }, { "epoch": 1.946524064171123, "grad_norm": 2.015941619873047, "learning_rate": 2.7170424228104207e-06, "loss": 0.0811, "step": 455 }, { "epoch": 1.950802139037433, "grad_norm": 2.00461745262146, "learning_rate": 2.697072693531637e-06, "loss": 0.08, "step": 456 }, { "epoch": 1.9550802139037433, "grad_norm": 1.9232676029205322, "learning_rate": 2.6771494826660782e-06, "loss": 0.0793, "step": 457 }, { "epoch": 1.9593582887700536, "grad_norm": 2.0206451416015625, "learning_rate": 2.6572731926563867e-06, "loss": 0.0898, "step": 458 }, { "epoch": 1.9636363636363636, "grad_norm": 2.0119080543518066, "learning_rate": 2.6374442249974214e-06, "loss": 0.0915, "step": 459 }, { "epoch": 1.9679144385026737, "grad_norm": 1.98100745677948, "learning_rate": 2.617662980228144e-06, "loss": 0.0777, "step": 460 }, { "epoch": 1.972192513368984, "grad_norm": 2.208132743835449, "learning_rate": 2.5979298579235276e-06, "loss": 0.1013, "step": 461 }, { "epoch": 1.9764705882352942, "grad_norm": 2.1247506141662598, "learning_rate": 2.578245256686488e-06, "loss": 0.0858, "step": 462 }, { "epoch": 1.9807486631016042, "grad_norm": 2.1127898693084717, "learning_rate": 2.558609574139829e-06, "loss": 0.0904, "step": 463 }, { "epoch": 1.9850267379679143, "grad_norm": 1.8930281400680542, "learning_rate": 2.539023206918212e-06, "loss": 0.0872, "step": 464 }, { "epoch": 1.9893048128342246, "grad_norm": 1.8810979127883911, "learning_rate": 2.5194865506601507e-06, "loss": 0.0843, "step": 465 }, { "epoch": 1.9935828877005348, "grad_norm": 2.0668907165527344, "learning_rate": 2.5000000000000015e-06, "loss": 0.0936, "step": 466 }, { "epoch": 1.9978609625668449, "grad_norm": 2.309532880783081, "learning_rate": 2.4805639485600087e-06, "loss": 0.0989, "step": 467 }, { "epoch": 2.002139037433155, "grad_norm": 4.147768497467041, "learning_rate": 2.4611787889423546e-06, "loss": 0.0848, "step": 468 }, { "epoch": 2.006417112299465, "grad_norm": 1.6421159505844116, "learning_rate": 2.441844912721209e-06, "loss": 0.057, "step": 469 }, { "epoch": 2.0106951871657754, "grad_norm": 1.4409747123718262, "learning_rate": 2.422562710434848e-06, "loss": 0.048, "step": 470 }, { "epoch": 2.0149732620320857, "grad_norm": 1.4703036546707153, "learning_rate": 2.403332571577738e-06, "loss": 0.0477, "step": 471 }, { "epoch": 2.0192513368983955, "grad_norm": 1.6000310182571411, "learning_rate": 2.3841548845926844e-06, "loss": 0.0568, "step": 472 }, { "epoch": 2.023529411764706, "grad_norm": 1.4772285223007202, "learning_rate": 2.365030036862988e-06, "loss": 0.0476, "step": 473 }, { "epoch": 2.027807486631016, "grad_norm": 1.7068934440612793, "learning_rate": 2.3459584147046e-06, "loss": 0.0545, "step": 474 }, { "epoch": 2.0320855614973263, "grad_norm": 1.4684138298034668, "learning_rate": 2.3269404033583443e-06, "loss": 0.0461, "step": 475 }, { "epoch": 2.036363636363636, "grad_norm": 1.6764367818832397, "learning_rate": 2.3079763869821176e-06, "loss": 0.0555, "step": 476 }, { "epoch": 2.0406417112299464, "grad_norm": 1.8675357103347778, "learning_rate": 2.2890667486431296e-06, "loss": 0.0553, "step": 477 }, { "epoch": 2.0449197860962567, "grad_norm": 1.8676857948303223, "learning_rate": 2.270211870310184e-06, "loss": 0.0491, "step": 478 }, { "epoch": 2.049197860962567, "grad_norm": 1.6475131511688232, "learning_rate": 2.251412132845933e-06, "loss": 0.0437, "step": 479 }, { "epoch": 2.053475935828877, "grad_norm": 1.9799823760986328, "learning_rate": 2.232667915999216e-06, "loss": 0.0498, "step": 480 }, { "epoch": 2.057754010695187, "grad_norm": 1.9527326822280884, "learning_rate": 2.2139795983973654e-06, "loss": 0.0487, "step": 481 }, { "epoch": 2.0620320855614973, "grad_norm": 2.1700730323791504, "learning_rate": 2.1953475575385618e-06, "loss": 0.0475, "step": 482 }, { "epoch": 2.0663101604278076, "grad_norm": 1.9394606351852417, "learning_rate": 2.1767721697842244e-06, "loss": 0.0541, "step": 483 }, { "epoch": 2.070588235294118, "grad_norm": 1.7941431999206543, "learning_rate": 2.1582538103513896e-06, "loss": 0.047, "step": 484 }, { "epoch": 2.0748663101604277, "grad_norm": 2.110564947128296, "learning_rate": 2.139792853305141e-06, "loss": 0.0501, "step": 485 }, { "epoch": 2.079144385026738, "grad_norm": 2.1480815410614014, "learning_rate": 2.121389671551054e-06, "loss": 0.0498, "step": 486 }, { "epoch": 2.083422459893048, "grad_norm": 2.0997166633605957, "learning_rate": 2.1030446368276547e-06, "loss": 0.0521, "step": 487 }, { "epoch": 2.0877005347593585, "grad_norm": 2.176145553588867, "learning_rate": 2.0847581196989277e-06, "loss": 0.057, "step": 488 }, { "epoch": 2.0919786096256683, "grad_norm": 1.9567787647247314, "learning_rate": 2.0665304895468114e-06, "loss": 0.0442, "step": 489 }, { "epoch": 2.0962566844919786, "grad_norm": 2.0823962688446045, "learning_rate": 2.04836211456375e-06, "loss": 0.0472, "step": 490 }, { "epoch": 2.100534759358289, "grad_norm": 1.7501331567764282, "learning_rate": 2.030253361745251e-06, "loss": 0.0483, "step": 491 }, { "epoch": 2.104812834224599, "grad_norm": 2.014772415161133, "learning_rate": 2.012204596882472e-06, "loss": 0.046, "step": 492 }, { "epoch": 2.109090909090909, "grad_norm": 1.8758680820465088, "learning_rate": 1.9942161845548334e-06, "loss": 0.0485, "step": 493 }, { "epoch": 2.113368983957219, "grad_norm": 2.0299036502838135, "learning_rate": 1.9762884881226535e-06, "loss": 0.0548, "step": 494 }, { "epoch": 2.1176470588235294, "grad_norm": 1.911232352256775, "learning_rate": 1.958421869719807e-06, "loss": 0.0486, "step": 495 }, { "epoch": 2.1219251336898397, "grad_norm": 1.9703571796417236, "learning_rate": 1.9406166902464128e-06, "loss": 0.0448, "step": 496 }, { "epoch": 2.1262032085561495, "grad_norm": 2.0307888984680176, "learning_rate": 1.922873309361542e-06, "loss": 0.0403, "step": 497 }, { "epoch": 2.13048128342246, "grad_norm": 2.503244161605835, "learning_rate": 1.9051920854759543e-06, "loss": 0.0594, "step": 498 }, { "epoch": 2.13475935828877, "grad_norm": 2.0204391479492188, "learning_rate": 1.887573375744856e-06, "loss": 0.0483, "step": 499 }, { "epoch": 2.1390374331550803, "grad_norm": 1.822661280632019, "learning_rate": 1.8700175360606882e-06, "loss": 0.0469, "step": 500 }, { "epoch": 2.1433155080213906, "grad_norm": 1.8638768196105957, "learning_rate": 1.8525249210459345e-06, "loss": 0.0443, "step": 501 }, { "epoch": 2.1475935828877004, "grad_norm": 2.0525190830230713, "learning_rate": 1.8350958840459665e-06, "loss": 0.058, "step": 502 }, { "epoch": 2.1518716577540107, "grad_norm": 2.026292562484741, "learning_rate": 1.8177307771218894e-06, "loss": 0.0398, "step": 503 }, { "epoch": 2.156149732620321, "grad_norm": 1.8052258491516113, "learning_rate": 1.8004299510434493e-06, "loss": 0.04, "step": 504 }, { "epoch": 2.160427807486631, "grad_norm": 1.9584407806396484, "learning_rate": 1.7831937552819345e-06, "loss": 0.0608, "step": 505 }, { "epoch": 2.164705882352941, "grad_norm": 1.9575603008270264, "learning_rate": 1.766022538003122e-06, "loss": 0.0532, "step": 506 }, { "epoch": 2.1689839572192513, "grad_norm": 1.8743293285369873, "learning_rate": 1.7489166460602496e-06, "loss": 0.0449, "step": 507 }, { "epoch": 2.1732620320855616, "grad_norm": 1.9783152341842651, "learning_rate": 1.7318764249869934e-06, "loss": 0.0499, "step": 508 }, { "epoch": 2.177540106951872, "grad_norm": 1.86168372631073, "learning_rate": 1.7149022189905041e-06, "loss": 0.0555, "step": 509 }, { "epoch": 2.1818181818181817, "grad_norm": 1.8643457889556885, "learning_rate": 1.697994370944452e-06, "loss": 0.0479, "step": 510 }, { "epoch": 2.186096256684492, "grad_norm": 1.849522352218628, "learning_rate": 1.6811532223820875e-06, "loss": 0.0552, "step": 511 }, { "epoch": 2.190374331550802, "grad_norm": 1.8427163362503052, "learning_rate": 1.6643791134893644e-06, "loss": 0.0514, "step": 512 }, { "epoch": 2.1946524064171125, "grad_norm": 1.877435564994812, "learning_rate": 1.6476723830980451e-06, "loss": 0.0572, "step": 513 }, { "epoch": 2.1989304812834223, "grad_norm": 1.8627737760543823, "learning_rate": 1.631033368678872e-06, "loss": 0.0482, "step": 514 }, { "epoch": 2.2032085561497325, "grad_norm": 1.6647427082061768, "learning_rate": 1.6144624063347514e-06, "loss": 0.0459, "step": 515 }, { "epoch": 2.207486631016043, "grad_norm": 2.0262389183044434, "learning_rate": 1.597959830793947e-06, "loss": 0.061, "step": 516 }, { "epoch": 2.211764705882353, "grad_norm": 1.9522355794906616, "learning_rate": 1.5815259754033407e-06, "loss": 0.0517, "step": 517 }, { "epoch": 2.216042780748663, "grad_norm": 1.7932432889938354, "learning_rate": 1.5651611721216865e-06, "loss": 0.0449, "step": 518 }, { "epoch": 2.220320855614973, "grad_norm": 1.9433608055114746, "learning_rate": 1.5488657515129001e-06, "loss": 0.0508, "step": 519 }, { "epoch": 2.2245989304812834, "grad_norm": 1.8138562440872192, "learning_rate": 1.5326400427394023e-06, "loss": 0.0445, "step": 520 }, { "epoch": 2.2288770053475937, "grad_norm": 2.202510356903076, "learning_rate": 1.5164843735554408e-06, "loss": 0.056, "step": 521 }, { "epoch": 2.233155080213904, "grad_norm": 1.8365262746810913, "learning_rate": 1.5003990703004994e-06, "loss": 0.0474, "step": 522 }, { "epoch": 2.237433155080214, "grad_norm": 2.0539376735687256, "learning_rate": 1.4843844578926863e-06, "loss": 0.055, "step": 523 }, { "epoch": 2.241711229946524, "grad_norm": 1.922347903251648, "learning_rate": 1.4684408598221722e-06, "loss": 0.0459, "step": 524 }, { "epoch": 2.2459893048128343, "grad_norm": 1.8658438920974731, "learning_rate": 1.452568598144668e-06, "loss": 0.0409, "step": 525 }, { "epoch": 2.250267379679144, "grad_norm": 1.8378267288208008, "learning_rate": 1.4367679934749085e-06, "loss": 0.05, "step": 526 }, { "epoch": 2.2545454545454544, "grad_norm": 2.15887188911438, "learning_rate": 1.421039364980178e-06, "loss": 0.0456, "step": 527 }, { "epoch": 2.2588235294117647, "grad_norm": 1.7189967632293701, "learning_rate": 1.405383030373867e-06, "loss": 0.044, "step": 528 }, { "epoch": 2.263101604278075, "grad_norm": 1.6648989915847778, "learning_rate": 1.3897993059090492e-06, "loss": 0.0428, "step": 529 }, { "epoch": 2.267379679144385, "grad_norm": 1.7326384782791138, "learning_rate": 1.374288506372099e-06, "loss": 0.0469, "step": 530 }, { "epoch": 2.271657754010695, "grad_norm": 1.892421841621399, "learning_rate": 1.3588509450763281e-06, "loss": 0.0384, "step": 531 }, { "epoch": 2.2759358288770053, "grad_norm": 1.907393217086792, "learning_rate": 1.3434869338556594e-06, "loss": 0.0396, "step": 532 }, { "epoch": 2.2802139037433156, "grad_norm": 1.8455674648284912, "learning_rate": 1.3281967830583264e-06, "loss": 0.0403, "step": 533 }, { "epoch": 2.284491978609626, "grad_norm": 1.8442206382751465, "learning_rate": 1.3129808015406064e-06, "loss": 0.0435, "step": 534 }, { "epoch": 2.2887700534759357, "grad_norm": 1.8563473224639893, "learning_rate": 1.297839296660579e-06, "loss": 0.0445, "step": 535 }, { "epoch": 2.293048128342246, "grad_norm": 1.820439100265503, "learning_rate": 1.2827725742719205e-06, "loss": 0.0447, "step": 536 }, { "epoch": 2.297326203208556, "grad_norm": 1.8624792098999023, "learning_rate": 1.267780938717722e-06, "loss": 0.0468, "step": 537 }, { "epoch": 2.3016042780748664, "grad_norm": 1.8183844089508057, "learning_rate": 1.252864692824346e-06, "loss": 0.0473, "step": 538 }, { "epoch": 2.3058823529411763, "grad_norm": 1.8453257083892822, "learning_rate": 1.2380241378953067e-06, "loss": 0.0449, "step": 539 }, { "epoch": 2.3101604278074865, "grad_norm": 1.7943516969680786, "learning_rate": 1.223259573705184e-06, "loss": 0.0432, "step": 540 }, { "epoch": 2.314438502673797, "grad_norm": 1.7877850532531738, "learning_rate": 1.2085712984935693e-06, "loss": 0.0497, "step": 541 }, { "epoch": 2.318716577540107, "grad_norm": 1.7979768514633179, "learning_rate": 1.1939596089590394e-06, "loss": 0.0415, "step": 542 }, { "epoch": 2.3229946524064173, "grad_norm": 1.8382468223571777, "learning_rate": 1.1794248002531644e-06, "loss": 0.0492, "step": 543 }, { "epoch": 2.327272727272727, "grad_norm": 1.638438105583191, "learning_rate": 1.1649671659745504e-06, "loss": 0.0469, "step": 544 }, { "epoch": 2.3315508021390374, "grad_norm": 2.0724680423736572, "learning_rate": 1.1505869981628953e-06, "loss": 0.0546, "step": 545 }, { "epoch": 2.3358288770053477, "grad_norm": 1.7388609647750854, "learning_rate": 1.1362845872931044e-06, "loss": 0.0472, "step": 546 }, { "epoch": 2.3401069518716575, "grad_norm": 1.8055834770202637, "learning_rate": 1.1220602222694166e-06, "loss": 0.0495, "step": 547 }, { "epoch": 2.344385026737968, "grad_norm": 1.8614943027496338, "learning_rate": 1.1079141904195662e-06, "loss": 0.0441, "step": 548 }, { "epoch": 2.348663101604278, "grad_norm": 1.6754151582717896, "learning_rate": 1.0938467774889883e-06, "loss": 0.0402, "step": 549 }, { "epoch": 2.3529411764705883, "grad_norm": 1.8570998907089233, "learning_rate": 1.0798582676350316e-06, "loss": 0.0513, "step": 550 }, { "epoch": 2.3572192513368986, "grad_norm": 1.6499812602996826, "learning_rate": 1.0659489434212323e-06, "loss": 0.0432, "step": 551 }, { "epoch": 2.3614973262032084, "grad_norm": 1.8310068845748901, "learning_rate": 1.0521190858116042e-06, "loss": 0.0417, "step": 552 }, { "epoch": 2.3657754010695187, "grad_norm": 1.8544480800628662, "learning_rate": 1.0383689741649516e-06, "loss": 0.0476, "step": 553 }, { "epoch": 2.370053475935829, "grad_norm": 2.061567544937134, "learning_rate": 1.0246988862292462e-06, "loss": 0.0522, "step": 554 }, { "epoch": 2.374331550802139, "grad_norm": 1.7920609712600708, "learning_rate": 1.0111090981359961e-06, "loss": 0.0514, "step": 555 }, { "epoch": 2.378609625668449, "grad_norm": 1.9858516454696655, "learning_rate": 9.975998843946811e-07, "loss": 0.0477, "step": 556 }, { "epoch": 2.3828877005347593, "grad_norm": 1.7108044624328613, "learning_rate": 9.841715178872092e-07, "loss": 0.0417, "step": 557 }, { "epoch": 2.3871657754010696, "grad_norm": 1.7552218437194824, "learning_rate": 9.708242698623898e-07, "loss": 0.0542, "step": 558 }, { "epoch": 2.39144385026738, "grad_norm": 1.790418028831482, "learning_rate": 9.575584099304735e-07, "loss": 0.0505, "step": 559 }, { "epoch": 2.3957219251336896, "grad_norm": 2.0047428607940674, "learning_rate": 9.443742060576916e-07, "loss": 0.0529, "step": 560 }, { "epoch": 2.4, "grad_norm": 1.7387304306030273, "learning_rate": 9.312719245608487e-07, "loss": 0.0445, "step": 561 }, { "epoch": 2.40427807486631, "grad_norm": 1.5612061023712158, "learning_rate": 9.182518301019466e-07, "loss": 0.0419, "step": 562 }, { "epoch": 2.4085561497326204, "grad_norm": 1.8490066528320312, "learning_rate": 9.053141856828274e-07, "loss": 0.0463, "step": 563 }, { "epoch": 2.4128342245989307, "grad_norm": 1.8447004556655884, "learning_rate": 8.924592526398762e-07, "loss": 0.0544, "step": 564 }, { "epoch": 2.4171122994652405, "grad_norm": 1.7815872430801392, "learning_rate": 8.796872906387299e-07, "loss": 0.0401, "step": 565 }, { "epoch": 2.421390374331551, "grad_norm": 1.815962314605713, "learning_rate": 8.669985576690371e-07, "loss": 0.0465, "step": 566 }, { "epoch": 2.425668449197861, "grad_norm": 1.73190438747406, "learning_rate": 8.54393310039246e-07, "loss": 0.0487, "step": 567 }, { "epoch": 2.429946524064171, "grad_norm": 1.7489674091339111, "learning_rate": 8.418718023714235e-07, "loss": 0.0474, "step": 568 }, { "epoch": 2.434224598930481, "grad_norm": 1.8730748891830444, "learning_rate": 8.29434287596122e-07, "loss": 0.0495, "step": 569 }, { "epoch": 2.4385026737967914, "grad_norm": 1.7025930881500244, "learning_rate": 8.170810169472593e-07, "loss": 0.0454, "step": 570 }, { "epoch": 2.4427807486631017, "grad_norm": 1.7497166395187378, "learning_rate": 8.04812239957049e-07, "loss": 0.0389, "step": 571 }, { "epoch": 2.447058823529412, "grad_norm": 1.9598861932754517, "learning_rate": 7.926282044509593e-07, "loss": 0.0487, "step": 572 }, { "epoch": 2.4513368983957218, "grad_norm": 1.6827856302261353, "learning_rate": 7.805291565427065e-07, "loss": 0.041, "step": 573 }, { "epoch": 2.455614973262032, "grad_norm": 1.7500789165496826, "learning_rate": 7.685153406292845e-07, "loss": 0.048, "step": 574 }, { "epoch": 2.4598930481283423, "grad_norm": 1.8886414766311646, "learning_rate": 7.56586999386027e-07, "loss": 0.0496, "step": 575 }, { "epoch": 2.4641711229946526, "grad_norm": 1.8937726020812988, "learning_rate": 7.447443737617066e-07, "loss": 0.0405, "step": 576 }, { "epoch": 2.4684491978609624, "grad_norm": 1.953205943107605, "learning_rate": 7.329877029736665e-07, "loss": 0.0476, "step": 577 }, { "epoch": 2.4727272727272727, "grad_norm": 1.8571252822875977, "learning_rate": 7.213172245029892e-07, "loss": 0.0495, "step": 578 }, { "epoch": 2.477005347593583, "grad_norm": 1.9600136280059814, "learning_rate": 7.097331740896995e-07, "loss": 0.0407, "step": 579 }, { "epoch": 2.481283422459893, "grad_norm": 2.01959490776062, "learning_rate": 6.98235785728002e-07, "loss": 0.0487, "step": 580 }, { "epoch": 2.485561497326203, "grad_norm": 1.9001308679580688, "learning_rate": 6.868252916615553e-07, "loss": 0.0465, "step": 581 }, { "epoch": 2.4898395721925133, "grad_norm": 1.7344181537628174, "learning_rate": 6.755019223787807e-07, "loss": 0.0472, "step": 582 }, { "epoch": 2.4941176470588236, "grad_norm": 2.2562477588653564, "learning_rate": 6.642659066082046e-07, "loss": 0.0558, "step": 583 }, { "epoch": 2.498395721925134, "grad_norm": 1.99720299243927, "learning_rate": 6.531174713138416e-07, "loss": 0.0574, "step": 584 }, { "epoch": 2.502673796791444, "grad_norm": 1.6997674703598022, "learning_rate": 6.420568416906059e-07, "loss": 0.0407, "step": 585 }, { "epoch": 2.506951871657754, "grad_norm": 1.9574675559997559, "learning_rate": 6.310842411597667e-07, "loss": 0.0502, "step": 586 }, { "epoch": 2.511229946524064, "grad_norm": 1.991170883178711, "learning_rate": 6.201998913644319e-07, "loss": 0.0462, "step": 587 }, { "epoch": 2.5155080213903744, "grad_norm": 1.9040491580963135, "learning_rate": 6.094040121650719e-07, "loss": 0.0537, "step": 588 }, { "epoch": 2.5197860962566843, "grad_norm": 1.7386451959609985, "learning_rate": 5.986968216350786e-07, "loss": 0.0493, "step": 589 }, { "epoch": 2.5240641711229945, "grad_norm": 1.9728233814239502, "learning_rate": 5.880785360563596e-07, "loss": 0.0449, "step": 590 }, { "epoch": 2.528342245989305, "grad_norm": 1.965031623840332, "learning_rate": 5.775493699149754e-07, "loss": 0.0422, "step": 591 }, { "epoch": 2.532620320855615, "grad_norm": 1.7499405145645142, "learning_rate": 5.671095358967926e-07, "loss": 0.043, "step": 592 }, { "epoch": 2.5368983957219253, "grad_norm": 1.814707636833191, "learning_rate": 5.56759244883206e-07, "loss": 0.0517, "step": 593 }, { "epoch": 2.541176470588235, "grad_norm": 1.7832578420639038, "learning_rate": 5.464987059468629e-07, "loss": 0.0432, "step": 594 }, { "epoch": 2.5454545454545454, "grad_norm": 1.9133834838867188, "learning_rate": 5.36328126347449e-07, "loss": 0.0469, "step": 595 }, { "epoch": 2.5497326203208557, "grad_norm": 1.5707026720046997, "learning_rate": 5.262477115275022e-07, "loss": 0.0376, "step": 596 }, { "epoch": 2.5540106951871655, "grad_norm": 1.8033281564712524, "learning_rate": 5.162576651082541e-07, "loss": 0.042, "step": 597 }, { "epoch": 2.558288770053476, "grad_norm": 1.8233810663223267, "learning_rate": 5.063581888855285e-07, "loss": 0.0506, "step": 598 }, { "epoch": 2.562566844919786, "grad_norm": 1.5576063394546509, "learning_rate": 4.965494828256573e-07, "loss": 0.0352, "step": 599 }, { "epoch": 2.5668449197860963, "grad_norm": 1.9136029481887817, "learning_rate": 4.868317450614407e-07, "loss": 0.051, "step": 600 }, { "epoch": 2.5711229946524066, "grad_norm": 1.714444637298584, "learning_rate": 4.772051718881532e-07, "loss": 0.0509, "step": 601 }, { "epoch": 2.5754010695187164, "grad_norm": 1.6594096422195435, "learning_rate": 4.676699577595667e-07, "loss": 0.0423, "step": 602 }, { "epoch": 2.5796791443850267, "grad_norm": 1.8420310020446777, "learning_rate": 4.582262952840355e-07, "loss": 0.0423, "step": 603 }, { "epoch": 2.583957219251337, "grad_norm": 1.8472838401794434, "learning_rate": 4.4887437522059487e-07, "loss": 0.0494, "step": 604 }, { "epoch": 2.588235294117647, "grad_norm": 1.652789831161499, "learning_rate": 4.3961438647511066e-07, "loss": 0.0444, "step": 605 }, { "epoch": 2.5925133689839575, "grad_norm": 1.6729520559310913, "learning_rate": 4.304465160964699e-07, "loss": 0.0452, "step": 606 }, { "epoch": 2.5967914438502673, "grad_norm": 1.744144082069397, "learning_rate": 4.2137094927279296e-07, "loss": 0.0426, "step": 607 }, { "epoch": 2.6010695187165775, "grad_norm": 2.081646680831909, "learning_rate": 4.1238786932769947e-07, "loss": 0.0492, "step": 608 }, { "epoch": 2.605347593582888, "grad_norm": 1.9060657024383545, "learning_rate": 4.0349745771660233e-07, "loss": 0.0438, "step": 609 }, { "epoch": 2.6096256684491976, "grad_norm": 1.9777454137802124, "learning_rate": 3.946998940230401e-07, "loss": 0.056, "step": 610 }, { "epoch": 2.613903743315508, "grad_norm": 1.8688209056854248, "learning_rate": 3.859953559550589e-07, "loss": 0.0392, "step": 611 }, { "epoch": 2.618181818181818, "grad_norm": 1.7810322046279907, "learning_rate": 3.7738401934161006e-07, "loss": 0.0501, "step": 612 }, { "epoch": 2.6224598930481284, "grad_norm": 1.5421664714813232, "learning_rate": 3.6886605812900766e-07, "loss": 0.0307, "step": 613 }, { "epoch": 2.6267379679144387, "grad_norm": 1.5893266201019287, "learning_rate": 3.604416443774117e-07, "loss": 0.0405, "step": 614 }, { "epoch": 2.6310160427807485, "grad_norm": 1.7493019104003906, "learning_rate": 3.5211094825735147e-07, "loss": 0.0445, "step": 615 }, { "epoch": 2.635294117647059, "grad_norm": 1.8685232400894165, "learning_rate": 3.4387413804628955e-07, "loss": 0.0483, "step": 616 }, { "epoch": 2.639572192513369, "grad_norm": 2.100968837738037, "learning_rate": 3.357313801252238e-07, "loss": 0.0452, "step": 617 }, { "epoch": 2.643850267379679, "grad_norm": 1.8143876791000366, "learning_rate": 3.276828389753234e-07, "loss": 0.044, "step": 618 }, { "epoch": 2.648128342245989, "grad_norm": 1.9584417343139648, "learning_rate": 3.197286771746094e-07, "loss": 0.0482, "step": 619 }, { "epoch": 2.6524064171122994, "grad_norm": 1.6974005699157715, "learning_rate": 3.118690553946685e-07, "loss": 0.0443, "step": 620 }, { "epoch": 2.6566844919786097, "grad_norm": 2.060076951980591, "learning_rate": 3.041041323974098e-07, "loss": 0.0478, "step": 621 }, { "epoch": 2.66096256684492, "grad_norm": 1.7227494716644287, "learning_rate": 2.964340650318548e-07, "loss": 0.0442, "step": 622 }, { "epoch": 2.6652406417112298, "grad_norm": 1.6479852199554443, "learning_rate": 2.8885900823097223e-07, "loss": 0.0425, "step": 623 }, { "epoch": 2.66951871657754, "grad_norm": 1.8982220888137817, "learning_rate": 2.813791150085454e-07, "loss": 0.047, "step": 624 }, { "epoch": 2.6737967914438503, "grad_norm": 1.6966097354888916, "learning_rate": 2.73994536456087e-07, "loss": 0.0371, "step": 625 }, { "epoch": 2.6780748663101606, "grad_norm": 1.635574221611023, "learning_rate": 2.6670542173977745e-07, "loss": 0.0359, "step": 626 }, { "epoch": 2.682352941176471, "grad_norm": 1.7045031785964966, "learning_rate": 2.5951191809746146e-07, "loss": 0.0375, "step": 627 }, { "epoch": 2.6866310160427807, "grad_norm": 1.6453510522842407, "learning_rate": 2.524141708356681e-07, "loss": 0.0349, "step": 628 }, { "epoch": 2.690909090909091, "grad_norm": 1.890796184539795, "learning_rate": 2.454123233266781e-07, "loss": 0.048, "step": 629 }, { "epoch": 2.695187165775401, "grad_norm": 1.8894267082214355, "learning_rate": 2.385065170056283e-07, "loss": 0.05, "step": 630 }, { "epoch": 2.699465240641711, "grad_norm": 1.8466600179672241, "learning_rate": 2.3169689136765038e-07, "loss": 0.046, "step": 631 }, { "epoch": 2.7037433155080213, "grad_norm": 1.8343045711517334, "learning_rate": 2.249835839650588e-07, "loss": 0.0486, "step": 632 }, { "epoch": 2.7080213903743315, "grad_norm": 1.7663116455078125, "learning_rate": 2.1836673040456947e-07, "loss": 0.0531, "step": 633 }, { "epoch": 2.712299465240642, "grad_norm": 1.6989845037460327, "learning_rate": 2.1184646434455947e-07, "loss": 0.042, "step": 634 }, { "epoch": 2.716577540106952, "grad_norm": 1.83659029006958, "learning_rate": 2.0542291749237053e-07, "loss": 0.0445, "step": 635 }, { "epoch": 2.720855614973262, "grad_norm": 1.953467845916748, "learning_rate": 1.9909621960164382e-07, "loss": 0.044, "step": 636 }, { "epoch": 2.725133689839572, "grad_norm": 1.8050445318222046, "learning_rate": 1.9286649846970318e-07, "loss": 0.0482, "step": 637 }, { "epoch": 2.7294117647058824, "grad_norm": 1.871206521987915, "learning_rate": 1.8673387993497383e-07, "loss": 0.0425, "step": 638 }, { "epoch": 2.7336898395721922, "grad_norm": 1.7672706842422485, "learning_rate": 1.8069848787443556e-07, "loss": 0.0445, "step": 639 }, { "epoch": 2.7379679144385025, "grad_norm": 1.7932413816452026, "learning_rate": 1.7476044420112637e-07, "loss": 0.046, "step": 640 }, { "epoch": 2.742245989304813, "grad_norm": 1.8379826545715332, "learning_rate": 1.689198688616761e-07, "loss": 0.0404, "step": 641 }, { "epoch": 2.746524064171123, "grad_norm": 1.8425133228302002, "learning_rate": 1.631768798338834e-07, "loss": 0.045, "step": 642 }, { "epoch": 2.7508021390374333, "grad_norm": 1.7943239212036133, "learning_rate": 1.5753159312433762e-07, "loss": 0.043, "step": 643 }, { "epoch": 2.755080213903743, "grad_norm": 1.7196213006973267, "learning_rate": 1.5198412276606622e-07, "loss": 0.0382, "step": 644 }, { "epoch": 2.7593582887700534, "grad_norm": 1.9109811782836914, "learning_rate": 1.465345808162427e-07, "loss": 0.0462, "step": 645 }, { "epoch": 2.7636363636363637, "grad_norm": 2.1393885612487793, "learning_rate": 1.4118307735391412e-07, "loss": 0.0442, "step": 646 }, { "epoch": 2.767914438502674, "grad_norm": 1.9865096807479858, "learning_rate": 1.3592972047777874e-07, "loss": 0.0499, "step": 647 }, { "epoch": 2.772192513368984, "grad_norm": 1.783246636390686, "learning_rate": 1.3077461630400967e-07, "loss": 0.0487, "step": 648 }, { "epoch": 2.776470588235294, "grad_norm": 1.6146647930145264, "learning_rate": 1.2571786896410144e-07, "loss": 0.0384, "step": 649 }, { "epoch": 2.7807486631016043, "grad_norm": 1.6205673217773438, "learning_rate": 1.2075958060277394e-07, "loss": 0.0431, "step": 650 }, { "epoch": 2.7850267379679146, "grad_norm": 1.7698028087615967, "learning_rate": 1.158998513759052e-07, "loss": 0.0502, "step": 651 }, { "epoch": 2.7893048128342244, "grad_norm": 1.8170526027679443, "learning_rate": 1.1113877944850804e-07, "loss": 0.0435, "step": 652 }, { "epoch": 2.7935828877005346, "grad_norm": 1.6055933237075806, "learning_rate": 1.0647646099275267e-07, "loss": 0.0454, "step": 653 }, { "epoch": 2.797860962566845, "grad_norm": 1.6979992389678955, "learning_rate": 1.0191299018601608e-07, "loss": 0.0433, "step": 654 }, { "epoch": 2.802139037433155, "grad_norm": 1.6803139448165894, "learning_rate": 9.744845920898527e-08, "loss": 0.0464, "step": 655 }, { "epoch": 2.8064171122994654, "grad_norm": 1.7149966955184937, "learning_rate": 9.308295824379365e-08, "loss": 0.038, "step": 656 }, { "epoch": 2.8106951871657753, "grad_norm": 1.7448934316635132, "learning_rate": 8.881657547219869e-08, "loss": 0.0377, "step": 657 }, { "epoch": 2.8149732620320855, "grad_norm": 1.6970432996749878, "learning_rate": 8.46493970738016e-08, "loss": 0.0476, "step": 658 }, { "epoch": 2.819251336898396, "grad_norm": 1.6256245374679565, "learning_rate": 8.058150722430658e-08, "loss": 0.0353, "step": 659 }, { "epoch": 2.8235294117647056, "grad_norm": 1.7968635559082031, "learning_rate": 7.661298809381878e-08, "loss": 0.0416, "step": 660 }, { "epoch": 2.827807486631016, "grad_norm": 1.766870379447937, "learning_rate": 7.274391984518736e-08, "loss": 0.0439, "step": 661 }, { "epoch": 2.832085561497326, "grad_norm": 1.977318525314331, "learning_rate": 6.897438063238393e-08, "loss": 0.0484, "step": 662 }, { "epoch": 2.8363636363636364, "grad_norm": 1.9380624294281006, "learning_rate": 6.530444659892443e-08, "loss": 0.0506, "step": 663 }, { "epoch": 2.8406417112299467, "grad_norm": 1.7131413221359253, "learning_rate": 6.173419187633201e-08, "loss": 0.0431, "step": 664 }, { "epoch": 2.8449197860962565, "grad_norm": 1.6274473667144775, "learning_rate": 5.82636885826382e-08, "loss": 0.0385, "step": 665 }, { "epoch": 2.8491978609625668, "grad_norm": 2.1234095096588135, "learning_rate": 5.4893006820926355e-08, "loss": 0.0597, "step": 666 }, { "epoch": 2.853475935828877, "grad_norm": 1.887045979499817, "learning_rate": 5.162221467791772e-08, "loss": 0.045, "step": 667 }, { "epoch": 2.8577540106951873, "grad_norm": 1.6769535541534424, "learning_rate": 4.8451378222592605e-08, "loss": 0.043, "step": 668 }, { "epoch": 2.8620320855614976, "grad_norm": 1.7363476753234863, "learning_rate": 4.5380561504858586e-08, "loss": 0.0452, "step": 669 }, { "epoch": 2.8663101604278074, "grad_norm": 1.816578984260559, "learning_rate": 4.240982655425552e-08, "loss": 0.0442, "step": 670 }, { "epoch": 2.8705882352941177, "grad_norm": 1.9129750728607178, "learning_rate": 3.953923337870147e-08, "loss": 0.0461, "step": 671 }, { "epoch": 2.874866310160428, "grad_norm": 1.8041390180587769, "learning_rate": 3.6768839963285395e-08, "loss": 0.0461, "step": 672 }, { "epoch": 2.8791443850267378, "grad_norm": 1.6570104360580444, "learning_rate": 3.409870226908863e-08, "loss": 0.048, "step": 673 }, { "epoch": 2.883422459893048, "grad_norm": 1.8463152647018433, "learning_rate": 3.1528874232059635e-08, "loss": 0.0477, "step": 674 }, { "epoch": 2.8877005347593583, "grad_norm": 1.7744309902191162, "learning_rate": 2.905940776192384e-08, "loss": 0.0452, "step": 675 }, { "epoch": 2.8919786096256686, "grad_norm": 1.6847875118255615, "learning_rate": 2.669035274113274e-08, "loss": 0.0448, "step": 676 }, { "epoch": 2.896256684491979, "grad_norm": 1.8154982328414917, "learning_rate": 2.4421757023859737e-08, "loss": 0.044, "step": 677 }, { "epoch": 2.9005347593582886, "grad_norm": 1.8036484718322754, "learning_rate": 2.2253666435029797e-08, "loss": 0.0451, "step": 678 }, { "epoch": 2.904812834224599, "grad_norm": 1.6241322755813599, "learning_rate": 2.0186124769396855e-08, "loss": 0.0437, "step": 679 }, { "epoch": 2.909090909090909, "grad_norm": 1.7207401990890503, "learning_rate": 1.8219173790658406e-08, "loss": 0.0491, "step": 680 }, { "epoch": 2.913368983957219, "grad_norm": 1.8342528343200684, "learning_rate": 1.6352853230609534e-08, "loss": 0.0533, "step": 681 }, { "epoch": 2.9176470588235293, "grad_norm": 1.7498728036880493, "learning_rate": 1.4587200788343524e-08, "loss": 0.0462, "step": 682 }, { "epoch": 2.9219251336898395, "grad_norm": 1.915738582611084, "learning_rate": 1.2922252129489165e-08, "loss": 0.056, "step": 683 }, { "epoch": 2.92620320855615, "grad_norm": 1.8094533681869507, "learning_rate": 1.1358040885490196e-08, "loss": 0.0466, "step": 684 }, { "epoch": 2.93048128342246, "grad_norm": 1.4755526781082153, "learning_rate": 9.894598652925857e-09, "loss": 0.0327, "step": 685 }, { "epoch": 2.93475935828877, "grad_norm": 1.829437255859375, "learning_rate": 8.53195499287196e-09, "loss": 0.0431, "step": 686 }, { "epoch": 2.93903743315508, "grad_norm": 1.8991645574569702, "learning_rate": 7.2701374303063565e-09, "loss": 0.0498, "step": 687 }, { "epoch": 2.9433155080213904, "grad_norm": 1.9571171998977661, "learning_rate": 6.109171453549944e-09, "loss": 0.0572, "step": 688 }, { "epoch": 2.9475935828877007, "grad_norm": 1.73331880569458, "learning_rate": 5.049080513752636e-09, "loss": 0.0436, "step": 689 }, { "epoch": 2.951871657754011, "grad_norm": 1.8351444005966187, "learning_rate": 4.089886024421508e-09, "loss": 0.0433, "step": 690 }, { "epoch": 2.9561497326203208, "grad_norm": 1.663426160812378, "learning_rate": 3.2316073609856e-09, "loss": 0.0363, "step": 691 }, { "epoch": 2.960427807486631, "grad_norm": 1.8152010440826416, "learning_rate": 2.474261860406779e-09, "loss": 0.0402, "step": 692 }, { "epoch": 2.9647058823529413, "grad_norm": 2.0324859619140625, "learning_rate": 1.817864820827242e-09, "loss": 0.0413, "step": 693 }, { "epoch": 2.968983957219251, "grad_norm": 1.9560918807983398, "learning_rate": 1.2624295012625409e-09, "loss": 0.0466, "step": 694 }, { "epoch": 2.9732620320855614, "grad_norm": 1.7552952766418457, "learning_rate": 8.079671213334639e-10, "loss": 0.0427, "step": 695 }, { "epoch": 2.9775401069518717, "grad_norm": 1.7103976011276245, "learning_rate": 4.5448686103732876e-10, "loss": 0.0383, "step": 696 }, { "epoch": 2.981818181818182, "grad_norm": 2.1250243186950684, "learning_rate": 2.0199586056590669e-10, "loss": 0.045, "step": 697 }, { "epoch": 2.986096256684492, "grad_norm": 1.7546443939208984, "learning_rate": 5.049922015887276e-11, "loss": 0.0429, "step": 698 }, { "epoch": 2.990374331550802, "grad_norm": 1.6765040159225464, "learning_rate": 0.0, "loss": 0.0426, "step": 699 }, { "epoch": 2.990374331550802, "step": 699, "total_flos": 4.915479180686131e+16, "train_loss": 0.10607939445345699, "train_runtime": 910.371, "train_samples_per_second": 24.626, "train_steps_per_second": 0.768 } ], "logging_steps": 1, "max_steps": 699, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 100, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4.915479180686131e+16, "train_batch_size": 8, "trial_name": null, "trial_params": null }