| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0014025245441796, | |
| "eval_steps": 134, | |
| "global_step": 535, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018700327255726976, | |
| "grad_norm": 13.994741439819336, | |
| "learning_rate": 2e-05, | |
| "loss": 3.2962, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0037400654511453952, | |
| "grad_norm": 16.602272033691406, | |
| "learning_rate": 4e-05, | |
| "loss": 3.6632, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005610098176718092, | |
| "grad_norm": 21.41153335571289, | |
| "learning_rate": 6e-05, | |
| "loss": 3.8852, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0074801309022907905, | |
| "grad_norm": 15.502434730529785, | |
| "learning_rate": 8e-05, | |
| "loss": 3.8645, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009350163627863487, | |
| "grad_norm": 14.494356155395508, | |
| "learning_rate": 0.0001, | |
| "loss": 3.868, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.011220196353436185, | |
| "grad_norm": 20.01993751525879, | |
| "learning_rate": 0.00012, | |
| "loss": 3.5737, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.013090229079008883, | |
| "grad_norm": 14.023553848266602, | |
| "learning_rate": 0.00014, | |
| "loss": 3.2279, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014960261804581581, | |
| "grad_norm": 15.476705551147461, | |
| "learning_rate": 0.00016, | |
| "loss": 3.8968, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.016830294530154277, | |
| "grad_norm": 14.212241172790527, | |
| "learning_rate": 0.00018, | |
| "loss": 3.401, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.018700327255726974, | |
| "grad_norm": 14.703544616699219, | |
| "learning_rate": 0.0002, | |
| "loss": 3.4936, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.020570359981299673, | |
| "grad_norm": 14.499024391174316, | |
| "learning_rate": 0.00019999820960091608, | |
| "loss": 2.5218, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02244039270687237, | |
| "grad_norm": 16.242597579956055, | |
| "learning_rate": 0.00019999283846777488, | |
| "loss": 3.0075, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02431042543244507, | |
| "grad_norm": 16.940950393676758, | |
| "learning_rate": 0.00019998388679290583, | |
| "loss": 2.8596, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.026180458158017766, | |
| "grad_norm": 13.73834228515625, | |
| "learning_rate": 0.00019997135489685034, | |
| "loss": 3.0382, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.028050490883590462, | |
| "grad_norm": 16.882173538208008, | |
| "learning_rate": 0.00019995524322835034, | |
| "loss": 3.4678, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.029920523609163162, | |
| "grad_norm": 15.432026863098145, | |
| "learning_rate": 0.00019993555236433213, | |
| "loss": 3.4976, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.031790556334735855, | |
| "grad_norm": 16.355684280395508, | |
| "learning_rate": 0.00019991228300988585, | |
| "loss": 3.3903, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.033660589060308554, | |
| "grad_norm": 13.33781909942627, | |
| "learning_rate": 0.00019988543599824005, | |
| "loss": 3.1913, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.035530621785881254, | |
| "grad_norm": 18.799007415771484, | |
| "learning_rate": 0.0001998550122907321, | |
| "loss": 3.2008, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03740065451145395, | |
| "grad_norm": 13.545116424560547, | |
| "learning_rate": 0.0001998210129767735, | |
| "loss": 2.8425, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03927068723702665, | |
| "grad_norm": 14.966565132141113, | |
| "learning_rate": 0.00019978343927381113, | |
| "loss": 2.8036, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04114071996259935, | |
| "grad_norm": 17.932737350463867, | |
| "learning_rate": 0.00019974229252728342, | |
| "loss": 2.9432, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.043010752688172046, | |
| "grad_norm": 16.62327766418457, | |
| "learning_rate": 0.0001996975742105723, | |
| "loss": 3.0234, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04488078541374474, | |
| "grad_norm": 17.949251174926758, | |
| "learning_rate": 0.00019964928592495045, | |
| "loss": 3.0715, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04675081813931744, | |
| "grad_norm": 18.255725860595703, | |
| "learning_rate": 0.00019959742939952392, | |
| "loss": 3.155, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04862085086489014, | |
| "grad_norm": 22.37006187438965, | |
| "learning_rate": 0.0001995420064911702, | |
| "loss": 3.0856, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05049088359046283, | |
| "grad_norm": 23.708791732788086, | |
| "learning_rate": 0.00019948301918447183, | |
| "loss": 3.1826, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05236091631603553, | |
| "grad_norm": 21.139902114868164, | |
| "learning_rate": 0.00019942046959164515, | |
| "loss": 3.1289, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05423094904160823, | |
| "grad_norm": 18.88515281677246, | |
| "learning_rate": 0.0001993543599524649, | |
| "loss": 3.2124, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.056100981767180924, | |
| "grad_norm": 22.860271453857422, | |
| "learning_rate": 0.00019928469263418374, | |
| "loss": 3.1165, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.057971014492753624, | |
| "grad_norm": 25.16404151916504, | |
| "learning_rate": 0.0001992114701314478, | |
| "loss": 3.6229, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.059841047218326324, | |
| "grad_norm": 20.945270538330078, | |
| "learning_rate": 0.00019913469506620707, | |
| "loss": 3.2879, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.061711079943899017, | |
| "grad_norm": 19.237075805664062, | |
| "learning_rate": 0.0001990543701876217, | |
| "loss": 2.9019, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06358111266947171, | |
| "grad_norm": 21.761228561401367, | |
| "learning_rate": 0.0001989704983719635, | |
| "loss": 2.7975, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06545114539504442, | |
| "grad_norm": 18.43031120300293, | |
| "learning_rate": 0.00019888308262251285, | |
| "loss": 2.6259, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06732117812061711, | |
| "grad_norm": 29.478282928466797, | |
| "learning_rate": 0.00019879212606945136, | |
| "loss": 3.4395, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0691912108461898, | |
| "grad_norm": 20.588619232177734, | |
| "learning_rate": 0.00019869763196974957, | |
| "loss": 2.4942, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.07106124357176251, | |
| "grad_norm": 25.75873374938965, | |
| "learning_rate": 0.0001985996037070505, | |
| "loss": 3.6293, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0729312762973352, | |
| "grad_norm": 23.140533447265625, | |
| "learning_rate": 0.00019849804479154837, | |
| "loss": 2.928, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0748013090229079, | |
| "grad_norm": 25.42648696899414, | |
| "learning_rate": 0.00019839295885986296, | |
| "loss": 3.1603, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0766713417484806, | |
| "grad_norm": 25.744718551635742, | |
| "learning_rate": 0.00019828434967490943, | |
| "loss": 3.3389, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0785413744740533, | |
| "grad_norm": 39.95504379272461, | |
| "learning_rate": 0.0001981722211257634, | |
| "loss": 3.6146, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.08041140719962599, | |
| "grad_norm": 32.589569091796875, | |
| "learning_rate": 0.00019805657722752202, | |
| "loss": 2.9932, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0822814399251987, | |
| "grad_norm": 48.44812774658203, | |
| "learning_rate": 0.00019793742212115978, | |
| "loss": 3.2977, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08415147265077139, | |
| "grad_norm": 46.82399368286133, | |
| "learning_rate": 0.00019781476007338058, | |
| "loss": 3.3127, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08602150537634409, | |
| "grad_norm": 35.45515441894531, | |
| "learning_rate": 0.00019768859547646478, | |
| "loss": 3.409, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08789153810191679, | |
| "grad_norm": 28.829622268676758, | |
| "learning_rate": 0.00019755893284811196, | |
| "loss": 3.4779, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08976157082748948, | |
| "grad_norm": 42.49812316894531, | |
| "learning_rate": 0.00019742577683127911, | |
| "loss": 3.7014, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.09163160355306219, | |
| "grad_norm": 77.00621795654297, | |
| "learning_rate": 0.0001972891321940145, | |
| "loss": 5.1841, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09350163627863488, | |
| "grad_norm": 46.6325798034668, | |
| "learning_rate": 0.00019714900382928675, | |
| "loss": 3.8944, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09537166900420757, | |
| "grad_norm": 118.46067810058594, | |
| "learning_rate": 0.0001970053967548098, | |
| "loss": 5.2813, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09724170172978028, | |
| "grad_norm": 94.66202545166016, | |
| "learning_rate": 0.0001968583161128631, | |
| "loss": 4.5682, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09911173445535297, | |
| "grad_norm": 60.450828552246094, | |
| "learning_rate": 0.00019670776717010767, | |
| "loss": 3.9013, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.10098176718092566, | |
| "grad_norm": 25.923925399780273, | |
| "learning_rate": 0.0001965537553173972, | |
| "loss": 3.6057, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10285179990649837, | |
| "grad_norm": 14.4677152633667, | |
| "learning_rate": 0.00019639628606958533, | |
| "loss": 3.0911, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10472183263207106, | |
| "grad_norm": 12.290410041809082, | |
| "learning_rate": 0.000196235365065328, | |
| "loss": 2.6739, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10659186535764376, | |
| "grad_norm": 14.281597137451172, | |
| "learning_rate": 0.0001960709980668816, | |
| "loss": 2.8039, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10846189808321646, | |
| "grad_norm": 12.305088996887207, | |
| "learning_rate": 0.0001959031909598966, | |
| "loss": 2.8797, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.11033193080878916, | |
| "grad_norm": 12.407671928405762, | |
| "learning_rate": 0.00019573194975320673, | |
| "loss": 2.402, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.11220196353436185, | |
| "grad_norm": 12.547879219055176, | |
| "learning_rate": 0.0001955572805786141, | |
| "loss": 2.5572, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11407199625993455, | |
| "grad_norm": 15.23267650604248, | |
| "learning_rate": 0.0001953791896906692, | |
| "loss": 3.1069, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.11594202898550725, | |
| "grad_norm": 14.644232749938965, | |
| "learning_rate": 0.00019519768346644737, | |
| "loss": 2.4441, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11781206171107994, | |
| "grad_norm": 14.574198722839355, | |
| "learning_rate": 0.00019501276840532016, | |
| "loss": 3.1865, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11968209443665265, | |
| "grad_norm": 15.284114837646484, | |
| "learning_rate": 0.00019482445112872264, | |
| "loss": 2.7122, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.12155212716222534, | |
| "grad_norm": 14.484579086303711, | |
| "learning_rate": 0.00019463273837991643, | |
| "loss": 2.9405, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.12342215988779803, | |
| "grad_norm": 13.775676727294922, | |
| "learning_rate": 0.00019443763702374812, | |
| "loss": 2.8403, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12529219261337074, | |
| "grad_norm": 14.479483604431152, | |
| "learning_rate": 0.0001942391540464035, | |
| "loss": 2.9306, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12716222533894342, | |
| "grad_norm": 13.77834701538086, | |
| "learning_rate": 0.00019403729655515737, | |
| "loss": 2.7441, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 17.22318458557129, | |
| "learning_rate": 0.0001938320717781191, | |
| "loss": 2.793, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.13090229079008883, | |
| "grad_norm": 14.486649513244629, | |
| "learning_rate": 0.00019362348706397373, | |
| "loss": 2.8446, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1327723235156615, | |
| "grad_norm": 19.49801254272461, | |
| "learning_rate": 0.0001934115498817189, | |
| "loss": 3.0295, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13464235624123422, | |
| "grad_norm": 17.23444938659668, | |
| "learning_rate": 0.00019319626782039734, | |
| "loss": 2.9814, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13651238896680692, | |
| "grad_norm": 18.949792861938477, | |
| "learning_rate": 0.00019297764858882514, | |
| "loss": 3.0234, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1383824216923796, | |
| "grad_norm": 18.541202545166016, | |
| "learning_rate": 0.00019275570001531578, | |
| "loss": 3.1777, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1402524544179523, | |
| "grad_norm": 21.238248825073242, | |
| "learning_rate": 0.00019253043004739968, | |
| "loss": 3.677, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.14212248714352502, | |
| "grad_norm": 24.113332748413086, | |
| "learning_rate": 0.00019230184675153976, | |
| "loss": 3.2631, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1439925198690977, | |
| "grad_norm": 17.838098526000977, | |
| "learning_rate": 0.00019206995831284242, | |
| "loss": 2.728, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1458625525946704, | |
| "grad_norm": 18.885387420654297, | |
| "learning_rate": 0.00019183477303476467, | |
| "loss": 2.9679, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.1477325853202431, | |
| "grad_norm": 19.87195587158203, | |
| "learning_rate": 0.00019159629933881666, | |
| "loss": 2.9025, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1496026180458158, | |
| "grad_norm": 19.265392303466797, | |
| "learning_rate": 0.0001913545457642601, | |
| "loss": 3.4623, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1514726507713885, | |
| "grad_norm": 19.50520133972168, | |
| "learning_rate": 0.00019110952096780258, | |
| "loss": 3.1419, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.1533426834969612, | |
| "grad_norm": 20.407455444335938, | |
| "learning_rate": 0.00019086123372328746, | |
| "loss": 2.5145, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.15521271622253388, | |
| "grad_norm": 42.79829788208008, | |
| "learning_rate": 0.00019060969292137992, | |
| "loss": 3.0625, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1570827489481066, | |
| "grad_norm": 20.8466796875, | |
| "learning_rate": 0.00019035490756924832, | |
| "loss": 2.9626, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1589527816736793, | |
| "grad_norm": 21.628929138183594, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 2.8822, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.16082281439925197, | |
| "grad_norm": 29.18246078491211, | |
| "learning_rate": 0.00018983563982356405, | |
| "loss": 3.4651, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.16269284712482468, | |
| "grad_norm": 26.07509422302246, | |
| "learning_rate": 0.0001895711760239413, | |
| "loss": 3.0619, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1645628798503974, | |
| "grad_norm": 39.00653076171875, | |
| "learning_rate": 0.00018930350486128856, | |
| "loss": 3.1362, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.16643291257597007, | |
| "grad_norm": 23.455015182495117, | |
| "learning_rate": 0.00018903263592036989, | |
| "loss": 3.4385, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16830294530154277, | |
| "grad_norm": 29.16703224182129, | |
| "learning_rate": 0.00018875857890045543, | |
| "loss": 3.1283, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.17017297802711548, | |
| "grad_norm": 29.75694465637207, | |
| "learning_rate": 0.00018848134361497385, | |
| "loss": 3.0399, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.17204301075268819, | |
| "grad_norm": 28.572084426879883, | |
| "learning_rate": 0.00018820093999116124, | |
| "loss": 3.8989, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 50.12794876098633, | |
| "learning_rate": 0.00018791737806970538, | |
| "loss": 3.6389, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.17578307620383357, | |
| "grad_norm": 23.859804153442383, | |
| "learning_rate": 0.00018763066800438636, | |
| "loss": 3.1504, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.17765310892940628, | |
| "grad_norm": 44.48038864135742, | |
| "learning_rate": 0.00018734082006171299, | |
| "loss": 3.1353, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17952314165497896, | |
| "grad_norm": 26.45856475830078, | |
| "learning_rate": 0.00018704784462055503, | |
| "loss": 2.7182, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.18139317438055166, | |
| "grad_norm": 30.120588302612305, | |
| "learning_rate": 0.00018675175217177175, | |
| "loss": 3.6129, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.18326320710612437, | |
| "grad_norm": 61.866886138916016, | |
| "learning_rate": 0.00018645255331783617, | |
| "loss": 5.1282, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.18513323983169705, | |
| "grad_norm": 39.70136260986328, | |
| "learning_rate": 0.00018615025877245523, | |
| "loss": 3.8535, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18700327255726976, | |
| "grad_norm": 51.10054397583008, | |
| "learning_rate": 0.00018584487936018661, | |
| "loss": 3.8204, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18887330528284246, | |
| "grad_norm": 160.09117126464844, | |
| "learning_rate": 0.00018553642601605068, | |
| "loss": 5.7671, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.19074333800841514, | |
| "grad_norm": 126.96265411376953, | |
| "learning_rate": 0.0001852249097851391, | |
| "loss": 5.543, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.19261337073398785, | |
| "grad_norm": 79.90239715576172, | |
| "learning_rate": 0.0001849103418222194, | |
| "loss": 4.1758, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.19448340345956056, | |
| "grad_norm": 48.66373062133789, | |
| "learning_rate": 0.00018459273339133537, | |
| "loss": 3.4729, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.19635343618513323, | |
| "grad_norm": 31.121978759765625, | |
| "learning_rate": 0.0001842720958654039, | |
| "loss": 3.2988, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.19822346891070594, | |
| "grad_norm": 15.884181022644043, | |
| "learning_rate": 0.00018394844072580773, | |
| "loss": 3.4934, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.20009350163627865, | |
| "grad_norm": 13.574240684509277, | |
| "learning_rate": 0.00018362177956198408, | |
| "loss": 2.6176, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.20196353436185133, | |
| "grad_norm": 11.254271507263184, | |
| "learning_rate": 0.00018329212407100994, | |
| "loss": 2.3752, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.20383356708742403, | |
| "grad_norm": 11.554123878479004, | |
| "learning_rate": 0.00018295948605718314, | |
| "loss": 2.295, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.20570359981299674, | |
| "grad_norm": 11.747841835021973, | |
| "learning_rate": 0.0001826238774315995, | |
| "loss": 2.378, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.20757363253856942, | |
| "grad_norm": 13.485661506652832, | |
| "learning_rate": 0.00018228531021172658, | |
| "loss": 2.6167, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.20944366526414213, | |
| "grad_norm": 12.09218692779541, | |
| "learning_rate": 0.0001819437965209732, | |
| "loss": 2.1478, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.21131369798971483, | |
| "grad_norm": 15.334200859069824, | |
| "learning_rate": 0.0001815993485882553, | |
| "loss": 2.7231, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2131837307152875, | |
| "grad_norm": 14.982645988464355, | |
| "learning_rate": 0.0001812519787475582, | |
| "loss": 2.7762, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.21505376344086022, | |
| "grad_norm": 15.121198654174805, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 2.6204, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.21692379616643293, | |
| "grad_norm": 12.098959922790527, | |
| "learning_rate": 0.0001805485232008601, | |
| "loss": 2.6039, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.2187938288920056, | |
| "grad_norm": 13.73119068145752, | |
| "learning_rate": 0.0001801924626841824, | |
| "loss": 2.7469, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2206638616175783, | |
| "grad_norm": 14.34265422821045, | |
| "learning_rate": 0.00017983353063727016, | |
| "loss": 3.0645, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.22253389434315102, | |
| "grad_norm": 13.765779495239258, | |
| "learning_rate": 0.00017947173991275555, | |
| "loss": 3.014, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2244039270687237, | |
| "grad_norm": 17.305545806884766, | |
| "learning_rate": 0.00017910710346563416, | |
| "loss": 2.9139, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2262739597942964, | |
| "grad_norm": 13.76815414428711, | |
| "learning_rate": 0.00017873963435280121, | |
| "loss": 2.6729, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2281439925198691, | |
| "grad_norm": 16.517879486083984, | |
| "learning_rate": 0.000178369345732584, | |
| "loss": 3.1773, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2300140252454418, | |
| "grad_norm": 18.648420333862305, | |
| "learning_rate": 0.00017799625086427064, | |
| "loss": 3.5063, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2318840579710145, | |
| "grad_norm": 19.074386596679688, | |
| "learning_rate": 0.00017762036310763532, | |
| "loss": 2.743, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.2337540906965872, | |
| "grad_norm": 16.391334533691406, | |
| "learning_rate": 0.00017724169592245995, | |
| "loss": 2.7288, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.23562412342215988, | |
| "grad_norm": 26.434553146362305, | |
| "learning_rate": 0.0001768602628680522, | |
| "loss": 2.6386, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.2374941561477326, | |
| "grad_norm": 18.820554733276367, | |
| "learning_rate": 0.00017647607760275987, | |
| "loss": 3.3189, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2393641888733053, | |
| "grad_norm": 18.246240615844727, | |
| "learning_rate": 0.00017608915388348187, | |
| "loss": 2.9794, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.24123422159887797, | |
| "grad_norm": 24.38311004638672, | |
| "learning_rate": 0.00017569950556517566, | |
| "loss": 2.9017, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.24310425432445068, | |
| "grad_norm": 17.729557037353516, | |
| "learning_rate": 0.00017530714660036112, | |
| "loss": 2.7735, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2449742870500234, | |
| "grad_norm": 20.780271530151367, | |
| "learning_rate": 0.00017491209103862084, | |
| "loss": 3.2586, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.24684431977559607, | |
| "grad_norm": 25.706838607788086, | |
| "learning_rate": 0.00017451435302609714, | |
| "loss": 2.9574, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.24871435250116877, | |
| "grad_norm": 17.675687789916992, | |
| "learning_rate": 0.0001741139468049855, | |
| "loss": 2.6356, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2505843852267415, | |
| "grad_norm": 23.035436630249023, | |
| "learning_rate": 0.0001737108867130245, | |
| "loss": 3.3455, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.2505843852267415, | |
| "eval_loss": 3.025357723236084, | |
| "eval_runtime": 12.856, | |
| "eval_samples_per_second": 17.579, | |
| "eval_steps_per_second": 8.79, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.25245441795231416, | |
| "grad_norm": 17.924945831298828, | |
| "learning_rate": 0.00017330518718298264, | |
| "loss": 2.8946, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.25432445067788684, | |
| "grad_norm": 20.890548706054688, | |
| "learning_rate": 0.00017289686274214118, | |
| "loss": 2.9531, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.25619448340345957, | |
| "grad_norm": 19.297199249267578, | |
| "learning_rate": 0.0001724859280117742, | |
| "loss": 2.7719, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 23.349380493164062, | |
| "learning_rate": 0.000172072397706625, | |
| "loss": 3.5567, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.25993454885460493, | |
| "grad_norm": 24.923961639404297, | |
| "learning_rate": 0.00017165628663437922, | |
| "loss": 3.113, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.26180458158017766, | |
| "grad_norm": 21.864042282104492, | |
| "learning_rate": 0.0001712376096951345, | |
| "loss": 2.9502, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.26367461430575034, | |
| "grad_norm": 30.322755813598633, | |
| "learning_rate": 0.00017081638188086697, | |
| "loss": 3.7115, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.265544647031323, | |
| "grad_norm": 23.991077423095703, | |
| "learning_rate": 0.0001703926182748945, | |
| "loss": 3.3406, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.26741467975689576, | |
| "grad_norm": 24.86219024658203, | |
| "learning_rate": 0.00016996633405133655, | |
| "loss": 3.347, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.26928471248246844, | |
| "grad_norm": 29.057273864746094, | |
| "learning_rate": 0.00016953754447457078, | |
| "loss": 2.6652, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2711547452080411, | |
| "grad_norm": 25.7908992767334, | |
| "learning_rate": 0.00016910626489868649, | |
| "loss": 3.1458, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.27302477793361385, | |
| "grad_norm": 27.767778396606445, | |
| "learning_rate": 0.00016867251076693482, | |
| "loss": 3.3319, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.27489481065918653, | |
| "grad_norm": 30.76727294921875, | |
| "learning_rate": 0.0001682362976111758, | |
| "loss": 3.682, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2767648433847592, | |
| "grad_norm": 25.062744140625, | |
| "learning_rate": 0.0001677976410513221, | |
| "loss": 3.0067, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.27863487611033194, | |
| "grad_norm": 48.073890686035156, | |
| "learning_rate": 0.00016735655679477979, | |
| "loss": 4.0274, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2805049088359046, | |
| "grad_norm": 44.6626091003418, | |
| "learning_rate": 0.00016691306063588583, | |
| "loss": 3.1489, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2823749415614773, | |
| "grad_norm": 73.71078491210938, | |
| "learning_rate": 0.0001664671684553426, | |
| "loss": 4.4787, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.28424497428705003, | |
| "grad_norm": 56.24951934814453, | |
| "learning_rate": 0.00016601889621964904, | |
| "loss": 3.7768, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2861150070126227, | |
| "grad_norm": 46.72991943359375, | |
| "learning_rate": 0.00016556825998052924, | |
| "loss": 3.5245, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2879850397381954, | |
| "grad_norm": 31.89575958251953, | |
| "learning_rate": 0.00016511527587435737, | |
| "loss": 3.4228, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 18.977670669555664, | |
| "learning_rate": 0.00016465996012157995, | |
| "loss": 3.0195, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2917251051893408, | |
| "grad_norm": 12.196249008178711, | |
| "learning_rate": 0.00016420232902613523, | |
| "loss": 2.8431, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2935951379149135, | |
| "grad_norm": 11.746070861816406, | |
| "learning_rate": 0.000163742398974869, | |
| "loss": 2.5194, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.2954651706404862, | |
| "grad_norm": 12.877666473388672, | |
| "learning_rate": 0.00016328018643694812, | |
| "loss": 2.695, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.2973352033660589, | |
| "grad_norm": 11.190670013427734, | |
| "learning_rate": 0.00016281570796327068, | |
| "loss": 3.0105, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2992052360916316, | |
| "grad_norm": 12.561261177062988, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 2.113, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3010752688172043, | |
| "grad_norm": 20.69705581665039, | |
| "learning_rate": 0.00016188001981733588, | |
| "loss": 2.7286, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.302945301542777, | |
| "grad_norm": 13.717788696289062, | |
| "learning_rate": 0.00016140884365018252, | |
| "loss": 2.9779, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.30481533426834967, | |
| "grad_norm": 12.364960670471191, | |
| "learning_rate": 0.00016093546855628084, | |
| "loss": 2.6577, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.3066853669939224, | |
| "grad_norm": 12.219094276428223, | |
| "learning_rate": 0.0001604599114862375, | |
| "loss": 2.3431, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3085553997194951, | |
| "grad_norm": 14.56771183013916, | |
| "learning_rate": 0.00015998218946879138, | |
| "loss": 2.7715, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.31042543244506776, | |
| "grad_norm": 14.490442276000977, | |
| "learning_rate": 0.00015950231961020373, | |
| "loss": 2.7294, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3122954651706405, | |
| "grad_norm": 13.455503463745117, | |
| "learning_rate": 0.00015902031909364564, | |
| "loss": 2.7422, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3141654978962132, | |
| "grad_norm": 14.774016380310059, | |
| "learning_rate": 0.00015853620517858276, | |
| "loss": 2.807, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.31603553062178585, | |
| "grad_norm": 15.136216163635254, | |
| "learning_rate": 0.00015804999520015734, | |
| "loss": 2.8907, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3179055633473586, | |
| "grad_norm": 14.85161018371582, | |
| "learning_rate": 0.00015756170656856737, | |
| "loss": 2.8376, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.31977559607293127, | |
| "grad_norm": 16.69162368774414, | |
| "learning_rate": 0.0001570713567684432, | |
| "loss": 2.8552, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.32164562879850395, | |
| "grad_norm": 13.032023429870605, | |
| "learning_rate": 0.00015657896335822147, | |
| "loss": 3.02, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3235156615240767, | |
| "grad_norm": 16.295326232910156, | |
| "learning_rate": 0.00015608454396951645, | |
| "loss": 3.0504, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.32538569424964936, | |
| "grad_norm": 15.998780250549316, | |
| "learning_rate": 0.00015558811630648846, | |
| "loss": 3.0265, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.32725572697522204, | |
| "grad_norm": 25.61530113220215, | |
| "learning_rate": 0.00015508969814521025, | |
| "loss": 2.8652, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3291257597007948, | |
| "grad_norm": 15.13475227355957, | |
| "learning_rate": 0.00015458930733303018, | |
| "loss": 2.5502, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.33099579242636745, | |
| "grad_norm": 13.900614738464355, | |
| "learning_rate": 0.00015408696178793331, | |
| "loss": 2.9654, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.33286582515194013, | |
| "grad_norm": 16.07615852355957, | |
| "learning_rate": 0.00015358267949789966, | |
| "loss": 2.7143, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.33473585787751287, | |
| "grad_norm": 18.99631118774414, | |
| "learning_rate": 0.0001530764785202603, | |
| "loss": 2.8096, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.33660589060308554, | |
| "grad_norm": 20.583770751953125, | |
| "learning_rate": 0.00015256837698105047, | |
| "loss": 2.9865, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3384759233286583, | |
| "grad_norm": 18.920778274536133, | |
| "learning_rate": 0.00015205839307436088, | |
| "loss": 3.3376, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.34034595605423096, | |
| "grad_norm": 18.719928741455078, | |
| "learning_rate": 0.00015154654506168585, | |
| "loss": 2.7643, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.34221598877980364, | |
| "grad_norm": 22.60373878479004, | |
| "learning_rate": 0.00015103285127126962, | |
| "loss": 2.9539, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.34408602150537637, | |
| "grad_norm": 19.586341857910156, | |
| "learning_rate": 0.00015051733009745013, | |
| "loss": 2.5908, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.34595605423094905, | |
| "grad_norm": 20.427947998046875, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 2.6959, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 19.461990356445312, | |
| "learning_rate": 0.000149480879503466, | |
| "loss": 2.9813, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.34969611968209446, | |
| "grad_norm": 23.931228637695312, | |
| "learning_rate": 0.00014895998719650526, | |
| "loss": 3.047, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.35156615240766714, | |
| "grad_norm": 19.459693908691406, | |
| "learning_rate": 0.00014843734173122002, | |
| "loss": 2.9688, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.3534361851332398, | |
| "grad_norm": 25.261966705322266, | |
| "learning_rate": 0.0001479129618224895, | |
| "loss": 2.8956, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.35530621785881256, | |
| "grad_norm": 23.84093475341797, | |
| "learning_rate": 0.00014738686624729986, | |
| "loss": 3.4096, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.35717625058438524, | |
| "grad_norm": 23.549619674682617, | |
| "learning_rate": 0.00014685907384407186, | |
| "loss": 2.9033, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.3590462833099579, | |
| "grad_norm": 21.206594467163086, | |
| "learning_rate": 0.00014632960351198618, | |
| "loss": 3.0757, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.36091631603553065, | |
| "grad_norm": 21.78272247314453, | |
| "learning_rate": 0.00014579847421030678, | |
| "loss": 2.9197, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.3627863487611033, | |
| "grad_norm": 24.570682525634766, | |
| "learning_rate": 0.00014526570495770194, | |
| "loss": 3.6499, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.364656381486676, | |
| "grad_norm": 28.85085105895996, | |
| "learning_rate": 0.00014473131483156327, | |
| "loss": 2.6047, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.36652641421224874, | |
| "grad_norm": 27.24856948852539, | |
| "learning_rate": 0.0001441953229673227, | |
| "loss": 3.0789, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3683964469378214, | |
| "grad_norm": 28.561187744140625, | |
| "learning_rate": 0.000143657748557767, | |
| "loss": 3.4813, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.3702664796633941, | |
| "grad_norm": 27.148723602294922, | |
| "learning_rate": 0.00014311861085235085, | |
| "loss": 3.0691, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.37213651238896683, | |
| "grad_norm": 31.670246124267578, | |
| "learning_rate": 0.00014257792915650728, | |
| "loss": 3.9914, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.3740065451145395, | |
| "grad_norm": 42.1184196472168, | |
| "learning_rate": 0.00014203572283095657, | |
| "loss": 4.6206, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3758765778401122, | |
| "grad_norm": 40.36967086791992, | |
| "learning_rate": 0.00014149201129101286, | |
| "loss": 3.2648, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3777466105656849, | |
| "grad_norm": 38.3328971862793, | |
| "learning_rate": 0.00014094681400588906, | |
| "loss": 3.248, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.3796166432912576, | |
| "grad_norm": 36.41666030883789, | |
| "learning_rate": 0.00014040015049799953, | |
| "loss": 3.2585, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.3814866760168303, | |
| "grad_norm": 23.713642120361328, | |
| "learning_rate": 0.00013985204034226115, | |
| "loss": 2.8715, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.383356708742403, | |
| "grad_norm": 13.198444366455078, | |
| "learning_rate": 0.00013930250316539238, | |
| "loss": 3.0464, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3852267414679757, | |
| "grad_norm": 11.255416870117188, | |
| "learning_rate": 0.0001387515586452103, | |
| "loss": 2.7254, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 11.445615768432617, | |
| "learning_rate": 0.00013819922650992625, | |
| "loss": 2.3275, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3889668069191211, | |
| "grad_norm": 11.428658485412598, | |
| "learning_rate": 0.0001376455265374392, | |
| "loss": 2.7019, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3908368396446938, | |
| "grad_norm": 11.425288200378418, | |
| "learning_rate": 0.00013709047855462765, | |
| "loss": 2.6879, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.39270687237026647, | |
| "grad_norm": 9.648893356323242, | |
| "learning_rate": 0.00013653410243663952, | |
| "loss": 2.1931, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3945769050958392, | |
| "grad_norm": 12.493631362915039, | |
| "learning_rate": 0.00013597641810618073, | |
| "loss": 2.5069, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.3964469378214119, | |
| "grad_norm": 11.868135452270508, | |
| "learning_rate": 0.0001354174455328015, | |
| "loss": 2.8156, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.39831697054698456, | |
| "grad_norm": 10.956955909729004, | |
| "learning_rate": 0.00013485720473218154, | |
| "loss": 2.3458, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4001870032725573, | |
| "grad_norm": 12.486101150512695, | |
| "learning_rate": 0.00013429571576541315, | |
| "loss": 2.559, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.40205703599813, | |
| "grad_norm": 13.160475730895996, | |
| "learning_rate": 0.00013373299873828303, | |
| "loss": 2.5989, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.40392706872370265, | |
| "grad_norm": 14.06461238861084, | |
| "learning_rate": 0.00013316907380055208, | |
| "loss": 2.9656, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4057971014492754, | |
| "grad_norm": 13.031517028808594, | |
| "learning_rate": 0.0001326039611452342, | |
| "loss": 2.494, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.40766713417484807, | |
| "grad_norm": 12.205702781677246, | |
| "learning_rate": 0.00013203768100787297, | |
| "loss": 2.4097, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.40953716690042075, | |
| "grad_norm": 13.180871963500977, | |
| "learning_rate": 0.0001314702536658172, | |
| "loss": 2.5468, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.4114071996259935, | |
| "grad_norm": 16.636207580566406, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 3.3479, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.41327723235156616, | |
| "grad_norm": 13.308319091796875, | |
| "learning_rate": 0.000130332038681685, | |
| "loss": 2.597, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.41514726507713884, | |
| "grad_norm": 13.627033233642578, | |
| "learning_rate": 0.00012976129179678988, | |
| "loss": 2.4171, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4170172978027116, | |
| "grad_norm": 15.449974060058594, | |
| "learning_rate": 0.00012918947922010336, | |
| "loss": 3.2259, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.41888733052828425, | |
| "grad_norm": 16.265350341796875, | |
| "learning_rate": 0.00012861662142707968, | |
| "loss": 2.7149, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.42075736325385693, | |
| "grad_norm": 14.941640853881836, | |
| "learning_rate": 0.00012804273893060028, | |
| "loss": 2.9722, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.42262739597942967, | |
| "grad_norm": 19.596858978271484, | |
| "learning_rate": 0.00012746785228023904, | |
| "loss": 2.9886, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.42449742870500234, | |
| "grad_norm": 23.8212947845459, | |
| "learning_rate": 0.00012689198206152657, | |
| "loss": 3.0503, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.426367461430575, | |
| "grad_norm": 16.560731887817383, | |
| "learning_rate": 0.0001263151488952132, | |
| "loss": 2.6271, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.42823749415614776, | |
| "grad_norm": 16.66946029663086, | |
| "learning_rate": 0.00012573737343653024, | |
| "loss": 2.804, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.43010752688172044, | |
| "grad_norm": 22.840084075927734, | |
| "learning_rate": 0.00012515867637445086, | |
| "loss": 2.7086, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4319775596072931, | |
| "grad_norm": 14.753687858581543, | |
| "learning_rate": 0.00012457907843094882, | |
| "loss": 2.5197, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.43384759233286585, | |
| "grad_norm": 14.791753768920898, | |
| "learning_rate": 0.0001239986003602566, | |
| "loss": 2.394, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.43571762505843853, | |
| "grad_norm": 22.011295318603516, | |
| "learning_rate": 0.00012341726294812238, | |
| "loss": 2.5769, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.4375876577840112, | |
| "grad_norm": 20.5321102142334, | |
| "learning_rate": 0.00012283508701106557, | |
| "loss": 3.0704, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.43945769050958394, | |
| "grad_norm": 16.91661262512207, | |
| "learning_rate": 0.00012225209339563145, | |
| "loss": 2.8309, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.4413277232351566, | |
| "grad_norm": 26.08629608154297, | |
| "learning_rate": 0.00012166830297764471, | |
| "loss": 3.0209, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.4431977559607293, | |
| "grad_norm": 19.537260055541992, | |
| "learning_rate": 0.00012108373666146191, | |
| "loss": 2.8018, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.44506778868630203, | |
| "grad_norm": 23.111852645874023, | |
| "learning_rate": 0.00012049841537922307, | |
| "loss": 3.0249, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.4469378214118747, | |
| "grad_norm": 19.322980880737305, | |
| "learning_rate": 0.00011991236009010183, | |
| "loss": 3.0254, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.4488078541374474, | |
| "grad_norm": 30.11720848083496, | |
| "learning_rate": 0.00011932559177955533, | |
| "loss": 3.9059, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4506778868630201, | |
| "grad_norm": 24.715557098388672, | |
| "learning_rate": 0.00011873813145857249, | |
| "loss": 3.1555, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4525479195885928, | |
| "grad_norm": 29.622190475463867, | |
| "learning_rate": 0.00011815000016292164, | |
| "loss": 2.6222, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.4544179523141655, | |
| "grad_norm": 24.93659019470215, | |
| "learning_rate": 0.00011756121895239753, | |
| "loss": 2.9839, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4562879850397382, | |
| "grad_norm": 26.67534637451172, | |
| "learning_rate": 0.00011697180891006689, | |
| "loss": 3.1704, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.4581580177653109, | |
| "grad_norm": 25.920337677001953, | |
| "learning_rate": 0.00011638179114151377, | |
| "loss": 2.653, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4600280504908836, | |
| "grad_norm": 29.56134796142578, | |
| "learning_rate": 0.0001157911867740836, | |
| "loss": 2.9653, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.4618980832164563, | |
| "grad_norm": 35.859981536865234, | |
| "learning_rate": 0.00011520001695612674, | |
| "loss": 3.6873, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.463768115942029, | |
| "grad_norm": 41.76987838745117, | |
| "learning_rate": 0.00011460830285624118, | |
| "loss": 3.7447, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.46563814866760167, | |
| "grad_norm": 38.14467239379883, | |
| "learning_rate": 0.0001140160656625146, | |
| "loss": 4.65, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.4675081813931744, | |
| "grad_norm": 55.1082649230957, | |
| "learning_rate": 0.00011342332658176555, | |
| "loss": 3.6235, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4693782141187471, | |
| "grad_norm": 34.86373519897461, | |
| "learning_rate": 0.00011283010683878423, | |
| "loss": 3.1883, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.47124824684431976, | |
| "grad_norm": 30.727251052856445, | |
| "learning_rate": 0.00011223642767557227, | |
| "loss": 3.3415, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.4731182795698925, | |
| "grad_norm": 32.8022575378418, | |
| "learning_rate": 0.00011164231035058228, | |
| "loss": 3.1891, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.4749883122954652, | |
| "grad_norm": 21.999401092529297, | |
| "learning_rate": 0.00011104777613795661, | |
| "loss": 3.1296, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.47685834502103785, | |
| "grad_norm": 20.68447494506836, | |
| "learning_rate": 0.00011045284632676536, | |
| "loss": 2.7139, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.4787283777466106, | |
| "grad_norm": 13.36279010772705, | |
| "learning_rate": 0.00010985754222024436, | |
| "loss": 2.4648, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.48059841047218327, | |
| "grad_norm": 10.14539909362793, | |
| "learning_rate": 0.00010926188513503215, | |
| "loss": 2.4138, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.48246844319775595, | |
| "grad_norm": 9.960616111755371, | |
| "learning_rate": 0.00010866589640040669, | |
| "loss": 2.7157, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.4843384759233287, | |
| "grad_norm": 10.990545272827148, | |
| "learning_rate": 0.00010806959735752174, | |
| "loss": 2.4142, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.48620850864890136, | |
| "grad_norm": 10.179009437561035, | |
| "learning_rate": 0.00010747300935864243, | |
| "loss": 2.0645, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.48807854137447404, | |
| "grad_norm": 9.950582504272461, | |
| "learning_rate": 0.00010687615376638093, | |
| "loss": 2.2024, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.4899485741000468, | |
| "grad_norm": 10.244935989379883, | |
| "learning_rate": 0.00010627905195293135, | |
| "loss": 2.1277, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.49181860682561945, | |
| "grad_norm": 11.46534252166748, | |
| "learning_rate": 0.00010568172529930447, | |
| "loss": 2.534, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.49368863955119213, | |
| "grad_norm": 13.022493362426758, | |
| "learning_rate": 0.00010508419519456219, | |
| "loss": 2.3155, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.49555867227676487, | |
| "grad_norm": 15.116632461547852, | |
| "learning_rate": 0.00010448648303505151, | |
| "loss": 3.0314, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.49742870500233755, | |
| "grad_norm": 13.014046669006348, | |
| "learning_rate": 0.0001038886102236385, | |
| "loss": 2.4966, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4992987377279102, | |
| "grad_norm": 11.42091178894043, | |
| "learning_rate": 0.00010329059816894186, | |
| "loss": 2.56, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.501168770453483, | |
| "grad_norm": 12.75170612335205, | |
| "learning_rate": 0.00010269246828456629, | |
| "loss": 2.5727, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.501168770453483, | |
| "eval_loss": 2.828613758087158, | |
| "eval_runtime": 12.8533, | |
| "eval_samples_per_second": 17.583, | |
| "eval_steps_per_second": 8.792, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.5030388031790556, | |
| "grad_norm": 13.714940071105957, | |
| "learning_rate": 0.0001020942419883357, | |
| "loss": 2.4806, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.5049088359046283, | |
| "grad_norm": 14.08552074432373, | |
| "learning_rate": 0.00010149594070152638, | |
| "loss": 2.7159, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.506778868630201, | |
| "grad_norm": 14.860525131225586, | |
| "learning_rate": 0.00010089758584809979, | |
| "loss": 2.5984, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.5086489013557737, | |
| "grad_norm": 14.785524368286133, | |
| "learning_rate": 0.00010029919885393563, | |
| "loss": 2.6805, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.5105189340813464, | |
| "grad_norm": 14.939064979553223, | |
| "learning_rate": 9.970080114606439e-05, | |
| "loss": 2.4628, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.5123889668069191, | |
| "grad_norm": 13.804859161376953, | |
| "learning_rate": 9.910241415190021e-05, | |
| "loss": 2.847, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.5142589995324918, | |
| "grad_norm": 14.139949798583984, | |
| "learning_rate": 9.850405929847366e-05, | |
| "loss": 2.3347, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 16.16015625, | |
| "learning_rate": 9.790575801166432e-05, | |
| "loss": 2.7167, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.5179990649836372, | |
| "grad_norm": 17.632705688476562, | |
| "learning_rate": 9.730753171543374e-05, | |
| "loss": 3.3824, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5198690977092099, | |
| "grad_norm": 20.08257484436035, | |
| "learning_rate": 9.670940183105812e-05, | |
| "loss": 3.1337, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 16.899017333984375, | |
| "learning_rate": 9.611138977636153e-05, | |
| "loss": 2.8295, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5236091631603553, | |
| "grad_norm": 16.086267471313477, | |
| "learning_rate": 9.551351696494854e-05, | |
| "loss": 2.2724, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.525479195885928, | |
| "grad_norm": 17.3554744720459, | |
| "learning_rate": 9.491580480543784e-05, | |
| "loss": 3.0389, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5273492286115007, | |
| "grad_norm": 15.312295913696289, | |
| "learning_rate": 9.431827470069558e-05, | |
| "loss": 1.8464, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5292192613370734, | |
| "grad_norm": 14.824910163879395, | |
| "learning_rate": 9.372094804706867e-05, | |
| "loss": 2.5675, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.531089294062646, | |
| "grad_norm": 62.370765686035156, | |
| "learning_rate": 9.312384623361909e-05, | |
| "loss": 3.6012, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5329593267882188, | |
| "grad_norm": 17.69074249267578, | |
| "learning_rate": 9.252699064135758e-05, | |
| "loss": 2.9718, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5348293595137915, | |
| "grad_norm": 18.726192474365234, | |
| "learning_rate": 9.193040264247829e-05, | |
| "loss": 2.6739, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5366993922393641, | |
| "grad_norm": 23.5485782623291, | |
| "learning_rate": 9.13341035995933e-05, | |
| "loss": 3.519, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.5385694249649369, | |
| "grad_norm": 23.890811920166016, | |
| "learning_rate": 9.073811486496788e-05, | |
| "loss": 3.0704, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5404394576905096, | |
| "grad_norm": 21.349260330200195, | |
| "learning_rate": 9.014245777975565e-05, | |
| "loss": 3.1382, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5423094904160822, | |
| "grad_norm": 27.911243438720703, | |
| "learning_rate": 8.954715367323468e-05, | |
| "loss": 3.3497, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.544179523141655, | |
| "grad_norm": 23.007844924926758, | |
| "learning_rate": 8.89522238620434e-05, | |
| "loss": 2.4557, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5460495558672277, | |
| "grad_norm": 28.537771224975586, | |
| "learning_rate": 8.835768964941773e-05, | |
| "loss": 3.3786, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5479195885928003, | |
| "grad_norm": 23.56622314453125, | |
| "learning_rate": 8.776357232442778e-05, | |
| "loss": 3.245, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5497896213183731, | |
| "grad_norm": 26.669605255126953, | |
| "learning_rate": 8.716989316121578e-05, | |
| "loss": 3.5067, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5516596540439458, | |
| "grad_norm": 25.95106315612793, | |
| "learning_rate": 8.657667341823448e-05, | |
| "loss": 2.7495, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5535296867695184, | |
| "grad_norm": 36.05848693847656, | |
| "learning_rate": 8.598393433748541e-05, | |
| "loss": 3.4859, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5553997194950911, | |
| "grad_norm": 25.490476608276367, | |
| "learning_rate": 8.539169714375885e-05, | |
| "loss": 3.1739, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5572697522206639, | |
| "grad_norm": 36.27049255371094, | |
| "learning_rate": 8.479998304387329e-05, | |
| "loss": 3.6866, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5591397849462365, | |
| "grad_norm": 33.86897277832031, | |
| "learning_rate": 8.420881322591642e-05, | |
| "loss": 2.8654, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5610098176718092, | |
| "grad_norm": 44.89165115356445, | |
| "learning_rate": 8.361820885848624e-05, | |
| "loss": 3.9259, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.562879850397382, | |
| "grad_norm": 20.971372604370117, | |
| "learning_rate": 8.302819108993312e-05, | |
| "loss": 2.7844, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.5647498831229546, | |
| "grad_norm": 24.576387405395508, | |
| "learning_rate": 8.243878104760249e-05, | |
| "loss": 3.0554, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.5666199158485273, | |
| "grad_norm": 25.367847442626953, | |
| "learning_rate": 8.184999983707837e-05, | |
| "loss": 2.9751, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.5684899485741001, | |
| "grad_norm": 20.629751205444336, | |
| "learning_rate": 8.126186854142752e-05, | |
| "loss": 2.8788, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.5703599812996727, | |
| "grad_norm": 17.573223114013672, | |
| "learning_rate": 8.067440822044469e-05, | |
| "loss": 2.3871, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5722300140252454, | |
| "grad_norm": 15.68101692199707, | |
| "learning_rate": 8.00876399098982e-05, | |
| "loss": 1.9594, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5741000467508182, | |
| "grad_norm": 13.078360557556152, | |
| "learning_rate": 7.950158462077697e-05, | |
| "loss": 2.3795, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.5759700794763908, | |
| "grad_norm": 11.211087226867676, | |
| "learning_rate": 7.891626333853812e-05, | |
| "loss": 2.7153, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5778401122019635, | |
| "grad_norm": 9.67572021484375, | |
| "learning_rate": 7.833169702235531e-05, | |
| "loss": 2.4531, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 12.671576499938965, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 2.2148, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5815801776531089, | |
| "grad_norm": 9.585877418518066, | |
| "learning_rate": 7.716491298893442e-05, | |
| "loss": 2.7332, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5834502103786816, | |
| "grad_norm": 10.074613571166992, | |
| "learning_rate": 7.658273705187761e-05, | |
| "loss": 2.2678, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5853202431042543, | |
| "grad_norm": 10.348039627075195, | |
| "learning_rate": 7.600139963974341e-05, | |
| "loss": 2.6394, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.587190275829827, | |
| "grad_norm": 10.6520414352417, | |
| "learning_rate": 7.542092156905123e-05, | |
| "loss": 2.1318, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.5890603085553997, | |
| "grad_norm": 11.883548736572266, | |
| "learning_rate": 7.484132362554915e-05, | |
| "loss": 2.5724, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5909303412809724, | |
| "grad_norm": 11.804710388183594, | |
| "learning_rate": 7.426262656346978e-05, | |
| "loss": 2.3502, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.5928003740065451, | |
| "grad_norm": 12.366445541381836, | |
| "learning_rate": 7.368485110478685e-05, | |
| "loss": 2.2743, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.5946704067321178, | |
| "grad_norm": 13.19687557220459, | |
| "learning_rate": 7.310801793847344e-05, | |
| "loss": 2.6187, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.5965404394576905, | |
| "grad_norm": 13.92205810546875, | |
| "learning_rate": 7.2532147719761e-05, | |
| "loss": 2.8836, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.5984104721832632, | |
| "grad_norm": 13.191638946533203, | |
| "learning_rate": 7.195726106939974e-05, | |
| "loss": 2.7437, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6002805049088359, | |
| "grad_norm": 12.879383087158203, | |
| "learning_rate": 7.138337857292034e-05, | |
| "loss": 2.3656, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.6021505376344086, | |
| "grad_norm": 18.539752960205078, | |
| "learning_rate": 7.081052077989667e-05, | |
| "loss": 2.7168, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.6040205703599812, | |
| "grad_norm": 15.435444831848145, | |
| "learning_rate": 7.023870820321017e-05, | |
| "loss": 2.8032, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.605890603085554, | |
| "grad_norm": 15.457225799560547, | |
| "learning_rate": 6.966796131831501e-05, | |
| "loss": 2.9741, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.6077606358111267, | |
| "grad_norm": 14.631234169006348, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 2.782, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6096306685366993, | |
| "grad_norm": 13.480672836303711, | |
| "learning_rate": 6.85297463341828e-05, | |
| "loss": 2.3114, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.6115007012622721, | |
| "grad_norm": 15.09902572631836, | |
| "learning_rate": 6.796231899212704e-05, | |
| "loss": 2.0404, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.6133707339878448, | |
| "grad_norm": 19.798410415649414, | |
| "learning_rate": 6.739603885476582e-05, | |
| "loss": 3.2275, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.6152407667134174, | |
| "grad_norm": 19.74947738647461, | |
| "learning_rate": 6.683092619944796e-05, | |
| "loss": 2.8039, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.6171107994389902, | |
| "grad_norm": 15.367700576782227, | |
| "learning_rate": 6.626700126171702e-05, | |
| "loss": 2.8833, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6189808321645629, | |
| "grad_norm": 18.47310447692871, | |
| "learning_rate": 6.570428423458687e-05, | |
| "loss": 2.8571, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.6208508648901355, | |
| "grad_norm": 17.92266273498535, | |
| "learning_rate": 6.51427952678185e-05, | |
| "loss": 2.3749, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.6227208976157083, | |
| "grad_norm": 17.863025665283203, | |
| "learning_rate": 6.458255446719854e-05, | |
| "loss": 2.6118, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.624590930341281, | |
| "grad_norm": 16.579729080200195, | |
| "learning_rate": 6.402358189381934e-05, | |
| "loss": 2.2111, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.6264609630668536, | |
| "grad_norm": 21.433856964111328, | |
| "learning_rate": 6.34658975633605e-05, | |
| "loss": 2.8057, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6283309957924264, | |
| "grad_norm": 17.18570899963379, | |
| "learning_rate": 6.290952144537241e-05, | |
| "loss": 2.7357, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6302010285179991, | |
| "grad_norm": 27.641077041625977, | |
| "learning_rate": 6.23544734625608e-05, | |
| "loss": 3.0753, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6320710612435717, | |
| "grad_norm": 25.628257751464844, | |
| "learning_rate": 6.180077349007376e-05, | |
| "loss": 3.2045, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6339410939691444, | |
| "grad_norm": 18.360332489013672, | |
| "learning_rate": 6.12484413547897e-05, | |
| "loss": 2.6208, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6358111266947172, | |
| "grad_norm": 52.43937683105469, | |
| "learning_rate": 6.069749683460765e-05, | |
| "loss": 3.2548, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6376811594202898, | |
| "grad_norm": 25.194255828857422, | |
| "learning_rate": 6.014795965773884e-05, | |
| "loss": 2.9652, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6395511921458625, | |
| "grad_norm": 33.8499870300293, | |
| "learning_rate": 5.9599849502000485e-05, | |
| "loss": 3.4296, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6414212248714353, | |
| "grad_norm": 33.058990478515625, | |
| "learning_rate": 5.9053185994110974e-05, | |
| "loss": 3.6159, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6432912575970079, | |
| "grad_norm": 22.76582145690918, | |
| "learning_rate": 5.8507988708987146e-05, | |
| "loss": 2.9352, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 27.470590591430664, | |
| "learning_rate": 5.796427716904347e-05, | |
| "loss": 3.0844, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6470313230481534, | |
| "grad_norm": 30.144134521484375, | |
| "learning_rate": 5.7422070843492734e-05, | |
| "loss": 3.3897, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.648901355773726, | |
| "grad_norm": 27.228771209716797, | |
| "learning_rate": 5.6881389147649176e-05, | |
| "loss": 3.0126, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6507713884992987, | |
| "grad_norm": 30.412782669067383, | |
| "learning_rate": 5.634225144223302e-05, | |
| "loss": 2.8863, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6526414212248715, | |
| "grad_norm": 43.44699478149414, | |
| "learning_rate": 5.5804677032677354e-05, | |
| "loss": 4.0037, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.6545114539504441, | |
| "grad_norm": 50.50189208984375, | |
| "learning_rate": 5.526868516843673e-05, | |
| "loss": 3.4202, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6563814866760168, | |
| "grad_norm": 17.37109375, | |
| "learning_rate": 5.47342950422981e-05, | |
| "loss": 2.9548, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.6582515194015895, | |
| "grad_norm": 18.439983367919922, | |
| "learning_rate": 5.420152578969326e-05, | |
| "loss": 2.7266, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.6601215521271622, | |
| "grad_norm": 20.2562313079834, | |
| "learning_rate": 5.3670396488013854e-05, | |
| "loss": 2.5423, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.6619915848527349, | |
| "grad_norm": 15.055497169494629, | |
| "learning_rate": 5.3140926155928136e-05, | |
| "loss": 2.6748, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.6638616175783076, | |
| "grad_norm": 13.408388137817383, | |
| "learning_rate": 5.261313375270014e-05, | |
| "loss": 2.9774, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.6657316503038803, | |
| "grad_norm": 14.892769813537598, | |
| "learning_rate": 5.208703817751053e-05, | |
| "loss": 2.2886, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.667601683029453, | |
| "grad_norm": 12.354819297790527, | |
| "learning_rate": 5.156265826877999e-05, | |
| "loss": 2.8214, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.6694717157550257, | |
| "grad_norm": 13.169706344604492, | |
| "learning_rate": 5.1040012803494795e-05, | |
| "loss": 2.4982, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.6713417484805985, | |
| "grad_norm": 12.36215591430664, | |
| "learning_rate": 5.0519120496534044e-05, | |
| "loss": 3.0511, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.6732117812061711, | |
| "grad_norm": 11.116745948791504, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 2.3823, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6750818139317438, | |
| "grad_norm": 11.059001922607422, | |
| "learning_rate": 4.9482669902549894e-05, | |
| "loss": 2.3138, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.6769518466573166, | |
| "grad_norm": 10.290990829467773, | |
| "learning_rate": 4.896714872873038e-05, | |
| "loss": 2.0308, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.6788218793828892, | |
| "grad_norm": 14.33850383758545, | |
| "learning_rate": 4.845345493831419e-05, | |
| "loss": 2.6556, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.6806919121084619, | |
| "grad_norm": 12.50505542755127, | |
| "learning_rate": 4.794160692563917e-05, | |
| "loss": 2.6198, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.6825619448340347, | |
| "grad_norm": 11.454310417175293, | |
| "learning_rate": 4.743162301894952e-05, | |
| "loss": 2.4727, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6844319775596073, | |
| "grad_norm": 11.56086254119873, | |
| "learning_rate": 4.692352147973973e-05, | |
| "loss": 2.1338, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.68630201028518, | |
| "grad_norm": 11.487324714660645, | |
| "learning_rate": 4.6417320502100316e-05, | |
| "loss": 2.4042, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.6881720430107527, | |
| "grad_norm": 11.415949821472168, | |
| "learning_rate": 4.591303821206673e-05, | |
| "loss": 2.6558, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.6900420757363254, | |
| "grad_norm": 12.308177947998047, | |
| "learning_rate": 4.541069266696984e-05, | |
| "loss": 2.3123, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.6919121084618981, | |
| "grad_norm": 11.585166931152344, | |
| "learning_rate": 4.491030185478976e-05, | |
| "loss": 2.7433, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6937821411874708, | |
| "grad_norm": 12.77014446258545, | |
| "learning_rate": 4.441188369351157e-05, | |
| "loss": 2.646, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 12.757980346679688, | |
| "learning_rate": 4.391545603048358e-05, | |
| "loss": 2.702, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.6975222066386162, | |
| "grad_norm": 15.349445343017578, | |
| "learning_rate": 4.3421036641778556e-05, | |
| "loss": 3.0098, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.6993922393641889, | |
| "grad_norm": 15.5166597366333, | |
| "learning_rate": 4.2928643231556844e-05, | |
| "loss": 3.2723, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.7012622720897616, | |
| "grad_norm": 13.189452171325684, | |
| "learning_rate": 4.2438293431432665e-05, | |
| "loss": 2.363, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7031323048153343, | |
| "grad_norm": 17.275712966918945, | |
| "learning_rate": 4.195000479984265e-05, | |
| "loss": 3.0557, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.705002337540907, | |
| "grad_norm": 16.319204330444336, | |
| "learning_rate": 4.146379482141723e-05, | |
| "loss": 2.6115, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.7068723702664796, | |
| "grad_norm": 14.946385383605957, | |
| "learning_rate": 4.097968090635439e-05, | |
| "loss": 2.6765, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.7087424029920524, | |
| "grad_norm": 19.0872745513916, | |
| "learning_rate": 4.049768038979631e-05, | |
| "loss": 2.2803, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.7106124357176251, | |
| "grad_norm": 15.21898078918457, | |
| "learning_rate": 4.001781053120863e-05, | |
| "loss": 2.4357, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7124824684431977, | |
| "grad_norm": 17.521011352539062, | |
| "learning_rate": 3.954008851376252e-05, | |
| "loss": 2.6634, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.7143525011687705, | |
| "grad_norm": 16.630462646484375, | |
| "learning_rate": 3.90645314437192e-05, | |
| "loss": 2.2544, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.7162225338943432, | |
| "grad_norm": 18.478824615478516, | |
| "learning_rate": 3.859115634981748e-05, | |
| "loss": 2.6528, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.7180925666199158, | |
| "grad_norm": 22.990461349487305, | |
| "learning_rate": 3.811998018266416e-05, | |
| "loss": 2.8017, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.7199625993454886, | |
| "grad_norm": 23.46438217163086, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 3.5866, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7218326320710613, | |
| "grad_norm": 19.73815155029297, | |
| "learning_rate": 3.718429203672936e-05, | |
| "loss": 2.6817, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.7237026647966339, | |
| "grad_norm": 20.134384155273438, | |
| "learning_rate": 3.671981356305191e-05, | |
| "loss": 3.4092, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.7255726975222067, | |
| "grad_norm": 19.134763717651367, | |
| "learning_rate": 3.6257601025131026e-05, | |
| "loss": 2.5706, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.7274427302477794, | |
| "grad_norm": 19.289335250854492, | |
| "learning_rate": 3.57976709738648e-05, | |
| "loss": 3.1928, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.729312762973352, | |
| "grad_norm": 21.083354949951172, | |
| "learning_rate": 3.534003987842005e-05, | |
| "loss": 3.1004, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7311827956989247, | |
| "grad_norm": 19.628206253051758, | |
| "learning_rate": 3.488472412564264e-05, | |
| "loss": 3.0638, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.7330528284244975, | |
| "grad_norm": 25.418609619140625, | |
| "learning_rate": 3.4431740019470774e-05, | |
| "loss": 3.3569, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.7349228611500701, | |
| "grad_norm": 24.504236221313477, | |
| "learning_rate": 3.398110378035098e-05, | |
| "loss": 3.3934, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.7367928938756428, | |
| "grad_norm": 32.00563430786133, | |
| "learning_rate": 3.353283154465746e-05, | |
| "loss": 3.3888, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7386629266012156, | |
| "grad_norm": 25.788663864135742, | |
| "learning_rate": 3.308693936411421e-05, | |
| "loss": 3.7214, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7405329593267882, | |
| "grad_norm": 29.49918556213379, | |
| "learning_rate": 3.264344320522024e-05, | |
| "loss": 3.6635, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7424029920523609, | |
| "grad_norm": 52.400569915771484, | |
| "learning_rate": 3.220235894867794e-05, | |
| "loss": 3.574, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.7442730247779337, | |
| "grad_norm": 31.379642486572266, | |
| "learning_rate": 3.1763702388824214e-05, | |
| "loss": 3.8696, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7461430575035063, | |
| "grad_norm": 33.01647186279297, | |
| "learning_rate": 3.132748923306522e-05, | |
| "loss": 3.5235, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.748013090229079, | |
| "grad_norm": 54.04301071166992, | |
| "learning_rate": 3.089373510131354e-05, | |
| "loss": 4.3509, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7498831229546518, | |
| "grad_norm": 9.012869834899902, | |
| "learning_rate": 3.0462455525429257e-05, | |
| "loss": 2.4901, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.7517531556802244, | |
| "grad_norm": 10.679055213928223, | |
| "learning_rate": 3.0033665948663448e-05, | |
| "loss": 2.0973, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.7517531556802244, | |
| "eval_loss": 2.625810146331787, | |
| "eval_runtime": 12.8607, | |
| "eval_samples_per_second": 17.573, | |
| "eval_steps_per_second": 8.786, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.7536231884057971, | |
| "grad_norm": 9.760929107666016, | |
| "learning_rate": 2.960738172510551e-05, | |
| "loss": 2.5254, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.7554932211313699, | |
| "grad_norm": 8.057788848876953, | |
| "learning_rate": 2.9183618119133062e-05, | |
| "loss": 1.8935, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.7573632538569425, | |
| "grad_norm": 9.041131019592285, | |
| "learning_rate": 2.876239030486554e-05, | |
| "loss": 2.5203, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.7592332865825152, | |
| "grad_norm": 9.86754322052002, | |
| "learning_rate": 2.8343713365620772e-05, | |
| "loss": 2.1501, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.7611033193080879, | |
| "grad_norm": 9.489944458007812, | |
| "learning_rate": 2.7927602293375e-05, | |
| "loss": 2.1061, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.7629733520336606, | |
| "grad_norm": 9.308277130126953, | |
| "learning_rate": 2.751407198822583e-05, | |
| "loss": 1.8263, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.7648433847592333, | |
| "grad_norm": 10.4146728515625, | |
| "learning_rate": 2.7103137257858868e-05, | |
| "loss": 2.5591, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.766713417484806, | |
| "grad_norm": 10.501506805419922, | |
| "learning_rate": 2.669481281701739e-05, | |
| "loss": 2.7704, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7685834502103787, | |
| "grad_norm": 12.172067642211914, | |
| "learning_rate": 2.6289113286975485e-05, | |
| "loss": 2.5226, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.7704534829359514, | |
| "grad_norm": 12.657646179199219, | |
| "learning_rate": 2.5886053195014538e-05, | |
| "loss": 2.7101, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.7723235156615241, | |
| "grad_norm": 11.94274616241455, | |
| "learning_rate": 2.5485646973902865e-05, | |
| "loss": 2.6979, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 11.538039207458496, | |
| "learning_rate": 2.508790896137918e-05, | |
| "loss": 2.7357, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.7760635811126695, | |
| "grad_norm": 11.735588073730469, | |
| "learning_rate": 2.4692853399638917e-05, | |
| "loss": 2.7582, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.7779336138382422, | |
| "grad_norm": 11.011154174804688, | |
| "learning_rate": 2.4300494434824373e-05, | |
| "loss": 2.5606, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.7798036465638148, | |
| "grad_norm": 12.485893249511719, | |
| "learning_rate": 2.391084611651816e-05, | |
| "loss": 2.2526, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.7816736792893876, | |
| "grad_norm": 13.266201972961426, | |
| "learning_rate": 2.352392239724016e-05, | |
| "loss": 2.7745, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.7835437120149603, | |
| "grad_norm": 12.729734420776367, | |
| "learning_rate": 2.3139737131947824e-05, | |
| "loss": 2.1154, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.7854137447405329, | |
| "grad_norm": 11.948831558227539, | |
| "learning_rate": 2.275830407754006e-05, | |
| "loss": 2.3556, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7872837774661057, | |
| "grad_norm": 15.799360275268555, | |
| "learning_rate": 2.237963689236472e-05, | |
| "loss": 3.1533, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.7891538101916784, | |
| "grad_norm": 14.989856719970703, | |
| "learning_rate": 2.200374913572939e-05, | |
| "loss": 2.8148, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.791023842917251, | |
| "grad_norm": 17.03190040588379, | |
| "learning_rate": 2.163065426741603e-05, | |
| "loss": 2.8717, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.7928938756428238, | |
| "grad_norm": 16.29947280883789, | |
| "learning_rate": 2.1260365647198798e-05, | |
| "loss": 3.1721, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.7947639083683965, | |
| "grad_norm": 14.097606658935547, | |
| "learning_rate": 2.0892896534365904e-05, | |
| "loss": 2.7177, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7966339410939691, | |
| "grad_norm": 18.629955291748047, | |
| "learning_rate": 2.0528260087244487e-05, | |
| "loss": 3.0088, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.7985039738195419, | |
| "grad_norm": 14.538827896118164, | |
| "learning_rate": 2.016646936272987e-05, | |
| "loss": 2.5514, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.8003740065451146, | |
| "grad_norm": 13.943222045898438, | |
| "learning_rate": 1.9807537315817604e-05, | |
| "loss": 2.4046, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.8022440392706872, | |
| "grad_norm": 17.54332160949707, | |
| "learning_rate": 1.9451476799139935e-05, | |
| "loss": 3.5145, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.80411407199626, | |
| "grad_norm": 16.975404739379883, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 2.6382, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8059841047218327, | |
| "grad_norm": 18.486377716064453, | |
| "learning_rate": 1.8748021252441817e-05, | |
| "loss": 2.5148, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.8078541374474053, | |
| "grad_norm": 15.601974487304688, | |
| "learning_rate": 1.8400651411744685e-05, | |
| "loss": 2.418, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.809724170172978, | |
| "grad_norm": 17.163917541503906, | |
| "learning_rate": 1.805620347902681e-05, | |
| "loss": 2.592, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.8115942028985508, | |
| "grad_norm": 18.376161575317383, | |
| "learning_rate": 1.771468978827343e-05, | |
| "loss": 3.3599, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.8134642356241234, | |
| "grad_norm": 20.212865829467773, | |
| "learning_rate": 1.7376122568400532e-05, | |
| "loss": 3.021, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.8153342683496961, | |
| "grad_norm": 26.279674530029297, | |
| "learning_rate": 1.7040513942816906e-05, | |
| "loss": 3.0855, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.8172043010752689, | |
| "grad_norm": 18.360366821289062, | |
| "learning_rate": 1.6707875928990058e-05, | |
| "loss": 2.2329, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.8190743338008415, | |
| "grad_norm": 25.305208206176758, | |
| "learning_rate": 1.6378220438015933e-05, | |
| "loss": 2.7794, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.8209443665264142, | |
| "grad_norm": 19.780479431152344, | |
| "learning_rate": 1.6051559274192275e-05, | |
| "loss": 2.8426, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.822814399251987, | |
| "grad_norm": 24.164377212524414, | |
| "learning_rate": 1.5727904134596083e-05, | |
| "loss": 3.2383, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8246844319775596, | |
| "grad_norm": 24.38163185119629, | |
| "learning_rate": 1.540726660866466e-05, | |
| "loss": 3.5773, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.8265544647031323, | |
| "grad_norm": 22.310827255249023, | |
| "learning_rate": 1.5089658177780653e-05, | |
| "loss": 3.2983, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.828424497428705, | |
| "grad_norm": 54.48847579956055, | |
| "learning_rate": 1.477509021486091e-05, | |
| "loss": 3.3077, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.8302945301542777, | |
| "grad_norm": 23.941181182861328, | |
| "learning_rate": 1.4463573983949341e-05, | |
| "loss": 3.283, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.8321645628798504, | |
| "grad_norm": 23.76746940612793, | |
| "learning_rate": 1.415512063981339e-05, | |
| "loss": 3.4818, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.8340345956054231, | |
| "grad_norm": 26.566600799560547, | |
| "learning_rate": 1.3849741227544777e-05, | |
| "loss": 2.6301, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.8359046283309958, | |
| "grad_norm": 30.18052864074707, | |
| "learning_rate": 1.3547446682163889e-05, | |
| "loss": 3.8662, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.8377746610565685, | |
| "grad_norm": 36.08682632446289, | |
| "learning_rate": 1.3248247828228245e-05, | |
| "loss": 3.9059, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.8396446937821412, | |
| "grad_norm": 39.56899642944336, | |
| "learning_rate": 1.2952155379444975e-05, | |
| "loss": 3.1025, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.8415147265077139, | |
| "grad_norm": 39.47623062133789, | |
| "learning_rate": 1.2659179938287035e-05, | |
| "loss": 4.5911, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8433847592332866, | |
| "grad_norm": 8.356927871704102, | |
| "learning_rate": 1.2369331995613665e-05, | |
| "loss": 2.3298, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.8452547919588593, | |
| "grad_norm": 7.952692985534668, | |
| "learning_rate": 1.2082621930294635e-05, | |
| "loss": 2.4439, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.847124824684432, | |
| "grad_norm": 9.349421501159668, | |
| "learning_rate": 1.1799060008838791e-05, | |
| "loss": 2.6627, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.8489948574100047, | |
| "grad_norm": 9.868120193481445, | |
| "learning_rate": 1.151865638502615e-05, | |
| "loss": 2.1623, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.8508648901355774, | |
| "grad_norm": 9.13068675994873, | |
| "learning_rate": 1.124142109954459e-05, | |
| "loss": 1.9143, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.85273492286115, | |
| "grad_norm": 7.523913860321045, | |
| "learning_rate": 1.0967364079630115e-05, | |
| "loss": 1.7452, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.8546049555867228, | |
| "grad_norm": 8.57178783416748, | |
| "learning_rate": 1.069649513871147e-05, | |
| "loss": 1.6952, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.8564749883122955, | |
| "grad_norm": 9.974442481994629, | |
| "learning_rate": 1.042882397605871e-05, | |
| "loss": 2.4002, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.8583450210378681, | |
| "grad_norm": 9.599936485290527, | |
| "learning_rate": 1.0164360176435961e-05, | |
| "loss": 2.5942, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.8602150537634409, | |
| "grad_norm": 9.855104446411133, | |
| "learning_rate": 9.903113209758096e-06, | |
| "loss": 1.904, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8620850864890136, | |
| "grad_norm": 11.599543571472168, | |
| "learning_rate": 9.6450924307517e-06, | |
| "loss": 1.6992, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.8639551192145862, | |
| "grad_norm": 9.953221321105957, | |
| "learning_rate": 9.39030707862013e-06, | |
| "loss": 2.1692, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.865825151940159, | |
| "grad_norm": 12.537870407104492, | |
| "learning_rate": 9.138766276712552e-06, | |
| "loss": 2.1261, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.8676951846657317, | |
| "grad_norm": 12.677627563476562, | |
| "learning_rate": 8.890479032197464e-06, | |
| "loss": 2.4732, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 12.54604434967041, | |
| "learning_rate": 8.645454235739903e-06, | |
| "loss": 2.1939, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.8714352501168771, | |
| "grad_norm": 11.750197410583496, | |
| "learning_rate": 8.403700661183355e-06, | |
| "loss": 2.1129, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.8733052828424498, | |
| "grad_norm": 11.671151161193848, | |
| "learning_rate": 8.165226965235328e-06, | |
| "loss": 2.4747, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.8751753155680224, | |
| "grad_norm": 14.79646110534668, | |
| "learning_rate": 7.930041687157607e-06, | |
| "loss": 2.1669, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.8770453482935952, | |
| "grad_norm": 11.03843879699707, | |
| "learning_rate": 7.698153248460271e-06, | |
| "loss": 2.1497, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.8789153810191679, | |
| "grad_norm": 12.181533813476562, | |
| "learning_rate": 7.46956995260033e-06, | |
| "loss": 2.5349, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8807854137447405, | |
| "grad_norm": 27.885995864868164, | |
| "learning_rate": 7.244299984684233e-06, | |
| "loss": 2.6204, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.8826554464703132, | |
| "grad_norm": 14.558568000793457, | |
| "learning_rate": 7.022351411174866e-06, | |
| "loss": 2.2863, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.884525479195886, | |
| "grad_norm": 14.143083572387695, | |
| "learning_rate": 6.803732179602684e-06, | |
| "loss": 2.2255, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.8863955119214586, | |
| "grad_norm": 15.334390640258789, | |
| "learning_rate": 6.5884501182811084e-06, | |
| "loss": 2.7062, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.8882655446470313, | |
| "grad_norm": 14.617964744567871, | |
| "learning_rate": 6.37651293602628e-06, | |
| "loss": 2.4742, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.8901355773726041, | |
| "grad_norm": 13.574972152709961, | |
| "learning_rate": 6.167928221880926e-06, | |
| "loss": 2.6229, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.8920056100981767, | |
| "grad_norm": 14.283699989318848, | |
| "learning_rate": 5.9627034448426545e-06, | |
| "loss": 2.2862, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.8938756428237494, | |
| "grad_norm": 16.076574325561523, | |
| "learning_rate": 5.760845953596527e-06, | |
| "loss": 2.9803, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.8957456755493222, | |
| "grad_norm": 18.35474967956543, | |
| "learning_rate": 5.562362976251901e-06, | |
| "loss": 2.6646, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.8976157082748948, | |
| "grad_norm": 16.39321517944336, | |
| "learning_rate": 5.367261620083575e-06, | |
| "loss": 2.596, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8994857410004675, | |
| "grad_norm": 16.686824798583984, | |
| "learning_rate": 5.175548871277358e-06, | |
| "loss": 2.5261, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.9013557737260403, | |
| "grad_norm": 16.57494354248047, | |
| "learning_rate": 4.9872315946798535e-06, | |
| "loss": 2.3015, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 17.73284339904785, | |
| "learning_rate": 4.80231653355262e-06, | |
| "loss": 2.7761, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.9050958391771856, | |
| "grad_norm": 16.7850284576416, | |
| "learning_rate": 4.620810309330803e-06, | |
| "loss": 2.2855, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.9069658719027583, | |
| "grad_norm": 18.645483016967773, | |
| "learning_rate": 4.442719421385922e-06, | |
| "loss": 3.0409, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.908835904628331, | |
| "grad_norm": 17.9419002532959, | |
| "learning_rate": 4.268050246793276e-06, | |
| "loss": 2.664, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.9107059373539037, | |
| "grad_norm": 17.279287338256836, | |
| "learning_rate": 4.096809040103444e-06, | |
| "loss": 2.2947, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.9125759700794764, | |
| "grad_norm": 25.570619583129883, | |
| "learning_rate": 3.9290019331184145e-06, | |
| "loss": 2.6834, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.9144460028050491, | |
| "grad_norm": 22.238800048828125, | |
| "learning_rate": 3.7646349346719955e-06, | |
| "loss": 3.1555, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.9163160355306218, | |
| "grad_norm": 18.883892059326172, | |
| "learning_rate": 3.6037139304146762e-06, | |
| "loss": 2.4276, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9181860682561945, | |
| "grad_norm": 20.898273468017578, | |
| "learning_rate": 3.446244682602817e-06, | |
| "loss": 3.0974, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.9200561009817672, | |
| "grad_norm": 19.29071044921875, | |
| "learning_rate": 3.292232829892361e-06, | |
| "loss": 2.7486, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.9219261337073399, | |
| "grad_norm": 23.598310470581055, | |
| "learning_rate": 3.1416838871368924e-06, | |
| "loss": 3.4687, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.9237961664329126, | |
| "grad_norm": 22.534082412719727, | |
| "learning_rate": 2.9946032451902194e-06, | |
| "loss": 3.0483, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.9256661991584852, | |
| "grad_norm": 20.772890090942383, | |
| "learning_rate": 2.8509961707132494e-06, | |
| "loss": 2.8497, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.927536231884058, | |
| "grad_norm": 27.105192184448242, | |
| "learning_rate": 2.7108678059855065e-06, | |
| "loss": 3.8906, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.9294062646096307, | |
| "grad_norm": 29.60268211364746, | |
| "learning_rate": 2.5742231687209017e-06, | |
| "loss": 4.1652, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.9312762973352033, | |
| "grad_norm": 30.220417022705078, | |
| "learning_rate": 2.4410671518880655e-06, | |
| "loss": 3.4544, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.9331463300607761, | |
| "grad_norm": 34.319881439208984, | |
| "learning_rate": 2.311404523535243e-06, | |
| "loss": 3.7756, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.9350163627863488, | |
| "grad_norm": 48.59423065185547, | |
| "learning_rate": 2.1852399266194314e-06, | |
| "loss": 4.6194, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9368863955119214, | |
| "grad_norm": 7.484718322753906, | |
| "learning_rate": 2.062577878840244e-06, | |
| "loss": 2.3524, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.9387564282374942, | |
| "grad_norm": 8.843035697937012, | |
| "learning_rate": 1.9434227724779984e-06, | |
| "loss": 2.4799, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.9406264609630669, | |
| "grad_norm": 7.7384161949157715, | |
| "learning_rate": 1.8277788742365965e-06, | |
| "loss": 1.8435, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.9424964936886395, | |
| "grad_norm": 9.377725601196289, | |
| "learning_rate": 1.7156503250905898e-06, | |
| "loss": 2.1838, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.9443665264142123, | |
| "grad_norm": 9.2131929397583, | |
| "learning_rate": 1.6070411401370334e-06, | |
| "loss": 2.0863, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.946236559139785, | |
| "grad_norm": 9.792388916015625, | |
| "learning_rate": 1.501955208451633e-06, | |
| "loss": 2.5053, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.9481065918653576, | |
| "grad_norm": 10.286921501159668, | |
| "learning_rate": 1.400396292949513e-06, | |
| "loss": 1.9311, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.9499766245909304, | |
| "grad_norm": 10.273000717163086, | |
| "learning_rate": 1.3023680302504338e-06, | |
| "loss": 2.4607, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.9518466573165031, | |
| "grad_norm": 9.218442916870117, | |
| "learning_rate": 1.207873930548653e-06, | |
| "loss": 1.5958, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.9537166900420757, | |
| "grad_norm": 12.184614181518555, | |
| "learning_rate": 1.1169173774871478e-06, | |
| "loss": 2.7458, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9555867227676484, | |
| "grad_norm": 12.237886428833008, | |
| "learning_rate": 1.0295016280365112e-06, | |
| "loss": 2.8767, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.9574567554932212, | |
| "grad_norm": 14.358260154724121, | |
| "learning_rate": 9.456298123782902e-07, | |
| "loss": 2.5067, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.9593267882187938, | |
| "grad_norm": 15.50924301147461, | |
| "learning_rate": 8.65304933792932e-07, | |
| "loss": 2.7335, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.9611968209443665, | |
| "grad_norm": 12.935508728027344, | |
| "learning_rate": 7.885298685522235e-07, | |
| "loss": 2.3211, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.9630668536699393, | |
| "grad_norm": 13.291459083557129, | |
| "learning_rate": 7.153073658162646e-07, | |
| "loss": 2.5455, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.9649368863955119, | |
| "grad_norm": 17.141666412353516, | |
| "learning_rate": 6.456400475351232e-07, | |
| "loss": 3.0588, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.9668069191210846, | |
| "grad_norm": 16.180906295776367, | |
| "learning_rate": 5.795304083548559e-07, | |
| "loss": 2.8284, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.9686769518466574, | |
| "grad_norm": 13.24155044555664, | |
| "learning_rate": 5.169808155281786e-07, | |
| "loss": 2.2862, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.97054698457223, | |
| "grad_norm": 15.022529602050781, | |
| "learning_rate": 4.579935088298015e-07, | |
| "loss": 2.5049, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.9724170172978027, | |
| "grad_norm": 18.232423782348633, | |
| "learning_rate": 4.025706004760932e-07, | |
| "loss": 3.3347, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9742870500233755, | |
| "grad_norm": 15.241950988769531, | |
| "learning_rate": 3.50714075049563e-07, | |
| "loss": 2.7126, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.9761570827489481, | |
| "grad_norm": 15.458436965942383, | |
| "learning_rate": 3.0242578942771825e-07, | |
| "loss": 2.6416, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.9780271154745208, | |
| "grad_norm": 18.882591247558594, | |
| "learning_rate": 2.577074727165951e-07, | |
| "loss": 2.461, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.9798971482000935, | |
| "grad_norm": 17.970928192138672, | |
| "learning_rate": 2.1656072618887468e-07, | |
| "loss": 2.9272, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.9817671809256662, | |
| "grad_norm": 18.435932159423828, | |
| "learning_rate": 1.7898702322648453e-07, | |
| "loss": 2.8594, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9836372136512389, | |
| "grad_norm": 19.39750099182129, | |
| "learning_rate": 1.449877092679075e-07, | |
| "loss": 2.1917, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.9855072463768116, | |
| "grad_norm": 20.87994384765625, | |
| "learning_rate": 1.1456400175994252e-07, | |
| "loss": 3.1765, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.9873772791023843, | |
| "grad_norm": 20.146398544311523, | |
| "learning_rate": 8.771699011416168e-08, | |
| "loss": 2.4422, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.989247311827957, | |
| "grad_norm": 21.26473045349121, | |
| "learning_rate": 6.444763566786361e-08, | |
| "loss": 2.6351, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.9911173445535297, | |
| "grad_norm": 24.013477325439453, | |
| "learning_rate": 4.475677164966774e-08, | |
| "loss": 3.721, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9929873772791024, | |
| "grad_norm": 29.331228256225586, | |
| "learning_rate": 2.86451031496604e-08, | |
| "loss": 3.3834, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.9948574100046751, | |
| "grad_norm": 20.945201873779297, | |
| "learning_rate": 1.6113207094181626e-08, | |
| "loss": 2.846, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.9967274427302478, | |
| "grad_norm": 22.861347198486328, | |
| "learning_rate": 7.161532225130607e-09, | |
| "loss": 2.9253, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.9985974754558204, | |
| "grad_norm": 30.596744537353516, | |
| "learning_rate": 1.7903990839229779e-09, | |
| "loss": 4.0554, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.0014025245441796, | |
| "grad_norm": 14.026424407958984, | |
| "learning_rate": 0.0, | |
| "loss": 2.7318, | |
| "step": 535 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 535, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 134, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.8753566709645312e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |