{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.99360341151386, "eval_steps": 500, "global_step": 1170, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0042643923240938165, "grad_norm": 5.762583588325692, "learning_rate": 3.4188034188034194e-07, "loss": 0.8544, "step": 1 }, { "epoch": 0.008528784648187633, "grad_norm": 5.762853501660702, "learning_rate": 6.837606837606839e-07, "loss": 0.8386, "step": 2 }, { "epoch": 0.01279317697228145, "grad_norm": 5.9677523180556635, "learning_rate": 1.0256410256410257e-06, "loss": 0.884, "step": 3 }, { "epoch": 0.017057569296375266, "grad_norm": 5.699941904390049, "learning_rate": 1.3675213675213678e-06, "loss": 0.8563, "step": 4 }, { "epoch": 0.021321961620469083, "grad_norm": 5.62168022349983, "learning_rate": 1.7094017094017097e-06, "loss": 0.852, "step": 5 }, { "epoch": 0.0255863539445629, "grad_norm": 5.493702356483672, "learning_rate": 2.0512820512820513e-06, "loss": 0.8753, "step": 6 }, { "epoch": 0.029850746268656716, "grad_norm": 4.303897748944961, "learning_rate": 2.3931623931623937e-06, "loss": 0.8191, "step": 7 }, { "epoch": 0.03411513859275053, "grad_norm": 3.9794175111660945, "learning_rate": 2.7350427350427355e-06, "loss": 0.7877, "step": 8 }, { "epoch": 0.03837953091684435, "grad_norm": 2.3816462501227518, "learning_rate": 3.0769230769230774e-06, "loss": 0.8057, "step": 9 }, { "epoch": 0.042643923240938165, "grad_norm": 2.0788392465970302, "learning_rate": 3.4188034188034193e-06, "loss": 0.7537, "step": 10 }, { "epoch": 0.046908315565031986, "grad_norm": 1.8863166395033435, "learning_rate": 3.760683760683761e-06, "loss": 0.7445, "step": 11 }, { "epoch": 0.0511727078891258, "grad_norm": 3.5536033903037505, "learning_rate": 4.102564102564103e-06, "loss": 0.7596, "step": 12 }, { "epoch": 0.05543710021321962, "grad_norm": 3.651015786009793, "learning_rate": 4.444444444444444e-06, "loss": 0.7396, "step": 13 }, { "epoch": 0.05970149253731343, "grad_norm": 3.5280670262633596, "learning_rate": 4.786324786324787e-06, "loss": 0.7307, "step": 14 }, { "epoch": 0.06396588486140725, "grad_norm": 3.1262409040889914, "learning_rate": 5.128205128205128e-06, "loss": 0.7104, "step": 15 }, { "epoch": 0.06823027718550106, "grad_norm": 2.89142473527489, "learning_rate": 5.470085470085471e-06, "loss": 0.6875, "step": 16 }, { "epoch": 0.07249466950959488, "grad_norm": 2.4647987498160244, "learning_rate": 5.8119658119658126e-06, "loss": 0.6755, "step": 17 }, { "epoch": 0.0767590618336887, "grad_norm": 1.8250668337679865, "learning_rate": 6.153846153846155e-06, "loss": 0.6897, "step": 18 }, { "epoch": 0.08102345415778252, "grad_norm": 1.4024377137744832, "learning_rate": 6.495726495726496e-06, "loss": 0.6613, "step": 19 }, { "epoch": 0.08528784648187633, "grad_norm": 1.3704415928007279, "learning_rate": 6.837606837606839e-06, "loss": 0.6393, "step": 20 }, { "epoch": 0.08955223880597014, "grad_norm": 1.6980228810730489, "learning_rate": 7.17948717948718e-06, "loss": 0.6399, "step": 21 }, { "epoch": 0.09381663113006397, "grad_norm": 1.6402395462904322, "learning_rate": 7.521367521367522e-06, "loss": 0.6195, "step": 22 }, { "epoch": 0.09808102345415778, "grad_norm": 1.2780132829684268, "learning_rate": 7.863247863247863e-06, "loss": 0.6235, "step": 23 }, { "epoch": 0.1023454157782516, "grad_norm": 1.023680566259576, "learning_rate": 8.205128205128205e-06, "loss": 0.6236, "step": 24 }, { "epoch": 0.10660980810234541, "grad_norm": 1.0592925294970703, "learning_rate": 8.547008547008548e-06, "loss": 0.6317, "step": 25 }, { "epoch": 0.11087420042643924, "grad_norm": 1.2196353358103003, "learning_rate": 8.888888888888888e-06, "loss": 0.6061, "step": 26 }, { "epoch": 0.11513859275053305, "grad_norm": 0.9600456404562788, "learning_rate": 9.230769230769232e-06, "loss": 0.6225, "step": 27 }, { "epoch": 0.11940298507462686, "grad_norm": 0.6975540635958875, "learning_rate": 9.572649572649575e-06, "loss": 0.5981, "step": 28 }, { "epoch": 0.12366737739872068, "grad_norm": 0.9833632649277261, "learning_rate": 9.914529914529915e-06, "loss": 0.5908, "step": 29 }, { "epoch": 0.1279317697228145, "grad_norm": 0.9787682562634072, "learning_rate": 1.0256410256410256e-05, "loss": 0.5978, "step": 30 }, { "epoch": 0.13219616204690832, "grad_norm": 0.7131374700897437, "learning_rate": 1.05982905982906e-05, "loss": 0.5638, "step": 31 }, { "epoch": 0.13646055437100213, "grad_norm": 0.776325884804065, "learning_rate": 1.0940170940170942e-05, "loss": 0.5913, "step": 32 }, { "epoch": 0.14072494669509594, "grad_norm": 0.8028754922062066, "learning_rate": 1.1282051282051283e-05, "loss": 0.6077, "step": 33 }, { "epoch": 0.14498933901918976, "grad_norm": 0.765715406571868, "learning_rate": 1.1623931623931625e-05, "loss": 0.5749, "step": 34 }, { "epoch": 0.14925373134328357, "grad_norm": 0.6112531001869583, "learning_rate": 1.1965811965811966e-05, "loss": 0.5503, "step": 35 }, { "epoch": 0.1535181236673774, "grad_norm": 0.6194576345031709, "learning_rate": 1.230769230769231e-05, "loss": 0.5765, "step": 36 }, { "epoch": 0.15778251599147122, "grad_norm": 0.6976498059588975, "learning_rate": 1.264957264957265e-05, "loss": 0.5592, "step": 37 }, { "epoch": 0.16204690831556504, "grad_norm": 0.5878952789888727, "learning_rate": 1.2991452991452993e-05, "loss": 0.553, "step": 38 }, { "epoch": 0.16631130063965885, "grad_norm": 0.721834531098246, "learning_rate": 1.3333333333333333e-05, "loss": 0.5866, "step": 39 }, { "epoch": 0.17057569296375266, "grad_norm": 0.572927984090807, "learning_rate": 1.3675213675213677e-05, "loss": 0.5439, "step": 40 }, { "epoch": 0.17484008528784648, "grad_norm": 0.5998067556843685, "learning_rate": 1.4017094017094018e-05, "loss": 0.5702, "step": 41 }, { "epoch": 0.1791044776119403, "grad_norm": 0.619560266999832, "learning_rate": 1.435897435897436e-05, "loss": 0.5709, "step": 42 }, { "epoch": 0.18336886993603413, "grad_norm": 0.5985922697985028, "learning_rate": 1.4700854700854703e-05, "loss": 0.5607, "step": 43 }, { "epoch": 0.18763326226012794, "grad_norm": 0.6755231307052864, "learning_rate": 1.5042735042735043e-05, "loss": 0.591, "step": 44 }, { "epoch": 0.19189765458422176, "grad_norm": 0.5611610809137622, "learning_rate": 1.5384615384615387e-05, "loss": 0.557, "step": 45 }, { "epoch": 0.19616204690831557, "grad_norm": 0.6547322304669131, "learning_rate": 1.5726495726495726e-05, "loss": 0.5436, "step": 46 }, { "epoch": 0.20042643923240938, "grad_norm": 0.5654181199007654, "learning_rate": 1.6068376068376072e-05, "loss": 0.5555, "step": 47 }, { "epoch": 0.2046908315565032, "grad_norm": 0.6043389720348237, "learning_rate": 1.641025641025641e-05, "loss": 0.5219, "step": 48 }, { "epoch": 0.208955223880597, "grad_norm": 0.604378677169402, "learning_rate": 1.6752136752136753e-05, "loss": 0.5316, "step": 49 }, { "epoch": 0.21321961620469082, "grad_norm": 0.6156791403933493, "learning_rate": 1.7094017094017095e-05, "loss": 0.5806, "step": 50 }, { "epoch": 0.21748400852878466, "grad_norm": 0.5311982638264362, "learning_rate": 1.7435897435897438e-05, "loss": 0.5119, "step": 51 }, { "epoch": 0.22174840085287847, "grad_norm": 0.6806043408167816, "learning_rate": 1.7777777777777777e-05, "loss": 0.5427, "step": 52 }, { "epoch": 0.2260127931769723, "grad_norm": 0.5268262684314252, "learning_rate": 1.8119658119658122e-05, "loss": 0.5113, "step": 53 }, { "epoch": 0.2302771855010661, "grad_norm": 0.5125041865303465, "learning_rate": 1.8461538461538465e-05, "loss": 0.5527, "step": 54 }, { "epoch": 0.2345415778251599, "grad_norm": 0.5544301707286969, "learning_rate": 1.8803418803418804e-05, "loss": 0.5293, "step": 55 }, { "epoch": 0.23880597014925373, "grad_norm": 0.5776950245039686, "learning_rate": 1.914529914529915e-05, "loss": 0.5365, "step": 56 }, { "epoch": 0.24307036247334754, "grad_norm": 0.5419276025054381, "learning_rate": 1.9487179487179488e-05, "loss": 0.5159, "step": 57 }, { "epoch": 0.24733475479744135, "grad_norm": 0.587019298382551, "learning_rate": 1.982905982905983e-05, "loss": 0.5579, "step": 58 }, { "epoch": 0.2515991471215352, "grad_norm": 0.558245566849885, "learning_rate": 2.0170940170940173e-05, "loss": 0.5227, "step": 59 }, { "epoch": 0.255863539445629, "grad_norm": 0.5636197758898119, "learning_rate": 2.0512820512820512e-05, "loss": 0.5242, "step": 60 }, { "epoch": 0.2601279317697228, "grad_norm": 0.5819473915079625, "learning_rate": 2.0854700854700857e-05, "loss": 0.4985, "step": 61 }, { "epoch": 0.26439232409381663, "grad_norm": 0.6585083765912015, "learning_rate": 2.11965811965812e-05, "loss": 0.5361, "step": 62 }, { "epoch": 0.26865671641791045, "grad_norm": 0.6369197447630097, "learning_rate": 2.153846153846154e-05, "loss": 0.5278, "step": 63 }, { "epoch": 0.27292110874200426, "grad_norm": 0.6947872378280153, "learning_rate": 2.1880341880341884e-05, "loss": 0.5544, "step": 64 }, { "epoch": 0.2771855010660981, "grad_norm": 0.7348638961603595, "learning_rate": 2.2222222222222227e-05, "loss": 0.5345, "step": 65 }, { "epoch": 0.2814498933901919, "grad_norm": 0.5897299352701553, "learning_rate": 2.2564102564102566e-05, "loss": 0.494, "step": 66 }, { "epoch": 0.2857142857142857, "grad_norm": 0.4833992909133447, "learning_rate": 2.2905982905982908e-05, "loss": 0.508, "step": 67 }, { "epoch": 0.2899786780383795, "grad_norm": 0.9014759038348941, "learning_rate": 2.324786324786325e-05, "loss": 0.541, "step": 68 }, { "epoch": 0.2942430703624733, "grad_norm": 0.766201358388276, "learning_rate": 2.3589743589743593e-05, "loss": 0.511, "step": 69 }, { "epoch": 0.29850746268656714, "grad_norm": 0.6763470514557186, "learning_rate": 2.393162393162393e-05, "loss": 0.5069, "step": 70 }, { "epoch": 0.302771855010661, "grad_norm": 0.7960304510774352, "learning_rate": 2.4273504273504274e-05, "loss": 0.5029, "step": 71 }, { "epoch": 0.3070362473347548, "grad_norm": 0.666138666582889, "learning_rate": 2.461538461538462e-05, "loss": 0.5145, "step": 72 }, { "epoch": 0.31130063965884863, "grad_norm": 0.5806980890439299, "learning_rate": 2.495726495726496e-05, "loss": 0.5213, "step": 73 }, { "epoch": 0.31556503198294245, "grad_norm": 0.6186803475054912, "learning_rate": 2.52991452991453e-05, "loss": 0.5092, "step": 74 }, { "epoch": 0.31982942430703626, "grad_norm": 0.737335241924766, "learning_rate": 2.5641025641025646e-05, "loss": 0.5408, "step": 75 }, { "epoch": 0.32409381663113007, "grad_norm": 0.7943077140830255, "learning_rate": 2.5982905982905985e-05, "loss": 0.5693, "step": 76 }, { "epoch": 0.3283582089552239, "grad_norm": 0.7311735920812683, "learning_rate": 2.6324786324786328e-05, "loss": 0.5214, "step": 77 }, { "epoch": 0.3326226012793177, "grad_norm": 0.650892143783386, "learning_rate": 2.6666666666666667e-05, "loss": 0.5414, "step": 78 }, { "epoch": 0.3368869936034115, "grad_norm": 0.8822230320163791, "learning_rate": 2.7008547008547012e-05, "loss": 0.523, "step": 79 }, { "epoch": 0.3411513859275053, "grad_norm": 0.8072091437291843, "learning_rate": 2.7350427350427355e-05, "loss": 0.5305, "step": 80 }, { "epoch": 0.34541577825159914, "grad_norm": 0.8096540761124144, "learning_rate": 2.7692307692307694e-05, "loss": 0.5179, "step": 81 }, { "epoch": 0.34968017057569295, "grad_norm": 0.7668789963017784, "learning_rate": 2.8034188034188036e-05, "loss": 0.533, "step": 82 }, { "epoch": 0.35394456289978676, "grad_norm": 0.8780947367628191, "learning_rate": 2.8376068376068378e-05, "loss": 0.5135, "step": 83 }, { "epoch": 0.3582089552238806, "grad_norm": 0.8644912878031398, "learning_rate": 2.871794871794872e-05, "loss": 0.518, "step": 84 }, { "epoch": 0.3624733475479744, "grad_norm": 0.8027366679366013, "learning_rate": 2.9059829059829063e-05, "loss": 0.5383, "step": 85 }, { "epoch": 0.36673773987206826, "grad_norm": 0.7951556664027158, "learning_rate": 2.9401709401709405e-05, "loss": 0.5177, "step": 86 }, { "epoch": 0.37100213219616207, "grad_norm": 1.0886293807161818, "learning_rate": 2.9743589743589747e-05, "loss": 0.5089, "step": 87 }, { "epoch": 0.3752665245202559, "grad_norm": 0.9060369709344223, "learning_rate": 3.0085470085470086e-05, "loss": 0.4971, "step": 88 }, { "epoch": 0.3795309168443497, "grad_norm": 0.8715123131973772, "learning_rate": 3.042735042735043e-05, "loss": 0.5417, "step": 89 }, { "epoch": 0.3837953091684435, "grad_norm": 1.0361588473316101, "learning_rate": 3.0769230769230774e-05, "loss": 0.4945, "step": 90 }, { "epoch": 0.3880597014925373, "grad_norm": 0.816413638766369, "learning_rate": 3.111111111111112e-05, "loss": 0.5044, "step": 91 }, { "epoch": 0.39232409381663114, "grad_norm": 0.9700049358178698, "learning_rate": 3.145299145299145e-05, "loss": 0.5202, "step": 92 }, { "epoch": 0.39658848614072495, "grad_norm": 1.268167713035034, "learning_rate": 3.1794871794871795e-05, "loss": 0.5295, "step": 93 }, { "epoch": 0.40085287846481876, "grad_norm": 0.7278122145603797, "learning_rate": 3.2136752136752144e-05, "loss": 0.5043, "step": 94 }, { "epoch": 0.4051172707889126, "grad_norm": 1.5360813524930013, "learning_rate": 3.247863247863248e-05, "loss": 0.5026, "step": 95 }, { "epoch": 0.4093816631130064, "grad_norm": 0.9753666203897736, "learning_rate": 3.282051282051282e-05, "loss": 0.5188, "step": 96 }, { "epoch": 0.4136460554371002, "grad_norm": 1.4100688600112532, "learning_rate": 3.316239316239317e-05, "loss": 0.5144, "step": 97 }, { "epoch": 0.417910447761194, "grad_norm": 1.1259322179404374, "learning_rate": 3.3504273504273506e-05, "loss": 0.507, "step": 98 }, { "epoch": 0.42217484008528783, "grad_norm": 1.2305441221747766, "learning_rate": 3.384615384615385e-05, "loss": 0.5241, "step": 99 }, { "epoch": 0.42643923240938164, "grad_norm": 1.2748472488214255, "learning_rate": 3.418803418803419e-05, "loss": 0.5124, "step": 100 }, { "epoch": 0.43070362473347545, "grad_norm": 0.9865825760085118, "learning_rate": 3.452991452991453e-05, "loss": 0.5106, "step": 101 }, { "epoch": 0.4349680170575693, "grad_norm": 0.9763228928414522, "learning_rate": 3.4871794871794875e-05, "loss": 0.4896, "step": 102 }, { "epoch": 0.43923240938166314, "grad_norm": 1.1225185277789729, "learning_rate": 3.521367521367522e-05, "loss": 0.4981, "step": 103 }, { "epoch": 0.44349680170575695, "grad_norm": 1.1094174750241743, "learning_rate": 3.555555555555555e-05, "loss": 0.5062, "step": 104 }, { "epoch": 0.44776119402985076, "grad_norm": 0.898347373542904, "learning_rate": 3.58974358974359e-05, "loss": 0.5007, "step": 105 }, { "epoch": 0.4520255863539446, "grad_norm": 0.773624200108943, "learning_rate": 3.6239316239316245e-05, "loss": 0.5025, "step": 106 }, { "epoch": 0.4562899786780384, "grad_norm": 1.0750778962974719, "learning_rate": 3.658119658119658e-05, "loss": 0.5208, "step": 107 }, { "epoch": 0.4605543710021322, "grad_norm": 0.9050953116899519, "learning_rate": 3.692307692307693e-05, "loss": 0.5276, "step": 108 }, { "epoch": 0.464818763326226, "grad_norm": 0.7021746088531706, "learning_rate": 3.726495726495727e-05, "loss": 0.5052, "step": 109 }, { "epoch": 0.4690831556503198, "grad_norm": 0.7199013432957416, "learning_rate": 3.760683760683761e-05, "loss": 0.5255, "step": 110 }, { "epoch": 0.47334754797441364, "grad_norm": 1.1160990501561305, "learning_rate": 3.794871794871795e-05, "loss": 0.4983, "step": 111 }, { "epoch": 0.47761194029850745, "grad_norm": 1.0555937817011538, "learning_rate": 3.82905982905983e-05, "loss": 0.5124, "step": 112 }, { "epoch": 0.48187633262260127, "grad_norm": 1.0150732063928598, "learning_rate": 3.8632478632478634e-05, "loss": 0.4883, "step": 113 }, { "epoch": 0.4861407249466951, "grad_norm": 0.8697583364801937, "learning_rate": 3.8974358974358976e-05, "loss": 0.5292, "step": 114 }, { "epoch": 0.4904051172707889, "grad_norm": 0.9185038001613525, "learning_rate": 3.931623931623932e-05, "loss": 0.5174, "step": 115 }, { "epoch": 0.4946695095948827, "grad_norm": 0.9745755865848704, "learning_rate": 3.965811965811966e-05, "loss": 0.5147, "step": 116 }, { "epoch": 0.4989339019189765, "grad_norm": 0.6983551544056755, "learning_rate": 4e-05, "loss": 0.5027, "step": 117 }, { "epoch": 0.5031982942430704, "grad_norm": 0.9208847354860332, "learning_rate": 3.9999910989204816e-05, "loss": 0.5286, "step": 118 }, { "epoch": 0.5074626865671642, "grad_norm": 1.2397233900658762, "learning_rate": 3.999964395761154e-05, "loss": 0.5085, "step": 119 }, { "epoch": 0.511727078891258, "grad_norm": 0.7942041370107185, "learning_rate": 3.9999198907597046e-05, "loss": 0.5293, "step": 120 }, { "epoch": 0.5159914712153518, "grad_norm": 1.5080290506230445, "learning_rate": 3.999857584312277e-05, "loss": 0.5222, "step": 121 }, { "epoch": 0.5202558635394456, "grad_norm": 0.9438008695990223, "learning_rate": 3.9997774769734646e-05, "loss": 0.5076, "step": 122 }, { "epoch": 0.5245202558635395, "grad_norm": 1.2363148275202405, "learning_rate": 3.9996795694563096e-05, "loss": 0.5265, "step": 123 }, { "epoch": 0.5287846481876333, "grad_norm": 0.8260779515257772, "learning_rate": 3.999563862632295e-05, "loss": 0.5052, "step": 124 }, { "epoch": 0.5330490405117271, "grad_norm": 1.0383706258239225, "learning_rate": 3.9994303575313356e-05, "loss": 0.4833, "step": 125 }, { "epoch": 0.5373134328358209, "grad_norm": 1.0731088546315215, "learning_rate": 3.999279055341771e-05, "loss": 0.5088, "step": 126 }, { "epoch": 0.5415778251599147, "grad_norm": 0.8305037164926722, "learning_rate": 3.9991099574103544e-05, "loss": 0.4969, "step": 127 }, { "epoch": 0.5458422174840085, "grad_norm": 1.1543261349619998, "learning_rate": 3.9989230652422407e-05, "loss": 0.4858, "step": 128 }, { "epoch": 0.5501066098081023, "grad_norm": 0.911733958800851, "learning_rate": 3.998718380500971e-05, "loss": 0.4867, "step": 129 }, { "epoch": 0.5543710021321961, "grad_norm": 0.9484447965879755, "learning_rate": 3.998495905008461e-05, "loss": 0.5108, "step": 130 }, { "epoch": 0.55863539445629, "grad_norm": 0.7766564927463866, "learning_rate": 3.9982556407449816e-05, "loss": 0.4779, "step": 131 }, { "epoch": 0.5628997867803838, "grad_norm": 0.8227852649914328, "learning_rate": 3.997997589849145e-05, "loss": 0.488, "step": 132 }, { "epoch": 0.5671641791044776, "grad_norm": 0.9370342689098532, "learning_rate": 3.9977217546178825e-05, "loss": 0.5293, "step": 133 }, { "epoch": 0.5714285714285714, "grad_norm": 0.8852132441099798, "learning_rate": 3.997428137506426e-05, "loss": 0.4949, "step": 134 }, { "epoch": 0.5756929637526652, "grad_norm": 0.7412646580333639, "learning_rate": 3.9971167411282835e-05, "loss": 0.5054, "step": 135 }, { "epoch": 0.579957356076759, "grad_norm": 0.751878959140086, "learning_rate": 3.9967875682552206e-05, "loss": 0.5219, "step": 136 }, { "epoch": 0.5842217484008528, "grad_norm": 0.6519295683576232, "learning_rate": 3.9964406218172305e-05, "loss": 0.5182, "step": 137 }, { "epoch": 0.5884861407249466, "grad_norm": 0.6755524662705334, "learning_rate": 3.99607590490251e-05, "loss": 0.5039, "step": 138 }, { "epoch": 0.5927505330490405, "grad_norm": 0.8347488695469496, "learning_rate": 3.995693420757435e-05, "loss": 0.4946, "step": 139 }, { "epoch": 0.5970149253731343, "grad_norm": 0.9427351754391033, "learning_rate": 3.9952931727865264e-05, "loss": 0.5128, "step": 140 }, { "epoch": 0.6012793176972282, "grad_norm": 1.055031036692266, "learning_rate": 3.9948751645524235e-05, "loss": 0.4946, "step": 141 }, { "epoch": 0.605543710021322, "grad_norm": 1.23006352166961, "learning_rate": 3.994439399775851e-05, "loss": 0.522, "step": 142 }, { "epoch": 0.6098081023454158, "grad_norm": 0.9715538904791429, "learning_rate": 3.993985882335584e-05, "loss": 0.5226, "step": 143 }, { "epoch": 0.6140724946695096, "grad_norm": 0.7501049697251858, "learning_rate": 3.9935146162684206e-05, "loss": 0.493, "step": 144 }, { "epoch": 0.6183368869936035, "grad_norm": 0.8701421851852503, "learning_rate": 3.993025605769135e-05, "loss": 0.4963, "step": 145 }, { "epoch": 0.6226012793176973, "grad_norm": 1.2467070738538217, "learning_rate": 3.992518855190449e-05, "loss": 0.4952, "step": 146 }, { "epoch": 0.6268656716417911, "grad_norm": 0.8206473558009, "learning_rate": 3.9919943690429906e-05, "loss": 0.5246, "step": 147 }, { "epoch": 0.6311300639658849, "grad_norm": 1.095920307115201, "learning_rate": 3.991452151995252e-05, "loss": 0.489, "step": 148 }, { "epoch": 0.6353944562899787, "grad_norm": 0.9862679683535909, "learning_rate": 3.9908922088735504e-05, "loss": 0.5258, "step": 149 }, { "epoch": 0.6396588486140725, "grad_norm": 1.0605259536794247, "learning_rate": 3.9903145446619837e-05, "loss": 0.4901, "step": 150 }, { "epoch": 0.6439232409381663, "grad_norm": 0.9963084800854008, "learning_rate": 3.989719164502388e-05, "loss": 0.4998, "step": 151 }, { "epoch": 0.6481876332622601, "grad_norm": 0.9149684222988724, "learning_rate": 3.989106073694289e-05, "loss": 0.5103, "step": 152 }, { "epoch": 0.652452025586354, "grad_norm": 0.9657524344033639, "learning_rate": 3.9884752776948564e-05, "loss": 0.5296, "step": 153 }, { "epoch": 0.6567164179104478, "grad_norm": 0.8189190963813775, "learning_rate": 3.987826782118855e-05, "loss": 0.497, "step": 154 }, { "epoch": 0.6609808102345416, "grad_norm": 0.8044373372459764, "learning_rate": 3.9871605927385976e-05, "loss": 0.5089, "step": 155 }, { "epoch": 0.6652452025586354, "grad_norm": 0.7698309044015865, "learning_rate": 3.9864767154838864e-05, "loss": 0.4952, "step": 156 }, { "epoch": 0.6695095948827292, "grad_norm": 0.9001207944326993, "learning_rate": 3.985775156441968e-05, "loss": 0.4991, "step": 157 }, { "epoch": 0.673773987206823, "grad_norm": 0.6240493911350159, "learning_rate": 3.985055921857477e-05, "loss": 0.4973, "step": 158 }, { "epoch": 0.6780383795309168, "grad_norm": 0.8228062170958025, "learning_rate": 3.9843190181323744e-05, "loss": 0.5416, "step": 159 }, { "epoch": 0.6823027718550106, "grad_norm": 0.7864730645574066, "learning_rate": 3.9835644518259015e-05, "loss": 0.5052, "step": 160 }, { "epoch": 0.6865671641791045, "grad_norm": 0.7692469509420303, "learning_rate": 3.982792229654512e-05, "loss": 0.4863, "step": 161 }, { "epoch": 0.6908315565031983, "grad_norm": 0.809929218928369, "learning_rate": 3.982002358491817e-05, "loss": 0.4896, "step": 162 }, { "epoch": 0.6950959488272921, "grad_norm": 0.8299232929459466, "learning_rate": 3.981194845368523e-05, "loss": 0.5059, "step": 163 }, { "epoch": 0.6993603411513859, "grad_norm": 1.429255919088972, "learning_rate": 3.980369697472367e-05, "loss": 0.5099, "step": 164 }, { "epoch": 0.7036247334754797, "grad_norm": 0.872846079343231, "learning_rate": 3.979526922148058e-05, "loss": 0.518, "step": 165 }, { "epoch": 0.7078891257995735, "grad_norm": 0.7368088843338214, "learning_rate": 3.978666526897204e-05, "loss": 0.49, "step": 166 }, { "epoch": 0.7121535181236673, "grad_norm": 0.9306889404234494, "learning_rate": 3.9777885193782534e-05, "loss": 0.5197, "step": 167 }, { "epoch": 0.7164179104477612, "grad_norm": 0.9830777592344603, "learning_rate": 3.9768929074064206e-05, "loss": 0.5035, "step": 168 }, { "epoch": 0.720682302771855, "grad_norm": 1.9067676890157668, "learning_rate": 3.9759796989536185e-05, "loss": 0.504, "step": 169 }, { "epoch": 0.7249466950959488, "grad_norm": 1.2261524546903462, "learning_rate": 3.975048902148388e-05, "loss": 0.4878, "step": 170 }, { "epoch": 0.7292110874200426, "grad_norm": 1.2878696480150222, "learning_rate": 3.9741005252758255e-05, "loss": 0.4701, "step": 171 }, { "epoch": 0.7334754797441365, "grad_norm": 1.1029539614135186, "learning_rate": 3.9731345767775105e-05, "loss": 0.5018, "step": 172 }, { "epoch": 0.7377398720682303, "grad_norm": 1.404966776096553, "learning_rate": 3.972151065251425e-05, "loss": 0.4853, "step": 173 }, { "epoch": 0.7420042643923241, "grad_norm": 1.1681953067483049, "learning_rate": 3.971149999451886e-05, "loss": 0.4736, "step": 174 }, { "epoch": 0.746268656716418, "grad_norm": 1.054326724521683, "learning_rate": 3.970131388289456e-05, "loss": 0.4931, "step": 175 }, { "epoch": 0.7505330490405118, "grad_norm": 1.3228718253060476, "learning_rate": 3.9690952408308775e-05, "loss": 0.4908, "step": 176 }, { "epoch": 0.7547974413646056, "grad_norm": 0.8356955846154145, "learning_rate": 3.9680415662989806e-05, "loss": 0.4866, "step": 177 }, { "epoch": 0.7590618336886994, "grad_norm": 1.1751465063213684, "learning_rate": 3.966970374072605e-05, "loss": 0.5181, "step": 178 }, { "epoch": 0.7633262260127932, "grad_norm": 0.8467287376160608, "learning_rate": 3.9658816736865183e-05, "loss": 0.502, "step": 179 }, { "epoch": 0.767590618336887, "grad_norm": 0.9233742749028502, "learning_rate": 3.9647754748313294e-05, "loss": 0.4942, "step": 180 }, { "epoch": 0.7718550106609808, "grad_norm": 0.778278596309507, "learning_rate": 3.9636517873534025e-05, "loss": 0.5079, "step": 181 }, { "epoch": 0.7761194029850746, "grad_norm": 0.717645374386566, "learning_rate": 3.9625106212547696e-05, "loss": 0.5039, "step": 182 }, { "epoch": 0.7803837953091685, "grad_norm": 0.7722997985001351, "learning_rate": 3.96135198669304e-05, "loss": 0.4899, "step": 183 }, { "epoch": 0.7846481876332623, "grad_norm": 1.006001772306961, "learning_rate": 3.960175893981312e-05, "loss": 0.4909, "step": 184 }, { "epoch": 0.7889125799573561, "grad_norm": 0.6651331133907898, "learning_rate": 3.958982353588081e-05, "loss": 0.4992, "step": 185 }, { "epoch": 0.7931769722814499, "grad_norm": 0.9754898318183376, "learning_rate": 3.957771376137144e-05, "loss": 0.5103, "step": 186 }, { "epoch": 0.7974413646055437, "grad_norm": 0.5728157064040578, "learning_rate": 3.956542972407509e-05, "loss": 0.4951, "step": 187 }, { "epoch": 0.8017057569296375, "grad_norm": 0.7305936260428764, "learning_rate": 3.955297153333294e-05, "loss": 0.4786, "step": 188 }, { "epoch": 0.8059701492537313, "grad_norm": 0.6733977139026949, "learning_rate": 3.954033930003634e-05, "loss": 0.4862, "step": 189 }, { "epoch": 0.8102345415778252, "grad_norm": 0.5879575682316353, "learning_rate": 3.952753313662581e-05, "loss": 0.4995, "step": 190 }, { "epoch": 0.814498933901919, "grad_norm": 0.7661837329326288, "learning_rate": 3.9514553157090027e-05, "loss": 0.4989, "step": 191 }, { "epoch": 0.8187633262260128, "grad_norm": 0.700425176093609, "learning_rate": 3.9501399476964806e-05, "loss": 0.4638, "step": 192 }, { "epoch": 0.8230277185501066, "grad_norm": 0.8958166865667093, "learning_rate": 3.948807221333212e-05, "loss": 0.4872, "step": 193 }, { "epoch": 0.8272921108742004, "grad_norm": 0.8807166064400732, "learning_rate": 3.947457148481899e-05, "loss": 0.4934, "step": 194 }, { "epoch": 0.8315565031982942, "grad_norm": 0.6091238805523633, "learning_rate": 3.946089741159648e-05, "loss": 0.5032, "step": 195 }, { "epoch": 0.835820895522388, "grad_norm": 0.617575065646615, "learning_rate": 3.94470501153786e-05, "loss": 0.4925, "step": 196 }, { "epoch": 0.8400852878464818, "grad_norm": 1.0204579677550079, "learning_rate": 3.943302971942124e-05, "loss": 0.4764, "step": 197 }, { "epoch": 0.8443496801705757, "grad_norm": 0.7837781333209204, "learning_rate": 3.9418836348521045e-05, "loss": 0.5101, "step": 198 }, { "epoch": 0.8486140724946695, "grad_norm": 0.811890861572903, "learning_rate": 3.940447012901435e-05, "loss": 0.4777, "step": 199 }, { "epoch": 0.8528784648187633, "grad_norm": 0.8205952847062596, "learning_rate": 3.9389931188776016e-05, "loss": 0.5, "step": 200 }, { "epoch": 0.8571428571428571, "grad_norm": 0.9742017513829521, "learning_rate": 3.937521965721831e-05, "loss": 0.4791, "step": 201 }, { "epoch": 0.8614072494669509, "grad_norm": 0.8221657370452279, "learning_rate": 3.9360335665289735e-05, "loss": 0.4904, "step": 202 }, { "epoch": 0.8656716417910447, "grad_norm": 0.7286486621235604, "learning_rate": 3.9345279345473894e-05, "loss": 0.497, "step": 203 }, { "epoch": 0.8699360341151386, "grad_norm": 0.6333922650560874, "learning_rate": 3.933005083178828e-05, "loss": 0.4892, "step": 204 }, { "epoch": 0.8742004264392325, "grad_norm": 0.8578052818677135, "learning_rate": 3.9314650259783116e-05, "loss": 0.5076, "step": 205 }, { "epoch": 0.8784648187633263, "grad_norm": 0.750399410827557, "learning_rate": 3.9299077766540106e-05, "loss": 0.4999, "step": 206 }, { "epoch": 0.8827292110874201, "grad_norm": 0.7444183300691442, "learning_rate": 3.928333349067125e-05, "loss": 0.5174, "step": 207 }, { "epoch": 0.8869936034115139, "grad_norm": 0.7390293670816017, "learning_rate": 3.926741757231761e-05, "loss": 0.5088, "step": 208 }, { "epoch": 0.8912579957356077, "grad_norm": 0.5895494036218917, "learning_rate": 3.925133015314804e-05, "loss": 0.5001, "step": 209 }, { "epoch": 0.8955223880597015, "grad_norm": 0.6960049209107905, "learning_rate": 3.923507137635792e-05, "loss": 0.4615, "step": 210 }, { "epoch": 0.8997867803837953, "grad_norm": 0.6037964081497676, "learning_rate": 3.9218641386667935e-05, "loss": 0.4833, "step": 211 }, { "epoch": 0.9040511727078892, "grad_norm": 0.6639130289439265, "learning_rate": 3.920204033032272e-05, "loss": 0.5172, "step": 212 }, { "epoch": 0.908315565031983, "grad_norm": 0.7469004136077626, "learning_rate": 3.9185268355089606e-05, "loss": 0.503, "step": 213 }, { "epoch": 0.9125799573560768, "grad_norm": 0.9525137178233742, "learning_rate": 3.916832561025727e-05, "loss": 0.5039, "step": 214 }, { "epoch": 0.9168443496801706, "grad_norm": 0.7093471875210987, "learning_rate": 3.915121224663443e-05, "loss": 0.4769, "step": 215 }, { "epoch": 0.9211087420042644, "grad_norm": 0.8092572396863047, "learning_rate": 3.913392841654851e-05, "loss": 0.4798, "step": 216 }, { "epoch": 0.9253731343283582, "grad_norm": 0.7715257865453511, "learning_rate": 3.9116474273844233e-05, "loss": 0.4801, "step": 217 }, { "epoch": 0.929637526652452, "grad_norm": 0.8072195224151741, "learning_rate": 3.909884997388234e-05, "loss": 0.5066, "step": 218 }, { "epoch": 0.9339019189765458, "grad_norm": 0.6747723474026673, "learning_rate": 3.9081055673538093e-05, "loss": 0.4869, "step": 219 }, { "epoch": 0.9381663113006397, "grad_norm": 0.5535598548600815, "learning_rate": 3.90630915312e-05, "loss": 0.4847, "step": 220 }, { "epoch": 0.9424307036247335, "grad_norm": 0.74876001608516, "learning_rate": 3.904495770676831e-05, "loss": 0.4779, "step": 221 }, { "epoch": 0.9466950959488273, "grad_norm": 0.9083484919348656, "learning_rate": 3.902665436165364e-05, "loss": 0.4863, "step": 222 }, { "epoch": 0.9509594882729211, "grad_norm": 0.6461488033648031, "learning_rate": 3.900818165877552e-05, "loss": 0.4784, "step": 223 }, { "epoch": 0.9552238805970149, "grad_norm": 0.6754320741347062, "learning_rate": 3.898953976256094e-05, "loss": 0.472, "step": 224 }, { "epoch": 0.9594882729211087, "grad_norm": 0.9580967189164344, "learning_rate": 3.897072883894291e-05, "loss": 0.5014, "step": 225 }, { "epoch": 0.9637526652452025, "grad_norm": 0.6874808198271747, "learning_rate": 3.895174905535896e-05, "loss": 0.4874, "step": 226 }, { "epoch": 0.9680170575692963, "grad_norm": 0.6832917066807791, "learning_rate": 3.893260058074964e-05, "loss": 0.475, "step": 227 }, { "epoch": 0.9722814498933902, "grad_norm": 0.6575987132879606, "learning_rate": 3.8913283585557054e-05, "loss": 0.4826, "step": 228 }, { "epoch": 0.976545842217484, "grad_norm": 0.6835011560464193, "learning_rate": 3.8893798241723306e-05, "loss": 0.4756, "step": 229 }, { "epoch": 0.9808102345415778, "grad_norm": 0.6203307688676428, "learning_rate": 3.8874144722689e-05, "loss": 0.4854, "step": 230 }, { "epoch": 0.9850746268656716, "grad_norm": 0.6892823863039171, "learning_rate": 3.885432320339167e-05, "loss": 0.495, "step": 231 }, { "epoch": 0.9893390191897654, "grad_norm": 0.540532340018327, "learning_rate": 3.883433386026422e-05, "loss": 0.4868, "step": 232 }, { "epoch": 0.9936034115138592, "grad_norm": 0.6964599188865301, "learning_rate": 3.88141768712334e-05, "loss": 0.4972, "step": 233 }, { "epoch": 0.997867803837953, "grad_norm": 0.5625723596199244, "learning_rate": 3.879385241571817e-05, "loss": 0.4927, "step": 234 }, { "epoch": 1.0031982942430704, "grad_norm": 1.020180894743182, "learning_rate": 3.877336067462812e-05, "loss": 0.7904, "step": 235 }, { "epoch": 1.007462686567164, "grad_norm": 0.7065695413618637, "learning_rate": 3.875270183036187e-05, "loss": 0.4162, "step": 236 }, { "epoch": 1.011727078891258, "grad_norm": 0.723667383252581, "learning_rate": 3.873187606680543e-05, "loss": 0.4699, "step": 237 }, { "epoch": 1.0159914712153517, "grad_norm": 0.6883131102397942, "learning_rate": 3.871088356933059e-05, "loss": 0.418, "step": 238 }, { "epoch": 1.0202558635394456, "grad_norm": 0.7771256561903841, "learning_rate": 3.8689724524793224e-05, "loss": 0.4763, "step": 239 }, { "epoch": 1.0245202558635393, "grad_norm": 0.7452973304187979, "learning_rate": 3.866839912153168e-05, "loss": 0.4056, "step": 240 }, { "epoch": 1.0287846481876333, "grad_norm": 0.6256137550373733, "learning_rate": 3.864690754936506e-05, "loss": 0.3945, "step": 241 }, { "epoch": 1.033049040511727, "grad_norm": 0.8299580033525263, "learning_rate": 3.862524999959157e-05, "loss": 0.45, "step": 242 }, { "epoch": 1.037313432835821, "grad_norm": 0.7637656058846947, "learning_rate": 3.860342666498677e-05, "loss": 0.4415, "step": 243 }, { "epoch": 1.0415778251599148, "grad_norm": 0.6215398485067458, "learning_rate": 3.85814377398019e-05, "loss": 0.4079, "step": 244 }, { "epoch": 1.0458422174840085, "grad_norm": 0.6490505807617634, "learning_rate": 3.8559283419762134e-05, "loss": 0.4176, "step": 245 }, { "epoch": 1.0501066098081024, "grad_norm": 0.555231955081115, "learning_rate": 3.853696390206484e-05, "loss": 0.3918, "step": 246 }, { "epoch": 1.0543710021321961, "grad_norm": 0.7216281242065635, "learning_rate": 3.8514479385377813e-05, "loss": 0.451, "step": 247 }, { "epoch": 1.05863539445629, "grad_norm": 0.6113808147518047, "learning_rate": 3.8491830069837526e-05, "loss": 0.3889, "step": 248 }, { "epoch": 1.0628997867803838, "grad_norm": 0.7048840459286184, "learning_rate": 3.846901615704734e-05, "loss": 0.4507, "step": 249 }, { "epoch": 1.0671641791044777, "grad_norm": 0.8160850539508353, "learning_rate": 3.84460378500757e-05, "loss": 0.4542, "step": 250 }, { "epoch": 1.0714285714285714, "grad_norm": 0.7961547389944886, "learning_rate": 3.842289535345435e-05, "loss": 0.4084, "step": 251 }, { "epoch": 1.0756929637526653, "grad_norm": 0.6472721075881466, "learning_rate": 3.839958887317649e-05, "loss": 0.4252, "step": 252 }, { "epoch": 1.079957356076759, "grad_norm": 0.7571976997791344, "learning_rate": 3.837611861669495e-05, "loss": 0.4391, "step": 253 }, { "epoch": 1.084221748400853, "grad_norm": 0.6417985486555089, "learning_rate": 3.835248479292037e-05, "loss": 0.4298, "step": 254 }, { "epoch": 1.0884861407249466, "grad_norm": 0.5842886170539788, "learning_rate": 3.832868761221926e-05, "loss": 0.4299, "step": 255 }, { "epoch": 1.0927505330490406, "grad_norm": 0.6130005274798123, "learning_rate": 3.830472728641225e-05, "loss": 0.4306, "step": 256 }, { "epoch": 1.0970149253731343, "grad_norm": 0.6292597107386093, "learning_rate": 3.828060402877209e-05, "loss": 0.4227, "step": 257 }, { "epoch": 1.1012793176972282, "grad_norm": 0.742136846771457, "learning_rate": 3.825631805402182e-05, "loss": 0.4385, "step": 258 }, { "epoch": 1.105543710021322, "grad_norm": 0.5582311466247738, "learning_rate": 3.823186957833282e-05, "loss": 0.4066, "step": 259 }, { "epoch": 1.1098081023454158, "grad_norm": 0.7107736966260204, "learning_rate": 3.8207258819322936e-05, "loss": 0.4485, "step": 260 }, { "epoch": 1.1140724946695095, "grad_norm": 0.5933308768453202, "learning_rate": 3.818248599605448e-05, "loss": 0.3744, "step": 261 }, { "epoch": 1.1183368869936035, "grad_norm": 0.664231277660076, "learning_rate": 3.8157551329032315e-05, "loss": 0.4399, "step": 262 }, { "epoch": 1.1226012793176972, "grad_norm": 0.5756320234685443, "learning_rate": 3.81324550402019e-05, "loss": 0.3954, "step": 263 }, { "epoch": 1.126865671641791, "grad_norm": 0.5450744362052974, "learning_rate": 3.810719735294731e-05, "loss": 0.4005, "step": 264 }, { "epoch": 1.1311300639658848, "grad_norm": 0.6811551113220277, "learning_rate": 3.808177849208922e-05, "loss": 0.4376, "step": 265 }, { "epoch": 1.1353944562899787, "grad_norm": 0.7297298114596543, "learning_rate": 3.8056198683882914e-05, "loss": 0.4441, "step": 266 }, { "epoch": 1.1396588486140724, "grad_norm": 0.6654168989431408, "learning_rate": 3.8030458156016326e-05, "loss": 0.4166, "step": 267 }, { "epoch": 1.1439232409381663, "grad_norm": 0.590097586325753, "learning_rate": 3.800455713760792e-05, "loss": 0.4459, "step": 268 }, { "epoch": 1.14818763326226, "grad_norm": 0.6520050326986123, "learning_rate": 3.797849585920474e-05, "loss": 0.3812, "step": 269 }, { "epoch": 1.152452025586354, "grad_norm": 0.5487525639612298, "learning_rate": 3.795227455278029e-05, "loss": 0.4349, "step": 270 }, { "epoch": 1.1567164179104479, "grad_norm": 0.8148597696250168, "learning_rate": 3.792589345173249e-05, "loss": 0.4159, "step": 271 }, { "epoch": 1.1609808102345416, "grad_norm": 0.5234208183844118, "learning_rate": 3.7899352790881636e-05, "loss": 0.4103, "step": 272 }, { "epoch": 1.1652452025586353, "grad_norm": 0.7783194060987326, "learning_rate": 3.787265280646825e-05, "loss": 0.4337, "step": 273 }, { "epoch": 1.1695095948827292, "grad_norm": 0.6077482932406173, "learning_rate": 3.784579373615102e-05, "loss": 0.4448, "step": 274 }, { "epoch": 1.1737739872068231, "grad_norm": 0.713559723726581, "learning_rate": 3.781877581900467e-05, "loss": 0.3954, "step": 275 }, { "epoch": 1.1780383795309168, "grad_norm": 0.6706429841933795, "learning_rate": 3.7791599295517825e-05, "loss": 0.4363, "step": 276 }, { "epoch": 1.1823027718550105, "grad_norm": 0.7750344644644377, "learning_rate": 3.776426440759088e-05, "loss": 0.435, "step": 277 }, { "epoch": 1.1865671641791045, "grad_norm": 0.6485505143173109, "learning_rate": 3.7736771398533855e-05, "loss": 0.3992, "step": 278 }, { "epoch": 1.1908315565031984, "grad_norm": 0.6278621684062865, "learning_rate": 3.7709120513064196e-05, "loss": 0.4532, "step": 279 }, { "epoch": 1.195095948827292, "grad_norm": 0.4993768641894848, "learning_rate": 3.768131199730465e-05, "loss": 0.4289, "step": 280 }, { "epoch": 1.199360341151386, "grad_norm": 0.5680603593838152, "learning_rate": 3.7653346098781016e-05, "loss": 0.4178, "step": 281 }, { "epoch": 1.2036247334754797, "grad_norm": 0.6166978263820861, "learning_rate": 3.762522306641998e-05, "loss": 0.4082, "step": 282 }, { "epoch": 1.2078891257995736, "grad_norm": 0.49268121002441434, "learning_rate": 3.759694315054689e-05, "loss": 0.4053, "step": 283 }, { "epoch": 1.2121535181236673, "grad_norm": 0.5952406524302505, "learning_rate": 3.7568506602883535e-05, "loss": 0.4392, "step": 284 }, { "epoch": 1.2164179104477613, "grad_norm": 0.7640007391300089, "learning_rate": 3.7539913676545874e-05, "loss": 0.4142, "step": 285 }, { "epoch": 1.220682302771855, "grad_norm": 0.566641974103758, "learning_rate": 3.7511164626041823e-05, "loss": 0.4077, "step": 286 }, { "epoch": 1.224946695095949, "grad_norm": 0.6259927094471025, "learning_rate": 3.748225970726897e-05, "loss": 0.4067, "step": 287 }, { "epoch": 1.2292110874200426, "grad_norm": 0.5599340338022601, "learning_rate": 3.745319917751229e-05, "loss": 0.4483, "step": 288 }, { "epoch": 1.2334754797441365, "grad_norm": 0.5263970037589443, "learning_rate": 3.742398329544187e-05, "loss": 0.4036, "step": 289 }, { "epoch": 1.2377398720682302, "grad_norm": 0.5983772515182831, "learning_rate": 3.7394612321110606e-05, "loss": 0.4485, "step": 290 }, { "epoch": 1.2420042643923241, "grad_norm": 0.7056768634699175, "learning_rate": 3.736508651595188e-05, "loss": 0.4423, "step": 291 }, { "epoch": 1.2462686567164178, "grad_norm": 0.6538542504257794, "learning_rate": 3.733540614277721e-05, "loss": 0.3845, "step": 292 }, { "epoch": 1.2505330490405118, "grad_norm": 0.5352314141573133, "learning_rate": 3.7305571465773976e-05, "loss": 0.455, "step": 293 }, { "epoch": 1.2547974413646055, "grad_norm": 0.8442015630816173, "learning_rate": 3.727558275050301e-05, "loss": 0.4175, "step": 294 }, { "epoch": 1.2590618336886994, "grad_norm": 0.7334448723876601, "learning_rate": 3.724544026389624e-05, "loss": 0.4193, "step": 295 }, { "epoch": 1.263326226012793, "grad_norm": 0.6837481469169735, "learning_rate": 3.721514427425435e-05, "loss": 0.4563, "step": 296 }, { "epoch": 1.267590618336887, "grad_norm": 0.4843766573199469, "learning_rate": 3.718469505124434e-05, "loss": 0.393, "step": 297 }, { "epoch": 1.271855010660981, "grad_norm": 0.5246589482230002, "learning_rate": 3.715409286589717e-05, "loss": 0.4223, "step": 298 }, { "epoch": 1.2761194029850746, "grad_norm": 0.502659816866133, "learning_rate": 3.7123337990605335e-05, "loss": 0.4147, "step": 299 }, { "epoch": 1.2803837953091683, "grad_norm": 0.5247676176001221, "learning_rate": 3.709243069912041e-05, "loss": 0.4625, "step": 300 }, { "epoch": 1.2846481876332623, "grad_norm": 0.45876509318226205, "learning_rate": 3.706137126655066e-05, "loss": 0.3995, "step": 301 }, { "epoch": 1.2889125799573562, "grad_norm": 0.4627163601994038, "learning_rate": 3.703015996935858e-05, "loss": 0.4193, "step": 302 }, { "epoch": 1.29317697228145, "grad_norm": 0.6686928675672329, "learning_rate": 3.699879708535838e-05, "loss": 0.4439, "step": 303 }, { "epoch": 1.2974413646055436, "grad_norm": 0.5922109448050908, "learning_rate": 3.6967282893713595e-05, "loss": 0.4263, "step": 304 }, { "epoch": 1.3017057569296375, "grad_norm": 0.5470627126862041, "learning_rate": 3.6935617674934554e-05, "loss": 0.4062, "step": 305 }, { "epoch": 1.3059701492537314, "grad_norm": 0.6136095476120296, "learning_rate": 3.69038017108759e-05, "loss": 0.4738, "step": 306 }, { "epoch": 1.3102345415778252, "grad_norm": 0.7612711898578125, "learning_rate": 3.687183528473404e-05, "loss": 0.4267, "step": 307 }, { "epoch": 1.3144989339019189, "grad_norm": 0.5910092852049679, "learning_rate": 3.683971868104468e-05, "loss": 0.4401, "step": 308 }, { "epoch": 1.3187633262260128, "grad_norm": 0.5005219840239118, "learning_rate": 3.680745218568026e-05, "loss": 0.4137, "step": 309 }, { "epoch": 1.3230277185501067, "grad_norm": 0.7165297595567666, "learning_rate": 3.677503608584743e-05, "loss": 0.4336, "step": 310 }, { "epoch": 1.3272921108742004, "grad_norm": 0.7862156021846395, "learning_rate": 3.674247067008447e-05, "loss": 0.4259, "step": 311 }, { "epoch": 1.331556503198294, "grad_norm": 0.7808877468122921, "learning_rate": 3.6709756228258735e-05, "loss": 0.4782, "step": 312 }, { "epoch": 1.335820895522388, "grad_norm": 0.6561500406466976, "learning_rate": 3.667689305156406e-05, "loss": 0.3878, "step": 313 }, { "epoch": 1.340085287846482, "grad_norm": 0.5372106013768749, "learning_rate": 3.664388143251822e-05, "loss": 0.4225, "step": 314 }, { "epoch": 1.3443496801705757, "grad_norm": 0.6924199058045664, "learning_rate": 3.6610721664960236e-05, "loss": 0.4364, "step": 315 }, { "epoch": 1.3486140724946696, "grad_norm": 0.7015133870310188, "learning_rate": 3.657741404404785e-05, "loss": 0.4536, "step": 316 }, { "epoch": 1.3528784648187633, "grad_norm": 0.693862690594011, "learning_rate": 3.654395886625484e-05, "loss": 0.4069, "step": 317 }, { "epoch": 1.3571428571428572, "grad_norm": 0.6224295371607119, "learning_rate": 3.65103564293684e-05, "loss": 0.4283, "step": 318 }, { "epoch": 1.361407249466951, "grad_norm": 0.6738546833298403, "learning_rate": 3.647660703248651e-05, "loss": 0.4368, "step": 319 }, { "epoch": 1.3656716417910448, "grad_norm": 0.5857243004349567, "learning_rate": 3.644271097601521e-05, "loss": 0.4281, "step": 320 }, { "epoch": 1.3699360341151385, "grad_norm": 0.5585286307033677, "learning_rate": 3.640866856166601e-05, "loss": 0.3689, "step": 321 }, { "epoch": 1.3742004264392325, "grad_norm": 0.6876061087292902, "learning_rate": 3.637448009245315e-05, "loss": 0.4533, "step": 322 }, { "epoch": 1.3784648187633262, "grad_norm": 0.6596849434007818, "learning_rate": 3.63401458726909e-05, "loss": 0.4347, "step": 323 }, { "epoch": 1.38272921108742, "grad_norm": 0.6690336155185622, "learning_rate": 3.6305666207990886e-05, "loss": 0.4667, "step": 324 }, { "epoch": 1.3869936034115138, "grad_norm": 0.6173155021258911, "learning_rate": 3.6271041405259354e-05, "loss": 0.4212, "step": 325 }, { "epoch": 1.3912579957356077, "grad_norm": 0.632565166171892, "learning_rate": 3.623627177269441e-05, "loss": 0.4105, "step": 326 }, { "epoch": 1.3955223880597014, "grad_norm": 0.5784228743084681, "learning_rate": 3.6201357619783336e-05, "loss": 0.4114, "step": 327 }, { "epoch": 1.3997867803837953, "grad_norm": 0.5604209350319785, "learning_rate": 3.616629925729977e-05, "loss": 0.4415, "step": 328 }, { "epoch": 1.4040511727078893, "grad_norm": 0.6626347666514213, "learning_rate": 3.613109699730099e-05, "loss": 0.4459, "step": 329 }, { "epoch": 1.408315565031983, "grad_norm": 0.496782761321269, "learning_rate": 3.609575115312511e-05, "loss": 0.4213, "step": 330 }, { "epoch": 1.4125799573560767, "grad_norm": 0.6361067708225957, "learning_rate": 3.60602620393883e-05, "loss": 0.3538, "step": 331 }, { "epoch": 1.4168443496801706, "grad_norm": 0.6654635023457927, "learning_rate": 3.602462997198198e-05, "loss": 0.4233, "step": 332 }, { "epoch": 1.4211087420042645, "grad_norm": 0.6225178321156866, "learning_rate": 3.598885526807003e-05, "loss": 0.4167, "step": 333 }, { "epoch": 1.4253731343283582, "grad_norm": 0.6043453020450956, "learning_rate": 3.595293824608591e-05, "loss": 0.4121, "step": 334 }, { "epoch": 1.429637526652452, "grad_norm": 0.5986851241982426, "learning_rate": 3.591687922572991e-05, "loss": 0.4422, "step": 335 }, { "epoch": 1.4339019189765458, "grad_norm": 0.5894944313822743, "learning_rate": 3.5880678527966224e-05, "loss": 0.4221, "step": 336 }, { "epoch": 1.4381663113006398, "grad_norm": 0.49054383198031515, "learning_rate": 3.584433647502015e-05, "loss": 0.3841, "step": 337 }, { "epoch": 1.4424307036247335, "grad_norm": 0.5274462614351788, "learning_rate": 3.580785339037519e-05, "loss": 0.4065, "step": 338 }, { "epoch": 1.4466950959488272, "grad_norm": 0.5278549805650488, "learning_rate": 3.577122959877017e-05, "loss": 0.4572, "step": 339 }, { "epoch": 1.450959488272921, "grad_norm": 0.6317113335812257, "learning_rate": 3.57344654261964e-05, "loss": 0.4191, "step": 340 }, { "epoch": 1.455223880597015, "grad_norm": 0.5600572445916931, "learning_rate": 3.569756119989467e-05, "loss": 0.399, "step": 341 }, { "epoch": 1.4594882729211087, "grad_norm": 0.594198415593923, "learning_rate": 3.566051724835245e-05, "loss": 0.4455, "step": 342 }, { "epoch": 1.4637526652452024, "grad_norm": 0.564190038407502, "learning_rate": 3.562333390130089e-05, "loss": 0.3932, "step": 343 }, { "epoch": 1.4680170575692963, "grad_norm": 0.5954898940274361, "learning_rate": 3.5586011489711934e-05, "loss": 0.45, "step": 344 }, { "epoch": 1.4722814498933903, "grad_norm": 0.48911744248144157, "learning_rate": 3.554855034579532e-05, "loss": 0.3832, "step": 345 }, { "epoch": 1.476545842217484, "grad_norm": 0.6441473712847673, "learning_rate": 3.551095080299568e-05, "loss": 0.4369, "step": 346 }, { "epoch": 1.480810234541578, "grad_norm": 0.6183527348725328, "learning_rate": 3.5473213195989526e-05, "loss": 0.4255, "step": 347 }, { "epoch": 1.4850746268656716, "grad_norm": 0.6003497153907605, "learning_rate": 3.5435337860682304e-05, "loss": 0.4118, "step": 348 }, { "epoch": 1.4893390191897655, "grad_norm": 0.7004108419888099, "learning_rate": 3.539732513420538e-05, "loss": 0.4425, "step": 349 }, { "epoch": 1.4936034115138592, "grad_norm": 0.6873870156526409, "learning_rate": 3.535917535491306e-05, "loss": 0.4296, "step": 350 }, { "epoch": 1.4978678038379531, "grad_norm": 0.5528330417053992, "learning_rate": 3.532088886237956e-05, "loss": 0.4146, "step": 351 }, { "epoch": 1.502132196162047, "grad_norm": 0.6819580069714934, "learning_rate": 3.5282465997396e-05, "loss": 0.4364, "step": 352 }, { "epoch": 1.5063965884861408, "grad_norm": 0.7074884540484468, "learning_rate": 3.524390710196735e-05, "loss": 0.4194, "step": 353 }, { "epoch": 1.5106609808102345, "grad_norm": 0.5952225677619433, "learning_rate": 3.520521251930941e-05, "loss": 0.3989, "step": 354 }, { "epoch": 1.5149253731343284, "grad_norm": 0.6477164505835896, "learning_rate": 3.516638259384573e-05, "loss": 0.4493, "step": 355 }, { "epoch": 1.5191897654584223, "grad_norm": 0.6137858732201392, "learning_rate": 3.512741767120457e-05, "loss": 0.4015, "step": 356 }, { "epoch": 1.523454157782516, "grad_norm": 0.5341242976683102, "learning_rate": 3.5088318098215805e-05, "loss": 0.4403, "step": 357 }, { "epoch": 1.5277185501066097, "grad_norm": 0.5910346237420299, "learning_rate": 3.5049084222907846e-05, "loss": 0.4203, "step": 358 }, { "epoch": 1.5319829424307037, "grad_norm": 0.6750588762077021, "learning_rate": 3.5009716394504526e-05, "loss": 0.4366, "step": 359 }, { "epoch": 1.5362473347547976, "grad_norm": 0.6282869693485486, "learning_rate": 3.497021496342203e-05, "loss": 0.4029, "step": 360 }, { "epoch": 1.5405117270788913, "grad_norm": 0.4327732025523846, "learning_rate": 3.493058028126572e-05, "loss": 0.3786, "step": 361 }, { "epoch": 1.544776119402985, "grad_norm": 0.5628905924554551, "learning_rate": 3.4890812700827065e-05, "loss": 0.442, "step": 362 }, { "epoch": 1.549040511727079, "grad_norm": 0.7672190095876459, "learning_rate": 3.485091257608047e-05, "loss": 0.4599, "step": 363 }, { "epoch": 1.5533049040511728, "grad_norm": 0.4315110702270802, "learning_rate": 3.48108802621801e-05, "loss": 0.4105, "step": 364 }, { "epoch": 1.5575692963752665, "grad_norm": 0.699119311503924, "learning_rate": 3.477071611545676e-05, "loss": 0.4625, "step": 365 }, { "epoch": 1.5618336886993602, "grad_norm": 0.62580213083298, "learning_rate": 3.473042049341474e-05, "loss": 0.3939, "step": 366 }, { "epoch": 1.5660980810234542, "grad_norm": 0.41085317867894716, "learning_rate": 3.4689993754728564e-05, "loss": 0.4227, "step": 367 }, { "epoch": 1.570362473347548, "grad_norm": 0.8583066478775195, "learning_rate": 3.464943625923984e-05, "loss": 0.4174, "step": 368 }, { "epoch": 1.5746268656716418, "grad_norm": 0.7483607145625111, "learning_rate": 3.4608748367954064e-05, "loss": 0.4246, "step": 369 }, { "epoch": 1.5788912579957355, "grad_norm": 0.4588313921101154, "learning_rate": 3.45679304430374e-05, "loss": 0.44, "step": 370 }, { "epoch": 1.5831556503198294, "grad_norm": 0.8463537495188219, "learning_rate": 3.452698284781343e-05, "loss": 0.4707, "step": 371 }, { "epoch": 1.5874200426439233, "grad_norm": 0.6714451469463699, "learning_rate": 3.4485905946759965e-05, "loss": 0.4293, "step": 372 }, { "epoch": 1.591684434968017, "grad_norm": 0.49084942527038844, "learning_rate": 3.4444700105505766e-05, "loss": 0.387, "step": 373 }, { "epoch": 1.5959488272921107, "grad_norm": 0.7339552481136006, "learning_rate": 3.44033656908273e-05, "loss": 0.436, "step": 374 }, { "epoch": 1.6002132196162047, "grad_norm": 0.9668658455117516, "learning_rate": 3.4361903070645484e-05, "loss": 0.446, "step": 375 }, { "epoch": 1.6044776119402986, "grad_norm": 0.6437283133723504, "learning_rate": 3.432031261402239e-05, "loss": 0.4056, "step": 376 }, { "epoch": 1.6087420042643923, "grad_norm": 0.5471114373733622, "learning_rate": 3.4278594691157985e-05, "loss": 0.3993, "step": 377 }, { "epoch": 1.613006396588486, "grad_norm": 0.8359743208014425, "learning_rate": 3.423674967338681e-05, "loss": 0.4337, "step": 378 }, { "epoch": 1.61727078891258, "grad_norm": 0.6077166341043039, "learning_rate": 3.419477793317471e-05, "loss": 0.4377, "step": 379 }, { "epoch": 1.6215351812366738, "grad_norm": 0.4846179838526592, "learning_rate": 3.415267984411547e-05, "loss": 0.381, "step": 380 }, { "epoch": 1.6257995735607675, "grad_norm": 0.5563839923571956, "learning_rate": 3.411045578092754e-05, "loss": 0.452, "step": 381 }, { "epoch": 1.6300639658848612, "grad_norm": 0.5298650018801719, "learning_rate": 3.406810611945065e-05, "loss": 0.4023, "step": 382 }, { "epoch": 1.6343283582089554, "grad_norm": 0.47036826245422314, "learning_rate": 3.402563123664251e-05, "loss": 0.43, "step": 383 }, { "epoch": 1.638592750533049, "grad_norm": 0.5918256778605155, "learning_rate": 3.398303151057543e-05, "loss": 0.414, "step": 384 }, { "epoch": 1.6428571428571428, "grad_norm": 0.4724509860010177, "learning_rate": 3.394030732043296e-05, "loss": 0.4261, "step": 385 }, { "epoch": 1.6471215351812367, "grad_norm": 0.6438266470974083, "learning_rate": 3.389745904650652e-05, "loss": 0.4461, "step": 386 }, { "epoch": 1.6513859275053306, "grad_norm": 0.5814714611362957, "learning_rate": 3.385448707019199e-05, "loss": 0.4159, "step": 387 }, { "epoch": 1.6556503198294243, "grad_norm": 0.4206338335496547, "learning_rate": 3.381139177398636e-05, "loss": 0.4136, "step": 388 }, { "epoch": 1.659914712153518, "grad_norm": 0.653347902879806, "learning_rate": 3.3768173541484276e-05, "loss": 0.4152, "step": 389 }, { "epoch": 1.664179104477612, "grad_norm": 0.5717054823575618, "learning_rate": 3.372483275737468e-05, "loss": 0.4158, "step": 390 }, { "epoch": 1.668443496801706, "grad_norm": 0.567865027735176, "learning_rate": 3.3681369807437316e-05, "loss": 0.4328, "step": 391 }, { "epoch": 1.6727078891257996, "grad_norm": 0.5670521478372824, "learning_rate": 3.363778507853938e-05, "loss": 0.4398, "step": 392 }, { "epoch": 1.6769722814498933, "grad_norm": 0.6187350261747493, "learning_rate": 3.359407895863199e-05, "loss": 0.445, "step": 393 }, { "epoch": 1.6812366737739872, "grad_norm": 0.41177560006752595, "learning_rate": 3.35502518367468e-05, "loss": 0.4106, "step": 394 }, { "epoch": 1.6855010660980811, "grad_norm": 0.7844489084961632, "learning_rate": 3.350630410299252e-05, "loss": 0.416, "step": 395 }, { "epoch": 1.6897654584221748, "grad_norm": 0.5969945265152646, "learning_rate": 3.34622361485514e-05, "loss": 0.4537, "step": 396 }, { "epoch": 1.6940298507462686, "grad_norm": 0.6009377011808636, "learning_rate": 3.3418048365675835e-05, "loss": 0.3976, "step": 397 }, { "epoch": 1.6982942430703625, "grad_norm": 0.7970282912858717, "learning_rate": 3.3373741147684776e-05, "loss": 0.3958, "step": 398 }, { "epoch": 1.7025586353944564, "grad_norm": 0.51714913947718, "learning_rate": 3.332931488896029e-05, "loss": 0.4424, "step": 399 }, { "epoch": 1.70682302771855, "grad_norm": 0.6571324014445984, "learning_rate": 3.328476998494405e-05, "loss": 0.4293, "step": 400 }, { "epoch": 1.7110874200426438, "grad_norm": 0.6733750404894205, "learning_rate": 3.3240106832133784e-05, "loss": 0.4012, "step": 401 }, { "epoch": 1.7153518123667377, "grad_norm": 0.591500333207376, "learning_rate": 3.319532582807977e-05, "loss": 0.4224, "step": 402 }, { "epoch": 1.7196162046908317, "grad_norm": 0.6174698680618833, "learning_rate": 3.315042737138128e-05, "loss": 0.4114, "step": 403 }, { "epoch": 1.7238805970149254, "grad_norm": 0.5557934635630427, "learning_rate": 3.3105411861683046e-05, "loss": 0.4227, "step": 404 }, { "epoch": 1.728144989339019, "grad_norm": 0.5300688328065255, "learning_rate": 3.30602796996717e-05, "loss": 0.3967, "step": 405 }, { "epoch": 1.732409381663113, "grad_norm": 0.5760369736255456, "learning_rate": 3.3015031287072225e-05, "loss": 0.4386, "step": 406 }, { "epoch": 1.736673773987207, "grad_norm": 0.4555624191417979, "learning_rate": 3.296966702664431e-05, "loss": 0.4274, "step": 407 }, { "epoch": 1.7409381663113006, "grad_norm": 0.5059665427249744, "learning_rate": 3.2924187322178865e-05, "loss": 0.4145, "step": 408 }, { "epoch": 1.7452025586353943, "grad_norm": 0.46756504233420204, "learning_rate": 3.287859257849434e-05, "loss": 0.4309, "step": 409 }, { "epoch": 1.7494669509594882, "grad_norm": 0.46285792608796633, "learning_rate": 3.2832883201433185e-05, "loss": 0.3712, "step": 410 }, { "epoch": 1.7537313432835822, "grad_norm": 0.6245072491599375, "learning_rate": 3.278705959785821e-05, "loss": 0.4762, "step": 411 }, { "epoch": 1.7579957356076759, "grad_norm": 0.6061486243901444, "learning_rate": 3.274112217564892e-05, "loss": 0.4067, "step": 412 }, { "epoch": 1.7622601279317696, "grad_norm": 0.5820768104988486, "learning_rate": 3.2695071343697996e-05, "loss": 0.4124, "step": 413 }, { "epoch": 1.7665245202558635, "grad_norm": 0.6819975740897306, "learning_rate": 3.2648907511907544e-05, "loss": 0.4446, "step": 414 }, { "epoch": 1.7707889125799574, "grad_norm": 0.5715841364919834, "learning_rate": 3.2602631091185506e-05, "loss": 0.4548, "step": 415 }, { "epoch": 1.775053304904051, "grad_norm": 0.5021404496551074, "learning_rate": 3.255624249344198e-05, "loss": 0.415, "step": 416 }, { "epoch": 1.779317697228145, "grad_norm": 0.6236286206409429, "learning_rate": 3.250974213158555e-05, "loss": 0.418, "step": 417 }, { "epoch": 1.783582089552239, "grad_norm": 0.5082012644538463, "learning_rate": 3.246313041951966e-05, "loss": 0.4172, "step": 418 }, { "epoch": 1.7878464818763327, "grad_norm": 0.5804182502381373, "learning_rate": 3.241640777213884e-05, "loss": 0.4134, "step": 419 }, { "epoch": 1.7921108742004264, "grad_norm": 0.5841814227406251, "learning_rate": 3.23695746053251e-05, "loss": 0.3799, "step": 420 }, { "epoch": 1.7963752665245203, "grad_norm": 0.561217410497316, "learning_rate": 3.232263133594419e-05, "loss": 0.4553, "step": 421 }, { "epoch": 1.8006396588486142, "grad_norm": 0.5791964104930211, "learning_rate": 3.2275578381841876e-05, "loss": 0.4318, "step": 422 }, { "epoch": 1.804904051172708, "grad_norm": 0.5015381227163118, "learning_rate": 3.222841616184025e-05, "loss": 0.4287, "step": 423 }, { "epoch": 1.8091684434968016, "grad_norm": 0.4931963971340661, "learning_rate": 3.218114509573397e-05, "loss": 0.4177, "step": 424 }, { "epoch": 1.8134328358208955, "grad_norm": 0.4139642925732555, "learning_rate": 3.213376560428656e-05, "loss": 0.3862, "step": 425 }, { "epoch": 1.8176972281449895, "grad_norm": 0.6053235803447229, "learning_rate": 3.208627810922665e-05, "loss": 0.4278, "step": 426 }, { "epoch": 1.8219616204690832, "grad_norm": 0.4928980958456602, "learning_rate": 3.20386830332442e-05, "loss": 0.4042, "step": 427 }, { "epoch": 1.8262260127931769, "grad_norm": 0.5447428327419589, "learning_rate": 3.199098079998676e-05, "loss": 0.4338, "step": 428 }, { "epoch": 1.8304904051172708, "grad_norm": 0.630636350668545, "learning_rate": 3.194317183405573e-05, "loss": 0.407, "step": 429 }, { "epoch": 1.8347547974413647, "grad_norm": 0.4388934662903153, "learning_rate": 3.189525656100248e-05, "loss": 0.4002, "step": 430 }, { "epoch": 1.8390191897654584, "grad_norm": 0.5477299121670924, "learning_rate": 3.184723540732469e-05, "loss": 0.3893, "step": 431 }, { "epoch": 1.8432835820895521, "grad_norm": 0.5001645364202892, "learning_rate": 3.1799108800462466e-05, "loss": 0.4561, "step": 432 }, { "epoch": 1.847547974413646, "grad_norm": 0.56411219365906, "learning_rate": 3.175087716879454e-05, "loss": 0.4391, "step": 433 }, { "epoch": 1.85181236673774, "grad_norm": 0.49624337673620295, "learning_rate": 3.170254094163454e-05, "loss": 0.4447, "step": 434 }, { "epoch": 1.8560767590618337, "grad_norm": 0.48073525858910976, "learning_rate": 3.1654100549227024e-05, "loss": 0.4207, "step": 435 }, { "epoch": 1.8603411513859274, "grad_norm": 0.42948274359118516, "learning_rate": 3.160555642274382e-05, "loss": 0.3848, "step": 436 }, { "epoch": 1.8646055437100213, "grad_norm": 0.5376792952401593, "learning_rate": 3.155690899428002e-05, "loss": 0.4532, "step": 437 }, { "epoch": 1.8688699360341152, "grad_norm": 0.41636533485267063, "learning_rate": 3.1508158696850275e-05, "loss": 0.3942, "step": 438 }, { "epoch": 1.873134328358209, "grad_norm": 0.5319610981633921, "learning_rate": 3.1459305964384855e-05, "loss": 0.4106, "step": 439 }, { "epoch": 1.8773987206823026, "grad_norm": 0.448232868334879, "learning_rate": 3.141035123172581e-05, "loss": 0.4116, "step": 440 }, { "epoch": 1.8816631130063965, "grad_norm": 0.5880866733448852, "learning_rate": 3.136129493462312e-05, "loss": 0.4367, "step": 441 }, { "epoch": 1.8859275053304905, "grad_norm": 0.4930255466134952, "learning_rate": 3.1312137509730776e-05, "loss": 0.4176, "step": 442 }, { "epoch": 1.8901918976545842, "grad_norm": 0.5224716593111842, "learning_rate": 3.126287939460293e-05, "loss": 0.4358, "step": 443 }, { "epoch": 1.8944562899786779, "grad_norm": 0.5031275852344921, "learning_rate": 3.121352102768998e-05, "loss": 0.431, "step": 444 }, { "epoch": 1.8987206823027718, "grad_norm": 0.4599363811170101, "learning_rate": 3.116406284833468e-05, "loss": 0.3975, "step": 445 }, { "epoch": 1.9029850746268657, "grad_norm": 0.5204739120479832, "learning_rate": 3.111450529676821e-05, "loss": 0.4095, "step": 446 }, { "epoch": 1.9072494669509594, "grad_norm": 0.41426066836339737, "learning_rate": 3.106484881410628e-05, "loss": 0.4312, "step": 447 }, { "epoch": 1.9115138592750534, "grad_norm": 0.5350188027407833, "learning_rate": 3.101509384234519e-05, "loss": 0.3846, "step": 448 }, { "epoch": 1.9157782515991473, "grad_norm": 0.5541430552086132, "learning_rate": 3.09652408243579e-05, "loss": 0.4402, "step": 449 }, { "epoch": 1.920042643923241, "grad_norm": 0.45616595483342576, "learning_rate": 3.091529020389009e-05, "loss": 0.3975, "step": 450 }, { "epoch": 1.9243070362473347, "grad_norm": 0.47195887268241116, "learning_rate": 3.086524242555621e-05, "loss": 0.4186, "step": 451 }, { "epoch": 1.9285714285714286, "grad_norm": 0.417937083067313, "learning_rate": 3.081509793483551e-05, "loss": 0.3991, "step": 452 }, { "epoch": 1.9328358208955225, "grad_norm": 0.5147323724222759, "learning_rate": 3.076485717806808e-05, "loss": 0.4352, "step": 453 }, { "epoch": 1.9371002132196162, "grad_norm": 0.4721363439626508, "learning_rate": 3.0714520602450896e-05, "loss": 0.4399, "step": 454 }, { "epoch": 1.94136460554371, "grad_norm": 0.5204879106863163, "learning_rate": 3.066408865603383e-05, "loss": 0.4044, "step": 455 }, { "epoch": 1.9456289978678039, "grad_norm": 0.47869603996362364, "learning_rate": 3.061356178771564e-05, "loss": 0.4344, "step": 456 }, { "epoch": 1.9498933901918978, "grad_norm": 0.5487456033934804, "learning_rate": 3.056294044723999e-05, "loss": 0.3805, "step": 457 }, { "epoch": 1.9541577825159915, "grad_norm": 0.517435524986642, "learning_rate": 3.051222508519148e-05, "loss": 0.4757, "step": 458 }, { "epoch": 1.9584221748400852, "grad_norm": 0.5445809033478819, "learning_rate": 3.0461416152991555e-05, "loss": 0.3884, "step": 459 }, { "epoch": 1.962686567164179, "grad_norm": 0.5736641325670427, "learning_rate": 3.0410514102894582e-05, "loss": 0.4237, "step": 460 }, { "epoch": 1.966950959488273, "grad_norm": 0.5181389483618426, "learning_rate": 3.0359519387983738e-05, "loss": 0.4501, "step": 461 }, { "epoch": 1.9712153518123667, "grad_norm": 0.7055217266557069, "learning_rate": 3.0308432462167045e-05, "loss": 0.4302, "step": 462 }, { "epoch": 1.9754797441364604, "grad_norm": 0.45222781902353903, "learning_rate": 3.0257253780173293e-05, "loss": 0.3897, "step": 463 }, { "epoch": 1.9797441364605544, "grad_norm": 0.5399300307884194, "learning_rate": 3.0205983797548e-05, "loss": 0.4031, "step": 464 }, { "epoch": 1.9840085287846483, "grad_norm": 0.48971933733689127, "learning_rate": 3.015462297064936e-05, "loss": 0.4133, "step": 465 }, { "epoch": 1.988272921108742, "grad_norm": 0.44738322674726755, "learning_rate": 3.010317175664416e-05, "loss": 0.395, "step": 466 }, { "epoch": 1.9925373134328357, "grad_norm": 0.5055000394229842, "learning_rate": 3.0051630613503776e-05, "loss": 0.4032, "step": 467 }, { "epoch": 1.9968017057569296, "grad_norm": 0.4866897113581119, "learning_rate": 3.0000000000000004e-05, "loss": 0.4457, "step": 468 }, { "epoch": 2.002132196162047, "grad_norm": 1.003557894755282, "learning_rate": 2.994828037570104e-05, "loss": 0.7268, "step": 469 }, { "epoch": 2.0063965884861408, "grad_norm": 0.8084026414756201, "learning_rate": 2.9896472200967377e-05, "loss": 0.3024, "step": 470 }, { "epoch": 2.0106609808102345, "grad_norm": 0.9696863524057489, "learning_rate": 2.98445759369477e-05, "loss": 0.3435, "step": 471 }, { "epoch": 2.014925373134328, "grad_norm": 1.0289459178017686, "learning_rate": 2.9792592045574772e-05, "loss": 0.3344, "step": 472 }, { "epoch": 2.0191897654584223, "grad_norm": 0.6422882221374078, "learning_rate": 2.9740520989561357e-05, "loss": 0.3192, "step": 473 }, { "epoch": 2.023454157782516, "grad_norm": 0.7686763465029794, "learning_rate": 2.9688363232396056e-05, "loss": 0.3685, "step": 474 }, { "epoch": 2.0277185501066097, "grad_norm": 0.7619507919081873, "learning_rate": 2.9636119238339213e-05, "loss": 0.3232, "step": 475 }, { "epoch": 2.0319829424307034, "grad_norm": 0.7077722853700464, "learning_rate": 2.958378947241878e-05, "loss": 0.3433, "step": 476 }, { "epoch": 2.0362473347547976, "grad_norm": 0.9555687435185575, "learning_rate": 2.9531374400426158e-05, "loss": 0.346, "step": 477 }, { "epoch": 2.0405117270788913, "grad_norm": 0.7037304850265937, "learning_rate": 2.9478874488912078e-05, "loss": 0.3333, "step": 478 }, { "epoch": 2.044776119402985, "grad_norm": 0.674207846993653, "learning_rate": 2.9426290205182418e-05, "loss": 0.3334, "step": 479 }, { "epoch": 2.0490405117270787, "grad_norm": 0.7820884567891954, "learning_rate": 2.9373622017294075e-05, "loss": 0.3309, "step": 480 }, { "epoch": 2.053304904051173, "grad_norm": 0.5305149128764548, "learning_rate": 2.9320870394050783e-05, "loss": 0.3287, "step": 481 }, { "epoch": 2.0575692963752665, "grad_norm": 0.7079912872295963, "learning_rate": 2.926803580499892e-05, "loss": 0.3194, "step": 482 }, { "epoch": 2.0618336886993602, "grad_norm": 0.5541579502858093, "learning_rate": 2.9215118720423375e-05, "loss": 0.3444, "step": 483 }, { "epoch": 2.066098081023454, "grad_norm": 0.6513668190534649, "learning_rate": 2.9162119611343324e-05, "loss": 0.3327, "step": 484 }, { "epoch": 2.070362473347548, "grad_norm": 0.5294612918314424, "learning_rate": 2.910903894950805e-05, "loss": 0.343, "step": 485 }, { "epoch": 2.074626865671642, "grad_norm": 0.4752949935850779, "learning_rate": 2.9055877207392752e-05, "loss": 0.3062, "step": 486 }, { "epoch": 2.0788912579957355, "grad_norm": 0.4781620758478508, "learning_rate": 2.900263485819431e-05, "loss": 0.3214, "step": 487 }, { "epoch": 2.0831556503198296, "grad_norm": 0.548093488067997, "learning_rate": 2.8949312375827116e-05, "loss": 0.3296, "step": 488 }, { "epoch": 2.0874200426439233, "grad_norm": 0.4938614632595294, "learning_rate": 2.8895910234918828e-05, "loss": 0.3309, "step": 489 }, { "epoch": 2.091684434968017, "grad_norm": 0.5185353183120329, "learning_rate": 2.8842428910806143e-05, "loss": 0.3318, "step": 490 }, { "epoch": 2.0959488272921107, "grad_norm": 0.5799498196416928, "learning_rate": 2.878886887953059e-05, "loss": 0.3296, "step": 491 }, { "epoch": 2.100213219616205, "grad_norm": 0.4942313754302595, "learning_rate": 2.873523061783426e-05, "loss": 0.3175, "step": 492 }, { "epoch": 2.1044776119402986, "grad_norm": 0.5236125945451315, "learning_rate": 2.8681514603155575e-05, "loss": 0.3013, "step": 493 }, { "epoch": 2.1087420042643923, "grad_norm": 0.4518285850453316, "learning_rate": 2.8627721313625073e-05, "loss": 0.3316, "step": 494 }, { "epoch": 2.113006396588486, "grad_norm": 0.5126760484367925, "learning_rate": 2.8573851228061084e-05, "loss": 0.3355, "step": 495 }, { "epoch": 2.11727078891258, "grad_norm": 0.5029294170658799, "learning_rate": 2.851990482596554e-05, "loss": 0.3416, "step": 496 }, { "epoch": 2.121535181236674, "grad_norm": 0.4288555624298432, "learning_rate": 2.846588258751964e-05, "loss": 0.3195, "step": 497 }, { "epoch": 2.1257995735607675, "grad_norm": 0.4298499268015924, "learning_rate": 2.8411784993579633e-05, "loss": 0.3052, "step": 498 }, { "epoch": 2.1300639658848612, "grad_norm": 0.4671221612093904, "learning_rate": 2.8357612525672503e-05, "loss": 0.3337, "step": 499 }, { "epoch": 2.1343283582089554, "grad_norm": 0.5498904625760632, "learning_rate": 2.830336566599169e-05, "loss": 0.3416, "step": 500 }, { "epoch": 2.138592750533049, "grad_norm": 0.40997801357104674, "learning_rate": 2.8249044897392814e-05, "loss": 0.3216, "step": 501 }, { "epoch": 2.142857142857143, "grad_norm": 0.5165614467304281, "learning_rate": 2.819465070338935e-05, "loss": 0.3189, "step": 502 }, { "epoch": 2.1471215351812365, "grad_norm": 0.37915760405651994, "learning_rate": 2.814018356814834e-05, "loss": 0.324, "step": 503 }, { "epoch": 2.1513859275053306, "grad_norm": 0.6558262452058821, "learning_rate": 2.80856439764861e-05, "loss": 0.3607, "step": 504 }, { "epoch": 2.1556503198294243, "grad_norm": 0.4499344644834719, "learning_rate": 2.8031032413863863e-05, "loss": 0.3473, "step": 505 }, { "epoch": 2.159914712153518, "grad_norm": 0.49760589957195217, "learning_rate": 2.7976349366383487e-05, "loss": 0.3408, "step": 506 }, { "epoch": 2.1641791044776117, "grad_norm": 0.4523387967822899, "learning_rate": 2.792159532078314e-05, "loss": 0.3222, "step": 507 }, { "epoch": 2.168443496801706, "grad_norm": 0.4724751053334817, "learning_rate": 2.786677076443292e-05, "loss": 0.3405, "step": 508 }, { "epoch": 2.1727078891257996, "grad_norm": 0.4128778956844833, "learning_rate": 2.7811876185330566e-05, "loss": 0.3303, "step": 509 }, { "epoch": 2.1769722814498933, "grad_norm": 0.4598923533907129, "learning_rate": 2.77569120720971e-05, "loss": 0.3263, "step": 510 }, { "epoch": 2.181236673773987, "grad_norm": 0.4398641540643089, "learning_rate": 2.770187891397246e-05, "loss": 0.3357, "step": 511 }, { "epoch": 2.185501066098081, "grad_norm": 0.39972559010107306, "learning_rate": 2.764677720081116e-05, "loss": 0.3167, "step": 512 }, { "epoch": 2.189765458422175, "grad_norm": 0.39291427360339126, "learning_rate": 2.7591607423077932e-05, "loss": 0.2892, "step": 513 }, { "epoch": 2.1940298507462686, "grad_norm": 0.43942382097226057, "learning_rate": 2.7536370071843353e-05, "loss": 0.3463, "step": 514 }, { "epoch": 2.1982942430703627, "grad_norm": 0.455479906539244, "learning_rate": 2.748106563877948e-05, "loss": 0.3309, "step": 515 }, { "epoch": 2.2025586353944564, "grad_norm": 0.47069509530231296, "learning_rate": 2.7425694616155474e-05, "loss": 0.3302, "step": 516 }, { "epoch": 2.20682302771855, "grad_norm": 0.46486150824863315, "learning_rate": 2.7370257496833206e-05, "loss": 0.3247, "step": 517 }, { "epoch": 2.211087420042644, "grad_norm": 0.52520003158657, "learning_rate": 2.7314754774262885e-05, "loss": 0.33, "step": 518 }, { "epoch": 2.2153518123667375, "grad_norm": 0.4133516757830205, "learning_rate": 2.7259186942478656e-05, "loss": 0.3136, "step": 519 }, { "epoch": 2.2196162046908317, "grad_norm": 0.4584557324610895, "learning_rate": 2.720355449609421e-05, "loss": 0.3336, "step": 520 }, { "epoch": 2.2238805970149254, "grad_norm": 0.5088846176274742, "learning_rate": 2.7147857930298377e-05, "loss": 0.3503, "step": 521 }, { "epoch": 2.228144989339019, "grad_norm": 0.4105166262617953, "learning_rate": 2.7092097740850712e-05, "loss": 0.372, "step": 522 }, { "epoch": 2.232409381663113, "grad_norm": 0.48064341561246865, "learning_rate": 2.7036274424077107e-05, "loss": 0.3284, "step": 523 }, { "epoch": 2.236673773987207, "grad_norm": 0.43023264705358255, "learning_rate": 2.698038847686533e-05, "loss": 0.3209, "step": 524 }, { "epoch": 2.2409381663113006, "grad_norm": 0.41178473939160515, "learning_rate": 2.692444039666066e-05, "loss": 0.3064, "step": 525 }, { "epoch": 2.2452025586353943, "grad_norm": 0.488463118984366, "learning_rate": 2.6868430681461385e-05, "loss": 0.3322, "step": 526 }, { "epoch": 2.2494669509594885, "grad_norm": 0.4149213055208161, "learning_rate": 2.681235982981444e-05, "loss": 0.3559, "step": 527 }, { "epoch": 2.253731343283582, "grad_norm": 0.4817511814973644, "learning_rate": 2.6756228340810946e-05, "loss": 0.3055, "step": 528 }, { "epoch": 2.257995735607676, "grad_norm": 0.39070396332909846, "learning_rate": 2.670003671408174e-05, "loss": 0.3168, "step": 529 }, { "epoch": 2.2622601279317696, "grad_norm": 0.5203601542790566, "learning_rate": 2.6643785449792967e-05, "loss": 0.3282, "step": 530 }, { "epoch": 2.2665245202558637, "grad_norm": 0.41130909380747965, "learning_rate": 2.6587475048641596e-05, "loss": 0.3468, "step": 531 }, { "epoch": 2.2707889125799574, "grad_norm": 0.4364319149963848, "learning_rate": 2.6531106011850988e-05, "loss": 0.2883, "step": 532 }, { "epoch": 2.275053304904051, "grad_norm": 0.4314737959819764, "learning_rate": 2.6474678841166426e-05, "loss": 0.3049, "step": 533 }, { "epoch": 2.279317697228145, "grad_norm": 0.558103408427805, "learning_rate": 2.6418194038850634e-05, "loss": 0.3682, "step": 534 }, { "epoch": 2.283582089552239, "grad_norm": 0.4584428049233054, "learning_rate": 2.6361652107679335e-05, "loss": 0.3401, "step": 535 }, { "epoch": 2.2878464818763327, "grad_norm": 0.3837451796840203, "learning_rate": 2.630505355093676e-05, "loss": 0.3038, "step": 536 }, { "epoch": 2.2921108742004264, "grad_norm": 0.48972122254900113, "learning_rate": 2.624839887241115e-05, "loss": 0.3486, "step": 537 }, { "epoch": 2.29637526652452, "grad_norm": 0.375527961031895, "learning_rate": 2.619168857639032e-05, "loss": 0.3316, "step": 538 }, { "epoch": 2.300639658848614, "grad_norm": 0.45411810854625184, "learning_rate": 2.6134923167657106e-05, "loss": 0.3169, "step": 539 }, { "epoch": 2.304904051172708, "grad_norm": 0.42531128407316904, "learning_rate": 2.607810315148494e-05, "loss": 0.346, "step": 540 }, { "epoch": 2.3091684434968016, "grad_norm": 0.41992642350814313, "learning_rate": 2.6021229033633303e-05, "loss": 0.3089, "step": 541 }, { "epoch": 2.3134328358208958, "grad_norm": 0.41898615604904854, "learning_rate": 2.5964301320343227e-05, "loss": 0.3595, "step": 542 }, { "epoch": 2.3176972281449895, "grad_norm": 0.4177885547418394, "learning_rate": 2.5907320518332827e-05, "loss": 0.313, "step": 543 }, { "epoch": 2.321961620469083, "grad_norm": 0.38229897718006295, "learning_rate": 2.5850287134792747e-05, "loss": 0.321, "step": 544 }, { "epoch": 2.326226012793177, "grad_norm": 0.4041612241717893, "learning_rate": 2.5793201677381675e-05, "loss": 0.3418, "step": 545 }, { "epoch": 2.3304904051172706, "grad_norm": 0.4274720630355263, "learning_rate": 2.5736064654221808e-05, "loss": 0.336, "step": 546 }, { "epoch": 2.3347547974413647, "grad_norm": 0.3861871646004023, "learning_rate": 2.5678876573894326e-05, "loss": 0.3482, "step": 547 }, { "epoch": 2.3390191897654584, "grad_norm": 0.4077636527683995, "learning_rate": 2.5621637945434876e-05, "loss": 0.3286, "step": 548 }, { "epoch": 2.343283582089552, "grad_norm": 0.3756224596945538, "learning_rate": 2.5564349278329056e-05, "loss": 0.3001, "step": 549 }, { "epoch": 2.3475479744136463, "grad_norm": 0.4109819828033757, "learning_rate": 2.5507011082507834e-05, "loss": 0.3305, "step": 550 }, { "epoch": 2.35181236673774, "grad_norm": 0.4383761223461391, "learning_rate": 2.544962386834306e-05, "loss": 0.3083, "step": 551 }, { "epoch": 2.3560767590618337, "grad_norm": 0.41464003342514316, "learning_rate": 2.539218814664288e-05, "loss": 0.349, "step": 552 }, { "epoch": 2.3603411513859274, "grad_norm": 0.5062569973089363, "learning_rate": 2.5334704428647227e-05, "loss": 0.3302, "step": 553 }, { "epoch": 2.364605543710021, "grad_norm": 0.36480421490785353, "learning_rate": 2.527717322602325e-05, "loss": 0.3174, "step": 554 }, { "epoch": 2.368869936034115, "grad_norm": 0.41474113980895305, "learning_rate": 2.521959505086075e-05, "loss": 0.3671, "step": 555 }, { "epoch": 2.373134328358209, "grad_norm": 0.36718360788116927, "learning_rate": 2.516197041566764e-05, "loss": 0.2863, "step": 556 }, { "epoch": 2.3773987206823026, "grad_norm": 0.42920079467315114, "learning_rate": 2.510429983336539e-05, "loss": 0.354, "step": 557 }, { "epoch": 2.3816631130063968, "grad_norm": 0.46613801110394354, "learning_rate": 2.5046583817284437e-05, "loss": 0.3276, "step": 558 }, { "epoch": 2.3859275053304905, "grad_norm": 0.35311255938462993, "learning_rate": 2.4988822881159627e-05, "loss": 0.3059, "step": 559 }, { "epoch": 2.390191897654584, "grad_norm": 0.39778364103851305, "learning_rate": 2.4931017539125648e-05, "loss": 0.3399, "step": 560 }, { "epoch": 2.394456289978678, "grad_norm": 0.3969131330840867, "learning_rate": 2.487316830571244e-05, "loss": 0.3441, "step": 561 }, { "epoch": 2.398720682302772, "grad_norm": 0.3998460096155394, "learning_rate": 2.481527569584064e-05, "loss": 0.347, "step": 562 }, { "epoch": 2.4029850746268657, "grad_norm": 0.3724128051878021, "learning_rate": 2.4757340224816964e-05, "loss": 0.3212, "step": 563 }, { "epoch": 2.4072494669509594, "grad_norm": 0.4014481595437383, "learning_rate": 2.4699362408329646e-05, "loss": 0.3335, "step": 564 }, { "epoch": 2.411513859275053, "grad_norm": 0.42703172622134855, "learning_rate": 2.4641342762443845e-05, "loss": 0.3464, "step": 565 }, { "epoch": 2.4157782515991473, "grad_norm": 0.3889377354808822, "learning_rate": 2.458328180359704e-05, "loss": 0.3391, "step": 566 }, { "epoch": 2.420042643923241, "grad_norm": 0.3966463724749422, "learning_rate": 2.4525180048594452e-05, "loss": 0.3274, "step": 567 }, { "epoch": 2.4243070362473347, "grad_norm": 0.37604211049525493, "learning_rate": 2.4467038014604402e-05, "loss": 0.3301, "step": 568 }, { "epoch": 2.4285714285714284, "grad_norm": 0.3860914267444212, "learning_rate": 2.4408856219153775e-05, "loss": 0.3415, "step": 569 }, { "epoch": 2.4328358208955225, "grad_norm": 0.4195346230813679, "learning_rate": 2.435063518012335e-05, "loss": 0.3278, "step": 570 }, { "epoch": 2.4371002132196162, "grad_norm": 0.4641696550558176, "learning_rate": 2.4292375415743235e-05, "loss": 0.3766, "step": 571 }, { "epoch": 2.44136460554371, "grad_norm": 0.4055853025049748, "learning_rate": 2.423407744458822e-05, "loss": 0.3359, "step": 572 }, { "epoch": 2.4456289978678036, "grad_norm": 0.38176150606644776, "learning_rate": 2.4175741785573177e-05, "loss": 0.3068, "step": 573 }, { "epoch": 2.449893390191898, "grad_norm": 0.4069042605717178, "learning_rate": 2.4117368957948452e-05, "loss": 0.3253, "step": 574 }, { "epoch": 2.4541577825159915, "grad_norm": 0.3860477934658931, "learning_rate": 2.4058959481295226e-05, "loss": 0.3612, "step": 575 }, { "epoch": 2.458422174840085, "grad_norm": 0.36453620340817394, "learning_rate": 2.4000513875520892e-05, "loss": 0.2868, "step": 576 }, { "epoch": 2.4626865671641793, "grad_norm": 0.38667475042155025, "learning_rate": 2.3942032660854434e-05, "loss": 0.3264, "step": 577 }, { "epoch": 2.466950959488273, "grad_norm": 0.3810779692705642, "learning_rate": 2.3883516357841797e-05, "loss": 0.3502, "step": 578 }, { "epoch": 2.4712153518123667, "grad_norm": 0.4029065335258986, "learning_rate": 2.3824965487341247e-05, "loss": 0.3164, "step": 579 }, { "epoch": 2.4754797441364604, "grad_norm": 0.3641659784797967, "learning_rate": 2.376638057051874e-05, "loss": 0.3689, "step": 580 }, { "epoch": 2.479744136460554, "grad_norm": 0.386320924313445, "learning_rate": 2.370776212884327e-05, "loss": 0.3223, "step": 581 }, { "epoch": 2.4840085287846483, "grad_norm": 0.3683303725219809, "learning_rate": 2.3649110684082258e-05, "loss": 0.3419, "step": 582 }, { "epoch": 2.488272921108742, "grad_norm": 0.3400504068210789, "learning_rate": 2.3590426758296873e-05, "loss": 0.3319, "step": 583 }, { "epoch": 2.4925373134328357, "grad_norm": 0.36853362002125634, "learning_rate": 2.35317108738374e-05, "loss": 0.3407, "step": 584 }, { "epoch": 2.49680170575693, "grad_norm": 0.44453763314473116, "learning_rate": 2.3472963553338614e-05, "loss": 0.3664, "step": 585 }, { "epoch": 2.5010660980810235, "grad_norm": 0.3896349047477031, "learning_rate": 2.3414185319715062e-05, "loss": 0.3057, "step": 586 }, { "epoch": 2.5053304904051172, "grad_norm": 0.4037747109455124, "learning_rate": 2.335537669615649e-05, "loss": 0.32, "step": 587 }, { "epoch": 2.509594882729211, "grad_norm": 0.42757034531523247, "learning_rate": 2.3296538206123134e-05, "loss": 0.3533, "step": 588 }, { "epoch": 2.5138592750533046, "grad_norm": 0.3590031237695879, "learning_rate": 2.3237670373341065e-05, "loss": 0.3362, "step": 589 }, { "epoch": 2.518123667377399, "grad_norm": 0.3564034419723339, "learning_rate": 2.3178773721797546e-05, "loss": 0.3356, "step": 590 }, { "epoch": 2.5223880597014925, "grad_norm": 0.3772097468768756, "learning_rate": 2.311984877573636e-05, "loss": 0.3423, "step": 591 }, { "epoch": 2.526652452025586, "grad_norm": 0.3572311012517997, "learning_rate": 2.3060896059653133e-05, "loss": 0.3298, "step": 592 }, { "epoch": 2.5309168443496803, "grad_norm": 0.48695055246766855, "learning_rate": 2.3001916098290684e-05, "loss": 0.3562, "step": 593 }, { "epoch": 2.535181236673774, "grad_norm": 0.36634244613995803, "learning_rate": 2.2942909416634326e-05, "loss": 0.3412, "step": 594 }, { "epoch": 2.5394456289978677, "grad_norm": 0.44425887232294786, "learning_rate": 2.2883876539907238e-05, "loss": 0.3492, "step": 595 }, { "epoch": 2.543710021321962, "grad_norm": 0.3804698065592149, "learning_rate": 2.2824817993565743e-05, "loss": 0.3192, "step": 596 }, { "epoch": 2.5479744136460556, "grad_norm": 0.4056116969076533, "learning_rate": 2.2765734303294666e-05, "loss": 0.3225, "step": 597 }, { "epoch": 2.5522388059701493, "grad_norm": 0.35778597376945903, "learning_rate": 2.2706625995002626e-05, "loss": 0.318, "step": 598 }, { "epoch": 2.556503198294243, "grad_norm": 0.38168555414145255, "learning_rate": 2.2647493594817375e-05, "loss": 0.3601, "step": 599 }, { "epoch": 2.5607675906183367, "grad_norm": 0.37162055381097153, "learning_rate": 2.2588337629081107e-05, "loss": 0.3131, "step": 600 }, { "epoch": 2.565031982942431, "grad_norm": 0.39569822220724193, "learning_rate": 2.2529158624345793e-05, "loss": 0.3355, "step": 601 }, { "epoch": 2.5692963752665245, "grad_norm": 0.3683693883610303, "learning_rate": 2.246995710736844e-05, "loss": 0.3281, "step": 602 }, { "epoch": 2.5735607675906182, "grad_norm": 0.3638653113004468, "learning_rate": 2.2410733605106462e-05, "loss": 0.3473, "step": 603 }, { "epoch": 2.5778251599147124, "grad_norm": 0.3432942852936682, "learning_rate": 2.2351488644712977e-05, "loss": 0.3174, "step": 604 }, { "epoch": 2.582089552238806, "grad_norm": 0.36441663266011637, "learning_rate": 2.2292222753532067e-05, "loss": 0.3404, "step": 605 }, { "epoch": 2.5863539445629, "grad_norm": 0.34359099977889773, "learning_rate": 2.2232936459094158e-05, "loss": 0.2855, "step": 606 }, { "epoch": 2.5906183368869935, "grad_norm": 0.41180627075263315, "learning_rate": 2.2173630289111267e-05, "loss": 0.3292, "step": 607 }, { "epoch": 2.594882729211087, "grad_norm": 0.40856953637927446, "learning_rate": 2.211430477147232e-05, "loss": 0.3148, "step": 608 }, { "epoch": 2.5991471215351813, "grad_norm": 0.4143158196510196, "learning_rate": 2.205496043423849e-05, "loss": 0.3582, "step": 609 }, { "epoch": 2.603411513859275, "grad_norm": 0.3688325646692266, "learning_rate": 2.1995597805638415e-05, "loss": 0.3033, "step": 610 }, { "epoch": 2.6076759061833688, "grad_norm": 0.44966974266720894, "learning_rate": 2.1936217414063584e-05, "loss": 0.336, "step": 611 }, { "epoch": 2.611940298507463, "grad_norm": 0.35613612940024736, "learning_rate": 2.1876819788063586e-05, "loss": 0.3155, "step": 612 }, { "epoch": 2.6162046908315566, "grad_norm": 0.4111271913056953, "learning_rate": 2.1817405456341412e-05, "loss": 0.3383, "step": 613 }, { "epoch": 2.6204690831556503, "grad_norm": 0.3659202167790694, "learning_rate": 2.1757974947748762e-05, "loss": 0.3215, "step": 614 }, { "epoch": 2.624733475479744, "grad_norm": 0.3856052575394882, "learning_rate": 2.16985287912813e-05, "loss": 0.31, "step": 615 }, { "epoch": 2.6289978678038377, "grad_norm": 0.41136358511221127, "learning_rate": 2.1639067516074004e-05, "loss": 0.3251, "step": 616 }, { "epoch": 2.633262260127932, "grad_norm": 0.3477988590548033, "learning_rate": 2.1579591651396412e-05, "loss": 0.2935, "step": 617 }, { "epoch": 2.6375266524520256, "grad_norm": 0.4337010275602838, "learning_rate": 2.1520101726647922e-05, "loss": 0.361, "step": 618 }, { "epoch": 2.6417910447761193, "grad_norm": 0.4624391197853646, "learning_rate": 2.146059827135309e-05, "loss": 0.333, "step": 619 }, { "epoch": 2.6460554371002134, "grad_norm": 0.3727233265161806, "learning_rate": 2.1401081815156894e-05, "loss": 0.296, "step": 620 }, { "epoch": 2.650319829424307, "grad_norm": 0.44643626666607605, "learning_rate": 2.1341552887820048e-05, "loss": 0.3495, "step": 621 }, { "epoch": 2.654584221748401, "grad_norm": 0.3970268512437942, "learning_rate": 2.1282012019214275e-05, "loss": 0.3386, "step": 622 }, { "epoch": 2.6588486140724945, "grad_norm": 0.37620787661454785, "learning_rate": 2.122245973931757e-05, "loss": 0.3234, "step": 623 }, { "epoch": 2.663113006396588, "grad_norm": 0.4494049986409132, "learning_rate": 2.1162896578209517e-05, "loss": 0.3425, "step": 624 }, { "epoch": 2.6673773987206824, "grad_norm": 0.3300487285841384, "learning_rate": 2.110332306606655e-05, "loss": 0.3558, "step": 625 }, { "epoch": 2.671641791044776, "grad_norm": 0.4730585433014445, "learning_rate": 2.104373973315724e-05, "loss": 0.3596, "step": 626 }, { "epoch": 2.6759061833688698, "grad_norm": 0.31988464666714095, "learning_rate": 2.0984147109837564e-05, "loss": 0.3197, "step": 627 }, { "epoch": 2.680170575692964, "grad_norm": 0.4024194409780812, "learning_rate": 2.092454572654621e-05, "loss": 0.3419, "step": 628 }, { "epoch": 2.6844349680170576, "grad_norm": 0.4185283654543116, "learning_rate": 2.0864936113799818e-05, "loss": 0.3554, "step": 629 }, { "epoch": 2.6886993603411513, "grad_norm": 0.4659433660171756, "learning_rate": 2.0805318802188307e-05, "loss": 0.3266, "step": 630 }, { "epoch": 2.6929637526652455, "grad_norm": 0.3663919667117319, "learning_rate": 2.0745694322370095e-05, "loss": 0.3287, "step": 631 }, { "epoch": 2.697228144989339, "grad_norm": 0.451511426664104, "learning_rate": 2.0686063205067426e-05, "loss": 0.3538, "step": 632 }, { "epoch": 2.701492537313433, "grad_norm": 0.3626527728578006, "learning_rate": 2.0626425981061608e-05, "loss": 0.3372, "step": 633 }, { "epoch": 2.7057569296375266, "grad_norm": 0.4221364252043816, "learning_rate": 2.056678318118832e-05, "loss": 0.3609, "step": 634 }, { "epoch": 2.7100213219616203, "grad_norm": 0.3636772833343896, "learning_rate": 2.050713533633287e-05, "loss": 0.3004, "step": 635 }, { "epoch": 2.7142857142857144, "grad_norm": 0.3508912160194277, "learning_rate": 2.0447482977425465e-05, "loss": 0.3473, "step": 636 }, { "epoch": 2.718550106609808, "grad_norm": 0.40925231854920796, "learning_rate": 2.038782663543649e-05, "loss": 0.3616, "step": 637 }, { "epoch": 2.722814498933902, "grad_norm": 0.3613534312982199, "learning_rate": 2.0328166841371796e-05, "loss": 0.3297, "step": 638 }, { "epoch": 2.727078891257996, "grad_norm": 0.38532149432963253, "learning_rate": 2.0268504126267952e-05, "loss": 0.3448, "step": 639 }, { "epoch": 2.7313432835820897, "grad_norm": 0.3267536319660621, "learning_rate": 2.020883902118753e-05, "loss": 0.3176, "step": 640 }, { "epoch": 2.7356076759061834, "grad_norm": 0.34048997891320393, "learning_rate": 2.014917205721437e-05, "loss": 0.3182, "step": 641 }, { "epoch": 2.739872068230277, "grad_norm": 0.3574451853759449, "learning_rate": 2.008950376544887e-05, "loss": 0.3194, "step": 642 }, { "epoch": 2.7441364605543708, "grad_norm": 0.3815211412430124, "learning_rate": 2.0029834677003235e-05, "loss": 0.3767, "step": 643 }, { "epoch": 2.748400852878465, "grad_norm": 0.34767904398176336, "learning_rate": 1.9970165322996768e-05, "loss": 0.3256, "step": 644 }, { "epoch": 2.7526652452025586, "grad_norm": 0.38620602061788006, "learning_rate": 1.9910496234551132e-05, "loss": 0.3109, "step": 645 }, { "epoch": 2.7569296375266523, "grad_norm": 0.3924985087086562, "learning_rate": 1.985082794278563e-05, "loss": 0.3488, "step": 646 }, { "epoch": 2.7611940298507465, "grad_norm": 0.3511647263545227, "learning_rate": 1.979116097881248e-05, "loss": 0.3176, "step": 647 }, { "epoch": 2.76545842217484, "grad_norm": 0.3804886933596516, "learning_rate": 1.9731495873732055e-05, "loss": 0.3346, "step": 648 }, { "epoch": 2.769722814498934, "grad_norm": 0.3711827433144586, "learning_rate": 1.967183315862821e-05, "loss": 0.3245, "step": 649 }, { "epoch": 2.7739872068230276, "grad_norm": 0.4040189766473357, "learning_rate": 1.9612173364563517e-05, "loss": 0.3413, "step": 650 }, { "epoch": 2.7782515991471213, "grad_norm": 0.35670183847640413, "learning_rate": 1.9552517022574542e-05, "loss": 0.3335, "step": 651 }, { "epoch": 2.7825159914712154, "grad_norm": 0.3375759136279141, "learning_rate": 1.9492864663667135e-05, "loss": 0.2882, "step": 652 }, { "epoch": 2.786780383795309, "grad_norm": 0.3975695376663641, "learning_rate": 1.9433216818811686e-05, "loss": 0.3295, "step": 653 }, { "epoch": 2.791044776119403, "grad_norm": 0.3290654794586691, "learning_rate": 1.93735740189384e-05, "loss": 0.3169, "step": 654 }, { "epoch": 2.795309168443497, "grad_norm": 0.37186538014031906, "learning_rate": 1.931393679493258e-05, "loss": 0.3502, "step": 655 }, { "epoch": 2.7995735607675907, "grad_norm": 0.35915428065348987, "learning_rate": 1.925430567762991e-05, "loss": 0.3391, "step": 656 }, { "epoch": 2.8038379530916844, "grad_norm": 0.37547914494206946, "learning_rate": 1.9194681197811703e-05, "loss": 0.3224, "step": 657 }, { "epoch": 2.8081023454157785, "grad_norm": 0.36867525713174365, "learning_rate": 1.9135063886200186e-05, "loss": 0.339, "step": 658 }, { "epoch": 2.8123667377398722, "grad_norm": 0.35008429854386747, "learning_rate": 1.9075454273453797e-05, "loss": 0.3089, "step": 659 }, { "epoch": 2.816631130063966, "grad_norm": 0.39355064952010693, "learning_rate": 1.901585289016244e-05, "loss": 0.3641, "step": 660 }, { "epoch": 2.8208955223880596, "grad_norm": 0.3537373717689814, "learning_rate": 1.8956260266842762e-05, "loss": 0.3354, "step": 661 }, { "epoch": 2.8251599147121533, "grad_norm": 0.41391996442389295, "learning_rate": 1.8896676933933455e-05, "loss": 0.3491, "step": 662 }, { "epoch": 2.8294243070362475, "grad_norm": 0.3569482915635849, "learning_rate": 1.8837103421790486e-05, "loss": 0.3226, "step": 663 }, { "epoch": 2.833688699360341, "grad_norm": 0.37516722859555784, "learning_rate": 1.8777540260682433e-05, "loss": 0.302, "step": 664 }, { "epoch": 2.837953091684435, "grad_norm": 0.4248830795261121, "learning_rate": 1.8717987980785728e-05, "loss": 0.3519, "step": 665 }, { "epoch": 2.842217484008529, "grad_norm": 0.37783359614073125, "learning_rate": 1.8658447112179952e-05, "loss": 0.321, "step": 666 }, { "epoch": 2.8464818763326227, "grad_norm": 0.41802490330254377, "learning_rate": 1.8598918184843116e-05, "loss": 0.3183, "step": 667 }, { "epoch": 2.8507462686567164, "grad_norm": 0.3718331088787641, "learning_rate": 1.8539401728646923e-05, "loss": 0.3411, "step": 668 }, { "epoch": 2.85501066098081, "grad_norm": 0.44945631637607075, "learning_rate": 1.8479898273352084e-05, "loss": 0.3516, "step": 669 }, { "epoch": 2.859275053304904, "grad_norm": 0.3579317342747698, "learning_rate": 1.8420408348603595e-05, "loss": 0.3389, "step": 670 }, { "epoch": 2.863539445628998, "grad_norm": 0.36099208067311495, "learning_rate": 1.8360932483926e-05, "loss": 0.3241, "step": 671 }, { "epoch": 2.8678038379530917, "grad_norm": 0.47001526249967157, "learning_rate": 1.83014712087187e-05, "loss": 0.3337, "step": 672 }, { "epoch": 2.8720682302771854, "grad_norm": 0.3714088126438469, "learning_rate": 1.824202505225125e-05, "loss": 0.3279, "step": 673 }, { "epoch": 2.8763326226012795, "grad_norm": 0.4175307068350295, "learning_rate": 1.818259454365859e-05, "loss": 0.3735, "step": 674 }, { "epoch": 2.8805970149253732, "grad_norm": 0.3550834520314073, "learning_rate": 1.8123180211936417e-05, "loss": 0.3165, "step": 675 }, { "epoch": 2.884861407249467, "grad_norm": 0.4383311757544316, "learning_rate": 1.806378258593642e-05, "loss": 0.3508, "step": 676 }, { "epoch": 2.8891257995735606, "grad_norm": 0.39587309166564166, "learning_rate": 1.800440219436159e-05, "loss": 0.3356, "step": 677 }, { "epoch": 2.8933901918976543, "grad_norm": 0.3556833358344844, "learning_rate": 1.794503956576152e-05, "loss": 0.3129, "step": 678 }, { "epoch": 2.8976545842217485, "grad_norm": 0.49312145714140465, "learning_rate": 1.7885695228527683e-05, "loss": 0.3646, "step": 679 }, { "epoch": 2.901918976545842, "grad_norm": 0.4062615415192085, "learning_rate": 1.782636971088874e-05, "loss": 0.3253, "step": 680 }, { "epoch": 2.906183368869936, "grad_norm": 0.3933273488643739, "learning_rate": 1.776706354090585e-05, "loss": 0.3456, "step": 681 }, { "epoch": 2.91044776119403, "grad_norm": 0.43585170578526733, "learning_rate": 1.7707777246467933e-05, "loss": 0.358, "step": 682 }, { "epoch": 2.9147121535181237, "grad_norm": 0.40241469625160387, "learning_rate": 1.7648511355287037e-05, "loss": 0.3261, "step": 683 }, { "epoch": 2.9189765458422174, "grad_norm": 0.4205344502308214, "learning_rate": 1.758926639489354e-05, "loss": 0.3076, "step": 684 }, { "epoch": 2.923240938166311, "grad_norm": 0.38789875148418834, "learning_rate": 1.7530042892631568e-05, "loss": 0.3185, "step": 685 }, { "epoch": 2.927505330490405, "grad_norm": 0.3516989414998489, "learning_rate": 1.7470841375654214e-05, "loss": 0.3049, "step": 686 }, { "epoch": 2.931769722814499, "grad_norm": 0.3839851773789998, "learning_rate": 1.7411662370918893e-05, "loss": 0.3595, "step": 687 }, { "epoch": 2.9360341151385927, "grad_norm": 0.42202276848131426, "learning_rate": 1.7352506405182635e-05, "loss": 0.3483, "step": 688 }, { "epoch": 2.9402985074626864, "grad_norm": 0.3429765228499871, "learning_rate": 1.7293374004997384e-05, "loss": 0.3195, "step": 689 }, { "epoch": 2.9445628997867805, "grad_norm": 0.35536478053153486, "learning_rate": 1.7234265696705344e-05, "loss": 0.3421, "step": 690 }, { "epoch": 2.9488272921108742, "grad_norm": 0.35560187786546854, "learning_rate": 1.717518200643426e-05, "loss": 0.3342, "step": 691 }, { "epoch": 2.953091684434968, "grad_norm": 0.3995220644927285, "learning_rate": 1.711612346009277e-05, "loss": 0.3359, "step": 692 }, { "epoch": 2.957356076759062, "grad_norm": 0.3695517373221299, "learning_rate": 1.7057090583365678e-05, "loss": 0.3403, "step": 693 }, { "epoch": 2.961620469083156, "grad_norm": 0.300686396282184, "learning_rate": 1.699808390170933e-05, "loss": 0.2988, "step": 694 }, { "epoch": 2.9658848614072495, "grad_norm": 0.4104459486229632, "learning_rate": 1.6939103940346874e-05, "loss": 0.3721, "step": 695 }, { "epoch": 2.970149253731343, "grad_norm": 0.30710749710833807, "learning_rate": 1.6880151224263646e-05, "loss": 0.3053, "step": 696 }, { "epoch": 2.974413646055437, "grad_norm": 0.4085034233515771, "learning_rate": 1.6821226278202458e-05, "loss": 0.3504, "step": 697 }, { "epoch": 2.978678038379531, "grad_norm": 0.3714715567431405, "learning_rate": 1.676232962665894e-05, "loss": 0.333, "step": 698 }, { "epoch": 2.9829424307036247, "grad_norm": 0.3676846794603304, "learning_rate": 1.6703461793876876e-05, "loss": 0.3326, "step": 699 }, { "epoch": 2.9872068230277184, "grad_norm": 0.4106369860379901, "learning_rate": 1.6644623303843518e-05, "loss": 0.3095, "step": 700 }, { "epoch": 2.9914712153518126, "grad_norm": 0.37751964437773083, "learning_rate": 1.6585814680284945e-05, "loss": 0.3751, "step": 701 }, { "epoch": 2.9957356076759063, "grad_norm": 0.3902738884075048, "learning_rate": 1.6527036446661396e-05, "loss": 0.3261, "step": 702 }, { "epoch": 3.0010660980810235, "grad_norm": 0.7206224820790956, "learning_rate": 1.64682891261626e-05, "loss": 0.5125, "step": 703 }, { "epoch": 3.0053304904051172, "grad_norm": 0.5281661795561857, "learning_rate": 1.6409573241703138e-05, "loss": 0.2907, "step": 704 }, { "epoch": 3.009594882729211, "grad_norm": 0.42118092897335085, "learning_rate": 1.635088931591775e-05, "loss": 0.2279, "step": 705 }, { "epoch": 3.013859275053305, "grad_norm": 0.6733763572008821, "learning_rate": 1.6292237871156734e-05, "loss": 0.2392, "step": 706 }, { "epoch": 3.018123667377399, "grad_norm": 0.4467857416127825, "learning_rate": 1.6233619429481268e-05, "loss": 0.2593, "step": 707 }, { "epoch": 3.0223880597014925, "grad_norm": 0.4960925783392073, "learning_rate": 1.6175034512658753e-05, "loss": 0.2384, "step": 708 }, { "epoch": 3.026652452025586, "grad_norm": 0.5148846864817478, "learning_rate": 1.6116483642158213e-05, "loss": 0.2707, "step": 709 }, { "epoch": 3.0309168443496803, "grad_norm": 0.3489966405865502, "learning_rate": 1.6057967339145572e-05, "loss": 0.2101, "step": 710 }, { "epoch": 3.035181236673774, "grad_norm": 0.48600165080284125, "learning_rate": 1.5999486124479115e-05, "loss": 0.2589, "step": 711 }, { "epoch": 3.0394456289978677, "grad_norm": 0.4671615074532655, "learning_rate": 1.594104051870478e-05, "loss": 0.2443, "step": 712 }, { "epoch": 3.0437100213219614, "grad_norm": 0.4565693482005637, "learning_rate": 1.588263104205155e-05, "loss": 0.2356, "step": 713 }, { "epoch": 3.0479744136460556, "grad_norm": 0.4030961726983731, "learning_rate": 1.5824258214426833e-05, "loss": 0.2489, "step": 714 }, { "epoch": 3.0522388059701493, "grad_norm": 0.4183617430166792, "learning_rate": 1.5765922555411793e-05, "loss": 0.2525, "step": 715 }, { "epoch": 3.056503198294243, "grad_norm": 0.4104838834580541, "learning_rate": 1.570762458425677e-05, "loss": 0.2444, "step": 716 }, { "epoch": 3.0607675906183367, "grad_norm": 0.39244414919578313, "learning_rate": 1.5649364819876655e-05, "loss": 0.2483, "step": 717 }, { "epoch": 3.065031982942431, "grad_norm": 0.36480709630972974, "learning_rate": 1.559114378084623e-05, "loss": 0.2216, "step": 718 }, { "epoch": 3.0692963752665245, "grad_norm": 0.3946748564748482, "learning_rate": 1.5532961985395605e-05, "loss": 0.2644, "step": 719 }, { "epoch": 3.0735607675906182, "grad_norm": 0.38069525301339247, "learning_rate": 1.547481995140556e-05, "loss": 0.241, "step": 720 }, { "epoch": 3.077825159914712, "grad_norm": 0.3544801582734647, "learning_rate": 1.5416718196402966e-05, "loss": 0.224, "step": 721 }, { "epoch": 3.082089552238806, "grad_norm": 0.39637110438337914, "learning_rate": 1.535865723755616e-05, "loss": 0.2661, "step": 722 }, { "epoch": 3.0863539445629, "grad_norm": 0.355117830176425, "learning_rate": 1.5300637591670357e-05, "loss": 0.2448, "step": 723 }, { "epoch": 3.0906183368869935, "grad_norm": 0.3712024712331717, "learning_rate": 1.5242659775183039e-05, "loss": 0.231, "step": 724 }, { "epoch": 3.094882729211087, "grad_norm": 0.34556003676946495, "learning_rate": 1.5184724304159368e-05, "loss": 0.2438, "step": 725 }, { "epoch": 3.0991471215351813, "grad_norm": 0.3833769510169898, "learning_rate": 1.5126831694287564e-05, "loss": 0.2547, "step": 726 }, { "epoch": 3.103411513859275, "grad_norm": 0.32399743754417243, "learning_rate": 1.5068982460874357e-05, "loss": 0.2392, "step": 727 }, { "epoch": 3.1076759061833688, "grad_norm": 0.32323430623785687, "learning_rate": 1.5011177118840376e-05, "loss": 0.2504, "step": 728 }, { "epoch": 3.111940298507463, "grad_norm": 0.32697679971993443, "learning_rate": 1.4953416182715566e-05, "loss": 0.2636, "step": 729 }, { "epoch": 3.1162046908315566, "grad_norm": 0.32920439795771206, "learning_rate": 1.4895700166634615e-05, "loss": 0.2428, "step": 730 }, { "epoch": 3.1204690831556503, "grad_norm": 0.3389965095480859, "learning_rate": 1.4838029584332364e-05, "loss": 0.2411, "step": 731 }, { "epoch": 3.124733475479744, "grad_norm": 0.3407193460469488, "learning_rate": 1.478040494913926e-05, "loss": 0.2659, "step": 732 }, { "epoch": 3.128997867803838, "grad_norm": 0.32204674701899744, "learning_rate": 1.4722826773976757e-05, "loss": 0.2289, "step": 733 }, { "epoch": 3.133262260127932, "grad_norm": 0.3246267785751812, "learning_rate": 1.4665295571352776e-05, "loss": 0.2608, "step": 734 }, { "epoch": 3.1375266524520256, "grad_norm": 0.33825862655046196, "learning_rate": 1.460781185335713e-05, "loss": 0.2376, "step": 735 }, { "epoch": 3.1417910447761193, "grad_norm": 0.2967323124289374, "learning_rate": 1.4550376131656952e-05, "loss": 0.2549, "step": 736 }, { "epoch": 3.1460554371002134, "grad_norm": 0.3098012403561765, "learning_rate": 1.449298891749217e-05, "loss": 0.2364, "step": 737 }, { "epoch": 3.150319829424307, "grad_norm": 0.2997550978061584, "learning_rate": 1.443565072167095e-05, "loss": 0.2293, "step": 738 }, { "epoch": 3.154584221748401, "grad_norm": 0.32943504430215304, "learning_rate": 1.4378362054565123e-05, "loss": 0.2881, "step": 739 }, { "epoch": 3.1588486140724945, "grad_norm": 0.3109316470570683, "learning_rate": 1.4321123426105686e-05, "loss": 0.221, "step": 740 }, { "epoch": 3.1631130063965887, "grad_norm": 0.34529106989101405, "learning_rate": 1.4263935345778202e-05, "loss": 0.2709, "step": 741 }, { "epoch": 3.1673773987206824, "grad_norm": 0.2975749768452129, "learning_rate": 1.420679832261833e-05, "loss": 0.2276, "step": 742 }, { "epoch": 3.171641791044776, "grad_norm": 0.29508491762785355, "learning_rate": 1.4149712865207256e-05, "loss": 0.2344, "step": 743 }, { "epoch": 3.1759061833688698, "grad_norm": 0.2973838926689209, "learning_rate": 1.409267948166718e-05, "loss": 0.242, "step": 744 }, { "epoch": 3.180170575692964, "grad_norm": 0.3144417744580649, "learning_rate": 1.4035698679656777e-05, "loss": 0.2497, "step": 745 }, { "epoch": 3.1844349680170576, "grad_norm": 0.30367019765704484, "learning_rate": 1.397877096636671e-05, "loss": 0.2429, "step": 746 }, { "epoch": 3.1886993603411513, "grad_norm": 0.2977699986334773, "learning_rate": 1.3921896848515064e-05, "loss": 0.2495, "step": 747 }, { "epoch": 3.192963752665245, "grad_norm": 0.30274038191648445, "learning_rate": 1.3865076832342897e-05, "loss": 0.23, "step": 748 }, { "epoch": 3.197228144989339, "grad_norm": 0.2916186731441911, "learning_rate": 1.3808311423609686e-05, "loss": 0.233, "step": 749 }, { "epoch": 3.201492537313433, "grad_norm": 0.32401373636645003, "learning_rate": 1.3751601127588849e-05, "loss": 0.2529, "step": 750 }, { "epoch": 3.2057569296375266, "grad_norm": 0.3465730918845462, "learning_rate": 1.3694946449063249e-05, "loss": 0.2621, "step": 751 }, { "epoch": 3.2100213219616203, "grad_norm": 0.32661862365231215, "learning_rate": 1.3638347892320671e-05, "loss": 0.2498, "step": 752 }, { "epoch": 3.2142857142857144, "grad_norm": 0.28989193446587946, "learning_rate": 1.3581805961149371e-05, "loss": 0.2154, "step": 753 }, { "epoch": 3.218550106609808, "grad_norm": 0.32094760076315715, "learning_rate": 1.3525321158833582e-05, "loss": 0.255, "step": 754 }, { "epoch": 3.222814498933902, "grad_norm": 0.3026223441819097, "learning_rate": 1.3468893988149014e-05, "loss": 0.2311, "step": 755 }, { "epoch": 3.227078891257996, "grad_norm": 0.34131523556890064, "learning_rate": 1.341252495135841e-05, "loss": 0.2379, "step": 756 }, { "epoch": 3.2313432835820897, "grad_norm": 0.3128286044976363, "learning_rate": 1.3356214550207038e-05, "loss": 0.2549, "step": 757 }, { "epoch": 3.2356076759061834, "grad_norm": 0.29134924946443214, "learning_rate": 1.3299963285918265e-05, "loss": 0.2423, "step": 758 }, { "epoch": 3.239872068230277, "grad_norm": 0.32994384460086157, "learning_rate": 1.324377165918906e-05, "loss": 0.2668, "step": 759 }, { "epoch": 3.2441364605543708, "grad_norm": 0.30399396622625136, "learning_rate": 1.3187640170185563e-05, "loss": 0.2388, "step": 760 }, { "epoch": 3.248400852878465, "grad_norm": 0.298330256235117, "learning_rate": 1.3131569318538629e-05, "loss": 0.216, "step": 761 }, { "epoch": 3.2526652452025586, "grad_norm": 0.3356191635268766, "learning_rate": 1.3075559603339354e-05, "loss": 0.2515, "step": 762 }, { "epoch": 3.2569296375266523, "grad_norm": 0.3278526977727378, "learning_rate": 1.3019611523134672e-05, "loss": 0.2514, "step": 763 }, { "epoch": 3.2611940298507465, "grad_norm": 0.3076746070268428, "learning_rate": 1.2963725575922896e-05, "loss": 0.2556, "step": 764 }, { "epoch": 3.26545842217484, "grad_norm": 0.302825164819859, "learning_rate": 1.2907902259149287e-05, "loss": 0.2343, "step": 765 }, { "epoch": 3.269722814498934, "grad_norm": 0.3013156949351398, "learning_rate": 1.2852142069701623e-05, "loss": 0.2541, "step": 766 }, { "epoch": 3.2739872068230276, "grad_norm": 0.2976355674558156, "learning_rate": 1.2796445503905797e-05, "loss": 0.2309, "step": 767 }, { "epoch": 3.2782515991471217, "grad_norm": 0.29706305491961443, "learning_rate": 1.274081305752135e-05, "loss": 0.2295, "step": 768 }, { "epoch": 3.2825159914712154, "grad_norm": 0.2881769876779232, "learning_rate": 1.268524522573712e-05, "loss": 0.2568, "step": 769 }, { "epoch": 3.286780383795309, "grad_norm": 0.30859507617667564, "learning_rate": 1.2629742503166797e-05, "loss": 0.277, "step": 770 }, { "epoch": 3.291044776119403, "grad_norm": 0.2700505406761647, "learning_rate": 1.2574305383844528e-05, "loss": 0.22, "step": 771 }, { "epoch": 3.295309168443497, "grad_norm": 0.2700835765523195, "learning_rate": 1.2518934361220523e-05, "loss": 0.2129, "step": 772 }, { "epoch": 3.2995735607675907, "grad_norm": 0.33992965453093965, "learning_rate": 1.2463629928156653e-05, "loss": 0.2461, "step": 773 }, { "epoch": 3.3038379530916844, "grad_norm": 0.295791608172303, "learning_rate": 1.2408392576922075e-05, "loss": 0.2656, "step": 774 }, { "epoch": 3.308102345415778, "grad_norm": 0.2911758696308667, "learning_rate": 1.2353222799188844e-05, "loss": 0.2465, "step": 775 }, { "epoch": 3.3123667377398722, "grad_norm": 0.27346284641679014, "learning_rate": 1.2298121086027543e-05, "loss": 0.2263, "step": 776 }, { "epoch": 3.316631130063966, "grad_norm": 0.32999783460577875, "learning_rate": 1.2243087927902905e-05, "loss": 0.2593, "step": 777 }, { "epoch": 3.3208955223880596, "grad_norm": 0.2830646454254038, "learning_rate": 1.2188123814669437e-05, "loss": 0.2335, "step": 778 }, { "epoch": 3.3251599147121533, "grad_norm": 0.32255253651431215, "learning_rate": 1.2133229235567085e-05, "loss": 0.2603, "step": 779 }, { "epoch": 3.3294243070362475, "grad_norm": 0.32944650851857604, "learning_rate": 1.2078404679216864e-05, "loss": 0.2538, "step": 780 }, { "epoch": 3.333688699360341, "grad_norm": 0.30124840761449095, "learning_rate": 1.2023650633616511e-05, "loss": 0.2263, "step": 781 }, { "epoch": 3.337953091684435, "grad_norm": 0.32447240154800916, "learning_rate": 1.1968967586136144e-05, "loss": 0.2648, "step": 782 }, { "epoch": 3.342217484008529, "grad_norm": 0.2975315219529145, "learning_rate": 1.1914356023513904e-05, "loss": 0.2279, "step": 783 }, { "epoch": 3.3464818763326227, "grad_norm": 0.3031477389481757, "learning_rate": 1.185981643185166e-05, "loss": 0.2596, "step": 784 }, { "epoch": 3.3507462686567164, "grad_norm": 0.30806045278790106, "learning_rate": 1.1805349296610658e-05, "loss": 0.2474, "step": 785 }, { "epoch": 3.35501066098081, "grad_norm": 0.28929666459434256, "learning_rate": 1.1750955102607193e-05, "loss": 0.2426, "step": 786 }, { "epoch": 3.359275053304904, "grad_norm": 0.3007594311300095, "learning_rate": 1.1696634334008317e-05, "loss": 0.2629, "step": 787 }, { "epoch": 3.363539445628998, "grad_norm": 0.3098618482796862, "learning_rate": 1.1642387474327505e-05, "loss": 0.229, "step": 788 }, { "epoch": 3.3678038379530917, "grad_norm": 0.2933436606084351, "learning_rate": 1.1588215006420374e-05, "loss": 0.2526, "step": 789 }, { "epoch": 3.3720682302771854, "grad_norm": 0.2716612613001274, "learning_rate": 1.1534117412480366e-05, "loss": 0.2185, "step": 790 }, { "epoch": 3.3763326226012795, "grad_norm": 0.3085235399779269, "learning_rate": 1.1480095174034464e-05, "loss": 0.2574, "step": 791 }, { "epoch": 3.3805970149253732, "grad_norm": 0.28565152972713814, "learning_rate": 1.1426148771938915e-05, "loss": 0.2276, "step": 792 }, { "epoch": 3.384861407249467, "grad_norm": 0.30010302424063223, "learning_rate": 1.1372278686374935e-05, "loss": 0.2564, "step": 793 }, { "epoch": 3.3891257995735606, "grad_norm": 0.30330171467680656, "learning_rate": 1.1318485396844427e-05, "loss": 0.2231, "step": 794 }, { "epoch": 3.3933901918976543, "grad_norm": 0.31861742745133986, "learning_rate": 1.1264769382165748e-05, "loss": 0.2343, "step": 795 }, { "epoch": 3.3976545842217485, "grad_norm": 0.30438593980087136, "learning_rate": 1.1211131120469413e-05, "loss": 0.2446, "step": 796 }, { "epoch": 3.401918976545842, "grad_norm": 0.28813768071806006, "learning_rate": 1.1157571089193856e-05, "loss": 0.2242, "step": 797 }, { "epoch": 3.406183368869936, "grad_norm": 0.30606850192411184, "learning_rate": 1.110408976508118e-05, "loss": 0.2552, "step": 798 }, { "epoch": 3.41044776119403, "grad_norm": 0.3145029450607833, "learning_rate": 1.105068762417289e-05, "loss": 0.2496, "step": 799 }, { "epoch": 3.4147121535181237, "grad_norm": 0.2924588998467964, "learning_rate": 1.0997365141805697e-05, "loss": 0.2464, "step": 800 }, { "epoch": 3.4189765458422174, "grad_norm": 0.30443844783069873, "learning_rate": 1.094412279260726e-05, "loss": 0.2395, "step": 801 }, { "epoch": 3.423240938166311, "grad_norm": 0.3123338581477662, "learning_rate": 1.0890961050491952e-05, "loss": 0.2351, "step": 802 }, { "epoch": 3.4275053304904053, "grad_norm": 0.2964017517166986, "learning_rate": 1.0837880388656687e-05, "loss": 0.2416, "step": 803 }, { "epoch": 3.431769722814499, "grad_norm": 0.3281476512317521, "learning_rate": 1.0784881279576635e-05, "loss": 0.2383, "step": 804 }, { "epoch": 3.4360341151385927, "grad_norm": 0.31829793297554937, "learning_rate": 1.073196419500109e-05, "loss": 0.2478, "step": 805 }, { "epoch": 3.4402985074626864, "grad_norm": 0.32602829624699914, "learning_rate": 1.067912960594923e-05, "loss": 0.2241, "step": 806 }, { "epoch": 3.4445628997867805, "grad_norm": 0.3191426788748806, "learning_rate": 1.0626377982705929e-05, "loss": 0.2418, "step": 807 }, { "epoch": 3.4488272921108742, "grad_norm": 0.3021368983804344, "learning_rate": 1.0573709794817586e-05, "loss": 0.2459, "step": 808 }, { "epoch": 3.453091684434968, "grad_norm": 0.3307143438925763, "learning_rate": 1.0521125511087927e-05, "loss": 0.2456, "step": 809 }, { "epoch": 3.4573560767590616, "grad_norm": 0.3175944122781918, "learning_rate": 1.0468625599573842e-05, "loss": 0.2478, "step": 810 }, { "epoch": 3.461620469083156, "grad_norm": 0.28439054844349576, "learning_rate": 1.0416210527581221e-05, "loss": 0.2457, "step": 811 }, { "epoch": 3.4658848614072495, "grad_norm": 0.33296617403496936, "learning_rate": 1.0363880761660786e-05, "loss": 0.2438, "step": 812 }, { "epoch": 3.470149253731343, "grad_norm": 0.29277732354416547, "learning_rate": 1.0311636767603952e-05, "loss": 0.2244, "step": 813 }, { "epoch": 3.474413646055437, "grad_norm": 0.28674974708683865, "learning_rate": 1.025947901043865e-05, "loss": 0.2279, "step": 814 }, { "epoch": 3.478678038379531, "grad_norm": 0.3469212992340864, "learning_rate": 1.020740795442523e-05, "loss": 0.2564, "step": 815 }, { "epoch": 3.4829424307036247, "grad_norm": 0.30741933996359333, "learning_rate": 1.0155424063052306e-05, "loss": 0.246, "step": 816 }, { "epoch": 3.4872068230277184, "grad_norm": 0.2985597325581848, "learning_rate": 1.0103527799032627e-05, "loss": 0.2449, "step": 817 }, { "epoch": 3.4914712153518126, "grad_norm": 0.32011526265805135, "learning_rate": 1.0051719624298962e-05, "loss": 0.233, "step": 818 }, { "epoch": 3.4957356076759063, "grad_norm": 0.30558408877185367, "learning_rate": 1.0000000000000006e-05, "loss": 0.226, "step": 819 }, { "epoch": 3.5, "grad_norm": 0.31726146235825886, "learning_rate": 9.948369386496231e-06, "loss": 0.243, "step": 820 }, { "epoch": 3.5042643923240937, "grad_norm": 0.311677236974838, "learning_rate": 9.89682824335584e-06, "loss": 0.2547, "step": 821 }, { "epoch": 3.5085287846481874, "grad_norm": 0.29139088680330627, "learning_rate": 9.84537702935065e-06, "loss": 0.2261, "step": 822 }, { "epoch": 3.5127931769722816, "grad_norm": 0.33737575572615713, "learning_rate": 9.794016202452003e-06, "loss": 0.2431, "step": 823 }, { "epoch": 3.5170575692963753, "grad_norm": 0.3141703384709648, "learning_rate": 9.742746219826715e-06, "loss": 0.251, "step": 824 }, { "epoch": 3.521321961620469, "grad_norm": 0.2934897661263017, "learning_rate": 9.691567537832964e-06, "loss": 0.2494, "step": 825 }, { "epoch": 3.525586353944563, "grad_norm": 0.32253152245304295, "learning_rate": 9.640480612016272e-06, "loss": 0.2679, "step": 826 }, { "epoch": 3.529850746268657, "grad_norm": 0.26852755792702954, "learning_rate": 9.589485897105431e-06, "loss": 0.2129, "step": 827 }, { "epoch": 3.5341151385927505, "grad_norm": 0.29815354851790776, "learning_rate": 9.538583847008452e-06, "loss": 0.238, "step": 828 }, { "epoch": 3.538379530916844, "grad_norm": 0.30229937242138655, "learning_rate": 9.487774914808536e-06, "loss": 0.2302, "step": 829 }, { "epoch": 3.542643923240938, "grad_norm": 0.3277955189869379, "learning_rate": 9.437059552760018e-06, "loss": 0.242, "step": 830 }, { "epoch": 3.546908315565032, "grad_norm": 0.3166866159369122, "learning_rate": 9.386438212284372e-06, "loss": 0.2438, "step": 831 }, { "epoch": 3.5511727078891258, "grad_norm": 0.2749474828352015, "learning_rate": 9.33591134396618e-06, "loss": 0.2152, "step": 832 }, { "epoch": 3.5554371002132195, "grad_norm": 0.30041265733845113, "learning_rate": 9.28547939754911e-06, "loss": 0.2362, "step": 833 }, { "epoch": 3.5597014925373136, "grad_norm": 0.3146999658258432, "learning_rate": 9.235142821931928e-06, "loss": 0.244, "step": 834 }, { "epoch": 3.5639658848614073, "grad_norm": 0.3293166515443863, "learning_rate": 9.1849020651645e-06, "loss": 0.2585, "step": 835 }, { "epoch": 3.568230277185501, "grad_norm": 0.2760295232594156, "learning_rate": 9.134757574443793e-06, "loss": 0.2258, "step": 836 }, { "epoch": 3.572494669509595, "grad_norm": 0.30975186669845683, "learning_rate": 9.084709796109907e-06, "loss": 0.2515, "step": 837 }, { "epoch": 3.576759061833689, "grad_norm": 0.33162696304477796, "learning_rate": 9.034759175642099e-06, "loss": 0.2595, "step": 838 }, { "epoch": 3.5810234541577826, "grad_norm": 0.2883332393947174, "learning_rate": 8.984906157654818e-06, "loss": 0.205, "step": 839 }, { "epoch": 3.5852878464818763, "grad_norm": 0.2978502889456044, "learning_rate": 8.93515118589373e-06, "loss": 0.2458, "step": 840 }, { "epoch": 3.58955223880597, "grad_norm": 0.29408879165914026, "learning_rate": 8.885494703231798e-06, "loss": 0.2392, "step": 841 }, { "epoch": 3.593816631130064, "grad_norm": 0.3249087183509099, "learning_rate": 8.835937151665328e-06, "loss": 0.257, "step": 842 }, { "epoch": 3.598081023454158, "grad_norm": 0.32232731451375923, "learning_rate": 8.786478972310023e-06, "loss": 0.2469, "step": 843 }, { "epoch": 3.6023454157782515, "grad_norm": 0.29034742011961323, "learning_rate": 8.737120605397071e-06, "loss": 0.2484, "step": 844 }, { "epoch": 3.6066098081023457, "grad_norm": 0.28525812573618703, "learning_rate": 8.687862490269232e-06, "loss": 0.248, "step": 845 }, { "epoch": 3.6108742004264394, "grad_norm": 0.2872684471511718, "learning_rate": 8.638705065376887e-06, "loss": 0.2351, "step": 846 }, { "epoch": 3.615138592750533, "grad_norm": 0.30255280187752015, "learning_rate": 8.589648768274192e-06, "loss": 0.2377, "step": 847 }, { "epoch": 3.6194029850746268, "grad_norm": 0.28771967192257747, "learning_rate": 8.54069403561515e-06, "loss": 0.2422, "step": 848 }, { "epoch": 3.6236673773987205, "grad_norm": 0.31269556887708466, "learning_rate": 8.491841303149728e-06, "loss": 0.2727, "step": 849 }, { "epoch": 3.6279317697228146, "grad_norm": 0.3013824576718503, "learning_rate": 8.443091005719987e-06, "loss": 0.2425, "step": 850 }, { "epoch": 3.6321961620469083, "grad_norm": 0.2648872302755811, "learning_rate": 8.394443577256192e-06, "loss": 0.2167, "step": 851 }, { "epoch": 3.636460554371002, "grad_norm": 0.2896543847103546, "learning_rate": 8.345899450772975e-06, "loss": 0.2321, "step": 852 }, { "epoch": 3.640724946695096, "grad_norm": 0.2966430163917612, "learning_rate": 8.297459058365472e-06, "loss": 0.2536, "step": 853 }, { "epoch": 3.64498933901919, "grad_norm": 0.29102715377273775, "learning_rate": 8.249122831205458e-06, "loss": 0.2352, "step": 854 }, { "epoch": 3.6492537313432836, "grad_norm": 0.276871037645944, "learning_rate": 8.200891199537549e-06, "loss": 0.2314, "step": 855 }, { "epoch": 3.6535181236673773, "grad_norm": 0.29679802277295453, "learning_rate": 8.152764592675317e-06, "loss": 0.2466, "step": 856 }, { "epoch": 3.657782515991471, "grad_norm": 0.3288786101624257, "learning_rate": 8.104743438997527e-06, "loss": 0.2481, "step": 857 }, { "epoch": 3.662046908315565, "grad_norm": 0.2810797169473959, "learning_rate": 8.056828165944282e-06, "loss": 0.2462, "step": 858 }, { "epoch": 3.666311300639659, "grad_norm": 0.27585540738489867, "learning_rate": 8.009019200013237e-06, "loss": 0.2356, "step": 859 }, { "epoch": 3.6705756929637525, "grad_norm": 0.32641357147184774, "learning_rate": 7.961316966755806e-06, "loss": 0.2456, "step": 860 }, { "epoch": 3.6748400852878467, "grad_norm": 0.31910225006889814, "learning_rate": 7.913721890773354e-06, "loss": 0.2721, "step": 861 }, { "epoch": 3.6791044776119404, "grad_norm": 0.3397767935316877, "learning_rate": 7.866234395713441e-06, "loss": 0.2709, "step": 862 }, { "epoch": 3.683368869936034, "grad_norm": 0.296901908055256, "learning_rate": 7.818854904266033e-06, "loss": 0.2503, "step": 863 }, { "epoch": 3.6876332622601278, "grad_norm": 0.2808208297289632, "learning_rate": 7.771583838159756e-06, "loss": 0.2399, "step": 864 }, { "epoch": 3.6918976545842215, "grad_norm": 0.29198514248012775, "learning_rate": 7.724421618158122e-06, "loss": 0.2666, "step": 865 }, { "epoch": 3.6961620469083156, "grad_norm": 0.29995463338764394, "learning_rate": 7.677368664055815e-06, "loss": 0.2396, "step": 866 }, { "epoch": 3.7004264392324093, "grad_norm": 0.28578829975709263, "learning_rate": 7.630425394674903e-06, "loss": 0.257, "step": 867 }, { "epoch": 3.704690831556503, "grad_norm": 0.28021430102270617, "learning_rate": 7.583592227861167e-06, "loss": 0.2236, "step": 868 }, { "epoch": 3.708955223880597, "grad_norm": 0.3268601383355693, "learning_rate": 7.536869580480351e-06, "loss": 0.2458, "step": 869 }, { "epoch": 3.713219616204691, "grad_norm": 0.290893866821192, "learning_rate": 7.49025786841445e-06, "loss": 0.2358, "step": 870 }, { "epoch": 3.7174840085287846, "grad_norm": 0.2944701896730605, "learning_rate": 7.443757506558033e-06, "loss": 0.2386, "step": 871 }, { "epoch": 3.7217484008528787, "grad_norm": 0.2986050086210331, "learning_rate": 7.397368908814499e-06, "loss": 0.2637, "step": 872 }, { "epoch": 3.7260127931769724, "grad_norm": 0.289696846898496, "learning_rate": 7.3510924880924575e-06, "loss": 0.2266, "step": 873 }, { "epoch": 3.730277185501066, "grad_norm": 0.3240013899409774, "learning_rate": 7.3049286563020036e-06, "loss": 0.2466, "step": 874 }, { "epoch": 3.73454157782516, "grad_norm": 0.3055316273372297, "learning_rate": 7.258877824351081e-06, "loss": 0.2439, "step": 875 }, { "epoch": 3.7388059701492535, "grad_norm": 0.2904274401241053, "learning_rate": 7.212940402141808e-06, "loss": 0.2157, "step": 876 }, { "epoch": 3.7430703624733477, "grad_norm": 0.3054868031454664, "learning_rate": 7.16711679856682e-06, "loss": 0.2483, "step": 877 }, { "epoch": 3.7473347547974414, "grad_norm": 0.2670625511172484, "learning_rate": 7.121407421505666e-06, "loss": 0.2206, "step": 878 }, { "epoch": 3.751599147121535, "grad_norm": 0.2721294014835585, "learning_rate": 7.075812677821145e-06, "loss": 0.2514, "step": 879 }, { "epoch": 3.7558635394456292, "grad_norm": 0.2808108535749494, "learning_rate": 7.030332973355696e-06, "loss": 0.2198, "step": 880 }, { "epoch": 3.760127931769723, "grad_norm": 0.29486840541948817, "learning_rate": 6.984968712927789e-06, "loss": 0.2333, "step": 881 }, { "epoch": 3.7643923240938166, "grad_norm": 0.2874118074599997, "learning_rate": 6.939720300328303e-06, "loss": 0.2518, "step": 882 }, { "epoch": 3.7686567164179103, "grad_norm": 0.2683580455012857, "learning_rate": 6.894588138316962e-06, "loss": 0.2286, "step": 883 }, { "epoch": 3.772921108742004, "grad_norm": 0.27089873290672944, "learning_rate": 6.84957262861873e-06, "loss": 0.2352, "step": 884 }, { "epoch": 3.777185501066098, "grad_norm": 0.28794297921319983, "learning_rate": 6.8046741719202385e-06, "loss": 0.2435, "step": 885 }, { "epoch": 3.781449893390192, "grad_norm": 0.29512771103869784, "learning_rate": 6.7598931678662205e-06, "loss": 0.2486, "step": 886 }, { "epoch": 3.7857142857142856, "grad_norm": 0.27422515206444054, "learning_rate": 6.7152300150559514e-06, "loss": 0.2302, "step": 887 }, { "epoch": 3.7899786780383797, "grad_norm": 0.28450358656356606, "learning_rate": 6.67068511103971e-06, "loss": 0.2641, "step": 888 }, { "epoch": 3.7942430703624734, "grad_norm": 0.27481427773374834, "learning_rate": 6.626258852315226e-06, "loss": 0.2443, "step": 889 }, { "epoch": 3.798507462686567, "grad_norm": 0.2924351460268186, "learning_rate": 6.581951634324164e-06, "loss": 0.2571, "step": 890 }, { "epoch": 3.802771855010661, "grad_norm": 0.2790682450802608, "learning_rate": 6.537763851448593e-06, "loss": 0.23, "step": 891 }, { "epoch": 3.8070362473347545, "grad_norm": 0.2930329051203839, "learning_rate": 6.493695897007488e-06, "loss": 0.2586, "step": 892 }, { "epoch": 3.8113006396588487, "grad_norm": 0.28549744285559625, "learning_rate": 6.4497481632532025e-06, "loss": 0.2447, "step": 893 }, { "epoch": 3.8155650319829424, "grad_norm": 0.29773280536775487, "learning_rate": 6.4059210413680175e-06, "loss": 0.2651, "step": 894 }, { "epoch": 3.819829424307036, "grad_norm": 0.25471857385762836, "learning_rate": 6.362214921460628e-06, "loss": 0.2212, "step": 895 }, { "epoch": 3.8240938166311302, "grad_norm": 0.2862555008696523, "learning_rate": 6.318630192562685e-06, "loss": 0.2438, "step": 896 }, { "epoch": 3.828358208955224, "grad_norm": 0.2902843603482996, "learning_rate": 6.275167242625331e-06, "loss": 0.261, "step": 897 }, { "epoch": 3.8326226012793176, "grad_norm": 0.2734916460539061, "learning_rate": 6.231826458515728e-06, "loss": 0.235, "step": 898 }, { "epoch": 3.836886993603412, "grad_norm": 0.2959908993657831, "learning_rate": 6.1886082260136486e-06, "loss": 0.2674, "step": 899 }, { "epoch": 3.8411513859275055, "grad_norm": 0.2629506747169864, "learning_rate": 6.145512929808013e-06, "loss": 0.2269, "step": 900 }, { "epoch": 3.845415778251599, "grad_norm": 0.2796514994593026, "learning_rate": 6.102540953493483e-06, "loss": 0.2539, "step": 901 }, { "epoch": 3.849680170575693, "grad_norm": 0.28247757622970854, "learning_rate": 6.0596926795670445e-06, "loss": 0.2505, "step": 902 }, { "epoch": 3.8539445628997866, "grad_norm": 0.28546889184805163, "learning_rate": 6.016968489424572e-06, "loss": 0.239, "step": 903 }, { "epoch": 3.8582089552238807, "grad_norm": 0.26614246481199944, "learning_rate": 5.974368763357494e-06, "loss": 0.2541, "step": 904 }, { "epoch": 3.8624733475479744, "grad_norm": 0.270890457656894, "learning_rate": 5.931893880549356e-06, "loss": 0.2469, "step": 905 }, { "epoch": 3.866737739872068, "grad_norm": 0.28586020360247877, "learning_rate": 5.889544219072465e-06, "loss": 0.2477, "step": 906 }, { "epoch": 3.8710021321961623, "grad_norm": 0.2680248514296973, "learning_rate": 5.847320155884535e-06, "loss": 0.2302, "step": 907 }, { "epoch": 3.875266524520256, "grad_norm": 0.25517931577869063, "learning_rate": 5.805222066825296e-06, "loss": 0.2215, "step": 908 }, { "epoch": 3.8795309168443497, "grad_norm": 0.27280219353131735, "learning_rate": 5.7632503266131925e-06, "loss": 0.2317, "step": 909 }, { "epoch": 3.8837953091684434, "grad_norm": 0.2716445422818712, "learning_rate": 5.721405308842023e-06, "loss": 0.2579, "step": 910 }, { "epoch": 3.888059701492537, "grad_norm": 0.27313578866815935, "learning_rate": 5.679687385977617e-06, "loss": 0.2282, "step": 911 }, { "epoch": 3.8923240938166312, "grad_norm": 0.2797256005452073, "learning_rate": 5.638096929354522e-06, "loss": 0.2157, "step": 912 }, { "epoch": 3.896588486140725, "grad_norm": 0.2892205252889863, "learning_rate": 5.596634309172704e-06, "loss": 0.2571, "step": 913 }, { "epoch": 3.9008528784648187, "grad_norm": 0.2701601040294437, "learning_rate": 5.555299894494237e-06, "loss": 0.2395, "step": 914 }, { "epoch": 3.905117270788913, "grad_norm": 0.2601323072881841, "learning_rate": 5.514094053240035e-06, "loss": 0.2149, "step": 915 }, { "epoch": 3.9093816631130065, "grad_norm": 0.32348241733730365, "learning_rate": 5.47301715218657e-06, "loss": 0.2475, "step": 916 }, { "epoch": 3.9136460554371, "grad_norm": 0.2805297373975744, "learning_rate": 5.432069556962602e-06, "loss": 0.2275, "step": 917 }, { "epoch": 3.917910447761194, "grad_norm": 0.2886640845993047, "learning_rate": 5.39125163204594e-06, "loss": 0.2384, "step": 918 }, { "epoch": 3.9221748400852876, "grad_norm": 0.28537996894175865, "learning_rate": 5.350563740760164e-06, "loss": 0.2397, "step": 919 }, { "epoch": 3.9264392324093818, "grad_norm": 0.3035362606109377, "learning_rate": 5.31000624527144e-06, "loss": 0.2383, "step": 920 }, { "epoch": 3.9307036247334755, "grad_norm": 0.2998584944750757, "learning_rate": 5.269579506585259e-06, "loss": 0.2456, "step": 921 }, { "epoch": 3.934968017057569, "grad_norm": 0.2843194883548224, "learning_rate": 5.229283884543235e-06, "loss": 0.2397, "step": 922 }, { "epoch": 3.9392324093816633, "grad_norm": 0.27890127535725073, "learning_rate": 5.189119737819912e-06, "loss": 0.2538, "step": 923 }, { "epoch": 3.943496801705757, "grad_norm": 0.2650375563185337, "learning_rate": 5.149087423919541e-06, "loss": 0.2119, "step": 924 }, { "epoch": 3.9477611940298507, "grad_norm": 0.284810452643455, "learning_rate": 5.109187299172938e-06, "loss": 0.2668, "step": 925 }, { "epoch": 3.9520255863539444, "grad_norm": 0.27726445813021233, "learning_rate": 5.069419718734283e-06, "loss": 0.241, "step": 926 }, { "epoch": 3.956289978678038, "grad_norm": 0.28755904463832055, "learning_rate": 5.029785036577976e-06, "loss": 0.2693, "step": 927 }, { "epoch": 3.9605543710021323, "grad_norm": 0.25514683912114794, "learning_rate": 4.99028360549548e-06, "loss": 0.1971, "step": 928 }, { "epoch": 3.964818763326226, "grad_norm": 0.2956950707929946, "learning_rate": 4.9509157770921625e-06, "loss": 0.2491, "step": 929 }, { "epoch": 3.9690831556503197, "grad_norm": 0.2767014770132041, "learning_rate": 4.911681901784198e-06, "loss": 0.2441, "step": 930 }, { "epoch": 3.973347547974414, "grad_norm": 0.28218722265046, "learning_rate": 4.87258232879543e-06, "loss": 0.2531, "step": 931 }, { "epoch": 3.9776119402985075, "grad_norm": 0.2724674673441032, "learning_rate": 4.83361740615427e-06, "loss": 0.2273, "step": 932 }, { "epoch": 3.981876332622601, "grad_norm": 0.2738688947167063, "learning_rate": 4.794787480690597e-06, "loss": 0.2465, "step": 933 }, { "epoch": 3.9861407249466954, "grad_norm": 0.27288919048483945, "learning_rate": 4.7560928980326535e-06, "loss": 0.2411, "step": 934 }, { "epoch": 3.990405117270789, "grad_norm": 0.26906162947359014, "learning_rate": 4.717534002604005e-06, "loss": 0.2293, "step": 935 }, { "epoch": 3.9946695095948828, "grad_norm": 0.2618466173879373, "learning_rate": 4.679111137620442e-06, "loss": 0.2064, "step": 936 }, { "epoch": 3.9989339019189765, "grad_norm": 0.6242272523532433, "learning_rate": 4.640824645086943e-06, "loss": 0.4147, "step": 937 }, { "epoch": 4.004264392324094, "grad_norm": 0.502124542538278, "learning_rate": 4.6026748657946226e-06, "loss": 0.182, "step": 938 }, { "epoch": 4.008528784648187, "grad_norm": 0.40214808756394727, "learning_rate": 4.5646621393177e-06, "loss": 0.1878, "step": 939 }, { "epoch": 4.0127931769722816, "grad_norm": 0.2964884308202411, "learning_rate": 4.5267868040104765e-06, "loss": 0.1827, "step": 940 }, { "epoch": 4.017057569296376, "grad_norm": 0.379223019030444, "learning_rate": 4.489049197004323e-06, "loss": 0.1759, "step": 941 }, { "epoch": 4.021321961620469, "grad_norm": 0.5008060151918551, "learning_rate": 4.451449654204685e-06, "loss": 0.1906, "step": 942 }, { "epoch": 4.025586353944563, "grad_norm": 0.45022251052851464, "learning_rate": 4.413988510288072e-06, "loss": 0.1764, "step": 943 }, { "epoch": 4.029850746268656, "grad_norm": 0.3183418187342008, "learning_rate": 4.376666098699112e-06, "loss": 0.177, "step": 944 }, { "epoch": 4.0341151385927505, "grad_norm": 0.3522770365054266, "learning_rate": 4.339482751647557e-06, "loss": 0.1705, "step": 945 }, { "epoch": 4.038379530916845, "grad_norm": 0.4263503005592876, "learning_rate": 4.302438800105335e-06, "loss": 0.188, "step": 946 }, { "epoch": 4.042643923240938, "grad_norm": 0.3946522385281874, "learning_rate": 4.265534573803607e-06, "loss": 0.1986, "step": 947 }, { "epoch": 4.046908315565032, "grad_norm": 0.31039206261225744, "learning_rate": 4.228770401229824e-06, "loss": 0.1807, "step": 948 }, { "epoch": 4.051172707889126, "grad_norm": 0.2698934513861876, "learning_rate": 4.1921466096248164e-06, "loss": 0.1644, "step": 949 }, { "epoch": 4.0554371002132195, "grad_norm": 0.3191366565183697, "learning_rate": 4.155663524979854e-06, "loss": 0.1818, "step": 950 }, { "epoch": 4.059701492537314, "grad_norm": 0.3494685301529987, "learning_rate": 4.119321472033779e-06, "loss": 0.1853, "step": 951 }, { "epoch": 4.063965884861407, "grad_norm": 0.3691215369886063, "learning_rate": 4.083120774270095e-06, "loss": 0.1892, "step": 952 }, { "epoch": 4.068230277185501, "grad_norm": 0.30967940037401875, "learning_rate": 4.04706175391409e-06, "loss": 0.1862, "step": 953 }, { "epoch": 4.072494669509595, "grad_norm": 0.28470587344113735, "learning_rate": 4.011144731929981e-06, "loss": 0.1635, "step": 954 }, { "epoch": 4.076759061833688, "grad_norm": 0.28866644852025736, "learning_rate": 3.975370028018022e-06, "loss": 0.195, "step": 955 }, { "epoch": 4.081023454157783, "grad_norm": 0.3005010899793622, "learning_rate": 3.9397379606117045e-06, "loss": 0.2031, "step": 956 }, { "epoch": 4.085287846481877, "grad_norm": 0.3223639228910986, "learning_rate": 3.904248846874894e-06, "loss": 0.2033, "step": 957 }, { "epoch": 4.08955223880597, "grad_norm": 0.2830213529504144, "learning_rate": 3.868903002699014e-06, "loss": 0.1836, "step": 958 }, { "epoch": 4.093816631130064, "grad_norm": 0.27857117999872993, "learning_rate": 3.833700742700237e-06, "loss": 0.1682, "step": 959 }, { "epoch": 4.098081023454157, "grad_norm": 0.28521159706834004, "learning_rate": 3.7986423802166705e-06, "loss": 0.1829, "step": 960 }, { "epoch": 4.1023454157782515, "grad_norm": 0.31579283722941986, "learning_rate": 3.7637282273055918e-06, "loss": 0.1881, "step": 961 }, { "epoch": 4.106609808102346, "grad_norm": 0.28461700269519835, "learning_rate": 3.7289585947406504e-06, "loss": 0.1681, "step": 962 }, { "epoch": 4.110874200426439, "grad_norm": 0.2946870658948828, "learning_rate": 3.694333792009115e-06, "loss": 0.1707, "step": 963 }, { "epoch": 4.115138592750533, "grad_norm": 0.26948058780522965, "learning_rate": 3.6598541273091035e-06, "loss": 0.2017, "step": 964 }, { "epoch": 4.119402985074627, "grad_norm": 0.26347028329932537, "learning_rate": 3.6255199075468595e-06, "loss": 0.1742, "step": 965 }, { "epoch": 4.1236673773987205, "grad_norm": 0.2655975197937126, "learning_rate": 3.5913314383339937e-06, "loss": 0.1903, "step": 966 }, { "epoch": 4.127931769722815, "grad_norm": 0.28973082824861224, "learning_rate": 3.5572890239847934e-06, "loss": 0.19, "step": 967 }, { "epoch": 4.132196162046908, "grad_norm": 0.28373915001437056, "learning_rate": 3.5233929675134993e-06, "loss": 0.1668, "step": 968 }, { "epoch": 4.136460554371002, "grad_norm": 0.27968718279785926, "learning_rate": 3.4896435706316e-06, "loss": 0.1792, "step": 969 }, { "epoch": 4.140724946695096, "grad_norm": 0.2684912141931692, "learning_rate": 3.456041133745163e-06, "loss": 0.1859, "step": 970 }, { "epoch": 4.144989339019189, "grad_norm": 0.25671222298384944, "learning_rate": 3.4225859559521536e-06, "loss": 0.1943, "step": 971 }, { "epoch": 4.149253731343284, "grad_norm": 0.2941742818421993, "learning_rate": 3.3892783350397675e-06, "loss": 0.1804, "step": 972 }, { "epoch": 4.153518123667378, "grad_norm": 0.26842044384478514, "learning_rate": 3.356118567481785e-06, "loss": 0.1875, "step": 973 }, { "epoch": 4.157782515991471, "grad_norm": 0.27257574650490113, "learning_rate": 3.3231069484359367e-06, "loss": 0.1614, "step": 974 }, { "epoch": 4.162046908315565, "grad_norm": 0.26576435827231454, "learning_rate": 3.290243771741275e-06, "loss": 0.2, "step": 975 }, { "epoch": 4.166311300639659, "grad_norm": 0.28364887161883395, "learning_rate": 3.2575293299155343e-06, "loss": 0.177, "step": 976 }, { "epoch": 4.1705756929637525, "grad_norm": 0.2648805323845748, "learning_rate": 3.2249639141525724e-06, "loss": 0.1761, "step": 977 }, { "epoch": 4.174840085287847, "grad_norm": 0.2772188319946492, "learning_rate": 3.1925478143197418e-06, "loss": 0.1652, "step": 978 }, { "epoch": 4.17910447761194, "grad_norm": 0.24545620530472215, "learning_rate": 3.160281318955325e-06, "loss": 0.182, "step": 979 }, { "epoch": 4.183368869936034, "grad_norm": 0.25783169525362326, "learning_rate": 3.1281647152659687e-06, "loss": 0.1783, "step": 980 }, { "epoch": 4.187633262260128, "grad_norm": 0.26773304598461506, "learning_rate": 3.0961982891241083e-06, "loss": 0.1845, "step": 981 }, { "epoch": 4.1918976545842215, "grad_norm": 0.2634763485660847, "learning_rate": 3.0643823250654447e-06, "loss": 0.1883, "step": 982 }, { "epoch": 4.196162046908316, "grad_norm": 0.27160922729899195, "learning_rate": 3.032717106286409e-06, "loss": 0.1804, "step": 983 }, { "epoch": 4.20042643923241, "grad_norm": 0.28052463542160555, "learning_rate": 3.001202914641628e-06, "loss": 0.1876, "step": 984 }, { "epoch": 4.204690831556503, "grad_norm": 0.27189232529742935, "learning_rate": 2.9698400306414353e-06, "loss": 0.1847, "step": 985 }, { "epoch": 4.208955223880597, "grad_norm": 0.261299366875251, "learning_rate": 2.9386287334493426e-06, "loss": 0.1674, "step": 986 }, { "epoch": 4.21321961620469, "grad_norm": 0.2685084307301793, "learning_rate": 2.907569300879596e-06, "loss": 0.1815, "step": 987 }, { "epoch": 4.217484008528785, "grad_norm": 0.2534128060130963, "learning_rate": 2.876662009394673e-06, "loss": 0.1924, "step": 988 }, { "epoch": 4.221748400852879, "grad_norm": 0.25673713632277023, "learning_rate": 2.8459071341028323e-06, "loss": 0.1912, "step": 989 }, { "epoch": 4.226012793176972, "grad_norm": 0.2468883201944162, "learning_rate": 2.815304948755664e-06, "loss": 0.1779, "step": 990 }, { "epoch": 4.230277185501066, "grad_norm": 0.25408696538411385, "learning_rate": 2.784855725745661e-06, "loss": 0.1872, "step": 991 }, { "epoch": 4.23454157782516, "grad_norm": 0.2781469282460988, "learning_rate": 2.7545597361037657e-06, "loss": 0.1934, "step": 992 }, { "epoch": 4.2388059701492535, "grad_norm": 0.2615576306598456, "learning_rate": 2.7244172494969978e-06, "loss": 0.1935, "step": 993 }, { "epoch": 4.243070362473348, "grad_norm": 0.2639612644746222, "learning_rate": 2.6944285342260277e-06, "loss": 0.1746, "step": 994 }, { "epoch": 4.247334754797441, "grad_norm": 0.2490668697614864, "learning_rate": 2.6645938572227946e-06, "loss": 0.1745, "step": 995 }, { "epoch": 4.251599147121535, "grad_norm": 0.25084455227336827, "learning_rate": 2.6349134840481294e-06, "loss": 0.1826, "step": 996 }, { "epoch": 4.255863539445629, "grad_norm": 0.25951044293340914, "learning_rate": 2.6053876788893952e-06, "loss": 0.1765, "step": 997 }, { "epoch": 4.2601279317697225, "grad_norm": 0.2616833813932719, "learning_rate": 2.5760167045581308e-06, "loss": 0.1779, "step": 998 }, { "epoch": 4.264392324093817, "grad_norm": 0.2683034823812298, "learning_rate": 2.546800822487714e-06, "loss": 0.178, "step": 999 }, { "epoch": 4.268656716417911, "grad_norm": 0.25617069908219275, "learning_rate": 2.5177402927310344e-06, "loss": 0.1685, "step": 1000 }, { "epoch": 4.272921108742004, "grad_norm": 0.2625568533886325, "learning_rate": 2.488835373958185e-06, "loss": 0.183, "step": 1001 }, { "epoch": 4.277185501066098, "grad_norm": 0.2526196797026667, "learning_rate": 2.4600863234541338e-06, "loss": 0.1788, "step": 1002 }, { "epoch": 4.281449893390192, "grad_norm": 0.2717109859468861, "learning_rate": 2.4314933971164734e-06, "loss": 0.1838, "step": 1003 }, { "epoch": 4.285714285714286, "grad_norm": 0.2596571713184107, "learning_rate": 2.4030568494531135e-06, "loss": 0.1744, "step": 1004 }, { "epoch": 4.28997867803838, "grad_norm": 0.25389529145774403, "learning_rate": 2.374776933580025e-06, "loss": 0.1912, "step": 1005 }, { "epoch": 4.294243070362473, "grad_norm": 0.24747405968954783, "learning_rate": 2.3466539012189913e-06, "loss": 0.1684, "step": 1006 }, { "epoch": 4.298507462686567, "grad_norm": 0.26068317199815344, "learning_rate": 2.318688002695355e-06, "loss": 0.184, "step": 1007 }, { "epoch": 4.302771855010661, "grad_norm": 0.2476855542290121, "learning_rate": 2.2908794869358044e-06, "loss": 0.1831, "step": 1008 }, { "epoch": 4.3070362473347545, "grad_norm": 0.2541200698145374, "learning_rate": 2.2632286014661477e-06, "loss": 0.1904, "step": 1009 }, { "epoch": 4.311300639658849, "grad_norm": 0.2504818694519344, "learning_rate": 2.2357355924091207e-06, "loss": 0.1588, "step": 1010 }, { "epoch": 4.315565031982943, "grad_norm": 0.25158539609157676, "learning_rate": 2.2084007044821764e-06, "loss": 0.1805, "step": 1011 }, { "epoch": 4.319829424307036, "grad_norm": 0.2550618363663431, "learning_rate": 2.181224180995336e-06, "loss": 0.1896, "step": 1012 }, { "epoch": 4.32409381663113, "grad_norm": 0.2547599198922931, "learning_rate": 2.154206263848986e-06, "loss": 0.1691, "step": 1013 }, { "epoch": 4.3283582089552235, "grad_norm": 0.2529811695628979, "learning_rate": 2.127347193531757e-06, "loss": 0.197, "step": 1014 }, { "epoch": 4.332622601279318, "grad_norm": 0.26719444732853537, "learning_rate": 2.1006472091183715e-06, "loss": 0.1767, "step": 1015 }, { "epoch": 4.336886993603412, "grad_norm": 0.25726098495581406, "learning_rate": 2.074106548267516e-06, "loss": 0.1745, "step": 1016 }, { "epoch": 4.341151385927505, "grad_norm": 0.2377016385868149, "learning_rate": 2.0477254472197237e-06, "loss": 0.1743, "step": 1017 }, { "epoch": 4.345415778251599, "grad_norm": 0.24345600501457476, "learning_rate": 2.021504140795265e-06, "loss": 0.1734, "step": 1018 }, { "epoch": 4.349680170575693, "grad_norm": 0.2461514291416044, "learning_rate": 1.995442862392081e-06, "loss": 0.1975, "step": 1019 }, { "epoch": 4.353944562899787, "grad_norm": 0.24447636461188885, "learning_rate": 1.96954184398368e-06, "loss": 0.1745, "step": 1020 }, { "epoch": 4.358208955223881, "grad_norm": 0.2406706927340542, "learning_rate": 1.943801316117089e-06, "loss": 0.1663, "step": 1021 }, { "epoch": 4.362473347547974, "grad_norm": 0.24261079701896185, "learning_rate": 1.918221507910789e-06, "loss": 0.1756, "step": 1022 }, { "epoch": 4.366737739872068, "grad_norm": 0.2560728213012035, "learning_rate": 1.8928026470526917e-06, "loss": 0.1909, "step": 1023 }, { "epoch": 4.371002132196162, "grad_norm": 0.24951492778723952, "learning_rate": 1.8675449597980978e-06, "loss": 0.1906, "step": 1024 }, { "epoch": 4.3752665245202556, "grad_norm": 0.2476830068344565, "learning_rate": 1.842448670967687e-06, "loss": 0.166, "step": 1025 }, { "epoch": 4.37953091684435, "grad_norm": 0.25954959096851477, "learning_rate": 1.817514003945524e-06, "loss": 0.1694, "step": 1026 }, { "epoch": 4.383795309168444, "grad_norm": 0.26570027932355017, "learning_rate": 1.792741180677069e-06, "loss": 0.2076, "step": 1027 }, { "epoch": 4.388059701492537, "grad_norm": 0.25639086209798745, "learning_rate": 1.76813042166718e-06, "loss": 0.1793, "step": 1028 }, { "epoch": 4.392324093816631, "grad_norm": 0.2564487505775299, "learning_rate": 1.743681945978184e-06, "loss": 0.1769, "step": 1029 }, { "epoch": 4.396588486140725, "grad_norm": 0.25370708499935873, "learning_rate": 1.7193959712279106e-06, "loss": 0.1768, "step": 1030 }, { "epoch": 4.400852878464819, "grad_norm": 0.24834886176780574, "learning_rate": 1.6952727135877501e-06, "loss": 0.1813, "step": 1031 }, { "epoch": 4.405117270788913, "grad_norm": 0.26274943101751796, "learning_rate": 1.6713123877807413e-06, "loss": 0.1729, "step": 1032 }, { "epoch": 4.409381663113006, "grad_norm": 0.24031929989576914, "learning_rate": 1.6475152070796396e-06, "loss": 0.184, "step": 1033 }, { "epoch": 4.4136460554371, "grad_norm": 0.2559954382971304, "learning_rate": 1.6238813833050504e-06, "loss": 0.1836, "step": 1034 }, { "epoch": 4.417910447761194, "grad_norm": 0.24487155650329637, "learning_rate": 1.6004111268235156e-06, "loss": 0.2054, "step": 1035 }, { "epoch": 4.422174840085288, "grad_norm": 0.2420735342734897, "learning_rate": 1.5771046465456574e-06, "loss": 0.1739, "step": 1036 }, { "epoch": 4.426439232409382, "grad_norm": 0.2608216346583349, "learning_rate": 1.5539621499243064e-06, "loss": 0.182, "step": 1037 }, { "epoch": 4.430703624733475, "grad_norm": 0.2482170636749505, "learning_rate": 1.5309838429526714e-06, "loss": 0.2013, "step": 1038 }, { "epoch": 4.434968017057569, "grad_norm": 0.25334395308346525, "learning_rate": 1.5081699301624819e-06, "loss": 0.1771, "step": 1039 }, { "epoch": 4.439232409381663, "grad_norm": 0.2564039736071258, "learning_rate": 1.4855206146221934e-06, "loss": 0.1826, "step": 1040 }, { "epoch": 4.443496801705757, "grad_norm": 0.2604551013420376, "learning_rate": 1.4630360979351644e-06, "loss": 0.1748, "step": 1041 }, { "epoch": 4.447761194029851, "grad_norm": 0.2503977756454165, "learning_rate": 1.4407165802378664e-06, "loss": 0.1891, "step": 1042 }, { "epoch": 4.452025586353945, "grad_norm": 0.2743350520560739, "learning_rate": 1.4185622601981042e-06, "loss": 0.1844, "step": 1043 }, { "epoch": 4.456289978678038, "grad_norm": 0.2692092257511118, "learning_rate": 1.396573335013236e-06, "loss": 0.186, "step": 1044 }, { "epoch": 4.460554371002132, "grad_norm": 0.3262914630858915, "learning_rate": 1.374750000408438e-06, "loss": 0.1894, "step": 1045 }, { "epoch": 4.464818763326226, "grad_norm": 0.25490515346975273, "learning_rate": 1.353092450634943e-06, "loss": 0.1781, "step": 1046 }, { "epoch": 4.46908315565032, "grad_norm": 0.25118910970659714, "learning_rate": 1.3316008784683265e-06, "loss": 0.1955, "step": 1047 }, { "epoch": 4.473347547974414, "grad_norm": 0.24598518384836726, "learning_rate": 1.3102754752067792e-06, "loss": 0.1787, "step": 1048 }, { "epoch": 4.477611940298507, "grad_norm": 0.2608257108213762, "learning_rate": 1.2891164306694148e-06, "loss": 0.185, "step": 1049 }, { "epoch": 4.481876332622601, "grad_norm": 0.2576580996322884, "learning_rate": 1.2681239331945695e-06, "loss": 0.1917, "step": 1050 }, { "epoch": 4.486140724946695, "grad_norm": 0.2504942501895405, "learning_rate": 1.2472981696381315e-06, "loss": 0.1828, "step": 1051 }, { "epoch": 4.490405117270789, "grad_norm": 0.2536034604981403, "learning_rate": 1.2266393253718812e-06, "loss": 0.1951, "step": 1052 }, { "epoch": 4.494669509594883, "grad_norm": 0.2525231036260257, "learning_rate": 1.2061475842818337e-06, "loss": 0.1762, "step": 1053 }, { "epoch": 4.498933901918977, "grad_norm": 0.25440738093351306, "learning_rate": 1.185823128766601e-06, "loss": 0.1809, "step": 1054 }, { "epoch": 4.50319829424307, "grad_norm": 0.24386942032621886, "learning_rate": 1.1656661397357815e-06, "loss": 0.1674, "step": 1055 }, { "epoch": 4.507462686567164, "grad_norm": 0.2415595398325273, "learning_rate": 1.1456767966083393e-06, "loss": 0.1723, "step": 1056 }, { "epoch": 4.5117270788912585, "grad_norm": 0.2511080328869282, "learning_rate": 1.1258552773110031e-06, "loss": 0.1729, "step": 1057 }, { "epoch": 4.515991471215352, "grad_norm": 0.2502308872486097, "learning_rate": 1.106201758276697e-06, "loss": 0.1752, "step": 1058 }, { "epoch": 4.520255863539446, "grad_norm": 0.2552077136895359, "learning_rate": 1.086716414442952e-06, "loss": 0.1796, "step": 1059 }, { "epoch": 4.524520255863539, "grad_norm": 0.23941103130674032, "learning_rate": 1.0673994192503633e-06, "loss": 0.169, "step": 1060 }, { "epoch": 4.528784648187633, "grad_norm": 0.2568336996961339, "learning_rate": 1.048250944641045e-06, "loss": 0.1826, "step": 1061 }, { "epoch": 4.533049040511727, "grad_norm": 0.25307944722264636, "learning_rate": 1.0292711610570904e-06, "loss": 0.1892, "step": 1062 }, { "epoch": 4.537313432835821, "grad_norm": 0.25965249103962207, "learning_rate": 1.0104602374390593e-06, "loss": 0.1703, "step": 1063 }, { "epoch": 4.541577825159915, "grad_norm": 0.24426354335989117, "learning_rate": 9.918183412244863e-07, "loss": 0.1771, "step": 1064 }, { "epoch": 4.545842217484008, "grad_norm": 0.2645766856661182, "learning_rate": 9.733456383463658e-07, "loss": 0.1809, "step": 1065 }, { "epoch": 4.550106609808102, "grad_norm": 0.25999228059596646, "learning_rate": 9.550422932316938e-07, "loss": 0.1882, "step": 1066 }, { "epoch": 4.554371002132196, "grad_norm": 0.2588645606625913, "learning_rate": 9.369084688000063e-07, "loss": 0.2033, "step": 1067 }, { "epoch": 4.55863539445629, "grad_norm": 0.24825190481868026, "learning_rate": 9.189443264619102e-07, "loss": 0.1724, "step": 1068 }, { "epoch": 4.562899786780384, "grad_norm": 0.23615188533726575, "learning_rate": 9.011500261176742e-07, "loss": 0.1894, "step": 1069 }, { "epoch": 4.567164179104478, "grad_norm": 0.2625288054723175, "learning_rate": 8.835257261557695e-07, "loss": 0.1835, "step": 1070 }, { "epoch": 4.571428571428571, "grad_norm": 0.24898546589666998, "learning_rate": 8.660715834514977e-07, "loss": 0.1734, "step": 1071 }, { "epoch": 4.575692963752665, "grad_norm": 0.24334105931233999, "learning_rate": 8.487877533655698e-07, "loss": 0.1796, "step": 1072 }, { "epoch": 4.5799573560767595, "grad_norm": 0.2636363968555462, "learning_rate": 8.316743897427315e-07, "loss": 0.1705, "step": 1073 }, { "epoch": 4.584221748400853, "grad_norm": 0.241712043228359, "learning_rate": 8.147316449103959e-07, "loss": 0.1795, "step": 1074 }, { "epoch": 4.588486140724947, "grad_norm": 0.2416859186454114, "learning_rate": 7.979596696772773e-07, "loss": 0.1888, "step": 1075 }, { "epoch": 4.59275053304904, "grad_norm": 0.2549025323127069, "learning_rate": 7.813586133320661e-07, "loss": 0.193, "step": 1076 }, { "epoch": 4.597014925373134, "grad_norm": 0.2595086426061621, "learning_rate": 7.649286236420806e-07, "loss": 0.1924, "step": 1077 }, { "epoch": 4.601279317697228, "grad_norm": 0.2517909149989371, "learning_rate": 7.486698468519682e-07, "loss": 0.184, "step": 1078 }, { "epoch": 4.605543710021322, "grad_norm": 0.26019528159860256, "learning_rate": 7.325824276823934e-07, "loss": 0.175, "step": 1079 }, { "epoch": 4.609808102345416, "grad_norm": 0.26605256455150367, "learning_rate": 7.166665093287539e-07, "loss": 0.1747, "step": 1080 }, { "epoch": 4.61407249466951, "grad_norm": 0.2830200475494057, "learning_rate": 7.009222334599019e-07, "loss": 0.1788, "step": 1081 }, { "epoch": 4.618336886993603, "grad_norm": 0.2467861629974839, "learning_rate": 6.853497402168896e-07, "loss": 0.1799, "step": 1082 }, { "epoch": 4.622601279317697, "grad_norm": 0.23737390883921275, "learning_rate": 6.69949168211721e-07, "loss": 0.1917, "step": 1083 }, { "epoch": 4.6268656716417915, "grad_norm": 0.253351168991941, "learning_rate": 6.547206545261109e-07, "loss": 0.1886, "step": 1084 }, { "epoch": 4.631130063965885, "grad_norm": 0.25325937730860365, "learning_rate": 6.396643347102683e-07, "loss": 0.1834, "step": 1085 }, { "epoch": 4.635394456289979, "grad_norm": 0.2614575479574234, "learning_rate": 6.247803427816945e-07, "loss": 0.1809, "step": 1086 }, { "epoch": 4.639658848614072, "grad_norm": 0.23936509091432043, "learning_rate": 6.10068811223985e-07, "loss": 0.1807, "step": 1087 }, { "epoch": 4.643923240938166, "grad_norm": 0.2503677362313397, "learning_rate": 5.955298709856516e-07, "loss": 0.1844, "step": 1088 }, { "epoch": 4.6481876332622605, "grad_norm": 0.24342072360801173, "learning_rate": 5.811636514789598e-07, "loss": 0.1772, "step": 1089 }, { "epoch": 4.652452025586354, "grad_norm": 0.24539213277480382, "learning_rate": 5.669702805787714e-07, "loss": 0.1989, "step": 1090 }, { "epoch": 4.656716417910448, "grad_norm": 0.2652265147460065, "learning_rate": 5.529498846214054e-07, "loss": 0.1863, "step": 1091 }, { "epoch": 4.660980810234541, "grad_norm": 0.2539211580897812, "learning_rate": 5.391025884035239e-07, "loss": 0.1677, "step": 1092 }, { "epoch": 4.665245202558635, "grad_norm": 0.24704196039425894, "learning_rate": 5.254285151810124e-07, "loss": 0.1751, "step": 1093 }, { "epoch": 4.669509594882729, "grad_norm": 0.2560366885967282, "learning_rate": 5.119277866678829e-07, "loss": 0.1935, "step": 1094 }, { "epoch": 4.673773987206823, "grad_norm": 0.24482466047481527, "learning_rate": 4.986005230351954e-07, "loss": 0.184, "step": 1095 }, { "epoch": 4.678038379530917, "grad_norm": 0.2364984365035076, "learning_rate": 4.854468429099801e-07, "loss": 0.1735, "step": 1096 }, { "epoch": 4.682302771855011, "grad_norm": 0.251430758880685, "learning_rate": 4.7246686337419247e-07, "loss": 0.1765, "step": 1097 }, { "epoch": 4.686567164179104, "grad_norm": 0.25154556154939406, "learning_rate": 4.5966069996365993e-07, "loss": 0.2019, "step": 1098 }, { "epoch": 4.690831556503198, "grad_norm": 0.25294495835219155, "learning_rate": 4.470284666670632e-07, "loss": 0.1736, "step": 1099 }, { "epoch": 4.6950959488272925, "grad_norm": 0.2592829877544315, "learning_rate": 4.345702759249171e-07, "loss": 0.1636, "step": 1100 }, { "epoch": 4.699360341151386, "grad_norm": 0.26789209418974264, "learning_rate": 4.22286238628562e-07, "loss": 0.1911, "step": 1101 }, { "epoch": 4.70362473347548, "grad_norm": 0.24541428505242718, "learning_rate": 4.101764641191963e-07, "loss": 0.1851, "step": 1102 }, { "epoch": 4.707889125799573, "grad_norm": 0.2503897444579661, "learning_rate": 3.982410601868858e-07, "loss": 0.1791, "step": 1103 }, { "epoch": 4.712153518123667, "grad_norm": 0.2518436880411174, "learning_rate": 3.8648013306960664e-07, "loss": 0.1772, "step": 1104 }, { "epoch": 4.7164179104477615, "grad_norm": 0.2541166033719267, "learning_rate": 3.748937874523062e-07, "loss": 0.1687, "step": 1105 }, { "epoch": 4.720682302771855, "grad_norm": 0.2474483136739983, "learning_rate": 3.634821264659727e-07, "loss": 0.168, "step": 1106 }, { "epoch": 4.724946695095949, "grad_norm": 0.2497137197767489, "learning_rate": 3.522452516867048e-07, "loss": 0.1805, "step": 1107 }, { "epoch": 4.729211087420042, "grad_norm": 0.2519870636421656, "learning_rate": 3.4118326313481887e-07, "loss": 0.1812, "step": 1108 }, { "epoch": 4.733475479744136, "grad_norm": 0.26608759792221703, "learning_rate": 3.3029625927395446e-07, "loss": 0.1812, "step": 1109 }, { "epoch": 4.73773987206823, "grad_norm": 0.2532206981956829, "learning_rate": 3.1958433701019697e-07, "loss": 0.1882, "step": 1110 }, { "epoch": 4.742004264392325, "grad_norm": 0.24329576225972888, "learning_rate": 3.09047591691225e-07, "loss": 0.1605, "step": 1111 }, { "epoch": 4.746268656716418, "grad_norm": 0.25515549709266605, "learning_rate": 2.9868611710543785e-07, "loss": 0.1673, "step": 1112 }, { "epoch": 4.750533049040512, "grad_norm": 0.24876718599525113, "learning_rate": 2.8850000548115155e-07, "loss": 0.1841, "step": 1113 }, { "epoch": 4.754797441364605, "grad_norm": 0.2510551497387393, "learning_rate": 2.7848934748574863e-07, "loss": 0.2004, "step": 1114 }, { "epoch": 4.759061833688699, "grad_norm": 0.24763934130476553, "learning_rate": 2.686542322248986e-07, "loss": 0.1912, "step": 1115 }, { "epoch": 4.7633262260127935, "grad_norm": 0.25906371020191316, "learning_rate": 2.5899474724174313e-07, "loss": 0.1726, "step": 1116 }, { "epoch": 4.767590618336887, "grad_norm": 0.26739724085295546, "learning_rate": 2.495109785161232e-07, "loss": 0.1842, "step": 1117 }, { "epoch": 4.771855010660981, "grad_norm": 0.2547089358431073, "learning_rate": 2.402030104638198e-07, "loss": 0.1869, "step": 1118 }, { "epoch": 4.776119402985074, "grad_norm": 0.24268462946214736, "learning_rate": 2.3107092593579905e-07, "loss": 0.1925, "step": 1119 }, { "epoch": 4.780383795309168, "grad_norm": 0.2542924795251583, "learning_rate": 2.2211480621746828e-07, "loss": 0.1822, "step": 1120 }, { "epoch": 4.7846481876332625, "grad_norm": 0.2464947967335171, "learning_rate": 2.133347310279632e-07, "loss": 0.1845, "step": 1121 }, { "epoch": 4.788912579957356, "grad_norm": 0.24770021042648263, "learning_rate": 2.0473077851942858e-07, "loss": 0.1832, "step": 1122 }, { "epoch": 4.79317697228145, "grad_norm": 0.25095916292893766, "learning_rate": 1.9630302527633427e-07, "loss": 0.19, "step": 1123 }, { "epoch": 4.797441364605544, "grad_norm": 0.2441082841134644, "learning_rate": 1.8805154631477808e-07, "loss": 0.1985, "step": 1124 }, { "epoch": 4.801705756929637, "grad_norm": 0.24377040955638918, "learning_rate": 1.799764150818306e-07, "loss": 0.1889, "step": 1125 }, { "epoch": 4.8059701492537314, "grad_norm": 0.23922657574755457, "learning_rate": 1.7207770345488263e-07, "loss": 0.2025, "step": 1126 }, { "epoch": 4.810234541577826, "grad_norm": 0.24300973562859218, "learning_rate": 1.6435548174098536e-07, "loss": 0.1651, "step": 1127 }, { "epoch": 4.814498933901919, "grad_norm": 0.250775024372348, "learning_rate": 1.5680981867625566e-07, "loss": 0.1999, "step": 1128 }, { "epoch": 4.818763326226013, "grad_norm": 0.24745461516925246, "learning_rate": 1.4944078142523854e-07, "loss": 0.1882, "step": 1129 }, { "epoch": 4.823027718550106, "grad_norm": 0.2419131887023878, "learning_rate": 1.4224843558031888e-07, "loss": 0.1663, "step": 1130 }, { "epoch": 4.8272921108742, "grad_norm": 0.24717510676764287, "learning_rate": 1.3523284516113955e-07, "loss": 0.1822, "step": 1131 }, { "epoch": 4.8315565031982945, "grad_norm": 0.24718117512657953, "learning_rate": 1.2839407261403091e-07, "loss": 0.1883, "step": 1132 }, { "epoch": 4.835820895522388, "grad_norm": 0.2569349094275073, "learning_rate": 1.2173217881144894e-07, "loss": 0.1968, "step": 1133 }, { "epoch": 4.840085287846482, "grad_norm": 0.24347494132677133, "learning_rate": 1.1524722305144231e-07, "loss": 0.1906, "step": 1134 }, { "epoch": 4.844349680170575, "grad_norm": 0.24501684057809445, "learning_rate": 1.0893926305711289e-07, "loss": 0.1822, "step": 1135 }, { "epoch": 4.848614072494669, "grad_norm": 0.2588930031400982, "learning_rate": 1.0280835497612052e-07, "loss": 0.1701, "step": 1136 }, { "epoch": 4.8528784648187635, "grad_norm": 0.23483922876779595, "learning_rate": 9.685455338016347e-08, "loss": 0.1945, "step": 1137 }, { "epoch": 4.857142857142857, "grad_norm": 0.24724560647261207, "learning_rate": 9.107791126450106e-08, "loss": 0.1802, "step": 1138 }, { "epoch": 4.861407249466951, "grad_norm": 0.26145293127090224, "learning_rate": 8.547848004748505e-08, "loss": 0.1687, "step": 1139 }, { "epoch": 4.865671641791045, "grad_norm": 0.23391703784279536, "learning_rate": 8.005630957010014e-08, "loss": 0.1779, "step": 1140 }, { "epoch": 4.869936034115138, "grad_norm": 0.2489454120460888, "learning_rate": 7.481144809551311e-08, "loss": 0.1659, "step": 1141 }, { "epoch": 4.8742004264392325, "grad_norm": 0.2487457225486811, "learning_rate": 6.974394230865766e-08, "loss": 0.1673, "step": 1142 }, { "epoch": 4.878464818763327, "grad_norm": 0.23375974218421045, "learning_rate": 6.485383731580142e-08, "loss": 0.1815, "step": 1143 }, { "epoch": 4.88272921108742, "grad_norm": 0.23773610901652706, "learning_rate": 6.014117664415953e-08, "loss": 0.1667, "step": 1144 }, { "epoch": 4.886993603411514, "grad_norm": 0.2418877562825462, "learning_rate": 5.560600224149948e-08, "loss": 0.1848, "step": 1145 }, { "epoch": 4.891257995735607, "grad_norm": 0.2509736533987665, "learning_rate": 5.1248354475768034e-08, "loss": 0.1977, "step": 1146 }, { "epoch": 4.895522388059701, "grad_norm": 0.24187318770204605, "learning_rate": 4.706827213473819e-08, "loss": 0.197, "step": 1147 }, { "epoch": 4.899786780383796, "grad_norm": 0.25000706255404975, "learning_rate": 4.3065792425651675e-08, "loss": 0.1752, "step": 1148 }, { "epoch": 4.904051172707889, "grad_norm": 0.2480693668890402, "learning_rate": 3.924095097489922e-08, "loss": 0.1769, "step": 1149 }, { "epoch": 4.908315565031983, "grad_norm": 0.2539622681438357, "learning_rate": 3.559378182770079e-08, "loss": 0.1857, "step": 1150 }, { "epoch": 4.912579957356077, "grad_norm": 0.2385544079469686, "learning_rate": 3.212431744779476e-08, "loss": 0.1762, "step": 1151 }, { "epoch": 4.91684434968017, "grad_norm": 0.2576425107348214, "learning_rate": 2.8832588717164766e-08, "loss": 0.1852, "step": 1152 }, { "epoch": 4.9211087420042645, "grad_norm": 0.24173764751107737, "learning_rate": 2.5718624935744395e-08, "loss": 0.1806, "step": 1153 }, { "epoch": 4.925373134328359, "grad_norm": 0.2558027038778884, "learning_rate": 2.2782453821177386e-08, "loss": 0.1905, "step": 1154 }, { "epoch": 4.929637526652452, "grad_norm": 0.2520433231603843, "learning_rate": 2.0024101508555604e-08, "loss": 0.1682, "step": 1155 }, { "epoch": 4.933901918976546, "grad_norm": 0.24293952334741303, "learning_rate": 1.7443592550190346e-08, "loss": 0.1802, "step": 1156 }, { "epoch": 4.938166311300639, "grad_norm": 0.25128505428584746, "learning_rate": 1.5040949915399173e-08, "loss": 0.1896, "step": 1157 }, { "epoch": 4.9424307036247335, "grad_norm": 0.25347230511353935, "learning_rate": 1.281619499029274e-08, "loss": 0.1926, "step": 1158 }, { "epoch": 4.946695095948828, "grad_norm": 0.24720562161341594, "learning_rate": 1.0769347577594958e-08, "loss": 0.1852, "step": 1159 }, { "epoch": 4.950959488272921, "grad_norm": 0.24835048400025717, "learning_rate": 8.900425896454234e-09, "loss": 0.1724, "step": 1160 }, { "epoch": 4.955223880597015, "grad_norm": 0.25080539790226697, "learning_rate": 7.209446582292501e-09, "loss": 0.1983, "step": 1161 }, { "epoch": 4.959488272921108, "grad_norm": 0.24987301745644983, "learning_rate": 5.6964246866497705e-09, "loss": 0.1728, "step": 1162 }, { "epoch": 4.963752665245202, "grad_norm": 0.24407422413019816, "learning_rate": 4.361373677055358e-09, "loss": 0.1821, "step": 1163 }, { "epoch": 4.968017057569297, "grad_norm": 0.24886407641643385, "learning_rate": 3.2043054369057523e-09, "loss": 0.1897, "step": 1164 }, { "epoch": 4.97228144989339, "grad_norm": 0.25408212065061436, "learning_rate": 2.2252302653558155e-09, "loss": 0.186, "step": 1165 }, { "epoch": 4.976545842217484, "grad_norm": 0.25089979448383903, "learning_rate": 1.4241568772321856e-09, "loss": 0.1861, "step": 1166 }, { "epoch": 4.980810234541578, "grad_norm": 0.2377807867704369, "learning_rate": 8.010924029533406e-10, "loss": 0.1736, "step": 1167 }, { "epoch": 4.985074626865671, "grad_norm": 0.24764069381058376, "learning_rate": 3.560423884629849e-10, "loss": 0.1725, "step": 1168 }, { "epoch": 4.9893390191897655, "grad_norm": 0.24390854347455568, "learning_rate": 8.901079518786048e-11, "loss": 0.1762, "step": 1169 }, { "epoch": 4.99360341151386, "grad_norm": 0.25811352424878503, "learning_rate": 0.0, "loss": 0.1743, "step": 1170 }, { "epoch": 4.99360341151386, "step": 1170, "total_flos": 2.47676830262254e+18, "train_loss": 0.3444926413588035, "train_runtime": 48626.6745, "train_samples_per_second": 3.084, "train_steps_per_second": 0.024 } ], "logging_steps": 1.0, "max_steps": 1170, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.47676830262254e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }