| { | |
| "best_metric": 0.10709268599748611, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 1.001088139281828, | |
| "eval_steps": 100, | |
| "global_step": 230, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004352557127312296, | |
| "grad_norm": 0.0974438488483429, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0846, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004352557127312296, | |
| "eval_loss": 0.2704615294933319, | |
| "eval_runtime": 31.6023, | |
| "eval_samples_per_second": 12.246, | |
| "eval_steps_per_second": 3.069, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008705114254624592, | |
| "grad_norm": 0.1328469216823578, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1194, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.013057671381936888, | |
| "grad_norm": 0.13327664136886597, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1057, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.017410228509249184, | |
| "grad_norm": 0.1369505226612091, | |
| "learning_rate": 2e-05, | |
| "loss": 0.11, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02176278563656148, | |
| "grad_norm": 0.1252405047416687, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0972, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.026115342763873776, | |
| "grad_norm": 0.10932479798793793, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0971, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.030467899891186073, | |
| "grad_norm": 0.1175251305103302, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.1163, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.03482045701849837, | |
| "grad_norm": 0.11143293976783752, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1114, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03917301414581066, | |
| "grad_norm": 0.12479252368211746, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.1104, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04352557127312296, | |
| "grad_norm": 0.10673598945140839, | |
| "learning_rate": 5e-05, | |
| "loss": 0.101, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04787812840043525, | |
| "grad_norm": 0.10495679080486298, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.0878, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.05223068552774755, | |
| "grad_norm": 0.10622629523277283, | |
| "learning_rate": 6e-05, | |
| "loss": 0.0869, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.056583242655059846, | |
| "grad_norm": 0.1247960776090622, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.111, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.060935799782372145, | |
| "grad_norm": 0.10907791554927826, | |
| "learning_rate": 7e-05, | |
| "loss": 0.0805, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06528835690968444, | |
| "grad_norm": 0.11206729710102081, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0861, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06964091403699674, | |
| "grad_norm": 0.1308465600013733, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0909, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.07399347116430903, | |
| "grad_norm": 0.12455932796001434, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.0886, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.07834602829162132, | |
| "grad_norm": 0.1294569969177246, | |
| "learning_rate": 9e-05, | |
| "loss": 0.1105, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.08269858541893363, | |
| "grad_norm": 0.09594004601240158, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.0918, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08705114254624592, | |
| "grad_norm": 0.10954522341489792, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0988, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09140369967355821, | |
| "grad_norm": 0.10947348922491074, | |
| "learning_rate": 9.999440509051368e-05, | |
| "loss": 0.1166, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0957562568008705, | |
| "grad_norm": 0.10741765797138214, | |
| "learning_rate": 9.997762161417517e-05, | |
| "loss": 0.09, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.10010881392818281, | |
| "grad_norm": 0.0934891551733017, | |
| "learning_rate": 9.994965332706573e-05, | |
| "loss": 0.0894, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1044613710554951, | |
| "grad_norm": 0.09388478845357895, | |
| "learning_rate": 9.991050648838675e-05, | |
| "loss": 0.0821, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1088139281828074, | |
| "grad_norm": 0.10594481974840164, | |
| "learning_rate": 9.986018985905901e-05, | |
| "loss": 0.0978, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.11316648531011969, | |
| "grad_norm": 0.10756991058588028, | |
| "learning_rate": 9.979871469976196e-05, | |
| "loss": 0.0993, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.117519042437432, | |
| "grad_norm": 0.11428479105234146, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.0979, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.12187159956474429, | |
| "grad_norm": 0.13125038146972656, | |
| "learning_rate": 9.964234631709187e-05, | |
| "loss": 0.1166, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.12622415669205658, | |
| "grad_norm": 0.11122190952301025, | |
| "learning_rate": 9.954748808839674e-05, | |
| "loss": 0.0875, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1305767138193689, | |
| "grad_norm": 0.145439013838768, | |
| "learning_rate": 9.944154131125642e-05, | |
| "loss": 0.1288, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13492927094668117, | |
| "grad_norm": 0.14494478702545166, | |
| "learning_rate": 9.932452969617607e-05, | |
| "loss": 0.1195, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.13928182807399347, | |
| "grad_norm": 0.14560213685035706, | |
| "learning_rate": 9.919647942993148e-05, | |
| "loss": 0.1134, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.14363438520130578, | |
| "grad_norm": 0.16578984260559082, | |
| "learning_rate": 9.905741916970864e-05, | |
| "loss": 0.1167, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.14798694232861806, | |
| "grad_norm": 0.17312835156917572, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.124, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.15233949945593037, | |
| "grad_norm": 0.17817851901054382, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.1078, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.15669205658324264, | |
| "grad_norm": 0.24239462614059448, | |
| "learning_rate": 9.857450191464337e-05, | |
| "loss": 0.1241, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.16104461371055495, | |
| "grad_norm": 0.22900445759296417, | |
| "learning_rate": 9.839173742253334e-05, | |
| "loss": 0.1397, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.16539717083786726, | |
| "grad_norm": 0.23307330906391144, | |
| "learning_rate": 9.819814303479267e-05, | |
| "loss": 0.1414, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.16974972796517954, | |
| "grad_norm": 0.25255629420280457, | |
| "learning_rate": 9.799376207714445e-05, | |
| "loss": 0.1491, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.17410228509249184, | |
| "grad_norm": 0.2130127251148224, | |
| "learning_rate": 9.777864028930705e-05, | |
| "loss": 0.1094, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.17845484221980412, | |
| "grad_norm": 0.28394246101379395, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.1465, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.18280739934711643, | |
| "grad_norm": 0.2802410423755646, | |
| "learning_rate": 9.731636918995821e-05, | |
| "loss": 0.1516, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.18715995647442873, | |
| "grad_norm": 0.32573777437210083, | |
| "learning_rate": 9.706932333304517e-05, | |
| "loss": 0.1737, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.191512513601741, | |
| "grad_norm": 0.3744868040084839, | |
| "learning_rate": 9.681174353198687e-05, | |
| "loss": 0.2084, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.19586507072905332, | |
| "grad_norm": 0.40487226843833923, | |
| "learning_rate": 9.654368743221022e-05, | |
| "loss": 0.2048, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.20021762785636563, | |
| "grad_norm": 0.343398779630661, | |
| "learning_rate": 9.626521502369984e-05, | |
| "loss": 0.1814, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.2045701849836779, | |
| "grad_norm": 0.38668176531791687, | |
| "learning_rate": 9.597638862757255e-05, | |
| "loss": 0.2336, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2089227421109902, | |
| "grad_norm": 0.52562016248703, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.2415, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2132752992383025, | |
| "grad_norm": 0.46004876494407654, | |
| "learning_rate": 9.536793472839325e-05, | |
| "loss": 0.2616, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2176278563656148, | |
| "grad_norm": 0.8537291288375854, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.2623, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2219804134929271, | |
| "grad_norm": 0.16656877100467682, | |
| "learning_rate": 9.471887038331685e-05, | |
| "loss": 0.0606, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.22633297062023938, | |
| "grad_norm": 0.18018873035907745, | |
| "learning_rate": 9.437928945022771e-05, | |
| "loss": 0.0753, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2306855277475517, | |
| "grad_norm": 0.1334458738565445, | |
| "learning_rate": 9.40297765928369e-05, | |
| "loss": 0.0806, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.235038084874864, | |
| "grad_norm": 0.10263438522815704, | |
| "learning_rate": 9.367041003085649e-05, | |
| "loss": 0.0816, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.23939064200217627, | |
| "grad_norm": 0.09806181490421295, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.0846, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.24374319912948858, | |
| "grad_norm": 0.09737102687358856, | |
| "learning_rate": 9.292243968009331e-05, | |
| "loss": 0.089, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.24809575625680086, | |
| "grad_norm": 0.09676666557788849, | |
| "learning_rate": 9.253400328436699e-05, | |
| "loss": 0.0774, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.25244831338411317, | |
| "grad_norm": 0.0946255475282669, | |
| "learning_rate": 9.213604793270196e-05, | |
| "loss": 0.0823, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.25680087051142547, | |
| "grad_norm": 0.08449132740497589, | |
| "learning_rate": 9.172866268606513e-05, | |
| "loss": 0.085, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.2611534276387378, | |
| "grad_norm": 0.08588384091854095, | |
| "learning_rate": 9.131193871579975e-05, | |
| "loss": 0.082, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26550598476605003, | |
| "grad_norm": 0.07667472958564758, | |
| "learning_rate": 9.088596928322158e-05, | |
| "loss": 0.0852, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.26985854189336234, | |
| "grad_norm": 0.09125405550003052, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.0978, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.27421109902067464, | |
| "grad_norm": 0.09058438986539841, | |
| "learning_rate": 9.000667740056032e-05, | |
| "loss": 0.1011, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.27856365614798695, | |
| "grad_norm": 0.08789756894111633, | |
| "learning_rate": 8.955355173281708e-05, | |
| "loss": 0.0913, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.28291621327529926, | |
| "grad_norm": 0.0933385044336319, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.0974, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.28726877040261156, | |
| "grad_norm": 0.10251226276159286, | |
| "learning_rate": 8.862084796122998e-05, | |
| "loss": 0.0987, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2916213275299238, | |
| "grad_norm": 0.09278054535388947, | |
| "learning_rate": 8.814147859311332e-05, | |
| "loss": 0.0954, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2959738846572361, | |
| "grad_norm": 0.08714459091424942, | |
| "learning_rate": 8.765357330018056e-05, | |
| "loss": 0.0885, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3003264417845484, | |
| "grad_norm": 0.09142354130744934, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.1035, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.30467899891186073, | |
| "grad_norm": 0.08519690483808517, | |
| "learning_rate": 8.665259359149132e-05, | |
| "loss": 0.0818, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.30903155603917304, | |
| "grad_norm": 0.08325667679309845, | |
| "learning_rate": 8.613974319136958e-05, | |
| "loss": 0.087, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.3133841131664853, | |
| "grad_norm": 0.10583508759737015, | |
| "learning_rate": 8.561880484756725e-05, | |
| "loss": 0.1153, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3177366702937976, | |
| "grad_norm": 0.09730641543865204, | |
| "learning_rate": 8.508989514419958e-05, | |
| "loss": 0.1031, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.3220892274211099, | |
| "grad_norm": 0.09521140158176422, | |
| "learning_rate": 8.455313244934324e-05, | |
| "loss": 0.0912, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.3264417845484222, | |
| "grad_norm": 0.09912557899951935, | |
| "learning_rate": 8.400863688854597e-05, | |
| "loss": 0.0981, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3307943416757345, | |
| "grad_norm": 0.10728394985198975, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.1066, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.33514689880304677, | |
| "grad_norm": 0.08985297381877899, | |
| "learning_rate": 8.289693629698564e-05, | |
| "loss": 0.0884, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3394994559303591, | |
| "grad_norm": 0.09841588884592056, | |
| "learning_rate": 8.232998006078997e-05, | |
| "loss": 0.0966, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.3438520130576714, | |
| "grad_norm": 0.10963951796293259, | |
| "learning_rate": 8.175578849210895e-05, | |
| "loss": 0.0995, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.3482045701849837, | |
| "grad_norm": 0.1173572987318039, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1032, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.352557127312296, | |
| "grad_norm": 0.12008894979953766, | |
| "learning_rate": 8.058621495575032e-05, | |
| "loss": 0.0989, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.35690968443960824, | |
| "grad_norm": 0.12407906353473663, | |
| "learning_rate": 7.999109473439569e-05, | |
| "loss": 0.0998, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.36126224156692055, | |
| "grad_norm": 0.14508754014968872, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.1313, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.36561479869423286, | |
| "grad_norm": 0.13731609284877777, | |
| "learning_rate": 7.878085328428369e-05, | |
| "loss": 0.1029, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.36996735582154516, | |
| "grad_norm": 0.15624341368675232, | |
| "learning_rate": 7.81660029031811e-05, | |
| "loss": 0.1135, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.37431991294885747, | |
| "grad_norm": 0.17237362265586853, | |
| "learning_rate": 7.754484907260513e-05, | |
| "loss": 0.1182, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3786724700761698, | |
| "grad_norm": 0.21730513870716095, | |
| "learning_rate": 7.691753080453412e-05, | |
| "loss": 0.1494, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.383025027203482, | |
| "grad_norm": 0.22640293836593628, | |
| "learning_rate": 7.628418849052523e-05, | |
| "loss": 0.1786, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.38737758433079433, | |
| "grad_norm": 0.21066981554031372, | |
| "learning_rate": 7.564496387029532e-05, | |
| "loss": 0.135, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.39173014145810664, | |
| "grad_norm": 0.31958574056625366, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.1758, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.39608269858541895, | |
| "grad_norm": 0.2449241578578949, | |
| "learning_rate": 7.434944122021836e-05, | |
| "loss": 0.111, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.40043525571273125, | |
| "grad_norm": 0.280224472284317, | |
| "learning_rate": 7.369343312364993e-05, | |
| "loss": 0.1551, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.4047878128400435, | |
| "grad_norm": 0.29259204864501953, | |
| "learning_rate": 7.303212252253162e-05, | |
| "loss": 0.1603, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.4091403699673558, | |
| "grad_norm": 0.32008427381515503, | |
| "learning_rate": 7.236565741578163e-05, | |
| "loss": 0.1527, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.4134929270946681, | |
| "grad_norm": 0.3090142011642456, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1446, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.4178454842219804, | |
| "grad_norm": 0.4006205201148987, | |
| "learning_rate": 7.101786141547828e-05, | |
| "loss": 0.2507, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.42219804134929273, | |
| "grad_norm": 0.5021695494651794, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 0.265, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.426550598476605, | |
| "grad_norm": 0.3973623514175415, | |
| "learning_rate": 6.965125158269619e-05, | |
| "loss": 0.1922, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4309031556039173, | |
| "grad_norm": 0.5247178673744202, | |
| "learning_rate": 6.896127313264643e-05, | |
| "loss": 0.2127, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4352557127312296, | |
| "grad_norm": 0.5562911033630371, | |
| "learning_rate": 6.826705121831976e-05, | |
| "loss": 0.2448, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4352557127312296, | |
| "eval_loss": 0.11724230647087097, | |
| "eval_runtime": 32.1058, | |
| "eval_samples_per_second": 12.054, | |
| "eval_steps_per_second": 3.021, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4396082698585419, | |
| "grad_norm": 0.08145804703235626, | |
| "learning_rate": 6.756874120406714e-05, | |
| "loss": 0.0543, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.4439608269858542, | |
| "grad_norm": 0.08369223773479462, | |
| "learning_rate": 6.686649936914152e-05, | |
| "loss": 0.0668, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.44831338411316646, | |
| "grad_norm": 0.07429920136928558, | |
| "learning_rate": 6.616048287272301e-05, | |
| "loss": 0.0694, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.45266594124047876, | |
| "grad_norm": 0.07733328640460968, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.0881, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.45701849836779107, | |
| "grad_norm": 0.07064362615346909, | |
| "learning_rate": 6.473775872054521e-05, | |
| "loss": 0.0777, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4613710554951034, | |
| "grad_norm": 0.06992431730031967, | |
| "learning_rate": 6.402136946530014e-05, | |
| "loss": 0.0802, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4657236126224157, | |
| "grad_norm": 0.0750981792807579, | |
| "learning_rate": 6.330184227833376e-05, | |
| "loss": 0.0868, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.470076169749728, | |
| "grad_norm": 0.06660962849855423, | |
| "learning_rate": 6.257933818722543e-05, | |
| "loss": 0.0796, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.47442872687704024, | |
| "grad_norm": 0.06947064399719238, | |
| "learning_rate": 6.185401888577488e-05, | |
| "loss": 0.0816, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.47878128400435255, | |
| "grad_norm": 0.07036959379911423, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.0802, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.48313384113166485, | |
| "grad_norm": 0.0723167285323143, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 0.0793, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.48748639825897716, | |
| "grad_norm": 0.06801431626081467, | |
| "learning_rate": 5.9662795889777666e-05, | |
| "loss": 0.0785, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.49183895538628947, | |
| "grad_norm": 0.06911415606737137, | |
| "learning_rate": 5.8927844739931834e-05, | |
| "loss": 0.0724, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4961915125136017, | |
| "grad_norm": 0.07668229192495346, | |
| "learning_rate": 5.819089557075689e-05, | |
| "loss": 0.0836, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.500544069640914, | |
| "grad_norm": 0.08050959557294846, | |
| "learning_rate": 5.745211330880872e-05, | |
| "loss": 0.0801, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5048966267682263, | |
| "grad_norm": 0.07494529336690903, | |
| "learning_rate": 5.6711663290882776e-05, | |
| "loss": 0.0771, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.5092491838955386, | |
| "grad_norm": 0.08858672529459, | |
| "learning_rate": 5.596971122701221e-05, | |
| "loss": 0.0914, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.5136017410228509, | |
| "grad_norm": 0.07994210720062256, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.0827, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.5179542981501633, | |
| "grad_norm": 0.09033293277025223, | |
| "learning_rate": 5.448196544517168e-05, | |
| "loss": 0.0911, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.5223068552774756, | |
| "grad_norm": 0.08702769875526428, | |
| "learning_rate": 5.373650467932122e-05, | |
| "loss": 0.095, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5266594124047879, | |
| "grad_norm": 0.08354932814836502, | |
| "learning_rate": 5.299020769725172e-05, | |
| "loss": 0.0756, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.5310119695321001, | |
| "grad_norm": 0.08543534576892853, | |
| "learning_rate": 5.2243241517525754e-05, | |
| "loss": 0.0846, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5353645266594124, | |
| "grad_norm": 0.082368865609169, | |
| "learning_rate": 5.149577330846993e-05, | |
| "loss": 0.0827, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5397170837867247, | |
| "grad_norm": 0.10136285424232483, | |
| "learning_rate": 5.074797035076319e-05, | |
| "loss": 0.1007, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.544069640914037, | |
| "grad_norm": 0.09355714172124863, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0842, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5484221980413493, | |
| "grad_norm": 0.0911218672990799, | |
| "learning_rate": 4.925202964923683e-05, | |
| "loss": 0.0938, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5527747551686616, | |
| "grad_norm": 0.10014523565769196, | |
| "learning_rate": 4.850422669153009e-05, | |
| "loss": 0.0899, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5571273122959739, | |
| "grad_norm": 0.11333120614290237, | |
| "learning_rate": 4.775675848247427e-05, | |
| "loss": 0.0994, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5614798694232862, | |
| "grad_norm": 0.10408790409564972, | |
| "learning_rate": 4.700979230274829e-05, | |
| "loss": 0.0997, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5658324265505985, | |
| "grad_norm": 0.10925531387329102, | |
| "learning_rate": 4.626349532067879e-05, | |
| "loss": 0.1019, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5701849836779108, | |
| "grad_norm": 0.12819620966911316, | |
| "learning_rate": 4.551803455482833e-05, | |
| "loss": 0.1163, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5745375408052231, | |
| "grad_norm": 0.10823093354701996, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.0873, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5788900979325353, | |
| "grad_norm": 0.13952915370464325, | |
| "learning_rate": 4.403028877298779e-05, | |
| "loss": 0.1062, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5832426550598476, | |
| "grad_norm": 0.14679069817066193, | |
| "learning_rate": 4.328833670911724e-05, | |
| "loss": 0.1156, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5875952121871599, | |
| "grad_norm": 0.14709697663784027, | |
| "learning_rate": 4.254788669119127e-05, | |
| "loss": 0.1068, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5919477693144722, | |
| "grad_norm": 0.1794285625219345, | |
| "learning_rate": 4.180910442924312e-05, | |
| "loss": 0.1481, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5963003264417845, | |
| "grad_norm": 0.16199396550655365, | |
| "learning_rate": 4.107215526006817e-05, | |
| "loss": 0.1008, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.6006528835690969, | |
| "grad_norm": 0.20580187439918518, | |
| "learning_rate": 4.0337204110222346e-05, | |
| "loss": 0.1383, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.6050054406964092, | |
| "grad_norm": 0.2128140777349472, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 0.1014, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.6093579978237215, | |
| "grad_norm": 0.2550433874130249, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.1657, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6137105549510338, | |
| "grad_norm": 0.23781460523605347, | |
| "learning_rate": 3.814598111422513e-05, | |
| "loss": 0.1541, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.6180631120783461, | |
| "grad_norm": 0.2926751375198364, | |
| "learning_rate": 3.742066181277458e-05, | |
| "loss": 0.1675, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.6224156692056583, | |
| "grad_norm": 0.3387477993965149, | |
| "learning_rate": 3.6698157721666246e-05, | |
| "loss": 0.2082, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.6267682263329706, | |
| "grad_norm": 0.28316712379455566, | |
| "learning_rate": 3.597863053469987e-05, | |
| "loss": 0.1789, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.6311207834602829, | |
| "grad_norm": 0.3266338109970093, | |
| "learning_rate": 3.5262241279454785e-05, | |
| "loss": 0.2029, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.6354733405875952, | |
| "grad_norm": 0.3332321345806122, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.186, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.6398258977149075, | |
| "grad_norm": 0.3611805737018585, | |
| "learning_rate": 3.383951712727701e-05, | |
| "loss": 0.1816, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.6441784548422198, | |
| "grad_norm": 0.4147430956363678, | |
| "learning_rate": 3.313350063085851e-05, | |
| "loss": 0.2133, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6485310119695321, | |
| "grad_norm": 0.48492154479026794, | |
| "learning_rate": 3.243125879593286e-05, | |
| "loss": 0.2802, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6528835690968444, | |
| "grad_norm": 0.7098520994186401, | |
| "learning_rate": 3.173294878168025e-05, | |
| "loss": 0.2229, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6572361262241567, | |
| "grad_norm": 0.06466802954673767, | |
| "learning_rate": 3.103872686735358e-05, | |
| "loss": 0.0642, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.661588683351469, | |
| "grad_norm": 0.07565873116254807, | |
| "learning_rate": 3.0348748417303823e-05, | |
| "loss": 0.0699, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6659412404787813, | |
| "grad_norm": 0.07711576670408249, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 0.0826, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6702937976060935, | |
| "grad_norm": 0.06918355822563171, | |
| "learning_rate": 2.8982138584521735e-05, | |
| "loss": 0.0698, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6746463547334058, | |
| "grad_norm": 0.07764638215303421, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0778, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6789989118607181, | |
| "grad_norm": 0.06955543160438538, | |
| "learning_rate": 2.7634342584218365e-05, | |
| "loss": 0.0755, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6833514689880305, | |
| "grad_norm": 0.07822169363498688, | |
| "learning_rate": 2.6967877477468397e-05, | |
| "loss": 0.0896, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6877040261153428, | |
| "grad_norm": 0.07690484076738358, | |
| "learning_rate": 2.630656687635007e-05, | |
| "loss": 0.0682, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6920565832426551, | |
| "grad_norm": 0.07423487305641174, | |
| "learning_rate": 2.5650558779781635e-05, | |
| "loss": 0.0728, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6964091403699674, | |
| "grad_norm": 0.07923182845115662, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.0781, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7007616974972797, | |
| "grad_norm": 0.07322314381599426, | |
| "learning_rate": 2.43550361297047e-05, | |
| "loss": 0.0846, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.705114254624592, | |
| "grad_norm": 0.07659196853637695, | |
| "learning_rate": 2.371581150947476e-05, | |
| "loss": 0.0837, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.7094668117519043, | |
| "grad_norm": 0.0735752210021019, | |
| "learning_rate": 2.3082469195465893e-05, | |
| "loss": 0.0797, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.7138193688792165, | |
| "grad_norm": 0.07915199548006058, | |
| "learning_rate": 2.245515092739488e-05, | |
| "loss": 0.0855, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.7181719260065288, | |
| "grad_norm": 0.07221949100494385, | |
| "learning_rate": 2.1833997096818898e-05, | |
| "loss": 0.0744, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.7225244831338411, | |
| "grad_norm": 0.08134427666664124, | |
| "learning_rate": 2.1219146715716332e-05, | |
| "loss": 0.0931, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.7268770402611534, | |
| "grad_norm": 0.08584997057914734, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.0956, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.7312295973884657, | |
| "grad_norm": 0.08483581990003586, | |
| "learning_rate": 2.0008905265604316e-05, | |
| "loss": 0.0951, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.735582154515778, | |
| "grad_norm": 0.08276212960481644, | |
| "learning_rate": 1.9413785044249678e-05, | |
| "loss": 0.0957, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.7399347116430903, | |
| "grad_norm": 0.08514475077390671, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0888, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7442872687704026, | |
| "grad_norm": 0.07838297635316849, | |
| "learning_rate": 1.8244211507891063e-05, | |
| "loss": 0.0737, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.7486398258977149, | |
| "grad_norm": 0.08513329923152924, | |
| "learning_rate": 1.7670019939210024e-05, | |
| "loss": 0.0859, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7529923830250272, | |
| "grad_norm": 0.08508553355932236, | |
| "learning_rate": 1.7103063703014372e-05, | |
| "loss": 0.0912, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7573449401523396, | |
| "grad_norm": 0.09576783329248428, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.1088, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.7616974972796517, | |
| "grad_norm": 0.09258586913347244, | |
| "learning_rate": 1.599136311145402e-05, | |
| "loss": 0.0915, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.766050054406964, | |
| "grad_norm": 0.10096845775842667, | |
| "learning_rate": 1.544686755065677e-05, | |
| "loss": 0.0967, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7704026115342764, | |
| "grad_norm": 0.10371614992618561, | |
| "learning_rate": 1.4910104855800427e-05, | |
| "loss": 0.1021, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7747551686615887, | |
| "grad_norm": 0.09611905366182327, | |
| "learning_rate": 1.438119515243277e-05, | |
| "loss": 0.0893, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.779107725788901, | |
| "grad_norm": 0.09496966004371643, | |
| "learning_rate": 1.3860256808630428e-05, | |
| "loss": 0.0871, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.7834602829162133, | |
| "grad_norm": 0.10661742091178894, | |
| "learning_rate": 1.3347406408508695e-05, | |
| "loss": 0.0838, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7878128400435256, | |
| "grad_norm": 0.12279023230075836, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 0.0967, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7921653971708379, | |
| "grad_norm": 0.11622121185064316, | |
| "learning_rate": 1.2346426699819458e-05, | |
| "loss": 0.0953, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7965179542981502, | |
| "grad_norm": 0.15825197100639343, | |
| "learning_rate": 1.1858521406886675e-05, | |
| "loss": 0.1235, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.8008705114254625, | |
| "grad_norm": 0.15962378680706024, | |
| "learning_rate": 1.137915203877003e-05, | |
| "loss": 0.1179, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.8052230685527747, | |
| "grad_norm": 0.17821140587329865, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.1458, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.809575625680087, | |
| "grad_norm": 0.1665080189704895, | |
| "learning_rate": 1.0446448267182952e-05, | |
| "loss": 0.1239, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.8139281828073993, | |
| "grad_norm": 0.1886962503194809, | |
| "learning_rate": 9.993322599439692e-06, | |
| "loss": 0.1275, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.8182807399347116, | |
| "grad_norm": 0.18211908638477325, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.1026, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.8226332970620239, | |
| "grad_norm": 0.24852445721626282, | |
| "learning_rate": 9.114030716778433e-06, | |
| "loss": 0.1498, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.8269858541893362, | |
| "grad_norm": 0.231397345662117, | |
| "learning_rate": 8.688061284200266e-06, | |
| "loss": 0.1242, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8313384113166485, | |
| "grad_norm": 0.26934152841567993, | |
| "learning_rate": 8.271337313934869e-06, | |
| "loss": 0.1469, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.8356909684439608, | |
| "grad_norm": 0.2680892050266266, | |
| "learning_rate": 7.863952067298042e-06, | |
| "loss": 0.1441, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.8400435255712732, | |
| "grad_norm": 0.33063918352127075, | |
| "learning_rate": 7.465996715633028e-06, | |
| "loss": 0.1923, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.8443960826985855, | |
| "grad_norm": 0.2599483132362366, | |
| "learning_rate": 7.077560319906695e-06, | |
| "loss": 0.1467, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.8487486398258978, | |
| "grad_norm": 0.30372336506843567, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.1614, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.85310119695321, | |
| "grad_norm": 0.36303284764289856, | |
| "learning_rate": 6.329589969143518e-06, | |
| "loss": 0.2805, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.8574537540805223, | |
| "grad_norm": 0.37215545773506165, | |
| "learning_rate": 5.9702234071631e-06, | |
| "loss": 0.2837, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8618063112078346, | |
| "grad_norm": 0.4205995500087738, | |
| "learning_rate": 5.620710549772295e-06, | |
| "loss": 0.312, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.8661588683351469, | |
| "grad_norm": 0.5164013504981995, | |
| "learning_rate": 5.281129616683167e-06, | |
| "loss": 0.2912, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8705114254624592, | |
| "grad_norm": 0.5630713105201721, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.2237, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8705114254624592, | |
| "eval_loss": 0.10709268599748611, | |
| "eval_runtime": 32.1216, | |
| "eval_samples_per_second": 12.048, | |
| "eval_steps_per_second": 3.02, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8748639825897715, | |
| "grad_norm": 0.04893573746085167, | |
| "learning_rate": 4.632065271606756e-06, | |
| "loss": 0.0464, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8792165397170838, | |
| "grad_norm": 0.054525069892406464, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.0653, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8835690968443961, | |
| "grad_norm": 0.06352470815181732, | |
| "learning_rate": 4.023611372427471e-06, | |
| "loss": 0.07, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8879216539717084, | |
| "grad_norm": 0.05715398117899895, | |
| "learning_rate": 3.734784976300165e-06, | |
| "loss": 0.0565, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8922742110990207, | |
| "grad_norm": 0.06614100933074951, | |
| "learning_rate": 3.4563125677897932e-06, | |
| "loss": 0.0706, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8966267682263329, | |
| "grad_norm": 0.07538996636867523, | |
| "learning_rate": 3.18825646801314e-06, | |
| "loss": 0.081, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.9009793253536452, | |
| "grad_norm": 0.07782811671495438, | |
| "learning_rate": 2.930676666954846e-06, | |
| "loss": 0.0761, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.9053318824809575, | |
| "grad_norm": 0.07623429596424103, | |
| "learning_rate": 2.6836308100417873e-06, | |
| "loss": 0.0808, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.9096844396082698, | |
| "grad_norm": 0.06650186330080032, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.0677, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.9140369967355821, | |
| "grad_norm": 0.0781843438744545, | |
| "learning_rate": 2.221359710692961e-06, | |
| "loss": 0.0835, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9183895538628944, | |
| "grad_norm": 0.08650253713130951, | |
| "learning_rate": 2.006237922855553e-06, | |
| "loss": 0.0895, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.9227421109902068, | |
| "grad_norm": 0.08480913937091827, | |
| "learning_rate": 1.8018569652073381e-06, | |
| "loss": 0.0849, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.9270946681175191, | |
| "grad_norm": 0.07606717199087143, | |
| "learning_rate": 1.6082625774666794e-06, | |
| "loss": 0.0723, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.9314472252448314, | |
| "grad_norm": 0.08265078812837601, | |
| "learning_rate": 1.4254980853566247e-06, | |
| "loss": 0.0782, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.9357997823721437, | |
| "grad_norm": 0.07758332788944244, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0664, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.940152339499456, | |
| "grad_norm": 0.08595745265483856, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.0833, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.9445048966267682, | |
| "grad_norm": 0.09878195077180862, | |
| "learning_rate": 9.42580830291373e-07, | |
| "loss": 0.1022, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.9488574537540805, | |
| "grad_norm": 0.1008504182100296, | |
| "learning_rate": 8.035205700685167e-07, | |
| "loss": 0.0786, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.9532100108813928, | |
| "grad_norm": 0.10530146956443787, | |
| "learning_rate": 6.75470303823933e-07, | |
| "loss": 0.0882, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.9575625680087051, | |
| "grad_norm": 0.11855965107679367, | |
| "learning_rate": 5.584586887435739e-07, | |
| "loss": 0.0952, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9619151251360174, | |
| "grad_norm": 0.129081130027771, | |
| "learning_rate": 4.52511911603265e-07, | |
| "loss": 0.0809, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.9662676822633297, | |
| "grad_norm": 0.16669493913650513, | |
| "learning_rate": 3.576536829081323e-07, | |
| "loss": 0.1204, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.970620239390642, | |
| "grad_norm": 0.16019439697265625, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 0.1158, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9749727965179543, | |
| "grad_norm": 0.2115449756383896, | |
| "learning_rate": 2.012853002380466e-07, | |
| "loss": 0.1401, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9793253536452666, | |
| "grad_norm": 0.23436211049556732, | |
| "learning_rate": 1.3981014094099353e-07, | |
| "loss": 0.1593, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9836779107725789, | |
| "grad_norm": 0.24148912727832794, | |
| "learning_rate": 8.949351161324227e-08, | |
| "loss": 0.1317, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9880304678998912, | |
| "grad_norm": 0.30770501494407654, | |
| "learning_rate": 5.0346672934270534e-08, | |
| "loss": 0.162, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9923830250272034, | |
| "grad_norm": 0.2943860590457916, | |
| "learning_rate": 2.237838582483387e-08, | |
| "loss": 0.1882, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9967355821545157, | |
| "grad_norm": 0.3997349441051483, | |
| "learning_rate": 5.594909486328348e-09, | |
| "loss": 0.2187, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.001088139281828, | |
| "grad_norm": 0.5895123481750488, | |
| "learning_rate": 0.0, | |
| "loss": 0.2661, | |
| "step": 230 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 230, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.7320134588104704e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |