loss,grad_norm,learning_rate,epoch,step,train_runtime,train_samples_per_second,train_steps_per_second,total_flos,train_loss
0.6262,0.1733129620552063,0.00019,0.050505050505050504,20,,,,,
0.5476,0.08093126118183136,0.00019874255503213152,0.10101010101010101,40,,,,,
0.5033,0.08490929752588272,0.0001947376831715892,0.15151515151515152,60,,,,,
0.5047,0.0910632386803627,0.00018809346758274012,0.20202020202020202,80,,,,,
0.4979,0.09581230580806732,0.0001789950127367606,0.25252525252525254,100,,,,,
0.4942,0.09885909408330917,0.00016769579696382597,0.30303030303030304,120,,,,,
0.4932,0.09994969516992569,0.000154510610675594,0.35353535353535354,140,,,,,
0.4902,0.09873974323272705,0.0001398067864642431,0.40404040404040403,160,,,,,
0.4863,0.10524720698595047,0.00012399396540305205,0.45454545454545453,180,,,,,
0.4748,0.1007479727268219,0.00010751268465418785,0.5050505050505051,200,,,,,
0.4808,0.09715110808610916,9.082210432715197e-05,0.5555555555555556,220,,,,,
0.4821,0.10728275775909424,7.438721551137367e-05,0.6060606060606061,240,,,,,
0.4844,0.11552736908197403,5.866588586066481e-05,0.6565656565656566,260,,,,,
0.4812,0.10522180050611496,4.409610363297211e-05,0.7070707070707071,280,,,,,
0.484,0.10800069570541382,3.108377556000037e-05,0.7575757575757576,300,,,,,
0.4779,0.10630737990140915,1.999141849185938e-05,0.8080808080808081,320,,,,,
0.4786,0.1103220134973526,1.11280598617714e-05,0.8585858585858586,340,,,,,
0.48,0.10225826501846313,4.740628338761255e-06,0.9090909090909091,360,,,,,
0.4753,0.10074862837791443,1.0070745203721532e-06,0.9595959595959596,380,,,,,
,,,1.0,396,758.3116,66.69,0.522,9.241770786768384e+16,0.49630829661783543