global_step,eval_loss,train_loss
3200,8.98454891900771,6.239049434661865
6400,12.710789349633243,6.5854692459106445
8242,11.180710541745055,6.323003768920898
9600,10.88143311352015,6.214521408081055
12800,10.264017345256333,6.225090980529785
16000,8.552877298666978,6.213644027709961
16484,9.861861456603116,6.396312236785889
19200,9.872641939147476,5.7665114402771
22400,11.80832740417214,6.291630744934082
24726,9.768011126455782,6.106754302978516
25600,10.007712807023717,6.074300765991211
28800,9.315486268474315,6.214197158813477
32000,11.036558783326202,6.228422164916992
32968,11.823396353279717,6.375258922576904
35200,9.315472150300094,6.675766468048096
38400,11.318372732918567,6.045510768890381
41210,11.067470789765455,6.386971473693848
41600,11.831813449480185,6.32453727722168
44800,12.112716560049117,6.689652919769287