step,epoch,train_loss,val_loss,learning_rate,grad_norm,step_time_ms
10,1,7.5848,,1.000000e-03,0.5097,204.11
20,1,7.1844,,9.728617e-04,0.6758,223.15
30,1,6.7289,,8.947200e-04,0.4708,210.48
40,1,6.5363,,7.750000e-04,0.3171,215.34
50,1,6.5,,6.281417e-04,0.4004,191.86
60,1,6.4886,,4.718583e-04,0.3225,215.4
70,1,6.4699,,3.250000e-04,0.3582,217.84
80,1,6.4572,,2.052800e-04,0.3365,209.91
90,1,6.4136,,1.271383e-04,0.3819,225.43
100,1,6.4002,,1.000000e-04,0.3906,214.31
50,20,5.3424,,1.435303e-03,0.2586,127.81
100,40,0.2568,,1.192831e-03,0.0383,126.41
50,20,5.358,,1.435303e-03,0.2871,105.68
100,40,0.263,,1.192831e-03,0.0492,119.16
150,60,0.0227,,8.378972e-04,0.0179,119.89
200,80,0.0091,,4.721141e-04,0.0191,154.0
250,100,0.0056,,2.002000e-04,0.0099,120.45
300,120,0.0044,,1.000000e-04,0.0132,98.86
50,12,5.8472,,1.435303e-03,0.3191,101.12
100,23,0.42,,1.192831e-03,0.0661,109.74
150,34,0.0356,,8.378972e-04,0.0342,132.65
200,45,0.015,,4.721141e-04,0.0231,109.05
250,56,0.0096,,2.002000e-04,0.0313,99.42
300,67,0.0078,,1.000000e-04,0.0228,103.03
50,10,6.1239,,1.435303e-03,0.3244,108.98
100,20,0.4734,,1.192831e-03,0.0611,109.15
150,30,0.0374,,8.378972e-04,0.0371,136.83
200,40,0.015,,4.721141e-04,0.0173,109.22
250,50,0.0092,,2.002000e-04,0.0267,94.8
300,60,0.0072,,1.000000e-04,0.0193,108.05
50,10,6.174,,1.435303e-03,0.2936,105.29
100,20,0.5316,,1.192831e-03,0.082,134.02
150,30,0.0418,,8.378972e-04,0.0453,103.7
200,40,0.0164,,4.721141e-04,0.0496,105.05
250,50,0.0102,,2.002000e-04,0.0265,103.54
300,60,0.008,,1.000000e-04,0.0303,102.89
50,1,2.4925,,1.435303e-03,0.0696,129.94
100,2,0.1492,,1.192831e-03,0.037,110.73
150,3,0.0361,,8.378972e-04,0.0185,226.67
200,4,0.0254,,4.721141e-04,0.0211,109.12
250,5,0.0233,,2.002000e-04,0.0141,93.04
300,6,0.0225,,1.000000e-04,0.0121,109.11