| step,train_loss,val_loss,lr,grad_norm,dt_ms,tok_per_sec,timer_data,timer_fwd,timer_loss,timer_bwd,timer_clip,timer_optim,mem_gpu_mb |
| 0,11.091398,-1.000000,0.000600,19.279064,7024.300781,149278.345654,9.299328,2776.847267,16.714656,4120.889542,0.003296,64.269791,94444.937500 |
| 1,9.663927,-1.000000,0.000600,31.903540,6492.857910,161496.834600,6.008896,2316.131859,13.071328,4093.511032,0.003744,51.003872,95660.937500 |
| 2,9.033846,-1.000000,0.000330,108.185303,6453.253906,162487.950301,6.007296,2409.168556,13.055424,3960.884277,0.003392,51.114334,95660.937500 |
|
|
