| 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 |
| 0,2.749795,-1.000000,0.000006,0.369926,19070.187500,54985.091258,28.417728,9397.607430,23.844800,9498.981613,0.003232,61.891327,52292.937500 |
| 1,2.768989,-1.000000,0.000012,0.369283,15180.346680,69074.575313,22.225088,5218.900940,15.223200,9833.401306,0.003392,48.674625,52708.937500 |
| 2,2.739640,-1.000000,0.000018,0.367964,15601.478516,67210.040314,22.717760,5712.806496,15.106144,9759.792923,0.003072,48.829056,52644.937500 |
| 3,2.713684,-1.000000,0.000024,0.336673,13912.223633,75370.841332,21.858464,4363.964279,15.188960,9419.757721,0.003360,48.711327,52644.937500 |
|
|
