| 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,-1.000000,0.001000,19.279182,8030.833008,130568.771506,8.701376,3881.846298,16.530784,4023.512527,0.003360,63.372448,94444.937500 | |
| 1,9.587653,-1.000000,0.001000,40.352005,7238.025391,144870.450628,5.636224,3138.602791,12.841120,4018.376205,0.003264,49.770977,95660.937500 | |