nyaya-3b-v3/training/figures/training_smoke.csv

1	step	epoch	loss	learning_rate	grad_norm	mean_token_accuracy
2	10	0.16	1.8803634643554688	9.678584095202468e-05	1.182433009147644	0.582773034274578
3	20	0.32	1.1098207473754882	8.202965893490878e-05	0.5895639657974243	0.7326480627059937
4	30	0.48	0.9109317779541015	5.896403794053679e-05	0.4854218065738678	0.772037148475647
5	40	0.64	0.8902366638183594	3.357288090445827e-05	0.4858759939670563	0.775420868396759
6	50	0.8	0.8645750999450683	1.2443403456474017e-05	0.4035021960735321	0.7814616322517395
7	60	0.96	0.9060385704040528	1.0572157452321097e-06	0.37886306643486023	0.7762055635452271