Qwen3-8B-Material-GRPO-TR/results/validation_mean16.csv

1	step	val_mean16
2	10	0.6795212765957447
3	20	0.706781914893617
4	30	0.7027925531914894
5	40	0.7240691489361702
6	50	0.7406914893617021
7	60	0.7533244680851063
8	70	0.7593085106382979
9	80	0.7506648936170213
10	90	0.7593085106382979
11	100	0.7799202127659575