Qwen3-4B-Chemical-GRPO-TR/results/validation_mean16.csv

1	step	val_mean16
2	10	0.4398809523809524
3	20	0.5059523809523809
4	30	0.5526785714285715
5	40	0.5904761904761905
6	50	0.6318452380952381
7	60	0.6482142857142857
8	70	0.655952380952381
9	80	0.6657738095238095
10	90	0.6788690476190476
11	100	0.6880952380952381