Qwen3-4B-Chemistry-SDPO-TR/results/validation_mean16.csv

1	step	val_mean16
2	10	0.558928571429
3	20	0.669345238095
4	30	0.689880952381
5	40	0.677976190476
6	50	0.677976190476
7	60	0.680952380952
8	70	0.682738095238
9	80	0.702380952381
10	90	0.743452380952
11	100	0.749107142857