💾 Reduce memory peak in GRPO by adding max_prompt_length and loop u…
#1058
| Job | Run time |
|---|---|
| 3m 40s | |
| 3m 40s |
max_prompt_length and loop u…
#1058
| Job | Run time |
|---|---|
| 3m 40s | |
| 3m 40s |