[LLM] Rewrite GSM8K reward function to follow standard GRPO conventions #8552
| Job | Run time |
|---|---|
| 29m 9s | |
| 1h 0m 0s | |
| 5m 52s | |
| 16m 31s | |
| 58m 54s | |
| 12m 44s | |
| 11m 40s | |
| 5m 37s | |
| 29m 36s | |
| 51m 23s | |
| 53m 48s | |
| 52m 58s | |
| 35m 54s | |
| 50m 31s | |
| 12m 41s | |
| 11m 27s | |
| 16m 23s | |
| 13s | |
| 8h 35m 21s |
| Job | Run time |
|---|---|
| 29m 9s | |
| 1h 0m 0s | |
| 5m 52s | |
| 16m 31s | |
| 58m 54s | |
| 12m 44s | |
| 11m 40s | |
| 5m 37s | |
| 29m 36s | |
| 51m 23s | |
| 53m 48s | |
| 52m 58s | |
| 35m 54s | |
| 50m 31s | |
| 12m 41s | |
| 11m 27s | |
| 16m 23s | |
| 13s | |
| 8h 35m 21s |