agentscope-ai / agentscope-ai/Trinity-RFT
Reward function中的Reward Model应该在哪里初始化？
Open
good first issue
- Dominant language
- Python
- Stars
- 701
- Forks
- 79
- Avg merge
- 8h 7m
- Merged PRs (30d)
- 1
Description
Reward function中的Reward Model在哪里初始化是最好的呢；我目前是在workflow类下初始化的，但是它只能加载到cpu，然后推理打分时会非常非常慢，以至于超时报错
Contributor guide
Assessment
This issue has not been assessed yet.