feat: Reward model #1271
feat: Reward model #1271
Conversation
camel/models/reward/nemetro_model.py
Outdated
| if logprobs | ||
| else {} | ||
| ) | ||
| except (KeyError, IndexError): |
There was a problem hiding this comment.
Maybe add a log here?
| except (KeyError, IndexError): | |
| except (KeyError, IndexError): | |
| logging.error(f"Error parsing scores: {e}") |
There was a problem hiding this comment.
Yeah, I think we shouldn't add the error msg into score dict, better raise error
There was a problem hiding this comment.
Thanks @Asher-hss , as mentioned in the issue, we also want to support skywork (Seq. Classifier), could you also add this model?
Thanks wendong，I plan to update this model today. |
There was a problem hiding this comment.
Thanks @Asher-hss ! Left some comments below and added one more commit here 58d7278 based on the review, feel free to review the change, I think we can add the skywork model in another PR to let this PR merged first, but please add unit test first before the merge, thanks!
camel/models/reward/nemetro_model.py
Outdated
| ) | ||
|
|
||
| @api_keys_required("NVIDIA_API_KEY") | ||
| def evaluate(self, messages: List[OpenAIMessage]) -> Dict[str, float]: |
There was a problem hiding this comment.
set messages type as List[Dict[str, str]] would suit for more general use case
|
The model is called |
I create a new pr to fix it now. |
Description
#889
Motivation and Context
#889
Types of changes
What types of changes does your code introduce? Put an
xin all the boxes that apply:Implemented Tasks
Checklist
Go over all the following points, and put an
xin all the boxes that apply.If you are unsure about any of these, don't hesitate to ask. We are here to help!