Improvement for the issue : Not working with LLM #396 #397
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Explicitly informing the model not to include unnecessary sections in its response. This is done because some models like llama2 70B outputs sections such as "Explanation:" or "Justification:" after the JSON response. This will cause the validate_response function to fail. And as a consequence, console will show the error "Invalid response from model , trying again" infinitely.
This is not a surefire fix though. The model (llama2) will still includes explanation and justification sections at times despite this explicit instruction in the prompt