Add tests for tools #209
Add tests for tools #209
Conversation
Tools working with gpt and claude models
richardblythman
force-pushed
the
chore/tool-tests
branch
from
befda44 to
006cc1d
Compare
richardblythman
force-pushed
the
chore/tool-tests
branch
from
006cc1d to
de0218f
Compare
| return response | ||
|
|
||
| if self.llm_provider == "openrouter": | ||
| # TODO investigate the transform parameter https://openrouter.ai/docs#transforms |
There was a problem hiding this comment.
You can remove already this comment
| ): | ||
| results.append(url) | ||
| except Exception: | ||
| pass | ||
| unique_results = list(set(results)) |
There was a problem hiding this comment.
what if all queries fail and you have no urls?
There was a problem hiding this comment.
Not they ever all fail
There was a problem hiding this comment.
I haven't seen it happen at least
|
|
||
|
|
||
| def extract_texts(urls: List[str], num_words: Optional[int] = None) -> List[Document]: | ||
| """Extract texts from URLs with improved error handling, excluding failed URLs.""" | ||
| extracted_texts = [] |
There was a problem hiding this comment.
if the list of urls is empty then there is nothing to batch, you could return an empty list and avoid the loop
There was a problem hiding this comment.
Would this make a difference if the loop is for len(urls).
I'm going to leave this out for now for code clarity since I don't think the list is ever empty.
|
|
||
|
|
||
| class TestPredictionCOT(BaseToolTest): | ||
| """Test Prediction RAG.""" |
There was a problem hiding this comment.
I guess you mean Test Prediction COT
There was a problem hiding this comment.
The whole tool has been deleted. Claude can now be used in prediction_request_rag.
This PR adds tests for tools.
To run the tests, you can use:
or alternatively: