Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

61 evalator finish code for lmeh #65

Merged
merged 4 commits into from
Jun 14, 2024
Merged

Conversation

AguirreNicolas
Copy link
Contributor

This PR concludes the Evaluator (hopefully). The main changes can be summarized as:

  1. Responses not properly formated or not answered at all are removed.
  2. Only those docs_ids filled by all its instances/responses are kept. This implies also filtering those docs_ids from self.dataset[self.eval_split].
  3. get_task_manager now initialize properly the logger into EvaluatePocketNetworkConfigurableTask.
  4. In MongoDB , the 'tokenizer' index now include inique=true.
  5. Fix metrics in packages/python/lmeh/utils/task_config.py.
  6. Code to save result in mongoDB in PocketNetworkMongoDBResultNumerical format.

* Remove responses not correct or not answered at all.
* Kept only those docs_ids filled by all its instances/responses
* get_task_manager handles properly the logger into EvaluatePocketNetworkConfigurableTask.
* Tokenizer index now include inique=true
* Fix metrics in packages/python/lmeh/utils/task_config.py
* Save result in mongoDB
@AguirreNicolas AguirreNicolas linked an issue Jun 14, 2024 that may be closed by this pull request
* Update Tokenizer_evaluate (by Rawthil)
* reconstruct_instances now return result_height and do not raise error when no doc_id
@RawthiL RawthiL merged commit 0ab0fbb into main Jun 14, 2024
3 checks passed
@RawthiL RawthiL deleted the 61-evalator-finish-code-for-lmeh branch June 14, 2024 21:18
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Evalator: Finish code for LMEH
2 participants