Skip to content

Conversation

@TosakaUCW
Copy link

Problem

src/generate_model_answers.py previously restricted --dataset via argparse choices=LIST_OF_DATASETS. This caused runs to fail at argument parsing time when using dataset names not present in the allowlist (e.g., triviaqa_test).

Change

  • Removed the choices restriction for --dataset
  • Made --dataset required and added a clearer help message

Impact

Allows running the script with both training and test dataset names (including *_test variants) without being blocked by argparse.

How to test

python src/generate_model_answers.py --model <model_name> --dataset triviaqa_test

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.

1 participant