search

castorini / rank_llm

RankLLM is a Python toolkit for reproducible information retrieval research using rerankers, with a focus on listwise reranking.
http://rankllm.ai
Apache License 2.0
312 stars 39 forks source link

Test coverage #73

Closed Richard5678 closed 8 months ago

Richard5678 commented 8 months ago
sahel-sh commented 8 months ago

Some high level comments: 1- Please pull and merge with the latest main 2- Make sure your imports are sorted and remove any unused imports 3- Looks like you have removed init.py files which are required for unittest automatic test discovery: (python3 -m unittest discover -p 'test_*.py') How do you run your tests? 4- Similar to 3, but about the import path, looks like you are removing 'src' from the import and patch paths. From where do you run your tests? 5- I have pulled your PR and running the tests fails because of the path config issues mention in 4 and 5, if you think your setting is correct, please share the command(s) that you use for running your tests and the starting folder that you run the tests from.

Richard5678 commented 8 months ago

run pip install -e . at source directory, then navigate to the test folder and run ./run_all.sh to run all tests

sahel-sh commented 8 months ago

run pip install -e . at source directory, then navigate to the test folder and run ./run_all.sh to run all tests

This is not how it should be done, please use __init__.py files to make tests discoverable.