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

[Bug]: Error in certain Docker images for SWE-Bench instance evaluations #4235

Closed
2 tasks done
sani903 opened this issue Oct 7, 2024 · 3 comments · Fixed by #4241
Closed
2 tasks done

[Bug]: Error in certain Docker images for SWE-Bench instance evaluations #4235

sani903 opened this issue Oct 7, 2024 · 3 comments · Fixed by #4241
Assignees
Labels
bug Something isn't working

Comments

@sani903
Copy link

sani903 commented Oct 7, 2024

Is there an existing issue for the same bug?

Describe the bug

When initializing the environment for instance-level SWE-Bench Docker images, the code to go to the desired directory as base throws an error. It might be related to swe-bench-instance.json. More specifically, for certain instances from repositories like pytest, matplotlib, sphinx, etc. the directory name to change directory to might not be constructed properly with an error concatenating the repo name and its version which leads to an error in making the required directory in line 31 of evaluation/swe_bench/scripts/setup/instance_swe_entry.sh

Some instances which throw this error:

  • matplotlib__matplotlib-23314
  • sphinx-doc__sphinx-7757
  • scikit-learn__scikit-learn-9288
  • pytest-dev__pytest-7432
  • pytest-dev__pytest-5262
  • matplotlib__matplotlib-26466

From all the instances of django and sympy that I have tested, none throw this error.

Current OpenHands version

0.9.7

Installation and Configuration

git clone
make build

Model and Agent

No response

Operating System

Linux

Reproduction Steps

  1. Create config.toml in evaluation/swe_bench and add the instance IDs mentioned as "selected_ids = [ "<error_instance_ids> "]
  2. Run ./evaluation/swe_bench/scripts/run_infer.sh <llm.config>

Logs, Errors, Screenshots, and Additional Context

Screenshot 2024-10-05 at 1 30 33 AM
@sani903 sani903 added the bug Something isn't working label Oct 7, 2024
@neubig
Copy link
Contributor

neubig commented Oct 7, 2024

@xingyaoww it'd be great if you could take a look

@xingyaoww
Copy link
Collaborator

Haven't test it yet (will do soon) - but this should fix it: #4241

@xingyaoww
Copy link
Collaborator

I tested #4241 on matplotlib__matplotlib-23314 which is working. I'll close this for now - @sani903 lmk if this works for you, feel free to re-open if not!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants