christian-muertz's picture
Upload eval/2025-09-06-05:07:00/README.md with huggingface_hub
83e59a3 verified

SWE-bench Report

This folder contains the evaluation results of the SWE-bench using the official evaluation docker containerization.

Summary

  • total instances: 500
  • submitted instances: 238
  • completed instances: 229
  • empty patch instances: 1
  • resolved instances: 105
  • unresolved instances: 124
  • error instances: 8

Resolved Instances

Unresolved Instances

Error Instances

Empty Patch Instances

Incomplete Instances