Safetensors
English
qwen3_5
Eval Results

Question about the SWE-bench Verified evaluation setup

#1
by maplebb - opened

Hi, could you share the detailed evaluation methodology for the reported SWE-bench Verified score, including the harness, agent scaffold, inference settings, and evaluation scripts? Thanks!

Hi, thank you so much for your interest in our work. We are actively preparing to open-source the training and evaluation framework.

Sign up or log in to comment