deepseek-ai/DeepSeek-Coder-V2

How is the result on SWE-bench obtained?

zkx06111 opened this issue · 1 comments

Is it SWE-bench or SWE-bench lite?
What's the agent frontend producing the results?

SWE-bench lite and CodeACT agent v1.3. We have updated the paper on GitHub and described the agent framework used.