Can it support SWE-bench Lite task parallelism?

princeton-nlp / SWE-agent

SWE-agent takes a GitHub issue and tries to automatically fix it, using GPT-4, or your LM of choice. It solves 12.47% of bugs in the SWE-bench evaluation set and takes just 1 minute to run.

https://princeton-nlp.github.io/SWE-agent/

MIT License

11.86k stars 1.18k forks source link

Can it support SWE-bench Lite task parallelism? #598

Closed rookielyb closed 1 week ago

rookielyb commented 1 week ago

It takes a lot of time to run it now. Thanks!

ofirpress commented 1 week ago

we do not have any support for parallelizing tasks right now. you could just open the agent multiple times and have each swe-agent instance solve 1 task using the filter argument

klieret commented 1 week ago

You could also do what @ofirpress said but filter by repository or a whole range of issues (since the --instance_filter supports any regex)