SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice. It can also be employed for offensive cybersecurity or competitive coding challenges. [NeurIPS 2024]
Change: Set SWEBenchInstances.evaluate=False by default
Kilian Lieret46feb1e50efe
diff --git a/sweagent/run/batch_instances.py b/sweagent/run/batch_instances.pyindex 55289a3f..c710a563 100644--- a/sweagent/run/batch_instances.py+++ b/sweagent/run/batch_instances.py@@ -272,7 +272,7 @@ class SWEBenchInstances(BaseModel, AbstractInstanceSource):shuffle: bool = False"""Shuffle the instances (before filtering and slicing)."""- evaluate: bool = True+ evaluate: bool = False"""Run sb-cli to evaluate"""def _get_huggingface_name(self) -> str: