SWE-smith: Scaling Data for Software Engineering Agents
Open paper
Benchmark
Model
Rank
Results
code-generation-on-swe-bench-multilingual
SWE-agent-LM-32B
#40
Accuracy: 8.4
code-generation-on-swe-bench-verified
SWE-agent-LM-32B
#80
Accuracy: 40.2
Rank counts only results with a code link.