FrogNano: Training a 4B Coding Agent via Online Task Synthesis
Open paper
Benchmark
Model
Rank
Results
code-generation-on-swe-bench-pro
FrogNano-4B-2609
#41
Resolved: 37.6
code-generation-on-swe-bench-verified
FrogNano-4B-2609 (Leaf)
#60
Accuracy: 61.5
Rank counts only results with a code link.