Share your thoughts, 1 month free Claude Pro on usSee more

Code Generation on HumanEval (HE)

48.7HumanEval Score

95% (CodeEnhance-gemma-4b+CodeQA) + 5% Raw

Updated 1mo ago

Evaluation Results

Method	Links
95% (CodeEnhance-gemma-4b+CodeQA) + 5% Raw 2026.06		48.7
Nemotron-RW 2026.06		48.2
80% CodeEnhance-gemma-4b + 15% CodeQA + 5% Raw 2026.06		46.3
90% (CodeEnhance-gemma-4b+CodeQA) + 10% Raw 2026.06		45.7
50% CodeEnhance + 50% Nemotron-RW 2026.06		45.7
CodeEnhance+Nemotron-RW+SwallowCode 2026.06		45.1
CodeEnhance+CodeQA 2026.06		45.1
SwallowCode 2026.06		43.9
85% CodeEnhance-gemma-4b + 10% CodeQA + 5% Raw 2026.06		43.3
75% CodeEnhance-gemma-4b + 15% CodeQA + 10% Raw 2026.06		43.3
95% CodeEnhance+CodeQA + 5% Raw 2026.06		42.7
95% Nemotron-RW + 5% Raw 2026.06		41.5
CodeEnhance-gemma-4b 2026.06		38.4
90% CodeEnhance+CodeQA + 10% Raw 2026.06		37.8
95% SwallowCode + 5% Raw 2026.06		37.2
CodeEnhance 2026.06		34.1
CodeEnhance-gemma-1b 2026.06		32.9