List your company

Code Generation on MBPP EvalPlus

84.9Pass@1Llama 3.1 Instruct
57.96464.95771.9578.943Apr 23, 2025

Evaluation results

MethodLinks
Llama 3.1 InstructModel Scale=70B2025.0484.9
Llama 3 InstructModel Scale=70B2025.0482.8
ParamΔModel Scale=70B2025.0480.7
ParamΔModel Scale=8B2025.0473.3
Llama 3 InstructModel Scale=8B2025.0470.6
Llama 3.1 InstructModel Scale=8B2025.0470.4
Llama 3.1 BaseModel Scale=70B2025.0470.4
Llama 3 BaseModel Scale=70B2025.0466.9
Llama 3.1 BaseModel Scale=8B2025.0459.3
Llama 3 BaseModel Scale=8B2025.0459