Coder Reviewer Reranking for Code Generation

Benchmark Model Rank Results
code-generation-on-mbppcode-davinci-002 175B + Reviewer#25Accuracy: 66.9
code-generation-on-mbppcode-davinci-002 175B + Coder-Reviewer#26Accuracy: 66.4
code-generation-on-mbppcode-davinci-002 175B + MBR-Exec#31Accuracy: 63
code-generation-on-mbppcode-cushman-001 12B + MBR-Exec#55Accuracy: 48.3
code-generation-on-mbppCodeGen 16B + MBR-Exec#58Accuracy: 47.3
code-generation-on-mbppCodeGen 16B + Coder-Reviewer#62Accuracy: 46.2
code-generation-on-mbppCodeGen 16B + Reviewer#67Accuracy: 44.1
code-generation-on-mbppInCoder 6.7B + MBR-Exec#81Accuracy: 26.7
code-generation-on-mbppInCoder 6.7B + Coder-Reviewer#82Accuracy: 26.1
code-generation-on-mbppInCoder 6.7B + Reviewer#83Accuracy: 24.4