OpenCodeInterpreter: Integrating Code Generation with Execution and Refinement

The introduction of large language models has significantly advanced codegeneration. However, open-source models often lack the execution capabilitiesand iterative refinement of advanced systems like the GPT-4 Code Interpreter.To address this, we introduce OpenCodeInterpreter, a family of open-source codesystems designed for generating, executing, and iteratively refining code.Supported by Code-Feedback, a dataset featuring 68K multi-turn interactions,OpenCodeInterpreter integrates execution and human feedback for dynamic coderefinement. Our comprehensive evaluation of OpenCodeInterpreter across keybenchmarks such as HumanEval, MBPP, and their enhanced versions from EvalPlusreveals its exceptional performance. Notably, OpenCodeInterpreter-33B achievesan accuracy of 83.2 (76.4) on the average (and plus versions) of HumanEval andMBPP, closely rivaling GPT-4's 84.2 (76.2) and further elevates to 91.6 (84.6)with synthesized human feedback from GPT-4. OpenCodeInterpreter brings the gapbetween open-source code generation models and proprietary systems like GPT-4Code Interpreter.