CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning

TL;DR

CodeRL integrates pretrained language models with deep reinforcement learning, using a critic network and feedback from unit tests to significantly improve program synthesis performance, especially on complex unseen tasks.

Contribution

The paper introduces CodeRL, a novel framework combining pretrained models and reinforcement learning with a critic network for improved code generation.

Findings

Achieves new SOTA on APPS benchmark.

Demonstrates strong zero-shot transfer on MBPP.

Enhances code generation with critic feedback and critical sampling.

Abstract

Program synthesis or code generation aims to generate a program that satisfies a problem specification. Recent approaches using large-scale pretrained language models (LMs) have shown promising results, yet they have some critical limitations. In particular, they often follow a standard supervised fine-tuning procedure to train a code generation model only from the pairs of natural-language problem descriptions and ground-truth programs. Such paradigm largely ignores some important but potentially useful signals in the problem specification such as unit tests, which thus often results in poor performance when solving complex unseen coding tasks. To address the limitations, we propose "CodeRL", a new framework for program synthesis tasks through pretrained LMs and deep reinforcement learning (RL). Specifically, during training, we treat the code-generating LM as an actor network, and introduce a critic network that is trained to predict the functional correctness of generated programs and provide dense feedback signals to the actor. During inference, we introduce a new generation procedure with a critical sampling strategy that allows a model to automatically regenerate programs based on feedback from example unit tests and critic scores. For the model backbones, we extended the encoder-decoder architecture of CodeT5 with enhanced learning objectives, larger model sizes, and better pretraining data. Our method not only achieves new SOTA results on the challenging APPS benchmark, but also shows strong zero-shot transfer capability with new SOTA results on the simpler MBPP benchmark.