Abstract:Large language models (LLMs) pretrained on vast source code have achieved prominent progress in code intelligence. However, existing code LLMs have two main limitations in terms of architecture and pretraining tasks. First, they often adopt a specific architecture (encoder-only or decoder-only) or rely on a unified encoder-decoder network for different downstream tasks. The former paradigm is limited by inflexibility in applications while in the latter, the model is treated as a single system for all tasks, leading to suboptimal performance on a subset of tasks. Secondly, they often employ a limited set of pretraining objectives which might not be relevant to some downstream tasks and hence result in substantial performance degrade. To address these limitations, we propose ``CodeT5+'', a family of encoder-decoder LLMs for code in which component modules can be flexibly combined to suit a wide range of downstream code tasks. Such flexibility is enabled by our proposed mixture of pretraining objectives to mitigate the pretrain-finetune discrepancy. These objectives cover span denoising, contrastive learning, text-code matching, and causal LM pretraining tasks, on both unimodal and bimodal multilingual code corpora. Furthermore, we propose to initialize CodeT5+ with frozen off-the-shelf LLMs without training from scratch to efficiently scale up our models, and explore instruction-tuning to align with natural language instructions. We extensively evaluate CodeT5+ on over 20 code-related benchmarks in different settings, including zero-shot, finetuning, and instruction-tuning. We observe state-of-the-art (SoTA) model performance on various code-related tasks, such as code generation and completion, math programming, and text-to-code retrieval tasks. Particularly, our instruction-tuned CodeT5+ 16B achieves new SoTA results on HumanEval code generation task against other open code LLMs.

ACECode: A Reinforcement Learning Framework for Aligning Code Efficiency and Correctness in Code Language Models

Effi-Code: Unleashing Code Efficiency in Language Models

EffiLearner: Enhancing Efficiency of Generated Code via Self-Optimization

CodeACT: Code Adaptive Compute-efficient Tuning Framework for Code LLMs

Should AI Optimize Your Code? A Comparative Study of Current Large Language Models Versus Classical Optimizing Compilers

Evaluating and Aligning CodeLLMs on Human Preference

SelfCodeAlign: Self-Alignment for Code Generation

E-code: Mastering Efficient Code Generation through Pretrained Models and Expert Encoder Group

LLM-Assisted Code Cleaning For Training Accurate Code Generators

Performance-Aligned LLMs for Generating Fast Code

How Efficient is LLM-Generated Code? A Rigorous & High-Standard Benchmark

aiXcoder-7B: A Lightweight and Effective Large Language Model for Code Completion

AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data

Aligning CodeLLMs with Direct Preference Optimization

CodeLutra: Boosting LLM Code Generation via Preference-Guided Refinement

CodeT5+: Open Code Large Language Models for Code Understanding and Generation

Measuring Code Efficiency Optimization Capabilities with ACEOB

Exploring the Capabilities of LLMs for Code Change Related Tasks

DolphCoder: Echo-Locating Code Large Language Models with Diverse and Multi-Objective Instruction Tuning