The rise of large language models (LLMs) has significantly enhanced various coding tasks, such as code completion, and tools like GitHub Copilot. However, the substantial size of these LLMs poses a significant challenge when it comes to fine-tuning them for specific downstream tasks. As an alternative approach, retrieval-based methods have emerged as a promising solution, augmenting model predictions without the need for fine-tuning. Despite their potential, a significant challenge is that the designs of these methods often rely on heuristics, leaving critical questions about what information should be stored or retrieved and how to interpolate such information for augmenting predictions. To tackle this challenge, we first perform a theoretical analysis of the fine-tuning process, highlighting the importance of Δlogits as a catalyst for improving model predictions. Building on this insight, we develop a novel retrieval-based method, FT2Ra, which aims to mimic genuine fine-tuning. While FT2Ra adopts a retrieval-based mechanism, it uniquely adopts a paradigm with a learning rate and multi-epoch retrievals, which is similar to fine-tuning. We conducted a comprehensive evaluation of FT2Ra in both token-level and line-level code completions. Our findings demonstrate the remarkable effectiveness of FT2Ra when compared to state-of-the-art methods and its potential to genuine fine-tuning. In token-level completion, which represents a relatively easier task, FT2Ra achieves a 4.29% improvement in accuracy compared to the best baseline method on UniXcoder. In the more challenging line-level completion task, we observe a substantial ∼2×+  increase in Exact Match (EM) performance, indicating the significant advantages of our theoretical analysis. Notably, even when operating without actual fine-tuning, FT2Ra exhibits competitive performance, aligning closely with the results achieved by fine-tuned UniXcoder model over 7 and 10 epochs in token-level and line-level completion, respectively.

1. Supplementary Experimental Data

2. Code and Dataset