# Code Auto-Completion

aiXcoder-7B is a large language model focused on code generation and understanding, trained on 1.2T unique tokens, excelling in code completion and multi-language code generation tasks.

CodeGen2.5 is a series of autoregressive language models for program synthesis, improved upon CodeGen2 and trained on StarCoderData, achieving high performance at a smaller scale.

CodeGen2.5 is a series of autoregressive language models for program synthesis, improved upon CodeGen2 and trained on StarCoderData, achieving performance comparable to larger models at a smaller scale.