CodeGemma: Open Code Models Based on Gemma
It provides specialized open models for code generation tasks, addressing needs in software development and AI applications, but is incremental as it builds on existing Gemma models.
The paper introduces CodeGemma, a collection of open code models based on Gemma, with variants including 7B models for resilient natural language understanding and mathematical reasoning, and a 2B model for state-of-the-art code completion in latency-sensitive settings.
This paper introduces CodeGemma, a collection of specialized open code models built on top of Gemma, capable of a variety of code and natural language generation tasks. We release three model variants. CodeGemma 7B pretrained (PT) and instruction-tuned (IT) variants have remarkably resilient natural language understanding, excel in mathematical reasoning, and match code capabilities of other open models. CodeGemma 2B is a state-of-the-art code completion model designed for fast code infilling and open-ended generation in latency-sensitive settings.