CodeGemma

Model Page : CodeGemma

Resources and Technical Documentation : Technical Report : Responsible Generative AI Toolkit

Terms of Use : Terms

Authors : Google

In llama.cpp, and other related tools such as Ollama and LM Studio, please make sure that you have these flags set correctly, especially repeat-penalty. Georgi Gerganov (llama.cpp's author) shared his experience in https://huggingface.co/google/gemma-7b-it/discussions/38#65d7b14adb51f7c160769fa1.

Description

CodeGemma is a collection of lightweight open code models built on top of Gemma. CodeGemma models are text-to-text and text-to-code decoder-only models and are available as a 7 billion pretrained variant that specializes in code completion and code generation tasks, a 7 billion parameter instruction-tuned variant for code chat and instruction following and a 2 billion parameter pretrained variant for fast code completion.

For detailed model card, refer to https://huggingface.co/google/codegemma-1.1-7b-it.

Sample Usage

$ cat non_prime
/// Write a rust function to identify non-prime numbers.
///
/// Examples:
/// >>> is_not_prime(2)
/// False
/// >>> is_not_prime(10)
/// True
pub fn is_not_prime(n: i32) -> bool {
$ main -m codegemma-1.1-7b-it.gguf --temp 0 --top-k 0 -f non_prime --log-disable --repeat-penalty 1.0
 /// Write a rust function to identify non-prime numbers.
///
/// Examples:
/// >>> is_not_prime(2)
/// False
/// >>> is_not_prime(10)
/// True
pub fn is_not_prime(n: i32) -> bool {
    if n <= 1 {
        return true;
    }
    for i in 2..=(n as f64).sqrt() as i32 {
        if n % i == 0 {
            return true;
        }
    }
    false
}

Coding Benchmarks

Natural Language Benchmarks