README.md

---
license: gemma
pipeline_tag: text-generation
tags:
- cortex.cpp
---

## Overview

The [Gemma](https://huggingface.co/google/gemma-2-2b-it), state-of-the-art open model trained with the Gemma datasets that includes both synthetic data and the filtered publicly available websites data with a focus on high-quality and reasoning dense properties. The model belongs to the Gemma family with the 4B, 7B version in two variants 8K and 128K which is the context length (in tokens) that it can support.

## Variants

| No | Variant | Cortex CLI command |
| --- | --- | --- |
| 1 | [Gemma2-2b](https://huggingface.co/cortexso/gemma2/tree/2b) | `cortex run gemma2:2b` |
| 2 | [Gemma2-9b](https://huggingface.co/cortexso/gemma2/tree/9b) | `cortex run gemma2:9b` |
| 3 | [Gemma2-27b](https://huggingface.co/cortexso/gemma2/tree/27b) | `cortex run gemma2:27b` |

## Use it with Jan (UI)

1. Install **Jan** using [Quickstart](https://jan.ai/docs/quickstart)
2. Use in Jan model Hub:
    ```bash
    cortexso/gemma2
    ```
    
## Use it with Cortex (CLI)

1. Install **Cortex** using [Quickstart](https://cortex.jan.ai/docs/quickstart)
2. Run the model with command:
    ```bash
    cortex run gemma2
    ```
    
## Credits

- **Author:** Go‌ogle
- **Converter:** [Homebrew](https://www.homebrew.ltd/)
- **Original License:** [License](https://ai.google.dev/gemma/terms)
- **Papers:** [Gemma Technical Report](https://arxiv.org/abs/2403.08295)