AI Gallery

AI Gallery is a collection of ready-to-run recipes for popular AI frameworks and AI models. It provides a simple way to package, share, and run AI projects using the simple interface of SkyPilot.

You can directly run these recipes on your own infrastructure, such as cloud VMs or Kubernetes.

Contents

Inference Frameworks

• vLLM
• Hugging Face TGI
• Ollama
• SGLang
• LoRAX

LLM Models

• Vision Llama 3.2 (Meta)
• Llama 3.1 (Meta)
• Llama 3 (Meta)
• Llama 2 (Meta)
• CodeLlama (Meta)
• Pixtral (Mistral AI)
• Mixtral (Mistral AI)
• Mistral 7B (Mistral AI)
• Qwen 2.5 (Alibaba)
• Gemma (Google)
• DBRX (Databricks)

Applications

• Tabby: Coding Assistant
• LocalGPT: Chat with PDF

Tutorials

• Train Your Own Vicuna on Llama-2

Adding an Example to the Gallery

We welcome contributions from the community. If you would like to contribute, please follow the guidelines below.

1. Fork the SkyPilot repository on GitHub.

2. Create a new folder for your own framework, LLM model, or tutorial under llm/.

3. Add a README.md, a SkyPilot YAML, and other necessary files to run the AI project.

4. Create a soft link to your README in docs/source/_gallery_original to the README file in one of the subfolders (frameworks, llms, tutorials), e.g., cd docs/source/_gallery_original/llms; ln -s ../../../../llm/mixtral/README.md mixtral.md.

5. Add the file path to the toctree above on this page.

6. Create a pull request to the SkyPilot repository.

If you have any questions, please feel free to ask in the SkyPilot Slack.