Welcome to dstack-examples. This repository contains a collection of examples
demonstrating how to use dstack.
pip install "dstack[all]" -U
dstack startThe dstack start command starts the Hub server, creating the default project that runs dev environments, pipelines,
and apps locally.
To run examples in the cloud (AWS, GCP, Azure, Lambda Cloud), make sure to create the corresponding project via the UI.
git clone https://github.com/dstackai/dstack-examples
cd dstack-examples
dstack init| Example | How to run |
|---|---|
| Default dev environment | dstack run . --gpu 16GB |
| Deploying LLMs using vLLM | dstack run . -f vllm/serve.dstack.yml --gpu 24GB |
| Deploying LLMs using TGI | dstack run . -f text-generation-inference/serve.dstack.yml --gpu 24GB |
| Custom dev environment for TGI | dstack run . -f text-generation-inference/.dstack.yml --build --gpu 24GB |
| Deploying SDXL using FastAPI | dstack run . -f stable-diffusion-xl/api.dstack.yml --gpu 16GB |
| Fine-tuning Llama 2 using QLoRA | dstack run . -f llama-2/train.dstack.yml --gpu 16GB |
| Deploying LLMs using Python API | streamlit run deploy-python/app.py |