This repo is for using the Superflows fine-tuned LLM on https://runpod.io.
You want to download the model weights from hugging face and put all model-0000X-of-0000Y.safetensors files in the /zephyr directory.
The adapter_model.safetensors file should live in the /superflows directory.
These files aren't included in this repo because they are too large.
You need to:
- make a RunPod account
- build the docker container using the Dockerfile in this repo
- push the docker container to a container registry
- deploy the container to RunPod
Reach out to [email protected] if you have any questions or need help with this process.