Created by: salanki
Patch Description Adds a Dockerfile to the project repository, based on the setup instructions to make it possible to run on any nvidia-docker enabled host and in Kubernetes.
Can be run locally with an interactive shell:
docker run --rm -it --gpus 1 coreweave/ml-images:metaseq-6cd6a708 bash
Testing steps Build and run tested. Not yet tested for large inference or training.