Working code with full server requirements

#24
by gmjolt - opened

Is there someone who can provide a working code with server instance and gpu requirements needed to run the model smoothly on aws/azure/google?
Thanks

See here: https://huggingface.co/tiiuae/falcon-40b/discussions/18

Contains code & requirements, and runs on any A100 80G instance.
(I'm personally using Datacrunch.io spot instances, but any A100 80G instance should do)

Technology Innovation Institute org

You could have a look to this blogpost from HuggingFace.

Specific to AWS, there is also a tutorial for deploying on SageMaker.

Sign up or log in to comment