Can I run HyperCLOVA X-omni on RunPod without using Docker?
#5
by
BenHyun - opened
Since RunPod does not support Docker Compose, I cannot build images using it. Is it possible to run this model without the Docker Compose process?
My goal is to use Speech-to-Text, so I'm planning to run only the audio encoder, chainer, and vLLM. Is this the right approach? Or would I need to run the entire stack, including the encoder, decoder, and vLLM?