This PR adds a Dockerfile for Image Generation.
It is a draft because it doesn't read the prompts from a file (to be done later today hopefully) but it does successfully use unified-io-inference to generate images from prompts and could be useful in exploring the bias of the model.
This PR adds a Dockerfile for Image Generation.
It is a draft because it doesn't read the prompts from a file (to be done later today hopefully) but it does successfully use unified-io-inference to generate images from prompts and could be useful in exploring the bias of the model.