hoan17/D500S100_github

TRL O2O Model

This is a diffusion model that has been fine-tuned with reinforcement learning to guide the model outputs according to a value, function, or human feedback. The model can be used for image generation conditioned with text.