mihirpd/alignprop-trl-aesthetics

AlignProp Aesthetics TRL

Prompt

Squirrel

Prompt

Lion

Model description

This is a text to image model finetuned on top of Stable Diffusion v1.5 using aesthetic reward function on animals dataset, via reward backpropagation as per AlignProp (https://align-prop.github.io/). The model was trained using the code in TRL repository.

Trigger words

You should use alignprop to trigger the image generation.

You should use reward backpropagation to trigger the image generation.

You should use rlhf to trigger the image generation.

You should use diffusion to trigger the image generation.

You should use reward gradients to trigger the image generation.

Download model

Weights for this model are available in Safetensors,PyTorch format.

Download them in the Files & versions tab.