crumb's picture
Update README.md
5eb3703
|
raw
history blame
664 Bytes
metadata
license: bigscience-bloom-rail-1.0
tags:
  - stable-diffusion
  - diffusion
model-index:
  - name: bloom-560m-RLHF-SD2-prompter
    results: []
datasets:
  - Gustavosta/Stable-Diffusion-Prompts
widget:
  - text: '<s>Prompt:'
inference:
  parameters:
    eos_token_id: 2
    max_length: 128

Using RLHF (Reinforcement Learning from Human Feedback) to finetune mrm8488/bloom-560m-finetuned-sd-prompts further for SD2.0

batch_size = 8
learning_rate = 0.001

Generate extension with "<s>Prompt: " and whatever your normal prompt is.

I did this myself. I sat down and just ranked images for so long.