This pipeline was finetuned from OFA-Sys/small-stable-diffusion-v0 on the MusicBrainz and Cover Art Archive datasets. Below are some example images generated with the finetuned pipeline using the following prompts: ['Cover art for a disco album titled "My Love", by "Meux Amis"']:

Pipeline usage

You can use the pipeline like so:

python

from diffusers import DiffusionPipeline
import torch

pipeline = DiffusionPipeline.from_pretrained("jpholanda/SD-coverart-v1", torch_dtype=torch.float16)
prompt = 'Cover art for a disco album titled "My Love", by "Meux Amis"'
image = pipeline(prompt).images[0]
image.save("my_image.png")

Training info

These are the key hyperparameters used during training:

Epochs: 5
Learning rate: 1e-05
Batch size: 32
Gradient accumulation steps: 4
Image resolution: 250
Mixed-precision: fp16

More information on all the CLI arguments and the environment are available on your wandb run page.

Training details

Used the MusicBrainz dataset for the metadata (title, genre, artist) and the Cover Art Archive for the cover arts.