Instructions to use crumb/eva-fusion-v2.22 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use crumb/eva-fusion-v2.22 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("crumb/eva-fusion-v2.22", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
EVA-Fusion 2.22 - You are (Not) HD
for an art project of mine
| model | init model | lr | bs | acc | total steps | training samples | training resolution | models trained |
|---|---|---|---|---|---|---|---|---|
| eva-fusion-v2 | compvis 1-4 | 2.057e-06 | 1 | 1 | 8,000 | 1,143 (captioned just "picture in the style of evangelion") | 512x384 (4:3) | unet |
| eva-fusion-v2.1 | eva-fusion-v2 | 1.313e-06 | 1 | 1 | 10,000 | + 560 (hand captioned) | 512x384 (4:3) | unet |
| eva-fusion-v2.22 "you are (not) HD" | previous models + compvis1-4 + wd1.3 at various ratios | x | x | x | x | x | x | x |
for generating let me suggest Whatchamacallit, suggested in the negative prompt: "blurry jpeg, multiple views, simple background, black and white reference sheet".
it wasn't trained on black and white character sheets, it was barely trained on any grayscale images, I do not know why it leans heavily towards them sometimes.
link to ckpt file: https://huggingface.co/crumb/eva-model-ckpt/blob/main/test%20model%205.ckpt
I also recommend "in the style of Evangelion" in the prompt
- Downloads last month
- 5