File size: 1,349 Bytes
c6ef664 829f49a c6ef664 829f49a 1fc910e 829f49a 1fc910e 829f49a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 |
---
license: cc0-1.0
language:
- en
---
# Generalis V1.1
<hr>
### Second attempt at merging several models v1.5 into one general purpose model.
Focus has been put into simple prompts, good one-off generation, slightly muted colours, low memory usage, small model size.
It is intended as easy model for use in larger projects where image generation is needed.
Published under CC0
<hr>
Use example:
```python
import torch # Tested with 2.0.1+cu118
from diffusers import StableDiffusionPipeline # <3
# Model location in HF
model = "https://huggingface.co/vluz/Generalis_V1.1/blob/main/Generalis_v1-1.safetensors"
# Create pipe
pipe = StableDiffusionPipeline.from_ckpt(model,
torch_dtype=torch.float16,
safety_checker=None,
feature_extractor=None,
requires_safety_checker=False,)
# Cleanup
del pipe.vae.encoder
torch.cuda.empty_cache()
# Send to GPU
pipe = pipe.to("cuda")
# Optimize for low vram use and clear cache again
pipe.enable_vae_tiling()
pipe.enable_attention_slicing("max")
pipe.enable_xformers_memory_efficient_attention(attention_op=None)
pipe.unet.to(memory_format=torch.channels_last)
pipe.enable_sequential_cpu_offload()
torch.cuda.empty_cache()
# Set a prompt
prompt = "a cat"
# Generate image based on prompt
image = pipe(prompt).images[0]
# Save result image to disk
image.save("cat.png")
``` |