|
--- |
|
license: mit |
|
--- |
|
# Nunchaku |
|
Python package for Nunchaku - post-training quantization with 4-bit weights and activations. |
|
|
|
## Installation |
|
```py |
|
pip install nunchaku-0.0.2b0-cp311-cp311-linux_x86_64.whl |
|
``` |
|
## Usage |
|
```py |
|
import torch |
|
from diffusers import FluxPipeline |
|
|
|
from nunchaku.models.transformer_flux import NunchakuFluxTransformer2dModel |
|
|
|
transformer = NunchakuFluxTransformer2dModel.from_pretrained("mit-han-lab/svdq-int4-flux.1-schnell") |
|
pipeline = FluxPipeline.from_pretrained( |
|
"black-forest-labs/FLUX.1-schnell", transformer=transformer, torch_dtype=torch.bfloat16 |
|
).to("cuda") |
|
image = pipeline("A cat holding a sign that says hello world", num_inference_steps=4, guidance_scale=0).images[0] |
|
image.save("example.png") |
|
``` |
|
|
|
refer to https://github.com/mit-han-lab/nunchaku for move details, this is not an official release of the pkg |
