---
library_name: diffusers
license: other
license_name: flux-1-dev-non-commercial-license
license_link: LICENSE.md
---

## To run

> [!TIP]
> Check out [sayakpaul/flux.1-dev-nf4-pkg](https://huggingface.co/sayakpaul/flux.1-dev-nf4-pkg) that shows how to run this checkpoint along with an NF4 T5 in a free-tier Colab Notebook.

Be mindful of the license of Flux.1-Dev here. 

Make sure you have the latest versions of `bitsandbytes` and `accelerate` installed. 

And then install `diffusers` from [this PR](https://github.com/huggingface/diffusers/pull/9213/):

```bash
pip install git+https://github.com/huggingface/diffusers@c795c82df39620e2576ccda765b6e67e849c36e7
```

```python
import torch
from diffusers import FluxTransformer2DModel, FluxPipeline

model_id = "black-forest-labs/FLUX.1-dev"
nf4_id = "sayakpaul/flux.1-dev-nf4-with-bnb-integration"
model_nf4 = FluxTransformer2DModel.from_pretrained(nf4_id, torch_dtype=torch.bfloat16)
print(model_nf4.dtype)
print(model_nf4.config.quantization_config)

pipe = FluxPipeline.from_pretrained(model_id, transformer=model_nf4, torch_dtype=torch.bfloat16)
pipe.enable_model_cpu_offload()

prompt = "A mystic cat with a sign that says hello world!"
image = pipe(prompt, guidance_scale=3.5, num_inference_steps=50, generator=torch.manual_seed(0)).images[0]
image.save("flux-nf4-dev-loaded.png")
```

![image/png](https://cdn-uploads.huggingface.co/production/uploads/5f7fbd813e94f16a85448745/lBpug2CXhXU5_GEgjl-cJ.png)