Image-to-Image
Diffusers
Diffusion Single File
English
fp8
unsloth
quantized
image-generation
image-editing
flux
Instructions to use unsloth/FLUX.2-dev-FP8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use unsloth/FLUX.2-dev-FP8 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("unsloth/FLUX.2-dev-FP8", dtype=torch.bfloat16, device_map="cuda") prompt = "cute sloth typing on a computer (INT8)" input_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cat.png") image = pipe(image=input_image, prompt=prompt).images[0] - Diffusion Single File
How to use unsloth/FLUX.2-dev-FP8 with Diffusion Single File:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Unsloth Desktop
Rename checkpoints to model-name files
Browse files
transformer_fp8.pt → FLUX.2-dev-FP8.pt
RENAMED
|
File without changes
|
transformer_int8.pt → FLUX.2-dev-INT8.pt
RENAMED
|
File without changes
|
README.md
CHANGED
|
@@ -113,8 +113,8 @@ dynamic activation quantization from the dense bf16 transformer. The official mo
|
|
| 113 |
is unchanged from the source repo.
|
| 114 |
|
| 115 |
Files:
|
| 116 |
-
-
|
| 117 |
-
-
|
| 118 |
|
| 119 |
Details:
|
| 120 |
- int8: Int8DynamicActivationInt8WeightConfig (per-token activation, per-channel weight, torch._int_mm).
|
|
|
|
| 113 |
is unchanged from the source repo.
|
| 114 |
|
| 115 |
Files:
|
| 116 |
+
- FLUX.2-dev-INT8.pt (33.0 GB)
|
| 117 |
+
- FLUX.2-dev-FP8.pt (32.2 GB)
|
| 118 |
|
| 119 |
Details:
|
| 120 |
- int8: Int8DynamicActivationInt8WeightConfig (per-token activation, per-channel weight, torch._int_mm).
|