Skip to content

Commit 28a6147

Browse files
leejetleejet
committed
docs: update docs/flux.md
1 parent c837c5d commit 28a6147

File tree

2 files changed

+9
-6
lines changed

2 files changed

+9
-6
lines changed

‎assets/flux/flux1-dev-q4_k.png

468 KB
Loading

‎docs/flux.md

Lines changed: 9 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -4,14 +4,17 @@ You can run Flux using stable-diffusion.cpp with a GPU that has 6GB or even 4GB
44

55
## Download weights
66

7-
- Download flux-dev from https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/flux1-dev.safetensors
8-
- Download flux-schnell from https://huggingface.co/black-forest-labs/FLUX.1-schnell/blob/main/flux1-schnell.safetensors
7+
- Download flux
8+
- If you don't want to do the conversion yourself, download the preconverted gguf model from [FLUX.1-dev-gguf](https://huggingface.co/leejet/FLUX.1-dev-gguf) or [FLUX.1-schnell](https://huggingface.co/leejet/FLUX.1-schnell-gguf)
9+
- Otherwise, download flux-dev from https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/flux1-dev.safetensors or flux-schnell from https://huggingface.co/black-forest-labs/FLUX.1-schnell/blob/main/flux1-schnell.safetensors
910
- Download vae from https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/ae.safetensors
1011
- Download clip_l from https://huggingface.co/comfyanonymous/flux_text_encoders/blob/main/clip_l.safetensors
1112
- Download t5xxl from https://huggingface.co/comfyanonymous/flux_text_encoders/blob/main/t5xxl_fp16.safetensors
1213

1314
## Convert flux weights
1415

16+
You can download the preconverted gguf weights from [FLUX.1-dev-gguf](https://huggingface.co/leejet/FLUX.1-dev-gguf) or [FLUX.1-schnell](https://huggingface.co/leejet/FLUX.1-schnell-gguf), this way you don't have to do the conversion yourself.
17+
1518
Using fp16 will lead to overflow, but ggml's support for bf16 is not yet fully developed. Therefore, we need to convert flux to gguf format here, which also saves VRAM. For example:
1619
```
1720
.\bin\Release\sd.exe -M convert -m ..\..\ComfyUI\models\unet\flux1-dev.sft -o ..\models\flux1-dev-q8_0.gguf -v --type q8_0
@@ -30,10 +33,10 @@ For example:
3033

3134
Using formats of different precisions will yield results of varying quality.
3235

33-
| Type | q8_0 | q4_0 | q3_k | q2_k |
34-
|---- | ---- |---- |---- |---- |
35-
| **Memory** | 12068.09 MB | 6394.53 MB | 4888.16 MB | 3735.73 MB |
36-
| **Result** | ![](../assets/flux/flux1-dev-q8_0.png) |![](../assets/flux/flux1-dev-q4_0.png) |![](../assets/flux/flux1-dev-q3_k.png) |![](../assets/flux/flux1-dev-q2_k.png)|
36+
| Type | q8_0 | q4_0 | q4_k | q3_k | q2_k |
37+
|---- | ---- |---- |---- |---- |---- |
38+
| **Memory** | 12068.09 MB | 6394.53 MB | 6395.17 MB | 4888.16 MB | 3735.73 MB |
39+
| **Result** | ![](../assets/flux/flux1-dev-q8_0.png) |![](../assets/flux/flux1-dev-q4_0.png) |![](../assets/flux/flux1-dev-q4_k.png) |![](../assets/flux/flux1-dev-q3_k.png) |![](../assets/flux/flux1-dev-q2_k.png)|
3740

3841

3942

0 commit comments

Comments
 (0)