Update README.md
Browse files
README.md
CHANGED
@@ -7,13 +7,16 @@ base_model:
|
|
7 |
- stabilityai/stable-diffusion-3.5-large-turbo
|
8 |
base_model_relation: quantized
|
9 |
---
|
|
|
10 |
These models are made to work with [stable-diffusion.cpp](https://github.com/leejet/stable-diffusion.cpp) release [master-ac54e00](https://github.com/leejet/stable-diffusion.cpp/releases/tag/master-ac54e00) onwards. Support for other inference backends is not guarenteed.
|
11 |
|
12 |
Quantized using this PR https://github.com/leejet/stable-diffusion.cpp/pull/447
|
13 |
|
14 |
-
|
15 |
|
16 |
-
|
|
|
|
|
17 |
|
18 |
- [sd3.5_large_turbo-q2_k_4_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q2_k_4_0.gguf): Smallest quantization yet. Use this if you can't afford anything bigger
|
19 |
- [sd3.5_large_turbo-q3_k_4_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q3_k_4_0.gguf): Smaller than q4_0, acceptable degradation.
|
@@ -21,12 +24,12 @@ Quantized using this PR https://github.com/leejet/stable-diffusion.cpp/pull/447
|
|
21 |
- [sd3.5_large_turbo-q4_k_4_1.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q4_k_4_1.gguf): Smaller than q4_1, and with comparable degradation. Recommended
|
22 |
- [sd3.5_large_turbo-q4_k_5_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q4_k_5_0.gguf): Smaller than q5_0, and with comparable degradation. Recommended
|
23 |
|
24 |
-
|
25 |
|
26 |
- [sd3.5_large_turbo-q4_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/legacy/sd3.5_large_turbo-q4_0.gguf): Same size as q4_k_4_0, Not recommended (use q4_k_4_0 instead)
|
27 |
-
(I wanted to upload more, but it's not working anymore, maybe i hit a rate limit)
|
28 |
|
29 |
-
|
30 |
|
31 |
| Name | Image | Image | Image |
|
32 |
| ------------------ | -------------------------------- | ---------------------------------- | ---------------------------------- |
|
|
|
7 |
- stabilityai/stable-diffusion-3.5-large-turbo
|
8 |
base_model_relation: quantized
|
9 |
---
|
10 |
+
## Overview
|
11 |
These models are made to work with [stable-diffusion.cpp](https://github.com/leejet/stable-diffusion.cpp) release [master-ac54e00](https://github.com/leejet/stable-diffusion.cpp/releases/tag/master-ac54e00) onwards. Support for other inference backends is not guarenteed.
|
12 |
|
13 |
Quantized using this PR https://github.com/leejet/stable-diffusion.cpp/pull/447
|
14 |
|
15 |
+
Normal K-quants are not working properly with SD3.5-Large models because over 90% of the weights are in tensors whose shape doesn't match the 256 superblock size of K-quants and can't be quantized. Mixing quantization types allow to take adventage of the better fidelity of k-quantto some extent while keeping the file size small.
|
16 |
|
17 |
+
## Files:
|
18 |
+
|
19 |
+
### Mixed Types:
|
20 |
|
21 |
- [sd3.5_large_turbo-q2_k_4_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q2_k_4_0.gguf): Smallest quantization yet. Use this if you can't afford anything bigger
|
22 |
- [sd3.5_large_turbo-q3_k_4_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q3_k_4_0.gguf): Smaller than q4_0, acceptable degradation.
|
|
|
24 |
- [sd3.5_large_turbo-q4_k_4_1.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q4_k_4_1.gguf): Smaller than q4_1, and with comparable degradation. Recommended
|
25 |
- [sd3.5_large_turbo-q4_k_5_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/sd3.5_large_turbo-q4_k_5_0.gguf): Smaller than q5_0, and with comparable degradation. Recommended
|
26 |
|
27 |
+
### Legacy types:
|
28 |
|
29 |
- [sd3.5_large_turbo-q4_0.gguf](https://huggingface.co/stduhpf/SD3.5-Large-Turbo-GGUF-mixed-sdcpp/blob/main/legacy/sd3.5_large_turbo-q4_0.gguf): Same size as q4_k_4_0, Not recommended (use q4_k_4_0 instead)
|
30 |
+
- (I wanted to upload more, but it's not working anymore, maybe i hit a rate limit)
|
31 |
|
32 |
+
## Outputs:
|
33 |
|
34 |
| Name | Image | Image | Image |
|
35 |
| ------------------ | -------------------------------- | ---------------------------------- | ---------------------------------- |
|