Update README.md
Browse files
README.md
CHANGED
|
@@ -11,6 +11,9 @@ library_name: diffusers
|
|
| 11 |
tags:
|
| 12 |
- diffusion-single-file
|
| 13 |
---
|
|
|
|
|
|
|
|
|
|
| 14 |
|
| 15 |
```python
|
| 16 |
import torch
|
|
@@ -57,4 +60,40 @@ image = pipe(
|
|
| 57 |
enable_prompt_rewrite=True
|
| 58 |
).images[0]
|
| 59 |
image.save('image longcat-image.png')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
```
|
|
|
|
| 11 |
tags:
|
| 12 |
- diffusion-single-file
|
| 13 |
---
|
| 14 |
+
### How to Use
|
| 15 |
+
|
| 16 |
+
#### `diffusers`
|
| 17 |
|
| 18 |
```python
|
| 19 |
import torch
|
|
|
|
| 60 |
enable_prompt_rewrite=True
|
| 61 |
).images[0]
|
| 62 |
image.save('image longcat-image.png')
|
| 63 |
+
```
|
| 64 |
+
|
| 65 |
+
#### ComfyUI
|
| 66 |
+
Currently, this model is not supported natively in ComfyUI. Do let me know if it receives native support, and I will get to supporting it.
|
| 67 |
+
|
| 68 |
+
### Compression details
|
| 69 |
+
|
| 70 |
+
This is the `pattern_dict` for compression:
|
| 71 |
+
|
| 72 |
+
```python
|
| 73 |
+
pattern_dict = {
|
| 74 |
+
r"transformer_blocks\.\d+": (
|
| 75 |
+
"norm1.linear",
|
| 76 |
+
"norm1_context.linear",
|
| 77 |
+
"attn.to_q",
|
| 78 |
+
"attn.to_k",
|
| 79 |
+
"attn.to_v",
|
| 80 |
+
"attn.to_out.0",
|
| 81 |
+
"attn.add_q_proj",
|
| 82 |
+
"attn.add_k_proj",
|
| 83 |
+
"attn.add_v_proj",
|
| 84 |
+
"attn.to_add_out",
|
| 85 |
+
"ff.net.0.proj",
|
| 86 |
+
"ff.net.2",
|
| 87 |
+
"ff_context.net.0.proj",
|
| 88 |
+
"ff_context.net.2",
|
| 89 |
+
),
|
| 90 |
+
r"single_transformer_blocks\.\d+": (
|
| 91 |
+
"norm.linear",
|
| 92 |
+
"proj_mlp",
|
| 93 |
+
"proj_out",
|
| 94 |
+
"attn.to_q",
|
| 95 |
+
"attn.to_k",
|
| 96 |
+
"attn.to_v",
|
| 97 |
+
),
|
| 98 |
+
}
|
| 99 |
```
|