CHAITron commited on
Commit
602ad3a
1 Parent(s): e536b60

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +26 -2
README.md CHANGED
@@ -17,18 +17,42 @@ please have a look at the [Stable Diffusion](https://huggingface.co/docs/diffuse
17
 
18
  You can also export the model to [ONNX](https://huggingface.co/docs/diffusers/optimization/onnx), [MPS](https://huggingface.co/docs/diffusers/optimization/mps) and/or [FLAX/JAX]().
19
 
 
 
 
 
20
  ```python
21
  from diffusers import StableDiffusionPipeline
22
  import torch
 
 
 
 
 
 
 
 
 
 
 
 
23
 
24
  model_id = "CHAITron/KranokThaiV1"
25
  pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
26
  pipe = pipe.to("cuda")
27
 
28
  prompt = "a colourful masterpiece thai dog in <knt-style> style, high-quality, professional art, realistic, thai style"
29
- image = pipe(prompt).images[0]
30
 
31
- image.save("./result.png")
 
 
 
 
 
 
 
 
 
32
  ```
33
 
34
  ## License
 
17
 
18
  You can also export the model to [ONNX](https://huggingface.co/docs/diffusers/optimization/onnx), [MPS](https://huggingface.co/docs/diffusers/optimization/mps) and/or [FLAX/JAX]().
19
 
20
+ ```shell
21
+ pip install torch diffusers accelerate pillow
22
+ ```
23
+
24
  ```python
25
  from diffusers import StableDiffusionPipeline
26
  import torch
27
+ from PIL import Image
28
+
29
+ def image_grid(imgs, rows, cols):
30
+ assert len(imgs) == rows*cols
31
+
32
+ w, h = imgs[0].size
33
+ grid = Image.new('RGB', size=(cols*w, rows*h))
34
+ grid_w, grid_h = grid.size
35
+
36
+ for i, img in enumerate(imgs):
37
+ grid.paste(img, box=(i%cols*w, i//cols*h))
38
+ return grid
39
 
40
  model_id = "CHAITron/KranokThaiV1"
41
  pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
42
  pipe = pipe.to("cuda")
43
 
44
  prompt = "a colourful masterpiece thai dog in <knt-style> style, high-quality, professional art, realistic, thai style"
 
45
 
46
+ num_samples = 4
47
+ num_rows = 1
48
+
49
+ all_images = []
50
+ for _ in range(num_rows):
51
+ images = pipe(prompt, num_images_per_prompt=num_samples, num_inference_steps=25, guidance_scale=9).images
52
+ all_images.extend(images)
53
+
54
+ grid = image_grid(all_images, num_rows, num_samples)
55
+ grid
56
  ```
57
 
58
  ## License