ffjefckds commited on
Commit
60da9ff
·
verified ·
1 Parent(s): 416593f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +0 -29
README.md CHANGED
@@ -14,35 +14,6 @@ configs:
14
  data_files:
15
  - split: train
16
  path: data/train-*
17
- dataset_name: "sd-class-beer-32"
18
- dataset_link: "https://huggingface.co/datasets/ffjefckds/sd-class-beer-32"
19
- size: "25 images"
20
- resolution: "1024x1024 px"
21
- creator: "Crayon (DALL·E Mini)"
22
- task: "Unconditional image generation"
23
- license: "MIT"
24
- intended_use: "This dataset is intended for training diffusion models on beer-related images."
25
- dataset_description: |
26
- This synthetic dataset contains 25 images of beer in 1024x1024 px resolution, created using Crayon (DALL·E Mini).
27
- The images depict various beer-related objects and scenes, designed for use in training diffusion models for unconditional image generation.
28
- features:
29
- - Images: The dataset consists of 25 images generated in the style of Crayon (DALL·E Mini). Each image is 1024x1024 px and portrays scenes related to beer.
30
-
31
- usage_instructions: |
32
- To use the dataset, you can load the images using standard image-processing libraries like PIL or OpenCV.
33
- For model training, this dataset can be used with diffusion models like DDPMPipeline from Hugging Face's diffusers library.
34
-
35
- Example usage:
36
- ```python
37
- from diffusers import DDPMPipeline
38
-
39
- pipeline = DDPMPipeline.from_pretrained('{hub_model_id}')
40
- image = pipeline().images[0]
41
- image
42
- ```
43
- additional_info: |
44
- The dataset is focused on the class of beer and can be used for training or fine-tuning image generation models.
45
- Since it is synthetic, the data may not represent real-world beer in diverse settings but is intended for the development of general beer-related imagery generation.
46
  ---
47
 
48
 
 
14
  data_files:
15
  - split: train
16
  path: data/train-*
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
17
  ---
18
 
19