---
dataset_info:
  features:
  - name: Unique_ID
    dtype: string
  - name: shape_0_row_0
    dtype: string
  - name: color_0_row_0
    dtype: string
  - name: shape_1_row_0
    dtype: string
  - name: color_1_row_0
    dtype: string
  - name: shape_2_row_0
    dtype: string
  - name: color_2_row_0
    dtype: string
  - name: shape_3_row_0
    dtype: string
  - name: color_3_row_0
    dtype: string
  - name: shape_0_row_1
    dtype: string
  - name: color_0_row_1
    dtype: string
  - name: shape_1_row_1
    dtype: string
  - name: color_1_row_1
    dtype: string
  - name: shape_2_row_1
    dtype: string
  - name: color_2_row_1
    dtype: string
  - name: shape_3_row_1
    dtype: string
  - name: color_3_row_1
    dtype: string
  - name: shape_0_row_2
    dtype: string
  - name: color_0_row_2
    dtype: string
  - name: shape_1_row_2
    dtype: string
  - name: color_1_row_2
    dtype: string
  - name: shape_2_row_2
    dtype: string
  - name: color_2_row_2
    dtype: string
  - name: shape_3_row_2
    dtype: string
  - name: color_3_row_2
    dtype: string
  - name: shape_0_row_3
    dtype: string
  - name: color_0_row_3
    dtype: string
  - name: shape_1_row_3
    dtype: string
  - name: color_1_row_3
    dtype: string
  - name: shape_2_row_3
    dtype: string
  - name: color_2_row_3
    dtype: string
  - name: shape_3_row_3
    dtype: string
  - name: color_3_row_3
    dtype: string
  - name: image
    dtype: string
  splits:
  - name: train
    num_bytes: 336856
    num_examples: 1000
  download_size: 39616
  dataset_size: 336856
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: apache-2.0
---

from datasets import load_dataset
from IPython.display import display
from PIL import Image

dataset = load_dataset("gaodrew/gemstone_eval-4by4-256by256px-4shapes-8colors-1k", split='train')

#shapes = ["circle", "triangle", "square", "star"]
#colors = ["red", "orange", "yellow", "green", "blue", "purple", "pink", "black"]

print(dataset)

# Display the first few images
for i in range(10):
    image_path = dataset[i]['image']
    print(f"Image {i+1}:")
    display(Image.open(image_path))