---
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: image
    dtype: image
  - name: serial_number
    dtype: string
  - name: toughness
    dtype: int64
  splits:
  - name: train
    num_bytes: 6074617.0
    num_examples: 55
  download_size: 6059493
  dataset_size: 6074617.0
---
```python
import requests
import io
import json
import datasets
import huggingface_hub

huggingface_hub.login(HF_TOKEN)
ds = datasets.load_dataset('kahua-ml/serial_number_dataset')['train']

headers = {
    'accept': 'application/json',
}

# Convert PIL Image to bytes
image_buffer = io.BytesIO()
ds[0]['image'].save(image_buffer, format='JPEG')
image_buffer.seek(0)

fields = [
    {"field_mapping":"serial_number","field_name":"serial_number","field_type":"text"},
]

files = {
    'file': ('image.jpg', image_buffer, 'image/jpeg'),
    'input': (None, '{"domain_id":"domain456","fields":' + json.dumps(fields) + ',"user_id":"user123"}'),
}

response = requests.post(
    AZURE_URL,
    headers=headers,
    files=files,
)
```
---

dataset_info:
  features:
  - name: image
    dtype: image
  - name: serial_number
    dtype: string
  splits:
  - name: train
    num_bytes: 8858770.0
    num_examples: 61
  download_size: 8670997
  dataset_size: 8858770.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---