Instructions to use datalab-to/chandra with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use datalab-to/chandra with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="datalab-to/chandra")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("datalab-to/chandra")
model = AutoModelForImageTextToText.from_pretrained("datalab-to/chandra")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use datalab-to/chandra with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "datalab-to/chandra"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "datalab-to/chandra",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/datalab-to/chandra

SGLang

How to use datalab-to/chandra with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "datalab-to/chandra" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "datalab-to/chandra",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "datalab-to/chandra" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "datalab-to/chandra",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use datalab-to/chandra with Docker Model Runner:
```
docker model run hf.co/datalab-to/chandra
```

Can this model produce layout-aware JSON (blocks, bbox, polygons, hierarchy) ?

by biswajitggiiygg - opened Jan 4

Discussion

biswajitggiiygg

Jan 4

Hi everyone 👋

I’m evaluating Chandra OCR for document OCR on scanned PDFs and images, and I had a question about the structure of the output it can produce.

What I’m trying to achieve

I’m looking for an output format similar to a layout-aware document DOM, for example:

Page → blocks → children hierarchy
Explicit block_type (Page, SectionHeader, Text, Table, TableCell, etc.)
Bounding boxes / polygons for each block
HTML serialization per block (paragraphs, tables, headers)
Stable IDs like /page/0/Table/4
Section hierarchy tracking

Example :

{
  "id": "/page/0/Table/4",
  "block_type": "Table",
  "html": "<table>...</table>",
  "bbox": [x1, y1, x2, y2],
  "polygon": [[...]],
  "children": [...]
}

Or is Chandra intended to provide semantic OCR only (Markdown / HTML / raw text) without explicit geometry, requiring a separate layout-detection step?
Just want to confirm the intended scope and best practice here.
Thanks!

tarun-menta

Datalab org Feb 2

We support a JSON output that does exactly this - https://github.com/datalab-to/chandra?tab=readme-ov-file#how-it-works

tarun-menta changed discussion status to closed Feb 2

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment