Instructions to use OraRL/Video-ORA-9B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OraRL/Video-ORA-9B with Transformers:

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

pipe = pipeline("image-text-to-text", model="OraRL/Video-ORA-9B")
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, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("OraRL/Video-ORA-9B")
model = AutoModelForMultimodalLM.from_pretrained("OraRL/Video-ORA-9B", device_map="auto")
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 Settings

vLLM

How to use OraRL/Video-ORA-9B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "OraRL/Video-ORA-9B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "OraRL/Video-ORA-9B",
		"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/OraRL/Video-ORA-9B

SGLang

How to use OraRL/Video-ORA-9B 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 "OraRL/Video-ORA-9B" \
    --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": "OraRL/Video-ORA-9B",
		"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 "OraRL/Video-ORA-9B" \
        --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": "OraRL/Video-ORA-9B",
		"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 OraRL/Video-ORA-9B with Docker Model Runner:
```
docker model run hf.co/OraRL/Video-ORA-9B
```

Video-ORA-9B

File size: 3,158 Bytes

53c10a4

# Copy this file, then replace every ../local_data placeholder.
#
# OraRL never downloads data or media. You must obtain each dataset under its
# license, keep the required attribution, and point input/media_root at your
# own local, licensed copy. The optional license and url values are audit
# metadata only and are never opened by the builder.
#
# The seven quotas reproduce the final paper mixture: 100,032 train prompts
# (1,563 batches of 64), with 62,656 structured prompts and 37,376 answer-only
# prompts. The canary is
# additional held-out data and is not subtracted from the train target.
version: 1
seed: 42
target: 100032
canary_size: 512
max_prompts_per_media: 2
require_media: true
allow_shortfall: false

# Supply JSON/JSONL benchmark records before a release build. Any candidate
# sharing a normalized prompt identity or media identity is removed.
benchmark_excludes:
  - ../local_data/exclusions/public_benchmarks.jsonl

sources:
  - name: temporal_grounding_train
    input: ../local_data/annotations/temporal_grounding.jsonl
    task: temporal grounding
    family: temporal
    quota: 20096
    media_root: ../local_data/media/temporal
    license: REPLACE_WITH_DATASET_LICENSE
    url: REPLACE_WITH_DATASET_HOME_PAGE

  - name: tracking_train
    input: ../local_data/annotations/tracking.jsonl
    task: tracking
    family: tracking
    quota: 13952
    media_root: ../local_data/media/tracking
    license: REPLACE_WITH_DATASET_LICENSE
    url: REPLACE_WITH_DATASET_HOME_PAGE

  - name: segmentation_train
    input: ../local_data/annotations/segmentation.jsonl
    task: segmentation
    family: segmentation
    quota: 12032
    media_root: ../local_data/media/segmentation
    license: REPLACE_WITH_DATASET_LICENSE
    url: REPLACE_WITH_DATASET_HOME_PAGE

  - name: spatial_grounding_train
    input: ../local_data/annotations/spatial_grounding.jsonl
    task: spatial grounding
    family: spatial
    quota: 7040
    media_root: ../local_data/media/spatial
    license: REPLACE_WITH_DATASET_LICENSE
    url: REPLACE_WITH_DATASET_HOME_PAGE

  - name: spatial_temporal_grounding_train
    input: ../local_data/annotations/spatial_temporal_grounding.jsonl
    task: spatial-temporal grounding
    family: stvg
    quota: 9536
    media_root: ../local_data/media/spatial_temporal
    license: REPLACE_WITH_DATASET_LICENSE
    url: REPLACE_WITH_DATASET_HOME_PAGE

  - name: video_qa_train
    input: ../local_data/annotations/video_qa.jsonl
    task: video_qa_mc
    family: video_qa
    quota: 20288
    media_root: ../local_data/media/video_qa
    license: REPLACE_WITH_DATASET_LICENSE
    url: REPLACE_WITH_DATASET_HOME_PAGE

  - name: spatial_intelligence_train
    input: ../local_data/annotations/spatial_intelligence.jsonl
    task: spatial intelligence
    family: spatial_intelligence
    # Keep subtype labels such as object_counting and object_rel_direction;
    # the reward adapter uses them to select the official scoring rule.
    preserve_problem_type: true
    quota: 17088
    media_root: ../local_data/media/spatial_intelligence
    license: REPLACE_WITH_DATASET_LICENSE
    url: REPLACE_WITH_DATASET_HOME_PAGE