Instructions to use OraRL/Video-ORA-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OraRL/Video-ORA-4B with Transformers:

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

pipe = pipeline("image-text-to-text", model="OraRL/Video-ORA-4B")
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-4B")
model = AutoModelForMultimodalLM.from_pretrained("OraRL/Video-ORA-4B", 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-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "OraRL/Video-ORA-4B"
# 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-4B",
		"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-4B

SGLang

How to use OraRL/Video-ORA-4B 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-4B" \
    --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-4B",
		"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-4B" \
        --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-4B",
		"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-4B with Docker Model Runner:
```
docker model run hf.co/OraRL/Video-ORA-4B
```

Video-ORA-4B / code /verl /models /monkey_patch.py

lyhisme

Add 4B model card, benchmark assets, and release code

0185029 verified about 13 hours ago

Raw

History Blame Contribute Delete

3.51 kB

	# Copyright 2024 Bytedance Ltd. and/or its affiliates
	#
	# Licensed under the Apache License, Version 2.0 (the "License");
	# you may not use this file except in compliance with the License.
	# You may obtain a copy of the License at
	#
	# http://www.apache.org/licenses/LICENSE-2.0
	#
	# Unless required by applicable law or agreed to in writing, software
	# distributed under the License is distributed on an "AS IS" BASIS,
	# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
	# See the License for the specific language governing permissions and
	# limitations under the License.

	from transformers.modeling_utils import ALL_ATTENTION_FUNCTIONS

	from ..utils.py_functional import is_transformers_version_greater_than
	from .transformers.flash_attention_utils import flash_attention_forward


	SUPPORTED_MODEL_TYPE = (
	"llama",
	"gemma",
	"gemma2",
	"mistral",
	"qwen2",
	"qwen2_moe",
	"qwen3",
	"qwen3_moe",
	"qwen2_vl",
	"qwen2_5_vl",
	"qwen3_vl",
	"qwen3_vl_moe",
	"qwen3_5",
	)

	QWEN2_VL_MODELS = ("qwen2_vl", "qwen2_5_vl")
	QWEN3_VL_MODELS = ("qwen3_vl", "qwen3_vl_moe")
	QWEN3_5_MODELS = ("qwen3_5",)


	def apply_ulysses_patch(model_type: str) -> None:
	if not is_transformers_version_greater_than("4.54.0"):
	raise RuntimeError("Only support transformers >= 4.54.0.")

	if model_type in SUPPORTED_MODEL_TYPE:
	ALL_ATTENTION_FUNCTIONS["flash_attention_2"] = flash_attention_forward
	else:
	raise NotImplementedError(f"Model architecture {model_type} is not supported yet.")

	if model_type in QWEN2_VL_MODELS:
	from transformers.models.qwen2_5_vl.modeling_qwen2_5_vl import (
	Qwen2_5_VLForConditionalGeneration,
	Qwen2_5_VLModel,
	)
	from transformers.models.qwen2_vl.modeling_qwen2_vl import Qwen2VLForConditionalGeneration, Qwen2VLModel

	from .transformers.qwen2_vl import qwen2_vl_base_forward, qwen2_vl_model_forward

	# fix text-image mixed data
	Qwen2VLModel.forward = qwen2_vl_base_forward
	Qwen2_5_VLModel.forward = qwen2_vl_base_forward
	# TODO: add linear cross entropy kernels
	Qwen2VLForConditionalGeneration.forward = qwen2_vl_model_forward
	Qwen2_5_VLForConditionalGeneration.forward = qwen2_vl_model_forward
	elif model_type in QWEN3_VL_MODELS:
	from transformers.models.qwen3_vl.modeling_qwen3_vl import Qwen3VLForConditionalGeneration, Qwen3VLModel
	from transformers.models.qwen3_vl_moe.modeling_qwen3_vl_moe import (
	Qwen3VLMoeForConditionalGeneration,
	Qwen3VLMoeModel,
	)

	from .transformers.qwen3_vl import qwen3_vl_base_forward, qwen3_vl_model_forward

	# fix text-image mixed data
	Qwen3VLModel.forward = qwen3_vl_base_forward
	Qwen3VLMoeModel.forward = qwen3_vl_base_forward
	# TODO: add linear cross entropy kernels
	Qwen3VLForConditionalGeneration.forward = qwen3_vl_model_forward
	Qwen3VLMoeForConditionalGeneration.forward = qwen3_vl_model_forward
	elif model_type in QWEN3_5_MODELS:
	from transformers.models.qwen3_5.modeling_qwen3_5 import Qwen3_5ForConditionalGeneration, Qwen3_5Model

	from .transformers.qwen3_5 import qwen3_5_base_forward, qwen3_5_model_forward

	# fix text-image mixed data
	Qwen3_5Model.forward = qwen3_5_base_forward
	# TODO: add linear cross entropy kernels
	Qwen3_5ForConditionalGeneration.forward = qwen3_5_model_forward