language:

• en library_name: transformers license: apache-2.0 metrics:
• accuracy tags:
• multimodal pipeline_tag: video-text-to-text base_model: Qwen/Qwen2-VL-7B-Instruct

💡 VideoChat-R1_7B_caption

[📂 GitHub]
[📜 Tech Report]

🚀 How to use the model

We provide a simple installation example below:

pip install transformers
pip install qwen_vl_utils

Then you could use our model:

from transformers import Qwen2_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info

model_path = "OpenGVLab/VideoChat-R1_7B_caption"
# default: Load the model on the available device(s)
model = Qwen2_VLForConditionalGeneration.from_pretrained(
    model_path, torch_dtype="auto", device_map="auto",
    attn_implementation="flash_attention_2"
)

# default processer
processor = AutoProcessor.from_pretrained(model_path)

video_path = "your_video.mp4"
question = "Describe the video in detail."

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "video",
                "video": video_path,
                "max_pixels": 360 * 420,
                "fps": 1.0,
            },
            {"type": "text", "text": f""""{question} First output the thinking process in <think> </think> tags and then output the final answer in <answer> </answer> tags"""},
        ],
    }
]



#In Qwen 2 VL, frame rate information is also input into the model to align with absolute time.
# Preparation for inference
text = processor.apply_chat_template(
    messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs, video_kwargs = process_vision_info(messages, return_video_kwargs=True)
inputs = processor(
    text=[text],
    images=image_inputs,
    videos=video_inputs,
    padding=True,
    return_tensors="pt",
    **video_kwargs,
)
inputs = inputs.to("cuda")

# Inference
generated_ids = model.generate(**inputs, max_new_tokens=512)
generated_ids_trimmed = [
    out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
    generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)

✏️ Citation

@article{li2025videochatr1,
  title={VideoChat-R1: Enhancing Spatio-Temporal Perception via Reinforcement Fine-Tuning},
  author={Li, Xinhao and Yan, Ziang and Meng, Desen and Dong, Lu and Zeng, Xiangyu and He, Yinan and Wang, Yali and Qiao, Yu and Wang, Yi and Wang, Limin},
  journal={arXiv preprint arXiv:2504.06958},
  year={2025}
}