

QVQ 72B Preview
Overview :
QVQ-72B-Preview is an experimental research model developed by the Qwen team, focusing on enhancing visual reasoning capabilities. The model demonstrates strong abilities in multidisciplinary understanding and reasoning, achieving significant advances especially in mathematical reasoning tasks. Although advancements have been made in visual reasoning, it does not completely replace the capabilities of Qwen2-VL-72B, and may gradually lose focus on image content in multi-step visual reasoning, leading to hallucinations. Furthermore, QVQ does not show significantly better performance in basic recognition tasks compared to Qwen2-VL-72B.
Target Users :
The target audience includes researchers and developers, particularly professionals seeking advanced solutions in the fields of visual reasoning, multidisciplinary understanding, and mathematical reasoning. QVQ-72B-Preview provides a powerful tool to assist them in processing complex visual and textual data, thereby advancing research and applications in relevant fields.
Use Cases
- Use the QVQ-72B-Preview model in multidisciplinary understanding and reasoning tasks in the MMMU benchmark.
- Utilize the model to handle mathematical reasoning tasks in the MathVision benchmark.
- Apply the model to solve challenging problems on OlympiadBench.
Features
- Multidisciplinary understanding and reasoning: Achieved scores as high as 70.3% in the MMMU benchmark test, showcasing strong capabilities in this area.
- Mathematical reasoning tasks: Significant improvements observed in the MathVision benchmark, highlighting the model's capabilities in mathematical reasoning tasks.
- Challenging problem-solving: Performance on OlympiadBench demonstrates the model's ability to solve challenging problems.
- Single-turn dialogue support: Currently, the model only supports single-turn dialogue and image output, with no video input.
- Safety and ethical considerations: Robust safety measures are needed to ensure reliable and secure performance.
- Performance and benchmark limitations: May gradually lose focus on image content in multi-step visual reasoning, leading to hallucinations.
- Basic recognition tasks: Does not show significantly better improvement than Qwen2-VL-72B in basic tasks such as recognizing people, animals, or plants.
How to Use
1. Install the qwen-vl-utils package for more convenient handling of various types of visual input.
2. Load the Qwen2VLForConditionalGeneration model using the transformers library.
3. Import the process_vision_info function from qwen_vl_utils to process visual information.
4. Prepare input messages, including messages from the system role and the user role, where the user message contains images and text.
5. Use the processor.apply_chat_template function to prepare the text required for inference.
6. Call the process_vision_info function to process visual information.
7. Pass the text and visual inputs to the processor to prepare the model's input.
8. Use the model.generate function to produce outputs.
9. Decode the generated IDs using the processor.batch_decode function to obtain the final output text.
Featured AI Tools

Gemini
Gemini is the latest generation of AI system developed by Google DeepMind. It excels in multimodal reasoning, enabling seamless interaction between text, images, videos, audio, and code. Gemini surpasses previous models in language understanding, reasoning, mathematics, programming, and other fields, becoming one of the most powerful AI systems to date. It comes in three different scales to meet various needs from edge computing to cloud computing. Gemini can be widely applied in creative design, writing assistance, question answering, code generation, and more.
AI Model
11.4M
Chinese Picks

Liblibai
LiblibAI is a leading Chinese AI creative platform offering powerful AI creative tools to help creators bring their imagination to life. The platform provides a vast library of free AI creative models, allowing users to search and utilize these models for image, text, and audio creations. Users can also train their own AI models on the platform. Focused on the diverse needs of creators, LiblibAI is committed to creating inclusive conditions and serving the creative industry, ensuring that everyone can enjoy the joy of creation.
AI Model
6.9M