From the model card
Fields below are copied from the tags and counters on the HuggingFace repository Qwen/Qwen2-VL-2B-Instruct at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.
- Publisher (HF namespace)
- Qwen
- Pipeline tag
- image-text-to-text
- Library
- Transformers
- Weight formats
- safetensors
- License tag
apache-2.0— read the license file in the repo before relying on it- Lineage
-
- base model Qwen/Qwen2-VL-2B
- fine-tune of Qwen/Qwen2-VL-2B
- Language tags
- English (en)
- Papers cited
- arXiv:2409.12191, arXiv:2308.12966
- Downloads (HF counter at last fetch)
- 1,821,310
- Likes (HF counter at last fetch)
- 518
- Model card
- https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct
Use cases
- Captioning product images in e-commerce pipelines
- Visual question answering over uploaded charts or diagrams
- Document OCR on edge devices with limited VRAM
- Lightweight VQA in mobile or embedded applications
Pros
- Runs in under 8GB VRAM making it edge-deployable
- Apache 2.0 license with no commercial restrictions
- Strong OCR and structured document understanding for its parameter count
Cons
- 2B scale trails larger VL models on complex visual reasoning tasks
- Shorter context window than Qwen2-VL-7B variant
- Video understanding limited compared to dedicated video-language models
Tags
transformerssafetensorsqwen2_vlimage-text-to-textmultimodalconversationalenarxiv:2409.12191arxiv:2308.12966base_model:Qwen/Qwen2-VL-2Bbase_model:finetune:Qwen/Qwen2-VL-2Blicense:apache-2.0text-generation-inferenceendpoints_compatibleregion:usdeploy:azure