Qwen3-VL-4B-Instruct Fine-Tuned Table Image-to-HTML Model

This model is a fine-tuned version of Qwen3-VL-4B-Instruct, specialized for converting table images into HTML code.

Fine-Tuning Details

  • Base model: Qwen3-VL-4B-Instruct
  • Task: Table image-to-HTML conversion
  • Data type: Image
  • File format: JSON
  • Training data: 500,777 examples
  • Data size: 15.7 GB
  • Epochs: 1
  • Dataset: PubTabNet-based table image data

Evaluation

Metric Before After
TEDS Struct Score 0.8527 0.8738

Government R&D Project

This software is a result of the following government research and development project:

Item Details
Project Name AI 기반 중소기업 업무 자동화 및 통합 검색 솔루션
Project No. RQT-25-090153
Supervisor 정보통신산업진흥원
Performer Proten Co., Ltd.
R&D Period 2026.01.01 ~ 2026.06.30
Downloads last month
15
Safetensors
Model size
4B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for ProtenLabs/Qwen3-VL-4B-Instruct-pubnet

Finetuned
(305)
this model