Loyca-Qwen3-VL-2B-Instruct-ImgAnalysis is a lightweight LoRA adapter built on top of Qwen/Qwen3-VL-2B-Instruct, fine-tuned for screen content understanding.
It enhances the base model’s ability to read and interpret text embedded in images — particularly screenshots and user interfaces — and respond with structured, instruction-following outputs.

Model Sources

Repository: https://huggingface.co/Vokturz/Loyca-Qwen3-VL-2B-Instruct-ImgAnalysis
Base model: https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct
Fine-tuning run: W&B Experiment

Uses

This model can be used for captioning or summarizing visual elements that contain textual information, and extracting structured text (e.g., app names, UI labels, error messages) from software screenshots.

The model is not designed for:

Handwritten OCR
Scene text in natural environments (e.g., street signs)
Legal or financial document processing without human review

Training Details

Training Data

The model was trained on Vokturz/sourceforge-app-screenshots-ocr (~1100 records), a custom dataset of annotated application screenshots containing readable text and UI elements. The dataset focuses on clean UI text extraction rather than general image captioning.

Training Hyperparameters

Parameter	Value
Epochs	6
Batch size	8
Learning rate	2.5e-4
LoRA rank	64
LoRA alpha	64
Precision	bfloat16 (mixed)
Optimizer	AdamW
Scheduler	Cosine decay
Gradient accumulation	2
Weight decay	0.01