Skip to content
ModelListedExternal
Download

Qwen3 VL 2B Persian Arabic Ocr v1.0

This is a 16-bit version of Qwen/Qwen3-VL-2B-Instruct fine-tuned specifically for Persian text recognition (OCR) on individual text lines. The model has been trained exclusively on cropped single-line text images and...

KindText modelBase modelQwen/Qwen3-VL-2B-InstructVersionmodel

Model source

Kind
Text model
File size
3.96 GB
Base model
Qwen/Qwen3-VL-2B-Instruct
Version
vmodel
Updated
Sep 17, 2026
Downloads
432
Tasks
Captioning
Source
Hugging Face

Source description

Persian/Arabic OCR - Qwen3-VL-2B-Instruct - v1.0

This is a 16-bit version of Qwen/Qwen3-VL-2B-Instruct fine-tuned specifically for Persian text recognition (OCR) on individual text lines. The model has been trained exclusively on cropped single-line text images and is not designed for full-page OCR.

Training Details

  • Dataset:
    • 56,000 real Persian text line images
    • 100,000 synthetic images (47 fonts in 3 diffrent sizes) with clean and noisy/colored backgrounds
  • Total Examples: 156k text line images
  • Epochs: 1
  • LoRA Rank: 512
  • Batch Size: 100
  • Learning Rate: 2e-4
  • Trainable Parameters: 759M Params

Training Performance

Training loss decreased steadily, ending at approximately 0.071.

StepTraining Loss
02.2102
1000.1577
2000.1384
3000.1330
4000.1208
5000.1078
6000.1080
7000.1065
8000.0980
9000.0871
10000.0827
11000.0802
12000.0898
13000.0866
14000.0811
15000.0774
15500.0714

Sources

1 source
Hugging FacePrimaryVersion f595ee658b9d9d9fb51d5d7a7647a44154051491
huggingface.co/mohajesmaeili/Qwen3-VL-2B-Persian-Arabic-Ocr-v1.0/tree/f595ee658b9d9d9fb51d5d7a7647a44154051491

Verified Sep 17

Model artifacts

1 artifact

model.safetensors

safetensors · 3.96 GB · SHA-256 ebee35c899fc…411e · Hugging Face

Download

Source excerpts

2 excerpts

--- base_model: - Qwen/Qwen3-VL-2B-Instruct tags: - ocr - qwen3_vl - persian - arabic - text-line-ocr - Optical Character Recognition - vision-language - vl - persian-ocr - arabic-ocr - farsi license: apache-2.0 language: - en - fa - ar datasets: - mohajesmaeili/Persian_Arabic_TextLine_Image_Ocr_Small pipeline_tag: image-to-text --- # Persian/Arabic OCR - Qwen3-VL-2B-Instruct - v1.0 This is a **16-bit version** of **Qwen/Qwen3-VL-2B-Instruct** fine-tuned specifically for Persian text recognition (OCR) on **individual text lines**. The model has been trained exclusively on cropped single-line text images and is **not designed for full-page OCR**. ## Training Details - **Dataset**: - 56,000 real Persian text line images - 100,000 synthetic images (47 fonts in 3 diffrent sizes) with clean and noisy/colored backgrounds - **Total Examples**: 156k text line images - **Epochs**: 1 - **LoRA Rank**: 512 - **Batch Size**: 100 - **Learning Rate**: 2e-4 - **Trainable Parameters**: 759M Params ## Training Performance Training loss decreased steadily, ending at approximately **0.071**. | Step | Training Loss | |--------|---------------| | 0 | 2.2102 | | 100 | 0.1577 | | 200 | 0.1384 | | 300 | 0.1330 | | 400 | 0.1208 | | 500 | 0.1078 | | 600 | 0.1080 | | 700 | 0.1065 | | 800 | 0.0980 | | 900 | 0.0871 | | 1000 | 0.0827 | | 1100 | 0.0802 | | 1200 | 0.0898 | | 1300 | 0.0866 | | 1400 | 0.0811 | | 1500 | 0.0774 | | 1550 | 0.0714 |

Source context: 432 downloads · 17 likes · Pipeline image-to-text · Repo mohajesmaeili/Qwen3-VL-2B-Persian-Arabic-Ocr-v1.0