Ba2han/qwen-TR-OCR
VISIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2.3BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jul 18, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold
The Ba2han/qwen-TR-OCR is a 2.3 billion parameter Qwen3.5 model developed by Ba2han, fine-tuned from Ba2han/tr_ocr-test. This model is specifically designed for Optical Character Recognition (OCR) tasks, focusing on Turkish text. It was trained using Unsloth, which enabled a 2x faster training process, making it efficient for OCR applications.
Loading preview...
Overview
The Ba2han/qwen-TR-OCR is a specialized 2.3 billion parameter model based on the Qwen3.5 architecture, developed by Ba2han. It has been fine-tuned from the Ba2han/tr_ocr-test model, indicating its primary focus on Optical Character Recognition (OCR).
Key Capabilities
- Turkish OCR: This model is specifically designed and optimized for processing and recognizing text in the Turkish language, making it suitable for applications requiring Turkish OCR capabilities.
- Efficient Training: The model was trained using Unsloth, which significantly accelerated the training process by a factor of two. This suggests an optimized and efficient development pipeline.
Good For
- Turkish Document Processing: Ideal for developers and organizations working with Turkish documents, images, or scanned texts that require accurate character recognition.
- Integration into OCR Systems: Can be integrated as a core component in larger systems that need to extract text from Turkish-language visual inputs.