hralamin6/gemma4e2b-ocr-finetuned

VISIONConcurrency Cost:1Model Size:5.1BQuant:BF16Ctx Length:32kTool Calling:SupportedPublished:Jun 4, 2026Architecture:Transformer Cold

The hralamin6/gemma4e2b-ocr-finetuned model is a 5.1 billion parameter Gemma-based language model, fine-tuned for OCR-related tasks. It features a 32768-token context length and is provided in GGUF format, optimized for efficient deployment. This model is specifically designed to enhance optical character recognition workflows and applications.

Loading preview...

Model Overview

hralamin6/gemma4e2b-ocr-finetuned is a 5.1 billion parameter model based on the Gemma architecture, specifically fine-tuned for Optical Character Recognition (OCR) related applications. It leverages a substantial 32768-token context window, making it suitable for processing longer documents or complex OCR tasks. The model is distributed in the GGUF format, which is optimized for efficient inference on various hardware.

Key Capabilities

  • OCR-focused Fine-tuning: The model has undergone specialized fine-tuning to improve performance in OCR contexts.
  • GGUF Format: Provided in GGUF, enabling compatibility with llama-cli and other tools for local deployment.
  • Unsloth Optimization: The fine-tuning process utilized Unsloth, indicating potential for faster training and efficient model conversion.
  • Multimodal Support: Includes a separate mmproj file for multimodal capabilities, though specific usage notes are provided for Ollama.

Deployment Notes

  • Ollama Compatibility: Users should be aware that Ollama currently does not support separate mmproj files for vision models. Instructions are provided to create a unified bf16 model for Ollama deployment.
  • Example Usage: Command-line examples are given for both text-only and multimodal llama-cli applications, facilitating quick setup for developers.

Use Cases

This model is particularly well-suited for applications requiring enhanced language understanding and generation within OCR workflows, such as document analysis, information extraction from scanned texts, and improving the accuracy of text recognition systems.