Skip to main content

MLX Nougat

MLX Nougat is a CLI tool for OCR using the Nougat model.

Installation

  1. Install ImageMagick:

    brew install imagemagick
    
  2. Configure environment variables for ImageMagick:

    Add the following lines to your shell configuration file (e.g., ~/.bashrc, ~/.zshrc):

    export MAGICK_HOME=$(brew --prefix imagemagick)
    export PATH=$MAGICK_HOME/bin:$PATH
    export DYLD_LIBRARY_PATH=$MAGICK_HOME/lib:$DYLD_LIBRARY_PATH
    

    After adding these lines, reload your shell configuration or restart your terminal.

  3. Install MLX Nougat:

    git clone git@github.com:mzbac/mlx-nougat.git
    cd mlx-nougat
    pip install .
    

Usage

After installation, you can use MLX Nougat from the command line:

mlx_nougat --input <path_to_image_or_pdf_or_url> [--output <output_file>] [--model <model_name_or_path>]

Arguments

  • --input: (Required) Path to the input image or PDF file, or a URL to an image or PDF.
  • --output: (Optional) Path to save the output text file. If not provided, the output will be printed to the console.
  • --model: (Optional) Name or path of the Nougat model to use. Default is "facebook/nougat-small".

Examples

  1. Process a local image:

    mlx_nougat --input path/to/your/image.png --output results.txt
    
  2. Process a local PDF:

    mlx_nougat --input path/to/your/document.pdf --output results.txt
    
  3. Process a remote image:

    mlx_nougat --input https://example.com/image.jpg --output results.txt
    
  4. Process a remote PDF:

    mlx_nougat --input https://example.com/document.pdf --output results.txt
    
  5. Use a different model:

    mlx_nougat --input path/to/your/image.png --model facebook/nougat-base --output results.txt
    
  6. Use a quantized model:

    mlx_nougat --input path/to/your/document.pdf --model mzbac/nougat-small-8bit-mlx
    

TODOs

  • Support quantized model to improve the performance.

Acknowledgements

This project is built upon several open-source projects and research works:

  • Nougat: The original Nougat model developed by Facebook AI Research.
  • faster-nougat: An optimized implementation of Nougat, which inspired this MLX-based version.
  • MLX: The machine learning framework developed by Apple, used for efficient model inference in this project.
  • Transformers: Hugging Face's state-of-the-art natural language processing library.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mlx_nougat-0.1.2.tar.gz (14.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

mlx_nougat-0.1.2-py3-none-any.whl (15.8 kB view details)

Uploaded Python 3

File details

Details for the file mlx_nougat-0.1.2.tar.gz.

File metadata

  • Download URL: mlx_nougat-0.1.2.tar.gz
  • Upload date:
  • Size: 14.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.9

File hashes

Hashes for mlx_nougat-0.1.2.tar.gz
Algorithm Hash digest
SHA256 4bdb02f05c8569725c2953ff659164a0bfc39685ca66951676d25927441ac33e
MD5 61204a69ad338679a7a41a8658f1d87b
BLAKE2b-256 b075ed31c1e8d958dc4e729db3fd4bce80df3af1c9c740318292b678a4eb4f7d

See more details on using hashes here.

File details

Details for the file mlx_nougat-0.1.2-py3-none-any.whl.

File metadata

  • Download URL: mlx_nougat-0.1.2-py3-none-any.whl
  • Upload date:
  • Size: 15.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.9

File hashes

Hashes for mlx_nougat-0.1.2-py3-none-any.whl
Algorithm Hash digest
SHA256 8ba15588cc98919ef35ab23945741a9824d870a7a54c76c3696c21ed0952acfb
MD5 7795d91e0c7a6661710bb068c921ad72
BLAKE2b-256 5d1c64877ad294120426975f0ad4722271df53db7490b99179df088e5c4e6c1d

See more details on using hashes here.

Release history Release notifications | RSS feed

This release

0.1.2 This release

2 files

0.1.1

2 files

0.1.0

2 files

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page