Skip to content

About

Complete pipeline for upscaling images with ONNX models on Google Colab. PyTorch to ONNX conversion, GPU tiled inference, OpenModelDB integration, and persistent storage via Drive. Works on any device with a browser.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

🚀 ONNX Image Upscaler — Complete Pipeline for Everyone

Author: [Jym Nils Caballero] Contact: [GitHub /jymnils@gmail.com]

Open In Colab Abrir en Colab License: CC BY-NC 4.0

The Reality

Image super-resolution has advanced enormously. Models capable of upscaling images to impressive resolutions are published every day. But most users don't have a powerful GPU in their computer. This creates a gap: repositories full of extraordinary models that only those with dedicated hardware can use. If you don't have a GPU, you're left out.

Sound familiar? It did to me too. That's why I built this.

The Solution

This notebook runs in the cloud. Google Colab gives you a free T4 GPU and this notebook does the rest. You don't need to install anything, you don't need special hardware, you don't need to know how to code. You only need a browser.

  • Windows PC → works.
  • Mac → works.
  • Chromebook → works.
  • Android phone → works.
  • iPhone → works.

If you have a browser, you have a GPU.

What's Included

This isn't a basic notebook. It's a professional pipeline with everything you need:

Section What it does
Inference Upscales images with smart GPU tiling for any resolution. Downloads in PNG and JPG with embedded metadata.
Conversion Transforms PyTorch models to ONNX with one click. Automatic architecture detection via Spandrel. fp32/fp16 precision selection.
Models Complete model info card with OpenModelDB data. Editable personal comments to organize your collection. Safe deletion.
Drive Persistent storage on Google Drive. Your models, info cards, and comments survive Colab session closures. Automatic sync between sessions.

Technical Features

  • Tile processing with overlap to prevent seam artifacts
  • PyTorch → ONNX conversion with dynamic axes for full tiling compatibility
  • OpenModelDB scraping to generate model cards with architecture, scale, dataset, license, and author info
  • Embedded metadata in ONNX files and in PNG/JPG output files
  • Overwrite protection with auto-rename
  • Professional web interface with 3 tabs (Gradio 6)

Key Dependencies

Gradio 6 · ONNX Runtime GPU · PyTorch 2.11 · Spandrel · BeautifulSoup4 · Google Drive API

Credits

Third-Party Licenses

This project uses the following open-source libraries:

Library License Copyright
Gradio Apache 2.0 Copyright (c) 2022 Gradio, Inc.
ONNX Runtime MIT Copyright (c) Microsoft Corporation
PyTorch BSD-style Copyright (c) 2016-2025 Facebook, Inc.
TorchVision BSD-style Copyright (c) 2016-2025 Facebook, Inc.
Spandrel MIT Copyright (c) chaiNNer-org
Beautiful Soup 4 MIT Copyright (c) 2004-2025 Leonard Richardson
Requests Apache 2.0 Copyright (c) 2019 Kenneth Reitz
ONNX Apache 2.0 Copyright (c) ONNX Project Contributors
NumPy BSD 3-Clause Copyright (c) 2005-2025 NumPy Developers
Pillow HPND Copyright (c) 1997-2025 Alex Clark and contributors
onnxconverter-common MIT Copyright (c) Microsoft Corporation
onnxscript Apache 2.0 Copyright (c) Microsoft Corporation

See THIRD_PARTY_LICENSES.md for full license texts.

License

This project is licensed under CC BY-NC 4.0.

You are free to:

  • Share — copy and redistribute the material in any medium or format
  • Adapt — remix, transform, and build upon the material

Under the following terms:

  • Attribution — You must give appropriate credit and indicate if changes were made
  • NonCommercial — You may not use the material for commercial purposes without explicit permission from the author

For commercial use inquiries, contact: [jymnils@gmail.com]

About

Complete pipeline for upscaling images with ONNX models on Google Colab. PyTorch to ONNX conversion, GPU tiled inference, OpenModelDB integration, and persistent storage via Drive. Works on any device with a browser.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages