thookham/DeOldify-on-Browser
By thookham · ↓ 0 · ♥ 1 · Updated 2025-12-08
What's in this repo
This page lists the files in thookham/DeOldify-on-Browser, including deoldify-art.onnx, deoldify-artistic.onnx, deoldify-quant.onnx and more.
- 5 ONNX models (.onnx files)
- 16 files, including 5 .onnx, 3 .pth, 3 .html, 3 .data, 1 .md
- Last updated 2025-12-08
ONNX files
| File | Size | Download |
|---|---|---|
| deoldify-art.onnx | — | Download |
| deoldify-artistic.onnx | — | Download |
| deoldify-quant.onnx | — | Download |
| deoldify-stable.onnx | — | Download |
| deoldify-video.onnx | — | Download |
How to use
These ONNX files can be loaded with ONNX Runtime. To inspect a model graph, upload the file to the Netron viewer. Download links point to the original files on Hugging Face.
README
DeOldify Browser Integration
This directory contains a browser-based implementation of DeOldify using ONNX Runtime Web. It allows you to run colorization models directly in your web browser without a backend server (after the model is downloaded).
⚠️ Important: How to Run
Due to browser security policies (CORS), you cannot simply double-click the HTML files to run them. You must serve them via a local HTTP server.
Quick Start (Windows/Linux/Mac)
If you have Python installed (which is likely if you're using DeOldify), simply run:
# Run this command in your terminal within this directory
python -m http.server 8000
Then open your browser and visit:
Using the Helper Script (Windows)
We've provided a PowerShell script to make this easy:
- Right-click
serve.ps1and select "Run with PowerShell" - Or run it from the terminal:
.\serve.ps1
Available Models
- Artistic Model: High quality, vibrant colors. Larger download (~243 MB).
- Quantized Model: Faster, smaller download (~61 MB), slightly lower quality.
Troubleshooting
"Failed to fetch" or CORS Errors:
If you see an error about CORS or "Failed to fetch" in the status log, it means you are likely trying to open the file directly (file:///...). Please use the HTTP server method described above.
Memory Issues: The artistic model requires a significant amount of RAM. If the tab crashes, try closing other tabs or using the Quantized model.
Files (16)
Related models
More from thookham
More models
No task type declared.