# Janspiry/Palette-Image-to-Image-Diffusion-Models

Unofficial implementation of Palette: Image-to-Image Diffusion Models by Pytorch

Repository: https://github.com/Janspiry/Palette-Image-to-Image-Diffusion-Models
Canonical: https://ross.abutalabs.com/products/palette-image-to-image-diffusion-models
Language: Python
License: MIT
License Family: permissive
Topics: ddpm, image-restoration, ddp, implementation, pytorch, diffusion-model
Last push: 2023-07-07T13:46:24+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1636, "days_push": 1153, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1833, forks 238 (observed 2026-08-28T04:05:42.386256+00:00)

## What it is
An unofficial PyTorch implementation of the Palette paper on image-to-image diffusion models, supporting tasks like inpainting, uncropping, and colorization. It includes a distributed training pipeline with DDP, EMA, metrics, and pretrained example models.

## Use cases
- implement palette image-to-image diffusion model in pytorch
- inpaint images with a diffusion model
- train a ddpm for image restoration
- colorize grayscale photos with deep learning
- outpaint or uncrop images using diffusion
- distributed multi-gpu training of diffusion models

## When to choose
- you want a working PyTorch reference implementation of Palette for research or experimentation
- you need diffusion-based inpainting with pretrained example checkpoints
- you want a distributed training template for DDPM-style models

## When to avoid
- you need a production-ready image restoration service with support
- you lack GPU resources, since training diffusion models is computationally heavy
- you need the unfinished tasks like uncropping or colorization fully validated

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing
- domain: deep-learning, computer-vision, image-processing
- platform: python
- tags: diffusion-models, ddpm, pytorch, image-to-image, inpainting, colorization, image-restoration, unofficial-implementation, gpu

## Member repositories
- Janspiry/Palette-Image-to-Image-Diffusion-Models (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:42.386256+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:18:37.061914+00:00, confidence not recorded.
  - readme: https://github.com/Janspiry/Palette-Image-to-Image-Diffusion-Models (fetched 2026-08-28T04:05:42.386256+00:00, sha 876f82c45960)
- Data as of 2026-08-30T08:39:29.467469+00:00.