Top AI Repos — open-source AI, indexed and scored

Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.

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

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

43MITPython

View on GitHub

Stars: 1,831
Forks: 236
Watchers: 16
Open issues: 34
Contributors: 3

Star history, last 90 days-1 over 21 snapshots

2026-07-31 · 1,8322026-11-04 · 1,831

Star history, last 90 days
Date	Stars
2026-07-31	1832
2026-08-09	1832
2026-08-18	1834
2026-08-21	1832
2026-08-22	1833
2026-08-29	1833
2026-09-04	1834
2026-09-09	1835
2026-09-20	1834
2026-09-23	1835
2026-09-27	1836
2026-09-29	1835
2026-09-30	1834
2026-10-03	1835
2026-10-05	1835
2026-10-06	1835
2026-10-10	1834
2026-10-13	1833
2026-10-19	1832
2026-10-28	1831
2026-11-04	1831

Today

— stars today

This week

— stars this week

This month

-4 stars this month

Momentum

0.0

growth rate 0.00%/day

From the README

# Palette: Image-to-Image Diffusion Models

[Paper](https://arxiv.org/pdf/2111.05826.pdf ) |  [Project](https://iterative-refinement.github.io/palette/ )

## Brief

This is an unofficial implementation of **Palette: Image-to-Image Diffusion Models** by **Pytorch**, and it is mainly inherited from its super-resolution version [Image-Super-Resolution-via-Iterative-Refinement](https://github.com/Janspiry/Image-Super-Resolution-via-Iterative-Refinement). The code template is from my another seed project: [distributed-pytorch-template](https://github.com/Janspiry/distributed-pytorch-template).

There are some implementation details with paper descriptions:

- We adapted the U-Net architecture used in  `Guided-Diffusion`, which give a substantial boost to sample quality.
- We used the attention mechanism in low-resolution features (16×16) like vanilla `DDPM`.
- We encode the $\gamma$ rather than $t$ in `Palette` and embed it with affine transformation.
- We fix the variance $Σ_\theta(x_t, t)$ to a constant during the inference as described in `Palette`.

## Status

### Code
- [x] Diffusion Model Pipeline
- [x] Train/Test Process
- [x] Save/Load Training State
- [x] Logger/Tensorboard
- [x] Multiple GPU Training (DDP)
- [x] EMA
- [x] Metrics (now for FID, IS)
- [x] Dataset (now for inpainting, uncropping, colorization)
- [x] Google colab script 🌟(now for inpainting)

### Task

I try to finish following tasks in order:
- [x] Inpainting on [CelebaHQ](https://drive.google.com/drive/folders/1CjZAajyf-jIknskoTQ4CGvVkAigkhNWA?usp=sharing)🚀 ([Google Colab](https://colab.research.google.com/drive/1wfcd6QKkN2AqZDGFKZLyGKAoI5xcXUgO#scrollTo=8VFpuekybeQK))
- [x] Inpainting on [Places2 with 128×128 centering mask](https://drive.google.com/drive/folders/1fLyFtrStfEtyrqwI0N_Xb_3idsf0gz0M?usp=sharing)🚀

The follow-up experiment is uncertain, due to lack of time and GPU resources:

- [ ] Uncropping on Places2
- [ ] Colorization on ImageNet val set 

## Results

The DDPM model requires significant computational resources, and we have only built a few example models to validate the ideas in this paper.

### Visuals

#### Celeba-HQ

Results with 200 epochs and 930K iterations, and the first 100 samples in [centering mask](https://drive.google.com/drive/folders/10zyHZtYV5vCht2MGNCF8WzpZJT2ae2RS?usp=sharing) and [irregular mask](https://drive.google.com/drive/folders/1vmSI-R9J2yQZY1cVkSSZlTYil2DprzvY?usp=sharing). 

| ![Process_02323](misc//image//Process_02323.jpg) |    ![Process_02323](misc//image//Process_26190.jpg)  |
| ------------------------------------------------ | ---- |

#### Places2 with 128×128 centering mask

Results with 16 epochs and 660K iterations, and the several **picked** samples in [centering mask](https://drive.google.com/drive/folders/1XusKO0_M6GUfPG-FOlID0Xcp0SiexKNe?usp=sharing).

| ![Mask_Places365_test_00209019.jpg](misc//image//Mask_Places365_test_00209019.jpg) | ![Mask_Places365_test_00143399.jpg](misc//image//Mask_Places365_test_00143399.jpg) | ![Mask_Places365_test_00263905.jpg](misc//image//Mask_Places365_test_00263905.jpg) |  ![Mask_Places365_test_00144085.jpg](misc//image//Mask_Places365_test_00144085.jpg)    |
| ------------------------------------------------------------ | ------------------------------------------------------------ | ------------------------------------------------------------ | ---- |
| ![Out_Places365_test_00209019](misc//image//Out_Places365_test_00209019.jpg) | ![Out_Places365_test_00143399.jpg](misc//image//Out_Places365_test_00143399.jpg) | ![Out_Places365_test_00263905.jpg](misc//image//Out_Places365_test_00263905.jpg) | ![Out_Places365_test_00144085.jpg](misc//image//Out_Places365_test_00144085.jpg)    |

#### Uncropping on Places2

Results with 8 epochs and 330K iterations, and the several  **picked** samples in [uncropping](https://drive.google.com/drive/folders/1tC3B8ayaadhXAJrOCTrw15R8t84REPWJ?usp=sharing).
| ![Process_Places365_test_00309553](misc//image//Process_Places365_test_003095

Excerpt of 8,673 characters

Read on GitHub

Top contributors3 shown

Janspiry
25
bruce-willis
Yury Belousov · Switzerland
1
PuAnysh
1

Quality score

Would you bet a product on this? Bounded 0–100 and slow moving.

43/10043

Maintenance: 5%; How recently the default branch was pushed to
Releases: 25%; Recency and cadence of tagged releases
Community: 14%; Contributor breadth and bus factor
Issue backlog: 85%; Open issues measured against the size of the audience
Documentation: 84%; README depth, plus a homepage and topics
Licensing: 100%; How freely the code can be adopted

What to know before adopting

No commits in over a year
No tagged releases
Bus factor of 1 — one contributor dominates
Usable in commercial products with attribution.

Details

Created: 5 years ago
Last push: 3 years ago
Last release: none
Release tag: —
Releases in the last year: 0
Default branch: main
Size: 7.0k KB
Owner type: User
Owner location: —
Discovered via: topic:pytorch
First indexed: 3 months ago
Last synced: 5 days ago

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

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

43MITPython

View on GitHub

Stars: 1,831
Forks: 236
Watchers: 16
Open issues: 34
Contributors: 3

Star history, last 90 days-1 over 21 snapshots

2026-07-31 · 1,8322026-11-04 · 1,831

Star history, last 90 days
Date	Stars
2026-07-31	1832
2026-08-09	1832
2026-08-18	1834
2026-08-21	1832
2026-08-22	1833
2026-08-29	1833
2026-09-04	1834
2026-09-09	1835
2026-09-20	1834
2026-09-23	1835
2026-09-27	1836
2026-09-29	1835
2026-09-30	1834
2026-10-03	1835
2026-10-05	1835
2026-10-06	1835
2026-10-10	1834
2026-10-13	1833
2026-10-19	1832
2026-10-28	1831
2026-11-04	1831

Today

— stars today

This week

— stars this week

This month

-4 stars this month

Momentum

0.0

growth rate 0.00%/day

From the README

# Palette: Image-to-Image Diffusion Models

[Paper](https://arxiv.org/pdf/2111.05826.pdf ) |  [Project](https://iterative-refinement.github.io/palette/ )

## Brief

This is an unofficial implementation of **Palette: Image-to-Image Diffusion Models** by **Pytorch**, and it is mainly inherited from its super-resolution version [Image-Super-Resolution-via-Iterative-Refinement](https://github.com/Janspiry/Image-Super-Resolution-via-Iterative-Refinement). The code template is from my another seed project: [distributed-pytorch-template](https://github.com/Janspiry/distributed-pytorch-template).

There are some implementation details with paper descriptions:

- We adapted the U-Net architecture used in  `Guided-Diffusion`, which give a substantial boost to sample quality.
- We used the attention mechanism in low-resolution features (16×16) like vanilla `DDPM`.
- We encode the $\gamma$ rather than $t$ in `Palette` and embed it with affine transformation.
- We fix the variance $Σ_\theta(x_t, t)$ to a constant during the inference as described in `Palette`.

## Status

### Code
- [x] Diffusion Model Pipeline
- [x] Train/Test Process
- [x] Save/Load Training State
- [x] Logger/Tensorboard
- [x] Multiple GPU Training (DDP)
- [x] EMA
- [x] Metrics (now for FID, IS)
- [x] Dataset (now for inpainting, uncropping, colorization)
- [x] Google colab script 🌟(now for inpainting)

### Task

I try to finish following tasks in order:
- [x] Inpainting on [CelebaHQ](https://drive.google.com/drive/folders/1CjZAajyf-jIknskoTQ4CGvVkAigkhNWA?usp=sharing)🚀 ([Google Colab](https://colab.research.google.com/drive/1wfcd6QKkN2AqZDGFKZLyGKAoI5xcXUgO#scrollTo=8VFpuekybeQK))
- [x] Inpainting on [Places2 with 128×128 centering mask](https://drive.google.com/drive/folders/1fLyFtrStfEtyrqwI0N_Xb_3idsf0gz0M?usp=sharing)🚀

The follow-up experiment is uncertain, due to lack of time and GPU resources:

- [ ] Uncropping on Places2
- [ ] Colorization on ImageNet val set 

## Results

The DDPM model requires significant computational resources, and we have only built a few example models to validate the ideas in this paper.

### Visuals

#### Celeba-HQ

Results with 200 epochs and 930K iterations, and the first 100 samples in [centering mask](https://drive.google.com/drive/folders/10zyHZtYV5vCht2MGNCF8WzpZJT2ae2RS?usp=sharing) and [irregular mask](https://drive.google.com/drive/folders/1vmSI-R9J2yQZY1cVkSSZlTYil2DprzvY?usp=sharing). 

| ![Process_02323](misc//image//Process_02323.jpg) |    ![Process_02323](misc//image//Process_26190.jpg)  |
| ------------------------------------------------ | ---- |

#### Places2 with 128×128 centering mask

Results with 16 epochs and 660K iterations, and the several **picked** samples in [centering mask](https://drive.google.com/drive/folders/1XusKO0_M6GUfPG-FOlID0Xcp0SiexKNe?usp=sharing).

| ![Mask_Places365_test_00209019.jpg](misc//image//Mask_Places365_test_00209019.jpg) | ![Mask_Places365_test_00143399.jpg](misc//image//Mask_Places365_test_00143399.jpg) | ![Mask_Places365_test_00263905.jpg](misc//image//Mask_Places365_test_00263905.jpg) |  ![Mask_Places365_test_00144085.jpg](misc//image//Mask_Places365_test_00144085.jpg)    |
| ------------------------------------------------------------ | ------------------------------------------------------------ | ------------------------------------------------------------ | ---- |
| ![Out_Places365_test_00209019](misc//image//Out_Places365_test_00209019.jpg) | ![Out_Places365_test_00143399.jpg](misc//image//Out_Places365_test_00143399.jpg) | ![Out_Places365_test_00263905.jpg](misc//image//Out_Places365_test_00263905.jpg) | ![Out_Places365_test_00144085.jpg](misc//image//Out_Places365_test_00144085.jpg)    |

#### Uncropping on Places2

Results with 8 epochs and 330K iterations, and the several  **picked** samples in [uncropping](https://drive.google.com/drive/folders/1tC3B8ayaadhXAJrOCTrw15R8t84REPWJ?usp=sharing).
| ![Process_Places365_test_00309553](misc//image//Process_Places365_test_003095

Excerpt of 8,673 characters

Read on GitHub

Top contributors3 shown

Janspiry
25
bruce-willis
Yury Belousov · Switzerland
1
PuAnysh
1

Quality score

Would you bet a product on this? Bounded 0–100 and slow moving.

43/10043

Maintenance: 5%; How recently the default branch was pushed to
Releases: 25%; Recency and cadence of tagged releases
Community: 14%; Contributor breadth and bus factor
Issue backlog: 85%; Open issues measured against the size of the audience
Documentation: 84%; README depth, plus a homepage and topics
Licensing: 100%; How freely the code can be adopted

What to know before adopting

No commits in over a year
No tagged releases
Bus factor of 1 — one contributor dominates
Usable in commercial products with attribution.

Details

Created: 5 years ago
Last push: 3 years ago
Last release: none
Release tag: —
Releases in the last year: 0
Default branch: main
Size: 7.0k KB
Owner type: User
Owner location: —
Discovered via: topic:pytorch
First indexed: 3 months ago
Last synced: 5 days ago

Top AI Repos — open-source AI, indexed and scored

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

From the README

Top contributors3 shown

Quality score

What to know before adopting

Categories

Details

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

From the README

Top contributors3 shown

Quality score

What to know before adopting

Categories

Details

Top AI Repos — open-source AI, indexed and scored

Trend

From the README

Top contributors3 shown

Quality score

What to know before adopting

Categories

Details

Trend

From the README

Top contributors3 shown

Quality score

What to know before adopting

Categories

Details