Commit
•
6b31a45
1
Parent(s):
2b240fa
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
{}
|
| 3 |
+
---
|
| 4 |
+
This repository contains the dataset of model weights and utility files for [Interpreting the Weight Space of Customized Diffusion Models](https://snap-research.github.io/weights2weights/). The code can be found at https://github.com/snap-research/weights2weights.
|
| 5 |
+
|
| 6 |
+
# Readout Head Weights
|
| 7 |
+
The `weights/` folder contains the pre-trained weights of the readout heads, named according to the following convention:
|
| 8 |
+
```
|
| 9 |
+
readout_<base-model>_<task-type>_<head-type>
|
| 10 |
+
```
|
| 11 |
+
|
| 12 |
+
**Spatially Aligned Control**
|
| 13 |
+
- `readout_sdxl_spatial_pose.pt`
|
| 14 |
+
- `readout_sdv15_spatial_pose.pt`
|
| 15 |
+
- Readout head trained with [OpenPose](https://arxiv.org/abs/1812.08008) pose skeletons as supervision on [PascalVOC](http://host.robots.ox.ac.uk/pascal/VOC) images, filtered only to those containing people.
|
| 16 |
+
- `readout_sdxl_spatial_depth.pt`
|
| 17 |
+
- `readout_sdv15_spatial_depth.pt`
|
| 18 |
+
- Readout head trained with [MiDaS](https://arxiv.org/abs/1907.01341) depth maps as supervision on [PascalVOC](http://host.robots.ox.ac.uk/pascal/VOC) images.
|
| 19 |
+
- `readout_sdxl_spatial_edge.pt`
|
| 20 |
+
- `readout_sdv15_spatial_edge.pt`
|
| 21 |
+
- Readout head trained with [HED](https://arxiv.org/abs/1504.06375) edge detections as supervision on [PascalVOC](http://host.robots.ox.ac.uk/pascal/VOC) images.
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
**Drag-Based Manipulation**
|
| 25 |
+
- `readout_sdxl_drag_correspondence.pt`
|
| 26 |
+
- `readout_sdv15_drag_correspondence.pt`
|
| 27 |
+
- Readout head trained with a contrastive loss with [CoTracker](https://co-tracker.github.io) point tracks across pairs of [DAVIS](https://davischallenge.org) video frames.
|
| 28 |
+
- `readout_sdxl_drag_appearance.pt`
|
| 29 |
+
- `readout_sdv15_drag_appearance.pt`
|
| 30 |
+
- Readout head trained with a triplet loss with real frames as positives and [SDEdit-ed](https://sde-image-editing.github.io) frames as negatives derived from [DAVIS](https://davischallenge.org) videos.
|