Commit
•
9a06fc4
1
Parent(s):
8b318ef
initial commit
Browse files- .gitignore +3 -1
- README.md +20 -4
.gitignore
CHANGED
@@ -5,4 +5,6 @@
|
|
5 |
|
6 |
weights*
|
7 |
models
|
8 |
-
sd-v2*
|
|
|
|
|
|
5 |
|
6 |
weights*
|
7 |
models
|
8 |
+
sd-v2*
|
9 |
+
|
10 |
+
venv/
|
README.md
CHANGED
@@ -1,13 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
# MVDream-diffusers
|
2 |
|
3 |
A **unified** diffusers implementation of [MVDream](https://github.com/bytedance/MVDream) and [ImageDream](https://github.com/bytedance/ImageDream).
|
4 |
|
5 |
We provide converted `fp16` weights on huggingface:
|
6 |
-
* [MVDream](https://huggingface.co/ashawkey/mvdream-sd2.1-diffusers)
|
7 |
-
* [ImageDream](https://huggingface.co/ashawkey/imagedream-ipmv-diffusers)
|
8 |
|
|
|
|
|
9 |
|
10 |
### Install
|
|
|
11 |
```bash
|
12 |
# dependency
|
13 |
pip install -r requirements.txt
|
@@ -27,6 +41,7 @@ python run_imagedream.py data/anya_rgba.png
|
|
27 |
### Convert weights
|
28 |
|
29 |
MVDream:
|
|
|
30 |
```bash
|
31 |
# download original ckpt (we only support the SD 2.1 version)
|
32 |
mkdir models
|
@@ -40,6 +55,7 @@ python convert_mvdream_to_diffusers.py --checkpoint_path models/sd-v2.1-base-4vi
|
|
40 |
```
|
41 |
|
42 |
ImageDream:
|
|
|
43 |
```bash
|
44 |
# download original ckpt (we only support the pixel-controller version)
|
45 |
cd models
|
@@ -53,7 +69,7 @@ python convert_mvdream_to_diffusers.py --checkpoint_path models/sd-v2.1-base-4vi
|
|
53 |
|
54 |
### Acknowledgement
|
55 |
|
56 |
-
|
57 |
```bibtex
|
58 |
@article{shi2023MVDream,
|
59 |
author = {Shi, Yichun and Wang, Peng and Ye, Jianglong and Mai, Long and Li, Kejie and Yang, Xiao},
|
@@ -68,4 +84,4 @@ python convert_mvdream_to_diffusers.py --checkpoint_path models/sd-v2.1-base-4vi
|
|
68 |
year={2023}
|
69 |
}
|
70 |
```
|
71 |
-
|
|
|
1 |
+
---
|
2 |
+
license: openrail
|
3 |
+
pipeline_tag: image-to-3d
|
4 |
+
---
|
5 |
+
|
6 |
+
This is a duplicate of [ashawkey/imagedream-ipmv-diffusers](https://huggingface.co/ashawkey/imagedream-ipmv-diffusers).
|
7 |
+
|
8 |
+
It is hosted here for the purpose of persistence and reproducibility for the ML for 3D course.
|
9 |
+
|
10 |
+
Original model card below.
|
11 |
+
|
12 |
+
---
|
13 |
+
|
14 |
# MVDream-diffusers
|
15 |
|
16 |
A **unified** diffusers implementation of [MVDream](https://github.com/bytedance/MVDream) and [ImageDream](https://github.com/bytedance/ImageDream).
|
17 |
|
18 |
We provide converted `fp16` weights on huggingface:
|
|
|
|
|
19 |
|
20 |
+
- [MVDream](https://huggingface.co/ashawkey/mvdream-sd2.1-diffusers)
|
21 |
+
- [ImageDream](https://huggingface.co/ashawkey/imagedream-ipmv-diffusers)
|
22 |
|
23 |
### Install
|
24 |
+
|
25 |
```bash
|
26 |
# dependency
|
27 |
pip install -r requirements.txt
|
|
|
41 |
### Convert weights
|
42 |
|
43 |
MVDream:
|
44 |
+
|
45 |
```bash
|
46 |
# download original ckpt (we only support the SD 2.1 version)
|
47 |
mkdir models
|
|
|
55 |
```
|
56 |
|
57 |
ImageDream:
|
58 |
+
|
59 |
```bash
|
60 |
# download original ckpt (we only support the pixel-controller version)
|
61 |
cd models
|
|
|
69 |
|
70 |
### Acknowledgement
|
71 |
|
72 |
+
- The original papers:
|
73 |
```bibtex
|
74 |
@article{shi2023MVDream,
|
75 |
author = {Shi, Yichun and Wang, Peng and Ye, Jianglong and Mai, Long and Li, Kejie and Yang, Xiao},
|
|
|
84 |
year={2023}
|
85 |
}
|
86 |
```
|
87 |
+
- This codebase is modified from [mvdream-hf](https://github.com/KokeCacao/mvdream-hf).
|