Point code links to LynnReal-AI/ReentryMem
Browse files- README.md +5 -5
- checksums.json +2 -2
README.md
CHANGED
|
@@ -23,7 +23,7 @@ tags:
|
|
| 23 |
|
| 24 |
<p align="center">
|
| 25 |
<a href="https://reentrymem-project.pages.dev/">🌐 Project Page</a> |
|
| 26 |
-
<a href="https://github.com/
|
| 27 |
<a href="https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B">Base Model</a>
|
| 28 |
</p>
|
| 29 |
|
|
@@ -60,10 +60,10 @@ The single-entity checkpoint was fine-tuned using entity-weighted flow matching
|
|
| 60 |
|
| 61 |
## Getting Started
|
| 62 |
|
| 63 |
-
Use the [ReentryMem inference repository](https://github.com/
|
| 64 |
|
| 65 |
```bash
|
| 66 |
-
git clone https://github.com/
|
| 67 |
cd ReentryMem
|
| 68 |
conda create -n reentrymem python=3.11 -y
|
| 69 |
conda activate reentrymem
|
|
@@ -91,7 +91,7 @@ python infer.py \
|
|
| 91 |
--output-dir outputs/demo
|
| 92 |
```
|
| 93 |
|
| 94 |
-
The default uses localized sampling, 50 steps, CFG 5, sigma shift 5, and correction strength 8. Add `--sampler euler` for ordinary sampling or `--save-masks` to save entity-region fields. The output includes the complete video, the generated continuation, and a settings receipt. See the [code documentation](https://github.com/
|
| 95 |
|
| 96 |
## Evaluation
|
| 97 |
|
|
@@ -109,7 +109,7 @@ The protocol separately reports entity reappearance, subject consistency, and te
|
|
| 109 |
|
| 110 |
This release targets single-entity re-entry with the stated observed/future clip lengths and camera controls. Use the custom ReentryMem inference implementation to load the camera-conditioned transformer. GPU memory use depends on the attention backend and runtime settings.
|
| 111 |
|
| 112 |
-
The research results above come from the project's evaluation runs. The standalone package has passed CPU tests, strict full-weight loading, and a small forward check with synthetic text features; full-resolution CUDA generation has not yet been re-run for this package. Validation details accompany the [code release](https://github.com/
|
| 113 |
|
| 114 |
## License and Acknowledgments
|
| 115 |
|
|
|
|
| 23 |
|
| 24 |
<p align="center">
|
| 25 |
<a href="https://reentrymem-project.pages.dev/">🌐 Project Page</a> |
|
| 26 |
+
<a href="https://github.com/LynnReal-AI/ReentryMem">💻 Code</a> |
|
| 27 |
<a href="https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B">Base Model</a>
|
| 28 |
</p>
|
| 29 |
|
|
|
|
| 60 |
|
| 61 |
## Getting Started
|
| 62 |
|
| 63 |
+
Use the [ReentryMem inference repository](https://github.com/LynnReal-AI/ReentryMem) with Python 3.10 or later and CUDA-enabled PyTorch:
|
| 64 |
|
| 65 |
```bash
|
| 66 |
+
git clone https://github.com/LynnReal-AI/ReentryMem.git
|
| 67 |
cd ReentryMem
|
| 68 |
conda create -n reentrymem python=3.11 -y
|
| 69 |
conda activate reentrymem
|
|
|
|
| 91 |
--output-dir outputs/demo
|
| 92 |
```
|
| 93 |
|
| 94 |
+
The default uses localized sampling, 50 steps, CFG 5, sigma shift 5, and correction strength 8. Add `--sampler euler` for ordinary sampling or `--save-masks` to save entity-region fields. The output includes the complete video, the generated continuation, and a settings receipt. See the [code documentation](https://github.com/LynnReal-AI/ReentryMem) for camera inputs and additional options.
|
| 95 |
|
| 96 |
## Evaluation
|
| 97 |
|
|
|
|
| 109 |
|
| 110 |
This release targets single-entity re-entry with the stated observed/future clip lengths and camera controls. Use the custom ReentryMem inference implementation to load the camera-conditioned transformer. GPU memory use depends on the attention backend and runtime settings.
|
| 111 |
|
| 112 |
+
The research results above come from the project's evaluation runs. The standalone package has passed CPU tests, strict full-weight loading, and a small forward check with synthetic text features; full-resolution CUDA generation has not yet been re-run for this package. Validation details accompany the [code release](https://github.com/LynnReal-AI/ReentryMem).
|
| 113 |
|
| 114 |
## License and Acknowledgments
|
| 115 |
|
checksums.json
CHANGED
|
@@ -10,8 +10,8 @@
|
|
| 10 |
"sha256": "d5b013a4561d1dd1b194840ca232dd46293c155ff4e75699c5f13d50e27c5bb8"
|
| 11 |
},
|
| 12 |
"README.md": {
|
| 13 |
-
"bytes":
|
| 14 |
-
"sha256": "
|
| 15 |
},
|
| 16 |
"config.json": {
|
| 17 |
"bytes": 297,
|
|
|
|
| 10 |
"sha256": "d5b013a4561d1dd1b194840ca232dd46293c155ff4e75699c5f13d50e27c5bb8"
|
| 11 |
},
|
| 12 |
"README.md": {
|
| 13 |
+
"bytes": 5827,
|
| 14 |
+
"sha256": "01c2c87ff22f69c4648d4235b1f9304267b66805bce041e76ebaa9c04fb04d20"
|
| 15 |
},
|
| 16 |
"config.json": {
|
| 17 |
"bytes": 297,
|