Update README.md
Browse files
README.md
CHANGED
|
@@ -37,7 +37,7 @@ python setup_vsa.py install
|
|
| 37 |
num_gpus=1
|
| 38 |
export FASTVIDEO_ATTENTION_BACKEND=VIDEO_SPARSE_ATTN
|
| 39 |
# change model path to local dir if you want to inference using your checkpoint
|
| 40 |
-
export MODEL_BASE=
|
| 41 |
# export MODEL_BASE=hunyuanvideo-community/HunyuanVideo
|
| 42 |
fastvideo generate \
|
| 43 |
--model-path $MODEL_BASE \
|
|
@@ -55,7 +55,7 @@ fastvideo generate \
|
|
| 55 |
--prompt "A beautiful woman in a red dress walking down a street" \
|
| 56 |
--negative-prompt "Bright tones, overexposed, static, blurred details, subtitles, style, works, paintings, images, static, overall gray, worst quality, low quality, JPEG compression residue, ugly, incomplete, extra fingers, poorly drawn hands, poorly drawn faces, deformed, disfigured, misshapen limbs, fused fingers, still picture, messy background, three legs, many people in the background, walking backwards" \
|
| 57 |
--seed 1024 \
|
| 58 |
-
--output-path
|
| 59 |
```
|
| 60 |
- Try it out on **FastVideo** — we support a wide range of GPUs from **H100** to **4090**
|
| 61 |
- We use [FastVideo 720P Synthetic Wan dataset](https://huggingface.co/datasets/FastVideo/Wan-Syn_77x768x1280_250k) for training.
|
|
|
|
| 37 |
num_gpus=1
|
| 38 |
export FASTVIDEO_ATTENTION_BACKEND=VIDEO_SPARSE_ATTN
|
| 39 |
# change model path to local dir if you want to inference using your checkpoint
|
| 40 |
+
export MODEL_BASE=FastVideo/Wan2.1-VSA-T2V-14B-720P-Diffusers
|
| 41 |
# export MODEL_BASE=hunyuanvideo-community/HunyuanVideo
|
| 42 |
fastvideo generate \
|
| 43 |
--model-path $MODEL_BASE \
|
|
|
|
| 55 |
--prompt "A beautiful woman in a red dress walking down a street" \
|
| 56 |
--negative-prompt "Bright tones, overexposed, static, blurred details, subtitles, style, works, paintings, images, static, overall gray, worst quality, low quality, JPEG compression residue, ugly, incomplete, extra fingers, poorly drawn hands, poorly drawn faces, deformed, disfigured, misshapen limbs, fused fingers, still picture, messy background, three legs, many people in the background, walking backwards" \
|
| 57 |
--seed 1024 \
|
| 58 |
+
--output-path VSA-DMD/sparsity_0.9/
|
| 59 |
```
|
| 60 |
- Try it out on **FastVideo** — we support a wide range of GPUs from **H100** to **4090**
|
| 61 |
- We use [FastVideo 720P Synthetic Wan dataset](https://huggingface.co/datasets/FastVideo/Wan-Syn_77x768x1280_250k) for training.
|