Add examples to README.md
Browse files
README.md
CHANGED
|
@@ -31,6 +31,12 @@ This is a **partial NVFP4 quantization** of [Wan2.1-I2V-14B-480P-StepDistill-Cfg
|
|
| 31 |
|
| 32 |
Wan2.1-I2V-14B-480P-StepDistill-CfgDistill-Lightx2v is an advanced image-to-video generation model built upon the Wan2.1-I2V-14B-480P foundation. This approach allows the model to generate videos with significantly fewer inference steps (4 steps) and without classifier-free guidance, substantially reducing video generation time while maintaining quality.
|
| 33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
## Quantization
|
| 35 |
The model weights have been partially quantized to **NVFP4** (NVIDIA Floating Point 4-bit), a quantization format supported on NVIDIA Blackwell architecture GPUs. Only a subset of the model layers have been quantized; the remaining layers are kept at their original precision to preserve output quality.
|
| 36 |
|
|
|
|
| 31 |
|
| 32 |
Wan2.1-I2V-14B-480P-StepDistill-CfgDistill-Lightx2v is an advanced image-to-video generation model built upon the Wan2.1-I2V-14B-480P foundation. This approach allows the model to generate videos with significantly fewer inference steps (4 steps) and without classifier-free guidance, substantially reducing video generation time while maintaining quality.
|
| 33 |
|
| 34 |
+
<div style="display: flex; align-items: center; gap: 16px;">
|
| 35 |
+
<img src="assets/wan21_input_cat.png" width="25%"/>
|
| 36 |
+
<span style="font-size: 2em;">➡️</span>
|
| 37 |
+
<video src="https://huggingface.co/InsecureErasure/Wan2.1-I2V-14B-480P-StepDistill-CfgDistill-Lightx2v-NVFP4/resolve/main/assets/wan21_output_cat.mp4" width="25%" controls></video>
|
| 38 |
+
</div>
|
| 39 |
+
|
| 40 |
## Quantization
|
| 41 |
The model weights have been partially quantized to **NVFP4** (NVIDIA Floating Point 4-bit), a quantization format supported on NVIDIA Blackwell architecture GPUs. Only a subset of the model layers have been quantized; the remaining layers are kept at their original precision to preserve output quality.
|
| 42 |
|