Update README.md

README.md

@@ -6,9 +6,9 @@ base_model:
 ---
 # ComfyUI-SeedVR2_VideoUpscaler
 
-A Non official custom nodes for ComfyUI that enables Upscale Video generation using [SeedVR2](https://github.com/ByteDance-Seed/SeedVR).
+[![View Code](https://img.shields.io/badge/📂_View_Code-GitHub-181717?style=for-the-badge&logo=github)](https://github.com/numz/ComfyUI-SeedVR2_VideoUpscaler)
 
-[![GitHub](https://img.shields.io/badge/GitHub-Code-black?style=for-the-badge&logo=github)](https://github.com/numz/ComfyUI-SeedVR2_VideoUpscaler)
+A Non official custom nodes for ComfyUI that enables Upscale Video generation using [SeedVR2](https://github.com/ByteDance-Seed/SeedVR).
 
 <video width="700px" controls>
   <source src="https://github.com/user-attachments/assets/8fbd6c1f-4246-4dbe-8819-4e684490c5f2" type="video/mp4">
@@ -60,6 +60,13 @@ python_embeded\python.exe -m pip install -r ComfyUI-SeedVR2_VideoUpscaler/requir
 python_embeded\python.exe -m pip install -r flash_attn
 ```
 
+3. Models
+
+   Will be automtically download into :
+   `models/SEEDVR2`
+
+   or can be found here ([MODELS](https://huggingface.co/numz/SeedVR2_comfyUI/tree/main))
+
 ## Usage
 
 1. In ComfyUI, locate the **SeedVR2 Video Upscaler** node in the node menu.
@@ -79,10 +86,14 @@ python_embeded\python.exe -m pip install -r flash_attn
 
 1. **NVIDIA H100 93GB VRAM**
 
+3B or 7B spike to 90+GB VRAM!! but fast!!
+
 - 3B Model, 97 images, from 512x768 to 1280x1920, batch_size=50 => Prompt executed in 338.63 seconds
 - 3B Model, 97 images, from 512x768 to 1280x1920, batch_size=10 => Prompt executed in 540.22 seconds
 - 3B Model, 97 images, from 512x768 to 720x1080, batch_size=10 => Prompt executed in 183.64 seconds
-- 7B Model, 50images, 512x768 to 1080x1620, batch_size=50, executed in 166.89 seconds
+- 7B Model, 50 images, 512x768 to 1080x1620, batch_size=50, Prompt executed in 166.89 seconds
+- 7B Model, 97 images, 512x768 to 1080x1620, batch_size=97, Prompt executed in 146.72 seconds
+- 7B Model, 200 images, 512x768 to 1080x1620, batch_size=200, Prompt executed in 266.14 seconds
 
 2. **NVIDIA RTX4090 24GB VRAM**
 
@@ -102,3 +113,4 @@ python_embeded\python.exe -m pip install -r flash_attn
 
 - The code in this repository is released under the MIT license as found in the [LICENSE file](LICENSE).
 
+