-
Notifications
You must be signed in to change notification settings - Fork 1.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ONNX Feature already worked. This is a result run on GPU with onnxruntime. I'll preview and plan to new feature updated is TensorRT and C++ Inference code in the next week 🔥🔥🔥 #142
Comments
Very nice work congrats |
Thank you very much for your efforts in making LivePortrait a better project! We will follow up on your thread next week. @aihacker111 |
rụng trứng mất thôi ✌️ |
@aihacker111 Where can I find the Colab notebook? I want to run this on Google Colab. |
@PointerSoftware I'll create notebook for that in this afternoon |
Eagerly waiting for that!, Please remember to create a notebook that supports Gradio, so anyone can use the project with a graphical interface. |
|
@aihacker111 Thanks for your great work!! I would like to know where can I find the C++ inference code. :)) |
@ChunChenLin I'm currently still not implement C++ code , maybe in the next week :) |
@aihacker111 Sure, thanks! Looking forward to it. Thanks again for your great work. :)) |
@aihacker111 Hey, how's everything :) Just wanna know is the C++ inference code available or coming soon? :)) |
@aihacker111 Hi! Thanks for your hard work. |
So during the My Journey to optimize this model run by ONNX GPU and CPU , we got the result like this:
This demo is run on P100 16gb in Kaggle Kernel
This optimiztion will reduce the time speed for run all process of Live-Portrait to 1s/1 frame with 354 frame in total 4 minute 30s
Also support running on Google Colab
Please following my source for run demo: https://github.com/aihacker111/Efficient-Live-Portrait
This source already support for run on CPU and Apple silicon , and CUDA
TensorRT models
I'm already converted to TensorRT and in progress building C++ code for inference and demo. Wait the great TensorRT in the next week
🔥🔥🔥🔥🔥
Have fun with result:
368220873_826368889022136_4472311944594836999_n--368220873_826368889022136_4472311944594836999_n_concat.mp4
368220873_826368889022136_4472311944594836999_n--368220873_826368889022136_4472311944594836999_n.mp4
The text was updated successfully, but these errors were encountered: