fq
fq
> > > > > As I understand, the papers 1 and 2 (mentioned in paragraph 2 of sec 4.1) are used to extract the face landmarks. Did you just...
> Check the code and the citing paper. for any camera model, we must know the depth of every z value of 2Dshoulder. from the paper, billboard model like a...
> > I recommend using the new state-of-the-art [DECA](https://github.com/YadiraF/DECA) as the reconstruction model, which requires an easier environment installation. > > Yeah, I ended up using DECA. Many thx! is...
> > When I run the test script: > # get reconstruction results of your custom images > > python test.py --name= --epoch=20 --img_folder= I faced this problem? Please, any...
let me join it. just for leraning.
可以参考这个： [SadTalkerTriton](https://github.com/foocker/SadTalkerTriton) 25 fps , 84 frames, cost 2.4s
> > 可以，尽量使用多块显卡，而且需要把预加载相关的代码从pipeline中摘出来 > > 大佬，怎么使用多块显卡呀？需要改哪里的代码呀? 参考下https://github.com/triton-inference-server/python_backend/tree/main/examples/instance_kind 或这里的一些文档。
> Hello, I wonder if AITemplate now supports LoRA for the Diffusion model (load LoRAa before or after compile) ? can you clone the code an do a test?
> > > Hello, I wonder if AITemplate now supports LoRA for the Diffusion model (load LoRAa before or after compile) ? > > > > > > can you...
is there only answer? i have download the file from the url, but not the describing in readme