Closed hardikdava closed 6 months ago
Hi, thanks for your advice. I will take a look.
@yzslab any update on this issue? Thanks
@yzslab any update on this issue? Thanks
Sorry for keeping you waiting. I am busy in doing other experiments, so this has not been scheduled yet.
nerfstudio-project/gsplat has been integrated.
@yzslab I dont see it in main branch. Is it implemented in other branch?
@yzslab I dont see it in main branch. Is it implemented in other branch?
I didn't add it to submodules.
Install it via pip pip install gsplat==0.1.8
, then use with configs/gsplat.yaml
:
python main.py fit \
--config configs/gsplat.yaml \
...
@yzslab I dont see it in main branch. Is it implemented in other branch?
I didn't add it to submodules. Install it via pip
pip install gsplat==0.1.8
, then use withconfigs/gsplat.yaml
:python main.py fit \ --config configs/gsplat.yaml \ ...
Besides, you need to make sure that nvcc
is executable, or gsplat will be disabled.
gsplat should work if you can see output from command which nvcc
:
$ which nvcc
/usr/local/cuda-11.8/bin/nvcc
Thanks @yzslab for great work.
@yzslab First of all great work, you have released the core codebase as MIT since submodule is private licensed. What do you think about using gsplat instead of inria based rendering submodule?