Fix packaging with using sdist (#2)

* sdist packaging

* Bump version

.github/workflows/release.yml

@@ -48,7 +48,6 @@ jobs:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           changelog: false
 
-
       - name: Publish package distributions to PyPI
         uses: pypa/gh-action-pypi-publish@release/v1
 

README.md

@@ -35,7 +35,7 @@ Follow this step, if you are already using Vanilla's [graphdeco-inria 3D Gaussia
 in your project and you want to replace forward/backward kernels
 with deepsense.ai open-source kernels.
 
-In your environment, simply install:
+Make sure CUDA compiler is installed in your environment and simply install:
 ```bash
 pip install ds-splat
 ```
@@ -212,4 +212,4 @@ this regard. Higher PSNR is better.
 
 📥 [Download](https://drive.google.com/file/d/1MADQzb6onTV6JBJQqTx8H9bN6It4Vcdj/view?usp=sharing) more benchmark plots from GDrive.
 
-![deepsense/gsplat PSNR to Vanilla. Bicycle Scene.](assets/psnr_to_vanilla/bicycle_1280x720_psnr.png)
+![deepsense/gsplat PSNR to Vanilla. Bicycle Scene.](assets/psnr_to_vanilla/bicycle_1280x720_psnr.png)

pyproject.toml

@@ -34,7 +34,7 @@ allow_zero_version = true
 no_git_verify = false
 tag_format = "v{version}"
 version_variables = ["setup.py:__version__"]
-build_command = "pip install build && python -m build"
+build_command = "pip install build && python -m build --sdist"
 
 [tool.semantic_release.branches.main]
 match = "(main|master)"

setup.py

@@ -3,7 +3,7 @@
 from setuptools import find_packages, setup
 from torch.utils.cpp_extension import BuildExtension, CUDAExtension
 
-__version__ = "0.0.0"
+__version__ = "0.0.1"
 
 def get_cuda_extension():
     extra_compile_args = {"cxx": ["-O3"], "nvcc": ["-O3", "--use_fast_math"]}