detailed libspconv example

README.md

@@ -74,6 +74,7 @@ Update Spconv: you **MUST UNINSTALL** all spconv/cumm/spconv-cuxxx/cumm-cuxxx fi
 * nvrtc support: kernel in old GPUs will be compiled in runtime.
 * [libspconv](docs/PURE_CPP_BUILD.md): pure c++ build of all spconv ops. see [example](example/libspconv/run_build.sh)
 * tf32 kernels, faster fp32 training, disabled by default. set ```import spconv as spconv_core; spconv_core.constants.SPCONV_ALLOW_TF32 = True``` to enable them.
+* all weights are KRSC layout, some old model can't be loaded anymore.
 
 
 ## Spconv 2.1 vs Spconv 1.x

docs/PURE_CPP_BUILD.md

@@ -2,6 +2,8 @@
 
 Spconv provide a way to generate sources to build a C++ library with all op needed for inference and train.
 
+The generated libspconv has been deployed to tensorrt and run in real automatic driving vehicles.
+
 ## Steps
 
 1. Install spconv, Install cumm, Install cumm cmake, or add it as a sub directory

example/libspconv/README.md

+# libspconv
+
+libspconv + pybindings = "core_cc.so" in spconv python package.
+
 ## libspconv Example
 
 run ```run_build.sh``` to get ```libspconv.so```.

example/libspconv/run_build.sh

@@ -4,7 +4,7 @@ SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
 
 git clone https://github.com/FindDefinition/cumm.git $SCRIPT_DIR/cumm
 
-export CUMM_CUDA_VERSION=11.4 # cuda version, required
+export CUMM_CUDA_VERSION=11.4 # cuda version, required but only used for flag selection when build libspconv.
 export CUMM_DISABLE_JIT=1
 export SPCONV_DISABLE_JIT=1
 export CUMM_INCLUDE_PATH="\${CUMM_INCLUDE_PATH}" # if you use cumm as a subdirectory, you need this to find cumm includes.
@@ -16,4 +16,4 @@ python -m spconv.gencode --include=$SCRIPT_DIR/spconv/include --src=$SCRIPT_DIR/
 mkdir -p $SCRIPT_DIR/build
 cd $SCRIPT_DIR/build
 cmake ..
-cmake --build $SCRIPT_DIR/build --config Release -j 8 --verbose
+cmake --build $SCRIPT_DIR/build --config Release -j 8 # --verbose