<!--
Copyright 2021 Yan Yan
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
-->
[pypi-ver-cpu]: https://img.shields.io/pypi/v/spconv
[pypi-ver-114]: https://img.shields.io/pypi/v/spconv-cu114
[pypi-ver-111]: https://img.shields.io/pypi/v/spconv-cu111
[pypi-ver-117]: https://img.shields.io/pypi/v/spconv-cu117
[pypi-ver-116]: https://img.shields.io/pypi/v/spconv-cu116
[pypi-ver-118]: https://img.shields.io/pypi/v/spconv-cu118
[pypi-ver-113]: https://img.shields.io/pypi/v/spconv-cu113
[pypi-ver-120]: https://img.shields.io/pypi/v/spconv-cu120
[pypi-ver-102]: https://img.shields.io/pypi/v/spconv-cu102
[pypi-url-102]: https://pypi.org/project/spconv-cu102/
[pypi-download-102]: https://img.shields.io/pypi/dm/spconv-cu102
[pypi-url-111]: https://pypi.org/project/spconv-cu111/
[pypi-download-111]: https://img.shields.io/pypi/dm/spconv-cu111
[pypi-url-113]: https://pypi.org/project/spconv-cu113/
[pypi-download-113]: https://img.shields.io/pypi/dm/spconv-cu113
[pypi-url-114]: https://pypi.org/project/spconv-cu114/
[pypi-download-114]: https://img.shields.io/pypi/dm/spconv-cu114
[pypi-url-117]: https://pypi.org/project/spconv-cu117/
[pypi-download-117]: https://img.shields.io/pypi/dm/spconv-cu117
[pypi-url-120]: https://pypi.org/project/spconv-cu120/
[pypi-download-120]: https://img.shields.io/pypi/dm/spconv-cu120
[pypi-url-cpu]: https://pypi.org/project/spconv/
[pypi-download-cpu]: https://img.shields.io/pypi/dm/spconv
[pypi-url-118]: https://pypi.org/project/spconv-cu118/
[pypi-download-118]: https://img.shields.io/pypi/dm/spconv-cu118
[pypi-url-116]: https://pypi.org/project/spconv-cu116/
[pypi-download-116]: https://img.shields.io/pypi/dm/spconv-cu116
# SpConv: Spatially Sparse Convolution Library
[](https://github.com/traveller59/spconv/actions?query=workflow%3Abuild)
| | PyPI | Install |Downloads |
| -------------- |:---------------------:| ---------------------:| ---------------------:|
| CPU (Linux Only) | [![PyPI Version][pypi-ver-cpu]][pypi-url-cpu] | ```pip install spconv``` | [![pypi monthly download][pypi-download-cpu]][pypi-url-cpu] |
| CUDA 10.2 | [![PyPI Version][pypi-ver-102]][pypi-url-102] | ```pip install spconv-cu102```| [![pypi monthly download][pypi-download-102]][pypi-url-102]|
| CUDA 11.3 | [![PyPI Version][pypi-ver-113]][pypi-url-113] | ```pip install spconv-cu113```| [![pypi monthly download][pypi-download-113]][pypi-url-113]|
| CUDA 11.4 | [![PyPI Version][pypi-ver-114]][pypi-url-114] | ```pip install spconv-cu114```| [![pypi monthly download][pypi-download-114]][pypi-url-114]|
| CUDA 11.6 | [![PyPI Version][pypi-ver-116]][pypi-url-116] | ```pip install spconv-cu116```| [![pypi monthly download][pypi-download-116]][pypi-url-116]|
| CUDA 11.7 | [![PyPI Version][pypi-ver-117]][pypi-url-117] | ```pip install spconv-cu117```| [![pypi monthly download][pypi-download-117]][pypi-url-117]|
| CUDA 11.8 | [![PyPI Version][pypi-ver-118]][pypi-url-118] | ```pip install spconv-cu118```| [![pypi monthly download][pypi-download-118]][pypi-url-118]|
| CUDA 12.0 | [![PyPI Version][pypi-ver-120]][pypi-url-120] | ```pip install spconv-cu120```| [![pypi monthly download][pypi-download-120]][pypi-url-120]|
```spconv``` is a project that provide heavily-optimized sparse convolution implementation with tensor core support. check [benchmark](docs/BENCHMARK.md) to see how fast spconv 2.x runs.
[Spconv 1.x code](https://github.com/traveller59/spconv/tree/v1.2.1). We won't provide any support for spconv 1.x since it's deprecated. use spconv 2.x if possible. <!--remove this message in spconv 2.2-->
Check [spconv 2.x algorithm introduction](docs/spconv2_algo.pdf) to understand sparse convolution algorithm in spconv 2.x!
## WARNING
Use spconv >= cu114 if possible. cuda 11.4 can compile greatly faster kernel in some situation.
Update Spconv: you **MUST UNINSTALL** all spconv/cumm/spconv-cuxxx/cumm-cuxxx first, use ```pip list | grep spconv``` and ```pip list | grep cumm``` to check all installed package. then use pip to install new spconv.
## NEWS
* spconv 2.3: int8 quantization support. see docs and examples for more details.
* spconv 2.2: ampere feature support (by [EvernightAurora](https://github.com/EvernightAurora)), pure c++ code generation, nvrtc, drop python 3.6
## Spconv 2.2 vs Spconv 2.1
* faster fp16 conv kernels (~5-30%) in ampere GPUs (tested in RTX 3090)
* greatly faster int8 conv kernels (~1.2x-2.7x) in ampere GPUs (tested in RTX 3090)
* drop python 3.6 support
* 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
* spconv now can be installed by **pip**. see install section in readme for more details. Users don't need to build manually anymore!
* Microsoft Windows support (only windows 10 has been tested).
* fp32 (not tf32) training/inference speed is increased (+50~80%)
* fp16 training/inference speed is greatly increased when your layer support tensor core (channel size must be multiple of 8).
* int8 op is ready, but we still need some time to figure out how to run int8 in pytorch.
* [doesn't depend on pytorch binary](docs/FAQ.md#What-does-no-dependency-on-pytorch-mean), but you may need at least pytorch >= 1.5.0 to run spconv 2.x.
* since spconv 2.x doesn't depend on pytorch binary (never in future), it's impossible to support torch.jit/libtorch inference.
## Usage
Firstly you need to use ```import spconv.pytorch as spconv``` in spconv 2.x.
Then see [this](docs/USAGE.md).
Don't forget to check [performance guide](docs/PERFORMANCE_GUIDE.md).
### Common Solution for Some Bugs
see [common problems](docs/COMMON_PROBLEMS.md).
## Install
You need to install python >= 3.7 first to use spconv 2.x.
You need to install CUDA toolkit first before using prebuilt binaries or build from source.
You need at least CUDA 11.0 to build and run spconv 2.x. We won't offer any support for CUDA < 11.0.
### Prebuilt
We offer python 3.7-3.11 and cuda 10.2/11.3/11.4/11.7/12.0 prebuilt binaries for linux (manylinux).
We offer python 3.7-3.11 and cuda 10.2/11.4/11.7/12.0 prebuilt binaries for windows 10/11.
For Linux users, you need to install pip >= 20.3 first to install prebuilt.
**WARNING**: spconv-cu117 may require CUDA Driver >= 515.
```pip install spconv``` for CPU only (**Linux Only**). you should only use this for debug usage, the performance isn't optimized due to manylinux limit (no omp support).
```pip install spconv-cu102``` for CUDA 10.2
```pip install spconv-cu113``` for CUDA 11.3 (**Linux Only**)
```pip install spconv-cu114``` for CUDA 11.4
```pip install spconv-cu117``` for CUDA 11.7
```pip install spconv-cu120``` for CUDA 12.0
**NOTE** It's safe to have different **minor** cuda version between system and conda (pytorch) in **CUDA >= 11.0** because of [CUDA Minor Version Compatibility](https://docs.nvidia.com/deploy/cuda-compatibility/#minor-version-compatibility). For example, you can use spconv-cu114 with anaconda version of pytorch cuda 11.1 in a OS with CUDA 11.2 installed.
**NOTE** In Linux, you can install spconv-cuxxx without install CUDA to system! only suitable NVIDIA driver is required. for CUDA 11, we need driver >= 450.82. You may need newer driver if you use newer CUDA. for cuda 11.8, you need to have driver >= 520 installed.
#### Prebuilt GPU Support Matrix
See [this page](https://arnon.dk/matching-sm-architectures-arch-and-gencode-for-various-nvidia-cards/) to check supported GPU names by arch.
If you use a GPU architecture that isn't compiled in prebuilt, spconv will use NVRTC to compile a slightly slower kernel.
| CUDA version | GPU Arch List |
| -------------- |:---------------------:|
| 11.1~11.7 | 52,60,61,70,75,80,86 |
| 11.8+ | 60,70,75,80,86,89,90 |
### Build from source for development (JIT, recommend)
The c++ code will be built automatically when you change c++ code in project.
For NVIDIA Embedded Platforms, you need to specify cuda arch before build: ```export CUMM_CUDA_ARCH_LIST="7.2"``` for xavier, ```export CUMM_CUDA_ARCH_LIST="6.2"``` for TX2, ```export CUMM_CUDA_ARCH_LIST="8.7"``` for orin.
You need to remove ```cumm``` in ```requires``` section in pyproject.toml after install editable ```cumm``` and before install spconv due to pyproject limit (can't find editable installed ```cumm```).
You need to ensure ```pip list | grep spconv``` and ```pip list | grep cumm``` show nothing before install editable spconv/cumm.
#### Linux
0. uninstall spconv and cumm installed by pip
1. install build-essential, install CUDA
2. ```git clone https://github.com/FindDefinition/cumm```, ```cd ./cumm```, ```pip install -e .```
3. ```git clone https://github.com/traveller59/spconv```, ```cd ./spconv```, ```pip install -e .```
4. in python, ```import spconv``` and wait for build finish.
#### Windows
0. uninstall spconv and cumm installed by pip
1. install visual studio 2019 or newer. make sure C++ development component is installed. install CUDA
2. set [powershell script execution policy](https://docs.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_execution_policies?view=powershell-7.1)
3. start a new powershell, run ```tools/msvc_setup.ps1```
4. ```git clone https://github.com/FindDefinition/cumm```, ```cd ./cumm```, ```pip install -e .```
5. ```git clone https://github.com/traveller59/spconv```, ```cd ./spconv```, ```pip install -e .```
6. in python, ```import spconv``` and wait for build finish.
### Build wheel from source (not recommend, this is done in CI.)
You need to rebuild ```cumm``` first if you are build along a CUDA version that not provided in prebuilts.
#### Linux
1. install build-essential, install CUDA
2. run ```export SPCONV_DISABLE_JIT="1"```
3. run ```pip install pccm cumm wheel```
4. run ```python setup.py bdist_wheel```+```pip install dists/xxx.whl```
#### Windows
1. install visual studio 2019 or newer. make sure C++ development component is installed. install CUDA
2. set [powershell script execution policy](https://docs.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_execution_policies?view=powershell-7.1)
3. start a new powershell, run ```tools/msvc_setup.ps1```
4. run ```$Env:SPCONV_DISABLE_JIT = "1"```
5. run ```pip install pccm cumm wheel```
6. run ```python setup.py bdist_wheel```+```pip install dists/xxx.whl```
## Citation
If you find this project useful in your research, please consider cite:
```latex
@misc{spconv2022,
title={Spconv: Spatially Sparse Convolution Library},
author={Spconv Contributors},
howpublished = {\url{https://github.com/traveller59/spconv}},
year={2022}
}
```
## Contributers
* [EvernightAurora](https://github.com/EvernightAurora): add ampere feature.
## Note
The work is done when the author is an employee at [Tusimple](https://www.tusimple.com/).
## LICENSE
Apache 2.0
Raw data
{
"_id": null,
"home_page": "https://github.com/traveller59/spconv",
"name": "spconv-cu102",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "",
"author": "Yan Yan",
"author_email": "yanyan.sub@outlook.com",
"download_url": "",
"platform": null,
"description": "\n<!--\n Copyright 2021 Yan Yan\n \n Licensed under the Apache License, Version 2.0 (the \"License\");\n you may not use this file except in compliance with the License.\n You may obtain a copy of the License at\n \n http://www.apache.org/licenses/LICENSE-2.0\n \n Unless required by applicable law or agreed to in writing, software\n distributed under the License is distributed on an \"AS IS\" BASIS,\n WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n See the License for the specific language governing permissions and\n limitations under the License.\n-->\n[pypi-ver-cpu]: https://img.shields.io/pypi/v/spconv\n[pypi-ver-114]: https://img.shields.io/pypi/v/spconv-cu114\n[pypi-ver-111]: https://img.shields.io/pypi/v/spconv-cu111\n[pypi-ver-117]: https://img.shields.io/pypi/v/spconv-cu117\n[pypi-ver-116]: https://img.shields.io/pypi/v/spconv-cu116\n[pypi-ver-118]: https://img.shields.io/pypi/v/spconv-cu118\n\n[pypi-ver-113]: https://img.shields.io/pypi/v/spconv-cu113\n[pypi-ver-120]: https://img.shields.io/pypi/v/spconv-cu120\n[pypi-ver-102]: https://img.shields.io/pypi/v/spconv-cu102\n\n[pypi-url-102]: https://pypi.org/project/spconv-cu102/\n[pypi-download-102]: https://img.shields.io/pypi/dm/spconv-cu102\n[pypi-url-111]: https://pypi.org/project/spconv-cu111/\n[pypi-download-111]: https://img.shields.io/pypi/dm/spconv-cu111\n[pypi-url-113]: https://pypi.org/project/spconv-cu113/\n[pypi-download-113]: https://img.shields.io/pypi/dm/spconv-cu113\n[pypi-url-114]: https://pypi.org/project/spconv-cu114/\n[pypi-download-114]: https://img.shields.io/pypi/dm/spconv-cu114\n[pypi-url-117]: https://pypi.org/project/spconv-cu117/\n[pypi-download-117]: https://img.shields.io/pypi/dm/spconv-cu117\n[pypi-url-120]: https://pypi.org/project/spconv-cu120/\n[pypi-download-120]: https://img.shields.io/pypi/dm/spconv-cu120\n[pypi-url-cpu]: https://pypi.org/project/spconv/\n[pypi-download-cpu]: https://img.shields.io/pypi/dm/spconv\n[pypi-url-118]: https://pypi.org/project/spconv-cu118/\n[pypi-download-118]: https://img.shields.io/pypi/dm/spconv-cu118\n\n[pypi-url-116]: https://pypi.org/project/spconv-cu116/\n[pypi-download-116]: https://img.shields.io/pypi/dm/spconv-cu116\n\n# SpConv: Spatially Sparse Convolution Library\n[](https://github.com/traveller59/spconv/actions?query=workflow%3Abuild) \n\n\n\n| | PyPI | Install |Downloads |\n| -------------- |:---------------------:| ---------------------:| ---------------------:| \n| CPU (Linux Only) | [![PyPI Version][pypi-ver-cpu]][pypi-url-cpu] | ```pip install spconv``` | [![pypi monthly download][pypi-download-cpu]][pypi-url-cpu] | \n| CUDA 10.2 | [![PyPI Version][pypi-ver-102]][pypi-url-102] | ```pip install spconv-cu102```| [![pypi monthly download][pypi-download-102]][pypi-url-102]| \n| CUDA 11.3 | [![PyPI Version][pypi-ver-113]][pypi-url-113] | ```pip install spconv-cu113```| [![pypi monthly download][pypi-download-113]][pypi-url-113]| \n| CUDA 11.4 | [![PyPI Version][pypi-ver-114]][pypi-url-114] | ```pip install spconv-cu114```| [![pypi monthly download][pypi-download-114]][pypi-url-114]|\n| CUDA 11.6 | [![PyPI Version][pypi-ver-116]][pypi-url-116] | ```pip install spconv-cu116```| [![pypi monthly download][pypi-download-116]][pypi-url-116]|\n| CUDA 11.7 | [![PyPI Version][pypi-ver-117]][pypi-url-117] | ```pip install spconv-cu117```| [![pypi monthly download][pypi-download-117]][pypi-url-117]| \n| CUDA 11.8 | [![PyPI Version][pypi-ver-118]][pypi-url-118] | ```pip install spconv-cu118```| [![pypi monthly download][pypi-download-118]][pypi-url-118]| \n| CUDA 12.0 | [![PyPI Version][pypi-ver-120]][pypi-url-120] | ```pip install spconv-cu120```| [![pypi monthly download][pypi-download-120]][pypi-url-120]| \n\n\n```spconv``` is a project that provide heavily-optimized sparse convolution implementation with tensor core support. check [benchmark](docs/BENCHMARK.md) to see how fast spconv 2.x runs.\n\n[Spconv 1.x code](https://github.com/traveller59/spconv/tree/v1.2.1). We won't provide any support for spconv 1.x since it's deprecated. use spconv 2.x if possible. <!--remove this message in spconv 2.2-->\n\nCheck [spconv 2.x algorithm introduction](docs/spconv2_algo.pdf) to understand sparse convolution algorithm in spconv 2.x!\n\n## WARNING\n\nUse spconv >= cu114 if possible. cuda 11.4 can compile greatly faster kernel in some situation.\n\nUpdate Spconv: you **MUST UNINSTALL** all spconv/cumm/spconv-cuxxx/cumm-cuxxx first, use ```pip list | grep spconv``` and ```pip list | grep cumm``` to check all installed package. then use pip to install new spconv.\n\n## NEWS\n\n* spconv 2.3: int8 quantization support. see docs and examples for more details.\n\n* spconv 2.2: ampere feature support (by [EvernightAurora](https://github.com/EvernightAurora)), pure c++ code generation, nvrtc, drop python 3.6\n\n## Spconv 2.2 vs Spconv 2.1\n\n* faster fp16 conv kernels (~5-30%) in ampere GPUs (tested in RTX 3090)\n* greatly faster int8 conv kernels (~1.2x-2.7x) in ampere GPUs (tested in RTX 3090)\n* drop python 3.6 support\n* nvrtc support: kernel in old GPUs will be compiled in runtime.\n* [libspconv](docs/PURE_CPP_BUILD.md): pure c++ build of all spconv ops. see [example](example/libspconv/run_build.sh)\n* tf32 kernels, faster fp32 training, disabled by default. set ```import spconv as spconv_core; spconv_core.constants.SPCONV_ALLOW_TF32 = True``` to enable them.\n* all weights are KRSC layout, some old model can't be loaded anymore.\n\n\n## Spconv 2.1 vs Spconv 1.x\n\n* spconv now can be installed by **pip**. see install section in readme for more details. Users don't need to build manually anymore!\n* Microsoft Windows support (only windows 10 has been tested).\n* fp32 (not tf32) training/inference speed is increased (+50~80%)\n* fp16 training/inference speed is greatly increased when your layer support tensor core (channel size must be multiple of 8).\n* int8 op is ready, but we still need some time to figure out how to run int8 in pytorch.\n* [doesn't depend on pytorch binary](docs/FAQ.md#What-does-no-dependency-on-pytorch-mean), but you may need at least pytorch >= 1.5.0 to run spconv 2.x.\n* since spconv 2.x doesn't depend on pytorch binary (never in future), it's impossible to support torch.jit/libtorch inference.\n\n## Usage\n\nFirstly you need to use ```import spconv.pytorch as spconv``` in spconv 2.x.\n\nThen see [this](docs/USAGE.md).\n\nDon't forget to check [performance guide](docs/PERFORMANCE_GUIDE.md).\n\n### Common Solution for Some Bugs\n\nsee [common problems](docs/COMMON_PROBLEMS.md).\n\n## Install\n\nYou need to install python >= 3.7 first to use spconv 2.x.\n\nYou need to install CUDA toolkit first before using prebuilt binaries or build from source.\n\nYou need at least CUDA 11.0 to build and run spconv 2.x. We won't offer any support for CUDA < 11.0.\n\n### Prebuilt\n\nWe offer python 3.7-3.11 and cuda 10.2/11.3/11.4/11.7/12.0 prebuilt binaries for linux (manylinux).\n\nWe offer python 3.7-3.11 and cuda 10.2/11.4/11.7/12.0 prebuilt binaries for windows 10/11.\n\nFor Linux users, you need to install pip >= 20.3 first to install prebuilt.\n\n**WARNING**: spconv-cu117 may require CUDA Driver >= 515.\n\n```pip install spconv``` for CPU only (**Linux Only**). you should only use this for debug usage, the performance isn't optimized due to manylinux limit (no omp support).\n\n```pip install spconv-cu102``` for CUDA 10.2\n\n```pip install spconv-cu113``` for CUDA 11.3 (**Linux Only**)\n\n```pip install spconv-cu114``` for CUDA 11.4\n\n```pip install spconv-cu117``` for CUDA 11.7\n\n```pip install spconv-cu120``` for CUDA 12.0\n\n**NOTE** It's safe to have different **minor** cuda version between system and conda (pytorch) in **CUDA >= 11.0** because of [CUDA Minor Version Compatibility](https://docs.nvidia.com/deploy/cuda-compatibility/#minor-version-compatibility). For example, you can use spconv-cu114 with anaconda version of pytorch cuda 11.1 in a OS with CUDA 11.2 installed.\n\n**NOTE** In Linux, you can install spconv-cuxxx without install CUDA to system! only suitable NVIDIA driver is required. for CUDA 11, we need driver >= 450.82. You may need newer driver if you use newer CUDA. for cuda 11.8, you need to have driver >= 520 installed.\n\n#### Prebuilt GPU Support Matrix\n\nSee [this page](https://arnon.dk/matching-sm-architectures-arch-and-gencode-for-various-nvidia-cards/) to check supported GPU names by arch.\n\nIf you use a GPU architecture that isn't compiled in prebuilt, spconv will use NVRTC to compile a slightly slower kernel.\n\n| CUDA version | GPU Arch List |\n| -------------- |:---------------------:|\n| 11.1~11.7 | 52,60,61,70,75,80,86 | \n| 11.8+ | 60,70,75,80,86,89,90 | \n\n### Build from source for development (JIT, recommend)\n\nThe c++ code will be built automatically when you change c++ code in project.\n\nFor NVIDIA Embedded Platforms, you need to specify cuda arch before build: ```export CUMM_CUDA_ARCH_LIST=\"7.2\"``` for xavier, ```export CUMM_CUDA_ARCH_LIST=\"6.2\"``` for TX2, ```export CUMM_CUDA_ARCH_LIST=\"8.7\"``` for orin.\n\nYou need to remove ```cumm``` in ```requires``` section in pyproject.toml after install editable ```cumm``` and before install spconv due to pyproject limit (can't find editable installed ```cumm```).\n\nYou need to ensure ```pip list | grep spconv``` and ```pip list | grep cumm``` show nothing before install editable spconv/cumm.\n\n#### Linux\n\n0. uninstall spconv and cumm installed by pip\n1. install build-essential, install CUDA\n2. ```git clone https://github.com/FindDefinition/cumm```, ```cd ./cumm```, ```pip install -e .```\n3. ```git clone https://github.com/traveller59/spconv```, ```cd ./spconv```, ```pip install -e .```\n4. in python, ```import spconv``` and wait for build finish.\n\n#### Windows\n0. uninstall spconv and cumm installed by pip\n1. install visual studio 2019 or newer. make sure C++ development component is installed. install CUDA\n2. set [powershell script execution policy](https://docs.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_execution_policies?view=powershell-7.1)\n3. start a new powershell, run ```tools/msvc_setup.ps1```\n4. ```git clone https://github.com/FindDefinition/cumm```, ```cd ./cumm```, ```pip install -e .```\n5. ```git clone https://github.com/traveller59/spconv```, ```cd ./spconv```, ```pip install -e .```\n6. in python, ```import spconv``` and wait for build finish.\n\n### Build wheel from source (not recommend, this is done in CI.)\n\nYou need to rebuild ```cumm``` first if you are build along a CUDA version that not provided in prebuilts.\n\n#### Linux\n\n1. install build-essential, install CUDA\n2. run ```export SPCONV_DISABLE_JIT=\"1\"```\n3. run ```pip install pccm cumm wheel```\n4. run ```python setup.py bdist_wheel```+```pip install dists/xxx.whl```\n\n#### Windows\n\n1. install visual studio 2019 or newer. make sure C++ development component is installed. install CUDA\n2. set [powershell script execution policy](https://docs.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_execution_policies?view=powershell-7.1)\n3. start a new powershell, run ```tools/msvc_setup.ps1```\n4. run ```$Env:SPCONV_DISABLE_JIT = \"1\"```\n5. run ```pip install pccm cumm wheel```\n6. run ```python setup.py bdist_wheel```+```pip install dists/xxx.whl```\n\n## Citation\n\nIf you find this project useful in your research, please consider cite:\n\n```latex\n@misc{spconv2022,\n title={Spconv: Spatially Sparse Convolution Library},\n author={Spconv Contributors},\n howpublished = {\\url{https://github.com/traveller59/spconv}},\n year={2022}\n}\n```\n## Contributers\n\n* [EvernightAurora](https://github.com/EvernightAurora): add ampere feature.\n\n## Note\n\nThe work is done when the author is an employee at [Tusimple](https://www.tusimple.com/).\n\n## LICENSE\n\nApache 2.0\n",
"bugtrack_url": null,
"license": "Apache License 2.0",
"summary": "spatial sparse convolution",
"version": "2.3.6",
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "435c5df126879349ee8650080875cef23dcecdbf6e5abca39260a81ab5f44bfd",
"md5": "2f90b0e40ae27a45b7e8b890f92ce8b9",
"sha256": "df8559c1ce7fe7f9ba0a331e6ec8a581d52d4ac0e7f17eb3333a4cbda4804bc9"
},
"downloads": -1,
"filename": "spconv_cu102-2.3.6-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl",
"has_sig": false,
"md5_digest": "2f90b0e40ae27a45b7e8b890f92ce8b9",
"packagetype": "bdist_wheel",
"python_version": "cp310",
"requires_python": ">=3.7",
"size": 51932661,
"upload_time": "2023-04-19T16:19:51",
"upload_time_iso_8601": "2023-04-19T16:19:51.137597Z",
"url": "https://files.pythonhosted.org/packages/43/5c/5df126879349ee8650080875cef23dcecdbf6e5abca39260a81ab5f44bfd/spconv_cu102-2.3.6-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "e95eb904105f94e7a6647ecf9a9a5cb925d3a6db107a24c7569db44f4ba0aaad",
"md5": "6ca6e2be4d336ed554230d94dec14a11",
"sha256": "18bc9af1273652efda0b38ada0ab04e1a4c855577fa332e0968b5d6e5ccc1342"
},
"downloads": -1,
"filename": "spconv_cu102-2.3.6-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl",
"has_sig": false,
"md5_digest": "6ca6e2be4d336ed554230d94dec14a11",
"packagetype": "bdist_wheel",
"python_version": "cp311",
"requires_python": ">=3.7",
"size": 51939272,
"upload_time": "2023-04-19T16:21:34",
"upload_time_iso_8601": "2023-04-19T16:21:34.221133Z",
"url": "https://files.pythonhosted.org/packages/e9/5e/b904105f94e7a6647ecf9a9a5cb925d3a6db107a24c7569db44f4ba0aaad/spconv_cu102-2.3.6-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "bed843bc55c41733595ef4fb639a8a893b9bd7b83644636c45ff1a2c05751291",
"md5": "8955ccf1e5d81be5bdb8a40a07176f18",
"sha256": "7619cd3b827a8f010d435ba22b7b49642e05e41f7734522fcd086af61b9eda84"
},
"downloads": -1,
"filename": "spconv_cu102-2.3.6-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl",
"has_sig": false,
"md5_digest": "8955ccf1e5d81be5bdb8a40a07176f18",
"packagetype": "bdist_wheel",
"python_version": "cp37",
"requires_python": ">=3.7",
"size": 51924045,
"upload_time": "2023-04-19T16:21:40",
"upload_time_iso_8601": "2023-04-19T16:21:40.564803Z",
"url": "https://files.pythonhosted.org/packages/be/d8/43bc55c41733595ef4fb639a8a893b9bd7b83644636c45ff1a2c05751291/spconv_cu102-2.3.6-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-04-19 16:19:51",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "traveller59",
"github_project": "spconv",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "spconv-cu102"
}
JSON
