# b10-kernel
Baseten Kernel Library - High-performance GPU kernels for AI inference workloads.
## Installation
### From PyPI
```bash
pip install b10-kernel
```
**Requirements:**
- Python >= 3.12
- CUDA-compatible GPU and drivers
- PyTorch >= 2.8.0 with CUDA support
### From Source
```bash
git clone <repository>
cd mp/kernels/b10-kernel
pip install -e .
```
### For Development
```bash
# Install with test dependencies
pip install -e .[test]
# Install with all development dependencies
pip install -e .[dev]
```
## Development guide
- Build the library from source
```bash
make build
make rebuild
```
- Run unit tests
```bash
make test
```
- Format code
```bash
make format
```
## Kernel Development Workflow
Steps to add a new kernel:
- Implement the kernel in `csrc`
- Expose the interface in `include/b10_kernel_ops.h`
- Create torch extension in `csrc/common_extension.cc`
- Update `CMakeLists.txt` to include new CUDA source
- Expose Python interface in `python/b10_kernel/xxx.py` and `python/b10_kernel/__init__.py`
- Add unit test for the kernel in `test/test_xxx.py`
- Add benchmark script for the kernel in `benchmark/bench_xxx.py`
- Format code with `make format`
Raw data
{
"_id": null,
"home_page": null,
"name": "b10-kernel",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.12",
"maintainer_email": "Ke Bao <ke.bao@baseten.co>, Pankaj Gupta <pankaj@baseten.co>, Yikai Zhu <yikai.zhu@baseten.co>, Shounak Ray <shounak.ray@baseten.co>",
"keywords": "machine learning, gpu, cuda, kernels, pytorch",
"author": null,
"author_email": "Ke Bao <ke.bao@baseten.co>",
"download_url": null,
"platform": null,
"description": "# b10-kernel\n\nBaseten Kernel Library - High-performance GPU kernels for AI inference workloads.\n\n## Installation\n\n### From PyPI\n```bash\npip install b10-kernel\n```\n\n**Requirements:**\n- Python >= 3.12\n- CUDA-compatible GPU and drivers\n- PyTorch >= 2.8.0 with CUDA support\n\n### From Source\n```bash\ngit clone <repository>\ncd mp/kernels/b10-kernel\npip install -e .\n```\n\n### For Development\n```bash\n# Install with test dependencies\npip install -e .[test]\n\n# Install with all development dependencies \npip install -e .[dev]\n```\n\n## Development guide\n- Build the library from source\n```bash\nmake build\nmake rebuild\n```\n- Run unit tests\n```bash\nmake test\n```\n- Format code\n```bash\nmake format\n```\n\n## Kernel Development Workflow\nSteps to add a new kernel:\n- Implement the kernel in `csrc`\n- Expose the interface in `include/b10_kernel_ops.h`\n- Create torch extension in `csrc/common_extension.cc`\n- Update `CMakeLists.txt` to include new CUDA source\n- Expose Python interface in `python/b10_kernel/xxx.py` and `python/b10_kernel/__init__.py`\n- Add unit test for the kernel in `test/test_xxx.py`\n- Add benchmark script for the kernel in `benchmark/bench_xxx.py`\n- Format code with `make format`\n\n",
"bugtrack_url": null,
"license": null,
"summary": "Baseten Kernel Library",
"version": "0.1.1",
"project_urls": null,
"split_keywords": [
"machine learning",
" gpu",
" cuda",
" kernels",
" pytorch"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "b849a231a4496b977ca9a08700034c03e1ad9e52e91b31259b69f94bf8104b3f",
"md5": "21b96e02a56e31ad913b042edae1e766",
"sha256": "046027b27b8d36e603d397065f0cb872d1d549a72a22edaaa5fdcd0560c91b89"
},
"downloads": -1,
"filename": "b10_kernel-0.1.1-py312-none-manylinux2014_x86_64.whl",
"has_sig": false,
"md5_digest": "21b96e02a56e31ad913b042edae1e766",
"packagetype": "bdist_wheel",
"python_version": "py312",
"requires_python": ">=3.12",
"size": 513650,
"upload_time": "2025-08-18T15:46:50",
"upload_time_iso_8601": "2025-08-18T15:46:50.373354Z",
"url": "https://files.pythonhosted.org/packages/b8/49/a231a4496b977ca9a08700034c03e1ad9e52e91b31259b69f94bf8104b3f/b10_kernel-0.1.1-py312-none-manylinux2014_x86_64.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-18 15:46:50",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "b10-kernel"
}
JSON