A Simple Pytorch version for CFU
Project description
pytorch_open_registration_example
Example of using pytorch's open device registration API. It covers:
(1) Writing custom kernels in C++, and registering them to the PyTorch dispatcher
(2) Providing a user API for your custom device, so users can invoke the custom code using torch.foo(..., device="custom_device")
(3) Registering a custom memory allocator
(4) Registering a custom device guard
This repo should be run using the latest pytorch nightly. If it fails on the latest nightly, please post an issue and I'll take a look!
python setup.py sdist bdist_wheel
twine upload dist/*
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
torch_cfu-0.0.2.tar.gz
(5.9 kB
view details)
File details
Details for the file torch_cfu-0.0.2.tar.gz.
File metadata
- Download URL: torch_cfu-0.0.2.tar.gz
- Upload date:
- Size: 5.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.1.1 pkginfo/1.4.2 requests/2.22.0 setuptools/65.3.0 requests-toolbelt/0.8.0 tqdm/4.30.0 CPython/3.8.10
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
584816c71c4450adef2faab7a3a7a91b5bf0c35cf5235b1dd69f2ec876ea5e98
|
|
| MD5 |
c1fe3ac330bcab5778eb6d09d83aca83
|
|
| BLAKE2b-256 |
337a90ce9dbbfc03f5709adc0cd1709659427ef2897856a1531211e6a5f0bfff
|
