A Library extending PyTorch for Personal Needs backed by C++/CUDA APIs
Project description
🔥 MyTorch 🔥
🐣 A Library extending PyTorch for Personal Needs backed by C++/CUDA APIs
🚧 WIP Forever 🚧 |
---|
Installation 👨💻
I have not included any dependencies in the setup.py
nor a requirements.txt
as I leave the hassle of setting up GPU support for torch on your own. It should work on torch>=1.4
and CUDA>=10.0
but I frankly have no clue. I use torch==1.7.1
and CUDA
Version of 11.2
To install it, just do
pip install git+https://github.com/Syzygianinfern0/MyTorch.git
Its also available on PyPi, but I wouldn't be very keen on maintaining it.
Documentation 📑
mytorch.ops
mytorch.ops.im2col
and mytorch.ops.col2im
- Rearrange image blocks into columns.
- The representation is used to perform GEMM-based convolution.
- Output is 5D (or 6D in case of minibatch) tensor.
- Minibatch implementation is inefficient, and could be done in a single CUDA kernel.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
python-mytorch-0.1.tar.gz
(4.1 kB
view hashes)
Built Distribution
Close
Hashes for python_mytorch-0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | c273a70c11c2ec4fb5e75d99d69cd6c5c260951225645f37317abfbb7a258cef |
|
MD5 | 32bed2d2559f2b10892dd602ad855e26 |
|
BLAKE2b-256 | b585d94432e234b22d323992f23792e631bb5d343cb66f8ad79f25ce0ee2d8d7 |