Toroidal - Lightweight Transformers for PyTorch
Project description
toroidal - a lightweight transformer library for PyTorch
Toroidal transformers are of smaller size and lower weight than the more common E-I types. This is the software equivalent.
This is a small and educational project, not big and professional like the transformers library.
-
Simplicity! We only cover very popular transformer types and keep implementations things simple and beautiful.
-
Hightlight similarities. We try not to copy-paste code between almost identical implementations.
-
PyTorch first and only. We do not provide everything for everyone, but focus on PyTorch and use PyTorchy coding style. Ideally, our models work well with TorchScript.
Important Note
For the time being, we will emphasize beautiful code over backward compatibility, so things will break. Don't use it for things you cannot fix.
FAQ
Why not minGPT? I love minGPT, but I needed BERT.
Why not transformers? I wanted small, but if you have to ask, you should use them instead.
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 Distributions
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file toroidal-0.0.1.post2-py3-none-any.whl.
File metadata
- Download URL: toroidal-0.0.1.post2-py3-none-any.whl
- Upload date:
- Size: 4.8 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.4.2 importlib_metadata/4.6.4 pkginfo/1.4.2 requests/2.25.1 requests-toolbelt/0.9.1 tqdm/4.57.0 CPython/3.9.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
6a4bd13ace87d95e4ca7560de2aed552097fc0fa0b2e75c4927a1362b5ff17a5
|
|
| MD5 |
67d5c23ad65d08d57a96f8963f8c6ee5
|
|
| BLAKE2b-256 |
7ac49a1c533ed59b7af8b2ca064978a9c4314f689ef6940395a07287e1ea9cec
|