Torchlite
Torchlite is a high level library on top of popular machine learning frameworks such as sklearn, Pytorch and Tensorflow. It gives a high layer abstraction of repetitive code used in machine learning for day-to-day data science tasks.
Installation
pip install torchlite
or if you want to run this lib directly to have access to the examples clone this repository and run:
pip install -r requirements.txt
to install the required dependencies.
By default Pytorch 0.4.0+ and Tensorflow-GPU 1.8.0+ are installed along with this library but it's recommended
to install them from source from here if you want to use the torchlite.torch
package and/or head over to the Tensorflow install page if you want to
use the torchlite.tf package.
Documentation
For now the library has no complete documentation but you can quickly get to know how
it works by looking at the examples in the examples-* folders. This library is still in
alpha and few APIs may change in the future. The only things which will evolve at the same
pace as the library are the examples, they are meant to always be up to date with
the library.
Few examples will generates folders/files such as saved models or tensorboard logs.
To visualize the tensorboard logs download Tensorflow's tensorboard as well as
Pytorch's tensorboard if you're interested by
the torchlite.torch package. Then execute:
tensorboard --logdir=./tensorboard
Packaging the project for Pypi deploy
pip install twine
pip install wheel
python setup.py sdist
python setup.py bdist_wheel
Create a pypi account and create $HOME/.pypirc with:
[pypi]
username = <username>
password = <password>
Then upload the packages with:
twine upload dist/*
Or just:
pypi_deploy.sh
1 maintainer
Metadata
1 maintainer
Release files for torchlite 0.2.0.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| torchlite-0.2.0.1.tar.gz | 43.2 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| torchlite-0.2.0.1-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 97.2 kB
Release files / torchlite-0.2.0.1.tar.gz
| Download URL | torchlite-0.2.0.1.tar.gz |
|---|---|
| Size | 43.2 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
1a5966e3fb6905a11d961d3f19ceb58577a5ecd6a5d4ccf9eb5bd97693d92b43
|
|
BLAKE2b-256 checksum How to use checksums |
f45ec23cd2e195fcc82868798f787be77107009177175fbe816fc799233650d2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
Release files / torchlite-0.2.0.1-py3-none-any.whl
| Download URL | torchlite-0.2.0.1-py3-none-any.whl |
|---|---|
| Size | 54.0 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
1a290036a0de982cf32c98ab710e6da4429be22bf69ed65faf2b1a4eb8b5eff1
|
|
BLAKE2b-256 checksum How to use checksums |
52cbf2c73a98f09dcae362e764f794c61761a1ea446dccc4b36e0a3b9e144164
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
