PyTorch

Created by Facebook, PyTorch is the second most popular deep learning framework (behind TensorFlow). It is more popular in the research community (vs. industry) and is consider to have a cleaner API.

Logo

../_images/pytorch_pytorch-small.png

Website

https://pytorch.org

Repository

https://github.com/pytorch/pytorch

Byline

Tensors and Dynamic neural networks in Python with strong GPU acceleration.

License

BSD 3-clause

Project age

10 years 2 months

Backers

FaceBook (Creator and maintainer)

Lastest News (2022-06-28)

PyTorch 1.12: TorchArrow, Functional API for Modules and nvFuser, are now available We are excited to announce the release of PyTorch … more

Size score (1 to 10, higher is better)

9.25

Trend score (1 to 10, higher is better)

8.25

Education Resources

URL

Resource Type

Description

https://pytorch.org/docs/stable/index.html

Documentation

This is the documentation site.

https://pytorch.org/tutorials/

Project Website

PyTorch has a really good tutorial site for beginners.

https://github.com/pytorch/tutorials

Example Code

PyTorch maintains its tutorials on Github. If you want code examples, it is worth to explore this repository.

https://www.youtube.com/watch?v=GIsg-ZUy0MY

Video

This is a 9-hour training on YouTube.

Git Commit Statistics

Statistics computed using Git data through March 31, 2022.

Statistic

Lifetime

Last 12 Months

Commits

116,826

41,952

Lines committed

25,250,912

10,620,681

Unique committers

3,119

871

Core committers

15

16

../_images/pytorch_pytorch-monthly-commits.png

Similar Projects

Project

Size Score

Trend Score

Byline

DeepLearning4j

6.75

6.0

Eclipse Deeplearning4j, ND4J, DataVec and more - deep learning & linear algebra for Java/Scala with GPUs + Spark

Keras

7.25

6.25

Deep Learning for humans

MXNet

7.75

4.25

A flexible and efficient library for deep learning.

TVM

7.0

9.25

Open deep learning compiler stack for cpu, gpu and specialized accelerators

TensorFlow

10.0

5.5

An open source machine learning framework for everyone.