Follow the wechat public number: Mai Microelectronics RESEARCH and development Club, get more exciting content, first in the personal public number.

@TOC

Since PyTorch launched in January 2017, its popularity has continued to grow. PyTorch was quickly adopted by many researchers and engineers because of its many advantages, such as Python, dynamic graph mechanism, flexible network construction, and strong community.

A recent visit to GitHub found a great list of Pytorch resources, including NLP/CV projects, sample code, libraries, paper implementations, and more. Here I have done a sort, recommend to you.

Coincidentally, I am familiar with this list of resources. Compared with the previous article, I found that there was an original English version on GitHub before, and this one was translated into Chinese

Original English version of GitHub project address: github.com/bharathgs/A…

GitHub address: github.com/xavier-zy/A…

1. Natural language processing and speech processing

This section contains 41 popular PyTorch NLP related projects, such as a cross-speaker speech generation method, speech to text end-to-end model implementation, fast WaveNet generation implementation; PyTorch NLP is a popular library related to PyTorch NLP, such as The PyTorch NLP library based on FastAI, LASER, which is used to calculate and use multi-language statement embedding. PyTorch NLP is a popular framework and tool related to PyTorch, such as PyTorch – SeQ2SeQ, PyTorch sequence-to-sequence framework nMTPyTorch, etc.

Computer vision

This section contains 25 popular PyTorch CV related projects and libraries. Examples include TorchVision, which includes popular data sets, model architectures, and image transformations commonly used in computer vision, Augmentor, an image enhancement library for machine learning, maskrCNN-Benchmark, a fast modular reference implementation for instance segmentation and object detection, PyTorch based 2D and 3D face alignment library ACE -alignment and more.

3. Probability library and generation library

4. Tutorials and examples

This section features 66 PyTorch classic tutorials, including reinforcement learning, NLP, and CV. Logistic, CNN, RNN, LSTM and other neural network models are implemented by several lines of code, and some advanced examples are implemented by complex models.

For example, PyTorch number 5 is a variety of tutorials, which, according to the official tutorial, is rich in content:

pytorch.org/tutorials/

Deep Learning with PyTorch: A 60 Minute Blitz is the best introduction to PyTorch.

5. Thesis realization

This section includes 338 PyTorch related paper implementations. For example, PyTorch implements a recursive variational autoencoder for generating sequence data, PyTorch implements v-Net: Full convolutional neural network for the application of body medical image segmentation, and PyTorch implements a simple implementation for generating adversarial network, focusing on cartoon face painting, etc.

6. Other

This section covers 37 PyTorch resources, including a list of tutorials, papers, projects, communities, forums, and Deep Learning templates. There are also some interesting projects, such as drawing with neural networks, a chatbot with PyTorch, and playing backgammon with AlphaZero.

Finally, attach the GitHub address of the project:

Github.com/xavier-zy/A…

Recommended reading

  1. Just now! PyTorch is available for free for a limited time!

  2. 23 k +, making the star from scratch the depth study of practical tutorial | PyTorch official recommendations

  3. GitHub Trend # 1: Detectron2, PyTorch target detection, faster training and more tasks

  4. Machine Learning Algorithms — Support Vector Machine (SVM)

  5. Machine learning algorithms — Logistic Regression algorithm and Python implementation

  6. Machine learning algorithm — Decision Tree Model algorithm and Python implementation

  7. Machine learning algorithm — K-nearest Neighbor (KNN) classification algorithm principle Explanation

  8. Machine learning algorithm — K-nearest Neighbor (KNN) algorithm implemented in Python


Follow the wechat public number: Mai Microelectronics RESEARCH and development Club, get more exciting content, first in the personal public number.

Knowledge Planet (paid group) : The community aims to share autumn/spring recruitment preparation tips (including brush questions), interview and internal promotion opportunities, learning routes, knowledge questions bank, etc.