Open source at the forefront （ID：OpenSourceTop） The BGF
10 month GitHub On the hottest Python The list of open source projects has been published , Let's take a look at the details of the list ：
https://github.com/soimort/you-get Star 36698
you-get It's actually a very small command line tool , Can help you from Web Download media content , Such as video 、 Audio 、 Image, etc , adopt you-get, You can ：
https://github.com/pytorch/fairseq Star 10205
Fairseq This translation model is made up of Facebook AI The laboratory is in 2017 in , And in the past with RNN Based on the translation model , In order to cnn The main model structure .Fairseq Model of , The structure of the model is shown in the figure below ：
https://github.com/CorentinJ/Real-Time-Voice-Cloning Star 21291
Real-Time-Voice-Cloning This is a speech synthesis project based on deep learning , It collects and analyzes a specific sound sample , Can be found in 5 Generate similar clone voice within seconds .
https://github.com/home-assistant/core Star 36968
Home Assistant Is a run in Python 3 Home automation platform on . It can be connected to a variety of home smart devices —— The lamp 、 switch 、 door 、 Fan 、 Ring the bell 、 Temperature and humidity meter 、 sound 、 Security monitoring equipment 、 Smart meters and so on , This project is very, very suitable for Raspberry Pi Or on the local server .
https://github.com/quantopian/zipline Star 12659
Zipline It's a Pythonic Algorithmic transaction Library . It's an event driven system , Support back testing and real-time trading .Zipline Currently used in production as Quantopian（ Hosting platform ） Testing and real-time trading engine .
https://github.com/sherlock-project/sherlock Star 16718
Sherlock It's an easy to use Python Distributed in process locking mechanism library , You can choose different background of lock synchronization .
https://github.com/rwightman/pytorch-image-models Star 5386
pytorch-image-models Integrated PyTorch Many image models of 、 Script and pre training parameters . Except from Torchvision And the existing warehouse model , The creators themselves provided some models .
https://github.com/facebookresearch/pifuhd Star 3853
adopt pifuhd, You can do it with just one photo 3D modeling , Its basic framework is ICCV 2019 Implicit function of pixel alignment on the scene PIFu. however ,PIFu At a resolution of 512×512 Image as input , Output 3D The resolution of the model is not high .
https://github.com/huggingface/transformers Star 36289
Transformers yes TensorFlow 2.0 and PyTorch The latest natural language processing library ,Transformers( Formerly known as pytorch-transformers and pytorch-pretrained-bert) For natural language understanding (NLU) And natural language generation (NLG) The most advanced model of (BERT,GPT-2,RoBERTa,XLM,DistilBert,XLNet,CTRL …) , Have more than 32 Kind of pre training model , Support 100 A variety of languages , And in TensorFlow 2.0 and PyTorch There is deep interoperability between .
https://github.com/Jack-Cherish/python-spider Star 11676
This article is from WeChat official account. - Machine learning algorithm and Python Study （MLPython）
The source and reprint of the original text are detailed in the text , If there is any infringement , Please contact the [email protected] Delete .
Original publication time ： 2020-11-10
Participation of this paper Tencent cloud media sharing plan , You are welcome to join us , share .