A repository of concepts related to neural networks for NLP

model compression based on pytorch (1、quantization: 8/4/2bits(dorefa)、ternary/binary value(twn/bnn/xnor-net)；2、 pruning: normal、regular and group convolutional channel pruning；3、 group convolution …

Different implementations of "Weighted Prediction Error" for speech dereverberation

albert_zh对应的pytorch版本

Pyroomacoustics is a package for audio signal processing for indoor applications. It was developed as a fast prototyping platform for beamforming algorithms in indoor scenarios.

Rainbow: Combining Improvements in Deep Reinforcement Learning

A Flow-based Generative Network for Speech Synthesis

The Implementation of FastSpeech based on pytorch.

台湾大学李宏毅老师机器学习

Efficient neural speech synthesis

Tacotron 2 - PyTorch implementation with faster-than-realtime inference

A speech dereverberation algorithm, also called wpe

A system for quickly generating training data with weak supervision

Reading Wikipedia to Answer Open-Domain Questions

fastNLP: A Modularized and Extensible NLP Framework. Currently still in incubation.

Code for the paper "Language Models are Unsupervised Multitask Learners"

A PyTorch implementation of OpenAI's finetuned transformer language model with a script to import the weights pre-trained by OpenAI

PyTorch original implementation of Cross-lingual Language Model Pretraining.

Code and model for the paper "Improving Language Understanding by Generative Pre-Training"

Facebook AI Research Automatic Speech Recognition Toolkit

A natural language modeling framework based on PyTorch

Make huge neural nets fit in memory

Flight rules for git

implement distributed machine learning with Pytorch + OpenMPI

Pre-training of Deep Bidirectional Transformers for Language Understanding: pre-train TextCNN

Import C++ files directly from Python!

Training RNNs as Fast as CNNs (https://arxiv.org/abs/1709.02755)

PyTorch version of Google AI's BERT model with script to load Google's pre-trained models

深度学习500问，以问答形式对常用的概率知识、线性代数、机器学习、深度学习、计算机视觉等热点问题进行阐述，以帮助自己及有需要的读者。 全书分为15个章节，近20万字。由于水平有限，书中不妥之处恳请广大读者批评指正。 未完待续............ 如有意合作，联系scutjy2015@163.com 版权所有，违权必究 Tan 2018.06

TensorFlow code and pre-trained models for BERT