Building a dockerized classifier for detecting dysplastic colon glands while guiding it's attention blocks by provided regions of interest. The attention block can be a soft attention in case of a CNN-based network and a self-attention in case of a ViT-based network ⚡.

We implemented QANet from scratch and improved baseline BiDAF. We also used an ensemble of BiDAF and QANet models to achieve EM/F1 of 69.47/71.96, ranking #3 on the leaderboard as of Mar 4, 2022.

A torch-trained network that leverages self-attention and LSTMs to generate piano notes from midi files.

Adversarial Transfer Learning for Chinese Named Entity Recognition with Self-Attention Mechanism

An open-source implementation of the paper ``A Structured Self-Attentive Sentence Embedding'' published by IBM and MILA.

This repository provides a PyTorch implementation of SAGAN cited by heykeetae/Self-Attention-GAN. This repository provide an efficient method to generate large resolution images and attention weights visualisation using tensorboard platform. Tensorboard is a robust platform to monitor generated images and learning weights in computer vision lear…

La pensée critique, c'est évaluer selon des critères pour porter un jugement. « La pensée critique est un processus qui consiste à examiner des idées ou des situations pour arriver à bien les comprendre, à en déterminer les implications ou les conséquences et à porter un jugement ou à éclairer une décision

Subjectivity removal and Polarity classification of movie reviews employing a shallow model (Multinomial Naive Bayes) and a deep model (Bidirectional LSTM with self-attention)

Investigation into Transformer self-attention building blocks, and the effects of pretraining.

This repository is reimplementation of Transformer model which was introduced in 2017 NeurIPS paper "Attention is all you need"

Unsupervised feed forward layer learning with a single pass.

The implementation of transformer as presented in the paper "Attention is all you need" from scratch.

SA-DETR: Saliency Attention-based DETR for Salienct Object Detection

Implementation of different object detection neural networks, in order to detect pedestrians.

Learning parities with various neural network architectures

Self-Attention Implementation using Pytorch

Unofficial Implementation of Universal Transformer https://arxiv.org/abs/1807.03819

Twitter Sentiment analysis using RNS like LSTMs, GRUs and enhancing the performance with ELMo embeddings and a self-attention model

Self-Interpretable Agent implemented on the Procgen game 'Dodgeball'.

Implementation of a variety of transformer/MSA based architectures.