README.md

cmu-11785-deeplearning-s19

HW codes in the course 11785-s19

INTRODUCTION

Each HW has two parts:
Part 1: Deep learning from scratch (No DL framework allowed)
Part 2: Kaggle (PyTorch, AWS or GCP)
My name on Kaggle leaderboard: zihaog

HW 0

Basic NumPy and PyTorch

HW 1

Part 1: MLP
Part 2 (Kaggle): Frame Level Classification of Wall Street Journal Speech via MLP
https://www.kaggle.com/c/11-785hw1p2-s19-slack/overview

HW 2

Part 1: CNN
Part 2 (Kaggle): Face Classification and Verification via Convolutional Neural Networks (ResNet50)
Classification: https://www.kaggle.com/c/11785-s19-hw2p2-classification/overview
Verification: https://www.kaggle.com/c/11785-s19-hw2p2-verification2

HW 3

Part 1: GRU
Part 2 (Kaggle): Utterance to Phoneme Mapping
https://www.kaggle.com/c/11785-s19-hw3p2v2/overview

HW 4

Part 1: Language modeling using RNNs
Part 2 (Kaggle): Attention-based End-to-End Speech-to-Text Deep Neural Network (Listen, Attend and Spell)
https://www.kaggle.com/c/11785-s19-hw4p2-slack/overview