chrisway613 / dogs_vs_cats

Udacity Machine Learning Capstone Project

dogs_vs_cats

Udacity Machine Learning Capstone Project.

Libs imported in this project

os; numpy; pandas; pathlib; pprint; seaborn; functools; matplotlib; keras; sklearn; tensorflow; IPython; gc.

Machine OS & Hardware

Machine: Amazon EC2(p2.xlarge instance); OS: Deep Learning AMI (Ubuntu) Version 23.0, Ubuntu 16.04.6 LTS (GNU/Linux 4.4.0-1088-aws x86_64v); Hardware: 4 vCPU, 1000G user filesystem, NVIDIA Tesla K80 GPUs, 12 GiB of memory per GPU.

Code illustration

data_processing.ipynb: data preprocessing, including data visualization, data cleaning, spliting trainig&validation set, data augmentation; model_training_x.ipynb: including model building, model training with different optimizer & parameter, model prediction.

Training Time used

model which combining Xception&InceptionV3: about 700s per epochs; model which combining NARSNetLarge&InceptionResNetV2: about 2580s per epochs.

Data sources

refer to https://www.kaggle.com/c/dogs-vs-cats-redux-kernels-edition/data.

Additional

file 'cheat.csv' which contains some error-labeled images in training set, comes from kaggle forum.