Skip to content

Latest commit

 

History

History
37 lines (34 loc) · 1022 Bytes

README.md

File metadata and controls

37 lines (34 loc) · 1022 Bytes
Raw

SEM Lab Session

Environment setup

conda env create -f environment.yml
conda activate sem-deep-learning
mkdir model_dir saved_models

Getting datasets

Use this to host the files online: https://www.station307.com/#/ - will make the files available to stream from host machine over network until browser tab is closed, very handy in a classroom setting.

Metmusem

mkdir -p data/metmuseum
wget <station307-url> -O data/metmuseum/MetObjects.csv

MNIST

  1. Download the MNIST as .jpg dataset from the station307 url
wget <station307-url> -O mnist.zip
  1. Extract to data dir
mkdir -p data/MNIST/raw
unzip mnist.zip -d data/MNIST
mv data/MNIST/trainingSet/trainingSet/* data/MNIST/raw
  1. Remove unneeded files
rm -r mnist.zip data/MNIST/testS* data/MNIST/trainingS*
  1. Create filepath to label mapping:
cd data/MNIST && chmod +x create_csv.sh
./create_csv.sh raw