Matrix Multiplication

0.1 First description

• This notebook is actually reprodcution fo course practical deep learning for coders part2. Matrix Multiplication starts with Lession 11 (1:08:47).
• To get the noebook shown in the video, one should clone the course repository and run the notebook 01_matmul.ipynb.

0.2 Get data

We first download the mnist data to work with it

from pathlib import Path
import gzip, time, os, pickle, math
from urllib.request import urlretrieve

HOME = os.getenv('HOME')
data_path = Path(fr'{HOME}/Schreibtisch/projects/git_data/course22p2/nbs/data')
MNIST_URL='https://github.com/mnielsen/neural-networks-and-deep-learning/blob/master/data/mnist.pkl.gz?raw=true'
path_data = data_path
path_data.mkdir(exist_ok=True)
path_gz = path_data/'mnist.pkl.gz'

with gzip.open(path_gz, 'rb') as f_in:
    ((x_train, y_train), (x_valid, y_valid),_ ) =pickle.load(f_in, encoding='latin1')

x_train.shape

(50000, 784)