coursera clone

coursera01w02/lr_utils.py

+import numpy as np
+import h5py
+from os import path
+
+def load_dataset():
+    train_path = path.join(path.dirname(path.realpath(__file__)), 'datasets/train_catvnoncat.h5')
+    train_dataset = h5py.File(train_path, "r")
+    train_set_x_orig = np.array(train_dataset["train_set_x"][:]) # your train set features
+    train_set_y_orig = np.array(train_dataset["train_set_y"][:]) # your train set labels
+
+    test_path = path.join(path.dirname(path.realpath(__file__)), 'datasets/test_catvnoncat.h5')
+    test_dataset = h5py.File(test_path, "r")
+    test_set_x_orig = np.array(test_dataset["test_set_x"][:]) # your test set features
+    test_set_y_orig = np.array(test_dataset["test_set_y"][:]) # your test set labels
+
+    classes = np.array(test_dataset["list_classes"][:]) # the list of classes
+
+    train_set_y_orig = train_set_y_orig.reshape((1, train_set_y_orig.shape[0]))
+    test_set_y_orig = test_set_y_orig.reshape((1, test_set_y_orig.shape[0]))
+
+    return train_set_x_orig, train_set_y_orig, test_set_x_orig, test_set_y_orig, classes
+

requirements.txt

+torch
+torchvision
+numpy
+matplotlib
+h5py
+scipy
+# PIL
\ No newline at end of file