week02

iYanjo · Apr 18, 2023 · c78290c · c78290c
1 parent 8d22ea8
commit c78290c
Show file tree

Hide file tree

Showing 6 changed files with 2,366 additions and 0 deletions.
diff --git a/week02_autodiff/README.md b/week02_autodiff/README.md
@@ -0,0 +1,31 @@
+[__slides__](https://yadi.sk/i/eRVlESjqlIPBGw)
+
+
+## Materials
+
+- __In english:__
+  * Deep learning frameworks - [video](https://www.youtube.com/watch?v=Vf_-OkqbwPo)
+  * [PyTorch tutorial](https://www.youtube.com/watch?v=VMcRWYEKmhw)
+  * [Tensorflow tutorial](https://www.youtube.com/watch?v=FQ660T4uu7k)
+
+- __In russian:__
+  * [Pytorch tutorial](https://yadi.sk/i/O3mQ76u43So3h9) __recommended__
+  * [Tensorflow tutorial](https://www.youtube.com/watch?v=FQ660T4uu7k) (english only for now. Links are welcome)
+
+## More on DL frameworks
+  - A lecture on nonlinearities, intializations and other tricks in deep learning (karpathy) - [video](https://www.youtube.com/watch?v=GUtlrDbHhJM)
+  - A lecture on activations, recap of adaptive SGD and dropout (karpathy) - [video](https://www.youtube.com/watch?v=KaR4lIdI1MQ)
+  - [a deep learning neophite cheat sheet](http://www.kdnuggets.com/2016/03/must-know-tips-deep-learning-part-1.html)
+  - [bonus video] Deep learning philosophy: [our humble take](https://www.youtube.com/watch?v=9qyE1Ev1Xdw) (english)
+  - [reading] on weight initialization: [blog post](http://andyljones.tumblr.com/post/110998971763/an-explanation-of-xavier-initialization)
+  - [reading] pretty much all the module 1 of http://cs231n.github.io/
+
+
+## Practice
+
+[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/yandexdataschool/Practical_DL/blob/spring33/week02_autodiff/seminar_pytorch.ipynb)
+
+As usual, go to `seminar_pytorch.ipynb` and follow instructions from there. You will also need to pass `homework_pytorch.ipynb` for full score.
+
+__Alternative (TensorFlow):__ a similar tutorial for tensorflow is provided in `tensorflow.ipynb`. From now on, you *can* submit assignments in any framework - but you will have to do some extra engineering in that case. However, unless you're already profficient with PyTorch, we recommend you stick to it.
+
diff --git a/week02_autodiff/homework.ipynb b/week02_autodiff/homework.ipynb
diff --git a/week02_autodiff/mnist.py b/week02_autodiff/mnist.py
@@ -0,0 +1,63 @@
+import sys
+import os
+import time
+
+import numpy as np
+
+__doc__="""taken from https://github.com/Lasagne/Lasagne/blob/master/examples/mnist.py"""
+
+def load_dataset():
+    # We first define a download function, supporting both Python 2 and 3.
+    if sys.version_info[0] == 2:
+        from urllib import urlretrieve
+    else:
+        from urllib.request import urlretrieve
+
+    def download(filename, source='http://yann.lecun.com/exdb/mnist/'):
+        print("Downloading %s" % filename)
+        urlretrieve(source + filename, filename)
+
+    # We then define functions for loading MNIST images and labels.
+    # For convenience, they also download the requested files if needed.
+    import gzip
+
+    def load_mnist_images(filename):
+        if not os.path.exists(filename):
+            download(filename)
+        # Read the inputs in Yann LeCun's binary format.
+        with gzip.open(filename, 'rb') as f:
+            data = np.frombuffer(f.read(), np.uint8, offset=16)
+        # The inputs are vectors now, we reshape them to monochrome 2D images,
+        # following the shape convention: (examples, channels, rows, columns)
+        data = data.reshape(-1, 1, 28, 28)
+        # The inputs come as bytes, we convert them to float32 in range [0,1].
+        # (Actually to range [0, 255/256], for compatibility to the version
+        # provided at http://deeplearning.net/data/mnist/mnist.pkl.gz.)
+        return data / np.float32(256)
+
+    def load_mnist_labels(filename):
+        if not os.path.exists(filename):
+            download(filename)
+        # Read the labels in Yann LeCun's binary format.
+        with gzip.open(filename, 'rb') as f:
+            data = np.frombuffer(f.read(), np.uint8, offset=8)
+        # The labels are vectors of integers now, that's exactly what we want.
+        return data
+
+    # We can now download and read the training and test set images and labels.
+    X_train = load_mnist_images('train-images-idx3-ubyte.gz')
+    y_train = load_mnist_labels('train-labels-idx1-ubyte.gz')
+    X_test = load_mnist_images('t10k-images-idx3-ubyte.gz')
+    y_test = load_mnist_labels('t10k-labels-idx1-ubyte.gz')
+
+    # We reserve the last 10000 training examples for validation.
+    X_train, X_val = X_train[:-10000], X_train[-10000:]
+    y_train, y_val = y_train[:-10000], y_train[-10000:]
+
+    # We just return all the arrays in order, as expected in main().
+    # (It doesn't matter how we do this as long as we can read them again.)
+    return X_train, y_train, X_val, y_val, X_test, y_test
+
+
+
+
diff --git a/week02_autodiff/notmnist.py b/week02_autodiff/notmnist.py
@@ -0,0 +1,52 @@
+import os
+from glob import glob
+
+import numpy as np
+from matplotlib.pyplot import imread
+from skimage.transform import resize
+from sklearn.model_selection import train_test_split
+
+
+def load_notmnist(path='./notMNIST_small', letters='ABCDEFGHIJ',
+                  img_shape=(28, 28), test_size=0.25, one_hot=False):
+    # download data if it's missing. If you have any problems, go to the urls
+    # and load it manually.
+    if not os.path.exists(path):
+        print("Downloading data...")
+        assert os.system(
+            'wget http://yaroslavvb.com/upload/notMNIST/notMNIST_small.tar.gz') == 0
+        print("Extracting ...")
+        assert os.system(
+            'tar -zxvf notMNIST_small.tar.gz > untar_notmnist.log') == 0
+
+    data, labels = [], []
+    print("Parsing...")
+    for img_path in glob(os.path.join(path, '*/*')):
+        class_i = img_path.split(os.sep)[-2]
+        if class_i not in letters:
+            continue
+        try:
+            data.append(resize(imread(img_path), img_shape))
+            labels.append(class_i,)
+        except BaseException:
+            print(
+                "found broken img: %s [it's ok if <10 images are broken]" %
+                img_path)
+
+    data = np.stack(data)[:, None].astype('float32')
+    data = (data - np.mean(data)) / np.std(data)
+
+    # convert classes to ints
+    letter_to_i = {l: i for i, l in enumerate(letters)}
+    labels = np.array(list(map(letter_to_i.get, labels)))
+
+    if one_hot:
+        labels = (np.arange(np.max(labels) + 1)
+                  [None, :] == labels[:, None]).astype('float32')
+
+    # split into train/test
+    X_train, X_test, y_train, y_test = train_test_split(
+        data, labels, test_size=test_size, random_state=42)
+
+    print("Done")
+    return X_train, y_train, X_test, y_test