Skip to content

Commit 49dc4bc

Browse files
andremicciandremicci
committed
Created using Colab
1 parent 9b1048d commit 49dc4bc

File tree

1 file changed

+393
-0
lines changed

1 file changed

+393
-0
lines changed

‎cnn.ipynb

Lines changed: 393 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,393 @@
1+
{
2+
"cells": [
3+
{
4+
"cell_type": "markdown",
5+
"metadata": {
6+
"id": "view-in-github",
7+
"colab_type": "text"
8+
},
9+
"source": [
10+
"<a href=\"https://colab.research.google.com/github/andremicci/Generative_Deep_Learning_2nd_Edition/blob/main/cnn.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
11+
]
12+
},
13+
{
14+
"cell_type": "code",
15+
"source": [],
16+
"metadata": {
17+
"id": "Mhp4GfwTjOee"
18+
},
19+
"execution_count": null,
20+
"outputs": []
21+
},
22+
{
23+
"cell_type": "code",
24+
"source": [
25+
"!git clone https://github.com/andremicci/Generative_Deep_Learning_2nd_Edition.git"
26+
],
27+
"metadata": {
28+
"colab": {
29+
"base_uri": "https://localhost:8080/"
30+
},
31+
"id": "KOnmreJbjLIz",
32+
"outputId": "a3a49f71-3b29-41cb-fc11-a031442b32a9"
33+
},
34+
"execution_count": null,
35+
"outputs": [
36+
{
37+
"output_type": "stream",
38+
"name": "stdout",
39+
"text": [
40+
"Cloning into 'Generative_Deep_Learning_2nd_Edition'...\n",
41+
"remote: Enumerating objects: 653, done.\u001b[K\n",
42+
"remote: Counting objects: 100% (195/195), done.\u001b[K\n",
43+
"remote: Compressing objects: 100% (70/70), done.\u001b[K\n",
44+
"remote: Total 653 (delta 142), reused 125 (delta 125), pack-reused 458 (from 1)\u001b[K\n",
45+
"Receiving objects: 100% (653/653), 37.09 MiB | 8.67 MiB/s, done.\n",
46+
"Resolving deltas: 100% (377/377), done.\n"
47+
]
48+
}
49+
]
50+
},
51+
{
52+
"cell_type": "markdown",
53+
"metadata": {
54+
"id": "TFy-fuvQjJJp"
55+
},
56+
"source": [
57+
"# 🏞 Convolutional Neural Network"
58+
]
59+
},
60+
{
61+
"cell_type": "markdown",
62+
"metadata": {
63+
"id": "wEIaaUnijJJr"
64+
},
65+
"source": [
66+
"In this notebook, we'll walk through the steps required to train your own convolutional neural network (CNN) on the CIFAR dataset"
67+
]
68+
},
69+
{
70+
"cell_type": "code",
71+
"execution_count": null,
72+
"metadata": {
73+
"colab": {
74+
"base_uri": "https://localhost:8080/",
75+
"height": 350
76+
},
77+
"id": "Ewl7DskxjJJs",
78+
"outputId": "72f7b930-03ff-4bab-97c7-8f337c0da519"
79+
},
80+
"outputs": [
81+
{
82+
"output_type": "error",
83+
"ename": "ModuleNotFoundError",
84+
"evalue": "No module named 'notebooks'",
85+
"traceback": [
86+
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
87+
"\u001b[0;31mModuleNotFoundError\u001b[0m Traceback (most recent call last)",
88+
"\u001b[0;32m<ipython-input-2-a230bc8cc151>\u001b[0m in \u001b[0;36m<cell line: 0>\u001b[0;34m()\u001b[0m\n\u001b[1;32m 2\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mlayers\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodels\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizers\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mutils\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdatasets\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mnotebooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mutils\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mdisplay\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
89+
"\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'notebooks'",
90+
"",
91+
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0;32m\nNOTE: If your import is failing due to a missing package, you can\nmanually install dependencies using either !pip or !apt.\n\nTo view examples of installing some common dependencies, click the\n\"Open Examples\" button below.\n\u001b[0;31m---------------------------------------------------------------------------\u001b[0m\n"
92+
],
93+
"errorDetails": {
94+
"actions": [
95+
{
96+
"action": "open_url",
97+
"actionText": "Open Examples",
98+
"url": "/notebooks/snippets/importing_libraries.ipynb"
99+
}
100+
]
101+
}
102+
}
103+
],
104+
"source": [
105+
"import numpy as np\n",
106+
"\n",
107+
"from tensorflow.keras import layers, models, optimizers, utils, datasets\n",
108+
"from notebooks.utils import display"
109+
]
110+
},
111+
{
112+
"cell_type": "markdown",
113+
"metadata": {
114+
"tags": [],
115+
"id": "5xHzrj8jjJJu"
116+
},
117+
"source": [
118+
"## 0. Parameters <a name=\"parameters\"></a>"
119+
]
120+
},
121+
{
122+
"cell_type": "code",
123+
"execution_count": null,
124+
"metadata": {
125+
"id": "4dTmq7jjjJJv"
126+
},
127+
"outputs": [],
128+
"source": [
129+
"NUM_CLASSES = 10"
130+
]
131+
},
132+
{
133+
"cell_type": "markdown",
134+
"metadata": {
135+
"id": "MTLZ1IG6jJJw"
136+
},
137+
"source": [
138+
"## 1. Prepare the Data <a name=\"prepare\"></a>"
139+
]
140+
},
141+
{
142+
"cell_type": "code",
143+
"execution_count": null,
144+
"metadata": {
145+
"id": "DR22w6jsjJJx"
146+
},
147+
"outputs": [],
148+
"source": [
149+
"(x_train, y_train), (x_test, y_test) = datasets.cifar10.load_data()"
150+
]
151+
},
152+
{
153+
"cell_type": "code",
154+
"execution_count": null,
155+
"metadata": {
156+
"id": "gkQBN7pojJJx"
157+
},
158+
"outputs": [],
159+
"source": [
160+
"x_train = x_train.astype(\"float32\") / 255.0\n",
161+
"x_test = x_test.astype(\"float32\") / 255.0\n",
162+
"\n",
163+
"y_train = utils.to_categorical(y_train, NUM_CLASSES)\n",
164+
"y_test = utils.to_categorical(y_test, NUM_CLASSES)"
165+
]
166+
},
167+
{
168+
"cell_type": "code",
169+
"execution_count": null,
170+
"metadata": {
171+
"id": "p6D1z7qPjJJz"
172+
},
173+
"outputs": [],
174+
"source": [
175+
"display(x_train[:10])\n",
176+
"print(y_train[:10])"
177+
]
178+
},
179+
{
180+
"cell_type": "markdown",
181+
"metadata": {
182+
"id": "uqly1nGYjJJ0"
183+
},
184+
"source": [
185+
"## 2. Build the model <a name=\"build\"></a>"
186+
]
187+
},
188+
{
189+
"cell_type": "code",
190+
"execution_count": null,
191+
"metadata": {
192+
"id": "rvLwhVf7jJJ1"
193+
},
194+
"outputs": [],
195+
"source": [
196+
"input_layer = layers.Input((32, 32, 3))\n",
197+
"\n",
198+
"x = layers.Conv2D(filters=32, kernel_size=3, strides=1, padding=\"same\")(\n",
199+
" input_layer\n",
200+
")\n",
201+
"x = layers.BatchNormalization()(x)\n",
202+
"x = layers.LeakyReLU()(x)\n",
203+
"\n",
204+
"x = layers.Conv2D(filters=32, kernel_size=3, strides=2, padding=\"same\")(x)\n",
205+
"x = layers.BatchNormalization()(x)\n",
206+
"x = layers.LeakyReLU()(x)\n",
207+
"\n",
208+
"x = layers.Conv2D(filters=64, kernel_size=3, strides=1, padding=\"same\")(x)\n",
209+
"x = layers.BatchNormalization()(x)\n",
210+
"x = layers.LeakyReLU()(x)\n",
211+
"\n",
212+
"x = layers.Conv2D(filters=64, kernel_size=3, strides=2, padding=\"same\")(x)\n",
213+
"x = layers.BatchNormalization()(x)\n",
214+
"x = layers.LeakyReLU()(x)\n",
215+
"\n",
216+
"x = layers.Flatten()(x)\n",
217+
"\n",
218+
"x = layers.Dense(128)(x)\n",
219+
"x = layers.BatchNormalization()(x)\n",
220+
"x = layers.LeakyReLU()(x)\n",
221+
"x = layers.Dropout(rate=0.5)(x)\n",
222+
"\n",
223+
"x = layers.Dense(NUM_CLASSES)(x)\n",
224+
"output_layer = layers.Activation(\"softmax\")(x)\n",
225+
"\n",
226+
"model = models.Model(input_layer, output_layer)\n",
227+
"\n",
228+
"model.summary()"
229+
]
230+
},
231+
{
232+
"cell_type": "markdown",
233+
"metadata": {
234+
"tags": [],
235+
"id": "_0JEpfyWjJJ3"
236+
},
237+
"source": [
238+
"## 3. Train the model <a name=\"train\"></a>"
239+
]
240+
},
241+
{
242+
"cell_type": "code",
243+
"execution_count": null,
244+
"metadata": {
245+
"id": "RQWD-6i6jJJ4"
246+
},
247+
"outputs": [],
248+
"source": [
249+
"opt = optimizers.Adam(learning_rate=0.0005)\n",
250+
"model.compile(\n",
251+
" loss=\"categorical_crossentropy\", optimizer=opt, metrics=[\"accuracy\"]\n",
252+
")"
253+
]
254+
},
255+
{
256+
"cell_type": "code",
257+
"execution_count": null,
258+
"metadata": {
259+
"tags": [],
260+
"id": "pajMd0T0jJJ5"
261+
},
262+
"outputs": [],
263+
"source": [
264+
"model.fit(\n",
265+
" x_train,\n",
266+
" y_train,\n",
267+
" batch_size=32,\n",
268+
" epochs=10,\n",
269+
" shuffle=True,\n",
270+
" validation_data=(x_test, y_test),\n",
271+
")"
272+
]
273+
},
274+
{
275+
"cell_type": "markdown",
276+
"metadata": {
277+
"tags": [],
278+
"id": "M2Jl9j7ljJJ5"
279+
},
280+
"source": [
281+
"## 4. Evaluation <a name=\"evaluate\"></a>"
282+
]
283+
},
284+
{
285+
"cell_type": "code",
286+
"execution_count": null,
287+
"metadata": {
288+
"id": "t7GXyVvYjJJ6"
289+
},
290+
"outputs": [],
291+
"source": [
292+
"model.evaluate(x_test, y_test, batch_size=1000)"
293+
]
294+
},
295+
{
296+
"cell_type": "code",
297+
"execution_count": null,
298+
"metadata": {
299+
"id": "Ttvkm2g_jJJ6"
300+
},
301+
"outputs": [],
302+
"source": [
303+
"CLASSES = np.array(\n",
304+
" [\n",
305+
" \"airplane\",\n",
306+
" \"automobile\",\n",
307+
" \"bird\",\n",
308+
" \"cat\",\n",
309+
" \"deer\",\n",
310+
" \"dog\",\n",
311+
" \"frog\",\n",
312+
" \"horse\",\n",
313+
" \"ship\",\n",
314+
" \"truck\",\n",
315+
" ]\n",
316+
")\n",
317+
"\n",
318+
"preds = model.predict(x_test)\n",
319+
"preds_single = CLASSES[np.argmax(preds, axis=-1)]\n",
320+
"actual_single = CLASSES[np.argmax(y_test, axis=-1)]"
321+
]
322+
},
323+
{
324+
"cell_type": "code",
325+
"execution_count": null,
326+
"metadata": {
327+
"id": "5SkFQ0pHjJJ7"
328+
},
329+
"outputs": [],
330+
"source": [
331+
"import matplotlib.pyplot as plt\n",
332+
"\n",
333+
"n_to_show = 10\n",
334+
"indices = np.random.choice(range(len(x_test)), n_to_show)\n",
335+
"\n",
336+
"fig = plt.figure(figsize=(15, 3))\n",
337+
"fig.subplots_adjust(hspace=0.4, wspace=0.4)\n",
338+
"\n",
339+
"for i, idx in enumerate(indices):\n",
340+
" img = x_test[idx]\n",
341+
" ax = fig.add_subplot(1, n_to_show, i + 1)\n",
342+
" ax.axis(\"off\")\n",
343+
" ax.text(\n",
344+
" 0.5,\n",
345+
" -0.35,\n",
346+
" \"pred = \" + str(preds_single[idx]),\n",
347+
" fontsize=10,\n",
348+
" ha=\"center\",\n",
349+
" transform=ax.transAxes,\n",
350+
" )\n",
351+
" ax.text(\n",
352+
" 0.5,\n",
353+
" -0.7,\n",
354+
" \"act = \" + str(actual_single[idx]),\n",
355+
" fontsize=10,\n",
356+
" ha=\"center\",\n",
357+
" transform=ax.transAxes,\n",
358+
" )\n",
359+
" ax.imshow(img)"
360+
]
361+
}
362+
],
363+
"metadata": {
364+
"kernelspec": {
365+
"display_name": "Python 3 (ipykernel)",
366+
"language": "python",
367+
"name": "python3"
368+
},
369+
"language_info": {
370+
"codemirror_mode": {
371+
"name": "ipython",
372+
"version": 3
373+
},
374+
"file_extension": ".py",
375+
"mimetype": "text/x-python",
376+
"name": "python",
377+
"nbconvert_exporter": "python",
378+
"pygments_lexer": "ipython3",
379+
"version": "3.8.10"
380+
},
381+
"vscode": {
382+
"interpreter": {
383+
"hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
384+
}
385+
},
386+
"colab": {
387+
"provenance": [],
388+
"include_colab_link": true
389+
}
390+
},
391+
"nbformat": 4,
392+
"nbformat_minor": 0
393+
}

0 commit comments

Comments
 (0)