Skip to content

Commit

Permalink
word embeddings
Browse files Browse the repository at this point in the history
  • Loading branch information
codebasics committed Mar 23, 2021
1 parent 409e7e2 commit abab5ab
Showing 1 changed file with 317 additions and 0 deletions.
317 changes: 317 additions & 0 deletions DeepLearningML/22_word_embedding/supervised_word_embeddings.ipynb
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,317 @@
{
"cells": [
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"from tensorflow.keras.preprocessing.text import one_hot\n",
"from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
"from tensorflow.keras.models import Sequential\n",
"from tensorflow.keras.layers import Dense\n",
"from tensorflow.keras.layers import Flatten\n",
"from tensorflow.keras.layers import Embedding\n",
"\n",
"reviews = ['nice food',\n",
" 'amazing restaurant',\n",
" 'too good',\n",
" 'just loved it!',\n",
" 'will go again',\n",
" 'horrible food',\n",
" 'never go there',\n",
" 'poor service',\n",
" 'poor quality',\n",
" 'needs improvement']\n",
"\n",
"sentiment = np.array([1,1,1,1,1,0,0,0,0,0])"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"[4, 23]"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"one_hot(\"amazing restaurant\",30)"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[[13, 21], [4, 23], [14, 17], [8, 15, 16], [22, 15, 29], [8, 21], [26, 15, 24], [16, 4], [16, 12], [4, 29]]\n"
]
}
],
"source": [
"vocab_size = 30\n",
"encoded_reviews = [one_hot(d, vocab_size) for d in reviews]\n",
"print(encoded_reviews)"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[[13 21 0 0]\n",
" [ 4 23 0 0]\n",
" [14 17 0 0]\n",
" [ 8 15 16 0]\n",
" [22 15 29 0]\n",
" [ 8 21 0 0]\n",
" [26 15 24 0]\n",
" [16 4 0 0]\n",
" [16 12 0 0]\n",
" [ 4 29 0 0]]\n"
]
}
],
"source": [
"max_length = 4\n",
"padded_reviews = pad_sequences(encoded_reviews, maxlen=max_length, padding='post')\n",
"print(padded_reviews)"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"scrolled": true
},
"outputs": [],
"source": [
"embeded_vector_size = 5\n",
"\n",
"model = Sequential()\n",
"model.add(Embedding(vocab_size, embeded_vector_size, input_length=max_length,name=\"embedding\"))\n",
"model.add(Flatten())\n",
"model.add(Dense(1, activation='sigmoid'))"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {},
"outputs": [],
"source": [
"X = padded_reviews\n",
"y = sentiment"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Model: \"sequential_1\"\n",
"_________________________________________________________________\n",
"Layer (type) Output Shape Param # \n",
"=================================================================\n",
"embedding (Embedding) (None, 4, 5) 150 \n",
"_________________________________________________________________\n",
"flatten_1 (Flatten) (None, 20) 0 \n",
"_________________________________________________________________\n",
"dense_1 (Dense) (None, 1) 21 \n",
"=================================================================\n",
"Total params: 171\n",
"Trainable params: 171\n",
"Non-trainable params: 0\n",
"_________________________________________________________________\n",
"None\n"
]
}
],
"source": [
"model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
"print(model.summary())"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"<tensorflow.python.keras.callbacks.History at 0x1bb8daa5a30>"
]
},
"execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"model.fit(X, y, epochs=50, verbose=0)"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"1/1 [==============================] - 0s 1ms/step - loss: 0.6384 - accuracy: 1.0000\n"
]
},
{
"data": {
"text/plain": [
"1.0"
]
},
"execution_count": 29,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# evaluate the model\n",
"loss, accuracy = model.evaluate(X, y)\n",
"accuracy"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"30"
]
},
"execution_count": 30,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"weights = model.get_layer('embedding').get_weights()[0]\n",
"len(weights)"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([-0.08330977, -0.06752131, -0.04629624, -0.00765801, -0.02024159],\n",
" dtype=float32)"
]
},
"execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"weights[13]"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {
"scrolled": false
},
"outputs": [
{
"data": {
"text/plain": [
"array([-0.07935128, -0.08574004, 0.06615968, -0.02349528, 0.00917289],\n",
" dtype=float32)"
]
},
"execution_count": 32,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"weights[4]"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"array([ 0.0128377 , 0.03549778, 0.05134471, -0.07147218, 0.03261041],\n",
" dtype=float32)"
]
},
"execution_count": 33,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"weights[16]"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.5"
}
},
"nbformat": 4,
"nbformat_minor": 4
}

0 comments on commit abab5ab

Please sign in to comment.