Skip to content

Instantly share code, notes, and snippets.

@graph226

graph226/TensorFlow+Test.ipynb

Created February 16, 2018 04:47
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save graph226/eefcaa673835916e30f338fc5a003c09 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save graph226/eefcaa673835916e30f338fc5a003c09 to your computer and use it in GitHub Desktop.
Download ZIP
えんべっでぃんぐるっくあっぷ
Display the source blob
Display the rendered blob
Raw
TensorFlow+Test.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"import tensorflow as tf\n",
"from tensorflow import contrib\n",
"import numpy as np\n",
"import pickle"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"keys = tf.Variable([\"1\", \"2\", \"3\"], dtype=tf.string)\n",
"values = tf.Variable([1, 2, 3], dtype=tf.float32)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"table = contrib.lookup.HashTable(\n",
" contrib.lookup.KeyValueTensorInitializer(keys, values), -1)\n",
"out = table.lookup(names_tf)\n",
"with tf.Session() as sess:\n",
" table.init.run()\n",
" print(out.eval())"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": true
},
"source": [
"## これはめも\n",
"\n",
"https://qiita.com/kzmssk/items/ddf2c0f956a5d26e992a"
]
},
{
"cell_type": "code",
"execution_count": 93,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"array([28918, 3294, 28918, 28918, 19501, 4847])"
]
},
"execution_count": 93,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"sess = tf.InteractiveSession()\n",
"mapping_strings = tf.constant(all_words)\n",
"table = tf.contrib.lookup.index_table_from_tensor(\n",
" mapping=mapping_strings, num_oov_buckets=1, default_value=-1)\n",
"features = tf.constant([\"emerson\", \"feel\", \"and\", \"palmer\", \"deep\", \"popcorn\"])\n",
"ids = table.lookup(features)\n",
"tf.tables_initializer().run()\n",
"\n",
"ids.eval()"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"with open('positive_example.pickle', 'rb') as f:\n",
" positive_example = pickle.load(f)\n",
"with open('all_words.pickle', 'rb') as f:\n",
" all_words = pickle.load(f)"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"TOTAL_SIZE = len(all_words)\n",
"DIM = 46\n",
"BATCH_SIZE = 2\n",
"np_x = np.random.rand(TOTAL_SIZE, DIM).astype(np.float32)\n",
"x = tf.Variable(np_x)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[[ 0.33407766 0.66344464 0.13722664 0.57478184 0.48289245 0.72210312\n",
" 0.47872588 0.71832132 0.22690746 0.9794398 0.23006117 0.77183402\n",
" 0.76896125 0.72980267 0.32537961 0.40896854 0.60130352 0.78982878\n",
" 0.03451943 0.72029352 0.63087696 0.63158554 0.25121796 0.94101858\n",
" 0.46409729 0.34709042 0.22485785 0.78404438 0.33605623 0.79317814\n",
" 0.04439238 0.22641681 0.4303872 0.34324634 0.06531932 0.80964732\n",
" 0.97386193 0.42601633 0.73470759 0.18472442 0.78394419 0.3850196\n",
" 0.20360945 0.81814951 0.34891433 0.97032487]\n",
" [ 0.36110488 0.25897187 0.23618111 0.97066313 0.38464648 0.70196772\n",
" 0.99959099 0.47679022 0.05685244 0.98868525 0.33194125 0.50208384\n",
" 0.75492752 0.96577746 0.40557855 0.46122807 0.76450753 0.0459351\n",
" 0.275621 0.24587847 0.67910224 0.63183731 0.61231363 0.17298323\n",
" 0.58105111 0.70024359 0.74448323 0.70820373 0.75346893 0.48200104\n",
" 0.85238212 0.60653716 0.52646261 0.26115027 0.15900189 0.70963442\n",
" 0.04609635 0.70944047 0.79803944 0.46148825 0.61928189 0.45201811\n",
" 0.88260871 0.56540591 0.14678472 0.64773971]]\n"
]
}
],
"source": [
"np_inds = np.array([0, 3])\n",
"inds = tf.placeholder(tf.int32, [BATCH_SIZE])\n",
"\n",
"h = tf.nn.embedding_lookup(x, inds)\n",
"with tf.Session() as sess:\n",
" sess.run(tf.global_variables_initializer())\n",
" print(sess.run(h, feed_dict={inds: np_inds}))"
]
},
{
"cell_type": "code",
"execution_count": 90,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([[ 0.61467636, 0.40071121, 0.99041861, ..., 0.77075809,\n",
" 0.20176312, 0.79317933],\n",
" [ 0.76744157, 0.52253145, 0.53163719, ..., 0.15996242,\n",
" 0.84362042, 0.72988248],\n",
" [ 0.43286699, 0.63465619, 0.28215456, ..., 0.30870688,\n",
" 0.45839456, 0.45931903],\n",
" ..., \n",
" [ 0.60267776, 0.55912673, 0.7641086 , ..., 0.53222561,\n",
" 0.79974645, 0.65310097],\n",
" [ 0.87345588, 0.31659752, 0.38947991, ..., 0.60199815,\n",
" 0.89751315, 0.31559569],\n",
" [ 0.56062859, 0.37648842, 0.68439686, ..., 0.83632123,\n",
" 0.73119318, 0.5668627 ]], dtype=float32)"
]
},
"execution_count": 90,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"np_x"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": []
}
],
"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": "3.6.1"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment