Ideone.com

fork download

copy

import tensorflow as tf
from tensorflow import keras
import numpy as np
 
# Define the model architecture
model = keras.Sequential([
    keras.layers.Dense(128, input_shape=(26,), activation='relu'),
    keras.layers.Dense(64, activation='relu'),
    keras.layers.Dense(1, activation='sigmoid')
])
 
# Compile the model
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
 
# Define the training data
english_words = ['hello', 'world', 'goodbye', 'python', 'programming']
non_english_words = ['hola', 'adios', 'merci', 'gracias', 'danke']
 
# Convert the words to one-hot encoding
def word_to_one_hot(word):
    one_hot = np.zeros((26,))
    for i, letter in enumerate(word.lower()):
        one_hot[ord(letter) - 97] = 1
    return one_hot
 
english_words_one_hot = np.array([word_to_one_hot(word) for word in english_words])
non_english_words_one_hot = np.array([word_to_one_hot(word) for word in non_english_words])
 
# Create the labels for the training data
english_labels = np.ones((len(english_words),))
non_english_labels = np.zeros((len(non_english_words),))
 
# Combine the training data and labels
x_train = np.concatenate((english_words_one_hot, non_english_words_one_hot))
y_train = np.concatenate((english_labels, non_english_labels))
 
# Train the model
model.fit(x_train, y_train, epochs=50, batch_size=2)
 
# Test the model on some new words
test_words = ['hello', 'world', 'bonjour', 'merci', 'python']
test_words_one_hot = np.array([word_to_one_hot(word) for word in test_words])
predictions = model.predict(test_words_one_hot)
for i, prediction in enumerate(predictions):
    if prediction > 0.5:
        print(test_words[i], 'is English')
    else:
        print(test_words[i], 'is not English')

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

Success #stdin #stdout #stderr 2.23s 225468KB

stdin

copy

stdout

copy

Epoch 1/50

 2/10 [=====>........................] - ETA: 1s - loss: 0.5744 - acc: 1.0000
10/10 [==============================] - 0s 49ms/sample - loss: 0.7265 - acc: 0.3000
Epoch 2/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.7413 - acc: 0.0000e+00
10/10 [==============================] - 0s 568us/sample - loss: 0.6666 - acc: 0.5000
Epoch 3/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.5506 - acc: 1.0000
10/10 [==============================] - 0s 565us/sample - loss: 0.6127 - acc: 0.9000
Epoch 4/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.5623 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.5716 - acc: 1.0000
Epoch 5/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.5656 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.5309 - acc: 1.0000
Epoch 6/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.4446 - acc: 1.0000
10/10 [==============================] - 0s 559us/sample - loss: 0.4982 - acc: 1.0000
Epoch 7/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.3537 - acc: 1.0000
10/10 [==============================] - 0s 559us/sample - loss: 0.4602 - acc: 1.0000
Epoch 8/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.3103 - acc: 1.0000
10/10 [==============================] - 0s 559us/sample - loss: 0.4216 - acc: 1.0000
Epoch 9/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.4057 - acc: 1.0000
10/10 [==============================] - 0s 561us/sample - loss: 0.3850 - acc: 1.0000
Epoch 10/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.2675 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.3475 - acc: 1.0000
Epoch 11/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.3733 - acc: 1.0000
10/10 [==============================] - 0s 561us/sample - loss: 0.3122 - acc: 1.0000
Epoch 12/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.2847 - acc: 1.0000
10/10 [==============================] - 0s 565us/sample - loss: 0.2770 - acc: 1.0000
Epoch 13/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.2578 - acc: 1.0000
10/10 [==============================] - 0s 560us/sample - loss: 0.2459 - acc: 1.0000
Epoch 14/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.1861 - acc: 1.0000
10/10 [==============================] - 0s 557us/sample - loss: 0.2111 - acc: 1.0000
Epoch 15/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.2205 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.1814 - acc: 1.0000
Epoch 16/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.1684 - acc: 1.0000
10/10 [==============================] - 0s 557us/sample - loss: 0.1568 - acc: 1.0000
Epoch 17/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.1407 - acc: 1.0000
10/10 [==============================] - 0s 565us/sample - loss: 0.1369 - acc: 1.0000
Epoch 18/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.1222 - acc: 1.0000
10/10 [==============================] - 0s 554us/sample - loss: 0.1161 - acc: 1.0000
Epoch 19/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.1137 - acc: 1.0000
10/10 [==============================] - 0s 572us/sample - loss: 0.0981 - acc: 1.0000
Epoch 20/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0792 - acc: 1.0000
10/10 [==============================] - 0s 562us/sample - loss: 0.0845 - acc: 1.0000
Epoch 21/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0754 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.0720 - acc: 1.0000
Epoch 22/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0633 - acc: 1.0000
10/10 [==============================] - 0s 557us/sample - loss: 0.0630 - acc: 1.0000
Epoch 23/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0863 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.0544 - acc: 1.0000
Epoch 24/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0191 - acc: 1.0000
10/10 [==============================] - 0s 562us/sample - loss: 0.0475 - acc: 1.0000
Epoch 25/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0639 - acc: 1.0000
10/10 [==============================] - 0s 556us/sample - loss: 0.0422 - acc: 1.0000
Epoch 26/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0195 - acc: 1.0000
10/10 [==============================] - 0s 577us/sample - loss: 0.0367 - acc: 1.0000
Epoch 27/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0677 - acc: 1.0000
10/10 [==============================] - 0s 583us/sample - loss: 0.0333 - acc: 1.0000
Epoch 28/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0161 - acc: 1.0000
10/10 [==============================] - 0s 561us/sample - loss: 0.0291 - acc: 1.0000
Epoch 29/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0455 - acc: 1.0000
10/10 [==============================] - 0s 559us/sample - loss: 0.0263 - acc: 1.0000
Epoch 30/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0644 - acc: 1.0000
10/10 [==============================] - 0s 556us/sample - loss: 0.0238 - acc: 1.0000
Epoch 31/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0130 - acc: 1.0000
10/10 [==============================] - 0s 560us/sample - loss: 0.0213 - acc: 1.0000
Epoch 32/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0285 - acc: 1.0000
10/10 [==============================] - 0s 561us/sample - loss: 0.0196 - acc: 1.0000
Epoch 33/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0160 - acc: 1.0000
10/10 [==============================] - 0s 559us/sample - loss: 0.0180 - acc: 1.0000
Epoch 34/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0454 - acc: 1.0000
10/10 [==============================] - 0s 566us/sample - loss: 0.0162 - acc: 1.0000
Epoch 35/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0079 - acc: 1.0000
10/10 [==============================] - 0s 561us/sample - loss: 0.0147 - acc: 1.0000
Epoch 36/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0061 - acc: 1.0000
10/10 [==============================] - 0s 563us/sample - loss: 0.0134 - acc: 1.0000
Epoch 37/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0243 - acc: 1.0000
10/10 [==============================] - 0s 557us/sample - loss: 0.0126 - acc: 1.0000
Epoch 38/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0050 - acc: 1.0000
10/10 [==============================] - 0s 559us/sample - loss: 0.0116 - acc: 1.0000
Epoch 39/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0042 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.0109 - acc: 1.0000
Epoch 40/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0081 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.0101 - acc: 1.0000
Epoch 41/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0105 - acc: 1.0000
10/10 [==============================] - 0s 562us/sample - loss: 0.0093 - acc: 1.0000
Epoch 42/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0164 - acc: 1.0000
10/10 [==============================] - 0s 556us/sample - loss: 0.0088 - acc: 1.0000
Epoch 43/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0066 - acc: 1.0000
10/10 [==============================] - 0s 573us/sample - loss: 0.0082 - acc: 1.0000
Epoch 44/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0061 - acc: 1.0000
10/10 [==============================] - 0s 567us/sample - loss: 0.0077 - acc: 1.0000
Epoch 45/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0036 - acc: 1.0000
10/10 [==============================] - 0s 558us/sample - loss: 0.0072 - acc: 1.0000
Epoch 46/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0095 - acc: 1.0000
10/10 [==============================] - 0s 564us/sample - loss: 0.0068 - acc: 1.0000
Epoch 47/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0042 - acc: 1.0000
10/10 [==============================] - 0s 677us/sample - loss: 0.0064 - acc: 1.0000
Epoch 48/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0183 - acc: 1.0000
10/10 [==============================] - 0s 572us/sample - loss: 0.0061 - acc: 1.0000
Epoch 49/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0041 - acc: 1.0000
10/10 [==============================] - 0s 567us/sample - loss: 0.0057 - acc: 1.0000
Epoch 50/50

 2/10 [=====>........................] - ETA: 0s - loss: 0.0027 - acc: 1.0000
10/10 [==============================] - 0s 556us/sample - loss: 0.0054 - acc: 1.0000
('hello', 'is English')
('world', 'is English')
('bonjour', 'is English')
('merci', 'is not English')
('python', 'is English')

stderr

copy

WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.
WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.cast instead.

Discover > Sphere Engine API

Discover > IDE Widget

Discover > Sphere Engine API

Discover > IDE Widget

Choose your language