40. Tensorflow를 통한 논리구현

텐서플로우를 통한 OR게이트

import tensorflow as tf
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import SGD
from tensorflow.keras.losses import mse
tf.random.set_seed(777)
# 데이터
data = np.array([[0,0],[1,0],[0,1],[1,1]])
# 라벨링
label = np.array([[0],[1],[1],[1]])
model = Sequential()
model.add(Dense(1, input_shape = (2,), activation = 'linear')) # 퍼셉트론
# 모델설정 
# GD 경사하강법, SGD stochastic :minibatch
# loss 손실함수, 비용함수
model.compile(optimizer = SGD(), loss = mse, metrics = ['acc'])
# epochs 100으로 하면 학습하다가 관둬서, 200으로 함
model.fit(data, label, epochs = 200)
# 그 값은 이거야
model.get_weights()
# 머신러닝은 값 하나하나 매겨주는데 딥러닝은 라벨링하고 주면 알아서 찾아감
model.predict(data)
model.evaluate(data, label) # 평가, 손실함수, 정확도

Epoch 1/200
1/1 [==============================] - 1s 1s/step - loss: 1.4290 - acc: 0.5000
Epoch 2/200
1/1 [==============================] - 0s 3ms/step - loss: 1.3602 - acc: 0.5000
Epoch 3/200
1/1 [==============================] - 0s 2ms/step - loss: 1.2956 - acc: 0.5000
Epoch 4/200
1/1 [==============================] - 0s 2ms/step - loss: 1.2349 - acc: 0.5000
Epoch 5/200
1/1 [==============================] - 0s 2ms/step - loss: 1.1779 - acc: 0.5000
Epoch 6/200
1/1 [==============================] - 0s 3ms/step - loss: 1.1242 - acc: 0.5000
Epoch 7/200
1/1 [==============================] - 0s 2ms/step - loss: 1.0738 - acc: 0.5000
Epoch 8/200
1/1 [==============================] - 0s 3ms/step - loss: 1.0264 - acc: 0.5000
Epoch 9/200
1/1 [==============================] - 0s 3ms/step - loss: 0.9819 - acc: 0.5000
Epoch 10/200
1/1 [==============================] - 0s 2ms/step - loss: 0.9399 - acc: 0.5000
Epoch 11/200
1/1 [==============================] - 0s 3ms/step - loss: 0.9005 - acc: 0.5000
Epoch 12/200
1/1 [==============================] - 0s 2ms/step - loss: 0.8634 - acc: 0.5000
Epoch 13/200
1/1 [==============================] - 0s 2ms/step - loss: 0.8284 - acc: 0.5000
Epoch 14/200
1/1 [==============================] - 0s 3ms/step - loss: 0.7955 - acc: 0.5000
Epoch 15/200
1/1 [==============================] - 0s 2ms/step - loss: 0.7646 - acc: 0.5000
Epoch 16/200
1/1 [==============================] - 0s 2ms/step - loss: 0.7354 - acc: 0.5000
Epoch 17/200
1/1 [==============================] - 0s 1000us/step - loss: 0.7079 - acc: 0.5000
Epoch 18/200
1/1 [==============================] - 0s 2ms/step - loss: 0.6820 - acc: 0.5000
Epoch 19/200
1/1 [==============================] - 0s 2ms/step - loss: 0.6576 - acc: 0.5000
Epoch 20/200
1/1 [==============================] - 0s 3ms/step - loss: 0.6346 - acc: 0.5000
Epoch 21/200
1/1 [==============================] - 0s 2ms/step - loss: 0.6129 - acc: 0.5000
Epoch 22/200
1/1 [==============================] - 0s 2ms/step - loss: 0.5925 - acc: 0.5000
Epoch 23/200
1/1 [==============================] - 0s 2ms/step - loss: 0.5732 - acc: 0.5000
Epoch 24/200
1/1 [==============================] - 0s 2ms/step - loss: 0.5549 - acc: 0.5000
Epoch 25/200
1/1 [==============================] - 0s 2ms/step - loss: 0.5377 - acc: 0.5000
Epoch 26/200
1/1 [==============================] - 0s 2ms/step - loss: 0.5215 - acc: 0.5000
Epoch 27/200
1/1 [==============================] - 0s 2ms/step - loss: 0.5061 - acc: 0.5000
Epoch 28/200
1/1 [==============================] - 0s 3ms/step - loss: 0.4916 - acc: 0.5000
Epoch 29/200
1/1 [==============================] - 0s 2ms/step - loss: 0.4778 - acc: 0.5000
Epoch 30/200
1/1 [==============================] - 0s 3ms/step - loss: 0.4648 - acc: 0.5000
Epoch 31/200
1/1 [==============================] - 0s 2ms/step - loss: 0.4525 - acc: 0.7500
Epoch 32/200
1/1 [==============================] - 0s 2ms/step - loss: 0.4409 - acc: 0.7500
Epoch 33/200
1/1 [==============================] - 0s 2ms/step - loss: 0.4298 - acc: 0.7500
Epoch 34/200
1/1 [==============================] - 0s 2ms/step - loss: 0.4193 - acc: 0.7500
Epoch 35/200
1/1 [==============================] - 0s 2ms/step - loss: 0.4094 - acc: 0.7500
Epoch 36/200
1/1 [==============================] - 0s 3ms/step - loss: 0.4000 - acc: 0.7500
Epoch 37/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3911 - acc: 0.7500
Epoch 38/200
1/1 [==============================] - 0s 3ms/step - loss: 0.3826 - acc: 0.7500
Epoch 39/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3745 - acc: 0.7500
Epoch 40/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3668 - acc: 0.7500
Epoch 41/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3595 - acc: 0.7500
Epoch 42/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3525 - acc: 0.7500
Epoch 43/200
1/1 [==============================] - 0s 3ms/step - loss: 0.3459 - acc: 0.7500
Epoch 44/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3396 - acc: 0.7500
Epoch 45/200
1/1 [==============================] - 0s 3ms/step - loss: 0.3336 - acc: 0.7500
Epoch 46/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3278 - acc: 0.7500
Epoch 47/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3223 - acc: 0.7500
Epoch 48/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3170 - acc: 0.7500
Epoch 49/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3120 - acc: 0.7500
Epoch 50/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3072 - acc: 0.7500
Epoch 51/200
1/1 [==============================] - 0s 2ms/step - loss: 0.3026 - acc: 0.7500
Epoch 52/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2982 - acc: 0.7500
Epoch 53/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2939 - acc: 0.7500
Epoch 54/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2898 - acc: 0.7500
Epoch 55/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2859 - acc: 0.7500
Epoch 56/200
1/1 [==============================] - 0s 1ms/step - loss: 0.2822 - acc: 0.7500
Epoch 57/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2785 - acc: 0.7500
Epoch 58/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2750 - acc: 0.7500
Epoch 59/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2717 - acc: 0.7500
Epoch 60/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2684 - acc: 0.7500
Epoch 61/200
1/1 [==============================] - 0s 3ms/step - loss: 0.2653 - acc: 0.7500
Epoch 62/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2623 - acc: 0.7500
Epoch 63/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2594 - acc: 0.7500
Epoch 64/200
1/1 [==============================] - 0s 1ms/step - loss: 0.2565 - acc: 0.7500
Epoch 65/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2538 - acc: 0.7500
Epoch 66/200
1/1 [==============================] - 0s 3ms/step - loss: 0.2511 - acc: 0.7500
Epoch 67/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2486 - acc: 0.7500
Epoch 68/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2461 - acc: 0.7500
Epoch 69/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2436 - acc: 0.7500
Epoch 70/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2413 - acc: 0.7500
Epoch 71/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2390 - acc: 0.7500
Epoch 72/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2368 - acc: 0.7500
Epoch 73/200
1/1 [==============================] - 0s 3ms/step - loss: 0.2346 - acc: 0.7500
Epoch 74/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2325 - acc: 0.7500
Epoch 75/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2304 - acc: 0.7500
Epoch 76/200
1/1 [==============================] - 0s 3ms/step - loss: 0.2284 - acc: 0.7500
Epoch 77/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2264 - acc: 0.7500
Epoch 78/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2245 - acc: 0.7500
Epoch 79/200
1/1 [==============================] - 0s 1ms/step - loss: 0.2226 - acc: 0.7500
Epoch 80/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2208 - acc: 0.7500
Epoch 81/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2190 - acc: 0.7500
Epoch 82/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2173 - acc: 0.7500
Epoch 83/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2155 - acc: 0.7500
Epoch 84/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2139 - acc: 0.7500
Epoch 85/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2122 - acc: 0.7500
Epoch 86/200
1/1 [==============================] - 0s 3ms/step - loss: 0.2106 - acc: 0.7500
Epoch 87/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2090 - acc: 0.7500
Epoch 88/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2074 - acc: 0.7500
Epoch 89/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2059 - acc: 0.7500
Epoch 90/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2044 - acc: 0.7500
Epoch 91/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2029 - acc: 0.7500
Epoch 92/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2015 - acc: 0.7500
Epoch 93/200
1/1 [==============================] - 0s 2ms/step - loss: 0.2000 - acc: 0.7500
Epoch 94/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1986 - acc: 0.7500
Epoch 95/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1973 - acc: 0.7500
Epoch 96/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1959 - acc: 0.7500
Epoch 97/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1946 - acc: 0.7500
Epoch 98/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1932 - acc: 0.7500
Epoch 99/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1919 - acc: 0.7500
Epoch 100/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1907 - acc: 0.7500
Epoch 101/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1894 - acc: 0.7500
Epoch 102/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1881 - acc: 0.7500
Epoch 103/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1869 - acc: 0.7500
Epoch 104/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1857 - acc: 0.7500
Epoch 105/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1845 - acc: 0.7500
Epoch 106/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1833 - acc: 0.7500
Epoch 107/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1822 - acc: 0.7500
Epoch 108/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1810 - acc: 0.7500
Epoch 109/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1799 - acc: 0.7500
Epoch 110/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1787 - acc: 0.7500
Epoch 111/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1776 - acc: 0.7500
Epoch 112/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1765 - acc: 0.7500
Epoch 113/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1755 - acc: 0.7500
Epoch 114/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1744 - acc: 0.7500
Epoch 115/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1733 - acc: 0.7500
Epoch 116/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1723 - acc: 0.7500
Epoch 117/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1712 - acc: 0.7500
Epoch 118/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1702 - acc: 0.7500
Epoch 119/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1692 - acc: 0.7500
Epoch 120/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1682 - acc: 0.7500
Epoch 121/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1672 - acc: 0.7500
Epoch 122/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1663 - acc: 0.7500
Epoch 123/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1653 - acc: 0.7500
Epoch 124/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1643 - acc: 0.7500
Epoch 125/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1634 - acc: 0.7500
Epoch 126/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1625 - acc: 0.7500
Epoch 127/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1615 - acc: 0.7500
Epoch 128/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1606 - acc: 0.7500
Epoch 129/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1597 - acc: 0.7500
Epoch 130/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1588 - acc: 0.7500
Epoch 131/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1579 - acc: 0.7500
Epoch 132/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1570 - acc: 0.7500
Epoch 133/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1562 - acc: 0.7500
Epoch 134/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1553 - acc: 0.7500
Epoch 135/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1545 - acc: 0.7500
Epoch 136/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1536 - acc: 0.7500
Epoch 137/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1528 - acc: 0.7500
Epoch 138/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1520 - acc: 0.7500
Epoch 139/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1511 - acc: 0.7500
Epoch 140/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1503 - acc: 0.7500
Epoch 141/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1495 - acc: 0.7500
Epoch 142/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1487 - acc: 0.7500
Epoch 143/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1480 - acc: 0.7500
Epoch 144/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1472 - acc: 0.7500
Epoch 145/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1464 - acc: 0.7500
Epoch 146/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1456 - acc: 0.7500
Epoch 147/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1449 - acc: 0.7500
Epoch 148/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1441 - acc: 0.7500
Epoch 149/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1434 - acc: 0.7500
Epoch 150/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1427 - acc: 0.7500
Epoch 151/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1419 - acc: 0.7500
Epoch 152/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1412 - acc: 0.7500
Epoch 153/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1405 - acc: 0.7500
Epoch 154/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1398 - acc: 0.7500
Epoch 155/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1391 - acc: 0.7500
Epoch 156/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1384 - acc: 0.7500
Epoch 157/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1377 - acc: 0.7500
Epoch 158/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1370 - acc: 0.7500
Epoch 159/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1364 - acc: 0.7500
Epoch 160/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1357 - acc: 0.7500
Epoch 161/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1350 - acc: 0.7500
Epoch 162/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1344 - acc: 0.7500
Epoch 163/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1337 - acc: 0.7500
Epoch 164/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1331 - acc: 0.7500
Epoch 165/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1325 - acc: 0.7500
Epoch 166/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1318 - acc: 0.7500
Epoch 167/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1312 - acc: 0.7500
Epoch 168/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1306 - acc: 0.7500
Epoch 169/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1300 - acc: 0.7500
Epoch 170/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1294 - acc: 0.7500
Epoch 171/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1288 - acc: 0.7500
Epoch 172/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1282 - acc: 0.7500
Epoch 173/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1276 - acc: 0.7500
Epoch 174/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1270 - acc: 0.7500
Epoch 175/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1264 - acc: 0.7500
Epoch 176/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1258 - acc: 0.7500
Epoch 177/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1253 - acc: 0.7500
Epoch 178/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1247 - acc: 0.7500
Epoch 179/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1242 - acc: 0.7500
Epoch 180/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1236 - acc: 0.7500
Epoch 181/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1231 - acc: 0.7500
Epoch 182/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1225 - acc: 0.7500
Epoch 183/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1220 - acc: 0.7500
Epoch 184/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1214 - acc: 0.7500
Epoch 185/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1209 - acc: 0.7500
Epoch 186/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1204 - acc: 0.7500
Epoch 187/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1199 - acc: 0.7500
Epoch 188/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1194 - acc: 0.7500
Epoch 189/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1189 - acc: 0.7500
Epoch 190/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1184 - acc: 0.7500
Epoch 191/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1179 - acc: 0.7500
Epoch 192/200
1/1 [==============================] - 0s 3ms/step - loss: 0.1174 - acc: 1.0000
Epoch 193/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1169 - acc: 1.0000
Epoch 194/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1164 - acc: 1.0000
Epoch 195/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1159 - acc: 1.0000
Epoch 196/200
1/1 [==============================] - 0s 1ms/step - loss: 0.1154 - acc: 1.0000
Epoch 197/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1150 - acc: 1.0000
Epoch 198/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1145 - acc: 1.0000
Epoch 199/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1140 - acc: 1.0000
Epoch 200/200
1/1 [==============================] - 0s 2ms/step - loss: 0.1136 - acc: 1.0000
1/1 [==============================] - 0s 78ms/step - loss: 0.1131 - acc: 1.0000


[0.11312820017337799, 1.0]

 

텐서플로우를 통한 AND게이트

import tensorflow as tf
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import SGD
from tensorflow.keras.losses import mse
tf.random.set_seed(777)
# 데이터
data = np.array([[0,0],[1,0],[0,1],[1,1]])
# 라벨링
label = np.array([[0],[0],[0],[1]]) # AND 구현
model = Sequential()

model.add(Dense(1, input_shape = (2,), activation = 'linear')) # 퍼셉트론
model.compile(optimizer = SGD(), loss = mse, metrics = ['acc'])
model.fit(data, label, epochs = 2000, verbose = 0)
model.get_weights()
model.predict(data).flatten()
model.evaluate(data, label) # 평가, 손실함수, 정확도

1/1 [==============================] - 0s 53ms/step - loss: 0.0625 - acc: 1.0000
[0.06250044703483582, 1.0]

 

from tensorflow import keras
import numpy
x = numpy.array([0,1,2,3,4])
y = x*2+1 # [1,3,5,7,9]
model = keras.models.Sequential()
# 1개층 만들기
model.add(keras.layers.Dense(1, input_shape=(1,)))
# 로스함수 : mse, 활성화함수는 default linear
model.compile('SGD','mse')
# verbose = 0 학습내용 안보여줌
model.fit(x[:2], y[:2], epochs=1000, verbose=0)
model.get_weights()
# [array([[1.9739313]], dtype=float32), 가중치 : array([1.0161117], dtype=float32)]
# 0, 1 중 어디에 더가까운가?

# 이후값
model.predict(x[2:])

array([[5.000806],
       [7.001389],
       [9.001972]], dtype=float32)

 

model.predict([5])

# array([[11.002556]], dtype=float32)

 

# 백터 내적
from tensorflow import keras
import numpy
# 난수발생기, 10행 5열 실수형 난수
#              난수의 갯수분포가 균등하게 생성
x = tf.random.uniform((10,5)) # 세로랑
w = tf.random.uniform((5,3)) # 가로랑 곱
# 행렬 곱
d = tf.matmul(x, w)
print(d.shape)

# (10, 3)

 

x

<tf.Tensor: shape=(10, 5), dtype=float32, numpy=
array([[0.1357429 , 0.07509017, 0.2639438 , 0.47604764, 0.39591897],
       [0.14548802, 0.17393434, 0.00936472, 0.8090905 , 0.617025  ],
       [0.8713819 , 0.558359  , 0.17226672, 0.50340676, 0.18701088],
       [0.9073597 , 0.717615  , 0.38108468, 0.8958354 , 0.59624827],
       [0.77847326, 0.4488796 , 0.14225698, 0.8686327 , 0.03972971],
       [0.3629743 , 0.55276537, 0.3255931 , 0.5238236 , 0.05080891],
       [0.01347697, 0.3558432 , 0.77311885, 0.48737752, 0.5625943 ],
       [0.02250803, 0.8551339 , 0.36489332, 0.5632981 , 0.09144831],
       [0.25097954, 0.5333061 , 0.426386  , 0.19805324, 0.28281295],
       [0.99601805, 0.4646746 , 0.0783782 , 0.66289246, 0.17973018]],
      dtype=float32)>

 

# LINEAR로 구현이 안됨
# 텐서플로우를 통한 XOR게이트
import tensorflow as tf
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import SGD
from tensorflow.keras.losses import mse
tf.random.set_seed(777)
# 데이터
data = np.array([[0,0],[1,0],[0,1],[1,1]])
# 라벨링
label = np.array([[0],[1],[1],[0]])
model = Sequential()

model.add(Dense(1, input_shape = (2,), activation = 'linear')) # 퍼셉트론
model.compile(optimizer = SGD(), loss = mse, metrics = ['acc'])
model.fit(data, label, epochs = 2000, verbose = 0)
model.get_weights()
model.predict(data).flatten()
model.evaluate(data, label) # 평가, 손실함수, 정확도
# 1/1 [==============================] - 0s 52ms/step - loss: 0.2500 - acc: 0.5000
# 손실도 0.25, 정확도 0.5 => LINEAR로 는 실패, 

1/1 [==============================] - 0s 52ms/step - loss: 0.2500 - acc: 0.5000
[0.25, 0.5]

 

텐서플로우를 통한 XOR게이트

import tensorflow as tf
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import RMSprop, SGD
from tensorflow.keras.losses import mse
tf.random.set_seed(777)
# 데이터
data = np.array([[0,0],[1,0],[0,1],[1,1]])
# 라벨링
label = np.array([[0],[1],[1],[0]])
model = Sequential()

# 2개층
# 32개층, 활성화함수 relu
# relu 음수는 0, 양수는 그대로
model.add(Dense(32, input_shape = (2,), activation = 'relu')) # 퍼셉트론
# sigmoid는 0 ~ 1.0 이상의 값리턴
model.add(Dense(1, activation = 'sigmoid')) # 퍼셉트론

# optimizer 최적 위치 찾아가는 방식
# RMSprop : adagrad 알고리즘 보완 : 이전값을 참조해서, 해당 보폭을 찾음
# adagrad : 처음 첩근시 큰보폭, 가본곳은 작은 보폭 =
#            많은 변동시에 학습률이 감소될 수 있음


# model.compile(optimizer = RMSprop(), loss = mse, metrics = ['acc'])
model.compile(optimizer = SGD(), loss = mse, metrics = ['acc'])
model.fit(data, label, epochs = 100)
model.get_weights()
predict = model.predict(data).flatten()
model.evaluate(data, label) # 평가, 손실함수, 정확도
print(predict)
# 1/1 [==============================] - 0s 56ms/step - loss: 0.2106 - acc: 1.0000
# [0.48657197 0.54643464 0.55219495 0.44657207]

Epoch 1/100
1/1 [==============================] - 0s 163ms/step - loss: 0.2646 - acc: 0.5000
Epoch 2/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2644 - acc: 0.2500
Epoch 3/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2643 - acc: 0.2500
Epoch 4/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2642 - acc: 0.2500
Epoch 5/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2640 - acc: 0.2500
Epoch 6/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2639 - acc: 0.2500
Epoch 7/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2638 - acc: 0.2500
Epoch 8/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2637 - acc: 0.2500
Epoch 9/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2635 - acc: 0.2500
Epoch 10/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2634 - acc: 0.2500
Epoch 11/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2633 - acc: 0.2500
Epoch 12/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2632 - acc: 0.2500
Epoch 13/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2630 - acc: 0.2500
Epoch 14/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2629 - acc: 0.2500
Epoch 15/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2628 - acc: 0.2500
Epoch 16/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2627 - acc: 0.2500
Epoch 17/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2625 - acc: 0.2500
Epoch 18/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2624 - acc: 0.2500
Epoch 19/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2623 - acc: 0.2500
Epoch 20/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2622 - acc: 0.2500
Epoch 21/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2620 - acc: 0.2500
Epoch 22/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2619 - acc: 0.2500
Epoch 23/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2618 - acc: 0.2500
Epoch 24/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2617 - acc: 0.2500
Epoch 25/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2616 - acc: 0.2500
Epoch 26/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2614 - acc: 0.2500
Epoch 27/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2613 - acc: 0.2500
Epoch 28/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2612 - acc: 0.2500
Epoch 29/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2611 - acc: 0.2500
Epoch 30/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2610 - acc: 0.2500
Epoch 31/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2608 - acc: 0.2500
Epoch 32/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2607 - acc: 0.2500
Epoch 33/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2606 - acc: 0.2500
Epoch 34/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2605 - acc: 0.2500
Epoch 35/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2604 - acc: 0.2500
Epoch 36/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2602 - acc: 0.2500
Epoch 37/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2601 - acc: 0.2500
Epoch 38/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2600 - acc: 0.2500
Epoch 39/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2599 - acc: 0.2500
Epoch 40/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2598 - acc: 0.2500
Epoch 41/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2597 - acc: 0.2500
Epoch 42/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2596 - acc: 0.2500
Epoch 43/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2594 - acc: 0.2500
Epoch 44/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2593 - acc: 0.2500
Epoch 45/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2592 - acc: 0.2500
Epoch 46/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2591 - acc: 0.2500
Epoch 47/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2590 - acc: 0.2500
Epoch 48/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2589 - acc: 0.2500
Epoch 49/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2588 - acc: 0.2500
Epoch 50/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2586 - acc: 0.2500
Epoch 51/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2585 - acc: 0.2500
Epoch 52/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2584 - acc: 0.2500
Epoch 53/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2583 - acc: 0.2500
Epoch 54/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2582 - acc: 0.2500
Epoch 55/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2581 - acc: 0.2500
Epoch 56/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2580 - acc: 0.2500
Epoch 57/100
1/1 [==============================] - 0s 5ms/step - loss: 0.2579 - acc: 0.2500
Epoch 58/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2578 - acc: 0.2500
Epoch 59/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2576 - acc: 0.2500
Epoch 60/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2575 - acc: 0.2500
Epoch 61/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2574 - acc: 0.2500
Epoch 62/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2573 - acc: 0.2500
Epoch 63/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2572 - acc: 0.2500
Epoch 64/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2571 - acc: 0.2500
Epoch 65/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2570 - acc: 0.2500
Epoch 66/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2569 - acc: 0.2500
Epoch 67/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2568 - acc: 0.2500
Epoch 68/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2567 - acc: 0.2500
Epoch 69/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2566 - acc: 0.2500
Epoch 70/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2565 - acc: 0.2500
Epoch 71/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2563 - acc: 0.2500
Epoch 72/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2562 - acc: 0.2500
Epoch 73/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2561 - acc: 0.2500
Epoch 74/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2560 - acc: 0.2500
Epoch 75/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2559 - acc: 0.2500
Epoch 76/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2558 - acc: 0.2500
Epoch 77/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2557 - acc: 0.2500
Epoch 78/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2556 - acc: 0.2500
Epoch 79/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2555 - acc: 0.2500
Epoch 80/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2554 - acc: 0.2500
Epoch 81/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2553 - acc: 0.2500
Epoch 82/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2552 - acc: 0.2500
Epoch 83/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2551 - acc: 0.2500
Epoch 84/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2550 - acc: 0.2500
Epoch 85/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2549 - acc: 0.2500
Epoch 86/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2548 - acc: 0.2500
Epoch 87/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2547 - acc: 0.2500
Epoch 88/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2546 - acc: 0.2500
Epoch 89/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2545 - acc: 0.2500
Epoch 90/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2544 - acc: 0.2500
Epoch 91/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2543 - acc: 0.2500
Epoch 92/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2542 - acc: 0.2500
Epoch 93/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2541 - acc: 0.2500
Epoch 94/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2540 - acc: 0.2500
Epoch 95/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2539 - acc: 0.2500
Epoch 96/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2538 - acc: 0.2500
Epoch 97/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2537 - acc: 0.2500
Epoch 98/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2536 - acc: 0.2500
Epoch 99/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2535 - acc: 0.2500
Epoch 100/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2534 - acc: 0.2500
WARNING:tensorflow:8 out of the last 9 calls to <function Model.make_predict_function.<locals>.predict_function at 0x0000026102109700> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
WARNING:tensorflow:7 out of the last 7 calls to <function Model.make_test_function.<locals>.test_function at 0x0000026102109B80> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
1/1 [==============================] - 0s 56ms/step - loss: 0.2533 - acc: 0.2500
[0.50530165 0.44862053 0.49225846 0.442587  ]

 

import tensorflow as tf
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import RMSprop, SGD
from tensorflow.keras.losses import mse
tf.random.set_seed(777)
# 데이터
data = np.array([[0,0],[1,0],[0,1],[1,1]])
# 라벨링
label = np.array([[0],[1],[1],[0]])
model = Sequential()

# 2개층
# 32개층, 활성화함수 relu
# Dense 층쌓기
# relu 음수는 0, 양수는 그대로
model.add(Dense(32, input_shape = (2,), activation = 'relu')) # 퍼셉트론
# sigmoid는 0 ~ 1.0 이상의 값리턴
model.add(Dense(1, activation = 'sigmoid')) # 퍼셉트론

# optimizer 최적 위치 찾아가는 방식
# RMSprop : adagrad 알고리즘 보완 : 이전값을 참조해서, 해당 보폭을 찾음
# adagrad : 처음 첩근시 큰보폭, 가본곳은 작은 보폭 =
#            많은 변동시에 학습률이 감소될 수 있음


# model.compile(optimizer = RMSprop(), loss = mse, metrics = ['acc'])
model.compile(optimizer = SGD(), loss = mse, metrics = ['acc'])
model.fit(data, label, epochs = 100)
model.get_weights()
predict = model.predict(data).flatten()
model.evaluate(data, label) # 평가, 손실함수, 정확도
print(predict)
# 1/1 [==============================] - 0s 249ms/step - loss: 0.2533 - acc: 0.2500
# [0.50530165 0.44862053 0.49225846 0.442587  ]

Epoch 1/100
1/1 [==============================] - 0s 161ms/step - loss: 0.2646 - acc: 0.5000
Epoch 2/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2644 - acc: 0.2500
Epoch 3/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2643 - acc: 0.2500
Epoch 4/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2642 - acc: 0.2500
Epoch 5/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2640 - acc: 0.2500
Epoch 6/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2639 - acc: 0.2500
Epoch 7/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2638 - acc: 0.2500
Epoch 8/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2637 - acc: 0.2500
Epoch 9/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2635 - acc: 0.2500
Epoch 10/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2634 - acc: 0.2500
Epoch 11/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2633 - acc: 0.2500
Epoch 12/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2632 - acc: 0.2500
Epoch 13/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2630 - acc: 0.2500
Epoch 14/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2629 - acc: 0.2500
Epoch 15/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2628 - acc: 0.2500
Epoch 16/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2627 - acc: 0.2500
Epoch 17/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2625 - acc: 0.2500
Epoch 18/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2624 - acc: 0.2500
Epoch 19/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2623 - acc: 0.2500
Epoch 20/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2622 - acc: 0.2500
Epoch 21/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2620 - acc: 0.2500
Epoch 22/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2619 - acc: 0.2500
Epoch 23/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2618 - acc: 0.2500
Epoch 24/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2617 - acc: 0.2500
Epoch 25/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2616 - acc: 0.2500
Epoch 26/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2614 - acc: 0.2500
Epoch 27/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2613 - acc: 0.2500
Epoch 28/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2612 - acc: 0.2500
Epoch 29/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2611 - acc: 0.2500
Epoch 30/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2610 - acc: 0.2500
Epoch 31/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2608 - acc: 0.2500
Epoch 32/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2607 - acc: 0.2500
Epoch 33/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2606 - acc: 0.2500
Epoch 34/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2605 - acc: 0.2500
Epoch 35/100
1/1 [==============================] - 0s 8ms/step - loss: 0.2604 - acc: 0.2500
Epoch 36/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2602 - acc: 0.2500
Epoch 37/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2601 - acc: 0.2500
Epoch 38/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2600 - acc: 0.2500
Epoch 39/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2599 - acc: 0.2500
Epoch 40/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2598 - acc: 0.2500
Epoch 41/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2597 - acc: 0.2500
Epoch 42/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2596 - acc: 0.2500
Epoch 43/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2594 - acc: 0.2500
Epoch 44/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2593 - acc: 0.2500
Epoch 45/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2592 - acc: 0.2500
Epoch 46/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2591 - acc: 0.2500
Epoch 47/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2590 - acc: 0.2500
Epoch 48/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2589 - acc: 0.2500
Epoch 49/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2588 - acc: 0.2500
Epoch 50/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2586 - acc: 0.2500
Epoch 51/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2585 - acc: 0.2500
Epoch 52/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2584 - acc: 0.2500
Epoch 53/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2583 - acc: 0.2500
Epoch 54/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2582 - acc: 0.2500
Epoch 55/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2581 - acc: 0.2500
Epoch 56/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2580 - acc: 0.2500
Epoch 57/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2579 - acc: 0.2500
Epoch 58/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2578 - acc: 0.2500
Epoch 59/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2576 - acc: 0.2500
Epoch 60/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2575 - acc: 0.2500
Epoch 61/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2574 - acc: 0.2500
Epoch 62/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2573 - acc: 0.2500
Epoch 63/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2572 - acc: 0.2500
Epoch 64/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2571 - acc: 0.2500
Epoch 65/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2570 - acc: 0.2500
Epoch 66/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2569 - acc: 0.2500
Epoch 67/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2568 - acc: 0.2500
Epoch 68/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2567 - acc: 0.2500
Epoch 69/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2566 - acc: 0.2500
Epoch 70/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2565 - acc: 0.2500
Epoch 71/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2563 - acc: 0.2500
Epoch 72/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2562 - acc: 0.2500
Epoch 73/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2561 - acc: 0.2500
Epoch 74/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2560 - acc: 0.2500
Epoch 75/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2559 - acc: 0.2500
Epoch 76/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2558 - acc: 0.2500
Epoch 77/100
1/1 [==============================] - 0s 4ms/step - loss: 0.2557 - acc: 0.2500
Epoch 78/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2556 - acc: 0.2500
Epoch 79/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2555 - acc: 0.2500
Epoch 80/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2554 - acc: 0.2500
Epoch 81/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2553 - acc: 0.2500
Epoch 82/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2552 - acc: 0.2500
Epoch 83/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2551 - acc: 0.2500
Epoch 84/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2550 - acc: 0.2500
Epoch 85/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2549 - acc: 0.2500
Epoch 86/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2548 - acc: 0.2500
Epoch 87/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2547 - acc: 0.2500
Epoch 88/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2546 - acc: 0.2500
Epoch 89/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2545 - acc: 0.2500
Epoch 90/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2544 - acc: 0.2500
Epoch 91/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2543 - acc: 0.2500
Epoch 92/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2542 - acc: 0.2500
Epoch 93/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2541 - acc: 0.2500
Epoch 94/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2540 - acc: 0.2500
Epoch 95/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2539 - acc: 0.2500
Epoch 96/100
1/1 [==============================] - 0s 3ms/step - loss: 0.2538 - acc: 0.2500
Epoch 97/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2537 - acc: 0.2500
Epoch 98/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2536 - acc: 0.2500
Epoch 99/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2535 - acc: 0.2500
Epoch 100/100
1/1 [==============================] - 0s 2ms/step - loss: 0.2534 - acc: 0.2500
WARNING:tensorflow:9 out of the last 10 calls to <function Model.make_predict_function.<locals>.predict_function at 0x000002610365D0D0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
WARNING:tensorflow:8 out of the last 8 calls to <function Model.make_test_function.<locals>.test_function at 0x000002610365DEE0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
1/1 [==============================] - 0s 249ms/step - loss: 0.2533 - acc: 0.2500
[0.50530165 0.44862053 0.49225846 0.442587  ]