數(shù)據(jù)初始化

import tensorflow as tf from tensorflow import keras from tensorflow.keras.layers import *((x_train, y_train), (x_test, y_test)) = keras.datasets.mnist.load_data()x_train = x_train.reshape(60000, -1) y_train = keras.utils.np_utils.to_categorical(y_train)

SimpleRNN

model1 = keras.Sequential() model1.add(Embedding(input_dim=256, output_dim=5)) model1.add(SimpleRNN(units=2)) model1.add(Dense(10))model1.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"]) model1.fit(x_train, y_train, batch_size=10)

GRU

model2 = keras.Sequential() model2.add(Embedding(input_dim=256, output_dim=5)) model2.add(GRU(units=2)) model2.add(Dense(10))model2.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"]) model2.fit(x_train, y_train, batch_size=10)

LSTM

model3 = keras.Sequential() model3.add(Embedding(input_dim=256, output_dim=5)) model3.add(LSTM(units=2)) model3.add(Dense(10))model3.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"]) model3.fit(x_train, y_train, batch_size=10)

encoder-decoder

model3 = keras.Sequential() model3.add(Embedding(input_dim=256, output_dim=5)) model3.add(LSTM(units=2)) model3.add(Dense(10))model3.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"]) model3.fit(x_train, y_train, batch_size=10)

總結(jié)

以上是生活随笔為你收集整理的tf 从RNN到BERT的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺(jué)得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。