您好,登錄后才能下訂單哦!
這篇文章主要介紹“Python怎么實(shí)現(xiàn)多對(duì)多網(wǎng)絡(luò)結(jié)構(gòu)”的相關(guān)知識(shí),小編通過實(shí)際案例向大家展示操作過程,操作方法簡(jiǎn)單快捷,實(shí)用性強(qiáng),希望這篇“Python怎么實(shí)現(xiàn)多對(duì)多網(wǎng)絡(luò)結(jié)構(gòu)”文章能幫助大家解決問題。
多對(duì)多網(wǎng)絡(luò)結(jié)構(gòu):
由于這次我們將預(yù)測(cè)一個(gè)序列而不是下一個(gè)標(biāo)記,因此 y
也應(yīng)該是一個(gè)序列。y
是從 X
左移 1
的序列。
NUM_INPUT_TOKENS = 10
step = 3
sequences = []
for i in range(0, len(tokenized) - NUM_INPUT_TOKENS-1, step):
sequences.append(tokenized[i: i + NUM_INPUT_TOKENS+1])
print('# of training sequences:', len(sequences))
X_temp = np.zeros((len(sequences), NUM_INPUT_TOKENS + 1, len(uniqueTokens)), dtype=np.bool)
X = np.zeros((len(sequences), NUM_INPUT_TOKENS, len(uniqueTokens)), dtype=np.bool)
y = np.zeros((len(sequences), NUM_INPUT_TOKENS, len(uniqueTokens)), dtype=np.bool)
for i, sequence in enumerate(sequences):
for t, char in enumerate(sequence):
X_temp[i, t, token_indices[char]] = 1
num_sequences = len(X_temp)
for i, vec in enumerate(X_temp):
y[i] = vec[1:]
X[i]= vec[:-1]
這是構(gòu)建多對(duì)多循環(huán)網(wǎng)絡(luò)的代碼。
model = Sequential()
model.add(LSTM(128, input_shape=(NUM_INPUT_TOKENS, len(uniqueTokens)), return_sequences=True))
model.add(TimeDistributed(Dense(len(uniqueTokens))))
model.add(Activation('softmax'))
optimizer = RMSprop(lr=0.01)
model.compile(loss='categorical_crossentropy', optimizer=optimizer)
print(model.summary())
你可以打印網(wǎng)絡(luò)結(jié)構(gòu):
_________________________________________________________________ Layer (type) Output Shape Param # ================================================================= lstm_1 (LSTM) (None, 10, 128) 670208 _________________________________________________________________ time_distributed_1 (TimeDist (None, 10, 1180) 152220 _________________________________________________________________ activation_1 (Activation) (None, 10, 1180) 0 =================================================================
就像我們?yōu)槎鄬?duì)一結(jié)構(gòu)所做的那樣,我們也可以輕松地多堆疊一層 LSTM,如下所示:
model = Sequential()
model.add(LSTM(128, return_sequences=True, input_shape=(NUM_INPUT_TOKENS, len(uniqueTokens))))
model.add(LSTM(128, return_sequences=True))
model.add(TimeDistributed(Dense(len(uniqueTokens))))
model.add(Activation('softmax'))
optimizer = RMSprop(lr=0.01)
model.compile(loss='categorical_crossentropy', optimizer=optimizer)
print(model.summary())
網(wǎng)絡(luò)結(jié)構(gòu)是這樣的:
_________________________________________________________________ Layer (type) Output Shape Param # ================================================================= lstm_1 (LSTM) (None, 10, 128) 670208 _________________________________________________________________ lstm_2 (LSTM) (None, 10, 128) 131584 _________________________________________________________________ time_distributed_1 (TimeDist (None, 10, 1180) 152220 _________________________________________________________________ activation_1 (Activation) (None, 10, 1180) 0 =================================================================
經(jīng)過幾次迭代,結(jié)果看起來比之前的多對(duì)一網(wǎng)絡(luò)要好。我強(qiáng)烈建議你在運(yùn)行代碼能夠有自己的觀察并思考原因。那將是一個(gè)很好的練習(xí)。
runattributes = numberelements [ i ] . offsets [ currindex ] ;
patternentry ucompactintarray ;
import sun . util . oldstart ;
關(guān)于“Python怎么實(shí)現(xiàn)多對(duì)多網(wǎng)絡(luò)結(jié)構(gòu)”的內(nèi)容就介紹到這里了,感謝大家的閱讀。如果想了解更多行業(yè)相關(guān)的知識(shí),可以關(guān)注億速云行業(yè)資訊頻道,小編每天都會(huì)為大家更新不同的知識(shí)點(diǎn)。
免責(zé)聲明:本站發(fā)布的內(nèi)容(圖片、視頻和文字)以原創(chuàng)、轉(zhuǎn)載和分享為主,文章觀點(diǎn)不代表本網(wǎng)站立場(chǎng),如果涉及侵權(quán)請(qǐng)聯(lián)系站長(zhǎng)郵箱:is@yisu.com進(jìn)行舉報(bào),并提供相關(guān)證據(jù),一經(jīng)查實(shí),將立刻刪除涉嫌侵權(quán)內(nèi)容。