Im trying to convert a sequence of length N to a sequence of around length N^2 using a pseudo seq2seq type model, but Im not sure how to implement the variable input length in my keras model
def LSTMModel():
input = Input(shape = (None,num_channels))
lstm_one = LSTM(75, return_sequences = True)
lstm_one_output = lstm_one(input)
BiLSTM = Bidirectional(LSTM(units = 100, return_sequences=True, recurrent_dropout = 0.1))
LSTM_outputs = BiLSTM(lstm_one_output)
output = LSTM(2, return_sequences = False)(LSTM_outputs)
return Model(input, output)
This code would produce a (None, 2) output, but I really want it to be a (None, None^2) output. Is there any way to somehow store the shape within the model and do some operations with it with keras layers, perhaps with a lambda function?
Related
I have a seq to seq model trained of some clever bot data:
justphrases_X is a list of sentences and justphrases_Y is a list of responses to those sentences.
maxlen = 62
#low is a list of all the unique words.
def Convert_To_Encoding(just_phrases):
encodings = []
for sentence in just_phrases:
onehotencoded = one_hot(sentence, len(low))
encodings.append(np.array(onehotencoded))
encodings_padded = pad_sequences(encodings, maxlen=maxlen, padding='post', value = 0.0)
return encodings_padded
encodings_X_padded = Convert_To_Encoding(just_phrases_X)
encodings_y_padded = Convert_To_Encoding(just_phrases_y)
model = Sequential()
embedding_layer = Embedding(len(low), output_dim=8, input_length=maxlen)
model.add(embedding_layer)
model.add(GRU(128)) # input_shape=(None, 496)
model.add(RepeatVector(numberofwordsoutput)) #number of characters?
model.add(GRU(128, return_sequences = True))
model.add(Flatten())
model.add(Dense(62, activation = 'softmax'))
model.compile(loss = 'categorical_crossentropy', optimizer= 'adam', metrics=['accuracy'])
model.summary()
model.fit(encodings_X_padded, encodings_y_padded, batch_size = 1, epochs=1) #, validation_data = (testX, testy)
model.save("cleverbottheseq-uel.h5")
When I use this model for prediction, the output will be between 0 and 1 because of my use of softmax. However as I have around 3000 unique words, each with a separate integer assigned to it, how do I essentially repeat what the model did during training and convert the output back to an integer which has a word assigned to it?
I dont think it is possible to create seq2seq with Sequential API. Try to create encoder and decoder separately with Functional API. You need two inputs - first for encoder, second - for decoder.
I try to create image embeddings for the purpose of deep ranking using a triplet loss function. The idea is that we can take a pretrained CNN (e.g. resnet50 or vgg16), remove the FC layers and add an L2 normalization function to retrieve unit vectors which can then be compared via a distance metric (e.g. cosine similarity). As far as I understand the predicted vectors that come out of a pretrained CNN are not optimal, but are a good start. By adding the triplet loss function we can re-train the network to keep similar pictures 'close' to each other and different pictures 'far' apart in the feature space. Inspired by this notebook , I tried to setup the following code, but I get an error ValueError: The name "conv1_pad" is used 3 times in the model. All layer names should be unique..
# Anchor, Positive and Negative are numpy arrays of size (200, 256, 256, 3), same for the test images
pic_size=256
def shared_dnn(inp):
base_model = ResNet50(weights='imagenet', include_top=False, input_shape=(3, pic_size, pic_size),
input_tensor=inp)
x = base_model.output
x = Flatten()(x)
x = Lambda(lambda x: K.l2_normalize(x,axis=1))(x)
for layer in base_model.layers[15:]:
layer.trainable = False
return x
anchor_input = Input((3, pic_size,pic_size ), name='anchor_input')
positive_input = Input((3, pic_size,pic_size ), name='positive_input')
negative_input = Input((3, pic_size,pic_size ), name='negative_input')
encoded_anchor = shared_dnn(anchor_input)
encoded_positive = shared_dnn(positive_input)
encoded_negative = shared_dnn(negative_input)
merged_vector = concatenate([encoded_anchor, encoded_positive, encoded_negative], axis=-1, name='merged_layer')
model = Model(inputs=[anchor_input,positive_input, negative_input], outputs=merged_vector)
#ValueError: The name "conv1_pad" is used 3 times in the model. All layer names should be unique.
model.compile(loss=triplet_loss, optimizer=adam_optim)
model.fit([Anchor,Positive,Negative],
y=Y_dummy,
validation_data=([Anchor_test,Positive_test,Negative_test],Y_dummy2), batch_size=512, epochs=500)
I am new to keras and I am not quite sure how to solve this. The author in the link above creates his own CNN from scratch, but I would like to build it upon resnet (or vgg16). How can I configure ResNet50 to use a triplet loss function (in the link above you find also the source code for the triplet loss function).
In your ResNet50 definition, you've written
base_model = ResNet50(weights='imagenet', include_top=False, input_shape=(3, pic_size, pic_size), input_tensor=inp)
Remove the input_tensor argument. Change input_shape=inp.
If you're using TF backend as you mentioned the input should be (256, 256, 3), then your input should be (pic_size, pic_size, 3).
def shared_dnn(inp):
base_model = ResNet50(weights='imagenet', include_top=False, input_shape=inp)
x = base_model.output
x = Flatten()(x)
x = Lambda(lambda x: K.l2_normalize(x,axis=1))(x)
for layer in base_model.layers[15:]:
layer.trainable = False
return x
img_shape=(256, 256, 3)
anchor_input = Input(img_shape, name='anchor_input')
positive_input = Input(img_shape, name='positive_input')
negative_input = Input(img_shape, name='negative_input')
encoded_anchor = shared_dnn(anchor_input)
encoded_positive = shared_dnn(positive_input)
encoded_negative = shared_dnn(negative_input)
merged_vector = concatenate([encoded_anchor, encoded_positive, encoded_negative], axis=-1, name='merged_layer')
model = Model(inputs=[anchor_input,positive_input, negative_input], outputs=merged_vector)
model.compile(loss=triplet_loss, optimizer=adam_optim)
model.fit([Anchor,Positive,Negative],
y=Y_dummy,
validation_data=([Anchor_test,Positive_test,Negative_test],Y_dummy2), batch_size=512, epochs=500)
The model plot is as follows:
model_plot
Normally, if you train with keras, model.fit expects the train data to have a shape of (samples, timesteps, input) and a label of (samples, outputs). Is there a way to change the matching label to (samples*timesteps, output) or (samples, timesteps, input). So one sample matches len(sample)*label and not only one label?
Yes. You can have whatever shape you want as the output layer. For instance auto-encoders will have the same output shape as input shape.
A toy example:
sequence_length = 20
n_features = 4
def make_model():
inp = Input(shape=(sequence_length, n_features,))
encoder = LSTM(16, return_sequences=True)(inp)
vector = LSTM(32)(encoder)
decoder_in = RepeatVector(sequence_length)(vector)
decoder = LSTM(16, return_sequences=True)(decoder_in)
out = Dense(4)(decoder)
model = Model(inp, out)
model.compile('adam', 'mse')
return model
model = make_model()
model.summary()
In this case the vector layer has shape (32,) (i.e. there is a dimensionality reduction compared to the input) and the output layer has the same dimensions as the input.
I’m trying to implement a Visual Storytelling model using Keras with a hierarchical RNN model, basically Neural Image Captioner style but over a sequence of photos with a bidirectional RNN on top of the decoder RNNs.
I implemented and tested the three parts of this model, CNN, BRNN and decoder RNN separately but got this error when trying to connect them:
ValueError: An operation has None for gradient. Please make sure that all of your ops have a gradient defined (i.e. are differentiable). Common ops without gradient: K.argmax, K.round, K.eval.
My code are as follows:
#vgg16 model with the fc2 layer as output
cnn_base_model = self.cnn_model.base_model
brnn_model = self.brnn_model.model
rnn_model = self.rnn_model.model
cnn_part = TimeDistributed(cnn_base_model)
img_input = Input((self.story_length,) + self.cnn_model.input_shape, name='brnn_img_input')
extracted_feature = cnn_part(img_input)
#[None, 5, 512], a 512 length vector for each picture in the story
brnn_feature = brnn_model(extracted_feature)
#[None, 5, 25], input groundtruth word indices fed as input when training
decoder_input = Input((self.story_length, self.max_length), name='brnn_decoder_input')
decoder_outputs = []
for i in range(self.story_length):
#separate timesteps for decoding
decoder_input_i = Lambda(lambda x: x[:, i, :])(decoder_input)
brnn_feature_i = Lambda(lambda x: x[:, i, :])(brnn_feature)
#the problem persists when using Dense instead of the Lambda layers above
#decoder_input_i = Dense(25)(Reshape((125,))(decoder_input))
#brnn_feature_i = Dense(512)(Reshape((5 * 512,))(brnn_feature))
decoder_output_i = rnn_model([decoder_input_i, brnn_feature_i])
decoder_outputs.append(decoder_output_i)
decoder_output = Concatenate(axis=-2, name='brnn_decoder_output')(decoder_outputs)
self.model = Model([img_input, decoder_input], decoder_output)
And codes for the BRNN:
image_feature = Input(shape=(self.story_length, self.img_feature_dim,))
image_emb = TimeDistributed(Dense(self.lstm_size))(image_feature)
brnn = Bidirectional(LSTM(self.lstm_size, return_sequences=True), merge_mode='concat')(image_emb)
brnn_emb = TimeDistributed(Dense(self.lstm_size))(brnn)
self.model = Model(inputs=image_feature, outputs=brnn_emb)
And RNN:
#[None, 512], the vector to be decoded
initial_input = Input(shape=(self.input_dim,), name='rnn_initial_input')
#[None, 25], the groundtruth word indices fed as input when training
decoder_inputs = Input(shape=(None,), name='rnn_decoder_inputs')
decoder_input_masking = Masking(mask_value=0.0)(decoder_inputs)
decoder_input_embeddings = Embedding(self.vocabulary_size, self.emb_size,
embeddings_regularizer=l2(regularizer))(decoder_input_masking)
decoder_input_dropout = Dropout(.5)(decoder_input_embeddings)
initial_emb = Dense(self.emb_size,
kernel_regularizer=l2(regularizer))(initial_input)
initial_reshape = Reshape((1, self.emb_size))(initial_emb)
initial_masking = Masking(mask_value=0.0)(initial_reshape)
initial_dropout = Dropout(.5)(initial_masking)
decoder_lstm = LSTM(self.hidden_dim, return_sequences=True, return_state=True,
recurrent_regularizer=l2(regularizer),
kernel_regularizer=l2(regularizer),
bias_regularizer=l2(regularizer))
_, initial_hidden_h, initial_hidden_c = decoder_lstm(initial_dropout)
decoder_outputs, decoder_state_h, decoder_state_c = decoder_lstm(decoder_input_dropout,
initial_state=[initial_hidden_h, initial_hidden_c])
decoder_output_dense_layer = TimeDistributed(Dense(self.vocabulary_size, activation='softmax',
kernel_regularizer=l2(regularizer)))
decoder_output_dense = decoder_output_dense_layer(decoder_outputs)
self.model = Model([decoder_inputs, initial_input], decoder_output_dense)
I’m using adam as optimizer and sparse_categorical_crossentropy as loss.
At first I thought the problem is with the Lambda layers used for splitting the timesteps but the problem persists when I replaced them with Dense layers (which are guarantee
I had a similar error and it turned out I was suppose to build the layers (in my custom layer or model) in the init() like so:
self.lstm_custom_1 = keras.layers.LSTM(128,batch_input_shape=batch_input_shape, return_sequences=False,stateful=True)
self.lstm_custom_1.build(batch_input_shape)
self.dense_custom_1 = keras.layers.Dense(32, activation = 'relu')
self.dense_custom_1.build(input_shape=(batch_size, 128))```
The issue is actually with the Embedding layer, I think. Gradients can't pass through an Embedding layer, so unless it's the first layer in the model it won't work.
I am working on a binary classification problem, where the network takes two inputs and output the label of this input pair.
Basically, I use an encoder layer to do embedding first and concatenate the embedding results. Next, I am going to use RNN structure to classify the concatenated result. But I can't figure out a proper way to write the code. I attach my code below.
input_size = n_feature # the number of features
encoder_size = 2000 # output dim for each encoder
dropout_rate = 0.5
X1 = Input(shape=(input_size, ), name='input_1')
X2 = Input(shape=(input_size, ), name='input_2')
encoder = Sequential()
encoder.add(Dropout(dropout_rate, input_shape=(input_size, )))
encoder.add(Dense(encoder_size, activation='relu'))
encoded_1 = encoder(X1)
encoded_2 = encoder(X2)
merged = concatenate([encoded_1, encoded_2])
#----------Need Help---------------#
comparer = Sequential()
comparer.add(LSTM(512, input_shape=(encoder_size*2, ), return_sequences=True))
comparer.add(Dropout(dropout_rate))
comparer.add(TimeDistributed(Dense(1)))
comparer.add(Activation('sigmoid'))
#----------Need Help---------------#
Y = comparer(merged)
model = Model(inputs=[X1, X2], outputs=Y)
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
It seems for the LSTM layer, the input should be (None, encoder_size*2). I tried to use Y = comparer(K.transpose(merged)) to reshape the input for the LSTM layer but I failed. BTW, for this network, the input shape is (input_size,) and output shape is (1,).
If the idea is to transform the input vector in a time series, you can simply reshape it:
comparer = Sequential()
#reshape the vector into a time series form: (None, timeSteps, features)
comparer.add(Reshape((2 * encoder_size,1), input_shape=(2*encoder_size,))
#don't return sequences, you don't want a sequence as result:
comparer.add(LSTM(512, return_sequences=False))
comparer.add(Dropout(dropout_rate))
#Don't use a TimeDistributed, you're not dealing with a series anymore
comparer.add(Dense(1))
comparer.add(Activation('sigmoid'))