merge function is not working for Skipgram code

[NK10]

in the below code with keras latest version, its not working.

from keras.layers import Merge
from keras.layers.core import Dense, Reshape
from keras.layers.embeddings import Embedding
from keras.models import Sequential

# build skip-gram architecture
word_model = Sequential()
word_model.add(Embedding(vocab_size, embed_size,
                         embeddings_initializer="glorot_uniform",
                         input_length=1))
word_model.add(Reshape((embed_size, )))

context_model = Sequential()
context_model.add(Embedding(vocab_size, embed_size,
                  embeddings_initializer="glorot_uniform",
                  input_length=1))
context_model.add(Reshape((embed_size,)))

model = Sequential()
model.add(Merge([word_model, context_model], mode="dot"))
model.add(Dense(1, kernel_initializer="glorot_uniform", activation="sigmoid"))
model.compile(loss="mean_squared_error", optimizer="rmsprop")

# view model summary
print(model.summary())

# visualize model structure
from IPython.display import SVG
from keras.utils.vis_utils import model_to_dot

SVG(model_to_dot(model, show_shapes=True, show_layer_names=False, 
                 rankdir='TB').create(prog='dot', format='svg'))

I have modified the code from sequential api to functional API. please find below the code:

from keras.layers import dot
from keras.layers import Dense, Input
from keras.layers.core import Reshape
from keras.layers import Embedding
from keras.models import  Model
from keras.backend import reshape

# build skip-gram architecture
inp = Input(shape=(1,),name = "first_input")
word_model22 = Embedding(input_dim=vocab_size, output_dim=embed_size,
                         embeddings_initializer="glorot_uniform",
                         input_length=1)
emb  = word_model22(inp)
word_model22 = Reshape(target_shape= (embed_size,))(emb)

inp1 = Input(shape=(1,),name = "2nd_input")
context_model = Embedding(input_dim=vocab_size, output_dim=embed_size,
                         embeddings_initializer="glorot_uniform",
                         input_length=1)
emb1  = context_model(inp1)
context_model = Reshape(target_shape= (embed_size,))(emb1)

mo = (dot([word_model22, context_model],axes=-1))
mo = (Dense(1, kernel_initializer="glorot_uniform", activation="sigmoid"))(mo)
model = Model(inputs = (inp, inp1), outputs =mo)
model.compile(loss="mean_squared_error", optimizer="rmsprop")

# view model summary
print(model.summary())

# visualize model structure
from IPython.display import SVG
from keras.utils.vis_utils import model_to_dot

SVG(model_to_dot(model, show_shapes=True, show_layer_names=False, 
                 rankdir='TB').create(prog='dot', format='svg'))

[NK10]

sorry, didn't mention earlier, it was the merge part that is not working.

[dipanjanS]

Thanks, yes we are aware of this happening due to the deprecation of the Merge API, I will be pushing this change soon in my 2nd edition of the text analytics book and once that is out, will be replacing the code here also!

[amraneabdeslam]

Hi, you can replace Merge by Concatenate
model.add(Merge([word_model, context_model], mode="dot"))
-->
model.add(Concatenate([word_model, context_model]))