Hfrxdjt

I tried to implement a model that takes as input sentences, which are hate_tweets and outputs exactly the same sentences. For this reason, I gave Input to the encoder and decoder exactly the same sentences. I ran the model for 10000 samples (sentences) and for 100 epochs. A small sample of the results I have, are shown below. The model does not seem to be going well. The only time that it manages to output the same sentence is when the sentence is empty or consists of only one word.

What can be blamed for this? Do you think more epochs are needed?

Ignore the predicted class, it is irrelevant to the problem.

-
Input sentence: !!! RT @mayasolovely As a woman you shouldn't complain about cleaning up your house. & as a man you should always take the trash out...
Predicted class: neutral
Decoded sentence: funnt texted that bitch and show up the careland for the face that bitch ass niggas to me in the stay and still less niggas gonna smack
Predicted class: hate
-
Input sentence: !!!!! RT @mleew17 boy dats cold...tyga dwn bad for cuffin dat hoe in the 1st place!!
Predicted class: neutral
Decoded sentence: for a bitch that bitch ass nigga all niggas and call httpt.co8ynkynFWIF #heersh #fan
Predicted class: hate
-
Input sentence: !!!!!!! RT @UrKindOfBrand Dawg!!!! RT @80sbaby4life You ever fuck a bitch and she start to cry You be confused as shit
Predicted class: hate
Decoded sentence: for a bitch when a colored man i got a mine of the lost to the whole of the way to fuck where the mestions
Predicted class: hate
-
Input sentence: !!!!!!!!! RT @C_G_Anderson @viva_based she look like a tranny
Predicted class: neutral
Decoded sentence: & These hoes ain't loyal 😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂&#12851
Predicted class: neutral
-
Input sentence: !!!!!!!!!!!!! RT @ShenikaRoberts The shit you hear about me might be true or it might be faker than the bitch who told it to ya 
Predicted class: hate
Decoded sentence: & These hoes ain't loyal 😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂&#12851
Predicted class: neutral
-
Input sentence: !!!!!!!!!!!!!!!!!!@T_Madison_x The shit just blows me..claim you so faithful and down for somebody but still fucking with hoes! 😂😂😂
Predicted class: hate
Decoded sentence: & These hoes ain't loyal 😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂&#12851
Predicted class: neutral
-
Input sentence: !!!!!!@__BrighterDays I can not just sit up and HATE on another bitch .. I got too much shit going on!
Predicted class: hate
Decoded sentence: & These hoes ain't loyal for the most his start will can't have a good and the bitches they are ho
Predicted class: hate
-
Input sentence: !!!!“@selfiequeenbri cause I'm tired of you big bitches coming for us skinny girls!!”
Predicted class: hate
Decoded sentence: for a bitch that bitch ass nigga all niggas and call httpt.co8ynkynFWIF #heersh #fanges
Predicted class: hate
-
Input sentence: & you might not get ya bitch back & thats that 
Predicted class: hate
Decoded sentence: it's trashet her and send this still like a pussy.
Predicted class: hate
-
Input sentence: @rhythmixx_ hobbies include fighting Mariam
Predicted class: neutral
Decoded sentence: ' I wonder it this bitch he still fucked.”
Predicted class: hate
-
Input sentence: 
Predicted class: neutral
Decoded sentence: 
Predicted class: neutral
-
Input sentence: bitch
Predicted class: hate
Decoded sentence: bitch
Predicted class: hate
-
Input sentence: Keeks is a bitch she curves everyone lol I walked into a conversation like this. Smh
Predicted class: hate
Decoded sentence: bitch who do you love me should vige of the words and say it out of them all u cant
Predicted class: hate

Following is the code I use :

from __future__ import print_function

import pickle

import numpy as np
from keras.layers import Input, LSTM, Dense
from keras.models import load_model

batch_size = 64 # Batch size for training.
epochs = 100 # Number of epochs to train for.
latent_dim = 256 # Latent dimensionality of the encoding space.
num_samples = 10000 # Number of samples to train on.
# Path to the data txt file on disk.
data_path = 'seq2seq_hate_tweets.txt'

# Vectorize the data.
input_texts = []
target_texts = []
input_characters = set()
target_characters = set()
with open(data_path, 'r', encoding='utf-8') as f:
 lines = f.read().split('n')
for line in lines[: min(num_samples, len(lines) - 1)]:
 input_text = line
 target_text = line
 # We use "tab" as the "start sequence" character
 # for the targets, and "n" as "end sequence" character.
 target_text = 't' + target_text + 'n'
 input_texts.append(input_text)
 target_texts.append(target_text)
 for char in input_text:
 if char not in input_characters:
 input_characters.add(char)
 for char in target_text:
 if char not in target_characters:
 target_characters.add(char)

input_characters = sorted(list(input_characters))
target_characters = sorted(list(target_characters))
num_encoder_tokens = len(input_characters)
num_decoder_tokens = len(target_characters)
max_encoder_seq_length = max([len(txt) for txt in input_texts])
max_decoder_seq_length = max([len(txt) for txt in target_texts])

input_token_index = dict(
 [(char, i) for i, char in enumerate(input_characters)])
target_token_index = dict(
 [(char, i) for i, char in enumerate(target_characters)])

encoder_input_data = np.zeros(
 (len(input_texts), max_encoder_seq_length, num_encoder_tokens),
 dtype='float32')
decoder_input_data = np.zeros(
 (len(input_texts), max_decoder_seq_length, num_decoder_tokens),
 dtype='float32')
decoder_target_data = np.zeros(
 (len(input_texts), max_decoder_seq_length, num_decoder_tokens),
 dtype='float32')

for i, (input_text, target_text) in enumerate(zip(input_texts, target_texts)):
 for t, char in enumerate(input_text):
 encoder_input_data[i, t, input_token_index[char]] = 1.
 for t, char in enumerate(target_text):
 # decoder_target_data is ahead of decoder_input_data by one timestep
 decoder_input_data[i, t, target_token_index[char]] = 1.
 if t > 0:
 # decoder_target_data will be ahead by one timestep
 # and will not include the start character.
 decoder_target_data[i, t - 1, target_token_index[char]] = 1.

# Define an input sequence and process it.
encoder_inputs = Input(shape=(None, num_encoder_tokens))
encoder = LSTM(latent_dim, return_state=True)
encoder_outputs, state_h, state_c = encoder(encoder_inputs)
# We discard `encoder_outputs` and only keep the states.
encoder_states = [state_h, state_c]

# Set up the decoder, using `encoder_states` as initial state.
decoder_inputs = Input(shape=(None, num_decoder_tokens))
# We set up our decoder to return full output sequences,
# and to return internal states as well. We don't use the
# return states in the training model, but we will use them in inference.
decoder_lstm = LSTM(latent_dim, return_sequences=True, return_state=True)
decoder_outputs, _, _ = decoder_lstm(decoder_inputs,
 initial_state=encoder_states)
decoder_dense = Dense(num_decoder_tokens, activation='softmax')
decoder_outputs = decoder_dense(decoder_outputs)

# Define the model that will turn
# `encoder_input_data` & `decoder_input_data` into `decoder_target_data`
#model = Model([encoder_inputs, decoder_inputs], decoder_outputs)

# Run training
#model.compile(optimizer='rmsprop', loss='categorical_crossentropy')
#model.fit([encoder_input_data, decoder_input_data], decoder_target_data,
# batch_size=batch_size,
# epochs=epochs,
# validation_split=0.2)

#model.save('my_model.h5') # creates a HDF5 file 'my_model.h5'
#del model # deletes the existing model

# returns a compiled model
# identical to the previous one
model = load_model('my_model.h5')

# Next: inference mode (sampling).

# Define sampling models
#encoder_model = Model(encoder_inputs, encoder_states)

decoder_state_input_h = Input(shape=(latent_dim,))
decoder_state_input_c = Input(shape=(latent_dim,))
decoder_states_inputs = [decoder_state_input_h, decoder_state_input_c]
decoder_outputs, state_h, state_c = decoder_lstm(
 decoder_inputs, initial_state=decoder_states_inputs)
decoder_states = [state_h, state_c]
decoder_outputs = decoder_dense(decoder_outputs)
#decoder_model = Model(
# [decoder_inputs] + decoder_states_inputs,
# [decoder_outputs] + decoder_states)

#encoder_model.save('my_encoder_model.h5')
#decoder_model.save('my_decoder_model.h5')
#del encoder_model
#del decoder_model

# returns a compiled model
# identical to the previous one
encoder_model = load_model('my_encoder_model.h5')
decoder_model = load_model('my_decoder_model.h5')

# Reverse-lookup token index to decode sequences back to
# something readable.
reverse_input_char_index = dict(
 (i, char) for char, i in input_token_index.items())
reverse_target_char_index = dict(
 (i, char) for char, i in target_token_index.items())


def decode_sequence(input_seq):
 # Encode the input as state vectors.
 states_value = encoder_model.predict(input_seq)

 #print("states_value is : ")
 #print(states_value)
 # Generate empty target sequence of length 1.
 target_seq = np.zeros((1, 1, num_decoder_tokens))
 # Populate the first character of target sequence with the start character.
 target_seq[0, 0, target_token_index['t']] = 1.

 # Sampling loop for a batch of sequences
 # (to simplify, here we assume a batch of size 1).
 stop_condition = False
 decoded_sentence = ''
 while not stop_condition:
 output_tokens, h, c = decoder_model.predict(
 [target_seq] + states_value)

 # Sample a token
 sampled_token_index = np.argmax(output_tokens[0, -1, :])
 sampled_char = reverse_target_char_index[sampled_token_index]
 decoded_sentence += sampled_char

 # Exit condition: either hit max length
 # or find stop character.
 if (sampled_char == 'n' or
 len(decoded_sentence) > max_decoder_seq_length):
 stop_condition = True

 # Update the target sequence (of length 1).
 target_seq = np.zeros((1, 1, num_decoder_tokens))
 target_seq[0, 0, sampled_token_index] = 1.

 # Update states
 states_value = [h, c]

 return decoded_sentence


for seq_index in range(100):
# Take one sequence (part of the training set)
# for trying out decoding.
input_seq = encoder_input_data[seq_index: seq_index + 1]
decoded_sentence = decode_sequence(input_seq)
print('-')
print('Input sentence:', input_texts[seq_index])
print('Decoded sentence:', decoded_sentence)