def predict(args):
with open(os.path.join(args.save_dir, 'config.pkl')) as f:
saved_args = cPickle.load(f)
with open(os.path.join(args.save_dir, 'combined_vocab.pkl')) as f:
_, vocab = cPickle.load(f)
model = Discriminator(saved_args, is_training=False)
with tf.Session() as sess:
tf.initialize_all_variables().run()
saver = tf.train.Saver(tf.all_variables())
ckpt = tf.train.get_checkpoint_state(args.save_dir)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
return model.predict(sess, args.text, vocab)
def predict(args):
with open(os.path.join(args.save_dir, 'config.pkl')) as f:
saved_args = cPickle.load(f)
with open(os.path.join(args.save_dir, 'combined_vocab.pkl')) as f:
_, vocab = cPickle.load(f)
model = Discriminator(saved_args, is_training = False)
with tf.Session() as sess:
tf.initialize_all_variables().run()
saver = tf.train.Saver(tf.all_variables())
ckpt = tf.train.get_checkpoint_state(args.save_dir)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
return model.predict(sess, args.text, vocab)
def main():
# f = open("model/comp.graphics_uni.pkl")
# Generating a unigram sentence
# with open("models/comp.graphics_uni.pkl", "rb") as f:
# model = pickle.load(f)
# print(len(model.model.keys()))
graph_uni_gen = Generator(model_path="models/comp.graphics_uni.pkl")
bikes_uni_gen = Generator(model_path="models/rec.motorcycles_uni.pkl")
g_unigram_sentence = graph_uni_gen.generate_sentence(threshold=50)
b_unigram_sentence = bikes_uni_gen.generate_sentence(threshold=50)
# Generating a bigram sentence
graph_bi_gen = Generator(model_path="models/comp.graphics_bi.pkl")
bikes_bi_gen = Generator(model_path="models/rec.motorcycles_bi.pkl")
g_bigram_sentence_1 = graph_bi_gen.generate_sentence(threshold=50)
b_bigram_sentence_1 = bikes_bi_gen.generate_sentence(threshold=50)
g_bigram_sentence_2 = graph_bi_gen.generate_sentence(threshold=50)
b_bigram_sentence_2 = bikes_bi_gen.generate_sentence(threshold=50)
graph_tri_gen = Generator(model_path="models/comp.graphics_tri.pkl")
bikes_tri_gen = Generator(model_path="models/rec.motorcycles_tri.pkl")
g_trigram_sentence_1 = graph_tri_gen.generate_sentence(threshold=50)
b_trigram_sentence_1 = bikes_tri_gen.generate_sentence(threshold=50)
g_trigram_sentence_2 = graph_tri_gen.generate_sentence(threshold=50)
b_trigram_sentence_2 = bikes_tri_gen.generate_sentence(threshold=50)
print("Corpus\t:\t" + graph_uni_gen.model.corpus)
print("\nGenerated sentences\n")
print("Unigram -- \n")
print(g_unigram_sentence + "\n")
print("Bigram -- \n")
print("1. " + g_bigram_sentence_1 + "\n")
print("2. " + g_bigram_sentence_2 + "\n")
print("Trigram -- \n")
print("1. " + g_trigram_sentence_1 + "\n")
print("2. " + g_trigram_sentence_2 + "\n")
print(
"==============================================================================="
)
print(
"===============================================================================\n"
)
print("Corpus\t:\t" + bikes_uni_gen.model.corpus)
print("\nGenerated sentences\n")
print("Unigram -- \n")
print(b_unigram_sentence + "\n")
print("Bigram -- \n")
print("1. " + b_bigram_sentence_1 + "\n")
print("2. " + b_bigram_sentence_2 + "\n")
print("Trigram -- \n")
print("1. " + b_trigram_sentence_1 + "\n")
print("2. " + b_trigram_sentence_2 + "\n")
print(
"==============================================================================="
)
print(
"===============================================================================\n"
)
print("\n\n")
print("Without UNK replacement")
d1 = Discriminator(model_paths=["models/comp.graphics_bi.pkl", "models/rec.motorcycles_bi.pkl"]\
, smoothing=True, smoothing_k=1, unk_tackle=False)
print("Predicting sentence:\n", g_bigram_sentence_1, "\n")
print("Results\n")
print(d1.predict(" ".join(g_bigram_sentence_1.split()[1:-1])))
print("\n\n")
print("Predicting sentence:\n", b_bigram_sentence_2, "\n")
print("Results\n")
print(d1.predict(" ".join(b_bigram_sentence_2.split()[1:-1])))
print("\n\n")
sent = input("Enter a sentence:\n").strip()
print("Predicting sentence:\n", sent, "\n")
print("Results\n")
print(d1.predict(sent))
print("\n\n")
print("With UNK Replacement")
print("\n\n")
d2 = Discriminator(model_paths=["models/comp.graphics_bi.pkl", "models/rec.motorcycles_bi.pkl"]\
, smoothing=True, smoothing_k=1, unk_tackle=True, unk_threshold=10)
print("Predicting sentence:\n", g_bigram_sentence_1, "\n")
print("Results\n")
print(d2.predict(" ".join(g_bigram_sentence_1.split()[1:-1])))
print("\n\n")
print("Predicting sentence:\n", b_bigram_sentence_2, "\n")
print("Results\n")
print(d2.predict(" ".join(b_bigram_sentence_2.split()[1:-1])))
print("\n\n")
sent = input("Enter a sentence:\n").strip()
print("Predicting sentence:\n", sent, "\n")
print("Results\n")
print(d2.predict(sent))
print("\n\n")
