def inference():
word_2_id = get_word_to_id(config.VOCAB_PATH)
id_2_word = {k: v for (v, k) in word_2_id.items()}
initializer = tf.random_uniform_initializer(-0.05, 0.05)
model = Seq2seq(w2i_target=word_2_id, initializer=initializer)
source_batch, source_lens, target_batch, target_lens = make_batches()
print_every = 100
batches = 10400
with tf.Session() as sess:
tf.summary.FileWriter('graph', sess.graph)
saver = tf.train.Saver()
sess.run(tf.global_variables_initializer())
losses = []
total_loss = 0
try:
for batch in range(batches):
feed_dict = {
model.seq_inputs: source_batch[batch],
model.seq_inputs_length: source_lens[batch],
model.seq_targets: target_batch[batch],
model.seq_targets_length: target_lens[batch]
}
loss, _ = sess.run([model.loss, model.train_op],
feed_dict=feed_dict)
total_loss += loss
if batch % print_every == 0 and batch > 0:
print_loss = total_loss if batch == 0 else total_loss / print_every
losses.append(print_loss)
total_loss = 0
print("-----------------------------")
print("batch:", batch, "/", batches)
print(
"time:",
time.strftime('%Y-%m-%d %H:%M:%S',
time.localtime(time.time())))
print("loss:", print_loss)
print(losses)
print(saver.save(sess, "checkpoint/model.ckpt"))
except Exception as e:
print(source_batch[batch], np.shape(source_batch[batch]))
print(source_lens[batch], np.shape(source_lens[batch]))
if __name__ == "__main__":
print("(1)load data......")
docs_source, docs_target = load_data("")
w2i_source, i2w_source = make_vocab(docs_source)
w2i_target, i2w_target = make_vocab(docs_target)
print("(2) build model......")
config = Config()
config.source_vocab_size = len(w2i_source)
config.target_vocab_size = len(w2i_target)
model = Seq2seq(config=config,
w2i_target=w2i_target,
useTeacherForcing=True,
useAttention=True,
useBeamSearch=1)
print("(3) run model......")
batches = 3000
print_every = 100
with tf.Session(config=tf_config) as sess:
tf.summary.FileWriter('graph', sess.graph)
saver = tf.train.Saver()
sess.run(tf.global_variables_initializer())
losses = []
total_loss = 0
for batch in range(batches):
if __name__ == "__main__":
print("(1)load data......")
docs_source, docs_target = load_data(
'seq2seq_dataset/train_data_question_pair')
eval_doc_source, eval_doc_target = load_data(
'seq2seq_dataset/dev_data_question_pair')
print("(2) build model......")
config = Config()
config.source_vocab_size = len(tokenizer.vocab)
config.target_vocab_size = 10000 #len(tokenizer.vocab)
model = Seq2seq(config=config,
tokenizer=tokenizer,
useTeacherForcing=True,
useAttention=True,
useBeamSearch=1)
print("(3) run model......")
print_every = 200
epoches = 3
train_data_size = len(docs_target)
batch_size = 32
batches = train_data_size // batch_size
acc = 0.0
with tf.Session(config=tf_config) as sess:
tf.summary.FileWriter('graph', sess.graph)
saver = tf.train.Saver()
sess.run(tf.global_variables_initializer())
percent = 0.8 # 80% 数据用来训练
config = Config()
print("(1)load data......")
f = open(input_name, 'rb').read()
data = f.decode().split('\n')
data = data[:-1]
sequence = []
for str in data:
sequence.append(list(map(float, str.split(','))))
x_train, y_train = load_train_data(sequence, config.input_length,
config.predict_length, percent)
print("(2) build model......")
model = Seq2seq(config=config, useTeacherForcing=True, useAttention=True)
print("(3) run model......")
batches = 3000
print_every = 500
with tf.Session(config=tf_config) as sess:
tf.summary.FileWriter('graph', sess.graph)
saver = tf.train.Saver()
sess.run(tf.global_variables_initializer())
losses = []
total_loss = 0
for batch in range(batches):
source_batch, source_lens, target_batch, target_time, target_lens = get_batch(
x_train, y_train, config.batch_size)