def __init__(self):
self.w2v_wr = data_helpers.w2v_wrapper(FLAGS.w2v_file) # 加载词向量
self.init_model()
self.refuse_classification_map = {
0: '可回收垃圾',
1: '有害垃圾',
2: '湿垃圾',
3: '干垃圾'
}
FLAGS.num_epochs)
def dev_test():
batches_dev = data_helpers.batch_iter(list(zip(x_dev, y_dev)),
FLAGS.batch_size, 1)
for batch_dev in batches_dev:
x_batch_dev, y_batch_dev = zip(*batch_dev)
dev_step(x_batch_dev,
y_batch_dev,
writer=dev_summary_writer)
# Training loop. For each batch...
for batch in batches:
x_batch, y_batch = zip(*batch)
train_step(x_batch, y_batch)
current_step = tf.train.global_step(sess, global_step)
# Training loop. For each batch...
if current_step % FLAGS.evaluate_every == 0:
print("\nEvaluation:")
dev_test()
if current_step % FLAGS.checkpoint_every == 0:
path = saver.save(sess,
checkpoint_prefix,
global_step=current_step)
print("Saved model checkpoint to {}\n".format(path))
if __name__ == "__main__":
w2v_wr = data_helpers.w2v_wrapper(FLAGS.w2v_file)
train(w2v_wr.model)
def dev_test():
# batches_dev = data_helpers.batch_iter(list(zip(x_dev, y_dev,entity_dev)), FLAGS.batch_size, 1)
# for batch_dev in batches_dev:
# x_batch_dev, y_batch_dev ,entity_batch_dev= zip(*batch_dev)
accuracy = dev_step(x_dev, y_dev, writer=dev_summary_writer)
return accuracy
# Training loop. For each batch...
for batch in batches:
x_batch, y_batch = zip(*batch)
train_step(x_batch, y_batch)
current_step = tf.train.global_step(sess, global_step)
# Training loop. For each batch...
if current_step % FLAGS.evaluate_every == 0:
print("\nEvaluation:")
accuracy = dev_test()
if accuracy > best_accurcy:
best_accurcy = accuracy
print("bset_accuracy {}\n".format(best_accurcy))
path = saver.save(sess, checkpoint_prefix, global_step=current_step)
path_pass = str(path)
path_pass = path_pass.split('\\')
print(path_pass)
print("Saved model checkpoint to {}\n".format(path))
return x_dev,y_dev,path_pass
if __name__ == "__main__":
w2v_wr = data_helpers.w2v_wrapper(FLAGS.pre_emb_file)
x_dev,y_dev,path = train(w2v_wr.model)
eval(x_dev,y_dev,path)
