def _text2ids(text, max_words):
word_ids = text2ids.text2ids(text,
seg_method=FLAGS.seg_method_,
feed_single=FLAGS.feed_single_,
allow_all_zero=True,
pad=False)
word_ids = gezi.pad(word_ids, max_words, 0)
return word_ids
def _text2ids(text, max_words):
word_ids = text2ids.text2ids(text,
seg_method=FLAGS.seg_method,
feed_single=FLAGS.feed_single,
allow_all_zero=True,
pad=False)
word_ids_length = len(word_ids)
word_ids = word_ids[:max_words]
word_ids = gezi.pad(word_ids, max_words, 0)
return word_ids
def _text2ids(text, max_words):
word_ids = text2ids.text2ids(text,
seg_method=FLAGS.seg_method_,
feed_single=FLAGS.feed_single_,
append_start=False,
append_end=False,
allow_all_zero=True,
pad=True,
max_words=max_words)
return word_ids
def _text2ids(text, max_words):
word_ids = text2ids.text2ids(text,
seg_method='basic',
feed_single=True,
allow_all_zero=True,
pad=False)
word_ids = word_ids[:max_words]
word_ids = gezi.pad(word_ids, max_words, 0)
return word_ids
def predict(text):
timer = gezi.Timer()
text_ids = text2ids.text2ids(text, FLAGS.seg_method_, feed_single=True)
print('text_ids', text_ids)
#seq_len = 50
#print('words', words)
argmax_encode = predictor.inference(
['text_importance'], feed_dict={'rnn/main/text:0': [text_ids]})
print('argmax_encode', argmax_encode[0])
argmax_encode = argmax_encode[0][0]
text_ids = text2ids.text2ids(text,
FLAGS.seg_method_,
feed_single=True,
append_start=True,
append_end=True)
words = text2ids.ids2words(text_ids)
seq_len = 0
for x in words:
if x != 0:
seq_len += 1
else:
break
print(text_ids)
# visualize model
import matplotlib.pyplot as plt
argmaxs = [np.sum((argmax_encode == k)) for k in range(seq_len)]
print('argmaxs', argmaxs, np.sum(argmaxs), seq_len)
x = range(len(argmax_encode))
y = [100.0 * n / np.sum(argmaxs) for n in argmaxs]
#print(words, y)
print(text)
for word, score in zip(words, y):
print(word, score)
def predict(self,
inputs,
seg_method='basic',
feed_single=True,
max_words=None):
if not isinstance(inputs, (list, tuple, np.ndarray)):
inputs = [inputs]
if isinstance(inputs[0][0], str):
word_ids = [
text2ids.text2ids(input,
seg_method=seg_method,
feed_single=feed_single,
max_words=max_words) for input in inputs
]
else:
word_ids = inputs
return self.predictor.predict(word_ids), word_ids
END_WORD = '</S>'
NUM_WORD = '<NUM>'
print('seg_method:', FLAGS.seg_method_, file=sys.stderr)
num = 0
for line in sys.stdin:
if num % 10000 == 0:
print(num, file=sys.stderr)
l = line.rstrip().split('\t')
texts = l[1].split('\x01')
for text in texts:
ids = text2ids.text2ids(text,
seg_method=FLAGS.seg_method_,
feed_single=FLAGS.feed_single_,
allow_all_zero=True,
pad=False,
append_start=True,
append_end=True,
to_lower=True,
norm_digit=True)
if num % 10000 == 0:
print(ids, file=sys.stderr)
print(text2ids.ids2text(ids), file=sys.stderr)
ids = map(str, ids)
if ids:
print('\t'.join(ids))
num += 1
dir = '/home/gezi/new/temp/makeup/title2name/tfrecord/seq-basic/'
text2ids.init(os.path.join(dir, 'vocab.txt'))
vocab = text2ids.vocab
embsim = melt.EmbeddingSim(os.path.join(dir, 'word2vec'), name='w_in')
corpus_pattern = os.path.join('/home/gezi/data/product/makeup/tb/title2name/valid/*')
max_words = 50
#itexts = ['ÑÅÊ«À¼÷ìË®Èó˪', 'ÑÅÊ«À¼÷ìС×ØÆ¿', 'ÑÅÊ«À¼÷ìºìʯÁñ', 'æÃÃÀ¿óÎïȪ²¹Ë®¾«»ª', 'Adidas°¢µÏ´ï˹ÄÐÊ¿ÏãË®ÄÐÊ¿¹ÅÁúµÏãË® ±ùµãÄÐÏã100ml¡¾¾©¶«³¬ÊС¿']
itexts = ['ÑÅÊ«À¼÷ìanrÐÞ»¤¼¡Í¸¾«»ªÂ¶']
left_ids = [text2ids.text2ids(x, seg_method='basic', feed_single=True, max_words=max_words) for x in itexts]
lids_ = tf.placeholder(dtype=tf.int32, shape=[None, max_words])
rids_ = tf.placeholder(dtype=tf.int32, shape=[None, max_words])
nids_ = embsim.top_sim(lids_, rids_)
sess = embsim._sess
corpus_text = []
for file in glob.glob(corpus_pattern):
corpus_text += open(file).readlines()
corpus_text = [x.strip() for x in corpus_text]
r_text = [x.split('\t')[1] for x in corpus_text]
r_text = list(set(r_text))
right_ids = [text2ids.text2ids(x, seg_method='basic', feed_single=True, max_words=max_words) for x in r_text]
image_file = raw_input('image_file like 6275b5349168ac3fab6a493c509301d023cf39d3.jpg:')
if image_file == 'q':
break
image_path = os.path.join(image_dir, image_file)
print('image_path:', image_path)
if not os.path.exists(image_path):
print('image path not find!')
continue
try:
hits = img2text[image_file]
texts = [text_strs[hit] for hit in hits]
for text in texts:
word_ids = text2ids.text2ids(text)
seg_text = text2ids.ids2text(word_ids, print_end=False)
print('label:', text, seg_text)
words_importance = sim_predictor.words_importance([word_ids])
words_importance = words_importance[0]
print('word importance:')
for i in range(len(word_ids)):
if word_ids[i] == 0:
break
print(vocab.key(int(word_ids[i])), words_importance[i], end='|')
print()
except Exception:
print(traceback.format_exc(), file=sys.stderr)
pass
image = melt.read_image(image_path)
from libword_counter import Vocabulary
from deepiu.util import text2ids
dir = '/home/gezi/new/temp/makeup/title2name/tfrecord/seq-basic/'
text2ids.init(os.path.join(dir, 'vocab.txt'))
vocab = text2ids.vocab
embsim = melt.EmbeddingSim(os.path.join(dir, 'word2vec'), name='w_in')
corpus_file = os.path.join('/home/gezi/data/product/makeup/tb/title2name/valid/name.filtered.rand.valid.txt_0')
max_words = 50
itext = 'ÑÅÊ«À¼÷ìË®Èó˪'
left_ids = text2ids.text2ids(itext, seg_method='basic', feed_single=True, max_words=max_words)
corpus_text = open(corpus_file).readlines()
corpus_text = [x.split()[0] for x in corpus_text]
right_ids = [text2ids.text2ids(x, seg_method='basic', feed_single=True, max_words=max_words) for x in corpus_text[:1000]]
print(right_ids)
lids_ = tf.placeholder(dtype=tf.int32, shape=[None, max_words])
rids_ = tf.placeholder(dtype=tf.int32, shape=[None, max_words])
nids_ = embsim.top_sim(lids_, rids_)
sess = embsim._sess
vocab = text2ids.vocab
embsim = melt.EmbeddingSim(os.path.join(dir, 'word2vec'), name='w_in')
corpus_pattern = os.path.join(
'/home/gezi/data/product/makeup/tb/title2name/valid/*')
max_words = 50
itexts = [
'ÑÅÊ«À¼÷ìË®Èó˪', 'ÑÅÊ«À¼÷ìС×ØÆ¿', 'ÑÅÊ«À¼÷ìºìʯÁñ', 'æÃÃÀ¿óÎïȪ²¹Ë®¾«»ª',
'Adidas°¢µÏ´ï˹ÄÐÊ¿ÏãË®ÄÐÊ¿¹ÅÁúµÏãË® ±ùµãÄÐÏã100ml¡¾¾©¶«³¬ÊС¿'
]
left_ids = [
text2ids.text2ids(x,
seg_method='basic',
feed_single=True,
max_words=max_words) for x in itexts
]
lids_ = tf.placeholder(dtype=tf.int32, shape=[None, max_words])
rids_ = tf.placeholder(dtype=tf.int32, shape=[None, max_words])
nids_ = embsim.top_sim(lids_, rids_)
sess = embsim._sess
corpus_text = []
for file in glob.glob(corpus_pattern):
corpus_text += open(file).readlines()
corpus_text = [x.strip() for x in corpus_text]
r_text = [x.split('\t')[1] for x in corpus_text]
r_text = list(set(r_text))
def deal_file(file, thread_index):
out_file = '{}/{}_{}'.format(
FLAGS.output_directory, FLAGS.name,
thread_index) if FLAGS.threads > 1 else '{}/{}'.format(
FLAGS.output_directory, FLAGS.name)
print('out_file:', out_file)
with melt.tfrecords.Writer(out_file) as writer:
num = 0
for line in open(file):
if num % 1000 == 0:
print(num)
l = line.rstrip().split('\t')
img = l[0]
img_end = IMAGE_FEATURE_LEN + 1
img_feature = [float(x) for x in l[1:img_end]]
texts = [x.split('\x01')[0] for x in l[img_end:]]
for text in texts:
if text.strip() == '':
continue
#@TODO from text -> ids should move out so online code can share it for evaluation or use for feed dict
#words = segmentor.Segment(text, FLAGS.seg_method)
#word_ids = [vocabulary.id(word) for word in words if vocabulary.has(word) or ENCODE_UNK]
word_ids = text2ids.text2ids(text,
seg_method=FLAGS.seg_method,
feed_single=FLAGS.feed_single,
allow_all_zero=True,
pad=False)
word_ids_length = len(word_ids)
if num % 1000 == 0:
print(text,
word_ids,
text2ids.ids2text(word_ids),
file=sys.stderr)
if len(word_ids) == 0:
continue
word_ids = word_ids[:TEXT_MAX_WORDS]
if FLAGS.pad:
word_ids = gezi.pad(word_ids, TEXT_MAX_WORDS, 0)
if FLAGS.np_save:
gtexts[thread_index].append(word_ids)
gtext_strs[thread_index].append(text)
#add pos info? weght info? or @TODO add click num info
example = tf.train.Example(features=tf.train.Features(
feature={
'image_name': melt.bytes_feature(img),
'image_feature': melt.float_feature(img_feature),
'text': melt.int_feature(word_ids),
'text_str': melt.bytes_feature(text),
}))
writer.write(example)
global counter, max_num_words, sum_words
with counter.get_lock():
counter.value += 1
if word_ids_length > max_num_words.value:
with max_num_words.get_lock():
max_num_words.value = word_ids_length
with sum_words.get_lock():
sum_words.value += word_ids_length
num += 1
texts_dict[thread_index] = gtexts[thread_index]
text_strs_dict[thread_index] = gtext_strs[thread_index]
