# Zipping back the files
cmd.append("gzip " + big_gw_data)
cmd.append("gzip " + gw_data)
for c in cmd:
print c
os.system(c)
# break
print
print "----------------"
print "Printing the image files"
# game_img_dict = {}
img_dict = {}
for (n, s) in options:
game_set = OracleDataset.load(args.data_dir, s)
for game in game_set.games:
img_dict[str(game.image.id) + ".jpg"] = 1
f = open(path + s + "_img_index.txt", "w")
for key in img_dict:
print >> f, key
f.close()
# game_img_dict[s] = img_dict
# print s
# print img_dict
print
print img_dict
# path of the data
raw_dir = path + "img/raw"
# Load image
image_builder, crop_builder = None, None
use_resnet = False
if config['inputs'].get('image', False):
logger.info('Loading images..')
image_builder = get_img_builder(config['model']['image'], args.img_dir)
use_resnet = image_builder.is_raw_image()
if config['inputs'].get('crop', False):
logger.info('Loading crops..')
crop_builder = get_img_builder(config['model']['crop'], args.crop_dir, is_crop=True)
use_resnet = crop_builder.is_raw_image()
# Load data
logger.info('Loading data..')
trainset = OracleDataset.load(args.data_dir, "train", image_builder, crop_builder)
validset = OracleDataset.load(args.data_dir, "valid", image_builder, crop_builder)
testset = OracleDataset.load(args.data_dir, "test", image_builder, crop_builder)
# Load dictionary
logger.info('Loading dictionary..')
tokenizer = GWTokenizer(os.path.join(args.data_dir, args.dict_file))
# Build Network
logger.info('Building network..')
network = OracleNetwork(config, num_words=tokenizer.no_words)
# Build Optimizer
logger.info('Building optimizer..')
optimizer, outputs = create_optimizer(network, config, finetune=finetune)
'<padding>': 0,
'<start>': 1,
'<stop>': 2,
'<stop_dialogue>': 3,
'<unk>': 4,
'<yes>': 5,
'<no>': 6,
'<n/a>': 7,
}
word2occ = collections.defaultdict(int)
tknzr = TweetTokenizer(preserve_case=False)
print("Processing train dataset...")
trainset = OracleDataset.load(args.data_dir, "train")
for game in trainset.games:
question = game.questions[0]
tokens = tknzr.tokenize(question)
for tok in tokens:
word2occ[tok] += 1
print("filter words...")
for word, occ in word2occ.items():
if occ >= args.min_occ and word.count('.') <= 1:
word2i[word] = len(word2i)
print("Number of words (occ >= 1): {}".format(len(word2occ)))
print("Number of words (occ >= {}): {}".format(args.min_occ, len(word2i)))
dict_path = os.path.join(args.data_dir, 'dict.json')
# CPU/GPU option
cpu_pool = Pool(args.no_thread, maxtasksperchild=1000)
gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=args.gpu_ratio)
with tf.Session(config=tf.ConfigProto(gpu_options=gpu_options, allow_soft_placement=True)) as sess:
saver = tf.train.Saver()
saver.restore(sess, args.ckpt)
features = dict()
for one_set in args.set_type:
print("Load dataset -> set: {}".format(one_set))
dataset = OracleDataset.load(args.data_dir, one_set, image_loader=image_loader, crop_loader=crop_loader)
batchifier = OracleBatchifier(tokenizer=None, sources=[source])
iterator = Iterator(dataset,
batch_size=args.batch_size,
pool=cpu_pool,
batchifier=batchifier)
for batch in tqdm(iterator):
feat = sess.run(end_points[feature_name], feed_dict={images: numpy.array(batch[source])})
for f, game in zip(feat, batch["raw"]):
f = f.squeeze()
if args.mode == "crop":
id = game.object_id
else:
id = game.picture.id
args = parser.parse_args()
all_question = []
word2occ = collections.defaultdict(int)
tknzr = TweetTokenizer(preserve_case=False)
# model = FastText()
# model.load_model("data/Embedding/wiki-simple.vec")
# print(model.nearest_neighbors('teacher'))
# exit()
print("Processing train/valid dataset...")
trainset = OracleDataset.load("data", "train")
validset = OracleDataset.load("data", "valid")
test = []
print("array append ...")
for game in trainset.games:
question = game.questions[0]
tokens = tknzr.tokenize(question)
all_question.append(tokens)
np.append(all_question, all_question)
for game in validset.games:
question = game.questions[0]
tokens = tknzr.tokenize(question)
np.append(all_question, all_question)