def get_embedding(vocab, args):
print("{}, Building embedding".format(
datetime.datetime.now().strftime('%02y/%02m/%02d %H:%M:%S')))
# check if loading pre-trained embeddings
if args.bert:
ebd = CXTEBD()
else:
ebd = WORDEBD(vocab)
if args.embedding == 'avg':
model = AVG(ebd, args)
elif args.embedding in ['idf', 'iwf']:
model = IDF(ebd, args)
elif args.embedding in ['meta', 'meta_mlp']:
model = META(ebd, args)
elif args.embedding == 'cnn':
model = CNN(ebd, args)
if args.snapshot != '':
# load pretrained models
print("{}, Loading pretrained embedding from {}".format(
datetime.datetime.now().strftime('%02y/%02m/%02d %H:%M:%S'),
args.snapshot + '.ebd'))
model.load_state_dict(torch.load(args.snapshot + '.ebd'))
if args.cuda != -1:
return model.cuda(args.cuda)
else:
return model
def get_embedding(vocab, args):
print("{}, Building embedding".format(datetime.datetime.now()), flush=True)
ebd = WORDEBD(vocab, args.finetune_ebd)
modelG = ModelG(ebd, args)
# modelD = ModelD(ebd, args)
print("{}, Building embedding".format(datetime.datetime.now()), flush=True)
if args.cuda != -1:
modelG = modelG.cuda(args.cuda)
# modelD = modelD.cuda(args.cuda)
return modelG # , modelD
else:
return modelG # , modelD
def get_embedding(vocab, args):
print("{}, Building embedding".format(
datetime.datetime.now().strftime('%02y/%02m/%02d %H:%M:%S')),
flush=True)
# check if loading pre-trained embeddings
if args.bert:
print('Embedding type: BERT')
ebd = CXTEBD(args.pretrained_bert,
cache_dir=args.bert_cache_dir,
finetune_ebd=args.finetune_ebd,
return_seq=(args.embedding != 'ebd'))
else:
print('Embedding type: WORDEBD')
# WORDEBD returns a neural network layer that maps word tokens to vectors
ebd = WORDEBD(vocab, args.finetune_ebd)
print('Using: ', args.embedding)
if args.embedding == 'avg':
model = AVG(ebd, args)
elif args.embedding in ['idf', 'iwf']:
model = IDF(ebd, args)
elif args.embedding in ['meta', 'meta_mlp']:
model = META(ebd, args)
elif args.embedding == 'cnn':
model = CNN(ebd, args)
elif args.embedding == 'lstmatt':
model = LSTMAtt(ebd, args)
elif args.embedding == 'ebd' and args.bert:
model = ebd # using bert representation directly
print("{}, Building embedding".format(
datetime.datetime.now().strftime('%02y/%02m/%02d %H:%M:%S')),
flush=True)
if args.snapshot != '':
# load pretrained models
print("{}, Loading pretrained embedding from {}".format(
datetime.datetime.now().strftime('%02y/%02m/%02d %H:%M:%S'),
args.snapshot + '.ebd'))
model.load_state_dict(torch.load(args.snapshot + '.ebd'))
if args.cuda != -1:
return model.cuda(args.cuda)
else:
return model
def load_dataset(args):
if args.dataset == '20newsgroup':
train_classes, val_classes, test_classes = _get_20newsgroup_classes(
args)
elif args.dataset == 'amazon':
train_classes, val_classes, test_classes = _get_amazon_classes(args)
elif args.dataset == 'fewrel':
train_classes, val_classes, test_classes = _get_fewrel_classes(args)
elif args.dataset == 'huffpost':
train_classes, val_classes, test_classes = _get_huffpost_classes(args)
elif args.dataset == 'reuters':
train_classes, val_classes, test_classes = _get_reuters_classes(args)
elif args.dataset == 'rcv1':
train_classes, val_classes, test_classes = _get_rcv1_classes(args)
else:
raise ValueError(
'args.dataset should be one of'
'[20newsgroup, amazon, fewrel, huffpost, reuters, rcv1]')
assert (len(train_classes) == args.n_train_class)
assert (len(val_classes) == args.n_val_class)
assert (len(test_classes) == args.n_test_class)
if args.mode == 'finetune':
# in finetune, we combine train and val for training the base classifier
train_classes = train_classes + val_classes
args.n_train_class = args.n_train_class + args.n_val_class
args.n_val_class = args.n_train_class
tprint('Loading data from {}'.format(args.data_path))
all_data = _load_json(args.data_path)
tprint('Loading word vectors')
path = os.path.join(args.wv_path, args.word_vector)
if not os.path.exists(path):
# Download the word vector and save it locally:
tprint('Downloading word vectors')
import urllib.request
urllib.request.urlretrieve(
'https://dl.fbaipublicfiles.com/fasttext/vectors-wiki/wiki.en.vec',
path)
vectors = Vectors(args.word_vector, cache=args.wv_path)
vocab = Vocab(collections.Counter(_read_words(all_data)),
vectors=vectors,
specials=['<pad>', '<unk>'],
min_freq=5)
# print word embedding statistics
wv_size = vocab.vectors.size()
tprint('Total num. of words: {}, word vector dimension: {}'.format(
wv_size[0], wv_size[1]))
num_oov = wv_size[0] - torch.nonzero(
torch.sum(torch.abs(vocab.vectors), dim=1)).size()[0]
tprint(('Num. of out-of-vocabulary words'
'(they are initialized to zeros): {}').format(num_oov))
# Split into meta-train, meta-val, meta-test data
train_data, val_data, test_data = _meta_split(all_data, train_classes,
val_classes, test_classes)
tprint('#train {}, #val {}, #test {}'.format(len(train_data),
len(val_data),
len(test_data)))
# Convert everything into np array for fast data loading
train_data = _data_to_nparray(train_data, vocab, args)
val_data = _data_to_nparray(val_data, vocab, args)
test_data = _data_to_nparray(test_data, vocab, args)
train_data['is_train'] = True
# this tag is used for distinguishing train/val/test when creating source pool
stats.precompute_stats(train_data, val_data, test_data, args)
if args.meta_w_target:
# augment meta model by the support features
if args.bert:
ebd = CXTEBD(args.pretrained_bert,
cache_dir=args.bert_cache_dir,
finetune_ebd=False,
return_seq=True)
else:
ebd = WORDEBD(vocab, finetune_ebd=False)
train_data['avg_ebd'] = AVG(ebd, args)
if args.cuda != -1:
train_data['avg_ebd'] = train_data['avg_ebd'].cuda(args.cuda)
val_data['avg_ebd'] = train_data['avg_ebd']
test_data['avg_ebd'] = train_data['avg_ebd']
# if finetune, train_classes = val_classes and we sample train and val data
# from train_data
if args.mode == 'finetune':
train_data, val_data = _split_dataset(train_data, args.finetune_split)
return train_data, val_data, test_data, vocab
