def build_fields(ds: Dataset):
# initialize the frequency counter
counters = defaultdict(Counter)
_src_vocab, _src_vocab_size = inputter._load_vocab(ds.vocab.source, 'src',
counters)
_tgt_vocab, _tgt_vocab_size = inputter._load_vocab(ds.vocab.target, 'tgt',
counters)
# initialize fields
src_nfeats, tgt_nfeats = 0, 0 # do not support word features for now
fields = inputter.get_fields(defaults.vocabulary["data_type"], src_nfeats,
tgt_nfeats)
return inputter._build_fields_vocab(fields, counters,
**defaults.vocabulary)
def maybe_load_vocab(corpus_type, counters, opt):
src_vocab = None
tgt_vocab = None
existing_fields = None
if corpus_type == "train":
if opt.src_vocab != "":
try:
logger.info("Using existing vocabulary...")
existing_fields = torch.load(opt.src_vocab)
except torch.serialization.pickle.UnpicklingError:
logger.info("Building vocab from text file...")
src_vocab, src_vocab_size = _load_vocab(
opt.src_vocab, "src", counters,
opt.src_words_min_frequency)
if opt.tgt_vocab != "":
tgt_vocab, tgt_vocab_size = _load_vocab(
opt.tgt_vocab, "tgt", counters, opt.tgt_words_min_frequency)
return src_vocab, tgt_vocab, existing_fields
def build_dynamic_fields(opts, src_specials=None, tgt_specials=None):
"""Build fields for dynamic, including load & build vocab."""
fields = _get_dynamic_fields(opts)
counters = defaultdict(Counter)
logger.info("Loading vocab from text file...")
_src_vocab, _src_vocab_size = _load_vocab(
opts.src_vocab, 'src', counters, min_freq=opts.src_words_min_frequency)
if opts.src_feats_vocab:
for feat_name, filepath in opts.src_feats_vocab.items():
_, _ = _load_vocab(filepath, feat_name, counters, min_freq=0)
if opts.tgt_vocab:
_tgt_vocab, _tgt_vocab_size = _load_vocab(
opts.tgt_vocab,
'tgt',
counters,
min_freq=opts.tgt_words_min_frequency)
elif opts.share_vocab:
logger.info("Sharing src vocab to tgt...")
counters['tgt'] = counters['src']
else:
raise ValueError("-tgt_vocab should be specified if not share_vocab.")
logger.info("Building fields with vocab in counters...")
fields = _build_fields_vocab(fields,
counters,
'text',
opts.share_vocab,
opts.vocab_size_multiple,
opts.src_vocab_size,
opts.src_words_min_frequency,
opts.tgt_vocab_size,
opts.tgt_words_min_frequency,
src_specials=src_specials,
tgt_specials=tgt_specials)
return fields
def build_save_dataset(corpus_type, fields, src_reader, tgt_reader, opt):
assert corpus_type in ['train', 'valid']
if corpus_type == 'train':
counters = defaultdict(Counter)
srcs = opt.train_src
tgts = opt.train_tgt
ids = opt.train_ids
else:
srcs = [opt.valid_src]
tgts = [opt.valid_tgt]
ids = [None]
logger.info(opt)
for src, tgt, maybe_id in zip(srcs, tgts, ids):
logger.info("Reading source and target files: %s %s." % (src, tgt))
src_shards = split_corpus(src, opt.shard_size)
tgt_shards = split_corpus(tgt, opt.shard_size)
shard_pairs = zip(src_shards, tgt_shards)
dataset_paths = []
if (corpus_type == "train" or opt.filter_valid) and tgt is not None:
filter_pred = partial(inputters.filter_example,
use_src_len=opt.data_type == "text",
max_src_len=opt.src_seq_length,
max_tgt_len=opt.tgt_seq_length)
else:
filter_pred = None
if corpus_type == "train":
existing_fields = None
if opt.src_vocab != "":
try:
logger.info("Using existing vocabulary...")
existing_fields = torch.load(opt.src_vocab)
except torch.serialization.pickle.UnpicklingError:
logger.info("Building vocab from text file...")
src_vocab, src_vocab_size = _load_vocab(
opt.src_vocab, "src", counters,
opt.src_words_min_frequency)
else:
src_vocab = None
if opt.tgt_vocab != "":
tgt_vocab, tgt_vocab_size = _load_vocab(
opt.tgt_vocab, "tgt", counters,
opt.tgt_words_min_frequency)
else:
tgt_vocab = None
for i, (src_shard, tgt_shard) in enumerate(shard_pairs):
assert len(src_shard) == len(tgt_shard)
logger.info("Building shard %d." % i)
# @memray: to be different from normal datasets
dataset = inputters.str2dataset[opt.data_type](
fields,
readers=([src_reader, tgt_reader]
if tgt_reader else [src_reader]),
data=([("src", src_shard),
("tgt", tgt_shard)] if tgt_reader else [("src",
src_shard)]),
dirs=([opt.src_dir, None] if tgt_reader else [opt.src_dir]),
sort_key=inputters.str2sortkey[opt.data_type],
filter_pred=filter_pred)
if corpus_type == "train" and existing_fields is None:
for ex in dataset.examples:
for name, field in fields.items():
try:
f_iter = iter(field)
except TypeError:
f_iter = [(name, field)]
all_data = [getattr(ex, name, None)]
else:
all_data = getattr(ex, name)
for (sub_n, sub_f), fd in zip(f_iter, all_data):
has_vocab = (sub_n == 'src' and
src_vocab is not None) or \
(sub_n == 'tgt' and
tgt_vocab is not None)
if (hasattr(sub_f, 'sequential')
and sub_f.sequential and not has_vocab):
val = fd
if opt.data_type == 'keyphrase' and sub_n == 'tgt':
# in this case, val is a list of phrases (list of strings (words))
for v in val:
counters[sub_n].update(v)
else:
counters[sub_n].update(val)
if maybe_id:
shard_base = corpus_type + "_" + maybe_id
else:
shard_base = corpus_type
data_path = "{:s}.{:s}.{:d}.pt".\
format(opt.save_data, shard_base, i)
dataset_paths.append(data_path)
logger.info(" * saving %sth %s data shard to %s. %d examples" %
(i, corpus_type, data_path, len(dataset.examples)))
dataset.save(data_path)
del dataset.examples
gc.collect()
del dataset
gc.collect()
if corpus_type == "train":
vocab_path = opt.save_data + '.vocab.pt'
if existing_fields is None:
fields = _build_fields_vocab(
fields, counters, opt.data_type, opt.share_vocab,
opt.vocab_size_multiple, opt.src_vocab_size,
opt.src_words_min_frequency, opt.tgt_vocab_size,
opt.tgt_words_min_frequency)
else:
fields = existing_fields
torch.save(fields, vocab_path)
def build_save_dataset(corpus_type, fields, src_reader, opt):
assert corpus_type in ['train', 'valid']
if corpus_type == 'train':
counters = defaultdict(Counter)
srcs = opt.train_src
ids = opt.train_ids
else:
srcs = [opt.valid_src]
ids = [None]
for src, maybe_id in zip(srcs, ids):
logger.info("Reading source files: %s." % src)
# src_shards = split_corpus(src, opt.shard_size)
src_shards = split_corpus(src, 0)
dataset_paths = []
# if (corpus_type == "train" or opt.filter_valid):
# filter_pred = partial(
# inputters.filter_example, use_src_len=opt.data_type == "text",
# max_src_len=opt.src_seq_length, max_tgt_len=opt.tgt_seq_length)
# else:
filter_pred = None
if corpus_type == "train":
existing_fields = None
if opt.src_vocab != "":
try:
logger.info("Using existing vocabulary...")
existing_fields = torch.load(opt.src_vocab)
except torch.serialization.pickle.UnpicklingError:
logger.info("Building vocab from text file...")
src_vocab, src_vocab_size = _load_vocab(
opt.src_vocab, "src", counters,
opt.src_words_min_frequency)
else:
src_vocab = None
for i, _src_shard in enumerate(src_shards): # not considered shard
logger.info("Building shard %d." % i)
src_shard = []
for j, line in enumerate(_src_shard):
if len(line.strip().split("\t")) == 6:
src_shard.append(line)
_id = [line.strip().split("\t")[0] for line in src_shard[1:]]
# _id = [i for i, line in enumerate(src_shard[1:], 1)]
sent1 = [line.strip().split("\t")[3] for line in src_shard[1:]]
sent2 = [line.strip().split("\t")[4] for line in src_shard[1:]]
prelogit1 = [0.0 for _ in src_shard[1:]]
prelogit2 = [0.0 for _ in src_shard[1:]]
label = []
for line in src_shard[1:]:
token = line.strip().split("\t")[5]
if token in ["Good", "entailment", "1", 1]:
label.append(1)
else:
label.append(0)
dataset = inputters.Dataset(
fields,
readers=([
src_reader, src_reader, src_reader, src_reader, src_reader,
src_reader
]),
data=([("id", _id), ("sent1", sent1), ("sent2", sent2),
("label", label), ("prelogit1", prelogit1),
("prelogit2", prelogit2)]),
# data=([("src", src_shard), ("tgt", tgt_shard)]
# if tgt_reader else [("src", src_shard)]),
dirs=([None, None, None, None, None, None]),
sort_key=inputters.str2sortkey[opt.data_type],
filter_pred=filter_pred)
if corpus_type == "train" and existing_fields is None:
for ex in dataset.examples:
for name, field in fields.items():
if name in ["label", "id", "prelogit1", "prelogit2"]:
continue
try:
f_iter = iter(field)
except TypeError:
f_iter = [(name, field)]
all_data = [getattr(ex, name, None)]
else:
all_data = getattr(ex, name)
for (sub_n, sub_f), fd in zip(f_iter, all_data):
has_vocab = (sub_n == 'src' and src_vocab)
if (hasattr(sub_f, 'sequential')
and sub_f.sequential and not has_vocab):
val = fd
counters[sub_n].update(val)
if maybe_id:
shard_base = corpus_type + "_" + maybe_id
else:
shard_base = corpus_type
data_path = "{:s}.{:s}.{:d}.pt".\
format(opt.save_data, shard_base, i)
dataset_paths.append(data_path)
logger.info(" * saving %sth %s data shard to %s." %
(i, shard_base, data_path))
dataset.save(data_path)
del dataset.examples
gc.collect()
del dataset
gc.collect()
if corpus_type == "train":
vocab_path = opt.save_data + '.vocab.pt'
if existing_fields is None:
fields = _build_fields_vocab(fields, counters, opt.data_type,
opt.share_vocab,
opt.vocab_size_multiple,
opt.src_vocab_size,
opt.src_words_min_frequency)
else:
fields = existing_fields
torch.save(fields, vocab_path)
