Spaces:
Runtime error
Runtime error
File size: 40,295 Bytes
62e9ca6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 |
# ----------------------------------------------------------------------------
# SpeechUT: Bridging Speech and Text with Hidden-Unit for Encoder-Decoder Based Speech-Text Pre-training (https://arxiv.org/abs/2210.03730)
# Github source: https://github.com/microsoft/SpeechT5/tree/main/SpeechUT
# Code based on fairseq: https://github.com/facebookresearch/fairseq/tree/272c4c5197250997148fb12c0db6306035f166a4
#
# Copyright (c) 2022 Microsoft
# Licensed under The MIT License [see LICENSE for details]
# ----------------------------------------------------------------------------
import logging
import os
import sys
from typing import Dict, List, Optional, Tuple
from pathlib import Path
import numpy as np
from argparse import Namespace
from collections import OrderedDict
import torch
from dataclasses import dataclass, field
from fairseq.data import (
Dictionary,
encoders,
data_utils,
StripTokenDataset,
PrependTokenDataset,
AppendTokenDataset,
DenoisingDataset,
ConcatDataset,
FairseqDataset,
iterators,
ResamplingDataset,
MaskTokensDataset,
LanguagePairDataset,
)
from fairseq.data.audio.speech_to_text_joint_dataset import S2TJointDataConfig
from fairseq.data.shorten_dataset import maybe_shorten_dataset
# from fairseq.data.encoders.utils import get_whole_word_mask
from fairseq.dataclass.configs import FairseqDataclass
from fairseq.tasks import register_task
from fairseq.tasks.fairseq_task import FairseqTask
from fairseq.dataclass.constants import ChoiceEnum
from omegaconf import MISSING
from speechut.data.multimodal_corpus_dataset import MultiCorpusDataset
from speechut.data.load_langpair_dataset import load_langpair_dataset
from speechut.data.language_trible_dataset import LanguageTripleDataset, load_langtriple_dataset
from speechut.data.hubert_dataset import HubertDataset
logger = logging.getLogger(__name__)
TOKENIZER_CHOICES = ChoiceEnum(["sentencepiece", "hubert_letters", "none"])
def _lang_token(lang: str):
return "<lang:{}>".format(lang)
def _lang_token_index(dic: Dictionary, lang: str):
"""Return language token index."""
idx = dic.index(_lang_token(lang))
assert idx != dic.unk_index, "cannot find language token for lang {}".format(lang)
return idx
class LabelEncoder(object):
def __init__(self, dictionary: Dictionary) -> None:
self.dictionary = dictionary
def __call__(self, label: str) -> List[str]:
return self.dictionary.encode_line(
label, append_eos=False, add_if_not_exist=False,
)
### wrap the initial get_whole_word_mask which needs bpe_tokenizer,
### here we just assume words are splited by "|" or "<SIL>"
def get_whole_word_mask(args, dictionary):
def is_beginning_of_word(i):
if i < dictionary.nspecial:
# special elements are always considered beginnings
return True
tok = dictionary[i]
if tok.startswith("madeupword"):
return True
elif tok in ["<unk>", "<s>", "</s>", "<pad>", "|", "<eps>"]:
return True
else:
return False
mask_whole_words = torch.ByteTensor(
list(map(is_beginning_of_word, range(len(dictionary))))
)
return mask_whole_words
def get_repeative_start(tokens):
"""
tokens: torch.Tensor with repeative tokens
"""
length = len(tokens)
rep_start_id = tokens[:-1] != tokens[1:]
return torch.cat([torch.tensor([True]), rep_start_id])
@dataclass
class TextPretrainingConfig(FairseqDataclass):
### added for joint pretraining
text_data: Optional[str] = field(
default=None,
metadata={
"help": "if set, path to text data directory",
},
)
seed: Optional[int] = field(
default=1,
metadata={
"help": "for ordered_indices in MulticorpusDataset",
},
)
tokens_per_sample: Optional[int] = field(
default=512,
metadata={
"help": "max number of total tokens over all segments per sample for dataset",
},
)
tokens_per_sample_tgt: Optional[int] = field(
default=512,
metadata={
"help": "max number of total tokens over all segments per target sample for dataset",
},
)
sample_break_mode: Optional[str] = field(
default="eos",
metadata={
"help": "mode for breaking sentence",
},
)
mask: Optional[float] = field(
default=0.3,
metadata={
"help": "fraction of words/subwords that will be masked",
},
)
leave_unmasked_prob: float = field(
default=0.1,
metadata={"help": "probability that a masked token is unmasked"},
)
mask_random: Optional[float] = field(
default=0.1,
metadata={
"help": "instead of using [MASK], use random token this often",
},
)
freq_weighted_replacement: bool = field(
default=False,
metadata={"help": "sample random replacement words based on word frequencies"},
)
mask_whole_words: bool = field(
default=True,
metadata={"help": "mask whole words; you may also want to set --bpe"},
)
mask_repeative_tokens: bool = field(
default=True,
metadata={"help": "mask repeative_tokens; if mask_whole_words=False"},
)
mask_multiple_length: int = field(
default=1,
metadata={"help": "repeat the mask indices multiple times"},
)
mask_stdev: float = field(
default=0.0,
metadata={"help": "stdev of the mask length"},
)
shorten_method: Optional[str] = field(
default="none",
metadata={
"help": "if not none, shorten sequences that exceed tokens_per_sample",
"choices": "none/truncate/random_crop"
},
)
shorten_data_split_list: Optional[str] = field(
default="",
metadata={
"help": "comma_separated list of dataset splits to apply shortening to, e.g., train,valid (default: all dataset splits)",
},
)
### below hypra-parameters is used in bart
insert: Optional[float] = field(
default=0.0,
metadata={
"help": "insert this percentage of additional random tokens",
},
)
permute: Optional[float] = field(
default=0.0,
metadata={
"help": "take this proportion of subwords and permute them",
},
)
rotate: Optional[float] = field(
default=0.0,
metadata={
"help": "rotate this proportion of inputs",
},
)
poisson_lambda: Optional[float] = field(
default=3.5,
metadata={
"help": "randomly shuffle sentences for this proportion of inputs",
},
)
permute_sentences: Optional[float] = field(
default=0.0,
metadata={
"help": "shuffle this proportion of sentences in all inputs",
},
)
mask_length: Optional[str] = field(
default="span-poisson",
metadata={
"help": "mask length to choose",
"choice": "subword/word/span-poisson"
},
)
replace_length: Optional[int] = field(
default=1,
metadata={
"help": "when masking N tokens, replace with 0, 1, or N tokens (use -1 for N)",
},
)
shuffle_instance: Optional[bool] = field(
default=False,
metadata={"help": "shuffle instance"},
)
max_source_positions: Optional[int] = field(
default=1024,
metadata={"help": "max number of tokens in the source sequence"},
)
max_target_positions: Optional[int] = field(
default=1024,
metadata={"help": "max number of tokens in the target sequence"},
)
bpe: Optional[str] = field(
default="",
metadata={
"help": "will wrapped by the text_data_config yaml",
},
)
data_config: Optional[str] = field(
default=None,
metadata={
"help": "a config yaml specify the bpe model of text data",
},
)
text_maxtokens_ratio: Optional[float] = field(
default=1.0,
metadata={
"help": "for text, max_tokens = max_tokens * text_maxtokens_ratio / 320 ",
},
)
prepend_tgt_lang_tag: bool = field(
default=False,
metadata={"help": "prepend tgt_lang_tag to replace <eos>"},
)
mask_text_ratio: Optional[float] = field(
default=0.0,
metadata={
"help": "mask_text_ratio, for paired data",
},
)
truncate_mono_source: bool = field(
default=True,
metadata={"help": "truncate mono source-side examples that exceed max-positions"},
)
@dataclass
class JointPretrainingConfig(FairseqDataclass):
data: str = field(
default=MISSING, metadata={"help": "path to speech data directory"}
)
fine_tuning: bool = field(
default=False, metadata={"help": "set to true if fine-tuning Hubert"}
)
labels: List[str] = field(
default_factory=lambda: ["ltr"],
metadata={
"help": (
"extension of the label files to load, frame-level labels for"
" pre-training, and sequence-level label for fine-tuning"
)
},
)
label_dir: Optional[str] = field(
default=None,
metadata={
"help": "if set, looks for labels in this directory instead",
},
)
label_rate: int = field(
default=-1,
metadata={"help": "label frame rate. -1 for sequence label"},
)
sample_rate: int = field(
default=16_000,
metadata={
"help": "target sample rate. audio files will be up/down "
"sampled to this rate"
},
)
normalize: bool = field(
default=False,
metadata={
"help": "if set, normalizes input to have 0 mean and unit variance"
},
)
enable_padding: bool = field(
default=False,
metadata={"help": "pad shorter samples instead of cropping"},
)
max_keep_size: Optional[int] = field(
default=None,
metadata={"help": "exclude sample longer than this"},
)
max_sample_size: Optional[int] = field(
default=None,
metadata={"help": "max sample size to crop to for batching"},
)
min_sample_size: Optional[int] = field(
default=None,
metadata={"help": "min sample size to crop to for batching"},
)
single_target: Optional[bool] = field(
default=False,
metadata={
"help": "if set, AddTargetDatasets outputs same keys "
"as AddTargetDataset"
},
)
random_crop: Optional[bool] = field(
default=True,
metadata={"help": "always crop from the beginning if false"},
)
pad_audio: Optional[bool] = field(
default=False,
metadata={"help": "pad audio to the longest one in the batch if true"},
)
store_labels: Optional[bool] = field(
default=True,
metadata={"help": "store spm labels in memory, should be true when fine-tune with bpe"},
)
add_decoder_target: bool = field(
default=False,
metadata={"help": "contral the model architecture, if set True, load reduced unit as target"},
)
split_modality_batch: bool = field(
default=False,
metadata={"help": "whether create all samples of different modalities in a batch"},
)
speech_tgt_lang: str = field(
default="",
metadata={"help": "prepend <tgt-id> to prev_output_tokens to replace <eos>, only used for decoder"},
)
speech_sampling_alpha: float = field(
default=0.2,
metadata={
"help": "Hyper-parameter alpha = 1/T for temperature-based speech resampling."
"(alpha = 1 for no resampling)"
},
)
text_sampling_alpha: float = field(
default=0.2,
metadata={
"help": "Hyper-parameter alpha = 1/T for temperature-based text resampling."
"(alpha = 1 for no resampling)"
},
)
hubert_tokenizer: Optional[TOKENIZER_CHOICES] = field(
default="none",
metadata={"help": "which tokenizer for processing text"},
)
sp_path: Optional[str] = field(
default=None,
metadata={"help": "sentencepiece model path if using bpe tokenizer"},
)
text_cfg: TextPretrainingConfig = TextPretrainingConfig()
# For inference
ctc_weight: float = field(
default=0.0,
metadata={"help": "ctc weight during inference"},
)
lm_dict: Optional[str] = field(
default="dict.txt",
metadata={"help": "dict used for decoding with language model, should be in cfg.data/"},
)
@register_task("joint_sc2t_pretraining", dataclass=JointPretrainingConfig)
class Jsc2tPretrainingTask(FairseqTask):
cfg: JointPretrainingConfig
def __init__(
self,
cfg: JointPretrainingConfig,
load_local_states: True,
) -> None:
super().__init__(cfg)
logger.info(f"current directory is {os.getcwd()}")
logger.info(f"JSTPretrainingTask Config {cfg}")
self.cfg = cfg
self.fine_tuning = cfg.fine_tuning
self.blank_symbol = "<s>"
if load_local_states:
self.state.add_factory("hubert_tokenizer", self.build_tokenizer)
if self.cfg.text_cfg.text_data is not None and os.path.exists(self.cfg.text_cfg.text_data):
self.state.add_factory("text_dictionary", self.load_text_dictionary)
self.state.add_factory("text_src_dictionary", self.load_text_src_dictionary)
if cfg.fine_tuning:
self.state.add_factory("target_dictionary", self.load_dictionaries)
else:
self.state.add_factory("dictionaries", self.load_dictionaries)
if cfg.text_cfg.data_config is not None:
self.text_data_cfg = S2TJointDataConfig(Path(f"{cfg.text_cfg.text_data}/{cfg.text_cfg.data_config}"))
self.cfg.text_cfg.bpe = self.text_data_cfg.bpe_tokenizer["bpe"]
else:
self.text_data_cfg = None
@property
def source_dictionary(self) -> Optional[Dictionary]:
return None
@property
def target_dictionary(self) -> Optional[Dictionary]:
return self.state.target_dictionary
@property
def dictionaries(self) -> List[Dictionary]:
return self.state.dictionaries
@property
def text_dictionary(self) -> Optional[Dictionary]:
return self.state.text_dictionary
@property
def text_src_dictionary(self) -> Optional[Dictionary]:
return self.state.text_src_dictionary
@property
def hubert_tokenizer(self):
return self.state.hubert_tokenizer
def load_dictionaries(self):
label_dir = self.cfg.data if self.cfg.label_dir is None else self.cfg.label_dir
dictionaries = [Dictionary.load(f"{label_dir}/dict.{label}.txt") for label in self.cfg.labels]
if not self.cfg.fine_tuning:
for dictionary in dictionaries:
dictionary.add_symbol("<mask>")
return dictionaries[0] if self.cfg.fine_tuning else dictionaries
def load_text_dictionary(self):
tgt_dict_path = f"{self.cfg.text_cfg.text_data}/{self.text_data_cfg.vocab_filename if self.text_data_cfg is not None else 'dict.txt'}"
if not os.path.isfile(tgt_dict_path):
raise FileNotFoundError(f"Dict not found: {tgt_dict_path}")
text_dictionary = Dictionary.load(tgt_dict_path)
self.mask_idx = text_dictionary.add_symbol("<mask>")
return text_dictionary
def load_text_src_dictionary(self):
src_dict_path = f"{self.cfg.text_cfg.text_data}/{self.text_data_cfg.src_vocab_filename if self.text_data_cfg is not None else 'dict.txt'}"
if not os.path.isfile(src_dict_path):
raise FileNotFoundError(f"Dict not found: {src_dict_path}")
src_text_dictionary = Dictionary.load(src_dict_path)
self.mask_idx = src_text_dictionary.add_symbol("<mask>")
return src_text_dictionary
@classmethod
def setup_task(
cls, cfg: JointPretrainingConfig, **kwargs
) -> "Jsc2tPretrainingTask":
load_local_states = kwargs.get("load_local_states", True)
return cls(cfg, load_local_states)
def get_label_dir(self) -> str:
if self.cfg.label_dir is None:
return self.cfg.data
return self.cfg.label_dir
def load_paired_dataset(self, text_split, truncate_source=False):
text_split, lp = text_split.rsplit('.', 1) # e.g. "libritext.ltr-ltr"
if len(lp.split("-")) == 2:
src, tgt = lp.split("-")
if src == tgt:
logger.warn(f"| trying to load monolingual dataset {text_split}.{lp}, please check your task is right.")
paired_dataset = self.load_char_bart_dataset(f"{text_split}.{lp}.{tgt}")
return paired_dataset
paired_dataset = load_langpair_dataset(
self.cfg.text_cfg.text_data,
text_split,
src,
self.text_src_dictionary,
tgt,
self.text_dictionary,
combine=True,
dataset_impl=None,
upsample_primary=1,
left_pad_source=False,
left_pad_target=False,
max_source_positions=self.cfg.text_cfg.tokens_per_sample,
max_target_positions=self.cfg.text_cfg.tokens_per_sample,
truncate_source=truncate_source,
prepend_bos=False,
load_alignments=False,
append_source_id=True if self.cfg.text_cfg.prepend_tgt_lang_tag else False,
lang_format="<lang:{}>" if self.cfg.text_cfg.prepend_tgt_lang_tag else "[{}]",
input_feeding=self.cfg.add_decoder_target,
)
if self.cfg.text_cfg.mask_text_ratio > 0:
# add mask
self.mask_idx = self.text_src_dictionary.index("<mask>")
mask_whole_words = None
if self.cfg.text_cfg.mask_whole_words:
mask_whole_words = get_whole_word_mask(self.cfg.text_cfg, self.text_src_dictionary)
elif self.cfg.text_cfg.mask_repeative_tokens:
mask_whole_words = get_repeative_start
src_dataset, src_unmasked_dataset = MaskTokensDataset.apply_mask(
paired_dataset.src,
self.text_src_dictionary,
pad_idx=self.text_src_dictionary.pad(),
mask_idx=self.mask_idx,
seed=self.cfg.text_cfg.seed,
mask_prob=self.cfg.text_cfg.mask_text_ratio,
leave_unmasked_prob=self.cfg.text_cfg.leave_unmasked_prob,
random_token_prob=self.cfg.text_cfg.mask_random,
freq_weighted_replacement=self.cfg.text_cfg.freq_weighted_replacement,
mask_whole_words=mask_whole_words,
mask_multiple_length=self.cfg.text_cfg.mask_multiple_length,
mask_stdev=self.cfg.text_cfg.mask_stdev,
)
tgt_dataset = paired_dataset.tgt if paired_dataset.tgt is not None else src_unmasked_dataset
paired_dataset = LanguageTripleDataset(
src_dataset,
src_dataset.sizes,
self.text_src_dictionary,
src_unmasked_dataset,
src_unmasked_dataset.sizes,
self.text_src_dictionary,
tgt_dataset,
tgt_dataset.sizes,
self.text_dictionary,
left_pad_source=False,
left_pad_target=False,
align_dataset=None,
eos=None,
num_buckets=0,
shuffle=True,
pad_to_multiple=1,
)
else:
src, ref, tgt = lp.split("-")
paired_dataset = load_langtriple_dataset(
self.cfg.text_cfg.text_data,
text_split,
src,
self.text_src_dictionary,
ref,
self.dictionaries[-1],
tgt,
self.text_dictionary,
combine=True,
dataset_impl=None,
upsample_primary=1,
left_pad_source=False,
left_pad_target=False,
max_source_positions=self.cfg.text_cfg.tokens_per_sample,
max_target_positions=self.cfg.text_cfg.tokens_per_sample,
truncate_source=truncate_source,
prepend_bos=False,
load_alignments=False,
append_source_id=True if self.cfg.text_cfg.prepend_tgt_lang_tag else False,
lang_format="<lang:{}>" if self.cfg.text_cfg.prepend_tgt_lang_tag else "[{}]",
)
return paired_dataset
def load_dataset(self, split: str, epoch=1, **kwargs) -> None:
"""
Create Wav dataset for audio, and Index dataset for phonemized text,
then concatenate them to by fairseq.data.multi_corpus_dataset.MultiCorpusDataset.
"""
speech_splits = split.split('+')[0].split(',')
### 1st, create a speech dataset using STSpeechDataset (modified from HubertDataset)
dicts = [self.target_dictionary] if self.cfg.fine_tuning else self.dictionaries
pad_list = [dict.pad() for dict in dicts]
eos_list = [dict.eos() for dict in dicts]
procs = [LabelEncoder(dict) for dict in dicts]
if self.cfg.speech_tgt_lang != "":
tgt_lang_idx = _lang_token_index(dicts[0], self.cfg.speech_tgt_lang)
logger.info(f"Will prepend <{tgt_lang_idx}> at the beginning of prev_output_tokens to replace <eos>")
else:
tgt_lang_idx = None
# hubert v1: pad_audio=True, random_crop=False;
speech_datasets = []
for speech_split in speech_splits:
paths = [
f"{self.get_label_dir()}/{speech_split}.{l}" for l in self.cfg.labels
]
speech_datasets.append(
HubertDataset(
f"{self.cfg.data}/{speech_split}.tsv",
sample_rate=self.cfg.sample_rate,
label_paths=paths,
label_rates=self.cfg.label_rate,
pad_list=pad_list,
eos_list=eos_list,
label_processors=procs,
max_keep_sample_size=self.cfg.max_keep_size,
min_keep_sample_size=self.cfg.min_sample_size,
max_sample_size=self.cfg.max_sample_size,
pad_audio=self.cfg.pad_audio,
normalize=self.cfg.normalize,
store_labels=self.cfg.store_labels,
random_crop=self.cfg.random_crop,
single_target=self.cfg.single_target,
tgt_dict=dicts[0],
add_decoder_target=self.cfg.add_decoder_target,
fine_tuning=self.cfg.fine_tuning,
tgt_lang_idx=tgt_lang_idx,
tokenizer=self.hubert_tokenizer,
)
)
if len(speech_datasets) > 1:
speech_dataset = ConcatDataset(speech_datasets)
else:
speech_dataset = speech_datasets[0]
has_text = len(split.split('+')) > 1
if not has_text:
assert speech_dataset is not None
self.datasets[split] = speech_dataset
return
### 2nd, create paired/mono text datasets using Langpairdataset
if split.split('+')[1] != '':
paired_splits = [paired_split for paired_split in split.split('+')[1].split(',') if paired_split != '']
paired_datasets = [self.load_paired_dataset(paired_split) for paired_split in paired_splits]
else:
paired_splits, paired_datasets = [], []
if len(split.split('+')) > 2 and split.split('+')[2] != '':
mono_splits = [mono_split for mono_split in split.split('+')[2].split(',') if mono_split != '']
mono_datasets = [self.load_paired_dataset(mono_split, truncate_source=self.cfg.text_cfg.truncate_mono_source) for mono_split in mono_splits]
else:
mono_splits, mono_datasets = [], []
assert len(mono_datasets + paired_datasets) > 0, f"split {split} has no text! you should check out for that"
### 3rd, if provided, create a supervised dataset with labeled data
if len(split.split('+')) > 3 and split.split('+')[3] != '':
assert len(paired_splits) > 0, f"supervised dataset can not be loaded without text paired dataset!"
tgt = paired_splits[0].rsplit('.', 1)[1].split("-")[1]
sup_split = split.split('+')[3]
sup_dataset = HubertDataset(
f"{self.cfg.data}/{sup_split}.tsv",
sample_rate=self.cfg.sample_rate,
label_paths=[f"{self.get_label_dir()}/{sup_split}.{tgt}"],
label_rates=[-1],
pad_list=[self.text_dictionary.pad()],
eos_list=[self.text_dictionary.eos()],
label_processors=[LabelEncoder(self.text_dictionary)],
max_keep_sample_size=self.cfg.max_keep_size,
min_keep_sample_size=None,
max_sample_size=None,
pad_audio=True,
normalize=self.cfg.normalize,
store_labels=self.cfg.store_labels,
random_crop=False,
single_target=True,
tgt_dict=self.text_dictionary,
add_decoder_target=self.cfg.add_decoder_target,
fine_tuning=True,
tgt_lang_idx=None,
tokenizer=None,
)
else:
sup_dataset = None
### 4th, compose a MultiCorpusDataset
dataset_dict, max_positions_dict, distributions, max_tokens_ratios = self.resample_multi_modality_dataset(
speech_dataset, sup_dataset, mono_datasets, paired_datasets, mono_splits, paired_splits, epoch=epoch,
)
self.datasets[split] = MultiCorpusDataset(
dataset_dict,
max_positions=max_positions_dict,
distribution=distributions,
max_tokens_ratio=max_tokens_ratios,
seed=self.cfg.text_cfg.seed,
sort_indices=True,
)
def max_positions(self) -> Tuple[int, int]:
return (sys.maxsize, sys.maxsize)
def filter_indices_by_size(
self, indices: np.array, *args, **kwargs
) -> np.array:
return indices
def get_batch_iterator(
self,
dataset,
max_tokens=None,
max_sentences=None,
max_positions=None,
ignore_invalid_inputs=False,
required_batch_size_multiple=1,
seed=1,
num_shards=1,
shard_id=0,
num_workers=0,
epoch=1,
data_buffer_size=0,
disable_iterator_cache=False,
skip_remainder_batch=False,
grouped_shuffling=False,
update_epoch_batch_itr=False,
):
"""
Get an iterator that yields batches of data from the given dataset.
Args:
dataset (~fairseq.data.FairseqDataset): dataset to batch
max_tokens (int, optional): max number of tokens in each batch
(default: None).
max_sentences (int, optional): max number of sentences in each
batch (default: None).
max_positions (optional): max sentence length supported by the
model (default: None).
ignore_invalid_inputs (bool, optional): don't raise Exception for
sentences that are too long (default: False).
required_batch_size_multiple (int, optional): require batch size to
be a multiple of N (default: 1).
seed (int, optional): seed for random number generator for
reproducibility (default: 1).
num_shards (int, optional): shard the data iterator into N
shards (default: 1).
shard_id (int, optional): which shard of the data iterator to
return (default: 0).
num_workers (int, optional): how many subprocesses to use for data
loading. 0 means the data will be loaded in the main process
(default: 0).
epoch (int, optional): the epoch to start the iterator from
(default: 1).
data_buffer_size (int, optional): number of batches to
preload (default: 0).
disable_iterator_cache (bool, optional): don't cache the
EpochBatchIterator (ignores `FairseqTask::can_reuse_epoch_itr`)
(default: False).
skip_remainder_batch (bool, optional): if set, discard the last
batch in each training epoch, as the last batch is often smaller than
local_batch_size * distributed_word_size (default: ``True``).
grouped_shuffling (bool, optional): group batches with each groups
containing num_shards batches and shuffle groups. Reduces difference
between sequence lengths among workers for batches sorted by length.
update_epoch_batch_itr (bool optional): if true then donot use the cached
batch iterator for the epoch
Returns:
~fairseq.iterators.EpochBatchIterator: a batched iterator over the
given dataset split
"""
if self.fine_tuning or not isinstance(dataset, MultiCorpusDataset):
return super().get_batch_iterator(
dataset,
max_tokens=max_tokens,
max_sentences=max_sentences,
max_positions=max_positions,
ignore_invalid_inputs=ignore_invalid_inputs,
required_batch_size_multiple=required_batch_size_multiple,
seed=seed,
num_shards=num_shards,
shard_id=shard_id,
num_workers=num_workers,
epoch=epoch,
data_buffer_size=data_buffer_size,
disable_iterator_cache=disable_iterator_cache,
skip_remainder_batch=skip_remainder_batch,
grouped_shuffling=grouped_shuffling,
update_epoch_batch_itr=update_epoch_batch_itr,
)
can_reuse_epoch_itr = (
not disable_iterator_cache
and not update_epoch_batch_itr
and self.can_reuse_epoch_itr(dataset)
)
if can_reuse_epoch_itr and dataset in self.dataset_to_epoch_iter:
logger.debug("reusing EpochBatchIterator for epoch {}".format(epoch))
return self.dataset_to_epoch_iter[dataset]
assert isinstance(dataset, FairseqDataset)
# initialize the dataset with the correct starting epoch
dataset.set_epoch(epoch)
# get indices ordered by example size
with data_utils.numpy_seed(seed):
indices = dataset.ordered_indices()
# filter examples that are too large
if max_positions is not None:
indices = self.filter_indices_by_size(
indices, dataset, max_positions, ignore_invalid_inputs
)
# create mini-batches with given size constraints
batch_sampler = dataset.get_batch_sampler(
indices,
num_shards,
seed,
max_tokens=max_tokens,
max_sentences=max_sentences,
required_batch_size_multiple=required_batch_size_multiple,
split_modality_batch=self.cfg.split_modality_batch,
)
# return a reusable, sharded iterator
epoch_iter = iterators.EpochBatchIterator(
dataset=dataset,
collate_fn=dataset.collater,
batch_sampler=batch_sampler,
seed=seed,
num_shards=num_shards,
shard_id=shard_id,
num_workers=num_workers,
epoch=epoch,
buffer_size=data_buffer_size,
skip_remainder_batch=skip_remainder_batch,
disable_shuffling=True,
grouped_shuffling=grouped_shuffling,
)
if can_reuse_epoch_itr:
self.dataset_to_epoch_iter[dataset] = epoch_iter
return epoch_iter
def build_generator(
self,
models,
args,
seq_gen_cls=None,
extra_gen_cls_kwargs=None,
):
"""Build ED-CTC generator for finet-tuned ASR model"""
from speechut.squence_generator import SequenceGenerator
extra_gen_cls_kwargs = {
"ctc_weight": self.cfg.ctc_weight,
"lm_dict": Dictionary.load(os.path.join(self.cfg.data, self.cfg.lm_dict)),
**extra_gen_cls_kwargs
}
return super().build_generator(
models, args, seq_gen_cls=SequenceGenerator, extra_gen_cls_kwargs=extra_gen_cls_kwargs
)
@classmethod
def _get_size_ratios(cls, ids: List[str], sizes: List[int], alpha: float = 1.0):
"""Size ratios for temperature-based sampling
(https://arxiv.org/abs/1907.05019)"""
_sizes = np.array(sizes)
prob = _sizes / _sizes.sum()
smoothed_prob = prob ** alpha
smoothed_prob = smoothed_prob / smoothed_prob.sum()
size_ratio = (smoothed_prob * _sizes.sum()) / _sizes
o_str = str({_i: f"{prob[i]:.3f}" for i, _i in enumerate(ids)})
logger.info(f"original sampling probability: {o_str}")
p_str = str({_i: f"{smoothed_prob[i]:.3f}" for i, _i in enumerate(ids)})
logger.info(f"balanced sampling probability: {p_str}")
sr_str = str({_id: f"{size_ratio[i]:.3f}" for i, _id in enumerate(ids)})
logger.info(f"balanced sampling size ratio: {sr_str}")
return size_ratio.tolist()
def resample_multi_modality_dataset(self, speech_dataset, sup_dataset, mono_datasets, paired_datasets, mono_splits, paired_splits, epoch=1, train=True):
assert len(mono_datasets+paired_datasets) > 0, f"No text data loaded!"
if len(mono_datasets) > 1 and self.cfg.text_sampling_alpha != 1.0:
size_ratios = self._get_size_ratios(
mono_splits, [len(s) for s in mono_datasets], alpha=self.cfg.text_sampling_alpha
)
mono_datasets = [
ResamplingDataset(
d, size_ratio=r, seed=0, epoch=epoch, replace=(r >= 1.0)
) for d, r in zip(mono_datasets, size_ratios)
]
if len(paired_datasets) > 1 and self.cfg.text_sampling_alpha != 1.0:
size_ratios = self._get_size_ratios(
paired_splits, [len(s) for s in paired_datasets], alpha=self.cfg.text_sampling_alpha
)
paired_datasets = [
ResamplingDataset(
d, size_ratio=r, seed=0, epoch=epoch, replace=(r >= 1.0)
) for d, r in zip(paired_datasets, size_ratios)
]
dataset_list = [speech_dataset, sup_dataset]
for datasets in [mono_datasets, paired_datasets]:
if len(datasets) > 1:
dataset_list.append(ConcatDataset(datasets))
elif len(datasets) == 1:
dataset_list.append(datasets[0])
else:
dataset_list.append(None)
### match speech/text datasets according to modality
dataset_dict = OrderedDict((name, d) for name, d in zip(["speech", "speech_sup", "text_mono", "text_paired"], dataset_list) if d is not None)
max_positions_dict = {
"speech": None,
"speech_sup": None,
"text_mono": (self.cfg.text_cfg.tokens_per_sample, self.cfg.text_cfg.tokens_per_sample),
"text_paired": (self.cfg.text_cfg.tokens_per_sample, self.cfg.text_cfg.tokens_per_sample),
}
max_positions_dict = OrderedDict((name, max_positions_dict[name]) for name in dataset_dict.keys())
max_tokens_ratios_dict = {
"speech": 1.0,
"speech_sup": 1.0,
"text_mono": 1.0 / 320 / self.cfg.text_cfg.text_maxtokens_ratio,
"text_paired": 1.0 / 320 / self.cfg.text_cfg.text_maxtokens_ratio,
}
max_tokens_ratios = [max_tokens_ratios_dict[name] for name in dataset_dict.keys()]
dataset_lens = np.array([len(dataset) for dataset in dataset_dict.values()])
dataset_avg_sample_lens = np.array([
sum([dataset.num_tokens(i) for i in np.random.randint(low=0, high=len(dataset), size=10000)]) / 10000.0
for dataset in dataset_dict.values()
])
if not "speech" in dataset_dict:
distributions = [l / sum(dataset_lens) for l in dataset_lens]
else:
## we just keep the batches of speech and non-speech the same, expand_coef is to ensure speech batches is less than others
first_ratio = dataset_lens[0] / sum(dataset_lens)
expand_coef = 1.2 if sup_dataset is None else 1.1 * sum(dataset_lens[0:2]) / dataset_lens[0]
distributions = [expand_coef * max_tokens_ratios[i] * dataset_avg_sample_lens[0] / l for (i, l) in enumerate(dataset_avg_sample_lens)]
distributions[0] = 1.0
if sup_dataset is not None:
distributions[1] = dataset_lens[1] / dataset_lens[0]
distributions = [first_ratio * d for d in distributions]
logging.info(f"Number samples of datasets is {dataset_lens}")
logging.info(f"Avg sample length of datasets is {dataset_avg_sample_lens}")
logging.info(f"Sampling distributions is {distributions}")
logging.info(f"Maxtokens ratio is {max_tokens_ratios}")
return dataset_dict, max_positions_dict, distributions, max_tokens_ratios
def build_tokenizer(self, cfg=None):
logger.info(f"tokenizer: {self.cfg.hubert_tokenizer}")
if self.cfg.hubert_tokenizer != "none":
return encoders.build_bpe(Namespace(**{"bpe": self.cfg.hubert_tokenizer, "sentencepiece_model": self.cfg.sp_path}))
else:
return None
def load_char_bart_dataset(self, split):
mono_dataset = data_utils.load_indexed_dataset(
f"{self.cfg.text_cfg.text_data}/{split}",
self.text_dictionary,
)
mono_dataset = StripTokenDataset(mono_dataset, self.text_dictionary.eos())
mono_dataset = maybe_shorten_dataset(
mono_dataset,
split,
self.cfg.text_cfg.shorten_data_split_list,
self.cfg.text_cfg.shorten_method,
self.cfg.text_cfg.tokens_per_sample - 2,
self.cfg.text_cfg.seed,
)
logger.info("loaded {} samples from: {}".format(len(mono_dataset), mono_dataset))
### prepend bos and eos to dataset
mono_dataset = PrependTokenDataset(mono_dataset, self.text_dictionary.bos())
mono_dataset = AppendTokenDataset(mono_dataset, self.text_dictionary.eos())
mask_whole_words = (
get_whole_word_mask(None, self.text_dictionary)
if self.cfg.text_cfg.mask_whole_words
else None
)
lang=self.cfg.speech_tgt_lang
mono_dataset = DenoisingDataset(
mono_dataset,
mono_dataset.sizes,
self.text_dictionary,
self.mask_idx,
mask_whole_words,
shuffle=self.cfg.text_cfg.shuffle_instance,
seed=self.cfg.text_cfg.seed,
args=self.cfg.text_cfg,
tgt_lang_idx=_lang_token_index(self.text_dictionary, lang) if self.cfg.text_cfg.prepend_tgt_lang_tag else None,
)
return mono_dataset
|