Commit 925a6c9f authored by Jan Kocoń's avatar Jan Kocoń

Megatron-LM adjusted to Polish texts

parent 78e7e33c
# Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of NVIDIA CORPORATION nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
------------- LICENSE FOR huggingface(transformer) repository --------------
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
This diff is collapsed.
{"url": "http://jankocon.clarin-pl.eu/share/gpt2-pl-merges-new.txt", "etag": "\"5e665606-9c30f\""}
\ No newline at end of file
{"url": "http://jankocon.clarin-pl.eu/share/gpt2-pl-vocab-new.json", "etag": "\"5e665606-c3dfc\""}
\ No newline at end of file
{"url": "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-merges.txt", "etag": "\"75a37753dd7a28a2c5df80c28bf06e4e\""}
\ No newline at end of file
{"url": "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-vocab.json", "etag": "\"dffec25a898b1f5e569bec4dffd7e5c0\""}
\ No newline at end of file
{"url": "http://jankocon.clarin-pl.eu/share/gpt2-pl-merges-new.txt", "etag": "\"5e665606-9c30f\""}
\ No newline at end of file
{"url": "http://jankocon.clarin-pl.eu/share/gpt2-pl-vocab-new.json", "etag": "\"5e665606-c3dfc\""}
\ No newline at end of file
# coding=utf-8
# Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""parses arguments and preps data loader"""
import copy
import torch
import data_utils
import mpu
class DataConfig:
def __init__(self, defaults={}):
super(DataConfig, self).__init__()
self.defaults = defaults
def apply(self, args):
if torch.distributed.get_rank() == 0:
print('configuring data')
self.apply_defaults(args)
return make_loaders(args)
def set_defaults(self, **kwargs):
for k, v in kwargs.items():
self.defaults[k] = v
def apply_defaults(self, args):
for k, v in self.defaults.items():
k = k.replace('-', '_')
if not hasattr(args, k):
setattr(args, k, v)
def make_data_loader(dataset, batch_size, args):
shuffle = args.shuffle
if shuffle:
sampler = data_utils.samplers.RandomSampler(dataset, replacement=True, num_samples=batch_size*args.train_iters)
else:
sampler = torch.utils.data.SequentialSampler(dataset)
world_size = torch.distributed.get_world_size(
group=mpu.get_data_parallel_group())
rank = torch.distributed.get_rank(group=mpu.get_data_parallel_group())
distributed = world_size > 1
drop_last = distributed
if distributed:
batch_sampler = data_utils.samplers.DistributedBatchSampler(sampler,
batch_size,
drop_last,
rank,
world_size)
else:
batch_sampler = torch.utils.data.BatchSampler(sampler,
batch_size,
drop_last)
data_loader = torch.utils.data.DataLoader(dataset,
batch_sampler=batch_sampler,
num_workers=args.num_workers,
pin_memory=True)
return data_loader
def make_tfrecord_loaders(args):
"""Load train/val/test dataset from shuffled TFRecords"""
import data_utils.tf_dl
data_set_args = {'batch_size': args.batch_size,
'max_seq_len': args.seq_length,
'max_preds_per_seq': args.max_preds_per_seq,
'train': True,
'num_workers': max(args.num_workers, 1),
'seed': args.seed + args.rank + 1,
'threaded_dl': args.num_workers > 0
}
train = data_utils.tf_dl.TFRecordDataLoader(args.train_data,
**data_set_args)
data_set_args['train'] = False
if args.eval_seq_length is not None:
data_set_args['max_seq_len'] = args.eval_seq_length
if args.eval_max_preds_per_seq is not None:
data_set_args['max_preds_per_seq'] = args.eval_max_preds_per_seq
valid = None
if args.valid_data is not None:
valid = data_utils.tf_dl.TFRecordDataLoader(args.valid_data,
**data_set_args)
test = None
if args.test_data is not None:
test = data_utils.tf_dl.TFRecordDataLoader(args.test_data,
**data_set_args)
tokenizer = data_utils.make_tokenizer(args.tokenizer_type,
train,
args.tokenizer_path,
args.vocab_size,
args.tokenizer_model_type,
cache_dir=args.cache_dir)
return (train, valid, test), tokenizer
def make_loaders(args):
"""makes training/val/test"""
if args.use_tfrecords:
return make_tfrecord_loaders(args)
world_size = torch.distributed.get_world_size(
group=mpu.get_data_parallel_group())
batch_size = args.batch_size * world_size
eval_batch_size = batch_size
if args.eval_batch_size is not None:
eval_batch_size = args.eval_batch_size * world_size
seq_length = args.seq_length
if seq_length < 0:
seq_length = seq_length * world_size
eval_seq_length = args.eval_seq_length
if eval_seq_length is not None and eval_seq_length < 0:
eval_seq_length = eval_seq_length * world_size
split = get_split(args)
data_set_args = {
'path': args.train_data,
'seq_length': seq_length,
'lazy': args.lazy_loader,
'delim': args.delim,
'text_key': args.text_key,
'label_key': 'label',
'non_binary_cols': None,
'ds_type': args.data_set_type,
'split': split,
'loose': args.loose_json,
'tokenizer_type': args.tokenizer_type,
'tokenizer_model_path': args.tokenizer_path,
'vocab_size': args.vocab_size,
'model_type': args.tokenizer_model_type,
'cache_dir': args.cache_dir,
'max_preds_per_seq': args.max_preds_per_seq,
'presplit_sentences': args.presplit_sentences,
'parallel_group': mpu.get_data_parallel_group()}
eval_set_args = copy.copy(data_set_args)
eval_set_args['split'] = [1.]
# if optional eval args were set then replace their
# equivalent values in the arg dict
if eval_seq_length:
eval_set_args['seq_length'] = eval_seq_length
if args.eval_max_preds_per_seq:
eval_set_args['max_preds_per_seq'] = args.eval_max_preds_per_seq
if args.eval_text_key is not None:
eval_set_args['text_key'] = args.eval_text_key
# make datasets splits and tokenizer
train = None
valid = None
test = None
if args.train_data is not None:
train, tokenizer = data_utils.make_dataset(**data_set_args)
if data_utils.should_split(split):
train, valid, test = train
eval_set_args['tokenizer'] = tokenizer
# make training and val dataset if necessary
if valid is None and args.valid_data is not None:
eval_set_args['path'] = args.valid_data
valid, tokenizer = data_utils.make_dataset(**eval_set_args)
eval_set_args['tokenizer'] = tokenizer
if test is None and args.test_data is not None:
eval_set_args['path'] = args.test_data
test, tokenizer = data_utils.make_dataset(**eval_set_args)
# wrap datasets with data loader
if train is not None and args.batch_size > 0:
train = make_data_loader(train, batch_size, args)
args.do_train = True
else:
args.do_train = False
eval_batch_size = eval_batch_size if eval_batch_size != 0 else batch_size
if valid is not None:
valid = make_data_loader(valid, eval_batch_size, args)
args.do_valid = True
else:
args.do_valid = False
if test is not None:
test = make_data_loader(test, eval_batch_size, args)
args.do_test = True
else:
args.do_test = False
return (train, valid, test), tokenizer
def get_split(args):
"""
Get dataset splits from comma separated string list
"""
splits = []
if args.split.find(',') != -1:
splits = [float(s) for s in args.split.split(',')]
elif args.split.find('/') != -1:
splits = [float(s) for s in args.split.split('/')]
else:
splits = [float(args.split)]
split_total = sum(splits)
if split_total < 1.:
splits.append(1-split_total)
while len(splits) < 3:
splits.append(0.)
splits = splits[:3]
if args.valid_data is not None:
splits[1] = 0.
if args.test_data is not None:
splits[2] = 0.
final_sum = sum(splits)
return [s/final_sum for s in splits]
def configure_data():
"""add cmdline flags for configuring datasets"""
# These are options that are used by data_utils, but are either
# deprecated or not meant to be exposed to the command line user.
# These options are intneded to be set in code by specific scripts.
defaults = {
'world_size': 1,
'rank': -1,
'persist_state': 0,
'lazy': False,
'transpose': False,
'data_set_type': 'supervised',
'seq_length': 256,
'eval_seq_length': 256,
'samples_per_shard': 100
}
return DataConfig(defaults=defaults)
# coding=utf-8
# Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""utils for creating datasets"""
import os
import math
import torch
from .samplers import DistributedBatchSampler
from .datasets import json_dataset, csv_dataset, split_ds, ConcatDataset, SplitDataset, bert_sentencepair_dataset, GPT2Dataset
from .lazy_loader import exists_lazy, make_lazy, lazy_array_loader
from .tokenization import Tokenization, CommandToken, Tokenizer, CharacterLevelTokenizer, BertWordPieceTokenizer, GPT2BPETokenizer, make_tokenizer
from . import corpora
TRAIN_DATA = 0
VAL_DATA = 1
TEST_DATA = 2
def should_split(split):
"""
given split proportions checks if should split
Examples:
>>> should_split([10,0,0])
False
>>> should_split([1,.1,.2])
True
"""
return max(split)/sum(split) != 1.
def get_ext(path):
"""gets path extension"""
return os.path.splitext(path)[1]
def get_dataset(path, **kwargs):
"""gets dataset object based on keyword args and file at `path`"""
if supported_corpus(path):
return corpora.NAMED_CORPORA[path](**kwargs)
ext = get_ext(path)
if '.json' in ext:
text = json_dataset(path, **kwargs)
elif ext in ['.csv', '.tsv']:
text = csv_dataset(path, **kwargs)
else:
raise NotImplementedError('data file type %s is not supported'%(ext))
return text
def supported_corpus(corpus_name):
"""checks if corpus name is defined in `corpora.py`"""
return corpus_name in corpora.NAMED_CORPORA
def make_dataset(path, seq_length, text_key, label_key, lazy=False, process_fn=None, split=[1.],
delim=',', loose=False, binarize_sent=False, drop_unlabeled=False, tokenizer=None,
tokenizer_type='CharacterLevelTokenizer', tokenizer_model_path=None, vocab_size=None,
model_type='bpe', pad_token=0, character_converage=1.0, non_binary_cols=None,
parallel_group=None, **kwargs):
"""function to create datasets+tokenizers for common options"""
if isinstance(process_fn, str):
process_fn = eval(process_fn)
if non_binary_cols is not None:
# multilabel dataset support (only for csvs)
label_key = non_binary_cols
def get_dataset_from_path(path_):
if lazy:
# get lazily loaded dataset
named_corpora = False
if supported_corpus(path_):
named_corpora = True
name