Spaces:
Running
Running
Joshua Lochner
commited on
Commit
•
490a61c
1
Parent(s):
e77b67b
Merge duplicated training dataclasses
Browse files- src/preprocess.py +0 -42
- src/shared.py +101 -1
- src/train.py +15 -82
- src/train_classifier.py +44 -152
src/preprocess.py
CHANGED
@@ -490,54 +490,12 @@ def download_file(url, filename):
|
|
490 |
|
491 |
@dataclass
|
492 |
class PreprocessingDatasetArguments(DatasetArguments):
|
493 |
-
|
494 |
-
train_file: Optional[str] = field(
|
495 |
-
default='train.json', metadata={'help': 'The input training data file (a jsonlines file).'}
|
496 |
-
)
|
497 |
-
validation_file: Optional[str] = field(
|
498 |
-
default='valid.json',
|
499 |
-
metadata={
|
500 |
-
'help': 'An optional input evaluation data file to evaluate the metrics on (a jsonlines file).'
|
501 |
-
},
|
502 |
-
)
|
503 |
-
test_file: Optional[str] = field(
|
504 |
-
default='test.json',
|
505 |
-
metadata={
|
506 |
-
'help': 'An optional input test data file to evaluate the metrics on (a jsonlines file).'
|
507 |
-
},
|
508 |
-
)
|
509 |
-
|
510 |
-
c_train_file: Optional[str] = field(
|
511 |
-
default='c_train.json', metadata={'help': 'The input training data file (a jsonlines file).'}
|
512 |
-
)
|
513 |
-
c_validation_file: Optional[str] = field(
|
514 |
-
default='c_valid.json',
|
515 |
-
metadata={
|
516 |
-
'help': 'An optional input evaluation data file to evaluate the metrics on (a jsonlines file).'
|
517 |
-
},
|
518 |
-
)
|
519 |
-
c_test_file: Optional[str] = field(
|
520 |
-
default='c_test.json',
|
521 |
-
metadata={
|
522 |
-
'help': 'An optional input test data file to evaluate the metrics on (a jsonlines file).'
|
523 |
-
},
|
524 |
-
)
|
525 |
-
|
526 |
# excess_file: Optional[str] = field(
|
527 |
# default='excess.json',
|
528 |
# metadata={
|
529 |
# 'help': 'The excess segments left after the split'
|
530 |
# },
|
531 |
# )
|
532 |
-
dataset_cache_dir: Optional[str] = field(
|
533 |
-
default=None,
|
534 |
-
metadata={
|
535 |
-
'help': 'Where to store the cached datasets'
|
536 |
-
},
|
537 |
-
)
|
538 |
-
overwrite_cache: bool = field(
|
539 |
-
default=False, metadata={'help': 'Overwrite the cached training and evaluation sets'}
|
540 |
-
)
|
541 |
|
542 |
positive_file: Optional[str] = field(
|
543 |
default='sponsor_segments.json', metadata={'help': 'File to output sponsored segments to (a jsonlines file).'}
|
|
|
490 |
|
491 |
@dataclass
|
492 |
class PreprocessingDatasetArguments(DatasetArguments):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
493 |
# excess_file: Optional[str] = field(
|
494 |
# default='excess.json',
|
495 |
# metadata={
|
496 |
# 'help': 'The excess segments left after the split'
|
497 |
# },
|
498 |
# )
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
499 |
|
500 |
positive_file: Optional[str] = field(
|
501 |
default='sponsor_segments.json', metadata={'help': 'File to output sponsored segments to (a jsonlines file).'}
|
src/shared.py
CHANGED
@@ -104,6 +104,10 @@ class DatasetArguments:
|
|
104 |
},
|
105 |
)
|
106 |
|
|
|
|
|
|
|
|
|
107 |
dataset_cache_dir: Optional[str] = field(
|
108 |
default=None,
|
109 |
metadata={
|
@@ -111,6 +115,35 @@ class DatasetArguments:
|
|
111 |
},
|
112 |
)
|
113 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
114 |
|
115 |
@dataclass
|
116 |
class OutputArguments:
|
@@ -178,7 +211,7 @@ def reset():
|
|
178 |
print(torch.cuda.memory_summary(device=None, abbreviated=False))
|
179 |
|
180 |
|
181 |
-
def load_datasets(dataset_args):
|
182 |
|
183 |
print('Reading datasets')
|
184 |
data_files = {}
|
@@ -240,6 +273,39 @@ class CustomTrainingArguments(OutputArguments, TrainingArguments):
|
|
240 |
# * :obj:`"steps"`: Evaluation is done (and logged) every :obj:`eval_steps`.
|
241 |
# * :obj:`"epoch"`: Evaluation is done at the end of each epoch.
|
242 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
243 |
|
244 |
logging.basicConfig()
|
245 |
logger = logging.getLogger(__name__)
|
@@ -279,3 +345,37 @@ def train_from_checkpoint(trainer, last_checkpoint, training_args):
|
|
279 |
trainer.save_model() # Saves the tokenizer too for easy upload
|
280 |
|
281 |
return train_result
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
104 |
},
|
105 |
)
|
106 |
|
107 |
+
overwrite_cache: bool = field(
|
108 |
+
default=False, metadata={'help': 'Overwrite the cached training and evaluation sets'}
|
109 |
+
)
|
110 |
+
|
111 |
dataset_cache_dir: Optional[str] = field(
|
112 |
default=None,
|
113 |
metadata={
|
|
|
115 |
},
|
116 |
)
|
117 |
|
118 |
+
train_file: Optional[str] = field(
|
119 |
+
default='train.json', metadata={'help': 'The input training data file (a jsonlines file).'}
|
120 |
+
)
|
121 |
+
validation_file: Optional[str] = field(
|
122 |
+
default='valid.json',
|
123 |
+
metadata={
|
124 |
+
'help': 'An optional input evaluation data file to evaluate the metrics on (a jsonlines file).'
|
125 |
+
},
|
126 |
+
)
|
127 |
+
test_file: Optional[str] = field(
|
128 |
+
default='test.json',
|
129 |
+
metadata={
|
130 |
+
'help': 'An optional input test data file to evaluate the metrics on (a jsonlines file).'
|
131 |
+
},
|
132 |
+
)
|
133 |
+
|
134 |
+
def __post_init__(self):
|
135 |
+
if self.train_file is None or self.validation_file is None:
|
136 |
+
raise ValueError(
|
137 |
+
"Need either a GLUE task, a training/validation file or a dataset name.")
|
138 |
+
else:
|
139 |
+
train_extension = self.train_file.split(".")[-1]
|
140 |
+
assert train_extension in [
|
141 |
+
"csv", "json"], "`train_file` should be a csv or a json file."
|
142 |
+
validation_extension = self.validation_file.split(".")[-1]
|
143 |
+
assert (
|
144 |
+
validation_extension == train_extension
|
145 |
+
), "`validation_file` should have the same extension (csv or json) as `train_file`."
|
146 |
+
|
147 |
|
148 |
@dataclass
|
149 |
class OutputArguments:
|
|
|
211 |
print(torch.cuda.memory_summary(device=None, abbreviated=False))
|
212 |
|
213 |
|
214 |
+
def load_datasets(dataset_args: DatasetArguments):
|
215 |
|
216 |
print('Reading datasets')
|
217 |
data_files = {}
|
|
|
273 |
# * :obj:`"steps"`: Evaluation is done (and logged) every :obj:`eval_steps`.
|
274 |
# * :obj:`"epoch"`: Evaluation is done at the end of each epoch.
|
275 |
|
276 |
+
preprocessing_num_workers: Optional[int] = field(
|
277 |
+
default=None,
|
278 |
+
metadata={'help': 'The number of processes to use for the preprocessing.'},
|
279 |
+
)
|
280 |
+
max_seq_length: int = field(
|
281 |
+
default=512,
|
282 |
+
metadata={
|
283 |
+
"help": "The maximum total input sequence length after tokenization. Sequences longer "
|
284 |
+
"than this will be truncated, sequences shorter will be padded."
|
285 |
+
},
|
286 |
+
)
|
287 |
+
max_train_samples: Optional[int] = field(
|
288 |
+
default=None,
|
289 |
+
metadata={
|
290 |
+
"help": "For debugging purposes or quicker training, truncate the number of training examples to this "
|
291 |
+
"value if set."
|
292 |
+
},
|
293 |
+
)
|
294 |
+
max_eval_samples: Optional[int] = field(
|
295 |
+
default=None,
|
296 |
+
metadata={
|
297 |
+
"help": "For debugging purposes or quicker training, truncate the number of evaluation examples to this "
|
298 |
+
"value if set."
|
299 |
+
},
|
300 |
+
)
|
301 |
+
max_predict_samples: Optional[int] = field(
|
302 |
+
default=None,
|
303 |
+
metadata={
|
304 |
+
"help": "For debugging purposes or quicker training, truncate the number of prediction examples to this "
|
305 |
+
"value if set."
|
306 |
+
},
|
307 |
+
)
|
308 |
+
|
309 |
|
310 |
logging.basicConfig()
|
311 |
logger = logging.getLogger(__name__)
|
|
|
345 |
trainer.save_model() # Saves the tokenizer too for easy upload
|
346 |
|
347 |
return train_result
|
348 |
+
|
349 |
+
|
350 |
+
def prepare_datasets(raw_datasets, dataset_args: DatasetArguments, training_args: CustomTrainingArguments, preprocess_function):
|
351 |
+
|
352 |
+
with training_args.main_process_first(desc="dataset map pre-processing"):
|
353 |
+
raw_datasets = raw_datasets.map(
|
354 |
+
preprocess_function,
|
355 |
+
batched=True,
|
356 |
+
load_from_cache_file=not dataset_args.overwrite_cache,
|
357 |
+
desc="Running tokenizer on dataset",
|
358 |
+
)
|
359 |
+
|
360 |
+
if 'train' not in raw_datasets:
|
361 |
+
raise ValueError('Train dataset missing')
|
362 |
+
train_dataset = raw_datasets['train']
|
363 |
+
if training_args.max_train_samples is not None:
|
364 |
+
train_dataset = train_dataset.select(
|
365 |
+
range(training_args.max_train_samples))
|
366 |
+
|
367 |
+
if 'validation' not in raw_datasets:
|
368 |
+
raise ValueError('Validation dataset missing')
|
369 |
+
eval_dataset = raw_datasets['validation']
|
370 |
+
if training_args.max_eval_samples is not None:
|
371 |
+
eval_dataset = eval_dataset.select(
|
372 |
+
range(training_args.max_eval_samples))
|
373 |
+
|
374 |
+
if 'test' not in raw_datasets:
|
375 |
+
raise ValueError('Test dataset missing')
|
376 |
+
predict_dataset = raw_datasets['test']
|
377 |
+
if training_args.max_predict_samples is not None:
|
378 |
+
predict_dataset = predict_dataset.select(
|
379 |
+
range(training_args.max_predict_samples))
|
380 |
+
|
381 |
+
return train_dataset, eval_dataset, predict_dataset
|
src/train.py
CHANGED
@@ -1,12 +1,17 @@
|
|
1 |
from preprocess import PreprocessingDatasetArguments
|
2 |
-
from shared import
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
from model import ModelArguments
|
4 |
import transformers
|
5 |
import logging
|
6 |
import os
|
7 |
import sys
|
8 |
-
from dataclasses import dataclass, field
|
9 |
-
from typing import Optional
|
10 |
from datasets import utils as d_utils
|
11 |
from transformers import (
|
12 |
DataCollatorForSeq2Seq,
|
@@ -35,38 +40,6 @@ logging.basicConfig(
|
|
35 |
)
|
36 |
|
37 |
|
38 |
-
|
39 |
-
@dataclass
|
40 |
-
class DataTrainingArguments:
|
41 |
-
"""
|
42 |
-
Arguments pertaining to what data we are going to input our model for training and eval.
|
43 |
-
"""
|
44 |
-
|
45 |
-
preprocessing_num_workers: Optional[int] = field(
|
46 |
-
default=None,
|
47 |
-
metadata={'help': 'The number of processes to use for the preprocessing.'},
|
48 |
-
)
|
49 |
-
|
50 |
-
max_train_samples: Optional[int] = field(
|
51 |
-
default=None,
|
52 |
-
metadata={
|
53 |
-
'help': 'For debugging purposes or quicker training, truncate the number of training examples to this value if set.'
|
54 |
-
},
|
55 |
-
)
|
56 |
-
max_eval_samples: Optional[int] = field(
|
57 |
-
default=None,
|
58 |
-
metadata={
|
59 |
-
'help': 'For debugging purposes or quicker training, truncate the number of evaluation examples to this value if set.'
|
60 |
-
},
|
61 |
-
)
|
62 |
-
max_predict_samples: Optional[int] = field(
|
63 |
-
default=None,
|
64 |
-
metadata={
|
65 |
-
'help': 'For debugging purposes or quicker training, truncate the number of prediction examples to this value if set.'
|
66 |
-
},
|
67 |
-
)
|
68 |
-
|
69 |
-
|
70 |
def main():
|
71 |
|
72 |
# See all possible arguments in src/transformers/training_args.py
|
@@ -76,10 +49,9 @@ def main():
|
|
76 |
hf_parser = HfArgumentParser((
|
77 |
ModelArguments,
|
78 |
PreprocessingDatasetArguments,
|
79 |
-
DataTrainingArguments,
|
80 |
CustomTrainingArguments
|
81 |
))
|
82 |
-
model_args, dataset_args,
|
83 |
|
84 |
log_level = training_args.get_process_log_level()
|
85 |
logger.setLevel(log_level)
|
@@ -128,7 +100,6 @@ def main():
|
|
128 |
# See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
|
129 |
# https://huggingface.co/docs/datasets/loading_datasets.html.
|
130 |
|
131 |
-
|
132 |
# Detecting last checkpoint.
|
133 |
last_checkpoint = get_last_checkpoint(training_args)
|
134 |
|
@@ -165,47 +136,8 @@ def main():
|
|
165 |
|
166 |
return model_inputs
|
167 |
|
168 |
-
|
169 |
-
|
170 |
-
preprocess_function,
|
171 |
-
batched=True,
|
172 |
-
num_proc=data_training_args.preprocessing_num_workers,
|
173 |
-
remove_columns=column_names,
|
174 |
-
load_from_cache_file=not dataset_args.overwrite_cache,
|
175 |
-
desc=desc, # tokenizing train dataset
|
176 |
-
)
|
177 |
-
# train_dataset # TODO shuffle?
|
178 |
-
|
179 |
-
# if training_args.do_train:
|
180 |
-
if 'train' not in raw_datasets: # TODO do checks above?
|
181 |
-
raise ValueError('Train dataset missing')
|
182 |
-
train_dataset = raw_datasets['train']
|
183 |
-
if data_training_args.max_train_samples is not None:
|
184 |
-
train_dataset = train_dataset.select(
|
185 |
-
range(data_training_args.max_train_samples))
|
186 |
-
with training_args.main_process_first(desc='train dataset map pre-processing'):
|
187 |
-
train_dataset = prepare_dataset(
|
188 |
-
train_dataset, desc='Running tokenizer on train dataset')
|
189 |
-
|
190 |
-
if 'validation' not in raw_datasets:
|
191 |
-
raise ValueError('Validation dataset missing')
|
192 |
-
eval_dataset = raw_datasets['validation']
|
193 |
-
if data_training_args.max_eval_samples is not None:
|
194 |
-
eval_dataset = eval_dataset.select(
|
195 |
-
range(data_training_args.max_eval_samples))
|
196 |
-
with training_args.main_process_first(desc='validation dataset map pre-processing'):
|
197 |
-
eval_dataset = prepare_dataset(
|
198 |
-
eval_dataset, desc='Running tokenizer on validation dataset')
|
199 |
-
|
200 |
-
if 'test' not in raw_datasets:
|
201 |
-
raise ValueError('Test dataset missing')
|
202 |
-
predict_dataset = raw_datasets['test']
|
203 |
-
if data_training_args.max_predict_samples is not None:
|
204 |
-
predict_dataset = predict_dataset.select(
|
205 |
-
range(data_training_args.max_predict_samples))
|
206 |
-
with training_args.main_process_first(desc='prediction dataset map pre-processing'):
|
207 |
-
predict_dataset = prepare_dataset(
|
208 |
-
predict_dataset, desc='Running tokenizer on prediction dataset')
|
209 |
|
210 |
# Data collator
|
211 |
data_collator = DataCollatorForSeq2Seq(
|
@@ -228,10 +160,11 @@ def main():
|
|
228 |
)
|
229 |
|
230 |
# Training
|
231 |
-
train_result = train_from_checkpoint(
|
|
|
232 |
|
233 |
metrics = train_result.metrics
|
234 |
-
max_train_samples =
|
235 |
train_dataset)
|
236 |
metrics['train_samples'] = min(max_train_samples, len(train_dataset))
|
237 |
|
@@ -240,7 +173,7 @@ def main():
|
|
240 |
trainer.save_state()
|
241 |
|
242 |
kwargs = {'finetuned_from': model_args.model_name_or_path,
|
243 |
-
|
244 |
|
245 |
if training_args.push_to_hub:
|
246 |
trainer.push_to_hub(**kwargs)
|
|
|
1 |
from preprocess import PreprocessingDatasetArguments
|
2 |
+
from shared import (
|
3 |
+
CustomTokens,
|
4 |
+
prepare_datasets,
|
5 |
+
load_datasets,
|
6 |
+
CustomTrainingArguments,
|
7 |
+
get_last_checkpoint,
|
8 |
+
train_from_checkpoint
|
9 |
+
)
|
10 |
from model import ModelArguments
|
11 |
import transformers
|
12 |
import logging
|
13 |
import os
|
14 |
import sys
|
|
|
|
|
15 |
from datasets import utils as d_utils
|
16 |
from transformers import (
|
17 |
DataCollatorForSeq2Seq,
|
|
|
40 |
)
|
41 |
|
42 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
43 |
def main():
|
44 |
|
45 |
# See all possible arguments in src/transformers/training_args.py
|
|
|
49 |
hf_parser = HfArgumentParser((
|
50 |
ModelArguments,
|
51 |
PreprocessingDatasetArguments,
|
|
|
52 |
CustomTrainingArguments
|
53 |
))
|
54 |
+
model_args, dataset_args, training_args = hf_parser.parse_args_into_dataclasses()
|
55 |
|
56 |
log_level = training_args.get_process_log_level()
|
57 |
logger.setLevel(log_level)
|
|
|
100 |
# See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
|
101 |
# https://huggingface.co/docs/datasets/loading_datasets.html.
|
102 |
|
|
|
103 |
# Detecting last checkpoint.
|
104 |
last_checkpoint = get_last_checkpoint(training_args)
|
105 |
|
|
|
136 |
|
137 |
return model_inputs
|
138 |
|
139 |
+
train_dataset, eval_dataset, predict_dataset = prepare_datasets(
|
140 |
+
raw_datasets, dataset_args, training_args, preprocess_function)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
141 |
|
142 |
# Data collator
|
143 |
data_collator = DataCollatorForSeq2Seq(
|
|
|
160 |
)
|
161 |
|
162 |
# Training
|
163 |
+
train_result = train_from_checkpoint(
|
164 |
+
trainer, last_checkpoint, training_args)
|
165 |
|
166 |
metrics = train_result.metrics
|
167 |
+
max_train_samples = training_args.max_train_samples or len(
|
168 |
train_dataset)
|
169 |
metrics['train_samples'] = min(max_train_samples, len(train_dataset))
|
170 |
|
|
|
173 |
trainer.save_state()
|
174 |
|
175 |
kwargs = {'finetuned_from': model_args.model_name_or_path,
|
176 |
+
'tasks': 'summarization'}
|
177 |
|
178 |
if training_args.push_to_hub:
|
179 |
trainer.push_to_hub(**kwargs)
|
src/train_classifier.py
CHANGED
@@ -3,14 +3,12 @@
|
|
3 |
|
4 |
import logging
|
5 |
import os
|
6 |
-
import random
|
7 |
import sys
|
8 |
from dataclasses import dataclass, field
|
9 |
from typing import Optional
|
10 |
|
11 |
import datasets
|
12 |
import numpy as np
|
13 |
-
from datasets import load_metric
|
14 |
|
15 |
import transformers
|
16 |
from transformers import (
|
@@ -18,96 +16,39 @@ from transformers import (
|
|
18 |
EvalPrediction,
|
19 |
HfArgumentParser,
|
20 |
Trainer,
|
21 |
-
default_data_collator,
|
22 |
set_seed,
|
23 |
)
|
24 |
from transformers.utils import check_min_version
|
25 |
from transformers.utils.versions import require_version
|
26 |
-
from shared import CATEGORIES, load_datasets, CustomTrainingArguments, train_from_checkpoint, get_last_checkpoint
|
27 |
-
from preprocess import PreprocessingDatasetArguments
|
28 |
from model import get_model_tokenizer, ModelArguments
|
29 |
|
30 |
# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
|
31 |
-
check_min_version(
|
32 |
-
require_version(
|
33 |
|
34 |
-
os.environ[
|
35 |
|
36 |
logger = logging.getLogger(__name__)
|
37 |
|
38 |
|
39 |
@dataclass
|
40 |
-
class
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
Using `HfArgumentParser` we can turn this class
|
45 |
-
into argparse arguments to be able to specify them on
|
46 |
-
the command line.
|
47 |
-
"""
|
48 |
-
|
49 |
-
max_seq_length: int = field(
|
50 |
-
default=512,
|
51 |
-
metadata={
|
52 |
-
"help": "The maximum total input sequence length after tokenization. Sequences longer "
|
53 |
-
"than this will be truncated, sequences shorter will be padded."
|
54 |
-
},
|
55 |
-
)
|
56 |
-
overwrite_cache: bool = field(
|
57 |
-
default=False, metadata={"help": "Overwrite the cached preprocessed datasets or not."}
|
58 |
-
)
|
59 |
-
pad_to_max_length: bool = field(
|
60 |
-
default=True,
|
61 |
-
metadata={
|
62 |
-
"help": "Whether to pad all samples to `max_seq_length`. "
|
63 |
-
"If False, will pad the samples dynamically when batching to the maximum length in the batch."
|
64 |
-
},
|
65 |
)
|
66 |
-
|
67 |
-
default=
|
68 |
metadata={
|
69 |
-
|
70 |
-
"value if set."
|
71 |
},
|
72 |
)
|
73 |
-
|
74 |
-
default=
|
75 |
metadata={
|
76 |
-
|
77 |
-
"value if set."
|
78 |
},
|
79 |
)
|
80 |
-
max_predict_samples: Optional[int] = field(
|
81 |
-
default=None,
|
82 |
-
metadata={
|
83 |
-
"help": "For debugging purposes or quicker training, truncate the number of prediction examples to this "
|
84 |
-
"value if set."
|
85 |
-
},
|
86 |
-
)
|
87 |
-
|
88 |
-
dataset_cache_dir: Optional[str] = PreprocessingDatasetArguments.__dataclass_fields__[
|
89 |
-
'dataset_cache_dir']
|
90 |
-
data_dir: Optional[str] = PreprocessingDatasetArguments.__dataclass_fields__[
|
91 |
-
'data_dir']
|
92 |
-
train_file: Optional[str] = PreprocessingDatasetArguments.__dataclass_fields__[
|
93 |
-
'c_train_file']
|
94 |
-
validation_file: Optional[str] = PreprocessingDatasetArguments.__dataclass_fields__[
|
95 |
-
'c_validation_file']
|
96 |
-
test_file: Optional[str] = PreprocessingDatasetArguments.__dataclass_fields__[
|
97 |
-
'c_test_file']
|
98 |
-
|
99 |
-
def __post_init__(self):
|
100 |
-
if self.train_file is None or self.validation_file is None:
|
101 |
-
raise ValueError(
|
102 |
-
"Need either a GLUE task, a training/validation file or a dataset name.")
|
103 |
-
else:
|
104 |
-
train_extension = self.train_file.split(".")[-1]
|
105 |
-
assert train_extension in [
|
106 |
-
"csv", "json"], "`train_file` should be a csv or a json file."
|
107 |
-
validation_extension = self.validation_file.split(".")[-1]
|
108 |
-
assert (
|
109 |
-
validation_extension == train_extension
|
110 |
-
), "`validation_file` should have the same extension (csv or json) as `train_file`."
|
111 |
|
112 |
|
113 |
def main():
|
@@ -115,14 +56,17 @@ def main():
|
|
115 |
# or by passing the --help flag to this script.
|
116 |
# We now keep distinct sets of args, for a cleaner separation of concerns.
|
117 |
|
118 |
-
|
119 |
-
|
120 |
-
|
|
|
|
|
|
|
121 |
|
122 |
# Setup logging
|
123 |
logging.basicConfig(
|
124 |
-
format=
|
125 |
-
datefmt=
|
126 |
handlers=[logging.StreamHandler(sys.stdout)],
|
127 |
)
|
128 |
|
@@ -135,10 +79,10 @@ def main():
|
|
135 |
|
136 |
# Log on each process the small summary:
|
137 |
logger.warning(
|
138 |
-
f
|
139 |
-
+ f
|
140 |
)
|
141 |
-
logger.info(f
|
142 |
|
143 |
# Detecting last checkpoint.
|
144 |
last_checkpoint = get_last_checkpoint(training_args)
|
@@ -148,7 +92,7 @@ def main():
|
|
148 |
|
149 |
# Loading a dataset from your local files.
|
150 |
# CSV/JSON training and evaluation files are needed.
|
151 |
-
raw_datasets = load_datasets(
|
152 |
|
153 |
# See more about loading any type of standard or custom dataset at
|
154 |
# https://huggingface.co/docs/datasets/loading_datasets.html.
|
@@ -158,69 +102,26 @@ def main():
|
|
158 |
'id2label': {k: str(v).upper() for k, v in enumerate(CATEGORIES)},
|
159 |
'label2id': {str(v).upper(): k for k, v in enumerate(CATEGORIES)}
|
160 |
}
|
161 |
-
model, tokenizer = get_model_tokenizer(
|
|
|
162 |
|
163 |
-
|
164 |
-
# Padding strategy
|
165 |
-
if data_args.pad_to_max_length:
|
166 |
-
padding = "max_length"
|
167 |
-
else:
|
168 |
-
# We will pad later, dynamically at batch creation, to the max sequence length in each batch
|
169 |
-
padding = False
|
170 |
-
|
171 |
-
if data_args.max_seq_length > tokenizer.model_max_length:
|
172 |
logger.warning(
|
173 |
-
f
|
174 |
-
f
|
175 |
)
|
176 |
-
max_seq_length = min(
|
|
|
177 |
|
178 |
def preprocess_function(examples):
|
179 |
# Tokenize the texts
|
180 |
result = tokenizer(
|
181 |
-
examples['text'], padding=
|
182 |
result['label'] = examples['label']
|
183 |
return result
|
184 |
|
185 |
-
|
186 |
-
raw_datasets
|
187 |
-
preprocess_function,
|
188 |
-
batched=True,
|
189 |
-
load_from_cache_file=not data_args.overwrite_cache,
|
190 |
-
desc="Running tokenizer on dataset",
|
191 |
-
)
|
192 |
-
if training_args.do_train:
|
193 |
-
if "train" not in raw_datasets:
|
194 |
-
raise ValueError("--do_train requires a train dataset")
|
195 |
-
train_dataset = raw_datasets["train"]
|
196 |
-
if data_args.max_train_samples is not None:
|
197 |
-
train_dataset = train_dataset.select(
|
198 |
-
range(data_args.max_train_samples))
|
199 |
-
|
200 |
-
if training_args.do_eval:
|
201 |
-
if "validation" not in raw_datasets:
|
202 |
-
raise ValueError("--do_eval requires a validation dataset")
|
203 |
-
eval_dataset = raw_datasets["validation"]
|
204 |
-
if data_args.max_eval_samples is not None:
|
205 |
-
eval_dataset = eval_dataset.select(
|
206 |
-
range(data_args.max_eval_samples))
|
207 |
-
|
208 |
-
if training_args.do_predict or data_args.test_file is not None:
|
209 |
-
if "test" not in raw_datasets:
|
210 |
-
raise ValueError("--do_predict requires a test dataset")
|
211 |
-
predict_dataset = raw_datasets["test"]
|
212 |
-
if data_args.max_predict_samples is not None:
|
213 |
-
predict_dataset = predict_dataset.select(
|
214 |
-
range(data_args.max_predict_samples))
|
215 |
-
|
216 |
-
# Log a few random samples from the training set:
|
217 |
-
if training_args.do_train:
|
218 |
-
for index in random.sample(range(len(train_dataset)), 3):
|
219 |
-
logger.info(
|
220 |
-
f"Sample {index} of the training set: {train_dataset[index]}.")
|
221 |
-
|
222 |
-
# Get the metric function
|
223 |
-
metric = load_metric("accuracy")
|
224 |
|
225 |
# You can define your custom compute_metrics function. It takes an `EvalPrediction` object (a namedtuple with a
|
226 |
# predictions and label_ids field) and has to return a dictionary string to float.
|
@@ -228,20 +129,11 @@ def main():
|
|
228 |
preds = p.predictions[0] if isinstance(
|
229 |
p.predictions, tuple) else p.predictions
|
230 |
preds = np.argmax(preds, axis=1)
|
231 |
-
|
232 |
-
result = metric.compute(predictions=preds, references=p.label_ids)
|
233 |
-
if len(result) > 1:
|
234 |
-
result["combined_score"] = np.mean(
|
235 |
-
list(result.values())).item()
|
236 |
-
return result
|
237 |
-
else:
|
238 |
-
return {"accuracy": (preds == p.label_ids).astype(np.float32).mean().item()}
|
239 |
|
240 |
# Data collator will default to DataCollatorWithPadding when the tokenizer is passed to Trainer, so we change it if
|
241 |
# we already did the padding.
|
242 |
-
if
|
243 |
-
data_collator = default_data_collator
|
244 |
-
elif training_args.fp16:
|
245 |
data_collator = DataCollatorWithPadding(
|
246 |
tokenizer, pad_to_multiple_of=8)
|
247 |
else:
|
@@ -264,24 +156,24 @@ def main():
|
|
264 |
|
265 |
metrics = train_result.metrics
|
266 |
max_train_samples = (
|
267 |
-
|
268 |
train_dataset)
|
269 |
)
|
270 |
-
metrics[
|
271 |
|
272 |
trainer.save_model() # Saves the tokenizer too for easy upload
|
273 |
|
274 |
-
trainer.log_metrics(
|
275 |
-
trainer.save_metrics(
|
276 |
trainer.save_state()
|
277 |
|
278 |
-
kwargs = {
|
279 |
-
|
280 |
if training_args.push_to_hub:
|
281 |
trainer.push_to_hub(**kwargs)
|
282 |
else:
|
283 |
trainer.create_model_card(**kwargs)
|
284 |
|
285 |
|
286 |
-
if __name__ ==
|
287 |
main()
|
|
|
3 |
|
4 |
import logging
|
5 |
import os
|
|
|
6 |
import sys
|
7 |
from dataclasses import dataclass, field
|
8 |
from typing import Optional
|
9 |
|
10 |
import datasets
|
11 |
import numpy as np
|
|
|
12 |
|
13 |
import transformers
|
14 |
from transformers import (
|
|
|
16 |
EvalPrediction,
|
17 |
HfArgumentParser,
|
18 |
Trainer,
|
|
|
19 |
set_seed,
|
20 |
)
|
21 |
from transformers.utils import check_min_version
|
22 |
from transformers.utils.versions import require_version
|
23 |
+
from shared import CATEGORIES, DatasetArguments, prepare_datasets, load_datasets, CustomTrainingArguments, train_from_checkpoint, get_last_checkpoint
|
|
|
24 |
from model import get_model_tokenizer, ModelArguments
|
25 |
|
26 |
# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
|
27 |
+
check_min_version('4.17.0')
|
28 |
+
require_version('datasets>=1.8.0', 'To fix: pip install -r requirements.txt')
|
29 |
|
30 |
+
os.environ['WANDB_DISABLED'] = 'true'
|
31 |
|
32 |
logger = logging.getLogger(__name__)
|
33 |
|
34 |
|
35 |
@dataclass
|
36 |
+
class ClassifierDatasetArguments(DatasetArguments):
|
37 |
+
train_file: Optional[str] = field(
|
38 |
+
default='c_train.json', metadata={'help': 'The input training data file (a jsonlines file).'}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
)
|
40 |
+
validation_file: Optional[str] = field(
|
41 |
+
default='c_valid.json',
|
42 |
metadata={
|
43 |
+
'help': 'An optional input evaluation data file to evaluate the metrics on (a jsonlines file).'
|
|
|
44 |
},
|
45 |
)
|
46 |
+
test_file: Optional[str] = field(
|
47 |
+
default='c_test.json',
|
48 |
metadata={
|
49 |
+
'help': 'An optional input test data file to evaluate the metrics on (a jsonlines file).'
|
|
|
50 |
},
|
51 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
52 |
|
53 |
|
54 |
def main():
|
|
|
56 |
# or by passing the --help flag to this script.
|
57 |
# We now keep distinct sets of args, for a cleaner separation of concerns.
|
58 |
|
59 |
+
hf_parser = HfArgumentParser((
|
60 |
+
ModelArguments,
|
61 |
+
ClassifierDatasetArguments,
|
62 |
+
CustomTrainingArguments
|
63 |
+
))
|
64 |
+
model_args, dataset_args, training_args = hf_parser.parse_args_into_dataclasses()
|
65 |
|
66 |
# Setup logging
|
67 |
logging.basicConfig(
|
68 |
+
format='%(asctime)s - %(levelname)s - %(name)s - %(message)s',
|
69 |
+
datefmt='%m/%d/%Y %H:%M:%S',
|
70 |
handlers=[logging.StreamHandler(sys.stdout)],
|
71 |
)
|
72 |
|
|
|
79 |
|
80 |
# Log on each process the small summary:
|
81 |
logger.warning(
|
82 |
+
f'Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}'
|
83 |
+
+ f'distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}'
|
84 |
)
|
85 |
+
logger.info(f'Training/evaluation parameters {training_args}')
|
86 |
|
87 |
# Detecting last checkpoint.
|
88 |
last_checkpoint = get_last_checkpoint(training_args)
|
|
|
92 |
|
93 |
# Loading a dataset from your local files.
|
94 |
# CSV/JSON training and evaluation files are needed.
|
95 |
+
raw_datasets = load_datasets(dataset_args)
|
96 |
|
97 |
# See more about loading any type of standard or custom dataset at
|
98 |
# https://huggingface.co/docs/datasets/loading_datasets.html.
|
|
|
102 |
'id2label': {k: str(v).upper() for k, v in enumerate(CATEGORIES)},
|
103 |
'label2id': {str(v).upper(): k for k, v in enumerate(CATEGORIES)}
|
104 |
}
|
105 |
+
model, tokenizer = get_model_tokenizer(
|
106 |
+
model_args, training_args, config_args=config_args, model_type='classifier')
|
107 |
|
108 |
+
if training_args.max_seq_length > tokenizer.model_max_length:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
109 |
logger.warning(
|
110 |
+
f'The max_seq_length passed ({training_args.max_seq_length}) is larger than the maximum length for the'
|
111 |
+
f'model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}.'
|
112 |
)
|
113 |
+
max_seq_length = min(training_args.max_seq_length,
|
114 |
+
tokenizer.model_max_length)
|
115 |
|
116 |
def preprocess_function(examples):
|
117 |
# Tokenize the texts
|
118 |
result = tokenizer(
|
119 |
+
examples['text'], padding='max_length', max_length=max_seq_length, truncation=True)
|
120 |
result['label'] = examples['label']
|
121 |
return result
|
122 |
|
123 |
+
train_dataset, eval_dataset, predict_dataset = prepare_datasets(
|
124 |
+
raw_datasets, dataset_args, training_args, preprocess_function)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
125 |
|
126 |
# You can define your custom compute_metrics function. It takes an `EvalPrediction` object (a namedtuple with a
|
127 |
# predictions and label_ids field) and has to return a dictionary string to float.
|
|
|
129 |
preds = p.predictions[0] if isinstance(
|
130 |
p.predictions, tuple) else p.predictions
|
131 |
preds = np.argmax(preds, axis=1)
|
132 |
+
return {'accuracy': (preds == p.label_ids).astype(np.float32).mean().item()}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
133 |
|
134 |
# Data collator will default to DataCollatorWithPadding when the tokenizer is passed to Trainer, so we change it if
|
135 |
# we already did the padding.
|
136 |
+
if training_args.fp16:
|
|
|
|
|
137 |
data_collator = DataCollatorWithPadding(
|
138 |
tokenizer, pad_to_multiple_of=8)
|
139 |
else:
|
|
|
156 |
|
157 |
metrics = train_result.metrics
|
158 |
max_train_samples = (
|
159 |
+
training_args.max_train_samples if training_args.max_train_samples is not None else len(
|
160 |
train_dataset)
|
161 |
)
|
162 |
+
metrics['train_samples'] = min(max_train_samples, len(train_dataset))
|
163 |
|
164 |
trainer.save_model() # Saves the tokenizer too for easy upload
|
165 |
|
166 |
+
trainer.log_metrics('train', metrics)
|
167 |
+
trainer.save_metrics('train', metrics)
|
168 |
trainer.save_state()
|
169 |
|
170 |
+
kwargs = {'finetuned_from': model_args.model_name_or_path,
|
171 |
+
'tasks': 'text-classification'}
|
172 |
if training_args.push_to_hub:
|
173 |
trainer.push_to_hub(**kwargs)
|
174 |
else:
|
175 |
trainer.create_model_card(**kwargs)
|
176 |
|
177 |
|
178 |
+
if __name__ == '__main__':
|
179 |
main()
|