-
Notifications
You must be signed in to change notification settings - Fork 1
/
utils.py
44 lines (40 loc) · 1.43 KB
/
utils.py
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
import transformers
import typing
import copy
import random
import torch
import numpy as np
def set_seed(seed: int):
"""Set the seed for reproducibility."""
random.seed(seed)
torch.manual_seed(seed)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(seed)
np.random.seed(seed)
transformers.set_seed(seed)
class TrainingMetricsCallback(transformers.trainer_callback.TrainerCallback):
"""A callback which logs the training metrics to W&B."""
def __init__(
self,
trainer: transformers.Trainer,
eval_to_train_ratio: float = 2, # TODO: Make this a flag.
) -> None:
super().__init__()
self._trainer = trainer
self._eval_to_train_ratio = eval_to_train_ratio
def on_epoch_end(
self,
args: transformers.training_args.TrainingArguments,
state: transformers.trainer_callback.TrainerState,
control: transformers.trainer_callback.TrainerControl,
**kwargs: typing.Any,
) -> transformers.trainer_callback.TrainerControl:
"""Log the training metrics to W&B."""
control_copy = copy.deepcopy(control)
if state.epoch % self._eval_to_train_ratio == 0:
self._trainer.evaluate(
eval_dataset=self._trainer.train_dataset,
metric_key_prefix="training",
)
# control_copy.should_evaluate = False
return control_copy