fix(templates): merge handlers.py into utils.py

templates/gan/test_all.py

@@ -8,7 +8,6 @@
 import torch
 from config import get_default_parser
 from datasets import get_datasets
-from handlers import get_handlers, get_logger
 from ignite.contrib.handlers import (
     ClearMLLogger,
     MLflowLogger,
@@ -28,7 +27,7 @@
 from torch import nn, optim
 from torch.utils.data import Dataset
 from trainers import create_trainers, train_function
-from utils import hash_checkpoint, log_metrics, resume_from, setup_logging
+from utils import hash_checkpoint, log_metrics, resume_from, setup_logging, get_handlers, get_logger
 
 
 class TestDataset(unittest.TestCase):

templates/gan/utils.py

@@ -21,6 +21,8 @@
 from torch.optim.optimizer import Optimizer
 
 from models import Generator, Discriminator
+{% include "_handlers.py" %}
+
 
 # we can use `idist.auto_model` to handle distributed configurations
 # for your model : https://pytorch.org/ignite/distributed.html#ignite.distributed.auto.auto_model

templates/image_classification/test_all.py

@@ -9,7 +9,6 @@
 import torch
 from config import get_default_parser
 from datasets import get_datasets
-from handlers import get_handlers, get_logger
 from ignite.contrib.handlers import (
     ClearMLLogger,
     MLflowLogger,
@@ -36,7 +35,7 @@
     train_events_to_attr,
     train_function,
 )
-from utils import hash_checkpoint, initialize, log_metrics, resume_from, setup_logging
+from utils import hash_checkpoint, initialize, log_metrics, resume_from, setup_logging, get_handlers, get_logger
 
 
 class TestDataset(unittest.TestCase):

templates/image_classification/utils.py

@@ -21,6 +21,8 @@
 from torch.optim import Optimizer, SGD
 
 from models import get_model
+{% include "_handlers.py" %}
+
 
 
 def initialize(config: Optional[Any]) -> Tuple[Module, Optimizer, Module, Union[_LRScheduler, ParamScheduler]]:

templates/single/test_all.py

@@ -26,15 +26,14 @@
 from torch import nn, optim
 
 from config import get_default_parser
-from handlers import get_handlers, get_logger
 from trainers import (
     TrainEvents,
     create_trainers,
     evaluate_function,
     train_events_to_attr,
     train_function,
 )
-from utils import hash_checkpoint, log_metrics, resume_from, setup_logging
+from utils import hash_checkpoint, log_metrics, resume_from, setup_logging, get_handlers, get_logger
 
 
 class TestHandlers(unittest.TestCase):

templates/single/utils.py

@@ -18,6 +18,7 @@
 from torch.nn import Module
 from torch.optim.lr_scheduler import _LRScheduler
 from torch.optim.optimizer import Optimizer
+{% include "_handlers.py" %}
 
 
 # we can use `idist.auto_model` to handle distributed configurations