GANDLF 0.1.3.dev20250217__tar.gz → 0.1.3.dev20250218__tar.gz

{gandlf-0.1.3.dev20250217 → gandlf-0.1.3.dev20250218}/GANDLF/cli/main_run.py

@@ -19,6 +19,7 @@ def main_run(
     device: str,
     resume: bool,
     reset: bool,
+    _profile: Optional[bool] = False,
     output_dir: Optional[str] = None,
 ) -> None:
     """
@@ -98,6 +99,7 @@ def main_run(
                 device=device,
                 resume=resume,
                 reset=reset,
+                _profile=_profile,
             )
     else:
         data_full, headers = parseTrainingCSV(file_data_full, train=train_mode)

{gandlf-0.1.3.dev20250217 → gandlf-0.1.3.dev20250218}/GANDLF/entrypoints/run.py

@@ -24,6 +24,7 @@ def _run(
     reset_flag: bool,
     resume_flag: bool,
     output_path: Optional[str],
+    _profile: Optional[bool] = False,
 ):
     if model_dir is None and output_path:
         model_dir = output_path
@@ -60,6 +61,7 @@ def _run(
     logging.debug(f"{reset_flag=}")
     logging.debug(f"{resume_flag=}")
     logging.debug(f"{output_path=}")
+    logging.debug(f"{_profile=}")
 
     main_run(
         data_csv=input_data,
@@ -70,6 +72,7 @@ def _run(
         resume=resume_flag,
         reset=reset_flag,
         output_dir=output_path,
+        _profile=_profile,
     )
     print("Finished.")
 
@@ -145,6 +148,12 @@ def _run(
     default=None,
     help="Output file which will contain the logs.",
 )
+@click.option(
+    "--profile",
+    "-pf",
+    is_flag=True,
+    help="Track the run time and memory consumption for each layer",
+)
 @append_copyright_to_help
 def new_way(
     config: str,
@@ -156,6 +165,7 @@ def new_way(
     resume: bool,
     output_path: str,
     raw_input: str,
+    profile: bool,
     log_file: str,
 ):
     """Semantic segmentation, regression, and classification for medical images using Deep Learning."""
@@ -170,6 +180,7 @@ def new_way(
         reset_flag=reset,
         resume_flag=resume,
         output_path=output_path,
+        _profile=profile,
     )
 
 

{gandlf-0.1.3.dev20250217 → gandlf-0.1.3.dev20250218}/GANDLF/training_manager.py

@@ -1,6 +1,7 @@
 import pandas as pd
 import os, pickle, shutil
 from pathlib import Path
+from torch.profiler import profile, ProfilerActivity
 
 from GANDLF.compute import training_loop
 from GANDLF.utils import get_dataframe, split_data
@@ -145,6 +146,7 @@ def TrainingManager_split(
     device: str,
     resume: bool,
     reset: bool,
+    _profile: bool,
 ):
     """
     This is the training manager that ties all the training functionality together
@@ -158,6 +160,8 @@ def TrainingManager_split(
         device (str): The device to perform computations on.
         resume (bool): Whether the previous run will be resumed or not.
         reset (bool): Whether the previous run will be reset or not.
+        _profile(bool):Whether the we want the profile activity or not.
+
     """
     currentModelConfigPickle = os.path.join(outputDir, "parameters.pkl")
     currentModelConfigYaml = os.path.join(outputDir, "config.yaml")
@@ -178,11 +182,27 @@ def TrainingManager_split(
         with open(currentModelConfigYaml, "w") as handle:
             yaml.dump(parameters, handle, default_flow_style=False)
 
-    training_loop(
-        training_data=dataframe_train,
-        validation_data=dataframe_validation,
-        output_dir=outputDir,
-        device=device,
-        params=parameters,
-        testing_data=dataframe_testing,
-    )
+    if _profile:
+        with profile(
+            activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA],
+            profile_memory=True,
+            record_shapes=True,
+        ) as prof:
+            training_loop(
+                training_data=dataframe_train,
+                validation_data=dataframe_validation,
+                output_dir=outputDir,
+                device=device,
+                params=parameters,
+                testing_data=dataframe_testing,
+            )
+        print(prof.key_averages().table(sort_by="cpu_time_total", row_limit=10))
+    else:
+        training_loop(
+            training_data=dataframe_train,
+            validation_data=dataframe_validation,
+            output_dir=outputDir,
+            device=device,
+            params=parameters,
+            testing_data=dataframe_testing,
+        )

gandlf-0.1.3.dev20250218/GANDLF/version.py

@@ -0,0 +1 @@
+__version__ = "0.1.3-dev20250218"

{gandlf-0.1.3.dev20250217 → gandlf-0.1.3.dev20250218}/GANDLF.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: GANDLF
-Version: 0.1.3.dev20250217
+Version: 0.1.3.dev20250218
 Summary: PyTorch-based framework that handles segmentation/regression/classification using various DL architectures for medical imaging.
 Author: MLCommons
 Author-email: gandlf@mlcommons.org

{gandlf-0.1.3.dev20250217 → gandlf-0.1.3.dev20250218}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: GANDLF
-Version: 0.1.3.dev20250217
+Version: 0.1.3.dev20250218
 Summary: PyTorch-based framework that handles segmentation/regression/classification using various DL architectures for medical imaging.
 Author: MLCommons
 Author-email: gandlf@mlcommons.org

gandlf-0.1.3.dev20250217/GANDLF/version.py

@@ -1 +0,0 @@
-__version__ = "0.1.3-dev20250217"