Enhance checkpoint and logging configurations; update test setups

- Update configs/callbacks/hnet_checkpoint.yaml to:
  - Save checkpoints under trained_on_/standard_resolution/
  - Append timestamp to filenames for better traceability

- Modify configs/logging/tensorboard.yaml to:
  - Change logger name to trained_on_/standard_resolution
  - Update version format to include only hnet_model_DOA

- Update pytest.ini to include new markers:
  - Add scenarios_run for tests that train the algorithm
  - Add scenarios_test for tests that evaluate the algorithm

- Improve documentation in tests/scenarios_tests/conftest.py:
  - Add Args and Returns sections to fixtures for clarity
  - Update descriptions to accurately reflect fixture behavior

- Clean up tests/scenarios_tests/generate_hnet_training_data/test_scenarios_generate_hnet_training_data.py:
  - Remove unused numpy import

- Update data paths and rename test function in tests/scenarios_tests/run/test_scenarios_run.py:
  - Rename test_train_hnetgru_under_various_distributions to test_run_under_various_distributions for clarity

configs/callbacks/hnet_checkpoint.yaml

@@ -2,8 +2,8 @@ checkpoint:
   _target_: lightning.pytorch.callbacks.ModelCheckpoint
   save_last: True # additionally, always save model from last epoch
   verbose: False
-  dirpath: "${hydra:runtime.cwd}/checkpoints/${now:%Y%m%d}/"
-  filename: "hnet_model_DOA${max_len}_${sample_range_used}_{epoch:01d}"
+  dirpath: "${hydra:runtime.cwd}/checkpoints/${now:%Y%m%d}/trained_on_${sample_range_trained_on}/standard_resolution/"
+  filename: "hnet_model_DOA${max_len}_${sample_range_used}_{epoch:01d}_{now:%H%M%S}"
   monitor: "validation_loss"
   save_top_k: 1
   mode: "min"

configs/logging/tensorboard.yaml

@@ -3,7 +3,7 @@ log_every_n_steps: 100
 logger:
   _target_: lightning.pytorch.loggers.tensorboard.TensorBoardLogger
   save_dir: "${hydra:runtime.cwd}/tb_logs/"
-  name: 'hnet_model'
-  version: "hnet_model_DOA${max_len}_${sample_range_used}"
+  name: 'trained_on_${sample_range_trained_on}/standard_resolution'
+  version: "hnet_model_DOA${max_len}"
   log_graph: False
   default_hp_metric: False

pytest.ini

@@ -3,5 +3,7 @@ addopts = --cache-clear --strict-markers -vv --capture=tee-sys --cov=hungarian_n
 markers =
     consistency: mark tests for consistency checks
     scenarios: mark tests for scenario-based tests
+    scenarios_run: mark tests for scenario-based tests that train the algorithm
+    scenarios_test: mark tests for scenario-based tests that test the algorithm
     scenarios_generate_data: mark tests for scenario-based tests that generate data
     nonregression: mark non-regression tests to ensure existing functionality is not broken

run.py

@@ -38,8 +38,7 @@ def main(cfg: DictConfig):
         dict(hydra.utils.instantiate(cfg.metrics))
     )
     lightning_module: L.LightningModule = hydra.utils.instantiate(
-        cfg.lightning_module,
-        metrics=metrics
+        cfg.lightning_module, metrics=metrics
     )
 
     # Instantiate Trainer

tests/scenarios_tests/conftest.py

@@ -37,6 +37,13 @@ def batch_size(request) -> int:
     and behavior with various batch sizes. This helps in ensuring that the model scales
     appropriately with different amounts of data processed in each training iteration.
 
+    Args:
+        request (FixtureRequest): Pytest's fixture request object that provides access to the
+                                  parameters specified in the `params` list.
+
+    Returns:
+        int: The current value of `batch_size` for the test iteration.
+
     Example:
         When used in a test, `batch_size` will sequentially take the values 64, 128, and 256.
     """
@@ -80,7 +87,7 @@ def nb_epochs(request) -> int:
                                   parameters specified in the `params` list.
 
     Returns:
-        int: The current `nb_epochs` for the test iteration.
+        int: The current number of training epochs for the test iteration.
     """
     return request.param
 

tests/scenarios_tests/generate_hnet_training_data/test_scenarios_generate_hnet_training_data.py

@@ -1,6 +1,5 @@
 # tests/scenarios_tests/generate_hnet_training_data/test_scenarios_generate_hnet_training_data.py
 
-import numpy as np
 import pytest
 
 from generate_hnet_training_data import main

tests/scenarios_tests/run/test_scenarios_run.py

@@ -1,46 +1,48 @@
 # tests/scenarios_tests/run/test_scenarios_run.py
 
 import os
-from pathlib import Path
 import re
+from pathlib import Path
 
 import pytest
 
+
 @pytest.mark.scenarios
+@pytest.mark.scenarios_run
 @pytest.mark.parametrize(
     "training_data, test_data",
     [
         (
-            "data/20241205/train/hung_data_train_DOA2_3000-5000-15000",
-            "data/20241205/test/hung_data_test_DOA2_3000-5000-15000",
+            "data/20241206/train/hung_data_train_DOA2_3000-5000-15000",
+            "data/20241206/test/hung_data_test_DOA2_3000-5000-15000",
         ),
         (
-            "data/20241205/train/hung_data_train_DOA2_5000-5000-5000",
-            "data/20241205/test/hung_data_test_DOA2_5000-5000-5000",
+            "data/20241206/train/hung_data_train_DOA2_5000-5000-5000",
+            "data/20241206/test/hung_data_test_DOA2_5000-5000-5000",
         ),
         (
-            "data/20241205/train/hung_data_train_DOA2_1000-3000-31000",
-            "data/20241205/test/hung_data_test_DOA2_1000-3000-31000",
+            "data/20241206/train/hung_data_train_DOA2_1000-3000-31000",
+            "data/20241206/test/hung_data_test_DOA2_1000-3000-31000",
         ),
         (
-            "data/20241205/train/hung_data_train_DOA2_2600-5000-17000",
-            "data/20241205/test/hung_data_test_DOA2_2600-5000-17000",
+            "data/20241206/train/hung_data_train_DOA2_2600-5000-17000",
+            "data/20241206/test/hung_data_test_DOA2_2600-5000-17000",
         ),
         (
-            "data/20241205/train/hung_data_train_DOA2_6300-4000-1500",
-            "data/20241205/test/hung_data_test_DOA2_6300-4000-1500",
+            "data/20241206/train/hung_data_train_DOA2_6300-4000-1500",
+            "data/20241206/test/hung_data_test_DOA2_6300-4000-1500",
         ),
         (
-            "data/20241205/train/hung_data_train_DOA2_2000-7000-14000",
-            "data/20241205/test/hung_data_test_DOA2_2000-7000-14000",
+            "data/20241206/train/hung_data_train_DOA2_2000-7000-14000",
+            "data/20241206/test/hung_data_test_DOA2_2000-7000-14000",
         ),
         (
-            "data/20241205/train/hung_data_train_DOA2_2500-8000-8500",
-            "data/20241205/test/hung_data_test_DOA2_2500-8000-8500",
+            "data/20241206/train/hung_data_train_DOA2_2500-8000-8500",
+            "data/20241206/test/hung_data_test_DOA2_2500-8000-8500",
         ),
     ],
 )
-def test_train_hnetgru_under_various_distributions(training_data, test_data):
+def test_run_under_various_distributions(training_data, test_data):
     """
     Train the HNetGRU model with various data distributions.