fix model storage issue

.gitignore

@@ -1,6 +1,6 @@
 /data
 .aim
-model_storage*
+model_storage/*
 *__pycache__*
 logs/*
 credentials.json

dvc.lock

@@ -1,7 +1,7 @@
 schema: '2.0'
 stages:
   train:
-    cmd: python src/train.py --multirun --config-name=train experiment=catdog_ex trainer.max_epochs=2
+    cmd: python src/train.py --multirun --config-name=train experiment=catdog_ex trainer.max_epochs=3
     deps:
     - path: configs/experiment/catdog_ex.yaml
       hash: md5
@@ -23,8 +23,8 @@ stages:
     deps:
     - path: configs/experiment/catdog_ex.yaml
       hash: md5
-      md5: 959c64c29a91c7f7b4448e0c5cbc54f1
-      size: 1215
+      md5: d82b5d6fb9758f53546f97bda014f6f2
+      size: 1243
     - path: src/eval.py
       hash: md5
       md5: 47a9e50a0213ee71ff67a97b71d1ca6e
@@ -34,9 +34,11 @@ stages:
     deps:
     - path: configs/experiment/catdog_ex.yaml
       hash: md5
-      md5: 959c64c29a91c7f7b4448e0c5cbc54f1
-      size: 1215
+      md5: d82b5d6fb9758f53546f97bda014f6f2
+      size: 1243
     - path: src/infer.py
       hash: md5
       md5: 1b360e17f7740c9bd6eba9c52e0dc5d2
       size: 4717
+  code-test-coverage:
+    cmd: pytest --cov-report term --cov=src/ tests/

gradio_demo/app.py

@@ -12,6 +12,7 @@ def __init__(self, model_path="model_storage/epoch-checkpoint_patch_size-8_embed
         self.model = self.model.to(self.device)
         self.model.eval()
         # test comment
+        # test comment 2
 
         # Define the same transforms used during training/testing
         self.transform = transforms.Compose([

scripts/multirun_metrics_fetch.py

@@ -41,6 +41,7 @@ def get_metrics(log_dir, timestamp, mode):
     for run in os.listdir(metrics_dir):
         metrics_file = os.path.join(metrics_dir, run, "csv", "version_0", 'metrics.csv')
         if os.path.exists(metrics_file):
+            print(metrics_file)
             metrics_csv_dict[int(run)] = read_csv_to_dict(metrics_file)
     return metrics_csv_dict
 
@@ -75,7 +76,7 @@ def main_run():
     assert list(train_metrics.keys()).sort() == list(hyperparams.keys()).sort()
 
     data = [["Exp No"]]
-    # print(train_metrics)
+    print(train_metrics)
     metrics_keys = train_metrics[0].keys()
     hyperparams_keys = hyperparams[0].keys()
     data[0].extend(metrics_keys)

src/train.py

@@ -54,7 +54,7 @@ def _save_checkpoint(self, trainer, filepath):
         # Export the model to TorchScript
         example_input = torch.randn(1, 3, 224, 224)  # Example input for tracing
         try:
-            # model_torch = copy.deepcopy(trainer.model)
+            model_torch = copy.deepcopy(trainer.model)
             torchscript_model = model_torch.to_torchscript(method="trace", example_inputs=example_input)
 
             # Save the TorchScript model