refactor: Reorganize project structure

- Moved data loader to src/sroie_loader.py
- Moved training and eval scripts to scripts/ folder
- Updated imports to fix python path
- Removed unused tests/utils.py

.dockerignore

@@ -29,6 +29,7 @@ env
 data/
 outputs/
 temp/
+models/
 
 # Tests (not needed in production)
 tests/

eval_new_dataset.py → scripts/eval_new_dataset.py

@@ -1,3 +1,8 @@
+import sys
+import os
+
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
+
 import torch
 from src.data_loader import load_unified_dataset
 from transformers import LayoutLMv3ForTokenClassification, LayoutLMv3Processor, DataCollatorForTokenClassification

explore_new_dataset.py → scripts/explore_new_dataset.py

@@ -1,3 +1,8 @@
+import sys
+import os
+
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
+
 from datasets import load_dataset
 import json
 import ast  # <--- Added for robust parsing

train_combined.py → scripts/train_combined.py

@@ -1,3 +1,8 @@
+import sys
+import os
+
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
+
 import torch
 from torch.utils.data import Dataset, DataLoader
 from transformers import LayoutLMv3ForTokenClassification, LayoutLMv3Processor, DataCollatorForTokenClassification
@@ -10,7 +15,7 @@ import random
 import os
 
 # --- IMPORTS ---
-from load_sroie_dataset import load_sroie
+from src.sroie_loader import load_sroie
 from src.data_loader import load_unified_dataset
 
 # --- CONFIGURATION ---

train_layoutlm.py → scripts/train_layoutlm.py

@@ -1,7 +1,12 @@
+import sys
+import os
+
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
+
 import torch
 from torch.utils.data import Dataset, DataLoader
 from transformers import LayoutLMv3ForTokenClassification, LayoutLMv3Processor, DataCollatorForTokenClassification
-from load_sroie_dataset import load_sroie # Assumes your helper script is in the root
+from src.sroie_loader import load_sroie
 from PIL import Image
 from tqdm import tqdm
 from seqeval.metrics import f1_score, precision_score, recall_score

tests/utils.py

@@ -1,7 +0,0 @@
-def save_image(image, path):
-
-def visualize_boxes(image, boxes, text):
-
-def validate_output(data):
-
-def format_currency(amount):