adding config files

config.json

@@ -0,0 +1,19 @@
+{
+    "model_type": "TinyVGG",
+    "num_epochs": 10,
+    "batch_size": 32,
+    "learning_rate": 0.001,
+    "input_shape": 3,        
+    "hidden_units": 64,              
+    "output_shape": 10,              
+    "kernel_size": 3,                
+    "stride": 1,                    
+    "padding": 0,                    
+    "pooling_kernel_size": 2,       
+    "pooling_stride": 2,             
+    "framework": "PyTorch",          
+    "image_size": 64,                
+    "author": "Ajit",           
+    "description": "TinyVGG model for image classification with 64x64 input images."
+}
+  

train.py

@@ -9,13 +9,24 @@ from torchvision import transforms
 from timeit import default_timer as timer 
 from pathlib import Path
 from get_data import fetch_data
+import json
+
+
+
 
 def train_torch():
+    # Load config.json
+    with open("config.json", "r") as f:
+        config = json.load(f)
+
+    # Access configuration parameters
+    input_shape = config["input_shape"]
+    hidden_units = config["hidden_units"]
+  
     # Setup hyperparameters
-    NUM_EPOCHS = 10
-    BATCH_SIZE = 32
-    HIDDEN_UNITS = 10
-    LEARNING_RATE = 0.001
+    NUM_EPOCHS = config["num_epochs"]
+    BATCH_SIZE = config["batch_size"]
+    LEARNING_RATE = config["learning_rate"]
     
     # Define the URL and paths
     # Define the URL and paths
@@ -46,8 +57,8 @@ def train_torch():
                                                                                     batch_size=BATCH_SIZE)
     # Create model with help from model_builder.py
     model = model_builder.TinyVGG(
-        input_shape = 3,
-        hidden_units=HIDDEN_UNITS,
+        input_shape = input_shape,
+        hidden_units=hidden_units,
         output_shape=len(class_names)
     ).to(device)