Delete train_model.py

train_model.py

@@ -1,52 +0,0 @@
-import os
-import numpy as np
-from tensorflow.keras.models import Sequential
-from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense
-from tensorflow.keras.preprocessing.image import ImageDataGenerator
-from tensorflow.keras.optimizers import Adam
-
-# Set paths to the dataset (adjust paths based on your directory structure)
-train_dir = './data/train'
-validation_dir = './data/validation'
-
-# Define the CNN model
-def create_cnn_model(input_shape=(224, 224, 3)):
-    model = Sequential()
-    model.add(Conv2D(32, (3, 3), activation='relu', input_shape=input_shape))
-    model.add(MaxPooling2D((2, 2)))
-    
-    model.add(Conv2D(64, (3, 3), activation='relu'))
-    model.add(MaxPooling2D((2, 2)))
-    
-    model.add(Conv2D(128, (3, 3), activation='relu'))
-    model.add(MaxPooling2D((2, 2)))
-    
-    model.add(Flatten())
-    model.add(Dense(128, activation='relu'))
-    model.add(Dense(1, activation='sigmoid'))  # Binary classification (Normal vs Abnormal)
-    
-    model.compile(optimizer=Adam(), loss='binary_crossentropy', metrics=['accuracy'])
-    return model
-
-# Create the CNN model
-model = create_cnn_model()
-
-# ImageDataGenerator for training and validation
-train_datagen = ImageDataGenerator(rescale=1./255, rotation_range=40, width_shift_range=0.2, 
-                                   height_shift_range=0.2, shear_range=0.2, zoom_range=0.2,
-                                   horizontal_flip=True, fill_mode='nearest')
-
-validation_datagen = ImageDataGenerator(rescale=1./255)
-
-# Flow training and validation data from directories
-train_generator = train_datagen.flow_from_directory(train_dir, target_size=(224, 224),
-                                                    batch_size=32, class_mode='binary')
-
-validation_generator = validation_datagen.flow_from_directory(validation_dir, target_size=(224, 224),
-                                                              batch_size=32, class_mode='binary')
-
-# Train the model
-history = model.fit(train_generator, epochs=10, validation_data=validation_generator)
-
-# Save the trained model
-model.save('classification_model.h5')