train_model.py

import tensorflow as tf
from tensorflow.keras.models import Model
from tensorflow.keras.layers import GlobalAveragePooling2D, Dense, Dropout
from tensorflow.keras.applications import MobileNetV2
from tensorflow.keras.preprocessing.image import ImageDataGenerator
import os
import matplotlib.pyplot as plt

DATASET_DIR = 'archive/chest_xray'
TRAIN_DIR = os.path.join(DATASET_DIR, 'train')
VAL_DIR = os.path.join(DATASET_DIR, 'val')
TEST_DIR = os.path.join(DATASET_DIR, 'test')
MODEL_FILENAME = 'pneumonia_detection_model.keras'

IMAGE_SIZE = (224, 224)
BATCH_SIZE = 32
EPOCHS = 10
NUM_CLASSES = 2 # Pneumonia aur Normal
CLASS_NAMES = ['NORMAL', 'PNEUMONIA']

print(f"TensorFlow Version: {tf.__version__}")
print(f"Looking for data in: {TRAIN_DIR}")

train_datagen = ImageDataGenerator(
    rescale=1./255, # Normalization
    rotation_range=20,
    width_shift_range=0.1,
    height_shift_range=0.1,
    shear_range=0.1,
    zoom_range=0.1,
    horizontal_flip=True,
    fill_mode='nearest'
)


val_test_datagen = ImageDataGenerator(rescale=1./255)
train_generator = train_datagen.flow_from_directory(
    TRAIN_DIR,
    target_size=IMAGE_SIZE,
    batch_size=BATCH_SIZE,
    class_mode='categorical',
    classes=CLASS_NAMES # Class order define karna
)


validation_generator = val_test_datagen.flow_from_directory(
    VAL_DIR,
    target_size=IMAGE_SIZE,
    batch_size=BATCH_SIZE,
    class_mode='categorical',
    classes=CLASS_NAMES
)


test_generator = val_test_datagen.flow_from_directory(
    TEST_DIR,
    target_size=IMAGE_SIZE,
    batch_size=BATCH_SIZE,
    class_mode='categorical',
    classes=CLASS_NAMES,
    shuffle=False # Test data ko shuffle nahi karna hai
)


base_model = MobileNetV2(
    input_shape=IMAGE_SIZE + (3,),
    include_top=False,
    weights='imagenet'
)


base_model.trainable = False 


x = base_model.output
x = GlobalAveragePooling2D()(x)
x = Dense(512, activation='relu')(x)
x = Dropout(0.5)(x) # Overfitting kam karne ke liye
predictions = Dense(NUM_CLASSES, activation='softmax')(x) 

model = Model(inputs=base_model.input, outputs=predictions)


model.compile(optimizer='adam',
              loss='categorical_crossentropy',
              metrics=['accuracy'])

print("\n--- Starting Model Training ---")


history = model.fit(
    train_generator,
    steps_per_epoch=train_generator.samples // BATCH_SIZE,
    epochs=EPOCHS,
    validation_data=validation_generator,
    validation_steps=validation_generator.samples // BATCH_SIZE
)


print("\n--- Evaluating Model on Test Data ---")
loss, acc = model.evaluate(test_generator, steps=test_generator.samples // BATCH_SIZE)
print(f"Final Test Accuracy: {acc*100:.2f}%")

model.save(MODEL_FILENAME)
print(f"\nModel saved successfully as {MODEL_FILENAME}")

plt.figure(figsize=(12, 4))

plt.subplot(1, 2, 1)
plt.plot(history.history['accuracy'], label='Training Accuracy')
plt.plot(history.history['val_accuracy'], label='Validation Accuracy')
plt.title('Training and Validation Accuracy')
plt.xlabel('Epoch')
plt.ylabel('Accuracy')
plt.legend()

plt.subplot(1, 2, 2)
plt.plot(history.history['loss'], label='Training Loss')
plt.plot(history.history['val_loss'], label='Validation Loss')
plt.title('Training and Validation Loss')
plt.xlabel('Epoch')
plt.ylabel('Loss')
plt.legend()

plt.show()