"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[1m59/59\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m30s\u001b[0m 502ms/step - accuracy: 0.4418 - loss: 1.6044\n",
"\n",
"Test Accuracy: 0.4418\n",
"Test Loss: 1.6044\n"
]
},
{
"data": {
"text/html": [
"Model: \"sequential\"\n",
"\n"
],
"text/plain": [
"\u001b[1mModel: \"sequential\"\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ rescaling (Rescaling) │ (None, 512, 512, 3) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d (Conv2D) │ (None, 510, 510, 32) │ 896 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d (MaxPooling2D) │ (None, 255, 255, 32) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_1 (Conv2D) │ (None, 253, 253, 64) │ 18,496 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_1 (MaxPooling2D) │ (None, 126, 126, 64) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten (Flatten) │ (None, 1016064) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense (Dense) │ (None, 128) │ 130,056,320 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout (Dropout) │ (None, 128) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (Dense) │ (None, 12) │ 1,548 │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"\n"
],
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ rescaling (\u001b[38;5;33mRescaling\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m512\u001b[0m, \u001b[38;5;34m512\u001b[0m, \u001b[38;5;34m3\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m510\u001b[0m, \u001b[38;5;34m510\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m896\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m255\u001b[0m, \u001b[38;5;34m255\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m253\u001b[0m, \u001b[38;5;34m253\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m18,496\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_1 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m126\u001b[0m, \u001b[38;5;34m126\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten (\u001b[38;5;33mFlatten\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1016064\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m130,056,320\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout (\u001b[38;5;33mDropout\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m) │ \u001b[38;5;34m1,548\u001b[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Total params: 390,231,782 (1.45 GB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m390,231,782\u001b[0m (1.45 GB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Trainable params: 130,077,260 (496.21 MB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m130,077,260\u001b[0m (496.21 MB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Non-trainable params: 0 (0.00 B)\n",
"\n"
],
"text/plain": [
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Optimizer params: 260,154,522 (992.41 MB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Optimizer params: \u001b[0m\u001b[38;5;34m260,154,522\u001b[0m (992.41 MB)\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# 3. Optimise data loading, using a TensorFlow tool\n",
"\n",
"AUTOTUNE = tf.data.AUTOTUNE\n",
"\n",
"train_ds = train_ds.prefetch(AUTOTUNE)\n",
"\n",
"valid_ds = valid_ds.prefetch(AUTOTUNE)\n",
"\n",
"test_ds = test_ds.prefetch(AUTOTUNE)\n",
"\n",
" \n",
"# Build a basic CNN\n",
"\n",
"num_classes = len(class_names)\n",
"model = models.Sequential([\n",
"\n",
" layers.Rescaling(1./255, input_shape=(512, 512, 3)),\n",
"\n",
"\n",
" layers.Conv2D(32, 3, activation='relu'),\n",
"\n",
" layers.MaxPooling2D(),\n",
"\n",
" \n",
"\n",
" layers.Conv2D(64, 3, activation='relu'),\n",
"\n",
" layers.MaxPooling2D(),\n",
"\n",
"\n",
" layers.Flatten(),\n",
"\n",
" layers.Dense(128, activation='relu'),\n",
"\n",
" layers.Dropout(0.5),\n",
"\n",
" layers.Dense(num_classes, activation='softmax')\n",
"\n",
"])\n",
"\n",
"\n",
"model.summary()\n",
"\n",
"\n",
"# 5. Compile the model\n",
"\n",
"model.compile(\n",
"\n",
" optimizer='adam',\n",
"\n",
" loss='sparse_categorical_crossentropy',\n",
"\n",
" metrics=['accuracy']\n",
"\n",
")\n",
"\n",
" \n",
"\n",
"# 6. Train the model\n",
"\n",
"history = model.fit(\n",
"\n",
" train_ds,\n",
"\n",
" validation_data=valid_ds,\n",
"\n",
" epochs=3\n",
")\n",
"\n",
" \n",
"plt.figure(figsize=(12,5))\n",
"\n",
"\n",
"# Check the results\n",
"\n",
" \n",
"\n",
"# Accuracy\n",
"\n",
"plt.subplot(1,2,1)\n",
"\n",
"plt.plot(history.history['accuracy'])\n",
"\n",
"plt.plot(history.history['val_accuracy'])\n",
"\n",
"plt.title(\"Model Accuracy\")\n",
"\n",
"plt.xlabel(\"Epoch\")\n",
"\n",
"plt.ylabel(\"Accuracy\")\n",
"\n",
"plt.legend(['Train','Validation'])\n",
"\n",
" \n",
"\n",
"# Loss\n",
"\n",
"plt.subplot(1,2,2)\n",
"\n",
"plt.plot(history.history['loss'])\n",
"\n",
"plt.plot(history.history['val_loss'])\n",
"\n",
"plt.title(\"Model Loss\")\n",
"\n",
"plt.xlabel(\"Epoch\")\n",
"\n",
"plt.ylabel(\"Loss\")\n",
"\n",
"plt.legend(['Train','Validation'])\n",
"\n",
"\n",
"plt.show()\n",
"\n",
"\n",
"# 7. Evaluate on the test set\n",
"\n",
"test_loss, test_accuracy = model.evaluate(test_ds)\n",
"\n",
"print(f\"\\nTest Accuracy: {test_accuracy:.4f}\")\n",
"\n",
"print(f\"Test Loss: {test_loss:.4f}\")\n",
"\n",
"model.summary()"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "9b4927c9-b3be-4663-82e9-38f01b680e55",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Found 8691 files belonging to 12 classes.\n",
"Found 1862 files belonging to 12 classes.\n",
"Found 1863 files belonging to 12 classes.\n"
]
},
{
"data": {
"text/html": [
"Model: \"sequential_2\"\n",
"\n"
],
"text/plain": [
"\u001b[1mModel: \"sequential_2\"\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ sequential_1 (Sequential) │ (None, 224, 224, 3) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ rescaling_1 (Rescaling) │ (None, 224, 224, 3) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_2 (Conv2D) │ (None, 224, 224, 32) │ 896 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ batch_normalization │ (None, 224, 224, 32) │ 128 │\n",
"│ (BatchNormalization) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_2 (MaxPooling2D) │ (None, 112, 112, 32) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_3 (Conv2D) │ (None, 112, 112, 64) │ 18,496 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ batch_normalization_1 │ (None, 112, 112, 64) │ 256 │\n",
"│ (BatchNormalization) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_3 (MaxPooling2D) │ (None, 56, 56, 64) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_4 (Conv2D) │ (None, 56, 56, 128) │ 73,856 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ batch_normalization_2 │ (None, 56, 56, 128) │ 512 │\n",
"│ (BatchNormalization) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_4 (MaxPooling2D) │ (None, 28, 28, 128) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ global_average_pooling2d │ (None, 128) │ 0 │\n",
"│ (GlobalAveragePooling2D) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (Dense) │ (None, 128) │ 16,512 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_1 (Dropout) │ (None, 128) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_3 (Dense) │ (None, 12) │ 1,548 │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"\n"
],
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ sequential_1 (\u001b[38;5;33mSequential\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ rescaling_1 (\u001b[38;5;33mRescaling\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m896\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ batch_normalization │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m128\u001b[0m │\n",
"│ (\u001b[38;5;33mBatchNormalization\u001b[0m) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_2 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_3 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m18,496\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ batch_normalization_1 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m256\u001b[0m │\n",
"│ (\u001b[38;5;33mBatchNormalization\u001b[0m) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_3 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_4 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m73,856\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ batch_normalization_2 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m512\u001b[0m │\n",
"│ (\u001b[38;5;33mBatchNormalization\u001b[0m) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_4 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ global_average_pooling2d │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"│ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m16,512\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_1 (\u001b[38;5;33mDropout\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_3 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m) │ \u001b[38;5;34m1,548\u001b[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Total params: 112,204 (438.30 KB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m112,204\u001b[0m (438.30 KB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Trainable params: 111,756 (436.55 KB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m111,756\u001b[0m (436.55 KB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Non-trainable params: 448 (1.75 KB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m448\u001b[0m (1.75 KB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m449s\u001b[0m 2s/step - accuracy: 0.3413 - loss: 1.8854 - val_accuracy: 0.0875 - val_loss: 2.3504 - learning_rate: 1.0000e-04\n",
"Epoch 2/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m461s\u001b[0m 2s/step - accuracy: 0.4072 - loss: 1.6692 - val_accuracy: 0.3894 - val_loss: 1.6650 - learning_rate: 1.0000e-04\n",
"Epoch 3/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m447s\u001b[0m 2s/step - accuracy: 0.4324 - loss: 1.5955 - val_accuracy: 0.4898 - val_loss: 1.4789 - learning_rate: 1.0000e-04\n",
"Epoch 4/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m474s\u001b[0m 2s/step - accuracy: 0.4541 - loss: 1.5454 - val_accuracy: 0.5081 - val_loss: 1.4385 - learning_rate: 1.0000e-04\n",
"Epoch 5/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m444s\u001b[0m 2s/step - accuracy: 0.4683 - loss: 1.5039 - val_accuracy: 0.4705 - val_loss: 1.4925 - learning_rate: 1.0000e-04\n",
"Epoch 6/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1342s\u001b[0m 5s/step - accuracy: 0.4791 - loss: 1.4711 - val_accuracy: 0.5011 - val_loss: 1.4671 - learning_rate: 1.0000e-04\n",
"Epoch 7/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m42541s\u001b[0m 157s/step - accuracy: 0.4894 - loss: 1.4478 - val_accuracy: 0.5263 - val_loss: 1.3882 - learning_rate: 1.0000e-04\n",
"Epoch 8/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m459s\u001b[0m 2s/step - accuracy: 0.5129 - loss: 1.4072 - val_accuracy: 0.5317 - val_loss: 1.3901 - learning_rate: 1.0000e-04\n",
"Epoch 9/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m378s\u001b[0m 1s/step - accuracy: 0.5281 - loss: 1.3765 - val_accuracy: 0.4882 - val_loss: 1.5264 - learning_rate: 1.0000e-04\n",
"Epoch 10/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m388s\u001b[0m 1s/step - accuracy: 0.5347 - loss: 1.3523 - val_accuracy: 0.5166 - val_loss: 1.3977 - learning_rate: 1.0000e-04\n",
"Epoch 11/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m377s\u001b[0m 1s/step - accuracy: 0.5539 - loss: 1.2976 - val_accuracy: 0.5559 - val_loss: 1.2993 - learning_rate: 5.0000e-05\n",
"Epoch 12/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m387s\u001b[0m 1s/step - accuracy: 0.5668 - loss: 1.2759 - val_accuracy: 0.6026 - val_loss: 1.2325 - learning_rate: 5.0000e-05\n",
"Epoch 13/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m380s\u001b[0m 1s/step - accuracy: 0.5716 - loss: 1.2663 - val_accuracy: 0.5736 - val_loss: 1.2765 - learning_rate: 5.0000e-05\n",
"Epoch 14/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m379s\u001b[0m 1s/step - accuracy: 0.5751 - loss: 1.2503 - val_accuracy: 0.5961 - val_loss: 1.2198 - learning_rate: 5.0000e-05\n",
"Epoch 15/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m376s\u001b[0m 1s/step - accuracy: 0.5820 - loss: 1.2288 - val_accuracy: 0.5714 - val_loss: 1.2819 - learning_rate: 5.0000e-05\n",
"Epoch 16/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m402s\u001b[0m 1s/step - accuracy: 0.5902 - loss: 1.2175 - val_accuracy: 0.5559 - val_loss: 1.2977 - learning_rate: 5.0000e-05\n",
"Epoch 17/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1465s\u001b[0m 5s/step - accuracy: 0.5958 - loss: 1.1998 - val_accuracy: 0.5789 - val_loss: 1.2392 - learning_rate: 5.0000e-05\n",
"Epoch 18/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m414s\u001b[0m 2s/step - accuracy: 0.6032 - loss: 1.1879 - val_accuracy: 0.6278 - val_loss: 1.1819 - learning_rate: 2.5000e-05\n",
"Epoch 19/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m411s\u001b[0m 2s/step - accuracy: 0.6064 - loss: 1.1716 - val_accuracy: 0.6085 - val_loss: 1.1939 - learning_rate: 2.5000e-05\n",
"Epoch 20/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m401s\u001b[0m 1s/step - accuracy: 0.6083 - loss: 1.1696 - val_accuracy: 0.5827 - val_loss: 1.2324 - learning_rate: 2.5000e-05\n",
"Epoch 21/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m369s\u001b[0m 1s/step - accuracy: 0.6213 - loss: 1.1544 - val_accuracy: 0.5897 - val_loss: 1.2206 - learning_rate: 2.5000e-05\n",
"Epoch 22/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m371s\u001b[0m 1s/step - accuracy: 0.6186 - loss: 1.1469 - val_accuracy: 0.6224 - val_loss: 1.1696 - learning_rate: 1.2500e-05\n",
"Epoch 23/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m634s\u001b[0m 2s/step - accuracy: 0.6155 - loss: 1.1441 - val_accuracy: 0.6294 - val_loss: 1.1524 - learning_rate: 1.2500e-05\n",
"Epoch 24/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m383s\u001b[0m 1s/step - accuracy: 0.6252 - loss: 1.1405 - val_accuracy: 0.6375 - val_loss: 1.1374 - learning_rate: 1.2500e-05\n",
"Epoch 25/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m428s\u001b[0m 2s/step - accuracy: 0.6267 - loss: 1.1275 - val_accuracy: 0.6047 - val_loss: 1.1976 - learning_rate: 1.2500e-05\n",
"Epoch 26/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m432s\u001b[0m 2s/step - accuracy: 0.6274 - loss: 1.1292 - val_accuracy: 0.6359 - val_loss: 1.1296 - learning_rate: 1.2500e-05\n",
"Epoch 27/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m542s\u001b[0m 2s/step - accuracy: 0.6219 - loss: 1.1415 - val_accuracy: 0.6251 - val_loss: 1.1587 - learning_rate: 1.2500e-05\n",
"Epoch 28/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m434s\u001b[0m 2s/step - accuracy: 0.6262 - loss: 1.1240 - val_accuracy: 0.6171 - val_loss: 1.1499 - learning_rate: 1.2500e-05\n",
"Epoch 29/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m462s\u001b[0m 2s/step - accuracy: 0.6358 - loss: 1.1123 - val_accuracy: 0.6235 - val_loss: 1.1603 - learning_rate: 1.2500e-05\n",
"Epoch 30/30\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m32358s\u001b[0m 119s/step - accuracy: 0.6342 - loss: 1.1116 - val_accuracy: 0.6359 - val_loss: 1.1266 - learning_rate: 6.2500e-06\n"
]
}
],
"source": [
"IMG_SIZE = (224, 224) # smaller, faster, still enough detail for leaf disease\n",
"BATCH_SIZE = 32\n",
"\n",
"train_ds = tf.keras.utils.image_dataset_from_directory(\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\train\",\n",
" image_size=IMG_SIZE,\n",
" batch_size=BATCH_SIZE,\n",
" label_mode=\"int\"\n",
")\n",
"\n",
"valid_ds = tf.keras.utils.image_dataset_from_directory(\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\valid\",\n",
" image_size=IMG_SIZE,\n",
" batch_size=BATCH_SIZE,\n",
" label_mode=\"int\"\n",
")\n",
"\n",
"test_ds = tf.keras.utils.image_dataset_from_directory(\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\test\",\n",
" image_size=IMG_SIZE,\n",
" batch_size=BATCH_SIZE,\n",
" label_mode=\"int\",\n",
" shuffle=False\n",
")\n",
"\n",
"class_names = train_ds.class_names\n",
"num_classes = len(class_names)\n",
"\n",
"AUTOTUNE = tf.data.AUTOTUNE\n",
"train_ds = train_ds.prefetch(AUTOTUNE)\n",
"valid_ds = valid_ds.prefetch(AUTOTUNE)\n",
"test_ds = test_ds.prefetch(AUTOTUNE)\n",
"\n",
"# Data augmentation — reduces overfitting significantly\n",
"data_augmentation = tf.keras.Sequential([\n",
" layers.RandomFlip(\"horizontal_and_vertical\"),\n",
" layers.RandomRotation(0.2),\n",
" layers.RandomZoom(0.2),\n",
" layers.RandomContrast(0.1),\n",
"])\n",
"\n",
"model = models.Sequential([\n",
" layers.Input(shape=(224, 224, 3)),\n",
" data_augmentation,\n",
" layers.Rescaling(1./255),\n",
"\n",
" layers.Conv2D(32, 3, activation='relu', padding='same'),\n",
" layers.BatchNormalization(),\n",
" layers.MaxPooling2D(),\n",
"\n",
" layers.Conv2D(64, 3, activation='relu', padding='same'),\n",
" layers.BatchNormalization(),\n",
" layers.MaxPooling2D(),\n",
"\n",
" layers.Conv2D(128, 3, activation='relu', padding='same'),\n",
" layers.BatchNormalization(),\n",
" layers.MaxPooling2D(),\n",
"\n",
" layers.GlobalAveragePooling2D(), # replaces Flatten -> huge param cut\n",
" layers.Dense(128, activation='relu'),\n",
" layers.Dropout(0.5),\n",
" layers.Dense(num_classes, activation='softmax')\n",
"])\n",
"\n",
"model.summary()\n",
"\n",
"model.compile(\n",
" optimizer=tf.keras.optimizers.Adam(learning_rate=1e-4),\n",
" loss='sparse_categorical_crossentropy',\n",
" metrics=['accuracy']\n",
")\n",
"\n",
"callbacks = [\n",
" tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True),\n",
" tf.keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=3)\n",
"]\n",
"\n",
"history = model.fit(\n",
" train_ds,\n",
" validation_data=valid_ds,\n",
" epochs=30,\n",
" callbacks=callbacks\n",
")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "6499a496-e025-443a-9d89-592d3204d3df",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Found 8691 files belonging to 12 classes.\n",
"Found 1862 files belonging to 12 classes.\n",
"Found 1863 files belonging to 12 classes.\n",
"Downloading data from https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5\n",
"\u001b[1m16705208/16705208\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 0us/step\n",
"Epoch 1/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m226s\u001b[0m 752ms/step - accuracy: 0.4859 - loss: 1.5149 - val_accuracy: 0.6208 - val_loss: 1.1973 - learning_rate: 0.0010\n",
"Epoch 2/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m187s\u001b[0m 686ms/step - accuracy: 0.6056 - loss: 1.2334 - val_accuracy: 0.6676 - val_loss: 1.0691 - learning_rate: 0.0010\n",
"Epoch 3/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m188s\u001b[0m 693ms/step - accuracy: 0.6263 - loss: 1.1527 - val_accuracy: 0.6826 - val_loss: 1.0199 - learning_rate: 0.0010\n",
"Epoch 4/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m185s\u001b[0m 681ms/step - accuracy: 0.6407 - loss: 1.1066 - val_accuracy: 0.6880 - val_loss: 0.9933 - learning_rate: 0.0010\n",
"Epoch 5/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m210s\u001b[0m 708ms/step - accuracy: 0.6457 - loss: 1.0848 - val_accuracy: 0.6933 - val_loss: 0.9695 - learning_rate: 0.0010\n",
"Epoch 6/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m193s\u001b[0m 711ms/step - accuracy: 0.6515 - loss: 1.0579 - val_accuracy: 0.7019 - val_loss: 0.9490 - learning_rate: 0.0010\n",
"Epoch 7/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m260s\u001b[0m 955ms/step - accuracy: 0.6530 - loss: 1.0457 - val_accuracy: 0.7111 - val_loss: 0.9316 - learning_rate: 0.0010\n",
"Epoch 8/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m215s\u001b[0m 791ms/step - accuracy: 0.6587 - loss: 1.0334 - val_accuracy: 0.7052 - val_loss: 0.9300 - learning_rate: 0.0010\n",
"Epoch 9/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m189s\u001b[0m 695ms/step - accuracy: 0.6606 - loss: 1.0389 - val_accuracy: 0.7170 - val_loss: 0.9000 - learning_rate: 0.0010\n",
"Epoch 10/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m186s\u001b[0m 685ms/step - accuracy: 0.6671 - loss: 1.0128 - val_accuracy: 0.7164 - val_loss: 0.8985 - learning_rate: 0.0010\n",
"Epoch 11/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m191s\u001b[0m 701ms/step - accuracy: 0.6682 - loss: 1.0099 - val_accuracy: 0.7202 - val_loss: 0.9004 - learning_rate: 0.0010\n",
"Epoch 12/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m203s\u001b[0m 746ms/step - accuracy: 0.6675 - loss: 1.0095 - val_accuracy: 0.7197 - val_loss: 0.8802 - learning_rate: 0.0010\n",
"Epoch 13/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m189s\u001b[0m 696ms/step - accuracy: 0.6689 - loss: 1.0058 - val_accuracy: 0.7143 - val_loss: 0.8934 - learning_rate: 0.0010\n",
"Epoch 14/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m194s\u001b[0m 713ms/step - accuracy: 0.6750 - loss: 0.9843 - val_accuracy: 0.7197 - val_loss: 0.8796 - learning_rate: 0.0010\n",
"Epoch 15/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m192s\u001b[0m 705ms/step - accuracy: 0.6703 - loss: 0.9994 - val_accuracy: 0.7266 - val_loss: 0.8655 - learning_rate: 0.0010\n",
"Epoch 1/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m251s\u001b[0m 832ms/step - accuracy: 0.4092 - loss: 1.8485 - val_accuracy: 0.5881 - val_loss: 1.3007 - learning_rate: 1.0000e-05\n",
"Epoch 2/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m943s\u001b[0m 3s/step - accuracy: 0.5256 - loss: 1.4831 - val_accuracy: 0.6654 - val_loss: 1.1094 - learning_rate: 1.0000e-05\n",
"Epoch 3/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m223s\u001b[0m 822ms/step - accuracy: 0.5896 - loss: 1.2996 - val_accuracy: 0.6923 - val_loss: 1.0085 - learning_rate: 1.0000e-05\n",
"Epoch 4/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m219s\u001b[0m 804ms/step - accuracy: 0.6257 - loss: 1.1927 - val_accuracy: 0.7057 - val_loss: 0.9473 - learning_rate: 1.0000e-05\n",
"Epoch 5/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m215s\u001b[0m 789ms/step - accuracy: 0.6446 - loss: 1.1216 - val_accuracy: 0.7175 - val_loss: 0.9095 - learning_rate: 1.0000e-05\n",
"Epoch 6/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m220s\u001b[0m 807ms/step - accuracy: 0.6602 - loss: 1.0742 - val_accuracy: 0.7207 - val_loss: 0.8836 - learning_rate: 1.0000e-05\n",
"Epoch 7/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m215s\u001b[0m 790ms/step - accuracy: 0.6707 - loss: 1.0364 - val_accuracy: 0.7293 - val_loss: 0.8619 - learning_rate: 1.0000e-05\n",
"Epoch 8/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m439s\u001b[0m 2s/step - accuracy: 0.6792 - loss: 0.9999 - val_accuracy: 0.7315 - val_loss: 0.8454 - learning_rate: 1.0000e-05\n",
"Epoch 9/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m262s\u001b[0m 965ms/step - accuracy: 0.6831 - loss: 0.9821 - val_accuracy: 0.7358 - val_loss: 0.8319 - learning_rate: 1.0000e-05\n",
"Epoch 10/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m216s\u001b[0m 795ms/step - accuracy: 0.6960 - loss: 0.9578 - val_accuracy: 0.7390 - val_loss: 0.8211 - learning_rate: 1.0000e-05\n",
"Epoch 11/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m413s\u001b[0m 786ms/step - accuracy: 0.7085 - loss: 0.9141 - val_accuracy: 0.7422 - val_loss: 0.8095 - learning_rate: 1.0000e-05\n",
"Epoch 12/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m924s\u001b[0m 3s/step - accuracy: 0.7067 - loss: 0.9164 - val_accuracy: 0.7481 - val_loss: 0.8006 - learning_rate: 1.0000e-05\n",
"Epoch 13/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m216s\u001b[0m 796ms/step - accuracy: 0.7150 - loss: 0.8896 - val_accuracy: 0.7524 - val_loss: 0.7928 - learning_rate: 1.0000e-05\n",
"Epoch 14/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m216s\u001b[0m 795ms/step - accuracy: 0.7182 - loss: 0.8801 - val_accuracy: 0.7556 - val_loss: 0.7826 - learning_rate: 1.0000e-05\n",
"Epoch 15/15\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m219s\u001b[0m 804ms/step - accuracy: 0.7237 - loss: 0.8618 - val_accuracy: 0.7562 - val_loss: 0.7755 - learning_rate: 1.0000e-05\n",
"\u001b[1m59/59\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m35s\u001b[0m 584ms/step - accuracy: 0.7687 - loss: 0.7463\n",
"\n",
"Test Accuracy: 0.7687\n",
"Test Loss: 0.7463\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import tensorflow as tf\n",
"from tensorflow.keras import layers, models\n",
"from tensorflow.keras.applications import EfficientNetB0\n",
"import matplotlib.pyplot as plt\n",
"\n",
"IMG_SIZE = (224, 224)\n",
"BATCH_SIZE = 32\n",
"\n",
"train_ds = tf.keras.utils.image_dataset_from_directory(\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\train\",\n",
" image_size=IMG_SIZE, batch_size=BATCH_SIZE, label_mode=\"int\"\n",
")\n",
"valid_ds = tf.keras.utils.image_dataset_from_directory(\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\valid\",\n",
" image_size=IMG_SIZE, batch_size=BATCH_SIZE, label_mode=\"int\"\n",
")\n",
"test_ds = tf.keras.utils.image_dataset_from_directory(\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\test\",\n",
" image_size=IMG_SIZE, batch_size=BATCH_SIZE, label_mode=\"int\", shuffle=False\n",
")\n",
"\n",
"class_names = train_ds.class_names\n",
"num_classes = len(class_names)\n",
"\n",
"AUTOTUNE = tf.data.AUTOTUNE\n",
"train_ds = train_ds.prefetch(AUTOTUNE)\n",
"valid_ds = valid_ds.prefetch(AUTOTUNE)\n",
"test_ds = test_ds.prefetch(AUTOTUNE)\n",
"\n",
"data_augmentation = tf.keras.Sequential([\n",
" layers.RandomFlip(\"horizontal_and_vertical\"),\n",
" layers.RandomRotation(0.2),\n",
" layers.RandomZoom(0.2),\n",
"])\n",
"\n",
"# ---- Phase 1: frozen base, train only the head ----\n",
"base_model = EfficientNetB0(include_top=False, weights='imagenet', input_shape=(224, 224, 3))\n",
"base_model.trainable = False\n",
"\n",
"inputs = layers.Input(shape=(224, 224, 3))\n",
"x = data_augmentation(inputs)\n",
"x = tf.keras.applications.efficientnet.preprocess_input(x)\n",
"x = base_model(x, training=False)\n",
"x = layers.GlobalAveragePooling2D()(x)\n",
"x = layers.Dropout(0.3)(x)\n",
"outputs = layers.Dense(num_classes, activation='softmax')(x)\n",
"model = models.Model(inputs, outputs)\n",
"\n",
"model.compile(\n",
" optimizer=tf.keras.optimizers.Adam(learning_rate=1e-3),\n",
" loss='sparse_categorical_crossentropy',\n",
" metrics=['accuracy']\n",
")\n",
"\n",
"callbacks = [\n",
" tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True),\n",
" tf.keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=2)\n",
"]\n",
"\n",
"history1 = model.fit(train_ds, validation_data=valid_ds, epochs=15, callbacks=callbacks)\n",
"\n",
"# ---- Phase 2: fine-tune top layers of the base model ----\n",
"base_model.trainable = True\n",
"for layer in base_model.layers[:-30]: # freeze all but last 30 layers\n",
" layer.trainable = False\n",
"\n",
"model.compile(\n",
" optimizer=tf.keras.optimizers.Adam(learning_rate=1e-5), # much lower LR for fine-tuning\n",
" loss='sparse_categorical_crossentropy',\n",
" metrics=['accuracy']\n",
")\n",
"\n",
"history2 = model.fit(train_ds, validation_data=valid_ds, epochs=15, callbacks=callbacks)\n",
"\n",
"# ---- Evaluate ----\n",
"test_loss, test_accuracy = model.evaluate(test_ds)\n",
"print(f\"\\nTest Accuracy: {test_accuracy:.4f}\")\n",
"print(f\"Test Loss: {test_loss:.4f}\")\n",
"\n",
"# ---- Plot combined history ----\n",
"acc = history1.history['accuracy'] + history2.history['accuracy']\n",
"val_acc = history1.history['val_accuracy'] + history2.history['val_accuracy']\n",
"loss = history1.history['loss'] + history2.history['loss']\n",
"val_loss = history1.history['val_loss'] + history2.history['val_loss']\n",
"\n",
"plt.figure(figsize=(12,5))\n",
"plt.subplot(1,2,1)\n",
"plt.plot(acc); plt.plot(val_acc)\n",
"plt.title(\"Model Accuracy\"); plt.xlabel(\"Epoch\"); plt.ylabel(\"Accuracy\")\n",
"plt.legend(['Train','Validation'])\n",
"\n",
"plt.subplot(1,2,2)\n",
"plt.plot(loss); plt.plot(val_loss)\n",
"plt.title(\"Model Loss\"); plt.xlabel(\"Epoch\"); plt.ylabel(\"Loss\")\n",
"plt.legend(['Train','Validation'])\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c763256c-2fad-44d9-97dd-15e42c956933",
"metadata": {},
"outputs": [],
"source": [
"\n",
"base_dir = r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\"\n",
"\n",
"\n",
"# 2. Load the datasets\n",
"\n",
"IMG_SIZE = (512, 512)\n",
"\n",
"BATCH_SIZE = 32\n",
"\n",
"train_ds = tf.keras.utils.image_dataset_from_directory(\n",
"\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\train\",\n",
"\n",
" image_size=IMG_SIZE,\n",
"\n",
" batch_size=BATCH_SIZE,\n",
"\n",
" label_mode=\"int\"\n",
")\n",
"\n",
" \n",
"\n",
"valid_ds = tf.keras.utils.image_dataset_from_directory(\n",
"\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\valid\",\n",
"\n",
" image_size=IMG_SIZE,\n",
"\n",
" batch_size=BATCH_SIZE,\n",
"\n",
" label_mode=\"int\"\n",
"\n",
")\n",
"\n",
" \n",
"\n",
"test_ds = tf.keras.utils.image_dataset_from_directory(\n",
"\n",
" r\"C:\\Users\\adina\\Downloads\\PlantPathology_Split\\test\",\n",
"\n",
" image_size=IMG_SIZE,\n",
"\n",
" batch_size=BATCH_SIZE,\n",
"\n",
" label_mode=\"int\",\n",
"\n",
" shuffle=False\n",
"\n",
")\n",
"\n",
" \n",
"\n",
"class_names = train_ds.class_names\n",
"\n",
"print(\"Classes:\")\n",
"\n",
"print(class_names)\n"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "21a6d324-fe59-48af-9e5c-acfd0104a584",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"Model: \"sequential_5\"\n",
"\n"
],
"text/plain": [
"\u001b[1mModel: \"sequential_5\"\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ rescaling_5 (Rescaling) │ (None, 224, 224, 3) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_7 (Conv2D) │ (None, 222, 222, 32) │ 896 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_7 (MaxPooling2D) │ (None, 111, 111, 32) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_8 (Conv2D) │ (None, 109, 109, 64) │ 18,496 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_8 (MaxPooling2D) │ (None, 54, 54, 64) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten_2 (Flatten) │ (None, 186624) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_7 (Dense) │ (None, 128) │ 23,888,000 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_4 (Dropout) │ (None, 128) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_8 (Dense) │ (None, 12) │ 1,548 │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"\n"
],
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ rescaling_5 (\u001b[38;5;33mRescaling\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_7 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m222\u001b[0m, \u001b[38;5;34m222\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m896\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_7 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m111\u001b[0m, \u001b[38;5;34m111\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_8 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m109\u001b[0m, \u001b[38;5;34m109\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m18,496\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_8 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m54\u001b[0m, \u001b[38;5;34m54\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten_2 (\u001b[38;5;33mFlatten\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m186624\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_7 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m23,888,000\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_4 (\u001b[38;5;33mDropout\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_8 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m) │ \u001b[38;5;34m1,548\u001b[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Total params: 23,908,940 (91.21 MB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m23,908,940\u001b[0m (91.21 MB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Trainable params: 23,908,940 (91.21 MB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m23,908,940\u001b[0m (91.21 MB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Non-trainable params: 0 (0.00 B)\n",
"\n"
],
"text/plain": [
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1539s/step - accuracy: 0.3140 - loss: 1.9188 \n",
"Epoch 1: val_accuracy improved from None to 0.40602, saving model to best_model.keras\n",
"\n",
"Epoch 1: finished saving model to best_model.keras\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m416995s\u001b[0m 1539s/step - accuracy: 0.3140 - loss: 1.9188 - val_accuracy: 0.4060 - val_loss: 1.7117\n",
"Epoch 2/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 578ms/step - accuracy: 0.4043 - loss: 1.6560\n",
"Epoch 2: val_accuracy improved from 0.40602 to 0.44361, saving model to best_model.keras\n",
"\n",
"Epoch 2: finished saving model to best_model.keras\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m166s\u001b[0m 609ms/step - accuracy: 0.4043 - loss: 1.6560 - val_accuracy: 0.4436 - val_loss: 1.5466\n",
"Epoch 3/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 618ms/step - accuracy: 0.4524 - loss: 1.5200\n",
"Epoch 3: val_accuracy improved from 0.44361 to 0.46724, saving model to best_model.keras\n",
"\n",
"Epoch 3: finished saving model to best_model.keras\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m177s\u001b[0m 649ms/step - accuracy: 0.4524 - loss: 1.5200 - val_accuracy: 0.4672 - val_loss: 1.5130\n",
"Epoch 4/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 576ms/step - accuracy: 0.5131 - loss: 1.3654\n",
"Epoch 4: val_accuracy improved from 0.46724 to 0.48657, saving model to best_model.keras\n",
"\n",
"Epoch 4: finished saving model to best_model.keras\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m166s\u001b[0m 609ms/step - accuracy: 0.5131 - loss: 1.3654 - val_accuracy: 0.4866 - val_loss: 1.5051\n",
"Epoch 5/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 581ms/step - accuracy: 0.5652 - loss: 1.2228\n",
"Epoch 5: val_accuracy did not improve from 0.48657\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m165s\u001b[0m 605ms/step - accuracy: 0.5652 - loss: 1.2228 - val_accuracy: 0.4748 - val_loss: 1.4716\n",
"Epoch 6/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 571ms/step - accuracy: 0.6186 - loss: 1.0616\n",
"Epoch 6: val_accuracy improved from 0.48657 to 0.50107, saving model to best_model.keras\n",
"\n",
"Epoch 6: finished saving model to best_model.keras\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m164s\u001b[0m 603ms/step - accuracy: 0.6186 - loss: 1.0616 - val_accuracy: 0.5011 - val_loss: 1.5838\n",
"Epoch 7/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 591ms/step - accuracy: 0.6739 - loss: 0.9061\n",
"Epoch 7: val_accuracy did not improve from 0.50107\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m206s\u001b[0m 618ms/step - accuracy: 0.6739 - loss: 0.9061 - val_accuracy: 0.4871 - val_loss: 1.6724\n",
"Epoch 8/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 602ms/step - accuracy: 0.7109 - loss: 0.7874\n",
"Epoch 8: val_accuracy improved from 0.50107 to 0.50376, saving model to best_model.keras\n",
"\n",
"Epoch 8: finished saving model to best_model.keras\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m173s\u001b[0m 636ms/step - accuracy: 0.7109 - loss: 0.7874 - val_accuracy: 0.5038 - val_loss: 1.8173\n",
"Epoch 9/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 580ms/step - accuracy: 0.7572 - loss: 0.6749\n",
"Epoch 9: val_accuracy did not improve from 0.50376\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m165s\u001b[0m 605ms/step - accuracy: 0.7572 - loss: 0.6749 - val_accuracy: 0.5038 - val_loss: 1.9628\n",
"Epoch 10/10\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 595ms/step - accuracy: 0.7881 - loss: 0.5950\n",
"Epoch 10: val_accuracy did not improve from 0.50376\n",
"\u001b[1m272/272\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m169s\u001b[0m 620ms/step - accuracy: 0.7881 - loss: 0.5950 - val_accuracy: 0.4694 - val_loss: 2.1204\n",
"Restoring model weights from the end of the best epoch: 8.\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[1m59/59\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 116ms/step - accuracy: 0.5013 - loss: 1.7312\n",
"\n",
"Test Accuracy: 0.5013\n",
"Test Loss: 1.7312\n"
]
},
{
"data": {
"text/html": [
"Model: \"sequential_5\"\n",
"\n"
],
"text/plain": [
"\u001b[1mModel: \"sequential_5\"\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ rescaling_5 (Rescaling) │ (None, 224, 224, 3) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_7 (Conv2D) │ (None, 222, 222, 32) │ 896 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_7 (MaxPooling2D) │ (None, 111, 111, 32) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_8 (Conv2D) │ (None, 109, 109, 64) │ 18,496 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_8 (MaxPooling2D) │ (None, 54, 54, 64) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten_2 (Flatten) │ (None, 186624) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_7 (Dense) │ (None, 128) │ 23,888,000 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_4 (Dropout) │ (None, 128) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_8 (Dense) │ (None, 12) │ 1,548 │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"\n"
],
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ rescaling_5 (\u001b[38;5;33mRescaling\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_7 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m222\u001b[0m, \u001b[38;5;34m222\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m896\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_7 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m111\u001b[0m, \u001b[38;5;34m111\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_8 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m109\u001b[0m, \u001b[38;5;34m109\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m18,496\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_8 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m54\u001b[0m, \u001b[38;5;34m54\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten_2 (\u001b[38;5;33mFlatten\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m186624\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_7 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m23,888,000\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_4 (\u001b[38;5;33mDropout\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_8 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m) │ \u001b[38;5;34m1,548\u001b[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Total params: 71,726,822 (273.62 MB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m71,726,822\u001b[0m (273.62 MB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Trainable params: 23,908,940 (91.21 MB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m23,908,940\u001b[0m (91.21 MB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Non-trainable params: 0 (0.00 B)\n",
"\n"
],
"text/plain": [
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Optimizer params: 47,817,882 (182.41 MB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Optimizer params: \u001b[0m\u001b[38;5;34m47,817,882\u001b[0m (182.41 MB)\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import tensorflow as tf\n",
"from tensorflow.keras import layers, models\n",
"from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
"import matplotlib.pyplot as plt\n",
"\n",
"# 3. Optimise data loading, using a TensorFlow tool\n",
"AUTOTUNE = tf.data.AUTOTUNE\n",
"train_ds = train_ds.prefetch(AUTOTUNE)\n",
"valid_ds = valid_ds.prefetch(AUTOTUNE)\n",
"test_ds = test_ds.prefetch(AUTOTUNE)\n",
"\n",
"# 4. Build a basic CNN\n",
"num_classes = len(class_names)\n",
"model = models.Sequential([\n",
" layers.Rescaling(1./255, input_shape=(224, 224, 3)), # was (512, 512, 3)\n",
" layers.Conv2D(32, 3, activation='relu'),\n",
" layers.MaxPooling2D(),\n",
"\n",
" layers.Conv2D(64, 3, activation='relu'),\n",
" layers.MaxPooling2D(),\n",
" layers.Flatten(),\n",
" layers.Dense(128, activation='relu'),\n",
" layers.Dropout(0.5),\n",
" layers.Dense(num_classes, activation='softmax')\n",
"])\n",
"model.summary()\n",
"\n",
"# 5. Compile the model\n",
"model.compile(\n",
" optimizer='adam',\n",
" loss='sparse_categorical_crossentropy',\n",
" metrics=['accuracy']\n",
")\n",
"\n",
"# 6. Set up callbacks to capture the BEST model, not just the last epoch\n",
"early_stop = EarlyStopping(\n",
" monitor='val_accuracy',\n",
" patience=3, # stop if no improvement for 3 epochs\n",
" restore_best_weights=True, # roll back to best epoch's weights\n",
" verbose=1\n",
")\n",
"\n",
"checkpoint = ModelCheckpoint(\n",
" 'best_model.keras',\n",
" monitor='val_accuracy',\n",
" save_best_only=True,\n",
" verbose=1\n",
")\n",
"\n",
"# 7. Train the model\n",
"history = model.fit(\n",
" train_ds,\n",
" validation_data=valid_ds,\n",
" epochs=10,\n",
" callbacks=[early_stop, checkpoint]\n",
")\n",
"\n",
"# 8. Plot accuracy and loss curves\n",
"plt.figure(figsize=(12,5))\n",
"\n",
"# Accuracy\n",
"plt.subplot(1,2,1)\n",
"plt.plot(history.history['accuracy'])\n",
"plt.plot(history.history['val_accuracy'])\n",
"plt.title(\"Model Accuracy\")\n",
"plt.xlabel(\"Epoch\")\n",
"plt.ylabel(\"Accuracy\")\n",
"plt.legend(['Train','Validation'])\n",
"\n",
"# Loss\n",
"plt.subplot(1,2,2)\n",
"plt.plot(history.history['loss'])\n",
"plt.plot(history.history['val_loss'])\n",
"plt.title(\"Model Loss\")\n",
"plt.xlabel(\"Epoch\")\n",
"plt.ylabel(\"Loss\")\n",
"plt.legend(['Train','Validation'])\n",
"plt.tight_layout()\n",
"plt.show()\n",
"\n",
"# 9. Evaluate on the test set (using the restored best weights)\n",
"test_loss, test_accuracy = model.evaluate(test_ds)\n",
"print(f\"\\nTest Accuracy: {test_accuracy:.4f}\")\n",
"print(f\"Test Loss: {test_loss:.4f}\")\n",
"model.summary()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "09520585-24c4-456a-9054-56a44f3184d8",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python [conda env:base] *",
"language": "python",
"name": "conda-base-py"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.13.9"
}
},
"nbformat": 4,
"nbformat_minor": 5
}