diff --git a/labworks/LW1/5-90.png b/labworks/LW1/5-90.png
new file mode 100644
index 0000000..4c7b244
Binary files /dev/null and b/labworks/LW1/5-90.png differ
diff --git a/labworks/LW1/5.png b/labworks/LW1/5.png
new file mode 100644
index 0000000..b7dea6f
Binary files /dev/null and b/labworks/LW1/5.png differ
diff --git a/labworks/LW1/IS_LR1.ipynb b/labworks/LW1/IS_LR1.ipynb
new file mode 100644
index 0000000..49b9227
--- /dev/null
+++ b/labworks/LW1/IS_LR1.ipynb
@@ -0,0 +1 @@
+{"cells":[{"cell_type":"code","execution_count":1,"metadata":{"executionInfo":{"elapsed":28,"status":"ok","timestamp":1758484819406,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"_h4DGjN7sHZa"},"outputs":[],"source":["import os\n","os.chdir('/content/drive/MyDrive/Colab Notebooks')"]},{"cell_type":"code","execution_count":2,"metadata":{"executionInfo":{"elapsed":8502,"status":"ok","timestamp":1758484829283,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"dyr70xmcsXQU"},"outputs":[],"source":["from tensorflow import keras\n","import matplotlib.pyplot as plt\n","import numpy as np\n","import sklearn"]},{"cell_type":"code","execution_count":3,"metadata":{"executionInfo":{"elapsed":403,"status":"ok","timestamp":1758484830646,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"1-4Nf4M-spPi"},"outputs":[],"source":["from keras.datasets import mnist\n","(X_train, y_train), (X_test, y_test) = mnist.load_data()"]},{"cell_type":"code","execution_count":4,"metadata":{"executionInfo":{"elapsed":82,"status":"ok","timestamp":1758484836270,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"Y0-OHY-GstmN"},"outputs":[],"source":["from sklearn.model_selection import train_test_split"]},{"cell_type":"code","execution_count":5,"metadata":{"executionInfo":{"elapsed":64,"status":"ok","timestamp":1758484837665,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"tmaCmlbdsw01"},"outputs":[],"source":["X = np.concatenate((X_train, X_test))\n","y = np.concatenate((y_train, y_test))"]},{"cell_type":"code","execution_count":6,"metadata":{"executionInfo":{"elapsed":60,"status":"ok","timestamp":1758484838516,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"7i0LOumLszkn"},"outputs":[],"source":["X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 10000, train_size = 60000, random_state = 7)"]},{"cell_type":"code","execution_count":7,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":31,"status":"ok","timestamp":1758484839301,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"pSHJE5y-tCaS","outputId":"3688b968-0b01-4abd-e310-682ea55d7db0"},"outputs":[{"output_type":"stream","name":"stdout","text":["Shape of X train: (60000, 28, 28)\n","Shape of y train: (60000,)\n"]}],"source":["print('Shape of X train:', X_train.shape)\n","print('Shape of y train:', y_train.shape)"]},{"cell_type":"code","execution_count":8,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":251},"executionInfo":{"elapsed":479,"status":"ok","timestamp":1758484841128,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"x5Aki0dYu9k8","outputId":"c55ee86d-572a-4736-8c40-e4063613ffc6"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1000x300 with 4 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["fig, axes = plt.subplots(1, 4, figsize=(10, 3))\n","\n","for i in range(4):\n","    axes[i].imshow(X_train[i], cmap=plt.get_cmap('gray'))\n","    axes[i].set_title(f'Label: {y_train[i]}')\n","\n","plt.show()"]},{"cell_type":"code","execution_count":9,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":245,"status":"ok","timestamp":1758484843822,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"CGOWqtNUz_VP","outputId":"7fabcc62-6138-43ad-f097-bd5104be8a67"},"outputs":[{"output_type":"stream","name":"stdout","text":["Shape of transformed X train: (60000, 784)\n"]}],"source":["num_pixels = X_train.shape[1] * X_train.shape[2]\n","X_train = X_train.reshape(X_train.shape[0], num_pixels) / 255\n","X_test = X_test.reshape(X_test.shape[0], num_pixels) / 255\n","print('Shape of transformed X train:', X_train.shape)"]},{"cell_type":"code","execution_count":10,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":63,"status":"ok","timestamp":1758484845939,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"GnlaKGw11f0w","outputId":"bc8532d6-7088-4140-80a9-0ed8ea65983e"},"outputs":[{"output_type":"stream","name":"stdout","text":["Shape of transformed y train: (60000, 10)\n"]}],"source":["from keras.utils import to_categorical\n","y_train = to_categorical(y_train)\n","y_test = to_categorical(y_test)\n","print('Shape of transformed y train:', y_train.shape)\n","num_classes = y_train.shape[1]"]},{"cell_type":"code","execution_count":11,"metadata":{"executionInfo":{"elapsed":4,"status":"ok","timestamp":1758484847258,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"JrUiXnVX4h7y"},"outputs":[],"source":["from keras.models import Sequential\n","from keras.layers import Dense"]},{"cell_type":"code","execution_count":13,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":170},"executionInfo":{"elapsed":150,"status":"ok","timestamp":1758484854548,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"G8M-v6-G3378","outputId":"670fca24-77ad-4330-eae7-4f0fb9781b6d"},"outputs":[{"output_type":"display_data","data":{"text/plain":["\u001b[1mModel: \"sequential_1\"\u001b[0m\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_1\"</span>\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"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","│ dense_1 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)             │         \u001b[38;5;34m7,850\u001b[0m │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n","┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n","│ dense_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)             │         <span style=\"color: #00af00; text-decoration-color: #00af00\">7,850</span> │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Total params: \u001b[0m\u001b[38;5;34m7,850\u001b[0m (30.66 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">7,850</span> (30.66 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m7,850\u001b[0m (30.66 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">7,850</span> (30.66 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n","</pre>\n"]},"metadata":{}}],"source":["model_01 = Sequential()\n","model_01.add(Dense(units=num_classes,input_dim=num_pixels, activation='softmax'))\n","model_01.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])\n","model_01.summary()\n"]},{"cell_type":"code","execution_count":14,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":60153,"status":"ok","timestamp":1758484917485,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"ut479Pn87OSB","outputId":"89a4d4ab-bdef-4c7f-c742-3caa5cbaff88"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 20ms/step - accuracy: 0.2095 - loss: 2.2063 - val_accuracy: 0.6653 - val_loss: 1.5891\n","Epoch 2/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.6920 - loss: 1.4764 - val_accuracy: 0.7613 - val_loss: 1.1972\n","Epoch 3/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7728 - loss: 1.1430 - val_accuracy: 0.7987 - val_loss: 0.9912\n","Epoch 4/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8014 - loss: 0.9647 - val_accuracy: 0.8177 - val_loss: 0.8671\n","Epoch 5/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8211 - loss: 0.8510 - val_accuracy: 0.8283 - val_loss: 0.7843\n","Epoch 6/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8318 - loss: 0.7777 - val_accuracy: 0.8390 - val_loss: 0.7248\n","Epoch 7/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8396 - loss: 0.7273 - val_accuracy: 0.8442 - val_loss: 0.6802\n","Epoch 8/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8461 - loss: 0.6806 - val_accuracy: 0.8497 - val_loss: 0.6450\n","Epoch 9/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8508 - loss: 0.6451 - val_accuracy: 0.8550 - val_loss: 0.6166\n","Epoch 10/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8543 - loss: 0.6222 - val_accuracy: 0.8587 - val_loss: 0.5931\n","Epoch 11/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8585 - loss: 0.5973 - val_accuracy: 0.8617 - val_loss: 0.5732\n","Epoch 12/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8626 - loss: 0.5734 - val_accuracy: 0.8660 - val_loss: 0.5562\n","Epoch 13/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8643 - loss: 0.5583 - val_accuracy: 0.8682 - val_loss: 0.5415\n","Epoch 14/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8670 - loss: 0.5490 - val_accuracy: 0.8715 - val_loss: 0.5286\n","Epoch 15/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8682 - loss: 0.5379 - val_accuracy: 0.8733 - val_loss: 0.5171\n","Epoch 16/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8707 - loss: 0.5242 - val_accuracy: 0.8753 - val_loss: 0.5068\n","Epoch 17/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8712 - loss: 0.5152 - val_accuracy: 0.8767 - val_loss: 0.4976\n","Epoch 18/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8739 - loss: 0.5033 - val_accuracy: 0.8768 - val_loss: 0.4892\n","Epoch 19/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8754 - loss: 0.4947 - val_accuracy: 0.8783 - val_loss: 0.4816\n","Epoch 20/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8790 - loss: 0.4828 - val_accuracy: 0.8792 - val_loss: 0.4745\n","Epoch 21/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8787 - loss: 0.4765 - val_accuracy: 0.8812 - val_loss: 0.4681\n","Epoch 22/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8805 - loss: 0.4713 - val_accuracy: 0.8823 - val_loss: 0.4622\n","Epoch 23/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8799 - loss: 0.4695 - val_accuracy: 0.8830 - val_loss: 0.4566\n","Epoch 24/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8828 - loss: 0.4591 - val_accuracy: 0.8832 - val_loss: 0.4515\n","Epoch 25/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8808 - loss: 0.4615 - val_accuracy: 0.8847 - val_loss: 0.4467\n","Epoch 26/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 7ms/step - accuracy: 0.8831 - loss: 0.4495 - val_accuracy: 0.8862 - val_loss: 0.4422\n","Epoch 27/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 12ms/step - accuracy: 0.8812 - loss: 0.4527 - val_accuracy: 0.8867 - val_loss: 0.4379\n","Epoch 28/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - accuracy: 0.8831 - loss: 0.4480 - val_accuracy: 0.8868 - val_loss: 0.4339\n","Epoch 29/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 8ms/step - accuracy: 0.8845 - loss: 0.4422 - val_accuracy: 0.8883 - val_loss: 0.4301\n","Epoch 30/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 7ms/step - accuracy: 0.8870 - loss: 0.4303 - val_accuracy: 0.8882 - val_loss: 0.4266\n","Epoch 31/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8880 - loss: 0.4299 - val_accuracy: 0.8888 - val_loss: 0.4232\n","Epoch 32/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8853 - loss: 0.4295 - val_accuracy: 0.8892 - val_loss: 0.4200\n","Epoch 33/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8874 - loss: 0.4265 - val_accuracy: 0.8902 - val_loss: 0.4170\n","Epoch 34/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8889 - loss: 0.4224 - val_accuracy: 0.8903 - val_loss: 0.4141\n","Epoch 35/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8898 - loss: 0.4177 - val_accuracy: 0.8912 - val_loss: 0.4113\n","Epoch 36/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8893 - loss: 0.4161 - val_accuracy: 0.8927 - val_loss: 0.4086\n","Epoch 37/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8871 - loss: 0.4173 - val_accuracy: 0.8927 - val_loss: 0.4061\n","Epoch 38/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8894 - loss: 0.4145 - val_accuracy: 0.8932 - val_loss: 0.4037\n","Epoch 39/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8911 - loss: 0.4061 - val_accuracy: 0.8932 - val_loss: 0.4014\n","Epoch 40/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8889 - loss: 0.4107 - val_accuracy: 0.8938 - val_loss: 0.3992\n","Epoch 41/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8902 - loss: 0.4036 - val_accuracy: 0.8937 - val_loss: 0.3970\n","Epoch 42/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8920 - loss: 0.4016 - val_accuracy: 0.8948 - val_loss: 0.3949\n","Epoch 43/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8933 - loss: 0.3972 - val_accuracy: 0.8950 - val_loss: 0.3930\n","Epoch 44/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8935 - loss: 0.4007 - val_accuracy: 0.8952 - val_loss: 0.3910\n","Epoch 45/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8944 - loss: 0.3934 - val_accuracy: 0.8958 - val_loss: 0.3892\n","Epoch 46/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8913 - loss: 0.4002 - val_accuracy: 0.8960 - val_loss: 0.3874\n","Epoch 47/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8919 - loss: 0.3979 - val_accuracy: 0.8965 - val_loss: 0.3857\n","Epoch 48/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8914 - loss: 0.3920 - val_accuracy: 0.8965 - val_loss: 0.3840\n","Epoch 49/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8940 - loss: 0.3909 - val_accuracy: 0.8968 - val_loss: 0.3824\n","Epoch 50/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8949 - loss: 0.3865 - val_accuracy: 0.8968 - val_loss: 0.3808\n","Epoch 51/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8951 - loss: 0.3862 - val_accuracy: 0.8970 - val_loss: 0.3793\n","Epoch 52/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8925 - loss: 0.3961 - val_accuracy: 0.8975 - val_loss: 0.3779\n","Epoch 53/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8960 - loss: 0.3798 - val_accuracy: 0.8978 - val_loss: 0.3765\n","Epoch 54/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8967 - loss: 0.3809 - val_accuracy: 0.8985 - val_loss: 0.3751\n","Epoch 55/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8967 - loss: 0.3788 - val_accuracy: 0.8988 - val_loss: 0.3737\n","Epoch 56/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8952 - loss: 0.3793 - val_accuracy: 0.8987 - val_loss: 0.3724\n","Epoch 57/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8937 - loss: 0.3808 - val_accuracy: 0.8987 - val_loss: 0.3712\n","Epoch 58/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8994 - loss: 0.3762 - val_accuracy: 0.8990 - val_loss: 0.3700\n","Epoch 59/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8941 - loss: 0.3793 - val_accuracy: 0.8997 - val_loss: 0.3688\n","Epoch 60/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8965 - loss: 0.3778 - val_accuracy: 0.8992 - val_loss: 0.3676\n","Epoch 61/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8972 - loss: 0.3743 - val_accuracy: 0.9000 - val_loss: 0.3664\n","Epoch 62/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8986 - loss: 0.3720 - val_accuracy: 0.8993 - val_loss: 0.3653\n","Epoch 63/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8992 - loss: 0.3693 - val_accuracy: 0.8995 - val_loss: 0.3643\n","Epoch 64/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9013 - loss: 0.3642 - val_accuracy: 0.9003 - val_loss: 0.3632\n","Epoch 65/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8990 - loss: 0.3690 - val_accuracy: 0.9008 - val_loss: 0.3622\n","Epoch 66/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8982 - loss: 0.3755 - val_accuracy: 0.9012 - val_loss: 0.3612\n","Epoch 67/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.9025 - loss: 0.3612 - val_accuracy: 0.9015 - val_loss: 0.3602\n","Epoch 68/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8996 - loss: 0.3693 - val_accuracy: 0.9022 - val_loss: 0.3592\n","Epoch 69/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.9001 - loss: 0.3653 - val_accuracy: 0.9025 - val_loss: 0.3583\n","Epoch 70/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8981 - loss: 0.3681 - val_accuracy: 0.9027 - val_loss: 0.3574\n","Epoch 71/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8998 - loss: 0.3668 - val_accuracy: 0.9027 - val_loss: 0.3565\n","Epoch 72/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8971 - loss: 0.3674 - val_accuracy: 0.9035 - val_loss: 0.3556\n","Epoch 73/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9010 - loss: 0.3587 - val_accuracy: 0.9038 - val_loss: 0.3548\n","Epoch 74/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9005 - loss: 0.3586 - val_accuracy: 0.9037 - val_loss: 0.3540\n","Epoch 75/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9006 - loss: 0.3586 - val_accuracy: 0.9042 - val_loss: 0.3531\n","Epoch 76/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9012 - loss: 0.3622 - val_accuracy: 0.9043 - val_loss: 0.3523\n","Epoch 77/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9017 - loss: 0.3592 - val_accuracy: 0.9045 - val_loss: 0.3516\n","Epoch 78/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8973 - loss: 0.3651 - val_accuracy: 0.9047 - val_loss: 0.3508\n","Epoch 79/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.9019 - loss: 0.3582 - val_accuracy: 0.9053 - val_loss: 0.3500\n","Epoch 80/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9030 - loss: 0.3522 - val_accuracy: 0.9053 - val_loss: 0.3493\n","Epoch 81/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9038 - loss: 0.3513 - val_accuracy: 0.9053 - val_loss: 0.3485\n","Epoch 82/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9009 - loss: 0.3582 - val_accuracy: 0.9053 - val_loss: 0.3479\n","Epoch 83/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9022 - loss: 0.3518 - val_accuracy: 0.9050 - val_loss: 0.3472\n","Epoch 84/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9016 - loss: 0.3538 - val_accuracy: 0.9053 - val_loss: 0.3465\n","Epoch 85/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9033 - loss: 0.3485 - val_accuracy: 0.9052 - val_loss: 0.3458\n","Epoch 86/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9025 - loss: 0.3505 - val_accuracy: 0.9057 - val_loss: 0.3451\n","Epoch 87/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.9023 - loss: 0.3536 - val_accuracy: 0.9058 - val_loss: 0.3445\n","Epoch 88/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.9038 - loss: 0.3499 - val_accuracy: 0.9057 - val_loss: 0.3438\n","Epoch 89/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 7ms/step - accuracy: 0.9030 - loss: 0.3479 - val_accuracy: 0.9057 - val_loss: 0.3433\n","Epoch 90/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 8ms/step - accuracy: 0.9039 - loss: 0.3473 - val_accuracy: 0.9058 - val_loss: 0.3426\n","Epoch 91/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.9029 - loss: 0.3489 - val_accuracy: 0.9057 - val_loss: 0.3420\n","Epoch 92/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.9023 - loss: 0.3500 - val_accuracy: 0.9057 - val_loss: 0.3414\n","Epoch 93/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.9031 - loss: 0.3477 - val_accuracy: 0.9060 - val_loss: 0.3408\n","Epoch 94/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9052 - loss: 0.3436 - val_accuracy: 0.9065 - val_loss: 0.3403\n","Epoch 95/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.9057 - loss: 0.3427 - val_accuracy: 0.9068 - val_loss: 0.3397\n","Epoch 96/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9030 - loss: 0.3457 - val_accuracy: 0.9068 - val_loss: 0.3392\n","Epoch 97/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.9044 - loss: 0.3381 - val_accuracy: 0.9068 - val_loss: 0.3386\n","Epoch 98/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9027 - loss: 0.3466 - val_accuracy: 0.9072 - val_loss: 0.3381\n","Epoch 99/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.9050 - loss: 0.3393 - val_accuracy: 0.9072 - val_loss: 0.3376\n","Epoch 100/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9071 - loss: 0.3384 - val_accuracy: 0.9073 - val_loss: 0.3371\n"]}],"source":["H = model_01.fit(\n","    X_train, y_train,\n","    validation_split=0.1,\n","    epochs=100,\n","    batch_size = 512\n",")"]},{"cell_type":"code","execution_count":15,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":487},"executionInfo":{"elapsed":351,"status":"ok","timestamp":1758484925312,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"YfhWOnxq70K3","outputId":"45e2df4d-6cfa-4986-b3b6-1ec22201aed1"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x500 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.figure(figsize=(12, 5))\n","\n","plt.subplot(1, 2, 1)\n","plt.plot(H.history['loss'], label='Обучающая ошибка')\n","plt.plot(H.history['val_loss'], label='Валидационная ошибка')\n","plt.title('Функция ошибки по эпохам')\n","plt.xlabel('Epochs')\n","plt.ylabel('loss')\n","plt.legend()\n","plt.grid(True)"]},{"cell_type":"code","execution_count":16,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":2769,"status":"ok","timestamp":1758484931222,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"3ZJ4HMej9UBE","outputId":"bfd4d038-1ec6-4ab6-be81-7c4d1234a12f"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - accuracy: 0.9079 - loss: 0.3455\n","Loss on test data: 0.3511466085910797\n","Accuracy on test data: 0.9067999720573425\n"]}],"source":["scores=model_01.evaluate(X_test,y_test)\n","print('Loss on test data:', scores[0])\n","print('Accuracy on test data:', scores[1])"]},{"cell_type":"code","execution_count":17,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":204},"executionInfo":{"elapsed":169,"status":"ok","timestamp":1758484937403,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"Xy4I-UYP91_a","outputId":"814dcfdd-2ed4-4605-84a9-61e2ff4a0d26"},"outputs":[{"output_type":"display_data","data":{"text/plain":["\u001b[1mModel: \"sequential_2\"\u001b[0m\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_2\"</span>\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"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","│ dense_2 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m100\u001b[0m)            │        \u001b[38;5;34m78,500\u001b[0m │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_3 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)             │         \u001b[38;5;34m1,010\u001b[0m │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n","┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n","│ dense_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">100</span>)            │        <span style=\"color: #00af00; text-decoration-color: #00af00\">78,500</span> │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)             │         <span style=\"color: #00af00; text-decoration-color: #00af00\">1,010</span> │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Total params: \u001b[0m\u001b[38;5;34m79,510\u001b[0m (310.59 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">79,510</span> (310.59 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m79,510\u001b[0m (310.59 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">79,510</span> (310.59 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n","</pre>\n"]},"metadata":{}}],"source":["model_01_100 = Sequential()\n","model_01_100.add(Dense(units=100,input_dim=num_pixels, activation='sigmoid'))\n","model_01_100.add(Dense(units=num_classes, activation='softmax'))\n","model_01_100.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])\n","model_01_100.summary()"]},{"cell_type":"code","execution_count":18,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":58118,"status":"ok","timestamp":1758485009570,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"},"user_tz":-180},"id":"IfOOffSABn_u","outputId":"821867e4-5d5b-400b-e290-0910e3510b3a"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 15ms/step - accuracy: 0.1144 - loss: 2.3654 - val_accuracy: 0.3688 - val_loss: 2.1933\n","Epoch 2/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.4250 - loss: 2.1612 - val_accuracy: 0.5125 - val_loss: 2.0693\n","Epoch 3/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.5401 - loss: 2.0408 - val_accuracy: 0.5837 - val_loss: 1.9510\n","Epoch 4/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6047 - loss: 1.9234 - val_accuracy: 0.6332 - val_loss: 1.8370\n","Epoch 5/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6477 - loss: 1.8073 - val_accuracy: 0.6737 - val_loss: 1.7282\n","Epoch 6/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6824 - loss: 1.7042 - val_accuracy: 0.6938 - val_loss: 1.6254\n","Epoch 7/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7016 - loss: 1.6027 - val_accuracy: 0.7125 - val_loss: 1.5291\n","Epoch 8/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7249 - loss: 1.5062 - val_accuracy: 0.7350 - val_loss: 1.4398\n","Epoch 9/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7420 - loss: 1.4167 - val_accuracy: 0.7503 - val_loss: 1.3581\n","Epoch 10/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7577 - loss: 1.3384 - val_accuracy: 0.7622 - val_loss: 1.2836\n","Epoch 11/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7648 - loss: 1.2675 - val_accuracy: 0.7778 - val_loss: 1.2161\n","Epoch 12/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7779 - loss: 1.2033 - val_accuracy: 0.7828 - val_loss: 1.1551\n","Epoch 13/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7861 - loss: 1.1414 - val_accuracy: 0.7915 - val_loss: 1.0999\n","Epoch 14/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.7918 - loss: 1.0921 - val_accuracy: 0.7960 - val_loss: 1.0504\n","Epoch 15/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.7972 - loss: 1.0415 - val_accuracy: 0.8032 - val_loss: 1.0054\n","Epoch 16/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8019 - loss: 1.0022 - val_accuracy: 0.8088 - val_loss: 0.9647\n","Epoch 17/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8095 - loss: 0.9625 - val_accuracy: 0.8145 - val_loss: 0.9277\n","Epoch 18/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8178 - loss: 0.9236 - val_accuracy: 0.8203 - val_loss: 0.8941\n","Epoch 19/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8203 - loss: 0.8889 - val_accuracy: 0.8260 - val_loss: 0.8635\n","Epoch 20/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8221 - loss: 0.8632 - val_accuracy: 0.8298 - val_loss: 0.8356\n","Epoch 21/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8258 - loss: 0.8351 - val_accuracy: 0.8335 - val_loss: 0.8099\n","Epoch 22/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8318 - loss: 0.8071 - val_accuracy: 0.8363 - val_loss: 0.7863\n","Epoch 23/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8335 - loss: 0.7909 - val_accuracy: 0.8380 - val_loss: 0.7646\n","Epoch 24/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8362 - loss: 0.7637 - val_accuracy: 0.8407 - val_loss: 0.7445\n","Epoch 25/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8395 - loss: 0.7427 - val_accuracy: 0.8448 - val_loss: 0.7258\n","Epoch 26/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8406 - loss: 0.7290 - val_accuracy: 0.8463 - val_loss: 0.7085\n","Epoch 27/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8455 - loss: 0.7081 - val_accuracy: 0.8498 - val_loss: 0.6924\n","Epoch 28/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8491 - loss: 0.6921 - val_accuracy: 0.8515 - val_loss: 0.6773\n","Epoch 29/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8484 - loss: 0.6778 - val_accuracy: 0.8533 - val_loss: 0.6634\n","Epoch 30/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8515 - loss: 0.6648 - val_accuracy: 0.8563 - val_loss: 0.6501\n","Epoch 31/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8505 - loss: 0.6549 - val_accuracy: 0.8575 - val_loss: 0.6377\n","Epoch 32/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8557 - loss: 0.6389 - val_accuracy: 0.8587 - val_loss: 0.6261\n","Epoch 33/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8562 - loss: 0.6298 - val_accuracy: 0.8607 - val_loss: 0.6150\n","Epoch 34/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8590 - loss: 0.6148 - val_accuracy: 0.8612 - val_loss: 0.6047\n","Epoch 35/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 10ms/step - accuracy: 0.8597 - loss: 0.6065 - val_accuracy: 0.8630 - val_loss: 0.5949\n","Epoch 36/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8602 - loss: 0.5975 - val_accuracy: 0.8652 - val_loss: 0.5856\n","Epoch 37/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8623 - loss: 0.5877 - val_accuracy: 0.8675 - val_loss: 0.5768\n","Epoch 38/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8639 - loss: 0.5819 - val_accuracy: 0.8703 - val_loss: 0.5683\n","Epoch 39/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8657 - loss: 0.5723 - val_accuracy: 0.8712 - val_loss: 0.5604\n","Epoch 40/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8644 - loss: 0.5693 - val_accuracy: 0.8715 - val_loss: 0.5528\n","Epoch 41/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8694 - loss: 0.5554 - val_accuracy: 0.8742 - val_loss: 0.5456\n","Epoch 42/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8707 - loss: 0.5489 - val_accuracy: 0.8738 - val_loss: 0.5387\n","Epoch 43/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8680 - loss: 0.5462 - val_accuracy: 0.8745 - val_loss: 0.5321\n","Epoch 44/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8710 - loss: 0.5413 - val_accuracy: 0.8758 - val_loss: 0.5257\n","Epoch 45/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8741 - loss: 0.5287 - val_accuracy: 0.8753 - val_loss: 0.5198\n","Epoch 46/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8735 - loss: 0.5253 - val_accuracy: 0.8768 - val_loss: 0.5139\n","Epoch 47/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8760 - loss: 0.5145 - val_accuracy: 0.8782 - val_loss: 0.5085\n","Epoch 48/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8737 - loss: 0.5136 - val_accuracy: 0.8783 - val_loss: 0.5031\n","Epoch 49/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8758 - loss: 0.5064 - val_accuracy: 0.8792 - val_loss: 0.4979\n","Epoch 50/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8763 - loss: 0.5030 - val_accuracy: 0.8800 - val_loss: 0.4930\n","Epoch 51/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8776 - loss: 0.4976 - val_accuracy: 0.8812 - val_loss: 0.4884\n","Epoch 52/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8780 - loss: 0.4931 - val_accuracy: 0.8817 - val_loss: 0.4837\n","Epoch 53/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8770 - loss: 0.4895 - val_accuracy: 0.8827 - val_loss: 0.4793\n","Epoch 54/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8774 - loss: 0.4899 - val_accuracy: 0.8827 - val_loss: 0.4752\n","Epoch 55/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8791 - loss: 0.4836 - val_accuracy: 0.8832 - val_loss: 0.4710\n","Epoch 56/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8800 - loss: 0.4794 - val_accuracy: 0.8835 - val_loss: 0.4671\n","Epoch 57/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8793 - loss: 0.4749 - val_accuracy: 0.8840 - val_loss: 0.4633\n","Epoch 58/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8813 - loss: 0.4680 - val_accuracy: 0.8845 - val_loss: 0.4596\n","Epoch 59/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8820 - loss: 0.4681 - val_accuracy: 0.8855 - val_loss: 0.4561\n","Epoch 60/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8833 - loss: 0.4603 - val_accuracy: 0.8860 - val_loss: 0.4526\n","Epoch 61/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8844 - loss: 0.4572 - val_accuracy: 0.8870 - val_loss: 0.4493\n","Epoch 62/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8832 - loss: 0.4597 - val_accuracy: 0.8875 - val_loss: 0.4461\n","Epoch 63/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8853 - loss: 0.4462 - val_accuracy: 0.8877 - val_loss: 0.4429\n","Epoch 64/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8835 - loss: 0.4553 - val_accuracy: 0.8885 - val_loss: 0.4399\n","Epoch 65/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8836 - loss: 0.4501 - val_accuracy: 0.8888 - val_loss: 0.4370\n","Epoch 66/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8866 - loss: 0.4395 - val_accuracy: 0.8887 - val_loss: 0.4342\n","Epoch 67/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8855 - loss: 0.4425 - val_accuracy: 0.8897 - val_loss: 0.4314\n","Epoch 68/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8869 - loss: 0.4374 - val_accuracy: 0.8903 - val_loss: 0.4287\n","Epoch 69/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8904 - loss: 0.4308 - val_accuracy: 0.8907 - val_loss: 0.4261\n","Epoch 70/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8888 - loss: 0.4320 - val_accuracy: 0.8912 - val_loss: 0.4235\n","Epoch 71/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8885 - loss: 0.4294 - val_accuracy: 0.8918 - val_loss: 0.4210\n","Epoch 72/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8880 - loss: 0.4278 - val_accuracy: 0.8920 - val_loss: 0.4187\n","Epoch 73/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8869 - loss: 0.4253 - val_accuracy: 0.8925 - val_loss: 0.4163\n","Epoch 74/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8895 - loss: 0.4194 - val_accuracy: 0.8920 - val_loss: 0.4141\n","Epoch 75/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8882 - loss: 0.4211 - val_accuracy: 0.8930 - val_loss: 0.4118\n","Epoch 76/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8900 - loss: 0.4162 - val_accuracy: 0.8930 - val_loss: 0.4097\n","Epoch 77/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8889 - loss: 0.4184 - val_accuracy: 0.8937 - val_loss: 0.4075\n","Epoch 78/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8916 - loss: 0.4116 - val_accuracy: 0.8937 - val_loss: 0.4054\n","Epoch 79/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8895 - loss: 0.4163 - val_accuracy: 0.8948 - val_loss: 0.4035\n","Epoch 80/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8917 - loss: 0.4078 - val_accuracy: 0.8950 - val_loss: 0.4015\n","Epoch 81/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8919 - loss: 0.4042 - val_accuracy: 0.8953 - val_loss: 0.3996\n","Epoch 82/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8933 - loss: 0.4036 - val_accuracy: 0.8960 - val_loss: 0.3977\n","Epoch 83/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8926 - loss: 0.4025 - val_accuracy: 0.8960 - val_loss: 0.3959\n","Epoch 84/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8931 - loss: 0.4006 - val_accuracy: 0.8955 - val_loss: 0.3941\n","Epoch 85/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8930 - loss: 0.3955 - val_accuracy: 0.8963 - val_loss: 0.3924\n","Epoch 86/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8921 - loss: 0.3990 - val_accuracy: 0.8967 - val_loss: 0.3907\n","Epoch 87/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8923 - loss: 0.4006 - val_accuracy: 0.8970 - val_loss: 0.3890\n","Epoch 88/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8934 - loss: 0.3962 - val_accuracy: 0.8970 - val_loss: 0.3874\n","Epoch 89/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8929 - loss: 0.3946 - val_accuracy: 0.8978 - val_loss: 0.3858\n","Epoch 90/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8930 - loss: 0.3918 - val_accuracy: 0.8982 - val_loss: 0.3843\n","Epoch 91/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8957 - loss: 0.3865 - val_accuracy: 0.8987 - val_loss: 0.3827\n","Epoch 92/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8957 - loss: 0.3871 - val_accuracy: 0.8987 - val_loss: 0.3812\n","Epoch 93/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8948 - loss: 0.3862 - val_accuracy: 0.8983 - val_loss: 0.3797\n","Epoch 94/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8953 - loss: 0.3856 - val_accuracy: 0.8992 - val_loss: 0.3784\n","Epoch 95/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8948 - loss: 0.3884 - val_accuracy: 0.8997 - val_loss: 0.3769\n","Epoch 96/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8962 - loss: 0.3833 - val_accuracy: 0.8997 - val_loss: 0.3755\n","Epoch 97/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8961 - loss: 0.3814 - val_accuracy: 0.8995 - val_loss: 0.3742\n","Epoch 98/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8946 - loss: 0.3817 - val_accuracy: 0.8997 - val_loss: 0.3728\n","Epoch 99/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8993 - loss: 0.3725 - val_accuracy: 0.8998 - val_loss: 0.3716\n","Epoch 100/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8956 - loss: 0.3770 - val_accuracy: 0.9008 - val_loss: 0.3703\n"]}],"source":["H_01_100 = model_01_100.fit(\n","    X_train, y_train,\n","    validation_split=0.1,\n","    epochs=100,\n","    batch_size = 512\n",")"]},{"cell_type":"code","execution_count":19,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":487},"id":"tKjsQBv9CFvt","executionInfo":{"status":"ok","timestamp":1758485015648,"user_tz":-180,"elapsed":447,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"38ba88f1-c62f-420b-9d18-6de273e1d7cb"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x500 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.figure(figsize=(12, 5))\n","\n","plt.subplot(1, 2, 1)\n","plt.plot(H_01_100.history['loss'], label='Обучающая ошибка')\n","plt.plot(H_01_100.history['val_loss'], label='Валидационная ошибка')\n","plt.title('Функция ошибки по эпохам')\n","plt.xlabel('Epochs')\n","plt.ylabel('loss')\n","plt.legend()\n","plt.grid(True)"]},{"cell_type":"code","execution_count":20,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"EL53RhenDJwj","executionInfo":{"status":"ok","timestamp":1758485019932,"user_tz":-180,"elapsed":1457,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"56ad3805-e7af-4b2c-cb65-42bfc273817e"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8996 - loss: 0.3781\n","Loss on test data: 0.3824511766433716\n","Accuracy on test data: 0.9000999927520752\n"]}],"source":["scores_01_100=model_01_100.evaluate(X_test,y_test)\n","print('Loss on test data:', scores_01_100[0])\n","print('Accuracy on test data:', scores_01_100[1])"]},{"cell_type":"code","execution_count":21,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":204},"id":"sa85CPpiD58n","executionInfo":{"status":"ok","timestamp":1758485023364,"user_tz":-180,"elapsed":202,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"c0ccea13-86da-41fe-a37b-935119b680d6"},"outputs":[{"output_type":"display_data","data":{"text/plain":["\u001b[1mModel: \"sequential_3\"\u001b[0m\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_3\"</span>\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"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","│ dense_4 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m300\u001b[0m)            │       \u001b[38;5;34m235,500\u001b[0m │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_5 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)             │         \u001b[38;5;34m3,010\u001b[0m │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n","┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n","│ dense_4 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">300</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">235,500</span> │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_5 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)             │         <span style=\"color: #00af00; text-decoration-color: #00af00\">3,010</span> │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Total params: \u001b[0m\u001b[38;5;34m238,510\u001b[0m (931.68 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">238,510</span> (931.68 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m238,510\u001b[0m (931.68 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">238,510</span> (931.68 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n","</pre>\n"]},"metadata":{}}],"source":["model_01_300 = Sequential()\n","model_01_300.add(Dense(units=300,input_dim=num_pixels, activation='sigmoid'))\n","model_01_300.add(Dense(units=num_classes, activation='softmax'))\n","model_01_300.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])\n","model_01_300.summary()"]},{"cell_type":"code","execution_count":22,"metadata":{"id":"vVMVL1BQEn6o","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485079791,"user_tz":-180,"elapsed":54178,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"3e249870-c81f-48e9-99ca-50aad4ad42f3"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 15ms/step - accuracy: 0.1505 - loss: 2.3045 - val_accuracy: 0.4097 - val_loss: 2.1516\n","Epoch 2/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.4658 - loss: 2.1130 - val_accuracy: 0.6090 - val_loss: 2.0029\n","Epoch 3/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.6184 - loss: 1.9658 - val_accuracy: 0.6613 - val_loss: 1.8630\n","Epoch 4/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.6724 - loss: 1.8277 - val_accuracy: 0.6930 - val_loss: 1.7323\n","Epoch 5/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7051 - loss: 1.6994 - val_accuracy: 0.7148 - val_loss: 1.6098\n","Epoch 6/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7326 - loss: 1.5800 - val_accuracy: 0.7342 - val_loss: 1.4971\n","Epoch 7/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7469 - loss: 1.4727 - val_accuracy: 0.7588 - val_loss: 1.3944\n","Epoch 8/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7653 - loss: 1.3697 - val_accuracy: 0.7695 - val_loss: 1.3020\n","Epoch 9/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7743 - loss: 1.2805 - val_accuracy: 0.7807 - val_loss: 1.2195\n","Epoch 10/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7847 - loss: 1.2033 - val_accuracy: 0.7938 - val_loss: 1.1460\n","Epoch 11/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7962 - loss: 1.1317 - val_accuracy: 0.8002 - val_loss: 1.0810\n","Epoch 12/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8011 - loss: 1.0689 - val_accuracy: 0.8062 - val_loss: 1.0232\n","Epoch 13/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8078 - loss: 1.0127 - val_accuracy: 0.8147 - val_loss: 0.9722\n","Epoch 14/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8136 - loss: 0.9662 - val_accuracy: 0.8175 - val_loss: 0.9268\n","Epoch 15/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8231 - loss: 0.9161 - val_accuracy: 0.8242 - val_loss: 0.8865\n","Epoch 16/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8246 - loss: 0.8816 - val_accuracy: 0.8273 - val_loss: 0.8504\n","Epoch 17/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8300 - loss: 0.8454 - val_accuracy: 0.8343 - val_loss: 0.8180\n","Epoch 18/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8367 - loss: 0.8115 - val_accuracy: 0.8368 - val_loss: 0.7888\n","Epoch 19/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8357 - loss: 0.7875 - val_accuracy: 0.8405 - val_loss: 0.7624\n","Epoch 20/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8418 - loss: 0.7579 - val_accuracy: 0.8433 - val_loss: 0.7383\n","Epoch 21/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8453 - loss: 0.7354 - val_accuracy: 0.8465 - val_loss: 0.7163\n","Epoch 22/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8475 - loss: 0.7119 - val_accuracy: 0.8480 - val_loss: 0.6967\n","Epoch 23/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8481 - loss: 0.6960 - val_accuracy: 0.8522 - val_loss: 0.6779\n","Epoch 24/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8495 - loss: 0.6779 - val_accuracy: 0.8538 - val_loss: 0.6611\n","Epoch 25/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8523 - loss: 0.6607 - val_accuracy: 0.8553 - val_loss: 0.6455\n","Epoch 26/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8566 - loss: 0.6454 - val_accuracy: 0.8575 - val_loss: 0.6311\n","Epoch 27/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8562 - loss: 0.6301 - val_accuracy: 0.8578 - val_loss: 0.6179\n","Epoch 28/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8585 - loss: 0.6245 - val_accuracy: 0.8613 - val_loss: 0.6053\n","Epoch 29/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8580 - loss: 0.6083 - val_accuracy: 0.8628 - val_loss: 0.5934\n","Epoch 30/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8598 - loss: 0.6007 - val_accuracy: 0.8637 - val_loss: 0.5829\n","Epoch 31/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8611 - loss: 0.5887 - val_accuracy: 0.8658 - val_loss: 0.5725\n","Epoch 32/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8633 - loss: 0.5788 - val_accuracy: 0.8663 - val_loss: 0.5629\n","Epoch 33/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8652 - loss: 0.5647 - val_accuracy: 0.8685 - val_loss: 0.5539\n","Epoch 34/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8662 - loss: 0.5548 - val_accuracy: 0.8703 - val_loss: 0.5454\n","Epoch 35/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8662 - loss: 0.5513 - val_accuracy: 0.8712 - val_loss: 0.5373\n","Epoch 36/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8719 - loss: 0.5373 - val_accuracy: 0.8728 - val_loss: 0.5297\n","Epoch 37/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8695 - loss: 0.5371 - val_accuracy: 0.8727 - val_loss: 0.5227\n","Epoch 38/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8693 - loss: 0.5309 - val_accuracy: 0.8758 - val_loss: 0.5157\n","Epoch 39/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8715 - loss: 0.5210 - val_accuracy: 0.8753 - val_loss: 0.5094\n","Epoch 40/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8704 - loss: 0.5181 - val_accuracy: 0.8760 - val_loss: 0.5032\n","Epoch 41/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8729 - loss: 0.5101 - val_accuracy: 0.8780 - val_loss: 0.4974\n","Epoch 42/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8767 - loss: 0.4976 - val_accuracy: 0.8792 - val_loss: 0.4918\n","Epoch 43/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8736 - loss: 0.5012 - val_accuracy: 0.8783 - val_loss: 0.4866\n","Epoch 44/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8749 - loss: 0.4963 - val_accuracy: 0.8803 - val_loss: 0.4815\n","Epoch 45/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8769 - loss: 0.4890 - val_accuracy: 0.8817 - val_loss: 0.4767\n","Epoch 46/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8757 - loss: 0.4865 - val_accuracy: 0.8827 - val_loss: 0.4719\n","Epoch 47/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8781 - loss: 0.4813 - val_accuracy: 0.8832 - val_loss: 0.4675\n","Epoch 48/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8791 - loss: 0.4729 - val_accuracy: 0.8835 - val_loss: 0.4633\n","Epoch 49/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8798 - loss: 0.4691 - val_accuracy: 0.8847 - val_loss: 0.4592\n","Epoch 50/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8831 - loss: 0.4595 - val_accuracy: 0.8852 - val_loss: 0.4553\n","Epoch 51/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8802 - loss: 0.4627 - val_accuracy: 0.8858 - val_loss: 0.4515\n","Epoch 52/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8839 - loss: 0.4528 - val_accuracy: 0.8867 - val_loss: 0.4479\n","Epoch 53/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8836 - loss: 0.4487 - val_accuracy: 0.8865 - val_loss: 0.4446\n","Epoch 54/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8825 - loss: 0.4520 - val_accuracy: 0.8875 - val_loss: 0.4412\n","Epoch 55/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8833 - loss: 0.4510 - val_accuracy: 0.8873 - val_loss: 0.4378\n","Epoch 56/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8821 - loss: 0.4460 - val_accuracy: 0.8870 - val_loss: 0.4349\n","Epoch 57/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8861 - loss: 0.4369 - val_accuracy: 0.8875 - val_loss: 0.4318\n","Epoch 58/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8851 - loss: 0.4383 - val_accuracy: 0.8875 - val_loss: 0.4289\n","Epoch 59/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8832 - loss: 0.4400 - val_accuracy: 0.8875 - val_loss: 0.4261\n","Epoch 60/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8842 - loss: 0.4347 - val_accuracy: 0.8888 - val_loss: 0.4233\n","Epoch 61/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8846 - loss: 0.4308 - val_accuracy: 0.8890 - val_loss: 0.4206\n","Epoch 62/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8871 - loss: 0.4306 - val_accuracy: 0.8892 - val_loss: 0.4182\n","Epoch 63/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8843 - loss: 0.4278 - val_accuracy: 0.8898 - val_loss: 0.4158\n","Epoch 64/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8876 - loss: 0.4205 - val_accuracy: 0.8905 - val_loss: 0.4132\n","Epoch 65/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8884 - loss: 0.4182 - val_accuracy: 0.8905 - val_loss: 0.4109\n","Epoch 66/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8884 - loss: 0.4183 - val_accuracy: 0.8908 - val_loss: 0.4088\n","Epoch 67/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8895 - loss: 0.4148 - val_accuracy: 0.8913 - val_loss: 0.4067\n","Epoch 68/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8890 - loss: 0.4125 - val_accuracy: 0.8922 - val_loss: 0.4044\n","Epoch 69/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8903 - loss: 0.4092 - val_accuracy: 0.8920 - val_loss: 0.4025\n","Epoch 70/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8879 - loss: 0.4104 - val_accuracy: 0.8923 - val_loss: 0.4004\n","Epoch 71/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8895 - loss: 0.4051 - val_accuracy: 0.8920 - val_loss: 0.3985\n","Epoch 72/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8900 - loss: 0.4037 - val_accuracy: 0.8915 - val_loss: 0.3967\n","Epoch 73/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8907 - loss: 0.4038 - val_accuracy: 0.8918 - val_loss: 0.3948\n","Epoch 74/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8895 - loss: 0.4043 - val_accuracy: 0.8938 - val_loss: 0.3929\n","Epoch 75/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8910 - loss: 0.4012 - val_accuracy: 0.8930 - val_loss: 0.3912\n","Epoch 76/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8904 - loss: 0.4014 - val_accuracy: 0.8933 - val_loss: 0.3895\n","Epoch 77/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8909 - loss: 0.3975 - val_accuracy: 0.8945 - val_loss: 0.3879\n","Epoch 78/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8911 - loss: 0.3982 - val_accuracy: 0.8955 - val_loss: 0.3862\n","Epoch 79/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8925 - loss: 0.3950 - val_accuracy: 0.8947 - val_loss: 0.3847\n","Epoch 80/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8912 - loss: 0.3954 - val_accuracy: 0.8965 - val_loss: 0.3830\n","Epoch 81/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8921 - loss: 0.3918 - val_accuracy: 0.8968 - val_loss: 0.3816\n","Epoch 82/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8967 - loss: 0.3809 - val_accuracy: 0.8962 - val_loss: 0.3801\n","Epoch 83/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8900 - loss: 0.3933 - val_accuracy: 0.8963 - val_loss: 0.3787\n","Epoch 84/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8927 - loss: 0.3867 - val_accuracy: 0.8967 - val_loss: 0.3774\n","Epoch 85/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8946 - loss: 0.3859 - val_accuracy: 0.8963 - val_loss: 0.3760\n","Epoch 86/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8948 - loss: 0.3826 - val_accuracy: 0.8983 - val_loss: 0.3746\n","Epoch 87/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8944 - loss: 0.3795 - val_accuracy: 0.8988 - val_loss: 0.3734\n","Epoch 88/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8955 - loss: 0.3813 - val_accuracy: 0.8993 - val_loss: 0.3721\n","Epoch 89/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8955 - loss: 0.3781 - val_accuracy: 0.8992 - val_loss: 0.3709\n","Epoch 90/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8985 - loss: 0.3721 - val_accuracy: 0.8990 - val_loss: 0.3696\n","Epoch 91/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8941 - loss: 0.3830 - val_accuracy: 0.8998 - val_loss: 0.3685\n","Epoch 92/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8962 - loss: 0.3748 - val_accuracy: 0.9000 - val_loss: 0.3672\n","Epoch 93/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8956 - loss: 0.3760 - val_accuracy: 0.8997 - val_loss: 0.3661\n","Epoch 94/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8959 - loss: 0.3739 - val_accuracy: 0.9012 - val_loss: 0.3650\n","Epoch 95/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8942 - loss: 0.3770 - val_accuracy: 0.9008 - val_loss: 0.3639\n","Epoch 96/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8986 - loss: 0.3678 - val_accuracy: 0.9012 - val_loss: 0.3628\n","Epoch 97/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8963 - loss: 0.3707 - val_accuracy: 0.9010 - val_loss: 0.3619\n","Epoch 98/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8978 - loss: 0.3643 - val_accuracy: 0.9017 - val_loss: 0.3606\n","Epoch 99/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8947 - loss: 0.3761 - val_accuracy: 0.9018 - val_loss: 0.3596\n","Epoch 100/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8975 - loss: 0.3661 - val_accuracy: 0.9013 - val_loss: 0.3588\n"]}],"source":["H_01_300 = model_01_300.fit(\n","    X_train, y_train,\n","    validation_split=0.1,\n","    epochs=100,\n","    batch_size = 512\n",")"]},{"cell_type":"code","execution_count":23,"metadata":{"id":"6bOcTrCEFjct","colab":{"base_uri":"https://localhost:8080/","height":487},"executionInfo":{"status":"ok","timestamp":1758485094903,"user_tz":-180,"elapsed":264,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"e65cc5a7-c364-4fe7-b968-06a2b6451d92"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x500 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.figure(figsize=(12, 5))\n","\n","plt.subplot(1, 2, 1)\n","plt.plot(H_01_300.history['loss'], label='Обучающая ошибка')\n","plt.plot(H_01_300.history['val_loss'], label='Валидационная ошибка')\n","plt.title('Функция ошибки по эпохам')\n","plt.xlabel('Epochs')\n","plt.ylabel('loss')\n","plt.legend()\n","plt.grid(True)"]},{"cell_type":"code","execution_count":24,"metadata":{"id":"nYzwB9pGGEQD","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485098629,"user_tz":-180,"elapsed":1478,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"ba8d8490-d9df-4985-9a50-191a11bad95f"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9016 - loss: 0.3667\n","Loss on test data: 0.37091827392578125\n","Accuracy on test data: 0.9013000130653381\n"]}],"source":["scores_01_300=model_01_300.evaluate(X_test,y_test)\n","print('Loss on test data:', scores_01_300[0])\n","print('Accuracy on test data:', scores_01_300[1])"]},{"cell_type":"code","execution_count":25,"metadata":{"id":"0kOcD8BSGslt","colab":{"base_uri":"https://localhost:8080/","height":204},"executionInfo":{"status":"ok","timestamp":1758485101393,"user_tz":-180,"elapsed":203,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"a1e2896b-aa12-449a-9655-e35be88cccfa"},"outputs":[{"output_type":"display_data","data":{"text/plain":["\u001b[1mModel: \"sequential_4\"\u001b[0m\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_4\"</span>\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"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","│ dense_6 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m500\u001b[0m)            │       \u001b[38;5;34m392,500\u001b[0m │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_7 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)             │         \u001b[38;5;34m5,010\u001b[0m │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n","┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n","│ dense_6 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">500</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">392,500</span> │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_7 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)             │         <span style=\"color: #00af00; text-decoration-color: #00af00\">5,010</span> │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Total params: \u001b[0m\u001b[38;5;34m397,510\u001b[0m (1.52 MB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">397,510</span> (1.52 MB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m397,510\u001b[0m (1.52 MB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">397,510</span> (1.52 MB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n","</pre>\n"]},"metadata":{}}],"source":["model_01_500 = Sequential()\n","model_01_500.add(Dense(units=500,input_dim=num_pixels, activation='sigmoid'))\n","model_01_500.add(Dense(units=num_classes, activation='softmax'))\n","model_01_500.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])\n","model_01_500.summary()"]},{"cell_type":"code","execution_count":26,"metadata":{"id":"LxLlRJdkH0zB","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485160222,"user_tz":-180,"elapsed":55645,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"72cd15e1-300e-4c6f-bc80-05a30c1d243a"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 22ms/step - accuracy: 0.2209 - loss: 2.2701 - val_accuracy: 0.4380 - val_loss: 2.1357\n","Epoch 2/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.5175 - loss: 2.0961 - val_accuracy: 0.5918 - val_loss: 1.9738\n","Epoch 3/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6221 - loss: 1.9347 - val_accuracy: 0.6730 - val_loss: 1.8232\n","Epoch 4/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6893 - loss: 1.7883 - val_accuracy: 0.7188 - val_loss: 1.6837\n","Epoch 5/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7225 - loss: 1.6534 - val_accuracy: 0.7382 - val_loss: 1.5557\n","Epoch 6/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7488 - loss: 1.5271 - val_accuracy: 0.7690 - val_loss: 1.4384\n","Epoch 7/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7682 - loss: 1.4134 - val_accuracy: 0.7788 - val_loss: 1.3334\n","Epoch 8/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7841 - loss: 1.3139 - val_accuracy: 0.7938 - val_loss: 1.2402\n","Epoch 9/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.7940 - loss: 1.2225 - val_accuracy: 0.8033 - val_loss: 1.1581\n","Epoch 10/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8008 - loss: 1.1415 - val_accuracy: 0.8057 - val_loss: 1.0859\n","Epoch 11/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8111 - loss: 1.0746 - val_accuracy: 0.8173 - val_loss: 1.0224\n","Epoch 12/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8193 - loss: 1.0101 - val_accuracy: 0.8240 - val_loss: 0.9669\n","Epoch 13/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8232 - loss: 0.9619 - val_accuracy: 0.8262 - val_loss: 0.9184\n","Epoch 14/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8259 - loss: 0.9162 - val_accuracy: 0.8298 - val_loss: 0.8753\n","Epoch 15/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8306 - loss: 0.8690 - val_accuracy: 0.8358 - val_loss: 0.8364\n","Epoch 16/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8357 - loss: 0.8388 - val_accuracy: 0.8395 - val_loss: 0.8025\n","Epoch 17/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8394 - loss: 0.7990 - val_accuracy: 0.8432 - val_loss: 0.7726\n","Epoch 18/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8423 - loss: 0.7732 - val_accuracy: 0.8427 - val_loss: 0.7453\n","Epoch 19/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8441 - loss: 0.7445 - val_accuracy: 0.8457 - val_loss: 0.7206\n","Epoch 20/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8468 - loss: 0.7210 - val_accuracy: 0.8510 - val_loss: 0.6982\n","Epoch 21/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8496 - loss: 0.7009 - val_accuracy: 0.8527 - val_loss: 0.6784\n","Epoch 22/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8515 - loss: 0.6786 - val_accuracy: 0.8543 - val_loss: 0.6599\n","Epoch 23/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8555 - loss: 0.6617 - val_accuracy: 0.8558 - val_loss: 0.6436\n","Epoch 24/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8575 - loss: 0.6447 - val_accuracy: 0.8565 - val_loss: 0.6278\n","Epoch 25/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8580 - loss: 0.6338 - val_accuracy: 0.8582 - val_loss: 0.6137\n","Epoch 26/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8591 - loss: 0.6179 - val_accuracy: 0.8620 - val_loss: 0.6005\n","Epoch 27/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8604 - loss: 0.6008 - val_accuracy: 0.8625 - val_loss: 0.5880\n","Epoch 28/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8629 - loss: 0.5892 - val_accuracy: 0.8652 - val_loss: 0.5767\n","Epoch 29/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8631 - loss: 0.5845 - val_accuracy: 0.8677 - val_loss: 0.5660\n","Epoch 30/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8634 - loss: 0.5741 - val_accuracy: 0.8682 - val_loss: 0.5562\n","Epoch 31/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8646 - loss: 0.5652 - val_accuracy: 0.8698 - val_loss: 0.5474\n","Epoch 32/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8664 - loss: 0.5540 - val_accuracy: 0.8700 - val_loss: 0.5387\n","Epoch 33/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8687 - loss: 0.5449 - val_accuracy: 0.8733 - val_loss: 0.5306\n","Epoch 34/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8687 - loss: 0.5377 - val_accuracy: 0.8742 - val_loss: 0.5227\n","Epoch 35/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8716 - loss: 0.5246 - val_accuracy: 0.8760 - val_loss: 0.5154\n","Epoch 36/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8715 - loss: 0.5207 - val_accuracy: 0.8765 - val_loss: 0.5088\n","Epoch 37/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8735 - loss: 0.5109 - val_accuracy: 0.8777 - val_loss: 0.5023\n","Epoch 38/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8714 - loss: 0.5143 - val_accuracy: 0.8770 - val_loss: 0.4963\n","Epoch 39/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8751 - loss: 0.5039 - val_accuracy: 0.8787 - val_loss: 0.4904\n","Epoch 40/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8738 - loss: 0.5006 - val_accuracy: 0.8798 - val_loss: 0.4847\n","Epoch 41/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8751 - loss: 0.4933 - val_accuracy: 0.8802 - val_loss: 0.4794\n","Epoch 42/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8800 - loss: 0.4798 - val_accuracy: 0.8810 - val_loss: 0.4745\n","Epoch 43/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8794 - loss: 0.4790 - val_accuracy: 0.8810 - val_loss: 0.4698\n","Epoch 44/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8782 - loss: 0.4756 - val_accuracy: 0.8812 - val_loss: 0.4654\n","Epoch 45/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8813 - loss: 0.4701 - val_accuracy: 0.8830 - val_loss: 0.4610\n","Epoch 46/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8797 - loss: 0.4657 - val_accuracy: 0.8832 - val_loss: 0.4566\n","Epoch 47/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8794 - loss: 0.4635 - val_accuracy: 0.8835 - val_loss: 0.4528\n","Epoch 48/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8837 - loss: 0.4544 - val_accuracy: 0.8838 - val_loss: 0.4489\n","Epoch 49/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8795 - loss: 0.4568 - val_accuracy: 0.8842 - val_loss: 0.4454\n","Epoch 50/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8823 - loss: 0.4523 - val_accuracy: 0.8848 - val_loss: 0.4418\n","Epoch 51/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8829 - loss: 0.4495 - val_accuracy: 0.8847 - val_loss: 0.4385\n","Epoch 52/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8851 - loss: 0.4457 - val_accuracy: 0.8853 - val_loss: 0.4354\n","Epoch 53/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8843 - loss: 0.4433 - val_accuracy: 0.8852 - val_loss: 0.4322\n","Epoch 54/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8843 - loss: 0.4355 - val_accuracy: 0.8875 - val_loss: 0.4292\n","Epoch 55/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8871 - loss: 0.4334 - val_accuracy: 0.8877 - val_loss: 0.4261\n","Epoch 56/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8850 - loss: 0.4304 - val_accuracy: 0.8877 - val_loss: 0.4232\n","Epoch 57/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8854 - loss: 0.4304 - val_accuracy: 0.8873 - val_loss: 0.4207\n","Epoch 58/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8859 - loss: 0.4251 - val_accuracy: 0.8877 - val_loss: 0.4181\n","Epoch 59/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8887 - loss: 0.4200 - val_accuracy: 0.8882 - val_loss: 0.4154\n","Epoch 60/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8866 - loss: 0.4241 - val_accuracy: 0.8888 - val_loss: 0.4131\n","Epoch 61/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8880 - loss: 0.4179 - val_accuracy: 0.8895 - val_loss: 0.4107\n","Epoch 62/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8878 - loss: 0.4179 - val_accuracy: 0.8887 - val_loss: 0.4084\n","Epoch 63/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8863 - loss: 0.4177 - val_accuracy: 0.8905 - val_loss: 0.4061\n","Epoch 64/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8871 - loss: 0.4175 - val_accuracy: 0.8910 - val_loss: 0.4041\n","Epoch 65/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8881 - loss: 0.4101 - val_accuracy: 0.8912 - val_loss: 0.4018\n","Epoch 66/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8898 - loss: 0.4067 - val_accuracy: 0.8920 - val_loss: 0.3998\n","Epoch 67/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8907 - loss: 0.4032 - val_accuracy: 0.8913 - val_loss: 0.3979\n","Epoch 68/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8908 - loss: 0.4072 - val_accuracy: 0.8918 - val_loss: 0.3960\n","Epoch 69/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8889 - loss: 0.4056 - val_accuracy: 0.8925 - val_loss: 0.3941\n","Epoch 70/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8921 - loss: 0.4025 - val_accuracy: 0.8925 - val_loss: 0.3924\n","Epoch 71/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8926 - loss: 0.3976 - val_accuracy: 0.8933 - val_loss: 0.3907\n","Epoch 72/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8918 - loss: 0.3939 - val_accuracy: 0.8937 - val_loss: 0.3889\n","Epoch 73/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8896 - loss: 0.3996 - val_accuracy: 0.8953 - val_loss: 0.3872\n","Epoch 74/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8920 - loss: 0.3886 - val_accuracy: 0.8938 - val_loss: 0.3857\n","Epoch 75/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8930 - loss: 0.3896 - val_accuracy: 0.8958 - val_loss: 0.3841\n","Epoch 76/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8906 - loss: 0.3949 - val_accuracy: 0.8962 - val_loss: 0.3827\n","Epoch 77/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8932 - loss: 0.3902 - val_accuracy: 0.8965 - val_loss: 0.3811\n","Epoch 78/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8905 - loss: 0.3979 - val_accuracy: 0.8963 - val_loss: 0.3796\n","Epoch 79/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8953 - loss: 0.3802 - val_accuracy: 0.8973 - val_loss: 0.3781\n","Epoch 80/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8928 - loss: 0.3888 - val_accuracy: 0.8975 - val_loss: 0.3768\n","Epoch 81/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8940 - loss: 0.3820 - val_accuracy: 0.8977 - val_loss: 0.3754\n","Epoch 82/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8945 - loss: 0.3804 - val_accuracy: 0.8978 - val_loss: 0.3740\n","Epoch 83/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8949 - loss: 0.3824 - val_accuracy: 0.8980 - val_loss: 0.3729\n","Epoch 84/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8950 - loss: 0.3808 - val_accuracy: 0.8987 - val_loss: 0.3714\n","Epoch 85/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8945 - loss: 0.3799 - val_accuracy: 0.8990 - val_loss: 0.3702\n","Epoch 86/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8923 - loss: 0.3821 - val_accuracy: 0.8993 - val_loss: 0.3691\n","Epoch 87/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8935 - loss: 0.3821 - val_accuracy: 0.8982 - val_loss: 0.3679\n","Epoch 88/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8962 - loss: 0.3727 - val_accuracy: 0.8992 - val_loss: 0.3668\n","Epoch 89/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8944 - loss: 0.3759 - val_accuracy: 0.8997 - val_loss: 0.3655\n","Epoch 90/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8963 - loss: 0.3701 - val_accuracy: 0.8990 - val_loss: 0.3645\n","Epoch 91/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8947 - loss: 0.3757 - val_accuracy: 0.8992 - val_loss: 0.3634\n","Epoch 92/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8951 - loss: 0.3707 - val_accuracy: 0.9005 - val_loss: 0.3622\n","Epoch 93/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8969 - loss: 0.3684 - val_accuracy: 0.8998 - val_loss: 0.3612\n","Epoch 94/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8970 - loss: 0.3658 - val_accuracy: 0.9002 - val_loss: 0.3602\n","Epoch 95/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8973 - loss: 0.3677 - val_accuracy: 0.9007 - val_loss: 0.3594\n","Epoch 96/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8948 - loss: 0.3723 - val_accuracy: 0.9002 - val_loss: 0.3582\n","Epoch 97/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8976 - loss: 0.3658 - val_accuracy: 0.9002 - val_loss: 0.3573\n","Epoch 98/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8954 - loss: 0.3669 - val_accuracy: 0.9013 - val_loss: 0.3564\n","Epoch 99/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8960 - loss: 0.3657 - val_accuracy: 0.9018 - val_loss: 0.3555\n","Epoch 100/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8961 - loss: 0.3659 - val_accuracy: 0.9030 - val_loss: 0.3544\n"]}],"source":["H_01_500 = model_01_500.fit(\n","    X_train, y_train,\n","    validation_split=0.1,\n","    epochs=100,\n","    batch_size = 512\n",")"]},{"cell_type":"code","execution_count":27,"metadata":{"id":"s6lj_lfsIIFp","colab":{"base_uri":"https://localhost:8080/","height":487},"executionInfo":{"status":"ok","timestamp":1758485168758,"user_tz":-180,"elapsed":213,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"15e02626-365b-4f54-9ff5-15e9cf6deb71"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x500 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.figure(figsize=(12, 5))\n","\n","plt.subplot(1, 2, 1)\n","plt.plot(H_01_500.history['loss'], label='Обучающая ошибка')\n","plt.plot(H_01_500.history['val_loss'], label='Валидационная ошибка')\n","plt.title('Функция ошибки по эпохам')\n","plt.xlabel('Epochs')\n","plt.ylabel('loss')\n","plt.legend()\n","plt.grid(True)"]},{"cell_type":"code","execution_count":28,"metadata":{"id":"zMa8AsG1IftR","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485172643,"user_tz":-180,"elapsed":1450,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"a5af30f6-3f2e-4139-8e97-c184d89b5fd0"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.9007 - loss: 0.3635\n","Loss on test data: 0.36660370230674744\n","Accuracy on test data: 0.9010000228881836\n"]}],"source":["scores_01_500=model_01_500.evaluate(X_test,y_test)\n","print('Loss on test data:',scores_01_500[0]) #значение функции ошибки\n","print('Accuracy on test data:',scores_01_500[1]) #значение метрики качества"]},{"cell_type":"code","execution_count":29,"metadata":{"id":"KDE5Vru8J7kp","colab":{"base_uri":"https://localhost:8080/","height":238},"executionInfo":{"status":"ok","timestamp":1758485196170,"user_tz":-180,"elapsed":154,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"0632517b-6ce2-4f20-c624-1a9da573ec07"},"outputs":[{"output_type":"display_data","data":{"text/plain":["\u001b[1mModel: \"sequential_5\"\u001b[0m\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_5\"</span>\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"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","│ dense_8 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m300\u001b[0m)            │       \u001b[38;5;34m235,500\u001b[0m │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_9 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m50\u001b[0m)             │        \u001b[38;5;34m15,050\u001b[0m │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_10 (\u001b[38;5;33mDense\u001b[0m)                │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)             │           \u001b[38;5;34m510\u001b[0m │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n","┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n","│ dense_8 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">300</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">235,500</span> │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_9 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">50</span>)             │        <span style=\"color: #00af00; text-decoration-color: #00af00\">15,050</span> │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_10 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)             │           <span style=\"color: #00af00; text-decoration-color: #00af00\">510</span> │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Total params: \u001b[0m\u001b[38;5;34m251,060\u001b[0m (980.70 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">251,060</span> (980.70 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m251,060\u001b[0m (980.70 KB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">251,060</span> (980.70 KB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n","</pre>\n"]},"metadata":{}}],"source":["model_01_300_50 = Sequential()\n","model_01_300_50.add(Dense(units=300, input_dim=num_pixels, activation='sigmoid'))\n","model_01_300_50.add(Dense(units=50, activation='sigmoid'))\n","model_01_300_50.add(Dense(units=num_classes, activation='softmax'))\n","model_01_300_50.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])\n","model_01_300_50.summary()"]},{"cell_type":"code","execution_count":30,"metadata":{"id":"SC5DsfcyLMVo","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485277730,"user_tz":-180,"elapsed":66773,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"ca48fe7c-bb60-4340-8655-bcda7db997dc"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 26ms/step - accuracy: 0.1039 - loss: 2.3373 - val_accuracy: 0.1325 - val_loss: 2.2951\n","Epoch 2/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step - accuracy: 0.1429 - loss: 2.2919 - val_accuracy: 0.1357 - val_loss: 2.2808\n","Epoch 3/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.1592 - loss: 2.2776 - val_accuracy: 0.1543 - val_loss: 2.2670\n","Epoch 4/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 7ms/step - accuracy: 0.1845 - loss: 2.2641 - val_accuracy: 0.2415 - val_loss: 2.2531\n","Epoch 5/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.2785 - loss: 2.2490 - val_accuracy: 0.3228 - val_loss: 2.2384\n","Epoch 6/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.3247 - loss: 2.2347 - val_accuracy: 0.3753 - val_loss: 2.2231\n","Epoch 7/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.3946 - loss: 2.2195 - val_accuracy: 0.4215 - val_loss: 2.2068\n","Epoch 8/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.4341 - loss: 2.2029 - val_accuracy: 0.4810 - val_loss: 2.1894\n","Epoch 9/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.4842 - loss: 2.1844 - val_accuracy: 0.5177 - val_loss: 2.1708\n","Epoch 10/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.5376 - loss: 2.1668 - val_accuracy: 0.5078 - val_loss: 2.1508\n","Epoch 11/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 8ms/step - accuracy: 0.5288 - loss: 2.1459 - val_accuracy: 0.5518 - val_loss: 2.1293\n","Epoch 12/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.5605 - loss: 2.1239 - val_accuracy: 0.5760 - val_loss: 2.1062\n","Epoch 13/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.5711 - loss: 2.0998 - val_accuracy: 0.5848 - val_loss: 2.0809\n","Epoch 14/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.5877 - loss: 2.0751 - val_accuracy: 0.5900 - val_loss: 2.0539\n","Epoch 15/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.5919 - loss: 2.0465 - val_accuracy: 0.6038 - val_loss: 2.0247\n","Epoch 16/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.6025 - loss: 2.0177 - val_accuracy: 0.6132 - val_loss: 1.9934\n","Epoch 17/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 10ms/step - accuracy: 0.6138 - loss: 1.9855 - val_accuracy: 0.6157 - val_loss: 1.9598\n","Epoch 18/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 9ms/step - accuracy: 0.6178 - loss: 1.9511 - val_accuracy: 0.6273 - val_loss: 1.9242\n","Epoch 19/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 14ms/step - accuracy: 0.6255 - loss: 1.9167 - val_accuracy: 0.6273 - val_loss: 1.8864\n","Epoch 20/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 7ms/step - accuracy: 0.6341 - loss: 1.8757 - val_accuracy: 0.6400 - val_loss: 1.8466\n","Epoch 21/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.6412 - loss: 1.8353 - val_accuracy: 0.6487 - val_loss: 1.8053\n","Epoch 22/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.6493 - loss: 1.7953 - val_accuracy: 0.6492 - val_loss: 1.7625\n","Epoch 23/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.6512 - loss: 1.7502 - val_accuracy: 0.6588 - val_loss: 1.7186\n","Epoch 24/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.6550 - loss: 1.7108 - val_accuracy: 0.6600 - val_loss: 1.6738\n","Epoch 25/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 8ms/step - accuracy: 0.6633 - loss: 1.6628 - val_accuracy: 0.6707 - val_loss: 1.6288\n","Epoch 26/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.6710 - loss: 1.6181 - val_accuracy: 0.6745 - val_loss: 1.5836\n","Epoch 27/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.6737 - loss: 1.5751 - val_accuracy: 0.6778 - val_loss: 1.5387\n","Epoch 28/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.6801 - loss: 1.5288 - val_accuracy: 0.6858 - val_loss: 1.4943\n","Epoch 29/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.6865 - loss: 1.4893 - val_accuracy: 0.6917 - val_loss: 1.4508\n","Epoch 30/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.6966 - loss: 1.4409 - val_accuracy: 0.6985 - val_loss: 1.4084\n","Epoch 31/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.6990 - loss: 1.4011 - val_accuracy: 0.7070 - val_loss: 1.3673\n","Epoch 32/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7073 - loss: 1.3601 - val_accuracy: 0.7098 - val_loss: 1.3275\n","Epoch 33/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7132 - loss: 1.3202 - val_accuracy: 0.7173 - val_loss: 1.2893\n","Epoch 34/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7199 - loss: 1.2800 - val_accuracy: 0.7238 - val_loss: 1.2527\n","Epoch 35/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7274 - loss: 1.2406 - val_accuracy: 0.7292 - val_loss: 1.2175\n","Epoch 36/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 14ms/step - accuracy: 0.7289 - loss: 1.2134 - val_accuracy: 0.7375 - val_loss: 1.1839\n","Epoch 37/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - accuracy: 0.7391 - loss: 1.1813 - val_accuracy: 0.7430 - val_loss: 1.1520\n","Epoch 38/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7429 - loss: 1.1453 - val_accuracy: 0.7475 - val_loss: 1.1214\n","Epoch 39/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7489 - loss: 1.1200 - val_accuracy: 0.7530 - val_loss: 1.0924\n","Epoch 40/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.7546 - loss: 1.0882 - val_accuracy: 0.7607 - val_loss: 1.0647\n","Epoch 41/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7572 - loss: 1.0657 - val_accuracy: 0.7635 - val_loss: 1.0385\n","Epoch 42/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7679 - loss: 1.0333 - val_accuracy: 0.7682 - val_loss: 1.0133\n","Epoch 43/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7692 - loss: 1.0102 - val_accuracy: 0.7732 - val_loss: 0.9894\n","Epoch 44/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7717 - loss: 0.9908 - val_accuracy: 0.7782 - val_loss: 0.9667\n","Epoch 45/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7800 - loss: 0.9665 - val_accuracy: 0.7797 - val_loss: 0.9451\n","Epoch 46/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7825 - loss: 0.9413 - val_accuracy: 0.7860 - val_loss: 0.9244\n","Epoch 47/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7860 - loss: 0.9243 - val_accuracy: 0.7898 - val_loss: 0.9047\n","Epoch 48/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7886 - loss: 0.9025 - val_accuracy: 0.7930 - val_loss: 0.8860\n","Epoch 49/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7955 - loss: 0.8839 - val_accuracy: 0.7958 - val_loss: 0.8681\n","Epoch 50/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7947 - loss: 0.8678 - val_accuracy: 0.8000 - val_loss: 0.8509\n","Epoch 51/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7996 - loss: 0.8523 - val_accuracy: 0.8013 - val_loss: 0.8346\n","Epoch 52/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7995 - loss: 0.8423 - val_accuracy: 0.8023 - val_loss: 0.8191\n","Epoch 53/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8023 - loss: 0.8245 - val_accuracy: 0.8078 - val_loss: 0.8041\n","Epoch 54/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8089 - loss: 0.8036 - val_accuracy: 0.8080 - val_loss: 0.7899\n","Epoch 55/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8099 - loss: 0.7926 - val_accuracy: 0.8105 - val_loss: 0.7761\n","Epoch 56/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8130 - loss: 0.7780 - val_accuracy: 0.8133 - val_loss: 0.7628\n","Epoch 57/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8194 - loss: 0.7630 - val_accuracy: 0.8178 - val_loss: 0.7504\n","Epoch 58/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8202 - loss: 0.7558 - val_accuracy: 0.8190 - val_loss: 0.7382\n","Epoch 59/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8216 - loss: 0.7402 - val_accuracy: 0.8207 - val_loss: 0.7268\n","Epoch 60/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8222 - loss: 0.7317 - val_accuracy: 0.8235 - val_loss: 0.7155\n","Epoch 61/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8265 - loss: 0.7202 - val_accuracy: 0.8262 - val_loss: 0.7046\n","Epoch 62/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8288 - loss: 0.7088 - val_accuracy: 0.8293 - val_loss: 0.6943\n","Epoch 63/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8293 - loss: 0.6998 - val_accuracy: 0.8303 - val_loss: 0.6845\n","Epoch 64/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8331 - loss: 0.6896 - val_accuracy: 0.8322 - val_loss: 0.6749\n","Epoch 65/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8351 - loss: 0.6782 - val_accuracy: 0.8365 - val_loss: 0.6655\n","Epoch 66/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8371 - loss: 0.6706 - val_accuracy: 0.8375 - val_loss: 0.6568\n","Epoch 67/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8417 - loss: 0.6563 - val_accuracy: 0.8400 - val_loss: 0.6481\n","Epoch 68/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8416 - loss: 0.6475 - val_accuracy: 0.8417 - val_loss: 0.6397\n","Epoch 69/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8420 - loss: 0.6445 - val_accuracy: 0.8427 - val_loss: 0.6317\n","Epoch 70/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8436 - loss: 0.6397 - val_accuracy: 0.8447 - val_loss: 0.6239\n","Epoch 71/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8446 - loss: 0.6303 - val_accuracy: 0.8445 - val_loss: 0.6166\n","Epoch 72/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8466 - loss: 0.6227 - val_accuracy: 0.8472 - val_loss: 0.6093\n","Epoch 73/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8463 - loss: 0.6187 - val_accuracy: 0.8490 - val_loss: 0.6023\n","Epoch 74/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8483 - loss: 0.6085 - val_accuracy: 0.8502 - val_loss: 0.5955\n","Epoch 75/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8531 - loss: 0.5998 - val_accuracy: 0.8512 - val_loss: 0.5892\n","Epoch 76/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8518 - loss: 0.5938 - val_accuracy: 0.8548 - val_loss: 0.5827\n","Epoch 77/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8518 - loss: 0.5921 - val_accuracy: 0.8552 - val_loss: 0.5765\n","Epoch 78/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8551 - loss: 0.5830 - val_accuracy: 0.8573 - val_loss: 0.5705\n","Epoch 79/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8575 - loss: 0.5744 - val_accuracy: 0.8583 - val_loss: 0.5648\n","Epoch 80/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8594 - loss: 0.5679 - val_accuracy: 0.8593 - val_loss: 0.5592\n","Epoch 81/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8592 - loss: 0.5645 - val_accuracy: 0.8607 - val_loss: 0.5538\n","Epoch 82/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8627 - loss: 0.5533 - val_accuracy: 0.8617 - val_loss: 0.5483\n","Epoch 83/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8612 - loss: 0.5535 - val_accuracy: 0.8617 - val_loss: 0.5434\n","Epoch 84/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8660 - loss: 0.5414 - val_accuracy: 0.8633 - val_loss: 0.5384\n","Epoch 85/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8651 - loss: 0.5382 - val_accuracy: 0.8633 - val_loss: 0.5337\n","Epoch 86/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8657 - loss: 0.5371 - val_accuracy: 0.8660 - val_loss: 0.5289\n","Epoch 87/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8664 - loss: 0.5331 - val_accuracy: 0.8662 - val_loss: 0.5244\n","Epoch 88/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8663 - loss: 0.5285 - val_accuracy: 0.8673 - val_loss: 0.5200\n","Epoch 89/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8675 - loss: 0.5247 - val_accuracy: 0.8680 - val_loss: 0.5155\n","Epoch 90/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8696 - loss: 0.5166 - val_accuracy: 0.8690 - val_loss: 0.5114\n","Epoch 91/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8695 - loss: 0.5157 - val_accuracy: 0.8702 - val_loss: 0.5073\n","Epoch 92/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8680 - loss: 0.5163 - val_accuracy: 0.8703 - val_loss: 0.5033\n","Epoch 93/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8706 - loss: 0.5092 - val_accuracy: 0.8727 - val_loss: 0.4994\n","Epoch 94/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8749 - loss: 0.5013 - val_accuracy: 0.8728 - val_loss: 0.4958\n","Epoch 95/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8711 - loss: 0.5031 - val_accuracy: 0.8733 - val_loss: 0.4921\n","Epoch 96/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8742 - loss: 0.4936 - val_accuracy: 0.8743 - val_loss: 0.4886\n","Epoch 97/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8718 - loss: 0.4968 - val_accuracy: 0.8750 - val_loss: 0.4850\n","Epoch 98/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8739 - loss: 0.4927 - val_accuracy: 0.8753 - val_loss: 0.4817\n","Epoch 99/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8746 - loss: 0.4847 - val_accuracy: 0.8758 - val_loss: 0.4783\n","Epoch 100/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8771 - loss: 0.4793 - val_accuracy: 0.8770 - val_loss: 0.4751\n"]}],"source":["H_01_300_50 = model_01_300_50.fit(\n","    X_train, y_train,\n","    validation_split=0.1,\n","    epochs=100,\n","    batch_size=512\n",")"]},{"cell_type":"code","execution_count":31,"metadata":{"id":"Wh6GsTrvLo0c","colab":{"base_uri":"https://localhost:8080/","height":487},"executionInfo":{"status":"ok","timestamp":1758485311049,"user_tz":-180,"elapsed":394,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"619a23e6-9324-4d4f-d15b-1020b2c824c3"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x500 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.figure(figsize=(12, 5))\n","\n","plt.subplot(1, 2, 1)\n","plt.plot(H_01_300_50.history['loss'], label='Обучающая ошибка')\n","plt.plot(H_01_300_50.history['val_loss'], label='Валидационная ошибка')\n","plt.title('Функция ошибки по эпохам')\n","plt.xlabel('Epochs')\n","plt.ylabel('loss')\n","plt.legend()\n","plt.grid(True)"]},{"cell_type":"code","execution_count":32,"metadata":{"id":"pP5dd_4LMZqn","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485317073,"user_tz":-180,"elapsed":2667,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"7a0b3c25-c98e-456a-d8e9-51ccba27cd0b"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step - accuracy: 0.8761 - loss: 0.4844\n","Loss on test data: 0.4881931245326996\n","Accuracy on test data: 0.8740000128746033\n"]}],"source":["scores_01_300_50=model_01_300_50.evaluate(X_test,y_test)\n","print('Loss on test data:',scores_01_300_50[0])\n","print('Accuracy on test data:',scores_01_300_50[1])"]},{"cell_type":"code","execution_count":33,"metadata":{"id":"2KGjZmc1NVSI","colab":{"base_uri":"https://localhost:8080/","height":238},"executionInfo":{"status":"ok","timestamp":1758485320784,"user_tz":-180,"elapsed":134,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"16a357cb-5197-4fcf-c024-09c14787637c"},"outputs":[{"output_type":"display_data","data":{"text/plain":["\u001b[1mModel: \"sequential_6\"\u001b[0m\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_6\"</span>\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"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","│ dense_11 (\u001b[38;5;33mDense\u001b[0m)                │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m300\u001b[0m)            │       \u001b[38;5;34m235,500\u001b[0m │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_12 (\u001b[38;5;33mDense\u001b[0m)                │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m100\u001b[0m)            │        \u001b[38;5;34m30,100\u001b[0m │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_13 (\u001b[38;5;33mDense\u001b[0m)                │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)             │         \u001b[38;5;34m1,010\u001b[0m │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n","┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n","│ dense_11 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">300</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">235,500</span> │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_12 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">100</span>)            │        <span style=\"color: #00af00; text-decoration-color: #00af00\">30,100</span> │\n","├─────────────────────────────────┼────────────────────────┼───────────────┤\n","│ dense_13 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)             │         <span style=\"color: #00af00; text-decoration-color: #00af00\">1,010</span> │\n","└─────────────────────────────────┴────────────────────────┴───────────────┘\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Total params: \u001b[0m\u001b[38;5;34m266,610\u001b[0m (1.02 MB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">266,610</span> (1.02 MB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m266,610\u001b[0m (1.02 MB)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">266,610</span> (1.02 MB)\n","</pre>\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"],"text/html":["<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n","</pre>\n"]},"metadata":{}}],"source":["model_01_300_100 = Sequential()\n","model_01_300_100.add(Dense(units=300, input_dim=num_pixels, activation='sigmoid'))\n","model_01_300_100.add(Dense(units=100, activation='sigmoid'))\n","model_01_300_100.add(Dense(units=num_classes, activation='softmax'))\n","model_01_300_100.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])\n","model_01_300_100.summary()"]},{"cell_type":"code","execution_count":34,"metadata":{"id":"_t8LqaEiOiqw","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485391291,"user_tz":-180,"elapsed":61754,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"9686c162-0c30-42d2-a4f1-ab91ca97559e"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 19ms/step - accuracy: 0.1070 - loss: 2.3687 - val_accuracy: 0.1328 - val_loss: 2.2869\n","Epoch 2/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 6ms/step - accuracy: 0.1536 - loss: 2.2832 - val_accuracy: 0.1165 - val_loss: 2.2717\n","Epoch 3/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.1559 - loss: 2.2679 - val_accuracy: 0.1893 - val_loss: 2.2564\n","Epoch 4/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 7ms/step - accuracy: 0.2401 - loss: 2.2525 - val_accuracy: 0.2463 - val_loss: 2.2406\n","Epoch 5/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.2753 - loss: 2.2360 - val_accuracy: 0.3690 - val_loss: 2.2241\n","Epoch 6/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.3593 - loss: 2.2189 - val_accuracy: 0.4323 - val_loss: 2.2067\n","Epoch 7/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.4321 - loss: 2.2018 - val_accuracy: 0.4517 - val_loss: 2.1882\n","Epoch 8/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.4562 - loss: 2.1821 - val_accuracy: 0.4837 - val_loss: 2.1682\n","Epoch 9/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.4822 - loss: 2.1624 - val_accuracy: 0.5333 - val_loss: 2.1469\n","Epoch 10/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.5166 - loss: 2.1417 - val_accuracy: 0.5207 - val_loss: 2.1236\n","Epoch 11/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.5365 - loss: 2.1172 - val_accuracy: 0.5345 - val_loss: 2.0984\n","Epoch 12/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.5490 - loss: 2.0925 - val_accuracy: 0.5278 - val_loss: 2.0709\n","Epoch 13/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 9ms/step - accuracy: 0.5436 - loss: 2.0629 - val_accuracy: 0.5803 - val_loss: 2.0408\n","Epoch 14/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.5827 - loss: 2.0324 - val_accuracy: 0.5922 - val_loss: 2.0079\n","Epoch 15/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.5978 - loss: 1.9996 - val_accuracy: 0.5823 - val_loss: 1.9725\n","Epoch 16/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.5958 - loss: 1.9629 - val_accuracy: 0.6122 - val_loss: 1.9342\n","Epoch 17/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.6200 - loss: 1.9249 - val_accuracy: 0.6068 - val_loss: 1.8931\n","Epoch 18/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.6200 - loss: 1.8817 - val_accuracy: 0.6195 - val_loss: 1.8492\n","Epoch 19/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.6285 - loss: 1.8358 - val_accuracy: 0.6505 - val_loss: 1.8028\n","Epoch 20/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.6448 - loss: 1.7902 - val_accuracy: 0.6435 - val_loss: 1.7546\n","Epoch 21/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6468 - loss: 1.7447 - val_accuracy: 0.6480 - val_loss: 1.7047\n","Epoch 22/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.6540 - loss: 1.6938 - val_accuracy: 0.6660 - val_loss: 1.6539\n","Epoch 23/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.6696 - loss: 1.6431 - val_accuracy: 0.6608 - val_loss: 1.6027\n","Epoch 24/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.6675 - loss: 1.5930 - val_accuracy: 0.6803 - val_loss: 1.5518\n","Epoch 25/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.6840 - loss: 1.5408 - val_accuracy: 0.6958 - val_loss: 1.5013\n","Epoch 26/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.6951 - loss: 1.4895 - val_accuracy: 0.7042 - val_loss: 1.4521\n","Epoch 27/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7064 - loss: 1.4433 - val_accuracy: 0.7098 - val_loss: 1.4044\n","Epoch 28/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7130 - loss: 1.3961 - val_accuracy: 0.7142 - val_loss: 1.3587\n","Epoch 29/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7169 - loss: 1.3499 - val_accuracy: 0.7298 - val_loss: 1.3147\n","Epoch 30/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7278 - loss: 1.3082 - val_accuracy: 0.7313 - val_loss: 1.2729\n","Epoch 31/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7352 - loss: 1.2638 - val_accuracy: 0.7373 - val_loss: 1.2332\n","Epoch 32/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.7420 - loss: 1.2247 - val_accuracy: 0.7425 - val_loss: 1.1955\n","Epoch 33/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.7468 - loss: 1.1908 - val_accuracy: 0.7470 - val_loss: 1.1600\n","Epoch 34/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.7526 - loss: 1.1557 - val_accuracy: 0.7595 - val_loss: 1.1263\n","Epoch 35/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.7569 - loss: 1.1232 - val_accuracy: 0.7653 - val_loss: 1.0945\n","Epoch 36/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7607 - loss: 1.0941 - val_accuracy: 0.7665 - val_loss: 1.0646\n","Epoch 37/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7689 - loss: 1.0604 - val_accuracy: 0.7748 - val_loss: 1.0362\n","Epoch 38/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7709 - loss: 1.0389 - val_accuracy: 0.7773 - val_loss: 1.0095\n","Epoch 39/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7743 - loss: 1.0101 - val_accuracy: 0.7808 - val_loss: 0.9844\n","Epoch 40/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7802 - loss: 0.9829 - val_accuracy: 0.7843 - val_loss: 0.9606\n","Epoch 41/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7838 - loss: 0.9628 - val_accuracy: 0.7887 - val_loss: 0.9379\n","Epoch 42/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.7869 - loss: 0.9390 - val_accuracy: 0.7933 - val_loss: 0.9162\n","Epoch 43/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.7921 - loss: 0.9161 - val_accuracy: 0.7962 - val_loss: 0.8955\n","Epoch 44/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.7966 - loss: 0.8950 - val_accuracy: 0.7988 - val_loss: 0.8759\n","Epoch 45/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8002 - loss: 0.8727 - val_accuracy: 0.8028 - val_loss: 0.8575\n","Epoch 46/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8020 - loss: 0.8602 - val_accuracy: 0.8050 - val_loss: 0.8397\n","Epoch 47/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8066 - loss: 0.8408 - val_accuracy: 0.8083 - val_loss: 0.8229\n","Epoch 48/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8055 - loss: 0.8284 - val_accuracy: 0.8122 - val_loss: 0.8068\n","Epoch 49/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8113 - loss: 0.8099 - val_accuracy: 0.8112 - val_loss: 0.7913\n","Epoch 50/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8149 - loss: 0.7931 - val_accuracy: 0.8158 - val_loss: 0.7764\n","Epoch 51/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8173 - loss: 0.7760 - val_accuracy: 0.8185 - val_loss: 0.7625\n","Epoch 52/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8190 - loss: 0.7667 - val_accuracy: 0.8225 - val_loss: 0.7487\n","Epoch 53/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8223 - loss: 0.7514 - val_accuracy: 0.8275 - val_loss: 0.7357\n","Epoch 54/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8255 - loss: 0.7384 - val_accuracy: 0.8280 - val_loss: 0.7234\n","Epoch 55/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8268 - loss: 0.7266 - val_accuracy: 0.8293 - val_loss: 0.7114\n","Epoch 56/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8308 - loss: 0.7069 - val_accuracy: 0.8328 - val_loss: 0.6999\n","Epoch 57/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8295 - loss: 0.7043 - val_accuracy: 0.8357 - val_loss: 0.6889\n","Epoch 58/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8309 - loss: 0.6982 - val_accuracy: 0.8367 - val_loss: 0.6784\n","Epoch 59/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8373 - loss: 0.6804 - val_accuracy: 0.8402 - val_loss: 0.6682\n","Epoch 60/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8363 - loss: 0.6759 - val_accuracy: 0.8412 - val_loss: 0.6583\n","Epoch 61/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8404 - loss: 0.6558 - val_accuracy: 0.8425 - val_loss: 0.6489\n","Epoch 62/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8392 - loss: 0.6574 - val_accuracy: 0.8430 - val_loss: 0.6399\n","Epoch 63/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8417 - loss: 0.6455 - val_accuracy: 0.8468 - val_loss: 0.6313\n","Epoch 64/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8427 - loss: 0.6393 - val_accuracy: 0.8468 - val_loss: 0.6226\n","Epoch 65/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8439 - loss: 0.6299 - val_accuracy: 0.8487 - val_loss: 0.6146\n","Epoch 66/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8475 - loss: 0.6208 - val_accuracy: 0.8500 - val_loss: 0.6070\n","Epoch 67/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8484 - loss: 0.6127 - val_accuracy: 0.8522 - val_loss: 0.5994\n","Epoch 68/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8505 - loss: 0.6047 - val_accuracy: 0.8533 - val_loss: 0.5921\n","Epoch 69/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8507 - loss: 0.5999 - val_accuracy: 0.8553 - val_loss: 0.5851\n","Epoch 70/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8554 - loss: 0.5851 - val_accuracy: 0.8557 - val_loss: 0.5786\n","Epoch 71/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8527 - loss: 0.5844 - val_accuracy: 0.8568 - val_loss: 0.5719\n","Epoch 72/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8574 - loss: 0.5742 - val_accuracy: 0.8583 - val_loss: 0.5656\n","Epoch 73/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8554 - loss: 0.5724 - val_accuracy: 0.8595 - val_loss: 0.5595\n","Epoch 74/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8562 - loss: 0.5662 - val_accuracy: 0.8598 - val_loss: 0.5538\n","Epoch 75/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8580 - loss: 0.5593 - val_accuracy: 0.8608 - val_loss: 0.5482\n","Epoch 76/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8605 - loss: 0.5549 - val_accuracy: 0.8623 - val_loss: 0.5428\n","Epoch 77/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8618 - loss: 0.5465 - val_accuracy: 0.8632 - val_loss: 0.5373\n","Epoch 78/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8612 - loss: 0.5462 - val_accuracy: 0.8652 - val_loss: 0.5321\n","Epoch 79/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8649 - loss: 0.5372 - val_accuracy: 0.8655 - val_loss: 0.5272\n","Epoch 80/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8635 - loss: 0.5326 - val_accuracy: 0.8682 - val_loss: 0.5224\n","Epoch 81/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.8665 - loss: 0.5271 - val_accuracy: 0.8683 - val_loss: 0.5177\n","Epoch 82/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 5ms/step - accuracy: 0.8641 - loss: 0.5286 - val_accuracy: 0.8693 - val_loss: 0.5132\n","Epoch 83/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8683 - loss: 0.5145 - val_accuracy: 0.8693 - val_loss: 0.5086\n","Epoch 84/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8694 - loss: 0.5124 - val_accuracy: 0.8705 - val_loss: 0.5044\n","Epoch 85/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8692 - loss: 0.5131 - val_accuracy: 0.8710 - val_loss: 0.5003\n","Epoch 86/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8731 - loss: 0.5021 - val_accuracy: 0.8718 - val_loss: 0.4963\n","Epoch 87/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8708 - loss: 0.5041 - val_accuracy: 0.8730 - val_loss: 0.4924\n","Epoch 88/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8707 - loss: 0.5002 - val_accuracy: 0.8733 - val_loss: 0.4885\n","Epoch 89/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8720 - loss: 0.4959 - val_accuracy: 0.8737 - val_loss: 0.4849\n","Epoch 90/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8737 - loss: 0.4936 - val_accuracy: 0.8738 - val_loss: 0.4812\n","Epoch 91/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8743 - loss: 0.4866 - val_accuracy: 0.8755 - val_loss: 0.4777\n","Epoch 92/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8738 - loss: 0.4834 - val_accuracy: 0.8772 - val_loss: 0.4744\n","Epoch 93/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8730 - loss: 0.4854 - val_accuracy: 0.8765 - val_loss: 0.4710\n","Epoch 94/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8760 - loss: 0.4751 - val_accuracy: 0.8772 - val_loss: 0.4679\n","Epoch 95/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8773 - loss: 0.4742 - val_accuracy: 0.8775 - val_loss: 0.4647\n","Epoch 96/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - accuracy: 0.8781 - loss: 0.4704 - val_accuracy: 0.8780 - val_loss: 0.4616\n","Epoch 97/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8746 - loss: 0.4729 - val_accuracy: 0.8785 - val_loss: 0.4588\n","Epoch 98/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8787 - loss: 0.4652 - val_accuracy: 0.8797 - val_loss: 0.4558\n","Epoch 99/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.8778 - loss: 0.4651 - val_accuracy: 0.8798 - val_loss: 0.4530\n","Epoch 100/100\n","\u001b[1m106/106\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.8780 - loss: 0.4608 - val_accuracy: 0.8803 - val_loss: 0.4502\n"]}],"source":["H_01_300_100 = model_01_300_100.fit(\n","    X_train, y_train,\n","    validation_split=0.1,\n","    epochs=100,\n","    batch_size=512\n",")"]},{"cell_type":"code","execution_count":35,"metadata":{"id":"uvTQqGxRO38O","colab":{"base_uri":"https://localhost:8080/","height":487},"executionInfo":{"status":"ok","timestamp":1758485399049,"user_tz":-180,"elapsed":487,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"019606fc-7864-484c-d7d5-c769db23d523"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x500 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.figure(figsize=(12, 5))\n","\n","plt.subplot(1, 2, 1)\n","plt.plot(H_01_300_100.history['loss'], label='Обучающая ошибка')\n","plt.plot(H_01_300_100.history['val_loss'], label='Валидационная ошибка')\n","plt.title('Функция ошибки по эпохам')\n","plt.xlabel('Epochs')\n","plt.ylabel('loss')\n","plt.legend()\n","plt.grid(True)"]},{"cell_type":"code","execution_count":36,"metadata":{"id":"RGl_z5fmPG7t","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485404216,"user_tz":-180,"elapsed":2727,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"bdfb1abf-1ca4-4eb5-9fbd-c940c2977569"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8814 - loss: 0.4605\n","Loss on test data: 0.4638420343399048\n","Accuracy on test data: 0.8795999884605408\n"]}],"source":["scores_01_300_100=model_01_300_100.evaluate(X_test,y_test)\n","print('Loss on test data:',scores_01_300_100[0])\n","print('Accuracy on test data:',scores_01_300_100[1])"]},{"cell_type":"code","execution_count":37,"metadata":{"id":"fVMBwoRJVXlZ","executionInfo":{"status":"ok","timestamp":1758485422219,"user_tz":-180,"elapsed":571,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["model_01_300.save(filepath='best_model.keras')\n"]},{"cell_type":"code","execution_count":38,"metadata":{"id":"yFKJ50yqW0jD","executionInfo":{"status":"ok","timestamp":1758485424981,"user_tz":-180,"elapsed":98,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["from keras.models import load_model\n","model = load_model('best_model.keras')"]},{"cell_type":"code","execution_count":39,"metadata":{"id":"WXonZCRTWLEr","colab":{"base_uri":"https://localhost:8080/","height":517},"executionInfo":{"status":"ok","timestamp":1758485427550,"user_tz":-180,"elapsed":450,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"53e59c68-4e34-4403-fa10-030ef090bc98"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 251ms/step\n","NN output: [[6.7044683e-03 6.5092892e-05 8.0898860e-03 3.6560427e-04 4.4942164e-04\n","  1.0991883e-02 9.6887839e-01 3.7091802e-06 4.2457585e-03 2.0581596e-04]]\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}},{"output_type":"stream","name":"stdout","text":["Real mark:  6\n","NN answer:  6\n"]}],"source":["n = 123\n","result = model.predict(X_test[n:n+1])\n","print('NN output:', result)\n","plt.imshow(X_test[n].reshape(28,28), cmap=plt.get_cmap('gray'))\n","plt.show()\n","print('Real mark: ', str(np.argmax(y_test[n])))\n","print('NN answer: ', str(np.argmax(result)))"]},{"cell_type":"code","execution_count":40,"metadata":{"id":"cKmqXXcOX80K","colab":{"base_uri":"https://localhost:8080/","height":517},"executionInfo":{"status":"ok","timestamp":1758485432276,"user_tz":-180,"elapsed":243,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"214bd2f9-bbb2-4bbb-a2e3-049385771de7"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 27ms/step\n","NN output: [[3.7733166e-04 3.6412096e-04 1.4499854e-03 9.2658949e-01 5.1390834e-04\n","  5.4276615e-02 3.5510810e-05 8.6189411e-04 1.2458544e-02 3.0724849e-03]]\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}},{"output_type":"stream","name":"stdout","text":["Real mark:  3\n","NN answer:  3\n"]}],"source":["n = 765\n","result = model.predict(X_test[n:n+1])\n","print('NN output:', result)\n","plt.imshow(X_test[n].reshape(28,28), cmap=plt.get_cmap('gray'))\n","plt.show()\n","print('Real mark: ', str(np.argmax(y_test[n])))\n","print('NN answer: ', str(np.argmax(result)))"]},{"cell_type":"code","execution_count":41,"metadata":{"id":"43z4eLyxiLs5","executionInfo":{"status":"ok","timestamp":1758485438169,"user_tz":-180,"elapsed":433,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["from PIL import Image\n","file_07_data = Image.open('7.png')\n","file_07_data = file_07_data.convert('L')\n","test_07_img = np.array(file_07_data)"]},{"cell_type":"code","execution_count":42,"metadata":{"id":"oZR5iCUtiYAh","colab":{"base_uri":"https://localhost:8080/","height":430},"executionInfo":{"status":"ok","timestamp":1758485444072,"user_tz":-180,"elapsed":188,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"ed0cf700-05cd-41b6-c22c-85934b6474f1"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAaAAAAGdCAYAAABU0qcqAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAGHxJREFUeJzt3X9MVff9x/HXVeFWW7gUES63IkVtNamVZU4ZcXVNJIpbTP3xh+v6h12MjfbaTF27xSVquyxhs0mzdDHr/qpZVm1nMjT1DxNFwWxDm1qNMeuIMDYwcnE14VxEQQOf7x+s97urIIL38r738nwkn6Tcc7j37fHIsweOV59zzgkAgHE2yXoAAMDERIAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAICJKdYD3GtgYEDXrl1TTk6OfD6f9TgAgFFyzqm7u1uhUEiTJg1/nZNyAbp27ZpKSkqsxwAAPKL29nbNnDlz2O0p9y24nJwc6xEAAAkw0tfzpAVo//79evrpp/XYY4+poqJCn3322UN9Ht92A4DMMNLX86QE6JNPPtHOnTu1d+9effHFFyovL9fKlSt1/fr1ZLwcACAduSRYsmSJC4fDsY/7+/tdKBRyNTU1I36u53lOEovFYrHSfHme98Cv9wm/Arpz547Onz+vqqqq2GOTJk1SVVWVGhsb79u/r69P0Wg0bgEAMl/CA/TVV1+pv79fRUVFcY8XFRUpEonct39NTY0CgUBscQccAEwM5nfB7dq1S57nxVZ7e7v1SACAcZDwvwdUUFCgyZMnq7OzM+7xzs5OBYPB+/b3+/3y+/2JHgMAkOISfgWUnZ2tRYsWqa6uLvbYwMCA6urqVFlZmeiXAwCkqaS8E8LOnTu1ceNGfetb39KSJUv0m9/8Rj09PfrRj36UjJcDAKShpARow4YN+s9//qM9e/YoEonoG9/4ho4fP37fjQkAgInL55xz1kP8r2g0qkAgYD0GAOAReZ6n3NzcYbeb3wUHAJiYCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACAiYQH6O2335bP54tb8+fPT/TLAADS3JRkPOlzzz2nkydP/v+LTEnKywAA0lhSyjBlyhQFg8FkPDUAIEMk5WdAV65cUSgU0uzZs/XKK6+ora1t2H37+voUjUbjFgAg8yU8QBUVFTpw4ICOHz+u3/3ud2ptbdULL7yg7u7uIfevqalRIBCIrZKSkkSPBABIQT7nnEvmC3R1dam0tFTvvfeeNm3adN/2vr4+9fX1xT6ORqNECAAygOd5ys3NHXZ70u8OyMvL07PPPqvm5uYht/v9fvn9/mSPAQBIMUn/e0A3b95US0uLiouLk/1SAIA0kvAAvfnmm2poaNC//vUv/e1vf9PatWs1efJkvfzyy4l+KQBAGkv4t+CuXr2ql19+WTdu3NCMGTP0ne98R2fPntWMGTMS/VIAgDSW9JsQRisajSoQCFiPAaScFPujOiH4fD7rEdLaSDch8F5wAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAICJpP+DdMhcvDlm6svEN9PkvMscXAEBAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABO+GjXGVie/OjNTHeZeauAICAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEzwZqQAzDjnrEeAIa6AAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATvBkpeENIpBWfz2c9AhKEKyAAgAkCBAAwMeoAnTlzRqtXr1YoFJLP59ORI0fitjvntGfPHhUXF2vq1KmqqqrSlStXEjUvACBDjDpAPT09Ki8v1/79+4fcvm/fPr3//vv64IMPdO7cOT3++ONauXKlent7H3lYAEAGcY9AkqutrY19PDAw4ILBoHv33Xdjj3V1dTm/3+8OHTr0UM/peZ6TxBrHNZ6sf62s1FqcQ5m9PM974O9lQn8G1NraqkgkoqqqqthjgUBAFRUVamxsHPJz+vr6FI1G4xYAIPMlNECRSESSVFRUFPd4UVFRbNu9ampqFAgEYqukpCSRIwEAUpT5XXC7du2S53mx1d7ebj0SAGAcJDRAwWBQktTZ2Rn3eGdnZ2zbvfx+v3Jzc+MWACDzJTRAZWVlCgaDqquriz0WjUZ17tw5VVZWJvKlAABpbtRvxXPz5k01NzfHPm5tbdXFixeVn5+vWbNmafv27frlL3+pZ555RmVlZdq9e7dCoZDWrFmTyLkBAOlutLdAnj59esjb7TZu3OicG7wVe/fu3a6oqMj5/X63fPly19TU9NDPz23Y47/Gk/WvlZVai3Mos9dIt2H7/vsbmjKi0agCgYD1GBPKWE8B3hQS/2u8vpRw3qUPz/Me+HN987vgAAATEwECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEyM+t8DApD5eGdrjAeugAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAE7wZKXhDSCQE5xFGiysgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEb0YKZDDnnPUIwLC4AgIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATPBmpADu4/P5rEfABMAVEADABAECAJgYdYDOnDmj1atXKxQKyefz6ciRI3HbX331Vfl8vrhVXV2dqHkBABli1AHq6elReXm59u/fP+w+1dXV6ujoiK1Dhw490pAAgMwz6psQVq1apVWrVj1wH7/fr2AwOOahAACZLyk/A6qvr1dhYaHmzZunrVu36saNG8Pu29fXp2g0GrcAAJkv4QGqrq7WH/7wB9XV1enXv/61GhoatGrVKvX39w+5f01NjQKBQGyVlJQkeiQAQAryOefcmD/Z51Ntba3WrFkz7D7//Oc/NWfOHJ08eVLLly+/b3tfX5/6+vpiH0ejUSIEJMhY/3jz94CQCJ7nKTc3d9jtSb8Ne/bs2SooKFBzc/OQ2/1+v3Jzc+MWACDzJT1AV69e1Y0bN1RcXJzslwIApJFR3wV38+bNuKuZ1tZWXbx4Ufn5+crPz9c777yj9evXKxgMqqWlRT/96U81d+5crVy5MqGDAwDSnBul06dPO0n3rY0bN7pbt265FStWuBkzZrisrCxXWlrqNm/e7CKRyEM/v+d5Qz4/i8Ua/Ror67lZmbE8z3vgefZINyEkQzQaVSAQsB4DyAhj/ePNTQhIBPObEAAAGAoBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgYor1AACSx+fzjenznHPj9lqYuLgCAgCYIEAAABOjClBNTY0WL16snJwcFRYWas2aNWpqaorbp7e3V+FwWNOnT9cTTzyh9evXq7OzM6FDAwDS36gC1NDQoHA4rLNnz+rEiRO6e/euVqxYoZ6entg+O3bs0KeffqrDhw+roaFB165d07p16xI+OAAgzblHcP36dSfJNTQ0OOec6+rqcllZWe7w4cOxfb788ksnyTU2Nj7Uc3qe5ySxWCzDNRbWM7NSb3me98Bz5pF+BuR5niQpPz9fknT+/HndvXtXVVVVsX3mz5+vWbNmqbGxccjn6OvrUzQajVsAgMw35gANDAxo+/btWrp0qRYsWCBJikQiys7OVl5eXty+RUVFikQiQz5PTU2NAoFAbJWUlIx1JABAGhlzgMLhsC5fvqyPP/74kQbYtWuXPM+Lrfb29kd6PgBAehjTX0Tdtm2bjh07pjNnzmjmzJmxx4PBoO7cuaOurq64q6DOzk4Fg8Ehn8vv98vv949lDABAGhvVFZBzTtu2bVNtba1OnTqlsrKyuO2LFi1SVlaW6urqYo81NTWpra1NlZWViZkYAJARRnUFFA6HdfDgQR09elQ5OTmxn+sEAgFNnTpVgUBAmzZt0s6dO5Wfn6/c3Fy98cYbqqys1Le//e2k/AIAAGkqEbdZfvjhh7F9bt++7V5//XX35JNPumnTprm1a9e6jo6Oh34NbsNmsezXWFjPzEq9NdJt2L7/njgpIxqNKhAIWI8BTGhj+bLAm5HiXp7nKTc3d9jtvBccAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYGKK9QAAUo/P57MeARMAV0AAABMECABgYlQBqqmp0eLFi5WTk6PCwkKtWbNGTU1Ncfu8+OKL8vl8cWvLli0JHRoAkP5GFaCGhgaFw2GdPXtWJ06c0N27d7VixQr19PTE7bd582Z1dHTE1r59+xI6NAAg/Y3qJoTjx4/HfXzgwAEVFhbq/PnzWrZsWezxadOmKRgMJmZCAEBGeqSfAXmeJ0nKz8+Pe/yjjz5SQUGBFixYoF27dunWrVvDPkdfX5+i0WjcAgBMAG6M+vv73fe//323dOnSuMd///vfu+PHj7tLly65P/7xj+6pp55ya9euHfZ59u7d6ySxWCwWK8OW53kP7MiYA7RlyxZXWlrq2tvbH7hfXV2dk+Sam5uH3N7b2+s8z4ut9vZ284PGYrFYrEdfIwVoTH8Rddu2bTp27JjOnDmjmTNnPnDfiooKSVJzc7PmzJlz33a/3y+/3z+WMQAAaWxUAXLO6Y033lBtba3q6+tVVlY24udcvHhRklRcXDymAQEAmWlUAQqHwzp48KCOHj2qnJwcRSIRSVIgENDUqVPV0tKigwcP6nvf+56mT5+uS5cuaceOHVq2bJkWLlyYlF8AACBNjebnPhrm+3wffvihc865trY2t2zZMpefn+/8fr+bO3eue+utt0b8PuD/8jzP/PuWLBaLxXr0NdLXft9/w5IyotGoAoGA9RgAgEfkeZ5yc3OH3c57wQEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATKRcgJxz1iMAABJgpK/nKReg7u5u6xEAAAkw0tdzn0uxS46BgQFdu3ZNOTk58vl8cdui0ahKSkrU3t6u3NxcowntcRwGcRwGcRwGcRwGpcJxcM6pu7tboVBIkyYNf50zZRxneiiTJk3SzJkzH7hPbm7uhD7BvsZxGMRxGMRxGMRxGGR9HAKBwIj7pNy34AAAEwMBAgCYSKsA+f1+7d27V36/33oUUxyHQRyHQRyHQRyHQel0HFLuJgQAwMSQVldAAIDMQYAAACYIEADABAECAJhImwDt379fTz/9tB577DFVVFTos88+sx5p3L399tvy+Xxxa/78+dZjJd2ZM2e0evVqhUIh+Xw+HTlyJG67c0579uxRcXGxpk6dqqqqKl25csVm2CQa6Ti8+uqr950f1dXVNsMmSU1NjRYvXqycnBwVFhZqzZo1ampqitunt7dX4XBY06dP1xNPPKH169ers7PTaOLkeJjj8OKLL953PmzZssVo4qGlRYA++eQT7dy5U3v37tUXX3yh8vJyrVy5UtevX7cebdw999xz6ujoiK2//OUv1iMlXU9Pj8rLy7V///4ht+/bt0/vv/++PvjgA507d06PP/64Vq5cqd7e3nGeNLlGOg6SVF1dHXd+HDp0aBwnTL6GhgaFw2GdPXtWJ06c0N27d7VixQr19PTE9tmxY4c+/fRTHT58WA0NDbp27ZrWrVtnOHXiPcxxkKTNmzfHnQ/79u0zmngYLg0sWbLEhcPh2Mf9/f0uFAq5mpoaw6nG3969e115ebn1GKYkudra2tjHAwMDLhgMunfffTf2WFdXl/P7/e7QoUMGE46Pe4+Dc85t3LjRvfTSSybzWLl+/bqT5BoaGpxzg7/3WVlZ7vDhw7F9vvzySyfJNTY2Wo2ZdPceB+ec++53v+t+/OMf2w31EFL+CujOnTs6f/68qqqqYo9NmjRJVVVVamxsNJzMxpUrVxQKhTR79my98soramtrsx7JVGtrqyKRSNz5EQgEVFFRMSHPj/r6ehUWFmrevHnaunWrbty4YT1SUnmeJ0nKz8+XJJ0/f153796NOx/mz5+vWbNmZfT5cO9x+NpHH32kgoICLViwQLt27dKtW7csxhtWyr0Z6b2++uor9ff3q6ioKO7xoqIi/eMf/zCaykZFRYUOHDigefPmqaOjQ++8845eeOEFXb58WTk5OdbjmYhEIpI05Pnx9baJorq6WuvWrVNZWZlaWlr085//XKtWrVJjY6MmT55sPV7CDQwMaPv27Vq6dKkWLFggafB8yM7OVl5eXty+mXw+DHUcJOmHP/yhSktLFQqFdOnSJf3sZz9TU1OT/vznPxtOGy/lA4T/t2rVqth/L1y4UBUVFSotLdWf/vQnbdq0yXAypIIf/OAHsf9+/vnntXDhQs2ZM0f19fVavny54WTJEQ6Hdfny5Qnxc9AHGe44vPbaa7H/fv7551VcXKzly5erpaVFc+bMGe8xh5Ty34IrKCjQ5MmT77uLpbOzU8Fg0Giq1JCXl6dnn31Wzc3N1qOY+foc4Py43+zZs1VQUJCR58e2bdt07NgxnT59Ou6fbwkGg7pz5466urri9s/U82G44zCUiooKSUqp8yHlA5Sdna1Fixaprq4u9tjAwIDq6upUWVlpOJm9mzdvqqWlRcXFxdajmCkrK1MwGIw7P6LRqM6dOzfhz4+rV6/qxo0bGXV+OOe0bds21dbW6tSpUyorK4vbvmjRImVlZcWdD01NTWpra8uo82Gk4zCUixcvSlJqnQ/Wd0E8jI8//tj5/X534MAB9/e//9299tprLi8vz0UiEevRxtVPfvITV19f71pbW91f//pXV1VV5QoKCtz169etR0uq7u5ud+HCBXfhwgUnyb333nvuwoUL7t///rdzzrlf/epXLi8vzx09etRdunTJvfTSS66srMzdvn3bePLEetBx6O7udm+++aZrbGx0ra2t7uTJk+6b3/yme+aZZ1xvb6/16AmzdetWFwgEXH19vevo6IitW7duxfbZsmWLmzVrljt16pT7/PPPXWVlpausrDScOvFGOg7Nzc3uF7/4hfv8889da2urO3r0qJs9e7ZbtmyZ8eTx0iJAzjn329/+1s2aNctlZ2e7JUuWuLNnz1qPNO42bNjgiouLXXZ2tnvqqafchg0bXHNzs/VYSXf69Gkn6b61ceNG59zgrdi7d+92RUVFzu/3u+XLl7umpibboZPgQcfh1q1bbsWKFW7GjBkuKyvLlZaWus2bN2fc/6QN9euX5D788MPYPrdv33avv/66e/LJJ920adPc2rVrXUdHh93QSTDScWhra3PLli1z+fn5zu/3u7lz57q33nrLeZ5nO/g9+OcYAAAmUv5nQACAzESAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmPg/fvtR1bWuPFIAAAAASUVORK5CYII=\n"},"metadata":{}}],"source":["plt.imshow(test_07_img, cmap=plt.get_cmap('gray'))\n","plt.show()"]},{"cell_type":"code","execution_count":43,"metadata":{"id":"PdnQPo_ziduu","executionInfo":{"status":"ok","timestamp":1758485446770,"user_tz":-180,"elapsed":48,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["test_07_img = test_07_img / 255\n","test_07_img = test_07_img.reshape(1, num_pixels)"]},{"cell_type":"code","execution_count":44,"metadata":{"id":"PJydX_A6ih1c","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485448538,"user_tz":-180,"elapsed":111,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"273ebf5b-6609-4992-c390-52afe8e38dab"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 30ms/step\n","I think it's  7\n"]}],"source":["result = model.predict(test_07_img)\n","print('I think it\\'s ', np.argmax(result))"]},{"cell_type":"code","execution_count":45,"metadata":{"id":"-vu4le7Ii1kD","executionInfo":{"status":"ok","timestamp":1758485451506,"user_tz":-180,"elapsed":468,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["from PIL import Image\n","file_05_data = Image.open('5.png')\n","file_05_data = file_05_data.convert('L') # перевод в градации серого\n","test_05_img = np.array(file_05_data)"]},{"cell_type":"code","execution_count":46,"metadata":{"id":"d9ddUAozi7Xe","colab":{"base_uri":"https://localhost:8080/","height":430},"executionInfo":{"status":"ok","timestamp":1758485452928,"user_tz":-180,"elapsed":179,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"23af732e-e652-46f5-eedb-2ed070e3d881"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.imshow(test_05_img, cmap=plt.get_cmap('gray'))\n","plt.show()"]},{"cell_type":"code","execution_count":47,"metadata":{"id":"Cg9NgtUZi-pN","executionInfo":{"status":"ok","timestamp":1758485456201,"user_tz":-180,"elapsed":47,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["test_05_img = test_05_img / 255\n","test_05_img = test_05_img.reshape(1, num_pixels)"]},{"cell_type":"code","execution_count":48,"metadata":{"id":"ZPBBAH1fjFEi","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485457453,"user_tz":-180,"elapsed":109,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"3bafabf1-4a3a-4d9c-da9c-db401fb1054d"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 28ms/step\n","I think it's  5\n"]}],"source":["result = model.predict(test_05_img)\n","print('I think it\\'s ', np.argmax(result))"]},{"cell_type":"code","execution_count":49,"metadata":{"id":"1Prj83mdlIqH","executionInfo":{"status":"ok","timestamp":1758485460398,"user_tz":-180,"elapsed":483,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["from PIL import Image\n","file_07_90_data = Image.open('7-90.png')\n","file_07_90_data = file_07_90_data.convert('L')\n","test_07_90_img = np.array(file_07_90_data)"]},{"cell_type":"code","execution_count":50,"metadata":{"id":"J-h0NoPblWHP","colab":{"base_uri":"https://localhost:8080/","height":430},"executionInfo":{"status":"ok","timestamp":1758485461709,"user_tz":-180,"elapsed":93,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"1bcc94ab-eb24-4153-9539-7ed2343754b5"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plt.imshow(test_07_90_img, cmap=plt.get_cmap('gray'))\n","plt.show()"]},{"cell_type":"code","execution_count":51,"metadata":{"id":"XDyQEozwlaan","executionInfo":{"status":"ok","timestamp":1758485464505,"user_tz":-180,"elapsed":4,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["test_07_90_img = test_07_90_img / 255\n","test_07_90_img = test_07_90_img.reshape(1, num_pixels)"]},{"cell_type":"code","execution_count":52,"metadata":{"id":"jTT1aO7HlgA0","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485466008,"user_tz":-180,"elapsed":130,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"ba458c09-2bc8-47cf-c8ff-e5fa95111baf"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 40ms/step\n","I think it's  2\n"]}],"source":["result = model.predict(test_07_90_img)\n","print('I think it\\'s ', np.argmax(result))"]},{"cell_type":"code","execution_count":53,"metadata":{"id":"r151rpruli75","executionInfo":{"status":"ok","timestamp":1758485468248,"user_tz":-180,"elapsed":435,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["from PIL import Image\n","file_05_90_data = Image.open('5-90.png')\n","file_05_90_data = file_05_90_data.convert('L')\n","test_05_90_img = np.array(file_05_90_data)"]},{"cell_type":"code","execution_count":54,"metadata":{"id":"voOAg_CslqvJ","colab":{"base_uri":"https://localhost:8080/","height":430},"executionInfo":{"status":"ok","timestamp":1758485469423,"user_tz":-180,"elapsed":254,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"9be9b21c-493c-4b96-e512-a54a33f0c3c6"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAaAAAAGdCAYAAABU0qcqAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAGJRJREFUeJzt3X9MVff9x/HXVeFWW7gUES63IkVtNamVZU4ZcXVNJIpbTP3xh+v6h12MjfbaTF27xSVquyxhs0mzdDHr/tIsq7YzGZr6h4miYLahTa3GmHVEGBsYubiacC6ioIHP9w/W+92tIAL38r73+nwkn6Tce7j37bknPHu5x6PPOecEAMAEm2Q9AADg0USAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACAiSnWA3zdwMCArl+/rpycHPl8PutxAACj5JxTd3e3QqGQJk0a/n1OygXo+vXrKikpsR4DADBO7e3tmjlz5rD3p9yv4HJycqxHAAAkwEg/z5MWoP379+vpp5/WY489poqKCn366acP9X382g0AMsNIP8+TEqCPP/5YO3fu1N69e/X555+rvLxcK1eu1I0bN5LxdACAdOSSYMmSJS4cDse+7u/vd6FQyNXU1Iz4vZ7nOUksFovFSvPled4Df94n/B3Q3bt3deHCBVVVVcVumzRpkqqqqtTY2Hjf9n19fYpGo3ELAJD5Eh6gL7/8Uv39/SoqKoq7vaioSJFI5L7ta2pqFAgEYosz4ADg0WB+FtyuXbvkeV5stbe3W48EAJgACf97QAUFBZo8ebI6Ozvjbu/s7FQwGLxve7/fL7/fn+gxAAApLuHvgLKzs7Vo0SLV1dXFbhsYGFBdXZ0qKysT/XQAgDSVlCsh7Ny5Uxs3btS3vvUtLVmyRL/5zW/U09OjH/3oR8l4OgBAGkpKgDZs2KD//Oc/2rNnjyKRiL7xjW/oxIkT952YAAB4dPmcc856iP8VjUYVCASsxwAAjJPnecrNzR32fvOz4AAAjyYCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGAi4QF6++235fP54tb8+fMT/TQAgDQ3JRkP+txzz+nUqVP//yRTkvI0AIA0lpQyTJkyRcFgMBkPDQDIEEn5DOjq1asKhUKaPXu2XnnlFbW1tQ27bV9fn6LRaNwCAGS+hAeooqJCBw8e1IkTJ/S73/1Ora2teuGFF9Td3T3k9jU1NQoEArFVUlKS6JEAACnI55xzyXyCrq4ulZaW6r333tOmTZvuu7+vr099fX2xr6PRKBECgAzgeZ5yc3OHvT/pZwfk5eXp2WefVXNz85D3+/1++f3+ZI8BAEgxSf97QLdu3VJLS4uKi4uT/VQAgDSS8AC9+eabamho0L/+9S/97W9/09q1azV58mS9/PLLiX4qAEAaS/iv4K5du6aXX35ZN2/e1IwZM/Sd73xH586d04wZMxL9VACANJb0kxBGKxqNKhAIWI8BABinkU5C4FpwAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYGKK9QBILOfcqL/H5/MlYRIAeDDeAQEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJrgYKfA/xnIxV0wsLp6bOXgHBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCY4GKkwDhxccyx4+KvjzbeAQEATBAgAICJUQfo7NmzWr16tUKhkHw+n44ePRp3v3NOe/bsUXFxsaZOnaqqqipdvXo1UfMCADLEqAPU09Oj8vJy7d+/f8j79+3bp/fff18ffPCBzp8/r8cff1wrV65Ub2/vuIcFAGQQNw6SXG1tbezrgYEBFwwG3bvvvhu7raury/n9fnf48OGHekzP85wk1hjXWF9HFvuP/c1K9PI874GvZUI/A2ptbVUkElFVVVXstkAgoIqKCjU2Ng75PX19fYpGo3ELAJD5EhqgSCQiSSoqKoq7vaioKHbf19XU1CgQCMRWSUlJIkcCAKQo87Pgdu3aJc/zYqu9vd16JADABEhogILBoCSps7Mz7vbOzs7YfV/n9/uVm5sbtwAAmS+hASorK1MwGFRdXV3stmg0qvPnz6uysjKRTwUASHOjvhTPrVu31NzcHPu6tbVVly5dUn5+vmbNmqXt27frl7/8pZ555hmVlZVp9+7dCoVCWrNmTSLnBgCku9GeAnnmzJkhT7fbuHGjc27wVOzdu3e7oqIi5/f73fLly11TU9NDPz6nYY9vjYX1zKm02H/sb1bi1kinYfv++4KmjGg0qkAgYD1G2hrLy5mJF9OcyMM6E/ffRJmo14nXyIbneQ/8XN/8LDgAwKOJAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJkb97wEBmYyrJk+ssezvFLuAP8aBd0AAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABNTrAfA8Jxz1iMAQNLwDggAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYGHWAzp49q9WrVysUCsnn8+no0aNx97/66qvy+Xxxq7q6OlHzAgAyxKgD1NPTo/Lycu3fv3/Ybaqrq9XR0RFbhw8fHteQAIDMM+p/EXXVqlVatWrVA7fx+/0KBoNjHgoAkPmS8hlQfX29CgsLNW/ePG3dulU3b94cdtu+vj5Fo9G4BQDIfAkPUHV1tf7whz+orq5Ov/71r9XQ0KBVq1apv79/yO1ramoUCARiq6SkJNEjAQBSkM8558b8zT6famtrtWbNmmG3+ec//6k5c+bo1KlTWr58+X339/X1qa+vL/Z1NBolQv81jpdmVHw+34Q8z0Qa677LxH2Racby2vK62vA8T7m5ucPen/TTsGfPnq2CggI1NzcPeb/f71dubm7cAgBkvqQH6Nq1a7p586aKi4uT/VQAgDQy6rPgbt26FfduprW1VZcuXVJ+fr7y8/P1zjvvaP369QoGg2ppadFPf/pTzZ07VytXrkzo4ACANOdG6cyZM07SfWvjxo3u9u3bbsWKFW7GjBkuKyvLlZaWus2bN7tIJPLQj+953pCP/yiuiWL950ylfWc9Nys5r631zI/q8jzvga/LuE5CSIZoNKpAIGA9RtpKsZcz7fBhdeobyzHO62rD/CQEAACGQoAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABOj/veAkNoy8aq/E3mFb660DEwc3gEBAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACa4GClSHhf7zFwTeaFZpB7eAQEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJrgYKYD7pPJFQrk4bebgHRAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIKLkQK4Dxf8xETgHRAAwAQBAgCYGFWAampqtHjxYuXk5KiwsFBr1qxRU1NT3Da9vb0Kh8OaPn26nnjiCa1fv16dnZ0JHRoAkP5GFaCGhgaFw2GdO3dOJ0+e1L1797RixQr19PTEttmxY4c++eQTHTlyRA0NDbp+/brWrVuX8MEBAGnOjcONGzecJNfQ0OCcc66rq8tlZWW5I0eOxLb54osvnCTX2Nj4UI/peZ6TxGKxWKw0X57nPfDn/bg+A/I8T5KUn58vSbpw4YLu3bunqqqq2Dbz58/XrFmz1NjYOORj9PX1KRqNxi0AQOYbc4AGBga0fft2LV26VAsWLJAkRSIRZWdnKy8vL27boqIiRSKRIR+npqZGgUAgtkpKSsY6EgAgjYw5QOFwWFeuXNFHH300rgF27dolz/Niq729fVyPBwBID2P6i6jbtm3T8ePHdfbsWc2cOTN2ezAY1N27d9XV1RX3Lqizs1PBYHDIx/L7/fL7/WMZAwCQxkb1Dsg5p23btqm2tlanT59WWVlZ3P2LFi1SVlaW6urqYrc1NTWpra1NlZWViZkYAJARRvUOKBwO69ChQzp27JhycnJin+sEAgFNnTpVgUBAmzZt0s6dO5Wfn6/c3Fy98cYbqqys1Le//e2k/AEAAGlqNKdda5hT7Q4cOBDb5s6dO+711193Tz75pJs2bZpbu3at6+joeOjn4DRsFovFyow10mnYvv+GJWVEo1EFAgHrMQAA4+R5nnJzc4e9n2vBAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAE6MKUE1NjRYvXqycnBwVFhZqzZo1ampqitvmxRdflM/ni1tbtmxJ6NAAgPQ3qgA1NDQoHA7r3LlzOnnypO7du6cVK1aop6cnbrvNmzero6Mjtvbt25fQoQEA6W/KaDY+ceJE3NcHDx5UYWGhLly4oGXLlsVunzZtmoLBYGImBABkpHF9BuR5niQpPz8/7vYPP/xQBQUFWrBggXbt2qXbt28P+xh9fX2KRqNxCwDwCHBj1N/f777//e+7pUuXxt3++9//3p04ccJdvnzZ/fGPf3RPPfWUW7t27bCPs3fvXieJxWKxWBm2PM97YEfGHKAtW7a40tJS197e/sDt6urqnCTX3Nw85P29vb3O87zYam9vN99pLBaLxRr/GilAo/oM6Cvbtm3T8ePHdfbsWc2cOfOB21ZUVEiSmpubNWfOnPvu9/v98vv9YxkDAJDGRhUg55zeeOMN1dbWqr6+XmVlZSN+z6VLlyRJxcXFYxoQAJCZRhWgcDisQ4cO6dixY8rJyVEkEpEkBQIBTZ06VS0tLTp06JC+973vafr06bp8+bJ27NihZcuWaeHChUn5AwAA0tRoPvfRML/nO3DggHPOuba2Nrds2TKXn5/v/H6/mzt3rnvrrbdG/D3g//I8z/z3liwWi8Ua/xrpZ7/vv2FJGdFoVIFAwHoMAMA4eZ6n3NzcYe/nWnAAABMECABgggABAEwQIACACQIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADABAECAJggQAAAEwQIAGCCAAEATBAgAIAJAgQAMEGAAAAmCBAAwAQBAgCYIEAAABMpFyDnnPUIAIAEGOnnecoFqLu723oEAEACjPTz3OdS7C3HwMCArl+/rpycHPl8vrj7otGoSkpK1N7ertzcXKMJ7bEfBrEfBrEfBrEfBqXCfnDOqbu7W6FQSJMmDf8+Z8oEzvRQJk2apJkzZz5wm9zc3Ef6APsK+2EQ+2EQ+2EQ+2GQ9X4IBAIjbpNyv4IDADwaCBAAwERaBcjv92vv3r3y+/3Wo5hiPwxiPwxiPwxiPwxKp/2QcichAAAeDWn1DggAkDkIEADABAECAJggQAAAE2kToP379+vpp5/WY489poqKCn366afWI024t99+Wz6fL27Nnz/feqykO3v2rFavXq1QKCSfz6ejR4/G3e+c0549e1RcXKypU6eqqqpKV69etRk2iUbaD6+++up9x0d1dbXNsElSU1OjxYsXKycnR4WFhVqzZo2ampritunt7VU4HNb06dP1xBNPaP369ers7DSaODkeZj+8+OKL9x0PW7ZsMZp4aGkRoI8//lg7d+7U3r179fnnn6u8vFwrV67UjRs3rEebcM8995w6Ojpi6y9/+Yv1SEnX09Oj8vJy7d+/f8j79+3bp/fff18ffPCBzp8/r8cff1wrV65Ub2/vBE+aXCPtB0mqrq6OOz4OHz48gRMmX0NDg8LhsM6dO6eTJ0/q3r17WrFihXp6emLb7NixQ5988omOHDmihoYGXb9+XevWrTOcOvEeZj9I0ubNm+OOh3379hlNPAyXBpYsWeLC4XDs6/7+fhcKhVxNTY3hVBNv7969rry83HoMU5JcbW1t7OuBgQEXDAbdu+++G7utq6vL+f1+d/jwYYMJJ8bX94Nzzm3cuNG99NJLJvNYuXHjhpPkGhoanHODr31WVpY7cuRIbJsvvvjCSXKNjY1WYybd1/eDc85997vfdT/+8Y/thnoIKf8O6O7du7pw4YKqqqpit02aNElVVVVqbGw0nMzG1atXFQqFNHv2bL3yyitqa2uzHslUa2urIpFI3PERCARUUVHxSB4f9fX1Kiws1Lx587R161bdvHnTeqSk8jxPkpSfny9JunDhgu7duxd3PMyfP1+zZs3K6OPh6/vhKx9++KEKCgq0YMEC7dq1S7dv37YYb1gpdzHSr/vyyy/V39+voqKiuNuLior0j3/8w2gqGxUVFTp48KDmzZunjo4OvfPOO3rhhRd05coV5eTkWI9nIhKJSNKQx8dX9z0qqqurtW7dOpWVlamlpUU///nPtWrVKjU2Nmry5MnW4yXcwMCAtm/frqVLl2rBggWSBo+H7Oxs5eXlxW2bycfDUPtBkn74wx+qtLRUoVBIly9f1s9+9jM1NTXpz3/+s+G08VI+QPh/q1ativ33woULVVFRodLSUv3pT3/Spk2bDCdDKvjBD34Q++/nn39eCxcu1Jw5c1RfX6/ly5cbTpYc4XBYV65ceSQ+B32Q4fbDa6+9Fvvv559/XsXFxVq+fLlaWlo0Z86ciR5zSCn/K7iCggJNnjz5vrNYOjs7FQwGjaZKDXl5eXr22WfV3NxsPYqZr44Bjo/7zZ49WwUFBRl5fGzbtk3Hjx/XmTNn4v75lmAwqLt376qrqytu+0w9HobbD0OpqKiQpJQ6HlI+QNnZ2Vq0aJHq6upitw0MDKiurk6VlZWGk9m7deuWWlpaVFxcbD2KmbKyMgWDwbjjIxqN6vz584/88XHt2jXdvHkzo44P55y2bdum2tpanT59WmVlZXH3L1q0SFlZWXHHQ1NTk9ra2jLqeBhpPwzl0qVLkpRax4P1WRAP46OPPnJ+v98dPHjQ/f3vf3evvfaay8vLc5FIxHq0CfWTn/zE1dfXu9bWVvfXv/7VVVVVuYKCAnfjxg3r0ZKqu7vbXbx40V28eNFJcu+99567ePGi+/e//+2cc+5Xv/qVy8vLc8eOHXOXL192L730kisrK3N37twxnjyxHrQfuru73ZtvvukaGxtda2urO3XqlPvmN7/pnnnmGdfb22s9esJs3brVBQIBV19f7zo6OmLr9u3bsW22bNniZs2a5U6fPu0+++wzV1lZ6SorKw2nTryR9kNzc7P7xS9+4T777DPX2trqjh075mbPnu2WLVtmPHm8tAiQc8799re/dbNmzXLZ2dluyZIl7ty5c9YjTbgNGza44uJil52d7Z566im3YcMG19zcbD1W0p05c8ZJum9t3LjROTd4Kvbu3btdUVGR8/v9bvny5a6pqcl26CR40H64ffu2W7FihZsxY4bLyspypaWlbvPmzRn3P2lD/fkluQMHDsS2uXPnjnv99dfdk08+6aZNm+bWrl3rOjo67IZOgpH2Q1tbm1u2bJnLz893fr/fzZ0717311lvO8zzbwb+Gf44BAGAi5T8DAgBkJgIEADBBgAAAJggQAMAEAQIAmCBAAAATBAgAYIIAAQBMECAAgAkCBAAwQYAAACYIEADAxP8B7e1+qRzfcncAAAAASUVORK5CYII=\n"},"metadata":{}}],"source":["plt.imshow(test_05_90_img, cmap=plt.get_cmap('gray'))\n","plt.show()"]},{"cell_type":"code","execution_count":55,"metadata":{"id":"goK2uhgwlvAF","executionInfo":{"status":"ok","timestamp":1758485472543,"user_tz":-180,"elapsed":59,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}}},"outputs":[],"source":["test_05_90_img = test_05_90_img / 255\n","test_05_90_img = test_05_90_img.reshape(1, num_pixels)"]},{"cell_type":"code","execution_count":56,"metadata":{"id":"uqtsomP7lxTu","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1758485473980,"user_tz":-180,"elapsed":146,"user":{"displayName":"Mi Ri","userId":"10370067304318068772"}},"outputId":"b7ceb4a9-05d2-432e-d147-8b757532c0f4"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 29ms/step\n","I think it's  4\n"]}],"source":["result = model.predict(test_05_90_img)\n","print('I think it\\'s ', np.argmax(result))"]}],"metadata":{"accelerator":"GPU","colab":{"gpuType":"T4","provenance":[],"mount_file_id":"150ibV5FVGjhfi9jyOYyFoMLj6A0FBqNq","authorship_tag":"ABX9TyNZQfksCD0/vXsU8GuGWaJ1"},"kernelspec":{"display_name":"Python 3","name":"python3"},"language_info":{"name":"python"}},"nbformat":4,"nbformat_minor":0}
\ No newline at end of file
diff --git a/labworks/LW1/Report.md b/labworks/LW1/Report.md
new file mode 100644
index 0000000..023824c
--- /dev/null
+++ b/labworks/LW1/Report.md
@@ -0,0 +1,692 @@
+# Отчёт по лабораторной работе №1
+## по теме: "Архитектура и обучение глубоких нейронных сетей"
+
+---
+Выполнили: Бригада 2, Мачулина Д.В., Бирюкова А.С.
+
+---
+### 1. Создание  блокнота в Google Collab и настройка директории
+```python
+import os 
+os.chdir('/content/drive/MyDrive/Colab Notebooks')
+```
+**Импорт библиотек** 
+```python
+from tensorflow import keras 
+import matplotlib.pyplot as plt 
+import numpy as np 
+import sklearn
+```
+
+---
+### 2. Загрузка набора данных MNIST
+```python
+from keras.datasets import mnist 
+(X_train, y_train), (X_test, y_test) = mnist.load_data()
+```
+___
+### 3. Разбиение набора данных на обучающие и тестовые данные
+```python
+from sklearn.model_selection import train_test_split 
+```
+**Объединение обучающих и тестовых данных в один набор**
+```python
+X = np.concatenate((X_train, X_test)) 
+y = np.concatenate((y_train, y_test))
+```
+**Разбиение набора случайным образом (номер бригады - 2)**
+```python
+X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 10000, train_size = 60000, random_state = 7) 
+```
+**Вывод размерностей**
+```python
+print('Shape of X train:', X_train.shape) 
+print('Shape of y train:', y_train.shape) 
+```
+*Shape of X train: (60000, 28, 28);
+Shape of y train: (60000,)*
+
+### 4. Вывод элементов обучающих данных
+```python
+fig, axes = plt.subplots(1, 4, figsize=(10, 3))
+
+for i in range(4):
+    axes[i].imshow(X_train[i], cmap=plt.get_cmap('gray'))
+    axes[i].set_title(f'Label: {y_train[i]}')  
+
+plt.show()
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/3576.png)
+
+---
+### 5. Предобработка данных
+**Преобразование данных из массива в вектор**
+```python
+num_pixels = X_train.shape[1] * X_train.shape[2] 
+X_train = X_train.reshape(X_train.shape[0], num_pixels) / 255 
+X_test = X_test.reshape(X_test.shape[0], num_pixels) / 255 
+print('Shape of transformed X train:', X_train.shape)
+```
+*Shape of transformed X train: (60000, 784)*
+
+**Кодировка метод цифр по принципу one-hot encoding**
+```python
+from keras.utils import to_categorical 
+y_train = to_categorical(y_train) 
+y_test = to_categorical(y_test) 
+print('Shape of transformed y train:', y_train.shape) 
+num_classes = y_train.shape[1]
+```
+*Shape of transformed y train: (60000, 10)*
+
+---
+### 6. Реализация модели нейронной сети
+```python
+from keras.models import Sequential 
+from keras.layers import Dense 
+```
+
+**Создание и компиляция модели**
+```python
+model_01 = Sequential()
+model_01.add(Dense(units=num_classes,input_dim=num_pixels, activation='softmax'))
+model_01.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy']) 
+model_01.summary()
+```
+*Model: "sequential_5"*
+<table>
+    <thead>
+        <tr>
+            <th>Layer (type)</th>
+            <th>Output Shape</th>
+            <th>Param #</th>
+        </tr>
+    </thead>
+    <tbody>
+        <tr>
+            <td rowspan=1 align="center">dense_10 (Dense)</td>
+            <td rowspan=2 align="center">(None, 10)</td>
+            <td align="center">7,850</td>
+        </tr>
+    </tbody>
+</table>
+
+*Total params: 7,850 (30.66 KB)*
+
+*Trainable params: 7,850 (30.66 KB)*
+
+*Non-trainable params: 0 (0.00 B)*
+
+**Обучение модели**
+```python
+H = model_01.fit(
+    X_train, y_train,
+    validation_split=0.1,
+    epochs=100,
+    batch_size = 512
+)
+```
+**Вывод графика ошибки**
+```python
+plt.figure(figsize=(12, 5))
+
+plt.subplot(1, 2, 1)
+plt.plot(H.history['loss'], label='Обучающая ошибка')
+plt.plot(H.history['val_loss'], label='Валидационная ошибка')
+plt.title('Функция ошибки по эпохам')
+plt.xlabel('Epochs')
+plt.ylabel('loss')
+plt.legend()
+plt.grid(True)
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/model_01.png)
+
+### 7. Применение модели к тестовым данным
+```python
+scores=model_01.evaluate(X_test,y_test)
+print('Loss on test data:', scores[0]) 
+print('Accuracy on test data:', scores[1])
+```
+*Loss on test data: 0.3511466085910797;*
+
+*Accuracy on test data: 0.9067999720573425*
+
+### 8. Повторные эксперименты с добавлением первого скрытого слоя
+**100 нейронов в первом скрытом слое:**
+```python
+model_01_100 = Sequential()
+model_01_100.add(Dense(units=100,input_dim=num_pixels, activation='sigmoid'))
+model_01_100.add(Dense(units=num_classes, activation='softmax'))
+model_01_100.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])
+model_01_100.summary()
+```
+*Model: "sequential_6"*
+<table>
+    <thead>
+        <tr>
+            <th>Layer (type)</th>
+            <th>Output Shape</th>
+            <th>Param #</th>
+        </tr>
+    </thead>
+    <tbody>
+        <tr>
+            <td align="center">dense_11 (Dense)</td>
+            <td align="center">(None, 100)</td>
+            <td align="center">78,500</td>
+        </tr>
+        <tr>
+        <td align="center">dense_12(Dense)</td>
+        <td align="center">(None,10)</td>
+        <td align="center">1,010</td>
+        </tr>
+    </tbody>
+</table>
+
+*Total params: 79,510 (310.59 KB)*
+
+*Trainable params: 79,510 (310.59 KB)*
+
+*Non-trainable params: 0 (0.00 B)*
+```python
+H_01_100 = model_01_100.fit(
+    X_train, y_train,
+    validation_split=0.1,
+    epochs=100,
+    batch_size = 512
+)
+```
+```python
+plt.figure(figsize=(12, 5))
+
+plt.subplot(1, 2, 1)
+plt.plot(H_01_100.history['loss'], label='Обучающая ошибка')
+plt.plot(H_01_100.history['val_loss'], label='Валидационная ошибка')
+plt.title('Функция ошибки по эпохам')
+plt.xlabel('Epochs')
+plt.ylabel('loss')
+plt.legend()
+plt.grid(True)
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/model_01_100.png)
+```python
+scores_01_100=model_01_100.evaluate(X_test,y_test)
+print('Loss on test data:', scores_01_100[0]) 
+print('Accuracy on test data:', scores_01_100[1])
+```
+*Loss on test data: 0.3824511766433716*
+
+*Accuracy on test data: 0.9000999927520752*
+
+**300 нейронов в первом скрытом слое**
+```python
+model_01_300 = Sequential()
+model_01_300.add(Dense(units=300,input_dim=num_pixels, activation='sigmoid'))
+model_01_300.add(Dense(units=num_classes, activation='softmax'))
+model_01_300.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])
+model_01_300.summary()
+```
+*Model: "sequential_7"*
+<table>
+    <thead>
+        <tr>
+            <th>Layer (type)</th>
+            <th>Output Shape</th>
+            <th>Param #</th>
+        </tr>
+    </thead>
+    <tbody>
+        <tr>
+            <td align="center">dense_13 (Dense)</td>
+            <td align="center">(None, 300)</td>
+            <td align="center">235,500</td>
+        </tr>
+        <tr>
+        <td align="center">dense_14(Dense)</td>
+        <td align="center">(None,10)</td>
+        <td align="center">3,010</td>
+        </tr>
+    </tbody>
+</table>
+
+*Total params: 238,510 (931.68 KB)*
+
+*Trainable params: 238,510 (931.68 KB)*
+
+*Non-trainable params: 0 (0.00 B)*
+
+```python
+H_01_300 = model_01_300.fit(
+    X_train, y_train,
+    validation_split=0.1,
+    epochs=100,
+    batch_size = 512
+)
+```
+```python
+plt.figure(figsize=(12, 5))
+
+plt.subplot(1, 2, 1)
+plt.plot(H_01_300.history['loss'], label='Обучающая ошибка')
+plt.plot(H_01_300.history['val_loss'], label='Валидационная ошибка')
+plt.title('Функция ошибки по эпохам')
+plt.xlabel('Epochs')
+plt.ylabel('loss')
+plt.legend()
+plt.grid(True)
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/model_01_300.png)
+```python
+scores_01_300=model_01_300.evaluate(X_test,y_test)
+print('Loss on test data:', scores_01_300[0]) 
+print('Accuracy on test data:', scores_01_300[1])
+```
+
+*Loss on test data: 0.37091827392578125*
+
+*Accuracy on test data: 0.9013000130653381*
+
+**500 нейронов в первом скрытом слое**
+```python
+model_01_500 = Sequential()
+model_01_500.add(Dense(units=500,input_dim=num_pixels, activation='sigmoid'))
+model_01_500.add(Dense(units=num_classes, activation='softmax'))
+model_01_500.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])
+model_01_500.summary()
+```
+*Model: "sequential_8"*
+<table>
+    <thead>
+        <tr>
+            <th>Layer (type)</th>
+            <th>Output Shape</th>
+            <th>Param #</th>
+        </tr>
+    </thead>
+    <tbody>
+        <tr>
+            <td align="center">dense_15 (Dense)</td>
+            <td align="center">(None, 500)</td>
+            <td align="center">392,500</td>
+        </tr>
+        <tr>
+        <td align="center">dense_16(Dense)</td>
+        <td align="center">(None,10)</td>
+        <td align="center">5,010</td>
+        </tr>
+    </tbody>
+</table>
+ 
+*Total params: 397,510 (1.52 MB)*
+
+*Trainable params: 397,510 (1.52 MB)*
+
+*Non-trainable params: 0 (0.00 B)*
+
+```python
+H_01_500 = model_01_500.fit(
+    X_train, y_train,
+    validation_split=0.1,
+    epochs=100,
+    batch_size = 512
+)
+```
+```python
+plt.figure(figsize=(12, 5))
+
+plt.subplot(1, 2, 1)
+plt.plot(H_01_500.history['loss'], label='Обучающая ошибка')
+plt.plot(H_01_500.history['val_loss'], label='Валидационная ошибка')
+plt.title('Функция ошибки по эпохам')
+plt.xlabel('Epochs')
+plt.ylabel('loss')
+plt.legend()
+plt.grid(True)
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/model_01_500.png)
+```python
+scores_01_500=model_01_500.evaluate(X_test,y_test)
+print('Loss on test data:',scores_01_500[0]) 
+print('Accuracy on test data:',scores_01_500[1]) 
+```
+*Loss on test data: 0.36660370230674744*
+
+*Accuracy on test data: 0.9010000228881836*
+
+Таким образом, наиболее точной архитектурой со скрытым слоем является архитектура со 300 нейронами в скрытом слое. Для дальнейшей работы будем использовать её.
+
+### 9. Повторные эксперименты с добавлением второго скрытого слоя
+**50 нейронов во втором скрытом слое**
+```python
+model_01_300_50 = Sequential()
+model_01_300_50.add(Dense(units=300, input_dim=num_pixels, activation='sigmoid'))
+model_01_300_50.add(Dense(units=50, activation='sigmoid'))
+model_01_300_50.add(Dense(units=num_classes, activation='softmax'))
+model_01_300_50.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])
+model_01_300_50.summary()
+```
+*Model: "sequential_11"*
+<table>
+    <thead>
+        <tr>
+            <th>Layer (type)</th>
+            <th>Output Shape</th>
+            <th>Param #</th>
+        </tr>
+    </thead>
+    <tbody>
+        <tr>
+            <td align="center">dense_23 (Dense)</td>
+            <td align="center">(None, 300)</td>
+            <td align="center">235,500</td>
+        </tr>
+        <tr>
+        <td align="center">dense_24(Dense)</td>
+        <td align="center">(None,50)</td>
+        <td align="center">15,050</td>
+        </tr>
+        <tr>
+        <td align="center">dense_25 (Dense)</td>
+        <td align="center">(None,10)</td>
+        <td align="center">510</td>
+        </tr>
+    </tbody>
+</table>
+
+*Total params: 251,060 (328.36 KB)*
+
+*Trainable params: 251,060 (328.36 KB)*
+
+*Non-trainable params: 0 (0.00 B)*
+
+```python
+H_01_300_50 = model_01_300_50.fit(
+    X_train, y_train,
+    validation_split=0.1,
+    epochs=100,
+    batch_size=512
+)
+```
+```python
+plt.figure(figsize=(12, 5))
+
+plt.subplot(1, 2, 1)
+plt.plot(H_01_300_50.history['loss'], label='Обучающая ошибка')
+plt.plot(H_01_300_50.history['val_loss'], label='Валидационная ошибка')
+plt.title('Функция ошибки по эпохам')
+plt.xlabel('Epochs')
+plt.ylabel('loss')
+plt.legend()
+plt.grid(True)
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/model_01_300_50.png)
+```python
+scores_01_300_50=model_01_300_50.evaluate(X_test,y_test)
+print('Loss on test data:',scores_01_300_50[0])
+print('Accuracy on test data:',scores_01_300_50[1])
+```
+*Loss on test data: 0.4881931245326996*
+
+*Accuracy on test data: 0.8740000128746033*
+
+**100 нейронов во втором скрытом слое**
+```python
+model_01_300_100 = Sequential()
+model_01_300_100.add(Dense(units=300, input_dim=num_pixels, activation='sigmoid'))
+model_01_300_100.add(Dense(units=100, activation='sigmoid'))
+model_01_300_100.add(Dense(units=num_classes, activation='softmax'))
+model_01_300_100.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])
+model_01_300_100.summary()
+```
+*Model: "sequential_12"*
+<table>
+    <thead>
+        <tr>
+            <th>Layer (type)</th>
+            <th>Output Shape</th>
+            <th>Param #</th>
+        </tr>
+    </thead>
+    <tbody>
+        <tr>
+            <td align="center">dense_26 (Dense)</td>
+            <td align="center">(None, 300)</td>
+            <td align="center">235,500</td>
+        </tr>
+        <tr>
+        <td align="center">dense_27(Dense)</td>
+        <td align="center">(None,100)</td>
+        <td align="center">30,100</td>
+        </tr>
+        <tr>
+        <td align="center">dense_28 (Dense)</td>
+        <td align="center">(None,10)</td>
+        <td align="center">1,010</td>
+        </tr>
+    </tbody>
+</table>
+
+*Total params: 266,610 (350.04 KB)*
+
+*Trainable params: 266,610 (350.04 KB)*
+ 
+*Non-trainable params: 0 (0.00 B)*
+
+```python
+H_01_300_100 = model_01_300_100.fit(
+    X_train, y_train,
+    validation_split=0.1,
+    epochs=100,
+    batch_size=512
+)
+```
+```python
+plt.figure(figsize=(12, 5))
+
+plt.subplot(1, 2, 1)
+plt.plot(H_01_300_100.history['loss'], label='Обучающая ошибка')
+plt.plot(H_01_300_100.history['val_loss'], label='Валидационная ошибка')
+plt.title('Функция ошибки по эпохам')
+plt.xlabel('Epochs')
+plt.ylabel('loss')
+plt.legend()
+plt.grid(True)
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/model_01_300_100.png)
+```python
+scores_01_300_100=model_01_300_100.evaluate(X_test,y_test)
+print('Lossontestdata:',scores_01_300_100[0])
+print('Accuracyontestdata:',scores_01_300_100[1])
+```
+*Loss on test data: 0.4638420343399048*
+
+*Accuracy on test data: 0.8795999884605408*
+
+**Сведём результаты в таблицу**
+<table>
+    <thead>
+        <tr>
+            <th>Количество скрытых слоёв (type)</th>
+            <th>Количество нейронов в первом скрытом слое</th>
+            <th>Количество нейронов во втором скрытом слое </th>
+            <th>Значение метрики качества классификации</th>
+        </tr>
+    </thead>
+    <tbody>
+        <tr>
+            <td align="center">0</td>
+            <td align="center">-</td>
+            <td align="center">-</td>
+            <td align="center">0.9067999720573425</td>
+        </tr>
+        <tr>
+        <td rowspan = 3 align="center">1</td>
+        <td align="center">100</td>
+        <td rowspan =3 align="center">-</td>
+        <td align="center">0.9000999927520752</td>
+        </tr>
+        <tr>
+        <td align="center">300</td>
+        <td align="center">0.9013000130653381</td>
+        </tr>
+        <tr>
+        <td align="center">500</td>
+        <td align="center">0.9010000228881836</td>
+        </tr>
+        <tr>
+        <td rowspan = 2 align="center">2</td>
+        <td rowspan = 2 align="center">300</td>
+        <td  align="center">50</td>
+        <td  align="center">0.8740000128746033</td>
+        </tr>
+        <tr>
+        <td  align="center">100</td>
+        <td  align="center">0.8795999884605408</td>
+        </tr>
+    </tbody>
+</table>
+
+### 11.Сохранение лучшей модели на диск
+```python
+model_01_300.save(filepath='best_model.keras')
+```
+
+### 12. Вывод тестовых изображений
+**Загрузка лучшей модели с диска**
+```python
+from keras.models import load_model 
+model = load_model('best_model.keras') 
+```
+**Вывод изображений**
+```python
+n = 123
+result = model.predict(X_test[n:n+1]) 
+print('NN output:', result) 
+plt.imshow(X_test[n].reshape(28,28), cmap=plt.get_cmap('gray')) 
+plt.show() 
+print('Real mark: ', str(np.argmax(y_test[n]))) 
+print('NN answer: ', str(np.argmax(result)))
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/12.1.png)
+
+*Real mark:  6*
+
+*NN answer:  6*
+
+```python
+n = 765
+result = model.predict(X_test[n:n+1]) 
+print('NN output:', result) 
+plt.imshow(X_test[n].reshape(28,28), cmap=plt.get_cmap('gray')) 
+plt.show() 
+print('Real mark: ', str(np.argmax(y_test[n]))) 
+print('NN answer: ', str(np.argmax(result)))
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/12.2.png)
+
+*Real mark:  3*
+
+*NN answer:  3*
+
+### 13. Тестирование на собственных изображениях
+**Загрузка собственного изображения**
+```python
+from PIL import Image 
+file_07_data = Image.open('7.png') 
+file_07_data = file_07_data.convert('L') 
+test_07_img = np.array(file_07_data) 
+```
+**Вывод изображения**
+```python
+plt.imshow(test_07_img, cmap=plt.get_cmap('gray')) 
+plt.show() 
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/7.png)
+
+**Распознавание изображения**
+```python
+test_07_img = test_07_img / 255 
+test_07_img = test_07_img.reshape(1, num_pixels) 
+```
+*I think it's  7*
+
+**Второе изображение**
+```python
+from PIL import Image 
+file_05_data = Image.open('5.png') 
+file_05_data = file_05_data.convert('L') 
+test_05_img = np.array(file_05_data) 
+```
+```python
+plt.imshow(test_05_img, cmap=plt.get_cmap('gray')) 
+plt.show()
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/5.png)
+
+
+```python
+test_05_img = test_05_img / 255 
+test_05_img = test_05_img.reshape(1, num_pixels) 
+```
+```python
+result = model.predict(test_05_img) 
+print('I think it\'s ', np.argmax(result))
+```
+*I think it's  5*
+
+Нейросеть распознала изображения корректно
+
+### 14. Тестирование на собственных перевёрнутых изображениях
+**Первое изображение**
+```python
+from PIL import Image 
+file_07_90_data = Image.open('7-90.png') 
+file_07_90_data = file_07_90_data.convert('L') 
+test_07_90_img = np.array(file_07_90_data) 
+```
+```python
+plt.imshow(test_07_90_img, cmap=plt.get_cmap('gray')) 
+plt.show() 
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/7-90.png)
+
+
+```python
+test_07_90_img = test_07_90_img / 255 
+test_07_90_img = test_07_90_img.reshape(1, num_pixels) 
+```
+```python
+result = model.predict(test_07_90_img) 
+print('I think it\'s ', np.argmax(result))
+```
+*I think it's  2*
+
+**Второе изображение**
+```python
+from PIL import Image 
+file_05_90_data = Image.open('5-90.png') 
+file_05_90_data = file_05_90_data.convert('L') 
+test_05_90_img = np.array(file_05_90_data) 
+```
+```python
+plt.imshow(test_05_90_img, cmap=plt.get_cmap('gray')) 
+plt.show() 
+```
+![](http://uit.mpei.ru/git/MachulinaDV/is_dnn/raw/branch/main/labworks/LW1/5-90.png)
+
+
+```python
+test_05_90_img = test_05_90_img / 255 
+test_05_90_img = test_05_90_img.reshape(1, num_pixels) 
+```
+```python
+result = model.predict(test_05_90_img) 
+print('I think it\'s ', np.argmax(result))
+```
+*I think it's  4*
+
+Нейросеть не смогла распознать изображения
+
+**Вывод по архитектуре**: анализируя полученные результаты, можем прийти к выводу, что с ростом количества нейронов точность сначала улучшается - сеть обучается лучше, а при 500 нейронах - немного падает качество классификации, что может свидетельствовать о том, что алгоритм «застревал» в каком-то локальном минимуме; либо слишком малое время обучения - сеть не успевает обучиться, из-за чего страдает качество конечного результата. В данном примере это не критично, так как переобучение не наблюдается, а сама по себе точность достаточно высокая.
+
+**Вывод по картинкам**: проанализировав результаты работы сети, делаем вывод, что нейросеть справилась только с прямыми изображениями, повёрнутые она распознать не смогла. Это логично, потому что обучали её только на прямых изображениях. Если необходимо, чтобы картинки распознавались в том числе перевёрнутыми, в обучающую выборку стоит включить изображения такого же характера.
\ No newline at end of file