diff --git a/Keras_Einfuehrung.pdf b/Dokumentation.pdf similarity index 88% rename from Keras_Einfuehrung.pdf rename to Dokumentation.pdf index 45e606570e38ab606d4b0f4b7da64bb073c4e78b..8312435386f0a8ac625ba52d8ba4b0e674f707e0 100644 Binary files a/Keras_Einfuehrung.pdf and b/Dokumentation.pdf differ diff --git a/MachineLearning.pptx b/MachineLearning.pptx index 7df594cf33a22db2804c8eb4bf918f5de22b956b..a50bc839fd8fe5d0548a3b7692f2f3224e7b5841 100644 Binary files a/MachineLearning.pptx and b/MachineLearning.pptx differ diff --git a/notebooks/digit_recognizer/accuracy_graph_6.png b/notebooks/digit_recognizer/accuracy_graph_6.png new file mode 100644 index 0000000000000000000000000000000000000000..a3ca6b9d27583c3b3d12e468596bf50a341a2573 Binary files /dev/null and b/notebooks/digit_recognizer/accuracy_graph_6.png differ diff --git a/notebooks/digit_recognizer/digit_recognizer.ipynb b/notebooks/digit_recognizer/digit_recognizer.ipynb index 7dd59285b3c9ad3882b6f63d641366555c66b61a..514b224261866074347b491e24aec09042871520 100644 --- a/notebooks/digit_recognizer/digit_recognizer.ipynb +++ b/notebooks/digit_recognizer/digit_recognizer.ipynb @@ -499,7 +499,7 @@ }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 14, "metadata": {}, "outputs": [], "source": [ @@ -520,7 +520,7 @@ }, { "cell_type": "code", - "execution_count": 11, + "execution_count": 15, "metadata": {}, "outputs": [], "source": [ @@ -32188,7 +32188,7 @@ }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 7, "metadata": {}, "outputs": [], "source": [ @@ -32207,7 +32207,7 @@ }, { "cell_type": "code", - "execution_count": 11, + "execution_count": 8, "metadata": {}, "outputs": [ { @@ -32224,7 +32224,7 @@ }, { "cell_type": "code", - "execution_count": 12, + "execution_count": 9, "metadata": {}, "outputs": [], "source": [ @@ -32242,7 +32242,7 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": 10, "metadata": {}, "outputs": [ { @@ -32282,7 +32282,7 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 11, "metadata": {}, "outputs": [], "source": [ @@ -32295,7 +32295,7 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 12, "metadata": { "scrolled": true }, @@ -32305,45 +32305,45 @@ "output_type": "stream", "text": [ "Epoch 1/20\n", - "469/469 [==============================] - 40s 84ms/step - loss: 0.7240 - accuracy: 0.7853 - val_loss: 0.2441 - val_accuracy: 0.9251\n", + "469/469 [==============================] - 39s 84ms/step - loss: 0.6765 - accuracy: 0.7986 - val_loss: 0.2589 - val_accuracy: 0.9215\n", "Epoch 2/20\n", - "469/469 [==============================] - 38s 80ms/step - loss: 0.3312 - accuracy: 0.9002 - val_loss: 0.1899 - val_accuracy: 0.9420\n", + "469/469 [==============================] - 39s 83ms/step - loss: 0.3452 - accuracy: 0.8957 - val_loss: 0.2064 - val_accuracy: 0.9390\n", "Epoch 3/20\n", - "469/469 [==============================] - 38s 80ms/step - loss: 0.2761 - accuracy: 0.9180 - val_loss: 0.1575 - val_accuracy: 0.9531\n", + "469/469 [==============================] - 39s 82ms/step - loss: 0.2952 - accuracy: 0.9131 - val_loss: 0.1724 - val_accuracy: 0.9494\n", "Epoch 4/20\n", - "469/469 [==============================] - 39s 84ms/step - loss: 0.2363 - accuracy: 0.9299 - val_loss: 0.1340 - val_accuracy: 0.9590\n", + "469/469 [==============================] - 38s 82ms/step - loss: 0.2552 - accuracy: 0.9243 - val_loss: 0.1455 - val_accuracy: 0.9559\n", "Epoch 5/20\n", - "469/469 [==============================] - 40s 86ms/step - loss: 0.2091 - accuracy: 0.9381 - val_loss: 0.1174 - val_accuracy: 0.9659\n", + "469/469 [==============================] - 38s 81ms/step - loss: 0.2215 - accuracy: 0.9336 - val_loss: 0.1241 - val_accuracy: 0.9625\n", "Epoch 6/20\n", - "469/469 [==============================] - 40s 85ms/step - loss: 0.1821 - accuracy: 0.9455 - val_loss: 0.1001 - val_accuracy: 0.9705\n", + "469/469 [==============================] - 39s 83ms/step - loss: 0.1946 - accuracy: 0.9433 - val_loss: 0.1066 - val_accuracy: 0.9679\n", "Epoch 7/20\n", - "469/469 [==============================] - 40s 84ms/step - loss: 0.1600 - accuracy: 0.9531 - val_loss: 0.0893 - val_accuracy: 0.9729\n", + "469/469 [==============================] - 39s 82ms/step - loss: 0.1688 - accuracy: 0.9505 - val_loss: 0.0944 - val_accuracy: 0.9716\n", "Epoch 8/20\n", - "469/469 [==============================] - 38s 82ms/step - loss: 0.1461 - accuracy: 0.9575 - val_loss: 0.0805 - val_accuracy: 0.9759\n", + "469/469 [==============================] - 36s 78ms/step - loss: 0.1494 - accuracy: 0.9560 - val_loss: 0.0844 - val_accuracy: 0.9746\n", "Epoch 9/20\n", - "469/469 [==============================] - 39s 84ms/step - loss: 0.1281 - accuracy: 0.9618 - val_loss: 0.0749 - val_accuracy: 0.9770\n", + "469/469 [==============================] - 39s 83ms/step - loss: 0.1364 - accuracy: 0.9602 - val_loss: 0.0739 - val_accuracy: 0.9765\n", "Epoch 10/20\n", - "469/469 [==============================] - 38s 82ms/step - loss: 0.1203 - accuracy: 0.9636 - val_loss: 0.0673 - val_accuracy: 0.9799\n", + "469/469 [==============================] - 41s 88ms/step - loss: 0.1264 - accuracy: 0.9627 - val_loss: 0.0695 - val_accuracy: 0.9788\n", "Epoch 11/20\n", - "469/469 [==============================] - 38s 81ms/step - loss: 0.1103 - accuracy: 0.9672 - val_loss: 0.0608 - val_accuracy: 0.9806\n", + "469/469 [==============================] - 39s 84ms/step - loss: 0.1150 - accuracy: 0.9657 - val_loss: 0.0633 - val_accuracy: 0.9810\n", "Epoch 12/20\n", - "469/469 [==============================] - 39s 84ms/step - loss: 0.1021 - accuracy: 0.9692 - val_loss: 0.0583 - val_accuracy: 0.9817\n", + "469/469 [==============================] - 39s 84ms/step - loss: 0.1075 - accuracy: 0.9685 - val_loss: 0.0610 - val_accuracy: 0.9811\n", "Epoch 13/20\n", - "469/469 [==============================] - 38s 81ms/step - loss: 0.0991 - accuracy: 0.9699 - val_loss: 0.0573 - val_accuracy: 0.9814\n", + "469/469 [==============================] - 37s 78ms/step - loss: 0.1011 - accuracy: 0.9700 - val_loss: 0.0578 - val_accuracy: 0.9814\n", "Epoch 14/20\n", - "469/469 [==============================] - 38s 81ms/step - loss: 0.0940 - accuracy: 0.9720 - val_loss: 0.0547 - val_accuracy: 0.9831\n", + "469/469 [==============================] - 36s 78ms/step - loss: 0.0974 - accuracy: 0.9715 - val_loss: 0.0538 - val_accuracy: 0.9834\n", "Epoch 15/20\n", - "469/469 [==============================] - 38s 81ms/step - loss: 0.0879 - accuracy: 0.9738 - val_loss: 0.0528 - val_accuracy: 0.9831\n", + "469/469 [==============================] - 37s 78ms/step - loss: 0.0928 - accuracy: 0.9726 - val_loss: 0.0527 - val_accuracy: 0.9834\n", "Epoch 16/20\n", - "469/469 [==============================] - 39s 82ms/step - loss: 0.0854 - accuracy: 0.9744 - val_loss: 0.0509 - val_accuracy: 0.9835\n", + "469/469 [==============================] - 36s 78ms/step - loss: 0.0878 - accuracy: 0.9739 - val_loss: 0.0504 - val_accuracy: 0.9837\n", "Epoch 17/20\n", - "469/469 [==============================] - 38s 80ms/step - loss: 0.0828 - accuracy: 0.9751 - val_loss: 0.0492 - val_accuracy: 0.9839\n", + "469/469 [==============================] - 36s 78ms/step - loss: 0.0836 - accuracy: 0.9753 - val_loss: 0.0494 - val_accuracy: 0.9838\n", "Epoch 18/20\n", - "469/469 [==============================] - 38s 82ms/step - loss: 0.0787 - accuracy: 0.9772 - val_loss: 0.0493 - val_accuracy: 0.9836\n", + "469/469 [==============================] - 36s 78ms/step - loss: 0.0815 - accuracy: 0.9758 - val_loss: 0.0475 - val_accuracy: 0.9854\n", "Epoch 19/20\n", - "469/469 [==============================] - 38s 81ms/step - loss: 0.0760 - accuracy: 0.9773 - val_loss: 0.0465 - val_accuracy: 0.9851\n", + "469/469 [==============================] - 37s 79ms/step - loss: 0.0792 - accuracy: 0.9763 - val_loss: 0.0477 - val_accuracy: 0.9846\n", "Epoch 20/20\n", - "469/469 [==============================] - 38s 82ms/step - loss: 0.0748 - accuracy: 0.9778 - val_loss: 0.0464 - val_accuracy: 0.9852\n" + "469/469 [==============================] - 37s 78ms/step - loss: 0.0766 - accuracy: 0.9767 - val_loss: 0.0457 - val_accuracy: 0.9852\n" ] } ], @@ -32354,12 +32354,12 @@ }, { "cell_type": "code", - "execution_count": 69, + "execution_count": 16, "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "<Figure size 432x288 with 1 Axes>" ] @@ -32371,7 +32371,7 @@ }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "<Figure size 432x288 with 1 Axes>" ] diff --git a/notebooks/digit_recognizer/digit_recognizer_model_conv.h5 b/notebooks/digit_recognizer/digit_recognizer_model_conv.h5 index bd1c82b7e78b6aa7cf2711614be81200ce08d3ae..ac3c5b364d6a446267b8791954f63019bf55bbc6 100644 Binary files a/notebooks/digit_recognizer/digit_recognizer_model_conv.h5 and b/notebooks/digit_recognizer/digit_recognizer_model_conv.h5 differ diff --git a/notebooks/digit_recognizer/loss_graph_6.png b/notebooks/digit_recognizer/loss_graph_6.png new file mode 100644 index 0000000000000000000000000000000000000000..d3e7ef06f0e489b5604ee1e7225349c29ebf0f9e Binary files /dev/null and b/notebooks/digit_recognizer/loss_graph_6.png differ diff --git a/notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/fits_to_npy_big-checkpoint.ipynb b/notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/fits_to_npy_gold_big-checkpoint.ipynb similarity index 100% rename from notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/fits_to_npy_big-checkpoint.ipynb rename to notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/fits_to_npy_gold_big-checkpoint.ipynb diff --git a/notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/spectral_classifier_from_npy_big-checkpoint.ipynb b/notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/spectral_classifier_from_npy_gold_big-checkpoint.ipynb similarity index 100% rename from notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/spectral_classifier_from_npy_big-checkpoint.ipynb rename to notebooks/spectral_classifier_gold_big/.ipynb_checkpoints/spectral_classifier_from_npy_gold_big-checkpoint.ipynb diff --git a/notebooks/spectral_classifier_gold_big/accuracy_graph_goldbig.png b/notebooks/spectral_classifier_gold_big/accuracy_graph_gold_big.png similarity index 100% rename from notebooks/spectral_classifier_gold_big/accuracy_graph_goldbig.png rename to notebooks/spectral_classifier_gold_big/accuracy_graph_gold_big.png diff --git a/notebooks/spectral_classifier_gold_big/fits_to_npy_big.ipynb b/notebooks/spectral_classifier_gold_big/fits_to_npy_gold_big.ipynb similarity index 100% rename from notebooks/spectral_classifier_gold_big/fits_to_npy_big.ipynb rename to notebooks/spectral_classifier_gold_big/fits_to_npy_gold_big.ipynb diff --git a/notebooks/spectral_classifier_gold_big/loss_graph_goldbig.png b/notebooks/spectral_classifier_gold_big/loss_graph_gold_big.png similarity index 100% rename from notebooks/spectral_classifier_gold_big/loss_graph_goldbig.png rename to notebooks/spectral_classifier_gold_big/loss_graph_gold_big.png diff --git a/notebooks/spectral_classifier_gold_big/spectral_classifier_from_npy_big.ipynb b/notebooks/spectral_classifier_gold_big/spectral_classifier_from_npy_gold_big.ipynb similarity index 91% rename from notebooks/spectral_classifier_gold_big/spectral_classifier_from_npy_big.ipynb rename to notebooks/spectral_classifier_gold_big/spectral_classifier_from_npy_gold_big.ipynb index c9e77c918dd5849db36dbb94e754994a99e2070b..1e728a0691fe63b0db266d7dd2796005270aa968 100644 --- a/notebooks/spectral_classifier_gold_big/spectral_classifier_from_npy_big.ipynb +++ b/notebooks/spectral_classifier_gold_big/spectral_classifier_from_npy_gold_big.ipynb @@ -9,7 +9,7 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": 2, "metadata": { "scrolled": true }, @@ -42,12 +42,12 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "## 2) Plotten" + "## 2) Daten Plotten" ] }, { "cell_type": "code", - "execution_count": 2, + "execution_count": 3, "metadata": {}, "outputs": [ { @@ -122,6 +122,66 @@ }, "output_type": "display_data" }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "<Figure size 432x288 with 1 Axes>" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "<Figure size 432x288 with 1 Axes>" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "<Figure size 432x288 with 1 Axes>" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "<Figure size 432x288 with 1 Axes>" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "<Figure size 432x288 with 1 Axes>" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, { "data": { "image/png": "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\n", @@ -136,7 +196,7 @@ } ], "source": [ - "for i in [1, 2, 3, 1002, 1534, 1550, 3999]: #1534 \n", + "for i in [1, 2, 3, 1002, 1534, 1550, 3500,3501,3502,3506,3508,3999]: #1534 \n", " plt.plot(wavelengths, data[i]) \n", " plt.xlabel('wavelength(Å)') \n", " plt.ylabel('flux (10-17 ergs/s/cm2/Å)') \n", @@ -154,7 +214,7 @@ }, { "cell_type": "code", - "execution_count": 19, + "execution_count": 3, "metadata": {}, "outputs": [], "source": [ @@ -167,7 +227,7 @@ }, { "cell_type": "code", - "execution_count": 20, + "execution_count": 4, "metadata": {}, "outputs": [ { @@ -186,7 +246,7 @@ }, { "cell_type": "code", - "execution_count": 21, + "execution_count": 5, "metadata": {}, "outputs": [], "source": [ @@ -214,7 +274,7 @@ }, { "cell_type": "code", - "execution_count": 22, + "execution_count": 6, "metadata": {}, "outputs": [], "source": [ @@ -229,7 +289,7 @@ }, { "cell_type": "code", - "execution_count": 23, + "execution_count": 7, "metadata": {}, "outputs": [], "source": [ @@ -241,7 +301,7 @@ }, { "cell_type": "code", - "execution_count": 24, + "execution_count": 8, "metadata": {}, "outputs": [ { @@ -260,7 +320,7 @@ }, { "cell_type": "code", - "execution_count": 25, + "execution_count": 13, "metadata": {}, "outputs": [], "source": [ @@ -284,11 +344,24 @@ "# ])\n", "\n", "# Am besten (schnell):\n", + "# model = Sequential([\n", + "# Conv1D(filters=64, kernel_size=80, strides=10, activation='relu', input_shape=(3522,1)), # stride\n", + "# MaxPooling1D(3), #Pooling verringert Accuracy leicht aber verhindert overfitting\n", + "# Dropout(0.35),\n", + "# Conv1D(filters=128, kernel_size=40, strides=10, activation='relu'),\n", + "# MaxPooling1D(3),\n", + "# Dropout(0.35),\n", + "# Flatten(),\n", + "# Dense(units=128, activation='relu'), # Droput, weniger neuronen\n", + "# Dropout(0.35),\n", + "# Dense(units=4, activation='softmax')\n", + "# ])\n", + "\n", "model = Sequential([\n", - " Conv1D(filters=64, kernel_size=80, strides=10, activation='relu', input_shape=(3522,1)), # stride\n", + " Conv1D(filters=32, kernel_size=20, strides=10, activation='relu', input_shape=(3522,1)), # stride\n", " MaxPooling1D(3), #Pooling verringert Accuracy leicht aber verhindert overfitting\n", " Dropout(0.35),\n", - " Conv1D(filters=128, kernel_size=40, strides=10, activation='relu'),\n", + " Conv1D(filters=64, kernel_size=40, strides=10, activation='relu'),\n", " MaxPooling1D(3),\n", " Dropout(0.35),\n", " Flatten(),\n", @@ -297,6 +370,8 @@ " Dense(units=4, activation='softmax')\n", "])\n", "\n", + "\n", + "\n", "# model = Sequential([\n", "# Conv1D(filters=64, kernel_size=40, strides=10, activation='relu', input_shape=(3522,1)), # stride\n", "# Dropout(0.2),\n", @@ -323,7 +398,7 @@ }, { "cell_type": "code", - "execution_count": 26, + "execution_count": 14, "metadata": {}, "outputs": [], "source": [ @@ -333,7 +408,7 @@ }, { "cell_type": "code", - "execution_count": 27, + "execution_count": 15, "metadata": {}, "outputs": [], "source": [ @@ -349,437 +424,119 @@ }, { "cell_type": "code", - "execution_count": 28, + "execution_count": 16, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Epoch 1/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 1.2416 - accuracy: 0.3875 - val_loss: 0.9693 - val_accuracy: 0.5267\n", - "Epoch 2/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.9215 - accuracy: 0.5674 - val_loss: 0.8352 - val_accuracy: 0.6233\n", - "Epoch 3/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.8234 - accuracy: 0.6267 - val_loss: 0.7642 - val_accuracy: 0.6456\n", - "Epoch 4/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.7559 - accuracy: 0.6537 - val_loss: 0.6841 - val_accuracy: 0.6667\n", - "Epoch 5/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.6795 - accuracy: 0.6841 - val_loss: 0.6025 - val_accuracy: 0.7167\n", - "Epoch 6/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.6175 - accuracy: 0.7127 - val_loss: 0.5634 - val_accuracy: 0.7189\n", - "Epoch 7/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.5633 - accuracy: 0.7364 - val_loss: 0.4972 - val_accuracy: 0.7578\n", - "Epoch 8/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.5292 - accuracy: 0.7527 - val_loss: 0.4805 - val_accuracy: 0.7656\n", - "Epoch 9/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.4991 - accuracy: 0.7730 - val_loss: 0.4704 - val_accuracy: 0.7667\n", - "Epoch 10/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.4775 - accuracy: 0.7822 - val_loss: 0.4483 - val_accuracy: 0.8078\n", - "Epoch 11/200\n", - "41/41 [==============================] - 3s 74ms/step - loss: 0.4635 - accuracy: 0.7830 - val_loss: 0.4327 - val_accuracy: 0.7856\n", - "Epoch 12/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.4481 - accuracy: 0.7958 - val_loss: 0.4111 - val_accuracy: 0.7989\n", - "Epoch 13/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.4354 - accuracy: 0.8010 - val_loss: 0.3972 - val_accuracy: 0.8233\n", - "Epoch 14/200\n", - "41/41 [==============================] - 3s 82ms/step - loss: 0.4186 - accuracy: 0.8089 - val_loss: 0.3755 - val_accuracy: 0.8378\n", - "Epoch 15/200\n", - "41/41 [==============================] - 4s 89ms/step - loss: 0.4069 - accuracy: 0.8119 - val_loss: 0.3781 - val_accuracy: 0.8022\n", - "Epoch 16/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.3953 - accuracy: 0.8263 - val_loss: 0.3574 - val_accuracy: 0.8578\n", - "Epoch 17/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.3800 - accuracy: 0.8330 - val_loss: 0.3469 - val_accuracy: 0.8522\n", - "Epoch 18/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.3718 - accuracy: 0.8398 - val_loss: 0.3361 - val_accuracy: 0.8700\n", - "Epoch 19/200\n", - "41/41 [==============================] - 3s 79ms/step - loss: 0.3608 - accuracy: 0.8360 - val_loss: 0.3365 - val_accuracy: 0.8578\n", - "Epoch 20/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.3494 - accuracy: 0.8404 - val_loss: 0.3158 - val_accuracy: 0.8789\n", - "Epoch 21/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.3470 - accuracy: 0.8462 - val_loss: 0.3214 - val_accuracy: 0.8667\n", - "Epoch 22/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.3423 - accuracy: 0.8530 - val_loss: 0.2996 - val_accuracy: 0.8767\n", - "Epoch 23/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.3240 - accuracy: 0.8563 - val_loss: 0.2985 - val_accuracy: 0.8800\n", - "Epoch 24/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.3182 - accuracy: 0.8642 - val_loss: 0.2886 - val_accuracy: 0.8933\n", - "Epoch 25/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.3055 - accuracy: 0.8678 - val_loss: 0.2848 - val_accuracy: 0.8789\n", - "Epoch 26/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.3049 - accuracy: 0.8695 - val_loss: 0.2803 - val_accuracy: 0.8833\n", - "Epoch 27/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.2954 - accuracy: 0.8735 - val_loss: 0.2674 - val_accuracy: 0.8844\n", - "Epoch 28/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.2949 - accuracy: 0.8764 - val_loss: 0.2705 - val_accuracy: 0.8944\n", - "Epoch 29/200\n", - "41/41 [==============================] - 3s 73ms/step - loss: 0.2847 - accuracy: 0.8779 - val_loss: 0.2601 - val_accuracy: 0.8911\n", - "Epoch 30/200\n", - "41/41 [==============================] - 3s 72ms/step - loss: 0.2844 - accuracy: 0.8807 - val_loss: 0.2693 - val_accuracy: 0.8900\n", - "Epoch 31/200\n", - "41/41 [==============================] - 3s 72ms/step - loss: 0.2771 - accuracy: 0.8801 - val_loss: 0.2515 - val_accuracy: 0.8978\n", - "Epoch 32/200\n", - "41/41 [==============================] - 3s 72ms/step - loss: 0.2744 - accuracy: 0.8811 - val_loss: 0.2548 - val_accuracy: 0.9011\n", - "Epoch 33/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.2677 - accuracy: 0.8874 - val_loss: 0.2431 - val_accuracy: 0.9022\n", - "Epoch 34/200\n", - "41/41 [==============================] - 3s 73ms/step - loss: 0.2640 - accuracy: 0.8902 - val_loss: 0.2507 - val_accuracy: 0.8956\n", - "Epoch 35/200\n", - "41/41 [==============================] - 3s 74ms/step - loss: 0.2606 - accuracy: 0.8914 - val_loss: 0.2412 - val_accuracy: 0.9011\n", - "Epoch 36/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.2635 - accuracy: 0.8911 - val_loss: 0.2419 - val_accuracy: 0.8989\n", - "Epoch 37/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.2558 - accuracy: 0.8969 - val_loss: 0.2464 - val_accuracy: 0.9011\n", - "Epoch 38/200\n", - "41/41 [==============================] - 3s 74ms/step - loss: 0.2503 - accuracy: 0.8957 - val_loss: 0.2370 - val_accuracy: 0.9122\n", - "Epoch 39/200\n", - "41/41 [==============================] - 3s 73ms/step - loss: 0.2435 - accuracy: 0.9002 - val_loss: 0.2312 - val_accuracy: 0.9067\n", - "Epoch 40/200\n", - "41/41 [==============================] - 3s 75ms/step - loss: 0.2498 - accuracy: 0.8977 - val_loss: 0.2521 - val_accuracy: 0.8956\n", - "Epoch 41/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.2462 - accuracy: 0.9002 - val_loss: 0.2309 - val_accuracy: 0.9067\n", - "Epoch 42/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.2364 - accuracy: 0.9014 - val_loss: 0.2381 - val_accuracy: 0.9011\n", - "Epoch 43/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.2318 - accuracy: 0.9073 - val_loss: 0.2356 - val_accuracy: 0.9078\n", - "Epoch 44/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.2272 - accuracy: 0.9098 - val_loss: 0.2175 - val_accuracy: 0.9122\n", - "Epoch 45/200\n", - "41/41 [==============================] - 3s 74ms/step - loss: 0.2301 - accuracy: 0.9077 - val_loss: 0.2213 - val_accuracy: 0.9111\n", - "Epoch 46/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.2337 - accuracy: 0.9043 - val_loss: 0.2233 - val_accuracy: 0.9078\n", - "Epoch 47/200\n", - "41/41 [==============================] - 3s 72ms/step - loss: 0.2251 - accuracy: 0.9053 - val_loss: 0.2153 - val_accuracy: 0.9200\n", - "Epoch 48/200\n", - "41/41 [==============================] - 3s 73ms/step - loss: 0.2343 - accuracy: 0.9028 - val_loss: 0.2283 - val_accuracy: 0.9122\n", - "Epoch 49/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.2258 - accuracy: 0.9069 - val_loss: 0.2219 - val_accuracy: 0.9089\n", - "Epoch 50/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.2156 - accuracy: 0.9107 - val_loss: 0.2088 - val_accuracy: 0.9156\n", - "Epoch 51/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.2159 - accuracy: 0.9117 - val_loss: 0.2259 - val_accuracy: 0.9122\n", - "Epoch 52/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.2162 - accuracy: 0.9112 - val_loss: 0.2145 - val_accuracy: 0.9167\n", - "Epoch 53/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.2209 - accuracy: 0.9085 - val_loss: 0.2156 - val_accuracy: 0.9156\n", - "Epoch 54/200\n", - "41/41 [==============================] - 3s 80ms/step - loss: 0.2168 - accuracy: 0.9132 - val_loss: 0.2195 - val_accuracy: 0.9100\n", - "Epoch 55/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.2084 - accuracy: 0.9147 - val_loss: 0.2283 - val_accuracy: 0.9011\n", - "Epoch 56/200\n", - "41/41 [==============================] - 3s 78ms/step - loss: 0.2160 - accuracy: 0.9091 - val_loss: 0.2111 - val_accuracy: 0.9167\n", - "Epoch 57/200\n", - "41/41 [==============================] - 3s 79ms/step - loss: 0.2129 - accuracy: 0.9109 - val_loss: 0.2238 - val_accuracy: 0.9089\n", - "Epoch 58/200\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "41/41 [==============================] - 3s 72ms/step - loss: 0.2089 - accuracy: 0.9162 - val_loss: 0.1993 - val_accuracy: 0.9178\n", - "Epoch 59/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.2068 - accuracy: 0.9185 - val_loss: 0.2069 - val_accuracy: 0.9156\n", - "Epoch 60/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.2000 - accuracy: 0.9178 - val_loss: 0.2094 - val_accuracy: 0.9189\n", - "Epoch 61/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.2027 - accuracy: 0.9163 - val_loss: 0.2124 - val_accuracy: 0.9211\n", - "Epoch 62/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.2056 - accuracy: 0.9140 - val_loss: 0.2029 - val_accuracy: 0.9211\n", - "Epoch 63/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.2049 - accuracy: 0.9109 - val_loss: 0.2055 - val_accuracy: 0.9267\n", - "Epoch 64/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.2021 - accuracy: 0.9170 - val_loss: 0.1994 - val_accuracy: 0.9222\n", - "Epoch 65/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1931 - accuracy: 0.9201 - val_loss: 0.2161 - val_accuracy: 0.9100\n", - "Epoch 66/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1960 - accuracy: 0.9223 - val_loss: 0.2116 - val_accuracy: 0.9200\n", - "Epoch 67/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1944 - accuracy: 0.9231 - val_loss: 0.2000 - val_accuracy: 0.9300\n", - "Epoch 68/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1885 - accuracy: 0.9230 - val_loss: 0.1959 - val_accuracy: 0.9200\n", - "Epoch 69/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1880 - accuracy: 0.9236 - val_loss: 0.2034 - val_accuracy: 0.9144\n", - "Epoch 70/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.1939 - accuracy: 0.9185 - val_loss: 0.1938 - val_accuracy: 0.9278\n", - "Epoch 71/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.1852 - accuracy: 0.9242 - val_loss: 0.1910 - val_accuracy: 0.9278\n", - "Epoch 72/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1883 - accuracy: 0.9232 - val_loss: 0.2040 - val_accuracy: 0.9289\n", - "Epoch 73/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1815 - accuracy: 0.9279 - val_loss: 0.1986 - val_accuracy: 0.9200\n", - "Epoch 74/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1859 - accuracy: 0.9270 - val_loss: 0.1924 - val_accuracy: 0.9233\n", - "Epoch 75/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1907 - accuracy: 0.9228 - val_loss: 0.1983 - val_accuracy: 0.9244\n", - "Epoch 76/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1826 - accuracy: 0.9275 - val_loss: 0.1965 - val_accuracy: 0.9222\n", - "Epoch 77/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1763 - accuracy: 0.9291 - val_loss: 0.1931 - val_accuracy: 0.9211\n", - "Epoch 78/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1821 - accuracy: 0.9246 - val_loss: 0.1977 - val_accuracy: 0.9256\n", - "Epoch 79/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1791 - accuracy: 0.9257 - val_loss: 0.1999 - val_accuracy: 0.9233\n", - "Epoch 80/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1795 - accuracy: 0.9252 - val_loss: 0.1980 - val_accuracy: 0.9211\n", - "Epoch 81/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1746 - accuracy: 0.9291 - val_loss: 0.1950 - val_accuracy: 0.9222\n", - "Epoch 82/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1766 - accuracy: 0.9277 - val_loss: 0.2117 - val_accuracy: 0.9233\n", - "Epoch 83/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1741 - accuracy: 0.9280 - val_loss: 0.1939 - val_accuracy: 0.9256\n", - "Epoch 84/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1755 - accuracy: 0.9267 - val_loss: 0.2104 - val_accuracy: 0.9244\n", - "Epoch 85/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1652 - accuracy: 0.9342 - val_loss: 0.1937 - val_accuracy: 0.9244\n", - "Epoch 86/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1702 - accuracy: 0.9269 - val_loss: 0.1978 - val_accuracy: 0.9267\n", - "Epoch 87/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1698 - accuracy: 0.9331 - val_loss: 0.1968 - val_accuracy: 0.9244\n", - "Epoch 88/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1705 - accuracy: 0.9316 - val_loss: 0.2042 - val_accuracy: 0.9189\n", - "Epoch 89/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1670 - accuracy: 0.9322 - val_loss: 0.1937 - val_accuracy: 0.9133\n", - "Epoch 90/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1688 - accuracy: 0.9289 - val_loss: 0.1943 - val_accuracy: 0.9256\n", - "Epoch 91/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1595 - accuracy: 0.9359 - val_loss: 0.1940 - val_accuracy: 0.9256\n", - "Epoch 92/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1621 - accuracy: 0.9344 - val_loss: 0.1929 - val_accuracy: 0.9289\n", - "Epoch 93/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1621 - accuracy: 0.9335 - val_loss: 0.1935 - val_accuracy: 0.9233\n", - "Epoch 94/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1608 - accuracy: 0.9357 - val_loss: 0.1902 - val_accuracy: 0.9300\n", - "Epoch 95/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1724 - accuracy: 0.9301 - val_loss: 0.1926 - val_accuracy: 0.9267\n", - "Epoch 96/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1652 - accuracy: 0.9342 - val_loss: 0.1835 - val_accuracy: 0.9289\n", - "Epoch 97/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1604 - accuracy: 0.9319 - val_loss: 0.1906 - val_accuracy: 0.9300\n", - "Epoch 98/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.1627 - accuracy: 0.9316 - val_loss: 0.1900 - val_accuracy: 0.9311\n", - "Epoch 99/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1583 - accuracy: 0.9365 - val_loss: 0.1910 - val_accuracy: 0.9278\n", - "Epoch 100/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1580 - accuracy: 0.9369 - val_loss: 0.2024 - val_accuracy: 0.9233\n", - "Epoch 101/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1620 - accuracy: 0.9384 - val_loss: 0.1986 - val_accuracy: 0.9211\n", - "Epoch 102/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1580 - accuracy: 0.9364 - val_loss: 0.1963 - val_accuracy: 0.9256\n", - "Epoch 103/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1539 - accuracy: 0.9353 - val_loss: 0.1923 - val_accuracy: 0.9289\n", - "Epoch 104/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1525 - accuracy: 0.9338 - val_loss: 0.2003 - val_accuracy: 0.9289\n", - "Epoch 105/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1595 - accuracy: 0.9346 - val_loss: 0.1945 - val_accuracy: 0.9289\n", - "Epoch 106/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1598 - accuracy: 0.9312 - val_loss: 0.1919 - val_accuracy: 0.9256\n", - "Epoch 107/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1448 - accuracy: 0.9420 - val_loss: 0.1920 - val_accuracy: 0.9244\n", - "Epoch 108/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1534 - accuracy: 0.9406 - val_loss: 0.1925 - val_accuracy: 0.9333\n", - "Epoch 109/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1515 - accuracy: 0.9381 - val_loss: 0.1983 - val_accuracy: 0.9244\n", - "Epoch 110/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1540 - accuracy: 0.9384 - val_loss: 0.1947 - val_accuracy: 0.9300\n", - "Epoch 111/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1423 - accuracy: 0.9422 - val_loss: 0.1998 - val_accuracy: 0.9256\n", - "Epoch 112/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1507 - accuracy: 0.9406 - val_loss: 0.1901 - val_accuracy: 0.9311\n", - "Epoch 113/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1452 - accuracy: 0.9417 - val_loss: 0.2047 - val_accuracy: 0.9322\n", - "Epoch 114/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.1533 - accuracy: 0.9375 - val_loss: 0.2186 - val_accuracy: 0.9256\n", - "Epoch 115/200\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "41/41 [==============================] - 3s 66ms/step - loss: 0.1428 - accuracy: 0.9428 - val_loss: 0.1911 - val_accuracy: 0.9289\n", - "Epoch 116/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1405 - accuracy: 0.9426 - val_loss: 0.1945 - val_accuracy: 0.9267\n", - "Epoch 117/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1403 - accuracy: 0.9410 - val_loss: 0.1922 - val_accuracy: 0.9289\n", - "Epoch 118/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1441 - accuracy: 0.9447 - val_loss: 0.2081 - val_accuracy: 0.9211\n", - "Epoch 119/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1546 - accuracy: 0.9395 - val_loss: 0.1906 - val_accuracy: 0.9322\n", - "Epoch 120/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1458 - accuracy: 0.9415 - val_loss: 0.2004 - val_accuracy: 0.9278\n", - "Epoch 121/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1426 - accuracy: 0.9436 - val_loss: 0.2107 - val_accuracy: 0.9267\n", - "Epoch 122/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1433 - accuracy: 0.9437 - val_loss: 0.1937 - val_accuracy: 0.9378\n", - "Epoch 123/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1476 - accuracy: 0.9417 - val_loss: 0.1817 - val_accuracy: 0.9289\n", - "Epoch 124/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1429 - accuracy: 0.9404 - val_loss: 0.1913 - val_accuracy: 0.9233\n", - "Epoch 125/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1359 - accuracy: 0.9442 - val_loss: 0.1983 - val_accuracy: 0.9233\n", - "Epoch 126/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1378 - accuracy: 0.9438 - val_loss: 0.2048 - val_accuracy: 0.9289\n", - "Epoch 127/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1385 - accuracy: 0.9433 - val_loss: 0.1900 - val_accuracy: 0.9322\n", - "Epoch 128/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1386 - accuracy: 0.9432 - val_loss: 0.1949 - val_accuracy: 0.9356\n", - "Epoch 129/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1327 - accuracy: 0.9465 - val_loss: 0.1981 - val_accuracy: 0.9289\n", - "Epoch 130/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1401 - accuracy: 0.9425 - val_loss: 0.2115 - val_accuracy: 0.9300\n", - "Epoch 131/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1450 - accuracy: 0.9422 - val_loss: 0.2013 - val_accuracy: 0.9267\n", - "Epoch 132/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1302 - accuracy: 0.9486 - val_loss: 0.2040 - val_accuracy: 0.9267\n", - "Epoch 133/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1345 - accuracy: 0.9464 - val_loss: 0.2095 - val_accuracy: 0.9311\n", - "Epoch 134/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1315 - accuracy: 0.9462 - val_loss: 0.1944 - val_accuracy: 0.9278\n", - "Epoch 135/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1324 - accuracy: 0.9489 - val_loss: 0.2036 - val_accuracy: 0.9267\n", - "Epoch 136/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1353 - accuracy: 0.9448 - val_loss: 0.1986 - val_accuracy: 0.9267\n", - "Epoch 137/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1354 - accuracy: 0.9452 - val_loss: 0.1876 - val_accuracy: 0.9278\n", - "Epoch 138/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1332 - accuracy: 0.9484 - val_loss: 0.1899 - val_accuracy: 0.9333\n", - "Epoch 139/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1253 - accuracy: 0.9485 - val_loss: 0.1918 - val_accuracy: 0.9322\n", - "Epoch 140/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1244 - accuracy: 0.9501 - val_loss: 0.1921 - val_accuracy: 0.9278\n", - "Epoch 141/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1293 - accuracy: 0.9480 - val_loss: 0.1992 - val_accuracy: 0.9322\n", - "Epoch 142/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1326 - accuracy: 0.9452 - val_loss: 0.2043 - val_accuracy: 0.9278\n", - "Epoch 143/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1278 - accuracy: 0.9489 - val_loss: 0.1905 - val_accuracy: 0.9322\n", - "Epoch 144/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1300 - accuracy: 0.9452 - val_loss: 0.2003 - val_accuracy: 0.9344\n", - "Epoch 145/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1282 - accuracy: 0.9509 - val_loss: 0.1946 - val_accuracy: 0.9322\n", - "Epoch 146/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1289 - accuracy: 0.9477 - val_loss: 0.2134 - val_accuracy: 0.9200\n", - "Epoch 147/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1346 - accuracy: 0.9438 - val_loss: 0.1940 - val_accuracy: 0.9311\n", - "Epoch 148/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1233 - accuracy: 0.9490 - val_loss: 0.1977 - val_accuracy: 0.9311\n", - "Epoch 149/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1247 - accuracy: 0.9501 - val_loss: 0.1963 - val_accuracy: 0.9333\n", - "Epoch 150/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1298 - accuracy: 0.9481 - val_loss: 0.2277 - val_accuracy: 0.9278\n", - "Epoch 151/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1293 - accuracy: 0.9472 - val_loss: 0.2038 - val_accuracy: 0.9311\n", - "Epoch 152/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1286 - accuracy: 0.9472 - val_loss: 0.1987 - val_accuracy: 0.9333\n", - "Epoch 153/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1333 - accuracy: 0.9483 - val_loss: 0.2113 - val_accuracy: 0.9300\n", - "Epoch 154/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1175 - accuracy: 0.9515 - val_loss: 0.1867 - val_accuracy: 0.9344\n", - "Epoch 155/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1272 - accuracy: 0.9486 - val_loss: 0.2030 - val_accuracy: 0.9344\n", - "Epoch 156/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1390 - accuracy: 0.9456 - val_loss: 0.1979 - val_accuracy: 0.9322\n", - "Epoch 157/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1210 - accuracy: 0.9511 - val_loss: 0.2128 - val_accuracy: 0.9278\n", - "Epoch 158/200\n", - "41/41 [==============================] - 3s 62ms/step - loss: 0.1222 - accuracy: 0.9480 - val_loss: 0.1888 - val_accuracy: 0.9278\n", - "Epoch 159/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1224 - accuracy: 0.9496 - val_loss: 0.1929 - val_accuracy: 0.9256\n", - "Epoch 160/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1270 - accuracy: 0.9488 - val_loss: 0.1879 - val_accuracy: 0.9344\n", - "Epoch 161/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1255 - accuracy: 0.9507 - val_loss: 0.1930 - val_accuracy: 0.9344\n", - "Epoch 162/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1241 - accuracy: 0.9478 - val_loss: 0.2095 - val_accuracy: 0.9256\n", - "Epoch 163/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1253 - accuracy: 0.9495 - val_loss: 0.1963 - val_accuracy: 0.9333\n", - "Epoch 164/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1162 - accuracy: 0.9535 - val_loss: 0.1998 - val_accuracy: 0.9289\n", - "Epoch 165/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1227 - accuracy: 0.9512 - val_loss: 0.1874 - val_accuracy: 0.9389\n", - "Epoch 166/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1145 - accuracy: 0.9553 - val_loss: 0.2136 - val_accuracy: 0.9256\n", - "Epoch 167/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1111 - accuracy: 0.9548 - val_loss: 0.1982 - val_accuracy: 0.9378\n", - "Epoch 168/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1209 - accuracy: 0.9486 - val_loss: 0.2017 - val_accuracy: 0.9311\n", - "Epoch 169/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1222 - accuracy: 0.9528 - val_loss: 0.1957 - val_accuracy: 0.9322\n", - "Epoch 170/200\n", - "41/41 [==============================] - 3s 68ms/step - loss: 0.1149 - accuracy: 0.9536 - val_loss: 0.1885 - val_accuracy: 0.9378\n", - "Epoch 171/200\n", - "41/41 [==============================] - 3s 75ms/step - loss: 0.1186 - accuracy: 0.9509 - val_loss: 0.1974 - val_accuracy: 0.9233\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 172/200\n", - "41/41 [==============================] - 3s 70ms/step - loss: 0.1089 - accuracy: 0.9559 - val_loss: 0.1989 - val_accuracy: 0.9333\n", - "Epoch 173/200\n", - "41/41 [==============================] - 3s 74ms/step - loss: 0.1149 - accuracy: 0.9530 - val_loss: 0.1933 - val_accuracy: 0.9367\n", - "Epoch 174/200\n", - "41/41 [==============================] - 3s 72ms/step - loss: 0.1175 - accuracy: 0.9546 - val_loss: 0.1858 - val_accuracy: 0.9389\n", - "Epoch 175/200\n", - "41/41 [==============================] - 3s 71ms/step - loss: 0.1196 - accuracy: 0.9521 - val_loss: 0.1984 - val_accuracy: 0.9389\n", - "Epoch 176/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1178 - accuracy: 0.9522 - val_loss: 0.1861 - val_accuracy: 0.9356\n", - "Epoch 177/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1137 - accuracy: 0.9575 - val_loss: 0.1888 - val_accuracy: 0.9356\n", - "Epoch 178/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1147 - accuracy: 0.9530 - val_loss: 0.1935 - val_accuracy: 0.9311\n", - "Epoch 179/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1101 - accuracy: 0.9577 - val_loss: 0.2080 - val_accuracy: 0.9344\n", - "Epoch 180/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1234 - accuracy: 0.9509 - val_loss: 0.1917 - val_accuracy: 0.9400\n", - "Epoch 181/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1085 - accuracy: 0.9567 - val_loss: 0.2062 - val_accuracy: 0.9311\n", - "Epoch 182/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.1114 - accuracy: 0.9569 - val_loss: 0.1821 - val_accuracy: 0.9367\n", - "Epoch 183/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1147 - accuracy: 0.9563 - val_loss: 0.1884 - val_accuracy: 0.9367\n", - "Epoch 184/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1100 - accuracy: 0.9556 - val_loss: 0.1942 - val_accuracy: 0.9322\n", - "Epoch 185/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1057 - accuracy: 0.9586 - val_loss: 0.1917 - val_accuracy: 0.9400\n", - "Epoch 186/200\n", - "41/41 [==============================] - 3s 69ms/step - loss: 0.1143 - accuracy: 0.9536 - val_loss: 0.2078 - val_accuracy: 0.9289\n", - "Epoch 187/200\n", - "41/41 [==============================] - 3s 72ms/step - loss: 0.1138 - accuracy: 0.9557 - val_loss: 0.1974 - val_accuracy: 0.9344\n", - "Epoch 188/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1118 - accuracy: 0.9551 - val_loss: 0.1934 - val_accuracy: 0.9367\n", - "Epoch 189/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1068 - accuracy: 0.9562 - val_loss: 0.2098 - val_accuracy: 0.9289\n", - "Epoch 190/200\n", - "41/41 [==============================] - 3s 64ms/step - loss: 0.1056 - accuracy: 0.9591 - val_loss: 0.2108 - val_accuracy: 0.9356\n", - "Epoch 191/200\n", - "41/41 [==============================] - 3s 62ms/step - loss: 0.1079 - accuracy: 0.9572 - val_loss: 0.2068 - val_accuracy: 0.9378\n", - "Epoch 192/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1041 - accuracy: 0.9577 - val_loss: 0.2015 - val_accuracy: 0.9389\n", - "Epoch 193/200\n", - "41/41 [==============================] - 3s 63ms/step - loss: 0.1084 - accuracy: 0.9570 - val_loss: 0.1963 - val_accuracy: 0.9333\n", - "Epoch 194/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1129 - accuracy: 0.9531 - val_loss: 0.2044 - val_accuracy: 0.9344\n", - "Epoch 195/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1037 - accuracy: 0.9585 - val_loss: 0.1972 - val_accuracy: 0.9322\n", - "Epoch 196/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.1067 - accuracy: 0.9569 - val_loss: 0.1967 - val_accuracy: 0.9344\n", - "Epoch 197/200\n", - "41/41 [==============================] - 3s 66ms/step - loss: 0.0989 - accuracy: 0.9614 - val_loss: 0.1974 - val_accuracy: 0.9389\n", - "Epoch 198/200\n", - "41/41 [==============================] - 3s 65ms/step - loss: 0.0995 - accuracy: 0.9626 - val_loss: 0.1995 - val_accuracy: 0.9322\n", - "Epoch 199/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1030 - accuracy: 0.9591 - val_loss: 0.2119 - val_accuracy: 0.9333\n", - "Epoch 200/200\n", - "41/41 [==============================] - 3s 67ms/step - loss: 0.1019 - accuracy: 0.9596 - val_loss: 0.2222 - val_accuracy: 0.9378\n" + "Epoch 1/50\n", + "41/41 [==============================] - 1s 32ms/step - loss: 1.3114 - accuracy: 0.3609 - val_loss: 1.0574 - val_accuracy: 0.5022\n", + "Epoch 2/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.9935 - accuracy: 0.5249 - val_loss: 0.8919 - val_accuracy: 0.6100\n", + "Epoch 3/50\n", + "41/41 [==============================] - 1s 30ms/step - loss: 0.9094 - accuracy: 0.5827 - val_loss: 0.8263 - val_accuracy: 0.6244\n", + "Epoch 4/50\n", + "41/41 [==============================] - 1s 29ms/step - loss: 0.8445 - accuracy: 0.6140 - val_loss: 0.7948 - val_accuracy: 0.6033\n", + "Epoch 5/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.7937 - accuracy: 0.6322 - val_loss: 0.6985 - val_accuracy: 0.6878\n", + "Epoch 6/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.7285 - accuracy: 0.6642 - val_loss: 0.6314 - val_accuracy: 0.7144\n", + "Epoch 7/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.6790 - accuracy: 0.6936 - val_loss: 0.5847 - val_accuracy: 0.7256\n", + "Epoch 8/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.6316 - accuracy: 0.7136 - val_loss: 0.5462 - val_accuracy: 0.7611\n", + "Epoch 9/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.5930 - accuracy: 0.7283 - val_loss: 0.5086 - val_accuracy: 0.7756\n", + "Epoch 10/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.5622 - accuracy: 0.7367 - val_loss: 0.4812 - val_accuracy: 0.7711\n", + "Epoch 11/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.5415 - accuracy: 0.7475 - val_loss: 0.4687 - val_accuracy: 0.7956\n", + "Epoch 12/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.5165 - accuracy: 0.7625 - val_loss: 0.4444 - val_accuracy: 0.7922\n", + "Epoch 13/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.4952 - accuracy: 0.7748 - val_loss: 0.4342 - val_accuracy: 0.7800\n", + "Epoch 14/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.4768 - accuracy: 0.7827 - val_loss: 0.4199 - val_accuracy: 0.7922\n", + "Epoch 15/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.4615 - accuracy: 0.7904 - val_loss: 0.4085 - val_accuracy: 0.8356\n", + "Epoch 16/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.4515 - accuracy: 0.7949 - val_loss: 0.3926 - val_accuracy: 0.8467\n", + "Epoch 17/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.4385 - accuracy: 0.8004 - val_loss: 0.3822 - val_accuracy: 0.8456\n", + "Epoch 18/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.4313 - accuracy: 0.8044 - val_loss: 0.3901 - val_accuracy: 0.8544\n", + "Epoch 19/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.4271 - accuracy: 0.8081 - val_loss: 0.3775 - val_accuracy: 0.8478\n", + "Epoch 20/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.4176 - accuracy: 0.8106 - val_loss: 0.3641 - val_accuracy: 0.8711\n", + "Epoch 21/50\n", + "41/41 [==============================] - 1s 29ms/step - loss: 0.4048 - accuracy: 0.8175 - val_loss: 0.3608 - val_accuracy: 0.8633\n", + "Epoch 22/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.3908 - accuracy: 0.8307 - val_loss: 0.3501 - val_accuracy: 0.8689\n", + "Epoch 23/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.3904 - accuracy: 0.8254 - val_loss: 0.3451 - val_accuracy: 0.8667\n", + "Epoch 24/50\n", + "41/41 [==============================] - 1s 29ms/step - loss: 0.3869 - accuracy: 0.8279 - val_loss: 0.3471 - val_accuracy: 0.8589\n", + "Epoch 25/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.3799 - accuracy: 0.8375 - val_loss: 0.3419 - val_accuracy: 0.8733\n", + "Epoch 26/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.3723 - accuracy: 0.8358 - val_loss: 0.3264 - val_accuracy: 0.8811\n", + "Epoch 27/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.3617 - accuracy: 0.8477 - val_loss: 0.3225 - val_accuracy: 0.8833\n", + "Epoch 28/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.3668 - accuracy: 0.8375 - val_loss: 0.3191 - val_accuracy: 0.8867\n", + "Epoch 29/50\n", + "41/41 [==============================] - 1s 30ms/step - loss: 0.3563 - accuracy: 0.8478 - val_loss: 0.3237 - val_accuracy: 0.8789\n", + "Epoch 30/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.3589 - accuracy: 0.8495 - val_loss: 0.3078 - val_accuracy: 0.8867\n", + "Epoch 31/50\n", + "41/41 [==============================] - 1s 30ms/step - loss: 0.3461 - accuracy: 0.8520 - val_loss: 0.3095 - val_accuracy: 0.8944\n", + "Epoch 32/50\n", + "41/41 [==============================] - 1s 33ms/step - loss: 0.3429 - accuracy: 0.8533 - val_loss: 0.3091 - val_accuracy: 0.8911\n", + "Epoch 33/50\n", + "41/41 [==============================] - 1s 32ms/step - loss: 0.3432 - accuracy: 0.8478 - val_loss: 0.3105 - val_accuracy: 0.8811\n", + "Epoch 34/50\n", + "41/41 [==============================] - 1s 31ms/step - loss: 0.3408 - accuracy: 0.8574 - val_loss: 0.3007 - val_accuracy: 0.8956\n", + "Epoch 35/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.3287 - accuracy: 0.8574 - val_loss: 0.3005 - val_accuracy: 0.8933\n", + "Epoch 36/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.3300 - accuracy: 0.8604 - val_loss: 0.3027 - val_accuracy: 0.8822\n", + "Epoch 37/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.3247 - accuracy: 0.8623 - val_loss: 0.2915 - val_accuracy: 0.8900\n", + "Epoch 38/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.3269 - accuracy: 0.8593 - val_loss: 0.2910 - val_accuracy: 0.8956\n", + "Epoch 39/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.3145 - accuracy: 0.8635 - val_loss: 0.2868 - val_accuracy: 0.9000\n", + "Epoch 40/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.3249 - accuracy: 0.8594 - val_loss: 0.2953 - val_accuracy: 0.8878\n", + "Epoch 41/50\n", + "41/41 [==============================] - 1s 30ms/step - loss: 0.3152 - accuracy: 0.8652 - val_loss: 0.2851 - val_accuracy: 0.8922\n", + "Epoch 42/50\n", + "41/41 [==============================] - 1s 30ms/step - loss: 0.3121 - accuracy: 0.8701 - val_loss: 0.2877 - val_accuracy: 0.8889\n", + "Epoch 43/50\n", + "41/41 [==============================] - 1s 26ms/step - loss: 0.3052 - accuracy: 0.8707 - val_loss: 0.2836 - val_accuracy: 0.8978\n", + "Epoch 44/50\n", + "41/41 [==============================] - 1s 29ms/step - loss: 0.3062 - accuracy: 0.8677 - val_loss: 0.2815 - val_accuracy: 0.9000\n", + "Epoch 45/50\n", + "41/41 [==============================] - 1s 29ms/step - loss: 0.3064 - accuracy: 0.8704 - val_loss: 0.2808 - val_accuracy: 0.8922\n", + "Epoch 46/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.3041 - accuracy: 0.8736 - val_loss: 0.2872 - val_accuracy: 0.9011\n", + "Epoch 47/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.3002 - accuracy: 0.8751 - val_loss: 0.2843 - val_accuracy: 0.8900\n", + "Epoch 48/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.3091 - accuracy: 0.8664 - val_loss: 0.2843 - val_accuracy: 0.8944\n", + "Epoch 49/50\n", + "41/41 [==============================] - 1s 27ms/step - loss: 0.2988 - accuracy: 0.8753 - val_loss: 0.2778 - val_accuracy: 0.9056\n", + "Epoch 50/50\n", + "41/41 [==============================] - 1s 28ms/step - loss: 0.2978 - accuracy: 0.8740 - val_loss: 0.2809 - val_accuracy: 0.9022\n" ] } ], "source": [ "history = model.fit(x_train, y_train,\n", - " epochs=200, validation_split=0.1,\n", + " epochs=50, validation_split=0.1,\n", " shuffle=True, batch_size=200,\n", " verbose=1)\n", "\n",