{"cells":[{"cell_type":"markdown","source":["#**Text Generation Using Different Types of RNN Model**"],"metadata":{"id":"V4-Ze5EziBte"}},{"cell_type":"code","source":["from google.colab import drive\n","drive.mount('/content/drive')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"bwGkITtH4b_w","executionInfo":{"status":"ok","timestamp":1684904311864,"user_tz":-480,"elapsed":28049,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"outputId":"b24b8078-a1f1-4f9f-e553-345780705f15"},"execution_count":2,"outputs":[{"output_type":"stream","name":"stdout","text":["Mounted at /content/drive\n"]}]},{"cell_type":"code","source":["import tensorflow as tf\n","from tensorflow.keras.preprocessing.sequence import pad_sequences\n","from tensorflow.keras.layers import Embedding, LSTM, GRU, Dense, Bidirectional\n","from tensorflow.keras.preprocessing.text import Tokenizer\n","from tensorflow.keras.models import Sequential\n","from tensorflow.keras.optimizers import Adam \n","from tensorflow.keras.callbacks import EarlyStopping\n","import numpy as np\n","import pandas as pd\n","import matplotlib.pyplot as plt"],"metadata":{"id":"ehxlGuKyiAdx","executionInfo":{"status":"ok","timestamp":1684904387525,"user_tz":-480,"elapsed":399,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}}},"execution_count":5,"outputs":[]},{"cell_type":"code","source":["# Convert CSV file to Txt File\n","\n","csv_data = \"/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/Reviews.csv\"  # Path of the CSV\n","\n","df = pd.read_csv(csv_data)\n","content = '\\n'.join(df['Reviews'].astype(str))\n","\n","with open('/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/my_file.txt', 'w') as file:\n","    file.write(content)"],"metadata":{"id":"KVKnJPXJCFxf","executionInfo":{"status":"ok","timestamp":1684904391028,"user_tz":-480,"elapsed":963,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}}},"execution_count":6,"outputs":[]},{"cell_type":"code","source":["# Create functions for every model\n","\n","def LSTM_Model(total_words, max_sequence_len):\n","    model = Sequential()\n","    model.add(Embedding(total_words, 100, input_length=max_sequence_len - 1))\n","    model.add(LSTM(128))\n","    model.add(Dense(100, activation='relu'))\n","    model.add(Dense(total_words, activation='softmax'))\n","    return model\n","\n","def GRU_Model(total_words, max_sequence_len):\n","    model = Sequential()\n","    model.add(Embedding(total_words, 100, input_length=max_sequence_len - 1))\n","    model.add(GRU(128))\n","    model.add(Dense(100, activation='relu'))\n","    model.add(Dense(total_words, activation='softmax'))\n","    return model\n","\n","def BiLSTM_Model(total_words, max_sequence_len):\n","    model = Sequential()\n","    model.add(Embedding(total_words, 100, input_length=max_sequence_len - 1))\n","    model.add(Bidirectional(LSTM(128)))\n","    model.add(Dense(100, activation='relu'))\n","    model.add(Dense(total_words, activation='softmax'))\n","    return model"],"metadata":{"id":"Lr_ESuj0hPEQ","executionInfo":{"status":"ok","timestamp":1684904394172,"user_tz":-480,"elapsed":17,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}}},"execution_count":7,"outputs":[]},{"cell_type":"code","source":["# Read data from file\n","data = open('/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/my_file.txt').read()\n","\n","# Preprocess the data\n","c_data = [line.strip().lower() for line in data.split(\"\\n\")]"],"metadata":{"id":"CbkKBRl0ica5","executionInfo":{"status":"ok","timestamp":1684904397973,"user_tz":-480,"elapsed":481,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}}},"execution_count":8,"outputs":[]},{"cell_type":"code","source":["# Tokenize the data\n","tokenizer = Tokenizer()\n","tokenizer.fit_on_texts(c_data)\n","total_words = len(tokenizer.word_index) + 1\n","\n","# Generate input sequences\n","input_sequences = []\n","for line in c_data:\n","    token_list = tokenizer.texts_to_sequences([line])[0]\n","    for i in range(1, len(token_list)):\n","        n_gram_sequence = token_list[:i + 1]\n","        input_sequences.append(n_gram_sequence)"],"metadata":{"id":"zplinDsUidm-","executionInfo":{"status":"ok","timestamp":1684904398337,"user_tz":-480,"elapsed":11,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}}},"execution_count":9,"outputs":[]},{"cell_type":"code","source":["# Pad input sequences\n","max_sequence_len = max([len(x) for x in input_sequences])\n","input_sequences = np.array(pad_sequences(input_sequences,\n","                                         maxlen=max_sequence_len,\n","                                         padding='pre'))\n","\n","# create predictors and label\n","xs, labels = input_sequences[:,:-1], input_sequences[:,-1]\n","ys = tf.keras.utils.to_categorical(labels, num_classes=total_words)"],"metadata":{"id":"YUE97uWFid0z","executionInfo":{"status":"ok","timestamp":1684904402272,"user_tz":-480,"elapsed":383,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}}},"execution_count":10,"outputs":[]},{"cell_type":"code","source":["# Build LSTM model\n","lstm_model = LSTM_Model(total_words, max_sequence_len)\n","\n","# Build GRU model\n","gru_model = GRU_Model(total_words, max_sequence_len)\n","\n","# Build Bidirectional LSTM model\n","bilstm_model = BiLSTM_Model(total_words, max_sequence_len)"],"metadata":{"id":"ME_A0ocqisTR"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["optimizer = 'rmsprop' # Adam(learning_rate=0.01)\n","epoch = 1000\n","patience = 20\n","size = 12\n","\n","# Compile models\n","lstm_model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n","gru_model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n","bilstm_model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy'])"],"metadata":{"id":"LFcWFF6-jFWE"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Define early stopping\n","earlystop = EarlyStopping(monitor='loss', min_delta=0, patience=patience, verbose=2)"],"metadata":{"id":"iTwDwPDki3Fr"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Train LSTM model\n","print(\"Training LSTM model...\")\n","lstm_history = lstm_model.fit(xs, ys, batch_size=size,\n","                              epochs=epoch, verbose=2, callbacks=[earlystop])"],"metadata":{"id":"PJz6Rrm4iyOF","executionInfo":{"status":"ok","timestamp":1684898160200,"user_tz":-480,"elapsed":725794,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"fdd483c6-9b3e-4975-c519-55c55f09b3d6"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Training LSTM model...\n","Epoch 1/1000\n","359/359 - 18s - loss: 6.0307 - accuracy: 0.0562 - 18s/epoch - 50ms/step\n","Epoch 2/1000\n","359/359 - 4s - loss: 5.7825 - accuracy: 0.0572 - 4s/epoch - 10ms/step\n","Epoch 3/1000\n","359/359 - 3s - loss: 5.6474 - accuracy: 0.0660 - 3s/epoch - 8ms/step\n","Epoch 4/1000\n","359/359 - 3s - loss: 5.5055 - accuracy: 0.0760 - 3s/epoch - 9ms/step\n","Epoch 5/1000\n","359/359 - 3s - loss: 5.3376 - accuracy: 0.0913 - 3s/epoch - 8ms/step\n","Epoch 6/1000\n","359/359 - 3s - loss: 5.1949 - accuracy: 0.1125 - 3s/epoch - 7ms/step\n","Epoch 7/1000\n","359/359 - 2s - loss: 5.0578 - accuracy: 0.1343 - 2s/epoch - 7ms/step\n","Epoch 8/1000\n","359/359 - 2s - loss: 4.9263 - accuracy: 0.1487 - 2s/epoch - 6ms/step\n","Epoch 9/1000\n","359/359 - 3s - loss: 4.8053 - accuracy: 0.1761 - 3s/epoch - 8ms/step\n","Epoch 10/1000\n","359/359 - 2s - loss: 4.6694 - accuracy: 0.1914 - 2s/epoch - 7ms/step\n","Epoch 11/1000\n","359/359 - 2s - loss: 4.5465 - accuracy: 0.2121 - 2s/epoch - 6ms/step\n","Epoch 12/1000\n","359/359 - 2s - loss: 4.4245 - accuracy: 0.2365 - 2s/epoch - 7ms/step\n","Epoch 13/1000\n","359/359 - 2s - loss: 4.3108 - accuracy: 0.2523 - 2s/epoch - 7ms/step\n","Epoch 14/1000\n","359/359 - 3s - loss: 4.1927 - accuracy: 0.2739 - 3s/epoch - 8ms/step\n","Epoch 15/1000\n","359/359 - 2s - loss: 4.0759 - accuracy: 0.2993 - 2s/epoch - 6ms/step\n","Epoch 16/1000\n","359/359 - 2s - loss: 3.9744 - accuracy: 0.3141 - 2s/epoch - 6ms/step\n","Epoch 17/1000\n","359/359 - 2s - loss: 3.8791 - accuracy: 0.3404 - 2s/epoch - 6ms/step\n","Epoch 18/1000\n","359/359 - 3s - loss: 3.7773 - accuracy: 0.3532 - 3s/epoch - 7ms/step\n","Epoch 19/1000\n","359/359 - 2s - loss: 3.6777 - accuracy: 0.3743 - 2s/epoch - 7ms/step\n","Epoch 20/1000\n","359/359 - 2s - loss: 3.5892 - accuracy: 0.3882 - 2s/epoch - 7ms/step\n","Epoch 21/1000\n","359/359 - 2s - loss: 3.5090 - accuracy: 0.4047 - 2s/epoch - 6ms/step\n","Epoch 22/1000\n","359/359 - 2s - loss: 3.4217 - accuracy: 0.4164 - 2s/epoch - 6ms/step\n","Epoch 23/1000\n","359/359 - 3s - loss: 3.3387 - accuracy: 0.4375 - 3s/epoch - 7ms/step\n","Epoch 24/1000\n","359/359 - 2s - loss: 3.2570 - accuracy: 0.4487 - 2s/epoch - 6ms/step\n","Epoch 25/1000\n","359/359 - 2s - loss: 3.1840 - accuracy: 0.4607 - 2s/epoch - 6ms/step\n","Epoch 26/1000\n","359/359 - 2s - loss: 3.1058 - accuracy: 0.4689 - 2s/epoch - 6ms/step\n","Epoch 27/1000\n","359/359 - 2s - loss: 3.0263 - accuracy: 0.4819 - 2s/epoch - 6ms/step\n","Epoch 28/1000\n","359/359 - 2s - loss: 2.9496 - accuracy: 0.4993 - 2s/epoch - 7ms/step\n","Epoch 29/1000\n","359/359 - 2s - loss: 2.8767 - accuracy: 0.5107 - 2s/epoch - 7ms/step\n","Epoch 30/1000\n","359/359 - 2s - loss: 2.8127 - accuracy: 0.5188 - 2s/epoch - 6ms/step\n","Epoch 31/1000\n","359/359 - 2s - loss: 2.7424 - accuracy: 0.5346 - 2s/epoch - 7ms/step\n","Epoch 32/1000\n","359/359 - 2s - loss: 2.6819 - accuracy: 0.5451 - 2s/epoch - 6ms/step\n","Epoch 33/1000\n","359/359 - 2s - loss: 2.6188 - accuracy: 0.5539 - 2s/epoch - 7ms/step\n","Epoch 34/1000\n","359/359 - 2s - loss: 2.5605 - accuracy: 0.5653 - 2s/epoch - 7ms/step\n","Epoch 35/1000\n","359/359 - 2s - loss: 2.4950 - accuracy: 0.5816 - 2s/epoch - 6ms/step\n","Epoch 36/1000\n","359/359 - 2s - loss: 2.4456 - accuracy: 0.5827 - 2s/epoch - 6ms/step\n","Epoch 37/1000\n","359/359 - 2s - loss: 2.3879 - accuracy: 0.5976 - 2s/epoch - 6ms/step\n","Epoch 38/1000\n","359/359 - 3s - loss: 2.3331 - accuracy: 0.6118 - 3s/epoch - 7ms/step\n","Epoch 39/1000\n","359/359 - 2s - loss: 2.2818 - accuracy: 0.6127 - 2s/epoch - 6ms/step\n","Epoch 40/1000\n","359/359 - 2s - loss: 2.2384 - accuracy: 0.6255 - 2s/epoch - 6ms/step\n","Epoch 41/1000\n","359/359 - 2s - loss: 2.1842 - accuracy: 0.6324 - 2s/epoch - 6ms/step\n","Epoch 42/1000\n","359/359 - 2s - loss: 2.1389 - accuracy: 0.6389 - 2s/epoch - 7ms/step\n","Epoch 43/1000\n","359/359 - 3s - loss: 2.0909 - accuracy: 0.6468 - 3s/epoch - 8ms/step\n","Epoch 44/1000\n","359/359 - 2s - loss: 2.0420 - accuracy: 0.6552 - 2s/epoch - 6ms/step\n","Epoch 45/1000\n","359/359 - 2s - loss: 2.0037 - accuracy: 0.6610 - 2s/epoch - 7ms/step\n","Epoch 46/1000\n","359/359 - 3s - loss: 1.9620 - accuracy: 0.6678 - 3s/epoch - 7ms/step\n","Epoch 47/1000\n","359/359 - 2s - loss: 1.9229 - accuracy: 0.6733 - 2s/epoch - 6ms/step\n","Epoch 48/1000\n","359/359 - 3s - loss: 1.8808 - accuracy: 0.6733 - 3s/epoch - 7ms/step\n","Epoch 49/1000\n","359/359 - 2s - loss: 1.8416 - accuracy: 0.6829 - 2s/epoch - 6ms/step\n","Epoch 50/1000\n","359/359 - 2s - loss: 1.8071 - accuracy: 0.6887 - 2s/epoch - 6ms/step\n","Epoch 51/1000\n","359/359 - 2s - loss: 1.7719 - accuracy: 0.6933 - 2s/epoch - 6ms/step\n","Epoch 52/1000\n","359/359 - 2s - loss: 1.7448 - accuracy: 0.6961 - 2s/epoch - 6ms/step\n","Epoch 53/1000\n","359/359 - 3s - loss: 1.7092 - accuracy: 0.7014 - 3s/epoch - 7ms/step\n","Epoch 54/1000\n","359/359 - 2s - loss: 1.6753 - accuracy: 0.7028 - 2s/epoch - 6ms/step\n","Epoch 55/1000\n","359/359 - 2s - loss: 1.6479 - accuracy: 0.7047 - 2s/epoch - 6ms/step\n","Epoch 56/1000\n","359/359 - 2s - loss: 1.6219 - accuracy: 0.7082 - 2s/epoch - 6ms/step\n","Epoch 57/1000\n","359/359 - 2s - loss: 1.5867 - accuracy: 0.7158 - 2s/epoch - 6ms/step\n","Epoch 58/1000\n","359/359 - 3s - loss: 1.5644 - accuracy: 0.7186 - 3s/epoch - 7ms/step\n","Epoch 59/1000\n","359/359 - 2s - loss: 1.5375 - accuracy: 0.7284 - 2s/epoch - 6ms/step\n","Epoch 60/1000\n","359/359 - 2s - loss: 1.5101 - accuracy: 0.7268 - 2s/epoch - 6ms/step\n","Epoch 61/1000\n","359/359 - 2s - loss: 1.4859 - accuracy: 0.7330 - 2s/epoch - 6ms/step\n","Epoch 62/1000\n","359/359 - 2s - loss: 1.4614 - accuracy: 0.7419 - 2s/epoch - 6ms/step\n","Epoch 63/1000\n","359/359 - 3s - loss: 1.4380 - accuracy: 0.7435 - 3s/epoch - 7ms/step\n","Epoch 64/1000\n","359/359 - 2s - loss: 1.4154 - accuracy: 0.7470 - 2s/epoch - 6ms/step\n","Epoch 65/1000\n","359/359 - 2s - loss: 1.3884 - accuracy: 0.7514 - 2s/epoch - 6ms/step\n","Epoch 66/1000\n","359/359 - 2s - loss: 1.3668 - accuracy: 0.7539 - 2s/epoch - 6ms/step\n","Epoch 67/1000\n","359/359 - 2s - loss: 1.3432 - accuracy: 0.7588 - 2s/epoch - 6ms/step\n","Epoch 68/1000\n","359/359 - 3s - loss: 1.3337 - accuracy: 0.7598 - 3s/epoch - 7ms/step\n","Epoch 69/1000\n","359/359 - 2s - loss: 1.3048 - accuracy: 0.7649 - 2s/epoch - 6ms/step\n","Epoch 70/1000\n","359/359 - 2s - loss: 1.2846 - accuracy: 0.7697 - 2s/epoch - 6ms/step\n","Epoch 71/1000\n","359/359 - 2s - loss: 1.2692 - accuracy: 0.7732 - 2s/epoch - 6ms/step\n","Epoch 72/1000\n","359/359 - 2s - loss: 1.2492 - accuracy: 0.7767 - 2s/epoch - 6ms/step\n","Epoch 73/1000\n","359/359 - 3s - loss: 1.2328 - accuracy: 0.7818 - 3s/epoch - 8ms/step\n","Epoch 74/1000\n","359/359 - 2s - loss: 1.2122 - accuracy: 0.7846 - 2s/epoch - 6ms/step\n","Epoch 75/1000\n","359/359 - 2s - loss: 1.2147 - accuracy: 0.7846 - 2s/epoch - 6ms/step\n","Epoch 76/1000\n","359/359 - 2s - loss: 1.2017 - accuracy: 0.7881 - 2s/epoch - 6ms/step\n","Epoch 77/1000\n","359/359 - 2s - loss: 1.1827 - accuracy: 0.7897 - 2s/epoch - 6ms/step\n","Epoch 78/1000\n","359/359 - 3s - loss: 1.1624 - accuracy: 0.7967 - 3s/epoch - 8ms/step\n","Epoch 79/1000\n","359/359 - 2s - loss: 1.1399 - accuracy: 0.8034 - 2s/epoch - 6ms/step\n","Epoch 80/1000\n","359/359 - 2s - loss: 1.1209 - accuracy: 0.8083 - 2s/epoch - 6ms/step\n","Epoch 81/1000\n","359/359 - 2s - loss: 1.1113 - accuracy: 0.8102 - 2s/epoch - 6ms/step\n","Epoch 82/1000\n","359/359 - 2s - loss: 1.0925 - accuracy: 0.8164 - 2s/epoch - 6ms/step\n","Epoch 83/1000\n","359/359 - 3s - loss: 1.0765 - accuracy: 0.8192 - 3s/epoch - 7ms/step\n","Epoch 84/1000\n","359/359 - 2s - loss: 1.0607 - accuracy: 0.8239 - 2s/epoch - 6ms/step\n","Epoch 85/1000\n","359/359 - 2s - loss: 1.0453 - accuracy: 0.8260 - 2s/epoch - 6ms/step\n","Epoch 86/1000\n","359/359 - 2s - loss: 1.0312 - accuracy: 0.8313 - 2s/epoch - 6ms/step\n","Epoch 87/1000\n","359/359 - 2s - loss: 1.0152 - accuracy: 0.8388 - 2s/epoch - 6ms/step\n","Epoch 88/1000\n","359/359 - 3s - loss: 1.0025 - accuracy: 0.8413 - 3s/epoch - 7ms/step\n","Epoch 89/1000\n","359/359 - 2s - loss: 1.0747 - accuracy: 0.8232 - 2s/epoch - 6ms/step\n","Epoch 90/1000\n","359/359 - 2s - loss: 0.9850 - accuracy: 0.8406 - 2s/epoch - 6ms/step\n","Epoch 91/1000\n","359/359 - 2s - loss: 0.9699 - accuracy: 0.8469 - 2s/epoch - 6ms/step\n","Epoch 92/1000\n","359/359 - 2s - loss: 0.9551 - accuracy: 0.8494 - 2s/epoch - 6ms/step\n","Epoch 93/1000\n","359/359 - 2s - loss: 0.9428 - accuracy: 0.8532 - 2s/epoch - 7ms/step\n","Epoch 94/1000\n","359/359 - 2s - loss: 0.9339 - accuracy: 0.8543 - 2s/epoch - 6ms/step\n","Epoch 95/1000\n","359/359 - 2s - loss: 0.9201 - accuracy: 0.8618 - 2s/epoch - 6ms/step\n","Epoch 96/1000\n","359/359 - 2s - loss: 0.9088 - accuracy: 0.8615 - 2s/epoch - 6ms/step\n","Epoch 97/1000\n","359/359 - 2s - loss: 0.8957 - accuracy: 0.8645 - 2s/epoch - 6ms/step\n","Epoch 98/1000\n","359/359 - 2s - loss: 0.8833 - accuracy: 0.8729 - 2s/epoch - 7ms/step\n","Epoch 99/1000\n","359/359 - 2s - loss: 0.8717 - accuracy: 0.8738 - 2s/epoch - 6ms/step\n","Epoch 100/1000\n","359/359 - 2s - loss: 0.8760 - accuracy: 0.8729 - 2s/epoch - 6ms/step\n","Epoch 101/1000\n","359/359 - 2s - loss: 0.8575 - accuracy: 0.8796 - 2s/epoch - 6ms/step\n","Epoch 102/1000\n","359/359 - 2s - loss: 0.8450 - accuracy: 0.8829 - 2s/epoch - 6ms/step\n","Epoch 103/1000\n","359/359 - 2s - loss: 0.8357 - accuracy: 0.8801 - 2s/epoch - 7ms/step\n","Epoch 104/1000\n","359/359 - 2s - loss: 0.8266 - accuracy: 0.8873 - 2s/epoch - 7ms/step\n","Epoch 105/1000\n","359/359 - 2s - loss: 0.8127 - accuracy: 0.8862 - 2s/epoch - 7ms/step\n","Epoch 106/1000\n","359/359 - 2s - loss: 0.8023 - accuracy: 0.8975 - 2s/epoch - 6ms/step\n","Epoch 107/1000\n","359/359 - 2s - loss: 0.7923 - accuracy: 0.8980 - 2s/epoch - 6ms/step\n","Epoch 108/1000\n","359/359 - 2s - loss: 0.7847 - accuracy: 0.9031 - 2s/epoch - 7ms/step\n","Epoch 109/1000\n","359/359 - 2s - loss: 0.7763 - accuracy: 0.8999 - 2s/epoch - 7ms/step\n","Epoch 110/1000\n","359/359 - 2s - loss: 0.7665 - accuracy: 0.9036 - 2s/epoch - 6ms/step\n","Epoch 111/1000\n","359/359 - 2s - loss: 0.7651 - accuracy: 0.9029 - 2s/epoch - 6ms/step\n","Epoch 112/1000\n","359/359 - 2s - loss: 0.7552 - accuracy: 0.9064 - 2s/epoch - 6ms/step\n","Epoch 113/1000\n","359/359 - 3s - loss: 0.7443 - accuracy: 0.9092 - 3s/epoch - 7ms/step\n","Epoch 114/1000\n","359/359 - 2s - loss: 0.7341 - accuracy: 0.9122 - 2s/epoch - 6ms/step\n","Epoch 115/1000\n","359/359 - 2s - loss: 0.7301 - accuracy: 0.9143 - 2s/epoch - 6ms/step\n","Epoch 116/1000\n","359/359 - 2s - loss: 0.7247 - accuracy: 0.9140 - 2s/epoch - 6ms/step\n","Epoch 117/1000\n","359/359 - 2s - loss: 0.7139 - accuracy: 0.9157 - 2s/epoch - 6ms/step\n","Epoch 118/1000\n","359/359 - 2s - loss: 0.7014 - accuracy: 0.9184 - 2s/epoch - 7ms/step\n","Epoch 119/1000\n","359/359 - 2s - loss: 0.6964 - accuracy: 0.9203 - 2s/epoch - 7ms/step\n","Epoch 120/1000\n","359/359 - 2s - loss: 0.6901 - accuracy: 0.9231 - 2s/epoch - 6ms/step\n","Epoch 121/1000\n","359/359 - 2s - loss: 0.6799 - accuracy: 0.9226 - 2s/epoch - 6ms/step\n","Epoch 122/1000\n","359/359 - 2s - loss: 0.6795 - accuracy: 0.9254 - 2s/epoch - 6ms/step\n","Epoch 123/1000\n","359/359 - 2s - loss: 0.6674 - accuracy: 0.9254 - 2s/epoch - 6ms/step\n","Epoch 124/1000\n","359/359 - 3s - loss: 0.6589 - accuracy: 0.9282 - 3s/epoch - 7ms/step\n","Epoch 125/1000\n","359/359 - 2s - loss: 0.6481 - accuracy: 0.9317 - 2s/epoch - 6ms/step\n","Epoch 126/1000\n","359/359 - 2s - loss: 0.6420 - accuracy: 0.9338 - 2s/epoch - 6ms/step\n","Epoch 127/1000\n","359/359 - 2s - loss: 0.6350 - accuracy: 0.9352 - 2s/epoch - 6ms/step\n","Epoch 128/1000\n","359/359 - 2s - loss: 0.6276 - accuracy: 0.9363 - 2s/epoch - 6ms/step\n","Epoch 129/1000\n","359/359 - 3s - loss: 0.6089 - accuracy: 0.9421 - 3s/epoch - 7ms/step\n","Epoch 130/1000\n","359/359 - 2s - loss: 0.6253 - accuracy: 0.9389 - 2s/epoch - 6ms/step\n","Epoch 131/1000\n","359/359 - 2s - loss: 0.6172 - accuracy: 0.9373 - 2s/epoch - 6ms/step\n","Epoch 132/1000\n","359/359 - 2s - loss: 0.6229 - accuracy: 0.9336 - 2s/epoch - 6ms/step\n","Epoch 133/1000\n","359/359 - 2s - loss: 0.6469 - accuracy: 0.9294 - 2s/epoch - 6ms/step\n","Epoch 134/1000\n","359/359 - 3s - loss: 0.6693 - accuracy: 0.9229 - 3s/epoch - 7ms/step\n","Epoch 135/1000\n","359/359 - 2s - loss: 0.6634 - accuracy: 0.9198 - 2s/epoch - 6ms/step\n","Epoch 136/1000\n","359/359 - 2s - loss: 0.6486 - accuracy: 0.9240 - 2s/epoch - 6ms/step\n","Epoch 137/1000\n","359/359 - 2s - loss: 0.6269 - accuracy: 0.9303 - 2s/epoch - 6ms/step\n","Epoch 138/1000\n","359/359 - 2s - loss: 0.6234 - accuracy: 0.9273 - 2s/epoch - 6ms/step\n","Epoch 139/1000\n","359/359 - 3s - loss: 0.6194 - accuracy: 0.9308 - 3s/epoch - 7ms/step\n","Epoch 140/1000\n","359/359 - 2s - loss: 0.6179 - accuracy: 0.9310 - 2s/epoch - 6ms/step\n","Epoch 141/1000\n","359/359 - 2s - loss: 0.6216 - accuracy: 0.9315 - 2s/epoch - 6ms/step\n","Epoch 142/1000\n","359/359 - 2s - loss: 0.5949 - accuracy: 0.9356 - 2s/epoch - 6ms/step\n","Epoch 143/1000\n","359/359 - 2s - loss: 0.5804 - accuracy: 0.9391 - 2s/epoch - 6ms/step\n","Epoch 144/1000\n","359/359 - 3s - loss: 0.5774 - accuracy: 0.9408 - 3s/epoch - 9ms/step\n","Epoch 145/1000\n","359/359 - 2s - loss: 0.5644 - accuracy: 0.9412 - 2s/epoch - 6ms/step\n","Epoch 146/1000\n","359/359 - 2s - loss: 0.5542 - accuracy: 0.9417 - 2s/epoch - 6ms/step\n","Epoch 147/1000\n","359/359 - 2s - loss: 0.5469 - accuracy: 0.9438 - 2s/epoch - 6ms/step\n","Epoch 148/1000\n","359/359 - 2s - loss: 0.5409 - accuracy: 0.9468 - 2s/epoch - 6ms/step\n","Epoch 149/1000\n","359/359 - 3s - loss: 0.5373 - accuracy: 0.9461 - 3s/epoch - 7ms/step\n","Epoch 150/1000\n","359/359 - 2s - loss: 0.5260 - accuracy: 0.9487 - 2s/epoch - 6ms/step\n","Epoch 151/1000\n","359/359 - 2s - loss: 0.5190 - accuracy: 0.9505 - 2s/epoch - 6ms/step\n","Epoch 152/1000\n","359/359 - 2s - loss: 0.5093 - accuracy: 0.9521 - 2s/epoch - 6ms/step\n","Epoch 153/1000\n","359/359 - 2s - loss: 0.5029 - accuracy: 0.9517 - 2s/epoch - 6ms/step\n","Epoch 154/1000\n","359/359 - 2s - loss: 0.4963 - accuracy: 0.9528 - 2s/epoch - 7ms/step\n","Epoch 155/1000\n","359/359 - 2s - loss: 0.4931 - accuracy: 0.9524 - 2s/epoch - 6ms/step\n","Epoch 156/1000\n","359/359 - 2s - loss: 0.4840 - accuracy: 0.9556 - 2s/epoch - 6ms/step\n","Epoch 157/1000\n","359/359 - 2s - loss: 0.4799 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 158/1000\n","359/359 - 2s - loss: 0.4735 - accuracy: 0.9566 - 2s/epoch - 6ms/step\n","Epoch 159/1000\n","359/359 - 2s - loss: 0.4697 - accuracy: 0.9561 - 2s/epoch - 7ms/step\n","Epoch 160/1000\n","359/359 - 2s - loss: 0.4646 - accuracy: 0.9575 - 2s/epoch - 7ms/step\n","Epoch 161/1000\n","359/359 - 2s - loss: 0.4620 - accuracy: 0.9559 - 2s/epoch - 6ms/step\n","Epoch 162/1000\n","359/359 - 2s - loss: 0.4613 - accuracy: 0.9556 - 2s/epoch - 6ms/step\n","Epoch 163/1000\n","359/359 - 2s - loss: 0.4533 - accuracy: 0.9568 - 2s/epoch - 6ms/step\n","Epoch 164/1000\n","359/359 - 2s - loss: 0.4780 - accuracy: 0.9498 - 2s/epoch - 6ms/step\n","Epoch 165/1000\n","359/359 - 2s - loss: 0.4674 - accuracy: 0.9503 - 2s/epoch - 7ms/step\n","Epoch 166/1000\n","359/359 - 2s - loss: 0.4613 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 167/1000\n","359/359 - 2s - loss: 0.4563 - accuracy: 0.9545 - 2s/epoch - 6ms/step\n","Epoch 168/1000\n","359/359 - 2s - loss: 0.4725 - accuracy: 0.9498 - 2s/epoch - 6ms/step\n","Epoch 169/1000\n","359/359 - 2s - loss: 0.4688 - accuracy: 0.9493 - 2s/epoch - 6ms/step\n","Epoch 170/1000\n","359/359 - 2s - loss: 0.4566 - accuracy: 0.9500 - 2s/epoch - 7ms/step\n","Epoch 171/1000\n","359/359 - 2s - loss: 0.4450 - accuracy: 0.9554 - 2s/epoch - 6ms/step\n","Epoch 172/1000\n","359/359 - 2s - loss: 0.4679 - accuracy: 0.9470 - 2s/epoch - 6ms/step\n","Epoch 173/1000\n","359/359 - 2s - loss: 0.4480 - accuracy: 0.9533 - 2s/epoch - 6ms/step\n","Epoch 174/1000\n","359/359 - 2s - loss: 0.4353 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 175/1000\n","359/359 - 3s - loss: 0.4184 - accuracy: 0.9579 - 3s/epoch - 7ms/step\n","Epoch 176/1000\n","359/359 - 2s - loss: 0.4165 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 177/1000\n","359/359 - 2s - loss: 0.4263 - accuracy: 0.9542 - 2s/epoch - 6ms/step\n","Epoch 178/1000\n","359/359 - 2s - loss: 0.4131 - accuracy: 0.9566 - 2s/epoch - 6ms/step\n","Epoch 179/1000\n","359/359 - 2s - loss: 0.4113 - accuracy: 0.9582 - 2s/epoch - 6ms/step\n","Epoch 180/1000\n","359/359 - 3s - loss: 0.3993 - accuracy: 0.9584 - 3s/epoch - 7ms/step\n","Epoch 181/1000\n","359/359 - 2s - loss: 0.4063 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 182/1000\n","359/359 - 2s - loss: 0.4050 - accuracy: 0.9556 - 2s/epoch - 6ms/step\n","Epoch 183/1000\n","359/359 - 2s - loss: 0.4033 - accuracy: 0.9547 - 2s/epoch - 6ms/step\n","Epoch 184/1000\n","359/359 - 2s - loss: 0.4002 - accuracy: 0.9549 - 2s/epoch - 6ms/step\n","Epoch 185/1000\n","359/359 - 3s - loss: 0.4098 - accuracy: 0.9526 - 3s/epoch - 8ms/step\n","Epoch 186/1000\n","359/359 - 2s - loss: 0.4324 - accuracy: 0.9489 - 2s/epoch - 6ms/step\n","Epoch 187/1000\n","359/359 - 2s - loss: 0.4208 - accuracy: 0.9487 - 2s/epoch - 6ms/step\n","Epoch 188/1000\n","359/359 - 2s - loss: 0.4048 - accuracy: 0.9519 - 2s/epoch - 6ms/step\n","Epoch 189/1000\n","359/359 - 2s - loss: 0.3928 - accuracy: 0.9545 - 2s/epoch - 6ms/step\n","Epoch 190/1000\n","359/359 - 3s - loss: 0.3891 - accuracy: 0.9559 - 3s/epoch - 7ms/step\n","Epoch 191/1000\n","359/359 - 2s - loss: 0.3766 - accuracy: 0.9554 - 2s/epoch - 6ms/step\n","Epoch 192/1000\n","359/359 - 2s - loss: 0.3796 - accuracy: 0.9531 - 2s/epoch - 6ms/step\n","Epoch 193/1000\n","359/359 - 2s - loss: 0.3760 - accuracy: 0.9549 - 2s/epoch - 6ms/step\n","Epoch 194/1000\n","359/359 - 2s - loss: 0.3710 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 195/1000\n","359/359 - 3s - loss: 0.3574 - accuracy: 0.9577 - 3s/epoch - 8ms/step\n","Epoch 196/1000\n","359/359 - 2s - loss: 0.3588 - accuracy: 0.9563 - 2s/epoch - 6ms/step\n","Epoch 197/1000\n","359/359 - 2s - loss: 0.3476 - accuracy: 0.9586 - 2s/epoch - 6ms/step\n","Epoch 198/1000\n","359/359 - 2s - loss: 0.3473 - accuracy: 0.9566 - 2s/epoch - 6ms/step\n","Epoch 199/1000\n","359/359 - 2s - loss: 0.3338 - accuracy: 0.9603 - 2s/epoch - 6ms/step\n","Epoch 200/1000\n","359/359 - 3s - loss: 0.3311 - accuracy: 0.9589 - 3s/epoch - 7ms/step\n","Epoch 201/1000\n","359/359 - 2s - loss: 0.3522 - accuracy: 0.9521 - 2s/epoch - 6ms/step\n","Epoch 202/1000\n","359/359 - 2s - loss: 0.3478 - accuracy: 0.9547 - 2s/epoch - 6ms/step\n","Epoch 203/1000\n","359/359 - 2s - loss: 0.3345 - accuracy: 0.9568 - 2s/epoch - 6ms/step\n","Epoch 204/1000\n","359/359 - 2s - loss: 0.3358 - accuracy: 0.9559 - 2s/epoch - 6ms/step\n","Epoch 205/1000\n","359/359 - 3s - loss: 0.3263 - accuracy: 0.9559 - 3s/epoch - 7ms/step\n","Epoch 206/1000\n","359/359 - 2s - loss: 0.3248 - accuracy: 0.9563 - 2s/epoch - 6ms/step\n","Epoch 207/1000\n","359/359 - 2s - loss: 0.3222 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 208/1000\n","359/359 - 2s - loss: 0.3220 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 209/1000\n","359/359 - 2s - loss: 0.3231 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 210/1000\n","359/359 - 3s - loss: 0.3250 - accuracy: 0.9549 - 3s/epoch - 7ms/step\n","Epoch 211/1000\n","359/359 - 2s - loss: 0.3217 - accuracy: 0.9533 - 2s/epoch - 6ms/step\n","Epoch 212/1000\n","359/359 - 2s - loss: 0.3174 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 213/1000\n","359/359 - 2s - loss: 0.3098 - accuracy: 0.9579 - 2s/epoch - 6ms/step\n","Epoch 214/1000\n","359/359 - 2s - loss: 0.3007 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 215/1000\n","359/359 - 2s - loss: 0.3039 - accuracy: 0.9563 - 2s/epoch - 7ms/step\n","Epoch 216/1000\n","359/359 - 2s - loss: 0.2863 - accuracy: 0.9589 - 2s/epoch - 7ms/step\n","Epoch 217/1000\n","359/359 - 2s - loss: 0.2964 - accuracy: 0.9554 - 2s/epoch - 6ms/step\n","Epoch 218/1000\n","359/359 - 2s - loss: 0.2979 - accuracy: 0.9542 - 2s/epoch - 6ms/step\n","Epoch 219/1000\n","359/359 - 2s - loss: 0.2868 - accuracy: 0.9568 - 2s/epoch - 6ms/step\n","Epoch 220/1000\n","359/359 - 2s - loss: 0.3079 - accuracy: 0.9517 - 2s/epoch - 6ms/step\n","Epoch 221/1000\n","359/359 - 2s - loss: 0.3290 - accuracy: 0.9468 - 2s/epoch - 7ms/step\n","Epoch 222/1000\n","359/359 - 2s - loss: 0.3246 - accuracy: 0.9477 - 2s/epoch - 6ms/step\n","Epoch 223/1000\n","359/359 - 2s - loss: 0.3706 - accuracy: 0.9428 - 2s/epoch - 6ms/step\n","Epoch 224/1000\n","359/359 - 2s - loss: 0.3702 - accuracy: 0.9417 - 2s/epoch - 6ms/step\n","Epoch 225/1000\n","359/359 - 2s - loss: 0.3740 - accuracy: 0.9394 - 2s/epoch - 6ms/step\n","Epoch 226/1000\n","359/359 - 3s - loss: 0.3574 - accuracy: 0.9424 - 3s/epoch - 7ms/step\n","Epoch 227/1000\n","359/359 - 2s - loss: 0.3647 - accuracy: 0.9421 - 2s/epoch - 6ms/step\n","Epoch 228/1000\n","359/359 - 2s - loss: 0.3408 - accuracy: 0.9454 - 2s/epoch - 6ms/step\n","Epoch 229/1000\n","359/359 - 2s - loss: 0.3229 - accuracy: 0.9512 - 2s/epoch - 6ms/step\n","Epoch 230/1000\n","359/359 - 2s - loss: 0.3024 - accuracy: 0.9528 - 2s/epoch - 6ms/step\n","Epoch 231/1000\n","359/359 - 3s - loss: 0.2939 - accuracy: 0.9538 - 3s/epoch - 7ms/step\n","Epoch 232/1000\n","359/359 - 2s - loss: 0.2880 - accuracy: 0.9549 - 2s/epoch - 6ms/step\n","Epoch 233/1000\n","359/359 - 2s - loss: 0.2706 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 234/1000\n","359/359 - 2s - loss: 0.2645 - accuracy: 0.9575 - 2s/epoch - 6ms/step\n","Epoch 235/1000\n","359/359 - 2s - loss: 0.2593 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 236/1000\n","359/359 - 3s - loss: 0.2546 - accuracy: 0.9589 - 3s/epoch - 7ms/step\n","Epoch 237/1000\n","359/359 - 2s - loss: 0.2466 - accuracy: 0.9582 - 2s/epoch - 6ms/step\n","Epoch 238/1000\n","359/359 - 2s - loss: 0.2386 - accuracy: 0.9605 - 2s/epoch - 6ms/step\n","Epoch 239/1000\n","359/359 - 2s - loss: 0.2362 - accuracy: 0.9610 - 2s/epoch - 6ms/step\n","Epoch 240/1000\n","359/359 - 2s - loss: 0.2483 - accuracy: 0.9531 - 2s/epoch - 6ms/step\n","Epoch 241/1000\n","359/359 - 3s - loss: 0.2562 - accuracy: 0.9524 - 3s/epoch - 7ms/step\n","Epoch 242/1000\n","359/359 - 2s - loss: 0.2383 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 243/1000\n","359/359 - 2s - loss: 0.2256 - accuracy: 0.9596 - 2s/epoch - 6ms/step\n","Epoch 244/1000\n","359/359 - 3s - loss: 0.2174 - accuracy: 0.9603 - 3s/epoch - 7ms/step\n","Epoch 245/1000\n","359/359 - 2s - loss: 0.2133 - accuracy: 0.9603 - 2s/epoch - 6ms/step\n","Epoch 246/1000\n","359/359 - 3s - loss: 0.2071 - accuracy: 0.9610 - 3s/epoch - 7ms/step\n","Epoch 247/1000\n","359/359 - 2s - loss: 0.2007 - accuracy: 0.9614 - 2s/epoch - 6ms/step\n","Epoch 248/1000\n","359/359 - 2s - loss: 0.1980 - accuracy: 0.9612 - 2s/epoch - 6ms/step\n","Epoch 249/1000\n","359/359 - 2s - loss: 0.1969 - accuracy: 0.9624 - 2s/epoch - 6ms/step\n","Epoch 250/1000\n","359/359 - 2s - loss: 0.1911 - accuracy: 0.9617 - 2s/epoch - 6ms/step\n","Epoch 251/1000\n","359/359 - 3s - loss: 0.1916 - accuracy: 0.9607 - 3s/epoch - 7ms/step\n","Epoch 252/1000\n","359/359 - 2s - loss: 0.1866 - accuracy: 0.9619 - 2s/epoch - 6ms/step\n","Epoch 253/1000\n","359/359 - 2s - loss: 0.1870 - accuracy: 0.9605 - 2s/epoch - 6ms/step\n","Epoch 254/1000\n","359/359 - 2s - loss: 0.1830 - accuracy: 0.9603 - 2s/epoch - 6ms/step\n","Epoch 255/1000\n","359/359 - 2s - loss: 0.1789 - accuracy: 0.9614 - 2s/epoch - 6ms/step\n","Epoch 256/1000\n","359/359 - 3s - loss: 0.1779 - accuracy: 0.9600 - 3s/epoch - 7ms/step\n","Epoch 257/1000\n","359/359 - 2s - loss: 0.1772 - accuracy: 0.9612 - 2s/epoch - 6ms/step\n","Epoch 258/1000\n","359/359 - 2s - loss: 0.1713 - accuracy: 0.9614 - 2s/epoch - 6ms/step\n","Epoch 259/1000\n","359/359 - 2s - loss: 0.1981 - accuracy: 0.9559 - 2s/epoch - 6ms/step\n","Epoch 260/1000\n","359/359 - 2s - loss: 0.1844 - accuracy: 0.9575 - 2s/epoch - 6ms/step\n","Epoch 261/1000\n","359/359 - 3s - loss: 0.1821 - accuracy: 0.9591 - 3s/epoch - 7ms/step\n","Epoch 262/1000\n","359/359 - 2s - loss: 0.1958 - accuracy: 0.9552 - 2s/epoch - 6ms/step\n","Epoch 263/1000\n","359/359 - 2s - loss: 0.2223 - accuracy: 0.9521 - 2s/epoch - 6ms/step\n","Epoch 264/1000\n","359/359 - 2s - loss: 0.2015 - accuracy: 0.9526 - 2s/epoch - 6ms/step\n","Epoch 265/1000\n","359/359 - 2s - loss: 0.2042 - accuracy: 0.9512 - 2s/epoch - 6ms/step\n","Epoch 266/1000\n","359/359 - 3s - loss: 0.2322 - accuracy: 0.9473 - 3s/epoch - 7ms/step\n","Epoch 267/1000\n","359/359 - 2s - loss: 0.2141 - accuracy: 0.9507 - 2s/epoch - 6ms/step\n","Epoch 268/1000\n","359/359 - 2s - loss: 0.2062 - accuracy: 0.9519 - 2s/epoch - 6ms/step\n","Epoch 269/1000\n","359/359 - 2s - loss: 0.2093 - accuracy: 0.9531 - 2s/epoch - 6ms/step\n","Epoch 270/1000\n","359/359 - 2s - loss: 0.1876 - accuracy: 0.9566 - 2s/epoch - 6ms/step\n","Epoch 271/1000\n","359/359 - 2s - loss: 0.1898 - accuracy: 0.9566 - 2s/epoch - 7ms/step\n","Epoch 272/1000\n","359/359 - 2s - loss: 0.1907 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 273/1000\n","359/359 - 2s - loss: 0.1983 - accuracy: 0.9542 - 2s/epoch - 6ms/step\n","Epoch 274/1000\n","359/359 - 2s - loss: 0.1836 - accuracy: 0.9586 - 2s/epoch - 6ms/step\n","Epoch 275/1000\n","359/359 - 2s - loss: 0.1731 - accuracy: 0.9589 - 2s/epoch - 6ms/step\n","Epoch 276/1000\n","359/359 - 2s - loss: 0.1734 - accuracy: 0.9589 - 2s/epoch - 7ms/step\n","Epoch 277/1000\n","359/359 - 2s - loss: 0.1687 - accuracy: 0.9605 - 2s/epoch - 6ms/step\n","Epoch 278/1000\n","359/359 - 2s - loss: 0.1767 - accuracy: 0.9582 - 2s/epoch - 6ms/step\n","Epoch 279/1000\n","359/359 - 2s - loss: 0.1707 - accuracy: 0.9584 - 2s/epoch - 6ms/step\n","Epoch 280/1000\n","359/359 - 2s - loss: 0.1638 - accuracy: 0.9582 - 2s/epoch - 6ms/step\n","Epoch 281/1000\n","359/359 - 2s - loss: 0.1693 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 282/1000\n","359/359 - 3s - loss: 0.1618 - accuracy: 0.9607 - 3s/epoch - 7ms/step\n","Epoch 283/1000\n","359/359 - 2s - loss: 0.1604 - accuracy: 0.9614 - 2s/epoch - 6ms/step\n","Epoch 284/1000\n","359/359 - 2s - loss: 0.1594 - accuracy: 0.9598 - 2s/epoch - 6ms/step\n","Epoch 285/1000\n","359/359 - 2s - loss: 0.1588 - accuracy: 0.9610 - 2s/epoch - 6ms/step\n","Epoch 286/1000\n","359/359 - 2s - loss: 0.1583 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 287/1000\n","359/359 - 3s - loss: 0.1604 - accuracy: 0.9605 - 3s/epoch - 7ms/step\n","Epoch 288/1000\n","359/359 - 2s - loss: 0.1593 - accuracy: 0.9619 - 2s/epoch - 6ms/step\n","Epoch 289/1000\n","359/359 - 2s - loss: 0.1586 - accuracy: 0.9607 - 2s/epoch - 6ms/step\n","Epoch 290/1000\n","359/359 - 2s - loss: 0.1565 - accuracy: 0.9614 - 2s/epoch - 6ms/step\n","Epoch 291/1000\n","359/359 - 2s - loss: 0.1556 - accuracy: 0.9621 - 2s/epoch - 6ms/step\n","Epoch 292/1000\n","359/359 - 3s - loss: 0.1584 - accuracy: 0.9607 - 3s/epoch - 7ms/step\n","Epoch 293/1000\n","359/359 - 2s - loss: 0.1567 - accuracy: 0.9605 - 2s/epoch - 6ms/step\n","Epoch 294/1000\n","359/359 - 2s - loss: 0.1605 - accuracy: 0.9603 - 2s/epoch - 6ms/step\n","Epoch 295/1000\n","359/359 - 2s - loss: 0.1652 - accuracy: 0.9575 - 2s/epoch - 6ms/step\n","Epoch 296/1000\n","359/359 - 2s - loss: 0.1841 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 297/1000\n","359/359 - 3s - loss: 0.1956 - accuracy: 0.9498 - 3s/epoch - 7ms/step\n","Epoch 298/1000\n","359/359 - 2s - loss: 0.1836 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 299/1000\n","359/359 - 2s - loss: 0.1762 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 300/1000\n","359/359 - 2s - loss: 0.1779 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 301/1000\n","359/359 - 2s - loss: 0.1738 - accuracy: 0.9575 - 2s/epoch - 6ms/step\n","Epoch 302/1000\n","359/359 - 3s - loss: 0.1692 - accuracy: 0.9596 - 3s/epoch - 7ms/step\n","Epoch 303/1000\n","359/359 - 2s - loss: 0.1938 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 304/1000\n","359/359 - 2s - loss: 0.1945 - accuracy: 0.9524 - 2s/epoch - 6ms/step\n","Epoch 305/1000\n","359/359 - 2s - loss: 0.1947 - accuracy: 0.9517 - 2s/epoch - 6ms/step\n","Epoch 306/1000\n","359/359 - 2s - loss: 0.1843 - accuracy: 0.9521 - 2s/epoch - 6ms/step\n","Epoch 307/1000\n","359/359 - 3s - loss: 0.1806 - accuracy: 0.9561 - 3s/epoch - 7ms/step\n","Epoch 308/1000\n","359/359 - 2s - loss: 0.1791 - accuracy: 0.9554 - 2s/epoch - 6ms/step\n","Epoch 309/1000\n","359/359 - 2s - loss: 0.1755 - accuracy: 0.9568 - 2s/epoch - 6ms/step\n","Epoch 310/1000\n","359/359 - 2s - loss: 0.1681 - accuracy: 0.9603 - 2s/epoch - 6ms/step\n","Epoch 311/1000\n","359/359 - 2s - loss: 0.1663 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 311: early stopping\n"]}]},{"cell_type":"code","source":["# Train GRU model\n","print(\"Training GRU model...\")\n","gru_history = gru_model.fit(xs, ys, batch_size=size,\n","                            epochs=epoch, verbose=2, callbacks=[earlystop])"],"metadata":{"id":"LoBLSx_Bj3iA","executionInfo":{"status":"ok","timestamp":1684898657195,"user_tz":-480,"elapsed":422146,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"743dfc6c-657a-47be-b3ff-879d63f970a9"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Training GRU model...\n","Epoch 1/1000\n","359/359 - 17s - loss: 6.1005 - accuracy: 0.0569 - 17s/epoch - 49ms/step\n","Epoch 2/1000\n","359/359 - 3s - loss: 5.7197 - accuracy: 0.0611 - 3s/epoch - 9ms/step\n","Epoch 3/1000\n","359/359 - 3s - loss: 5.5440 - accuracy: 0.0774 - 3s/epoch - 8ms/step\n","Epoch 4/1000\n","359/359 - 3s - loss: 5.3273 - accuracy: 0.1001 - 3s/epoch - 10ms/step\n","Epoch 5/1000\n","359/359 - 3s - loss: 5.0864 - accuracy: 0.1255 - 3s/epoch - 8ms/step\n","Epoch 6/1000\n","359/359 - 3s - loss: 4.8267 - accuracy: 0.1594 - 3s/epoch - 7ms/step\n","Epoch 7/1000\n","359/359 - 2s - loss: 4.5718 - accuracy: 0.2061 - 2s/epoch - 7ms/step\n","Epoch 8/1000\n","359/359 - 2s - loss: 4.3594 - accuracy: 0.2214 - 2s/epoch - 7ms/step\n","Epoch 9/1000\n","359/359 - 3s - loss: 4.1635 - accuracy: 0.2486 - 3s/epoch - 8ms/step\n","Epoch 10/1000\n","359/359 - 2s - loss: 3.9831 - accuracy: 0.2742 - 2s/epoch - 6ms/step\n","Epoch 11/1000\n","359/359 - 3s - loss: 3.8112 - accuracy: 0.3023 - 3s/epoch - 7ms/step\n","Epoch 12/1000\n","359/359 - 2s - loss: 3.6630 - accuracy: 0.3253 - 2s/epoch - 7ms/step\n","Epoch 13/1000\n","359/359 - 2s - loss: 3.5108 - accuracy: 0.3474 - 2s/epoch - 7ms/step\n","Epoch 14/1000\n","359/359 - 3s - loss: 3.3687 - accuracy: 0.3680 - 3s/epoch - 7ms/step\n","Epoch 15/1000\n","359/359 - 2s - loss: 3.1913 - accuracy: 0.3957 - 2s/epoch - 6ms/step\n","Epoch 16/1000\n","359/359 - 2s - loss: 3.0311 - accuracy: 0.4136 - 2s/epoch - 6ms/step\n","Epoch 17/1000\n","359/359 - 2s - loss: 2.8855 - accuracy: 0.4440 - 2s/epoch - 6ms/step\n","Epoch 18/1000\n","359/359 - 2s - loss: 2.7349 - accuracy: 0.4665 - 2s/epoch - 7ms/step\n","Epoch 19/1000\n","359/359 - 2s - loss: 2.5943 - accuracy: 0.4935 - 2s/epoch - 6ms/step\n","Epoch 20/1000\n","359/359 - 2s - loss: 2.4565 - accuracy: 0.5181 - 2s/epoch - 6ms/step\n","Epoch 21/1000\n","359/359 - 2s - loss: 2.3525 - accuracy: 0.5337 - 2s/epoch - 6ms/step\n","Epoch 22/1000\n","359/359 - 2s - loss: 2.2226 - accuracy: 0.5539 - 2s/epoch - 6ms/step\n","Epoch 23/1000\n","359/359 - 2s - loss: 2.1026 - accuracy: 0.5760 - 2s/epoch - 7ms/step\n","Epoch 24/1000\n","359/359 - 2s - loss: 1.9740 - accuracy: 0.5960 - 2s/epoch - 7ms/step\n","Epoch 25/1000\n","359/359 - 2s - loss: 1.8440 - accuracy: 0.6206 - 2s/epoch - 7ms/step\n","Epoch 26/1000\n","359/359 - 2s - loss: 1.7032 - accuracy: 0.6355 - 2s/epoch - 7ms/step\n","Epoch 27/1000\n","359/359 - 2s - loss: 1.5633 - accuracy: 0.6587 - 2s/epoch - 6ms/step\n","Epoch 28/1000\n","359/359 - 3s - loss: 1.4291 - accuracy: 0.6780 - 3s/epoch - 7ms/step\n","Epoch 29/1000\n","359/359 - 2s - loss: 1.3080 - accuracy: 0.6991 - 2s/epoch - 7ms/step\n","Epoch 30/1000\n","359/359 - 2s - loss: 1.1747 - accuracy: 0.7210 - 2s/epoch - 6ms/step\n","Epoch 31/1000\n","359/359 - 2s - loss: 1.0627 - accuracy: 0.7400 - 2s/epoch - 6ms/step\n","Epoch 32/1000\n","359/359 - 2s - loss: 0.9485 - accuracy: 0.7598 - 2s/epoch - 6ms/step\n","Epoch 33/1000\n","359/359 - 2s - loss: 0.8380 - accuracy: 0.7837 - 2s/epoch - 7ms/step\n","Epoch 34/1000\n","359/359 - 3s - loss: 0.7595 - accuracy: 0.8039 - 3s/epoch - 7ms/step\n","Epoch 35/1000\n","359/359 - 2s - loss: 0.6827 - accuracy: 0.8178 - 2s/epoch - 6ms/step\n","Epoch 36/1000\n","359/359 - 2s - loss: 0.6213 - accuracy: 0.8360 - 2s/epoch - 7ms/step\n","Epoch 37/1000\n","359/359 - 2s - loss: 0.5631 - accuracy: 0.8525 - 2s/epoch - 6ms/step\n","Epoch 38/1000\n","359/359 - 3s - loss: 0.5223 - accuracy: 0.8608 - 3s/epoch - 7ms/step\n","Epoch 39/1000\n","359/359 - 2s - loss: 0.4707 - accuracy: 0.8801 - 2s/epoch - 6ms/step\n","Epoch 40/1000\n","359/359 - 2s - loss: 0.4243 - accuracy: 0.8885 - 2s/epoch - 6ms/step\n","Epoch 41/1000\n","359/359 - 2s - loss: 0.3952 - accuracy: 0.9013 - 2s/epoch - 6ms/step\n","Epoch 42/1000\n","359/359 - 2s - loss: 0.3515 - accuracy: 0.9124 - 2s/epoch - 6ms/step\n","Epoch 43/1000\n","359/359 - 3s - loss: 0.3343 - accuracy: 0.9198 - 3s/epoch - 7ms/step\n","Epoch 44/1000\n","359/359 - 2s - loss: 0.3152 - accuracy: 0.9280 - 2s/epoch - 6ms/step\n","Epoch 45/1000\n","359/359 - 2s - loss: 0.2909 - accuracy: 0.9303 - 2s/epoch - 6ms/step\n","Epoch 46/1000\n","359/359 - 2s - loss: 0.2770 - accuracy: 0.9349 - 2s/epoch - 6ms/step\n","Epoch 47/1000\n","359/359 - 2s - loss: 0.2638 - accuracy: 0.9373 - 2s/epoch - 6ms/step\n","Epoch 48/1000\n","359/359 - 3s - loss: 0.2625 - accuracy: 0.9396 - 3s/epoch - 8ms/step\n","Epoch 49/1000\n","359/359 - 2s - loss: 0.2502 - accuracy: 0.9408 - 2s/epoch - 6ms/step\n","Epoch 50/1000\n","359/359 - 2s - loss: 0.2498 - accuracy: 0.9414 - 2s/epoch - 6ms/step\n","Epoch 51/1000\n","359/359 - 2s - loss: 0.2404 - accuracy: 0.9440 - 2s/epoch - 6ms/step\n","Epoch 52/1000\n","359/359 - 2s - loss: 0.2289 - accuracy: 0.9447 - 2s/epoch - 6ms/step\n","Epoch 53/1000\n","359/359 - 3s - loss: 0.2312 - accuracy: 0.9480 - 3s/epoch - 7ms/step\n","Epoch 54/1000\n","359/359 - 2s - loss: 0.2203 - accuracy: 0.9482 - 2s/epoch - 6ms/step\n","Epoch 55/1000\n","359/359 - 2s - loss: 0.2094 - accuracy: 0.9514 - 2s/epoch - 6ms/step\n","Epoch 56/1000\n","359/359 - 2s - loss: 0.2193 - accuracy: 0.9514 - 2s/epoch - 6ms/step\n","Epoch 57/1000\n","359/359 - 2s - loss: 0.2081 - accuracy: 0.9510 - 2s/epoch - 6ms/step\n","Epoch 58/1000\n","359/359 - 2s - loss: 0.2101 - accuracy: 0.9524 - 2s/epoch - 7ms/step\n","Epoch 59/1000\n","359/359 - 2s - loss: 0.2106 - accuracy: 0.9505 - 2s/epoch - 7ms/step\n","Epoch 60/1000\n","359/359 - 2s - loss: 0.2083 - accuracy: 0.9507 - 2s/epoch - 6ms/step\n","Epoch 61/1000\n","359/359 - 2s - loss: 0.2082 - accuracy: 0.9500 - 2s/epoch - 6ms/step\n","Epoch 62/1000\n","359/359 - 2s - loss: 0.2064 - accuracy: 0.9517 - 2s/epoch - 6ms/step\n","Epoch 63/1000\n","359/359 - 2s - loss: 0.2008 - accuracy: 0.9535 - 2s/epoch - 7ms/step\n","Epoch 64/1000\n","359/359 - 2s - loss: 0.1974 - accuracy: 0.9519 - 2s/epoch - 7ms/step\n","Epoch 65/1000\n","359/359 - 2s - loss: 0.1991 - accuracy: 0.9535 - 2s/epoch - 7ms/step\n","Epoch 66/1000\n","359/359 - 2s - loss: 0.1987 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 67/1000\n","359/359 - 2s - loss: 0.1906 - accuracy: 0.9533 - 2s/epoch - 6ms/step\n","Epoch 68/1000\n","359/359 - 2s - loss: 0.1883 - accuracy: 0.9545 - 2s/epoch - 7ms/step\n","Epoch 69/1000\n","359/359 - 2s - loss: 0.1909 - accuracy: 0.9547 - 2s/epoch - 7ms/step\n","Epoch 70/1000\n","359/359 - 2s - loss: 0.1908 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 71/1000\n","359/359 - 2s - loss: 0.1883 - accuracy: 0.9549 - 2s/epoch - 6ms/step\n","Epoch 72/1000\n","359/359 - 2s - loss: 0.1889 - accuracy: 0.9542 - 2s/epoch - 6ms/step\n","Epoch 73/1000\n","359/359 - 2s - loss: 0.1884 - accuracy: 0.9545 - 2s/epoch - 7ms/step\n","Epoch 74/1000\n","359/359 - 2s - loss: 0.1859 - accuracy: 0.9561 - 2s/epoch - 7ms/step\n","Epoch 75/1000\n","359/359 - 2s - loss: 0.1827 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 76/1000\n","359/359 - 2s - loss: 0.1853 - accuracy: 0.9535 - 2s/epoch - 6ms/step\n","Epoch 77/1000\n","359/359 - 2s - loss: 0.1845 - accuracy: 0.9531 - 2s/epoch - 6ms/step\n","Epoch 78/1000\n","359/359 - 2s - loss: 0.1801 - accuracy: 0.9547 - 2s/epoch - 6ms/step\n","Epoch 79/1000\n","359/359 - 2s - loss: 0.1849 - accuracy: 0.9514 - 2s/epoch - 7ms/step\n","Epoch 80/1000\n","359/359 - 2s - loss: 0.1810 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 81/1000\n","359/359 - 2s - loss: 0.1767 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 82/1000\n","359/359 - 2s - loss: 0.1801 - accuracy: 0.9542 - 2s/epoch - 6ms/step\n","Epoch 83/1000\n","359/359 - 2s - loss: 0.1775 - accuracy: 0.9556 - 2s/epoch - 6ms/step\n","Epoch 84/1000\n","359/359 - 3s - loss: 0.1751 - accuracy: 0.9561 - 3s/epoch - 7ms/step\n","Epoch 85/1000\n","359/359 - 2s - loss: 0.1793 - accuracy: 0.9568 - 2s/epoch - 6ms/step\n","Epoch 86/1000\n","359/359 - 2s - loss: 0.1824 - accuracy: 0.9540 - 2s/epoch - 6ms/step\n","Epoch 87/1000\n","359/359 - 2s - loss: 0.1798 - accuracy: 0.9528 - 2s/epoch - 6ms/step\n","Epoch 88/1000\n","359/359 - 2s - loss: 0.1710 - accuracy: 0.9556 - 2s/epoch - 6ms/step\n","Epoch 89/1000\n","359/359 - 3s - loss: 0.1778 - accuracy: 0.9561 - 3s/epoch - 7ms/step\n","Epoch 90/1000\n","359/359 - 2s - loss: 0.1784 - accuracy: 0.9535 - 2s/epoch - 7ms/step\n","Epoch 91/1000\n","359/359 - 2s - loss: 0.1743 - accuracy: 0.9559 - 2s/epoch - 7ms/step\n","Epoch 92/1000\n","359/359 - 2s - loss: 0.1718 - accuracy: 0.9547 - 2s/epoch - 6ms/step\n","Epoch 93/1000\n","359/359 - 2s - loss: 0.1727 - accuracy: 0.9563 - 2s/epoch - 6ms/step\n","Epoch 94/1000\n","359/359 - 3s - loss: 0.1662 - accuracy: 0.9566 - 3s/epoch - 7ms/step\n","Epoch 95/1000\n","359/359 - 2s - loss: 0.1734 - accuracy: 0.9566 - 2s/epoch - 6ms/step\n","Epoch 96/1000\n","359/359 - 2s - loss: 0.1669 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 97/1000\n","359/359 - 2s - loss: 0.1719 - accuracy: 0.9538 - 2s/epoch - 6ms/step\n","Epoch 98/1000\n","359/359 - 2s - loss: 0.1731 - accuracy: 0.9559 - 2s/epoch - 6ms/step\n","Epoch 99/1000\n","359/359 - 3s - loss: 0.1759 - accuracy: 0.9552 - 3s/epoch - 7ms/step\n","Epoch 100/1000\n","359/359 - 2s - loss: 0.1709 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 101/1000\n","359/359 - 2s - loss: 0.1675 - accuracy: 0.9566 - 2s/epoch - 7ms/step\n","Epoch 102/1000\n","359/359 - 2s - loss: 0.1706 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 103/1000\n","359/359 - 2s - loss: 0.1700 - accuracy: 0.9554 - 2s/epoch - 6ms/step\n","Epoch 104/1000\n","359/359 - 3s - loss: 0.1652 - accuracy: 0.9563 - 3s/epoch - 7ms/step\n","Epoch 105/1000\n","359/359 - 2s - loss: 0.1668 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 106/1000\n","359/359 - 2s - loss: 0.1706 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 107/1000\n","359/359 - 2s - loss: 0.1667 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 108/1000\n","359/359 - 2s - loss: 0.1674 - accuracy: 0.9582 - 2s/epoch - 6ms/step\n","Epoch 109/1000\n","359/359 - 3s - loss: 0.1614 - accuracy: 0.9582 - 3s/epoch - 7ms/step\n","Epoch 110/1000\n","359/359 - 2s - loss: 0.1666 - accuracy: 0.9556 - 2s/epoch - 6ms/step\n","Epoch 111/1000\n","359/359 - 2s - loss: 0.1682 - accuracy: 0.9563 - 2s/epoch - 6ms/step\n","Epoch 112/1000\n","359/359 - 2s - loss: 0.1657 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 113/1000\n","359/359 - 2s - loss: 0.1613 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 114/1000\n","359/359 - 3s - loss: 0.1683 - accuracy: 0.9568 - 3s/epoch - 7ms/step\n","Epoch 115/1000\n","359/359 - 2s - loss: 0.1682 - accuracy: 0.9538 - 2s/epoch - 6ms/step\n","Epoch 116/1000\n","359/359 - 2s - loss: 0.1637 - accuracy: 0.9568 - 2s/epoch - 6ms/step\n","Epoch 117/1000\n","359/359 - 2s - loss: 0.1643 - accuracy: 0.9579 - 2s/epoch - 6ms/step\n","Epoch 118/1000\n","359/359 - 2s - loss: 0.1672 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 119/1000\n","359/359 - 3s - loss: 0.1654 - accuracy: 0.9579 - 3s/epoch - 7ms/step\n","Epoch 120/1000\n","359/359 - 2s - loss: 0.1695 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 121/1000\n","359/359 - 2s - loss: 0.1648 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 122/1000\n","359/359 - 2s - loss: 0.1632 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 123/1000\n","359/359 - 2s - loss: 0.1578 - accuracy: 0.9559 - 2s/epoch - 6ms/step\n","Epoch 124/1000\n","359/359 - 3s - loss: 0.1621 - accuracy: 0.9568 - 3s/epoch - 7ms/step\n","Epoch 125/1000\n","359/359 - 2s - loss: 0.1633 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 126/1000\n","359/359 - 2s - loss: 0.1714 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 127/1000\n","359/359 - 2s - loss: 0.1656 - accuracy: 0.9566 - 2s/epoch - 6ms/step\n","Epoch 128/1000\n","359/359 - 2s - loss: 0.1630 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 129/1000\n","359/359 - 3s - loss: 0.1614 - accuracy: 0.9582 - 3s/epoch - 7ms/step\n","Epoch 130/1000\n","359/359 - 2s - loss: 0.1617 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 131/1000\n","359/359 - 2s - loss: 0.1593 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 132/1000\n","359/359 - 2s - loss: 0.1633 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 133/1000\n","359/359 - 2s - loss: 0.1597 - accuracy: 0.9586 - 2s/epoch - 6ms/step\n","Epoch 134/1000\n","359/359 - 2s - loss: 0.1588 - accuracy: 0.9552 - 2s/epoch - 7ms/step\n","Epoch 135/1000\n","359/359 - 2s - loss: 0.1605 - accuracy: 0.9563 - 2s/epoch - 6ms/step\n","Epoch 136/1000\n","359/359 - 2s - loss: 0.1592 - accuracy: 0.9579 - 2s/epoch - 6ms/step\n","Epoch 137/1000\n","359/359 - 2s - loss: 0.1618 - accuracy: 0.9556 - 2s/epoch - 6ms/step\n","Epoch 138/1000\n","359/359 - 2s - loss: 0.1619 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 139/1000\n","359/359 - 2s - loss: 0.1724 - accuracy: 0.9549 - 2s/epoch - 7ms/step\n","Epoch 140/1000\n","359/359 - 2s - loss: 0.1593 - accuracy: 0.9570 - 2s/epoch - 7ms/step\n","Epoch 141/1000\n","359/359 - 2s - loss: 0.1554 - accuracy: 0.9579 - 2s/epoch - 6ms/step\n","Epoch 142/1000\n","359/359 - 2s - loss: 0.1585 - accuracy: 0.9568 - 2s/epoch - 6ms/step\n","Epoch 143/1000\n","359/359 - 2s - loss: 0.1538 - accuracy: 0.9593 - 2s/epoch - 6ms/step\n","Epoch 144/1000\n","359/359 - 2s - loss: 0.1583 - accuracy: 0.9589 - 2s/epoch - 6ms/step\n","Epoch 145/1000\n","359/359 - 2s - loss: 0.1585 - accuracy: 0.9584 - 2s/epoch - 7ms/step\n","Epoch 146/1000\n","359/359 - 2s - loss: 0.1597 - accuracy: 0.9566 - 2s/epoch - 6ms/step\n","Epoch 147/1000\n","359/359 - 2s - loss: 0.1548 - accuracy: 0.9575 - 2s/epoch - 6ms/step\n","Epoch 148/1000\n","359/359 - 2s - loss: 0.1563 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 149/1000\n","359/359 - 2s - loss: 0.1560 - accuracy: 0.9605 - 2s/epoch - 6ms/step\n","Epoch 150/1000\n","359/359 - 2s - loss: 0.1587 - accuracy: 0.9577 - 2s/epoch - 7ms/step\n","Epoch 151/1000\n","359/359 - 2s - loss: 0.1566 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 152/1000\n","359/359 - 2s - loss: 0.1610 - accuracy: 0.9591 - 2s/epoch - 6ms/step\n","Epoch 153/1000\n","359/359 - 2s - loss: 0.1573 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 154/1000\n","359/359 - 2s - loss: 0.1575 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 155/1000\n","359/359 - 3s - loss: 0.1567 - accuracy: 0.9572 - 3s/epoch - 7ms/step\n","Epoch 156/1000\n","359/359 - 2s - loss: 0.1563 - accuracy: 0.9589 - 2s/epoch - 6ms/step\n","Epoch 157/1000\n","359/359 - 2s - loss: 0.1531 - accuracy: 0.9596 - 2s/epoch - 6ms/step\n","Epoch 158/1000\n","359/359 - 2s - loss: 0.1537 - accuracy: 0.9572 - 2s/epoch - 6ms/step\n","Epoch 159/1000\n","359/359 - 2s - loss: 0.1631 - accuracy: 0.9561 - 2s/epoch - 6ms/step\n","Epoch 160/1000\n","359/359 - 3s - loss: 0.1626 - accuracy: 0.9526 - 3s/epoch - 7ms/step\n","Epoch 161/1000\n","359/359 - 2s - loss: 0.1637 - accuracy: 0.9542 - 2s/epoch - 6ms/step\n","Epoch 162/1000\n","359/359 - 2s - loss: 0.1644 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 163/1000\n","359/359 - 2s - loss: 0.1671 - accuracy: 0.9549 - 2s/epoch - 6ms/step\n","Epoch 164/1000\n","359/359 - 2s - loss: 0.1622 - accuracy: 0.9577 - 2s/epoch - 6ms/step\n","Epoch 165/1000\n","359/359 - 3s - loss: 0.1560 - accuracy: 0.9538 - 3s/epoch - 7ms/step\n","Epoch 166/1000\n","359/359 - 2s - loss: 0.1591 - accuracy: 0.9547 - 2s/epoch - 6ms/step\n","Epoch 167/1000\n","359/359 - 2s - loss: 0.1629 - accuracy: 0.9579 - 2s/epoch - 6ms/step\n","Epoch 168/1000\n","359/359 - 2s - loss: 0.1564 - accuracy: 0.9547 - 2s/epoch - 6ms/step\n","Epoch 169/1000\n","359/359 - 2s - loss: 0.1566 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 170/1000\n","359/359 - 3s - loss: 0.1569 - accuracy: 0.9552 - 3s/epoch - 7ms/step\n","Epoch 171/1000\n","359/359 - 2s - loss: 0.1667 - accuracy: 0.9554 - 2s/epoch - 6ms/step\n","Epoch 172/1000\n","359/359 - 2s - loss: 0.1755 - accuracy: 0.9531 - 2s/epoch - 6ms/step\n","Epoch 173/1000\n","359/359 - 2s - loss: 0.1591 - accuracy: 0.9554 - 2s/epoch - 6ms/step\n","Epoch 174/1000\n","359/359 - 2s - loss: 0.1633 - accuracy: 0.9559 - 2s/epoch - 6ms/step\n","Epoch 175/1000\n","359/359 - 3s - loss: 0.1618 - accuracy: 0.9575 - 3s/epoch - 7ms/step\n","Epoch 176/1000\n","359/359 - 2s - loss: 0.1587 - accuracy: 0.9570 - 2s/epoch - 6ms/step\n","Epoch 177/1000\n","359/359 - 2s - loss: 0.1587 - accuracy: 0.9586 - 2s/epoch - 6ms/step\n","Epoch 177: early stopping\n"]}]},{"cell_type":"code","source":["# Train Bidirectional LSTM model\n","print(\"Training Bidirectional LSTM model...\")\n","bilstm_history = bilstm_model.fit(xs, ys, batch_size = size,\n","                                  epochs=epoch, verbose=2, callbacks=[earlystop])"],"metadata":{"id":"muQ0QD-nj6_I","executionInfo":{"status":"ok","timestamp":1684899819841,"user_tz":-480,"elapsed":713200,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"575ed3c2-d1fe-4663-a2fb-64ed8700f362"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Training Bidirectional LSTM model...\n","Epoch 1/1000\n","359/359 - 20s - loss: 6.0609 - accuracy: 0.0565 - 20s/epoch - 56ms/step\n","Epoch 2/1000\n","359/359 - 5s - loss: 5.7669 - accuracy: 0.0602 - 5s/epoch - 15ms/step\n","Epoch 3/1000\n","359/359 - 5s - loss: 5.6473 - accuracy: 0.0602 - 5s/epoch - 14ms/step\n","Epoch 4/1000\n","359/359 - 4s - loss: 5.5490 - accuracy: 0.0697 - 4s/epoch - 10ms/step\n","Epoch 5/1000\n","359/359 - 4s - loss: 5.4473 - accuracy: 0.0769 - 4s/epoch - 12ms/step\n","Epoch 6/1000\n","359/359 - 5s - loss: 5.3354 - accuracy: 0.0864 - 5s/epoch - 13ms/step\n","Epoch 7/1000\n","359/359 - 3s - loss: 5.1988 - accuracy: 0.0969 - 3s/epoch - 10ms/step\n","Epoch 8/1000\n","359/359 - 3s - loss: 5.0527 - accuracy: 0.1108 - 3s/epoch - 9ms/step\n","Epoch 9/1000\n","359/359 - 4s - loss: 4.8982 - accuracy: 0.1313 - 4s/epoch - 11ms/step\n","Epoch 10/1000\n","359/359 - 4s - loss: 4.7567 - accuracy: 0.1494 - 4s/epoch - 10ms/step\n","Epoch 11/1000\n","359/359 - 4s - loss: 4.5988 - accuracy: 0.1740 - 4s/epoch - 11ms/step\n","Epoch 12/1000\n","359/359 - 4s - loss: 4.4325 - accuracy: 0.2066 - 4s/epoch - 11ms/step\n","Epoch 13/1000\n","359/359 - 4s - loss: 4.2793 - accuracy: 0.2212 - 4s/epoch - 10ms/step\n","Epoch 14/1000\n","359/359 - 4s - loss: 4.0996 - accuracy: 0.2500 - 4s/epoch - 10ms/step\n","Epoch 15/1000\n","359/359 - 4s - loss: 3.9295 - accuracy: 0.2783 - 4s/epoch - 11ms/step\n","Epoch 16/1000\n","359/359 - 3s - loss: 3.7065 - accuracy: 0.3032 - 3s/epoch - 10ms/step\n","Epoch 17/1000\n","359/359 - 4s - loss: 3.5576 - accuracy: 0.3264 - 4s/epoch - 10ms/step\n","Epoch 18/1000\n","359/359 - 4s - loss: 3.3819 - accuracy: 0.3513 - 4s/epoch - 11ms/step\n","Epoch 19/1000\n","359/359 - 4s - loss: 3.2725 - accuracy: 0.3810 - 4s/epoch - 10ms/step\n","Epoch 20/1000\n","359/359 - 4s - loss: 3.1591 - accuracy: 0.4017 - 4s/epoch - 10ms/step\n","Epoch 21/1000\n","359/359 - 4s - loss: 2.9856 - accuracy: 0.4247 - 4s/epoch - 11ms/step\n","Epoch 22/1000\n","359/359 - 3s - loss: 2.8513 - accuracy: 0.4524 - 3s/epoch - 9ms/step\n","Epoch 23/1000\n","359/359 - 3s - loss: 2.6912 - accuracy: 0.4800 - 3s/epoch - 10ms/step\n","Epoch 24/1000\n","359/359 - 4s - loss: 2.5462 - accuracy: 0.4988 - 4s/epoch - 11ms/step\n","Epoch 25/1000\n","359/359 - 4s - loss: 2.4234 - accuracy: 0.5160 - 4s/epoch - 10ms/step\n","Epoch 26/1000\n","359/359 - 3s - loss: 2.3386 - accuracy: 0.5376 - 3s/epoch - 10ms/step\n","Epoch 27/1000\n","359/359 - 3s - loss: 2.1931 - accuracy: 0.5537 - 3s/epoch - 10ms/step\n","Epoch 28/1000\n","359/359 - 4s - loss: 2.0117 - accuracy: 0.5774 - 4s/epoch - 11ms/step\n","Epoch 29/1000\n","359/359 - 3s - loss: 1.9174 - accuracy: 0.5967 - 3s/epoch - 10ms/step\n","Epoch 30/1000\n","359/359 - 4s - loss: 1.8649 - accuracy: 0.6194 - 4s/epoch - 10ms/step\n","Epoch 31/1000\n","359/359 - 4s - loss: 1.7773 - accuracy: 0.6373 - 4s/epoch - 11ms/step\n","Epoch 32/1000\n","359/359 - 3s - loss: 1.7368 - accuracy: 0.6452 - 3s/epoch - 10ms/step\n","Epoch 33/1000\n","359/359 - 3s - loss: 2.3545 - accuracy: 0.5504 - 3s/epoch - 9ms/step\n","Epoch 34/1000\n","359/359 - 4s - loss: 1.9226 - accuracy: 0.6011 - 4s/epoch - 11ms/step\n","Epoch 35/1000\n","359/359 - 4s - loss: 1.5831 - accuracy: 0.6689 - 4s/epoch - 10ms/step\n","Epoch 36/1000\n","359/359 - 3s - loss: 1.4789 - accuracy: 0.6836 - 3s/epoch - 9ms/step\n","Epoch 37/1000\n","359/359 - 4s - loss: 1.3289 - accuracy: 0.7100 - 4s/epoch - 11ms/step\n","Epoch 38/1000\n","359/359 - 3s - loss: 1.2080 - accuracy: 0.7191 - 3s/epoch - 10ms/step\n","Epoch 39/1000\n","359/359 - 4s - loss: 1.0690 - accuracy: 0.7423 - 4s/epoch - 10ms/step\n","Epoch 40/1000\n","359/359 - 4s - loss: 0.9161 - accuracy: 0.7681 - 4s/epoch - 11ms/step\n","Epoch 41/1000\n","359/359 - 3s - loss: 0.8021 - accuracy: 0.7909 - 3s/epoch - 9ms/step\n","Epoch 42/1000\n","359/359 - 3s - loss: 0.6906 - accuracy: 0.8141 - 3s/epoch - 10ms/step\n","Epoch 43/1000\n","359/359 - 4s - loss: 0.5997 - accuracy: 0.8381 - 4s/epoch - 10ms/step\n","Epoch 44/1000\n","359/359 - 4s - loss: 0.5328 - accuracy: 0.8641 - 4s/epoch - 10ms/step\n","Epoch 45/1000\n","359/359 - 3s - loss: 0.4711 - accuracy: 0.8778 - 3s/epoch - 10ms/step\n","Epoch 46/1000\n","359/359 - 3s - loss: 0.4342 - accuracy: 0.8880 - 3s/epoch - 10ms/step\n","Epoch 47/1000\n","359/359 - 4s - loss: 0.3923 - accuracy: 0.9075 - 4s/epoch - 11ms/step\n","Epoch 48/1000\n","359/359 - 3s - loss: 0.3732 - accuracy: 0.9096 - 3s/epoch - 10ms/step\n","Epoch 49/1000\n","359/359 - 4s - loss: 0.3398 - accuracy: 0.9175 - 4s/epoch - 10ms/step\n","Epoch 50/1000\n","359/359 - 4s - loss: 0.3171 - accuracy: 0.9273 - 4s/epoch - 11ms/step\n","Epoch 51/1000\n","359/359 - 3s - loss: 0.2974 - accuracy: 0.9331 - 3s/epoch - 10ms/step\n","Epoch 52/1000\n","359/359 - 3s - loss: 0.2849 - accuracy: 0.9377 - 3s/epoch - 9ms/step\n","Epoch 53/1000\n","359/359 - 4s - loss: 0.2789 - accuracy: 0.9359 - 4s/epoch - 11ms/step\n","Epoch 54/1000\n","359/359 - 3s - loss: 0.2700 - accuracy: 0.9391 - 3s/epoch - 10ms/step\n","Epoch 55/1000\n","359/359 - 3s - loss: 0.2592 - accuracy: 0.9426 - 3s/epoch - 10ms/step\n","Epoch 56/1000\n","359/359 - 4s - loss: 0.2490 - accuracy: 0.9433 - 4s/epoch - 11ms/step\n","Epoch 57/1000\n","359/359 - 3s - loss: 0.2483 - accuracy: 0.9484 - 3s/epoch - 10ms/step\n","Epoch 58/1000\n","359/359 - 3s - loss: 0.2398 - accuracy: 0.9456 - 3s/epoch - 10ms/step\n","Epoch 59/1000\n","359/359 - 4s - loss: 0.2388 - accuracy: 0.9480 - 4s/epoch - 10ms/step\n","Epoch 60/1000\n","359/359 - 4s - loss: 0.2343 - accuracy: 0.9505 - 4s/epoch - 10ms/step\n","Epoch 61/1000\n","359/359 - 3s - loss: 0.2268 - accuracy: 0.9482 - 3s/epoch - 9ms/step\n","Epoch 62/1000\n","359/359 - 3s - loss: 0.2323 - accuracy: 0.9480 - 3s/epoch - 10ms/step\n","Epoch 63/1000\n","359/359 - 4s - loss: 0.2214 - accuracy: 0.9503 - 4s/epoch - 11ms/step\n","Epoch 64/1000\n","359/359 - 3s - loss: 0.2182 - accuracy: 0.9512 - 3s/epoch - 9ms/step\n","Epoch 65/1000\n","359/359 - 4s - loss: 0.2129 - accuracy: 0.9535 - 4s/epoch - 11ms/step\n","Epoch 66/1000\n","359/359 - 4s - loss: 0.2175 - accuracy: 0.9517 - 4s/epoch - 11ms/step\n","Epoch 67/1000\n","359/359 - 3s - loss: 0.2155 - accuracy: 0.9500 - 3s/epoch - 9ms/step\n","Epoch 68/1000\n","359/359 - 3s - loss: 0.2080 - accuracy: 0.9503 - 3s/epoch - 9ms/step\n","Epoch 69/1000\n","359/359 - 4s - loss: 0.2078 - accuracy: 0.9507 - 4s/epoch - 11ms/step\n","Epoch 70/1000\n","359/359 - 3s - loss: 0.2093 - accuracy: 0.9521 - 3s/epoch - 10ms/step\n","Epoch 71/1000\n","359/359 - 3s - loss: 0.2051 - accuracy: 0.9514 - 3s/epoch - 9ms/step\n","Epoch 72/1000\n","359/359 - 4s - loss: 0.2054 - accuracy: 0.9512 - 4s/epoch - 11ms/step\n","Epoch 73/1000\n","359/359 - 3s - loss: 0.2026 - accuracy: 0.9519 - 3s/epoch - 10ms/step\n","Epoch 74/1000\n","359/359 - 3s - loss: 0.2076 - accuracy: 0.9524 - 3s/epoch - 10ms/step\n","Epoch 75/1000\n","359/359 - 4s - loss: 0.2068 - accuracy: 0.9510 - 4s/epoch - 11ms/step\n","Epoch 76/1000\n","359/359 - 3s - loss: 0.2056 - accuracy: 0.9514 - 3s/epoch - 10ms/step\n","Epoch 77/1000\n","359/359 - 4s - loss: 0.1999 - accuracy: 0.9505 - 4s/epoch - 10ms/step\n","Epoch 78/1000\n","359/359 - 4s - loss: 0.2017 - accuracy: 0.9524 - 4s/epoch - 11ms/step\n","Epoch 79/1000\n","359/359 - 4s - loss: 0.2081 - accuracy: 0.9498 - 4s/epoch - 10ms/step\n","Epoch 80/1000\n","359/359 - 4s - loss: 0.2006 - accuracy: 0.9517 - 4s/epoch - 10ms/step\n","Epoch 81/1000\n","359/359 - 4s - loss: 0.2041 - accuracy: 0.9498 - 4s/epoch - 10ms/step\n","Epoch 82/1000\n","359/359 - 4s - loss: 0.1963 - accuracy: 0.9500 - 4s/epoch - 11ms/step\n","Epoch 83/1000\n","359/359 - 3s - loss: 0.1934 - accuracy: 0.9538 - 3s/epoch - 10ms/step\n","Epoch 84/1000\n","359/359 - 3s - loss: 0.1900 - accuracy: 0.9542 - 3s/epoch - 10ms/step\n","Epoch 85/1000\n","359/359 - 4s - loss: 0.1961 - accuracy: 0.9531 - 4s/epoch - 11ms/step\n","Epoch 86/1000\n","359/359 - 4s - loss: 0.1910 - accuracy: 0.9533 - 4s/epoch - 10ms/step\n","Epoch 87/1000\n","359/359 - 3s - loss: 0.1968 - accuracy: 0.9514 - 3s/epoch - 9ms/step\n","Epoch 88/1000\n","359/359 - 4s - loss: 0.1854 - accuracy: 0.9533 - 4s/epoch - 11ms/step\n","Epoch 89/1000\n","359/359 - 3s - loss: 0.1901 - accuracy: 0.9498 - 3s/epoch - 9ms/step\n","Epoch 90/1000\n","359/359 - 3s - loss: 0.1852 - accuracy: 0.9540 - 3s/epoch - 9ms/step\n","Epoch 91/1000\n","359/359 - 4s - loss: 0.1902 - accuracy: 0.9545 - 4s/epoch - 11ms/step\n","Epoch 92/1000\n","359/359 - 3s - loss: 0.1857 - accuracy: 0.9533 - 3s/epoch - 9ms/step\n","Epoch 93/1000\n","359/359 - 3s - loss: 0.1924 - accuracy: 0.9524 - 3s/epoch - 9ms/step\n","Epoch 94/1000\n","359/359 - 4s - loss: 0.1788 - accuracy: 0.9556 - 4s/epoch - 10ms/step\n","Epoch 95/1000\n","359/359 - 4s - loss: 0.1793 - accuracy: 0.9542 - 4s/epoch - 10ms/step\n","Epoch 96/1000\n","359/359 - 3s - loss: 0.1835 - accuracy: 0.9540 - 3s/epoch - 9ms/step\n","Epoch 97/1000\n","359/359 - 4s - loss: 0.1753 - accuracy: 0.9540 - 4s/epoch - 10ms/step\n","Epoch 98/1000\n","359/359 - 4s - loss: 0.1834 - accuracy: 0.9535 - 4s/epoch - 11ms/step\n","Epoch 99/1000\n","359/359 - 3s - loss: 0.1813 - accuracy: 0.9542 - 3s/epoch - 9ms/step\n","Epoch 100/1000\n","359/359 - 3s - loss: 0.1792 - accuracy: 0.9540 - 3s/epoch - 9ms/step\n","Epoch 101/1000\n","359/359 - 4s - loss: 0.1755 - accuracy: 0.9538 - 4s/epoch - 11ms/step\n","Epoch 102/1000\n","359/359 - 3s - loss: 0.1726 - accuracy: 0.9535 - 3s/epoch - 10ms/step\n","Epoch 103/1000\n","359/359 - 3s - loss: 0.1686 - accuracy: 0.9559 - 3s/epoch - 10ms/step\n","Epoch 104/1000\n","359/359 - 4s - loss: 0.1679 - accuracy: 0.9528 - 4s/epoch - 11ms/step\n","Epoch 105/1000\n","359/359 - 3s - loss: 0.1725 - accuracy: 0.9561 - 3s/epoch - 10ms/step\n","Epoch 106/1000\n","359/359 - 4s - loss: 0.1702 - accuracy: 0.9538 - 4s/epoch - 10ms/step\n","Epoch 107/1000\n","359/359 - 4s - loss: 0.1751 - accuracy: 0.9535 - 4s/epoch - 11ms/step\n","Epoch 108/1000\n","359/359 - 3s - loss: 0.1705 - accuracy: 0.9547 - 3s/epoch - 9ms/step\n","Epoch 109/1000\n","359/359 - 3s - loss: 0.1737 - accuracy: 0.9549 - 3s/epoch - 9ms/step\n","Epoch 110/1000\n","359/359 - 4s - loss: 0.1658 - accuracy: 0.9568 - 4s/epoch - 10ms/step\n","Epoch 111/1000\n","359/359 - 4s - loss: 0.1719 - accuracy: 0.9545 - 4s/epoch - 10ms/step\n","Epoch 112/1000\n","359/359 - 3s - loss: 0.1681 - accuracy: 0.9554 - 3s/epoch - 10ms/step\n","Epoch 113/1000\n","359/359 - 3s - loss: 0.1708 - accuracy: 0.9561 - 3s/epoch - 9ms/step\n","Epoch 114/1000\n","359/359 - 4s - loss: 0.1705 - accuracy: 0.9547 - 4s/epoch - 11ms/step\n","Epoch 115/1000\n","359/359 - 3s - loss: 0.1697 - accuracy: 0.9540 - 3s/epoch - 10ms/step\n","Epoch 116/1000\n","359/359 - 3s - loss: 0.1715 - accuracy: 0.9538 - 3s/epoch - 10ms/step\n","Epoch 117/1000\n","359/359 - 4s - loss: 0.1726 - accuracy: 0.9554 - 4s/epoch - 11ms/step\n","Epoch 118/1000\n","359/359 - 3s - loss: 0.1720 - accuracy: 0.9528 - 3s/epoch - 9ms/step\n","Epoch 119/1000\n","359/359 - 3s - loss: 0.1714 - accuracy: 0.9545 - 3s/epoch - 9ms/step\n","Epoch 120/1000\n","359/359 - 4s - loss: 0.1678 - accuracy: 0.9531 - 4s/epoch - 11ms/step\n","Epoch 121/1000\n","359/359 - 3s - loss: 0.1656 - accuracy: 0.9572 - 3s/epoch - 10ms/step\n","Epoch 122/1000\n","359/359 - 3s - loss: 0.1654 - accuracy: 0.9556 - 3s/epoch - 9ms/step\n","Epoch 123/1000\n","359/359 - 4s - loss: 0.1673 - accuracy: 0.9566 - 4s/epoch - 11ms/step\n","Epoch 124/1000\n","359/359 - 3s - loss: 0.1671 - accuracy: 0.9566 - 3s/epoch - 9ms/step\n","Epoch 125/1000\n","359/359 - 3s - loss: 0.1652 - accuracy: 0.9582 - 3s/epoch - 10ms/step\n","Epoch 126/1000\n","359/359 - 4s - loss: 0.1648 - accuracy: 0.9568 - 4s/epoch - 10ms/step\n","Epoch 127/1000\n","359/359 - 4s - loss: 0.1651 - accuracy: 0.9568 - 4s/epoch - 11ms/step\n","Epoch 128/1000\n","359/359 - 4s - loss: 0.1709 - accuracy: 0.9563 - 4s/epoch - 10ms/step\n","Epoch 129/1000\n","359/359 - 3s - loss: 0.1608 - accuracy: 0.9572 - 3s/epoch - 10ms/step\n","Epoch 130/1000\n","359/359 - 4s - loss: 0.1611 - accuracy: 0.9591 - 4s/epoch - 11ms/step\n","Epoch 131/1000\n","359/359 - 3s - loss: 0.1637 - accuracy: 0.9566 - 3s/epoch - 10ms/step\n","Epoch 132/1000\n","359/359 - 3s - loss: 0.1693 - accuracy: 0.9542 - 3s/epoch - 9ms/step\n","Epoch 133/1000\n","359/359 - 4s - loss: 0.1632 - accuracy: 0.9559 - 4s/epoch - 11ms/step\n","Epoch 134/1000\n","359/359 - 3s - loss: 0.1626 - accuracy: 0.9579 - 3s/epoch - 9ms/step\n","Epoch 135/1000\n","359/359 - 3s - loss: 0.1608 - accuracy: 0.9561 - 3s/epoch - 9ms/step\n","Epoch 136/1000\n","359/359 - 4s - loss: 0.1621 - accuracy: 0.9575 - 4s/epoch - 11ms/step\n","Epoch 137/1000\n","359/359 - 3s - loss: 0.1606 - accuracy: 0.9575 - 3s/epoch - 10ms/step\n","Epoch 138/1000\n","359/359 - 3s - loss: 0.1600 - accuracy: 0.9572 - 3s/epoch - 10ms/step\n","Epoch 139/1000\n","359/359 - 4s - loss: 0.1598 - accuracy: 0.9561 - 4s/epoch - 11ms/step\n","Epoch 140/1000\n","359/359 - 4s - loss: 0.1614 - accuracy: 0.9570 - 4s/epoch - 10ms/step\n","Epoch 141/1000\n","359/359 - 3s - loss: 0.1602 - accuracy: 0.9568 - 3s/epoch - 9ms/step\n","Epoch 142/1000\n","359/359 - 4s - loss: 0.1695 - accuracy: 0.9540 - 4s/epoch - 11ms/step\n","Epoch 143/1000\n","359/359 - 4s - loss: 0.1572 - accuracy: 0.9579 - 4s/epoch - 10ms/step\n","Epoch 144/1000\n","359/359 - 3s - loss: 0.1604 - accuracy: 0.9568 - 3s/epoch - 9ms/step\n","Epoch 145/1000\n","359/359 - 3s - loss: 0.1669 - accuracy: 0.9545 - 3s/epoch - 9ms/step\n","Epoch 146/1000\n","359/359 - 4s - loss: 0.1674 - accuracy: 0.9545 - 4s/epoch - 11ms/step\n","Epoch 147/1000\n","359/359 - 3s - loss: 0.1601 - accuracy: 0.9549 - 3s/epoch - 10ms/step\n","Epoch 148/1000\n","359/359 - 3s - loss: 0.1574 - accuracy: 0.9575 - 3s/epoch - 9ms/step\n","Epoch 149/1000\n","359/359 - 4s - loss: 0.1563 - accuracy: 0.9556 - 4s/epoch - 11ms/step\n","Epoch 150/1000\n","359/359 - 3s - loss: 0.1556 - accuracy: 0.9572 - 3s/epoch - 9ms/step\n","Epoch 151/1000\n","359/359 - 3s - loss: 0.1574 - accuracy: 0.9561 - 3s/epoch - 9ms/step\n","Epoch 152/1000\n","359/359 - 4s - loss: 0.1600 - accuracy: 0.9572 - 4s/epoch - 11ms/step\n","Epoch 153/1000\n","359/359 - 3s - loss: 0.1585 - accuracy: 0.9559 - 3s/epoch - 9ms/step\n","Epoch 154/1000\n","359/359 - 3s - loss: 0.1554 - accuracy: 0.9575 - 3s/epoch - 9ms/step\n","Epoch 155/1000\n","359/359 - 4s - loss: 0.1587 - accuracy: 0.9577 - 4s/epoch - 11ms/step\n","Epoch 156/1000\n","359/359 - 3s - loss: 0.1604 - accuracy: 0.9566 - 3s/epoch - 10ms/step\n","Epoch 157/1000\n","359/359 - 3s - loss: 0.1592 - accuracy: 0.9575 - 3s/epoch - 9ms/step\n","Epoch 158/1000\n","359/359 - 3s - loss: 0.1573 - accuracy: 0.9561 - 3s/epoch - 10ms/step\n","Epoch 159/1000\n","359/359 - 4s - loss: 0.1577 - accuracy: 0.9556 - 4s/epoch - 11ms/step\n","Epoch 160/1000\n","359/359 - 3s - loss: 0.1619 - accuracy: 0.9540 - 3s/epoch - 10ms/step\n","Epoch 161/1000\n","359/359 - 3s - loss: 0.1584 - accuracy: 0.9572 - 3s/epoch - 9ms/step\n","Epoch 162/1000\n","359/359 - 4s - loss: 0.1596 - accuracy: 0.9563 - 4s/epoch - 11ms/step\n","Epoch 163/1000\n","359/359 - 3s - loss: 0.1581 - accuracy: 0.9570 - 3s/epoch - 10ms/step\n","Epoch 164/1000\n","359/359 - 3s - loss: 0.1561 - accuracy: 0.9584 - 3s/epoch - 10ms/step\n","Epoch 165/1000\n","359/359 - 4s - loss: 0.1599 - accuracy: 0.9519 - 4s/epoch - 11ms/step\n","Epoch 166/1000\n","359/359 - 3s - loss: 0.1532 - accuracy: 0.9579 - 3s/epoch - 9ms/step\n","Epoch 167/1000\n","359/359 - 3s - loss: 0.1568 - accuracy: 0.9554 - 3s/epoch - 9ms/step\n","Epoch 168/1000\n","359/359 - 4s - loss: 0.1575 - accuracy: 0.9561 - 4s/epoch - 11ms/step\n","Epoch 169/1000\n","359/359 - 3s - loss: 0.1603 - accuracy: 0.9552 - 3s/epoch - 9ms/step\n","Epoch 170/1000\n","359/359 - 3s - loss: 0.1679 - accuracy: 0.9556 - 3s/epoch - 10ms/step\n","Epoch 171/1000\n","359/359 - 4s - loss: 0.1614 - accuracy: 0.9566 - 4s/epoch - 10ms/step\n","Epoch 172/1000\n","359/359 - 4s - loss: 0.1522 - accuracy: 0.9579 - 4s/epoch - 10ms/step\n","Epoch 173/1000\n","359/359 - 3s - loss: 0.1572 - accuracy: 0.9568 - 3s/epoch - 9ms/step\n","Epoch 174/1000\n","359/359 - 3s - loss: 0.1541 - accuracy: 0.9570 - 3s/epoch - 9ms/step\n","Epoch 175/1000\n","359/359 - 4s - loss: 0.1609 - accuracy: 0.9556 - 4s/epoch - 11ms/step\n","Epoch 176/1000\n","359/359 - 3s - loss: 0.1564 - accuracy: 0.9570 - 3s/epoch - 9ms/step\n","Epoch 177/1000\n","359/359 - 3s - loss: 0.1588 - accuracy: 0.9542 - 3s/epoch - 10ms/step\n","Epoch 178/1000\n","359/359 - 4s - loss: 0.1563 - accuracy: 0.9552 - 4s/epoch - 11ms/step\n","Epoch 179/1000\n","359/359 - 4s - loss: 0.1531 - accuracy: 0.9572 - 4s/epoch - 10ms/step\n","Epoch 180/1000\n","359/359 - 3s - loss: 0.1588 - accuracy: 0.9563 - 3s/epoch - 10ms/step\n","Epoch 181/1000\n","359/359 - 4s - loss: 0.1557 - accuracy: 0.9584 - 4s/epoch - 11ms/step\n","Epoch 182/1000\n","359/359 - 3s - loss: 0.1554 - accuracy: 0.9577 - 3s/epoch - 9ms/step\n","Epoch 183/1000\n","359/359 - 3s - loss: 0.1572 - accuracy: 0.9568 - 3s/epoch - 10ms/step\n","Epoch 184/1000\n","359/359 - 4s - loss: 0.1568 - accuracy: 0.9577 - 4s/epoch - 11ms/step\n","Epoch 185/1000\n","359/359 - 4s - loss: 0.1580 - accuracy: 0.9579 - 4s/epoch - 10ms/step\n","Epoch 186/1000\n","359/359 - 3s - loss: 0.1567 - accuracy: 0.9563 - 3s/epoch - 10ms/step\n","Epoch 187/1000\n","359/359 - 4s - loss: 0.1592 - accuracy: 0.9561 - 4s/epoch - 10ms/step\n","Epoch 188/1000\n","359/359 - 4s - loss: 0.1606 - accuracy: 0.9561 - 4s/epoch - 11ms/step\n","Epoch 189/1000\n","359/359 - 3s - loss: 0.1581 - accuracy: 0.9596 - 3s/epoch - 9ms/step\n","Epoch 190/1000\n","359/359 - 4s - loss: 0.1553 - accuracy: 0.9584 - 4s/epoch - 11ms/step\n","Epoch 191/1000\n","359/359 - 4s - loss: 0.1577 - accuracy: 0.9559 - 4s/epoch - 11ms/step\n","Epoch 192/1000\n","359/359 - 3s - loss: 0.1529 - accuracy: 0.9577 - 3s/epoch - 9ms/step\n","Epoch 192: early stopping\n"]}]},{"cell_type":"code","source":["\n","# Define the models dictionary\n","models = {\n","    \"LSTM\": lstm_model,\n","    \"GRU\": gru_model,\n","    \"Bidirectional LSTM\": bilstm_model\n","}\n","\n","# Evaluate and obtain scores for each model\n","for model_name, model in models.items():\n","    scores = model.evaluate(xs, ys, verbose=0)\n","    print(f\"Scores for {model_name}:\")\n","    print(\"Loss:\", scores[0])\n","    print(\"Accuracy:\", scores[1])\n","    print(\"\\n\")"],"metadata":{"id":"cSWA22FNVHAF"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Save the models\n","lstm_model.save(\"/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/lstm_model.h5\")\n","gru_model.save(\"/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/gru_model.h5\")\n","bilstm_model.save(\"/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/bilstm_model.h5\")"],"metadata":{"id":"CEPeNDUL2hlU"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["def generate_predictions(seed_text, tokenizer, models, max_sequence_len, next_words):\n","    for model_name, model in models.items():\n","        print(f\"Predictions for {model_name}:\")\n","        generated_text = seed_text\n","\n","        for _ in range(next_words):\n","            token_list = tokenizer.texts_to_sequences([generated_text])[0]\n","            token_list = pad_sequences([token_list], maxlen=max_sequence_len-1, padding='pre')\n","            predicted = np.argmax(model.predict(token_list), axis=1)\n","\n","            output_word = \" \"\n","            for word, index in tokenizer.word_index.items():\n","                if index == predicted:\n","                    output_word = word\n","                    break\n","\n","            generated_text += \" \" + output_word\n","\n","        print(generated_text)\n","        print(\"\\n\")"],"metadata":{"id":"OlNWKninonx2"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Set seed text and number of next words\n","seed_text = \"Their team\"\n","next_words = 10\n","\n","# Generate predictions for each model\n","generate_predictions(seed_text, tokenizer, models, max_sequence_len, next_words)"],"metadata":{"id":"9qEHVqejpj7T","executionInfo":{"status":"ok","timestamp":1684900018981,"user_tz":-480,"elapsed":2194,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"7af7a525-3814-419b-e6f4-71907ed5c4e3"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Predictions for LSTM:\n","1/1 [==============================] - 0s 21ms/step\n","1/1 [==============================] - 0s 21ms/step\n","1/1 [==============================] - 0s 19ms/step\n","1/1 [==============================] - 0s 22ms/step\n","1/1 [==============================] - 0s 19ms/step\n","1/1 [==============================] - 0s 22ms/step\n","1/1 [==============================] - 0s 21ms/step\n","1/1 [==============================] - 0s 19ms/step\n","1/1 [==============================] - 0s 18ms/step\n","1/1 [==============================] - 0s 20ms/step\n","Their team was incredibly knowledgeable about the real estate market and provided\n","\n","\n","Predictions for GRU:\n","1/1 [==============================] - 0s 20ms/step\n","1/1 [==============================] - 0s 19ms/step\n","1/1 [==============================] - 0s 21ms/step\n","1/1 [==============================] - 0s 23ms/step\n","1/1 [==============================] - 0s 19ms/step\n","1/1 [==============================] - 0s 20ms/step\n","1/1 [==============================] - 0s 22ms/step\n","1/1 [==============================] - 0s 21ms/step\n","1/1 [==============================] - 0s 19ms/step\n","1/1 [==============================] - 0s 20ms/step\n","Their team was knowledgeable professional and went above and beyond to provide\n","\n","\n","Predictions for Bidirectional LSTM:\n","1/1 [==============================] - 0s 22ms/step\n","1/1 [==============================] - 0s 23ms/step\n","1/1 [==============================] - 0s 20ms/step\n","1/1 [==============================] - 0s 20ms/step\n","1/1 [==============================] - 0s 22ms/step\n","1/1 [==============================] - 0s 25ms/step\n","1/1 [==============================] - 0s 23ms/step\n","1/1 [==============================] - 0s 21ms/step\n","1/1 [==============================] - 0s 23ms/step\n","1/1 [==============================] - 0s 21ms/step\n","Their team was incredibly knowledgeable about the real estate market and provided\n","\n","\n"]}]},{"cell_type":"code","source":["\n","def plot_graphs(history_dict, string):\n","    plt.figure(figsize=(12, 4))\n","    \n","    for model_name, history in history_dict.items():\n","        plt.plot(history.history[string], label=model_name)\n","\n","    plt.xlabel(\"Epochs\")\n","    plt.ylabel(string)\n","    plt.legend()\n","    plt.show()\n","\n","history_dict = {\n","    \"LSTM\": lstm_history,\n","    \"GRU\": gru_history,\n","    \"Bidirectional LSTM\": bilstm_history\n","}\n","\n","plot_graphs(history_dict, 'accuracy')\n","plot_graphs(history_dict, 'loss')"],"metadata":{"id":"v4cXeHREtXil","executionInfo":{"status":"ok","timestamp":1684900090783,"user_tz":-480,"elapsed":1464,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"colab":{"base_uri":"https://localhost:8080/","height":269},"outputId":"c3934c83-c1b8-47c3-d726-2965a1e2074e"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x400 with 1 Axes>"],"image/png":"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\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x400 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["def plot_graphs(history_dict, string):\n","    plt.figure(figsize=(12, 8))\n","    num_models = len(history_dict)\n","    plot_index = 1\n","    \n","    for model_name, history in history_dict.items():\n","        plt.subplot(num_models, 1, plot_index)\n","        plt.plot(history.history[string])\n","        plt.xlabel(\"Epochs\")\n","        plt.ylabel(f\"{model_name} {string}\")\n","        plt.title(model_name)\n","        plot_index += 1\n","\n","    plt.tight_layout()\n","    plt.show()\n","\n","plot_graphs(history_dict, 'accuracy')\n","plot_graphs(history_dict, 'loss')"],"metadata":{"id":"W-sHqRv0ty_K","executionInfo":{"status":"ok","timestamp":1684900138368,"user_tz":-480,"elapsed":2082,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"colab":{"base_uri":"https://localhost:8080/","height":464},"outputId":"87ed00b6-a06c-4e9f-ef0f-69fdd234e11b"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x800 with 3 Axes>"],"image/png":"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\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x800 with 3 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["# Load the models\n","load_bilstm_model = tf.keras.models.load_model('/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/bilstm_model.h5')\n","load_lstm_model = tf.keras.models.load_model('/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/lstm_model.h5')\n","load_gru_model = tf.keras.models.load_model('/content/drive/MyDrive/Data_Mining_NLP_Group_4/Codes_Documentation/TextGeneration/gru_model.h5')\n","\n","# Check its architecture\n","bilstm_model.summary()\n","lstm_model.summary()\n","gru_model.summary()"],"metadata":{"id":"aJmmcXFjPMVM"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["def evaluate_model(model, xs, ys):\n","    loss, accuracy = model.evaluate(xs, ys, verbose=2)\n","    return loss, accuracy\n","\n","# Evaluate the models\n","bilstm_loss, bilstm_acc = evaluate_model(bilstm_model, xs, ys)\n","lstm_loss, lstm_acc = evaluate_model(lstm_model, xs, ys)\n","gru_loss, gru_acc = evaluate_model(gru_model, xs, ys)\n","\n","# Print the accuracy and loss of each model\n","\n","print(\"\\nLSTM Model:\")\n","print(\"Loss:\", lstm_loss)\n","print(\"Accuracy:\", lstm_acc)\n","print(\"\")\n","print(\"GRU Model:\")\n","print(\"Loss:\", gru_loss)\n","print(\"Accuracy:\", gru_acc)\n","print(\"\")\n","print(\"Bidirectional LSTM Model:\")\n","print(\"Loss:\", bilstm_loss)\n","print(\"Accuracy:\", bilstm_acc)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"NXFZr7kKMCEp","executionInfo":{"status":"ok","timestamp":1684906284127,"user_tz":-480,"elapsed":42121,"user":{"displayName":"DAVE FAGARITA","userId":"05095540274383433925"}},"outputId":"8a2da909-99d9-47e2-a8ea-d5eae9acea24"},"execution_count":30,"outputs":[{"output_type":"stream","name":"stdout","text":["135/135 - 16s - loss: 0.1233 - accuracy: 0.9649 - 16s/epoch - 116ms/step\n","135/135 - 14s - loss: 0.1467 - accuracy: 0.9631 - 14s/epoch - 102ms/step\n","135/135 - 4s - loss: 0.1355 - accuracy: 0.9638 - 4s/epoch - 31ms/step\n","\n","LSTM Model:\n","Loss: 0.14669956266880035\n","Accuracy: 0.9630576372146606\n","\n","GRU Model:\n","Loss: 0.13550086319446564\n","Accuracy: 0.9637546539306641\n","\n","Bidirectional LSTM Model:\n","Loss: 0.12325380742549896\n","Accuracy: 0.9649163484573364\n"]}]},{"cell_type":"code","source":["\n","models = {\n","    \"LSTM\": load_lstm_model,\n","    \"GRU\": load_gru_model,\n","    \"Bidirectional LSTM\": load_bilstm_model\n","\n","def generate_predictions(seed_text, tokenizer, models, max_sequence_len, next_words):\n","    for model_name, model in models.items():\n","        print(f\"Predictions for {model_name}:\")\n","        generated_text = seed_text\n","\n","        for _ in range(next_words):\n","            token_list = tokenizer.texts_to_sequences([generated_text])[0]\n","            token_list = pad_sequences([token_list], maxlen=max_sequence_len-1, padding='pre')\n","            predicted = np.argmax(model.predict(token_list), axis=1)\n","\n","            output_word = \" \"\n","            for word, index in tokenizer.word_index.items():\n","                if index == predicted:\n","                    output_word = word\n","                    break\n","\n","            generated_text += \" \" + output_word\n","\n","        print(generated_text)\n","        print(\"\\n\")\n","\n","\n","# Set seed text and number of next words\n","seed_text = \"Their team\"\n","next_words = 10\n","\n","# Generate predictions for each model\n","generate_predictions(seed_text, tokenizer, models, max_sequence_len, next_words)"],"metadata":{"id":"F3yd3AlPPca6"},"execution_count":null,"outputs":[]}],"metadata":{"accelerator":"GPU","colab":{"provenance":[{"file_id":"1tRwS6uh1HgDuwYfd3AZMa4NcAqUH85GA","timestamp":1684839113836},{"file_id":"1fuKXuOmu4wyWAaOO3Bg49Yot8moJe_yv","timestamp":1684836313602}],"gpuType":"T4"},"gpuClass":"standard","kernelspec":{"display_name":"Python 3","name":"python3"},"language_info":{"name":"python"}},"nbformat":4,"nbformat_minor":0}