"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"vectorized_text = vectorizer(X.values)"
],
"metadata": {
"id": "VMEoKm6lXSNW",
"execution": {
"iopub.status.busy": "2024-06-25T14:53:19.627901Z",
"iopub.execute_input": "2024-06-25T14:53:19.628255Z",
"iopub.status.idle": "2024-06-25T14:53:23.730808Z",
"shell.execute_reply.started": "2024-06-25T14:53:19.628228Z",
"shell.execute_reply": "2024-06-25T14:53:23.730033Z"
},
"trusted": true
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"vectorized_text.dtype"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "G-o7uCoWZNUG",
"outputId": "8168d0de-37bd-4731-8678-2ece48a484a5",
"execution": {
"iopub.status.busy": "2024-06-25T14:53:31.269373Z",
"iopub.execute_input": "2024-06-25T14:53:31.270310Z",
"iopub.status.idle": "2024-06-25T14:53:31.275970Z",
"shell.execute_reply.started": "2024-06-25T14:53:31.270269Z",
"shell.execute_reply": "2024-06-25T14:53:31.275061Z"
},
"trusted": true
},
"execution_count": null,
"outputs": [
{
"execution_count": 21,
"output_type": "execute_result",
"data": {
"text/plain": "tf.int64"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"vectorized_text.shape"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "mXYxar2pZUsV",
"outputId": "060e9018-c36e-4bbb-eed7-8b4f0beb4e1b",
"execution": {
"iopub.status.busy": "2024-06-25T14:53:38.231329Z",
"iopub.execute_input": "2024-06-25T14:53:38.231712Z",
"iopub.status.idle": "2024-06-25T14:53:38.238041Z",
"shell.execute_reply.started": "2024-06-25T14:53:38.231683Z",
"shell.execute_reply": "2024-06-25T14:53:38.237006Z"
},
"trusted": true
},
"execution_count": null,
"outputs": [
{
"execution_count": 22,
"output_type": "execute_result",
"data": {
"text/plain": "TensorShape([1600000, 20])"
},
"metadata": {}
}
]
},
{
"cell_type": "markdown",
"source": [
"## Prepare Dataset"
],
"metadata": {
"id": "PgiE9_Yxulln"
}
},
{
"cell_type": "code",
"source": [
"#MCSHBAP - map, chache, shuffle, batch, prefetch from_tensor_slices, list_file\n",
"dataset = tf.data.Dataset.from_tensor_slices((vectorized_text, Y))\n",
"dataset = dataset.cache()\n",
"dataset = dataset.shuffle(1000000)\n",
"dataset = dataset.batch(16)\n",
"dataset = dataset.prefetch(4)"
],
"metadata": {
"id": "aEWYamp9Z7nc",
"execution": {
"iopub.status.busy": "2024-06-25T14:53:50.931596Z",
"iopub.execute_input": "2024-06-25T14:53:50.932259Z",
"iopub.status.idle": "2024-06-25T14:53:50.960596Z",
"shell.execute_reply.started": "2024-06-25T14:53:50.932225Z",
"shell.execute_reply": "2024-06-25T14:53:50.959710Z"
},
"trusted": true
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"train = dataset.take(int(len(dataset)*.8))\n",
"val = dataset.skip(int(len(dataset)*.8)).take(int(len(dataset)*.1))\n",
"test = dataset.skip(int(len(dataset)*.9)).take(int(len(dataset)*.1))"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:53:54.228079Z",
"iopub.execute_input": "2024-06-25T14:53:54.228960Z",
"iopub.status.idle": "2024-06-25T14:53:54.245007Z",
"shell.execute_reply.started": "2024-06-25T14:53:54.228922Z",
"shell.execute_reply": "2024-06-25T14:53:54.244234Z"
},
"trusted": true,
"id": "JNqpMDGVulln"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"## Create Model"
],
"metadata": {
"id": "gPgfQL0Oulln"
}
},
{
"cell_type": "code",
"source": [
"from tensorflow.keras.models import Sequential\n",
"from tensorflow.keras.layers import LSTM, Dropout, Bidirectional, Dense, Embedding"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:53:58.500182Z",
"iopub.execute_input": "2024-06-25T14:53:58.501050Z",
"iopub.status.idle": "2024-06-25T14:53:58.507412Z",
"shell.execute_reply.started": "2024-06-25T14:53:58.501016Z",
"shell.execute_reply": "2024-06-25T14:53:58.506548Z"
},
"trusted": true,
"id": "uf2qEhrBullo"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"model = Sequential()\n",
"# Create the embedding layer\n",
"model.add(Embedding(MAX_FEATURES + 1, 32))\n",
"# Bidirectional LSTM Layer with specified parameters\n",
"model.add(Bidirectional(LSTM(50, activation='tanh', dropout=0.2, recurrent_dropout=0.2)))\n",
"# Dense layer with Dropout\n",
"model.add(Dense(128, activation='relu'))\n",
"model.add(Dropout(0.1))\n",
"# Output layer\n",
"model.add(Dense(1, activation='sigmoid'))"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:54:10.259729Z",
"iopub.execute_input": "2024-06-25T14:54:10.260096Z",
"iopub.status.idle": "2024-06-25T14:54:10.295468Z",
"shell.execute_reply.started": "2024-06-25T14:54:10.260068Z",
"shell.execute_reply": "2024-06-25T14:54:10.294744Z"
},
"trusted": true,
"id": "ddUNfyl3ullo"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"model.build(input_shape=(None, None))"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:54:13.901270Z",
"iopub.execute_input": "2024-06-25T14:54:13.902006Z",
"iopub.status.idle": "2024-06-25T14:54:14.087392Z",
"shell.execute_reply.started": "2024-06-25T14:54:13.901975Z",
"shell.execute_reply": "2024-06-25T14:54:14.086630Z"
},
"trusted": true,
"id": "JnhwdMa9ullo"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from tensorflow.keras.optimizers import Adam\n",
"from tensorflow.keras.losses import BinaryCrossentropy\n",
"from tensorflow.keras.models import load_model"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:54:16.930948Z",
"iopub.execute_input": "2024-06-25T14:54:16.931270Z",
"iopub.status.idle": "2024-06-25T14:54:16.938622Z",
"shell.execute_reply.started": "2024-06-25T14:54:16.931246Z",
"shell.execute_reply": "2024-06-25T14:54:16.937714Z"
},
"trusted": true,
"id": "-IA6WFlpullo"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from tensorflow.keras.callbacks import ModelCheckpoint\n",
"\n",
"# Define the filepath pattern for saving the model weights\n",
"filepath = \"/kaggle/working/Twitter_epoch_{epoch:02d}_val_loss:{val_loss:.2f}_val_acc:{val_accuracy:.2f}.weights.h5\"\n",
"\n",
"# Create a list of callbacks\n",
"callbacks = [ModelCheckpoint(filepath=filepath,save_weights_only=True,save_freq=\"epoch\")]"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:54:28.712934Z",
"iopub.execute_input": "2024-06-25T14:54:28.713295Z",
"iopub.status.idle": "2024-06-25T14:54:28.720603Z",
"shell.execute_reply.started": "2024-06-25T14:54:28.713267Z",
"shell.execute_reply": "2024-06-25T14:54:28.719758Z"
},
"trusted": true,
"id": "znphNNi9ullo"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"model.compile(optimizer=Adam(learning_rate=0.001),loss=BinaryCrossentropy(), metrics=['accuracy'])"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:54:34.161816Z",
"iopub.execute_input": "2024-06-25T14:54:34.162210Z",
"iopub.status.idle": "2024-06-25T14:54:34.176415Z",
"shell.execute_reply.started": "2024-06-25T14:54:34.162180Z",
"shell.execute_reply": "2024-06-25T14:54:34.175537Z"
},
"trusted": true,
"id": "hnN4envQullo"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"model.summary()"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:54:36.370257Z",
"iopub.execute_input": "2024-06-25T14:54:36.370987Z",
"iopub.status.idle": "2024-06-25T14:54:36.390660Z",
"shell.execute_reply.started": "2024-06-25T14:54:36.370951Z",
"shell.execute_reply": "2024-06-25T14:54:36.389810Z"
},
"trusted": true,
"id": "aXlzo4D5ullp",
"outputId": "378fc4f2-785d-45a1-e7bd-ce14027b7ad3"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": "\u001b[1mModel: \"sequential\"\u001b[0m\n",
"text/html": "Model: \"sequential\"\n
\n"
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n│ embedding (\u001b[38;5;33mEmbedding\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m2,400,032\u001b[0m │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ bidirectional (\u001b[38;5;33mBidirectional\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m100\u001b[0m) │ \u001b[38;5;34m33,200\u001b[0m │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m12,928\u001b[0m │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ dropout (\u001b[38;5;33mDropout\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ dense_1 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m129\u001b[0m │\n└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"text/html": "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n│ embedding (Embedding) │ (None, None, 32) │ 2,400,032 │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ bidirectional (Bidirectional) │ (None, 100) │ 33,200 │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ dense (Dense) │ (None, 128) │ 12,928 │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ dropout (Dropout) │ (None, 128) │ 0 │\n├─────────────────────────────────┼────────────────────────┼───────────────┤\n│ dense_1 (Dense) │ (None, 1) │ 129 │\n└─────────────────────────────────┴────────────────────────┴───────────────┘\n
\n"
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m2,446,289\u001b[0m (9.33 MB)\n",
"text/html": " Total params: 2,446,289 (9.33 MB)\n
\n"
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m2,446,289\u001b[0m (9.33 MB)\n",
"text/html": " Trainable params: 2,446,289 (9.33 MB)\n
\n"
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n",
"text/html": " Non-trainable params: 0 (0.00 B)\n
\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"history = model.fit(train, epochs=5, validation_data=val, callbacks=callbacks)"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-25T14:55:10.330140Z",
"iopub.execute_input": "2024-06-25T14:55:10.330506Z"
},
"trusted": true,
"id": "ZSTZWH0Hullp",
"outputId": "745998f4-13e4-45a9-871d-b60ec1972385"
},
"execution_count": null,
"outputs": [
{
"name": "stdout",
"text": "Epoch 1/5\n\u001b[1m72347/80000\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━━\u001b[0m \u001b[1m5:30\u001b[0m 43ms/step - accuracy: 0.7937 - loss: 0.4444",
"output_type": "stream"
},
{
"text": "IOPub message rate exceeded.\nThe notebook server will temporarily stop sending output\nto the client in order to avoid crashing it.\nTo change this limit, set the config variable\n`--NotebookApp.iopub_msg_rate_limit`.\n\nCurrent values:\nNotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\nNotebookApp.rate_limit_window=3.0 (secs)\n\n",
"name": "stderr",
"output_type": "stream"
},
{
"name": "stdout",
"text": "\u001b[1m80000/80000\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3552s\u001b[0m 44ms/step - accuracy: 0.8271 - loss: 0.3845 - val_accuracy: 0.8192 - val_loss: 0.3976\nEpoch 4/5\n\u001b[1m80000/80000\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3547s\u001b[0m 44ms/step - accuracy: 0.8321 - loss: 0.3761 - val_accuracy: 0.8243 - val_loss: 0.3898\nEpoch 5/5\n\u001b[1m 2284/80000\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m56:07\u001b[0m 43ms/step - accuracy: 0.8639 - loss: 0.3181",
"output_type": "stream"
}
]
},
{
"cell_type": "code",
"source": [
"Y.shape"
],
"metadata": {
"execution": {
"iopub.status.busy": "2024-06-23T18:35:56.636125Z",
"iopub.execute_input": "2024-06-23T18:35:56.636495Z",
"iopub.status.idle": "2024-06-23T18:35:56.644631Z",
"shell.execute_reply.started": "2024-06-23T18:35:56.636460Z",
"shell.execute_reply": "2024-06-23T18:35:56.643704Z"
},
"trusted": true,
"id": "bCQlcFbzullp",
"outputId": "95a728da-e6d6-4ea0-88ac-e1b20ec6bb01"
},
"execution_count": null,
"outputs": [
{
"execution_count": 37,
"output_type": "execute_result",
"data": {
"text/plain": "(1600000,)"
},
"metadata": {}
}
]
},
{
"cell_type": "markdown",
"source": [
"## Make Prediction"
],
"metadata": {
"id": "e04-q7UB3DGD"
}
},
{
"cell_type": "code",
"source": [
"from tensorflow.keras.models import load_model"
],
"metadata": {
"id": "MKxQgs3m58H-"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"model = load_model('/content/drive/MyDrive/twitter_sentiment_analysis/twitter_sentiment_analysis_epoch4.h5')"
],
"metadata": {
"id": "HTqzhL298MM1"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"input_text = vectorizer(\"6 years later and nothing changes… And he blames the other driver each time.\")"
],
"metadata": {
"id": "qehiN3QDullp"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"input_text = tf.expand_dims(input_text, axis=0)"
],
"metadata": {
"id": "MPBPiHeq86fM"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"input_text.shape"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "0huBcwIQ88wt",
"outputId": "db15e490-f048-4026-ad40-4313bfe2e4a5"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"TensorShape([1, 20])"
]
},
"metadata": {},
"execution_count": 41
}
]
},
{
"cell_type": "code",
"source": [
"res = model.predict(input_text)"
],
"metadata": {
"id": "zvzeF7pK45Lg",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "5d81930a-e6dc-4479-8370-19d289c064d6"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"1/1 [==============================] - 1s 801ms/step\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"res.shape"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "-MBu6wRe8jv2",
"outputId": "086ac5fc-0aed-4558-802d-a64fa5058e5d"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"(1, 1)"
]
},
"metadata": {},
"execution_count": 43
}
]
},
{
"cell_type": "code",
"source": [
"res"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "nI-ryBTn9CSG",
"outputId": "6488c02d-a050-4fd6-8eec-8fe267caf475"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"array([[0.60668284]], dtype=float32)"
]
},
"metadata": {},
"execution_count": 44
}
]
},
{
"cell_type": "code",
"source": [
"if res<0.5:\n",
" print('Negative')\n",
"else:\n",
" print('Positive')"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "yIImIbYF9D3d",
"outputId": "aadfcdb4-2e79-4f1a-a220-404c7dcf42bd"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Positive\n"
]
}
]
},
{
"cell_type": "markdown",
"source": [
"## Save the vectorizer"
],
"metadata": {
"id": "AxK1G-IXBG39"
}
},
{
"cell_type": "code",
"source": [
"\n",
"model = tf.keras.models.Sequential()\n",
"model.add(tf.keras.Input(shape=(1,), dtype=tf.string))\n",
"model.add(vectorizer)"
],
"metadata": {
"id": "6moK73ijCOTZ"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"model.save('/content/drive/MyDrive/twitter_sentiment_analysis/vectorizer_model', save_format='tf')"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "-YNccjKeDQoD",
"outputId": "33ca7c92-63e5-45e0-c258-19f0653b107e"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"WARNING:tensorflow:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"loaded_model = tf.keras.models.load_model('/content/drive/MyDrive/twitter_sentiment_analysis/vectorizer_model')"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "lWRVvOJZJB-1",
"outputId": "fe446de4-35fb-429b-81ef-8cf0ffcfb293"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"WARNING:tensorflow:No training configuration found in save file, so the model was *not* compiled. Compile it manually.\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"loaded_vectorizer = loaded_model.layers[0]"
],
"metadata": {
"id": "IlaNfQAsJWL0"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"vectorizer('Hello world, life is pretty awesome')"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "eqrPTycoJhkE",
"outputId": "d78dc9f3-91a4-4f67-9ec5-10d80a773fe8"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
""
]
},
"metadata": {},
"execution_count": 33
}
]
},
{
"cell_type": "code",
"source": [
"loaded_vectorizer('Hello world, life is pretty awesome')"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "LiGC60zSJnOr",
"outputId": "767d59f5-60e2-4df1-e468-6a3e044906be"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
""
]
},
"metadata": {},
"execution_count": 9
}
]
},
{
"cell_type": "code",
"source": [],
"metadata": {
"id": "wBPbQIIeihSt"
},
"execution_count": null,
"outputs": []
}
]
}