noshot 0.1.0__py3-none-any.whl

noshot/data/AIDS CN NLP/NLP/NLP 2/Lab10(LSTM)/Lab_10_Text_generator_using_LSTM.ipynb

@@ -0,0 +1,480 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "ab178016",
+   "metadata": {
+    "id": "ab178016"
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "2023-11-01 13:49:31.029384: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used.\n",
+      "2023-11-01 13:49:33.145555: E tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:9342] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n",
+      "2023-11-01 13:49:33.145680: E tensorflow/compiler/xla/stream_executor/cuda/cuda_fft.cc:609] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n",
+      "2023-11-01 13:49:33.151894: E tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:1518] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n",
+      "2023-11-01 13:49:34.451107: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used.\n",
+      "2023-11-01 13:49:34.453035: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n",
+      "To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
+      "2023-11-01 13:49:39.145697: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\n"
+     ]
+    }
+   ],
+   "source": [
+    "# keras module for building LSTM\n",
+    "from keras_preprocessing.sequence import pad_sequences\n",
+    "from keras.layers import Embedding, LSTM, Dense, Dropout\n",
+    "from keras.preprocessing.text import Tokenizer\n",
+    "from keras.callbacks import EarlyStopping\n",
+    "from keras.models import Sequential\n",
+    "import keras.utils as ku\n",
+    "\n",
+    "# set seeds for reproducability\n",
+    "import tensorflow\n",
+    "tensorflow.random.set_seed(2)\n",
+    "\n",
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "import string, os\n",
+    "\n",
+    "import warnings\n",
+    "warnings.filterwarnings(\"ignore\")\n",
+    "warnings.simplefilter(action='ignore', category=FutureWarning)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "c1475d1b",
+   "metadata": {
+    "id": "c1475d1b",
+    "outputId": "6b1e36ce-a1e5-45a9-e117-31e18b86c6d7"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "831"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "\n",
+    "all_headlines = []\n",
+    "article_df = pd.read_csv('ArticlesApril2017.csv')\n",
+    "all_headlines.extend(list(article_df.headline.values))\n",
+    "all_headlines = [h for h in all_headlines if h != \"Unknown\"]\n",
+    "len(all_headlines)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "d272dc69",
+   "metadata": {
+    "id": "d272dc69",
+    "outputId": "288dee58-7dcc-4ea4-b0aa-c71438b485be"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "['finding an expansive view  of a forgotten people in niger',\n",
+       " 'and now  the dreaded trump curse',\n",
+       " 'venezuelas descent into dictatorship',\n",
+       " 'stain permeates basketball blue blood',\n",
+       " 'taking things for granted',\n",
+       " 'the caged beast awakens',\n",
+       " 'an everunfolding story',\n",
+       " 'oreilly thrives as settlements add up',\n",
+       " 'mouse infestation',\n",
+       " 'divide in gop now threatens trump tax plan']"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "def clean_text(txt):\n",
+    "    txt = \"\".join(v for v in txt if v not in string.punctuation).lower()\n",
+    "    txt = txt.encode(\"utf8\").decode(\"ascii\",'ignore')\n",
+    "    return txt\n",
+    "\n",
+    "corpus = [clean_text(x) for x in all_headlines]\n",
+    "corpus[:10]\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "d5f40739",
+   "metadata": {
+    "id": "d5f40739",
+    "outputId": "336a5531-034e-47c9-885e-70bf9998a915"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[[169, 17],\n",
+       " [169, 17, 665],\n",
+       " [169, 17, 665, 367],\n",
+       " [169, 17, 665, 367, 4],\n",
+       " [169, 17, 665, 367, 4, 2],\n",
+       " [169, 17, 665, 367, 4, 2, 666],\n",
+       " [169, 17, 665, 367, 4, 2, 666, 170],\n",
+       " [169, 17, 665, 367, 4, 2, 666, 170, 5],\n",
+       " [169, 17, 665, 367, 4, 2, 666, 170, 5, 667],\n",
+       " [6, 80]]"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tokenizer = Tokenizer()\n",
+    "\n",
+    "def get_sequence_of_tokens(corpus):\n",
+    "    ## tokenization\n",
+    "    tokenizer.fit_on_texts(corpus)\n",
+    "    total_words = len(tokenizer.word_index) + 1\n",
+    "\n",
+    "    ## convert data to sequence of tokens\n",
+    "    input_sequences = []\n",
+    "    for line in corpus:\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)\n",
+    "    return input_sequences, total_words\n",
+    "\n",
+    "inp_sequences, total_words = get_sequence_of_tokens(corpus)\n",
+    "inp_sequences[:10]\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "c4c05b1d",
+   "metadata": {
+    "id": "c4c05b1d"
+   },
+   "outputs": [],
+   "source": [
+    "def generate_padded_sequences(input_sequences):\n",
+    "    max_sequence_len = max([len(x) for x in input_sequences])\n",
+    "    input_sequences = np.array(pad_sequences(input_sequences, maxlen=max_sequence_len, padding='pre'))\n",
+    "\n",
+    "    predictors, label = input_sequences[:,:-1],input_sequences[:,-1]\n",
+    "    label = ku.to_categorical(label, num_classes=total_words)\n",
+    "    return predictors, label, max_sequence_len\n",
+    "\n",
+    "predictors, label, max_sequence_len = generate_padded_sequences(inp_sequences)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "f8ad0539",
+   "metadata": {
+    "id": "f8ad0539",
+    "outputId": "42efafee-02b8-46e6-ef00-91a7629ceebe"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model: \"sequential\"\n",
+      "_________________________________________________________________\n",
+      " Layer (type)                Output Shape              Param #   \n",
+      "=================================================================\n",
+      " embedding (Embedding)       (None, 18, 10)            24220     \n",
+      "                                                                 \n",
+      " lstm (LSTM)                 (None, 100)               44400     \n",
+      "                                                                 \n",
+      " dropout (Dropout)           (None, 100)               0         \n",
+      "                                                                 \n",
+      " dense (Dense)               (None, 2422)              244622    \n",
+      "                                                                 \n",
+      "=================================================================\n",
+      "Total params: 313242 (1.19 MB)\n",
+      "Trainable params: 313242 (1.19 MB)\n",
+      "Non-trainable params: 0 (0.00 Byte)\n",
+      "_________________________________________________________________\n"
+     ]
+    }
+   ],
+   "source": [
+    "def create_model(max_sequence_len, total_words):\n",
+    "    input_len = max_sequence_len - 1\n",
+    "    model = Sequential()\n",
+    "\n",
+    "    # Add Input Embedding Layer\n",
+    "    model.add(Embedding(total_words, 10, input_length=input_len))\n",
+    "\n",
+    "    # Add Hidden Layer 1 - LSTM Layer\n",
+    "    model.add(LSTM(100))\n",
+    "    model.add(Dropout(0.1))\n",
+    "\n",
+    "    # Add Output Layer\n",
+    "    model.add(Dense(total_words, activation='softmax'))\n",
+    "\n",
+    "    model.compile(loss='categorical_crossentropy', optimizer='adam')\n",
+    "\n",
+    "    return model\n",
+    "\n",
+    "model = create_model(max_sequence_len, total_words)\n",
+    "model.summary()\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "a1fe10c4",
+   "metadata": {
+    "id": "a1fe10c4",
+    "outputId": "01dab848-fc9a-43f6-ba9f-019187b12b96"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/100\n",
+      "Epoch 2/100\n",
+      "Epoch 3/100\n",
+      "Epoch 4/100\n",
+      "Epoch 5/100\n",
+      "Epoch 6/100\n",
+      "Epoch 7/100\n",
+      "Epoch 8/100\n",
+      "Epoch 9/100\n",
+      "Epoch 10/100\n",
+      "Epoch 11/100\n",
+      "Epoch 12/100\n",
+      "Epoch 13/100\n",
+      "Epoch 14/100\n",
+      "Epoch 15/100\n",
+      "Epoch 16/100\n",
+      "Epoch 17/100\n",
+      "Epoch 18/100\n",
+      "Epoch 19/100\n",
+      "Epoch 20/100\n",
+      "Epoch 21/100\n",
+      "Epoch 22/100\n",
+      "Epoch 23/100\n",
+      "Epoch 24/100\n",
+      "Epoch 25/100\n",
+      "Epoch 26/100\n",
+      "Epoch 27/100\n",
+      "Epoch 28/100\n",
+      "Epoch 29/100\n",
+      "Epoch 30/100\n",
+      "Epoch 31/100\n",
+      "Epoch 32/100\n",
+      "Epoch 33/100\n",
+      "Epoch 34/100\n",
+      "Epoch 35/100\n",
+      "Epoch 36/100\n",
+      "Epoch 37/100\n",
+      "Epoch 38/100\n",
+      "Epoch 39/100\n",
+      "Epoch 40/100\n",
+      "Epoch 41/100\n",
+      "Epoch 42/100\n",
+      "Epoch 43/100\n",
+      "Epoch 44/100\n",
+      "Epoch 45/100\n",
+      "Epoch 46/100\n",
+      "Epoch 47/100\n",
+      "Epoch 48/100\n",
+      "Epoch 49/100\n",
+      "Epoch 50/100\n",
+      "Epoch 51/100\n",
+      "Epoch 52/100\n",
+      "Epoch 53/100\n",
+      "Epoch 54/100\n",
+      "Epoch 55/100\n",
+      "Epoch 56/100\n",
+      "Epoch 57/100\n",
+      "Epoch 58/100\n",
+      "Epoch 59/100\n",
+      "Epoch 60/100\n",
+      "Epoch 61/100\n",
+      "Epoch 62/100\n",
+      "Epoch 63/100\n",
+      "Epoch 64/100\n",
+      "Epoch 65/100\n",
+      "Epoch 66/100\n",
+      "Epoch 67/100\n",
+      "Epoch 68/100\n",
+      "Epoch 69/100\n",
+      "Epoch 70/100\n",
+      "Epoch 71/100\n",
+      "Epoch 72/100\n",
+      "Epoch 73/100\n",
+      "Epoch 74/100\n",
+      "Epoch 75/100\n",
+      "Epoch 76/100\n",
+      "Epoch 77/100\n",
+      "Epoch 78/100\n",
+      "Epoch 79/100\n",
+      "Epoch 80/100\n",
+      "Epoch 81/100\n",
+      "Epoch 82/100\n",
+      "Epoch 83/100\n",
+      "Epoch 84/100\n",
+      "Epoch 85/100\n",
+      "Epoch 86/100\n",
+      "Epoch 87/100\n",
+      "Epoch 88/100\n",
+      "Epoch 89/100\n",
+      "Epoch 90/100\n",
+      "Epoch 91/100\n",
+      "Epoch 92/100\n",
+      "Epoch 93/100\n",
+      "Epoch 94/100\n",
+      "Epoch 95/100\n",
+      "Epoch 96/100\n",
+      "Epoch 97/100\n",
+      "Epoch 98/100\n",
+      "Epoch 99/100\n",
+      "Epoch 100/100\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "<keras.src.callbacks.History at 0x7f26142dd480>"
+      ]
+     },
+     "execution_count": 8,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "model.fit(predictors, label, epochs=100, verbose=5)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "a76a078d",
+   "metadata": {
+    "id": "a76a078d"
+   },
+   "outputs": [],
+   "source": [
+    "def generate_text(seed_text, next_words, model, max_sequence_len):\n",
+    "    for _ in range(next_words):\n",
+    "        token_list = tokenizer.texts_to_sequences([seed_text])[0]\n",
+    "        token_list = pad_sequences([token_list], maxlen=max_sequence_len-1, padding='pre')\n",
+    "#         predicted = model.predict_classes(token_list, verbose=0)\n",
+    "        predict_x=model.predict(token_list)\n",
+    "        classes_x=np.argmax(predict_x,axis=1)\n",
+    "\n",
+    "        output_word = \"\"\n",
+    "        for word,index in tokenizer.word_index.items():\n",
+    "            if index == classes_x:\n",
+    "                output_word = word\n",
+    "                break\n",
+    "        seed_text += \" \"+output_word\n",
+    "    return seed_text.title()\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "d1c3679e",
+   "metadata": {
+    "id": "d1c3679e",
+    "outputId": "4814b0be-d18c-40bc-e827-6f3d119b2ea8"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "1/1 [==============================] - 0s 380ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "United States Erode Shorelines Tasmania Shows What\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "Preident Trump Is Wimping Out On\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "Donald Trump Is May So Contagious\n",
+      "1/1 [==============================] - 0s 16ms/step\n",
+      "1/1 [==============================] - 0s 16ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "India And China Station Rail Mishap Spurs\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 16ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "New York Today A Holocaust Survivors\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 14ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "1/1 [==============================] - 0s 15ms/step\n",
+      "Science And Technology Knocking In The Original News\n"
+     ]
+    }
+   ],
+   "source": [
+    "print (generate_text(\"united states\", 5, model, max_sequence_len))\n",
+    "print (generate_text(\"preident trump\", 4, model, max_sequence_len))\n",
+    "print (generate_text(\"donald trump\", 4, model, max_sequence_len))\n",
+    "print (generate_text(\"india and china\", 4, model, max_sequence_len))\n",
+    "print (generate_text(\"new york\", 4, model, max_sequence_len))\n",
+    "print (generate_text(\"science and technology\", 5, model, max_sequence_len))"
+   ]
+  }
+ ],
+ "metadata": {
+  "colab": {
+   "provenance": []
+  },
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}