File size: 6,497 Bytes
d23fad5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 |
{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"id": "3d8c593f",
"metadata": {
"ExecuteTime": {
"end_time": "2023-03-26T07:31:34.213141Z",
"start_time": "2023-03-26T07:31:14.082603Z"
},
"scrolled": true
},
"outputs": [],
"source": [
"!pip install huggingface_hub\n",
"!pip install datasets\n",
"!pip install keras"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "bca92d1d",
"metadata": {
"ExecuteTime": {
"end_time": "2023-03-26T14:03:42.287776Z",
"start_time": "2023-03-26T14:03:39.989670Z"
}
},
"outputs": [],
"source": [
"from huggingface_hub import notebook_login\n",
"from datasets import load_dataset\n",
"import pandas as pd\n",
"from datasets import load_dataset\n",
"import tensorflow as tf\n",
"from tensorflow.keras.applications.vgg16 import VGG16\n",
"from tensorflow.keras.models import Model\n",
"from tensorflow.keras.layers import Dense, GlobalAveragePooling2D\n",
"from tensorflow.keras.optimizers import Adam\n",
"from tensorflow.keras.utils import to_categorical\n",
"from PIL import Image\n",
"import numpy as np"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "62254f94",
"metadata": {
"ExecuteTime": {
"end_time": "2023-03-26T14:03:42.317000Z",
"start_time": "2023-03-26T14:03:42.289947Z"
}
},
"outputs": [],
"source": [
"notebook_login()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "57308b59",
"metadata": {
"ExecuteTime": {
"end_time": "2023-03-26T14:03:52.591875Z",
"start_time": "2023-03-26T14:03:48.476822Z"
}
},
"outputs": [],
"source": [
"# load dataset from hugging face\n",
"# prepare data for training, validation and testing\n",
"train_ds, val_ds = load_dataset('competitions/aiornot', split=\"train\").train_test_split(test_size=0.15).values()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "b83b1536",
"metadata": {
"ExecuteTime": {
"end_time": "2023-03-26T14:04:10.210069Z",
"start_time": "2023-03-26T14:03:53.833533Z"
}
},
"outputs": [],
"source": [
"data_sz = 1000\n",
"X_train = train_ds[:data_sz]['image']\n",
"X_val = val_ds[:data_sz]['image']\n",
"Y_train = to_categorical(train_ds[:data_sz]['label'])\n",
"Y_val = to_categorical(val_ds[:data_sz]['label'])\n",
"# Convert the input data to a NumPy array\n",
"X_train = np.stack([np.array(image) for image in X_train])\n",
"X_val = np.stack([np.array(image) for image in X_val])"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "72df9419",
"metadata": {
"ExecuteTime": {
"start_time": "2023-03-26T14:04:33.658Z"
}
},
"outputs": [],
"source": [
"with tf.device('/device:GPU:3'):\n",
" # Load the VGG16 model pre-trained on ImageNet\n",
" base_model = VGG16(weights='imagenet', include_top=False)\n",
"\n",
" # Add a global spatial average pooling layer\n",
" x = base_model.output\n",
" x = GlobalAveragePooling2D()(x)\n",
"\n",
" # Add a fully-connected layer\n",
" x = Dense(1024, activation='relu')(x)\n",
"\n",
" # Add a logistic layer with the number of classes of target variable\n",
" num_classes = 2\n",
" predictions = Dense(num_classes, activation='softmax')(x)\n",
"\n",
" # Create the final model\n",
" model = Model(inputs=base_model.input, outputs=predictions)\n",
"\n",
" # Freeze all layers in the base VGG16 model\n",
" for layer in base_model.layers:\n",
" layer.trainable = False\n",
"\n",
" # Compile the model\n",
" model.compile(optimizer=Adam(lr=0.001), loss='categorical_crossentropy', metrics=['accuracy'])\n",
"\n",
"# Train the model on your new dataset\n",
"model.fit(X_train, Y_train, epochs=10, validation_data=(X_val, Y_val))"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "bbf079b7",
"metadata": {
"ExecuteTime": {
"start_time": "2023-03-26T14:05:03.786Z"
}
},
"outputs": [],
"source": [
"# Generate predictions for the data\n",
"y_pred = model.predict(X_val)\n",
"# Convert predictions and true labels to class indices\n",
"y_pred_classes = y_pred.argmax(axis=1)\n",
"y_true_classes = Y_val.argmax(axis=1)\n",
"# Find the indices of the misclassified samples\n",
"misclassified_indices = np.where(y_pred_classes != y_true_classes)[0]\n",
"\n",
"# Get the misclassified samples\n",
"# x_misclassified = X_val[misclassified_indices]\n",
"# y_misclassified_true = Y_val[misclassified_indices]\n",
"# y_misclassified_pred = y_pred[misclassified_indices]"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "1e639f6b",
"metadata": {
"ExecuteTime": {
"start_time": "2023-03-26T14:05:06.090Z"
}
},
"outputs": [],
"source": [
"# a helper function to view missclassfied data with the image and prediction\n",
"def checkMiss(idx):\n",
" print(\"\\ncorrect:\", Y_val[idx])\n",
" print(\"miss:\", y_pred[idx])\n",
" img = Image.fromarray(X_val[idx])\n",
" img.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "951ff24e",
"metadata": {
"ExecuteTime": {
"start_time": "2023-03-26T14:05:07.650Z"
}
},
"outputs": [],
"source": [
"# view 5 miss classified data to see what could be improved\n",
"for i in range(10):\n",
" checkMiss(misclassified_indices[i])"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"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": "3.8.10"
},
"toc": {
"base_numbering": 1,
"nav_menu": {},
"number_sections": false,
"sideBar": true,
"skip_h1_title": false,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": false,
"toc_position": {},
"toc_section_display": true,
"toc_window_display": false
}
},
"nbformat": 4,
"nbformat_minor": 5
}
|